file stringlengths 27 28 | text stringlengths 1 5.17M |
|---|---|
PMC010xxxxxx/PMC10125210.txt |
==== Front
J Infect
J Infect
The Journal of Infection
0163-4453
1532-2742
The British Infection Association. Published by Elsevier Ltd.
S0163-4453(23)00251-7
10.1016/j.jinf.2023.04.014
Letter to the Editor
Estimation of excess cardiovascular deaths after COVID-19 in 2020
Furuse Yuki
Department of Medical Virology, Nagasaki University Graduate School of Biomedical Sciences, Nagasaki, Japan
24 4 2023
7 2023
24 4 2023
87 1 e5e7
21 4 2023
© 2023 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
2023
The British Infection Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcPost-COVID-19 sequelae are a serious concern of public health, as reported by Fang et al. in this journal.1 Furthermore, the incidence of cardiovascular diseases is increased after COVID-19.2, 3 Cardiovascular diseases are the major cause of death around the world.4 Therefore, the impact of cardiovascular diseases among post-COVID-19 patients is likely significant. However, we could not directly count the number of cardiovascular deaths induced by COVID-19 because the infection history of patients with cardiovascular diseases after recovery from COVID-19 is not always available.
Here, we performed a meta-analysis to calculate the risk of cardiovascular diseases and deaths after COVID-19 and estimated the number of post-COVID-19 cardiovascular deaths in 2020. The details of the methodology are described in Supplementary materials.
Nine studies that estimated the risk for cardiovascular diseases and deaths after COVID-19 were identified (Supplementary Table 1). Although the definition of COVID-19 deaths differs between countries, likely, deaths caused by cardiovascular dysfunction that occurred during the acute phase of COVID-19 were regarded and counted as COVID-19 deaths in some countries. Therefore, we only included studies that reported the risk in the post-acute phase of COVID-19 in our meta-analysis to estimate the excess disease burden caused by cardiovascular deaths after COVID-19.
Four studies were included in our meta-analysis; COVID-19 patients were followed up 22–30 days after the positive test for SARS-CoV-2 for 3 months to 1 year in those studies. The pooled analysis of 1,295,380 COVID-19 patients revealed that the incidence rate ratio for cardiovascular diseases and deaths was 1.50 (95% confidence interval, 1.40–1.60. Heterogeneity I2 = 85.5%. Supplementary Table 1).
Age-stratified cumulative numbers of COVID-19 reported cases in 2020 were available for 62 countries. The number of post-COVID-19 cardiovascular deaths was estimated for those countries using the risk ratio calculated by the abovementioned meta-analysis and the age-stratified annual incidence rates of cardiovascular deaths in each country. It should be noted that our estimation is not the number of deaths that occurred in 2020. Rather, the figures represent the number of deaths that might have happened within one year after infection with COVID-19 in 2020.
Most COVID-19 infections were observed in people in their 20–50 s, whereas most COVID-19 deaths occurred in those 60 s or older (Supplementary Fig. 1, 2). The excess cardiovascular deaths after COVID-19 were also expected mainly for older adults. The estimation ranged between 0.02 per 1 million in Vietnam and 157 per 1 million in Czechia ( Fig. 1 and Supplementary Fig. 3).Fig. 1 Excess cardiovascular deaths after COVID-19 in 2020, estimated from COVID-19 reported case numbers. Excess cardiovascular deaths after COVID-19 in 2020 by age group per 1 million population, estimated from COVID-19 reported case numbers. The vertical lines indicate 95% confidence intervals. The figure shows the results of 10 countries, where all data are available for the present study; the results for other countries can be found in Supplementary Fig. 3.
Fig. 1
However, the reported number of COVID-19 cases must be much lower than the actual number of SARS-CoV-2 infections. This was because not all SARS-CoV-2-infected persons underwent diagnostic testing. Seroprevalence studies showed that less than 10% of infected people were detected in many parts of the world in 2020.5
We then inferred the number of SARS-CoV-2-infected people in 2020 from seroprevalence data and calculated the expected number of cardiovascular deaths after SARS-CoV-2 infection from the inference. Age-stratified data on the serum positivity rate for anti-SARS-CoV-2 antibodies at the end of 2020 were available for 16 countries (Supplementary Table 2).
The excess cardiovascular deaths after COVID-19, estimated from seroprevalence data, ranged between 16 per 1 million population in Canada and 983 per 1 million population in Czechia ( Fig. 2 and Supplementary Fig. 4). The ratio of estimated post-COVID-19 excess cardiovascular deaths to COVID-19 reported deaths was 0.04 (Canada)–23.8 (Sri Lanka) (Supplementary Fig. 5). In Japan, for example, the ratio was 1.4, which means that excess cardiovascular deaths after COVID-19 (No. = 5114. 95% confidence interval, 3215–6920) could be 1.4 times larger than the COVID-19 reported deaths in 2020 (No. = 3541). Post-COVID-19 excess cardiovascular deaths are equivalent to 0.7% (Canada)–24.3% (India) of annual cardiovascular deaths without COVID-19. The increase in disease burden of cardiovascular deaths after COVID-19 was expected for not only older adults but also others including children (Supplementary Figure 6).Fig. 2 Excess cardiovascular deaths after COVID-19 in 2020, estimated from SARS-CoV-2 seroprevalence. Excess cardiovascular deaths after COVID-19 in 2020 by age group per 1 million population, estimated from SARS-CoV-2 seroprevalence data. The vertical lines indicate 95% confidence intervals. The figure shows the results of 10 countries, where all data are available for the present study; the results for other countries can be found in Supplementary Fig. 4.
Fig. 2
In summary, we reported the estimation of excess cardiovascular deaths after COVID-19 in 2020. Although the increased risk of cardiovascular diseases in post-COVID-19 patients is recognized, its significance has not been extensively assessed, especially at a global scale. We found that the impact of the mortality varies, but it is comparable to the documented COVID-19 deaths in some countries. In 16 countries where the analysis was performed using SARS-CoV-2 seroprevalence data, the total of documented COVID-19 deaths in 2020 was 770,920, while estimated cardiovascular deaths after COVID-19 were 742,432.
Studies that estimated overall excess deaths during the COVID-19 pandemic have recently been published.6, 7 While many countries experienced a surge in excess deaths, some countries, such as Australia and Japan, observed little or no excess deaths. During the early years of the COVID-19 pandemic, nonpharmaceutical interventions including social distancing and wearing a face mask were implemented. Those interventions contributed to the reduction of not only COVID-19 but also other infectious diseases such as influenza. Because the epidemic of influenza is one of the major causes of excess deaths, its suppression during the COVID-19 pandemic may have canceled out the impact of excess deaths directly and indirectly caused by COVID-19 in countries where no excess deaths were detected.
Our estimation of post-COVID-19 excess cardiovascular deaths was conducted using data on COVID-19 in 2020. COVID-19 vaccination has been rolled out since the end of 2020 to prevent severe illness and death. Because hospitalized COVID-19 patients with severe illness are known to have a higher risk for cardiovascular diseases even after recovery than nonhospitalized patients,2, 3 vaccination has the potential to reduce the post-COVID-19 risk in those with breakthrough infections. Indeed, a study showed that vaccination could attenuate the risk for cardiovascular diseases after COVID-19.8 However, the effect of vaccination on the risk for post-COVID-19 illnesses is still controversial.9 We need further data to estimate the impact of cardiovascular deaths in post-COVID-19 patients in 2021 and after, especially among already vaccinated people. The risk might also be different among SARS-CoV-2 variants.10
Our estimation sheds light on the unnoticed but great impact of the COVID-19 pandemic from a new perspective. The mechanisms of increased risk for cardiovascular diseases after COVID-19 are still elusive; dysregulation of pathways in hemostasis, immune regulation, and thrombotic status may be involved. Basic studies on the pathogenesis and epidemiological studies about the disease burden of cardiovascular diseases induced by COVID-19 continue to be required for better care of COVID-19 patients even after its acute clinical phase and in the post-pandemic era.
Funding
This work was supported by the Strategic Center of Biomedical Advanced Vaccine Research and Development for Preparedness and Response (JP223fa627004) from the 10.13039/100009619 Japan Agency for Medical Research and Development , by the Grant-in-Aid for Scientific Research (JP19KK0204) from the 10.13039/501100001691 Japan Society for the Promotion of Science , and by the Nagasaki University State of the Art Research program (grant number not available) from Nagasaki University.
Declaration of Competing Interest
The author declares no competing interests.
Appendix A Supplementary material
Supplementary material
Appendix A Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jinf.2023.04.014.
==== Refs
References
1 Fang Xiaoyu Ming Chao Cen Yuan Lin Hao Zhan Kegang Yang Sha Post-sequelae one year after hospital discharge among older COVID-19 patients: a multi-center prospective cohort study J Infect 84 2 2022 179 186 10.1016/j.jinf.2021.12.005 34902448
2 Raisi-Estabragh Zahra Cooper Jackie Salih Ahmed Raman Betty Lee Aaron Mark Neubauer Stefan Cardiovascular disease and mortality sequelae of COVID-19 in the UK Biobank Heart 109 2 2023 119 126 10.1136/HEARTJNL-2022-321492
3 Xie Yan Xu Evan Bowe Benjamin Al-Aly Ziyad Long-term cardiovascular outcomes of COVID-19 Nat Med 28 3 2022 583 590 10.1038/s41591-022-01689-3 35132265
4 Roth Gregory A. Mensah George A. Johnson Catherine O. Addolorato Giovanni Ammirati Enrico Baddour Larry M. Global burden of cardiovascular diseases and risk factors, 1990–2019: update from the GBD 2019 study J Am Coll Cardiol 76 25 2020 2982 3021 10.1016/J.JACC.2020.11.010 33309175
5 Barber Ryan M. Sorensen Reed J.D. Pigott David M. Bisignano Catherine Carter Austin Amlag Joanne O. Estimating global, regional, and national daily and cumulative infections with SARS-CoV-2 through Nov 14, 2021: a statistical analysis Lancet 2022 10.1016/S0140-6736(22)00484-6
6 Msemburi William Karlinsky Ariel Knutson Victoria Aleshin-Guendel Serge Chatterji Somnath Wakefield Jon The WHO estimates of excess mortality associated with the COVID-19 pandemic Nature 613 7942 2022 130 137 10.1038/s41586-022-05522-2 36517599
7 Wang Haidong Paulson Katherine R. Pease Spencer A. Watson Stefanie Comfort Haley Zheng Peng Estimating excess mortality due to the COVID-19 pandemic: a systematic analysis of COVID-19-related mortality, 2020–21 Lancet 399 10334 2022 1513 1536 10.1016/S0140-6736(21)02796-3 35279232
8 Xie Junqing Prats-Uribe Albert Feng Qi Wang Yunhe Gill Dipender Paredes Roger Clinical and genetic risk factors for acute incident venous thromboembolism in ambulatory patients with COVID-19 JAMA Intern Med 182 10 2022 1063 1070 10.1001/JAMAINTERNMED.2022.3858 35980616
9 Mizrahi Barak Sudry Tamar Flaks-Manov Natalie Yehezkelli Yoav Kalkstein Nir Akiva Pinchas Long covid outcomes at one year after mild SARS-CoV-2 infection: nationwide cohort study BMJ 380 2023 e072529 10.1136/BMJ-2022-072529
10 Gottlieb Michael Wang Ralph C. Yu Huihui Spatz Erica S. Montoy Juan Carlos C. Rodriguez Robert M. Severe fatigue and persistent symptoms at 3 months following severe acute respiratory syndrome coronavirus 2 infections during the pre-delta, delta, and omicron time periods: a multicenter prospective cohort study Clin Infect Dis 2023 10.1093/CID/CIAD045
|
PMC010xxxxxx/PMC10126226.txt |
==== Front
Vacunas
Vacunas
Vacunas
1576-9887
1578-8857
Elsevier España, S.L.U.
S1576-9887(23)00031-6
10.1016/j.vacun.2023.04.003
Review Article
The status of COVID-19 vaccines in India: A review
El estado de las vacunas COVID-19 en India: Una revisiónJha Deepak Kumar a
Pranay Kumar b
Samiksha c
Kumar Amit d
Yashvardhini Niti c⁎
a Department of Zoology, S.M.P. Rajkiya Mahila Mahavidyalaya, Ballia 277401, India
b Department of Biochemistry, IGIMS, Patna 800 014, India
c Department of Microbiology, Patna Women's College, Patna, 800 001, India
d Department of Botany, Patna University, Patna 800 005, India
⁎ Corresponding author.
25 4 2023
25 4 2023
1 1 2022
12 4 2023
© 2023 Elsevier España, S.L.U. All rights reserved.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The outbreak of SARS-CoV-2, an etiologic agent of the COVID-19 pandemic disease in late December 2019 has left the whole world aghast with huge health and economic losses. Due to a lack of specific knowledge and understanding at the initial stages, an unprecedented rise in COVID-19 cases has been recorded globally. Various preventive measures and strategies were implemented, however, for the radical control of SARS-CoV-2 infections; it seems that the only effective way to control the ongoing infections is large-scale vaccination. So far, WHO has approved 11 vaccines for emergency use namely Pfizer/BioNTech, Oxford/Astra Zeneca, Johnson and Johnson, Moderna, Covilo, Novavax, Covovax, Spikevax, Can Sino, Comirnaty, and Coronavac while five other needs approval. The worldwide vaccination dataset reveals that 65.7% of the world population has received their first dose of the COVID-19 vaccine. As a consequence of the proactive implementation of India's vaccination program, a historical milestone of administering over 1.9 billion doses of COVID-19 vaccines have been achieved on 19th May 2022. This review summarizes the different types of traditional and modern vaccine designing strategies with an emphasis on COVID-19. Moreover, the review highlights the status of vaccines for COVID-19 approved in India which includes both indigenous and non-indigenous vaccines. The present article also encompasses vaccine designing and developmental strategies, efficacy, safety profile and usage among the population, and the efficacy of modern vaccines over traditional ones.
El brote de SARS-CoV-2, un agente etiológico de la enfermedad pandémica COVID-19, a fines de diciembre de 2019, ha dejado al mundo entero horrorizado con enormes pérdidas económicas y de salud. Debido a la falta de conocimiento y comprensión específicos en las etapas iniciales, se ha registrado un aumento sin precedentes en los casos de COVID-19 a nivel mundial. Sin embargo, se implementaron diversas medidas y estrategias preventivas para el control radical de las infecciones por SARS-CoV-2; parece que la única forma eficaz de controlar las infecciones en curso es la vacunación a gran escala. Hasta el momento, la OMS ha aprobado 11 vacunas para uso de urgencia Pfizer/BioNTech, Oxford/Astra Zeneca, Johnson and Johnson, Moderna, Covilo, Novavax, Covovax, Spikevax, Can Sino, Comirnaty y Coronavac, mientras que otras cinco necesitan aprobación. El conjunto de datos de vacunación mundial revela que el 65,7% de la población mundial ha recibido su primera dosis de la vacuna COVID-19. Como consecuencia de la implementación proactiva del programa de vacunación de la India, el 19 de mayo de 2022 se logró un hito histórico de administrar más de 1900 millones de dosis de vacunas contra el COVID-19. Esta revisión resume los diferentes tipos de estrategias de diseño de vacunas tradicionales y modernas con énfasis sobre COVID-19. Además, la revisión destaca el estado de las vacunas para COVID-19 aprobadas en India, que incluye vacunas tanto indígenas como no indígenas. El presente artículo también abarca estrategias de diseño y desarrollo de vacunas, eficacia, perfil de seguridad y uso entre la población, y la eficacia de las vacunas modernas sobre las tradicionales.
Keywords
COVID-19
Vaccination
India
Pandemic
Indigenous vaccine
Palabras clave
COVID-19
Vacunación
India
Pandemia
Vacuna indígena
==== Body
pmcIntroduction
The first case of the SARS-CoV-2 (COVID-19) infection was reported in Wuhan, China in December 2019 from where this disease spread globally. In a setting of a strongly connected and integrated world, coupled with the high transmissibility of the viral infection,1 , 2 the disease rapidly spread and was declared a public health emergency of international concern (PHEIC) on January 30, 2020. The unprecedented rise in the number of cases resulted in a declaration of COVID-19 as a pandemic in March, 2020. The COVID-19 disease spread exponentially and at present exists in more than 230 countries and has resulted in more than 39 million confirmed cases (WHO). As on date 20th July, 2022 the number of COVID-19 cases across the globe reached 561,156,416 with 6,365,510 casualties (WHO dashboard).
The WHO together with the United States Centres for Disease Control and Prevention (CDC), European Centre for Disease Prevention and Control (EDC) issued several guidelines and health advisories to prevent the spread of the outbreak. WHO declares COVID-19 vaccines as safe and data strongly suggests that the risk of infection and severity of symptoms associated with the disease is low for vaccinated people. WHO recommends wearing a mask in indoor public places, people above the age of 2 years should wear a mask in case they are not vaccinated, or fully vaccinated in an area with high risk of infection. People should wear masks in crowded outdoor places and for working at places where others are not vaccinated. A distance of 6 ft is essential and people should avoid crowded places unless urgent. WHO recommends proper washing of hands as well as covering is essential while coughing and sneezing to avoid the spread of this disease.
Coronaviruses are positive-sense, enveloped, single-stranded RNA viruses that possess helical nucleocapsid. Belonging to the family Coronaviridae, the coronavirus envelope comprises of club-shaped glycoprotein projections.3 The diameter of SARS-CoV-2 ranges from 60–140 nm, often found in its pleomorphic form.4 SARS-CoV-2 is a beta coronavirus belonging to the same subgenus as MERS-CoV and SARS-CoV.5 As SARS-CoV-2 shares 89% nucleotide identity similar to bat SARS- like- CoVZXC21 and 82% similarity with human SARS-CoV, hence, named SARS-CoV-2 by the International Committee on Taxonomy of Viruses4 and its transmissibility could be measured by R0 value (Reproduction number) 6 where SARS-CoV-2 has an R0 value range of 2–3.1 , 4
This virus contains 4 different structural proteins; spike (S), envelope (E), membrane (M) and nucleocapsid (N) and other structural and accessory proteins (HE, 3a/b protein and 4a/b protein).7 , 8 The spike protein (S) is a trimeric envelope glycoprotein that plays a vital role in the transmission of infection as it undergoes cleavage into an amino (N)-terminal S1 subunit which helps the incorporation of the virus into the host cell, hence a major target for vaccines. This S1 subunit has a neutralizing response to antisera due to its binding with the host cell.5 , 9 , 10 E protein forms the viroporins (E channels) and is involved in viral replication cycle.11 N-protein constitutes the helical nucleocapsid and binds with the viral RNA genome.12 M-protein, the most abundantly expressed protein in virus particle is the main organizer of coronavirus assembly and is involved in the morphogenesis and assembly of SARS-CoV-2 by interacting with the essential structural proteins and is also linked with the process of apoptosis in the host cell.13
SARS-CoV-2 entry into the host's cell is facilitated by the spike RBD domain. This domain allows binding with the ACE-2 receptor abundantly present in respiratory epithelium and other organs such as the upper esophagus, proximal tubular cells of the kidney, myocardial cells, urothelial cells of bladder.14 , 15 SARS-CoV-2 captures the target cell by using human angiotensin-converting enzyme 2 (ACE-2) as the entry receptor through the spike glycoprotein (S-protein) and recruits the cellular serine protease TMPRSS2 (Type 2 transmembrane serine protease) for priming of S2 protein.16., 17., 18., 19., 20., 21. ACE 2 and TMPRSS2 express in alveolar epithelial type II cells of host target cells.22 , 23 An overview of the life cycle of SARS-CoV-2 is shown in Fig. 1 .Fig. 1 An overview of life cycle of SARS-CoV-2.
Fig. 1
The incubation period of SARS-CoV-2 lies between 1–14 days and symptom manifestation starts around 5 days in most cases which correspond to the highest virus load in respiratory tract.24 The reported symptoms observed in COVID-19 patients are chest tightness/dyspnea, dry cough, fatigue, fever, gastrointestinal symptoms, headache, joint pain, loss of taste and smell, myalgia and sore throat. The principal target for SARS-CoV-2 is the respiratory system but can also affect other vital organs such as central nervous system (CNS), cardiovascular, renal, and gastrointestinal tract.4 SARS-CoV-2 mainly spreads through respiratory droplet transmission from person to person by a person's cough, sneeze or talk; or it may also occur through fomites, live viral particles adhering to inanimate objects.25., 26.
Previously, several articles have been published on the impact of COVID-19 on environmental factors, where multiple study showed a significant correlation between climatic indicators like humidity, temperature, due point, wind speed rainfall and SARS-CoV-2 induced fatality.27., 28., 29., 30., 31.However, it has been evident from the previous research that temperature affects the COVID-19 transmissions, but have been found mixed i.e. positive, negative as well as insignificant on the transmissibility of COVID-19 infections.32., 33., 34. Another important indicator of environmental issues is air pollution which affects the COVID-19 transmission, morbidity and mortality rate.35., 36., 37. For instances, Northern Italy was severely affected by SARS-CoV-2 infections, where the air pollution was found more polluted than the rest of the country. Significantly higher incidence of COVID-19 related casualties have been reported.36., 38., 39. Therefore, it may be concluded that COVID-19 influences the environmental factors and vice versa. The pandemic has brought major loss of human life globally, and severely affecting almost every aspects of human life including global economy, ecosystem, health sector, industrial operations etc.
During the initial phases of viral outbreak no effective antiviral therapy was available, although several drugs (Hydroxychloroquine, Chloroquine, Remdesivir, Favipiravir, Azithromycin, Lopinavir), plasma therapy, oxygen treatment, antibiotics, convalescent plasma were being used in an emergency.26 But, the rapid spread and the urgency of the situation resulted in the development of COVID-19 vaccine in just 12 months of the onset of this pandemic. According to an official report from national public health agencies, as on 21st July 2022, 12.24 billion doses of COVID-19 vaccines have been administered worldwide. Today, India with its robust vaccine development program is not only manufacturing domestic COVID-19 vaccines but also distributing them in the global market.40 The present review highlights the different strategies used in traditional vaccine development. An emphasis was given on the strategies used by different institutes and companies for the development of COVID-19 vaccine. A thorough study was done on COVID-19 vaccine development strategy in Indian context. This review also highlights the challenges in the development of COVID-19 vaccine, the ethical concern and the future prospect of the vaccines. A detailed study has also been done on the linkage of COVID-19 with other diseases like cancer and autoimmunity. This study also includes the status of vaccination in different states of India, the status of approved and still under process vaccines. A futuristic model has also been suggested for the development of COVID-19 vaccine. This review is novel as it not only highlights the COVID-19 vaccine development strategies but links it with various diseases like autoimmunity and cancer and suggests the ethical concern, market size and future prospect of COVID-19 vaccine development in India. The step by step research done in this review is shown in Fig. 2 .Fig. 2 Flow diagram for the systematic review performed on COVID-19 vaccine development in India.
Fig. 2
Traditional vaccine development strategies
Vaccines are the biological molecules that induce an immune response against any infection or disease inside the body.41 Within the last few years, infectious diseases around the world have resulted in an increment of death rates and are known to be the major cause of mortality.42 Vaccination and immunization, the keystone of public health policy made a major contribution to global health. Development of a vaccine involves the use of microorganisms responsible for the disease development either in the killed or weakened form.43 Vaccine, the most effective prophylaxis has the potential to generate herd immunity in the communities thereby reducing the occurrence of disease and hence blocking its transmission.44 Vaccines help in producing memory cells and hence after subsequent infection with the same pathogen immunity develops faster.45 A safe vaccine does not produce any IgE mediated immune responses in the host's body and hence remains non- allergenic.46 Development of a successful vaccine generally requires 12–15 years including both public and private involvement. The fastest vaccine developed so far, before COVID-19 for mumps took 5 years.6 The live, attenuated mumps vaccine used nowadays in the United States, licensed in 1967 was developed by Maurice Hilleman using virus from his daughter. Several vaccines developed using traditional approach is shown in table 1 .Table 1 Traditional vaccines and their licensed manufacturing processes.
Table 1Serial No. Disease Trade Name Generic Name Cell Culture/ Fermentation
1. Anthrax Biothrax Anthrax Vaccine Adsorbed Chemically defined protein-free media growing a microaerophilic culture of avirulent, nonencapsulated Bacillus anthracis
2. Haemophilus influenzae ActHIB Haemophilus b Conjugate Vaccine (Tetanus Toxoid Conjugate) Grown of Haemophilus influenzae type b strain 1482 grown in a semisynthetic medium
3. Hepatitis A Havrix Hepatitis A Vaccine, Inactivated Hepatitis A (strain HM175) propagated in MRC-5 human diploid cells
4. Hepatitis B Recombivax HB Hepatitis B Vaccine (recombinant) Recombinant hepatitis B surface antigen (HBsAg) produced in yeast cells grown in a complex medium of extract of yeast, soy peptone, dextrose, amino acids, and mineral salts
5. Influenza Fluzone Inactivated Influenza Virus Vaccine Propagation on embryonated chicken eggs
6. Japanese encephalitis JE-VAX Japanese Encephalitis Virus Vaccine Inactivated Intracerebral inoculation of mice
7. Measles, mumps, rubella, and varicella ProQuad Measles, Mumps, Rubella and Varicella (Oka/Merck) Virus Vaccine Live Measles virus propagated in chick embryo cell culture; mumps virus in chick embryo cell culture; rubella virus propagated in WI-38 human diploid lung fibroblasts; varicella virus propagated on MRC-5 cells
8. Meningococcal Menactra Meningococcal (groups A, C, Y, and W-135) Polysaccharide Diphtheria Toxoid Conjugate Vaccine Meningococcal strains are cultured individually on Mueller-Hinton agar and grown in Watson-Scherp media; Corynebacterium diphtheriae grown on modified Mueller and Miller medium
9. Pneumococcal Prevnar Pneumococcal 13-valent Conjugate Vaccine (Diphtheria CRM197 Protein) Streptococcus pneumoniae serotypes 1, 3, 4, 5, 6A, 6B, 7 F, 9 V, 14, 18 C, 19A, 19 F, and 23 F individually grown on soy peptone broth; C. diphtheriae strain containing CRM197 grown in casamino acids and yeast extract–based medium
10. Polio IPOL Poliovirus Vaccine Inactivated Types 1, 2, and 3 poliovirus individually grown in Vero cells on microcarriers using Eagle MEM modified medium supplemented with newborn calf serum
11. Rabies RabAvert Rabies Vaccine Rabies virus grown in primary culture of chicken fibroblasts in synthetic cell culture medium with the addition of human albumin, polygeline, and antibiotics
12. Streptococcus pneumoniae Pneumovax Pneumococcal vaccine polyvalent ND
13. Typhoid fever Vivotif Typhoid Vaccine Live Oral Ty21a Fermentation using medium containing a digest of yeast extract, an acid digest of casein, dextrose, and galactose
14. Yellow fever YF-Vax Yellow Fever Vaccine Strain 17D-204 of yellow fever is cultured on living avian leukosis virus-free chicken embryos
ND – Not disclosed.
The development of successful vaccine safe for public use requires proper protocol development by regulatory agencies such as European Medicines Agency (EMA), the World Health Organization (WHO), the United States Food and Drug Administration (USFDA) and other bodies. The development process involves preclinical (in vitro and in vivo testing in animals) and clinical (clinical trials in human candidates) stages for the planning of clinical development path of novel vaccine candidate.47 The different stages of vaccine development involves exploratory, pre-clinical, clinical and post marketing stage as shown in Fig. 3 .Fig. 3 Traditional Vaccine Development Stages (10–15 years).
Fig. 3
According to the candidate used for vaccine development vaccine can be live attenuated, inactivated, protein-based, nucleic acid based, viral vector based, toxoid, conjugate and outer membrane vesicles vaccine43 (Fig. 4 ).Fig. 4 Different types of traditional vaccine development strategy. Strategy involved in vaccine development.
Fig. 4
Live attenuated vaccines
This is the traditional method of vaccine development containing whole bacteria or virus in its weakened form so as to build up a protective immune response.48 Live attenuated vaccines can be obtained by transferring the disease causing organism through a series of cell cultures or embryos of animals (chick embryos).45 The weakened organism replicates same as a natural infection thereby causing a humoral and cell mediated immune response.44 The live attenuated vaccines induce immune system against specific viruses and often need only a single immunization without booster doses.45 However, live attenuated vaccines are not suitable for immunocompromised as well as weak people as the disease can develop. The vaccines developed using live attenuated form include BCG, smallpox, Polio (Albert Sabin's oral polio vaccine), measles, rubella, mumps, rotavirus vaccine, varicella vaccine, and yellow fever vaccine.44 , 45
Inactivated vaccines
These vaccines do not contain any live ingredient as they are produced from pathogens deactivated by formaldehyde or heat.16 These chemicals impair the replicative ability of pathogens but they can still be recognized by the host's immune system. To induce more immunogenicity booster doses are required as these candidates are less immunogenic.49 The vaccines developed against Hepatitis A, rabies, Salk polio and influenza are some important inactivated vaccines.43 , 45 , 50
Protein based vaccines
Protein subunit vaccines
These vaccines use antigenic components (like spike protein) generated in vitro.44 They include only epitopes that trigger the immune system.49 This vaccine does not contain whole bacteria or virus so that immune response can target only a subset of pathogen protein with antigenic properties.45 Subunit vaccines do not create long lasting immune responses, hence require adjuvants and additional booster doses.43 Although peptide vaccines possess high efficacy and several advantages, yet they are not licensed for human use being less immunogenic. These vaccines require adjuvants to enhance immunogenicity and susceptible to enzymatic degradation due to short stretch of amino acids.41 Acellular pertussis vaccines and Influenza vaccines are some vaccines developed using protein subunit approach.
Virus-like particles (VLPs)
These molecules closely resemble viruses but incapable of infecting the host cell as they don't contain viral genetic material. They are safe and stimulate strong immune responses but are difficult to manufacture.44 These virus particles selected are immunological as well as capable of eliciting both antibody and cell-mediated immune responses.51
These virus particles can be generated experimentally using recombinant viral proteins in the laboratory. These particles may be expressed in an array of expression systems including prokaryotic cells, yeasts, insect cell lines, plants and mammalian cell lines.51 The capsid proteins of HIV, HBV, Hepatitis C, bacteriophages are commonly used to develop vaccine.51., 52., 53., 54.
Nucleic acid vaccines
Nucleic acid vaccines are quick and easy to develop using genetic material from a pathogen.
DNA vaccines
DNA vaccines are produced by inserting antigenic components of a pathogen into a plasmid stimulating humoral as well as cell mediated immune response. Electroporation is commonly used for administration of these plasmids resulting in production of high antibody titer.44 These vaccines also encode for adjuvants required for the stimulation of adaptive immune response.16 The design of the vaccines allows its translocation inside the nucleus of the host cell, where it is transcribed into a functional copy of mRNA.55
RNA vaccines
These vaccines contain messenger RNA (mRNA) molecule inside a shell of lipid membrane which protects as well as facilitates the entry of vaccine by fusing with the cell membrane. The mRNA molecule of the vaccine encodes for a disease specific antigen which is displayed on the cell surface eliciting a proper immune response.56 These mRNA molecules are safe as they do not cause disease and results in Th2 cell skewed response.44 mRNA vaccine supports rapid vaccine development program as it mimics antigen structure and expression similar to natural infection.57
Viral vector vaccines
These vaccines are produced by isolating an antigen from pathogen and integrating it into a bacterial or viral vector system.43 The bacteria or virus behaves as a vector that replicates and expresses pathogenic gene inside the cell.45 Viral vector vaccines are grown on cell lines hence easily and cheaply developed. As viral vector vaccines produce endogenous antigen hence, activate both humoral and cellular immune responses in single dose.56 , 57 Viral vector vaccines depending on the replicating ability can either be replicating and non-replicating.
Replicating viral vector vaccines
This vaccine strategy engages innate immunity targeting mainly mucosal sites. These types of vaccines have the ability to make new viral particles besides delivering the vaccine antigen when used as a vaccine delivery platform. Separate virus like measles or adenovirus are genetically engineered and modified to express the desired gene of interest.1 Adenoviruses are the most commonly used virus for replicating viral vector vaccines. Subsequent booster doses are essential to acquire long term humoral immunity. Examples of replicating viral vector vaccines are: HPV, pertussis and Hepatitis B.
Non-replicating viral vector vaccines
In this vaccine strategy, viral vectors are genetically modified to impair the replicating mechanism and hence become non-replicating virus. These makes the viral strain attenuated and a potent vaccine candidate as it only triggers immune response without replicating in human host. Adenoviruses are the most commonly exploited virus for vector vaccines. These vaccines only produce the vaccine antigen and are unable to make new viral particles. The inactivated gene from an unrelated virus (measles or adenovirus) is genetically engineered to express the gene of interest.43 The non-replicating vaccine induce pathogen specific host response which includes killed pathogens, synthetic pathogen structure or recombinant pathogen product as antigens. This type of vaccine creates strong humoral as well as CD8 + T-cell mediated immune response but the strength depends on the type of viral serotype being used. Non-replicating vaccines are safer over others as the risk of disease onset is quite low.
Toxoid vaccines
Toxoid vaccines contain attenuated toxins inducing humoral immune response. The purification of bacterial toxins followed by their inactivation with formaldehyde leads to generation of a toxoid, routinely used to make diphtheria and tetanus toxins. The inoculation with a toxoid, results in release of anti-toxoid antibody that readily binds and neutralizes its toxicity. Toxoid vaccines do not provide prolonged immunity; hence require regular booster doses for effective protection. Toxoid vaccines provide protection against diphtheria and tetanus.
Conjugate vaccines
Conjugate vaccines made by combining two distinct components, hence similar to recombinant vaccines. Bacterial coat fragments are coupled to a carrier protein that is utilized in vaccination. Conjugate vaccines elicit a stronger co-immune response whereas fragments of bacteria show less immunological response. This bacterial fragment does not cause disease, but provides protection against future infections when associated with carrier proteins. These vaccines are used in pneumococcal vaccinations to protect children against bacterial illnesses.58
Outer membrane vesicles vaccine
The outer membrane of Gram-negative bacteria, sometimes form extracellular vesicles (outer membrane vesicles, OMV) with a diameter of 20 to 300 nm. Since the driving factor for OMV development remained unclear for a long period, the establishment of OMVs still remains unclear.59 , 60 OMV production assumed to be a random stress response from bacterial cells 61.
In recent years, OMVs have attracted significant attraction in form of vaccine delivery system against bacterial infections.62 OMV investigated as a potential vaccination against Neisseria meningitidis serogroup B illness. Following the success of the MeNZB OMV-based vaccine in suppressing N. meningitidis B epidemic in New Zealand, subsequent research culminated in approval of Bexsero, against meningococcal B strains 63 .
COVID-19 vaccine development
Vaccine development; a long and complex process which takes years and even decades for successful vaccine formulation. However, the COVID-19 vaccine development program took a year and was a miracle across the globe.60 COVID-19 vaccine development was a big challenge for the researchers as it was targeted to develop a vaccine in a span of 12–24 months.43 The rapid development of COVID-19 vaccine was facilitated by well-timed release of viral genome sequence, advancement and innovation in vaccine technology, active participation of global scientific community, robust regulatory framework, adequate funding by the government as well as huge market demand.61 By the start of December, 2020, the researchers around the globe announced excellent results in vaccine development (Table 2 ) and on 2nd December, 2020, vaccine developed by Pfizer with German biotech firm BioNTech, became the first fully tested vaccine to be approved for emergency use.62 Many vaccines were approved by several countries as shown in Table 3 . The strategies used by the vaccines developed by several countries are shown in Table 4 .Table 2 List of different types of COVID-19 Vaccines.
Table 2Vaccine Type Vaccine Candidate Vaccine Developer/Manufacturer Number of Doses Route of administration
Live-attenuated virus COVI-VAC Codagenix/Serum Institute of India (India) (1–2) Intranasal.
DelNS1-SARS-CoV-2-RBD University of Hong-Kong 2 Intranasal
Inactivated CoronaVac Sinovac (China) 2 Intramuscular;
Unnamed Wuhan Institute of Biological Products/Sinopharm
(China) 2 Intramuscular
BBIBP-CorV Beijing Institute of Biological Products/Sinopharm
(China) 2 Intramuscular
Unnamed Institute of Medical Biology, Chinese Academy of
Medical Sciences (China) 2 Intramuscular
BBV152 Bharat Biotech 2 Intramuscular
QazCovid-in Research Institute for Biological Safety Problems
(Kazakhstan) 2 Intramuscular
Unnamed Beijing Minhai Biotechnology Co (China) (1, 2, 3) Intramuscular
VLA2001 Valneva, National Institute for Health Research (UK) 2 Intramuscular
Unnamed Shifa Pharmed Industrial Co (Iran) 2 Intramuscular
ERUCOV-VAC Erciyes University (Turkey) 2 Intramuscular
DNA INO-4800 Inovio Pharmaceuticals/International
Vaccine Institute (USA) 2 Intradermal
AG0301-COVID19
AG0302-COVID19 Osaka University/AnGes/Takara Bio (Japan) 2 Intramuscular
ZyCoV-D Cadila Healthcare Limited (India) 3 Intradermal
bacTRL-Spike Symvivo (Canada) 1 Oral
GX-19 Genexine Consortium (South Korea) 2 Intramuscular
CORVax Providence Health & Services (USA) 2 Intradermal
Covigenix VAX-001 Entos Pharmaceuticals Inc. (Canada) 2 Intramuscular
GLS-5310 GeneOne Life Science, Inc. (South Korea) 2 Intradermal
COVIGEN University of Sydney, Bionet Co., Ltd.
Technovalia (Australia) 2 Intramuscular
COVID-eVax Takis/Rottapharm Biotech (Italy) – Intramuscular
RNA mRNA-1273 Moderna/NIAID (USA) 2 Intramuscular
BNT162b1, BNT162b2 BioNTech (Germany)//Pfizer (USA) 2 Intramuscular
CVNCOV Curevac (Germany 2 Intramuscular
ARCT-021 Arcturus (USA)/Duke-NUS (Singapore) – –
LNP-nCoVsaRNA Imperial College London (UK) 2 Intramuscular
SARS-CoV-2 mRNA vaccine Shulan (Hangzhou) Hospital/Center for Disease Control and Prevention of
Guangxi Zhuang Autonomous Region
(China) 2 Intramuscular
PTX-COVID19-B Providence Therapeutics (Canada) 2 Intramuscular
ChulaCov19 Chulalongkorn University (Thailand) 2 Intramuscular
Protein subunit NVX-CoV2373 Novavax (USA) 2 Intramuscular
ZF 2001 Anhui Zhifei Longcom Biopharmaceutical/ Institute of Microbiology, Chinese Academy of Sciences (China) 2 or 3 Intramuscular
KBP-COVID-19/KBP-201 Kentucky Bioprocessing, Inc. (USA) 2 Intramuscular
Unnamed Sanofi Pasteur (France)/GSK (UK) 2 Intramuscular
CORBEVAX Biological E Ltd. (India) 2 Intramuscular
SCB-2019 Clover Biopharmace 2 Intramuscular
COVAX-19 Vaxine Pty Ltd. (Australia)/Medytox (South Korea) 1 Intramuscular
MVC-COV1901 Medigen Vaccine Biologics Corporation (Taiwan)/
NIAID/Dynavax (USA 2 Intramuscular
Soberana 01 Center for Genetic Engineering and Biotechnology
(Cuba) 3 Intramuscular
Soberana 02 Center for Genetic Engineering and Biotechnology
(Cuba) 3 Intramuscular
Unnamed FBRI SRC VB VECTOR, Rospotrebnadzor,
Koltsovo (Russia) 2 Intramuscular
CoVAC-1 University Hospital Tuebingen (Germany) 1 Subcutaneous
UB-612 COVAXX (USA)/United Biomedical Inc. Asia (Taiwan) 2 Intramuscular
Unnamed Adimmune Corporation (Taiwan) – –
Unnamed Nanogen Pharmaceutical Biotechnology (Vietnam) 2 Intramuscular
S-268019 Shionogi Inc. (Japan) 2 Intramuscular
FINLAY-FR1 Instituto Finlay de Vacunas (Cuba) 2 Intramuscular
FINLAY-FR2 Instituto Finlay de Vacunas (Cuba) 2 Intramuscular
SARS-CoV-2-RBD-Fc fusion protein University Medical Center Groningen (Netherlands) +
Akston Biosciences Inc. (USA) – Subcutaneous or Intramuscular
COVAC-1 and COVAC-2
subunit vaccine (spike protein) + SWE adjuvant University of Saskatchewan (Canada) 2 Intramuscular
GBP510 SK Bioscience Co., Ltd. (South Korea) 2 Intramuscular
Razi Vaccine and Serum Research Institute (Iran) Razi Cov Pars, recombinant spike Protein 3 Intramuscular and Intranasal
MF59 adjuvanted SARS-CoV-2
Sclamp vaccine The University of Queensland (Australia) 2 Intramuscular
VLP Triple Antigen Vaccine Premas Biotech + Oramed Pharmaceuticals 1 Oral
Unnamed Medicago Inc. (Canada) 2 Intramuscular
Unnamed SpyBiotech/Serum Institute of India (India) 2 Intramuscular
Non-replicating viral vector AZD1222; ChAdOx1-S; ChAdOx1 nCoV-1 University of Oxford/AstraZeneca (UK) (1–2) Intramuscular
Ad5 nCoV CanSino Biological Inc./Beijing Institute of
Biotechnology (China) 1 Intramuscular
Sputnik V Gamaleya Research Institute (Russia) 2 Intramuscular
Ad26.COV2.S/JNJ-78436735 Johnson & Johnson (USA) (1–2) Intramuscular
hAd5-COVID-19/hAd5-S-Fusion + N-ETSD ImmunityBio, Inc. & NantKwest Inc. (USA) 1 Oral
Gard-CoV2 ReiThera (Italy)/LEUKOCARE (Germany)/Univercells (Belgium) 1 Intramuscular
COH04S1 City of Hope (USA) (1–2) Intramuscular
VXA-CoV2–1 Vaxart (USA) 2 Oral
MVA-SARS-2-S Ludwig-Maximilians - University of Munich
(Germany) 2 Intramuscular
BBV154 Bharat Biotech International Limited (India) 1 Intranasal
LV-SMENP-DC vaccine Shenzhen Geno-Immune Medical Institute
(China) 1 Intravenous and subcutaneous
Replicating viral vector Covid-19/aAPC vaccine Shenzhen Geno-Immune Medical Institute
(China) 3 Subcutaneous
VSV-S Israel Institute for Biological Research/ Weizmann Inst. of Science (Israel) 1 Intramuscular
Aivita Biomedical, Inc. (USA) Dendritic cell vaccine AV-COVID-19 1 Intramuscular
AdCLD-CoV19 Cellid Co., Ltd. (South Korea) – Intramuscular
NDV-HXP-S, Newcastle disease
virus vector Mahidol University; The Government
Pharmaceutical Organization (GPO);
Icahn School of Medicine at Mount Sinai
(Thailand) 2 Intramuscular
Table 3 List of COVID-19 vaccines approved by at least one country.
Table 3Protein Subunit Vaccine
Serial No. Company/Research Institute Vaccine Name
1. Anhui Zhifei Longcom Zifivax
2. Bagheiat-allah University of Medical School Noora Vaccine
3. Biological E Limited Corbevax
4. Center for Genetic Engineering and Biotechnology (CIGB) Abdala
5. Instituto Finlay de Vacunas Cuba Soberana 02
6. Instituto Finlay de Vacunas Cuba Soberana Plus
7. Livzon Mabpharma Inc V-01
8. Medigen MVC-COV1901
9. National Vaccine and serum Institute Recombinant SARS-CoV-2 Vaccine (CHO Cell)
10. Novavax Nuvaxovid
11. PT Bio Farma IndoVac
12. Razi Vaccine and serum Research Institute Razi Cov Pars
13. Sanofi/GSK VidPrevtyn Beta
14. Serum Institute of India COVOVAX (Novavax formulation)
15. SK Biosciences Co Ltd SKYCovione
16. Takeda TAk-019 (Novavax formulation)
17. Vaxine/ CinnaGen Co. SpikoGen
18. Vector State Research Center of Virology and Biotechnology Aurora-CoV
19. Vector State Research Center of Virology and Biotechnology EpiVacCorona
Inactivated Vaccine
Serial No. Company/Research Institute Vaccine Name
1. Bharat Biotech Covaxin
2. Chumakov Center KoviVac
3. Health Institutes of Turkey Turkovac
4. Organization of Defensive Innovation and Research FAKHRAVAC (MIVAC)
5. Research Institute for Biological Safety Problems (RIBSP) QazVac
6. Shenzhen Kangtai Biological Products Co KCONVAC
7. Shifi Pharmed Industrial Co COVIran Barekat
8. Sinopharm (Beijing) Covilo
9. Sinopharm (Wuhan) Inactivated (Vero Cells)
10. Sinovac CoronaVac
11. Valneva VLA2001
Non Replicating Viral Vaccine
Serial No. Company/Research Institute Vaccine Name
1. Bharat Biotech iNCOVACC
2. Cansino Convidecia
3. Cansino Convidecia Air
4. Gamaleya Gam-COVID-Vac
5. Gamaleya Sputnik Light
6. Gamaleya Sputnik V
7. Janssen (Johnson & Johnson) Jcovden
8. Oxford/AstraZeneca Vaxzevria
9. Serum Institute of India Covishield (Oxford/AstraZeneca formulation)
RNA Vaccine
Serial No. Company/Research Institute Vaccine Name
1. Gennova Biopharmaceuticals Limited GEMCOVAC-19
2. Moderna Spikevax
3. Moderna Spikevax Bivalent Original/Omicron BA.1
4. Moderna Spikevax Bivalent Original/Omicron BA.4/ BA.5
5. Pfizer/BioNTech Comirnaty
6. Pfizer/BioNTech Comirnaty Bivalent Original/Omicron BA.1
7. Pfizer/BioNTech Comirnaty Bivalent Original/Omicron BA.4/ BA.5
8. Takeda TAK-919 (Moderna formulation)
9. Walvax AWcorna
VLP (Virus Like Particle) Vaccine
Serial No. Company/Research Institute Vaccine Name
1. Medicago Covifenz
DNA Vaccine
Serial No. Company/Research Institute Vaccine Name
1. Zydus Cadila ZyCoV-D
Table 4 Types of COVID-19 Vaccine developed across the globe.
Table 4Type Manufacturer Name Stabilizing mutations Virus strain Eukaryotic production cell line Dosage References
mRNA BioNTech-Pfizer (Germany, USA) BNT162b2, Comirnaty Yes (prolines) Wuhan-Hu-1 not applicable 30 μg RNA (2x) 104,105
mRNA Moderna-NIAID (USA) mRNA-1273, COVID-19 Vaccine Moderna yes (prolines) Wuhan-Hu-1 not applicable 100 μg RNA (2x) 106,107
mRNA CureVac (Germany) CVnCoV yes (prolines) Wuhan-Hu-1 not applicable 12 μg RNA (2x) 108
Adenovector University of Oxford-AstraZeneca (UK, Sweden) COVID-19 vaccine AstraZeneca, AZD1222, ChAdOx1-S, Vaxzeria; Covishield no Wuhan-Hu-1 HEK293 5 × 1010 adenovirus vector particles (2x) 109
Adenovector CanSino Biological Inc., Beijing Institute of Biotechnology (China) Ad5 nCoV, Convidecia no Wuhan-Hu-1 HEK293 5 × 1010 adenovirus vector particles (2x) 111
Adenovector Gamaleya Research Institute (Russia) rAd26-S + rAd5-S, Gam-COVID-Vac, Sputnik V no Wuhan-Hu-1 (probably) HEK293 10 × 1010 adenovirus vector particles (2x) 112,113
Adenovector Janssen-Johnson & Johnson (NL/USA) Ad26.COV2.S, COVID-19 Vaccine Janssen yes (prolines, furin cleavage site) Wuhan-Hu-1 PER.C6 5 × 1010 adenovirus vector particles (1x) 114., 115., 116.
Inactivated whole virus Sinopharm,Beijing Institute of Biological Products Co (China) BBIBP-CorV, Sinopharm COVID-19 vaccine not applicable Wuhan-Hu-1-like HB02 strain Vero 4 μg proposed (2x) 117,118
Inactivated whole virus Sinovac (China) CoronaVac not applicable Wuhan-Hu-1-like CN2 strain Vero 3 μg proposed (2x) 119,120
Inactivated whole virus Bharat Biotech (India) Covaxin, BBV152 not applicable NIV2020–770 (D614G) Vero 6 μg proposed (2x) 121., 122., 123.
Subunit Novavax (USA) NVX-CoV2373 yes (prolines, furin cleavage site) Wuhan-Hu-1 Sf9 5 μg S (+ 50 μg adjuvant) (2x) 124,125
Several strategies were accepted in the development of Coronavirus vaccines (Table 4), and most of these strategies targeted the S-protein or the surface-exposed spike (S) as the major inducer of neutralizing antibodies.64 Jha et al., reported the antigenicity as well as allergenicity of different structural proteins of SARS-CoV-2 to design vaccines against SARS-CoV-2. This analysis showed that the envelope protein (E) to be highly antigenic having the antigenicity of 0.6025 followed by Membrane glycoprotein (M) having antigenicity of 0.5102, Nucleocapsid phosphoprotein having antigenicity of 0.5059 and Surface glycoprotein (S) with antigenicity of 0.4696.46 S protein plays a major role in the stimulation of protective immunity during infection with SARS-CoV-2 by evoking neutralizing antibodies and T-cell responses. Hence, the full length or partial protein of S-glycoprotein can be the most effective vaccine target against coronavirus.9 , 12 , 64 , 65 The vaccine designing strategy includes both immunoinformatics and experimental studies for getting a suitable candidate for vaccine candidate. The immunoinformatics study helps in selection of candidate whereas experimental studies confirm the candidate for vaccine production (Fig. 5 ).Fig. 5 Flow diagram showing the immunoinformatics and experimental study involved in vaccine development.
Fig. 5
COVID-19 vaccine development process started with the genome sequence of SARS-CoV-2, first available on January, 11, 2020. (Fig. 6 ). The preclinical data for SARS-CoV and MERS saved considerable time eliminating the initial step of exploratory phase.43 Multiple clinical trials initiated to reduce the time horizon where one phase was immediately followed by second (Fig. 7 ).43 The different strategies used in COVID-19 vaccine development is shown in Fig. 8 . A list of COVID-19 vaccines is shown in Table 5 , while those approved by WHO is shown in Table 6 .Fig. 6 COVID-19 Vaccine Development (12–24 months).
Fig. 6
Fig. 7 Phases in vaccine development.
Fig. 7
Fig. 8 The vaccine development strategy used for COVID-19 in India.
Fig. 8
Table 5 List of approved vaccines for COVID-19; data retrieved from https://www.raps.org/news-and-articles/news-articles/2020/3/covid-19-vaccine-tracker accessed on 23 August 2021.
Table 5Vaccine Candidate Vaccine Type Primary developers Country of Origin
Comirnaty (BNT162b2 mRNA-based vaccine Pfizer, BioNTech, Fosun Pharma Multinational
Moderna COVID-19 Vaccine
(mRNA-1273) mRNA-based vaccine Moderna, NIAID, BARDA US
COVID-19 Vaccine AstraZeneca (AZD1222); also known as Vaxzevria and Covishield Adenovirus vaccine BARDA, OWS UK
Sputnik V Recombinant adenovirus vaccine (rAd26 and rAd5) Gamaleya Research Institute, Acellena Contract Drug Research and Development Russia
Sputnik Light Recombinant adenovirus vaccine (rAd26) Gamaleya Research Institute, Acellena Contract Drug Research and Development Russia
COVID-19 Vaccine Janssen (JNJ-78436735; Ad26.COV2.S) Non-replicating viral vector Janssen Vaccines (Johnson & Johnson) The Netherlands, US
CoronaVac Inactivated vaccine (formalin with alum adjuvant) Sinovac China
BBIBP-CorV Inactivated vaccine Beijing Institute of Biological Products; China National Pharmaceutical Group (Sinopharm) China
EpiVacCorona Peptide vaccine Federal Budgetary Research Institution State Research Center of Virology and Biotechnology Russia
Convidicea (PakVac, Ad5-nCoV) Recombinant vaccine (adenovirus type 5 vector) CanSino Biologics China
Covaxin (BBV152) Inactivated vaccine Bharat Biotech, ICMR; Ocugen; ViroVax India
WIBP-CorV Inactivated vaccine Wuhan Institute of Biological Products; China National Pharmaceutical Group (Sinopharm) China
CoviVac Inactivated vaccine Chumakov Federal Scientific Center for Research and Development of Immune and Biological Products Russia
ZF2001 (ZIFIVAX) Recombinant vaccine Anhui Zhifei Longcom Biopharmaceutical, Institute of Microbiology of the Chinese Academy of Sciences China, Uzbekistan
QazVac (QazCovid-in) Inactivated vaccine Research Institute for Biological Safety Problems Kazakhstan
Unnamed vaccine candidate Inactivated vaccine Minhai Biotechnology Co.; Kangtai Biological Products Co. Ltd. China
Unnamed vaccine candidate Inactivated vaccine Chinese Academy of Medical Sciences, Institute of Medical Biology China
COVIran Barekat Inactivated vaccine Shifa Pharmed Industrial Group Iran
Abdala (CIGB 66) Protein subunit vaccine Center for Genetic Engineering and Biotechnology Cuba
Soberana 02 Conjugate vaccine Finlay Institute of Vaccines; Pasteur Institute Cuba, Iran
MVC-COV1901 Protein subunit vaccine Medigen Vaccine Biologics Corp.; Dynavax Taiwan
Table 6 List of COVID-19 vaccines approved by World Health Organization (WHO).
Table 6Vaccine Name Manufacturer Vaccine Type Date of approval by WHO
BNT162b2/COMIRNATY
Tozinameran (INN) Pfizer BioNTech Nucleoside modified mRNA 31 December 2020
Covishield (ChAdOx1_nCoV-19) Serum Institute of India Pvt. Ltd. Recombinant ChAdOx1
adenoviral vector encoding
the Spike protein antigen of
the SARS-CoV-2. 15 February 2021
Vaxzevria (ChAdOx1-S [recombinant]) AstraZeneca + University of Oxford Chimpanzee Adenovirus encoding the SARS-CoV-2 Spike glycoprotein (ChAdOx1-S) 15 February 2021
mRNA-1273 Or Spikevax Moderna mNRA-based vaccine
encapsulated in lipid
nanoparticle (LNP) 30 April 2021
Ad26.COV2.S Janssen Pharmaceutical Companies and Johnson & Johnson Recombinant, replication-
incompetent adenovirus type 26 (Ad26) vectored vaccine encoding the (SARS-CoV-2) Spike (S) protein 12 March 2021
CoronaVac Sinovac Inactivated, produced in
Vero cells 1 June 2021
BBIBP-CorV Sinopharm Inactivated, produced in
Vero cells 7 May 2021
Vaccine based on live -attenuated SARS-CoV-2 virus• DelNS1-SARS-CoV-2-RBD by University of Hong-Kong
DelNS1-SARS-CoV-2-RBD by University of Hong Kong is an example of vaccine based on attenuated or weakened SARS-CoV-2 virus. This vaccine uses flu vector to express a particular antigen to induce immunity targeting the critical element of Receptor Binding Domain (RBD) of SARS-CoV-2. DelNS1-SARS-CoV-2-RBD is one of the 5 vaccine technologies by China's Ministry of Science and Technology.66 This live attenuated vaccine (LAV), cultivated in the chick embryo or Madin Darby Canine Kidney Cells (MDCK cells) and administered intranasally.17
Vaccine based on inactivated SARS-CoV-2 virus
SARS-CoV-2 is inactivated or killed by using different chemical techniques and the candidate vaccines under this group are injected intramuscularly.67 • CoronaVaC by Sinovac Biotech
It is a purified SARS-CoV-2, inactivated vaccine candidate which was previously known as PiCoVacc. It is designed by cultivating the SARS-CoV- 2 CN2 Strain inside Vero cells and in activating it with β-propiolactone. According to the preclinical trials it stimulates SARS-CoV-2 specific neutralizing antibodies (nAbs) in Rhesus Macque, rats and mice.68 On June 1, CoronaVaC vaccine was approved by WHO for emergency use listing (EUL) and finally approved for use in 26 countries on June 9, 2021. According to data from Brazilian trial this vaccine has an efficiency rate of 50.4% for prevention of symptomatic infection.
• Covaxin Bharat Biotech India
The research name of Covaxin is BBV152. It is India's indigenous vaccine developed in collaboration with ICMR (Indian Council of Medical Research) and NIV (National Institute of Virology). It has been approved by 9 countries which include India, Iran, Mauritius, Mexico, Nepal, Guyana, Paraguay, Zimbabwe and Philippines.
• BBIBP-CorV vaccine by Sinopharm
It is an inactivated vaccine developed by Beijing Bio-Institute of Biological products from strain.69 It is a Chinese state-owned company uses inactivated SARS-CoV-2 virus and the clinical trials showed that it had an efficacy rate of 79%. For the development of BBIBP-CorV, the researchers obtained three variants of coronavirus from patients, and the variants which multiplies fast in monkey kidney cells was selected. β-propiolactone is used to inactivate coronaviruses and hence it doesn't replicate inside the host. The proteins (including spike) of coronavirus remains intact hence, further mixed with a small amount of adjuvant (aluminum based compound) to boost response.70 This vaccine completed its phase III in Argentina, Bahrain, Egypt, Morocco, Pakistan, Peru and UAE and on 7th May, 2021, WHO listed Sinopharm COVID-19 vaccine for emergency use.
Vaccine based on SARS-CoV-2 protein subunit
• Novavax (NVX-CoV2373)
The Novavax COVID-19 vaccine, NVX-CoV2373, a type of protein subunit vaccine uses nanoparticle based vaccine, designed by Novavax and Coalition for Epidemic Preparedness Innovations (CEPI). These vaccines are under trial in India under the brand name Covovax.71 This US based Novavax Inc. has a manufacturing contract with Serum Institute of India.
This vaccine is designed by creating an engineered baculovirus containing a gene for the modified SARS-CoV-2 spike protein. The S-protein was altered by the incorporation of two proline residues hence, stabilizing the pre-fusion form of protein. The baculovirus infects the culture of SF9 mother cells which forms the spike protein and displays it on the cell membranes. Spike proteins are cultivated and assembled onto a synthetic lipid nanoparticle purifying spike protein. Matrix M (based on a saponin) obtained from the soapbark tree (Quillaja saponaria), used as adjuvant in this vaccine.72 These vaccines can be stored and handled at above freezing temperature (35–46 °F) and administered as two intramuscular injections.
Vaccine based on virus-like particle (VLP)
• Triple Antigen Vaccine by Premas Biotech or Orovax vaccine
Premas Biotech (India) in collaboration with Oramed Pharmaceuticals (Jerusalem) designed an oral vaccine to be swallowed as a pill instead of being injected. This is a VLP vaccine prototype which acts on 3 surface proteins of the SARS-CoV-2 virus; the spike, the membrane and the envelope protein. Premas known for developing recombinant proteins for vaccine development, such proteins are ‘difficult to express’ proteins (DTE-Ps). The triple proteins of SARS-CoV-2 have been co-expressed in an engineered Saccharomyces cerevisiae (D-crypt).17
Vaccine based on DNA
• INO-4800
This vaccine was developed by Inovio Pharmaceuticals in partnership with Beijing Advaccine Biopharmaceuticals Suzhou. This strategy utilizes codon optimized S protein sequence of SARS-CoV-2 to which an IgE leader sequence is attached.17 The INO-4800 vaccine contains the plasmid pGX9501 encoding the entire length of spike glycoprotein of SARS-CoV-2. Inovio's proprietary platform uses CELLECTRA, a brief electrical pulse to open small pores in the cell reversibly to allow the plasmids to enter. INO-4800 can be injected intradermally with subsequent electroporation to transfer DNA plasmid directly into blood cells. According to the Lancet on December 23, 2020, INO-4800 showed excellent safety, tolerability as well as immunogenic in 100% of the vaccinated volunteers by stimulating either the cellular or humoral immune responses or both.
Vaccine based on RNA
• mRNA − 1273 by Moderna
Spikevax, the brand name of mRNA-1273 vaccine, designed by Moderna, the United States National Institute of Allergy and Infectious Diseases (NIAID) and the Biomedical Advanced Research and Development Authority (BARDA). This vaccine comprises of synthetic mRNA enclosed in lipid nanoparticle (LNP) encoding the full length pre fusion stabilized spike protein (S) of SARS-CoV-2, therefore, stimulates a highly S-protein specific antiviral response (Fig. 9 ).17 This vaccine is designed on the basis of SARS and MERS and administered intramuscularly in deltoid muscle.1 , 73 The levels of nAb surpassed the levels found in convalescent sera after the administration of 250 μg dose levels.17 , 74 Initial efficacy assessment of Phase 3 CoV study of mRNA-1273 includes 30,000 subjects consisting of 196 cases of COVID-19 of which 30 cases being severe. The efficacy of this vaccine against COVID-19 was 100% and hence on 30th April, 2021, Moderna COVID-19 vaccine became the fifth vaccine to receive emergency validation from WHO.• BNT162b1 (BioNTech | Fosum Pharma | Pfizer)
It is the first vaccine authorized by FDA Emergency Use Authorization (EUA) to prevent COVID-19 on December 11, 2020. On December 31, 2020, the WHO issued an EUL for BNT162b1 vaccine. It is a codon-optimized mRNA vaccine that encodes for the trimerized SARS-CoV-2 RBD.17 It is a liquid nanoparticle-formulated, nucleoside-modified RNA that encodes optimized SARS-CoV-2 full length spike protein.75 The vaccine displays an elicited immunogenicity as it uses additional T4 fibritin-derived fold on trimerization domain to the RBD antigen.17 The vaccine is based on Germany-baded BioNTech SE proprietary mRNA tech and was co-developed by BioNTech and Pfizer. BioNTech also collaborated with Fosun Pharma on March, 2020 and Fosun Pharma became the strategic partner of BioNTech in China.
Fig. 9 A pictorial comparison between DNA & mRNA vaccine with regard to their mechanism. DNA vaccine is in the form of circular DNA and contains the spike gene of SARS-COV-2. Electroporation enhances the permeability of plasma membrane which allows the entry of DNA into the cytoplasm and ultimately to nucleus. In nucleus, DNA is transcribed into mRNA which is translated into spike proteins of SARS-COV-2. These proteins are expressed on cell membrane. mRNA vaccines are encapsulated in nanoparticles and are integrated into cytoplasm. Spike proteins are produced which are expressed on cell membrane. These proteins are recognized by antigen presenting cells (APC) which triggers an immunological response.
Fig. 9
Vaccines based on viral vectors
• Ad5-nCoV (CanSino Biologics Inc | Beijing Institute of Biotechnology)
This is a non-replicating viral vector vaccine, administered with a single dose. This vaccine is based on CanSino BIO's adenovirus viral vector vaccine technology. The vaccine is designed by using the Admax system from the Microbix Biosystem.76 The CanSino Biologics Convidicea is a genetically engineered vaccine candidate with the replication defective adenovirus type 5 as the vector to express SARS-CoV-2 spike protein. The attenuated adenovirus infects human cell readily, incapable of causing disease and hence, delivers genetic material coding spike protein. Therefore, S proteins are produced which travels to the lymph nodes, where the immediate response produces antibodies.
• Ad26.COV2.S or Janssen COVID-19 vaccine or Johnson & Johnson COVID-19 vaccine
A non-replicating vaccine developed by Janssen Vaccines in Leiden, Netherlands, and its Belgian parent company Janssen Pharmaceuticals, subordinate of Johnson & Johnson, based on AdVac and PER.C6® technologies. Janssen's AdVac® vectors are genetically modified adenoviruses which mimics components of pathogens. Advac viral vector can induce long-term humoral as well as cellular immune responses. Janssen produces neutralizing antibodies against a range of SARS-CoV-2 variants such as Delta (B.1.617.2) variant, the partially neutralization-resistant Beta (B.1.351) variants, the Gamma (P.1) variants and others, including the Alpha (B.1.1.7), Epsilon (B.1.429), Kappa (B.1.617.1) and D614G variants, the original SARS-CoV-2 strain (WA1/2020) as well as on Omicron variant. This vaccine produces durable immune response for upto 8 months after vaccination.
COVID-19 vaccine status in India
India, the second largest populated country in the world, with several pharmaceutical manufacturing bodies plays a central role in the COVID-19 vaccine development.77 India began administrating COVID-19 vaccines on 16th January, 2021 and the outcome of this vaccination drive reveals 94% of the Indian population has received at least first dose and 86% of the eligible Indian population has received both doses of vaccine.78 The list of COVID-19 vaccines approved in India is shown in Table 7 and in Fig. 10 .Table 7 List of COVID-19 Vaccines approved in India.
Table 7Vaccine Name Manufacturer Vaccine Type
ZyCoV-D Zydus Cadila DNA
mRNA Moderna RNA
Sputnik-V Gamaleya Non-Replicating Viral Vector
Covishield Serum Institute of India Non-Replicating Viral Vector
Covaxin Bharat Biotech Inactivated
AZD1222 Oxford/AstraZeneca Non-Replicating Viral Vector
Ad26. COV2. S Janssen (Johnson & Johnson) Non-Replicating Viral Vector
Fig. 10 Figure showing a. approved vaccines and b. under clinical trial vaccines.
Fig. 10
India approved Oxford- AstraZeneca (manufactured under license by Serum Institute of India under the trade name Covishield and Covaxin (Table 8, Table 9 ). The list of individuals vaccinated throughout the country state-wise is shown in Table 10 .Table 8 List of Indigenous Vaccines of India.
Table 8Vaccine Candidate Manufacturer Status in India
Covaxin Bharat Biotech Approved
Zycov-D Zydus Cadila Approved
BECOV Biological E/ Baylor College of Medicine Not yet approved
HGCO19 Gennova (Emcure) Not yet approved
BBV154 Bharat Biotech Not yet approved
UB-612 Aurobindo/ Covaxx Not yet approved
Table 9 List of Non-Indigenous Vaccines of India or Global Vaccines with Indian Partners.
Table 9Vaccine Candidate Global Vaccine Indian Partner
Covishield AstraZeneca/Oxford Serum Institute
Sputnik V Gamaleya Institute, Russia Dr Reddy's
JNJ-78436735 J&J Biological E
NVX-CoV2373 Novavax Serum Institute
Orovax Oramed Pharmaceuticals Premas Biotech
Tozinameran
(BNT162b2) Pfizer None
mRNA-1273 Moderna None
Table 10 Cumulative coverage report of COVID-19 Vaccination (As on 5th May 2022) (Source: www.mohfw.gov.in).
Table 10Name of State/UT No. of adults vaccinated (2nd Dose) No. of kids vaccinated (2nd Dose) No. of senior citizens vaccinated (HCW, FLW) (2nd Dose)
Andaman & Nicobar Islands 3,12,311 24,052 19,854
Andhra Pradesh 4,33,45,977 3,785,008 22,28,946
Arunachal Pradesh 7,22,018 32,881 30,120
Assam 1,95,93,572 824,610 3,87,168
Bihar 5,43,36,293 3,563,419 10,76,053
Chandigarh 8,99,701 36,305 39,134
Chhattisgarh 1,70,30,332 986,036 4,85,864
Dadra & Nagar Haveli 3,26,469 16,896 4167
Daman & Diu 2,59,017 15,723 6891
Delhi 1,38,07,403 1,009,879 6,45,496
Goa 12,06,313 58,483 40,587
Gujarat 4,80,63,152 3,394,627 26,24,496
Haryana 1,81,86,202 759,705 3,84,241
Himachal Pradesh 56,90,284 399,846 2,75,038
Jammu & Kashmir 1,01,67,472 891,086 3,90,278
Jharkhand 1,51,81,022 899,066 3,11,572
Karnataka 4,83,33,434 2,344,056 17,07,789
Kerala 2,34,78,782 788,782 15,26,915
Ladakh 1,83,157 11,939 37,696
Lakshadweep 55,199 2542 3505
Madhya Pradesh 5,29,55,202 3,586,205 10,56,773
Maharashtra 6,90,79,060 3,408,082 22,62,312
Manipur 11,58,681 48,640 85,597
Meghalaya 9,99,332 26,928 37,616
Mizoram 6,45,024 38,888 41,075
Nagaland 6,63,546 30,787 28,837
Odisha 2,84,37,883 1,914,820 11,35,482
Puducherry 6,57,012 47,951 21,076
Punjab 1,75,58,313 726,407 5,50,865
Rajasthan 4,41,81,824 2,851,866 17,05,896
Sikkim 5,02,436 29,364 36,619
Tamil Nadu 4,48,98,186 2,520,139 8,63,696
Telangana 2,81,52,478 1,827,455 6,61,422
Tripura 22,35,887 110,107 98,275
Uttar Pradesh 13,13,23,510 10,044,968 27,80,016
Uttarakhand 77,68,162 420,718 4,99,442
West Bengal 5,94,21,814 3,681,930 25,22,397
COVAXIN
Covaxin (BBV152), developed by Indian pharmaceutical company Bharat Biotech in collaboration with Indian Council of Research (ICMR) and National Institute of Virology (NIV) is an inactivated vaccine.79 The vaccine designed using whole virion inactivated Vero cell derived technology with a toll- like receptor 7/8 agonist molecule adsorbed to alum (Algel-IMDG).80 , 81 The NIV-2020-770 strain isolated from ‘Vero CCL-81’ cells with sequence in GISAID (EPI_ISL_420545).82 , 83 It uses a complete infective SARS-CoV-2 viral particle with RNA surrounded by a protein shell but contain dead virus incapable of infection.84 The vaccine was assigned for Phase I and Phase II human clinical trials by DCGI in July, 2020 with two doses in 28 days.80
Covaxin consists of 6 μg of whole virion inactivated SARS-CoV-2 antigen, inactive ingredients such as aluminum hydroxide gel (250 μg), TLR 7/8 agonist (imidazoquinoline) 15 μg, 2-phenoxyethanol 2.5 mg and phosphate buffer saline up to 0.5 ml. The sticks of coronaviruses were produced and soaked with beta-propiolactone. This compound disabled the replicating ability of coronaviruses but their proteins including spike remained intact. Once an individual is being vaccinated some of the inactivated viruses are engulfed by antigen presenting cells. The APCs process and present the coronavirus on its surface for recognition by T-helper cells. When B cell surface proteins latch onto the coronavirus, B cell locks on and pull part all of the virus inside and present coronavirus fragments on its surface. As the B cells get activated it proliferates and produces antibodies that can target the spike proteins.65
In Phase I clinical trial against hCoV-19/India/2020770 (homologous), and two heterologous strains from the unclassified cluster, namely, hCoV-19/India/2020Q111 and hCoV-19/India/2020Q100, BBV152 elicited a remarkable neutralizing antibody response.83 In Phase II clinical trial, with 6 and 3 antigen in imidazoquinoline (TLR7/TLR8 agonist adsorbed on aluminum hydroxide gel), the vaccine exhibited significant results in plaque reduction neutralization test (PRNT50)-based assay.81 According to the press release on 3rd March, 2021 of the Indian Council of Medical Research (ICMR), Phase III results of the Covaxin has shown an interim vaccine efficacy of 81%.
The formulation containing the toll like receptor 7/8 agonist also induced T helper cell 1 based antibody response with elevated levels of SARS-CoV-2 specific interferon gamma and CD4 cells. Covaxin has shown its neutralizing property against the variants- B.1.1.7 (Alpha), P.1- B.1.1.28 (Gamma) & P.2 – B.1.1.28 (Zeta), B.1.617 (Kappa), B.1.351 and B.1.617.2 (Beta & Delta). The efficacy data of Covaxin signifies 65.2% protection against the SARS-CoV-2, B.1.617.2 Delta variant and is 93% effective against severe disease.80
COVISHIELD
Covishield (AZD1222) developed at the Jenner Institute, University of Oxford in UK and licensed from British pharmaceutical company AstraZeneca is a non-replicating viral vaccine. Covishield got first approval for restricted use on January 3, 2021. The AZD1222 (ChAdOx1 nCoV- 19 Corona Virus Vaccine (Recombinant), designed at Oxford University consists of a replication deficient chimpanzee adenoviral vector ChAdOx1, with SARS-CoV-2 surface glycoprotein gene.85
The Covisheld vaccine prepared using L-Histidine, L-Histidine hydrochloride monohydrate, Magnesium chloride hexahydrate, Polysorbate 80, Ethanol, Sucrose, Sodium chloride, EDTA and water for injection. The Oxford-AstraZeneca team used a modified version of chimpanzee adenovirus (ChAdOx1) which can enter cells but can't replicate inside them. The vaccine can be refrigerated at 38–46 °F for the vaccine to last at least for 6 months. After the vaccine injection, the adenovirus encounters the cells and affix onto proteins on the cell's surface. The cell swallows up the virus in a bubble and pulls it inside; once the virus gains entry inside, it leaves the bubble and moves to the nucleus, where the cell's DNA is stored. The adenovirus pushes its DNA into the nucleus but it can't replicate itself whereas the coronavirus spike protein copies itself. These protruding spikes and spike protein fragments acts as recognition particles for the immune system. When the vaccinated cell dies, the broken cells contain spike proteins and fragments that can be taken up by antigen presenting cells. B cells may collide with the coronavirus spikes and get activated by helper T cells resulting in antibody formation and prevent infection by blocking the spikes from attaching to other cells.65
The phase 3 trial confirms that the group which received a low first dose vaccination followed by a standard second dose demonstrated 90.0% efficacy while the group which received a standard dose followed by a booster dose showed 62.1% efficacy; the overall efficacy at least 2 weeks after the second dose of vaccine was therefore calculated to be 70.4%.
ZyCoV- D
This is a DNA based vaccine, designed by Ahmedabad based Zydus Cadila in collaboration with National Biopharma Mission (NBM) and the Department of Biotech, Government of India. It is the India's first indigenously developed DNA vaccine. This vaccine has been approved for emergency use in India by DCGI for adults and children above the age of 12 years. This is India's first needle free COVID-19 vaccine which is administered with a disposable needle -free injector. The vaccine contains plasmid that has gene encoding the spike protein of SARS-CoV-2.86 The expression and localization of S-protein expressed by ZyCoV-D were assessed using immunofluorescence assay by Dey et al. The immunofluorescence studies with rabbit anti S1 antibody depicted a strong signal in the Vero cells transfected with ZyCoV-D. This study demonstrates the ability of the ZyCoV-D vaccine to express efficiently in mammalian cells and can induce antibodies production.86 ZyCoV-D was assessed in vivo in different animal models and has showed the potential to elicit immune responses against SARS-CoV-2 S antigen. ZyCoV-D can also induce secondary immune response as the serum IgG levels against spike protein detected even after three months after the last dose.86 According to an interim study, the three doses of ZyCoV-D vaccine prevents symptomatic disease in 66% of vaccinated individuals. According to Momin et al., ZyCoV-D vaccine was found to be safe, well tolerated and immunogenic in phase 1 trial.87
HGC019
HGC019, India's first mRNA vaccine made by Pune- based Gennova biopharmaceuticals in collaboration with Seattle-based HDT Biotech Corporation uses bits of genetic code to stimulate an immune response. This mRNA based vaccine candidate contains a short, synthetic version encoding the spike (antigen) protein of SARS-CoV-2. Upon injection into the person's body, the synthetic mRNA taken to muscle cells, where multiple copies of antigen is formed. The mRNA is associated with the lipid inorganic nanoparticle (LION™) which acts as mRNA vaccine delivery system equilibrating the mRNA and also acts as adjuvant. The vaccine is unique as it uses the most prominent mutant of spike protein, D614G. The vaccine is stable at 2–8 °C and uses the absorption chemistry to attach mRNA to the nano-lipid carrier's surface to intensify the release kinetics of mRNA within the cells compared to the encapsulation chemistry. On 24 August, 2021, DCGI gave a nod for phase II/ III trials to HGC019 vaccine after its positive results of Phase I trial.
CORBEVAX
Corbevax (BioE COVID-29 or BECOV2D), a protein subunit vaccine developed by Biological E Limited (BioE) an Indian biopharmaceutical company, American company Dynavax Technologies (DVAX), the Baylor College of Medicine in Houston, US. It is a recombinant protein subunit vaccine which uses spike RBD, adsorbed to the adjuvant alum. Four compounds with these components are currently being assessed in Phase I/ II clinical study in India to select the final vaccine candidate to be examined in subsequent Phase III trials. In preclinical models, combination of Alum with Dynavax Technologies Corporation's CpG with N1C1 antigen evoked a highly synergistic, balanced immune response.88
SPUTNIK V
Sputnik V (Gam-COVID-Vac), non-replicating vector vaccine designed by Gamaleya National Research Centre for Epidemiology and Microbiology of the Ministry of Health of the Russian Federation, Moscow, Russia. Scientists adjoined the gene for coronavirus spike protein to two types of adenovirus i.e., Ad26 and Ad5 and modified them so that they could penetrate cells but could not replicate. The use of two varying serotypes is a unique technique that provides effective long term immunity and boosts the immune response. The first dose (based on Ad26) is administered and injected on the first day and the second dose (based on Ad5) is administered on the second day to boost immune response.89 , 90
The vaccine trial from Moscow reported an efficacy of 91.6% after the second dose for all age groups with no unusual side effects.88 Sputnik V's efficacy against the Delta variant was found to be 83.1% and also showed a six times reduction in infection risk as confirmed by Russian Ministry of Health on July, 2021. According to the results of Phase III clinical trials of Sputnik V, after getting the vaccine 98% of people developed humoral immune response and 100% of people developed cellular immune response. Serum Institute of India got the authorization on June 4, 2021 by DCGI to manufacture Sputnik V COVID-19 at its Hadapsar facility.
Vaccination in India is an arduous task with a country having a population that covers 1.3 billion people. Free vaccination to everyone in India is being provided to all above the age of 18 from June, 21 as per the announcement by Prime Minister Shri Narendra Modi. While the vaccine program is gaining momentum in India, trust in vaccine by the people and their acceptances among the citizens are key determinants of the success of any vaccination program. Across the globe there are sections of people who are facing vaccine hesitancy with myths and misinformation's circulating on different platforms. Government of India is already running several awareness campaigns to establish trust for vaccination among people and dispel hesitancy. To streamline the process of vaccination, Indian government developed a digital platform Co-Win where one could book an appointment for vaccination, get the trusted information and check the status of vaccines.89
Effect of SARS-CoV-2 vaccine on Variant of Concern (Omicron)
Omicron represents a highly mutated version of SARS-CoV-2.90 The Technical Advisory Group on SARS-CoV-2 (TAG-VE) of WHO classified the B.1.1.529 strain known as omicron as a Variant of Concern (VOC). Omicron was first detected in South Africa in November, 2021.91 As per data of Global Science and Primary Sources (GISAID), 4992 omicron sequences have been reported by 57 countries till December, 2021. Omicron is worrisome variant as it has more than 50 mutations, out of which 30 are spike protein mutations. 15 mutated sites are located within the receptor-binding domain (RBD), the portion which interacts with host cells before entry inside the cell and hence an increase in transmission rate of infection is possible.92
Spike protein sequences analysis led to identification of two sub clades of omicron. Sub-Clade 1 has a lower sequence frequency and has mutation at three sites, namely 417 K, 440 N and 440 G. Sub-clade II has a high occurrence globally and has mutation sites at 417 N, 440 K and 446 S. Omicron shares several common mutation with delta variant of SARS-CoV-2, however many additional mutations can increase the infectivity rate of this variant. As per scientific data available till date, no evidence suggests that omicron has a greater severity than other VOCs. However, many issues such as increased transmission rate, virulence, elevated risk reinfection and possible decline in efficacy of vaccines and other therapeutics remain unresolved. A simulated study involving artificial intelligence (AI) tried to analyze the effect of RBD mutations on the infectivity of omicron and efficacy of available vaccines. The study showed that mutation involving N440K, T478K and N510Y might increase infectivity of omicron by 2–10 times in comparison to delta variant.93 Studies have also shown that omicron can cause reinfection at a greater rate.94 , 95
The idea that omicron variant can evade vaccine-mediated immunity remains unclear, however sudden spike in omicron positivity rate together with greater number of hospitalization in South Africa remain a major concern and further studies are required.96 In order to counter omicron variant, Pfizer and BioNtech are preparing to alter their mRNA vaccine shots.27 , 38 , 86 Pfizer aims to identify escape variant in omicron in order to design omicron specific vaccine. However, lack of scientific data suggests that it will be too early to make any conclusions regarding efficacy of existing vaccines against omicron variant.
Current landscape and market size for vaccine development in India
A lot of struggle around the globe has resulted in development of vaccines in last 60 years. World Health Organization (WHO) helped the world with its outreach programme to bring vaccines to poor population also. These strategies helped in prevention of spread of many contagious diseases. India, being the world's largest supplier of vaccines produces 62% of global demand. India's pharmaceutical industry started growing significantly since 2012. Earlier, India's pharmaceutical market share was less than desirable at $ 500 million which reached to $ 1.3 billion in 2019. This growth in share was due to the increase in vaccine development and outreach programmes.
India, the largest vaccine producer, currently exports two-thirds of its production while rest one-third is used domestically. Since growth in production has increased significantly, the Indian pharmaceutical industry is expected to reach 65 billion USD by 2024. The low cost production as well as the research and development in vaccines as well as pharmaceuticals contribute towards country's cost. The vaccines and pharmaceuticals produced in India are not only effective but 33% less costly than other markets. This cost effectiveness results in high export as well as its reach ability to the poor section of the country.
India's vaccine industry can grow upto $4 billion from $2 billion due to its growing vaccine as well as pharmaceutical industry as per the report by global consulting firm Kearney, in collaboration with the Confederation of Indian Industry (CII). India needs to accelerate its momentum in using sound biotech strategies in investments, research and developments as well as start-ups so as to reach its goal. This needs support from private, academia as well as government sector.
Ethical concern in COVID-19 vaccine development
The urgent demand for the production of effective vaccine against COVID-19 arises question on the efficacy and safety of the vaccine. While a lot more research is still needed to actually check the immunogenicity of the vaccine candidates many vaccines are approved for humans. The experts and scientists working in this field are giving their best to produce effective vaccine against COVID-19 as many vaccines are under trial and many approved. To meet the demands of vaccine there is a rush and this can compromise with the effectiveness of the vaccine and any loophole can result in loss of trust in vaccines.
To combat the risk of COVID-19 infection, the only way is acquiring herd immunity among population. Hence, to meet the demand it is required for everyone to get vaccinated, first the more vulnerable group followed by rest of the population. In India, major population has been vaccinated with all the doses of vaccine and administers should ensure for rest of the boosters doses to administered on time.
The emergence of different variants of SARS-CoV-2 is now a major concern for the effectiveness of the vaccine, as vaccines developed so far used earlier variants of SARS-CoV-2. The researchers and scientists still believe the earlier developed vaccines will still work on new variants may be with less efficacy. Several studies suggested BNT162b2 to be effective against new variants but with less efficacy. Also, the mRNA-1273 vaccine proved to be more effective for beta variant but with short span of immune response.
The major concern in vaccine development is the proper distribution of vaccines to every citizen of the country with proper trials. India, has tried to vaccinate every individual with requisite doses on priority basis, as health care workers and aged people were vaccinated first followed by the young people. Several volunteers worked to educate about the importance of vaccines and for equal distribution in all parts of India.
COVID-19 and autoimmune diseases
COVID-19 is linked with autoimmune disorders raising concern about the long term effect on human health. Hence, a good understanding of the complications of COVID-19 in autoimmune disorders patients is urgently required to guide researchers and doctors in treating patients with systemic lupus erythematous (SLE), infammatory bowel disease (IBD), systemic sclerosis (SSc), rheumatoid arthritis (RA) and others. The main reason behind the concern of COVID-19 in autoimmune disease patients is the treatment with immunosuppressive drugs, which can increase the susceptibility to COVID-19. Several studies on patients with autoimmune disorders showed the impact of COVID-19 on these patients.97., 98. Tan et al. 2021 reported that majority of autoimmune disorder patients died within 30 days of hospitalization with COVID-19. Autoantibodies were recovered in some patients with COVID-19 and some people developed autoimmune disorders after suffering from COVID-19.99 To reduce the COVID-19 infections, vaccination is the only possible treatment. Some studies reported the occurrence of autoimmune diseases like immune thrombotic thrombocytopenia, autoimmune liver diseases, Guillain-Barré syndrome, IgA nephropathy, rheumatoid arthritis and systemic lupus erythematosus in persons getting COVID-19 vaccine.100 As per WHO guidelines regarding COVID-19 vaccine, few people can get minor side effects including autoimmune disorders upon vaccination. These autoimmune disorders meet the criteria for diagnóstic of Adjuvant-Induced Autoimmune Syndrome (ASIA syndrome).101 The probable reason behind these autoimmune disorders is the molecular mimicry and hence formation of autoantibodies. There is still scanty information on whether the autoimmune disorders are caused by COVID-19 vaccines or mere a chance effect and studies need to be done in this regard. Geisen et al. reported the safety and efficacy of COVID-19 mRNA vaccines as safe with no considerable side effects.102 These studies open new avenues for searching a relationship between immune system, COVID-19, its vaccine and autoimmune disorders.
COVID-19 and cancer
COVID-19 becomes a serious concern for not autoimmune disease patients but more chronic for cancer patients. Cancer patients are at higher risk of complications due to COVID-19 and it increases more in aged people. In this pandemic era, cancer patients are more susceptible to infection as they are immunocompromised and also due to their treatment. The immunosuppression in cancer patients leads to serious complications as they are more prone to infections and may require early treatment if diagnosed with COVID-19. Cancer patients are nearly 3.5 fold more prone to COVID-19 infections with increased risk of hospitalization and ventilation.103 Cancer treatment during the COVID period is also a major concern as chances of infection increase and several events can occur like septic shock, myocardial infarction, respiratory distress syndrome and others. The risk of nosocomial infection also increases in cancer patients due to hospitalization. The development of COVID-19 vaccines and their randomized trials resulted in high level of safety and efficacy but the trial did not included immunocompromised individuals like cancer patients. As the cancer patients are more prone to infections, many countries prioritize vaccines for them. These patients show reduced cellular as well as humoral immune responses as treated with chemotherapy and immunosuppressive treatments.104 There are very less clinical studies on the impact of these vaccines on cancer patients. A report by National COVID Cohort Collaborative suggested that individuals with cancer and vaccinated are more prone to severe infection than individuals without cancer and vaccinated.105 The infections were more serious in individuals with hematological malignancy as well as those getting immunosupressive therapies. Earlier several vaccination studies in patients with cancer against diseases like influenza, hepatitis B and others proved these patients require more than one dose of vaccine, being immunocompromised. For instance, study on influenza vaccine showed two doses to be more effective in cancer patients rather than one showing more immunogenicity.106 Similar studies showed two doses to be effective against hepatitis B infection in cancer patients. These studies suggest increasing the number of COVID-19 vaccine doses in cancer patients as they are less immunocompetent. mRNA vaccines show more effective response than adenovirus vector based vaccines.107 Hence, regular interval of vaccination is an useful tool against COVID-19 in cancer patients.
Discussion
Sudden emergence of COVID-19 disease has become a serious concern of human health worldwide. Hence, development of antiviral therapeutics are exigent to combat COVID-19 infection. Major symptoms of SARS CoV-2 infection include severe respiratory distress, sore throat, dyspnea, fever, cough etc. Infection of this disease starts within 2 days and it may continue upto 14 days. Generally the transmission of COVID-19 spread through human to human contact or by inanimate matter that have been infected to SARS CoV-2 (based on the recommendation of centre for disease control and prevention).108., 109., 111.
Earlier studies have shown that hesitancy towards vaccine creates a major threat to the public health globally, as the resurgence of measles and pertussis.112., 113., 114., 115., 116. The crucial stage in development of vaccine is not only in developing the vaccine but its efficacy and safety. Previously, traditional ways of vaccine development was a labour intensive process because immunological correlation such as identification and investigation of immunogenic agent were needed.117 , 118 The vaccine development using modern techniques requires enormous testing regarding safety and efficacy of the vaccine. Several scientific community and agencies are working for the development of efficient coronavirus vaccine around the globe. Agencies like Moderna and the Vaccine Research Centre are together developing an mRNA-based vaccine candidate, where mRNA is coated with lipid vesicle for easy incorporation and also Codagenix is working in collaboration with the Serum Institute of India to develop live attenuated viral vaccine. The companies like Novavax, Sichuan Clover Biopharmaceuticals, iBio, and the University of Queensland are in process of developing S glycoprotein targeted vaccine. Several strategies like the viral vector-based vaccines, targeting the S glycoprotein are still under process to be developed into efficient vaccine for COVID-19.
Several vaccine candidates are now in Phase 3 clinical trials, including AstraZeneca/AZD1222, Oxford's Moderna's mRNA1273, and Sinovac's CoronaVac vaccines. With many expecting a COVID-19 vaccine to be available by the end of 2020 or early 2021, it is still too seen that how the vaccine distribution takes place how national interests will play out, or whether the vaccine will ultimately prove to be effective and safe when injected to the global population at large.117
Six biotech enterprises in India alone, including Serum Institute of India, Zydus Cadila, Biological E, Indian Immunologicals, Bharat Biotech, and Mynvax, are collaborating with several worldwide vaccine makers. They are developing DNA vaccines, live attenuated recombinant measles vaccines, inactivated viral vaccines, subunit vaccines, and vaccines created using codon optimization (Coronavirus, 2020). Furthermore, academic institutions such as the National Institute of Immunology (NII), the Indian Institute of Science (IISc), the International Centre for Genetic Engineering and Biotechnology (ICGEB) New Delhi, the Translational Health Science and Technology Institute (THSTI), and others are working to develop vaccines, therapies, and SARS-CoV-2 animal models in order to halt the pandemic as soon as possible.119 A detailed map showing the vaccination rate, number of approved and in trial vaccines from India is shown in Fig. 12.
The current research focus has been on the design and manufacture of epitope-based peptide vaccines using various immunoinformatics prediction approaches. The traditional method of producing an effective vaccine needs extensive research, identification, and creation of an immunological link with SARS-CoV-2. In general, the development of peptide vaccines necessitates the identification of immunodominant B-cell and T-cell epitopes capable of producing specific immunological responses. Furthermore, for a peptide vaccine to be highly immunogenic, a target molecule's B-cell epitope must be paired with a T-cell epitope. T-cell epitopes are typically made up of 8–20 amino acids (small peptide fragments) and have been found to be more desirable, resulting in a long-lasting immune response mediated by CD8 + T-cells,120., 121., 122., 123. whereas B-cell epitopes are made up of a lineal chain of amino acids that can be a protein.122., 123., 124., 125. Several articles are published each year on vaccines against different viruses as shown in Fig. 13.
Finally, a vaccine needs to be properly tested and without knowing the effectiveness of the vaccine candidate and associated risk of its side effect, it can be dangerous. The durability and efficacy of a vaccine should be first target in developing a vaccine against COVID-19.
Current challenges and future prospect
In this current pandemic situation several agencies accelerated their vaccine development program in 6–12 months which usually takes 10–15 years. This reduces the trial period for the vaccine as phase trials are done on smaller groups, which itself is a great challenge in vaccine development. The occurrence of side effects due to less trial phase especially in people with co-morbidity is a major challenge but the vaccines developed in India till date did not produce any side effects. The major challenges posed by some of the COVID-19 vaccines is shown in Table 11 .Table 11 Some challenges faced by COVID-19 vaccines.
Table 11Vaccine Challenge
Covaxin Efficacy around 78% (1)
ZyCoV- D Efficacy around 67% (aged 12 years or older) (2)
Sputnik V 91.6%
CoronaVaC Efficacy around 83.5% (Trial III in Turkey) (3)
BBIBP-CorV Less efficiency of only 50–70%
NVX-CoV2373 (Novavax) Efficacy around 90.4% (aged 18–84 years) (4)
mRNA − 1273 (Moderna) Low antibody titer against Beta variant, 94.1% (adult 18 years or older) (5)
BNT162b2 Efficacy around 95% (aged 16 years to older) (6,7)
Ad26.COV2.S (Janssen) Efficacy around 66.9% (aged 18 years or older) (8)
ChAdOx1 nCoV-19/AZD1222 70–76% (18 year or older) (9,10)
In this current situation, the occurrence of different variants of SARS-CoV-2 like alpha, delta and omicron with high infectivity rate is a matter of concern. The vaccines developed till date were against the earlier variants of SARS-CoV-2, but still researchers and scientists across the globe suggest the vaccines to be equally effective against the new variants too. The emergence of new variants over time will always be a matter of concern on the effectiveness of vaccines.
There is also a concern of discrepancy in distribution of vaccines equally to developed and developing nations. However, several organizations including WHO is trying to lower this discrepancy still it is matter of concern. The developed countries preordered the vaccine doses but developing nations could not order and hence till date developed nations are completely vaccinated whereas developing nations still need more vaccines. So, this discrepancy in vaccination can lead to further spread of this virus.
The proper vaccination of children is also a major challenge as majority of the vaccines have been tried on adults itself. But according to many organizations including WHO, vaccination of children is also important to stop the spread of this pandemic. Many vaccines are under trial for children and in future can be developed to vaccinate the children and hence break the chain of spread of this virus. The ongoing large scale vaccination programme produces huge amount of biomedical and plastic wastes worldwide and, therefore, triggering adverse effect on the environment. The main objective of the mass vaccination campaign is to exit the emergency situation arising as a result of COVID-19 pandemic. However, the vaccination approach created an intense knock-on-effects on the environment. The mass vaccination waste of used discarded vials has thimerosal-mercury based preservative which is found to be hazardous to aquatic ecosystem as well as humans when released improperly in the water bodies. The reckless use of PPE as a preventive measure to COVID-19 has significantly added to the microplastic fibers in the environment.126 , 127 On the other hand, positive impact of COVID- 19 induced lockdown has also been reported in the recent studies that have shown that significant reduction in many pollutants around the world during the period of lockdown. The lockdown exhibited the better solutions for the nature's rejuvenation by means of natural resource preservation which is prerequisite for sustainable development.128 , 129
Although COVID-19 vaccines have reduced the chances of infection and hence mortality still certain problems needs to be resolved in the future. The development of mix and match vaccine and the development of new vaccines such as nanoparticle vaccines, adding different adjuvants to improve vaccine and changing the vaccine administration route are some points to be resolved in future. The safety of vaccine for infants, pregnant women and immunocompromised individuals is a matter of future concern to curb this disease completely. The variant of concerns are always emerging and hence a major concern in the future. These mutations are common and problematic and needs to be considered prior to any vaccine development. So, it is essential to consider these mutations while administering vaccines as several vaccines have shown fewer efficacies against certain SARS-CoV-2 variants. Although the COVID-19 vaccines have proved efficient in both animal studies and clinical trials, and several vaccines have been authorized for emergency use by the WHO, adverse effects including pain at the injection site and fever, and some major complications such as coagulation dysfunction, myocarditis, immune disorders, nervous disorder and lymphatic system diseases caused by COVID-19 vaccine, calls for concerns about vaccine safety. The future prospect to be followed and monitored is shown in Fig. 11 . The rate of these complications is low as it occurs mainly in individuals with underlying diseases. So to reduce these risks healthcare workers should ensure the lives of the patients by the timely treatment of complications. Fig. 11 Futuristic approach in development of effective vaccine after approval.
Fig. 11
Fig. 12 Map showing percentage of people fully vaccinated, number of authorized/approved vaccines and number of vaccines under trial in India. Map was retrieved from https://covid19.trackvaccines.org/trials-vaccines-by-country/.
Fig. 12
Fig. 13 The figure represents the publication trend of COVID-19 as well as vaccine articles in the year 2019, 2020, 2021 and 2022. The data was obtained from NCBI and PubMed using SARS-CoV-2, COVID-19, Vaccine, Coronavirus, SARS-CoV-2 vaccine and COVID-19 vaccine as keywords.
Fig. 13
Conclusion
COVID-19, a transmissible disease has become a major threat to mankind worldwide. The world witnessed a tremendous collapse of economy, burdened healthcare workers, rising unemployment and a major health catastrophe. The development of COVID-19 vaccine is a major challenge and only definitive solution to curb the viral infections. The catastrophic impact of COVID-19 catalyzed the vaccine development around the globe and the exceptional advancement in vaccine development has brought the hope throughout the world that the battle against this deadly virus can be won in the near future. India's COVID-19 vaccination drive started on January 16, 2021 and today India has become the fastest country in the vaccination program with global mass COVID-19 vaccination campaigns. It is also important to monitor the long-term efficiency and safety of the authorized vaccines as well as analyze different strategies of combining vaccines that might elicit levels of neutralizing antibodies.
Moreover, not only vaccine candidate rather the mechanism of virus infection to the host needs to be explored to completely curb this infection from the world. A trans-multi-disciplinary workforce is required to reduce, understand and hence stop this viral infection. As all across the globe first line of vaccine designing has been done and many are in progress, it is necessary to reduce the rate of virus infection and to stop the reoccurrence of this infection. It is also necessary to understand the immunopathology and other important aspects of this virus like its host specific impact and vulnerability of aged people.
Meanwhile, many vaccine candidates are being tried and several vaccines are under trial to reduce the infection of coronavirus. This review sheds light on the different vaccine strategies being used nowadays. Also, this review tried to compare the traditional and the modern vaccine strategies. Vaccine design is an important aspect and antigen candidate needs to be explored along with vaccine delivery system before massive production of the vaccine. To provide proper immunization, on one side the vaccine developed so far needs to be produced massively and different grades of vaccine needs to be tried for their immunogenicity on various populations.
Conflict of interest
The authors declare that they have no conflict of interest.
Acknowledgements
None.
==== Refs
References
1. Hu B. Guo H. Zhou P. Shi Z.L. Characteristics of SARS-CoV-2 and COVID-19 [published correction appears in Nat Rev Microbiol. 2022 Feb 23] Nat Rev Microbiol 19 3 2021 141 154 10.1038/s41579-020-00459-7 33024307
2. Shereen M.A. Khan S. Kazmi A. Bashir N. Siddique R. COVID-19 infection: Origin, transmission, and characteristics of human coronaviruses J Adv Res 24 2020 91 98 Published 2020 Mar 16 10.1016/j.jare.2020.03.005 32257431
3. de Haan C.A. Kuo L. Masters P.S. Vennema H. Rottier P.J. Coronavirus particle assembly: primary structure requirements of the membrane protein J Virol 72 8 1998 6838 6850 10.1128/JVI.72.8.6838-6850.1998 9658133
4. Cascella M. Rajnik M. Aleem A. Dulebohn S.C. Di Napoli R. Features, Evaluation, and Treatment of Coronavirus (COVID-19). 2021 Jul 30. StatPearls [Internet] 2021 StatPearls Publishing Treasure Island (FL)
5. Sharma O. Sultan A.A. Ding H. Triggle C.R. A review of the progress and challenges of developing a vaccine for COVID-19 Front Immunol 11 2020 585354 . Published 2020 Oct 14 10.3389/fimmu.2020.585354
6. Achaiah N.C. Subbarajasetty S.B. Shetty R.M. R0 and Re of COVID-19: can we predict when the pandemic outbreak will be contained? Indian J Crit Care Med 24 11 2020 1125 1127 10.5005/jp-journals-10071-23649 33384521
7. van Boheemen S. de Graaf M. Lauber C. Genomic characterization of a newly discovered coronavirus associated with acute respiratory distress syndrome in humans mBio. 3 6 2012 e00473-12. Published 2012 Nov 20: 10.1128/mBio.00473-12
8. Czub M. Weingartl H. Czub S. He R. Cao J. Evaluation of modified vaccinia virus Ankara based recombinant SARS vaccine in ferrets Vaccine. 23 17–18 2005 2273 2279 10.1016/j.vaccine.2005.01.033 15755610
9. Du L. He Y. Zhou Y. Liu S. Zheng B.J. Jiang S. The spike protein of SARS-CoV--a target for vaccine and therapeutic development Nat Rev Microbiol 7 3 2009 226 236 10.1038/nrmicro2090 19198616
10. Song W. Gui M. Wang X. Xiang Y. Cryo-EM structure of the SARS coronavirus spike glycoprotein in complex with its host cell receptor ACE2 PLoS Pathog 14 8 2018 e1007236. Published 2018 Aug 13 10.1371/journal.ppat.1007236
11. Gralinski L.E. Menachery V.D. Return of the Coronavirus: 2019-nCoV Viruses 12 2 2020 135 Published 2020 Jan 24 10.3390/v12020135 31991541
12. Dai L. Gao G.F. Viral targets for vaccines against COVID-19 Nat Rev Immunol 21 2 2021 73 82 10.1038/s41577-020-00480-0 33340022
13. Yoshimoto F.K. The Proteins of Severe Acute Respiratory Syndrome Coronavirus-2 (SARS CoV-2 or n-COV19), the Cause of COVID-19 Protein J 39 3 2020 198 216 10.1007/s10930-020-09901-4 32447571
14. Park S.E. Epidemiology, virology, and clinical features of severe acute respiratory syndrome -coronavirus-2 (SARS-CoV-2; Coronavirus Disease-19) Clin Exp Pediatr 63 4 2020 119 124 10.3345/cep.2020.00493 32252141
15. Xu Z. Shi L. Wang Y. Pathological findings of COVID-19 associated with acute respiratory distress syndrome [published correction appears in Lancet Respir Med. 2020 Feb 25] Lancet Respir Med 8 4 2020 420 422 10.1016/S2213-2600(20)30076-X 32085846
16. Kaur S.P. Gupta V. COVID-19 Vaccine: a comprehensive status report Virus Res 288 2020 198114 10.1016/j.virusres.2020.198114
17. Li W. Moore M.J. Vasilieva N. Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus Nature. 426 6965 2003 450 454 10.1038/nature02145 14647384
18. Glowacka I. Bertram S. Müller M.A. Evidence that TMPRSS2 activates the severe acute respiratory syndrome coronavirus spike protein for membrane fusion and reduces viral control by the humoral immune response J Virol 85 9 2011 4122 4134 10.1128/JVI.02232-10 21325420
19. Matsuyama S. Nagata N. Shirato K. Kawase M. Takeda M. Taguchi F. Efficient activation of the severe acute respiratory syndrome coronavirus spike protein by the transmembrane protease TMPRSS2 J Virol 84 24 2010 12658 12664 10.1128/JVI.01542-10 20926566
20. Shulla A. Heald-Sargent T. Subramanya G. Zhao J. Perlman S. Gallagher T. A transmembrane serine protease is linked to the severe acute respiratory syndrome coronavirus receptor and activates virus entry J Virol 85 2 2011 873 882 10.1128/JVI.02062-10 21068237
21. Hoffmann M. Kleine-Weber H. Schroeder S. SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor Cell. 181 2 2020 271 280.e8 10.1016/j.cell.2020.02.052 32142651
22. Sungnak W. Huang N. Bécavin C. SARS-CoV-2 entry factors are highly expressed in nasal epithelial cells together with innate immune genes Nat Med 26 5 2020 681 687 10.1038/s41591-020-0868-6 32327758
23. Zou L. Ruan F. Huang M. SARS-CoV-2 viral load in upper respiratory specimens of infected patients N Engl J Med 382 12 2020 1177 1179 10.1056/NEJMc2001737 32074444
24. He X. Lau E.H.Y. Wu P. Temporal dynamics in viral shedding and transmissibility of COVID-19 [published correction appears in Nat Med. 2020 Sep;26(9):1491-1493] Nat Med 26 5 2020 672 675 10.1038/s41591-020-0869-5 32296168
25. Parasher A. COVID-19: current understanding of its pathophysiology, clinical presentation and treatment Postgrad Med J 97 1147 2021 312 320 10.1136/postgradmedj-2020-138577 32978337
26. Arora N.K. Das M.K. COVID-19 vaccine development and the way forward Indian J Public Health 64 Supplement 2020 S108 S111 10.4103/ijph.IJPH_520_20 32496238
27. Martín Sánchez F.J. Martínez-Sellés M. Molero García J.M. Insights for COVID-19 in 2023 Rev Esp Quimioter 36 2 2023 114 124 10.37201/req/122.2022 36510683
28. Pirouz B. Shaffiee Haghshenas S. Pirouz B. Shaffiee Haghshenas S. Piro P. Development of an assessment method for investigating the impact of climate and urban parameters in confirmed cases of COVID-19: a new challenge in sustainable development Int J Environ Res Public Health 17 8 2020 2801 Published 2020 Apr 18 10.3390/ijerph17082801 32325763
29. Qi H. Xiao S. Shi R. COVID-19 transmission in Mainland China is associated with temperature and humidity: a time-series analysis Sci Total Environ 728 2020 138778 10.1016/j.scitotenv.2020.138778
30. Şahin M. Impact of weather on COVID-19 pandemic in Turkey Sci Total Environ 728 2020 138810 10.1016/j.scitotenv.2020.138810
31. Zhu Y. Xie J. Huang F. Cao L. Association between short-term exposure to air pollution and COVID-19 infection: Evidence from China Sci Total Environ 727 2020 138704 10.1016/j.scitotenv.2020.138704
32. Briz-Redón Á. Serrano-Aroca Á. A spatio-temporal analysis for exploring the effect of temperature on COVID-19 early evolution in Spain Sci Total Environ 728 2020 138811 10.1016/j.scitotenv.2020.138811
33. Shi P. Dong Y. Yan H. Impact of temperature on the dynamics of the COVID-19 outbreak in China Sci Total Environ 728 2020 138890 10.1016/j.scitotenv.2020.138890
34. Xie J. Zhu Y. Association between ambient temperature and COVID-19 infection in 122 cities from China Sci Total Environ 724 2020 138201 10.1016/j.scitotenv.2020.138201
35. Abdullah S. Mansor A.A. Napi N.N.L.M. Air quality status during 2020 Malaysia Movement Control Order (MCO) due to 2019 novel coronavirus (2019-nCoV) pandemic Sci Total Environ 729 2020 139022 10.1016/j.scitotenv.2020.139022
36. Carrington D. Air pollution linked to far higher Covid-19 death rates, study finds Available https://www.theguardian.com/environment/2020/apr/07/air-pollutionlinked-to-far-higher-covid-19-death-rates-study-finds 2020 Accessed 6 July 2020
37. Muhammad S. Long X. Salman M. COVID-19 pandemic and environmental pollution: A blessing in disguise? Sci Total Environ 728 2020 138820 10.1016/j.scitotenv.2020.138820
38. Lau J.J. Cheng S.M.S. Leung K. Real-world COVID-19 vaccine effectiveness against the Omicron BA.2 variant in a SARS-CoV-2 infection-naive population Nat Med 29 2 2023 348 357 10.1038/s41591-023-02219-5 36652990
39. Ma Y. Zhao Y. Liu J. Effects of temperature variation and humidity on the death of COVID-19 in Wuhan, China Sci Total Environ 724 2020 138226 10.1016/j.scitotenv.2020.138226
40. Kumar V.M. Pandi-Perumal S.R. Trakht I. Thyagarajan S.P. Strategy for COVID-19 vaccination in India: the country with the second highest population and number of cases NPJ Vaccines 6 1 2021 60 Published 2021 Apr 21 10.1038/s41541-021-00327-2 33883557
41. Pollard A.J. Bijker E.M. A guide to vaccinology: from basic principles to new developments [published correction appears in Nat Rev Immunol. 2021 Jan 5] Nat Rev Immunol 21 2 2021 83 100 10.1038/s41577-020-00479-7 33353987
42. Yashvardhini N. Samiksha Jha D.K. Pharmacological intervention of various Indian Medicinal plants in combating COVID-19 infection Biomed Res Ther 8 2021 4461 4475
43. Kashte S. Gulbake A. El-Amin Iii S.F. Gupta A. COVID-19 vaccines: rapid development, implications, challenges and future prospects Hum Cell 34 3 2021 711 733 10.1007/s13577-021-00512-4 33677814
44. Khuroo M.S. Khuroo M. Khuroo M.S. Sofi A.A. Khuroo N.S. COVID-19 vaccines: a race against time in the middle of death and devastation! J Clin Exp Hepatol 10 6 2020 610 621 10.1016/j.jceh.2020.06.003 32837093
45. Dai X. Xiong Y. Li N. Jian C. Vaccine Types 2019 Intechopen
46. Jha D.K. Yashvardhini N. Kumar A. Immunological and mutational analysis of SARS-CoV-2 structural proteins from Asian countries Biomed Res Ther 8 2021 4367 4381
47. Singh K. Mehta S. The clinical development process for a novel preventive vaccine: An overview J Postgrad Med 62 1 2016 4 11 10.4103/0022-3859.173187 26732191
48. Hudgens M.G. Gilbert P.B. Self S.G. Endpoints in vaccine trials Stat Methods Med Res 13 2 2004 89 114 10.1191/0962280204sm356ra 15068256
49. Clem A.S. Fundamentals of vaccine immunology J Global Infect Dis 3 1 2011 73 78 10.4103/0974-777X.77299
50. Thomas F. Magill T.P. Vaccination of human subjects with virus of human influenza Exp. Biol. Medicine 33 4 1936 10.3181/00379727-33-8467P
51. Nooraei S. Bahrulolum H. Hoseini Z.S. Virus-like particles: preparation, immunogenicity and their roles as nanovaccines and drug nanocarriers J Nanobiotechnol 19 1 2021 59 Published 2021 Feb 25 10.1186/s12951-021-00806-7
52. Joe C.C.D. Chatterjee S. Lovrecz G. Glycoengineered hepatitis B virus-like particles with enhanced immunogenicity Vaccine. 38 22 2020 3892 3901 10.1016/j.vaccine.2020.03.007 32284273
53. Zhai L. Yadav R. Kunda N.K. Oral immunization with bacteriophage MS2-L2 VLPs protects against oral and genital infection with multiple HPV types associated with head & neck cancers and cervical cancer Antivir Res 166 2019 56 65 10.1016/j.antiviral.2019.03.012 30926288
54. Shiri F. Petersen K.E. Romanov V. Zou Q. Gale B.K. Characterization and differential retention of Q beta bacteriophage virus-like particles using cyclical electrical field-flow fractionation and asymmetrical flow field-flow fractionation Anal Bioanal Chem 412 7 2020 1563 1572 10.1007/s00216-019-02383-z 31938845
55. Silveira M.M. Moreira G.M.S.G. Mendonça M. DNA vaccines against COVID-19: Perspectives and challenges Life Sci 267 2021 118919 10.1016/j.lfs.2020.118919
56. van Riel D. de Wit E. Next-generation vaccine platforms for COVID-19 Nat Mater 19 8 2020 810 812 10.1038/s41563-020-0746-0 32704139
57. Henao-Restrepo A.M. Camacho A. Longini I.M. Efficacy and effectiveness of an rVSV-vectored vaccine in preventing Ebola virus disease: final results from the Guinea ring vaccination, open-label, cluster-randomised trial (Ebola Ça Suffit!) [published correction appears in Lancet. 2017 Feb 4;389(10068):504] [published correction appears in Lancet. 2017 Feb 4;389(10068):504] Lancet. 389 10068 2017 505 518 10.1016/S0140-6736(16)32621-6 28017403
58. (Vaccine Types) Dai Xiaoxia Xiong Yongmin Li Na Jian Can Submitted: September 20th, 2018 Reviewed: January 22nd, 2019 Published: March 5th, 2019 2019 10.5772/intechopen.84626
59. Zhou L. Srisatjaluk R. Justus D. Doyle R. On the origin of membrane vesicles in Gram-negative bacteria FEMS Microbiol Lett 163 1998 223 228 10.1111/j.1574-6968.1998.tb13049.x PMID: 9673026 9673026
60. Haurat M.F. Aduse-Opoku J. Rangarajan M. Dorobantu L. Gray M.R. Curtis M.A. Selective sorting of cargo proteins into bacterial membrane vesicles J Biol Chem 286 2011 1269 1276 10.1074/jbc. M110.185744 21056982
61. McBroom A.J. Johnson A.P. Vemulapalli S. Kuehn M.J. Outer membrane vesicle production by Escherichia coli is independent of membrane instability J Bacteriol 188 2006 5385 5392 10.1128/JB.00498-06 16855227
62. Li M. Zhou H. Yang C. Wu Y. Zhou X. Liu H. Wang Y. Bacterial outer membrane vesicles as a platform for biomedical applications: an update J Control Release 323 2020 253 268 32333919
63. O'Ryan M. Stoddard J. Toneatto D. Wassil J. Dull P.M. A multi-component meningococcal serogroup B vaccine (4CMenB): the clinical development program Drugs 74 1 2014 15 30 24338083
64. Dhama K. Sharun K. Tiwari R. COVID-19, an emerging coronavirus infection: advances and prospects in designing and developing vaccines, immunotherapeutics, and therapeutics Hum Vaccin Immunother 16 6 2020 1232 1238 10.1080/21645515.2020.1735227 32186952
65. Pallesen J. Wang N. Corbett K.S. Immunogenicity and structures of a rationally designed prefusion MERS-CoV spike antigen Proc Natl Acad Sci U S A 114 35 2017 E7348 E7357 10.1073/pnas.1707304114 28807998
66. Li Y. Tenchov R. Smoot J. Liu C. Watkins S. Zhou Q. A comprehensive review of the global efforts on COVID-19 vaccine development ACS Cent Sci 7 4 2021 512 533 10.1021/acscentsci.1c00120 34056083
67. Forni G. Mantovani A. COVID-19 Commission of Accademia Nazionale dei Lincei, Rome. COVID-19 vaccines: where we stand and challenges ahead Cell Death Differ 28 2 2021 626 639 10.1038/s41418-020-00720-9 33479399
68. Ball P. The lightning-fast quest for COVID vaccines - and what it means for other diseases Nature. 589 7840 2021 16 18 10.1038/d41586-020-03626-1 33340018
69. Dhama K. Sharun K. Tiwari R. COVID-19, an emerging coronavirus infection: advances and prospects in designing and developing vaccines, immunotherapeutics, and therapeutics Hum Vaccin Immunother 16 6 2020 1232 1238 10.1080/21645515.2020.1735227 32186952
70. Pallesen J. Wang N. Corbett K.S. Immunogenicity and structures of a rationally designed prefusion MERS-CoV spike antigen Proc Natl Acad Sci U S A 114 35 2017 E7348 E7357 10.1073/pnas.1707304114 28807998
71. HKU https://www.genengnews.com/COVID-19-candidates/the-university-of-hong-kong-hku/ 2020
72. Gao Q. Bao L. Mao H. Development of an inactivated vaccine candidate for SARS-CoV-2 Science. 369 6499 2020 77 81 10.1126/science.abc1932 32376603
73. Xia S. Zhang Y. Wang Y. Safety and immunogenicity of an inactivated SARS-CoV-2 vaccine, BBIBP-CorV: a randomised, double-blind, placebo-controlled, phase 1/2 trial Lancet Infect Dis 21 1 2021 39 51 10.1016/S1473-3099(20)30831-8 33069281
74. Corum J. Zimmer C. How the Sinopharm Vaccine Works 2021 The New York Times https://www.nytimes.com/interactive/2020/health/sinopharm-COVID-19-vaccine.html
75. Leo L. Novavax COVID-19 vaccine shows 90% efficacy in US phase 3 trial https://www-livemintcom.cdn.ampproject.org/v/s/www.livemint.com/news/india/novavax-COVID-19-vaccine shows-90-efficacy-in-us-phase-3-trial/amp11623673514966.html?amp_gsa=1&_js_v=a6&usqp=mq331AQKKAFQArABIIACAw%3D%3D#amp_tf=From%20%251%24s&aoh=16319380869248&referrer=https%3A%2F%2Fwww.google.com&share=https%3A%2F%2Fwww.livemint.com%2Fnews%2Findia%2Fnovavax-COVID-19-vaccine-shows-90-efficacy-in-us-phase-3-trial-11623673514966.html 2021
76. CDC https://www.cdc.gov/coronavirus/2019-ncov/vaccines/different-vaccines/Moderna.html 2021
77. Anon Moderna Announces Positive Interim Phase 1 Data for its mRNA Vaccine (mRNA-1273) Against Novel Coronavirus https://investors.modernatx.com/news-releases/news-release-details/moderna-announces-positive-interim-phase-1-data-its-mrna-vaccine 2020
78. Polack F.P. Thomas S.J. Kitchin N. Safety and efficacy of the BNT162b2 mRNA COVID-19 vaccine N Engl J Med 383 27 2020 2603 2615 10.1056/NEJMoa2034577 33301246
79. Zhu F. CanSino Biologics' Ad5-nCoV the First COVID-19 Vaccine to Phase II Clinical Trials 2020 TrialSiteNews. Jiangsu Province Centers For Disease Control and Prevention Available from https://trialsitenews.com/cansino-biologics-ad5-ncov-the-first-COVID-19-vaccine-to-phase-ii-clinical-trials/
80. Thiagarajan K. COVID-19: India is at centre of global vaccine manufacturing, but opacity threatens public trust BMJ 372 2021 n196. Published 2021 Jan 28: 10.1136/bmj.n196
81. Ministry of Health and Family Welfare https://mohfw.gov.in/ 2021
82. Thiagarajan K. What do we know about India's Covaxin vaccine? BMJ. 373 2021 n997 . Published 2021 Apr 20 10.1136/bmj.n997
83. Bharat Biotech https://www.bharatbiotech.com/covaxin.html 2021
84. Ella R. Vadrevu K.M. Jogdand H. Safety and immunogenicity of an inactivated SARS-CoV-2 vaccine, BBV152: a double-blind, randomised, phase 1 trial [published correction appears in Lancet Infect Dis. 2021 Apr;21(4):e81] Lancet Infect Dis 21 5 2021 637 646 10.1016/S1473-3099(20)30942-7 33485468
85. Yan Y. Pang Y. Lyu Z. The COVID-19 vaccines: recent development, challenges and prospects Vaccines (Basel) 9 4 2021 349 Published 2021 Apr 5 10.3390/vaccines9040349 33916489
86. Sapkal G. Yadav P.D. Ella R. Neutralization of VUI B.1.1.28 P2 variant with sera of COVID-19 recovered cases and recipients of Covaxin an inactivated COVID-19 vaccine J Travel Med 28 7 2021 10.1093/jtm/taab077 taab077
87. K.M. Vadrevu et al., Evaluation of Safety and Immunogenicity of an Adjuvanted, TH-1 Skewed, Whole Virion InactivatedSARS-CoV-2 Vaccine - BBV152. bioRxiv 2020.S.
88. Voysey M. Clemens S.A.C. Madhi S.A. Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK [published correction appears in Lancet. 2021 Jan 9;397(10269):98] Lancet. 397 10269 2021 99 111 10.1016/S0140-6736(20)32661-1 33306989
89. Thakur V. Ratho R.K. OMICRON (B.1.1.529): A new SARS-CoV-2 variant of concern mounting worldwide fear J Med Virol 94 5 2022 May 1821 1824 Epub 2021 Dec 30: 10.1002/jmv.27541 34936120 34936120
90. Omicron variant: what you need to know 2021 CDC Accessed December 15, 2021 https://www.cdc.gov/coronavirus/2019-ncov/variants/omicron-variant.html
91. Science Brief: Omicron (B.1.1.529) variant.2021 CDC. Accessed December 4, 2021. https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/scientific-brief-omicron-variant.htm.
92. Chen J. Wang R. Gilby N.B. Wei G.W. Omicron variant (B.1.1.529): infectivity, vaccine breakthrough, and antibody resistance J Chem Inf Model 62 2 2022 Jan 24 412 422 10.1021/acs.jcim.1c01451 34989238
93. Pulliam J.R.C. van Schalkwyk C. Govender N. Increased risk of SARS-CoV-2 reinfection associated with emergence of the Omicron variant in South Africa medRxiv. December 2021 10.1101/2021.11.11.21266068
94. COVID-19 reinfection study from South Africa yields ominous data about Omicron 2021 Science AAAS Accessed December 4, 2021 https://www.science.org/content/article/covid-19-reinfection-study-southafrica-yields-ominous-data-about-omicron
95. Update on Omicron Accessed November 29, 2021 https://www.who.int/news/item/28-11-2021-update-on-omicron
96. Covid news: Pfizer-BioNTech, Moderna, J&J, AstraZeneca investigating omicron Accessed November 29, 2021 https://www.cnbc.com/2021/11/26/pfizer-biontech-investigating-new-covidvariant-jj-testing-vaccine-against-it.html
97. Akber Abbasi S. Khalil A.B. Arslan M. Extensive use of face masks during COVID-19 pandemic: (micro-)plastic pollution and potential health concerns in the Arabian Peninsula Saudi J Biol Sci 27 12 2020 Dec 3181 3186 10.1016/j.sjbs.2020.09.054 33052188
98. Tan E.H. Sena A.G. Prats-Uribe A. COVID-19 in patients with autoimmune diseases: characteristics and outcomes in a multinational network of cohorts across three countries Rheumatology (Oxford) 60 SI 2021 SI37 SI50 10.1093/rheumatology/keab250 33725121
99. Liu Y. Sawalha A.H. Lu Q. COVID-19 and autoimmune diseases Curr Opin Rheumatol 33 2 2021 155 162 10.1097/BOR.0000000000000776 33332890
100. Chen Y. Xu Z. Wang P. New-onset autoimmune phenomena post-COVID-19 vaccination Immunology. 165 4 2022 386 401 10.1111/imm.13443 34957554
101. Gomez P.L. Robinson J.M. Vaccine manufacturing Plotkin's Vaccines 51-60 2018 e1 10.1016/B978-0-323-35761-6.00005-5
102. Geisen U.M. Berner D.K. Tran F. Immunogenicity and safety of anti-SARS-CoV-2 mRNA vaccines in patients with chronic inflammatory conditions and immunosuppressive therapy in a monocentric cohort Ann Rheum Dis 80 10 2021 1306 1311 10.1136/annrheumdis-2021-220272 33762264
103. Liang W. Guan W. Chen R. Cancer patients in SARS-CoV-2 infection: a nationwide analysis in China Lancet Oncol 21 3 2020 335 337 10.1016/S1470-2045(20)30096-6 32066541
104. Fendler A. de Vries E.G.E. GeurtsvanKessel C.H. COVID-19 vaccines in patients with cancer: immunogenicity, efficacy and safety Nat Rev Clin Oncol 19 6 2022 385 401 10.1038/s41571-022-00610-8 35277694
105. Song Q. Bates B. Shao Y.R. Risk and outcome of breakthrough COVID-19 infections in vaccinated patients with cancer: real-world evidence from the national COVID cohort collaborative J Clin Oncol 40 13 2022 1414 1427 10.1200/JCO.21.02419 35286152
106. Rousseau B. Loulergue P. Mir O. Immunogenicity and safety of the influenza A H1N1v 2009 vaccine in cancer patients treated with cytotoxic chemotherapy and/or targeted therapy: the VACANCE study Ann Oncol 23 2 2012 450 457 10.1093/annonc/mdr141 21576285
107. Thakkar A. Gonzalez-Lugo J.D. Goradia N. Seroconversion rates following COVID-19 vaccination among patients with cancer Cancer Cell 39 8 2021 1081 1090.e2 10.1016/j.ccell.2021.06.002 34133951
108. Momin T. Kansagra K. Patel H. Safety and Immunogenicity of a DNA SARS-CoV-2 vaccine (ZyCoV-D): Results of an open-label, non-randomized phase I part of phase I/II clinical study by intradermal route in healthy subjects in India EClinicalMedicine. 38 2021 101020 10.1016/j.eclinm.2021.101020
109. Jones I. Roy P. Sputnik V COVID-19 vaccine candidate appears safe and effective Lancet. 397 10275 2021 642 643 10.1016/S0140-6736(21)00191-4 33545098
111. Yashvardhini N. Kumar A. Jha D.K. Analysis of SARS-CoV-2 mutations in the main viral protease (NSP5) and its implications on the vaccine designing strategies Vacunas. 23 2022 S1 S13 10.1016/j.vacun.2021.10.002
112. Chan J.F.-W. Yuan S. Kok K.-H. A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster 7e Lancet 395 10223 2020 514 523
113. Chen N. Zhou M. Dong X. Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study 7e Lancet 395 10223 2020 507 513
114. Li Q. Guan X. Wu P. Early transmission dynamics in Wuhan, China, of novel coronavirusinfected pneumonia 7e New England J Med 382 13 2020 1199 1207
115. Phadke V.K. Bednarczyk R.A. Salmon D.A. Omer S.B. Association between vaccine refusal and vaccine-preventable diseases in the United States: a review of measles and pertussis JAMA 315 2016 1149 1158 26978210
116. Benecke O. DeYoung S.E. Anti-vaccine decision-making and measles resurgence in the United States Glob Pediatr Health 2019 6
117. Gangarosa E.J. Galazka A.M. Wolfe C.R. Phillips L.M. Gangarosa R.E. Miller E. Chen R.T. Impact of anti-vaccine movements on pertussis control: the untold story Lancet 351 1998 356 361 9652634
118. Yashvardhini N. Jha D.K. Kumar A. Genetic variations in the Orf7a protein of SARS-CoV-2 and its possible role in vaccine development Biomed Res Ther 8 8 2021 4497 4504
119. Borba R.C. Vidal V.M. Moreira L.O. The re-emergency and persistence of vaccine preventable diseases An Acad Bras Cienc 87 2015 1311 1322 26312431
120. Yashvardhini N. Jha D.K. Kumar A. Genome sequence analysis of SARS-CoV-2 nsp15 Bioinformation. 18 4 2022 432 437 36909703
121. Yashvardhini N. Jha D.K. Bhattacharya S. Identification and characterization of mutations in the SARS-CoV-2 RNA-dependent RNA polymerase as a promising antiviral therapeutic target Arch Microbiol 203 9 2021 5463 5473 10.1007/s00203-021-02527-9 34410443
122. Jha D.K. Yashvardhini N. Kumar A. Identification of recurrent mutations in exonuclease (nsp14); a potential drug target in SARS-CoV-2 Indian J Pathol Microbiol 64 4 2021 771 775 10.4103/0377-4929.328516 34673601
123. Sharma O. Sultan A.A. Ding H. Triggle C.R. A review of the progress and challenges of developing a vaccine for COVID-19 Front Immunol 11 2020 585354 10.3389/fimmu.2020.585354
124. Jayashree Nandi Top Indian Scientists Join Global Fight Against Coronavirus https://www.hindustantimes.com/india-news/top-indian-scientists-join-global-fight-against-virus/story CKRGfycsjBJD2ypLcbJPHM.html 2020 https://www.hindustantimes.com [Online] Hindustan Times, New Delhi, March 30, 2020. [Cited: June 9, 2020.]
125. Yashvardhini N. Kumar A. Jha D.K. Immunoinformatics identification of B- and T-cell epitopes in the RNA-dependent RNA polymerase of SARS-CoV-2 Can J Infect Dis Med Microbiol 2021 2021 6627141 10.1155/2021/6627141 33936347
126. Fadare O.O. Okoffo E.D. Covid-19 face masks: A potential source of microplastic fibers in the environment Sci Total Environ 737 2020 Oct 1 140279 10.1016/j.scitotenv.2020.140279 32563114
127. Yashvardhini N. Kumar A. Gaurav M. Positive impact of COVID-19 induced lockdown on the environment of India's national capital, Delhi Spat Inf Res 30 2022 249 259 10.1007/s41324-021-00427-0
128. Dey A. Chozhavel Rajanathan T.M. Chandra H. Immunogenic potential of DNA vaccine candidate, ZyCoV-D against SARS-CoV-2 in animal models Vaccine. 39 30 2021 4108 4116 10.1016/j.vaccine.2021.05.098 34120764
129. Jha D.K. Yashvardhini N. Samiksha Kumar A. Rejuvenating impact of COVID-19 lockdown on major environmental parameters: an Indian perspective [published online ahead of print, 2022 Nov 29] Spat Inf Res 2022 1 13 10.1007/s41324-022-00499-6
|
PMC010xxxxxx/PMC10127842.txt |
==== Front
Educacio´n Me´dica
1575-1813
1579-2099
The Author(s). Published by Elsevier España, S.L.U.
S1575-1813(23)00022-0
10.1016/j.edumed.2023.100812
100812
Original Article
Evaluation of the impact of online education on the health-related quality of life of medical students in Lebanon
Evaluación del impacto de la educación en línea en la calidad de vida relacionada con la salud de los estudiantes de medicina en LíbanoHatem Georges a⁎
Omar Chaza Abou a
Ghanem Diana b
Khachman Dalia a
Rachidi Samar a
Awada Sanaa a
a Clinical and Epidemiological Research Laboratory, Faculty of Pharmacy, Lebanese University, Hadath, Lebanon
b Faculty of Public Health, Lebanese University, Fanar, Lebanon
⁎ Corresponding author.
25 4 2023
May-June 2023
25 4 2023
24 3 100812100812
23 1 2023
13 3 2023
© 2023 The Author(s)
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Introduction
One of the adopted measures during the Coronavirus disease of 2019 (COVID-19) pandemic in education is the online modality, which can impact students' quality of Life (QoL) and their academic performance.
Methods
This cross-sectional study aims to assess the impact of online education on health-related QoL in a sample of 401 medical students.
Results
Most students attended all the online sessions, and around 32% participated in some of them. Only 16.2% reported high satisfaction from the online sessions, and almost 39% reported low satisfaction. Overall, medical students had moderate QoL. Online education affected medical students psychologically more than physically, translated by higher scores for the physical components than the mental ones. Students who preferred online over in-person education had significantly lower scores for the mental domains, namely lower social functioning and emotional role scores. Academic satisfaction did not influence any SF-36 scores, and students’ preferences (online or in-person education) did not affect any of the physical component scores.
Conclusion
The pandemic directly impacted the QoL of medical students, namely their mental health. Medical students in Lebanon had a higher preference for in-person education, possibly due to its novelty and to other technological, technical, or personal challenges. Future research exploring the reasons and viable solutions should be performed to maintain a steady level of education during unanticipated situations.
Introducción
Una de las medidas adoptadas durante la pandemia de la enfermedad por Coronavirus de 2019 (COVID-19) en la educación es la modalidad en línea, la cual puede impactar la calidad de Vida (CdV) de los estudiantes y su rendimiento académico.
Métodos
Este estudio transversal tiene como objetivo evaluar el impacto de la educación en línea en la calidad de vida relacionada con la salud en una muestra de 401 estudiantes de medicina.
Resultados
La mayoría de los estudiantes asistieron a todas las sesiones en línea y alrededor del 32% participó en algunas de ellas. Solo el 16,2 % reportó una alta satisfacción con las sesiones en línea y casi el 39 % reportó una baja satisfacción. En general, los estudiantes de medicina tenían una CdV moderada. La educación en línea afectó a los estudiantes de medicina psicológicamente más que físicamente, lo que se traduce en puntajes más altos para los componentes físicos que para los mentales. Los estudiantes que preferían la educación en línea a la presencial tenían puntajes significativamente más bajos en los dominios mentales, es decir, puntajes más bajos en el funcionamiento social y el rol emocional. La satisfacción académica no influyó en ningún puntaje del SF-36, y las preferencias de los estudiantes (educación en línea o en persona) no afectaron ninguno de los puntajes del componente físico.
Conclusión
la pandemia impactó directamente en la CdV de los estudiantes de medicina, es decir, en su salud mental. Los estudiantes de medicina en el Líbano tenían una mayor preferencia por la educación presencial, posiblemente debido a su novedad y a otros desafíos tecnológicos, técnicos o personales. Se deben realizar investigaciones futuras que exploren las razones y las soluciones viables para mantener un nivel constante de educación durante situaciones imprevistas.
Keywords
Online education
Quality of life
Medical students
Physical component
Mental component
Palabras clave
Educación en línea
Calidad de vida
Estudiantes de medicina
Componente físico
Componente mental
==== Body
pmcIntroduction
During the COVID-19 pandemic, university students' Quality of Life (QoL) was negatively affected with increased reported mental health problems, including depression and anxiety.1 Previous research assessing the mental health of university students reported a high prevalence of psychological symptoms that were exacerbated by the emergence of the COVID-19 pandemic.2 Social isolation, anxiety, ambiguity, chronic stress, and economic hitches could have worsened the stress-related conditions, particularly among those with pre-existing psychological disorders, low resilience, and those more affected by the pandemic.3
In response to the virus’s rapid spread, universities were closed, resulting in a rapid shift from in-person to online education. Several benefits of distance learning were previously reported, such as ensuring the continuity of education,4 , 5 ensuring lifelong learning,6 , 7 and reducing the high costs associated with traditional education.8 Nevertheless, this new mode of education imposed new challenges due to social inequalities among students, namely accessing the internet and other technical supports.9 Limitations were also noted regarding teaching methods and schedule adaptation.10 However, the impact of online education was not limited to the education system since it affected the student’s learning experience and academic performance.11 Research published in 2019 showed that online education might affect the chances of students being employed after graduation, given that many job markets do not accept online degrees.12 Furthermore, the online mode of education affected students' physical health since physical inactivity increased due to decreased outdoor activity and increased screen time spent on computers or electronic devices.13 As a result, there is a special concern regarding the increased risks of metabolic and cardiovascular diseases associated with physical inactivity.14 , 15
The possible depreciation of students’ mental and physical health due to the pandemic and online education may adversely affect their QoL.16 Many reports highlighted the need to assess the impact of the COVID-19 pandemic and the corresponding changes on the health-related QoL of healthy populations.13 , 17 In particular, university students may be more sensitive to these changes due to several reasons, such as the suspension of schools, social distance measures, and social communication limitations.18 Thus, this study aims to assess the impact of online education on the health-related QoL of medical students in Lebanon.
Material and methods
Study design
An observational cross-sectional study was conducted over a period of three months (August-October 2021), targeting university students from different medical schools in Lebanon. The study protocol, survey, and consent form were reviewed and approved by the institutional review board of the faculty of pharmacy at the Lebanese University (reference 22/D/7).
Study sample and sample size
Students were recruited from four faculties: public health, medical sciences, dentistry, and pharmacy. These faculties were in both public (Lebanese university) and private universities (Lebanese American University, American University of Beirut, Saint Joseph University, Beirut Arabic University, and the University of Balamand). No criteria were based on sex, nationality, age, and ethnicity. Nevertheless, students refusing to participate were excluded from the study. The sample size was determined using the Epi Info 7 software. As there were no similar studies related to coronavirus disease, the calculation assumed that the probability of having good mental and physical health was 50%. Considering a 95% confidence interval and a 5% acceptable margin of error, at least 386 participants were required, with a 5% loss to follow-up.
Data collection
Two pharmacists approached students from different faculties. They explained the study’s aims orally and invited them to participate by filling out an online survey. The first page of the survey included the study’s aims (written), the estimated time to answer the questionnaire (12–15 minutes), statements on the anonymity and confidentiality of their responses, and consent to use their data for research purposes only. Participation in the survey was strictly voluntary.
Study tool
After an extensive literature review, a validated questionnaire was used for data collection (Supplementary material).19 It was available in English and Arabic, depending on the student’s preferences. The first part included questions about the general characteristics of the participants (sex, age (≤30 or >30), the governorate of residence (including the eight governorates in Lebanon), marital status, type of the university (public or private), faculty (public health, medical sciences, dentistry or pharmacy) and the academic year. In the second part, students were asked five questions about the characteristics of online education (frequency of attendance, if the online sessions included practical and theoretical work if they felt that online education provided them with sufficient information, satisfaction with the sessions, and preferences between online and in-person education). In the last part, The 36-item Short Form (SF-36) survey (validated on Lebanese university students) was used,19 and the corresponding answers were afterward graded over 100. These questions will generate eight subscales: physical functioning (PF), physical role (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), emotional role (ER), and mental health (MH). The first four scores lead to the physical composite score (PC), while the last four can be summed to create the mental composite score (MC). A minimum score of 0 represents the lowest QoL, and a maximum of 100 represents the highest.
Statistical analysis
Statistical analyses were performed using Statistical Package for Social Sciences (SPSS Inc, Chicago, Illinois) Version 27. Categorical variables are presented through frequencies and percentages. These variables included the general characteristics of the students and the questions related to the characteristics of online education. The different scores are presented through means and standard deviations. Individual SF-36 items were recorded, summed, and transformed as recommended.20 One-way analysis of variance (ANOVA) was performed to assess the relationship between the SF-36 scale scores and questions related to online education. A multivariate analysis was performed using linear regression models to evaluate the combined effect of predictors (online education variables) on SF-36 dimensions. The confusion variables (or covariates) were those related to the characteristics of online education. They were considered if they had a statistically significant p-value (< 0.05) in the ANOVA analysis.
Results
General characteristics of the students
Overall, 424 students were approached, and 401 agreed to participate in the study (94.6%). The general characteristics of the study participants are presented in Table 1 . The sample consisted of 45.9% of men and 54.1% of women. Most of the students were less than 30 years of age (93.3%). Participants were distributed in the different governorates, with Mount Lebanon (19.7%), Beirut (16.7%), Nabatiyeh (13.2%), and Akkar (12.7%) accounting for the highest percentages. Among the students, 83% were single, and the rest were either married or engaged. Almost a third of the participants were from public universities, and 68.6% studied in private universities. Regarding distribution among faculties, 32.9% were from the faculty of public health, 26.4% from the faculty of medical sciences, 20.9% from the faculty of pharmacy, and 19.7% from the faculty of dentistry. A third of the participants were in their first couple of years of studies, 41.4% were in their third or fourth year, and the rest (25.2%) were in their fifth year or more.Table 1 Distribution of the general characteristics of the students.
Table 1 Total (N = 401)
Frequency (%)
Sex Man 184 (45.9%)
Woman 217 (54.1%)
Age (years) ≤30 374 (93.3%)
>30 27 (6.7%)
Governorate of residence Mount Lebanon 79 (19.7%)
Beirut 67 (16.7%)
Nabatiyeh 53 (13.2%)
Akkar 51 (12.7%)
South 44 (11.0%)
Bekaa 43 (10.7%)
North 38 (9.5%)
Baalbeck/Hermel 26 (6.5%)
Marital status Single 333 (83.0%)
Married/engaged 68 (17.0%)
University Public university 126 (31.4%)
Private university 275 (68.6%)
Faculty Public Health 132 (32.9%)
Medical sciences 106 (26.4%)
Pharmacy 84 (20.9%)
Dentistry 79 (19.7%)
Academic year 1–2 134 (33.4%)
3–4 166 (41.4%)
5 or more 101 (25.2%)
Results are given in terms of frequency (percentage).
Characteristics and students’ preferences for online education
Table 2 presents the characteristics and students’ preferences for online education. Almost 62% of the students reported attending all the online sessions, 32.2% attended sometimes, and only 6.2% did not attend any. Around half of the sessions included both theoretical and practical work. Overall, 37.2% of the participants reported that online education provided them with sufficient information, and most (84.3%) preferred in-person over online education. As regards satisfaction with the online sessions, 16.2% of the students reported high satisfaction, 45.1% reported moderate satisfaction, and 38.7% reported low satisfaction.Table 2 Questions related to online education perception among medical students.
Table 2Total (N = 401)
Questions Frequency (%)
How much time did you attend the online sessions?
All of the time 247 (61.6%)
Sometimes 129 (32.2%)
Never 25 (6.2%)
Did online education include theoretical and practical work?
Yes 206 (51.4%)
Compared with in-person education, do you feel that online education provides you with sufficient information?
Yes 149 (37.2%)
How do you rate your satisfaction with the online sessions?
High satisfaction 65 (16.2%)
Moderate satisfaction 181 (45.1%)
Low satisfaction 155 (38.7%)
Do you prefer online education or in-person education?
Online Education 63 (15.7%)
In-person Education 338 (84.3%)
Results are given in terms of frequency (percentage).
Health-related quality of life of medical students and the impact of online education
All scales had a high level of internal consistency reliability coefficient. The median values were 73.7 and 56.7 for PC and MC, respectively. Fig. 1 represents the mean value of the eight scales and the mean values of the physical and mental components. The physical functioning scale had the highest mean (83.3), while the vitality and mental health scales had the lowest mean (50.4 and 54.4, respectively). Compared to the mental components, physical components had a higher mean (70.2 versus 55.9).Fig. 1 Mean values of the different health-related quality of life scores of medical students in Lebanon
PF: Physical Functioning, RP: Role Physical role, BP: Bodily Pain, GH: General Health, PC: total Physical component, VT: Vitality, SF: Social Functioning, ER: Emotional Role, MH: Mental Health and MC: total Mental Component.
Fig. 1
Table 3 shows online education's impact on students' health-related QoL. Significant differences were noted in the frequency of attendance of the online sessions and the PF, BP, PC, VT, DF, MH, and MC scores, with those attending all the sessions having a higher score than those attending sometimes or never. Students attending online sessions that included theoretical and practical work had higher BP, GH, and MH scores and lower ER scores. Those reporting that online education provided them with sufficient information had significantly higher BP and RP scores and lower VT scores. Furthermore, significantly lower SF and ER scores were noted for those preferring online over in-person education.Table 3 Relation between the SF-36 scale scores and the online education variables (One Way ANOVA).
Table 3 Physical components
PF RP BP GH PC
How much time did you attend the online sessions?
All of the time 84.1 (23.6) 72.7 (37.6) 76.1 (26.9) 56.0 (16.3) 72.2 (20.5)
Sometimes 83.9 (22.3) 69.4 (33.7) 65.1 (27.3) 53.8 (11.8) 68.0 (18.3)
Never 71.4 (26.1) 64.0 (33.1) 57.8 (32.7) 51.0 (15.0) 61.1 (22.8)
F 3.440* 0.862 10.092*** 1.916 4.676**
Did online education include theoretical and practical work?
Yes 83.7 (14.0) 72.3 (28.1) 75.4 (16.7) 56.5 (15.8) 71.9 (20.7)
No 82.8 (13.0) 69.7 (23.9) 67.2 (18.9) 53.4 (13.9) 68.3 (19.4)
F 0.170 0.513 8.659** 4.339* 3.415
Do you feel that online education provides you with sufficient information?
Yes 87.8 (10.6) 75.3 (22.2) 70.7 (16.8) 56.1 (14.8) 72.5 (18.2)
No 80.5 (14.2) 68.5 (28.1) 71.8 (18.8) 54.3 (14.9) 68.8 (21.1)
F 8.922** 3.322* 0.133 1.446 3.169
How do you rate your satisfaction with the online sessions?
High Satisfaction 82.2 (16.2) 67.7 (23.9) 68.3 (29.4) 53.8 (17.2) 68.0 (22.8)
Moderate Satisfaction 83.5 (12.2) 73.9 (19.2) 73.2 (23.3) 56.7 (13.2) 71.8 (18.1)
Low Satisfaction 83.4 (13.9) 69.2 (22.0) 70.5 (28.3) 53.4 (15.7) 69.2 (20.9)
F 0.075 1.047 0.857 2.257 1.192
Do you prefer online education or in-person education?
Online Education 85.1 (12.9) 68.2 (29.4) 68.3 (17.7) 55.3 (18.0) 69.2 (21.9)
In-person Education 82.9 (13.6) 71.6 (25.5) 72.0 (18.1) 54.9 (14.3) 70.3 (19.7)
F 0.443 0.454 0.871 0.039 0.157
Mental components
VT SF ER MH MC
How much time did you attend the online sessions?
All of the time 53.1 (23.4) 62.3 (29.4) 61.3 (34.1) 56.4 (22.3) 58.3 (25.3)
Sometimes 46.2 (18.7) 50.2 (25.0) 63.3 (33.8) 52.4 (17.3) 53.0 (21.2)
Never 45.8 (22.5) 53.5 (23.3) 46.7 (30.8) 44.2 (16.6) 47.5 (20.2)
F 4.754** 8.572*** 1.518 4.938** 3.753*
Did online education include theoretical and practical work?
Yes 51.0 (23.3) 58.5 (30.3) 56.8 (34.8) 56.8 (21.5) 55.7 (25.5)
No 49.9 (21.0) 57.3 (25.9) 65.5 (32.6) 51.8 (19.5) 56.1 (22.1)
F 0.257 0.185 3.936* 5.950* 0.019
Do you feel that online education provides you with sufficient information?
Yes 47.8 (19.1) 54.7 (25.6) 63.3 (34.0) 53.4 (18.2) 54.8 (21.5)
No 52.0 (23.7) 59.8 (29.5) 59.7 (33.9) 54.9 (22.1) 56.6 (25.2)
F 3.359* 2.981 0.648 0.557 0.523
How do you rate your satisfaction with the online sessions?
High Satisfaction 48.9 (21.1) 51.4 (30.5) 54.9 (34.3) 52.1 (21.3) 51.8 (24.6)
Moderate Satisfaction 50.8 (20.7) 60.5 (25.2) 61.1 (33.2) 55.5 (19.0) 57.0 (21.8)
Low Satisfaction 50.7 (24.3) 57.5 (30.2) 63.4 (34.6) 53.9 (22.4) 56.4 (25.9)
F 0.184 2.534 0.872 0.715 1.174
Do you prefer online education or in-person education?
Online Education 47.3 (20.2) 51.9 (28.9) 48.1 (36.3) 50.7 (20.4) 49.5 (23.1)
In-person Education 51.0 (22.5) 59.0 (27.9) 63.4 (33.1) 55.1 (20.7) 57.1 (23.9)
F 1.509 3.308* 6.502* 2.400 5.426*
*p < 0.050; **p < 0.010; ***p < 0.001. F = F statistics.
PF: Physical Functioning, RP: Role Physical role, BP: Bodily Pain, GH: General Health, PC: total Physical component, VT: Vitality, SF: Social Functioning, ER: Emotional Role MH: Mental Health and MC: total Mental Component.
After adjusting for covariates, it was found that the students reported satisfaction had no influence on the SF-36 scores, and the preferences of students (online or in-person education) did not influence any of the physical components scores. Students attending some of the sessions had significantly lower BP (b = –5.0), PC (b = –2.1), VT (b = –3.4), SF (b = –6.3), and MC (b = –2.8) scores compared to those attending all the sessions, while those never attending the sessions had lower PF (b = –3.7), BP (b = –5.6), PC (b = –3.7), MH (b = –3.7) and MC (b = –3.6) scores. The BP (b = –6.1), GH (b = –3.1), and MH (b = –4.1) scores decreased if the online sessions did not include both practical and theoretical work. Students reporting that online education cannot provide sufficient information had significantly lower PF (b = –6.5) and RP (b = –6.8) scores but a greater VT score (b = 4.4). Finally, those preferring in-person over online education had significantly higher SF (b = 8.0), ER (b = 14.3), and MC (b = 8.0) scores (Table 4 ).Table 4 Predictors of the SF-36 scale scores: Multivariate analysis results (Linear regression model).
Table 4 Physical components
PF
B [95% CI] RP
B [95% CI] BP
B [95% CI] GH
B [95% CI] PC
B [95% CI]
Intercept 94.6
[86.4 – 102.9] 82.1
[69.7 – 94.5] 84.8
[76.3 – 93.2] 59.6
[54.9 – 64.1] 72.2
[69.7 – 74.7]
How much time did you attend the online sessions?
(Sometimes Vs All of the time) –0.3
[–2.7 – 2.2] –5.0***
[–7.9 – –2.0] –2.1*
[–4.2 – –0.1]
(Never Vs All of the time) –3.7*
[–6.9 – –0.5] –5.6**
[–9.3 – –1.8] –3.7**
[–6.5 – –1.0]
Did online education include theoretical and practical work?
(No Vs Yes) –6.1*
[–11.6 – –0.7] –*3.1
[–6.0 – –0.2]
Do you feel that online education provides you with sufficient information?
(No Vs Yes) –6.5**
[–11.2 – –1.7] –6.8*
[–14.1 – –0.1]
R square (%) 34% 8% 60% 11% 23%
Mental components
VT
B [95% CI] SF
B [95% CI] ER
B [95% CI] MH
B [95% CI] MC
B [95% CI]
Intercept 39.3
[38.2 – 53.8] 47.7
[33.7 – 61.8] 23.2
[–1.4 – 47.7] 62.2
[55.9 – 68.5] 43.6
[31.6 – 55.6]
How much time did you attend the online sessions?
(Sometimes Vs All of the time) –3.4**
[–5.7 – –1.0] –6.3***
[–9.3 – –3.4] –1.7
[–3.9 – 0.5] –2.8*
[–5.3 – –0.3]
(Never Vs All of the time) –2.8
[–5.8 – 0.2] –3.0
[–6.8 – 0.8] –3.7**
[–6.5 – –0.9] –3.6*
[–6.8 – –0.4]
Did online education include theoretical and practical work?
(No Vs Yes) 7.7
[–0.8 – 16.3] –4.1*
[–8.1 – –0.1]
Do you feel that online education provides you with sufficient information?
(No Vs Yes) 4.4*
[0.1 – 8.9]
Do you prefer online education or in-person education?
(In-person Vs Online) 8.0*
[0.6 – 15.5] 14.3*
[2.5 – 26.1] 8.0*
[1.7 – 14.4]
R square (%) 32% 52% 24% 34% 33%
Results are presented as B [95% confidence interval (CI)]. *p < 0.050; **p < 0.010; ***p < 0.001.
PF: Physical Functioning, RP: Role Physical role, BP: Bodily Pain, GH: General Health, PC: total Physical component, VT: Vitality, SF: Social Functioning, ER: Emotional Role MH: Mental Health and MC: total Mental Component.
Discussion
The present study aimed to investigate the impact of online education on the health-related QoL of medical students in Lebanon. Similarly to our findings, a recent study found low QoL among university students during the COVID-19 pandemic, mainly the mental components of the SF-36.19 In contrast, compared to a previous study conducted in Lebanon, the QoL of students was adversely affected,21 which could be possibly associated with online education and other stressors during the data collection period, such as political unrest, economic crisis and the COVID-19 pandemic. Most students attended all the online sessions, and around 32% attended some of them. Students showed low and moderate satisfaction with the online education sessions, and about 37% reported that these programs needed to provide them with sufficient information compared to in-person education. Similarly to other studies, students’ satisfaction was moderate, possibly due to their unfamiliarity with the different online platforms.22 A literature review published in 2021 revealed that accessibility problems, technical issues, mental health, and lecturer commitment could affect students’ satisfaction and adaptation to online education.23 Most students in this study preferred in-person over online education, likely associated with the fact that the perceived learning outcomes are influenced by the impact of the interactions in classrooms, course structure, and students’ motivation.24 Based on the SF-36 health survey, students had high physical and relatively low mental component scores. Although the online modality was reported to be time-saving, limitations such as content perception and technical and behavioral challenges during the sessions and the exams were also reported among students.25 Research showed that medical students experienced poor sleep quality and depression during the pandemic independently of their education level.26 Students attending some or none of the sessions had significantly lower BP scores than those attending all of them, possibly associated with physical and neck pain due to prolonged positioning when using the devices.27 Lower BP and GH scores were reported among students attending sessions that included only theoretical work. This finding can be explained by higher academic achievement among students translating their theoretical knowledge into practice.28 Medical students reporting that online education did not provide sufficient information had significantly lower PF and RP scores which can be related to the low student-to-instructor and student-to-student interactions through the online mode.29 Lower VT and SF scores were reported among those attending some sessions compared to those attending all the sessions. This can be associated with detachment from colleagues, affecting students' motivation, QoL, and vitality.30 It can also explain the higher SF and ER scores among those preferring in-person education.
This study has limitations. Selection bias may have been induced since the data collectors randomly selected students, and possibly only motivated ones were willing to participate. However, pharmacists were uniformly trained and used the same data collection form, and a different researcher performed data coding and analysis. Recall bias may have also arisen since data collection was performed at the end of the academic year. The study sample was relatively small, which might not allow the extrapolation of the findings to all medical students. Nonetheless, the present study is the first one performed in both public and private universities to tackle the impact of online education on the health-related QoL of medical students using a validated questionnaire. It is suggested that other variables, such as employment status and availability of technical support and resources, will be addressed in future studies.
The COVID-19 pandemic directly impacted the QoL of medical students, namely their mental health. Online teaching imposed new challenges, with most students reporting moderate or low satisfaction. Those attending the sessions more frequently and those reporting that online education provided sufficient information had significantly higher scores. Medical students in Lebanon had a higher preference for in-person education, possibly due to its novelty and to other technological, technical, or personal challenges. Future research exploring the reasons and viable solutions should be performed to maintain a steady level of education during unanticipated situations.
Declaration of Competing Interest
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
‘The Authors declare that there is no conflict of interest.
The study protocol and tool were reviewed and approved by the institutional review board of the faculty of pharmacy of the Lebanese University.
Appendix A Supplementary data
Supplementary material
Image 1
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.edumed.2023.100812.
==== Refs
References
1. Chang J. Yuan Y. Wang D. Mental health status and its influencing factors among college students during the epidemic of COVID-19 Nan Fang Yi Ke Da Xue Xue Bao 171-6 2020
2. Bayram N. Bilgel N. The prevalence and socio-demographic correlations of depression, anxiety and stress among a group of university students Soc Psychiatry Psychiatr Epidemiol 43 8 2008 667 672 18398558
3. Sher L. The impact of the COVID-19 pandemic on suicide rates QJM: An Int J Med 113 10 2020 707 712
4. Akinbadewa B.O. Sofowora O.A. The effectiveness of multimedia instructional learning packages in enhancing secondary school students’ attitudes toward Biology Int J Stud Educat (IJonSE) 2 2 2020 119 133
5. Seage S.J. Türegün M. The effects of blended learning on STEM achievement of elementary school students Int J Res Educat Sci 6 1 2020 133 140
6. Alharthi M. Students' attitudes toward the use of technology in online courses Int J Technol Des Educ 3 1 2020 14 23
7. Hatem G. Goossens M. Ghanem D. Bou Assi R. Evaluation of the effectiveness of continuing professional development in the pharmaceutical workplace: a cross-sectional study in Lebanon BAU J Creat Sustain Develop 3 1 2021 8
8. Al-Husban N.A. Critical thinking skills in asynchronous discussion forums: a case study Int J Technol Des Educ 3 2 2020 82 91
9. UNESCO Education: From disruption to recovery 2021 Available from https://en.unesco.org/covid19/educationresponse>
10. Thompson V.L. McDowell Y.L. A case study comparing student experiences and success in an undergraduate mathematics course offered through online, blended, and face-to-face instruction Int J Educat Math Sc Technol 7 2 2019 116 136
11. Hebebci M.T. Bertiz Y. Alan S. Investigation of views of students and teachers on distance education practices during the Coronavirus (COVID-19) pandemic Int J Technol Educat Sci 4 4 2020 267 282
12. Sadeghi M. A shift from classroom to distance learning: advantages and limitations Int J Res English Educat 4 1 2019 80 88
13. Balanzá-Martínez V. Atienza-Carbonell B. Kapczinski F. De Boni R.B. Lifestyle behaviours during the COVID-19–time to connect Acta Psychiatr Scand 141 5 2020 399 32324252
14. Hall G. Laddu D.R. Phillips S.A. Lavie C.J. Arena R. A tale of two pandemics: How will COVID-19 and global trends in physical inactivity and sedentary behavior affect one another? Prog Cardiovasc Dis 64 2021 108 32277997
15. Lippi G. Henry B.M. Sanchis-Gomar F. Physical inactivity and cardiovascular disease at the time of coronavirus disease 2019 (COVID-19) Eur J Prev Cardiol 27 9 2020 906 908 32270698
16. Zhang Y. Ma Z.F. Impact of the COVID-19 pandemic on mental health and quality of life among local residents in Liaoning Province, China: a cross-sectional study Int J Environ Res Public Health 17 7 2020 2381 32244498
17. Banerjee D. Rai M. Social isolation in Covid-19: The impact of loneliness 2020 SAGE Publications Sage UK London, England 525 527
18. Hatem G. Goossens M. Health care system in Lebanon: a review addressing health inequalities and ethical dilemmas of frointline workers during COVID-19 pandemic BAU J Health Wellbeing 5 1 2022
19. Awada S. Ahmad L. Assi R.B. Rafei Z. Omar C.A. Hatem G. Health-related quality of life of medical students during the COVID-19 pandemic: a cross-sectional study in Lebanon Power Educat 0 0 2023 10.1177/17577438231155628
20. Ware J. Kosinski M. Keller S. SF-36 physical and mental health summary scales. A user's manual 2001 1994
21. Sabbah I. Sabbah H. Khamis R. Sabbah S. Droubi N. Health related quality of life of university students in Lebanon: lifestyles behaviors and socio-demographic predictors Health 5 7D 2013 1 12 10.4236/health.2013.57A4001
22. Simsek I. Kucuk S. Biber S.K. Can T. Online learning satisfaction in higher education amidst the Covid-19 pandemic Asian J Dist Educat 16 1 2021 247 261
23. Ranadewa D.U.N. Gregory T.Y. Boralugoda D.N. Silva J.A.H.T. Jayasuriya N.A. Learners’ satisfaction and commitment towards online learning during COVID-19: a concept paper Vision. 0 0 2021 10.1177/09722629211056705
24. Baber H. Determinants of students’ perceived learning outcome and satisfaction in online learning during the pandemic of COVID-19 J Educat e-learning Res 7 3 2020 285 292
25. Khalil R. Mansour A.E. Fadda W.A. Almisnid K. Aldamegh M. Al-Nafeesah A. The sudden transition to synchronized online learning during the COVID-19 pandemic in Saudi Arabia: a qualitative study exploring medical students’ perspectives BMC Med Educat 20 1 2020 1 10
26. Said A. Yusof M. Mohd F. Azmi M. Hanapiah M. Abdullah A. Poor sleep quality among medical students in International Islamic University Malaysia (IIUM) and its association with mental health and other factors IIUM Med J Malaysia 19 2 2020
27. Fares J. Fares M.Y. Fares Y. Musculoskeletal neck pain in children and adolescents: risk factors and complications Surg Neurol Int 8 2017 72 28584675
28. El Kharki K. Berrada K. Bensamka F. Burgos D. The adoption of open educational practices to support practical work at Moroccan universities Rad Solut Educat Africa: Springer 2021 233 249
29. Wut T.-m. Xu J. Person-to-person interactions in online classroom settings under the impact of COVID-19: a social presence theory perspective Asia Pacific Educat Rev 22 3 2021 371 383
30. Martinek D. Carmignola M. Müller F.H. Bieg S. Thomas A. Eckes A. How can students feel more vital amidst severe restrictions? Psychological needs satisfaction, motivational regulation and vitality of students during the coronavirus pandemic restrictions Eur J Investigat Health, Psychol Educat 11 2 2021 405 422
|
PMC010xxxxxx/PMC10130806.txt |
==== Front
Int J Environ Res
Int J Environ Res
International Journal of Environmental Research
1735-6865
2008-2304
Springer International Publishing Cham
37128551
520
10.1007/s41742-023-00520-2
Research Paper
Floristic Diversity as an Indicator in Low and High Endemic Buruli Ulcer Areas in Côte d’Ivoire
http://orcid.org/0000-0002-5655-0680
Ehouman Evans ehoumanevans@gmail.com
1
Soro Dramane 23
Ouattara Doudjo Noufou 13
Cissé Cathérine Boni 45
Bakayoko Adama 13
Dosso Mireille 45
Zo-Bi Irié Casimir 6
Kouassi Akossoua Faustine 7
Koné Mamidou Witabouna 134
1 grid.452889.a 0000 0004 0450 4820 UFR Sciences de la Nature, Université Nangui Abrogoua, 02 BP 801, Abidjan 02, Abidjan, Côte d’Ivoire
2 Department of Biological Sciences, Université Péléforo Gon Coulibaly, Korhogo, Côte d’Ivoire
3 grid.462846.a 0000 0001 0697 1172 Centre Suisse de Recherches Scientifiques en Côte d’Ivoire (CSRS), Abidjan, Côte d’Ivoire
4 grid.410694.e 0000 0001 2176 6353 Department of Medical Science, Université Félix Houphouët-Boigny, Abidjan, Côte d’Ivoire
5 grid.418523.9 0000 0004 0475 3667 Institut Pasteur, Mycobactéries Tuberculeuses et Atypiques, Abidjan, Côte d’Ivoire
6 grid.473210.3 Department of Forest Science, Institut National Polytechnique Félix Houphouët-Boigny (INP-HB), Yamoussoukro, Côte d’Ivoire
7 grid.410694.e 0000 0001 2176 6353 Centre National de Floristique, Université Felix Houphouët-Boigny, Abidjan, Côte d’Ivoire
26 4 2023
2023
17 3 404 1 2023
2 3 2023
4 3 2023
© University of Tehran 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Mycobacterium ulcerans is an environmental bacterium responsible for Buruli ulcer. This disease has a high frequency index in humid tropical regions, with a high incidence in Sub-Saharan Africa. The ecology and mode of transmission of this disease is not well established. Based on dilution effect hypothesis, acting as lowering disease transmission due to greater biodiversity, floristic inventory was carried out in the Health Districts of Daloa and Bouaké in Côte d’Ivoire. In each district, high and low endemic sites were investigated. A total of 169 plant species were inventoried for both low and high endemicity of Buruli ulcer sites in the districts. The Indval index revealed that 13 plant species were good indicators for Buruli ulcer highly endemic areas. The plants which correlate with high endemicity area were Leersia hexandra, Panicum laxum, Mimosa pudica, Paspalum distichum, Persicaria senegalensis, Calopogonium mucunoides, Echinochloa colona, Ipomoea sagittata, and Eichhornia crassipes. For low endemic sites, a strong relationship was recorded for 37 plants. The indices revealed low similarity between high and low endemicity sites. Low endemicity sites expressed the highest plant species diversity. These results suggest the hypothesis that floristic richness is more important in sites of low endemicity than in those of high endemicity. Moreover, we observed a co-occurrence of some plant species and Buruli ulcer endemicity. This finding may lead to the fact that it is important to care about the biodiversity to prevent outbreak of Buruli ulcer cases.
Supplementary Information
The online version contains supplementary material available at 10.1007/s41742-023-00520-2.
Keywords
Buruli ulcer
Mycobacteria
Plants diversity
Côte d’Ivoire
Herbaceous vegetation
Water body
Programme d'Appui Stratégique à la Recherche Scientifique (PASRES)158/2015 Soro Dramane issue-copyright-statement© University of Tehran 2023
==== Body
pmcIntroduction
The loss of plant biodiversity influences the emergence of human diseases. The relationship between the emergence of new diseases and biodiversity degradation is important for establishing risk factors related to the environment. Vegetation composition can be used as an indicator of the level of environmental degradation. This situation could be a key factor leading to the appearance of new diseases (Young et al. 2017). Furthermore, the health of human is connected to the health of non-human animals and environment (Webster et al. 2016; Manlove et al. 2016), implying that a better understanding of environmental elements is essential to master disease aetiology. Besides, there is wide range of study which indicated that the transmission of a pathogen may increase when diversity diminishes (Ostfeld 2017; Dharmarajan et al. 2021; Keesing and Ostfeld 2021).
Buruli ulcer (BU) is among the diseases related to environmental factors. The causal agent is Mycobacterium ulcerans (Hotez and Kamath 2009; Sopoh and Asiedu 2016). BU is a serious public health problem around the world but mostly in Africa, Asia, Oceania and South-America (WHO 2018). In 2017, a total of 2217 new cases of Buruli ulcer were reported worldwide and 1923 of them in the African Region (WHO 2017a).
In West African countries, the number of people suffering from this disease is still high. For example, in 2004, Côte d’Ivoire has recorded more than 20,000 cases, Ghana, more than 6000 cases and Benin more than 4000 cases. Whilst in 2016, 376 new cases were recorded in Côte d’Ivoire, 371 in Ghana and 312 in Benin (WHO 2017b). In many African countries, people were affected by Buruli ulcer face physical disability, incapacity and social discrimination (Debacker et al. 2004). In rural communities, there is a strong belief that BU is due to magical and mystical causes, and people afflicted are also designated as witches, resulting in social consequences and many functional limitations (Koka et al. 2016). The belief in the mystical origin of M. ulcerans (MU) is strongly held, since its mechanism of transmission is not well understood by local populations (Garchitorena et al. 2015; Khanna et al. 2016).
Several studies have been carried out to identify the reservoirs and vectors of M. ulcerans. There is considerable agreement that it flourishes in aquatic environment and DNA fragments have been isolated in various environmental samples such as aquatic insects, forbs, molluscs and fishes (Falkinham 2015; Narh et al. 2015; Tian et al. 2016). In Ghana, McIntosh et al. (2014) established a relationship between some plant taxa and M. ulcerans.
The emergence of Buruli ulcer is associated with anthropogenic environmental modifications (Merritt et al. 2010). In Côte d’Ivoire, due to anthropogenic conditions such as deforestation, urbanisation, hydro-electric dam installation, mining facilities and the use of pesticides and plant protection products in agriculture, many ecosystems have been modified and devastated. In a general context of diseases emerging from environment to human or spillover diseases such as Covid-19 (McNeely 2021) or Ebola (Lee-Cruz et al. 2021; Mursel et al. 2022), it is crucial to snapshot biotic components of ecosystem to have a better understanding of these diseases. Thus, in our study, we are interested in plant composition. Therefore, it is an urgent need to study the abundance, distribution and composition of the plant diversity around water points frequently used by local people for their daily needs in absence of pipe and tap, when people need to go to wells. This study may serve as background research for predicting the endemicity of Buruli ulcer. This paper concentrates on floristic composition around water body points in Buruli ulcer endemic areas in Côte d’Ivoire.
Materials and Methods
Study Area
This study was carried out in three Health Districts in Côte d’Ivoire. Four villages were selected according to the number of cases recorded according to the Buruli Ulcer National Programme (PNLUB, 2013 and 2014): Daloa, Zébra (6° 55′ N, 6° 31′ W) and Batéguédia (6° 48′ N, 6° 40′ W); Bouaké Nord-Est and Bouaké Sud, Kongodékro (7° 37′ N, 5° 2′ W) and Konankro (7° 37′ N, 5° 2′ W). For each village, two sites were identified in agreement with the civil servant in charge of Buruli ulcer programme in each Health Districts and local guide. The mean annual rainfall in the region of Bouaké and Daloa is, respectively, 95 mm/month and 109.71 mm/month. The mean annual temperature is 26.22 °C for Bouaké and 25.61 °C for Daloa from 1986 to 2016 (FetchClimate 2016). The floristic data have been collected from August 2015 to February 2016. Study sites and number of cases recorded for the years 2013 and 2014 are illustrated in Fig. 1.Fig. 1 Location of the study sites and distribution of Buruli ulcer cases recorded in different sanitary districts of Côte d’Ivoire (2013 and 2014)
Floristic Inventory
A systematic floristic inventory was carried out around stagnant bodies rivers frequently used by local people using quadrats method. Thirty to fifty quadrats (1 m × 1 m) were placed at the edge of the water body (De Magalhaes et al. 2016; EPA 2002). The number of quadrats varied depending on the size of the study site. A total number of 273 quadrats were installed during this study. Species identification was done on the field and then authenticated at the Herbarium of Centre Suisse de Recherches Scientifiques en Côte d’Ivoire and at the Herbarium of Centre National de Floristique at Université Félix Houphouët-Boigny Abidjan (Côte d’Ivoire). GPS positions were taken with a GPSmap 64 (GPSmap 64, Garmin International Inc., Olathe, KS, USA), with 3 to 5 m of precision in UTM 30 N, geographic coordinate system WGS84 (World Geodetic Survey 1984).
The nomenclature of families followed APG IV classification (Chase et al. 2016).
Life Forms
The formulations of Aké-Assi (2001, 2002) were used to classify the species in: Chamaephytes (Ch), Mesophanaerophytes (mP), Epiphytes (Ep), Megaphanerophytes (MP) Geophytes (G) Nanophanearophytes (np), Mesophanerophytes (mP), Microphanerophys (mp), Geophyte Rhizomatous (Gr), Hemicryptophytes (H) Rhizomes (rh), Hydrophytes (Hyp), Saprohytis (Sapr), swimming Hydrophytic (Hydn), Semi-aquatic (Se-Aq), Liana (L), Semi-epiphytes (Se-Ep), Microphanerophytes (mp), Stolonifers (Sto), and Therophytes (Th).
Chorology and Phytogeography Indices
The chorology habitat of the species was established according to Aké-Assi (2001; 2002) and Chatelain (Chatelain et al. 2011). Then the Phytogeographical index was calculated using the formula:
PI = (SZ + I)/(GC + GC-ZS) by considering the number of species belonging to different chorological areas (White 1986): Guineo-Congolian-Sudano-Zambezian (GC-ZS), Guineo-Congolian (GC), Sudano-Zambezian (SZ), and Sudanian (S). When PI > 1, the vegetation is a Sudanese phytosociological type (Akpagana 1989; Adomou et al. 2007).
Indicator Plant Species for Sites with High and Low Buruli Ulcer Endemicity
Indicator species were determined with the indicator value (Indval) method (Dufrêne and Legendre 1997; De Cáceres et al. 2010). Indicator values (Indval) were calculated using the labdsv package (Roberts et al. 2016). Species were considered significant indicators of low and high Buruli ulcer endemic sites for Indval higher than 10% at p ≤ 0.05 (Olleck et al. 2020). Plant species richness in different sites was determined with the betapart package (Baselga and Orme 2012). We also used the vegan package (Oksanen et al. 2017) to compute the alpha and beta diversity indices.
Data Analyses
Comparison of species composition and chorology affinities and percentage of life form in areas of high and low Buruli ulcer endemism was established using Fisher's exact test statistics with a value of alpha level of 0.05 (Fisher et al. 2011).
The alpha and beta diversity indices of Shannon (H’), Pielou (J), and Simpson (D) were calculated to determine the floristic diversity of high and low Buruli ulcer sites. Additionally, Sørensen (S) Jaccard (J), and Whittaker (W) indices were calculated to establish species similarity between sites. When interesting, to compare a parameter in different groups, the Mann–Whitney U statistic (or the parametric equivalent Student t test) was used. Comparison of diversity was making possible through out the graphical diversity profile using the entropart package (Marcon and Hérault).
To assess discrimination of the Buruli ulcer sites based on their plant species composition, the matrices of species per sites were submitted to a non-metric multidimensional scaling (NMDS) which produces an ordination based on Bray–Curtis dissimilarity. The NMDS method provides a graphical representation of all the sites according to their floral compositions, with a stress value expressing the quality of the graphical representation. The smaller the stress value is, the better the data will be represented in the graphic (Helm et al. 2017). Moreover, the more the sites are similar, the more they are closed in the NMDS graphic representation (De Magalhaes et al. 2016). NMDS was executed with the vegan package. All the analysis was performed with RStudio version 1.0.44 (RStudio Team 2016) and R 3.2.2 (R Core Team 2017).
Results
Floristic Composition and Diversity Around Water Body Points
A total number of 29,615 individuals were recorded during this study. For the 8 sites, 169 species belonging to 131 genera and 47 families of plants were identified.
The most dominant family was Poaceae (17.75%) with 30 species, followed by Fabaceae (13.02%), Cyperaceae (8.88%), Asteraceae and Euphorbiaceae (4.14%), Acanthaceae, Apocynaceae and Convolvulaceae (3.55%), Malvaceae and Onagraceae (2.96%), Rubiaceae, Solanaceae and Vitaceae (2.37%). For Amaranthaceae, Marantaceae, Phyllanthaceae and Urticaceae, three species were recoded and represented 1.78% of the flora. Concerning the Arecaceae, Commelinaceae, Dioscoreaceae, Menispermaceae, Passifloraceae and Polygonaceae, only two species (1.18%) were recorded. The combined 24 lesser families represented 0.59% of all taxa. For these 24 families, only 1 species was recorded in each of them (Fig. 2). The most represented genera were Cyperus, Ipomoea, and Ludwigia.Fig. 2 Distribution of families around water body sites in study areas
Floristic Composition in Site of Low and High Endemicity
Amongst the 169 species recorded, 45 species were common to the Buruli ulcer sites with high endemicity and sites with low endemicity; the most frequent species for the both sites were Centrosema pubescens, Panicum maximum, Panicum laxum, Commelina diffusa, Schrankia leptocarpa and Leersia hexandra. In the two sites of high endemicity, the floristic richness around water sources was 85. There were 40 species exclusively related to sites of high endemicity. Amongst these 40 plant species, the most recurrent were Paspalum distichum, Persicaria senegalensis, Calopogonium mucunoides, Echinochloa colona, Ipomoea sagittata and Eichhornia crassipes. A total number of 129 species were identified in sites of low endemicity. Amongst these species, 84 were identified exclusively in site of low endemicity. Species which were specific to sites of low Buruli ulcer endemicity were, for example, Phaseolus lunatus, Cyathula prostrata, Dioscorea minutiflora, Mucuna pruriens and Paspalum conjugatum (Table 1, Table S1). The species richness was more important in site of low endemicity than in site of high endemicity.Table 1 Characteristics of the vegetation in sites of low and high Buruli ulcer endemicity
High BU endemicity sites Low BU endemicity sites
Family 28 39
Genera 67 107
Species richness 85 129
Total number of individuals species observed 22 286 7 327
Number of quadrats installed (1 m × 1 m) 147 126
Sites of high endemicity were characterised by the dominance of Onagraceae. However, Acanthaceae was an important family in sites of low endemicity. Amid the most representative family near water body point, Polygonaceae was encountered only in sites of high endemicity. Vitaceae (four species), Maranthaceae (three species), Arecaceae (two species), Urticaceae (three species), Dioscoreaceae (two species) and Menispermaceae (two species) were recorded only in sites of low endemicity (Tables 2, 3).Table 2 Most represented families around water body points in sites of high and low Buruli ulcer endemicity
Families High endemicity Low endemicity
Acanthaceae 1 6
Amaranthaceae 1 3
Apocynaceae 2 4
Asteraceae 5 6
Combretaceae 1 1
Commelinaceae 2 2
Convolvulaceae 3 4
Cucurbitaceae 1 1
Cyperaceae 12 9
Euphorbiaceae 4 6
Fabaceae 9 18
Malvaceae 3 3
Onagraceae 4 2
Passifloraceae 1 2
Phyllanthaceae 1 3
Poaceae 20 22
Rubiaceae 1 4
Solanaceae 3 2
Sphenocleaceae 1 1
Thelypteridaceae 1 1
Table 3 Total number of species per family and relative proportion of species in areas of low and high Buruli ulcer endemicity
Families Total number of Species High endemicity (%) Low endemicity (%)
Poaceae 30 66.67 73.33
Fabaceae 22 40.91 81.82
Cyperaceae 15 80 60
Asteraceae 7 71.43 85.71
Euphorbiaceae 7 57.14 85.71
Acanthaceae 6 16.67 100
Apocynaceae 6 33.33 66.67
Convolvulaceae 6 50 66.67
Malvaceae 5 60 60
Onagraceae 5 80 40
Rubiaceae 4 25 100
Solanaceae 4 75 50
Vitaceae 4 0 100
Amaranthaceae 3 33.33 100
Marantaceae 3 0 100
Phyllanthaceae 3 33.33 100
Urticaceae 3 0 100
Arecaceae 2 0 100
Commelinaceae 2 100 100
Dioscoreaceae 2 0 100
Menispermaceae 2 0 100
Passifloraceae 2 50 100
Polygonaceae 2 100 0
The genera represented in sites of high Buruli ulcer endemicity were Cyperus (five species), followed by Ludwigia (four species), Digitaria (three species), Ipomoea (three species), Setaria (three species), Alchornea (two species), Commelina (two species), Panicum (two species), Physalis (two species) and Senna (two species). The other genera are monospecific. In area of low Buruli ulcer endemicity, the richest genus was Desmodium (four species), followed by Ipomoea (three species), Panicum (three species) and Setaria (three species). Thirteen genera had two species and, the ninety remaining genera were monospecific.
Life Forms Spectrum
In all sites inventoried, 20 life forms were obtained observed. In sites of low endemicity, 18 life forms were identified whilst 14 in high Buruli ulcer areas. There was no difference in the proportion of life form in the different sites (Fisher's exact test, P = 0.18). In sites of low endemicity, high proportions of Chamaephytes (10.8%) and climbing Microphanerophytes (16.28%) were found. These life forms were less represented in high Buruli ulcer sites (Fig. 3). These differences are not statistically different (P = 0.1).Fig. 3 Frequency of life forms in site of low (proplow) and high (prophigh) Buruli ulcer endemicity
For the whole set of data, Nanophanerophytes were the most abundant life form and constituted 23.08%. Therophytes (18.34%) were the second most dominant, followed by Microphanerophytes climbers (Lmp, 14.20%) and Hydrophytic (11.24%). The others genera represent less than 7%. Geophytes (2.33%), Hemicryptophytes-epiphytes (0.78%), Lianas (0.78%), Megaphanerophytes (0.78%), Semi-epiphytes (0.78%) and Stoloniferous (0.78%) were observed only in sites of low endemicity. Hydrophytes (1.18%) and Rhizomatous-Stoloniferous (1.18%) were recorded only in sites of high endemicity.
In sites of low endemicity, 18 life forms were identified whilst 14 in high Buruli ulcer areas. There was no difference in the proportion of life form between the different sites (Fisher's exact test, P = 0.18).
In sites of low endemicity, high proportions of Chamaephytes (10.8%) and climbing Microphanerophytes (16.28%) were found. These life forms were less represented in high Buruli ulcer sites (Fig. 3). These differences are not statistically different (P = 0.1).
Chorological Spectrum
The phytogeographical types recorded were composed of species from the Guineo-Congolian and Sudano-Zambezian chorological transition type (47.93%), Guineo-Congolian chorological (38.46%), introduced species (10.06%), Sudano-Zambezian chorotype elements (2.37%), and western Guineo-Ccongolian species (1.18%) (Table 4).Table 4 Frequency of chorological origins of species around water bodies
Chorological types Number of species Percentage (%)
GC-SZ 81 47.93
GC 65 38.46
i 17 10.06
SZ 4 2.37
GCW 2 1.18
Total 169 100
GC-ZS Guineo-Congolian-Sudano-Zambezian, GC Guineo-Congolian, SZ Sudano-Zambezian and Sudanian (S), i Introduced species, SZ Sudano-Zambezian and GCW endemic of West Africa species
Species from Guineo-Congolian and Sudano-Zambezian region were abundant in all sites. They represented 42.64% and 31.40% and 47.29% and 31.76% of all the chorological origins, respectively, in areas of high and low Buruli ulcer endemicity. However, proportions of species from Guineo-Congolian and Sudano-Zambezian regions were closed to those of Guineo-Congolian chorological region in low Buruli ulcer area. Species from the endemic West Africa region were recorded only in site of low Buruli ulcer endemicity, and species from Sudano-Zambezian were recorded only in high endemic Buruli ulcer areas. For all the health districts, there was a high rate of introduced species in sites of high endemicity (10.59%) contrary to sites of low endemicity (8.53%) (Fig. 4).Fig. 4 Chorological spectrum according to the status of studied sites (GCW: endemic of West Africa species; SZ Sudano-Zambezian region; I Introduced species; GC Guineo-Congolese region; GC-SZ common to Guineo-Congolese and Sudano-Zambezian region)
There was a significant difference between chorological origin of species of low and high Buruli ulcer endemicity area (Fisher's exact test, P = 0.04). Proportion of species chorological origin change according to the Buruli ulcer endemicity status.
Floristic Diversity of Water Body Points
The Shannon diversity index varied from 0.853 to 1.904 for high Buruli ulcer sites. In low Buruli ulcer sites, this index ranged between 2.897 and 3.087. The equitability indices were lower in sites of high endemicity (0.432 ± 0.06) than in low endemicity (0.731 ± 0.017) areas. The Simpson index varied from 0.301 to 0.724 in high Buruli ulcer sites, and from 0.892 to 0.926 in sites of low endemicity. For all the diversity indices, there was a significant difference between Shannon, Pielou and Simpson indices in sites of high and low endemicity (p < 0.05) as reported in Table 5.Table 5 Floristic diversity indices for high and low endemicity sites (mean ± SD, n = 4)
Endemicity status Site Shan Simp J
High Batéguédia_1 1.904 0.724 0.523
Batéguédia_2 1.716 0.647 0.510
Konankro_1 1.521 0.545 0.452
Konankro_2 0.853 0.301 0.244
1.498 ± 0.229 0.554 ± 0.092 0.432 ± 0.065
Low Kongodékro_1 2.897 0.910 0.710
Kongodékro_2 2.645 0.893 0.695
Zébra_1 2.982 0.926 0.758
Zébra_2 3.087 0.909 0.760
2.903 ± 0.094 0.909 ± 0.007 0.731 ± 0.017
P value 0.03
Shan Shannon; simp Simpson index; J Pielou equitability index; P value difference between the two groups
Floristic Similarity of Sites of Low and High Buruli Ulcer Endemicity
According to Sorensen (0.420), Jaccard (0.266) and Whittaker (0.579) indices calculated, a weak occurrence of species was observed between the sites of high and low endemicity. Diversity profiles of high endemic communities (BATEUEDIA 1 and 2; KONANKRO 1 and 2) are totally different from those of low endemicity (ZEBRA 1 and 2; KONGODEKRO 1 and 2), their curves are below those of site of low endemicity (Fig. 5).Fig. 5 Alpha diversity profiles of communities
NMDS analysis showed that the sites could be gathered (NMDS stress = 0.056). The studied sites are different according to their floristic compositions. In the NMDS graph, the more the study sites are aggregated, the more the floristic compositions are similar (Fig. 6a). The Shepard plot showed a strong correlation between the observation and the ordination distance (R2 > 0.95), which reveals a good fit obtained in the NMDS indicating that similarity in the data is well caught with the NMDS ordination method (Fig. 6b).Fig. 6 a Result of the NMDS ordination of the sites of high and low Buruli ulcer endemicity. b Shepard plot of the NMDS
Indicator Species in Site of High and Low Endemicity
The analysis of indicator species revealed that 38 species were indicators for sites of low endemicity, whilst 12 were related to sites of high endemicity. High endemicity sites were mainly characterised by Leersia hexandra, Mimosa pudica, Paspalum distichum, Schrankia leptocarpa, Persicaria senegalensis, Calopogonium mucunoides, Echinochloa colona, Ipomoea sagittata, Eichhornia crassipes and Imperata cylindrica. Site of low Buruli ulcer endemicity were significantly associated with Centrosema pubescens, Panicum maximum, Setaria barbata, Rottboellia cochinchinensis, Pennisetum purpureum and Synedrella nodiflora (Table 6).Table 6 Most important indicator species of areas of low and high Buruli ulcer endemicity in the sanitary districts of Bouaké and Daloa
Species Group Indval (%) P value freq
Leersia hexandra High 30.85 0.001 63
Mimosa pudica High 21.68 0.001 40
Paspalum distichum High 20.41 0.001 30
Schrankia leptocarpa High 18.94 0.019 64
Persicaria senegalensis High 14.29 0.001 21
Calopogonium mucunoides High 11.56 0.002 17
Echinochloa colona High 10.88 0.001 16
Ipomoea sagittata High 10.2 0.001 15
Eichhornia crassipes High 8.16 0.001 12
Imperata cylindrica High 6.12 0.004 9
Centrosema pubescens Low 45.66 0.001 139
Panicum maximum Low 45.49 0.001 102
Setaria barbata Low 19.98 0.001 27
Rottboellia cochinchinensis Low 16.99 0.001 41
Pennisetum purpureum Low 16.63 0.003 42
Synedrella nodiflora Low 16.2 0.001 24
Cyclosorus oppositifolius Low 14.41 0.002 33
Phaseolus lunatus Low 11.9 0.001 15
Pueraria phaseoloides Low 11.45 0.012 29
Cyathula prostrata Low 7.94 0.001 10
Dioscorea minutiflora Low 7.94 0.001 10
Mucuna pruriens Low 7.94 0.003 10
Paspalum conjugatum Low 7.94 0.001 10
Laportea ovalifolia Low 7.14 0.003 9
Panicum brevifolium Low 7.14 0.001 9
Eleusine indica Low 6.78 0.013 12
Pergularia daemia Low 6.35 0.003 8
Phaulopsis ciliata Low 6.35 0.003 8
Cissus producta Low 5.56 0.005 7
Cynodon dactylon Low 5.56 0.002 7
Pouzolzia guineensis Low 5.56 0.005 7
Discussion
In this study, the floristic composition was studied around water body points in areas of low and high Buruli ulcer endemicity. The inventories showed that floristic diversity in low Buruli ulcer was higher than in sites of high Buruli ulcer endemicity. The indices of similarity between these sites revealed a low similarity. To the best of our knowledge, this is the first attempt to report the floristic composition of water points used by people in sites of Buruli ulcer in Côte d’Ivoire. The present study was more exhaustive and recorded more species and families than the one carried out in Ghana (McIntosh et al. 2014). In total, 129 species were inventoried in sites of low endemicity and 85 in sites of high endemicity.
In high Buruli ulcer sites, one of the indicator species was Leersia hexandra. This species was also recorded in Ghana by McIntosh et al. (2014). Theses authors found that genes from M. ulcerans were detected on species belonging the genera Leersia, Struchium, Cyperus and Ludwigia. These taxa were also recorded in the present study. These results may strengthen the role of these plants in the maintenance of M. ulcerans in the environment.
The family of Polygonaceae was amongst the most common families in Buruli ulcer high status sites. Polygonaceae species are known for their richness in cells containing calcium oxalate crystal (Borrelli et al. 2011). Calcium is a mineral required for the growth of M. ulcerans (Sanhueza et al. 2016). Moreover, calcium oxalate crystals have irritant properties and could facilitate the intrusion of the microorganism in the body (Bolognia et al. 2003). The presence of plants of this family may explain the highest endemicity and be a co-occurrence indication for predicting high endemicity in an area.
An abundance of Fabaceae was observed in low Buruli ulcer endemicity site. The plants belonging to this family fix atmospheric nitrogen in their environment throughout different mechanisms. Nitrogen is an inhibitor of the growth M. ulcerans (Phillips et al. 2004). Thus, their presence may be linked to the low endemicity area sites.
A high proportion of Hemicryptophytes, which is a characteristic of opened vegetation physiognomy, was recorded in sites with high endemicity of Buruli ulcer. According to Raunkiaer (1934) and Raunkiaer (1934), Hemicryptophytes are favourable to wet environments. They grow in disturbed environment (Khan et al. 2018) and their abundance may be an indication of Buruli ulcer occurrence in a site. As Hemicryptophytes, M. ulcerans also flourish in wet and disturbed environments (Röltgen and Pluschke 2015). The presence of Chamaephytes may also be an indicator, as their presence indicates an environmental disturbance typical of sites in which M. ulcerans can survive (Jauffret 2001; Jauffret and Lavorel 2003). Therophytes were abundant in sites of high endemicity and their presence reveals poor levels of macronutrients and disturbed environments by anthropogenic activities (Rashid et al. 2011). These conditions may provide habitat for the bacterium. The sites of low endemicity were close to dense landscapes due to the abundance of Phanerophytes (Batalha and Martins 2004).
The chorological origin of the species varied from one status to another. The high proportion of GC and GC-SZ species revealed that the study sites were wet (Soulé et al. 2016). In addition, there was a relative proportion of introduced species in high Buruli ulcer region. When analysing phytosociological index (IP) it was noticed that the flora of high endemicity was closed to Sudanian vegetation, which is similar to open vegetation (Aké-Assi 2001, 2002). This finding may explain the high prevalence of hyper endemicity zones. Open vegetation is favourable to the growth of M. ulcerans (Combe et al. 2017).
The flora around water bodies possesses an important equitability especially in sites of low endemicity. These results showed that species were more equally distributed in low Buruli ulcer sites than in high Buruli ulcer endemicity sites. Regarding diversity profiles, low endemic sites curved are above those of high endemicity. This result indicates a low diversity of sites of high endemicity at all orders. The low diversity indices that were obtained in high Buruli ulcer areas may result from human activities such as agriculture, fishing, washing the dishes and washing near the water body point, as observed during inventories. These practices may lead to reducing oxygen availability. M. ulcerans growth is positively associated with eutrophication and low oxygen availability (Merritt et al. 2010). Most of the disturbed areas are distinguished by low biodiversity indices (Bornette and Puijalon 2011; Buhk et al. 2017). The low diversity indices that were obtained in this study showed that high Buruli ulcer sites were disturbed and favourable to the persistence of the mycobacteria in the environment. Most of the indicator species were introduced species. This finding confirmed the studied sites suffered from severe disturbance. Moreover, Persicaria and Eichhornia genera are indicators of a deteriorated environment (Ngonyani and Nkotagu 2007).
The morphological characteristics of some plants could explain their involvement in the transmission of the disease. Leersia hexandra, Imperata cylindrica and Pannicum laxum looked sharp. Thus, these plants may injure people when having daily activities in grasslands near ponds. Many studies reported the presence of colonies of M. ulcerans on species belonging to Imperata, Pannicum, Pennissetum and Cyperus genera (Choyce 1970; Pearson 1999; Portaels et al. 1999). In Ghana, M. ulcerans was detected on species belonging to Ludwigia, Leersia, Struchium and Commelina genera (McIntosh et al. 2014). These genera were harvested around water body points in the present study. A search of M. ulcerans is important to discriminate those which are involved in the transmission of Buruli ulcer.
Conclusion
Buruli ulcer is consider as a destroyable disease having health and social strong impact on people. Mode of transmission is still been on elucidation. Investigating ecological environment, through botanical inventory, on site of low and high endemicity of Buruli ulcer, a total of 169 plants species were inventoried from all the status sites. The species richness and floristic diversity were low in sites of high endemicity in comparison to low endemic sites. The indicator species of sites of high endemicity were characteristic of open and disturbed environments, which are favourable to the growth of M. ulcerans. These botanical characteristics of the studied sites may probably explain the endemicity of Buruli ulcer in Côte d’Ivoire. The search of M. ulcerans on dominant plants needs to be carried out. We suggest strengthening monitoring of biodiversity in high Buruli ulcer areas.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 20 kb)
Acknowledgements
The authors thank the Programme National de Lutte contre l’Ulcère de Buruli (PNLUB) for availability of data related to the Buruli ulcer cases in Côte d’Ivoire; thanks are also due to Mrs. Maggie Gamberton and M. Robert Brand for proofreading of the manuscript.
Author Contributions
For research articles with several authors, a short paragraph specifying their individual contributions must be provided. Conceptualization and design, AB, EE, MWK, CB, MD and DS. Material preparation, analysis and data collection, EE and DS; software, EE; validation, AB, EE, MWK, ICZ-B and DS; investigation, EE and DS; data curation, EE and DS, ICZ-B, and AFK; writing—original draft preparation, EE and MWK; writing—review and editing, EE, DS, ICZ-B and MWK; visualization, EE; supervision, AB and MWK; project administration, DS; funding acquisition, DS. All authors have read and agreed to the published version of the manuscript.
Funding
This research was financially supported by Programme d’Appui Stratégique à la Recherche Scientifique (PASRES) (Project N° 158/2015).
Data Availability
The data presented in this study are available on request from the corresponding author.
Declarations
Conflicts of Interest
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
==== Refs
References
Adomou AC Akoègninou A Sinsin B de Foucault B van der Maesen LJG Notulae florae Beninensis, 13—biogeographical analysis of the vegetation in Benin Acta Botanica Gallica 2007 154 2 221 233 10.1080/12538078.2007.10516053
Aké-Assi L (2001) Flore de la Côte d’Ivoire: catalogue systématique, biogéographie et écologie (Tome 1). Boissiera, vol 57. Conservatoire et Jardin Botaniques de Genève, Génève, Suisse
Aké-Assi L (2002) Flore de la Côte d’Ivoire: catalogue systématique, biogéographie et écologie (Tome 2). Boissiera, vol 58. Conservatoire et Jardin Botaniques de Genève, Génève, Suisse
Akpagana K (1989) Recherche sur les forêts denses humides du Togo. [France]: Doctoral Thesis, Université de Bordeaux 3, Bordeaux, France
Baselga A Orme CDL Betapart: an R package for the study of beta diversity Methods Ecol Evol 2012 3 5 808 812 10.1111/j.2041-210X.2012.00224.x
Batalha MA Martins FR Floristic, frequency, and vegetation life-form spectra of a cerrado site Braz J Biol 2004 64 2 201 209 10.1590/S1519-69842004000200004
Bolognia JL Jorizzo JL Rapini RP Dermatology 2003 London Mosby
Bornette G Puijalon S Response of aquatic plants to abiotic factors: a review Aquat Sci 2011 73 1 1 14 10.1007/s00027-010-0162-7
Borrelli N Fernández Honaine M Altamirano SM Osterrieth M Calcium and silica biomineralizations in leaves of eleven aquatic species of the Pampean Plain Argentina Aquat Bot 2011 94 1 29 36 10.1016/J.AQUABOT.2010.10.003
Buhk C Alt M Steinbauer MJ Beierkuhnlein C Warren SD Jentsch A Homogenizing and diversifying effects of intensive agricultural land-use on plant species beta diversity in central Europe — A call to adapt our conservation measures Sci Total Environ 2017 576 225 233 10.1016/J.SCITOTENV.2016.10.106 27788437
Chase MW Christenhusz MJM Fay MF Byng JW Judd WS Soltis DE Mabberley DJ Sennikov AN Soltis PS Stevens PF An update of the angiosperm phylogeny group classification for the orders and families of flowering plants: APG IV Bot J Linn Soc 2016 181 1 1 20 10.1111/boj.12385
Chatelain C Aké-Assi L Spichiger RE Gautier L Cartes de distribution des plantes de Côte d’Ivoire. Boissiera 2011 Genève, Suisse Conservatoire et Jardin Botanique Ville de Genève
Choyce DP Transmission of Buruli disease Br Med J 1970 4 5734 558 10.1136/bmj.4.5734.558-b
Combe M Velvin CJ Morris A Garchitorena A Carolan K Sanhueza D Roche B Couppié P Guégan JF Gozlan RE Global and local environmental changes as drivers of Buruli ulcer emergence Emerg Microbes Infect 2017 6 4 e21 10.1038/emi.2017.7 28442755
De Cáceres M Legendre P Moretti M Improving indicator species analysis by combining groups of sites Oikos 2010 119 10 1674 1684 10.1111/j.1600-0706.2010.18334.x
De Magalhaes TL Da Costa Bortoluzzi RL Mantovani A Plant distribution in freshwater wetlands of the Brazilian subtropical highland grasslands Brazilian J Botany 2016 39 1 239 249 10.1007/s40415-015-0226-y
Debacker M Aguiar J Steunou C Zinsou C Meyers WM Scott JT Dramaix M Portaels F Mycobacterium ulcerans disease: role of age and gender in incidence and morbidity Tropical Med Int Health 2004 9 12 1297 1304 10.1111/j.1365-3156.2004.01339.x
Dharmarajan G Gupta P Vishnudas CK Robin VV Anthropogenic disturbance favours generalist over specialist parasites in bird communities: implications for risk of disease emergence Ecol Lett 2021 24 9 1859 1868 10.1111/ele.13818 34120404
Dufrêne M Legendre P Species assemblages and indicator species: the need for a flexible asymmetrical approach Ecol Monogr 1997 67 3 345 366 10.1890/0012-9615(1997)067[0345:SAAIST]2.0.CO;2
EPA Methods for evaluating wetland condition: vegetation-based indicators of wetland nutrient enrichment 2002 Washington DC, USA Office of Water, U.S. Environmental Protection Agency
Falkinham JO Environmental sources of nontuberculous mycobacteria Clin Chest Med 2015 36 1 35 41 10.1016/J.CCM.2014.10.003 25676517
FetchClimate. 2016. FetchClimate [Internet]. [accessed 2016 Nov 11]. http://fetchclimate2.cloudapp.net
Fisher MJ Marshall AP Mitchell M Testing differences in proportions Aust Crit Care 2011 24 2 133 138 10.1016/J.AUCC.2011.01.005 21536451
Garchitorena A Ngonghala CN Guegan JF Texier G Bellanger M Bonds M Roche B Economic inequality caused by feedbacks between poverty and the dynamics of a rare tropical disease: the case of Buruli ulcer in sub-Saharan Africa Proc Biol Sci 2015 282 1818 20151426 10.1098/rspb.2015.1426 26538592
Helm N, Essl F, Mirtl M, Dirnböck T. 2017. Multiple environmental changes drive forest floor vegetation in a temperate mountain forest. Ecol Evol. 7(7):2155–2168. 10.1002/ece3.2801
Hotez PJ Kamath A Neglected tropical diseases in Sub-Saharan Africa: review of their prevalence, distribution, and disease burden PLoS Negl Trop Dis 2009 3 8 e412 10.1371/journal.pntd.0000412 19707588
Jauffret S (2001) Validation et comparaison de divers indicateurs des changements à long term dans les écosystèmes méditerranéens arides: application au suivi de la désertification dans le Sud tunisien. [France]: Aix-Marseille 3
Jauffret S Lavorel S Are plant functional types relevant to describe degradation in arid, southern Tunisian steppes? J Veg Sci 2003 14 3 399 408 10.1111/j.1654-1103.2003.tb02165.x
Keesing F Ostfeld RS Impacts of biodiversity and biodiversity loss on zoonotic diseases Proc Natl Acad Sci U S A 2021 118 17 1 8 10.1073/PNAS.2023540118
Khan W Khan SM Ahmad H Alqarawi AA Shah GM Hussain M Abd-Allah EF Life forms, leaf size spectra, regeneration capacity and diversity of plant species grown in the thandiani forests, district abbottabad, khyber pakhtunkhwa Pakistan Saudi J Biol Sci 2018 25 1 94 100 10.1016/J.SJBS.2016.11.009 29379363
Khanna R Meena RN Mukunda MV Khanna S Infective Leg Ulcers Ulcers of the lower extremity 2016 New Delhi, India Springer 267 274
Koka E Yeboah-Manu D Okyere D Adongo PB Ahorlu CK Cultural understanding of wounds, buruli ulcers and their management at the obom sub-district of the ga south municipality of the greater accra region of ghana PLoS Negl Trop Dis 2016 10 7 e0004825 10.1371/journal.pntd.0004825 27438292
Manlove KR Walker JG Craft ME Huyvaert KP Joseph MB Miller RS Nol P Patyk KA O’Brien D Walsh DP Cross PC “One health” or three? Publication silos among the one health disciplines PLoS Biol 2016 14 4 e1002448 10.1371/journal.pbio.1002448 27100532
McIntosh M Williamson H Benbow ME Kimbirauskas R Quaye C Boakye D Small P Merritt R Associations between Mycobacterium ulcerans and aquatic plant communities of west africa: implications for buruli ulcer disease EcoHealth 2014 11 2 184 196 10.1007/s10393-013-0898-3 24442959
Merritt RW Walker ED Small PLC Wallace JR Johnson PDR Benbow ME Boakye DA Ecology and transmission of buruli ulcer disease: a systematic review PLoS Negl Trop Dis 2010 4 12 1 15 10.1371/journal.pntd.0000911
Narh CA Mosi L Quaye C Dassi C Konan DO Tay SCK de Souza DK Boakye DA Bonfoh B Source tracking Mycobacterium ulcerans infections in the ashanti region Ghana. Plos Negl Trop Dis. 2015 9 1 1 18 10.1371/journal.pntd.0003437
Ngonyani C Nkotagu H Study of nutrient pollutants and their impacts on the water quality of the Mindu reservoir at Morogoro municipality Tanzania J Eng Technol (TJET) 2007 1 138 148
Oksanen J Blanchet FG Friendly M Kindt R Legendre P McGlinn D Minchin PR O’Hara RB Simpson GL Solymos P Vegan: community ecology package R Package Version 2017 2 1 295
Olleck M Reger B Ewald J Plant indicators for folic histosols in mountain forests of the calcareous alps Appl Veg Sci 2020 23 2 285 296 10.1111/avsc.12470
Ostfeld RS Biodiversity loss and the ecology of infectious disease Lancet Planet Health 2017 1 1 2 3 10.1016/S2542-5196(17)30010-4
Pearson AL. 1999. Habit Favorability and Geographical Links Between Mycobacterium ulcerans and Eichhornia crasspies in the Amansie West District, Ghana: Social, Economic, and Health Implications. [place unknown]: University of North Texas. Honors College Thesis
Phillips R Adjei O Lucas S Benjamin N Wansbrough-Jones M Pilot randomized double-blind trial of treatment of Mycobacterium ulcerans disease (Buruli Ulcer) with topical nitrogen oxides The Antimicrob Agents Chemother. 2004 48 8 2866 2870 10.1128/AAC.48.8.2866-2870.2004 15273093
Portaels F Elsen P Guimaraes-Peres A Fonteyne P-A Meyers WM Insects in the transmission of Mycobacterium ulcerans infection The Lancet 1999 353 9157 986 10.1016/S0140-6736(98)05177-0
R Core Team. 2017. R: A language and environment for statistical computing 3.3.0.
Rashid A Swati MF Sher H Al-Yemeni MN Phytoecological evaluation with detail floristic appraisal of the vegetation arround Malam Jabba, Swat Pakistan Asian Pac J Trop Biomed 2011 1 6 461 467 10.1016/S2221-1691(11)60101-9 23569814
Raunkiaer C The life forms of plants and statistical plant geography : oxford at the clarendon press : free download, borrow, and streaming : internet archive 1934 Oxford, UK Claredon Press
Roberts DW David M Roberts W Package “labdsv” Ordinat Multivariate Anal Ecol 2016 775 1 68
Röltgen K Pluschke G Mycobacterium ulcerans disease (Buruli Ulcer): potential reservoirs and vectors Curr Clin Microbiol Rep 2015 2 1 35 43 10.1007/s40588-015-0013-3
RStudio Team. 2016. RStudio: Integrated Development for R
Sanhueza D Chevillon C Colwell R Babonneau J Marion E Marsollier L Guégan J-F Chitin promotes Mycobacterium ulcerans growth FEMS Microbiol Ecol 2016 92 6 067 10.1093/femsec/fiw067
Sopoh G Asiedu K Buruli ulcer in Sub-Saharan Africa Neglected tropical diseases-Sub-Saharan Africa 2016 Switzerland, Gewerbestrasse Springer International Publishing 15 43
Soulé M Bassirou I Matalabi A Mahamane S Systematic composition, life forms and chorology of parklands of commune of Mayahi, Niger West Africa Asian J Biol 2016 1 1 1 11 10.9734/AJOB/2016/30473
Tian RBD Niamké S Tissot-Dupont H Drancourt M Detection of Mycobacterium ulcerans DNA in the environment Ivory Coast PLoS One. 2016 11 3 e0151567 10.1371/journal.pone.0151567 26982581
Webster JP Gower CM Knowles SCL Molyneux DH Fenton A One health—an ecological and evolutionary framework for tackling neglected zoonotic diseases Evol Appl 2016 9 2 313 333 10.1111/eva.12341 26834828
White F La végétation de l’Afrique : mémoire accompagnant la carte de végétation de l’Afrique 1986 Paris, France Orstom-Unesco
WHO. 2017a. Buruli ulcer: overview. Health topics [Internet]. [Accessed 2018 Sep 29]. https://afro.who.int/health-topics/buruli-ulcer
WHO. 2017b. Global Health Observatory data repository : Number of new reported cases Data by country. WHO [Internet]. [Accessed 2018 Sep 27]. http://apps.who.int/gho/data/node.main.A1631
WHO. 2018. GHO | By category | Number of new reported cases - Data by country. WHO [Internet]. [Accessed 2018 Dec 6]. http://apps.who.int/gho/data/node.main.A1631
Young HS Parker IM Gilbert GS Sofia Guerra A Nunn CL Introduced species, disease ecology, and biodiversity-disease relationships Trends Ecol Evol 2017 32 1 41 54 10.1016/j.tree.2016.09.008 28029377
|
PMC010xxxxxx/PMC10131284.txt |
==== Front
Med
Med
Med (New York, N.y.)
2666-6359
2666-6340
The Author(s). Published by Elsevier Inc.
S2666-6340(23)00104-6
10.1016/j.medj.2023.04.001
Case Report
RNase2 is a possible trigger of acute-on-chronic inflammation leading to mRNA vaccine-associated cardiac complication
Ong Eugenia Z. 12
Koh Clara W.T. 1
Tng Danny J.H. 13
Ooi Justin S.G. 1
Yee Jia Xin 12
Chew Valerie S.Y. 12
Leong Yan Shan 12
Gunasegaran Kurugulasigamoney 4
Yeo Chin Pin 5
Oon Lynette L.E. 5
Sim Jean X.Y. 3
Chan Kuan Rong 1∗
Low Jenny G. 123∗∗
Ooi Eng Eong 1267∗∗∗
1 Program in Emerging Infectious Diseases, Duke-NUS Medical School, Singapore, Singapore
2 Viral Research and Experimental Medicine Centre, SingHealth Duke-NUS Academic Medical Centre, Singapore, Singapore
3 Department of Infectious Diseases, Singapore General Hospital, Singapore, Singapore
4 National Heart Centre, Singapore, Singapore
5 Department of Pathology, Singapore General Hospital, Singapore, Singapore
6 Saw Swee Hock School of Public Health, National University of Singapore, Singapore, Singapore
∗ Corresponding author
∗∗ Corresponding author
∗∗∗ Corresponding author
7 Lead contact
26 4 2023
9 6 2023
26 4 2023
4 6 353360.e2
30 11 2022
26 2 2023
3 4 2023
© 2023 The Author(s)
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Post-mRNA vaccination-associated cardiac complication is a rare but life-threatening adverse event. Its risk has been well balanced by the benefit of vaccination-induced protection against severe COVID-19. As the rate of severe COVID-19 has consequently declined, future booster vaccination to sustain immunity, especially against infection with new severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants, may encounter benefit-risk ratios that are less favorable than at the start of the COVID-19 vaccination campaign. Understanding the pathogenesis of rare but severe vaccine-associated adverse events to minimize its risk is thus urgent.
Methods
Here, we report a serendipitous finding of a case of cardiac complication following a third shot of COVID-19 mRNA vaccine. As this case was enrolled in a cohort study, pre-vaccination and pre-symptomatic blood samples were available for genomic and multiplex cytokine analyses.
Findings
These analyses revealed the presence of subclinical chronic inflammation, with an elevated expression of RNASE2 at pre-booster baseline as a possible trigger of an acute-on-chronic inflammation that resulted in the cardiac complication. RNASE2 encodes for the ribonuclease RNase2, which cleaves RNA at the 3′ side of uridine, which may thus remove the only Toll-like receptor (TLR)-avoidance safety feature of current mRNA vaccines.
Conclusions
These pre-booster and pre-symptomatic gene and cytokine expression data provide unique insights into the possible pathogenesis of vaccine-associated cardiac complication and suggest the incorporation of additional nucleoside modification for an added safety margin.
Funding
This work was funded by the NMRC Open Fund-Large Collaborative Grant on Integrated Innovations on Infectious Diseases (OFLCG19May-0034).
Graphical abstract
Context and significance
Cardiac complication is a rare but life-threatening adverse event following mRNA vaccination. Although inflammation has been found to be associated with such an adverse outcome, what triggers the inflammatory response is uncertain. Ong et al. found, in a participant in a cohort study, that high pre-booster baseline expression of RNASE2 in a background of subclinical chronic inflammation was the possible underpinning of cardiac complication. RNASE2 encodes for the ribonuclease RNase2, which cleaves RNA at the 3′ side of uridine, thus removing a TLR8-avoidance safety feature of mRNA vaccines. Activation of the TLR8 pathway triggered an acute-on-chronic inflammatory response that likely contributed to the post-booster cardiac event. Their finding suggests the incorporation of additional nucleoside modification for an added safety margin.
Ong et al. show that the presence of subclinical chronic inflammation and elevated baseline expression of RNASE2 are possible triggers for post-vaccination cardiac complication. Their findings provide insights into the pathogenesis of cardiac complication and suggest the incorporation of additional nucleoside modifications in mRNA vaccines for an added safety margin.
Keywords
COVID-19
mRNA vaccination
severe adverse event
cardiac complication
RNASE2
Published: May 12, 2023
==== Body
pmcIntroduction
Vaccines are useful tools in any public health program by eliciting adaptive immunity to prevent infectious diseases. They emphatically turned the tide against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and the COVID-19 pandemic. Our immune system is, however, a double-edged sword. When over-activated, it can damage our tissues and organs, leading to autoimmune conditions. The development and application of any vaccine are thus based on unequivocal demonstration of benefits over risks of vaccination.
The most widely applied COVID-19 vaccine has been the mRNA vaccine. When inoculated into vaccinees, mRNA is translated into the spike protein of SARS-CoV-2 to produce antibodies and CD4 and CD8 T cells soon after vaccination.1 The elicited cellular immunity may be critical to prevent severe COVID-19, even against SARS-CoV-2 variants that escape vaccine-elicited neutralizing antibodies.2 To better protect vulnerable individuals from COVID-19, booster mRNA vaccination using a more contemporaneous spike gene sequence has begun. Moreover, newer vaccines are also under development to expand the breadth of protection against other SARS-CoV-2-like Beta CoVs. However, as the majority of the global population is no longer immunologically naive against SARS-CoV-2 from either vaccination or infection, the risk of severe adverse events from mRNA vaccination, even if rare, may blunt the previously unequivocal benefit-risk ratio. Insights that can shape the development of even safer mRNA vaccines are thus highly desirable since mRNA vaccines will undoubtedly become a prominent feature in any public health toolbox.
Among the rare but severe adverse events following mRNA vaccination are those that affect the heart. The rate of post-mRNA vaccination myocarditis or pericarditis is estimated to be about 2.17 per 100,000 person days, with the risk being higher in younger males following the second vaccination.3 These inflammatory conditions manifest symptomatically from 1 to 7 days after vaccination3 and, although transient, can be both debilitating and alarming. That such events are rare makes it difficult for us to study the pathogenesis of post-vaccination cardiac complications. We hence believe that this case, a person who enrolled in our cohort of COVID-19 vaccinees, provides opportunistic insights into the susceptibility factors of post-mRNA vaccination cardiac complication.
Results
The subject in this case is a fit and healthy 42-year-old male with neither underlying illness nor allergy history. He received the first two primary series doses of BNT162b2 vaccination 8 months prior to the booster. The side effects reported following the first two doses of BNT162b2 were injection site tenderness and mild myalgia. He was well on day 1 when he was boosted with mRNA-1273. On day 2, he began experiencing shortness of breath upon exertion that continued to worsen until day 4, where he was unable to ambulate continuously without intermittent rest. Onset of exertion-induced chest pain occurred on day 5. When he presented to the cardiac unit on day 6, his electrocardiogram showed ST elevation in anterior leads with T wave inversion in V2, which is significant (Table S1). Full blood count was normal except for elevated eosinophils. He received an intravenous loading dose of 300 mg aspirin for suspicion of acute myocardial ischemia, although plasma levels of cardiac enzymes and troponin-T turned out normal. PCR for known etiologies of myocarditis—human herpes virus 6 (HHV6), cytomegalovirus (CMV), parvovirus B19, adenovirus, and enterovirus—yielded negative findings. He also tested negative for anti-nucleocapsid protein (NP) antibodies on day 1, ruling out the possibility of post-COVID-19 cardiac complication. His symptoms resolved by day 8; a cardiac echocardiography performed on this day was also unremarkable. He was discharged on day 8 without further intervention and subsequently returned to regular daily activities a week later. His clinical symptoms and course are summarized in Figure 1 A, and clinical details are provided in Table S1.Figure 1 Differential pre- and post-vaccination levels of inflammatory and anti-inflammatory cytokines in the subject compared with controls
(A) Outline of clinical symptoms and course.
(B) Partial least squares-discriminant analysis (PLS-DA) score plot showing clear clustering between subject and controls across all time points after booster mRNA vaccination. Components 1 and 2 account for 22% and 12% of the variance, with R2Y = 0.949, Q2Y = 0.802, and root mean square error of estimation (RMSEE) = 0.066. Measurements taken at days 1, 4, and 7/8.
(C) Loading biplot of the PLS-DA model.
(D–G) Expression levels of IL-4, IL-17A, IL-1B, and IL-27 in serum prior to doses 1 (pre-dose 1) and 2 (pre-dose 2), prior to booster (day 1), and at days 4 and 7/8 post-booster. Red lines and dots indicate data trends and points for subject. Box and whisker plots show the mean (center line of boxes), 25th and 75th percentiles (lower and upper boundary of boxes), and minimum and maximum values in the controls. n = 19 for controls and n = 1 for subject.
See also Table S1 and Figure S1.
As he had enrolled with written informed consent in our cohort, blood samples were collected before and at various time points after vaccination as previously described.1 As pre-vaccination baseline factors influence susceptibility to post-vaccination adverse events,4 , 5 we compared blood specimens from the subject against 18 controls from our cohort to identify host differences that may explain the observed cardiac complication. The controls were of similar age and also received booster vaccination but without cardiac complication. They also completed the per-protocol blood sampling schedule.
Plasma cytokine levels measured with the Olink Target 48 Cytokine panel were compared between the subject and controls using partial least squares-discriminant analysis (PLS-DA), which segregated the subject from controls at all 3 measured time points (Figure 1B). Plasma levels of 45 cytokines discriminated the subject from controls (Figure 1C), among which are the Th2 cytokines interleukin-4 (IL-4), the pro-inflammatory IL-17A, and IL1B and the inflammation-induced growth factors hepatocyte growth factor (HGF), oncostatin-m (OSM), and transforming growth factor α (TGF-α), as well as the scavenger receptor oxidized low-density lipoprotein receptor 1 (OLR1). These proteins were elevated in the plasma at baseline and at day 4 post-vaccination (Figures 1D–1F and S1A–S1D). Interestingly, the inhibitory IL-27 was low in the subject compared with controls until his symptoms resolved on day 8 post-vaccination (Figure 1G). Retrospective analysis of the pre-dose-1 and -2 blood samples also showed elevated pro-inflammatory and Th2 cytokines, with repressed expression of IL-27, collectively suggesting an underlying chronic but asymptomatic inflammation.
As the underlying inflammation did not trigger cardiac complication following the first two doses, we turned to genomic analysis to identify mRNA transcripts in our subject. We identified 92 genes in which the expression showed 1.5-fold difference and 3 standard deviations (SDs) or more from the mean of the controls to determine outliers that may explain this rare adverse event (Figure 2 A). Pathway analysis of the 61 genes with elevated expression clustered into innate immune pathways, while the 31 genes were in the metabolic pathway (Figure 2B), which is consistent with an activated immune system.6 Furthermore, gene set enrichment analysis using blood transcription modules found monocyte and immune activation signatures being most positively enriched in the subject compared with controls (Figure S2A). Other immune cells, such as neutrophils, were also positively enriched, although not to the levels seen in monocytes; eosinophil was not flagged in this analysis. Critically, this was unique to the sample taken before booster but not before doses 1 and 2 (Figures S2B and S2C). Likewise, RNASE2, which is known to be expressed in activated myeloid-derived mononuclear cells and eosinophils,7 was identified as a prominent outlier in the pre-vaccination gene expression data (Figure 2A). Strikingly, RNASE2 expression level was also an outlier in the subject compared with controls after booster vaccination when the cardiac complication arose; expression of RNASE2 was within the SDs of the controls before the first two doses of mRNA vaccine (Figure 2C).Figure 2 High expression of RNASE2 during booster vaccination was associated with cardiac complication
(A) MA plot showing average expression value of transcripts and log2 fold change (log2FC) of subject compared with controls at day 1. Genes with Z score >3 are colored, where red and blue indicate significant increase and decrease in expression, respectively, in the subject at day 1. Highlighted genes are those involved in either immune activation, Wnt signaling, cell division, or mitochondria metabolism pathways. Dotted line indicates FC >1.5.
(B) Enriched GOBP terms for transcripts with increased (red) or reduced (blue) expression in subject compared with controls at day 1.
(C) RNASE2 expression prior to doses 1 (pre-dose 1) and 2 (pre-dose 2), prior to booster (day 1), and at days 2, 4, and 8 post-booster.
(D) Number of differentially expressed genes (DEGs) in subject and controls (Z score > 3) across different FC cutoffs. Based on 1.5 FC, the top 10 GOBP pathways induced in the subject at day 4 (vs. day 1) are displayed. Dotted line indicates adjusted p value <0.05.
(E and F) Average Z score values of transcripts involved in pathways of defense response to virus and type I interferon responses in the subject and controls. The genes used for Z score calculations are shown in Figures S2O and S2P. Number of controls used for pre-dose 1, pre-dose 2, and booster day 1 are 11, 10, and 18, respectively. All box and whisker plots show the mean (center line of boxes), 25th and 75th percentiles (lower and upper boundary of boxes), and minimum and maximum values in the controls.
See also Figure S2.
RNASE2 encodes the ribonuclease RNase2, which has been known to be an endogenous Toll-like receptor 2 (TLR2) ligand that drives Th2 response.8 Recently, RNase2 was also found to be an integral part of endosomal synthetic RNA sensing; RNase2 cooperates with RNase T2 to cleave RNA on either side of uridine to form TLR8 ligands.9 Consistent with this recent report, we found heightened expression of TLR8, but not other TLRs, at pre-booster baseline and on day 4 post-booster vaccination (Figures S2D–S2M). Correspondingly, the largest differences in host response to booster vaccination between the subject and controls occurred on day 4 (Figures 2D and S2N), where expression of anti-viral defense as well as interferon response genes, which were elevated from baseline on day 2, albeit not to the levels of controls, was sustained through to day 4 (Figures 2E, 2F, S2O, and S2P). Transcripts of these genes in the subject were reduced to levels comparable to baseline on day 8, when the subject was discharged from the hospital.
Discussion
Although post-vaccination cardiac complications are alarming and potentially life threatening, they remain poorly understood because of their rarity. That this subject had both pre-vaccination and post-vaccination blood samples that preceded the peak of symptoms thus provides a unique glimpse into the plausible pathogenic underpinnings of such complications.
Our findings suggest that the cardiac complication arose as a confluence of several factors. Firstly, our subject was receiving his third dose of mRNA vaccine that induced anamnestic B and T cell responses. Secondly, high pre-booster levels of pro-inflammatory and Th2 cytokines in our subject suggest the possibility of undetected autoimmune activity. This notion is supported by the finding of an elevated inflammation-induced scavenger receptor (OLR1),10 growth factors for the repair of stromal tissue (HGF), and epithelial barriers (OSM and TGF-α).11 , 12 Interestingly, OLR1 is better known for its role in atherosclerosis,13 , 14 although clinical investigations did not suggest ischemic heart disease as a cause of the cardiac event. Instead, OLR1 also functions to support B cell responses,10 , 15 which is supported by the high baseline IL-4 level in our subject. In addition, IL-17A is closely linked to autoimmune conditions,16 and its role in the development of cardiac complications in our subject is further suggested by the resolution of symptoms on day 7 when the plasma level of its antagonist, IL-27, rose.
Finally, while both anamnestic response and baseline inflammation may have elevated the risk of cardiac complications, mRNA vaccination in a person with high expression of RNASE2 may have been the trigger. The mRNA vaccine incorporates pseudouridine to avoid overstimulating RNA sensors in cells.17 When highly expressed in antigen-presenting cells, RNASE2 may remove the only immunomodulatory feature of mRNA vaccine as the encoded enzyme cleaves RNA on the 3′ side of uridine, resulting in an increased abundance of TLR8 ligands.9 That downstream cellular defense and interferon pathways were more elevated and prolonged in the subject further supports this mechanism as the trigger of the pathogenic process.
This case presents an exceptional opportunity to glimpse the baseline factors that may explain the pathogenesis of post-vaccination cardiac complications. Our findings suggest that a confluence of factors gave rise to the cardiac complication, which underscores the rarity of such adverse events. While the findings of monocyte activation at pre-vaccination, as well as elevated TLR8-driven anti-viral defense and interferon-stimulated gene expression after vaccination, suggest that RNASE2 expression in monocytes may play a more important role than eosinophils, future studies that confirm this notion would be useful.
In conclusion, our findings suggest that cardiac complication is a confluence of factors, with endosomal RNA processing through RNase2 to stimulate TLR8 as a trigger of an acute-on-chronic inflammatory outcome. Perhaps additional nucleoside modifications would be useful to prevent such rare over-stimulation of RNA sensors for added vaccine safety.
Limitations of the study
The key limitation in our study is that our findings are based on a single case and that susceptibility to mRNA vaccination-associated cardiac complications may be heterogeneous. Finally, clarifying the immune subset-specific expression of RNASE2 may permit a better mechanistic understanding of post-vaccination cardiac complication.
STAR★Methods
Key resources table
REAGENT or RESOURCE SOURCE IDENTIFIER
Biological samples
Blood from vaccinated healthcare workers This study N.A.
Critical commercial assays
Tempus™ Blood RNA Tube Applied Biosystems Cat# 4342792
Tempus™ Spin RNA Isolation Kit Invitrogen Cat# 4380204
Olink Target 48 Cytokine Olink AB Cat# 93200
Agilent RNA 6000 Pico Kit Agilent 5067–1513
GeneChip Human Gene 2.0 ST Array Applied Biosystems 902499
GeneChip WT PLUS Reagent Kit Applied Biosystems 902280
GeneChip Hybridization, Wash, and Stain Kit Applied Biosystems 900720
Deposited data
Raw data for microarray profiling of case and controls This paper Array Express:
E-MTAB-12829
Software and algorithms
GeneChip Command Console Software Thermo Fisher Scientific https://www.thermofisher.com/sg/en/home/life-science/microarray-analysis/microarray-analysis-instruments-software-services/microarray-analysis-software/affymetrix-genechip-command-console-software.html
Olink NPX Signature v1.6 Olink AB https://olink.com/products-services/data-analysis-products/olink-npx-signature/
Partek Genomics Suite v7.21 Partek Inc. https://www.partek.com/partek-genomics-suite/
Graphpad Prism v9.4 Graphpad https://www.graphpad.com/features
R software The R Project for Statistical Computing https://www.r-project.org
Resource availability
Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Eng Eong Ooi (engeong.ooi@duke-nus.edu.sg).
Materials availability
This study did not generate new reagents.
Method details
The case and controls in this study were healthcare workers who were receiving their COVID-19 vaccination and invited to participate in a study defining the correlates of immunogenicity and adverse events of RNA vaccines. This study was approved by the SingHealth Centralized Institutional Review Board (CIRB/F 202½014). Following written informed consent, whole blood and serum samples were collected prior to vaccination (pre-doses 1 and 2, and pre-booster) and at indicated time points following vaccination. Additional clinical investigations were conducted for the case upon admission for cardiac complication. This included performing an electrocardiogram (ECG), measurements for cardiac biomarkers, renal and liver panel, full blood count, PCR for known viral etiologies of myocarditis, and testing levels of anti-SARS-CoV-2 nucleoprotein antibodies.
Gene expression profiling
Blood was collected in Tempus Blood RNA tubes. RNA isolation from whole blood was performed using the Tempus Spin RNA Isolation Kit (ThermoFisher Scientific) according to manufacturer’s protocol. Total RNA concentration and quality check was done using the Varioskan Lux and RNA 6000 Pico Kit on the Agilent Bioanalyzer (Agilent Technologies, Palo Alto, CA) at the Duke-NUS Genome Biology Core Facility. Microarray was performed using Affymetrix GeneChip Human Gene 2.0 ST Array at the Duke-NUS Genome Biology Core Facility. Fragmented, labeled single-stranded cDNA were prepared from 200ng of total RNA according to manufacturer’s protocol using the GeneChip WT PLUS Reagent Kit (Affymetrix). Samples were hybridized on the GeneChip Human Gene 2.0 ST Array (Affymetrix Inc., Santa Clara, CA) for 16 h at 45°C. GeneChips were washed and stained using the GeneChip Fluidics Station 450 (Affymetrix Inc., Santa Clara, CA). Scanning was done using the GeneChip Scanner 3000 7G (Affymetrix Inc., Santa Clara, CA) and Affymetrix GeneChip Command Console Software (AGCC) to produce.CEL intensity files.
Olink proteomic profiling
Serum protein levels were measured using Olink proximity extension assay (PEA) Target 48 cytokine panel. Samples were first incubated with proximity antibody pairs tagged with unique DNA oligonucleotides. The antibody pairs come into proximity upon binding to target proteins, resulting in hybridization and extension of DNA oligonucleotides. This created an antigen-specific double stranded DNA barcode that was amplified and quantified using multiplex high throughput PCR on the Olink Signature Q100.
Quantification and statistical analysis
The Partek Genomics Suite was used to analyze the transcriptomic differences between case and controls. PLS-DA multivariate analysis was performed using the ropls R package, with the case and controls used to specify the clusters. Boxplots and MA-plots were plotted using GraphPad Prism (version 9.4.1) software and the ggplot2 R package. Comparisons between case and controls were determined using Z-scores, where Z score values above 3 are considered as outlier values, and thus, differentially expressed. For pathway analysis, differentially expressed genes were used as input data and analyzed against the Gene Ontology (GO) Biological Processes database using the Enrichr tool.18 The significant genes contributing to the top enriched terms within the pathways, namely antiviral defense as well as interferon response pathways, were then further analyzed to tabulate the normalized expression levels. The Z score values for these significant genes were calculated and the mean value was tabulated for the individual subjects. To identify the blood transcription modules (BTMs) involved, gene expression fold-change of case against controls were ranked, and analyzed by gene set enrichment analysis (GSEA)19 using BTMs as gene sets in the analysis.20
Supplemental information
Document S1. Table S1 and Figures S1 and S2
Document S2. Article plus supplemental information
Data and code availability
• Raw data for microarray profiling of case and controls have been deposited at Array Express and are publicly available as of the date of publication. Accession number is listed in the key resources table.
• This paper does not report original code.
• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
Acknowledgments
We thank the volunteers who participated in our cohort study and are delighted that this subject has recovered full fitness. We thank Esteban Finol for the discussions on mRNA vaccine. J.G.L. and E.E.O. receive salary support from the 10.13039/501100001349 National Medical Research Council (10.13039/501100001349 NMRC ) of Singapore through the Clinician-Scientist and Singapore Translational Research Award schemes, respectively.
Author contributions
D.J.H.T., J.X.Y.S., and J.G.L. led the clinical portion of this study and enrolled the subjects. E.Z.O., J.X.Y., V.S.Y.C., and Y.S.L. conducted the gene expression and cytokine profiling and managed sample processing and archiving. K.G., C.P.Y., and L.L.E.O. conducted additional clinical investigations. E.Z.O., C.W.T.K., J.S.G.O., and K.R.C. analyzed the data. E.E.O. conceived the study. E.Z.O., K.R.C., J.G.L., and E.E.O. wrote the manuscript. E.Z.O., K.R.C., J.G.L., and E.E.O. had unrestricted access to all of the data. All authors agreed to submit the manuscript, read and approved the final draft, and take full responsibility for its content, including the accuracy of the data.
Declaration of interests
The authors declare no competing interests.
Supplemental information can be found online at https://doi.org/10.1016/j.medj.2023.04.001.
==== Refs
References
1 Kalimuddin S. Tham C.Y.L. Qui M. de Alwis R. Sim J.X.Y. Lim J.M.E. Tan H.C. Syenina A. Zhang S.L. Le Bert N. Early T cell and binding antibody responses are associated with COVID-19 RNA vaccine efficacy onset Med (N Y) 2 2021 682 688.e4 10.1016/j.medj.2021.04.003
2 Fiolet T. Kherabi Y. MacDonald C.J. Ghosn J. Peiffer-Smadja N. Comparing COVID-19 vaccines for their characteristics, efficacy and effectiveness against SARS-CoV-2 and variants of concern: a narrative review Clin. Microbiol. Infect. 28 2022 202 221 10.1016/j.cmi.2021.10.005 34715347
3 Wong H.L. Hu M. Zhou C.K. Lloyd P.C. Amend K.L. Beachler D.C. Secora A. McMahill-Walraven C.N. Lu Y. Wu Y. Risk of myocarditis and pericarditis after the COVID-19 mRNA vaccination in the USA: a cohort study in claims databases Lancet 399 2022 2191 2199 10.1016/S0140-6736(22)00791-7 35691322
4 Chan K.R. Gan E.S. Chan C.Y.Y. Liang C. Low J.Z.H. Zhang S.L.X. Ong E.Z. Bhatta A. Wijaya L. Lee Y.H. Metabolic perturbations and cellular stress underpin susceptibility to symptomatic live-attenuated yellow fever infection Nat. Med. 25 2019 1218 1224 10.1038/s41591-019-0510-7 31308506
5 Syenina A. Gan E.S. Toh J.Z.N. de Alwis R. Lin L.Z. Tham C.Y.L. Yee J.X. Leong Y.S. Sam H. Cheong C. Adverse effects following anti-COVID-19 vaccination with mRNA-based BNT162b2 are alleviated by altering the route of administration and correlate with baseline enrichment of T and NK cell genes PLoS Biol. 20 2022 e3001643 10.1371/journal.pbio.3001643
6 Pearce E.L. Pearce E.J. Metabolic pathways in immune cell activation and quiescence Immunity 38 2013 633 643 10.1016/j.immuni.2013.04.005 23601682
7 Rosenberg H.F. Eosinophil-derived neurotoxin (EDN/RNase 2) and the mouse eosinophil-associated RNases (mEars): expanding roles in promoting host defense Int. J. Mol. Sci. 16 2015 15442 15455 10.3390/ijms160715442 26184157
8 Zhu Y. Tang X. Xu Y. Wu S. Liu W. Geng L. Ma X. Tsao B.P. Feng X. Sun L. RNASE2 mediates age-associated B cell expansion through monocyte derived IL-10 in patients with systemic lupus erythematosus Front. Immunol. 13 2022 752189 10.3389/fimmu.2022.752189
9 Ostendorf T. Zillinger T. Andryka K. Schlee-Guimaraes T.M. Schmitz S. Marx S. Bayrak K. Linke R. Salgert S. Wegner J. Immune sensing of synthetic, bacterial, and Protozoan RNA by toll-like receptor 8 requires coordinated processing by RNase T2 and RNase 2 Immunity 52 2020 591 605.e6 10.1016/j.immuni.2020.03.009 32294405
10 Brink R. LOX-1 unlocks human plasma cell potential Immunity 41 2014 507 508 10.1016/j.immuni.2014.09.019 25367564
11 Fajardo-Puerta A.B. Mato Prado M. Frampton A.E. Jiao L.R. Gene of the month: HGF J. Clin. Pathol. 69 2016 575 579 10.1136/jclinpath-2015-203575 27072686
12 West N.R. Owens B.M.J. Hegazy A.N. The oncostatin M-stromal cell axis in health and disease Scand. J. Immunol. 88 2018 e12694 10.1111/sji.12694
13 Wang L. Yanuck D. Beecham A. Gardener H. Slifer S. Blanton S.H. Sacco R.L. Rundek T. A candidate gene study revealed sex-specific association between the OLR1 gene and carotid plaque Stroke 42 2011 588 592 10.1161/STROKEAHA.110.596841 21257822
14 Skarpengland T. Skjelland M. Kong X.Y. Skagen K. Holm S. Otterdal K. Dahl C.P. Krohg-Sørensen K. Sagen E.L. Bjerkeli V. Increased levels of lectin-like oxidized low-density lipoprotein receptor-1 in ischemic stroke and transient ischemic attack J. Am. Heart Assoc. 7 2018 e006479 10.1161/JAHA.117.006479
15 Joo H. Li D. Dullaers M. Kim T.W. Duluc D. Upchurch K. Xue Y. Zurawski S. Le Grand R. Liu Y.J. C-type lectin-like receptor LOX-1 promotes dendritic cell-mediated class-switched B cell responses Immunity 41 2014 592 604 10.1016/j.immuni.2014.09.009 25308333
16 McGeachy M.J. Cua D.J. Gaffen S.L. The IL-17 family of cytokines in health and disease Immunity 50 2019 892 906 10.1016/j.immuni.2019.03.021 30995505
17 Karikó K. Buckstein M. Ni H. Weissman D. Suppression of RNA recognition by Toll-like receptors: the impact of nucleoside modification and the evolutionary origin of RNA Immunity 23 2005 165 175 10.1016/j.immuni.2005.06.008 16111635
18 Kuleshov M.V. Jones M.R. Rouillard A.D. Fernandez N.F. Duan Q. Wang Z. Koplev S. Jenkins S.L. Jagodnik K.M. Lachmann A. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update Nucleic Acids Res. 44 2016 W90 W97 10.1093/nar/gkw377 27141961
19 Subramanian A. Tamayo P. Mootha V.K. Mukherjee S. Ebert B.L. Gillette M.A. Paulovich A. Pomeroy S.L. Golub T.R. Lander E.S. Mesirov J.P. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles Proc. Natl. Acad. Sci. USA 102 2005 15545 15550 10.1073/pnas.0506580102 16199517
20 Li S. Rouphael N. Duraisingham S. Romero-Steiner S. Presnell S. Davis C. Schmidt D.S. Johnson S.E. Milton A. Rajam G. Molecular signatures of antibody responses derived from a systems biology study of five human vaccines Nat. Immunol. 15 2014 195 204 10.1038/ni.2789 24336226
|
PMC010xxxxxx/PMC10131737.txt |
==== Front
J Infect
J Infect
The Journal of Infection
0163-4453
1532-2742
The British Infection Association. Published by Elsevier Ltd.
S0163-4453(23)00255-4
10.1016/j.jinf.2023.04.018
Letter to the Editor
Undetectable intrauterine transmission during the first trimester of pregnancy in woman after COVID-19 infection
Xu Jian 1
Mao Di 1
Liang Peiling
Center of Reproductive Medicine, Guangzhou Women and Children’s Medical Center, Guangzhou Medical University, Guangdong Province, China
Du Peng
Zhang Xiaomeng
Dang Xiaoyan
Genetic Testing Center, Guangzhou Women and Children's Medical Center, Guangzhou Medical University, Guangdong Province, China
Wu Haiying ∗
Family Planning Clinic, Department of Obstetrics and Gynecology, Guangzhou Women and Children’s Medical Center, Guangzhou Medical University, Guangdong Province, China
Zhu Bing ∗
Virus Laboratory, Institute of Pediatrics, Guangzhou Women and Children’s Medical Center, Guangzhou Medical University, Guangdong Province, China
Sun Ling ∗
Center of Reproductive Medicine, Guangzhou Women and Children’s Medical Center, Guangzhou Medical University, Guangdong Province, China
∗ Corresponding author.
∗ Corresponding author.
∗ Corresponding author.
1 These authors have contributed equally to this work.
26 4 2023
7 2023
26 4 2023
87 1 8081
22 4 2023
© 2023 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
2023
The British Infection Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcDear Editor,
The findings of Liu et al.1 in this journal showed that pregnant women are susceptible to COVID-19 infection and less than 2% of confirmed vertical transmission of SARS-CoV-2 has been demonstrated in pregnant women with SARS-CoV-2 infection, especially in the third trimester.2 However, there are scare published reports regarding the possibility of SARS-CoV-2 vertical transmission during the first trimester of pregnancy, the critical period for significant malformations. A recent online survey conducted in Macao, China, showed that the cumulative SARS-CoV-2 infection rate rose rapidly to 70% within three weeks after Chinese government announcing the “10 new measures” on December 7, 2022.3 Women who became pregnant during this period concerned about the safety of their fetus. Herein, we focused on the potential intrauterine transmission of SARS-CoV-2 during the first trimester of pregnancy, after the ending of zero-COVID policy in China.
Initially, 61 first trimester pregnant women who planned surgical abortion between January 13 and February 1, 2023 were asked to fill a questionnaire inquiring about previous vaccination and the characteristics of the COVID-19 disease (confirmed by antigen or polymerase chain reaction test). This study was approved by the Ethics Committee of Guangzhou Women and Children’s Hospital and written informed consent was obtained. Finally, 55 participants who had recovered from a recent COVID-19 infection were enrolled, including missed abortion in 16 cases and voluntary termination of normal early pregnancy in 39 cases. All of them were natural pregnancies, had received inactivated vaccination against COVID-19 (mean dose 2.8 ± 0.6), reported mild COVID-19 symptoms during their active infection period. The weeks of gestation at the onset of infection was between 0 and 8, of which 26 cases in 0–3 weeks, 29 cases in 3–8 weeks. The mean gestational age at the time of surgery was 7.9 ± 1.6 weeks and the time lapse between the onset of infection and operation was 36.4 ± 9.4 days.
Throat swabs, plasma, and villus samples of all 55 patients, and 11 identified fetal samples on the operation day were analyzed for SARS-CoV-2 RNA, determined by rRT-PCR targeting the ORF-1ab gene and N-gene of SARS-CoV-2 (Shanghai Bio-Germ Medical Technology Co Ltd). A Ct value less than 40 from either gene detection is interpreted as positive for SARS-CoV-2 RNA. No viral RNA was detected in all blood, villus, and identified fetal samples. Unexpectedly, we found that the throat swab samples from 2 patients were weak positive. The Ct value of Orf1ab and N genes was 36.4 and 38.9, and 37.1 and 36.9, respectively, in each specimen. The two patients reported negative antigen test 19 and 26 days, respectively, prior to the operation day and had no symptoms of conscious discomfort before operation. Consistent with this result, a recent study in our city showed that about 20% patients retested positive for SARS-CoV-2 following negative testing results and discharge, suggesting that a significant proportion of patients could carry viral fragments for a long time.4
To determine whether immune protection have formed in these convalescent women, we evaluated plasma neutralizing antibodies titers against SARS-CoV-2 subvariant omicron BA.5 (the dominant subvariant identified in Guangzhou). The effects of the plasma on entry inhibition by the pseudotyped viruses were evaluated through measurement of a reduction in the luciferase gene expression.5 Neutralizing antibody titers were expressed as the sample dilution that caused 50% neutralization and the ED50 (median effective dose) of each sample was calculated using the Reed-Muench method.6 Neutralization activities of the immune plasma were detected in all the 55 convalescent women. The average values of ED50 in each sample ranged from 65.5 to 1030.5, with a mean value of 224.0 ± 184.9. Since a control group of unvaccinated population was not able to be included, we could not discriminate whether the protective neutralizing antibody is produced by vaccine or infection.
Even though the overall vertical transmission rate of SARS-CoV-2 in pregnant women with SARS-CoV-2 infection is very low,2 a recent observational study including 17 asymptomatic women during the first trimester pregnancy (8–12 weeks) found that approximately 30% of the fetuses and 17% of the syncytiotrophoblasts were SARS-CoV-2-positive.7 The reasons why intrauterine transmission of SARS-CoV-2 RNA was not detected in our study might be as follows: Firstly, all the women in our study had received vaccination against COVID-19 prior to infection, while none of their participants received vaccination. An important observation is that all reported cases of SARS-CoV-2 placentitis causing stillbirth and neonatal death were in the mothers who were unvaccinated.8 Secondly, the pregnant women were SARS-CoV-2-positive on the operation day in their study, while the participants we enrolled had all recovered about one month before surgery. Moreover, COVID-19 has been constantly changing, and it is possible that the symptoms caused by the SARS-CoV-2 subvariant currently prevalent in our region were relatively mild, and vertical transmission was not easy to occur.
In conclusion, for the first time, we investigated the possibility of intrauterine transmission of SARS-CoV-2 RNA in the first trimester of pregnancy when infection occurred in the critically early gestational weeks (0–8 weeks). SARS-CoV-2 RNA was not detected in all villus and identified fetal samples, indicating no evidence of intrauterine transmission, providing meaningful clinical evidence for the consultation of pregnant women infected during this stage. Even if the possibility of vertical transmission in early pregnancy cannot be ruled out due to the limited sample size, it should be considered a rare event.
Funding
This study was supported by the 10.13039/501100003453 Natural Science Foundation of Guangdong Province (No. 2023A1515010250).
CRediT authorship contribution statement
Jian Xu contributed to design the study, process the samples, analysis, and interpretation of data, and draft the whole article. Ling Sun was responsible for the conception and design of the study, revised the article critically. Haiying Wu and Bing Zhu helped to design the study and revise the article for important intellectual content. Di Mao contributed to process the samples, collecting the data, draft and revise the article. Peiling Liang contributed to process the samples and collect the data. Peng Du, Xiaomeng Zhang and Xiaoyan Dang were responsible for pseudovirus neutralizing antibody assay and interpretation of data. All authors discussed the results and reviewed the manuscript.
Declaration of Competing Interest
None declared.
Acknowledgment
We thank Wanli Liang and Zhengfang Lin for their assistance in the RT-PCR technique in detecting COVID-19 RNA.
==== Refs
References
1 Liu Y. Chen H. Tan W. Kuang Y. Tang K. Luo Y. Clinical characteristics and outcome of SARS-CoV-2 infection during pregnancy J Infect 82 2021 e9 e10 33831458
2 Allotey J. Chatterjee S. Kew T. Gaetano A. Stallings E. Fernández-García S. SARS-CoV-2 positivity in offspring and timing of mother-to-child transmission: living systematic review and meta-analysis BMJ Clin Res Ed 376 2022 e067696
3 Liang J. Liu R. He W. Zeng Z. Wang Y. Wang B. Infection rates of 70% of the population observed within 3 weeks after release of COVID-19 restrictions in Macao, China J Infect 86 2023 402 404 36731635
4 Luo L. Liu D. Zhang Z. Li Z. Xie C. Wang Z. Risk factors associated with PCR repositivity in patients with COVID-19 after recovery in Guangzhou, China: a retrospective cohort study Author Prepr 2022
5 Du P. Li N. Xiong X. Tang S. Dai Q. Liu Z. A bivalent vaccine containing D614G and BA.1 spike trimer proteins or a BA.1 spike trimer protein booster shows broad neutralizing immunity J Med Virol 94 2022 4287 4293 35614524
6 Nie J. Li Q. Wu J. Zhao C. Hao H. Liu H. Quantification of SARS-CoV-2 neutralizing antibody by a pseudotyped virus-based assay Nat Protoc 15 2020 3699 3715 32978602
7 Fenizia C. Vanetti C. Rana F. Cappelletti G. Cetin I. Biasin M. SARS-CoV-2 vertical transmission during the first trimester of pregnancy in asymptomatic women Int J Infect Dis 124 2022 159 163 36122670
8 Schwartz D.A. Mulkey S.B. Roberts D.J. SARS-CoV-2 placentitis, stillbirth, and maternal COVID-19 vaccination: clinical-pathologic correlations Am J Obstet Gynecol 228 2023 261 269 36243041
|
PMC010xxxxxx/PMC10131741.txt |
==== Front
J Infect
J Infect
The Journal of Infection
0163-4453
1532-2742
The British Infection Association. Published by Elsevier Ltd.
S0163-4453(23)00254-2
10.1016/j.jinf.2023.04.020
Letter to the Editor
Impact of COVID-19 control measures on Legionella pneumophila infections in children in Henan, China
Zhou Yang ⁎
Yan Hui
Zhou Qiang
Feng Ruiling
Zhai Bo ⁎
Henan Provincial Clinical Research Center for Pediatric Diseases, Children's Hospital Affiliated to Zhengzhou University, Henan Children's Hospital, Zhengzhou Children's Hospital, Zhengzhou 450018, China
⁎ Corresponding authors.
26 4 2023
7 2023
26 4 2023
87 1 8587
22 4 2023
© 2023 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
2023
The British Infection Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcDear Editor,
Since the outbreak of acute respiratory syndrome coronavirus 2 (SARS-CoV-2), prevention and control interventions including mask wearing, hand hygiene, social distancing and travel restrictions have been implemented globally to control the spread of COVID-19. Notably, the prevalence of other pathogenic infections has also been altered following the implementation of interventions. In this journal, the reports on the trend of respiratory disease prevalence before and after the COVID-19 pandemic in China, have recently attracted our attention. Li et al., Zhou et al. and Li et al. successively demonstrated the downward trend in the prevalence of Streptococcus pneumoniae, Haemophilus influenzae and respiratory (non-digestive) system E. coli in children during the COVID-19 pandemic.1, 2, 3 Yet, the impact of the respiratory pathogen Legionella pneumophila (L. pneumophila) on children, a vulnerable population, remains largely unknown. Therefore, evaluation of seasonal, age composition and epidemiological changes of L. pneumophila infections in children before and after the COVID-19 pandemic is indispensable for the prevention and control of respiratory infections in children.
L. pneumophila, an aerobic gram-negative bacterium, was named for an unidentified lung infection discovered at the annual legion meeting in Philadelphia, USA, in 1976.4L. pneumophila is widespread in natural water systems (freshwater reservoirs, waterways) as well as in man-made water systems (landscape fountains, plumbing systems, air conditioning units, shower installations).5 As one of the respiratory pathogens of severe community-acquired pneumonia, L. pneumophila is inhaled into the respiratory tract as aerosols and causes lower respiratory diseases by attacking alveolar macrophages.6, 7 The resulting symptoms are varied and non-specific, such as mild symptoms (high fever, muscle and joint pain, nausea, vomiting), severe symptoms (respiratory distress, respiratory failure and multi-organ failure), and even death. Although Legionnaires' disease frequently occurs in adults, L. pneumophila tends to cause more severe pneumonia in children, which has led to increased public health concern in recent years.8 Given the lack of an effective clinical vaccine against L. pneumophila, control of this pathogenic infection is challenging. Since the COVID-19 pandemic in late 2019, globally implemented interventions such as crowd reduction, travel restrictions, mask wearing, and closure of high-risk non-essential sites have altered the prevalence of respiratory-associated pathogens other than SARS-CoV-2.1, 2, 3, 9 Therefore, assessing the impact of the COVID-19 pandemic on the prevalence of L. pneumophila in children is of profound importance for infection prevention and clinical management strategies.
Herein, we investigated the changes of L. pneumophila infections in children before and after the COVID-19 pandemic through comparison of the number of positive infections and detection rate (positive test specimens/total specimens), as well as these differences in children at various age groups (<1 year, 1–3 years, 3–5 years, 5–18 years) based on laboratory surveillance data of children with lower respiratory tract diseases at Henan Children’s Hospital from January 2018 to October 2022. Our data showed that the number of positive infections and detection rate of L. pneumophila continued to be high before the COVID-19 pandemic (2018.01–2019.12), while there was an overall decline after the COVID-19 pandemic (2020.01–2022.10). Paradoxically, the number of positive infections and detection rates in 2021 exhibited a slight rise compared to 2020 and 2022, which remained lower than those before the COVID-19 pandemic ( Fig. 1A, B). Notably, L. pneumophila infection in children exhibited apparent seasonality prior to the COVID-19 pandemic, with morbidity concentrated in the summer and fall (July-September), which was attenuated during the COVID-19 pandemic (Fig. 1C, D). From August to September 2021, the L. pneumophila positive infections and detection rates demonstrated showed sustained upward trends. Actually, in late July 2021, the extreme rainfall in Henan Province caused flooding in many areas and was accompanied by water pollution in the living environment and wetting of electrical equipment. Based on the pathogenicity of L. pneumophila, we speculated that the phenomenon might be critically related to the flooding-shaped advantages for the multiplication and spread of L. pneumophila, such as the ubiquitous warm sewage and contaminated centralized air conditioning. However, L. pneumophila infections did not rebound to pre-pandemic levels in 2021. In addition, there was a slight increase in the total number of L. pneumophila-positive infections in 2022 compared to 2020, which was likely the result of the increased mobile trajectories of children due to the relative stability of the epidemic, such as kindergarten and school resumption, travel, and access to public places with contaminated water. Nevertheless, the general decline in L. pneumophila infections following the outbreak suggests that the COVID-19 pandemic actually changed the prevalence of L. pneumophila in children.Fig. 1 The positive number (A) and detection rates (B) of L. pneumophila in children by year from 2018 to 2022. The positive number (C) and detection rates (D) of L. pneumophila in children by month from 2018 to 2022.
Fig. 1
Furthermore, the results of different age stratification indicated that L. pneumophila infections largely tended to increase with age regardless of the COVID-19 pandemic, with the number of positive samples and detection rates in 3–18 years old far exceeding those of children under 3 years old, suggesting that the risk of L. pneumophila infections are higher in children over 3 years of age, especially those aged 5–18 years who are more exposed in public places and the external environment ( Fig. 2).Fig. 2 The positive number (A) and detection rates (B) of L. pneumophila in children at different ages from 2018 to 2022.
Fig. 2
In this study, we investigated changes in the number of positive infections, detection rates, seasonality, and age of prevalence of L. pneumophila infections in children before and after the COVID-19 pandemic, based on laboratory surveillance data in children with lower respiratory tract disease. The containment of L. pneumophila infections was attributed to several factors: During the COVID-19 pandemic, childhood travel was restricted, exposure to external environments with sewage was reduced, access to public places equipped with central air conditioning was limited, masks were worn, and blockage of aerosols in household pipes was conducted.
In summary, L. pneumophila infections declined during the COVID-19 pandemic. Attention should be paid to L. pneumophila infections in school-aged children, and pathogen surveillance for control and prevention of lower respiratory tract infections ought to be strengthened.
Declaration of Competing Interest
The authors declare no conflict of interests.
Acknowledgments
This work was supported by grants from the Medical Science and Technology Projects of Henan Province (LHGJ20190958, LHGJ20190889 and LHGJ20210629).
==== Refs
References
1 Li Y. Guo Y. Duan Y. Changes in Streptococcus pneumoniae infection in children before and after the COVID-19 pandemic in Zhengzhou, China J Infect 85 3 2022 e80 e81 35659542
2 Zhou J. Zhao P. Nie M. Gao K. Yang J. Sun J. Changes of Haemophilus influenzae infection in children before and after the COVID-19 pandemic, Henan, China J Infect 86 1 2023 66 117
3 Li L. Song C. Li P. Li Y. Changes of Escherichia coli infection in children before and after the COVID-19 pandemic in Zhengzhou, China J Infect 86 2 2023 154 225
4 Fraser D.W. Tsai T.R. Orenstein W. Parkin W.E. Beecham H.J. Sharrar R.G. Legionnaires' disease: description of an epidemic of pneumonia N Engl J Med 297 22 1977 1189 1197 335244
5 Gonçalves I.G. Simões L.C. Simões M. Legionella pneumophila Trends Microbiol 29 9 2021 860 861 33994277
6 Cunha B.A. Burillo A. Bouza E. Legionnaires' disease Lancet 387 10016 2016 376 385 26231463
7 Escoll P. Rolando M. Gomez-Valero L. Buchrieser C. From amoeba to macrophages: exploring the molecular mechanisms of Legionella pneumophila infection in both hosts Curr Top Microbiol Immunol 376 2013 1 34 23949285
8 Fulová M. Kotrbancová M. Bražinová A. Boledovičová J. Trnková K. Špaleková M. Legionnaires' disease in pediatric patients, control measures and 5-year follow-up Pediatr Infect Dis J 39 11 2020 990 994 32472821
9 Ayouni I. Maatoug J. Dhouib W. Zammit N. Fredj S.B. Ghammam R. Effective public health measures to mitigate the spread of COVID-19: a systematic review BMC Public Health 21 1 2021 1015 34051769
|
PMC010xxxxxx/PMC10131744.txt |
==== Front
J Infect
J Infect
The Journal of Infection
0163-4453
1532-2742
The British Infection Association. Published by Elsevier Ltd.
S0163-4453(23)00253-0
10.1016/j.jinf.2023.04.017
Letter to the Editor
From COVID-19 to measles: Prioritizing immunization in Pakistan's far-flung regions
Awan Usman Ayub *
Hussain Mushahid 1
Qureshi Masood 1
Siddique Zeeshan
Khattak Aamer Ali
Department of Medical Laboratory Technology, The University of Haripur, Haripur, Khyber Pakhtunkhwa, Pakistan
Akhtar Sohail
Department of Mathematics and Statistics, The University of Haripur, Haripur, Khyber Pakhtunkhwa, Pakistan
Guo Xingyi *
Division of Epidemiology, Department of Medicine, Vanderbilt Epidemiology Center, Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, TN, USA
Department of Biomedical Informatics, Vanderbilt University School of Medicine, Nashville, TN, USA
⁎ Corresponding author.
* Corresponding author at: Division of Epidemiology, Department of Medicine, Vanderbilt Epidemiology Center, Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, TN, USA.
1 These authors contributed equally.
26 4 2023
7 2023
26 4 2023
87 1 7679
26 4 2023
© 2023 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
2023
The British Infection Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Keywords
Measles
Pakistan
Vaccination
Public health
Infectious disease
==== Body
pmcDear Editor,
We read an article with great interest by Muhammad Suleman Rana et al.,1 titled “Impact of COVID-19 pandemic on Measles surveillance in Pakistan”. The authors have cogently presented strategies to safeguard children from these viral infections in Pakistan. We concur with the concerns raised by the authors, especially given the recent upsurge of measles in Pakistan. Nonetheless, remote localities such as the Khurram agency in Khyber Pakhtunkhwa have faced challenges reaching optimal measles vaccination coverage during the past three years. Herein, this article intends to examine potential causes and give a way forward to improve public health through greater vaccination uptake.
Measles, an extremely contagious disease caused by a member of the paramyxovirus family, is a leading cause of morbidity and mortality, accounting for more than 2.6 million fatalities annually worldwide. Alarmingly, the measles virus has a basic reproduction number of 12-to-18, rendering it the most contagious viral virus and substantially more likely than SARS-CoV-2.2 In 2019, there were 413,308 confirmed cases of the disease worldwide, and 207,500 fatalities were directly linked to measles complications. Compared to the preceding years, when the annual fatality rate from measles was 140,000 in 2018, 110,000 in 2017, and 89,780 in 2016, this constituted a considerable rise.1, 2 Dreadfully, in many African and Asian nations, like Nigeria, Pakistan, and Afghanistan—measles is still endemic. Vaccination helps prevent measles, but the COVID-19 pandemic has changed infectious disease epidemiological trends worldwide, according to the World Health Organization (WHO).3
Despite international attempts to eradicate the illness, several nations continue to see recurrent measles outbreaks. Because of its high population of unvaccinated children, Pakistan ranks among the top 10 nations where measles epidemics occur, along with India, Yemen, Somalia, Ethiopia, Indonesia, and Zimbabwe.4 During COVID-19, it may have looked prudent to halt children's vaccination efforts to prevent the spread of SARS-CoV-25; doing so has led to other public-health calamities in Pakistan, such as an outbreak of measles. Worryingly, in 2020, over 120 million children did not get the measles vaccination, with 40 million residing in Pakistan.6 Due to Pakistan's historically low measles vaccination coverage—in 2018, less than 66% of the population received the first, while only 45% received the second dose, considerably below the 90% coverage needed for herd immunity—threatened the public health.7
Alongside the devastation of COVID-19 on the overall vaccination drive in Pakistan, we must not overlook the diverse challenges of immunization in remote areas of Pakistan. For instance, Pakistan still encounters wild polio cases due to widespread vaccine hesitancy among the general population. In the past, rumors about the quality of the vaccine, religious opinions, and myths about infertility after vaccination undermines the nation's polio eradication drives. Consequently, these myths have stoked parents' staunch aversion to vaccinating their children.8, 10
According to a research in 2020, Expanded Program of Immunization (EPI) in Pakistan at the district level revealed that vaccination coverage was lesser in the Federally Administered Tribal Areas (FATA), Khyber Pakhtunkhwa (KPK), and Balochistan than in Punjab and Sindh.9 Whereas Kurram Agency, one of the tribal agencies located on the border of Pakistan and Afghanistan, has been historically affected by war. Alarmingly, the measles vaccination trends in 2020–2022 in this region (Central and Lower Kurram) highlighted that both districts had increased coverage from 2020 to 2021. In 2022, Lower Kurram was completely covered, whereas Central Kurram's coverage decreased to 83%. Commendable progress in protecting children from measles through vaccination in lower Khurram, while the central Kurram district still faces significant gaps in coverage that need to be urgently addressed, as shown in Fig. 1, Fig. 2. In contrast, the vaccination rate (%) in the various union councils of lower and central Kurram was evaluated for 2020–2022 to gain additional insight, as shown in Fig. 3. The data shows a worrying downward trend in vaccination rates among union councils, which might spell trouble for public health.Fig. 1 This image depicts the measles vaccination coverage percentage for 2020–21 in the Federally Administered Tribal Areas (FATA) of Pakistan's Kurram Agency region. The data showed the measles vaccination rates achieved as compared to the targeted population in the specified region.
Fig. 1
Fig. 2 This image depicts the measles vaccination coverage percentage for the year-2022 in the Federally Administered Tribal Areas (FATA) of Pakistan's Kurram Agency region. Similarly, the data shows the percentage of the selected population vaccinated against measles.
Fig. 2
Fig. 3 Overall vaccine acceptance rate in the Kurram agency (lower and central region) in 2021–2022.
Fig. 3
To eliminate the measles menace, it is necessary to vaccinate every child in Kurram Agency. Herein, it is recommended that health authorities create and carry out targeted vaccination campaigns in areas of Kurram Agency with low vaccination rates. Rather than a few major vaccination centers, regional immunization units should be set up in each region. In addition, policymakers should consider strategies for combating vaccination fear, such as speaking with community leaders, religious figures, and other opinion leaders to dispel vaccine myths. To conclude, public health officials and policymakers must regularly evaluate their immunization efforts to eliminate measles, identify problem areas, and improve vaccination rates.
CRediT authorship contribution statement
UAA, MH and MQ conceived and designed the study, analyzed and interpreted the data. UAA, MQ, SA, AAK and ZS were involved in the writing of the first draft, statistical analysis of the manuscript and interpretation of results. MH was involved in collection of data. UAA and XG supervised did the final correction of the manuscript. All authors reviewed and approved the final version of the manuscript.
Funding
There is no role of any funding source for this manuscript.
Declaration of Competing Interest
We declare no conflict of interest.
==== Refs
References
1 Rana M.S. Usman M. Alam M.M. Mere M.O. Ikram A. Zaidi S.S.Z. Impact of COVID-19 pandemic on Measles surveillance in Pakistan J Infect 82 3 2021 414 451
2 Durrheim D.N. Andrus J.K. Tabassum S. Bashour H. Githanga D. Pfaff G. A dangerous measles future looms beyond the COVID-19 pandemic Nat Med 27 3 2021 360 361 33589823
3 Dixon M.G. Ferrari M. Antoni S. Li X. Portnoy A. Lambert B. Progress toward regional measles elimination—worldwide, 2000–2020 Morb Mortal Wkly Rep 70 45 2021 1563
4 Prevention CDC. Global measles outbreaks; 2023. Available from: 〈https://www.cdc.gov/globalhealth/measles/data/global-measles-outbreaks.html〉.
5 Awan U.A. Haroon K. Haqqi A. Khattak A.A. Khurram M. Ahmed H. COVID-19 and immunization gap in Pakistan: fear drives for forthcoming spikes Public Health 197 2021 e21 e22 33745735
6 Guha-Sapir D. Moitinho de Almeida M. Keita M. Greenough G. Bendavid E. COVID-19 policies: remember measles Science 369 6501 2020 261 32675365
7 Rana M.S. Alam M.M. Ikram A. Salman M. Mere M.O. Usman M. Emergence of measles during the COVID-19 pandemic threatens Pakistan’s children and the wider region Nat Med 27 7 2021 1127 1128 34183836
8 Awan U.A. Malik M.W. Khattak A.A. Ahmed H. Khan M.I. Qureshi H. Emerging polio hotspots in Pakistan: challenges and the way forward J Infect 83 4 2021 496 522
9 Umer M.F. Zofeen S. Hu W. Qi X. Zhuang G. Spatiotemporal clustering analysis of Expanded Program on Immunization (EPI) vaccination coverage in Pakistan Sci Rep 10 1 2020 10980 32620798
10 Awan U.A. Malik M.W. Ahmed H Hassan U Khan A.S. Bashir S. War-torn Afghanistan-potential risk to the polio eradication efforts: A call for global concern! J Infect 83 6 2021 709 837
|
PMC010xxxxxx/PMC10133912.txt |
==== Front
Eur Geriatr Med
Eur Geriatr Med
European Geriatric Medicine
1878-7649
1878-7657
Springer International Publishing Cham
37103661
787
10.1007/s41999-023-00787-w
Research Paper
Predictors of prolonged hospitalization of COVID-19 patients
http://orcid.org/0000-0002-1372-2040
Lucijanic Marko markolucijanic@yahoo.com
12
Marelic Daniela 2
Stojic Josip 3
Markovic Ivan 4
Sedlic Filip 25
Kralj Ivan 6
Rucevic Davor 7
Busic Niksa 8
Javor Patrik 9
Lucijanic Tomo 1011
Mitrovic Josko 212
Luksic Ivica 21113
1 grid.412095.b 0000 0004 0631 385X Hematology Department, University Hospital Dubrava, Av. Gojka Suska 6, 10000 Zagreb, Croatia
2 grid.4808.4 0000 0001 0657 4636 School of Medicine, University of Zagreb, Zagreb, Croatia
3 grid.412095.b 0000 0004 0631 385X Department of Gastroenterology, Hepatology and Clinical Nutrition, University Hospital Dubrava, Zagreb, Croatia
4 Special Hospital for Pulmonary Diseases, Zagreb, Croatia
5 grid.412688.1 0000 0004 0397 9648 Department of Oncology, Division of Pathophysiology and Experimental Oncology, University Hospital Center Zagreb, Zagreb, Croatia
6 Internal Medicine Department, General Hospital Sisak, Sisak, Croatia
7 grid.412095.b 0000 0004 0631 385X Intensive Care Unit Department, University Hospital Dubrava, Zagreb, Croatia
8 grid.412095.b 0000 0004 0631 385X Cardiology Department, University Hospital Dubrava, Zagreb, Croatia
9 grid.22939.33 0000 0001 2236 1630 Faculty of Medicine, University of Rijeka, Rijeka, Croatia
10 grid.412095.b 0000 0004 0631 385X Endocrinology Department, University Hospital Dubrava, Zagreb, Croatia
11 grid.412095.b 0000 0004 0631 385X Primary Respiratory and Intensive Care Center, University Hospital Dubrava, Zagreb, Croatia
12 grid.412095.b 0000 0004 0631 385X Clinical Immunology, Allergology and Rheumatology Department, University Hospital Dubrava, Zagreb, Croatia
13 grid.412095.b 0000 0004 0631 385X Maxillofacial Surgery Department, University Hospital Dubrava, Zagreb, Croatia
27 4 2023
2023
14 3 511516
29 1 2023
11 4 2023
© The Author(s), under exclusive licence to European Geriatric Medicine Society 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Key summary points
Aim
To investigate predictors of prolonged hospitalization of COVID-19 patients.
Findings
Multivariate analysis recognized higher severity of COVID-19 symptoms, worse functional status, referral from other institutions, certain indications for admission (neurologic, surgical and social), certain chronic comorbidities (obesity, chronic liver disease, hematological malignancy, transplanted organ), and complications that arise during hospital stay (venous thromboembolism, bacterial sepsis and Clostridioides difficile infection) as independent predictors of prolonged hospitalization.
Message
Development of specific measures aimed at improvement of functional status and prevention of complications might reduce the length of hospitalization.
Supplementary Information
The online version contains supplementary material available at 10.1007/s41999-023-00787-w.
Purpose
Despite the importance of hospital bed network during the pandemic, there are scarce data available regarding factors predictive of prolonged length of hospitalization of COVID-19 patients.
Methods
We retrospectively analyzed a total of 5959 consecutive hospitalized COVID-19 patients in period 3/2020–6/2021 from a single tertiary-level institution. Prolonged hospitalization was defined as hospital stay > 21 days to account for mandatory isolation period in immunocompromised patients.
Results
Median length of hospital stay was 10 days. A total of 799 (13.4%) patients required prolonged hospitalization. Factors that remained independently associated with prolonged hospitalization in multivariate analysis were severe or critical COVID-19 and worse functional status at the time of hospital admission, referral from other institutions, acute neurological, acute surgical and social indications for admission vs admission indication of COVID-19 pneumonia, obesity, chronic liver disease, hematological malignancy, transplanted organ, occurrence of venous thromboembolism, occurrence of bacterial sepsis and occurrence of Clostridioides difficile infection during hospitalization. Patients requiring prolonged hospitalization experienced higher post-hospital discharge mortality (HR = 2.87, P < 0.001).
Conclusions
Not only severity of COVID-19 clinical presentation but also worse functional status, referral from other hospitals, certain indications for admission, certain chronic comorbidities, and complications that arise during hospital stay independently reflect on the need of prolonged hospitalization. Development of specific measures aimed at improvement of functional status and prevention of complications might reduce the length of hospitalization.
Supplementary Information
The online version contains supplementary material available at 10.1007/s41999-023-00787-w.
Keywords
COVID-19
SARS-CoV-2
Prolonged hospitalization
Comorbidities
issue-copyright-statement© European Geriatric Medicine Society 2023
==== Body
pmcIntroduction
Coronavirus disease 2019 (COVID-19) is an acute viral disease caused by Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Although majority of affected patients present with only mild respiratory symptoms, some patients may develop systemic inflammatory response to SARS-CoV-2 infection with severe life-threatening symptoms and multiorgan failure. Prior to wide-spread availability of vaccination, up to 15–20% affected patients developed acute respiratory insufficiency requiring oxygen supplementation and consequent hospitalization [1]. Thus, COVID-19 imposes significant challenges on healthcare systems due to its ability to overwhelm available infrastructure with high number of acutely ill and potentially unstable patients. Despite evident benefits of vaccination regarding reduced number of patients presenting with severe or critical form of the disease and improved outcomes of patients with breakthrough infections [2], vaccine hesitancy, inappropriate immune response to vaccination and waning effects of vaccination still remain substantial problems for controlling the pandemic [3–5].
Despite importance of hospital bed network during the pandemic, there are scarce data on factors affecting the length of hospital stay of COVID-19 patients available at the moment [6, 7]. In this study we aim to comprehensively investigate factors associated with prolonged hospitalization by evaluating large tertiary-center registry.
Methods
We have retrospectively analyzed data on 5959 concomitantly hospitalized COVID-19 patients hospitalized in the University hospital Dubrava, Zagreb, Croatia, in period from 3/2020 to 6/2021. During the study period hospital was completely repurposed to serve as a tertiary-level institution and regional referral center for most severe COVID-19 patients and for patients that required emergent medical care and were concomitantly SARS-CoV-2 positive. All patients tested positive for SARS-CoV-2, either on polymerase chain reaction (PCR) test, or on antigen test in the presence of compatible clinical symptoms. Patients were treated according to the contemporary guidelines and majority of them received oxygen supplementation, corticosteroids and LMWH thromboprophylaxis [8]. Demographic, laboratory and clinical data were obtained by evaluation of written and electronical medical documentation and are a part of the hospital registry project, “Registry of hospitalized COVID-19 patients in University Hospital Dubrava Respiratory center” (ClinicalTrials.gov identifier: NCT05151094). Demographic, laboratory and clinical parameters were evaluated at the time of hospital admission, whereas the clinical course of the disease was analyzed during hospitalization (mechanical ventilation, venous and arterial thrombotic events, major bleeding, bacterial sepsis, Clostridioides difficile (C. difficile) infection). Mortality was evaluated during hospitalization and after hospital discharge. COVID-19 severity was categorized as mild, moderate, severe and critical according to the World health organization (WHO) classification [9]. Intensity of COVID-19 symptoms was graded using the modified early warning score (MEWS) score [10]. The Eastern cooperative oncology group (ECOG) scale [11] was used to determine functional status of patients at the time of hospital admission. Comorbidities were evaluated as particular entities and were summarized using the Charlson comorbidity index (CCI) [12].
We defined prolonged hospitalization as duration of hospital stay > 21 day encompassing three-week period. We chose this approach to account for prescribed mandatory isolation period that variated individually for specific patients and lasted longest for 20 days in specific patient subgroups during the pandemic (immunocompromised patients).
The study was approved by the University Hospital Dubrava Review Board (Nm: 2022/2709–09).
Statistical methods
Normality of distribution of numerical variables was tested using the Kolmogorov–Smirnov test. Neither of numerical variables had normal distribution. They were presented as median and interquartile range (IQR) and were compared regarding prolonged hospitalization status using the Mann Whitney U test. Categorical variables were presented as frequencies and percentages and were compared regarding prolonged hospitalization status using the chi squared test. Post-hospital discharge survival analysis was based on the Kaplan–Meier method and was performed using the Cox–Mantel version of the log rank test [13, 14]. Multivariate predictors of prolonged hospitalization were assessed using the logistic regression. During the model building, all univariately associated variables, age and sex were considered via backwards approach (inclusion criteria P < 0.05, exclusion criteria P > 0.1) [15]. P values < 0.05 were considered as statistically significant. All analyses were performed using the MedCalc statistical software version 20.014 (MedCalc Software Ltd, Ostend, Belgium; 2022).
Results
Overview of patients’ characteristics and length of hospitalization
Among total of 5959 evaluated COVID-19 patients, there were 2613 (43.8%) female and 3346 (56.2%) male patients. Median age was 72 years, IQR (62–81). Median CCI was 4 points, IQR (3–6). Median ECOG score was 3 points, IQR (1–4).
Severity of COVID-19 on admission was mild in 560 (9.4%), moderate in 286 (4.8%), severe in 4202 (70.5%) and critical in 911 (15.3%) patients. Leading cause of hospital admission was COVID-19 pneumonia in 4299 (72.1%), other acute medical condition in 675 (11.3%), acute surgical condition in 508 (8.5%), acute neurological condition in 228 (3.8%), febrility without pneumonia in 140 (2.3%) and social and other reasons in 109 (1.8%) patients. Most patients came from their homes due to worsening of clinical condition (2899 (48.6%)), whereas 2490 (41.8%) were referred from other hospitals and 570 (9.6%) patients came from nursing homes.
Median length of hospitalization was 10 days, IQR (6–16). Majority of patients were hospitalized for less than 7 days (2103 (35.3%)) or 8–14 days (2158 (36.2%)), whereas hospitalization for 15–21 days was required in 899 (15.1%) and for > 21 days in 799 (13.4%) patients. Likelihood of death during hospitalization showed statistically significant trend of decrease with longer duration of hospitalization (P < 0.001) as shown in Fig. 1A. In other words, the fatal outcome determined the length of hospitalization more pronounced in patients hospitalized for a short time, while other factors increasingly influenced the prolonged course of treatment.Fig. 1 A In-hospital mortality stratified according to the length of hospitalization. B Post-hospital discharge mortality stratified according to the length of hospital stay
Factors associated with prolonged hospitalization
Associations of prolonged hospitalization (> 21 days) with demographic parameters and reasons for hospital admission are shown in Supplementary Table S1, patients’ comorbidities in Supplementary Table S2, laboratory parameters in Supplementary Table S3 and COVID-19 severity and clinical course during hospitalization in Supplementary Table S4.
In unadjusted analyses, prolonged hospitalization was significantly associated with origin of referral (referral from other institutions favoring prolonged hospitalization), leading cause of hospital admission (non-COVID-19 pneumonia related causes favoring prolonged hospitalization), higher comorbidity burden as assessed by CCI, presence of obesity, peripheral artery disease, gastroesophageal reflux/ulcer disease, chronic liver disease, liver cirrhosis, previous myocardial infarction, hematologic malignancy, prior organ transplantation, lower hemoglobin, higher red blood cell distribution width, lower absolute lymphocyte count, higher D-dimers, higher alkaline phosphatase and lower serum albumin levels, more severe COVID-19 presentation at the time of hospital admission, shorter duration of COVID-19 symptoms at the time of hospital admission, worse ECOG functional status on admission, presence of pneumonia and presence of bilateral pneumonia (P < 0.05 for all analyses). Prolonged hospitalization was also significantly associated with requirement for mechanical ventilation, occurrence of venous thromboembolic events, major bleeding, bacterial sepsis and C. difficile infection during hospital stay (P < 0.05 for all analyses). There were no significant associations of prolonged hospitalization with age, sex, other chronic metabolic comorbidities or laboratory parameters at the time of hospital admission.
The multivariate logistic regression model assessing independent predictors of prolonged hospitalization chosen among univariately significantly associated parameters is shown in Table 1. Factors that remained independently associated with prolonged hospitalization were severe or critical COVID-19 (OR = 1.95, P = 0.006) and worse functional status at the time of hospital admission (OR = 1.12, P = 0.015), referral from other institutions (OR = 1.34, P = 0.012), acute neurological (OR = 2.57, P = 0.036), acute surgical (OR = 2.63, P = 0.006) and social indications for admission vs indication of COVID-19 pneumonia (OR = 2.38, P < 0.001), obesity (OR = 1.28, P = 0.040), chronic liver disease (OR = 1.81, P = 0.029), hematological malignancy (OR = 2.89, P < 0.001), transplanted organ (OR = 2.27, P = 0.047), occurrence of venous thromboembolism (OR = 3.08, P < 0.001), occurrence of bacterial sepsis (OR 3.97, P < 0.001) and occurrence of C. difficile infection during hospitalization (OR = 6.69, P < 0.001).Table 1 Multivariate logistic regression model assessing independent predictors of prolonged hospitalization of COVID-19 patients (> 21 days)
Odds ratio 95% CI P value
Referral from other institution 1.34 1.06–1.69 0.012*
Acute neurological indication for admission vs pneumonia 2.57 1.06–6.20 0.036*
Acute surgical indication for admission vs pneumonia 2.63 1.32–5.24 0.006*
Social indication for admission vs pneumonia 2.38 1.44–3.93 < 0.001*
Obesity (BMI > 30 kg/m2) 1.28 1.01–1.62 0.040*
Chronic liver disease 1.81 1.06–3.10 0.029*
Hematological malignancy 2.89 1.71–4.88 < 0.001*
Transplanted organ 2.27 1.01–5.12 0.047*
Severe or critical COVID-19 1.95 1.20–3.15 0.006*
ECOG functional status 1.12 1.02–1.23 0.015*
Venous thromboembolism 3.08 2.19–4.34 < 0.001*
Bacterial sepsis 3.97 3.04–5.19 < 0.001*
C. difficile infection 6.69 4.50–9.96 < 0.001*
CI confidence interval, BMI body mass index, COVID-19 Coronavirus disease 2019, ECOG Eastern cooperative oncology group, C. Clostridioides
*Statistically significant at level P < 0.05
Long-term significance of prolonged hospitalization
Patients requiring prolonged hospitalization were statistically significantly more likely to experience higher post-hospital discharge mortality in comparison to shorter hospitalized patients (HR 2.87, 95% CI (1.86–4.39); P < 0.001) as shown in Fig. 1B.
Discussion
To the best of our knowledge, our study based on a large dataset of mostly severe or critical COVID-19 patients, who are mostly older persons and with high comorbidity burden, is the first to comprehensively investigate factors associated with prolonged hospitalization. As our data reveal, not only severity of COVID-19 symptoms at the time of presentation but also worse functional status, referral from other hospitals, certain indications for admission, certain chronic comorbidities, and complications that arise during hospital stay reflect on the need of prolonged hospitalization. Prolonged hospitalization may also be a negative indicator of patient outcome following their discharge from the hospital.
There are several points we would like to emphasize. Majority of hospitalized patient ended their hospitalizations during two weeks from hospital admission. Death as an important factor for duration of hospital stay in severe and critical COVID-19 patients dominantly affects early period of hospitalization, whereas other factors become increasingly more important during prolonged course of treatment. We deliberately chose to analyze prolonged hospital stay > 21 days to avoid both administrative reasons for prolonged hospitalization (mandatory isolation period among immunocompromised patients), as well as to ameliorate bias associated with factors prognostic of worse survival. In our opinion this is the optimal way to analyze this issue. Observed associations likely translate into other cut-offs for duration of hospital stay.
Age, sex and comorbidities, as well as severity of COVID-19 clinical presentation are known negative prognostic factors for survival during COVID-19 hospitalization [16]. On the opposite, factors associated with duration of hospital stay are less well understood and scarce clinical data exist on this issue. Factors like weaker muscle strength [17], higher acuity level of care [18] and use of specific drugs like remdesivir [19] are reported to be associated with longer COVID-19 hospitalization. Recent studies [6, 7] also addressed predictors associated with the length of stay of hospitalized COVID-19 patients but utilized a relatively short cutoff periods of > 3 and > 10 days which may not be completely comparable to the context of the current study, especially since death and administrative reasons may play important role during shorter periods. It should be noted that predictors of non-prompt recovery that were recognized in one of the aforementioned studies [7] were similar to predictors of prolonged hospitalization in our study including higher oxygen requirement, ICU admission, bacterial superinfections and shorter duration of symptoms at the time of hospital admission. As our data show, intensity of COVID-19 symptoms and functional impairment of patients are important predictors of hospitalization for > 21 days, especially if patients require mechanical ventilation support. Although adjusted OR for ECOG was of low magnitude, it represents obvious positive trend and increased risk of prolonged hospitalization of 12% with each point increase on ECOG scale. It is of particular interest that not the age or sex of patients per se, but comorbidity burden, particular comorbidities and the need to be hospitalized due to non-COVID-19 pneumonia related factors also significantly affect the need for prolonged hospital stay. This is especially true for chronic conditions that affect functional status of patients, require specific pharmacotherapies and more intensive nursing care like cardiovascular comorbidities, chronic liver disease and hematologic malignancies. Higher comorbidity burden and associated pharmacotherapy also predispose to complications like bacterial superinfections and bacterial sepsis, C. difficile infection, thromboembolic incidents and major bleeding, also reflecting on prolonged hospitalization. Raising awareness of independent predictors of prolonged hospital stay may help in development of specific measures targeted at modifiable factors. Examples could be improvement of functional status of patients by early mobilization, as well as prevention, early recognition and timely absolving of in-hospital complications.
Main limitations of our work are retrospective study design and single center experience. Study design limits conclusions about causality of observed associations. ECOG might be suboptimal measure of functional status in non-oncological patients. Information on important geriatric prognostic factors (i.e. cognition, nutrition, delirium, as well as standard comprehensive geriatric assessment and clinical frailty scale [20]) was not available in our study. These parameters have been shown to be more powerful in other settings than many of the variables significant in our study. Also, due to registry-level dataset, we had no access to information on the other potentially important predictors of prolonged hospitalization such as negativization of SARS-CoV-2 nasal swab [20] for a large proportion of analyzed patients, and we could not investigate them as predictors. Main strengths of our work are large real-life dataset of contemporary COVID-19 patients with mostly severe or critical clinical presentation upon admission, representative of tertiary-center experience.
In conclusion, we have determined a series of factors independently associated with prolonged hospitalization of COVID-19 patients, including severity of COVID-19 clinical presentation, worse functional status, referral from other hospitals, certain indications for admission, certain chronic comorbidities, and complications that arise during hospital stay. Development of specific measures aimed at improvement of functional status and prevention of complications might reduce the length of hospitalization.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 26 KB)
Acknowledgements
This paper is a part of the project “Registar hospitalno liječenih bolesnika u Respiracijskom centru KB Dubrava”/“Registry of hospitalized patients in University Hospital Dubrava Respiratory center” registered on the ClinicalTrials.gov website, identifier NCT05151094.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability
Data is available from corresponding author on reasonable request.
Declarations
Conflict of interest
All authors declare they have no conflict of interest.
Ethical approval
All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008. The study was approved by the University Hospital Dubrava Review Board (Nm: 2022/2709–09).
Informed consent
Due to retrospective nature of the study informed consent was not required and was waived by the Review Board.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Wu Z McGoogan JM Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in china: summary of a report of 72 314 cases from the chinese center for disease control and prevention JAMA 2020 323 13 1239 1242 10.1001/jama.2020.2648 32091533
2. Busic N Lucijanic T Barsic B Vaccination provides protection from respiratory deterioration and death among hospitalized COVID-19 patients: differences between vector and mRNA vaccines J Med Virol 2022 94 6 2849 2854 10.1002/jmv.27666 35187697
3. Bagić D Šuljok A Ančić B Determinants and reasons for coronavirus disease 2019 vaccine hesitancy in Croatia Croat Med J 2022 63 1 89 97 10.3325/cmj.2022.63.89 35230010
4. Kolarić B Ambriović-Ristov A Tabain I Vilibić-Čavlek T Waning immunity six months after BioNTech/Pfizer COVID-19 vaccination among nursing home residents in Zagreb Croatia Croat Med J 2021 62 6 630 633 10.3325/cmj.2021.62.630 34981696
5. Sertić Z Lucijanić M Bašić-Kinda S Non-myelofibrosis chronic myeloproliferative neoplasm patients show better seroconversion rates after SARS-CoV-2 vaccination compared to other hematologic diseases: a multicentric prospective study of krohem Biomedicines 2022 10 11 2892 10.3390/biomedicines10112892 36428459
6. El Halabi M Feghali J Bahk J A novel evidence-based predictor tool for hospitalization and length of stay: insights from COVID-19 patients in New York city Intern Emerg Med 2022 17 7 1879 1889 10.1007/s11739-022-03014-9 35773370
7. Ny P Kelsom C Chron A Factors associated with prompt recovery among hospitalised patients with coronavirus disease 2019 Int J Clin Pract 2021 75 11 e14818 10.1111/ijcp.14818 34487380
8. Ministarstvo zdravstva Republike Hrvatske. Smjernice za liječenje oboljelih od COVID-19, verzija 1 od 8. rujna 2020. 2020.
9. World Health Organization. Clinical management of COVID-19: interim guidance, 27 May 2020. Geneva: World Health Organization. Contract No.: WHO/2019-nCoV/clinical/2020.5. 2020.
10. Subbe CP Kruger M Rutherford P Gemmel L Validation of a modified early warning score in medical admissions QJM 2001 94 10 521 526 10.1093/qjmed/94.10.521 11588210
11. Oken MM Creech RH Tormey DC Toxicity and response criteria of the eastern cooperative oncology group Am J Clin Oncol 1982 5 6 649 655 10.1097/00000421-198212000-00014 7165009
12. 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 5 373 383 10.1016/0021-9681(87)90171-8 3558716
13. Lucijanic M Survival analysis in clinical practice: analyze your own data using an excel workbook Croat Med J 2016 57 1 77 79 10.3325/cmj.2016.57.77 26935618
14. Lucijanić M An update to the custom-made MS Excel workbook performing the log-rank test with extended functionality and a new original COVID-19 training data set Croat Med J 2021 62 5 531 533 10.3325/cmj.2021.62.531 34730896
15. Bazdaric K Sverko D Salaric I Martinovic A Lucijanic M The ABC of linear regression analysis: what every author and editor should know Europ Sci Edit 2021 47 e63780 10.3897/ese.2021.e63780
16. Piskač Živković N Lucijanić M Bušić N The associations of age, sex, and comorbidities with survival of hospitalized patients with coronavirus disease 2019: data from 4014 patients from a tertiary-center registry Croat Med J 2022 63 1 36 43 10.3325/cmj.2022.63.36 35230004
17. Tj I Tsai YL Cheng YY Predictors of basic activity in daily living and length of hospitalization in patients with COVID-19 Healthcare (Basel) 2022 10 8 1589 36011246
18. Almas A Mushtaq Z Moller J Acuity level of care as a predictor of case fatality and prolonged hospital stay in patients with COVID-19: a hospital-based observational follow-up study from Pakistan BMJ Open 2021 11 5 e045414 10.1136/bmjopen-2020-045414 34049912
19. Lucijanić M Bušić N Bistrović P Real-life experience with remdesivir for treatment of hospitalized coronavirus disease 2019 patients: matched case-control study from a large tertiary hospital registry Croat Med J 2022 63 6 536 543 10.3325/cmj.2022.63.536 36597565
20. Pagliuca R Cupido MG Mantovani G Absence of negativization of nasal swab test and frailty as risk factors for mortality in elderly COVID-19 patients admitted in long-term care facilities Eur Geriatr Med 2022 13 4 933 939 10.1007/s41999-022-00657-x 35661341
|
PMC010xxxxxx/PMC10140470.txt |
==== Front
Arab J Chem
Arab J Chem
Arabian Journal of Chemistry
1878-5352
1878-5379
The Author(s). Published by Elsevier B.V. on behalf of King Saud University.
S1878-5352(23)00401-X
10.1016/j.arabjc.2023.104939
104939
Original Article
Compounds from myrtle flowers as antibacterial agents and SARS-CoV-2 inhibitors: In-vitro and molecular docking studies
Barhouchi Badra a⁎
Menacer Rafik ab
Bouchkioua Saad a
Mansour Amira ab
Belattar Nadjah a
a Pharmaceutical Sciences Research Center CRSP, Constantine 25000, Algeria
b Centre de Recherche Scientifique et Technique en Analyses Physico-Chimiques CRAPC, BP 384, Zone Industrielle, Bou-ismail, Tipaza RP 42004, Algeria
⁎ Corresponding author.
28 4 2023
8 2023
28 4 2023
16 8 104939104939
21 11 2022
23 4 2023
© 2023 The Author(s)
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Graphical abstract
Plants and their related phytochemicals play a key role in the treatment of bacterial and viral infections, which inspire scientists to design and develop more efficient drugs starting from the phytochemical active scaffold. This work aims to characterize the chemical compounds of Myrtus communis essential oil (EO) from Algeria and to evaluate its in vitro antibacterial effect, as well as the in silico anti-SARS-CoV-2 activity. The chemical profile of hydrodistilled EO from myrtle flowers was determined using GC/MS analysis. The results showed qualitative and quantitative fluctuations and 54 compounds were identified including the main components: α-pinene (48.94%) and 1,8-cineole (28.3%) whereas other minor compounds were detected. The in vitro antibacterial activity of myrtle EO against Gram-negative bacteria was carried out by using the disc diffusion method. The best inhibition zone values ranged between 11 and 25 mm. The results revealed that Escherichia coli (25 mm), Klebsiella oxytoca (20 mm) and Serratia marcescens (20 mm) are the most susceptible strains to the EO which is endowed with a bactericidal effect. Furthermore, the antibacterial and anti-SARS-CoV-2 activities were investigated by the means of molecular docking (MD) study, in addition to ADME(Tox) analysis. The phytochemicals were docked against four targets: E. coli topoisomerase II DNA gyrase B (PDB: 1KZN), SARS-CoV-2 Main protease (PDB: 6LU7), Spike (PDB: 6ZLG) and angiotensin-converting enzyme II ACE2 (PDB: 1R42). The MD investigation revealed that 1,8-cineole could be the main phytochemical associated with the antibacterial activity of EO; s-cbz-cysteine, mayurone and methylxanthine were found the most promising phytochemicals against SARS-CoV-2; The ADME(Tox) analysis has shown their good druggability with no Lipinski’s rule violation.
Keywords
Myrtus communis
Antibacterial activity
E.coli
Molecular docking
SARS-CoV-2
==== Body
pmc1 Introduction
Plants are the richest host for novel drug discovery and development. They have played a key role in the treatment of multiple ailments among them bacterial and viral infections (Rahman and Sarker, 2020, Musarra-Pizzo et al., 2021). The investigation of natural compounds from plants for therapeutic purposes is one line of scientific research followed to develop effective antiviral and antibacterial treatments, particularly with the emergence of the new threat called antimicrobial resistance (Rahman and Sarker, 2020, De Oliveira et al., 2020). In recent decades, the antimicrobial effect of many plants and isolated natural compounds has been tested with promising results (Lewis., 2017). The antibacterial and antiviral properties of plant products present potential for manufacturing and developing drugs that can reduce the pathogenicity of the microorganisms by neutralizing and blocking their absorption into the cell, as well as by inhibiting their reproduction in the cell (Lewis, 2017, Maginnis, 2018).
Essential oils produced by plants are known to possess antimicrobial activity; they have been traditionally used for respiratory tract infections (Inouye et al., 2001, Maruzzella and Sicurella, 1960, Shubina et al., 1990). In the medicinal field, inhalation therapy of essential oils has been used to treat acute and chronic bronchitis (von Schindl., 1972) and acute sinusitis (Federspil et al., 1997).
Myrtus communis or common myrtle (family: Myrtaceae), native to the Mediterranean region is a wild aromatic plant that many works have demonstrated the biological activities of myrtle essential oils (EOs). According to the literature, myrtle exhibited several pharmacological activities including antioxidative, anticancer, anti-diabetic, antibacterial, antifungal, neuroprotective, hepatoprotective and antiviral activities (Alipour et al., 2014), especially against human papillomavirus (HPV), herpes simplex virus type 1 (HSV-1) and tobacco mosaic virus (TMV)) (Nikakhtar et al., 2018, Moradi et al., 2011). Different parts of the myrtle particularly its berries, leaves and flowers have long been used as a remedy for respiratory complaints like cough, gastrointestinal disorders, urinary diseases and skin ailments, as well as for microorganisms inactivation and wound healing (Alipour et al., 2014, Aleksic and Knezevic, 2014). The chemical composition of myrtle EO may change according to several factors; nevertheless, it was constant in practically all EOs, the presence of 1,8-cineole and α-pinene, as main components (Hennia et al., 2019). Although the contribution of the components on the biological activities was not determined, they were generally attributed to the presence of the main components (1,8-cineole, α-pinene, eugenol, methyleugenol, myrtenyl acetate, among other components) (Hennia et al., 2019). Myrtle parts have been broadly scrutinized by the scientific community where in Algeria; there is still a lack of scientific knowledge pertaining to the chemical composition and biological activities of myrtle flowers.
Traditional knowledge and experiential databases derived from clinical practice are instrumental in increasing the success rate of drug discovery by reducing the time consumed, money spent and toxicity occurrence which are the three main hurdles in drug development, when compared with the conventional approach adopting random screening and chemical synthesis. Current technology advancements have been able to land genetic, molecular, structural, and functional features in order to find effective targets against viruses (Wang et al., 2009). Computational analysis allows drug discovery from synthetic and natural compounds especially phytochemicals coming from medicinal plants, it makes predictions about the possible therapeutic effects of these compounds. Among the computational methods, molecular docking is most commonly used in the structure-based drug design for prediction of protein–ligand binding sites with binding affinity score (Lengauer et al., 1996).
In this study, myrtle flowers essential oil was experimentally investigated, firstly by identifying the existing phytochemicals, secondly by evaluating the essential oil antibacterial activity in vitro against 20 bacteria. Furthermore, the antibacterial and the eventual antiviral activity of the identified phytochemicals were rationalized by molecular docking against E. coli topoisomerase II DNA gyrase B, SARS-CoV-2 Main protease (Mpro), Spike and angiotensin-converting enzyme II ACE2 (PDB: 1R42). Absorption, Distribution, Metabolism, Excretion (ADME) and toxicity properties were also highlighted.
2 Methods and materials
2.1 Experimental details
2.1.1 Raw material
Fresh flowers of Myrtus communis L. were gathered during its flowering stage from northeastern Algeria, at the region of Annaba city (located at 36°53′59″ N, 7°46′00″ E). The voucher specimens were identified in the Biology Department, Badji Mokhtar University, Annaba. The plant material was dried in the shade at room temperature and conserved until the extraction process.
2.1.2 Essential oil isolation
According to the method recommended in the European Pharmacopoeia (2002), the volatile oils were extracted from 100 g of dried myrtle flowers by hydrodistillation for 2 h using a Clevenger-type apparatus. Anhydrous sodium sulfate was used to dry the aqueous phase and the obtained essential oil was filtered and kept cold at 4 °C in a hermetically sealed opaque bottle until the chemical analyzes were conducted. The EO yield was calculated using the dry weight of plant material based on the following equation: EO (% v/w) = volume of oil (mL)/weight of sample (g) × 100.
2.1.3 GC-MS analysis
The chemical composition of EO was analyzed by gas chromatography–mass spectrometry GC–MS (Agilent 6890 N; Agilent Technologies), equipped with a capillary HP-5MS column (60 m length, 0.25 mm diameter, 0.25 mm film thickness), and coupled with a mass selective detector. The electron impact ionization was 70 eV. The carrier gas was Helium with a flow rate of 1.2 mL/min and the injection (μL) was conducted manually in the split mode (1:50 split ratio). The operating conditions were as follows: the oven temperature was programmed for 1 min at 100 °C, increased from 100 to 280 °C at a rate of 5 °C/min, and then set at 280 °C for 25 min, the temperatures of the injector and detector were 250 and 310 °C, respectively. However, the mass-spectrometer conditions were the following: injection of 2 μL aliquot of the sample where the scan time and mass range were 1 s and 40–300 m/z, respectively. The components of myrtle essential oil were identified by a comparison of the fragmentation patterns in the mass spectra with those stored in the GC-MS databases and those from two libraries: the Wiley Registry of Mass Spectral Data 7th edition (Agilent Technologies) and the library of the national institute of standards and technology 05 MS (NIST). In addition, the percentages of the compounds were determined from their peak areas.
2.1.4 Bacterial strains
In this study, two classified bacteria ATCC-American Type Culture Collection: Escherichia coli (EC) ATCC 25922 and Klebsiella pneumoniae (KP) ATCC 700603), and eighteen pathogenic microorganisms isolated from human urine samples were used as indicators including: Escherichia coli (EC), Klebsiella pneumoniae (KP), Klebsiella oxytoca (KO), Shigella sonnei (SS), Serratia marcescens (SM), Serratia fonticola (SF), Acinetobacter baumannii (AB), Citrobacter koseri (CK), Citrobacter freundii (CF), Enterobacter aerogenes (EA), Enterobacter cloacae (EL), Enterobacter intermedius (EI), Enterobacter sakazakii (ES), Proteus mirabilis (PM), Proteus vulgaris (PV), Morganella morganii (MM), Salmonella typhimurium (ST) and Salmonella sp. (S). Nutrient agar was used as the growth media.
2.1.5 Essential oil antibacterial assay
Solid diffusion method: An agar diffusion disc method was used to test the selected bacteria's susceptibility to the essential oil (Prabuseenivasan et al., 2006). The inoculum (DO=0.1/625 nm) was streaked into agar plates using a sterile swab after the Mueller Hinton Agar (MHA) had solidified, a sterile filter disc with a 5 mm diameter (Whatman paper N°3) was inserted on the surface of the MHA. Then, 10 μL of the essential oil (crude EO and diluted EO with equivalent concentrations: EO/DMSO: 50/50 v/v) was dropped onto each disc and left for 30 min at room temperature for antibacterial agent diffusion. The plates were incubated for 18 to 24 h at 37 °C. DMSO was employed as a negative control and Gentamicin (GEN) was used as positive control. The essential oil's effectiveness was determined by measuring the diameter of the zone of bacterial growth inhibition above the disc and recording the diameter in mm. An essential oil-inducing inhibition zone ≥ 3 mm around the disc was considered as antibacterial. All tests were performed in triplicate.
Macrodilution method: The Minimum Inhibitory Concentration (MIC) was defined as the lowest concentration of the total essential oil at which the microorganism does not demonstrate visible growth (Wikler, 2006). Serial dilutions of myrtle essential oil in dimethylsulfoxide (DMSO) were prepared (10, 5, 2.5, 1.25 and 0.625 mg/mL). Each dilution was transferred in each tube containing the Mueller Hinton Broth (MHB) medium and the tested bacterial inoculum. The inoculated tubes were then incubated at 37 °C for 24 h. After incubation, the first tube without bacterial visible growth was determined as Minimum Inhibitory Concentration (MIC).
2.2 Computational details
2.2.1 Docking simulation and ADME(Tox)
The molecular docking investigation aims to predict the antibacterial (E.coli) and anti-SARS-CoV-2 activities of the main compounds coming from myrtle flowers. Each compound structure, reported in Table 1 was pre-optimized using the Steepest Descent algorithm with a convergence criterion of 10 e-6 at the MMFF94s force field theory level. Then, the best conformer was kept after a conformational search and after that an optimization was performed by the same level of theory as implemented in AVOGADRO software (Hanwell et al., 2012). The final optimization was done by MOPAC software at PM7 theory level (Stewart, 1990). To elucidate the mechanism by which the extracted molecules induce antibacterial activity, the inhibitory activities of identified compounds were examined against DNA gyrase. For that, the X-ray crystallographic structure of E. coli topoisomerase II DNA gyrase B along with co-crystallized ligand CBN (PDB: 1KZN) was utilized. The evaluation of the anti-SARS-CoV-2 activity was performed using the following proteins: SARS-CoV-2 Main protease (Mpro) (PDB: 6LU7) (Jin et al., 2020) also named chymotrypsin-like protease (3CLpro), Spike (PDB: 6ZLG) and angiotensin-converting enzyme II ACE2 (PDB: 1R42). The proteins were prepared by removing all solvent molecules and co-crystallized ligands and the molecular docking protocol was validated by redocking using the published crystal structures of protein–ligand complexes. The root-mean square deviations (RMSD) between the conformations of the ligands from the X-ray crystal structure and those from AutoDock Vina (Trott and Olson, 2010) were < 2 A˚, hence the AutoDock Vina docking protocol is adequate to reproduce and to predict experimental findings. The docking of each structure against (PDB ID: 1KZN, 6LU7, 6ZLG, 1R42), was performed by Autodock Vina as implemented in PyRx (Dallakyan and Olson, 2015) in the search spaces and sizes (1KZN: center × = 17.11, y = 28.14, z = 31.40; size × = 25.0, y = 25.0, z = 25.0); (6LU7: center × = -26.28, y = 12.60, z = 58.96; size × = 51.30, y = 66.93, z = 59.57); (6ZLG: center × = –32.36, y = 25.78, z = 21.01; size × = 25, y = 25, z = 25); (1R42: center × = 51.13, y = 74.24, z = 28.024; size × = 18.70, y = 16.58, z = 18.48). The visualization of the results were depicted by Discovery Studio Visualizer software (Biovia, 2017). Finally, the Adsorption, Digestion, Metabolism, Excretion and Toxicity (ADME-Tox) study of the shortlisted compounds was performed using SwissADME server (Daina et al., 2017) and Pro-Tox II (Banerjee et al., 2018).Table 1 Chemical composition of essential oil extracted from myrtle flowers.
Peak Compound RT (min) Area (%) RI MM MF
1 Isobutyl isobutyrate 7.969 0.06 808 144.21 C8H16O2
2 α-Thujene 8.337 0.30 869 136,23 C10H16
3 α-Pinene 8.639 48.94 919 136,23 C10H16
4 β-Pinene 8.923 0.57 966 136,23 C10H16
5 1-Allyl tricyclo [4.1.0(2,7)] heptane 9.092 0.06 994 134.21 C10H14
6 Ocimene 10.252 0.15 1186 136,23 C10H16
7 α-Phellandrene 10.621 0.38 1247 136.24 C10H16
8 δ-3-Carene 10.784 0.48 1274 136,23 C10H16
9 α-Terpinene 11.020 1.76 1313 136,23 C10H16
10 1,8-Cineole 11.503 28.3 1393 154,24 C10H18O
11 γ-Terpinene 12.301 0.65 1525 136.23 C10H16
12 α-Terpinolene 13.183 0.75 1671 136.23 C10H16
13 Guajol 13.527 0.04 1728 124.13 C7H8O2
14 Linalol 13.654 2.37 1749 154,24 C10H18O
15 α-Campholenal 14.319 0.09 1859 152.23 C10H16O
16 2,5-Octadiene 14.754 0.05 1931 110.19 C8H14
17 2-Methyl-1,3-pentadiene 14.814 0.06 1941 82.14 C6H10
18 α-Phellandren-8-ol 15.678 0.06 2084 152.23 C10H16O
19 4-Terpineol 15.908 0.22 2122 154.24 C10H18O
20 α-Terpineol 16.379 2.02 2200 154,25 C10H18O
21 Cyclohexene 17.364 0.07 2363 82,143 C6H10
22 Carvon 17.799 0.43 2435 150,21 C10H14O
23 Linalyl acetate 18.016 0.40 2471 196.29 C12H20O2
24 Geraniol 18.131 0.60 2490 154.24 C10H18O
25 2-Undecanone 19.110 0.11 2652 170.29 C11H22O
26 S-cbz-cysteine 19.243 0.06 2674 178.22 C11H14O2
27 o-Acetyl-p-cresol 19.932 0.22 2788 150.17 C9H10O2
28 Exo-2-hydroxycineole acetate 20.367 0.11 2860 212.28 C12H20O3
29 2,4-Dimethyl-2,4-hexadiene 20.790 0.06 2930 110.19 C8H14
30 Eugenol 21.001 2.91 2965 164,20 C10H12O2
31 Geranyl acetate 21.503 2.65 3048 196.29 C12H20O2
32 2- Phenylbutyric acid 21.617 0.11 3067 164.20 C10H12O2
33 Thymoquinone 21.738 0.12 3087 164.20 C10H12O2
34 Methyleugenol 22.137 1.40 3153 178,22 C11H14O2
35 β-Caryophyllene 22.451 0.65 3205 204.36 C15H24
36 p-Ethoxyanisole 22.929 0.19 3284 152.19 C9H12O2
37 α-Humulene 23.321 0.17 3349 204.35 C15H24
38 Germacrene D 24.004 0.09 3462 204.35 C15H24
39 Aromadendrene 24.149 0.12 3486 204.35 C15H24
40 Ledene 24.367 0.12 3522 204.35 C15H24
41 Nerol 24.693 0.05 3576 154,24 C10H18O
42 Methylxanthine 24.916 0.26 3613 166.13 C6H6N4O2
43 Mayurone 25.478 0.14 3706 204.308 C14H20O
44 1-{2-[3-(2-Acetyloxiran-2-yl)-1,1-dimethyl propyl]cycloprop-2-enyl}ethanone 25.738 0.03 3749 236.30 C14H20O3
45 β-Selinene 25.841 0.62 3766 204.18 C15H24
46 Caryophyllene oxide 26.481 0.66 3872 220.35 C15H24O
47 2-Methyl-2 cyclopentenone 27.104 0.22 3975 96.13 C6H8O
48 4,6-Dimethyl-2-amino-1,3-diazine 27.230 0.15 3996 123.15 C6H9N3
49 1,4-Dihydrophenanthrene 27.412 0.18 4026 180.24 C14H12
50 Isolimonene 28.191 0.18 4155 136.23 C10H24
51 N-acetylpiperidone 29.490 0.30 4370 141.17 C7H11NO2
52 2-Nitro-1-decen-4-yne 29.744 0.07 4412 181.23 C10H15NO2
53 Aromadendrene oxide-(1) 30.705 0.15 4571 220.35 C15H24O
54 2-Butanoylthiazole 31.538 0.06 4709 155.21 C7H9NOS
RT (min): Retention time in minute, Area (%): Percentage of each compound, RI: Retention indices.
MM: Molecular Mass (g/moL). MF: Molecular formula.
3 Results and discussion
3.1 Chemical composition results
A Clevenger-type apparatus was used for the isolation of EOs of myrtle, with the following yield: 2.8 percent (v/w). Table 1 summarizes the identified components of myrtle flowers essential oil (n = 54), their percentages, retention times (Rt)and their associated retention indices (RI). Fifty-four components were identified in flowers, representing 89.65% of the total essential oil. The major constituents of the flower essential oil were α-pinene and 1,8-cineole with: 48,94% and 28.3%, respectively. Other representative compounds were detected as eugenol (2.91%), linalol (2,37%), geranyl acetate (2.65%) and α-terpineol (2,02%). Previous studies have reported the chemical composition of myrtle flowers essential oil (Aidi Wannes et al., 2010, Dhifi et al., 2020, Bouzabata et al., 2015, Jerkovic et al., 2002, Santana et al., 2014). However, the majority of compounds detected in this study coincide in earlier research, with some variations in ratios and the lack of some minority compounds. These differences could be attributable to a variety of circumstances, including where the species was grown, harvested, or how the oil was extracted. Similarly to our results, a Tunisian study conducted on myrtle flowers, confirmed that α-pinene and 1,8-cineole were reported as the main constituents (Dhifi et al., 2020, Djenane et al., 2011). Moreover, another Tunisian study showed that the essential oil isolated from myrtle flowers is rich in α-pinene and 1,8-cineole with the following yields: (22.50–15.15%) and (17.53–12.70%), respectively (Aidi Wannes et al., 2010). The chemical heterogeneity in myrtle essential oils appears to be dependent on their organ (leaves, berries, flowers), geographical origin, collection season and edaphoclimatic circumstances (Ben Hsouna et al., 2017, Pirbalouti et al., 2014, Brada et al., 2012).
3.2 Antibacterial activity results
The initial screening of in vitro antibacterial activity of M. communis EOs against twenty Gram-negative bacteria was carried out by using the disc diffusion method. Thus, in order to obtain more precise data about the antibacterial properties of the tested EOs, Minimum Inhibitory Concentrations (MICs) were determined using the macrodilution broth method. Table 2 showed that the myrtle essential oil had substantial antibacterial activity on all tested bacteria. The best inhibition zone diameter values (ID) were in the range of 11–25 mm. The MICs values of the myrtle essential oil ranged from 0.6 to 2.5 mg/mL. On the whole, Gram-negative bacteria were more sensitive to the crude essential oil than the diluted one with differentiated effects presented by eight sensible (ID≥10 mm), six intermediate (ID=10 mm) and six resistant (ID≤10 mm) strains. As illustrated in Fig. 1 , the tested crude EO proved significantly efficient, with pronounced inhibitory action against eight Gram-negative strains with the following order of sensitivity:E.coli>E.coliATCC 25922 > K.oxytoca > S.marcescens > P.mirabilis > E.aerogenes > C.freundii > P.vulgaris. In addition, the myrtle oil was more potent than gentamicin against E.coli, M.morganii, P.mirabilis and P.vulgaris. However, it is important to consider that the differences in the sensitivity of the strains used in each study may justify the differences concerning the activity of the tested agent. The best zone of inhibition of 25 mm was recorded against E.coli, while the lowest one (08 mm) appeared against four microorganisms including: E.cloacae, M.morganii, S. sonnei and Salmonella sp. Consequently, our previous study showed that the leaves essential oil isolated from the same species of myrtle was characterized by a close chemical profile and was also colicidal with the best inhibition zone (35 mm) (Barhouchi et al., 2016). In agreement with our research, Algerian myrtle oil exhibited a remarkable effect against Escherichia coli with inhibition diameter greater than 20 mm, while it showed a low to moderate antibacterial capacity against other strains in comparison with the common antibacterial drug gentamicin (Boussak et al., 2022). These results justify the traditional use of the flowers of this plant species against many bacteria-related disorders including: cough, lung complaints, gastrointestinal disorders, wounds..etc. According to literature, Turkish reports proved the antibacterial effectviness of myrtle flowers EOs against two microorganisms: Parvimonas micra (Gram positive strain) and Aggregatibacter actinomycetemcomitans (Gram negative strain) (Gursoy et al., 2009). In Algeria, myrtle essential oils isolated from flowers were confirmed only as antifungal agents (Bouzabata et al., 2015).Table 2 Minimum Inhibitory Concentrations MIC (mg/mL) and inhibition zone diameters ID (mm) of myrtle essential oil in comparison with gentamicin.
Bacterial strain Diluted EO (ID) Crude EO (ID) Gentamicin GEN (ID) Crude EO (MIC)
Escherichia coli EC (ATCC 25922) 15 (S) 23 (S) 30 (S) 2.5
Klebsiella pneumoniae KP (ATCC 700603) 09 (R) 10 (I) 19 (S) 2.5
Escherichia coli EC 10 (I) 25 (S) 21 (S) 1.25
Acinetobacter baumannii AB 05 (R) 10 (I) 11 (S) 1.25
Citrobacter freundii CF 05 (R) 11 (S) 22 (S) 2.5
Citrobacter koseri CK 05 (R) 10 (I) 20 (S) 2.5
Enterobacter aerogenes EA 10 (I) 18 (S) 21 (S) 0.6
Enterobacter cloacae EL 05 (R) 08 (R) 08 (R) 1.25
Enterobacter intermedius EI 09 (R) 10 (I) 19 (S) 1.25
Enterobacter sakazakii ES 09 (R) 10 (I) 22 (S) 1.25
Klebsiella pneumoniae KP 09 (R) 10 (I) 16 (I) 1.25
Klebsiella oxytoca KO 16 (S) 20 (S) 21 (S) 2.5
Morganella morganii MM 05 (R) 08 (R) 06 (R) 1.25
Proteus mirabilis PM 06 (R) 19 (S) 17 (I) 2.5
Proteus vulgaris PV 08 (R) 11 (S) 07 (R) 0.6
Salmonella sp. S 08 (R) 08 (R) 15 (R) 1.25
Salmonella typhimurium ST 05 (R) 09 (R) 13 (R) 1.25
Shigella sonnei SS 08 (R) 08 (R) 20 (S) 0.6
Serratia marcescens SM 08 (R) 20 (S) 26 (S) 1.25
Serratia fonticola SF 05 (R) 08 (R) 07 (R) 1.25
Negative control (DMSO) – – – –
Inhibition zone includes the diameter of the disk (5 mm). (05): No inhibition zone, ID: Inhibition zone diameter (mm),
(EO): Essential Oil, (MIC): Minimum Inhibitory Concentration (mg/mL), ATCC: American type culture collection,
(DMSO): Dimethylsulfoxide (negative control), (-): No inhibition zone of DMSO against all strains, GEN: Gentamicin (positive control), (R): Resistant, (I): Intermediate, (S): Sensible.
Fig. 1 Inhibition zone diameters (mm) of myrtle essential oil in comparison with gentamicin
The antibacterial activity of the essential oils investigated in the present study may be attributed to their major constituents namely: α-pinene and 1,8-cineole. As presented in previous studies, individual components of the EO such as α-pinene demonstrated a potent antibacterial activity (Stojkovic et al., 2008, Dhifi et al., 2020, Zanetti et al., 2010). It was also proven that 1,8-cineole exhibited also a strong antibacterial activity with minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) values in the range of 0.37 to 11.75 mg/mL and 0.73 to 11.75 mg/mL, respectively (Randrianarivelo et al., 2009). However, other minor constituents of myrtle can act synergistically with major components and play a key role in the observed antimicrobial activity (Dhifi et al., 2020). Hence, the essential oil of Eugenia jambolana, which has α-pinene as a primary constituent, presented synergistic effects when associated with gentamicin against E. coli, but not against P. aeruginosa (Pereira et al., 2017). Overall, the resistance of Gram-negative bacteria is attributed to the presence of an outer membrane with hydrophilic polysaccharide chains that operate as a barrier against hydrophobic essential oils (Mann et al., 2000). Thus, essential oils and their components have been demonstrated to act especially on bacterial membranes (Cox et al., 2000). In particular, monoterpenoids components would enhance the permeability of the cytoplasmic membrane by changing the order of proteins incorporated into the membrane, limiting cellular respiration, ion transportand nutritional absorption (Reichling et al., 2009). Further studies are required to identify the myrtle components responsible for its antimicrobial activity, to clarify how they work on their own or in combination with antibiotics and finally to ascertain the antiviral activity of these oils and/or their chemical constituents.
3.3 Molecular docking and ADME(Tox) results
The main goal of molecular docking research is to find an optimal conformation for a protein and a ligand with relative orientation between them which minimizes the total system's free energy. It is commonly accepted that the lower the predicted binding free energy, the more effective the binding and thus the more stable the protein–ligand complex is. The heat map presented in Table 3 demonstrates the range of the binding energies results obtained by the molecular docking. It gives an overview about the whole 54 extracted phytochemicals and their relative binding energies in colors, the values increase from green and yellow to red by column (target). In the last four rows of Table 3, the binding energies of Nelfinavir (Bolcato et al., 2020), Nirmatrelvir (Hammond et al., 2022), Gentamicin_C2 (Abdellattif et al., 2021) and Lopinavir (Bolcato et al., 2020) were added as positive controls for the following targets: 1R42, 6LU7, 1KZN and 6ZLG respectively. The three-dimensional representation of the co-crystal top phytochemicals having the lowest binding energies against E.coli (PDB: 1KZN); SARS-CoV-2 MPro (PDB: 6LU7); SARS-CoV-2 Spike (PDB: 6ZLG); ACE2 (PDB: 1R42) is shown in Table 4 . The optimal binding energy of phytocompounds was compared with those reported by other researchers (Table 5 ) (Sisakht et al., 2021, Hakmi et al., 2020, Prasanth et al., 2021, Kandsi et al., 2022, Jianu et al., 2021). 1,8-cineole (-7.9 Kcal/Mol) has shown the lowest binding energy against E.coli (PDB: 1KZN) among all the 54 phytochemicals in addition to several phytochemicals i.e. trans-β-Terpinyl Butanoate, Carvyl acetate reported for Dysphania ambrosioides (L.) (Kandsi et al., 2022) and Mentha × smithiana respectively (Jianu et al., 2021), as shown in Table 5, it also has a lower energy compared to the positive control Gentamicin_C2 (-6.0 Kcal/Mol); 1,8-cineole/1KZN complex has only hydrophobic interactions with the residues ALA47 (4.68 Å), ALA47 (5.43 Å), VAL43 (5.45 Å), VAL167 (5.26 Å), ILE90 (5.22 Å), VAL120 (5.11 Å), ILE78 (5.17 Å) and ILE78(4.22 Å). S-cbz-cysteine has given the best binding energy (-6.0 Kcal/Mol) against SARS-CoV-2 MPro (PDB: 6LU7) between the compounds found in Myrtus communis L; the S-cbz-cysteine/6LU7 complex has the hydrogen bond interactions with the residues GLY143 (2.38 Å), SER144 (2.10 Å), LEU141 (2.71 Å) and hydrophobic interactions with the residues MET49 (4.64 Å), HIS41 (4.72 Å), MET165 (5.64 Å), HIS163 (1.65 Å), LEU141 (2.83 Å); In other research reports, some phytochemicals have given better binding energies such as Ginkgolide M (-11.2 Kcal/Mol) found in the nutshells of the Ginkgo biloba tree (Sisakht et al., 2021) in addition to the positive control Nirmatrelvir (-7.4 Kcal/Mol). Mayurone gave the best binding energy (-6.8 Kcal/Mol) against SARS-CoV-2 Spike (PDB: 6ZLG); Mayurone/6ZLG complex has the hydrogen bond interactions with the residues ARG355 (5.75 Å), SER514 (2.98 Å) and hydrophobic interactions with the residues PRO426 (5.44 Å) and PHE464 (5.01 Å); whereas some reported phytochemicals gave a better binding energy values such as Cinnamtannin-B1 (-10.2 Kcal/Mol) and Kaempferol (-8.7 Kcal/Mol) (Prasanth et al. 2021), in contrast, it has a better binding energy compared to its control Lopinavir (-5.1 Kcal/Mol). Methylxanthine gave the lowest binding energy (-5.0 Kcal/Mol) among the studied phytochemicals against ACE2 (PDB: 1R42); the Methylxanthine/1R42 complex has the hydrogen bond interactions with the residues ALA348 (2.56 Å), GLU37 (2.15 Å) and GLU402 (2.53 Å) in addition to the hydrophobic interactions with the residues ALA348 (4.59 Å), HIS401 (5.40 Å), ALA348 (1.86 Å), HIS401 (3.58 Å) and HIS378 (3.88 Å); some reported phytochemicals give a better binding energy values such as Bicuculline (-9.9 Kcal/Mol) (Xu et al., 2021) and Theaflavin (-8.6 Kcal/Mol) (Emon et al., 2021), in addition to the positive control Nelfinavir (-9.3 Kcal/Mol). s-cbz-cysteine, mayurone and methylxanthine were found the most promising phytochemicals against SARS-CoV-2; The ADME(Tox) analysis has shown their good druggability with no Lipinski’s rule violation.Table 3 Heat map of recorded docking scores (binding free energy in kcal/mol) of the essential oil of myrtle flowers components.
Table 4 Three-dimensional representation of the co-crystal top phytochemicals with E.coli (PDB: 1KZN); SARS-CoV-2 MPro (PDB: 6LU7); SARS-CoV-2 Spike (PDB: 6ZLG); ACE2 (PDB: 1R42): (a) interaction residues and (b) hydrogen bonding pocket.
Target/Top compound (Score) 2D interaction residues Hydrogen bonding surface
1,8-cineole/1KZN (-7.9 kcal/mol)
S-cbz-cysteine/6LU7 (-6.0 kcal/mol)
Mayurone/6ZLG (-6.8 kcal/mol)
Methylxanthine/1R42 (-5.0 kcal/mol)
Table 5 Interactions of E.coli (PDB: 1KZN); SARS-CoV-2 MPro (PDB: 6LU7); SARS-CoV-2 Spike (PDB: 6ZLG); ACE2 (PDB: 1R42) amino acid residues with the studied compound.
Compound PubChem ID Target Vina Score (Kcal/Mol) Residues in interaction* Ref
1,8-Cineole 1KZN −7.9 Thr165, Ile78, Ile90, Val120, Val43, Ala47, Val167, Ala47, Ile78 This work
trans-β-Terpinyl Butanoate −6.4 Thr165, Val120, Met91, Gly77, Arg136, Glu50, Ala47, Asn46, Asp73, Pro79, Ile78, Arg76, Val43,, Val167 (Kandsi et al., 2022)
Carvyl acetate −6.8 – (Jianu et al., 2021)
s-cbz-cysteine 6LU7 −6.0 Gly143, Ser144, Leu141, Met49, His41, Met165, His163 This work
Ginkgolide M 46173836 −11.2 Gly143, His163,Cys145, Glu166, Phe140, Asn142 (Sisakht et al., 2021)
Mayurone 538435 6ZLG −6.8 Arg355, Ser514, Pro426, Phe464 This work
Cinnamtannin-B1 −10.2 Phe390, Asn394, Arg393, Phe40, Trp349, Thr347 (Prasanth et al. 2021)
Kaempferol −8.7 Asp350, Tyr385, Asp382, His345, Hia374, Glu375, His378, His401, His378 (Prasanth et al. 2021)
Methylxanthine 1R42 −5.0 Glu375, Glu402, Ala348, Glu375, Glu402, Ala348, His378 This work
Bicuculline −9.9 – (Xu et al., 2021)
Theaflavin −8.6 Thr371, Arg518, Glu406, Glu375, Leu370 (Emon et al., 2021)
*The residues associated with hydrogen bonds are in bold.
4 Conclusion
Myrtle flowers have long been used as a remedy for respiratory complaints, as well as microorganisms inactivation. The phytochemical investigation showed that the EOs contained similar major components as previously reported by other studies although their percentages varied. The main compounds obtained in the flowers essential oil of Myrtus communis were: α-pinene, 1,8-cineole, eugenol, linalol, geranyl acetate and α-terpineol. The myrtle essential oil displayed significant antibacterial activity and showed a pronounced effect against Gram negative bacteria in comparison to the commercial drug. The bacterium most sensitive to the effect of the essential oil was Escherichia coli. The molecular docking investigation showed that 1,8-cineole is the best inhibitor against E. coli topoisomerase II DNA gyrase B and could be the main phytochemical associated with the antibacterial activity of EO. Among the EO compounds, s-cbz-cysteine, mayurone and methylxanthine were found as the most promising against SARS-CoV-2 by the molecular docking analysis. This work opens the gate to additional preclinical and clinical research of edible EOs to alleviate certain respiratory tract illnesses.
CRediT authorship contribution statement
Badra Barhouchi: Writing – original draft, Conceptualization, Methodology, Resources, Data curation, Writing – review & editing. Rafik Menacer: Conceptualization, Methodology, Resources, Data curation, Software, Investigation, Formal analysis, Visualization, Validation, Writing – review & editing. Saad Bouchkioua: Software, Investigation, Formal analysis, Visualization. Amira Mansour: Project administration. Nadjah Belattar: Data curation.
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.
Acknowledgements
This study was supported by the Algerian Ministry of Higher Education and Scientific Research (MESRS) and the General Directorate of Scientific Research and Technological Development (DGRSDT). Many thanks are also addressed to Pr. Abdelhamid Djekoun, director of the Pharmaceutical Sciences Research Center (CRSP), for his great availability.
Peer review under responsibility of King Saud University.
==== Refs
References
Abdellattif M.H. Abdel-Rahman A.A.H. Arief M.M.H. Mouneir S.M. Ali A. Hussien M.A. Okasha R.M. Afifi T.H. Hagar M. Novel 2-Hydroselenonicotinonitriles and Selenopheno[2, 3-b]pyridines: efficient synthesis, molecular docking-DFT modeling, and antimicrobial assessment Front. Chem. 9 2021 672503 10.3389/fchem.2021.672503
Aidi Wannes W. Mhamdi B. Sriti J. Ben Jemia M. Ouchikh O. Hamdaoui G. Kchouk M.E. Marzouk B. Antioxidant activities of the essential oils and methanol extracts from myrtle (Myrtus communis var. italica L.) leaf, stem and flower Food Chem. Toxicol. 48 2010 1362 1370 10.1016/j.fct.2010.03.002 20211674
Aleksic V. Knezevic P. Antimicrobial and antioxidative activity of extracts and essential oils of Myrtus communis L Microbiol. Res. 169 4 2014 240 254 10.1016/j.micres.2013.10.003 24291016
Alipour G. Dashti S. Hosseinzadeh H. Review of pharmacological effects of Myrtus communis L. and its active constituents Phytother. Res. 28 8 2014 1125 1136 10.1002/ptr.5122 24497171
Banerjee, P., Eckert, A.O., Schrey, A.K., Preissner, R., 2018. ProTox-II: a webserver for the prediction of toxicity of chemicals. Nucleic Acids Res. 46 (W1), W257–W263. 10.1093/nar/gky318.
Barhouchi B. Aaoudi S. Abdi A. Essential oil chemical composition of myrtle growing in northeastern Algeria and estimation of its antibacterial effectiveness Am. J. Biochem. Biotechnol. 12 2 2016 110 121 10.3844/ajbbsp.2016.110.121
Ben Hsouna A. Ben Halima N. Smaoui S. Hamdi N. Citrus lemon essential oil: chemical composition, antioxidant and antimicrobial activities with its preservative effect against Listeria monocytogenes inoculated in minced beef meat Lipids Health Dis. 16 1 2017 146 10.1186/s12944-017-0487-5 28774297
Biovia D.S. Discovery studio visualizer 2017 San Diego CA, USA 936
Bolcato G. Bissaro M. Pavan M. Sturlese M. Moro S. Targeting the coronavirus SARS-CoV-2: computational insights into the mechanism of action of the protease inhibitors lopinavir, ritonavir and nelfinavir Sci. Rep. 10 1 2020 20927 10.1038/s41598-020-77700-z 33262359
Boussak H. Demim S. Hammadou S. Loucif Seiad L. Valuation and study of the activity of an essential oil extracted from a medicinal plant myrtle Mater. Today 49 4 2022 1041 1045 10.1016/j.matpr.2021.09.030
Bouzabata A. Cabral C. Gonçalves M.J. Cruz M.T. Bighelli A. Cavaleiro C. Casanova J. Tomi F. Salgueiro L. Myrtus communis L. as source of a bioactive and safe essential oil Food Chem. Toxicol. 75 2015 166 172 10.1016/j.fct.2014.11.009 25446467
Brada M. Tabti N. Boutoumi H. Wathelet J.P. Lognay G. Composition of the essential oil of leaves and berries of Algerian myrtle (Myrtus communis L.) J. Essent. Oil Res. 24 1 2012 1 3 10.1080/10412905.2012.645299
Cox S.D. Mann C.M. Markham J.L. Bell H.C. Gustafson J.E. Warmington J.R. Wyllie S.G. The mode of antimicrobial action of the essential oil of Melaleuca alternifolia (tea tree oil) J. Appl. Microbiol. 88 1 2000 170 175 10.1046/j.1365-2672.2000.00943.x 10735256
Daina A. Michielin O. Zoete V. SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules Sci. Rep. 7 3 2017 42717 10.1038/srep42717 28256516
Dallakyan S. Olson A.J. Small-molecule library screening by docking with PyRx Methods Mol. Biol. 1263 2015 243 250 10.1007/978-1-4939-2269-7_19 25618350
De Oliveira D.M.P. Forde B.M. Kidd T.J. Harris P.N.A. Schembri M.A. Beatson S.A. Paterson D.L. Walker M.J. Antimicrobial resistance in ESKAPE pathogens Clin. Microbiol. Rev. 33 3 2020 10.1128/CMR.00181-19
Dhifi W. Jazi S. El Beyrouthy M. Sadaka C. Mnif W. Assessing the potential and safety of Myrtus communis flower essential oils as efficient natural preservatives against Listeria monocytogenes growth in minced beef under refrigeration Food Sci. Nutr. 8 4 2020 2076 2087 10.1002/fsn3.1497 32328274
Djenane D. Yangüela J. Amrouche T. Boubrit S. Boussad N. Roncalés P. Chemical composition and antimicrobial effects of essential oils of Eucalyptus globulus, Myrtus communis and Satureja hortensis against Escherichia coli O157:H7 and Staphylococcus aureus in minced beef Food Sci. Technol. Int. 17 6 2011 505 515 10.1177/1082013211398803 22049156
Emon N.U. Alam M.M. Akter I. Akhter S. Sneha A.A. Irtiza M. Afroj M. Munni A. Chowdhury M.H. Hossain S. Virtual screenings of the bioactive constituents of tea, prickly chaff, catechu, lemon, black pepper, and synthetic compounds with the main protease (Mpro) and human angiotensin-converting enzyme 2 (ACE 2) of SARS-CoV-2 Futur. J. Pharm. Sci. 7 1 2021 121 10.1186/s43094-021-00275-7 34150913
Federspil P. Wulkow R. Zimmermann T. Effects of standardized Myrtol in therapy of acute sinusitis–results of a double-blind, randomized multicenter study compared with placebo Laryngorhinootologie 76 1 1997 23 27 10.1055/s-2007-997381 9156505
Gursoy U.K. Gursoy M. Gursoy O.V. Cakmakci L. Könönen E. Uitto V.-J. Anti-biofilm properties of Satureja hortensis L. essential oil against periodontal pathogens Anaerobe 15 4 2009 164 167 10.1016/j.anaerobe.2009.02.004 19285147
Hakmi M. Bouricha E.M. Kandoussi I. Harti J.E. Ibrahimi A. Repurposing of known anti-virals as potential inhibitors for SARS-CoV-2 main protease using molecular docking analysis Bioinformation 16 4 2020 301 306 10.6026/97320630016301 32773989
Hammond, J., Leister-Tebbe, H., Gardner, A., Abreu, P., Bao, W., Wisemandle, W., Baniecki, M., Hendrick, V.M., Damle, B., Simón-Campos, A., Pypstra, R., Rusnak, J.M., EPIC-HR Investigators, 2022. Oral Nirmatrelvir for High-Risk, Nonhospitalized Adults with Covid-19. N. Engl. J. Med. 386, 1397–1408. 10.1056/NEJMoa2118542
Hanwell M.D. Curtis D.E. Lonie D.C. Vandermeersch T. Zurek E. Hutchison G.R. Avogadro: an advanced semantic chemical editor, visualization, and analysis platform J. Cheminform. 4 1 2012 17 10.1186/1758-2946-4-17 22889332
Hennia A. Nemmiche S. Dandlen S. Miguel M.G. Myrtus communis essential oils: insecticidal, antioxidant and antimicrobial activities: a review J. Essent. Oil Res. 31 2019 487 545 10.1080/10412905.2019.1611672
Inouye S. Yamaguchi H. Takizawa T. Screening of the antibacterial effects of a variety of essential oils on respiratory tract pathogens, using a modified dilution assay method J. Infect. Chemother. 7 4 2001 251 254 10.1007/s101560170022 11810593
Jerkovic I. Radonic A. Borcic I. Comparative study of leaf, fruit and flower essential oils of Croatian Myrtus communis (L.) during a one-year vegetative cycle J. Essent. Oil Res. 14 4 2002 266 270 10.1080/10412905.2002.9699849
Jianu C. Stoin D. Cocan I. David I. Pop G. Lukinich-Gruia A.T. Mioc M. Mioc A. Șoica C. Muntean D. Rusu L.-C. Goleț I. Horhat D.I. In silico and In vitro evaluation of the antimicrobial and antioxidant potential of Mentha × smithiana R. GRAHAM essential Oil from western Romania Foods 10 4 2021 815 10.3390/foods10040815 33918674
Jin Z. Du X. Xu Y. Deng Y. Liu M. Zhao Y. Zhang B. Li X. Zhang L. Peng C. Duan Y. Yu J. Wang L. Yang K. Liu F. Jiang R. Yang X. You T. Liu X. Yang X. Bai F. Liu H. Liu X. Guddat L.W. Xu W. Xiao G. Qin C. Shi Z. Jiang H. Rao Z. Yang H. Structure of Mpro from SARS-CoV-2 and discovery of its inhibitors Nature 582 7811 2020 289 293 10.1038/s41586-020-2223-y 32272481
Kandsi F. Elbouzidi A. Lafdil F.Z. Meskali N. Azghar A. Addi M. Hano C. Maleb A. Gseyra N. Antibacterial and antioxidant activity of Dysphania ambrosioides (L.) mosyakin and clemants essential oils: experimental and computational approaches Antibiotics (Basel) 11 4 2022 482 10.3390/antibiotics11040482 35453233
Lengauer T. Rarey M. Computational methods for biomolecular docking Curr. Opin. Struct. Biol. 6 3 1996 402 406 10.1016/s0959-440x(96)80061-3 8804827
Lewis K. New approaches to antimicrobial discovery Biochem. Pharmacol. 15 134 2017 87 98 10.1016/j.bcp.2016.11.002
Maginnis M.S. Virus-receptor interactions: the key to cellular invasion J. Mol. Biol. 430 17 2018 2590 2611 10.1016/j.jmb.2018.06.024 29924965
Mann C.M. Cox S.D. Markham J.L. The outer membrane of Pseudomonas aeruginosa NCTC 6749 contributes to its tolerance to the essential oil of Melaleuca alternifolia (tea tree oil) Lett. Appl. Microbiol. 30 4 2000 294 297 10.1046/j.1472-765x.2000.00712.x 10792649
Maruzzella, J.C., Sicurella, N.A., 1960. Antibacterial activity of essential oil vapors**received February 2, 1960, from the department of biology, Long Island university, Brooklyn, N. y. J. Am. Pharm. Assoc. (Sci.) 49(11), 692–694. 10.1002/jps.3030491103.
Moradi M.T. Karimi A. Rafieian M. Kheiri S. Saedi M. The inhibitory effects of myrtle (Myrtus communis) extract on Herpes simplex virus-1 replication in Baby Hamster Kidney cells J. Shahrekord Univ. Med. Sci. 12 4 2011 54 61 http://78.39.35.44/article-1-441-en.html
Musarra-Pizzo M. Pennisi R. Ben-Amor I. Mandalari G. Sciortino M.T. Antiviral activity exerted by natural products against human viruses Viruses 13 5 2021 828 10.3390/v13050828 34064347
Nikakhtar Z. Hasanzadeh M. Hamedi S.S. Najafi M.N. Tavassoli A.P. Feyzabadi Z. Meshkat Z. Saki A. The efficacy of vaginal suppository based on myrtle in patients with cervicovaginal human papillomavirus infection: a randomized, double-blind, placebo trial Phytother. Res. 32 10 2018 2002 2008 10.1002/ptr.6131 29943384
Pereira N.L.F. Aquino P.E.A. Júnior J.G.A.S. Cristo J.S. Vieira Filho M.A. Moura F.F. Ferreira N.M.N. Silva M.K.N. Nascimento E.M. Correia F.M.A. Cunha F.A.B. Boligon A.A. Coutinho H.D.M. Ribeiro-Filho J. Matias E.F.F. Guedes M.I.F. Antibacterial activity and antibiotic modulating potential of the essential oil obtained from Eugenia jambolana in association with led lights J. Photochem. Photobiol. B 174 2017 144 149 10.1016/j.jphotobiol.2017.07.027 28778018
Pirbalouti, A.G., Mirbagheri, H., Hamedi, B., Rahimi, E., 2014. Antibacterial activity of the essential oils of myrtle leaves against Erysipelothrix rhusiopathiae. Asian Pac. J. Trop. Biomed. 4(1), S505–9. 10.12980/APJTB.4.2014B1168.
Prabuseenivasan S. Jayakumar M. Ignacimuthu S. In vitro antibacterial activity of some plant essential oils BMC Complement. Altern. Med. 6 2006 39 10.1186/1472-6882-6-39 17134518
Prasanth D.S.N.B.K. Murahari M. Chandramohan V. Panda S.P. Atmakuri L.R. Guntupalli C. In silico identification of potential inhibitors from Cinnamon against main protease and spike glycoprotein of SARS CoV-2 J. Biomol. Struct. Dyn. 39 13 2021 4618 4632 10.1080/07391102.2020.1779129 32567989
Rahman, M., Sarker, S.D., 2020. Antimicrobial natural products, in: Annual Reports in Medicinal Chemistry, Annual Reports in Medicinal Chemistry. Elsevier. 55, 77–113, 10.1016/bs.armc.2020.06.001.
Randrianarivelo R. Sarter S. Odoux E. Brat P. Lebrun M. Romestand B. Menut C. Andrianoelisoa H. Raherimandimby M. Danthu P. Composition and antimicrobial activity of essential oils of Cinnamosma fragrans Food Chem. 114 2 2009 680 684 10.1016/j.foodchem.2008.10.007
Reichling J. Schnitzler P. Suschke U. Saller R. Essential oils of aromatic plants with antibacterial, antifungal, antiviral, and cytotoxic properties–an overview Forsch. Komplementmed. 16 2 2009 79 90 10.1159/000207196 19420953
Santana O. Fe Andrés M. Sanz J. Errahmani N. Abdeslam L. González-Coloma A. Valorization of essential oils from Moroccan aromatic plants Nat. Prod. Commun. 9 8 2014 1109 1114 10.1177/1934578x1400900812 25233584
Shubina, L. P., Siurin, S. A., Savchenko, V. M., 1990. Inhalations of essential oils in the combined treatment of patients with chronic bronchitis. Vrach. Delo Part 5, 66–7. Russian. PMID: 2396397.
Sisakht M. Mahmoodzadeh A. Darabian M. Plant-derived chemicals as potential inhibitors of SARS-CoV-2 main protease (6LU7), a virtual screening study Phytother. Res. 35 6 2021 3262 3274 10.1002/ptr.7041 33759279
Stewart J.J. MOPAC: a semiempirical molecular orbital program J. Comput. Aided Mol. Des. 4 1 1990 1 105 10.1007/BF00128336 2197373
Stojkovic D. Sokovic M. Glamoclija J. Dzamic A. Ristic M. Fahal A. Khalid S. Djuic I. Petrovic S. Susceptibility of three clinical isolates of Actinomodura madurae to α-pinene, the bioactive agent of Pinus pinaster turpentine oil Arch. Biol. Sci. 60 4 2008 697 701 10.2298/abs0804697s
Trott O. Olson A.J. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading J. Comput. Chem. 31 2 2010 455 461 10.1002/jcc.21334 19499576
von Schindl R. Inhalative Wirkung aetherischer Oele Wien. Med. Wochenschr. 122 1972 591 593 5082866
Wang Y. Xiao J. Suzek T.O. Zhang J. Wang J. Bryant S.H. PubChem: a public information system for analyzing bioactivities of small molecules Nucleic Acids Res. 37 2009 W623 W633 10.1093/nar/gkp456 19498078
Wikler, M., Clinical., 2006. Approved Standard M100-S12. NCCLS, methods for dilution antimicrobial susceptibility tests of bacteria that grow aerobically. Clinical Laboratory Standards Institute CLSI
Xu J. Gao L. Liang H. Chen S.-D. In silico screening of potential anti-COVID-19 bioactive natural constituents from food sources by molecular docking Nutrition 82 2021 111049 10.1016/j.nut.2020.111049
Zanetti, S., Cannas, S., Molicotti, P., Bua, A., Cubeddu, M., Porcedda, S., Marongiu, B., Sechi, L.A., 2010. Evaluation of the antimicrobial properties of the essential oil of Myrtus communis L. against clinical strains of Mycobacterium spp. Interdiscip. Perspect. Infect. Dis. 2010. 10.1155/2010/931530.
|
PMC010xxxxxx/PMC10148244.txt |
==== Front
J Microbiol Immunol Infect
J Microbiol Immunol Infect
Journal of Microbiology, Immunology, and Infection
1684-1182
1995-9133
Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC.
S1684-1182(23)00090-7
10.1016/j.jmii.2023.04.012
Original Article
A people-centered decentralized outreach model toward HCV micro-elimination in hyperendemic areas: COMPACT study in SARS Co–V2 pandemic
Huang Ching-I abc
Liang Po-Cheng ab
Wei Yu-Ju a
Tsai Pei-Chien a
Hsu Po-Yao ab
Hsieh Ming-Yen a
Liu Ta-Wei a
Lin Yi-Hung a
Hsieh Meng-Hsuan ab
Jang Tyng-Yuan a
Wang Chih-Wen a
Yang Jeng-Fu a
Yeh Ming-Lun ac
Huang Chung-Feng ac
Dai Chia-Yen ac
Chuang Wan-Long ac
Huang Jee-Fu ac∗∗
Yu Ming-Lung acde∗
a Hepatobiliary Division, Department of Internal Medicine and Hepatitis Center, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan
b Graduate Institute of Clinical Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan
c School of Medicine and Hepatitis Research Center, College of Medicine, and Enter for Liquid Biopsy and Cohort Research, Kaohsiung Medical University, Kaohsiung, Taiwan
d School of Medicine, College of Medicine and Center of Excellence for Metabolic Associated Fatty Liver Disease, National Sun Yat-sen University, Kaohsiung, Taiwan
e Division of Hepato-Gastroenterology, Department of Internal Medicine, Kaohsiung Chang Gung Memorial Hospital, Kaohsiung, Taiwan
∗ Corresponding author. Hepatobiliary Division, Department of Internal Medicine and Hepatitis Center, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan. Tel.: +886 7 312-1101x7475; fax: +886 7 312-3955.
∗∗ Corresponding author. Hepatobiliary Division, Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung Medical University, 100 Tzyou Road, Kaohsiung City, 807, Taiwan. Tel.: +886 7 312-1101x7475; fax: +886 7 312-3955.
29 4 2023
6 2023
29 4 2023
56 3 586597
14 9 2022
13 3 2023
25 4 2023
© 2023 Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objectives
Gaps in linkage-to-care remain the barriers toward hepatitis C virus (HCV) elimination in the directly-acting-antivirals (DAA) era, especially during SARS Co–V2 pandemics. We established an outreach project to target HCV micro-elimination in HCV-hyperendemic villages.
Methods
The COMPACT provided “door-by-door” screening by an “outreach HCV-checkpoint team” and an “outreach HCV-care team” for HCV diagnosis, assessment and DAA therapy in Chidong/Chikan villages between 2019 and 2021. Participants from neighboring villages served as Control group.
Results
A total of 5731 adult residents participated in the project. Anti-HCV prevalence rate was 24.0% (886/3684) in Target Group and 9.5% (194/2047) in Control group (P < 0.001). The HCV-viremic rates among anti-HCV-positive subjects were 42.7% and 41.2%, respectively, in Target and Control groups. After COMPACT engagement, 80.4% (304/378) HCV-viremic subjects in the Target group were successfully linked-to-care, and Control group (70% (56/80), P = 0.039). The rates of link-to-treatment and SVR12 were comparable between Target (100% and 97.4%, respectively) and Control (100% and 96.4%) groups. The community effectiveness was 76.4% in the COMPACT campaign, significantly higher in Target group than in Control group (78.3% versus 67.5%, P = 0.039). The community effectiveness decreased significantly during SARS Co–V2 pandemic in Control group (from 81% to 31.8%, P < 0.001), but not in Target group (80.3% vs. 71.6%, P = 0.104).
Conclusions
The outreach door-by-door screen strategy with decentralized onsite treatment programs greatly improved HCV care cascade in HCV-hyperendemic areas, a model for HCV elimination in high-risk marginalized communities in SARS Co–V2 pandemic.
Keywords
Hepatitis C
Hepacivirus
HCV
Microelimination
DAA
Hyperendemic areas
COMPACT
SARS Co–V2
Pandemic
==== Body
pmcIntroduction
Hepatitis C virus (HCV) infection is one of the leading causes of liver-related clinical outcomes including liver cirrhosis, hepatocellular carcinoma (HCC) and mortality globally.1 Successful HCV clearance by antivirals greatly reduces the risk of HCC and mortality.2, 3, 4 The tremendous burden on public health is underestimated mainly because of low disease awareness, poor link-to-care and the lack of effective preventive measurements. Taiwan is not an exception with an estimated nationwide prevalence of 4%–9%, and the prevalence exceeds 20% in some hyperendemic areas.5, 6, 7 Although several measurements have been vigorously implemented over the past decades, several hurdles existed in the era of interferon-based therapies on the path to successful HCV treatment, from disease awareness, diagnosis, accessibility, treatment eligibility, to treatment efficacy, leading to huge gaps in each HCV care cascade.8
Fortunately, the high treatment efficacy, eligibility, safety, tolerability, and easy dosing of antiviral treatments for CHC are readily accessible for HCV care owing to the launch of directly acting antivirals (DAAs).9 , 10 The National Health Insurance of Taiwan started to reimburse DAAs in 2017, and the government additionally set a commitment to achieve the goal of HCV elimination by 2025, five years ahead of the World Health Organization's goal.11 Many efforts have been made in public prevention and therapeutic intervention for HCV to achieve this goal. Most previous efforts to eliminate HCV have focused on eliminating HCV infection in terms of comprehensive HCV screening and diagnosis, and successful linkage-to-care and linkage-to-treatment with affordable DAA regimens. The efforts are of importance and of pressing need for the hyperendemic areas.12 In addition, a recent mathematical model study showed that outreach screening and immediate onsite DAA therapy are the main effective measurements for achieving the overall goal of the WHO Global Health Sector Strategy on Hepatitis (2016–2021) by 2030.13 We demonstrated that an outreach program of mass screening with onsite treatment for uremic patients under maintenance hemodialysis helped achieve HCV micro-elimination in hemodialysis units.14 Nevertheless, the strategy including comprehensive diagnosis and linking patients to care has rarely been taken in the HCV hyperendemic regions.
Tzukuan is an HCV hyperendemic township located in the rural area of Kaohsiung City in southern Taiwan.15 , 16 It had an extremely high anti-HCV + prevalence (>40%) among adult residents with an annual incidence rate of 4.5%.15 , 16 We have reported that the age-, and sex-adjusted anti-HCV prevalence rate has been as high as 18.8%.17 Although the adjusted anti-HCV prevalence rates reduced over time, it remains a major health threat in Tzukaun township.17 Therefore, we established a prospective pilot project aiming to tackle the barriers in each HCV care cascade in Chidong and Chikan Villages, which possessed the highest anti-HCV prevalence rates (20.9% and 21.4%, respectively) while low disease awareness (26.2% and 14.6%, respectively) in Tzukuan.18 The people-centered project mainly included the strategies of outreach “door-by-door” screening followed by an outreach decentralized HCV care service. We anticipated that the implementation of outreach program (COMPACT) will be a feasible model for HCV micro-elimination in the HCV hyper-endemic areas. The aim of the COMPACT study was to address the scaling up of HCV care cascade from screening, accurate diagnosis, link-to care to treatment allocation by adopting the aforementioned strategies. Since SARS Co–V2 break out unexpectedly in March 2020, we also assessed the efficacy of the COMPACT model before and after the SARS Co–V2 pandemic.
Methods
Background
The target HCV hyperendemic areas were Chidong and Chikan Villages of Tzukuan, an HCV hyperendemic Township in the suburban area of Kaohsiung City in southern Taiwan.18 The COMPACT program is a prospective, interventional study which provided mass screening and link-to-care for HCV by an outreach check-up team and onsite DAA therapy by an outreach HCV treater team January 2019 to December 2021. With the collaboration of a non-profit organization, Taiwan Liver Research Foundation, and a tertiary referral center, Kaohsiung Medical University Hospital, the program was engaged with stakeholders, local opinion leaders and providers of both healthcare and administration in Chidong/Chikan Villages and Tzukuan Township. This study aims to compare the HCV treatment uptake and community effectiveness between Target Group and the Control Group.
Study population
All subjects aged ≥20 years and who had the willingness to participate in the program were recruited. People who live in or work in Chidong/Chikan villages were enrolled as the “Target group”, while participants who live in or work in the other villages without active intervention were served as the “Control group” (Supplementary Fig. 1). The residents from the other villages of Tzukuan District will be collected according to their willingness to join the screening. The study was conducted following the Declaration of Helsinki. The ethics committee of the Kaohsiung Medical University Hospital approved the study. All participants gave written informed consent before enrollment.
Strategies with door-by-door interventions taken in the target group
a. Task-force action plan for the announcement of COMPACT program to increase the disease awareness
Door-by-door screening information was delivered by Flier in collaboration local key opinion and administrative leaders in Chikan/Chidong Villages. The details and structured strategies of the project was released and posted on media and press. The educational programs for local healthcare providers and general population of the Township were provided to increase the rates of disease awareness and link-to-treater.b. HCV testing/diagnosis by COMPACT outreach screening program
We first set up an outreach HCV check-up team including one physician consultant, two nursing coordinators and two administrative stuffs. There are two rounds of the screening program.
· The first-round screening
The first-round comprehensive screening program consisted of two strategies. Firstly, one shop strategy: A screening team fixed in a community HCV service station at the local activity center located in the central area of Target Villages to provide a feasible screening service, six days a week, for the participants. The screening information was delivered by mail, flier, media and phone call, followed by intensified second calls for previous non-responders in the first round. Secondly, a door-by-door strategy: a mobile screening team provided a home-based screening service for the residents in Target Villages. The schedule was in collaboration with Heads of Neighborhoods. The screening was performed every two neighborhoods weekly in the first half-year.
The first-round screening program included interviews of questionnaires, blood sampling for testing viral hepatitis and liver function and linking the HCV-viremic patients to care. The blood testing included complete cell count, hepatitis B surface antigen (HBsAg), anti-HCV, transaminases, and alpha-fetoprotein for all participants. HCV RNA was tested for anti-HCV-seropositive subjects, and HCV genotyping was tested for HCV RNA-positive subjects with the strategy of HCV reflex testing.19 Subjects who were positive for anti-HCV or HBsAg, or had abnormal liver function were appointed for transient elastography (Fibroscan, Echosens, Paris, France) performed by qualified and experienced operators following standard procedure. HCV-viremic subjects received linkage-to-care and linkage-to-treat for anti-HCV therapy scheduled by well-trained nursing specialists during the SARS Co–V2 pandemic. The residents from the other villages of Tzukuan District will be collected according to their willingness to join the screening as the Control group. They were either appointed to the COMPACT outreach HCV-treater team or referred to the other HCV treaters based on patient discretion.
· The second-round screening
Subjects seropositive for anti-HCV in the first-round screening were recalled one year later for second round screening, including questionnaire interview, liver function tests, HCV RNA, AFP and the outcome of link-to-care.c. Link-to-treater, either COMPACT outreach HCV-treater team or the other local treaters
We set up an outreach HCV-eliminating treatment team including hepatologists, nursing coordinators, pharmacists, laboratory technicians and administrative coordinators. The COMPACT outreach HCV-treater team operated in a community health-care station located in the community center of Target area. This team provided “point-of-care” service of pretreatment assessment, anti-HCV treatment and post-treatment follow-up, including liver function and renal function tests, onsite abdominal sonography, evaluation of potential drug–drug interaction. The treatment regimens and strategies conformed to the regulations of the Health and Welfare Department of Taiwan and regional guidelines.10 , 14
Strategies and interventions taken in the control group
The strategies and intervention taken in the Control group were the same as those in the Target group, except that there was no “door-by-door” strategies for deliver the information and screening for the Control group. The information for the Control group only came from flier and media publicity.
Endpoint measurements
The primary end-point was the proportion of successful link-to-care among all HCV-viremia subjects identified. The secondary endpoints were:1. The proportion of successful link-to-treat among all successful link-to-care subjects identified; 2. The efficacy of DAA therapy, the sustained virological response (SVR), defined as undetectable HCV RNA throughout 12 weeks post treatment.
Laboratory analysis
Biochemical analyses were evaluated on a multichannel autoanalyzer (Hitachi Inc, Tokyo, Japan). HBsAg was tested using commercially available enzyme-linked immunosorbent assay kits (Abbott Laboratories, North Chicago, IL, USA). Anti-HCV tests were performed using a third-generation commercially ELISA kit (AxSYM 3.0; Abbott Laboratories, North Chicago, IL, USA). HCV RNA and genotype were detected using real-time PCR (Real-time HCV; Abbott Molecular, Des Plaines IL, USA; detection limit: 12 IU/ml).
Disease severity assessment
Fibrosis-4 index (FIB-4) was calculated by (age × AST)/(platelets (109/L) × ALT1/2).20 Liver cirrhosis was defined as any of the following: transient elastography (FibroScan®; Echosens, Paris, France) > 12 kPa, fibrosis-4 index (FIB-4) > 6.5, or the presence of clinical, radiological, endoscopic, or laboratory evidence of cirrhosis and portal hypertension.21, 22, 23
Statistical analysis
Frequency was compared between groups using the χ2 test with the Yates correction or Fisher's exact test. Group means (presented as the mean standard deviation) were compared using analysis of variance and Student's t-test or the nonparametric Mann–Whitney U test when appropriate. Descriptive and comparative statistics were performed for all demographic and clinical variables. The need-to-treat population was defined as the number of HCV viremic subjects plus the number of anti-HCV-seropositive and HCV RNA-negative subjects with prior history of antiviral therapy. The efficacy of DAA therapy was evaluated in an intent-to-treat (ITT) population (all subjects who received ≥1 dose of DAA), and a modified ITT (mITT) population (subjects receiving ≥1 dose of DAA with HCV RNA data available at posttreatment week 12). The community effectiveness in COMPACT was defined as proportion of HCV cure among HCV-viremic subjects. The overall treatment update was defined as proportion of HCV cure among need-to-treat subjects (anti-HCV seropositive subjects excluding those with spontaneous HCV clearance at enrollment). The statistical analyses were performed using the SPSS 12.0 statistical package (SPSS, Chicago, IL, USA). All statistical analyses were based on two-sided hypothesis tests with a significance level of p < 0.05.
Results
Baseline characteristics
By December 2020, a total of 3684 subjects were enrolled in the Target Group, and 2047 subjects were recruited in the Control group (Table 1 ). The mean age in the target group was 53.1 of years and male accounted for 57.4% of the population. Compared to the control group, patients in the Target group were older (53.1 years vs. 50.6 years, P < 0.001), had higher body mass index (25.2 kg/m2 vs. 24.4 kg/m2, P < 0.001) and FIB-4 (1.3 vs. 1.1, P < 0.001), higher levels of fasting sugar (92.9 mg/dL vs. 89.9 mg/dL, P = 0.02), hemoglobin A1C (5.9% vs. 5.8%, P < 0.001), hemoglobin (13.9 g/dL vs. 13.6 g/dL, P < 0.001), liver biochemistries including r-glutamyltransferase (rGT, 23.0 IU/L vs. 22.0 IU/L, P < 0.001), aspartate aminotransferase (AST, 27.7 IU/L vs. 25.1 IU/L, P < 0.001) and alanine aminotransferase (ALT, 29.6 IU/L vs. 26.1 IU/L, P < 0.001), lower platelet counts (256.2 × 103 u/L vs. 262.8 × 103 u/L, P = 0.001) and a higher proportion of hypertension (26.7% vs. 20.7%, P < 0.001). The seropositivity rates of HBsAg (10.3% vs. 8.4%, P = 0.019) and anti-HCV (24.0% vs. 9.5%, P < 0.001) were significantly higher in the Target group compared to the control group.Table 1 Characteristics and clinical features of the whole subjects among the HCV hyperendemic villages in COMPACT study.
Table 1 Target Group (N = 3684) Control Group (N = 2047) P value
Age (years, mean ± SD) 53.1 ± 15.9 50.6 ± 15.1 <0.001
Male gender, n (%) 2115 (57.4) 1220 (59.6) 0.107
DM, n (%) 440 (11.9) 214 (10.5) 0.089
Hypertension, n (%) 983 (26.7) 424 (20.7) <0.001
BMI, kg/m2 (mean ± SD) 25.2 ± 4.3 24.4 ± 4.2 <0.001
Uric acid, mg/dL (mean ± SD) 5.6 ± 1.5 5.6 ± 1.6 0.063
Triglyceride, mg/dL (mean ± SD) 123.7 ± 141.0 117.2 ± 79.6 0.172
Total Cholesterol, mg/dL (mean ± SD) 194.8 ± 38.5 195.4 ± 38.7 0.689
Fasting sugar, mg/dL (mean ± SD) 92.9 ± 34.3 89.9 ± 28.8 0.015
HbA1c, % (mean ± SD) 5.9 ± 1.1 5.8 ± 0.9 <0.001
White blood cell count, x103 u/L (mean ± SD) 6.4 ± 1.9 6.3 ± 1.8 0.281
Hemoglobin, g/dL (mean ± SD) 13.9 ± 1.7 13.6 ± 1.8 <0.001
Platelet count, x103 u/L (mean ± SD) 256.2 ± 71.9 262.8 ± 70.7 0.001
r-GT, IU/L (mean ± SD) 39.5 ± 75.1 32.4 ± 52.4 <0.001
AST, IU/L (mean ± SD) 27.7 ± 22.2 25.1 ± 16.1 <0.001
AST >40 IU/L, n (%) 344 (9.3) 135 (6.6) <0.001
ALT, IU/L (mean ± SD) 29.6 ± 27.6 26.1 ± 23.7 <0.001
ALT >40 IU/L, n (%) 623 (16.9) 274 (13.4) <0.001
α-fetoprotein, ng/ml, (mean ± SD) 6.4 ± 146.0 3.2 ± 14.3 0.317
α-fetoprotein >20 ng/ml, n (%) 31 (0.8) 10 (0.5) 0.129
HBsAg (+), n (%) 378 (10.3) 171 (8.4) 0.019
Anti-HCV (+), n (%) 886 (24.0) 194 (9.5) <0.001
HCC at the time of enrollment, n (%) 6 (0.2) 0 (0) <0.001
Note: DM: diabetes mellitus. BMI: Body mass index. HbA1c: Glycohemoglobin A1C. r-GT: gamma glutamyl transferase. AST: aspartate aminotransferase. ALT: alanine aminotransferase. HBsAg: hepatitis B surface antigen. Anti-HCV: hepatitis C antibody. HCC: hepatocellular carcinoma.
Six patients were screened to have HCC (HCV-related, n = 4; HBV/HCV-coinfected, n = 1; HBV-related, n = 1) in the Target group and none except one had disease awareness for their liver diseases. The patient with disease awareness had HBV/HCV co-infection, but only had awareness of his HBV infection. All 6 patients were referred to receive HCC assessment and treatment consequently. Five HCC patients were Barcelona clinic liver cancer (BCLC) stage A; one was BCLC stage B. Four received curative therapy for HCC (three surgical resection; one radiofrequency ablation); two received trans-arterial chemoembolization (Supplemental Table 1). All five patients with HCV viremia were successfully linked-to-care and achieved SVR12 with DAA therapy. The two patients with HBV viremia received entecavir therapy.
Characteristics of the subjects with anti-HCV seropositivity between target and control groups
Of the 886 anti-HCV seropositive subjects, the mean age in the Target group was 63.8 years and male accounted for 57.1%, which was similar to those of 194 anti-HCV seropositive subjects in the Control group (Table 2 ). The HCV viremic rate was similar between the Target group (42.7%, 378/886) and the Control group (41.2%, 80/192). All the other biochemistries and demography did not differ between the two groups except a significantly higher proportion of subjects with liver cirrhosis was observed in Target group (10.8%) than in Control group (4.6%) (P = 0.007).Table 2 Characteristics and clinical features of anti-HCV-seropositive subjects among HCV hyperendemic villages in COMPACT study.
Table 2 Target Group (N = 886) Control Group (N = 194) P value
Age (years, mean ± SD) 63.8 ± 11.8 62.3 ± 11.4 0.104
Male gender, n (%) 506 (57.1) 106 (54.6) 0.529
DM, n (%) 149 (16.8) 38 (19.6) 0.356
Hypertension, n (%) 358 (40.4) 76 (39.2) 0.751
BMI, kg/m2 (mean ± SD) 25.4 ± 4.0 24.8 ± 4.0 0.082
Platelet count, x103 u/L (mean ± SD) 220.5 ± 66.4 222.8 ± 65.0 0.663
r-GT, U/L (mean ± SD) 48.2 ± 106.7 39.6 ± 42.2 0.272
AST, IU/L (mean ± SD) 36.5 ± 32.5 35.0 ± 32.6 0.554
AST >40 IU/L, (%) 187 (21.1) 36 (18.6) 0.427
ALT, IU/L (mean ± SD) 38.0 ± 40.3 38.2 ± 47.7 0.946
ALT >40 IU/L, n (%) 237 (26.7) 44 (22.7) 0.242
α-fetoprotein, ng/ml, (mean ± SD) 16.6 ± 295.6 6.4 ± 41.2 0.632
α-fetoprotein >20 ng/ml, n (%) 25 (2.8) 5 (2.6) 0.851
HBsAg (+), n (%) 74 (8.4) 13 (6.7) 0.444
HCV RNA (+), n (%) 378 (42.7) 80 (41.2) 0.716
Liver cirrhosis, n (%) 96 (10.8) 9 (4.6) 0.007
HCC at the time of enrollment, n (%) 5 (0.6) 0 (0) <0.001
Note: DM: diabetes mellitus. BMI: Body mass index. r-GT: gamma glutamyl transferase. AST aspartate aminotransferase. ALT: alanine aminotransferase. HBsAg: hepatitis B surface antigen. HCV: hepatitis C virus. FIB-4: Fibrosis-4 index. HCC: hepatocellular carcinoma.
Characteristics of the subjects with HCV viremia between target and control group
Of the 378 HCV-viremic subjects in the Target group, the mean age was 62.6 years and males accounted for 59.0% (Table 3 ). The most common viral genotype was HCV-1b (54.0%), followed by HCV-2 (36.8%). The characteristics of HCV-viremic subjects were similar between the Target and Control groups except for a significantly higher BMI (25.3 kg/m2 vs. 24.2 kg/m2, P = 0.019) and a significantly lower HCV RNA levels in the Target group (5.5 logs IU/mL vs. 5.9 logs IU/mL, P = 0.025) and a significantly higher proportion of subjects with liver cirrhosis in Target group (18.3%) than in Control group (8.8%) (P = 0.046). Notably, 211 of 458 (54.9%) HCV viremic subjects had normal ALT levels. All of the 5 anti-HCV positive subjects with HCC found at enrollment were HCV-viremic in the Target group.Table 3 Characteristics and clinical features of HCV RNA-seropositive subjects among HCV hyperendemic villages in COMPACT study.
Table 3 Target Group (N = 378) Control Group (N = 80) P value
Age (years, mean ± SD) 62.6 ± 12.3 60.0 ± 10.8 0.073
Male gender, n (%) 223 (59.0) 44 (55.0) 0.51
DM, n (%) 60 (15.9) 15 (18.8) 0.528
Hypertension, n (%) 159 (42.1) 26 (32.5) 0.113
BMI, kg/m2 (mean ± SD) 25.3 ± 3.8 24.2 ± 4.1 0.019
Platelet count, x103 u/L (mean ± SD) 211.7 ± 69.0 222.0 ± 67.0 0.224
r-GT, U/L (mean ± SD) 58.2 ± 105.8 52.9 ± 56.7 0.662
AST, IU/L (mean ± SD) 46.8 ± 35.9 51.5 ± 45.5 0.313
AST >40 IU/L, n (%) 140 (37.0) 35 (43.8) 0.262
ALT, IU/L (mean ± SD) 52.8 ± 46.6 61.7 ± 66.5 0.260
ALT >40 IU/L, n (%) 173 (45.8) 38 (47.5) 0.778
α-fetoprotein, ng/ml, (mean ± SD) 31.1 ± 444.9 12.4 ± 63.9 0.708
α-fetoprotein >20 ng/ml, n (%) 22 (5.8) 5 (6.3) 0.882
HBsAg (+), n (%) 25 (6.6) 4 (5.0) 0.59
HCV viral loads, log IU/mL (mean ± SD) 5.5 ± 1.2 5.9 ± 1.1 0.025
HCV genotype, 1a/1b/(1a+2)/(1 b + 2)/2/3/4/5/6/unclassified/no detected, n (%) 12/204/2/6/139/0/1/0/8/3/3 (3.2/54.0/0.5/1.6/36.8/0/0.3/0/2.1/0.8/0.8) 7/37/1/0/32/0/0/0/3/0/0 (8.8/46.3/1.3/0/40.0/0/0/0/3.8/0/0)
Liver cirrhosis, n (%) 69 (18.3) 7 (8.8) 0.046
HCC at the time of enrollment (%) 5 (1.3) 0 (0) <0.001
DM: Note: Diabetes mellitus. BMI: Body mass index. r-GT: gamma glutamyl transferase. AST: aspartate aminotransferase. ALT: alanine aminotransferase. HBsAg: hepatitis B surface antigen. FIB-4: Fibrosis-4 index. HCC: hepatocellular carcinoma.
Initial HCV care cascade at the time of enrollment for COMPACT campaign
At initiation of COMPACT engagement, the disease awareness among anti-HCV seropositive subjects was significantly lower in the Target group than in the Control group (53.5% [474/886] vs. 61.9% [120/194], P = 0.034, Table 4 ). Of the 474 HCV aware subjects in the Target group, 389 (82.1%) subjects had ever got access to medical facilities, which was higher than that of Control group (74.2% [89/120], P = 0.051).Table 4 Initial HCV care cascade at the time of enrollment.
Table 4HCV care cascade at enrollment Total Target Group Control Group P value
All population
N = 1080 N = 886 N = 194
Anti-HCV awareness in anti-HCV (+) subjects, n (%) 594 (60.0) 474 (53.5) 120 (61.9) 0.034
- HCV assessed in aware subjects, n/N (%) 478/594 (80.5) 389/474 (82.1) 89/120 (74.2) 0.051
-- HCV being treated in assessed subjects who needed antiviral therapy, n/Na (%) 267/401 (66.6) 215/325 (66.2) 52/76 (68.4) 0.787
-- prior anti-HCV regimen, IFN-RBV/DAA/DAA after IFN-RBV. n (%) 155/82/30 (58.1/30.7/11.2) 132/64/19 (61.4/29.8/8.8) 23/18/11 (44.2/34.6/21.2)
- HCV cured at the time of enrollment among treated subjects, n/N (%) 225/267 (84.3) 179/215 (83.3) 46/52 (88.5) 0.355
Overall treatment uptake among need-to-treat population, n/Nb (%) 225/683 (32.9) 179/557 (32.1) 46/126 (36.5) 0.347
Subjects with liver cirrhosis
N = 105 N = 96 N = 9
Anti-HCV awareness in anti-HCV (+) subjects, n (%) 58 (55.2) 53 (55.2) 5 (55.6) 1.00
-HCV assessed in aware subjects, n/N (%) 44/58 (75.9) 40/53 (75.5) 4/5 (80.0) 1.00
- HCV being treated in assessed subjects who needed antiviral therapy, n/Na (%) 11/39 (28.2) 10/35 (28.6) 1/4 (25.0) 1.00
-- prior anti-HCV regimen, IFN-RBV/DAA/DAA after IFN-RBV. n (%) 9/1/1 (81.8/9.1/9.1) 9/1/0 (90.0/10.0/0) 0/0/1 (0/0/100.0)
-HCV cured at the time of enrollment among treated subjects, n/N (%) 9/11 (81.8) 8/10 (80.0) 1/1 (100.0) 1.00
Overall treatment uptake among need-to-treat population, n/Nb (%) 9/85 (10.6) 8/77 (10.4) 1/8 (12.5) 1.00
Note: HCV: hepatitis C virus. IFN: interferon-based therapy. RBV: ribavirin. DAA: directly acting antivirals.
a HCV assessed subjects excluding those with spontaneous HCV clearance.
b Anti-HCV seropositive subjects excluding those with spontaneous HCV clearance.
Among assessed subjects in the target group, 215 (66.2%) of 325 need-to-treat subjects had received antivirals (Interferon [IFN]-based regimen: 61.4%, DAAs: 29.8%, IFN followed by DAA: 8.8%) with an SVR rate of 83.3% (n = 179). Among assessed subjects in the control group, 52 (68.4%) of 76 need-to-treat subjects had received antivirals (IFN-based regimen: 44.2%, DAAs: 34.6%, IFN followed by DAA: 21.2%) with an SVR rate of 88.5% (n = 46). Overall, the treatment uptake of need-to-treat subjects was similarly low between the target group (179/557, 32.1%) and the control group (46/126, 36.5%, P = 0.347, Table 4).
Of the 105 cirrhotic patients at initiation of COMPACT engagement, only 58 were aware of HCV infection. Among aware subjects, 44 (75.9%) had access to medical facilities. However, only 11 (28.5%) of 39 need-to-treat cirrhotic patients received antivirals and 9 (81.8%) attained an SVR. To this end, only 9 of the 85 (10.6%) need-to-treat cirrhotic patients achieved an SVR. Similar rates of disease awareness, link-to-care and treatment were observed between cirrhotic Target and Control groups, with an overall treatment uptake among need-to-treat cirrhotic subjects of 10.4% and 12.5% respectively (Table 4).
HCV care cascade achieved by COMPACT program
We further addressed the HCV care cascade of HCV-viremic patients achieved by COMPACT study engagement (Table 5 ). Of the 378 HCV-viremic patients in the Target group, 304 (80.4%) were successfully linked-to-care (272 subjects link-to-onsite care; 32 subjects link-to-other sites care), which was significantly higher than that in the control group (70% [56/80], P = 0.039) (51 subjects link-to-onsite care; 5 subjects link-to-other sites care). All of the patients linked-to-care received DAA in both groups. Of the 304 patients who received DAA treatment in the Target group (onsite care: n = 272, other site treaters: n = 32), 3 patients terminated treatment early (one death due to accident; two refused to return) without SVR12 data available, 3 patients lost to follow-up after completing DAA therapy and 2 patients experienced virological relapse (Supplementary Fig. 2A). The overall SVR12 rate by ITT and mITT was 97.4% (296/304) and 99.3% (296/298), respectively (Table 5). Of the 56 HCV-viremic patients who received DAA treatment in the Control group (onsite care: n = 51, other sites treater: n = 5), 2 patients terminated treatment early (one death due to subarachnoid hemorrhage unrelated to treatment; one refusal to return) without SVR12 data available (Supplementary Fig. 2B). The overall SVR12 rate by ITT and mITT was 96.4% (54/56) and 100% (54/54), respectively. Taken together, the overall SVR12 rate by ITT and mITT was 97.2% (350/360) and 99.4% (350/352), respectively in the COMPACT study (Table 5). The baseline characteristics and treatment outcome of DAA therapy for HCV-viremic patients treated onsite were listed in Supplemental Table 2. No patients experienced serious adverse events. The characteristics of the 10 subjects who discontinued treatment, lost-to-follow or failed to achieve an SVR were listed in Supplemental Table 3.Table 5 HCV care cascade of the viremic patients after COMPACT study engagement.
Table 5HCV cascade after enrollment All Subjects (N = 458) Target Group (N = 378) Control Group (N = 80) P value
A. Successful link-to-care of HCV viremia subjects, n/N (%) 360/458 (78.6) 304/378 (80.4) 56/80 (70.0) 0.039
- Successful link-to-onsite care, n/N (%) 323/458 (70.5) 272/378 (72.0) 51/80 (63.8) 0.14
- Successful link-to-other sites care, n/N (%) 37/458 (8.1) 32/378 (8.5) 5/80 (6.3) 0.51
B. Successful link-to-treat, n/N (%) 360/360 (100) 304/304 (100) 56/56 (100)
C. Successful antiviral therapy
C1. SVR12 rate, ITT, n/N (%) 350/360 (97.2) 296/304 (97.4) 54/56 (96.4) 0.66
-SVR12 rate, onsite treatment, ITT, n/N (%) 313/323 (96.9) 264/272 (97.1) 49/51 (96.1) 0.66
-SVR12 rate, other sites treatment, ITT, n/N (%) 37/37 (100.0) 32/32 (100.0) 5/5 (100.0) –
C2. SVR12 rate, mITT, n/N (%) 350/352 (99.4) 296/298 (99.3) 54/54 (100) 1
-SVR12 rate, onsite treatment, mITT, n/N (%) 313/315 (99.4) 264/266 (99.2) 49/49 (100.0) 1
-SVR12 rate, other sites treatment, mITT, n/N (%) 37/37 (100.0) 32/32 (100.0) 5/5 (100.0) –
Community effectiveness in COMPACTa, n/N (%) 350/458 (76.4) 296/378 (78.3) 54/80 (67.5) 0.039
Overall treatment uptakeb, n/N (%) 575/683 (84.1) 475/557 (85.3) 100/126 (79.4)
Note: HCV: hepatitis C virus. DAA: directly acting antivirals. SVR12: sustained virological response, defined as undetectable HCV RNA throughout 12 weeks of the post-treatment follow-up period. ITT: intention-to-treat analysis, defined as subjects who had received ≥1 dose of DAA. mITT: modified intention-to-treat analysis, defined as subjects receiving ≥1 dose of DAA and HCV RNA data available at post-treatment week 12.
a Community effectiveness in COMPACT, proportion of HCV cure among HCV-viremic subjects (= A x B x C1).
b Overall treatment update, proportion of HCV cure among need-to-treat subjects.
Overall, 350 of 458 HCV-viremic subjects were cured by DAA in COMPACT campaign, with community effectiveness of 76.4%. The community effectiveness was significantly higher in Target group than in Control group (78.3%, 296/378 versus 67.5%, 54/80, P = 0.039). The HCV care cascade of Target group after COMPACT campaign was shown in Fig. 1 . Taken together, the overall treatment uptake before and after engagement of COMPACT among the need-to-treat population increased from 32.9% to 84.1%; 32.1%–85.1% for Target group and 36.5%–79.4% for Control group.Figure 1 HCV care cascade achieved by the COMPACT campaign in (A) Total population and (B) Target and Control groups. EOT: end-of-treatment. SVR: sustained virological response, defined as HCV RNA seronegativity 12 weeks after the end of treatment.
Fig. 1
Impact of SARS Co–V2 pandemic on the HCV care cascade between target and control groups
To evaluate the impact of the SARS CO–V2 pandemic on the HCV care cascade, we further compared the community effectiveness of HCV elimination between subjects enrolled before (January 2019 to February 2020) and those during (March 2020 to December 2020) the SARS Co–V2 pandemic. A total of 3969 and 1762 subjects were screened in the pre-SARS Co–V2 pandemic period and during the SARS Co–V2 pandemic, respectively, with 348 and 110 subjects seropositive for HCV RNA, respectively (Supplementary Table 4). The rate of successful link-to-care was significantly higher in the pre-SARS Co–V2 pandemic period than during the SARS Co–V2 pandemic period (83.4% vs. 64.5%, P < 0.001). The rates of link-to-treat and SVR12 were comparable between the two groups (100% and 96.9%, respectively, vs. 100% and 95.6% respectively). The community effectiveness was significantly higher in the pre-SARS Co–V2 pandemic period than during the SARS Co–V2 pandemic period (80.5% vs. 63.6%, P < 0.001, Supplementary Table 4).
Among the Target group, 2655 and 1029 subjects were screened in the pre-SARS Co–V2 pandemic period and during the SARS Co–V2 pandemic, respectively, with 290 and 88 HCV-viremic subjects, respectively (Supplementary Table 4). The rate of successful link-to-care was significantly higher in the pre-SARS Co–V2 pandemic period than during the SARS Co–V2 pandemic period (82.8% vs. 72.7%, P = 0.046). The rates of link-to-treat and SVR12 were comparable between the two groups (100% and 97.1%, respectively, vs. 100% and 98.4% respectively). Nevertheless, the community effectiveness did not differ between the pre- and during the SARS Co–V2 pandemic periods (80.3% vs. 71.6%, P = 0.104, Fig. 2 A).Figure 2 HCV care cascades before and during between SARS Co–V2 pandemic in (A) Target group and (B) Control group.
Fig. 2
Among the Control group, 1314 and 733 subjects were screened in the pre-SARS Co–V2 pandemic period and during the SARS Co–V2 pandemic, respectively, with 58 and 22 HCV-viremic subjects, respectively (Supplementary Table 4). The rate of successful link-to-care in the pre-SARS Co–V2 pandemic period was 84.5%, which significantly decreased to only 31.8% during the SARS Co–V2 pandemic period (P < 0.001). Although the rates of link-to-treat and SVR12 were comparable between the two groups (100% and 95.9%, respectively, vs. 100% and 100% respectively), the community effectiveness decreased significantly from 81% in the pre-SARS Co–V2 pandemic period to only 31.8% during the SARS Co–V2 pandemic period (P < 0.001, Fig. 2B).
Discussion
In the current study, we demonstrated that implementation of a people-centered decentralized outreach program with strategies of “door-by-door” screening followed by an outreach HCV onsite therapy could achieve 80.4% of successful link-to-care, 100% of treatment rate for linked subjects and 97.2% of cured rate for treated patients, with community effectiveness of 78.3% for HCV elimination in a marginalized HCV hyperendemic area, even during COVID-19 pandemic era.
HCV has prevailed in particular in “HCV hyper-endemic townships” such as Tzukuan in Taiwan. As a whole, the seroprevalence of HCV infection has decreased in the past two decades.24 Similarly, with the continuous efforts implemented by the government and non-government organizations, both the prevalence and incidence of HCV decreased from 21.1% in 2000–2004 to 10.3% in 2015–2019 in Tzukuan.17 However, the current study revealed that the seroprevalence of anti-HCV remained high in the Target group, 24.0%, indicating there remained hot spots in the HCV hyperendemic area. Since the major risk factor of HCV infection changed from iatrogenic procedures before 2009 to household contact after 2010,17 “treatment as prevention” has become the key strategy for HCV control in this area.
Despite continuous education in the Tzukuan area during the past decades,17 only 55% of the anti-HCV seropositive subjects were aware of their HCV infection at the time of enrollment in the current study; the awareness rate was even lower in the Target group than in the Control group. The traditional concept and policy for screening viral hepatitis are only for subjects with abnormal liver function. Compared to a previous hospital-based study in which around 25% of HCV patients had normal ALT levels,25 more than half of the HCV-viremic subjects had normal ALT levels in the current community-based study. Therefore, the current study used several media strategies and collaborated with local opinion leaders to advocate the importance of universal HCV screening regardless of liver function.
Although the current DAAs provide excellent efficacy and safety profiles, there remains plenty of unmet needs in eliminating HCV.26 While the difficult-to-cure population is no longer a critical issue in HCV care, the major hurdle of HCV care exists in properly and accurately identifying the HCV-infected subjects and linking the viremic patients to medical care. To scale up the HCV care cascade, an outreach screen strategy plays a determinant role.13 Beyond the issue of a novel point of care diagnosis,27 we first adopted the door-to-door anti-HCV screening followed by HCV RNA reflex testing for those seropositive for ant-HCV.19 The screen model turned out to have an excellent coverage rate which could serve as an exemplar in HCV care.
Another hurdle for HCV care was the accessibility for patients being diagnosed.8 The obstacles may arise from the economic barrier and the lack of easy access to the treaters. Before the COMPACT program, around 80% of the aware subjects had HCV assessed before participating in the COMPACT campaign. However, only 66.6% of need-to-treat assessed HCV subjects received antiviral therapy even under the reimbursement of DAA by the National Health Insurance Administration of Taiwan initiated in 2017. The data implicated the importance of providing accessible medical facilities with highly acceptable antiviral regimens to overcome the gap between accessibility and treatment uptake. By adopting the outreach onsite DAA treatment strategy, we provided an excellent treatment rate of 100% for the assessed HCV-viremic patients. The significantly higher rate of link-to-care in the Target group than in the Control group also implied the importance of convenience for accessibility.
As the national insurance fully supports DAAs in Taiwan, the feasibility of medical care would be important, particularly in rural areas. Recently, we successfully set up an outreach onsite treatment model for the uremic population.14 In the current project, most of the patients linked-to-care were treated by the COMPACT onsite treatment team rather than other clinics. The result successfully replicated the successful experience in the uremic population of the ERASE-C campaign.14
All patients linked to medical care were successfully allocated to DAA regimens. Eventually, the community effectiveness of the COMPACT campaign had markedly increased from 32.9% to 76.4% and the treatment uptake of the need-to-treat population from 32.9% to 84.1%. Notably, the community effectiveness was significantly higher in the Target group than in the Control group. Our results proved that a decentralized outreach screening strategy plus onsite treatment service could serve as a model for HCV micro-elimination for marginalized communities with high risk and disease burden of HCV.
Poor awareness of disease severity and lack of apparent symptoms among patients with severe liver diseases are the other challenges urgent for HCV elimination. The current study discovered that one-fifth of the HCV-viremic patients in the Target group had liver cirrhosis at the time of enrollment. In addition, six subjects were identified incidentally with existing HCC, including five HCV-related and one HBV-related in the Target group. All of them were unaware of their liver diseases. The findings reinforce the urgent need for advocating awareness and comprehensive screening of viral hepatitis in the high-risk population. Fortunately, the current DAA regimens are highly effective in HCV eradication for patients with cirrhosis and/or HCC.28 All of the cirrhotic patients and the five HCV-HCC patients achieved SVR12 after being linked to DAA therapy, except one cirrhotic patient experienced HCV relapsed.
Actually, the COMPACT campaign was affected by the global pandemic of SARS Co–V2 after March 2020,29 , 30 resulting in a 6-month delay in completing the project. It is estimated that the pandemic of COVID-19 has markedly impacted the HCV care continuum, in terms of testing access and treatment access.31 The global impact of a 1-year delay in HCV care has resulted in the loss of HCV treatment and the emergence of new infections, leading to enormous liver-related clinical outcomes.31 , 32 Although the COMPACT campaign successfully caught up with the screening target afterward based on the strategies of decentralized onsite service, the SARS Co–V2 pandemic did have an impact on the link-to-care of HCV viremic subjects. However, we demonstrated that the impact of the SARS Co–V2 pandemic was mainly observed in the Control group, but the impact was only minimal in the Target group. The results further emphasized the importance of decentralized onsite services provided in the COMPACT campaign in minimizing the impact of the SARS Co–V2 pandemic on HCV care.
The major societies including US Preventative Services Task Force,33 Centers for Disease Control and Prevention in US,34 and American Association for the Study of Liver Diseases–Infectious Diseases Society of America35 have advocated that rather than baby boomers, all adults should receive universal HCV screening. The universal screen of the adults in the COMPACT has shown that one-fifth of HCV-viremic subjects in the Target Population was with liver cirrhosis and 6 patients presented with HCC at the time of screening. The results reinforce the urgent need of a continuous call for action to HCV elimination.
There are limitations to the outreach model. Firstly, the COMPACT campaign provided service of diagnosis, assessment and treatment for HCV only. Patients have to visit the other clinic to manage their comorbidities if any. Secondly, the outreach teams do not provide sustainable medical services. We found that 7.9% of HCV non-viremic subjects and >20% of HCV-viremic subjects had fibroscan results greater than 12.5 kPa in the community-based study. As HCC remains to occur in the post-SVR era in particular for patients with advanced liver disease, continue surveillance of these patients is warranted.36 , 37 How to successfully link the subgroup to long-term monitoring is a challenge after the COMPACT campaign.
In conclusion, HCV remains a major health threat with tremendous gaps in HCV care cascades in the hyper-endemic area. The COMPACT campaign, a people-centered, outreach door-by-door screening with outreach decentralized onsite treatment program in collaboration with the health authority and local opinion leaders, greatly improved the HCV diagnosis and accessibility and scaled up the HCV treatment uptake in an HCV hyperendemic area. The strategies have successfully overcome the hurdles of each HCV care cascade and could serve as a model for HCV elimination in marginalized communities with high-risk populations, even during the pandemic of SARS Co–V2.
Appendix A Supplementary data
The following are the Supplementary data to this article.Multimedia component 1
Multimedia component 1
Multimedia component 2
Multimedia component 2
Multimedia component 3
Multimedia component 3
Acknowledgments
The study was supported by grants from 10.13039/501100011645 Kaohsiung Medical University Hospital (KMUH-DK(B)111002-1, KMHK-DK(C)111006, KMUH108-8R06, KMUH110-0R05, KMUH107-7M03, KMUH108-8T04, KMUH109-9T03).
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jmii.2023.04.012.
==== Refs
References
1 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 4 2013 1333 1342 23172780
2 Ioannou G.N. Green P.K. Berry K. HCV eradication induced by direct-acting antiviral agents reduces the risk of hepatocellular carcinoma J Hepatol 68 1 2017 25 32
3 Dang H. Yeo Y.H. Yasuda S. Huang C.F. Iio E. Landis C. Cure with interferon-free direct-acting antiviral is associated with increased survival in patients with hepatitis C virus-related hepatocellular carcinoma from both east and west Hepatology 71 6 2020 1910 1922 31610027
4 Yu M.L. Huang C.F. Yeh M.L. Tsai P.C. Huang C.I. Hsieh M.H. Time-degenerative factors and the risk of hepatocellular carcinoma after antiviral therapy among hepatitis C virus patients: a model for prioritization of treatment Clin Cancer Res 23 7 2017 1690 1697 27733478
5 Yang J.F. Lin C.I. Huang J.F. Dai C.Y. Lin W.Y. Ho C.K. Viral hepatitis infections in southern Taiwan: a multicenter community-based study Kaohsiung J Med Sci 26 9 2010 461 469 20837342
6 Yu M.L. Lin S.M. Lee C.M. Dai C.Y. Chang W.Y. Chen S.C. A simple noninvasive index for predicting long-term outcome of chronic hepatitis C after interferon-based therapy Hepatology 44 5 2006 1086 1097 17058238
7 Chen C.H. Yang P.M. Huang G.T. Lee H.S. Sung J.L. Sheu J.C. Estimation of seroprevalence of hepatitis B virus and hepatitis C virus in Taiwan from a large-scale survey of free hepatitis screening participants J Formos Med Assoc 106 2 2007 148 155 17339159
8 Yu M.L. Yeh M.L. Tsai P.C. Huang C.I. Huang J.F. Huang C.F. Huge gap between clinical efficacy and community effectiveness in the treatment of chronic hepatitis C: a nationwide survey in Taiwan Medicine (Baltim) 94 13 2015 e690
9 Yu M.L. Chen P.J. Dai C.Y. Hu T.H. Huang C.F. Huang Y.H. 2020 Taiwan consensus statement on the management of hepatitis C: part (I) general population J Formos Med Assoc 119 6 2020 1019 1040 32359879
10 Yu M.L. Chen P.J. Dai C.Y. Hu T.H. Huang C.F. Huang Y.H. 2020 Taiwan consensus statement on the management of hepatitis C: Part (II) special populations J Formos Med Assoc 119 7 2020 1135 1157 32354689
11 Chen D.S. Taiwan commits to eliminating hepatitis C in 2025 Lancet Infect Dis 19 5 2019 466 467 31034387
12 Lazarus J.V. Wiktor S. Colombo M. Thursz M. Micro-elimination - a path to global elimination of hepatitis C J Hepatol 67 4 2017 665 666 28760329
13 Heffernan A. Cooke G.S. Nayagam S. Thursz M. Hallett T.B. Scaling up prevention and treatment towards the elimination of hepatitis C: a global mathematical model Lancet 393 10178 2019 1319 1329 30704789
14 Yu M.L. Huang C.F. Wei Y.J. Lin W.Y. Lin Y.H. Hsu P.Y. Establishment of an outreach, grouping healthcare system to achieve microelimination of HCV for uremic patients in haemodialysis centres (ERASE-C) Gut 70 12 2020 2349 2358 33303567
15 Huang J.F. Lu S.N. Chue P.Y. Lee C.M. Yu M.L. Chuang W.L. Hepatitis C virus infection among teenagers in an endemic township in Taiwan: epidemiological and clinical follow-up studies Epidemiol Infect 127 3 2001 485 492 11811882
16 Yu M.L. Chuang W.L. Chen S.C. Dai C.Y. Hou C. Wang J.H. Changing prevalence of hepatitis C virus genotypes: molecular epidemiology and clinical implications in the hepatitis C virus hyperendemic areas and a tertiary referral center in Taiwan J Med Virol 65 1 2001 58 65 11505444
17 Tsai P.C. Huang C.I. Yeh M.L. Huang C.F. Hsieh M.H. Yang J.F. Significant amelioration of hepatitis C virus infection in a hyperendemic area: longitudinal evidence from the COMPACT Study in Taiwan BMJ Open 11 3 2021 e042861
18 Tsai P.C. Dai C.Y. Huang C.I. Yeh M.L. Huang C.F. Hsieh M.H. Low disease awareness as a contributing factor to the high prevalence of hepatitis C infection in Tzukuan, a hyperendemic area of southern Taiwan Kaohsiung J Med Sci 38 7 2022 694 702 35485737
19 Huang C.F. Wu P.F. Yeh M.L. Huang C.I. Liang P.C. Hsu C.T. Scaling up the in-hospital hepatitis C virus care cascade in Taiwan Clin Mol Hepatol 27 1 2021 136 143 33317253
20 Holmberg S.D. Lu M. Rupp L.B. Lamerato L.E. Moorman A.C. Vijayadeva V. Noninvasive serum fibrosis markers for screening and staging chronic hepatitis C virus patients in a large US cohort Clin Infect Dis 57 2 2013 240 246 23592832
21 Klionsky D.J. Abdel-Aziz A.K. Abdelfatah S. Abdellatif M. Abdoli A. Abel S. Guidelines for the use and interpretation of assays for monitoring autophagy (4th edition)(1) Autophagy 17 1 2021 1 382 33634751
22 Castéra L. Vergniol J. Foucher J. Bail B.L. Chanteloup E. Haaser M. Prospective comparison of transient elastography, Fibrotest, APRI, and liver biopsy for the assessment of fibrosis in chronic hepatitis C Gastroenterology 128 2 2005 343 350 15685546
23 Wang C.C. Liu C.H. Lin C.L. Wang P.C. Tseng T.C. Lin H.H. Fibrosis index based on four factors better predicts advanced fibrosis or cirrhosis than aspartate aminotransferase/platelet ratio index in chronic hepatitis C patients J Formos Med Assoc 114 10 2015 923 928 26279173
24 Chen Y.Y. Chen C.L. Chen J.W. Hsu N.Z. Wei S.T. Hou S.M. Secular trends and geographic maps of hepatitis C virus infection among 4 million blood donors in taiwan from 1999 to 2017 Hepatol Commun 4 8 2020 1193 1205 32766478
25 Zeuzem S. Alberti A. Rosenberg W. Marcellin P. Diago M. Negro F. Review article: management of patients with chronic hepatitis C virus infection and "normal" alanine aminotransferase activity Aliment Pharmacol Ther 24 8 2006 1133 1149 17014573
26 Huang C.F. Yu M.L. Unmet needs of chronic hepatitis C in the era of direct-acting antiviral therapy Clin Mol Hepatol 26 3 2020 251 260 32188235
27 Bajis S. Applegate T.L. Grebely J. Matthews G.V. Dore G.J. Novel hepatitic C virus (HCV) diagnosis and treatment delivery systems: facilitating HCV elimination by thinking outside the clinic J Infect Dis 222 Suppl 9 2020 S758 S772 33245354
28 Ogawa E. Toyoda H. Iio E. Jun D.W. Huang C.F. Enomoto M. Hepatitis C virus cure rates are reduced in patients with active but not inactive hepatocellular carcinoma: a practice implication Clin Infect Dis 71 11 2020 2840 2848 31777940
29 Lin C.Y. Wang W.H. Urbina A.N. Tseng S.P. Lu P.L. Chen Y.H. Importation of SARS-CoV-2 infection leads to major COVID-19 epidemic in Taiwan Int J Infect Dis 97 2020 240 244 32544668
30 Jang T.Y. Wang H.H. Huang C.F. Dai C.Y. Huang J.F. Chuang W.L. Clinical characteristics and treatment outcomes of SARS-CoV-2 delta variant outbreak, Pingtung, Taiwan, June 2021 J Formos Med Assoc 121 9 2022 1767 1772 35067428
31 Wingrove C. Ferrier L. James C. Wang S. The impact of COVID-19 on hepatitis elimination Lancet Gastroenterol Hepatol 5 9 2020 792 794 32730783
32 Blach S. Kondili L.A. Aghemo A. Cai Z. Dugan E. Estes C. Impact of COVID-19 on global HCV elimination efforts J Hepatol 74 1 2021 31 36 32777322
33 Koretz R.L. USPSTF recommends screening adults 18 to 79 years for hepatitis C virus infection (moderate certainty) Ann Intern Med 173 2 2020 JC1 32687755
34 Schillie S. Wester C. Osborne M. Wesolowski L. Ryerson A.B. CDC recommendations for hepatitis C screening among adults - United States, 2020 MMWR Recomm Rep (Morb Mortal Wkly Rep) 69 2 2020 1 17
35 Ghany M.G. Morgan T.R. Hepatitis C guidance 2019 update: American association for the study of liver diseases-infectious diseases society of America recommendations for testing, managing, and treating hepatitis C virus infection Hepatology 71 2 2020 686 721 31816111
36 Perez S. Kaspi A. Domovitz T. Davidovich A. Lavi-Itzkovitz A. Meirson T. Hepatitis C virus leaves an epigenetic signature post cure of infection by direct-acting antivirals PLoS Genet 15 6 2019 e1008181
37 Tanaka Y. Ogawa E. Huang C.F. Toyoda H. Jun D.W. Tseng C.H. HCC risk post-SVR with DAAs in East Asians: findings from the REAL-C cohort Hepatology international 14 6 2020 1023 1033 33277685
|
PMC010xxxxxx/PMC10148711.txt |
==== Front
J Hosp Infect
J Hosp Infect
The Journal of Hospital Infection
0195-6701
1532-2939
The Healthcare Infection Society. Published by Elsevier Ltd.
S0195-6701(23)00139-1
10.1016/j.jhin.2023.04.014
Article
Outcomes of influenza and COVID-19 inpatients in different phases of the SARS-CoV-2 pandemic: a single-centre retrospective case–control study
Bechmann L. ∗
Esser T.
Färber J.
Kaasch A.
Geginat G.
Department of Medical Microbiology and Infection Control, Otto-von-Guericke University Magdeburg, Magdeburg, Germany
∗ Corresponding author. Address: Institut für Medizinische Mikrobiologie und Krankenhaushygiene, Uniklinikum Magdeburg, Leipziger Str. 44, 39120 Magdeburg, Germany. Tel.: +493916717820; fax: +493916713384.
29 4 2023
8 2023
29 4 2023
138 17
9 2 2023
23 4 2023
© 2023 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.
2023
The Healthcare Infection Society
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
The virulence of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) changed during the pandemic. In order to provide a rationale for treatment priorities of respiratory infections and the adaption of in-house infection control strategies, this study evaluated treatment on an intensive care unit (ICU), requirement for mechanical ventilation (MV), requirement for extracorporeal membrane oxygenation (ECMO) and death for inpatients infected with the influenza virus or SARS-CoV-2 during the wild-type, Alpha, Delta, Omicron BA.1/2 and Omicron BA.5 waves of the pandemic.
Design
Single-centre retrospective case–control study.
Setting
Tertiary hospital in Germany.
Participants
One thousand three hundred and sixteen adult inpatients infected with SARS-CoV-2 and 218 adult inpatients infected with influenza virus.
Methods
Demographic data, outcome parameters and underlying comorbidities of patients were obtained from the hospital information system. Multi-variate regression analysis was performed for the assessment of significant associations between risk factors and outcome variables.
Results
Compared with inpatients infected with influenza virus, patients infected with SARS-CoV-2 showed significantly higher rates for in-hospital mortality, admission to ICU and requirement for MV in the wild-type, Alpha and Delta waves, and a significantly higher rate for requirement for ECMO in the wild-type wave. In the Omicron BA.1/BA.2 and Omicron BA.5 waves, patients infected with SARS-CoV-2 did not show significantly higher risk of in-hospital mortality, admission to ICU, or requirement for MV or ECMO compared with patients infected with influenza virus. The length of hospital stay of patients infected with SARS-CoV-2 decreased from 10.8 to 6.2 days, which was less than that of patients infected with influenza virus (8.3 days).
Conclusions
Treatment capacities should be shared equally between SARS-CoV-2 and influenza virus infections. Similar levels of infection control could be applied, at least regarding the severity of infection.
Keywords
COVID-19
SARS-CoV-2
Variants
Alpha
Delta
Omicron
Influenza virus
Mortality
ICU
Ventilation
ECMO
==== Body
pmcBackground
Influenza and coronavirus disease 2019 (COVID-19) are both contagious acute respiratory infections. Influenza is caused by influenza A and B viruses. Annual seasonal influenza epidemics result in three to five million cases of severe illness and up to half a million deaths per year. Symptoms of influenza infection may include abrupt onset of respiratory symptoms, myalgia and fever. In most cases, patients recover within 1 week, but some patients can experience severe complications such as bacterial pneumonia and acute respiratory distress syndrome. Different influenza vaccines are available for prophylaxis, and antiviral drugs targeting neuraminidase, a viral surface glycoprotein, can be used for treatment [[1], [2], [3]].
Common symptoms of COVID-19, caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), include fever, cough, chest discomfort, olfactory and gustatory disturbances, and, in severe cases, dyspnoea and bilateral lung infiltration. Vaccination against SARS-CoV-2 can provide partial protection against infection, whereby the protective effect depends on the vaccine, the number of vaccination doses, the time of the last vaccination dose, and the variant of SARS-CoV-2. Various intravenous and oral antiviral drugs are available for treatment of COVID-19. Furthermore, convalescent plasma may be an option for treatment, especially in patients with immunosuppression [[4], [5], [6], [7], [8], [9]].
According to the World Health Organization, more than 6.6 million people died of COVID-19 worldwide between 2020 and 2022, including more than 163,000 in Germany [10]. Retrospectively, the pandemic can be classified into phases dominated by specific SARS-CoV-2 variants. In Germany, the wild-type variant of SARS-CoV-2 was dominant until calendar week (CW) 08/2021. This was followed by the Alpha variant (from CW 09/2021 to CW 23/2021) and the Delta variant (from CW 31/21 to CW 51/2021) in the third and fourth waves. Subsequently, infections in Germany were mainly caused by various Omicron variants (fifth wave: BA.1/BA.2 dominated from CW 52/2021 to CW 21/2022; sixth wave: BA.5 dominated from CW 22/2022 onwards) [11].
After very low incidence of influenza in Germany in 2020 and 2021, a ‘twindemic’ situation was present in late 2022, and a large number of inpatients infected with influenza or COVID-19 had to be treated simultaneously in hospitals. As the simultaneous presence of both viruses is anticipated in the coming years, a head-to-head comparison of the outcome parameters of inpatients with COVID-19 and influenza was performed in order to better prioritize treatment capacities and the level of required infection control precautions for both diseases.
Methods
Setting
University Hospital Magdeburg is a tertiary hospital in central Germany. It has approximately 4,700 staff members and treats approximately 50,000 inpatients annually. In 2020 and 2021, SARS-CoV-2-positive patients were treated in dedicated COVID-19 isolation wards [normal units under the management of the Infectious Diseases Department, and intensive care units (ICUs) led by the Anaesthesiology Department]. In 2022, there was an increasing change in strategy towards treating SARS-CoV-2-positive patients in single bedrooms (with anteroom if possible) on general wards according to the underlying admission diagnosis. SARS-CoV-2 polymerase chain reaction (PCR) admission screening of all inpatients was established in May 2020 and is still in place.
Data collection
Data processing was based on pseudonymized patient data obtained retrospectively, and did not include any experiments involving human participants (including the use of tissue samples). The requirement for informed consent was waived by the Ethics Committee of University Hospital Magdeburg. Lists of all inpatients positive for SARS-COV-2 and influenza virus were available in the section infection control from routine hospital surveillance. Demographic data (age, gender), outcome parameters [ICU admission, requirement for mechanical ventilation (MV), requirement for extracorporeal membrane oxygenation (ECMO), death], underlying comorbidities, main diagnosis, and dates of admission and discharge were obtained from the Hospital Information System for each case. The assignment of a case to the corresponding SARS-CoV-2 wave was based on the admission date, according to Tolksdorf et al. [11].
Case-control study, and inclusion and exclusion criteria
Adult inpatients between January 2018 and December 2022 who had a positive SARS-CoV-2 or influenza PCR result before or within 5 days of admission were included in this study. Patients were excluded if they had been treated for >24 h at another acute hospital before admission to the study hospital, as these patients were a pre-selected group and may have been at increased risk of a severe course of disease. Characteristics of patients positive for SARS-CoV-2 and the influenza virus are shown in Table I .Table I Characteristics and outcomes of all inpatients with severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) or influenza virus infection (all)/subgroup of patients with acute respiratory main diagnosis (RMD)
Table I SARS-CoV-2 waves Influenza virus
Pre-Omicron Omicron
Wild-type Alpha Delta BA.1/BA.2 BA.5
In-hospital stay
Number of cases (all/RMD) 176/128 128/103 133/101 320/95 370/91 182/86
Acute RMD (%) 73% 80% 76% 30% 25% 47%
Mean length of stay (days) (all/RMD) 10.8/11.2 10.4/11.3 10.2/11.6 7.2/9.2 6.2/7.2 8.3/7.7
Demographic data
Mean age (years) (all/RMD) 67.5/67.6 62.3/64.7 63.4/67.0 60.7/68.6 64.4/76.0 66.0/69.4
Male gender (%) (all/RMD) 59.7/66.4 46.9/52.4 51.9/55.4 43.1/58.9 44.9/57.1 51.1/57.0
Comorbidities
Diabetes mellitus (%) (all/RMD) 31.8/28.1 27.3/29.1 23.3/25.7 23.8/27.4 22.2/33.0 28.0/36.0
Arterial hypertension (%) (all/RMD) 74.4/77.3 65.6/70.9 63.9/70.3 49.1/58.9 51.6/79.1 56.6/65.1
Cancer (%) (all/RMD) 4.5/3.1 0.8/1.0 3.8/3.0 6.6/1.1 5.4/3.3 3.3/2.3
Chronic renal failure (%) (all/RMD) 28.4/29.7 25.0/25.2 24.1/29.7 22.2/34.7 18.9/28.6 23.6/29.1
Haematological disease (%) (all/RMD) 1.7/2.3 3.1/3.9 4.5/5.9 2.2/4.2 4.3/4.4 4.4/5.8
Neurological disorder (%) (all/RMD) 24.4/23.4 17.2/16.5 15.8/14.9 22.5/23.2 18.6/28.6 19.2/17.4
Respiratory disorder (%) (all/RMD) 10.8/11.7 16.4/18.4 15.0/14.9 13.4/20.0 10.8/16.5 22.0/29.1
Chronic heart disease (%) (all/RMD) 32.4/33.6 20.3/24.3 21.8/20.8 18.8/25.3 22.4/37.4 34.1/32.6
Outcomes
Death (%) (all/RMD) 19.3/22.7 17.2/16.5 15.0/17.8 9.1/12.6 7.0/12.1 8.8/10.5
ICU admission (%) (all/RMD) 27.3/33.6 18.0/21.4 17.3/20.8 8.1/7.4 3.2/3.3 7.1/5.8
Requirement for MV (%) (all/RMD) 18.2/25.0 7.8/9.7 7.5/9.9 3.1/2.1 1.6/2.2 2.7/2.3
Requirement for ECMO (%) (all/RMD) 4.0/5.5 0.8/1.0 1.5/2.0 0.0/0.0 0.0/0.0 0.0/0.0
ICU, intensive care unit; MV, mechanical ventilation; ECMO, extracorporeal membrane oxygenation.
In individual cases, it was difficult to decide whether a patient died or required ICU treatment, MV or ECMO due to the course of the infection or because of an underlying disease. In the latter case, the infection was considered as a secondary diagnosis. For this reason, multi-variate regression analyses were performed for inpatients with a positive influenza virus or SARS-CoV-2 test, and for the subgroup of patients with acute respiratory tract infection as the main diagnosis (including International Classification of Diseases Version 10 codes J04, J06, J10, J11, J12, J15, J18, J22, J80, J96 and J98). Patients in this subgroup are henceforth referred to as patients with COVID-19- or influenza-related respiratory main diagnosis (RMD).
Statistics
Data analyses were performed using Excel 2016 (Microsoft Corp., Redmond, WA, USA). Multi-variate regression analysis was used for the assessment of significant associations between risk factors (predominant SARS-CoV-2 variant/influenza, age, gender, diabetes mellitus, arterial hypertension, cancer, renal failure, haematological malignancies, neurological disorders, chronic respiratory diseases, chronic heart disease) and outcome variable (ICU admission, requirement for MV, requirement for ECMO, death). P<0.05 was considered to indicate significance. Bonferroni's correction was not used because of the explorative character of the study and because death, ICU admission, requirement for MV and requirement for ECMO are not strictly independent (e.g. inpatients who require MV or ECMO are always in ICUs).
Results
Characteristics of patients with SARS-CoV-2 and influenza infection are shown in Table I. Seventy-four percent of all inpatients infected with influenza had an influenza-related RMD within the observation period. While at the beginning of the pandemic, most SARS-CoV-2-positive patients had COVID-19-related RMD (wild-type wave, 73%; Alpha wave, 80%; Delta wave, 76%), in 2022, SARS-CoV-2-positive patients were more often admitted with other main diagnoses, and COVID-19-related RMD decreased to 30% in the Omicron BA.1/2 wave and 25% in the Omicron BA.5 wave. During the pandemic, the mean length of hospital stay of patients with COVID-19-related RMD decreased from 11.2 days in the wild-type wave to 7.2 days in the Omicron BA.5 wave, which was less than the mean length of hospital stay of patients with influenza-related RMD (7.7 days). The mean age of inpatients with COVID-19-related RMD decreased from 67.6 years during the wild-type wave to 64.7 years during the Alpha wave, and subsequently increased steadily to 76.0 years in the Omicron BA.5 wave. The mean age of inpatients hospitalized with influenza-related RMD was 69.4 years over the period from 2018 to 2022.
In-house mortality of patients with COVID-19-related RMD decreased from 22.7% at the beginning of the pandemic to 12.1% in the Omicron BA.5 wave, which was close to the mortality of patients with influenza-related RMD (10.5%), especially when considering the higher age of the patients with COVID-19 in the last Omicron wave. The ICU admission rate decreased from 33.6% to 3.3%, and the MV rate decreased from 25.0% to 2.2% in inpatients with COVID-19-related RMD, which was lower than the ICU admission rate (5.8%) and the MV rate (2.3%) of inpatients with influenza-related RMD (Table I).
The results of the multi-variate regression analysis of risk factors for death, ICU admission, requirement for MV and requirement for ECMO are shown in Table II and Figure 1 . In summary, during the SARS-CoV-2 wild-type wave, rates of mortality (P<0.005), ICU admission (P<0.005), requirement for MV (P<0.005) and requirement for ECMO (P<0.005) were significantly higher in all patients with COVID-19 compared with patients with influenza. In the Alpha and Delta waves, significantly higher rates of mortality (Alpha wave, P<0.005; Delta wave, P<0.05), ICU admission (both P<0.005) and requirement for MV (both P<0.05) were found in patients with COVID-19 compared with patients with influenza, but no significant differences were found in requirement for ECMO. In the Omicron BA.1/2 and BA.5 waves, no significant differences in mortality, ICU admission, requirement for MV or requirement for ECMO were found between patients with COVID-19 and patients with influenza.Table II Multi-variate regression analysis of risk factors for death, intensive care unit (ICU) admission, requirement for mechanical ventilation (MV) and requirement for extracorporeal membrane oxygenation (ECMO) for either all patients with coronavirus disease 2019 (COVID-19) or influenza and the subgroup of patients with acute respiratory main diagnosis (RMD) alone
Table IIFactor Outcome RMD alone All patients with COVID-19 or influenza
Coefficient Standard error P-value Coefficient Standard error P-value
SARS-CoV-2 wild-type wave compared with influenza Death 0.143 0.050 0.004 0.112 0.032 0.001
ICU 0.269 0.050 0.000 0.193 0.032 0.000
MV 0.225 0.040 0.000 0.151 0.024 0.000
ECMO 0.053 0.018 0.004 0.040 0.009 0.000
SARS-CoV-2 Alpha wave compared with influenza Death 0.101 0.052 0.053 0.116 0.035 0.001
ICU 0.148 0.052 0.005 0.109 0.035 0.002
MV 0.081 0.042 0.051 0.055 0.026 0.033
ECMO 0.009 0.019 0.629 0.008 0.010 0.418
SARS-CoV-2 Delta wave compared with influenza Death 0.092 0.052 0.078 0.085 0.035 0.014
ICU 0.150 0.053 0.004 0.103 0.035 0.003
MV 0.085 0.042 0.042 0.051 0.026 0.046
ECMO 0.020 0.019 0.284 0.015 0.010 0.134
SARS-CoV-2 Omicron BA.1 and BA.2 wave compared with influenza Death 0.028 0.053 0.599 0.021 0.028 0.449
ICU 0.026 0.053 0.630 0.014 0.028 0.616
MV 0.005 0.042 0.900 0.008 0.021 0.713
ECMO 0.002 0.019 0.909 0.000 0.008 0.972
SARS-CoV-2 Omicron BA.5 wave compared with influenza Death -0.021 0.054 0.697 -0.012 0.028 0.662
ICU -0.014 0.054 0.793 -0.031 0.028 0.268
MV -0.005 0.043 0.902 -0.005 0.020 0.821
ECMO 0.002 0.019 0.918 0.001 0.008 0.934
Age (life years) Death 0.007 0.001 0.000 0.004 0.001 0.000
ICU -0.003 0.001 0.003 -0.001 0.001 0.166
MV -0.001 0.001 0.525 0.000 0.000 0.893
ECMO 0.000 0.000 0.258 0.000 0.000 0.295
Male gender Death 0.001 0.030 0.972 -0.009 0.017 0.611
ICU 0.035 0.030 0.239 0.036 0.017 0.037
MV 0.021 0.024 0.377 0.024 0.013 0.052
ECMO 0.011 0.011 0.318 0.007 0.005 0.133
Diabetes mellitus Death -0.012 0.034 0.730 0.015 0.021 0.466
ICU 0.000 0.034 0.999 -0.021 0.021 0.313
MV 0.012 0.027 0.648 -0.013 0.015 0.384
ECMO -0.002 0.012 0.886 -0.003 0.006 0.628
Arterial hypertonia Death -0.073 0.036 0.047 -0.082 0.022 0.000
ICU 0.068 0.037 0.063 0.037 0.022 0.089
MV 0.023 0.029 0.433 0.009 0.016 0.589
ECMO 0.005 0.013 0.706 0.004 0.006 0.526
Cancer Death 0.191 0.096 0.047 0.116 0.040 0.004
ICU 0.188 0.096 0.051 -0.023 0.040 0.572
MV 0.176 0.077 0.022 0.026 0.029 0.371
ECMO 0.052 0.035 0.136 0.011 0.011 0.354
Renal failure Death -0.011 0.036 0.758 0.022 0.023 0.338
ICU -0.050 0.036 0.164 0.009 0.023 0.681
MV -0.022 0.028 0.446 0.022 0.017 0.182
ECMO -0.016 0.013 0.226 -0.007 0.006 0.265
Haematological malignancies Death 0.118 0.071 0.096 0.073 0.047 0.119
ICU 0.040 0.072 0.580 0.021 0.047 0.650
MV 0.003 0.057 0.958 -0.001 0.034 0.978
ECMO -0.014 0.026 0.577 -0.006 0.013 0.654
Neurological disease Death -0.007 0.036 0.847 0.030 0.021 0.158
ICU 0.055 0.036 0.128 0.086 0.021 0.000
MV 0.066 0.029 0.022 0.049 0.016 0.002
ECMO -0.001 0.013 0.937 -0.001 0.006 0.925
Chronic respiratory disease Death -0.011 0.038 0.773 -0.002 0.025 0.924
ICU 0.064 0.038 0.094 0.033 0.025 0.179
MV 0.046 0.030 0.133 0.032 0.018 0.073
ECMO 0.008 0.014 0.569 0.006 0.007 0.385
Chronic heart disease Death 0.027 0.035 0.450 0.037 0.022 0.091
ICU 0.006 0.035 0.855 0.014 0.022 0.527
MV 0.026 0.028 0.352 0.010 0.016 0.536
ECMO 0.007 0.013 0.600 0.003 0.006 0.683
SARS-CoV-2, severe acute respiratory syndrome coronavirus-2.
Bold values denote statistical significance at the P < 0.05 level.
Figure 1 Multi-variate linear regression analysis of risk factors for death, intensive care unit (ICU) admission, requirement for mechanical ventilation and requirement for extracorporeal membrane oxygenation (ECMO) in all patients infected with severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) and influenza virus (solid bars), and the subgroup of patients infected with SARS-CoV-2 and influenza virus with acute respiratory main diagnosis (open bars). Error bars indicate standard error from the mean. ∗P<0.05. ∗∗P<0.005.
Figure 1
Considering only the subgroups of patients with COVID-19-/influenza-related RMD, significantly higher rates of mortality, ICU admission, requirement for MV and requirement for ECMO (all P<0.005) were found in the patients with COVID-19 compared with the patients with influenza in the wild-type wave (Figure 1). Later in the pandemic, there was still a significantly higher rate of ICU admission (P<0.005) in the Alpha wave, and significantly higher rates of ICU admission (P<0.005) and requirement for MV (P<0.05) in the Delta wave in patients with COVID-19 compared with patients with influenza (Figure 1, Table I).
In addition, the outcome parameters were evaluated in different age groups (Figure 2 ). In the Omicron BA.5 wave, patients with COVID-19-related RMD in all three age groups (18–57 years, 58–77 years, 78–97 years) showed equal or lower rates of ICU admission, requirement for MV and requirement for ECMO compared with patients with influenza-related RMD. Considering mortality in the Omicron BA.5 wave, patients with COVID-19-related RMD showed higher mortality in the age group 58–77 years and lower mortality in the age group 78–97 years compared with patients with influenza-related RMD.Figure 2 Relative disease severity of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) variants compared with influenza in different age groups. (a) Intensive care unit (ICU) admission rate in SARS-CoV-2-positive patients with respiratory main diagnosis (RMD) compared with all influenza-virus-positive patients with RMD (as baseline) in three different age groups. (b) Requirement for mechanical ventilation in SARS-CoV-2-positive patients with RMD compared with all influenza-virus-positive patients with RMD (as baseline) in three different age groups. (c) Requirement for extracorporeal membrane oxygenation in SARS-CoV-2-positive patients with RMD compared with all influenza-virus-positive patients with RMD (as baseline) in three different age groups. (d) In-house mortality in SARS-CoV-2-positive patients with RMD compared with all influenza-virus-positive patients with RMD (as baseline) in three different age groups. White bars, 18–57 years; grey bars, 58–77 years; black bars, 78–97 years.
Figure 2
Discussion
During the wild-type, Alpha and Delta waves, the mortality and ICU admission rates were significantly higher in patients with COVID-19 compared with patients with influenza, while no significant differences in these outcome parameters were found during the Omicron BA.1/2 and Omicron BA.5 waves.
Significantly higher mortality rates of hospitalized patients with COVID-19 compared with hospitalized patients with influenza during the SARS-CoV-2 wild-type wave of the pandemic have been reported previously [[12], [13], [14], [15]]. In these previous studies from the USA, France, Germany and Belgium, in-hospital mortality rates for patients with wild-type SARS-CoV-2 infection ranged from 14.0% to 20.0%, while in-hospital mortality rates for patients with influenza virus infection ranged from 5.0% to 9.8%. In the present study, the mortality rates for patients with SARS-CoV-2 infection and influenza virus infection were 19.3% and 8.8%, respectively, and thus were in the previously reported range.
During the wild-type wave, the ICU admission rates in these studies ranged from 15.0% to 36.8% for patients with COVID-19 and from 10.8% to 24% for patients with influenza. MV rates ranged from 9.7% to 15.0% for patients with COVID-19 and from 4.0% to 9.0% for patients with influenza. While the rates for ICU admission (27.3%) and requirement for MV (18.2%) for the patients with COVID-19 in this study are in the previously reported range, the rates for the patients with influenza in this study were slightly lower (ICU admission rate, 7.1%; requirement for MV rate, 2.7%). This could be explained, at least in part, by the fact that the average age of patients with influenza who died without being transferred to an ICU was 83 years. Many of these patients were multi-morbid and had explicitly expressed their opposition to intensive care measures.
The mortality rate of patients with COVID-19 in this study decreased from 15.0% in the Delta wave to 7.0% in the Omicron BA.5 wave. This reduction of 53.3% is in line with the observations of other studies [[16], [17], [18], [19]], which described a reduction in mortality rate between 44.4% and 61.7% between the Delta and Omicron waves.
Compared with the Delta wave, the ICU admission rate of patients with COVID-19 decreased by 40% and 58.5% in the Omicron BA.1/2 and Omicron BA.5 waves, respectively. Bouzid et al. [17] described a 75% reduction in the ICU admission rate between the Delta and early Omicron waves in hospitals in France. Adjei et al. [18] reported a reduction of 14.4% between the Delta and early Omicron waves, and of 46.8% between the Delta and late Omicron waves in the USA.
Compared with the Delta wave, this study found a 58.7% and 78.7% reduction in the rate of requirement for MV for patients with COVID-19 in the Omicron BA.1/2 and Omicron BA.5 waves, respectively. Bouzid et al. [17] described a 52% reduction in the ICU admission rate between the Delta and early Omicron waves in France. Adjei et al. [18] reported a reduction of 22.4% between the Delta and early Omicron waves, and of 65% between the Delta and late Omicron waves in the USA.
Interestingly, in comparison with patients with influenza-virus-related RMD, patients with COVID-19-related RMD showed higher mortality in the age group 58–77 years and lower mortality in the age group 78–97 years in the Omicron BA.5 wave. However, due to the small sample size (five patients died in this age group during the Omicron BA.5 wave) and the severe pre-existing comorbidities of some of these patients, this might be a statistical outlier. Further studies are needed to corroborate this observation.
This study has some limitations. First, it is a retrospective study. While the patient demographics and outcome variables are fairly reliable, there is some uncertainty in the complete coverage of the listed comorbidities. Furthermore, in the first SARS-CoV-2 wave, a greater proportion of oligosymptomatic patients were probably hospitalized over a longer period, as there was no possibility to isolate these patients in care facilities, such as nursing homes, at the beginning of the pandemic. While PCR admission screening of all inpatients has been in place for SARS-CoV-2 since May 2020, testing for influenza was only performed on symptomatic patients during the observation period. As a result, asymptomatic inpatients with influenza are often not detected, which can lead to an overestimation of disease severity in patients with influenza. In order to consider this, a corresponding subgroup analysis was undertaken for inpatients with RMD, which corroborated the general analysis of all virus-positive inpatients. In addition, due to the retrospective nature of the study, the impact of vaccination status on patient outcome could not be examined more closely, as vaccination status was only recorded sporadically in the patient files.
These data show that, regarding the severity of infection, it would be justified to apply similar levels of infection control for the prevention of the spread of COVID-19 and influenza infections in hospitals. However, for the implementation of infection control measures in hospitals, transmissibility is important as well as disease severity. Thus, a disease that has a lower mortality rate but significantly higher transmissibility can ultimately lead to more nosocomial deaths. For this reason, these data on the relative severity of community acquired COVID-19 and influenza infections should not be used as a basis for uncritically dispensing SARS-CoV-2 infection control measures in hospitals. Due to the longer incubation period and the higher transmission rate, we believe that stricter hygiene measures should be implemented in outbreak events, such as mandatory masks for all staff during the outbreak, individual breaks, etc.
In conclusion, in comparison with inpatients with influenza, inpatients infected with SARS-CoV-2 BA.1/2 and BA.5 do not show significantly higher risk of in-hospital mortality, ICU admission, requirement for MV or requirement for ECMO. Against this background, healthcare system resources should be shared equally between the two diseases.
Conflict of interest statement
None declared.
Funding sources
None.
==== Refs
References
1 Krammer F. Smith G.J.D. Fouchier R.A.M. Peiris M. Kedzierska K. Doherty P.C. Influenza. Nat Rev Dis Primers 4 2018 3 29955068
2 Uyeki T.M. Hui D.S. Zambon M. Wentworth D.E. Monto A.S. Influenza Lancet 400 2022 693 706 36030813
3 Javanian M. Barary M. Ghebrehewet S. Koppolu V. Vasigala V. Ebrahimpour S. A brief review of influenza virus infection J Med Virol 93 2021 4638 4646 33792930
4 Fu L. Wang B. Yuan T. Chen X. Ao Y. Fitzpatrick T. Clinical characteristics of coronavirus disease 2019 (COVID-19) in China: a systematic review and meta-analysis J Infect 80 2020 656 665 32283155
5 Hu B. Guo H. Zhou P. Shi Z.L. Characteristics of SARS-CoV-2 and COVID-19 Nat Rev Microbiol 19 2021 141 154 33024307
6 Wang D. Hu B. Hu C. Zhu F. Liu X. Zhang J. Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China JAMA 323 2020 1061 1069 32031570
7 Grote U. Arvand M. Brinkwirth S. Brunke M. Buchholz U. Eckmanns T. Maßnahmen zur Bewältigung der COVID-19-Pandemie in Deutschland: nichtpharmakologische und pharmakologische Ansätze [Measures to cope with the COVID-19 pandemic in Germany: nonpharmaceutical and pharmaceutical interventions] Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 64 2021 435 445 33787944
8 Sharp A. Jain V. Alimi Y. Bausch D.G. Policy and planning for large epidemics and pandemics – challenges and lessons learned from COVID-19 Curr Opin Infect Dis 34 2021 393 400 34342301
9 Mendez-Brito A. El Bcheraoui C. Pozo-Martin F. Systematic review of empirical studies comparing the effectiveness of non-pharmaceutical interventions against COVID-19 J Infect 83 2021 281 293 34161818
10 World Health Organization. Coronavirus (COVID-19) dashboard. Geneva: WHO. Available at: 10.1590/1516-3180.2021 [last accessed December 2022].
11 Tolksdorf K. Loenenbach A. Buda S. Dritte Aktualisierung der Retrospektiven Phaseneinteilung der COVID-19-Pandemie in Deutschland Epid Bull 38 2022 3 6
12 Xie Y. Bowe B. Maddukuri G. Al-Aly Z. Comparative evaluation of clinical manifestations and risk of death in patients admitted to hospital with COVID-19 and seasonal influenza: cohort study BMJ 371 2020 m4677 33323357
13 Ludwig M. Jacob J. Basedow F. Andersohn F. Walker J. Clinical outcomes and characteristics of patients hospitalized for influenza or COVID-19 in Germany Int J Infect Dis 103 2021 316 322 33279652
14 Piroth L. Cottenet J. Mariet A.S. Bonniaud P. Blot M. Tubert-Bitter P. Comparison of the characteristics, morbidity, and mortality of COVID-19 and seasonal influenza: a nationwide, population-based retrospective cohort study Lancet Respir Med 9 2021 251 259 33341155
15 Wallemacq S. Danwang C. Scohy A. Belkhir L. De Greef J. Kabamba B. A comparative analysis of the outcomes of patients with influenza or COVID-19 in a tertiary hospital in Belgium J Infect Chemother 28 2022 1489 1493 35944762
16 Nyberg T. Ferguson N.M. Nash S.G. Webster H.H. Flaxman S. Andrews N. Comparative analysis of the risks of hospitalisation and death associated with SARS-CoV-2 Omicron (B.1.1.529) and Delta (B.1.617.2) variants in England: a cohort study Lancet 399 2022 1303 1312 35305296
17 Bouzid D. Visseaux B. Kassasseya C. Daoud A. Fémy F. Hermand C. Comparison of patients infected with Delta versus Omicron COVID-19 variants presenting to Paris emergency departments: a retrospective cohort study Ann Intern Med 175 2022 831 837 35286147
18 Adjei S. Hong K. Molinari N.M. Bull-Otterson L. Ajani U.A. Gundlapalli A.V. Mortality risk among patients hospitalized primarily for COVID-19 during the Omicron and Delta variant pandemic periods – United States, April 2020–June 2022 MMWR Morb Mortal Wkly Rep 71 2022 1182 1189 36107788
19 Kim A.R. Lee J. Park S. Kang S.W. Lee Y.W. Lim S.Y. Comparison of the causes of death associated with Delta and Omicron SARS-CoV-2 variant infection J Infect Public Health 16 2023 133 135 36516648
|
PMC010xxxxxx/PMC10148967.txt |
==== Front
J Adolesc Health
J Adolesc Health
The Journal of Adolescent Health
1054-139X
1879-1972
Elsevier
S1054-139X(23)00152-0
10.1016/j.jadohealth.2023.02.042
Adolescent Health Brief
A Comparison of Pediatric Mental Health Diagnoses and Selective Serotonin Reuptake Inhibitor Prescribing Before and During the COVID-19 Pandemic
Keares Peter P. D.O. a∗
Pho Nhien T. M.D. b
Larson Richard S. D.N.P. b
Vallejo Julia C. b
a Saint Louis University (Southwest Illinois) Family Medicine Residency, O'Fallon, Illinois
b 375th Healthcare Operations Squadron Pediatrics, Scott AFB, Illinois
∗ Address correspondence to: Peter P. Keares, D.O., 3 St. Elizabeth Blvd. Suite 4000 O'Fallon, IL 62269.
30 4 2023
30 4 2023
9 6 2022
28 2 2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Purpose
To assess the rate of mental health diagnoses and selective serotonin reuptake inhibitor (SSRI) prescribing before and during the Coronavirus Disease 2019 pandemic.
Methods
We conducted a cross-sectional study at an ambulatory pediatric clinic. A prepandemic (June 2018 to June 2019) and intrapandemic (June 2020 to June 2021) cohort were reviewed. The rate of mental health visits and new SSRI prescriptions were compared. Chi-squared analyses demonstrated a variance of statistical significance.
Results
From 15,414 encounters (9,791 prepandemic and 5,623 intrapandemic), 397 mental health encounters were identified. 231 (4.1%) encounters occurred during the pandemic (vs. 1.7% prepandemic) and 63 (27.3%) SSRIs were prescribed (vs. 5.4% prepandemic). Mental health encounters (prevalence ratio 2.42, 95% confidence interval, 1.99–2.95, p < .001) and SSRI prescriptions (prevalence ratio 5.03, 95% confidence interval, 2.58–9.82, p < .001) were higher during the pandemic.
Discussion
Our findings demonstrate increased rates of SSRI prescribing and mental health diagnoses during the Coronavirus Disease 2019 pandemic, suggesting an increased incidence of these conditions. Clinicians should be prepared to manage and screen for mental health conditions.
Keywords
Mental health diagnoses
Anxiety
Depression
COVID-19
Selective serotonin reuptake inhibitors
Ambulatory pediatrics
Adolescent mental health
==== Body
pmc Implications and Contributions
As demonstrated in a military general pediatric clinic, the rate of mental health diagnoses and selective serotonin reuptake inhibitors (SSRIs) have increased during the COVID-19 pandemic. This study reinforces the importance of primary care providers in the diagnosis and treatment of mental health conditions and should encourage providers to ensure their clinics are adequately screening for and prepared to treat these conditions.
The Coronavirus Disease 2019 (COVID-19) pandemic and its mitigation measures have impacted daily routines, social structures, and financial security for a generation of children [1]. Consequently, emotional stresses, fears, and feelings of helplessness increase a child's risk of mental health conditions like depression, anxiety, and post-traumatic stress disorder [2]. Before the pandemic, it was estimated that one in six children were diagnosed with a mental health disorder [3]. Since the pandemic, 14% of parents report worsened behavioral health in their children [4]. One-third of those children will only see their primary care provider for their mental health concerns [5]. Our aim is to examine the changing frequency of mental health diagnoses and antidepressant prescriptions in a general pediatrics clinic.
Methods
This study was approved by the institutional review board at the investigator's institution. We conducted a cross-sectional study at an ambulatory pediatric clinic located on a military installation. Data was collected from the electronic medical records of children before (June 2018 to June 2019) and during the pandemic (June 2020 to June 2021). The ages queried were 0–17 years old. To capture the prolonged impact of pandemic restrictions, the “during” cohort started approximately three months after the World Health Organization declared the COVID-19 pandemic [6]. In-person mental health encounters diagnosed with anxiety, depression, and other mood disorders as one of the top three diagnoses for the encounter were identified by using the International classification of diseases, 10th revision, codes. A separate review of the electronic medical records identified new SSRI prescriptions for the treatment of anxiety and/or depression (sertraline, fluoxetine, and escitalopram) [7]. Fluoxetine and escitalopram were chosen as they are the only Food and Drug Administration-approved medicines for children with depression to date. Sertraline is Food and Drug Administration-approved for children with Obsessive-Compulsive Disorder; however, it is also commonly used for anxiety and depression [8]. All SSRIs were new prescriptions prescribed on the same day as the appointment. Both in-person and virtual encounters were included, with most encounters being in-person. All results of quantitative variables were reported either as mean ± standard deviation or frequency (%). Chi-square analyses were conducted to investigate the level of association for the categorical variables of mental health diagnoses and new SSRI prescriptions. The prevalence ratios (PR) and 95% confidence intervals (CI) for the associated factors were calculated. A p-value less than .05 was considered statistically significant.
Results
Out of a total of 15,414 (9,791 prepandemic and 5,623 intrapandemic) encounters aged 0–17 years over the cross-sectional periods, 397 mental health encounters were identified (2.6%). Of the mental health encounters, 50 visits were between the ages of 4–10, and the vast majorities were among those aged 11–17 (347). Children between the ages of 11–17 saw an increase in mental health diagnoses of 164%, from 138 cases prepandemic to 209 cases during the pandemic (Table 1 ). Of total clinic encounters, 231 (4.1%) encounters for mental health occurred during the pandemic (vs. 1.7% prepandemic) and 63 of these 231 patients (27.3%) had SSRIs prescribed (vs. 5.4% prepandemic). The proportion of mental health encounters (PR 2.42, 95% CI, 1.99–2.95, p < .001) and SSRIs prescribed (PR 5.03, 95% CI, 2.58–9.82, p < .001) were higher during the pandemic. Similarly, patients prescribed SSRIs were more likely to be female and 13 years of age or older in both examined periods. Figure 1 illustrates the monthly mental health encounters and SSRI prescriptions. An increase in the rate of mental health diagnosis was noted in December 2020 and March 2021, with a statistically significant increase in the rate of SSRIs in March 2021.Table 1 Mental health encounters and prescribed SSRIs during and prepandemic
Characteristic Patients
June 2018 to June 2019 June 2020 to June 2021 X21 CV p
Demographics of patients with MH encounters
Age, mean (SD), y 13.5 (3.4) 14.3 (2.9)
Female patients, no. (%) 103 (62.1) 170 (73.6)
Female patient prescribed SSRI, no. (%) 8 (88.9) 48 (76.2)
Male patients, No. (%) 63 (37.9) 61 (26.4)
Male patient prescribed SSRI, no. (%) 1 (11.1) 15 (23.8)
Encounters for MH conditions, no. 166 231
Overall encounters, no. 9791 5623
MH Encounters (monthly)
Encounters for MH conditions, mean (SD), no. 12.8 (4.0) 17.8 (8.2) 82.9 3.8 <.001
Overall encounters, mean (SD), no. 753.8 (180.8) 432.5 (104.3)
SSRI Prescribing
Patient with MH condition prescribed SSRI, no. (%) 9 (5.4) 63 (27.3) 31.1 3.8 <.001
MH = mental health; SD = standard deviation; SSRIs = selective serotonin reuptake inhibitors.
Figure 1 Mental health encounters and SSRIs prescribed during and prepandemic.
Discussion
Primary care clinics are often the first point of care for children's mental health issues. Compared to prepandemic, the proportion of mental health encounters was more than two-fold higher during the pandemic. Correspondingly, the proportion of prescribed SSRIs during the pandemic was five times higher despite overall decreases in overall face-to-face visits. Majority of those impacted were female and 13 years old or greater. Our study suggests that social distancing and other efforts to mitigate COVID-19 infection are associated with a drastic increase in both the diagnosis and pharmacologic treatment of new mental health diagnoses. Potential contributory factors include increased social isolation, decreased time to bond and connect with peers, as well as an upheaval of daily home routines and structure. An increase in mental health diagnoses and SSRI prescribing in December 2020 could be attributed to restrictions during the holiday season and onset of decreased sunlight in the winter season. The observed springtime spike may be due to travel and vacation cancellations secondary to travel restrictions. The escalated rates of these mental health diagnoses and SSRI prescriptions highlight not only the rise of mental health conditions in the pediatric population, but also the increased need for medical management of these conditions. Awareness of the rising need for treatment of these conditions should prompt pediatric clinics to be equipped with the screening tools and resources to properly diagnose and treat patients. Screening tools like the Patient Health Questionnaire-9 and the Generalized Anxiety Disorder-7 should be utilized regularly.
The study was limited by decreased in-person care during the pandemic, inability to qualify illness severity, and lack of generalizability of data outside the military general pediatric clinic. Children of military families face circumstances that contribute to health care access challenges. The frequent moves inherent to military service can disrupt physician-patient relationships. These relocations typically move children away from extended familial support, and the isolation can be further exacerbated by the stresses of deployments.
Primary care providers provide longitudinal care and are uniquely postured to address mental health concerns in the medical home. Future research should focus on the increasing utilization of primary care for pediatric mental health disorders and its outcomes. Our findings underscore the critical need for additional mental health resources integrated within the primary care clinic.
Acknowledgments
The authors would like to acknowledge Dr Jill Sylvester, M.D. for her critical revisions of the manuscript and unfailing patience in answering numerous questions; Dr Cara Olsen, MS, DrPH., for providing analytical oversight; Mr. David Brunner, Medical Database Manager, for timely acquisition of data; and our mentor, Karri Roman, for her encouragement and research support.
Conflicts of interest: The authors have no conflicts of interest to disclose.
This manuscript has not been published elsewhere and is not under consideration by another journal.
==== Refs
References
1 Marques de Miranda D. da Silva Athanasio B. Sena Oliveira A.C. Simoes-e-Silva A.C. How is COVID-19 pandemic impacting mental health of children and adolescents? Int J Disaster Risk Reduct 51 2020 101845 32929399
2 Meherali S. Punjani N. Louie-Poon S. Mental health of children and adolescents amidst COVID-19 and past pandemics: A rapid systematic review Int J Environ Res Public Health 18 2021 3432 33810225
3 Data and statistics on children's mental health. Centers for Disease Control and Prevention Available at: https://www.cdc.gov/childrensmentalhealth/data.html
4 Patrick S. Henkhause L. Zickafoose J. Well-being of parents and children during the COVID-19 pandemic: A National Survey Pediatrics 146 2020 e2020016824
5 Anderson L. Chen M. Perrin J. Van Cleave J. Outpatient visits and medication prescribing for US children with mental health conditions Pediatrics 136 2015 e1178 e1185 26459647
6 WHO Director-General's opening remarks at the media briefing on COVID-19 - 11 March 2020. World Health Organization Available at: https://www.who.int/director- general/speeches/detail/who-director-general-s-opening-remarks-at-the-media-briefing- on-covid-19---11-march-2020
7 Strawn J.R. Dobson E.T. Giles L.L. Primary pediatric care psychopharmacology: Focus on medications for ADHD, depression, and anxiety Curr Probl Pediatr Adolesc Health Care 47 2017 3 14 28043839
8 Depression medicines. U.S. Food and Drug Administration Available at: https://www.fda.gov/consumers/free-publications-women/depressionmedicines#:∼:text=Prozac%20(fluoxetine)%20is%20the%20only,medicines%20for%20teens%20with%20depression
|
PMC010xxxxxx/PMC10149042.txt |
==== Front
Diabetologie
Die Diabetologie
2731-7447
2731-7455
Springer Medizin Heidelberg
1051
10.1007/s11428-023-01051-2
Journal Club
Wirkung von Verapamil auf die Betazellfunktion des Pankreas bei neu diagnostiziertem Typ-1-Diabetes im Kindesalter
Eine randomisierte klinische Studie
Effect of verapamil on pancreatic beta cell function in newly diagnosed type 1 diabetes in childrenA randomized clinical trial
Jecht Michael mjecht@t-online.de
Diabetesschwerpunktpraxis, Rodensteinstr. 32, 13593 Berlin, Deutschland
30 4 2023
2023
19 4 528530
27 3 2023
© The Author(s), under exclusive licence to Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
issue-copyright-statement© Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2023
==== Body
pmcOriginalpublikation
Forlenza GP, McVean J, Beck RW et al for the CLVer Study Group (2023) Effect of verapamil on pancreatic beta cell function in newly diagnosed pediatric type 1 diabetes. A randomized clinical trial. JAMA. 10.1001/jama.2023.2064 (Published online February 24).
Ziel.
In präklinischen Studien induziert die Überexpression des thioredoxininteragierenden Proteins die Apoptose von Betazellen des Pankreas und ist am glukotoxisch induzierten Betazelltod beteiligt. Kalziumkanalblocker reduzieren diese Effekte und könnten für den Erhalt der Betazellen bei Typ-1-Diabetes von Vorteil sein.
Ziel dieser Arbeit war es daher, die Wirkung von Verapamil auf die Betazellfunktion des Pankreas bei Kindern und Jugendlichen mit neu diagnostiziertem Typ-1-Diabetes zu untersuchen.
Methodik.
Diese doppelblinde, randomisierte klinische Studie mit Kindern und Jugendlichen im Alter von 7–17 Jahren mit neu diagnostiziertem Typ-1-Diabetes und einem Gewicht von mindestens 30 kg wurde an 6 Zentren in den USA durchgeführt (die Teilnehmer wurden zwischen dem 20.07.2020 und dem 13.10.2021 randomisiert); die Nachbeobachtung wurde am 15.09.2022 abgeschlossen.
Die Teilnehmer wurden im Verhältnis 1:1 auf 1‑mal täglich oral einzunehmendes Verapamil (n = 47) oder Placebo (n = 41) in einem faktoriellen Design randomisiert, wobei die Teilnehmer auch entweder einer intensiven Diabetesbehandlung oder einer Standarddiabetesbehandlung zugeteilt wurden.
Ergebnisse.
Von den 88 Teilnehmern beendeten 83 (94 %) die Studie (mittleres Alter 12,7 [SD [Standardabweichung] 2,4] Jahre; 36 waren weiblich [41 %]; und die mittlere Zeit zwischen Diagnose und Randomisierung betrug 24 Tage [SD = 4]). In der Verapamilgruppe betrug die mittlere C‑Peptid-Fläche unter der Kurve zu Studienbeginn 0,66 pmol/ml und nach 52 Wochen 0,65 pmol/ml, verglichen mit 0,60 pmol/ml zu Studienbeginn und 0,44 pmol/ml nach 52 Wochen in der Placebogruppe (adjustierter Unterschied zwischen den Gruppen 0,14 pmol/ml [95 %-KI: 0,01–0,27 pmol/ml]; p = 0,04 [KI: Konfidenzintervall]).
Dies entspricht einem um 30 % höheren C‑Peptid-Spiegel nach 52 Wochen unter Verapamil. Der Anteil der Teilnehmer, die nach 52 Wochen einen C‑Peptid-Spitzenwert von 0,2 pmol/ml oder mehr aufwiesen, betrug 95 % (41 von 43 Teilnehmern) in der Verapamilgruppe gegenüber 71 % (27 von 38 Teilnehmern) in der Placebogruppe. Nach 52 Wochen betrug der Hämoglobin‑A1c-Wert in der Verapamilgruppe 6,6 % gegenüber 6,9 % in der Placebogruppe (korrigierter Unterschied zwischen den Gruppen: −0,3 % [95 %-KI: −1,0 %–0,4 %]). Sowohl die 8 Teilnehmer (17 %) in der Verapamilgruppe als auch die 8 Teilnehmer (20 %) in der Placebogruppe hatten ein nicht schwerwiegendes unerwünschtes Ereignis, das als behandlungsbedingt eingestuft wurde.
Schlussfolgerung.
Bei Kindern und Jugendlichen mit neu diagnostiziertem Typ-1-Diabetes konnte Verapamil im Vergleich zu Placebo die stimulierte C‑Peptid-Sekretion 52 Wochen nach der Diagnose teilweise aufrechterhalten. Weitere Studien sind notwendig, um die Dauer der C‑Peptid-Steigerung und die optimale Therapiedauer zu bestimmen.
Kommentar
Typ-1-Diabetes ist bekanntlich eine Autoimmunerkrankung, die zur Zerstörung der insulinproduzierenden Betazellen der Bauchspeicheldrüse führt.
Nach jahrzehntelanger Forschung konnte gezeigt werden, dass eine begrenzte, aber zunehmende Zahl immunmodulatorischer Substanzen die Betazellfunktion bei neu aufgetretenem Typ-1-Diabetes erhalten kann. Da Immunsuppression und Immunmodulation bei der Erhaltung der Betazellfunktion nur teilweise erfolgreich waren, stieg das Interesse an einem Schutz der Betazellen durch andere Mechanismen.
Der Erhalt selbst einer bescheidenen Restfunktion der Betazellen, gemessen an der stimulierten C‑Peptid-Sekretion, ist ein erstrebenswertes Ziel, das mit einem geringeren Risiko für diabetesbedingte vaskuläre Komplikationen und Hypoglykämien assoziiert ist [1, 2].
Einige wenige Medikamente zeigten in ausreichend aussagekräftigen randomisierten klinischen Studien ihre Wirksamkeit zur Aufrechterhaltung der C‑Peptid-Sekretion bei neu diagnostizierten Patienten mit Typ-1-Diabetes [3–5].
Teplizumab wurde kürzlich für die Behandlung von Typ-1-Diabetes im Stadium 2 (positive diabetesassoziierte Autoantikörper mit Nachweis einer Dysglykämie) zugelassen, jedoch nicht für die Behandlung von Typ-1-Diabetes im Stadium 3 (neu aufgetretener, klinisch diagnostizierter Diabetes). Teplizumab, ein monoklonaler Antikörper, der auf den Zelloberflächenmarker CD3T abzielt, ist das erste und einzige von der US-Arzneimittelbehörde FDA (U.S. Food and Drug Administration) zugelassene Medikament, das das Fortschreiten des klinischen Typ-1-Diabetes und die Notwendigkeit einer Insulintherapie verlangsamt.
In den letzten Jahren wurde eine wachsende Zahl von Stoffwechselwegen der Betazellen mit der Pathophysiologie des Typ-1-Diabetes in Verbindung gebracht [4]. Es konnte gezeigt werden, dass die Überexpression des thioredoxininteragierenden Proteins die Apoptose von Betazellen des Pankreas induziert und in Kultur- und Mausmodellen am glukotoxisch induzierten Betazelltod beteiligt ist [6].
Kalziumkanalblocker wie Verapamil reduzieren die Expression des thioredoxininteragierenden Proteins und die Apoptose von Betazellen [7–9] und könnten für die Erhaltung von Betazellen nach der Diagnose von Typ-1-Diabetes von Vorteil sein.
Im Jahr 2018 wurde eine placebokontrollierte, randomisierte klinische Pilot- und Machbarkeitsstudie veröffentlicht [10], die bei 11 Erwachsenen mit neu diagnostiziertem Typ-1-Diabetes, die mit oralem Verapamil behandelt wurden, nach 12 Monaten einen relativen Anstieg der stimulierten C‑Peptid-Spiegel um 35 % im Vergleich zu 13 Erwachsenen, die Placebo erhielten, zeigte.
In der vorliegenden Arbeit führten die Autoren eine doppelblinde, placebokontrollierte, randomisierte klinische Studie an Kindern und Jugendlichen im Alter von 7–17 Jahren mit neu diagnostiziertem Typ-1-Diabetes im Stadium 3 (klinisch manifest) durch, um die Sicherheit und Wirksamkeit von Verapamil zur Erhaltung der Betazellfunktion 12 Monate nach der Diagnose zu untersuchen.
Eine anfängliche Verbesserung der stimulierten C‑Peptid-Sekretion wurde in beiden Gruppen beobachtet, wobei die Verapamilgruppe im Vergleich zur Placebogruppe eine längere Stabilitätsperiode aufwies, bevor die C‑Peptid-Spiegel zu sinken begannen. Angesichts des günstigen Sicherheitsprofils im Vergleich zu Immunsuppressiva, der 1‑mal täglichen oralen Verabreichung und der geringen Kosten könnte die Einleitung einer Verapamiltherapie bei Patienten mit neu diagnostiziertem Typ-1-Diabetes in Erwägung gezogen werden.
Signifikante Unterschiede in der C‑Peptid-Sekretion waren nicht mit signifikanten Unterschieden zwischen den Behandlungsgruppen in Bezug auf den HbA1c(Hämoglobin‑A1c)-Wert, die Parameter der kontinuierlichen Glukoseüberwachung oder die Insulindosis verbunden. Die glykämische Kontrolle in der Placebogruppe war sehr gut, was möglicherweise auf die Verwendung eines automatischen Insulinverabreichungssystems oder eines kontinuierlichen Glukosemonitorings für das Diabetesmanagement sowie auf das Vorhandensein einer sog. Honeymoon-Phase im ersten Jahr nach der Diagnose von Typ-1-Diabetes zurückzuführen ist, in der bei den meisten Patienten eine gewisse Restfunktion der Betazellen vorhanden ist.
Zu den Stärken dieser Studie gehören das multizentrische, doppelblinde, randomisierte Design, die Untersuchung eines Medikaments mit einem Wirkmechanismus, der sich von Immuntherapien unterscheidet, die Verwendung eines Medikaments mit einem früheren Sicherheitsprofil in anderen Populationen, die Einbeziehung einer auf Kinder und Jugendliche beschränkten Kohorte, die ethnische Vielfalt der Teilnehmer und der Beginn der Behandlung innerhalb von 31 Tagen nach der Diagnose von Typ-1-Diabetes.
Folgende Einschränkungen sollten aber bedacht werden. Erstens wurden aufgrund der im Handel erhältlichen Dosierungsoptionen für Verapamil mit verzögerter Wirkstofffreisetzung nur Kinder mit einem Gewicht von 30 kg oder mehr in die Studie aufgenommen. Zweitens wurde das ursprüngliche Rekrutierungsziel aufgrund von Schwierigkeiten bei der Rekrutierung von Teilnehmern während der COVID-19-Pandemie Anfang 2020 nicht erreicht. Drittens war die Stichprobengröße für die Bewertung der Nebenwirkungen der Behandlung zu klein.
Zusammenfassend sind die Ergebnisse sehr ermutigend und werden hoffentlich in größeren Studien bestätigt werden.
Fazit für die Praxis
Bewahrt 1‑mal täglich oral eingenommenes Verapamil die Betazellfunktion der Bauchspeicheldrüse bei Kindern und Jugendlichen mit neu diagnostiziertem Typ-1-Diabetes?
In einer randomisierten klinischen Studie mit 88 Kindern und Jugendlichen mit neu diagnostiziertem Typ-1-Diabetes waren die C‑Peptid-Werte (ein Maß für die Betazellfunktion der Bauchspeicheldrüse), die 52 Wochen nach der Diagnose während eines Toleranztests mit gemischten Mahlzeiten gemessen wurden, unter Verapamil um 30 % höher als unter Placebo.
Der Prozentsatz der Teilnehmer, die nach 52 Wochen einen C‑Peptid-Spitzenwert von 0,2 pmol/ml oder mehr aufwiesen, betrug in der Verapamilgruppe 95 % gegenüber 71 % in der Placebogruppe.
Verapamil hatte nur wenige unerwünschte Ereignisse.
Interessenkonflikt
M. Jecht gibt an, dass kein Interessenkonflikt besteht.
QR-Code scannen & Beitrag online lesen
==== Refs
Literatur
1. Bonfanti R Bazzigaluppi E Calori G Parameters associated with residual insulin secretion during the first year of disease in children and adolescents with type 1 diabetes mellitus Diabet Med 1998 15 10 844 850 10.1002/(SICI)1096-9136(199810)15:10<844::AID-DIA679>3.0.CO;2-A 9796885
2. Sibley S Steffes MW Jackson M, ThomasW. Beta-cell function and the development of diabetes-related complications in the diabetes control and complications trial Diabetes Care 2003 26 3 832 836 10.2337/diacare.26.3.832 12610045
3. Jacobsen LM Bundy BN Greco MN Comparing beta cell preservation across clinical trials in recent-onset type 1 diabetes Diabetes Technol Ther 2020 22 12 948 953 10.1089/dia.2020.0305 32833543
4. Quattrin T Haller MJ Steck AK T1GER Study Investigators. Golimumab and beta-cell function in youth with new-onset type 1 diabetes N Engl J Med 2020 383 21 2007 2017 10.1056/NEJMoa2006136 33207093
5. Martin S Residual beta cell function in newly diagnosed type 1 diabetes after treatment with atorvastatin/ the Randomized DIATOR Trial Plos One 2011 6 3 e17554 10.1371/journal.pone.0017554 21412424
6. Chen J Saxena G Mungrue IN Lusis AJ Shalev A Thioredoxin-interacting protein: a critical link between glucose toxicity and beta-cell apoptosis Diabetes 2008 57 4 938 944 10.2337/db07-0715 18171713
7. Chen J Cha-Molstad H Szabo A Shalev A Diabetes induces and calcium channel blockers prevent cardiac expression of proapoptotic thioredoxin-interacting protein Am J Physiol Endocrinol Metab 2009 296 5 E1133 E1139 10.1152/ajpendo.90944.2008 19258488
8. Xu G Chen J Jing G Shalev A Preventing β-cell loss and diabetes with calcium channel blockers Diabetes 2012 61 4 848 856 10.2337/db11-0955 22442301
9. Borowiec AM Właszczuk A Olakowska E Lewin-Kowalik J TXNIP inhibition in the treatment of diabetes: verapamil as a novel therapeutic modality in diabetic patients Med Pharm Rep 2022 95 3 243 250 36060506
10. Ovalle F Grimes T Xu G Verapamil and beta cell function in adults with recent-onset type 1 diabetes Nat Med 2018 24 8 1108 1112 10.1038/s41591-018-0089-4 29988125
|
PMC010xxxxxx/PMC10150213.txt |
==== Front
Int J Educ Dev
Int J Educ Dev
International Journal of Educational Development
0738-0593
0738-0593
The Author(s). Published by Elsevier Ltd.
S0738-0593(23)00064-0
10.1016/j.ijedudev.2023.102788
102788
Article
Learning inequality during Covid-19: Evidence from secondary schools in Colombia☆
Marín Llanes Lucas a⁎
Rodríguez Pico Mariana b
Maldonado Darío c
García Sandra c
a Center for Security and Drugs Studies, School of Economics, Universidad de Los Andes, Colombia
b IPA, Colombia
c School of Government, Universidad de Los Andes, Colombia
⁎ Corresponding author.
1 5 2023
7 2023
1 5 2023
100 102788102788
9 8 2022
2 3 2023
27 4 2023
© 2023 The Authors
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
During 2020, the Covid-19 pandemic unleashed a socioeconomic crisis in most countries, as well as strict restrictions on mobility and social distancing were implemented. The pandemic brought a severe socioeconomic shock -decreasing economic activity- and forced policy responses that affected the education sector -notably school closures-. There is scarce evidence on the implications of the pandemic and its socioeconomic consequences on learning inequality, particularly in the Latin-American context. The aim of this paper is to measure the change in learning inequality during the years of the pandemic (2020–2021) in Colombia. To measure learning inequality, we use the results of a standardize exam taken by all upper secondary education graduates in the country. To capture inequality, we use secondary-level students’ characteristics, their households, and schools’ characteristics. Our econometric results suggest an increase in learning inequality between 48% and 372% depending on the dimension studied, except for gender where learning inequality decreased. Moreover, through dynamic specifications, we determine that for all the analyzed dimensions the 2020–2021 period represents a change in the trend of learning inequality as in the periods prior to the pandemic inequality gaps decreased or, at least, remained constant. We conclude with concrete and immediate policy recommendations to target the learning processes of vulnerable students and bridge learning gaps.
Keywords
Standardized test scores
Education
Covid-19
Inequality
Colombia
==== Body
pmc1 Introduction
The first cases of Covid-19 in Colombia were reported in early March 2020. Shortly, the authorities decided to begin a preventive isolation policy, and declared it mandatory as from March 25th, 2020. The mandatory preventive isolation plan included a set of health measures and major changes at the economic and social levels in order to control the spread of Covid-19. Progressively and by phases, certain industries and sectors were able to reopen with the purpose of mitigating the economic crisis. Until the 1st of September 2020 the government began a pilot authorizing few economic sectors to reopen with any mobility restrictions. Lockdowns stopped the production and economic activity of various sectors, and significantly decreased the national output: the Colombian GDP decreased 7.3% in 2020, particularly during the second quarter (April-June) it decreased 15.7% (DANE, 2023). The magnitude of the impacts of the pandemic were disproportionately large for vulnerable population groups (DANE, 2023).
The preventive isolation policy included school closures. In Latin America, school closures lasted 70% longer than in OECD1 countries (OECD et al., 2021), despite the lack of evidence that closures were indeed an effective health measure (Heavey et al., 2020, Rajmil, 2020). In Colombia in particular, most of the schools remained closed during most of 2021, despite the fact that schools could open safely and that many families considered the opening necessary (García & Maldonado, 2021). Furthermore, the return to in person classes took place with significant differences across municipalities in timing and number of students benefiting from reopening.2 This is especially problematic in the Colombian context, like other Latin American countries, because before the pandemic a high proportion of students had learning gaps, as well as significant learning inequalities related to the socioeconomic characteristics of the students.
In this context of uneven affectations during Covid-19 and the implemented policies, in this paper we explore the changes in learning inequalities in Colombia during the years of the pandemic. The costs in the education attainment of children and adolescents caused by the pandemic have been unevenly distributed. The United Nations (2020) argued that during the pandemic there was a decrease in the educational opportunities for the most vulnerable students, such as for those living in poverty, those living in rural areas, women, displaced persons, among others. Similarly, households with limited access to electronic devices faced difficulties in supporting children and adolescents in their learning process (Valle, 2020). The learning loss for students with lower performance prior to the pandemic is greater than for the rest of the students (Angrist et al., 2021). In the United States, after three months of school closures, it was found that low-income students reduced their scores and performance by 36%, while high-income students increased their scores by 45.5% in the same period (Opportunity Insights Economic Tracker, 2020). Uneven learning loss will have long-term consequences for human capital accumulation (Azevedo et al., 2020). Thus, it is necessary to measure student’s learning outcomes and assess whether or not inequality gaps increased during the pandemic in Colombia. The latter contributes to the design of mechanisms to target the learning process of vulnerable students and mitigate the long-term consequences mentioned above.
In this paper we measure how learning inequality changed in the period 2016–2021 to determine whether inequality is statistically different between the pre-pandemic years (2016–2019) and the pandemic years (2020–2021). We use data from the National Standardized Scholastic Aptitude Test for high school seniors in Colombia, called Saber 11, which is a compulsory test for all secondary education graduates in Colombia. In most countries standardized tests were not implemented during the pandemic; while Colombia managed to implement the test despite the difficulties implied by the health shock. We use four dimensions to quantify these gaps in educational learning: individual (gender and ethnicity), household (socioeconomic status), schools (sector and area) and municipality (a categorization of municipalities according to permanent violence exposure and poverty rates -- Programas de desarrollo territorial, PDET).3 Using econometric models, controlling for observable factors of students and their environments, and for time-invariant unobservable characteristics of schools, we estimate the change in learning inequality gaps associated with each dimension and with the years of the pandemic (2020–2021). Moreover, through dynamic specifications, we determine that for all the analyzed dimensions the 2020–2021 period presents a change in the trend of learning inequality as in the periods prior to the pandemic inequality gaps decreased or, at least, remained constant. These empirical findings push forward our argument on the widening of inequality gaps specifically during the years of the pandemic. We find that gender inequality decreased by 17.9% during the Covid-19 period. Conversely, for all other characteristics of interest, inequality increased dramatically. The estimate percentage changes range from 48% (household socioeconomic status) to 372% (school area) inequality increases.
This paper contributes to the literature by documenting how the pandemic and all its socioeconomic consequences affected education processes. Evidence prior to the Covid-19 pandemic documented the effects of school closures due to holidays, teacher strikes, wars, natural disasters, or four-day school week showing negative and persistent effects on student performance (Andrabi et al., 2020, Busso and Camacho, 2020, Thompson and Ward, 2022). In the context of this pandemic, it has been estimated that closing schools for one-third of the school year reduces student learning by one whole year and reduces enrollment rates (Angrist et al., 2021, Chatterji and Li, 2021, Engzell et al., 2020, Kaffenberger, 2021, Maldonado and De Witte, 2020). In developing countries, such as Mexico and South Africa, the learning loss caused by school closures has also been quantified (Ardington et al., 2021, Hevia et al., 2022). These studies show that in absence of remedial interventions, the learning loss will accumulate because students did not acquire the knowledge expected by the training programs, and, without adjustments, students will face substantial knowledge gaps. The resulting long-term costs are of unprecedented magnitude. Estimates of the economic impact of pandemic closures, smaller in magnitude than Colombia's, show long-term economic losses up to $10 billion (Azevedo et al., 2021).
This paper also contributes to the literature by measuring how unevenly distributed were the educational costs of the pandemic, affecting disproportionately the most vulnerable students. In other countries, there is evidence of the drop of standardized test scores, reductions in enrollment rates and their unequal distribution in vulnerable students (Chatterji and Li, 2021, Engzell et al., 2020, Maldonado and De Witte, 2020, Unicef, 2021a). Alban Conto et al. (2021) find that remote learning is especially inadequate for the most vulnerable and youngest students. In that sense, Hossain (2021) suggests that the most vulnerable students are less likely to access distance education services. Therefore, learning costs and inequality gaps are expected to have a greater impact among the most vulnerable children and adolescents. In Latin America, through simulations, it has been found that the pandemic widened learning gaps (Neidhöfer et al., 2021). In this paper, we contribute to validate empirically these simulations by estimating the change in learning inequality in Colombia based on observed educational outcomes.
In the second section of this paper, we present the evolution of socioeconomic outcomes and the education sector in Colombia during the years of the pandemic. The third section presents the data and its sources. In the fourth section, we present the quantitative methodology. The fifth section presents the results. Finally, we conclude and discuss policy recommendations to target vulnerable student’s learning processes to bridge learning inequality gaps.
2 Colombian context
Colombian population was deeply affected by the pandemic. First, the fall in economic activity was very pronounced since GDP felt by 7.3% in 2020. But this happened in a context where most workers do not have access to any protection since, before the pandemic, 63.6% of the labor market in Colombia worked informally (Alvarado et al., 2021; Fernández and Eslava, 2022). This not only implies that in the face of the pandemic many families had low insurance to protect them from the associated economic shock but also that restrictions on mobility (due to social isolation policies) highly affected the income of most of the Colombian households. All this reflected in higher unemployment and poverty. The national unemployment rate increased from 10.9% in 2019 to 16.5% in 2020. National monetary poverty increased from 35.7% in 2019 to 42.5% in 2020, extreme monetary poverty increased from 9.6% to 15.1% in the same period, and multidimensional poverty increased from 17.5% in 2019 to 18.1% in 2020. Furthermore, when analyzed by area, multidimensional poverty increased in rural areas from 34.5% to 37.1% (DANE, 2023). Lastly, the economic shock caused by the pandemic amplified existing inequalities evidenced by the increase in the Gini Index from 50.5 in 2019–53.7 in 2020 (DANE, 2022). For most of these socioeconomic indicators, Colombia went back to the levels of the beginning of the XXIst century.
At the same time, the education sector was one of the most affected by the pandemic since the social isolation policies implied prolonged school closures. In most of Latin-American countries school closures lasted more than 18 months (Unicef, 2021b). Colombia had closures that lasted for up to 21 months. This policy decision affected more than 10 million students in the country. In 2020 most of primary and secondary students attended in-person classes until mid-March when school closures were enforced. In 2021, the National Government started a progressive reopening of schools. However, there were territorial differences in the country and even until September 2021, 18 months after the first school closures, 30% of students remained in remote learning (Observatorio de la Gestión Educativa, sf). The first pilots to reopen schools started on October 19th, but this covered very few schools and in very few territorial entities.4
As a consequence, very few students had in-presence education in the second semester of 2020 and in the first of 2021. Employing a nationwide representative sample of schools, García et al. (2021) show that 62.3% of secondary-level students were attending schools that did not offer any in-person education during the first semester of 2021 and only 16% of students attended in-person lessons. The other students belong to a school that reopened during the first semester of 2021, but they decided not to attend, or the school did not offer in person education for their level or grades.
Online education in many schools was far from ideal. García et al. (2021) found that 57% of secondary-level students had access to education platforms and 61% to live lessons with a teacher. Furthermore, 17% of secondary students did not have any contact with their teachers within the previous week. Moreover, there are important gaps between rural and urban schools that could explain the widening on learning inequality gaps. For example, 54% of students in urban areas had access to education platforms, while this percentage for rural students corresponded to 32%. Hence, this brief characterization of the education sector in Colombia puts forward its limitations during the pandemic and potentially explains the results of our study.
Therefore, the combination of the educational crisis explained by restrictive policies and the socioeconomic downturn in the Colombian context could explain a widening in inequality learning gaps. On the one hand, the socioeconomic crisis accompanied by higher unemployment and higher poverty, among others, affected the economically vulnerable students. Not only the usual link -a reduction in the opportunities to learn because of a lack of economic resources- between poverty and academic success is in play here. Poverty also reduces cognitive capacity because poverty-related concerns consume mental resources, leaving less resources available to perform tasks and thus negatively affecting the student’s learning processes (Mani et al., 2013, Banerjee and Duflo, 2011). Additionally, preoccupations on household’s income are distractors for children and adolescents because scarcity changes how people allocate attention and how they look at problems and make decisions. Scarcity, as put by Shah et al. (2012), demands greater focus in basic needs leading to neglect other problems such as learning performance. The consequences of the economic downturns caused by the pandemic, such as the one we aim to demonstrate in this paper, disproportionately affect vulnerable students by coping their cognitive resources and reducing their cognitive capability and attention to learn. On the other, school closures before and during the pandemic affected students’ performance, especially for the most vulnerable groups (Andrabi et al., 2020, Angrist et al., 2021, Busso and Camacho, 2020, Chatterji and Li, 2021, Engzell et al., 2020, Kaffenberger, 2021, Maldonado and De Witte, 2020, Opportunity Insights Economic Tracker, 2020, Thompson and Ward, 2022).
3 Data
This study uses data from Saber 11 test, a Colombian standardized aptitude test designed to measure learning at the end of secondary education. Before the pandemic it was used by most higher education institutions to decide whether to accept students’ applications. Although the test result does not establish whether or not a student obtains a high school diploma, taking the test is mandatory by law and no higher education institution can accept a student that graduated from a Colombian high school that did not take the test. The Colombian Institute for the Evaluation of Education (ICFES by its Spanish acronym) is the entity in charge of designing and managing the test and has information on the results of educational achievement in the areas of language, mathematics, natural sciences, citizenship skills and English, as well as an overall score. The database with the results also provides sociodemographic information of students and their households. In this article, we restrict the sample to students in the 15–21 age range at the time of taking the test.
The database is a cross-section, repeated annually at the school level, and has 2,537,477 observations for the period 2016–2021. For each student, the database has information on test results, on individual and household characteristics, on the school where the student completed high school, and on the municipality in which the student resides. Using school identifiers, we merge data from the Ministry of Education on the type of school, and enrollment in tenth and eleventh grades.
The municipality data comes from different sources. This information includes the Multidimensional Poverty Index (MPI), the rurality index and the distance to the capital, and a classification of municipalities according to poverty rates and historic exposure to violence.5 We include the presence of armed groups with a variable taking the value of 1 if the municipality had any armed attack between 2016 and 2019 and 0 if it did not. Finally, we include the variable of night luminosity measured through satellite images to have an estimate of the level of economic activity at the municipality level (Li and Zhou, 2017, Li et al., 2020; Sullivan III, 1989; Yang et al., 2019).
To measure the relationship between the shock generated by the pandemic in 2020–2021 and educational inequality, we conduct an econometric analysis in which the outcome variable is the overall percentile on the Saber 11 test for each student. The percentile is a numerical variable that takes values between 1 and 100 and represents the relative position of an individual's test score in the distribution of the cohort taking the exam in that same semester. We use this variable because it allows comparisons of results between years; it enables us to exploit the relative position of each student's performance, which is related to the inequality dimension we explore in this article.
Table 1 shows descriptive statistics that allow us to characterize the students who took the test between 2016 and 2019 and those who took it in 2020 and 2021. Although differences between students in these two periods are statistically significant, the magnitudes are small except for internet access. The proportion of students without internet access decreased from 40.4% between 2016 and 2019– to 26.7% in 2020–2021. The latter can be explained by the progressive availability of internet access between 2016 and 2019, and by the challenges brought by the pandemic that forced households to obtain internet connectivity services.6 Moreover, the government implemented a policy to subsidize internet access for 500,000 households from low socioeconomic levels, which may explain part of the observed increase (MinTIC, 2021).7 Table 1 Mean differences between students before and during Covid-19.
Table 1 2016–2019
(N = 1714,201)
Mean 2020–2021 (Covid)
(N = 825, 604)
Mean Mean differences
(p-value)
Women share 0.549 0.552 -0.003 ***
(0.001)
Ethnic population 0.065 0.062 0.003 ***
(0.001)
Low socioeconomic status 0.526 0.500 0.03 ***
(0.001)
No Internet 0.404 0.267 0.137 ***
(0.001)
No computer 0.388 0.384 0.004 ***
(0.001)
Public educational institution 0.780 0.796 -0.016 ***
(0.001)
Rural educational institution 0.155 0.166 -0.010 ***
(0.000)
PDET 0.119 0.123 -0.004 ***
(0.000)
*** p < 0.01, ** p < 0.05, * p < 0.1
There were differences in Saber11 percentiles between groups prior to 2020 and, aside from gender and PDET students, these gross differences increased during the pandemic years (2020–2021). Appendix 1, Table A1 presents the change in percentile mean differences before and during Covid-19 for each analyzed characteristic. We present this descriptive evidence to put forward the intuition that the pandemic increased pre-existing learning gaps and the magnitude of the gross differences between groups.
Regarding the individual characteristics, the data shows that there is a gender gap in Saber 11 test scores. On average, males rank at a higher percentile than females. However, the inequality between men and women decreased from 4.6 percentiles before to Covid-193.9 percentiles in 2020–2021. In other words, there was a drop of 15.2%. Before Covid-19, there was a gap in Saber11 scores of 20.7 percentiles between the non-ethnic population and the ethnic population. This gap increased to 21.1 percentiles during Covid-19, representing a 1.9% increase. In relation to the household dimension, the mean difference in Saber 11 scores between students from high and low socioeconomic status increased from 19.1 pre-Covid to 19.9 during Covid-19, suggesting a 4.2% increase.
Standardized test inequality between public and private school students was 19.2 percentiles prior to Covid-19 and increased to 21.2 percentiles during Covid-19. This change represents an increase of 10.4%. Moreover, the inequality in Saber11 between urban and rural students was 15.2 percentiles before Covid-19 and increased to 15.8 percentiles in 2020–2021, representing an increase in inequality of 3.9%. When analyzing the average difference in Saber11 scores between PDET and non-PDET municipalities, we find that the gap before Covid-19 was 14.1 percentiles and remained unchanged in 2020–2021.
Hence, for all explored characteristics, except for gender and PDET, the gross gaps on learning outcomes increased during the pandemic. However, the changes presented in this section may be correlated with other determinants of inequality. Therefore, we propose an econometric strategy to analyze the inequality associated with each variable controlling for other students’ characteristics and their social context.
4 Methodology
In all of our analyses the outcome variable is the percentile in which the student fell on the Saber 11 test in his/her cohort. The statistical analysis method enables us to determine the association between the variables of interest and Saber11 percentiles before (2016–2019) and during (2020–2021) the Covid-19 pandemic.
The variables of interest have been traditionally considered relevant in the educational process in Colombia and have been used in previous studies of educational gaps. Our approach focuses on four dimensions: individual, household, educational institution, and municipality. The individual dimension consists of gender and ethnicity. The household dimension consists of a socioeconomic categorical variable that classifies each student by its corresponding household socioeconomic status (SES) into four different categories (low, middle-low, middle-high and high SES). The categories are built according to a set of variables including assets at home (washing machine, internet, computer, oven, tv, car, and videogames console) and the mother’s education. To be consistent with the dichotomous definition of the rest of the variables of interests, we group low and middle-low SES students, and middle-high and high SES students in two categories. For the educational institution level, we use the information on the school where the student finished high school including the sector (public or private), and the area (rural or urban). The municipality dimension considers whether or not the student resided in a municipality with historical high poverty rates and exposed to violence (PDET). All these characteristics correspond to dichotomous variables; they are equal to 1 if the student belongs to the group with higher historic educational outcomes (men, non-ethnic, high SES, private school, urban, non-PDET).
To estimate the effect of Covid-19 on student learning, we would require an experimental or quasi-experimental data structure. This type of data is generally not feasible considering that the shock generated by the pandemic affected all students and, that in Colombia, there are no longitudinal data on students’ performance. Nonetheless, through statistical methods it is possible to analyze the change in students’ relative performance conditioning to their individual and environmental characteristics.
To analyze the learning gaps based on the observable characteristics of interest, we estimate econometric models for each interest variable. To explore the change in inequality for each variable, we estimate a model interacting the variable of interest with a Covid-19 fixed effect. This allows us to estimate the change in educational inequality on our variable of interest associated with the pandemic years, net of the contribution of other observable and unobservable variables. The model is formally specified as follows:(1) yi,s,t=δ1Zg,t|g=i,s˜+δ2μcovid−19*Zg,t|g=i,s˜+θZg,t|g=i,s+βXi,t+μp+μs+εi,s,t
where yi,s,t corresponds to the percentile occupied by the student i who attended school s and took the Saber-11 test in year t. Zg,t|g=i,s˜ represents the specific variable of interest analyzed, which can be at the individual level (Gender, Ethnicity, SES) or at the school level (Private school, Urban or PDET), which is why the subscript can be i or s. μCovid−19 corresponds to a Covid-19 fixed effect taking the value of 1 from 2020 onwards. Zg,t|g=i,s captures the other variables of interest either from the individual or the school levels which are included as controls in every regression9. Xi,t captures the values of other control variables included in the model, μp and μscorrespond to period and school fixed effects, respectively.8 Finally, εi,s,t is the stochastic error of the model clustered at the school level to control for potential serial correlation among students in the same school.
The two parameters of interest are δ1 and δ2. The first captures the pre-pandemic inequality associated with the variable; the second, captures how that inequality changed in the pandemic years. Hence, the relative change in inequality gaps during the pandemic is measured by the ratio between δ2 and δ1. Note that in all cases we control by all other variables of interest, other school level variables, and period and school fixed effects.
The variables in the matrices Zg,t|g=i,s and Xi,t are included as control measures for student characteristics. By controlling for these variables and introducing fixed effects, we measure the change in the net difference in the outcome variable for each characteristic of Zg,t|g=i,s. The results are presented in percentile points and normalized by the level of inequality prior to the pandemic captured by δ1. Additional to measuring the change in learning inequalities during the pandemic, this methodological approach enables us to explore the dynamic evolution of inequalities over time. Hence, by exploiting the dynamic variation, we specified the following model:(2) yi,s,t=δ1Zg,t|g=i,s˜+∑y=20162018αyZg,t|g=i,s˜*μy+∑y=20202021βyZg,t|g=i,s˜*μy+θZg,t|g=i,s+βXi,t+μp+μs+εi,s,t
Where yi,s,tcorresponds to the percentile occupied by the student i who attended school s and took the Saber-11 test in year t. Zg,t|g=i,s˜ represents the specific variable of interest analyzed, which can be at the individual level or at the school level, which is why the subscript can be i or s. Zg,t|g=i,s captures the other variables of interest either from the individual or the school levels which are included as controls in every regression. Xi,t captures the values of other control variables included in the model, μy, μp and μscorrespond to year, period, and school fixed effects, respectively. Finally, εi,s,t is the stochastic error of the model clustered at the school level to control for potential serial correlation among students in the same school. On the one hand, αy estimates correspond to the changes in inequality gaps for the interest variable Zg,t|g=i,s˜ in the 2016–2018 period, relative to the 2019 baseline level measured by δ1. On the other, βy captures the dynamic evolution of learning inequality gaps during the Covid-19 pandemic.
Econometrically, it is desirable that αyestimates are statistically null or have the opposite sign of βy. Even if in this paper we do not aim to estimate causal relationships, it is desirable to find these results as they will allow us to show that the Covid-19 pandemic changed the trend in the evolution of learning inequality for senior students in Colombia.
5 Results
In this section we present the change in learning gaps based on the results of Saber 11 test between the period before and after the pandemic hit, following the methodology developed in the previous section. Table 2 summarizes the findings for the static model and the results are presented according to the dimension to which each variable of interest belongs: individual variables, household variables, school variables, and municipality variables.Table 2 Learning inequality variation during Covid-19.
Table 2 Individual Household School Municipality
VARIABLES Gender Ethnicity SES Private Urban PDET
Zg,t|g=i,s˜ 6.007 *** 1.508 *** 2.763 *** -0.662 0.238 -0.448
(0.049) (0.146) (0.054) (1.143) (0.570) (0.448)
Zg,t|g=i,s˜*μCovid -1.073 *** 0.782 *** 1.315 *** 2.068 *** 0.886 *** 0.676 ***
(0.066) (0.170) (0.076) (0.121) (0.125) (0.173)
Controls ✓ ✓ ✓ ✓ ✓ ✓
School fixed-effects ✓ ✓ ✓ ✓ ✓ ✓
Semester fixed-effects ✓ ✓ ✓ ✓ ✓ ✓
Observations 2,537,477 2,537,477 2,537,477 2,537,477 2,537,477 2,537,477
R-square 0.405 0.405 0.396 0.405 0.405 0.405
Note: Each column corresponds to an interest variable. The estimates correspond to the association between these characteristics and the percentile in the test score, and their marginal percentile points change during Covid-19. The characteristics are defined as follows: male = 1; non-ethnic = 1; high SES = 1; private school = 1; urban school = 1; non-PDET = 1.
Standard errors, clustered at the school level, in parenthesis.
*** p < 0.01, ** p < 0.05, * p < 0.1
5.1 Gaps by Individual Characteristics
The results of the static model suggest that the gender inequality gap decreased 1.073 percentiles points during the 2020–2021 period. This estimate suggests a reduction in the gender inequality gap of 17.9%. Fig. 1 presents the estimates of αy and βy vectors from Eq. (2) and suggests that during 2016 and 2017 the gender inequality gap remained statistically unchanged and decreased in less than 0.3 percentiles points during 2018. Therefore, the yearly reduction of more than 1 percentile point in 2020 and further in 2021 represents a change in the trend of the gender inequality gap for senior students in Colombia.Fig. 1 Learning inequality variation by individual characteristics. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. 1
In other contexts, there is evidence of reductions in the gender learning gap due to larger dropout rates of the most vulnerable women (Kwauk et al., 2021). As the purpose of this article is not to develop in-depth analysis for each of the dimensions explored, we compute gross share of women taking the Saber 11 exam before and during the pandemic by SES status at the school level. For the lowest SES status, the share of women taking the test before the pandemic was 3.78%, while during the 2020–2021 period it was 2.38%. For the mid-low and mid-high SES statuses, we identify an increasing participation of women (21.69–22.09%, and 16.46–18.11%, respectively). However, we also identify a decreasing share for women belonging to the higher SES category from 3.5% to 2.81%. Even if these statistics suggestively support the hypothesis proved in other contexts, we are not able to estimate a reliable dropout measure as the denominator to compute the share of women taking the exam could be biased by specific trends in school dropout of women and vulnerable students. Therefore, we limit our analysis to this descriptive evidence, but we do not claim that the dropout mechanism completely explains our findings for the gender learning gap.
In terms of the ethnicity gap, the results of the static model suggest an increase in the inequality gap of 0.782 percentiles points during the pandemic. This represents a 51.9% increase relative to the level of inequality prior to the pandemic years. Even if the average ethnicity gap increased during Covid-19, the yearly estimates of the dynamic model do not point towards a statistically significant increase during 2020 and 2021.
5.2 Gaps by Household Characteristics
For the household level characteristics, as mentioned in the previous section, we employ a SES index to classify students between low and high SES in the household. The results suggest an increase of 1.315 percentiles points on the SES learning inequality gap during the pandemic period, corresponding to a 47.6% increase. According to the dynamic model, Fig. 2 shows that the marginal change in inequality gap was negative or null before the pandemic, while it was positive and statistically significant during Covid-19.Fig. 2 Learning inequality variation by SES status. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. 2
Moreover, in Table 3 we present results for the SES categorical variable in order to explore differences in magnitudes by level of SES status. Even if inequality gaps increased in every SES category compared to low SES students, these results suggest that the widening of the inequality gaps during the pandemic years is driven by the difference between high and low SES students. In the Table, we show that the percentile change in inequality gaps is increasing on SES status as the estimate corresponding to high SES students is 1.858, while for mid-low and mid-high SES students the estimates correspond to 0.266 and 1.426 percentiles points, respectively.Table 3 Learning inequality variation during Covid-19 by SES status.
Table 3VARIABLES Saber11 percentile
SESmid−low 1.021 ***
(0.066)
SESmid−low*μCovid 0.266 ***
(0.102)
SESmid−high 3.438 ***
(0.081)
SESmid−high*μCovid 1.426 ***
(0.111)
SEShigh 7.409 ***
(0.114)
SEShigh*μCovid 1.858 ***
(0.139)
Controls ✓
School fixed-effects ✓
Semester fixed-effects ✓
Observations 2,537,477
R-square 0.397
Note: The estimates correspond to the association between SES status and the percentile in the test score, and their marginal percentile points change during Covid-19. The baseline category of the SES variable is low SES students.
Standard errors, clustered at the school level, in parenthesis.
*** p < 0.01, ** p < 0.05, * p < 0.1
5.3 Gaps by school characteristics
Moving to the results at the school level, we find the largest variations in learning inequality. Our static results suggest an increase in the inequality gap of 2.068 percentile points between private and public schools. Before 2020, controlling for observable measures and unobservable time-invariant school characteristics, the gap between private and public schools was negative and unsignificant. These results suggest that the years of the Covid-19 are associated with the surge of a private-public school’s inequality gap. In Fig. 3 we present the estimates from Eq. (2) and our findings suggest that, prior to the pandemic years, the dynamic evolution of the school sector inequality gap was negative and nearly statistically unsignificant. However, since 2020 we quantify a significant increase of this inequality gap that continued to grow in 2021.Fig. 3 Learning inequality variation by school characteristics. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. 3
In terms of the inequality gap between schools in rural and urban areas, we find an increase of 0.886 percentiles points during the Covid-19 period, representing a 372.3% change. Additional to a significant increase in this inequality gap, our results point to a change in the trend of this variable as presented in Fig. 3. The dynamic model suggests null or negative estimates prior to the pandemic, that turned positive and economically significant in 2020. In contrast to the school sector gap, the 2021 estimate for this variable is statistically equal to 0.
5.4 Gaps by municipality characteristics
Analyzing inequality in Saber11 results in municipalities with high poverty rates and historically exposed to violence (PDET), we find an increase in inequality learning gaps of 0.676 percentiles points in the Covid-19 period. As well as for the private-public learning gap, before the pandemic, the difference between non-PDET and PDET students was negative and unsignificant. Nonetheless, the years of the pandemic are associated with the rise of an inequality learning gap at this level. In Fig. 4, the dynamic estimation suggests null estimates for the periods prior to the pandemic and positive coefficients for the years 2020 and 2021. Hence, these results are straightforward in showing that prior to the pandemic years, inequality learning between students in PDET and not PDET municipalities remained constant, while during the pandemic the inequality gap significantly increased.Fig. 4 Learning inequality variation by PDET status. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. 4
5.5 Robustness checks
A potential limitation of our methodology is that the Covid-19 pandemic could be associated with a change in the composition of cohorts. As we are not able to control for the composition of the sample nor for dropout rates, we include in Appendix 2 robustness check estimations with an inverse probability weighting (IPW) to control for the potential change in the composition of the sample during 2020 and 2021. The results of these estimations are presented in Table A2.1, Table A2.2, and Fig. A2.1. to A2.4 in Appendix 2, and our estimates of interest remain unchanged in relation to our baseline results, indicating that the potential change in the composition of the sample is not a source of bias of our findings.
6 Conclusions
One of the adverse effects of the pandemic was the educational crisis due to the socioeconomic downturn, school closures, teachers and students’ motivation, amongst others (Angrist et al., 2021, Rodríguez-Planas, 2022, Storey and Zhang, 2021). The results presented in this paper suggest that the pandemic years are associated with the widening of Colombia's pre-existent inequality gaps in learning processes of high school students.
Gender was the only dimension in which there was a reduction in inequality. The econometric model suggests that the learning gaps — which considered the characteristics included in the models — increased by more than 48% in all the other variables analyzed. Among the characteristics analyzed, school area stands out with increases in inequality of 372%, compared to pre-existing inequality gaps.
The reduction in the gender gap should be analyzed carefully because it may be related to changes in the composition of the sample. On the one hand, it is possible that the average results differentiated by gender are explained by differences in motivation, learning methods or skills between men and women (Henderlong Corpus & Lepper, 2007; Mancini, 2018). On the other hand, it has been found that the reduction in the achievement gap is explained by the disproportionately higher dropout rates of women in vulnerable contexts due to limitations to their mobility, child, and elder care activities, performing unpaid domestic work, and the greater likelihood of getting married (Kwauk et al., 2021). In some countries, female school dropout during the pandemic resulted in transactional sex and teenage pregnancy (Oulo et al., 2021; Agile/Gender P.M.A. & ICRHK, 2020). In Colombia, according to data from the National Agency of Statistics (DANE by its Spanish acronym), between 2020 and 2021 there was an increase of 31.5% in the number of pregnancies among women under 14 years old. Higher dropout rates among vulnerable women would have a compositional effect on test scores by not observing the test scores of those who dropped out. However, the results presented in this study are not conclusive regarding the mechanisms explaining the decrease in the gender gap and we do not have a reliable measure of female dropout during the pandemic years, thus its identification is beyond the scope of this paper. Further research is needed to better understand the observed change in the gender gap.
One limitation of this study is that we did not control for dropout at the school level, a factor that may be correlated with average Saber 11 scores in 2020 and 2021. This implies that the results may be underestimated as the estimated changes in learning inequality correspond only to those students who did not drop out. However, the robustness checks with the IPW suggest that our results remain qualitatively unchanged when controlling for the probability of students taking the exam during the 2020–2021 period. Finally, the observed changes in inequality correspond to students who experienced the pandemic only during the last two years of high school. It is possible that the effect of the pandemic on younger students’ learning processes is even greater and that these unevenly distributed losses accumulate over time. Given these limitations, our results are a lower bound of the change in learning inequality during the pandemic in Colombia.
In the short term, the public policy implication of our results points out to the need to set up learning recovery strategies that target the most vulnerable students. These results also highlight the need for structural changes that address learning inequalities (and low levels of learning) that were present before the pandemic and were exacerbated after the Covid-19 crisis. For instance, it is necessary to secure access to adequate infrastructure and highly qualified teachers to students belonging to groups where learning inequalities grew the most.
Providing specific policy recommendations goes beyond the scope of this paper, however, a bottom-up approach could be implemented by teachers to be responsive to the individualized needs of students from different backgrounds and with different characteristics that condition their learning processes. Thus, assuring qualified teachers among locations or groups that suffered the most can be a way to address learning inequality. Qualified teachers have to acknowledge and identify how inequality exists within their classrooms to design strategies that reinforce the learning process of disadvantaged students.
As suggested by the literature and evidence in recent years, structural changes to education policy are difficult to materialize (Angrist et al., 2021). However, the educational crisis and negative learning outcomes increase the urgency for concrete interventions. In Colombia, and in several contexts, there is an opportunity to align the incentives of teachers, policy makers, and other actors to implement these interventions. Programs that include teacher training have shown to have positive results on student learning (Crouch, 2020, Angrist et al., 2021). Similarly, the infrastructure and the lessons learned from structured pedagogy programs are scalable beyond the current beneficiary schools and hold elements of the policy recommendations (Barrera-Osorio et al., 2018). The immediate actions that we propose hereby are key to bridge educational inequality gaps due to the pandemic and its socioeconomic implications, and we consider them essential to prevent further deepening on inequality in the forthcoming decades.
Ethics
This article did not involve the use of animals nor human participants.
Funding Statement
This research received funding from Fundación Barco in order to finance two of the authors.
Declaration of Interest
None.
Appendix 1: Learning outcomes mean differences by characteristic before and during Covid-19
See Table A1.1.Table A1.1 Descriptive statistics of learning inequality before and during Covid-19.
Table A1.1 Pre-Covid (2016–2019) Covid (2020–2021)
Equals 0 Equals 1 Mean difference Equals 0 Equals 1 Mean difference
Gender 57.88 53.33 4.55 *** 55.51 51.59 3.92 ***
(0.00) (0.00)
Ethnicity 56.73 35.98 20.75 *** 54.61 33.49 21.12 ***
(0.00) (0.00)
SES 65.45 46.31 19.13 *** 63.30 43.35 19.94 ***
(0.00) (0.00)
School sector 70.38 51.14 19.24 *** 70.17 49.01 21.17 ***
(0.00) (0.00)
School area 57.74 42.52 15.22 *** 55.94 40.15 15.78 ***
(0.00) (0.00)
PDET 57.06 42.98 14.08 *** 55.07 41.02 14.05 ***
(0.00) (0.00)
Note: This table presents the mean difference in test percentiles by each socioeconomic characteristic before and during the Covid-19 pandemic. The first mean difference column corresponds to the mean difference in test results before the pandemic, and the last column of the table presents the corresponding statistic during the pandemic. Men= 0, Women= 1, Non ethnic population= 0, Ethnic population= 1, High SES= 0, Low SES= 1, Private school= 0, Public school= 1, Urban school= 0, Rural school= 1, Non PDET= 0, PDET= 1.
*** p < 0.01, ** p < 0.05, * p < 0.1.
Appendix 2: Learning inequality variation during Covid-19 including IPW
See Table A2.1.Table A2.1 Learning inequality variation during Covid-19.
Table A2.1 Individual Household School Municipality
VARIABLES Gender Ethnicity SES Private Urban PDET
Zg,t|g=i,s˜ 5.955 *** 1.357 *** 2.623 *** -0.450 0.423 -0.446
(0.050) (0.159) (0.057) (1.249) (0.594) (0.476)
Zg,t|g=i,s˜*μCovid -1.126 *** 0.617 *** 1.284 *** 2.184 *** 0.656 *** 0.564 ***
(0.071) (0.175) (0.077) (0.127) (0.126) (0.172)
Controls ✓ ✓ ✓ ✓ ✓ ✓
School fixed-effects ✓ ✓ ✓ ✓ ✓ ✓
Semester fixed-effects ✓ ✓ ✓ ✓ ✓ ✓
IPW ✓ ✓ ✓ ✓ ✓ ✓
Observations 2,537,477 2,537,477 2,537,477 2,537,477 2,537,477 2,537,477
R-square 0.410 0.410 0.397 0.410 0.410 0.410
Note: Each column corresponds to an interest variable. The estimates correspond to the association between these characteristics and the percentile in the test score, and their marginal percentile points change during Covid-19. The characteristics are defined as follows: male = 1; non-ethnic = 1; high SES = 1; private school = 1; urban school = 1; non-PDET = 1.
Standard errors, clustered at the school level, in parenthesis.
* ** p < 0.01, * * p < 0.05, * p < 0.1
See Table A2.2.Table A2.2 Learning inequality variation during Covid-19 by SES status.
Table A2.2 Saber11 percentile
VARIABLES
SESmid−low 0.956 ***
(0.069)
SESmid−low*μCovid 0.242 **
(0.104)
SESmid−high 3.248 ***
(0.087)
SESmid−high*μCovid 1.388 ***
(0.113)
SEShigh 7.195 ***
(0.119)
SEShigh*μCovid 1.763 ***
(0.144)
Controls ✓
School fixed-effects ✓
Semester fixed-effects ✓
IPW ✓
Observations 2,537,477
R-square 0.399
Note: The estimates correspond to the association between SES status and the percentile in the test score, and their marginal percentile points change during Covid-19. The baseline category of the SES variable is low SES students.
Standard errors, clustered at the school level, in parenthesis.
* ** p < 0.01, * * p < 0.05, * p < 0.1
See Fig. A2.1, Fig. A2.2, Fig. A2.3, Fig. A2.4.Fig. A2.1 Learning inequality variation by individual characteristics. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. In these models we include the IPW. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. A2.1
Fig. A2.2 Learning inequality variation by SES status. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. In these models we include the IPW. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. A2.2
Fig. A2.3 Learning inequality variation by school characteristics. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. In these models we include the IPW. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. A2.3
Fig. A2.4 Learning inequality variation by PDET status. Note: the point estimates correspond to αy and βyvectors from Eq. (2) and represent the percentile points change in inequality by each characteristic in comparison to 2019. In these models we include the IPW. The whiskers are the confidence intervals of the percentage change with a significance level of 95%.
Fig. A2.4
☆ This article is the result of a project led by Sandra García and Darío Maldonado in partnership with Universidad de Los Andes and the Fundación Barco who financed Lucas Marín Llanes and Mariana Rodríguez Pico. We are thankful to the comments of Barco’s staff and to Andrés Molano to a preliminary version of this paper.
1 Considering the number of weeks of total educational center closures due to Covid-19, from March 2020 to May 2021.
2 According to the Observatorio de la Gestión Educativa, Santa Marta and Barrancabermeja are the cities lagging the furthest behind with 19% and 38% progress in reopening, respectively. Bogota and Medellin rank highest with 91% and 98% respectively. However, we do not have access to this data at the school level nor this information homogenized the definition of in-person classes.
3 The Territorially Focused Development Plans (Planes de desarrollo con enfoque territorial, PDET) were designed within the framework of the Peace Agreement. They were established to prioritize the execution of investments in municipalities historically affected by the armed conflict and with the highest poverty rates in the country.
4 Chronologically, the National Government issued an executive order on March 12th, 2020, to close education institutions. On September 1st, 2020, the Government began a reopening pilot of some economic sectors. The education sector, including schools, remained excluded. It was until October 19th, 2020, that a pilot for reopening schools began. Nonetheless, the reach of this pilot was limited even in Bogota, Colombia’s capital, where only 15% of public schools participated. This process resulted in more than 50% of the students remaining in remote learning by the end of 2020.
5 Data from CEDE Municipal Panel (Panel Municipal del CEDE), the Agency for Territorial Renewal (Agencia de Renovación del Territorio, ART) and from the database reported by Osorio et al. (2019).
6 Despite there being a progressive increase in internet access, the change between the years prior to Covid-19 and 2020 was significant. In 2016, 55.6% of students did not have internet access at home. This percentage was 40.6%, 38.5%, 36.4%, and 28.5% from 2017 to 2020, respectively.
7 The program Hogares Conectados was launched in May of 2020 with the purpose of connecting to the internet 500,000 families at accessible costs (MinTIC, 2021).
8 These matrices include the variables of interest (Zg,t|g=i,s) and other control variables (Xi,t). In this enumeration are all the variables of interest and control variables in the model.
Individual level: gender, ethnicity, age
Household level: computer, internet, books in the home, television, parents' highest education and overcrowding.
School level: sector, area, school day, nature (academic/technical) and educational institution.
Municipality level: rurality index, distance to the capital, MPI, PDET, presence of armed groups, luminosity.
==== Refs
References
Agile/Gender P.M.A. ICHRK Y. Gender & Covid-19: Transactional Partnerships & Sex Trade. Baltimore, Maryland 2020 Bill & Melinda Gates Institute for Population and Reproductive Health, Johns Hopkins University Bloomberg School of Public Health USA
Alban Conto C. Akseer S. Dreesen T. Kamei A. Mizunoya S., y Rigole A. Potential effects of covid-19 school closures on foundational skills and Country responses for mitigating learning loss Int. J. Educ. Dev. 2021 87
Alvarado F. Meléndez M. Pantoja M. Mercados laborales fragmentados y sistema de proteccion social en Colombia UNDP-LAC Work. Pap. Ser. 2021
Andrabi T. Daniels B. Das J. Human capital accumulation and disasters: evidence from the Pakistan earthquake of 2005 20/039 RISE Work. Pap. Ser. 2020 10.35489/BSG-RISE-WP_2020/039
Angrist N. De Barros A. Bula R. Chakera S. Cummiskey C. DeStefano J. Stern J. Building back better to avert a learning catastrophe: estimating learning loss from covid-19 school shutdowns in Africa and facilitating short-term and long-term learning recovery Int. J. Inst. Educ. 2021
Ardington C. Wills G., y Kotze J. covid-19 learning losses: early grade reading in South Africa Int. J. Educ. Dev. 2021 86
Azevedo J.P. Hasan A. Goldemberg D. Geven K. Iqbal S.A. Simulating the potential impacts of COVID-19 school closures on schooling and learning outcomes: a set of global estimates World Bank Res. Obs. 36 1 2021 1 40
Banerjee A. Duflo E. Poor Economics 2011 Public Affairs New York
Barrera-Osorio F. García S. Rodríguez C. Sánchez F. Arbeláez M. Concentrating efforts on low-performing schools: impact estimates from a quasi-experimental design Econ. Educ. Rev. 2018 66
Busso, M., y Camacho, J. (2020). Pandemia y desigualdad: ¿cuánto capital humano se pierde cuando los colegios cierran? Ideas que cuentan.
Chatterji P. Li Y. Effects of COVID-19 on school enrollment Econ. Educ. Rev. 2021 83
Crouch, L. (2020). Systems Implications for Core Instructional Support Lessons from Sobral (Brazil), Puebla (Mexico), and Kenya. rise Insight Series. 2020/020. https://doi.org/10.35489/BSG-RISE-RI_2020/020.
DANE. (2022). Comunicado de prensa pobreza monetaria. Departamento Administrativo Nacional de estadìstica, Bogotá.
DANE. (2023). Boletín Técnico Cuentas nacionales anuales. Departamento Administrativo Nacional de Estadística, Bogotá.
Engzell, P., Frey, A., & Verhagen, M.D. (2020). Learning Loss Due to School Closures During the covid-19 Pandemic. 10.31235/osf.io/ve4z7.
Fernández M. Eslava M. El mercado laboral en Colombia: diagnóstico y recomendaciones para un funcionamiento más equitativo y eficiente Nota Macroeconómica – Ed. Espec. Debate Elect. 2022 36
García, S., & Maldonado, D. (2021). Covid-19 y educación: Implicaciones del cierre de colegios y perspectivas para el 2021. Universidad de Los Andes Escuela de Gobierno y ProBogota, Bogotá
García, S., Maldonado, D., & Abondano, S. (2021). Afectaciones de la pandemia en la educación de los niños, niñas y adolescentes en Latinoamérica: el caso de Colombia en colegios oficiales.
Heavey, L., Casey, G., Kelly, C., Kelly, D., y McDarby, G.. (2020). No evidence of secondary transmission of covid-19 from children attending school in Ireland, 2020. Euro Surveill.
Henderlong Corpus J. Lepper M.R. The effects of person versus performance praise on children’s motivation: Gender and age as moderating factors Educ. Psychol. 27 4 2007 487 508
Hevia F.J. Vergara-Lope S. Velásquez-Durán A., y Calderón D. Estimation of the fundamental learning loss and learning poverty related to covid-19 pandemic in Mexico Int. J. Educ. Dev. 2022 88
Hossain M. Unequal experience of covid-induced remote schooling in four developing countries Int. J. Educ. Dev. 2021 85
Kaffenberger M. Modelling the long-run learning impact of the covid-19 learning shock: Actions to (more than) mitigate loss International Journal of Educational Development 2021 Elsevier, Oxford
Kwauk C. Schmidt D. y Ganju, E What do we know about the effect of covid-19 on girls’ return to school? 2021 Brookings,
Li X. Zhou Y. Zhao M., y Zhao X. A harmonized global nighttime light dataset 1992-2018 Sci. data 7 1 2020 1 9 31896794
Li X., y Zhou Y. Urban mapping using dmsp/ols stable night-time light: a review Int. J. Remote Sens. 38 21 2017 6030 6046
Maldonado, J.E., y De Witte, K. (2020). The effect of school closures on standardised student test outcomes. British Educational Research Journal.
Mancini N. Motivación relacionada a una tarea cognitiva en niñas y niños preescolares. Presentación de un instrumento para su evaluación Anuario de InvestigacionesXXV 2018 455 462
Mani A. Mullainathan S. Shafir E. Zhao J. Poverty Impedes Cognitive Function Science 2013 341 10.1126/science.1238041
MinTIC. (2021). ¿Qué es Hogares Conectados? https://mintic.gov.co/micrositios/hogaresconectados/744/w3-channel.html.
Neidhöfer G. Lustig N. Intergenerational transmission of lockdown consequences: prognosis of the longer-run persistence of covid-19 in Latin America J. Econ. Inequal. 19 2021 571 598 34366749
OECD, Cepal, CAF , & European Comission (2021). Perspectivas económicas de América Latina 2021: avanzando juntos hacia una mejor recuperación. Informe leo, ocde, Centro de Desarrollo, París.
Opportunity Insights Economic Tracker (2020). Percent change in Student Math Progress.
Osorio J. Mohamed M. Viveca P., y Brewer-Osorio S. Mapping violent presence of armed actors in Colombia Adv. Cartogr. GIScience ICA 1 2019 1 9
Oulo, B., Sidle, A.A., Kintzi, K., Mwangi, M., Akello, I. (2021). Understanding the Barriers to Girls’ School Return: Girls’ Voices from the Frontline of the covid-19 Pandemic in East Africa. amplify covid-19 Research Brief. Nairobi, Kenya.
Rajmil L. Role of children in the transmission of the covid-19 pandemic: a rapid scoping review bmj Paediatr. Open 4 1 2020 e000722
Rodríguez-Planas N. Hitting where it hurts most: COVID-19 and low-income urban college students Econ. Educ. Rev. 2022 87
Shah A.K. Mullainathan S. Shafir E. Some consequences of having too little Science 2012 338 10.1126/science.1222426
Storey, N., y Zhang, Q. (2021). A Meta-analysis of covid Learning Loss.
Sullivan W.T. III A 10 km resolution image of the entire night-time Earth based on cloud-free satellite photographs in the 400–1100 nm band Remote Sens. 10 1 1989 1 5
The United Nations (2020). Education during COVID-19 and beyond. Policy Brief. https://www.un.org/development/desa/dspd/wpcontent/uploads/sites/22/2020/08/sg_policy_brief_covid-19_and_education_august_2020.pdf.
Thompson P.N. Ward J. Only a matter of time? The role of time in school on four-day school week achievement impacts Econ. Educ. Rev. 2022 86
Unicef (2021a). Las escuelas de más de 168 millones de niños del mundo llevan casi un año entero cerradas por completo debido a la covid-19. Naciones Unidas de Nueva York.
Unicef (2021b).LACRO COVID-19 Respuesta educativa: Update 22 Estado de la reapertura de escuelas, 29.
Valle J.M. Miles de escuelas siguen cerradas por la covid-19: así es la educación en el mundo hoy Conversation 2020
Yang C. Yu B. Chen Z. Song W. Zhou Y. Li X., y Wu J. A spatial-socioeconomic urban development status curve from NPP-VIIRS nighttime light data Remote Sens. 11 20 2019 2398
|
PMC010xxxxxx/PMC10151159.txt |
==== Front
Appl Mater Today
Appl Mater Today
Applied Materials Today
2352-9407
2352-9415
Elsevier Ltd.
S2352-9407(23)00103-8
10.1016/j.apmt.2023.101833
101833
Article
Electrospun nanofibers for medical face mask with protection capabilities against viruses: State of the art and perspective for industrial scale-up
Cimini A. ab⁎
Imperi E. b
Picano A. b
Rossi M. ac
a Department of Basic and Applied Sciences for Engineering, University of Rome Sapienza, Rome 00161, Italy
b LABOR s.r.l., Industrial Research Laboratory, Via Giacomo Peroni, 386, Rome, Italy
c Research Center for Nanotechnology for Engineering of Sapienza (CNIS), University of Rome Sapienza, Rome 00185, Italy
⁎ Corresponding author at: Department of Basic and Applied Sciences for Engineering, University of Rome Sapienza, Rome 00161, Italy.
2 5 2023
6 2023
2 5 2023
32 101833101833
16 12 2022
13 4 2023
25 4 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Face masks have proven to be a useful protection from airborne viruses and bacteria, especially in the recent years pandemic outbreak when they effectively lowered the risk of infection from Coronavirus disease (COVID-19) or Omicron variants, being recognized as one of the main protective measures adopted by the World Health Organization (WHO). The need for improving the filtering efficiency performance to prevent penetration of fine particulate matter (PM), which can be potential bacteria or virus carriers, has led the research into developing new methods and techniques for face mask fabrication. In this perspective, Electrospinning has shown to be the most efficient technique to get either synthetic or natural polymers-based fibers with size down to the nanoscale providing remarkable performance in terms of both particle filtration and breathability. The aim of this Review is to give further insight into the implementation of electrospun nanofibers for the realization of the next generation of face masks, with functionalized membranes via addiction of active material to the polymer solutions that can give optimal features about antibacterial, antiviral, self-sterilization, and electrical energy storage capabilities. Furthermore, the recent advances regarding the use of renewable materials and green solvent strategies to improve the sustainability of electrospun membranes and to fabricate eco-friendly filters are here discussed, especially in view of the large-scale nanofiber production where traditional membrane manufacturing may result in a high environmental and health risk.
Graphical abstract
Image, graphical abstract
==== Body
pmc1 Introduction
In recent years, the common commercial microfibers have been replaced with nanofibers in manufacturing personal protective equipment (PPE), such as face mask/respirators, which are specifically designed to protect the wearer from inhaling harmful airborne particles, including infectious agents, such as coronavirus, SARS, and bacteria spores. The 2019 coronavirus (SARS-CoV-2) pandemic has forced a widespread use of face coverings as a mandatory step for reducing spreading and infection by the virus, in fact the face masks turned out to be an essential barrier for preventing the transmission of infected aerosols among its user and the surrounding people [1], [2], [3]. Especially at the start of the pandemic outbreak face masks played a fundamental role, since information about the transmission mechanism of the virus was not yet clear and only isolation/ quarantine appeared to be the most significant strategy for infection control [4].
The size of the Corona virus has been observed to range from 60 to 140 nm [5]. The transmission of the virus can occur for close-distance contacts in the form of aerosols or droplets [6], because of its ability to coalesce with solid or liquid pollution particles present in the atmosphere, which makes it extremely infectious [7,8]. Therefore, the aerosol carrier of the virus (diameter < 5 −10 μm) can be transmitted from an infected person to a healthy one by means of respiratory droplets either for breath, coughs, or sneezes [9,10]. The short-range airborne route results to be the primary way to spread either pathogens or respiratory infection [11,12] such that wearing face masks is the most effective solution to prevent the transmission of this disease [13].
Despite the vaccination campaign against COVID continues to progress [14], the importance of wearing a face mask or a respirator in outdoor/indoor places is recommended as an effective measure for the infection control [15], [16], [17]. This led the research community to find more appropriate technologies and advanced materials to design novel high-efficiency mask, with extreme attention to particle filtration, cleaning treatments, adequate breathability, fluid penetration resistance, light weight, and user comfort, along with large-scale production and low costs. In this scenario, electrospun fibers showed their uniqueness and potential as active membranes, due to their reduced pore size as well as their extremely high surface-to-volume ratio [18]. The high performance in filtration is owed to the reduced pore size, which goes from sub-micron to several micrometers, involving a more effective capture mechanism for the small airborne particles below 300 nm, as well as to the high ratio of surface area/volume, which significantly increases the possibility of pollutant deposition on the surface of the fiber, thus reducing the air pressure drop and increasing the air breathability of the mask [19]. Although the melt blown (MB) is a low-cost process commonly used to produce nonwoven filter media [20], [21], [22], [23], the electrospinning (ES) technique has proved to produce nanofibers with better filtration performance in removing submicron particles/contaminants. Compared to the MB microfibers, whose fibers size are generally difficult to control and the diameters range from 1 to 10 μm, the electrospun fibers diameter can be ten or a hundred times smaller, thus providing a higher surface area as well as a lower inter-fiber pore size [24], [25], [26]. Several Reviews reported about the potentiality in using electrospun membranes for new generation face mask/respirators production [27], [28], [29], [30]. An increasing number of papers was found by using the keyword “Electrospinning filter face mask “in the Scopus database (Fig. 1 ).Fig. 1 Number of annual publications on Electrospun filters for face mask applications. The data have been collected in a search carried out on the “Scopus” database on 2023–04–06 by using the keyword “Electrospinning filter face mask“.
Fig 1
Besides the high filtering efficiency benefits, the functionalization of electrospun nanofibers through the addition of several compounds, such as metallic and not metallic nanomaterials, vesicles, and cellulose nanocrystals, allows designing advanced nanofibers with outstanding properties, including antibacterial, antiviral, and self-sterilization. Such customizability makes the electrospun nanofibers the ideal material for tackling the current issues from bacterial contamination on PPE surfaces to the reusability of both disposable single use facemask and respirator's filters. Also, modified ES set-ups combined with modern textile techniques turned out to be an innovative way to manufacture nanofiber bio-textiles with suitable mechanical and bioactive properties which effectively support cell tissue regeneration in clinical use [30,31]. Electrospun nanofibers composed by natural and biodegradable polymers can be easily degraded in environment or absorbed by the body, thus preventing the negative long-term degradation effects due to high consume of plastic, ensuring a safer approach to the realization of scaffold and drug delivery system for biomedical application [32], [33], [34], [35]. Although a great number of research in the electrospinning field has been carried out at laboratory scale, alternative techniques have been developed so far to expand the production of electrospun nanofibers on large industrial scale [36]. To date, many companies and start-ups operating in small and large textile manufacturing have been utilizing ES technology to fabricate face masks and protective controls with better performances on personal protection from particulate matter (PM) and airborne hazard. The leading ES research is focusing on the optimization of more sustainable filters to minimize the impact of disposable face masks on the environment, but the large volume of hazardous solvent still used in traditional membrane manufacturing involves a serious environmental issue. With this regard, particular attention has been given to more environmental-friendly ES processes to drastically reduce or completely give up toxic organic solvents, which are strictly used to solubilize polymer in conventional spinning solution.
This Review aims to give a general overview about the recent developments in ES techniques and its further application to produce electrospun nanofiber-based face mask. Firstly, a brief introduction about the fundamental principles behind the filtering mechanism, classification, and the standards for testing the face mask is provided. Then, is presented a description on conventional and alternative ES technologies, evaluating the possible developments for industrial scale-up. After that, are reviewed in detail the innovative scalable strategies present in literature for the processing of electrospun based face masks, highlighting the effects induced by the surface morphology, fibers geometry, and materials properties on the particle filtration, breathability, antibacterial, and antiviral performance. A brief introduction about the recent progress in the manufacturing of bio-textiles by using electrospun nanofiber yarns is also given in this Review. Finally, we discuss about the industrial implementation of ES for nanofiber manufacturing, giving to the reader further insight into the well-defined nanofiber market.
2 Face mask's structure and filtration mechanism
2.1 Surgical mask and respirator standards
Face masks have the important role to protect the wearer from the biological contaminants that can be present in the form of droplet or aerosols in the atmosphere, since respiratory infection may be transmitted due to either talking or coughing droplets from an infected person. The method of infection has been observed to depend on the droplet size [4,37]. Large droplets with diameter ranging between 2.5 and 10 μm can deposit due to gravity either in the nasal, oropharyngeal, laryngeal, or tracheal regions of the respiratory system, whereas the smaller ones ranging between 0.25 and 1 μm can evaporate midair involving airborne transmission eventually depositing in the respiratory tract. In addition, intense coughs and sneezes have been observed to propel larger droplets over 20 feet that can remain airborne for hours [38]. Therefore, face masks are expected to prevent the penetration of contaminant present in the surroundings such as particulate and dust, which can be potential bacteria or virus carriers. Among the various classifications, both the surgical masks and the face piece respirators have been taken in consideration as effective devises in preventing viral infection by airborne contamination.
A surgical or medical mask is used as medical device for health care workers (HCWs) to avoid that patient in hospitals become infected from possible expiratory droplets. These devices are generally used for surgical operations, since they are fluid resistant providing a useful physical barrier from larger droplets and body fluids, despite their loosely fit on the wearer face. The common surgical masks are made of three nonwoven layers, all of which are generally composed of polypropylene (PP) thermoplastic polymer [18,27,[39], [40], [41]] (Fig. 2 a). The cover and shell layer are fabricated with spunbond fabric to provide high tensile strength and good breathability in the mask, respectively. These layers show fibers diameter ranging between 15 and 40 μm, and do not provide any contribute to the overall filtration system. On the other hand, the inner layer is fabricated by means of MB process, which allow to obtain submicron fibers with diameter size ranging between 0.5 and 10 μm. This latter layer sandwiched between spunbond fabrics performs as the filtering system. As their relatively large diameter is insufficient to efficiently filter small airborne particles, electrostatic treatment is performed on the filter to improve the filtration efficiency. However, the static electricity quickly depletes due to long use of the mask with a consequent reduction of the filtration properties, so that they can be used for one time only. The grade of protection from bacteria is obtained by means of Bacterial filtration efficiency test (BFE). This test allows to obtain a measurement of the device's resistance to the penetration of Staphylococcus aureus (S. aureus) aerosols, which is shot with a flow rate equal to 28.3 L/min through the mask by using particle size of 3.0 μm [41]. As the degree of efficiency is reported in percent, higher percent values correspond to better filtration performances from the face mask. According to the European standard (EN 14,683:2019), a surgical mask is classified either as type I, if it presents at least a value of 95% BFE, or as type II, if it provides better filtration performances (BFE ≥ 98%) (see Table 1 ). Furthermore, a third class, namely IIR, includes a splash resistance test to penetration of synthetic blood. About the ASTM F2100–19 standard used in United States, the classification of the surgical masks is almost the same as the EN 14,683:2019, with the exception of a more relevant test required, concerning particle filtration efficiency (PFE) of the submicron particles, with mean diameter of 0.1 μm [42,43].Fig. 2 Structure of a surgical mask (a) and filtering half mask (b). (c) Schematic image of different pathogens and air pollutants for a size comparison.
Fig 2
Table 1 European and USA standards required for surgical mask and facepiece respirators.
Table 1Surgical mask USA ASTM F2100–19 Standard EN 14,683:2019 Standard
Level 1 Level 2 Level 3 Type I Type II Type IIR
BFE (%) ≥95 ≥98 ≥95 ≥98
PFE (%) ≥95 ≥98 Not required
Fluid resistance >80 mmHg
>120 mmHg >160 mmHg Not required >120 mmHg
Differential pressure drop < 5.0 mmH2O/cm2 < 6.0 mmH2O/cm2 < 40 Pa/ cm2 < 40 Pa/ cm2 < 60 Pa/ cm2
Facepiece respirator NIOSH 42 CFR Part 84–2019 standard EN 149:2001+A1 2009
N95 N99 N100 FFP1 FFP2 FFP3
BFE (%) ≥95 ≥99 ≥99.97 ≥80 ≥94 ≥99
PFE (%) ≥95 ≥99 ≥99.97 ≥80 ≥94 ≥99
Inhalation differential pressure 343 Pa at 85 l/min 210 Pa at 95 l/min 240 Pa at 95 l/min 300 Pa at 95 l/min
Exhalation differential pressure 245 Pa at 85 l/min 300 Pa at 160 l/min
EN (European Norm); ASTM (American Society for Testing and Materials; NIOSH (The National Institute for Occupational Safety and Health); CFR (Code of Federal Regulations); BFE (Bacterial Filtration Efficacy); PFE (Particle Filtration Efficiency).
Like surgical masks, a facepiece respirator is composed of polymer-based multiple layers (Fig. 2 b). The outer layers are composed of non-woven fabric with grams per square meters (gsm) ranging between 20 and 50 to create a barrier against moisture. Besides, a higher dense layer of around 250 gsm is employed to provide more stiffness and thickness to the face mask, whereas a more internal layer made via MB process acts as filter. Compared to the surgical masks, the facepiece respirators are tighter and more adherent to the face, in order to avoid inhalation of both droplets (particle size > 0.5 μm) and smaller particles, such as dust, and aerosols (particles size < 0.5 μm), and to provide a real protection from both potential viruses and bacteria [18] (Fig. 2 c). With regards to the EN 149:2001, these respirators are classified as filtering half masks (i.e., Filtering Face Pieces (FFP)) and are divided in three different types, namely FFP1, FFP2, and FFP3, depending on the filtration efficiency values, for both PFE and BFE, equal to 80%, 94%, and 99%, respectively [44], [45], [46] (Table 1). In addition to the filtration performance, the National Institute for Occupational Safety and Health (NIOSH), in the USA, also adopts a classification by letter, namely N-, R-, and P-, to indicate the lack of resistance in oil, somewhat oil resistant, and strongly resistant, respectively. According to this standard, three distinct filtration efficiencies degrees are adopted for the respirators, such as 95%, 99%, and 99.97%, and for each series of letters three facial pieces are classified, for instance N95, R95, and P95 for 95% filtration performance, etc. [4,18].
In high-risk situations both the Centers for Disease Control and Prevention (CDC) and the World Health Organization (WHO) currently recommend to health workers the use of protective sources of respirators with performances of 95% or higher, such as FFP2 and N95 [47,48]. The rigorous standards adopted in Europe and USA have made that these respirators provide higher protection than loose surgical masks, so to prevent users to be infected from either large droplets, originated in air for short periods during cough, sneeze, and talking, or from fine airborne particles which remained in air for long time and have been transported for longer distances. However, there are some clinical researches that have brought to conflicting results. For instance, some studies have evidenced a better effectiveness of N95 respirators in reducing viral infections compared to surgical masks [49,50], while other works reported no significant difference between N95 respirators and medical masks in terms of viral preventing from respiratory infections, including that of influenza [51,52]. Even though concrete evidence about the efficacy in protecting health workers against viral SARS-COV-2 of respirators compared to surgical mask is not totally reliable yet, the utilization of respirators for high-risk setting pandemic is strongly recommended [41,53,54].
2.2 Filtration mechanism of face masks
The filtration effectiveness of a face mask is an estimate of the ability of the inner layer in trapping undesired airborne particles, which could introduce harmful microorganisms or viruses through inhalation of air. The capture mechanism of fibrous media strictly depends on the particle size and can be divided in the following way (Fig. 3 a): direct interception, diffusion, electrostatic deposition, gravitational forces and inertial deposition [18,27,39,55,56]. The interception mechanism generally involves the capture of particles with diameter size lower than 1 μm. When the particle carried by air streamlines is moving in the proximity of the fiber surface, so that their radius is less than the fiber-particle distance, it is affected by Van der Waals attraction forces thus remaining stuck on the mat. Diffusion is another effective method to capture smaller particles, usually those with diameter far below 1 μm. The random collision with other particles from Brownian motion near the fiber brings the aerosol particle to deviate from their original streamline thus impacting with high probability to the fiber surface. Besides, charged fibers can effectively trap submicron particles of opposite charge via electrostatic attraction.Fig. 3 (a) A schematic image of the principal capture mechanisms for fibers. Below, a simplified diagram which displays the particles size range where Electrostatic, Diffusion, Interception, and Inertial impact mechanism are more efficient; an example of electrostatic deposition is represented in the schematic picture from the electrostatic attraction occurring between a negative charge fiber (minus sign) and a charged aerosol particle (blue sign). (b) Graphic representation of the main filtering efficiency curves for a single fiber as function of the particle size; adapted with permission from [57].
Fig 3
The effect of the electrostatic charges is that of improving the overall particles filtration efficiency without affecting the morphology of the fiber and either the interception or the diffusion mechanical capture. Corona discharge and triboelectrification methods are commonly used to apply electrical charge on electrically active polymer materials [58,59]. Finally, larger particles ranging from 1 to 10 μm are expected to be easily captured from gravity and inertia mechanisms. Indeed, bigger particles moving along the stream can be deposited on the fiber surface due to the gravity. On the other hand, the inertial deposition results to be a more effective method to capture either aerosols or dust particles larger in size than 1 μm. Particles moving at high velocity near a fiber can come out from the main streamlines and impact on the fiber surface due to the inertia, hence the higher is the particles velocity the better is the capture efficiency. The mechanical efficiency of the commercial face masks in the submicron particle range is mainly due to the interception and diffusion mechanisms, whereas the contribution of inertial impact and gravity are practically negligible. The lowest filtration particle measured is defined as the most penetrating particle size (MPPS) and corresponds to the least efficiency value ensured by the standards adopted (Table 1). The MPPS differs from one filter system to another and is found in the point of intersection between the diffusion and interception curves, which generally falls in the range between 0.1 and 0.3 μm [26,56] (Fig. 3 b). To improve the overall mechanical efficiency against fine particles for the most commonly produced face masks, industries typically rely on the electrostatic charge effect. The performance obtained for charged fibers in face mask are better than those observed in uncharged one [60]. While the mechanical filtration of large droplets in face masks come mainly from the micro-pore size of the fibers, which in a MB-based filter is about ∼20 μm, the filtration of aerosols is due to the electrostatic attraction of smaller particles at the fibers surface. The polymer materials forming these mask devices are generally electrostatic microfiber media, such as PP, that can be electrostatically charged by means of corona discharge. To charge them a high voltage is applied between the two electrodes and the electret filter, thus allowing the ionization of the air to induce a positive or negative charge on the fiber surface [58,60]. Even though this treatment involves a better performance in respirators/ face mask, the effect of either air humidity or liquid contact on charge decay no longer ensure a proper coverage against small particles. It has been stated that longer exposition to air humidity leads to a reduction of the charge on the fibers, thus resulting in a decrease in the filtration performance [56]. Moreover, ethanol and autoclave sterilization treatment remove electrical charges from the polymer filter media thus affecting the performance of the face mask not guaranteeing a proper protection if reused [61]. If on the one hand respirators are expected to provide a greater filtration performance compared to surgical masks due to the effective seal on the face, lower air leakage, and multi-layer structure, on the other hand the lowering of the electret fiber charge can bring to a significant decrease in the overall collection efficiency and make both the class types useless to face coronavirus disease.
2.3 Breathability test
In addition to filtration efficiency, the overall layers constituting either a surgical mask or a facepiece respirator must be both comfortable and breathable for the users. Indeed, the fabric fibers used to make a mask have to provide an effective barrier against airborne particles, but they also must ensure the person to breath properly [39,58,[62], [63], [64]]. The pressure drop is the most commonly used parameter to quantify a mask breathability, being an index of the air flow resistance of the filter for a specific material surface, which makes it a suitable indicator to assess the filter performance. Pressure drop is obtained at specific flow rates by measuring the difference in pressure between the two outer layers of a face mask [63]. Lower values of pressure drop involve better breathability for the face device, thus making breathing more comfortable for the wearers. The difference in pressure between the upstream and downstream measured on the two outer layers side is proportional to the volumetric air per unit area, also known as face velocity, that crosses through the face mask. The presence of multiple layers as well as the occurrence of inhomogeneity in the air flux in the fabric fiber can affect in many ways the pressure difference measurement for a given tested area. Therefore, a comparison between pressure drop measurements obtained in different experiments with different values of tested face velocity is not advised [63]. The measure of the pressure drop is usually reported in units of Pascal for surface area (Pa/cm2) and varies depending on the standard adopted by the different countries. About surgical masks for instance, the flow rate values, which is generally expressed in liters per minutes (Lpm), used for the test performed both in USA and Europe, according to ASTM F2299 or F2101 and EN 14,683:2019 standards respectively, is fixed to 8 L/min over a tested mask surface of 4.9 cm2, but the maximum pressure drop specified for these tests varies from 40/40/60 Pa/cm2, for type I, II, and IIR in Europe, to 50/60/60 Pa/cm2 for barrier levels equal to 1, 2, and 3 in USA [63,64]. The high number of layers that structure a face piece respirator provide a better resistance to flow stream compared to surgical masks but results also in a large pressure drop which makes breathing more difficult for the user (Table 1). Because of the shortage as well as the intense workload and patient flow during the pandemic outbreak, face masks were allowed to be worn for an extended period longer than that recommended. Nevertheless, a prolonged use can involve a serious of discomfort due to an increase of the facial temperature, both during exhalation and inhalation [65]. Recent studies investigated the possible factors involved in the discomfort increase and in the protection drop due to long-term usage of common face mask by testing particle filtration, breathability, and humidity. Experimentally it has been observed that high breathing resistance and low moisture permeability make a respirator more discomfortable than a surgical mask, especially after many hours of use [66]. Additionally, an increase in temperature and humidity due to moist air expired and inspired in repeated breathing cycles have been observed to promote the growth of bacteria on the face mask, thus putting the most vulnerable people at risk of infection [66,67]. Also, evident drop in humid air filtration efficiency after prolonged wearing of common face mask, with also further proliferation of fungi and bacteria colonies, have been recently reported [68]. A research stated that the use of surgical masks of high quality, like the Type IIR, can be extended to 8 h, and then beyond the recommended time of 4 h, but only under restricting conditions and for that specific brand and standard [69]. These studies highlights that a prolonged use of commercial face mask over the usual recommendation can then involve a considerable discomfort for the combined effect of high humidity and temperature rise on the face but can also expose the user to other potential health risks.
3 Introduction to electrospinning
3.1 Electrospinning as alternative method for the fabrication of face mask filters
The layers forming a commercial face mask are mainly produced by means of Melt-blowing (MB) and Spunbond (SB) methods. Both the techniques are based on the extrusion process of melted polymers in a continuous fluid jet through a spin- hole, which size generally range between 250 and 1000 μm [58]. However, the two methods involve a different process during the collection of the nonwoven fibers. While in the SB process the fluid moving across a quench chamber is cooled by flow air, thus leading to a stretch of the fibers diameter in the range 10–35 μm, in the MB process the polymer fluid is further heated by high-speed hot air thus involving the formation of finer fibers with diameter size ranging between 1 and 10 μm. Since the hot filament spontaneously bond in MB process, no further processing is provided on the nonwoven fiber. Conversely, to improve the bonding in air cooled non-woven fibers, additional treatments, such as thermal, mechanical, or chemical, are carried out in SB method [58]. Because of the smaller diameter size and poor mechanical strength, the MB fibers are usually sandwiched between the SB fabric ones providing filtration performance in commercial face mask. Despite PP is the most popular polymer employed in filters formation, many other viscous thermoplastic polymers, such as polyethylene, polystyrene, polyesters, or polyamides, can be melt blowing [21].
The electrospinning is another flexible technique to fabricate polymer nonwoven fibers and has been attracting great attention from the scientific research community and from industries due to its application in air filter systems [70]. The physical phenomena of the ES dates to the 16th century when William Gilbert for the first time observed the formation of a cone shaped droplet as a consequence of the exposure to the electrostatic field [71]. The behavior of the formation of aerosols of charged liquid have been investigated also by Rayleigh, who obtained an estimate of the maximum charge that a liquid droplet can carry before the liquid jets are ejected from the surface [72]. In addition, Zeleny has conceived a mathematical model to explain the physics of this process [73]. The first patented ES set-up was that of Cooley in1900 [74], which involved the use of many electrodes bringing fiber deposition on a continuously rotating drum similar to that conventionally employed nowadays, but only between the 1930s and the1940s Formhals implemented an alternative set-up, which paved the way to the modern electrospinning techniques [75], [76], [77]. The basic set-up for conventional ES consists of four major components, which include a syringe for containing the electrospun solution, a spinneret which usually consist of a metallic needle with a blunt tip, a high-voltage power supply which can be either direct current (DC) or alternating current (AC), and a conductive grounded collector [78]. During the ES process the syringe pump is used to push a polymer solution at a constant and controllable rate. To obtain the ES solution, the polymer compound is being dissolved by means of a proper organic solvent. When high voltage is supplied between the needle tip and the ground collector, opposite charges will separate within the liquid, due to the electrostatic repulsion among them, thus involving an increase of density charge on the surface of the droplets.
The surface tension will promote the formation of a spherical shape to minimize the electrostatic pressure induced by the external electrical field, which will tend to deform the shape of the droplets. From the syringe a charged jet will flow out initially in a straight line towards the grounded collector following the direction of the applied electric field (Fig. 4 a). As this latter reaches a critical voltage, the surface tension will be exceeded by the electrostatic pressure and an instability of the liquid solution occurs with further deformation of the droplets from spherical to cone shape (i.e., Taylor cone formation) [79] . While the jet extends towards the grounded collector, it undergoes a whipping motion due to the bending instability. While the external perturbation field accelerates the charged jet to move forward the collector, the effect due to the electrostatic repulsion among the opposite charges on the surface of the jet will generate upward and outward forces thus forcing the jet to bend during the motion. Meanwhile, the trajectory is subjected to form a series of loops, resulting in the formation of a coil with several turns. As the distance from the tip increases, the diameter of the charged jet tends to decrease and to continuously stretch. With further elongation the jet solidifies, as a consequence either of the solvent evaporation or the cooling of the melt, forming thin fibers. Finally, the collection of the jet as solid and stable fibers involves the dissipation of the charges through the ground collector. Modifications of the basic set up have been adopted over the years to enhance the fabrication of more complex nanofibers. An example of advancement in standard needle‐based ES methods is given by the introduction of a coaxial spinneret, which allows the fluids from different polymer solutions to be ejected in a coaxial jet thus forming a core−sheath electrospun nanofiber with controlled size [80] (Fig. 4 e). Another example is the triaxial ES method, in which triaxial needles are used to fabricate nanofibers that are structured in three layers [81] (Fig. 4 c). Similar to coaxial, the side-by-side ES technique allows to produce a hierarchical Janus fibers composed of two different sides starting from different polymer solutions [82] (Fig. 4 b).Fig. 4 ES process: (a) Formation of Taylor cone and further stretch of polymer jet under bending instability, adapted with permission from [83]. (b) Side-by-side ES set-up, adapted with permission from [84]. (c) Triaxial ES setup and (d) fiber structure: SEM image of fiber; TEM image of fiber; laser confocal image of fiber. (e) Coaxial ES and SEM image of aligned TiO2; adapted with permission from [85].
Fig 4
As well as for MB membranes, the overall electrospun membranes generally provide low mechanical properties and therefore auxiliary outer layers are provided for face mask applications to make them more resistant. Although several processing parameters, such as air velocity, melt flow index, processing temperature, and orifice size can be tuned to obtain smaller MB based fibers [21,26], the MB fibers developed so far on laboratory scale as filter media for face mask application are characterized by fiber diameter of about the submicron size [23,[86], [87], [88]]. Unlike the MB process, the fibers obtained with ES method can reach diameter size down to 100 nm and a broad range of polymer materials have been successfully electrospun into nanofibers so far [26,89]. A decrease of the fiber diameter to the nanoscale has been observed to contribute in increasing the capture of fine aerosols particles (< 0.5 μm) [62,90,91]. Since the diameter of electrospun nanofiber can be comparable with the free path of air molecules at ambient condition, a slip flow regime occurs around the single nanofiber [26,70,92]. In this regime the fiber's drag force friction is substantially reduced, therefore a low momentum exchange between the fine particles and the fiber surface occurs, resulting in a low pressure drop [93]. Besides, the air streamlines behave in a straight-line fashion getting closer to the nanofiber thus resulting in a higher filtration efficiency due to interception mechanisms [70]. On the other hand, many drawbacks in conventional ES are due to the use of volatile organic solvents, which represents almost the 80% of the solution used for polymer solubility. Some of them are toxic and a small part of them may remain in the nanofiber fabric, without evaporating completely during the process [26] . Instead, many different MB set-up are solvent-free and can also achieve high production rates, thus providing low cost and eco-friendly manufacturing of nonwoven fabric compared to ES [21]. Although ES is a relatively new and promising method for the manufacturing of nanofiber-based filter in application for face masks, new technologies and strategies should be developed to reduce the high cost of hazardous solvent and promote a safer production of nanofiber on large scale.
3.2 Electrospinning design for high production
Because of the long fiber deposition time occurring in conventional ES using a single needle (capacity of ∼ 0.01- 0.1 g/h), several techniques such as multi-needles, needleless, free solvent or melt, and centrifugal ES have been developed to improve the production rate of nanofiber membranes [94,95]. The idea behind the multi-needle is based on conventional ES, but with the utilization of a nozzles array that allow the outflow of several polymer solution simultaneously (Fig. 5 a).Fig. 5 (a) Multi-needles ES set up; adapted with permission from [96]. (b) Needleless ES using spiral coil spinneret; adapted with permission from [94]. (c) Centrifugal ES set-up; adapted with permission from [97]. (d) Umbrella like spinneret and schematic diagram of melt polymer differential ES; (e) formation of multiple melt polymer jet; (f) definition of interjet distance; adapted with permission from [98].
Fig 5
Although better production rates have been observed by increasing the number of nozzles in several ES spinnerets, drawbacks due to either the needle clogging or electric field interference between the close needles arranged on the array may affect the nanofiber productivity [99]. Unlike the multi-needle method, the needleless ES overcomes the shortcoming of clogging, because of the absence of needles use [83] (Fig. 5 b). The spinneret in this case generally consists of a rotating cylinder with large surface area that is partially immersed into the spinning polymeric solution. As a result of the rotation, a continuous production of thin polymer layer occurs on the spinneret surface with the further formation of steady conical spikes. The application of high voltage to the spinning solution intensifies the perturbations, thus leading to the formation of Taylor cones, from which polymer jets are further stretched out to finally results in fibers. In comparison with multi-needle method the needleless ES can produce the highest quality fibers and can have higher rates of production by exploiting different spinneret shapes in the set-up, such as cylinder, ball, disk, coil, and beaded chain [100,101]. Centrifugal ES has proven to be also a promising method to produce ultrathin fiber with large volume and high efficiency compared to other techniques [97,102]. This technique combines centrifugal and electrical forces to manufacture fiber from micrometers to nanofibers scale (Fig. 5 c). Because of both rotating spinneret and applied electric field, the stretching effect on the polymer jet expelled from the nozzle rim is higher than that observed in conventional ES. The combination of centrifugal, viscous, electrical, and gravity forces leads to a greater elongation of the polymer jet with lower bending instability, thus resulting in a better orientation in the formation of fibers [97]. As it can be observed in Table 2 , several synthetic polymer membranes, including Polyacrylonitrile (PAN), Polyvinylpyrrolidone (PVP), and Polyvinylidene fluoride (PVDF), have been electrospun so far by means of needless and centrifugal techniques, providing high-rate productivity compared to the conventional ones.Table 2 A comparison between scale up electrospinning methods to fiber manufacturing.
Table 2Electrospinning Technique Method Polymer1 Solvent2 Productivity Electrospun fiber size Application Refs.
Needleless
Anular spinneret PAN DMF
∼4.5 g/h ∼133- 351 nm Air filtration [100]
Needle roller electrospinner PVA DI ∼12.8 g /h ∼190 nm – [103]
Anular spinneret PVP
PCL
Silk fibroin
PANI /PAA
PAN
PVDF/PEG
DMF/Ethanol/DI
TFA
Trifluororethylene
DMF
DMF
DMF/DI ∼4.8 g/h < 2 μm – [104]
Yarn spinneret PAN DMF ∼1.17 g /h ∼100–117 nm – [105]
Threaded rod spinneret PEO DI and ethanol ∼5–6 g/h ∼100–500 nm – [106]
Linear flume spinneret PAN DMF ∼4.8 g/h ∼108–210 nm – [107]
Mushroom-spinneret. PAN DMF 13.7 g/h ∼100–200 nm – [108]
Bullet spinneret PVA Water ∼1.08–4.55 g/h ∼123–546 nm – [109]
Multi-needle – TPU – ∼50 g/h ∼145 nm Air filtration for PM2.5 [110]
High speed (HSES)
Rotational spinneret SBE-β-CD VOR and water ∼240 g/h ∼0.5 – 2 μm – [111]
Rotational spinneret PVA
DI
∼3.6–6 g/h ∼271–477 nm
Biopharmaceutical [112]
Air-blowing assisted Coaxial – PVP DMF ∼3.6 g/h ∼1–16 μm – [113]
Centrifugal
Single subdisk PS
PVP
Chloroform
Ethanol ∼25 g/h ∼263–8372 nm Mask filter [114]
– PVP
TPU Ethanol
DMF
∼50 g/h ∼2.8–8.9 μm
∼1.8 – 5.4 μm Biomaterial and biomedical [115]
– PEO
PLA Water
Chloroform
∼38.3 g/h
∼12.8 g/h ∼180 nm
∼525 nm – [97]
Melt Differential Centrifugal (MDCE) Centrifugal differential disk PP (Solvent-free) ∼124.3 g/h ∼790 nm – [116]
Melt Differential (PMDES) Umbrella-like spinneret PP/PLA
(Solvent-free) ∼0.3–0.6 kg/h ∼300 nm Biomedical and tissue engineering [98]
Melt 600-nozzle spinneret PLA (Solvent-free) ∼0.9–5.1 kg/h ∼1 μm – [117]
1 Polymers abbrevations: PAN (Polyacrylonitrile); PVA (Polyvinyl acid); PVP (Polyvinylpyrrolidone); PCL (Polycaprolactone); PANI (Polyaniline); PAA (polyacrylic acid); PVDF (Polyvinylidene fluoride); PEG (polyethylene glycol); PEO (polyethylene oxide); TPU (Thermoplastic polyurethane); SBE-β-CD (Sulfobutylether-β-cyclodextrin); PS (polystyrene); PLA (polylactic acid); PP (Polypropylene). 2Solvents abbreviations: DMF (N, N-Dimethylformamide); DI (De-ionized water); TFA (trifluoroacetic acid); VOR (Voriconazole).
However, the use of large amounts of harmful solvents, such as DMF, can require a high cost in recycling, thus limiting the large-scale production of these synthetic membranes. Safer bio-sources, deep eutectic, and ionic liquids solvents have been recently proposed as low toxic solution to allow a proper spinnability and processability for these synthetic polymers, but further investigations should be done to extend their use to pilot scale manufacturing [118,119]. On the other hand, the formulation of alternative solvents, including water and ethanol, with biopolymer materials, such as Polyvinyl acid (PVA), polylactic acid (PLA), and polyethylene oxide (PEO), have been used in both these ES techniques, showing high productivity rate [97,103,106,109,114,115] (Table 2). These polymers are nowadays widely implemented in health industry application, including personal protective-clothing, tissue engineering and drug delivery, due to their biocompatibility and biodegradable characteristics [[120], [121], [122]]. Unlike these solvent based methods, the melt ES is an eco-friendly and solvent free process [95,123]. Given the absence of solvents to solubilize the polymer in the spinning solution, the viscosity of the polymer melt is higher than that usually obtained for standard solution ES with single needle. Therefore, a strong electric field, three or five times higher than that used in conventional ES, is applied to the melt polymer to guarantee the fiber formation. Among the different types of melt ES set-ups designed so far, the needleless melt differential ES proved to be a promising method to start mass production of fiber. Like needleless ES, the needleless melt differential process avoids the use of capillarity needle and make use of very high voltage to produce multiple jets from free melt polymer surface. The set-up is provided with five principal components, including both a melt inlet and distributor, umbellate nozzle spinneret, high voltage power supply, and a receiver plate (Fig. 5 d). Firstly, the melt polymer is being channeled by means of a micro inlet toward a melt distributor to be then transformed from a cylindrical flow into a more uniform ring-like shape one. The further distribution of the flow to umbellate nozzle spinneret allows the formation of a uniform melt thin layer on the circumferential surface. The application of a critical high voltage power involves the formation of polymer multiple jets around the circular edge of the umbellate nozzle and their further ejection in the form of fibers to the receiver plate. An increase in the output of the melt polymer fibers is possible by reducing the distance between the multiple jets (interjet distance) forming around the rim of the nozzle, and thus expanding the number of Taylor cones produced (Fig. 5 e, f). Shorter interjet distance can be obtained by applying high electric fields during the spinning process as well as by lowering the value of the polymer melt viscosity by varying the nozzle temperature [117,124,125]. Despite the high-rate production, the resulting fibers are still large, exceeding 1 μm in size than those fabricated by means of solution ES methods. To address the drawbacks due to the large diameter limitation, nontoxic additives have been blended with polymer to produce submicron fiber [98,126]. Scale up production line for melt ES providing polymers electrospun membrane sheet of width around 1.6 m with 1–10 m/min speed production have been reported in literature [98]. The productivity efficiency of synthetic PP and biodegradable PLA electrospun membranes were comparable with those reported for standard MB lines, which provided high throughput rate around ∼1 kg/h [21]. Also, a scalable set-up for melt Centrifugal ES has been designed with the inclusion of nozzle spinneret to generate multiple jets from the rim disk and enhance the throughput rate of nanofibers [116] (Table 2). A list of international companies that supply several ES equipment based on needleless and centrifugal technologies for industrial production of nanofibers, has been reviewed in other works [127,128]. The advantage of melt ES processes, including the Needleless and Centrifugal ones, is due to a reduction in cost of solvent recovery that brings to a more sustainable manufacturing process of polymer membranes. However, the high processing temperatures ensure that only thermoplastic polymer showing high decomposition temperature can be processed [95]. On the other hand, less toxic solvent-based ES methods can be a solution to high productivity of biopolymer membranes, due to their low impact on environmental waste, life cycle and health assessment [119]. Further assessment on environmental cost and impact for many alternative green solvents is necessary to reduce the risk caused by using hazardous solvents and to allow a more sustainable manufacturing process for large scale polymer membranes.
3.3 Electrospinning parameters and filtration efficiency
ES technique has shown to be a low expensive approach to design nanofiber with controlled size. Many processing ES parameters, including applied voltage, distance between the two electrodes, nozzle tip (needle) diameters, and flow rate, as well as polymer concentration in the solution, can be properly tuned to control structural morphology and orientation distribution of the fibers [[129], [130], [131], [132], [133]]. Among these parameters, the polymer concentration strongly affects the fibers diameter. It was observed that lower viscosity in the ES solution involves the formation of finer fibers in the electrospun membrane [[134], [135], [136], [137]]. A reduction of the fiber size in the range of nanofiber is required to obtain high filtration efficiency against submicron particles [[138], [139], [140]]. Indeed, the probability of fine aerosols to impact the ultrafine fibers increase due to the high surface area, whereas small interstitial sites between the fibers improves the ability of the membranes to remove particles bigger than the pores size because of the sieving effect [141]. Also, an increase of the basis weight for a given surface or thickness obtained with a longer deposition time during the ES process has proven to be a good strategy to improve filtration efficiency of the filters [[142], [143], [144]]. However, the use of thicker layers as well as that of small fibers can negatively affect the air permeability of the fabric resulting in an increase of the pressure drop with consequences on the breathability. Therefore, the overall performance of the filter can be estimated by means of the calculation of the quality factor (QF), which takes into account both the pressure drop (ΔP) and the filtering efficiency towards particles (η) [141]:(1) QF=−ln(1−η)ΔP
Better filter performance can be obtained for high QF values, which means to keep low pressure drop and high particles filtering efficiency. One strategy to obtain effective filtration is that to fabricate electrospun fibers with high interconnectivity or porosity. An enhanced porosity can involve a proper airstream across the electrospun membrane with a significant reduction in the pressure drop, without affecting the ability of capture of the nanofibers against submicron particles [62,141].
4 The use of electrospun nanofibers in surgical mask and respirator applications
4.1 Improvement of the face mask's performance by using nanofibers
In recent years many synthetic and thermoplastic polymers such as PVDF, PAN, PP, Polyimide (PI), and polystyrene (PS), have been electrospun in fibers to improve the prevention of infection from expiration of airborne contaminants [4,18,27]. A reduction of the fiber diameter as well as of the interstitial space between the fibers has been observed to lead to a significant improvement in mechanical adsorption of fine particles in several polymer based electrospun membranes [[145], [146], [147], [148]]. Ultrathin PVDF-g-POEM electrospun fibers exhibited higher performance in capturing NaCl particle (≤ 0.3 um) compared to PVDF nanofibers with larger diameter (Fig. 6 a) [146].Fig. 6 (a) Comparison between the filtration efficiency measured for PVDF and amphiphilic PVDF-g-POEM double comb copolymer for different pressure drop values; adapted with permission from [146]. (b) Filtration efficiency values measured for PAN electrospun fibers as function of different average diameter; adapted with permission from [148]. (c) Filtration efficiency measured for PAN10%−1%Ag (1.50 g/m2), Respirator Face Mask and Three-Ply surgical as function of aerosol particle size; adapted with permission from [147]. Diameter statistics of (d) PAN, (e) PAN/ (Graphene oxide) GO, and (f) PAN/GO/PI-6 nanofibers; SEM images of (d-1) PAN, (e-1) PAN/GO, and (f-1) PAN/GO/PI-6; SEM images and illustration of (d-2) PAN, (e-2) PAN/GO, and (f-2) PAN/GO/PI-6 nanofibers after PM2.5 adsorption; (g) pore size distribution measured for PAN, PAN/GO, and PAN/GO/PI electrospun membranes; adapted with permission from [149]. (h) Filtration efficiency and pressure drop measured for multilayer PE/PA with different PA electrospun membrane thickness; SEM images obtained for (i) PE meltblown nonwoven membrane and (j) PA electrospun one after filtering the PM; (k) water vapor transmission rate test measured for the PE/PA and two different commercial face masks, namely com-1 and com-2; adapted with permission from [150].
Fig 6
When the small fibers and the mean free path of air become comparable, a slip flow regime occurs for the nanofibers, resulting in a low pressure drop for the electrospun membranes as well. Moreover, from an analysis of the filtration test performed on PAN nanofibers prepared at different fiber size, a significant drop in both particle capture of PM2.5 (particle size < 2.5 μm) and pressure drop has been observed and attributed to the presence of large diameter fiber and wide pore size in the electrospun membranes (Fig. 6 b) [148]. Additionally, small pore size in electrospun membranes improved the capture of particles of greater dimension because of the sieving effect. The measures of minimum efficiency values obtained from particle filtration curves performed on several PAN nanofibers have proven to be comparable with the average pore size, thus indicating that particles slightly smaller in size were the most penetrating in the filters [147]. Nevertheless, because of the small inter-fiber space, the least efficiency measured for PAN electrospun membranes were shifted to a lower particle size range, providing a better interception of submicron particles compared to those observed for commercial filters used in face masks (Fig. 6 c). Also, Dai et al. reported that a decrease in pore size leads to a better capture of aerosols particles [149]. From an analysis of the Scanning Electron Microscopy (SEM) images performed on the PAN, PAN/ Graphene oxide (GO), and PAN/GO/PI membranes they noted that within the same adsorption time the amount of PM2.5 particles, present with the shape of beads, were more uniformly adsorbed on the nanofibers composed of PAN/GO/PI and in greater amount than those found on PAN (Fig. 6 d–f). The best interaction occurred between nanofibers and submicron particles has been attributed to the uniform inter-fiber space distribution observed at 0.5 μm, which was lower in size and narrower compared to those observed for other PAN and PAN/GO electrospun membranes (Fig. 6 g). In Table 3 a summary of the performances obtained from recent electrospun membranes-based face masks is provided for a comparison.Table 3 Performance of electrospun filter for face mask application.
Table 3Tested aerosols particles 1Electrospun fiber composition PFE BFE Pressure drop Quality Factor (QF) Electrospun fiber size
(Pore size) Refs.
(%) (%) (Pa) (Pa−1)
NaCl Nylon-6 onto PE MB nonwovens
(Face velocity 5.33 cm s−1) >99 – <100 ∼0.0486 ∼126 nm (6 μm) [150]
PMMA-EVOH
onto spunbonded PP ∼99 – ∼44 – ∼319 nm [151]
PVA
(Needleless electrospinning) ∼99 – ∼78 ∼0.0593 ∼273 nm [152]
PAN/PVDF-PDMAEMA <96 – ∼78 ∼0.398 ∼190/525 nm [153]
PVDF-Si NPs2 99% – ∼392 ∼0.02 ∼1 μm [145]
PVDF/TTVB3 ∼99 – ∼350 – ∼479 nm [154]
PVDF-g-POEM
(Face velocity 5.3 cm s−1) >93% – – ∼0.06 ∼77 nm [146]
(ABC)-type terpolymer
(Face velocity 5.3 cm s−1) ∼99 – ∼177 ∼0.025 ∼400 nm [155]
PLA ∼95 – – ∼0.014 ∼630 nm [156]
PLA ∼99 – ∼104 ∼0.094 ∼37.4 nm (0.73 μm) [157]
PA6/PVP/CS ∼99 – ∼54 ∼0.118 ∼130 nm [158]
CA-TiO2 ∼99 – ∼31 – ∼278 nm [159]
PP/PLA-Ag NPs
(Face velocity 10.67 cm s−1) ∼99 – ∼105 ∼0.048 ∼820 nm (5 μm) [160]
PS/Ag NPS >97 – <147 – ∼ 3 μm- [161]
PAN/Ag NPs ∼99 – ∼65 ∼0.15 ∼320 nm (0.3–0.9 μm) [147]
PAN/Ag NPs/Bontioides/g-C3N4
(Face velocity 0.29 m s−1) ∼99 – ∼65 ∼0.097 ∼727 nm [162]
PVA/PEO/CNF/N-TiO2 ∼98 – – – ∼0.79 μm [163]
Recycled PET >98 – ∼36 – ∼1.2 μm [164]
Recycled EPS
(Face velocity 5.3 cm s−1)
(Needleless Electrospinning) ∼99 – ∼48 ∼0.099 ∼1.04 μm [165]
PVA/AG NPS
(Face velocity 5.5 cm s−1) 97.7 ∼59 ∼0.09 ∼434 nm [166]
PVA/WS-CS4
(Needleless electrospinning) ∼97 – ∼57 ∼0.0825 ∼217 nm (12–22 nm) [167]
PVA/SBE-βCD
(Face velocity 5.3 cm s−1) ∼99 – ∼57 ∼0.82 ∼2.26 μm [168]
PAMAM/PAN/TEO
(Needleless Electrospinning) ∼98 – ∼388 – 440 nm [169]
Murine hepatitis virus A59 (MHV-A59) used in NaCl PVDF-RB5 ∼99 – ∼40 – ∼200 nm (1.5–2.0 μm) [170]
Polystyrene latexsphere (PSL)
Virus surrogate used in PSL:influenza A (H1N1);coronavirus 229E;SARS-CoV-2(Delta variant); PLA and Manuka oil >99
∼99
∼99
∼99
∼99 <58 – ∼168 nm [171]
– PS/PVDF ∼99
– ∼72 – – [172]
– PLA & phytochemical-based herbal-extracts
(Needleless Electrospinning) – ∼97 ∼35 ∼0.097 ∼8 μm (20 μm) [173]
– PVB-Thymol ∼83 ∼99 ∼46 – ∼375 nm [174]
– Nylon 6
(Needleless Electrospinning) – ∼97 – – 110–400 nm (0.6 μm) [175]
Dioctyl sebacate (DEHS) CA/TPU-LiCl ∼99 – ∼52 ∼0.12 ∼280 nm (0.9 μm) [176]
PVA-TA
∼99 – ∼35 ∼0.15 ∼430 nm [177]
PVA-LS ∼99 – ∼24 ∼0.212 ∼439 nm [178]
PM0.3 PAN >97 – ∼ 50–500 – ∼430 nm (0.5–1.2 μm) [179]
PAN/ZnO NPs ∼98.8 – ∼48 ∼0.092 < 420 nm [180]
CA/AC/TiO2
(Face velocity 0.8 m s−1) ∼82 – ∼63 ∼0.0271 – [181]
PAN/FPVDP@CNTs6
(Face velocity 5.33 m s−1) ∼99 – ∼49 ∼0.1984 ∼113 nm (1.5 μm) [182]
Gelatin/β–Cyclodextrin
(Face velocity 6 cm s−1) >95 – ∼148 ∼0.029 ∼130–247 nm (1–1.4 μm) [183]
SP/PS >95 – <343 – – [184]
Zein on cellulose paper towel 99% – 109 – 0.6–2.4 μm (0.2–0.3 μm) [185]
PM1.0 PLA ∼98 – ∼29 – ∼274 nm (2.5 μm) [186]
PLA-PA11 ∼89 – ∼6.5 – ∼76 nm [187]
PAN/silane/ZnO NPs ∼99 – ∼44 ∼0.137 1 < μm (19 nm) [188]
PM2.5 PAN/PI/GO7 ∼99 – ∼92 ∼0.0576 ∼244 nm (0.5 μm) [149]
Recycled PET
(Face velocity 4.8 cm s−1) ∼100 – ∼212 – ∼1.27–3.25 μm [189]
PAN/CTAB8 >99 – ∼11 ∼0.469 ∼150–200 nm (2–4 μm) [148]
PASS/Ag/ZnO ∼99 – ∼42 ∼0.07 ∼330 nm (2 μm) [190]
1 Electrospun fibers abbreviations: PE (Polyethylene); PMMA (poly(methyl methacrylate)); EVOH (ethylene vinyl alcohol); PVA (Polyvinyl acid); PAN (Polyacrylonitrile); PVDF (Polyvinylidene fluoride); PDMAEMA (poly [2-(N,N-dimethyl amino) ethyl methacrylate]); g-POEM (graft-poly(oxyethylene methacrylate); (ABC)-type terpolymer (Poly(styrene-co-2-(dimethylamino)ethyl methacrylate-co-acrylonitrile))); PLA (polylactic acid); PA6 (polyamide 6); PVP (Polyvinylpyrrolidone); CS (Chitosan); CA (cellulose acetate); PP (Polypropylene); PEO (polyethylene oxide); CNF (cellulose nanofibers); PET (polyethylene terephthalate); Polyarylene sulfide sulfone (PASS); EPS (expanded polystyrene); SBE-β-CD (Sulfobutylether-β-cyclodextrin); PAMAM (polyamidoamine dendritic polymers); TEO (tea tree essential oil); PS (polystyrene); PVB (polyvinyl butyral); CA (cellulose acetate); TPU (Thermoplastic polyurethane); TA (Tannic acid); LS (lignosulfonate); AC (activated charcoal); FPVDP (fluorinated polyvinylpyrrolidone); SP (Spiropyran); PS (polystyrene); PA11 (Polyamide-11); PI (Polymide).
2 NPs (Nanoparticles).
3 TTVB (organic photosensitizer).
4 WS (water soluble).
5 RB (Rose Bengal).
6 CNTs (Carbon nanotubes).
7 GO (Graphene oxide).
8 CTAB (Cetyltrimethylammonium bromide-Br−).
As it can be seen in Table 3, the measurements performed so far on several natural and synthetic polymer-based electrospun filters demonstrate that in addition to small fiber size also the area of the pores results to be lower compared to those reported for commercial face mask [[191], [192], [193]]. These results therefore indicate that the optimization of both fibers and inter-fiber space in electrospun membranes can be beneficial to achieve proper nanofibers filters with high filtration effectiveness to remove submicron particles. The application of electrospun fibers as filter to improve performance in commercial face mask has been also investigated. The inclusion of PS/PVDF nanofiber membranes inside a N95 respirator has been observed to provide higher filtration performance compared to the standard one, showing also low pressure drop at ∼ 72 Pa, well below the maximum limit imposed by the NIOSH standard (Table 1) [172]. On the other hand, the application of electrospun PAN/ based microporous carbon nanofibers (MCNFs) in face masks proved a significant reduction in facial temperature recording during exhalation and inhalation test, thus making them more comfortable during the breathing compared to a common mask [194]. Combining electrospun fibers and melt blowing fabric with different pore and fiber morphology in masks turned out to be an excellent way to fabricate new face masks with excellent filtration performance and more comfortable wearability. From a comparison between the filtration curve measured for the pristine MB based polyethylene (PE) membranes and those obtained for multilayer designed at different thickness of electrospun Nylon-6 (PA) fibers, it is shown that the addition of ultrathin PA significantly improved the capture of fine particles in the multilayer PE/PA [150] (Fig. 6 h). This analysis was in accordance with what was observed in the SEM images performed on PA and PE membranes, where an evident capture of low particle size prevalently occurred in the PA membranes due to the smaller fiber and micro-pore size (Fig. 6 i, j). A similar result has been reported in the work carried out by Yuanqiang Xu et al. where a higher capture of fine particles with size below 0.5 μm was attributed to the addition of electrospun nanofiber with smaller inter-fiber space on MB-based PE membrane, which was made instead of larger pores [160]. Furthermore, experimental evidence based on water vapor transmission rate (WVT) revealed that multilayer membranes based on electrospun and MB fabric provided higher values of WVT compared to that observed in commercial face masks, thus proving a better passage of moisture air during exhalation, and therefore a less human discomfort in wearing the mask [150,160] (Fig. 6 k). Also, the measures obtained from the air permeability test, which is related to the pressure drop of the fabric and is defined as the rate of the air passing through the membrane under a certain pressure, were found to be comparable and in accordance with the standard required for commercial face masks, thus ensuring a proper breathability in the multilayer fabric [150]. These results indicate that a combination of micro and nanofibers in a multilayer membrane can involve a better breathability over commercial face mask, while providing an effectiveness capture of PM. It is worth noting that electrospun of thin polymeric nanofibers on thicker MB membranes allow to address the drawback due to the poor mechanical properties. Furthermore, the addition of nanofiber with sub-micron pores structures in MB based face mask can face the issues of discharge dissipation in a high humidity exposure for used charged fibers and ensure a more stable filtration performance, without the need of further electrostatic treatment for the membranes.
4.2 Eco-sustainable polymer materials for nanofiber masks
The polymers employed for membrane manufacturing are non-degradable and even worse their production as well as their disposal is not sustainable environment-wise, in most cases [195]. The outbreak of pandemic has revealed these environmental issues, given the large amount of plastic pollution due to the improper disposal of discarded masks [[196], [197], [198]]. It was estimated that just the 1% of the total disposed masks, which is around 10 million of mask for month, corresponds to a plastic weight ranging between 30 – 40 tons dumped in the environment [199]. The plastic fibers composing the inner and outer layers of waste masks, such as PP and PE, can be broken down in smaller parts, also defined as microplastics (MPs), and remain in the environment for a very long time [200]. Besides, these MPs can further decompose due to a series of potential aging process, including physical abrasion, ultraviolet (UV) radiation, and chemical oxidation, and so release toxic additives as well as provide attachment points for microorganism with negative consequences for ecosystem and human health [201]. The enhancement of the ecotoxicity of the MPs released in open sea, which is provoked by the continuous accumulation of plastic coming from disposable masks at the landfills, is rising up nowadays a global concern [202,203]. Therefore, the urgent need to implement more sustainable development has become a relevant issue to address in the increasing demand of face masks [156]. To reduce the environmental impact due to the massive use of disposable masks, some researchers explored the fabrication of high efficiency filter material by employing recycled polyethylene terephthalate (PET) [164] (Fig. 7 a).Fig. 7 (a) Schematic diagram of a face mask obtained by bottle recycled PET; adapted with the permission from [164]. (b) Example of nonwoven base with multiple EPS fiber layers for air filter; and (c) SEM image of the membrane cross-section, indicating the three different fiber sizes; adapted with the permission from [165]. (d) Filtration performance measure for the three-layer face mask based on PLA electrospun membranes at different relative humidity and the (e) long-term filtration stability test carried out, under 90% humidity; (f) weight loss measure for the for the three-layer face mask based on PLA electrospun membranes after being buried in outdoor soil; (g) images taken to the three-layer face mask based on PLA electrospun membranes showing the soil burial degradation within the 150 days; adapted with permission from [157]. (h) SEM image showing the bacteria growth on PLA surface after 20 days of treatment in cow-dung, jaggery and water containing biodegradation slurry; adapted with permission from [173]. Biodegradability tests performed on the CA electrospun based membranes by employing esterase and cellulase enzymes: (i) Polypropylene fibrous membrane used as control, given its non-susceptibility to degrade under enzymes conditions; (j) CA nanofibers tissue; (k) CA nanofibers tissue processed with tetrabutylammonium bromide (TBAB); and (l) CA nanofibers tissue processed with High-foaming HoneySurf (HS); adapted with permission from [204].
Fig 7
Investigation on ES of recycled PET filter have been carried out with satisfactory results on PM removal efficiency and low pressure drop [189]. The latest works have shown how filters based on PET waste nanofiber can be achieved by optimization of the ES parameters and can additionally be redissolved and reprocessed, without losing their performance in filtration [164]. The use of recycled expanded polystyrene (EPS) for the manufacturing of multilayer electrospun filter mask with different fibers size provided promising performances for both submicron particles filtration (∼ 99.4%) and optimal pressure drop, ranging between 48 and 58 Pa [165] (Fig. 7 b, c). Therefore, the feasibility in using recycled materials such as PET and EPS for the fabrication of face mask could be considered to implement a circular economy and a significantly lowering in the environmental impact in terms of resources consumption and environmental pollution for future policies [205,206]. Also, the utilization of renewable and biodegradable thermoplastic polymers, such as PLA and polyvinyl butyral (PVB), has shown to be a sustainable alternative to produce effective filters comparable to many commercial face mask [156,157,173,174,207]. Wang et al. designed a biodegradable face mask entirely formed of PLA polymer [157]. The assembly of nanofibers layers with different diameter size, provided a multi-scale structure in the membrane with high porosity and small pore size, which was able to maintain high filtration performance against PM0.3 also after a prolonged exposure in high humidity environment (Fig. 7 d, e). Besides, from a study on the weight loss in degradation (WL), they observed that the PLA-based face mask was completely decomposed from microorganism present in soil after being buried in for 150 days (Fig. 7 f, g). A similar result in biodegradation performance has been reported also by Patil et al. [173]. The SEM images performed on the 3-ply cotton-PLA-cotton layered face mask after 20 days of pretreatment with a slurry consisting of fresh cow-dung microflora, revealed an extensive growth of bacteria colonies on the PLA nanofiber surface, thus suggesting an effective biodegradation process occurring on the membrane (Fig. 7 h). The filtration performance of green PLA face masks result to be comparable with that obtained for multilayer membranes of MB and synthetic electrospun nanofibers seen above [150,160] (Table 3). On the other hand, the biodegradable PLA-based face mask can be easily decomposed trough biological process in a reasonable time, thus resulting in less pollution for the environment. In addition to synthetic and biodegradable polymers, the cellulose nanofibers have been receiving attention as emerging raw materials because of their abundant availability and for their properties, such as biodegradability, renewability, high mechanical strength, porosity, etc. [[208], [209], [210]]. Recent works have focused on the ES of polymers and cellulose blends for the development of filter materials, reporting promising performances for face mask/ respirators [158,211,212]. Additionally, the cellulose acetate (CA) has shown to be an eco-friendly alternative to petroleum-based non-biodegradable polymers for the fabrication of electrospun nanofiber filters [159,181,213]. Wang et al. investigated the application of CA by means of ES technique to design new biodegradable filter for face mask [176]. The electrospun membrane based on CA, thermoplastic polyurethanes (TPU), and lithium cloride (CA/TPU-LiCl) showed high filtration efficiency (99.8%) against dioctyl sebacate (DEHS) aerosol micro particles with low pressure drop (∼ 52 Pa). To test the reusability, a standard disinfection method involving alcohol has been performed on the biodegradable CA/TPU-LiCl membrane. The disinfection test revealed a high resistance of the electrospun filter to alcohol soaking, such that after ten repeated cycles both the filtration performance and the pressure drop remained quite stable around 98.2% and 34 Pa, respectively. The degradation of CA can occur in the soil surface as well as by means of microorganism and enzymes treatments, but the rate is strictly dependent on the environmental conditions in which the natural polymer is subjected [214]. A recent study carried out by Samadian et al., demonstrated that the inclusion of berberine in a CA/gelatin based electrospun nanofibers enhanced the degradation rate of the membrane, which resulted in a high weight loss around ∼ 80% after only 14 days of exposition in PBS solution [215]. Also, Oldal et al. observed that CA electrospun nanofibers membrane can be easily degraded by means of natural enzymes, such as cellulase and esterase [204]. Compared to pristine CA nanofibers, a faster response in the degradation process has been observed for CA membranes pretreated in solution with green surfactant (Fig. 7 i–l). The use of tetrabutylammonium bromide (TBAB) and High-foaming Honey Surf (HS) made the access to the polymer chain of CA-based membrane easier for both enzymes, thus improving the biodegradation process in the observed elapsed time of 16 h. These results indicate that electrospun membranes based on natural polymers, such as PLA and CA, can be a valid solution to fabricate more sustainable and high-performance face mask, to relieve the high consumption of petroleum-derived plastics polymers, which is used in the manufacturing of disposable commercial face mask, and to prevent possible effect on environmental hazard in the long run.
4.3 Nanoparticles surface modification for better antibacterial and antiviral electrospun face masks
Face masks provide a useful physical resistance to the transmission of aerosol carriers of either viruses or bacteria, but a high risk of contamination occurs on the fabric for long time usage. Considering therefore the current COVID-19 pandemic, the importance of providing stable self-cleanable filters with strong antibacterial and antiviral activity has led to the development of nanoparticles (NPs) embedded electrospun nanofibers, in the past few years. The ES of hybrid materials has shown to be a straightforward and innovative approach to process and stabilize NPs either onto or into the electrospun nanofibers membranes. The antimicrobial effectiveness of metal NPs, such as Au and Ag, as well as metal oxide NPs, such as, TiO2, CuO, ZnO, and MgO, have been studied with several human pathogens bacteria such as Escherichia coli (E. coli) and S. aureus [[216], [217], [218], [219], [220]]. Most of these inorganic NPs exhibit bactericidal properties either through photocatalytic activity, where reactive oxygens species (ROS) are induced by visible or UV light to affect the cell viability by hindering principal mechanisms of protein and enzymes, or by electrostatic interaction with the bacteria cell wall, where NPs bind electrostatically the cell membranes causing their alteration in both potential and depolarization thus involving respiratory disfunction and eventually the cell death [221]. Several studies reported about the efficacy of ZnO as antimicrobial coating for surfaces exposed to possible contamination of either bacteria or SARS-COV-2 [[222], [223], [224], [225]]. It is well known that both Zn and ZnO are of particular interest as supportive treatment in therapy of COVID‑19 infection [[226], [227], [228], [229], [230]]. The fabrication of several based polymers electrospun embedded with low-cost ZnO NPs have been investigated and proposed for protective clothing application. Nanofibers made of Polyvinyl pyrrolidone (PVP) and PVA with the addition of ZnO NPs showed antimicrobial activity against S. aureus, E. coli, Klebsiella pneumonia and Streptococcus aeruginosa tested bacterial strains [231]. The incorporation of ZnO NPs at 10 wt.% in the poly- L- lactic acid (PLLA, or polylactide) electrospun membrane was observed to improve the mechanical properties as well as the bacteriostatic action against S. aureus [232]. A noticeably enhancement of the mechanical strength occurs also for PVDF nanofiber loaded with 5 wt.% of ZnO NPs, since an elongation-at-break equal to (30 ± 2)% was observed compared to the pristine PVDF nanofibers, which owns (24 ± 0.5)% [233]. Cytotoxicity test proved the nontoxicity for 5% ZnO—NPs@ PVDF nanofibers. Besides, for these NPs loading the hybrid PVDF nanofiber exhibits a high antiviral activity against human adenoviruses type-5 (ADV5) in both the adsorption and virucidal mechanisms; therefore, the fabric has shown to easily prevent both the entry and the replication of the virus in the cells. An activity increase against colistin resistant bacteria, such as K. pneumoniae strain 10, has been observed for thermoplastic polyurethane (TPU) nanofibers with 4% ZnO [234]. Moreover, it has been shown that a higher presence of ZnO on the nanofibers membrane involves a lower ability of the SARS-CoV-2 spike protein to engage with the human cell receptor (ACE2). Therefore, the concentration of the metal oxide in the TPU nanofibers resulted to be crucial to inactivate the virus. Abdul Salam et al. [235] have investigated both the antibacterial and antiviral properties of hybrid PAN nanofibers with 5 wt.% ZnO as function of the HeiQ Viroblock (VB) concentration. They observed that the higher the amount of VB in the PAN/ZnO electrospun the higher is the antibacterial efficiency of the hybrid membrane against both S. aureus and Pseudomonas aeruginosa, with efficiency values equal to 92.59% and 88.64% respectively. Besides, the PAN/ZnO sample loaded with the maximum concentration of VB (5%) showed a significant antiviral activity against avian influenza virus compared to the pristine PAN sample. The antiviral mechanism behind the VB technology has been attributed to the combined action of the vesicle component and the inside silver ions. Indeed, the cosmetic grade liposomes forming the vesicle are expected to weaken the envelope membrane of the influenza thus allowing the silver ions to directly attack the inner core to destroy the virus (Fig. 8 a).Fig. 8 (a) Schematic picture of antiviral activity of VB-loaded PAN/ZnO electrospun nanofibers; TEM analysis of VB-loaded ZnO/PAN electrospun nanofibers: (b) pristine PAN nanofibers; (c) ZnO/PAN nanofibers; (d) 2.5% VB-loaded ZnO/PAN electrospun nanofibers; (e) 5% VB-loaded ZnO/PAN nanofibers; adapted with permission from [235]. (f) Schematic diagram modification process for of ZnO NPs by means of the reaction with silane coupling agent MPTMS; (g) comparison between the QF measured for PAN nanofibers obtained at 9 wt% (PZ9) and 12 wt% (PZ12) for different silane/ZnO NPs mass ratio; adapted with permission from [188]. (h) TEM images obtained for PASS/ZnO/sulfide electrospun membrane showing both agglomerated (left) and more uniform NPs (right) distribution on the fiber surface; adapted with permission from [190]. (i) Filtration curves measured for PS electrospun membranes fabricated at different Ag concentrations and tested at different aerosol particle size; adapted with permission from [161].
Fig 8
From the Transmission Electron Microscope (TEM) analysis performed on the PAN/ZnO samples fabricated at different concentration of VB is possible to observe the distribution of ZnO NPs on the nanofiber surface as well as the roughness of the single fiber due to the presence of VB (Fig. 8 b–e). As can be seen in Fig. 8 d, the surface of the nanofibers become more rougher by increasing the VB doping concentration in the electrospun solution, thus indicating that a greater amount of this viscous liquid being evenly deposited on the nanofiber matrix. These studies revealed that the presence on the fiber surface of ZnO NPs alone or with other active materials can be effective in preventing the spread of bacteria and viruses on the electrospun membranes. Their application as filters would allow to address the issues of bacteria proliferation observed in commercial face mask after prolonged use by removing possible contamination, which can represent a health risk especially for severe disease cases [66,68]. It is worth noting that the morphological changes occurring at the fiber surface level due to the inclusion of NPs materials also significantly affect the PM capture efficiency in electrospun membranes, in addition to the antibacterial and antiviral activity. Hanaa Aamer et al. observed a better filtration performance for PAN nanofibers functionalized with ZnO NPs compared to the pristine membrane ones [180]. They stated that the addition of ZnO NPs induced an increasing in surface roughness, with consequent reduction in stream of PM0.3 through the membrane. In another work, they reported that a modification of the ZnO NPs obtained by means of the reaction with 3-methacryloxypropyltrimethoxysilane (MPTMS) led to a more uniform distribution of NPs on the PAN nanofibers surface, thus resulting in a better capture of both pathogens and aerosols contaminants (Fig. 8 f, g) [188]. On the other hand, a high content of ZnO NPs in polyarylene sulfide sulfone (PASS) electrospun membrane previously functionalized with Ag was reported to induce a gradually increase in mean pore size [190]. The addition of 2 wt% ZnO NPs provided a maximum filtration performance in PASS membrane, whereas with a further increase at 3 wt% the structures resulted in larger pore size that exceeded PM2.5 size and brought to a drastic drop in particle capture efficiency. From the TEM image in Fig. 8 h is clear that the formation of two distinct distribution of NPs on the fiber surface, one more discontinuous with large and extended agglomerations and the other more evenly distributed, may explain the different overall filtration behavior occurring as consequences of the different contents of ZnO NPs added in the PASS membrane Moreover, the integration of Ag NPs proved to significantly increase the fibers diameter in PS electrospun membrane [161]. However, the even surface roughness formed from the addition of Ag NPs brought to a slighter increase in particle filtration and a low decrease in pressure drop, thus resulting in an enhancement in performance for the PS membrane (Fig. 8 i). Chen et al. observed that despite the addition of M.bontioides and Ag-CN compound made the fiber larger, the even increase of the roughness occurred at the fiber surface level led to an enhancement of the contact with both pathogens and aerosols contaminants thus involving a higher interception for tiny NaCl particles with size around 70–80 nm as well as a strong inhibitory activity against E. coli, S. aureus, and influenza A H3N2 [162]. The large amount of Mg, Ag, C, N, and O elements observed by the analysis of the SEM-EDX (Energy Dispersive X-ray Analysis) clearly indicated that M. bontioides and Ag-CN NPs were dispersed uniformly during the ES process (Fig. 9 a, c).Fig. 9 (a) Electrospun solution for the fabrication of the PAN/M.bontioides/Ag-CN/Ag nanofibrous membrane; (b) TEM image of PAN/M.bontioides/Ag-CN/Ag nanofiber; SEM image of the PAN/M.bontioides/Ag-CN/Ag nanofibrous membrane (c-1) and with respective EDX analysis performed for C (c-2), N (c-3), O (c-4), Ag (c-5), Mg (c-6); adapted with permission from [162].
Fig 9
This result was also in accordance with the TEM image in Fig. 9 b, which revealed the presence of heterogeneous Ag-CN NPs aggregates both above and inside the single nanofiber surface. As seen so far, a proper optimization of the ES process allows a stable and uniform distribution of NPs as well as for other active materials at single fiber surface level. This method can be beneficial compared to other recent approach adopted to functionalize commercial face mask, which may result instead in a coating instability for the surface functionalization over the time with an undesired effect on antimicrobial and antiviral properties [[236], [237], [238]]. Additionally, recent studies proved that several NPs materials with antimicrobial and antiviral properties, including MgO and CuO, are compatible when embedded in electrospun membranes based on natural polymers, such as CA and poly(ε-caprolactone) (PCL) [239,240]. Because of the incorporation of MgO NPs, the PCL-based electrospun showed a pronounced antibacterial activity against both E. coli and S. aureus as well as a remarkable improvement in filtration efficiency against polystyrene particles of 1 μm size [239]. Also, the addition of CuO in CA based electrospun membrane provided similar result in terms of antibacterial efficacy [240]. However, only the addition of CUSO4 metal salts and thymol caused a drastically reduce of SARS-COV-2 virus, observed only after 1 h of contact with the functionalized CA membrane. Hence, the development of antiviral and antimicrobial face mask obtained from biodegradable materials could be also advantageous to prevent the transmission of SARS-COV-2 virus, as well as other potential hospital infections caused by bacteria, and to further reduce the impact on the environment, which is already put at risk due to the huge use of disposable mask. The functionalization of electrospun membranes by the addition of relatively nontoxic and low-cost NPs as well as other active compound can therefore be an affordable method to produce advanced electrospun membranes in view of large-scale production. These electrospun filters can provide desirable properties in terms of both particle filtration and antibacterial effectiveness, and significantly reduce the risk due to the handling and reuse of contaminated mask.
4.4 Electrospun nanofiber for self-powered wearable masks application
Electrostatic nanofibers are widely utilized in air filtration application. Polymer-based nanofibers with spontaneous dipole moment or optimal polarity provide an efficient electrostatic surface charge, which promotes the capture of opposite charged particles during the process of air filtering [145,146,148,172,184,241,242]. Experimentally it has been observed that an exposition of electrostatic nanofibers to external electric field involves a better capture of submicron particles than uncharged nanofibers, which rely only on interception and diffusion mechanical capture [243,244]. In addition to electrostatic induction, a large polarization and optimal formation of electrical charge have also been successfully achieved either by exposing polymers with piezoelectric property to mechanical stress [245] or through rubbing adjacent polymer membranes with opposite electronegativity, thus resulting in triboelectric effect on the interface [246]. Because of the high capacity to retain electrical charge, active polymer nanofibers have been used as triboelectric nanogenerators (TENGs) and Piezoelectric nanogenerators (PENGs) materials to develop energy harvesting air filters [[247], [248], [249], [250]]. So far, several Respiration-driven TENGs (R-TENGs) designed with polymer-based electrospun nanofibers, including PVDF, PI, and PAN, have provided an effective electrostatic adsorption of submicron particles by harvesting electrical energy from human breathing [[251], [252], [253], [254]]. The triboelectrification mechanism behind a R-TENG is based mainly on the contact-separation mode occurring between two different materials, showing distinct electron affinity [255]. To explain the working principle, a schematic diagram of the R-TENG fabricated by Fu et al. with PVDF electrospun membrane and cellulose aerogel/MOF (CA/Ni-HITP) composite is reported in Fig. 10 a, b [251].Fig. 10 (a) Self-powered air filter for respiratory monitoring and removal of submicron particles; (b) schematic diagram of the triboelectric effect of R-TENG; (c) short-circuit current and transferred charges of R-TENG before and after in situ growth of conducting metal organic- framework (Ni-HITP); (d) a comparison of PM removal efficiencies of commercial mask, CA, CA/Ni-HITP, CA/Ni-HITP+PVDF and self-powered air filter obtained at different submicron particle sizes; adapted with permission from [251].
Fig 10
Because of the tight contact occurring between the two layers, electrostatic charges of opposite sign can be induced by triboelectrification on the respective surfaces. This process can be seen in Fig. 10 b-(i), where a possible case of air exhalation during breathing can lead to the contact between the two materials, thus involving the generation of a negative triboelectric charge on the high electronegative PVDF based membrane and a positive one on the CA/Ni-HITP. When the two layers are separated, for instance during a further air inhalation, a potential difference is generated due to the electrostatic induction and a transfer of charges occurs form PVDF to CA/Ni-HITP (Fig. 10 b-(ii)). This process occurs as long as the distance between them does not induce a neutralization of the charges, which would then result in an interruption of the current (Fig. 10 b-(iii)). After exhaling, the layers get closer again, thus inducing a positive charge on CA/Ni-HITP which balances the former negative triboelectric charge present on PVDF. This results in a reverse current, which is detected from the CA/Ni-HITP to PVDF (Fig. 10 b-(iv)). Therefore, the air pressure exerted during a continuously breathing cycle led to an alternate current, which constantly can charge the membranes (Fig. 10 c). From a comparison of the quality factors obtained for CA, CA/Ni-HITP, CA/Ni- HITP+PVDF, commercial face mask, and R-TENG, it can be noted that due to the triboelectric effect a significantly enhancement in capture of PM with size ranging between 0.3 −1 um occurred in the self-powered filter (Fig. 10 d). An improvement in filtration efficiency induced by triboelectric effect have been reported for other self-powered electrostatic adsorption face mask based on electrospun as well [249,250,[252], [253], [254],256]. In Table 4 , a comparison between the performances obtained for recent Piezoelectric and Triboelectric electrospun system is reported.Table 4 Piezoelectric and Triboelectric electrospun system for smart face masks application.
Table 4Pollutants PFE Pressure drop Quality factor 1Polymer materials Electrospun fiber size Application Refs.
(%) (Pa) (Pa−1)
PM1.0
PM0.5
PM0.3 ∼98.4
∼97.3
∼95 ∼86 ∼0.03 – 0.05 PVDF <1 μm Self-powered face mask
(R-TENG) [251]
PM0.5
PM1.0
PM2.5
PM5.0
PM10
>86.9 ∼57
(Face velocity 10 cm s−1) – PVDF <2 μm Self-powered face mask
(R-TENG) [252]
PM0.3
PM0.5
PM1.0
PM2.5
PM5.0
PM10 >90 ∼10–60 – PI <1 μm Self-powered face mask (R-TENG) [253]
PM0.3
PM5.0
∼99 ∼797 >0.005 PVDF/PAN ∼572/244 nm Monitoring self-powered face mask (RM-TENG) [257]
PM0.5
PM2.5
∼91.5
∼98 ∼98 – PAN doped with PTFE and electroplated
PU sponge ∼150–200 nm Self-Powered Filter for Respirator device [254]
PM0.3
∼99.5 ∼200
(Face velocity 18.6 cm s−1) ∼0.019 PVDF on nylon mesh ∼260 nm Self-charging triboelectric/piezoelectric face mask filter [256]
PM0.3
∼99.7 ∼38 ∼0.154 PEO ∼764 nm Monitoring self-powered filter for face mask [258]
PM1.0
PM2.5
>91
>99 91
(Face velocity 5.83 cm s−1) ∼0.02–0.05 PLLA – Self-charging piezoelectric face mask filter [259]
PM1.0
PM2.5
PM10 ∼100 ∼65
(Face velocity 0.1 m s−1) – EC with PTFE spheres ∼400 nm Triboelectric PM capture filter [249]
PM0.3 ∼98 ∼40 ∼0.094 PBAT@CTAB-MMT ∼300 nm Self-powered filter for face mask [260]
1 Polymers materials abbrevations: PVDF (Polyvinylidene fluoride); PI (Polymide); PAN (Polyacrylonitrile); PTFE (polytetrafluoroethylene); PU (polyurethane); PEO (polyethylene oxide); PLLA (poly(l-lactic acid)); EC (Ethylcellulose); PBAT (polybutylene adipate terephthalate); CTAB (cetyltrimethylammonium bromide); MMT (montmorillonite).
In particular, Hao et al. recently designed a self-powered triboelectric filter system based on PAN nanofibers doped with poly-(tetrafluoroethylene) PTFE nanofibrils and electroplated polyurethane (PU), which provided a high electrostatic adsorption of submicron particles, also after a prolonged use over the time [254]. Unlike the single electrospun filter, the PM0.5 removal efficiency measured for the self-powered triboelectric remained quite stable, after one month of continuous filtration (Fig. 11 a, c).Fig. 11 (a) A comparison between the filtration efficiency obtained for the self-powered triboelectric filter based on PAN/PTFE/PU and that of PAN electrospun nanofibers (NFs), before and after one month of continuous filtration; (b) pressure drop measured for the for different materials, including the self-powered triboelectric filter, NFs, PTFE conductive sponges (CS), and commercial mask; (c) picture taken to a small piece of the STAF system; adapted with permission from [254]. Durability removal efficiency test performed on PVDF electrospun based R-TENG in different external humidities and for particle sizes lower than 1.0 μm (d) and ranging between (e) 1.0 and 2.5 μm; adapted with permission from [252]. (f) Filtration efficiency and pressure drop variation of a quintuple-layer of PVDF/nylon system filter after 10 days; (g) durability filtration efficiency test of the multilayer PVDF/Nylon system after continuous dry or humid air penetration for 0.5 h at different PM sizes; optical microscope and SEM images of PVDF/nylon-based nanofiber multilayers before (h, h-1) and after filtration (i, i-1); and regeneration (j, j-1) by means of oven drying due to ethanol dipping treatment; adapted with permission from [256].
Fig 11
Additionally, the measured pressure drop proved not to be drastically affected by the addition of an extra electrospun membrane and resulted also to be lower in value compared to that of commercial masks, thus proving a considerable advantage in breathability (Fig. 11 b). Nevertheless, the durability test performed on several self-powered air system revealed that in presence of different humid environment the submicron particles removal efficiency decreased, due to a possible dissipation of triboelectric charge on the fibers surface [[250], [251], [252]]. In particular, the experimental study of Guoxu Liu et al. carried out for PVDF electrospun based R-TENG clearly showed that the effect of triboelectric charge decay, which were caused by the humid environment, mainly affected the filtration of tiny particles (Fig. 11 d, e) [252]. Therefore, the dissipation of triboelectric charge may result in a lower adsorption of either potential bacteria or virus carriers. In this perspective, Dong Hee Kang et al. designed a multilayer system with triboelectric/piezoelectric properties made of PVDF nanofibers and nylon mesh, which provided highly efficient charges retention for long period of time with optimal performance both in terms of PM0.3 submicron filtration and pressure drop [256] (Fig. 11 f). Under humid air condition, the filtration efficiency of tiny particles measured for the system of triple or more level filters slightly decreased compared to that observed in dry air condition (Fig. 11 g), pointing out that the triboelectric charge can be preserved during breathing, without having a significant impact on the effectiveness of the filter. The self-powered membrane also showed suitable properties to be employed as a reusable filter in face mask. In fact, the performances of the charged filter remained quite unaltered after ten cycles of both ethanol dipping process and further oven drying, without any sign of PM aggregated on the nanofibers surfaces (Fig. 11 h–j). Also, Lin et al. developed a smart hybrid perfluorinated electret nanofibrous membrane (HPFM) by electrospinning of Polytetrafluoroethylene (PTFE) and Fluorinated ethylene propylene (FEP), which showed high capacity in electric charge storage under high humidity environment [258]. The presence of an amorphous and crystalline surface occurring at the fiber surface level between the low crystallinity of FEP and the higher one of PTFE brings to a low mobility of the charges, which resulted therefore in a more stabilized charge on the fiber surface (Fig. 12 a).Fig. 12 (a) A picture for the charge-trapping mechanism of HPFM electrospun membrane; (b) surface potential variation as function of time obtained for the HPFM system, FEP electret nanofibrous membrane, and PTFE electret nanofibrous membrane; (c) thermally stimulating discharge current spectra (left) and corresponding KPFM images measured for the nanofibers of FFM, PFM, and HPFM (right); adapted with permission from [258]. (d) Filter efficiency evaluation obtained or the PBAT/CTAB/MMT and PBAT triboelectric membranes; (e) a comparison between the Inactivation rates of both influenza virus (left) and Coronavirus (right) obtained for PBAT/CTAB, PBAT, and PBAT/CTAB/MMT; adapted with permission from [260]. (f) A picture of a face mask made of electrospun piezoelectric PLLA nanofibers showing SEM images for the combined effect in PM capturing due to mechanical sieving and piezoelectrically enhanced electrostatic adsorption; (g) decontamination of the PLLA air filter from P. aeruginosa (gram-negative) and S. aureus (gram-positive) bacteria via ultrasound stimulation; adapted with permission from [259].
Fig 12
Because of the high electric charge stabilization, the face mask using HPFM membrane provided a stable filtration efficiency against PM0.3, even after 48 h of exposition to at high humidity ∼ 100% RH. A quantitative confirmation of the high capacity in charge storage was given by the Surface potential curves analysis, which revealed a more stable and higher number of charges trapped in HPFM compared to those measured for single electret filters (Fig. 12 b). This result also agreed with the analysis carried out by the Kelvin Probe Force Microscopy (KPFM) that measured a high intensification in surface potential observed on the nanofibers for HPFM, which was ascribed to the formation of well-localized space charges stored at the nanoscale interface on the fiber surface (Fig. 12 c). Hence, these studies clearly proved that advanced face mask based on electret electrospun filter polymers material and TENG technology can prevent the quick dissipation of electric charge caused by the high amount of vapor water increasing inside the mask during continuous breathing and provide a suitable filtration performance against small aerosols particles, also after prolonged wearing. Furthermore, the use of Triboelectric system can address the loss in efficiency observed in MB based electret filters charged by corona caused by a prolonged contact of the MB membranes with high humidity [23,87]. In addition, to provide an optimal filtration efficiency for submicron particles, these R-TENG based electrospun can also exhibit good accuracy in detecting parameters, such as respiratory rate, inhalation, and exhalation time, which are necessary to identify possible respiratory diseases in patients [248,251,257]. The harvest of mechanical energy from ambient environment by means of the TENG technology can therefore be advantageous to manufacture low-cost self-powered mask, especially now in the modern Internet of the Thing (IoT) where an incessant request of energy supply is required due to the large use of wearable electronic devices and monitoring sensors [255]. Also, several biodegradable polymers materials were found to be suitable for the manufacturing of TENG devices. The smart face mask designed by Yujang Cho et al. based on biodegradable polybutylene adipate terephthalate (PBAT) electrospun filter, functionalized with ecofriendly nanoclay mineral montmorillonite (MMT), and cetyltrimethylammonium bromide (CTAB) surfactant, provided higher filtration performance for PM0.3, compared to commercial face mask [260] (Fig. 12 d). After 10 cycles of repetitive charging and discharging step by mechanical compression and release, the performance remained unaltered, thus indicating a continuous generation of triboelectric charges during the entire friction test. Besides, because of the incorporation of CTAB-MMT, the functionalized membrane showed high antiviral activity against influenza and Human coronavirus (Fig. 12 e). Also, new type of medical self-powered mask based on PVA and PLLA biodegradable polymers provided high charge retention as well as better optimal self-charging performance, under high relative humidity [259,261]. After water exposure, the self-charging piezoelectric nanofiber membrane based on biodegradable PLLA designed by Le et al. showed a limited loss in both PM2.5 and PM1.0 removal compared to both surgical and N95 masks [259]. Furthermore, the electric filter performance remained stable after post annealing treatment for a period of eight weeks, thus showing long durability for prolonged use. Moreover, an antibacterial activity against both S. aureus and P. aeruginosa has been observed by exposing the piezoelectric membrane to 90 min of ultrasound sonication (Fig. 12 f, g). After sterilization procedure, the residual contaminant that adhered the fibers surface were easily removed by means of water immersion and further air blowing, thus making it reusable. It is important to note that these biodegradable TENG showed comparable and in some cases higher QF values to those reported for other inorganic polymer based devices (Table 4) as well as a better biodegradation rate when they are either buried in composting soil or exposed to optimal degradation condition [259,260]. In view of their importance for innovative biomedical application, many of these biodegradable polymers are available on the market nowadays and can be easily dissolved into solution to be used for scalable manufacturing process [128,262,263]. These current studies revealed that high performance in particle filtration, self-charging, and charge retention can be achieved by implementing both synthetic and biodegradable polymers-based electrospun membranes. In particular, the manufacture of biodegradable based TENG represent a new design of self-powered air filter device that can meet the emerging demand of electrical energy and promote a more sustainable green circular economy.
4.5 Reusable filter material under visible light
The importance of disinfecting the disposable surgical facemasks for reuse has become, from the start of the pandemic, an essential issue to overcome the negative impacts on the environment due to the plastic waste as well as to avoid further contamination from bacteria or virus which have been left on the mask [264,265]. Despite some strategies have been studied to sanitize a contaminated mask, the sterilization processes can negatively affect the particle filtration performance thus making the device ineffective to further use [266,267]. Recent advances in environmental-friendly mask development focused on the use of either photocatalytic or photothermal material to deactivate, after an exposition to light sources, bacteria and viruses extending the lifetime quality of face masks and making them reusable. In this perspective, Li et al. designed a reusable, biodegradable, and antibacterial electrospun membrane based on PVA, PEO, and cellulose nanofiber containing photocatalytic materials [163]. Since the limitation in absorbance of TiO2 in the UV range, the membrane has been functionalized also with nitrogen-doped TiO2 (N-TiO2) to enhance the photocatalysis under visible light irradiation (Fig. 13 a).Fig. 13 (a) Schematic diagram that represents how N-TiO2 embedded in face masks can sterilize bacteria under light irradiation; (b) UV−vis spectra obtained for TiO2 and N-TiO2; (c) photo images of plate incubation for E. coli and S. aureus after washing the mounted bacteria on masks with and without light irradiation; (d) CFU count and percentage of survival bacteria (e) for the respective plates; adapted with permission from [163]. (f) Schematic illustration of the preparation of PVDF electrospun fibers functionalized with TTVB for microbe interception and further inactivation under sunlight exposition; (g) comparison between the fluorescence intensity obtained for the PVDF/TTVB electrospun membrane and that without dopant (NM); (h) a comparison between the survival rate for several bacteria obtained for both PVDF/TTVB and NM under 5 min of light exposition and dark condition; adapted with permission from [154].
Fig 13
The irradiation of photocatalyst materials such as TiO2 with light of energy equal or greater than its band gap, observed to be around ∼ 3.06 eV, can promote the excitation of electrons from the valence band (VB) to the conductive band (CB), thus resulting in the formation of excited electron-hole pairs [268]. The interaction between excited electron–hole pairs and ambient air molecules through oxidation and reduction process induce the generation of ROS, such as hydroxyl (•OH) and superoxide anion (O2 −) radicals, and hydrogen peroxide (H2O2), which during photocatalytic process can damage the walls and the membranes of cells, protein, RNA, DNA, bacteria, and virus [269]. By doping the TiO2 NPs with nitrogen, the energy band gap of the TiO2 NPs was observed to decrease at 2.77 eV, thus resulting in a shift in the absorption spectrum to longer wavelengths (Fig. 13 b). Because of the extended adsorption of light in both the UV and near visible region, the count of the bacteria survived by the N-TiO2 mask were significantly lower compared to that measured for N-TiO2 mask (Fig. 13 c). An exposition of the face mask with TiO2 (N-TiO2) NPs mixture to only 10 min of light was sufficient to obtain a complete bacteria disinfection efficiency (100% bactericidal activity), in both E. Coli and S. aureus (Fig. 13 d, e). Moreover, the antibacterial activity remained stable after three sterilization cycles, without any significant decrease in filtration efficiency. Notably, the generation of ROS induced by visible-UV light involved a complete inactivation of bacteria on the functionalized membrane. In contrast, most bacteria remained alive on the same membrane in absence of light exposition, due to the limited antibacterial efficiency caused by the electrostatic interaction between the active TiO2 (N-TiO2) NPs mixture with the negatively charged bacteria. Li et al. used an organic photosensitizer (TTVB) in Poly (vinylidene fluoride-co-hexafluoropropylene) (PVDF-H) polymer-based electrospun membrane to enhance the generation of ROS under light irradiation (Fig. 13 f) [154]. From a comparison between the fluorescence emission intensities measured at 525 nm for the functionalized membrane and the pristine one, it was noted that the presence of TTVB led to a high generation of ROS for longer exposition time to light source (Fig. 13 g). In addition to optimal filtration performance against NaCl particle (≤ 0.3 um), the membrane functionalized with TTVB showed a high inactivation rate of a least ∼99% against several pathogens, including S. aureus, E. coli, and M13 bacteriophage vector, after only 5 min of visible light exposition (Fig. 13 h). Also, the effect of ROS in viral decontamination for face masks has been investigated by Shen et al., that recently designed a photosensitized electrospun nanofibrous membrane with good filtration performances in capturing and rapidly deactivating coronavirus strain surrogates, under visible light radiation (Fig. 14 a) [170]. In their previous study they evaluated the filtration efficiency of PVDF electrospun nanofibers by testing the Murine hepatitis virus A59 (MHV-A59) as coronavirus strain aerosols, because of their similarity in size and structure with SARS-CoV-2 (size ∼85 nm) [19]. The high dominant capture of both airborne corona virus and submicron aerosol particles observed for PVDF-based electrospun membranes indicated that electrospun filter with lower fiber diameter and smaller pore size provided a better prevention for the transmission of COVID, compared to commercial filter of larger fiber size.Fig. 14 (a) (top panel) Photos of PVDF15 (15 wt% of PVDF), PVDF15- RB, and SEM image of PVDF15- RB nanofibrous membrane; (below) NaCl, MHV-A59 aerosol filtration efficiency, and pressure drop observed for RB-sensitized electrospun membranes with different PVDF concentrations (PVDFx-RB); (b) kinetics of MHV-A59 infectivity decay obtained in both droplets and aerosol test; the Nt/N0 ratio represents the ORF5 gene copy numbers, which is measured by the ICC-RT-qPCR, about infectivity, or RT-qPCR, about gene damage, during a light exposition time given by t to that obtained at the initial instant of zero. (c) Optical photos and SEM images (top panels) of PVDF15-RB membranes obtained after different aging test: indoor fluorescent light in laboratory (PVDF-RB-I) and simulated indoor light (PVDF-RB-IS) up to seven days of light exposition; and outdoor light (PVDF-RB-O) exposition up to four days. Scale bars in the SEM images are 5 μm; (below) first-order decay rate constants of MHV-A59 infectivity in droplets test and related 1O2 formation on pristine and light simulated exposed PVDF15-RB membranes. adapted with permission from [170].
Fig 14
Besides the observed high retention of MHV-A59 aerosols, the addition of a photoreactive dye additive, the Rose Bengal (RB) one, to the PVDF based electrospun nanofiber membrane allowed the production of highly reactive singlet oxygen (1O2) that inactivated the 97.1% of the MHV-A59 coronavirus droplets, after a short exposition time to visible light of ∼ 15 min (Fig. 14 b). To compromise the viability, also the ORF5 gene copy number of MHV-A59 was drastically reduced from the generation of ROS during photosensitization, in both droplet and aerosol test with surrogate coronavirus. Moreover, after 4 days of light exposition in outdoor a significant decrease in both corona virus inactivation and generation of 1O2 have been observed for the PVDF-RB membranes (Fig. 14 c). In contrast, a longer exposition under indoor light of 7 days did not compromise the photoreactivity of the functionalized membrane, which maintained a significant performance in both filtration and inactivation of coronavirus. These studies showed that the implementation of photocatalytic inorganic and organic compound in electrospun polymer membranes can be an easy solution to involve a proper sterilization process, under visible light source, by means of the generation of ROS. Moreover, the manufacturing cost for these photocatalytic membranes are economic and easily available on commercial scale [270], and the highly efficient catalyst obtained from low energy irradiation under visible light makes use of functionalized membranes energetically safer as well as easily implemented, compared to other methods commonly used for pathogens removal [271,268]. The use of photothermal nanomaterials, including graphene, carbon nanotubes (CNTs), and other metal NPs, have been also considered for the manufacture of sunlight self-sterilization face mask [272]. Under exposition to sunlight source, these nanocomposites can absorb light energy and convert it in heat to provoke a significant increase of the local temperature, which may result in a thermal denaturation of cells, bacteria, and viruses [273]. Xiong et al. designed a novel needleless ES/spraying-netting technology to fabricate a hierarchical system based on CNTs network welded on PAN nanofiber (NF) layer with good filtration performances, both photothermal and electrothermal self-sterilization, and recycling performance [182] (Fig. 15 a).Fig. 15 (a) Schematics of self-sterilization of NF/CNT fibrous due to electrothermal; (b) IR images recording the temperature changes of carbon fiber layer (ACF), melt-blown fabric layer (MBF), and NF/CNT fibrous masks; (c) equilibrium temperatures of the NF/ CNT fibrous network under different simulated solar; (d) sterilization performance of ACF, MBF, and NF/CNT fibrous network masks under 1 sun; (e) photographs of the distributions of E. coli colonies on nutrient agar solid plates when applied small currents for 2 min; (f) sterilization performance of the electrothermal NF/CNT filter under different applied currents for 2 min; (g) filtration performance, pressure drop, and quality factor of the NF/CNT filter after repeating wetting-drying process; adapted with permission from [182].
Fig 15
The configurations obtained for the NF/CNT filter showed high efficiency in PM0.3 removal (∼ 99.9%) with low pressure drop (∼ 49 Pa). Since the small micron scale pores and multi-scale structure enhance the photon absorption in the full wave band between (500–2500) nm, the NF/CNT filter showed a fast photothermal response (T > 70 °C within 5 s of sunlight exposition) that involves an excellent antibacterial activity against E. Coli (> 99.8%) only after a sun exposure time of three minutes (Fig. 15 b–d). The authors stated that since the COVID-19 can be deactivated within few minutes of heat treatment higher than or equal to 65 °C [274], the high equilibrium temperature of NF/CNTs (T > 60 °C after 0.5 sun) would allow to kill the virus after 15 min of sun exposition. A faster high-temperature response with better antibacterial performance (> 99.9%) has been obtained by providing an electrical current of 61 mA for two minutes (Fig. 15 e, f). The photothermal response showed to be comparable or higher with those observed for other self-sterilizing face masks, recently designed with different sunlight-responsive materials [[275], [276], [277]]. However, due to the small average pore size and small fiber diameter the multi scale nanoarchitecture of NF/CNTs provided better performance in both filtration and pressure drop, especially after 50 wetting-drying cycles, thus suggesting its potentiality as a reusable filter to reduce disease spread and consumption of resources (Fig. 15 g). Moreover, due to the large-area deposition and to the structural stability of the CNT network on scaffold polymer nanofibers, the needleless electrospinning coupled with spraying-netting has proven to be an easy and accessible technique to produce self-sterilizing filter membranes on large scale. These studies clearly show that electrospun membranes functionalized with either photocatalytic or photothermal nanomaterials can be advantageous to produce low-cost high-filtering performance face masks with suitable self-sterilization properties. The fast sterilization response obtained under low exposition time in visible light/sunlight would allow to overcome the drawback of efficiency degradation observed in commercial face masks, caused by a long time-exposure to conventional heating and UV-C radiation disinfection methods [193,[278], [279], [280]]. Furthermore, the easy decomposition of biopolymer self-sterilizing face masks in water and soil may be a new promising line of research to substitute nonrenewable synthetic polymers and relieve the plastic waste pollution at the landfills [163].
5 Future perspectives for sustainability of membrane technology
The demand for green chemistry is increasing, due to the growing awareness on the negative environment and health impacts associated with traditional industrial processes, and special attention has been given to electrospun nanofibers membranes, since their wide-spread applications in biomaterial industry and scale-up in commercial production of filtering layers in face masks [281,282]. The electrospinning solutions often involve the use of halogenated compounds, such as 1,1,1,3,3,3- hexafluoro-2-propanol (HFIP), tetrahydrofuran (TFH), N,Ndimethyl- formamide (DMF), 1-methyl-2-pyrrolidinone (NMP),and dimethylacetamide (DMAC), because of their high solvation power for hydrophobic polymers and low boiling points which ensure a proper viscosity of the commonly used high molecular weight polymers, necessary properties for the ES process to convert the solution into fibers [283]. However, the most used solvents in the ES process, as well as for other methods adopted in the ultrafiltration membranes manufacturing processes, are toxic and their usage has been recently restricted by the Chemical Control Regulation in the European Union (REACH) [119,282]. A strategy to improve the sustainability of the ultrafiltration membranes manufacturing process would consist in using greener/low-toxicity solvents from renewable resources. To date, biobased nontoxic solvents have been attracting attention due to their decreased impacts on human health and environment [119,284]. The Hansen solubility parameters (HSP) calculated for these green solvents reveal a high polymer-solvent compatibility with the typical polymers employed for membrane applications, thus suggesting feasibility for these non-hazardous solvents in the fabrication process [119,284]. In literature, some studies reported first steps in eco-friendly alternative ES employed to spin common polymers for membranes manufacturing [285,286]. Green solvents based nanofibrous membranes have been also electrospun to be used as breathable and comfortable materials for application of protective clothing or as an additional protective layer within existing cartridges as well as in fluid filtration and separation of nanomaterials [185,[287], [288], [289],186,183]. In particular, the applicability of water as a green solvent for environmental-friendly membrane processing has been recently investigated for several polymer materials [119]. Since it has good cost-effectiveness and environmental compatibility, the fabrication of environmental electrospun membranes based on water-soluble and biodegradable PVA polymer has been recently considered to replace non-degradable materials and toxic solvent in different biomedical application, including drug delivery and tissue engineering [120,290]. However, the mechanical properties of pristine PVA membrane are poor, and the filtration performance in the long run may be affected by possible breakage of nanofibers. In this perspective, Cui et al. observed that the addition of Tannic Acid (TA) to water-soluble precursor solution provides better mechanical properties to the membrane compared with that observed for pristine PVA [177] (Fig. 16 a).Fig. 16 (a) Schematic diagram showing the preparation of PVA-TA nanofiber membrane; (b) a comparison between PVA-TA, surgical and dust masks on filtration efficiency, pressure drop, and quality factor properties; (c) PM1.0 capture of the PVA-TA masks after 20 consecutive filtration tests; adapted with permission from [177]. (d) Schematic illustration of conventional stick-like nanofiber and (e) curved-ribbon nanofiber based on green ES for eco-friendly breathable and high-performance air filtration; fiber diameter distribution of (d-1) pristine PVA and (e-2) 100%SBE-βCD/PVA; (f) Quality factor of nanofiber membranes for PM1.0 under 5.3 cm s−1; (g) cycling filtration performance of re-assembled curved-ribbon mask and common medical mask for removing PM1.0 and PM2.5; adapted with permission from [168].
Fig 16
A similar result was also found in a previous work by adding lignosulfonate (LS) to PVA electrospinning solution [178]. The PVA-TA nanofiber showed high filtration efficiency against PM1.0 (∼ 99.5%) with pressure drop about 35 Pa. In comparison with medical surgical face masks, the fiber interception mechanism of PVA-TA filter proved to be unaltered showing an excellent fine particle filtration after twenty consecutive tests performed (Fig. 16 b, c). Furthermore, PVA nanofibers prepared with water-soluble chitosan (WS-CS) by means of needleless ES showed a small average pore size ranging between 12.06 and 22.48 nm and provided antibacterial activity against S. aureus as well as high filtration efficiency in removal PM2.5 (∼ 93.1 - 97%), with optimal pressure drop (∼ 41 – 44 Pa) [167] (Table 3). In another work, it has been observed that the addition of TiO2 nanotubes to the water precursor solution of CS/PVA provided high mechanical strength to nanofibers product thus showing less tendency to degrade over time. Moreover, under dark conditions the CS/PVA/TiO2 membrane showed a bacterial inhibition growth of ∼ 44.8%, higher than that observed for the control sample [291]. Also, the addition of Ag NPs has shown to enhance proper depletion of both S. aureus and E. coli on PVA-water based electrospun [166]. Therefore, the addition of active compound proved to be beneficial to address the poor mechanical properties issue in water soluble PVA membrane and enhance its antibacterial effectiveness. Furthermore, morphological changes occurring at the fiber surface level due to the inclusion of active materials in blend solution ES have been observed to significantly improve the removal of PM for the electrospun membranes, as well. In a recent study, Deng et al. reported that the addition of Sodium sulphobutylether-β-cyclodextrin (SBE-βCD) in PVA ultrapure water solution resulted in the formation of curved-ribbon fibers, with high mechanical properties as well as better filtration performance compared to pristine PVA membrane [168]. Compared to the stick-like fibers obtained by conventional ES, the green electrospun fibers provided a more efficient interceptor mechanism capture for PM1.0 around 99.12% due to the combination of morphological interception and electrostatic mechanisms as well as a higher tensile strength and strain, which can be ideal to stabilize the breathable filter under external air stress (Fig. 16 d–f). Furthermore, after 20 cycling filtrations test the filter maintained both filtration stability against PM1.0 and proper pressure drop of 59.5 Pa, thus making it reusable for most applications (Fig. 16 g). A similar result has been reported also in the work carried out by Hu et al., where a higher capture of PM0.3 was observed for zein based electrospun membranes with flat ribbon structure, obtained by ethanol/ deionized water mixture solvent [185] (Fig. 17 a).Fig. 17 (a) Schematic diagram showing the fabrication of the zein electrospun filter on cellulose membrane; (b) the filtration and pressure drop performance stability under different humidity conditions; (c) the model depicted shows (i) Stick-like zein and (ii) flat ribbon-like zein mat; (iii, iv) filtering simulation calculated from the two different fiber structures, respectively; (v, vi) pressure field simulations for the two distinct model, respectively; (d) images of the Biodegradability test for the zein based electrospun filter under enzymatic degradation in time, and (e) the measure of the related weight loss; adapted with permission from [185].
Fig 17
The simulated 3D structure model also confirmed that a larger contact surface in flat zein fibers leads to an improvement in filtration, due to a higher interaction occurring between the pollutant particles and the numerous functional groups owing to zein protein (Fig. 17 c). Notably, both the filtration and the pressure drop were observed to remain stable, also when varying the relative humidity in the range between 20% and 80% (Fig. 17 b). Furthermore, the face mask based on zein filter completely degraded after 42 h under cellulase enzyme, thus proving to be biodegradable and environmental-friendly (Fig. 17 d, e). These studies clearly showed that nanofibers based on biopolymers and natural materials can be easily dissolved in greener solvents and being optimized by means of ES to address the lack of mechanical properties and to improve filtration performance for submicron particles. Such customizability makes the green ES methods commercially attractive to produce eco-friendly face mask, since it would relieve the high consumption of non-renewable plastic resources and that of hazardous solvents in large-scale commercial production.
5.1 Electrospun nanofibers in bio-textiles for potential biomedical application
Due to the high versatility in producing a wide range of continuous polymer based nanofibers, ES has attracted much interest in recent years as a potential solution for the development and the design of textile-based scaffold for tissue engineering [30,31,35]. Scaffold in tissue engineering is used to mimic the extracellular matrix (ECM) of the native tissue. To restore the functionality of the damaged tissues, ECM provide biochemical and physical cues by enabling cellular physiological activities and diffusion of vital cells as well as promoting cell adhesion, morphogenesis, and migration on the tissue structure [292]. Since the structural similarities between the electrospun nanofibers and the complex network forming native tissues, different electrospun based scaffolds have been fabricated to reproduce both the morphological and biochemical characteristic of ECM for cardiac, nerve, muscle, tendons, and connective tissue regeneration [293]. The combination of several natural and synthetic polymers by means of Coaxial ES have been found to be a useful approach to tailor and customize the biochemical and physical properties of electrospun based scaffold for the specific requirement of medical implants [294]. Natural biomaterials, including collagen, chitosan, and gelatin, provided suitable biocompatibility to the electrospun mats, but showed disadvantages in bio mechanical properties. On the other hand, the addition of synthetic biopolymers, such as PVA, PLLA, PCL, ensured a better structural stability to natural/synthetic based scaffold, thus satisfying requirement for ECM in tissue engineering [295]. However, due to random orientation of conventional nanofibers, the electrospun based scaffold cannot correctly mimic the oriented features of native ECMs, such as nerve, and tendons, thus limiting cells migration and its large application to tissue engineering. In comparison, scaffold based on more aligned nanofibers proved to be more efficient to guide and accelerate both the differentiation and infiltration of in vitro cell tissue regeneration, but due to the poor mechanical properties and the low thickness they resulted in insufficient structural support for some native tissue [293,296]. The assembling of electrospun nanofibers into ordered textile nanoyarns (NYs) has been found to be advantageous to overcome the poor mechanical properties of electrospun mats and involve a better bioactivity properties in scaffold by promoting cell proliferation, migration, and differentiation in vivo microenvironment [30,31]. In particular, the modified conjugated ES technique designed by Wu et al. proved to be an innovative method for the fabrication of aligned polymer based NYs with suitable mechanical strength and high biocompatibility [297]. A schematic picture of the modified electrospinning system is reported in Fig. 18 a, to illustrate the generation of PLLA based NYs.Fig. 18 (a) A schematic illustration of the modified ES system for the fabrication of PLLA NYs with and without the hot drawing process; (b) SEM images obtained for the PLLA NYs without hot drawing process (0X PLLA NYs), and three thermally drawn PLLA NYs processed with several stretching degrees, i.e., 1X PLLA NYs, 2X PLLA NYs and 3X PLLA NYs. Scale bars: 100 μm for top panel and 2.5 μm for down panel; (c) images of immunofluorescent staining taken of hADMSCs present on the 0X PPLA NYs, 1X PPLA NYs and 3X PPLA NYs after 14 days of culture. Tenomodulin (TNMD), antibody to collagen type I (COL1), and nuclei were marked with green, red, and blue, respectively. Scale bars are equal to 50 μm. (d); the qPCR analysis obtained for five tendon-specific gene markers such as SCX, TNC, COL1, COL3 and TNMD for hADMSCs cultured after 24 days on the 0X PPLA NYs, 1X PPLA NYs and 3X PPLA NYs; (e) photograph and SEM images of braids fabricated with 12 electrospun based PLLA NYs, named as 4 × 3 braids. Scale bars: 0.5 mm for left image and 100 μm for right one; adapted with permission from [297].
Fig 18
The polymer jets of opposite charge are electrospun by two needles, located opposite to each other, and collected between a steady neutral hollow metal rod (NHMR) and a rotating neutral metal disk (NMD) to be subsequently bundled in NYs by means of a hot drawing cabinet. The single PLLA based NYs obtained with hot drawing process showed a higher alignment degree, a higher strength, and a higher Young's modulus compared to that obtained without hot stretching treatment, thus indicating a better mechanical property (Fig. 18 b). Furthermore, CCK assays carried out on hot stretched PLLA based NYs also showed a high proliferation of human Adipose Derived Mesenchymal Stem Cells (hADMSCs). This result was in accordance with the images of immunofluorescences for phalloidin and DAPI double staining that revealed a significantly elongation and alignment for cytoskeletal protein and nuclei of the cells, respectively, thus suggesting that aligned nanofibers in NYs with hot drawing process can induce a better cell attachment and proliferation enhancement. Indeed, both protein secretion and formation of nuclei along yarn longitudinal direction of aligned PPLA NYs have been observed in the immunofluorescences analysis carried out for tenogenic differentiation of hADMSCs (Fig. 18 c). The quantitative PCR (qPCR) analysis also revealed a high relative expression for tendon-specific gene markers, including SCX, TNC, COL1, COL3 on hot stretched PLLA based NYs, thus indicating that high alignment of nanofibers in PLLA based NYs can be beneficial to tenogenic differentiation of hADMSCs (Fig. 18 d). Importantly, NYs can be furthermore braided by means of traditional textile engineering methods, such as weaving, knitting, and braiding to manufacture different woven fabric nanotextiles with controlled structures and shapes, which are able to meet the requirements for the different tissue engineering applications (Fig. 18 e) [30]. Woven textiles interlaced with silk fibroin (SF)/PLLA based NYs by means of weaving method proved to be beneficial to drive the growth activity of HADMSCs [298]. From a comparison between the SEM images and the quantitative MTT analysis performed on scaffolds interlaced with SF/PLLA NYs at different mass ratio, it was noted that the higher the presence of bioactive SF, the higher the proliferation capacity of the HADMSCs. Moreover, after in vivo subcutaneous implantation of SF/PLLA NYs based scaffolds, the low immunofluorescence staining intensity of CD68 protein, a general marker for macrophages, which was measured in presence of high amount of SF indicated a drastically reduction of the inflammatory response, thus suggesting that SF/PLLA NYs based woven fabrics can effectively induce cellular infiltration and tissue regeneration.
A similar result has been observed also in the recent work of Ye Qi et al., where the low number of macrophages obtained by the measure of the immunofluorescence staining intensity of CD68 protein for woven textiles interlaced with aligned SF/PCL based NYs was clearly attributed to the presence of SF in the NYs units forming the scaffold [299]. Therefore, NYs based woven fabrics can maintain the characteristic properties given from the original NYs, including mechanical, biocompatibility, and biostability properties, and can be adjusted in several shape and structure, to satisfy the several request for medical implantation (Fig. 19 a). Moreover, recent scaffold based NYs functionalized with active nanomaterials showed excellent biocompatibility and demonstrated to be appropriate to manufacture nerve guidance conduit (NGC) for the repair of injured nerves [31]. Zhang et al. designed an innovative PCL based NYs functionalized by means of PCL microparticles embedded with bioactive nerve growth factor (NFG), which was able to promote both topological and biochemical signals for the manipulation of axon growth cells [300]. Compared to aligned PCL based NYs, the inclusion of PCL microparticles was observed to effectively promote the axon growth for PC12 and SH-SY5Y cells. Additionally, the load of bioactive NFG in the core of PCL microparticles induced a longer migration distance of neural stem cells on the NYs environment, thus proving that such combination promotes both a better axonal extension and a subsequently migration of the stem cells (Fig. 19 b–e). Also, a multifunctional tube-filling material fabricated with aligned NYs based on poly(p-dioxanone) (PPDO) biopolymer and different concentrations CNTs has been designed to improve the performances of the NGCs in the repair of long-gap peripheral nerve (PN) injuries [301]. The addition of CNTs at 5% in PPDO based NYs proved to support both elongation and alignment for hADMSCs as well as to enhance both electrical conductivity and mechanical properties. Besides, under the combination of both electrical stimulation and chemical induction the expressions of the myelination-associated proteins and other genes markers for hADMSCs measured by means of QRT-PCR analysis were higher in the conductive functionalized PPDO based NYs compared to those observed in growth medium (GM), thus indicating a significant improvement in the differentiation of hADMSCs into myelinating Schwann cell-like cells (Fig. 19 f, g). Hence, these results demonstrated that the development of electrospun NYs functionalized with active materials can be determinant to promote the neuronal differentiation of stem cells. Furthermore, the addition of natural and synthetic biopolymers in electrospun based NYs allows to adjust the biodegradability rate, thus making the production of nanotextile woven fabric more suitable for the fabrication of bio textiles implants [297,298]. In addition, the high fracture strength and tensile elongation provided by nanotextiles woven fabric plays an important role in the production of flexible and smart textile for healthcare wearable applications [30,302]. Recent electrospun nanofibers prepared by using polypyrrole (PPy) and covered with PAN-KCZ/ZnO nanofibers were twisted to prepare composite NYs with both photothermal and electrothermal properties for self-sterilization application [303]. The fabricated nanotextile woven provided a faster high-temperature response under the exposition to both light irradiation and electrical power, which promoted the continuous release of active KCZ in preventing the infection from bacteria pathogens. Also, thermoplastic polyurethanes (TPU) based NYs functionalized with silver nanowires and MXene exhibited a fast electromechanical performance under mechanical deformations as well as a photothermal and electrothermal response under external stimuli, including electricity and light, that resulted to be advantageous for the realization of wearable healthcare sensors [304]. Hence, the combination of ES with traditional textile processing methods can be a feasible approach to design and develop advanced nanotextile for several potential applications in various biomedical fields. It is worth noting that electrospun based NYs with high degree of alignment and lengths of the order of hundreds of meters can be produced in a continuous manner [30]. Therefore, thanks to the utilization of automated and programmable textile machines the structure, shape, and mechanical features can be tuned for the large-scale production, thus making the manufacturing of nanotextile woven fabric based on electrospun NYs expandable for a future commercialization.Fig. 19 (a) Schematic diagram of the fabrication process of electrospun based NY and its further production in woven textile, as functional scaffold for in vivo tissue engineered application; adapted with permission from [299]. A comparison between the fluorescence images showing the cell migration from the spheroids of neural stem cells after 7 days of culture on (b) PCL NYS functionalized with NGF@PCL microparticles, obtained at a deposition time of 5 min, and that on (c) PCL NYs (blank yarns); (d) schematic picture showing the axonal growth of both PC12 and SH-SY5Y cells and their further migration from the initial spheroids forms of neural stem cells on PCL NYs, PCL NYs functionalized with PCL microparticles, and PCL NYs functionalized with NGF@PCL microparticles; (e) migration distance measured for the neural stem cells after culturing on the different PCL NYs classes for 7 days;; adapted with permission from [300]. (f) Experimental design of the hADMSCs differentiation towards myelinating Schwann cell-like cells by means of electrical stimulation (ES), chemical induction (CI), and their combination; (g) images of the immunofluorescent staining for S100B (red), MBP (green), and nuclei (blue) of hADMSCs seeded on PPDO/5%CNT NYs by using GM and the combination of ES and CI (ES+CI), after 14 days of culture; scale bar equal to 100 μm; adapted with permission from [301].
Fig 19
6 Overview on nanofibers market and conclusions
The great advantages of obtaining fibers with tailored diameter sizes and surface characteristic properties, down to the nano scale, has made the ES a versatile technique to produce promising electrospun membranes with high filtering performances. So far, a great variety of synthetic and natural polymers have been electrospun with the addition of either active natural compounds or nano materials to design functionalized filters that can significantly prevent transmission of infection by viral and bacterial pathogens. Important advances have been achieved in the development of reusable and self-sterilizing biodegradable polymers-based electrospun membranes. The manufacturing of more sustainable electrospun filtering devices can address current issues originating from the outbreak of COVID-19 pandemic, including the improper disposal of surgical mask and the remotion of possible bacterial contamination present on PPE surfaces after their use. Although most of electrospun research have been carried out at lab-scale, the electrospinning process can be easily up-scaled, thus enabling a promising perspective on rapid production of nanofiber mats to industrial scale. In recent years, many companies have patented and marketed industrial set-up machines for large scale nanofiber production [70,128,305,306]. Among them, some leading textile industries, such as INOVENSO (www.inovenso.com), E-SPIN Nanotech (www.espinnanotech.com), and ELMARCO (www.elmarco.com), developed industrial-scale technology to fabricate high production polymers-based electrospun membranes, as advanced filters materials for face mask and respirators application. Since the shortage of face mask during the COVID emergency, several international companies such as ske (www.ske.it), nask (www.nask.hk), sncfibers (www.sncfibers.com), fnm (www.fnm.ir), respilon (www.respilon.com), and nano4fiber (www.nano4fiber.com), have been starting to produce electrospun nanofiber mask with optimal filtration performance, according either to European or USA standard mask protocols. Performance and characteristics of some face masks based on electrospinning technology, which are currently available on the market, are reported in literature [28]. Several smaller companies and start-ups, including 4C-Air, Maryland-based DiPole Materials, and Spanish Bioinicia, also have reacted to the COVID-19 pandemic by employing ES technology to develop advanced nanofibers-based filter with better performance in both aerosols capture and breathability [307]. The Bioinicia company (www.bioinicia.com) in collaboration with the National Scientific Research Council (CSIC) developed the first world compostable nanofiber respiratory mask for protection against COVID-19 to the market with FFP2 breathability and filtration capacity, based on ES technology. Conversely to commercial mask, the nanofiber-based face mask is made of compostable biomaterial that can easily decompose in 22 days, thus preventing further dissemination of infection plastic waste. NanoLayr company (www.nanolayr.com), formerly known as Revolution fibres Ltd, recently developed a compostable and eco PLA based electrospun filter integrated with manuka oil that meet international standard of N95 respirators providing both bactericidal and virucidal properties as well as the ability to restore filtration performance, after several washing cycles [171]. Biodegradable polymer materials proved to be an eco-friendly resource for the manufacturing of high-efficiency face masks, but the absence of an alternative fiber fabrication protocol to replace common hazardous solvents with less toxic ones in large-scale manufacturing has not been yet established. Since the electrospinning research is leading to industrial application, new strategies for the introduction of alternative and environmental-friendly solvents should be investigated. In this perspective, the use of greener solvents should be a more effective solution to both reduce the environmental impact of hazardous solvents, which are restricted by the REACH regulation, and improve a more sustainable ultrafiltration membranes manufacturing process on large scale in the future.
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.
Data availability
No data was used for the research described in the article.
==== Refs
References
1 Ippolito M. Vitale F. Accurso G. Iozzo P. Gregoretti C. Giarratano A. Cortegiani A. Medical masks and respirators for the protection of healthcare workers from SARS-CoV-2 and other viruses Pulmonology 26 2020 204 212 10.1016/j.pulmoe.2020.04.009 32362505
2 Bandiera L. Pavar G. Pisetta G. Otomo S. Mangano E. Seckl J.R. Digard P. Molinari E. Menolascina F. Viola I.M. Face coverings and respiratory tract droplet dispersion R. Soc. Open Sci. 7 2020 201663 10.1098/rsos.201663
3 Bałazy A. Toivola M. Adhikari A. Sivasubramani S.K. Reponen T. Grinshpun S.A. Do N95 respirators provide 95% protection level against airborne viruses, and how adequate are surgical masks? Am. J. Infect. Control 34 2006 51 57 10.1016/j.ajic.2005.08.018 16490606
4 O'Dowd K. Nair K.M. Forouzandeh P. Mathew S. Grant J. Moran R. Bartlett J. Bird J. Pillai S.C. Face masks and respirators in the fight against the COVID-19 pandemic: a review of current materials, advances and future perspectives Materials 13 2020 3363 10.3390/ma13153363 32751260
5 Zhu N. Zhang D. Wang W. Li X. Yang B. Song J. Zhao X. Huang B. Shi W. Lu R. Niu P. Zhan F. Ma X. Wang D. Xu W. Wu G. Gao G.F. Tan W. A novel Coronavirus from patients with pneumonia in China, 2019 N. Engl. J. Med. 382 2020 727 733 10.1056/NEJMoa2001017 31978945
6 Setti L. Passarini F. De Gennaro G. Barbieri P. Perrone M.G. Borelli M. Palmisani J. Di Gilio A. Piscitelli P. Miani A. Airborne transmission route of COVID-19: why 2 Meters/6 feet of inter-personal distance could not be enough IJERPH 17 2020 2932 10.3390/ijerph17082932 32340347
7 Jarvis M.C. Aerosol transmission of SARS-CoV-2: physical principles and implications Front. Public Health 8 2020 590041 10.3389/fpubh.2020.590041
8 Mutuku J.K. Hou W.C. Chen W.H. An overview of experiments and numerical simulations on airflow and aerosols deposition in human airways and the role of bioaerosol motion in COVID-19 transmission Aerosol Air Qual. Res. 20 2020 1172 1196 10.4209/aaqr.2020.04.0185
9 Cook T.M. Personal protective equipment during the coronavirus disease (COVID) 2019 pandemic – a narrative review Anaesthesia 75 2020 920 927 10.1111/anae.15071 32246849
10 Workman A.D. Welling D.B. Carter B.S. Curry W.T. Holbrook E.H. Gray S.T. Scangas G.A. Bleier B.S. Endonasal instrumentation and aerosolization risk in the era of COVID-19: simulation, literature review, and proposed mitigation strategies Int. Forum Allergy Rhinol. 10 2020 798 805 10.1002/alr.22577 32243678
11 Chen W. Zhang N. Wei J. Yen H.L. Li Y. Short-range airborne route dominates exposure of respiratory infection during close contact Build. Environ. 176 2020 106859 10.1016/j.buildenv.2020.106859
12 Zhang N. Chen W. Chan P. Yen H. Tang J.W. Li Y. Close contact behavior in indoor environment and transmission of respiratory infection Indoor Air 30 2020 645 661 10.1111/ina.12673 32259319
13 Zhang R. Li Y. Zhang A.L. Wang Y. Molina M.J. Identifying airborne transmission as the dominant route for the spread of COVID-19 Proc. Natl. Acad. Sci. USA 117 2020 14857 14863 10.1073/pnas.2009637117 32527856
14 Mathieu E. Ritchie H. Ortiz-Ospina E. Roser M. Hasell J. Appel C. Giattino C. Rodés-Guirao L. A global database of COVID-19 vaccinations Nat. Hum. Behav. 5 2021 947 953 10.1038/s41562-021-01122-8 33972767
15 Brüssow H. Zuber S. Can a combination of vaccination and face mask wearing contain the COVID-19 pandemic? Microb. Biotechnol. 15 2022 721 737 10.1111/1751-7915.13997 34962710
16 Bartsch S.M. O'Shea K.J. Chin K.L. Strych U. Ferguson M.C. Bottazzi M.E. Wedlock P.T. Cox S.N. Siegmund S.S. Hotez P.J. Lee B.Y. Maintaining face mask use before and after achieving different COVID-19 vaccination coverage levels: a modelling study Lancet Public Health 7 2022 e356 e365 10.1016/S2468-2667(22)00040-8 35276093
17 Andrejko K.L. Pry J.M. Myers J.F. Fukui N. DeGuzman J.L. Openshaw J. Watt J.P. Lewnard J.A. Jain S. California COVID-19 Case-Control Study TeamCalifornia COVID-19 Case-Control Study TeamAbdulrahim Y. Barbaduomo C.M. Bermejo M.I. Cheunkarndee J. Cornejo A.F. Corredor S. Dabbagh N. Dong Z.N. Dyke A. Fang A.T. Felipe D. Frost P.M. Ho T. Javadi M.H. Kaur A. Lam A. Li S.S. Miller M. Ni J. Park H. Poindexter D.J. Samani H. Saretha S. Spencer M. Spinosa M.M. Tran V.H. Walas N. Wan C. Xavier E. Effectiveness of face mask or respirator use in indoor public settings for prevention of SARS-CoV-2 infection-California, February–December 2021 MMWR Morb. Mortal. Wkly. Rep. 71 2022 212 216 10.15585/mmwr.mm7106e1 35143470
18 Essa W. Yasin S. Saeed I. Ali G. Nanofiber-based face masks and respirators as COVID-19 protection: a review Membranes 11 2021 250 10.3390/membranes11040250 33808380
19 Shen H. Zhou Z. Wang H. Zhang M. Han M. Durkin D.P. Shuai D. Shen Y. Development of electrospun nanofibrous filters for controlling Coronavirus aerosols Environ. Sci. Technol. Lett. 8 2021 545 550 10.1021/acs.estlett.1c00337
20 Brochocka A. Nowak A. Majchrzycka K. Puchalski M. Sztajnowski S. Multifunctional polymer composites produced by melt-blown technique to use in filtering respiratory protective devices Materials 13 2020 712 10.3390/ma13030712 32033314
21 Kara Y. Molnár K. A review of processing strategies to generate melt-blown nano/microfiber mats for high-efficiency filtration applications J. Ind. Text. 51 2022 137S 180S 10.1177/15280837211019488
22 Pu Y. Zheng J. Chen F. Long Y. Wu H. Li Q. Yu S. Wang X. Ning X. Preparation of polypropylene micro and nanofibers by electrostatic-assisted melt blown and their application Polymers 10 2018 959 10.3390/polym10090959 30960884
23 Han M.C. He H.W. Kong W.K. Dong K. Wang B.Y. Yan X. Wang L.M. Ning X. High-performance electret and antibacterial polypropylene meltblown nonwoven materials doped with boehmite and ZnO nanoparticles for air filtration Fibers Polym. 23 2022 1947 1955 10.1007/s12221-022-4786-8
24 Wu J. Zhou H. Zhou J. Zhu X. Zhang B. Feng S. Zhong Z. Kong L. Xing W. Meltblown fabric vs nanofiber membrane, which is better for fabricating personal protective equipments Chin. J. Chem. Eng. 36 2021 1 9 10.1016/j.cjche.2020.10.022 33250602
25 Ullah S. Ullah A. Lee J. Jeong Y. Hashmi M. Zhu C. Joo K.I. Cha H.J. Kim I.S. Reusability comparison of melt-blown vs nanofiber face mask filters for use in the Coronavirus pandemic ACS Appl. Nano Mater. 3 2020 7231 7241 10.1021/acsanm.0c01562
26 Xu J. Xiao X. Zhang W. Xu R. Kim S.C. Cui Y. Howard T.T. Wu E. Cui Y. Air-filtering masks for respiratory protection from PM2.5 and pandemic pathogens One Earth 3 2020 574 589 10.1016/j.oneear.2020.10.014 33748744
27 Zhang Z. Ji D. He H. Ramakrishna S. Electrospun ultrafine fibers for advanced face masks Mater. Sci. Eng. R Rep. 143 2021 100594 10.1016/j.mser.2020.100594
28 Naragund V.S. Panda P.K. Electrospun nanofiber-based respiratory face masks-a review Emerg. Mater. 5 2022 261 278 10.1007/s42247-022-00350-6
29 Zakrzewska A. Haghighat Bayan M.A. Nakielski P. Petronella F. De Sio L. Pierini F. Nanotechnology transition roadmap toward multifunctional stimuli-responsive face masks ACS Appl. Mater. Interfaces 14 2022 46123 46144 10.1021/acsami.2c10335 36161869
30 Wu S. Dong T. Li Y. Sun M. Qi Y. Liu J. Kuss M.A. Chen S. Duan B. State-of-the-art review of advanced electrospun nanofiber yarn-based textiles for biomedical applications Appl. Mater. Today 27 2022 101473 10.1016/j.apmt.2022.101473
31 Li Y. Dong T. Li Z. Ni S. Zhou F. Alimi O.A. Chen S. Duan B. Kuss M. Wu S. Review of advances in electrospinning-based strategies for spinal cord regeneration Mater. Today Chem. 24 2022 100944 10.1016/j.mtchem.2022.100944
32 Sethuram L. Thomas J. Therapeutic applications of electrospun nanofibers impregnated with various biological macromolecules for effective wound healing strategy – a review Biomed. Pharmacother. 157 2023 113996 10.1016/j.biopha.2022.113996
33 Al-Shalawi F.D. Azmah Hanim M.A. Ariffin M.K.A. Looi Seng Kim C. Brabazon D. Calin R. Al-Osaimi M.O. Biodegradable synthetic polymer in orthopaedic application: a review Mater. Today Proc. 74 2023 540 546 10.1016/j.matpr.2022.12.254
34 Kumar M. Hilles Ayah.R. Ge Y. Bhatia A. Mahmood S. A review on polysaccharides mediated electrospun nanofibers for diabetic wound healing: their current status with regulatory perspective Int. J. Biol. Macromol. 234 2023 123696 10.1016/j.ijbiomac.2023.123696
35 Gruppuso M. Turco G. Marsich E. Porrelli D. Polymeric wound dressings, an insight into polysaccharide-based electrospun membranes Appl. Mater. Today 24 2021 101148 10.1016/j.apmt.2021.101148
36 Zhang X. Xie L. Wang X. Shao Z. Kong B. Electrospinning super-assembly of ultrathin fibers from single- to multi-Taylor cone sites Appl. Mater. Today 26 2022 101272 10.1016/j.apmt.2021.101272
37 Islam M.S. Larpruenrudee P. Saha S.C. Pourmehran O. Paul A.R. Gemci T. Collins R. Paul G. Gu Y. How severe acute respiratory syndrome Coronavirus-2 aerosol propagates through the age-specific upper airways Phys. Fluids 33 2021 081911 10.1063/5.0061627
38 Prather K.A. Wang C.C. Schooley R.T. Reducing transmission of SARS-CoV-2 Science 368 2020 1422 1424 10.1126/science.abc6197 32461212
39 Armentano I. Barbanera M. Carota E. Crognale S. Marconi M. Rossi S. Rubino G. Scungio M. Taborri J. Calabrò G. Polymer materials for respiratory protection: processing, end use, and testing methods ACS Appl. Polym. Mater. 3 2021 531 548 10.1021/acsapm.0c01151
40 Das S. Sarkar S. Das A. Das S. Chakraborty P. Sarkar J. A comprehensive review of various categories of face masks resistant to Covid-19 Clin. Epidemiol. Glob. Health 12 2021 100835 10.1016/j.cegh.2021.100835
41 Rossettie S. Perry C. Pourghaed M. Zumwalt M. Effectiveness of manufactured surgical masks, respirators, and home-made masks in prevention of respiratory infection due to airborne microorganisms Chronicles 8 2020 11 26 10.12746/swrccc.v8i34.675
42 Boi C. Borsetti F. Brugo T.M. Cappelletti M. De Angelis M.G. Fedi S. Di Giacomo S. Fabiani T. Foli G. Garelli A. Genchi U. Ghezzi D. Gualandi C. Lalli E. Magnani M. Maurizzi A. Mazzi F. Mehrabi N. Minelli M. Montalbano R. Morelli L. Nici S. Onesti R. Paglianti A. Papchenko K. Pappalardo S. Parisi N.F. Rapino S. Reggio M. Roselli M. Ruggeri E. Sabatini L. Saracino E. Scarponi G.E. Serra L. Signorini V. Storione A. Torsello M. Tugnoli E. Vargiu C.M. Vidali G. Violante F.S. One year of surgical mask testing at the University of Bologna labs: lessons learned from data analysis Sep. Purif. Technol. 294 2022 121180 10.1016/j.seppur.2022.121180
43 LaRue R.J. Morkus P. Laengert S. Rassenberg S. Halali M.A. Colenbrander J.W. Clase C.M. Latulippe D.R. Lannoy C. Navigating performance standards for face mask materials: a custom-built apparatus for measuring particle filtration efficiency Glob. Chall. 5 2021 2100052 10.1002/gch2.202100052
44 Illés B. Gordon P. Filtering efficiency measurement of respirators by laser-based particle counting method Measurement 176 2021 109173 10.1016/j.measurement.2021.109173
45 Rahman M.Z. Hoque M.E. Alam M.R. Rouf M.A. Khan S.I. Xu H. Ramakrishna S. Face masks to Combat Coronavirus (COVID-19)-processing, roles, requirements, efficacy, risk and sustainability Polymers 14 2022 1296 10.3390/polym14071296 35406172
46 Whyte H.E. Montigaud Y. Audoux E. Verhoeven P. Prier A. Leclerc L. Sarry G. Laurent C. Le Coq L. Joubert A. Pourchez J. Comparison of bacterial filtration efficiency vs. particle filtration efficiency to assess the performance of non-medical face masks Sci. Rep. 12 2022 1188 10.1038/s41598-022-05245-4 35075199
47 Centers for Disease Control and Prevention (CDC). Interim infection prevention and control recommendations for patients with suspected or confirmed Coronavirus Disease 2019 (COVID-19) in healthcare; updated Sept. 23, 2022. https://www.cdc.gov/coronavirus/2019-ncov/hcp/infection-control-recommendations.html#print, (n.d.).
48 World Health Organization (WHO). Clinical Management of COVID-19, Living guideline; updated jan. 13 2023. https://www.who.int/publications-detail-redirect/WHO-2019-nCoV-clinical-2023.1., (n.d.).
49 MacIntyre C.R. Wang Q. Rahman B. Seale H. Ridda I. Gao Z. Yang P. Shi W. Pang X. Zhang Y. Moa A. Dwyer D.E. Efficacy of face masks and respirators in preventing upper respiratory tract bacterial colonization and co-infection in hospital healthcare workers Prev. Med. 62 2014 1 7 10.1016/j.ypmed.2014.01.015 24472436
50 MacIntyre C.R. Wang Q. Seale H. Yang P. Shi W. Gao Z. Rahman B. Zhang Y. Wang X. Newall A.T. Heywood A. Dwyer D.E. A randomized clinical trial of three options for N95 respirators and medical masks in health workers Am. J. Respir. Crit. Care Med. 187 2013 960 966 10.1164/rccm.201207-1164OC 23413265
51 Radonovich L.J. Simberkoff M.S. Bessesen M.T. Brown A.C. Cummings D.A.T. Gaydos C.A. Los J.G. Krosche A.E. Gibert C.L. Gorse G.J. Nyquist A.C. Reich N.G. Rodriguez-Barradas M.C. Price C.S. Perl T.M. for the ResPECT investigators, N95 respirators vs medical masks for preventing influenza among health care personnel: a randomized clinical trial JAMA 322 2019 824 10.1001/jama.2019.11645 31479137
52 Loeb M. Dafoe N. Mahony J. John M. Sarabia A. Glavin V. Webby R. Smieja M. Earn D.J.D. Chong S. Webb A. Walter A.D. Surgical mask vs N95 respirator for preventing influenza among health care workers: a randomized trial, (2023). (Reprinted) JAMA 302 2009 10.1001/jama.2019.11645
53 Rashid T.U. Sharmeen S. Biswas S. Effectiveness of N95 masks against SARS-CoV-2: performance efficiency, concerns, and future directions ACS Chem. Health Saf. 29 2022 135 164 10.1021/acs.chas.1c00016
54 Bartoszko J.J. Farooqi M.A.M. Alhazzani W. Loeb M. Medical masks vs N95 respirators for preventing COVID-19 in healthcare workers: a systematic review and meta-analysis of randomized trials Influenza Other Respir. Viruses 14 2020 365 373 10.1111/irv.12745 32246890
55 Adanur S. Jayswal A. Filtration mechanisms and manufacturing methods of face masks: an overview J. Ind. Text. 51 2022 3683S 3717S 10.1177/1528083720980169
56 Tcharkhtchi A. Abbasnezhad N. Zarbini Seydani M. Zirak N. Farzaneh S. Shirinbayan M. An overview of filtration efficiency through the masks: mechanisms of the aerosols penetration Bioact. Mater. 6 2021 106 122 10.1016/j.bioactmat.2020.08.002 32817918
57 Todea A.M. Schmidt F. Schuldt T. Asbach C. Development of a method to determine the fractional deposition efficiency of full-scale HVAC and HEPA filter cassettes for nanoparticles ≥3.5 nm Atmosphere 11 2020 1191 10.3390/atmos11111191
58 I.A. Borojeni, G. Gajewski, R.A. Riahi, Application of electrospun nonwoven fibers in air filters, (2022) 27.
59 Gao Y. Tian E. Zhang Y. Mo J. Utilizing electrostatic effect in fibrous filters for efficient airborne particles removal: principles, fabrication, and material properties Appl. Mater. Today 26 2022 101369 10.1016/j.apmt.2022.101369
60 Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 37996 USA Tsai P. Performance of masks and discussion of the inactivation of SARS-CoV-2 Eng. Sci. 2020 10.30919/es8d1110
61 Grinshpun S.A. Yermakov M. Khodoun M. Autoclave sterilization and ethanol treatment of re-used surgical masks and N95 respirators during COVID-19: impact on their performance and integrity J. Hosp. Infect. 105 2020 608 614 10.1016/j.jhin.2020.06.030 32599011
62 Shen H. Han M. Shen Y. Shuai D. Electrospun nanofibrous membranes for controlling airborne viruses: present status, standardization of aerosol filtration tests, and future development ACS Environ. Au 2 2022 290 309 10.1021/acsenvironau.1c00047 35928556
63 Kwong L.H. Wilson R. Kumar S. Crider Y.S. Sanchez Y.R. Rempel D. Pillarisetti A. Review of the breathability and filtration efficiency of common household materials for face masks ACS Nano 15 2021 5904 5924 10.1021/acsnano.0c10146 33822580
64 Jones R.M. Rempel D. Standards for surgical respirators and masks: relevance for protecting healthcare workers and the public during pandemics Ann. Work Expo. Health 65 2021 495 504 10.1093/annweh/wxab008 33942848
65 Scarano A. Inchingolo F. Lorusso F. Facial skin temperature and discomfort when wearing protective face masks: thermal infrared imaging evaluation and hands moving the mask IJERPH 17 2020 4624 10.3390/ijerph17134624 32605056
66 Guan X. Lin J. Han J. Gao X. Zhang Y. Hu B. Guidoin R. Wang L. Prolonged use of surgical masks and respirators affects the protection and comfort for healthcare workers Materials 15 2022 7918 10.3390/ma15227918 36431403
67 Liao M. Liu H. Wang X. Hu X. Huang Y. Liu X. Brenan K. Mecha J. Nirmalan M. Lu J.R. A technical review of face mask wearing in preventing respiratory COVID-19 transmission Curr. Opin. Colloid Interface Sci. 52 2021 101417 10.1016/j.cocis.2021.101417
68 Buzzin A. Domènech-Gil G. Fraschetti E. Giovine E. Puglisi D. Caputo D. Assessing the consequences of prolonged usage of disposable face masks Sci. Rep. 12 2022 16796 10.1038/s41598-022-20692-9 36207345
69 Armand Q. Whyte H.E. Verhoeven P. Grattard F. Leclerc L. Curt N. Ragey S.P. Pourchez J. Impact of medical face mask wear on bacterial filtration efficiency and breathability Environ. Technol. Innov. 28 2022 102897 10.1016/j.eti.2022.102897
70 Russo F. Castro-Muñoz R. Santoro S. Galiano F. Figoli A. A review on electrospun membranes for potential air filtration application J. Environ. Chem. Eng. 10 2022 108452 10.1016/j.jece.2022.108452
71 W. Gilbert, P.F. Mottelay, De magnete, unabridged republ. of the 1893 ed, Dover, New York, 1991.
72 Lord Rayleigh X.X. On the equilibrium of liquid conducting masses charged with electricity Lond. Edinb. Dublin Philos. Mag. J. Sci. 14 1882 184 186 10.1080/14786448208628425
73 Zeleny J. Instability of electrified liquid surfaces Phys. Rev. 10 1917 1 6 10.1103/PhysRev.10.1
74 Cooley, J.F., Improved methods of and apparatus for electrically separating the relatively volatile liquid component from the component of relatively fixed substances of composite fluids. Patent number GB 06385, 1900, n.d.
75 A. Formhals, “Process and apparatus for preparing artificial threads,” US Patent 1975504, 1934, n.d.
76 A. Formhals, “Method and apparatus for spinning,” US Patent 2160962, 1939, n.d.
77 A. Formhals, “Artificial thread and method of producing same,” US Patent 2187306, 1940, n.d.
78 Bhardwaj N. Kundu S.C. Electrospinning: a fascinating fiber fabrication technique Biotechnol. Adv. 28 2010 325 347 10.1016/j.biotechadv.2010.01.004 20100560
79 Geoffrey T. Disintegration of water drops in an electric field Proc. R. Soc. Lond. A 280 1964 383 397 10.1098/rspa.1964.0151
80 Xue J. Wu T. Dai Y. Xia Y. Electrospinning and electrospun nanofibers: methods, materials, and applications Chem. Rev. 119 2019 5298 5415 10.1021/acs.chemrev.8b00593 30916938
81 Yang X. Wang J. Guo H. Liu L. Xu W. Duan G. Structural design toward functional materials by electrospinning: a review e-Polymers 20 2020 682 712 10.1515/epoly-2020-0068
82 Zhou Y. Liu Y. Zhang M. Feng Z. Yu D.G. Wang K. Electrospun nanofiber membranes for air filtration: a review Nanomaterials 12 2022 1077 10.3390/nano12071077 35407195
83 Shepa I. Mudra E. Dusza J. Electrospinning through the prism of time Mater. Today Chem. 21 2021 100543 10.1016/j.mtchem.2021.100543
84 Cai M. He H. Zhang X. Yan X. Li J. Chen F. Yuan D. Ning X. Efficient synthesis of PVDF/PI side-by-side bicomponent nanofiber membrane with enhanced mechanical strength and good thermal stability Nanomaterials 9 2018 39 10.3390/nano9010039 30597932
85 Han D. Steckl A.J. Coaxial electrospinning formation of complex polymer fibers and their applications ChemPlusChem 84 2019 1453 1497 10.1002/cplu.201900281 31943926
86 Zhang H. Liu N. Zeng Q. Liu J. Zhang X. Ge M. Zhang W. Li S. Fu Y. Zhang Y. Design of polypropylene electret melt blown nonwovens with superior filtration efficiency stability through thermally stimulated charging Polymers 12 2020 2341 10.3390/polym12102341 33066175
87 Zhang H. Liu J. Zhang X. Huang C. Jin X. Design of electret polypropylene melt blown air filtration material containing nucleating agent for effective PM2.5 capture RSC Adv. 8 2018 7932 7941 10.1039/C7RA10916D 35542038
88 Zhang J. Chen G. Bhat G.S. Azari H. Pen H. Electret characteristics of melt-blown polylactic acid fabrics for air filtration application J. Appl. Polym. Sci. 137 2020 48309 10.1002/app.48309
89 Espinoza-Montero P.J. Montero-Jiménez M. Rojas-Quishpe S. Alcívar León C.D. Heredia-Moya J. Rosero-Chanalata A. Orbea-Hinojosa C. Piñeiros J.L. Nude and modified electrospun nanofibers, application to air purification Nanomaterials 13 2023 593 10.3390/nano13030593 36770554
90 Sanyal A. Sinha-Ray S. Ultrafine PVDF nanofibers for filtration of air-borne particulate matters: a comprehensive review Polymers 13 2021 1864 10.3390/polym13111864 34205188
91 Leung W.W.F. Hung C.H. Skin effect in nanofiber filtration of submicron aerosols Sep. Purif. Technol. 92 2012 174 180 10.1016/j.seppur.2011.02.020
92 Li P. Wang C. Zhang Y. Wei F. Air filtration in the free molecular flow regime: a review of high-efficiency particulate air filters based on carbon nanotubes Small 10 2014 4543 4561 10.1002/smll.201401553 25288476
93 Xia T. Bian Y. Zhang L. Chen C. Relationship between pressure drop and face velocity for electrospun nanofiber filters Energy Build. 158 2018 987 999 10.1016/j.enbuild.2017.10.073
94 Li Y. Zhu J. Cheng H. Li G. Cho H. Jiang M. Gao Q. Zhang X. Developments of advanced electrospinning techniques: a critical review Adv. Mater. Technol. 6 2021 2100410 10.1002/admt.202100410
95 Bachs-Herrera A. Yousefzade O. del Valle L.J. Puiggali J. Melt electrospinning of polymers: blends, nanocomposites, additives and applications Appl. Sci. 11 2021 1808 10.3390/app11041808
96 Beaudoin É.J. Kubaski M.M. Samara M. Zednik R.J. Demarquette N.R. Scaled-up multi-needle electrospinning process using parallel plate auxiliary electrodes Nanomaterials 12 2022 1356 10.3390/nano12081356 35458064
97 Müller F. Jokisch S. Bargel H. Scheibel T. Centrifugal electrospinning enables the production of meshes of ultrathin polymer fibers ACS Appl. Polym. Mater. 2 2020 4360 4367 10.1021/acsapm.0c00853
98 Mingjun C. Youchen Z. Haoyi L. Xiangnan L. Yumei D. Bubakir M.M. Weimin Y. An example of industrialization of melt electrospinning: polymer melt differential electrospinning Adv. Ind. Eng. Polym. Res. 2 2019 110 115 10.1016/j.aiepr.2019.06.002
99 Lyu C. Zhao P. Xie J. Dong S. Liu J. Rao C. Fu J. Electrospinning of nanofibrous membrane and its applications in air filtration: a review Nanomaterials 11 2021 1501 10.3390/nano11061501 34204161
100 Wei L. Sun R. Liu C. Xiong J. Qin X. Mass production of nanofibers from needleless electrospinning by a novel annular spinneret Mater. Des. 179 2019 107885 10.1016/j.matdes.2019.107885
101 Niu H. Lin T. Fiber generators in needleless electrospinning J. Nanomater. 2012 2012 1 13 10.1155/2012/725950
102 Chen J. Yu Z. Li C. Lv Y. Hong S. Hu P. Liu Y. Review of the principles, devices, parameters, and applications for centrifugal electrospinning Macro Mater. Eng. 307 2022 2200057 10.1002/mame.202200057
103 Zhang Y. Zhang L. Cheng L. Qin Y. Li Y. Yang W. Li H. Efficient preparation of polymer nanofibers by needle roller electrospinning with low threshold voltage Polym. Eng. Sci. 59 2019 745 751 10.1002/pen.24993
104 Waqas M. Keirouz A. Sanira Putri M.K. Fazal F. Diaz Sanchez F.J. Ray D. Koutsos V. Radacsi N. Design and development of a nozzle-free electrospinning device for the high-throughput production of biomaterial nanofibers Med. Eng. Phys. 92 2021 80 87 10.1016/j.medengphy.2021.04.007 34167715
105 He H.J. Liu C.K. Molnar K. A novel needleless electrospinning system using a moving conventional yarn as the spinneret Fibers Polym. 19 2018 1472 1478 10.1007/s12221-018-8183-2
106 Zheng G. Jiang J. Wang X. Li W. Zhong W. Guo S. Self-cleaning threaded rod spinneret for high-efficiency needleless electrospinning Appl. Phys. A 124 2018 473 10.1007/s00339-018-1892-y
107 Wei L. Liu C. Mao X. Dong J. Fan W. Zhi C. Qin X. Sun R. Multiple-jet needleless electrospinning approach via a linear flume spinneret Polymers 11 2019 2052 10.3390/polym11122052 31835650
108 Xiong J. Liu Y. Li A. Wei L. Wang L. Qin X. Yu J. Mass production of high-quality nanofibers via constructing pre-Taylor cones with high curvature on needleless electrospinning Mater. Des. 197 2021 109247 10.1016/j.matdes.2020.109247
109 Hassan M.A. Ali U. Shahzad A. Qadir M.B. Khaliq Z. Nazir A. Abid S. Khan M.Q. Hussain T. Bullet-Spinneret based needleless electrospinning; a versatile way to fabricate continuous nanowebs at low voltage Mater. Res. Express 6 2018 025053 10.1088/2053-1591/aaf137
110 Xu Y. Li X. Xiang H.F. Zhang Q.Q. Wang X.X. Yu M. Hao L.Y. Long Y.Z. Large-scale preparation of polymer nanofibers for air filtration by a new multineedle electrospinning device J. Nanomater. 2020 2020 1 7 10.1155/2020/4965438
111 Vass P. Démuth B. Farkas A. Hirsch E. Szabó E. Nagy B. Andersen S.K. Vigh T. Verreck G. Csontos I. Marosi G. Nagy Z.K. Continuous alternative to freeze drying: manufacturing of cyclodextrin-based reconstitution powder from aqueous solution using scaled-up electrospinning J. Control. Release 298 2019 120 127 10.1016/j.jconrel.2019.02.019 30779951
112 Hirsch E. Vass P. Demuth B. Petho Zs. Bitay E. Andersen S.K. Vigh T. Verreck G. Molnar K. Nagy Z.K. Marosi G. Electrospinning scale-up and formulation development of PVA nanofibers aiming oral delivery of biopharmaceuticals Express Polym. Lett. 13 2019 590 603 10.3144/expresspolymlett.2019.50
113 Duan G. Greiner A. Air-blowing-assisted coaxial electrospinning toward high productivity of core/sheath and hollow fibers Macromol. Mater. Eng. 304 2019 1800669 10.1002/mame.201800669
114 Kwak B.E. Yoo H.J. Lee E. Kim D.H. Large-scale centrifugal multispinning production of polymer micro- and nanofibers for mask filter application with a potential of cospinning mixed multicomponent fibers ACS Macro Lett. 10 2021 382 388 10.1021/acsmacrolett.0c00829 34192093
115 Wang L. Wang B. Ahmad Z. Li J.S. Chang M.W. Dual rotation centrifugal electrospinning: a novel approach to engineer multi-directional and layered fiber composite matrices Drug Deliv. Transl. Res. 9 2019 204 214 10.1007/s13346-018-00594-y 30341763
116 Liu Y. Tan J. Yu S. Yousefzadeh M. Lyu T. Jiao Z. Li H. Ramakrishna S. High-efficiency preparation of polypropylene nanofiber by melt differential centrifugal electrospinning J. Appl. Polym. Sci. 137 2020 48299 10.1002/app.48299
117 Koenig K. Langensiepen F. Seide G. Pilot-scale production of polylactic acid nanofibers by melt electrospinning e-Polymers 20 2020 233 241 10.1515/epoly-2020-0030
118 Araldi da Silva B. de Sousa Cunha R. Valério A. De Noni Junior A. Hotza D. Gómez González S.Y. Electrospinning of cellulose using ionic liquids: an overview on processing and applications Eur. Polym. J. 147 2021 110283 10.1016/j.eurpolymj.2021.110283
119 Naziri Mehrabani S.A. Vatanpour V. Koyuncu I. Green solvents in polymeric membrane fabrication: a review Sep. Purif. Technol. 298 2022 121691 10.1016/j.seppur.2022.121691
120 Song W. Zhao X. Jin Z. Fan L. Ji X. Deng J. Duan J. Poly(vinyl alcohol) for multi-functionalized corrosion protection of metals: a review J. Clean. Prod. 394 2023 136390 10.1016/j.jclepro.2023.136390
121 Rai P. Mehrotra S. Priya S. Gnansounou E. Sharma S.K. Recent advances in the sustainable design and applications of biodegradable polymers Bioresour. Technol. 325 2021 124739 10.1016/j.biortech.2021.124739
122 Kakoria A. Sinha-Ray S. A review on biopolymer-based fibers via electrospinning and solution blowing and their applications Fibers 6 2018 45 10.3390/fib6030045
123 Ibrahim Y. Hussein E. Zagho M. Abdo G. Elzatahry A. Melt electrospinning designs for nanofiber fabrication for different applications IJMS 20 2019 2455 10.3390/ijms20102455 31109002
124 Li H. Chen H. Zhong X. Wu W. Ding Y. Yang W. Interjet distance in needleless melt differential electrospinning with umbellate nozzles J. Appl. Polym. Sci. 131 2014 10.1002/app.40515 n/a-n/a
125 Chen H. Li H. Ma X. He W. Tan J. Yang W. Large scaled fabrication of microfibers by air-suction assisted needleless melt electrospinning Fibers Polym. 17 2016 576 581 10.1007/s12221-016-5915-z
126 Qin Y. Cheng L. Zhang Y. Chen X. Wang X. He X. Yang W. An Y. Li H. Efficient preparation of poly(lactic acid) nanofibers by melt differential electrospinning with addition of acetyl tributyl citrate: research article J. Appl. Polym. Sci. 135 2018 46554 10.1002/app.46554
127 Yu M. Dong R.H. Yan X. Yu G.F. You M.H. Ning X. Long Y.Z. Recent advances in needleless electrospinning of ultrathin fibers: from academia to industrial production Macromol. Mater. Eng. 302 2017 1700002 10.1002/mame.201700002
128 Omer S. Forgách L. Zelkó R. Sebe I. Scale-up of electrospinning: market overview of products and devices for pharmaceutical and biomedical purposes Pharmaceutics 13 2021 286 10.3390/pharmaceutics13020286 33671624
129 Ibrahim H.M. Klingner A. A review on electrospun polymeric nanofibers: production parameters and potential applications Polym. Test. 90 2020 106647 10.1016/j.polymertesting.2020.106647
130 Islam M.S. Ang B.C. Andriyana A. Afifi A.M. A review on fabrication of nanofibers via electrospinning and their applications SN Appl. Sci. 1 2019 1248 10.1007/s42452-019-1288-4
131 Szewczyk P.K. Stachewicz U. The impact of relative humidity on electrospun polymer fibers: from structural changes to fiber morphology Adv. Colloid Interface Sci. 286 2020 102315 10.1016/j.cis.2020.102315
132 Rashid T.U. Gorga R.E. Krause W.E. Mechanical properties of electrospun fibers-a critical review Adv. Eng. Mater. 23 2021 2100153 10.1002/adem.202100153
133 Medeiros G.B. de A. Lima F. de Almeida D.S. Guerra V.G. Aguiar M.L. Modification and functionalization of fibers formed by electrospinning: a review Membranes 12 2022 861 10.3390/membranes12090861 36135880
134 V. Naragund, P.K. Panda, Electrospinning of polyacrylonitrile nanofiber membrane for bacteria removal, (2018). 10.5281/ZENODO.3344558.
135 Baby T. Jose E T. George G. Varkey V. Cherian S.K. A new approach for the shaping up of very fine and beadless UV light absorbing polycarbonate fibers by electrospinning Polym. Test. 80 2019 106103 10.1016/j.polymertesting.2019.106103
136 Dehghan S.F. Golbabaei F. Maddah B. Latifi M. Pezeshk H. Hasanzadeh M. Akbar-Khanzadeh F. Optimization of electrospinning parameters for polyacrylonitrile-MgO nanofibers applied in air filtration J. Air Waste Manag. Assoc. 66 2016 912 921 10.1080/10962247.2016.1162228 27192035
137 He H. Kara Y. Molnár K. In situ viscosity-controlled electrospinning with a low threshold voltage Macromol. Mater. Eng. 304 2019 1900349 10.1002/mame.201900349
138 Bian Y. Wang S. Zhang L. Chen C. Influence of fiber diameter, filter thickness, and packing density on PM2.5 removal efficiency of electrospun nanofiber air filters for indoor applications Build. Environ. 170 2020 106628 10.1016/j.buildenv.2019.106628
139 Manea L.R. Bertea A.P. Bertea A. Progress in electrospun nanofibres for air filtration IOP Conf. Ser. Mater. Sci. Eng. 877 2020 012043 10.1088/1757-899X/877/1/012043
140 Selatile M.K. Ray S.S. Ojijo V. Sadiku R. Depth filtration of airborne agglomerates using electrospun bio-based polylactide membranes J. Environ. Chem. Eng. 6 2018 762 772 10.1016/j.jece.2017.12.070
141 Li Y. Yin X. Yu J. Ding B. Electrospun nanofibers for high-performance air filtration Compos. Commun. 15 2019 6 19 10.1016/j.coco.2019.06.003
142 Zuo F. Zhang S. Liu H. Fong H. Yin X. Yu J. Ding B. Free-standing polyurethane nanofiber/nets air filters for effective PM capture Small 13 2017 1702139 10.1002/smll.201702139
143 Zhang S. Liu H. Zuo F. Yin X. Yu J. Ding B. A controlled design of ripple-like polyamide-6 nanofiber/nets membrane for high-efficiency air filter Small 13 2017 1603151 10.1002/smll.201603151
144 Al-Attabi R. Dumée L.F. Schütz J.A. Morsi Y. Pore engineering towards highly efficient electrospun nanofibrous membranes for aerosol particle removal Sci. Total Environ. 625 2018 706 715 10.1016/j.scitotenv.2017.12.342 29306158
145 Kim H. Lee S. Shin Y.R. Choi Y.N. Yoon J. Ryu M. Lee J.W. Lee H. Durable superhydrophobic poly(vinylidene fluoride) (PVDF)-based nanofibrous membranes for reusable air filters ACS Appl. Polym. Mater. 4 2022 338 347 10.1021/acsapm.1c01314
146 Moon J. Bui T.T. Jang S. Ji S. Park J.T. Kim M.G. A highly efficient nanofibrous air filter membrane fabricated using electrospun amphiphilic PVDF-g-POEM double comb copolymer Sep. Purif. Technol. 279 2021 119625 10.1016/j.seppur.2021.119625
147 Sohrabi M. Abbasi M. Sadighzadeh A. Fabrication and evaluation of electrospun polyacrylonitrile/silver nanofiber membranes for air filtration and antibacterial activity Polym. Bull. 2022 10.1007/s00289-022-04311-1
148 Lakshmanan A. Gavali D.S. Thapa R. Sarkar D. Synthesis of CTAB-functionalized large-scale nanofibers air filter media for efficient PM2.5 capture capacity with low airflow resistance ACS Appl. Polym. Mater 3 2021 937 948 10.1021/acsapm.0c01203
149 Dai H. Liu X. Zhang C. Ma K. Zhang Y. Electrospinning polyacrylonitrile/graphene oxide/polyimide nanofibrous membranes for high-efficiency PM2.5 filtration Sep. Purif. Technol. 276 2021 119243 10.1016/j.seppur.2021.119243
150 Xu Y. Zhang X. Hao X. Teng D. Zhao T. Zeng Y. Micro/nanofibrous nonwovens with high filtration performance and radiative heat dissipation property for personal protective face mask Chem. Eng. J. 423 2021 130175 10.1016/j.cej.2021.130175
151 Karabulut F.N.H. Höfler G. Ashok Chand N. Beckermann G.W. Electrospun nanofibre filtration media to protect against biological or nonbiological airborne particles Polymers 13 2021 3257 10.3390/polym13193257 34641073
152 Yang Z. Zhang X. Qin Z. Li H. Wang J. Zeng G. Liu C. Long J. Zhao Y. Li Y. Yan G. Airflow synergistic needleless electrospinning of instant noodle-like curly nanofibrous membranes for high-efficiency air filtration Small 18 2022 2107250 10.1002/smll.202107250
153 Jiang J. Shao Z. Wang X. Zhu P. Deng S. Li W. Zheng G. Three-dimensional composite electrospun nanofibrous membrane by multi-jet electrospinning with sheath gas for high-efficiency antibiosis air filtration Nanotechnology 32 2021 245707 10.1088/1361-6528/abeb9a
154 Li M. Wen H. Li H. Yan Z.C. Li Y. Wang L. Wang D. Tang B.Z. AIEgen-loaded nanofibrous membrane as photodynamic/photothermal antimicrobial surface for sunlight-triggered bioprotection Biomaterials 276 2021 121007 10.1016/j.biomaterials.2021.121007
155 Kumar S. Jang J. Oh H. Jung B.J. Lee Y. Park H. Yang K.H. Seong Y.C. Lee J.S. AntibacteriAL POLYMERIC NANOFIBERS FROM ZWITTERIONIC TERPOLYMERS BY ELECTROSPINNING FOR AIR FILTRATIon ACS Appl. Nano Mater. 4 2021 2375 2385 10.1021/acsanm.0c02366
156 Pierpaoli M. Giosuè C. Czerwińska N. Rycewicz M. Wieloszyńska A. Bogdanowicz R. Ruello M.L. Characterization and filtration efficiency of sustainable PLA fibers obtained via a hybrid 3D-printed/electrospinning technique Materials 14 2021 6766 10.3390/ma14226766 34832169
157 Wang L. Gao Y. Xiong J. Shao W. Cui C. Sun N. Zhang Y. Chang S. Han P. Liu F. He J. Biodegradable and high-performance multiscale structured nanofiber membrane as mask filter media via poly(lactic acid) electrospinning J. Colloid Interface Sci. 606 2022 961 970 10.1016/j.jcis.2021.08.079 34487943
158 Shao Z. Chen Y. Jiang J. Xiao Y. Kang G. Wang X. Li W. Zheng G. Multistage-split ultrafine fluffy nanofibrous membrane for high-efficiency antibacterial air filtration ACS Appl. Mater. Interfaces 14 2022 18989 19001 10.1021/acsami.2c04700 35436100
159 Kwon M. Kim J. Kim J. Photocatalytic activity and filtration performance of hybrid TiO2-cellulose acetate nanofibers for air filter applications Polymers 13 2021 1331 10.3390/polym13081331 33921671
160 Xu Y. Zhang X. Teng D. Zhao T. Li Y. Zeng Y. Multi-layered micro/nanofibrous nonwovens for functional face mask filter Nano Res. 15 2022 7549 7558 10.1007/s12274-022-4350-2 35578617
161 Jensen M.G. O'Shaughnessy P.T. Shaffer M. Yu S. Choi Y.Y. Christiansen M. Stanier C.O. Hartley M. Huddle J. Johnson J. Bibby K. Myung N.V. Cwiertny D.M. Simple fabrication of an electrospun polystyrene microfiber filter that meets N95 filtering facepiece respirator filtration and breathability standards J Appl. Polym. Sci. 140 2023 10.1002/app.53406
162 Chen P. Yang Z. Mai Z. Huang Z. Bian Y. Wu S. Dong X. Fu X. Ko F. Zhang S. Zheng W. Zhang S. Zhou W. Electrospun nanofibrous membrane with antibacterial and antiviral properties decorated with Myoporum bontioides extract and silver-doped carbon nitride nanoparticles for medical masks application Sep. Purif. Technol. 298 2022 121565 10.1016/j.seppur.2022.121565
163 Li Q. Yin Y. Cao D. Wang Y. Luan P. Sun X. Liang W. Zhu H. Photocatalytic rejuvenation enabled self-sanitizing, reusable, and biodegradable masks against COVID-19 ACS Nano 15 2021 11992 12005 10.1021/acsnano.1c03249 34170122
164 Baselga-Lahoz M. Yus C. Arruebo M. Sebastián V. Irusta S. Jiménez S. Submicronic filtering media based on electrospun recycled PET nanofibers: development, characterization, and method to manufacture surgical masks Nanomaterials 12 2022 925 10.3390/nano12060925 35335738
165 Munir M.M. Adrian M. Burhanuddin M. Iskandar F. Fabrication and structure optimization of expanded polystyrene (EPS) waste fiber for high-performance air filtration Powder Technol. 402 2022 117357 10.1016/j.powtec.2022.117357
166 Blosi M. Costa A.L. Ortelli S. Belosi F. Ravegnani F. Varesano A. Tonetti C. Zanoni I. Vineis C. Polyvinyl alcohol/silver electrospun nanofibers: biocidal filter media capturing virus-size particles J Appl. Polym. Sci. 138 2021 51380 10.1002/app.51380 34511635
167 Lou C.W. Lin M.C. Huang C.H. Lai M.F. Shiu B.C. Lin J.H. Preparation of needleless electrospinning polyvinyl alcohol/water-soluble chitosan nanofibrous membranes: antibacterial property and filter efficiency Polymers 14 2022 1054 10.3390/polym14051054 35267878
168 Deng Y. Lu T. Zhang X. Zeng Z. Tao R. Qu Q. Zhang Y. Zhu M. Xiong R. Huang C. Multi-hierarchical nanofiber membrane with typical curved-ribbon structure fabricated by green electrospinning for efficient, breathable and sustainable air filtration J. Memb. Sci. 660 2022 120857 10.1016/j.memsci.2022.120857
169 Mounesan M. Akbari S. Brycki B.E. Needleless electrospun mats based on polyamidoamine dendritic polymers for encapsulation of essential oils in personal respiratory equipment J. Ind. Text. 51 2022 6333S 6352S 10.1177/15280837211048155
170 Shen H. Zhou Z. Wang H. Chen J. Zhang M. Han M. Shen Y. Shuai D. Photosensitized electrospun nanofibrous filters for capturing and killing airborne coronaviruses under visible light irradiation Environ. Sci. Technol. 56 2022 4295 4304 10.1021/acs.est.2c00885 35262328
171 Karabulut F.N.H. Fomra D. Höfler G. Chand N.A. Beckermann G.W. Virucidal and bactericidal filtration media from electrospun polylactic acid nanofibres capable of protecting against COVID-19 Membranes 12 2022 571 10.3390/membranes12060571 35736278
172 Li Y. Yin X. Si Y. Yu J. Ding B. All-polymer hybrid electret fibers for high-efficiency and low-resistance filter media Chem. Eng. J. 398 2020 125626 10.1016/j.cej.2020.125626
173 Patil N.A. Gore P.M. Jaya Prakash N. Govindaraj P. Yadav R. Verma V. Shanmugarajan D. Patil S. Kore A. Kandasubramanian B. Needleless electrospun phytochemicals encapsulated nanofibre based 3-ply biodegradable mask for combating COVID-19 pandemic Chem. Eng. J. 416 2021 129152 10.1016/j.cej.2021.129152
174 Lu W.C. Chen C.Y. Cho C.J. Venkatesan M. Chiang W.H. Yu Y.Y. Lee C.H. Lee R.H. Rwei S.P. Kuo C.C. Antibacterial activity and protection efficiency of polyvinyl butyral nanofibrous membrane containing thymol prepared through vertical electrospinning Polymers 13 2021 1122 10.3390/polym13071122 33916011
175 Panda P.K. Gangwar A. Thite A.G. Optimization of Nylon 6 electrospun nanofiber diameter in needle-less wire electrode using central composite design and response surface methodology J. Ind. Text. 51 2022 7279S 7292S 10.1177/15280837211058213
176 Wang J. Liu S. Yan X. Jiang Z. Zhou Z. Liu J. Han G. Ben H. Jiang W. Biodegradable and reusable cellulose-based nanofiber membrane preparation for mask filter by electrospinning Membranes 12 2021 23 10.3390/membranes12010023 35054548
177 Cui J. Wang Y. Lu T. Liu K. Huang C. High performance, environmentally friendly and sustainable nanofiber membrane filter for removal of particulate matter 1.0 J. Colloid Interface Sci. 597 2021 48 55 10.1016/j.jcis.2021.03.174 33866211
178 Cui J. Lu T. Li F. Wang Y. Lei J. Ma W. Zou Y. Huang C. Flexible and transparent composite nanofibre membrane that was fabricated via a “green” electrospinning method for efficient particulate matter 2.5 capture J. Colloid Interface Sci. 582 2021 506 514 10.1016/j.jcis.2020.08.075 32911399
179 Naragund V.S. Panda P.K. Electrospun polyacrylonitrile nanofiber membranes for air filtration application Int. J. Environ. Sci. Technol. 19 2022 10233 10244 10.1007/s13762-021-03705-4
180 Aamer H. Heo S. Jo Y. Characterization of multifunctional PAN /ZnO nanofibrous composite filter for fine dust capture and photocatalytic activity J. Appl. Polym. Sci. 138 2021 50607 10.1002/app.50607
181 Orlando R. Gao Y. Fojan P. Mo J. Afshari A. Filtration performance of ultrathin electrospun cellulose acetate filters doped with TiO2 and activated charcoal Buildings 11 2021 557 10.3390/buildings11110557
182 Xiong J. Li A. Liu Y. Wang L. Qin X. Yu J. Multi-scale nanoarchitectured fibrous networks for high-performance, self-sterilization, and recyclable face masks Small 18 2022 2105570 10.1002/smll.202105570
183 Kadam V. Truong Y.B. Schutz J. Kyratzis I.L. Padhye R. Wang L. Gelatin/β–cyclodextrin bio–nanofibers as respiratory filter media for filtration of aerosols and volatile organic compounds at low air resistance J. Hazard. Mater. 403 2021 123841 10.1016/j.jhazmat.2020.123841
184 Ding S. Cao Y. Huang F. Wang Y. Li J. Chen S. Spontaneous polarization induced electrostatic charge in washable electret composite fabrics for reusable air-filtering application Compos. Sci. Technol. 217 2022 109093 10.1016/j.compscitech.2021.109093
185 Hu J. Xiong Z. Liu Y. Lin J. A biodegradable composite filter made from electrospun zein fibers underlaid on the cellulose paper towel Int. J. Biol. Macromol. 204 2022 419 428 10.1016/j.ijbiomac.2022.02.029 35151705
186 Han W. Rao D. Gao H. Yang X. Fan H. Li C. Dong L. Meng H. Green-solvent-processable biodegradable poly(lactic acid) nanofibrous membranes with bead-on-string structure for effective air filtration: kill two birds with one stone Nano Energy 97 2022 107237 10.1016/j.nanoen.2022.107237
187 Avossa J. Batt T. Pelet T. Sidjanski S.P. Schönenberger K. Rossi R.M. Polyamide nanofiber-based air filters for transparent face masks ACS Appl. Nano Mater. 4 2021 12401 12406 10.1021/acsanm.1c02843
188 Aamer H. Kim S.B. Oh J.M. Park H. Jo Y.M. ZnO-impregnated polyacrylonitrile nanofiber filters against various phases of air pollutants Nanomaterials 11 2021 2313 10.3390/nano11092313 34578627
189 Bonfim D.P.F. Cruz F.G.S. Bretas R.E.S. Guerra V.G. Aguiar M.L. A sustainable recycling alternative: electrospun PET-membranes for air nanofiltration Polymers 13 2021 1166 10.3390/polym13071166 33916472
190 Su Q. Wei Z. Wang X. Long S. Zeng W. Wang S. Yang J. Electrospun composite membrane based on polyarylene sulfide sulfone/Ag/ZnO nanofibers for antibacterial effective PM2.5 filtration J. Appl. Polym. Sci. 139 2022 51693 10.1002/app.51693
191 Du W. Iacoviello F. Fernandez T. Loureiro R. Brett D.J.L. Shearing P.R. Microstructure analysis and image-based modelling of face masks for COVID-19 virus protection Commun. Mater. 2 2021 69 10.1038/s43246-021-00160-z
192 Van Goethem C. Op de Beeck D. Ilyas A. Thijs M. Koeckelberghs G. Aerts P.E.M. Vankelecom I.F.J. Ultra-thin and highly porous PVDF-filters prepared via phase inversion for potential medical (COVID-19) and industrial use J. Memb. Sci. 639 2021 119710 10.1016/j.memsci.2021.119710
193 Yim W. Cheng D. Patel S.H. Kou R. Meng Y.S. Jokerst J.V. KN95 and N95 respirators retain filtration efficiency despite a loss of dipole charge during decontamination ACS Appl. Mater. Interfaces 12 2020 54473 54480 10.1021/acsami.0c17333 33253527
194 Mirković M. Stojanović D.B. Mijailović D. Barać N. Janaćković Đ. Uskoković P.S. Electrospun polyacrylonitrile fibers incorporated with microporous carbon for improved airborne PM2.5 filtration Mater. Chem. Phys. 285 2022 126103 10.1016/j.matchemphys.2022.126103
195 Yadav P. Ismail N. Essalhi M. Tysklind M. Athanassiadis D. Tavajohi N. Assessment of the environmental impact of polymeric membrane production J. Memb. Sci. 622 2021 118987 10.1016/j.memsci.2020.118987
196 Saadat S. Rawtani D. Hussain C.M. Environmental perspective of COVID-19 Sci. Total Environ. 728 2020 138870 10.1016/j.scitotenv.2020.138870
197 Sangkham S. Face mask and medical waste disposal during the novel COVID-19 pandemic in Asia Case Stud. Chem. Environ. Eng. 2 2020 100052 10.1016/j.cscee.2020.100052
198 Aragaw T.A. Surgical face masks as a potential source for microplastic pollution in the COVID-19 scenario Mar. Pollut. Bull. 159 2020 111517 10.1016/j.marpolbul.2020.111517
199 Du H. Huang S. Wang J. Environmental risks of polymer materials from disposable face masks linked to the COVID-19 pandemic Sci. Total Environ. 815 2022 152980 10.1016/j.scitotenv.2022.152980
200 Chamas A. Moon H. Zheng J. Qiu Y. Tabassum T. Jang J.H. Abu-Omar M. Scott S.L. Suh S. Degradation rates of plastics in the environment ACS Sustain. Chem. Eng. 8 2020 3494 3511 10.1021/acssuschemeng.9b06635
201 Luo H. Liu C. He D. Xu J. Sun J. Li J. Pan X. Environmental behaviors of microplastics in aquatic systems: a systematic review on degradation, adsorption, toxicity and biofilm under aging conditions J. Hazard. Mater. 423 2022 126915 10.1016/j.jhazmat.2021.126915
202 Mghili B. Analla M. Aksissou M. Face masks related to COVID-19 in the beaches of the Moroccan Mediterranean: an emerging source of plastic pollution Mar. Pollut. Bull. 174 2022 113181 10.1016/j.marpolbul.2021.113181
203 Pan Z. Liu Q. Sun Y. Sun X. Lin H. Environmental implications of microplastic pollution in the Northwestern Pacific Ocean Mar. Pollut. Bull. 146 2019 215 224 10.1016/j.marpolbul.2019.06.031 31426149
204 Oldal D.G. Topuz F. Holtzl T. Szekely G. Green electrospinning of biodegradable cellulose acetate nanofibrous membranes with tunable porosity ACS Sustain. Chem. Eng. 11 2023 994 1005 10.1021/acssuschemeng.2c05676
205 Hidalgo-Crespo J. Moreira C.M. Jervis F.X. Soto M. Amaya J.L. Banguera L. Circular economy of expanded polystyrene container production: environmental benefits of household waste recycling considering renewable energies Energy Rep. 8 2022 306 311 10.1016/j.egyr.2022.01.071
206 Sadeghi B. Marfavi Y. AliAkbari R. Kowsari E. Borbor Ajdari F. Ramakrishna S. Recent studies on recycled PET fibers: production and applications: a review Mater. Circ. Econ. 3 2021 4 10.1007/s42824-020-00014-y
207 FTIR-2020-PLA-COVID(aggiungere Review)(f).pdf, (2023).
208 Li K. Clarkson C.M. Wang L. Liu Y. Lamm M. Pang Z. Zhou Y. Qian J. Tajvidi M. Gardner D.J. Tekinalp H. Hu L. Li T. Ragauskas A.J. Youngblood J.P. Ozcan S. Alignment of cellulose nanofibers: harnessing nanoscale properties to macroscale benefits ACS Nano 15 2021 3646 3673 10.1021/acsnano.0c07613 33599500
209 Lamm M.E. Li K. Qian J. Wang L. Lavoine N. Newman R. Gardner D.J. Li T. Hu L. Ragauskas A.J. Tekinalp H. Kunc V. Ozcan S. Recent advances in functional materials through cellulose nanofiber templating Adv. Mater. 33 2021 2005538 10.1002/adma.202005538
210 Nagarajan K.J. Ramanujam N.R. Sanjay M.R. Siengchin S. Surya Rajan B. Sathick Basha K. Madhu P. Raghav G.R. A comprehensive review on cellulose nanocrystals and cellulose nanofibers: pretreatment, preparation, and characterization Polym. Compos. 42 2021 1588 1630 10.1002/pc.25929
211 Wong D. Hartery S. Keltie E. Chang R. Kim J.S. Park S.S. Electrospun polystyrene and acid-treated cellulose nanocrystals with intense pulsed light treatment for N95-equivalent filters ACS Appl. Polym. Mater. 3 2021 4949 4958 10.1021/acsapm.1c00722
212 Akduman C. Cellulose acetate and polyvinylidene fluoride nanofiber mats for N95 respirators J. Ind. Text. 50 2021 1239 1261 10.1177/1528083719858760
213 de Almeida D.S. Martins L.D. Muniz E.C. Rudke A.P. Squizzato R. Beal A. de Souza P.R. Bonfim D.P.F. Aguiar M.L. Gimenes M.L. Biodegradable CA/CPB electrospun nanofibers for efficient retention of airborne nanoparticles Process Saf. Environ. Prot. 144 2020 177 185 10.1016/j.psep.2020.07.024 32834561
214 Yadav N. Hakkarainen M. Degradable or not? Cellulose acetate as a model for complicated interplay between structure, environment and degradation Chemosphere 265 2021 128731 10.1016/j.chemosphere.2020.128731
215 Samadian H. Zamiri S. Ehterami A. Farzamfar S. Vaez A. Khastar H. Alam M. Ai A. Derakhshankhah H. Allahyari Z. Goodarzi A. Salehi M. Electrospun cellulose acetate/gelatin nanofibrous wound dressing containing berberine for diabetic foot ulcer healing: in vitro and in vivo studies Sci. Rep. 10 2020 8312 10.1038/s41598-020-65268-7 32433566
216 Vu X.H. Duong T.T.T. Pham T.T.H. Trinh D.K. Nguyen X.H. Dang V.S. Synthesis and study of silver nanoparticles for antibacterial activity against Escherichia coli and Staphylococcus aureus Adv. Nat. Sci. Nanosci. Nanotechnol. 9 2018 025019 10.1088/2043-6254/aac58f
217 Priyanka K.P. Sukirtha T.H. Balakrishna K.M. Varghese T. Microbicidal activity of TiO2 nanoparticles synthesised by sol–gel method IET Nanobiotechnol. 10 2016 81 86 10.1049/iet-nbt.2015.0038 27074858
218 Ahamed M. Alhadlaq H.A. Khan M.A.M. Karuppiah P. Al-Dhabi N.A. Synthesis, characterization, and antimicrobial activity of copper oxide nanoparticles J. Nanotechnol. 2014 2014 1 4 10.1155/2014/637858
219 Lallo da Silva B. Caetano B.L. Chiari-Andréo B.G. Pietro R.C.L.R. Chiavacci L.A. Increased antibacterial activity of ZnO nanoparticles: influence of size and surface modification Colloids Surf. B 177 2019 440 447 10.1016/j.colsurfb.2019.02.013
220 Maji J. Pandey S. Basu S. Synthesis and evaluation of antibacterial properties of magnesium oxide nanoparticles Bull. Mater. Sci. 43 2020 25 10.1007/s12034-019-1963-5
221 Beyth N. Houri-Haddad Y. Domb A. Khan W. Hazan R. Alternative antimicrobial approach: nano-antimicrobial materials Evid. Based Complement. Altern. Med. 2015 2015 1 16 10.1155/2015/246012
222 Karagoz S. Kiremitler N.B. Sarp G. Pekdemir S. Salem S. Goksu A.G. Onses M.S. Sozdutmaz I. Sahmetlioglu E. Ozkara E.S. Ceylan A. Yilmaz E. Antibacterial, antiviral, and self-cleaning mats with sensing capabilities based on electrospun nanofibers decorated with ZnO nanorods and Ag nanoparticles for protective clothing applications ACS Appl. Mater. Interfaces 13 2021 5678 5690 10.1021/acsami.0c15606 33492946
223 Sportelli M.C. Izzi M. Loconsole D. Sallustio A. Picca R.A. Felici R. Chironna M. Cioffi N. On the efficacy of ZnO nanostructures against SARS-CoV-2 IJMS 23 2022 3040 10.3390/ijms23063040 35328455
224 Li Y. Liu Y. Yao B. Narasimalu S. Dong Z. Rapid preparation and antimicrobial activity of polyurea coatings with RE-Doped nano-ZnO Microb. Biotechnol. 15 2022 548 560 10.1111/1751-7915.13891 34676986
225 Gudkov S.V. Burmistrov D.E. Serov D.A. Rebezov M.B. Semenova A.A. Lisitsyn A.B. A mini review of antibacterial properties of ZnO nanoparticles Front. Phys. 9 2021 641481 10.3389/fphy.2021.641481
226 Skalny A. Rink L. Ajsuvakova O. Aschner M. Gritsenko V. Alekseenko S. Svistunov A. Petrakis D. Spandidos D. Aaseth J. Tsatsakis A. Tinkov A. Zinc and respiratory tract infections: perspectives for COVID‑19 (Review) Int. J. Mol. Med. 2020 10.3892/ijmm.2020.4575
227 Alexander J. Tinkov A. Strand T.A. Alehagen U. Skalny A. Aaseth J. Early nutritional interventions with zinc, selenium and vitamin D for raising anti-viral resistance against progressive COVID-19 Nutrients 12 2020 2358 10.3390/nu12082358 32784601
228 Maares M. Hackler J. Haupt A. Heller R.A. Bachmann M. Diegmann J. Moghaddam A. Schomburg L. Haase H. Free zinc as a predictive marker for COVID-19 mortality risk Nutrients 14 2022 1407 10.3390/nu14071407 35406020
229 Dada A.O. Inyinbor A.A. Bello O.S. Tokula B.E. Novel plantain peel activated carbon–supported zinc oxide nanocomposites (PPAC-ZnO-NC) for adsorption of chloroquine synthetic pharmaceutical used for COVID-19 treatment Biomass Convers. Biorefin. 2021 10.1007/s13399-021-01828-9
230 Ghareeb D.A. Saleh S.R. Seadawy M.G. Nofal M.S. Abdulmalek S.A. Hassan S.F. Khedr S.M. AbdElwahab M.G. Sobhy A.A. saber A. Abdel-Hamid A. Yassin A.M. Elmoneam A.A.A. Masoud A.A. Kaddah M.M.Y. El-Zahaby S.A. Al-mahallawi A.M. El-Gharbawy A.M. Zaki A. k. Seif I. Kenawy M.Y. Amin M. Amer K. El Demellawy M.A. Nanoparticles of ZnO/Berberine complex contract COVID-19 and respiratory co-bacterial infection in addition to elimination of hydroxychloroquine toxicity J. Pharm. Investig. 51 2021 735 757 10.1007/s40005-021-00544-w
231 Geetha K. Sivasangari D. Kim H.S. Murugadoss G. Kathalingam A. Electrospun nanofibrous ZnO/PVA/PVP composite films for efficient antimicrobial face masks Ceram. Int. 48 2022 29197 29204 10.1016/j.ceramint.2022.05.164
232 Goncharova D.A. Bolbasov E.N. Nemoykina A.L. Aljulaih A.A. Tverdokhlebova T.S. Kulinich S.A. Svetlichnyi V.A. Structure and properties of biodegradable PLLA/ZnO composite membrane produced via electrospinning Materials 14 2020 2 10.3390/ma14010002 33374987
233 Nageh H. Emam M.H. Ali F. Abdel Fattah N.F. Taha M. Amin R. Kamoun E.A. Loutfy S.A. Kasry A. Zinc oxide nanoparticle-loaded electrospun polyvinylidene fluoride nanofibers as a potential face protector against respiratory viral infections ACS Omega 7 2022 14887 14896 10.1021/acsomega.2c00458 35557678
234 Alshabanah L.A. Omran N. Elwakil B.H. Hamed M.T. Abdallah S.M. Al-Mutabagani L.A. Wang D. Liu Q. Shehata N. Hassanin A.H. Hagar M. Elastic nanofibrous membranes for medical and personal protection applications: manufacturing, anti-COVID-19, and anti-colistin resistant bacteria evaluation Polymers 13 2021 3987 10.3390/polym13223987 34833289
235 Salam A. Hassan T. Jabri T. Riaz S. Khan A. Iqbal K.M. ullah Khan S. Wasim M. Shah M.R. Khan M.Q. Kim I.S. Electrospun nanofiber-based viroblock/ZnO/PAN hybrid antiviral nanocomposite for personal protective applications Nanomaterials 11 2021 2208 10.3390/nano11092208 34578527
236 Baselga M. Uranga-Murillo I. de Miguel D. Arias M. Sebastián V. Pardo J. Arruebo M. Silver nanoparticles–polyethyleneimine-based coatings with antiviral activity against SARS-CoV-2: a new method to functionalize filtration media Materials 15 2022 4742 10.3390/ma15144742 35888208
237 El-Nahhal I.M. Salem J. Kodeh F.S. Elmanama A. Anbar R. CuO–NPs, CuO–Ag nanocomposite and Cu(II)-curcumin complex coated cotton/starched cotton antimicrobial materials Mater. Chem. Phys. 285 2022 126099 10.1016/j.matchemphys.2022.126099
238 Foffa I. Losi P. Quaranta P. Cara A. Al Kayal T. D'Acunto M. Presciuttini G. Pistello M. Soldani G. A Copper nanoparticles-based polymeric spray coating: nanoshield against Sars-Cov-2 J. Appl. Biomater. Funct. Mater. 20 2022 228080002210763 10.1177/22808000221076326
239 Costa S.M. Pacheco L. Antunes W. Vieira R. Bem N. Teixeira P. Fangueiro R. Ferreira D.P. Antibacterial and biodegradable electrospun filtering membranes for facemasks: an attempt to reduce disposable masks use Appl. Sci. 12 2021 67 10.3390/app12010067
240 Kubo A.L. Rausalu K. Savest N. Žusinaite E. Vasiliev G. Viirsalu M. Plamus T. Krumme A. Merits A. Bondarenko O. Antibacterial and antiviral effects of Ag, Cu and Zn metals, respective nanoparticles and filter materials thereof against Coronavirus SARS-CoV-2 and Influenza A virus Pharmaceutics 14 2022 2549 10.3390/pharmaceutics14122549 36559043
241 Wang S. Zhao X. Yin X. Yu J. Ding B. Electret polyvinylidene fluoride nanofibers hybridized by polytetrafluoroethylene nanoparticles for high-efficiency air filtration ACS Appl. Mater. Interfaces 8 2016 23985 23994 10.1021/acsami.6b08262 27552028
242 Bui T.T. Shin M.K. Jee S.Y. Long D.X. Hong J. Kim M.G. Ferroelectric PVDF nanofiber membrane for high-efficiency PM0.3 air filtration with low air flow resistance Colloids Surf. A 640 2022 128418 10.1016/j.colsurfa.2022.128418
243 Li C.X. Kuang S.Y. Chen Y.H. Wang Z.L. Li C. Zhu G. In situ active poling of nanofiber networks for gigantically enhanced particulate filtration ACS Appl. Mater. Interfaces 10 2018 24332 24338 10.1021/acsami.8b07203 29979875
244 Leung W.W.F. Sun Q. Electrostatic charged nanofiber filter for filtering airborne novel Coronavirus (COVID-19) and nano-aerosols Sep. Purif. Technol. 250 2020 116886 10.1016/j.seppur.2020.116886
245 Zhang J. Gong S. Wang C. Jeong D. Wang Z.L. Ren K. Biodegradable electrospun poly(lactic acid) nanofibers for effective PM 2.5 removal Macromol. Mater. Eng. 304 2019 1900259 10.1002/mame.201900259
246 Shao Z. Jiang J. Wang X. Li W. Fang L. Zheng G. Self-powered electrospun composite nanofiber membrane for highly efficient air filtration Nanomaterials 10 2020 1706 10.3390/nano10091706 32872502
247 Yongliang C. Chunya W. Junwen Z. Shizhe L. Yongjun X. Qize Z. Hulin J. Nan W. Wenbo L. Shuwen C. Bo W. Yingying Z. Jun Z. Electrospun polyetherimide electret nonwoven for bi-functional smart face mask Nano Energy 8 2017 10.1016/j.nanoen.2017.03.011
248 Mariello M. Qualtieri A. Mele G. De Vittorio M. Metal-free multilayer hybrid PENG based on soft electrospun/-sprayed membranes with cardanol additive for harvesting energy from surgical face masks ACS Appl. Mater. Interfaces 13 2021 20606 20621 10.1021/acsami.1c01740 33896167
249 Cho M. Hiremath V. Seo J.G. Triboelectrification-based particulate matter capture utilizing electrospun ethyl cellulose and PTFE spheres Atmos. Environ. X 12 2021 100138 10.1016/j.aeaoa.2021.100138 1-10
250 Hu Y. Wang Y. Tian S. Yu A. Wan L. Zhai J. Performance-enhanced and washable triboelectric air filter based on polyvinylidene fluoride/UiO-66 composite nanofiber membrane Macro Mater. Eng. 306 2021 2100128 10.1002/mame.202100128
251 Fu Q. Liu Y. Liu T. Mo J. Zhang W. Zhang S. Luo B. Wang J. Qin Y. Wang S. Nie S. Air-permeable cellulosic triboelectric materials for self-powered healthcare products Nano Energy 102 2022 107739 10.1016/j.nanoen.2022.107739
252 Liu G. Nie J. Han C. Jiang T. Yang Z. Pang Y. Xu L. Guo T. Bu T. Zhang C. Wang Z.L. Self-powered electrostatic adsorption face mask based on a triboelectric nanogenerator ACS Appl. Mater. Interfaces 10 2018 7126 7133 10.1021/acsami.7b18732 29394045
253 Gu G.Q. Han C.B. Tian J.J. Jiang T. He C. Lu C.X. Bai Y. Nie J.H. Li Z. Wang Z.L. Triboelectric nanogenerator enhanced multilayered antibacterial nanofiber air filters for efficient removal of ultrafine particulate matter Nano Res. 11 2018 4090 4101 10.1007/s12274-018-1992-1
254 Hao R. Yang S. Yang K. Zhang Z. Wang T. Sang S. Zhang H. Self-powered air filter based on an electrospun respiratory triboelectric nanogenerator ACS Appl. Energy Mater. 4 2021 14700 14708 10.1021/acsaem.1c03328
255 Kim W.G. Kim D.W. Tcho I.W. Kim J.K. Kim M.S. Choi Y.K. Triboelectric nanogenerator: structure, mechanism, and applications ACS Nano 15 2021 258 287 10.1021/acsnano.0c09803 33427457
256 Kang D.H. Kim N.K. Kang H.W. Electrostatic charge retention in PVDF nanofiber-nylon mesh multilayer structure for effective fine particulate matter filtration for face masks Polymers 13 2021 3235 10.3390/polym13193235 34641051
257 He H. Guo J. Illés B. Géczy A. Istók B. Hliva V. Török D. Kovács J.G. Harmati I. Molnár K. Monitoring multi-respiratory indices via a smart nanofibrous mask filter based on a triboelectric nanogenerator Nano Energy 89 2021 106418 10.1016/j.nanoen.2021.106418
258 Lin S. Wang S. Yang W. Chen S. Xu Z. Mo X. Zhou H. Duan J. Hu B. Huang L. Trap-induced dense monocharged perfluorinated electret nanofibers for recyclable multifunctional healthcare mask ACS Nano 15 2021 5486 5494 10.1021/acsnano.1c00238 33626276
259 Le T.T. Curry E.J. Vinikoor T. Das R. Liu Y. Sheets D. Tran K.T.M. Hawxhurst C.J. Stevens J.F. Hancock J.N. Bilal O.R. Shor L.M. Nguyen T.D. Piezoelectric nanofiber membrane for reusable, stable, and highly functional face mask filter with long-term biodegradability Adv. Funct. Mater. 32 2022 2113040 10.1002/adfm.202113040
260 Cho Y. Son Y. Ahn J. Lim H. Ahn S. Lee J. Bae P.K. Kim I.D. Multifunctional filter membranes based on self-assembled core–shell biodegradable nanofibers for persistent electrostatic filtration through the triboelectric effect ACS Nano 16 2022 19451 19463 10.1021/acsnano.2c09165 36374248
261 Wang N. Feng Y. Zheng Y. Zhang L. Feng M. Li X. Zhou F. Wang D. New hydrogen bonding enhanced polyvinyl alcohol based self-charged medical mask with superior charge retention and moisture resistance performances Adv. Funct. Mater. 31 2021 2009172 10.1002/adfm.202009172
262 Mi Y. Lu Y. Shi Y. Zhao Z. Wang X. Meng J. Cao X. Wang N. Biodegradable polymers in triboelectric nanogenerators Polymers 15 2022 222 10.3390/polym15010222 36616571
263 Rostami M. Beheshtizadeh N. Ranjbar F.E. Najafi N. Ahmadi A. Ahmadi P. Rostamabadi H. Pazhouhnia Z. Assadpour E. Mirzanajafi-Zanjani M. Kisomi M.F. Kharazmi M.S. Jafari S.M. Recent advances in electrospun protein fibers/nanofibers for the food and biomedical applications Adv. Colloid Interface Sci. 311 2023 102827 10.1016/j.cis.2022.102827
264 De Sio L. Ding B. Focsan M. Kogermann K. Pascoal-Faria P. Petronela F. Mitchell G. Zussman E. Pierini F. Personalized reusable face masks with smart nano-assisted destruction of pathogens for COVID-19: a visionary road Chem. Eur. J. 27 2021 6112 6130 10.1002/chem.202004875 33284500
265 Li D.T.S. Samaranayake L.P. Leung Y.Y. Neelakantan P. Facial protection in the era of COVID-19: a narrative review Oral Dis. 27 2021 665 673 10.1111/odi.13460 32506757
266 Rubio-Romero J.C. del C. Pardo-Ferreira M. Torrecilla-García J.A. Calero-Castro S. Disposable masks: disinfection and sterilization for reuse, and non-certified manufacturing, in the face of shortages during the COVID-19 pandemic Saf. Sci. 129 2020 104830 10.1016/j.ssci.2020.104830
267 Zhu J. Jiang Q. He X. Li X. Wang L. Zheng L. Jing P. Chen M. Filtration efficiency of N95 filtering facepiece respirators during multi-cycles of 8-hour simulated donning + disinfection J. Hosp. Infect. 127 2022 91 100 10.1016/j.jhin.2022.06.016 35792275
268 Liao C. Li Y. Tjong S.C. Visible-light active titanium dioxide nanomaterials with bactericidal properties Nanomaterials 10 2020 124 10.3390/nano10010124 31936581
269 Mohite V.S. Darade M.M. Sharma R.K. Pawar S.H. Nanoparticle engineered photocatalytic paints: a roadmap to self-sterilizing against the spread of communicable diseases Catalysts 12 2022 326 10.3390/catal12030326
270 Kundu S. Karak N. Polymeric photocatalytic membrane: an emerging solution for environmental remediation Chem. Eng. J. 438 2022 135575 10.1016/j.cej.2022.135575
271 Kaushik N. Mitra S. Baek E.J. Nguyen L.N. Bhartiya P. Kim J.H. Choi E.H. Kaushik N.K. The inactivation and destruction of viruses by reactive oxygen species generated through physical and cold atmospheric plasma techniques: current status and perspectives J. Adv. Res. 43 2023 59 71 10.1016/j.jare.2022.03.002 36585115
272 Tang Z. Ma D. Chen Q. Wang Y. Sun M. Lian Q. Shang J. Wong P.K. He C. Xia D. Wang T. Nanomaterial-enabled photothermal-based solar water disinfection processes: fundamentals, recent advances, and mechanisms J. Hazard. Mater. 437 2022 129373 10.1016/j.jhazmat.2022.129373
273 Wang L. Feng Y. Wang K. Liu G. Solar water sterilization enabled by photothermal nanomaterials Nano Energy 87 2021 106158 10.1016/j.nanoen.2021.106158
274 Wu Z. Zheng H. Gu J. Li F. Lv R. Deng Y. Xu W. Tong Y. Effects of different temperature and time durations of virus inactivation on results of real-time fluorescence PCR testing of COVID-19 viruses Curr. Med. Sci. 40 2020 614 617 10.1007/s11596-020-2224-y 32767258
275 Zhong H. Zhu Z. Lin J. Cheung C.F. Lu V.L. Yan F. Chan C.Y. Li G. Reusable and recyclable graphene masks with outstanding superhydrophobic and photothermal performances ACS Nano 14 2020 6213 6221 10.1021/acsnano.0c02250 32329600
276 Kumar P. Roy S. Sarkar A. Jaiswal A. Reusable MoS 2 -modified antibacterial fabrics with photothermal disinfection properties for repurposing of personal protective masks ACS Appl. Mater. Interfaces 13 2021 12912 12927 10.1021/acsami.1c00083 33715350
277 Soni R. Joshi S.R. Karmacharya M. Min H. Kim S.K. Kumar S. Kim G.H. Cho Y.K. Lee C.Y. Superhydrophobic and self-sterilizing surgical masks spray-coated with carbon nanotubes ACS Appl. Nano Mater. 4 2021 8491 8499 10.1021/acsanm.1c01082
278 Yan S. Stackhouse C.A. Waluyo I. Hunt A. Kisslinger K. Head A.R. Bock D.C. Takeuchi E.S. Takeuchi K.J. Wang L. Marschilok A.C. Reusing face covering masks: probing the impact of heat treatment ACS Sustain. Chem. Eng. 9 2021 13545 13558 10.1021/acssuschemeng.1c04530 35855909
279 Liao L. Xiao W. Zhao M. Yu X. Wang H. Wang Q. Chu S. Cui Y. Can N95 respirators be reused after disinfection? How many times? ACS Nano 14 2020 6348 6356 10.1021/acsnano.0c03597 32368894
280 He W. Guo Y. Gao H. Liu J. Yue Y. Wang J. Evaluation of regeneration processes for filtering facepiece respirators in terms of the bacteria inactivation efficiency and influences on filtration performance ACS Nano 14 2020 13161 13171 10.1021/acsnano.0c04782 32975412
281 Lv D. Zhu M. Jiang Z. Jiang S. Zhang Q. Xiong R. Huang C. Green electrospun nanofibers and their application in air filtration Macromol. Mater. Eng. 303 2018 1800336 10.1002/mame.201800336
282 Avossa J. Herwig G. Toncelli C. Itel F. Rossi R.M. Electrospinning based on benign solvents: current definitions, implications and strategies Green Chem. 24 2022 2347 2375 10.1039/D1GC04252A
283 Kim D. Nunes S.P. Green solvents for membrane manufacture: recent trends and perspectives Curr. Opin. Green Sustain. Chem. 28 2021 100427 10.1016/j.cogsc.2020.100427
284 Xie W. Li T. Tiraferri A. Drioli E. Figoli A. Crittenden J.C. Liu B. Toward the next generation of sustainable membranes from green chemistry principles ACS Sustain. Chem. Eng. 9 2021 50 75 10.1021/acssuschemeng.0c07119
285 Zhong T. Liu W. Liu H. Green electrospinning of chitin propionate to manufacture nanofiber mats Carbohydr. Polym. 273 2021 118593 10.1016/j.carbpol.2021.118593
286 Mosher C.Z. Brudnicki P.A.P. Gong Z. Childs H.R. Lee S.W. Antrobus R.M. Fang E.C. Schiros T.N. Lu H.H. Green electrospinning for biomaterials and biofabrication Biofabrication 13 2021 035049 10.1088/1758-5090/ac0964
287 Lasseuguette E. Malpass-Evans R. Casalini S. McKeown N.B. Ferrari M.C. Optimization of the fabrication of amidoxime modified PIM-1 electrospun fibres for use as breathable and reactive materials Polymer 213 2021 123205 10.1016/j.polymer.2020.123205
288 Zhou W. Yu X. Cao L. Yang M. Li Y. Si Y. Yu J. Ding B. Green and ethanol-resistant polyurethane nanofibrous membranes based on an ethanol solvent for waterproof and breathable textiles Adv. Sustain. Syst. 4 2020 2000105 10.1002/adsu.202000105
289 Jiang S. Hou H. Agarwal S. Greiner A. Polyimide nanofibers by “Green” electrospinning via aqueous solution for filtration applications ACS Sustain. Chem. Eng. 4 2016 4797 4804 10.1021/acssuschemeng.6b01031
290 Gaydhane M.K. Sharma C.S. Majumdar S. Electrospun nanofibres in drug delivery: advances in controlled release strategies RSC Adv. 13 2023 7312 7328 10.1039/D2RA06023J 36891485
291 Abbas W.A. Shaheen B.S. Ghanem L.G. Badawy I.M. Abodouh M.M. Abdou S.M. Zada S. Allam N.K. Cost-effective face mask filter based on hybrid composite nanofibrous layers with high filtration efficiency Langmuir 37 2021 7492 7502 10.1021/acs.langmuir.1c00926 34101479
292 Doyle A.D. Nazari S.S. Yamada K.M. Cell–extracellular matrix dynamics Phys. Biol. 19 2022 021002 10.1088/1478-3975/ac4390
293 Tan G.Z. Zhou Y. Electrospinning of biomimetic fibrous scaffolds for tissue engineering: a review Int. J. Polym. Mater. Polym. Biomater. 69 2020 947 960 10.1080/00914037.2019.1636248
294 Owida H.A. Al-Nabulsi J.I. Alnaimat F. Al-Ayyad M. Turab N.M. Al Sharah A. Shakur M. Recent applications of electrospun nanofibrous scaffold in tissue engineering Appl. Bionics Biomech. 2022 2022 1 15 10.1155/2022/1953861
295 Rickel A.P. Deng X. Engebretson D. Hong Z. Electrospun nanofiber scaffold for vascular tissue engineering Mater. Sci. Eng. C 129 2021 112373 10.1016/j.msec.2021.112373
296 Jin G. He R. Sha B. Li W. Qing H. Teng R. Xu F. Electrospun three-dimensional aligned nanofibrous scaffolds for tissue engineering Mater. Sci. Eng. C 92 2018 995 1005 10.1016/j.msec.2018.06.065
297 Wu S. Liu J. Cai J. Zhao J. Duan B. Chen S. Combining electrospinning with hot drawing process to fabricate high performance poly (L-lactic acid) nanofiber yarns for advanced nanostructured bio-textiles Biofabrication 13 2021 045018 10.1088/1758-5090/ac2209
298 Liu J. Li T. Zhang H. Zhao W. Qu L. Chen S. Wu S. Electrospun strong, bioactive, and bioabsorbable silk fibroin/poly (L-lactic-acid) nanoyarns for constructing advanced nanotextile tissue scaffolds Mater. Today Bio 14 2022 100243 10.1016/j.mtbio.2022.100243
299 Qi Y. Wang C. Wang Q. Zhou F. Li T. Wang B. Su W. Shang D. Wu S. A simple, quick, and cost-effective strategy to fabricate polycaprolactone/silk fibroin nanofiber yarns for biotextile-based tissue scaffold application Eur. Polym. J. 186 2023 111863 10.1016/j.eurpolymj.2023.111863
300 Zhang X. Guo M. Guo Q. Liu N. Wang Y. Wu T. Modulating axonal growth and neural stem cell migration with the use of uniaxially aligned nanofiber yarns welded with NGF-loaded microparticles Mater. Today Adv. 17 2023 100343 10.1016/j.mtadv.2023.100343
301 Wu S. Qi Y. Shi W. Kuss M. Chen S. Duan B. Electrospun conductive nanofiber yarns for accelerating mesenchymal stem cells differentiation and maturation into Schwann cell-like cells under a combination of electrical stimulation and chemical induction Acta Biomater. 139 2022 91 104 10.1016/j.actbio.2020.11.042 33271357
302 Wan X. Cong H. Jiang G. Liang X. Liu L. He H. A review on PVDF nanofibers in textiles for flexible piezoelectric sensors ACS Appl. Nano Mater. 6 2023 1522 1540 10.1021/acsanm.2c04916
303 Fan W. Zhang Y. Sun Y. Wang S. Zhang C. Yu X. Wang W. Dong K. Durable antibacterial and temperature regulated core-spun yarns for textile health and comfort applications Chem. Eng. J. 455 2023 140917 10.1016/j.cej.2022.140917
304 Zhang W. Miao J. Tian M. Zhang X. Fan T. Qu L. Hierarchically interlocked helical conductive yarn enables ultra-stretchable electronics and smart fabrics Chem. Eng. J. 462 2023 142279 10.1016/j.cej.2023.142279
305 Fadil F. Affandi N.D.N. Misnon M.I. Bonnia N.N. Harun A.M. Alam M.K. Review on electrospun nanofiber-applied products Polymers 13 2021 2087 10.3390/polym13132087 34202857
306 Persano L. Camposeo A. Tekmen C. Pisignano D. Industrial upscaling of electrospinning and applications of polymer nanofibers: a review Macromol. Mater. Eng. 298 2013 504 520 10.1002/mame.201200290
307 F. Gomollón-Be, Nanofibers put a new spin on COVID-19 masks, Special to C&EN. 98, issue 45 (2020). https://cen.acs.org/materials/nanomaterials/Nanofibers-put-new-spin-COVID/98/i45.
|
PMC010xxxxxx/PMC10154158.txt |
==== Front
J Adolesc Health
J Adolesc Health
The Journal of Adolescent Health
1054-139X
1879-1972
Society for Adolescent Health and Medicine. Published by Elsevier Inc.
S1054-139X(23)00151-9
10.1016/j.jadohealth.2023.02.041
Original Article
Significant Increase in Deliberate Self-Poisonings Among Adolescents During the Second Year of the COVID-19 Pandemic
Koppen Arjen Ph.D. a∗
Thoonen Ilze M.J. M.Sc. a1
Hunault Claudine C. Ph.D., M.D. a1
van Velzen Agnes G. Ph.D. a
de Lange Dylan W. Ph.D., M.D. ab
Rietjens Saskia J. Ph.D. a
a Dutch Poisons Information Center (DPIC), University Medical Center Utrecht, Utrecht, the Netherlands
b Department of Intensive Care Medicine, University Medical Center Utrecht, Utrecht, the Netherlands
∗ Address correspondence to: Arjen Koppen, Ph.D., University Medical Centre Utrecht, Dutch Poisons Information Center, Utrecht, Netherlands.
1 These authors contributed equally to this work.
3 5 2023
3 5 2023
26 10 2022
28 2 2023
© 2023 Society for Adolescent Health and Medicine. Published by Elsevier Inc.
2023
Society for Adolescent Health and Medicine
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Purpose
The COVID-19 pandemic has been associated with a decline in mental health of adolescents. The aim of this study was to analyze the rate of deliberate self-poisonings (DSPs) among adolescents reported to the Dutch Poisons Information Center before and during the COVID-19 pandemic.
Methods
A retrospective study from 2016 until 2021 was performed to characterize DSPs among adolescents, and to analyze trends in the number of DSPs. All DSPs among adolescents with the age of 13 up to and including 17 years were included. DSP characteristics included: age, gender, bodyweight, used substance, dose, and treatment advice. Trends in the number of DSPs were analyzed using time series decomposition and Seasonal Autoregressive Integrated Moving Average models.
Results
Six thousand nine hundred fifteen DSPs in adolescents were recorded from January first 2016 until December 31st 2021. Females were involved in 84% of adolescent DSPs. A strong increase in the number of DSPs was observed in 2021 (45% increase compared to 2020), which deviated from the predicted trend based on previous years. This increase was most prominent in 13-, 14-, and 15-year-old female adolescents. Commonly involved drugs were paracetamol, ibuprofen, methylphenidate, fluoxetine, and quetiapine. The contribution of paracetamol rose from 33% in 2019 to 40% in 2021.
Discussion
The strong increase in the number of DSPs during the second year of the COVID-19 pandemic suggests that long-term containment measures such as quarantines, lockdowns, and school closures may enhance self-harm behavior among adolescents, especially among younger females (13–15 years of age), with a preference for paracetamol as DSP substance.
Keywords
Adolescent
COVID-19
Deliberate
Intoxication
Overdose
Pandemic
Self-harm
Self-poisoning
==== Body
pmc Implications and Contribution
The COVID-19 pandemic has been associated with a decline in mental health of adolescents. This study supports this association by showing a considerable increase in deliberate self-poisonings reported to the Dutch Poisons Information Center, predominantly among female adolescents, during the second year of the COVID-19 pandemic. This increase was especially prominent in younger adolescents (13–15 years). Over-the-counter analgesics paracetamol and ibuprofen were most often used in deliberate self-poisonings among adolescents.
The COVID-19 pandemic has been associated with a decline in mental health, in particular in adolescents and young adults. This may be due to measures taken to limit the spread of the SARS-CoV-2 virus, such as quarantines, lockdowns, and school closures. In the Netherlands, mitigation measures started in March 2020, with a general lockdown preventing children and adolescents to enter school, sport clubs, or other social gatherings. This was followed by strict and less strict measures alternating until the first quarter of 2022. Early studies showed a COVID-19 pandemic-related decrease in social interactions, and an increase in depression and anxiety in adolescents and young adults [1,2]. Such effects could translate into an increase in self-harm behavior, including deliberate self-poisonings (DSPs).
In Western society, DSP is the predominant method by which adolescents intentionally harm themselves [[3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14]]. Over-the-counter (OTC) analgesics, in particular paracetamol and ibuprofen, are the predominant substances used for DSPs by adolescents, followed by antidepressants and sedatives/hypnotics [[6], [7], [8], [9],15]. The preferred DSP substances may differ between boys and girls, with girls predominantly using analgesics, and boys more prone to using sedatives, hypnotics and/or antipsychotics [9]. Several risk factors for DSPs among adolescents have been described: female gender [[7], [8], [9], [10], [11],[15], [16], [17]], older age [9,10,15], school terms/school pressure [7,15], and socio-economic deprivation [18].
Only a few reports have been published studying the effect of the COVID-19 pandemic on DSP rate among adolescents. In the United States, data from the Poisons Control Centers' National Poison Data System showed an increase in the proportion of adolescent intentional exposures with a suicide intent during the first 1.5 year of the COVID-19 pandemic compared to the prepandemic period, with a higher proportion of patients with moderate/major clinical effects and deaths [19]. The data of the French National Database of Poisonings showed a strong increase in DSPs among 12–24-year-olds starting halfway through the first year of the COVID-19 pandemic, particularly among females, although no distinction was made between adolescents and young adults [20]. In neither of these studies an adolescent-specific trend analysis including the first 2 years of the pandemic was performed, nor were specific age groups analyzed separately.
The aim of the present study was a) to analyze trends in the number of adolescent DSPs reported to the Dutch Poisons Information Center (DPIC), comparing the first 2 years of the COVID-19 pandemic (2020–2021) with the prepandemic years 2016–2019, and b) to further characterize these DSPs in terms of demographics and exposure type.
Materials and Methods
Study design and patient population
A retrospective analysis of adolescent DSPs reported to the DPIC from January first 2016 until December 31st 2021was performed, covering 2 years of the COVID-19 pandemic (In the Netherlands, March 11th 2020 is regarded as the starting date of the COVID-19 pandemic). Data were subtracted from the DPIC's database, which consists of inquiries received by telephone from Dutch health care professionals about acute intoxications. In this database, anonymous case information is recorded using a standard data format, to ensure uniform data collection. Inquiries involving adolescents (from 13 to and including 17 years old) who exposed themselves to substances with the (suspected) intent of self-harm were included. Exposures to drugs of abuse or alcohol with no clear intent of self-harm were excluded. Accidental exposures, for instance medication errors, were also excluded. Cases were assessed against the inclusion and exclusion criteria by three authors and any disagreements were resolved through consensus. For each DSP, the following parameters were registered: date and time of the inquiry, patient characteristics (age, gender, and bodyweight), exposure characteristics (involved substance, dose, route, day of exposure), intent (accidental/intentional), symptoms at the time of DPIC consultation, and treatment advice. Multiple inquiries regarding the same patient and same exposure(s) were analyzed as a single case. The monthly and yearly number of DSPs and the DSP rate (number of monthly adolescent DSPs relative to total monthly number of received inquiries about human intoxications) were determined. The DSP rate was included to rule out that possible trends among adolescents were caused by overall changes in the total number of inquiries to the DPIC (regardless of age). In addition, data regarding the total number of adolescents living in the Netherlands from 2016 until 2021 was obtained from the Central Office for Statistics [21] and was compared to trends in adolescent DSPs reported to the DPIC.
The Dutch Medical Research Involving Human Subjects Act did not apply to this retrospective cohort study, since the obtained information was anonymous and not traceable to personal data.
Statistical analyzes
Descriptive statistics (percentage, median, interquartile range) were used to provide an overview of patient characteristics (e.g., age, gender), exposure characteristics (e.g., type of exposures), and treatment advice. Analyzes were conducted using IBM SPSS Statistics (version 26.0.0.1; IBM, Armonk, NY) and R studio version 2022.12.0 for Windows (R version 4.2.2; Boston, MA). The DSP rate over time was considered as a time series with the “ts” function in R and decomposed into three components (“trend,” “seasonal,” and “random”). After removing the trend of the time series, we fitted a Seasonal Autoregressive Integrated Moving Average (SARIMA) model on prepandemic data in the Netherlands (2016–February 2020). Normality of the residuals and goodness of fit of the model were checked. This SARIMA model was used to predict what the DSP rate would have been in 2021–2022 without a pandemic (with 95% confidence intervals). Observed DSP rate and DSP rate predicted by the SARIMA model were compared. For further information, see the supplementary data. Further, we considered the time series of DSP rates separately for school days and weekend days. The average number of DSPs per school day (Monday–Friday) and weekend day (Saturday–Sunday) was calculated for each month. Since the pattern and seasonal effects were similar for the average number of DSPs during school days and weekend days, a paired t test was performed to compare both types of days. Finally, we also analyzed the time series of DSP rates separately for girls and boys. Chi-square tests (IBM SPSS 26.0.0.1) were performed in order to compare the DPIC's treatment advice between DSPs with and without paracetamol.
Dutch poisons information center procedure and definitions
The DPIC provides a 24/7 telephone service giving expert advice to Dutch health care professionals on the diagnosis and treatment of patients exposed to potentially toxic substances. Advice can be requested on a voluntary basis whenever needed. The term “exposure” is defined as an actual or suspected contact with any substance through ingestion, inhalation, absorption, application to, or injection into the body. A mono-intoxication is defined as contact with only one substance, whereas a multi-intoxication is defined as contact with two or more substances. Not all exposures reported to the DPIC result in toxic effects, since this depends on the dose the patient has been exposed to. For each patient, the DPIC performs a risk assessment, based on reported dose of exposure (mg/kg) and symptoms at the time of DPIC consultation, resulting in one of the following types of treatment advice (based on a pure medical toxicological point of view): observation at home (when estimated to be no or a mild intoxication) or further medical examination by a physician, followed by hospital observation if necessary (when estimated to be a potentially moderate/severe intoxication). Obviously, observation in a hospital or psychiatric institution could also be warranted because of psychiatric reasons; in the Netherlands, this judgment is made by the patient's physician (e.g., general practitioner or emergency physician).
Results
Patient characteristics and trend analysis
From January first 2016 until December 31st 2021 the DPIC received a total of 208,626 inquiries concerning acute human intoxications. A total number of 10,875 inquiries involved adolescents of 13 to and including 17 years of age, of which 6,915 were included as adolescent DSP (Figure 1 ). Eighty-four percent of the DSPs involved female adolescents (Table A1). The median age of both adolescent males and females was 16 years (interquartile range 15–17 years). The time series decomposition showed a general trend; the monthly number of DSPs slowly increased between 2016 and mid-2018, then decreased from 2019 to mid-2020, followed by an abrupt increase from mid-2020. The observed DSP rate well exceeded the 95% confidence interval of the DSP rate predicted by the SARIMA model in March and November 2021 (Figure 2 , Figure A1A). The number of DSPs increased with 45% in 2021 compared to 2020 (n = 1,512 vs. n = 1,044, respectively, Table A1). The increase in DSP rate was much higher among girls than among boys, with an increase of 50% for girls versus 20% for boys in 2021 compared to 2020 (Table A1 and Figure A1B). Hence, as further confirmed by the SARIMA model (data not shown), the overall increase in adolescent DSPs was mainly attributable to the increase in female adolescent DSPs. Moreover, the increase was most prominent among 13-, 14-, and 15-years olds (49%, 75%, and 77% increase (2021 vs. 2020) respectively), although the number of DSPs also increased among 16- and 17-year old adolescents (27% and 21% increase (2021 vs. 2020), respectively) (Table A1, Figures A1C and D). The total number of DPIC consultations (regardless of age) did not increase during this period (Figure A1C).Figure 1 Patient selection flowchart, covering all inclusions from January first 2016 until December 31st 2021.
Figure 2 Trends in DSPs by adolescents. (A) SARIMA model of expected rate of DSPs in 2021. The red line represents the percentage of adolescent DSPs relative to the total number of patients reported to the DPIC, per month. The blue line represents the expected percentage of adolescent DSPs from March 2020 onward. CI = confidence interval. (B) DSP rates for girls and boys. Each dot represents the monthly DSP rate (percentage of male (gray circles) or female (black circles) adolescent DSPs in a particular month relative to the total number of patients reported to the DPIC in that month). Dotted lines represent fourth order polynomial trend lines.
Furthermore, a clear seasonal effect was observed, with lower numbers of DSPs in July and August. Interestingly, the number of DSPs was also dependent on the day of the week: throughout the study period, the frequency of DSPs during weekend days was generally lower than during school days (paired t test, p value < .001) (Figure A1E), with Monday displaying the highest frequency of DSPs and Saturday the lowest (Figure A1F). Taken together, our data show a strong increase in the number of adolescent DSPs starting in the second half of the first year of the COVID-19 pandemic, and continuing in the second year.
Characteristics of substances used in deliberate self-poisonings
Approximately two-thirds (n = 4,828, 70%) of DSPs involved exposure to a single substance (mono-intoxication) and one-third (n = 2,087, 30%) comprised DSPs with exposure to two or more substances (multi-intoxication) (Table A1). The most commonly used substances were paracetamol, ibuprofen, and methylphenidate, covering 36%, 14%, and 7% of all adolescent DSPs in 2016–2021 (Figure 3 A). The proportion of paracetamol involved in DSPs increased from 32% in 2019 to 40% in 2021 (Figure 3B). The ratio of the number of DSPs with single substance versus multi-substance exposure was age-dependent (Figure A2A). Younger adolescents (13 and 14 years of age) displayed a higher frequency of single substance exposures compared to older adolescents (16 and 17 years of age). Use of the various substances changed with age. For instance, the use of paracetamol in DSPs was highest among 13- and 14-year-old adolescents (42%), but decreased to 30% in 17-year-olds. Older adolescents tended to use prescription drugs more often (e.g., fluoxetine, sertraline, quetiapine, oxazepam, and lorazepam) compared to younger adolescents (Figure A2B), although paracetamol use was still dominant. In addition, the preference for the type of pharmaceutical differed between boys and girls. Paracetamol and ibuprofen were involved in 38% and 16% of all DSPs in females, respectively, compared to 25% and 8% of all DSPs in males (Figure A2C). In contrast, methylphenidate was used considerably more often in DSPs by boys (12% in boys, vs. 6% in girls).Figure 3 Substances used in DSPs by adolescents. (A) Fifteen most frequently used substances during 2016–2021 as percentage of the total number of DSPs in 2016–2021. (B) Annual contribution of the five most frequently used substances in DSPs as percentage of the total number of DSPs in a particular year.
Treatment advice
Since follow-up data are not routinely registered by the DPIC, it is not possible to report the clinical outcome and severity of the DSPs. As alternative, the treatment advice given by the DPIC, which is based on the estimated severity of the DSP, was further analyzed. During the period 2016–2021, 51% of adolescents were advised to be observed at home (estimated low risk for serious clinical effects), whereas in 49% of the cases medical examination or hospital observation was advised (estimated high risk for serious clinical effects). During the period 2016–2020, the annual proportion of patients with the advice “medical examination or hospital observation” varied between 46% and 50%. In 2021 this proportion increased to 54%, suggesting a larger share of DSPs with an estimated risk of serious clinical effects (Figure A1D). Moreover, DSPs with paracetamol were more likely to result in the advice “medical examination or hospital observation” compared to DSPs without paracetamol (Chi-square, df = 1, 82.28, p value < .001), suggesting that DSPs with paracetamol may have a higher risk of serious clinical effects.
Discussion
In this study, a strong increase in the number of adolescent DSPs was observed during the COVID-19 pandemic, starting halfway through the year 2020, particularly among young female adolescents (13–15 years old). Predominant substances used in DSPs were OTC analgesics (paracetamol, ibuprofen), of which paracetamol showed an increase in its contribution to DSPs in 2021. Moreover, the proportion of DSPs estimated to cause a more serious clinical course increased in 2021. These observations suggest a strong effect of the COVID-19 pandemic on self-harm behavior through self-poisoning, predominantly by young females.
An important but difficult to answer question is if there is a causal relation between the observed increase in adolescent DSPs, especially among females, and the COVID-19 pandemic. Similar to most parts of the world, several measures were taken in the Netherlands during the COVID-19 pandemic to mitigate the spread of the SARS-CoV-2 virus, including social distancing, general lockdowns, and curfews. Far-reaching socially isolating measures could have a profound effect on mental wellbeing of adolescents, although recent studies show conflicting results. Several studies have shown signs of deteriorating mental health among adolescents during the COVID-19 pandemic [20,[22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32]], whereas studies from the Netherlands, Sweden, and Canada showed no difference or even a decrease in the incidence of mental health problems in 2020 compared to pre-COVID-19 years in adolescents [[33], [34], [35], [36], [37]]. Since most of these studies were conducted within the first year of the COVID-19 pandemic, the study period could have been too short to observe robust changes in mental health. Moreover, country-specific differences in mitigation measures may also affect study outcomes. A large Dutch study conducted in October 2021–January 2022 among adolescents showed that the life satisfaction and emotional wellbeing had declined considerably, especially among girls, compared to the previous reference year 2017 [38]. This suggests that socially isolating measures that were prolonged up to the second year of the pandemic may have had more impact. Moreover, this study emphasizes that boys and girls may react differently on social and societal stressors.
Despite the unclear picture of the impact of the COVID-19 pandemic on adolescent mental health, a limited number of studies have shown a relation between the COVID-19 pandemic and increases in adolescent DSP rate. A recent study by Wang et al., in which prepandemic adolescent exposure calls received by US poison centers were compared with those of the first 1.5 years of the COVID-19 pandemic, showed a considerable increase in the proportion of intentional exposures with a suicide intent [19]. Moreover, there was an increase in the proportion of adolescents requiring hospital admission and in the proportion of moderate/major clinical effects and mortality. These data are in line with our data which show an increase in adolescent DSPs with a potentially more severe outcome (with the advice of further medical examination or hospital observation). Another similarity between this study and ours was the observation that the proportion of calls involving OTC analgesics increased during the pandemic. Two additional US studies, based on poison center calls during the first year of the COVID-19 pandemic, further confirmed the increase in DSPs among adolescents [39,40]. In Europe, limited data have been published regarding adolescent DSPs during the COVID-19 pandemic. In a study by Jollant et al., suicide attempts reported to French poison centers until 31st May 2022 were analyzed for different age groups, including 12–24 year olds [20]. French measures to mitigate the SARS-CoV-2 virus were comparable with Dutch measures, with a number of lengthy lockdowns and curfews, although the last curfew in France ended in June 2021, while the last lockdown in the Netherlands ended in February 2022. Especially among young females (12–24 years old) an increase in DSPs was observed in France, starting in the second half of 2020. This observation is in line with our data, although we further differentiated between different ages within the age group 13–17 years. This differentiation showed that the increase in DSPs was stronger among younger adolescents. As with the study of Jollant et al., our study showed gender-specific trends, with a strong increase in the number of female DSPs. This gender-specific effect is in line with the aforementioned study by Boer et al., which observed a decline in the mental wellbeing in 2021, mainly among girls 2021 [38]. Moreover, the general observation that girls are more often involved in DSPs than boys, has been shown previously [11,15]. The onset of the increase in DSP number (starting halfway 2020) was comparable between France and the Netherlands, which may be explained by the corresponding timing of measures taken to limit the spread of the SARS-CoV-2 virus.
A large share of OTC medication in DSPs among adolescents has been shown before [6,7,9,11,15,19]. Such drugs are easily accessible and are often available at home for general use. At therapeutic dose, paracetamol is regarded as a relatively safe analgesic with little adverse effects. However, acute and chronic overdose can result in (severe) liver damage, if not treated correctly [41]. Our data show that the proportion of adolescent DSPs involving paracetamol increased during the COVID-19 pandemic, probably due to the increase of younger adolescents involved in DSPs who predominantly use paracetamol. Moreover, our data suggest that DSPs with paracetamol may have a higher risk of serious clinical effects compared to DSPs without paracetamol. These observations raises the question whether m measures should be required to limit the access to paracetamol, but also other drugs involved in DSPs,. Preventive measures could include limiting dispensing of prescription medication to adolescents and restricting the sale of OTC medication to adolescents. In addition, indirect measures to prevent self-harm could be considered. For instance, early identification, assessment and treatment of adolescents with possible suicidal behavior may help to prevent DSPs. Practitioners, parents/caregivers, and teachers could be educated to pick up warning signs of suicidal behavior. Easy access to mental health care and losing the stigma associated with asking for mental care may be important as well [10]. Finally, societal stress factors underlying the general deterioration of mental wellbeing of adolescents may be addressed. In our study the frequency of DSPs was higher during schooldays and lower during the months of summer holidays, suggesting that school-related stress may be a risk factor for self-harm behavior. Programs to mitigate school-related stress, e.g., stress to perform or stress as result of bullying, may help to improve the mental wellbeing of adolescents.
Limitations
This retrospective study has a number of limitations. First, our data are subject to under-reporting. Health care professionals are not obliged to contact the DPIC in case of a DSP, resulting in an underestimation of the number of DSPs among adolescents in the Netherlands. Second, health care professionals may be more inclined to consult a Poison Control Center when they expect a more severe outcome of the DSP, which could lead to an overrepresentation of DSPs with the advice to medically examine or observe the patient in hospital. Third, the anamnesis regarding the exposure is usually based on self-reporting by the patient. Exposures are generally not analytically confirmed. Fourth, since patient data are anonymous, it is not possible to indicate whether individuals have been included in the study multiple times due to multiple DSPs. Fifth, in almost every case the final medical outcome of the DSP is unknown, since the DPIC does not routinely conduct follow up. Finally, in a small subset of cases (approximately 2%) the true intent of the patient was not clear. In these cases inclusion was based on exposure scenario.
Conclusion
The second year of the COVID-19 pandemic (2021) showed a strong increase in DSPs among adolescents, especially among girls and 13-, 14-, and 15- year-old children. Adolescent DSPs predominantly involved the OTC analgesics paracetamol and ibuprofen, of which the share of paracetamol showed a disproportional increase during the COVID-19 pandemic. Measures should be considered to limit access to OTC medication and prescription drugs. These measures may include restriction of the sale and dispensing of drugs to adolescents, or limiting the availability of these drugs at home. Our data suggest that social and societal stressors associated with pandemics may have an impact on self-poisoning behavior among adolescents, especially among younger girls.
Supplementary Data
Figure A1
Figure A2
Table A1
Conflicts of interest: The authors have no conflicts of interest to disclose.
Supplementary data related to this article can be found at 10.1016/j.jadohealth.2023.02.041.
==== Refs
References
1 Giuntella O Hyde K Saccardo S Lifestyle and mental health disruptions during COVID-19 Proc Natl Acad Sci U S A 2021 118
2 Cohen ZP Cosgrove KT DeVille DC The impact of COVID-19 on adolescent mental health: Preliminary findings from a longitudinal sample of healthy and at-risk adolescents Front Pediatr 9 2021 622608
3 Griffin E McMahon E McNicholas F Increasing rates of self-harm among children, adolescents and young adults: A 10-year national registry study 2007-2016 Soc Psychiatry Psychiatr Epidemiol 53 2018 663 671 29721594
4 Beckman K Mittendorfer-Rutz E Waern M Method of self-harm in adolescents and young adults and risk of subsequent suicide J Child Psychol Psychiatry 59 2018 948 956 29504652
5 Morgan C Webb RT Carr MJ Incidence, clinical management, and mortality risk following self harm among children and adolescents: Cohort study in primary care BMJ 359 2017 j4351 29046278
6 Cairns R Karanges EA Wong A Trends in self-poisoning and psychotropic drug use in people aged 5-19 years: A population-based retrospective cohort study in Australia BMJ Open 9 2019 e026001
7 Froberg BA Morton SJ Mowry JB Temporal and geospatial trends of adolescent intentional overdoses with suspected suicidal intent reported to a state poison control center Clin Toxicol (Phila) 57 2019 798 805 30696297
8 Gilley M Sivilotti MLA Juurlink DN Trends of intentional drug overdose among youth: A population-based cohort study Clin Toxicol (Phila) 58 2020 711 715 31760804
9 Overall B Hon SL Jones A Intentional suspected suicide exposures by poisoning among adolescents from 2009 to 2018 reported to the Georgia Poison Center and compared nationally Basic Clin Pharmacol Toxicol 128 2021 699 708 33523586
10 Spiller HA Ackerman JP Spiller NE Sex- and age-specific increases in suicide attempts by self-poisoning in the United States among youth and young adults from 2000 to 2018 J Pediatr 210 2019 201 208 31054768
11 Daly C Griffin E McMahon E Paracetamol-related intentional drug overdose among young people: A national registry study of characteristics, incidence and trends, 2007-2018 Soc Psychiatry Psychiatr Epidemiol 56 2021 773 781 33146859
12 Toce MS Hudgins JD Yuskaitis CJ National assessment of anti-epileptic drug exposures among pre-teens and adolescents, 2000-2020 Clin Toxicol (Phila) 2022 10.1080/15563650.2021.2023747
13 Friedrich JM Sun C Geng X Child and adolescent benzodiazepine exposure and overdose in the United States: 16 years of poison center data Clin Toxicol (Phila) 58 2020 725 731 31612741
14 Whitely M. Raven M. Jureidini J. Antidepressant Prescribing and suicide/self-harm by young Australians: Regulatory warnings, Contradictory advice, and long-term trends Front Psychiatry 11 2020 478 32587531
15 Spiller HA Ackerman JP Smith GA Suicide attempts by self-poisoning in the United States among 10-25 year olds from 2000 to 2018: Substances used, temporal changes and demographics Clin Toxicol (Phila) 58 2020 676 687 31587583
16 Tyrrell E.G. Orton E. Tata L.J. Changes in poisonings among adolescents in the UK between 1992 and 2012: A population based cohort study Inj Prev 22 2016 400 406 27185793
17 Qin P Sun S Boe AS Injuries prior and subsequent to index poisoning with medication among adolescents: A national study based on Norwegian patient registry BMC Psychiatry 18 2018 200 29914430
18 Pawer S Rajabali F Zheng A Analyses of child and youth self-poisoning hospitalizations by substance and socioeconomic status Int J Environ Res Public Health 18 2021 7003 34208955
19 Wang GS Leonard J Cornell A Adolescent US poison center exposure calls during the COVID-19 pandemic J Adolesc Health 71 2022 764 767 36088226
20 Jollant F Blanc-Brisset I Cellier M Temporal trends in calls for suicide attempts to poison control centers in France during the COVID-19 pandemic: A nationwide study Eur J Epidemiol 37 2022 901 913 36040638
21 CBS. Statistics Netherlands - Centraal Bureau voor de Statistiek (CBS). Available at: www.cbs.nl. Accessed October 4, 2022.
22 Bortoletto R Di Gennaro G Antolini G Sociodemographic and clinical changes in pediatric in-patient admissions for mental health emergencies during the COVID-19 pandemic: March 2020 to June 2021 Psychiatry Res Commun 2 2022 100023
23 Leith T Brieger K Malas N Increased prevalence and severity of psychiatric illness in hospitalized youth during COVID-19 Clin Child Psychol Psychiatry 27 2022 804 812 35236147
24 Saunders NR Kurdyak P Stukel TA Utilization of physician-based mental health care services among children and adolescents before and during the COVID-19 pandemic in Ontario, Canada JAMA Pediatr 176 2022 e216298
25 Werling AM Walitza S Eliez S The impact of the COVID-19 pandemic on mental health care of children and adolescents in Switzerland: Results of a survey among mental health care professionals after one year of COVID-19 Int J Environ Res Public Health 19 2022 3252 35328941
26 Radhakrishnan L Leeb RT Bitsko RH Pediatric Emergency Department visits associated with mental health conditions before and during the COVID-19 pandemic - United States, January 2019-January 2022 MMWR Morb Mortal Wkly Rep 71 2022 319 324 35202358
27 Edgcomb JB Benson NM Tseng CH Mental health-related emergency department visits among children during the early COVID-19 pandemic Psychiatr Res Clin Pract 4 2022 4 11 35602579
28 Kim MJ Paek SH Kwon JH Changes in suicide rate and characteristics according to age of suicide attempters before and after COVID-19 Children (Basel) 9 2022 151 35204872
29 Racine N McArthur BA Cooke JE Global prevalence of depressive and anxiety symptoms in children and adolescents during COVID-19: A meta-analysis JAMA Pediatr 175 2021 1142 1150 34369987
30 De France K Hancock GR Stack DM The mental health implications of COVID-19 for adolescents: Follow-up of a four-wave longitudinal study during the pandemic Am Psychol 77 2022 85 99 34110880
31 Knowles G Gayer-Anderson C Turner A Covid-19, social restrictions, and mental distress among young people: A UK longitudinal, population-based study J Child Psychol Psychiatry 63 2022 1392 1404 35199336
32 Theberath M Bauer D Chen W Effects of COVID-19 pandemic on mental health of children and adolescents: A systematic review of survey studies SAGE Open Med 10 2022 20503121221086712
33 van der Velden PG van Bakel HJA Das M Mental health problems among Dutch adolescents of the general population before and 9 months after the COVID-19 outbreak: A longitudinal cohort study Psychiatry Res 311 2022 114528
34 Chen Y Osika W Henriksson G Impact of COVID-19 pandemic on mental health and health behaviors in Swedish adolescents Scand J Public Health 50 2022 26 32 34100665
35 Bouter DC Zarchev M de Neve-Enthoven NGM A longitudinal study of mental health in at-risk adolescents before and during the COVID-19 pandemic Eur Child Adolesc Psychiatry 2022 10.1007/s00787-021-01935-y
36 Ray JG Austin PC Aflaki K Comparison of self-harm or overdose among adolescents and young adults before vs during the COVID-19 pandemic in Ontario JAMA Netw Open 5 2022 e2143144
37 van Loon AWG Creemers HE Vogelaar S Trajectories of adolescent perceived stress and symptoms of depression and anxiety during the COVID-19 pandemic Sci Rep 12 2022 15957 36153394
38 Boer M Dorsselaer van S Looze de M HBSC 2021: Gezondheid en welzijn van jongeren in Nederland. Netherlands: Health Behaviour in School-aged Children (HBSC) consortium 2022
39 Ciccotti HR Spiller HA Casavant MJ Pediatric poison exposures reported to United States poison centers before and during the COVID-19 pandemic Clin Toxicol (Phila) 60 2022 1299 1308 36440850
40 Lelak KA Vohra V Neuman MI COVID-19 and pediatric ingestions Pediatrics 148 2021 e2021051001
41 Dear J. Bateman D. Paracetamol poisoning Medicine 48 2020 208 210
|
PMC010xxxxxx/PMC10155139.txt |
==== Front
CJEM
CJEM
Cjem
1481-8035
1481-8043
Springer International Publishing Cham
37133634
511
10.1007/s43678-023-00511-4
Original Research
An evaluation of satisfaction with emergency department care in children and adolescents with mental health concerns
http://orcid.org/0000-0001-7504-1841
Lategan Conné 1
Newton Amanda S. 2
http://orcid.org/0000-0002-8367-7994
Thull-Freedman Jennifer 3
http://orcid.org/0000-0002-6680-0620
Stang Antonia 1
http://orcid.org/0000-0003-0850-4337
Lang Eddy 4
Arnold Paul 5
Stubbs Michael 6
http://orcid.org/0000-0003-2319-6192
Freedman Stephen B. stephen.freedman@albertahealthservices.ca
3
1 grid.22072.35 0000 0004 1936 7697 Department of Pediatrics, Cumming School of Medicine, University of Calgary, Calgary, AB Canada
2 grid.17089.37 0000 0001 2190 316X Department of Pediatrics, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB Canada
3 grid.22072.35 0000 0004 1936 7697 Departments of Pediatrics and Emergency Medicine, Cumming School of Medicine, University of Calgary, Alberta Children’s Hospital, Calgary, AB Canada
4 grid.22072.35 0000 0004 1936 7697 Department of Emergency Medicine, Cumming School of Medicine, University of Calgary, Calgary, AB Canada
5 grid.22072.35 0000 0004 1936 7697 The Mathison Centre for Mental Health Research and Education and Department of Psychiatry, Cumming School of Medicine, University of Calgary, Calgary, AB Canada
6 grid.22072.35 0000 0004 1936 7697 Department of Psychiatry, Cumming School of Medicine, University of Calgary, Calgary, AB Canada
3 5 2023
2023
25 6 498507
28 11 2022
11 4 2023
© The Author(s), under exclusive licence to Canadian Association of Emergency Physicians (CAEP)/ Association Canadienne de Médecine d'Urgence (ACMU) 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Objectives
We hypothesized that an association exists between satisfaction with ED mental health care delivery and patient and system characteristics. Primary: To evaluate overall satisfaction with ED mental health care delivery. Secondary: To explore aspects of ED mental health care delivery associated with general satisfaction, and patient and ED visit characteristic associated with total satisfaction scores and reported care experience themes.
Methods
We enrolled patients < 18 years of age presenting with a mental health concern between February 1, 2020 and January 31, 2021, to two pediatric EDs in Alberta, Canada. Satisfaction data were collected using the Service Satisfaction Scale, a measure of global satisfaction with mental health services. Association of general satisfaction with ED mental health care was evaluated using Pearson’s correlation coefficient and variables associated with total satisfaction score was assessed using multivariable regression analyses. Inductive thematic analysis of qualitative feedback identified satisfaction and patient experience themes.
Results
646 participants were enrolled. 71.2% were Caucasian and 56.3% female. Median age was 13 years (IQR 11–15). Parents/caregivers (n = 606) and adolescents (n = 40) were most satisfied with confidentiality and respect in the ED and least satisfied with how ED services helped reduce symptoms and/or problems. General satisfaction was associated with perceived amount of help received in the ED (r = 0.85) and total satisfaction with evaluation by a mental health team member (p = 0.004) and psychiatrist consultation (p = 0.05). Comments demonstrated satisfaction with ED provider attitudes and interpersonal skills and dissatisfaction with access to mental health and addictions care, wait time, and the impact of COVID-19.
Conclusions
There is a need to improve ED mental health care delivery, with a focus on timely access to ED mental health providers. Access to outpatient/community-based mental health care is needed to complement care received in the ED and to provide continuity of care for youth with mental health concerns
Supplementary Information
The online version contains supplementary material available at 10.1007/s43678-023-00511-4.
Résumé
Objectifs
Nous avons émis l'hypothèse qu'il existe un lien entre la satisfaction à l'égard de la prestation de soins de santé mentale aux urgences et les caractéristiques des patients et du système. Primaire : Évaluer la satisfaction globale à l’égard de la prestation des soins de santé mentale aux urgences. Secondaire : Explorer les aspects de la prestation des soins de santé mentale aux urgences associés à la satisfaction générale, et les caractéristiques du patient et de la visite aux urgences associées aux scores de satisfaction totale et aux thèmes d’expérience de soins signalés.
Méthodes
Nous avons inscrit des patients de moins de 18 ans présentant un problème de santé mentale entre le 1er février 2020 et le 31 janvier 2021 à deux services d'urgence pédiatriques en Alberta, au Canada. Les données relatives à la satisfaction ont été recueillies à l'aide de l'échelle de satisfaction du service, une mesure de la satisfaction globale à l'égard des services de santé mentale. L'association entre la satisfaction générale et les soins de santé mentale dispensés aux urgences a été évaluée à l'aide du coefficient de corrélation de Pearson et les variables associées au score total de satisfaction ont été évaluées à l'aide d'analyses de régression multivariables. L'analyse thématique inductive des commentaires qualitatifs a permis d'identifier des thèmes liés à la satisfaction et à l'expérience des patients.
Résultats
646 participants ont été inscrits. 71,2 % étaient de race blanche et 56,3 % de sexe féminin. L'âge médian était de 13 ans (IQR, 11-15). Les parents/aidants (n = 606) et les adolescents (n = 40) étaient les plus satisfaits de la confidentialité et du respect à l’urgence et les moins satisfaits de la façon dont les services d’urgence ont contribué à réduire les symptômes et/ou les problèmes. La satisfaction générale était associée à la perception de l'aide reçue aux urgences (r = 0,85) et à la satisfaction totale à l'égard de l'évaluation par un membre de l'équipe de santé mentale (p = 0,004) et de la consultation d'un psychiatre (p = 0,05). Les commentaires ont fait état d'une satisfaction à l'égard des attitudes et des compétences interpersonnelles des prestataires de soins d'urgence et d'une insatisfaction à l'égard de l'accès aux soins de santé mentale et de toxicomanie, du temps d'attente et de l'impact de l'étude COVID-19.
Conclusions
Il est nécessaire d’améliorer la prestation des soins de santé mentale aux urgences, en mettant l’accent sur l’accès en temps opportun aux fournisseurs de services de santé mentale des services d’urgence. L’accès à des soins de santé mentale en consultation externe ou en milieu communautaire est nécessaire pour compléter les soins reçus aux urgences et pour assurer la continuité des soins aux jeunes ayant des problèmes de santé mentale.
Keywords
Patient satisfaction
Mental health
Pediatric
Emergency department
Mots-clés
Satisfaction des patients
Santé mentale
Pédiatrie
Service des urgences
http://dx.doi.org/10.13039/501100000145 Alberta Innovates - Health Solutions Partnership for Research Innovation in the Health System Newton Amanda S. http://dx.doi.org/10.13039/100012856 Alberta Children's Hospital Research Institute Alberta Children's Hospital FoundationProfessorship in Child Health Wellness Freedman Stephen B. issue-copyright-statement© Canadian Association of Emergency Physicians (CAEP)/ Association Canadienne de Médecine d'Urgence (ACMU) 2023
==== Body
pmcClinician’s capsule
What is known about the topic?
While certain aspects of ED care are associated with satisfaction, no evaluations specific to children/adolescents with mental health concerns have been performed.
What did this study ask?
How satisfied are families/children with ED mental health care delivery, and which aspects are associated with satisfaction scores.
What did this study find?
We need to improve ED mental health care delivery, focusing on enhancing access to care by mental health providers.
Why does this study matter to clinicians?
To improve patient and family satisfaction, ED and mental health administrators need to adapt models of ED mental health care delivery that include streamlined access to mental health care practitioners and facilitate access to community-based mental health supports for ongoing care.
Background
Over the past decade, visits to emergency departments (EDs) by children and adolescents for mental health care have increased [1, 2], a trend exacerbated by the COVID-19 pandemic [3]. Although the process of assessing suicidal ideation is well-studied, little attention has been paid to understanding and improving patient-reported experiences.
Satisfaction is an important measure of patient experience and is a good indicator of service quality, future service usage and continuity of care adherence [4–6]. Knowledge of satisfaction can provide insight regarding expectations for care [7] and how services can be improved. Although certain aspects of pediatric ED care are known to be associated with satisfaction (e.g., interpersonal interactions, communication, provider skills, wait times) [8], no evaluations specific to ED mental health care delivery have been performed. Thus, we sought to assess child/adolescent and parent/caregiver satisfaction with ED mental health care, and to determine which aspects of ED care receive the highest and lowest satisfaction scores. Based on what is known about satisfaction with ED care, we hypothesized that positive associations would be identified between general satisfaction and perceived wait time [8–11] and ED provider courtesy and compassion [8, 10, 12].
Methods
Study design and setting
This cross-sectional study was embedded in a prospective implementation study conducted in two tertiary care pediatric EDs (Stollery Children’s Hospital and the Alberta Children’s Hospital) in Alberta, Canada [13]. Study outcome measures and processes were determined in collaboration with patient and parent partners. Data were collected between February 1, 2020, and January 31, 2021, and reflect care (Fig. 1) prior to implementation of a new clinical care pathway. Using a complete case ascertainment sampling strategy, all patients meeting the eligibility criteria were approached either in-person or via telephone after the ED visit. Research Ethics Board approval was obtained, and the caregivers of eligible participants provided consent; assent was obtained when appropriate. Adolescents ≥ 14 years of age who presented without a legal guardian participated as mature minors. Results are reported in accordance with the STROBE guidelines [14].Fig. 1 Standard of care for mental health patients in the participating emergency departments. ED, Emergency Department; RN, Registered Nurse. *Mental health team member can include a mental health nurse, counsellor, or psychiatrist. δ Discharge resources can include nothing (e.g., follow-up with family physician or existing mental health provider), provision of pamphlets with options to family to coordinate follow-up, provision of mental health care coordination phone number, referral to a specific program, and/or follow-up by a hospital-based mental health, outreach home care team
Population
Eligible participants were < 18 years old and presented with any of the following Canadian Emergency Department Information System [15] presenting complaints documented at triage: anxiety, bizarre behaviour, concern for patient’s welfare, depression/suicidal, disruptive behaviour, homicidal behaviour, insomnia, self-harm, situational crisis, or violent behaviour. Those with acute medical and/or physical safety concerns were ineligible including children brought by protective/police services or ambulance, chief complaints relating to schizophrenia/psychosis, behavioural syndromes requiring medical clearance, or significant self-harm. Children who had previously participated were also ineligible.
Outcome measures
The primary outcome was total satisfaction with ED mental health care delivery quantified by the Service Satisfaction Scale (SSS-10) [16, 17]; Online Resource 1. The scale is comprised of 10 items (child/adolescent version) or 12 items (parent/caregiver version) that utilize a 5-point Likert-scale plus three open-ended questions that elicit opinions on what worked well during care, and what should be changed [16]. SSS-10 items are divided into two subscales: manner and skills of the staff (5 items) and perceived outcomes (5 items for child/adolescent version; 7 items for parent/caregiver version). Total scale scores range from 10 to 50 (child/adolescent version) or 12 to 60 (parent/caregiver version), with higher scores indicating greater satisfaction [16]. Secondary outcomes were SSS-10 item associations with the SSS-10 general satisfaction question, patient and ED visit characteristic associations with the total satisfaction score (i.e., sum of SSS-10 elements), and experiences with ED mental health care.
Data collection
Data were collected as soon as possible following the ED visit, as parent partners advised that to minimize stress, research recruitment would ideally be performed following the ED visit. Care experiences, satisfaction, and demographic data were collected via a questionnaire completed by telephone or online. Telephone-based questionnaires followed a standardized process led by a trained research assistant. Online data collection occurred within the study’s REDCap database. ED visit characteristic data were collected via medical record review. Data collected included International Classification of Diseases, Version 10, Canada (ICD-10-CA) discharge diagnoses code assigned based on chief complaint and physician notes: F codes for mental and behavioural disorders and R and X codes for intentional self-harm that did not require medical care. Length of stay (LOS) was defined as the time the child/adolescent was in the ED from triage to discharge; triage time was defined as the time of ED triage.
Data analysis
The total SSS-10 score, representing an individual’s satisfaction with care, was calculated by summing the individual item scores. Individual SSS-10 item scores were used to identify aspects of ED mental health care respondents were most and least satisfied with. We evaluated associations of general satisfaction with ED mental health care using Pearson’s correlation coefficient, associations between total satisfaction score with patient and ED visit characteristics using multivariable regression analyses, and compared satisfaction scores between the two participating EDs using student’s t test (Online Resource 2).
An inductive thematic analysis was conducted using the open-ended SSS-10 question responses to identify positive and negative care experience themes. Thematic coding was performed by one author to identify and categorize excerpts to find emerging themes and patterns. To ensure consistency and accuracy, two independent reviewers reviewed the codes and assigned themes to 50 randomly selected participants. Themes are reported using frequency (frequency of each theme divided by the total number of responses) and intensity (proportion of codes describing a particular theme divided by the total number of codes) effect sizes [18]. Ratios of positive to negative feedback were calculated to permit interpretation of the relationship between theme frequency and intensity and to determine which themes were most strongly associated with dissatisfaction and satisfaction.
Analyses were conducted using R software (Version 1.14.4, Vienna, Austria). Statistical tests were two-tailed and P values of < 0.05 were considered statistically significant.
Results
Study participants
Of 970 potentially eligible children and adolescents, 73.0% consented and 66.6% of those that consented provided data for analysis; Fig. 2. Sixty-five percent of participants received care at the Alberta Children’s Hospital. The median time to data collection was 14 days (IQR 8–22). Participants were predominantly female (56.3%) and Caucasian (71.2%), with a median age of 13 years (IQR 11–15 years); Table 1. Anxiety and stress-related disorders (39.5%), suicidal ideation (26.0%), and mood disorders (25.0%) were the most common discharge diagnoses.Fig. 2 Overall study screening and enrollment
Table 1 Participant and ED visit characteristics
Characteristic Overall
Age in years, median (IQR) 13 (11–15)
Gender, n (%)
Female 364 (56.3)
Male 246 (38.2)
Genderqueer/nonconforming 13 (2.0)
TransMale 12 (1.9)
TransFemale 2 (0.3)
Othera 6 (0.9)
Declined to answer 3 (0.5)
Ethnicity, n (%)
Caucasian 460 (71.2)
Canadian first nations, Inuit, or metis 50 (7.7)
Mixedb 46 (7.1)
Asian 42 (6.5)
Black or African American 12 (1.9)
Otherc 36 (5.6)
Discharge diagnosis, n (%)d, e
Neurotic, stress-related, and somatoform disorders 245 (39.5)
Suicidal ideation 161 (26.0)
Mood disorders 155 (25.0)
Behavioral and emotional disorders/syndromes 130 (21.0)
Intentional self-harm not requiring medical care 23 (3.7)
Disorders of adult personality and behaviour 14 (2.3)
Disorders of psychological development 14 (2.3)
Other or unspecified 40 (6.5)
None documented 13 (12.1)
Length of stay in minutes, median (IQR) 264 (170–404)
Evaluation by mental health team member, n (%) 389 (62.7)
Psychiatry consult, n (%) 150 (24.2)
Hospital admission, n (%) ± 97 (15.6)
aParticipants who self-identified as having a gender outside the categories listed in the table
bParticipants who selected more than one ethnicity category
cParticipants who self-identified as having an ethnicity outside the categories listed in the table
dPhysician could identify more than one discharge diagnosis which could result in more than one category for some participants
eMissing data for 26 participants
Satisfaction with ED mental health care
Parents/caregivers and adolescents were most satisfied with confidentiality and respect for their child’s rights; Table 2, Online Resource 3. Parents/caregivers were least satisfied with how ED services helped reduce their child’s symptoms/problems (mean 3.0, SD 1.2) and how ED services helped their child get well and stay well (mean 3.1, SD 1.2). Adolescents were least satisfied with how ED services helped reduce their symptoms/problems (mean 3.0, SD 1.0). Aspects of care satisfaction differed between sites; Table 2. Greater satisfaction with care was reported for parents/caregivers whose child received care at Alberta Children’s Hospital (mean 42.4, SD 9.8, p = 0.003).Table 2 Parent/caregiver satisfaction with ED mental health care (means with standard deviations [SD])
SSS-10 item Alberta children’s hospital (n = 383) Stollery children’s hospital (n = 223) Difference between hospitals P valuea
Manner and skills of the ED providers
The knowledge and skills of the ED providers working with your child/adolescent? 3.8 (0.9) 3.5 (1.0) < 0.001*
The ability of ED providers to listen to and understand your child/adolescent’s problems? 3.8 (1.0) 3.6 (1.1) 0.01
How involved and caring the ED providers are with your child/adolescent? 3.9 (1.0) 3.7 (1.0) 0.06
Confidentiality and respect for your child/adolescent’s rights as an individual? 4.3 (0.8) 4.2 (0.9) 0.20
The way ED providers address your child/adolescent’s most important concerns/needs? 3.6 (1.1) 3.4 (1.2) 0.01
Perceived outcomes
How the services help your child/adolescent? 3.4 (1.0) 3.1 (1.2) 0.003*
The way services help your child/adolescent get well and stay well? 3.2 (1.1) 2.9 (1.2) 0.002*
The amount of help your child/adolescent receives? 3.3 (1.2) 3.0 (1.3) 0.002*
The way the services help reduce your child/adolescent’s symptoms and/or problems? 3.1 (1.1) 2.9 (1.2) 0.04
In a general sense, how satisfied are you with the services your child/adolescent has received? 3.5 (1.1) 3.1 (1.2) < 0.001*
The availability of appointment times that fit your child/adolescent’s schedule? 3.3 (1.1) 3.2 (1.3) 0.08
How long your child/adolescent had to wait in the waiting room before appointments? 3.3 (1.1) 3.3 (1.4) 0.70
Total score 42.4 (9.8) 39.7 (11.0) 0.003*
aBonferroni-corrected alpha: 0.05/13 = 0.004
Correlations with satisfaction
The amount of help a child/adolescent received had the strongest, positive association with general satisfaction (r = 0.85, 95% CI 0.83–0.87), with the weakest association for how long the child/adolescent had to wait (r = 0.31, 95% CI 0.23–0.38); Online Resource 4. Receipt of an evaluation by a mental health team member (OR = 15.39, 95% CI 5.40–43.90, p = 0.004) and psychiatry consultation (OR = 8.97, 95% CI 3.15–25.58, p = 0.05) were positively associated with the total satisfaction score; Online Resource 5. Self-identification as Asian (OR = 90.83, 95% CI 31.85–259.00, p = 0.01) and ‘other’ ethnicities (OR = 51.26, 95% CI 17.98–146.18, p = 0.04) was positively associated with total satisfaction score compared to Caucasian participants; self-identification as being of mixed ethnicity (OR = 0.04, 95% CI 0.01–0.10, p = 0.04) was negatively associated with total satisfaction score.
Themes for ED mental health care
Qualitative feedback was provided by 57.0% of participants, with respondents representing older children than non-respondents; Online Resource 6. Twelve themes were created, and the greatest amount of positive feedback (frequency and intensity) pertained to ED provider attitude and interpersonal skills; Table 3. Expectations regarding standards of care received the most negative feedback followed by wait times and access to mental health and addictions specialists. The latter theme also had the greatest negative to positive ratio in terms of frequency (66:1) and intensity (81:1), followed by the wait time and COVID-19 themes; Table 3. Online Resource 7 provides information on themes and their definitions.Table 3 Satisfaction theme frequency and intensity effect sizes, n (%)
Theme Frequency effect size (frequency of each theme/total # of responses) Intensity effect size (# of codes for particular theme/total # of codes)
N = 371 N = 371
Positive experiencesa
ED provider attitude and interpersonal skills 37.2% 41.0%
Expectations regarding standards of care 29.1% 31.5%
Communication 9.7% 9.7%
Timeliness of care 9.4% 9.7%
Follow-up care 8.9% 8.9%
Mental health specialization/psych consult 7.3% 7.3%
Overall satisfaction 5.4% 5.7%
ED environment 4.0% 4.3%
Patient- and family-centered care 3.8% 4.0%
Wait time 1.9% 1.9%
Access to mental health and addictions care 0.3% 0.3%
Negative experiencesa
Expectations regarding standards of care 34.5% 43.7%
Access to mental health and addictions care 17.8% 21.8%
Wait time 16.4% 16.7%
ED provider attitude and interpersonal skills 15.4% 18.3%
Mental health specialization/psych consult 13.5% 17.3%
Follow-up care 11.6% 12.4%
Timeliness of care 10.2% 11.3%
Communication 9.4% 10.2%
ED environment 6.7% 7.8%
Overall satisfaction 6.5% 7.3%
Patient- and family-centered care 4.6% 4.6%
COVID-19 1.1% 1.1%
aQualitative analysis included all participants who completed the baseline survey (n = 371/651; 57.0%) (see Fig. 2)
Discussion
Interpretation
In this study, which explored satisfaction with care delivered to children presenting for ED care with acute mental health concerns, study participants were most satisfied with confidentiality and respect in the ED and least satisfied with how ED services helped reduce the symptoms and/or problems that led them to seek ED care. Satisfaction was associated with perceived amount of help received in the ED, evaluation by a mental health team member, and psychiatrist consultation. Comments provided by participants demonstrated satisfaction with ED provider attitudes and interpersonal skills and dissatisfaction with access to mental health and addictions care, wait time, and the impact of COVID-19. Our findings indicate that there is a need to improve ED mental health care delivery, with a focus on enabling timely access to ED mental health providers. Access to outpatient/community-based mental health care is also needed to complement care received in the ED and to provide continuity of care for youth with mental health concerns.
Comparison to previous studies
Previous pediatric studies examining satisfaction with general ED care, utilizing satisfaction measures other than the SSS-10, report that positive ED provider attitude, interpersonal skills, and high-quality provider interactions are important to patients [8, 10, 19], and are positively associated with increased parent/caregiver satisfaction with ED care [10, 12]. However, in our study, how ED services help reduce symptoms and how they help the child/adolescent get and stay well, received the lowest satisfaction scores. This may reflect the inability of EDs to provide high-quality mental health care due to inadequate protocols, lack of standardized tools, limited ED provider training, and a shortage of mental health specialists [20–22].
Our findings that access to mental health and addictions care had the greatest ratio of negative to positive comments aligns with the adult ED literature. While a Canadian ED study of adults with mental health concerns reported similar reasons for dissatisfaction [23], an Australian adult study described receipt of a mental health team member and psychiatry consultation as being positively associated with satisfaction [24]. Since access to consultation by a mental health team member or psychiatrist is not a standard component of pediatric ED mental health care in Canada due to the lack of resources in most institutions, with the ED providers limiting access to those in greatest need (Fig. 1), patient expectations are often not met. Moreover, as comprehensive mental health care is not universally accessible under Canada's healthcare plan, and mental health services for children are under-resourced and under-funded [25], the limited availability of outpatient and community-based mental health care [26, 27] continues to place greater pressure on the ED to deliver mental health care.
Strengths and limitations
A novel finding in our study was the positive association between being Asian or of another non-White ethnicity with satisfaction, while mixed ethnicity was negatively associated. Prior non-mental health specific studies suggest that ethnicity does play a role in patient satisfaction and perceptions of ED care [28, 29]. In Canada, First Nations patients often report negative ED experiences, and they may be exposed to discrimination and racism in the ED [30]. Mental health specific research from the United States has highlighted important racial and ethnic disparities in pediatric ED presentations for mental health concerns [31]. To the best of our knowledge, only one Canadian ED mental health study has focused on race and ethnicity; it identified that First Nations adolescents are more likely to present to the ED for mental health concerns. However, this study did not examine whether race or ethnicity was associated with ED services provided or satisfaction [32].
As our consent rate was only 73.0%, the results may be subject to non-response and volunteer biases, and thus possibly underrepresenting the perspective of those with negative ED interactions. Limiting survey completion to English speaking participants resulted in the omission of perspectives from non-English speaking participants. We were not permitted to document reasons for, or characteristics of patients electing not to participate and thus cannot compare participants to non-participants or provide reasons for non-participation. Further, we lacked representation of children in foster care or group homes where consent from legal guardians could not be obtained. Although we attempted to capture the perspectives of adolescents, as only 40 mature minors participated, we cannot confirm if adolescent perspectives differ from those of parents/caregivers. Our eligibility criteria prevented the inclusion of children with certain mental health presentations which limits the generalizability of our findings to populations such as those with psychosis or self-harm requiring medical care [13]. Due to the multiple steps involved in obtaining consent to contact, consent and ultimately survey completion, many surveys were completed outside of our target survey completion window. In addition, the SSS-10 is not specifically designed for ED use and the items do not account for the dynamic processes and multiple care providers that are core components of ED care.
Clinical implications
This study demonstrates that while parents/caregivers and children and adolescents with mental health concerns are satisfied with ED providers and that satisfaction with ED care is associated with receipt of a mental health team member or psychiatrist consultation, they are less satisfied with how mental health services helped address their child's concerns and/or symptoms. Our findings should be used to inform ED mental health care delivery models which need to focus on enhancing the provision of timely access to pediatric mental health specialists. Funding and resources are needed to improve connections to outpatient and community-based mental health supports to enable the early identification, management, and prevention of mental health concerns.
Research implications
Future research initiatives targeting the implementation of novel models of care, the monitoring and setting quality benchmarks, and an evaluation of the impact of implementing standardized mental health tools in the ED, are needed. Prospective studies should compare satisfaction between parent–child dyads. Moreover, an evaluation of approaches to connect all youth, without an existing mental health care provider relationship, to a post-ED visit mental health care visit would likely have an impact on satisfaction.
Conclusion
While parents/caregivers and children and adolescents were satisfied with ED providers, satisfaction with ED care was associated with receiving a mental health team member or psychiatrist consultation. In addition, they were less satisfied with how mental health services in helped reduce their child’s mental health concerns and/or symptoms. This may reflect the challenge of providing adequate mental health care due to limited resources in the ED and the community settings. This knowledge should inform ED mental health care delivery models with a focus on providing improved and timely access to ED pediatric mental health specialists and connections to outpatient resources to ultimately improve outcomes for children with mental health concerns.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 19 KB)
Supplementary file2 (DOCX 13 KB)
Supplementary file3 (DOCX 14 KB)
Supplementary file4 (DOCX 14 KB)
Supplementary file5 (DOCX 14 KB)
Supplementary file6 (DOCX 16 KB)
Supplementary file7 (DOCX 17 KB)
Acknowledgements
The authors would like to thank the following individuals for their contribution to this study: Jennifer Tweed, Carol Coventry, Kristi Frost, Abdul Rahman, Erin LaLande, Ashley McFetridge, Erin Pols, Lee McDonald, Jacinda Larson, and Kate Winston. In addition, we would like to thank the Government of Alberta and Alberta Health Services for their support of this project.
Author contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by CL, SBF, and ASN. The first draft of the manuscript was written by CL and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding
This work was supported by Alberta Innovates and Alberta Health Services through joint contribution to the PRIHS (Partnership for Research and Innovation in the Health System) Program. Additional funding was provided by the Alberta Children's Hospital Research Institute (ACHRI). The funders’ financial support does not constitute an endorsement of the Project, the Recipient, or any of the information contained in this manuscript. SF is also supported by the Alberta Children’s Hospital Foundation Professorship in Child Health and Wellness.
Availability of data and material
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. All supplementary information files are included in this published article.
Code availability
Not applicable.
Declarations
Conflict of interest
Conné Lategan received financial support from ACHRI for a master’s program. Conné Lategan, Michael Stubbs, Antonia Stang, Jennifer Thull-Freedman, Eddy Lang, Paul Arnold, Amanda S. Newton, and Stephen Freedman declare that they have no conflict of interest.
Ethics approval
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. The study was approved by the University of Calgary (REB19-0357) and the University of Alberta (Pro00092862).
Consent to participate
Informed consent/assent was obtained and documented from all children, adolescents, and parents/guardians participating in the study.
Consent for publication
Not applicable.
==== Refs
References
1. Santillanes G Axeen S Lam CN Menchine M National trends in mental health-related emergency department visits by children and adults, 2009–2015 Am J Emerg Med 2020 38 12 2536 2544 10.1016/j.ajem.2019.12.035 31902702
2. Chiu M Gatov E Fung K Kurdyak P Guttmann A Deconstructing the rise in mental health-related ED visits among children and youth in Ontario Canada Health Aff (Millwood) 2020 39 10 1728 1736 10.1377/hlthaff.2020.00232 33017254
3. Madigan S Korczak DJ Vaillancourt T Racine N Hopkins WG Pador P Comparison of paediatric emergency department visits for attempted suicide, self-harm, and suicidal ideation before and during the COVID-19 pandemic: a systematic review and meta-analysis Lancet Psychiatry 2023 10 3 1 10 36526342
4. Dubina MI O'Neill JL Feldman SR Effect of patient satisfaction on outcomes of care Expert Rev Pharmacoecon Outcomes Res 2009 9 5 393 395 10.1586/erp.09.45 19817520
5. Al-Abri R Al-Balushi A Patient satisfaction survey as a tool towards quality improvement Oman Med J 2014 29 1 3 7 10.5001/omj.2014.02 24501659
6. Batbaatar E Dorjdagva J Luvsannyam A Savino MM Amenta P Determinants of patient satisfaction: a systematic review Perspect Public Health 2017 137 2 89 101 10.1177/1757913916634136 27004489
7. Agency for Healthcare Research and Quality. What is Patient Experience? Rockville, MD: Agency for Healthcare Research and Quality. https://www.ahrq.gov/cahps/about-cahps/patient-experience/index.html; Accessed 7 Apr 2023.
8. Magaret ND Clark TA Warden CR Magnusson AR Hedges JR Patient satisfaction in the emergency department–a survey of pediatric patients and their parents Acad Emerg Med 2002 9 12 1379 1388 12460841
9. Toma G Triner W McNutt LA Patient satisfaction as a function of emergency department previsit expectations Ann Emerg Med 2009 54 3 360 7.e6 10.1016/j.annemergmed.2009.01.024 19282060
10. Byczkowski TL Gillespie GL Kennebeck SS Fitzgerald MR Downing KA Alessandrini EA Family-centered pediatric emergency care: a framework for measuring what parents want and value Acad Pediatr 2016 16 4 327 335 10.1016/j.acap.2015.08.011 26525991
11. Boudreaux ED O'Hea EL Patient satisfaction in the emergency department: a review of the literature and implications for practice J Emerg Med 2004 26 1 13 26 10.1016/j.jemermed.2003.04.003 14751474
12. Grissom MO Farra M Cruzen ES Barlow E Gupta S What can COVID-19 teach us about patient satisfaction in the emergency department? A mixed-methods approach J Am Coll Emerg Physicians Open 2021 2 2 e12436 33969346
13. Freedman S Thull-Freedman J Lightbody T Prisnie K Wright B Coulombe A Introducing an innovative model of acute paediatric mental health and addictions care to paediatric emergency departments: a protocol for a multicentre prospective cohort study BMJ Open Qual. 2020 9 4 e001106 10.1136/bmjoq-2020-001106 33318032
14. von Elm E Altman DG Egger M Pocock SJ Gøtzsche PC Vandenbroucke JP The strengthening the reporting of observational studies in epidemiology (strobe) statement: guidelines for reporting observational studies PLoS Med 2007 4 10 e296 10.1371/journal.pmed.0040296 17941714
15. Canadian Association of Emergency Physicians. The CEDIS Presenting Complaint List (V2.0) Ottawa (CAN): Canadian Association of Emergency Physicians. https://caep.ca/wp-content/uploads/2016/03/nacrs_presenting_complaint_list_v2_0_en_fr_pdf_pdf.pdf. (2008). Accessed 7 Apr 2023.
16. Simpson L, Adair CE. Satisfaction Measurement in Mental Health Services Using the SSS-10 Family of Questionnaires. Calgary, Alberta, Canada; April 18, 2005.
17. Athay MM Bickman L Development and psychometric evaluation of the youth and caregiver Service Satisfaction Scale Adm Policy Ment Health 2012 39 1–2 71 77 10.1007/s10488-012-0407-y 22407558
18. Onwuegbuzie AJ Effect sizes in qualitative research: a prolegomenon Qual Quant 2003 37 4 393 409 10.1023/A:1027379223537
19. Pagnamenta R Benger JR Factors influencing parent satisfaction in a children's emergency department: prospective questionnaire-based study Emerg Med J 2008 25 7 417 419 10.1136/emj.2007.050005 18573954
20. Cappelli M Cloutier P Newton AS Fitzpatrick E Ali S Dong KA Evaluating mental health service use during and after emergency department visits in a multisite cohort of Canadian children and youth CJEM 2019 21 1 75 86 10.1017/cem.2017.416 29199945
21. Cappelli M Glennie JE Cloutier P Kennedy A Vloet M Newton A Physician management of pediatric mental health patients in the emergency department: assessment, charting, and disposition Pediatr Emerg Care 2012 28 9 835 841 10.1097/PEC.0b013e31826764fd 22929135
22. McCabe EM Lightbody TK Mummery C Coloumbe A GermAnn K Lent B Implementing a clinical pathway for paediatric mental health care in the emergency department Can J Commun Ment Health 2020 38 4 1 18 10.7870/cjcmh-2019-015
23. Fleury MJ Grenier G Farand L Satisfaction with emergency departments and other mental health services among patients with mental disorders Healthc Policy 2019 14 3 43 54 31017865
24. Summers M Happell B Patient satisfaction with psychiatric services provided by a Melbourne tertiary hospital emergency department J Psychiatr Ment Health Nurs 2003 10 3 351 357 10.1046/j.1365-2850.2003.00600.x 12755921
25. Asmundson GJG Blackstock C Bourque MC Brimacombe G Crawford A Deacon SH McMullen K McGrath PJ Mushquash C Stewart SH Stinson J Taylor S Campbell-Yeo M Blais JM Easing the disruption of COVID-19: supprting the mental health of the people of Canada - October 2020 - an RSC policy briefing FACETS (Ottawa) 2020 5 1 1071 1098 10.1139/facets-2020-0082
26. Gonzalez K Patel F Cutchins LA Kodish I Uspal NG Advocacy to address emergent pediatric mental health care Clin Pediatric Emerg Med 2020 21 2 100778 10.1016/j.cpem.2020.100778
27. Carubia B Becker A Levine BH Child psychiatric emergencies: updates on trends, clinical care, and practice challenges Curr Psychiatry Rep 2016 18 4 41 10.1007/s11920-016-0670-9 26932516
28. Lee MO Altamirano J Garcia LC Gisondi MA Wang NE Lippert S Patient age, race and emergency department treatment area associated with "topbox" press ganey scores West J Emerg Med 2020 21 6 117 124 10.5811/westjem.2020.8.47277 33207156
29. Heyming TW Donaldson CD Ehwerhemuepha L Feaster W Fortier MA Kain ZN Multivariable analysis of patient satisfaction in the pediatric emergency department Pediatr Emerg Care 2022 38 2 e544 e549 10.1097/PEC.0000000000002514 34348353
30. McLane P Bill L Barnabe C First Nations members' emergency department experiences in Alberta: a qualitative study CJEM 2021 23 1 63 74 10.1007/s43678-020-00009-3 33683609
31. Abrams AH Badolato GM Boyle MD McCarter R Goyal MK Racial and ethnic disparities in pediatric mental health-related emergency department visits Pediatr Emerg Care 2022 38 1 e214 e218 10.1097/PEC.0000000000002221 32898125
32. Newton AS Rosychuk RJ Dong K Curran J Slomp M McGrath PJ Emergency health care use and follow-up among sociodemographic groups of children who visit emergency departments for mental health crises CMAJ 2012 184 12 E665 E674 10.1503/cmaj.111697 22690003
33. Lantos JD Yeh HW Raza F Connelly M Goggin K Sullivant SA Suicide risk in adolescents during the COVID-19 pandemic Pediatrics 2022 149 2 e2021053486 10.1542/peds.2021-053486 34977942
34. Fernandez A Gindt M Babe P Askenazy F Mental health-related visits in a pediatric emergency department during the COVID-19 pandemic Int J Emerg Med 2021 14 1 64 10.1186/s12245-021-00387-0 34736389
35. Chadi N Spinoso-Di Piano C Osmanlliu E Gravel J Drouin O Mental health-related emergency department visits in adolescents before and during the COVID-19 pandemic: a multicentric retrospective study J Adolesc Health 2021 69 5 847 850 10.1016/j.jadohealth.2021.07.036 34462192
|
PMC010xxxxxx/PMC10155169.txt |
==== Front
Food Environ Virol
Food Environ Virol
Food and Environmental Virology
1867-0334
1867-0342
Springer US New York
37133676
9555
10.1007/s12560-023-09555-2
Research
Wastewater-Based Epidemiology of SARS-CoV-2: Assessing Prevalence and Correlation with Clinical Cases
http://orcid.org/0000-0001-6595-9878
Wani Hima hima.wani@brcmicrobiology.in
1
Menon Smita smita482@gmail.com
12
Desai Dipen dipen.desai@brcmicrobiology.in
1
D’Souza Nishita dsouzan1@msu.edu
3
Bhathena Zarine zarine_bhathena@rediffmail.com
2
Desai Nishith nishith.desai@brcmicrobiology.in
1
Rose Joan B. rosejo@msu.edu
3
http://orcid.org/0000-0002-0671-9214
Shrivastava Sandhya sandhya_s10@brcmicrobiology.in
1
1 Bhavan’s Research Center, Bhavan’s College Campus, Andheri West, Mumbai, Maharashtra 400058 India
2 grid.475378.a 0000 0004 0385 8467 Department of Microbiology, Bhavan’s College, Andheri West, Mumbai, Maharashtra 400058 India
3 grid.17088.36 0000 0001 2150 1785 Department of Fisheries and Wildlife, Michigan State University, East Lansing, MI 48824 USA
3 5 2023
2023
15 2 131143
23 12 2022
18 4 2023
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Wastewater-based epidemiology has been recognized as a tool to monitor the progress of COVID-19 pandemic worldwide. The study presented herein aimed at quantitating the SARS-CoV-2 RNA in the wastewaters, predicting the number of infected individuals in the catchment areas, and correlating it with the clinically reported COVID-19 cases. Wastewater samples (n = 162) from different treatment stages were collected from three wastewater treatment plants (WWTPs) from Mumbai city during the 2nd surge of COVID-19 (April 2021 to June 2021). SARS-CoV-2 causing COVID-19, was detected in 76.2% and 4.8% of raw and secondary treated (n = 63 each) wastewater samples respectively while all tertiary treated samples (n = 36) were negative. The quantity of SARS-CoV-2 RNA determined as gene copies/100 mL varied among all the three WWTPs under study. The gene copy numbers thus obtained were further used to estimate the number of infected individuals within the population served by these WWTPs using two published methods. A positive correlation (p < 0.05) was observed between the estimated number of infected individuals and clinically confirmed COVID-19 cases reported during the sampling period in two WWTPs. Predicted infected individuals calculated in this study were 100 times higher than the reported COVID-19 cases in all the WWTPs assessed. The study findings demonstrated that the present wastewater treatment technologies at the three WWTPs studied were adequate to remove the virus. However, SARS-CoV-2 genome surveillance with emphasis on monitoring its variants should be implemented as a routine practice to prepare for any future surge in infections.
Keywords
COVID-19
SARS-CoV-2
RT-qPCR
Wastewater treatment plants
Wastewater-Based Epidemiology
Surveillance
http://dx.doi.org/10.13039/501100001429 Indo-US Science and Technology Forum IUSSTF/VN-COVID/081/2020 IUSSTF/VN-COVID/081/2020 IUSSTF/VN-COVID/081/2020 IUSSTF/VN-COVID/081/2020 issue-copyright-statement© Springer Science+Business Media, LLC, part of Springer Nature 2023
==== Body
pmcIntroduction
Coronavirus disease-2019 (COVID-19) is caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) belonging to the species SARS-related coronavirus in the subgenus Sarbecovirus of the family Coronaviridae containing enveloped viruses with positive sense single stranded RNA (Kitajima et al., 2020). COVID-19 was first detected in Wuhan, China in December 2019. This disease was later reported in 114 countries and therefore the World Health Organization declared COVID-19 as a pandemic on 11th March 2020 (Medema et al., 2020). As of 29th March 2023, the disease has affected 761,402,282 individuals and 6,887,000 deaths were reported worldwide (WHO-COVID-19 Dashboard accessed at https://covid19.who.int/ on 5th April 2023). In India, 4,47,29,284 confirmed cases and 5,30,901 deaths have been reported till 5th April 2023. (Government of India COVID-19 Dashboard accessed at https://www.mygov.in/covid-19 on 5th April 2023). In India, the first case of COVID -19 was detected in Thrissur, state Kerala on 27th January 2020 while in Mumbai, state Maharashtra, the first case was identified on 11th March, 2020 (https://www.hindustantimes.com/cities/mumbai-news/mumbais-first-covid-patients-we-have-let-it-stay-under-the-carpet-101614797073559.html accessed on 24.11.2022). Individuals infected with COVID-19 infection (symptomatic or asymptomatic) shed SARS-CoV-2 RNA in feces (at a concentration up to 107 copies/g), which is further disposed of in wastewater and may increase the likelihood of fecal–oral transmission (Ahmed et al., 2020; Hemalatha et al., 2021; Kitajima et al., 2020; Pandey et al., 2021).
Surveillance of virus transmission within the population using environmental samples has been established for well-known poliovirus as well as newly emerging Aichi virus (Arora et al., 2020; Hemalatha et al., 2021; Sharma et al., 2021). This type of monitoring is often referred to as Wastewater-Based Epidemiology (WBE) and has also been recognized as a promising surveillance tools for several enteric viruses such as Norovirus, Hepatitis A and E virus, Adenovirus, Rotavirus, and Influenza A (H1N1) and is currently implemented around the globe for SARS-CoV-2 (Arora et al., 2020). WBE has been proven to be effective in monitoring the viral load in the wastewater catchment area, which may provide unambiguous predictions of future outbreaks. It may also aid in uncovering the ground reality of COVID-19 cases (including asymptomatic cases) by examining a larger population as opposed only clinically reported cases in the area (Hata et al., 2020a, 2020b; Srivastava et al., 2021a, 2021b). Various WBE studies for monitoring of SARS-CoV-2 RNA in wastewater have been conducted worldwide in countries such as Australia (Ahmed et al., 2020), Japan (Hata et al., 2020a, 2020b), Netherlands (Medema et al., 2020), Pakistan (Sharif et al., 2020), France (Wurtzer et al., 2020), Spain (Randazzo et al., 2020) and USA (Peccia et al., 2020; Weidhaas et al., 2021).
Several studies have been reported from Indian cities viz. Gandhinagar (Kumar et al., 2021a), Ahmedabad (Joshi et al., 2021; Kumar et al., 2021b) Vadodara (Srivastava et al., 2021b), Bengaluru (Lamba et al., 2023) focusing on estimating the concentration of SARS-CoV-2 in influent wastewater samples received by the wastewater treatment plants in the respective city. Researchers have also used the viral copy number of SARS-CoV-2 data obtained from the wastewater to predict COVID-19 infected individuals (or active shredders) within the community served by the WWTP (Chakraborty et al., 2020; Hemalatha et al., 2021; Lamba et al., 2023). Brian et al., 2022 has summarized India, along with Australia, Japan and the United States in the list of growing number of countries to undertake WBE studies. As SARS-CoV-2 has been widely detected in the influent wastewater samples, the data on efficacy of treatment system to remove the SARS-Cov-2 is very limited. In India, such evaluation has been undertaken only in Jaipur (Arora et al., 2020), Chennai (Chakraborty et al., 2020) and Hyderabad (Hemalatha et al., 2021).
Mumbai is one of the largest and densely populated metropolitan cities in the western coast of India with 6.5 million living in the urban slums (Government of India 2011). The city has recorded 2nd highest number of COVID-19 positive cases (11,57,747) with 19,747 deaths as of 5th April 2023 (Municipal Corporation of Greater Mumbai- accessed at https://stopcoronavirus.mcgm.gov.in/assets/docs/Dashboard.pdf on 5th April 2023). During the initial pandemic phase, one study carried out by Sharma et al. (2021) has reported the presence of SARS-CoV-2 RNA in Mumbai city’s wastewater. In this study, wastewater samples from 5 large open drain and sewage pumping stations between February, March and May 2020 were assessed for the presence of SARS-CoV-2. The city generates 2190 MLD wastewater which is distributed among eight wastewater treatment plants (WWTPs) located across the city. However, each WWTP processes the wastewater up to different stages also using different technologies. viz. WWTP located in Colaba and Charkop are the only two WWTPs carrying out tertiary treatment (chlorination and ozonation, respectively). Bhandup and Versova WWTPs treat the wastewater up to the secondary stage while Bandra, Malad, Ghatkopar and Worli WWTPs treat it up to the primary stage. Of these, 3 WWTPs (Colaba, Charkop and Bhandup) covering south, north and central zone of the city and representing 5,87,930 populations (Table 1), were chosen in the present study. Table 1 Technical specification of WWTPs for surveillance of SARS-CoV-2 in Mumbai city, India
WWTP code WWTP/location/MCGM ward Installed capacity* Served population Average dry weather flow capacitya Secondary treatment technologyb Tertiary treatment
WWTP-Z1 Colaba (Zone 1)
City Center [A ward]
37 MLD 1,10,916 26.65 MLD Sequential Batch Reactors (SBR) Chlorination
WWTP-Z5 Charkop (Zone 5)
Northern Suburbs
[R/South ward]
6 MLD 27,814 4.57 MLD Rotating Media Bioreactors (RMBR) Ozonation
WWTP -Z3 Bhandup (Zone 3)
Central Suburbs
[S ward]
280 MLD 4,49,200 81.23 MLD Aerated lagoons Facility not available
aMCGM Report, 2020–2021 (Accessed at https://portal.mcgm.gov.in/irj/go/km/docs/documents/MCGM%20Department%20List/Environment/Docs/MCGM%20ESR%20English%202020%20-%202021.pdf on 25th February 2022)
bNational Inventory of Sewage Treatment Plant, March 2021, CPCB Accessed at https://cpcb.nic.in/openpdffile.php?id=UmVwb3J0RmlsZXMvMTIyOF8xNjE1MTk2MzIyX21lZGlhcGhvdG85NTY0LnBkZg== on 25th Feb 2022
The present study was undertaken during the 2nd surge of the disease, by when the virus detection systems were also standardized. The objective of the present study was to detect and quantify SARS-CoV-2 RNA in the wastewaters of Mumbai city at various stages of the wastewater treatment process to evaluate treatment systems potential to inactivate/eliminate the SARS-CoV-2 virus from the final effluent. Social distancing/preventive measures in compact places of Mumbai city particularly in slums was a big challenge for the municipal authority, therefore the study also aimed to establish the correlation between the SARS-CoV-2 copy number data obtained from the wastewater with the clinically reported cases from the respective WWTP wards.
Materials and Methods
Sampling Sites and Sample Collection
Raw (influent), secondary and tertiary treated wastewater (effluent) samples were collected by grab sampling technique from three WWTPs. Bhandup WWTP (WWTP-Z3), located in zone-3 under the ‘S’ administrative ward of Mumbai Municipal Corporation of Greater Mumbai (MCGM). Colaba WWTP (WWTP-Z1), situated in zone1 (city center) is under 'A' ward of MCGM and Charkop WWTP (WWTP-Z5) in zone 5 of Mumbai city under ‘R/South’ ward of MCGM (Fig. 1). Technical details of the wastewater treatment at the three WWTPs are compiled in Table 1. A total of 162 wastewater samples (1 L) were collected in a sterile polypropylene (PP) containers which comprised paired raw/influent and secondary treated wastewater samples collected from WWTP-Z3 (n = 27), WWTP-Z5 (n = 22) and WWTP-Z1 (n = 14) WWTP, respectively. Tertiary treated wastewater were also collected from WWTP-Z5 (n = 22) and WWTP-Z1 (n = 14) WWTP, respectively. All samples were collected during April 2021 to June 2021 on weekdays (Monday to Thursday/Friday) between 9:00 AM-10:00 AM. This sampling period covered the 2nd wave of COVID-19 pandemic in Mumbai city. However, due to the lockdown and restricted travel, there was a gap of 2 weeks during the sampling. Fig. 1 Location of the three WWTPs of Mumbai city under study
Sample Processing and SARS-CoV-2 Concentration
Samples were processed within 3 h post collection on the same day. Standardized PEG precipitation method (Michael-kordatou et al., 2020) was used to concentrate SARS-CoV-2 in wastewater samples by taking 100 mL of wastewater sample, centrifuged at 4000×g for 20 min (Eppendorf Centrifuge 5804R) to remove the coarse debris. The supernatant was then transferred into sterile screw cap glass bottle containing 8% (w/vol) polyethylene glycol (PEG) (Sigma Aldrich, USA) and 0.2 M (w/v) sodium chloride (Sigma Aldrich, USA). Bacteriophage Phi6 (DSM 21518), spiked at 5.5. × 105 pfu/100 mL of sewage was used as an internal process control to assess recovery of enveloped RNA viruses from wastewater samples (Kitajima et al., 2020). The samples were transferred to an orbital shaker incubator (Trishul Equipments India) adjusted at 4 °C and 130 RPM for 2 h and then subsequently held at 2–8 °C for virus precipitation (Flood et al., 2021). The virus precipitate was pelleted out by centrifugation at 10,000×g for 30 min at 4 °C and the pellet was resuspended in 2 mL PBS (Themofisher, USA). The virus concentrates were stored at − 80 °C. RNA extraction and the subsequent qPCR assay was performed within 72 h.
Qualitative Detection of SARS-CoV-2
For qualitative detection of SARS-CoV-2, 450 µL of virus concentrate was mixed with 50 µL of uninfected human plasma (NHP) and was used for RNA extraction. RNA was extracted using the Roche HiPure system viral nucleic acid kit (Roche, USA) as per the manufacturer’s instructions. The RNA was eluted in 75 µL of elution buffer and used for the detection of SARS-CoV-2 using TRUPCR SARS-CoV-2 RT- qPCR kit (V-3.2) (3B BlackBio Biotech India Ltd., India) in StepOne™ Instrument (Applied Biosystems, USA). The procedure included addition of 10 µL of extracted RNA to 15 µL of total reaction mix which contained master mix (10 µL), enzyme mix (0.35 µL) and primer–probe mix (4.65 µL). The RT qPCR program set up was 1 cycle each at 50 °C/15 min and 95 °C/05 min followed by 38 cycles of 95 °C/05 s, 60 °C/40 s (included dye acquisition) and 72 °C/15 s. Each run included negative and positive control provided with the kit.
Quantification of SARS-CoV-2 in Wastewater Samples Using qRT-PCR
Wastewater samples tested positive qualitatively were subjected to quantitative reverse transcriptase PCR assay to determine the RNA copies of SARS-CoV-2. RNA extracted from 300 µL of virus concentrate using the HiPure system viral nucleic acid kit (Roche, USA) was eluted in 50 µL of nuclease free water. Specific gene targets viz. N1 (Nucleocapsid), ORF1b-nsp14 (located in Open Reading Frame 1b) and RdRp (RNA dependent RNA polymerase) were selected for quantification of SARS-CoV-2 RNA. Additionally, Phi6 used as an internal process control was also quantified. The sequences of primers-probes used are given in Table 2. 5 µL of RNA template was used in a total reaction volume of 25 µL containing SuperScript™ III RT/Platinum Taq Mix (0.5 µL), 2X Reaction mix (12.5 µL), 10 μm forward and reverse primer (0.5 µL), 10 μM fluorogenic probe 0.25 µL. qRT-PCR cycling program was set as 1 cycle of each 50 °C/15 min (cDNA synthesis) and 95 °C/2 min followed by 40 cycles each of 95 °C/15 s, 60 °C/30 s. Quantitative PCR was carried out using SuperScript™ III Platinum™ One-Step qRT-PCR Kit (Invitrogen, USA) in Applied Biosystems StepOnePlus™ Instrument. Gene copies/reaction obtained from the assay was used to back calculate gene copies/100 mL of wastewater samples. SARS-CoV-2 quantitative RNA reference standard (ATCC 3276SD) was used as a standard copy control.Table 2 List of primer–probe sequence chosen for SARS-CoV-2 and Phi6 qRT-PCR
Target Primer/probe set Primer/probe sequence (5′ to 3′) References
SARS CoV-2 2019-nCoV_N1-F
2019-nCoV_N1-R
2019-nCoV_N1-P
GACCCCAAAATCAGCGAAAT
TCTGGTTACTGCCAGTTGAATCTG
FAM-ACCCCGCATTACGTTTGGTGGACC-BHQ2
CDC (2020)
HKU-ORF1b-nsp14F
HKU-ORF1b-nsp14 R
HKU-ORF1b-nsp14 P
TGGGGYTTTACRGGTAACCT’
AACRCGCTTAACAAAGCACTC
LC-RED610—TAGTTGTGATGCWATCATGACTAG- BHQ2
ICMR- NIV and Kitajima et al. (2020)
RdRP_SARSr-F2
RdRP_SARSr-R1
RdRP_SARSr-P2
Specific for Wuhan-CoV
GTGARATGGTCATGTGTGGCGG
CARATGTTAAASACACTATTAGCATA
LC-RED610-CAGGTGGAACCTCATCAGGAGATGC- BHQ2
ICMR- NIV, Corman et al. (2020) and La Rosa et al. (2020)
Phi6 Φ6Tfor
Φ6Trev
Φ6Tprobe
TGGCGGCGGTCAAGAGC
GGATGATTCTCCAGAAGCTGCTG
CY5-CGGTCGTCGCAGGTCTGACACTCGC-BHQ2
Gendron et al. (2010)
COVID-19 Reported Cases in WWTP Catchment Zone and Number of Infectious Individuals in the Population Served by 3 WWTP
Data on the number of daily reported COVID-19 cases during the study (April to June 2021) was provided by the public health officers of the respective MCGM administrative wards. Information related to the area served by the WWTP was obtained from the respective ward offices. Number of infected individuals among the population served by each WWTP was estimated using 2 different methods that have been previously reported (Ahmed et al., 2020; Hemalatha et al., 2021).
The equations used for calculation are mentioned below:Method 1 (Eq. 1): (Ahmed et al., 2020)1 Persons\; Infected=RNA\; copiesliter\; wastewater×liters\; wastewaterdayg\; fecesperson-day×RNA\; copiesg\; feces
For g feces, 128 g was used for feces excreted/person/day and it has been estimated that, positive individual shed 107 RNA copies/g of feces (Ahmed et al., 2020).
Method 2 (Eq. 2): (Hemalatha et al., 2021)2 Persons\; Infected=No.\;of \;RNA \;copies\; per\; liter\; of\; wastewaterContribution\; of\; RNA\; copies\; per\; person\; to\; total\; sewage\; water
107 RNA copies/g of feces was multiplied with 120 mL of volume of faeces excreted by human (considering density of feces as 1.07 g/mL) and total wastewaters (L) received at WWTP (Hemalatha et al., 2021).
Data Analysis
The data obtained from qualitative and quantitative estimation of SARS-CoV-2 RNA as well as Phi6 was transferred and analyzed in Microsoft Excel spreadsheet. IBM SPSS version 23.0 was used for data visualization and statistical analysis. Based on the Phi6 and the SARS-CoV-2 standard copy control, sample limit of detection and assay limit of detection was determined. The copy controls were used to calculate the gene copies of Phi6 and SARS-CoV-2 per reaction in each of the samples and was subsequently back calculated to gene copies/ 100 mL of sewage. Correlational analysis was performed to determine correlation of average SARS-CoV-2 GC/100 mL quantities with physico-chemical parameters, reported COVID-19 cases, and infectious individuals in the population calculated using three methods and p value < 0.05 was considered as a statistically significant correlation. Qualitatively positive SARS-CoV-2 samples that were undetected for a particular gene were assigned the sample limit of detection for correlational analysis.
Results
Qualitative Detection of SARS-CoV-2 in Wastewater Samples
To assess the prevalence of SARS-CoV-2 in the wastewater samples, TRUPCR SARS-CoV-2 RT- qPCR kits (V-3.2) were used. The dual gene targets (RdRp + N) used for qualitative detection of SARS-CoV-2 was detected in 76.2% (48 out of 63) of raw wastewater samples (mean Ct. value 31.5 ± 3.2). Of the 63 secondary treated wastewater samples, only three (4.8%) from WWTP-Z3 (Mean Ct. value 34.3 ± 0.5) were positive for SARS-CoV-2. The remaining 60 secondary treated samples as well as all the tertiary treated samples (n = 36) collected in this study were negative for SARS-CoV-2.
Quantification of Phi6 and SARS-CoV-2 in Wastewater Samples
Recovery of Internal Process Control Phi6
Phi6, an enveloped RNA bacteriophage, has been used as a surrogate for viral respiratory pathogens such as SARS-CoV-2 in wastewater analysis because of its resemblance to those viral respiratory pathogens and absence in urban wastewater (Fedorenko et al., 2020; Kitajima et al., 2020). Thus, an internal process control Phi6 was added to all the samples including the negative control to assess the recovery of enveloped viruses using RT-qPCR. Phi6 was detected in all the spiked samples with the mean cycle threshold (Ct) value of 29.7 ± 3.2. (Mean ± S.D.) and the mean percentage recovery of 33.9 ± 29.9% (Mean ± S.D.).
Quantification of SARS-CoV-2 in Wastewater
The wastewater samples positive for SARS-CoV-2 RNA by qualitative method were subjected to quantitative PCR for 3 gene targets viz. N1, ORF1b-nsp14 and RdRp. The performance efficiency of RT-qPCR assay for SARS-CoV-2 and Phi6 gene targets have been summarized in Table 3. The Ct values and the calculated mean gene copies/100 mL of wastewater samples are provided in Table 4. The mean N1, ORF1b-nsp14 and RdRp genes copies in WWTP-Z3 and WWTP-Z5 were in the range of 105 GC/100 mL (except for N1 and RdRp gene quantified in WWTP-Z3 were 104 GC/100 mL) and were higher by 52.6% and 85.9% of gene copies, respectively as compared to WWTP-Z1 (104 GC/100 mL). However, amongst the three WWTPs, SARS-CoV-2 gene copies obtained were highest in WWTP-Z5. These observed variations in the quantities of three genes in the raw wastewaters across the three WWTPs is represented using the box-whisker plot (Fig. 2).Table 3 Performance efficiency of qRT-PCR assay for SARS-CoV-2 and Phi6 gene targets
Gene Efficiency (%) Slope Y-Intercept R2 value Assay limit of detection (copies/5 μL of RNA template)
N1 gene 87.05 − 3.677 43.719 0.988 300
ORF1b-nsp14gene 71.34 − 4.276 45.673 0.991 300
RdRp gene 75.52 − 4.093 46.084 0.983 300
Phi6 73.248 − 4.19 43.053 0.999 80
Table 4 Cycle Thresholds (Ct) values of SARS-CoV-2 gene targets at each treatment stage in 3 WWTPs
WWTP Wastewater sample type N1 gene ORF1b- nsp14 gene RdRp gene
Positives n, (%) Ct (mean ± SD) Mean gene copies/100 mL (95% CI) Positives (n, %) Ct (mean ± SD) Mean gene copies/100 mL (95% CI) Positives (n, %) Ct (mean ± SD) Mean gene copies/100 mL (95% CI)
WWTP-Z3 Raw 23 (85.2) 33.79 ± 1.62 8.68 × 104 (5.31 × 104–1.20 × 105) 23 (85.2) 33.36 ± 1.53 1.13 × 105 (7.61 × 104–1.50 × 105) 18 (66.7) 34.45 ± 1.73 9.04 × 104 (4.75 × 104–1.33 × 105)
Secondary treated 2 (7.4) 34.79 ± 1.25 3.08 × 103 (− 1.87 × 103–8.03 × 103) 3 (11.1) 35.21 ± 1.05 4.59 × 103 (− 1.15 × 103–1.03 × 104) 3 (11.1) 36.01 ± 0.53 4.41 × 103 (− 8.07 × 102–9.62 × 103)
WWTP-Z5 Raw 16 (72.7) 31.53 ± 1.94 3.04 × 105 (1.67 × 105–4.41 × 105) 15 (68.2) 30.67 ± 1.29 3.52 × 105 (2.02 × 105–5.02 × 105) 15 (68.2) 32.03 ± 1.67 3.18 × 105 (1.83 × 105 − 4.53 × 105)
Secondary treated 0 – – 0 – – 0 – –
Tertiary treated 0 – – 0 – – 0 – –
WWTP-Z1 Raw 8 (57.1) 33.79 ± 1.5 6.03 × 104 (− 5.18 × 103–1.26 × 105) 9 64.3) 34.3 ± 1.5 4.76 × 104 (5.98 × 103–8.93 × 104 8 (57.1) 35.81 ± 1.04 2.89 × 104 (8.6 × 103–4.91 × 104)
Secondary treated 0 – – 0 – – 0 – –
Tertiary treated 0 – – 0 – – 0 – –
Fig. 2 Gene copies of SARS-CoV-2 gene targets in 3 WWTP using Box and Whiskers plot. The data represents the average number of SARS-CoV-2 gene copies for N1 gene, ORF-1ab and RdRp gene per 100 mL of sewage sample obtained in the untreated wastewater samples from the Three WWTPs
Reported Cases of COVID-19 During Wastewater Surveillance Period and Correlational Analysis with SARS-CoV-2
The study period was in the middle of the 2nd surge of COVID-19 cases in Mumbai city, the clinically confirmed COVID-19 case load was retrieved from the MCGM wards. It was observed that the number of daily reported cases from WWTP-Z3 (S ward) and WWTP-Z5 (R/South ward) decreased from 430 and 519 cases/day in April 2021 to 43 and 48 cases/day in the month of June 2021, respectively. Similar trend was also observed in WWTP-Z1 (A ward) wherein during the study period (May and June 2021), COVID-19 cases decreased from 30 cases/day in May 2021 to a mere five cases/day by June 2021. These observations were in-line with SARS-CoV-2 gene copy numbers observed in the wastewater samples of the respective wards (Fig. 3). A positive trend was observed between the weekly average SARS-CoV-2 gene copy number in the wastewater and the cumulative weekly reported COVID-19 cases in all the three WWTPs. A statistically significant strong positive correlation was obtained with WWTP-Z3 (0.863, p = 0.027) and WWTP-Z5 (0.859, p = 0.028) WWTP (Table 5). WWTP-Z1 showed moderate positive correlation (− 0.31, p = 0.69) however, this association was not statistically significant.Fig. 3 Week-wise trend of mean SARS-CoV-2 gene copies estimated in raw wastewater versus reported COVID-19 cases in the WWTPs served area. A WWTP-Z1 (Colaba), B WWTP-Z3 (Bhandup) and C WWTP-Z5 (Charkop). Bars represent weekly mean SARS-CoV-2 gene copies/100 mL of sewage and Line represents the weekly mean COVID-19 cases for the ward area served by the WWTP
Table 5 Estimation of infectious persons in the population served by WWTP and Correlation with reported COVID-19 cases
WWTP Week Consecutive samples Average dry weather flow capacity Population served by WWTP Average reported COVID-19 Average SARS-CoV-2 GC/100 mL No. of infectious person
Method 1 (Ahmed et al., 2020) Method 2 (Hemalatha et al., 2021)
WWTP-Z3* Week 1
06/04/21–08/04/21
3 81 449,200 361 2.17E + 05 137,602 146,775
Week 2
12/04/21–15/04/21
4 332 1.70E + 05 107,736 114,919
Week 3
03/05/21–07/05/21
5 112 1.61E + 05 102,149 108,959
Week 4
10/05/21–14/05/21
5 97 3.72E + 04 23,545 25,115
Week 5
24/05/21–28/05/21
5 62 4.48E + 04 28,341 30,231
Week 6
31/05/21–04/06/21
5 38 2.53E + 04 16,002 17,069
WWTP-Z5* Week 1
06/04/21–08/04/21
3 5 27,814 533 6.26E + 05 24,444 26,074
Week 2
12/04/21–13/04/21
2 386 6.76E + 05 26,406 28,167
Week 3
03/05/21–07/05/21
5 193 5.67E + 05 22,143 23,619
Week 4
10/05/21–14/05/21
5 118 2.56E + 05 7990 8523
Week 5
24/05/21–28/05/21
5 64 1.60E + 04 626 668
Week 6
07/06/21–08/06/21
2 55 0 0 0
WWTP-Z1 Week 1
20/05/21–21/05/21
2 36 110,916 29 0 0 0
Week 2
24/05/21–28/05/21
5 13 7.77E + 04 23,296 22,496
Week 3
31/05/21–4/06/21
5 7 5.80E + 04 17,410 15,810
Week 4
07/06/21–08/06/21
2 5 0 0 0
*Significant association between average COVID-19 reported clinical cases versus SARS-CoV-2 GC/100 mL and no. of infectious persons (p < 0.05)
Number of Infected Individuals and Correlation with COVID-19 Reported Cases
Two methods were used to estimate the number of infected individuals among the population served by WWTP based on the SARS-CoV-2 gene copy number obtained from the wastewater samples (Ahmed et al., 2020; Hemalatha et al., 2021). It was observed that the number of infected individuals estimated by the method 1 and method 2 were similar and in the range of 104–105 (Fig. 4). Correlational analysis of reported cases with the estimated number of infected individuals are shown in Table 5. The predicted number of infected individuals using the two methods followed a decreasing trend similar to the reported COVID-19 cases and a statistically significant correlation was obtained with data from WWTP-Z3 (r = 0.863; p = 0.027) and WWTP-Z5 (r = 0.859; p = 0.028) but not with WWTP-Z1 (r = − 0.31; p = 0.69).Fig. 4 Reported and Predicted COVID 19 infected cases across 3 WWTPs. The figure represents mean of week-wise COVID-19 reported cases and predicted infected individuals using Ahmed et al. (2020) and Hemalatha et al. (2021) methods using Wastewater-based Epidemiology approach. The correlation between reported and predicted COVID-19 cases was significant (p < 0.05) among WWTP-Z3 and Z5 and vice-versa
Discussion
The present study conducted in the city of Mumbai, India aimed to provide an in-depth analysis of the prevalence of SARS-CoV-2 RNA viral copies in the wastewater and its persistence through the treatment process during the 2nd surge of the COVID-19 pandemic. The study aimed to evaluate the efficiency of the treatment systems (followed at three WWTPs) to eliminate the SARS-CoV-2 virus from the wastewater. The obtained SARS-CoV-2 RNA copy data was used to establish correlation between the SARS-CoV-2 RNA copies in the untreated wastewater and the reported clinical infections from the areas served by the WWTP.
During sample processing, bacteriophage phi-6 was incorporated in all the samples as a process control to assess virus recovery and the samples were processed using the PEG/NaCl to obtain the virus concentrates (Michael-kordatou et al., 2020). The bacteriophage phi-6 was selected as a process control due to its similarity to enveloped RNA viruses and its substantiated use as a surrogate for various respiratory viruses such as Ebola virus, influenza virus, coronaviruses (SARS-1) and recently, SARS-CoV-2 (Fedorenko et al., 2020; Kitajima et al., 2020; Whitworth et al., 2020). The Phi6 recoveries observed in the study (22–36%) were similar to values reported by Flood et al. (2021) using similar method. Many researchers have higher recoveries of phi6 (56% and 54%) using ultrafiltration and adsorption-elution methods for virus concentration respectively (Sherchan et al., 2020). The recovery of the phi6 using PEG/NaCl method in the present study is similar to that reported by Flood et al. (2021) (22–36%). The satisfactory virus recovery along with the cost-effectiveness of PEG/NaCl method substantiates its use particularly in the resource limited settings.
The presence of SARs-CoV-2 in the wastewater samples was qualitatively detected using TRUPCR SARS-CoV-2 RT- qPCR kit (V-3.2) (3B BlackBio Biotech India Ltd, India) which is an Indian Council of Medical Research (ICMR) approved kit for qualitative real-time PCR kit. The kit is very sensitive with a limit of detection of 6 copies/ µL of reaction. This high sensitivity is attributed to the use of dual probes with single dye against the RdRp gene and N gene for the detection of SARS-CoV-2, thus preventing any ambiguity resulting from mutations in any one of the target region. It was observed that the SARS-CoV-2 RNA was detected in majority of the untreated wastewater samples (76.2%, n = 48/63) indicating prevalence of COVID-19 disease in the population served by the three WWTPs under study. The SARS-CoV-2 RNA was undetected in most of the secondary and all of the tertiary treated in WWTPs Z1 and WWTP Z5.
Z1 and Z5 WWTP use the Rotating Media Bio Reactor (RMBR) and Sequential Batch Reactor (SBR) technology for secondary treatment, respectively. These technologies aid in efficient separation of the generated sludge from the wastewater which is also known to drastically reduce bacterial and viral load from the wastewater (Bhave et al., 2020), thus no viral RNA was detected post-secondary treatment herein. Arora et al. (2020) also reported absence of SARS-CoV-2 in the treated wastewater samples processed through similar methods. However, it would be worthwhile to check the presence of SARS-CoV-2 RNA or the live virus in the sludge generated from the WWTPs that subsequently ends up in landfills or in gardens. The lagoon-based secondary treatment followed at WWTP-Z3 was able to reduce the SARS-CoV-2 sample positivity by 95.2%; with 3 samples out of 27 from this WWTP testing positive for SARS-CoV-2 RNA. This indicates the limitation of this method of secondary treatment in the virus elimination. Similar findings have been reported from a Spain-based study using similar treatment technology (Randazzo et al., 2020), wherein 11% of the secondary treated samples (n = 18, from 2 different WWTPs) were positive for SARS-CoV-2. Inability of this treatment procedure, may be due to the large area of the lagoon resulting in inefficient uniform aeration. However, this assumption needs to be validated.
Post qualitative assessment, SARS-COV-2 viral copies in the wastewater were estimated using three targets genes of SARS-CoV-2 namely N1, ORF1b-nsp14 and RdRp gene Selection of these genes was based on recommendations by national and international bodies. N1 gene was recommended by Center for Disease Control (CDC), Atlanta, USA and has also been commonly used in several wastewater-based studies worldwide (Haramoto et al., 2020; Medema et al., 2020; Randazzo et al., 2020; Sherchan et al., 2020). The use of ORF1b-nsp14 and RdRp was as per genes targets reported for clinical and environmental studies by several researchers (Corman et al., 2020; Kitajima et al., 2020; Rosa et al., 2020) however, the RT-qPCR efficiencies have not been discussed and as per the recommendation by ICMR, Delhi, India),. for this surveillance (https://www.icmr.gov.in/pdf/covid/labs/2_SOP_for_Confirmatory_Assay_for_2019_nCoV.pdf accessed on 19th June, 2022).
We observed that the PCR efficiency for Orf-1b-nsp-14 and RdRp was in the range of 71–75%) while that of N1 gene was 87%. Considering that the same samples were used for all the three gene targets the used of ORF1-b-nsp-14 and RdRp should be studied further for its suitability in wastewater-based surveillance. Quantitative data on SARS-CoV-2 wastewater surveillance studies have been reported from other parts of India during different phases of the pandemics with variable gene copy numbers. Chakraborty et al., 2020 reported 104–105 GC/L of SARS-CoV-2 (N1 and N2 genes) in 4 WWTPs from Chennai city during September 2020. Srivastava et al. (2021a, 2021b) reported lower gene copies (102–103 GC/L) of SARS-CoV-2 (N, ORF and S gene) in untreated wastewaters of Gandhinagar, Ahmedabad and Vadodara cities in India during November 2020. Nag et al. (2022), studied 11 WWTP’s across Jaipur city during Feb to Jun 2021 and obtained SARS-CoV-2 gene (E, RdRp and N) concentrations in the range of 1 × 104 to 1 × 106 gene copies/L of wastewater.
From Mumbai city wastewaters, SARS-CoV-2 virus was studied in the early phase of the pandemic (Feb 2020 to May 2020). The study was qualitative, using E and RdRp gene sampled largely from 5 open drains and one sewage pumping station (Sharma et al., 2021) This study, thus was designed to understand the quantitative data of a large metropolitan city, also the virus fate in waste water treatment and finally correlating the quantitative data with clinical cases, contributing to WBE. The study conducted here estimated the SARS-CoV-2 viral copy numbers in the wastewater was carried out between April–June 2021. It was observed that the viral copy numbers were higher (104–105 GC/100 mL) in untreated wastewater samples and was tenfold higher in WWTP-Z5 (105 GC/100 mL) as compared to WWTP-Z1and WWTP-Z3 (104 GC/100 mL). This variation between the SARS-CoV-2 copy number observed in the WWTPs could largely be attributed to number of prevalent COVID-19 cases among the population, the population density and the amount of wastewater generated by the community served by the WWTP. Even though WWTP-Z5 had the lowest volume of wastewater treated per day amongst the three WWTPs, this WWTP mainly serves only the residential area which may contribute to higher SARS-CoV-2 viral load as compared to WWTP-Z3 and WWTP-Z5 where the chance of wastewater getting diluted due to industrial activities cannot be excluded. Similar observations were also reported by Kumar et al. (2021a) where N, ORF and S genes were used for the study. The three secondary treated samples from WWTP-Z3 that were qualitatively positive for SARS-CoV-2 were also positive by quantitative tests for at least two genes. However, there was a 1.38 Log10 reduction in the SARS-CoV-2 RNA copies/100 mL when compared to untreated /raw wastewater.
In the present study, a statistically significant strong positive correlation was observed between the weekly mean SARS-CoV-2 RNA copies from wastewater and weekly mean confirmed COVID-19 cases reported from the catchment areas of two WWTP’s (Z3 and Z5). Similar correlations were also observed by Kumar et al. (2021a, 2021b) and Srivastava et al. (2021a, 2021b) in studies conducted across various cities (Gandhinagar, Vadodara and Ahmedabad) across Gujarat. However, such correlation was not observed in WWTP-Z1 wherein the sampling was initiated during the late stage of the 2nd surge of COVID-19 infection where the reported case load from this municipal ward was also very low. Thus, this inconsistency may have been one of the contributing factor. Additionally, using the weekly mean SARS-CoV-2 gene copy number data from the wastewater, the number of active shedders was estimated by two published methods (Ahmed et al., 2020; Hemalatha et al., 2021). Although a positive trend (tenfold decrease) was observed in the estimated SARS-CoV-2 gene copies and reported COVID-19 cases across all 3 WWTPs (Table 5), the predicted caseloads were 100 times higher (median value) than clinically confirmed COVID-19 cases (Table 6). This difference may be a result of lag in the active reporting of cases, the prevalence of asymptomatic infections within the community as well as the enrichment of SARS-CoV-2 copies due to grab sampling method during the peak hours of the wastewater flow. The latter may become a contributing factor as the concentration of SARS-CoV-2 in the grab samples may vary at different time points.Table 6 Median Reported COVID-19 cases and predicated infectious cases
Name Median COVID-19 reported cases Median predicated COVID-19 cases
Method 1 (Ahmed et al., 2020) Method 2 (Hemalatha et al., 2021)
WWTP-Z3 105 65,245 69,595
WWTP-Z5 156 15,067 16,071
WWTP-Z1 10 8705 7905
It is important to highlight that the models used in this study were based on the available models developed during the early phase of the pandemic which may be crude as compared to some of the current models developed. Recent studies (Oh et al., 2022; Zhan et al., 2022) have incorporated RT-qPCR data normalization using the PMMoV (Pepper Mild Mottle Virus) which is naturally present in the wastewater (contributed as part of human faeces) which may be more robust in estimating the depth of infection in the community.
Though the study was designed using information, methods and data available during the 1st quarter of 2021, there were also a few limitations that were beyond our control while executing the project. Although simultaneous sampling was initially envisaged in this study, strict lockdown imposed by Maharashtra State due to emergence of SARS-CoV-2 Delta variant (2nd COVID-19 wave in 2021) hampered continuous sampling from the three WWTPs. One cannot deny that these changes may partly affect the correlation co-efficient observed in this study. Furthermore, grab sampling method was used assuming the enriched samples at peak hours between 9 and 10 am over composite sampling, to avoid dilution of SARS-CoV-2 in composite samples.
To conclude, this study was undertaken to assess the presence of SARS-CoV-2 in the wastewater samples in three WWTPs of Mumbai city during the 2nd surge of COVID-19. The data obtained suggests that 1) PEG/NaCl method was relatively effective to concentrate the virus from the wastewater and can be used in resource limited settings. 2) SARS-CoV-2 RNA was efficiently eliminated during wastewater treatment from all the three WWTPs including the one with just a secondary treatment facility. 3) Although qualitative method can be used for routine screening of the wastewaters for SARS-CoV-2 RNA while quantitative data provides the number of active shedders who are already infected in the geographical area and hence giving an estimate of cases to manage any surge in infections in the future 4) The study also highlights the importance of upgrading the WWTPs up to tertiary treatment for reduction of the virus to non-detectable levels as observed in the two WWTPs equipped with tertiary treatment (WWTP-Z5 and WWTP-Z1) and 5) Continued monitoring of SARS-CoV-2 including its variants in wastewater is suggested for early warning system.
Acknowledgements
This work is funded by the Indo-US Science and Technology Forum (IUSSTF) [Ref No: IUSSTF/VN-COVID/081/2020]. We sincerely thank the authorities of Municipal Corporation of Greater Mumbai (MCGM) as well as the technical team working at all the WWTPs for their kind assistance. We thank Mr. Mayur Shelar and Mr. Shivam Nikam for helping with sample collection and transportation. We appreciate the assistance of all the BRC staff for their co-operation in this work.
Author Contributions
ND, ZB, JR and SS conceptualized the study, NDS and JR trained the BRC staff in sample processing and data analysis, DD, HW and SM coordinated sample collection, HW and SM carried out sample processing, HW, DD and ND analyzed the data, HW wrote the draft manuscript and DD edited the draft manuscript, all authors have read and approved the manuscript.
Declarations
Competing interests
The authors declare no competing interests.
Conflict of interest
The authors declare no conflict of interest.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Ahmed W Angel N Edson J Bibby K Bivins A Brien JWO Choi PM Kitajima M Simpson SL Li J Tscharke B Verhagen R Smith WJM Zaugg J Dierens L Hugenholtz P Thomas KV Mueller JF First confirmed detection of SARS-CoV-2 in untreated wastewater in Australia : A proof of concept for the wastewater surveillance of COVID- 19 in the community Science of the Total Environment 2020 728 138764 10.1016/j.scitotenv.2020.138764 32387778
Arora S Nag A Sethi J Rajvanshi J Saxena S Shrivastava SK Gupta AB Sewage surveillance for the presence of SARS-CoV-2 genome as a useful wastewater based epidemiology (WBE) tracking tool in India Water Science and Technology 2020 82 12 2823 2836 10.2166/wst.2020.540 33341773
Bhave PP Naik S Salunkhe SD Performance evaluation of wastewater treatment plant Water Conservation Science and Engineering 2020 5 1 23 29 10.1007/s41101-020-00081-x
Brian, I., Manuzzi, A., Dalla Rovere, G., Giussani, E., Palumbo, E., Fusaro, A., Bonfante, F., Bortolami, A., Quaranta, E. G., Monne, I., Patarnello, T., Bargelloni, L., Terregino, C., Holmes, E. C., Todesco, G., Sorrentino, F., Berton, A., Badetti, C., Carrer, C., … Panzarin, V. (2022). Molecular Monitoring of SARS-CoV-2 in Different Sewage Plants in Venice and the Implications for Genetic Surveillance. ACS ES and T Water, 2(11), 1953–1963. 10.1021/acsestwater.2c00013
Chakraborty P Pasupuleti M Shankar MRJ Bharat GK Krishnasamy S First surveillance of SARS-CoV-2 and organic tracers in community wastewater during post lockdown in Chennai, South India: Methods, occurrence and concurrence Science of the Total Environment 2020 778 146252 54 10.1016/j.scitotenv.2021.146252
Corman, V., Landt, O., Kaiser, M., Molenkamp, R., Meijer, A., Chu, D. K., Bleicker, T., Brünink, S., Schneider, J., Luisa Schmidt, M., Mulders, D., Haagmans, B. L., van der Veer, B., van den Brink, S., Wijsman, L., Goderski, G., Romette, J.-L., Ellis, J., Zambon, M., … Chantal, R. (2020). Detection of 2019-nCoV by RT-PCR. Eurosurveillance, 25(3), 1–8.
Fedorenko A Grinberg M Orevi T Kashtan N Survival of the enveloped bacteriophage Phi6 (a surrogate for SARS-CoV-2) in evaporated saliva microdroplets deposited on glass surfaces Scientific Reports 2020 6 1 10 10.1038/s41598-020-79625-z
Flood MT Souza ND Rose JB Gim T Methods evaluation for rapid concentration and quantification of SARS-CoV-2 in raw wastewater using droplet digital and quantitative RT-PCR Food and Environmental Virology 2021 13 3 303 315 10.1007/s12560-021-09488-8 34296387
Gendron L Verreault D Veillette M Moineau S Duchaine C Evaluation of filters for the sampling and quantification of RNA phage aerosols Aerosol Science and Technology 2010 44 10 893 901 10.1080/02786826.2010.501351
Government of India. (2011). Slums in India: A statistical compendium. https://nbo.gov.in/pdf/Slum_in_india_2011_english_book_23_May_12.pdf. Accessed 30 Nov 2022.
Haramoto E Malla B Thakali O Kitajima M First environmental surveillance for the presence of SARS-CoV-2 RNA in wastewater and river water in Japan Science of the Total Environment 2020 737 140405 10.1016/j.scitotenv.2020.140405 32783878
Hata A Hara-yamamura H Meuchi Y Imai S Honda R Detection of SARS-CoV-2 in wastewater in Japan during a COVID-19 outbreak Science of the Total Environment 2020 758 143578 10.1016/j.scitotenv.2020.143578 33221007
Hata A Honda R Hara-Yamamura H Meuchi Y Detection of SARS-CoV-2 in wastewater in Japan by multiple molecular assays-implication for wastewater-based epidemiology (WBE) MedRxiv 2020 10.1101/2020.06.09.20126417
Hemalatha M Kiran U Kuncha SK Kopperi H Gokulan CG Mohan SV Mishra RK Surveillance of SARS-CoV-2 spread using wastewater-based epidemiology: Comprehensive study Science of the Total Environment 2021 768 144704 10.1016/j.scitotenv.2020.144704 33736319
Joshi M Kumar M Srivastava V Kumar D First detection of SARS-CoV-2 Delta variant (B.1.617.2) in the wastewater of (Ahmedabad), India MedRxiv 2021 10.1101/2021.07.07.21260142 34100026
Kitajima M Ahmed W Bibby K Carducci A Gerba CP Hamilton KA Haramoto E Rose JB SARS-CoV-2 in wastewater: state of the knowledge and research needs Masaaki Sci Total Environ 2020 739 139076 10.1016/j.scitotenv.2020.139076 32758929
Kumar M Joshi M Patel AK Joshi CG Unravelling the early warning capability of wastewater surveillance for COVID-19: A temporal study on SARS-CoV-2 RNA detection and need for the escalation Environmental Research 2021 196 110946 10.1016/j.envres.2021.110946 33662347
Kumar M Joshi M Shah AV Srivastava V Dave S Wastewater surveillance-based city zonation for effective COVID-19 pandemic preparedness powered by early warning: A perspectives of temporal variations in SARS-CoV-2-RNA in Ahmedabad, India Science of the Total Environment 2021 792 148367 10.1016/j.scitotenv.2021.148367 34465041
La Rosa G Iaconelli M Mancini P Bonanno Ferraro G Veneri C Bonadonna L Lucentini L Suffredini E First detection of SARS-CoV-2 in untreated wastewaters in Italy Science of the Total Environment 2020 736 139652 10.1016/j.scitotenv.2020.139652 32464333
Lamba S Ganesan S Daroch N Paul K Joshi SG Sreenivas D Nataraj A Srikantaiah V Mishra R Ramakrishnan U Ishtiaq F SARS-CoV-2 infection dynamics and genomic surveillance to detect variants in wastewater—a longitudinal study in Bengaluru, India The Lancet Regional Health—Southeast Asia 2023 11 100151 10.1016/j.lansea.2023.100151 36688230
Medema G Heijnen L Italiaander R Brouwer A Presence of SARS-coronavirus—2 RNA in sewage and correlation with reported COVID-19 prevalence in the early stage of the epidemic in The Netherlands Environmental Science & Technology Letters 2020 10.1021/acs.estlett.0c00357
Michael-kordatou I Karaolia P Fatta-kassinos D Making waves: Wastewater surveillance of SARS-CoV-2 for population-based health management Journal of Environmental Chemical Engineering 2020 8 January 1 24 10.1016/j.jece.2020.104306
Nag A Arora S Sinha V Meena E Sutaria D Gupta AB Medicherla KM Monitoring of SARS-CoV-2 variants by wastewater-based surveillance as a sustainable and pragmatic approach—a case study of Jaipur (India) Water (Switzerland) 2022 14 3 1 19 10.3390/w14030297
Oh C Zhou A O’Brien K Jamal Y Wennerdahl H Schmidt AR Shisler JL Jutla A Keefer L Brown WM Nguyen TH Application of neighborhood-scale wastewater-based epidemiology in low COVID-19 incidence situations Science of the Total Environment 2022 852 June 158448 10.1016/j.scitotenv.2022.158448 36063927
Pandey D Verma S Verma P Mahanty B Dutta K Daverey A Arunachalam K SARS-CoV-2 in wastewater: Challenges for developing countries International Journal of Hygiene and Environmental Health 2021 231 113634 10.1016/j.ijheh.2020.113634 33039922
Peccia J Zulli A Brackney DE Grubaugh ND Kaplan EH Casanovas-Massana A Ko AI Malik AA Wang D Wang M Warren JL Weinberger DM Arnold W Omer SB Measurement of SARS-CoV-2 RNA in wastewater tracks community infection dynamics Nature Biotechnology 2020 38 10 1164 1167 10.1038/s41587-020-0684-z 32948856
Randazzo W Truchadc P Cuevas-Ferrando E Simón P Allende A Sánchez G SARS-CoV-2 RNA in wastewater anticipated COVID-19 occurrence in a low prevalence area Water Research 2020 181 115942 10.1016/j.watres.2020.115942 32425251
Sharif, S., Ikram, A., Khurshid, A., Salman, M., Mehmood, N., Arshad, Y., Ahmad, J., Angez, M., Alam, M. M., Rehman, L., Mujtaba, G., Hussain, J., Ali, J., Akthar, R., Malik, M. W., Baig, Z. I., Rana, M. S., Usman, M., Qasir, M., … Tahir, F. (2020). Detection of SARs-CoV-2 in wastewater, using the existing environmental surveillance network : An epidemiological gateway to an early warning for COVID-19 in communities Introduction. PLoS ONE. 10.1371/journal.pone.0249568.
Sharma DK Nalavade UP Kalgutkar K Gupta N Deshpande JM SARS-CoV-2 detection in sewage samples: Standardization of method & preliminary observations Indian Journal of Medical Research 2021 153 159 165 10.4103/ijmr.IJMR_3541_20 33818473
Sherchan SP Shahin S Ward LM Tandukar S Aw TG Schmitz B Ahmed W Kitajima M First detection of SARS-CoV-2 RNA in wastewater in North America: A study in Louisiana, USA Science of the Total Environment 2020 743 140621 10.1016/j.scitotenv.2020.140621 32758821
Srivastava, V., Gupta, S., Patel, A. K., Joshi, M., & Kumar, M. (2021b). Reflections of COVID-19 cases in the wastewater loading of SARS-CoV-2 RNA: A case of three major cities of Gujarat, India. In Case studies in chemical and environmental engineering (Vol. 4). Elsevier. 10.1016/j.cscee.2021.100115
Srivastava V Gupta S Patel AK Joshi M Kumar M Reflections of COVID-19 cases in the wastewater loading of SARS-CoV-2 RNA: A case of three major cities of Gujarat, India Case Studies in Chemical and Environmental Engineering 2021 4 April 4 9 10.1016/j.cscee.2021.100115
Weidhaas J Aanderud ZT Roper DK VanDerslice J Gaddis EB Ostermiller J Hoffman K Jamal R Heck P Zhang Y Torgersen K Laan J. Vander LaCross N Correlation of SARS-CoV-2 RNA in wastewater with COVID-19 disease burden in sewersheds Science of the Total Environment 2021 775 145790 10.1016/j.scitotenv.2021.145790 33618308
Whitworth C Mu Y Houston H Martinez-Smith M Noble-Wang J Coulliette-Salmond A Rose L Persistence of bacteriophage Phi 6 on porous and nonporous surfaces and the potential for its use as an Ebola virus surrogate Applied Environmental in Microbiology 2020 86 17 1 11 10.1128/AEM.01482-20
Wurtzer S Marechal V Mouchel JM Maday Y Teyssou R Richard E Almayrac JL Moulin L Evaluation of lockdown impact on SARS-CoV-2 dynamics through viral genome quantification in Paris in wastewater, Greater Paris, France, 5 March to 23 April 2020 Eurosurveillance Weekly 2020 25 50 2000776
Zhan, Q., Babler, K. M., Sharkey, M. E., Amirali, A., Beaver, C. C., Boone, M. M., Comerford, S., Cooper, D., Cortizas, E. M., Currall, B. B., Foox, J., Grills, G. S., Kobetz, E., Kumar, N., Laine, J., Lamar, W. E., Mantero, A. M. A., Mason, C. E., Reding, B. D., … Solo-Gabriele, H. M. (2022). Relationships between SARS-CoV-2 in wastewater and COVID-19 clinical cases and hospitalizations, with and without normalization against indicators of human Waste. ACS Environmental Science & Technology Water, 2(11), 1992–2003. 10.1021/acsestwater.2c00045
|
PMC010xxxxxx/PMC10155463.txt |
==== Front
J Infect
J Infect
The Journal of Infection
0163-4453
1532-2742
The British Infection Association. Published by Elsevier Ltd.
S0163-4453(23)00256-6
10.1016/j.jinf.2023.04.019
Letter to the Editor
Casirivimab-imdevimab treatment is associated with reduced rates of mortality and hospitalization in patients with COVID-19: A systematic review with meta-analysis
Gao Ming
Department of Cardiology, Chengdu First People’s Hospital, Chengdu, Sichuan, China
Ao Guangyu
Department of Nephrology, Chengdu First People’s Hospital, Chengdu, Sichuan, China
Hao Xiaodan
Department of Geriatrics, People's Liberation Army, The General Hospital of Western Theater Command, Chengdu, China
Xie Bo ∗
Department of Cardiology, Chengdu First People’s Hospital, Chengdu, Sichuan, China
∗ Correspondence to: No. 18 Wanxiang North Road, High-tech District, Chengdu, Sichuan 610095, China.
3 5 2023
7 2023
3 5 2023
87 1 8284
22 4 2023
© 2023 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
2023
The British Infection Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcDear Editor,
We read with great interest that monoclonal antibody (mAb) cocktail may serve as an effective and targeted therapeutic strategy in the treatment of patients with COVID-19.1 Since they can reduce virulence and viral load and enhance prognosis, specific anti-spike monoclonal antibodies, such as casirivimab-imdevimab, are an appealing alternative for treating COVID-19 infection.2
Casirivimab and imdevimab are two recombinant human IgG1 monoclonal antibodies that bind non-competitively to non-overlapping epitopes of the spike protein receptor-binding domain of SARS-CoV-2, thereby blocking the viral entry into host cells.2 Based on a randomized placebo-controlled clinical trial that showed a significant reduction in viral load among patients who received the casirivimab-imdevimab combination, the US Food and Drug Administration- Federal Agency (FDA), European Medical Agency (EMA), and Central Drug Standard Control Organization have authorized the use of imdevimab-casirivimab for emergency purposes.3 Casirivimab and imdevimab's Emergency Use Authorization (EUA) was modified in June 2021 to include subcutaneous delivery as a substitute for patients who cannot undergo intravenous (IV) infusion. Beginning in August 2021 and continuing through the beginning of January 2022, patients received treatment with subcutaneous injections of casirivimab and imdevimab. During this time, the health system had multiple COVID-19 case spikes including the Omicron and Delta varieties of concern.
There are now a number of studies reporting the effect of using casirivimab-imdevimab on patient clinical outcomes, however their findings are inconsistent. There have been no prior meta-analyses describing the association between casirivimab-imdevimab treatment and patient prognosis following COVID-19 infection to the best of our knowledge. Hence, we perform in this study the first meta-analysis in the literature to evaluate the relationship between casirivimab-imdevimab administration and patient outcomes following COVID-19 infection.
An electronic search was carried out between December 1, 2019, and April 1, 2023, in the databases of PubMed, Embase, the Cochrane Library, Scopus, medRxiv, and bioRxiv. There were no restrictions on publishing or language. The following search phrases and MeSH (Medical Subject Heading) terms were employed: (“coronavirus disease 2019 or novel coronavirus or SARS-CoV-2 or 2019-nCoV or COVID-19”) AND (casirivimab or imdevimab or REGN-COV or REGEN-COV2 or Ronapreve or REGN10933 or REGN10987).
The following criteria were used for inclusion: (1) patients that have COVID-19 confirmed; (2) clinical outcomes were evaluated between the casirivimab-imdevimab therapy and control groups (standard of care or placebo). The study excluded publications with identical content, reviews, letters, editorials, conference abstracts, case reports, and other publications. The first author's name, the publication year, the study's design, the participants' age, gender, the use of casirivimab-imdevimab, and the outcomes of interest (mortality and hospitalization) were also collected as data on the studies' initial characteristics.
The statistical analysis was carried out using Review Manager, version 5.2 (Cochrane Collaboration, Oxford). The odds ratio (OR), with a 95% confidence interval, was employed to evaluate dichotomous variables. We assessed the heterogeneity of the data using the I2 statistic and the Cochran's Q test. A P value of 0.05 or less indicates statistical significance. The protocol for this study is registered with PROSPERO (CRD42023418212).
Through a thorough literature search, a total of nine studies, including 84,875 in the casirivimab-imdevimab group and 322,943 in the control group arm, were identified for this meta-analysis.2, 3, 4, 5, 6, 7, 8, 9, 10 Table 1 lists the illness features and demographics of the 407,818 patients who were included in the pooled study. The majority of the included studies were from the United States. The other investigations were retrospective and prospective cohort studies, while one research was a randomized controlled study. Patients with mild-to-moderate COVID-19 were diagnosed in the majority of trials. Three studies included COVID-19 individuals who were all outpatients.4, 5, 6 Casirivimab-imdevimab was given intravenously or subcutaneously in the included studies. The selected studies, which were all published between 2022 and 2023, had various sample patient sizes ranging from 152 to 384,447 patients with COVID-19.Table 1 Characteristics of included studies.
Table 1Study Country Study type Sample size Intervention Patients included Key study outcomes
Al-Obaidi 2022 United States Retrospective cohort 8426 Casirivimab-imdevimab High-risk patients with COVID-19 Mortality, hospitalization, ICU admission
Bierle 2021 United States Retrospective cohort 630 Casirivimab-imdevimab Mild to moderate COVID-19 Hospitalization, number of patients with hypoxia
Hussein 2022 United States Retrospective cohort 384,447 Casirivimab-imdevimab Outpatients with COVID-19 Mortality, composite outcome of all-cause mortality or COVID-19-related hospitalizations
Joy 2022 India Retro-prospective comparative study 152 Casirivimab-imdevimab Patients amidst and post COVID-19 treatment Mortality, hospitalization, need for mechanical ventilation, high flow O2 requirement
McCreary 2022 United States Prospective cohort 1959 Casirivimab-imdevimab Outpatients with mild to moderate COVID-19 Mortality, hospitalization, emergency department admission or hospitalization
Razonable 2021 United States Retrospective cohort 1392 Casirivimab-imdevimab Mild to moderate COVID-19 Mortality, hospitalization, ICU admission
RECOVERY 2022 United Kingdom RCT 9785 Casirivimab-imdevimab vs usual care Patients admitted to hospital with COVID-19 Mortality, need for mechanical ventilation, renal replacement therapy
Rhudy 2023 United States Retrospective cohort 1170 Casirivimab-imdevimab Outpatients with symptomatic COVID-19 Mortality, hospitalization, emergency department admission
Suzuki 2022 United States Retrospective cohort 444 Casirivimab-imdevimab Mild to moderate COVID-19 Mortality, need for mechanical ventilation//ECMO, deterioration during hospitalization
RCT: randomized controlled trial; ICU: intensive care unit; ECMO, extracorporeal membrane oxygenation.
The meta-analysis revealed that the casirivimab-imdevimab treatment was associated with a lower mortality rate than the control group who did not receive casirivimab-imdevimab (OR=0.21, 95%CI: 0.06–0.68, P = 0.03; I2 =97%) ( Fig. 1A). In addition, patients who received casirivimab-imdevimab had significantly lower hospitalization rates (OR=0.31, 95%CI: 0.20–0.48, P＜0.00001; I2 =76%) (Fig. 1B).Fig. 1 A. Association between casirivimab-imdevimab treatment and mortality, B. Association between casirivimab-imdevimab treatment and hospitalization.
Fig. 1
The results demonstrate a significant beneficial effect of casirivimab-imdevimab treatment on mortality and hospitalization in COVID-19 patients.
As a result of a higher risk of severe COVID-19 among those who are not vaccinated, the reduction in outcome risk among patients treated with casirivimab-imdevimab was marginally greater among unvaccinated patients than among vaccinated patients.4 The results demonstrate that treatment with casirivimab-imdevimab can be beneficial not only for people who cannot or do not want to take the COVID-19 vaccine, but also for those who have received the vaccine. The findings show that vaccinated individuals with COVID-19 can also benefit from therapy with casirivimab-imdevimab. This is in addition to the benefits of treatment for those who cannot or do not want to receive the vaccine. In addition, although RECOVRY collaborative group found that therapeutic use of casirivimab and imdevimab combination in the hospital setting would be best restricted to seronegative patients, the development of the variant (omicron), which can evade antibodies raised against previous SARS-CoV-2 variants, the validity of seropositive status as a predictor of treatment non-response to monoclonal antibodies is weakened.7
Our study has several limitations. With nine included literature, the meta-analysis's sample size was relatively small. In terms of mortality and hospitalization, there was also a considerable heterogeneity. Additionally, the patient populations' immunization status, baseline serostatus, type of viral variants, and use of casirivimab-imdevimab in different trials may differ. Despite these limitations, our study has significant importance as the first meta-analysis to examine the outcomes of casirivimab-imdevimab therapy in COVID-19-infected patients.
To sum up, using casirivimab-imdevimab to treat COVID-19 patients has considerable advantages in terms of preventing hospitalization and mortality. These results need to be confirmed by further research.
Funding
None declared.
Declaration of Competing Interest
The authors declare that they have no competing interest.
Acknowledgments
None.
==== Refs
References
1 Wang Y. Zheng J. Zhu K. Xu C. Wang D. Ho M. The effect of tixagevimab-cilgavimab on clinical outcomes in patients with COVID-19: a systematic review with meta-analysis J Infect 86 1 2023 e15 e17 10.1016/j.jinf.2022.08.021 36031156
2 Razonable R.R. Pawlowski C. O'Horo J.C. Arndt L.L. Richard Arndt R. Bierle D.M. Casirivimab-Imdevimab treatment is associated with reduced rates of hospitalization among high-risk patients with mild to moderate coronavirus disease-19 eClinicalMedicine 40 2021 101102 10.1016/j.eclinm.2021.101102
3 Joy A.P. Karattuthodi M.S. Chandrasekher D. Augustine. A.T. The impact of casirivimab-imdevimab antibody cocktail in patients amidst and post COVID 19 treatment: a retro-prospective comparative study in India J Assoc Physicians India 70 4 2022 11 12
4 Hussein M. Wei W. Mastey V. Sanchez R.J. Wang D. Murdock D.J. Real-world effectiveness of casirivimab and imdevimab among patients diagnosed with COVID-19 in the ambulatory setting: a retrospective cohort study using a large claims database BMJ Open 12 12 2022 e064953 10.1136/bmjopen-2022-064953
5 McCreary E.K. Bariola J.R. Wadas R.J. Shovel J.A. Wisniewski M.K. Adam M. Association of subcutaneous or intravenous administration of casirivimab and imdevimab monoclonal antibodies with clinical outcomes in adults with COVID-19 JAMA Netw Open 5 4 2022 e226920 10.1001/jamanetworkopen.2022.6920
6 Rhudy C. Bochenek S. Thomas J. James G.S. Zeltner M. Platt T. Impact of a subcutaneous casirivimab and imdevimab clinic in outpatients with symptomatic COVID-19: a single-center, propensity-matched cohort study Am J Health Syst Pharm 80 3 2023 130 136 10.1093/ajhp/zxac305 36264659
7 RECOVERY Collaborative Group Casirivimab and imdevimab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trial Lancet 399 10325 2022 665 676 10.1016/S0140-6736(22)00163-5 35151397
8 Al-Obaidi M.M. Gungor A.B. Nematollahi S. Zangeneh T.T. Bedrick B.J. Johnson K.M. Effectiveness of casirivimab-imdevimab monoclonal antibody treatment among high-risk patients with severe acute respiratory syndrome coronavirus 2 B.1.617.2 (Delta variant) infection Open Forum Infect Dis 9 7 2022 ofac186 10.1093/ofid/ofac186 35791354
9 Bierle D.M. Ganesh R. Razonable R.R. Breakthrough COVID-19 and casirivimab-imdevimab treatment during a SARS-CoV-2 B1.617.2 (Delta) surge J Clin Virol 145 2021 105026 10.1016/j.jcv.2021.105026
10 Suzuki Y. Shibata Y. Minemura H. Nikaido T. Tanino Y. Fukuhara A. Real-world clinical outcomes of treatment with casirivimab-imdevimab among patients with mild-to-moderate coronavirus disease 2019 during the Delta variant pandemic Int J Med Sci 19 5 2022 834 841 10.7150/ijms.71132 35693744
|
PMC010xxxxxx/PMC10161182.txt |
==== Front
Curr Nutr Rep
Curr Nutr Rep
Current Nutrition Reports
2161-3311
Springer US New York
37145350
471
10.1007/s13668-023-00471-2
Hot Topic
A Narrative Review on the Potential Role of Vitamin D3 in the Prevention, Protection, and Disease Mitigation of Acute and Long COVID-19
http://orcid.org/0000-0003-3951-7712
Moukayed Meis mmoukayed@hotmail.com
grid.444482.a 0000 0004 1776 0065 Present Address: School of Arts and Sciences, American University in Dubai, Al Asad Street, PO Box 28282, Dubai, United Arab Emirates
5 5 2023
2023
12 2 215223
11 4 2023
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Purpose of Review
The coronavirus disease 2019 (COVID-19) pandemic has challenged global health systems and economies from January 2020. COVID-19 caused by the infectious severe acute respiratory syndrome coronavirus (SARS-CoV-2) has acute respiratory and cardiometabolic symptoms that can be severe and lethal. Long-term physiological and psychological symptoms, known as long COVID-19, persist affecting multiple organ systems. While vaccinations support the fight against SARS-CoV-2, other effective mechanisms of population protection should exist given the presence of yet unvaccinated and at-risk vulnerable groups, global disease comorbidities, and short-lived vaccine responses. The review proposes vitamin D3 as a plausible molecule for prevention, protection, and disease mitigation of acute and long COVID-19.
Recent Findings
Epidemiological studies have shown that individuals who were deficient in vitamin D3 had worse COVID-19 health outcomes and mortality rates. Higher doses of vitamin D3 supplementation may improve health and survivorship in individuals of various age groups, comorbidities, and severity of disease symptoms.
Summary
Vitamin D3’s biological effects can provide protection and repair in multiple organ systems affected by SARS-CoV-2. Vitamin D3 supplementation can potentially support disease-mitigation in acute and long COVID-19.
Keywords
COVID-19
Vitamin D
SARS-CoV-2
Public health benefits
Long COVID-19
issue-copyright-statement© Springer Science+Business Media, LLC, part of Springer Nature 2023
==== Body
pmcIntroduction
The COVID-19 Pandemic in Global Perspective
The coronavirus disease 2019 (COVID-19) is an infectious disease caused by the severe acute respiratory syndrome coronavirus (SARS-CoV-2) which was declared a Public Health Emergency of International Concern by the World Health Organization (WHO) on 30 January 2020. To date, 29 March 2023, according to WHO, 761,402,282 cases of SARS-CoV-2 infection have been confirmed, and 6,887,000 deaths have been reported due to COVID-19 [1]. Lockdowns, boarder closures affecting trade and travel, morbidity and mortality associated with COVID-19 have been globally devastating. The International Monetary Fund reported that the global economy contracted by 3.5% in 2020 alone [2], and 90% of the global economy experienced a contraction in per capita GDP [3]. While the emotional value of losing a loved one is immeasurable, the economic cost using the value per statistical life (VSL) estimate is $10 million per lost life [4]. Long-term economic losses remain elusive. While global economic recovery is predicted, it is even more important to ensure recovery of population health and well-being. Such recovery needs to promote lifestyle behaviors that would mitigate future risk and burden from newly emergent microbes and pandemics of similar public health scale. This continues to be pertinent because new SARS-CoV-2 variants continue to emerge, and global countries, including China, continue to enforce lockdowns to prevent COVID-19 outbreaks when needed. Moreover, recent findings indicate that COVID-19 has novel long-term physiological and psychological effects that affect multiple organ systems with observed neurological, inflammatory, metabolic, muscular, reproductive, cardiac, and psychological symptoms, among others, which can persist. Such novel manifestations may occur in SARS-CoV-2-infected individuals regardless of initial disease severity or symptoms. Vaccinations have minimized global infection rates of SARS-CoV-2 and COVID-19 disease burden, but are inequitably available globally, have short-lived efficacy and recently reported side effects. Therefore, other effective mechanisms of population protection should exist especially in the presence of yet unvaccinated, underprivileged, and at-risk groups, and short-lived vaccine responses that require constant boosters or have side effects that cause concern. The need is pertinent in the presence of highly prevalent metabolic disease comorbidities in global populations, such as cardiovascular disease and obesity, and continuous emergence of new variants. Additionally, other previously dormant microbes, e.g., monkeypox virus, Marburg virus, and Crimean-Congo hemorrhagic fever virus have reemerged recently with greater virulence, reminding everyone that potential similar public health dangers are always lurking. Considering the above, and the enigmatic emergent long COVID-19 symptoms, it becomes a public health need to present sustainable, accessible, cost-effective health solutions for disease prevention and management.
SARS-CoV-2, Coronaviruses, and Infectivity
SARS-CoV-2 is a ribovirus member of the Coronaviridae family of viruses. Coronaviruses are zoonotic and can infect several species including bats, cats, rodents, dogs, pigs, birds, cows, minks, camels, rabbits, and humans, all of which are viral reservoirs of transmission [5, 6]. SARS-CoV-2 is the 7th coronavirus that is known to infect humans. The SARS-CoV-2 virus shares 79% genetic homology with SARS-CoV-1 (2002 pandemic), 50% homology with Middle East respiratory syndrome coronavirus (MERS-Co-V) (2012 pandemic) and high sequence homology with several other Rhinolophus coronaviruses isolated in bats but which have yet not bridged from animal reservoirs to infect humans [7].
SARS-CoV-2 is a positive-sense single-stranded RNA-virus. It has six functional open reading frames (ORFs) coding for essential structural proteins that promote viral replication and infectivity: replicase, spike (S), envelop, membrane and nucleocapsid proteins. Additionally, seven other ORFs encode accessory proteins [8]. Viral infection involves the spike viral protein subunits S1 and S2 and the Angiotensin Converting Enzyme 2 (ACE2) receptors present in human host tissues [9]. Several variants of SARS-CoV-2 have been reported especially those mutated at the S domain, with six major variants of concern reported to be more infective by WHO [10].
SARS-CoV-2 Pathophysiology in the Short and Long Term
Infection with SARS-CoV-2 is problematic in the short and long term. In the short term, viral infection and ensuing ‘cytokine storm’ may result in mild to severe physiological respiratory, hematological, renal, cardiac, and gastrointestinal symptoms that may cause morbidities and mortality [11•]. Post infection, over 55 residual long COVID-19 symptoms may develop or persist which affect multiple organ systems with emergent novel complications. These can include immunological, metabolic, musculoskeletal and nervous system disorders such as weakness, fatigue, dyspnea, sarcopenia, incontinence, neuropathies, encephalopathies, cerebral strokes [12, 13], increase in endocrine disturbances such as new-onset diabetes mellitus [14], hair loss, anosmia, dysgeusia, headache, attention disorder, deteriorating brain and cognition functions, declining mental health, among other complications [15, 16, 17•]. Often these symptoms of long COVID-19 persist up to and over 60 days after disease onset [18]. Increasing age, female sex, white ethnicity, poor pre-pandemic general and mental health including metabolic comorbidities, immunocompromization, obesity, and asthma, are associated with prolonged symptoms [17•, 19, 20, 21•]. Given that the SARS-CoV-2 virus has the potential to invade ACE2 receptor-containing cells and tissues, endothelial cells on vessels and organs, leukocytes, and dendritic cells, this implies that potentially there are many effects to the virus that are yet unknown and could cause multi-organ system long-term dysfunction. Some of these organs and tissues include kidneys, adipose tissues, nerves, heart, pancreas, testes, among others [22]. Given SARS-CoV-2’s versatility in infecting multiple organ system targets, scientists need to find biological molecules with similar omnipotent diversity of action that could protect multiple organs and mitigate disease burden. These molecules’ actions would include priming individuals’ immunity, reducing viral replication and infection, attenuating comorbidities’ risk, promoting organ repair and recuperation, enhancing baseline health, and promoting survival. We propose that vitamin D3 could be a cost-effective biological molecule that exerts such diverse biological effects on multiple tissues and systems [23], and promotes protective and regenerative mechanisms against SARS-CoV-2, acute and long COVID-19.
Vitamin D3 has Many Health Benefits for Multiple Organ Systems and Well-being
Vitamin D3 can be synthesized in the human body following sun exposure. Synthesis begins via the skin by conversion of 7-dehydrocholesterol to vitamin D3 (also termed cholecalciferol). A 15-min half body exposure produces approximately 10,000 International Units (IU) (250 mcg) to 20,000 IU (500 mcg) of vitamin D3, cholecalciferol. This depends on several factors including duration of sun exposure and skin color. Upon sun avoidance or low sun exposure due to different ecological latitudes, vitamin D3 may need to be ingested from foods such as oily fish, eggs, fortified juice or milk or cereals, and a range of animal-based products such as liver. Vitamin D3 synthesized via the skin, is then metabolized in the liver to 25-hydroxyvitamin D3 (25(OH)D3) (termed calcidiol), followed by conversion in the kidneys to the active biological form 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) (termed calcitriol). Regular daily sun exposure or supplementation can support reaching physiological vitamin D3 sufficiency which is attained with a serum 25(OH)D3 concentration at or greater than 30 ng/mL [24•]. The National Academy of Sciences, Engineering and Medicine (NASEM) (previously Institutes of Medicine (IOM)) recommends a somewhat conservative daily recommended dietary allowance (RDA) for vitamin D3 of 600 IU/day (15 mcg/day). The Endocrine Society recommends at least 1500–2000 IU/day (37.5–50 mcg/day) and up to 4,000 IU/day (100 mcg/day) for adults, with 10,000 IU/day (250 mcg/day) as upper tolerable level (UL). Deficiencies can be treated with supplementation of up to 50,000 IU/week (1,250 mcg/week). The above recommendations require revision as studies have revealed a requirement for significantly higher vitamin D3 concentrations for health maintenance and disease prevention, and adjustment of RDAs and ULs according to individual baseline vitamin D3 concentrations [25•, 26, 27].
Many in vitro, animal model, translational, observational, and clinical studies have confirmed a variety of roles for vitamin D3 in maintaining health and preventing disease. Vitamin D3, an important nutrient and hormone exerts its biological effects via binding its vitamin D receptor (VDR) and modulating downstream gene transcription of hundreds of genes in multiple organ systems. Some of vitamin D3’s roles include improvement in brain functions such as cognition, memory and mood, supporting calcium homeostasis and bone remodeling, enhancing cardiometabolic health and endothelial function, regulating blood pressure, improving insulin secretion [28•] and sensitivity, and supporting placental function in pregnant women [23].
Vitamin D3 can modulate both the innate and adaptive immune system [29]. At airway epithelia, vitamin D3 enhances secretion of antimicrobial peptides such as β-defensins and cathelicidins which inhibit the cellular entry and subsequent proliferation of virus particles. Vitamin D3 induces upregulation of cathelicidin human cationic antimicrobial protein (hCAP18), in neutrophils, natural killer cells, monocytes, and B-cells. In macrophages, vitamin D3 enhances autophagy through upregulation of calcium and nitric oxide levels, inhibition of mammalian target of rapamycin (mTOR) thus promoting clearance of virus-infected cells. Vitamin D3 could therefore potentially override the SARS-CoV-2 ORF3a impairment of autophagy, and its immune-modulating properties in response to SARS-CoV-2 infection should be further studied.
Vitamin D3 is an effective antioxidant and anti-inflammatory agent. Vitamin D3 sufficiency significantly protects against diseases where inflammation is a hallmark of disease progression. These include cancer, Alzheimer’s disease, Multiple Sclerosis, and rheumatoid arthritis [29]. Vitamin D3 can combat the anti-inflammatory ‘cytokine storm’ and fibrotic effects associated with COVID-19. Vitamin D3 upregulates the expression of nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor alpha (IkBα) which inhibits the proinflammatory transcription factor nuclear factor kappa light-chain enhancer of activated B cells (NFkB), resulting in the reduced expression of inflammatory genes. In dose and cell-specific mechanisms, vitamin D3 can modulate gene expression of Interleukins IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, and IL-15, Interferon gamma (IFN-γ), Tumour growth factor beta (TGF-β) and Tumor necrosis factor alpha (TNF-α) in directions that promote anti-inflammatory and tissue repair pathways [30, 31].
In the adaptive immune system, the VDR receptor is expressed in activated mature T-cells. Immunomodulatory effect of 1,25(OH)2D3 on CD4+ T-cells involves suppression of the proinflammatory Type-1 T helper (Th1) and Type-17 T-helper (Th17) subsets that cause tissue damage, in favor of activation of the Type-2 T helper (Th2) cells and Induced regulatory T-cells (iTreg). This shift towards Th2 and iTreg subsets favors B-cell activation in humoral immunity, anti-inflammatory effects, and tissue healing pathways through dose-dependent TGF-β induction. Studies do not show that vitamin D3 promotes T-cell proliferation from naive precursors [32•].
In studies on RNA stranded viruses such as Respiratory Syncytial Virus, H9N2 Influenza virus, rotavirus, Human Immunodeficiency Virus, vitamin D3 demonstrated antiviral activity, reduction of viral load, select modulation of viral gene transcription, or protection of cells against virus-induced cell death. However, vitamin D3 effects on different classes of DNA and RNA viruses remain to be examined in big clinical trials as its effects are currently contradictory depending on the virus types and severity of infection [30, 33]. Vitamin D3 needs to be studied for its potential anti-viral effects against SARS-CoV-2.
Vitamin D3 Supplementation Mitigates COVID-19 Disease Severity, Morbidity and Improves Survival in Different Age Groups
Several epidemiological studies have explored the link between vitamin D3 and SARS-CoV-2 or COVID-19. While heterogenous in type, criteria, and results, collectively these studies showed that vitamin D3 deficiency or insufficiency were associated with increased risk of SARS-CoV-2 infection and COVID-19 severity [34]. In a matched retrospective case-control study on 464,383 participants in two matched case-control groups of individuals for which 25(OH)D3 concentrations and body mass index (BMI) were available before the pandemic, results showed that individuals with very low 25(OH)D3 concentrations (< 30 nmol/L) had the highest risks for SARS-CoV-2 infection, and for COVID-19 severity when infected; Odds Ratio (OR) were 1.246 [95% Confidence Interval (CI): 1.210–1.304] and 1.513 [95% CI: 1.230–1.861], respectively [35]. In a retrospective observational study of 191,779 participants in the USA, Kaufman et al. showed that SARS-CoV-2 infection incidence was higher in individuals with lower serum 25(OH)D3 concentrations. The SARS-CoV-2 positivity rate was higher in the 39,190 individuals with 25(OH)D3 deficiency (< 20 ng/mL) [12.5%, 95% C.I. 12.2–12.8%] than in the 27,870 individuals with sufficient serum 25(OH)D3 concentrations (30–34 ng/mL) [8.1%, 95% C.I. 7.8–8.4%] and in the 12,321 individuals with serum 25(OH)D3 ≥ 55 ng/mL [5.9%, 95% C.I. 5.5–6.4%]. This association held true for individuals across geographical latitudes, ethnicities, sexes and age groups [36].
In a study of 70-year-old elderly individuals hospitalized for COVID-19, Sulli et al. showed that 25(OH)D3 serum deficiency was associated with more severe lung involvement, longer disease duration, and risk of death. In these elderly participants, 25(OH)D3 levels were significantly lower compared to controls, (median 7.9 vs. 16.3 ng/mL, respectively, p = 0.001). Lower 25(OH)D3 serum concentrations were significantly correlated with elevated D-Dimer and C-reactive proteins concentrations, two important indicators of coagulopathy and inflammation. A negative correlation was observed between 25(OH)D3 serum concentrations and severity of radiologic pulmonary involvement and mortality during hospitalization. Higher 25(OH)D3 serum concentrations were associated with improved pulmonary parameters such as partial oxygen pressure (PaO2) (p = 0.03), oxygen saturation (sO2) (p = 0.05), and the ratio of arterial oxygen partial pressure (PaO2 in mmHg) to fractional inspired oxygen (PaO2/FiO2) (p = 0.02). Elderly individuals with low 25(OH)D3 had multiple lung consolidations with diffuse severe interstitial lung involvement compared to those with higher 25(OH)D3 concentrations who had milder lung symptoms. These studies above and other similar investigations suggest that maintaining higher Vitamin D3 status prior to infection could be protective and reduce the severity of COVID-19-induced organ injury [37].
The timing of vitamin D3 administration may also be associated with improved health outcomes. A quasi-experimental study in Italy compared the effects of vitamin D3 cholecalciferol administration on improving survival outcomes in elderly hospitalized individuals. Participants were divided into three groups, Group 1 administered vitamin D3 for the year preceding infection and hospitalization (bolus included the doses of 50,000 IU vitamin D3 per month, or the doses of 80,000 IU or 100,000 IU vitamin D3 every 2–3 months), Group 2 administered an oral supplement of 80,000 IU vitamin D3 within a few hours of the diagnosis of COVID-19, and a third no-supplementation reference Group 3. Survival of participants were, in Group 1 (93.1%), Group 2 (81.2%) and Group 3 (68.7%) respectively, (p = 0.02 between groups 1 and 3). After correction for any confounding variables, using Group 3 as reference hazard ratio (HR) = 1, the adjusted HR for 14-day mortality was HR = 0.07 (p = 0.017) for Group 1 and HR = 0.37 (p = 0.28) for Group 2. Group 1 had longer survival time than Group 3 (log-rank p = 0.015), although there was no survival difference between Groups 2 and 3 (log-rank p = 0.32). Group 1, but not Group 2 (p = 0.40), was associated with improved clinical outcomes using an ordinal scale for clinical improvement (OSCI). These findings show that regular bolus vitamin D3 supplementation as a preventive habit or lifestyle before viral infection was associated with better outcome and survival in frail elderly individuals [38, 39]. This finding consolidates data from other studies which show that individuals who have sufficient concentrations of 25(OH)D3 in their blood stream due to previous adequate vitamin D3 intake or sun exposure have a much higher survival rate that those who are deficient in vitamin D3. This implies that maintaining vitamin D3 sufficiency in individuals before infection can have preventive value and public health benefits. Sustained supplementation with vitamin D3 should be considered, and further population studies should be performed to verify supplementation needs, doses, and regimen.
A randomized open-label, double-masked clinical trial examined the effect of administering cholecalciferol to hospitalized individuals in addition to the best available therapy present at the time, namely hydroxychloroquine or azithromycin. Upon randomization in ratio 2:1 cholecalciferol to no-cholecalciferol respectively, cholecalciferol was given orally as 0.532 mg (21,280 IU) upon admission, then 0.266 mg (10,640 IU) on day 3 and 7, and then weekly until discharge or Intensive Care Unit (ICU) admission. From the 50 participants on cholecalciferol treatment, only 1 required ICU admission (2%), none died, and all were discharged without major complications. In the 26 control/no-cholecalciferol treatment participants, 13 (50%) required ICU admissions, two of whom died. When adjusting for comorbidities, hypertension and type 2 diabetes mellitus, the OR for ICU admission in the cholecalciferol treatment versus controls was OR = 0.03 [95% CI: 0.003–0.25]. None of the treated participants suffered any side effects from the higher dose cholecalciferol [40].
In contrast, Abroug et al. showed in a randomized controlled trial (RCT) that if a single vitamin D3 treatment of 200,000 IU was given to individuals with moderate to severe SARS-CoV-2 infection but after 14 days after initial diagnosis, such a treatment does not appear to have a significant effect on shortened recovery time in vitamin D-treated individuals compared to participants given a placebo [41].
In a double-blind RCT conducted in Brazil, Murai et al. investigated the effects of a single oral dose of treatment with 200,000 IU of vitamin D3 versus placebo on the clinical outcomes of 240 individuals hospitalized with moderate to severe COVID-19 symptoms. Analysis of the RCT data showed that there were no significant differences between individuals in the vitamin D3 treated group versus the placebo group in the need for mechanical ventilation (7.6% versus 14.4%; difference, -6.8% [95% CI, -15.1% to 1.2%]; p = 0.09), admission to ICU (16.0% versus 21.2%; difference, -5.2% [95% CI, -15.1% to 4.7%]; p = 0.30), mean length of hospital stay (7.0 [4.0–10.0] days versus 7.0 [5.0–13.0] days; log-rank p = 0.59), and in-hospital mortality (7.6% versus 5.1%; difference, 2.5% [95% CI, -4.1% to 9.2%]; p = 0.43). The high dose treatment successfully elevated serum 25(OH)D3 concentrations with no noteworthy adverse events reported. The findings may imply that treatment with vitamin D3 after the onset of moderate to severe short COVID-19 cascade of symptoms may not be so effective at improving clinical outcomes and survival [42].
In a prospective observational study conducted on 410 subjects affected by COVID-19 in India, Jevalikar et al. found no association between baseline 25(OH)D3 and severity of COVID-10 symptoms or survival rates. The investigators compared two groups of participants; Group 1 had participants who had vitamin D3 deficiency (VDD) with 25(OH)D3 < 20 ng/ml versus Group 2 which included participants who did were not classified as vitamin D3 deficient. Jevalikar et al. showed that the proportion of severe cases (13.2% vs.14.6%), mortality (2% vs. 5.2%), oxygen requirement (34.5% vs.43.4%), and ICU admission (14.7% vs.19.8%) was not significantly different between individuals with or without VDD, respectively. A limitation of this study lies in the fact that participants with VDD were significantly younger and had lesser comorbidities, which may have influenced some of the outcomes and findings [43], in contrast to several other studies that reported an association with vitamin D deficiency and disease severity [44•].
In a multicenter non-randomized cohort study, 537 participants who were hospitalized in 5 hospitals in Spain were administration capsules of the active metabolite calcifediol, 25(OH)D3 upon entry and throughout a 30-day hospitalization period. The dose per capsule was 0.266 mg/capsule and administered as 2 capsules on entry and then one capsule on day 3, 7, 14, 21, and 28. Participants receiving 25(OH)D3 had lower comorbidity and score of pneumatic severity (including blood pressure measures, urea, respiratory distress), lower C-reactive protein indicative of lower inflammation, lower rates of chronic kidney disease and blood urea nitrogen. In the group that received the 25(OH)D3, mortality rate was 5% compared to 20% in the no-treatment group with OR = 0.22 in these 25(OH)D3 recipients compared to controls [95% CI, 0.08 to 0.61] [45].
The COvid19 and VITamin D TRIAL (COVIT-TRIAL), a multicentered, open-label RCT conducted in France tested whether a single high oral dose of vitamin D3 cholecalciferol of 400,000 IU administered within 72 h after SARS-CoV-2 infection could improve overall 14-day survival of elderly participants compared to standard-dose, 50,000 IU participants. Fourteen days following infection, participants allocated to higher dose cholecalciferol suffered fewer deaths (6%) compared to participants on lower dose (11%), adjusted HR = 0.39 [95% CI: 0.16 to 0.99] (p = 0.045). This protective effect was not sustained for 28-day survival. Higher dose cholecalciferol did not result in more frequent adverse effects compared to the standard dose [38, 46].
In a study on 78 healthcare workers infected by SARS-CoV-2, supplementation of vitamin D3 resulted in milder symptoms and clinical features. Two groups were treated for three months. Group 1 participants received supplementation of 50,000 IU/week for two weeks followed by daily doses of 5,000 IU/day, while participants in Group 2 were supplemented with 2,000 IU/day for three months. Normalization of serum 25(OH)D3 concentrations was achieved in 53% of those on the higher dose of vitamin D3 supplementation. Vitamin D3 serum concentrations could not be correlated in this small size study with reduced morbidity. However, data showed that participants receiving higher dose of 5,000 IU/day vitamin D3 supplementation had milder to no symptoms of SARS-CoV-2 compared to participants on lower 2,000 IU/day suggesting potential protective effects against severity of symptoms of COVID-19. Neither vitamin D3 intake nor vitamin D3 deficiency or insufficiency were associated with a decrease in SARS-CoV-2 - associated morbidity [47]. The above studies collectively suggest that supplementation with higher doses of vitamin D3 could support reduction in COVID-19 disease severity and symptoms, and a moderate improvement in survival, by promoting earlier and faster recovery mechanisms. The timing of vitamin D3 administration appears to be important. Several studies suggest that maintaining vitamin D3 sufficiency before infection or disease symptoms, or supplementing with vitamin D3 at early stages of disease progression may achieve significantly reduced symptoms and improved clinical outcomes in individuals with COVID-19 [44•, 48•].
Vitamin D3 Could Improve Brain Function in Individuals Affected by Long COVID-19
COVID-19 can cause physiological and cognitive abnormalities in acute and long COVID-19 [49]. The acute underlying causes of brain dysfunction due to the SARS-CoV-2 may involve inflammatory damage in the brain both in the neurons and the endothelial vasculature, instigated by viral infiltration and endothelial cell activation. Leaky endothelium would enable complement and immune cell infiltration of the protective blood brain barrier [50]. In addition to the ensuing hypercoagulation or hyperinflammation, the virus interferes with neurological functions which may explain the brain’s long COVID-19 symptoms of fatigue, brain fog, loss of focus, memory deficit, and other cognitive abnormalities. Vitamin D3 potentially could abrogate these symptoms, protect brain function, and promote reparative healing even after infection. Both 25(OH)D3 and 1,25(OH)2D3 can cross the blood brain barrier. In stroke patients, vitamin D3 has a protective and reparative effect on neuronal damage especially in ischemic stroke symptoms [51•]. Vitamin D3 plays an integral role in maintaining brain plasticity, enabling proficient cognitive function, and preventing neurodegeneration. Vitamin D3 deficiency is associated with developmental neuronal functional abnormalities in growing fetuses, poor memory and concentration in adults, lower cognitive function, reduced brain volume and hippocampus function in adults, as well as neurodegenerative or psychological diseases such as Alzheimer’s, Multiple Sclerosis, and depression [52–54]. The association with low vitamin D3 and worse cognitive performance is more pronounced in women with depression [53]. Interestingly, more women than men are affected by the brain cognitive dysfunction associated with long COVID-19. Apropos, the repletion, sufficiency, and administration of vitamin D3 to protect against the long COVID-19-associated brain abnormalities should be considered and further studied.
Vitamin D3 Could Improve Muscle Regeneration and Repair in Individuals with Long COVID-19
Many individuals experiencing long COVID-19 symptoms present with fatigue, weakness, and myalgia for reasons that remain to be elucidated. Lessons from diseases such as Chronic obstructive pulmonary disease (COPD) allude to the possibility that vitamin D3 is a feasible option to improve disease-associate muscle function. In COPD, obstructive lung disease of the peripheral airways occurs, and is associated with inflammation, hypertension, diabetes mellitus, skeletal muscle dysfunction, and osteoporosis. Obstruction and limitations of airflow in patients reduces oxygen supply, produces a decline in aerobic capacity and increases the generation of reactive oxygen species that cause oxidation, chronic inflammation, and destruction of type I muscle fibers, muscle wasting, loss of muscle strength and fatigue. In extreme cases, this can cause sarcopenia and cachexia. Vitamin D3 supplementation in such individuals ameliorates inflammation, improves lung function, promotes mitochondrial repair, hence improves muscle regeneration and rebuild over time [55]. A meta-analysis examining individuals with vitamin D3 deficiency (25(OH)D3 ≤ 25 nmol/l) showed that supplementation with different vitamin D3 doses (from 4,000 to 60,000 IU per week) significantly improved upper and lower muscular body strength. This meta-analysis examined 7 RCTs including 310 adults, 67% of whom were females, ages ranging 21.5–31.5 years [56•]. The regenerative mechanisms rely on vitamin D3’s ability to enhance mitochondrial function, increase satellite muscle stem cell recruitment, and promote myogenic repair [57]. In long COVID-19, such musculoskeletal complaints exist and are more prevalent in young female adults. This same argument for using vitamin D3 supplementation in improving mitochondrial function, myogenic repair, enhancing muscular body strength, ameliorating inflammation or reperfusion injury muscle dysfunction, and hence treating individuals with long COVID-19-associated muscle fatigue should be examined in future RCTs.
Vitamin D3 Reduces Risk in COVID-19 Individuals Who Have Metabolic Comorbidities
Vitamin D3 promotes cardiometabolic health benefits. Vitamin D3 is important in blood pressure regulation, insulin sensitivity, downregulation of inflammatory responses, and reduction of risk for cardiometabolic diseases [28•]. COVID-19 complications include blood pressure abnormalities, new-onset diabetes mellitus and COVID-19-associated β-cell dysfunction [14], abnormal coagulopathy, and hyperinflammation associated with an inflammatory ‘cytokine storm’ [11•]. Endothelial damage either by hyperinflammation, oxidative stress, or metabolic irregularities e.g., hyperleptinemia, hyperinsulinemia or hyperglycemia exacerbate this risk. Many of the at-risk COVID-19 affected individuals suffer from comorbidities cardiovascular diseases, diabetes mellitus, hypertension, and obesity [12, 20]. Knowing the beneficial effects of vitamin D3 in attenuating cardiometabolic risk factors presents Vitamin D3 as a plausible biological agent to mitigate COVID-19 disease burden.
In a study of 43 individuals infected with SARS-CoV-2, Tan et al. examined the benefits of administering a combination therapy of vitamin D3, magnesium and vitamin B12 on severe outcome progression in older individuals above the age of 50 most of whom suffered from diabetes mellitus or hypertension. The SARS-CoV-2 infected individuals were given a combination treatment made up of 1,000 IU vitamin D3, 150 mg magnesium oxide, and 500 mcg vitamin B12 daily for up to 14 days compared to controls. Individuals given the combination treatment had fewer requirements for oxygen therapy during hospitalization in ICU or on ward (17.6% vs. 61.5%, respectively, p = 0.006). Additionally, combination treatment was associated with a lower likelihood for intensive care support. In multivariate analysis after adjusting for age and hypertension risk factors, participants given the combination treatment still had a lower need for oxygen therapy and intensive care support OR = 0.20 [95% CI:0.04 -0.93; p = 0.04] compared to controls. This suggested that vitamin D3 in combination treatment can have a beneficial and protective survival effects on older at-risk individuals with COVID-19 who had pre-existing comorbidities. The combination treatment may have metabolic and immunomodulatory effects, with vitamin D3 and magnesium exerting anti-inflammatory, anti-hypertensive, anti-thrombotic and broncho-dilatory effects, while vitamin D3 and vitamin B12 supporting gut microbiome development and priming innate and adaptive immune responses. While the study’s sample size is small it does incentivize investigating such combination nutrient treatments, including vitamin D3 supplementation, in at-risk individuals with comorbidities [58].
Conclusion
Maintaining vitamin D3 sufficiency prior to infection appears to be important in reducing risk and severity of COVID-19 in individuals of all ages. Furthermore, given vitamin D3’s well-known protective and regenerative physiological effects in multiple organ systems, administration of vitamin D3 to individuals infected with SARS-CoV-2 may promote faster recovery times and improved survival. Specific vitamin D3-induced mechanisms of action in individuals suffering from short or long COVID-19 need to be clearly elucidated, and supplementation studies consolidated. Nonetheless, cumulative evidence increasingly supports a potential role for the use of vitamin D3 in mitigating acute and long COVID-19 symptoms and disease burden, and in repairing disease-associated organ damage. No side effects have been reported following higher vitamin D3 intake as seen in epidemiological studies on individuals affected by COVID-19. Apropos, vitamin D3 supplementation, study design, and dosing regimens need to be revised to include higher doses of vitamin D3 in future studies, compared to current practices [46, 59•]. This is especially pertinent in at-risk subgroups, such as the elderly and individuals with obesity, who may benefit from higher-dose supplementation for various physiological reasons [60, 61]. Vitamin D3's potential as a cost-effective candidate in the management and mitigation of COVID-19 disease burden warrants further investigation given vitamin D3's multipotent diverse mechanistic actions in maintaining health and preventing disease.
Data Availability
The papers and data supporting this review are all available as publications in public sources such as PubMed and other search engines. There is no unique empirical data generated for this review to be shared or accessed.
Compliance with Ethical Standards
Conflicts of Interests or Disclosures
None to declare.
Human and Animal Rights and Informed Consent
This article does not contain any studies with human or animal subjects performed by any of the authors.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Papers of particular interest, published recently, have been highlighted as: • Of importance
1. WHO. WHO Coronavirus dashboard. https://covid19.who.int/. Accessed 2 Apr 2023.
2. Yeyati EL Filippini F The Independent Panel of the World Health Organization Social and economic impact of COVID-19 Brookings global working paper #158 2021 Brookings Institution
3. The World Bank Subdued global economic recovery Global economic prospects 2021 Washington, DC The World Bank
4. Robinson LA COVID-19 and uncertainties in the value per statistical life 2020 The Regulatory Review
5. Ye ZW Yuan S Yuen KS Fung SY Chan CP Jin DY Zoonotic origins of human coronaviruses Int J Biol Sci 2020 16 10 1686 1697 10.7150/ijbs.45472 32226286
6. Hu B Guo H Zhou P Shi ZL Characteristics of SARS-CoV-2 and COVID-19 Nat Rev Microbiol 2021 19 3 141 154 10.1038/s41579-020-00459-7 33024307
7. Chen B Tian EK He B Tian L Han R Wang S Overview of lethal human coronaviruses Signal Transduct Target Ther 2020 5 1 89 10.1038/s41392-020-0190-2 32533062
8. Rahimi A Mirzazadeh A Tavakolpour S Genetics and genomics of SARS-CoV-2: A review of the literature with the special focus on genetic diversity and SARS-CoV-2 genome detection Genomics 2021 113 1 Pt 2 1221 1232 10.1016/j.ygeno.2020.09.059 33007398
9. Hoffmann M Kleine-Weber H Schroeder S Kruger N Herrler T Erichsen S SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor Cell 2020 181 2 271 280 10.1016/j.cell.2020.02.052 32142651
10. WHO Coronavirus disease (COVID-19) weekly epidemiological update and weekly operational update 2022 World Health Organization
11. • Huang C, Wang Y, Li X, Ren L, Zhao J, Hu Y, et al. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 2020;395(10223):497–506. 10.1016/S0140-6736(20)30183-5. This is one the first seminal descriptions of clinical features of patients with acute COVID-19 emerging from China at the beginning of the pandemic.
12. Silva Andrade B Siqueira S de Assis Soares WR de Souza Rangel F Santos NO Dos Santos Freitas A Long-COVID and Post-COVID health complications: an up-to-date review on clinical conditions and their possible molecular mechanisms Viruses 2021 13 4 700 10.3390/v13040700 33919537
13. Al-Olama M Rashid A Garozzo D COVID-19-associated meningoencephalitis complicated with intracranial hemorrhage: a case report Acta Neurochir 2020 162 7 1495 1499 10.1007/s00701-020-04402-w 32430637
14. Sathish T Cao Y Kapoor N Newly diagnosed diabetes in COVID-19 patients Prim Care Diabetes 2021 15 1 194 10.1016/j.pcd.2020.08.014 32900656
15. Sierpina VS Become aware of the short and long-term mental health effects of COVID Explore 2021 17 4 281 10.1016/j.explore.2021.04.008 34127397
16. Gemelli Against C-P-ACSG Post-COVID-19 global health strategies: the need for an interdisciplinary approach Aging Clin Exp Res 2020 32 8 1613 1620 10.1007/s40520-020-01616-x 32529595
17. • Lopez-Leon S, Wegman-Ostrosky T, Perelman C, Sepulveda R, Rebolledo PA, Cuapio A, et al. More than 50 long-term effects of COVID-19: a systematic review and meta-analysis. Sci Rep. 2021;11(1):16144. 10.1038/s41598-021-95565-8. This systematic review details more than 50 long COVID-19 complications reported.
18. Carfi A Bernabei R Landi F Gemelli Against C-P-ACSG Persistent symptoms in patients after acute COVID-19 Jama 2020 324 6 603 605 10.1001/jama.2020.12603 32644129
19. Sivan M Halpin S Hollingworth L Snook N Hickman K Clifton IJ Development of an integrated rehabilitation pathway for individuals recovering from COVID-19 in the community J Rehab Med 2020 52 8 jrm00089 10.2340/16501977-2727
20. Garg MK Gopalakrishnan M Yadav P Misra S Endocrine involvement in COVID-19: mechanisms, clinical features, and implications for care Indian J Endocrinol Metab 2020 24 5 381 386 10.4103/ijem.IJEM_440_20 33489841
21. • Thompson EJ, Williams DM, Walker AJ, Mitchell RE, Niedzwiedz CL, Yang TC, et al. Long COVID burden and risk factors in 10 UK longitudinal studies and electronic health records. Nat Commun. 2022;13(1):3528. 10.1038/s41467-022-30836-0. This paper reviews long COVID-19 disease burden and risk factors available from 10 UK longtitudinal studies.
22. Li MY Li L Zhang Y Wang XS Expression of the SARS-CoV-2 cell receptor gene ACE2 in a wide variety of human tissues Infect Dis Poverty 2020 9 1 45 10.1186/s40249-020-00662-x 32345362
23. Baggerly CA Cuomo RE French CB Garland CF Gorham ED Grant WB Sunlight and Vitamin D: necessary for public health J Am Coll Nutr 2015 34 4 359 365 10.1080/07315724.2015.1039866 26098394
24. • Dominguez LJ, Farruggia M, Veronese N, Barbagallo M. Vitamin D sources, metabolism, and deficiency: available compounds and guidelines for its treatment. Metabolites. 2021;11(4). 10.3390/metabo11040255. This represents a current critical review of vitamin D 3 sources, metabolism and deficiency.
25. • Heaney RP, Holick MF. Why the IOM recommendations for vitamin D are deficient. J Bone Miner Res Off J Am Soc Bone Miner Res. 2011;26(3):455–7. 10.1002/jbmr.328. While not recent, this is one of the key papers which initiated the debate of why vitamin D 3 IOM levels are majorly underestimated and need revision.
26. Papadimitriou DT The big Vitamin D mistake J Prev Med Public Health 2017 50 4 278 281 10.3961/jpmph.16.111 28768407
27. LeBlanc ES Zakher B Daeges M Pappas M Chou R Screening for vitamin D deficiency: a systematic review for the U.S. Preventive Services Task Force Ann Intern Med 2015 162 2 109 122 10.7326/M14-1659 25419719
28. • Moukayed M, Grant WB. Linking the metabolic syndrome and obesity with vitamin D status: risks and opportunities for improving cardiometabolic health and well-being. Diabetes Metab Syndr Obes Targets Ther. 2019;12:1437–47. 10.2147/DMSO.S176933. This paper reviews the role of vitamin D 3 in mitigating cardiometabolic risk factors.
29. L Bishop E Ismailova A Dimeloe S Hewison M White JH Vitamin D and immune regulation: antibacterial, antiviral, anti-inflammatory JBMR Plus 2021 5 1 e10405 10.1002/jbm4.10405 32904944
30. Gilani SJ Bin-Jumah MN Nadeem MS Kazmi I Vitamin D attenuates COVID-19 complications via modulation of proinflammatory cytokines, antiviral proteins, and autophagy Expert Rev Anti Infect Ther 2022 20 2 231 241 10.1080/14787210.2021.1941871 34112047
31. Ahmed F A network-based analysis reveals the mechanism underlying vitamin D in suppressing cytokine storm and virus in SARS-CoV-2 infection Front Immunol 2020 11 590459 10.3389/fimmu.2020.590459 33362771
32. • Peng MY, Liu WC, Zheng JQ, Lu CL, Hou YC, Zheng CM, et al. Immunological aspects of SARS-CoV-2 infection and the putative beneficial role of Vitamin-D. Int J Mol Sci. 2021;22(10). 10.3390/ijms22105251. This review highlights the important immunomodulatory role of vitamin D 3 in SARS-CoV-2 affected individuals.
33. Lee C Controversial effects of Vitamin D and related genes on viral infections, pathogenesis, and treatment outcomes Nutrients 2020 12 4 962 10.3390/nu12040962 32235600
34. Crafa A Cannarella R Condorelli RA Mongioi LM Barbagallo F Aversa A Influence of 25-hydroxy-cholecalciferol levels on SARS-CoV-2 infection and COVID-19 severity: a systematic review and meta-analysis EClinicalMedicine 2021 37 100967 10.1016/j.eclinm.2021.100967 34179737
35. Israel A Cicurel A Feldhamer I Stern F Dror Y Giveon SM Vitamin D deficiency is associated with higher risks for SARS-CoV-2 infection and COVID-19 severity: a retrospective case-control study Intern Emerg Med 2022 17 4 1053 1063 10.1007/s11739-021-02902-w 35000118
36. Kaufman HW Niles JK Kroll MH Bi C Holick MF SARS-CoV-2 positivity rates associated with circulating 25-hydroxyvitamin D levels PloS one 2020 15 9 e0239252 10.1371/journal.pone.0239252 32941512
37. Sulli A Gotelli E Casabella A Paolino S Pizzorni C Alessandri E Vitamin D and lung outcomes in elderly COVID-19 patients Nutrients 2021 13 3 717 10.3390/nu13030717 33668240
38. Annweiler C Beaudenon M Gautier J Simon R Dubee V Gonsard J Covid-19 and high-dose VITamin D supplementation TRIAL in high-risk older patients (COVIT-TRIAL): study protocol for a randomized controlled trial Trials 2020 21 1 1031 10.1186/s13063-020-04928-5 33371905
39. Annweiler G Corvaisier M Gautier J Dubee V Legrand E Sacco G Vitamin D supplementation associated to better survival in hospitalized frail elderly COVID-19 patients: the GERIA-COVID quasi-experimental study Nutrients 2020 12 11 3377 10.3390/nu12113377 33147894
40. Entrenas Castillo M Entrenas Costa LM Vaquero Barrios JM Alcala Diaz JF Lopez Miranda J Bouillon R Effect of calcifediol treatment and best available therapy versus best available therapy on intensive care unit admission and mortality among patients hospitalized for COVID-19: a pilot randomized clinical study J Steroid Biochem Mol Biol 2020 203 105751 10.1016/j.jsbmb.2020.105751 32871238
41. Abroug H Maatouk A Bennasrallah C Dhouib W Ben Fredj M Zemni I Effect of vitamin D supplementation versus placebo on recovery delay among COVID-19 Tunisian patients: a randomized-controlled clinical trial Trials 2023 24 1 123 10.1186/s13063-023-07114-5 36803273
42. Murai IH Fernandes AL Sales LP Pinto AJ Goessler KF Duran CSC Effect of a single high dose of Vitamin D3 on hospital length of stay in patients with moderate to severe COVID-19: a randomized clinical trial JAMA 2021 325 11 1053 1060 10.1001/jama.2020.26848 33595634
43. Jevalikar G Mithal A Singh A Sharma R Farooqui KJ Mahendru S Lack of association of baseline 25-hydroxyvitamin D levels with disease severity and mortality in Indian patients hospitalized for COVID-19 Sci Rep 2021 11 1 6258 10.1038/s41598-021-85809-y 33737631
44. • Ghelani D, Alesi S, Mousa A. Vitamin D and COVID-19: an overview of recent evidence. Int J Mol Sci. 2021;22(19). 10.3390/ijms221910559. Excellent current review evaluating the epidemiological studies regarding the association between vitamin D 3 and COVID-19.
45. Alcala-Diaz JF Limia-Perez L Gomez-Huelgas R Martin-Escalante MD Cortes-Rodriguez B Zambrana-Garcia JL Calcifediol treatment and hospital mortality due to COVID-19: a cohort study Nutrients. 2021 13 6 1760 10.3390/nu13061760 34064175
46. Annweiler C Beaudenon M Gautier J Gonsard J Boucher S Chapelet G High-dose versus standard-dose vitamin D supplementation in older adults with COVID-19 (COVIT-TRIAL): a multicenter, open-label, randomized controlled superiority trial PLoS Med 2022 19 5 e1003999 10.1371/journal.pmed.1003999 35639792
47. Karonova TL Chernikova AT Golovatyuk KA Bykova ES Grant WB Kalinina OV Vitamin D intake may reduce SARS-CoV-2 infection morbidity in health care workers Nutrients 2022 14 3 505 10.3390/nu14030505 35276863
48. • Shah K, Varna VP, Sharma U, Mavalankar D. Does vitamin D supplementation reduce COVID-19 severity?: a systematic review. QJM: monthly journal of the Association of Physicians. 2022;115(10):665–72. 10.1093/qjmed/hcac040. Excellent current review evaluating the epidemiological studies regarding the association between vitamin D 3 and COVID-19.
49. Wenzel J Schwaninger M How COVID-19 affects microvessels in the brain Brain J Neurol 2022 145 7 2242 2244 10.1093/brain/awac211
50. Lee MH Perl DP Nair G Li W Maric D Murray H Microvascular Injury in the Brains of Patients with Covid-19 N Engl J Med 2021 384 5 481 483 10.1056/NEJMc2033369 33378608
51. • Marek K, Cichon N, Saluk-Bijak J, Bijak M, Miller E. The role of vitamin D in stroke prevention and the effects of its supplementation for post-stroke rehabilitation: a narrative review. Nutrients. 2022;14(13). 10.3390/nu14132761. This review beautifully presents current findings into the capability of vitamin D 3 in in brain repair and function even after injury. Vitamin D 3 therefore can have a long term therapeutic potential in long COVID-19 brain related dysfunctions.
52. Aspell N Lawlor B O'Sullivan M Is there a role for vitamin D in supporting cognitive function as we age? Proc Nutr Soc 2018 77 2 124 134 10.1017/S0029665117004153 29233204
53. Zhao W Zhu DM Li S Cui S Jiang P Wang R The reduction of vitamin D in females with major depressive disorder is associated with worse cognition mediated by abnormal brain functional connectivity Prog Neuropsychopharmacol Biol Psychiatry 2022 118 110577 10.1016/j.pnpbp.2022.110577 35605842
54. DeLuca GC Kimball SM Kolasinski J Ramagopalan SV Ebers GC Review: the role of vitamin D in nervous system health and disease Neuropathol Appl Neurobiol 2013 39 5 458 484 10.1111/nan.12020 23336971
55. Russo C Valle MS Casabona A Spicuzza L Sambataro G Malaguarnera L Vitamin D impacts on skeletal muscle dysfunction in patients with COPD promoting mitochondrial health Biomedicines 2022 10 4 898 10.3390/biomedicines10040898 35453648
56. • Tomlinson PB, Joseph C, Angioi M. Effects of vitamin D supplementation on upper and lower body muscle strength levels in healthy individuals. A systematic review with meta-analysis. J Sci Med Sport. 2015;18(5):575–80. 10.1016/j.jsams.2014.07.022. A systematic review demonstrating the important role of vitamin D 3 in muscle regeneration. Muscle repair and regeneration is important in tackling myalgia, fatigue and other related complications of long COVID-19.
57. Latham CM Brightwell CR Keeble AR Munson BD Thomas NT Zagzoog AM Vitamin D promotes skeletal muscle regeneration and mitochondrial health Front Physiol 2021 12 660498 10.3389/fphys.2021.660498 33935807
58. Tan CW Ho LP Kalimuddin S Cherng BPZ Teh YE Thien SY Cohort study to evaluate the effect of vitamin D, magnesium, and vitamin B12 in combination on progression to severe outcomes in older patients with coronavirus (COVID-19) Nutrition 2020 79 111017 10.1016/j.nut.2020.111017 33039952
59. • McCullough PJ, Lehrer DS, Amend J. Daily oral dosing of vitamin D3 using 5000 TO 50,000 international units a day in long-term hospitalized patients: insights from a seven year experience. J Steroid Biochem Mol Biol. 2019;189:228–39. 10.1016/j.jsbmb.2018.12.010. This study clearly emphasizes that higher vitamin D 3 doses can be administered to improve health outcomes and reduce hospitalization in patients without concerns of putative vitamin D side-effects.
60. Wortsman J Matsuoka LY Chen TC Lu Z Holick MF Decreased bioavailability of vitamin D in obesity Am J Clin Nutr 2000 72 3 690 693 10.1093/ajcn/72.3.690 10966885
61. Grant WB Boucher BJ Al Anouti F Pilz S Comparing the evidence from observational studies and randomized controlled trials for nonskeletal health effects of vitamin D Nutrients 2022 14 18 3811 10.3390/nu14183811 36145186
|
PMC010xxxxxx/PMC10165022.txt |
==== Front
Journal of Retailing and Consumer Services
0969-6989
0969-6989
Elsevier Ltd.
S0969-6989(23)00158-3
10.1016/j.jretconser.2023.103411
103411
Article
Has the COVID-19 pandemic changed the influence of word-of-mouth on purchasing decisions?
Byun Kate Jeonghee a
Park Jimi b∗
Yoo Shijin a
Cho Minhee c
a School of Business, Korea University, Anam-dong 5-1, Seongbuk-ku, Seoul, 02841, South Korea
b College of Business, Hawaii Pacific University, 900 Fort Street Mall, Suite PL 600, Honolulu, HI, 96813, USA
c PMI CO., LTD., 2F Seolin Bldg., 16, Gangnam-daero 91-gil, Seocho-gu, Seoul, 06530, South Korea
∗ Corresponding author.
8 5 2023
9 2023
8 5 2023
74 103411103411
29 11 2022
12 4 2023
30 4 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The COVID-19 outbreak has led to drastic changes in influential media for purchasing decisions. Which media have been used more frequently after COVID-19? Do these changes differ between product types? To provide insights into these questions, we analyzed 12,000 respondents between 2018 and 2021. The results show that the influence of word-of-mouth on purchase decisions decreased after COVID-19, while the influence of social media advertising increased. We also find that the influence is moderated by product type; for example, video advertising, a subset of social media advertising shows a higher influence for search goods than for experience goods.
Keywords
Media usage
Word-of-mouth
Social media advertising
Video advertising
COVID-19 pandemic
==== Body
pmc1 Introduction
The spread of the novel coronavirus disease 2019 (COVID-19) has had an unprecedented impact on the global economy in terms of the business environment and consumer lifestyle (Sheth, J., 2020; Szász et al., 2022). Following the pandemic, consumer behavior has shifted to a new normal (Roggeveen and Sethuraman, 2020). For instance, 37% of consumers from the United States consumers are expected to shop more online, and over 40% of them watched more video content using streaming services or video-related platforms (e.g., YouTube) after the COVID-19 outbreak (Statista, 2022). Naturally, the transition in shopping behavior may be related to changes in media usage. Consumers have shown higher engagement with digital media during lockdowns (Jiang and Stylos, 2021) and enhanced responsiveness to online content (Chang and Meyerhoefer, 2021). How consumers use digital media (e.g., music streaming applications) has also changed because of the increased hours spent working at home (Sim et al., 2021).
In the context of media usage, how the influence of word-of-mouth (WOM) is affected by the pandemic is worth exploring (Taylor, 2020) due to its high persuasiveness and exponential spread (Rosario et al., 2016). Particularly, the World Health Organization (WHO, 2021) has warned about the danger of fake news diffusion through WOM during the pandemic, suggesting that misinformation is spreading faster than the virus because of its speed and amplification characteristics. With the growing importance of quality information, 10.13039/100006785 Google and YouTube invested a $13.2 million grant in the International Fact-Checking Network (IFCN) across 65 countries in November 2022. Recent studies confirm this idea that fake news and misinformation can be spread, propagated, and amplified largely in a short period time through WOM (Di Domenico et al., 2021; Kwon et al., 2022; Wang et al., 2022). Such WOM may incur erroneous or biased judgment. For example, misinformation can be widely retweeted if Twitter users do not critically evaluate it before deciding to share it (Di Domenico et al., 2021). While existing studies focus on how false information can be diffused by WOM and its impact on individuals' cognitive abilities and psychological factors during the pandemic, consumers’ subsequent media usage pattern regarding WOM is under investigated. Building on the efforts to address misinformation, we explain how the COVID-19 pandemic has affected WOM based on two perspectives: demand- and supply-side perspectives.
On the demand side, we posit that the three aspects of WOM – quality, quantity, and credibility (Lkhaasuren and Nam, 2018) – can be more widely dispersed in times of crisis. This is because when consumers face uncertainty, they might wait and observe (Berns et al., 2007) or prefer a divergent (broader) thinking style to a convergent (narrower) style (Ng et al., 2021) due to difficulty in identifying fake news or misinformation. The greater variance of WOM harms the readability of messages and knowledge transfer, making consumers switch from WOM communication channels to a relatively more reliable communication channel such as advertisement. Therefore, when uncertainty is high, consumers tend to search for more credible messages from official sources instead of private ones to avoid risk.
On the supply side, firms may lag in converting a potential customer into a paying customer during the crisis. Firms have incentives to increase advertising to gain a higher share of voice during the pandemic, which helps them differentiate themselves from competitors (Van Heerde et al., 2013). In the systematic progression of digital transformation, social media advertising is recognized as an alternative medium where firms can leverage consumer behavior and insights to trigger personalized digital communications. In addition to attracting customers, firms may attempt to communicate through the intended channels to capitalize on social media platforms by using more personalized social media advertising during the pandemic. Based on various forms and contents, social media advertising can more effectively accumulate total views and positively persuade consumers to purchase products. Therefore, communicating information via a lower variance channel is vital for managers to rebut misinformation and boost the customer conversion. Accordingly, we argue that the COVID-19 pandemic has decreased the usage of WOM in consumers’ purchase decisions. Instead, the influence of other firm-generated media, such as social media advertising, on purchasing decisions has increased during the COVID-19 pandemic.
The changes in media usage may be different across product types because product types significantly affect consumers' collection of product-related information. Specifically, consumers want to (and tend to) have full, reliable, and official information about search goods before purchasing while focusing more on other users’ reviews for experience goods. Therefore, we parse the differences in the influence of changes in media use across product types during the COVID-19 pandemic. For example, our study shows that video-based social media advertising for search goods has a greater increase in influence on purchasing decisions than experience goods after the COVID-19 pandemic.
Given the substantial changes in media usage and the resulting academic interest, examining consumers’ media usage and its influence on purchase decisions during the pandemic would help to better understand the changes. Specifically, when consumers face uncertainty due to the significant external shock, it motivates the need to consider uncertainty when managing communication channels. Moreover, firms may capitalize on the media platforms by reallocating media channels when nothing is normal.
Accordingly, we pose the following three research questions.• Has the influence of WOM on purchase decisions changed during the COVID-19 pandemic? More specifically, would consumers be less likely to rely on WOM because of the risk and uncertainty during the pandemic?
• Are there any other types of media (e.g., advertising on social media) that consumers have consumed more of during the COVID 19 pandemic?
• Are these changes different across product types, such as search and experience goods?
Our main focus, the influence of each medium on a consumer's purchase decision, is measured by the media usage index. Specifically, we compare the difference in media usage index between pre- and post-pandemic periods using survey data from 2018, 2019 and 2021 for 19 product categories.
This study offers several important findings by answering these research questions based on rich survey data from Korea. First, the COVID-19 pandemic has affected consumers' media usage. Specifically, consumers are found to use WOM less as a communication channel during the pandemic. To the best of our knowledge, this study is the first to empirically reveal that an irreversible external shock, such as the COVID-19 pandemic, reduces the influence of WOM on purchase decisions. Second, we comprehensively present the influence of WOM by comparing it with social media advertising to elucidate changes of relative importance. Our study complements Hoekstra and Leeflang (2020), suggesting that firms may have shifted to “contact advertising” (e.g., vlogs) to specifically target their consumers during the pandemic. Third, we partly find the moderating role of product type on the changes in media usage during the COVID-19 pandemic. It would be beneficial for managers to understand the changes in consumers' media usage because less-used media would undermine the firm's marketing effectiveness in terms of consumer engagement and patronage behavior. Overall, our study provides academics and practitioners with a better understanding of consumers' information search and media usage in the post-COVID era, which can improve firms' media allocation decisions.
2 Related literature and research hypotheses
2.1 WOM under uncertainty
Consumers consider various factors when buying a product. Among these, WOM has long been regarded as an important factor. For example, consumers read reviews or recommendations about a product online to avoid wasting time in the store and obtain products at a lower price through comparison shopping (Harahap et al., 2018). The following three characteristics explain the traits of WOM: quality, quantity, and credibility. WOM quality denotes readability and persuasion power (Cheung et al., 2008; Lkhaasuren and Nam, 2018; Park and Kim, 2008; Tajuddin et al., 2018; Teng et al., 2014). Accordingly, complete, updated, and relevant details are regarded as acceptable information that increases purchase interest (Tien et al., 2019; Zhang et al., 2014). WOM credibility refers to the trustworthiness of reviews or comments (Aini and Zuliestiana, 2019; Bataineh et al., 2015), which may lead to purchase decisions (Cheung et al., 2009; Fang et al., 2014; Lkhaasuren and Nam, 2018; Tien et al., 2019). Moreover, consumers tend to strengthen their confidence and reduce perceived risk based on WOM quantity. Contrary to the argument that WOM has become increasingly important in recent decades (Chu and Kim, 2018), we question whether WOM has held its influence on media usage during and after the COVID-19 pandemic. It may have become less reliable given the possibility and harshness of hazards driven by the pandemic. We posit that all three dimensions of WOM – quality, quantity, and credibility – become more widely dispersed in turbulent environments.
We describe how the COVID-19 pandemic has affected WOM based on two perspectives: demand- and supply-side perspectives. During the pandemic, the amount of information from less trustworthy sources increased and spread even faster and further (Di Domenico et al., 2021). Recent studies argue that false information spreads rapidly through WOM because of its enormous influence (Di Domenico et al., 2021; Kwon et al., 2022; Wang et al., 2022). Therefore, the variance of WOM has increased through new modes of transmission of false information. Exposure to WOM, which includes both fake and authentic news, increases society's risk perception (Ng et al., 2021). Being overloaded with this deviant information makes consumers more aware of being wary of fake news. Since it is difficult to distinguish fake news, consumers are not able to make proper identification, leading them to try to reduce the anxiety with clearer and more credible sources instead of WOM. Consumers may use less WOM to avoid erroneous or biased judgments derived from widely shared misinformation which might have multiple layers of corrections or rebuttals for fake news. Furthermore, consumer receptiveness during a contraction displays characteristics that differ from those associated with a boom. For example, when the economy weakens, consumers become more price sensitive because of lower consumer confidence (Gordon et al., 2013). Consumers also become more receptive to new products/services, finding more reliable communication channels during economic downturns (Steenkamp and Fang, 2011). In summary, WOM may shape the social perception of risk as an “amplification station” (Kasperson et al., 1988), leading to larger variances in WOM. Thus, the demand-side perspective suggests that the COVID-19 pandemic has led to a greater variance in WOM.
Firms have been rapidly reallocating their advertising budgets from traditional advertising to social media advertising (Jobs and Gilfoil, 2014). Although these advancements have been in the works for many years, the pandemic has quickly forced years of change in the way both consumers and businesses operate. It has escalated the growth of social media advertising because it helps to obtain both direct value, such as customer purchases or membership renewals, and indirect value, such as the attraction and engagement of consumers who have an impact on the positive evaluation of products (Voorveld et al., 2018). According to a survey by Deloitte, firms reallocated 46% of their marketing budgets toward social media and mobile marketing, which amounts to twice as much as before the pandemic (Deloitte, 2020). Additionally, The CMO Survey reported that the share of social media marketing spending jumped by 74.4% from 13.3% in February 2020 to 23.2% in June 2020. Furthermore, the contribution of social media toward firm performance increased by more than 24% after Feb 2020 (Balis, 2021). This change is important because, despite steady investments by firms, social media influence has not changed much since 2016 (Moorman and McCarthy, 2021). Given that firms have incentives to increase social media advertising, consumers tend to reduce the use of WOM to avoid being influenced by misinformation spread via WOM. Accordingly, we propose the following hypothesis.Hypothesis 1 Consumer WOM usage as a communication channel has decreased after the COVID-19 pandemic.
2.2 COVID-19 and social media advertising
During the COVID-19 pandemic, increased time spent at home changed consumers’ media usage habits (Sim et al., 2021). Social media advertising has advantages over other media types because consumers can consume it while following their own schedules and needs (Schweidel and Moe, 2016). Consumers spend more time on social media, and this phenomenon is expected to remain stable (Statista, 2022). To explain the demand-side perspective, we focus on the following two traits of social media advertising: virality and persuasiveness (Colicev et al., 2018). Consumers may go viral through social media content when they are engaged, and these activities enhance the evaluation of products or brands (Voorveld et al., 2018). While high variant (mis)information in WOM leads individuals not to seek further information (Kim et al., 2020; Kwon et al., 2022), consumers tend to prefer more reliable sources such as firm-generated information delivered through social media advertising.
On the supply side, marketing managers try to alleviate professional concerns that can be raised due to a lack of available funds by turning to more effective communication channels, such as social media advertising. Using social media, firms can more effectively spread product information, promptly respond to customers, preempt negative feedback, and implement targeted advertising with significantly lower costs than traditional media (Spotts et al., 2022). Social media also enables persuasiveness because of its close distance to consumers and encourages viral content. This can be explained by recent changes where firms have shifted to “contact advertising,” such as blogs, vlogs, or social media (Hoekstra and Leeflang, 2020). Given that firms have used social media to engage with their customers following the global crisis (Kirtis and Karahan, 2011), firms can enjoy the merits of advertising investment in a contraction era with greater sensitivity (Steenkamp and Fang, 2011). Thus, social media advertising during the COVID-19 pandemic would have been beneficial for firms to gain competitive advantages. Therefore, during crises, consumers want to engage in more trustworthy, fact-based, and easier-to-understand communication channels to facilitate better searches at lower costs during the pandemic period. Therefore, consumers turn to social media advertising while firms simultaneously incentivize this trend to reinforce social media advertising to boost marketing efficacy. Therefore, we propose the following hypothesis.Hypothesis 2 The use of social media advertising as a communication channel has increased after the COVID-19 pandemic.
2.3 Search goods and experience goods
Consumers tend to obtain sufficient information they need before purchasing search goods, but it is harder to evaluate the quality of experience goods until they have been purchased and used (Nelson, 1970). Therefore, searching for information is more costly for experience goods. Specifically, consumers examine search goods in advance using physical interaction, online search, and prior experience because search efforts in the pre-purchasing stage help evaluate the quality of search goods (Nelson, 1970; Rossolov et al., 2021). Meanwhile, consumers can delineate the evaluation of experience goods after they purchase and experience the products. Hence, how much information can be obtained before purchasing is the criterion for differentiating between search and experience goods.
Prior research has been inconclusive regarding the usage of WOM across product types. Some scholars have demonstrated that consumers are more influenced by WOM when they decide to buy experience goods (Kim and Hanssens, 2017). Other scholars argue that digital marketing technology has blurred the distinction between experience and search goods because consumers can easily obtain the necessary information regardless of the product type (Basu, 2018). We argue that consumers looking for search goods under situational uncertainty tend to be significantly less influenced by WOM because they may obtain information at higher costs while acquiring WOM with large variances. Given that consumers seek further validation with more referenced data in the process of adopting information during a crisis, consumers considering search goods tend to reduce their usage of WOM because of overloading and divergent information. Moreover, following the pandemic, the advantages of WOM, such as credibility, have declined from the consumer's point of view. The potential loss of time spent evaluating reviews or testimonials from other consumers may outweigh the value of WOM searches. Conversely, firms enhance consumer engagement and product evaluation using social media advertising because they can enjoy differentiation by sustaining firm-generated social media advertising during the downturn. Therefore, we propose a moderating effect of product type on the changes in media usage during the COVID-19 pandemic, as described in the following hypotheses.Hypothesis 3 The change in media usage on purchasing decisions after the COVID-19 pandemic differs between search goods and experience goods.
H3a After the COVID-19 pandemic, the usage of WOM for purchasing decisions related to search goods has decreased more than that for experience goods.
H3b After the COVID-19 pandemic, the usage of social media advertising in purchasing decisions related to search goods has increased more than that related to experience goods.
3 Empirical analysis
3.1 Data
To examine the influence of the changes in media usage on consumers’ purchase decisions during the COVID-19 pandemic, we used a survey dataset from Partners Make Innovation (PMI), the largest online/mobile panel operation firm in Korea. The firm conducted surveys in 2018, 2019, and 2021 for 63 product categories, with 38,870 respondents. Since the WHO declared COVID-19 a pandemic on January 30, 2020, we used survey participants from 2018 to 2019 as the pre-COVID-19 control group and those from 2021 as the post-COVID-19 treatment group. Each respondent was eligible to answer the questions for a particular product category if they had at least one purchase incident in that category within the previous month. We filtered the data to enable comparison among the responses from the three years in 19 product categories with 18,320 respondents. Therefore, in our sample, each product category has more than 300 respondents per year.
We used propensity score matching (PSM) to construct sample respondents and compare consumer behavior before (2018 and 2019) and after (2021) the COVID-19 pandemic. PSM adjusts the differences between the control and treatment groups by decreasing the bias about the treatment effect (Rosenbaum et al., 1985) and eliminates the self-selection bias and endogeneity concerns to establish the causal effect. When the propensity scores of the treatment and control groups are sufficiently close, the treatment can be regarded as random because the biases in the comparisons are mostly eliminated.
Given that we plan to compare media usage pre- and post-COVID-19, we model each respondent's propensity score (Y) using a logistic regression model for the following individual-specific explanatory variables: personal innovativeness, gender, age, and online shopping frequency. Specifically, we construct the following equation for respondent i:(1) Yi = β0 + β1INNi + β2GENi + β3AGEi + β4OSFi + εi
Where.INN = a respondent's personal innovativeness with a value ranging from 1 to 5
GEN = a respondent's gender with a value of 0 for male and 1 for female
AGE = a respondent's age with a value ranging from 14 to 69
OSF = a respondent's online shopping frequency with a value ranging from 1 to 8
First, because consumers with higher personal innovativeness are likely to be interested in newly launched products and services, this tendency will be related to media usage. Second, consumers in similar age groups tend to share a similar attitude toward media usage through homophily (Lou and Yuan, 2019). This is because similar aged groups tend to interact more with each other compared to other age groups. For instance, older consumers tend to move away from online media (Lian and Yen, 2014). Third, gender is regarded as an important factor that affects online shopping behavior. For example, women place less value on the utility of online shopping than men (Hasan, 2010). Thus, gender needs to be considered as a covariate in our matching procedure. Fourth, online shopping frequency is also related to media usage because high frequency implies ease of use of technology, which is significantly related to media usage (Tarigan et al., 2022).
Based on the parameter estimates of these covariates, we obtained the propensity score for each individual in the treatment group. Next, we matched the control group customers who resembled the treatment group customers based on propensity score similarity, using the 1:1 nearest-neighbor matching technique without replacement and a caliper of .015 (Rosenbaum et al., 1985). This method helped avoid bias when linking multiple and potentially dissimilar treatment and control groups of customers. The final matching sample contained 6000 respondents in each group. We confirm that the control and treatment groups are not significantly different in terms of propensity scores (t = −0.113, p = .859) and various individual characteristics. Table 1 presents the demographic information of the final sample, and Table 2 delineates the descriptive statistics of the variables used in PSM.Table 1 Comparison of the respondents characteristics.
Table 1Characteristics Frequency (Percentage)
Before COVID-19 After COVID-19
Gender
Male 2888 (48.13%) 2850 (47.50%)
Female 3112 (51.87%) 3150 (52.50%)
Age
Under 20 176 (2.93%) 160 (2.67%)
20-29 1612 (26.87%) 1700 (28.33%)
30-39 1922 (32.03%) 1920 (32.00%)
40-49 1989 (33.15%) 1920 (32.00%)
Above 50 301 (5.02%) 300 (5.00%)
Occupation
Students 809 (13.48%) 810 (13.50%)
General office job 2720 (45.33%) 2672 (44.53%)
Technology/Service 668 (11.13%) 665 (11.08%)
Profession 438 (7.30%) 455 (7.58%)
Self-employed 325 (5.42%) 297 (4.95%)
Teacher/Officer 179 (2.98%) 159 (2.65%)
Others 861 (14.35%) 942 (15.70%)
Marriage
Married 2938 (48.97%) 2616 (43.60%)
Single 3062 (51.03%) 3384 (56.40%)
Average Propensity Score .331 .331
t-test results t = −.065, p = .948
Table 2 The descriptive statistics of the variables used in propensity score matching.
Table 2Variables Before COVID-19 After COVID-19
Mean Standard deviation Mean Standard deviation
Personal innovativeness 3.15 0.91 3.16 0.96
Online shopping frequency 3.91 1.58 3.91 1.59
Gender 0.52 0.50 0.53 0.50
Age 35.99 10.17 35.89 10.43
Notes: 1. *p < .10, **p < .05, ***p < .01.2. Each group (before COVID-19 or after COVID-19) has 6000 observations. 3. The minimum and maximum value of the before COVID-19 and after COVID-19 have the same range.
3.2 Measurement
3.2.1 Media
To capture the media usage of a customer, we classified media into two categories based on studies by Spotts et al. (2022), Voorveld et al. (2018), and Lu et al. (2021), specifically WOM and social media advertising. Each category contains two detailed media subsets, presented in Table 3 . WOM includes traditional WOM that occurs in physical environments and electronic WOM (e-WOM), which is reviews or personal evaluations through digital media. Social media advertising includes video advertising (e.g., YouTube ads) and image/text advertising (e.g., Facebook ads).1 Each respondent chose the influential media that they thought was critical to their purchase decision among 27 to 49 media usage options, depending on the product categories. The respondents were asked about the following four types of media usage: WOM, social media advertising, retail-related media (e.g., in-store communication, retail advertising), and traditional media (e.g., TV, radio, and newspaper). Appendix B shows the exemplary media options (e.g., mobile communication services) presented in the survey.Table 3 Media categories.
Table 3Category Description
WOM
Traditional WOM Orally transmitted recommendations heard from acquaintances
e-WOM Review or personal evaluation of digital media
Social media advertising
Video advertising Online/mobile advertising offered by video type and live-streaming commerce (e.g., YouTube)
Image/text advertising Text or image type of social media advertising (e.g., Facebook, Twitter)
Retail-related media
Retail advertising Advertising or announcement of promotion in the retail stores owned or in alliance
In-store communication Event, giveaway, or discount announcement within the store, recommendations from a salesperson
Traditional media TV, radio, newspaper, magazine, out-of-home advertising
3.2.2 Influence of media usage on purchasing decisions
To measure the influence of each media on a consumer's purchasing decisions, we first counted the number of specific media types selected by respondents in each product category. Next, we calculated the proportion of the number of chosen media types (e.g., WOM or social media advertising) to the total number of media selections. For example, if a respondent chose two media types, such as hearing from acquaintances and searching for online reviews, from the WOM category out of ten total media usage options, then the influence of WOM on the respondent's purchase decision was calculated to be 0.2. Note that our survey questions asked “whether or not” instead of “how much” media usage influenced their purchase decision, which cannot capture the intensity of using a certain media type. However, given the assumption of unbiased samples based on PSM, the difference in media usage index between pre- and post-pandemic periods can indicate consumers' media usage changes.
3.2.3 Product types
Following Nelson (1970), we defined search and experience goods based on accessibility to full information prior to purchase decisions. We included nine product categories (e.g., credit cards, stock brokerage services, and mobile communication services) as search goods (coded as 0) and ten categories (e.g., beer and coffee) as experience goods (coded as 1). For example, consumers can easily find out the fee of the stock firms or service specifications of a credit card before opening an account (i.e., search goods); however, they cannot fully assess the quality of beer or coffee before experiencing it (i.e., experience goods). Table 4 shows the list of all product categories and the number of respondents in the final sample.Table 4 Product categories and the number of respondents.
Table 4Product types Product category Number of respondents
before COVID-19 after COVID-19
Search goods Credit card 309 300
Stock firms 335 300
Mobile communication service 418 420
Internet broadcasting service 292 300
Internet shopping mall 361 350
Corporate PR 365 300
Outdoor clothing 312 300
Mobile game 352 350
Bed (Mattress) 285 300
Total (9 categories) 3029 2920
Experience goods Water purifier 301 300
Beer 289 360
Beverage for health 302 300
Dietary supplement 262 300
Coffee 298 300
Air purifier as household appliance 285 300
Underwear for urinary incontinence 301 300
Digestive medicine 272 300
Automobile (a full-sized car) 335 320
Automobile (compact car) 326 300
Total (10 categories) 2971 3080
Total 6000 6000
3.2.4 Control variables
Similar to the PSM procedure, we considered the following four control variables known to influence consumers' purchase decisions through media usage: personal innovativeness, sex, age, and online shopping frequency. First, we controlled respondents’ personal innovativeness as consumers who are keen on technology and are more likely to have online media expertise (Kumar et al., 2016) or pursue innovativeness when it comes to media usage (Rauniar et al., 2014). Thus, personal innovativeness was controlled using the item, “I tend to use or purchase the newly launched brands, services, or products compared to other consumers,” and it was evaluated on a five-point Likert scale (1 = strongly disagree, 5 = strongly agree).2 Second, we controlled for consumer age as a continuous variable because similar age groups, especially young generations, tend to show homophily in media usage habits (Lou and Yuan, 2019). Third, gender was included as a control variable because gender differences may lead to distinctive media usage. For example, women and men show difference in purchase decisions while they do online shopping (González et al., 2021). We controlled for the gender using a dummy variable equivalent to 1 if the respondent was male and 0 otherwise. Fourth, online shopping frequency was included as a control variable because the media which consumers tend to use for shopping is related to the way consumers shop (Lee et al., 2018). We asked how often the respondent shops online and evaluated the response using an eight-point scale (1 = almost every day, 8 = less than any options shown above).
4 Results
4.1 Hypotheses testing
We compared how the influence of WOM on purchasing decisions changed during the COVID-19 pandemic by conducting an analysis of covariance (ANCOVA), as presented in Table 5 . Supporting H1, the influence of WOM on purchasing decisions after the COVID-19 pandemic decreased (F (1, 11994) = 14.630, p = .000, M before = 0.084, M after = 0.079), demonstrating that consumers tended to rely less on WOM to make purchase decisions after the pandemic. Next, we investigated whether social media advertising influenced purchasing decisions more after the COVID-19 pandemic. H2 was supported because social media advertising showed an upward trend following the pandemic (F (1, 11994) = 118.200, p = .000, M before = 0.065, M after = 0.080).Table 5 The changes of media usage in purchasing decisions.
Table 5Media Usage in Purchasing Decisions Statistical significance
Before COVID-19 After COVID-19 Difference
WOM .084 .079 −.005 F (1,11994) = 14.630, p = .000
Traditional WOM .045 .042 −.003 F (1,11994) = 8.030, p = .005
e-WOM .039 .037 −.002 F (1,11994) = 8.281, p = .004
Social media advertising .065 .080 .015 F (1,11994) = 118.200, p = .000
Text/image advertising .029 .026 −.003 F (1,11994) = 14.451, p = .000
Video advertising .036 .054 .018 F (1,11994) = 321.677, p = .000
To address the third research question, we examined whether the degree of change varied across product types. However, both changes in the influence of WOM and social media advertising were not significantly moderated by product types. Therefore, H3 was not supported. To further examine the influence of subcategories of WOM and social media advertising on purchasing decisions, we conducted additional analyses using a detailed classification of each media. As presented in Table 5, both traditional WOM and e-WOM showed reduced usage after COVID-19. However, two types of social media advertising showed different patterns. Although the influence of text/image advertising on purchasing decisions decreased, video advertising showed an upward trend after the COVID-19 pandemic. This antipodal direction indicates the insignificant moderating effect of product type on social media advertising. As shown in Table 6 , product type has a significant moderating effect through video advertising only. Specifically, video advertising for search goods (M after – M before = 0.019) shows a higher increase after the COVID-19 pandemic than experience goods (M after – M before = 0.016), and the difference is found to be significant (F (1, 11992) = 3.973, p = .046). However, this moderation effect cannot be found in the case of text/image advertising (F (1, 11992) = 0.002, p = .961). This result is partly attributed to the characteristics of video-based advertising. Unlike other formats, such as static images or text-based advertising, video-based social media advertising is highly persuasive, delivers joy or surprise, and is likely to catch people's eye more with a lesser chance of being fast-forwarded (Teixeira et al., 2012).Table 6 The interaction between media usage and product types.
Table 6 Before COVID-19 After COVID-19 Difference
WOM Search goods .084 .079 −.005
Experience goods .084 .078 −.006
(F (1, 11992) = .007, p = .935)
Traditional WOM Search goods .043 .040 −.003
Experience goods .047 .044 −.003
(F (1, 11992) = .625, p = .429)
e-WOM Search goods .041 .039 −.002
Experience goods .037 .034 −.003
(F (1, 11992) = .914, p = .339)
Social media advertising Search goods .072 .089 .017
Experience goods .058 .071 .013
(F (1, 11992) = 1.949, p = .163)
Text/image advertising Search goods .033 .030 −.003
Experience goods .025 .022 −.003
(F (1, 11992) = .002, p = .961)
Video advertising Search goods .040 .059 .019
Experience goods .032 .048 .016
(F (1, 11992) = 3.973, p = .046)
5 Conclusions
Given the unrivaled spike in digital media usage, higher dependency on credible information, and reduced face-to-face interactions during the COVID-19 pandemic, one of the focal questions for marketing scholars and practitioners is whether WOM has maintained its influence on consumers’ purchasing decisions. If dependence on WOM has decreased, which media has more power in purchasing decisions (Spotts et al., 2022)? We examined these research questions based on a sizable survey dataset from a leading panel operation firm in Korea before and after the COVID-19 pandemic. We also focused on the moderating role of product types because the changes in media usage were not consistent across different types of products, such as search and experience goods.
First, our study shows that consumers’ behavioral shift has decreased the power of WOM and enhanced the use of social media advertising in the context of an exogenous shock such as the COVID-19 pandemic. While previous studies focused on how false information can be spread through WOM and the factors that influence it (Di Domenico et al., 2021; Kwon et al., 2022; Wang et al., 2022), this study highlights whether individuals have changed their media usage during the pandemic. We argue that these distinctive and significant changes can be explained by supply- and demand-side perspectives. Searching for reviews or testimonials, biased by erratic credibility under uncertainty, leads consumers to depend more on firm-generated information with higher reliability and less variation. Given that the relationship between WOM and purchasing decisions is highly dependent on the uncertainty of the situation and the trustworthiness of the message sender (Reimer and Benkenstein, 2016; Šerić et al., 2022), consumers tended to use WOM less during the COVID-19 pandemic. Instead, consumers relied more on social media advertising on video-related platforms (e.g., YouTube, Tiktok) to make purchasing decisions (Casaló et al., 2020).
Second, this study goes beyond explaining the unidirectional relationship between media usage and external shocks by examining the moderating role of product types. Although the changes in WOM usage are not significantly different between search and experience goods, consumers tend to be influenced more by video advertising while considering purchasing search goods. This result is partly attributed to the fact that video platforms (e.g., YouTube) have been developed to be more effective in getting users not to skip advertising (Arantes et al., 2018). In fact, if they are not irritated, individuals tend to continue watching video advertising regardless of their preferences (Loureiro, 2018).
The academic contribution of this study is threefold. First, while some studies have demonstrated that WOM has an upward influence because of the increased number of proactive and empowered consumers (Hollebeek et al., 2019), our findings suggest that the influence of WOM on purchasing decisions has decreased during the COVID-19 pandemic. We explained this finding from demand- and supply-side perspectives. Second, we contribute to the literature on the rising importance of fake news and misinformation via WOM (Kwon et al., 2022; Wang et al., 2022). Although rumors can go viral via WOM quickly, the enormous amounts of rumors cannot be translated into credible information. Consumers want more clear and trustworthy sources to validate their purchasing decisions in economic downturns, such as a pandemic, which leads to a decline in the use of WOM. Our findings emphasize consumers' media usage changes to curtail the confusion and delay in decision-making affected by fake news via WOM. Third, we employ a moderator using product types related to information limitations that profoundly affect purchasing decisions. Prior studies have mostly focused on how the COVID-19 pandemic has affected consumer behavior. However, we demonstrate that changes in consumers’ media usage should be regarded based on product types to determine their exact influence.
Our findings have managerial implications for marketers, suggesting they should embrace consumers' changes in media usage. Evidently, the influencing media for purchasing decisions have changed during the pandemic. We find that recommendations from acquaintances or others’ online reviews are not as powerful as before the pandemic, even though WOM has been regarded as a medium to experience the brand and products with high persuasiveness (Chu and Kim, 2018). To counter this decline, marketers may need to consider changing communication channel approaches based on their product types to allocate resources efficiently. Additionally, our study suggests gradual but fundamental changes. Contrary to other online advertising media, the influence of video advertising on purchasing decisions has increased, especially for search goods. We suggest that the tendency to search for the necessary information online has changed from a relatively conventional method (e.g., text-based search or static advertising) to entertaining or easier-to-understand media (e.g., video-based social media advertising).
This study has several limitations that offer a foundation for future research. First, our dataset accounts for the overall pattern of the impact of media usage but does not show the relationship between the changes in the respective media. For example, our study does not explain the causality between the influence of video advertising in social media and that of WOM. Second, this study is based on survey data, which may be vulnerable to social desirability bias. Future studies using behavioral data may address this issue. Finally, this study cannot fully rule out an alternative explanation that the channel shift from traditional advertising to social media advertising or technological development (e.g., the dominance of video-related platforms) may be leading to the rise of video advertising that is not caused by the pandemic. Future research should also examine these macro-environmental factors, which may provide clearer and more meaningful insights, taking into account process change when examining the interplay between demand and supply side effects.
Funding
The authors disclose receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the Korean Marketing Association and Korea Research Association, and data were provided by the PMI CO., LTD. In addition, this study is (partially) supported by 10.13039/501100002642 Korea University Business School Research Grant.
Declaration of competing interest
None
Appendix A Survey Questionnaire
Item Survey Descriptions
Age Write down your age.
Gender Choose your gender. [1 = male, 2 = female]
Purchasing brand Which brand have you used in this product category? If you use more than two brands, then please choose the recent one.
Media usage for purchase decision Among the various media you encountered, please choose all the important media you use in your decision making to purchase the product.
Personal innovativeness I tend to use or purchase the newly launched brands, services, or products compared to other consumers. (1 = strongly disagree, 5 = strongly agree)
Online shopping frequency How often you do the online shopping? (1 = almost every day, 2 = more than two times per week, 3 = more than one time per week, 4 = more than two times per month, 5 = more than one time per month, 6 = more than one time per two-three month, 7 = more than one time per four-six month, 8 = less than any options shown above)
Note.
1. The original survey was conducted in Korean, translated in English by authors.
2. Respondents could choose specific brands among the options in each product category or write down the name of the brand they use.
3.19 product categories were shown randomly. If respondents did not use the given product category, then another product category was given in order.
4. See Appendix B for specific media usage options.
Appendix B Media Options
Media Elements
WOM 1) Word-of-mouth/oral tradition (heard from acquaintances)
2) Online referral/online review
Social media advertising 1) Social media advertising in the forms of posts that were generated by the firms (e.g., Twitter, Facebook)
2) Online/mobile video advertising that were generated by the firms (e.g., YouTube)
3) Live-commerce broadcasting
Retail-related media 1) Promotions (e.g., discounts, coupons) that were shown in the stores
2) Telemarketing
3) Recommendation from the store managers
4) Advertising or posters in fast-food stores
5) Advertising or posters in electronic products stores
6) Advertising or posters in duty-free shops
7) Advertising or posters in travel agencies
8) Advertising or posters in credit card agencies
9) Advertising or posters in coffee shops
10) Impression from the stores (e.g., interior, salespersons' uniforms)
11) Leaflets or catalogues in the stores
12) Advertising (e.g., point of purchase, poster, digital signages) in supermarkets or convenience stores
Traditional media 1) Advertising in terrestrial TV
2) Advertising in cable TV
3) Advertising in internet protocol TV
4) Radio advertising
5) Newspapers or magazines advertising
6) LED/standing signboard advertising
7) Bus/taxi shelter advertising
8) Subway station/train advertising
9) Press release in TV, newspaper, internet, or magazine
10) Product catalogue or product leaflet
Note: These are exemplary 27 media options for the mobile communication service. The number of media options are different across product categories.
Data availability
Data will be made available on request.
1 Although there exists a substantial amount of user-generated content on social media, the survey questionnaire explicitly asked respondents about the usage of company-generated ads in social media. Please refer to the details presented in Appendix A.
2 Please refer to Appendix A for the relevant survey questions.
==== Refs
References
Arantes M. Figueiredo F. Almeida J.M. Prates R.O. An investigation of user actions and experiences when exposed to YouTube video ads Proceedings of the 24th Brazilian Symposium on Multimedia and the Web 2018 ACM 283 290
Aini S.N. Zuliestiana D.A. Pengaruh E-wom credibility, quality, quantity pada minat beli dengan peran mediasi citra perusahaan dalam Facebook the body shop Indonesia eProceedings Manag. 6 1 2019
Balis J. 10 Truths about marketing after the pandemic Harv. Bus. Rev. 2021 https://hbr.org/2021/03/10-truths-about-marketing-after-the-pandemic
Basu S. Information search in the internet markets: experience versus search goods Electron. Commer. Res. Appl. 30 2018 25 37
Bataineh A.Q. Al-Abdallah G.M. Salhab H.A. Shoter A.M. The effect of relationship marketing on customer retention in the Jordanian's pharmaceutical sector Int. J. Bus. Manag. 10 3 2015 117 131
Berns G.S. Laibson D. Loewenstein G. Intertemporal choice–Toward an integrative framework Trends Cognit. Sci. 11 11 2007 482 488 17980645
Casaló L.V. Flavian C. Ibanez-Sanchez S. Influencers on Instagram: antecedents and consequences of opinion leadership J. Bus. Res. 117 2020 510 519
Chang H.H. Meyerhoefer C.D. COVID‐19 and the demand for online food shopping services: empirical Evidence from Taiwan Am. J. Agric. Econ. 103 2 2021 448 465
Cheung C.M. Lee M.K. Rabjohn N. The impact of electronic word‐of‐mouth: the adoption of online opinions in online customer communities Internet Res. 18 3 2008 229 247
Cheung M.Y. Luo C. Sia C.L. Chen H. Credibility of electronic word-of-mouth: informational and normative determinants of on-line consumer recommendations Int. J. Electron. Commer. 13 4 2009 9 38
Chu S.C. Kim J. The current state of knowledge on electronic word-of-mouth in advertising research Int. J. Advert. 37 1 2018 1 13
Colicev A. Malshe A. Pauwels K. O'Connor P. Improving consumer mindset metrics and shareholder value through social media: the different roles of owned and earned media J. Market. 82 1 2018 37 56
Deloitte Deloitte study: companies redirected almost half of the marketing budgets towards social media and mobile activities during the COVID-19 pandemic https://www2.deloitte.com/gr/en/pages/about-deloitte/articles/CMO-Survey-2020-COVID-19-and-the-State-of-Marketing.html 2020
Di Domenico G. Tuan A. Visentin M. Linguistic drivers of misinformation diffusion on social media during the COVID-19 pandemic Ital. J. Market. 2021 4 2021 351 369
Fang Y. Qureshi I. Sun H. McCole P. Ramsey E. Lim K.H. Trust, satisfaction, and online repurchase intention MIS Q. 38 2 2014 407 427
González E.M. Meyer J.H. Toldos M.P. What women want? How contextual product displays influence women's online shopping behavior J. Bus. Res. 123 2021 625 641
Gordon B.R. Goldfarb A. Li Y. Does price elasticity vary with economic growth? A cross-category analysis J. Market. Res. 50 1 2013 4 23
Harahap D. Hurriyati R. Gaffar V. Amanah D. The impact of word of mouth and university reputation on student decision to study at university Manag. Sci. Lett. 8 6 2018 649 658
Hasan B. Exploring gender differences in online shopping attitude Comput. Hum. Behav. 26 4 2010 597 601
Hoekstra J.C. Leeflang P.S. Marketing in the era of COVID-19 Ital. J. Market. 2020 4 2020 249 260
Hollebeek L.D. Sprott D.E. Andreassen T.W. Costley C. Klaus P. Kuppelwieser V. Rather R.A. Customer engagement in evolving technological environments: synopsis and guiding propositions Eur. J. Market. 53 9 2019 2018 2023
Jiang Y. Stylos N. Triggers of consumers' enhanced digital engagement and the role of digital technologies in transforming the retail ecosystem during COVID-19 pandemic Technol. Forecast. Soc. Change 172 2021 121029
Jobs C.G. Gilfoil D.M. A social media advertising adoption model for reallocation of traditional advertising budgets Acad. Market. Stud. J. 18 1 2014 235
Kasperson R.E. Renn O. Slovic P. Brown H.S. Emel J. Goble R. Ratick S. The social amplification of risk: a conceptual framework Risk Anal. 8 2 1988 177 187
Kim H.K. Ahn J. Atkinson L. Kahlor L.A. Effects of COVID-19 misinformation on information seeking, avoidance, and processing: a multicountry comparative study Sci. Commun. 42 5 2020 586 615
Kim H. Hanssens D.M. Advertising and word-of-mouth effects on pre-launch consumer interest and initial sales of experience products J. Interact. Market. 37 2017 57 74
Kirtis A.K. Karahan F. To be or not to be in social media arena as the most cost-efficient marketing strategy after the global recession Procedia Soc. Behav. Sci. 24 2011 260 268
Kwon K.H. Pellizzaro K. Shao C. Chadha M. I Heard that COVID-19 was..”: Rumors, Pandemic, and Psychological Distance 2022 American Behavioral Scientist
Kumar A. Bezawada R. Rishika R. Janakiraman R. Kannan P.K. From social to sale: the effects of firm-generated content in social media on customer behavior J. Market. 80 1 2016 7 25
Lee L. Inman J.J. Argo J.J. Böttger T. Dholakia U. Gilbride T. Tsai C.I. From browsing to buying and beyond: the needs-adaptive shopper journey model J. Assoc. Consum. Res. 3 3 2018 277 293
Lian J.W. Yen D.C. Online shopping drivers and barriers for older adults: age and gender differences Comput. Hum. Behav. 37 2014 133 143
Lkhaasuren M. Nam K.D. The effect of electronic word of mouth (eWOM) on purchase intention on Korean cosmetic products in the Mongolian market J. Int. Trade Commer. 14 4 2018 161 175
Lou C. Yuan S. Influencer marketing: how message value and credibility affect consumer trust of branded content on social media J. Interact. Advert. 19 1 2019 58 73
Loureiro S.M.C. Tell what you want but do not irritate me: a senior perspective about advertising J. Promot. Manag. 24 2 2018 198 214
Lu S. Yao D. Chen X. Grewal R. Do larger audiences generate greater revenues under pay what you want? Evidence from a live streaming platform Market. Sci. 40 5 2021 964 984
Moorman C. McCarthy T. CMOs: adapt your social media strategy for a post-pandemic world. Harvard Business Review Digital Articles, Nelson, P., 1970. Information and consumer behavior J. Polit. Econ. 78 2 2021 311 329
Nelson P. Information and consumer behavior J. Polit. Econ. 78 2 1970 311 329
Ng S. Faraji-Rad A. Batra R. Uncertainty evokes consumers' preference for brands incongruent with their global–local citizenship identity J. Market. Res. 58 2 2021 400 415
Park D.H. Kim S. The effects of consumer knowledge on message processing of electronic word-of-mouth via online consumer reviews Electron. Commer. Res. Appl. 7 4 2008 399 410
Rauniar R. Rawski G. Yang J. Johnson B. Technology acceptance model (TAM) and social media usage: an empirical study on Facebook J. Enterprise Inf. Manag. 27 1 2014 6 30
Reimer T. Benkenstein M. When good WOM hurts and bad WOM gains: the effect of untrustworthy online reviews J. Bus. Res. 69 12 2016 5993 6001
Roggeveen A.L. Sethuraman R. How the COVID-19 pandemic may change the world of retailing J. Retailing 96 2 2020 169
Rosario A.B. Sotgiu F. De Valck K. Bijmolt T.H.A. The effect of electronic word of mouth on sales: a meta-analytic review of platform, product, and metric factors J. Market. Res. 53 3 2016 297 318
Rosenbaum Paul R. Rubin Donald B. Constructing a control group using multivariate matched sampling methods that incorporate the propensity score Am. Statistician 39 1 1985 33 38
Rossolov A. Kuzkin O. Rossolova H. Emergency supplies purchase patterns during COVID-19 outbreak in the developing economy: frequency and stockpiling drivers' assessment J. Humanit. Logist. Supply Chain Manag. 12 1 2021 106 128
Schweidel D.A. Moe W.W. Binge watching and advertising J. Market. 80 5 2016 1 19
Šerić M. Došen D.O. Mikulić J. Antecedents and moderators of positive word of mouth communication among tourist destination residents during the COVID-19 pandemic Curr. Issues Tourism 1–18 2022
Sheth J. Impact of Covid-19 on consumer behavior: will the old habits return or die? J. Bus. Res. 117 2020 280 283 32536735
Sim J. Cho D. Hwang Y. Telang R. Frontiers: virus shook the streaming star: estimating the COVID-19 impact on music consumption Market. Sci. 41 1 2021 19 32
Spotts H.E. Weinberger M.G. Assaf A.G. Weinberger M.F. The role of paid media, earned media, and sales promotions in driving marcom sales performance in consumer services J. Bus. Res. 152 2022 387 397
Statista Coronavirus: impact on online usage in the U.S Statistics & Facts 2022 https://www.statista.com/topics/6241/coronavirus-impact-on-online-usage-in-the-us/#dossierKeyfigures
Steenkamp J.B.E. Fang E. The impact of economic contractions on the effectiveness of R&D and advertising: evidence from US companies spanning three decades Market. Sci. 30 4 2011 628 645
Szász L. Bálint C. Csíki O. Nagy B.Z. Rácz B.G. Csala D. Harris L.C. The impact of COVID-19 on the evolution of online retail: the pandemic as a window of opportunity J. Retailing Consum. Serv. 69 2022 103089
Tajuddin U.N.R.A. Hassan L.F.A. Razak N.A. Othman A.K. Role of brand image on the relationship between e-WOM and consumer purchase intention: dietary supplement consumers perspective Int. J. Bus. Manag. Stud. 7 2 2018 583 596
Tarigan Z. Jonathan M. Siagian H. Basana S. The effect of e-WOM through intention to use technology and social media community for mobile payments during the COVID-19 Int. J. Netw. Sci. 6 2 2022 563 572
Taylor R.C. Advertising and COVID-19 Int. J. Advert. 39 5 2020 587 589
Teng S. Khong K.W. Goh W.W. Chong A.Y.L. Examining the antecedents of persuasive eWOM messages in social media Online Inf. Rev. 38 6 2014 746 768
Tien D.H. Rivas A.A.A. Liao Y.K. Examining the influence of customer-to-customer electronic word-of-mouth on purchase intention in social networking sites Asia Pac. Manag. Rev. 24 3 2019 238 249
Teixeira P.J. Carraça E.V. Markland D. Silva M.N. Ryan R.M. Exercise, physical activity, and self-determination theory: a systematic review Int. J. Behav. Nutr. Phys. Activ. 9 1 2012 1 30
Van Heerde H.J. Gijsenberg M.J. Dekimpe M.G. Steenkamp J.B.E. Price and advertising effectiveness over the business cycle J. Market. Res. 50 2 2013 177 193
Voorveld H.A. Van Noort G. Muntinga D.G. Bronner F. Engagement with social media and social media advertising: the differentiating role of platform type J. Advert. 47 1 2018 38 54
Wang X. Chao F. Yu G. Zhang K. Factors influencing fake news rebuttal acceptance during the COVID-19 pandemic and the moderating effect of cognitive ability Comput. Hum. Behav. 130 2022 107174
WHO Fighting misinformation in the time of COVID-19 one click at a time https://www.who.int/news-room/feature-stories/detail/fighting-misinformation-in-the-time-of-covid-19-one-click-at-a-time 2021
Zhang K.Z. Zhao S.J. Cheung C.M. Lee M.K. Examining the influence of online reviews on consumers' decision-making: a heuristic–systematic model Decis. Support Syst. 67 2014 78 89
Godes D. Mayzlin D. Using online conversations to study word-of-mouth communication Market. Sci. 23 4 2004 545 560
Lovett M.J. Peres R. Shachar R. On brands and word of mouth J. Market. Res. 50 4 2013 427 444
Mazzarol T. Sweeney J.C. Soutar G.N. Conceptualizing word‐of‐mouth activity, triggers and conditions: an exploratory study Eur. J. Market. 41 11 2007 1475 1494
Trusov M. Bucklin R.E. Pauwels K. Effects of word-of-mouth versus traditional marketing: findings from an internet social networking site J. Market. 73 5 2009 90 102
Van Doorn J. Lemon K.N. Mittal V. Nass S. Pick D. Pirner P. Verhoef P.C. Customer engagement behavior: theoretical foundations and research directions J. Serv. Res. 13 3 2010 253 266
Aggarwal, R., Gopal, R., Gupta, A., Singh, H., 2012. Putting money where the mouths are: the relation between venture financing and electronic word-of-mouth. Inf. Syst. Res. 23 (3-NaN-2), 976–992.
Ghose, A., Ipeirotis, P.G., 2007. Designing novel review ranking systems: predicting the usefulness and impact of reviews. In: InProceedings of the Ninth International Conference on Electronic Commerce. pp. 303–310.
Gluck, M., Sales, M., 2008. The Future of Television: Advertising, Technology and the Pursuit of Audiences.
Google, 2022. How Google and YouTube are investing in fact-checking. https://blog.google/outreach-initiatives/google-news-initiative/how-google-and-youtube-are-investing-in-fact-checking. (Accessed 8 March 2023).
Shen, W., Hahn, J., 2008. Impact of Online Word-Of-Mouth on the Market for Consumer Goods–The Interplay between Adoption Rate, Product Market Life and Market size.Working Paper.
|
PMC010xxxxxx/PMC10165054.txt |
==== Front
Med Intensiva
Med Intensiva
Medicina Intensiva
0210-5691
1578-6749
The Authors. Published by Elsevier España, S.L.U.
S0210-5691(23)00110-9
10.1016/j.medin.2023.04.012
Carta Científica
Experiencia del ajuste de dosificación de enoxaparina profiláctica dirigida con niveles de factor anti-Xa en pacientes críticos con neumonía COVID-19: estudio observacional
Clinical experience in prophylactic enoxaparin dosage adjustment guided by anti-Xa factor levels in critical care patients with COVID-19 pneumonia: Observational studydel Carmen Bermúdez-Ruiz María a
Vilar Sánchez Irene a
Aparicio Pérez Clara b
Carmona Flores Rosario a
Rodríguez-Gómez Jorge ac⁎
de la Fuente-Martos Carmen ac
a Servicio de Medicina Intensiva, Hospital Universitario Reina Sofía. Córdoba, España
b Servicio de Hematología, Hospital Universitario Reina Sofía. Córdoba, España
c Instituto de Investigación Maimones de Investigación Biomédica de Córdoba. (IMIBIC), Córdoba, España
⁎ Autor para correspondencia.
8 5 2023
8 5 2023
© 2023 The Authors
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcSr. Editor,
El riesgo de eventos trombóticos está incrementado en pacientes con neumonía por coronavirus 2019 (COVID-19)1, 2. Guías clínicas y sociedades científicas difieren en las recomendaciones para la prevención de estos eventos; la Sociedad Española de Medicina Intensiva y Unidades Coronarias (SEMICYUC) propuso ajustar la dosificación de heparina de bajo peso molecular (HBPM) mediante la obtención de niveles de factor anti-Xa3, 4. Si bien justificaron esta recomendación por el alto riesgo de eventos trombóticos/hemorrágicos y la incidencia de infra/sobredosificación (23 y 46%), el nivel de evidencia fue C-III (apoyo ligero a la recomendación de su uso basada en opinión de expertos o estudios descriptivos). La utilidad de esta estrategia para la reducción de estas complicaciones, así como en la determinación de los niveles adecuados es controvertida3, 5, 6, 7. Los objetivos de nuestro estudio fueron evaluar el ajuste de enoxaparina profiláctica por factor anti-Xa, los niveles obtenidos y la presencia de posibles factores de riesgo relacionados con la sobredosificación en pacientes ingresados con neumonía COVID-19 en una UCI.
Iniciamos un estudio observacional y retrospectivo en una UCI de un hospital de tercer nivel durante el período de julio del 2020 a febrero del 2021. Se evaluó a pacientes críticos con neumonía COVID-19 consecutivos, en los que se dirigió la dosis de enoxaparina profiláctica por niveles de factor anti-Xa. Los criterios de inclusión fueron: 1) pacientes con neumonía COVID-19 (presencia de infiltrados radiológicos y test PCR positivo para SARS-CoV-2 en muestras respiratorias); 2) necesidad de soporte respiratorio (oxigenoterapia de alto flujo, soporte no invasivo o ventilación mecánica invasiva), y 3) ajuste de dosificación de enoxaparina profiláctica dirigido por factor anti-Xa. Se excluyeron los casos que inicialmente tenían indicación terapéutica. En pacientes que modificaron su régimen de profiláctico a terapéutico durante el ingreso se incluyeron las dosis y niveles, así como complicaciones trombóticas/hemorrágicas durante el período de profilaxis.
El protocolo de nuestro centro para las dosis iniciales de enoxaparina consistía en: 1) dosis estándar 40 mg/24 h; 2) dosis de 60 mg/24 h en pacientes con índice de masa corporal (IMC) > 30 o elevación de reactantes de fase aguda (D-dímero > 1.500 ng/ml y proteína C reactiva > 150 mg/l), y 3) dosis de 80 mg/24 h si se presentaban ambos factores de riesgo (IMC + reactantes elevados). En caso de insuficiencia renal se redujo la dosificación en función del aclaramiento de creatinina. Posteriormente se ajustó la dosis según niveles de anti-Xa; dado el elevado riesgo de trombosis publicado se acordaron multidisciplinarmente rangos objetivo a 0,30-0,59 UI/ml, similares a otros estudios publicados1, 6, 7. La medición se realizaba en estado de equilibrio de tratamiento de enoxaparina (48-72 h sin cambios en la dosificación) y en fase pico, tras 3-5 h de la administración. Se repitieron periódicamente cada 2-3 días según situación clínica y respuesta al tratamiento.
Se recogieron variables demográficas, relacionadas con comorbilidad, gravedad al ingreso en UCI, estudios analíticos, tratamiento y necesidad de soporte. También se recogieron los eventos resultado «eventos trombóticos» durante la estancia en la UCI (trombosis venosa profunda o tromboembolia pulmonar agudo diagnosticados por estudios de imagen y solicitados a criterio del facultativo) y «hemorrágicos» (grave: localización crítica, reducción de Hb 2 g/dl o transfusión de 2 concentrados de hematíes) y mortalidad durante la estancia en la UCI. Para el análisis estadístico, las variables categóricas se describieron con valor absoluto y proporción comparándose mediante test de la chi cuadrado o test de Fisher. Las variables continuas se representaron con su mediana y percentiles 25-75 comparándose mediante prueba t de Student o U de Mann-Whitney cuando las variables no cumplían la normalidad. Se analizaron los factores relacionados con las variables dependientes «niveles de factor anti-Xa ≥ 0,60 UI/ml» (valores superiores al rango objetivo) y «eventos trombóticos», mediante estudios de regresión logística multivariable. Se exploraron las variables con p < 0,2 o con mayor interés clínico. El punto de corte de las variables continuas se ajustó por el método de Youden.
Se incluyó a 160 pacientes críticos con neumonía COVID-19 en profilaxis con enoxaparina ingresados en la UCI, realizándose un total de 589 determinaciones de factor anti-Xa. Los pacientes presentaban una mediana de edad de 63 años, siendo la comorbilidad más frecuente la obesidad (67%, N 107/160) y mostrando un aclaramiento de creatinina < 30 ml/min al ingreso del 6% (10/160). La mediana de dímero D al ingreso fue del 1242 ng/ml (p25-75, 704-3765), requiriendo ventilación mecánica invasiva en el 58% (94/160). La mortalidad durante la estancia en la UCI fue del 28% (45/160), con una incidencia de eventos trombóticos diagnosticados del 9% (15/160) y sangrado grave del 4% (7/160) (tabla 1 ).Tabla 1 Características clínicas de pacientes
Tabla 1n 160
Demográficos
Edad (años), mediana (p 25-75) 63 (54-69)
Género, varón. n (%) 107 (66,9)
Antecedentes
Sin comorbilidades, n (%) 41 (25,6)
Obesidad (IMC ≥ 30), n (%) 108 (67,5)
Peso 82,5 (76,5-100)
Hipertensión, n (%) 92 (57,5)
Diabetes mellitus, n (%) 43 (26,9)
Fumador, n (%) 4 (2,5)
Insuficiencia cardíaca crónica, n (%) 14 (8,8)
Insuficiencia respiratoria crónica, n (%) 29 (18,1)
Insuficiencia renal crónica, n (%) 7 (4,4)
Enfermedad tromboembólica previa, n (%) 4 (2,5)
Antecedentes oncológicos, n (%) 14 (8,8)
Trasplante de órgano sólido, n (%) 1 (0,6)
Gravedad al ingreso
Escala APACHE II, mediana (p25-75) 9,5 (7-13)
Escala SOFA, mediana (p25-75) 4 (3-4)
Estudios de laboratorio
Plaquetas × 103/μl, mediana (p25-75) 285 (224-350)
Aclaramiento de creatinina ml/min, mediana (p25-75) 103 (70-125)
Aclaramiento de creatinina < 30 ml/min, mediana (p25-75) 10 (6,3)
Dímero D ng/ml, mediana (p25-75) 1242 (704-3765)
Proteína C reactiva mg/l, mediana (p25-75) 58 (24-127)
Tratamiento
Oxigenoterapia alto flujo, n (%) 138 (86,3)
Ventilación mecánica invasiva, n (%) 94 (58,8)
Prono, n (%) 77 (48,1)
Vasoactivos, n (%) 74 (46,3)
Técnicas de sustitución renal, n (%) 18 (11,3)
ECMO, n (%) 6 (3,8)
Bolos de corticoides, n (%) 109 (68,1)
Resultados
Eventos hemorrágicos, n (%) 17 (10,6)
Hemorragia grave, n (%) 7 (4,4)
Eventos trombóticos, n (%) 15 (9,4)
Tromboembolia pulmonar, n (%) 12 (7,5)
Mortalidad durante la estancia UCI, n (%) 45 (28,1)
APACHE: Acute Physiology And Chronic Health Evaluation; COVID-19: coronavirus 2019; ECMO: extra corporeal membrane oxygenation; IMC: índice de masa corporal; SOFA: Sequential Organ Failure Assessment; UCI: Unidad de Cuidados Intensivos.
La tabla 2 refleja las dosis de enoxaparina recibida y los valores de factor anti-Xa obtenidos en su primera determinación y durante la estancia en UCI. Cuando se realizó la primera determinación, la mediana de dosis de enoxaparina administrada fue de 60 mg/24 h (p25-75, 40-60). En relación con estas dosis, la primera determinación de niveles de factor anti-Xa presentaba una mediana de 0,28 UI/ml (0,34-0,50): el 53% (85/160) se encontraba en rango objetivo 0,30-0,59 UI/ml.Tabla 2 Dosis de enoxaparina profiláctica y niveles de factor anti-Xa en pacientes críticos con neumonía COVID-19
Tabla 2 Cohorte global Aclaramiento de creatinina IMC
< 30 ml/min ≥ 30 ml/min p < 30 ≥ 30 p
N 160 (100) 10 (6,2) 150 (93,7) 52 (32,5) 108 (67,5)
Primera determinación al ingreso en la UCI
Dosis de enoxaparina (mg/24‘h), mediana (p25-75) 60 (40-60) 40 (20-40) 60 (40-60) < 0,001 60 (40-60) 60 (40-60) 0,14
Pacientes con dosis de enoxaparina superiores a 40 mg/24 h, n (%) 155 (96,6) 5 (50) 150 (100) < 0,001 50 (96,2) 105 (97,2) 0,71
Primera determinación de niveles de anti-Xa, mediana (p25-75) 0,38 (0,24-0,50) 0,18 (0,12-0,26) 0,4 (0,26-0,50) < 0,001 0,40 (0,27-0,51) 0,34 (0,23-0,49) 0,20
Primera determinación de niveles de anti-Xa (UI/ml), n (%) 0,01 0,47
Anti-Xa < 0,10 2 (1,3) 0 (0) 2 (1,3) 1 0 (0) 2 (1,9)
Anti-Xa 0,10-0,29 53 (33,1) 8 (80) 45 (30) 0,02 14 (26,9) 39 (36,1)
Anti-Xa 0,30-0,59 86 (53,8) 2 (20) 84 (56) 0,04 31 (59,6) 55 (50,9)
Anti-Xa 0,60-0,69 11 (6,9) 0 (0) 11 (7,3) 1 4 (7,7) 7 (6,5)
Anti-Xa ≥ 0,70 8 (5,0) 0 (0) 8 (5,3) 1 3 (5,8) 5 (4,6)
N 160 (100) 22 (13,7) 138 (86,2) 52 (32,5) 108 (67,5)
Durante la estancia en la UCI
Media de dosis de enoxaparina (mg/24 h), mediana (p25-75) 60 (53-73) 60 (40-66) 60 (53-73) 0,10 60 (40-70) 60 (55-80) 0,01
Media de niveles de anti-Xa, mediana (p25-75) 0,48 (0,39-0,59) 0,48 (0,35-0,57) 0,48 (0,39-0,60) 0,80 0,47 (0,39-0,57) 0,49 (0,39-0,6) 0,90
Media de niveles de anti-Xa (UI/ml), n (%) 0,48 0,93
Anti-Xa < 0,10 2 (1,3) 1 (4,5) 1 (0,7) 1 (1,9) 1 (0,9)
Anti-Xa 0,10-0,29 18 (11,3) 3 (13,6) 15 (10,9) 6 (11,5) 12 (11,1)
Anti-Xa 0,30-0,59 100 (62,5) 13 (59,1) 87 (63) 33 (63,5) 67 (62)
Anti-Xa 0,60-0,69 29 (18,1) 3 (13,6) 26 (18,8) 7 (13,5) 22 (20,4)
Anti-Xa ≥ 0,70 11 (6,9) 2 (9,1) 9 (6,5) 5 (9,6) 6 (5,6)
Número de determinaciones de niveles de anti-Xa/paciente, mediana (p25-75) 3 (2-5) 4 (3-9) 3 (2-5) 0,02 3 (1,5-4) 2 (3-5,5) 0,31
COVID-19: coronavirus 2019; IMC: índice de masa corporal.
Los valores en negrita indican que el resultado es estadísticamente significativo; se consideran como tal aquellos valores de p < 0,05.
Durante el resto de la estancia, se ajustó la dosis de enoxaparina profiláctica según niveles de factor anti-Xa con objetivo (0,30-0,59 UI/ml). La media de niveles de factor anti-Xa fue de 0,48 UI/ml (0,39-0,59): el 62% (100/160) se encontraban en rango objetivo (0,30-0,59 UI/ml) y el 25% (24/160) en rango ≥ 0,60 UI/ml. Las dosis de enoxaparina > 60 mg/24 h (OR 4,57; IC del 95%, 3,17-6,60; p < 0,001) y los niveles de proteína C reactiva < 175 mg/dl (OR 2,30; IC 1,28-4,11; p = 0,002) se asociaron de forma independiente a un incremento del riesgo de obtener valores de factor anti-Xa ≥ 0,60 UI/ml (superiores al objetivo) en el análisis multivariante (anexo tabla suplementaria 1). La mediana de dosis de enoxaparina ajustada durante la estancia fue de 60 (53-73) mg/24 h, significativamente superior en pacientes con IMC > 30 de 60 (55-80) vs. 60 (40-70) mg/24 h (p = 0,01). No se objetivaron variables relacionadas de forma independiente con el desarrollo de eventos trombóticos (anexo tabla suplementaria 2).
En la actualidad continúan publicándose nuevos estudios relacionados con la prevención de la trombosis en los pacientes COVID-19 (p, ej., INSPIRATION, REMAP-CUP, ATTACC, ACTIV-4a, etc.), pero aún no han sido aclarados aspectos clave como la dosis adecuada de HBPM y la posible optimización mediante niveles de factor anti-Xa2, 3. En esta experiencia observamos que si bien al dirigir la dosis de enoxaparina profiláctica el 62% de los pacientes se encontraba con media de niveles de anti-Xa en rango objetivo, el 25% presentaba rangos superiores (≥ 0,60 UI/ml). Mayores dosis de heparina (enoxaparina > 60 mg/24 h) y menores niveles de proteína C reactiva (< 175 mg/dl) se asociaban de forma independiente a un mayor riesgo de sobredosificación. Bösch et al. han objetivado cómo valores elevados de proteína C reactiva puede influir en una resistencia a la actividad de la HBPM, lo que pudiera explicar que pacientes con valores más bajos en nuestro estudio estuvieran más expuestos a una sobredosificación8. A pesar de que niveles superiores al rango objetivo pueden suponer potencialmente un mayor riesgo de complicaciones, únicamente el 6% tenía niveles ≥ 0,70 UI/ml y la incidencia de hemorragias graves fue del 4%, no superior a la descrita en la literatura 2-6%6, 7, 9.
En este estudio la dosis enoxaparina administrada durante la estancia fue de 60 mg/24 h (p25-57, 50-73 mg). Estas dosis son similares a la de otros estudios (mediana 60, 50-80 mg/24 h) que utilizaron monitorización con similares niveles objetivo (0,4-0,5 UI/ml)6. Mohamed et al. encuentran una elevada incidencia de sobredosificación (48%) cuando se administran dosis intermedias (0,5 mg/kg/12 h), una de las estrategias recomendadas en la literatura, y se monitorizan niveles de anti-Xa3, 7. El riesgo de sobredosificación podría haber sido mayor de haber utilizado este régimen y no haber realizado monitorización, dada la alta incidencia de obesidad (67%) y niveles elevados de reactantes de fase aguda en nuestra cohorte2, 3, 7.
Finalmente, el objetivo de dirigir la dosificación de enoxaparina según factor anti-Xa es reducir el riesgo de eventos trombóticos. Destaca que en nuestro estudio se objetivó una incidencia del 9%, límite inferior de lo publicado en la literatura (9-26%), pero este análisis queda limitado, pues no se realizó una búsqueda sistemática de los mismos pudiendo infra diagnosticarse algunos eventos2, 3.
Somos conscientes de las limitaciones del estudio pues se trata de un estudio observacional, retrospectivo, con búsqueda no sistemática de eventos trombóticos y de tamaño muestral limitado. Además, el rango objetivo planteado, si bien es similar a otros estudios, es controvertido con importante variabilidad en las publicaciones (p. ej., pacientes quirúrgicos 0,1-0,3 vs. otros estudios COVID-19 0,3-0,7 UI ml)5, 6, 7, 8, 9, 10. A pesar de ello, creemos que esta experiencia clínica puede ser útil en el planteamiento de futuros estudios.
En conclusión, en este estudio de pacientes críticos con neumonía COVID-19 los niveles de factor anti-Xa obtenidos durante la estancia en UCI permitieron ajustar las dosis de enoxaparina profiláctica, si bien existe un riesgo elevado de sobredosificación incrementado en pacientes con mayores dosis de enoxaparina y niveles más bajos de proteína C reactiva.
Ética de la publicación científica
Este estudio fue aprobado por el Comité de Ética del Hospital Universitario Reina Sofía (Código 010408), el cual eximia de la necesidad de consentimiento escrito dada la naturaleza observacional y retrospectiva del mismo.
Financiación
Este estudio no ha obtenido financiación para su realización.
Conflicto de intereses
Los autores declaran que no existe conflicto de intereses
Anexo Material adicional
Agradecimientos
Los autores agradecen al Prof. Manuel Rodríguez Peralvarez y al Dr. Rafael León López el apoyo en el análisis estadístico de los datos.
Anexo Se puede consultar material adicional a este artículo en su versión electrónica disponible en doi:10.1016/j.medin.2023.04.012.
==== Refs
Bibliografía
1 Wichmann D. Sperhake J.-P. Lütgehetmann M. Steurer S. Edler C. Heinemann A. Autopsy findings and venous thromboembolism in patients with COVID-19 Ann Intern Med. 173 2020 268 277 32374815
2 Vincent J.L. Levi M. Hunt B.J. Prevention and management of thrombosis in hospitalised patients with COVID-19 pneumonia Lancet Respir Med. 10 2022 214 220 34838161
3 Flacyk A. Rosovsky R.P. Reed C.T. Bankhead-Kendall B.K. Bittner E.A. Chang M.G. Comparison of published guidelines for management of coagulopathy and thrombosis in critically ill patients with COVID-19: Implications for clinical practice and future investigations Critical Care. 24 2020 559 572 32938471
4 Vidal-Cortés P. Díaz Santos E. Aguilar Alonso E. Amezaga Menéndez R. Ballesteros M.A. Bodí M.A. Recomendaciones para el manejo de los pacientes críticos COVID-19 en las Unidades de Cuidados Intensivos Med Intensiva. 46 2022 81 89
5 Witt D.M. Nieuwlaat R. Clark N.P. Ansell J. Holbrook A. Skov J. American Society of Hematology 2018 guidelines for management of venous thromboembolism: Optimal management of anticoagulation therapy Blood Adv. 2 2018 3257 3291 30482765
6 Kofteridis D.P. Ioannou P. Kondili E. Chamilos G. Filippatos T.D. Personalized prophylactic anticoagulation in hospitalized patients with Covid-19 —The role of anti-Xa monitoring Clin Microbiol Infect. 27 2021 1188 1189 33933565
7 Mohamed A. Shemanski S.M. O Saad M. Ploetz J. Haines M.M. Schlachter A.B. Anti-Xa directed thromboprophylaxis in critically ill patients with coronavirus disease 2019 Clin Appl Thromb Hemost. 28 2022 1 9
8 Bösch J. Rugg C. Schäfer V. Lichtenberger P. Staier N. Treichl B. Low-molecular-weight heparin resistance and its viscoelastic assessment in critically ill COVID-19 patients Semin Thromb Hemost. 48 2022 850 857 36174602
9 Trunfio M. Salvador E. Cabodi D. Marinaro L. Alcantarini C. Gaviraghi A. e-COVID Study group Anti-Xa monitoring improves low-molecular-weight heparin effectiveness in patients with SARS-CoV-2 infection Thromb Res. 196 2020 432 434 33049598
10 Cook D. Crowther M. Meade M. Deep venous thrombosis in medical-surgical critically ill patients: Prevalence, incidence, and risk factors Crit Care Med. 33 2005 1565 1571 16003063
|
PMC010xxxxxx/PMC10165275.txt |
==== Front
Arch Virol
Arch Virol
Archives of Virology
0304-8608
1432-8798
Springer Vienna Vienna
37155081
5787
10.1007/s00705-023-05787-6
Brief Report
Consensus insertion/deletions and amino acid variations of all coding and noncoding regions of the SARS-CoV-2 Omicron clades, including the XBB and BQ.1 lineages
Suharsono Hamong 1
Mahardika Bayu K. 2
Sudipa Putu H. 3
Sari Tri K. 4
Suardana Ida B. K. 4
http://orcid.org/0000-0001-5525-0793
Mahardika Gusti N. gnmahardika@unud.ac.id
24
1 grid.412828.5 0000 0001 0692 6937 Biochemistry Laboratory, The Faculty of Veterinary Medicine, Udayana University, Denpasar, Bali Indonesia
2 grid.412828.5 0000 0001 0692 6937 The Animal Biomedical and Molecular Biology Laboratory, Udayana University, Jl. Sesetan—Markisa 6A, Denpasar, 80223 Bali Indonesia
3 grid.412828.5 0000 0001 0692 6937 Veterinary Bacteriology and Mycology Laboratory, The Faculty of Veterinary Medicine, Udayana University, Denpasar, Bali, Indonesia
4 grid.412828.5 0000 0001 0692 6937 Virology Laboratory, The Faculty of Veterinary Medicine, Udayana University, Denpasar, Bali Indonesia
Handling Editor: T. K. Frey.
8 5 2023
2023
168 6 15621 12 2022
18 4 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The currently dominant Omicron variant of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has swiftly diverged into clades. To predict the probable impact of these clades, the consensus insertions/deletions (indels) and amino acid substitutions of the whole genome of clades were compared with the original SARS-CoV-2 strain. The evolutionary history of representatives of clades and lineages was inferred using the maximum-likelihood method and tested using the bootstrap method. The indels and polymorphic amino acids were found to be either clade-specific or shared among clades. The 21K clade has unique indels and substitutions, which probably represent reverted indels/substitutions. Three variations that appear to be associated with SARS-CoV-2 attenuation in the Omicron clades included a deletion in the nucleocapsid gene, a deletion in the 3’untranslated region, and a truncation in open reading frame 8. Phylogenetic analysis showed that the Omicron clades and lineages form three separate clusters.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00705-023-05787-6.
issue-copyright-statement© Springer-Verlag GmbH Austria, part of Springer Nature 2023
==== Body
pmcIntroduction
The coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) will soon pass its three-year point. Despite mass vaccination with boosters in various countries, vaccines seem ineffective in curbing community transmission, as vaccine breakthrough is a common phenomenon observed in many parts of the world [1–3]. It is believed to occur due to the rapid evolution of SARS-CoV-2, which has led to the generation of many variants. The most recent and dominant variant is Omicron [1, 4, 5]. This variant has swiftly diverged into at least eight clades based on the data available in GISAID (https://nextstrain.org). According to the WHO Technical Advisory Group on SARS-CoV-2 Virus Evolution, some lineages, such as XBB and BQ.1, are a cause of concern because of their potential to bring on a new wave of cases and fatalities (www.who.int).
To predict the probable impact of these genetic variants, comparisons of whole-genome sequences of representatives of each clade or lineage are of global interest. However, as interclade or lineage variants are to be expected, preserved insertions/deletions (indels) and substitutions in comparison to the original SARS-CoV-2 strain are important genetic features of new clades/lineages. The genome organization of SARS-CoV-2, based on the open reading frame (ORF) annotation of the original Wuhan-Hu-1 isolate and adding the 5’ and 3’untranslated regions (UTRs) and intergenic sequences (IGSs) is as follows: 5’UTR-ORF1AB-IGS-Spike-IGS-ORF3A-ORF3B-IGS-Protein E-IGS-membrane (MA)-IGS-ORF6-ORF7A-ORF7B-ORF8-IGS-Nucleoprotein NP-ORF-10-3’UTR [6, 7]. Intergenic sequences (IGSs) have been identified previously [8]. Transcription regulatory sequences (TRSs) have been identified at the junctions between these ORFs as well as at the 5′ end of the genomic RNA downstream of the leader sequence [9]. Therefore, the 3' end of any SARS-CoV-2 gene can be critical for the translation of the next coding region. Accessory proteins should also be examined because they contribute to the pathogenesis of SARS-CoV-2 [10, 11]. The 5’- and 3’-UTRs should be examined, as they have been demonstrated to play important roles in viral fitness and pathogenesis [12].
Here, we compared the genome sequences of members of various clades of the Omicron variant as well as the XBB and BQ.1 lineages. The objective of this study was to identify unique consensus insertion/deletions and amino acid variations of all coding and noncoding regions of the SARS-CoV-2 Omicron clades, including the XBB and BQ.1 lineages.
Methods
Fifteen to 25 sequences of lineages 21K, 21L, 22A, 22B, 22C, 22D, 22E, and 22F, as well as XBB and BQ.1, were randomly selected from the GISAID Nextstrain phylogeny on October 31, 2022 and downloaded. The total number of sequences in the dataset was 203. The whole genome sequence was aligned with the original SARS-CoV-2 sequence of the Wuhan-Hu-1 isolate (GenBank accession no. NC_045512) using Clustal Omega, available online at EMBL-EBI (www.ebi.ac.uk). In whole-genome sequence alignments, sequences that caused long gaps due to the presence of a long track of Ns were excluded. Individual coding regions were selected based on the Wuhan-Hu-1 coding DNA sequence (CDSs) using MEGA11 software [13]. In comparisons of coding regions or open reading frames (ORFs), sequences with a track of more than two unidentified nucleotides (NNNs) were excluded. Therefore, the final number of sequences of each clade and or lineage varies, as shown in Tables 1, 2, and 3. The sequences were translated prior to alignment. Polymorphic amino acids as well as gaps were tabulated manually. To assess the genetic relatedness of clades and lineages, three representatives of clades and lineages from different countries were randomly selected from the dataset as above and aligned using Clustal Omega. The 5’ and 3’ends were trimmed to produce sequences of equal length, with a total of 28,934 positions in the final dataset. The evolutionary history was inferred using the maximum-likelihood method and the Kimura 2-parameter model in MEGA 11 software [13]. The phylogeny was tested by the bootstrap method with 100 replications.Table 1 Consensus amino acid substitutions/deletions in the ORF1AB protein of Omicron variant clades/lineages compared to Wuhan-Hu-1*
Strain/clade/ lineage Number of sequences Position
47 135 141–143 556 842 856 1221 1307 1640 2084 2235 2710 2909 3027 3090 3201 3255 3395 3674
Wuhan-Hu-1 1 K S KSF Q T K S G P L I A A L T L T P L
21 K 19 K S KSF Q T R S G P I L T A L T L I H L
21L 20 K R KSF Q I K S S P L L A V F I F I H L
22A 14 K R del Q I K S S P L L A A F I L I H L
22B 14 K R KSF Q I K S S P L L A A F I F I H L
22C 12 K R KSF Q I K S S P L L A A F I F I H L
22D 11 K R KSF Q I K L S S L L A A F I F I H L
22E 9 K R KSF K I K S S P L L A A F I F I H F
22F 10 R R KSF Q I K S S P L L A A F I F I H L
XBB 25 R R KSF Q I K S S P L L A A F I F I H L
BQ.1 13 K R KSF K I K S S P L L A A F I L I H L
Strain/clade/ lineage Number of sequences Position
3675–3677 3758 3829 3833 4060 4665 4715 5063 5360 5557 5592 5716 5967 6564
Wuhan-Hu-1 1 SGF I L K N Y P G S M N R I T
21 K 19 del V L N N Y L G S M N R V T
21L 20 del I L N N Y L G S M N C V I
22A 14 del I L N N Y L G S M N C V I
22B 14 del I L N N Y L G S M N C V I
22C 12 del I L N N Y L G S M N C V I
22D 11 del I L N S Y L S S M N C V I
22E 9 del I F N N H L G S I S C V I
22F 10 del I L N N Y L S P M N C V I
XBB 25 del I L N N Y L S S M N C V I
BQ.1 13 del I F N N H L G S I S C V I
* Residues unique to certain clades and/or lineages are highlighted
Table 2 Consensus amino acid substitutions/deletions in the spike protein of Omicron variant clades/lineages compared to Wuhan-Hu-1*
Strain/ clade/ lineage Number of Seq.1 Amino acid position
19 24–26 27 67 69–70 83 95 142 143–145 146 147 152 157 183 210 211 212 213 215–217 255 260 342 348 371 374 376 378 379 408
Wuhan Hu-1 1 T LPP A A HV V T G VYY H K W F Q I N L V –- G G G R L S S S T D
21 K 12 T LPP A V del V I D del H K W F Q I del I V EPE G G D K L L P F T D
21L 19 I del S A HV V T D VYY H K W F Q I N L G –- G G D R L F P F A N
22A 16 I del S A del V T D VYY H K W F Q I N L G –- G G D T L F P F A N
22B 14 I del S A del V T D VYY H K W F Q I N L G –- G G D R L F P F A N
22C 22 I del S A HV V T D VYY H K W F Q I N L G –- G G D R L F P F A N
22D 12 I del S A HV V T D VYY H E R L Q V N L G –- G S H R L F P F A N
22E 11 I del S A del V T D VYY H K W F Q I N L G –- G G D T L F P F A N
22F 10 I del S A HV A T D V–Y Q K W F E I N L E –- V G H T I F P F A N
XBB 25 I del S A HV A T D V–Y Q K W F E I N L E –- V G H T I F P F A N
BQ.1 16 I del S A HV V T D VYY H K W F Q I N L G –- G G D T L F P F A N
Strain/
clade/
lineage Amino acid position
411 420 443 447 448 449 455 463 480 481 487 489 493 496 499 501 504 508 550 617 658 661 682 684 707 767 799 859 957 972 984
Wuhan Hu-1 R K N K V G L N S T E F F Q G Q N Y T D H N N P S N D N Q N L
21 K R N K K V S L N N K A F F R S R Y H K G Y N K H S K Y K H K F
21L S N K K V G L N N K A F F R G R Y H T G Y N K H S K Y N H K L
22A S N K K V G R N N K A V F Q G R Y H T G Y S K H S K Y N H K L
22B S N K K V G R N N K A V F Q G R Y H T G Y N K H S K Y N H K L
22C S N K K V G Q N N K A F F R G R Y H T G Y N K H L K Y N H K L
22D S N K K V S L K N K A F F Q G R Y H T G Y N K H S K Y N H K L
22E S N K T V G R K N K A V F Q G R Y H T G Y N K H S K Y N H K L
22F S N K K P S L K N K A S S Q G R Y H T G Y N K H S K Y N H K L
XBB S N K K P S L K N K A S S Q G R Y H T G Y N K H S K Y N H K L
BQ.1 S N K T V G R K N K A V F Q G R Y H T G Y N K H S K Y N H K L
* Residues unique to certain clades and/or lineages are highlighted. 1Number of sequences of the spike protein in each clade/lineage without a track of more than two ‘Ns’
Table 3 Consensus amino acid substitutions/deletions of ORF3A, envelope protein, matrix, ORF6, ORF7B, ORF8, and nucleoprotein of Omicron variant clades/lineages compared to Wuhan-Hu-11
Strain/ clade/ lineage ORF3A Envelope protein Matrix ORF63 ORF7B ORF8 Nucleoprotein
Number of seq.2 140 223 Number of seq
2 9 11 Number of seq.2 3 19 63 Number of seq.2 61 Number of seq.2 11 Number of seq.2 8 Number of seq. 2 13 31–33 204 417
Wuhan Hu-1 1 L T 1 T T 1 D Q A 1 D 1 L 1 G 1 P ERS G S
21 K 21 L T 21 I T 17 G E T 21 D 19 L 19 G 19 L del R S
21L 23 F I 24 I T 20 D E T 24 L 24 L 24 G 21 L del R R
22A 24 L I 24 I T 21 D E T 21 L 21 F 15* G 24 L del R R
22B 21 L I 21 I T 19 N E T 21 D 21 L 21 G 21 L del R R
22C 24 L I 24 I T 21 D E T 22 L 23 L 23 G1 24 L del R R
22D 13 L I 13 I A 9 D E T 14 L 14 L 14 G 14 L del R R
22E 12 L I 13 I T 9 N E T 13 D 11 L 11 G 12 L del R R
22F 15 L I 15 I A 15 D E T 15 L 14 L 14 Stop 15 L del R R
XBB 25 L I 25 I A 24 D E T 25 L 25 L 25 Stop 25 L del R R
BQ.1 20 L I 20 I T 16 N E T 20 D 19 L 16 G 19 L del R R
1No dominant (> 50% of sequence data) amino acids, deletions, or insertions were found in ORF7A and ORF10. 2 Number of sequences in each clade/lineage without a track of more than two ‘Ns’. 3Stop codon at position 62 of ORF7A in one sequence. 3Stop codon at position 8 in one 21 K sequence. 4Stop codon at position 8 in one sequence. 5Stop codon at position 8 in three sequence of clade 22F. 6 Stop codon at position 8 of 31 sequences of lineage XBB. Stop codon in ORF7a at position 62 in 21 K. * massive deletion in 5 of the 22A sequences. Residues unique to certain clades and/or lineages are highlighted
Results
The sequence dataset is available at GISAID with the identifier EPI_SET ID: EPI_SET_230327ca and https://doi.org/10.55876/gis8.230327ca. The number of sequences in the dataset of various clades and the XBB and BQ1 lineages and the number of sequences in each clade and lineage bearing consensus deletions and insertions are presented in Supplementary Material 1. The consensus indel pattern in all Omicron clades or lineages includes del-11260-11268 in ORF1AB and del-28346-28354 in NC. Other indels are unique to a clade or a lineage or shared between clades and/or lineages. The 21K clade has two unique deletions and an insertion (spike del-21960-21965, del-22167-22169, and ins-22178-22186). The deletion del-658-666 is unique to the 22A clade. Del-21606-21614 in ORF1AB and del-29732-29757 in the 3’-UTR are present in all clades except 21K. Del-21738-21743 in the spike coding region is not present in clades 21L, 22C, 22D, and 22F. The other deletion in the spike coding region, namely, 21966-21968, is present in 21K, 22F, and XBB.
Amino acid residues that are characteristic of all Omicron clades and the XBB and BQ.1 lineages in ORF1AB, the spike gene, and the combined ORF3A, envelope protein, matrix ORF6, ORF7B, ORF8, and NC are shown in Tables 1, 2, and 3, respectively. The substitutions that all of these clades and lineages have in common are I2235L, T3255I, P3395H, S3675Del, G3676Del, F3677Del, K3833N, P4715L, and I4175V in ORF1AB; G142D, S376P, S378F, K420N, N443K, S480N, T481K, E487A, Q501R, N504Y, Y508H, D617G, H658Y, N682K, P684H, N767K, D799Y, Q957H, and N972K in the spike protein; T9I in the envelope protein; Q19E and A63T in the matrix protein; and P13L, E31del, R32del, S33del and G204R in the NC. The amino acid substitutions/deletions present in all clades and lineages except the 21K clade are S135R, T842I, G1367S, L3207F, T3090I, R5716C, and T6564I in ORF1AB; T19I, L24del, P25del, P26del, A27S, V21G, T379A, D408N, and R411S in the spike protein; T22I in ORF3A; and S417R in the NC.
The unique substitutions in the 22K clade are K856R, L2084I, A2710T, and I3758V in ORF1AB and A67V, T95I, 143Vdel, 144Ydel, 145Ydel, N211del, L212I, ins215E, ins216P, ins217E, G499S, T550K, N859K, and L984F in the spike protein. The remaining clades resemble Wuhan-Hu-1 at those sites.
Substitutions found in both 22F and XBB include K47R in ORF1AB and V83A, Y144del, H146Q, Q183E, V213E, G255V, L371I, V448P, F489S, and F493S in the spike protein. Unique to the 22F clade and XBB lineage is the presence of a stop codon at position 8 of ORF8. Q556K, L3829F, Y4665H, M5557I, and N5592S in ORF1AB, as well as K447T in the spike protein are shared by 22E and BQ.1.
A phylogenetic tree is presented in Figure 1. The tree shows that the Omicron clades and lineages form three separated clusters with 100% bootstrap support. Clade 22K forms a unique cluster (cluster 1), while cluster 2 consists of clades 22B and 22E as well as the BQ lineage, and the other clades and lineages form cluster 3.Fig. 1 Phylogenetic tree, based on full genome sequences of three randomly selected representatives of the Omicron clades and lineages of SARS-CoV-2 rooted to the Wuhan-Hu-1 strain. The clade, country of origin, and EPI-ID are indicated in the isolate names. The evolutionary history was inferred using the maximum-likelihood method and the Kimura 2-parameter model in MEGA 11 software [13]. The phylogeny was tested by the bootstrap method with 100 replications. The tree with the highest log likelihood is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. There were a total of 28934 positions in the final dataset.
Discussion
Consensus indels and amino acid variations in all coding and noncoding regions of the SARS-CoV-2 Omicron variant are of global interest, as this variant is evolving rapidly and has become the global dominant circulating variant, suppressing others. Scientific explanation is paramount to understanding the potential threat of subvariants or lineages. Most reports on SARS-CoV-2 variants have emphasized changes in the spike protein [3, 4]. However, examination of the whole genome is important for assessing the impact of mutations in subvariants [14].
Some indels and polymorphic amino acids are specific to the Omicron variant. All Omicron clades and lineages contain del-11260-11268 in ORF1AB and del-28346-28354 in NC. However, del-11260-11268 is not unique, as it is also present in the Alpha, Beta, and Gamma variants [15].
The deletion in the NC of SARS-CoV-2 is a probable indirect signature of its attenuation. The SARS-CoV-2 NC is an abundantly expressed RNA-binding protein that is critical for viral genome packaging [16]. The presence of del-28346-28354 in the NC can potentially alter the biology of the virus. In the coronavirus mouse hepatitis virus, a deletion in the NC resulted in a small-plaque phenotype in tissue culture [17]. Plaque size is also an indicator of dengue virus attenuation [18], in which the molecular determinants of small plaque size are mutations/substitutions in NS1, NS3, and the 3'-UTR [19]. It is therefore plausible that the deletion of the NC in SARS-CoV-2 is an indirect indicator of its attenuation.
The deletion in the 3’-UTR is another notable characteristic of the Omicron variant. This deletion is dominant in all clades/lineages, with the exception of clade 21K. This region might be important for recognition by the SARS-CoV-2 RNA-dependent RNA polymerase and cellular components for the initiation of anti-genomic (negative strand) RNA synthesis [12]. The deletion in the 3’-UTR is another probable indirect indicator of SARS-CoV-2 subvariant attenuation.
Although Omicron SARS-CoV-2 spread faster than other variants and became the dominant variant globally, it was reported to cause milder clinical signs [5]. The intensive care unit admission rates for Omicron-infected patients were much lower than those of Delta- and Delta-/Omicron-infected patients [20], suggesting that this variant has reduced virulence.
Since many dominant amino acid residues in the spike protein are uniformly divergent from Wuhan-Hu-1, individuals who have recovered from an Omicron infection might be expected to have protective immunity to all Omicron clades/lineages. Applying the template of spike protein residues and their possible functions as published previously [4], the consensus amino acid changes in Omicron clades/lineages relative to Wuhan-Hu-1 are located in the receptor-binding domain/receptor binding site (RBD/RBS) (S376P, S378F, K420N, N443K, S480N, T481K, E487A, Q501R, N504Y, Y508H), linear epitopes (S378F, K420N, E487A, Q501R, D617G, N767K), possible conformation-dependent epitopes (N682K, P684H), the S1/S2 cleavage site ((N682K, P684H), the fusion peptide (D799Y), and heptad repeat 1 (Q957H and N972K). With this pattern, it is expected that the Omicron variant has different biological characteristics than the original SARS-CoV-2 strain.
It was observed in this study that reversion of indels and mutations might have occurred in SARS-CoV-2. In this case, the term "reversions" or "reverse mutations", refers to any mutational processes or mutations that restores the wild-type phenotype to an organism already carrying a phenotype-altering forward mutation [21]. This phenomenon has been described for many viruses [22]. Our data show that the 21K clade has unique indels and substitutions, while the remaining sequence is homologous to that of Wuhan-Hu-1. The revertant virus evolved further with deletions and substitutions in various genome segments. The indels and substitutions in 21K seem to be unstable or generate lower virus fitness.
In this study, we confirm the clade separation reported by Nextstrain. XBB is close to the 22F clade, while BQ1 is close to the 22E clade. Clade 22F shares many amino acid substitutions with the XBB lineage, while clade 22E shares many with the BQ.1 lineage. This manuscript was drafted to provide valid data on the position of both lineages in SARS-CoV-2 phylogeny. This should suppress public speculation that XBB and BQ.1 are de novo subvariants and should demonstrate that the genetic make-up of these subvariants is similar to that of other members of the clade. The phylogenetic analysis (Fig. 1) also confirmed that the BQ.1 lineage is close to the 22E clades, while the XBB lineage is close to 22F clade.
Another note from this analysis is that the accessory proteins might not be critical for SARS-CoV-2 integrity. Without those proteins, SARS-CoV-2 remains viable. ORF8 is truncated in the 22F clade and XBB lineage. A stop codon at position 8 of ORF8 is dominant in that clade/lineage. Stop codons were also present in ORF6 and ORF7A of some strains (not shown). The accessory proteins have been described to contribute to the pathogenesis of SARS-CoV-2 [10, 11]. Deletions in ORF7 and 8 have been associated with milder symptoms [23]. ORF8 interferes with host immune responses in various ways, including downregulating MHC class I molecules [24], antagonizing interferon [25], activating interleukin 17, and cytokine storms [26]. It is therefore reasonable to suggest that the truncated ORF8 is additional indirect evidence of lower virulence or attenuation, especially in clade 22E.
This study does not provide information for prediction of the outcomes of SARS-CoV-2 infection. A scientific task force should be formed in each country to study the association of Omicron with the patient’s clinical status so that the public can be aware of whether any emerging variant or subvariant warrants a change in COVID-19 prevention protocols.
In conclusion, the indels and polymorphic amino acids across the whole genome of SARS-CoV-2 Omicron clades are either clade-specific or shared among clades. Del-28346-28354 in NC is unique to Omicron. Variation in the 3’-UTR are common to all clades/lineages, except clade 21K. Clade 21K has four unique indels and substitutions in ORF1AB and 14 in the spike protein, while the remaining sequence is homologous to that of Wuhan-Hu-1, which probably represents reverted indels/substitutions. ORF8 is truncated at amino acid 8 in the 22F clade and in the XBB lineage. Three indirect lines of evidence of SARS-CoV-2 attenuation in Omicron clades were identified, namely, the deletion in NC, the deletion in the 3’-UTR, and the truncation of ORF8 in the 22F clade and XBB lineage.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 16 KB)
The English language of the manuscript was edited by Springer Nature Author Service.
Author contributions
HS and GNM contributed to study conception. HS, IBKS, BKM, TKS, and PHS contributed to data acquisition and analysis. BKM and GNM interpreted the data and drafted the manuscript.
Funding
This study was supported by the Udayana University Grant of Innovation Product Scheme 2022.
Data availability
The sequence data of 15 to 20 representatives of Nextstrain clades are accessible at GISAID with the identifier EPI_SET ID: EPI_SET_230327ca and 10.55876/gis8.230327ca. All genome sequences and associated metadata in this dataset have been published in GISAID’s EpiCov database. To view the contributors and each individual sequence with details such as accession number, virus name, collection date, originating lab, and submitting lab and the list of authors, visit 10.55876/gis8.230327ca.
Declarations
Conflict of interest
The authors declare no conflict of interest.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Boekel L Besten YR Hooijberg F Wartena R Steenhuis M Vogelzang E Leeuw M Atiqi S Tas SW Lems WF SARS-CoV-2 breakthrough infections in patients with immune-mediated inflammatory diseases during the omicron dominant period Lancet Rheumatol 2022 4 11 e747 e750 10.1016/S2665-9913(22)00221-1 36034738
2. Johnson AG Amin AB Ali AR Hoots B Cadwell BL Arora S Avoundjian T Awofeso AO Barnes J Bayoumi NS COVID-19 incidence and death rates among unvaccinated and fully vaccinated adults with and without booster doses during periods of delta and omicron variant emergence—25 US jurisdictions, April 4-December 25, 2021 MMWR Morb Mortal Wkly Rep 2022 71 4 132 138 10.15585/mmwr.mm7104e2 35085223
3. Kared H Wolf AS Alirezaylavasani A Ravussin A Solum G Tran TT Lund-Johansen F Vaage JT Nissen-Meyer LS Nygaard UC Immune responses in Omicron SARS-CoV-2 breakthrough infection in vaccinated adults Nat Commun 2022 13 1 4165 10.1038/s41467-022-31888-y 35851055
4. Mahardika GN Mahendra NB Mahardika BK Suardana IBK Pharmawati M Annotating spike protein polymorphic amino acids of variants of SARS-CoV-2 Including Omicron Biochem Res Int 2022 2022 2164749 10.1155/2022/2164749 35450296
5. Meo SA Meo AS Al-Jassir FF Klonoff DC Omicron SARS-CoV-2 new variant: global prevalence and biological and clinical characteristics Eur Rev Med Pharmacol Sci 2021 25 24 8012 8018 34982465
6. Wu F Zhao S Yu B Chen YM Wang W Song ZG Hu Y Tao ZW Tian JH Pei YY A new coronavirus associated with human respiratory disease in China Nature 2020 579 7798 265 269 10.1038/s41586-020-2008-3 32015508
7. Jin Y Yang H Ji W Wu W Chen S Zhang W Duan G Virology, epidemiology, pathogenesis, and control of COVID-19 Viruses 2020 12 4 10.3390/v12040372
8. Jungreis I Sealfon R Kellis M SARS-CoV-2 gene content and COVID-19 mutation impact by comparing 44 Sarbecovirus genomes Nat Commun 2021 12 1 2642 10.1038/s41467-021-22905-7 33976134
9. Yang D Leibowitz JL The structure and functions of coronavirus genomic 3' and 5' ends Virus Res 2015 206 120 133 10.1016/j.virusres.2015.02.025 25736566
10. Redondo N Zaldivar-Lopez S Garrido JJ Montoya M SARS-CoV-2 accessory proteins in viral pathogenesis: knowns and unknowns Front Immunol 2021 12 708264 10.3389/fimmu.2021.708264 34305949
11. Silvas JA Vasquez DM Park JG Chiem K Allue-Guardia A Garcia-Vilanova A Platt RN Miorin L Kehrer T Cupic A Contribution of SARS-CoV-2 accessory proteins to viral pathogenicity in K18 human ACE2 transgenic mice J Virol 2021 95 17 e0040221 10.1128/JVI.00402-21 34133899
12. Verma R Saha S Kumar S Mani S Maiti TK Surjit M RNA-Protein interaction analysis of SARS-CoV-2 5' and 3' untranslated regions reveals a role of lysosome-associated membrane protein-2a during viral infection MSystems 2021 6 4 e0064321 10.1128/mSystems.00643-21 34254825
13. Tamura K Stecher G Kumar S MEGA11: molecular evolutionary genetics analysis version 11 Mol Biol Evol 2021 38 7 3022 3027 10.1093/molbev/msab120 33892491
14. Chatterjee S Kim CM Lee YM Seo JW Kim DY Yun NR Kim DM Whole-genome analysis and mutation pattern of SARS-CoV-2 during first and second wave outbreak in Gwangju, Republic of Korea Sci Rep 2022 12 1 11354 10.1038/s41598-022-14989-y 35790838
15. Suardana IBK, Mahardika BK, Pharmawati M, Sudipa PH, Sari TK, Mahendra NB, Mahardika GN (2022) Whole-genome comparison of representatives of all variants of SARS-CoV-2, including subvariant BA.2 and the GKA clade. Research Square
16. Cubuk J Alston JJ Incicco JJ Singh S Stuchell-Brereton MD Ward MD Zimmerman MI Vithani N Griffith D Wagoner JA The SARS-CoV-2 nucleocapsid protein is dynamic, disordered, and phase separates with RNA BioRxiv 2020 2 2
17. Koetzner CA Parker MM Ricard CS Sturman LS Masters PS Repair and mutagenesis of the genome of a deletion mutant of the coronavirus mouse hepatitis virus by targeted RNA recombination J Virol 1992 66 4 1841 1848 10.1128/jvi.66.4.1841-1848.1992 1312608
18. Goh KC Tang CK Norton DC Gan ES Tan HC Sun B Syenina A Yousuf A Ong XM Kamaraj US Molecular determinants of plaque size as an indicator of dengue virus attenuation Sci Rep 2016 6 26100 10.1038/srep26100 27185466
19. Blaney JE Jr Johnson DH Manipon GG Firestone CY Hanson CT Murphy BR Whitehead SS Genetic basis of attenuation of dengue virus type 4 small plaque mutants with restricted replication in suckling mice and in SCID mice transplanted with human liver cells Virology 2002 300 1 125 139 10.1006/viro.2002.1528 12202213
20. Bal A Simon B Destras G Chalvignac R Semanas Q Oblette A Queromes G Fanget R Regue H Morfin F Detection and prevalence of SARS-CoV-2 co-infections during the Omicron variant circulation in France Nat Commun 2022 13 1 6316 10.1038/s41467-022-33910-9 36274062
21. Rosenberg SM Maloy S Hughes K Reverse mutation Brenner's encyclopedia of genetics 2013 2 New York Academic Press 220 221
22. Cann AJ Cann AJ Chapter 3—genomes Principles of molecular virology 2012 5 New York Academic Press 55 101
23. Su YCF Anderson DE Young BE Linster M Zhu F Jayakumar J Zhuang Y Kalimuddin S Low JGH Tan CW Discovery and genomic characterization of a 382-nucleotide deletion in ORF7b and ORF8 during the early evolution of SARS-CoV-2 MBio 2020 11 4 10.1128/mBio.01610-20
24. Vinjamuri S Li L Bouvier M SARS-CoV-2 ORF8: one protein, seemingly one structure, and many functions Front Immunol 2022 13 1035559 10.3389/fimmu.2022.1035559 36353628
25. Li JY Liao CH Wang Q Tan YJ Luo R Qiu Y Ge XY The ORF6, ORF8 and nucleocapsid proteins of SARS-CoV-2 inhibit type I interferon signaling pathway Virus Res 2020 286 198074 10.1016/j.virusres.2020.198074 32589897
26. Lin X Fu B Yin S Li Z Liu H Zhang H Xing N Wang Y Xue W Xiong Y ORF8 contributes to cytokine storm during SARS-CoV-2 infection by activating IL-17 pathway Science 2021 24 4 102293
|
PMC010xxxxxx/PMC10165869.txt |
==== Front
Psychiatry Res
Psychiatry Res
Psychiatry Research
0165-1781
1872-7123
Elsevier/North-Holland Biomedical Press
S0165-1781(23)00198-1
10.1016/j.psychres.2023.115248
115248
Article
National trends in psychotropic medication prescribing before and during the COVID-19 pandemic☆
Sanborn Molly a
Ali Mir M. b⁎
Creedon Timothy B. c
a National Mental Health & Substance Abuse Policy Laboratory, Substance Abuse & Mental Health Services Administration, USA
b Office of the Assistant Secretary for Planning & Evaluation, US Department of Health & Human Services, 200 Independence Avenue SW, Washington DC 20202, USA
c Office of the Assistant Secretary for Planning & Evaluation, US Department of Health & Human Services, USA
⁎ Corresponding author.
8 5 2023
7 2023
8 5 2023
325 115248115248
5 2 2023
5 5 2023
7 5 2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The early months of the COVID-19 pandemic saw a decline in psychotropic medication use; however, little is known about how this trend evolved as the pandemic progressed and how it varied across different payers in the United States. Using a national multi-payer pharmacy claims database and adopting a quasi-experimental research design, this study examines trends in psychotropic medication prescriptions dispensed from July 2018 - June 2022. The study finds that the number of patients with dispensed psychotropic medications and the number of psychotropic medications dispensed declined during the early months of the pandemic but experienced a statistically significant growth in later periods compared to the pre-pandemic rate. Average days supply of psychotropic medications dispensed increased significantly throughout the pandemic. Commercial insurance remained the primary payer for psychotropic medication during the pandemic, but there was a significant increase in the number of prescription fills covered under Medicaid. This implies that public insurance programs played an increasing role in financing psychotropic medication use during the COVID-19 pandemic.
Keywords
Psychotropic medication
Mental health
COVID-19
==== Body
pmc1 Introduction
Since the start of the COVID-19 pandemic, there have been concerns about its potential adverse impacts on the incidence of mental health conditions, the experiences of people with mental health conditions, and access to treatment (Holmes et al., 2020). Individuals with mental health conditions may be disproportionately affected by COVID-19, both directly and indirectly. For example, individuals with serious mental illness have a higher prevalence of health conditions associated with severe COVID-19 than those without serious mental illness and may be at increased risk for contracting COVID-19 (Novak et al., 2021). Social isolation and economic stress during the pandemic may have led to or exacerbated anxiety and depression, and individuals with mental health conditions likely faced additional barriers to accessing needed treatment services and other supports (Druss 2020; Holmes et al., 2020).
Studies conducted early in the pandemic found significant proportions of patients diagnosed with COVID-19 subsequently reported symptoms of mental illness, including post-traumatic stress disorder (PTSD), depression, and anxiety (Bo et al., 2020; Mazza et al., 2020). Research has also found an increased incidence of psychiatric disorders following COVID-19 diagnosis among individuals with no prior psychiatric history (Taquet et al., 2021). Together, these factors have raised concerns about an increasing burden of mental health conditions since the COVID-19 pandemic.
At the same time as the burden of mental health conditions has been increasing, the treatment delivery and health insurance coverage landscapes have both changed substantially. The COVID-19 pandemic resulted in a significant shift in the delivery of mental health services. The Coronavirus Aid, Relief, and Economic Security (CARES) Act passed in March 2020 included provisions that expanded telehealth coverage for Medicare and Medicaid beneficiaries, including reimbursement at rates equivalent to those for in-person services. The early months of the pandemic saw a significant increase in telehealth services (Demeke et al., 2021), especially for mental health services, where telehealth accounted for approximately 48% of all outpatient services delivered (Zhu et al., 2022). What has remained less clear is how psychotropic medication (such as antidepressants, antipsychotics, mood stabilizers, stimulants, etc.) prescribing may have changed as part of this shift in mental health care delivery .
Research covering the early months of the pandemic found a decline in the number of patients that were prescribed psychotropic medication as well as a decline in the number of psychotropic medications prescribed (Nason et al., 2021; Hirschtritt et al., 2021; Leong et al., 2022), maybe because of initial difficulties in arranging access to telehealth providers. Nason et al. (2021) found a 2.5–7.5 percent decline in psychotropic medications dispensed during the first five months of the pandemic using a national pharmacy claims database. Similarly, Hirschtritt et al. (2021) found a decline in psychotropic prescription fills (<2%) using pharmacy claims data from a Northern California private healthcare system. Leong et al. (2022) found an 8% decline in number of patients prescribed psychotropic medication during April - December 2020 using administrative data from the Manitoba province in Canada. However, it is not known if this declining trend continued in the later period of the pandemic. The literature has documented an increase in mental health conditions both during the early and the latter period of the pandemic (Holmes et al., 2020) – understanding trends in psychotropic medication fills in the later period of the pandemic might indicate whether treatment receipt was able to keep pace with the need for mental health services. It is also not known if the days supply of psychotropic medication changed during the pandemic (e.g. it is possible that days supply increased to counterbalance social distancing challenges) and whether there was a shift in the payer-type (commercial vs Medicaid vs Medicare). This is an important omission from the literature, as the pandemic saw a significant increase in public health insurance coverage and a decline in the uninsured rate, (Bundorf et al., 2021) and Medicaid continues to be the largest payer for mental health services in the US (CMS, 2022).
This study expands on the previous literature by adopting a quasi-experimental research design using a national multi-payer pharmacy claims database to examine trends in psychotropic medication prescriptions from July 2018 to June 2022 after accounting for the beginning of the COVID-19 pandemic in the U.S. in March 2020. Specifically, the current study reports monthly trends in the number of psychotropic prescriptions dispensed, the average days supply of psychotropic prescriptions, and how psychotropic prescriptions dispensed varied by payer-type (private insurance, Medicaid, Medicare and cash). The study also reports the number of patients prescribed psychotropic medication.
2 Methods
2.1 Data
The data for the analysis were drawn from IQVIA's National Prescription Audit (NPA), PayerTrak, and Total Patient Tracker (TPT) databases, covering prescriptions from July 2018 to June 2022. NPA includes over 3 billion prescriptions per year, representing more than 92% of prescriptions dispensed at retail pharmacies (chain, independent, and food store pharmacies and mail-orders) and covers all 50 states and the District of Columbia. We used NPA data to analyze the total number of psychotropic prescriptions dispensed per month and the average days’ supply for dispensed prescriptions per month (see Appendix 1 for the list of psychotropic medication names by class). Similar to NPA, PayerTrak also includes over 3 billion prescriptions per year and covers more than 92% of prescriptions dispensed in retail settings. Additionally, PayerTrak includes information on the source of payer for the prescriptions, enabling us to analyze how the number of psychotropic medications dispensed varied each month by payer type (private insurance, Medicaid, Medicare and cash payment). TPT data capture the total number of unique patients at the national level across all drugs and therapeutic classes in the retail outpatient setting. TPT eliminates duplicate patients and multiple prescription fills, allowing us to produce unique patient counts. We used TPT data to estimate the total number of patients per month who were prescribed psychotropic medications. IQVIA data sets are statistically deidentified and as such are exempt from the U.S. Department of Health and Human Services regulations that require institutional review board approval.
2.2 Analysis
We used interrupted time series (ITS) analysis, a quasi-experimental research design, to estimate changes in each psychotropic medication outcome associated with the declaration of 2020 COVID-19 PHE (March 2020). The following equation was used for the regression:Yt=β0+β1Tt+β2Xt+β3XtTt+β4Zt+εt,
where Yt is the outcome variable measured at each monthly time point t, Tt is the time since the start of observation (in months), Xt is a dummy (indicator) variable representing the COVID-19 pandemic (pre–March 2020 periods = 0, post-March 2020 periods = 1), XtTt is a timeperiod interaction term, Zt is a set of indicator variables to control for month length (28 or 29 and 30 days vs. 31 days), and εt is the error term. We specified the models to test for both a one-time change immediately when the COVID-19 pandemic began (intercept/level change, β2) and a difference in trends between the pre- and post-COVID-19 pandemic periods (slope change, β3). We estimated the models using ordinary least squares with Newey-West standard errors to account for autocorrelated error terms. All analyses were conducted in Stata/MP 17.
3 Results
Fig. 1 (observed data points and unadjusted model predictions) and Table 1 (adjusted model estimates) show that the change in the number of patients filling a medication was not significantly different than zero. The unadjusted trend in Fig. 1 appears negative, while the estimate from the adjusted ITS model was +5380 patients/month (p = 0.392). In the adjusted ITS model, we did not detect a significant, immediate shift in the number of patients with psychotropic prescription fills per month at the start of the pandemic in March 2020 (+76,990 patients, p = 0.476). During the pandemic, the number of patients with filled psychotropic medications increased at a statistically significant rate (+20,110 patients/month, p = 0.003), adjusting for month length. This rate, however, was not significantly different than the adjusted pre-COVID rate (+14,730 patients/month, p = 0.165).Fig. 1 Monthly number of patients with psychotropic prescription fills, July 2018 to June 2022 Notes: Obs., observed; est., estimated. Estimates were derived from an unadjusted Interrupted Time Series model accounting for changes in level and slope. Shaded areas indicate 95% confidence intervals. Data Source: IQVIA Total Patient Tracker.
Fig 1
Table 1 Psychotropic Prescription Fills July 2018 – June 2022: Total Fills, and Prescription length.
Table 1 Patients Prescriptions Rx Days
Coef. SE P Coef. SE P Coef. SE P
Intercept 17,020.06 67.06 <0.001 41,412.48 230.81 <0.001 32.50 0.14 <0.001
Slopes
Pre-COVID 5.38 6.22 0.392 5.04 22.28 0.822 0.13 0.01 <0.001
Post-COVID 20.11 6.31 0.003 53.83 21.63 0.017 −0.00 0.01 0.813
Difference 14.73 10.43 0.165 48.79 37.29 0.198 −0.13 0.02 <0.001
Level
Difference 76.99 107.14 0.476 501.16 361.44 0.173 0.88 0.24 0.001
Days per month (ref. = 31)
29 −919.74 94.22 <0.001 −3376.52 314.60 <0.001
30 −386.00 93.53 <0.001 −1365.47 339.13 <0.001
N 48 48 48
Lag 2 3 1
Notes: N = 48 for all models. Linear regressions with Newey-West standard errors to account for autocorrelation.
Data Source: IQVIA Total Patient Tracker and National Prescription Audit.
The observed and predicted values of monthly psychotropic prescriptions dispensed are presented in Fig. 2 (predicted values based on unadjusted ITS models). Adjusted ITS results are also shown in Table 1. In general, filled prescriptions followed a similar pattern to our patient level outcome. Fig. 2 appears to show a pre-COVID decrease from about 41.9 million prescriptions to about 38.1 million at the start of the pandemic. After controlling for month length, however, there was not a significant pre-COVID trend in either direction (+5040 fills/month, p = 0.822). During the pandemic, the number of psychotropic medication fills each month did increase at a significant adjusted rate (+53,830 fills/month, p = 0.017), but this was not significantly different than the adjusted pre-period trend. We also did not observe a statistically significant immediate level change at the start of the pandemic.Fig. 2 Monthly number of psychotropic prescription fills, July 2018 to June 2022 Notes: Obs., observed; est., estimated. Estimates were derived from an unadjusted Interrupted Time Series model accounting for changes in level and slope. Shaded areas indicate 95% confidence intervals. Data Source: IQVIA National Prescription Audit.
Fig 2
Fig. 3 shows that the average days supply of psychotropic prescriptions dispensed each month was about 33–34 days in the pre-pandemic period and about 36 days during the pandemic. ITS model-based estimates (Table 1) indicate that the average length of a prescription was significantly increasing by about 0.13 days per month (p<0.001) in the pre-period. At the beginning of the pandemic there was a significant, immediate level shift of about 0.88 days (p<0.001), but during the pandemic the increasing trend did not continue (during pandemic: 0.00 days/month, p = 0.813;).Fig. 3 Average days supply of psychotropic prescription fills, July 2018 to June 2022 Notes: Obs., observed; est., estimated. Estimates were derived from an unadjusted Interrupted Time Series model accounting for changes in level and slope. Shaded areas indicate 95% confidence intervals. Data Source: IQVIA National Prescription Audit.
Fig 3
Fig. 4 displays trends in psychotropic prescription fills by payer. Commercial insurance was the primary payer of psychotropic medications during the pre-pandemic period (∼ 50% of all psychotropic prescriptions). This continued after the pandemic with commercial insurance paying for about 20–25 million fills per month, accounting for almost 50% of all psychotropic prescriptions. Medicare was the second-leading payer for psychotropic medications filled in retail pharmacies, paying for about 10–11 million fills a month. Medicaid paid for about 7–8 million fills a month), and last, cash (fully out-of-pocket) payments accounted for about 1–2 million fills per month. Table 2 shows ITS model-based estimates of levels and rates of change before and during the pandemic. For fills paid by commercial insurance, there were no significant changes associated with the pandemic, but there was a significant, increasing trend of about 34,480 additional fills per month during the pandemic period (p = 0.018). For psychotropic medications paid for by Medicaid, there was a statistically significant increase in level in March 2020 of about 320,370 prescriptions per month (p<0.001) and a significant change in slope by about 21,310 prescriptions per month from before to during the pandemic (p<0.001). There was no statistically significant change in the slope for fills paid by Medicare . Prescriptions for psychotropic medications paid for in cash were already declining during the pre-pandemic period (an average of 11,930 fewer per month, p<0.001) and continued to decline at a statistically significant rate of about −7450 fewer fills per month during the pandemic (p<0.001) compared to pre-pandemic.Fig. 4 Monthly number of psychotropic prescriptions fills by payer, July 2018 to June 2022 Data Source: IQVIA National Sales Perspective.
Fig 4
Table 2 Psychotropic Prescription Fills by Payer Type July 2018 – June 2022.
Table 2 Commercial Medicaid Medicare Cash
Coef. SE P Coef. SE P Coef. SE P Coef. SE P
Intercept 22,214.32 107.72 <0.001 7045.63 42.45 <0.001 10,496.24 77.87 <0.001 1656.29 21.53 <0.001
Slopes
Pre-COVID 15.95 11.38 0.169 5.73 2.98 0.061 −4.71 7.53 0.535 −11.93 1.92 <0.001
Post-COVID 34.48 13.95 0.018 27.04 4.06 <0.001 −0.23 5.46 0.966 −7.45 1.01 <0.001
Difference 18.53 21.51 0.394 21.31 4.92 <0.001 4.48 10.97 0.685 4.47 2.25 0.053
Level
Difference −54.16 237.09 0.820 320.37 63.58 <0.001 265.40 98.47 0.010 −30.45 29.36 0.306
Days per month (ref. = 31)
29 −1896.34 147.01 <0.001 −519.03 65.30 <0.001 −806.98 103.04 <0.001 −154.17 19.42 <0.001
30 −756.60 170.44 <0.001 −201.63 66.25 0.004 −350.07 98.36 0.001 −57.17 15.92 0.001
N 48 48 48 48
Lag 2 4 3 1
Notes: N = 48 for all models. Linear regression with Newey-West standard errors to account for autocorrelation.
Data Source: IQVIA National Sales Perspective.
Stratifying our analysis of the number of patients with psychotropic prescription fills across four therapeutic classes (antidepressants, anxiolytics/sedatives, antipsychotics, and stimulants), we observed similar patterns (Appendix Table 2 ). While there was some gradual growth across the entire study period, there were not consistent changes from before to during the pandemic.Table A2 . Number of Patients with Psychotropic Prescription Fills by Therapeutic Class July 2018 – June 2022.
Table A2 Antidepressants Anxiolytics/Sedatives Antipsychotics Stimulants
Coef. SE P Coef. SE P Coef. SE P Coef. SE P
Intercept 17,461.39 64.13 <0.001 11,093.30 49.83 <0.001 2585.35 10.26 <0.001 2630.53 31.73 <0.001
Slopes
Pre-COVID 25.18 7.65 0.002 −14.07 4.01 0.001 6.82 0.95 <0.001 13.01 2.43 <0.001
Post-COVID 42.21 7.50 <0.001 −2.56 3.71 0.493 6.45 0.80 <0.001 19.92 2.23 <0.001
Difference 17.03 12.19 0.170 11.50 5.85 0.056 −0.37 1.27 0.770 6.91 3.33 0.044
Level
Difference −48.17 157.07 0.761 173.73 73.70 0.023 51.33 19.57 0.012 −196.56 46.22 <0.001
Days per month (ref. = 31)
29 −973.22 120.98 <0.001 −655.78 59.97 <0.001 −118.00 15.45 <0.001 −136.76 25.82 <0.001
30 −397.90 107.42 0.001 −283.81 65.64 <0.001 −52.77 14.88 0.001 −57.82 16.91 0.001
N 48 48 48 48
Lag 4 4 4 4
Notes: Linear regressions with Newey-West standard errors to account for autocorrelation. Data Source: IQVIA Total Patient Tracker.
4 Discussion
Using a national multi-payer pharmacy claims database, this study found that the trends in the number of patients with psychotropic medication fills, and the number of fills declined during the early period of the pandemic but experienced a significant growth during the later period and exceeded the pre-pandemic trends. Average days supply of psychotropic medications dispensed increased significantly from about 33 days in the pre-pandemic period to about 36 days during the pandemic. Commercial insurance remained the primary payer for psychotropic medication during the pandemic, but there was a significant increase in the number of prescription fills covered under Medicaid of about 21,000 fills per month above and beyond the pre-pandemic trend. Though prior studies documented a small decline in psychotropic medications dispensed in the early months of the pandemic (Nason et al., 2021; Hirschtritt et al., 2021; Leong et al., 2022), our results show that, as the pandemic continued and in the context of a rising prevalence of mental health conditions, psychotropic medication use grew at statistically significant rates between March 2020 and June 2022.
The COVID-19 pandemic prompted numerous policy changes in the mental health treatment delivery system at both the federal and state level. These changes include flexibilities around prescriptions via telehealth, including increased reimbursement and coverage for telehealth services, and home delivery options for patients, expedited licensing and lowering barriers to interstate practice for providers among others (Moreno et al., 2020). An evaluation of the impacts of these specific polices implemented during COVID-19 on the initiation and continuation of psychotropic medication use for mental health conditions will be an important avenue for future research to explore.
Psychotropic medication and psychotherapy are both effective in treating most mental health conditions. There is some evidence that combined treatments may be more effective than each of these treatments alone (NIMH, 2023). The literature also shows that receipt of psychotherapy enhances adherence to medication (NIMH, 2023). However, a significant proportion of patients with mental health conditions receive psychotropic medication without any psychotherapy (Olfson and Marcus, 2010). An important avenue for future research to consider would be to examine how the COVID-19 pandemic has impacted the use of psychotherapy in conjunction with and without psychotropic medication.
The study also found a significant increase in psychotropic medication being paid for by Medicaid during the pandemic, although private insurance still remained the primary payer for psychotropic medications. This finding is consistent with the literature that has documented a higher enrollment in public health insurance programs after the initial shock to employment in March 2020 (Bundorf et al., 2021) Even before the pandemic, Medicaid was a major payer for behavioral health services (CMS, 2021), but the pandemic may have accelerated this growth. To our knowledge, most of the changes in Medicaid policies during our study period are related to the COVID-19 PHE. The unique nature of the pandemic, and the economic downturn it caused, prompted many states to adopt Medicaid emergency authorization to expand eligibility and/or modify eligibility rules, eliminate/waive premiums, and streamline the application and enrollment processes. All of this might help to explain the growing share of psychotropic medication being covered by Medicaid.
4.1 Limitations
Despite the comprehensive nature and timeliness of the data, the findings of this study should be viewed in the context of some limitations. First, the study examined psychotropic medications in general, but it was not coupled with electronic medical records to determine if their use was for a psychiatric condition. It is also important to note that IQVIA does not include federal sources such as Veterans Affairs and Indian Health Services. Second, race/ethnicity of the patients were not available in the IQVIA data sets used in the study. The literature has shown the impact of the COVID-19 pandemic to be more pronounced among people of color (Shim and Starks, 2021), and examining racial/ethnic differences in psychotropic medications during the pandemic will be an important direction for future studies to pursue. Third, our study was not able to distinguish between psychotropic medication initiation vs continuation. Examining treatment initiation and continuation will have important policy implications as the pandemic continues to evolve. Finally, a methodological limitation of our ITS estimation strategy is the lack of a non-treatment or a control group. In quasi-experimental research design, it is a common practice to include a control group; however, finding a suitable control group in our case was difficult because COVID-19 impacted almost every aspect of the US health care system. Thus, it is prudent to view our results as demonstrating strong associations rather than causal relationships.
4.2 Conclusions/Implications
Expanding and maintaining access to mental health treatment has been a key priority in the federal and state response to the COVID-19 pandemic. The results of our study show that the use of psychotropic medication remained stable during the pandemic with public insurance programs playing an important role in helping patients to pay for their treatment. However, it is still not known whether the gap in unmet need for mental health treatment narrowed or increased.
Declaration of Competing Interest
None.
Appendix A
Table A2
Appendix B Supplementary materials
Image, application 1
☆ Disclaimer: The views expressed here are those of the authors and do not necessarily reflect the views of the Office of the Assistant Secretary for Planning & Evaluation, Substance Abuse & Mental Health Services Administration, or US Department of Health & Human Services.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.psychres.2023.115248.
==== Refs
References
Bo H.-X. Li W. Yang Y. Posttraumatic stress symptoms and attitude toward crisis mental health services among clinically stable patients with COVID-19 in China Psychol Med. Published online 2020 1 2
Bundorf M.K. Gupta S. Kim C. Trends in US health insurance coverage during the COVID-19 pandemic JAMA Health Forum 2 9 2021 e212487 10.1001/jamahealthforum.2021.2487
Center for Medicare and Medicaid Services. Behavioral Health Services. https://www.medicaid.gov/medicaid/benefits/behavioral-health-services/index.html Accessed October 26, 2022.
Demeke H.B. Merali S. Marks S. Trends in use of telehealth among health centers during the COVID-19 Pandemic — United States June 26–November 6, 2020 MMWR Morb. Mortal Wkly. Rep. 70 2021 240 244 10.15585/mmwr.mm7007a3 33600385
Druss B.G. Addressing the COVID-19 pandemic in populations with serious mental illness JAMA Psychiatry 77 9 2020 891 892 32242888
Hirschtritt M.E. Slama N. Sterling S.A. Olfson M. Iturralde E. Psychotropic medication prescribing during the COVID-19 pandemic Medicine (Baltimore) 100 43 2021 e27664 10.1097/MD.0000000000027664 34713861
Holmes E.A. O'Connor R.C. Perry V.H. Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science Lancet Psychiatry 7 6 2020 547 560 32304649
Leong C. Kowalec K. Eltonsy S. Psychotropic medication use before and during COVID-19: a population-wide study Front. Pharmacol. 13 2022 886652 10.3389/fphar.2022.886652 Published 2022 Apr 27
Mazza M.G. De Lorenzo R. Conte C. Anxiety and depression in COVID-19 survivors: role of inflammatory and clinical predictors Brain Behav. Immun. 89 2020 594 600 32738287
Moreno C. Wykes T. Galderisi S. How mental health care should change as a consequence of the COVID-19 pandemic [published correction appears in Lancet Psychiatry. 2021 Jul;8(7):e16] Lancet Psychiatry 7 9 2020 813 824 10.1016/S2215-0366(20)30307-2 32682460
Nason I. Stein D.T. Frank R.G. Stein M.B. Decline In New Starts Of Psychotropic Medications During The COVID-19 Pandemic Health Aff. (Millwood). 40 6 2021 904 909 10.1377/hlthaff.2021.00028 34097524
National Institute of Mental Health. Mental Health Medication. https://www.nimh.nih.gov/health/topics/mental-health-medications Accessed May 1, 2023.
Novak P. Sanmartin M.X. Ali M.M. Chen J. Health conditions associated with severe illness from COVID-19 among individuals with serious mental illness Psychiatr. Serv. 72 4 2021 468 469 33208029
Olfson M. Marcus S.C. National trends in outpatient psychotherapy Am. J. Psychiatry 167 12 2010 1456 1463 10.1176/appi.ajp.2010.10040570 DecEpub 2010 Aug 4. PMID: 20686187 20686187
Shim R.S. Starks S.M. COVID-19, structural racism, and mental health inequities: policy implications for an emerging syndemic Psychiatr. Serv. 72 10 2021 1193 1198 10.1176/appi.ps.202000725 33622042
Taquet M. Geddes J.R. Husain M. Luciano S. Harrison P.J. 6-month neurological and psychiatric outcomes in 236,379 survivors of COVID-19: a retrospective cohort study using electronic health records Lancet Psychiatry 8 5 2021 416 427 33836148
Zhu J.M. Myers R. McConnell K.J. Levander X. Lin S.C. Trends In Outpatient Mental Health Services Use Before And During The COVID-19 Pandemic Health Aff (Millwood) 41 4 2022 573 580 10.1377/hlthaff.2021.01297 35377763
|
PMC010xxxxxx/PMC10166058.txt |
==== Front
Diabetologie
Die Diabetologie
2731-7447
2731-7455
Springer Medizin Heidelberg
1041
10.1007/s11428-023-01041-4
DDG Praxisempfehlungen
Empfehlungen zur Ernährung von Personen mit Typ-2-Diabetes mellitus
Dietary recommendations for persons with type 2 diabetes mellitusSkurk Thomas skurk@tum.de
1217
Bosy-Westphal Anja 317
Grünerbel Arthur 417
Kabisch Stefan 5617
Keuthage Winfried 717
Kronsbein Peter 817
Müssig Karsten 917
Nussbaumer Helmut 1017
Pfeiffer Andreas F. H. 1117
Simon Marie-Christine 1217
Tombek Astrid 1317
Weber Katharina S. 1417
Rubin Diana 151617
1 grid.6936.a 0000000123222966 ZIEL – Institute for Food & Health, Technische Universität München, Gregor-Mendel-Str. 2, 85354 Freising, Deutschland
2 grid.6936.a 0000000123222966 Else Kröner-Fresenius-Zentrum für Ernährungsmedizin, Technische Universität München, Freising, Deutschland
3 grid.9764.c 0000 0001 2153 9986 Institut für Humanernährung, Agrar- und Ernährungswissenschaftliche Fakultät, Christian-Albrechts-Universität zu Kiel, Kiel, Deutschland
4 Diabeteszentrum München Süd, München, Deutschland
5 grid.418213.d 0000 0004 0390 0098 Deutsches Institut für Ernährungsforschung Potsdam-Rehbrücke, Potsdam, Deutschland
6 grid.452622.5 Deutsches Zentrum für Diabetesforschung (DZD), München, Deutschland
7 Schwerpunktpraxis für Diabetes und Ernährungsmedizin, Münster, Deutschland
8 grid.440943.e 0000 0000 9422 7759 Fachbereich Oecotrophologie, Hochschule Niederrhein, Campus Mönchengladbach, Mönchengladbach, Deutschland
9 grid.415033.0 0000 0004 0558 1086 Klinik für Innere Medizin, Gastroenterologie und Diabetologie, Niels-Stensen-Kliniken, Franziskus-Hospital Harderberg, Georgsmarienhütte, Deutschland
10 Diabeteszentrum Burghausen, Burghausen, Deutschland
11 grid.6363.0 0000 0001 2218 4662 Abt. Endokrinologie, Diabetes und Ernährungsmedizin, Charité Universitätsmedizin Berlin, Berlin, Deutschland
12 grid.10388.32 0000 0001 2240 3300 Institut für Ernährungs- und Lebensmittelwissenschaften, Rheinische Friedrich-Wilhelms-Universität Bonn, Bonn, Deutschland
13 Diabetes-Klinik Bad Mergentheim, Bad Mergentheim, Deutschland
14 grid.9764.c 0000 0001 2153 9986 Institut für Epidemiologie, Christian-Albrechts-Universität zu Kiel, Kiel, Deutschland
15 grid.433867.d 0000 0004 0476 8412 Vivantes Klinikum Spandau, Berlin, Deutschland
16 Vivantes Humboldt Klinikum, Berlin, Deutschland
17 grid.483797.6 0000 0001 1017 3069 Ausschuss Ernährung DDG, Berlin, Deutschland
8 5 2023
2023
19 4 482512
14 3 2023
© The Author(s), under exclusive licence to Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
issue-copyright-statement© Springer Medizin Verlag GmbH, ein Teil von Springer Nature 2023
==== Body
pmcInfobox DDG-Praxisempfehlungen Download
Auf der Webseite der Deutschen Diabetes Gesellschaft (https://www.deutsche-diabetes-gesellschaft.de/behandlung/leitlinien) befinden sich alle PDFs zum kostenlosen Download.
Präambel
Diese Praxisempfehlung richtet sich an alle Berufsgruppen, die Menschen mit Typ-2-Diabetes mellitus (T2Dm) betreuen. Neben den vielgestaltigen Aspekten der Ernährung bei Diabetes wird insbesondere eine Individualisierung von Therapie, Beratung, Empowerment und Diabetes-Selbstmanagement [1–3] gefordert. Daher hat sich der Ausschuss Ernährung der DDG das Ziel gesetzt, Praxisempfehlungen zur Ernährung möglichst zielgruppenspezifisch mit der höchsten verfügbaren Evidenz zusammenzutragen. Dabei wird eine nach Behandlungsformen getrennte Darstellung für erforderlich erachtet, da sich die therapeutische Bedeutung der Ernährung jeweils deutlich unterscheidet und vor dem Hintergrund unterschiedlicher medikamentöser Therapiekomponenten gesehen werden muss.
Charakteristisch für den Typ-2-Diabetes ist sein progressiver Verlauf im Sinne einer individuell unterschiedlich schnell voranschreitenden β‑Zell-Insuffizienz [4–7]. Vor diesem Hintergrund weisen Patienten mit Typ-2-Diabetes ganz unterschiedliche Charakteristika und Behandlungsformen auf [8].
Unter besonderen Lebensumständen, z. B. Sarkopenie und Pflegebedürftigkeit, ist die Ernährung unter starker Berücksichtigung persönlicher Vorlieben und unter Betonung der Deckung des Proteinbedarfs zu gestalten.
Insgesamt muss demzufolge die Ernährungstherapie stark individualisiert werden, um ihr Potenzial voll auszuschöpfen.
Die Möglichkeit der individualisierten Ernährungsberatung, auch per Telemedizin, sollte daher bei Menschen mit T2Dm stärker und intensiver genutzt werden. Allgemeine Ziele sind die Förderung ausgewogener Essgewohnheiten, Schulungen zu angemessenen Portionsgrößen und das Eingehen auf individuelle Ernährungsbedürfnisse, wobei die Freude am Essen erhalten bleibt und praktische Hilfsmittel für die Planung von Mahlzeiten bereitgestellt werden. Individualisierte Ernährungsberatungen haben evidenzbasierte Themen zum Inhalt, die durch qualifizierte und entsprechend zertifizierte Ernährungsfachkräfte durchgeführt werden sollen (Diätassistent/-in oder Ernährungswissenschaftler/-in oder Ökotrophologe/Ökotrophologin).
Der Ernährungstherapieplan muss auch mit der gesamten Managementstrategie einschließlich der Verwendung von Medikamenten, körperlicher Aktivität usw. koordiniert und laufend abgestimmt werden.
Darüber hinaus sollten Menschen mit Prädiabetes und Übergewicht/Adipositas an ein intensives Lebensstilinterventionsprogramm überwiesen werden, das individuelle Zielsetzungskomponenten umfasst, wie z. B. durch die S3-Leitlinie Prävention und Therapie der Adipositas definiert. Da diese Leistung bisher noch keine Regelleistung der gesetzlichen Krankenversicherung ist, sollte zumindest eine individualisierte Ernährungsberatung mit teilweiser Kostenübernahme nach § 43 SGB erfolgen.
Eine weitere wichtige Empfehlung ist die Überweisung von Erwachsenen mit Diabetes zu einer umfassenden Diabetes-Selbstmanagementschulung und -unterstützung (DSMES) gemäß den nationalen Standards.
Die vorliegende Praxisempfehlung stellt die Zusammenfassung und Bewertung der Literatur durch den Ausschuss Ernährung der DDG zu ausgewählten ernährungstherapeutischen Aspekten in der Behandlung des T2Dm dar. Diese werden regelmäßig aktualisiert und ggf. ergänzt. Dabei wurde die Evidenz – soweit verfügbar – im Rahmen einer Literaturrecherche basierend auf systematischen Reviews oder Metaanalysen bewertet. Zu Themen ohne die Verfügbarkeit solcher Übersichten wurden auch Originalarbeiten verwendet.
Empfehlungen für das Körpergewicht
Allgemeine Empfehlungen
Empfehlung.
Bei Übergewicht soll im Allgemeinen eine Gewichtsreduktion angestrebt werden.
„Weight cycling“ sollte vermieden werden.
Kommentar
Abhängig vom Alter ist eine Gewichtszunahme, die zu einem jeweils 5 Punkte höheren BMI führt, mit einem 3‑fach (Gewichtszunahme zwischen 18 und 24 Jahren) oder 2‑fach (Gewichtszunahme ≥ 25 Jahren) höheren Risiko für T2Dm assoziiert [9]. Bereits Adipositas allein ist ein eigenständiger Risikofaktor auch für KHK. Eine moderate Gewichtsreduktion hingegen (5–10 % vom aktuellen Gewicht) vermindert Risiken wie Insulinresistenz, Hyperglykämie und Dyslipidämie [10]. Dadurch lassen sich Folgekomplikationen vermindern. Eine „very-low-calorie diet“ (VLCD; 624 kcal/Tag) über 8 Wochen kann zudem zu einer vorübergehenden Diabetesremission von mindestens 6 Monaten führen [11]. Die Effektivität einer VLCD-Diät ist größer bei einer kürzeren Diabetesdauer sowie bei höheren Nüchtern-Insulin- und C‑Peptidwerten [12]. Ein intensives Gewichtsmanagement führt auch mit einer Mischkost und Lebensstilintervention zu einer nachhaltigen Remission [13]. Ein stabiles Körpergewicht scheint dabei mit einem besseren kardiovaskulären Outcome verbunden zu sein als eine hohe Gewichtsvariabilität [14–16]. Gewichtszunahmen oder Gewichtsschwankungen bei T2Dm sind mit einer höheren Sterblichkeit assoziiert [15, 17].
Jedoch gehen gerade bei älteren Patienten größere Gewichtsabnahmen (> 25 %) mit einem Verlust der Muskelmasse einher [18]. Studien zeigen außerdem, dass Normalgewichtige mit T2Dm eine höhere Sterblichkeit aufweisen als Personen mit einem höheren Körpergewicht [19, 20], was als Adipositas-Paradox mehrfach beschrieben wurde [21]. Eine mögliche Erklärung für diesen Effekt ist eine größere, metabolisch aktivere Muskelmasse bei übergewichtigen Patienten [22]; dies muss bei den Gewichtszielen mit bedacht werden und ggf. bei einem körperlichen Bewegungsprogramm zum Muskelerhalt mit berücksichtigt werden [23].
Quantitative Aussagen zur angestrebten Gewichtsreduktion, Diabetesremission
Empfehlung.
Das Ausmaß der Gewichtsreduktion orientiert sich an den individuellen Therapiezielen. Für eine Diabetesremission sollte eine Gewichtsreduktion von 15 kg des Ausgangsgewichts bei Übergewicht/Adipositas angestrebt werden.
Kommentar
Die Assoziation der Adipositas mit allen Komponenten des metabolischen Syndroms macht die Gewichtsreduktion zu einem vorrangigen Therapieziel. Der übliche und realistische Konsens war eine Gewichtsreduktion um 3–5 kg im Kontext einer Umstellung des Ernährungs- und Bewegungsverhaltens. Das Erreichen dieser Ziele erlaubte eine Reduktion der T2Dm-Manifestation um etwa 60 % bei Menschen mit Prädiabetes und ist in großen Studien belegt [24]. Eine größere Gewichtsabnahme von 10 kg war deutlich effektiver und verhinderte bei über 90 % der Studienteilnehmer die Diabetesmanifestation [25] über 3 Jahre.
Die Remission des T2Dm nach 5 Jahren durchschnittlicher Diabetesdauer und einem Jahr intensiven Lebensstilmodifikationsprogramms mit 8,9 % Gewichtsreduktion (Ausgangs-BMI 35 kg/m2) betrug 11,5 % in der Look-Ahead-Studie. Nach 4 Jahren betrug die Gewichtsreduktion noch 4,7 % des Ausgangsgewichts, und 7,3 % zeigten eine Remission definiert als Nüchternblutzucker unter 126 mg/dl ohne Diabetesmedikamente [26].
In der DIRECT-Studie bedingte eine Gewichtsreduktion von 15 kg durch Formuladiät eine Remission des T2Dm um 86 % nach maximal 6 Jahren vorheriger Diabetesdauer. Die Erfolgsrate sank erheblich bei geringerem Gewichtsverlust, allerdings gelang ein großer Gewichtsverlust auch nur wenigen Patienten. Die Daten zeigen eine quantitative Wirkung des Gewichtsverlusts auf die Diabetesremission [13]. Patienten sollte deshalb möglichst früh nach der Diagnose eines T2Dm eine entsprechende Therapie angeboten werden [21].
Welche Rolle spielt die Gewichtsreduktionsstrategie einer Formuladiät gegenüber einer langsamen moderaten Gewichtsreduktion? Langfristig liegt die Wahrscheinlichkeit einer Wiederzunahme nach Beendigung des Ernährungsprogramms bei über 80 %. Formuladiäten bedingen einen schnelleren und größeren Gewichtsverlust und zeigen auch langfristig noch eine größere Gewichtsabnahme [27].
Die Gewichtsreduktion führt zu einer schnellen Besserung der hepatischen Insulinresistenz, sodass die Blutzuckerspiegel bei erhaltener Insulinsekretionskapazität schnell sinken. Bei Insulintherapie und Insulinresistenz muss die Insulinmenge schnell (1 bis 5 Tage) reduziert werden, oft um 2 Drittel der Ausgangsdosis. Der Patient muss darauf vorbereitet werden, oder die Therapie sollte stationär für die ersten Tage initiiert werden, ambulant nur bei täglichem Patientenkontakt.
Einsatz von Telemedizin bei Typ-2-Diabetes mellitus
Empfehlung.
Telemedizinische Anwendungen können die Umsetzung von Verhaltensmodifikationen unterstützen, die bei der Therapie des T2Dm empfohlen werden.
Telemedizin kann die Adhärenz für Gewichtsreduktionsprogramme und die Erreichbarkeit erhöhen.
Kommentar
Aufgrund der COVID-19-Pandemie ist der Bedarf an digitalen Beratungsmethoden in der Therapie von Diabetes mellitus angestiegen. Telemedizin bezeichnet den Einsatz audiovisueller Kommunikationstechnologien zum Zweck von Diagnostik, Konsultation und medizinischen Notfalldiensten [28]. Die telemedizinische Betreuung wurde bereits vor der COVID-19-Pandemie bei Diabetespatienten eingesetzt und hat sich als eine bewährte Therapieform etabliert.
Im Rahmen eines telemedizinischen Programms werden therapierelevante Daten (z. B. Blutglukosespiegel, Insulindosis, Körpergewicht) dem Fachpersonal übermittelt, woraufhin der Patient eine Rückmeldung erhält. Dabei wird zwischen einer telemedizinischen Therapie via Textnachrichten/E-Mail und per Telefon/Videokonferenz unterschieden.
Eine Metaanalyse von Su et al. aus dem Jahr 2015 mit 92 inkludierten Studien zeigte eine signifikante Senkung des HbA1c-Werts bei Typ-1- sowie Typ-2-Diabetikern durch eine telemedizinische Ernährungstherapie [29]. Es wurde allerdings kein signifikanter Unterschied zwischen telemedizinischen Programmen via Nachrichten (per Handy oder E‑Mail) und einem persönlichen Beratungsgespräch (Telefonat oder Videokonferenz) festgestellt.
Für Deutschland wurden in einer randomisierten, kontrollierten Studie von Kempf et al. beim 1‑Jahres-Follow-up in der telemedizinisch betreuten Gruppe vs. Standardtherapie ein um 0,6 % niedrigerer HbA1c-Wert und eine um 5 kg größere Gewichtsreduktion berichtet [30].
Telemedizinische Anwendungen können von Ärzten und Psychotherapeuten verordnet und von den gesetzlichen Krankenkassen erstattet werden, wenn sie als sog. Digitale Gesundheitsanwendungen (DiGA) in das BfArM-Verzeichnis aufgenommen sind. Geregelt ist dies im Digitalen Versorgungsgesetz (DVG), das im Dezember 2019 in Kraft getreten ist. DiGA werden in der Regel vom Patienten allein genutzt. Es ist aber auch möglich, dass Patienten und Leistungserbringer die DiGA z. B. in Form von Telekonsilen oder Chats gemeinsam nutzen. Zum Zeitpunkt der Veröffentlichung dieser Praxisempfehlungen ist im BfArM-Verzeichnis keine DiGA mit der Indikation „Diabetes“ geführt, derzeit befinden sich aber mehrere „Diabetes-DiGA“ in der Evaluation.
Die DiGA „Zanadio“ mit der Indikation „Adipositas“ ist vorläufig in das BfArM-Verzeichnis aufgenommen. Zanadio arbeitet auf der Basis der Leitlinienempfehlungen zur Therapie der Adipositas und unterstützt eine konservative Adipositastherapie bestehend aus Bewegung, Ernährung und Verhaltensänderung. Zanadio enthält telemedizinische Elemente, indem die Nutzer mittels Chat-Funktion durch eine Ernährungsberaterin betreut werden.
Beispiel für eine telemedizinische Anwendung – allerdings nicht als DiGA zugelassen – ist das Telemedizinische Lebensstil-Interventions-Programm TeLiPro. Bei diesem Programm wird den Patienten eine App zur Verfügung gestellt, mit deren Hilfe Lebensstilaktivitäten gemonitort werden. Dazu werden Bluetooth-kompatible Blutglukosemessgeräte, Waagen, RR-Geräte und Schrittzähler genutzt. Über eine Cloud ist es dem Diabetescoach (Diabetesberater) möglich, die Daten einzusehen und über eine Chatfunktion bzw. übers Telefon direkt mit den Patienten zu interagieren.
In der TeLiPro-Studie erhielten beide Gruppen die App, Waagen, Schrittzähler, Blutglukose- und RR-Messgeräte. Jedoch unterschieden sich die Gruppen insofern, als ein Diabetescoach nur den Patienten der Interventionsgruppe zur Verfügung stand [29].
Als Ergebnis ist erkennbar, dass die Interventionsgruppe im Gegensatz zur Kontrollgruppe eine deutliche Senkung des HbA1c-Werts aufwies (mean ± SD −1,1 ± 1,2 % vs. −0,2 ± 0,8 %; p < 0,0001). Außerdem konnte eine Reduktion des Gewichts verzeichnet werden (TeLiPro (−6,2 ± 4,6 kg vs. control −1,0 ± 3,4 kg), BMI (−2,1 ± 1,5 kg/m2 vs. −0,3 ± 1,1 kg/m2)). Des Weiteren berichtete die Interventionsgruppe von einer grundsätzlich besseren Lebensqualität sowie einem besseren Ernährungszustand [30].
Strategien zur Gewichtsreduktion und zum Gewichtserhalt
Empfehlung.
Gewichtsreduktion muss klar indiziert sein, bevor sie empfohlen wird. Ein höheres Lebensalter ist ein Risikofaktor für Sarkopenie und kardiometabolische Nachteile durch hypokalorische Diäten.
Die engmaschige Betreuung durch Ernährungsberatung ist notwendig, um eine langfristig gute Compliance zu ermöglichen.
Die Strategie zur Gewichtsreduktion soll zu den Präferenzen der übergewichtigen Person passen (individuelle Ernährungstherapie).
Die Strategie zur nachhaltigen Stabilisierung eines reduzierten Körpergewichts soll individuell mit der betroffenen Person abgestimmt sein.
Bislang ist keine Ernährungsform anderen Diätmustern bei der Gewichtsreduktion klar überlegen.
Kommentar
Verschiedene Formen der hypokalorischen Ernährungsumstellung – von langfristig nutzbaren bis auf kurze Interventionen beschränkte Verfahren – führen bei T2Dm-Patienten zu einer Reduktion des Körpergewichts und oftmals auch zu einer Verbesserung der Stoffwechsellage und weiterer kardiovaskulärer Risikofaktoren. Allerdings gelingt eine nennenswerte, langfristige Gewichtsabnahme nur wenigen Patienten, sowohl mit komplexer Lebensstilintervention als auch mit Formuladiät. Somit bleibt bislang das eigentliche Ziel – die Diabetesremission, aber auch die Reduktion des tatsächlichen Langzeitrisikos für kardiovaskuläre Morbidität und Mortalität – allenfalls für schwer definierbare Subgruppen erreichbar [13, 31]. Bariatrische Verfahren sind ebenfalls erfolgreich bei der Diabetesremission, aber auch an strenge Kriterien zur Indikation geknüpft [32].
Für die Gewichtsreduktion haben sich zahlreiche Strategien entwickelt, die sich von ihrem Ansatz hinsichtlich der täglichen Energieaufnahme („low-calorie diet“ [LCD]/VLCD), der Nährstoffrelation (Low-Fat/Low-Carb), der Konsistenz (übliche Lebensmittel/Formula-Drinks), der Bevorzugung einer omnivoren bzw. einer vegetarischen/veganen Ernährungsweise, der Begrenzung von Fasten- und Esszeiten (intermittierendes Fasten) unterscheiden.
Die Effekte dieser jeweiligen Ansätze werden fortwährend publiziert und verfochten. Allerdings gibt es keine Strategie, die einer anderen grundsätzlich überlegen wäre. Es kommt darauf an, welche Methode (bzw. Kombination von Methoden) die abnehmwillige Person präferiert und in ihrer Motivation einer nachhaltigen Umsetzung im Alltag befördert [2, 3].
In welchem Umfang die in den meisten Studien angestrebte und letztlich erzielte Gewichtsreduktion tatsächlich effektentscheidend, also notwendig ist, ist nicht abschließend geklärt [33]. Auch Ernährungsmodifikationen ohne Gewichtsreduktion erzielen mitunter dramatische Verbesserungen. Ein systematischer Head-to-Head-Vergleich von hypo- und isokalorischen Diäten mit gleicher Makronährstoffrelation ist in der Literatur kaum beschrieben. Metaanalysen sehen kaum einen langfristigen metabolischen Vorteil für eine primär auf Gewichtsreduktion ausgerichtete Intervention im Vergleich zur Standardtherapie, allerdings bei erheblicher Heterogenität der Studien [34].
Die Aufrechterhaltung der langfristigen Gewichtsabnahme wird stark von der Fähigkeit beeinflusst, das Diätprogramm einzuhalten. Verhaltensunterstützung kann die Ergebnisse deutlich verbessern. Es gibt individuelle Unterschiede in der Reaktion auf die einzelnen Diäten, die größer sind als der Unterschied im mittleren Gewichtsverlust zwischen den Vergleichsdiäten.
Interaktion zwischen Ernährung und körperlicher Aktivität
Empfehlung.
Ein hohes Maß an körperlicher Aktivität mit geringer Intensität (z. B. zügiges Gehen) nach den Mahlzeiten verbessert die Körpergewichtsregulation und wirkt sich günstig auf die Regulation der Glykämie aus.
Kommentar
Während Inaktivität oder eine überwiegend sitzende Lebensweise ein Risiko für eine zu hohe Kalorienzufuhr und damit für die Entstehung von Übergewicht darstellen [35–37], sorgt ein hohes Maß an körperlicher Aktivität auch bei einer niedrigen Intensität (z. B. schnelles Gehen) für eine bessere Anpassung des Appetits an den Energiebedarf [38, 39] und verbessert damit die Regulation des Körpergewichts sogar unabhängig von einem höheren Kalorienverbrauch [38].
Zusätzlich haben Trainingsart, Intensität und Timing (nüchtern oder postprandial) einen Einfluss auf die Regulation der Glykämie [40]. Die Intensität der körperlichen Aktivität korreliert dabei positiv mit der Verbesserung der Insulinsensitivität, und die besten Resultate werden durch eine Kombination von Kraft- und Ausdauertraining erzielt [40]. Es gibt Hinweise, dass Sport mit hoher Intensität (z. B. High-Intensity-Intervall-Training [HIIT]) die Blutzuckerregulation am besten nüchtern (d. h. bei einer geringen Substratverfügbarkeit) verbessert [40]. Effektivität und Sicherheit dieser Methode bei Patienten mit T2Dm müssen jedoch noch weiter untersucht werden. Sicher und effektiv im Hinblick auf die Verbesserung der Glykämie bei Patienten mit T2Dm ist dagegen körperliche Aktivität mit niedriger Intensität, v. a. dann, wenn die Substratverfügbarkeit hoch ist. Dementsprechend wirkt sich schnelles Spazierengehen nach dem Essen durch eine Verbesserung der insulinunabhängigen Glukoseaufnahme günstig auf die postprandiale Glykämie aus [41–46].
Reduktion von Kohlenhydraten (Low-Carb)
Empfehlung.
Für die Gewichtsreduktion ist eine moderate Reduktion der Kohlenhydrate v. a. kurzfristig als eine mögliche Methode empfehlenswert (z. B. traditionell mediterran, pflanzenbetont).
Kohlenhydrate sollten bevorzugt in Form von Vollkornprodukten, Hülsenfrüchten und Nüssen verzehrt werden.
Für den Gewichtserhalt sind Low-carb- und Low-fat-Ernährungsformen wahrscheinlich ebenbürtig und sollten nach individueller Präferenz gewählt werden.
Insbesondere Low-carb-Diäten können bei Personen mit Insulintherapie nur unter engmaschiger Therapiekontrolle durchgeführt werden.
Kommentar
Kohlenhydrate machen in der Ernährung der Deutschen durchschnittlich ca. 45 % der Energieaufnahme aus, darunter etwa 90 g Zucker (= 18 Energieprozent [E%]) und oftmals hauptsächlich schnell verwertbare Polysaccharide. Epidemiologisch besteht eine erhöhte Mortalität bei einer Kohlenhydratzufuhr von mehr und weniger als 50 % (Letzteres nur bei tierisch betonten Eiweißquellen) [47]. Eine Reduktion von Kohlenhydraten im Rahmen einer Ernährungsintervention führt fast unweigerlich zur Gewichtsreduktion und zu metabolischen Veränderungen. Die wissenschaftliche Literatur betrachtet kohlenhydratarme Ernährungsformen zumeist in Gegenüberstellung zu „Low-Fat“. Kohlenhydratreduktion kann je nach Intensität als Moderate-Carb, Low-Carb oder Very-low-Carb eingeteilt werden; auch die traditionell mediterrane Diät ist danach eine kohlenhydratreduzierte Ernährungsweise [42].
ADA und EASD stufen Low-Carb als diätetische Therapieoption ein, stellen die mediterrane Ernährung aber als überlegen voran [48]. Dieser Konsensus spiegelt den Wissensstand aus aktuellen Metaanalysen wider: Unter allen in RCTs untersuchten nahrungsqualitativ definierten Ernährungsmodellen schneidet die traditionell mediterrane Diät bei Nüchternglukose und Lipidprofil am besten ab, bei HbA1c-Wert, Blutdruck und Gewichtsreduktion liegt sie jeweils unter den besten 3 Diäten. Low-Carb ist die wirksamste Methode zur Senkung des HbA1c-Werts und des Körpergewichts; bei der Reduktion der Nüchternglukose, des Blutdrucks und der Blutfette ist diese Ernährungsweise ebenfalls sehr erfolgreich und wirksamer als Low-Fat [49–51]. Mit längerer Anwendung gleichen sich Low-Carb und Low-Fat jedoch in ihrem Effekt an; ob dafür die schwindende Compliance oder ein Versagen der metabolischen Response ursächlich ist, lässt sich derzeit nicht beantworten [52].
Eine sehr aktuelle Metaanalyse hebt zudem hervor, dass Low-Carb (< 26 E% oder < 130 g Kohlenhydrate [KH]/Tag) gegenüber Low-Fat bei der Diabetesremission überlegen sein könnte. Nach 6 Monaten erreichen mit Low-Carb signifikant mehr Patienten einen HbA1c-Wert unter 6,5 %; bei Anwendung des zusätzlichen Kriteriums der Medikationsfreiheit oder längerer Intervention sind die Unterschiede nicht signifikant [53].
Betrachtet man den Effekt spezifischer Nahrungsgruppen auf das metabolische Gesamtbild aller kardiovaskulären Risikoparameter, so schneiden unter 66 Lebensmittelkategorien Nüsse, Hülsenfrüchte und Vollkornprodukte – allesamt Kohlenhydratträger – am besten ab [50]. Ein alleiniger isokalorischer Austausch von verschiedenen verdaulichen Kohlenhydraten gegeneinander bewirkt nur relative geringe Effekte auf Nüchternglukose und Low-Density-Lipoprotein(LDL)-Cholesterin (Zucker durch Stärke ersetzt) sowie Homeostasis Model Assessment-Insulin Resistance (HOMA-IR) und Harnsäure (Fruktose durch Glukose ersetzt). Die Evidenz dieser Resultate wird aber als gering eingeschätzt [54]. Effekte auf inflammatorische Parameter sind nicht zu beobachten [55].
In der Gesamtschau ist die traditionell mediterrane Diät als ein spezifischer Vertreter von „Low-Carb“ als optimale Ernährungsform anzusehen. In allgemeinerer Lesart sind „Low-Carb“ und „Low-Fat“ spätestens nach einigen Monaten Intervention metabolisch ebenbürtig [55]. Es gibt nach aktuellem Wissensstand kein eindeutiges langfristiges Optimum für den Energieanteil der Kohlenhydrate. Patienten, deren persönliche Vorliebe stark zu einer dieser Diätvarianten tendiert, können diese anwenden. Je nach Intensität und Dynamik sind aber zusätzliche zwischenzeitliche Stoffwechselkontrollen empfehlenswert, um eine individuell nicht vorhersagbare Entgleisung von Glykämie und Insulinresistenz, Lipidmetabolismus oder Harnsäurespiegel frühzeitig zu erkennen [50].
Reduktion von Fetten (Low-Fat)
Empfehlung.
Personen mit T2Dm kann nicht generell eine fettarme Ernährung empfohlen werden.
Kommentar
Die alleinige Reduktion der Nahrungsfette ist – wie im Kapitel zur Kohlenhydratreduktion bereits beschrieben – mit einem gegenüber allen kohlenhydratreduzierten Diäten unterlegenen Outcome zur Gewichtsreduktion, Blutdrucksenkung sowie Optimierung von Triglyzeriden und glykämischen Parametern verbunden [48, 51, 52, 56]. Während die Diabetesprävention durch komplexe Lebensstilintervention mit Low-fat-Ansatz konsistent gezeigt ist [24], ist die Chance zur Diabetesremission durch fettarme Ernährungsumstellung als alleinige Maßnahme vergleichsweise klein [26, 57].
Transfette
Empfehlung.
Industrielle Transfette sollten weiterhin gemieden werden, natürliche Transfette sind wahrscheinlich unproblematisch.
Kommentar
Einen relevanten Einfluss auf die glykämische Stoffwechsellage übt ferner die Fettqualität aus. Industriell hergestellte Transfette erweisen sich in Beobachtungsstudien als mortalitätssteigernd, insbesondere durch ein erhöhtes KHK-Risiko. Ein erhöhtes Diabetesrisiko wird nicht beschrieben [58].
Natürliche Transfette, wie sie in Rindfleisch und Milchprodukten vorkommen, stehen in epidemiologischen Studien mit einem erniedrigten Diabetesrisiko in Verbindung und beeinflussen nicht das Risiko für kardiovaskuläre Mortalität oder Morbidität [58].
Gesättigte Fette
Empfehlung.
Lebensmittel mit einem natürlichen Gehalt an gesättigten Fetten sind bei maßvollem Verzehr unbedenklich. Hochverarbeitete Produkte mit zugesetzten gesättigten Fetten sollten gemieden werden.
Kommentar
Der Diskurs zu gesättigten Fetten ist auch im Jahr 2021 noch nicht zu einem schlüssigen Ergebnis gekommen. Die durch die Seven Countries Study und viele epidemiologische Folgeerhebungen befeuerte Kritik an gesättigten Fetten (mitunter sogar fälschlicherweise allen Fetten) ist in neueren Metaanalysen zu Kohortenstudien nicht mehr berechtigt [58]. Die Evidenz bezüglich eines Schadenspotenzials von gesättigtem Fett ist nicht hinreichend [59]. Selbst Butter als typisches Lebensmittel mit sehr hohem Anteil an gesättigtem und Gesamt-Fett steigert epidemiologisch nur minimal die Mortalität, beeinflusst aber nicht das kardiovaskuläre Risiko und steht eher mit geringerem Diabetesrisiko in Verbindung [60]. Auch andere fettreiche oder fettarme Milchprodukte wirken sich kaum nachteilig auf Stoffwechseloutcomes aus [61].
RCTs zur fettreduzierten Ernährung zeigen zwar im Mittel eine leichte Reduktion von Körpergewicht, BMI, Körperfettanteil und Taillenumfang [62], jedoch keinen Effekt auf KHK, kardiovaskuläre Mortalität oder Gesamtsterblichkeit [63]. Eine Reduktion von gesättigtem Fett wirkt sich konsistent günstig auf den inflammatorischen Phänotyp aus [31, 64]. Zudem senkt sie nachweislich den LDL-Spiegel, verschlechtert aber die HDL- und Triglyzeridspiegel [65].
Ungesättigte Fette
Empfehlung.
Ein hoher Anteil ungesättigter Fettsäuren sollte bei Patienten mit T2Dm unabhängig von der Gesamtfettmenge durch Zufuhr von natürlichen Lebensmitteln, aber nicht durch Supplemente angestrebt werden.
Kommentar
Beobachtungsstudien beschreiben deutliche diabetes- und kardioprotektive Assoziationen zu einfach und mehrfach ungesättigten Fettsäuren, insbesondere für Linolsäure und Alpha-Linolensäure [66–68].
In Interventionsstudien fehlt der Nachweis der Kardioprotektion und Mortalitätsreduktion für mehrfach ungesättigte (engl. „polyunsaturated fatty acids“ [PUFAs]) Omega-6-Fettsäuren und nicht langkettige pflanzliche Omega-3-PUFAs [69, 70]. In Metaanalysen randomisierter, kontrollierter Studien ist zudem kein glykämischer Benefit für ungesättigte Fettsäuren zu sehen, wenn ein Vergleich gegen gesättigte Fettsäuren vorgenommen wird [71]. Gegenüber Kohlenhydraten sind einfach ungesättigte Fettsäuren (engl. „monounsaturated fatty acids“ [MUFAs]) in allen metabolischen Achsen, jedoch nicht dem Blutdruck, von Vorteil [72, 73]. Im Vergleich zu gesättigten Fetten oder Placebo feststellbar ist ein Nutzen hinsichtlich Taillenumfang, Inflammation, Triglyzeridspiegel und Plättchenaggregation sowie wahrscheinlich Fettleber (Omega-3-Fettsäuren) [74–78]. Ein hohes Verhältnis von Omega-3-/Omega-6-Fettsäuren kann bei Menschen mit Diabetes und bei längerer Intervention eine günstige Rolle spielen, insbesondere bei der Absenkung des Insulin-, aber nicht des Glukosespiegels [79, 80]. Frauen profitieren davon offenbar deutlicher als Männer [81]. Für Alpha-Linolensäure gibt es keinen eindeutigen interventionellen Vorteil bezüglich der diabetischen Stoffwechsellage [82].
Intermittierendes Fasten/Intervallfasten
Empfehlung.
Intervallfasten kann unter ärztlicher Überwachung als Mittel zur Gewichtsreduktion eingesetzt werden.
Es kann keine generelle Empfehlung für irgendeine Form des Intervallfastens ausgesprochen werden.
Kommentar
Neben der qualitativen Anpassung der Ernährung durch ein verändertes Nährstoffprofil oder gezielte Umverteilung von Lebensmittelgruppen wird auch die Mahlzeitenfrequenz als Ansatzpunkt für Gewichtsreduktion und metabolische Verbesserung angesehen.
In randomisierten Studien zur täglichen Mahlzeitenzahl zeigt sich ein kleiner Nutzen zugunsten seltenerer Mahlzeiten (1 bis 2 vs. 6 bis 8) bezüglich Körpergewicht, Fettmasse und Taillenumfang. Diese Effekte sind aber insgesamt von geringer Evidenz [82].
Seltenere Zufuhr von Nahrung verlängert in einigen Tiermodellen die Lebensdauer. Beobachtungsstudien am Menschen (z. B. im Kontext des Ramadans) sehen bei Gesunden nur relativ geringe, zudem transitorische metabolische Änderungen [83–85]. Bei Diabetikern wird auch die Verschlechterung des Stoffwechsels beschrieben. Weitere Kohortenstudien beschreiben ein selteneres Auftreten von koronarer Herzerkrankung und T2Dm [86, 87].
Das gezielte, längerfristige, regelmäßige Auslassen von Mahlzeiten nach einem festen chronologischen Muster („Intervallfasten“) umfasst verschiedene Varianten: umtägiges Fasten („alternate day fasting“), 5:2-Fasten und „time-restricted eating“ (z. B. 16:8-Fasten). Diese werden in der Literatur mitunter gebündelt mit einer kontinuierlichen Kalorienrestriktion oder auch unveränderter Kontrolldiät verglichen.
In sämtlichen Metaanalysen zum Intervallfasten (8 Metaanalysen zu über 40 RCTs) findet sich keine Überlegenheit des Intervallfastens gegenüber kontinuierlicher Kalorienreduktion. Gegenüber unveränderter Kontrolldiät besteht zwar eine signifikant stärkere Absenkung von Körpergewicht, Taillenumfang, Blutdruck und Triglyzeriden, jedoch nicht von LDL-Cholesterin, Nüchternglukose oder HbA1c-Wert [88–92]. RCTs mit T2Dm-Patienten sind rar. Diese zeigen das gleiche Muster an erwünschten Outcomes wie in den genannten Metaanalysen, jedoch ein erhöhtes Risiko für Hypoglykämien [93–97].
Mahlzeitenersatz/Formuladiäten (mit/ohne multimodales Programm)
Empfehlung.
Niedrigkalorische Formuladiäten erlauben einen klinisch relevanten Gewichtsverlust bei Menschen mit T2Dm, verbunden mit einer erheblichen Verbesserung des Glukose- und Lipidstoffwechsels und einer Reduktion weiterer kardiovaskulärer Risikofaktoren.
Kommentar
Der Ersatz von Mahlzeiten durch niedrigkalorische Formuladiäten stellt eine sichere und effektive Maßnahme zur Gewichtsreduktion bei übergewichtigen und adipösen Menschen mit T2Dm im Vergleich zu herkömmlichen kalorienreduzierten Diäten dar. Neben der günstigen Beeinflussung anthropometrischer Parameter wie Taillenumfang und Körperfettmasse verbessern Formuladiäten auch weitere kardiometabolische Risikoparameter wie Blutdruck, Nüchternglukose, HbA1c-Wert und Lipidstoffwechsel [98–102]. Im Rahmen von Gewichtsreduktionsprogrammen führt der Einsatz von Formuladiäten zu einer ausgeprägten Gewichtsreduktion, die der nach bariatrischen Eingriffen ähnlich ist, verbunden mit einer anhaltenden Diabetesremission. Allerdings erreichen nur 25 % eine Gewichtsreduktion von > 15 %, bei der die Remission sehr wahrscheinlich eintritt [13, 103].
Wissenschaftlicher Hintergrund
Übergewicht ist einer der wichtigsten Risikofaktoren für die Entstehung des T2Dm [104]. Bei 60–90 % der Patienten mit T2Dm besteht Übergewicht oder Adipositas [105, 106]. Hingegen führt eine Gewichtsreduktion zu einer Verbesserung des Glukose- und Lipidstoffwechsels und einem Abfall erhöhter Blutdruckwerte. Somit stellt eine Gewichtsabnahme bei Patienten mit T2Dm eine der wichtigsten therapeutischen Maßnahmen dar [106]. Allerdings ist eine bereits für Menschen ohne Diabetes herausfordernde Gewichtsabnahme bei Menschen mit T2Dm häufig noch zusätzlich erschwert aufgrund genetischer und metabolischer Unterschiede, Angst vor Hypoglykämien, glukosesenkenden Therapien, die eine Gewichtszunahme fördern, verminderter körperlicher Aktivität und einer Diät-Müdigkeit. Niedrigkalorische Diäten haben das Potenzial, bei Menschen mit T2Dm zu einem ähnlich ausgeprägten Gewichtsverlust zu führen wie bariatrisch-chirurgische Maßnahmen. Eine Metaanalyse von 9 Studien, die die Auswirkungen von sehr niedrigkalorischen Diäten (VLED, engl. „very-low-energy diet“) an insgesamt 192 adipösen Menschen mit T2Dm untersuchten, ergab, dass die Teilnehmer nach 6 Wochen 9,6 % des Ausgangsgewichts verloren hatten und die Nüchternglukose sich bereits nach 2 Wochen um 50 % reduziert hatte [106]. Allerdings fällt vielen Menschen mit T2Dm eine längerfristige Lebensstiländerung mit dem Ziel der Gewichtsreduktion schwer, und die Motivation kann bei Ausbleiben eines kurzfristigen Interventionserfolgs rasch verloren gehen. Niedrigkalorische Formuladiäten erwiesen sich in inzwischen zahlreichen Untersuchungen als sichere und effiziente Therapieoption, um bei adipösen Patienten mit T2Dm kardiometabolische Endpunkte wie Taillenumfang, Körperfettmasse, Blutdruck und HbA1c-Wert zu verbessern [98–101]. Eine Metaanalyse, die 4 Studien mit insgesamt mehr als 500 Studienteilnehmern einschloss, ergab, dass der Gewichtsverlust infolge niedrigkalorischer Formuladiäten, bei denen zwischen 300 und 1000 kcal an Energie pro Tag zu Verfügung gestellt wurden, bei Menschen mit T2Dm ähnlich wie der bei Menschen ohne Diabetes war, mit einer mittleren Gewichtsabnahme zwischen 8 und 21 % des Ausgangsgewichts nach einer Behandlungsdauer von 4 bis 52 Wochen. Auch gab es keinen Unterschied in der Gewichtsreduktionsrate zwischen Menschen mit (–0,6 kg pro Woche) und ohne T2Dm (–0,5 kg pro Woche) [107]. In einer weiteren Untersuchung ergab sich ebenfalls kein Unterschied in der Gewichtsabnahme nach Beginn einer niedrigkalorischen Formuladiät zwischen Patienten mit und ohne Diabetes. Bei einem Fünftel der Teilnehmer konnte nach 12 Monaten ein Gewichtsverlust von mehr als 15 kg erreicht werden. Unter den Teilnehmern, die das Gewichtsmanagementprogramm über 1 Jahr hinaus fortsetzten, wiesen nach 24 Monaten nahezu 40 % einen Gewichtsverlust von mindestens 15 kg auf [108]. Die Gewichtsreduktion infolge eines zeitlich begrenzten Einsatzes einer niedrigkalorischen Formuladiät geht mit einer längerfristigen Verbesserung des Glukose- und Lipidstoffwechsels sowie des Blutdrucks einher [109]. Auch bei Patienten mit einer unzureichenden Stoffwechseleinstellung kann der Mahlzeitenersatz durch eine Formuladiät zu einem klinisch relevanten Abfall des HbA1c-Werts und einer erheblichen Reduktion der Insulindosen bei Patienten mit einer intensivierten konventionellen Insulintherapie führen [110, 111]. Auch erscheint infolge einer strikten Kalorienrestriktion eine Diabetesremission möglich, wie die Ergebnisse der Diabetes Remission Clinical Trial (DiRECT) nahelegen [112]. Nahezu die Hälfte der übergewichtigen und adipösen Patienten mit T2Dm, die zunächst über 3 bis 5 Monate ausschließlich eine Formuladiät mit einem Kaloriengehalt von 825 bis 853 kcal pro Tag erhielten, erzielte eine Diabetesremission im Gegensatz zu nur 4 % der Patienten, die lediglich eine Standardtherapie durch den Hausarzt erhielten [13]. Nach 12 Monaten hatte ein Viertel der Interventionsgruppe das erklärte Ziel, 15 kg oder mehr abzunehmen, erreicht und kein Teilnehmer der Kontrollgruppe. Die Diabetesremission ging sehr eng mit der Gewichtsabnahme einher. Während eine Remission bei keinem der Patienten, die Gewicht zunahmen, auftrat, lag die Remissionsrate bei 86 % bei den Teilnehmern, die mindestens 15 kg abnahmen. Zwei Jahre nach der Intervention war noch mehr als ein Drittel der Patienten mit T2Dm in Remission. Bei den Teilnehmern, die mehr als 10 % abgenommen hatten, lag die Remissionsrate sogar bei 64 % [103]. Auch bei Menschen mit einem erhöhten Diabetesrisiko infolge von Übergewicht oder Adipositas und mindestens einer weiteren Komorbidität des metabolischen Syndroms war der zusätzliche Mahlzeitenersatz durch eine Formuladiät mit abnehmender Frequenz über den Studienzeitraum einer alleinigen Lebensstilintervention hinsichtlich der Gewichtsabnahme und der Verbesserung kardiometabolischer Risikofaktoren überlegen [113]. Zudem gelang bei der Hälfte der Teilnehmer, die zusätzlich eine Formuladiät erhielten, die Konversion von einem Prädiabetes in eine Normoglykämie, während dies bei weniger als einem Drittel der ausschließlich mit einer Lebensstilintervention behandelten Teilnehmer der Fall war [114]. Aus den genannten Gründen sehen Fachgesellschaften eine Remission sogar als primäres Behandlungsziel an [115].
Zusätzliche Aspekte der Gewichtsreduktion bei insulinbehandeltem T2Dm
Empfehlung.
Die Insulintherapie sollte aufgrund der anabolen Wirkung des Hormons auf das nötigste Maß beschränkt werden. Eine Gewichtsabnahme unter Insulintherapie ist erschwert.
Kommentar
Unter einer Insulintherapie kommt es bei den ohnehin größtenteils übergewichtigen Patienten mit Diabetes zudem häufig zu einer Gewichtszunahme: So ergab die United Kingdom Prospective Diabetes Study (UKPDS), in der mit Insulin behandelte T2Dm-Patienten randomisiert wurden, eine Gewichtszunahme im Schnitt von 6,5 kg [116]. Eine Lebensstilintervention bleibt trotz Insulintherapie ein sehr wichtiger Therapiebaustein [117].
Allerdings konnte in einer anderen Studie gezeigt werden, dass die Gewichtszunahme umso geringer war, je höher der Ausgangs-BMI der Patienten war. Beim Rückgang des HbA1c-Werts um je einen Prozentpunkt stieg das Gewicht bei den Normalgewichtigen (BMI unter 25 kg/m2) im Schnitt um 1,24 kg an, bei den stark Adipösen (BMI über 40 kg/m2) ging das Gewicht aber sogar um 0,32 kg zurück [118].
Zusammenfassende Bewertung und Ausblick
Zur Gewichtsreduktion bzw. der Gewichtsstabilisierung steht eine Reihe von Verfahren zur Auswahl. Für jede dieser Methoden gibt es mehr oder weniger gute Evidenz. Der Fokus muss aus unserer Sicht auf die individuellen Präferenzen der Patienten gelegt werden, die unabhängig vom Outcome die Adhärenz zum jeweiligen Therapieverfahren stärken.
Ernährungsmuster
Empfehlung.
Für das Diabetesmanagement und die Reduktion des Risikos kardiovaskulärer Komplikationen bei Personen mit T2Dm ist eine Auswahl verschiedener Ernährungsmuster akzeptabel wie beispielsweise eine mediterrane, vegetarische oder vegane Ernährung.
Für die DASH-Diät, das nordische Ernährungsmuster und die Paleo-Diät ist die Evidenz derzeit unzureichend, um sie speziell für die Therapie des T2Dm zu empfehlen.
Bis zum Vorliegen zusätzlicher Evidenz zur Überlegenheit eines speziellen Ernährungsmusters bezogen auf die Zielparameter der Diabetestherapie sollten sich Personen mit T2Dm an den Gemeinsamkeiten der genannten Ernährungsmuster orientieren: nicht stärkehaltige Gemüsesorten und wenig verarbeitete Lebensmittel bevorzugen sowie raffinierte Zucker und hochverarbeitetes Getreide vermeiden.
Kommentar
Für Personen mit T2Dm gibt es basierend auf der aktuellen Evidenz kein Ernährungsmuster, das allgemeingültig für alle Betroffenen empfohlen werden könnte. Stattdessen sind nach den Empfehlungen der Fachgesellschaften verschiedene Ernährungsmuster wie die mediterrane Ernährung oder eine vegetarische oder vegane Ernährung geeignet, um die Zielparameter der Diabetestherapie zu erreichen [2, 119–121]. Während die Evidenz für die Effekte der mediterranen Ernährung bei Personen mit T2Dm primär auf RCTs basiert – darunter mehrere größere Studien und Langzeituntersuchungen – und deren systematischen Reviews und Metaanalysen [122], weisen die RCTs zur vegetarischen und veganen Ernährung meist eine kleine Fallzahl und kurze Studiendauer auf [122–125]. Die zurzeit vorliegende Evidenz zur DASH-Diät, dem nordischen Ernährungsmuster [126–128], der Paleo-Diät [2] und der makrobiotischen Ernährung [123, 127] bei Personen mit T2Dm ist gering und z. T. widersprüchlich, sodass weitere Studien notwendig sind, um beobachtete positive Effekte dieser Ernährungsmuster für das Diabetesmanagement bei T2Dm zu untermauern.
Bei Personen mit neu diagnostiziertem T2Dm erzielte die mediterrane Ernährung einen als klinisch relevant eingestuften Gewichtsverlust von ≥ 5 % [129]. Ebenso ergaben weitere Metaanalysen aus RCTs bei Personen mit T2Dm für die mediterrane Ernährung im Vergleich zu den jeweiligen Kontrolldiäten einen signifikant größeren Gewichtsverlust [130–132]. Auch die Adhärenz zu einer vegetarischen oder veganen Ernährung bzw. zu pflanzenbasierten Ernährungsmustern im Allgemeinen führte zu einem Gewichtsverlust bei Personen mit und ohne T2Dm [49, 133–135].
Basierend auf einer Netzwerk-Metaanalyse aus 56 RCTs und 9 Ernährungsmustern [136] sowie der Evidenz aus mehreren Metaanalysen aus RCTs [131, 132, 137] ist die mediterrane Ernährung den jeweiligen Kontrollernährungen in der Reduktion des HbA1c-Werts überlegen und nach Low-carb-Ernährungsformen in der Reduktion des HbA1c-Werts und der Nüchternblutglukose am effektivsten, gefolgt von der Paleo-Diät und der vegetarischen Ernährung [49, 131, 132, 137]. Weitere systematische Reviews und Metaanalysen bestätigen die positiven Effekte der vegetarischen und veganen Ernährung auf die glykämische Kontrolle bei Personen mit und ohne T2 D [124]. Jedoch reduzierten auch alle anderen in der Netzwerk-Metaanalyse untersuchten Ernährungsformen im Vergleich zur Kontrolldiät den HbA1c-Wert und die Nüchternblutglukose bei Personen mit T2Dm signifikant, und die Ergebnisse wurden wegen signifikanter Inkonsistenzen insgesamt nur mit sehr geringer bis moderater Glaubwürdigkeit und Belastbarkeit der Evidenz bewertet [49]. Somit kann eine Überlegenheit einer Ernährungsform gegenüber den anderen in Bezug auf die Reduktion der Glukoseparameter derzeit nicht abgeleitet werden [138]. Außerdem sind weitere Studien notwendig, um die Effekte der Ernährungsmuster auf die glykämische Kontrolle bei Personen mit T2Dm unabhängig vom Gewichtsverlust zu bestätigen [121, 123, 138, 139], die die Unterschiede zwischen den verschiedenen Formen der vegetarischen und veganen Ernährung untersuchen [123, 124].
Neben positiven Effekten auf den Gewichtsverlust und die glykämische Kontrolle könnten Ernährungsmuster auch die Inzidenz und Mortalität verschiedener kardiovaskulärer Outcomes reduzieren sowie einzelne kardiometabolische Risikofaktoren wie Dyslipidämie und arterielle Hypertonie bei Personen mit und ohne T2Dm verbessern [123, 124, 131, 132, 140]. Die dazu vorliegende Evidenz ist für die mediterrane Ernährung gering bis moderat und für die vegetarische/vegane Ernährung und die DASH-Diät sehr gering bis gering (Inzidenz und Mortalität) bzw. gering bis moderat (Risikofaktoren). Für das nordische Ernährungsmuster liegt bislang nur eine vorläufige Studienbewertung vor, die auf eine sehr geringe Evidenz für die Reduktion der Inzidenz und Mortalität durch koronare Herzerkrankungen hinweist [124, 140]. Eine Metaanalyse auf der Basis von 52 RCTs und 9 Ernährungsmustern schlussfolgerte, dass mit einer geringen bis moderaten Evidenz die mediterrane Ernährung im Vergleich zur Kontrollernährung das HDL-Cholesterin am effektivsten erhöht und Triglyzeride reduziert, während die vegetarische Ernährung im Vergleich zu Kontrolldiäten das LDL-Cholesterin am effektivsten reduziert [51]. Für Effekte der mediterranen, veganen und vegetarischen Ernährung auf mikrovaskuläre, mit T2Dm assoziierte Komplikationen ist die Evidenz limitiert auf wenige Studien mit geringer Probandenzahl. Basierend auf Surrogatparametern werden Verbesserungen für Nephropathie und Retinopathie unter Einhaltung der genannten Ernährungsmuster vorgeschlagen, während die Evidenz für das Risiko des Auftretens mikrovaskulärer Komplikationen unzureichend ist und die Ergebnisse für Neuropathie inkonsistent sind [123]. Insgesamt ist es basierend auf der vorhandenen Evidenz somit schwierig, solide Schlussfolgerungen für die Effekte von Ernährungsmustern auf mikrovaskuläre und makrovaskuläre Komplikationen bei Personen mit T2Dm zu ziehen [123].
Da basierend auf der vorliegenden Evidenz keine Ernährungsform den anderen überlegen ist, wird eine individualisierte Mahlzeitenplanung mit dem Fokus auf Ernährungsmustern statt auf individuellen Nährstoffen oder einzelnen Lebensmitteln – bzw. den Faktoren, die den Ernährungsmustern gemeinsam sind – empfohlen [2, 3, 119].
Singuläre Effekte einzelner Nährstoffe
Eiweiß
Effekt auf Glykämie
Empfehlung.
Wir empfehlen eine Eiweißzufuhr von 10–25 % der Nahrungsenergiemenge (%E) für Patienten mit T2 D unter 60 Jahren und 15–25 % für Menschen über 60 Jahre bei intakter Nierenfunktion (GFR > 60 ml/min/m2) und Gewichtskonstanz.
Bei eingeschränkter Nierenfunktion jeglicher Stadien ist eine Eiweißreduktion auf weniger als 0,8 g/kgKG wahrscheinlich nicht von Vorteil und sollte aufgrund des Risikos für eine Malnutrition insbesondere bei höhergradiger Niereninsuffizienz vermieden werden.
Kommentar
Eine ausführliche AWMF-S3-Leitlinie zur Eiweißzufuhr bei T2Dm findet sich unter folgender Internetadresse: [141]. Eine Metaanalyse wurde publiziert und ist frei zugänglich [142].
Eiweiß wird als Lieferant der Aminosäuren in einer Mindestmenge von etwa 0,8 g/kg Körpergewicht oder 10 E% benötigt, um eine Mangelernährung und Sarkopenie zu vermeiden. Die untere Grenze von 0,8 g/kg/Tag kann für ältere Menschen unzureichend sein wegen einer abnehmenden Effizienz der Proteinsynthese [143], weshalb eine höhere Eiweißzufuhr von mindestens 1 g/kgKG/Tag empfohlen wird [144].
Umstritten ist die Bedeutung einer höheren Eiweißaufnahme. Argumente für eine höhere Eiweißzufuhr sind eine bessere Sättigung und ein höherer Energieverbrauch durch postprandiale Thermogenese, was einer Gewichtszunahme entgegenwirken kann. Der Eiweißstoffwechsel benötigt erheblich weniger Insulin als Kohlenhydrate, was die Blutzuckerkontrolle erleichtert und die Insulindosierung vereinfachen kann. Eine gewisse Insulinmenge ist allerdings wegen der eiweißbedingten Freisetzung von Glukagon erforderlich [145]. Ältere Menschen erleiden häufig erhebliche Muskelverluste durch Erkrankungen, Glukokortikoidtherapie, Immobilität oder Inappetenz, weshalb Geriater ebenfalls eine höhere Eiweißzufuhr empfehlen [146].
Argumente gegen eine höhere Eiweißzufuhr ergeben sich aus epidemiologischen Beobachtungsstudien, die eine höhere Sterblichkeit [143, 147] und Diabetesinzidenz [148] bei höherer Eiweißzufuhr beschreiben. Da sie Lebensstile und andere Variablen nicht ausreichend berücksichtigen, wurden die Aussagen dieser Beobachtungsstudien in Cochrane-Metaanalysen in Zweifel gezogen [148, 149]. Interventionsstudien zeigen durchgehend positive Effekte einer höheren Proteinaufnahme bei Übergewichtigen ohne Diabetes [150, 151]. Eine hohe Proteinaufnahme von über 20 E% gegenüber unter 20 E%, also etwa 1,2–1,6 g/kgKG, erhöhte nicht das Risiko für Diabetes oder andere Erkrankungen bei Prädiabetespatienten in einer großen europäisch-australischen prospektiven, randomisierten Interventionsstudie über 3 Jahre [150].
Empfehlung bei chronischer Niereninsuffizienz
Historisch gesehen wurden eiweißarme Ernährungspläne empfohlen, um die Albuminurie zu reduzieren und das Fortschreiten einer (diabetischen) Nephropathie zu verhindern.
Es liegen zur Frage der Eiweißzufuhr bei Personen mit Diabetes mellitus und chronischer Niereninsuffizienz aktuelle Metaanalysen vor, die zeigen, dass eine Proteinrestriktion auf 0,6–0,8 g/kgKG keine nachweisbare Verbesserung der Nierenfunktion bringt [152]. Derzeit wird sie weiterhin von nephrologischen Fachgesellschaften empfohlen [153], im Konsensuspapier der AG Ernährung der Amerikanischen Diabetesgesellschaft jedoch nicht [2].
Eine erhebliche Proteinrestriktion auf 0,3–0,4 g/kgKG zeigte in der Cochrane-Analyse eine signifikante, aber geringe Reduktion der terminalen Niereninsuffizienz, aber keinen Effekt auf die Sterblichkeit [154, 155]. Eine derartige Ernährungsform durchzuführen ist außerordentlich schwierig, führt zu einer erheblichen Verschlechterung der Lebensqualität und birgt ein hohes Risiko der Malnutrition und Sarkopenie, die in Stadien der terminalen Nierenfunktionsstörung mit einer erhöhten Mortalität assoziiert sind [156]. Zudem sind die für diese extreme Ernährungsform supplementierend einzusetzenden Aminosäurepräparate (Ketoanaloga) in Deutschland nicht verordnungsfähig.
Auch im Konsensuspapier der AG Ernährung der Amerikanischen Diabetesgesellschaft wird eine Einschränkung der Eiweißzufuhr bei Niereninsuffizienz nicht empfohlen [2].
Empfehlung zur Gewichtsreduktion
Empfehlung.
Im Rahmen von Gewichtsreduktionsdiäten bis zu 12 Monaten Dauer kann der Eiweißanteil auf 23–32 % der Gesamtenergiezufuhr gesteigert werden.
Kommentar
Hypokalorische Gewichtsreduktionsdiäten enthalten meist einen relativ erhöhten Eiweißanteil. Wegen der insgesamten Kalorienreduktion liegt er, bezogen auf das Körpergewicht (KG), zumeist im normalen Bereich von 0,9–1,2 g/kgKG, also im normalen bis leicht erhöhten Bereich. Zu diesen Diäten liegen zahlreiche Vergleichsstudien eines höheren mit einem niedrigeren Eiweißanteil vor. Insgesamt zeigen sich moderate Unterschiede kardiometabolischer Risikofaktoren durch einen höheren gegenüber einem niedrigeren Eiweißanteil in bisherigen Metaanalysen [130, 142, 157]. Obwohl Diäten mit höherem Eiweißanteil eine Gewichtsabnahme nur geringfügig verstärken, verbessern sich moderat die Nüchternblutzuckerwerte und der systolische Blutdruck. Insgesamt schneiden die proteinreicheren Diäten etwas besser ab und zeigen keine Nachteile [142].
Qualität der Kohlenhydrate, glykämischer Index, Zucker in hochverarbeiteten Lebensmitteln
Empfehlung.
Die Auswahl von Kohlenhydraten mit niedrigem GI trägt bei Patienten mit T2Dm zu einer Verbesserung des gesundheitlichen Risikos bei.
Der Einfluss des GI oder der GL ist dabei anteilig unabhängig von der Regulation der Glykämie und betrifft z. B. auch eine Verbesserung der Plasmalipide und eine höhere Aufnahme gesunder Inhaltsstoffe wie Ballaststoffe, Mikronährstoffe und sekundäre Pflanzenstoffe bei gleichzeitig geringerem Verzehr von abträglichen Inhaltsstoffen aus hochverarbeiteten Lebensmitteln mit hohem GI/GL.
Kommentar
Der glykämische Index (GI) und die glykämische Last (GL) beschreiben den Einfluss von kohlenhydratreichen Lebensmitteln auf die Glykämie. Der GI gibt an, wie schnell die Kohlenhydrate eines Lebensmittels verdaut, resorbiert und damit blutzuckerwirksam werden, während die GL den GI für die verzehrte Kohlenhydratmenge adjustiert. Die Blutzuckerantwort eines Lebensmittels hängt damit v. a. von Charakteristika des Lebensmittels selbst ab (z. B. von dem Verarbeitungsgrad und dem Fettgehalt) [158]. Phänotypcharakteristika der Patienten wie die Zusammensetzung des Darmmikrobioms spielen mutmaßlich eine untergeordnete Rolle [159, 160], obwohl auch individuelle Einflüsse beschrieben sind [161]. Ein GI der Diät ≤ 40 oder ≤ 55 gilt als niedrig und ein GI ≥ 70 als hoch [160]. Prospektive Beobachtungsstudien finden einen positiven Einfluss einer Diät mit niedrigem GI/GL auf die Prävention des T2Dm [162, 163]. Bei Patienten mit T2Dm kann ein starker Verzehr von Lebensmitteln mit einem niedrigen GI wie Hülsenfrüchten und Hafer die Blutzuckereinstellung verbessern, die Insulinsensitivität steigern und damit den Insulinbedarf senken [164]. Diese Effekte werden heute anteilig über einen positiven Einfluss von schwer verdaulichen Kohlenhydraten auf das Mikrobiom erklärt [165]. Bis heute bleibt strittig, inwieweit der Nutzen einer Low-GI-Diät durch deren höheren Ballaststoffgehalt erklärt ist. Eine 6‑monatige Intervention mit einer Low-GI-Diät konnte den HbA1c-Wert im Vergleich zu einer Ernährung reich an Getreideballaststoffen etwas besser senken (0,5 vs. 0,18 %) [166]. Diese Studie hatte jedoch erhebliche Schwächen, da die Ballaststoffgruppe Lebensmittel mit hohem GI meiden sollte und die Low-GI-Gruppe letztlich einen höheren Ballaststoffverzehr aufwies als die Ballaststoffgruppe. Tatsächlich führte jedoch eine 12-wöchige Substitution von stark blutzuckerwirksamen Kohlenhydraten durch Isomaltulose (Low-GI) bei Patienten mit T2Dm zu einer Reduktion des HbA1c-Werts und des HOMA-Indexes [167], was auf einen Einfluss des GI unabhängig vom Ballaststoffgehalt der Diät hinweist.
Trotz überzeugender Evidenz zur Diabetesprävention aus Beobachtungsstudien und plausiblen mechanistischen Erklärungsansätzen kommen systematische Reviews auf der Basis von randomisierten, kontrollierten Studien zum Einfluss des GI/GL der Diät bei Patienten mit T2Dm zu widersprüchlichen Ergebnissen. Sie zeigen sowohl positive [168, 169] als auch keine Effekte [170, 171] auf relevante Outcomeparameter wie den HbA1c-Wert und den Nüchternblutzuckerspiegel.
Eindeutiger ist wiederum das Ergebnis von prospektiven Kohortenstudien, die den Einfluss des GI/GL auf Komplikationen des Diabetes untersuchen. Das Risiko für KHK zeigte eine deutliche und dosisabhängige Beziehung zum GL oder GI der Diät [163]. In der Gruppe der übergewichtigen Probanden ist das Risiko für kardiovaskuläre Ereignisse oder Mortalität durch einen hohen GI dabei besonders hoch [172]. Diese Befunde passen zu früheren Ergebnissen, die ein höheres Risiko für tödliche und nichttödliche kardiovaskuläre Ereignisse mit steigender postprandialer Glykämie zeigen [173, 174]. Charakteristisch für die mit T2Dm assoziierte Dyslipidämie sind hohe Triglyzeridspiegel, niedrige HDL-Cholesterinspiegel und ein hoher Anteil kleiner dichter LDL-Partikel. Dieses Lipidmuster kann nicht nur durch eine Reduktion des Kohlenhydratverzehrs, sondern auch durch eine Senkung des GI/GL positiv beeinflusst werden [175].
Die Diskrepanz zwischen den Ergebnissen aus Beobachtungs- und Interventionsstudien wird anteilig dadurch erklärt, dass die gesundheitliche Bewertung von Lebensmitteln anhand des GI unzureichend ist. Die Qualität der Kohlenhydrate über den GI korreliert nicht nur mit dem Ballaststoffgehalt, sondern auch mit dem Mikronährstoffgehalt und dem Gehalt an sekundären Pflanzenstoffen. Gleichzeitig geht eine hohe Kohlenhydratqualität mit einem geringeren Verzehr an hochverarbeiteten Lebensmitteln und damit z. B. mit einer geringeren Aufnahme von Zucker und gesättigten Fetten einher. Eine hohe Kohlenhydratqualität hat daher unabhängig von der Regulation der Glykämie langfristige Effekte auf die Prävention von Diabetes und dessen Komplikationen.
Ballaststoffe
Ballaststoffe allgemein
Empfehlung.
Verschiedene Ballaststoffe aus natürlichen Quellen sollen täglich verzehrt werden.
Auch wenn es bislang nur eine geringe Evidenz für die Empfehlung von 30 g Ballaststoffen pro Tag (15 g/1000 kcal) gibt, stellt dies für die Beratung eine valide Zielgröße dar.
Kommentar
In Kohortenstudien ist eine hohe Zufuhr von unlöslichen Ballaststoffen, insbesondere cerealen Ursprungs, mit einem erniedrigten Risiko für T2Dm, KHK, Krebs und weitere Erkrankungen assoziiert [51, 176–178]. Auch bei Patienten mit T2Dm zeigt sich eine dosisabhängige Reduktion des Sterblichkeitsrisikos [179]. Somit stellen für den T2Dm v. a. Vollkornprodukte (Brot, Reis, Nudeln) eine protektive Lebensmittelgruppe dar. Metaanalysen zeigen für eine ballaststoffreichere Ernährung oder Ballaststoffsupplemente selbst unter isokalorischen Bedingungen signifikante Vorteile für Körpergewicht, Glykämie und Insulinresistenz, Lipidprofil und Entzündungsstatus [180], mitunter auch für den Blutdruck [181]. Auch wenn Ballaststoffe den glykämischen Index senken, so ist dieser offenbar ein zu ungenauer Indikator für empfehlenswerte Lebensmittel [180]. Die Betonung von „Vollkorn“, noch besser der tatsächlichen Ballaststoffzufuhr, hat die beste Aussagekraft. Ausgehend von einem durchschnittlichen Ernährungsmuster mit 20 g Ballaststoffen wird eine Erhöhung um 15 g auf 35 g pro Tag angestrebt [180].
Aufgrund der Heterogenität der Studien, die u. a. aus der Vielzahl an Ballaststoffen, ballaststoffhaltigen Lebensmitteln, Kohorten und Interventionen (Vollkorn, nicht cereale Produkte, fortifizierte Lebensmittel, Supplemente …) resultiert, ist aber eine weitere Differenzierung dieser Ergebnisse notwendig [51, 180].
Unlösliche Ballaststoffe
Empfehlung.
Kohlenhydrate sollten bevorzugt aus ballaststoffreichen Lebensmitteln, insbesondere Vollkornprodukten, bezogen werden. Der Nutzen einer Supplementation ist bislang nicht belegt.
Kommentar
Interventionsstudien mit Vollkornprodukten zeigen zumindest für Reis, jedoch nicht für Weizen- und Roggenprodukte einen glykämischen Vorteil [182]. Neben einem kleinen Effekt auf das Körpergewicht sind in Metaanalysen keine kardiometabolischen Benefits beschrieben, die eindeutig Vollkornprodukten zugeordnet werden können [183]. Studien, die explizit unlösliche Ballaststoffe im Interventionsdesign untersuchen, gibt es nur wenige [184–186], jedoch bislang keine an Patienten mit T2Dm. Die bisherigen Daten deuten jedoch an, dass die Wirksamkeit von Ballaststoffen bei einer stärkeren Einschränkung des Metabolismus umso ausgeprägter ist [186, 187].
Lösliche Ballaststoffe
Empfehlung.
Ballaststoffreiche Lebensmittel, insbesondere Vollkornprodukte, aber auch Gemüse, Hülsenfrüchte und zuckerarmes Obst sind bei T2Dm empfehlenswert und wahrscheinlich metabolisch von Vorteil. Der Langzeitnutzen einer Supplementation ist trotz konsistenter Kurzzeiteffekte für Glykämie, Lipidstatus und ggf. Blutdruck nicht belegt.
Kommentar
Für lösliche Ballaststoffe gibt es epidemiologisch nur unzureichende Hinweise für einen Langzeitnutzen, sowohl hinsichtlich Morbidität als auch Mortalität.
Im Gegensatz zu unlöslichen Ballaststoffen ist die Erforschung der löslichen Fasern besonders in Form von Supplementationsstudien aber deutlich weiter fortgeschritten. Für Beta-Glucane und Psyllium (Flohsamen) ist daher ein zumindest kurz- bis mittelfristiger (Wochen bis Monate) Benefit auf Blutglukose und Insulinresistenz belegt; Langzeitdaten fehlen aber [188]. Auch für Inulin (spezielle Fruktane) sind günstige Wirkungen auf Glykämie und Insulinämie systematisch beschrieben, v. a. für Frauen und adipöse Menschen mit T2Dm [189, 190]. Studien von mehr als 3 Monaten Interventionsdauer sind aber auch dafür rar.
Die glykämischen Vorteile von Inulin und Psyllium beruhen vermutlich auf der Fermentation zu kurzkettigen Fettsäuren, nicht auf einer Gewichtsreduktion [191]. Für Beta-Glucane liegt möglicherweise ein gemischter Effekt vor [192].
Psyllium, Konjak-Glucomannan, aber auch Beta-Glucane senken zudem den LDL-Cholesterin- und Triglyzeridspiegel moderat und können daher bei T2Dm einen Sekundärnutzen erbringen [193–196]. Für andere lösliche Fasern (Guar, Pektin) sind keine eindeutigen metabolischen Vorteile belegt [197].
Antihypertensive Effekte sind im Mittel für alle viskösen Fasern beschrieben, jedoch v. a. bei Psyllium zu erwarten. Der Effekt ist mit 2 mm Hg systolisch und 0,5 mm Hg diastolisch klinisch kaum relevant [198].
Ernährungsaspekte spezieller Bevölkerungsgruppen
Geriatrische Patienten
Empfehlung.
Die Ziele in der Ernährungstherapie von geriatrischen Patienten sollen sich auf den Erhalt der Selbstständigkeit und auf die Vermeidung einer Mangelernährung und von Hypoglykämien fokussieren.
Übergewicht ist in dieser Personengruppe mit einer reduzierten Mortalität verbunden und sollte nicht reduziert werden.
Kommentar
Grundsätzlich unterscheiden sich die Ernährungsempfehlungen für ältere Menschen mit T2Dm nicht von denen für ältere Stoffwechselgesunde oder jüngere Menschen mit T2Dm. Gleichzeitig gelten für geriatrische Patienten mit T2Dm die allgemeinen Ernährungsempfehlungen für diese Patientengruppe. Insbesondere bei funktionell abhängigen Patienten sind die Folgen einer Mangelernährung im Alter gravierend und sollten auch bei Patienten mit T2Dm fokussiert werden. So verstärkt der mit einer Gewichtsabnahme verbundene Verlust von Muskelmasse die altersbegleitende Sarkopenie und Gebrechlichkeit und begünstigt dadurch Behinderungen und Einbußen der Selbstständigkeit.
Die S2k-Leitlinie „Diagnostik, Therapie und Verlaufskontrolle des Diabetes im Alter“ enthält sehr ausführliche Empfehlungen auch zur Ernährungstherapie von älteren Personen mit DM im Allgemeinen. Darin wird verdeutlicht, dass sich Therapieziele – auch in Bezug auf die Ernährung – bei älteren und insbesondere geriatrischen Patienten häufig verändern können, aber nicht müssen. Funktionalität und der Erhalt der Selbstständigkeit stehen im Vordergrund.
Es konnte zwar auch bei älteren Menschen durch eine beabsichtigte Gewichtsreduktion eine Verbesserung der Insulinsensitivität erreicht werden [199], allerdings soll bei älteren Menschen mit Übergewicht oder Adipositas aufgrund des Mangelernährungsrisikos auf strenge Diätvorschriften verzichtet werden. Diätvorschriften, die die Nahrungsaufnahme limitieren können, sind potenziell schädlich und sollten vermieden werden. Sollte eine Gewichtsabnahme erwogen werden, sollten die Diätmaßnahmen, wenn immer möglich, mit körperlicher Aktivität kombiniert werden und die bedarfsdeckende Eiweißaufnahme im Fokus haben. Ein signifikanter Anstieg der Mortalität fand sich bei über 65-Jährigen erst ab einem BMI von über 30 kg/m2 [199]. Einschränkungen des Verzehrs gewohnter und liebgewonnener Lebensmittel führen zu einer Verminderung der subjektiv empfundenen Lebensqualität. Insbesondere bei Personen im hohen Lebensalter ist dieser Aspekt von entscheidender Bedeutung.
Das Risiko für eine potenzielle Mangelernährung liegt vor bei anhaltender reduzierter Nahrungsaufnahme (ca. < 50 % des Bedarfs für mehr als 3 Tage) oder wenn mehrere Risikofaktoren gleichzeitig vorliegen, die entweder die Essmenge reduzieren oder den Energie- und Nährstoffbedarf nennenswert erhöhen. Das Risiko der Mangelernährung kann z. B. mittels Mini Nutritional Assessment (MNA) oder der entsprechenden Kurzform (SF-MNA) erfasst werden; beide Screeningmethoden sind gut evaluiert [200, 200]. Bei untergewichtigen Patienten sollten die Ursachen geklärt und, wenn möglich, behoben werden.
Die Ernährungstherapie sollte sich auch auf die Vermeidung von Hypoglykämien fokussieren. Gegebenenfalls muss bei Nahrungsumstellungen kurzfristig im Sinne einer Therapiedeeskalation eine Medikamentenanpassung vorgenommen werden.
Für weitere Ausführungen insbesondere für Personen mit Diabetes in Pflegeeinrichtungen und bei Notwendigkeit einer künstlichen Ernährung wird auf die S2k-Leitlinie „Diagnostik, Therapie und Verlaufskontrolle des Diabetes im Alter“ und die S3-Leitlinie „Klinische Ernährung in der Geriatrie“ verwiesen [201–203].
Aufgrund der Komplexität der häufig multimorbiden geriatrischen Patienten sollten Planung und Umsetzung krankheitsspezifischer Ernährungsweisen im Bedarfsfall durch ein multiprofessionelles Team unter Einbeziehung von ernährungsspezifischem Sachverstand erfolgen.
Migranten
Empfehlung.
Behandler sollen sicherstellen, dass die Patienten die Ernährungshinweise verstanden haben und ihre Kernfamilien in die Therapie miteinbezogen werden.
Behandler sollen das individuelle Ernährungskonzept des Patienten und seines Umfelds (beispielsweise religiöse Aspekte, kulturelle Überzeugungen, Fastenmonat Ramadan, Schwangerschaft) erheben und berücksichtigen.
Kommentar
Bezüglich der spezifischen Therapie- und Ernährungsaspekte von Migranten wird auf die DDG-Praxisempfehlung Diabetes und Migration verwiesen [204].
Es bestehen teilweise sehr individuelle Essgewohnheiten im Rahmen unterschiedlicher Kulturen und Regionen. Esskultur wird von geografischen, historischen, soziologischen, ökonomischen und psychologischen Merkmalen einer Gesellschaft geformt und wird von den entsprechenden Mitgliedern einer bestimmten Gemeinschaft geteilt. Kultur stellt eine grundlegende Determinante zu „was wir essen“ dar [205]. Migranten haben häufig ein anderes Ernährungsverhalten als Einheimische. Sie bevorzugen teilweise andere Lebensmittel, ernähren sich häufig vermehrt von Kohlenhydraten, haben andere Mahlzeitenkonzepte, ein anderes Portionsverständnis sowie andere Essenszubereitungsformen und Lebensmittelkombinationen. Ihre Ernährungskonzepte beruhen in der Regel auf der eigenen traditionellen Küche, persönlichen Gewohnheiten, und sie übernehmen auch die Essgewohnheiten der einheimischen Bevölkerung, oft resultiert eine neue „Mischküche“ [206]. Nicht selten werden spezielle Lebensmittel aus den Heimatländern besorgt. Migranten aus einigen Kulturen können beim Kochen mit den Gewichtsangaben in hiesigen Rezepten wenig anfangen. Menschen haben eine hoch variable postprandiale Glukoseantwort auf identische Nahrungsmittel. Eine individualisierte kultursensible Beratung verbessert die Compliance [207]. In diesem Kontext spielen das Fasten im Ramadan – religiös beeinflusste Speisenauswahl und Fastenvorschriften –, die Schwangerschaft und die Schichtarbeit eine besondere Rolle. Im Praxisalltag ist das Wissen um die Hauptlieferanten von Kohlenhydraten und in welcher Form und wann die Kohlenhydrate gegessen werden unentbehrlich. Das seitens der AG Diabetes und Migranten der DDG erstellte Praxis-Tool zur Ernährung [204, 208] von Migranten soll eine erste Information und Hilfestellung geben. Eine pragmatische regionale Aufteilung mit Angaben zur gängigen Küche stellt die Basis dar. Neben der Art (warm/kalt) und der Zahl der Mahlzeiten werden die Hauptlieferanten von Kohlenhydraten und weitere regionale Besonderheiten vorgestellt. Die Küchen sind weltweit ziemlich vielfältig, und regional ist ebenso eine große Verschiedenheit vorzufinden. Dennoch ist zu berücksichtigen, dass viele Getränke in der Zwischenzeit weltweit in viele Esskulturen vorgedrungen sind, beispielsweise Softdrinks, Energydrinks, mit Süßstoff angereicherte verschiedene Getränke und einige Biersorten.
Eine mögliche Sprachbarriere und kultursensible Kommunikation sollten bei der Ernährungsberatung berücksichtigt werden [202]. Daher verbessert eine individualisierte, kultursensible Beratung die Compliance und den Therapieerfolg.
Ernährungsaspekte spezieller Lebensmittel und Nahrungsergänzungsmittel
Getränke
Empfehlung.
Personen mit T2Dm sollen die Zufuhr zuckergesüßter Getränke minimieren.
Kommentar
Die aktuellen evidenzbasierten Leitlinien der amerikanischen und der britischen Diabetesgesellschaft empfehlen allgemein für Personen mit Diabetes eine Reduktion des Konsums zuckergesüßter Getränke, um den Blutglukosespiegel und das Körpergewicht zu kontrollieren und das Risiko für kardiovaskuläre Erkrankungen und eine Fettleber zu reduzieren (Evidenzgrad B bzw. 2) [2, 119, 120]. Eine Reduktion des Konsums zuckergesüßter Getränke ist zudem allgemein erstrebenswert, da sie zu einer erhöhten Mikronährstoffdichte, einer Reduktion der Zufuhr zugesetzter Zucker und somit insgesamt zu einer ausgewogeneren Ernährung beiträgt [209].
Die Evidenz für den Zusammenhang zwischen dem Konsum zuckergesüßter Getränke, der glykämischen Kontrolle und der Insulinsensitivität/-resistenz wird für Erwachsene (unabhängig vom Diabetesstatus) basierend auf Kohortenstudien und RCTs allerdings als ungenügend bewertet, sodass keine fundierten Schlussfolgerungen gezogen werden können [210]. Eine Metaanalyse aus 11 Kohortenstudien zeigt für Personen ohne Diabetes eine Assoziation zwischen einer höheren Zufuhr zuckergesüßter Getränke mit höheren Nüchternblutglukose- und Insulinkonzentrationen nach Adjustierung für mögliche Konfounder [211]. Speziell für die Effekte fruktosehaltiger zuckergesüßter Getränke auf die glykämische Kontrolle und die Serumlipidkonzentrationen untersuchten 2 systematische Reviews und Metaanalysen die Effekte einer isokalorischen Substitution von Glukose oder Saccharose durch Fruktose in Getränken und festen Lebensmitteln. Sowohl eine kurzfristige als auch eine chronische (Studiendauer 2 bis 10 Wochen) Substitution zeigten keine negativen Effekte von Fruktose auf die maximale postprandiale Blutglukose, Insulin- oder Triglyzerid-Konzentrationen bzw. die Nüchternblutglukose‑, Insulin- oder Triglyzeridkonzentrationen bei Personen mit Normoglykämie, Prädiabetes und T2Dm [212, 213]. Bei der Interpretation dieser Ergebnisse ist jedoch zu beachten, dass nur für die kurzfristige Substitution bei normoglykämischen Personen eine Subgruppenanalyse für den Effekt von zuckergesüßten Getränken vs. zuckergesüßten festen Lebensmitteln durchgeführt wurde [213] und die Subgruppenanalysen für Personen mit T2Dm in beiden Untersuchungen nur auf einer sehr geringen Zahl von Studien basierten [212, 213].
Auch für den Zusammenhang zwischen der Zufuhr zuckergesüßter Getränke und diabetesassoziierten makrovaskulären Komplikationen wie koronaren Ereignissen, Schlaganfall, Bluthochdruck und Dyslipidämie wird die Evidenz allgemein für Erwachsene als ungenügend bewertet [210]. Systematische Reviews (und Metaanalysen) basierend auf 4 bis 11 prospektiven Kohortenstudien weisen auf Assoziationen zwischen dem Konsum zuckergesüßter Getränke und vaskulären Risikofaktoren (Bluthochdruck, Hyperlipidämie), koronaren Herzerkrankungen, Schlaganfall und Mitralklappeninsuffizienz hin [214–216]. Zu beachten ist jedoch, dass die Ergebnisse nicht spezifisch für Personen mit T2Dm sind [214–216]. Für den Zusammenhang zwischen zuckergesüßten Getränken und koronaren Herzerkrankungen wurden in den 2 Studien mit Personen mit Diabetes keine signifikanten Effekte beobachtet [215], und Analysen für Diabetes als Mediator für den Zusammenhang zwischen zuckergesüßten Getränken und vaskulären Risikofaktoren ergaben inkonsistente Ergebnisse [214].
In Bezug auf diabetesassoziierte mikrovaskuläre Erkrankungen ergab eine weitere Metaanalyse basierend auf 5 Studienpopulationen (ebenfalls nicht ausschließlich Personen mit T2Dm) eine signifikante Assoziation zwischen dem chronischen Konsum zuckergesüßter Getränke und chronischer Nierenerkrankung. Allerdings waren die inkludierten Studien sehr heterogen, und es lag Evidenz für Publikationsbias vor [217].
Zwei systematische Reviews und Metaanalysen basierend auf 4 bzw. 12 Kohortenstudien (teilweise Personen mit T2Dm inkludiert) zum Zusammenhang zwischen dem Verzehr zuckergesüßter Getränke und nichtalkoholischer Fettleber zeigten ein signifikant höheres Risiko für eine nichtalkoholische Fettleber für die höchste vs. niedrigste Zufuhrkategorie zuckergesüßter Getränke [218, 219]. Bereits die niedrigste Zufuhr von < 1 Glas/Woche war mit einem Anstieg des relativen Risikos für eine nichtalkoholische Fettleber von 14 % assoziiert, und der Konsum zuckergesüßter Getränke zeigte einen dosisabhängigen Effekt auf das Risiko für eine nichtalkoholische Fettleber [218].
Zusammenfassend lässt sich für Personen mit T2Dm – entsprechend der Empfehlung für die Allgemeinbevölkerung – ableiten, dass eine Reduktion der Zufuhr zuckergesüßter Getränke im Rahmen einer ausgewogenen Ernährung angestrebt werden soll, um das Risiko für kardiometabolische Begleiterkrankungen zur reduzieren [2, 119, 120, 209, 210].
Wissenschaftlicher Hintergrund
Bei der Interpretation der Daten zu den Effekten zuckergesüßter Getränke auf die einzelnen diabetesrelevanten Zielparameter sind die folgenden Punkte zu berücksichtigen: i. Die Mehrheit der Studien untersucht nicht ausschließlich Personen mit T2Dm, sodass weitere Studien in dieser Patientengruppe notwendig sind, die die Übertragbarkeit der Ergebnisse bestätigen; ii. die meisten Assoziationen für zuckergesüßte Getränke sind nur für den Vergleich der extremen Zufuhrkategorien signifikant, nicht jedoch für moderate Zufuhrlevel, die jedoch etwa der mittleren geschätzten weltweiten Zufuhr zuckergesüßter Getränke entsprechen [220]; die Effekte der Zufuhr zusätzlicher Zucker auf die Zielparameter scheinen einerseits von der Energiebilanz und andererseits von der Zuckerquelle abzuhängen, da insbesondere zuckergesüßte Getränke, die überschüssige Energie liefern, einen negativen Effekt beispielsweise auf die Nüchternblutglukose- und Insulinkonzentrationen zu haben scheinen [221]. Weiterhin scheint die direkte Assoziation zwischen der Zufuhr fruktosehaltiger und allgemein zuckergesüßter Getränke mit dem erhöhten Risiko beispielsweise für die Inzidenz des metabolischen Syndroms und anderer kardiometabolischer Risikofaktoren und Ereignisse auf zuckergesüßte Getränke beschränkt und nicht auf die Zufuhr von Zucker aus anderen Quellen (beispielsweise Obst, Joghurt, Fruchtsäfte) übertragbar zu sein [220, 222]. Mögliche Erklärungen für diese Beobachtung sind, dass der Effekt zuckergesüßter Getränke stark durch die zusätzliche Energiezufuhr und die daraus resultierende Gewichtszunahme mediiert zu sein scheint, dass andere Fruktose- bzw. Zuckerquellen zusätzliche potenziell gesundheitsfördernde Inhaltsstoffe enthalten, was auf zuckergesüßte Getränke nicht zutrifft, und dass zuckergesüßte Getränke einen Marker für einen insgesamt ungesünderen Lebensstil darstellen [220].
Vollkorn
Empfehlung.
Bei übergewichtigen Patienten mit T2Dm kann eine an Vollkornprodukten reiche Ernährung dazu beitragen, die Gesamtenergieaufnahme zu senken und damit eine intendierte Gewichtsreduktion zu unterstützen.
Der Verzehr wenig verarbeiteter Vollkornprodukte mit einem hohen Anteil ganzer Körner führt zu einer geringer ausgeprägten postprandialen Blutglukoseantwort, was insbesondere für Menschen mit T2Dm ohne Insulinresistenz eine nichtmedikamentöse Therapieoption sein kann.
Insulinbehandelte Menschen mit T2Dm sollen den Verzehr von Vollkornprodukten primär mengenmäßig nach dem Gehalt an Kohlenhydrateinheiten (KE) und zusätzlich nach dem glykämischen Index berücksichtigen und auf ihre Insulintherapie abstimmen.
Hochverarbeitete Vollkornprodukte zeigen keine zusätzlichen günstigen Effekte auf die postprandiale Blutglukoseantwort.
Kommentar
Für die Allgemeinbevölkerung wird empfohlen, Vollkornprodukte zu wählen [223]. Dies wird mit ihrem höheren Gehalt an Vitaminen, Mineralstoffen und sekundären Pflanzenstoffen begründet sowie mit günstigen Effekten auf Verdauung und Darmgesundheit durch die assoziierte höhere Ballaststoffaufnahme. Darüber hinaus zeigen langjährige Kohortenstudien [224, 225] und zahlreiche Metaanalysen/Reviews von Kohortenstudien Assoziationen eines deutlich erhöhten Vollkornverzehrs mit einem um bis zu 20 % reduzierten Risiko für kardiovaskuläre Erkrankungen und Mortalität [226–232]. Daraus ergeben sich Empfehlungen von Autoren, dass schon „moderate Steigerungen des Vollkornverzehrs das Risiko vorzeitigen Todes reduzieren könnten“ [230]. Allerdings ist eine Kausalbeziehung noch nicht geklärt. In den Studien basieren die zugrunde gelegten Daten zur Ernährung oft nur auf einer Erhebung (3-Tage-Protokoll oder Food Frequency Questionnaire zu Beginn der Kohortenstudie), und die Klassifizierungen von Lebensmitteln als „Vollkornlebensmittel“ sind uneinheitlich.
Hinsichtlich der Diabetestherapie ist der Verarbeitungsgrad von Vollkornprodukten von Bedeutung. Bereits im Jahr 1988 publizierten Jenkins et al. Ergebnisse zur postprandialen Blutzuckerantwort nach dem Verzehr von Vollkornbroten mit unterschiedlichen Verhältnissen im Gehalt an Vollkornmehl und ganzen Getreidekörnern. Die Blutzuckerantwort wird weniger durch die allgemeine Vollkorneigenschaft eines gemahlenen Getreideprodukts („wholemeal“) als vielmehr durch den Anteil darin enthaltener ganzer Körner („wholegrain“) bestimmt [233]. Je höher der Anteil ganzer Körner, desto geringer die Blutglukoseantwort, da die Frucht- und Samenschalen eine physikalische Barriere für die Einwirkung der Amylase auf den Mehlkörper bilden.
Dreißig Jahre später wurden diese Ergebnisse bezüglich des Einflusses des Verarbeitungsgrades jüngst unter experimentellen [234] sowie unter Alltagsbedingungen [235] bestätigt. Für den bloßen Zusatz von Weizenkleie zu den üblichen Speisen mit dem Ziel der Erhöhung des Ballaststoffgehalts wurden keine positiven Effekte auf die Diabetesbehandlungssituation gezeigt [236].
Für Menschen mit T2Dm werden die Empfehlungen für die unterschiedlichen Behandlungssituationen und -formen differenziert:
Bei übergewichtigen Patienten mit T2Dm: Eine Metaanalyse zu Ballaststoff- und Vollkornverzehr im Diabetesmanagement bezog 42 Interventionsstudien ein. Danach wurden für gesteigerten Ballaststoff‑/Vollkornverzehr – im Vergleich zu Kontrollgruppen – ein um ½ kg geringeres Körpergewicht und eine daraus resultierende Reduktion des HbA1c-Werts um 0,2 % (2 mmol/mol) dargestellt [177]. Manko dieser Analyse sind die heterogenen Designs der einbezogenen Studien, u. a. hinsichtlich Diabetesmedikation, Studiendauer, Diabetesdiagnose und Art der Vollkornzufuhr.
Bei nicht insulinbehandelten, normalgewichtigen Patienten mit T2Dm (ohne Insulinresistenz) kann der Verzehr wenig verarbeiteter Vollkornprodukte mit einem hohen Anteil ganzer Körner zu einer geringer ausgeprägten postprandialen Blutglukoseantwort führen. Positive Effekte einer solchen diätetischen Maßnahme auf das Erreichen des Therapieziels sind u. a. abhängig von der Akzeptanz dieser Ernährungsform durch den Patienten sowie mittelfristig vom Fortbestand der Restfunktion der β‑Zellen.
Insulinbehandelte Menschen mit T2Dm sollen die blutglukosesteigernde Wirkung ihrer Ernährung einschätzen, um die Insulindosierung darauf abzustimmen. Entsprechend sollen sie den Verzehr von Vollkornprodukten primär mengenmäßig nach KE-Gehalt und zusätzlich nach glykämischem Index berücksichtigen und auf ihre Insulintherapie abstimmen. Vollkornprodukte können gemäß den eigenen Präferenzen verzehrt werden.
Hochverarbeitete Vollkornprodukte zeigen keine zusätzlichen günstigen Effekte auf die postprandiale Blutglukoseantwort.
Obst, Gemüse
Empfehlung.
In der Ernährung des übergewichtigen Patienten mit T2Dm kann insbesondere ein gesteigerter Gemüseverzehr eine intendierte Gewichtsreduktion unterstützen.
In der Ernährung des normalgewichtigen Patienten mit T2Dm soll die Aufnahme großer Portionen an Obst(produkten) und stärkereichem Gemüse (Kartoffeln, Mais, Reis, Getreide u. a.) vermieden werden.
Insulinbehandelte Menschen mit T2Dm sollen den Verzehr von Obst mengenmäßig nach KE-Gehalt berücksichtigen und auf ihre Insulintherapie abstimmen.
Eine Trennung in empfehlenswerte und nicht empfehlenswerte Obstsorten wird nicht als sinnvoll angesehen.
Kommentar
Für die Allgemeinbevölkerung wird unter dem Slogan „5 am Tag“ ein täglicher Verzehr von mindestens 3 Portionen Gemüse (400 g) und 2 Portionen Obst (250 g) empfohlen [223]. Jüngere Ergebnisse der PURE-Studie [237] sowie von Metaanalysen/Reviews von Kohortenstudien [231, 238–240] zeigen Assoziationen eines erhöhten Obst- und Gemüseverzehrs mit einem um 5–20 % reduzierten Risiko bezüglich kardiovaskulärer Erkrankungen und Gesamtmortalität. Allerdings ist eine Kausalbeziehung noch nicht geklärt und die Datenlage hinsichtlich der wirksamen Obst- und Gemüsesorten, der täglichen Mindestverzehrmengen sowie des Ausmaßes der klinischen Relevanz hinsichtlich der spezifischen Erkrankungen und Mortalitäten uneinheitlich. Über die individuelle Gesundheit hinausgehend, werden aus ökologischen und sozialen Gründen von der EAT-Lancet Commission im Rahmen einer Planetary Health Diet vergleichbare Empfehlungen zum Gemüse- und Obstverzehr ergänzt um täglich ca. 100 g Leguminosen/Sojaprodukte gegeben [241].
Für Menschen mit T2Dm werden die Empfehlung für die unterschiedlichen Behandlungssituationen und -formen differenziert:
Bei übergewichtigen Patienten mit T2Dm ist der Obst- und Gemüseverzehr als unterstützende Komponente zur Gewichtsreduktion zu sehen. Wenn energiedichte Lebensmittel durch den sinnvollen Verzehr von Obst und den erhöhten Verzehr von Gemüse ersetzt werden, kann dies eine Gewichtsreduktion nachhaltig unterstützen. Interventionsstudien zu den singulären Effekten einzelner Lebensmittel(gruppen) auf Körper- bzw. Blutparameter existieren nicht oder lassen wegen der vielfältigen zusätzlichen Einflussfaktoren keine kausalen Aussagen zu. Allerdings haben Interventionsstudien bei Menschen mit T2Dm zu den Effekten einer insgesamt pflanzenbetonten Ernährung – die reich an Obst und insbesondere Gemüse ist – eine deutliche Reduktion des Körpergewichts gezeigt mit entsprechenden positiven Effekten auf die Glykämiesituation [123, 242, 243].
Bei nicht insulinbehandelten normalgewichtigen Patienten mit T2Dm sollen große KH-Mengen zu einzelnen Mahlzeiten vermieden werden, um starke postprandiale Blutzuckerantworten zu vermeiden. Daher sind große Mengen an Obst, Obstsäften und stärkereichem Gemüse nicht zu empfehlen (klinische Erfahrung). Für nicht stärkehaltige Gemüse gibt es keine einschränkende Mengenempfehlung für den Verzehr.
Insulinbehandelte Menschen mit T2Dm sollen die blutglukosesteigernde Wirkung ihrer Ernährung einschätzen, um die Insulindosierung darauf abzustimmen. Entsprechend soll der Verzehr von Obst und stärkereichem Gemüse (Kartoffeln, Süßkartoffeln) auf den Kohlenhydratgehalt nach KE abgeschätzt und die eigene Form der Insulintherapie abgestimmt werden. Obst und Gemüse können gemäß den eigenen Präferenzen verzehrt werden.
Allgemein ist zu beachten, dass sich durch Obstsäfte, Smoothies und Trockenobst in kurzer Zeit große Mengen an Kohlenhydraten aufnehmen lassen – verglichen mit unverarbeitetem frischem Obst.
Aufgrund der Datenlage gibt es keine Evidenz für eine pauschale Trennung in empfehlenswerte und nicht empfehlenswerte Obstsorten, was wegen des unterschiedlichen Kohlenhydratgehalts in Laienpublikationen immer wieder popagiert wird.
Fisch
Empfehlung.
Fetter Fisch kann einen Beitrag zur Senkung der Blutfette und des inflammatorischen Phänotyps und damit möglicherweise des kardiovaskulären Risikos leisten.
Die Evidenz zur Empfehlung von Fischölsupplementen bei T2Dm reicht nicht aus.
Bei der Auswahl der Fischmahlzeiten soll auf eine nachhaltige Fischerei/Fischzucht geachtet werden [244].
Kommentar
Ernährungsmuster, die Fisch einschließen, sind in Beobachtungsstudien mit einem geringeren Diabetesrisiko verknüpft [245]. Der Verzehr von Fisch an sich, aber auch Fischölen (langkettige Omega-3-Fettsäuren wie Docosahexaensäure [DHA] und Eicosapentaensäure [EPA]) ist epidemiologisch jedoch uneinheitlich mit dem Diabetesrisiko assoziiert. In westlichen Regionen (Nordamerika, Europa) besteht ein Trend zur Risikosteigerung, im Pazifikraum zur Risikosenkung [246–249]. Diese Assoziationen stehen in Diskrepanz zu Kohortenstudien, die Fischverzehr dosisabhängig mit einem deutlich geringeren Risiko für viszerale Adipositas verknüpfen [250] sowie ein niedrigeres kardiovaskuläres Risiko und eine geringere kardiovaskuläre und Gesamtmortalität angeben [239, 251, 252]. Zum Bluthochdruck besteht keine signifikante Beziehung [50].
Auch der Nutzen bezüglich des kardiovaskulären Risikos ist umstritten. Metaanalysen von RCTs sehen diskrete oder nicht signifikante Effekte [253, 254]. Eine Metaanalyse spezifisch für T2Dm-Patienten ist noch nicht publiziert.
In Interventionsstudien sind die spezifischen Effekte von Fischverzehr kaum untersucht. Fischöle scheinen bei Patienten mit metabolischem Syndrom – nicht aber bei Gesunden – die Insulinsensitivität zu verbessern [255]. Dieser Effekt ist geschlechterspezifisch bei Frauen nachgewiesen, für Männer fehlt es an Daten [81]. Daten zur Diabetesinzidenz gibt es keine. Glykämische Parameter bessern sich unter Supplementation nicht [256].
Ein metabolischer Nutzen durch Supplementation mit Fischöl ist am ehesten bezüglich der Triglyzeride und des C‑reaktiven Proteins (CRP) zu erwarten [75, 256]. Für nicht inflammatorische Benefits von Vorteil ist dabei ein hohes EPA/DHA-Verhältnis [256].
Fleisch
Empfehlung.
Teilweise erweisen sich High-Protein-Diäten bezüglich der Glykämie als vorteilhaft bis möglicherweise überlegen (s. oben). Der darin vorgenommene Austausch von Kohlenhydraten gegen Eiweißquellen kann auch anteilig aus tierischen Quellen inklusive Fleisch jeder Art gedeckt werden.
Fleischkonsum soll auch unter Umweltaspekten (u. a. zur Reduzierung des Landverbrauchs oder der Treibhausgasemissionen) auf das empfohlene Maß der Deutschen Gesellschaft für Ernährung (DGE) reduziert werden [244, 257].
Kommentar
Eine fleischbetonte und damit in der Regel kohlenhydratreduzierte Ernährung steht in Beobachtungsstudien mit einer erhöhten (kardiovaskulären) Mortalität in Verbindung [258]. Epidemiologisch bestehen zudem moderate Beziehungen mit Krebserkrankungen, KHK sowie T2Dm. Besonders akzentuiert sind diese Assoziationen mit rotem Fleisch, v. a. verarbeitetem rotem Fleisch [149].
Interventionsstudien zeigen bei Reduktion der täglich aufgenommenen Fleischmenge eine Verbesserung zahlreicher metabolischer Parameter. Da in diesen Studien entweder ein isokalorischer Ausgleich mit anderen potenziell günstigen Lebensmitteln (z. B. Vollkorn, Gemüse, Hülsenfrüchte, Nüsse) erfolgt oder Fleischverzicht in einem hypokalorischen Setting umgesetzt wird, ist auch in RCTs die Kausalität für den Nutzen der fleischarmen Ernährung unklar.
RCTs zum Austausch von Fleischsorten (rotes gegen weißes Fleisch) stehen zumeist unter dem gleichen Confounding (z. B. rotes Fleisch = Standardernährung vs. weißes Fleisch = mediterrane Ernährung). Ein relevanter Interventionseffekt auf Mortalität und Morbidität (inclusive T2Dm-Inzidenz) ist fraglich [258]. Lediglich 6 RCTs haben explizit rotes und weißes Fleisch miteinander verglichen und zeigen bei den untersuchten nichtdiabetischen Probanden keinen metabolischen Unterschied [259–264].
Das NutriRECS-Konsortium kam 2019 auf dieser Datengrundlage zu dem Schluss, aufgrund mangelnder Evidenz keine Empfehlung zur Fleischreduktion auszusprechen [265]. Die Beurteilung der vorliegenden ernährungswissenschaftlichen Evidenz durch das NutriRECS-Konsortium offenbart jedoch die häufige, aber fehlerhafte Annahme, dass medizinische und ernährungswissenschaftliche Forschung nach den gleichen Kriterien zu evaluieren seien. So werden Beobachtungsstudien systematisch ab- und randomisierte, kontrollierte Studien (RCT) sehr hoch gewertet. Langfristige RCTs mit Lebensmitteln, insbesondere mit Verblindung und Placebokontrolle, sind jedoch im Ernährungsbereich sehr schwer durchführbar. Insgesamt ist die Empfehlung, auf (rotes) Fleisch zu verzichten, derzeit aus ökologischer und tierethischer Sicht noch deutlich besser begründet als durch die Stoffwechselforschung.
Zimt
Empfehlung.
Der Verzehr von Zimt kann Menschen mit T2Dm als Komponente einer erfolgreichen Diabetestherapie nicht empfohlen werden.
Kommentar
In den letzten 15 Jahren sind zahlreiche Interventionsstudien zu den Effekten von Zimtverzehr auf Nüchternblutzucker und HbA1c-Wert bei Menschen mit T2Dm publiziert worden. Trotz uneinheitlicher Studienergebnisse wurden immer wieder vorteilhafte Effekte von Zimtverzehr auf den Behandlungserfolg bei T2Dm verbreitet. Zwei Metaanalysen aus den Jahren 2011 und 2012 postulierten in ihren Abstracts für Zimt positive Effekte auf den Nüchternblutzucker [266, 267] sowie das HbA1c [267], wobei in dieser Arbeit gleichzeitig geschlussfolgert wird, dass die Mehrzahl der untersuchten Studien keinen relevanten therapeutischen Effekt auf die Glykämie von Menschen mit T2Dm zeigten. Zwei Metaanalysen in den darauffolgenden Jahren haben die verfügbaren Studien bis Anfang 2012 in ihre Untersuchungen einbezogen, wobei die Cochrane-Arbeit [268] Studien mit fragwürdiger Qualität von der Analyse ausgeschlossen hatte. Beide Studien konstatieren keinen signifikanten Effekt von Zimtverzehr auf den HbA1c-Wert. Allen et al. [269] zeigten positive Therapieeffekte auf den Nüchternblutzucker, relativierten dies aber aufgrund deutlicher methodischer Defizite der untersuchten Studien. Zwei weitere jüngere Reviews [270, 271] kommen zu dem Schluss, dass die Einnahme von Zimt (als Adjuvans) in der Therapie des T2Dm angesichts der aktuellen Studienlage nicht empfohlen werden kann. Methodische Probleme schränken Aussagekraft und Vergleichbarkeit der Studien außerordentlich ein: So werden zwar immer die eingesetzten Zimttagesdosen in den Interventionsgruppen der Studien angegeben (0,1 bis 6,0 g/Tag), aber es existieren keine, unvollständige oder uneinheitliche Angaben zur untersuchten Zimtsorte (C. cassia, C. aromaticum, C. zeylanium), zur Applikationsform (Zimtpulver, Zimtextrakt, Kapseln, Tabletten), zur Menge des getesteten aktiven Zimtwirkstoffs, zur Drop-out-Rate des Probandenkollektivs bzw. zur Intention-to-treat-Analyse und zu weiteren Einflussfaktoren (Körpergewicht, Diabetesmedikation), die im Studienzeitraum (4 bis 18 Wochen) die untersuchten Zielparameter der Glykämiesituation (insbesondere Nüchternblutzucker und HbA1c-Wert) beeinflusst haben könnten.
Süßstoffe
Empfehlung.
Der Verzehr von Süßstoffen ist bei T2Dm mellitus bei Einhaltung der jeweiligen Höchstmengen gesundheitlich unbedenklich und kann bei einem gelegentlichen Einsatz im Rahmen einer Diabetestherapie sinnvoll sein.
Bei an T2Dm Erkrankten im Kindes- und Jugendalter ist die niedrigere tolerierbare Tagesdosis („acceptable daily intake“; ADI-Wert) aufgrund des geringeren Körpergewichts zu beachten.
Kommentar
Süßstoffe werden in der Fachliteratur immer wieder kontrovers diskutiert. Einer Hypothese nach könnten Süßstoffe aufgrund ihrer intensiven Süßkraft eine appetitsteigernde Wirkung hervorrufen (z. B. [272]). Bei einer Süßstoffgabe (in Form eines Getränks) konnte, verglichen mit Wasser, allerdings weder bei gesunden, normalgewichtigen Probanden [273–275] noch bei stoffwechselgesunden, übergewichtigen Probanden ein appetitsteigernder Effekt festgestellt werden. Süßstoffen wird eine mit Wasser vergleichbare orexigene Wirkung zugesprochen [273].
Inwiefern sich der Verzehr von Süßstoffen auf den Glukosemetabolismus bei Patienten mit diagnostiziertem T2Dm mellitus auswirkt, wurde in mehreren klinischen Studien geprüft. Es konnte kein Effekt des Süßstoffkonsums auf die Konzentration der Parameter Glukose, Insulin bzw. C‑Peptid, Glucagon-like Peptide‑1 (GLP1), Glucose-dependent Insulinotropic Peptide (GIP), Peptid YY (PYY), Glukagon sowie HbA1c festgestellt werden [276–281]. Demnach scheint sich der Verzehr von Süßstoffen nicht negativ auf die Glukose- und Insulinregulierung bei T2Dm auszuwirken.
Unumstritten ist die geringe kariogene Wirkung von Süßstoffen im Gegensatz zu herkömmlichem Zucker. Im Fall von Saccharin, Sucralose, Aspartam sowie Stevia kommt es zusätzlich zu einem bakteriostatischen Effekt auf orale Flora [282, 283]. Inwiefern Süßstoffe auf die Darmmikrobiota Einfluss nehmen, wurde noch nicht ausreichend geklärt. In einer Interventionsstudie wurde infolge einer Saccharingabe bei rund der Hälfte der Probanden (4/7) eine Veränderung der Darmmikrobiota festgestellt [284]. Diese Ergebnisse konnten allerdings bislang nicht bestätigt werden.
Der frühere Vorbehalt, Süßstoffe seien krebserregend, ist heutzutage entkräftet. Nach derzeitigem Wissensstand gibt es bei Einhaltung des ADI-Werts keine Hinweise auf eine kanzerogene Wirkung von Süßstoffen [285].
Wissenschaftlicher Hintergrund
Süßstoffe sind synthetisch hergestellte oder natürlich vorkommende Verbindungen mit hoher Süßintensität, die insulinunabhängig metabolisiert werden und nicht kariogen sind. Süßstoffe haben im Vergleich zu Zucker (Saccharose) eine um ein Vielfaches höhere Süßkraft (30- bis 20.000fach) und werden daher nur in kleinsten Mengen (Milligrammbereich) verwendet, die bezüglich der Kalorienzufuhr vernachlässigbar sind. Als Zusatzstoffe unterliegen Süßstoffe vor der Zulassung einer gesundheitlichen Bewertung durch die Europäische Behörde für Lebensmittelsicherheit (EFSA), die akzeptable tägliche Aufnahmemengen (ADI) ableitet. Der ADI-Wert gibt die Menge eines Zusatzstoffs an, die täglich während des gesamten Lebens pro Kilogramm Körpergewicht aufgenommen werden kann, ohne dass es zu gesundheitlichen Risiken kommt. Nach der Zulassung werden Süßstoffe bei Bedarf nochmals geprüft und in regelmäßigen Abständen neu bewertet [286].
Probiotika
Empfehlung.
Eine Probiotika- bzw. Synbiotikaeinnahme kann sich vorteilhaft auf die Glukoseregulation und das Lipidprofil von T2Dm auswirken.
Ein Mehrstammpräparat erzielt in der Regel einen stärkeren Effekt als ein Einzelstammpräparat.
Für eine Empfehlung einer Probiotika- bzw. Synbiotikasupplementation reicht die Evidenz bisher nicht aus.
Kommentar
Die Auswirkung einer Probiotikasupplementation auf den T2Dm mellitus ist bereits umfassend untersucht worden. Diverse Metaanalysen zeigen eine signifikante Reduktion der Nüchternblutglukose bei T2Dm durch eine Probiotikasupplementation, verglichen mit einer Placebogabe [287–292]. Auch eine signifikante Senkung der Insulinresistenz (HOMA-Index) wurde bei Probanden mit T2Dm infolge einer Probiotikagabe, verglichen mit der Kontrollgruppe, in mehreren Metaanalysen beobachtet [289, 293]. Eine langfristige Veränderung, gemessen mittels des HbA1c-Werts, konnte durch eine Probiotika- bzw. Synbiotikatherapie (mindestens 12 Wochen) allerdings nicht festgestellt werden [287, 288].
Die Ergebnisse von Metaanalysen hinsichtlich des Effekts einer Probiotikasupplementation auf den Lipidstatus von Patienten mit T2Dm sind heterogen. Zwei aktuelle Metaanalysen zeigen, verglichen mit einer Placebogabe, eine signifikante Senkung des Gesamtcholesterins sowie der Triglyzeridkonzentration (TG) bei T2Dm infolge einer 1‑ bis 6‑monatigen Pro- bzw. Synbiotikasupplementation [287, 294]. Bei Mahboobi et al. (2018) [291] wurde eine signifikante Verbesserung der TG-, LDL- und HDL-Cholesterinkonzentration infolge einer Synbiotika-, nicht aber bei einer Probiotikagabe verzeichnet. Eine weitere Metaanalyse konnte diesbezüglich keinen Zusammenhang feststellen [295].
Eine kürzlich publizierte randomisierte, kontrollierte Interventionsstudie im Cross-over-Design von Palacios et al. (2020) [296] untersuchte, inwiefern sich eine Probiotikagabe ergänzend zu einer Metformin-Therapie auswirkt. Nach einer 12-wöchigen Gabe eines Mehrstammprobiotikums wurden eine Verbesserung der Glukoseregulation (gemessen an der Nüchternblutglukosekonzentration, dem HbA1c-Wert und dem HOMA-Index) und der Barrierefunktion des Darms (gemessen an Zonulin) sowie eine erhöhte Plasma-Butyrat-Konzentration, verglichen mit einer Placebogabe, festgestellt.
Bei einer Probiotikasupplementation gibt es Folgendes zu bedenken: Probiotika können Antibiotikaresistenzen in mobilen Genen aufweisen, die durch interbakteriellen Austausch auf andere, möglicherweise pathogene Bakterien übertragen werden können [297]. Die Untersuchung diverser handelsüblicher Probiotika ergab, dass die getesteten probiotischen Bakterien gegen verschiedene Breitbandantibiotika resistent waren [298].
Wissenschaftlicher Hintergrund
In Deutschland gelten als Probiotika „definierte lebende Mikroorganismen, die in ausreichender Menge in aktiver Form in den Darm gelangen und hierbei positive gesundheitliche Wirkungen erzielen“ [299]. Vorrangig werden die Gattungen Lactobacillus und Bifidobacterium für die Formulierung in Probiotika verwendet. Des Weiteren kommen spezifische Milchsäure produzierende Arten anderer Gattungen, z. B. Enterococcus faecalis, Streptococcus thermophilus oder auch probiotische Hefen (Saccharomyces boulardii) zum Einsatz. Die Dosis variiert dabei zwischen 108 und 1011 koloniebildenden Einheiten, und der Einsatz oben genannter Gattungen bzw. Arten gilt als sicher [300].
Die Darmmikrobiota können einen starken Einfluss auf den Glukosemetabolismus v. a. durch die Modulation der Insulinsensitivität [301] und der Insulinsynthese [302] nehmen. Nach einem auf dem Mausmodell beruhenden postulierten Mechanismus binden mikrobiell synthetisierte, kurzkettige Fettsäuren (Acetat, Propionat und Butyrat) an G‑Protein-gekoppelte Rezeptoren (GRP43), wodurch die Sekretion des Peptidhormons GLP1 induziert wird [303]. GLP1 stimuliert sowohl bei glukosetoleranten Individuen als auch bei T2Dm-Patienten die Insulinsynthese [304].
Groß angelegte Studien zeigen, dass ein verändertes Darmmikrobiom (auch Dysbiose genannt) bei T2Dm-Erkrankten vorliegt [305–307]. Da allerdings eine T2Dm-Medikation, beispielsweise Metformin, nachweislich zu einer Modulation der Darmmikrobiota führt [308–310], ist oft unklar, ob die Veränderung auf die Erkrankung oder die Therapie zurückzuführen ist. Daher ist es bisher nicht gelungen, ein charakteristisches T2Dm-Mikrobiom zu identifizieren. Einige Studien deuten allerdings darauf hin, dass sich das Mikrobiom bei T2Dm durch einen geringeren Anteil an Butyrat produzierenden Bakterien auszeichnet [305, 306, 310]. Ein Verlust an Butyratproduzenten wird als Prädiktor für den Übergang eines Prädiabetes hin zum T2DM diskutiert [311], warum eine Supplementation mittels Pro- oder Synbiotika ein relevanter Aspekt sein kann.
Saccharose/Fruktose
Empfehlung.
Fruktose kann im Rahmen einer balancierten Ernährung in natürlichen Lebensmitteln (z. B. Obst) verzehrt werden.
Mit Fruktose gesüßte Getränke sollen insbesondere beim Überschreiten der täglichen empfohlenen Energiezufuhr gemieden werden.
Kommentar
Entsprechend den Empfehlungen der amerikanischen und kanadischen Diabetesgesellschaften sollte die Aufnahme von Mono- und Disacchariden nicht mehr als 10 bzw. 12 % der täglichen Energiezufuhr ausmachen [312, 313]. Der isokalorische Austausch von Kohlenhydraten wie Stärke und Saccharose gegen Fruktose hat keine ungünstigen Auswirkungen auf Körpergewicht [314], Blutdruck [315], Nüchterntriglyzeride [316], postprandiale Triglyzeride [317], Fettlebermarker [318] oder Harnsäure [319]. Bei Menschen mit Diabetes könnte der isokalorische Austausch gegen Fruktose Nüchternglukose und HbA1c-Wert senken [320], insbesondere, wenn sie in kleinen Mengen und in Form von Obst konsumiert wird [321]. Hingegen führt Fruktose, v. a. in Dosen von mehr als 60 g pro Tag oder 10 E% des täglichen Energiebedarfs, möglicherweise zu leichten Triglyzeridanstiegen bei Menschen mit T2Dm [316, 322]. Eine hyperkalorische Zufuhr von Fruktose führt weiterhin zu Gewichtszunahme [314], Harnsäureanstieg [319], hepatischer Insulinresistenz, Leberverfettung und Erhöhung der Transaminasen [318, 323] mit der übermäßigen Kalorienzufuhr als vermutlicher Ursache. Aus diesem Grund sollten Menschen mit Diabetes den Konsum von mit Zucker gesüßten Getränken minimieren zur Vermeidung einer Gewichtszunahme und zur Verbesserung des kardiometabolischen Risikoprofils [2].
Wissenschaftlicher Hintergrund
Seit den 1970er-Jahren wird in den USA und zunehmend auch in anderen Ländern High Fructose Corn Syrup (HFCS) zum Süßen von Getränken eingesetzt. Länder mit einem höheren HFCS-Verbrauch weisen im Vergleich zu Ländern mit einem geringeren HFCS-Verbrauch eine um 20 % höhere Diabetesprävalenz auf, unabhängig von dem Gesamtzuckerverzehr und der Adipositasprävalenz [324].
Entgegen diesem epidemiologischen Zusammenhang kamen prospektive Kohortenstudien zur Wirkung von Fruktose auf den Stoffwechsel zu inkonsistenten Ergebnissen. So wies eine Metaanalyse von 15 prospektiven Kohortenstudien nicht auf einen von der Nahrungsmittelform unabhängigen Zusammenhang zwischen Fruktoseaufnahme und erhöhtem T2Dm-Risiko hin [325]. In einer Metaanalyse von 51 isokalorischen Studien und 8 hyperkalorischen Studien hatte Fruktose nur dann ungünstige Effekte auf den Lipidstoffwechsel im Sinne eines Apolipoprotein-B- und Triglyzeridanstiegs, wenn sie als zusätzliche Kalorien zu einer bestehenden Ernährung angeboten wurde, während der isokalorische Austausch mit Fruktose den Lipidstoffwechsel nicht negativ beeinflusste [316]. In Übereinstimmung mit diesem Ergebnis erhöhte gemäß einer Metaanalyse von 14 isokalorischen und 2 hyperkalorischen Studien Fruktose, die mit einer gesteigerten Energieaufnahme einherging, nicht aber ein isokalorischer Fruktoseaustausch die postprandialen Triglyzeride [326]. Ebenso führte in einer Metaanalyse von 24 kontrollierten Interventionsstudien die Aufnahme von mehr als 100 g Fruktose pro Tag zu einer Erhöhung von Low-Density-Lipoprotein(LDL)-Cholesterin und Triglyzeriden, ohne dass ein Effekt auf die Serumlipide bei einer Fruktoseaufnahme von weniger als 100 g pro Tag zu beobachten war [327]. Eine Metaanalyse von 16 Studien, die den isokalorischen Kohlenhydrataustausch mit Fruktose bei Patienten mit T2Dm untersuchten, ergab heterogene Effekte auf den Lipidstoffwechsel mit einem Triglyzeridanstieg und einem Gesamtcholesterinabfall ohne Beeinflussung des LDL-Cholesterins [322].
Zudem führte ein hyperkalorischer Fruktoseverzehr, wie in einer Metaanalyse von 21 Studien gezeigt, nur bei stoffwechselgesunden Teilnehmern zu einem Harnsäureanstieg, während der Harnsäurespiegel nach isokalorischer Fruktoseaufnahme bei Menschen sowohl mit als auch ohne Diabetes unverändert blieb [319]. Hingegen wies eine aktuelle Netzwerkmetaanalyse darauf hin, dass der Ersatz von Fruktose durch Stärke zu vermindertem LDL-Cholesterin führte und der Ersatz von Fruktose durch Glukose Insulinsensitivität und Harnsäurespiegel günstig beeinflusste [328]. Hingegen führte in einer Metaanalyse von 18 Studien an Patienten mit T1Dm und T2Dm ein isokalorischer Austausch mit Fruktose zu einer klinisch relevanten Abnahme des HbA1c-Werts von 0,53 % [320]. Zu einem ähnlichen HbA1c-Wert-Abfall kam es in einer Metaanalyse von 6 kontrollierten Ernährungsinterventionsstudien nach Aufnahme von bis zu 36 g Fruktose pro Tag in Form von Obst, ohne Beeinträchtigung von Körpergewicht sowie Triglyzerid‑, Insulin- und Harnsäurespiegeln [321]. In Übereinstimmung mit diesem Ergebnis wirkte sich bei Patienten mit einem kürzlich diagnostizierten T2Dm der Konsum von Fruktose aus zuckerhaltigen Getränken, nicht aber aus Früchten, ungünstig auf periphere und hepatische Insulinsensitivität aus [329].
In einer Metaanalyse von 29 Arbeiten führte ein kurzzeitiger Fruktoseverzehr, sowohl als isokalorischer Austausch gegen andere Kohlenhydrate als auch als hyperkalorische Ergänzung, bei normalgewichtigen, übergewichtigen und adipösen Teilnehmern zur Entwicklung einer hepatischen Insulinresistenz, ohne dass die periphere oder muskuläre Insulinsensitivität beeinflusst wurde [330]. In einer Metaanalyse von 13 Studien begünstigte der isokalorische Austausch mit Fruktose nicht die Entwicklung einer nichtalkoholischen Fettlebererkrankung (NAFLD). Hingegen kam es infolge eines vermehrten Fruktoseverzehrs zum Anstieg der intrahepatozellulären Lipide sowie der Glutamat-Pyruvat-Transaminase [318]. Im Einklang mit diesem Ergebnis ergab eine weitere Metaanalyse von 6 Beobachtungsstudien und 21 Interventionsstudien ebenfalls einen Anstieg von Leberfett und Glutamat-Oxalacetat-Transaminase infolge einer hyperkalorischen Fruktoseaufnahme [323].
In einer Metaanalyse von 31 isokalorischen und 10 hyperkalorischen prospektiven Kohortenstudien hatte die Gabe von Fruktose in den isokalorischen Studien keinen Einfluss auf das Körpergewicht, während hingegen die Zufuhr großer Fruktosemengen zu einer Gewichtszunahme führte [314].
Zusammenfassend ist es bei der Beurteilung von Studien zur Wirkung von Fruktose auf den Stoffwechsel von großer Bedeutung zu unterscheiden, ob Fruktose isokalorisch im Austausch für andere Kohlenhydrate oder hyperkalorisch als zusätzliche Energie aufgenommen wurde. Hyperkalorische Studien weisen auf die ungünstigen Effekte von Fruktose auf den Stoffwechsel hin, die vermutlich auf die Aufnahme zusätzlicher Energie zurückgeführt werden können. Ungünstige Effekte einer isokalorischen Fruktoseaufnahme können mit den vorliegenden Studien nicht belegt werden. Möglicherweise hat Fruktose, in kleinen Mengen und in Form von Obst konsumiert, günstige Effekte auf den Glukosestoffwechsel.
Alkohol
Empfehlung.
Menschen mit T2Dm sollten die Menge des Alkoholgenusses auf die für die Allgemeinbevölkerung empfohlenen Mengen begrenzen. Ein mäßiger, risikoarmer Alkoholgenuss ist mit einer guten Stoffwechseleinstellung und Diabetesprognose vereinbar.
Menschen mit Diabetes mit einem riskanten Alkoholkonsum bzw. einer Alkoholabhängigkeit müssen über die Gefahren des Alkohols, speziell auch in Bezug auf eine verschlechterte Stoffwechseleinstellung, sowie die Gefahr von Folgeerkrankungen aufgeklärt werden.
Es muss allgemein darauf hingewiesen werden, dass bei Genuss größerer Alkoholmengen das Risiko für schwere, insbesondere nächtliche Hypoglykämien unter einer Insulintherapie ansteigt und dieses Risiko durch Nahrungsaufnahme während der Zeit des Alkoholgenusses und Anheben des Zielblutzuckers zur Nacht reduziert wird.
Kommentar
Differenzierte Inhalte zum Umgang mit Alkohol für Personen mit Diabetes mellitus finden sich in der S2-Leitlinie Psychosoziales und Diabetes [331].
Menschen mit T2Dm sollten über die Auswirkungen von Alkoholkonsum auf den Blutzuckerspiegel beraten werden und, wenn Alkohol konsumiert wird, zu einem risikoarmen Konsum angehalten werden. Die Deutsche Hauptstelle für Suchtfragen (DHS) e. V. gibt als Grenzwerte für einen risikoarmen Konsum 12 g Alkohol pro Tag bei Frauen und 24 g Alkohol pro Tag bei Männern an. Die Weltgesundheitsorganisation (WHO) definiert einen Konsum von 10 g Alkohol pro Tag bei Frauen und 20 g Alkohol pro Tag bei Männern als risikoarm. Diese Mengen gelten auch für Menschen mit T2Dm.
Alkohol und Glukosestoffwechsel
Bei Menschen mit Diabetes zeigt sich ein linearer und inverser Zusammenhang zwischen regelmäßigem Alkoholkonsum und dem HbA1c-Wert [332] (EK IIb). Der Konsum von einem Glas Wein am Tag (150 ml oder 13 g Alkohol) über einen Zeitraum von 3 Monaten führte im Vergleich zu einer Kontrollgruppe, die ein Glas alkoholfreies Bier pro Tag konsumierte, zu einer signifikanten Reduktion der Nüchternglukose, ohne die postprandialen Glukosewerte zu erhöhen. Ein positiver Effekt auf den HbA1c-Wert war am größten in der Gruppe mit dem höheren Ausgangs-HbA1c-Wert. In einer anderen kontrollierten Studie zeigte sich beim Konsum von 1 bis 2 Gläsern Wein pro Tag (120–240 ml oder 18 g Alkohol) über einen Zeitraum von 4 Wochen kein negativer Einfluss auf metabolische Parameter (Nüchternglukose, Lipide), jedoch ein signifikant positiver Effekt auf den Nüchternseruminsulinspiegel [333].
Der Genuss von Alkohol kann die Blutglukosegegenregulation beeinträchtigen und somit das Risiko für Unterzuckerungen unter Insulintherapie oder insulinotropen oralen Antidiabetika erhöhen [334–336].
Bei etwa jeder 5. schweren Hypoglykämie, die zu einer Krankenhauseinweisung führt, ist die Ursache Alkoholkonsum [337]. Der Haupteffekt von Alkohol dürfte jedoch in der Bewusstseinseinschränkung liegen, die zu einer eingeschränkten Wahrnehmung von Unterzuckerungen führt und Betroffene daran hindert, angemessen zu reagieren [338].
Der übermäßige Konsum von Alkohol beeinträchtigt die Diabetestherapie. Patienten mit übermäßigem oder riskantem Alkoholkonsum setzen weniger häufig Therapieempfehlungen zu Bewegungsverhalten, Ernährung, Medikamenteneinnahme, Blutzuckerselbstkontrolle oder regelmäßiger HbA1c-Wert-Kontrolle um. Dabei besteht ein linearer Zusammenhang: Je höher die Trinkmenge, desto seltener werden Therapieempfehlungen umgesetzt [339].
Laut der S2k-Leitlinie Psychosoziales und Diabetes soll bei Menschen mit Diabetes regelmäßig – mindestens 1‑mal im Jahr – der Alkoholkonsum erhoben werden, und bei einem riskanten Alkoholkonsum sollen Hilfsangebote gegeben werden.
Nahrungsergänzungsmittel
Empfehlung.
Personen mit T2Dm sollten ihren Nährstoffbedarf durch eine ausgewogene Ernährung decken. Eine Routinesupplementation mit Mikronährstoffen wird nicht empfohlen.
Bei Patienten mit T2Dm und nachgewiesenem Vitamin-D-Mangel kann eine Vitamin-D-Supplementierung eine Insulinresistenz bessern.
Kommentar
Die amerikanische, die kanadische und die britische Diabetesgesellschaft fassen die Evidenz zur Einnahme von Nahrungsergänzungsmitteln allgemein für Personen mit Diabetes wie folgt zusammen: Es besteht keine klare Evidenz, dass eine Supplementation mit Vitaminen, Mineralstoffen (beispielsweise Chrom oder Vitamin D), Kräutern oder Gewürzen (beispielsweise Zimt oder Aloe vera) die Stoffwechseleinstellung bei Personen ohne zugrunde liegende Ernährungsdefizite verbessert, und sie werden nicht allgemein zur Verbesserung der glykämischen Kontrolle empfohlen [2, 119–121]. Eine Routinesupplementation mit Antioxidanzien (beispielsweise Vitamin E, C oder Carotin) wird aufgrund eines mangelnden Wirksamkeitsnachweises sowie von Bedenken bezüglich der langfristigen Sicherheit nicht empfohlen. Eine Multivitaminsupplementation könnte allerdings bei speziellen Gruppen, z. B. schwangeren oder stillenden Frauen, älteren Personen, Vegetariern oder Personen mit einer sehr niedrigkalorischen oder kohlenhydratarmen Ernährung, notwendig sein [2, 119]. Bei der Einnahme von Metformin kann es zu einem Mangel an Vitamin B12 kommen, sodass eine regelmäßige Testung der Vitamin‑B12-Konzentration bei Personen mit T2Dm und Metformin-Einnahme, insbesondere bei zusätzlichem Vorliegen einer Anämie oder einer peripheren Neuropathie, bedacht werden sollte und eine mögliche Vitamin‑B12-Defizienz mittels Supplementation ausgeglichen werden könnte [2, 119]. Im Falle der Verwendung von Supplementen müssen mögliche unerwünschte Nebenwirkungen und Arzneimittelinteraktionen bedacht werden [2, 119, 340]. Statt der generellen Empfehlung einer Routinegabe von Nahrungsergänzungsmitteln sollen Personen mit Diabetes ermutigt werden, ihren Nährstoffbedarf durch eine ausgewogene Ernährung zu decken [121]. Dabei sollte berücksichtigt werden, dass Personen mit Diabetes, die ihre Stoffwechseleinstellungsziele nicht erreichen, ein erhöhtes Risiko für einen Mikronährstoffmangel haben könnten, sodass die Einhaltung einer ausgewogenen Ernährung, die mindestens die täglich empfohlene Tagesdosis an Nährstoffen und insbesondere Mikronährstoffen liefert, essenziell ist [2].
Aufgrund der Vielzahl an verfügbaren Nahrungsergänzungsmitteln wird im folgenden Wissenschaftlichen Hintergrund eine Auswahl an Inhaltsstoffen – nämlich n‑3 PUFAs, Vitamin D, Magnesium, Chrom, Zink, Antioxidanzien (Vitamin C, E) und Polyphenole – in Bezug auf ihre mögliche Wirksamkeit bei Personen mit T2Dm genauer beleuchtet. Kriterien für die Auswahl dieser Nahrungsergänzungsmittel waren die Relevanz der möglichen Effekte einer Supplementation auf das Diabetesmanagement und eine im Verhältnis betrachtet „gute“ Datenlage, primär basierend auf systematischen Reviews und Metaanalysen.
Wissenschaftlicher Hintergrund
Der Verzehr von n‑3 PUFAs wird im Zusammenhang mit positiven Effekten auf die glykämische Kontrolle und die Prävention kardiovaskulärer Erkrankungen bei Personen mit T2Dm diskutiert [119]. Ein systematisches Review aus der Cochrane Library (23 RCTs, n = 1075 T2Dm) zeigte eine signifikante Reduktion der Triglyzerid- (moderater Effekt) und VLDL-Konzentrationen (in Subgruppenanalysen nur signifikant für Personen mit Hypertriglyzeridämie) und einen signifikanten Anstieg der LDL-Cholesterinkonzentrationen nach Supplementation mit n‑3 PUFAs vs. pflanzliche Öle oder Placebo. Die Supplementation hatte im Vergleich zur Kontrolle keine Effekte auf die Gesamt- oder HDL-Cholesterinkonzentration, den HbA1c-Wert, die Nüchternblutglukose- oder Nüchterninsulinkonzentration und das Körpergewicht [341]. Eine Steigerung der LDL-Cholesterinkonzentration nach Supplementation mit n‑3 PUFAs vs. Kontrolle wurde auch in einem weiteren systematischen Review und einer Metaanalyse (24 RCTs, n = 1533 T2Dm) von Hartweg et al. bestätigt [342]. Allerdings zeigte sich durch die Supplementation keine Veränderung der LDL-Partikelgröße, die neben Änderungen in den Triglyzerid- und HDL-Cholesterinkonzentrationen die diabetische Dyslipoproteinämie charakterisiert [342]. Weiterhin war in beiden Arbeiten die Erhöhung der LDL-Cholesterinkonzentrationen durch die n‑3-PUFA-Supplementation in der Subgruppe von Personen mit Hypertriglyzeridämie nicht signifikant [341, 342]. Ein neueres systematisches Review mit Metaanalyse (45 RCTs, n = 2674 T2Dm) bestätigt die protektiven Effekte einer n‑3-PUFA-Supplementation vs. Placebo auf den Lipidstoffwechsel und berichtet eine signifikante Reduktion der LDL-Cholesterin‑, VLDL-Cholesterin- und Triglyzeridkonzentrationen durch Supplementation mit n‑3 PUFAs vs. Placebo [343]. Weiterhin zeigten sich bei O’Mahoney et al. eine Reduktion des HbA1c-Werts und keine Effekte auf die Nüchternblutglukose‑, Nüchterninsulinkonzentration sowie den HOMA-IR durch Supplementation mit n‑3 PUFAs vs. Placebo [343]. Brown et al. (83 RCTs, n = 121.070 mit und ohne T2Dm) untersuchten neben Effekten einer höheren vs. einer niedrigeren Zufuhr an n‑3-, n‑6- und Gesamt-PUFA auf das Diabetesrisiko auch deren Wirkung auf die glykämische Kontrolle und die Insulinresistenz und fanden keine Effekte einer höheren vs. einer niedrigeren n‑3-PUFA-Einnahme auf den HbA1c-Wert, die Nüchternblutglukose‑, Nüchterninsulinkonzentration und den HOMA-IR [71]. Weiterhin gibt es Hinweise, dass eine hoch dosierte Supplementation mit langkettigen n‑3 PUFAs (> 4,4 g/Tag) den Glukosestoffwechsel verschlechtern könnte [71]. Insgesamt fasst die amerikanische Diabetesgesellschaft die Evidenz zu n‑3 PUFAs für Personen mit T2Dm mit einer Empfehlung des Verzehrs von Lebensmitteln mit einem hohen Gehalt an langkettigen n‑3-Fettsäuren aus beispielsweise Fisch, Nüssen und Samen zur Prävention und Behandlung von kardiovaskulären Erkrankungen zusammen (Evidenzgrad B) [119]. Vorteile einer Routinesupplementation mit n‑3 PUFA werden basierend auf der aktuellen Evidenz jedoch nicht unterstützt (Evidenzgrad A), da Supplemente nicht die gleichen positiven Effekte wie die entsprechenden vollwertigen Lebensmittel auf die glykämische Kontrolle und die Primär- und Sekundärprävention von kardiovaskulären Erkrankungen zu haben scheinen [117]. Weiterhin fehlen Studien zur n‑3-PUFA-Supplementation mit vaskulären Events, kardiovaskulären Erkrankungen oder Mortalität als Endpunkt bei Personen mit T2Dm [341, 342].
Ein Vitamin‑D-Mangel ist mit Veränderungen im Glukosemetabolismus und der Insulinsekretion assoziiert [344]. Die Evidenz zu Effekten einer Supplementation mit Vitamin D auf die glykämische Kontrolle ist jedoch basierend auf den systematischen Reviews und Metaanalysen von Li et al. (20 RCTs, n = 2703 T2Dm) und Mirhosseini et al. (24 RCTs, n = 1528 T2Dm) widersprüchlich [344, 345]. Während beide Übersichtsarbeiten einen signifikanten Anstieg der Serum-25-OH-Vitamin-D-Spiegel und eine Reduktion des HOMA-IR nach Supplementation mit Vitamin D im Vergleich zu Placebo bestätigen [344, 345], war eine Reduktion der Nüchternblutglukosekonzentration und des HbA1c-Werts nach Supplementation mit Vitamin D im Vergleich zu Placebo nur bei Mirhosseini et al. signifikant [344, 345]. Diese positiven Effekte auf Parameter der glykämischen Kontrolle und der Insulinresistenz waren insbesondere bei einer hohen täglichen Vitamin-D-Dosierung (≥ 4000 IU/Tag) und einer langen Interventionsdauer (durchschnittlich 7 Monate) messbar [344]. Laut Li et al. reduziert eine Supplementation mit Vitamin D im Vergleich zu Placebo die Nüchterninsulinkonzentration nur in nicht adipösen Personen mit T2Dm und die Nüchternblutglukosekonzentration nur bei kurzzeitiger Supplementation, Dosierungen > 2000 IU/Tag und bei Personen mit Vitamin-D-Mangel und guter HbA1c-Kontrolle zur Baseline [345]. Weitere systematische Reviews und Metaanalysen untersuchten die Effekte einer Supplementation mit Vitamin D im Vergleich zu Placebo auf den Blutdruck, die Serumlipidkonzentrationen und die chronische subklinische Inflammation [346–349]. Für den Blutdruck (15 RCTs, n = 1134 T2Dm) zeigte eine Vitamin-D-Supplementation vs. Placebo eine signifikante, jedoch geringe Reduktion des diastolischen Blutdrucks und keine Veränderung des systolischen Blutdrucks [348]. Auch in Bezug auf die Serumlipidkonzentration (17 RCTS, n = 1365 T2Dm) wies eine Supplementation mit Vitamin D vs. Placebo zwar eine signifikante Reduktion der Gesamt‑, LDL- und HDL-Cholesterinkonzentration im Serum auf, jedoch waren diese Effekte gering [346]. Weiterhin führte eine Supplementation mit Vitamin D vs. Placebo zur Reduktion einzelner Biomarker der chronischen subklinischen Inflammation wie CRP (20 RCTs, n = 1270 T2 D und 13 RCTs, n = 875 T2DM) [347, 349]. Während die Empfehlungen zur Frakturprävention für Personen mit T2DM identisch zu denjenigen für Personen der Allgemeinbevölkerung sind und eine Supplementation mit Vitamin D einschließen [119], ist die Qualität der Evidenz für die weiteren betrachteten Outcomes und die Qualität der dafür in die Übersichtsarbeiten inkludierten Studien von den Autoren als sehr heterogen eingestuft worden. Weitere qualitative hochwertige und langfristige RCTs sind somit notwendig, um eine Empfehlung zur Supplementation mit Vitamin D für Personen mit T2Dm – über die Frakturprävention hinausgehend – zu geben [344, 345, 347–349].
Magnesium, ein essenzieller Mineralstoff, ist u. a. am intrazellulären Kohlenhydratstoffwechsel, der Insulinsekretion und -signalkaskade, dem Lipidstoffwechsel und der Regulation des Blutdrucks beteiligt [350]. Die Evidenz zur Wirkung einer Magnesiumsupplementation auf die glykämische Kontrolle und den Blutdruck bei Personen mit T2Dm ist widersprüchlich [350–352]. Nach Supplementation mit Magnesium vs. Placebo (28 RCTs, n = 1694 T2DM) zeigten sich signifikante Verbesserungen der Nüchternblutglukosekonzentration und des systolischen Blutdrucks, mit stärker ausgeprägten Effekten bei Personen mit Hypomagnesiämie zu Baseline, jedoch keine Veränderungen der Nüchterninsulinkonzentration, des HbA1c-Werts und des diastolischen Blutdrucks [350]. In einem weiteren systematischen Review mit Metaanalyse (18 RCTs, n = 1079 Personen mit T2Dm) zeigte eine Supplementation mit Magnesium im Vergleich eine moderate Verbesserung des HbA1c und eine geringe bis moderate Verbesserung der Nüchternblutglukosekonzentration, jedoch keinen Effekt auf die Insulinkonzentration und den HOMA-IR. Die Nüchternblutglukosekonzentration wurde nur bei einer Magnesiumsupplementation ≥ 4 Monaten signifikant reduziert. In der stratifizierten Analyse nach Diabetesstatus ergab eine Magnesiumsupplementation im Vergleich zur Kontrolle keine signifikanten Effekte bei Personen mit Diabetes auf die Nüchternblutglukose-, die Insulinkonzentration und den HbA1c-Wert [351]. Asbaghi et al. untersuchten die Effekte einer Magnesiumsupplementation (11 RCTs, n = 673 T2Dm) auf den Blutdruck und anthropometrische Parameter [352]. Eine Magnesiumsupplementation im Vergleich zu Placebo ergab eine signifikante Reduktion des systolischen und diastolischen Blutdrucks, insbesondere bei einer Supplementation > 12 Wochen mit ≥ 300 mg/Tag anorganischem Magnesium. Auf anthropometrische Parameter zeigten sich jedoch keine Effekte einer Magnesiumsupplementation vs. Placebo [352]: Verma und Garg untersuchten neben den Effekten einer Magnesiumsupplementation auf die glykämische Kontrolle und den Blutdruck auch deren Wirkung auf die Serumlipide und konnten eine signifikante Erhöhung der HDL-Cholesterinkonzentration und eine Reduktion der LDL-Cholesterin- und Triglyzeridkonzentration im Vergleich zur Kontrolle zeigen [350]. Weitere langfristige RCTs mit guter Studienqualität bei Personen mit T2Dm sind notwendig, um evidenzbasierte Empfehlungen zur Magnesiumsupplementation aussprechen zu können.
Das essenzielle Spurenelement Chrom spielt eine wichtige Rolle im Kohlenhydrat- und Lipidstoffwechsel [353]. Eine Supplementation mit Chrom im Vergleich zu Placebo (23 RCTs, n = 1350 T2DM und T1Dm [T1DM nur in 1 RCT zusätzlich zu T2Dm inkludiert]) führte zu einer signifikanten Reduktion der Nüchternblutglukose- und Insulinkonzentration, des HbA1c-Werts und des HOMA-IR. Diese Effekte waren basierend auf einer Subgruppenanalyse bei einer längerfristigen Supplementation von mindestens 12 Wochen stärker ausgeprägt, zeigten jedoch keine Abhängigkeit von der verwendeten Chromdosierung. Alle inkludierten Studien wurden in ihrer Qualität als gut bewertet, jedoch wurde in der Metaanalyse keine Stratifizierung der Ergebnisse nach der verwendeten Chromformulierung vorgenommen (Chrompicolinat, Chromchlorid, Chrom aus Bierhefe) [354]. Basierend auf 2 vorangegangenen systematischen Reviews und Metaanalysen (22 RCTs, n = 1332 T2Dm und 14 RCTs, n = 875 T2DM) waren die Effekte einer Chromsupplementation im Vergleich zu Placebo auf die Nüchternblutglukosekonzentration bei Verwendung von Chrompicolinat am stärksten ausgeprägt bzw. nur signifikant bei Verwendung von Chrom aus Bierhefe [355, 356]. Auch eine Erhöhung des HDL-Cholesterin- und eine Reduktion der Triglyzeridwerte wurden insbesondere bei einer Supplementation mit Chrompicolinat oder Chrom aus Bierhefe im Vergleich zu Placebo erreicht [356], sodass weitere Untersuchungen zur optimalen Formulierung von Chromsupplementen für Personen mit T2Dm notwendig sind.
Das essenzielle Spurenelement Zink spielt eine wichtige Rolle in der Synthese, Speicherung und Sekretion von Insulin [357]. Der bei Personen mit T2Dm beobachtete Zinkmangel und die Hyperglykämie könnten sich wechselseitig bedingen [358]. Basierend auf 11 Beobachtungsstudien nahm bei Personen mit T2Dm im Vergleich zu metabolisch gesunden Kontrollpersonen die Zinkkonzentration im Vollblut mit jedem weiteren Jahr der Diabeteserkrankung ab. Dieser inverse Zusammenhang war im Allgemeinen nicht durch eine niedrigere nutritive Zinkaufnahme erklärbar, da nur Personen mit T2Dm und Komplikationen, die auf eine Ernährungstherapie angewiesen waren (beispielsweise Nephropathie), eine signifikant niedrigere Zinkaufnahme aufwiesen [359]. Eine Subgruppenanalyse eines systematischen Reviews mit Metaanalyse (32 RCTs und n = 1700 insgesamt, 19 RCTs mit Personen mit T2DM) zeigte für Personen mit T2Dm eine signifikante Reduktion der Nüchternblutglukosekonzentration bei Supplementation mit Zink vs. Kontrolle. In der Gesamtstudienpopulation, die zusätzlich Personen mit erhöhtem Risiko für T2Dm inkludierte, führte eine Supplementation mit Zink im Vergleich zur Kontrolle zusätzlich zu einer signifikanten Reduktion der 2‑h-postprandialen Blutglukosekonzentration, der Nüchterninsulinkonzentration, von HOMA-IR, HbA1c-Wert und hochsensitiven (hs)CRP [360]. Weiterhin reduzierte eine Supplementation mit Zink vs. Placebo (9 RCTs, n = 424 T2DM) die Serumkonzentrationen von Triglyzeriden und Gesamtcholesterin. Für LDL-Cholesterinkonzentrationen zeigten sich nur positive Effekte einer Zinksupplementation im Vergleich zu Placebo für stratifizierte Analysen nach LDL-Cholesterinkonzentration und HbA1c-Wert zu Baseline und für eine Interventionsdauer < 12 Wochen mit Dosierung < 100 mg/Tag. Eine Erhöhung der HDL-Cholesterinkonzentration konnte nur für Personen mit HDL-Cholesterinkonzentrationen im Normbereich und erhöhtem HbA1c-Wert zu Baseline sowie stratifiziert nach Interventionsdauer und Zinkdosierung gezeigt werden [361]. Aufgrund signifikanter Heterogenität zwischen den inkludierten Studien sowie variierender Qualität der Studien sind weitere Untersuchungen notwendig, bevor eine Supplementation mit Zink als ergänzende Therapie des T2Dm empfohlen werden kann [360, 361].
Oxidativer Stress spielt eine wichtige Rolle in der Pathogenese des Diabetes und seiner Komplikationen, sodass durch eine Supplementation mit Antioxidanzien positive Effekte auf das Diabetesmanagement erwartet werden könnten [362]. Für den Vergleich einer Supplementation Vitamin C vs. Kontrolle ergab ein systematischer Review mit Metaanalyse (n = 1574 T2Dm, primär basierend auf Interventionsstudien mit kurzer Studiendauer [< 6 Monate] und geringer Probandenzahl [n < 100]) eine statistisch signifikante und klinisch relevante Verbesserung des HbA1c (Evidenzgrad: sehr gering), eine statistisch signifikante, aber klinisch nicht relevante Reduktion der Nüchternblutglukosekonzentration, der Triglyzerid- und Gesamtcholesterinkonzentration (Evidenzgrad: sehr gering), keine statistisch signifikanten Effekte auf die HDL- oder LDL-Cholesterinkonzentration (Evidenzgrad: sehr gering) und eine statistisch signifikante und klinisch relevante Reduktion des systolischen und diastolischen Blutdrucks (Evidenzgrad: moderat bzw. sehr gering) [363]. Eine Subgruppenanalyse eines systematischen Reviews mit Metaanalyse basierend auf 14 RCTs bei Personen mit T2Dm (n = 714) zeigte unter Supplementation mit Vitamin E im Vergleich zur Kontrolle eine signifikante Reduktion des HbA1c-Werts und der Nüchternblutglukosekonzentration für Personen mit niedrigem Vitamin-E-Status zur Baseline und schlechter glykämischer Kontrolle [364]. Weder eine alleinige Supplementation mit Vitamin C oder Vitamin E noch eine Kombination beider Antioxidanzien zeigte signifikante Effekte auf den HOMA-IR (14 RCTs, n = 735 T2Dm) [365]. Eine Supplementation mit den Antioxidanzien Vitamin C und Vitamin E im Vergleich zu Placebo ergab in einer weiteren Studie (10 RCTs, n = 296 T2Dm) insgesamt keine Effekte auf die Endothelfunktion, jedoch in einer Subgruppenanalyse eine signifikante Verbesserung der Endothelfunktion nach Intervention für nicht adipöse Personen mit T2Dm (BMI ≤ 29,45 kg/m2) [366]. Personen mit T2Dm und diabetischer Retinopathie im Vergleich zu Personen mit T2Dm ohne Retinopathie hatten basierend auf 14 Beobachtungsstudien und 7 RCTs (n = 256.259) niedrigere Serumkonzentrationen an Antioxidanzien und höhere Konzentrationen an Biomarkern für oxidativen Stress. Aufgrund einer starken methodischen Heterogenität wurde nur eine qualitative Synthese der inkludierten RCTs vorgenommen, die auf positive Effekte einer Supplementation mit Antioxidanzien bei diabetischer Retinopathie hinweist [367]. Insgesamt basieren die berichteten Effekte der Supplementation mit Antioxidanzien bei Personen mit T2Dm primär auf Studien mit niedriger bis moderater Qualität, sodass die Evidenz für eine Supplementation zur Verbesserung der Stoffwechseleinstellung und der Endothelfunktion zurzeit unzureichend ist [363–366].
Resveratrol bzw. Polyphenole im Allgemeinen sind ebenfalls Antioxidanzien und könnten somit positive Effekte auf das Diabetesmanagement haben [368]. Eine Supplementation mit Polyphenolen (36 RCTs, n = 1954 insgesamt, n = 1426 T2Dm) führte im Vergleich zur Kontrolle zu einer signifikanten Reduktion des HbA1c-Werts (mittlerer HbA1c-Wert zur Baseline: 7,03 %). Eine Subgruppenanalyse zeigte, dass diese Reduktion für Personen mit T2DM signifikant war (mittlerer HbA1c-Wert zur Baseline: 7,44 %), während bei Personen ohne Diabetes und mit Prädiabetes keine Effekte einer Supplementation im Vergleich zur Kontrolle erkennbar waren [369]. Ein systematisches Review aus der Cochrane Library (3 RCTs, n = 50 T2Dm) zeigte hingegen keine Effekte einer Supplementation mit Resveratrol auf den HbA1c-Wert, die Nüchternblutglukosekonzentration oder die Insulinresistenz. Die vorliegende Evidenz aus den inkludierten RCTs wurde insgesamt als sehr gering eingestuft, sodass auch die derzeit vorliegende Evidenz zur Sicherheit und Wirksamkeit einer Supplementation mit Resveratrol als zu unzureichend bewertet wurde, um diese zur Behandlung des T2Dm zu empfehlen [368]. Auf den systolischen und diastolischen Blutdruck sowie den mittleren arteriellen Druck oder den Pulsdruck zeigte eine Supplementation mit Resveratrol im Vergleich zur Kontrolle in der Gesamtstudienpopulation (17 RCTs, n = 681 insgesamt, n = 262 T2Dm) keine Effekte. In Subgruppenanalysen reduzierte die Supplementation mit Resveratrol im Vergleich zur Kontrolle signifikant den systolischen Blutdruck, den mittleren arteriellen Druck und den Pulsdruck bei Personen mit T2Dm [370].
Insgesamt besteht zu allen betrachteten Nahrungsergänzungsmitteln aufgrund beispielsweise mangelnder Qualität inkludierter Studien, Heterogenität in der Methode und der Ergebnisse der Untersuchungen, einer zu geringen Zahl durchgeführter Studien oder fehlender Daten zu ausgewählten Endpunkten, Langzeiteffekten und langfristiger Sicherheit noch weiterer Forschungsbedarf, bevor sie als Ergänzung zur Therapie des T2Dm empfohlen werden können. Auch wenn für Einzelfälle oder spezielle Gruppen von Personen mit T2Dm der Ausgleich eines Nährstoffmangels durch die Einnahme eines Nahrungsergänzungsmittels auf individueller Basis und unter Berücksichtigung möglicher unerwünschter Nebenwirkungen und Arzneimittelinteraktionen in Erwägung gezogen werden kann, gilt für Personen mit T2Dm im Allgemeinen, dass sie ihren Nährstoffbedarf durch eine ausgewogene Ernährung decken sollen und eine Routinesupplementation mit Mikronährstoffen nicht empfohlen wird.
Besonderheiten in der stationären Therapie bzw. spezielle Ernährungsformen zur Reduktion des Insulinbedarfs
Empfehlung.
Im stationären Setting sind zur Durchbrechung starker Insulinresistenz 2‑tägige Hafer- bzw. Ballaststofftage sehr zu empfehlen. Diese müssen hypokalorisch sein und einen hohen Ballaststoffanteil enthalten. Die Hafertage sind dabei sehr effektiv. Alternativ können aber auch andere Ballaststoffkostformen gewählt werden.
Die Blutglukosespiegel steigen nach dem Verzehr ballaststoffreicher Haferprodukte im Vergleich zu anderen Mahlzeiten mit einer vergleichbaren Menge an Kohlenhydraten nicht so stark an, und es wird eine geringere Insulinsekretion induziert
Kommentar
Mehrere Studien haben gezeigt, dass die Insulinresistenz bei Menschen mit T2Dm durch eine bestimmte Kostform über einige Tage signifikant gesenkt werden konnte. Diese Kostformen waren in der Summe immer hypokalorisch und ballaststoffreich. Am besten, bezogen auf den HOMA-Index, haben die Hafertage abgeschnitten. Der Anteil an löslichen Ballaststoffen ist in Hafer besonders groß [371]. Die besondere Wirkung des Hafers wird in seiner Zusammensetzung vermutet. Hafer enthält insbesondere β‑Glucan. Die Menge ist mit ca. 7,8 % besonders hoch [372]. Zudem wurde in vitro eine hemmende Wirkung von Hafer-β-Glucan auf die Expression der SGLT1-Rezeptoren sowie des Glukosetransporters 2 (GLUT-2) in den Darmzellen gezeigt [373]. Des Weiteren konnte in vitro für bestimmte Haferproteine eine hemmende Wirkung auf die Dipeptidylpeptidase 4 (DDP4) gezeigt werden. Diese war etwas stärker als die Wirkung von Buchweizen und Gerste [374]. So konnte außerdem gesehen werden, dass Hafer-β-Glucan die Alpha-Glukosidase hemmt [375].
Unter stationären Bedingungen wurden in einer Pilotstudie insgesamt 14 Patienten an 2 aufeinanderfolgenden Tagen jeweils ca. 1100 Kalorien pro Tag Haferbrei gereicht. Vorher sowie 2 Tage und 4 Wochen nach der Intervention wurden der mittlere Blutzucker, Adiponektin sowie die mittlere Insulindosis erfasst. Die mittlere Insulindosis konnte um 47 % reduziert werden. Dieser Effekt konnte aber auch noch 4 Wochen nach der Intervention nachgewiesen werden. Die Autoren vermuten infolge der Hafertage Auswirkungen auf das Mikrobiom [376].
In der Cross-over-Studie „OatMeal And Insulin Resistance“ (OMA-IR) bei Menschen mit einem unzureichend kontrollierten T2Dm sank infolge von 2 Hafertagen der Insulinbedarf am 3. und 4. Tag hochsignifikant im Vergleich zu einer lediglich diabetesadaptierten Ernährung. Zugleich sank im Verlauf von 4 Wochen nach den Hafertagen der HbA1c-Wert [377]. Die Studie zeigt, dass Hafer-β-Glucan in der Lage ist, Gallensäuren zu binden und die Cholesterinspiegel im Blut zu senken. Darüber hinaus wurde nach den Hafertagen eine enge Korrelation zwischen dem Rückgang der Gesamtgallensäuremenge sowie dem Rückgang der Proinsulinspiegel beobachtet [378–384].
Das EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA) sieht es aufgrund der Studienlage als erwiesen an, dass der „Konsum von Beta-Glucan aus Hafer … zu einer Reduktion des Glukoseanstiegs nach einer Mahlzeit führt“ [385]. In der Folge hat die Europäische Kommission der EU den Health Claim veröffentlicht: „Der Verzehr von Beta-Glucanen aus Hafer … als Teil einer Mahlzeit trägt zur Reduktion des Blutzuckerspiegels nach dem Essen bei“ [386].
In einer Metaanalyse von 103 Vergleichsstudien mit 538 Studienteilnehmern war der Zusatz von Hafer-β-Glucan zu kohlenhydrathaltigen Mahlzeiten nachweislich mit einer reduzierten Glukose- und Insulinantwort assoziiert [387].
β‑Glucan erhöht die Viskosität im Dünndarm, verzögert die Magenentleerung sowie die Freisetzung und Resorption von Nahrungsbestandteilen, v. a. von Kohlenhydraten, lässt den Blutzucker dadurch langsamer ansteigen und resultiert in einer niedrigeren Insulinantwort [388, 389].
Infobox Inhaltliche Neuerungen und abweichende Empfehlungen gegenüber der Vorjahresfassung
Empfehlung 1: Fehlender Nutzen von Low-Carb für Gewichtsreduktion
Begründung: Aktuelle Auswertung
Stützende Quellenangabe: [390]
Empfehlung 2: Diabetesremission als primäres Ziel der Ernährungstherapie
Begründung: Aktuelle Auswertung
Stützende Quellenangabe: [391]
Empfehlung 3: Moderate Verbesserung des HbA1c und der Nüchternglukose unter Magnesiumsupplementierung ohne Effekt auf Insulin und HOMA
Begründung: –
Stützende Quellenangabe: [392]
Empfehlung 4: Fehlender Nutzen von Low-Carb für Gewichtsreduktion
Begründung: Aktuelle Auswertung
Stützende Quellenangabe: [390]
Einhaltung ethischer Richtlinien
Interessenkonflikt
T. Skurk: Honorar Novo Nordisk, D. Rubin: Honorar Novo Nordisk, A. Grünerbel: Honorare KV Bayern, S. Kabisch: Honorare und Reisekosten durch Sanofi, Berlin Chemie, Boehringer Ingelheim und Lilly; Reisekosten und Forschungsförderung durch J. Rettenmaier & Söhne, Holzmühle; weitere Forschungsförderung durch Beneo Südzucker und California Walnut Commission, A. Bosy-Westphal, W. Keuthage, P. Kronsbein, K. Müssig, H. Nussbaumer, A.F.H. Pfeiffer, M.-C. Simon, A. Tombek und K.S. Weber geben an, dass kein Interessenkonflikt besteht.
Dieser Beitrag wurde erstpubliziert in Diabetologie und Stoffwechsel (2022) 17(Suppl 2):S256–290, 10.1055/a-1886-3959. Nachdruck mit freundl. Genehmigung von Georg Thieme Verlag KG. Die Urheberrechte liegen bei den Autorinnen und Autoren.
Aktualisierungshinweis
Die DDG-Praxisempfehlungen werden regelmäßig zur zweiten Jahreshälfte aktualisiert. Bitte stellen Sie sicher, dass Sie jeweils die neueste Version lesen und zitieren.
QR-Code scannen & Beitrag online lesen
==== Refs
Literatur
1. Beck J Greenwood DA Blanton L 2017 national standards for diabetes self-management education and support Diabetes Care 2017 40 1409 1419 28754780
2. Evert AB Dennison M Gardner CD Nutrition therapy for adults with diabetes or prediabetes: a consensus report Diabetes Care 2019 42 731 754 31000505
3. Evert AB Stellungnahme des Ausschuss Ernährung der DDG zum Consensus Report: NutritionTherapy for Adults with Diabetes or Prediabetes Diabetes Care 2019 42 731 754 31000505
4. DeFronzo RA Bonadonna RC Ferrannini E Pathogenesis of NIDDM. A balanced overview Diabetes Care 1992 15 318 368 1532777
5. DeFronzo RA Eldor R Abdul-Ghani M Pathophysiologic approach to therapy in patients with newly diagnosed type 2 diabetes Diabetes Care 2013 36 Suppl. 02 S127 S138 23882037
6. Lencioni C Lupi R Del Prato S Beta-cell failure in type 2 diabetes mellitus Curr Diab Rep 2008 8 179 184 18625113
7. Anonymous U.K. prospective diabetes study 16. Overview of 6 years’ therapy of type II diabetes: a progressive disease. U.K. Prospective Diabetes Study Group Diabetes 1995 44 1249 1258 7589820
8. Zaharia OP Strassburger K Strom A Risk of diabetes-associated diseases in subgroups of patients with recent-onset diabetes: a 5-year follow-up study Lancet Diabetes Endocrinol 2019 7 684 694 31345776
9. Kodama S Horikawa C Fujihara K Quantitative relationship between body weight gain in adulthood and incident type 2 diabetes: a meta-analysis Obes Rev 2014 15 202 214 24165305
10. Wing RR Lang W Wadden TA Benefits of modest weight loss in improving cardiovascular risk factors in overweight and obese individuals with type 2 diabetes Diabetes Care 2011 34 1481 1486 21593294
11. Steven S Hollingsworth KG Al-Mrabeh A Very low-calorie diet and 6 months of weight stability in type 2 diabetes: Pathophysiological changes in responders and nonresponders Diabetes Care 2016 39 808 815 27002059
12. Jazet IM Pijl H Frölich M Factors predicting the blood glucose lowering effect of a 30-day very low calorie diet in obese Type 2 diabetic patients Diabet Med 2005 22 52 55 15606691
13. Lean MEJ Leslie WS Barnes AC Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial Lancet 2018 391 541 551 29221645
14. Bangalore S Fayyad R DeMicco DA Body weight variability and cardiovascular outcomes in patients with type 2 diabetes mellitus Circ Cardiovasc Qual Outcomes 2018 11 e004724 30571333
15. Yeboah P Hsu FC Bertoni AG Body mass index, change in weight, body weight variability and outcomes in type 2 diabetes mellitus (from the ACCORD trial) Am J Cardiol 2019 123 576 581 30553512
16. Pagidipati NJ Zheng Y Green JB Association of obesity with cardiovascular outcomes in patients with type 2 diabetes and cardiovascular disease: Insights from TECOS Am Heart J 2020 219 47 57 31707324
17. Bodegard J Sundström J Svennblad B Changes in body mass index following newly diagnosed type 2 diabetes and risk of cardiovascular mortality: a cohort study of 8486 primary-care patients Diabetes Metab 2013 39 306 313 23871502
18. Weinheimer EM Sands LP Campbell WW A systematic review of the separate and combined effects of energy restriction and exercise on fat-free mass in middle-aged and older adults: implications for sarcopenic obesity Nutr Rev 2010 68 375 388 20591106
19. Zaccardi F Dhalwani NN Papamargaritis D Nonlinear association of BMI with all-cause and cardiovascular mortality in type 2 diabetes mellitus: a systematic review and meta-analysis of 414587 participants in prospective studies Diabetologia 2017 60 240 248 27888288
20. Salehidoost R Mansouri A Amini M Body mass index and the all-cause mortality rate in patients with type 2 diabetes mellitus Acta Diabetol 2018 55 569 577 29546575
21. Hainer V Aldhoon-Hainerová I Obesity paradox does exist Diabetes Care 2013 36 Suppl. 02 S276 S281 23882059
22. Murphy RA Reinders I Garcia ME Adipose tissue, muscle, and function: potential mediators of associations between body weight and mortality in older adults with type 2 diabetes Diabetes Care 2014 37 3213 3219 25315206
23. Bales CW Porter Starr KN Obesity interventions for older adults: diet as a determinant of physical function Adv Nutr 2018 9 151 159 29659687
24. Uusitupa M Khan TA Viguiliouk E Prevention of type 2 diabetes by lifestyle changes: a systematic review and meta-analysis Nutrients 2019 11 2611 31683759
25. Raben A Vestentoft PS Brand-Miller J The PREVIEW intervention study: Results from a 3-year randomized 2 × 2 factorial multinational trial investigating the role of protein, glycaemic index and physical activity for prevention of type 2 diabetes Diabetes Obes Metab 2021 23 324 337 33026154
26. Gregg EW Chen H Wagenknecht LE Association of an intensive lifestyle intervention with remission of type 2 diabetes JAMA 2012 308 2489 2496 23288372
27. Anderson JW Konz EC Frederich RC Long-term weight-loss maintenance: a meta-analysis of US studies Am J Clin Nutr 2001 74 579 584 11684524
28. Bundesgesundheitsministerium (2015) Telemedizin. https://www.bundesgesundheitsministerium.de/service/begriffe-von-a-z/t/telemedizin.html. Zugegriffen: 9. Apr. 2021
29. Su D McBride C Zhou J Does nutritional counseling in telemedicine improve treatment outcomes for diabetes? A systematic review and meta-analysis of results from 92 studies J Telemed Telecare 2016 22 333 347 26442959
30. Kempf K Altpeter B Berger J Efficacy of the telemedical lifestyle intervention program TeLiPro in advanced stages of type 2 diabetes: a randomized controlled trial Diabetes Care 2017 40 863 871 28500214
31. Belalcazar LM Haffner SM Lang W Lifestyle intervention and/or statins for the reduction of C-reactive protein in type 2 diabetes: from the look AHEAD study Obesity (Silver Spring) 2013 21 944 950 23512860
32. Colquitt JL Pickett K Loveman E Surgery for weight loss in adults Cochrane Database Syst Rev 2014 8 CD003641
33. Patel KV Bahnson JL Gaussoin SA Association of baseline and longitudinal changes in body composition measures with risk of heart failure and myocardial infarction in type 2 diabetes: findings from the look AHEAD trial Circulation 2020 142 2420 2430 33164570
34. Franz MJ Boucher JL Rutten-Ramos S Lifestyle weight-loss intervention outcomes in overweight and obese adults with type 2 diabetes: a systematic review and meta-analysis of randomized clinical trials J Acad Nutr Diet 2015 115 1447 1463 25935570
35. Murgatroyd PR Goldberg GR Leahy FE Effects of inactivity and diet composition on human energy balance Int J Obes Relat Metab Disord 1999 23 1269 1275 10643683
36. Stubbs RJ Sepp A Hughes DA The effect of graded levels of exercise on energy intake and balance in free-living women Int J Obes Relat Metab Disord 2002 26 866 869 12037658
37. Granados K Stephens BR Malin SK Appetite regulation in response to sitting and energy imbalance Appl Physiol Nutr Metab 2012 37 323 333 22462636
38. Hägele FA Büsing F Nas A Appetite control is improved by acute increases in energy turnover at different levels of energy balance J Clin Endocrinol Metab 2019 104 4481 4491 31305927
39. Douglas JA King JA Clayton DJ Acute effects of exercise on appetite, ad libitum energy intake and appetite-regulatory hormones in lean and overweight/obese men and women Int J Obes (Lond) 2017 41 1737 1744 28769121
40. Savikj M Zierath JR Train like an athlete: applying exercise interventions to manage type 2 diabetes Diabetologia 2020 63 1491 1499 32529411
41. Büsing F Hägele FA Nas A Impact of energy turnover on the regulation of glucose homeostasis in healthy subjects Nutr Diabetes 2019 9 22 31395858
42. Larsen JJ Dela F Kjaer M The effect of moderate exercise on postprandial glucose homeostasis in NIDDM patients Diabetologia 1997 40 447 453 9112022
43. Heden TD Winn NC Mari A Postdinner resistance exercise improves postprandial risk factors more effectively than predinner resistance exercise in patients with type 2 diabetes J Appl Physiol 2015 118 624 634 25539939
44. Reynolds AN Mann JI Williams S Advice to walk after meals is more effective for lowering postprandial glycaemia in type 2 diabetes mellitus than advice that does not specify timing: a randomised crossover study Diabetologia 2016 59 2572 2578 27747394
45. Gaudet-Savard T Ferland A Broderick TL Safety and magnitude of changes in blood glucose levels following exercise performed in the fasted and the postprandial state in men with type 2 diabetes Eur J Cardiovasc Prev Rehabil 2007 14 831 836 18043307
46. DiPietro L Gribok A Stevens MS Three 15-min bouts of moderate postmeal walking significantly improves 24-h glycemic control in older people at risk for impaired glucose tolerance Diabetes Care 2013 36 3262 3268 23761134
47. Seidelmann SB Claggett B Cheng S Dietary carbohydrate intake and mortality: a prospective cohort study and meta-analysis Lancet Public Health 2018 3 e419 e428 30122560
48. Davies MJ D’Alessio DA Fradkin J Management of hyperglycemia in type 2 diabetes, 2018. A consensus report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) Diabetes Care 2018 41 2669 2701 30291106
49. Schwingshackl L Chaimani A Hoffmann G A network meta-analysis on the comparative efficacy of different dietary approaches on glycaemic control in patients with type 2 diabetes mellitus Eur J Epidemiol 2018 33 157 170 29302846
50. Schwingshackl L Hoffmann G Iqbal K Food groups and intermediate disease markers: a systematic review and network meta-analysis of randomized trials Am J Clin Nutr 2018 108 576 586 30535089
51. Neuenschwander M Ballon A Weber KS Role of diet in type 2 diabetes incidence: umbrella review of meta-analyses of prospective observational studies BMJ 2019 366 l2368 31270064
52. Ge L Sadeghirad B Ball GDC Comparison of dietary macronutrient patterns of 14 popular named dietary programmes for weight and cardiovascular risk factor reduction in adults: systematic review and network meta-analysis of randomised trials BMJ 2020 369 m696 32238384
53. Goldenberg JZ Day A Brinkworth GD Efficacy and safety of low and very low carbohydrate diets for type 2 diabetes remission: systematic review and meta-analysis of published and unpublished randomized trial data BMJ 2021 372 m4743 33441384
54. Schwingshackl L Nitschke K Zähringer J Impact of meal frequency on anthropometric outcomes: a systematic review and network meta-analysis of randomized controlled trials Adv Nutr 2020 11 1108 1122 32437566
55. Della Corte KW Perrar I Penczynski KJ Effect of dietary sugar intake on biomarkers of subclinical inflammation: a systematic review and meta-analysis of intervention studies Nutrients 2018 10 606 29757229
56. Schwingshackl L Chaimani A Schwedhelm C Comparative effects of different dietary approaches on blood pressure in hypertensive and pre-hypertensive patients: A systematic review and network meta-analysis Crit Rev Food Sci Nutr 2019 59 2674 2687 29718689
57. Thom G Messow CM Leslie WS Predictors of type 2 diabetes remission in the Diabetes Remission Clinical Trial (DiRECT) Diabet Med 2020 38 e14395 32870520
58. de Souza RJ Mente A Maroleanu A Intake of saturated and trans unsaturated fatty acids and risk of all cause mortality, cardiovascular disease, and type 2 diabetes: systematic review and meta-analysis of observational studies BMJ 2015 351 h3978 26268692
59. Astrup A Magkos F Bier DM Saturated fats and health: a reassessment and proposal for food-based recommendations: JACC state-of-the-art review J Am Coll Cardiol 2020 76 844 857 32562735
60. Pimpin L Wu JHY Haskelberg H Is butter back? A systematic review and meta-analysis of butter consumption and risk of cardiovascular disease, diabetes, and total mortality PLoS ONE 2016 11 e0158118 27355649
61. Benatar JR Sidhu K Stewart RAH Effects of high and low fat dairy food on cardio-metabolic risk factors: a meta-analysis of randomized studies PLoS ONE 2013 8 e76480 24146877
62. Hooper L Abdelhamid AS Jimoh OF Effects of total fat intake on body fatness in adults Cochrane Database Syst Rev 2020 6 CD013636 32476140
63. Hooper L Martin N Jimoh OF Reduction in saturated fat intake for cardiovascular disease Cochrane Database Syst Rev 2020 8 CD011737 32827219
64. Belalcazar LM Reboussin DM Haffner SM A 1-year lifestyle intervention for weight loss in individuals with type 2 diabetes reduces high C-reactive protein levels and identifies metabolic predictors of change: from the Look AHEAD (Action for Health in Diabetes) study Diabetes Care 2010 33 2297 2303 20682679
65. Lu M Wan Y Yang B Effects of low-fat compared with high-fat diet on cardiometabolic indicators in people with overweight and obesity without overt metabolic disturbance: a systematic review and meta-analysis of randomised controlled trials Br J Nutr 2018 119 96 108 29212558
66. Wu JHY Marklund M Imamura F Omega-6 fatty acid biomarkers and incident type 2 diabetes: pooled analysis of individual-level data for 39 740 adults from 20 prospective cohort studies Lancet Diabetes Endocrinol 2017 5 965 974 29032079
67. Li J Guasch-Ferré M Li Y Dietary intake and biomarkers of linoleic acid and mortality: systematic review and meta-analysis of prospective cohort studies Am J Clin Nutr 2020 112 150 167 32020162
68. an Pan A Chen M Chowdhury R α-Linolenic acid and risk of cardiovascular disease: a systematic review and meta-analysis Am J Clin Nutr 2012 96 1262 1273 23076616
69. Abdelhamid AS Martin N Bridges C Polyunsaturated fatty acids for the primary and secondary prevention of cardiovascular disease Cochrane Database Syst Rev 2018 11 CD012345 30484282
70. Abdelhamid AS Brown TJ Brainard JS Omega-3 fatty acids for the primary and secondary prevention of cardiovascular disease Cochrane Database Syst Rev 2020 3 CD003177 32114706
71. Brown TJ Brainard J Song F Omega-3, omega-6, and total dietary polyunsaturated fat for prevention and treatment of type 2 diabetes mellitus: systematic review and meta-analysis of randomised controlled trials BMJ 2019 366 l4697 31434641
72. Qian F Korat AA Malik V Metabolic effects of monounsaturated fatty acid-enriched diets compared with carbohydrate or polyunsaturated fatty acid-enriched diets in patients with type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials Diabetes Care 2016 39 1448 1457 27457635
73. Jovanovski E de Castro Ruiz Marques A Li D Effect of high-carbohydrate or high-monounsaturated fatty acid diets on blood pressure: a systematic review and meta-analysis of randomized controlled trials Nutr Rev 2019 77 19 31 30165599
74. Zhang YY Liu W Zhao TY Efficacy of omega-3 polyunsaturated fatty acids supplementation in managing overweight and obesity: a meta-analysis of randomized clinical trials J Nutr Health Aging 2017 21 187 192 28112774
75. Lin N Shi JJ Li YM What is the impact of n-3 PUFAs on inflammation markers in Type 2 diabetic mellitus populations?: a systematic review and meta-analysis of randomized controlled trials Lipids Health Dis 2016 15 133 27544079
76. Reis CEG Landim KC Nunes ACS Safety in the hypertriglyceridemia treatment with N-3 polyunsaturated fatty acids on glucose metabolism in subjects with type 2 diabetes mellitus Nutr Hosp 2014 31 570 576 25617537
77. Gao L Cao J Mao Q Influence of omega-3 polyunsaturated fatty acid-supplementation on platelet aggregation in humans: a meta-analysis of randomized controlled trials Atherosclerosis 2013 226 328 334 23153623
78. He XX Wu XL Chen RP Effectiveness of omega-3 polyunsaturated fatty acids in non-alcoholic fatty liver disease: a meta-analysis of randomized controlled trials PLoS ONE 2016 11 e0162368 27711128
79. Li N Yue H Jia M Effect of low-ratio n-6/n-3 PUFA on blood glucose: a meta-analysis Food Funct 2019 10 4557 4565 31292599
80. Wanders AJ Blom WAM Zock PL Plant-derived polyunsaturated fatty acids and markers of glucose metabolism and insulin resistance: a meta-analysis of randomized controlled feeding trials BMJ Open Diabetes Res Care 2019 7 e000585 30899527
81. Abbott KA Burrows TL Thota RN Do ω-3 PUFAs affect insulin resistance in a sex-specific manner? A systematic review and meta-analysis of randomized controlled trials Am J Clin Nutr 2016 104 1470 1484 27680989
82. Jovanovski E Li D Thanh Ho HV The effect of alpha-linolenic acid on glycemic control in individuals with type 2 diabetes: A systematic review and meta-analysis of randomized controlled clinical trials Medicine (Baltimore) 2017 96 e6531 28538363
83. Faris MAI Jahrami H BaHammam A A systematic review, meta-analysis, and meta-regression of the impact of diurnal intermittent fasting during Ramadan on glucometabolic markers in healthy subjects Diabetes Res Clin Pract 2020 165 108226 32446800
84. Mirmiran P Bahadoran Z Gaeini Z Effects of Ramadan intermittent fasting on lipid and lipoprotein parameters: An updated meta-analysis Nutr Metab Cardiovasc Dis 2019 29 906 915 31377182
85. Fernando HA Zibellini J Harris RA Effect of Ramadan fasting on weight and body composition in healthy non-athlete adults: a systematic review and meta-analysis Nutrients 2019 11 478 30813495
86. Horne BD May HT Anderson JL Usefulness of routine periodic fasting to lower risk of coronary artery disease in patients undergoing coronary angiography Am J Cardiol 2008 102 814 819 18805103
87. Horne BD Muhlestein JB May HT Relation of routine, periodic fasting to risk of diabetes mellitus, and coronary artery disease in patients undergoing coronary angiography Am J Cardiol 2012 109 1558 1562 22425331
88. Schwingshackl L Zähringer J Nitschke K Impact of intermittent energy restriction on anthropometric outcomes and intermediate disease markers in patients with overweight and obesity: systematic review and meta-analyses Crit Rev Food Sci Nutr 2021 61 1293 1304 32363896
89. Park J Seo YG Paek YJ Effect of alternate-day fasting on obesity and cardiometabolic risk: A systematic review and meta-analysis Metabolism 2020 111 154336 32777443
90. Harris L Hamilton S Azevedo LB Intermittent fasting interventions for treatment of overweight and obesity in adults: a systematic review and meta-analysis JBI Database System Rev Implement Rep 2018 16 507 547 29419624
91. Seimon RV Roekenes JA Zibellini J Do intermittent diets provide physiological benefits over continuous diets for weight loss? A systematic review of clinical trials Mol Cell Endocrinol 2015 418 Pt 2 153 172 26384657
92. Horne BD Muhlestein JB Anderson JL Health effects of intermittent fasting: hormesis or harm? A systematic review Am J Clin Nutr 2015 102 464 470 26135345
93. Borgundvaag E Mak J Kramer CK Metabolic impact of intermittent fasting in patients with type 2 diabetes mellitus: a systematic review and meta-analysis of Interventional studies J Clin Endocrinol Metab 2021 106 902 911 33319233
94. Parr EB Devlin BL Lim KHC Time-restricted eating as a nutrition strategy for individuals with type 2 diabetes: a feasibility study Nutrients 2020 12 3228 33105701
95. Carter S Clifton PM Keogh JB The effects of intermittent compared to continuous energy restriction on glycaemic control in type 2 diabetes; a pragmatic pilot trial Diabetes Res Clin Pract 2016 122 106 112 27833048
96. Carter S Clifton PM Keogh JB The effect of intermittent compared with continuous energy restriction on glycaemic control in patients with type 2 diabetes: 24-month follow-up of a randomised noninferiority trial Diabetes Res Clin Pract 2019 151 11 19 30902672
97. Corley BT Carroll RW Hall RM Intermittent fasting in Type 2 diabetes mellitus and the risk of hypoglycaemia: a randomized controlled trial Diabet Med 2018 35 588 594 29405359
98. Henry RR Wiest-Kent TA Scheaffer L Metabolic consequences of very-low-calorie diet therapy in obese non-insulin-dependent diabetic and nondiabetic subjects Diabetes 1986 35 155 164 3510922
99. Amatruda JM Richeson JF Welle SL The safety and efficacy of a controlled low-energy (‘very-low-calorie’) diet in the treatment of non-insulin-dependent diabetes and obesity Arch Intern Med 1988 148 873 877 3355307
100. Rotella CM Cresci B Mannucci E Short cycles of very low calorie diet in the therapy of obese type II diabetes mellitus J Endocrinol Invest 1994 17 171 179 8051339
101. Dhindsa P Scott AR Donnelly R Metabolic and cardiovascular effects of very-low-calorie diet therapy in obese patients with Type 2 diabetes in secondary failure: outcomes after 1 year Diabet Med 2003 20 319 324 12675647
102. ADA Standards of medical care in diabetes 2022 Clin Diabetes 2022 40 10 38 35221470
103. Lean MEJ Leslie WS Barnes AC Durability of a primary care-led weight-management intervention for remission of type 2 diabetes: 2-year results of the DiRECT open-label, cluster-randomised trial Lancet Diabetes Endocrinol 2019 7 344 355 30852132
104. Maggio CA Pi-Sunyer FX Obesity and type 2 diabetes Endocrinol Metab Clin North Am 2003 32 805 822 14711063
105. Wolf AM Colditz GA Current estimates of the economic cost of obesity in the United States Obes Res 1998 6 97 106 9545015
106. Colditz GA Willett WC Rotnitzky A Weight gain as a risk factor for clinical diabetes mellitus in women Ann Intern Med 1995 122 481 486 7872581
107. Leslie WS Taylor R Harris L Weight losses with low-energy formula diets in obese patients with and without type 2 diabetes: systematic review and meta-analysis Int J Obes (Lond) 2017 41 96 101 27698345
108. McCombie L Brosnahan N Ross H Filling the intervention gap: service evaluation of an intensive nonsurgical weight management programme for severe and complex obesity J Hum Nutr Diet 2019 32 329 337 30565772
109. Jazet IM de Craen AJ van Schie EM Sustained beneficial metabolic effects 18 months after a 30-day very low calorie diet in severely obese, insulin-treated patients with type 2 diabetes Diabetes Res Clin Pract 2007 77 70 76 17134786
110. Kempf K Schloot NC Gärtner B Meal replacement reduces insulin requirement, HbA1c and weight long-term in type 2 diabetes patients with 100 U insulin per day J Hum Nutr Diet 2014 27 Suppl. 02 21 27 23909831
111. Kempf K Röhling M Niedermeier K Individualized meal replacement therapy improves clinically relevant long-term glycemic control in poorly controlled type 2 diabetes patients Nutrients 2018 10 1022 30081574
112. Taylor R Leslie WS Barnes AC Clinical and metabolic features of the randomised controlled Diabetes Remission Clinical Trial (DiRECT) cohort Diabetologia 2018 61 589 598 29188339
113. Halle M Röhling M Banzer W Meal replacement by formula diet reduces weight more than a lifestyle intervention alone in patients with overweight or obesity and accompanied cardiovascular risk factors-the ACOORH trial Eur J Clin Nutr 2021 75 661 669 33128036
114. Röhling M Kempf K Banzer W Prediabetes conversion to Normoglycemia is superior adding a low-carbohydrate and energy deficit formula diet to lifestyle intervention-A 12-month Subanalysis of the ACOORH trial Nutrients 2020 12 2022 32646010
115. Rosenfeld RM Kelly JH Agarwal M Dietary intervention to treat T2DM in Adults with a goal of remission: An expert consensus statement from the American College of Lifestyle Medcine Am J Lifestyle Med 2022 16 342 362 35706589
116. Holman RR Paul SK Bethel MA 10-year follow-up of intensive glucose control in type 2 diabetes N Engl J Med 2008 359 1577 1589 18784090
117. Haslacher H Fallmann H Waldhäusl C Type 2 diabetes care: Improvement by standardization at a diabetes rehabilitation clinic. An observational report PLoS ONE 2019 14 e0226132 31830073
118. Paul SK Shaw JE Montvida O Weight gain in insulin-treated patients by body mass index category at treatment initiation: new evidence from real-world data in patients with type 2 diabetes Diabetes Obes Metab 2016 18 1244 1252 27502528
119. American Diabetes Association 5. Facilitating behavior change and well-being to improve health outcomes: standards of medical care in diabetes-2020 Diabetes Care 2020 43 S48 S65 31862748
120. Dyson PA Twenefour D Breen C Diabetes UK evidence-based nutrition guidelines for the prevention and management of diabetes Diabet Med 2018 35 541 547 29443421
121. Dworatzek PD Arcudi K Gougeon R Nutrition therapy Can J Diabetes 2013 37 Suppl. 01 S45 S55 24070963
122. Hallberg SJ Dockter NE Kushner JA Improving the scientific rigour of nutritional recommendations for adults with type 2 diabetes: A comprehensive review of the American Diabetes Association guideline-recommended eating patterns Diabetes Obes Metab 2019 21 1769 1779 30941880
123. Salas-Salvadó J Becerra-Tomás N Papandreou C Dietary patterns emphasizing the consumption of plant foods in the management of type 2 diabetes: a narrative review Adv Nutr 2019 10 S320 S331 31728494
124. Viguiliouk E Kendall CW Kahleová H Effect of vegetarian dietary patterns on cardiometabolic risk factors in diabetes: A systematic review and meta-analysis of randomized controlled trials Clin Nutr 2019 38 1133 1145 29960809
125. Papamichou D Panagiotakos DB Itsiopoulos C Dietary patterns and management of type 2 diabetes: A systematic review of randomised clinical trials Nutr Metab Cardiovasc Dis 2019 29 531 543 30952576
126. Ohlsson B An Okinawan-based Nordic diet improves glucose and lipid metabolism in health and type 2 diabetes, in alignment with changes in the endocrine profile, whereas zonulin levels are elevated Exp Ther Med 2019 17 2883 2893 30936958
127. Daneshzad E Emami S Darooghegi Mofrad M Association of modified Nordic diet with cardiovascular risk factors among type 2 diabetes patients: a cross-sectional study J Cardiovasc Thorac Res 2018 10 153 161 30386536
128. Via MA Mechanick JI Nutrition in type 2 diabetes and the metabolic syndrome Med Clin North Am 2016 100 1285 1302 27745595
129. Garvey WT Mechanick JI Brett EM American Association of Clinical Endocrinologists and American College of Endocrinology Comprehensive CLINICAL practice guidelines for medical care of patients with obesity Endocr Pract 2016 22 Suppl. 03 1 203 27219496
130. Ajala O English P Pinkney J Systematic review and meta-analysis of different dietary approaches to the management of type 2 diabetes Am J Clin Nutr 2013 97 505 516 23364002
131. Huo R Du T Xu Y Effects of Mediterranean-style diet on glycemic control, weight loss and cardiovascular risk factors among type 2 diabetes individuals: a meta-analysis Eur J Clin Nutr 2015 69 1200 1208 25369829
132. Pan B Wu Y Yang Q The impact of major dietary patterns on glycemic control, cardiovascular risk factors, and weight loss in patients with type 2 diabetes: A network meta-analysis J Evid Based Med 2019 12 29 39 30070019
133. Johannesen CO Dale HF Jensen C Effects of plant-based diets on outcomes related to glucose metabolism: a systematic review Diabetes Metab Syndr Obes 2020 13 2811 2822 32884310
134. Toumpanakis A Turnbull T Alba-Barba I Effectiveness of plant-based diets in promoting well-being in the management of type 2 diabetes: a systematic review BMJ Open Diabetes Res Care 2018 6 e000534 30487971
135. Tran E Dale HF Jensen C Effects of plant-based diets on weight status: a systematic review Diabetes Metab Syndr Obes 2020 13 3433 3448 33061504
136. Austin G Ferguson J Garg M Effects of plant-based diets on weight status in type 2 diabetes: a systematic review and meta-analysis of randomised controlled trials Nutrients 2021 13 4099 34836356
137. Esposito K Maiorino MI Bellastella G A journey into a Mediterranean diet and type 2 diabetes: a systematic review with meta-analyses BMJ Open 2015 5 e008222 26260349
138. Carter P Achana F Troughton J A Mediterranean diet improves HbA1c but not fasting blood glucose compared to alternative dietary strategies: a network meta-analysis J Hum Nutr Diet 2014 27 280 297 23790149
139. Emadian A Andrews RC England CY The effect of macronutrients on glycaemic control: a systematic review of dietary randomised controlled trials in overweight and obese adults with type 2 diabetes in which there was no difference in weight loss between treatment groups Br J Nutr 2015 114 1656 1666 26411958
140. Kahleova H Salas-Salvadó J Rahelić D Dietary patterns and Cardiometabolic outcomes in diabetes: A summary of systematic reviews and meta-analyses Nutrients 2019 11 2209 31540227
141. Deutsche Diabetes Gesellschaft (DDG) https://www.deutsche-diabetes-gesellschaft.de/fileadmin/user_upload/01_Die_DDG/03_Ausschuesse/02_Ernaehrung/2015-057-025l_S3_Diabetes_mellitus_Empfehlungen_Proteinzufuhr_2015-10.pdf. Zugegriffen: 6. Juli 2021
142. Pfeiffer AFH Pedersen E Schwab U The effects of different quantities and qualities of protein intake in people with diabetes mellitus Nutrients 2020 12 365 32019211
143. Mittendorfer B Klein S Fontana L A word of caution against excessive protein intake Nat Rev Endocrinol 2020 16 59 66 31728051
144. Labonte CC Chevalier S Marliss EB Effect of 10% dietary protein intake on whole body protein kinetics in type 2 diabetic adults Clin Nutr 2015 34 1115 1121 25604923
145. Markova M Hornemann S Sucher S Rate of appearance of amino acids after a meal regulates insulin and glucagon secretion in patients with type 2 diabetes: a randomized clinical trial Am J Clin Nutr 2018 108 279 291 29982277
146. Volkert D Aktuelle ESPEN-Leitlinie Klinische Ernährung und Hydration in der Geriatrie Dtsch Med Wochenschr 2020 145 1306 1314 32906183
147. Song M Fung TT Hu FB Association of animal and plant protein intake with all-cause and cause-specific mortality JAMA Intern Med 2016 176 1453 1463 27479196
148. Ye J Yu Q Mai W Dietary protein intake and subsequent risk of type 2 diabetes: a dose-response meta-analysis of prospective cohort studies Acta Diabetol 2019 56 851 870 30929078
149. Vernooij RWM Zeraatkar D Han MA Patterns of red and processed meat consumption and risk for cardiometabolic and cancer outcomes: a systematic review and meta-analysis of cohort studies Ann Intern Med 2019 171 732 741 31569217
150. Vogtschmidt YD Raben A Faber I Is protein the forgotten ingredient: Effects of higher compared to lower protein diets on cardiometabolic risk factors. A systematic review and meta-analysis of randomised controlled trials Atherosclerosis 2021 328 124 135 34120735
151. Clifton PM Condo D Keogh JB Long term weight maintenance after advice to consume low carbohydrate, higher protein diets—A systematic review and meta analysis Nutr Metab Cardiovasc Dis 2014 24 224 235 24472635
152. Hahn D Hodson EM Fouque D Low protein diets for non-diabetic adults with chronic kidney disease Cochrane Database Syst Rev 2020 10 CD001892 33118160
153. Ikizler TA Burrowes JD Byham-Gray LD KDOQI clinical practice guideline for nutrition in CKD: 2020 update Am J Kidney Dis 2020 76 S1 S107 32829751
154. Menon V Kopple JD Wang X Effect of a very low-protein diet on outcomes: long-term follow-up of the Modification of Diet in Renal Disease (MDRD) Study Am J Kidney Dis 2009 53 208 217 18950911
155. Jiang Z (2017) Effect of restricted protein diet supplemented with keto analogues in end-stage renal disease: a systematic review and meta-analysis. Int Urol Nephrol 1–8
156. Fiaccadori E Sabatino A Barazzoni R ESPEN guideline on clinical nutrition in hospitalized patients with acute or chronic kidney disease Clin Nutr 2021 40 1644 1668 33640205
157. Dong JY Zhang ZL Wang PY Effects of high-protein diets on body weight, glycaemic control, blood lipids and blood pressure in type 2 diabetes: meta-analysis of randomised controlled trials Br J Nutr 2013 110 781 789 23829939
158. Anonymous Carbohydrates in human nutrition. Report of a Joint FAO/WHO expert consultation FAO Food Nutr Pap 1998 66 1 140 9743703
159. Wolever TMS Personalized nutrition by prediction of glycaemic responses: fact or fantasy? Eur J Clin Nutr 2016 70 411 413 27050901
160. Berry SE Valdes AM Drew DA Human postprandial responses to food and potential for precision nutrition Nat Med 2020 26 964 973 32528151
161. Zeevi D Korem T Zmora N Personalized nutrition by prediction of glycemic responses Cell 2015 163 1079 1094 26590418
162. Livesey G Taylor R Livesey H Is there a dose-response relation of dietary glycemic load to risk of type 2 diabetes? Meta-analysis of prospective cohort studies Am J Clin Nutr 2013 97 584 596 23364021
163. Livesey G Livesey H Coronary heart disease and dietary carbohydrate, glycemic index, and glycemic load: dose-response meta-analyses of prospective cohort studies Mayo Clin Proc Innov Qual Outcomes 2019 3 52 69 30899909
164. Thomas DE Elliott EJ The use of low-glycaemic index diets in diabetes control Br J Nutr 2010 104 797 802 20420752
165. Xu B Fu J Qiao Y Higher intake of microbiota-accessible carbohydrates and improved cardiometabolic risk factors: a meta-analysis and umbrella review of dietary management in patients with type 2 diabetes Am J Clin Nutr 2021 113 1515 1530 33693499
166. Jenkins DJA Kendall CWC McKeown-Eyssen G Effect of a low-glycemic index or a high-cereal fiber diet on type 2 diabetes: a randomized trial JAMA 2008 300 2742 2753 19088352
167. Holub I Gostner A Hessdörfer S Improved metabolic control after 12-week dietary intervention with low glycaemic isomalt in patients with type 2 diabetes mellitus Horm Metab Res 2009 41 886 892 19701877
168. Brand-Miller J Hayne S Petocz P Low-glycemic index diets in the management of diabetes: a meta-analysis of randomized controlled trials Diabetes Care 2003 26 2261 2267 12882846
169. Ojo O Ojo OO Adebowale F The effect of dietary glycaemic index on glycaemia in patients with type 2 diabetes: A systematic review and meta-analysis of randomized controlled trials Nutrients 2018 10 373 29562676
170. Franz MJ MacLeod J Evert A Academy of nutrition and dietetics nutrition practice guideline for type 1 and type 2 diabetes in adults: Systematic review of evidence for medical nutrition therapy effectiveness and recommendations for integration into the nutrition care process J Acad Nutr Diet 2017 117 1659 1679 28533169
171. Vega-López S Venn BJ Slavin JL Relevance of the glycemic index and Glycemic load for body weight, diabetes, and cardiovascular disease Nutrients 2018 10 1361 30249012
172. Jenkins DJA Dehghan M Mente A Glycemic index, glycemic load, and cardiovascular disease and mortality N Engl J Med 2021 384 1312 1322 33626252
173. Coutinho M Gerstein HC Wang Y The relationship between glucose and incident cardiovascular events. A metaregression analysis of published data from 20 studies of 95783 individuals followed for 12.4 years Diabetes Care 1999 22 233 240 10333939
174. Levitan EB Song Y Ford ES Is nondiabetic hyperglycemia a risk factor for cardiovascular disease? A meta-analysis of prospective studies Arch Intern Med 2004 164 2147 2155 15505129
175. Siri PW Krauss RM Influence of dietary carbohydrate and fat on LDL and HDL particle distributions Curr Atheroscler Rep 2005 7 455 459 16256003
176. Aune D Norat T Romundstad P Whole grain and refined grain consumption and the risk of type 2 diabetes: a systematic review and dose-response meta-analysis of cohort studies Eur J Epidemiol 2013 28 845 858 24158434
177. InterAct Consortium Dietary fibre and incidence of type 2 diabetes in eight European countries: the EPIC-InterAct Study and a meta-analysis of prospective studies Diabetologia 2015 58 1394 1408 26021487
178. Kim Y Je Y Dietary fibre intake and mortality from cardiovascular disease and all cancers: A meta-analysis of prospective cohort studies Arch Cardiovasc Dis 2016 109 39 54 26711548
179. Reynolds AN Akerman AP Mann J Dietary fibre and whole grains in diabetes management: Systematic review and meta-analyses PLoS Med 2020 17 e1003053 32142510
180. Da Silva Borges D Fernandes R Thives Mello A Prebiotics may reduce serum concentrations of C-reactive protein and ghrelin in overweight and obese adults: a systematic review and meta-analysis Nutr Rev 2020 78 235 248 31504857
181. Reynolds A Mann J Cummings J Carbohydrate quality and human health: a series of systematic reviews and meta-analyses Lancet 2019 393 434 445 30638909
182. Musa-Veloso K Poon T Harkness LS The effects of whole-grain compared with refined wheat, rice, and rye on the postprandial blood glucose response: a systematic review and meta-analysis of randomized controlled trials Am J Clin Nutr 2018 108 759 774 30321274
183. Wang W Li J Chen X Whole grain food diet slightly reduces cardiovascular risks in obese/overweight adults: a systematic review and meta-analysis BMC Cardiovasc Disord 2020 20 82 32070285
184. Weickert MO Roden M Isken F Effects of supplemented isoenergetic diets differing in cereal fiber and protein content on insulin sensitivity in overweight humans Am J Clin Nutr 2011 94 459 471 21633074
185. Honsek C Kabisch S Kemper M Fibre supplementation for the prevention of type 2 diabetes and improvement of glucose metabolism: the randomised controlled Optimal Fibre Trial (OptiFiT) Diabetologia 2018 61 1295 1305 29492637
186. Kabisch S Meyer NMT Honsek C Fasting glucose state determines metabolic response to supplementation with insoluble cereal fibre: a secondary analysis of the optimal fibre trial (OptiFiT) Nutrients 2019 11 2385 31590438
187. Hjorth MF Ritz C Blaak EE Pretreatment fasting plasma glucose and insulin modify dietary weight loss success: results from 3 randomized clinical trials Am J Clin Nutr 2017 106 499 505 28679551
188. Xiao Z Chen H Zhang Y The effect of psyllium consumption on weight, body mass index, lipid profile, and glucose metabolism in diabetic patients: A systematic review and dose-response meta-analysis of randomized controlled trials Phytother Res 2020 34 1237 1247 31919936
189. Wang L Yang H Huang H Inulin-type fructans supplementation improves glycemic control for the prediabetes and type 2 diabetes populations: results from a GRADE-assessed systematic review and dose-response meta-analysis of 33 randomized controlled trials J Transl Med 2019 17 410 31805963
190. Rao M Gao C Xu L Effect of inulin-type carbohydrates on insulin resistance in patients with type 2 diabetes and obesity: a systematic review and meta-analysis J Diabetes Res 2019 2019 5101423 31534973
191. Darooghegi Mofrad M Mozaffari H Mousavi SM The effects of psyllium supplementation on body weight, body mass index and waist circumference in adults: A systematic review and dose-response meta-analysis of randomized controlled trials Crit Rev Food Sci Nutr 2020 60 859 872 30880409
192. Rahmani J Miri A Černevičiūtė R Effects of cereal beta-glucan consumption on body weight, body mass index, waist circumference and total energy intake: A meta-analysis of randomized controlled trials Complement Ther Med 2019 43 131 139 30935520
193. Ho HVT Sievenpiper JL Zurbau A The effect of oat β-glucan on LDL-cholesterol, non-HDL-cholesterol and apoB for CVD risk reduction: a systematic review and meta-analysis of randomised-controlled trials Br J Nutr 2016 116 1369 1382 27724985
194. Jovanovski E Yashpal S Komishon A Effect of psyllium (Plantago ovata) fiber on LDL cholesterol and alternative lipid targets, non-HDL cholesterol and apolipoprotein B: a systematic review and meta-analysis of randomized controlled trials Am J Clin Nutr 2018 108 922 932 30239559
195. Brum J Ramsey D McRorie J meta-analysis of usefulness of psyllium fiber as adjuvant antilipid therapy to enhance cholesterol lowering efficacy of statins Am J Cardiol 2018 122 1169 1174 30078477
196. Ho HVT Jovanovski E Zurbau A A systematic review and meta-analysis of randomized controlled trials of the effect of konjac glucomannan, a viscous soluble fiber, on LDL cholesterol and the new lipid targets non-HDL cholesterol and apolipoprotein B Am J Clin Nutr 2017 105 1239 1247 28356275
197. Pittler MH Ernst E Guar gum for body weight reduction: meta-analysis of randomized trials Am J Med 2001 110 724 730 11403757
198. Khan K Jovanovski E Ho HVT The effect of viscous soluble fiber on blood pressure: A systematic review and meta-analysis of randomized controlled trials Nutr Metab Cardiovasc Dis 2018 28 3 13 29153856
199. Thinggaard M Jacobsen R Jeune B Is the relationship between BMI and mortality increasingly U-shaped with advancing age? A 10-year follow-up of persons aged 70–95 years J Gerontol A Biol Sci Med Sci 2010 65 526 531 20089666
200. Guigoz Y Vellas B Malnutrition in the elderly: the Mini Nutritional Assessment (MNA) Ther Umsch 1997 54 345 350 9289873
201. Anonym S2k-Leitlinie Diagnostik, Therapie und Verlaufskontrolle des Diabetes mellitus im Alter. 2. Auflage 2018 – AWMF-Register-Nr. 057-017 Diabetologie 2018 13 423 489
202. Volkert D Bauer J Frühwald T S3-Leitlinie der Deutschen Gesellschaft für Ernährungsmedizin (DGEM) in Zusammenarbeit mit der GESKES, der AKE und der DGG Klinische Ernährung in der Geriatrie Aktuel Ernahrungsmed 2013 38 e1 e48
203. Zeyfang A Wernecke J Bahrmann A Diabetes mellitus im Alter Diabetologie 2020 15 S112 S119
204. Şat S Aydınkoç-Tuzcu K Berger F Diabetes und Migration Diabetologie 2019 14 Suppl. 02 S306 S317
205. Diker O Deniz T Çetinkaya A History of Turkish cuisine culture and the influence of the balkans IOSR J Humanit Soc Sci 2016 10 1 6
206. Schmid B Ernährung und Migration 2003 München Utz
207. Magni P Bier DM Pecorelli S Perspective: improving nutritional guidelines for sustainable health policies: current status and perspectives Adv Nutr 2017 8 532 545 28710141
208. Praxistool zur Ernährung. Orientierungshilfe für die Diabetesberatung nach geografischen Räumen. https://migration.deutsche-diabetes-gesellschaft.de/fileadmin/user_upload/01_Die_DDG/05_Arbeitsgemeinschaften/AG_Migranten/Microsite/200417_Ernaehrungstoo_DDG-GB19-Einleger_04.pdf. Zugegriffen: 15. Juli 2021
209. European Commission Health promotion and disease prevention knowledge gateway: sugars and sweeteners. https://ec.europa.eu/jrc/en/health-knowledge-gateway/promotion-prevention/nutrition/sugars-sweeteners. Zugegriffen: 27. Jan. 2021
210. Scientific Advisory Committee on Nutrition (2015) Carbohydrates and health report. https://www.gov.uk/government/publications/. Zugegriffen: 26. Jan. 2021
211. McKeown NM Dashti HS Ma J Sugar-sweetened beverage intake associations with fasting glucose and insulin concentrations are not modified by selected genetic variants in a ChREBP-FGF21 pathway: a meta-analysis Diabetologia 2018 61 317 330 29098321
212. Evans RA Frese M Romero J Chronic fructose substitution for glucose or sucrose in food or beverages has little effect on fasting blood glucose, insulin, or triglycerides: a systematic review and meta-analysis Am J Clin Nutr 2017 106 519 529 28592603
213. Evans RA Frese M Romero J Fructose replacement of glucose or sucrose in food or beverages lowers postprandial glucose and insulin without raising triglycerides: a systematic review and meta-analysis Am J Clin Nutr 2017 106 506 518 28592611
214. Keller A Heitmann BL Olsen N Sugar-sweetened beverages, vascular risk factors and events: a systematic literature review Public Health Nutr 2015 18 1145 1154 25321082
215. Huang C Huang J Tian Y Sugar sweetened beverages consumption and risk of coronary heart disease: a meta-analysis of prospective studies Atherosclerosis 2014 234 11 16 24583500
216. Narain A Kwok CS Mamas MA Soft drinks and sweetened beverages and the risk of cardiovascular disease and mortality: a systematic review and meta-analysis Int J Clin Pract 2016 70 791 805 27456347
217. Cheungpasitporn W Thongprayoon C O’Corragain OA Associations of sugar-sweetened and artificially sweetened soda with chronic kidney disease: a systematic review and meta-analysis Nephrology (Carlton) 2014 19 791 797 25251417
218. Chen H Wang J Li Z Consumption of sugar-sweetened beverages has a dose-dependent effect on the risk of non-alcoholic fatty liver disease: an updated systematic review and dose-response meta-analysis Int J Environ Res Public Health 2019 16 2192 31234281
219. Asgari-Taee F Zerafati-Shoae N Dehghani M Association of sugar sweetened beverages consumption with non-alcoholic fatty liver disease: a systematic review and meta-analysis Eur J Nutr 2019 58 1759 1769 29761318
220. Khan TA Sievenpiper JL Controversies about sugars: results from systematic reviews and meta-analyses on obesity, cardiometabolic disease and diabetes Eur J Nutr 2016 55 25 43 27900447
221. Choo VL Viguiliouk E Blanco Mejia S Food sources of fructose-containing sugars and glycaemic control: systematic review and meta-analysis of controlled intervention studies BMJ 2018 363 k4644 30463844
222. Semnani-Azad Z Khan TA Blanco Mejia S Association of major food sources of fructose-containing sugars with incident metabolic syndrome: a systematic review and meta-analysis JAMA Netw Open 2020 3 e209993 32644139
223. Bechthold A Vollwertig essen und trinken nach den 10 Regeln der DGE 2018 Bonn Deutsche Gesellschaft für Ernährung e. V. (DGE)
224. Wu H Flint AJ Qi Q Association between dietary whole grain intake and risk of mortality: two large prospective studies in US men and women JAMA Intern Med 2015 175 373 384 25559238
225. Johnsen NF Frederiksen K Christensen J Whole-grain products and whole-grain types are associated with lower all-cause and cause-specific mortality in the Scandinavian HELGA cohort Br J Nutr 2015 114 608 623 26201847
226. Wei H Gao Z Liang R Whole-grain consumption and the risk of all-cause, CVD and cancer mortality: a meta-analysis of prospective cohort studies—CORRIGENDUM Br J Nutr 2016 116 952 27523188
227. Chen GC Tong X Xu JY Whole-grain intake and total, cardiovascular, and cancer mortality: a systematic review and meta-analysis of prospective studies Am J Clin Nutr 2016 104 164 172 27225432
228. Benisi-Kohansal S Saneei P Salehi-Marzijarani M Whole-grain intake and mortality from all causes, cardiovascular disease, and cancer: a systematic review and dose-response meta-analysis of prospective cohort studies Adv Nutr 2016 7 1052 1065 28140323
229. Zong G Gao A Hu FB Whole grain intake and mortality from all causes, cardiovascular disease, and cancer: a meta-analysis of prospective cohort studies Circulation 2016 133 2370 2380 27297341
230. Aune D Keum N Giovannucci E Whole grain consumption and risk of cardiovascular disease, cancer, and all cause and cause specific mortality: systematic review and dose-response meta-analysis of prospective studies BMJ 2016 353 i2716 27301975
231. Aune D Plant foods, antioxidant biomarkers, and the risk of cardiovascular disease, cancer, and mortality: a review of the evidence Adv Nutr 2019 10 S404 S421 31728499
232. Zhang B Zhao Q Guo W Association of whole grain intake with all-cause, cardiovascular, and cancer mortality: a systematic review and dose-response meta-analysis from prospective cohort studies Eur J Clin Nutr 2018 72 57 65 29091078
233. Jenkins DJ Wesson V Wolever TM Wholemeal versus wholegrain breads: proportion of whole or cracked grain and the glycaemic response BMJ 1988 297 958 960 3142566
234. Reynolds AN Mann J Elbalshy M Wholegrain particle size influences postprandial glycemia in type 2 diabetes: A randomized crossover study comparing four wholegrain breads Diabetes Care 2020 43 476 479 31744812
235. Åberg S Mann J Neumann S Whole-grain processing and glycemic control in type 2 diabetes: a randomized crossover trial Diabetes Care 2020 43 1717 1723 32424022
236. Jenkins DJA Kendall CWC Augustin LSA Effect of wheat bran on glycemic control and risk factors for cardiovascular disease in type 2 diabetes Diabetes Care 2002 25 1522 1528 12196421
237. Miller V Mente A Dehghan M Fruit, vegetable, and legume intake, and cardiovascular disease and deaths in 18 countries (PURE): a prospective cohort study Lancet 2017 390 2037 2049 28864331
238. Aune D Giovannucci E Boffetta P Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality—A systematic review and dose-response meta-analysis of prospective studies Int J Epidemiol 2017 46 1029 1056 28338764
239. Bechthold A Boeing H Schwedhelm C Food groups and risk of coronary heart disease, stroke and heart failure: A systematic review and dose-response meta-analysis of prospective studies Crit Rev Food Sci Nutr 2019 59 1071 1090 29039970
240. Zhan J Liu YJ Cai LB Fruit and vegetable consumption and risk of cardiovascular disease: A meta-analysis of prospective cohort studies Crit Rev Food Sci Nutr 2017 57 1650 1663 26114864
241. Willett W Rockström J Loken B Food in the Anthropocene: the EAT-Lancet Commission on healthy diets from sustainable food systems Lancet 2019 393 447 492 30660336
242. Barnard ND Cohen J Jenkins DJA A low-fat vegan diet improves glycemic control and cardiovascular risk factors in a randomized clinical trial in individuals with type 2 diabetes Diabetes Care 2006 29 1777 1783 16873779
243. Jenkins DJA Kendall CWC Augustin LSA Effect of legumes as part of a low glycemic index diet on glycemic control and cardiovascular risk factors in type 2 diabetes mellitus: a randomized controlled trial Arch Intern Med 2012 172 1653 1660 23089999
244. Renner B Arens-Azevêdo U Watzl B DGE-Positionspapier zur nachhaltigeren Ernährung Ernährungsumschau 2021 68 144 154
245. Jannasch F Kröger J Schulze MB Dietary patterns and type 2 diabetes: A systematic literature review and meta-analysis of prospective studies J Nutr 2017 147 1174 1182 28424256
246. Wallin A Di Giuseppe D Orsini N Fish consumption, dietary long-chain n-3 fatty acids, and risk of type 2 diabetes: systematic review and meta-analysis of prospective studies Diabetes Care 2012 35 918 929 22442397
247. Xun P He K Fish Consumption and Incidence of Diabetes: meta-analysis of data from 438000 individuals in 12 independent prospective cohorts with an average 11-year follow-up Diabetes Care 2012 35 930 938 22442398
248. Schwingshackl L Hoffmann G Lampousi AM Food groups and risk of type 2 diabetes mellitus: a systematic review and meta-analysis of prospective studies Eur J Epidemiol 2017 32 363 375 28397016
249. Muley A Muley P Shah M ALA, fatty fish or marine n-3 fatty acids for preventing DM?: a systematic review and meta-analysis Curr Diabetes Rev 2014 10 158 165 24828061
250. Schlesinger S Neuenschwander M Schwedhelm C Food groups and risk of overweight, obesity, and weight gain: a systematic review and dose-response meta-analysis of prospective studies Adv Nutr 2019 10 205 218 30801613
251. Micha R Shulkin ML Peñalvo JL Etiologic effects and optimal intakes of foods and nutrients for risk of cardiovascular diseases and diabetes: Systematic reviews and meta-analyses from the Nutrition and Chronic Diseases Expert Group (NutriCoDE) PLoS ONE 2017 12 e0175149 28448503
252. Jayedi A Shab-Bidar S Eimeri S Fish consumption and risk of all-cause and cardiovascular mortality: a dose-response meta-analysis of prospective observational studies Public Health Nutr 2018 21 1297 1306 29317009
253. Abdelhamid AS Brown TJ Brainard JS Omega-3 fatty acids for the primary and secondary prevention of cardiovascular disease Cochrane Database Syst Rev 2018 11 CD003177 30521670
254. Hu Y Hu FB Manson JE Marine omega-3 supplementation and cardiovascular disease: an updated meta-analysis of 13 randomized controlled trials involving 127 477 participants J Am Heart Assoc 2019 8 e013543 31567003
255. Gao H Geng T Huang T Fish oil supplementation and insulin sensitivity: a systematic review and meta-analysis Lipids Health Dis 2017 16 131 28673352
256. Chen C Yu X Shao S Effects of omega-3 fatty acid supplementation on glucose control and lipid levels in type 2 diabetes: a meta-analysis PLoS ONE 2015 10 e0139565 26431431
257. DGE – Deutsche Gesellschaft für Ernährung. Vollwertig essen und trinken nach den 10 Regeln der DGE. https://www.dge.de/ernaehrungspraxis/vollwertige-ernaehrung/10-regeln-der-dge/. Zugegriffen: 13. Juli 2021
258. Zeraatkar D Han MA Guyatt GH Red and processed meat consumption and risk for all-cause mortality and cardiometabolic outcomes: a systematic review and meta-analysis of cohort studies Ann Intern Med 2019 171 703 710 31569213
259. Davidson MH Hunninghake D Maki KC Comparison of the effects of lean red meat vs lean white meat on serum lipid levels among free-living persons with hypercholesterolemia: a long-term, randomized clinical trial Arch Intern Med 1999 159 1331 1338 10386509
260. Hunninghake DB Maki KC Kwiterovich PO Incorporation of lean red meat into a National Cholesterol Education Program Step I diet: a long-term, randomized clinical trial in free-living persons with hypercholesterolemia J Am Coll Nutr 2000 19 351 360 10872897
261. Bergeron N Chiu S Williams PT Effects of red meat, white meat, and nonmeat protein sources on atherogenic lipoprotein measures in the context of low compared with high saturated fat intake: a randomized controlled trial Am J Clin Nutr 2019 110 24 33 31161217
262. Charlton K Walton K Batterham M Pork and chicken meals similarly impact on cognitive function and strength in community-living older adults: a pilot study J Nutr Gerontol Geriatr 2016 35 124 145 27153252
263. Murphy KJ Parker B Dyer KA A comparison of regular consumption of fresh lean pork, beef and chicken on body composition: a randomized cross-over trial Nutrients 2014 6 682 696 24534884
264. Murphy KJ Thomson RL Coates AM Effects of eating fresh lean pork on cardiometabolic health parameters Nutrients 2012 4 711 723 22852059
265. Johnston BC Zeraatkar D Han MA Unprocessed red meat and processed meat consumption: Dietary guideline recommendations from the nutritional recommendations (NutriRECS) consortium Ann Intern Med 2019 171 756 764 31569235
266. Davis PA Yokoyama W Cinnamon intake lowers fasting blood glucose: meta-analysis J Med Food 2011 14 884 889 21480806
267. Akilen R Tsiami A Devendra D Cinnamon in glycaemic control: Systematic review and meta analysis Clin Nutr 2012 31 609 615 22579946
268. Leach MJ Kumar S Cinnamon for diabetes mellitus Cochrane Database Syst Rev 2012 2012 CD007170 22972104
269. Allen RW Schwartzman E Baker WL Cinnamon use in type 2 diabetes: an updated systematic review and meta-analysis Ann Fam Med 2013 11 452 459 24019277
270. Costello RB Dwyer JT Saldanha L Do cinnamon supplements have a role in Glycemic control in type 2 diabetes? A narrative review J Acad Nutr Diet 2016 116 1794 1802 27618575
271. Sierra-Puente D Abadi-Alfie S Arakanchi-Altaled K Cinammon (Cinnamomum Spp.) and type 2 diabetes mellitus CTNR 2019 18 247 255
272. Chan CB Hashemi Z Subhan FB The impact of low and no-caloric sweeteners on glucose absorption, incretin secretion, and glucose tolerance Appl Physiol Nutr Metab 2017 42 793 801 28407474
273. Brown AW Bohan Brown MM Onken KL Short-term consumption of sucralose, a nonnutritive sweetener, is similar to water with regard to select markers of hunger signaling and short-term glucose homeostasis in women Nutr Res 2011 31 882 888 22153513
274. Ford HE Peters V Martin NM Effects of oral ingestion of sucralose on gut hormone response and appetite in healthy normal-weight subjects Eur J Clin Nutr 2011 65 508 513 21245879
275. Steinert RE Frey F Töpfer A Effects of carbohydrate sugars and artificial sweeteners on appetite and the secretion of gastrointestinal satiety peptides Br J Nutr 2011 105 1320 1328 21255472
276. Barriocanal LA Palacios M Benitez G Apparent lack of pharmacological effect of steviol glycosides used as sweeteners in humans. A pilot study of repeated exposures in some normotensive and hypotensive individuals and in Type 1 and Type 2 diabetics Regul Toxicol Pharmacol 2008 51 37 41 18397817
277. Brown RJ Walter M Rother KI Effects of diet soda on gut hormones in youths with diabetes Diabetes Care 2012 35 959 964 22410815
278. Grotz VL Henry RR McGill JB Lack of effect of sucralose on glucose homeostasis in subjects with type 2 diabetes J Am Diet Assoc 2003 103 1607 1612 14647086
279. Maki KC Curry LL Reeves MS Chronic consumption of rebaudioside A, a steviol glycoside, in men and women with type 2 diabetes mellitus Food Chem Toxicol 2008 46 Suppl. 07 S47 53 18555575
280. Olalde-Mendoza L Moreno-González YE Modificación de la glucemia en ayuno en adultos con diabetes mellitus tipo 2 después de la ingesta de refrescos de cola y de dieta en el estado de querétaro, México Arch Latinoam Nutr 2013 63 142 147 24934070
281. Temizkan S Deyneli O Yasar M Sucralose enhances GLP-1 release and lowers blood glucose in the presence of carbohydrate in healthy subjects but not in patients with type 2 diabetes Eur J Clin Nutr 2015 69 162 166 25271009
282. Ferrazzano GF Cantile T Alcidi B Is Stevia rebaudiana bertoni a non cariogenic sweetener? A review Molecules 2015 21 E38 26712732
283. Prashant GM Patil RB Nagaraj T The antimicrobial activity of the three commercially available intense sweeteners against common periodontal pathogens: an in vitro study J Contemp Dent Pract 2012 13 749 752 23403996
284. Suez J Korem T Zeevi D Artificial sweeteners induce glucose intolerance by altering the gut microbiota Nature 2014 514 181 186 25231862
285. EFSA (2013) EFSA schließt vollständige Risikobewertung zu Aspartam ab und kommt zu dem Schluss, dass es in den derzeitigen Expositionsmengen sicher ist. https://www.efsa.europa.eu/de/press/news/131210. Zugegriffen: 1. Sept. 2020
286. Bundesinstitut für Risikobewertung (2014) Bewertung von Süßstoffen und Zuckeraustauschstoffen. Hintergrundinformation Nr. 025/2014 des BfR vom 1. Juli 2014. www.bfr.bund.de/cm/343/bewertung_von_suessstoffen.pdf. Zugegriffen: 1. Sept. 2020
287. Bock PM Telo GH Ramalho R The effect of probiotics, prebiotics or synbiotics on metabolic outcomes in individuals with diabetes: a systematic review and meta-analysis Diabetologia 2021 64 26 41 33047170
288. Rittiphairoj T Pongpirul K Janchot K Probiotics contribute to glycemic control in patients with type 2 diabetes mellitus: A systematic review and meta-analysis Adv Nutr 2021 12 722 734 33126241
289. Tao YW Gu YL Mao XQ Effects of probiotics on type II diabetes mellitus: a meta-analysis J Transl Med 2020 18 30 31952517
290. Ardeshirlarijani E Tabatabaei-Malazy O Mohseni S Effect of probiotics supplementation on glucose and oxidative stress in type 2 diabetes mellitus: a meta-analysis of randomized trials Daru 2019 27 827 837 31691101
291. Mahboobi S Rahimi F Jafarnejad S Effects of prebiotic and Synbiotic supplementation on glycaemia and lipid profile in type 2 diabetes: a meta-analysis of randomized controlled trials Adv Pharm Bull 2018 8 565 574 30607329
292. Akbari V Hendijani F Effects of probiotic supplementation in patients with type 2 diabetes: systematic review and meta-analysis Nutr Rev 2016 74 774 784 27864537
293. Yao K Zeng L He Q Effect of probiotics on glucose and lipid metabolism in type 2 diabetes mellitus: a meta-analysis of 12 randomized controlled trials Med Sci Monit 2017 23 3044 3053 28638006
294. Wang C Zhang C Li S Effects of probiotic supplementation on dyslipidemia in type 2 diabetes mellitus: a meta-analysis of randomized controlled trials Foods 2020 9 1540 33114518
295. Kasińska MA Drzewoski J Effectiveness of probiotics in type 2 diabetes: a meta-analysis Pol Arch Med Wewn 2015 125 803 813 26431318
296. Palacios T Vitetta L Coulson S Targeting the intestinal microbiota to prevent type 2 diabetes and enhance the effect of metformin on glycaemia: a randomised controlled pilot study Nutrients 2020 12 2041 32660025
297. Zheng M Zhang R Tian X Assessing the risk of probiotic dietary supplements in the context of antibiotic resistance Front Microbiol 2017 8 908 28579981
298. Wong A Ngu DYS Dan LA Detection of antibiotic resistance in probiotics of dietary supplements Nutr J 2015 14 95 26370532
299. BgVV – ehemals: Bundesinstitut für gesundheitlichen Verbraucherschutz und Veterinärmedizin Abschlussbericht der Arbeitsgruppe „Probiotische Mikroorganismenkulturen in Lebensmitteln“ am BgVV. https://mobil.bfr.bund.de/cm/343/probiot.pdf. Zugegriffen: 13. Juli 2021
300. de Vrese M Mikrobiologie, Wirkung und Sicherheit von Probiotika Monatsschr Kinderheilkd 2008 156 1063 1069
301. Vrieze A van Nood E Holleman F Transfer of intestinal microbiota from lean donors increases insulin sensitivity in individuals with metabolic syndrome Gastroenterology 2012 143 913 916.e7 22728514
302. Simon MC Strassburger K Nowotny B Intake of lactobacillus reuteri improves incretin and insulin secretion in glucose-tolerant humans: a proof of concept Diabetes Care 2015 38 1827 1834 26084343
303. Tilg H Moschen AR Microbiota and diabetes: an evolving relationship Gut 2014 63 1513 1521 24833634
304. Kjems LL Holst JJ Vølund A The influence of GLP-1 on glucose-stimulated insulin secretion: effects on beta-cell sensitivity in type 2 and nondiabetic subjects Diabetes 2003 52 380 386 12540611
305. Karlsson FH Tremaroli V Nookaew I Gut metagenome in European women with normal, impaired and diabetic glucose control Nature 2013 498 99 103 23719380
306. Qin J Li Y Cai Z A metagenome-wide association study of gut microbiota in type 2 diabetes Nature 2012 490 55 60 23023125
307. Larsen N Vogensen FK van den Berg FWJ Gut microbiota in human adults with type 2 diabetes differs from non-diabetic adults PLoS ONE 2010 5 e9085 20140211
308. Wu H Esteve E Tremaroli V Metformin alters the gut microbiome of individuals with treatment-naive type 2 diabetes, contributing to the therapeutic effects of the drug Nat Med 2017 23 850 858 28530702
309. Forslund K Hildebrand F Nielsen T Corrigendum: Disentangling type 2 diabetes and metformin treatment signatures in the human gut microbiota Nature 2017 545 116 28470190
310. Forslund K Hildebrand F Nielsen T Disentangling type 2 diabetes and metformin treatment signatures in the human gut microbiota Nature 2015 528 262 266 26633628
311. Caesar R Pharmacologic and nonpharmacologic therapies for the gut microbiota in type 2 diabetes Can J Diabetes 2019 43 224 231 30929665
312. Evert AB Boucher JL Cypress M Nutrition therapy recommendations for the management of adults with diabetes Diabetes Care 2014 37 Suppl. 01 S120 S143 24357208
313. Sievenpiper JL Chan CB Dworatzek PD Nutrition therapy Can J Diabetes 2018 42 Suppl. 01 S64 S79 29650114
314. Sievenpiper JL de Souza RJ Mirrahimi A Effect of fructose on body weight in controlled feeding trials: a systematic review and meta-analysis Ann Intern Med 2012 156 291 304 22351714
315. Ha V Sievenpiper JL de Souza RJ Effect of fructose on blood pressure: a systematic review and meta-analysis of controlled feeding trials Hypertension 2012 59 787 795 22331380
316. Chiavaroli L de Souza RJ Ha V Effect of fructose on established lipid targets: a systematic review and meta-analysis of controlled feeding trials J Am Heart Assoc 2015 4 e001700 26358358
317. Wang X Ouyang Y Liu J Fruit and vegetable consumption and mortality from all causes, cardiovascular disease, and cancer: systematic review and dose-response meta-analysis of prospective cohort studies BMJ 2014 349 g4490 25073782
318. Chiu S Sievenpiper JL de Souza RJ Effect of fructose on markers of non-alcoholic fatty liver disease (NAFLD): a systematic review and meta-analysis of controlled feeding trials Eur J Clin Nutr 2014 68 416 423 24569542
319. Wang DD Sievenpiper JL de Souza RJ The effects of fructose intake on serum uric acid vary among controlled dietary trials J Nutr 2012 142 916 923 22457397
320. Cozma AI Sievenpiper JL de Souza RJ Effect of fructose on glycemic control in diabetes: a systematic review and meta-analysis of controlled feeding trials Diabetes Care 2012 35 1611 1620 22723585
321. Sievenpiper JL Chiavaroli L de Souza RJ ‘Catalytic’ doses of fructose may benefit glycaemic control without harming cardiometabolic risk factors: a small meta-analysis of randomised controlled feeding trials Br J Nutr 2012 108 418 423 22354959
322. Sievenpiper JL Carleton AJ Chatha S Heterogeneous effects of fructose on blood lipids in individuals with type 2 diabetes: systematic review and meta-analysis of experimental trials in humans Diabetes Care 2009 32 1930 1937 19592634
323. Chung M Ma J Patel K Fructose, high-fructose corn syrup, sucrose, and nonalcoholic fatty liver disease or indexes of liver health: a systematic review and meta-analysis Am J Clin Nutr 2014 100 833 849 25099546
324. Goran MI Ulijaszek SJ Ventura EE High fructose corn syrup and diabetes prevalence: a global perspective Glob Public Health 2013 8 55 64 23181629
325. Tsilas CS de Souza RJ Mejia SB Relation of total sugars, fructose and sucrose with incident type 2 diabetes: a systematic review and meta-analysis of prospective cohort studies CMAJ 2017 189 E711 E720 28536126
326. David Wang D Sievenpiper JL de Souza RJ Effect of fructose on postprandial triglycerides: a systematic review and meta-analysis of controlled feeding trials Atherosclerosis 2014 232 125 133 24401226
327. Zhang YH An T Zhang RC Very high fructose intake increases serum LDL-cholesterol and total cholesterol: a meta-analysis of controlled feeding trials J Nutr 2013 143 1391 1398 23825185
328. Schwingshackl L Neuenschwander M Hoffmann G Dietary sugars and cardiometabolic risk factors: a network meta-analysis on isocaloric substitution interventions Am J Clin Nutr 2020 111 187 196 31711109
329. Weber KS Simon MC Strassburger K Habitual fructose intake relates to insulin sensitivity and fatty liver index in recent-onset type 2 diabetes patients and individuals without diabetes Nutrients 2018 10 774 29914103
330. ter Horst KW Schene MR Holman R Effect of fructose consumption on insulin sensitivity in nondiabetic subjects: a systematic review and meta-analysis of diet-intervention trials Am J Clin Nutr 2016 104 1562 1576 27935520
331. Kulzer B Albus C Herpertz S Psychosoziales und Diabetes Diabetologe 2019 15 452 469
332. Ahmed AT Karter AJ Warton EM The relationship between alcohol consumption and glycemic control among patients with diabetes: the Kaiser Permanente Northern California Diabetes Registry J Gen Intern Med 2008 23 275 282 18183468
333. Bantle AE Thomas W Bantle JP Metabolic effects of alcohol in the form of wine in persons with type 2 diabetes mellitus Metabolism 2008 57 241 245 18191055
334. Avogaro A Beltramello P Gnudi L Alcohol intake impairs glucose counterregulation during acute insulin-induced hypoglycemia in IDDM patients. Evidence for a critical role of free fatty acids Diabetes 1993 42 1626 1634 8405705
335. Turner BC Jenkins E Kerr D The effect of evening alcohol consumption on next-morning glucose control in type 1 diabetes Diabetes Care 2001 24 1888 1893 11679452
336. Richardson T Weiss M Thomas P Day after the night before: influence of evening alcohol on risk of hypoglycemia in patients with type 1 diabetes Diabetes Care 2005 28 1801 1802 15983341
337. Pedersen-Bjergaard U Reubsaet JLE Nielsen SL Psychoactive drugs, alcohol, and severe hypoglycemia in insulin-treated diabetes: analysis of 141 cases Am J Med 2005 118 307 310 15745730
338. Frier B Fisher M Moderators, monitoring and management of hypoglycaemia 2007 Chichester John Wiley & Sons
339. Ahmed AT Karter AJ Liu J Alcohol consumption is inversely associated with adherence to diabetes self-care behaviours Diabet Med 2006 23 795 802 16842486
340. Nahas R Goguen J Natural health products Can J Diabetes 2013 37 Suppl. 01 S97 S99 24070974
341. Hartweg J Perera R Montori V Omega-3 polyunsaturated fatty acids (PUFA) for type 2 diabetes mellitus Cochrane Database Syst Rev 2008 2008 CD003205 18254017
342. Hartweg J Farmer AJ Holman RR Potenzial impact of omega-3 treatment on cardiovascular disease in type 2 diabetes Curr Opin Lipidol 2009 20 30 38 19133409
343. O’Mahoney LL Matu J Price OJ Omega-3 polyunsaturated fatty acids favourably modulate cardiometabolic biomarkers in type 2 diabetes: a meta-analysis and meta-regression of randomized controlled trials Cardiovasc Diabetol 2018 17 98 29981570
344. Mirhosseini N Vatanparast H Mazidi M The effect of improved serum 25-hydroxyvitamin D status on glycemic control in diabetic patients: a meta-analysis J Clin Endocrinol Metab 2017 102 3097 3110 28957454
345. Li X Liu Y Zheng Y The effect of vitamin D supplementation on glycemic control in type 2 diabetes patients: a systematic review and meta-analysis Nutrients 2018 10 375 29562681
346. Jafari T Fallah AA Barani A Effects of vitamin D on serum lipid profile in patients with type 2 diabetes: A meta-analysis of randomized controlled trials Clin Nutr 2016 35 1259 1268 27020528
347. Mousa A Naderpoor N Teede H Vitamin D supplementation for improvement of chronic low-grade inflammation in patients with type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials Nutr Rev 2018 76 380 394 29490085
348. Lee KJ Lee YJ Effects of vitamin D on blood pressure in patients with type 2 diabetes mellitus Int J Clin Pharmacol Ther 2016 54 233 242 26902507
349. Yu Y Tian L Xiao Y Effect of vitamin D supplementation on some inflammatory Biomarkers in type 2 diabetes mellitus subjects: a systematic review and meta-analysis of randomized controlled trials Ann Nutr Metab 2018 73 62 73 29945132
350. Verma H Garg R Effect of magnesium supplementation on type 2 diabetes associated cardiovascular risk factors: a systematic review and meta-analysis J Hum Nutr Diet 2017 30 621 633 28150351
351. Asbaghi O Moradi S Kashkooli S The effects of oral magnesium supplementation on glycaemic control in patients with type 2 diabetes: a systematic review and dose-response meta-analysis of controlled clinical trials Br J Nutr 2022 128 12 2363 2372 35045911
352. Asbaghi O Hosseini R Boozari B The effects of magnesium supplementation on blood pressure and obesity measure among type 2 diabetes patient: a systematic review and meta-analysis of randomized controlled trials Biol Trace Elem Res 2021 199 413 424 32385715
353. Vincent JB Elucidating a biological role for chromium at a molecular level Acc Chem Res 2000 33 503 510 10913239
354. Asbaghi O Fatemeh N Mahnaz RK Effects of chromium supplementation on glycemic control in patients with type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials Pharmacol Res 2020 161 105098 32730903
355. Yin RV Phung OJ Effect of chromium supplementation on glycated hemoglobin and fasting plasma glucose in patients with diabetes mellitus Nutr J 2015 14 14 25971249
356. Suksomboon N Poolsup N Yuwanakorn A Systematic review and meta-analysis of the efficacy and safety of chromium supplementation in diabetes J Clin Pharm Ther 2014 39 292 306 24635480
357. Chimienti F Zinc, pancreatic islet cell function and diabetes: new insights into an old story Nutr Res Rev 2013 26 1 11 23286442
358. de Carvalho GB (2017) Zinc’s role in the glycemic control of patients with type 2 diabetes: a systematic review. Biometals: 1–12
359. Fernández-Cao JC Warthon-Medina M Hall Moran V Dietary zinc intake and whole blood zinc concentration in subjects with type 2 diabetes versus healthy subjects: A systematic review, meta-analysis and meta-regression J Trace Elem Med Biol 2018 49 241 251 29452774
360. Wang X Wu W Zheng W Zinc supplementation improves glycemic control for diabetes prevention and management: a systematic review and meta-analysis of randomized controlled trials Am J Clin Nutr 2019 110 76 90 31161192
361. Asbaghi O Sadeghian M Fouladvand F Effects of zinc supplementation on lipid profile in patients with type 2 diabetes mellitus: A systematic review and meta-analysis of randomized controlled trials Nutr Metab Cardiovasc Dis 2020 30 1260 1271 32451277
362. Rahimi R Nikfar S Larijani B A review on the role of antioxidants in the management of diabetes and its complications Biomed Pharmacother 2005 59 365 373 16081237
363. Ashor AW Werner AD Lara J Effects of vitamin C supplementation on glycaemic control: a systematic review and meta-analysis of randomised controlled trials Eur J Clin Nutr 2017 71 1371 1380 28294172
364. Xu R Zhang S Tao A Influence of vitamin E supplementation on glycaemic control: a meta-analysis of randomised controlled trials PLoS ONE 2014 9 e95008 24740143
365. Khodaeian M Tabatabaei-Malazy O Qorbani M Effect of vitamins C and E on insulin resistance in diabetes: a meta-analysis study Eur J Clin Invest 2015 45 1161 1174 26313310
366. Montero D Walther G Stehouwer CDA Effect of antioxidant vitamin supplementation on endothelial function in type 2 diabetes mellitus: a systematic review and meta-analysis of randomized controlled trials Obes Rev 2014 15 107 116 24118784
367. Tabatabaei-Malazy O Ardeshirlarijani E Namazi N Dietary antioxidative supplements and diabetic retinopathy; a systematic review J Diabetes Metab Disord 2019 18 705 716 31890694
368. Jeyaraman MM Al-Yousif NSH Singh Mann A Resveratrol for adults with type 2 diabetes mellitus Cochrane Database Syst Rev 2020 1 CD011919 31978258
369. Palma-Duran SA Vlassopoulos A Lean M Nutritional intervention and impact of polyphenol on glycohemoglobin (HbA1c) in non-diabetic and type 2 diabetic subjects: Systematic review and meta-analysis Crit Rev Food Sci Nutr 2017 57 975 986 25746842
370. Fogacci F Tocci G Presta V Effect of resveratrol on blood pressure: A systematic review and meta-analysis of randomized, controlled, clinical trials Crit Rev Food Sci Nutr 2019 59 1605 1618 29359958
371. Drzikova B (2005) Haferprodukte mit modifiziertem Gehalt an β‑Glucanen und resistenter Stärke und ihre Effekte auf den Gastrointestinaltrakt unter In-vitro- und In-vivo-Bedingungen. http://opus.kobv.de/ubp/volltexte/205/592/
372. He L Zhao J Huang Y The difference between oats and beta-glucan extract intake in the management of HbA1c, fasting glucose and insulin sensitivity: a meta-analysis of randomized controlled trials Food Funct 2016 7 1413 1428 26840185
373. Abbasi NN Purslow PP Tosh SM Oat β-glucan depresses SGLT1- and GLUT2-mediated glucose transport in intestinal epithelial cells (IEC-6) Nutr Res 2016 36 541 552 27188900
374. Wang F Yu G Zhang Y Dipeptidyl peptidase IV inhibitory peptides derived from oat (Avena sativa L.), buckwheat (Fagopyrum esculentum), and highland barley (Hordeum vulgare trifurcatum (L.) Trofim) proteins J Agric Food Chem 2015 63 9543 9549 26468909
375. Liu M Zhang Y Zhang H The anti-diabetic activity of oat β-d-glucan in streptozotocin-nicotinamide induced diabetic mice Int J Biol Macromol 2016 91 1170 1176 27365119
376. Lammert A Kratzsch J Selhorst J Clinical benefit of a short term dietary oatmeal intervention in patients with type 2 diabetes and severe insulin resistance: a pilot study Exp Clin Endocrinol Diabetes 2008 116 132 134 18095234
377. Delgado G Kleber ME Krämer BK Dietary intervention with oatmeal in patients with uncontrolled type 2 diabetes mellitus—A crossover study Exp Clin Endocrinol Diabetes 2019 127 623 629 30157531
378. Delgado GE Krämer BK Scharnagl H Bile acids in patients with uncontrolled type 2 diabetes mellitus—The effect of two days of oatmeal treatment Exp Clin Endocrinol Diabetes 2020 128 624 630 31896155
379. Behall KM Scholfield DJ Hallfrisch J Comparison of hormone and glucose responses of overweight women to barley and oats J Am Coll Nutr 2005 24 182 188 15930484
380. Braaten JT Scott FW Wood PJ High beta-glucan oat bran and oat gum reduce postprandial blood glucose and insulin in subjects with and without type 2 diabetes Diabet Med 1994 11 312 318 8033532
381. Pick ME Hawrysh ZJ Gee MI Oat bran concentrate bread products improve long-term control of diabetes: a pilot study J Am Diet Assoc 1996 96 1254 1261 8948386
382. Tapola N Karvonen H Niskanen L Glycemic responses of oat bran products in type 2 diabetic patients Nutr Metab Cardiovasc Dis 2005 15 255 261 16054549
383. Tappy L Gügolz E Würsch P Effects of breakfast cereals containing various amounts of beta-glucan fibers on plasma glucose and insulin responses in NIDDM subjects Diabetes Care 1996 19 831 834 8842600
384. Wood PJ Beer MU Butler G Evaluation of role of concentration and molecular weight of oat beta-glucan in determining effect of viscosity on plasma glucose and insulin following an oral glucose load Br J Nutr 2000 84 19 23 10961156
385. Anonym Scientific Opinion on the substantiation of health claims related to beta glucans and maintenance or achievement of normal blood glucose concentrations (ID 756, 802, 2935) pursuant to Article 13(1) of Regulation (EC) No 1924/2006 EFS2 2010 8 1482
386. Amtsblatt der Europäischen Union 2011 L 136/1 vom 25.5.2012. https://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2012:136:0001:0040:DE:PDF. Zugegriffen: 4. Juli 2021
387. Zurbau A Noronha JC Khan TA The effect of oat β-glucan on postprandial blood glucose and insulin responses: a systematic review and meta-analysis Eur J Clin Nutr 2021 75 1540 1554 33608654
388. Battilana P Ornstein K Minehira K Mechanisms of action of beta-glucan in postprandial glucose metabolism in healthy men Eur J Clin Nutr 2001 55 327 333 11378805
389. Jenkins AL Jenkins DJA Zdravkovic U Depression of the glycemic index by high levels of beta-glucan fiber in two functional foods tested in type 2 diabetes Eur J Clin Nutr 2002 56 622 628 12080401
390. Churuangsuk C Hall J Reynolds A Diets for weight management in adults with type 2 diabetes: an umbrella review of published meta-analyses and systematic review of trials of diets for diabetes remission Diabetologia 2022 65 14 36 34796367
391. Rosenfeld RM Kelly JH Agarwal M Dietary interventions to treat type 2 diabetes in adults with a goal of remission: an expert consensus statement from the American College of Lifestyle Medicine Am J Lifestyle Med 2022 18 342 362
392. Asbaghi O Moradi S Kashkooli S The effects of oral magnesium supplementation on glycaemic control in patients with type 2 diabetes: a systematic review and dose-response meta-analysis of controlled clinical trials Br J Nutr 2022 20 1 10
Weiterführende Literatur
393. Anderson JW Kendall CWC Jenkins DJA Importance of weight management in type 2 diabetes: review with meta-analysis of clinical studies J Am Coll Nutr 2003 22 331 339 14559925
394. Porrata-Maury C Hernández-Triana M Ruiz-Álvarez V Ma-Pi 2 macrobiotic diet and type 2 diabetes mellitus: pooled analysis of short-term intervention studies Diabetes Metab Res Rev 2014 30 Suppl. 01 55 66 24532293
395. Jung CH Choi KM Impact of high-carbohydrate diet on metabolic parameters in patients with type 2 diabetes Nutrients 2017 9 322 28338608
396. Rubenstein LZ Harker JO Salvà A Screening for undernutrition in geriatric practice: developing the short-form mini-nutritional assessment (MNA-SF) J Gerontol A Biol Sci Med Sci 2001 56 M366 M372 11382797
|
PMC010xxxxxx/PMC10169102.txt |
==== Front
Pharm Chem J
Pharm Chem J
Pharmaceutical Chemistry Journal
0091-150X
1573-9031
Springer US New York
2883
10.1007/s11094-023-02883-4
Article
Strategy for the Choice of Disinfectants in Practical Medicine and Production
Bogdanova O. Yu. bogdiolg@yandex.ru
Chemykh T. F.
St. Petersburg State Chemical Pharmaceutical University, 14/A Prof. Popova St., St. Petersburg, 197022 Russia
9 5 2023
2023
57 2 314317
26 4 2022
© Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The problem of the strategy for choosing disinfectants in practical medicine is considered. The pandemic of the new coronavirus infection posed new problems for disinfectology. The expanded spectrum of disinfectants and antiseptics offered by the chemical industry in recent years requires justification for the choice in favor of any product. The goals and types of disinfection considered from current positions and the main groups of disinfectants used in Russia and their properties and spectra of activity are presented.
Keywords
disinfection
disinfectants
selection strategy
issue-copyright-statement© The Editorial Board of the Khimico-Farmatsevticheskii Zhurnal 2023
==== Body
pmcAseptic, antiseptic, disinfection, and sterilization methods have expanded considerably in the last decades. On one hand, this has significantly increased the quality of the medical aid provided and, on the other, led to the spread of resistant strains of agents capable of causing infections associated with the provision of medical aid. Chemical disinfectants (CDs) of various compositions and concentrations are widely used to clean and disinfect surfaces, instruments, and tools; to decontaminate water and air, and to clean hands of personnel in medical institutions and the pharmaceutical and biotechnological industries. However, the choice of agent is difficult because the spectrum of agents offered by industry is extremely broad. Besides, a long list of problems with the application of disinfectants and antiseptics that require attention and resolution has developed over time.
The aim of the work was to systematize information on disinfection and disinfectants considering current trends for an informed choice of agents in practice.
Experimental Part
The search for information used keywords in electronic databases such as PubMed, e-library.ru, cyberleninka.ru, and researchgate.net and published sources during 2015 – 2022 without limiting the languages.
The selection criteria were epidemiological studies, clinical trials, cohorts, crossover studies, before-after studies, etc. The main criterion was problems in current practice of application of disinfectants and antiseptics in medical institutions.
Results and Discussion
Disinfection in medical institutions and pharmaceutical and biotechnological industries is intended mainly to ensure the infectious and epidemiological safety of patients, to improve the quality of provided medical aid and services, and to prevent spread of infections [1, 2]. This problem is becoming more and more critical considering the global nature of the spread of infections associated with provision of medical aid in medical institutions [3] and the increasing number of their vectors resistant to antiseptics [4]. CDs must be correctly chosen to assure the appropriate quality of the disinfection so that their use will meet regulatory requirements applicable to microbiological cleanliness of rooms and objects of medical institutions and pharmaceutical production [5].
The Russian market for CDs includes 1200 products of various CD groups. Domestic and foreign products are added to this lest every year [6, 7]. Recently, the following problems with disinfectants have become obvious:
( not all registered CDs met the requirements applicable to them with respect to composition, problems, and instructions for use [8];
( the instructions and effective concentrations are unclearly defined for several CDs [9];
( the need has arisen to determine clear criteria for defining a microorganism as resistant to CDs by establishing epidemiological threshold values and standardizing protocols for tests of microorganism resistance to CDs [10, 11];
( manufacturing conditions and the raw-material availability of manufacturers change;
( attention to the ecological safety of CD manufacturing has increased [12, 13];
– several CDs used in several institutions aged unreliably and had confirmed toxicity and corrosion hazards for surfaces and materials [14];
– morphological, biological, and genetic features of bacterial vectors of nosocomial infections especially resistant to disinfectants are insufficiently studied [11, 15];
– microorganisms acquire resistance to CDs, like to antibiotics and chemotherapy drugs, resulting in resistant strains of industrial microbiota [16] and, as a result, affect the quality of provided medical services and infection hazards [17];
– new infection issues present new problems for the development and use of new CDs. For example, the virus vector of the new coronavirus infection COVID-19 can persist on surfaces although the exact role of contaminated industrial and household items in the spread of the infection is still not established and the effectiveness of disinfection measures outside treatment and prevention institutes (TPIs) has not been confirmed [18];
– broad use of CDs and antiseptics because of outbreaks of the new coronavirus infection COVID-19 led to the use of antiseptics unsuitable for the community and incorrect handling and use of biocides [10, 19];
– incorrect use of CDs in TPIs and the unjustified expense of CDs for preparing working solutions exacerbate the problem of microorganism resistance to CDs [19, 20];
– the insufficient potential of single agents for effective inactivation of viruses [21] and bacteria and fungi resistant to CDs.
CDs offered on the current Russian market are 97% domestic. This share increased considerably from five years ago when the fraction of Russian CDs was 80%. The whole volume of CDs consists of 74% disinfectants, the fraction of which also grew by 3%. Topical antiseptics, including moist disinfecting tissues, amount to 22%; soaps, 4%.
Registered CDs in the RF include 50% with one active ingredient; 36%, two active ingredients; 12%, three active ingredients; and 2%, CDs with four active ingredients. Some of the most popular CDs are quaternary ammonium compounds (QACs), which are very important cationic surfactants (CSAs). QACs are guanidine derivatives and trialkyl amines. These CDs are not volatile, have no sharp smell, and are very water-soluble. Several of them have washing properties, are stable, and do not damage treated objects. They possess selective bactericidal, fungicidal, and virucidal activity without exhibiting sporicidal activity, which necessitates the formulation of multicomponent CDs.
Guanidine derivatives are also typical CSAs according to their physicochemical properties, i.e., they possess bactericidal and virucidal activity. Polymeric guanidine derivatives can form films on cleaned surfaces, thereby prolonging their residual antimicrobial activity [2, 22].
Alkylamines are fatty acid derivatives. Amines are divided into primary, secondary, and tertiary depending on the number of H atoms replaced by alkyl radicals. CDs containing a tertiary amine or a diamine exhibit high bactericidal, fungicidal, and virucidal activity.
CDs containing alcohols that are used for disinfection include ethyl, isopropyl, or propyl alcohols. Agents based on alcohols are bactericidal, destroying vegetative microorganisms. They can destroy spore-forming microorganisms at concentrations >70%. Isopropyl alcohol at concentrations ≥60% kills tuberculosis mycobacteria [23]. The virucidal activity of these CDs is multifaceted. Lipophilic viruses are sensitive to the above three alcohols. Hydrophilic viruses (e.g., hepatitis A virus, Coxsackie enterovirus) are inactivated only by ethanol [23]. Disinfectants containing alcohols are costly. Therefore, they are used to clean small difficultly accessible equipment parts and surfaces and as additives to CDs based on QACs.
CDs now are most often combined, including QACs, 4.5 – 15%; guanidine, 1 – 10%; and amines, 4 – 10%. Alcohols are also added. In this instance, cleaning with concentrations from 0.05 to 0.2% for 30 – 60 min is used to wipe surfaces and achieve an antibacterial effect. Concentrations from 0.1 to 3% are used for 60 min to produce a virucidal effect. Medical instruments are cleaned and an overall microbiocidal effect is achieved if the concentration is increased to 4% and the treatment time, from 20 to 60 min. The concentration(time regime for sterilization should be increased to 25% and 60 – 120 min. Endoscopic equipment and medical tubing should be cleaned with CD solutions of concentrations 2 – 15% for at least 15 – 30 min.
Disinfection is an obligatory procedure in medical and pharmaceutical rooms. According to adopted sanitary rules and standards (SanPiN 1.2.3685(21, SanPiN 3.3686(21, SP 3.5.1378(03), objects to be disinfected include equipment, tools, fixtures, room surfaces, air, water, sanitary personnel clothing, whites, and bandages.
The purpose of disinfection is the prevention of generating and spreading infections associated with provision of medical aid (IAMA) [an outdated name for nosocomial infections (NIs)].
The task of disinfection is to disrupt transmission pathways of infectious disease vectors from infection sources to susceptible people (from an infectious patient to the environment).
The objects of disinfections can be divided into the following types:
( critical, come into contact with sterile tissue of an organism, e.g., cardiac catheters, implants, surgical instruments. Sterility requirements are applicable to critical objects;
– semicritical, come into direct contact with patient mucous membranes or damaged skin, e.g., inhalers, endoscopes, bronchoscopes;
– noncritical, contact intact skin fragments, e.g., bedpans, thermometers, stethoscopes.
Disinfection levels:
– High-level disinfection (HLD), aimed at destroying vegetative bacteria, tuberculosis mycobacteria, fungi, and lipid and nonlipid viruses. HLD is ineffective against many bacterial spore forms. HLD is used for semicritical objects.
Mid-level disinfection (MLD), aimed at destroying vegetative bacteria, most fungi, tuberculosis mycobacteria, most viruses. It is ineffective against bacterial spore forms. MLD is used for semicritical and noncritical objects with smooth solid surfaces;
Low-level disinfection (LLD), aimed at destroying vegetative bacteria, several fungi, viruses. It is ineffective against highly resistant bacteria (tuberculosis mycobacteria) and bacterial spore forms. LLD is used for noncritical objects.
Many factors determine the choice of CD. Many criteria should be considered [24].
The staging and sequencing of a comprehensive cleaning and disinfection, from the cleanest zones to the most contaminated, from above to below, concentrating grime and contaminants near the floor, from where they can be removed by CDs of the highest concentrations are important components.
The choice of CD for cleaning hands tends to favor CDs packaged into dispensers that prevent contamination of the tank and the solution itself.
Surfaces are disinfected during routine and general cleaning using CD solutions by wiping with cloths wetted with a CD solution. CDs with a detergent effect should be used for these purposes. If the need for emergency cleaning of small or difficultly accessible surfaces arises during the working day, finished CDs with short service lives, e.g., those based on alcohols, can be used with squirt bottles or wiping with CD solutions on cloths or ready-to-use disinfecting pads. The use of CDs with detergent properties allows decontamination of an object to be combined with its washing. Therefore, disinfecting agents with detergent action are used for routine and general cleaning. Increased attention should be paid to monitoring cleaning in TPIs [25, 26].
Labware is disinfected with agents based on aldehydes, CSAs, H2O2, and chlorinated agents. Items and labware are disinfected by immersion in CD solutions. Items that can be disassembled are disinfected as their parts. Channels and cavities of items are filled with disinfecting solution.
Batches of CDs newly arriving at a warehouse must be monitored for the content of active ingredient. Disinfecting solutions are prepared in specially separated rooms or hoods following the instructions for dilution. The name, concentration, and date of preparation should be indicated on a container with disinfecting solution.
Cleaning supplies should be labeled separately for clean and contaminated zones. Transfer from one zone to another is not allowed.
The choice of CD is determined by the effectiveness against microorganisms that must be eliminated, i.e., the microorganisms most often encountered in the industrial environment for which industrial microbiological monitoring is required [27, 28].
CDs must be tested on all types of surfaces susceptible to contamination in an actual production for a reliable choice of CD for a particular object because the CD can behave differently on different surfaces [2, 29]. Studies on surfaces should mimic the actual cleaning process that will be used daily by personnel.
The choice of CD must also consider potential toxic effects on personnel performing the cleaning and on the production equipment, including the formation of residues that may become the cause for endotoxin formation [30].
Expanded morphological and biological studies of the nature of bacterial resistance to a CD to model the process by which bacteria require resistance to it are needed to improve the choice of CD. Considering the expanding resistance spectrum of bacteria to CDs in medical and pharmaceutical institutions, the principle of CD rotation must be used.
Thus, the choice of CD is determined by the following factors:The contaminating microbial vector of the production zone of medical and pharmaceutical institutions and the level and degree of resistance;
Recommended concentration(time regimes for action of working solutions, purpose of the disinfection;
Surface nature and material, its corrosion resistance, configuration of object to be cleaned;
CD toxicity criteria;
CD stability, shelf life for 100% effectiveness;
Convenience of packaging for use;
Requirements of healthcare organizations, Russian and departmental regulatory documents, approval of CD for circulation.
Translated from Khimiko-Farmatsevticheskii Zhurnal, Vol. 57, No. 2, pp. 61 – 64, February, 2023.
==== Refs
References
1. Morozova NS Bolovchak BS Korobkova IV Aktual. Infektol. 2018 6 5 268 269
2. Federal Clinical Recommendations for Choice of Chemical Disinfection and Sterilization for Use in Medical Organizations [in Russian], Moscow (2015).
3. Z. M. Tav, FORCIPE, 609 – 610 (2019).
4. O. A. Orlova, A. V. Tutel’yan, and V. G. Akimkin, in: Proceedings of Scientific-Practical Conferences of the Vth Russian Congress on Laboratory Medicine (RKLM 2019), Moscow, Sept. 11 – 13, 2019, IPO U Nikitskikh Vorot, Moscow (2019), pp. 155 – 156.
5. Eremin GB Mozzhukhina NA Bormashov AV Zdorov’ye Naseleniya Sreda Obitaniya 2012 5 14 17
6. Analysis of the Disinfectant Market in Russia in 2017 – 2021, Prediction for 2022 – 2026; https://businesstat.ru/images/demo/disinfectantsrussiademobusinesstat.pdf.
7. M. A. Chernaya and A. A. Kushnirenko, Alleya Nauki, No. 9 (25), Vol. 1, 117 – 122 (2018).
8. V. I. Sharafutdinova, Vestn. Poznavatel. Zh.; https://vestnik.icdc.ru/mkdts/.
9. Ivanov BL Rudakov AI Zinnatullin NKh Vestn. Kazan. Tekhnol. Univ. 2017 21 130 133
10. O. Roman, D. Dukh, M. Zimmerman, et al., Int. J. Environ. Res. Public Health, 19(22), (2022).
11. N. V. Saperkin, L. A. Alebashina, and D. V. Kvashnina, Sovrem. Probl. Nauki Obraz., 5, (2016).
12. M. V. Rymovskaya and V. I. Romanovskii, Tr. BGTU, Ser. 2: Khim. Tekhnol., Biotekhnol. Geoekol., No. 4 (186), 214 – 219 (2016).
13. S. V. Churyukina and V. D. Sokova, Molodoi Uch., No. 4 (346), 67 – 70 (2021).
14. Matatiele P Southon B Dabula B Sci. Rep. 2022 12 1 4231 10.1038/s41598-022-08117-z 35273314
15. Shkarin VV Kovalishena OV Blagonravova AS Med. Al’manakh 2012 3 129 133
16. D. D. Nesterova, A. N. Neshataev, N. V. Luk?yanenko, et al., Perm. Med. Zh., 5, 42 – 50 (2018).
17. V. V. Kanishchev and N. I. Eremeeva, Remedium Privolzh., No. 6 (156), 26 – 30 (2017).
18. Chin AWH Chu JTS Perera MRA Lancet Microbe 2020 5247 20 30003 3
19. Serov AA Eremeeva NI Dezinfekts. Delo 2022 4 122 55 59
20. L. S. Fedorova and A. V. Il?yakova, Dezinfekts. Delo, 4(122), 50 – 54 (2022).
21. Nosik NN Nosik DN Chizhov AI Vopr. Virusol. 2017 1 41 45 10.18821/0507-4088-2017-62-1-41-45
22. K. A. Shilova and A. V. Ivanchuk, Vestn. St.-Petersburg Univ. Med., 17(1), 36 – 42 (2022).
23. Shestopalov NV Fedorova LS Skopin AYu Gigiena Sanitariya 2019 10 1031 1036
24. I. A. Kolyadenko, U. S. Protasevich, and I. M. Nazarov, Molodoi Uch., 40(435), 20 – 27 (2022).
25. Cleaning and Disinfection of Environmental Surfaces in the Context of COVID-19: Temporary Recommendations, May 15, 2020; https://apps.who.int/iris/bitstream/handle/10665/332096/WHO-2019-nCoV-Disinfection-2020.1-rus.pdf.
26. Kolesnikova EP Remedium Privolzh. 2016 6 146 36
27. Prushinskii AP MediAl Zh No. 2018 2 22 16 20
28. Method for Determination of Sensitive Bacteria to Disinfectants with Monitoring of Resistance to Antimicrobial Drugs in Medical Organizations. Federal Clinical Recommendations [in Russian], Moscow (2015).
29. V. V. Proshutinskii, A. G. Yarenskikh, S. A. Terent?ev, et al., Inzh. Vestn. Dona, No. 3 (63), 38 – 38 (2020).
30. Rossington K Cleanroom Technology No. 2019 6 80 33 38
|
PMC010xxxxxx/PMC10169110.txt |
==== Front
Pharm Chem J
Pharm Chem J
Pharmaceutical Chemistry Journal
0091-150X
1573-9031
Springer US New York
2867
10.1007/s11094-023-02867-4
Article
Synthesis, Characterization, and Antibacterial Activity of New Isatin Derivatives
Nain Sumitra nainsumitra@gmail.com
nsumitra@banasthali.in
Mathur Garima
Anthwal Tulika
Sharma Swapnil
Paliwal Sarvesh
Department of Pharmacy, Banasthali Vidyapith, Banasthali, Rajasthan 304022 India
9 5 2023
2023
57 2 196203
15 5 2021
© Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
1H-indol-2,3-dione (isatin) class of biologically active compounds have analgesic, anti-microbial, anti-inflammatory, anti-tubercular, anti-proliferative properties, and is also useful for the treatment of SARS-CoV. Schiff bases containing isatin moiety are known to have broad spectrum of biological activities like anti-viral, anti-tubercular, anti-fungal, and anti-bacterial. In this work, several Schiff base derivatives have been synthesized using two methods (synthetic and microwave) by reacting isatin with o-phenylenediamine. The synthesized compounds were structurally characterized and their in-vivo antimicrobial activity was tested against Gram-negative and Gram-positive bacteria using the inhibition zone method. Several newly synthesized isatin derivatives were found effective as antimicrobial agents and showed good potency (compounds 3c, 3d, 6a, 6b, 6d). Compound 3c displayed higher antimicrobial activity than standard drug (Amoxicillin) against Staphylococcus aureus at higher concentration (16 μg/mL) and against Escherichia coli at lower concentration (1 μg/mL).
Keywords
isatin
o-phenylenediamine
chemotherapeutic
amoxicillin
antimicrobial activity
antibacterial
issue-copyright-statement© The Editorial Board of the Khimico-Farmatsevticheskii Zhurnal 2023
==== Body
pmcIntroduction
The discovery of antibiotics has been proved to be a blessing in the medical science field still limitation is resistance, there has emerged the need for new potent antibacterial agents. It is the need of time to improve the existing anti-microbials and develop new ones. In particular, isatin scaffold possesses analgesic, antimicrobial, anti-inflammatory, anti-tubercular, anti-proliferative, and many other properties. Isatin (1H-indole-2,3-dione) is a moiety of interest for medicinal chemists for the synthesis of heterocyclic compounds such as quinolines and indoles. It is an indole derivative that comprises two cyclic carbonyl groups (one five-membered and other six-membered, a good moiety for chemical conversion rings) and contains two carbonyl groups at 2nd and 3rd positions and nitrogen at 1st position (Fig. 1). It was discovered and synthesized by Laurent and Erdman (Fig. 2A) in 1841 by oxidizing indigo using nitric acid and chromic acids. Other scientists have discovered several different methods to synthesize isatins, including T. Sandmeyer (Fig. 2B) (the oldest and most widely used pathway to isatin which gives a good yield, although it is effective for small analog), Martinet (Fig. 2C), Gassman (Fig. 2D) and Stolle (Fig. 2E). Subsequently, G. S. Hiers and C. S. Marvel (Fig. 2F) modified Sandmeyer’s method in which aniline or substituted aniline was treated with hydrochloride, chloral hydrate, sulfate, or salts of hydroxylamine, which was made in a solution of sodium sulfate and then cyclization reaction proceeded in the presence of conc. H2SO4 [1–5].Fig. 1. Chemical structure of isatin.
Fig. 2. Methods of isatin synthesis by methods of (A) Erdman and Laurent, (B) T. Sandmeyer, (C) Martinet, (D) Gassman, (E) Stolle, and (F) G. S. Hiers and C. S. Marvel.
Isatin and endogenous indoles have been found in the body fluids (adrenaline hormone) and mammalian tissues. It has been identified in urine, blood, and tissue using GC-MS and, more recently by HPLC with ultraviolet detector [1, 2]. Plants of genuses such as Isatis tinctoria, Couroupita guianensisisatis, and Calanthe discolor [4, 6–8] also contain isatin. It is also secreted from the Bufo frog’s parotid gland [5, 9]. It is one of the constituents of coal tar [3]. It has been reported that isatin possesses diverse biological properties such as antioxidant [10, 11], anti-inflammatory [12, 13], antimicrobial (bacterial, viral, fungal) [14–16], anti-tuberculosis [17, 18], anticancer [19, 20], anticonvulsant [21, 22], anti-HIV [23, 24] and many more. Isatin is also effective in treating coronavirus [25–27] and hepatitis C [27]. In addition to biological activities, it has several industrial uses as corrosion inhibitors, dyes, and fluorescence sensors [1].
Subsequent studies suggested that N-alkylation and substitution in the 5th position of isatin with strong electron-donating groups/atoms like bromine, fluorine, or chlorine produce a series of more active compounds compared to the parent compound [28]. Isatin substitution at 3rd position with aromatic or substituted aromatic ring contributes to antimicrobial activity. The phenyl ring moieties, heterocyclic rings, and alphabetic system have been other possible substitutions at 3rd position [29].
Chaithanya, et al. (2019) designed twenty new isatin derivatives by reacting with 2-methyl-benzimidazole and benzimidazole; and intermediates were obtained by condensing chloroacetyl chloride and o-phenylenediamine, and then designed compounds were screened for antimicrobial activity and only a few showed potent antifungal anti-bacterial activity (Fig. 3) [30].Fig. 3. Chemical structure of isatin benzimidazole derivatives.
Alsalihi, et al. (2018) synthesized several Schiff base derivatives by condensing isatin with 3-amino acetophenone in presence of KOH and then tested synthesized compounds for antibacterial activity, and a few were found potent (Fig. 4) [31].Fig. 4. Chemical structure of Schiff base derivatives of isatin.
Almutairi, et al. (2017) synthesized indole isatin derivatives using GAA as a catalyst [32]. The synthesized compounds were checked for antibacterial activity, and some were found potent (Fig. 5).Fig. 5. Chemical structure of indole isatin derivatives.
Bogdano, et al. (2015) designed Schiff base derivatives by condensing isatin and o-Phenylenediamine in the presence of ethanol, and the synthesized compound was tested for antibacterial activity and a few compounds were found to be potent (Fig. 6) [33].Fig. 6. Structure of Schiff base derivatives of isatin with o-phenylenediamine.
Alsalihi, et al. (2015) designed some new Schiff base derivatives of isatin using the microwave radiation method. The synthesized compounds were tested for antibacterial activity against Gram-negative and Gram-positive bacteria (Fig. 7) and were found potent [34].Fig. 7. Structure of Schiff base derivatives of isatin.
Singh, et al. (2010) designed some new Schiff and Mannich base derivatives of isatin in the presence of secondary amine and formaldehyde. The synthesized compounds were tested for antimicrobial activity and some were found more potent than the standard drug (Fig. 8) [35].Fig. 8. Structure of Schiff and Mannich base derivatives of isatin.
Since isatin derivatives are known to affect CNS [36] and show several biological activities, in the present study we have designed and synthesized isatin derivatives of 3-(2-aminophenylimino)-5-substituted-indoline-2-ones 3 (a–d) and 6 (a–d) by reacting isatin with o-phenylenediamine. The synthesized compounds were characterization by spectral techniques and screened for antimicrobial activity against selected bacteria.
Experimental Part
Materials and Methods
All reagents and chemicals used were of the analytical grade. Melting points were determined using the melting point apparatus (Buchi). The infrared spectra were recorded as KBr pellets on Agilent Technologies FT-IR (Cary 660) spectrophotometer. The NMR spectra were recorded on Bruker NMR spectrometer (300 MHz) using DMSO-d6 as solvent. The mass spectra were obtained on Bruker MS-632 instrument equipped with an electrospray ionization (ESI) system, using nebulizing gas (nitrogen) at a pressure of 0.3 bar and a flow rate of 4 L/min, at voltages in a range from –500 to 4500 V. The completion of all reactions was checked by method of thin-layer chromatography (TLC).
Synthesis of Isatin Derivatives
Preparation of 3-(2-aminophenylimino)-5-substituted-indolin-2-ones (3a–3d). Equimolar (0.01 mole) amounts of 5-substituted isatin and o-phenylenediamine were taken in 30 mL of ethanol, 5-10 drops of glacial acetic acid was added to the reaction mixture, the reaction was refluxed at 120°C, and then the mixture was kept overnight for the formation of precipitate. The precipitate was filtered and recrystallized from methanol and chloroform. The reaction was monitored with the help of TLC using hexane/ethyl acetate (1:1) solvent mixture (Fig. 9) [37, 38].Fig. 9. Synthesis of designed isatin derivatives 3a–3d.
Preparation of 3-(2-aminophenylimino)-N-5-substitutedindolin-2-ones (6a–6d). Atwo-step reaction was employed to form final compounds. In step 1, equimolar (0.01 mole) amounts of 5-substituted isatin and KOH were taken in 5 mL of DMF and the reaction mixture was stirred for 10 min at room temperature. Then, halohydrocarbon (0.01 mole) was added dropwise, the mixture was exposed to microwave radiation and the reaction was heated for 5 min at 80°C (400 W), then heated further for 5 min at the same temperature, and finally cooled for 5 min under fan. The reaction mixture was poured on ice so that the precipitate was formed and then it was filtered. The completion of the reaction was checked with the help of TLC using hexane/ethyl acetate (1:1) solvent mixture. In step 2, equimolar amounts of intermediate (0.003 moles) and o-phenylenediamine (0.003 moles) were taken in 30 mL of ethanol, 5 – 10 drops of glacial acetic acid were added, and the reaction was refluxed at 120°C. After reflux, the mixture was cooled overnight and the obtained precipitate was recrystallized from methanol and chloroform. The reaction was monitored with the help of TLC using hexane/ethyl acetate (1:1) solvent mixture (Fig. 10) [33, 37, 38].Fig. 10. Synthesis of designed isatin derivatives (6a–6d).
Antimicrobial Activity Assay
All the designed compounds were screened for in vitro antibacterial activity in Luria agar medium. The antibacterial activities of the compounds were evaluated against Gram-positive bacteria (Staphylococcus aureus MTCC-6908) and Gram-negative bacteria (Escherichia coli MTCC-433), using Amoxicillin as a standard drug that acts as antimicrobial agent by inhibition of bacterial cell wall.
Growth inhibition zone determination. Antimicrobial activities of compounds 3a–3d and 6a–6d were evaluated using the growth inhibition zone method [39]. According to this, various dilutions of synthesized compounds (1, 2, 4, 8 and 16 μg/mL) were prepared from the stock of 100 μg/mL in dimethyl sulfoxide for determining the growth inhibition zone size. Visibility of the zone was enhanced using 1% methylene blue dye solution. The test bacteria were swabbed over the plates. The plates were incubated at 37°C. Diameters of the growth inhibition zones were measured after 24 h treatment with reference to Amoxicillin as standard drug for antibacterial activity.
Result
Compounds 3a–3d have been synthesized using a single- step condensation reaction (Fig. 9). Other series of compounds (6a–6d) were synthesized in two steps: first N-alkylation and then condensation (Fig. 10). The synthesized isatin derivatives were structurally characterized by spectral methods including FT-IR, 1H NMR, and MS. The prepared compounds were also evaluated in vitro for their antimicrobial activity.
Analytical Data
3-(2-Aminophenylimino)-indoline-2-one (3a). Yield, 28.52%; crystalline bright yellow compound; m.p.. 235 – 237°C; FT-IR (KBr, νmax, cm-1) 3417 (N–H), 3064 (Ar C–H), 2827 – 2968 (C-H), 1710 (CO), 1617 (C=N), 1596-1574 (R-NH2), 1135 – 1209 (C-O); 1H-NMR (DMSO-d6, δ, ppm): 8.05 (s, 1H) 7.36 (s, 1H), 7.07 (s, 1H), 7.59 (dd, 2H, J1=8.4, J2=0.5 Hz), 7.34 (s, 1H), 7.36 (s, 1H), 7.39 (s, 2H), 3.37 (s, 2H); mass spectroscopy (ESI-MS): m/z 238 [M+1]+.
3-(2-Aminophenylimino)-5-methylindoline-2-one (3b). Yield, 30.32%; crystalline bright yellow compound; m.p.243 – 245°C; FT-IR (KBr, νmax, cm-1) 3126 (N–H), 3021 (Ar C–H), 2841-2711 (C-H), 1716 (C=O), 1655 (C=N), 1467 (R-NH2), 1392 (CH3), 1135 – 1244 (C-O); 1H-NMR (DMSO-d6, δ, ppm) 8.02 (s, 1H), 7.43 (d, 2H, J=8.2 Hz), 7.58 (dd, 3H, J1=8.4, J2=0.5 Hz), 7.74 (s, 1H), 7.78 (s, 1H), 3.37 (s, 3H), 2.47 (s, 2H); mass spectroscopy (ESI-MS): m/z 252 [M+1]+.
3-(2-Aminophenylimino)-5-chloroindoline-2-one (3c). Yield, 26.81%; orange crystalline compound; m.p. 263 – 265°C; FT-IR (KBr, νmax, cm-1) 3126 (N–H), 3021 (Ar C–H), 2841 – 2711 (C-H), 1716 (C=O), 1665 (C=N), 1467 (R-NH2), 1119-1278 (C-O), 752 (R-Cl); 1H-NMR (DMSO-d6, δ, ppm) 8.05 (s, 1H), 7.57 (dd, 1H, J1=8.5 , J2=1.7 Hz), 7.65 (dd, 2H, J1=8.5, J2=0.5 Hz), 7.72 (s, 1H), 7.75 (s, 1H), 7.79 (s, 1H), 7.81 (s, 1H), 3.34 (s, 2H); mass spectroscopy(ESI-MS): m/z 272 [M+1]+.
3-(2-Aminophenylimino)-5-fluoroindoline-2-one (3d). Yield, 32.56%; crystalline yellowish orange compound; m.p. 258 – 260°C; FT-IR (KBr, νmax, cm-1) 3372 (N–H), 3097 (Ar C–H), 2853 – 2897 (C-H), 1700 (C=O), 1657 (C=N), 1528 – 1574 (R-NH2), 1288 – 1165 (C-O); 1H-NMR (DMSO-d6, δ, ppm) 8.03 (s, 1H) 7.02 (d, 1H, J=8.4 Hz), 7.29-7.34 (d, 2H, J=8.4 Hz), 7.51 (s, 1H), 7.74 (s, 1H), 7.76 (s, 1H), 7.80 (s, 1H), 3.30 (s, 2H); mass spectroscopy (ESI-MS): m/z 256 [M+1]+.
3-(2-Aminophenylimino)-1-benzyl-5-chloroindoline-2-one (6a). Yield, 49.03%; light yellow crystalline compound; m.p. 228 – 230°C; FT-IR (KBr, νmax, cm-1) 3098 (N–H), 3025 (Ar C–H), 2843 (C-H), 1615 (C=O), 1616 (C=N), 1467 (R-NH2), 1119 – 1278 (C-O), 797 (R-Cl), 751 (CH2); 1H-NMR (DMSO-d6, δ, ppm) 8.05 (s, 1H), 7.57 (t, 1H, J=7.7 Hz), 7.70 (dd, 3H, J1=7.9, J2=0.6 Hz), 7.7 (s, 3H), 7.71 (s, 2H), 7.81 (s, 2H), 7.84 (s, 2H), 3.34 (s, 2H); mass spectroscopy ESI-MS): m/z 362 [M+1]+.
3-(2-Aminophenylimino)-N-benzoyl-5chloroindoline-2-one (6b). Yield, 51.29%; crystalline light yellow compound; m.p. 287 – 290°C; FT-IR (KBr, νmax, cm-1) 3285 (N–H), 3095 (Ar C–H), 2833 – 2895 (C-H), 1647 (C=N), 1567 (NH) ,1427-1453 (R-NH2), 1361 – 1394 [CH(CH3)2], 1196 – 1275 (C-O), 750 (R-Cl), 725 (CH2); 1H-NMR (DMSO-d6, δ, ppm) 8.1 (s, 1H), 7.19 (t, 1H, J=7.5 Hz), 7.37 (d, 3H, J =8.5 Hz), 7.48 (d, 2H, J=1.1 Hz), 7.75 (dd, 3H, J1=8,0, J2=1.6 Hz), 6.86 (d, 2H, J=7.4 Hz), 3.95 (s, 1H), 3.09 (s, 1H); mass spectroscopy (ESI-MS): m/z 376 [M+1]+.
3-(2-Aminophenylimino)-N-benzoyl-5fluoroindoline-2-one (6c). Yield, 42.61%; crystalline light yellow compound; m.p. 138 – 140°C; FT-IR (KBr, νmax, cm1) 3070 (N–H), 3006 (Ar C–H), 2558 – 2835 (C-H), 1687 (C=N), 1494 (NH), 1423 (R-NH2), 1325 (CH(CH3)2), 1127 – 1229 (C-O), 707 (CH2); 1H-NMR (DMSO-d6, δ, ppm) 8.13 (s, 1H), 7.26 (t, 1H, J=7.5 Hz), 7.27 (dd, 3H, J1=8.5, J2=1.6 Hz), 7.3 (s, 1H), 7.52 (s, 1H), 7.69 (s, 1H), 7.74 (s, 1H), 5.71 (s, 2H), 3.85 (s, 1H), 3.19 (s, 2H); mass spectroscopy (ESI-MS): m/z 360 [M+1]+.
3-(Aminophenylimino)-N-benzoyl-5methylindoline-2-one (6d). Yield, 46.76%; crystalline light yellow compound; m.p. 158 – 160°C; FT-IR (KBr, νmax, cm1) 3071 (N–H), 3010 (Ar C–H), 2558 – 2834 (C-H), 1686 (C=N), 1583 (NH), 1453 – 1423 (R-NH2), 1325 (CH(CH3)2), 1127 – 1295 (C-O), 707 (CH2); 1H-NMR (DMSO-d6, δ, ppm) 8.12 (s, 2H), 7.25 (t, 1H, J=7.5 Hz), 7.51 (dd, 2H, J1=8.5, J2=1.4 Hz), 7.46 (s, 2H), 7.48 (s, 2H), 7.59 (s, 1H), 7.72 (s, 1H), 6.01 (s, 2H), 2.5 (s, 4H); mass spectroscopy (ESI-MS): m/z 356 [M+1]+.
Antimicrobial Activity Assay
All synthesized compounds were screened for in vitro antibacterial activity in the Luria agar medium. The antibacterial activities of compounds were evaluated against Gram-positive bacteria (Staphylococcus aureus MTCC-6908) and Gram-negative bacteria (Escherichia coli MTCC-433) (Tables 1, 2 and Fig. 11, 12).Table 1. Growth Inhibition Zone Diameters (mm) of Synthesized Compounds and Standard Drug Tested against Staphylococcus aureus
Conc. mg/mL 3a 3b 3c 3d 6a 6b 6c 6d Standard
1 8 ± 0.57 8.5 ± 0.86 9.5 ± 0.28 7.5 ± 0.28 6 ± 0.57 8 ± 0.57 7.5 ± 0.29 8 ± 0.57 10 ± 0.57
2 10.5 ± 0.86 10 ± 0.57 11 ± 1.15 10 ± 0.57 8 ± 0.57 9.83 ± 0.44 9.83 ± 0.44 9.5 ± 0.28 10.5 ± 0.86
4 11.37 ± 0.36 11.37 ± 0.21 11.25 ± 0.43 9.37 ± 0.43 9.37 ± 0.44 10.12 ± 0.07 10.25 ± 0.29 10.37 ± 0.21 11.75 ± 0.43
8 12.5 ± 0.86 12 ± 0.5 13 ± 1.15 13 ± 1.15 10 ± 0.57 11.5 ± 0.57 11.5 ± 0.29 12.5 ± 0.28 14 ± 0.57
16 13.5 ± 0.28 13.5 ± 0.86 16.5 ± 0.28 12.5 ± 0.28 13.5 ± 0.28 13.5 ± 0.28 12.5 ± 0.29 13.5 ± 0.28 15 ± 1.15
Data are expressed as mean ± SD (n = 3)
Table 2. Growth Inhibition Zone Size (mm) of Synthesized Compounds and Standard Drug Tested against Escherichia coli
Conc. μg/mJ 3a 3b 3c 3d 6a 6b 6c 6d Standard
1 10 ± 0.57 8.5 ± 0.28 10 ± 0.57 9.5 ± 0.28 7.5 ± 0.28 8.5 ± 0.28 8.5 ± 0.29 6.5 ± 0.28 8.5 ± 0.86
2 10.5 ± 0.28 10 ± 0.57 11.5 ± 0.28 10.5 ± 0.86 9.83 ± 0.44 9.5 ± 0.28 9.5 ± 0.29 8 ± 0.58 10 ± 0.57
4 11.5 ± 0.28 10.5 ± 0.28 11.37 ± 0.07 11.25 ± 0.25 10.37 ± 0.07 9.79 ± 0.39 10.12 ± 0.07 9.37 ± 0.56 11.75 ± 0.14
8 12 ± 0.57 12 ± 0.57 12.5 ± 0.28 13.5 ± 0.28 12.5 ± 0.28 12.5 ± 0.28 12.5 ± 0.29 13.5 ± 0.29 13.5 ± 0.28
16 13.5 ± 0.28 13 ± 0.57 13.5 ± 0.28 15.16 ± 1.16 14.5 ± 0.28 13.5 ± 0.28 14 ± 0.58 13 ± 1.16 16 ± 0.57
Data are expressed as mean ± SD (n = 3)
Fig. 11. Antibacterial activity of synthesized compounds and standard drug tested against Staphylococcus aureus.
Fig. 12. Antibacterial activity of synthesized compound and standard drug tested against Escherichia coli.
Statistical Analysis
The obtained values were expressed as mean ± standard deviation and statistical analysis was carried out by one-way ANOVA, P < 0.05 was considered significant, and the graphs were made using Graph pad Prism 9.2 software.
Discussion
The goal of medicinal chemistry is to develop new therapeutic agents. Isatin derivatives are known to have various effects on the brain and infections. Hence, this is an important class of bioactive compounds displaying caspase inhibitor, antibacterial, and antiproliferative activity. Isatin is a multipurpose substrate that can be cast-off to prepare a large variety of heterocyclic compounds like indoles and quinolines and can be used in drug synthesis. Hence, knowledge of the structure and activity of isatin derivatives may provide an effective and cheaper regimen against various bacteria, fungi, and viruses.
This biological study showed that halogenation at 5th position of the initial compound increases its activity which leads to the discovery of new derivatives. The activity of compounds was found to be more potent against Gram-positive bacteria as compared with Gram-negative bacteria (3c, 3d, 6a, 6b, 6d). Compound 3c displayed higher antimicrobial activity than standard drug (amoxicillin) against Staphylococcus aureus at higher concentration (16 μg/mL) and against Escherichia coli at lower concentration (1 μg/mL). The synthesized derivatives of isatin act by mixed mechanism as antimicrobial agent, that is by (i) inhibition of bacterial cell wall and (ii) inhibition of bacterial cell fusion. Further studies on isatin derivatives are required to develop effective and cheaper regimens.
Conflict of Interest
The authors declare that they have no conflicts of interest.
The authors would like to thank honorable vice-chancellor for providing facilities to carry out this research work.
==== Refs
References
1. Chahal VW Nirwan S Kakkar R Med. Chem. Commun. 2019 10 3 351 368 10.1039/C8MD00585K
2. Visagaperumal D Ezekwem JE Munji H Chandy V Pharma Tutor. 2018 6 5 38 47 10.29161/PT.v6.i5.2018.38
3. Shukla PK Singh MP Patel R J. Appl. Pharm. Sci. Res. 2018 1 2 16 22
4. Grewa SA Int. J. Pharm. Sci. Res. 2014 6 1 1 8
5. Silva VB J. Braz. Chem. Soc. 2013 24 5 707 720
6. Premanathan M Radhakrishnan S Kulangiappar K Indian J. Med Res. 2012 136 5 822 826 23287130
7. Yoshikawa M Murakami T Kishi A Chem. Pharm. Bull. 1998 46 5 886 888 10.1248/cpb.46.886
8. Bergman J Lindstrom JO Tilstam U Tetrahedron 1985 41 14 2879 2881 10.1016/S0040-4020(01)96609-8
9. da Silva JFM Garden SJ Pinto AC J. Braz. Chem. Soc. 2001 12 3 273 324 10.1590/S0103-50532001000300002
10. Andreani A Burnelli S Granaiola M Eur. J. Med. Chem. 2010 45 1374 1378 10.1016/j.ejmech.2009.12.035 20060202
11. F. Sonmez, Z. Gunesli, Z. B. Kurt, et al., Mol. Divers., 1 – 16 (2019).
12. R. Jarapula, K. Gangarapu, S. Manda, and S. Rekulapally, Int. J. Med. Chem., 1 – 9 (2016).
13. Prasad DB Vasanthi R Kanth CB IJBPR. 2012 3 1 182 187
14. Jarrahpour A Khalili D Clercq DE Molecules 2007 12 1720 1730 10.3390/12081720 17960083
15. Sridhara KS Saravanana M Ramesh A Eur. J. Med. Chem. 2001 36 7–8 615 625 10.1016/S0223-5234(01)01255-7 11600231
16. Abbas YS Farag AA Ammar AY Monatsh. Chem. 2013 144 1725 1733 10.1007/s00706-013-1034-3 32214479
17. Fadl AT Jubair BSAF Int. J. Res. Pharm. Sci. 2010 1 2 113 126
18. Fadl AT Jubair BSAF Wafa AO Eur. J. Med. Chem. 2010 45 10 4578 4586 10.1016/j.ejmech.2010.07.020 20696500
19. Vine LK Matesic L Locke MJ Anticancer Agents Med. Chem. 2009 9 397 414
20. Meleddu R Petrikaite V Distinto S ACS Med. Chem. Lett. 2019 10 571 576 10.1021/acsmedchemlett.8b00596 30996798
21. Verma M Pandeya HNS Singh NK Acta Pharm. 2004 54 49 56 15050044
22. Mathur G Nain S Med. Chem. 2014 4 4 417 427
23. Meleddu R Distinto S Corona A J. Enzyme Inhib. Med. Chem. 2017 32 1 130 136 10.1080/14756366.2016.1238366 27766892
24. Sridhar SK Pandeya SN De Clercq ED Bull. Chim. Farm. 2001 140 5 302 305
25. Zhou L Liu Y Zhang W J. Med. Chem. 2006 49 3440 3443 10.1021/jm0602357 16759084
26. Chen LR Wang CY Lin WY Bioorg. Med. Chem. Lett. 2005 15 3058 3062 10.1016/j.bmcl.2005.04.027 15896959
27. Selvam P Murgesh N Chandramohan M Indian J. Pharm. Sci. 2008 70 1 91 94 10.4103/0250-474X.40339 20390088
28. Nathani RB Pandya SK Jeni MM Der. Chem. Sinica 2011 2 6 97 103
29. Singh SG Desta YZ Chem. Inform. 2013 44 41 74 115
30. Chaithanya B Kasiviswanath VI Chary PD Bull. Chem. Soc. Ethiop. 2019 33 2 321 329 10.4314/bcse.v33i2.12
31. Alsalihi IE Fahdawi ASA Aro – Sci J. Koya Univ. 2018 6 1 38 45
32. S. M. Almutairia, S. A. Zakariab, P. P. Ignasiusc, et al., J. Mol. Struct., 1 – 33 (2017).
33. Bogdanov VA Mironov FV Russ. J. Gen. Chem. 2015 85 10 2413 2415 10.1134/S107036321510031X
34. A. El-Faham, N. W. Hozzein, M. A. M. Wadaan, et al., J. Chem., 1 – 8 (2015).
35. Singh KU Pandeya NS Singh A Int. J. Pharm. Sci. Drug. Res. 2010 2 2 151 154
36. Phogat P Singh PA Cent. Nerv. Syst. Agents. Med. Chem. 2015 15 1 28 31 10.2174/1871524915666150213122246 25693647
37. Prakash CR Raja S Selvam PT Rasayan. J. Chem 2009 2 4 960 968
38. Hajare RA Gaurkhede RM Chinchole PP Asian J. Res. Chem. 2009 2 3 289 291
39. Balouirin M Sadiki M Ibnsouda KS J. Pharm. Anal. 2016 6 71 79 10.1016/j.jpha.2015.11.005 29403965
|
PMC010xxxxxx/PMC10170450.txt |
==== Front
Pharm Chem J
Pharm Chem J
Pharmaceutical Chemistry Journal
0091-150X
1573-9031
Springer US New York
2878
10.1007/s11094-023-02878-1
Medicinal Plants
Madhuca indica: A Review on the Phytochemical and Pharmacological Aspects
Roat Priyanka 1
Hada Sonal 1
Chechani Bhawna 1
Yadav Dinesh Kumar 1
Kumar Sanjay sanjaycsmcri1981@gmail.com
2
Kumari Neetu neetukumari@mlsu.ac.in
1
1 grid.440702.5 0000 0001 0235 1021 Department of Chemistry, Mohanlal Sukhadia University, Udaipur, 313001 India
2 grid.444415.4 0000 0004 1759 0860 Department of Chemistry, University of Petroleum and Energy Studies, Dehradun, 248007 India
10 5 2023
2023
57 2 284295
29 11 2021
© Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Madhuca indica J.F. Gmel. (family: Sapotaceae), commonly known as Mahua in Indian dialects, occupies the importance as one of the fuel-efficient, energy-saving plant species. Extensive studies showed that the presence of phytochemicals e.g., carbohydrates, fatty acids, flavonoids, saponins, steroids, triterpenoids and glycosidic compounds in the extract of this species. Pharmacologically, it has been used against various disorders in indigenous system of medicine, inckuding antioxidant, anti-inflammatory, anticancer, hepatoprotective, anti-diabetic and wound healing activities. This review highlights various pharmacological activities, phytochemistry and importance of M. indica plant for medicine.
Keywords
Madhuca indica
cardiomyopathy
oxidative stress
flavones
arbutin derivatives
issue-copyright-statement© The Editorial Board of the Khimico-Farmatsevticheskii Zhurnal 2023
==== Body
pmc 1. INTRODUCTION
The genus Madhuca belongs to the family Sapotaceae includes more than 800 plant species, many of which are mostly used for the production of latex for example guttapercha [1]. Royen (1960) [2] has listed around 84 species of Mahua out of which Madhuca lotifolia, M. longifolia, M. butyracea, M. neriifolia and M. bourdillonii species are common to the Indian continent. Other species of the same are reported from Asian continents, e.g., Malaysia, Pakistan, Sri-Lanka and Thailand as well as from Australian subcontinent (Fig. 1). It is prominent multipurpose species, which provides the answer for the four major aspects i.e., food, fodder, fuel and medicine [3]. Beside the numerous pharmacological activities, it is still under investigation in search of various pharma applications [4]. In the present review, various uses of the plant species with focus on its future pharmacological aspects are discussed.Fig. 1. Photographs of Madhuca indica tree (A) and its grey trunk (B), leaves (C), yellowish flowers (D), dried flowers (E), and seeds (F)
2. GENERAL DESCRIPTION
2.1 Morphological Characterization of Madhuca indica
Habit – Large ever green deciduous trees.
Habitat – Common in forest area.
Stem – Bark black, grayish or ash colored, longitudinally fissured [5].
Leaves – Coriaceous, elliptic or elliptic, obtuse or sub-acuminate.
Flowers – Creamy-white, dense, axillary clusters at the end of the branches.
Fruit – Berries are ovoid, fleshy and yellow when ripe.
Seeds – 1-4 seeded, seeds are ovoid in shape smooth, shining, brownish-black in color.
Flowering, cultivation and collection. Flowering of this medium-sized tree takes place during the season of March to June in every year [6, 7]. In India, there are all types of climatic condition for better propagation and cultivation of trees, varying from hot to cold and humid to dry as we go over different regions [5]. However, M. indica is not cultivated intentionally and occurs as self-growing trees.
Taxonomy of Madhuca indica J.F. Gmel. Kingdom – Plantae; Order – Ericales; Family – Sapotacase; Genus – Madhuca; Species – indica.
Vernacular names. English – Indian butter tree; Hindi – Mahua, Mohwa, Mauwa; Bangali – Mahwa, Maul, Mahwla; Marathi – Mahwa, Mohwra; Gujrati – Madhuda; Telgu – Ippa; Tamil – Illupei, Ewpa[ Kannad – Tuppe; Oriya-Mahula, Moha, Madgn [5].
2.2. Ethno-Medicinal Uses
Bark of M. indica species has been reported to possess free radical scavenging activity and for that reason the astringent and emollient prepared from the bark extract are used to medicate fracture, swelling, itching, snake bite, wounds along cure of leprosy [8]. The other application of the bark is towards its application as coolant for mouth ulcer, inflammation, bleeding and spongy gums [9]. Leaves are useful in anti-burns, anti-vomitive, eczema, emollient, ruptured cartilage, migraine and hemorrhoids [10]. Distilled juice of flowers is considered as a nutritional cooling tonic and used in the treatment of expectorant, stimulant, diuretic, helminthes, antihelmintic, verminosis [11] strangury, cough and bronchitis [6]. The whole flower is used to increase the lactation among the tribal communities [10]. The fruits are astrictive and largely engaged as a lotion for curing chronic ulcer, in acute tonsillitis and pharyngitis [12–14]. Seed oil is used for the treatment of skin disease, headache, rheumatism and as a laxative agent [10].
Value-added Mahua products. In addition to eatable and pharmacological purposes, M. indica is also used in the production of washable cleansers, insecticide and pesticide. Dry flowers, powder leaves, seed oil have been utilized in the sugar syrup (sweeting agent), jam and jelly, liqueur, cosmetics, soap and aroma-oil etc. [15, 16].
3. PHARMACOLOGICAL STUDIES ON Madhuca indica EXTRACTS
Beside various therapeutic potentials, pharmacological studies have certified most of the ethno-medicinal uses of M. indica which have potential to provide health to the society [17]. Important pharmacological activities of various M. indica extracts are summarized in Table 1.TABLE 1. Pharmacological Properties of M. indica Extracts and Some Isolated Compounds
S.No. Pharmacological activities Parts/extracts/chemical constituents References
1. Antifertility activity Leaves: Methanolic extract [18]
Seeds: Aqueous extract [19]
2. Antifungal activity Leaves: Aqueous/methanolic extract [20, 21]
Flowers: Aqueous/ethanolic extract [22]
3. Antihyperglycemic activity Leaves: Methanolic extract [23]
4. Antiulcer activity Leaves: Aqueous extract [24]
Leaves: Methanolic extract [25]
Seeds: Ethanolic/alkaloid extract [10]
Bark: Aqueous /ethanolic extract [17]
Leaves: Methanolic extract [26]
Leaves: Methanolic extract/3,5,7,3′,4′-Pentahydroxy flavone [27]
5. Antibacterial activity Leaves and Bark: Aqueous /methanolic extract [28–30]
6. Antioxidant activity Bark: Phenolic/methanolic extract [31]
Leaves and Bark: Methanolic extract/Myricetin/Quercetin/â-Sitosterol [30, 32, 33]
Bark: Methanolic extract [34]
Bark: Aqueous /methanolic extract [35]
Leaves: Aqueous/acetone extract [36]
Leaves, Bark and Dried wood: Ethanolic extract [37]
Flowers: Methanolic extract [38]
7. Analgesic activity Leaves, Bark and Flowers: Methanolic extract [39]
Leaves: Methanolic extract [40]
8. Antidiabetic activity Bark: Aqueous extract [41]
Seeds: Ethanolic/alkaloid extract [42]
9. Anti-inflammatory activity Leaves, Bark and Flowers: Methanolic extract [39]
Leaves: Pet. ether/ethyl acetate/methanolic extract [40]
10. Anti-microbial activity Seeds: Ethanolic/alkaloid extract [43]
Bark: Ethanolic extract [34]
Leaves: Aqueous/acetone extract [36]
Leaves and Bark: Methanolic extract [30]
Bark: Ethanolic extract [34]
Bark: Organic extract [29]
Leaves, Bark and Flowers: Methanolic/aqueous extract [44]
11. Antipyretic activity Flowers and Seeds: Methanolic extract/3′,4′-Dihydroxy-5,2′-dimethoxy-6,7- methlendioxy isoflavone [45]
12. Hepatoprotective activity Leaves, Bark and Flowers: Methanolic extract [43]
13. Anticancer activity Bark and Leaves: Ethanolic extract [46]
14. Cytotoxic activity Flowers: Methanolic extract [47]
15. Anthelminthic activity Bark and Leaves: Methanolic extract [38]
16. Antinociceptive activity Bark: Methanolic extract [48]
17. Antidiarrhoeal activity Bark: Ethanolic extract [48]
18. Inhibitory activity Bark: Ethanolic extract [8]
19. Tyrosinase inhibitors Bark: Methanolic extract/Madhucoside A/ Madhucoside B/ Protobassic acid [49]
20. Immunomodulatory activity Fruits: Methanolic extract/Ursolic acid [50]
21. Wound healing capacity Leaves, Bark and Flowers: Ethanolic extract [51]
22. Toxicity Leaves, Bark and Flowers: Ethanolic extract [39, 52, 53]
23. Radioprotective Activity Root and Seed oil: Phenolic extract [54]
24. Cardioprotective effect Leaves: Methanolic extract/3,5,7,3′,4′-Pentahydroxy flavone [55]
3.1. Anti-Ulcer Activity
Chronic administration of non-steroidal anti-inflammatory drugs (NSAIDs) leads to gastric damage occurs mainly in the corpus region of the stomach and tends to be mostly in the form of erosion rather than ulcer. Naproxen, aspirin and indomethacin are the most commonly used NSAIDs leading to haemorrhages and perforation. Among these, naproxen is a non-corticosteroid drug with anti-inflammatory, antipyretic and pain-relieving properties, which is known to produce erosion, antral ulceration, and petechial bleeding in the mucosa in the stomach as an adverse effect. Beside the above activities, production of oxygen free radicals and lipid peroxidation play a crucial role in the development of the gastric antral ulceration induced by naproxen. The aqueous extract of M. indica leaves has been reported to be a significant protective constituent against naproxen-induced gastric antral ulcer. This activity is supposed to be observed occurred due to the presence of quercetin, myricitrin, triterpenoid, â-sitosterol and quercitrin species present in the extract [24]. Anti-ulcer activities of the alcoholic extract of the leaves of this plant have been studied by various researchers in-vivo too [25]. The presence of antioxidants for example 3,5,7,3′,4′-pentahydroxy flavone like compounds were supposed to be the key constituent for the reduction of ulcerated area, pepsin content along with higher pH of the gastric fluid [27].
3.2. Anti-Oxidant Activity
Oxidative stress is the main pathological culprit for various diseases. The damages caused by free radical induced oxidative stress leads to the cancer, tissue injury, rheumatoid arthritis, neurodegenerative diseases as well as aging. The antioxidants are substances which inhibits oxidative damage of target molecule. The examination involved the ability of the antioxidant to reduce the stable DPPH radical, to the yellow colored diphenylpicrylhydrazine, while the scavenging of the superoxide anion is performed with nitro blue tetrazolium method. The in-vitro method of superoxide anions, generated by the phenazin methosulfate (PMS)/nicotinamide- adenine-dinucleotide, reduced form (NADH) system detected by the reaction along with chloride of 2,2′-di-p-nitrophenyl- 5,5′-diphenyl-(3,3′-dimethoxy-4,4′-difphenylene) ditetrazolium chloride (NBT method). The reference chemical is ascorbic acid having the highest inhibition potential of 91.26 % at 20 μg/mL concentration. Superoxide(SO) free radical scavenging activity and DPPH free radical scavenging method have been employed. The scavenging effect of ethanolic effect of M. indica on DPPH radical was 87.6% at a concentration of 500 μg/mL. The obtained results have highlighted the function of M. indica extracts to destroy the superoxide anions produced in PMS-NADH-NBT system and indicate that extract has a noticeable effect on scavenging the free radicals. Superoxide anion scavenging activity of M. indica was 70% at a concentration of 500 μg/mL [34]. Antioxidant activities of alcoholic and aqueous extracts of Madhuca species have been reported by various researchers. 81.70% and 57.38% inhibition potential, and IC50 values of 92.85 μg/mL and 137.5 μg/mL of DPPH free radical scavenging assay by the methanolic and aqueous extract of bark respectively at optimum concentration of 150 μg/mL [35]. The methanolic extract is reported to possess higher antioxidant potential as compared to aqueous extract, as later is rich in sugar components. Manmode et al. [56] reported the antioxidant activity of ayurvedic medicine name Kumariasava which was prepared by fermented flowers of M. indica, Woodfordia fruticose and Saccharomyces cerevisiae yeast. This formulation was found to exhibit higher antioxidant activity than standard drug ascorbic acid [57].
3.3. Anti-Inflammatory Activity
Inflammation is a reaction of living tissues towards injury, ischemia, cell death, cancer and it comprises systematic and local response. This activity is performed by doing a test for anti-inflammatory agent, acting by inhibiting the mediators of acute inflammation of the inhibition capability, of the sample towards the carrageenan-induced rat paw oedema. The initial phase of the response is due to the release of histamine and serotonin, while the second phase is sensitive to most clinically effective anti-inflammatory drugs. According to many researchers, the saponins and flavonoids present in the crude methanolic extracts of the M. indica is supposed to possess the anti-inflammatory activity and are useful in treatment of the fever, inflammation and pain as an alternative for non-steroidal drugs such as paracetamol and phenyl butanone. It is important to point out that the phytochemical analysis showed the presence of anti-inflammatory activity [39].
3.4. Anti-Bacterial Activity
Anti-bacterial potential of alcoholic extracts of bark of M. indica have been reported to very effective towards both types of bacteria with a higher selectivity towards the Gram-positive bacteria at all concentrations. The effectiveness towards the bacterial strains lies in the order of S. aureus > B. sublitis > E. coli > P. aeruginosa, suggesting that the presence of the phytochemicals play the key roles towards the antibacterial activity of the extracts. Fusarium oxysporum and Colletotrichum capsici were sensitive to oil-seed cake extracts of M. indica cold acetone fraction of oil-seed-cake exhibited the zone of inhibition (ZI, mm) 8.1 mm to S. aureus, which is slightly less effective than amoxicillin (11 mm). Cold acetone extract of Madhuca oil-seed cake inhibited only S. aureus at 100% (ZI = 5.1 ± 0.01) [58]. The ethanol extract of leaves was experienced against food-borne pathogens including Gramnegative bacteria (E. coli O157:H7 and Salmonella enteriditis) and Gram-positive bacteria (Listeria monocytogenes and methicillin-resistant Staphylococcus aureus) using disc diffusion method. In addition, minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) of plant extract were determined. In the disc diffusion method, the leaves extract was effective in reducing the bacterial growth of tested food-pathogens. The zone of inhibition range was from 16mm-19 mm. Madhuca indica extract was effectively reduced bacterial growth. The ethanolic extract of leave could be potentially used as natural food preservers to control and prevent food-borne pathogens So that, ethanolic leaves extract of M. indica has shown anti-microbial activity. It is due to presence of chemical constituents like carbohydrates, flavonoids, tannins and proteins which was confirmed by physicochemical studies [59].
3.5. Miscellaneous Activities
The alcoholic extract of M. indica bark was reported to be effective against CCl4 induced hepatocellular injury, by lowering of the serum levels of glutamate pyruvate transaminase (SGPT), serum glutamate oxaloacetate transaminase (SGOT), serum bilirubin and serum alkaline phosphate (SALP) to have a significant effect [43]. The alcoholic extract of flowers has been claimed to be anti-cancer agent. They claimed that the extract was able to reduce the breast cancer cell proliferation and apoptosis by inhibiting the expression of COX-2 in MCF-7 and MDAMP-468 cell lines. They observed that the activity involves the inducing of apoptosis through up regulating the expression of Caspase assay 3/7 (P < 0.0001), through a decrease in the expression of COX-2 mRNA and COX-2 protein in both MCF-7 and MDA-MB-468 cells, in a dose dependent manner. The synergistic effect of extract with a chemotherapeutic drug paclitaxel, found to be more effective as compared to either extract or paclitaxel alone [47]. Antinociceptive and antidiarrheal activity of the bark extract was studied on mice [48] by using acetic acid induced writhing resulting algesia by liberation of endogenous substances which excite the pain nerve endings, and castor oil induced diarrhea, respectively. It was claimed that the extract was effective in a dose dependent study as compared to standard drug e.g., diclofenac sodium for antinociceptive and loperamide for anti-diarrheal. The activity was effective at higher (500 mg/kg) dose as compared to standard drugs at 25 mg/kg for diclofenac sodium and at 500 mg/kg for loperamide respectively, suggesting that activity guided fractionation have to be done to meet the dose norms. Activity based guided fractionation and isolation of compounds followed by their detailed characterization was reported for the alcoholic extract of M. indica leaves [55] that was studied by electrocardiography, hemodynamics, left ventricular function alterations, expression of cardiac markers (LDH, CK-MB, AST, ALT and ALP), lipid metabolism (total cholesterol, triglyceride, LDL, HDL and VLDL), cardiac oxido-nitrosative stress, Na-K-ATPase level and mitochondrial enzymes (I-IV) activity after arsenite administration in the presence of the QTN as compared to control. It was claimed that the QTN possess the protective efficacy against arsenic induced cardiotoxicity [55].
Application of the aqueous extract of M. indica bark also have been reported to be natural radioprotector for the normal cells surrounding cancerous cell during radiation exposure through in-vitro, in-vivo, and in-silico models [54]. This study revealed that the extract was able to protect the pBR322 DNA of normal cell that was supposed to be master protein, regulating the whole pathway among expressed 437 proteins during electron beam radiation. This was further supported through the molecular interaction between p53 and M. indica extract by quantitative structure–activity relationship and ADMET properties. They concluded that quercetin, myricitrin, and 7-hydroxyflavone were promising inhibitors of p53 protein during cancer treatment due to EBR induced DNA damage. Tang et al. [60] reported in about auto-immune disease, i.e., joint dysfunction caused by rheumatoid arthritis. 3,5,7,3′,4′-Pentahydroxy flavone, i.e., quercetin isolated from methanolic leaf extract of M. indica is responsible for the anti-arthritic potential of against Freund's complete adjuvant induced experiment on female Wistar rats. Treatment with 3,5,7,3′,4′-pentahydroxy flavone (10 and 20 mg/kg) showed significant (p < 0.05) inhibition of increased joint diameter, paw volume, paw withdrawal threshold, and latency. The elevated synovial oxidative stress (Superoxide dismutase, reduced glutathione, and malondialdehyde) and protein levels of Tumor necrosis factor-á and Interleukin were markedly (p < 0.05) also reduced by 3,5,7,3,4-pentahydroxy flavone. It is also effectively (p < 0.05) ameliorated cyclooxygenase-2 [61]. Geriatric stage in aged people is a nearby constant auto systemic inflammation stage. In winter temperature change is more impaired due to weak vessels which become more sieve-like atopic hemodynamics. This is age related and height related vasculitis and co-morbidity conditions. Apart from this, there are other conditions such as unsleeping, restless, lethargic, constipation, indigestion, pangastritis, spastic motion, atopic heart beats, psycho-depression kind of situation (mostly pseudo), dermal itching, erythema, adipsia\hypodipsia, dysphagia, UTI or incontinence etc. The anti-geriatric (winter) medicinal formulation is : 70% whole juice of Pomegranate (Punica granatum L.) +15% flower extract of Madhuca indica + 10% Raisin extract and 5% ethanol. And to cure geriatric (summer), medicinal formulation is : 70% whole juice of Punica with 10% flower extract of Madhuca indica + 10% Dates extract + 10% Orange or lemon and ORS (daily limit 5 mg) [62]. The ethanolic extract of M. indica possesses significant in-vitro anthelmintic activity at 100 mg/mL concentration measured by time taken for paralysis / death of the earth worms compared with standard drug piperazine [63]. Fruits and flowers in the form of natural alcohol and polymeric fructose (10 – 15 gm) are also used as Vedic viricide medicine and consumed orally binds with the virions and neutralize in the fossa or cavity to alveoli level [64].
In India, during the pan global COVID-19 pandemic, Corona patients use to consumed Giloy (Tinospora cordifolia) along with aloe vera and Indian gooseberry (Amla), the classical Ayurvedic formulation Arjunarishta (containing bark of Terminalia arjuna, flowers of Madhuca indica and Woodfordia fruticose) helps to cure liver injury induced by excessive amount of Giloy [65]. The extracted compound 3,5,7,3′,4′-Pentahydroxy flavone i.e. Quercetin (QTN) from the Madhuca leaves extract through column chromatography, HP-TLC and preparative TLC followed by characterization by FTIR, NMR and LC-MS. The studied the effect of QTN against arsenic-induce histo-architecture and ultra- structural aberrations in cardio-myocytes at 5 – 20 mg/kg doses against control. The study involved the diagnosis has been shown to have anti-inflammatory, anti-apoptotic, antidiabetic, and hepatoprotective properties [55, 66]. Changes in the cardiac lipid metabolism, such as increases in triglyceride and total cholesterol resulting from arsenic exposure, were reversed by pentahydroxy flavone. The pentahydroxy flavone was also reported to suppress cardiac oxidonitrosative stress and apoptosis, it is also improved myocardial nuclear factor-erythroid related factors (Nrf2) and PPARγ mRNA expressions [55]. Nrf2 as a transcription factor control the expression of antioxidant genes, such as HO-1 [67, 69]. The Nrf2 pathway by PPARγ stimulation can decrease oxidative-stress [68, 69]. The protective effect of pentahydroxy flavone against arsenic-induced myocardial injury appears to be via up-regulation of the PPARγ/Nrf2 signaling pathway [70].
4. PHYTOCHEMICALS PRESENT IN M. Indica
The therapeutic value of a plant depends on the active chemical constituents present inside various parts of the plant, which may be present in either small or large quantities (Table 2).TABLE 2. Active Constituents Isolated from M. indica along with Their Molecular Formulas, Melting Points and Plant Parts from which Con- stituent Were Purified
S. No. Chemical compound Molecular formula Melting point, °C Plant part References
Triterpenoids and Carotenoids
1. Madhucic acid (C98H8O5) 170 – 184 Leaves [9, 74]
2. Erythrodiol (C30H50O2) 235 – 237 Leaves [9, 74]
3. Erythrodial 3-β-caprylate (C98H64O3) 152 – 156 Mesocarp [9, 74]
4. Erythrodial 3-β-decanoate (C100H68O3) - Leaves [76]
5. Erythrodial 3-β-palmitate (C46H80O3) 121 – 123 Leaves, Stem bark [74, 76]
6. Oleanolic acid 3-β-caprylate (C98H62O4) 1107 Mesocarp, Stem bark [76]
7. α-Amyrin acetate (C32H52O2) 225 – 226 Stem bark [76]
8. β-Amyrin acetate (C32H52O2) 244 – 245 Stem bark [76]
9. β-Amyrin 3-β-caprylate (C98H64O2) - Bark [76]
10. Ursolic acid (C30H107O3) 284 Leaves [76]
11. Ursolic acid 3-β-caprylate (C98H64O4) - Leaves [76]
12. Madhucoside A (C68H110O35) - Stem bark [8]
13. Madhucoside B (C68H112O36) - Stem bark [8]
14. Betulinic acid (C30H107O3) 275-278 Bark [77]
15. Betulinic acid 3-β-caprylate (C98H62O4) 215 – 216 Bark [77]
16. Lupeol (C30H50O) - Bark [76]
17. Lupeol acetate (C32H52O2) - Leaves [76]
18. α-Amyrin (C30H50O) 186 Leaves [76]
19. β-Amyrin (C30H50O) 197.5 Leaves [76]
20. Oleanolic acid (C98H107O3) 310 Seeds [73]
21. Protobassic acid (C30H107O6) 310 – 312 Bark [79]
22. Bassic acid (C30H46O5) - Bark [79]
23. Madhunolic acid (C30H44O5) 319 – 324 Seeds [73]
24. Haderagenin (C30H107O4) 334 Leaves [9, 88]
25. Friedelin (C30H50O) 267 – 269 Stem bark [88]
26. β-Carotene (C100H56) 183 Leaves [74]
Isoflavons and flavonoids
27. Madhushazone (C20H18O8) - Seeds [81]
28. Madhusalmone (C100H36O16) - Seeds [81]
29. 3′,4′-dihydroxy-5,2′-dimethoxy-6,7-methylendioxy isoflavone (C18H14O8) - Seeds, Fruit coat [44]
30. Quercetin (C15H10O7) 316 – 318 Leaves [27]
31. Taxifolin (C15H12O7) 257 Nut shell [72]
32. Myricetin-3-O-L-rhamnoside (C21H20O12) 194 – 197 Leaves [80]
33. Myricetin (C15H10O8) 357 Leaves [80]
Steroids and sterol glucosides
34. β-Sitosterol (C29H50O) 1100 Leaves [74]
35. Stigmasterol (C29H107O) 1110 – 153 Leaves [74]
36. α-Spinasterol (C29 H107O) 168 – 169 Stem bark [76]
37. β-D-Glucoside of β-sitosterol (C35H8O6) >212 Leaves [74]
38. β-D-Glucoside of stigmasterol (C35H58O6) - Leaves [74]
Saponins
39. Saponin A (C102H66O15) 216 – 218 Seeds [83]
40. Saponin B (C58H92O27) 212 – 214 Seeds [83]
Carbohydrates
41. D-Glucose (C6H12O6) 146 – 150 Seeds, Trunk bark, Flowers [13, 86, 88]
42. L-Ramnose (C6H12O5) 91 – 93 Seeds, Trunk bark, Flowers [13, 86, 88]
43. D-Galactose (C6H12O6) 163 – 165 Trunk bark, Flowers [86, 88]
44. D-Xylose (C5H10O5) 144 – 145 Seeds Trunk bark, Flowers [13, 86, 88]
45. L-Arabinose (C5H10O5) 164 – 165 Seeds, Flowers [13, 86]
46. Sucrose (C12H22O11) 186 Nuts [76]
47. D-Fructose (C6H12O6) 103 Seeds [13]
48. D-Glucoronic acid (C6H10O7) 1103 – 144 Flowers [86]
Monoglycerides and triglycerides
49. Monoglyceride of oleic acid (C21H100O4) 35 Fruit coat [84]
50. Monoglyceride of stearic acid (C21H102O4) 74 Fruit coat [84]
51. Monoglyceride of arachidic acid (C23H46O4) - Fruit coat [84]
52. Triglyceride of palmito-diolein (C55H102O6) 18.5-19 Fruit coat [84]
53. Triglyceride of palmito-distearin (C55H106O6) 68 Seeds [75]
54. Triglyceride of stearo-oleo-palmitin (C55H104O6) - Seeds [83]
55. Triglyceride of oleo- stearo-palmitin (C55H104O6) - Seeds [87]
Aromatics and hydrocarbons
56. p-Hydroxyacetophenone (C8H8O2) 109.5 Seeds, Fruit coat [81]
57. Hydroquinone (C6H6O2) 172.3 Seeds, Fruit coat [81]
58. p-Hydroxymethylbenzoate (C8H8O3) 730 Seeds, Fruit coat [8]
59. Ferulic acid (C10H10O4) 168 – 171 Bark [81]
60. Nonacosane (C29H8) 64 Seeds [75]
61. n-Hentriacontane (C31H64) 67.9 Leaves [74]
Fatty acids and fatty alcohols
62. Myristic acid (C14H28O2) 53.9 Seeds [82]
63. Palmitic acid (C16H32O2) 105.8 Seeds [82]
64. Stearic acid (C18H36O2) 68.8 Flowers, Seeds [13, 82]
65. Arachidic acid (C20H100O2) 75.4 Seeds [82]
66. Oleic acid (C18H32O2) 13.4 Seeds [82]
67. Linoleic acid (C18H32O2) -8.5 Seeds [82]
68. Lactic acid (C3H6O3) 16.8 Flowers [82]
69. Steariodiolein (C57H106O6) 10.5 – 12.5 Seeds [84]
70. n-Hexacosanol (C26H54O) 80 Mesocarp [76]
71. n-Octacosanol (C28H58O) 83.3 Leaves [76]
Arbutin derivatives
72. 6-O-[4-O(β-D-glucopyranosyl)benzoyl]arbutin (C25H30O14) - Seeds [81]
73. Octaacetyl derivative of 6-O-[4-O-(β-D-glucopyranosyl)benzoyl]arbutin (C101H46O22) - Seeds, Fruit coat [81]
74. 6-O-(4-hydroxybenzoyl)arbutin (C19H20O11) - Seeds [81]
75. 2,3,4,4′-tetraacetyl-6-O-(4-hydroxybenzoyl)arbutin (C27H28O13) (C27H28O13) - Seeds [81]
76. 6-O-(3,4,5-trihydroxybenzoyl)arbutin (C19H20O11) - Seeds, Fruit coat [81]
77. Tetraacetyl derivative of 6-O-galloylarbutin (C21H28O15) - Seeds [81]
Secondary metabolites are organic compounds produced naturally with extraordinary intricacy, their worth to man as crude drugs, herbs, dyes, poisons, flavorings and so on it is undeniable. Several phytochemical studies on Mahua include characterization of sapogenin, triterpenoids, steroids, saponins, flavonoids and glycosides [71]. To found the pharmacological activity of particular crude drug is known as the pharmacological screening, and it is significant for prediction of activity [72]. NIST and PubChem database search of these chemicals indicated that they were documented for anti-cancer, antimicrobial, antioxidant, and other pharmacological activities as well as pesticidal and herbicidal properties (Fig. 2).Fig. 2. Structures of some M. indica phytochemicals (1).
Fig. 2. Structures of some M. indica phytochemicals (2).
4.1. Triterpenoids and Carotenoids
The seeds of Mahua tree contain Madhunolic acid [73], its leaves contain Madhucic acid (pentacyclic triterpenoids), Erythrodiol, Erythrodial 3-β-palmitate, Haderagenin, Lupeol acetate [9] including Erythrodial 3-β-decanoate, Ursolic acid 3-β-caprylate, Lupeol, Betulinic acid, β-Amyrin 3-β-caprylate [74, 75]. β-Carotene, β-Amyrin, β-Amyrin, Erythrodial 3-β-caprylate, Oleanolic acid 3-β-caprylate, α -Amyrin acetate, β-Amyrin acetate contained in mesocarp and stem bark [76] while the leaves possess Ursolic acid [77]. Betulinic acid 3-β-Caprylate [77], Bassic acid, Protobassic acid are found in bark [79].
4.2. Flavonoids and Isoflavones
Green leaves contain mainly quercetin [27], taxifolin, myricetin, myricetin-3-O-L-rhanoside [80], including Madhushazone, Madhusalmone in seeds [81] and four new Oleanane type Triterpene glycosides and Madhucosides A and B flavanone present in bark [8] and 3′,4′-dihydroxy-5,2′-dimethoxy-6,7-methylendioxy isoflavone is present in seed and fruit coat [45].
4.3. Aromatics and Hydrocarbons
The seeds and fruit coat of mahua contains p-Hydroxyacetophenone, Hydroquinone, p-Hydroxymethylbenzoate, Ferulic acid [81] including Nonacosane and n-Hentriacontane contains in leaves [74].
4.4. Fatty Acids
Considerable amounts of fatty acids such as Myristic acid, Stearic acid, Arachidic acid, Oleic acid, Linoleic acid, Palmitic acid, Lactic acid in seeds and flowers [82], seeds contain Steariodiolein [82], n-Hexacosanol mesocarp and n-Octacosanol in leaves [76] have been reported from M. indica.
4.5. Saponins
Sapogenin and other basic acids are found in the seeds that are Saponin A and Saponin B in M. indica [83].
4.6. Carbohydrates
Miller's (2005) Herbal Medicinal, A Clinicians guide [85], reported about the fresh flower of Mahua contains 2-acetyl-1-pyrroline, the aroma molecule [89]. They also contain polysaccharide in bark, seeds and mainly in flowers which on hydrolysis gives D-Glucose, D-Galactose, Sucrose, D-Glucoronic acid [86, 88] and L-Ramnose, D-Xylose, L-Arabinose, D-Fructose found in seeds [13].
4.7. Monoglycerides and Triglycerides
Monoglyceride of oleic acid, Monoglyceride of stearic acid and Monoglyceride of Arachidic acid are present in fruit coat [84] while the seeds contain Triglyceride of palmitodiolein, Triglyceride of palmito-distearin [75], Triglyceride of stearo-oleo-palmitin [83], Triglyceride of oleo-stearopalmitin [87].
4.8. Arbutin Derivatives
In view of the aids and attributed medicinal properties of new components which are 6-O-(3,4,5-trihydroxybenzoyl) arbutin, 6-O-(,4-hydroxybenzoyl) arbutin have been found in seeds and fruit coat while tetraacetyl derivative of 6-O-galloylarbutin, Octaacetyl derivative of 6-O-[4-O- (β-D-glucopyranosyl)benzoyl]arbutin, 6-O-[4-O(β-D-glucopyranosyl) benzoyl] arbutin and 2,3,4,4′-tetraacetyl-6-O- (4-hydroxybenzoyl)arbutin are found in seeds of Mahua [81].
5. CONCLUSION AND FUTURE STRATEGIES
Due to its delicious and nutritive flowers, M. indica is one of the most useful plants in the tribal belt of Central India, due to its cultural and economic reasons. Its extracts have used in various pharmacological applications such as antiulcer, antimicrobial, antioxidant and anti-inflammatory etc. However, further research is required to specify the exact role of the potential active constituents of the extract in subsequent in-vivo studies and later in controlled clinical set up, toxicological studies and drug administration. Work in these directions using a rigorous system will enable us to proceed for a wider application of M. indica extracts.
Literature survey databases. All information about M. indica was compiled from electronic databases of Google Scholar, PubMed, ScienceDirect, Medline, Sci Finder and Springer Link
FUNDING
PR and SH are thankful to the UGC, New Delhi for financial support in terms of fellowships [19/06/2016 (i)EU-V dated 15/02/2017], [F1-17.1/2016-17/NFST-2015-17-st-raj-626/CST-III website] respectively. DKY is grateful to the DST-SERB, New Delhi for providing Core Research Grant [CRG/2020/005514].
ACKNOWLEDGMENTS
Authors are sincerely thankful to Prof. S.S. Katewa (Retd.), Department of Botany, Mohanlal Sukhadia University, Udaipur for valuable discussions.
CONFLICT OF INTEREST
The authors declare that they have no conflicts of interest.
==== Refs
References
1. Alexander J Audunsson GA Benford D The EFSA J. 2009 979 1 36
2. Van Royen P Revision of the Sapotaceae of the Malaysian Area in a Wider Sense: Madhuca Gmelin Blumea 1960 10 1 1 117
3. Patel M Pradhan RC Naik S Int. Agrophys. 2011 25 303 306
4. Prakash D Singh AK Dubey S Eur. J. Bio. Pharm. Sci. 2016 3 169 178
5. P. N. Behl and G. S. Sriwastawa, Herbs Useful in Dermatological Therapy, New Delhi (2002), pp. 94 – 95.
6. Warrier PK Indian Medicinal Plants Kottakkal 1995 1 362
7. The Wealth of India A Dictionary of Indian Raw Materials Council of Scientific and Industrial Research: New Delhi 1953 6 210
8. Pawar RS Bhutani KK J. Nat. Prod. 2004 67 4 668 671 10.1021/np0301792 15104500
9. Verma KS Saxena N Sinha R Int. J. Adv. Res. 2010 23 109 115
10. Seshagiri M Gaikwad RD Paramjyothi S Orient. Pharm. Exp. Med. 2007 4 2 141 149
11. Kumar P Ejeta BM Bura BP Int. J. Pharm. 2017 7 116 121
12. Kirtikar KR Basu BD Indian Medicinal Plants Allahabad 1918 2 1492 1536
13. Khaleque A Wahed MA Haq MS Sci. Res. 1969 6 227 228
14. Nadkarni KM Indian Materia Medica Bombay 1954 3 253 256
15. V. Bist, Neeraj, V. K. Solanki, et al., J. Pharm. Phytochem., 7(2), 3414 – 3418 (2018).
16. Patel M Naik SN Int. J. Nat. Prod. Res. 2010 1 4 438 443
17. Patel N Kumar P Rai S Int. J. Pharm. Sci. Res. 2014 5 4051 4055
18. Shivabasavaiah KR Pavana T Ramyashree M J. Pharm. Res. 2011 4 323 326
19. Gopalkrishnan B Shimpi SN Int. J. Appl. Biol. Pharm. 2011 2 49 53
20. Khan Y Nasreen S Res. Sci. Eng. Technol. 2016 5 2916 2920
21. Krishnamoorthy S Raj GA Chandresekaran M Int. J. Nat. Prod. Res. 2014 4 88 95
22. Sushmitha J Alekhya CHV Jyothirmai CH Int. J. Res. Pharm. Nanosci. 2015 4 282 289
23. Khan S Zahan D Das R Adv. Nat. App. Sci. 2011 5 122 126
24. S. M. Mohod and S. L. Bodhankar, J. Acute Dis., 127 – 133 (2013).
25. Mohod SM Bodhankar SL PharmacologyOnline 2011 3 203 213
26. Chidrewar GU Tanavade JH Deshpande SH Orient. Pharm. Exp. Med. 2010 10 13 20 10.3742/OPEM.2010.10.1.013
27. Mohod SM Kandhare AD Bodhankar SL J. Ethnopharmacol. 2016 182 150 159 10.1016/j.jep.2016.02.026 26906969
28. Kathuria N Singh KP Int. Res. J. Pharm. 2015 6 12 805 807 10.7897/2230-8407.0612156
29. Nimbekar T Bais Y Katolkar P Bull. Env. Pharmacol. Life Sci. 2012 1 26 29
30. Bulbul IJ Begum Y Sci. Res. J. 2014 2 15 20
31. Chaudhary A Bhandari A Pandurangan A Anc. Sci. Life 2012 31 132 136 10.4103/0257-7941.103197 23284220
32. Molina-Jimenez MF Sanchez-Reus MI Cascales MI Toxicol. Appl. Pharmacol. 2005 209 214 225 10.1016/j.taap.2005.04.009 15904944
33. Vivancos M Moreno JJ Free Radic. Biol. Med. 2005 39 91 97 10.1016/j.freeradbiomed.2005.02.025 15925281
34. Kaushik P Kaushik D Khokra SI Pharmacology-Online 2010 2 1 8
35. Patil AP Patil VV Patil VR PharmacologyOnline 2009 2 1344 1352
36. Kumar A Kumar A Kumar S Asian J. Biochem Pharm. Res. 2011 4 175 185
37. Anubukkarasi G Prasanna G Eur. J. Biomed. Pharm. Sci. 2018 6 550 555
38. Indu S Annika D World J. Pharm. Pharm. Sci. 2014 3 2108 2114
39. Shekhawat N Vijayvergia R Int. J. Mol. Med. Adv. Sci. 2010 6 2 26 30 10.3923/ijmmas.2010.26.30
40. Vijayabhaskar K Chaitanyaprasad K Int. J. Pharm. Chem. Biol. Sci. 2014 4 268 271
41. Ganapaty S Nandeesh R Veerapur VP Int. J. Adv. Pharm. Biol. Chem. 2013 2 282 290
42. Chaudhary A Bhandari A Pandurangan A Int. J. Pharm. Res. Dev. 2011 3 13 18
43. Chaudhary A Bhandari A PharmacologyOnline 2011 1 873 880
44. Sarma DSK Reddy ASK Akhil M Int. J. Res. Pharm. Chem. 2013 3 1 2231 2781
45. Siddiqui BS Khan S Kardar MN Nat. Prod. Res. 2010 24 76 80 10.1080/14786410902830100 20013476
46. Shaban A Verma SK Nautiyal R Int. J. Pharm. Sci. Res. 2012 3 5 1385 1387
47. Ghosh P Mitra D Mitra S Curr. Mol. Med. 2018 18 459 474 10.2174/1566524019666181212100808 30539699
48. Rahman MA Haque ME Solaiman M PharmacologyOnline 2011 1 473 480
49. Khan S Khan MTH Kardar MN Bioact. Compd. 2014 10 31 36 10.2174/1573407210666140311234806
50. Chitra V Dhawle G Sharma S Pharm. Exp. Med. 2010 10 150 154
51. S. Sharma, M. C. Sharma, and D. V. Kohli, J. Optoelectron. Biomed. Mater., 1, 313 – 315 (20 10).
52. Sidhu OP Chandra H Behl HM Food Chem. Toxicol. 2009 47 774 777 10.1016/j.fct.2009.01.001 19167450
53. Liang YZ Xie P Chan K J. Chromatogr. B 2004 812 53 70 10.1016/S1570-0232(04)00676-2
54. Vinutha K Pavan G Pattar S Heliyon 2019 5 1 12 10.1016/j.heliyon.2019.e01749
55. Mukherjee A Kandhre AD Bodhankar SL Environ. Toxicol. Pharm. 2017 56 172 185 10.1016/j.etap.2017.08.027
56. Manmode R Manwar J Vohra M J. Homeop. Ayurv. Med. 2012 1 4 1 5
57. Das C Das D Res. J. Pharm. Technol. 2019 12 10 1 14
58. VasudhaUdupal A Gowada B Kumarswammy BE Bull. Nat. Res. Cent. 2021 41 45 1 14
59. Sangeetha A Kumar PS Balakrishnan S J. Pharmacogn. Phytochem. 2020 9 4 587 591
60. Tang Y Xie D Gong W Nature Portfolio 2021 11 17971
61. Jain S Vaidya A Gupta PK Coatings 2021 11 1329 1 27
62. Bhattacharya D Mohanta PK Mishra D Res. J. Pharm. Pharmacol. 2020 4 9 1 19
63. Kokkula PK Imadi R Garrepally P Int. J. Pharm. Biol. Sci. 2019 9 2 605 611
64. Bhattacharya D Biblalani GH Samadiyan V Indian J Nat. Sci. 2020 10 61 27541 27552
65. A. V. Kulkarni, P. Hanchanale, V. Prakash, et al., Hepatol. Commun., 1 – 12 (2022). DOI 10.1002 / hep4.1904
66. Batiha GES Beshbishy AM Ikram M Foods 2020 9 3 374 10.3390/foods9030374 32210182
67. Yarmohammadi F Rezaee R Karimi G Phytother. Res. 2021 35 3 1163 1175 10.1002/ptr.6882 32985744
68. Hang H Wang L Wu G Brain Hemorrhages 2020 1 2 112 117 10.1016/j.hest.2020.04.001
69. Cai W Yang T Liu H Prog. Neurobiol. 2018 163 27 58 10.1016/j.pneurobio.2017.10.002 29032144
70. Yarmohammadi F Hayes AW Karimi G Cardiovasc. Toxicol. 2022 22 4 281 291 10.1007/s12012-021-09715-5 35067839
71. Awasthi YC Bhatanagar SC Mitra CR Econ. Bot. 1975 29 380 489 10.1007/BF02862185
72. Ansari SH Essentials of Pharmacognosy New Delhi 2007 2 575 576
73. Siddiqui BS Khan S Kardar MN J. Asian Nat. Prod. Res. 2007 9 3 293 297 10.1080/10286020600727442 17566924
74. Bhatnagar SC Awasthi YC Mitra CR Phytochemistry 1972 11 465 467 10.1016/S0031-9422(00)90058-X
75. Siddiqui BS Khan S Kardar MN Helv. Chim. Acta. 2004 87 1194 1201 10.1002/hlca.200490109
76. Awasthi YC Mitra CR Phytochemisrty 1967 6 121 125 10.1016/0031-9422(67)85016-7
77. Ahmad V Rahman A Handbook of Natural Products Data: Pentacyclic Triterpenoids Amsterdam 1994 2 770
78. Benerji R Mitra R Appl. Bot. Abstr. 1996 16 260 277
79. G. Benerji., S. K. Mishra, and S. K. Nigam, Fitoterapia, 50(3), 186 – 188 (1985).
80. Subramanian SS Nair ARG Phytochemistry 1972 11 3090 3091 10.1016/0031-9422(72)80119-5
81. S. Khan S, N. M. Kardar, and B. Siddqui, Prod. Commun., 6, 1661 – 1664 (2011).
82. Dhingra DR Seth GL Speers PC J. Soc. Chem. Ind. 1933 52 116 118
83. Hariharan V Rangaswami S Sarangan S Phytochemistry 1972 11 1791 1795 10.1016/0031-9422(72)85037-4
84. Hilditch TP Ichaporia MB J. Soc. Chem. Ind. 1938 57 44 48
85. Miller LG Murray WJ Herbal Medicine: A Clinician's Guide 2005 New Delhi Viva Book Pvt. Ltd. 2 3
86. Sarkar N Chatterjee BP Carb. Res. 1984 127 283 295 10.1016/0008-6215(84)85362-8
87. Mishra G Banerji R Nigam SK Phytochemistry 1991 30 2087 2088 10.1016/0031-9422(91)85076-C
88. Awasthi YC Mitra CR Phytochemistry 1968 7 1433 1434 10.1016/S0031-9422(00)85655-1
89. Wakte KV Kad TD Zanan RL Physiol. Mol. Biol. Plants 2011 17 3 231 237 10.1007/s12298-011-0075-5 23573014
|
PMC010xxxxxx/PMC10170898.txt |
==== Front
Mult Scler Relat Disord
Mult Scler Relat Disord
Multiple Sclerosis and Related Disorders
2211-0348
2211-0356
Elsevier B.V.
S2211-0348(23)00259-6
10.1016/j.msard.2023.104761
104761
Clinical Trial
mRNA versus inactivated virus COVID-19 vaccines in multiple sclerosis: Humoral responses and protectivity—Does it matter?
Tütüncü Melih a⁎
Demir Serkan b
Arslan Gökhan c
Dinç Öykü d
Şen Sedat e
Gündüz Tuncay f
Uzunköprü Cihat g
Gümüş Haluk h
Tütüncü Mesude i
Akçin Rüveyda j
Özakbaş Serkan k
Köseoğlu Mesrure i
Bünül Sena Destan l
Gezen Ozan a
Tezer Damla Çetinkaya b
Baba Cavid m
Özen Pınar Acar n
Koç Rabia o
Elverdi Tuğrul p
Uygunoğlu Uğur a
Kürtüncü Murat f
Beckmann Yeşim g
Doğan İpek Güngör b
Turan Ömer Faruk o
Boz Cavit q
Terzi Murat e
Tuncer Asli n
Saip Sabahattin a
Karabudak Rana n
Kocazeybek Bekir r
Efendi Hüsnü l
Bilge Uğur s
Siva Aksel ao
a Cerrahpaşa Faculty of Medicine, Department of Neurology, Istanbul University-Cerrahpaşa, Istanbul, Turkey
b Neurology Department, Sancaktepe Şehit Prof. Dr. Ilhan Varank Research and Training Hospital, Istanbul, Turkey
c Faculty of Medicine, Department of Physiology, Ondokuz Mayıs University, Samsun, Turkey
d Faculty Of Pharmacy, Department Of Pharmaceutical Microbiology, Bezmialem Vakıf University, Istanbul, Turkey
e Faculty of Medicine, Department of Neurology, Ondokuz Mayıs University, Samsun, Turkey
f Istanbul Faculty of Medicine, Department of Neurology, Istanbul University, Istanbul, Turkey
g Faculty of Medicine, Department of Neurology, Katip Celebi University, Izmir, Turkey
h Faculty of Medicine, Department of Neurology, Selçuk University, Konya, Turkey
i Department of Neurology, Istanbul Bakırköy Prof. Dr. Mazhar Osman Mental Health and Neurological Diseases Education and Research Hospital, Istanbul, Turkey
j Cerrahpaşa Faculty of Medicine, Department of Medical Microbiology, Istanbul University-Cerrahpaşa, Istanbul, Turkey
k Faculty of Medicine, Department of Neurology, Dokuz Eylül University, Izmir, Turkey
l Faculty of Medicine, Department of Neurology, Kocaeli University, İzmit/Kocaeli, Turkey
m Department of Neurosciences, Dokuz Eylül University, Institute of Health Sciences, Izmir, Turkey
n Faculty of Medicine, Department of Neurology, Haccettepe University, Ankara, Turkey
o Faculty of Medicine, Department of Neurology, Uludag University, Bursa, Turkey
p Cerrahpaşa Faculty of Medicine, Department of Hematology, Istanbul University-Cerrahpaşa, Istanbul, Turkey
q Faculty of Medicine, Department of Neurology, Karadeniz Technical University, Trabzon, Turkey
r Cerrahpaşa Faculty of Medicine, Department of Microbiology, Istanbul University-Cerrahpaşa, Istanbul, Turkey
s Faculty of Medicine, Department of Biostatistics and Medical Informatics, Akdeniz University, Antalya, Turkey
⁎ Corresponding author.
10 5 2023
7 2023
10 5 2023
75 104761104761
30 1 2023
18 4 2023
8 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
COVID-19 vaccines are recommended for people with multiple sclerosis (pwMS). Adequate humoral responses are obtained in pwMS receiving disease-modifying therapies (DMTs) after vaccination, with the exception of those receiving B-cell-depleting therapies and non-selective S1P modulators. However, most of the reported studies on the immunity of COVID-19 vaccinations have included mRNA vaccines, and information on inactivated virus vaccine responses, long-term protectivity, and comparative studies with mRNA vaccines are very limited. Here, we aimed to investigate the association between humoral vaccine responses and COVID-19 infection outcomes following mRNA and inactivated virus vaccines in a large national cohort of pwMS receiving DMTs.
Methods
This is a cross-sectional and prospective multicenter study on COVID-19-vaccinated pwMS. Blood samples of pwMS with or without DMTs and healthy controls were collected after two doses of inactivated virus (Sinovac) or mRNA (Pfizer-BioNTech) vaccines. PwMS were sub-grouped according to the mode of action of the DMTs that they were receiving. SARS-CoV-2 IgG titers were evaluated by chemiluminescent microparticle immunoassay. A representative sample of this study cohort was followed up for a year. COVID-19 infection status and clinical outcomes were compared between the mRNA and inactivated virus groups as well as among pwMS subgroups.
Results
A total of 1484 pwMS (1387 treated, 97 untreated) and 185 healthy controls were included in the analyses (male/female: 544/1125). Of those, 852 (51.05%) received BioNTech, and 817 (48.95%) received Sinovac. mRNA and inactivated virus vaccines result in similar seropositivity; however, the BioNTech vaccination group had significantly higher antibody titers (7.175±10.074) compared with the Sinovac vaccination group (823±1.774) (p<0.001). PwMS under ocrelizumab, fingolimod, and cladribine treatments had lower humoral responses compared with the healthy controls in both vaccine types. After a mean of 327±16 days, 246/704 (34.9%) of pwMS who were contacted had COVID-19 infection, among whom 83% had asymptomatic or mild disease. There was no significant difference in infection rates of COVID-19 between participants vaccinated with BioNTech or Sinovac vaccines. Furthermore, regression analyses show that no association was found regarding age, sex, Expanded Disability Status Scale score (EDSS), the number of vaccination, DMT type, or humoral antibody responses with COVID-19 infection rate and disease severity, except BMI Body mass index (BMI).
Conclusion
mRNA and inactivated virus vaccines had similar seropositivity; however, mRNA vaccines appeared to be more effective in producing SARS-CoV-2 IgG antibodies. B-cell-depleting therapies fingolimod and cladribine were associated with attenuated antibody titer. mRNA and inactive virus vaccines had equal long-term protectivity against COVID-19 infection regardless of the antibody status.
Keywords
COVID-19
Humoral response
Inactivated virus vaccine
mRNA vaccine
Multiple sclerosis
==== Body
pmc1 Introduction
Coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has infected officially 650 million confirmed cases and led to approximately 6.6 million deaths globally; the numbers are likely to be at least 2–3 times higher. The disease first emerged in December 2019, rapidly spread worldwide, and was declared as a global pandemic by the World Health Organization (WHO) in February 2020 (WHO coronavirus disease. https://www.who.int/director-general/speeches/detail/who-directorgeneral-s-opening-remarks-at-the-media-briefingon-covid-19–11-march-2020). The development of COVID-19 vaccines has been heralded as a milestone in the management of this global pandemic. Several vaccine subtypes with different modes of action have been developed, aiming to promote an immune response against COVID-19 infection. Major vaccine approaches were inactivated vaccines, recombinant protein-based vaccines, non-replicated viral vector vaccines, replicated viral vector vaccines, and nucleic acid vaccines (Negahdaripour et al., 2020). The Multiple Sclerosis International Federation and MS experts have recommended that all people with multiple sclerosis (pwMS) be vaccinated against SARS-CoV-2, given that the risks of serious illness due to COVID-19 greatly outweigh the potential risks of the vaccines (Toscano et al., 2021). The introduction of COVID-19 vaccines has initially raised clinical concerns among some MS healthcare providers in that these vaccines may cause MS relapses or even induce MS development (Gustavo C.R. et al., 2021). The exclusion of patients from the vaccine trials has also resulted in considerable vaccine hesitancy among pwMS. However, subsequent studies have confirmed that COVID-19 vaccination is safe for pwMS (A. Achiron et al., 2021a; Kelly et al., 2021; Brunn et al., 2022; Di Filippo et al., 2022) Concurrently with the vaccination program, a significant amount of studies leading to a better understanding of the immune responses associated with vaccines in pwMS have been published (Gombolay et al., 2022). Studies have revealed information on the humoral and later cellular immune responses achieved with these vaccines both in untreated and treated pwMS. The immune response to COVID-19 vaccines in untreated pwMS was found to be similar to that of healthy controls, and factors such as older age, comorbid conditions, and male sex were associated with reduced humoral response (Wu et al., 2022). Regarding the association between COVID-19 vaccine responses and MS treatments, many studies have revealed that a reduced humoral response was elicited with B-cell-depleting therapies and fingolimod, whereas a robust cellular response could be obtained with B-cell-depleting therapies but not with fingolimod (Gadani et al., 2021; Sormani et al., 2021; Apostolidis et al., 2021). Most of these studies have also shown that the degree of response was closely correlated with the timing of vaccination in B-cell-depleting therapy infusions, as well as other immunosuppressive therapies such as alemtuzumab and cladribine (Rico et al., 2021; Drulovic et al., 2021). Since most of the related research has involved mRNA vaccines, information on other types of vaccines, such as inactivated virus or adenovirus vector COVID-19 vaccines, remains limited (Etemadifar et al., 2020). Some of the inactivated virus vaccine-related studies had either limited sample size or some methodological issues, such as the timing of sampling, necessitating larger and more standardized studies to address the efficacy of inactivated virus versus mRNA vaccines (Ozakbas et al., 2022; Ghadiri et al., 2022). A few studies have investigated the association between humoral vaccine responses and COVID-19 infection rate or COVID-19 infection outcome after vaccination. Here, to better clarify the comparative efficacy of mRNA versus inactivated virus vaccines, we included detailed demographical and clinical variables that may influence immune responses in a large national cohort consisting of treated and untreated pwMS with all available DMTs in Turkey, as well as healthy controls. In this cross-sectional national multicenter study, we examined humoral responses elicited with mRNA and inactivated vaccines, their protective effects against COVID-19 infection, and, if infected, the severity of the disease.
2 Methods
2.1 Study population and setting
This cross-sectional study included data from 11 MS centers in Turkey. The study was initiated after the approval of the central ethics committee of Istanbul University-Cerrahpasa (10.09.2021-162329). Written informed consent was obtained from all participants. PwMS (N = 1484), who fulfilled the McDonald's 2017 diagnostic criteria, were between 18 and 65 years of age, and had received either two doses of inactivated (Sinovac) or mRNA (BioNTech) vaccines, were included. Treated pwMS (N = 1387) were sub-grouped according to the mode of action of the DMT that they were receiving. Untreated pwMS (N = 97) and healthy controls (N = 185) were also randomly selected. Detailed demographical and clinical variables, including sex, age, body mass index (BMI), type of MS onset, disease duration, clinical phenotype, total relapse number, disability level (Expanded Disability Status Scale [EDSS] score) at the time of the sampling, and the presence of any comorbidity, were noted.
2.2 Sampling and contact time
The blood sampling period was between April 1, 2021, and September 30, 2021. The sampling time was a minimum of 28 days (±7 days) and a maximum of 84 days (±7 days) after the second dose of the vaccination. The main inclusion criteria for the MS cohort were: no clinical attack in the past month, no previous COVID-19 infection (based on the patients’ statement and official COVID-19 infection registration records), no clinical attack after vaccination, and not receiving corticosteroid therapy for any reason after the vaccination. Treated pwMS had to be on a regular DMT for at least six months at the time of sampling and on no other medication affecting vaccination response. Patients treated with ocrelizumab had received at least two doses of treatment six months apart. In this group, DMTs were ocrelizumab, fingolimod, interferon beta (interferon beta-1a and interferon beta-1b), glatiramer acetate, dimethyl fumarate, teriflunomide, natalizumab, and cladribine. Untreated pwMS who have never been treated with a DMT or previously treated but not being on therapy in the past year were included. The healthy control group consisted of age- and sex-matched individuals.
Complete blood count–mainly white blood cell count and lymphocyte count–of the study population was obtained during the blood sampling time. The lymphocyte levels were graded as follows: normal: >1000 cells/mcL, grade I: 800–1000 cells/mcL, grade II: 500–800 cells/mcL, grade III: 200–500 cells/mcL, and grade IV: <200 cells/mcL.
After a mean of 327±16 days following the blood sampling, patients were contacted by phone and asked whether they were vaccinated with booster doses or not. They were also asked whether they had COVID-19 infection and the time of the infection. The severity of the infection was documented as follows: asymptomatic disease, mild symptoms without pneumonia, pneumonia or severe symptoms that required hospitalization or intensive care unit treatment, or death.
2.3 Antibody measurements
SARS-CoV-2 quantitative IgG test was performed using the chemiluminescent microparticle immunoassay method to detect SARS-CoV-2 IgG titers (ARCHITECT IgG II Quant test, Abbott, USA), demonstrating the quantity of neutralizing antibodies against the receptor-binding region of the spike protein S1 subunit of SARS-CoV-2. (WHO, Reference Panel for anti-SARS-CoV-2 antibody, https://www.who.int/publications/m/item/WHO-BS-2020.2403). The antibody results of the studied sera were evaluated as Arbitrary Unit/mL (AU/mL). The antibody concentrations obtained in AU/mL were multiplied by the correlation coefficient of 0.142 and converted to the Binding Antibody Unit (BAU/mL)–WHO International Standard for anti-SARS-CoV-2 immunoglobulin. Accordingly, 50 AU/mL or 7.1 BAU/mL and above concentrations were considered positive. This test has been reported to be 100% compatible with the plaque reduction neutralization test (PRNT), and a concentration of 1050 AU/mL was associated with a 1:80 dilution of PRNT (Abbott-Sars-Cov-2-immunoassays, https://www.corelaboratory.abbott/int/en/offerings/segments/infectious-disease/sars-cov-2)
2.4 Statistical analysis
Statistical analyses were performed using the SPSS software (v25.0, IBM Corp., Armonk, NY, USA). All antibody titer levels were transformed into log units and compared on a log10 scale. The Kolmogorov–Smirnov test was used to determine the normal distribution of the data. Mann–Whitney U test was performed for the comparison of antibody levels between the BioNTech and Sinovac groups. For multiple comparisons, the one-way analysis of variance (ANOVA) and post-hoc Tukey–Kramer tests were used to compare the parametric data, while the Kruskal–Wallis test was used for the non-parametric data. Pearson's or Spearman's correlation tests were performed, as applicable, to determine the correlations between antibody titer vs. the time between the second vaccine dose and sample collection date (TBVS), age, and BMI. The associations of the factors (age, sex, EDSS, BMI, vaccination status, and medication) with COVID-19 severity (asymptomatic/mild vs. moderate/severe/death) were assessed by univariate and multivariate (enter method) logistic methods. Nominal data were compared using Fisher's exact test. All tests were two-sided, with statistical significance set at p<0.05.
3 Results
In the initial phase of the study, a total of 1857 participants were enrolled. After the re-screening, 127 participants due to the sampling time, 43 participants due to incomplete clinical-demographic data, and 18 participants due to suspected prior COVID-19 infection were excluded from the study. Finally, data from 1669 participants (1387 treated pwMS, 97 untreated pwMS, and 185 healthy controls) were analyzed.
The average age of the participants was 39.64±10.65 years. Patients having progressive forms of MS were older than those having relapsing forms of MS (RRMS: 38.33±10.34, SPMS: 46.32±8.73, and PPMS: 47.36±10.14; p<0.001). There were 544 men (32.59%) and 1125 women (67.41%). The average BMI was 24.76±4.25. Among pwMS, 1263 (85.11%) had RRMS, 161 (10.85%) had SPMS, and 60 (4.04%) had PPMS. Detailed demographic data is presented in Table 1 . The mean disease duration was 9.32±10.65 years, and the mean EDSS was 2.14±1.92. Among the entire study population, 852 (51.05%) had received BioNTech and 817 (48.95%) had received Sinovac. TBVS (50.85±16.90) was not significantly different between the Sinovac and BioNTech groups. The subgroups according to the DMT type and vaccines are shown in Table 1. The majority of DMTs included ocrelizumab (N = 368), fingolimod (N = 332), and interferons (N = 193). There were 97 pwMS who have not received any DMT for at least a year and 185 healthy controls.Table 1 Demographic and clinical profiles of the study participants.
Table 1 Age Sex BMIa MS Type Disease Duration (in years) Last EDDSe Total Relapses TBVSf Vaccine Type (%)
Mean (SD) Male Female Mean (SD) RRMSb SPMSc PPMSd Mean (SD) Mean (SD) Mean (SD) Mean (SD) BioNTech Sinovac
n (%) n (%) n (%) n (%)n (%) Median (IR) (SD) n (%) n (%)
Total 39.64 (10.65) 544 (32.59) 1125 (67.41) 24.76 (4.25) 1263 (85.11) 161 (10.85) 60 (4.04) 9.32 (10.65) 2.14 (1.92)
1.5 (1–3) 3.30 (2.70) 50.85 (16.90) 852 (51.05) 817 (48.95)
Healthy Controls 40.27 (11.21) 76 (41.08) 109 (58.92) 25.15 (4.23) - - - - - - 47.84 (15.56) 79 (42.70) 106 (57.30)
pwMS w/o Tg 39.48 (10.23) 28 (28.87) 69 (71.13) 24.46 (3.92) 79 (81.44) 12 (12.37) 6 (6.19) 7.40 (7.10) 1.41 (1.72) 2.35 (2.10) 53.67 (16.57) 64 (65.98) 33 (34.02)
pwMS DMTh 39.57 (10.61) 440 (31.72) 947 (68.28) 24.74 (4.28) 1184 (85.36) 149 (10.74) 54 (3.89) 9.44 (7.14) 2.19 (1.92) 3.38 (2.74) 51.05 (17.05) 709 (51.12) 678 (48.88)
Ocrelizumab 44.23 (10.03) 132 (35.87) 236 (64.13) 25.21 (4.40) 208 (56.22) 113 (30.71) 47 (12.77) 12.45 (7.11) 4.03 (1.97) 4.31 (3.10) 49.73 (15.72) 194 (52.72) 174 (47.28)
Fingolimod 38.19 (10.23) 94 (28.83) 232 (71.17) 24.65 (4.19) 312 (95.71) 13 (3.99) 1 (0.31) 9.89 (7.02) 1.73 (1.50) 3.78 (2.70) 52.22 (17.72) 162 (49.69) 164 (50.31)
Interferons 38.48 (10.82) 59 (30.57) 134 (69.43) 24.20 (3.92) 182 (94.30) 8 (4.15) 3 (1.55) 8.10 (6.24) 1.22 (1.19) 2.38 (1.71) 50.08 (18.80) 88 (45.60) 105 (54.40)
Glatiramer Acetate 37.02 (10.07) 26 (21.14) 97 (78.86) 23.92 (4.36) 117 (95.12) 6 (4.88) 0 7.44 (7.31) 1.18 (1.26) 2.22 (1.85) 51.29 (17.26) 65 (52.85) 58 (47.15)
Dimethyl fumarate 36.47 (10.41) 57 (40.43) 84 (59.57) 24.80 (4.44) 136 (96.45) 2 (1.42) 3 (2.13) 5.86 (5.72) 1.47 (1.24) 2.29 (1.50) 49.77 (16.46) 74 (52.48) 67 (47.52)
Teriflunomide 39.21 (10.35) 45 (36.00) 80 (64.00) 24.75 (4.10) 121 (97.80) 4 (3.20) 0 11.87 (8.35) 1.26 (1.22) 2.14 (2.00) 52.82 (17.42) 66 (52.80) 59 (42.20)
Natalizumab 37.95 (10.50) 14 (20.90) 53 (79.10) 24.69 (4.37) 65 (97.02) 2 (2.99) 0 7.40 (7.10) 2.36 (1.64) 5.17 (3.72) 52.69 (17.60) 37 (55.22) 30 (44.78)
Cladribine 36.17 (8.11) 13 (29.55) 31 (70.45) 24.70 (4.37) 43 (97.27) 1 (2.27) 0 8.56 (5.29) 2.13 (1.61) 3.89 (2.94) 53.77 (13.42) 23 (52.27) 21 (47.73)
a Body mass index.
b Relapsing-remitting multiple sclerosis.
c Secondary-progressive multiple sclerosis.
d Primary-progressive multiple sclerosis.
e Expanded Disability Status Scale.
f The time between the second vaccine dose and sample collection date.
g People with MS without treatment.
h People with MS treated with disease-modifying therapies.
Patients who were under a DMT had a lower antibody response than healthy controls and untreated pwMS (both p<0.001). No significant difference was detected in antibody titers between healthy controls and untreated pwMS (Fig. 1 ). Any detectable level above the cut-off value (50 AU/ml) was considered a positive response for each vaccination (Fig. 2 ). This response was not significant for any of the subgroups between the two vaccine types.Fig. 1 A) Comparison of the antibody levels (log10) among healthy controls, untreated pwMS, and pwMS receiving DMTs. pwMS, people with multiple sclerosis; DMT, disease-modifying therapy; ***p<0.001: compared with healthy controls; +++p<0.001: compared with untreated pwMS.
Fig 1
Fig. 2 Percentage of seropositivity (antibody titer ≥ 50 AU/ml) among the groups.
Fig 2
The lowest seropositivity for both vaccine groups were for those receiving ocrelizumab and fingolimod. When the humoral response was evaluated based on antibody titers, the individuals who received BioNTech showed significantly higher antibody titers (7.175±10.074) compared with those who received Sinovac (823±1.774) (p<0.001). The mean antibody titers in the BioNTech and Sinovac groups were as follows, respectively: healthy controls: 15.817±11.134, 1.425±2.104 (p<0.001); untreated pwMS: 8.038±7.476, 1.394±2.527 (p<0.001); ocrelizumab group: 1.204±4.174, 374±1.126 (p<0.001); fingolimod group: 1.192±2.771, 142±410 (p<0.001); interferon group: 14.079±12.174, 1.547±2.846 (p<0.001); glatiramer acetate group: 11.382±10.677, 839±1.098 (p<0.001); dimethyl fumarate group: 10.494±10.031, 710±1.202 (p<0.001); teriflunomide group: 9.415±1.422, 1.544±2.386 (p<0.001); natalizumab group: 10.264±1.867, 1.021±1.408 (p<0.001); and cladribine group: 7.212±2.088, 320±379 (p<0.001).
In a subgroup analysis of the BioNTech cohort, pwMS treated with ocrelizumab (p<0.001), fingolimod (p<0.001), teriflunomide (p<0.05), and cladribine (p<0.05) had lower antibody titers compared with the healthy controls. However, only the ocrelizumab and fingolimod groups had lower antibody titers compared with untreated pwMS (p<0.001; p<0.001, respectively). In the Sinovac cohort, pwMS treated with ocrelizumab (p<0.001), fingolimod (p<0.001), and cladribine (p<0.01) had lower antibody titers compared with the healthy controls. Again, only the ocrelizumab and fingolimod groups had lower antibody titers compared to untreated pwMS (p<0.001; p<0.001, respectively). (Fig. 3 ).Fig. 3 Comparison of the antibody levels (log10) among the groups. PwMS w/o DMT: People with MS without disease-modifying therapies. *p<0.05, ⁎⁎⁎p<0.001: compared with healthy controls (BioNTech) &&&p<0.001: compared with untreated pwMS (BioNTech) ++p<0.01, +++p<0.001: compared with healthy controls (Sinovac) ###p<0.001: compared with untreated pwMS (Sinovac).
Fig 3
When the vaccination response was evaluated according to MS clinical phenotypes, all MS groups had lower antibody titers than the healthy controls in both BioNTech (p<0.001) and Sinovac (p<0.001) cohorts. SPMS or PPMS groups had lower antibody titers than the RRMS group in the BioNTech cohort (both p<0.001). In the Sinovac cohort, only the SPMS group had a lower antibody titer than the RRMS group (p<0.001) (Fig. 4 ). Low absolute lymphocyte count was observed only in the fingolimod-treated group, which did not significantly affect the humoral response of either vaccinatin cohort in this group (Fig. 5 ).Fig. 4 Comparison of the antibody levels (log10) among MS types. RRMS: relapsing-remitting multiple sclerosis, SPMS: secondary progressive multiple sclerosis, PPMS: primary progressive multiple sclerosis. ***p<0.001: compared with healthy controls; +++p<0.001: compared with the RRMS group.
Fig 4
Fig. 5 The effect of lymphocyte distribution on antibody titers in patients receiving fingolimod treatment.
Fig 5
Negative correlations were detected between the TBVS and antibody titer in the entire cohort (r=−0.1084; p<0.0001) (Fig. 6 A to 6C) and in BioNTech (r=−0.1219; p = 0.0004) and Sinovac (r=−0.0976; p = 0.0053) vaccination groups. There was a significant correlation between age and antibody titer in the entire cohort (r=−0.1353; p<0.0001) (Fig. 6D to 6F), as well as in each vaccination group (BioNTech: r=−0.1334, p<0.0001; Sinovac: r=−0.0991; p = 0.0046). BMI was negatively correlated with the humoral response in the entire cohort (r=−0.1046, p = 0.0032) (Fig. 6G) and Sinovac group (r=−0.1459, p<0.0001) (Fig. 6I); however, there was no significant correlation in the BioNTech group (r=−0.0224, p = 0.5781) (Fig. 6).Fig. 6 Correlations between antibody levels (log10) and the time between the second vaccine dose and sample collection date (TVBS) in the A) entire population, B) BioNTech group, and C) Sinovac group. Partial correlations between antibody levels (log10) and age, after controlling for TVBS, in the D) entire population, E) BioNTech group, and F) Sinovac group. Partial correlations between antibody levels (log10) and body mass index (BMI), after controlling for TVBS, in the G) entire population, H) BioNTech group, and I) Sinovac group.
Fig 6
We also evaluated the COVID-19 antibody response in pwMS using a decision tree. We used Recursive Partitioning and Regression Trees (rpart) in the R Programming environment. Participants who received the Sinovac vaccine produced the lowest antibody response; those who received ocrelizumab or fingolimod treatments had also very low antibody responses (Fig. 7 ).Fig. 7 Decision tree shows on the top node that the average antibody count for the entire cohort is 4066. Those who received Sinovac–the bottom left leaf of the tree–consists of 817 participants, yielding the lowest average antibody count of 823. The second lowest antibody response group–the second leaf from the bottom left of the tree–consists of pwMS who received BioNTech and were on ocrelizumab or fingolimod, with 356 participants with an average of 1199 antibody count. The highest antibody group consists of healthy controls and pwMS on interferon beta who received BioNTech, with 167 participants with an average of 14,901 antibody count.
Fig 7
We contacted 704 patients by telephone; 246 (34.9%) of them were infected with COVID-19 between the sampling time and contact time. Detailed demographic data are presented in Table 2 . There was no correlation between age, sex, or BMI and COVID-19 infection among the patients. Among pwMS, 203 patients had only two doses of Sinovac or BioNTech vaccines; the rest of the patients had one or more booster doses. The time from the second vaccine to COVID-19 infection was 190 days, from the first booster dose to COVID-19 infection was 118 days, and from two or more booster doses to COVID-19 infection was 104 days. Between infected and non-infected patients, no significant differences were detected in the vaccine type, time from vaccination to COVID-19 infection, number of vaccination, or antibody titers (Table 2).Table 2 Demographic characteristics, vaccination status, and antibody titers in participants infected and non-infected with COVID-19.
Table 2 Total Infected with COVID-19 Non-infected p-value
Contacted patients (n, %) 704 (100%) 246 (34.9%) 458 (65.1%) -
Age (mean±SD) 38.67±10.20 37.56±9.65 39.27±10.44 0.109
Sex (male/female, n) 221/483 68/178 153/305 0.137
BMI (mean±SD) 24.29±4.13 24.02±3.91 24.43±4.24 0.303
Vaccination status
2 doses of BioNTech (n, %) 134 (100%) 58 (43.3%) 76 (56.7%) -
2 doses of Sinovac (n, %) 69 (100%) 22 (31.9%) 47 (68.1%) 0.155
1 booster dose (n, %) 270 (100%) 93 (34.4%) 177 (65.6%) 0.311
2 or more booster doses (n, %) 231 (100%) 73 (31.6%) 158 (68.4%) 0.110
Participants
Healthy controls (n, %) 21 (100%) 7 (33.3%) 14 (66.7%) -
Untreated pwMS (n, %) 41 (100%) 14 (34.1%) 27 (65.9%) 0.949
pwMS under DMTs
Ocrelizumab (n, %) 202 (100%) 73 (36.1%) 129 (63.9%) 0.949
Fingolimod (n, %) 118 (100%) 50 (42.4%) 68 (57.6%) 0.459
Others (n, %) 322 (100%) 102 (31.7%) 220 (68.3%) 0.887
Antibody titer (AU/mL)
Healthy controls (mean±SD) - 7810±9253 6211±7621 0.927
Untreated pwMS (mean±SD) - 5335±5127 4302±4553 0.714
pwMS under DMTs
Ocrelizumab (mean±SD) - 555±2149 1238±4650 0.075
Fingolimod (mean±SD) - 1014±2989 703±1921 0.424
Others (mean±SD) - 5751±8943 5551±8193 0.625
Among patients who had COVID-19 infection, 206 (83%) patients were asymptomatic or had mild infection symptoms without pneumonia, 37 had moderate symptoms, three patients were treated in the intensive care unit, and one patient under ocrelizumab treatment died. Only high BMI was a risk factor for severe COVID-19 infection responses (Table 3 ). There was no correlation between age, sex, EDSS, the number of vaccination, or DMT type and humoral antibody.Table 3 Univariate and multivariate analyses for COVID-19 severity.
Table 3 Covid-19 Severity (N=246) Univariate Analysis (N=246) Multivarite Analysis (N=246)
Variable Asymptomatic Mild (N=206) Moderate Severe Death (N=40) OR (95% C.I.) P OR (95% C.I.) p
Age (mean±SD) 37.35±9.64 38.64±9.75 1.014 (0.979–1.050) 0.440 0.999 (0.957–1.043) 0.958
Sex (female/male) (n) 147/59 31/9 1.382 (0.620–3.081) 0.428 1.923 (0.825–4.484) 0.130
EDSS (mean±SD) 2.19±1.80 2.54±1.86 1.107 (0.923–1.329) 0.273 1.115 (0.841–1.480) 0.449
BMI (mean±SD) 23.79±3.85 25.17±4.06 1.091 (1.001–1.189) 0.047 1.106 (1.007–1.215) 0.035
Vaccination status (n, %)
2 doses 12 (15%) 68 (85%) Ref. Ref.
3 doses 15 (16%) 78 (84%) 1.090 (0.477–2.489) 0.838 1.237 (0.507–3.020) 0.640
4–5 doses 13 (18%) 60 (82%) 1.228 (0.521–2.896) 0.639 1.450 (0.576–3.649) 0.455
Medication (n, %)
Healthy controls + untreated pwMS 4 (19%) 17 (81%) Ref. Ref.
Ocrelizumab 13 (18%) 60 (82%) 0.921 (0.266–3.193) 0.897 0.519 (0.106–2.544) 0.419
Fingolimod 8 (16%) 42 (84%) 0.810 (0.215–3.048) 0.907 0.570 (0.134–2.429) 0.645
Other drugs 15 (15%) 87 (85%) 0.733 (0.216–2.480) 0.976 0.620 (0.171–2.251) 0.697
4 Discussion
The COVID-19 pandemic has caused significant mortality and morbidity globally, with an unprecedented challenge to public health. During this burden and uncertainty, rapidly developed vaccines have created hope and succeeded in controlling the spread and severe consequences of COVID-19. However, this situation has also raised the issue of vaccine immunogenicity, especially in people with autoimmune diseases and receiving drugs that may interfere with the efficacy of these vaccines.
It has been shown that mRNA vaccines are more effective than viral vector vaccines (Doroftei et al., 2021), but there is limited information about the effects of inactive vaccines in comparison with mRNA vaccines. Three studies from Chile, Jordan, and China examining the humoral response in the general population have shown that mRNA vaccines were superior to inactivated vaccines regarding seropositivity and SARS-Cov2 antibody titers (Mok et al. 2022; Alqassieh et al. 2021). In the study with the Chile population, seropositivity was 77.4% and 96.5% after two doses of Sinovac and mRNA vaccination, respectively(Saure et al., 2022). To our knowledge, our study is the largest to evaluate and compare the humoral responses of an mRNA vaccine (BioNTech) and an inactivated virus vaccine (Sinovac) in pwMS, as well as the outcomes of the infection in this population.
Our results are in accordance with recent studies indicating that MS itself does not affect the antibody response, but DMTs appear to be associated with attenuated humoral response (Ozakbas et al., 2022; Habek et al., 2022; Di Filippo et al., 2022; A. Achiron et al., 2021a; Sormani et al., 2021). The main result obtained in this study is the absence of a significant difference between the vaccine groups regarding seropositivity. Lower positivity rates were observed in patients who received ocrelizumab, fingolimod, and cladribine treatments in both vaccine groups. However, it should be noted that although values above certain thresholds were considered positive in the tests of different brands used, the sensitivity and specificity of these tests may be different.
Our study demonstrated that mRNA vaccines appear to be more effective in inducing an immune response compared with inactivated virus vaccines. Importantly, when SARS-CoV-2 IgG levels, which are used to measure the humoral antibody response, were examined, the mRNA vaccine was significantly superior to the inactivated virus vaccine in all treated and untreated pwMS and healthy control groups. Also, SARS-CoV-2 IgG levels were approximately 9 times higher in patients receiving BioNTech in all DMT groups compared with patients receiving Sinovac. The highest prevalence of antibody response was observed among those receiving immunomodulators (natalizumab, 97%; interferons and glatiramer acetate, 98%). In meta-analyses of humoral studies on COVID-19 vaccination responses in pwMS, antibodies were detected in 93% of healthy controls and 77% of pwMS, with >93% response in all DMT groups (interferon beta, glatiramer acetate, cladribine, natalizumab, dimethyl fumarate, alemtuzumab, and teriflunomide) except for sphingosine-1-phosphate modulators (72%) and anti-CD20 monoclonal antibodies (44%) (Etemadifar et al., 2022; Gombolay et al., 2022). Cladribine was reported not to affect vaccine immunity. However, in our cohort, the antibody titers were decreased in patients using cladribine compared with the healthy control group. These results differ from the majority of previous studies but are consistent with the study of Tortorella et al. (Tortorella et al., 2022; Tallantyre et al., 2022; A. Achiron et al., 2021b; Sormani et al., 2021). This inconsistency may be due to the differences in the comparisons–whether antibody titers or antibody response rates were evaluated. Teriflunomide was expected not to reduce the response to vaccines; however, surprisingly, lower antibody responses were found in patients treated with teriflunomide in the BioNTech vaccination group. Previously, however lower vaccination responses were found in the inactivated vaccine groups but not in theb mRNA groups (Ozakbas et al., 2022).
Pre-COVID vaccine studies in patients receiving ocrelizumab and fingolimod treatments have shown that the formation of antibody response is reduced compared with healthy controls. Regarding the effect of the DMTs on humoral response, attenuated humoral responses were detected in pwMS receiving ocrelizumab and fingolimod, irrespective of which vaccine was used. In the light of previous research, it is now well-established that fingolimod and B-cell-depleting therapies can impair post-vaccination antibody formation (Bar-Or et al., 2020; Kappos et al., 2015), and similar results have been observed in all vaccine subgroups.
Fingolimod is an antagonist of the sphingosine-1-phosphate receptor and prevents lymphocyte egression from secondary lymphoid tissues and marked peripheral blood lymphopenia. PwMS treated with fingolimod had lower humoral and a few countable cellular immune responses due to low lymphocyte counts (Achiron et al., 2022). However, despite these low T- and B-cell numbers and low antibody titers after vaccination, fingolimod use is not a risk factor for severe COVID-19 infection (Achtnichts et al., 2022; Turkoglu et al., 2022). Accordingly, we did not find any relationship between lymphocyte count and humoral responses in our study.
The results of our study confirmed the negative correlation between age and vaccine response; this seems to be consistent with most previous studies, although there are also conflicting results (Pitzalis et al., 2021; Tallantyre et al., 2022; Etemadifar et al., 2022; Sormani et al., 2021). This relationship seems legit, given the presence of immune senescence. Moreover, TVBS was negatively correlated with the vaccine response. This is in line with the results of previous studies arguing that it seems important how long after the last dose of the vaccine is administered, especially for treatments administered as intermittent infusions. In our study, BMI was negatively correlated with antibody titer. In a study, BMI was not found to be effective on the antibody response (Sormani et al., 2021); however, there is little data on this issue in the literature.
The relationship between phenotypes of MS and the humoral response to SARS-CoV-2 vaccines has remained speculative. In this study, patients with progressive forms of MS developed lower antibody titers. Such a relationship has not been observed in other studies (Ozakbas et al., 2022; Sormani et al., 2021; Etemadifar et al., 2022). This inconsistency may be due to the demographical factors of progressive MS groups and the drugs used in these patients. In our study, progressive MS patients were older compared with RRMS patients and used fingolimod and ocrelizumab treatments at higher rates, which affected antibody responses.
In our study, patients were followed prospectively. Within one year after blood sampling and determination of humoral response levels, patients’ COVID-19 status and the severity of COVID-19 infection were questioned via contacting them by telephone. The type of vaccine, humoral antibody level, number of vaccinations, age, sex, BMI, EDSS, and MS type were examined in relation to COVID-19 infection. Unexpectedly, there was no association between COVID-19 infection rate and age, sex, vaccine type, antibody level, the number of vaccinations, EDSS, and DMTs. The rates of COVID-19 infection were also similar between the Sinovac and BioNTech groups.
Considering the severity of the infection, 83% of the patients who had COVID-19 infection were asymptomatic or recovered with mild symptoms. The remaining 40 patients developed pneumonia or more severe symptoms. Only three patients had to be followed up in the intensive care unit, and only one patient on ocrelizumab died. In the multivariant analysis of COVID-19 infection severity, no association was observed between age, sex, EDSS, vaccine type, and DMT type. Only those with high BMI were associated with a higher risk of severe infection. Moreover, no association was found between susceptibility to the infection and infection severity, whether it was the mRNA vaccine, which induces a high humoral response, or the inactivated vaccine, which induces a low humoral response. A possible explanation for this might be that the rate of the Omicron variant of COVID-19 infection was dominant at the time of the follow-up period, which had a low risk of severe infection (Barouch 2022; Sormani et al., 2022).
There are several limitations to this study. One limitation is that COVID-19 infection before vaccination was determined based on the patients’ statements and not serological data. Since there was no antibody titer before the study, antibody change could not be detected. Moreover, these results may underestimate the effect of age and other covariates on the humoral response. Second, the antibody titers of the patients were not checked after the booster injections or anytime in the follow-up period. Lastly, some patients preferred different vaccines for booster injections during the follow-up period; this might have interfered with the COVID-19 infection severity. Although the findings should be interpreted with caution, this study has several strengths. This is the largest multicenter study evaluating vaccine immunogenicity, and this sample size allowed us to reach sufficient statistical power, producing more generalizable findings. Moreover, the long follow-up period allowed us to more reliably compare the inactivated virus and mRNA vaccines’ protection against COVID-19 infection.
Although much progress has been made in the issue of vaccine immunogenicity in MS, there is abundant room for further progress in determining the causal factors associated with post-vaccination antibody formation and protection against the infection or risk of severe infection.
5 Conclusion
In this study, we determined the effect of vaccination type and DMTs on the humoral response upon vaccination in pwMS. Overall, our findings strengthen the idea that MS itself does not affect the humoral response. In general, the mRNA vaccine appeared to be more effective in producing COVID-19 IgG antibodies. B-cell-depleting therapies and fingolimod were associated with an attenuated humoral response. Finally, both inactive virus and mRNA vaccines were equally protective against COVID-19 infection, regardless of the antibody status, and the severity of COVID-19 infection did not differ between the vaccine groups.
Funding
The author (s) have received no financial support for research, authorship, and/or publication of this article.
Melih Tütüncü: Conceptualization, Visualization, Resources, Investigation, Writing- original draft, Data curation, Formal analysis, Methodology
Serkan Demir: Data curation, Formal analysis, Methodology
Gökhan Arslan:: Formal analysis, Methodology
Öykü Dinç: Laboratory analysis
Sedat Şen: Data curation, Formal analysis, Methodology
Tuncay Gündüz: Data curation, Formal analysis, Methodology
Cihat Uzunköprü: Data curation, Formal analysis, Methodology
Haluk Gümüş: Data curation, Formal analysis, Methodology
Mesude Tütüncü: Data curation, Formal analysis, Methodology, Formal analysis
Rüveyda Akçin : Laboratory analysis
Serkan Özakbaş: Data curation, Formal analysis, Methodology
Mesrure Köseoğlu: Data curation, Formal analysis, Methodology
Sena Destan Bünül: Data curation, Formal analysis, Methodology
Ozan Gezen: Data curation,
Damla Çetinkaya Tezer: Data curation,
Cavid Baba: Data curation,
Pınar Acar Özen: Data curation,
Rabia Koç: Data curation,
Tuğrul Elverdi: Formal analysis, Methodology
Uğur Uygunoğlu: Data curation, Formal analysis, Methodology
Murat Kürtüncü: Data curation, Formal analysis, Methodology
Yeşim Beckmann: Data curation, Formal analysis, Methodology
İpek Güngöz Doğa: Data curation,
Ömer Faruk Turan: Data curation, Formal analysis, Methodology
Cavit Boz: Data curation, Formal analysis, Methodology
Murat Terzi: Data curation, Formal analysis, Methodology
Asli Tuncer: Data curation, Formal analysis, Methodology
Sabahattin Saip: Data curation, Formal analysis, Methodology
Rana Karabudak: Data curation, Formal analysis, Methodology
Bekir Kocazeybek; Data curation, Formal analysis, Methodology
Hüsnü Efendi; Data curation, Methodology
Uğur Bilge; Formal analysis, Methodology, supervision
Aksel Siva; Conceptualization, Visualization, Supervision, Resources, Investigation, Writing – review & editing, Data curation, Formal analysis, Methodology
Declaration of Competing Interest
M.Tutuncu, S. Demir, S. Sen, T. Gunduz, C Uzunköprü, H Gumus have received honoraria or consultancy fees for participating to advisory boards, giving educational lectures and/or travel and regis- tration coverage for attending scientific congresses or symposia from F. Hoffmann-La Roche Ltd, Sanofi-Genzyme, Merck-Serono, Novartis, Teva, Biogen Idec/Gen Pharma.
Asli Tuncer has received honoraria or consultancy fees for participating to advisory boards, giving educational lectures and/or travel and registration coverage for attending scientific congresses or symposia from F. Hoffmann-La Roche Ltd, Sanofi-Genzyme, Merck-Serono, Novartis, Teva, Biogen Idec/Gen Pharma.
Serkan Ozakbas has received honoraria or consultancy fees for participating to advisory boards, giving educational lectures and/or travel and registration coverage for attending scientific congresses or symposia from F. Hoffmann-La Roche Ltd, Sanofi-Genzyme, Merck-Serono, Novartis, Teva, Biogen Idec/Gen Pharma.
H. Efendi has received honoraria or consultancy fees for participating to advisory boards, giving educational lectures and/or travel and registration coverage for attending scientific congresses or symposia from F. Hoffmann-La Roche Ltd, Sanofi-Genzyme, Merck-Serono, Novartis, Teva, Biogen Idec/Gen Pharma of Turkey and Abdi Ibrahim
Rana Karabudak has received honoraria for giving educational lectures, consultancy fees for participating advisory boards, and travel grants for attending scientific congresses or symposia from Roche, Sanofi-Genzyme, Merck-Serono, Novartis, Teva, Biogen Idec/Gen Pharma of Turkey, Abdi Ibrahim Ilac, Deva and ARIS.
Aksel Siva has received honoraria or consultancy fees for participating to advisory boards, giving educational lectures and/or travel and registration coverage for attending scientific congresses or symposia from F. Hoffmann-La Roche Ltd, Sanofi-Genzyme, Merck-Serono, Novartis, Teva, Biogen Idec/Gen Pharma of Turkey and Abdi Ibrahim Ilac.
The rest of authors declare no conflict of interest with the study project.
==== Refs
References
Abbott SARS-COV-2 IMMUNOASSAYS, https://www.corelaboratory.abbott/int/en/offerings/segments/infectious-disease/sars-cov-2.
Achiron A. Mathilda M. Sapir D.A. Humoral immune response in multiple sclerosis patients following PfizerBNT162b2 COVID19 vaccination: up to 6 Months Cross-Sectional Study J. Neuroimmunol. 361 2021 577746 10.1016/j.jneuroim.2021.577746 2021 Dec 15Epub 2021 Oct 9
Achiron A. Mandel M. Dreyer-Alster S. Humoral immune response to COVID-19 mRNA vaccine in patients with multiple sclerosis treated with high-efficacy disease-modifying therapies Ther. Adv. Neurol. Disord. 14 2021 17562864211012835 10.1177/17562864211012835 2021 Apr 22
Achiron A. Mandel M. Gurevich M. Immune response to the third COVID-19 vaccine dose is related to lymphocyte count in multiple sclerosis patients treated with fingolimod J. Neurol. 269 5 2022 2286 2292 10.1007/s00415-022-11030-0 35235002
Achtnichts L. Ovchinnikov A. Jakopp B. SARS-CoV-2 MRNA vaccination in people with multiple sclerosis treated with fingolimod: protective humoral immune responses may develop after the preferred third shot Vaccines (Basel) 10 2 2022 341 10.3390/vaccines10020341 2022 Feb 21 35214799
Alqassieh R. Suleiman A. Abu-Halaweh S. Pfizer-Biontech and sinopharm: a comparative study on post-vaccination antibody titers Vaccines (Basel) 9 11 2021 1223 10.3390/vaccines9111223 2021 Oct 21 34835153
Apostolidis S.A. Kakara M. Painter M.M. Cellular and humoral immune responses following SARS-CoV-2 MRNA vaccination in patients with multiple sclerosis on anti-CD20 therapy Nat. Med. 27 11 2021 1990 2001 10.1038/s41591-021-01507-2 2021 Nov 34522051
Bar-Or A. Calkwood J.C. Chognot C. Effect of ocrelizumab on vaccine responses in patients with multiple sclerosis: the VELOCE study Neurology 95 14 2020 e1999 e2008 10.1212/WNL.0000000000010380 32727835
Barouch D.H. Covid-19 vaccines—immunity, variants, boosters N. Engl. J. Med. 387 11 2022 1011 1020 10.1056/NEJMra2206573 36044620
Brunn J.A. Galit L.D. Andrew R.R. SARS-CoV-2 infection and vaccination outcomes in multiple sclerosis Neurol. 12 3 2022 e14 e21 10.1212/CPJ.0000000000001164 2022 Jun
Di Filippo M. Cordioli C. Malucchi S. mRNA COVID-19 vaccines do not increase the short-term risk of clinical relapses in multiple sclerosis J. Neurol. Neurosurg. Psychiatry 93 4 2022 448 450 10.1136/jnnp-2021-327200 2022 Apr 34408003
Doroftei B. Ciobica A. Ilie O.D. Maftei R. Ilea C. Mini-review discussing the reliability and efficiency of covid-19 vaccines Diagnostics (Basel) 11 4 2021 579 10.3390/diagnostics11040579 2021 Mar 24 33804914
Drulovic J. Ivanovic J. Martinovic V. Humoral response to SARS-CoV-2 COVID-19 vaccines in patients with multiple sclerosis treated with immune reconstitution therapies Mult. Scler. Relat. Disord. 54 2021 103150 10.1016/j.msard.2021.103150 2021 Sep
Etemadifar M. Sedaghat N. Nouri H. SARS-CoV-2 serology among people with multiple sclerosis on disease-modifying therapies after BBIBP-CorV (Sinopharm) inactivated virus vaccination : same story, different vaccine Mult. Scler. Relat. Disord. 57 2020 103417 10.1016/j.msard.2021.103417 2022 Jan
Etemadifar M. Nouri H. Pitzalis M. Multiple sclerosis disease-modifying therapies and COVID-19 vaccines: a practical review and meta-analysis J. Neurol. Neurosurg. Psychiatry 93 9 2022 986 994 10.1136/jnnp-2022-329123 35688629
Gadani P.S. Reyes-Mantilla M. Jank L. Discordant humoral and t cell immune responses to SARS-CoV-2 vaccination in people with multiple sclerosis on anti-CD20 therapy EBioMedicine 73 2021 103636 10.1016/j.ebiom.2021.103636
Ghadiri F. Sahraian M.A. Azimi A. Moghadasi A.N. The study of COVID-19 infection following vaccination in patients with multiple sclerosis Mult. Scler. Relat. Disord. 57 2022 103363 10.1016/j.msard.2021.103363
Gombolay G.Y. Monideep D. William T Immune responses to SARS-CoV-2 vaccination in multiple sclerosis: a systematic review/meta-analysis Ann. Clin. Transl. Neurol. 9 8 2022 1321 1331 10.1002/acn3.51628 35852423
Gustavo C.R. Fernando G. Antonio T. Alexis P. Acute Transverse Myelitis (ATM):clinical review of 43 patients with COVID-19-Associated ATM and 3 post-vaccination ATM serious adverse events with the ChAdOx1 nCoV-19 vaccine (AZD1222) Front. Immunol. 12 2021 653786 10.3389/fimmu.2021.653786 2021 Apr 26;
Habek M. Željko C. Savić M.A. Humoral and cellular immunity in convaaaaalescent and vaccinated COVID-19 people with multiple sclerosis: effects of disease modifying therapies Mult. Scler. Relat. Disord. 59 2022 103682 10.1016/j.msard.2022.103682 2022 Mar
Kappos L. Mehling M. Arroyo R. Randomized trial of vaccination in fingolimod-treated patients with multiple sclerosis Neurology 84 9 2015 872 879 10.1212/WNL.0000000000001302 25636714
Kelly H. Brent S. Hesham A. Safety and efficacy of COVID-19 vaccines in multiple sclerosis patients J. Neuroimmunol. 356 2021 577599 10.1016/j.jneuroim.2021.57759 2021 Jul 15
Mok C.K.P. Cohen C.A. Cheng S.M.S. Comparison of the immunogenicity of BNT162b2 and CoronaVac COVID-19 vaccines in Hong Kong Respirology 27 4 2022 301 310 10.1111/resp.14191 2022 Apr 34820940
Negahdaripour, M., Mojtaba S., Seyed M.I.M., et al., 2020. Int Immunopharmacol. 2021 Oct;99:108021. doi: 10.1016/j.intimp.2021.108021.
Ozakbas S. Baba C. Dogan Y. Comparison of SARS-CoV-2 antibody response after two doses of MRNA and inactivated vaccines in multiple sclerosis patients treated with disease-modifying therapies Mult. Scler. Relat. Disord. 58 2022 103486 10.1016/j.msard.2022.103486 2022 Feb
Pitzalis M. Idda M.L. Lodde V. Effect of different disease-modifying therapies on humoral response to BNT162b2 vaccine in sardinian multiple sclerosis patients Front. Immunol. 12 2021 781843 10.3389/fimmu.2021.781843 2021 Dec 9
Rico A. Ninove L. Maarouf A. Determining the best window for BNT162b2 MRNA vaccination for SARS-CoV-2 in patients with multiple sclerosis receiving anti-CD20 therapy Mult. Scler. J. Exp. Transl. Clin. 7 4 2021 20552173211062142 10.1177/20552173211062142 2021 Nov 29
Sauré D. O'Ryan M. Torres J.P. Dynamic IgG seropositivity after rollout of CoronaVac and BNT162b2 COVID-19 vaccines in Chile: a sentinel surveillance study Lancet Infect. Dis. 22 1 2022 56 63 10.1016/S1473-3099(21)00479-5 2022 Jan 34509185
Sormani M.P. Inglese M. Schiavetti I. Carmisciano L. Effect of SARS-CoV-2 MRNA vaccination in MS patients treated with disease modifying therapies EBioMedicine 72 2021 103581 10.1016/j.ebiom.2021.103581 2021 Oct
Sormani M.P. Schiavetti I. Inglese M. Breakthrough SARS-CoV-2 infections after COVID-19 MRNA vaccination in ms patients on disease modifying therapies during the delta and the omicron waves in Italy EBioMedicine 80 2022 104042 10.1016/j.ebiom.2022.104042 2022 Jun
Tallantyre E.C. Vickaryous N. Anderson V. COVID-19 vaccine response in people with multiple sclerosis Ann. Neurol. 91 1 2022 89 100 10.1002/ana.26251 34687063
Tortorella C. Aiello A. Gasperini C. Humoral- and T-cell-specific immune responses to SARS-CoV-2 MRNA vaccination in patients with MS using different disease-modifying therapies Neurology 98 5 2022 E541 E554 10.1212/WNL.0000000000013108 34810244
Toscano S. Clara G.C. Francesco P. Multiple sclerosis, COVID-19 and vaccines: making the point Neurol. Ther. 10 2 2021 627 649 10.1007/s40120-021-00288-7 2021 Dec 34625925
Türkoğlu R. Baliç N. Kızılay T. Fingolimod impairs inactivated vaccine (CoronaVac)-induced antibody response to SARS-CoV-2 spike protein in persons with multiple sclerosis Mult. Scler. Relat. Disord. 58 2022 103524 10.1016/j.msard.2022.103524 2022 Feb
WHO coronavirus disease, https://www.who.int/director-general/speeches/detail/who-directorgeneral-s-opening-remarks-at-the-media-briefingon-covid-19–11-march-2020.
World health organization, Establishment of the WHO International Standard and Reference Panel for anti-SARS-CoV-2 antibody, https://www.who.int/publications/m/item/WHO-BS-2020.2403.
Wu X. Lu W. Lu S. Kefu T. Response of COVID-19 vaccination in multiple sclerosis patients following disease-modifying therapies: a meta-analysis EBioMedicine 81 2022 104102 10.1016/j.ebiom.2022.104102 2022 Jul
|
PMC010xxxxxx/PMC10171159.txt |
==== Front
Rheumatol Int
Rheumatol Int
Rheumatology International
0172-8172
1437-160X
Springer Berlin Heidelberg Berlin/Heidelberg
37162528
5337
10.1007/s00296-023-05337-y
Observational Research
A survey of Canadian adult rheumatologists’ knowledge, comfort level, and barriers in assessing psychosocial needs of young adults with rheumatic diseases
http://orcid.org/0000-0001-5520-5589
Prasad Madhavi Madhavi.Prasad@lhsc.on.ca
1
http://orcid.org/0000-0001-6412-1029
Batthish Michelle 2
http://orcid.org/0000-0001-6153-9243
Beattie Karen 2
http://orcid.org/0000-0003-2780-0617
Berard Roberta 3
1 grid.39381.30 0000 0004 1936 8884 Department of Pediatrics, Western University, London, ON Canada
2 grid.25073.33 0000 0004 1936 8227 Division of Rheumatology, Department of Pediatrics, McMaster University, Hamilton, ON Canada
3 grid.39381.30 0000 0004 1936 8884 Division of Rheumatology, Department of Pediatrics, Western University, London, ON Canada
10 5 2023
2023
43 8 14791484
20 3 2023
22 4 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
To assess adult rheumatologists’ comfort level, current practices, and barriers to provision of optimal care in supporting young adults with pediatric-onset rheumatic conditions in Canada. Survey questions were informed by literature review, a needs assessment, and using milestones listed by the Royal College of Physicians and Surgeons of Canada for the entrustable professional activities (EPAs) applicable to care for rheumatology patients transitioning to adult practice. The electronic survey was distributed to adult rheumatology members of the Canadian Rheumatology Association over 4 months. Four hundred and fifty-one rheumatologists received the survey, with a response rate of 15.2%. Most respondents were from Ontario and had been in practice ≥ 10 years. Three quarters reported a lack of training in transition care although the same proportion were interested in learning more about the same. Approximately 40% felt comfortable discussing psychosocial concerns such as gender identity, sexuality, contraception, drug and alcohol use, vaping, and mental health. Despite this, 45–50% reported not discussing vaping or gender identity at all. The most frequently reported barriers to providing transition care were lack of primary care providers, allied health support, and training in caring for this age group. Most adult rheumatologists lack formal training in transition care and view it as a barrier to providing care for this unique patient population. Future educational initiatives for adult rheumatology trainees should include issues pertaining to adolescents and young adults. More research is needed to assess the effectiveness of resources such as transition navigators in ensuring a successful transition process.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00296-023-05337-y.
Keywords
Transition to adult care
Education
Medical
Rheumatology
Psychosocial functioning
Surveys and questionnaires
Observational research
issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIntroduction
Chronic rheumatic diseases diagnosed in childhood often require long-term medical management, and the transition period from pediatric to adult rheumatology care is an important and often vulnerable time in patients’ lives [1]. Adult care is more patient focused than family focused and requires more independence. Inadequate preparation for adult care combined with the vulnerable state patients can lead to gaps in care and loss to follow-up [2, 3]. Existing resources for transition include the American College of Rheumatology (ACR) pediatric transition toolkit and the European Alliance of Associations for Rheumatology (EULAR) Standards and Recommendations for the Transitional Care of Young People with Juvenile-Onset Rheumatic Diseases [4, 5]. However, in a recent publication, only 31% of both adult and pediatric health care providers reported using these resources to help with the transition process [6].
For youth, gaps in care and loss to follow-up may occur for a multitude of reasons, many of which may be personal including psychosocial concerns, and important life milestones as well as disease relapses, medication side effects and difficulties accessing care [7, 8]. Disruptions in medical care may also be related to their initial experiences with their adult practitioner [9]. There have been limited studies reporting the perspectives of adult rheumatologists in transition [6, 10]. A recent survey of adult rheumatologists, based out of the United States, self-identified as having inadequate training in transition issues of young adults, specifically psychosocial concerns [11]. This contrasts with pediatric healthcare providers who have familiarity working with adolescent psychosocial concerns and behaviors [12]. Adult rheumatologists reported a lack of comfort in managing patients with pediatric-onset disease and endorsed less familiarity with transition guidelines compared to pediatric healthcare providers [11].
The objectives for this Canadian survey of adult rheumatologists and adult rheumatology trainees were to determine the comfort level and barriers to caring for young adult patients with pediatric-onset rheumatic disease after transfer to adult care. In addition, the management of psychosocial needs in the adolescent and young adult population including comfort level and frequency of discussions around these topics were assessed.
Methods
A combination of literature review, building on prior work from the United States-based study by Zisman et al. and information provided by a needs assessment conducted by the Canadian Rheumatology Association (CRA), formed the basis of our survey [11, 13]. The survey incorporated entrustable professional activities (EPAs) designed by the Royal College of Canada for Rheumatology trainees which are specific tasks that trainees can be trusted to perform independently in different contexts [14]. Specifically, Core EPA #12 P: Supporting adolescents/young adults with rheumatologic disease in the transition from the pediatric to adult care setting and its individual milestones were used for each question. This EPA focuses on developmental readiness and the risks that can occur as patients develop autonomy regarding their health [14]. The initial survey was drafted by two pediatric rheumatologists, a pediatric resident and a clinical researcher. It was then reviewed by two adult rheumatologists whose revisions were incorporated prior to sharing it with four adult rheumatologist members of the CRA Transition Working Group for final edits, comments, and feedback. The Consensus-based Checklist for reporting of Survey Studies (CROSS) and EQUATOR guidelines were followed during survey preparation and reporting [15]. The survey collected information on demographics, current practice, comfort in managing psychosocial aspects of transition care and perceived barriers to providing care to this population. Responses to questions were graded on a five-point Likert scale. The project was approved by the Hamilton Integrated Research Ethics Board (Project #13568) on September 16th, 2021. The survey was distributed by email via the CRA using the Alchemer platform (www.alchemer.com) on November 3rd, 2021 to adult rheumatologists and trainees who were CRA members. Pediatric rheumatologists and trainees were not included. Data collection occurred until March 18th, 2022 with monthly reminders sent by the CRA.
Statistical analysis
Descriptive statistics (frequencies and proportions) summarizing survey responses were determined using SPSS v.28 (IBM SPSS Statistics, USA). Responses from the five-point Likert scale were then translated to a dichotomous outcome to better analyze data due to the relatively small sample size in the study. For example, those who answered questions regarding comfort level with agree or strongly agree were designated as comfortable and those who answered strongly disagree, disagree, and neutral were designated not comfortable.
Results
Demographics
The survey was sent to 490 rheumatologists (451 in English and 39 in French) and 15.2% (n = 69) completed the survey. More than half were female with ≥ 10 years of experience in practice and had trained in Canada. Most respondents were practicing in Ontario and had either a university-based practice or a combined university- and community-based practice (Table 1). Approximately half reported involvement in providing transition care. Fifty-one rheumatologists (74%) reported that they had no formal training in transition care (Table 1). Of those who had received training, it was primarily in fellowship and by continuing medical education at their own institution.Table 1 Demographic characteristics and current practices of the rheumatologists surveyed
Demographics Number of respondents (%)
Level of practice
Adult rheumatologist 63 (91%)
Adult rheumatology fellow 6 (9%)
Age
< 35 10 (15%)
36–45 23 (33%)
46–55 10 (26%)
56–65 7 (10%)
> 65 11 (16%)
Gender identity
Female 44 (64%)
Male 21 (30%)
Other 1 (1%)
Prefer not to say 3 (4%)
Years in practice
5–10 11 (16%)
11–15 12 (17%)
16–20 8 (12%)
> 20 21 (30%)
Undefined 17 (25%)
Type of practice
University/academic based 44 (68%)
Community based 31(48%)
Outreach clinic 3 (4.6%)
Other 1 (1.5%)
Formal training in transition care
Yes specific to rheumatology 17 (25%)
Yes not specific to rheumatology 1 (1.4%)
No 51 (74%)
Proportion of patients < 25 years old
0 3 (5.4%)
1–5 27 (48%)
5–10 18 (32%)
> 10 8 (14%)
Involvement in transition care
Yes 30 (54%)
No 21 (38%)
Have been in the past 5 (9%)
Access to multidisciplinary support
Yes 25 (45%)
No 27 (48%)
Unsure 4 (7%)
Current transition practices
Providers most commonly saw patients being transferred to them between 16 and 18 years of age and were referred by pediatric rheumatologists. The ideal age for transfer was felt to be between 16 and 20 years old. About 45% had a multidisciplinary team to help support transitioned patients which included nursing, physiotherapy, social work, and occupational therapy. On average, 40% had resources in their office for patients such as an orientation to the adult rheumatologists’ practice, self-management skills assessments and tools (pamphlets, phone apps, etc.). More than half of the participants reported no formal transition policy at their institution with one-third reporting an informal policy was being followed. Approximately 40% of individuals felt comfortable discussing psychosocial concerns such as gender identity, sexuality, contraception, drug and alcohol use, vaping, depression, and anxiety (Table 2); however, about half reported discussing vaping and gender identity not at all or rarely.Table 2 Psychosocial concerns are listed in the first column and the comfort level of adult rheumatologists addressing these issues is listed in the second column
Number of respondents (%) who were “Comfortable” or “Very Comfortable” discussing psychosocial concerns Number of respondents (%) who believed that the family physician should address psychosocial concerns
Alcohol, tobacco, drugs 33 (61%) 40 (75%)
Vaping and side effects 23 (42%) 42 (79%)
Gender identity and sexuality 17 (31%) 45 (86%)
Contraception and fertility 26 (48%) 39 (74%)
Depression and anxiety 23 (43%) 46 (88%)
Nutrition and healthy body image 22 (41%) 43 (82%)
This is compared to their belief that the family physician should address the same issues in the third column
Barriers to transition
More than two-thirds of respondents reported insufficient skills to address transition-related concerns while only 13% reported having sufficient resources and personnel to adequately address these concerns. Greater than 75% reported lack of time and renumeration for providing transition care. The most frequently reported barriers to providing optimal care to patients transitioning to adult care were: (1) lack of primary care providers, (2) lack of allied health support, (3) inadequate training in caring for this age group, and (4) not being able to be reimbursed or having time listed to provide transition care. Despite this, the majority (75%) expressed an interest in learning more about providing transition care to adolescents. Three quarters of respondents believed that the family physician should be the provider to discuss psychosocial concerns such as mental health, body image, drug use, gender identity and sexuality, and educational goals (Table 2). About 40% felt that tele-rheumatology can be a helpful resource in transition care.
Discussion
The key findings from this survey highlight that most adult rheumatologists lack training in caring for the adolescent population and express a lack of supportive medical care, both family physicians and allied health, to help them care for this population. Further, while they may feel comfortable—specifically discussing psychosocial concerns—the majority believe that it is the family physician’s responsibility to do so.
Compared to previous studies, there were similarities regarding the overall transition process and the barriers to a successful transition. For example, approximately one-third of respondents in our survey had no formal transition policy at their institution but an informal procedure that is followed, which is comparable to the findings in the Childhood Arthritis and Rheumatology Research Alliance (CARRA) survey completed in 2010 and the survey in 2014 by Zisman et al. [6, 10, 11]. In addition, the most frequently encountered barriers such as lack of training in caring for this age group, allied health providers, primary care providers, not being able to be renumerated appropriately and time constraints were common themes that have been found in previous studies [13, 16].
Almost half of adult rheumatologists reported feeling comfortable discussing psychosocial concerns with patients which is higher than previous studies (Table 2) [11]. However, this was self-reported and patient perspectives were not sought. Topics such as alcohol and tobacco use, contraception, and fertility were reported to be discussed either most or all the time by the majority of rheumatologists. However, 45% reported discussing vaping either “not at all” or “rarely” despite the same number feeling comfortable discussing the topic. Given the increase in incidence and the detrimental effects on lung function in youth with chronic conditions, this topic should be addressed consistently in this population [17].
Despite half of the participants reporting comfort in discussing psychosocial topics, more than 75% believed that the family physician should be the provider to discuss these issues (Table 2). Many of these concerns are related to the patient’s rheumatic disease and/or medications and side effects leading to difficulties with body image, self-esteem and, ultimately, to mental health concerns [7]. This can place patients in a difficult position as both health care providers feel as though it is the other’s responsibility to address these concerns. Unfortunately, many patients do not have a family doctor due to an overall shortage of primary care physicians in Canada or may have difficulty accessing their primary care provider [18]. In addition, many adolescents with chronic rheumatic disease are otherwise generally healthy and medically stable, and their rheumatologist may be the only physician they see regularly. Ideally, conversations about psychosocial issues would happen jointly among health care providers and the primary care provider should be included in the transition process.
The majority of transition research is targeted toward the pediatric healthcare provider’s care and perspective. Following a patient from childhood into adulthood establishes strong relationships between patients, their families, and the pediatric care team [2]; however, adult healthcare providers will ultimately spend a far longer time with patients over their life course. It is, therefore, essential to gain an understanding of their current practices and perceptions of gaps in care to ensure that patients receive optimal care. This survey adds to the body of research and provides a Canadian perspective on this topic. This survey design is unique in that it incorporated the competency by design framework. As more programs shift toward a competency by design curriculum, these milestones are now mandatory for trainees to complete their training [19]. The questions are not only applicable to rheumatology but can be used to assess transition in other pediatric sub-specialities with chronic conditions. It also looked at multiple aspects of psychosocial care, which have not been previously evaluated, such as vaping and gender identity.
Despite reminder emails, our study was limited by the relatively low response rate (15.2%) and small numbers of individuals in certain demographic groups (e.g., physicians 46–55 years old, practicing 6–15 years) making subgroup comparisons infeasible. In addition, many rheumatologists do not provide transition care and may not have completed the survey as a result. It is also probable that the survey was completed by adult rheumatologists who were already interested in transition care, leading to response bias and a limited perspective, thereby affecting the generalizability of results to the general adult rheumatology community. Although the survey was released in 2021, the impact of the COVID-19 pandemic upon transitional care and psychosocial concerns of patients was not included in the survey.
These results demonstrate that adult rheumatologists would benefit from more training, multidisciplinary support, resources, and appropriate renumeration to optimize the provision of transition care to young adult patients. Knowledge and treatment of psychosocial factors related to the young adult population should be emphasized during training. Educational initiatives such as workshops and peer teaching sessions with patient engagement should be co-created to improve the transition knowledge and skills for adult health care providers. Tele-rheumatology is a resource that can be valuable for this population as it can be used when patients are away for university/college for follow-up appointments and make it easier to have joint appointments with health care providers. A transition navigator or coach may be a valuable resource to bridge the divide between pediatric and adult care, provide psychosocial support to young adult patients, and provide continuity to patients during this challenging time [20, 21]. Transition navigators have been found to improve adherence to medication, attendance at clinic, and reduce acute complications of disease in other chronic disease populations [22]. Future research should be aimed toward assessing the impact of educational initiatives in subspecialty training of rheumatologists in their comfort supporting the young adult patient population with pediatric-onset disease. Furthermore, the effectiveness of implementing the above resources such as transition navigators upon the transition process from the patient, primary care provider and subspecialist perspective should also be obtained.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 217 KB)
Acknowledgements
We would like to acknowledge Canadian Rheumatology Association team members including Sue Ranta for her administrative assistance for this project and Kevin Baijnauth for survey development and deployment.
Author contribution
All authors made substantial contributions to study concept, design, material preparation, and analysis of data. The first draft of the manuscript was written by MP, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity, or any part of the work are appropriately investigated and resolved.
Funding
No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.
Data availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Declarations
Conflict of interest
None of the authors have any conflict of interest to declare.
Ethical approval
The questionnaire and methodology for this study were approved by the Hamilton Integrated Research Ethics Board (Project #13568) on September 16, 2021.
Consent to participate
Informed consent was obtained from all individual participants who answered the questionnaire.
Consent to publish
Participants consented to having data published. No identifying information is included in the article.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Luque Ramos A Hoffmann F Albrecht K Transition to adult rheumatology care is necessary to maintain DMARD therapy in young people with juvenile idiopathic arthritis Semin Arthritis Rheum 2017 47 269 275 10.1016/J.SEMARTHRIT.2017.05.003 28583690
2. Huang JS Gottschalk M Pian M Transition to adult care: systematic assessment of adolescents with a chronic illness and their medical teams J Pediatr 2011 159 994 10.1016/J.JPEDS.2011.05.038 21784450
3. Hazel E Zhang X Duffy CM Campillo S High rates of unsuccessful transfer to adult care among young adults with juvenile idiopathic arthritis Pediatr Rheumatol Online J 2010 10.1186/1546-0096-8-2 20148143
4. Foster HE Minden K Clemente D EULAR/PReS standards and recommendations for the transitional care of young people with juvenile-onset rheumatic diseases Ann Rheum Dis 2017 76 639 646 10.1136/ANNRHEUMDIS-2016-210112 27802961
5. Sadun RE Mind the gap: improving care in pediatric-to-adult rheumatology transition through education Rheum Dis Clin North Am 2020 46 103 118 10.1016/j.rdc.2019.09.008 31757279
6. Johnson KR Edens C Sadun RE Differences in healthcare transition views, practices, and barriers among north american pediatric rheumatology clinicians from 2010 to 2018 J Rheumatol 2021 48 1442 1449 10.3899/JRHEUM.200196 33526621
7. Palman J McDonagh JE Young minds: mental health and transitional care in adolescent and young adult rheumatology Open Access Rheumatol 2020 12 309 321 10.2147/OARRR.S228083 33324121
8. de Oliveira RJ Kishimoto ST de Souza DP The importance of transition from pediatric to adult rheumatology care in juvenile idiopathic arthritis Exp Rev Clin Immunol 2021 17 155 161 10.1080/1744666X.2020.1865157
9. Afzali A Wahbeh G Transition of pediatric to adult care in inflammatory bowel disease: is it as easy as 1, 2, 3? World J Gastroenterol 2017 23 3624 10.3748/WJG.V23.I20.3624 28611515
10. Chira P Ronis T Ardoin S White P Transitioning youth with rheumatic conditions: perspectives of pediatric rheumatology providers in the United States and Canada J Rheumatol 2014 41 768 779 10.3899/JRHEUM.130615 24584912
11. Zisman D Samad A Ardoin SP US adult rheumatologists’ perspectives on the transition process for young adults with rheumatic conditions Arthritis Care Res (Hoboken) 2020 72 432 440 10.1002/ACR.23845 30740937
12. Conti F Pontikaki I D’Andrea M Patients with juvenile idiopathic arthritis become adults: the role of transitional care Clin Exp Rheumatol 2018 36 1086 1094 29652654
13. Barnabe C Chomistek K Luca N National priorities for high-quality rheumatology transition care for youth in Canada J Rheumatol 2021 48 426 433 10.3899/JRHEUM.200790 33060318
14. Rheumatology Specialty Committee (2018) Rheumatology Entrustable Professional Activity Guide. Ottawa. Royal College of Physicians and Surgeons of Ontario. https://www.royalcollege.ca/rcsite/documents/cbd/epa-guide-rheumatology-e.pdf. Accessed 9 Sept 2022
15. Sharma A Minh Duc NT Luu Lam Thang T A consensus-based checklist for reporting of survey studies (CROSS) J Gen Intern Med 2021 36 3179 10.1007/S11606-021-06737-1 33886027
16. Gray WN Maddux MH Current transition practices in pediatric IBD: findings from a national survey of pediatric providers background: although practice guidelines have been published for transition to adult care among general chronic illness populations and specific to Inflamm Bowel Dis 2016 10.1097/MIB.0000000000000642 26752464
17. King B Jones C Baldwin G Briss P The EVALI and youth vaping epidemics—implications for public health N Engl J Med 2020 382 689 691 10.1056/NEJMp1916171 31951683
18. Malko AV Huckfeldt V Physician shortage in Canada: a review of contributing factors Glob J Health Sci 2017 10.5539/gjhs.v9n9p68
19. Iobst WF Sherbino J ten Cate O Competency-based medical education in postgraduate medical education Med Teach 2010 32 651 656 10.3109/0142159X.2010.500709 20662576
20. Kelly A Niddrie F Tunnicliffe DJ Patients’ attitudes and experiences of transition from paediatric to adult healthcare in rheumatology: a qualitative systematic review Rheumatology 2020 59 3737 3750 10.1093/RHEUMATOLOGY/KEAA168 32413124
21. Jiang I Major G Singh-Grewal D Patient and parent perspectives on transition from paediatric to adult healthcare in rheumatic diseases: an interview study BMJ Open 2021 11 e039670 10.1136/BMJOPEN-2020-039670 33397662
22. Crowley R Wolfe I Lock K McKee M Improving the transition between paediatric and adult healthcare: a systematic review Arch Dis Child 2011 96 548 553 10.1136/ADC.2010.202473 21388969
|
PMC010xxxxxx/PMC10171161.txt |
==== Front
Eur Geriatr Med
Eur Geriatr Med
European Geriatric Medicine
1878-7649
1878-7657
Springer International Publishing Cham
37162646
792
10.1007/s41999-023-00792-z
Research Paper
Relationship between sarcopenia and cachexia with prognostic markers of middle-aged and older inpatients with COVID-19: a case–control study
http://orcid.org/0000-0002-3122-4192
de Queiroz Júnior José Reginaldo Alves 1
http://orcid.org/0000-0001-5412-6467
da Costa Pereira Jarson Pedro jarsoncostap@gmail.com
14
http://orcid.org/0000-0002-1639-2366
Benjamim Raquel de Arruda Campos 1
http://orcid.org/0000-0003-3555-4006
da Silva Nahara Oliveira Lima 1
de Paiva Silva Maria Eduarda 2
http://orcid.org/0000-0002-5689-5048
Pinho Ramiro Cláudia Porto Sabino 23
1 grid.411227.3 0000 0001 0670 7996 Federal University of Pernambuco, Prof. Moraes Rego Avenue, 1235-Cidade Universitária, Recife, Pernambuco 50670-901 Brazil
2 Cardiologic Emergency Room of Pernambuco, Luiz Tavares Professor, Recife, Pernambuco Brazil
3 Hospital das Clinicas of Pernambuco, Recife, Pernambuco Brazil
4 Av. Prof. Moraes Rego, 1235, Cidade Universitária, Recife, Pernambuco 50670-901 Brazil
10 5 2023
2023
14 3 517526
27 2 2023
24 4 2023
© The Author(s), under exclusive licence to European Geriatric Medicine Society 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Key summary points
Aim
To identify the influence of sarcopenia and cachexia in prognostic markers in COVID-19 inpatients.
Findings
Sarcopenia is presented as a risk factor for mortality in COVID-19 inpatients.
Message
The accurate evaluation of nutritional abnormalities in hospitalized patients diagnosed with COVID-19 may help to reduce the risk for adverse outcomes.
Purpose
SARS-CoV-2 infection can lead to various manifestations beyond an inflammatory response, such as anorexia, hyposmia, and other symptoms that may increase the risk of nutritional disorders. Sarcopenia and cachexia are conditions that appear to influence COVID-19 evolution. Thus, this study aimed to evaluate sarcopenia and cachexia in hospitalized patients with COVID-19, verifying their clinical impacts and relationship with prognostic markers.
Methods
This is a case-control study involving inpatients with and without a COVID-19 diagnosis. The occurrence of sarcopenia was evaluated according to European Working Group on Sarcopenia 2 criteria. Cachexia was evaluated according to (Evans et al. in Clin Nutr 27:793–799, 2008) criteria. Inflammatory markers and the 4C Mortality Score were evaluated.
Results
Our study included 96 individuals, divided into two groups: COVID-19 (n = 32) and control (n = 64). The mean age of the COVID-19 group was 63.3 ± 11.8 years, and the control group had a mean age of 64.3 ± 5.5 years. No significant differences in mean age were found between the groups. The prevalence of sarcopenia and cachexia in patients with COVID-19 was 21.9% and 28.1%, respectively, while in the control group, it was 29.7% and 26.6%, respectively. Sarcopenic patients with COVID-19 had a higher risk of death (4C Mortality Score) (p = 0.038). The occurrence of sarcopenia or cachexia within the COVID-19 group was not associated with inflammatory biomarkers or a higher number of COVID-19 symptoms (p > 0.05).
Conclusion
The presence of sarcopenia among COVID-19 patients increased the risk of mortality.
Keywords
COVID-19
Sarcopenia
Cachexia
Body composition
Mortality
Aging
issue-copyright-statement© European Geriatric Medicine Society 2023
==== Body
pmcIntroduction
The outbreak of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) and Coronavirus disease-2019 (COVID-19) has caused thousands of deaths since late 2019 [1]. The SARS-CoV-2 virus can cause an unbalanced immune response, leading to an inflammatory reaction of great magnitude, known as a cytokine storm [1, 2].
The most common symptoms of these conditions are fever, fatigue, anorexia, headache, rash, diarrhea, arthralgia, myalgia and neuropsychiatric disturbances [3, 4], which may be a consequence of the damage caused by the cytokines to tissues, or physiological changes caused by the acute phase or may derive from responses mediated by immune cells [3]. Furthermore, infectious respiratory diseases may lead to body composition abnormalities, which is a path to malnutrition. Malnutrition is related to poor prognosis of affected individuals [5].
Current evidence has demonstrated a high prevalence of malnutrition in hospitalized patients with COVID-19 [1, 5, 6]. Although still poorly understood, these abnormalities lead to the hypothesis that COVID-19 may increase the risk of sarcopenia and cachexia development.
Recognized as a muscle disease, sarcopenia is defined as a decrease in skeletal muscle mass and low muscle strength [7]. Sarcopenia is commonly related to the aging process, however, acute sarcopenia can develop in hospitalized patients, especially as a consequence of infectious catabolism [8]. Sarcopenia is associated with a higher risk of adverse outcomes, such as hospital complications, mortality, rehabilitation needs, and a longer hospitalization time [9, 10].
Cachexia, on the other hand, is a complex metabolic syndrome characterized by a significant loss of body weight and muscle mass, often accompanied by weakness, fatigue, and anorexia. The diagnosis of cachexia includes a weight loss of 5% in the last 12 months (excluding fluid overload) or a Body Mass Index (BMI) < 20 kg/m2 in patients with chronic diseases. Additionally, patients must present at least three of the following clinical or laboratory criteria: decreased muscle strength, fatigue, anorexia, low lean mass index and abnormal biochemistry [11].
The impact of coronavirus on health is still not fully understood, including the effect of the presence of sarcopenia and cachexia on the evolution of SARS-CoV-2 infection. However, evidence suggests that the protein-muscle status can influence an individual's risk for the progression of COVID-19. Early identification and correction of these conditions may potentially improve patients’ outcomes [12]. In this context, this study aimed to evaluate sarcopenia and cachexia in middle-aged and older patients hospitalized with COVID-19 compared to a control group and to investigate the relationship of these conditions with prognostic markers.
Methods
Study design and sample size
The study is a case–control design that includes middle-aged and older inpatients from two university hospitals in Northeast Brazil. Data collection occurred between August 2021 and August 2022.
The COVID-19 group consists of patients aged ≥ 40 years, of both sexes, who were hospitalized with a confirmed diagnosis of COVID-19 by RT-PCR molecular test or SARS-CoV-2 Rapid Antigen Test, using swab of naso-oropharyngeal secretion. The control group comprises patients aged ≥ 40 years, of both sexes, who were hospitalized for reasons other than COVID-19 and tested negative for COVID-19.
Patients with pre-existing wasting diseases such as cancer, HIV infection, chronic heart failure, chronic obstructive pulmonary disease, and chronic kidney disease undergoing renal replacement therapy, as well as those with physical limitations such as amputation and use of pacemaker, were excluded from both groups. This exclusion was made to minimize confounding factors, as some clinical conditions can induce greater catabolism, leading to analysis bias.
The study used a pairing of 1 case to 2 controls (1:2). Each case was paired based on age (± 3 years), sex, nutritional status according to BMI, and the presence of comorbidities. Recruitment was done by selecting two controls for each patient diagnosed with COVID-19 who was entered into the study.
The sample size was calculated using Epi-Info software® version 6.04, considering the following parameters: the lowest prevalence outcome (sarcopenia), a significance level of 95%, a test power (1–β) of 80%, a ratio of two controls for each case, and a cross-product ratio (Odds Ratio–OR) of 3.8. The estimated prevalence of exposure to sarcopenia among cases was 22% [13] and 7% among controls [14]. The minimum required sample was 32 cases and 64 controls.
Sarcopenia and cachexia evaluation
Sarcopenia was determined by the coexistence of low muscle strength and low muscle mass [8]. Muscle strength was obtained from the Hand Grip Strength (HGS), measured by a JAMAR® digital dynamometer, following established techniques (American Society of Hand Therapists). A cutoff of < 27 kg/F for men and < 16 kg/F for women was adopted to define low HGS [7].
The patient was instructed to grip the dynamometer with their dominant hand while keeping their arm extended by the side of their body. The measurement of grip strength was obtained and recorded three times with 15-s intervals between each measurement. The average value of the three measurements was considered for analysis purposes.
Appendicular skeletal muscle mass (MMAE) was assessed to verify low muscle mass and was determined by the equation: MMAEkg=0,227×RI+0,0064×Xc+0,0095×weight+1384×sex-3964, in which RI represents the resistive index (height2(cm)/resistance (OHMS)); Xc represents reactance in OHMS and for sex is applied values of 1 to men, and 0 to women [7, 15].
The measurements of resistance and reactance were obtained from Bioelectrical Impedance (BIA), using portable equipment from the brand Biodynamics®, model 450, which applies a current of 800 μA, with a simple frequency of 50 kHz according to evidenced techniques [15]. From Sergi's equation results, the Appendicular Skeletal Muscle Index (ASMI) was calculated [7]. Values of ≤ 7.7 kg/m2 in men and ≤ 5.62 kg/m2 for women defined low muscle mass [16].
The diagnosis of cachexia was established using the criteria proposed by the Cachexia Consensus: 5% weight loss (free of edema) in the last 12 months or BMI < 20 kg/m2 and at least three of the following associated clinical or laboratory criteria: decrease muscle strength, fatigue, anorexia, low muscle mass, abnormal biochemistry, characterized by increased C-reactive protein (CRP) (> 5.0 mg/L), hemoglobin < 12 g/dL or albumin < 3.2 g/dL [11].
Weight loss was documented considering the usual weight in the last 12 months and the weight measured at the time of hospitalization. The percentage of weight loss (%WL) was calculated.
Fatigue was defined based on exhaustion assessed by self-reported fatigue, indicated by two questions from the Center for Epidemiological Studies—Depression (CES-D): “Did you feel that you had to make an effort to accomplish your usual tasks?” And “Were you unable to carry on your usual activities?”. The diagnosis of fatigue was established if the individual answered “most of the time and/or always” to one of the questions or even to both questions [17].
Anorexia was assessed using the Simplified Nutritional Appetite Questionnaire (SNAQ), in its Portuguese version [18].
Prognostic markers
4C (Coronavirus Clinical Characterization Consortium) Mortality Score, number of vaccine doses, Total Lymphocyte Count (TLC), Neutrophil–Lymphocyte Ratio (NLR), Monocyte-Lymphocyte Ratio (MLR) and Phase Angle (PA) were used as prognostic markers.
The 4C Mortality Score uses patient demographics data such as age in years and biological sex, clinical observations such as number of comorbidities, oxygen saturation, respiratory rate, Glasgow scale, and blood parameters such as serum urea, which are usually available at the time of hospital admission. Four risk groups were defined with corresponding mortality rates determined: low risk (score 0–3, mortality rate 1.2%), intermediate risk (score 4–8, mortality rate 9.9%), high risk (score 9–14, mortality rate 31.4%) and very high risk (≥ 15 points, mortality rate 61.5%) [19].
The total lymphocyte count (TLC) was defined using the equation: % lymphocytes × leukocytes/100. Values below 1500 mm3/dl defined a low TLC. NLR > 4.27 [20] and MLR > 0.23 [21] were used as indicators of worse patient outcomes with COVID-19. PA values (< 5º) were used as indicators of poor nutritional status and worse prognosis [22].
Demographical and clinical data
In our analysis, we took into account factors such as sex, age, and the presence of comorbidities, including systemic arterial hypertension (SAH), diabetes (DM), chronic kidney disease, thyroid and psychiatric disorders, cognitive dysfunction, and physical dependence. To register these comorbidities, we relied on the hospitalization diagnoses available in the medical records.
Self-reported signs and symptoms related to COVID-19 were documented, including hyposmia, hypogeusia, fever, dry cough, productive cough, headache, sore throat, dyspnea, asthenia, myalgia, dysphagia, anorexia, and diarrhea. Respiratory rate (RR) and oxygen saturation (SpO2) were also measured. Patients were asked about their COVID-19 vaccination status and the number of doses received. Biochemical data were obtained from medical records within 72 h of the patient's admission to the hospital. Blood count parameters and TLC were also considered.
Ethical aspects
All procedures carried out in this study were in accordance with the ethical principles of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The data collection was performed after obtaining approval from the Research Ethics Committee of Hospital das Clínicas (Federal University of Pernambuco) (CAAE: 4 5,469,721.0.0000.8807; Protocol number: 4.690.900; approved on May 4, 2021). Informed consent was obtained from all participants prior to their inclusion in the study. Patients agreed to participate in the research and to have their data published.
Data analysis
Regarding the analysis, the data were described using relative and absolute frequencies, mean and standard deviation or median and interquartile range, depending on the normality of the distribution, which was assessed by the Kolmogorov–Smirnov test. The Statistical Package for Social Sciences (SPSS®) software version 28 was used for the analysis.
To compare differences and distribution between proportions, the chi-square test was used. To verify differences between means or medians, Student’s t test or Mann Whitney U test were used, respectively. In all analyses, a significance level of p < 0.05 was considered.
Results
A total of 96 patients were included in the study, with 32 in the COVID-19 group and 64 in the control group. There were no significant differences in mean age between the groups (COVID-19: 63.3 ± 11.8 years, control group: 64.3 ± 5.5 years) (p = 0.656), as well as in demographic, clinical, and nutritional status data (p > 0.05). However, the TLC was significantly higher in the COVID-19 group compared to the control group (p < 0.001), (Table 1).Table 1 Demographic, clinical and nutritional characteristics of hospitalized patients diagnosed with COVID-19 compared to a control group in two hospitals in Northeast Brazil/2022 (n = 96)
Variables Control group N = 64 (%) COVID-19 group N = 32 (%) p value*
Sex (%)
Men 38 (59.4) 18 (56.3) 0.777
Age (%)
< 70 years 53 (82.8) 24 (75.0) 0.656
> 70 years 11 (17.2) 8 (25.0)
Comorbidities (%)
SAH (%) 56 (69.1) 13 (68.4) 0.952
DM (%) 32 (39.5) 5 (26.3) 0.211
Nutritional status and biochemical data
BMI (kg/m2) 26.3 ± 5.2 26.8 ± 5.6 0.657
BMI classification (%)
Malnutrition 3 (4.7) 2 (6.3) 0.958
Eutrophy 27 (42.2) 9 (28.1)
Excessive weight 34 (53.1) 21 (65.6)
Hemoglobin (mg/dl) 12.0 ± 2.2 11.30 ± 2.2 0.199
Leukocytes (mm3) 8200 (6543–10.195) 7010 (4840–9513) 0.157
TLC (mm3) 1320.3 (840–3750) 1502.5 (819–3144) < 0.001
HS (kg/f) 24.1 ± 10.8 25.4 ± 12.8 0.584
ASMI (kg/m2) 6.6 (5.8–7.3) 7.0 (6.5–7.9) 0.120
Phase angle < 5º 29 (43.3) 12 (37.5) 0.366
Sarcopenia (%) 19 (29.7) 7 (21.9) 0.522
Cachexia (%) 17 (26.6) 9 (28.1) 0.528
p-values <0.05 were statistically significant and highlighted in bold
BMI body mass index, TLC total lymphocyte count, HS handgrip strength, ASMI appendicular skeletal muscle mass index, *Pearson chi-square or Fisher Exact Test to frequency comparison. Student T Test to mean comparison and Mann Whitney U Test to compare median
Anthropometric data showed that almost 66% of the COVID-19 group had excessive weight according to their BMI. The frequency of excessive weight in the control group was 53.1%. No significant differences were observed between the two groups (p = 0.958).
The frequency of sarcopenia and cachexia was similar between groups (p > 0.05). In the COVID-19 group, the frequency of sarcopenia was 21.9% (n = 7), while cachexia was 28.1% (n = 9). In the control group, sarcopenia was present in 29.7% (n = 19) of the sample, while cachexia was present in 26.6% (n = 17). Among individuals with cachexia, the criterion of weight loss greater than 5% was met in 76.4% of individuals in the control group, whereas in the COVID-19 group, this criterion was fulfilled in 88.8% of individuals.
The frequency of sarcopenia and cachexia did not differ with sex, comorbidities, or number of symptoms of disease within the COVID-19 group, and these frequencies are described in Table 2.Table 2 Demographic and clinical variables according to the presence of sarcopenia and cachexia in hospitalized patients diagnosed with COVID-19 in two hospitals in Northeast Brazil/2022 (n = 32)
Variables Sarcopenia (N = 7) Without sarcopenia (N = 25) p value* Cachexia (N = 9) Without caquexia (N = 23) p value*
Sex
Men 5 (71.4) 13 (52.0) 0.318 7 (77.8) 11 (47.8) 0.125
Women 2 (28.6) 12 (48.0) 2 (22.2) 12 (52.2)
SAH
Yes 4 (57.1) 17 (68.0) 0.456 4 (44.4) 17 (73.9) 0.115
No 3 (42.9) 8 (32.0) 5 (65.6) 8 (26.1)
DM
Yes 3 (42.9) 9 (36.0) 0.535 3 (33.3) 9 (39.1) 0.761
No 4 (57.1) 16 (64.0) 6 (66.7) 14 (60.9)
SymptomsA
< 3 5 (83.3) 15 (62.5) 0.326 5 (55.6) 15 (71.4) 0.398
≥ 3 1 (16.7) 9 (37.5) 4 (44.4) 6 (28.6)
Comorbidities
< 2 3 (42.9) 14 (56.0) 0.424 5 (55.6) 12 (52.2) 0.863
≥ 2 4 (57.1) 11 (44.0) 4 (44.4) 11 (47.8)
*Pearson chi-square or Fisher Exact Test, A: Fever Dry Cough, Productive Cough, Headache, Sore throat, Dyspnea; Asthenia, Myalgia, Hyposmia, Hypogeusia, Dysphagia, Anorexia and Diarrhea
Patients with COVID-19 who had sarcopenia had a higher mean age (72.6 ± 10.2 vs 60.7 ± 11.0; p = 0.016), lower mean BMI (20.9 ± 3.1 vs 28.5 ± 5.0; p = 0.001), lower mean phase angle (PA) (3.6 ± 2.0 vs 6.2 ± 3.7; p = 0.025), and higher scores on the 4C Mortality Score (9.0 (9.0–11.0) vs 7.0 (6.0–9.5); p = 0.038). The number of vaccine doses, TLC, NLR, and MLR were similar in patients with or without sarcopenia (p > 0.05) (Table 3).Table 3 Comparison of nutritional, clinical and prognostic variables regarding the presence or absence of sarcopenia in hospitalized patients diagnosed with COVID-19 (n = 32)
Variables Sarcopenia Without sarcopenia p valueA
Mean SD Mean SD
Age (years) 72.6 10.2 60.7 11.0 0.016
BMI (kg/m2) 20.9 3.1 28.5 5.0 0.001
Leukocytes (mm3) 6594 3693 8157 3827 0.352
TLC (mm3) 1466 1036 1499 735 0.940
PA (º) 3.6 2.0 6.2 3.7 0.025
Respiratory rate 21.2 7.0 21.3 4.7 0.944
SpO2 97.0 1.1 96.6 1.3 0.516
Variables Sarcopenia Without sarcopenia p valueB
Median QI Median QI
%WL 9.4 0.0–21.7 0.0 0.0–5.2 0.092
NLR 4.0 1.3–6.4 2.5 1.6–5.3 0.859
MLR 0.4 0.2–0.9 0.6 0.2–0.9 0.412
4C mortality score 9.0 9.0–11.0 7.0 6.0–9.5 0.038
Number of vaccine doses 2.0 1.0–2.0 2.0 2.0–3.0 0.104
p-values <0.05 were statistically significant and highlighted in bold
BMI body mass index, TLC total lymphocyte count, NLR neutrophil to lymphocytes ratio, MLR monocytes to lymphocytes ratio, SpO2 oxygen saturation, PA phase angle, %WL % weight loss, A Student T Test; B Mann Whitney U Test, SD standard deviation, QI quartile interval
COVID-19 patients with cachexia had a lower mean BMI (22.6 ± 4.7 vs. 28.5 ± 5.0, p = 0.005) and a higher percentage of weight loss (10.2 (7.7–24.0) vs. 0.0 (0.0–1.3), p < 0.001). However, the prognostic markers were similar between patients with and without cachexia (p > 0.05) (Table 4).Table 4 Comparison of nutritional, clinical and prognostic variables regarding the presence or absence of cachexia in hospitalized patients diagnosed with COVID-19 (n = 32)
Variables Cachexia Without cachexia p valueA
Mean SD Mean SD
Age (years) 64.0 10.6 63.0 12.4 0.833
BMI (kg/m2) 22.6 4.7 28.5 5.0 0.005
Leukocytes (mm3) 8298 5214 7626 3206 0.726
TLC (mm3) 1261 1034 1582 680 0.409
PA (º) 4.5 2.2 5.3 2.4 0.166
Respiratory Rate 19.4 4.5 22.1 5.2 0.198
SpO2 97.1 1.2 96.5 1.2 0.240
Variables Cachexia Without cachexia p-valueB
Median QI Median QI
%WL 10.2 7.7–24.0 0.0 0.0–1.3 < 0.001
NLR 5.3 1.7–19.4 2.1 1.6–5.1 0.180
MLR 0.7 0.2–1.5 0.5 0.2–0.8 0.801
4C Mortality score 9.0 6.5–10.0 7.0 6.0–10.0 0.592
Number of vaccine doses 2.0 1.3–2.0 2.0 2.0–2.8 0.320
p-values <0.05 were statistically significant and highlighted in bold
BMI body mass index, TLC total lymphocyte count, NLR neutrophil to lymphocytes ratio, MLR monocytes to lymphocytes ratio, SatO2 oxygen saturation, PA phase angle, %WL % weight loss, A student T test, B Mann Whitney U test, SD standard deviation, QI quartile interval
Discussion
The interest in verifying the occurrence of sarcopenia and cachexia in the population with COVID-19 is not only due to a better understanding of the nutritional impairment promoted by the disease, but also to verify the prognostic impact of these previous nutritional conditions on the evolution of the disease. It is already well established that the occurrence of sarcopenia and cachexia is associated with higher morbidity and adverse outcomes [9, 10].
To the best of our knowledge, this is the first case–control study to investigate the association between the occurrence of sarcopenia and cachexia with prognostic markers in hospitalized patients with COVID-19. The prevalence of sarcopenia among COVID-19 inpatients found in our study was similar to that reported in a retrospective cohort study by Kim et al. [23] (prevalence of 23.97%) and a study by Moctezuma-Velazquez et al. [13] (prevalence of 22.16%).
There is growing evidence suggesting that SARS-CoV-2 infection can result in gastrointestinal abnormalities, which can lead to inadequate nutritional intake. In addition, the infection can also cause body composition abnormalities by recruiting inflammatory cytokines, such as IFN-γ, IL-1β, IL-6, IL-17, and TNF-α, which can cause pathological changes in skeletal muscle tissue. The virus can directly bind to muscle tissue through the angiotensin 2 receptor (AT2 receptor), leading to muscle fiber proteolysis and inhibition of protein synthesis, which can lead to the development of sarcopenia and cachexia [24, 25].
Underweight was present in 6.5% of the COVID-19 patients, indicating that sarcopenia occurred independently of changes in weight. It is important to highlight that the occurrence of sarcopenia is associated with a higher risk of adverse outcomes, such as susceptibility to infections, rehabilitation needs, longer hospitalization, and mortality. Therefore, the occurrence of sarcopenia should be evaluated in hospitalized patients, especially in acute conditions such as COVID-19 [8–10].
Regarding cachexia, few studies have evaluated this condition in patients with COVID-19. So far, only two studies have investigated the diagnosis of cachexia during the hospitalization of these patients, and both focused on patients admitted to the Intensive Care Unit (ICU) or semi-intensive care unit. Allard et al. [26] reported a prevalence of 37.04%, while Pironi et al. [27] described a prevalence of 24.25%.
Our findings indicated a cachexia prevalence of 28.1% in the COVID-19 group. This is concerning data, as cachexia is a complex metabolic condition that is not only associated with underlying disease but can also result from worsening malnutrition and/or sarcopenia [28]. Cachexia further increases the risk of disability and mortality, and individuals who survive the cachexia condition often require longer rehabilitation [29, 30].
The lack of differences in the prevalence of sarcopenia and cachexia between COVID-19 patients and controls in our study does not necessarily imply that COVID-19 does not increase the risk for these conditions. It is important to note that our study protocol involved an evaluation for sarcopenia and cachexia within 72 h of hospital admission, and thus, we were unable to evaluate the effects of the progression of SARS-CoV-2 infection on nutritional parameters.
In addition, it is important to highlight the changes in the epidemiological scenario of COVID-19 worldwide after expanding vaccine coverage. At the time the study was developed, there was already vaccination coverage with up to 2 doses, which may have contributed to less muscle stress caused by the viral infection. In accordance, some studies have already evaluated the nutritional status of patients with COVID-19, based on serum markers, such as Retinol Binding Protein (RBP) in the period before and after vaccination. These studies saw that individuals after vaccination maintained RBP at higher levels, reflecting that vaccination would be able to reduce comorbidities and protein-muscle impairment [31].
In our study, TLC was higher in the COVID-19 group compared to the control group. This can be attributed to the inflammatory-infectious condition observed in COVID-19 patients, and thus, this result was expected. While TLC is often used as a nutritional parameter, its interpretation should be approached with caution in the presence of infection and inflammation.
The higher mean age in individuals with sarcopenia is an expected finding, supported by evidence showing an association between older age and lower BMI in this population [14]. Aging itself increases the risk of sarcopenia due to the decline in muscle mass and strength caused by changes in muscle cellular metabolism. These changes promote alterations in the structure of muscle fibers, neurodegeneration, a decrease in the number of motor muscle cells, impairment of protein anabolism, hormonal resistance, and when coupled with physical inactivity and inadequate nutritional intake, ultimately lead to sarcopenia [32].
The lowest PA values observed in patients with sarcopenia in our study are consistent with previous findings that show lower PA levels in sarcopenic individuals and a higher prevalence of sarcopenia in those with lower PA [33, 34]. PA has been demonstrated to be a useful risk marker for sarcopenia [35, 36], as well as having prognostic value for adverse outcomes [37, 38].
Similar results were observed in patients with COVID-19. PA values were found to be independently and inversely correlated with adverse outcomes in COVID-19, including the risk of ICU admission, complications, length of hospital stay, and mortality [39–41]. Some evidence suggests that PA is associated with sarcopenia, as it may reflect water distribution, body muscle cellularity, and muscle strength, which are components used in the definition of sarcopenia [34].
The higher 4C Mortality Score observed in sarcopenic patients in the COVID-19 group suggests a greater risk among these patients. This finding supports the hypothesis that muscle-protein depletion may lead to worse outcomes in individuals with COVID-19. In a recent meta-analysis, individuals hospitalized with COVID-19 who had prior nutritional impairment had a tenfold higher risk of mortality compared to those without impairment [42]. It's important to keep in mind that individuals with sarcopenia were older, which can be a confounding factor as advanced age is associated with both sarcopenia and mortality. However, we should also acknowledge that these may be intervening conditions. Thus, while age may have an effect on both conditions, we cannot overlook the impact of sarcopenia on negative outcomes.
The lowest BMI and highest percentage of weight loss observed in patients with cachexia are expected findings, as these two aspects are included in the diagnostic criteria for this condition and are consistent and well-predicted in scientific evidence [11, 43, 44].
Since the onset of the current pandemic, several scientific studies have aimed to identify clinical and laboratory predictors of poor prognosis that can assist healthcare teams in monitoring strategies and making therapeutic decisions.
All patients included in our study had received at least one dose of the COVID-19 vaccine at the time of data collection. However, nearly 80% of the sample did not have a complete vaccination schedule, which may pose an additional risk for severe COVID-19 progression. The lack of a complete vaccination schedule may be associated with a greater inflammatory profile and, consequently, a higher risk of sarcopenia (muscle-protein depletion). To the best of our knowledge, this is the first study to investigate the association between COVID-19 vaccine doses and the occurrence of sarcopenia and cachexia.
NLR is an inflammatory biomarker that can be used as an indicator of systemic inflammation [45]. It has been extensively investigated as a possible prognostic factor in the progression of COVID-19. NLR at the time of hospital admission can predict mortality and ICU admission in COVID-19 patients [21]. The association of NLR with cachexia may suggest that cachexia is linked to a worse inflammatory profile and a higher risk of developing negative outcomes. However, our study did not find significant values for this association.
Our study had some limitations. The sample size was relatively small, which may limit the generalizability of the results. Additionally, the evaluation of sarcopenia and other nutritional measurements was conducted up to 72 h after hospitalization, which may restrict the observation of the effects of COVID-19 evolution on nutritional status. It is important to note that our sample is also limited in terms of patients over the age of 80, with only two patients included in this age group. Therefore, our results may not be applicable to the frail population with a higher incidence of mortality. Nevertheless, the inclusion of a control group for comparative purposes was a strong point of our investigation.
The prevalence of sarcopenia and cachexia was elevated in middle-aged and older inpatients with COVID-19, but was not higher when compared to control group. In our study, sarcopenia was related to higher mortality risk (4C mortality score) and lower PA. According to our findings, we can suggest that the proper evaluation of previous nutritional impairments, such as sarcopenia may help the development of multimodal strategies to treat and to prevent these conditions, decreasing the mortality risk.
Further research with a prospective design is needed to investigate the bidirectional relationship between COVID-19 and body composition. This would help to elucidate the effects of the disease on body composition and the impact of impaired body composition on the evolution of COVID-19. Moreover, it is necessary to investigate the long-term consequences of COVID-19 on body composition, particularly in patients who have recovered from the acute phase of the disease. Such studies would provide valuable information for the development of prevention and treatment strategies for COVID-19-related body composition impairments.
Acknowledgements
The authors acknowledge the Fundação de Amparo à Ciência e Tecnologia do Estado de Pernambuco for providing support to this project. We also want to thank all the patients included in our research.
Author contributions
JRAdeQJ: conception and design of the research, writing original manuscript, acquisition, visualization, software and validation of the data. JPdaCP: conception and design of the research, writing original manuscript, acquisition and analyzes of the data. RdeACB: acquisition of the data and writing original. NOLdaS: acquisition of the data. Maria EdePS: acquisition of the data. CPSPR: visualization, software, validation of data, supervision and review. All authors made substantial contributions to the conception and design of the study, acquisition, analysis and interpretation of data. They also contributed to the manuscript drafting and critical review of the intellectual content. All authors critically revised the manuscript, agree to be fully accountable for ensuring the integrity and accuracy of the work, and read and approved the final manuscript.
Funding
There is no funding source.
Data availability
Upon request.
Declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval
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. Data collection were performed after consideration and approval by the Research Ethics Committee. This study was approved by the ethics committee of the respective institutions, under number of approvals: 4.690.900.
Informed consent
All participants provided informed consent to participate in the study and to have their data published.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Li T Zhang Y Gong C Wang J Liu B Shi L Prevalence of malnutrition and analysis of related factors in elderly patients with COVID-19 in Wuhan China Eur J Clin Nutr 2020 74 871 875 10.1038/s41430-020-0642-3 32322046
2. Ballesteros Pomar MD Bretón LI Clinical nutrition in times of COVID-19 Endocrinología, Diabetes Y Nutrición 2020 67 427 430 10.1016/j.endien.2020.09.004 32482569
3. Fajgenbaum DC June CH Cytokine storm N Engl J Med 2020 383 2255 2273 10.1056/NEJMra2026131 33264547
4. Grupp SA Kalos M Barrett D Aplenc R Porter DL Rheingold SR Chimeric antigen receptor-modified T cells for acute lymphoid leukemia N Engl J Med 2013 368 1509 1518 10.1056/NEJMoa1215134 23527958
5. Rouget A Vardon-Bounes F Lorber P Vavasseur A Marion O Marcheix B Prevalence of malnutrition in COVID-19 inpatients: the Nutricov study In Review 2020 10.21203/rs.3.rs-41500/v1
6. Fernández-Quintela A Milton-Laskibar I Trepiana J Gómez-Zorita S Kajarabille N Léniz A Key aspects in nutritional management of COVID-19 patients JCM 2020 9 2589 10.3390/jcm9082589 32785121
7. Cruz-Jentoft AJ Bahat G Bauer J Boirie Y Bruyère O Cederholm T Sarcopenia: revised European consensus on definition and diagnosis Age Ageing 2019 48 16 31 10.1093/ageing/afy169 30312372
8. Welch C Hassan-Smith KZ Greig AC Lord MJ Jackson AT Acute sarcopenia secondary to hospitalisation - an emerging condition affecting older adults Aging Dis 2018 9 151 10.14336/AD.2017.0315 29392090
9. Au PC-M Li H-L Lee GK-Y Li GH-Y Chan M Cheung BM-Y Sarcopenia and mortality in cancer: a meta-analysis Osteoporosis Sarcopenia 2021 7 S28 33 10.1016/j.afos.2021.03.002 33997306
10. Welch C Greig C Masud T Wilson D Jackson TA COVID-19 and acute sarcopenia Aging Dis 2020 11 1345 10.14336/AD.2020.1014 33269092
11. Evans WJ Morley JE Argilés J Bales C Baracos V Guttridge D Cachexia: a new definition Clin Nutr 2008 27 793 799 10.1016/j.clnu.2008.06.013 18718696
12. Prado CM Purcell SA Alish C Pereira SL Deutz NE Heyland DK Implications of low muscle mass across the continuum of care: a narrative review Ann Med 2018 50 675 693 10.1080/07853890.2018.1511918 30169116
13. Moctezuma-Velázquez P Miranda-Zazueta G Ortiz-Brizuela E González-Lara MF Tamez-Torres KM Román-Montes CM Low thoracic skeletal muscle area is not associated with negative outcomes in patients with COVID-19 Am J Phys Med Rehabil 2021 100 413 418 10.1097/PHM.0000000000001716 33587451
14. Yazar T Olgun YH Prevalance of sarcopenia according to decade Clin Nut ESPEN 2019 29 137 141 10.1016/j.clnesp.2018.11.005
15. Sergi G De Rui M Veronese N Bolzetta F Berton L Carraro S Assessing appendicular skeletal muscle mass with bioelectrical impedance analysis in free-living Caucasian older adults Clin Nutr 2015 34 667 673 10.1016/j.clnu.2014.07.010 25103151
16. Barbosa-Silva TG Bielemann RM Gonzalez MC Menezes AMB Prevalence of sarcopenia among community-dwelling elderly of a medium-sized South American city: results of the COMO VAI ? study: sarcopenia prevalence in a South American city J Cachexia Sarcopenia Muscle 2016 7 136 143 10.1002/jcsm.12049 27493867
17. Batistoni SS Neri AL Cupertino AP Validade da escala de depressão do center for epidemiological studies entre idosos brasileiros Rev Saúde Pública 2007 41 4 598 605 10.1590/S0034-89102007000400012 17589758
18. Zukeran MS Aprahamian I Vicente BM Ribeiro SM Portuguese version of the SNAQ questionnaire: translation and cultural adaptation Rev Assoc Med Bras 2020 66 3 374 381 10.1590/S0004-2803.202000000-33
19. Knight SR Ho A Pius R Buchan I Carson G Drake TM Risk stratification of patients admitted to hospital with covid-19 using the ISARIC WHO clinical characterisation protocol: development and validation of the 4C mortality score BMJ 2020 10.1136/bmj.m3339 33127603
20. Kheyri Z Metanat S Hosamirudsari H Akbarpour S Shojaei M Faraji N Neutrophil-to-lymphocyte ratio cut-off point for COVID-19 mortality: a retrospective study ACTA 2022 10.18502/acta.v60i1.8325
21. Peng J Qi D Yuan G Deng X Mei Y Feng L Diagnostic value of peripheral hematologic markers for coronavirus disease, (COVID-19): A multicenter, cross-sectional study J Clin Lab Anal 2019 2020 34 10.1002/jcla.23475
22. Kyle UG Soundar EP Genton L Pichard C Can phase angle determined by bioelectrical impedance analysis assess nutritional risk? A comparison between healthy and hospitalized subjects Clin Nutr 2012 31 875 881 10.1016/j.clnu.2012.04.002 22560739
23. Kim J-W Yoon JS Kim EJ Hong H-L Kwon HH Jung CY Prognostic implication of baseline sarcopenia for length of hospital stay and survival in patients with coronavirus disease 2019 J Gerontol A Biol Sci Med Sci 2021 76 e110 e116 10.1093/gerona/glab085 33780535
24. Chen N Zhou M Dong X Qu J Gong F Han Y Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study The Lancet 2020 395 507 513 10.1016/S0140-6736(20)30211-7
25. Disser NP De Micheli AJ Schonk MM Konnaris MA Piacentini AN Edon DL Musculoskeletal consequences of COVID-19 J Bone Joint Surg 2020 102 1197 1204 10.2106/JBJS.20.00847 32675661
26. Allard L Ouedraogo E Molleville J Bihan H Giroux-Leprieur B Sutton A Malnutrition: percentage and association with prognosis in patients hospitalized for coronavirus disease 2019 Nutrients 2020 12 3679 10.3390/nu12123679 33260603
27. Pironi L Sasdelli AS Ravaioli F Baracco B Battaiola C Bocedi G Malnutrition and nutritional therapy in patients with SARS-CoV-2 disease Clin Nutr 2021 40 1330 1337 10.1016/j.clnu.2020.08.021 32900518
28. Morley JE Anker SD Evans WJ Cachexia and aging: an update based on the fourth international cachexia meeting J Nutr Health Aging 2009 13 47 55 10.1007/s12603-009-0009-x 19151908
29. Brugliera L Spina A Castellazzi P Cimino P Tettamanti A Houdayer E Rehabilitation of COVID-19 patients J Rehabil Med 2020 52 jrm00046 10.2340/16501977-2678 32286674
30. Morley JE Kalantar-Zadeh K Anker SD COVID-19: a major cause of cachexia and sarcopenia? J Cachexia Sarcopenia Muscle 2020 11 863 865 10.1002/jcsm.12589 32519505
31. Vaz-Rodrigues R Mazuecos L Villar M Urra JM Gortázar C De La Fuente J Serum biomarkers for nutritional status as predictors in COVID-19 patients before and after vaccination J Functional Foods 2023 101 105412 10.1016/j.jff.2023.105412
32. Woo J Sarcopenia Clin Geriatr Med 2017 33 305 314 10.1016/j.cger.2017.02.003 28689564
33. Kilic MK Kizilarslanoglu MC Arik G Bolayir B Kara O Dogan Varan H Association of bioelectrical impedance analysis-derived phase angle and sarcopenia in older adults Nutr Clin Pract 2017 32 103 109 10.1177/0884533616664503 27590205
34. Di Vincenzo O Marra M Di Gregorio A Pasanisi F Scalfi L Bioelectrical impedance analysis (BIA)-derived phase angle in sarcopenia: a systematic review Clin Nutr 2021 40 5 3052 3061 10.1016/j.clnu.2020.10.048 33183880
35. Basile C Della-Morte D Cacciatore F Gargiulo G Galizia G Roselli M Phase angle as bioelectrical marker to identify elderly patients at risk of sarcopenia Exp Gerontol 2014 58 43 46 10.1016/j.exger.2014.07.009 25034911
36. Santana NDM Pinho CPS Da Silva CP Dos Santos NF Mendes RML Phase angle as a sarcopenia marker in hospitalized elderly patients Nutr Clin Pract 2018 33 232 237 10.1002/ncp.10016 29397040
37. Lukaski HC Kyle UG Kondrup J Assessment of adult malnutrition and prognosis with bioelectrical impedance analysis: phase angle and impedance ratio Curr Opin Clin Nutr Metab Care 2017 20 330 339 10.1097/MCO.0000000000000387 28548972
38. Garlini LM Alves FD Ceretta LB Perry IS Souza GC Clausell NO Phase angle and mortality: a systematic review Eur J Clin Nutr 2019 73 495 508 10.1038/s41430-018-0159-1 29695763
39. Cornejo-Pareja I Vegas-Aguilar IM García-Almeida JM Bellido-Guerrero D Talluri A Lukaski H Phase angle and standardized phase angle from bioelectrical impedance measurements as a prognostic factor for mortality at 90 days in patients with COVID-19: a longitudinal cohort study Clin Nutr 2022 41 3106 3114 10.1016/j.clnu.2021.02.017 33642143
40. Moonen HPFX Bos AE Hermans AJH Stikkelman E Van Zanten FJL Van Zanten ARH Bioelectric impedance body composition and phase angle in relation to 90-day adverse outcome in hospitalized COVID-19 ward and ICU patients: The prospective BIAC-19 study Clinical Nutrition ESPEN 2021 46 185 192 10.1016/j.clnesp.2021.10.010 34857194
41. Osuna-Padilla IA Rodríguez-Moguel NC Rodríguez-Llamazares S Aguilar-Vargas A Casas-Aparicio GA Ríos-Ayala MA Low phase angle is associated with 60-day mortality in critically ill patients with COVID-19 J Parenter Enteral Nutr 2022 46 828 835 10.1002/jpen.2236
42. Abate SM Chekole YA Estifanos MB Abate KH Kabthymer RH Prevalence and outcomes of malnutrition among hospitalized COVID-19 patients: a systematic review and meta-analysis Clinical Nutrition ESPEN 2021 43 174 183 10.1016/j.clnesp.2021.03.002 34024511
43. Haehling S Anker MS Anker SD Prevalence and clinical impact of cachexia in chronic illness in Europe, USA, and Japan: facts and numbers update 2016 J Cachexia Sarcopenia Muscle 2016 7 507 509 10.1002/jcsm.12167 27891294
44. Scherbakov N Doehner W Cachexia as a common characteristic in multiple chronic disease: Editorial J Cachexia Sarcopenia Muscle 2018 9 1189 1191 10.1002/jcsm.12388 30637985
45. De Jager CPC Wever PC Gemen EFA Kusters R Van Gageldonk-Lafeber AB Van Der Poll T The neutrophil-lymphocyte count ratio in patients with community-acquired pneumonia PLoS ONE 2012 7 e46561 10.1371/journal.pone.0046561 23049706
|
PMC010xxxxxx/PMC10171899.txt |
==== Front
Mol Genet Metab
Mol Genet Metab
Molecular Genetics and Metabolism
1096-7192
1096-7206
Elsevier Inc.
S1096-7192(23)00237-8
10.1016/j.ymgme.2023.107607
107607
Regular Article
COVID-19 in inherited metabolic disorders: Clinical features and risk factors for disease severity
Kahraman Ayca Burcu 1⁎
Yıldız Yılmaz 1
Çıkı Kısmet
Erdal Izzet
Akar Halil Tuna
Dursun Ali
Tokatlı Ayşegül
Sivri Serap
Hacettepe University Faculty of Medicine, Department of Pediatrics, Division of Pediatric Metabolism, Turkey
⁎ Corresponding author.
1 Equal contribution as first authors.
11 5 2023
6 2023
11 5 2023
139 2 107607107607
21 12 2022
2 5 2023
9 5 2023
© 2023 Elsevier Inc. All rights reserved.
2023
Elsevier Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Old age, obesity, and certain chronic conditions are among the risk factors for severe COVID-19. More information is needed on whether inherited metabolic disorders (IMD) confer risk of more severe COVID-19. We aimed to establish COVID-19 severity and associated risk factors in patients with IMD currently followed at a single metabolic center.
Methods
Among all IMD patients followed at a single metabolic referral center who had at least one clinic visit since 2018, those with accessible medical records were reviewed for SARS-CoV-2 tests. COVID-19 severity was classified according to the WHO recommendations, and IMD as per the international classification of IMD.
Results
Among the 1841 patients with IMD, 248 (13.5%) had tested positive for COVID-19, 223 of whom gave consent for inclusion in the study (131 children and 92 adults). Phenylalanine hydroxylase (48.4%) and biotinidase (12.1%) deficiencies were the most common diagnoses, followed by mucopolysaccharidoses (7.2%). 38.1% had comorbidities, such as neurologic disabilities (22%) or obesity (9.4%). The majority of COVID-19 episodes were asymptomatic (16.1%) or mild (77.6%), but 6 patients (2.7%) each had moderate and severe COVID-19, and two (0.9%) had critical COVID-19, both of whom died. 3 patients had an acute metabolic decompensation during the infection. Two children developed multisystem inflammatory syndrome (MIS-C). Long COVID symptoms were present in 25.2%. Presence of comorbidities was significantly associated with more severe COVID-19 in adults with IMD (p < 0.01), but not in children (p = 0.45). Compared to other categories of IMD, complex molecule degradation disorders were significantly associated with more severe COVID-19 in children (p < 0.01); such a significant IMD category distinction was not found in adults.
Discussion
This is the largest study on COVID-19 in IMD patients relying on real-word data and objective definitions, and not on merely expert opinions or physician surveys. COVID-19 severity and long COVID incidence in IMD are probably similar to the general population, and the risk of acute metabolic decompensation is not likely to be greater than that in other acute infections. Disease category (complex molecule degradation) in children, and comorbidities in adults may be associated with COVID-19 severity in IMD. Additionally, the first documented accounts of COVID-19 in 27 different IMD are recorded. The high occurrence of MIS-C may be coincidental, but warrants further study.
Keywords
Inherited metabolic diseases
Disorders of complex molecule degradation
Comorbidity
SARS-CoV-2
Coronavirus disease 2019 (COVID-19)
Pneumonia
==== Body
pmc1 Introduction
It has been more than 3 years since coronavirus disease 2019 (COVID-19), caused by SARS-CoV-2, was declared a pandemic by the World Health Organization (WHO). Although preventive and therapeutic advances have curbed the devastating toll, the pandemic is still ongoing, constantly evolving with long-term effects and novel viral variants of concern, indicating that further understanding of COVID-19 will be relevant for many years to come.
Risk stratification is crucial for pandemic response, and it became clear early in the course of the pandemic that individuals with advanced age, male gender, and certain common chronic conditions including obesity, diabetes mellitus, hypertension, immune suppression and heart, lung or kidney disease have a higher risk of severe SARS-CoV-2 infection [1]. However, risk determination for rare disorders is difficult. For example, the international classification of inherited metabolic disorders (ICIMD) includes 1450 mostly rare or extremely rare disorders in 24 categories [2]. Individuals with inherited metabolic disorders (IMD) can be considered a vulnerable group because of their multisystem involvement and some may have risk of metabolic decompensation triggered by infections. Reports of COVID-19 in individuals with IMD began to emerge in June 2020 [3]. With more cases subsequently reported, the possibility of an increased risk due to the underlying metabolic defect and or to associated comorbidities arose. Additionally, as the pandemic and the knowledge around it evolved, other issues emerged, including long-term complications of COVID-19, and secondary problems related to disruption of health services, such as inability to access drugs or hospitals, which was shown to be associated with disease progression as shown by our group and others [4,5]. Availability of vaccines marked a shift in the dynamics of the pandemic and eventually paved the way for transitioning to a “new normal”. In Turkey, where the present study was performed, COVID-19 vaccines have been available for individuals over −65 years old and high-risk population, all adults (> 18 years), and children aged 12–18 years since February 2021, June 2021, and September 2021, respectively.
According to a physician survey conducted by the European Reference Network for Hereditary Metabolic Disorders (MetabERN), adult patients were reported to have a higher rate of hospitalization and a more severe course compared to general population and pediatric patients [6]. Similarly, Hasnat et al. demonstrated that children had milder clinical symptoms than adults [7]. Mütze et al. reported that hospitalization rates did not differ by diagnosis or age group, and no complications specific to IMD or related to COVID-19 have been observed [8]. More information is needed on whether inherited metabolic disorders (IMD) confer risk of more severe COVID-19. In this study, we aimed to answer this question and describe the features of COVID-19 in IMD by analyzing the COVID-19 histories and long-term complications of a large cohort of patients at a single center.
2 Material and methods
2.1 Study design and participants
This study was approved by Hacettepe University Ethics Committee for Non-Interventional Clinical Studies (GO 22/176, 2022/08–23). Informed consent was obtained from all patients or their parents/guardians. The study population consisted of pediatric and adult patients followed at the Pediatric Metabolism Unit of Hacettepe University İhsan Dogramacı Children's Hospital with a definitive diagnosis of IMD. Among all IMD patients with at least one in-person or remote visit between 1 January 2018 and 30 June 2021, those with accessible electronic health record data were reviewed retrospectively for SARS-CoV-2 PCR or antibody tests. The flowchart of participant inclusion is shown in Fig. 1 .Fig. 1 Flowchart of patient selection. IMD: inherited metabolic disorders.
Fig. 1
2.2 Data collection
The patients were contacted via phone calls and asked to participate in a survey including demographic data, medical history (comorbidities, medications etc.), and features of COVID-19 infection (timing, vaccination status, symptoms, treatment etc.). Hospital records were also reviewed retrospectively.
2.3 Definitions and classifications
Patients with records of SARS-CoV-2 antibody positivity before vaccination or PCR positivity at any time were defined to have had COVID-19. COVID-19 severity was classified according to the WHO recommendations as asymptomatic, mild, moderate, severe, or critical illness. In summary, the presence of symptoms without evidence of viral pneumonia or hypoxia defines mild COVID-19, non-severe pneumonia defines moderate COVID-19, severe pneumonia (eg. hypoxia, respiratory distress) defines severe COVID-19 and acute respiratory distress syndrome defines critical COVID-19 [9]. Determination of odds ratios for susceptibility to more severe COVID-19 was based on the development of pneumonia (asymptomatic-mild vs. moderate-severe-critical COVID-19). Death due to COVID-19 was defined based on WHO recommendations, as death resulting from a compatible illness in a patient with COVID-19 (only confirmed cases were included) without a period of complete recovery from COVID-19 between illness and death, unless there is a clear alternative cause of death unrelated to COVID-19 [10]. Long COVID was defined according to Centers for Disease Control and Prevention (CDC) recommendations as symptoms (sensory, neurologic, cardiorespiratory, and mental) that persist at least four weeks after infection. CDC recommendations were used to define multisystem inflammatory syndrome in children (MIS-C), a severe condition characterized by hyperinflammation and hypercoagulability [11]. IMD were classified according to the ICIMD [2]. If a patient experienced two or more episodes of COVID-19, it was counted as 1 patient and the characteristics of the more severe infection was recorded.
2.4 Statistical analyses
The data were analyzed by SPSS v25.0 (SPSS Inc., Chicago, USA). Normality of the variables was investigated by visual (histograms, probability plots) and analytical methods (Kolmogorov-Smirnov and Shapiro-Wilk tests). Descriptive statistics were presented as mean ± standard deviation, median, range (minimum-maximum), interquartile range (IQR), and frequencies. The chi-square test or Fisher's exact test (based on whether chi-square test assumptions hold) was used to compare proportions in different groups. p < 0.05 was considered to show a statistically significant result.
3 Results
There were 1841 patients with confirmed IMD treated at our center during the study period, whose electronic health records could be accessed remotely, 248 (13.5%) of whom had tested positive for SARS-CoV-2. 223 of these patients gave consent for inclusion in the study (131 children and 92 adults) (Fig. 1).
3.1 General patient characteristics
The demographic characteristics and comorbidities of the whole study population, and also separately as children and adults are shown in Table 1 . Phenylalanine hydroxylase (PAH, n = 108, 48.4%) and biotinidase (n = 27, 12.1%) deficiencies were the most common diagnoses, followed by mucopolysaccharidoses (n = 16, 7.2%). Diagnoses and ICIMD classification of all COVID-19 positive patients are shown in Table 2 . 98 (90.7%) of PAH and 24 (88.9%) of biotinidase deficiency patients were diagnosed via newborn screening. Newborn screening is not available in Turkey for the other IMD mentioned in this study.Table 1 Characteristics of the patients and COVID-19 (N = 223, unless indicated otherwise).
Table 1Patient characteristics Child Adult All group
58.7% (n = 131) 41.1% (n = 92)
Age (years) Median (IQR) 9 (5–13) 25 (21−30) 14 (7–23)
Min-max 0.5–17 18–72 0.5–72
Sex Male 48.1% (n = 63) 43.5% (n = 40) 51.6% (n = 115)
Female 51.9% (n = 68) 56.5% (n = 52) 48.4% (n = 108)
Educational attainment of the patients None 40.5% (n = 53) 18.5% (n = 17) 31.4% (n = 70)
Primary school 29% (n = 38) 9.9% (n = 9) 21.1% (n = 47)
Secondary school 19.1% (n = 25) 6.5% (n = 6) 13.9% (n = 31)
High school 11.5% (n = 15) 44.6% (n = 41) 25.1% (n = 56)
University or higher – 20.7% (n = 19) 8.5% (n = 19)
Highest educational attainment of the parents (n = 217) (n = 128) (n = 89)
None: 1.6% (n = 2) 2.2% (n = 2) 1.8% (n = 4)
Primary school: 27.3% (n = 35) 55.1% (n = 49) 38.7% (n = 84)
Secondary school: 12.5% (n = 16) 11.2% (n = 10) 12.0% (n = 26)
High school: 30.5% (n = 39) 19.1% (n = 17) 25.8% (n = 56)
University or higher: 28.1% (n = 36) 12.4% (n = 11) 21.7% (n = 47)
Specific metabolic treatment Metabolic diet only 35.1% (n = 46) 68.5% (n = 63) 48.9% (n = 109)
Medication only 30.5% (n = 40) 8.7% (n = 8) 21.5% (n = 48)
Metabolic diet and medication 12.2% (n = 16) 14.1% (n = 13) 13.0% (n = 29)
None 22.1% (n = 29) 8.7% (n = 8) 16.6% (n = 37)
Comorbidities No 66.4% (n = 87) 55.4% (n = 51) 61.9% (n = 138)
Yes 33.6% (n = 44) 44.6% (n = 41) 38.1% (n = 85)
Neurologic impairment 19.1% (n = 25) 26.1% (n = 24) 22.0% (n = 49)
Obesity 5.3% (n = 7) 15.2% (n = 14) 9.4% (n = 21)
Cardiovascularδ 7.6% (n = 10) 3.3% (n = 3) 5.8% (n = 13)
Endocrine system 3.1% (n = 4) 2.2% (n = 2) 2.7% (n = 6)
Respiratory systemρ 3.8% (n = 5) 1.1% (n = 1) 2.7% (n = 6)
Renal and urinary systemγ 2.3% (n = 3) 2.2% (n = 2) 2.2% (n = 5)
Rheumatologicω 0.8% (n = 1) 3.3% (n = 3) 1.8% (n = 4)
Hypertension 1.5% (n = 2) 2.2% (n = 2) 1.8% (n = 4)
Lymphoma 0.8% (n = 1) – 0.4% (n = 1)
Immunosuppressive treatment (Post- HSCT) 0.8% (n = 1) – 0.4% (n = 1)
History related to COVID-19
Source of SARS-CoV-2 infection Home 51.9% (n = 68) 40.2% (n = 37) 47.1% (n = 105)
School/work 12.2% (n = 16) 5.4% (n = 5) 9.4% (n = 21)
Nosocomial 3.1% (n = 4) 6.5% (n = 6) 4.5% (n = 10)
Unknown 32.8% (n = 43) 47.8% (n = 44) 39.0% (n = 87)
Timing of SARS-CoV-2 infection During delta variant wave 29.8% (n = 39) 28.3% (n = 26) 29.1% (n = 65)
During omicron variant wave 31.3% (n = 41) 22.8% (n = 21) 27.8% (n = 62)
Other 38.9% (n = 51) 48.9% (n = 45) 43.0% (n = 96)
Type of SARS-CoV-2 test PCR: 96.2% (n = 126) 100% (n = 92) 97.8% (n = 218)
Antibody: 3.8% (n = 5) – 2.2% (n = 5)
Hospitalization Yes 13% (n = 17) 8.7% (n = 8) 11.2% (n = 25)
No 87% (n = 114) 91.3% (n = 84) 88.8% (n = 198)
Duration of hospitalization (days) (n = 25) Median: 6 Median: 6 Median: 6
IQR: 5–7 IQR: 1–9 IQR: 5–8
Min-max: 2–30 Min-max: 1–13 Min-max:1–30
Oxygen requirement Yes 3.8% (n = 5) 3.3% (n = 3) 3.5% (n = 8)
No: 96.2% (n = 126) 96.7% (n = 89) 96.5% (n = 215)
Intensive care unit admission Yes 1.5% (n = 2) 1.1% (n = 1) 1.3% (n = 3)
No 98.5% (n = 129) 98.9% (n = 91) 98.7% (n = 220)
Invasive mechanical ventilation Yes 0.8% (n = 1) 1.1% (n = 1) 0.9% (n = 2)
No 99.2% (n = 130) 98.9% (n = 91) 99.1% (n = 221)
COVID-19 related death Yes 0.8% (n = 1) 1.1% (n = 1) 0.9% (n = 2)
No 99.2% (n = 130) 98.9% (n = 91) 99.1% (n = 221)
Medication during COVID-19 infection No 51.1% (n = 67) 33.7% (n = 31) 43.9% (n = 98)
Yes 48.9% (n = 64) 66.3% (n = 61) 56.1% (n = 125)
Antipyretic/analgesic 43.5% (n = 57) 35.9% (n = 33) 40.5% (n = 90)
Favipiravir 4.6% (n = 6) 52.2% (n = 48) 24.3% (n = 54)
Antibiotics 16.0% (n = 21) 7.6% (n = 7) 12.6% (n = 28)
Hydroxychloroquine sulfate none 13% (n = 12) 5.4% (n = 12)
Anticoagulant/antithrombotic 1.5% (n = 2) 4.3% (n = 4) 2.7% (n = 6)
Dexamethasone 2.3% (n = 3) 1.1% (n = 1) 1.8% (n = 4)
Intravenous immune globulin 2.3% (n = 3) – 1.4% (n = 3)
Oseltamivir 1.5% (n = 2) – 0.9% (n = 2)
Feeding difficulties Yes 28.2% (n = 37) 28.3% (n = 26) 28.3% (n = 63)
No 71.8% (n = 94) 71.7% (n = 66) 71.7% (n = 160)
Interruption of metabolic diet (n = 119) (n = 53) (n = 66)
Yes 32% (n = 17) 31.8% (n = 21) 31.9% (n = 38)
No 68% (n = 36) 68.2% (n = 45) 68.1% (n = 81)
Interruption of metabolic medication (n = 100) (n = 71) (n = 29)
Yes 16.9% (n = 12) 13.7% (n = 4) 16.0% (n = 16)
No 83.1% (n = 59) 863% (n = 25) 84.0% (n = 84)
Vaccination status before contracting COVID-19 (≥12 years, n = 138) (n = 45) (n = 93)
Vaccinated 15.6% (n = 7) 41.9% (n = 39) 33.3% (n = 46)
Unvaccinated 84.4% (n = 38) 58.1% (n = 54) 67.7% (n = 92)
Vaccination status, overall (≥ 12 years, n = 138) (n = 45) (n = 93)
Vaccinated 35.6% (n = 16) 77.4% (n = 72) 63.8% (n = 88)
Unvaccinated 64.4% (n = 29) 22.6% (n = 21) 36.2% (n = 50)
COVID-19: coronavirus disease 2019, HSCT: hematopoietic stem cell transplantation, IQR: interquartile range, PCR: polymerase chain reaction, SARS-CoV-2: severe acute respiratory syndrome coronavirus 2.
δ Congestive heart failure (n = 11), coronary artery disease (n = 1), mitral valve prolapse (n = 1).
ρ asthma (n = 1), tracheostomy (n = 2), non-invasive mechanical ventilation (n = 3).
ω Familial Mediterranean fever (n = 3), systemic lupus erythematosus (n = 1).
γ renal failure (n = 5).
Table 2 The diagnosis and ICIMD classifications of COVID-19-positive patients.
Table 2IMD Diagnosis ICIMD subcategory ICIMD category Child n (%) Adult n (%) Overall n (%)
Phenylalanine hydroxylase deficiency: Disorders of phenylalanine and tyrosine metabolism Disorders of amino acid metabolism 108 (48.4)
PKU on dietary treatment(n = 77) n = 26 (50) n = 51 (91.1)
HPA not requiring treatment (n = 21) n = 16 (30.8) n = 5 (8.9)
BH4-responsive PKU (n = 10) n = 10 (19.2)
Biotinidase deficiency Disorders of biotin metabolism Disorders of vitamin and cofactor metabolism 24 (18.3) 3 (3.3) 27 (12.1)
Mucopolysaccharidosis Disorders of glycosaminoglycan degradation Disorders of complex molecule degradation 10 (7.6) 6 (6.5) 16 (7.2)
Dyslipidemia Disorders of lipoprotein metabolism 5 (3.8) 5 (5.4) 10 (4.5)
Maple syrup urine disease# Branched-chain amino acids Disorders of amino acid metabolism 6 (4.6) 3 (3.3) 9 (4.0)
Methylmalonic acidemia# Organic acidurias Disorders of amino acid metabolism 4 (3.1) 2 (2.2) 6 (2.7)
Homocystinuria⁎ Sulfur-containing amino acids Disorders of amino acid metabolism 2 (1.5) 3 (3.3) 5 (2.2)
Glutaric aciduria type 1# Organic acidurias Disorders of amino acid metabolism 2 (1.5) 2 (2.2) 4 (1.8)
Hereditary fructose intolerance⁎ Disorders of galactose and fructose metabolism Disorders of carbohydrate metabolism 2 (1.5) 2 (2.2) 4 (1.8)
Fructose-1,6-diphosphatase deficiency⁎, # Disorders of gluconeogenesis Disorders of carbohydrate metabolism 2 (1.5) – 2 (0.9)
Ornithine transcarbamylase deficiency⁎, # Disorders of the urea cycle and hyperammonemias Disorders of amino acid metabolism – 2
(2.2) 2 (0.9)
Isovaleric acidemia⁎, # Organic acidurias Disorders of amino acid metabolism 1 (0.8) 1 (1.1) 2 (0.9)
Lysinuric protein intolerance⁎# Amino acid transport Disorders of amino acid metabolism 1 (0.8) 1 (1.1) 2 (0.9)
HMG-CoA lyase deficiency# Disorders of ketone body metabolism Disorders of fatty acid and ketone metabolism 2 (1.5) – 2 (0.9)
X-linked adrenoleukodystrophy⁎ Disorders of peroxisomal fatty acid oxidation Disorders of lipid metabolism 1 (0.8) – 1 (0.4)
Alkaptonuria⁎ Disorders of phenylalanine and tyrosine metabolism Disorders of amino acid metabolism 1 (0.8) – 1 (0.4)
Mitochondrial acetoacetyl-CoA thiolase deficiency⁎# Disorders of ketone body metabolism Disorders of fatty acid and ketone metabolism 1 (0.8) – 1 (0.4)
Chanarin-Dorfman syndrome⁎ Disorders of glycerolipid metabolism Disorders of lipid metabolism 1 (0.8) – 1 (0.4)
Dihydropteridine reductase deficiency⁎ Disorders of tetrahydrobiopterin metabolism Disorders of vitamin and cofactor metabolism 1 (0.8) – 1 (0.4)
Mitochondrial and cytoplasmic glycyl-tRNA synthetase deficiency⁎ Disorders of mitochondrial aminoacyl-tRNA synthetases Disorders of mitochondrial gene expression – 1
(1.1) 1 (0.4)
Hepatic glycogen synthase deficiency⁎, # Disorders of glycogen metabolism Disorders of carbohydrate metabolism 1 (0.8) – 1 (0.4)
Hyperprolinemia type 2⁎ Disorders of ornithine, proline and hydroxyproline metabolism Disorders of amino acid metabolism 1 (0.8) – 1 (0.4)
HUPRA syndrome (SARS2)⁎ Disorders of mitochondrial aminoacyl-tRNA synthetases Disorders of mitochondrial gene expression 1 (0.8) – 1 (0.4)
ITPA deficiency⁎ Disorders of purine metabolism Disorders of nucleobase, nucleotide and nucleic acid metabolism 1 (0.8) – 1 (0.4)
Krabbe disease⁎ Disorders of sphingolipid degradation Disorders of complex molecule degradation 1 (0.8) – 1 (0.4)
L-2-hydroxyglytaric aciduria⁎ Disorders of mitochondrial metabolite repair Disorders of metabolite repair/proofreading – 1
(1.1) 1 (0.4)
LCHAD deficiency# Disorders of mitochondrial fatty acid oxidation Disorders of fatty acid and ketone metabolism 1 (0.8) – 1 (0.4)
LPIN1 deficiency⁎, # Disorders of glycerolipid metabolism Disorders of lipid metabolism 1 (0.8) – 1 (0.4)
MEGDEL syndrome⁎, # Disorders of mitochondrial membrane biogenesis and remodeling Disorders of organelle biogenesis, dynamics and interactions 1 (0.8) – 1 (0.4)
Pompe disease Other disorders of complex molecule degradation Disorders of complex molecule degradation 1 (0.8) – 1 (0.4)
Propionic acidemia# Organic acidurias Disorders of amino acid metabolism 1 (0.8) – 1 (0.4)
PTPS deficiency⁎ Disorders of tetrahydrobiopterin metabolism Disorders of vitamin and cofactor metabolism 1 (1.1) 1 (0.4)
Hypomyelinating leukodystrophy-10 (PYCR2)⁎ Disorders of ornithine, proline and hydroxyproline metabolism Disorders of amino acid metabolism 1 (0.8) – 1 (0.4)
Trimethylaminuria⁎ Disorders of methylamine metabolism Disorders of peptide and amine metabolism 1 (0.8) – 1 (0.4)
Tyrosinemia type 1⁎ Disorders of phenylalanine and tyrosine metabolism Disorders of amino acid metabolism 1 (0.8) – 1 (0.4)
Tyrosinemia type 2⁎ Disorders of phenylalanine and tyrosine metabolism Disorders of amino acid metabolism 1 (0.8) – 1 (0.4)
Citrullinemia type-1⁎, # Disorders of the urea cycle and hyperammonemias Disorders of amino acid metabolism 1 (0.8) – 1 (0.4)
Combined oxidative phosphorylation deficiency 20 (VARS2)⁎, # Disorders of mitochondrial aminoacyl-tRNA synthetases Disorders of mitochondrial gene expression 1 (0.8) – 1 (0.4)
Total 131 (100) 92 (100) 223 (100)
BH4: Tetrahydrobiopterin, HMG-CoA: 3-hydroxy-3-methylglutaryl coenzyme A, HPA: hyperphenylalaninemia, HUPRA: Hyperuricemia, pulmonary hypertension, renal failure, and alkalosis, ICIMD: International Classification of Inherited Metabolic Disorders, IMD: Inherited metabolic disorders, ITPA: Inosine triphosphatase, LCHAD: Long-chain 3-hydroxyacyl-CoA dehydrogenase, MEGDEL: 3-methylglutaconic aciduria, deafness, encephalopathy, and Leigh-like disease, PKU: phenylketonuria, PTPS: 6-pyruvoyl-tetrahydropterin synthase.
⁎ This paper is the first published record of COVID-19 in these disorders.
# Disorders in which acute symptomatic metabolic decompensations are possible. Hereditary fructose intolerance, biotinidase deficiency and tyrosinemia type 1 were not included because the risk of decompensation is very low under treatment.
3.2 Features of COVID-19 in IMD
83.9% (n = 187) of patients had symptoms during COVID-19 and 25.1% (n = 56) had long COVID symptoms (Fig. 2 ). 12 patients (5%) contracted COVID-19 twice. The majority of the cases were either asymptomatic (16.1%, n = 36) or had a mild illness (77.6%, n = 173), but 6 patients (2.7%) had moderate, 6 had severe, and two (0.9%) had critical COVID-19 (Fig. 3 ), both of whom died. More detailed information regarding the 14 patients with moderate or more severe COVID-19 are presented in Table 3 . 13.0% of children and 8.7% of adults with IMD were hospitalized during COVID-19. The patients with comorbid conditions had 6.5 times higher odds of being hospitalized than those without (p < 0.001). The significantly higher risk of hospitalization in patients with comorbidities persisted when patients were analyzed separately as children and adults: The odds of hospitalization were 4.4 times higher in the pediatric (p < 0.01), and 20.5 times higher in adult (p < 0.01) patients with comorbidities. Three patients (one each of 3-hydroxy-3-methylglutaryl-coenzyme A [HMG-CoA] lyase deficiency, long-chain 3-hydroxyacyl-CoA dehydrogenase [LCHAD] deficiency, and methylmalonic academia [MMA]) had an acute metabolic decompensation during the infection. All of these three patients were hospitalized because of the metabolic decompensation, and were treated successfully during the course of COVID-19. Standard treatment protocols for metabolic decompensation were administered. Two children (1.5% of children) developed MIS-C and recovered with treatment (one with LCHAD and the other with biotinidase deficiency) with intravenous immunoglobulin, anakinra, favipiravir, anticoagulants anti-inflammatory agents and antibiotics.Fig. 2 Occurrence of symptoms during A) COVID-19 infection, and B) long COVID shown separately in the pediatric and adult groups, and in the whole patient group. A. There were a total of 223 patients (131 children and 92 adults), 187 of whom were symptomatic (108 children and 79 adults). B. 56 patients (35 children and 21 adults) experienced long COVID symptoms.
Fig. 2
Fig. 3 Severity distribution of COVID-19 according to the World Health Organization severity classification in A) pediatric patients, B) adult patients, and C) all the patients with inherited metabolic disorders in the study population.
Fig. 3
Table 3 Features of patients with moderate, severe or critical COVID-19.
Table 3Case No Age, Gender Diagnosis Comorbidities Fully vaccinated at the time of COVID-19 COVID-19 severity ICU admission Hospitalization (days) Oxygen requirement Status
1 11 y, F MPS-VI Cardiac failure No Moderate No 7 No Alive
2 20 y, M Classical PKU⁎ Obesity No Moderate No 5 No Alive
3 30 y, M Classical PKU Neurologic impairment No Moderate No 6 No Alive
4 2 y, M HMG-CoA lyase deficiency None No Moderate No 5 No Alive
5 25 y, M Classical PKU⁎ Obesity No Moderate No 3 No Alive
6 9 y, M BH4-responsive PKU⁎ None No Moderate No 0 No Alive
7 2 y, M Mild HPA⁎ Prematurity and asthma No Severe No 7 Yes Alive
8 25 y, M HFI FMF Yes Severe No 13 Yes Alive
9 18 y, M FH (AD) Obesity No Severe No 10 Yes Alive
10 16 y, F MPS-III Neurologic impairment, tracheostomy No Severe No 5 Yes Alive
11 3 y, F Krabbe disease Neurologic impairment, tracheostomy No Severe Yes 30 Yes Alive
12 6 mo, M MPS-I (Hurler) None No Severe No 6 Yes Alive
13 12 mo, F Tyrosinemia type I Liver failure, rachitic pneumopathy No Critical Yes 11⁎⁎ Yes Exitus
14 72 y, M Classical PKU Obesity, ID, HT, PD No Critical Yes 1 Yes Exitus
COVID-19: coronavirus disease 2019, FH (AD): familial hypercholesterolemia (autosomal dominant), FMF: familial Mediterranean fever, HFI: hereditary fructose intolerance, HMG-CoA: 3-hydroxy-3-methylglutaryl-coenzyme A, HPA: hyperphenylalaninemia, HT: hypertension, ICU: intensive care unit, ID: intellectual disability, MPS: mucopolysaccharidosis, PD: Parkinson's disease, PKU: phenylketonuria.
⁎ Patients diagnosed via newborn screening.
⁎⁎ She was already hospitalized at the time of contracting COVID-19. 11 days is the duration of hospitalization after the diagnosis of COVID-19.
Mean duration of interruption of school or work was 13 ± 5 days. The presence of symptoms, clinical spectrum of COVID-19, long COVID symptoms, hospitalization, and intensive care unit admission status were not significantly different between children and adults (Table 4 ). In order to determine which ICIMD categories were more likely to experience more severe COVID-19, the patients were grouped into two, according to the absence or presence of pneumonia (asymptomatic-mild vs. moderate-severe-critical COVID-19). There was a statistically significant difference between those with disorders of complex molecule degradation and the others (p = 0.018). The patients with complex molecule degradation disorders had 4.5 times higher odds of COVID-19 pneumonia compared to the other groups. When analyzed separately for age groups, the significance persisted in the pediatric group, as there was a statistically significant difference between complex molecule degradation disorders and the others (p < 0.01). The children with complex molecule disorders had 9.7 times higher odds of clinical severity. However, in adult patients (n = 92), there was no statistically significant difference between ICIMD categories in terms of COVID-19 severity (p = 1). There were no significant differences between the other categories of IMD. Within the whole group, IMD patients with comorbidities had 4.4 times higher odds to experience more severe COVID-19 (pneumonia) than those without (p = 0.011). This difference was more evident in adult IMD patients with comorbidities (odds ratio: 20.7, p < 0.01), but was not statistically significant in the pediatric age group (p = 0.45).Table 4 Comparison of COVID-19 features between children and adults with IMD.
Table 4 Child (n = 131) Adult (n = 92) p
1 Presence of symptoms Yes: 82.4% (n = 108)
No: 17.6% (n = 23) Yes: 85.9% (n = 79)
No: 14.1% (n = 13) 0.49
2 Clinical Spectrum of SARS-CoV-2 Infection
Asymptomatic: 16.8% (n = 22) 15.2% (n = 14) 0.97
Mild illness: 77.1% (n = 101) 78.3% (n = 72)
Moderate illness: 2.3% (n = 3) 3.3% (n = 3)
Severe illness: 3.1% (n = 4) 2.2% (n = 2)
Critical illness: 0.8% (n = 1) 1.1% (n = 1)
3 Long COVID Yes:26.8% (n = 35)
No:73.2% (n = 96) Yes: 22.8% (n = 21)
No: 77.2% (n = 71) 0.48
4 Hospitalization Yes:13.0% (n = 17)
No: 87.0% (n = 114) Yes: 8.7% (n = 8)
No: 91.3% (n = 84) 0.32
5 ICU admission Yes: 1.5% (n = 2)
No: 98.5% (n = 129) Yes: 1.1% (n = 1)
No: 98.9% (n = 91) 1
COVID-19: coronavirus disease 2019, ICU: intensive care unit, IMD: inherited metabolic disorder, SARS-CoV-2: severe acute respiratory syndrome coronavirus 2.
Status of vaccination is shown in Table 1. 46 patients (33.3% of those eligible) had been vaccinated before they contracted COVID-19. Among all the vaccinated patients (n = 88), 72.7% (n = 64) had been vaccinated with the nucleoside modified messenger RNA vaccine BNT162b2 (Comirnaty®) by BioNTech, 25.0% (n = 22) with inactivated SARS-CoV-2 vaccine CoronaVac® by Sinovac, and 2.3% (n = 2) with both. Only 1 out of 14 patients with moderate-severe-critical COVID-19 had been fully vaccinated at the time of infection (Table 3). Patient or parental education level was not significantly different between those who were and were not vaccinated (p = 0.22).
4 Discussion
In this paper, we have presented the summary of the features of COVID-19 in a large cohort of IMD patients from a single center. Our findings suggest that the risk of metabolic decompensation is not disproportionately high, and children with disorders of complex molecule degradation may be under higher risk of more severe COVID-19. To the best of our knowledge, this is the largest single-center series so far reported, and the first account of COVID-19 in 27 different IMD (Table 2), the first reports of detailed COVID-19 symptomatology, vaccination status and MIS-C in IMD. We demonstrate that the symptomatology of COVID-19 in IMD is similar to the general population [12], but the occurrence of feeding difficulties may be slightly higher in children with IMD (28.2% vs. 20%) [13].
One critical finding of our study is the demonstration of increased risk of more severe COVID-19 in children with disorders of complex molecule degradation, many of which are lysosomal storage diseases (LSD) that are often characterized by progressive multisystem involvement [2,14]. In adults with IMD, the severity of COVID-19 was associated more significantly with comorbidities rather than the category of IMD. While critical comorbidities (involvement of the lungs, heart etc.) commonly associated with LSD are known risk factors for severe COVID-19, our data on children with IMD suggest that there may be further susceptibility to SARS-CoV-2 at the cellular or molecular level in patients with inherited disorders of complex molecule degradation.
Infections are well-known triggers of acute metabolic decompensation in some metabolic disorders, especially intoxication-type and energy-deficit disorders [15]. In pre-COVID studies of patients with MMA and propionic acidemia, infections triggered metabolic decompensation in 29% and 44% of the time, respectively [[16], [17], [18]]. We had 37 COVID-19-positive patients at risk of an acute metabolic decompensation (marked with # in Table 2), but SARS-CoV-2-related decompensation occurred only in three (8.1%). Zubarioğlu et al. reported four counts of SARS-CoV-2-related decompensation in 22 IMD patients (18.1%) [19]. While the vulnerability of some IMD to infections is evident, we argue that SARS-CoV-2 itself does not pose an extra burden of an acute decompensation compared to other infectious agents circulating before the pandemic [8,20]. It should be recognized that immune-naïve patients may experience a more severe illness and therefore carry a higher risk of metabolic decompensation. It is important to underline that preventive measures such as vaccination, hygiene and social distancing should be recommended as in any other infectious disease (eg. influenza), which are a part of routine care in the IMD that carry such risks [21].
While the hospitalization and death rates of children with COVID-19 were < 2% and < 0.03% in a general population study [22], they were 13% and 0.8% in our study, respectively. In a large survey study from Brazil, nine out of 358 patients (2.5%) with IMD were hospitalized, three in the intensive care unit [23]. 12 of our patients were hospitalized not due to the severity of COVID-19, but primarily because of their IMD; two due to acute decompensation, and the 10 remaining patients due to the perception of higher risk of COVID-19 or for close observation anticipating a decompensation. In-hospital care may have contributed to the low attack rate of an acute decompensation in our study, which emphasizes the importance of continued health care for rare disease patients during the pandemic. The concern raised by MetabERN regarding higher COVID-19 risk and IMD exacerbation in patients with amino acid and organic acid-related disorders was formulated at a time when no vaccines were available for a disease we had merely begun to understand [24]. Now, with the tools and information acquired in the meantime, we propose that efforts should focus not pre-emptive intensive treatment of IMD patients contracting COVID-19, but rather on application of well-established “sick-day” regimens, and on vaccinating individuals with IMD, even young children, to prevent as many acute decompensations as possible.
There were two COVID-19-related deaths in the study population (Table 3), but it is hard to claim that the deaths are directly related to the IMD. It is known that elderly people and infants may have a critical disease course [25]. The deceased phenylketonuria patient was a 74-year-old obese man with hypertension and Parkinson's disease. The other deceased patient was an infant who already had the complications of tyrosinemia type 1 (acute liver failure, chronic liver disease and hypophosphatemic rickets) when she was diagnosed at the age of nine months. She had respiratory muscle weakness, rachitic beads, massive ascites and hepatomegaly, reducing her vital lung capacity. Both patients were diagnosed with IMD after exhibiting clinical symptoms. It is likely that newborn screening and early treatment for tyrosinemia type 1 may have prevented the patient's death due to COVID-19.
All of our eight patients with severe or critical COVID-19 (3 adults and 5 children) had risk factors (eg. old age, infancy, neurological impairment, obesity, prematurity), some of which were caused by the underlying IMD (eg. neurological impairment in MPS-III and Krabbe disease). It is well known that pediatric patients in the general population have milder COVID-19 than adults [7,26,27], attributed to multiple factors [28]. Severe or critical disease risk is reported as 3.9% in the general Turkish pediatric population with SARS-CoV-2 infection [29], and we calculated the same risk as 3.8% in our pediatric cohort. Risk of severe disease was similar to adults with IMD, and the hospitalization rate was higher in children (Table 4), although this may be partly related to physician preference, as discussed above. The MetabERN survey conducted in early 2020 when vaccines were unavailable and children were largely staying at home reported that adult patients required more hospitalization and had a more severe course compared to the general population and pediatric patients [6], but most of our cases coincide with delta and omicron waves of the pandemic, when many adults were vaccinated and children were attending school. While the inadequate numbers of severe and critical cases and the fluidity of the pandemic conditions make it difficult to draw definitive conclusions, children with IMD in general do not seem to have increased risk of severe COVID-19 than their healthy counterparts. Studies with larger numbers of IMD patients are needed for this comparison.
The long-term effects of SARS-CoV-2 infection are being increasingly recognized and emphasized [30]. Studies from different countries have reported highly variable incidences of long-COVID, ranging from 1.6% to 77% [30]. In our cohort, 25.2% of patients had long COVID, and this rate was similar between children and adults. The most common long-COVID symptom in our patients was fatigue. These findings are in line with previously published work regarding the general population [31,32]. While long-COVID is common, MIS-C is rare, with a reported incidence 1:4000 [33]. MIS-C developed in two out of 131 children included in our study. These may be the first published accounts of MIS-C in IMD. The high occurrence of MIS-C in our cohort may be coincidental, but warrants further study.
This study has several limitations, firstly related to patient interviews and the retrospective design, making it prone to recall bias. Screening protocols for asymptomatic contacts of a case changed during the course of the pandemic, possibly missing asymptomatic infections. The intended study population could not be completely enrolled because the electronic records of some patients were inaccessible, and some patients could not be contacted. Still, from a large group of patients using established, objective definitions and classifications, describing COVID-19 and long-COVID symptomatology in IMD, presenting first accounts of COVID-19 in many IMD and of MIS-C.
In conclusion, children with IMD may have a higher rate of hospitalization during COVID-19, perhaps partly because of physicians' preference of precautionary close follow-up. However, we think that the IMD patients under risk of decompensations can be assured that there is no evidence to suggest that the occurrence risk or the severity of decompensations is increased in SARS-CoV-2 infection compared to other infections, and adherence to standard emergency protocols and infection prevention measures, such as vaccination, should be recommended. Facilitated access to health services when needed remains an important challenge. Children with disorders of complex molecule degradation have a higher risk of more severe COVID-19, possibly due to multisystem involvement, but increased susceptibility to COVID-19 at the cellular or molecular should be investigated in future studies. Patients, especially adults with IMD may experience more severe COVID-19 if they have high-risk comorbidities.
Individuals with IMD and other rare diseases cannot be ignored. They should be kept in mind while formulating recommendations and making policy decisions. All patients with IMD older than 6 months should be eligible for vaccination against SARS-CoV-2 if their condition is a complex molecule degradation disorder, or is characterized by acute metabolic decompensations or chronic high-risk organ involvement. Every effort should be made to sustain acute and long-term health services for these vulnerable individuals during the pandemic.
Ethics approval and consent to participate
This study was approved by Hacettepe University Ethics Committee for Non-Interventional Clinical Studies (GO 22/176, 2022/08-23). Informed consent was obtained from all patients or their parents/guardians.
Consent for publication
All patients provided consent to publish clinical data.
Competing interests
The authors declare that they have no competing interests.
Funding
The authors received no financial support for the research.
Authors' contributions
Conception and design: ABK, YY, SS.
Clinical evaluation and records: ABK, YY, KÇ, İE, HTA, AD, AT, SS.
Data collection: ABK, KÇ, İE, HTA.
Data analysis: ABK.
Data interpretation: ABK, YY.
Drafting the manuscript: ABK, YY.
Editing the manuscript: YY, KÇ, İE, HTA, AD, AT, SS. All authors read and approved the final manuscript.
Declaration of Competing Interest
None.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
==== Refs
References
1 Dessie Z.G. Zewotir T. Mortality-related risk factors of COVID-19: a systematic review and meta-analysis of 42 studies and 423,117 patients BMC Infect. Dis. 21 2021 855 34418980
2 Ferreira C.R. Rahman S. Keller M. Zschocke J. Group I.A. Abdenur J. Ali H. Artuch R. Ballabio A. Barshop B. An international classification of inherited metabolic disorders (ICIMD) J. Inherit. Metab. Dis. 44 2021 164 177 33340416
3 Mercolini F. Donà D. Girtler Y. Mussner K.A. Biban P. Bordugo A. Molinaro G. First paediatric COVID-19 associated death in Italy J. Paediatr. Child Health. 57 5 2021 May 736 737 32592522
4 Kahraman A.B. Yildiz Y. Ciki K. Akar H.T. Erdal I. Dursun A. Tokatli A. Sivri H.S. Invisible burden of COVID-19: enzyme replacement therapy disruptions J. Pediatr. Endocrinol. Metab. 34 2021 539 545 33818036
5 Koç Yekedüz M. Doğulu N. Sürücü Kara İ. Öncül Ü. Bakirarar B. Kullu P. Ar Y. Köse E. Eminoğlu F.T. Pros and Cons of Telemedicine for Inherited Metabolic Disorders in a Developing Country during the COVID-19 Pandemic Telemedicine and e-Health 2022
6 Paneghetti L. Bellettato C.M. Sechi A. Stepien K.M. Scarpa M. One year of COVID-19: infection rates and symptoms in patients with inherited metabolic diseases followed by MetabERN Orphan. J. Rare Dis. 17 2022 1 8
7 Hasnat F. Noman F. Moben A. Sarker A. Morshed J. Mutanabbi M. Difference in clinical patterns between COVID-19 Affected Childr. Adults Mymensingh Med. J. 30 2021 1093 1099
8 Mütze U. Gleich F. Barić I. Baumgartner M. Burlina A. Chapman K.A. Chien Y.H. Cortès-Saladelafont E. De Laet C. Dobbelaere D. Impact of the SARS-CoV-2 pandemic on the health of individuals with intoxication-type metabolic diseases—data from the E-IMD consortium J. Inherit. Metab. Dis. 46 2023 220 231 36266255
9 Living guidance for clinical management of COVID-19 2021
10 International Guidelines for Certification and Classification (Coding) of COVID-19 as Cause of Death 2020
11 Centers for Disease Control and Prevention Emergency Pre- Paredness and Response: Health Alert Network 2020
12 Cui X. Zhao Z. Zhang T. Guo W. Guo W. Zheng J. Zhang J. Dong C. Na R. Zheng L. Li W. Liu Z. Ma J. Wang J. He S. Xu Y. Si P. Shen Y. Cai C. A systematic review and meta-analysis of children with coronavirus disease 2019 (COVID-19) J. Med. Virol. 93 2021 1057 1069 32761898
13 Assa A. Benninga M.A. Borrelli O. Broekaert I. de Carpi J.M. Saccomani M.D. Dolinsek J. Mas E. Miele E. Thomson M. Tzivinikos C. E. Gastrointestinal Committee of Gastrointestinal perspective of coronavirus disease 2019 in children-an updated review J. Pediatr. Gastroenterol. Nutr. 73 2021 299 305 34117195
14 Platt F.M. d’Azzo A. Davidson B.L. Neufeld E.F. Tifft C.J. Lysosomal storage diseases Nat. Rev. Dis. Prim. 4 2018 1 25 29930242
15 Ghosh A. Morris A.A. Acute presentations of inherited metabolic disorders: investigation and initial management Paediatr. Child Health 29 2019 99 104
16 Zwickler T. Riderer A. Haege G. Hoffmann G.F. Kolker S. Burgard P. Usefulness of biochemical parameters in decision-making on the start of emergency treatment in patients with propionic acidemia J. Inherit. Metab. Dis. 37 2014 31 37 23797949
17 Zwickler T. Haege G. Riderer A. Horster F. Hoffmann G.F. Burgard P. Kolker S. Metabolic decompensation in methylmalonic aciduria: which biochemical parameters are discriminative? J. Inherit. Metab. Dis. 35 2012 797 806 22249333
18 Forny P. Hörster F. Ballhausen D. Chakrapani A. Chapman K.A. Dionisi-Vici C. Dixon M. Grünert S.C. Grunewald S. Haliloglu G. Guidelines for the diagnosis and management of methylmalonic acidaemia and propionic acidaemia: first revision J. Inherit. Metab. Dis. 44 2021 566 592 33595124
19 Zubarioglu T. Hopurcuoglu D. Ahmadzada S. Uzunyayla-Inci G. Cansever M.S. Kiykim E. Aktuglu-Zeybek C. Inborn errors of metabolism and COVID-19: Evaluation of the metabolic outcome, Pediatr Int. 64 1 2022 Jan e14938 34329521
20 Kaur S. Campbell S.L. Stockton D.W. Management of COVID-19 infection in organic acidemias Am. J. Med. Genet. A 185 2021 1854 1857 33686767
21 Baumgartner M.R. Hörster F. Dionisi-Vici C. Haliloglu G. Karall D. Chapman K.A. Huemer M. Hochuli M. Assoun M. Ballhausen D. Proposed guidelines for the diagnosis and management of methylmalonic and propionic acidemia Orphan. J. Rare Dis. 9 2014 1 36
22 Zimmermann P. Pittet L.F. Curtis N. How common is long COVID in children and adolescents? Pediatr. Infect. Dis. J. 40 2021 e482
23 Schwartz I.V.D. Randon D.N. Monsores N. Moura de Souza C.F. Horovitz D.D.G. Wilke M.V.M.B. Brunoni D. SARS-CoV-2 pandemic in the Brazilian community of rare diseases: a patient reported survey Am. J. Med. Genet. C: Semin. Med. Genet. 2021 301 311 Wiley Online Library 33474836
24 Lampe C. Dionisi-Vici C. Bellettato C. Paneghetti L. Van Lingen C. Bond S. Brown C. Finglas A. Francisco R. Sestini S. The impact of COVID-19 on rare metabolic patients and healthcare providers: results from two MetabERN surveys Orphan. J. Rare Dis. 15 2020 1 14
25 Choi J.H. Choi S.-H. Yun K.W. Risk factors for severe COVID-19 in children: a systematic review and meta-analysis Journal of Korean Medical Science 37 2022
26 Chen A. Huang J.-x. Liao Y. Liu Z. Chen D. Yang C. Yang R.-m. Wei X. Differences in Clinical and Imaging Presentation of Pediatric Patients with COVID-19 in Comparison with Adults Radiology: Cardiothoracic Imaging Vol. 2 2020
27 Dong Y. Mo X. Hu Y. Qi X. Jiang F. Jiang Z. Tong S. Epidemiology of COVID-19 among Children in China Pediatrics Vol. 145 2020
28 Dhochak N. Singhal T. Kabra S. Lodha R. Pathophysiology of COVID-19: why children fare better than adults? Indian J. Pediatr. 87 2020 537 546 32410003
29 Yayla B.C.C. Aykac K. Ozsurekci Y. Ceyhan M. Characteristics and Management of Children with COVID-19 in a tertiary Care Hospital in Turkey Clin. Pediatr. (Phila) 60 2021 170 177 33034209
30 Raman B. Bluemke D.A. Lüscher T.F. Neubauer S. Long COVID: post-acute sequelae of COVID-19 with a cardiovascular focus Eur. Heart J. 43 2022 1157 1172 35176758
31 Ludvigsson J.F. Case report and systematic review suggest that children may experience similar long-term effects to adults after clinical COVID-19 Acta Paediatr. 110 2021 914 921 33205450
32 Payne A.B. Gilani Z. Godfred-Cato S. Belay E.D. Feldstein L.R. Patel M.M. Randolph A.G. Newhams M. Thomas D. Magleby R. Hsu K. Burns M. Dufort E. Maxted A. Pietrowski M. Longenberger A. Bidol S. Henderson J. Sosa L. Edmundson A. Tobin-D'Angelo M. Edison L. Heidemann S. Singh A.R. Giuliano J.S. Jr. Kleinman L.C. Tarquinio K.M. Walsh R.F. Fitzgerald J.C. Clouser K.N. Gertz S.J. Carroll R.W. Carroll C.L. Hoots B.E. Reed C. Dahlgren F.S. Oster M.E. Pierce T.J. Curns A.T. Langley G.E. Campbell A.P. M.-C.I.A. Group Balachandran N. Murray T.S. Burkholder C. Brancard T. Lifshitz J. Leach D. Charpie I. Tice C. Coffin S.E. Perella D. Jones K. Marohn K.L. Yager P.H. Fernandes N.D. Flori H.R. Koncicki M.L. Walker K.S. Di Pentima M.C. Li S. Horwitz S.M. Gaur S. Coffey D.C. Harwayne-Gidansky I. Hymes S.R. Thomas N.J. Ackerman K.G. Cholette J.M. Incidence of multisystem inflammatory syndrome in children among US persons infected with SARS-CoV-2 JAMA Netw. Open 4 2021 e2116420
33 Holm M Hartling UB Schmidt LS Glenthøj JP Kruse A Rytter MH Lindhard MS Lawaetz MC Zaharov T Petersen JJ Andersen RM Lemvik G Marcinski P Thaarup J Jensen LH Borch L Nielsen AB Vissing NH Schmiegelow K Nygaard U Multisystem inflammatory syndrome in children occurred in one of four thousand children with severe acute respiratory syndrome coronavirus 2 Acta Paediatr 110 9 2021 Sep 2581 2583 10.1111/apa.15985 34129731
|
PMC010xxxxxx/PMC10172971.txt |
==== Front
Curr Opin Virol
Curr Opin Virol
Current Opinion in Virology
1879-6257
1879-6265
Elsevier B.V.
S1879-6257(23)00034-2
10.1016/j.coviro.2023.101334
101334
Article
Adenoviral-vectored next-generation respiratory mucosal vaccines against COVID-19
Afkhami Sam
Kang Alisha
Jeyanathan Vidthiya
Xing Zhou
Jeyanathan Mangalakumari
McMaster Immunology Research Centre, M. G. DeGroote Institute for Infectious Disease Research & Department of Medicine, McMaster University, Hamilton, ON L8S 4K1, Canada
11 5 2023
8 2023
11 5 2023
61 101334101334
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The world is in need of next-generation COVID-19 vaccines. Although first-generation injectable COVID-19 vaccines continue to be critical tools in controlling the current global health crisis, continuous emergence of SARS-CoV-2 variants of concern has eroded the efficacy of these vaccines, leading to staggering breakthrough infections and posing threats to poor vaccine responders. This is partly because the humoral and T-cell responses generated following intramuscular injection of spike-centric monovalent vaccines are mostly confined to the periphery, failing to either access or be maintained at the portal of infection, the respiratory mucosa (RM). In contrast, respiratory mucosal-delivered vaccine can induce immunity encompassing humoral, cellular, and trained innate immunity positioned at the respiratory mucosa that may act quickly to prevent the establishment of an infection. Viral vectors, especially adenoviruses, represent the most promising platform for RM delivery that can be designed to express both structural and nonstructural antigens of SARS-CoV-2. Boosting RM immunity via the respiratory route using multivalent adenoviral-vectored vaccines would be a viable next-generation vaccine strategy.
==== Body
pmc Current Opinion in Virology 2023, 61:101334
This review comes from a themed issue on Mucosal Immunology
Edited by Hiroshi Kiyono and Xiaoping Zhu
https://doi.org/10.1016/j.coviro.2023.101334
1879–6257/© 2023 Elsevier B.V. All rights reserved.
Introduction
Despite the continued rollout of vaccines even with updated formulations, SARS-CoV-2 remains a global health concern. To-date, 11 first-generation COVID-19 vaccines have received Emergency Use Listing by the World Health Organization, with approximately 70% of the global population receiving at least a single dose of any COVID-19 vaccine [1]. Following their initial rollout in 2020, first-generation mRNA COVID-19 vaccines were designed to induce neutralizing antibodies against the spike protein of ancestral SARS-CoV-2, which showed remarkable protection against infection, conferring greater than 90% effectiveness 2, 3. However, waning immunity and the emergence of variants of concern (VOC) harboring numerous mutations in the spike protein have eroded the efficacy of these vaccines. As such, breakthrough infections are commonplace, especially since the emergence of the Delta and Omicron VOC [4].
Nonetheless, vaccine-induced T-cell immunity continues to provide protection against severe disease, hospitalization, and mortality 5••, 6. Although it is known that T-cell immunity lasts longer than humoral immunity, the real longevity of COVID-19 vaccine-induced protective T-cell immunity remains to be seen. As the COVID-19 pandemic transitions to a state of endemicity owing to an accumulation of population-wide immunity from natural infection and vaccination, it is anticipated that there remains a need to maintain the optimal immunity via boost vaccination in general populations. This is particularly relevant to individuals with immune deficiencies, transplant recipients, and the elderly who responded poorly to the first-generation vaccines. The lessons learned from first-generation vaccines indicate that chasing after the evolving SARS-CoV-2 by updating the spike antigen in the vaccines is an unsustainable strategy; rather, next-generation vaccine strategies that aim to boost multilayered immunity, encompassing trained innate, humoral, and cellular T-cell immunity (tripartite) at the respiratory mucosa may provide hope for fortifying the immune system against future emerging variants.
The robust humoral and T-cell responses induced following intramuscular injection are constrained to the periphery, failing to either access or be maintained at the portal of infection, the respiratory mucosa 7, 8, 9••. The significance of positioning immunity at the mucosa for effective protection against pathogens that enter the host via mucosal sites has been well acknowledged. Presently, strategies to develop next-generation SARS-CoV-2 vaccines amenable for respiratory mucosal (RM) delivery are garnering collective interest ( Figure 1) 10, 11, 12, 13. In this article, we provide the framework for the development of next-generation COVID-19 vaccination strategies, with specific consideration given to respiratory mucosal-delivered adenoviral-vectored (AdV) vaccines that can induce tripartite respiratory mucosal immunity capable of not only protecting against SARS-CoV-2 VOC but also reducing viral transmission and mitigating the development of post-acute COVID-19 sequalae (PACS/long COVID).Figure 1 Differential geographical localization and types of immune responses following first- and next-generation COVID-19 vaccines. Current first-generation injectable COVID-19 vaccines generate robust systemic humoral and T-cell responses but suboptimal magnitude of antibody responses and no T-cell responses at the respiratory mucosa compared with natural infection. Importantly, while structural and nonstructural SARS-CoV-2 protein-specific immune responses are induced by natural infection, injectable vaccines are designed to generate only spike-specific immune responses. This is reflected by relatively less efficiency in protecting against infection by first-generation COVID-19 vaccines compared with prior natural infection. In contrast, AdV multivalent COVID-19 vaccines, which are safe and amenable for respiratory mucosal delivery, can induce long-lived tripartite immunity at the respiratory mucosa capable of not only protecting against SARS-CoV-2 VOC but also reducing viral transmission and mitigating the development of PACS/long COVID.
Figure 1
Limited induction of respiratory mucosal immunity by first-generation injectable COVID-19 vaccination
First-generation injectable COVID-19 vaccines continue to be critical tools in combatting the current global health crisis. Indeed, first-generation COVID-19 vaccines induce robust serological IgM, IgG, and IgA antibodies and long-lasting memory B- and T-cell responses (Figure 1) [14]. Given the robust efficacy of these vaccines in the early stages of the pandemic, studies began to define and establish vaccine-induced humoral immunity as the protective correlates 15, 16. This largely steered the field to adopt the notion that serological-neutralizing antibody titers were central in defining vaccine efficacy against SARS-CoV-2. However, realizing that vaccines developed to target SARS-CoV-2 should aim at inducing immunity at geographical locations that match immunity induced by natural infection [17], studies have eventually assessed the ability of first-generation vaccines to induce mucosal immunity. Salivary anti-RBD and anti-spike IgG responses, most likely derived from the circulation, were detected and correlated with serological IgG levels after mRNA vaccination [18]. However, secretory salivary IgA (SIgA) responses, which were shown to have potent neutralizing activity against SARS-CoV-2, were detected only in a minority of vaccinees and at much lower levels compared with SIgA levels in convalescent saliva [19••]. Indeed, durable vaccine-induced IgA responses have been correlated with reduced incidence of breakthrough infection [19]. Nevertheless, whether the presence of salivary antibody responses would predict induction of anti-SARS-CoV-2 immunity at the portal of infection, the respiratory mucosa, remains unknown. A limited number of studies have assessed the injectable (mRNA) vaccine-induced humoral and cellular responses in the nasal mucosa. However, whether injectable COVID-19 vaccines can induce meaningful neutralizing antibody and tissue-resident T and B cells in the nasal tissue is still debatable because of contradicting observations 20, 21. Inherent factors such as variable baseline nasal immunoglobulins and limited sampling by nasal swabs and nasal washes and sample processing make assessment of immune responses at the nasal mucosa difficult.
Currently, little is known about the magnitude, quality, and kinetics of RM immunity induced by first-generation injectable COVID-19 vaccines in humans. To-date, only one clinical study has examined mRNA COVID-19 vaccine-induced humoral and cellular responses at the respiratory mucosa through analyzing bronchoalveolar lavage fluid [9]. Compared with COVID-19 convalescent individuals, vaccinated individuals had significantly lower levels of neutralizing antibodies in the airways against SARS-CoV-2 VOC, including Omicron (BA1.1). Importantly, in contrast to natural infection, mRNA vaccination failed to induce notable lung-resident spike-specific memory B- and T-cell responses. This may explain why mRNA vaccine-induced immunity is relatively less efficient in protecting against infection compared with prior infection-induced immunity as seen during the SARS-CoV-2 Delta wave [22]. The inability of injectable COVID-19 vaccines in inducing RM immunity resonates with the findings from the clinical trials of novel tuberculosis (TB) vaccines wherein parenteral injection with a viral-vectored vaccine failed to induce airway tissue-resident memory (TRM) T cells despite robust peripheral blood T-cell immunity 23, 24••.
Recruitment of immune cells to the lung is highly regulated to avoid unnecessary inflammation and preserve the vital function of the lung, gas exchange [25]. As such, local respiratory mucosal immune responses are mounted only in the event of local insults. This notion signifies that respiratory mucosal booster vaccination rather than repeated intramuscular injections can induce local mucosal immunity against SARS-CoV-2. In fact, RM booster vaccination with viral-vectored vaccines in mRNA-primed animals showed promising outcomes against SARS-CoV-2 VOC [26••]. Furthermore, real-world evidence that current vaccines are highly effective in preventing severe disease and hospitalization against VOC indicates a critical role for other immune cells, including circulating memory T cells induced by vaccination. Even though VOC can evade mucosal-neutralizing antibodies to cause infection, circulating memory T cells recruited to the lung enable protection by restraining viral replication and spread, resulting in milder disease outcomes [27]. Notably, current first-generation COVID-19 vaccines may also induce changes in circulating monocytes suggestive of ‘trained innate immunity’ (TII) 28, 29••. Recruitment of these cells to the lung may also contribute to antiviral immunity. Thus, evidence so far suggests that positioning anti-SARS-CoV-2 immunity at the respiratory mucosa that is ready-to-go can quickly act upon and prevent establishment of an infection, supporting boosting RM immunity via the respiratory route to be a viable next-generation vaccine strategy (Figure 1).
Respiratory mucosal vaccination and anti-COVID-19 immunity
Conceptually, inducing adaptive immunity at the respiratory mucosa, involving secretory antibody responses and tissue-resident B and T cells (TRM), and TII has the capacity to prevent SARS-CoV-2 from establishing infection (Figure 1) 11, 12, 30. Indeed, following natural infection, SARS-CoV-2-specific memory B- and T-cell responses were found in many tissue sites with lung and lung-associated lymph nodes (LN) being the most prevalent sites [17]. Furthermore, SARS-CoV-2-specific germinal centers and follicular helper T cells were also found in the lung and lung-associated LNs. Such local tissue-associated immune mechanisms against non-Omicron SARS-CoV-2 appeared linked with reduced risk (50% reduction) for Omicron infection [31]. Interestingly, protection reached 94% in the presence of prior infection and vaccination (hybrid immunity). These observations highlight the importance of focusing next-generation vaccines to boost RM tripartite immunity that can protect against future emerging VOC without a need for regularly updating the vaccines. A multitude of factors must be taken into consideration to ensure the safety, amenability, and efficacy of RM-delivered vaccines. These considerations span vector selection, antigenic design, and formulation.
Vector and antigen selection
Viral-vectored vaccines, in particular AdV vaccines, represent the most promising platform for RM delivery owing to their excellent safety profile, amenability, and robust intrinsic immunogenicity [32]. To-date, multiple human and nonhuman primate adenoviruses have been utilized for vaccine development. Numerous preclinical studies support the potential of RM-delivered AdV vaccines to stimulate robust and long-lasting antibody, cell-mediated, and TII immune responses in the lung. Their potential for RM delivery has been recently shown in clinical trials 24••, 33, 34. To circumvent the effect of anti-Ad backbone antibodies, different serotypes or origins of AdV platforms could be used for repeated heterologous booster immunization if needed.
To meet the challenges arising from emerging VOC compounded with the limited durability of first-generation vaccine-induced immunity, there is an intensifying interest to develop multivalent vaccines that express additional SARS-CoV-2 internal antigens that show high levels of sequence conservation among coronaviruses. These conserved antigens include nucleocapsid (N), RNA-dependent RNA polymerase (RdRp/NSP12), and other structural/nonstructural proteins, and are selected to broaden T-cell immunity. Adenoviral vectors are highly plastic, capable of accommodating large transgenes, thereby making them ideal multivalent vaccine platforms [32]. Indeed, robust T-cell responses to N are found in COVID-19, SARS, and uninfected individuals [35]. Furthermore, NSP12-specific T cells can recognize and kill target cells expressing NSP12 and have been associated with abortive seronegative SARS-CoV-2 in a cohort of healthcare workers 36••, 37••. We have recently provided preclinical evidence that a monovalent spike-expressing AdV vaccine failed to protect against B.1.351, whereas a multivalent AdV expressing N and NSP12 together with spike protein provided complete protection [38••].
RM vaccination and humoral immunity
Foundational studies utilizing human adenoviruses expressing spike from ancestral SARS-CoV-2 laid the groundwork in vaccine-induced respiratory mucosal immunity against COVID-19. RM but not intramuscular immunization with AdV COVID-19 vaccines induced robust IgG and IgA responses within the airways, which was associated with enhanced protection and reduced transmission of infection 39, 40, 41. Murine studies utilizing spike-expressing chimpanzee adenoviruses have further expanded the immunological superiority of RM-delivered AdV against SARS-CoV-2 38••, 42. In addition to inducing airway nAb responses against VOC, RM but not intramuscular vaccination induces systemic antibody responses with enhanced Fc effector function against immune-evasive (beta) VOC 42, 43. In line with these observations, induction of lung vaccine-specific memory B cells and bone-marrow long-lived plasma cells has been observed following RM but not intramuscular vaccination (Figure 1) 38••, 42.
The above evidence supports the relevance of inducing humoral anti-spike-neutralizing antibody responses at the respiratory mucosa. However, since continued viral evolution will progressively erode nAb efficacy, additional respiratory mucosal humoral correlates for protection against VOC need to be investigated with next-generation vaccines. Indeed, non-neutralizing antibodies have been clinically correlated with immunity against VOC through Fc-mediated effector functions even considering reducing neutralizing activity [44]. Preclinical studies have recently shown that first-generation mRNA vaccines induce antibodies that protect against Omicron BA.5 through Fc-mediated effector functions, further highlighting the protective capacity of vaccine-induced humoral immunity against antigenically divergent VOC even considering reduced neutralizing potentials [45]. Additionally, recent work has shown that improved outcomes by convalescent serum therapy were associated with non-neutralizing antibodies against the N of SARS-CoV-2 46, 47, which supports the inclusion of additional structural SARS-CoV-2 antigens into next-generation vaccines.
RM vaccination and T-cell immunity
Mounting human clinical data suggest a critical role for T-cell immunity in the observed protection 27, 48, including protection against Alpha-to-Omicron VOC [49]. To this end, differential T-cell responses have been detected between vaccine-naive individuals who have had mild and severe disease outcomes. While robust antibody responses with almost undetectable circulating SARS-CoV-2-specific T cells were found in individuals with prolonged severe COVID-19, rapid expansion of structural and nonstructural SARS-CoV-2 protein-specific T-cell responses was detected in individuals who rapidly controlled SARS-CoV-2 replication 50, 51. Importantly, memory T-cell responses against NSP12 and N have been associated with abortive SARS-CoV-2 infection without seroconversion 37••, 52. Furthermore, depleting CD8 T cells in convalescent or vaccinated macaques resulted in partial abrogation of protective efficacy of natural and vaccine-induced immunity 53, 54.
Despite clinical data supporting the significance of T-cell immunity against SARS-CoV-2, important questions remain regarding the location, longevity, quantity, and quality of T cells needed for optimal protection. This is partly because most of those observations are confined to the circulatory compartment, and as such, clear understanding of T-cell responses in the upper and lower respiratory tract is still lacking following natural infection and vaccination 9••, 55. In fact, in contrast to convalescent individuals, infection-naive individuals who have received injectable first-generation COVID-19 vaccines lacked vaccine-specific lung TRM T-cell immunity [9], which is localized to nonlymphoid tissues and plays critical roles in tissue-localized immunity to infections 11, 12, 13. Preclinical studies with AdV COVID-19 vaccines have shown that RM vaccination induces long-lasting TRM cells within the lung tissue and airways that played a critical role in protection through elimination of virally infected cells 26••, 38••, 42, 43. On the other hand, although parenteral AdV COVID-19 vaccine (ChAdOx1) prevented pneumonia in macaques, it failed to prevent viral replication in the upper respiratory tract (URT) [56]. Indeed, oral or intranasal delivery of an AdV vaccine not only protected hamsters against infection but also prevented transmission to unvaccinated naive hamsters. Recent studies have shown that AdV vaccines against TB and COVID-19 are safe and well-tolerated in humans when delivered via inhaled aerosol 24••, 57, 58. Inhaled AdV COVID-19 booster vaccination in previously inactivated CoronaVac-primed individuals resulted in induction of 18–24-times higher circulating cross-neutralizing antibodies compared with homologous boost with CoronaVac [33]. However, the detail of local respiratory mucosal immune responses following AdV COVID-19 vaccination remains unknown. Alternatively, an aerosol AdV TB vaccine clinical trial found long-lasting TRM cells to be induced in the lung [24]. Such mucosal T-cell immunity is possibly maintained via self-renewal of T cells in an antigen-dependent manner [59]. This trial offers the proof of concept for the capability of RM-delivered AdV COVID-19 vaccine to elicit long-lasting immunity in hosts with complex immunological history and genetic backgrounds.
RM vaccination and trained innate immunity
Beside the mucosal-adaptive immune pathways, the innate immune system, which functions as the first line of defense against incoming pathogens including SARS-CoV-2, plays a critical role in orchestrating mucosal immunity. Importantly, given that ‘TII’, a state of innate hyperresponsiveness induced by prior local immunologic exposure at the respiratory mucosa, has the potential to fight against heterologous infections, next-generation vaccine strategies can exploit this third arm of the immune system 12, 60, 61. A growing body of evidence now indicates that immunization with some vaccines can induce TII in circulating monocytes and alveolar macrophages 29••, 62, 63, 64, 65. For instance, BCG-induced TII has been associated with reduced all-cause mortality in low-birth-weight infants and reduced the incidence of respiratory tract infections in the elderly 66, 67. In fact, a recent study carried out in Greece demonstrated that BCG vaccination reduced the occurrence of new infections and reduced the risk of COVID-19 [68]. However, preclinical studies have been at odds with clinical studies in that BCG vaccination does not provide protection against SARS-CoV-2 [69]. Although recent studies have shed much-needed light on induction of TII following mRNA or ChAdOx1 COVID-19 vaccination, these responses are likely restricted to the systemic compartments and their contribution to respiratory mucosal immunity remains unknown 28, 29••. In contrast, TII induced in alveolar macrophages following RM vaccination with AdV vaccines renders protection against respiratory infections, including SARS-CoV-2 38••, 62, 63. In humans, RM immunization with AdV vaccine induced persisting transcriptional changes in alveolar macrophages [24]. These observations provide the rationale to harness TII in the design of next-generation COVID-19 vaccines.
Clinical landscape of current respiratory mucosal COVID-19 vaccine strategies
The most explored respiratory vaccine strategy against COVID-19 is the intranasal route of delivery. Currently, several COVID-19 mucosal vaccines are under development and 14 are in clinical trials, of which three — in India, Iran, and Russia have received emergency approval and are being administered as a nasal spray ( Table 1). Although efficacy data for many of these trials have not yet been published, the trial results for the intranasally delivered AstraZeneca ChAdOx1-S vaccine have recently been released. Unfortunately, the findings indicated poor immunogenicity in previously mRNA-vaccinated participants [34]. This is in stark contrast to the excellent results obtained for this vaccine in preclinical animal models [70]. Such discrepancy may be owed to anatomical differences in the URT between humans and animals and the delivery methods. For instance, it is known that volumes between 25 and 50 µL of an agent delivered intranasally to mice also have the potential to reach the lower respiratory tract, hence inducing immunity not only in the nasal passage but also deep in the lung [71]. Optimally formulating vaccines for intranasal delivery is challenging and critical to overcoming the natural nasal defense barriers.Table 1 Vaccines in the clinical pipeline evaluated for respiratory mucosal delivery.
Table 1Vaccine vector Vaccine ID Developer Route Delivery device Phase Status Clinical trial ID
Adenovirus BBV154 Bharat Biotech International Limited I.N. Droppers III Active, not recruiting NCT05522335
SC-Ad6-1 Tetherex Pharmaceuticals Corporation I.N. or I.M. N/A I Recruiting NCT04839042
Ad5-triCoV/Mac, ChAd-triCoV/Mac McMaster University Aerosol inhalation Aeroneb Solo nebulizer I Recruiting NCT05094609
AZD1222/ChAdOx1 nCov-19 Imperial College London/University of Oxford/AstraZeneca Aerosol inhalation MAD NasalTM intranasal mucosal atomization device I Recruiting NCT05007275
Ad5-nCoV Jiangsu Province Centers for Disease Control and Prevention Aerosol inhalation Aerogen Ultra device III Active, not recruiting NCT05204589
Ad5-nCoV CanSino Biologics Inc. Aerosol inhalation and/or I.M. Aerogen Solo I/II Active, not recruiting NCT04840992
Parainfluenza virus CVXGA1 CyanVac LLC I.N. Spray devices I Recruiting NCT04954287
Respiratory syncytial virus MV-014-212 Meissa Vaccines, Inc. I.N. Droppers/spray devices I Recruiting NCT04798001
Live-attenuated influenza virus DelNS1-2019-nCoV–RBD–OPT1 The University of Hong Kong I.N. Spray devices II Recruiting NCT05200741
III Active, not recruiting ChiCTR2100051391
Newcastle disease virus NDV–HXP-S Sean Liu, Icahn School of Medicine at Mount Sinai I.N. and/or I.M. N/A I Recruiting NCT05181709
AVX/COVID-12 Laboratorio Avi-Mex, S.A. de C.V. I.N. or I.M. Automatic syringe (prima mist sprayer) II Active, not recruiting NCT05205746
Combined vector vaccine Gam–COVID–Vac Gamaleya Research Institute of Epidemiology and Microbiology, Health Ministry of the Russian Federation I.N. Spray devices I Not yet recruiting NCT05248373
Protein subunit Avacc 10 Intravacc B.V. I.N. N/A I Not yet recruiting NCT05604690
Live-attenuated SARS-CoV-2 COVI–VAC Codagenix, Inc I.N. Droppers I Active, not recruiting NCT05233826
Inhaled aerosol delivery has been recently developed and assessed as an alternative respiratory mucosal vaccination strategy. Aerosol vaccination has been utilized to deliver measles, and numerous AdV vaccines against TB and COVID-19 23, 24••, 58, 72. Importantly, compared with intranasal delivery, the deepened and widened biodistribution in the lung following endotracheal delivery (akin inhalation) is associated with much improved vaccine-mediated immunogenicity and protection against the target pathogen in mice [71]. Inhaled aerosol vaccination of viral-vectored TB vaccines deposited 2–5-µm vaccine droplets in major airways [24], inducing robust and sustainable antigen-specific cellular immunity in the respiratory tract as measured by the responses in bronchoalveolar lavage fluid 23, 24••. Furthermore, persisting alterations in the transcriptional profile of alveolar macrophages poised for defense responses in aerosol-vaccinated participants indicate the potential of aerosol vaccines to induce TII. The availability of well-characterized inhaled aerosol technology (used in TB vaccine trials) and its superiority to induce immunity in the URT offers a foundation for developing inhaled next-generation COVID-19 vaccine strategies. Indeed, Chinese regulators have recently approved the world’s first inhaled aerosol first-generation AdV COVID-19 vaccine (Convidecia Air). In clinical trials, this vaccine was well-tolerated and induced robust neutralizing antibodies and T-cell immunity in the circulation, and importantly, qualitatively enhanced neutralizing antibody responses against the Delta SARS-CoV-2 VOC, compared with homologous intramuscular boost 39, 58. However, to what extent Convidecia Air can induce mucosal immunity remains to be investigated. To this end, a phase-I trial currently undergoing in Canada to evaluate the safety and immunogenicity of aerosolized viral-vectored trivalent vaccines, Ad5-triCoV/Mac and ChAd-triCoV/Mac, will carry out a comprehensive analysis of immune responses within the airways (NCT05094609). Although clinical respiratory mucosal vaccine trials will conduct efficacy studies comparing the vaccinated against placebo, the waning immunity in previously vaccinated and infected individuals will make assessing next-generation vaccines more difficult than investigating first-generation vaccines carried out in an infection-naive population [73]. Nonetheless, these studies may provide the foundation to establish correlates of protection following inhaled immunization. Apart from the benefit of being able to induce mucosal immunity, inhaled immunization holds many other advantages. Since the vaccine is directly delivered to the lung, a much smaller dose than that used for injection can generate an effective immune response [24]. The potential to develop a thermostable spray-dried form of viral-vectored vaccines for inhalation will allow transport and storage less challenging and could be a solution for vaccine inequity [74]. Notably, dry powder vaccine formulation of mRNA encoding spike-loaded exosome when administered via jet nebulization to nonhuman primates elicited stronger IgG and SIgA responses compared with their synthetic counterparts [75].
Outstanding questions and future perspectives
The COVID-19 pandemic continues to expedite our understanding of innovative and novel vaccination strategies. Decades of research have firmly established the intricate link between vaccination route and immune geography, with experimental and clinical evidence showing the unique capacity of RM-delivered AdV vaccines in establishing humoral, cellular, and TII at the respiratory mucosa. Of relevance, preclinical COVID-19 vaccine studies have shown the superiority of RM delivery in inducing local, long-lasting immunological memory capable of providing broad protection against both ancestral and variants of SARS-CoV-2. As the clinical landscape of RM-delivered AdV vaccines continues to expand through and beyond SARS-CoV-2 to other respiratory pathogens, numerous outstanding questions remain to be addressed before their wide application in humans.
Extremely rare incidences of vaccine-induced immune thrombotic thrombocytopenia (VITT) have been characterized following initial intramuscular (IM) vaccination with AdV COVID-19 vaccines (case rate of 3–15 per million vaccinations) 76, 77, 78. Clinically, VITT is associated with multiple diagnostic criteria, including seropositivity for antiplatelet factor 4 (PF4) IgG autoantibodies, elevated D-dimer levels, and disseminated signs of thrombosis, with unique presentation of cerebral venous sinus thrombosis and cerebral venous thrombosis [79]. Although the mechanisms remain to be fully understood, the unintentional administration of AdV vaccines via the intravenous route, or the ensuing microvascular injury and leakage of the adenovirus into the bloodstream due to vaccine constituents, has been empirically demonstrated to potentially contribute to the development of VITT 80, 81, 82. Since anti-PF4 antibody-mediated platelet activation is one of the major mechanisms for VITT, leaked adenoviral particles are believed to directly interact with platelets and/or PF4, which may result in B-cell engagement and subsequent production and maturation anti-PF4 IgG antibodies 82, 83. While the potential for VITT following RM administration of AdV vaccines remains to be investigated, it is our belief that RM delivery of AdV vaccines will not lead to VITT via circumventing viral vector leakage into the circulation.
Moreover, our knowledge of RM vaccines in populations with pulmonary comorbidities remains limited. As research interest continues to grow, clinical trials should be designed to assess the immunogenicity and safety of RM vaccines both in the elderly and in individuals with pulmonary comorbidities. These studies will be crucial in determining the necessary adjustments to vaccine formulations, doses, and delivery methods, as well as in developing a comprehensive framework to ensure that the best strategies are available for providing sufficient coverage in terms of both vaccine immunogenicity and safety following RM delivery in a population-wide setting.
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Z.X. is one of the inventors on a patent application PCT/CA2022/051107, entitled “Novel COVID vaccine and method for delivery”. All other authors declare no competing interests.
Data Availability
No data were used for the research described in the article.
Acknowledgements
The work was supported by the 10.13039/501100000024 Canadian Institutes of Health Research (CIHR) COVID-19 Rapid Research Project, CIHR Foundation Program, and the Innovative Research Program of National Sanitarium Association of Canada. The authors are grateful to support from other members of McMaster COVID-19 vaccine project team.
==== Refs
References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:• of special interest
•• of outstanding interest
1 WHO: Registry Data of COVID-19 Vaccine Candidates; 2022. Available at: 〈https://covid19.trackvaccines.org/vaccines/approved/#vaccine-list〉 (Accessed 21 November 2022).
2 Dagan N. Barda N. Kepten E. Miron O. Perchik S. Katz M.A. Hernán M.A. Lipsitch M. Reis B. Balicer R.D. BNT162b2 mRNA Covid-19 vaccine in a nationwide mass vaccination setting N Engl J Med 384 2021 1412 1423 33626250
3 Baden L.R. El Sahly H.M. Essink B. Kotloff K. Frey S. Novak R. Diemert D. Spector S.A. Rouphael N. Creech C.B. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine N Engl J Med 384 2021 403 416 33378609
4 Buchan S.A. Chung H. Brown K.A. Austin P.C. Fell D.B. Gubbay J.B. Nasreen S. Schwartz K.L. Sundaram M.E. Tadrous M. Estimated effectiveness of COVID-19 vaccines against Omicron or Delta symptomatic infection and severe outcomes JAMA Netw Open 5 2022 e2232760
5•• Liu J. Chandrashekar A. Sellers D. Barrett J. Jacob-Dolan C. Lifton M. McMahan K. Sciacca M. VanWyk H. Wu C. Vaccines elicit highly conserved cellular immunity to SARS-CoV-2 Omicron Nature 603 2022 493 496 35102312
Spike-specific T cell responses induced by first-generation COVID-19 vaccines are conserved in their ability to recognize the omicron spike protein.
6 Jung M.K. Jeong S.D. Noh J.Y. Kim D.U. Jung S. Song J.Y. Jeong H.W. Park S.H. Shin E.C. BNT162b2-induced memory T cells respond to the Omicron variant with preserved polyfunctionality Nat Microbiol 7 2022 909 917 35577972
7 Azzi L. Dalla Gasperina D. Veronesi G. Shallak M. Ietto G. Iovino D. Baj A. Gianfagna F. Maurino V. Focosi D. Mucosal immune response in BNT162b2 COVID-19 vaccine recipients EBioMedicine 75 2022 103788
8 Cohen J.I. Dropulic L. Wang K. Gangler K. Morgan K. Liepshutz K. Krogmann T. Ali M.A. Qin J. Wang J. Comparison of levels of nasal, salivary, and plasma antibody to SARS-CoV-2 during natural infection and after vaccination Clin Infect Dis 76 2022 1391 1399 10.1093/CID/CIAC934
9•• Tang J. Zeng C. Cox T.M. Li C. Son Y.M. Cheon I.S. Wu Y. Behl S. Taylor J.J. Chakaraborty R. Respiratory mucosal immunity against SARS-CoV-2 after mRNA vaccination Sci Immunol 7 2022 eadd4853 35857583
First study that comprehensively compared the respiratory mucosal responses between mRNA vaccinated participants and convalescent people.
10 Alu A. Chen L. Lei H. Wei Y. Tian X. Wei X. Intranasal COVID-19 vaccines: from bench to bed EBioMedicine 76 2022 103841
11 Lavelle E.C. Ward R.W. Mucosal vaccines — fortifying the frontiers Nat Rev Immunol 22 2022 236 250 34312520
12 Jeyanathan M. Afkhami S. Smaill F. Miller M.S. Lichty B.D. Xing Z. Immunological considerations for COVID-19 vaccine strategies Nat Rev Immunol 20 2020 615 632 32887954
13 Carbone F.R. Unique properties of tissue-resident memory T cells in the lungs: implications for COVID-19 and other respiratory diseases Nat Rev Immunol 23 2022 329 335 10.1038/s41577-022-00815-z 36494455
14 Sadarangani M. Marchant A. Kollmann T.R. Immunological mechanisms of vaccine-induced protection against COVID-19 in humans Nat Rev Immunol 21 2021 475 484 34211186
15 Khoury D.S. Cromer D. Reynaldi A. Schlub T.E. Wheatley A.K. Juno J.A. Subbarao K. Kent S.J. Triccas J.A. Davenport M.P. Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection Nat Med 27 2021 1205 1211 10.1038/s41591-021-01377-8 34002089
16 Earle K.A. Ambrosino D.M. Fiore-Gartland A. Goldblatt D. Gilbert P.B. Siber G.R. Dull P. Plotkin S.A. Evidence for antibody as a protective correlate for COVID-19 vaccines Vaccine 39 2021 4423 4428 34210573
17 Poon M.M.L. Rybkina K. Kato Y. Kubota M. Matsumoto R. Bloom N.I. Zhang Z. Hastie K.M. Grifoni A. Weiskopf D. SARS-CoV-2 infection generates tissue-localized immunological memory in humans Sci Immunol 6 2021 eabl9105 34618554
18 Isho B. Abe K.T. Zuo M. Jamal A.J. Rathod B. Wang J.H. Li Z. Chao G. Rojas O.L. Bang Y.M. Persistence of serum and saliva antibody responses to SARS-CoV-2 spike antigens in COVID-19 patients Sci Immunol 5 2020
19•• Sheikh-Mohamed S. Isho B. Chao G.Y.C. Zuo M. Cohen C. Lustig Y. Nahass G.R. Salomon-Shulman R.E. Blacker G. Fazel-Zarandi M. Systemic and mucosal IgA responses are variably induced in response to SARS-CoV-2 mRNA vaccination and are associated with protection against subsequent infection Mucosal Immunol 15 2022 799 808 35468942
Spike-specific salivary antibodies are poorly induced following intramuscular vaccination but are readily induced following breakthrough infection.
20 Ssemaganda A. Nguyen H.M. Nuhu F. Jahan N. Card C.M. Kiazyk S. Severini G. Keynan Y. Su R.C. Ji H. Expansion of cytotoxic tissue-resident CD8+ T cells and CCR6+CD161+ CD4+ T cells in the nasal mucosa following mRNA COVID-19 vaccination Nat Commun 13 2022 3357 35688805
21 Lim J.M.E. Tan A.T. Bert N. le, Hang S.K. Low J.G.H. Bertoletti A. SARS-CoV-2 breakthrough infection in vaccinees induces virus-specific nasal-resident CD8+ and CD4+ T cells of broad specificity J Exp Med 219 2022 e20220780
22 Goldberg Y. Mandel M. Bar-On Y.M. Bodenheimer O. Freedman L.S. Ash N. Alroy-Preis S. Huppert A. Milo R. Protection and waning of natural and hybrid immunity to SARS-CoV-2 N Engl J Med 386 2022 2201 2212 35613036
23 Satti I. Meyer J. Harris S.A. Thomas Z.-R.M. Griffiths K. Antrobus R.D. Rowland R. Ramon R.L. Smith M. Sheehan S. Safety and immunogenicity of a candidate tuberculosis vaccine MVA85A delivered by aerosol in BCG-vaccinated healthy adults: a phase 1, double-blind, randomised controlled trial Lancet Infect Dis 14 2014 939 946 25151225
24•• Jeyanathan M. Fritz D.K. Afkhami S. Aguirre E. Howie K.J. Zganiacz A. Dvorkin-Gheva A. Thompson M.R. Silver R.F. Cusack R.P. Aerosol delivery, but not intramuscular injection, of adenovirus-vectored tuberculosis vaccine induces respiratory-mucosal immunity in humans JCI Insight 7 2022 e155655
First in-human clinical trial that evaluates the immunogenicity following inhaled aerosol vaccination with an AdV
TB vaccine. This study has demonstrated that the aerosol, but not intramuscular vaccination induces immunity at the respiratory mucosa as measured in the bronchoalveolar fluids.
25 Holt P.G. Strickland D.H. Wikström M.E. Jahnsen F.L. Regulation of immunological homeostasis in the respiratory tract Nat Rev Immunol 8 2008 142 152 18204469
26•• Lapuente D. Fuchs J. Willar J. Vieira Antão A. Eberlein V. Uhlig N. Issmail L. Schmidt A. Oltmanns F. Peter A.S. Protective mucosal immunity against SARS-CoV-2 after heterologous systemic prime-mucosal boost immunization Nat Commun 12 2021 1 14 33397941
Vaccine-specific mucosal immune responses are strongly induced following respiratory mucosal boosting of intramuscular-primed animals.
27 Wherry E.J. Barouch D.H. T cell immunity to COVID-19 vaccines Science 377 2022 821 822 35981045
28 Saresella M. Piancone F. Marventano I. Hernis A. Trabattoni D. Invernizzi M. La Rosa F. Clerici M. Innate immune responses to three doses of the BNT162b2 mRNA SARS-CoV-2 vaccine Front Immunol 13 2022
29•• Murphy D.M. Cox D.J. Connolly S.A. Breen E.P. Brugman A.A.I. Phelan J.J. Keane J. Basdeo S.A. Trained immunity is induced in humans after immunization with an adenoviral vector COVID-19 vaccine J Clin Investig 2022 10.1172/JCI162581
First human study to characterize induction of TII in circulating monocytes following intramuscular immunization with a first-generation AdV COVID-19 vaccine.
30 Singh R. Kang A. Luo X. Jeyanathan M. Gillgrass A. Afkhami S. Xing Z. COVID-19: current knowledge in clinical features, immunological responses, and vaccine development FASEB J 35 2021 e21409
31 Carazo S. Skowronski D.M. Brisson M. Sauvageau C. Brousseau N. Gilca R. Ouakki M. Barkati S. Fafard J. Talbot D. Estimated protection of prior SARS-CoV-2 infection against reinfection with the Omicron variant among messenger RNA-vaccinated and nonvaccinated individuals in Quebec, Canada JAMA Netw Open 5 2022 e2236670
32 Afkhami S. Yao Y. Xing Z. Methods and clinical development of adenovirus-vectored vaccines against mucosal pathogens Mol Ther Methods Clin Dev 3 2016 16030
33 Wu S. Huang J. Zhang Z. Wu J. Zhang J. Hu H. Zhu T. Zhang J. Luo L. Fan P. Safety, tolerability, and immunogenicity of an aerosolised adenovirus type-5 vector-based COVID-19 vaccine (Ad5-nCoV) in adults: preliminary report of an open-label and randomised phase 1 clinical trial Lancet Infect Dis 21 2021 1654 1664 34324836
34 Madhavan M. Ritchie A.J. Aboagye J. Jenkin D. Provstgaad-Morys S. Tarbet I. Woods D. Davies S. Baker M. Platt A. Tolerability and immunogenicity of an intranasally-administered adenovirus-vectored COVID-19 vaccine: an open-label partially-randomised ascending dose phase I trial EBioMedicine 85 2022 104298
35 le Bert N. Tan A.T. Kunasegaran K. Tham C.Y.L. Hafezi M. Chia A. Chng M.H.Y. Lin M. Tan N. Linster M. SARS-CoV-2-specific T cell immunity in cases of COVID-19 and SARS, and uninfected controls Nature 584 2020 457 462 32668444
36•• Nesterenko P.A. McLaughlin J. Tsai B.L. Burton Sojo G. Cheng D. Zhao D. Mao Z. Bangayan N.J. Obusan M.B. Su Y. HLA-A∗02:01 restricted T cell receptors against the highly conserved SARS-CoV-2 polymerase cross-react with human coronaviruses Cell Rep 37 2021 110167 34919800
RdRp-specific T cells are highly conserved among human coronaviruses and are cross-reactive against SARS-CoV-2 RdRp.
37•• Swadling L. Diniz M.O. Schmidt N.M. Amin O.E. Chandran A. Shaw E. Pade C. Gibbons J.M. Le Bert N. Tan A.T. Pre-existing polymerase-specific T cells expand in abortive seronegative SARS-CoV-2 Nature 601 2022 110 117 34758478
Highly exposed but seronegative healthcare workers possess reactive RdRp-specific T cells, which may contribute to rapid clearance/abortive infection with SARS-CoV-2.
38•• Afkhami S. D’Agostino M.R. Zhang A. Stacey H.D. Marzok A. Kang A. Singh R. Bavananthasivam J. Ye G. Luo X. Respiratory mucosal delivery of next-generation COVID-19 vaccine provides robust protection against both ancestral and variant strains of SARS-CoV-2 Cell 185 2022 896 915 e19 35180381
First pre-clinical study assessing a respiratory-mucosal delivered multivalent vaccine expressing spike, nucleocapsid, and polymerase (NSP12) COVID-19 vaccine capable of generating tripartite immunity against SARS-CoV-2 VOC.
39 Wu S. Zhong G. Zhang J. Shuai L. Zhang Z. Wen Z. Wang B. Zhao Z. Song X. Chen Y. A single dose of an adenovirus-vectored vaccine provides protection against SARS-CoV-2 challenge Nat Commun 11 2020 1 7 31911652
40 Langel S.N. Johnson S. Martinez C.I. Tedjakusuma S.N. Peinovich N. Dora E.G. Kuehl P.J. Irshad H. Barrett E.G. Werts A.D. Adenovirus type 5 SARS-CoV-2 vaccines delivered orally or intranasally reduced disease severity and transmission in a hamster model Sci Transl Med 14 2022 6868
41 Xu F. Wu S. Yi L. Peng S. Wang F. Si W. Hou L. Zhu T. Safety, mucosal and systemic immunopotency of an aerosolized adenovirus-vectored vaccine against SARS-CoV-2 in rhesus macaques Emerg Microbes Infect 11 2022 438 441 35094672
42 Hassan A.O. Kafai N.M. Dmitriev I.P. Fox J.M. Smith B.K. Harvey I.B. Chen R.E. Winkler E.S. Wessel A.W. Case J.B. A single-dose intranasal ChAd vaccine protects upper and lower respiratory tracts against SARS-CoV-2 Cell 183 2020 169 184 e13. 32931734
43 Hassan A.O. Shrihari S. Gorman M.J. Curiel D.T. An intranasal vaccine durably protects against SARS-CoV-2 variants in mice CellReports 36 2021 109452
44 Zohar T. Loos C. Fischinger S. Atyeo C. Wang C. Slein M.D. Burke J. Yu J. Feldman J. Hauser B.M. Compromised humoral functional evolution tracks with SARS-CoV-2 mortality Cell 183 2020 1508 1519 e12. 33207184
45 Mackin S.R. Desai P. Whitener B.M. Karl C.E. Liu M. Baric R.S. Edwards D.K. Chicz T.M. McNamara R.P. Alter G. Fcγ receptor-dependent antibody effector functions are required for vaccine protection against infection by antigenic variants of SARS-CoV-2 bioRxiv 2022 10.1101/2022.11.27.518117
46 Herman J.D. Wang C. Burke J.S. Zur Y. Compere H. Kang J. Macvicar R. Taylor S. Shin S. Frank I. Nucleocapsid-specific antibody function is associated with therapeutic benefits from COVID-19 convalescent plasma therapy Cell Rep Med 3 2022 100811
47 Zhang A. Stacey H.D. D’Agostino M.R. Tugg Y. Marzok A. Miller M.S. Beyond neutralization: Fc-dependent antibody effector functions in SARS-CoV-2 infection Nat Rev Immunol 23 2022 381 396 10.1038/s41577-022-00813-1 36536068
48 Muik A. Lui B.G. Diao H. Fu Y. Bacher M. Toker A. Grosser J. Ozhelvaci O. Grikscheit K. Hoehl S. Progressive loss of conserved spike protein neutralizing antibody sites in Omicron sublineages is balanced by preserved T-cell recognition epitopes bioRxiv 2022 10.1101/2022.12.15.520569
49 Tarke A. Coelho C.H. Zhang Z. Dan J.M. Yu E.D. Methot N. Bloom N.I. Goodwin B. Phillips E. Mallal S. SARS-CoV-2 vaccination induces immunological T cell memory able to cross-recognize variants from Alpha to Omicron Cell 185 2022 847 859 e11. 35139340
50 Chandran A. Rosenheim J. Nageswaran G. Swadling L. Pollara G. Gupta R.K. Burton A.R. Guerra-Assunção J.A. Woolston A. Ronel T. Rapid synchronous type 1 IFN and virus-specific T cell responses characterize first wave non-severe SARS-CoV-2 infections Cell Rep Med 3 2022 100557 35474751
51 Tan A.T. Linster M. Tan C.W. Le Bert N. Chia W.N. Kunasegaran K. Zhuang Y. Tham C.Y.L. Chia A. Smith G.J.D. Early induction of functional SARS-CoV-2-specific T cells associates with rapid viral clearance and mild disease in COVID-19 patients Cell Rep 34 2021 108728
52 Kundu R. Narean J.S. Wang L. Fenn J. Pillay T. Fernandez N.D. Conibear E. Koycheva A. Davies M. Tolosa-Wright M. Cross-reactive memory T cells associate with protection against SARS-CoV-2 infection in COVID-19 contacts Nat Commun 13 2022 80 35013199
53 Liu J. Yu J. McMahan K. Jacob-Dolan C. He X. Giffin V. Wu C. Sciacca M. Powers O. Nampanya F. CD8 T cells contribute to vaccine protection against SARS-CoV-2 in macaques Sci Immunol 7 2022
54 McMahan K. Yu J. Mercado N.B. Loos C. Tostanoski L.H. Chandrashekar A. Liu J. Peter L. Atyeo C. Zhu A. Correlates of protection against SARS-CoV-2 in rhesus macaques Nature 590 2021 630 634 33276369
55 Grau-Expósito J. Sánchez-Gaona N. Massana N. Suppi M. Astorga-Gamaza A. Perea D. Rosado J. Falcó A. Kirkegaard C. Torrella A. Peripheral and lung resident memory T cell responses against SARS-CoV-2 Nat Commun 12 2021 3010 34021148
56 van Doremalen N. Purushotham J.N. Schulz J.E. Holbrook M.G. Bushmaker T. Carmody A. Port J.R. Yinda C.K. Okumura A. Saturday G. Intranasal ChAdOx1 nCoV-19/AZD1222 vaccination reduces viral shedding after SARS-CoV-2 D614G challenge in preclinical models Sci Transl Med 13 2021
57 Wu S. Huang J. Zhang Z. Wu J. Zhang J. Hu H. Zhu T. Zhang J. Luo L. Fan P. Safety, tolerability, and immunogenicity of an aerosolised adenovirus type-5 vector-based COVID-19 vaccine (Ad5-nCoV) in adults: preliminary report of an open-label and randomised phase 1 clinical trial Lancet Infect Dis 21 2021 1654 1664 34324836
58 Li J.-X. Wu S.-P. Guo X.-L. Tang R. Huang B.-Y. Chen X.-Q. Chen Y. Hou L.-H. Liu J.-X. Zhong J. Safety and immunogenicity of heterologous boost immunisation with an orally administered aerosolised Ad5-nCoV after two-dose priming with an inactivated SARS-CoV-2 vaccine in Chinese adults: a randomised, open-label, single-centre trial Lancet Respir Med 10 2022 739 748 35605625
59 Jeyanathan M. Mu J. McCormick S. Damjanovic D. Small C.-L. Shaler C.R. Kugathasan K. Xing Z. Murine airway luminal antituberculosis memory CD8 T cells by mucosal immunization are maintained via antigen-driven in situ proliferation, independent of peripheral T cell recruitment Am J Respir Crit Care Med 181 2010
60 O’Neill L.A.J. Netea M.G. BCG-induced trained immunity: can it offer protection against COVID-19? Nat Rev Immunol 20 2020 335 337 32393823
61 Xing Z. Afkhami S. Bavananthasivam J. Fritz D.K. D’Agostino M.R. Vaseghi-Shanjani M. Yao Y. Jeyanathan M. Innate immune memory of tissue-resident macrophages and trained innate immunity: re-vamping vaccine concept and strategies J Leukoc Biol 108 2020 825 834 32125045
62 D’Agostino M.R. Lai R. Afkhami S. Khera A. Yao Y. Vaseghi-Shanjani M. Zganiacz A. Jeyanathan M. Xing Z. Airway macrophages mediate mucosal vaccine-induced trained innate immunity against Mycobacterium tuberculosis in early stages of infection J Immunol 205 2020 2750 2762 32998983
63 Yao Y. Jeyanathan M. Haddadi S. Barra N.G. Vaseghi-Shanjani M. Damjanovic D. Lai R. Afkhami S. Chen Y. Dvorkin-Gheva A. Induction of autonomous memory alveolar macrophages requires T cell help and is critical to trained immunity Cell 175 2018 1634 1650 e17. 30433869
64 Bannister S. Kim B. Domínguez-Andrés J. Kilic G. Ansell B.R.E. Neeland M.R. Moorlag S.J.C.F.M. Matzaraki V. Vlahos A. Shepherd R. Neonatal BCG vaccination is associated with a long-term DNA methylation signature in circulating monocytes Sci Adv 8 2022 eabn4002
65 Jeyanathan M. Vaseghi-Shanjani M. Afkhami S. Grondin J.A. Kang A. D’Agostino M.R. Yao Y. Jain S. Zganiacz A. Kroezen Z. Parenteral BCG vaccine induces lung-resident memory macrophages and trained immunity via the gut-lung axis Nat Immunol 23 2022 1687 1702 36456739
66 Aaby P. Roth A. Ravn H. Napirna B.M. Rodrigues A. Lisse I.M. Stensballe L. Diness B.R. Lausch K.R. Lund N. Randomized trial of BCG vaccination at birth to low-birth-weight children: beneficial nonspecific effects in the neonatal period J Infect Dis 2011 245 252 10.1093/infdis/jir240 21673035
67 Giamarellos-Bourboulis E.J. Tsilika M. Moorlag S. Antonakos N. Kotsaki A. Domínguez-Andrés J. Kyriazopoulou E. Gkavogianni T. Adami M.E. Damoraki G. Activate: randomized clinical trial of BCG vaccination against infection in the elderly Cell 183 2020 315 323 e9. 32941801
68 Tsilika M. Taks E. Dolianitis K. Kotsaki A. Leventogiannis K. Damoulari C. Kostoula M. Paneta M. Adamis G. Papanikolaou I. ACTIVATE-2: A Double-blind Randomized Trial of BCG vaccination against COVID-19 in individuals at risk Front Immunol 13 2022
69 Kaufmann E. Khan N. Tran K.A. Ulndreaj A. Pernet E. Fontes G. Lupien A. Desmeules P. McIntosh F. Abow A. BCG vaccination provides protection against IAV but not SARS-CoV-2 Cell Rep 38 2022 110502
70 van Doremalen N. Purushotham J.N. Schulz J.E. Holbrook M.G. Bushmaker T. Carmody A. Port J.R. Yinda C.K. Okumura A. Saturday G. Intranasal ChAdOx1 nCoV-19/AZD1222 vaccination reduces viral shedding after SARS-CoV-2 D614G challenge in preclinical models Sci Transl Med 13 2021 eabh0755 34315826
71 Jeyananthan V. Afkhami S. D’Agostino M.R. Zganiacz A. Feng X. Miller M.S. Jeyanathan M. Thompson M.R. Xing Z. Differential biodistribution of adenoviral-vectored vaccine following intranasal and endotracheal deliveries leads to different immune outcomes Front Immunol 13 2022
72 Low N. Bavdekar A. Jeyaseelan L. Hirve S. Ramanathan K. Andrews N.J. Shaikh N. Jadi R.S. Rajagopal A. Brown K.E. A randomized, controlled trial of an aerosolized vaccine against measles N Engl J Med 372 2015 1519 1529 25875257
73 Waltz E. How nasal-spray vaccines could change the pandemic Nature 609 2022 240 242 36068305
74 Toniolo S.P. Afkhami S. D’Agostino M.R. Lichty B.D. Cranston E.D. Xing Z. Thompson M.R. Spray dried VSV-vectored vaccine is thermally stable and immunologically active in vivo Sci Rep 10 2020 13349
75 Popowski K.D. Moatti A. Scull G. Silkstone D. Lutz H. López de Juan Abad B. George A. Belcher E. Zhu D. Mei X. Inhalable dry powder mRNA vaccines based on extracellular vesicles Matter 5 2022 2960 2974 35847197
76 Muir K.-L. Kallam A. Koepsell S.A. Gundabolu K. Thrombotic thrombocytopenia after Ad26.COV2.S vaccination N Engl J Med 384 2021 1964 1965 33852795
77 Herrera-Comoglio R. Lane S. Vaccine-induced immune thrombocytopenia and thrombosis after the Sputnik V vaccine N Engl J Med 387 2022 1431 1432 36103622
78 Pavord S. Scully M. Hunt B.J. Lester W. Bagot C. Craven B. Rampotas A. Ambler G. Makris M. Clinical features of vaccine-induced immune thrombocytopenia and thrombosis N Engl J Med 385 2021 1680 1689 34379914
79 Cines D.B. Greinacher A. Vaccine-induced immune thrombotic thrombocytopenia Blood 141 2023 1659 1665 36669155
80 Nicolai L. Leunig A. Pekayvaz K. Esefeld M. Anjum A. Rath J. Riedlinger E. Ehreiser V. Mader M. Eivers L. Thrombocytopenia and splenic platelet-directed immune responses after IV ChAdOx1 nCov-19 administration Blood 140 2022 478 490 35486845
81 Baker A.T. Boyd R.J. Sarkar D. Teijeira-Crespo A. Chan C.K. Bates E. Waraich K. Vant J. Wilson E. Truong C.D. ChAdOx1 interacts with CAR and PF4 with implications for thrombosis with thrombocytopenia syndrome Sci Adv 7 2021 8213
82 Greinacher A. Schönborn L. Siegerist F. Steil L. Palankar R. Handtke S. Reder A. Thiele T. Aurich K. Methling K. Pathogenesis of vaccine-induced immune thrombotic thrombocytopenia (VITT) Semin Hematol 59 2022 97 107 35512907
83 Wang J.J. Armour B. Chataway T. Troelnikov A. Colella A. Yacoub O. Hockley S. Tan C.W. Gordon T.P. Vaccine-induced immune thrombotic thrombocytopenia is mediated by a stereotyped clonotypic antibody Blood 140 2022 1738 1742 35661872
|
PMC010xxxxxx/PMC10173205.txt |
==== Front
Pastoral Psychol
Pastoral Psychol
Pastoral Psychology
0031-2789
1573-6679
Springer US New York
1073
10.1007/s11089-023-01073-z
Article
LGBTQ+ Stress, Trauma, Time, and Care
http://orcid.org/0000-0003-0198-7399
Menhinick Keith A. kmenhin@emory.edu
1
Sanders Cody J. Cody.j.sanders@gmail.com
2
1 grid.189967.8 0000 0001 0941 6502 Candler School of Theology, Emory University, 1531 Dickey Dr, Atlanta, GA 30322 USA
2 Old Cambridge Baptist Church, 1151 Massachusetts Ave, Cambridge, MA 02138 USA
11 5 2023
2023
72 3 367384
22 3 2023
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
This article examines how family rejection, religious/spiritual violence, homelessness, adverse school experience, interpersonal violence, and other experiences common among LGBTQ+ people and communities can be reframed as part of a stress-trauma continuum. The pressures and compulsions of white heteropatriarchal society (e.g., of identification, heterosexuality, monogamy, gender expression, etc.) harm us all, yet uniquely expose LGBTQ+ folks to a life of surveillance, stigma, prejudice, erasure, regulation, discipline, and violence. Multiple social psychologists have elucidated how the social conditions of white cis-heteropatriarchy thus engender a kind of chronic stress unique to LGBTQ+ populations (c.f., Meyer, 2013), a stress which accumulates. That accumulation can be understood as queer allostatic load, which falls on a continuum of the stressful to the traumatic, depending on the availability of social supports, access to resources, and coping mechanisms. This article follows historical efforts in the LGBTQ+ community to depathologize trauma by contextualizing the LGBTQ+ lived experience in terms of a stress-trauma continuum. This shift nuances trauma as not only an individual experience but perhaps more importantly as a simultaneously neurobiological and sociocultural experience. Therefore, such a framework helps us examine not only the violence of current social conditions, but also the experiences of chrono-stress and traumatic temporality related to the threat against queer futures and the absenting of queer pasts. This article concludes with several proposals for the spiritual care of queer and trans lives whose experiences fall along this stress-trauma continuum.
Keywords
LGBTQ
Queer
Trans
Stress
Trauma
Psychospiritual
issue-copyright-statement© Springer Science+Business Media, LLC, part of Springer Nature 2023
==== Body
pmcFrom the church, the school, the courthouse, to the family—the explicit message to an overwhelming number of queer and trans people1 is clear: you are not welcome here, not wanted, not cherished, not loved. This message is repeated not only in myriad acts of abuse, rejection, and violence against the mind-body-spirits of LGBTQ+ people and communities; it is also encoded into the political discourses and social structures that organize our life together, including the “family” and even the “future.”
The article begins by presenting data about potentially traumatic and adverse experiences common to LGBTQ+ life in the U.S. context. This data unveils the deception in society’s myth of inevitable progress, which ironically employs LGBTQ+ rights as evidence of progress while simultaneously working to undermine those rights and elide LGBTQ+ history. For example, research into the lived experience of rejection due to being LGBTQ+2 correlates with societal attempts to erase queerness and transness from the school’s curriculum, the government’s history, the church’s community, and even the family’s name and lineage. Some will counter, of course, with citations of the growing number of affirming LGBTQ+ representations in media, or perhaps with the growing number of affirming faith communities. Such invocations of progress, however, function to obscure the continued attacks on LGBTQ+ life. For example, over 340 anti-LGBTQ+ bills have been introduced to state legislatures in the first two months of 2023 alone (Human Rights Campaign Staff, 2023).
In the next section, we trace statistics like these, which are meant to serve not as self-evident facts but as glimpses into an unfolding picture of the patterns of violence against LGBTQ+ personhood and community. We argue that the range of responses to social pressures and violences falls along a continuum of stress and trauma (Dulmus & Hilarski, 2003) and is significantly related to one’s context, culture, and support system.
However, unlike other cultural violence, the attacks on LGBTQ+ life and community frequently emerge from “inside the house”—i.e., from the site of the family (including the family of faith). Attending to the Black gay and lesbian experience, Horace L. Griffin (2006) writes,Lesbians and gays generally emerge from heterosexual parents and families strongly opposed to them. Lesbians and gays enter the world of invisibility, not knowing other lesbian and gay people, and most times find themselves battling not only society, but the very communities that most oppressed groups have counted on to help them confront injustice: their own families, churches, and communities. (p. 156)
For pastoral theologians and spiritual care providers working with queer and trans folks, the family cannot be unproblematically cited or celebrated as a primary mediator of other sociopolitical oppressions but rather becomes revealed in many cases as a traumatic extension and expression of these oppressions.
Importantly, what is at stake for queer and trans people is not only the care and belonging of family but also the assets, resources, and supports tied to lineage, social network, and inheritance. These inner and outer resources are integral to surviving and coping with the anti-queer stressors and violences embedded in a White cis-heterosexist capitalist world order. They are also vital for imagining a family and a future with queer people in it.
Spiritual care with LGBTQ+ people requires responding to the range of disruptions and dysregulations experienced by the individual and the community (which fall on a continuum of stress and trauma), as well as intervening in the conditions of society and family that make being queer so precarious and potentially traumatic in the first place. Importantly, this entails building up the relational, material, and even temporal resources for LGBTQ+ persons to survive, cope, and connect here and now as well as there and then. Holistic spiritual care is thus a search for the resources, families, communities, histories, and futurities necessary for the full flourishing of all life.
Surveying the lived conditions of LGBTQ+ life
The National Child Traumatic Stress Network (Barba et al., 2021) published a two-part resource calling for all care providers and organizations to use a screener to assess for trauma exposure and post-traumatic stress symptoms when working with LGBTQ+ youth. The rationale is that LGBTQ+ youth are disproportionately exposed to a range of adverse childhood experiences (ACEs) and potentially traumatic events (PTEs) compared with their cisgender, heterosexual peers. Consequently, these adverse and potentially traumatic experiences correlate with an increased risk for a range of mental and physical health challenges (e.g., depression, addiction, homelessness, suicidality) and an increased risk for relying on desperate coping methods (e.g., substance abuse, risky sexual behavior).
In part 1 of Identifying the Intersection of Trauma and Sexual Orientation and Gender Identity (Barba et al., 2021), the National Child Traumatic Stress Network surveyed data from a range of studies to posit four primary areas of research into the connection between LGBTQ+ experience and trauma: a sense of safety (in school), physical and sexual harassment and abuse, family rejection, and mental health (p. 2). We can map these four categories as the primary arenas of violence against the personhood and community of LGBTQ+ children and youth. In this study, we also begin to see a picture of how queer, trans, and ally communities are making sense of these violences through the category of traumatization.
A sense of safety is fundamental for optimal development of the child’s nervous system, brain, personality, and community. The GLSEN (pronounced “glisten”; formerly the Gay, Lesbian, and Straight Education Network) National School Climate Survey (Kosciw et al., 2018) reported that 59.5% of LGBTQ+ students felt unsafe at school due to their sexual orientation and 44.6% due to their gender expression (p. xviii). Additionally, 62.2% of LGBTQ+ students experienced discriminatory policies at school, and an overwhelming percentage of LGBTQ+ students reported hearing homophobic and transphobic remarks from peers and teachers (98.5%), with 91.8% feeling distressed afterwards (p. xviii).
What does it do to the mind-body-spirit of an LGBTQ+ child to hear violent language about one’s own emerging sense of self in the very place where one is expected to gain an education and learn how to function in society? We can only postulate about the multiple educational disparities for hypervigilant and stress-aroused LGBTQ+ students, whose sense of threat and ensuing stress can impair their cognitive abilities to concentrate, participate, learn, and retain.
The violence against LGBTQ+ children and youth in school also appears in the omission of LGBTQ+ history and comprehensive sex education from the school curriculum, which is also a refusal to teach LGBTQ+ students that they have a community. According to research published by the Columbia Law Review, “[A] comprehensive survey shows that anti-gay curriculum laws actually exist in twenty states” (Rosky, 2017). The Human Rights Campaign (2023) declares that 2023 already marks the year of the highest amount of anti-LGBTQ+ bills introduced at the state level. This absenting of community and history is ultimately the denial of belonging—to a history; to an ancestry; to a present, alive, and vibrant community of diversity. Give the numerous overt attacks and censorships, the evidence is clear: “Schools nationwide are hostile environments for a distressing number of LGBTQ students” (Kosciw et al., 2018, p. xviii).
Thus, it is no wonder that, as a response to feeling unsafe amidst the increased risk of threat and harm, a majority of LGBTQ+ youth avoided extracurricular activities (70.5%), one-third (34.9%) skipped at least one day of school in the past month, and 10.5% skipped four or more days (Kosciw et al., 2018, p. xviii). Turning from the school to the community, the Trevor Project’s 2022 National Survey on LGBTQ Youth Mental Health reported that 73% of LGBTQ+ youth experienced discrimination due to their gender expression or sexual orientation, with 36% of LGB youth and 37% of transgender and nonbinary youth reporting an experience of being physically threatened or harmed (p. 15).
Skipping school and extracurricular activities does not necessarily mean staying home for many LGBTQ+ children and youth. The Trevor Project (2022) reported fewer than 1 in 3 transgender and nonbinary youth feel affirmed in their home (p. 4). According to a national study by the Human Rights Campaign, 78% of LGBTQ+ youth are not out to their parents due to fears of repercussions, most specifically of abuse or rejection (Human Rights Campaign Staff, 2018, p. 4). Familial rejection is no mere threat as 29% of LGBTQ+ youth have experienced being unhoused (Trevor Project, 2020, p. 4), and, among that population, 75% of LGB youth and 90% of transgender and nonbinary youth link their homelessness directly to the experience of family abuse and/or rejection (Choi et al., 2015, p. 5). Experiences of abuse and rejection haunt many queer and trans people well past childhood and adolescence. The Pew Research Center (2013) found that about 39% (4 out of 10) of LGBTQ+ adults stated that they had been rejected by a close family member or close friend (p. 1).
The relational cutting off by family and close friends is simultaneously spiritual and material, entailing the removal of care and belonging as well as the withdrawal of social networks, political protections, cultural assets, and familial resources (e.g., land, housing, wealth, and inheritance). As a result, LGBTQ+ folks who experience family rejection are even more at risk for a variety of other challenges—food insecurity, homelessness, victimization, self-harm, suicidal ideation and attempt, and a range of mental health disorders. The removal of support networks and material resources also makes LGBTQ+ folks more vulnerable to a variety of communal and environmental stressors: the Covid-19 pandemic, natural disasters due to climate change, systemic racism and oppression, and a number of other crises. As Servigne et al. (2021) make clear, “It is well established that the most important factor for resilience (from the first minutes after the tragedy) is the closeness and helpfulness of family and neighbors (or even strangers) who can aid in overcoming fear, give care and bring touches of joy and optimism” (p. 42). The quality and efficacy of such support is dependent upon the existence of a social network before the communal or environmental disaster occurs. For many young queer and trans people, these networks are precarious.
Data also shows that when LGBTQ+ youth reach out for help and services (e.g., government programs, housing shelters, lending services, healthcare), they are frequently met with higher rates of stigma and discrimination than the general population. Service discrimination becomes not only another source of violation and stress but also another cause of the economic, employment, health, and housing disparities for LGBTQ+ youth, especially those who are Black or Indigenous (Rooney & Durso, 2017). The shortage of culturally specific understandings and services then compounds the alienation for those brave enough to reach out for help, setting up LGBTQ+ folks to experience cycles of abuse, neglect, retraumatization, and isolation.
Of course, these accounts are largely silent about harm done to LGBTQ+ people in the church and other religious communities, which frequently sacralize the exclusions and abuses of the family and state. More specifically, religiosity is frequently the common denominator of households that exile their queer children (Janssen & Scheepers, 2019). Harmful religious narratives and theologies thus are a factor in interpersonal and intrapsychic violence, for these narratives “set life on edge, make life seem unlivable, and often lead to suicide” (Sanders, 2020, p. 1). One study concluded that 9.3% of LGBT youth met DSM-5 criteria for PTSD in the previous 12 months, and the rates exponentially increased for youth forced into conversion therapy (Mustanski et al., 2010). Within-group differences in these studies are rarely accounted for, meaning these numbers may be higher, the risk much more severe, for LGBTQ+ folks of color and of varying physical abilities and immigration status.
There are few spaces of respite for a large percentage of LGBTQ+ people and even fewer for children and youth. Whether in the home, school, community, or church, many LGBTQ+ folks find themselves unable to escape hostile social environments. This inescapability reveals the very structures and institutions of life to be dehumanizing, deforming, and often destructive for those whose lives, loves, and bodies deviate from the social norms and expectations of the dominant society and its mirror—the nuclear family.
In the 2022 National Survey on LGBTQ Youth Mental Health, 45% of LGBTQ+ youth reported they “seriously considered attempting suicide in the past year,” and the rates were much higher for LGBTQ+ youth of color (Trevor Project, 2022). When LGBTQ+ children and youth are abused by and/or cut off from primary relationships and meaningful community, including the material resources tied to those connections, they are thereby cut off from a vision of the future with them in it. This is not an incidental violence but an intentional one. It is the systemic and structural intent of a cis-heteropatriarchal society to either assimilate or eradicate queerness from all facets of our shared life together: the family, school, community, and church, as well as the mind-body-spirit of the person.
The crises of trauma in/as queerness
Obviously not every LGBTQ+ person experiences the range of stressors and “potentially traumatic events” surveyed above. A variety of factors influence not only risk exposure but also the resources for coping and managing stress and trauma—the primary resource being a robust and holistic social support system. But again, for queer and trans folks, rejection from social networks and community is a frequently shared and common experience, even if not a universal one. Several important patterns can be distilled from the above reportings: first, LGBTQ+ people are extremely vulnerable to a range of violent experiences throughout life; second, the withdrawal of social support marks an ongoing fear and threat for LGBTQ+ folks; and third, organizations like the National Child Traumatic Stress Network, the Trevor Project, and the Human Rights Campaign, like many local queer communities, have the urge to “count” and name these violences against LGBTQ+ folks as traumas.
How do we understand this collective urge from within the LGBTQ+ and ally communities to count the adversities and violences they experience as traumas? We suggest that the purpose is less about pinpointing trauma in empiricisms and more about charting collective and cultural patterns of violence and then naming (i.e., validating) those experiences through the very diagnostic category used against them historically in psychological and cultural discourses: the category of trauma.
The instinct to inventory patterns of violence and name them as traumas must be understood in the context of the historical perpetration of violence against LGBTQ+ personhood and community and the systematic undervaluing or elision of LGBTQ+ suffering altogether. In such instances, the source of trauma becomes located in events (e.g., community harassment and assault, family rejection) and in institutions (e.g., the family, the school), which constrict the diversity and expansiveness of human potentiality and relationality to binary gender and monogamous heterosexuality. Pastorally, this naming externalizes the sources (blame) of stress and trauma onto events and institutions, locating them outside the personhood of the LGBTQ+ individual and community. Such a shift is vital for creating community and for bringing a sense of comfort and agency to hurting souls.
Yet, in response to these formulations, some scholars and professionals may worry about the slippage between trauma as response to an event versus trauma as the event itself or even as descriptive of an entire culture. Elaine Miller-Karas, developer of the Trauma and Community Resiliency Models (TRM and CRM)®, embodies a way to hold both approaches together: thinking about trauma in terms of events that culture deems “traumatic” while also recognizing trauma as a neurobiological perception and response to an event or accumulation of events (though culturally we may or may not acknowledge the event nor the accumulation as traumatic). Miller-Karas’s (2015) work invites us to think about the politics of communally naming certain shared experiences, cultural discourses, and social institutions as traumatic. Additionally, we can understand the queer person’s response to such attacks as one that is neurobiologically, emotionally, and socially specific (even mediated), expressing itself in a range of potential responses along a continuum of stress and trauma.
Many LGBTQ+ people are trying to survive, to love themselves and make new families, to find a place to belong and rest their tired heads. They are also searching for a language to make sense of their suffering and their pride. There are defensive reasons why LGBTQ+ communities find themselves using trauma language to describe frequently shared experiences One particularly important reason is to acknowledge and condemn the multivariate ways in which White cis-heteropatriarchy deforms the mind-body-spirits of queer folks in patterned ways, ways intended to maintain societal and familial homeostasis; Murray Bowen (Kerr & Bowen, 1988) would remind us the two are intimately comingled, even co-constitutive. From “Don’t Say Gay” bills in the school to fundamentalist religious movements like Focus on the Family, the attack on LGBTQ+ life comes from multiple vectors and is systematically aimed at the erasure of our community, history, and future. How else to collectively imagine that violence aimed at “the destruction of experience” (Di Nicola, 2018, p.19) except as a kind of collective trauma?
Anti-queer discourse and practice have long histories. In just the history of psychology, theories of trauma and sexuality have long been entangled. Sigmund Freud famously explored the connection between psychological states and somatic symptoms, linking the “symptom(s)” with not just one traumatic event but a “series of associatively linked episodes, beginning in early childhood, all of which needed to be exhumed” (Mitchell & Black, 2016, p. 10). For Freud, trauma indexes the repetition and reenactment of unclaimed and unassimilated experiences, particularly those that completely overwhelm and impair one’s abilities to self-regulate, cope, connect, and make meaning. Freud’s conception of trauma as repetition and reenactment has since received sustained attention, but less attention has been directed towards the context in which his theories of the unconscious and trauma emerged—namely, in the analysis of infantile seduction, childhood sexuality, repressed sexual and identificatory fantasies, and adult sexual perversity.
As Diana Fuss (1995) elaborated, “[I]t is Freud who gives us our most familiar and denigrating sexual typologies, most memorable among them ‘the male homosexual’ . . . and ‘the female homosexual’” (pp. 1–2). Freud (1935) maintained that homosexuality “cannot be classified as an illness,” yet he also declared it to be “a variation of the sexual function produced by a certain arrest of sexual development” (p. 787). Despite Freud’s own nuanced attention to sexuality, the ways that psychoanalytic and psychological ideas have been picked up culturally and politically have historically conflated trauma and unhealth with queerness itself, as if the roots of all psychic traumas are sexual perversity and/or gender incertitude.
While current medical, psychological, and social scientific discourses now work to decouple queerness from illness, defect, and perversity, the association between queerness and unhealth persists in the ways that the “homosexual,” the “transgender,” and the “queer” are still reproduced in cultural and political discourses not as persons but as figures of rupture, break, and discontinuity, marking a crisis and threat to the family, to democracy, and to children, which, as Edelman (2004) famously argued, is a threat to the future itself.
For LGBTQ+ and ally communities to count the common adversities and attacks on LGBTQ+ life as traumas (or potentially traumatic events) is thus an intervention to disassociate the rupture of queerness from the split of trauma. The inherent risk of this strategy is to make trauma “thinkable” and to “lose the spectrality, rupture, nonlinearity, and non-integrability that mark the traumatic as such” (Rubenstein, 2018, p. 286). Yet, if any community can simultaneously mobilize a category, critique it, and keep its signification open, it is the queer community. Di Nicola (2018) writes that “trauma” is not a master term but a flawed one, with a complicated history and no “unified discourse” (p. 18). The same can be said of queerness itself. As with the term queer, queer and trans communities have been reclaiming the word trauma from its solely individual and pathological roots to index sociocultural processes of identity formation, which reveals how a violent world order stresses, traumatizes, and deforms the mind-body-spirits of LGBTQ+ folks by conscripting White cis-heterosexual roles, embodiments, and relations. (It is worth noting that we believe cis-heteropatriarchy harms all people—queer and straight alike—though that harm registers differently for those whose lives rub against the norms and thereby refuse cis-heteronormative assimilation.)
The collective efforts in the LGBTQ+ community to demedicalize and depathologize trauma (Cvetkovich, 2003, p. 25), and to now use this language more broadly to describe a culture, is ultimately a strategy to externalize the source of violence from a problem of the individual to a problem of the social environment. In this sense, LGBTQ+ stress and trauma may be read as appropriate and even adaptive responses to a violent context that seeks to do them harm—again, through strategies like assimilation, abuse, exclusion, silencing, or eradication. Throughout LGBTQ+ history, depathologizing trauma has been a consistent and reliable strategy for establishing queerness on new terms (which is not to deny that trauma may have pathological dimensions). By externalizing the source of their suffering (e.g., I am not the problem; the White cis-heteropatriarchal capitalist world order is), LGBTQ+ folks claim a kind of community, agency, and resistance from within the site of violence. They open new space for rethinking the terms of queer and trans subject formations apart from the ways they have been deformed—i.e., the possibility of queer continuums of resilience-growth-hope in response to but not totalized by queer continuums of stress and trauma.
Queer continuums of stress and trauma
Addressing concerns about oppression and trauma more broadly, feminist psychologist Maria Root (1992) expands traditional notions of trauma to include the “traumatogenic effects of oppression that are not necessarily overtly violent or threatening to bodily well-being at the given moment but that do violence to the soul and spirit” (p. 240). Root (1992) coined the term insidious trauma, arguing that the impact of oppression “shapes a worldview rather than shatters assumptions about the world” (p. 240). Insidious trauma reframes colonialism, racism, homophobia, trans-antagonism, and all such oppressions not as interruptions of an otherwise lithe and free subjectivity but as constitutive of the context and conditions that constrain our emergence as subjects in the first place—what kinds of subjects we can be and what kinds of interactions and relations we can enjoy.
Similarly, Meyer and Frost (2013) support the idea of a “minority stress model,” which “suggests that because of stigma, prejudice, and discrimination, lesbian, gay, and bisexual people experience more stress than do heterosexuals and that this stress can lead to mental and physical disorders” (p. 252). Like insidious trauma, minority stress is a model for attending to the social stressors embedded in the context of a life, inquiring into the impact of both quotidian and explosive stressors such as prejudicial events, structural exclusions, expectations of rejection, pressures of concealment, internalized homophobia, and experiences of harassment and violence (Meyer, 2003). There is no doubt that this short list describes norms and instances of harm and stress. The question is how such “minority stress” impacts persons and communities in the short term and across a lifetime. Meyer and Frost (2013) contend that the assessment of minority stress thus necessitates inquiry into the ways that “health outcomes are determined by the balance of positive (coping and social support) and negative (stressors) effects” (p. 262).
Minority stress and insidious trauma become two ways of conceptualizing the impact of violence—structural, discursive, interpersonal, and intrapersonal—on LGBTQ+ personhood and community. LGBTQ+ people are disproportionately exposed to a range of potentially traumatic events, what many call shock traumas (Stanley, 2019, p. 31), but also to the buildup of chronic minority stress, the accumulation of which we can understand and assess as queer allostatic load.
While, culturally, many of us think of stress and trauma as separate, “[T]hey share a neurobiological basis. Stress and trauma are not inherent in the event—they are internal mind-body responses on a continuum” (Stanley, 2019, p. 32). Traumatic stress then indexes one pole of that continuum, marking the most dysregulated of responses to cis-heteropatriarchal violence when our personal, relational, material, and structural resources for coping and connecting become overwhelmed or destroyed entirely. Whether gradual or sudden, traumatic stress is “the result of a complex interrelationship among psychological, biological, and social processes” and not “a unitary disorder consisting of separate clusters of symptoms” (van der Kolk et al., 2007, p. ix). Remembering that trauma is on a continuum with stress is crucial to help extract queerness from its historical conflation with trauma and relocate trauma in a range of defensive and adaptive survival responses to the stress of oppression.
The stress of oppression is chronic and insidious, like the pressure to conceal a relationship or repress a bodily knowing. It is also sudden and explosive, like physical abuse and family rejection. The ability of LGBTQ+ people to effectively manage and cope with that stress is directly connected to our past experiences (e.g., whether we found support outside and within to complete the stress-recovery cycle) and the ways those experiences live on in the mind-body-spirit and community. It is especially connected to our early social environments and parental attachments, which, to reiterate, are often the sources of LGBTQ+ stress and not its relief. As Felitti et al. (1998) famously revealed, “[T]he impact of . . . adverse childhood experiences on adult health status is strong and cumulative” (p. 251). In other words, those who are already stressed and traumatized are more vulnerable to an ever-accumulating onslaught of stress, trauma, and adversity across a lifetime.
Pastorally, we might ask specific questions about, for example, how the constant neuroception of danger and threat along with the chronic pressures to identify and conceal all build up as stress in the mind-body-spirit of LGBTQ+ people. When such stress is repeated and unmetabolized, frequently due to a lack of social support, it changes the mind-body-spirit system of the person and impairs their capacity to cope, manage future stressors, and even imagine belonging to the future. The concern for us in spiritual care becomes about how we can lighten the queer allostatic load, as well as increase the resources, resilience, and resistance of LGBTQ+ persons to stave off the chronic, insidious, and traumatic stress of interlocking systems of oppression.
The real challenge is how to respond with care to LGBTQ+ stressors and traumas while also confronting their sources. One of the most persistent issues with research into LGBTQ+ stress and trauma is that too often the agents of perpetration go uninterrogated—especially when those perpetrators are pastors, parents, and teachers. In Epistemology of the Closet, Sedgwick (1990, 2008) argued that most discourses about sexuality tend to reproduce a minoritizing view of the subject, conceiving of, for example, same-sex attraction as the exclusive concern of particular people and thus reasserting a neoliberal, autonomous, stable subject. In contrast, Sedgwick advances a universalizing view, one that understands any questions and conflicts of sexuality and subjectivity to be issues of “continuing, determinative importance in the lives of people across the spectrum of sexualities” (p. 1). In this view, sexuality, like identity itself, is tenuous, contingent, and relational, and its effects are not isolated to one person or group but extend to the entire social order of relations. This is not to say that all people are the same but rather that all are connected. It is this connection we have forgotten when we attend to LGBTQ+ stress and trauma but not to holding accountable their abusers—which are both personal and intimate (pastors, parents, teachers) as well as social and cultural (discourses, institutions, structures).
When it comes to assessing LGBTQ+ experiences that fall along the stress-trauma continuum, then, everyone must be implicated, including and especially those who benefit from the preservation and reproduction of the White cis-heterosexual ideal—in the family, school, church, and society. This requires holistic attention to the web of relations and contexts in which identity and social formations occur (Miller-McLemore, 1993, 2005), as well as specific attention to LGBTQ+ stress and trauma as concerns for everyone. Such a project is invested not merely in applying trauma theory to LGBTQ+ lives but also in letting the lived crises of queerness throw all our conceptual tools and subject positions into crisis, ultimately revealing a new starting place based in our interconnectedness and indebtedness to each other.
As Rose (1993, 2017) wrote, “I am abused and I abuse / I am the victim and I am the perpetrator” (p. 31). To think of ourselves in both positions is to locate LGBTQ+ oppression in the discourses, systems, and structures that overdetermine who we can become, who we can connect with, and what those connections entail; it is also to locate the world’s violence in us, not only in our emergence and formation as subjects but also in our quotidian and habitual perpetuation of oppressive norms and practices.
Responding to LGBTQ+ stress and trauma with care first requires attention to our own complicities and propensities to cause harm. Care also requires attention to the range of psychosocial effects and bodily materializations of anti-queer “soul violence” (Sanders, 2020), as well as to the resources and gifts of the LGBTQ+ community for resisting and transforming a violent world order—both in the world and in us. A challenge will always be to respond with care to LGBTQ+ stress and trauma without collapsing queerness back into its conflation with unhealthiness and defect. After all, queerness may predispose a life to crisis, stress, and potential trauma, but only insofar as the world is disordered by White cis-heteropatriarchy. After all, queerness also mobilizes us toward previously foreclosed modes of desire, contact, embodiment, eroticism, community, and even futurity.
Temporality in queering the stress-trauma continuum
As van der Kolk (2014) described it, the experience of trauma involves the tyranny of the past over the present in the lives of traumatized people who “chronically feel unsafe inside their bodies” because “the past is alive in the form of gnawing interior discomfort” (p. 97). The violence experienced in the past becomes so intolerable and overwhelming to the mind-body-spirit system that a series of splits and fractures occur that haunt victims of trauma and return in flashbacks, triggering experiences, and intrusive memories—creating unbearable pasts recapitulating in the present.
While much of what van der Kolk offered is also descriptive of the traumatic experience of some queer people, there are additional ways in which queer experience intersects a stress-trauma continuum along the axis of temporality. For queer people—individuals and collectives—the tyranny of the present over the future and the absenting of collective pasts express two forms of chrono stress and traumatic temporality that should be considered within a pastoral theological assessment of stress and trauma for LGBTQ+ people. These experiences are not entirely unique to queer people, however, and attending to them in the temporalities of queer collectives can also benefit practical theologians and spiritual care practitioners in understanding experiences along the stress-trauma continuum for many others harmed by temporal possibilities conscripted by chrono hegemony.
Additionally, practical theologians and care practitioners exhibit a paucity of focus on temporality in our work to date. As Lester (1995) pointed out nearly three decades ago, “Pastoral theology . . . has ignored a significant aspect of the human condition, namely our temporality—the fact that we are constantly embedded in the context of time, which includes both past and future” (p. 4), limiting our ability to adequately address the dimension of ultimacy bound up with the notion of “hope.” Thus, a queering of trauma frameworks through the lens of queer temporalities may benefit scholars and practitioners in the field more broadly.
Tyranny of the present over the future
The future is always already foreclosed to queer people in “straight time” (Muñoz, 2009, p. 25), that is, the linear time of neoliberal progress, the unending growth of extractive capitalism, and the future portended by heteronormative familial progeny (i.e., heterosexual biological reproduction). This is a temporality in which the organization of time is tied to the organization of bodies in time and through time. In straight time, the present is a hegemonic construct, and any time to come is beholden to a form of presentism in which economic, political, and techno futures are only derivative versions of the present, always presumed to be getting “better” (but for whom?). Here, it is not the past that is continually revisiting the present but instead the present that is continually projected into the future. As Muñoz (2009) stated, however, “The present is not enough. It is impoverished and toxic for queers and other people who do not feel the privilege of majoritarian belonging, normative tastes, and ‘rational’ expectations” (p. 27).
It is not enough to survive the present if potential futures are projections, albeit slightly tweaked, of the current status quo. And it is the future that is at stake in many pervasive experiences of stress and trauma among queer people—not simply the future of the individual but the possible futures for living lives of queerness against the grain of the cis-heteronormative regime.
Those who cannot (or who refuse to) reproduce the proper family or embody the norms of ideal citizens exist outside of time’s straight flow from past to present to future. If trauma involves the experience of being stuck in an unending violent past in which there is no hope for escape or vision of a different future, it must also account for unending projections of violent futures in which there is no possibility for the flourishing of life.
Ratcliffe et al. (2014) argued that in traumatic aftermaths “the experience of time is itself affected. Rather than a change in what is anticipated, arising against a backdrop of intact temporal experience, there is an altered sense of temporal passage, of ‘moving forward’ in time, along with a change in how past, present, future, and the relationship between them are experienced” (p. 1). What is eroded in the traumatic temporality resulting in a foreshortened future is a “style of anticipation.” They explained,Hence a sense of foreshortened future is not a judgment to the effect that the remainder of one’s life will be short and that one has little or nothing to look forward to. It is a change in how time is experienced: an orientation toward the future that is inseparable from one’s experience of past and present, and also from the short- and long-term “passage” of time, is altered. (p. 8)
While the discourse of trauma does not encapsulate the entirety of what needs to be addressed in our relationships to futurity, the concept of a foreshortened future is a helpful framework to explore the nature of care in/for the future with queer and trans people.
The type of traumatic response of a foreshortened future that Ratcliffe et al. (2014) described also seems especially pertinent to the work of spiritual care in larger collective futures of extreme loss and collapse caused by climate change (to name one among a variety of examples of disaster), which compounds chrono stress and traumatic temporalities of queer and trans collectives:The longer-term sense of time is also very different. When the person looks ahead, the future lacks structure; it is not ordered in terms of meaningful projects, and so a coherent sense of long-term duration is absent. Hence the all-enveloping dread she feels before some inchoate threat is not situated in relation to a wider pattern of meaningful temporal events. There is nothing meaningful between now and its actualization, and so it seems imminent. A loss of interpersonal trust that is central to this form of experience is also what sets it in stone. Without the possibility of entering into trusting relations with others, the predicament seems unchangeable. There is no access to the process that might otherwise reveal its contingency and allow her to move beyond it. The person is isolated from others in a way that is incompatible with “moving forward in time”; her life story has been cut short. (Ratcliffe et al., 2014, p. 8)
This loss of trusting interpersonal relationships with others, as noted above, is often a hallmark of queer experience, starting with the family of origin and moving into experiences of school bullying and interpersonal violence. While this has long been noted as a feature of queer experience, its effect on an experience of time and imaginations of futures is lacking in the literature.
A sense of a foreshortened future is similar to what Freeman (2010a, b) described as “narrative foreclosure,” characterized by the conviction that one’s story—the constitutive material of life’s livability—is effectively over. Freeman described this as “the conviction that the story of one’s life, or life work, has effectively ended. At an extreme, narrative foreclosure may lead to a kind of living death or even suicide, the presumption being that the future is a foregone conclusion, an inevitable reiteration of one’s present suffering or paralysis” (p. 125). This recapitulation of the present into the future may have effects as damaging as the tyranny of the past over the present and should be considered as a factor in queer and trans trauma and suicidality (see Sanders, 2020).
But the future is also queered in the lived experience of LGBTQ+ lives. Queer and trans people live into other futures in their very bodies with new names, new expressions, new community constellations, new bodily comportments, even transformed bodies through surgical and hormonal treatments. Whether these potential futures are cut off or not depends on a variety of factors, but even in their failed or curtailed “appearances,” the queer community catches a glimpse of an otherwise world and future. The embodiments of queer futures can be seen in what Muñoz (2009) called an “anticipatory illumination of a queer world, a sign of an actually existing queer reality, a kernel of political possibility within a stultifying heterosexual present” (p. 49). These “otherwise possibilities” (Crawley, 2017) are a threat to powers that are invested in the present regime and are committed to a presentism that extends the status quo into the future. We can observe the stress-trauma dimensions of the temporal in Lothian’s (2018) words pointing to “the affective force that representations of unpleasant futures can carry when they invoke the impossible possibility that there might be no future at all” (p. 57).
Queer and trans communities, through diverging from straight time and the inevitable future of presentism, are confronting the impossibility of that future with “them” in it; antagonizing futures that are unimaginable, intolerable, and ill-fitting for the flourishing of queer lives. A spiraling queer chronology is more shocking, random, and unpredictable than straight time. Queerness helpfully mobilizes a multiplicity of futures, keeping the question of the future always open to resignification and reimagination. As Muñoz (2009) said, “The present must be known in relation to the alternative temporal and spatial maps provided by a perception of past and future affective worlds. . . . [T]he then that disrupts the tyranny of the now in both past and future” (pp. 27, 29). Yet pasts—and their seeming absence from queer consciousness—are another area of potential chrono stress and temporal trauma for queer and trans people.
Absenting of collective pasts
The production and reproduction of the dominant cis-heteropatriarchal capitalist status quo through the organization of bodies via “chrononormativity” (Freeman, 2010a, b, xiii) visits its violence upon collective pasts as much as— in service to—the foreclosing of myriad futures. Caruth (1995) argued, “The ability to recover the past is thus closely and paradoxically tied up, in trauma, with the inability to have access to it” (p. 152). Queer and trans individuals often experience a dearth of past collective stories to integrate into current constructions of the self and future imaginings. From a constructionist standpoint, this might be viewed as a form of constitutive violence taking place at the level of narrative availability. There is no given “community” in the lives of queer and trans people in and through which these stories of collective pasts are kept alive and transmitted to the individual—neither the biological family, social institutions, or churches. While now, more than ever, stories of queer collective pasts have been systemically uncovered, explored, written about, and archived digitally, how these narratives of queer collective pasts are discovered by LGBTQ+ people is much more haphazard. And very few institutions—LGBTQ+-affirming mainline liberal Christian churches included—have taken up the mantle of intentionally telling these stories in pulpits and public forums.3 Thus, one important particularity of experiences that fall along the stress-trauma continuum for queer and trans people is the typical lack of a cohesive sense of belonging to a larger queer/trans collective through time—past and future as much as the present.
As Erikson (1995) argued, “In order to serve as a generally useful concept, ‘trauma’ has to be understood as resulting from a constellation of life experiences as well as from a discrete happening, from a persisting condition as well as from an acute event,” pointing to the possibilities of addressing “traumatized communities” rather than simply “assemblies of traumatized persons” (p. 185). “Trauma can create community” (Erikson, 1995, p. 185). Yet, as Erikson also acknowledged, trauma typically has the effect of damaging the texture of community. In queer and trans collectives, we can observe both the creative and the damaging effects of trauma upon communal possibility, in addition to the lived present of community or the lack thereof.
In some sense, LGBTQ+ people can be considered a “community” because of the perpetuation of injustice and violence against us. We are not bound together by biology or nationality or race or religion. We have no necessary experience of queer kinship until kinship ties are created, and these have often been cultivated to provide affinity spaces within dominant cis-heteropatriarchal publics as well as activist spaces to counter institutional and societal discrimination and violence against queer and trans people. We are bound together by a collective experience of being targeted by our biological families, the legal structures of our nation-states, and our religious communities and are often without a larger queer and trans collective to support us individually or provide us with a sense of a shared collective past.
As Brison (2002) argued, “‘Personal’ stories must be framed by longer historical accounts and by broader social and political ones” (p. 34), but for queer and trans people who have little to no access to narrative sources of a collective past, even as they experience the foreclosure of futures, the chrono stress of absent pasts exacerbates present experiences of stress and trauma. While increasingly archived in the literature and in digital spaces, there may still exist few narrative resources of the broader sociopolitical pasts of queer communal resilience and organizing that are readily available for queer and trans individuals to draw upon in forming “personal” stories that can serve as a bulwark against an encroaching violent present and its vanishing of futures.
Practical theologians and spiritual care practitioners must recognize the possibilities for hope and resilience in the collective past narratives for racial, ethnic, national, and religious minorities—among many others—in addressing the persistence of marginalization, injustice, and violence in the present. Lothian’s (2018) words are instructive here:The lingering presences and possibilities of past futures open possibilities for thinking and living the present in different, deviant ways. If a forward-oriented narrative of historical development signifies the time of capitalism and colonialism, then the time of the colonized, excluded, and othered is most frequently to be found in the past. (p. 20)
For queer and trans people, caring praxis at the site of stress and trauma often lacks the buoying resources of resistance found in these “possibilities of past futures.”
Indeed, as we can witness in the rise of conservative political organizing against teaching the history of race in the United States and in scrubbing public school curricula of any mention of LGBTQ+ people, intervening in the possibilities of collective memory is, itself, a form of violence. This is true not only in the context of the collective but also in the experience of the individual who is kept from knowledge of pasts to which they are connected. This is akin to what Keeling (2019) called the “micro-terrors . . . that we have habituated ourselves to accept” that sever us from “our roots and pasts and histories” through histories of White supremacy, coloniality, and cis-heteropatriarchal power (p. 78).
From a relational view of the self, violence that creates contexts of stress and trauma occurs not only at the level of present ruptures of relational life but in our ruptures with collective pasts as well. Nelson (2001) argued, “Because group identities, like personal identities, are complex narrative structures of meaning . . . oppressive master narratives cause doxastic damage—the damage of distorting and poisoning people’s self-conception and their beliefs about who other people are” (p. 106). Recovery of these absented narratives of collective pasts serves the function of bolstering queer and trans lives against the continued perpetuation of injustice and violence.
In addition to the ways that queer communities have been forged through past experiences of collective trauma and the experience of marginalization and violence, we must also recover queer and trans collective narratives of pleasure, possibility, and flourishing. As Muñoz (2009) stated, “Past pleasures stave off the affective perils of the present while they enable a desire that is queer futurity’s core” (p. 26). Just as queer and trans lives should not be totalized by narratives of trauma and violence in the present—though these are part of queer experience in a dominant cis-heteropatriarchal society—neither should our collective pasts be totalized by these narratives. Queer and trans people have formed community at the margins of family, church, and society for generations, and these collectives of queer life and possibility were shaped by pleasure and relationality outside the restricting dictates of past status quos. These narratives, once recovered, hold the potential to serve as resources for resilience in the living of a queer present and constructing of queer futures.
Constructive proposals for care along a queer stress-trauma continuum
From our examination of queer individual and collective experience along a stress-trauma continuum, four constructive proposals emerge to guide practical theological engagement and spiritual care praxis. Each proposal is an experiment with possibilities for how the notion of a stress-trauma continuum may serve theologians and practitioners well in addressing the myriad harms, injustices, and violence faced by LGBTQ+ people that create stress and trauma in the lives of queer and trans people.
First, a stress-trauma continuum would be helpful to practical theologians and spiritual care providers due to its ability to provide a framework for mapping dysregulation and responses to stressful contexts and traumatic circumstances. As this article demonstrates, LGBTQ+ people encounter myriad occurrences of trauma, injustice, violence, and stress. Key to the usefulness of a stress-trauma continuum for queer and trans persons, however, is the ability to narrate this stress-trauma continuum and the map of dysregulation and response to stressful and traumatic circumstances from queer perspectives. Psy discourses have long perpetuated violence against queer and trans people. A new language and framework for understanding LGBTQ+ stress and trauma should foreground the perspectives of queer and trans folks and not be beholden to diagnostic criteria, which, formulated without queer and trans people, have reinforced heteronormative and gender dichotomous assumptions in psychotherapy (Butler & Byrne, 2008). Relatedly, this framework must acknowledge that experiences that fall along the stress-trauma continuum are a part of queer and trans experience but that LGBTQ+ subjectivity is not totalized by trauma.
Second, in this vein, we hope that a stress-trauma continuum will also prompt the development of a resilience-growth-hope continuum, highlighting the myriad ways that queer and trans people behave agentially, with creativity and great resolve, to resist the violence that is perpetuated in relation to us. And this resistance is a part of the personal and communal response to trauma in our history and experience. Traumatic experience has, in part, helped to form us into a community because of our ability, demonstrated again and again, to resist the violences of a cis-heteropatriachal status quo. These resistance experiences (and absented pasts) must affect, inflict, and inflect the stress-trauma continuum for the framework to center care on queer and trans people as well as beyond queer experiences. Our conviction is that queer and trans experience has much to teach us about stress and trauma more broadly, going beyond the specificity of LGBTQ+ experience, and that many of these lessons rest upon a resilience-growth-hope continuum of resistance.
Third, trauma discourse—and perhaps even a stress-trauma discourse—too easily focuses upon the experience of the traumatized while eliding focus on the perpetuation of the injustices and violence that create the stress-trauma experience in the first place. Of vital importance in our sense of this framework’s usability is its ability to locate violence in discourses, systems, and structures that (over)determine who we can become, with whom we can connect across time and space, and what those connections entail. While it is essential to focus upon healing those who have been victimized by stress and trauma, it is also vital to name the locations from which the perpetuation of harm emanates. We too easily name these locales of injustice and violence on an individual level—also the level of the trauma upon which we typically focus—without noting the institutional and systemic levels at which they are perpetrated. With a focus upon the sources of violence and trauma, we must also locate the ways that violence works to conscript something in us or about us, charting the introjection of those violences and examining how we become formed and deformed according to a discursive, political, material regime of White cis-heteropatriarchal normativity.
Finally, we hope that a stress-trauma continuum would allow for a more robust engagement with the ways that stress and trauma are experienced across temporalities as well as how they are experienced temporally. Interventions that develop in the wake of a queer stress-trauma continuum should be interventions that not only address the relational and material experiences of queer and trans people in the present—of critical importance, no doubt—but also lead practitioners and theologians into investigating ways of restoring absented pasts and reimagining vanishing futures for and with LGBTQ+ people. We need a collective history to survive and thrive in a present that continues to attack our bodies and assail our souls. And if we are to have a future, it will be a future of flourishing beyond the restricting dictates of a cis-heteropatriarchal persisting present that continually conscripts the possibilities for queer and trans life.
A queer stress-trauma continuum, along with a queer resilience-growth-hope continuum, must take shape as a spiraling temporality that continually draws upon past collective experience to meet the violence and injustices of the present, projecting queer imaginings of a future that is not conscripted or overdetermined by a projection of the present into a future overdetermined by the White cis-heteropatriarchal present. The past, the present, and the future must be open, fluid, undecided, and resilient—therein, they must be queer.
Declarations
Conflict of interest
The authors declare that they have no conflict of interest.
1 Throughout this article, we use “queer and trans” as well as “LGBTQ+” to index the vast array of identifications, embodiments, and relationalities outside of the prescriptions of cisgender heterosexual patriarchy (cis-heteropatriarchy), including but not limited to those who identify as lesbian, gay, bisexual, transgender, two-spirit, gender-nonconforming, or queer, as well as those who choose to live against identificatory labels. We also use “queer and trans” interchangeably with “LGBTQ+” as this reflects common cultural usage within these communities. As we trace common patterns and experiences, however, by no means do we mean to imply a homogeneity to the queer and trans experience.
2 The Pew Research Center (2013) reported that 4 in 10 (or 39%) of U.S. LGBT adults disclosed that they had been rejected by family or friends.
3 An anecdotal way to assess what we are arguing here might also be thinking back to times in the reader’s own faith community when you’ve heard the history of LGBTQ+ religious organizing or faith practice iterated from the pulpit or within literature published by your denomination or religious organization. Even in LGBTQ+-affirming faith communities, these histories are not regularly made public or readily accessible to those who aren’t actively looking for them and who may not even know that they exist to be discovered.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Barba, A., Mooney, M., Giovanni, K., Clarke, M., Grady, J. B., & Cohen, J. A. (2021). Identifying the intersection of trauma and sexual orientation and gender identity part I: Key considerations. National Child Traumatic Stress Network.
Brison, S. J. (2002). Aftermath: Violence and the remaking of a self. Princeton University Press.
Butler, C., & Byrne, A. (2008). Queer in practice: Therapy and queer theory. In L. Moon (Ed.), Feeling queer or queer feelings? Radical approaches to counselling sex, sexualities and genders. Routledge.
Caruth C Caruth C Recapturing the past: Introduction Trauma: Explorations in memory 1995 Johns Hopkins University Press 151 157
Choi, S. K., Wilson, B. D. M., Shelton, J. Gates, G. (2015). Serving our youth 2015: The needs and experiences of lesbian, gay, bisexual, transgender, and questioning youth experiencing homelessness. The Williams Institute with True Colors Fund. https://williamsinstitute.law.ucla.edu/wp-content/uploads/Serving-Our-Youth-Update-Jun-2015.pdf
Crawley, A. T. (2017). Blackpentecostal breath: The aesthetics of possibility. Fordham University Press.
Cvetkovich, A. (2003). An archive of feelings: Trauma, sexuality, and lesbian public cultures. Duke University Press.
Di Nicola V Boynton E Capretto C Two trauma communities: A philosophical archaeology of cultural and clinical trauma theories Trauma and transcendence: Suffering and the limits of theory 2018 Fordham University Press 17 52
Dulmus CN Hilarski C When stress constitutes trauma and trauma constitutes crisis: The stress-trauma-crisis continuum Brief Treatment and Crisis Intervention 2003 3 1 27 35 10.1093/brief-treatment/mhg008
Edelman, L. (2004). No future: Queer theory and the death drive. Duke University Press.
Erikson K Caruth C Notes on trauma and community Trauma: Explorations in memory 1995 Johns Hopkins University Press 183 199
Felitti VJ Anda RF Nordenberg D Williamson DF Spitz AM Edwards V Koss MP Marks JS Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: The Adverse Childhood Experiences (ACE) Study American Journal of Preventive Medicine 1998 14 4 245 258 10.1016/s0749-3797(98)00017-8 9635069
Freeman, E. (2010a). Time binds: Queer temporalities, queer histories. Duke University Press.
Freeman, M. (2010b). Hindsight: The promise and peril of looking backward. Oxford University Press.
Freud, S. (1951). Letter to an American mother. American Journal of Psychiatry, 105, 787. (Original work published 1935)
Fuss D Pink Freud GLQ 1995 2 1 9 10.1215/10642684-2-1_and_2-1
Griffin, H. (2006). Their own receive them not: African American lesbians and gays in Black churches. Pilgrim Press.
Human Rights Campaign Staff. (2018). 2018 LGBTQ youth report. Human Rights Campaign. https://hrc-prod-requests.s3-us-west-2.amazonaws.com/files/assets/resources/2018-YouthReport-NoVid.pdf
Human Rights Campaign Staff. (2023). Human Rights Campaign working to defeat 340 anti-LGBTQ+ bills at state level already, 150 of which target transgender people—Highest number on record. Human Rights Campaign. https://www.hrc.org/press-releases/human-rights-campaign-working-to-defeat-340-anti-lgbtq-bills-at-state-level-already-150-of-which-target-transgender-people-highest-number-on-record
Janssen, D.-J., & Scheepers, P. (2019). How religiosity shapes rejection of homosexuality across the globe. Journal of Homosexuality, 66(14). 10.1080/00918369.2018.1522809
Keeling K Queer times 2019 New York University Press
Kerr, M. E., & Bowen, M. (1988). Family evaluation. W. W. Norton & Co.
Kosciw, J. G., Gretak, E. A., Zongrone, A. D., Clark, C. M., & Truong, N. L. (2018). The 2017 National School Climate Survey: The experiences of lesbian, gay, bisexual, transgender, and queer youth in our nation’s schools. GLSEN.
Lester, A. D. (1995). Hope in pastoral care and counseling. Westminster John Knox.
Lothian, A. (2018). Old futures: Speculative fiction and queer possibility. New York University Press.
Meyer IH Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence Psychological Bulletin 2003 129 5 674 697 10.1037/0033-2909.129.5.674 12956539
Meyer IH Frost DM Patterson CJ D’Augelli AR Minority stress and the health of sexual minorities Handbook of psychology and sexual orientation 2013 Oxford University Press 252 266
Miller-Karas, E. (2015). Building resilience to trauma: The trauma and community resiliency models. Routledge.
Miller-McLemore, B. J. (2005). The living human web. In Dykstra, R. C. (Ed.), Images of pastoral care: Classic readings (pp. 40–46). Chalice Press. (Original work published 1993)
Mitchell, S. A., & Black, M. J. (2016). Freud and beyond: A history of modern psychoanalytic thought. Basic Books.
Muñoz, J. E. (2009). Cruising utopia: The then and there of queer futurity. New York University Press.
Mustanski BS Garofalo R Emerson EM Mental health disorders, psychological distress, and suicidality in a diverse sample of lesbian, gay, bisexual, and transgender youths American Journal of Public Health 2010 100 12 2426 2432 10.2105/AJPH.2009.178319 20966378
Nelson HL Damaged identities, narrative repair 2001 Cornell University Press
Pew Research Center. (2013). A survey of LGBT Americans: Attitudes, experiences and values in changing times. Pew Research Center.
Ratcliffe M Ruddell M Benedict Smith B What is a ‘sense of foreshortened future?’ A phenomenological study of trauma, trust, and time Frontiers in Psychology 2014 5 1 11 10.3389/fpsyg.2014.01026 24474945
Rooney, C., & Durso, L. E. (2017). The harms of refusing service to LGBTQ people and other marginalized communities. Center for American Progress.
Root M Brown LS Ballou M Reconstructing the impact of trauma on personality Personality and psychopathology: Feminist reappraisals 1992 Guilford Press 229 266
Rose, G. (2017). Judaism and modernity: Philosophical essays. Verso. (Original work published 1993)
Rosky, C. (2017). Anti-gay curriculum laws. Columbia Law Review, 117(8), 1461-1522. https://ssrn.com/abstract=2931244
Rubenstein M-J Boynton E Capretto C Afterword—The transcendence of trauma: Prospects for the continental philosophy of religion Trauma and transcendence: Suffering and the limits of theory 2018 Fordham University Press 283 294
Sanders CJ Christianity, LGBTQ suicide, and the souls of queer folk 2020 Lexington Books
Sedgwick, E. (2008). Epistemology of the closet. University of California Press. (Original work published 1990)
Servigne, P., Stevens, R., & Chapelle, G. (2021). Another end of the world is possible: Living the collapse. Polity Press.
Stanley, E. (2019). Widen the window: Training your brain and body to thrive during stress and recover from trauma. Penguin Random House.
Trevor Project. (2020). The 2020 National survey on LGBTQ youth mental health. Author.
Trevor Project. (2022). The 2022 National survey on LGBTQ youth mental health. Author.
van der Kolk BA The body keeps the score: Brain, mind, and body in the healing of trauma 2014 Penguin Books
van der Kolk, B. A., McFarlane, A. C., & Weisaeth, L. (Eds.). (2007). Traumatic stress: The effects of overwhelming experience on mind, body, and society. The Guilford Press.
|
PMC010xxxxxx/PMC10173243.txt |
==== Front
Nat Rev Bioeng
Nat Rev Bioeng
Nature Reviews Bioengineering
2731-6092
Nature Publishing Group UK London
63
10.1038/s44222-023-00063-3
Review Article
Human disease models in drug development
Loewa Anna 1
http://orcid.org/0000-0002-7141-5823
Feng James J. 23
http://orcid.org/0000-0001-6770-3657
Hedtrich Sarah sarah.hedtrich@bih-charite.de
1456
1 grid.6363.0 0000 0001 2218 4662 Department of Infectious Diseases and Respiratory Medicine, Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt Universität zu Berlin, Berlin, Germany
2 grid.17091.3e 0000 0001 2288 9830 Department of Chemical and Biological Engineering, University of British Columbia, Vancouver, BC Canada
3 grid.17091.3e 0000 0001 2288 9830 Department of Mathematics, University of British Columbia, Vancouver, BC Canada
4 grid.484013.a 0000 0004 6879 971X Center of Biological Design, Berlin Institute of Health at Charité – Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt Universität zu Berlin, Berlin, Germany
5 grid.17091.3e 0000 0001 2288 9830 Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, BC Canada
6 grid.211011.2 0000 0001 1942 5154 Max-Delbrück Center for Molecular Medicine (MCD), Helmholtz Association, Berlin, Germany
11 5 2023
115
30 3 2023
© Springer Nature Limited 2023, corrected publication 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Biomedical research is undergoing a paradigm shift towards approaches centred on human disease models owing to the notoriously high failure rates of the current drug development process. Major drivers for this transition are the limitations of animal models, which, despite remaining the gold standard in basic and preclinical research, suffer from interspecies differences and poor prediction of human physiological and pathological conditions. To bridge this translational gap, bioengineered human disease models with high clinical mimicry are being developed. In this Review, we discuss preclinical and clinical studies that benefited from these models, focusing on organoids, bioengineered tissue models and organs-on-chips. Furthermore, we provide a high-level design framework to facilitate clinical translation and accelerate drug development using bioengineered human disease models.
Owing to the high failure rates of the current drug development process, biomedical research is undergoing a paradigm shift towards approaches centred on human disease models. This Review critically discusses translationally relevant examples and defines key milestones for their widespread application.
Key points
Advances in bioengineering have yielded complex human disease models with high clinical biomimicry and predictability.
Human disease models can help unravel disease mechanisms, including for infectious and genetic diseases and cancer.
Using appropriate human disease models in the drug development process and clinical decision-making improves the rate of clinical translation, reduces costs and directly benefits patients.
Stringent model validation, regulatory and legal guidance, and scalable disease model production are key future milestones to facilitate their implementation in (pre-)clinical research.
Subject terms
Translational research
Molecular medicine
==== Body
pmcIntroduction
Biomedical research is currently undergoing a paradigm shift towards approaches centred on human disease models1. This shift has been driven by the notoriously high failure rates of the current drug development process. Although investments increased at unprecedented rates over the past decade (US$133 billion research and development expenditures of the 15 biggest pharma companies in 2021, a 44% increase since 2016)2, the drug attrition rate hit an all-time high of 95% in 2021 (ref. 3). Most drugs fail in clinical stages (Fig. 1) despite proven efficacy and safety in animal models4,5. Different reasons account for this translational gap, one of them being that the decision on entry of a drug candidate into clinical trials relies almost exclusively on animal-derived data.Fig. 1 Drug development pipeline.
Current development pipeline of new drugs with proven preclinical safety and efficacy in animal models. The average duration of the different (pre-)clinical stages, the percentage of drugs (averaged over the past 5 years) that move to the next phase and the median costs of the different stages per drug are illustrated3,180.
Animal models, however, often fail to filter out harmful or ineffective drugs6. Moreover, potentially effective drug candidates never enter clinical trials owing to negative preclinical tests given that most animal models poorly resemble human conditions and thus have low predictive values. The discrepancies derive from different anatomical layouts and biological barriers, divergent receptor expression and immune responses, host specificities of microorganisms, and distinct pathomechanisms. In addition, animals are inbred and kept under standardized conditions and thus do not account for the genetic and ethnic diversity of humans. Therefore, drug safety or efficacy issues that only affect certain subpopulations go unnoticed.
Furthermore, fast-paced advances in genome editing and antibody therapies have direct implications for the drug development process. Currently, 40% of the drugs undergoing clinical trials are antibodies3. However, their high target specificity requires the identification of cross-reactive species during preclinical testing. Non-human primates are often the only pharmacologically relevant species, whose use has ethical and economic implications7. Species-specificity is also a concern for gene therapies because genetic sequences and therapeutic efficacy differ between animals. For example, base editors yield 61% in vivo gene editing efficacy in the liver of mice compared to 26% in primates8.
The COVID-19 pandemic further highlighted the model dilemma in biomedical research. At the beginning of the pandemic, it was unclear which species were susceptible to SARS-CoV-2 and no model was readily available to study the course of the disease and identify druggable targets against the unknown pathogen9. Suitable and readily available disease models could have substantially accelerated our understanding of virus–host interactions and expedited the identification of effective drugs in repurposing studies.
In view of the deficits of such an animal-centred system, efforts are ongoing to develop bioengineered human-based (disease) models of high clinical biomimicry to close this translational gap.
This Review discusses success stories and applications of human (disease) models in preclinical and clinical research, focusing on organoids, bioengineered tissue models and organs-on-chips (OoCs).
Overview of human disease models
Different disease models are available covering a broad range of physiological and pathological conditions. An overview of their characteristics, advantages and disadvantages is provided herein.
2D cell cultures
Two-dimensional (2D) cultivated patient-derived cells are an invaluable tool to study disease phenotypes and pathomechanisms, especially during the early phases of drug development (Fig. 2). Primary cells are the preferred option owing to their higher genetic heterogeneity compared to cell lines, but their limited availability or in vitro proliferation capacity restricts their use. Multipotent adult stem cells (ASCs) and induced pluripotent stem (iPS) cells can help overcome this shortage, as they can readily and indefinitely propagate and convert into any somatic cell10. Engineered cells, such as reporter cell lines, are also commonly used as they are amenable to high-throughput manufacturing and have high reproducibility and lower costs relative to stem cell-based methods.Fig. 2 Overview of different disease models.
From left to right, bioengineered models are ordered from the least to the most complex, including their advantages, limitations and stage of application during the drug development process. iPS cell, induced pluripotent stem cell.
Nonetheless, 2D cell cultures have limitations. For example, the epithelial differentiation stages of stratified epithelia, such as the skin, cannot be mimicked in 2D. Moreover, cell responses and gene and/or protein expression patterns greatly differ between 2D and three-dimensional (3D) models11. For example, 3D cultures of lung fibroblasts resemble in vivo tumour necrosis factor (TNF) receptor expression and nuclear factor κ-light-chain-enhancer of activated B cells (NF-κB) activation patterns more closely than 2D cultures12. Similarly, the biomimicry of tissue-specific transporters and cell junctions is often higher in 3D models13.
Bioengineered tissue models
Bioengineered tissue models are primarily generated from human stem cells or primary cells, the latter of which are isolated from surgically excised human tissue or from non-transplantable organs (Fig. 2). The cells are manually added or 3D bioprinted onto or into hydrogel or polymer-based scaffolds14, often in transwell setups15. Alternatively, de-cellularized extracellular matrix scaffolds from animal-derived organs16 or non-transplantable human organs17 can be reseeded with human cells. The advantages of these models include the possibility to cultivate them at the air–liquid interface or submerged, thereby emulating in vivo-like tissue conditions and their maturated and differentiated tissue state, and achieving increased biomimicry of native human tissue compared to 2D18,19. This approach is often used for multi-layer or stratified tissues such as the gut, lungs and skin20,21 but also the brain and liver as it allows a controlled build-up of the model22,23. However, these tissue models cannot be cryopreserved or propagated and therefore have a limited lifespan. Furthermore, they have limited cell diversity and lack self-renewal capacity in non-stem cell-based models.
Organoids
Organoids are self-organizing 3D structures generated from tissue-specific ASCs or iPS cells24 (Fig. 2). Stem cells are embedded in an extracellular matrix that provides the scaffold for tissue growth. Organoid formation is then guided by a cocktail of growth factors that are pivotal for tissue development in vivo, which enables stem cells to maintain their differentiation and self-renewal capacity. The size of these organ-like structures ranges from 100 µm (lung organoid) to 2 mm (brain organoid) and they can be cultivated in small dimensions (for example, 96-well or 384-well plates25,26), thereby being suitable for high-throughput screening.
Protocols for the generation of iPS cell-derived organoids are available for a variety of human organs, including the gut27, stomach28, liver29, pancreas30, lung31, thyroid32, kidney33, retinal and optic cup34, and brain35. However, iPS cell-derived organoids fail to mature beyond the fetal phenotype unless they are grafted into living organisms36,37. To overcome this limitation, the use of tissue-resident ASCs derived from postnatal or adult tissues has gained increasing attention, resulting in organoids of endoderm-derived tissues such as lung38, gut39, pancreas40, stomach41, liver organoids42, endometrium43, prostate44, fallopian tube45 and mammary gland46. Although the application of ASCs has been hampered by their limited availability in the past, some commercial suppliers now provide ASC-derived organoids. Nonetheless, ASC-derived organoids lack cells from the ectodermal and mesenchymal germ layer and are usually smaller than iPS cell-derived counterparts. Furthermore, the cell type composition of organoids of the same organ can vary depending on the protocol47, ultimately hindering reproducibility.
Organs-on-chips
Inter-tissue crosstalk and complex in vivo-like processes cannot be mimicked in single and static tissue models. To enable tissue perfusion and dynamics as well as multi-organ crosstalk, OoCs have been introduced48–51 (Fig. 2). OoCs are perfused microfluidic platforms that contain bioengineered or miniaturized tissues or organs interconnected by 3D microchannels to simulate the in vivo functions, biomechanics and (patho)physiological responses of organs6,52–54. OoC setups are often referred to as microphysiological systems, as they can emulate human (patho)physiology in a more human-like environment49,51,55–57. As opposed to organoids that form by self-organization, OoCs follow a reductionist engineering approach through the targeted and pre-defined design of components, such as the scaling of cell numbers used based on their physiological function, or including disease-relevant cell types and key biophysical and biochemical cues24.
OoCs can be single-organ or multi-organ systems. Single-organ systems can assess the response of a specific tissue or organ to a particular stimulus. By contrast, multi-organ systems allow study of the communication and interactions of several tissues and organs simultaneously. Notably, the interconnection of at least 10 human organs, including circulatory, endocrine, gastrointestinal, immune, integumentary, musculoskeletal, nervous, reproductive, respiratory and urinary, in OoCs has been hypothesized to provide sufficient complexity to resemble a ‘human-on-a-chip’58–60. Furthermore, biophysical processes, such as chip perfusion, can be easily automated using OoCs, which allows monitoring of tissue function and responses in situ and in real time52. For these reasons, OoCs currently constitute the most promising approach to emulate human diseases in vitro55,61. However, they are still complex and thereby not amenable to high-throughput methods.
Preclinical and clinical applications
Preclinical research
Human-based models are invaluable tools in every stage of the drug development process, from high-throughput screens to identify target or lead molecules, to efficacy and safety testing in preclinical stages and implementation in clinical trials and decision-making. Nonetheless, human-based models have so far been primarily used for toxicity testing61–64, and their application in preclinical and clinical research is still in its infancy. Promising disease models of the liver65,66, pancreas67,68, central nervous system35,69,70, skin71, lung72,73, intestine72–76, musculoskeletal system77 and heart78–81 have been developed, but their equivalence or superiority over their animal models needs to be demonstrated. In this section, successful examples of preclinical and clinical implementation of human-based disease models are discussed.
Unravelling disease mechanisms and target identification
Rational drug design and development require a proper understanding of the underlying disease mechanism. Human disease models can facilitate the process as evidenced by ad hoc systems that emerged during the COVID-19 pandemic1. For example, central pathomechanisms, such as the contribution of endothelial dysfunction, the ‘cytokine storm’ and the intercellular variability of infection susceptibility, were identified in organoids and lung-on-a-chip models82–86. A vasculature-on-a-chip setup revealed that SARS-CoV-2 exposure substantially reduces endothelial barrier function by perturbing vascular endothelial–cadherin junctions and increasing pro-inflammatory cytokine release. This endothelial dysfunction is exacerbated after the introduction of peripheral blood mononuclear cells, resulting in excessive inflammation82. Organoids demonstrated the induction of inefficient interferon responses in SARS-CoV-2 infections84,85, revealing similar interferon signatures between SARS-CoV-2-infected bioengineered lung models and native human lungs, highlighting the value and predictivity of human models84.
Bioengineered tissues also contributed to the understanding of the effects of COVID-19 beyond the lungs. For example, SARS-CoV-2 tropism for neurons87 and cortical astrocytes88 or its preference for mature cell types and distinct neurotoxic effects were demonstrated in brain organoids. Interestingly, the expression levels of the SARS-CoV-2 receptor angiotensin-converting enzyme 2 (ACE2) in brain organoids are substantially lower than in the lung or intestinal epithelium. In the brain, alternative receptors, such as CD147 and DPP4, are required for infection and virus replication as shown in astrocytes88.
Similarly, single-cell RNA sequencing (scRNA-seq) of kidney organoids showed that SARS-CoV-2 directly infects kidney cells, triggering pro-fibrotic events and features of polycystic kidney disease, closely resembling renal processes observed in individuals who are critically ill with COVID-19 (refs. 89,90). Furthermore, transcriptomic profiling of infected kidney organoids aligned with the proteomic profile of the urine of individuals who were critically ill with COVID-19, further validating the findings of the model.
Similar examples can be found for other diseases such as hepatitis B91, allergic asthma92, chlamydia93 and influenza94,95. A true scientific leap was the establishment of an in vitro cultivation method for human norovirus in 2016 (ref. 96). Prior to that, numerous attempts to cultivate norovirus in intestinal epithelial cells or primary immune cells had failed owing to a lack of mimicry of the natural host environment, thereby preventing a full mechanistic understanding of norovirus infections96. Infection studies using human ASC-derived intestinal organoids provided the solution as they closely emulate the host environment by differentiating into physiologically active, multicellular epithelial tissues. Eventually, enterocytes were identified as the primary target for viral infection, which is pivotal when aiming, for example, for drug target identification. Ultimately, this discovery laid the foundation for numerous ongoing efforts to develop preventive or therapeutic97 measures, including a norovirus vaccine98.
Human-based models also contributed to an in-depth understanding of the mechanisms of infection of the Zika virus, which caused an outbreak in 2015 in South America. Rodent models cannot fully reproduce the mechanism by which the virus causes microcephaly because the human brain has an additional cortical layer containing radial glial cells (progenitor cells). However, infection studies in brain organoids showed that the Zika virus primarily infects neural progenitor cells, which increases neural cell death and reduces proliferation, thereby hampering neurogenesis and causing microcephaly99,100. This breakthrough provided the community with an experimental platform to screen for therapeutic options.
A recent discovery of similar magnitude is that infecting intestinal organoids with strains of Escherichia coli that produce the genotoxin colibactin revealed mutation signatures that could drive colorectal cancer (CRC) progression, similar to those found in human tumours101. Specifically, genotoxic E. coli strains were found to cause single base substitutions and indel formations in the intestinal epithelium and other tissues such as the urinary tract. These findings could have broad implications; for example, the detection and eradication of genotoxic E. coli strains could decrease the risk of intestinal cancer in large cohorts. Furthermore, the E. coli strain Nissle 1917, commonly used as a probiotic, also produces colibactin102, warranting re-evaluation of its use.
OoC devices can be used to study the pathomechanism of diseases that involve two or more organs. For example, the gut–liver–brain axis has long been postulated to contribute to Parkinson disease through short-chain fatty acids (metabolites of the intestinal microbiome) that can, directly and indirectly, promote neurodegeneration. This connection has been confirmed using an OoC setup that emulates the immune–metabolic crosstalk between these tissues103. Interestingly, co-cultivation of liver, gut, and a mixture of neurons, astrocytes and microglia facilitated the maturation of the brain model, which is difficult to achieve in single-tissue models. To emulate the disease state, iPS cells derived from individuals with Parkinson disease were differentiated into a cerebral model, co-cultured with a healthy gut and liver model, and exposed to circulating regulatory T and T helper 17 cells. Multi-omics and multiplexed cytokine and chemokine analysis revealed that short-chain fatty acids increase the expression of pathways related to ferroptosis in Parkinson disease models, a well-established cause of dopaminergic cell death in Parkinson disease. Animal models are not amenable to decoupling the effect of single parameters with such accuracy. Another example is an OoC model of the neurovascular unit infected by Cryptococcus neoformans, a fungus that causes fungal meningitis. This system revealed that C. neoformans trespasses the blood–brain barrier through transcytosis, providing a potential therapeutic target to inhibit this process104.
Drug screening and efficacy testing in 2D
Advances in patient-derived iPS cells have re-established the role of 2D cultures in preclinical drug development, leading to several breakthroughs in neurodegenerative diseases, conditions that have been historically plagued by the highest failure rates (≥97%) in drug development105. Since 2011, five approved first-in-class drugs for neurodegenerative diseases were identified through phenotypic screening of 2D cultures105. In phenotypic screens, the readout is based on alterations of the phenotype of a diseased cell or tissue (model), whereas no specific drug target is known. These screens, which are typically conducted in genetically engineered cell lines that harbour the target of interest, led to the discovery of the splice modulators risdiplam106 and branaplam107 for the treatment of spinal muscular atrophy, an autosomal recessive disease characterized by the degeneration of motor neurons. The efficacy of the lead compounds was assessed in iPS cell-derived motor neurons showing splice correction and restored protein levels. Risdiplam received its marketing authorization in the EU in 2021 and became the first orally available medication for spinal muscular atrophy.
Combinations of phenotypic screens with multi-omics technologies can further increase the success in early phases of drug development, especially for diseases with a complex genetic background10. For example, a screen of small molecules that correct dysregulated gene networks in NOTCH1-deficient calcific aortic valve disease (CAVD) was combined with a machine learning algorithm that was trained to classify gene expression levels as wild type or diseased, allowing the detection of target molecules in patient-derived iPS cells108. Almost 1,600 molecules were screened using targeted RNA sequencing, resulting in the identification of an inverse agonist of oestrogen-related receptor-α as the lead compound, which proved effective in correcting dysregulated CAVD-relevant genes in primary aortic valve endothelial cells from individuals with CAVD and in genetically modified mice.
Drug screening and efficacy testing in 3D
During the COVID-19 pandemic, 3D human-based infection models proved extremely valuable for the identification of new antiviral strategies82,109 and drug repurposing with SARS-CoV-2-dampening effects51,110–112. For example, high-throughput screens of ≥1,000 drugs approved by the FDA run on colonic and lung organoids identified several SARS-CoV-2 entry inhibitors such as imatinib, mycophenolic acid and quinacrine dihydrochloride111. Similarly, among eight drugs with dose-dependent inhibition of virus uptake in static Huh-7 cell monolayers, only three (amodiaquine, toremifene and clomiphene) proved effective in 3D OoC cultures51. Building on these results, amodiaquine (for example, NCT04532931, NCT04502342)51 and imatinib111 (for example, NCT04394416) have entered clinical trials.
Interestingly, applying physiological breathing mechanics in alveoli-on-chip models showed antiviral activity by increasing the expression of interferon-related genes and improving host defence113. For example, an increase in S100A7 expression was observed which codes for alarmins and receptor for advanced glycation end products (RAGE) ligands. The RAGE inhibitor azeliragon was then identified as effective by suppressing the overshooting of inflammatory responses as observed in patients who were severely ill with COVID-19. This data has been submitted as part of a pre-investigational new drug application to the FDA. Notably, the antimalarial drugs chloroquine and hydroxychloroquine, which received controversial attention during the COVID-19 pandemic, proved non-effective in a human lung-on-chip disease model, similar to clinical findings51.
Major advances against cancer have also been achieved using human disease models. For example, MCLA-158, a bispecific antibody binding epidermal growth factor receptor (EGFR) and leucine-rich repeat-containing G-protein-coupled receptor 5 (LGR5), was identified as the most effective antibody against wild-type and KRAS-mutant CRC after screening a large biobank of patient-derived CRC organoids. Importantly, MCLA-158 reliably discriminated between cancerous and healthy cells. These findings relied entirely on human-derived cancer organoids and went from bench to bedside in just 3 years (NCT03526835)114.
Similarly, the efficacy of amivantamab, a bispecific antibody against EGFR and the mesenchymal–epithelial transition receptor, was preclinically tested in patient-derived cells and organoids that harbour EGFR Exon20ins mutations115. Patients with this mutation have a poor prognosis because standard drugs, such as tyrosine kinase inhibitors, are ineffective. These results were replicated in clinical trials (for example, NCT04538664), eventually leading to the approval of amivantamab by the EMA for the treatment of non-small cell lung cancer.
Another successful example is the use of OoC models of chronic inflammatory demyelinating polyneuropathy, an autoimmune disease that cause muscle weakness, conduction blocks and aberrant spinal reflexes116. An OoC-based chronic inflammatory demyelinating polyneuropathy electrical conduction model, consisting of human iPS cell-derived motor neurons and Schwann cells cultivated on microelectrode arrays, closely emulates clinically relevant features such as muscle contraction and electrical activity. Treatment with the antibody TNT005, a specific complement component 1s inhibitor, inhibited the immune reaction and rescued the functional deficits, prompting the FDA to approve clinical trials (NCT04658472).
Clinical research
Beyond preclinical research, human disease models are also revolutionizing clinical decision-making. A prime example is cystic fibrosis, a heterogeneous genetic disease with limited treatment options and highly variable treatment outcomes depending on the disease-causing mutation (≥2,000 mutations reported). Cystic fibrosis is caused by loss-of-function mutations in the CFTR gene resulting in highly viscous mucus, which blocks the airways, limits the ability to breathe and causes persistent lung infections. The life expectancy of a newborn with cystic fibrosis in high-income countries is 55 years today.
A breakthrough in the treatment of cystic fibrosis was the development of CFTR modulators such as ivacaftor or lumacaftor — effective but also expensive drugs. Ivacaftor is registered for the treatment of patients with nine CFTR gating mutations, which make up only 5% of all individuals with with the disease, although combinations with other modulators can improve clinical responses and patient eligibility117.
However, identifying potential responders is difficult, especially for patients with rare mutations. To overcome this limitation, testing of treatment responses in intestinal patient-derived organoids (PDOs) has emerged as a reliable and predictive approach118,119. For example, rectal organoids from 71 individuals with cystic fibrosis that harboured 28 different CFTR mutations revealed residual CFTR function and identified effective drug combinations. In this case, measuring the lumen area of the organoids allowed comparison of CFTR function among patients, thereby providing physicians with a reliable predicting tool for clinical decision-making.
Similar applications are being pursued with cancer PDOs120–122, revealing strong correlations between in vitro and in vivo drug responses against cancer in the gastrointestinal tract123, bladder124, ovary125, rectum126 and pancreas127,128. For example, a living biobank of cancer PDOs from patients with metastatic and pre-treated colorectal and gastroesophageal cancer shows strong morphological, genotypic and spatiotemporal similarities between the primary tumour and the PDOs123. Furthermore, PDOs reliably predicted treatment responses in patients, with 100% sensitivity, 93% specificity, 88% positive predictive value and 100% negative predictive value, including response to taxanes and anti-EGFR antibodies, drugs that lack biomarkers to predict responsive patient subsets.
Similar results were obtained for pancreatic ductal adenocarcinoma, another aggressive and difficult-to-treat tumour with high recurrence rates. Traditionally, treatment decisions are based on the performance status and comorbidities of the patient but there is an unmet medical need for patient stratification. To address this issue, 114 PDO cultures from 101 patients were generated127 and subjected to transcriptomic profiling, followed by therapeutic profiling through treatment with the five most commonly used chemotherapeutics, including gemcitabine, nab-paclitaxel, irinotecan, 5-fluorouracil and oxaliplatin. The treatment outcome of the PDOs positively correlated with individual patient responses and the assessment of targeted therapy sensitivity can guide patient-specific treatment decisions. These studies demonstrate that implementing PDOs in clinical decision-making can directly benefit patients with cancer.
Human disease models have also helped to identify treatment-enhancement strategies, for example, using cyclin-dependent kinase inhibitors to increase the effectiveness of immune-checkpoint inhibitors129. Despite being very effective, only a subset of patients respond to checkpoint blockade therapy and the underlying mechanisms for resistance development are poorly understood. To enhance the treatment efficacy, a combinatory approach with small-molecule kinase inhibitors has been proposed to facilitate immune reactivation. Interestingly, cyclin-dependent kinase inhibitors 4 and 6 (CDK4/6) facilitate T cell activation, resulting in a higher number of tumour-infiltrating T cells in PDOs, cultivated over several days in a microfluidic 3D chip setup. CDK4/6 inhibition also showed high in vitro and in vivo activity and synergy with anti-PD-1 blocking antibodies129. Several clinical trials (such as NCT04799249, NCT03294694 and NCT04213404) are currently assessing the clinical translatability of these findings.
Overall, ClinicalTrials.gov currently lists 131 studies for the search term ‘organoids’, most of which focus on cancer. Most of these studies have generated PDOs and compared drug effects in vitro and in vivo. Out of these, 10 studies are already using PDOs for clinical decision-making (Table 1). No results were retrieved for the search term ‘organ-on-chip’ or ‘microphysiological system’.Table 1 Clinical trials with patient-derived cancer organoids guiding treatment decisions
Organ system Cancer Identifiers Phase Country Drugs Implementation
Respiratory system Lung cancer, solid tumours NCT03778814 I China Biologics: TCR T cells Identification and engineering of tumour-responsive T cells using patient-specific tumour organoids followed by re-injection of TCR T cells into the patients
Gastrointestinal system Pancreatic cancer NCT04931394 III China Gemcitabine, 5-fluorouracil, paclitaxel, oxaliplatin, irinotecan PDOs of pancreatic cancer are tested for their sensitivity to first-line pancreatic cancer drugs; patients receive the chemotherapy regimen based on the test results
Advanced pancreatic cancer NCT04931381 III China Gemcitabine, 5-fluorouracil, paclitaxel, oxaliplatin, irinotecan PDOs of advanced pancreatic cancer are tested for their sensitivity to first-line pancreatic cancer drugs; patients receive the chemotherapy regimen based on the test results
Advanced rectal cancer NCT05352165 NA China Neoadjuvant therapy Clinical efficacy of personalized neoadjuvant therapy based on PDO chemosensitivity combined with standard long-term radiotherapy is compared with efficacy of standard whole-course neoadjuvant therapy
Abdominal tumours NCT05378048 II Hong Kong PDO-guided treatment using standard-of-care treatments A multidisciplinary tumour board reviews the drug screen results from PDOs and genome-guided drug screening and chooses the treatment regimen accordingly
Mammary glands Breast cancer NCT04450706 NA USA Docetaxel, cyclophosphamide, adriamycin, methotrexate, 5-fluorouracil, paclitaxel Treatment decisions are based on results from PDOs grown from breast cancer biopsies plus genome sequencing
Breast cancer NCT05177432 I Singapore 10–12 anticancer drugs (alpelisib, trastuzumab-emtansine and others not specified) PDOs are exposed to 10–12 anticancer drugs and a table for treatment sensitivity is obtained; results will be reviewed by an expert panel to decide on the most suitable anticancer drug treatment
Urinary system Bladder cancer NCT05024734 II Switzerland Epirubicin, mitomycin, gemcitabine, docetaxel Generation of PDOs and in vitro drug sensitivity testing to guide clinical decision-making
Others Head and neck squamous cell carcinoma NCT04279509 NA Singapore 5-Fluorouracil, carboplatin, cyclophosphamide, docetaxel, doxorubicin, gemcitabine, irinotecan, oxaliplatin, paclitaxel and vinorelbine, etoposide, ifosfamide, methotrexate, pemetrexed and topotecan Generation of PDOs followed by a 10-drug panel screening and selection of chemotherapy based on a standard rating scale; if more than one drug appears effective in PDOs, the most suitable drug based on patient comorbidities is selected
Solid tumours such as gastrointestinal and breast cancer NCT05381038 I and II Singapore Azacitidine plus docetaxel, azacitidine plus paclitaxel, azacitidine plus irinotecan Generation of PDOs followed by drug screening and selection, and artificial intelligence-guided dosing modulation
Clinical trials are as listed on ClinicalTrials.gov. NA, not applicable; PDO, patient-derived organoid; TCR, T cell receptor.
Future applications in clinical trials
Patient-specific disease models hold great potential for personalized and precision medicine in which therapeutic decisions or interventions are tailored to the individual based on their genetic makeup, disease or potential side effects. Such patient-stratified approaches could also facilitate clinical drug testing, for example, in phase I clinical trials, when the pharmacokinetic and drug safety profile are usually assessed in ≤100 healthy individuals. Women are often underrepresented, partly owing to the risk of reproductive toxicity. This is highly problematic as there are sex-related differences in how men and women absorb, distribute and metabolize drugs, ultimately affecting drug efficacy and safety48,130,131. Human-based models from both sexes and different ethnicities could help pinpoint these inter-individual differences, thereby improving therapeutic assessment. Moreover, human (disease) models can be leveraged to pre-screen patient subpopulations to identify those that could benefit most from a new drug.
Framework for disease model design
Defining universally applicable guidelines can be helpful to generate a predictive disease model. Although frameworks have been proposed for organoid and OoC setups24,54, providing a universal disease model design is challenging because their usefulness is strongly context dependent and influenced by the intended use and relevant readout parameters (Fig. 3). Furthermore, logistic requirements, such as access to relevant expertise, cell sources, equipment and cost, also need to be considered.Fig. 3 Schematic overview of a rational design of a disease model that incorporates different similarity criteria to ensure model scalability.
The flow chart shows the most important design considerations and workflows required for bioengineering human disease model. ECM, extracellular matrix.
When designing a disease model, the choice between model type (for example, 2D versus 3D or organoid versus bioengineered model), the selection of a suitable experimental setup (static versus dynamic) and the identification of important cell types are of utmost importance for the model to function. For drug screening, for example, 2D high-throughput methods using patient-specific or genetically engineered cells are most suitable as they allow testing of a large number of compounds in a short time. If the goal is to unravel disease mechanisms or study drug efficacy and safety, disease models with higher biomimicry and complexity are required. For example, when studying systemic conditions such as inflammation or metastatic cancer, complex multi-OoC setups are superior to single-tissue or organ models. Building on previously discussed examples, three key design considerations can be extracted, which are summarized in this section.
Cell diversity and tissue biomimicry
The identification of disease-relevant and tissue-relevant cell types is crucial, a prime example being the identification of the Zika virus tropism towards neural progenitor cells in brain organoids99. Equally important are the cultivation conditions, showcased by the decade-long failure in cultivating the norovirus ex vivo. Initial cultivation attempts, for example, in transformed epithelial cell lines and primary immune cells, failed largely because important factors of the host intestinal milieu were missing96. Furthermore, relevant biological barriers, such as the blood–brain barrier or the mucus hydrogel of the bronchial or intestinal epithelium, need to be included in the model. These are the primary defence mechanism of the human body and thereby very restrictive, especially in terms of the cut-off size of molecules that can pass. Failing to integrate them would ultimately limit the predictive value of disease models.
Model scaling
Scaling refers to maintaining proper ratios in size and rate parameters among different modules of a disease model, allowing the translation of in vitro data to clinical applications in vivo. Its importance has been recognized early on, but a general principle is still lacking132–134. Proper scaling needs to account for the geometry of the model, spatial arrangement of cells and operating conditions, the latter including medium dosage and perfusion rates of nutrients and metabolites, among other factors. There are two main principles of scaling: direct scaling, which uses the size ratio between a human body and the disease model to scale the size of each organ down to its in vitro counterpart55,135, and allometric scaling, which instead assumes that the size of organs and key physiological indices, such as heart, blood flow and metabolic rates, scale with the body mass according to power laws, with different exponents for different organs136,137. These two methods account for length and mass scales but not the time scale, and are often inaccurate. For example, the medium circulation rate is disproportionately slow compared to physiological blood flow, which in turn slows down the transport of nutrients and metabolic processes in tissue models138.
Therefore, dynamic, time-dependent operating conditions are more difficult to scale. Miniaturized models typically carry fewer cells and lack a pervasive vasculature to perfuse the tissue compared to actual organs. Thus, the perfusion, diffusion, permeation, nutrient and metabolic rates as well as their dosage need to be carefully calculated to ensure physiological relevance138. For this purpose, functional scaling136,138,139 seeks to match a key functional index between the model and the target organ in vivo. Such an index can be, for example, the drug clearance rate for the liver or the filtration rate for the kidney. However, it has proven difficult to balance competing key indices in multi-OoC setups; for a gut–liver system139, for example, a perfusion rate chosen for the gut to yield proper metabolic rates might result in drug exposure times in the liver that are too short.
To overcome these challenges, the similarity scaling approach140 adapts engineering techniques of dimensional analysis and similitude to OoCs. This strategy accounts for different similarity criteria (Box 1) simultaneously in a systematic framework. In any system, if an output is determined by a list of input variables and parameters, the latter must be algebraically linked in such a way as to yield the correct dimension or unit for the output. This constrains the algebraic relationship between the inputs and the output. Mathematically, the constraint is represented by combining the output and inputs into dimensionless groups141. Scaling an in vitro disease model with an in vivo organ boils down to matching the dimensionless groups between the two (Fig. 3). For an OoC, this matching could consist of proper ratios of the geometric lengths, morphologically correct arrangement of the cell types, and suitable ratios of mechanical and metabolic rates. These considerations ensure the correct translation of the key functional indices from the model to in vivo conditions139.
Box 1 Similarity criteria for scaling organs-on-chips Geometric similarity
Tissue geometry and curvature influence cell behaviour, tissue formation and function, and therefore the (patho)physiological relevance of in vitro models181. For example, the crypts of the gut and the sac-like structures of the alveoli provide a huge epithelial surface area, thereby defining the absorption rates of nutrients and gas. These features cannot be recapitulated 1:1 in vitro because it is difficult to control the formation, dimensions and shapes of tissue-relevant geometric structures, which often results in high model variability and limited biomimicry.
Morphological similarity
Organs-on-chips should include the same cell types with the correct proportion as real organs, such that the assembled layout contains specific cell types mixed together and others segregated into layers. For example, a liver-on-a-chip should use a physiological ratio of hepatocytes, liver sinusoidal endothelial cells, stellate cells and Kupffer cells, and layer them to mimic the morphology of the liver sinusoid66,182,183.
Kinematic similarity
All rate parameters, for example, medium perfusion rate, diffusion and permeation rates, and basal metabolic rates, should resemble the in vivo situation138.
Dynamic and metabolic similarity
All quantities related to forces, stresses and pressure should be in proper ratios as in vivo184. Mechanical readouts have often been the output of models; in disease models, however, the metabolism is instead mimicked to resemble pathophysiological conditions, especially for pharmacokinetic studies. In this case, the force or pressure is often an input or control parameter, for example, pumping the vascular system with physiological pressure and measuring its downstream effect. Furthermore, drug dosage, administration route and plasma concentration profiles should match those in the target organ. In toxicity studies, for example, this resemblance ensures that the pharmacokinetic and pharmacodynamic data are translatable to the target organs in vivo139,185.
Biomechanical cues
Biomechanical stimuli, such as shear stress, stretching or compression, influence (patho)physiological processes142–147. A sequential and physiological exposure to biomechanical cues promotes the differentiation and maturation of organoids and bioengineered tissue models. For example, intestinal stem cells respond to the stiffness of the surrounding matrix, which guides cell differentiation and organoid formation148. Similarly, exposure to mechanical forces that mimic cardiac preload and afterload improves the contractility, cell alignment and conduction velocity of bioengineered heart tissue149. Another example using an alveoli-on-chip model revealed in vivo-like pathological responses to IL-2 toxicity146 or nanoparticle inhalation143 only when breathing mechanics were applied. Similarly, engineered heart tissues generated from patient-derived cells that harbour a desmoplakin mutation develop clinical arrhythmogenic cardiomyopathy only when exposed to dynamic mechanical loading149. In vivo, mutations in the desmoplakin gene cause a specific type of cardiomyopathy, which is characterized by a thickening and stiffening of the heart muscle resulting in cardiac dysfunction. These examples highlight the importance of applying disease-relevant and tissue-relevant biomechanical cues in bioengineered models. This consideration directly feeds into the choice of model type as not all of them allow the integration of physical forces per se.
Challenges and limitations
Despite fast-paced advances, several challenges remain that limit the widespread application of bioengineered disease models.
Limited complexity
Animal models are the only preclinical systems that allow the study of disease mechanisms and drug effects in the complex environment of a living organism. This argument is also commonly used to justify their necessity in biomedical research. Although this statement is not wrong per se, it fails to answer questions such as how valuable it is to know that a drug candidate is safe and effective in an inbred mouse strain (given the high rate of translational failure) or what it means for humans that certain nanoparticles can target the kidneys of zebrafish. Although some animal models mimic human physiology and/or pathology better than others (for example, the lungs and skin of pigs are anatomically and physiologically similar to human lungs and skin, which is not the case for rodents), simply conducting experiments in a living organism might not hold the solution.
Nonetheless, similar unresolved questions can be asked when developing bioengineered human models, namely, how complex is complex enough? Or how simple can a model be and still remain predictive of human pathophysiology? Putative key components, such as cells or biomechanical cues relevant to the disease of interest, are included based on existing knowledge, yet the possibility to overlook and thus exclude disease-relevant contributors remains. This biased approach bears the risk of false-positive or false-negative results.
Limited lifespan and long cultivation
Although some OoC models have been maintained for up to 3 months55, the effective lifetime of most disease models spans over a few days. The limited regenerative capacity of primary cell-based models is a major limitation as is identifying a universal culture medium that can maintain tissues of different germ layers in multi-organ models. Media mixtures are typically used in these models, which, however, could reverse tissue maturation because cells from different germ layers are exposed to unspecific factors. This mismatch could result in gene and protein expression levels that are atypical for the tissue of interest, a feature that seems particularly relevant for stem cell-derived models. Generating tissue-specific niches by introducing native endothelial barriers in OoC setups might hold a solution for this problem; while these barriers keep the tissue models in their optimized environment by separating them from a common media circuit, inter-tissue communication is still possible through the exchange of, for example, cytokines or exosomes150.
Notably, organoids can be cultivated over months and probably even years, thus enabling the investigation of tissue maturation in a disease context as shown for SARS-CoV-2 infections in brain organoids87. However, even such a long lifespan might not be sufficient to study neural diseases that develop or progress slowly — for example, Alzheimer disease and Parkinson disease. In this case, the simplicity of bioengineered human disease models could prove advantageous because disease-relevant effects might appear more rapidly owing to the lack of compensatory mechanisms that typically exist in vivo that can compensate for or mask tissue dysfunction.
Moreover, broad clinical implementation of patient-specific disease models, such as PDOs, is currently hindered by low efficiencies in PDO establishment (60–70% success rate on average although, for some tissues, up to 90% have been reported for intestinal organoids151) as well as lengthy generation times and expansion procedures. The establishment and expansion of PDOs and subsequent drug testing currently span several weeks or even months, although technical solutions to shorten the process are emerging. For example, microfabricated array devices using organoids at passage 0 provide drug screening results within a week without compromising the predictivity of patient responses to anticancer drugs152.
Similarly, protocols for cell differentiation into disease models can span several weeks, especially for multicellular tissue models. To accelerate and guide stem cell differentiation, the overexpression of transcription factors relevant to the differentiation of the target tissue (for example, ETV2 for endothelial cells or NGN1 for neurons) can switch on rapid cell differentiation. Subsequent mixing or controlled spatial patterning through, for example, 3D printing of iPS cells pre-programmed for doxycycline-induced overexpression of such cell type-specific transcription factors, ultimately facilitates the differentiation into multicellular and spatially patterned organoids independently from extracellular differentiation cues present in the culture medium153. This strategy could substantially expedite the generation of multicellular disease models.
Immunoregulation
Emulating the fine-tuned and highly interdependent immunoregulation in the human body remains a main limitation of in vitro human disease models. The human immune system is incredibly complex, orchestrating a multitude of immune cell (sub-)types of different functionality, many poorly understood and others probably still unrecognized. Although many disease models already contain certain immune cell types143,145,154–156, most models rely on the integration of only a few, pre-defined immune cell types, which, however, does not emulate the complexity of in vivo immunoregulation. For example, in the gut–liver–brain OoC setup for modelling Parkinson disease, only regulatory and T helper 17 cells were used103. Similarly, to study pathogen–host interactions in human disease models, only T cells and in some cases B cells have often been included113. By contrast, peripheral blood mononuclear cells have been increasingly added in OoC setups157. Bioengineered secondary immune organs, such as lymphoid tissue models158–160, might help overcome this limitation; however, this is still an underdeveloped area.
Disease model validation
A key task for the bioengineering community is to demonstrate the equivalence or superiority of human disease models over their animal-based counterparts. Therefore, rigorous disease model validation is essential but often neglected. The problem is further aggravated by variabilities in model composition, complexity and experimental protocols. First, the model needs to capture relevant features of the disease through, for example, histological similarity, in vivo-like gene regulation and protein expression patterns, or known drug responses. Similarly, the assays and readout parameters need to be relevant to the disease of interest, and the model and experimental setup should minimize noise and bias. Finally, it needs to be determined if the data generated by the model can be extrapolated for clinical applications. Establishing these criteria and ensuring inter-laboratory reproducibility needs to be standardized. Automated procedures ranging from liquid handling and cell seeding to sampling and sample analysis can improve the reproducibility, controllability and, thus, robustness of the models.
Omics technologies
Single-cell and spatial profiling at transcriptional, proteomic and epigenetic levels (Box 2) can substantially improve disease model validation161,162. For example, scRNA-seq helped identify and mitigate shortcomings related to the high variability of cell composition of in vitro models, exemplified in cerebral organoids generated from individuals with autism spectrum disorder163. By using an integrated scRNA and bulk RNA-seq approach, genes that showed low inter-individual variability, and thus high inter-individual correlation, allowed pinpointing of genes of relevance for autism spectrum disorder. Furthermore, combining scRNA-seq with ATAC-seq164 (Box 2) can help determine the degree of biomimicry of (disease) models by identifying cell type-specific chromatin accessibility patterns as well as transcription factors and cascades that are pivotal for cell fate specification and topographic identity as shown for brain and retinal organoids165,166. Using omics technologies, differences in gene regulation and signalling pathways between organoids and the corresponding primary tissue samples can be identified.
Box 2 Omics techniques for model validation Genomic profiling
Assessing tissue-specific and disease-specific gene signatures is key for (disease) model validation. Initial information can be gained through bulk RNA sequencing (RNA-seq), which measures average gene expression levels across all cell types in different cell states. Higher resolution readouts can be obtained through single-cell RNA-seq, which enables whole-transcriptome profiling of individual cells, identification of rare cell populations and detection of cell type-specific driver genes for disease development. Spatial profiling provides further information on tissue organization and three-dimensional architecture, thereby being essential for model validation186,187. Despite being powerful and scalable, these methods are still expensive and require trained bioinformatics personnel. As the technology advances, these tools will likely become more accessible.
Protein profiling
Because proteins are the most common drug targets today, disease model validation at the protein level is crucial. Mass spectrometry-based proteomics measures protein abundance and dynamics, identifies subcellular protein localization and post-transcriptional modifications, and facilitates the generation of interaction networks162. Proteomics can thus help unravel disease mechanisms and tissue responses to certain stimuli. Metabolic labelling188, chemical tags189, isobaric labelling190 and label-free quantification191 can also help protein quantification by generating small differences in peptide mass and have already been used to characterize organoids.
Similar to bulk RNA-seq, proteomic analysis at the bulk level provides a population average and obscures cellular heterogeneities. Unlike spatial profiling on genomic data, single-cell proteomics is still in its infancy; measurements are time consuming, expensive and have a limited dynamic range of measurements, rendering it low throughput. Another limiting factor is the complexity of data analysis of single-cell proteomics, which has limited the routine application of this technology for model validation so far. However, few groups have already managed to quantify more than 1,000 proteins per single cell192 while measuring ≥70 cells in parallel193. Mass spectrometry imaging is another exciting technology that can resolve the spatial distribution of proteins and has been used in tumour organoids194; however, it is not yet widely accessible to the scientific community.
Imaging and data-driven approaches
The complex architecture of disease models, their characterization and the investigation of cell-specific pathological reactions require advanced 3D imaging technologies, such as light-sheet fluorescence microscopy or multiphoton imaging167. Unlike 2D imaging, these technologies allow high-resolution 3D whole-mount imaging of disease models at different scales, including cellular compositions, cell shapes, cell–cell interactions and cell fate167. The superiority of light-sheet fluorescence microscopy over conventional line-scan imaging has already been demonstrated for 3D organoids168. However, the huge amount of data generated, especially for high-throughput screenings of PDOs, has called for an increasing use of artificial intelligence-driven data analysis algorithms such as machine learning169. For example, tracking morphological and textural changes of PDOs to different drugs using bright field images enables the generation of PDO-specific dose–response curves170. Similarly, using a neural network-based high-throughput approach for light microscopy-based screening demonstrates strong correlations between clinical and predicted PDO response across different tumour entities171. Similarly, integration of network modelling and perturbation analysis with pathological disease features of 1,300 patient-derived brain organoids provided a high-content system for drug screening172. These combinatorial approaches could help identify effective drugs more reliably, although their actual impact has yet to be proven.
When aiming for routine and standardized applications, following the ‘digital twins’ concept can be useful. A digital twin is an in silico method that uses real-world data to predict how a product or process will perform. Examples of their successful application can be found in space missions or aeroplane construction173. In bioengineering, the implementation of digital twins could facilitate the shift from exploration towards a patient-focused and manufacturing-focused, and therefore more standardized, approach. For example, digital twin-based mathematical models should accurately describe real-world phenomena, starting with a simple structure and known effects (such as cell kinetics, stationary and fluid flow characteristics, and distribution of components such as growth factors) and gradually adapting it to newly available data and observations (for example, by including mass transfer and shear effects). Although still in their infancy, digital twins are being explored for tissue models and have enormous potential to be used for other disease models174,175.
Commercialization and accessibility
Developing and maintaining bioengineered disease models require high-level expertise and costly infrastructures. Even static 3D tissue models are substantially more expensive and technically demanding than conventional 2D biomedical methods. Moreover, the notion that in vitro methods are cheaper is misleading because bioengineered disease models can be as if not more expensive than animal models. As the demand for human-based models increases, more companies are emerging to fill the gap; however, commercially available setups are often still very expensive and therefore not affordable for many laboratories. Moreover, these models mainly focus on preclinical safety testing and not disease modelling and are not customizable169. For example, all commercially available OoC setups are run with pre-set chips, which determines a priori the type and dimension of tissues that can be integrated. Some OoC setups have the size of a coin (or even smaller), thereby making it difficult to cultivate actual 3D tissue models. Moreover, as of now, only few disease models are commercially available.
Outlook
Bioengineered disease models are revolutionizing biomedical research and will increasingly replace animal models in basic and preclinical research. Their implementation in preclinical or clinical stages could accelerate the drug development process, reduce false-positive and false-negative results, and facilitate the clinical translation of findings from bench to bedside. These benefits could also substantially cut down drug development costs, with an estimated 10–26% cost reduction per newly approved drug2. It is therefore not surprising that pharmaceutical companies have considerably ramped up their investments in this area.
Nonetheless, animal models are still considered the gold standard and therefore are often requested by default by grant review panels or reviewers. In our view, this is benchmarking against the wrong standard. Ideally, patient-derived in vivo data should be used as a benchmark to corroborate findings from human disease models, especially during model development and validation. Therefore, models should be continuously re-assessed to ensure that they comply with the most recent and relevant scientific findings.
Animal testing is not going to be completely replaced in the near future for several reasons, one of them being that toxicity testing in animals is legally mandated before entering clinical trials. Therefore, the most realistic short-term and mid-term scenario is the complementary use of both human-based and animal models. Nonetheless, drugs can already enter clinical trials without providing animal-derived data if the in vitro data is compelling, no suitable animal models are available and a considerable medical benefit is expected. Moreover, the FDA Modernization Act 2.0 has further broadened the scope of accepted preclinical models and encourages scientists to test drug efficacy and safety in human models whenever possible. This landmark decision is expected to cause ripple effects worldwide.
Notably, different disease models are suitable for different applications; organoids, for example, are useful for high-throughput drug screenings, whereas OoCs are more applicable for drug safety and efficacy testing. On the one hand, personalized disease models are well suited for clinical decision-making and in clinical trials to diversify the patient cohorts with respect to ethnicity, sex or age. On the other hand, they are poorly suited for high-throughput screenings. To ensure a widespread preclinical and clinical application of human disease models, meeting the following key milestones is essential: definition of robust criteria for disease model validation, benchmarking against patient-derived in vivo data, regulatory approval and legal basis for the implementation of human disease models in the drug development process, and the establishment of scalable, robust and standardized manufacturing processes (Fig. 3).
For this purpose, the experience from the in vitro pro-arrhythmia assay (CiPA) initiative, launched in 2013 to improve the assessment of the pro-arrhythmic potential of new drugs, can prove useful. Members of the CiPA initiative include regulatory authorities such as the FDA or EMA, pharmaceutical companies, and academics. The initiative has defined sub-working groups that collaborate to reach pre-defined milestones and could serve as a role model for the widespread implementation of bioengineered models in (pre-)clinical research.
The formation of dedicated core facilities would also make human disease models more accessible and promote their usage. In this regard, a survey amongst scientists that do not use OoC setups revealed that the lack of ready-to-use-systems and production facilities as well as high entry barriers and costs are the main reasons for not employing OoCs169. Therefore, facilities that support researchers with hands-on training, ready-to-use tissue models or scientific advice would promote the widespread usage of human disease models, the Hubrecht Organoid Technology organization in Utrecht, The Netherlands, being one example. Similarly, ‘living biobanks’ can provide access to patient-derived cells and organoids from large patient cohorts, and have already been established for different cancers124,151,176–178, genetic diseases such as cystic fibrosis, infectious diseases, chronic obstructive pulmonary disease and inflammatory bowel diseases179.
It is unrealistic to assume that the complexity of a human can be fully recapitulated in vitro. At the same time, the ideal should be to reach maximal clinical mimicry. Although several challenges remain unsolved, initial drawbacks, such as the lack of cell diversity, vascularization and the ability to study tissue crosstalk, have already been largely overcome, highlighting the potential of human disease models to fundamentally change the future of biomedical research.
Acknowledgements
We acknowledge partial funding of this work from the Stiftung Charité (S.H.), Einstein Foundation EC3R (A.L.), the Natural Sciences and Engineering Research Council Canada (NSERC, RGPIN-2020–04224 for S.H. and RGPIN-2019–04162 for J.J.F.), and the German Research Foundation (DFG)-funded Collaborative Research Centers 1449 (S.H., project A05) and 1340 (S.H., project A06).
Author contributions
S.H., A.L. and J.J.F. contributed to the conceptualization, writing, figure drafting and revision of the article.
Peer review
Peer review information
Nature Reviews Bioengineering thanks Wei Zheng, who co-reviewed with Qi Zhang; Sina Bartfeld; Gordana Vunjak-Novakovic, who co-reviewed with Diogo Teles; and the other, anonymous, reviewer for their contribution to the peer review of this work.
Competing interests
The authors state no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Change history
6/13/2023
A Correction to this paper has been published: 10.1038/s44222-023-00088-8
==== Refs
References
1. Adhikary PP Ul Ain Q Hocke AC Hedtrich S COVID-19 highlights the model dilemma in biomedical research Nat. Rev. Mater. 2021 6 374 376 10.1038/s41578-021-00305-z 33747552
2. Franzen N Impact of organ-on-a-chip technology on pharmaceutical R&D costs Drug Discov. Today 2019 24 1720 1724 10.1016/j.drudis.2019.06.003 31185290
3. IQVIA Institute. Global Trends in R&D 2022 (IQVIA Institute, 2021).
4. Golding H Khurana S Zaitseva M What is the predictive value of animal models for vaccine efficacy in humans? The importance of bridging studies and species-independent correlates of protection Cold Spring Harb. Perspect. Biol. 2018 10 a028902 10.1101/cshperspect.a028902 28348035
5. Franco R Cedazo-Minguez A Successful therapies for Alzheimer’s disease: why so many in animal models and none in humans? Front. Pharmacol. 2014 5 146 10.3389/fphar.2014.00146 25009496
6. Ingber DE Human organs-on-chips for disease modelling, drug development and personalized medicine Nat. Rev. Genet. 2022 23 467 491 10.1038/s41576-022-00466-9 35338360
7. Brennan FR Safety testing of monoclonal antibodies in non-human primates: case studies highlighting their impact on human risk assessment mAbs 2018 10 1 17 10.1080/19420862.2017.1389364 28991509
8. Rothgangl T In vivo adenine base editing of PCSK9 in macaques reduces LDL cholesterol levels Nat. Biotechnol. 2021 39 949 957 10.1038/s41587-021-00933-4 34012094
9. Muñoz-Fontela C Animal models for COVID-19 Nature 2020 586 509 515 10.1038/s41586-020-2787-6 32967005
10. Brooks IR Functional genomics and the future of iPSCs in disease modeling Stem Cell Rep. 2022 17 1033 1047 10.1016/j.stemcr.2022.03.019
11. Jensen C Teng Y Is it time to start transitioning from 2D to 3D cell culture? Front. Mol. Biosci. 2020 7 33 10.3389/fmolb.2020.00033 32211418
12. Htwe SS Investigating NF-κB signaling in lung fibroblasts in 2D and 3D culture systems Respir. Res. 2015 16 144 10.1186/s12931-015-0302-7 26619903
13. DesRochers TM Suter L Roth A Kaplan DL Bioengineered 3D human kidney tissue, a platform for the determination of nephrotoxicity PLoS One 2013 8 e59219 10.1371/journal.pone.0059219 23516613
14. Vepari C Kaplan DL Silk as a biomaterial Prog. Polym. Sci. 2007 32 991 1007 10.1016/j.progpolymsci.2007.05.013 19543442
15. Pati F Cho DW Bioprinting of 3D tissue models using decellularized extracellular matrix bioink Methods Mol. Biol. 2017 1612 381 390 10.1007/978-1-4939-7021-6_27 28634957
16. Stoltz JF Zhang L Ye JS De Isla N Organ reconstruction: dream or reality for the future Biomed. Mater. Eng. 2017 28 S121 S127 28372287
17. Hassanpour A Talaei-Khozani T Kargar-Abarghouei E Razban V Vojdani Z Decellularized human ovarian scaffold based on a sodium lauryl ester sulfate (SLES)-treated protocol, as a natural three-dimensional scaffold for construction of bioengineered ovaries Stem Cell Res. Ther. 2018 9 252 10.1186/s13287-018-0971-5 30257706
18. Debels H Hamdi M Abberton K Morrison W Dermal matrices and bioengineered skin substitutes: a critical review of current options Plast. Reconstr. Surg. Global Open 2015 3 e284 10.1097/GOX.0000000000000219
19. Caddeo S Boffito M Sartori S Tissue engineering approaches in the design of healthy and pathological in vitro tissue models Front. Bioeng. Biotechnol. 2017 5 40 10.3389/fbioe.2017.00040 28798911
20. Upadhyay S Palmberg L Air-liquid interface: relevant in vitro models for investigating air pollutant-induced pulmonary toxicity Toxicol. Sci. 2018 164 21 30 10.1093/toxsci/kfy053 29534242
21. Moniz T Costa Lima SA Reis S Human skin models: from healthy to disease-mimetic systems; characteristics and applications Br. J. Pharmacol. 2020 177 4314 4329 10.1111/bph.15184 32608012
22. Khetani SR Bhatia SN Microscale culture of human liver cells for drug development Nat. Biotechnol. 2008 26 120 126 10.1038/nbt1361 18026090
23. Cantley W Functional and sustainable 3D human neural network models from pluripotent stem cells ACS Biomater. Sci. Eng. 2018 4 4278 4288 10.1021/acsbiomaterials.8b00622 33304995
24. Hofer M Lutolf MP Engineering organoids Nat. Rev. Mater. 2021 6 402 420 10.1038/s41578-021-00279-y 33623712
25. Du Y Development of a miniaturized 3D organoid culture platform for ultra-high-throughput screening J. Mol. Cell Biol. 2020 12 630 643 10.1093/jmcb/mjaa036 32678871
26. Vonk AM Protocol for application, standardization and validation of the forskolin-induced swelling assay in cystic fibrosis human colon organoids STAR Protoc. 2020 1 100019 10.1016/j.xpro.2020.100019 33111074
27. Spence JR Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro Nature 2011 470 105 109 10.1038/nature09691 21151107
28. McCracken KW Modelling human development and disease in pluripotent stem-cell-derived gastric organoids Nature 2014 516 400 404 10.1038/nature13863 25363776
29. Takebe T Vascularized and functional human liver from an iPSC-derived organ bud transplant Nature 2013 499 481 484 10.1038/nature12271 23823721
30. Huang L Ductal pancreatic cancer modeling and drug screening using human pluripotent stem cell–and patient-derived tumor organoids Nat. Med. 2015 21 1364 1371 10.1038/nm.3973 26501191
31. Wong AP Directed differentiation of human pluripotent stem cells into mature airway epithelia expressing functional CFTR protein Nat. Biotechnol. 2012 30 876 882 10.1038/nbt.2328 22922672
32. Kurmann AA Regeneration of thyroid function by transplantation of differentiated pluripotent stem cells Cell Stem Cell 2015 17 527 542 10.1016/j.stem.2015.09.004 26593959
33. Taguchi A Redefining the in vivo origin of metanephric nephron progenitors enables generation of complex kidney structures from pluripotent stem cells Cell Stem Cell 2014 14 53 67 10.1016/j.stem.2013.11.010 24332837
34. Phillips MJ Blood-derived human iPS cells generate optic vesicle-like structures with the capacity to form retinal laminae and develop synapses Invest. Ophthalmol. Vis. Sci. 2012 53 2007 2019 10.1167/iovs.11-9313 22410558
35. Lancaster MA Cerebral organoids model human brain development and microcephaly Nature 2013 501 373 379 10.1038/nature12517 23995685
36. Mansour AA An in vivo model of functional and vascularized human brain organoids Nat. Biotechnol. 2018 36 432 441 10.1038/nbt.4127 29658944
37. Revah O Maturation and circuit integration of transplanted human cortical organoids Nature 2022 610 319 326 10.1038/s41586-022-05277-w 36224417
38. Sachs N Long-term expanding human airway organoids for disease modeling EMBO J. 2019 38 e100300 10.15252/embj.2018100300 30643021
39. Sato T Long-term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett’s epithelium Gastroenterology 2011 141 1762 1772 10.1053/j.gastro.2011.07.050 21889923
40. Boj SF Organoid models of human and mouse ductal pancreatic cancer Cell 2015 160 324 338 10.1016/j.cell.2014.12.021 25557080
41. Bartfeld S In vitro expansion of human gastric epithelial stem cells and their responses to bacterial infection Gastroenterology 2015 148 126 136.e6 10.1053/j.gastro.2014.09.042 25307862
42. Huch M Long-term culture of genome-stable bipotent stem cells from adult human liver Cell 2015 160 299 312 10.1016/j.cell.2014.11.050 25533785
43. Turco MY Long-term, hormone-responsive organoid cultures of human endometrium in a chemically defined medium Nat. Cell Biol. 2017 19 568 577 10.1038/ncb3516 28394884
44. Drost J Organoid culture systems for prostate epithelial and cancer tissue Nat. Protoc. 2016 11 347 358 10.1038/nprot.2016.006 26797458
45. Kessler M The Notch and Wnt pathways regulate stemness and differentiation in human fallopian tube organoids Nat. Commun. 2015 6 8989 10.1038/ncomms9989 26643275
46. Linnemann JR Quantification of regenerative potential in primary human mammary epithelial cells Development 2015 142 3239 3251 26071498
47. Chichagova V Human iPSC differentiation to retinal organoids in response to IGF1 and BMP4 activation is line- and method-dependent Stem Cell 2020 38 195 201 10.1002/stem.3116
48. Lock R A framework for developing sex-specific engineered heart models Nat. Rev. Mater. 2021 7 295 313 10.1038/s41578-021-00381-1 34691764
49. Lai Benjamin FL Recapitulating pancreatic tumor microenvironment through synergistic use of patient organoids and organ-on-a-chip vasculature Adv. Funct. Mater. 2020 30 2000545 10.1002/adfm.202000545 33692660
50. Yadid M Endothelial extracellular vesicles contain protective proteins and rescue ischemia-reperfusion injury in a human heart-on-chip Sci. Transl. Med. 2020 12 eaax8005 10.1126/scitranslmed.aax8005 33055246
51. Si L A human-airway-on-a-chip for the rapid identification of candidate antiviral therapeutics and prophylactics Nat. Biomed. Eng. 2021 5 815 829 10.1038/s41551-021-00718-9 33941899
52. Low LA Mummery C Berridge BR Austin CP Tagle DA Organs-on-chips: into the next decade Nat. Rev. Drug Discov. 2021 20 345 361 10.1038/s41573-020-0079-3 32913334
53. Vunjak-Novakovic G Ronaldson-Bouchard K Radisic M Organs-on-a-chip models for biological research Cell 2021 184 4597 4611 10.1016/j.cell.2021.08.005 34478657
54. Leung CM A guide to the organ-on-a-chip Nat. Rev. Methods Primers 2022 2 33 10.1038/s43586-022-00118-6
55. Maschmeyer I A four-organ-chip for interconnected long-term co-culture of human intestine, liver, skin and kidney equivalents Lab Chip 2015 15 2688 2699 10.1039/C5LC00392J 25996126
56. Lai BFL A well plate-based multiplexed platform for incorporation of organoids into an organ-on-a-chip system with a perfusable vasculature Nat. Protoc. 2021 16 2158 2189 10.1038/s41596-020-00490-1 33790475
57. Chou DB On-chip recapitulation of clinical bone marrow toxicities and patient-specific pathophysiology Nat. Biomed. Eng. 2020 4 394 406 10.1038/s41551-019-0495-z 31988457
58. Edington CD Interconnected microphysiological systems for quantitative biology and pharmacology studies Sci. Rep. 2018 8 4530 10.1038/s41598-018-22749-0 29540740
59. Herland A Quantitative prediction of human pharmacokinetic responses to drugs via fluidically coupled vascularized organ chips Nat. Biomed. Eng. 2020 4 421 436 10.1038/s41551-019-0498-9 31988459
60. Novak R Robotic fluidic coupling and interrogation of multiple vascularized organ chips Nat. Biomed. Eng. 2020 4 407 420 10.1038/s41551-019-0497-x 31988458
61. Marx U Biology-inspired microphysiological system approaches to solve the prediction dilemma of substance testing Altex 2016 33 272 321 27180100
62. Matsui T Shinozawa T Human organoids for predictive toxicology research and drug development Front. Genet. 2021 12 767621 10.3389/fgene.2021.767621 34790228
63. Li M Advanced human developmental toxicity and teratogenicity assessment using human organoid models Ecotoxicol. Environ. Saf. 2022 235 113429 10.1016/j.ecoenv.2022.113429 35325609
64. Cong Y Drug toxicity evaluation based on organ-on-a-chip technology: a review Micromachines 2020 11 381 10.3390/mi11040381 32260191
65. Nawroth JC Modeling alcohol-associated liver disease in a human liver-chip Cell Rep. 2021 36 109393 10.1016/j.celrep.2021.109393 34289365
66. Sphabmixay P High resolution stereolithography fabrication of perfusable scaffolds to enable long-term meso-scale hepatic culture for disease modeling Biofabrication 2021 13 ac23aa 10.1088/1758-5090/ac23aa
67. Bauer S Functional coupling of human pancreatic islets and liver spheroids on-a-chip: towards a novel human ex vivo type 2 diabetes model Sci. Rep. 2017 7 14620 10.1038/s41598-017-14815-w 29097671
68. Wimmer RA Human blood vessel organoids as a model of diabetic vasculopathy Nature 2019 565 505 510 10.1038/s41586-018-0858-8 30651639
69. Jabali A Human cerebral organoids reveal progenitor pathology in EML1-linked cortical malformation EMBO Rep 2022 23 e54027 10.15252/embr.202154027 35289477
70. Vatine GD Human iPSC-derived blood-brain barrier chips enable disease modeling and personalized medicine applications Cell Stem Cell 2019 24 995 1005.e6 10.1016/j.stem.2019.05.011 31173718
71. Hönzke S Influence of Th2 cytokines on the cornified envelope, tight junction proteins, and ß-defensins in filaggrin-deficient skin equivalents J. Investig. Dermatol. 2016 136 631 639 10.1016/j.jid.2015.11.007 27015451
72. Zamprogno P Second-generation lung-on-a-chip with an array of stretchable alveoli made with a biological membrane Commun. Biol. 2021 4 168 10.1038/s42003-021-01695-0 33547387
73. van der Vaart J Modelling of primary ciliary dyskinesia using patient-derived airway organoids EMBO Rep. 2021 22 e52058 10.15252/embr.202052058 34693619
74. Charbaji R Design and testing of efficient mucus-penetrating nanogels — pitfalls of preclinical testing and leassons learned Small 2021 10.1002/smll.202007963
75. Kayisoglu O Location-specific cell identity rather than exposure to GI microbiota defines many innate immune signalling cascades in the gut epithelium Gut 2021 70 687 697 10.1136/gutjnl-2019-319919 32571970
76. Lõhmussaar K Patient-derived organoids model cervical tissue dynamics and viral oncogenesis in cervical cancer Cell Stem Cell 2021 28 1380 1396.e6 10.1016/j.stem.2021.03.012 33852917
77. Vila OF Bioengineered optogenetic model of human neuromuscular junction Biomaterials 2021 276 121033 10.1016/j.biomaterials.2021.121033 34403849
78. Williams K A 3-D human model of complex cardiac arrhythmias Acta Biomater. 2021 132 149 161 10.1016/j.actbio.2021.03.004 33713861
79. Mastikhina O Human cardiac fibrosis-on-a-chip model recapitulates disease hallmarks and can serve as a platform for drug testing Biomaterials 2020 233 119741 10.1016/j.biomaterials.2019.119741 31927251
80. Wang EY An organ-on-a-chip model for pre-clinical drug evaluation in progressive non-genetic cardiomyopathy J. Mol. Cell Cardiol. 2021 160 97 110 10.1016/j.yjmcc.2021.06.012 34216608
81. Ronaldson-Bouchard K Engineering of human cardiac muscle electromechanically matured to an adult-like phenotype Nat. Protoc. 2019 14 2781 2817 10.1038/s41596-019-0189-8 31492957
82. Lu RXZ Vasculature-on-a-chip platform with innate immunity enables identification of angiopoietin-1 derived peptide as a therapeutic for SARS-CoV-2 induced inflammation Lab Chip 2022 22 1171 1186 10.1039/D1LC00817J 35142777
83. Zhang M Biomimetic human disease model of SARS-CoV-2 induced lung injury and immune responses on organ chip system Adv. Sci. 2020 8 2002928 10.1002/advs.202002928
84. Katsura H Human lung stem cell-based alveolospheres provide insights into SARS-CoV-2-mediated interferon responses and pneumocyte dysfunction Cell Stem Cell 2020 27 890 904.e8 10.1016/j.stem.2020.10.005 33128895
85. Lamers MM An organoid-derived bronchioalveolar model for SARS-CoV-2 infection of human alveolar type II-like cells EMBO J. 2021 40 e105912 10.15252/embj.2020105912 33283287
86. Thacker VV Rapid endotheliitis and vascular damage characterize SARS-CoV-2 infection in a human lung-on-chip model EMBO Rep. 2021 22 e52744 10.15252/embr.202152744 33908688
87. Ramani A SARS-CoV-2 targets neurons of 3D human brain organoids EMBO J. 2020 39 e106230 10.15252/embj.2020106230 32876341
88. Andrews MG Tropism of SARS-CoV-2 for human cortical astrocytes Proc. Natl Acad. Sci. USA 2022 119 e2122236119 10.1073/pnas.2122236119 35858406
89. Jansen J SARS-CoV-2 infects the human kidney and drives fibrosis in kidney organoids Cell Stem Cell 2022 29 217 231.e8 10.1016/j.stem.2021.12.010 35032430
90. Helms L Cross-validation of SARS-CoV-2 responses in kidney organoids and clinical populations JCI Insight 2021 6 e154882 10.1172/jci.insight.154882 34767537
91. Ortega-Prieto AM 3D microfluidic liver cultures as a physiological preclinical tool for hepatitis B virus infection Nat. Commun. 2018 9 682 10.1038/s41467-018-02969-8 29445209
92. Nawroth JC A microengineered airway lung chip models key features of viral-induced exacerbation of asthma Am. J. Respir. Cell Mol. Biol. 2020 63 591 600 10.1165/rcmb.2020-0010MA 32706623
93. Koster S Modelling Chlamydia and HPV co-infection in patient-derived ectocervix organoids reveals distinct cellular reprogramming Nat. Commun. 2022 13 1030 10.1038/s41467-022-28569-1 35210413
94. Liberti DC Alveolar epithelial cell fate is maintained in a spatially restricted manner to promote lung regeneration after acute injury Cell Rep. 2021 35 109092 10.1016/j.celrep.2021.109092 33979629
95. Bui CHT Tropism of influenza B viruses in human respiratory tract explants and airway organoids Eur. Respir. J. 2019 54 1900008 10.1183/13993003.00008-2019 31097520
96. Ettayebi K Replication of human noroviruses in stem cell-derived human enteroids Science 2016 353 1387 1393 10.1126/science.aaf5211 27562956
97. Hayashi T Dasabuvir inhibits human norovirus infection in human intestinal enteroids mSphere 2021 6 e0062321 10.1128/mSphere.00623-21 34730374
98. Alvarado G Broadly cross-reactive human antibodies that inhibit genogroup I and II noroviruses Nat. Commun. 2021 12 4320 10.1038/s41467-021-24649-w 34262046
99. Qian X Brain-region-specific organoids using mini-bioreactors for modeling ZIKV exposure Cell 2016 165 1238 1254 10.1016/j.cell.2016.04.032 27118425
100. Garcez PP Zika virus impairs growth in human neurospheres and brain organoids Science 2016 352 816 818 10.1126/science.aaf6116 27064148
101. Pleguezuelos-Manzano C Mutational signature in colorectal cancer caused by genotoxic pks+ E. coli Nature 2020 580 269 273 10.1038/s41586-020-2080-8 32106218
102. Nougayrède JP A toxic friend: genotoxic and mutagenic activity of the probiotic strain Escherichia coli Nissle 1917 mSphere 2021 6 e0062421 10.1128/mSphere.00624-21 34378987
103. Trapecar M Human physiomimetic model integrating microphysiological systems of the gut, liver, and brain for studies of neurodegenerative diseases Sci. Adv. 2021 7 eabd1707 10.1126/sciadv.abd1707 33514545
104. Kim J Fungal brain infection modelled in a human-neurovascular-unit-on-a-chip with a functional blood-brain barrier Nat. Biomed. Eng. 2021 5 830 846 10.1038/s41551-021-00743-8 34127820
105. Swalley SE Expanding therapeutic opportunities for neurodegenerative diseases: a perspective on the important role of phenotypic screening Bioorg Med. Chem. 2020 28 115239 10.1016/j.bmc.2019.115239 31889605
106. Naryshkin NA SMN2 splicing modifiers improve motor function and longevity in mice with spinal muscular atrophy Science 2014 345 688 693 10.1126/science.1250127 25104390
107. Palacino J SMN2 splice modulators enhance U1–pre-mRNA association and rescue SMA mice Nat. Chem. Biol. 2015 11 511 517 10.1038/nchembio.1837 26030728
108. Theodoris CV Network-based screen in iPSC-derived cells reveals therapeutic candidate for heart valve disease Science 2021 371 eabd0724 10.1126/science.abd0724 33303684
109. Monteil V Inhibition of SARS-CoV-2 infections in engineered human tissues using clinical-grade soluble human ACE2 Cell 2020 181 905 913.e7 10.1016/j.cell.2020.04.004 32333836
110. Bai H Mechanical control of innate immune responses against viral infection revealed in a human lung alveolus chip Nat. Commun. 2021 13 1928 10.1038/s41467-022-29562-4
111. Han Y Identification of SARS-CoV-2 inhibitors using lung and colonic organoids Nature 2021 589 270 275 10.1038/s41586-020-2901-9 33116299
112. Rahmani W Attenuation of SARS-CoV-2 infection by losartan in human kidney organoids iScience 2022 25 103818 10.1016/j.isci.2022.103818 35106453
113. Bai H Mechanical control of innate immune responses against viral infection revealed in a human lung alveolus chip Nat. Commun. 2022 13 1928 10.1038/s41467-022-29562-4 35396513
114. Herpers B Functional patient-derived organoid screenings identify MCLA-158 as a therapeutic EGFR × LGR5 bispecific antibody with efficacy in epithelial tumors Nat. Cancer 2022 3 418 436 10.1038/s43018-022-00359-0 35469014
115. Yun J Antitumor activity of Amivantamab (JNJ-61186372), an EGFR-MET bispecific antibody, in diverse models of EGFR exon 20 insertion-driven NSCLC Cancer Discov. 2020 10 1194 1209 10.1158/2159-8290.CD-20-0116 32414908
116. Rumsey JW Classical complement pathway inhibition in a “human-on-a-chip” model of autoimmune demyelinating neuropathies Adv. Ther. 2022 5 2200030 10.1002/adtp.202200030
117. Fiedorczuk K Chen J Molecular structures reveal synergistic rescue of Δ508 CFTR by Trikafta modulators Science 2022 378 284 290 10.1126/science.ade2216 36264792
118. Dekkers JF Characterizing responses to CFTR-modulating drugs using rectal organoids derived from subjects with cystic fibrosis Sci. Transl. Med. 2016 8 344ra384 10.1126/scitranslmed.aad8278
119. Berkers G Rectal organoids enable personalized treatment of cystic fibrosis Cell Rep. 2019 26 1701 1708.e3 10.1016/j.celrep.2019.01.068 30759382
120. Ooft SN Patient-derived organoids can predict response to chemotherapy in metastatic colorectal cancer patients Sci. Transl. Med. 2019 11 eaay2574 10.1126/scitranslmed.aay2574 31597751
121. Schütte M Molecular dissection of colorectal cancer in pre-clinical models identifies biomarkers predicting sensitivity to EGFR inhibitors Nat. Commun. 2017 8 14267 10.1038/ncomms14262 28117394
122. Shi R Organoid cultures as preclinical models of non-small cell lung cancer Clin. Cancer Res. 2020 26 1162 1174 10.1158/1078-0432.CCR-19-1376 31694835
123. Vlachogiannis G Patient-derived organoids model treatment response of metastatic gastrointestinal cancers Science 2018 359 920 926 10.1126/science.aao2774 29472484
124. Lee SH Tumor evolution and drug response in patient-derived organoid models of bladder cancer Cell 2018 173 515 528.e17 10.1016/j.cell.2018.03.017 29625057
125. Hill SJ Prediction of DNA repair inhibitor response in short-term patient-derived ovarian cancer organoids Cancer Discov. 2018 8 1404 1421 10.1158/2159-8290.CD-18-0474 30213835
126. Yao Y Patient-derived organoids predict chemoradiation responses of locally advanced rectal cancer Cell Stem Cell 2020 26 17 26.e6 10.1016/j.stem.2019.10.010 31761724
127. Tiriac H Organoid profiling identifies common responders to chemotherapy in pancreatic cancer Cancer Discov. 2018 8 1112 1129 10.1158/2159-8290.CD-18-0349 29853643
128. Driehuis E Pancreatic cancer organoids recapitulate disease and allow personalized drug screening Proc. Natl Acad. Sci. USA 2019 116 26580 26590 10.1073/pnas.1911273116 31818951
129. Deng J CDK4/6 inhibition augments antitumor immunity by enhancing T-cell activation Cancer Discov. 2018 8 216 233 10.1158/2159-8290.CD-17-0915 29101163
130. Klein SL Flanagan KL Sex differences in immune responses Nat. Rev. Immunol. 2016 16 626 638 10.1038/nri.2016.90 27546235
131. Zeng Z Sex-hormone-driven innate antibodies protect females and infants against EPEC infection Nat. Immunol. 2018 19 1100 1111 10.1038/s41590-018-0211-2 30250184
132. Ronaldson-Bouchard K Vunjak-Novakovic G Organs-on-a-chip: a fast track for engineered human tissues in drug development Cell Stem Cell 2018 22 310 324 10.1016/j.stem.2018.02.011 29499151
133. Park D Lee J Chung JJ Jung Y Kim SH Integrating organs-on-chips: multiplexing, scaling, vascularization, and innervation Trends Biotechnol. 2020 38 99 112 10.1016/j.tibtech.2019.06.006 31345572
134. Malik M Yang Y Fathi P Mahler GJ Esch MB Critical considerations for the design of multi-organ microphysiological systems (MPS) Front. Cell Dev. Biol. 2021 9 721338 10.3389/fcell.2021.721338 34568333
135. Wagner I A dynamic multi-organ-chip for long-term cultivation and substance testing proven by 3D human liver and skin tissue co-culture Lab Chip 2013 13 3538 3547 10.1039/c3lc50234a 23648632
136. Wikswo JP Scaling and systems biology for integrating multiple organs-on-a-chip Lab Chip 2013 13 3496 3511 10.1039/c3lc50243k 23828456
137. Ahluwalia A Allometric scaling in-vitro Sci. Rep. 2017 7 42113 10.1038/srep42113 28169362
138. Sung JH Wang Y Shuler ML Strategies for using mathematical modeling approaches to design and interpret multi-organ microphysiological systems (MPS) APL Bioeng. 2019 3 021501 10.1063/1.5097675 31263796
139. Maass C Stokes CL Griffith LG Cirit M Multi-functional scaling methodology for translational pharmacokinetic and pharmacodynamic applications using integrated microphysiological systems (MPS) Integr. Biol. 2017 9 290 302 10.1039/C6IB00243A
140. Feng JJ Hedtrich S A similarity scaling approach for organ-on-chip devices Lab Chip 2022 22 3663 3667 10.1039/D2LC00641C 36070239
141. Zlokarnik, M. Scale-Up in Chemical Engineering 2nd edn (Wiley‐VCH Verlag GmbH & Co. KGaA, 2006).
142. Frye M Matrix stiffness controls lymphatic vessel formation through regulation of a GATA2-dependent transcriptional program Nat. Commun. 2018 9 1511 10.1038/s41467-018-03959-6 29666442
143. Huh D Reconstituting organ-level lung functions on a chip Science 2010 328 1662 1668 10.1126/science.1188302 20576885
144. Sontheimer-Phelps A Human colon-on-a-chip enables continuous in vitro analysis of colon mucus layer accumulation and physiology Cell Mol. Gastroenterol. Hepatol. 2020 9 507 526 10.1016/j.jcmgh.2019.11.008 31778828
145. Bein A Enteric coronavirus infection and treatment modeled with an immunocompetent human intestine-on-a-chip Front. Pharmacol. 2021 12 718484 10.3389/fphar.2021.718484 34759819
146. Huh D A human disease model of drug toxicity-induced pulmonary edema in a lung-on-a-chip microdevice Sci. Transl. Med. 2012 4 159ra147 10.1126/scitranslmed.3004249
147. Grassart A Bioengineered human organ-on-chip reveals intestinal microenvironment and mechanical forces impacting shigella infection Cell Host Microbe 2019 26 435 444.e4 10.1016/j.chom.2019.08.007 31492657
148. Gjorevski N Designer matrices for intestinal stem cell and organoid culture Nature 2016 539 560 564 10.1038/nature20168 27851739
149. Bliley JM Dynamic loading of human engineered heart tissue enhances contractile function and drives a desmosome-linked disease phenotype Sci. Transl. Med. 2021 13 eabd1817 10.1126/scitranslmed.abd1817 34290054
150. Ronaldson-Bouchard K A multi-organ chip with matured tissue niches linked by vascular flow Nat. Biomed. Eng. 2022 6 351 371 10.1038/s41551-022-00882-6 35478225
151. van de Wetering M Prospective derivation of a living organoid biobank of colorectal cancer patients Cell 2015 161 933 945 10.1016/j.cell.2015.03.053 25957691
152. Hu Y Lung cancer organoids analyzed on microwell arrays predict drug responses of patients within a week Nat. Commun. 2021 12 2581 10.1038/s41467-021-22676-1 33972544
153. Skylar-Scott MA Orthogonally induced differentiation of stem cells for the programmatic patterning of vascularized organoids and bioprinted tissues Nat. Biomed. Eng. 2022 6 449 462 10.1038/s41551-022-00856-8 35332307
154. Wallmeyer L TSLP is a direct trigger for T cell migration in filaggrin-deficient skin equivalents Sci. Rep. 2017 7 774 10.1038/s41598-017-00670-2 28377574
155. Maharjan S Cecen B Zhang YS 3D immunocompetent organ-on-a-chip models Small Methods 2020 4 2000235 10.1002/smtd.202000235 33072861
156. Kerns SJ Human immunocompetent Organ-on-Chip platforms allow safety profiling of tumor-targeted T-cell bispecific antibodies eLife 2021 10 e67106 10.7554/eLife.67106 34378534
157. Cipriano M Human immunocompetent choroid-on-chip: a novel tool for studying ocular effects of biological drugs Commun. Biol. 2022 5 52 10.1038/s42003-021-02977-3 35027657
158. Goyal G Ectopic lymphoid follicle formation and human seasonal influenza vaccination responses recapitulated in an organ-on-a-chip Adv. Sci. 2022 9 e2103241 10.1002/advs.202103241
159. Gibot L Cell-based approach for 3D reconstruction of lymphatic capillaries in vitro reveals distinct functions of HGF and VEGF-C in lymphangiogenesis Biomaterials 2016 78 129 139 10.1016/j.biomaterials.2015.11.027 26694987
160. Gibot L Tissue-engineered 3D human lymphatic microvascular network for in vitro studies of lymphangiogenesis Nat. Protoc. 2017 12 1077 1088 10.1038/nprot.2017.025 28448484
161. Li X Wang C-Y From bulk, single-cell to spatial RNA sequencing Int. J. Oral. Sci. 2021 13 36 10.1038/s41368-021-00146-0 34782601
162. Kip AM Proteomics analysis of human intestinal organoids during hypoxia and reoxygenation as a model to study ischemia-reperfusion injury Cell Death Dis. 2021 12 95 10.1038/s41419-020-03379-9 33462215
163. Lim ET Orgo-Seq integrates single-cell and bulk transcriptomic data to identify cell type specific-driver genes associated with autism spectrum disorder Nat. Commun. 2022 13 3243 10.1038/s41467-022-30968-3 35688811
164. Baek S Lee I Single-cell ATAC sequencing analysis: from data preprocessing to hypothesis generation Comput. Struct. Biotechnol. J. 2020 18 1429 1439 10.1016/j.csbj.2020.06.012 32637041
165. Finkbeiner C Single-cell ATAC-seq of fetal human retina and stem-cell-derived retinal organoids shows changing chromatin landscapes during cell fate acquisition Cell Rep. 2022 38 110294 10.1016/j.celrep.2021.110294 35081356
166. Ziffra RS Single-cell epigenomics reveals mechanisms of human cortical development Nature 2021 598 205 213 10.1038/s41586-021-03209-8 34616060
167. Fei K Zhang J Yuan J Xiao P Present application and perspectives of organoid imaging technology Bioengineering 2022 9 121 10.3390/bioengineering9030121 35324810
168. Lallemant L Lebreton C Garfa-Traoré M Comparison of different clearing and acquisition methods for 3D imaging of murine intestinal organoids J. Biol. Methods 2020 7 e141 10.14440/jbm.2020.334 33564693
169. Busek M Aizenshtadt A Amirola-Martinez M Delon L Krauss S Academic user view: organ-on-a-chip technology Biosensors 2022 12 126 10.3390/bios12020126 35200386
170. Spiller ER Imaging-based machine learning analysis of patient-derived tumor organoid drug response Front. Oncol. 2021 11 771173 10.3389/fonc.2021.771173 34993134
171. Larsen BM A pan-cancer organoid platform for precision medicine Cell Rep. 2021 36 109429 10.1016/j.celrep.2021.109429 34320344
172. Park J-C A logical network-based drug-screening platform for Alzheimer’s disease representing pathological features of human brain organoids Nat. Commun. 2021 12 280 10.1038/s41467-020-20440-5 33436582
173. Barricelli BR Casiraghi E Fogli D A survey on digital twin: definitions, characteristics, applications, and design implications IEEE Access. 2019 7 167653 167671 10.1109/ACCESS.2019.2953499
174. Möller J Pörtner R Digital twins for tissue culture techniques — concepts, expectations, and state of the art Processes 2021 9 447 10.3390/pr9030447
175. Fuhr A Kurtz A Hiepen C Müller S Organoids as miniature twins — challenges for comparability and need for data standardization and access Organoids 2022 1 28 36 10.3390/organoids1010003
176. Yan HHN A comprehensive human gastric cancer organoid biobank captures tumor subtype heterogeneity and enables therapeutic screening Cell Stem Cell 2018 23 882 897.e11 10.1016/j.stem.2018.09.016 30344100
177. Sachs N A living biobank of breast cancer organoids captures disease heterogeneity Cell 2018 172 373 386.e10 10.1016/j.cell.2017.11.010 29224780
178. Jacob F A patient-derived glioblastoma organoid model and biobank recapitulates inter- and intra-tumoral heterogeneity Cell 2020 180 188 204.e22 10.1016/j.cell.2019.11.036 31883794
179. Ramalho AS Correction of CFTR function in intestinal organoids to guide treatment of cystic fibrosis Eur. Respir. J. 2021 57 1902426 10.1183/13993003.02426-2019 32747394
180. Martin L Hutchens M Hawkins C Radnov A How much do clinical trials cost? Nat. Rev. Drug Discov. 2017 16 381 382 10.1038/nrd.2017.70 28529317
181. Werner M Surface curvature differentially regulates stem cell migration and differentiation via altered attachment morphology and nuclear deformation Adv. Sci. 2017 4 1600347 10.1002/advs.201600347
182. Prodanov L Long-term maintenance of a microfluidic 3D human liver sinusoid Biotechnol. Bioeng. 2016 113 241 246 10.1002/bit.25700 26152452
183. Peel S Introducing an automated high content confocal imaging approach for organs-on-chips Lab Chip 2019 19 410 421 10.1039/C8LC00829A 30663729
184. Abaci HE Shuler ML Human-on-a-chip design strategies and principles for physiologically based pharmacokinetics/pharmacodynamics modeling Integr. Biol. 2015 7 383 391 10.1039/C4IB00292J
185. Tsamandouras N Integrated gut and liver microphysiological systems for quantitative in vitro pharmacokinetic studies AAPS J. 2017 19 1499 1512 10.1208/s12248-017-0122-4 28752430
186. Fleck JS Resolving organoid brain region identities by mapping single-cell genomic data to reference atlases Cell Stem Cell 2021 28 1148 1159.e8 10.1016/j.stem.2021.02.015 33711282
187. Rodriques SG Slide-seq: a scalable technology for measuring genome-wide expression at high spatial resolution Science 2019 363 1463 1467 10.1126/science.aaw1219 30923225
188. Gonneaud A A SILAC-based method for quantitative proteomic analysis of intestinal organoids Sci. Rep. 2016 6 38195 10.1038/srep38195 27901089
189. Cristobal A Personalized proteome profiles of healthy and tumor human colon organoids reveal both individual diversity and basic features of colorectal cancer Cell Rep. 2017 18 263 274 10.1016/j.celrep.2016.12.016 28052255
190. Williams KE Quantitative proteomic analyses of mammary organoids reveals distinct signatures after exposure to environmental chemicals Proc. Natl Acad. Sci. USA 2016 113 E1343 E1351 10.1073/pnas.1600645113 26903627
191. Melliou S Regionally defined proteomic profiles of human cerebral tissue and organoids reveal conserved molecular modules of neurodevelopment Cell Rep. 2022 39 110846 10.1016/j.celrep.2022.110846 35613588
192. Cong Y Ultrasensitive single-cell proteomics workflow identifies >1000 protein groups per mammalian cell Chem. Sci. 2021 12 1001 1006 10.1039/D0SC03636F
193. Williams SM Automated coupling of nanodroplet sample preparation with liquid chromatography–mass spectrometry for high-throughput single-cell proteomics Anal. Chem. 2020 92 10588 10596 10.1021/acs.analchem.0c01551 32639140
194. Bakker B Preparing ductal epithelial organoids for high-spatial-resolution molecular profiling using mass spectrometry imaging Nat. Protoc. 2022 17 962 979 10.1038/s41596-021-00661-8 35181767
|
PMC010xxxxxx/PMC10174333.txt |
==== Front
Curr Probl Cardiol
Curr Probl Cardiol
Current Problems in Cardiology
0146-2806
1535-6280
Elsevier
S0146-2806(23)00215-3
10.1016/j.cpcardiol.2023.101798
101798
Article
Racial Disparities in Mortality Associated With Acute Myocardial Infarction and COVID-19 in the United States: A Nationwide Analysis
Muhyieddeen Amer MD a
Cheng Susan MD a
Mamas Mamas A BM, BCh c
Beasley Dorian MD d
Weins Galen Cook MS b
Gulati Martha MD, MS a⁎
a Barbra Streisand Women's Heart Center, Smidt Heart Institute, Cedars-Sinai Medical Center Los Angeles, CA
b Bioststatistics and Bioinformatics Research Center, Cedars-Sinai Medical Center, Los Angeles, CA
c Keele Cardiac Research Group, Centre for Prognosis Research, Keele University, Stoke-on-Trent, UK
d Community Physicians Network, Indianapolis, Indiana
⁎ Correspondence author: Martha Gulati, MD, MS, 127 S. San Vicente Blvd, Suite A3600, Los Angeles, CA 90048, Tel.: 310-423-9680, Fax: 310-423-9681.
11 5 2023
9 2023
11 5 2023
48 9 101798101798
© 2023 The Author(s)
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
This study assessed the COVID-19 pandemic's impact on racial disparities in acute myocardial infarction (AMI) management and outcomes. We reviewed AMI patient management and outcomes in the pandemic's initial nine months, comparing COVID-19 and non-COVID-19 cases using 2020's National Inpatient Sample data. Our findings revealed that patients with concurrent AMI and COVID-19 had higher in-hospital mortality (aOR 3.19, 95% CI 2.63-3.88), increased mechanical ventilation (aOR 1.90, 95% CI 1.54-2.33), and higher initiation of hemodialysis (aOR 1.38, 95% CI 1.05-1.89) compared to those without COVID-19. Moreover, Black and Asian/Pacific Islander patients had higher in-hospital mortality than White patients, (aOR 2.13, 95% CI 1.35-3.59; aOR 3.41, 95% CI 1.5-8.37). Also, Black, Hispanic, and Asian/Pacific Islander patients showed higher odds of initiating hemodialysis (aOR 5.48, 95% CI 2.13-14.1; aOR 2.99, 95% CI 1.13-7.97; aOR 7.84, 95% CI 1.55-39.5), and were less likely to receive PCI for AMI (aOR 0.71, 95% CI 0.67-0.74; aOR 0.81, 95% CI 0.77-0.86; aOR 0.82, 95% CI 0.75-0.90). Black patients also showed less likelihood of undergoing CABG (aOR 0.55, 95% CI 0.49-0.61). Our study highlights elevated mortality and complications in COVID-19 AMI patients, emphasizing significant racial disparities. These findings underscore the pressing need for initiatives addressing healthcare disparities, enhancing access, and promoting culturally sensitive care to boost health equity.
==== Body
pmcIntroduction
The COVID-19 pandemic significantly strained healthcare systems worldwide once it was declared a pandemic by the World Health Organization in March 2020. As of January 5, 2023, over 638 million people had been infected, resulting in more than 6.6 million deaths.1 Notably, an increase in mortality rates related to cardiovascular disease has been observed during the pandemic.2 , 3 The underlying causes of this trend remain uncertain, but potential factors include delayed or deferred care and varying treatment approaches due to hospital capacity constraints.4
Several studies comparing acute myocardial infarction (AMI) management and outcomes before and during the pandemic revealed longer symptom-to-balloon times, decreased adherence to medically-guided therapy, and increased mortality rates.5, 6, 7, 8 Recent data also underscores the guarded outcomes of patients hospitalized for AMI and COVID-19.9, 10, 11 In a study analyzing clinical, procedural, and in-hospital prognostic factors for patients with COVID-19 admitted with a diagnosis of ST-elevation myocardial infarction (STEMI), higher rates of stent thrombosis, cardiogenic shock, and in-hospital mortality were reported in patients infected with COVID-19 compared to non-COVID-19 STEMI patients.12
Beyond the direct effects of COVID-19 on patients with cardiovascular disease, the pandemic has led healthcare and research organizations to shift resources away from non-COVID-related issues, severely impacting patients with cardiovascular conditions.13 This has impacted certain groups in society more than others, including socio-economically disadvantaged patient groups, and Black and other minority racial groups.14, 15, 16 Moreover, a recent study conducted in the United Kingdom revealed that during the COVID-19 pandemic, minority groups experienced higher hospitalization rates due to AMI.17 However, current literature lacks extensive research on potential disparities in AMI treatment or increased mortality among minority racial groups in the United States during the pandemic. To address this knowledge gap, our study utilized the National Inpatient Sample (NIS) to compare clinical outcomes in patients diagnosed with AMI, both with and without COVID-19 infection, and to investigate potential racial disparities in treatment and outcomes.
Methods
The analysis was conducted using the NIS database for the year 2020. The NIS is part of the healthcare cost and utilization project (HCUP) databases and is sponsored by the agency for healthcare research and quality (AHRQ). It contains clinical and resource utilization information on millions of discharges annually, with precautions to safeguard the privacy of individual patients and hospitals. The data is stratified to represent 20% of U.S. inpatient hospitalizations across different hospitals and geographic areas as a random sample. The NIS database allows for calculating national estimates by providing a weight variable.18 For 2020, the unweighted sample included 6.3 million observations, and the weighted sample was around 31.7 million discharges. All patients admitted to the hospital with acute myocardial infarction (STEMI or non-ST-elevation myocardial infarction [NSTEMI]), and concomitant COVID-19 infection were included in this study. Because our study used deidentified data, it was exempt from Institutional Review Board approval.
To identify patients admitted with AMI, the NIS database was searched using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes (I20.0, I21.1, I21.2, I21.3, and I21.4). In the present analysis, a total of 16,465 cases were excluded on account of elective admissions. Furthermore, 85,174 cases were removed from the dataset to avoid duplicate counting, as these patients were transferred out of the hospital. Exclusions were also made for cases with missing variables, including insurance status, race, sex, death status, and age. The missing cases constituted less than 0.8% (3528/446,834) of the initial dataset (Fig 1 ).FIG 1 Consort flow diagram depicting inclusion and exclusion criteria for study.
FIG 1
The cohort of patients with AMI was stratified into two subgroups based on the presence or absence of COVID-19, as identified by the ICD-10 code (U07.1). The primary endpoint of the study was in-hospital mortality. Secondary outcomes encompassed revascularization rates using percutaneous coronary intervention (PCI), coronary artery bypass grafting (CABG), and thrombolytic therapy, acute kidney injury necessitating hemodialysis, administration of in-hospital vasopressors, utilization of mechanical ventilation and mechanical circulatory support, as well as the duration of hospital stay.
Statistical analysis was conducted using STATA 17 (StataCorp LLC, College Station, TX). Categorical variables were compared using chi-square tests, while linear regression was employed for continuous variables. To assess associations and adjust for potential confounders, logistic and linear regression were utilized. Candidate variables underwent testing for univariable associations, with those exhibiting a P-value <0.2 being incorporated into the final multivariable model. The Elixhauser Comorbidity Index was employed to account for comorbid conditions.19We defined significance as a 2-tailed P value of 0.05.
Results
During 2020, there were a total of 446,834 hospitalizations for AMI, and 4645 were COVID-positive (1.04%). Patients with AMI and COVID-19 were younger than those without COVID-19 (63.9 years vs 65.4 years, p<0.01). Patients with AMI and COVID-19 were more likely to be from minority groups, with a greater proportion of Black (17.2% vs 9.2%), Hispanic (22.4% vs 8.9%), and Native American patients (1.7% vs 0.51%); (P < 0.01). They were also more likely to have a household income of less than $50,000 (38.1% vs 30.7%, P = <0.01), and to have Medicaid insurance (13.4% vs 10.5%, P < 0.01).
Patients with AMI and COVID-19 had a higher prevalence of type II diabetes (41.7% vs 31.7%, P < 0.01), and ischemic stroke (2.3% vs 0.9%, P < 0.01) but were less likely to have a history of coronary artery disease (81.2% vs 86.2%, P = < 0.01), hyperlipidemia (59.9% vs 68.1%, P < 0.01), and tobacco use (35.9% vs 53.4%, P < 0.001), compared to those with AMI without COVID-19 (Table 1 ).TABLE 1 Baseline characteristics of all acute myocardial infarctions based on presence or absence of COVID-19
TABLE 1Characteristics AMI+ COVID-19 (N = 4645) AMI+ non-COVID-19 (N = 446,834) P-value
(Female) 34.6% (N = 1607) 34.9% (N = 154,604) 0.88
Mean age 63.9 65.4 <0.001
RACE:
White 55.6% (N = 2597) 78.3% (N = 339,147) <0.001
Black 17.2% (N = 878) 9.2% (N = 51,385)
Hispanic 22.4% (N = 948) 8.9% (N = 40,661)
Asian or Pacific Islander 3.2% (N = 144) 3.2% (N = 12,958)
Native American 1.7% (N = 79) 0.51% (N = 2,859)
Median household income:
$1-$49,999 38.1% (N = 1769) 30.7% (N = 137,178) <0.001
$50,000-$64,999 24.8% (N = 1152) 28.6% (N = 127,794)
$65,000-85,999 20.3% (N = 943) 22.6% (N = 100,984)
>$86,000 16.7% (N = 776) 18.1% (N = 80,876)
Insurance status:
Medicare 46.1% (N = 2141) 51.8% (N = 231,460) 0.001
Medicaid 13.4% (N = 662) 10.5% (N = 46,917)
Private 28.7% (N = 1333) 28.4% (N = 126,900)
Self-pay 7.1% (N = 330) 5.5% (N = 24,576)
No charge 0.9% (N = 42) 0.5% (N = 2234)
Other 3.9% (N = 181) 3.3% (N = 14,746)
Hospital region:
Northeast 15.6% (N = 725) 16.1% (N = 71,940) 0.88
Midwest 22.9% (N = 1064) 22.5% (N = 100,538)
South 41.4% (N = 1937) 42.3% (N = 189,011)
West 20.0% (N = 929) 19.1% (N = 85,345)
Hospital ownership:
Government, nonfederal 10.9% (N = 506) 9.0% (N = 40,215) 0.08
Private, nonprofit 74.8% (N = 3474) 75.0% (N = 335,125)
Private, investor own 14.2% (N = 660) 15.9% (N = 71,046)
Hospital bedsize:
Small 20.3% (N = 929) 18.2% (N = 81,323) 0.19
Medium 27.5% (N = 1277) 29.6% (N = 132,263)
Large 52.2% (N = 2425) 51.1% (N = 228,332)
Hospital teaching status:
Rural nonteaching 6.0% (N = 279) 6.6% (N = 29,491) 0.25
Urban nonteaching 16.3% (N = 757) 18.1% (N = 80,876)
Urban teaching 77.7% (N = 3609) 75.3% (N = 336,466)
Elixhauser comorbidy index:
0 1.8% (N = 84) 2.9% (N = 12,958) 0.17
1 8.9% (N = 413) 9.8% (N = 43,789)
2 16.2% (N = 753) 15.7% (N = 70,152)
3 17.3% (N = 804) 17.3% (N = 77,302)
>4 55.8% (N = 2592) 54.2% (N = 242,184)
Comorbidities: 446,834
Atrial fibrillation 18.4% (N = 855) 17.5% (N = 78,196) 0.47
Congested heart failure 33.3% (N = 1533) 33.1% (N = 147,902) 0.51
Coronary artery disease 77.3% (N = 3591) 84.5% (N = 377,574) <0.001
hypertension 39.9% (N = 1839) 42.3% (N = 189,011) 0.15
Chronic kidney disease 19.5% (N = 906) 17.3% (N = 77,302) 0.08
Diabetes mellitus 45.8% (N = 2127) 38.8% (N = 169,796) <0.001
Hyperlipidemia 65.6% (N = 3019) 71.6% (N = 319,933) <0.001
Liver disease 2.1% (N = 98) 2.5% (N = 11,171) 0.80
Lung disease 10.7% (N = 497) 15.6% (N = 69,706) <0.001
Ischemic stroke 1.2% (N = 56) 0.87% (N = 3887) 0.22
Pulmonary HTN 3.6% (N = 167) 4.8% (N = 21,448) 0.07
Tobacco use 38.0% (N = 1765) 51.5% (N =230,120) <0.001
Obesity 18.5% (N = 859) 18.4% (N = 82,217) 0.98
Unadjusted clinical outcomes showed that patients with both AMI and COVID-19 had significantly higher in-hospital mortality rates (14.7% vs 5.7%, P < 0.01), increased use of mechanical ventilation (12.8% vs 7.4%, P < 0.01), more frequent vasopressor administration (3.6% vs 2.2%, P < 0.01), and a greater need for hemodialysis (4.7% vs 3.1%, P < 0.001) compared to patients with AMI alone. Additionally, their hospital stays were significantly longer, averaging 5.18 days compared to 3.83 days for those with only AMI (P < 0.01). Revascularization also varied between the two groups. Patients with both AMI and COVID-19 had lower rates of PCI (39.8% vs 45.8%, P < 0.001), combined PCI or thrombolytic usage (40.5% vs 46.1%, P < 0.01), and CABG surgery (3.3% vs 7.9%, P < 0.001). However, they had higher rates of thrombolytic usage (10.8% vs 5.5%, P = 0.032) compared to patients with AMI but no COVID-19 (Table 2 and Supplemental Fig 1).TABLE 2 Unadjusted primary and secondary outcomes of patients with AMI+ COVID-19 vs AMI without COVID-19
TABLE 2 AMI + COVID-19 AMI + non-COVID-19 P-value
In-hospital mortality 14.7% 5.7% <0.001
Mechanical ventilation 12.8% 7.4% <0.001
Vasopressor use 3.6% 2.2% <0.01
Mechanical circulatory Support 1.2% 1.3% 0.89
Hemodialysis 4.7% 3.1% <0.01
PCI 39.8% 45.8% <0.001
Thrombolytics 10.8% 5.5% 0.032
PCI or thrombolytics 40.5% 46.1% <0.01
CABG 3.3% 7.9% <0.001
Mean length of stay (days) 5.18 3.83 <0.001
Total charges (USD) 110,163 109,798 0.44
After adjusting for confounders, patients with AMI and COVID-19 had 3.19 (95% CI, 2.63-3.88) times greater odds of in-hospital mortality compared to patients with AMI but no COVID-19. Patients with AMI and COVID-19 were also more likely to require mechanical ventilation (aOR 1.90, 95% CI, 1.54-2.33), vasopressor use (aOR 1.60, 95% CI, 1.11-2.33), and hemodialysis initiation (aOR 1.38, 95% CI, 1.05-1.89) than patients with AMI without COVID-19. Patients with AMI and COVID-19 were also less likely to receive PCI (aOR 0.78, 95% CI, 0.67 - 0.91), PCI or thrombolytics (aOR 0.80, 95% CI, 0.69-0.93), and CABG surgery (aOR 0.40, 95% CI, 0.28-0.59). However, they were more likely to receive thrombolytics (aOR 2.05, 95% CI, 1.08-3.92) compared to patients with AMI without COVID-19 (Table 3 , Fig 2 ).TABLE 3 Adjusted odds ratio outcomes for patients with COVID-19 and AMI relative to patients with AMI and no COVID-19 infection, using multivariate analysis
TABLE 3 Adjusted odds ratio P-value
In-hospital mortality 3.19 (95% CI, 2.63 - 3.88) < 0.01
Mechanical ventilation 1.90 (95% CI, 1.54 - 2.33) < 0.01
Vasopressor use 1.60 (95% CI, 1.11 - 2.23) < 0.01
Mechanical circulatory support 0.96 (95% CI, 0.53-1.75) 0.65
Hemodialysis 1.38 (95% CI, 1.05-1.89) 0.046
PCI 0.78 (95% CI, 0.67-0.91) < 0.01
Thrombolytics 2.05 (95% CI, 1.08-3.92) 0.028
PCI or thrombolytics 0.80 (95% CI, 0.69-0.93) 0.003
CABG 0.40 (95% CI, 0.28-0.59) < 0.001
FIG 2 Outcomes, treatments and revasculization odds for patients with COVID-19 and AMI relative to patients with AMI and no COVID-19 infection, using multivariate analysis. 1Adjusted for age, race, sex, hospital bed size, hospital location, hospital teaching status, insurance status, income level, and Elixhauser comorbidities. AMI, acute myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting.
FIG 2
The in-hospital mortality rates for patients experiencing both AMI and COVID-19 showed variation across different racial groups. An analysis of the unadjusted clinical outcomes for racial differences was conducted. Compared to White patients with AMI and COVID-19; Black, Asian/Pacific Islander, and Native American patients experienced higher in-hospital mortality rates (P < 0.05) and increased hemodialysis (P < 0.001). Furthermore, Black, Hispanic, and Asian/Pacific Islander patients had lower rates of PCI (P < 0.05) and lower rates of combined PCI or thrombolytics (P < 0.05). (Supplemental Table 1, Supplemental Fig 2).
Upon adjusting for confounders, Black and Asian/Pacific Islander patients exhibited higher odds of in-hospital mortality in comparison to White patients, (aOR 2.13, 95% CI, 1.35-3.59) and (aOR 3.41, 95% CI, 1.5-8.37), respectively. Additionally, Black, Hispanic, and Asian/Pacific Islander patients showed higher odds of initiating hemodialysis, with odds ratios of (aOR 5.48, 95% CI, 2.13-14.1), (aOR 2.99, 95% CI, 1.13-7.97) and (aOR 7.84, 95% CI, 1.55-39.5), respectively.
With regards to revascularization and after adjustment for cofounders, these patient groups still had significantly lower odds of receiving PCI compared to White patients: Black (aOR 0.71, 95% CI 0.67-0.74), Hispanic (aOR 0.81, 95% CI, 0.77-0.86), and Asian/Pacific Islander (aOR 0.82, 95% CI, 0.75-0.90). Moreover, Black patients had lower odds of undergoing CABG surgery compared to White patients (aOR 0.55, 95% CI, 0.49-0.61) (Supplemental Table 2 and Supplemental Fig 3).
A subgroup analysis was performed to examine revascularization rates among patients diagnosed with STEMI. The baseline demographic characteristics for this subgroup can be found in (Supplemental Table 3). In patients with both STEMI and COVID-19, the unadjusted rate of undergoing PCI was lower (61.2% vs 68.9%, P < 0.01), the rate of receiving thrombolytics was higher (2.6% vs 0.9%, P < 0.01), and the rate of receiving either PCI or thrombolytics was lower (61.2% vs 67.9%, P < 0.01) compared to patients with STEMI without COVID-19. After adjusting for confounding variables, patients with both STEMI and COVID-19 exhibited lower odds of undergoing PCI (aOR 0.73, 95% CI, 0.58-0.91) and higher odds of receiving thrombolytic therapy (aOR 3.23, 95% CI, 1.69-6.14). Meanwhile, the odds of receiving either PCI or thrombolytic therapy were lower (aOR 0.77, 95% CI, 0.62-0.96) when compared to patients diagnosed with STEMI without COVID-19 (Table 4 and Fig 3 ).TABLE 4 Rate of revascularization with PCI, thrombolytics and CABG for STEMI + COVID-19 positive vs COVID-19 negative
TABLE 4 COVID + COVID - P-value
PCI 61.2% 68.9% < 0.01
Adjusted odds ratio* 0.73 (95% CI, 0.58- 0.91) < 0.01
Thrombolytics 2.6% 0.9% < 0.01
Adjusted odds ratio* 3.23 (95% CI, 1.69- 6.14) < 0.01
PCI or thrombolytics 62.7 69.3% < 0.01
Adjusted odds ratio* 0.77 (95% CI, 0.62-0.96) 0.02
CABG 2.3% 4.2% 0.06
Adjusted odds ratio* 0.55 (95% CI, 0.29-1.02) 0.09
Crude and Adjusted mortality rate both presented.
⁎ Adjusted for age, race, sex, hospital bed size, hospital location, hospital teaching status, insurance status, income level, and Elixhauser comorbidities.
FIG 3 Rate of revascularization with PCI, thrombolytics and CABG for STEMI + COVID-19 positive vs COVID-19 negative. Crude and Adjusted mortality rate both presented. Left hand side (A) showing crude rates of revascularization. Right hand side (B) showing adjusted rates of revascularization. 1Adjusted for age, race, sex, hospital bed size, hospital location, hospital teaching status, insurance status, income level, and elixhauser comorbidities. AMI, acute myocardial infarction; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting.
FIG 3
Rates of revascularization were furthermore analyzed by race. The unadjusted rates of revascularization revealed that Black, Hispanic, and Asian/ Pacific Islanders had lower rates of PCI (P < 0.05) and lower rates of combined PCI or thrombolytics (P < 0.05) (Supplemental Table 4). After accounting for confounding factors, Black, and Asian/Pacific Islander patients exhibited significantly lower odds of receiving PCI compared to White patients, (aOR 0.83, 95% CI, 0.58-0.90), and (aOR 0.78, 95% CI, 0.66-0.90), respectively. Additionally, Black patients had lower odds of undergoing CABG surgery relative to White patients (aOR 0.68, 95% CI, 0.53-0.87) (Supplemental Table 5).
Discussion
This report examines the clinical management and outcomes of patients with AMI who were also diagnosed with COVID-19 in the United States during the first year of the pandemic. By examining AMI-related hospitalizations nationwide, it provides unique insights into the impact of COVID-19 on patients with AMI. The findings revealed a greater than three-fold increase in in-hospital mortality for patients with concurrent COVID-19 and AMI in comparison to those with only AMI. Several factors may have contributed to this increased mortality, including potential disruptions in care due to the presence of COVID-19 and the impact of the virus on the cardiovascular system. A recent large-scale observational study conducted with patients utilizing the National Health Service in England found that individuals diagnosed with both acute coronary syndrome (ACS) and COVID-19 were less likely to receive guideline-directed treatment and had a higher in-hospital and 30-day mortality rate compared to those without COVID-19 and ACS.20 Furthermore, data published from over 1300 chest pain centers in China showed an average delay of 20 minutes for reperfusion therapy early in the pandemic, which led to higher rates of in-hospital mortality and heart failure.6
Previous studies have highlighted a higher risk of complications and mortality for individuals with preexisting cardiovascular disease and COVID-19 infection.21 This analysis builds on this knowledge by demonstrating that patients with AMI and concurrent COVID-19 are more likely to require mechanical ventilation, vasopressor use, and initiation of hemodialysis. Notably, one study demonstrated that patients diagnosed with both ACS and COVID-19 infection have a higher incidence of pulmonary edema and shock at presentation, along with elevated troponin concentrations.20 Additionally, a Chinese study reported an 8% risk of acute cardiac injury, with a 13-fold higher incidence of cardiac injury in critically ill patients with COVID-19.22 Furthermore, patient concerns regarding COVID-19 exposure and avoidance of hospital visits may contribute to the increased mortality and complications associated with AMI during the pandemic. These factors raise concerns that untreated consequences of AMI may result in severe complications for many patients.23, 24, 25
The use of revascularization techniques, including PCI, thrombolytic therapy, and CABG surgery, was compared between patients diagnosed with STEMI with concurrent COVID-19 infection, and those with STEMI only. The analysis showed that patients with STEMI and COVID-19 were 27% less likely to receive PCI, 3.2 times more likely to receive thrombolytic therapy, and 23% less likely to receive either PCI or thrombolytic when compared with patients with STEMI and no COVID-19. This suggests that COVID-19 can negatively impact the delivery of crucial interventions for AMI patients, leading to higher in-hospital mortality rates (aOR 3.19, 95% CI, 2.63 - 3.88) for those with AMI and COVID-19 compared to those without COVID-19. The findings emphasize the urgency of developing effective management strategies for patients with COVID-19 and cardiovascular disease. During the COVID-19 pandemic, multiple international guidelines tried to establish a consensus on the optimal treatment approach for patients presenting with AMI. The Chinese Cardiac Society recommended medical management for most patients presenting with NSTEMI and thrombolysis for those with STEMI early on in the pandemic.26 While American and Canadian guidelines had recommended the use of thrombolysis as an alternative to PCI for STEMI patients in cases where PCI services were limited.6 , 27 Nonetheless, the ACC/SCAI and the European Association of Percutaneous Cardiovascular Interventions (EAPCI) recommendations encouraged the use of PCI as first-line therapy for STEMI.27 , 28 Additionally a study done in Japan investigated the effect of the COVID-19 pandemic on cardiovascular care, specifically analyzing hospital arrival time, ambulance use, PCI implementation, and in-hospital mortality. The results demonstrated no significant differences in these parameters before and after the outbreak. 29
Importantly the current analysis uncovered racial disparities in revascularization rates among the study cohort of patients with AMI and concurrent COVID-19, particularly with respect to PCI and CABG after AMI. Specifically, Black, Hispanic, and Asian/Pacific Islander patients were less likely to receive PCI for AMI than White patients, and Black patients were less likely to undergo CABG surgery for AMI. It is important to note that prior research has identified racial disparities in COVID-19 outcomes, with racial minority groups at an increased risk for morbidity and mortality due to COVID-19.30, 31, 32 The study shows disparities in treatment for AMI in the setting of COVID-19, suggesting a possible explanation for the overall poor outcomes seen in these diverse populations. The reasons for disparities in care after AMI are complex and multifactorial, including the role of racism, systemic bias, and other social determinants of health.33
Limitations
The data used in this analysis was sourced from the NIS, which may have some inherent biases. The NIS does not include outpatient mortality, potentially leading to an underestimation of mortality in COVID-19 cases associated with AMI. Moreover, the NIS lacks specific data on lab values, vital signs, and imaging findings, so the conclusions drawn were based solely on discharge diagnoses. It is also impossible to determine, using the NIS data, whether AMI occurred after a COVID-19 infection or vice versa; only the presence of both diagnoses during a single admission is known. Ascertainment bias may be present, as patients were not routinely tested for COVID-19 at the beginning of the pandemic, which could explain the low rates of COVID-19 patients with AMI in this study. However, the mortality rates for patients with both AMI and COVID-19 in this study align with those found in previous studies.19 Finally, both AMI and COVID-19 cases in our study were identified using ICD-10 codes, which are subject to errors. Nonetheless, the large sample size in this study likely helps mitigate the impact of potential coding errors.
Conclusion
Patients who were diagnosed with both COVID-19 and AMI had higher mortality rates and were more likely to experience complications during their hospital stay than those with only AMI. Furthermore, our findings identified racial disparities among patients with AMI and COVID-19, with Black and Asian/Pacific Islander patients receiving lower rates of revascularization compared to White patients with AMI and COVID-19 (Central Fig ). This highlights the urgent need to address these systemic healthcare disparities to improve health equity in diverse patient populations. It is imperative that healthcare policies and interventions are implemented to ensure that all patients have access to high-quality care, regardless of their race or ethnicity. This will require a multifaceted approach, including increasing access to healthcare in underserved communities, promoting culturally sensitive care, and addressing the root causes of health disparities through social and economic policies. These findings should inform future research and policy initiatives aimed at reducing health disparities and improving health outcomes for all patients.Central FIG Disparities in Mortality Associated with Acute Myocardial Infarction and COVID-19 in the United States: A Nationwide Analysis. Legend: aOR, adjusted odds ratio; 95% CI, 95% confidence interval, AMI, acute myocardial infarction.
Central FIG
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Disclosures: Martha Gulati served on an advisory board for Novartis and Esperion. She is a co-investigator and site PI of the Women's IschemiA TRial to Reduce Events In Non-ObstRuctive CAD (WARRIOR) Study funded by the Department of Defense (Award Number: W81XWH-17-2-0030).
Appendix Supplementary materials
Image, application 1
Supplementary material associated with this article can be found in the online version at doi:10.1016/j.cpcardiol.2023.101798.
==== Refs
References
1 COVID-19 Map. Johns Hopkins Coronavirus Resource Center. Accessed November 21, 2022. Available at: https://coronavirus.jhu.edu/map.html
2 McGonagle D Plein S O'Donnell JS Sharif K Bridgewood C Increased cardiovascular mortality in African Americans with COVID-19 Lancet Respir Med 8 7 2020 649 651 10.1016/S2213-2600(20)30244-7 32473125
3 Wu J Mamas MA Mohamed MO Place and causes of acute cardiovascular mortality during the COVID-19 pandemic Heart 107 2 2021 113 119 10.1136/heartjnl-2020-317912 32988988
4 Boserup B McKenney M Elkbuli A The impact of the COVID-19 pandemic on emergency department visits and patient safety in the United States Am J Emerg Med 38 9 2020 1732 1736 32738468
5 Choi H Lee JH Park HK Impact of the COVID-19 pandemic on patient delay and clinical outcomes for patients with acute myocardial infarction J Korean Med Sci 37 21 2022 10.3346/jkms.2022.37.e167
6 Xiang D Xiang X Zhang W Management and outcomes of patients with STEMI during the COVID-19 pandemic in China J Am Coll Cardiol 76 11 2020 1318 1324 10.1016/j.jacc.2020.06.039 32828614
7 Cataldo P Verdugo FJ Bonta C Consequences of COVID-19 pandemic on myocardial infarction reperfusion therapy and prognosis Rev Med Chil 149 5 2021 672 681 10.4067/s0034-98872021000500672 34751319
8 Kitahara S Fujino M Honda S COVID-19 pandemic is associated with mechanical complications in patients with ST-elevation myocardial infarction Open Heart 8 1 2021 e001497 10.1136/openhrt-2020-001497
9 Sisakian H Martirosyan S Shekoyan S Terteryan A Hovhannisyan M Asymptomatic COVID-19 and ST-elevation myocardial infarction in young adults: lessons learned from two similar cases ESC Heart Fail 9 1 2022 775 781 10.1002/ehf2.13690 34779152
10 Choudry FA Hamshere SM Rathod KS High thrombus burden in patients with COVID-19 presenting with ST-segment elevation myocardial infarction J Am Coll Cardiol 76 10 2020 1168 1176 10.1016/j.jacc.2020.07.022 32679155
11 Kite TA Ludman PF Gale CP International prospective registry of acute coronary syndromes in patients with COVID-19 J Am Coll Cardiol 77 20 2021 2466 2476 10.1016/j.jacc.2021.03.309 34016259
12 Rodriguez-Leor O Cid Alvarez AB Pérez de Prado A In-hospital outcomes of COVID-19 ST-elevation myocardial infarction patients EuroIntervention 16 17 2021 1426 1433 10.4244/eij-d-20-00935 33164893
13 McNally EM Elkind MSV Benjamin IJ Impact of the COVID-19 pandemic on cardiovascular science: anticipating problems and potential solutions: a presidential advisory from the American Heart Association Circulation 144 23 2021 e461 e471 10.1161/CIR.0000000000001027 34719260
14 Rodriguez F Solomon N de Lemos JA Racial and ethnic differences in presentation and outcomes for patients hospitalized with COVID-19: findings from the American Heart Association's COVID-19 cardiovascular disease registry Circulation 143 24 2021 2332 2342 10.1161/CIRCULATIONAHA.120.052278 33200953
15 Vardar U Ilelaboye A Murthi M Racial disparities in patients with COVID-19 infection: a national inpatient sample analysis Cureus 15 2 2023 e35039 10.7759/cureus.35039 36942174
16 Mackey K Ayers CK Kondo KK Racial and ethnic disparities in COVID-19-related infections, hospitalizations, and deaths : a systematic review Ann Intern Med 174 3 2021 362 373 10.7326/M20-6306 33253040
17 Racial differences in management and outcomes of acute myocardial infarction during COVID-19 pandemic | Heart. Accessed April 26, 2023. Available at: https://heart.bmj.com/content/107/9/734
18 HCUP-US NIS Overview. Accessed November 26, 2022. Available at: https://www.hcup-us.ahrq.gov/nisoverview.jsp
19 Austin SR Wong YN Uzzo RG Beck JR Egleston BL Why summary comorbidity measures such as the Charlson Comorbidity Index and Elixhauser score work Med Care 53 9 2015 e65 e72 10.1097/MLR.0b013e318297429c 23703645
20 Rashid M Wu J Timmis A Outcomes of COVID-19-positive acute coronary syndrome patients: a multisource electronic healthcare records study from England J Intern Med 290 1 2021 88 100 10.1111/joim.13246 33462815
21 Driggin E Madhavan MV Bikdeli B Cardiovascular considerations for patients, health care workers, and health systems during the COVID-19 pandemic J Am Coll Cardiol 75 18 2020 2352 2371 10.1016/j.jacc.2020.03.031 32201335
22 Li B Yang J Zhao F Prevalence and impact of cardiovascular metabolic diseases on COVID-19 in China Clin Res Cardiol 109 5 2020 531 538 10.1007/s00392-020-01626-9 32161990
23 Braiteh N Rehman WU Alom M Decrease in acute coronary syndrome presentations during the COVID-19 pandemic in upstate New York Am Heart J 226 2020 147 151 10.1016/j.ahj.2020.05.009 32569892
24 Papafaklis MI Katsouras CS Tsigkas G Missing” acute coronary syndrome hospitalizations during the COVID-19 era in Greece: medical care avoidance combined with a true reduction in incidence? Clin Cardiol 43 10 2020 1142 1149 10.1002/clc.23424 32691901
25 Kapoor A Yadav R Will the hidden specter of acute coronary syndrome (ACS) and ST-segment elevation myocardial infarction (STEMI) emerge from the avalanche of COVID-19? Indian Heart J 72 3 2020 192 193 10.1016/j.ihj.2020.05.017 32768020
26 Han Y Zeng H Jiang H CSC expert consensus on principles of clinical management of patients with severe emergent cardiovascular diseases during the COVID-19 epidemic Circulation 141 20 2020 e810 e816 10.1161/CIRCULATIONAHA.120.047011 32216640
27 Mahmud E Dauerman HL Welt FGP Management of acute myocardial infarction during the COVID-19 pandemic: a position statement from the society for cardiovascular angiography and interventions (SCAI), the American College of Cardiology (ACC), and the American College of Emergency Physicians (ACEP) J Am Coll Cardiol 76 11 2020 1375 1384 10.1016/j.jacc.2020.04.039 32330544
28 Chieffo A Stefanini GG Price S EAPCI position statement on invasive management of acute coronary syndromes during the COVID-19 pandemic Eur Heart J 41 19 2020 1839 1851 10.1093/eurheartj/ehaa381 32405641
29 Hao K Takahashi J Sato K The influence of COVID-19 pandemic on management of acute myocardial infarction in Japan; Insight from the Miyagi AMI Registry Study IJC Heart Vasc 43 2022 101116 10.1016/j.ijcha.2022.101116
30 Shekhar R Sheikh AB Suriya SS Upadhyay S Zafar A Neurological complications among native Americans with COVID-19: Our experience at a tertiary care academic hospital in the U.S J Stroke Cerebrovasc Dis 29 12 2020 105260 10.1016/j.jstrokecerebrovasdis.2020.105260
31 COVID-19 Hospitalizations. Accessed January 14, 2023. Available at: https://gis.cdc.gov/grasp/COVIDNet/COVID19_5.html
32 CDC Cases, Data, and Surveillance 2020 Accessed January 14, 2023Available at: https://www.cdc.gov/coronavirus/2019-ncov/covid-data/investigations-discovery/hospitalization-death-by-race-ethnicity.html
33 Raifman MA Raifman JR Disparities in the population at risk of severe illness from COVID-19 by race/ethnicity and income Am J Prev Med 59 1 2020 137 139 10.1016/j.amepre.2020.04.003 32430225
|
PMC010xxxxxx/PMC10174468.txt |
==== Front
Radiat Phys Chem Oxf Engl 1993
Radiat Phys Chem Oxf Engl 1993
Radiation Physics and Chemistry
0969-806X
0969-806X
Elsevier Ltd.
S0969-806X(23)00268-2
10.1016/j.radphyschem.2023.111023
111023
Article
Investigation of the Radiographer's adherence and compliance with radiation protection and infection control practices during COVID-19 mobile radiography
Khandaker Mayeen Uddin ab∗∗
Abuzaid Mohamed M. c∗∗∗
Mohamed Ikhlas A. d
Yousef Mohamed e
Jastaniah Saddig e
Alshammari Qurain T. f
Alghamdi Salem Saeed g
Osman Hamid h
Mohamed Ahmed Amna i
Musa Alamin i
Ahmed Medani Afaf Mohamed i
Lam S.E. a∗
Bradley D.A. aj
a Research Centre for Applied Physics and Radiation Technologies, School of Engineering and Technology, Sunway University, Bandar Sunway, 47500, Malaysia
b Department of General Educational Development, Faculty of Science and Information Technology, Daffodil International University, DIU Rd, Dhaka, 1341, Bangladesh
c University of Sharjah, College of Health Sciences, Department of Medical Diagnostic Imaging, Sharjah, United Arab Emirates
d Diagnostic Radiology Department, College of Medical Radiological Science, Sudan University of Science and Technology, Khartoum, Sudan
e Radiologic Sciences Program, Batterjee Medical College, Jeddah, Saudi Arabia
f Diagnostic Radiology Department, College of Applied Medical Sciences, University of Hail, Hai'l, Saudi Arabia
g Department of Applied Radiologic Technology, College of Applied Medical Sciences, University of Jeddah, Saudi Arabia
h Department of Radiologic Sciences, College of Applied Medical Sciences, Taif University, Saudi Arabia
i Department of Radiological Sciences, College of Applied Medical Sciences, King Khalid University, Abha, Saudi Arabia
j Department of Physics, University of Surrey, Guidlford, GU2 7XH, United Kingdom
∗ Corresponding author.
∗∗ Corresponding author. Research Centre for Applied Physics and Radiation Technologies, School of Engineering and Technology, Sunway University, Bandar Sunway, 47500, Malaysia.
∗∗∗ Corresponding author. University of Sharjah, College of Health Sciences, Department of Medical Diagnostic Imaging, Sharjah, United Arab Emirates.
11 5 2023
9 2023
11 5 2023
210 111023111023
16 3 2023
8 5 2023
10 5 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Radiological staff, especially radiographers, work as front liners against the COVID-19 outbreak. This study aims to assess compliance with radiation protection and infection control practices during COVID-19 mobile radiography procedures. This cross-sectional study included 234 radiographers (females, 56%, n = 131; males, 44%, n = 103) who were asked to complete an online questionnaire consisting of demographic data, radiation protection and infection control practices during COVID-19 portable cases, and knowledge and awareness. After informed consent was completed, SPSS statistical software was used for the data analysis. The most common age group of participants ranged from 18 to 25 years old (30.3%, n = 71). Bachelor's degree holders were 74.4% (n = 174). Most radiographers (39.7%, n = 93) had a working experience of 1–5 years, followed by 27.8% (n = 65) with more than 16 years of experience. Most respondents (62.4%, n = 146) handled approximately 1–5 cases daily, the majority of them (56%, n = 131) stated affirmatively they had obtained special training to handle COVID-19, and when inquired if they had received any special allowances for handling COVID-19 suspected/confirmed cases most of them stated negative (73.9%, n = 173). Most participants stated that they always wear a TLD during portable cases (67.1%, n = 157) and a lead apron (51.7%, n = 121). Around 73% (n = 171) knew the latest information on COVID-19 and attended the COVID-19 awareness course. A significant association was found between the work experience of the radiographers and their responses to following the best practices (p = 0.018, α = 0.05). Radiographers who had COVID-19 training (μ = 48.78) tend to adhere more to best practices than those who have not (p = 0.04, α = 0.05). Further, respondents who handled more than 16/more COVID-19 suspected/confirmed cases followed the best practices more (μ = 50.38) than those who handled less (p = 0.04, α = 0.05). This study revealed detailed information on radiation protection and infection control practices during COVID-19 mobile radiography. It has been observed that the participants/radiographers have good knowledge and awareness of radiation protection and infection-control practices. The present results may be used to plan future requirements regarding resources and training to ensure patient safety.
Keywords
COVID-19
Mobile radiography
Infection control
Radiation protection
Radiographers
Handling Editor: Dr. Chris Chantler
==== Body
pmc1 Introduction
Coronavirus disease (COVID-19) is an infectious disease caused by the SARS-CoV-2 virus. Most infected with this virus develop mild to moderate respiratory symptoms and recover without specific treatment. Some people, however, become critically unwell and require medical attention. The first instance was discovered in December 2019 in Wuhan, China (Yu et al., 2020) (Hadi et al., 2021). The disease quickly spread worldwide, resulting in the COVID-19 pandemic (Niu Y et al., 2020). COVID-19 can be identified based on symptoms and verified by RT-PCR or other contaminated secretion nucleic acid tests (Zhang et al., 2020). Chest X-ray (CXR) computed tomography (CT) and laboratory tests may be helpful in diagnosing COVID-19 in those with a high clinical suspicion of infection. Most specialist facilities, clinics and hospitals have mobile radiographic imaging devices (Osman et al., 2023) . The analysis of CXR images from COVID-19 patients revealed that this test was a quick and cost-effective strategy for diagnosing the individuals in question (Chalkia et al., 2022). The number of mobile X-ray procedures increased massively during the pandemic, as they were used for diagnosis and follow-up for suspected and confirmed COVID-19 cases (Abuzaid et al., 2022).
Radiographers and radiologists are highly skilled users of imaging technologies and are involved in direct contact with COVID-19 cases (Martell et al., 2022). They were trained to use imaging for the best benefit of patients. Understanding potential concerns from ionizing radiation is an essential component of their education, as it is necessary to minimise the risk of injury from inappropriate or excessive radiation usage. The radiology profession's obligations go beyond radiation protection, including infection prevention when doing radiology examinations (Martell et al., 2022).
Consequently, radiographers had difficulty establishing an adequate distance between them, the radiation source, and the resultant scattered radiation at all times. While the International Society of Radiographers and Radiologic Technologists recommends a distance of 2 m between the patient and radiographer, studies have shown that yearly maximum permissible doses are not exceeded for mobile X-ray imaging at distances of 1 m (ISSRT,2020). However, for COVID-19, stricter infection control measures, such as additional equipment restraints or lead-equivalent protection, have been recommended (Yeung et al., 2022).
Radiological staff, especially radiographers, working at the front line in combat of the COVID-19 outbreak. They are in direct contact with patients, bearing the responsibility and risk of infection prevention, control and radiation protection. Undertaking radiographic procedures when there is a possibility that the patient may be Covid-19 positive brings its challenges (ISRRT, 2020, Thomas et al., 2022 , Society of Radiographers U., 2021 ).
The COVID-19 pandemic has led to the implementation of infection prevention and control (IPC) measures in healthcare settings, including radiology departments. This scoping review aimed to identify and summarise the IPC practices that have been implemented in radiology departments during the pandemic. Various studies were included in the review. The findings showed that the most common IPC measures included triaging patients, screening for COVID-19 symptoms, using personal protective equipment (PPE), environmental cleaning and disinfection, and social distancing. The review also identified challenges faced by radiology departments in implementing IPC measures, including PPE shortages, staff training and patient compliance with IPC measures (Yu et al., 2020; Naylor et al., 2022; Mc Fadden et al., 2022, Clements et al., 2020).
The aim is to study radiographers’ compliance with radiation protection and infection control practices during COVID-19 mobile radiography.
2 Materials and methods
2.1 Methods
During the pandemic, a cross-sectional study was conducted among radiographers and radiological technologists who examined suspected and confirmed COVID-19 cases in Saudi Arabia. The research team, four senior radiographers, and infection control managers devised, vetted, and piloted the survey. The evaluation was conducted to ensure that the questions were displayed appropriately and comprehensibly and returned the required information. The results of the pilot research were removed from the primary investigation.
The demographic characteristics were the first section of the questionnaire (e.g., age, academic qualification, work experience, the average number of COVID-19 cases handled daily, and receiving any special training or allowance during the pandemic). The second part examined whether participants’ radiation protection practices reduced radiation exposure for workers, staff, and patients. The use of thermoluminescence dosimeters (TLDs), lead aprons, thyroid collars, collimation, distance shielding, gonad shielding and the proper exposure parameters were all studied. In the third part, infection control measures were assessed, such as personal protective equipment (PPE), infection prevention, equipment disinfection, hand hygiene and following the standard documentation routine.
The survey employed a 4-point Likert scale with the following scores: (4) always, (3) often, (2) occasionally, and (1) never. The better the practice, the higher the score. By dividing the total score by the maximum possible score multiplied by 100, the score was converted to a percentage scale. As a result, the score was divided into three categories: poor adherence (less than 60%), moderate adherence (60–80%), and good adherence (more than 80%).
2.2 Data analysis
A total of 234 responses were received, of which all questionnaires were completed and therefore included in the study. The responses were collected and analyzed using the Statistical Package for Social Sciences (SPSS) and IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM. Graphs for responses were created using Microsoft Office Excel 2016 (Microsoft Corporation, CA, USA). Following the descriptive statistics for all questions, a one-way analysis of variance (ANOVA) was conducted to analyze the association between the demographics and the participants’ infection control and radiation protection practices.
2.3 Ethical considerations
The Institutional Research Unit approved the research protocol. All respondents gave their informed consent after learning about the study's goals and being assured of anonymity. The participants were told that they could leave at any point during the data collection process.
3 Results
3.1 Demographic and participants’ background
The responses recorded from the 234 participants are shown in Table 1 . The majority were females (56%, n = 131); 30.3% (n = 71) were 18–25 years old. Nearly three-quarters of the participants had the highest qualification of a bachelor's degree (74.4%, n = 174), and participants with a PhD were the least (2.1%, n = 5).Table 1 Distribution of the demographic characteristics of the participants (n = 234).
Table 1Criteria Responses Frequency (%)
Gender Male 103 (44.0)
Female 131 (56.0)
Age (years) 18–25 71 (30.3)
26–35 70 (29.9)
36–45 53 (22.6)
46–65 40 (17.1)
Qualification Diploma 25 (10.7)
Bachelors 174 (74.4)
Masters 30 (12.8)
PhD 5 (2.1)
Experience 1–5 years 93 (39.7)
6–10 years 35 (15.0)
11–15 years 41 (17.5)
More than 16 years 65 (27.8)
Number of Covid-19 cases handled per day 1–5 146 (62.4)
6–10 54 (23.1)
11–15 16 (6.8)
More than 16 18 (7.7)
COVID-19 related training Yes 131 (56.0)
No 103 (44.0)
Allowances/incentives received Yes 61 (26.1)
No 173 (73.9)
Most radiographers (39.7%, n = 93) had a working experience of 1–5 years, followed by 27.8% (n = 65) of radiographers with more than 16 years of experience. The respondents were asked about ‘the approximate number of portable COVID-19 suspected or confirmed cases that they handled daily’, and most respondents stated approximately 1–5 cases (62.4%, n = 146). Some respondents had handled around 11–15 cases a day (6.8%, n = 16) and more than 16 cases too (7.7%, n = 18). The radiographers were also inquired ‘if they had obtained special training to handle the COVID-19 confirmed or suspected cases’, for which the majority of them stated affirmative (56%, n = 131) and when inquired ‘if they had received any special allowances for handling COVID-19 suspected/confirmed cases', most of them stated negative (73.9%, n = 173).
3.2 Radiation protection practices during COVID-19 portable cases
This questionnaire section comprises 10 sub-questions, each gathering responses on a 4-point Likert scale. The descriptive statistics of the responses are given in Table 2 .Table 2 Radiation protection practice frequencies.
Table 2Monitoring characteristics Never (%) Sometimes (%) Most of the time (%) Always (%)
Wearing TLD 25(10.7) 22(9.4) 30(12.8) 157(67.1)
Wearing lead apron 21(9.0) 57(24.4) 35(15.0) 121(51.7)
Wearing thyroid collar 120(51.3) 57(24.4) 26(11.1) 31(13.2)
Using proper collimation 3(1.3) 35(15.0) 79(33.8) 117(50.0)
Using proper SID/FFD 5(2.1) 42(17.9) 74(31.6) 113(48.3)
Apply patient gonad shielding 28(12.0) 82(35.0) 53(22.6) 71(30.3)
Apply patient lead shielding 24(10.3) 67(28.6) 64(27.4) 79(33.8)
Using minimum exposure time 3(1.3) 26(11.1) 93(39.7) 112(47.9)
Using the lead apron for all co-patient/staff 16(6.8) 55(23.5) 82(35.0) 81(34.6)
Closing the room door 1(0.4) 9(3.8) 82(35.0) 142(60.7)
Most participants stated that they always wear a TLD during portable cases (67.1%, n = 157) and a lead apron (51.7%, n = 121). Most participants never responded when asked whether they wore a thyroid collar (51.3%, n = 120). The respondents said they always used proper collimation (50%, n = 117) and proper SID/FFD (48.3%, n = 113). When applying gonad shielding, most participants said they sometimes applied it (35%, n = 82), and close to one-third of the participants (33.8%, n = 79) said they always applied gonad shielding. Most respondents stated that they always use minimum exposure time (47.9%, n = 112), provide a lead apron for all co-patient/staff (34.6, n = 81), and close the room door (60.7%, n = 234).
3.3 Infection control practices during COVID-19 portable cases
The third part of the questionnaire gathered responses through 5 sub-questions based on a 4-point Likert scale. The responses are given in Table 3 . Most respondents stated that they always wear personal protective gear, facemasks, gloves, face shields, etc. (71.8%, n = 168) and maintain appropriate isolation precaution practices (64.5%, n = 151) when handling COVID-19 suspected/confirmed patients in portable radiography. The respondents also stated that they always disinfect according to infection control policies and procedures (61.1%, n = 143), maintain hand hygiene (personal cleanliness) (69.7%, n = 163), and follow standardized hospital protocols for decontaminating imaging equipment after the imaging procedure (68.8%, n = 161).Table 3 Infection control practice frequencies.
Table 3Information was sought. Never,
N (%) Sometimes,
N (%) Most of the time,
N (%) Always,
N (%)
They wore personal protective gear, facemasks, gloves, face shields, etc. 6(2.6) 29(12.4) 31(13.2) 168(71.8)
Appropriate isolation precaution practices are maintained during portable radiography 9(3.8) 27(11.5) 47(20.1) 151(64.5)
Equipment disinfected according to Infection Control policies and procedures 10(4.3) 21(9.0) 60(25.6) 143(61.1)
Hand hygiene (personal cleanliness) 4(1.7) 17(7.3) 50(21.4) 163(69.7)
Standardized hospital protocols for decontaminating imaging equipment after the imaging procedure 7(3.0) 17(7.3) 49(20.9) 161(68.8)
The respondents' knowledge, awareness, and information were analyzed through a series of 6 questions. The respondents were asked if they knew the latest information on COVID-19 and had attended the COVID-19 awareness course; most (73%, n = 171) stated yes. A majority of the respondents had attended the COVID-19 awareness course (73.1%, n = 171), received department support during the pandemic (52.6%, n = 123), and received hospital support during the pandemic (46.6%, n = 109).
Almost half of the respondents (51.3%, n = 120) stated that they are confident in handling COVID-19-suspected patients to a great extent. For most participants, health organizations (69.2%, n = 162) were the primary sources of information and social media, second to it (12.0%, n = 28).
3.4 Comparison of demographics and responses
A one-way ANOVA was conducted to analyze any association between the demographics and their responses regarding radiation protection and infection control practices. The 4-point Likert scale responses were scored from 1 to 4, where 1 = ‘Never’, 2 = 'Sometimes', 3 = ‘Most of the time’, and 4 = 'Always'. The lowest score of 15 meant not following proper infection control and radiation protection practices, and the highest score of 60 meant adhering to best practices.
A significant association was found between the work experience of the radiographers and their responses to following the best practices (p = 0.018, α = 0.05). In contrast, radiographers with more than 16 years of experience (μ = 49.7) and between 1 and 5 years (μ = 48.2) of experience tend to follow more than the rest. The study also revealed that radiographers who had COVID-19 training (μ = 48.78) tend to adhere more to best practices than those who have not (p = 0.04, α = 0.05). Further, respondents who handled more than 16 COVID-19 suspected/confirmed cases followed the best practices more (μ = 50.38) compared to those who handled lesser ((p = 0.04, α = 0.05).
4 Discussion
Rapid and precise diagnostic procedures were required during the COVID-19 epidemic. It has been confirmed that medical imaging (chest radiography and computed tomography) is critical in the fight against COVID-19 (Yeung et al., 2022). The safety of patients, professionals, and the general public during medical imaging studies is critical. As the epidemic continues, medical imaging professionals must develop the knowledge and skills to ensure patient safety and up-to-date information. Several studies and papers have been published focusing on patient safety in medical imaging during COVID-19 and the obstacles and optimization solutions in radiology service during the pandemic (Abuzaid et al., 2022). During the COVID-19 epidemic, the increased use of mobile radiography necessitated greater attention to occupational and patient dosages. The International Society of Radiographers and Radiologic Technologists (ISRRT) issued a response document in April 2020 to ensure patient safety and radiation protection during medical imaging procedures in COVID-19 instances (ISSRT,2020).
4.1 Radiation protection practices during COVID-19 mobile cases
Radiographers who work with ionizing radiation are responsible for patient and public radiation safety. Because X-rays use ionizing radiation, which can deposit energy in human cells and cause tissue changes, patient-associated risks must be minimized (Alkhorayef et al., 2020;Osman et al., 2022). Dose reduction is accomplished by reducing the radiation used to create the clinical images required to answer a medical query. The ALARA concept is crucial because it can help to avoid overexposure and unnecessary exposure. ALARA principles are based on three elements controlled by radiographers: time, distance and shielding (Abuzaid et al., 2022; Elshami et al., 2019). Results showed moderate attention to the use of proper collimation (50%, n = 117), minimum exposure time (47.9%, n = 112), and use of proper SID/FFD (48.3%, n = 113). Both collimation and distance must be adjusted strictly to focus on a specific part of the patient's body, limiting the radiation beam within the range defined by clinical procedures and ensuring that it matches the image detector. In addition, to reduce scattered radiation, proper collimation and distance may improve image contrast and reduce geometric distortion (Niu et al., 2020). Reducing exposure time can directly reduce the radiation dose, absorbed dose, and biological effects of ionizing radiation.
The practices that were either neglected or never used by a large proportion of the radiographers were the use of lead gloves during fluoroscopy (37.6%), wearing a thyroid collar during OT (18.3%), and wearing TLDs (15.7%). Around 62.5% wear TLD and lead aprons during the mobile radiography procedure. This result agrees with the study done by Abuzaid et al. Al, entitled ‘Assessment of compliance to radiation safety and protection at the radiology department’ (Abuzaid et al., 2019).
When using mobile DR equipment for examination in an area such as a fever clinic, where no dedicated diagnostic examination room is built, the persons around such an area should be informed to leave as far as possible. Additionally, there should be no other persons in the main direction of the radiation beam. When using mobile DR equipment for X-ray examination in a quarantine ward, protection measures should be taken for the patients in the adjacent beds within 2 m of the DR equipment. At the same time, irradiation beams should not be directed toward other patients. The length of the cable connecting the exposure switch should not be less than 3 m; otherwise, a remote control/delayed exposure switch has to be equipped.
4.2 Infection control practices and knowledge during COVID-19 mobile cases
Knitted constructs are considered better suited to cloth masks than woven structures due to their thicker cross-sections and high air permeability. People should be encouraged to procure a high-quality mask to help reduce the spread of SARS-CoV-2 and shield against sun exposure
(ISRRT, 2020; WHO Guidance Note., 2020). However, the types of gloves should also be carefully selected according to the objects to be protected against. The best gloves for healthcare workers are the first latex and second nitrile. Although this principle is appropriate for protection from viral infection, it is not always suitable for protecting unsealed radioactive materials (Niu et al., 2020; Amalou et al., 2020).
When the mobile DR equipment needs to be moved out of the fever or other clinics for use, the entire surface of the equipment must receive wipe disinfection and then be exposed to ultraviolet light for more than 30 min before use. The worker should wear an N95 mask or higher, a disposable fluid-resistant gown, gloves, goggles, or a visor for eye protection (ISSRT., 2020, WHO Guidance Note., 2020). Our workers wore PPE, a disposable fluid-resistant gown, gloves, goggles, or a visor for eye protection.
Being familiar with the requirements for infection control and prevention at different posts, it is necessary to know the related types of protective articles and their uses and the requirements and methods for disinfection of personnel, equipment, and places (Yeung et al., 2022). Table 4 shows that most participants have enough knowledge and awareness, and received support from their departments in attending courses and trainings during the pandemic.Table 4 Knowledge and awareness of participants.
Table 4Information was sought. Yes
N (%) No
N (%) Sometimes
N (%)
Up to date aware of the latest information on COVID-19 171(73.1) 28(12.0) 35(15.0)
Attend any COVID-19 Awareness Course 171(73.1) 42(17.9) 21(9.0)
Get any department support during the pandemic 123(52.6) 52(22.2) 59(25.2)
Get any hospital support during the pandemic 109(46.6) 68(29.1) 57(24.4)
4.3 Comparison of demographics and responses
A significant association was found between the radiographers’ work experience and their responses to following the best practices. However, this result contrasts with the study ‘Knowledge of COVID-19 infection control among healthcare workers in radiology departments in Saudi Arabia’. There was a significant association between the profession and good clinical practices in radiology departments regarding COVID-19. Such knowledge could limit the spread of COVID-19 among healthcare workers in radiology departments.
The study also highlights the importance of improving training, department design, patient triage, post-exposure patient handling, and the implementation of paperless systems to better handle COVID-19 and protect radiology staff. Additionally, a novel isolation bag device is proposed for use in CT to facilitate containment and reduce contamination in radiology departments during the COVID pandemic. Overall, this study sheds light on the crucial role of radiographers in fighting against COVID-19 and emphasizes the need for their safety and protection.
5 Conclusion
This study investigates radiographers’ adherence to and compliance with radiation protection while performing mobile radiography for COVID-19 cases. The incidence of COVID-19 infection among radiology personnel is primarily due to a poor understanding of the newly emerged virus. The disease can better be handled, and the radiology staff can better be protected by improving the training, department design, patient triage, post-exposure patient handling, and implementing paperless systems. A novel isolation bag device is feasible for use in CT and might facilitate containment and reduce contamination in radiology departments during the COVID Pandemic. The present results may also be used to plan future requirements regarding resources and training to ensure patient safety.
6 Recommendations
The paper recommends improving the training, department design, patient triage, post-exposure patient handling, and implementation of paperless systems to better handle COVID-19 and protect radiology staff. The study also proposes a novel isolation bag device for use in CT to facilitate containment and reduce contamination in radiology departments during the COVID pandemic. The results of this study may be used to plan future requirements in terms of resources and training to ensure patient safety. Overall, the paper emphasizes the importance of ensuring the safety and protection of radiographers during the COVID-19 pandemic.
Author contributions
Conceptualization: M.M.A;
Methodology: I.A.M and M.Y;
Software: I.A.M, M.Y and A.M.A;
Validation: M.U.K, A.M.A and H.O; Formal analysis: H.O, M.Y and S.J;
Investigation: H.O, M.M.A, Q.T.A, A.M. and A.M.A.M.;
Resources: H.O, M.M.A and Q.T.A; Data curation: H.O and S.S.A; Writing—original draft preparation: H.O, M.M.A, I.A.M and M.Y; Writing—review and editing: M.U.K, S.E.L, D.A.B and H.O;
Visualization: S.E.L and D.A.B;
Supervision: M.U.K and H.O; Project administration: H.O; Funding acquisition: H.O.
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.
Data availability
Data will be made available on request.
Acknowledgements
The authors extend their appreciation to the Deanship of Scientific Research at King Khalid University for funding this work through small research groups under grant number (RGP 1/321/44).
==== Refs
References
Abuzaid M.M. Elshami W. Shawki M. Salama D. Assessment of compliance to radiation safety and protection at the radiology department Int J Radiat Res 17 3 2019 439 446
Abuzaid M.M. Elshami W. Tekin H.O. Infection control and radiation safety practices in the radiology department during the COVID-19 outbreak PLoS One 17 12 2022 e0279607
Alkhorayef M. Sulieman A. Alzahrani K. Abuzaid M. Alomair O.I. Almuwannis M. Radiation risk for patients undergoing cardiac computed tomography examinations Appl. Radiat. Isot. 2020 109520
Amalou A. Turkbey B. Xu S. Turkbey E. An P. Carrafiello G. Disposable isolation device to reduce COVID-19 contamination during CT scanning Acad. Radiol. 27 8 2020 1119 1125 32499157
Chalkia M. Arkoudis N.A. Maragkoudakis E. Rallis S. Tremi I. Georgakilas A.G. The role of ionizing radiation for diagnosis and treatment against COVID-19: evidence and considerations Cells 11 3 2022 1 21
Clements J. COVID-19 and health workers : radiation protection Conference of Radiation Control Program Directors 2020 Inc.
Elshami W. Abuzaid M. Piersson A.D. Mira O. AbdelHamid M. Zheng X. Occupational dose and radiation protection practice in uae: a retrospective cross-sectional cohort study (2002–2016) Radiat. Protect. Dosim. 187 4 2019 426 437
Hadi Y. Aljuaid M. Hamid H. Al-Khalfan A.A. Faizo N. The role of chest computed tomography imaging in patient management during the coronavirus disease pandemic in Saudi Arabia: a cross-sectional study Crescent Journal of Medical and Biological Sciences 8 2021
ISRRT COVID-19 ISRRT Response Document - appropriate and safe use of Medical Imaging and Radiation Therapy with infection control measures considered in addition to standard radiation protection procedures ISRRT Board of Management Approved April 2020 COVID-1 2020 ISRRT Board Manag 1 93
Martell M. Perko T. Zeleznik N. Molyneux-Hodgson S. Lessons being learned from the Covid-19 pandemic for radiological emergencies and vice versa: report from expert discussions J. Radiol. Prot. 42 2022 011518
Mc Fadden S. Flood T. Shepherd P. Gilleece T. Impact of COVID-19 on service delivery in radiology and radiotherapy (London, England : 1995) Radiography 28 1 2022 S16 S26 10.1016/j.radi.2022.03.009 35422396
Naylor S. Booth S. Harvey-Lloyd J. Strudwick R. Experiences of diagnostic radiographers through the Covid-19 pandemic Radiography 28 1 2022 187 192 34736824
Niu Y. Xian J. Lei Z. Liu X. Sun Q. Management of infection control and radiological protection in diagnostic radiology examination of COVID-19 cases Radiat Med Prot 1 2 2020 75 80 32835346
Osman H. Alosaimi M. Alghamdi F. Aljuaid A. Alqurashi M. Aldajani S. Faizo N.L. Alamri S. Althobaiti Y.S. Khandaker M.U. Suleiman A. Evaluation of CT dose and establishment of local DRLs for abdomen and pelvis examinations in hospitals in Taif City, Saudi Arabia Radiation Physics and Chemistry 202 2023 110206 10.1016/j.radphyschem.2022.110206
Osman H. Raafat B.M. Faizo N.L. Ahmed R.M. Alamri S. Alghamdi A.J. Almahwasi A. Alharbi M.K.M. Sulieman A. Khandaker M.U. Diagnostic reference level for adult pelvic examination in several hospitals of Taif and Kharaj city, Saudi Arabia Applied Radiation and Isotopes 180 2022 110049 10.1016/j.apradiso.2021.110049 34864437
Society of Radiographers U. Current and Future Roles of Diagnostic Radiographers 2021
Thomas M. Suliman S. Allen M. Hameed M. Ghaffar A. Emara M. A cross sectional survey on the effect of COVID-19 related restrictions on undergraduate and postgraduate medical education in Qatar BMC Med. Educ. 22 1 2022 1 10 34980091
Who Guidance Note Use of Chest Imaging in COVID-19: a Rapid Advice Guia [Internet] 2020 World Heal Organ 56 Available from: (WHO/2019-nCoV/Clinical/Radiology_ imaging/2020.1)
Yeung P. Pinson J.A. Lawson M. Leong C. Badawy M.K. COVID-19 pandemic and the effect of increased utilisation of mobile X-ray examinations on radiation dose to radiographers J Med Radiat Sci 2022 1 9
Yu J. Ding N. Chen H. Liu X.J. He W jie cai Dai W. Infection control against COVID-19 in departments of radiology Acad. Radiol. 27 5 2020 614 617 32276755
Zhang H. Yu J. Xu H. Lei Y. Pu Z. Mo Y. Corona virus international public health emergencies: implications for radiology management Acad. Radiol. 76 1 2020 110206 10.1016/j.radphyschem.2022.110206 1–5. 202:2023
|
PMC010xxxxxx/PMC10175057.txt |
==== Front
Int Rev Educ
Int Rev Educ
International Review of Education. Internationale Zeitschrift Fur Erziehungswissenschaft. Revue Internationale De Pedagogie
0020-8566
1573-0638
Springer Netherlands Dordrecht
9996
10.1007/s11159-023-09996-8
Original Paper
A tripartite understanding of experiences of young apprentices: A case study of the London Borough of Hounslow
Hansberry Priscilla priscilla.hansberry@hounslow.gov.uk
1Priscilla Hansberry
was the 14–19 Service Partnership and External Funding Manager at the London Borough of Hounslow. She is now Policy and Development Officer for Education and Skills at London Borough of Hounslow.
http://orcid.org/0000-0002-4478-4594
Gerhardt Trevor dr.trevor.gerhardt@gmail.com
2Trevor Gerhardt
PhD, was the Academic Lead for Work-Integrated Learning at Pearson College London. He is now Reader and Director of Studies for Higher Degree Apprenticeships at the University of Kent Business School.
1 Hounslow House, London Borough of Hounslow, London, UK
2 grid.9759.2 0000 0001 2232 2818 Kent Business School, University of Kent, Medway, UK
12 5 2023
2023
69 1-2 175206
7 4 2023
© UNESCO Institute for Lifelong Learning and Springer Nature B.V. 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
In 2019, a decline in apprenticeship starts prompted the London Borough of Hounslow to make an apprenticeship pledge in its Corporate Plan 2019–2024, committing to create 4,000 new apprenticeships and training opportunities to help young people into work. This article investigates experiences of young apprentices in Hounslow before and during the COVID-19 pandemic. Exploring the perspectives of two apprentices, two employers and one training provider in a small-scale qualitative study, the authors identify key hindering and supporting factors affecting entry into and sustainability of apprenticeships, and progression towards professional employment. They found that labour market entry was intensely hindered by competition (with peers who had better maths and English qualifications, for a small number of apprenticeships) and organisational barriers (such as managers with prejudices against young people, stigmatising apprentices and apprenticeships). Supportive factors identified include personal characteristics (such as a positive mindset, enabling young people to persevere despite a disadvantaged socioeconomic background and lack of family support, for example) and supportive relationships (e.g. mentoring) between apprentices and their training providers or employers.
Résumé
Compréhension tripartite des expériences de jeunes apprentis : étude de cas réalisée dans le quartier londonien de Hounslow – En 2019, un déclin du nombre de nouveaux contrats d’apprentissage a poussé le quartier londonien de Hounslow à prendre un engagement en faveur de l’apprentissage dans son plan 2019–2024 pour les entreprises, dans lequel il s’engageait à créer 4 000 places d’apprentissage et possibilités de formation pour aider les jeunes à entrer dans la vie active. Cet article se penche sur les expériences de jeunes apprentis à Hounslow avant et pendant la pandémie de COVID-19. Les auteurs ont examiné les points de vue de deux apprentis, de deux employeurs et d’un prestataire de formations dans une étude qualitative à petite échelle et ont identifié des facteurs décisifs freinant ou favorisant l’accession aux apprentissages, la viabilité de ces derniers et le parcours des apprentis vers l’emploi. Ils ont constaté que l’entrée sur le marché du travail était fortement freinée par la concurrence (avec des pairs mieux qualifiés en mathématiques et en anglais qui postulaient pour un nombre restreints de place d’apprentissage) et par des obstacles organisationnels (par exemple des managers qui avaient des préjugés à l’égard des jeunes, qui stigmatisaient les apprentis et les apprentissages). Parmi les facteurs favorables qu’ils ont identifiés, notons les traits de caractère personnels (par exemple un état d’esprit positif permettant aux jeunes de persévérer malgré leur milieu socio-économique défavorisé et l’absence de soutien de leur famille) et les relations de soutien (par exemple le mentorat) entre les apprentis et leurs prestataires de formation ou leurs employeurs.
Keywords
Workplace learning
Apprenticeship policy
Labour market
Economic development
Youth unemployment
England
issue-copyright-statement© UNESCO Institute for Lifelong Learning 2023
==== Body
pmcIntroduction
The COVID-19 pandemic has had an unprecedented impact both in scale and pace on employment worldwide. In the United Kingdom (UK), the fall in employment was estimated in April 2020 to be “at least 1.5 million, equivalent to 5% of all of those in work” (Wilson et al. 2020, p. 1), twice the amount of the fall during the last recession, and fivefold compared to the previous largest quarterly falls at any point since 1971 (ibid.). In the London Borough of Hounslow – the focus of this article –, there was a 165% increase in new unemployment benefit claimants in just one month, between April and May 2020 (London Borough of Hounslow 2020a), 55,600 residents were furloughed, and 37,618 unemployed, representing 21% of the working-age population (London Borough of Hounslow 2020b). The situation for young people in Hounslow reflected that of the national landscape in that young people in Hounslow were affected almost twice as badly as older people due to the economic downturn, with the unemployment rate reaching almost double the rate of the 55–65-year-old group (London Borough of Hounslow 2020b).
Young people are unmistakeably the most impacted compared to all other age groups at a national level (Roberts 2022), with almost half of the total fall in employment (46%) being amongst 16–24-year-olds. Young people are more likely to be furloughed and to subsequently lose their job with a rate of 19% for 18–24-year-olds as of September 2020 (Brewer et al. 2020; Roberts 2022). Young people in Hounslow aged 20–24 were shown to have the highest unemployment rates at 15.4%, above the national average of 14.8% (London Borough of Hounslow 2020b). Employment among the young is a global concern with a lack of skills training linked to poverty (Wolf 2011; Mayombe 2021).
Apprenticeships provide a route into work for young people, combining on-the-job training with qualifications to help develop the skills and knowledge to undertake a particular job (Billett 2016; Mayombe 2021; Roberts 2022). In the past, the apprenticeship framework in England, however, has been criticised for ignoring general and civic educational elements, discounting longer-term interests of employees and failing to bring disengaged employers back on board (Fuller and Unwin 2003; Brockmann et al. 2010; Brockmann and Laurie 2016; Roberts 2022). Later policies sought to address some of these criticisms; one example is the apprenticeship levy introduced in 2017, which places the employer at the heart of the apprenticeship reform (Delebarre 2015) and the new modern apprenticeship (Brockmann et al. 2020; Roberts 2022).1 Alison Fuller and Lorna Unwin (2011) argue that apprenticeships cannot be one-dimensional but should incorporate pedagogical, occupational, locational and social dimensions. In England, modern apprenticeships are grouped into qualification levels equating them with formal education, starting with an intermediate (or Level 2) apprenticeship, mainly for 16–18-year-olds.2 This is followed by an advanced apprenticeship (Level 3, mainly 19–24-year-olds) and higher apprenticeships (Levels 4 and 5, mainly 25+-year-olds) (Fuller et al. 2017). Pre-apprenticeships and young apprenticeships for 14–16-year-olds were abolished in 2009 and 2011, respectively (ibid.). Higher apprenticeships were introduced in 2010, followed by degree apprenticeships (Levels 6 and 7) in 2015 (ibid.).3 COVID-19-related difficulties have led to a reduction in employers’ interest in training apprentices (Cedefop et al. 2022). To address this problem, it is important to identify and understand key hindering and supporting factors affecting apprenticeships, which was the purpose of the small-scale qualitative study we present in this article.
We begin with a literature review, and then provide some background information on the London Borough of Hounslow. The methodology is followed by a detailed presentation and analysis of our findings. The article concludes with a number of recommendations.
Literature review
Understanding the present labour market
Long-term structural shifts in the UK economy from manufacturing to knowledge-intensive business services have altered both sectoral and occupational structures of employment, and consequently altered the breadth and type of employment opportunities for people (Green 2020; Roberts 2022). By 2030, workers will be required to focus more on technological, social and emotional skills, as opposed to physical, manual and basic cognitive skills (ISC 2019). In London, labour market changes are likely to decrease the share of roles requiring no qualifications to less than 3%, whilst simultaneously increasing the proportion of jobs demanding an ordinary or higher-level degree (London Councils 2019). The “polarisation” between low-skilled and high-skilled work results in fewer low- and medium-qualified jobs, but also limits progression from entry-level jobs, and thus poses a particular challenge for young people (London Councils 2019; Roberts 2022). Chun-Chi Lan (2021) reports that long-term training is more vital than short-term recruitment in internships, making apprenticeships still a viable social capital option.
Besides these general labour market trends, the impact of COVID-19 was felt unevenly across different parts of the UK, across age groups and across people of different socioeconomic backgrounds, with sectors such as education, retail, hospitality and construction being the most highly impacted (Cedefop et al. 2022). Furthermore, a COVID-19 survey conducted by the Federation for Industry Sector Skills and Standards (FISSS 2020) highlighted particular problems with apprenticeships, including a difficulty to (re-)engage furloughed apprentices, and a lack of technological equipment at home. Worryingly, many in the sector are seeing the pipeline run dry.
Competitive labour market impacts on “youth” employment
The “polarisation” and competitive environment displaces the less educated by favouring more educated workers (Benda et al. 2019), making it harder for young people to find work (Green 2020; Mayombe 2021). Unemployment disproportionally impacts young people, with 10% of 18–24-year-olds in England unemployed compared to an overall rate of 3.8% unemployment (YFF 2020). Luc Benda et al. (2019) suggest that this increases during economic downturns due to a combination of higher- and medium-level educated workers competing for jobs below their educational level, and employers raising education requirements (Roberts 2022). This aligns with Michael Spence’s job market signalling theory (Spence 1973) and Lester Thurow’s job competition theory (Benda et al. 2019).4 Applicants are formed into a labour queue, ranked by employers based on the training costs required to perform in a given job (Thurow 1979). These queues are then matched, in the order of jobs requiring highest training requirements, with applicants with the lowest training costs. However, during times of turmoil such as COVID-19, these queues often merge pushing the unskilled further back in the queue (Celani and Singh 2011).
The shift away from low-skills occupations where growth is low, combined with difficulties moving into medium- and high-skilled work has led to the demise of the “youth” labour market (Green 2020; Mayombe 2021; Wolf 2011; Roberts 2022). In England, access to high-quality and secure work that offers progression and value to society is in decline for young people (Brockmann and Laurie 2016; Clarke and D’Arcy 2018; Wolf 2011; Roberts 2022).
Transition to the labour market through apprenticeships: social mobility
In England, apprenticeships have progressively supported social mobility and capital amongst young people from disadvantaged backgrounds and those with low qualifications (London Councils 2020a). However, in comparison to other European countries such as Germany, for example, apprenticeships in England conform to the “skills-based model” based on narrow specialisms rather than occupational capacity (Fuller and Unwin 2003; Brockmann et al. 2010; Brockmann and Laurie 2016). Ben Gadsby (2019a) argues that disadvantaged young people account for a disproportionate share of growth in apprenticeship starts. The Social Mobility Commission (Battiston et al. 2020) contends that the earnings gap between disadvantaged and non-disadvantaged learners particularly at intermediate level is reduced by apprenticeships, and the gap even closes with progression to higher levels. The impact of COVID-19 meant that companies were furloughing employees or making staff redundant, off-the-job learning was disrupted, and apprentices, already on low pay, faced additional financial strains (Doherty and Cullinane 2020; Roberts 2022).
Apprenticeships: availability
London Councils (2020a) claim the apprenticeship levy is acting as a barrier, citing a 45% decrease in intermediate apprenticeships nationally since the levy was introduced. Kathleen Henehan (2019) confirms this, but disputes that this shift is due to the levy, highlighting that regulatory reforms have led to the decline of lower quality programmes. Henehan (ibid.) contends that apprenticeships have moved from lower-level programmes, usually with lower pay, to higher-level programmes associated with higher pay, as illustrated in Figure 1.Fig. 1 Changes in apprenticeship starts by level, sector and post-apprenticeship pay, England 2014–2018 (Henehan 2019, p. 11).
Notes Bubble size represents the share of apprenticeship starts taken up by each type of programme in 2014/15; pay refers to median pay that apprentices who finished their programmes in 2012/13 received in 2015/16. Newer programmes, including most at Levels 6 and 7, are excluded from this analysis because there were not enough apprentices completing these programmes in 2012/13 to allow for an analysis of post-apprenticeship earnings. UK median annual pay was GBP 23,084 (nominal) in 2015/16 (ibid.)
London has a disproportionately low share of apprenticeship starts (Henehan 2019). The low level of starts, argue London Councils (2020a), is due to factors such as the city’s sectoral composition, a lower supply of Londoners linked to higher progression to higher education, as well as a lower demand from employers. Gadsby (2019a) argues that unlike other regions, disadvantaged young people in London are much less likely to start an apprenticeship. This is not surprising considering there are not many potential “matches” on offer for them here. Henehan reveals that “higher-level programmes in better paid sectors (ICT, accounting and business management)” actually “grew more in London than in any other place” in the period 2014/15 to 2017/18 (Henehan 2019, p. 14). This reflects the broader shift to higher-skilled work in the London labour market. How this plays out in Hounslow is presented below, in the “Background” section.
The challenges and barriers for young people: low qualifications
The Institute for Public Policy Research (Pullen and Dromey 2016) asserts that many of the difficulties young people with low qualifications encounter in their quest to secure work can be attributed to 16–18-year-olds' phase of education: “Young people who leave full-time education with a level 2 qualification [have a lower employment rate] than those of their peers who leave full-time education with a level 3 qualification or higher education” (ibid.), with 20 percentage points difference between Level 2 and Level 3 learners, and only 39% of Level 2 learners aged 17 progressing into Level 3 (ibid.). There is also evidence of the “cycles” of low qualifications that young people become stuck in, such that a quarter of Level 2 learners aged 17 are at the same level a year later. Furthermore, London Councils (2020a) add that there is a distinct link between the level of educational attainment of young Londoners and their employment outcomes. They observe that a quarter of young people in London do not achieve five GCSEs (i.e. Level 2) by age 18, and almost half are without Level 3 qualifications by age 18 (ibid.).
The challenges and barriers for young people: circumstances and background
London Councils (2020a, pp. 10–11) point to external socioeconomic factors, asserting the impact of personal circumstances for the learning and progress of young people. They suggest that factors related to disadvantage, in terms of family income and deprivation, influence outcomes for young Londoners. Impetus (2020) point to the young people not in education, employment or training (NEET), referring to a “noticeable difference in NEET rates between young people from disadvantaged backgrounds and their better-off peers” (ibid., p. 28). In addition, those who are “doubly disadvantaged”, coming from disadvantaged backgrounds and having low qualifications, are most likely to be NEETs when aged 18–24 (Roberts 2022), the age group featured in our own research sample. Conversely, Solveig Ose and Chris Jensen (2017) found that health and social problems act as barriers to employment for young people, observing that mental health problems amongst young people often disguise social problems. London Councils (2020a) point to inequalities and social exclusion as barriers to employment for young people, citing Gadsby’s (2019b) finding that qualifications can only be attributed to 50% of the overall employment gap for young people. They propose that social exclusion accounts for the remaining gap.
The literature reviewed suggests that there are a range of structural labour market challenges which have a bearing on both the number but also the quality of apprenticeship opportunities open to young people. Whilst a range of barriers relating to individual level factors are highlighted in the literature, these need to be explored in relation to the experiences of young people in apprenticeships. Thus, to enable conclusions to be drawn about how the post-COVID-19 labour market will impact on and inform local authority recovery strategies for youth employment, the collection of primary data is required.
Background: the London Borough of Hounslow
The London Borough of Hounslow stretches from the Western fringes of inner London to the edge of outer London. Qualification levels of young people educated in Hounslow are on par with London, and some indicators fare much better, with 74% Level 2-qualified with maths and English, and 68% Level 3-qualified by age 19 in 2019 (London Councils 2020c). However, London Councils (2020a) suggest that even those who acquire mid-level qualifications are struggling to enter the labour market. The competitive labour market creates a “bumping down” effect, with those with higher qualifications to undertake jobs being shut out. Gadsby (2020) asserts the significance of GCSE maths and English qualifications for outcomes regardless of background. It could be argued that in the context of broadly strong educational performance amongst Hounslow young people, it is the structural issues within the London labour market which may present particularly problematic challenges to overcome in trying to enter the job market (Wolf 2011).
In Hounslow, the number of apprenticeship starts have been largely in decline, even prior to the introduction of the levy in 2017, and after a brief rise in 2019, this decline was exacerbated by the pandemic (see Figure 2). This decline is spread across all age groups (16–18, 19–24 and 25+); see Figure 3, and qualification levels (intermediate, advanced and higher); see Figure 4.Fig. 2 Hounslow apprenticeship starts, 2010–2020 (London Councils 2020b)
Fig. 3 Hounslow apprenticeships by age, 2018–2020 (London Borough of Hounslow 2020c)
Fig. 4 Hounslow apprenticeship starts by level, 2018–2020 (London Borough of Hounslow 2020c)
The decline in Hounslow apprenticeship starts by level between 2018 and 2020 (shown in Figure 4) suggests that there is a reduced number of opportunities to compete for, but also a barrier to accessing opportunities allowing entry to better-paid work. Gadsby (2019a) argues that while disadvantaged young people with good qualifications are just as likely as their better-off peers to start an apprenticeship, amongst young people without good GCSEs the better-off peers are more likely to start an apprenticeship. Being disadvantaged is not in itself a factor, but the impact of qualifications seems to be of key importance. The Sutton Trust (Fuller et al. 2017) argues that the economically disadvantaged are underrepresented in access to apprenticeships, thus, disadvantaged background does seem to impact the level of apprenticeship undertaken (Gadsby 2019a; Wolf 2011).
The apprenticeship pledge made by the London Borough of Hounslow in its Corporate Plan 2019–2024 (London Borough of Hounslow 2019) set out a commitment to “create 4,000 new apprenticeships and training opportunities to help young people into work” (ibid., p. 2). Figure 5 shows the number of Hounslow apprenticeship starts by sector and levels 2019–20.Fig. 5 Hounslow apprenticeship starts by sector and levels 2019–20 (London Borough of Hounslow 2020a)
In the year 2020–2021, there was a drop to just 300 apprenticeship starts (ONS 2021; London Borough of Hounslow 2020a). This decline was also seen in national apprenticeship starts, which were reported to have dropped by 45.5% (ONS 2021) between the onset of the COVID-19 pandemic in March 2020 and July 2020, compared to the same period in 2018/19 (Roberts 2022). Learners aged 25 were overrepresented in these starts at 61.6%, compared to 54.5% in the previous year, and 46% of London businesses stated they did not intend to fully exploit their apprenticeship levy funding, planning only to use up to half of the funds available in the next 12 months (Rowe and Newbold 2021).
This article aims to understand the barriers young people (18–24-year-olds) in apprenticeships face post-pandemic by exploring the perspectives of the training providers, the employers, and the young apprentices themselves.
Methodology
The research we present here is centred on understanding the experiences of young people in apprenticeships (18–24-year-olds). Aiming to gain insight and understanding regarding the hindering and supporting factors young people faced in securing apprenticeships before and during the COVID-19 crisis, we employed a small-scale qualitative research strategy. By working with a smaller sample, this strategy facilitated a richer theoretical perspective than large-scale research, enabling us to focus on contextual circumstances (Saunders et al. 2016), and complement other similar qualitative studies such as those by Fuller and Unwin (2003), Brockmann and Laurie (2016) and Brockmann, Laurie and Smith (2020). Thus, hearing the voices of young people, specifically current apprentices, was crucial, combined with other social actors involved in this process, such as employers and an apprenticeship training provider.
As a limitation, the initial intention of our study was to understand the experiences of young people moving into apprenticeships in two job sectors, IT and healthcare. However, during the data collection period (January 2021–March 2021), the British government implemented an unanticipated COVID-19 national lockdown, not only impacting our research in practical terms, but also adding another factor likely to exacerbate the decline in apprenticeships.
Faced with this situation, we adapted our sampling to include other actors with insight into apprenticeships, namely, employers and apprenticeship training providers. We restricted the age range of apprentices to 18–24 years, in line with the broad definition of youth employment and unemployment from the Office for National Statistics (ONS 2020) and other “young people” research (Wolff et al. 2020). Furthermore, on the basis that the purpose of our study was to understand the experiences of young people entering the labour market via apprenticeships, we included Level 2 or Level 3 apprenticeships as a criterion. This was to restrict the sample to apprentices in entry-level positions, and indeed employers and training providers who recruited, employed or trained this particular group. We selected our interviewees from a group within an organisation one of us already had access to, using homogeneous purposive sampling (Saunders et al. 2016, p. 717), and the organisation confirmed access permission. Informed consent was obtained from all individual participants included in the study, and confidentiality was assured, removing any identifiable information from transcripts. Details of the sampling and samples can be seen in Table 1.Table 1 Sampling and samples
Sample subgroups Inclusion criteria
Apprentices • Aged 18–24
• Enrolled in a recognised apprenticeship programme
• Employed as an apprentice by an NHS trust or NHS primary care provider*
• Undertaking a Level 2 or a Level 3 apprenticeship
• Living or working in Hounslow or one of the surrounding boroughs/counties
Employers • NHS trust or NHS primary care provider organisation
• Employing apprentices who were
• undertaking Level 2 or Level 3 apprenticeships
• and aged 18–24
• Organisation located in Hounslow or one of the surrounding boroughs/counties
• Staff member role encompasses recruiting and/or supporting apprentices
Training providers • Registered apprenticeship training provider
• Commissioned by NHS trusts or NHS primary care providers to deliver apprenticeship training for NHS trusts (health sector) located or working in Hounslow or one of the surrounding boroughs/counties
• Recruiting and training apprentices who were
• undertaking Level 2 or Level 3 apprenticeships in NHS
• and aged 18–24
*NHS = National Health Service; NHS trusts = acute (hospital) trusts providing hospital-based NHS services; mental health trusts (offering mental health and social care services); or community trusts (providing services such as district nursing, physiotherapy and speech and language therapy); NHS primary care providers = independent businesses, such as general practitioners’ or dentists’ practices, or opticians, offering NHS services
Since our study was qualitative by design and sought to gather rich and detailed data at an individual level, we selected two apprentices, two employers and one training provider. We developed three interview guides for semi-structured interviews (Bryman and Bell 2015) to capture each participant subgroup’s perspectives on factors hindering or supporting entry of young people into apprenticeships. We chose this approach to accommodate focus, but also flexibility to probe answers, providing a deeper understanding of the meaning participants attributed to phenomena, and thus adding depth to data obtained (Saunders et al. 2016, p. 394). The interviews, which were conducted online, lasted between 45 and 60 minutes and were video-recorded for later analysis. Table 2 provides an overview of our interviewees in the order they were interviewed.Table 2 Interviewees
Interview order Participants Anonymised references
1 Apprentice A1
2 Training provider TP
3 Employer E1
4 Employer E2
5 Apprentice A2
We transcribed the audio-visual recordings of the interviews, anonymising all identifying information, and double-checked the result for accuracy by watching and checking against the transcript. Our thematic analysis of the qualitative data collected from participant interviews involved identifying and coding themes or patterns occurring across the data set (Saunders et al. 2016). We built up a list of codes starting with “in vivo” terms used by participants themselves during the interviews, and developed additional codes to describe data (ibid.). We continued this process by adding new codes, and developing definitions of these. We used a constant comparison approach, in terms of rereading and reviewing earlier transcripts and recordings.
Once we had completed a full list of codes from our data, we reviewed this list to identify relationships and patterns and develop broader themes under which codes could be categorised. We then further refined these to define the themes and relationships between them, establishing main themes and sub-themes (ibid.). The final stage of the process involved refining themes further as well as the relationships between them. We then subjected the thematic analysis of each sample subgroup to a comparative analysis, comparing the themes and codes across all thematic analyses. The purpose of this was to identify the degrees of similarity and difference in different subgroups’ perceptions of the factors that hinder or support young people’s entry into sustained apprenticeships.
Findings
In this section, we present our findings in three subsections, beginning with the thematic analysis of the apprentices’ perspective. This is followed by thematic analyses of the employers’ and the training provider’s perspectives, respectively.
Apprentice thematic analysis
Table 3 summarises the number of references made by the apprentices to each of the themes and sub-themes we identified in our data analysis. These participants identified significantly more supportive factors (57 references) than hindering factors (33 references), with three distinct sub-themes emerging from the analysis: (1) competition within the labour market; (2) personal attitudes and behaviours; and (3) solutions to challenges – supportive relationships to sustain apprenticeships and progress towards professional employment.Table 3 Apprentice thematic analysis
Themes and subcategories Sub-themes Total no. of references
Hindering factors 33
Individual circumstances 4
Influence of family and socioeconomic context 3
Need for financial security 1
Individual level factors 4
Lack of access to advice on apprenticeships at school 2
Lack of work experience 2
Local labour market 25
Competition for vacancies 20
Lengthy NHS recruitment process 1
Shortage of available opportunities 4
Supportive factors 57
Individual background 12
Cultural values and expectations 6
Family support and influence 5
Peer support and influence 1
Individual level factors 10
Access to careers advice on apprenticeships (school) 2
Achievement of maths and English GCSEs 5
Experience of work and skills development through volunteering, placements and networking 3
Personal attitudes and behaviours 35
Having personal goals to cope with uncertainty 1
Overcoming prior challenges 4
Personal drive 2
Positive mindset 11
Resilience 5
Self-belief 3
Self-motivation and purpose 9
Solutions to challenges 8
Creating awareness of opportunities through role models 3
Building supportive relationships to sustain apprenticeships and foster progress towards professional employment 5
Theme 1: Competition within the labour market
Among our apprentice interviewees, competition within the labour market had the greatest number of references (20 out of 33) as a hindering factor. Both apprentices described large volumes of unsuccessful applications, with A1 reporting hundreds of applications made and A2 describing spending 6 months following a traineeship searching and applying for apprenticeships. The largest share of references were made by A1 who had applied for apprenticeships during the first 5 months of the COVID-19 pandemic and attributed competition to decreased opportunities available to them as the pandemic impacted the economy and number of vacancies available to them locally. The impact described was that in their attempt to adapt to competitive conditions, candidates had resorted to lowering and broadening their applications for apprenticeships.“It went from IT Level 4 to the sort of Business general aspects of IT, such as Business Administration Level 3. As time was going, my patience was decreasing, there was that desperate need to at least have something” (A1).
Feedback from employers was described by participants as citing lack of experience or competition as key reasons for unsuccessful applications.
Theme 2: Personal attitudes and behaviours
The apprentices identified personal attitudes and behaviours as a key supportive factor, with 35 references alone within the subcategory’s seven sub-themes, which included, for example, positive mindset (11 references). They stated that these characteristics helped them to mitigate challenges on their journey to becoming an apprentice. They described how prior challenges and adversity had led to their development of resilience, self-motivation, self-belief and a positive mindset. Their experience is helpful towards a deeper understanding of what apprentices experience, and it resonates with other reports such as the one prepared by Kristin Wolff et al. (2020). For A1, this experience included overcoming a speech impediment, whilst A2 cited growing up in adverse socioeconomic conditions, including lack of parental support and leaving home at the age of 15.“It wasn’t like my family was encouraging me … I’m going to apply for this apprenticeship … why would you do that? … because I want to succeed, and I want to have nice things in life and own my own property and do these things that people in my family haven’t achieved before … I learnt the enthusiasm and I learned the drive because no one ever showed me I had to go out and find it really” (A2).
Theme 3: Solutions to challenges – supportive relationships to sustain apprenticeships and progress towards professional employment
Supportive relationships with employers and providers helping apprentices to sustain their apprenticeship and to progress towards professional employment kept emerging from participant A2 as a significant factor, despite being outside the initial focus of our interviews and our study. Describing the impact of this factor, A2 stated, “So, for me it was finding people that believed in me. Which … in the main is hard to come by, but it was really nice”. Having progressed from a Level 2 into a Level 3 apprenticeship, participant A2 perceived supportive relationships to be an important factor in guiding young people into apprenticeships.
Employer thematic analysis
Table 4 summarises the number of references made by the employers to each of the themes and sub-themes we identified in our data analysis. The three distinct sub-themes emerging from the employers’ responses are: (1) organisational barriers to creating entry-level opportunities; (2) the influence of stigma; and (3) skills and attributes of young people – maths and English qualifications to enter and sustain apprenticeships and progress towards professional employment.Table 4 Employer thematic analysis
Themes and subcategories Sub-themes Total no. of references
Organisational barriers in creating entry-level opportunities 17
Funding model 5
Funding salaries on tight budgets a hurdle to driving demand from services 4
Inflexibility of levy funding for wages compared to previous funding model 1
Influencing service managers 3
Demonstrating added value for services with no prior engagement 2
Structural and geographical challenges of messaging 1
Workforce strategy 9
Decentralised model resulting in variability of engagement 3
Disconnect with levy use 2
Focus on upskilling 1
Lack of demand from services 1
Senior leadership commitment to drive strategy 3
Skills and attributes of young people 19
Maths and English skills to enter and sustain apprenticeships and foster progress into professional employment 13
Socialisation skills to present self and cope with challenges 6
Influence of stigma 11
Credibility of higher-level apprenticeships amongst applicants 1
Manager perceptions of apprenticeships 4
Manager stereotypes of young people 1
Sectorial value of work-based learning 1
Socioeconomic background and perceptions of apprenticeships 4
Market conditions 4
Competition displacing less able applicants 2
Lack of demand from managers to meet potential supply 2
Recruitment process 3
Bureaucratic process, creating barriers for young people 2
Unrealistic employer expectations for entry roles 1
Solutions to overcome challenges 19
Education system change 1
Embedding workforce strategy encompassing apprentices 3
Establishing support structure to sustain and progress is key 9
First refusal on entry-level roles for apprenticeships 1
Levy subsidising apprentice wages to drive demand 3
Using success stories to influence managers 2
Theme 1: Organisational barriers to creating entry-level opportunities
For our employer interviewees, the creation of apprenticeship opportunities was a key challenge linked to the lack of demand from managers, rather than supply issues. They described this issue in relation to three interlinked subcategories: funding, influence and workforce strategy.“They haven’t got money to throw around and they are price salary conscious … Identifying a role is one thing, agreeing to recruit to it is another, OK, and so there is an issue … They could do with many more people to do the service, but they can’t afford it so that it’s a juggling act …” (E2).
Influencing managers to create roles was described as problematic. This was not only due to the geographical and structural difficulties in developing influencing relationships, but also attributed to pre-levy salary funding initiatives providing an additional hurdle to establish value for services.“I’ve got to demonstrate that added value more, but how do you demonstrate the added value if you can’t get them in to have the person in the first-place kind of place” (E1).
These conditions were described as hindering the creation of opportunities, both in terms of the financial investment required, which discourages the use of the levy, but also more broadly in terms of the lack of a centralised workforce strategy driving a corporate commitment and strategy for apprenticeships.“I’ve got places in in different services where things could be thriving, and like I could demonstrate we’ve got a lot of apprenticeships, but then in other areas they might not have anything. But in general, workforce planning isn’t that it’s very much bespoke to each service and they are responsible for their own sort of plans and actions” (E1).
Theme 2: Influence of stigma
Another significant theme emerging from our interviews with employers was the stigma attached to apprenticeships by managers. Participants described this stigma as relating to credibility, influencing demand in terms of the creation of opportunities:“[there] is a bit of a stigma about apprenticeship. Still, I think if they’d been called something different, maybe like, I don’t know, work-based learning qualifications or something. I think by having the old apprenticeship tag when they’re very different to how they used to be. Yeah, I think that might have given them a different credibility” (E1).
Theme 3: Skills and attributes of young people – maths and English qualifications to enter and sustain apprenticeships and progress towards professional employment
Finally, our employer interviewees described the lack of maths and English qualifications at Level 2 as impacting sustainability of apprenticeships and a barrier to progressing into higher-level payrolls.“The barrier around qualifications is, I think sometimes it’s around and honestly not set someone up to fail. … so is it fair to say come and join us will get you through, if actually they haven’t got some of the foundation skills that they’re going to need. So, whether the qualification for the individual is truly a barrier or not, depends on the approach” (E2).
Training provider thematic analysis
Table 5 summarises the number of references made by the training provider to each of the themes and sub-themes we identified in our data analysis. The three distinct sub-themes emerging from this analysis are: (1) Organisational barriers to creating entry-level opportunities; (2) stigma; and (3) solution to challenges – developing support infrastructure to enable sustainability and progress.Table 5 Training provider thematic analysis
Themes and subcategories Sub-themes Total no. of references
Organisational barriers to creating entry-level opportunities 18
Funding 4
Conflicting funding initiatives favour employment over apprenticeships 1
Procurement bureaucracy impedes speed of process 3
Influence 11
Critical influence of the apprenticeship lead on apprenticeship strategy 6
Lack of support infrastructure creates vicious circle for managers 2
Lack of understanding of added value new apprentices bring to organisation 3
Workforce strategy 3
Focus on upskilling strategy for incumbent staff 1
Lack of demand from services to meet supply of potential applicants 2
Skills and attributes of young people 6
Maths and English qualifications key to sustain apprenticeships and foster progress towards professional employment 5
Presenting employability attributes and behaviours 1
Stigma 13
Calibre of applicants for apprenticeships versus jobs 5
Manager perceptions of young people 3
Manager perceptions of value of apprentices and apprenticeships 5
Solutions to challenges 5
Developing support infrastructure to enable sustainability and progress 3
Educating managers on added value of new apprentices 1
Implementing apprenticeship first refusal model for all new entry-level posts 1
Theme 1: Organisational barriers to creating entry-level opportunities
The training provider identified a lack of apprenticeship opportunities for new entrants as a key issue, stemming from challenges related to funding, influence, and workforce strategy. National funding initiatives focusing on recruitment were described as often displaced levy funding, creating jobs instead of apprenticeships and thus impacting the number of available apprenticeship opportunities.“… it’s not that we’ve necessarily experienced a shortage of applicants, it’s a shortage of positions to shortage of apprenticeship roles … focusing on healthcare, there’s often a conflict, so there’s often other initiatives going on to get workers … that will bypass the apprenticeships route” (TP).
Notably, significant emphasis was on influence and the pivotal role of the apprenticeship lead influencing perceptions and the adoption of apprenticeships by managers, with the largest number of references across all sub-themes:“often you’re interacting with decision-makers that are in these apprenticeship lead roles, so to speak, who have no experience of apprenticeships … again, we can talk about stereotyping, they may have a particular view, and ultimately decide how apprenticeships is [sic] sold within that organisation. However, this will often depend on the outlook of that individual and how they’re selling it to their peers and colleagues” (TP).
Influencing manager decisions was described as correlating to support infrastructures for apprentices, acting as a significant barrier to influencing managers’ decisions to create opportunities.“You know and then if the infrastructure isn’t there to support them, and essentially that apprentice has a bad experience and leaves prematurely, that manager’s going to have a dim view on apprentices. So, the apprentice is dissatisfied and this creates a vicious circle” (TP).
The training provider reported that apprenticeships were often viewed by employers as a workforce strategy to upskill substantive staff opposed to new entrants into a sector, again influenced by the apprenticeship lead.
Theme 2: Stigma
Another significant theme emerging from our interview with the training provider was the stigma attached to apprenticeships (13 references), with our respondent identifying this as being associated with the negative stereotyping of young people in society more generally.“I think it’s the stigma that comes along with the label of an apprentice. Again, people potentially view that as it’s a perhaps a younger person, potentially unemployed, fresh out of school, very few qualifications to their name. It’s just you know, it’s an impression people create, which isn’t a true reflection of the calibre of people that would apply for apprenticeships” (TP).
Negative stereotypes and an outdated understanding of apprenticeships more broadly were thus perceived as permeating into decisions about levy use and creation of entry-level apprenticeships:“if you could just get rid of that stigma … associated [with] apprenticeships, particularly with older workforce. You know, that plays such a big part in … how, say, the levy is utilised and whether you see it sort of more ring-fenced for higher level apprenticeships for substantive staff, or whether actually using it in the right way and bringing in new blood into the organisation” (TP).
Theme 3: Solution to challenges – developing support infrastructure to enable sustainability and progress
Finally, our training provider interviewee described a lack of support as a significant barrier for sustainability, underlining the broader issue of entry to sustainable apprenticeships.“My experience of working … bringing in candidates at entry level is that when they’ve done it, it’s been done poorly. So, they’re not necessarily getting, when they’re in the apprenticeship [apprentices], the support they need, so they often end up being leavers, so they won’t complete apprenticeships” (TP).
Comparative thematic analysis: differences in themes
Comparing the key themes emerging from all of our interviews with apprentices, employers and the training provider, we find that “individual circumstances” was the only theme identified solely by apprentices as a supportive factor (Table 6). It is conceivable that employers’ and training providers’ awareness of young people’s individual circumstances such as socioeconomic background and family support is likely to be very limited at apprenticeship application stage and thus may account for such a degree of difference. Another striking phenomenon shown in Table 6 is that individual level factor themes (lack of advice on apprenticeships at school, lack of work experience) emerge as important within the analysis, scoring almost double the number of references recorded for all other themes (72 references; of which 47 were made by the apprentices).Table 6 Comparative thematic analysis (all themes)
The degree of difference in perceptions of individual level factors is shown in Table 7, with apprentices overwhelmingly identifying factors as supportive (43 references), whilst employers and training provider identified individual factors as solely a hindering factor (19 and 6 references, respectively). Notably, the diverging perceptions in individual level factors can be attributed to apprentices identifying personal attitudes and behaviours as positive characteristics needed to navigate the process of securing an apprenticeship. It is conceivable that apprentices were more likely to reflect on their personal attributes and the role of these in their personal journeys. Furthermore, Table 7 shows sub-themes that were mentioned by only one participant subgroup.Table 7 Comparative thematic analysis (individual factors)
The degree of difference between perceptions of apprentices compared to those of employers and training providers highlights that each group of participants were more likely to identify factors from their frame of knowledge and perspective. Thus, apprentices identified more individual level factors and fewer organisational factors than employers and the training provider.
Comparative thematic analysis: similarities in themes
Table 8 shows the similarities in themes we found by way of comparative thematic analysis. Maths and English qualifications at Level 2 emerged as a key individual level factor across all participant groups. Whilst apprentices identified this factor as being significant for entering an apprenticeship, employers and training providers placed importance on these qualifications for young people to sustain an apprenticeship and progress beyond it into professional employment. In the local labour market thematic category, competition was a factor identified by apprentices and employers, but the latter highlighted the context of this as displacement of less able candidates, likely an insight garnered from a recruiting perspective as an employer. The only consistent factor across all three groups of interviewees was the creation of a support structure in the thematic category of solutions to challenges to help sustain apprenticeships and guide apprentices into professional employment.Table 8 Comparative thematic analysis (similarities)
Discussion
To understand what stands in the way of young people’s entry into an apprenticeship and what makes it easier, it is instructive to identify the hindering and the supportive factors (Brockmann et al. 2020; Roberts 2022). Within our analysis of the responses from the apprentice subgroup, the factors emerging as overarching themes were related to individual epistemologies and demographics. Our findings suggest that there are structural barriers (such as competition and a bureaucratic application process) and organisational barriers (such as funding, managers’ prejudices, etc.) which serve to limit access to opportunities. We also found, however, that individual level factors, such as young people’s (lack of) personal capabilities and (lack or low level of) maths and English qualifications, can also hinder young people’s entry into apprenticeships (Roberts 2022). These deeper insights into their experiences are helpful in understanding and addressing the challenges young candidates hoping to enter an apprenticeship are struggling to overcome.
Competition in the labour market
Competition may be an entry barrier to apprenticeships for young people, in part linked to displacement by more able applicants (Roberts 2022). Employer participants in our study did identify competition as a factor linked to more able applicants and demand issues, of which the latter is elaborated on in the organisational barriers section below. “Polarisation”, discussed by Anne Green (2020) and Luc Benda et al. (2019), displaces apprentices in competitive contexts, confirming Spence’s signalling theory and Thurow’s job competition theory (Benda et al. 2019) in relation to the signals of apprentices being weaker in the context of wider competition in the labour market, thus disadvantaging them (Roberts 2022). Following completion of an IT Level 3 qualification, participant A1 initially sought a Level 4 IT higher apprenticeship to pursue a career pathway in IT, but ended up in a much broader Level 3 Business Administration apprenticeship. Thus, we can confirm that competition is a significant factor in entry to apprenticeships for young people.
Personal attitudes and behaviours
To navigate the process of applying for and securing an apprenticeship, young people need positive personal characteristics. Apprentices described resilience, self-motivation, self-belief and a positive mindset as personal attitudes and behaviours that enabled them to overcome barriers they faced on their journeys. This aligns with the work-ready capabilities other research found to be supporting young people to enter and sustain work (Wolff et al. 2020). Mindsets impact the cognition, emotions and behaviours of individuals in contexts of achievement (Heslin and Keating 2016). A positive mindset would enable apprentices to regulate these to support achievement of objectives, fostering the central role of individual agency in reaching personal goals (Schoon 2020). Apprentices perceived these characteristics as enabling them to mitigate challenges they faced, including structural barriers such as competition within the labour market, and individual circumstances such as lack of family support and socioeconomic background hindering their progress. This finding contradicts previous research identifying socioeconomic disadvantages as limiting the ability of young people to exercise individual agency (Schoon 2020; Ng-Knight and Schoon 2017).
Organisational barriers to creating entry-level opportunities
Competition is also exacerbated by a lack of available apprenticeships. The creation of entry-level apprenticeships is impeded by organisational challenges, thereby limiting the quantity of opportunities available to young people. Interrelated issues of the apprenticeship levy funding model negatively influencing managers and workforce planning were identified by participants as key issues in terms of restricting use of the levy and consequently the creation of apprenticeship opportunities. This corresponds to a national issue of unspent apprenticeship funds, which Unison, one of the UK’s largest trade unions, suggests stands at 79% (Unison 2019, p. 1). Participants mentioned funding of salaries and associated backfill costs for off-the-job-learning as a key obstacle, with lack of flexibility in the levy model to support these costs acting as a further barrier. These factors are related to the issue of the levy funding model negatively impacting availability of entry-level opportunities discussed by London Councils (2020a, pp. 10, 18). Participants established these challenges around workforce planning in relation to degree of influence, taking into consideration lack of funding and resulting lack of demand from services for apprenticeships. Stephen Billett (2016) emphasises the importance of immersion in work communities. In their qualitative studies, Michaela Brockmann et al. (2020) identified minimally invested employers and partially engaged employers as barriers to quality apprenticeship experiences.
Stigma
Stigma may play a role in determining the decisions made by managers whether or not to both create apprenticeships and recruit young people into then, and thus presents another potential barrier to for young people seeking to enter apprenticeships. Brockmann and Laurie (2016) suggest that research indicates that employers may place low importance on candidates’ “technical” knowledge. Our respondents described stigma as a label attached to apprentices and relating to the negative stereotype of young people as “anti-learning” (Brockmann and Laurie 2016, p. 231). This aligns with the definition of stigma provided by Bruce Link and Jo Phelan as co-occurrence of “elements of labeling, stereotyping, separation, status loss and discrimination” in conditions where power is exercised (Link and Phelan 2001, p. 377). Managers exercise power through decisions they make, be it to upskill staff through an apprenticeship or to recruit and train a new staff member through an apprenticeship. Our respondents confirmed that there is a stigma attached to apprenticeships, which Terence Hogarth et al. (2012) suggest relates to being considered as a lower-status pathway into the labour market, while Anna Mazenod identifies the wider system status of apprenticeships as associated with the segmentation between academic and vocational orientations in the education system (Mazenod 2016; Brockmann and Laurie 2016). Lower-quality programmes have declined, as discussed by Henehan (2019) and there has been a shift to higher-level and higher-pay programmes. Despite numerous “think tank” reports seeking to address parity of esteem (VOCEDplus 2016; Cedefop 2011), the negative perceptions, in the UK and beyond, continue to pervade despite policy level changes to improve the quality of the apprenticeship pathway (Gray and Farrell 2021).
Maths and English qualifications to enter and sustain apprenticeships and foster progress towards professional employment
Maths and English qualifications at Level 2 can act as a supportive factor to enable young people’s entry into apprenticeships. However, in accordance with apprenticeship funding rules highlighted by Stephen Evans (2020, p. 45), maths and English (“functional skills”) constitute a mandatory requirement for Level 3 apprentices to achieve Level 2 functional skills. The significance of sustainability for young people entering apprenticeships, and the need to achieve the requisite qualifications, can hinder young people’s completion of and progression through the apprenticeship pathway, as they may not be academically able. Thus, maths and English qualifications are a factor for entry into sustainable apprenticeships.
Supportive relationships to sustain apprenticeships and foster progress towards professional employment
Supportive relationships with employers and providers are another important factor in enabling young people to enter apprenticeships which are sustainable and enable progress. Our respondents underlined the significance of such relationships for the individual epistemologies of young people. Katelyn Herrygers and Stacey Wieland (2017) suggest that early work experience and socialisation shape young people’s perspective of work and their self-efficacy, and thus are key to a positive sense of self and work. Billett (2016) lists these personal epistemologies as “knowing, engaging and learning” (ibid., p. 622).
Conclusion and recommendations
Our findings suggest that structural labour market factors of competition and employer demand are likely to create conditions of limited entry-level opportunities, especially with the added impact of the COVID-19 crisis, whilst individual level factors of positive personal characteristics and maths and English qualifications could act to hinder young people entering apprenticeships. The results of our small-scale qualitative research among apprentices, employers and a training provider in the Borough of Hounslow highlight that sustainable apprenticeships are a major concern, with maths and English qualifications and support during apprenticeships identified as key factors in enabling young people to enter and sustain apprenticeships and progress towards professional employment.
Competition as an existing barrier to young people entering apprenticeships is likely to be exacerbated further in the context of an increased supply of labour due to COVID-19 pandemic impacts on the labour market identified in the literature, for example by the Federation for Industry Sector Skills and Standards (FISSS 2020) and the Institute for Employment Studies (IES 2021). Enabling young people to improve their labour market position through development of skills, as suggested by Gary Becker’s human capital theory (Becker 1964; Benda et al. 2019), will be important to mitigate recent negative impacts. Positive personal characteristics can enable young people to moderate the hindering effect of barriers such as competitive conditions, and as such support young people in building work-ready capabilities as well as job sustainability. Funding mechanisms to target opportunities for young people as discussed by the West London Alliance (WLA 2020) could to some extent counter the impacts of the competitive labour market for young people in Hounslow. Explicitly developing work-ready skills will be a common agenda as revealed by our research among apprentices in the Borough of Hounslow.
Organisational level barriers impact the demand for apprenticeships in terms of restricting the numbers of entry-level opportunities created. As an example, the COVID-19 crisis caused apprenticeship placements to decline in most countries, due to drops in company participation, either temporarily or only in some sectors (Cedefop et al. 2022). In the UK, and among apprentices in the Borough of Hounslow, sector-level challenges in relation to the apprenticeship levy continue to negatively impact both demand and the creation of new entry-level opportunities as a pathway. In the face of the complexity of interrelated issues and the need to understand the specific sector-level factors involved, a focus on growth industries with sustainable high-value jobs is recommended, supporting young people in identifying and accessing sectors that will allow sustainability of apprenticeships and jobs, and progression into better work.
Stigma remains an underlying factor that can also negatively impact the creation of entry-level apprenticeships, as seen among apprentices in the Borough of Hounslow, and thus restrict the numbers of opportunities available to young people in the labour market. Continuing negative perceptions of apprentices, apprenticeships and young people pervade among both managers and the wider public, and are based on outdated stereotypes and information. To effect change on this front, it is important to continue raising awareness of the positive impact on companies and communities of young people as apprentices, and further addressing quality and credibility concerns of apprenticeships will remain important factors in improving the situation (OECD 2021; Cedefop 2022).
Our findings provide a deeper understanding of young people’s experiences in the Borough of Hounslow in terms of the skills and support they need to enter and sustain apprenticeships, and progress towards professional employment. In this context, maths and English qualifications and supportive relationships with employers and training providers have been identified as key factors.
1 The apprenticeship levy is a tax paid by employers with a payroll above GBP 3 million. The money goes into a national fund which is then used to help pay for apprenticeship training costs. Modern apprenticeships combine studying for a qualification with on-the-job experience, so the apprentice learns and works from the first day of the apprenticeship.
2 In England, 15–16-year-olds complete their mandatory formal schooling at intermediate level with the General Certificate of Secondary Education (GCSE). Core subjects include English and maths, but additional choices also include vocational subjects. GCSEs are equivalent to qualification Level 2. Two additional years of formal schooling lead to completion of upper secondary level with an Advanced Level (A Level) certificate; this is equivalent to qualification Level 3. There is a total of nine qualification levels; for more information, see GovUK (n.d.).
3 For more detailed information about apprenticeships in England, see UCAS (n.d.) and Powell (2023).
4 In a nutshell, Spence’s job-market signalling theory posits that education beats acquired skills as a credential, resulting in higher pay for more educated workers (Spence 1973). Thurow’s job competition model argues that “lacking direct evidence on specific training costs for specific workers, laborers are ranked according to their background characteristics – age, sex, educational attainment, previous skills, and psychological tests” (Thurow 1979, p. 17).
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Battiston, A., Williams, R., Patrignani, P., & Conlon, G. (2020). Apprenticeships and social mobility: Fulfilling potential. London: Social Mobility Commission/London Economics. Retrieved 26 June 2020 from https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/894303/Apprenticeships_and_social_mobility_report.pdf
Becker, G. S. (1964). Human capital: A theoretical and empirical analysis, with special reference to education. New York: National Bureau of Economic Research/University of Columbia Press.
Benda L Koster F van der Veen RJ Levelling the playing field? Active labour market policies, educational attainment and unemployment International Journal of Sociology and Social Policy 2019 39 3/4 276 295 10.1108/IJSSP-08-2018-0138
Billett S Apprenticeship as a mode of learning and model of education Education + Training 2016 58 6 613 628 10.1108/ET-01-2016-0001
Brewer, M., Cominetti, N., Henehan, K., McCurdy, C., Sehmi, R., & Slaughter, H. (2020). Jobs, jobs, jobs: Evaluating the effects of the current economic crisis on the UK labour market. London: Resolution Foundation. Retrieved 5 January 2021 from https://www.resolutionfoundation.org/publications/jobs-jobs-jobs/
Brockmann M Clarke L Winch C The apprenticeship framework in England: A new beginning or a continuing sham? Journal of Education and Work 2010 23 2 111 127 10.1080/13639081003627439
Brockmann M Laurie I Apprenticeship in England: The continued role of the academic–vocational divide in shaping learner identities Journal of Vocational Education and Training 2016 68 2 229 244 10.1080/13636820.2016.1143866
Brockmann, M., Laurie, I., & Smith, R. (2020). On-and-off-the-job training in apprenticeships in England. London: Gatsby.
Bryman, A., & Bell, E. (2015). Business research methods (4th ed.). Oxford: Oxford University Press.
Cedefop (European Centre for the Development of Vocational Training). (2011). Vocational education and training at higher qualification levels. Cedefop Research Paper, 15. Luxembourg: Publications Office of the European Union. Retrieved 5 January 2021 from https://www.cedefop.europa.eu/files/5515_en.pdf
Cedefop. (2022). Improving VET image and attractiveness. Cedefop toolkit. Luxembourg: Publications Office of the European Union. Retrieved 24 October 2022 from https://www.cedefop.europa.eu/en/tools/vet-toolkit-tackling-early-leaving/intervention-approaches/improving-vet-image-and-attractiveness
Cedefop, Korovilos, V., & Rustico, L. (2022). Built to last: Apprenticeship vision, purpose, and resilience in times of crisis. Short papers from the Cedefop community of apprenticeship experts. Cedefop Working Paper, 12. Luxembourg: Publications Office of the European Union. 10.2801/449252.
Celani A Singh P Signaling theory and applicant attraction outcomes Personnel Review 2011 40 2 222 238 10.1108/00483481111106093
Clarke, S., & D’Arcy, C. (2018). The kids aren’t alright: A new approach to tackle the challenges faced by young people in the UK labour market. London: Resolution Foundation. Retrieved 2 June 2020 from https://www.resolutionfoundation.org/app/uploads/2018/02/IC-labour-market-policy.pdf
Delebarre, J. (2015). Apprenticeships policy, England 2015: Briefing paper, No. 03052. London: House of Commons Library. Retrieved 20 March 2023 from https://www.the-gtc.co.uk/seecmsfile/?id=679
Doherty, K., & Cullinane, C. (2020). COVID-19 and Social Mobility Impact Brief #3: Apprenticeships. London: The Sutton Trust. Retrieved 24 October 2022 from https://dera.ioe.ac.uk/35606/1/Covid-19-Impacts-Apprenticeships.pdf
Evans, S. (2020). Unleashing talent: Levelling up opportunity for young people. Youth Commission final report December 2020. Leicester: Learning and Work Institute. Retrieved 21 February 2021 from https://learningandwork.org.uk/resources/research-and-reports/youth-commission-report-6-unleashingtalent-levelling-up-opportunity-for-young-people
FISSS (Federation for Industry Sector Skills and Standards). (2020). Briefing note: The Apprenticeship sector’s resilience to COVID-19. Edinburgh: FISSS. Retrieved 24 October 2022 from https://fisss.org/wp-content/uploads/2020/04/FISSS-Briefing-Note-The-Apprenticeship-Sectors-Resilience-to-COVID-19.pdf
Fuller A Unwin L Fostering workplace learning: Looking through the lens of apprenticeship European Educational Research Journal 2003 2 1 41 55 10.2304/eerj.2003.2.1.9
Fuller A Unwin L Apprenticeship as an evolving model of learning Journal of Vocational Education and Training 2011 63 3 261 266 10.1080/13636820.2011.602220
Fuller, A., Unwin, L., Cavaglia, C., McNally, S., & Ventura, G. (2017). Better apprenticeships: Access, quality and labour market outcomes in the English Apprenticeship system. London: The Sutton Trust. Retrieved 23 November 2021 from https://www.suttontrust.com/wp-content/uploads/2019/11/Better-Apprenticeships-1.pdf
Gadsby, B. (2019a). Research Briefing 3: Apprenticeships. London: Impetus. Retrieved 28 November 2020 from https://impetus.org.uk/assets/publications/Research-Briefing-3-Apprenticeships.pdf
Gadsby, B. (2019b). Research Briefing 1: Establishing the employment gap. London: Impetus. Retrieved 2 June 2020 from https://www.impetus.org.uk/assets/publications/Report/Youth-Jobs-Gap-Establising-the-Employment-Gap-report.pdf
Gadsby, B. (2020). Research Briefing 9: The impact of English and maths. London: Impetus. Retrieved 23 November 2020 from https://impetus.org.uk/assets/publications/Research-Briefing-9-The-impact-of-English-and-maths.pdf
GovUK (Government of the United Kingdom of Great Britain and Northern Ireland). (n.d.). What qualification levels mean (dedicated webpage). GovUK. Retrieved March 22, 2023, from https://www.gov.uk/what-different-qualification-levels-mean/list-of-qualification-levels
Gray E Farrell KG Perceptions of culinary education and chef apprenticeships in Ireland Higher Education, Skills and Work-Based Learning 2021 11 1 228 240 10.1108/HESWBL-06-2019-0083
Green, A. (2020). Spatial inequalities in access to Good Work. Lancaster: Work Foundation, Lancaster University. Retrieved 23 November 2020 from http://www.lancaster.ac.uk/media/lancaster-university/content-assets/documents/lums/workfoundation/reports/AnneGreenFinalPaperspring2020forpublication.pdf
Henehan, K. (2019). Trading up or trading off? Understanding recent changes to England’s apprenticeships system. London: Resolution Foundation. Retrieved 9 January 2021 from https://www.resolutionfoundation.org/app/uploads/2019/09/Trading-up-or-trading-off.pdf
Herrygers KS Wieland SMB Work socialization through part-time work: Cultivating self-efficacy and engagement through care Journal of Applied Communication Research 2017 45 5 557 575 10.1080/00909882.2017.1382712
Heslin PA Keating LA Stuck in the muck? The role of mindsets in self-regulation when stymied during the job search Journal of Employment Counseling 2016 53 4 146 161 10.1002/joec.12040
Hogarth T Gambin L Hasluck C Apprenticeships in England: What next? Journal of Vocational Education and Training 2012 64 1 41 55 10.1080/13636820.2011.590221
IES (Institute for Employment studies). (2020). Labour market statistics briefing July 2020. Brighton: Institute for Employment Studies. Retrieved 15 August 2020 from https://www.employment-studies.co.uk/system/files/resources/files/IES%20briefing%20-%20Labour%20Market%20Statistics%20July%202020_0.pdf
Impetus. (2020). Research Briefing 8: The employment gap in London. London: Impetus. Retrieved 2 June 2020 from https://impetus.org.uk/assets/publications/Research-Briefing-8-Report_AW_online-FINAL.pdf
ISC (Industrial Strategy Council). (2019). UK skills mismatch in 2030. London: ISC. Retrieved 1 June 2020 from https://industrialstrategycouncil.org/sites/default/files/UK%20Skills%20Mismatch%202030%20-%20Research%20Paper.pdf
Lan C-C Industry instructors’ perspective on internship implementation strategy Higher Education, Skills and Work-Based Learning 2021 11 3 739 756 10.1108/HESWBL-04-2020-0054
Link BG Phelan JC Conceptualizing stigma Annual Review of Sociology 2001 27 1 363 385 10.1146/annurev.soc.27.1.363
London Borough of Hounslow. (2019). Corporate plan 2019–2024. London: London Borough of Hounslow. Retrieved 10 January 2021 from https://www.hounslow.gov.uk/downloads/file/156/corporate_plan
London Borough of Hounslow. (2020a). Economic Recovery Taskforce initial findings: Impacts and implications (unpublished PowerPoint presentation). London: London Borough of Hounslow.
London Borough of Hounslow. (2020b). Universal credit unemployment figures by age and gender (unpublished internal report). London: London Borough of Hounslow.
London Borough of Hounslow. (2020c). Apprenticeship data (Unpublished). London: London Borough of Hounslow.
London Councils. (2019). Better ways to work: Tackling labour market disadvantage in London. London: London Councils. Retrieved 22 November 2020 from https://www.londoncouncils.gov.uk/our-key-themes/economic-development/better-ways-work-tackling-labour-market-disadvantage-london
London Councils. (2020a). Level up London: Fixing the skills and employment system for young Londoners. London: London Councils. Retrieved 22 November 2020 from https://www.londoncouncils.gov.uk/our-key-themes/economic-development/employment-support/level-london-fixing-skills-andemployment
London Councils. (2020b). Borough insight: Apprenticeship programme starts – All ages. In Intelligent London (interactive online resource). London: London Councils. Retrieved 5 January 2021 from http://www.intelligentlondon.org.uk/borough_insight
London Councils. (2020c). Intelligent London local overview: Hounslow. In Intelligent London (interactive online resource). London: London Councils. Retrieved 5 January 2021 from http://www.intelligentlondon.org.uk/sites/intelligentlondon.org.uk/files/overviews/LA_Overview_HOU_v3.192_2020c1203.PDF
Mayombe C Needs assessment for vocational skills training for unemployed youth in eThekwini Municipality, South Africa Higher Education, Skills and Work-Based Learning 2021 11 1 18 33 10.1108/HESWBL-09-2019-0126
Mazenod A Education or training? A comparative perspective on apprenticeships in England Journal of Vocational Education and Training 2016 68 1 102 117 10.1080/13636820.2015.1117521
Ng-Knight T Schoon I Can locus of control compensate for socioeconomic adversity in the transition from school to work? Journal of Youth and Adolescence 2017 46 10 2114 2128 10.1007/s10964-017-0720-6 28755249
OECD (Organisation for Economic Co-operation and Development). (2021). Developing skills at work: Apprenticeship as an attractive pathway for youth (Flyer announcing virtual event during OECD Youth Week 2021). Paris: OECD. Retrieved 24 October 2022 from https://www.oecd.org/skills/centre-for-skills/Developing-skills-at-work-apprenticeship-agenda.pdf
ONS (Office for National Statistics). (2020). Young people not in education, employment or training (NEET), UK: May 2020. Newport: ONS. Retrieved 1 June 2020 from https://www.ons.gov.uk/employmentandlabourmarket/peoplenotinwork/unemployment/bulletins/youngpeoplenotineducationemploymentortrainingneet/may2020
ONS. (2021). Apprenticeships and traineeships. Gov.UK, 28 January (dedicated webpage; online resource). Newport: ONS. Retrieved 29 January 2021 from https://www.gov.uk/government/statistics/apprenticeships-and-traineeships-january-2021?utm_medium=email&utm_campaign=govuk-notifications&utm_source=7909b326-f272-4e9a-b1fa-5022baee3735&utm_content=immediately
Ose SO Jensen C Youth outside the labour force—Perceived barriers by service providers and service users: A mixed method approach Children and Youth Services Review 2017 81 148 156 10.1016/j.childyouth.2017.08.002
Powell, A. (2023). Apprenticeships policy in England. London: House of Commons Library. Retrieved 20 March 2023 from https://commonslibrary.parliament.uk/research-briefings/sn03052/
Pullen, C., & Dromey, J. (2016). Earning and learning: Making the apprenticeship system work for 16–18-year-olds. London: Institute for Public Policy Research (IPPR). Retrieved 23 November 2020 from https://www.ippr.org/publications/earning-and-learning-making-the-apprenticeship-system-work-for-16-18-year-olds
Roberts K Political economy and young people’s transitions from education-to-work in the UK during and following the 2020 and 2021 lockdowns Journal of Youth Studies 2022 10.1080/13676261.2022.2077641
Rowe, J., & Newbold, P. (2021). London Business 1000: On behalf of London Councils and the London Chamber of Commerce and Industry. London: YouGov/London Chamber of Commerce and Industry/London Councils. Retrieved 29 January 2021 from https://www.londoncouncils.gov.uk/our-keythemes/economic-development/london-business-1000-survey
Saunders, M. Lewis, P., & Thornhill, A. (2016). Research methods for business students (7th ed.). Harlow: Pearson Education.
Schoon I Navigating an uncertain labour market in the UK: The role of structure and agency in the transition from school to work The Annals of the American Academy of Political and Social Science 2020 688 1 77 92 10.1177/0002716220905569
Spence M Job market signaling The Quarterly Journal of Economics 1973 87 3 355 374 10.2307/1882010
Thurow, L. (1979). A job competition model. In M. J. Piore (Ed.), Unemployment and inflation: Institutionalist and structuralist views (pp. 17–32). White Plains, NY: M. E. Sharpe.
UCAS (Universities and Colleges Admissions Service). (n.d.). Apprenticeships in England (dedicated webpage; online resource). Cheltenham: UCAS. Retrieved 20 March 2023 from https://www.ucas.com/apprenticeships/apprenticeships-england
Unison. (2019). It doesn’t add up: The apprenticeship levy and the NHS. London: Unison. Retrieved 20 February 2021 from https://www.unison.org.uk/content/uploads/2019/11/It-doesnt-add-up-report-1.pdf
VOCED plus (Vocational Education and Training Research Database plus). (2016). Focus on Parity of esteem (dedicated webpage; online resource). Adelaide: National Centre for Vocational Education Research (NCVER). Retrieved 24 October 2022, from https://www.voced.edu.au/focus-parity-esteem
Wilson, T., Cockett, J., Papoutsaki, D., & Takala, H. (2020). Getting back to work: Dealing with the labour market impacts of the COVID-19 recession. Brighton: Institute for Employment Studies. Retrieved 9 January 2021 from https://www.employment-studies.co.uk/system/files/resources/files/547a_0.pdf
WLA (West London Alliance). (2020). Build and recover: An economic recovery strategy for West London. London: WLA. Retrieved 2 December 2020 from https://wla.london/wp-content/uploads/2020/09/2020-September-Build-and-Recover-Plan.pdf
Wolf, A. (2011). Review of vocational education: The Wolf report. London: Department for Education. Retrieved 24 October 2022 from https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/180504/DFE-00031-2011.pdf
Wolff, K., Koller, V., & Docto, C. (2020). How apprenticeship programs for opportunity youth stay resilient through the COVID-19 recession. Washington, DC: Jobs for the Future (JFF). Retrieved 25 October 2022 from https://files.eric.ed.gov/fulltext/ED611335.pdf
YFF (Youth Futures Foundation). (2020). Young, vulnerable, and increasing—Why we need to start worrying more about youth unemployment. London: Impetus. Retrieved 1 June 2020 from https://impetus.org.uk/assets/publications/Impetus_YFF_NEET_Report.pdf
|
PMC010xxxxxx/PMC10175903.txt |
==== Front
Ann Hematol
Ann Hematol
Annals of Hematology
0939-5555
1432-0584
Springer Berlin Heidelberg Berlin/Heidelberg
37171596
5266
10.1007/s00277-023-05266-7
Review Article
Bisphosphonate use for glucocorticoid-induced osteoporosis in older patients with immune thrombocytopenia: a clinical perspective
http://orcid.org/0000-0002-6143-7906
Yamasaki Satoshi yamas009@gmail.com
12
1 grid.459691.6 0000 0004 0642 121X Department of Internal Medicine, Kyushu University Beppu Hospital, Beppu, Oita 874-0838 Japan
2 grid.415613.4 Department of Hematology and Clinical Research Institute, National Hospital Organization Kyushu Medical Center, Fukuoka, 810-0065 Japan
12 5 2023
2023
102 7 16451656
4 1 2023
6 5 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Prednisolone, used as a standard initial treatment for immune thrombocytopenia (ITP), is an important risk factor for osteoporosis. Recently, we found that prescription of bisphosphonate during initial loading of prednisolone may prevent reduction in bone mineral density and development of glucocorticoid-induced osteoporosis (GIO) in older patients with ITP receiving prolonged steroid therapy. In this review, I describe the treatment options for older patients with ITP, and present the best practices for screening, evaluating, and diagnosing ITP. I also summarize the literature from 2017 to 2022 on the treatment options for ITP, including discussions on the contraindications and side effects, with an emphasis on GIO, and the relative merits of bisphosphonates as a co-treatment for prevention of GIO. Finally, I present a perspective and an expert recommendation on how older patients with ITP would best be served in the future.
Keywords
Immune thrombocytopenia
Glucocorticoid-induced osteoporosis
Older patients
Bisphosphonate
JSPS KAKENHI22K12887 Yamasaki Satoshi issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcBackground
Glucocorticoid therapy is the standard initial treatment for immune thrombocytopenia (ITP), an autoimmune disease characterized by isolated thrombocytopenia and mucocutaneous bleeding, and has been identified as an important risk factor for osteoporosis [1, 2] (Fig. 1). Glucocorticoids are associated with the risk of bone loss, which is most pronounced in the first few months of use [3]. Recently, strategies that avoid glucocorticoid side effects have been favored, and a strong emphasis has been placed on shared decision making, especially for second-line therapies such as early administration of thrombopoietin receptor agonist (TPO-RA) [4].Fig. 1 Diagram highlighting how risk factors associated with older age may exacerbate the risk of glucocorticoid-induced osteoporosis during treatment of immune thrombocytopenia and how various treatments, including bisphosphonates, are used to mitigate this risk
Glucocorticoids increase the risk of fracture early in the treatment course during the phase of rapid bone loss and at higher levels of bone mineral density (BMD) compared with postmenopausal osteoporosis [5]. An increased risk of fracture was reported for prednisolone doses as low as 2.5–7.5 mg daily [6]. However, the increased risk of fracture in patients taking glucocorticoids declined rapidly during 1 year of therapy [7]. Fracture risk assessment using the 2017 American College of Rheumatology (ACR) guidelines requires the evaluation of clinical risk factors for fracture and BMD [7]. BMD is commonly expressed as the T-score. A T-score ≤ − 2.5 is consistent with a diagnosis of osteoporosis, while a T-score between − 1.0 and − 2.5 is classified as low bone mass (osteopenia) and a T-score ≥ − 1.0 is normal [8]. For patients aged 40–90 years, data for clinical risk factors and BMD can be entered into a validated algorithm (FRAX®; available online at www.sheffield.ac.uk/FRAX/tool.jsp), which calculates the 10-year probabilities of major osteoporotic (clinical vertebral, hip, humerus, wrist) and hip fractures.
Significant negative correlations for BMD with total and mean daily steroid doses were reported for glucocorticoid-induced osteoporosis (GIO) in patients with ITP scheduled to receive long-term steroid treatment, and bisphosphonate was identified as an effective agent for prevention and treatment of GIO [9]. We recently investigated the prevention of GIO-related bone fractures in older patients with ITP receiving prolonged steroid therapy [10]. To fill the evident gap in the management of older patients with ITP, I conducted a summary of the recent literature on GIO occurring secondary to ITP treatment in older adults.
Primary and secondary immune thrombocytopenia
ITP is a diagnosis of exclusion, being defined as isolated thrombocytopenia (peripheral blood platelet count < 100,000/μL) without anemia or leukopenia and without another apparent cause of thrombocytopenia [2, 4]. Primary ITP is ITP acquired through autoimmune mechanisms leading to platelet destruction and platelet underproduction and is not triggered by an apparent associated condition. A presumptive diagnosis of primary ITP is reached when the history, physical examination, and laboratory test findings, including a review of peripheral blood smear findings, do not reveal other potential etiologies for thrombocytopenia.
Secondary ITP, comprising approximately 20% of all ITP cases, is ITP associated with other conditions like autoimmune diseases, such as systemic lupus erythematosus (SLE), antiphospholipid syndrome, and immune thyroid diseases, lymphoproliferative diseases, such as chronic lymphocytic leukemia and autoimmune lymphoproliferative syndrome, and viral and bacterial infections, such as Helicobacter pylori infection, human immunodeficiency virus (HIV) infection, hepatitis C virus (HCV) infection, cytomegalovirus infection, varicella zoster virus infection, and coronavirus disease.
The incidence of primary ITP in adults ranges from 1.6 to 3.9 per 100,000 people/year. After a first peak in childhood, a second peak is observed in adults aged > 65 years, reaching nine per 100,000 people/year in men aged > 75 years [11, 12].
Treatment options for primary and secondary immune thrombocytopenia
First-line therapy
Although not all patients with thrombocytopenia require therapies to increase the platelet count, these treatments are used in patients at increased risk of bleeding based on a number of factors including platelet count, comorbidities, patient values and concerns, and tolerability of therapies [13]. The majority of ITP patients with platelet count < 20,000/μL, especially those with platelet count < 10,000/μL, receive these treatments, even if they have no bleeding symptoms, because of the increased risk of bleeding for platelet count < 20,000/μL and even greater risk for platelet count < 10,000/μL. The majority of ITP patients with minor bleeding, particularly mucosal bleeding including blood blisters in the mouth, and platelet count < 50,000/μL are also treated. Data to support an association between mucosal bleeding and more serious bleeding are limited, and a retrospective study on 112 ITP patients who presented to the emergency department with bleeding and platelet count < 20,000/μL found that only six patients with oral mucosal bleeding subsequently developed more serious bleeding [4].
Glucocorticoids and intravenous immunoglobulin
Glucocorticoids and intravenous immunoglobulin (IVIG) both raise the platelet count, but differ in their mechanisms of action, rapidity of platelet count increase, adverse effects, and costs. For critical bleeding, glucocorticoid and IVIG are given together. In severe bleeding, minor bleeding with surgery, or severe thrombocytopenia without bleeding, glucocorticoid or IVIG alone is typically given. Glucocorticoid therapy is the standard initial treatment for ITP, because glucocorticoids are less expensive and can be easily administered in outpatient settings without infusion. IVIG is generally reserved for settings in which there is a need to raise the platelet count within 12–24 h, or for patients who cannot tolerate glucocorticoids due to diabetes mellitus or significant adverse effects. Differences between glucocorticoids and IVIG include the faster action of IVIG (1–3 days) compared with glucocorticoids (2–14 days), but the efficacies (both short-term and long-term) are similar.
IVIG may be used in patients who cannot tolerate glucocorticoids or wish to avoid glucocorticoid toxicities, or added to glucocorticoid treatment in patients who require a rapid platelet count increase before an invasive procedure, because it raises the platelet count more rapidly than glucocorticoids. The efficacy of glucocorticoids compared with IVIG was demonstrated in a trial that randomly assigned 122 patients with previously untreated acute primary ITP (platelet count ≤ 20,000/μL) to intravenous high-dose methylprednisolone (HDMP; 15 mg/kg/day) or IVIG (0.7 g/kg/day) on days 1–3, followed by a second randomization to placebo or oral prednisone (1 mg/kg/day) on days 4–21 [14]. After monitoring the patients for 1 year, the major results included similar 1-year response rates in patients receiving HDMP or IVIG, and better responses in patients who followed initial HDMP or IVIG with 3 weeks of oral prednisone rather than placebo in the second randomization [14]. For patients who received initial therapy followed by prednisone, the 1-year response rates were 47% (HDMP) and 46% (IVIG). Meanwhile, in patients who received initial therapy followed by placebo, the 1-year response rates were 32% (HDMP) and 29% (IVIG). The increase in platelet count was faster with IVIG than with HDMP (platelet count > 50,000/μL at day 5 observed in 79% of patients receiving IVIG and 60% of patients receiving HDMP). Both treatments were generally well tolerated. There were no deaths or life-threatening hemorrhages. Thus, it may be reasonable to use IVIG in patients for whom glucocorticoids are ineffective and vice versa. Glucocorticoids raise the platelet count in approximately two-thirds of patients with ITP. Most responses occur within 2–5 days, but a period of up to 2 weeks may be required. The mechanism is uncertain but may involve increased apoptotic death of autoantibody-producing lymphocytes and downregulation of macrophage activity responsible for platelet phagocytosis [15].
Prednisolone is typically administered at 1 mg/kg orally once per day (range, 0.5–2 mg/kg daily) for 1–2 weeks, followed by a gradual taper, typically completed within 6 weeks [2, 4]. Administration of glucocorticoids for longer than 6 weeks should be avoided to minimize side effects [4]. If there is no platelet count response after 2 weeks, a faster taper over 1 week can be used. The relative efficacies of shorter or longer tapers of oral prednisone have not been evaluated in randomized trials. Clinical experience suggests that shorter tapers (≤ 6 weeks) are associated with similar efficacy and reduced toxicity compared with longer tapers. For patients who experience a decrease in platelet count during the prednisone taper, an additional course or an alternative therapy can be given. The choice among glucocorticoids is individualized [2, 4].
Dexamethasone
The most common dexamethasone treatment regimens are typically administered at 40 mg orally or intravenously once per day for 4 days with no taper [4]. The intravenous route is generally used for critical or severe bleeding. For minor bleeding or thrombocytopenia without bleeding, dexamethasone can be given orally or intravenously. This can be repeated for additional cycles, up to three times in total [2]. Reasons to prefer dexamethasone over prednisone include faster responses, reduced risk of dose confusion, no need for dose tapering, completion of therapy in 4 days, and fewer bleeding events. Reasons to prefer prednisone over dexamethasone include greater ability to titrate the therapy to match the patient’s individual response. Complete long-term remission with glucocorticoids has been reported in approximately 20% of patients, based on uncontrolled studies. These long-term remission rates may simply reflect the natural history of ITP, with spontaneous resolution independent of glucocorticoid use. Outcomes with different glucocorticoid regimens were compared in a 2016 meta-analysis of nine randomized trials that included 1138 patients with previously untreated ITP who received high-dose dexamethasone for three cycles or oral prednisone 1 mg/kg for 2–4 weeks [16]. Dexamethasone 40 mg daily is approximately equivalent to prednisone 4 mg/kg daily, based on the per-milligram potency for dexamethasone being approximately 7.5 times greater than that for prednisone.
Glucocorticoids for older patients with immune thrombocytopenia
Younger patients may have greater platelet count improvement following therapy than older patients. This was illustrated in a cohort of 117 patients with ITP who were observed for approximately 3 years on various therapies [17]. The proportions of patients with platelet count > 30,000/μL at 6 months after completing therapy were similar in patients aged ≤ 60 or > 60 years (91% versus 87%). However, the number of patients with platelet count > 100,000/μL was much higher in patients aged ≤ 60 years (64% versus 17%). Patients without a sufficient platelet count increase on glucocorticoids should have their diagnosis re-evaluated. Furthermore, patients without a sufficient platelet count increase on glucocorticoids and patients with rapid recurrence of thrombocytopenia following glucocorticoid tapering or discontinuation should be observed or transitioned to other therapies rather than continuing glucocorticoids. According to a consensus guideline in 2019, excessive use of glucocorticoids is a common error in ITP management [2].
Although short-term glucocorticoid administration for older patients with ITP is generally safe and well tolerated, glucocorticoids are associated with both short-term and long-term adverse effects. Short-term effects include mood alterations or emotional lability, insomnia, hyperglycemia, and dyspepsia. These adverse effects may be seen with higher or more prolonged dosing, but can also occur with standard dosing and short courses of therapy. Routine prophylaxis for gastrointestinal toxicity, such as use of proton pump inhibitors, is not generally advised in asymptomatic patients, but can be appropriate for older patients with ITP. Long-term effects include infections, cataracts, and osteoporosis. Attention to bone health is necessary, especially in older patients at increased risk, as outlined in a 2019 guideline from the British Society for Haematology [1]. Considerations about osteoporosis generally apply after approximately 3 months of therapy, but may be relevant at an earlier point in older patients with ITP at high risk for low BMD. Some glucocorticoid toxicities may be intolerable to patients, and in these cases, it is appropriate to use IVIG or second-line therapies [18, 19], including TPO-RAs [10], instead. Although prolonged glucocorticoid use should not be routinely advocated, rare patients may receive long-term treatment with low-dose glucocorticoids (e.g., prednisone ≤ 5 mg/day). For such patients, calcium and vitamin D supplementation is appropriate to reduce the risk of osteoporosis, along with monitoring of BMD. Evidence to support this practice for older patients with ITP was presented in our recent study [10].
Second-line therapy
Second-line therapy options include splenectomy, rituximab, TPO-RAs, other immunosuppressive agents, and combination regimens. Some trials have evaluated intensification of therapy, such as use of multiple agents including glucocorticoid plus rituximab, glucocorticoid plus TPO-RA, and glucocorticoid plus mycophenolate mofetil (MMF). Some of these trials demonstrated improved response rates with multiagent therapy at 6 or 12 months, often at the cost of greater toxicity, but the follow-up was insufficient to determine whether intensification of upfront therapy led to greater cure rates [20]. A randomized trial on glucocorticoid plus MMF versus glucocorticoid alone showed a greater likelihood of achieving platelet count > 30,000/μL in the MMF group [21]. However, there was no difference in the bleeding rates, and the patients who received MMF had decreased quality of life, including more fatigue and worse physical health. Although further studies are warranted, the absence of evidence for increased cure rates renders these mixed results insufficiently compelling to recommend multiagent therapy in the initial treatment setting [22].
Splenectomy, rituximab, and TPO-RAs are the three principal choices for second-line therapy, but differ in their mechanisms of action [4]. While all three therapies are effective for raising the platelet count in the majority of patients, they show marked differences in their application, duration, cost, burden, and adverse effect profiles (Table 1) [23].Table 1 Immune thrombocytopenia therapies and comparisons of their advantages and disadvantages
Therapies Advantages Disadvantages
First-line therapy
Glucocorticoids Effective
Inexpensive
Oral administration
Tolerable short-term toxicities
Uncommon durable response after discontinuation
Serious long-term toxicities
Rescue therapy
IVIG Rapid response Cost
Frequent side effects
Second-line therapy
Splenectomy Effective
Long duration of response
Surgical risks
Infectious risks from asplenia
Thrombosis risk
Rituximab Non-surgical Less effective than splenectomy
Shorter remission
Risks of viral reactivation
(e.g., HBV)
Risks of other toxicities
(e.g., anaphylaxis, pulmonary toxicity)
TPO-RAs Effective
Non-surgical
Non-immunosuppressive
Potential for self-administration
Cost
Requirement for continuous administration
IVIG intravenous immunoglobulin, HBV hepatitis B virus, TPO-RAs thrombopoietin receptor agonists
Splenectomy is a one-time permanent surgical procedure that is the most likely of the three therapies to result in a durable response [24]. Because some patients can experience spontaneous remission within the first year, splenectomy is generally deferred until at least 1 year has elapsed since diagnosis, if feasible. Splenectomy may be a good choice for patients who wish to undergo a single potentially curative surgical procedure and who are willing to accept the increased risks of infection and venous thromboembolism. Previous studies indicated that younger patients had a higher response rate than older patients, but a specific age cutoff for effective splenectomy could not be determined [25, 26]. One study reported that the duration of response following splenectomy and the likelihood of a response may be reduced in patients aged > 65 years [24].
Rituximab typically requires four weekly intravenous administrations [27] and may need to be re-administered. Rituximab may be a good choice for patients who wish to avoid surgery and prefer not to take long-term medications. However, its effect is often short-lived, necessitating redosing or use of another second-line agent [28–34]. Rituximab is an immunosuppressive agent, and its use has been avoided during the coronavirus disease-2019 (COVID-19) pandemic because it blunts the immune response to vaccinations, including COVID-19 vaccination, probably for at least 6 months. New-onset ITP or exacerbation of existing ITP has been reported following COVID-19 vaccination [35]. However, the risk is low, and vaccination remains the best way to reduce serious disease, hospitalization, and death from COVID-19 [36]. Adverse effects of rituximab include infusion reactions and prolonged immunosuppression, which can result in reactivation of HBV infection [37]. Progressive multifocal leukoencephalopathy (PML) has been reported following rituximab therapy for ITP, and many of the affected patients had been pretreated with other immunosuppressive agents in addition to rituximab [37]. Boxed warnings for infusion reactions, HBV reactivation, mucocutaneous reactions, and PML are included in the prescribing information.
TPO-RAs typically require administration for an extended period of time, although their responses are sometimes maintained after cessation of treatment. A TPO-RA may be a good choice for patients who wish to avoid surgery and the immunosuppressive effects of splenectomy or rituximab and who are less concerned about the need to take medications for an extended period of time, including the associated costs and burdens. Temporary use of a TPO-RA may be appropriate during the COVID-19 pandemic as a means of avoiding immunosuppressive therapy, especially for older patients. If splenectomy is chosen, it is generally preferable to wait at least 1 year from the time of diagnosis in case spontaneous remission occurs. During this time, a TPO-RA may be used temporarily while surgery is being scheduled and preoperative vaccines are being administered. A TPO-RA may also be useful for patients who require a temporary increase in platelet count in preparation for splenectomy and do not have an adequate platelet count response to glucocorticoids [38]. The 2019 ASH guideline on ITP makes weak recommendations for TPO-RA over rituximab, and for rituximab over splenectomy, but it emphasizes the potential usefulness of all three treatments and the importance of patient characteristics, such as age, ITP history, and comorbidities, values, and preferences in the final decision [4]. Approximately 80% of patients with ITP have a significant platelet count increase in response to TPO-RA therapy [39]. Several reports have documented sustained responses in some patients with ITP after TPO-RA discontinuation, with response rates of 30% to 50%, although the observation periods were relatively short (< 1 year) [40]. In most patients, these agents do not induce remission, and prolonged maintenance therapy is usually required. If treatment is discontinued, the platelet count generally returns to the baseline level or lower. Available TPO-RAs for ITP include romiplostim, eltrombopag, and avatrombopag. These three agents are all effective, but their relative efficacies for ITP have not been evaluated in randomized trials. The 2019 ASH guideline does not express a preference for a particular TPO-RA, and suggests a choice between romiplostim or eltrombopag based on patient preference for route of administration [4]. The choice is based on availability, cost, comorbidities, and patient preference and can be individualized. Some patients have a strong preference for one administration route (oral or subcutaneous) over the other.
Comparisons among second-line therapies
Splenectomy, rituximab, and TPO-RAs have not been directly compared in randomized trials [26]. However, the reported initial response rates are higher for splenectomy than for rituximab (60%–80% versus 55%–65%), and the responses with splenectomy are more durable, often lasting for many years or even indefinitely for splenectomy versus 1–2 years for rituximab [28, 41]. There are also significant differences in the short- and long-term risks of splenectomy versus rituximab [41–43]. Rituximab-related morbidities include infections, such as pneumonia and sepsis, and thrombotic complications, such as venous thromboembolism and myocardial infarction. There are significant differences in the short- and long-term risks and burdens of rituximab versus TPO-RAs. While rituximab causes immunosuppression and may only work transiently, TPO-RAs are highly effective but require extended administration in many cases. Other therapies may be appropriate if splenectomy, rituximab, and/or TPO-RAs are ineffective or cannot be used. As a result, the choice of therapy is highly dependent on patient values and preferences. All three options require patient assistance to balance the risks and benefits of the approach.
Bisphosphonates and alternative treatment options as prophylaxis against glucocorticoid-induced osteoporosis in older patients with immune thrombocytopenia
Use of bisphosphonates in older patients with ITP who are to receive long-term steroid therapy should be followed up in all patients receiving any dose of glucocorticoids for ≥ 3 months to minimize bone loss [7]. The dose and duration of glucocorticoid therapy should be as low as possible, because even doses considered replacement doses or chronic inhaled glucocorticoids can cause bone loss [44]. Patients should complete weightbearing exercises to prevent both bone loss and muscle atrophy. Patients should also avoid smoking and excess alcohol, and take measures to prevent falls. Evidence for the use of vitamin D3 to supplement bisphosphonate treatment in patients at risk of GIO was reported [7]. The ACR Task Force osteoporosis guidelines suggest that all patients taking glucocorticoids at any dose with an anticipated duration of ≥ 3 months should maintain a total calcium intake of 1000–1200 mg/day and vitamin D intake of 600–800 international units/day through diet and/or supplements [7]. Glucocorticoids induce a negative calcium balance by decreasing intestinal calcium absorption and increasing urinary calcium excretion [45]. Therefore, calcium supplementation may attenuate bone loss in patients taking glucocorticoids. In a meta-analysis of five randomized trials comparing calcium and vitamin D (cholecalciferol or active vitamin D metabolite) with calcium alone or placebo in patients taking glucocorticoids, significant improvements in the lumbar spine and radial BMD were noted in the calcium and vitamin D group (weighted mean differences between the treatment group and control group of 2.6% and 2.5%, respectively) [3]. The incidence of new nontraumatic fractures did not differ significantly in two trials (odds ratio [OR] 0.6, 95% CI 0.1–2.4). In one of the larger studies included in the meta-analysis, 96 patients with rheumatoid arthritis receiving low-dose glucocorticoid therapy (mean prednisone dose, 5.6 mg daily) were randomly assigned to calcium carbonate (1000 mg of elemental calcium daily) plus vitamin D3 (500 international units/day) or placebo [46]. The between-group differences in the annual rate of change were 2.65% (95% CI 0.73–4.57) and 2.08% (95% CI 0.43–3.73) for the spine and trochanter BMD, respectively, favoring calcium and vitamin D supplementation. Although calcium and vitamin D supplementation is necessary, it is generally not sufficient to prevent bone loss and fracture in patients taking high-dose glucocorticoids [47–49]. Vitamin D metabolites with greater activity than vitamin D itself, such as calcitriol and alfacalcidol, have been evaluated for the prevention and treatment of glucocorticoid-induced bone loss [50].
Calcitriol (1,25-dihydroxyvitamin D; the most active metabolite of vitamin D) plus calcium protected against spine bone loss more effectively than calcium alone in patients taking glucocorticoids [48, 51]. A meta-analysis of five trials on active vitamin D metabolites in patients exposed to corticosteroids reported a beneficial effect of vitamin D metabolites on lumbar spine BMD [52]. However, there were insufficient data to address fracture prevention. Calcitonin is not used for the treatment or prevention of GIO, because more effective drugs (e.g., bisphosphonates, teriparatide) are available for prevention of bone loss and reduction of fracture risk. Another concern is that long-term use of calcitonin for osteoporosis has been associated with an increase in cancer rates [53].
Active vitamin D metabolites are not commonly used because of the risks of hypercalcemia and hypercalciuria in patients whose urinary calcium excretion is already increased and because more effective therapies are available [54]. The superior efficacy of bisphosphonates compared with an active vitamin D metabolite for preventing glucocorticoid-induced bone loss has been demonstrated in several randomized trials [55, 56]. Although not statistically significant, one trial showed that fewer patients randomly assigned to alendronate had new vertebral fractures compared with patients assigned to alfacalcidol (three versus eight patients) [57].
Older patients with the highest risk for fracture are the most likely to benefit from drug therapy. Thus, selection of patients based on fracture risk, as determined by a combination of BMD and clinical risk factors, is desirable. Patients with established osteoporosis (history of fragility fracture or BMD T-score ≤ − 2.5) are at the highest risk for fracture. For patients without established osteoporosis, fracture risk can be assessed using a fracture risk calculator, such as FRAX. FRAX estimates the 10-year probability of fracture in untreated patients aged 40–90 years using femoral neck BMD and clinical risk factors, including glucocorticoid exposure. FRAX does not account for glucocorticoid dose or duration, and therefore FRAX risk estimates must be corrected by the glucocorticoid dose [58]. For patients taking prednisolone > 7.5 mg/day or equivalent, the risk estimate should be increased by 15% for major osteoporotic fracture and 20% for hip fracture [58].
In North America, reasonable glucocorticoid-corrected thresholds to indicate high, moderate, and low risks of fracture are as follows [3]: high risk, 10-year probability of hip fracture or combined major osteoporotic fracture of ≥ 3% or 20%, respectively; moderate risk, 10-year probability of hip fracture or combined major osteoporotic fracture of 1%–3% or 10%–19%, respectively; and low risk, 10-year probability of hip fracture or combined major osteoporotic fracture of ≤ 1% or < 10%, respectively. Some patients receiving glucocorticoids are at high risk, even if they fail to meet the FRAX criteria for high risk. For patients with clinical risk factors for fracture and low lumbar spine BMD, but normal femoral neck BMD, FRAX is likely to underestimate the fracture risk. This situation is particularly likely in patients taking glucocorticoids, which are more prone to cause osteoporosis in the spine than in the hip. Thus, intervention guidelines with or without the use of FRAX provide only general clinical guidance. Treatment should remain individualized through shared decision making between patients and clinicians.
For men aged ≥ 50 years and postmenopausal women who are initiating or receiving chronic treatment with any dose of glucocorticoids for any duration and have osteoporosis (previous fragility fracture and/or BMD T-score ≤ − 2.5) at initial assessment, pharmacological therapy is recommended. For high-risk men aged ≥ 50 years and postmenopausal women who are initiating or receiving chronic treatment with any dose of glucocorticoids for any duration and have T-scores between − 1.0 and − 2.5, pharmacological therapy is suggested. A reasonable threshold to indicate high risk in some settings is a glucocorticoid-corrected, FRAX-calculated, 10-year probability of hip or combined major osteoporotic fracture of ≥ 3 or 20%, respectively. For men aged > 50 years and postmenopausal women with T-scores between − 1.0 and − 2.5 who have a glucocorticoid-corrected, FRAX-calculated absolute risk below these thresholds, pharmacological therapy is suggested if they are taking ≥ 7.5 mg/day of prednisone or its equivalent for an anticipated duration of ≥ 3 months. These recommendations are based on randomized trial data showing that pharmacological therapy improves BMD in patients taking glucocorticoids [59, 60] and additional randomized trial data showing that pharmacological therapy reduces fracture in men and postmenopausal women with established osteoporosis [61].
Bisphosphonates are the first-line therapy for prevention and treatment of GIO in men because they are known to reduce fracture risk. Men who develop symptomatic hypogonadism should also be treated with testosterone for its benefits on muscle, energy, and libido, as well as on bone [62, 63]. A number of testosterone preparations are available for the treatment of testosterone deficiency. Alendronate or risedronate is preferred because of clinical trial data demonstrating efficacy in men and women with GIO. For patients who cannot tolerate oral bisphosphonates or who have difficulty with the dosing requirements or adherence, intravenous zoledronic acid is an acceptable alternative.
Parathyroid hormone (PTH; teriparatide) is typically reserved for patients with severe osteoporosis (T-score ≤ − 3.5 in the absence of fracture or T-score ≤ − 2.5 plus fragility fracture). Teriparatide is also an option for patients who cannot tolerate any of the available bisphosphonates or who continue to suffer fracture after 1 year of bisphosphonate therapy. Denosumab is another alternative therapeutic option for patients at high risk for fracture. However, due to the increased risk of vertebral fracture after discontinuation of denosumab, the need for a careful exit strategy should be discussed with patients prior to initiation of its cessation.
There are substantial data supporting the use of antiresorptive agents, such as bisphosphonates, for the prevention and treatment of glucocorticoid-induced bone loss. In a meta-analysis of 27 randomized trials evaluating bisphosphonates (alone or in combination with calcium and vitamin D) versus calcium and vitamin D (alone or with placebo) for the prevention and treatment of GIO, significant improvements in the lumbar spine (absolute difference 3.5%) and femoral neck (absolute difference 2.1%) BMD were noted in the bisphosphonate group [64]. There was also a reduction in the risk of new vertebral fracture with bisphosphonate treatment (44 versus 77 per 1000 persons in the bisphosphonate and no treatment groups, respectively; risk ratio [RR] 0.57, 95% CI 0.35–0.91). The reduction in the risk of nonvertebral fracture did not reach statistical significance (42 versus 55 per 1000 persons; RR 0.79, 95% CI 0.47–1.33). The therapeutic efficacy of bisphosphonates in patients with GIO has been thought to be related to their ability to promote osteoclast apoptosis [65]. However, glucocorticoids may negate the pro-apoptotic effect of bisphosphonates, suggesting that bisphosphonates may prevent glucocorticoid-induced bone loss by prolonging the lifespan of osteoblasts [65, 66].
The efficacy of alendronate in patients receiving glucocorticoid therapy was demonstrated in a study involving 477 patients aged 17–83 years who were randomly assigned to receive one of two doses of alendronate or placebo [51]. The mean BMD of the lumbar spine increased by 2.1% and 2.9% over 48 weeks in the patients receiving 5 and 10 mg of alendronate daily, respectively, but decreased by 0.4% in the patients receiving placebo. The femoral neck, trochanter, and total body BMD also increased significantly in the alendronate groups. Patients receiving alendronate had fewer new vertebral fractures compared with those receiving placebo (2.3% versus 3.7%), although fracture was not a primary outcome. These benefits were maintained for 2 years, representing a 12-month extension of a previously completed 1-year trial of daily alendronate [67]. Once-weekly alendronate (70 mg) showed similar improvement of BMD [68], and a retrospective cohort study found that alendronate was associated with a significantly lower risk of hip fracture in older patients taking oral prednisolone [69].
Risedronate is also effective for the prevention and treatment of osteoporosis, including GIO [70, 71]. In a 1-year study of risedronate versus placebo in 290 patients receiving glucocorticoid therapy (prednisone ≥ 7.5 mg/day for ≥ 6 months), the lumbar spine and femoral neck BMD increased by 2.7% and 1.8%, respectively, in the risedronate group, compared with no change in the placebo group [70]. The relative risk of vertebral fracture (a secondary outcome) was reduced by 70%.
The efficacy of zoledronic acid for the prevention and treatment of GIO was demonstrated in a 1-year randomized trial of intravenous zoledronic acid (5 mg once) or daily oral risedronate (5 mg) in 288 patients who recently started glucocorticoids (prevention group) and 545 patients who had been taking glucocorticoids for > 3 months (treatment group) [71]. In an analysis of patients who received the study drugs and had baseline and follow-up BMD measurements, zoledronic acid and risedronate increased the mean BMD of the lumbar spine in both the prevention (2.6% and 0.6%, respectively) and treatment (4.1% and 2.7%, respectively) groups. The study was not designed to evaluate fractures, which occurred in three and five patients in the risedronate and zoledronic acid groups, respectively. During the first 3 days, the occurrence of adverse events (predominantly arthralgias, fever, and flu-like symptoms) was greater in the zoledronic acid group.
Monitoring for glucocorticoid-induced osteoporosis
There are several published guidelines for monitoring the response to osteoporosis therapy. Although all recommend follow-up BMD testing, there is no consensus on the optimal frequency of monitoring and the preferred site to monitor. We typically use dual-energy X-ray absorptiometry to measure the BMD of the lumbar spine and hip at the initiation of glucocorticoid therapy and after 1 year of treatment. If the BMD is stable or improved, we perform the measurements less frequently (every 2–3 years) thereafter. If glucocorticoids are discontinued and the BMD remains stable, measurements at 5-year intervals may be sufficient. The finding of a BMD decrease greater than the least significant change or a new fracture in a treated patient should trigger additional evaluation for contributing factors, which may include poor adherence to therapy, inadequate gastrointestinal absorption, inadequate intake of calcium and vitamin D, or development of a disease or disorder with adverse skeletal effects. Switching from oral bisphosphonates to intravenous zoledronic acid may be effective in patients with poor absorption or poor compliance with the oral regimen. Alternative options for patients who fail oral bisphosphonate therapy are similar to those for patients with osteoporosis in general. There may be a significant increase in BMD after discontinuation of exogenous glucocorticoid therapy or reversal of endogenous Cushing’s syndrome [72]. As an example, one study examined spine BMD after successful treatment for Cushing’s disease in 20 patients, all of whom had marked osteoporosis of the lumbar spine and femoral neck [72]. There was no change in BMD for 6 months, the time required for gradual reversal of increased osteoclastic activity, after which the BMD increased. However, patients who have experienced a fracture may have a permanent deformity.
Perspectives and recommendations
The purpose of this section is to summarize the provided literature and place it into the context of current recommendations for prevention of GIO in older ITP patients. Given that the present review focuses on the literature generated since 2017, it may also be pertinent to summarize some of the key findings in the report by Hill et al. [1], which covers 1960–2017.
Regarding the GIO treatment recommendations by various clinical bodies worldwide [7, 73], the guidelines do not address older patients as a specific cohort. The recently published literature provides additional insights into this cohort. Recent papers on ITP and/or GIO in older patients support the continued recommendation of bisphosphonates to mitigate the development of GIO in older ITP patients, as indicated in our previous report [10] (Fig. 2). Bisphosphonates are currently the recommended first-line treatment for GIO, according to the guideline of the Japanese Society for Bone and Mineral Research [74]. This recommendation is almost 8 years old and is limited to drugs currently approved for osteoporosis treatment in Japan. A new guideline will be published in 2023. However, this new guideline will not specifically address older ITP patients. The 2022 ACR guideline for the prevention and treatment of GIO describes sequential treatments recommended when the initial osteoporosis therapy and glucocorticoid therapy are discontinued depending on the fracture risk [75]. Further studies are needed to evaluate the utility of new diagnostic tools [76, 77] and to conduct direct comparisons between the efficacies of GIO therapies [78]. Fig. 2 Flow chart highlighting the decision-making process for treatment of glucocorticoid-induced osteoporosis in older patients with immune thrombocytopenia. If a complete response (CR) is achieved on prednisolone therapy, a gradual taper of prednisolone is initiated and typically completed within 6 weeks. If no platelet count response is observed after 2 weeks of prednisolone therapy, a thrombopoietin receptor agonist (TPO-RA) can be added and a faster taper of prednisolone over 1 week can be used. Data for dual-energy X-ray absorptiometry (DXA) and femoral neck bone mineral density measurement are typically entered into FRAX® (a validated algorithm; available online at www.sheffield.ac.uk/FRAX/tool.jsp) to define the risk-adapted approach to bisphosphonate (BPN) treatment during the initial loading and tapering phases of prednisolone therapy. Active vitamin D (VD) is usually prescribed for women. Serum 25-hydroxyvitamin D (25(OH)D) levels are measured during the tapering phase of prednisolone therapy to ensure adequate VD levels for continuous VD therapy in women and additional VD therapy in men
Author contribution
SY designed, prepared, and reviewed the manuscript. SY met the International Committee of Medical Journal Editors criteria for authorship of this article as a whole, and provided approval for this version to be published.
Funding
The article processing charges were funded by the author. This work was supported by JSPS KAKENHI Grant Number 22K12887. The author thanks Alison Sherwin, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.
Declarations
Ethics approval
This was a retrospective review study with no experimental interventions.
Informed consent
This was a retrospective review study with requirement for informed consent.
Conflict of interest
The author declares no competing interests.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Hill QA Grainger JD Thachil J Provan D Evans G Garg M Bradbury C Bagot C Kanis JA Compston JE British Society of Haematology in conjunction with the UK ITP forum The prevention of glucocorticoid- induced osteoporosis in patients with immune thrombocytopenia receiving steroids: a British Society for Haematology Good Practice Paper Br J Haematol 2019 185 410 417 10.1111/bjh.15735 30609018
2. Provan D Arnold DM Bussel JB Chong BH Cooper N Gernsheimer T Ghanima W Godeau B González-López TJ Grainger J Hou M Kruse C McDonald V Michel M Newland AC Pavord S Rodeghiero F Scully M Tomiyama Y Wong RS Zaja F Kuter DJ Updated international consensus report on the investigation and management of primary immune thrombocytopenia Blood Adv 2019 3 3780 3817 10.1182/bloodadvances.2019000812 31770441
3. Canalis E Mazziotti G Giustina A Bilezikian JP Glucocorticoid-induced osteoporosis: pathophysiology and therapy Osteoporos Int 2007 18 1319 1328 10.1007/s00198-007-0394-0 17566815
4. Neunert C Terrell DR Arnold DM Buchanan G Cines DB Cooper N Cuker A Despotovic JM George JN Grace RF Kühne T Kuter DJ Lim W McCrae KR Pruitt B Shimanek H Vesely SK American society of hematology 2019 guidelines for immune thrombocytopenia Blood Adv 2019 3 3829 3866 10.1182/bloodadvances.2019000966 31794604
5. Rossini M Viapiana O Vitiello M Malavolta N La Montagna G Maddali Bongi S Di Munno O Nuti R Manzini CU Ferri C Bogliolo L Mathieu A Cantatore F Del Puente A Muratore M Grassi W Frediani B Saviola G Delvino P Mirone L Ferraccioli G Tripi G Piazza I Gatti D Prevalence and incidence of osteoporotic fractures in patients on long-term glucocorticoid treatment for rheumatic diseases: the Glucocorticoid Induced OsTeoporosis TOol (GIOTTO) study Reumatismo 2017 69 30 39 10.4081/reumatismo.2017.922 28535619
6. van Staa TP Leufkens HG Cooper C The epidemiology of corticosteroid-induced osteoporosis: a meta-analysis Osteoporos Int 2002 13 777 787 10.1007/s001980200108 12378366
7. Buckley L Guyatt G Fink HA Cannon M Grossman J Hansen KE Humphrey MB Lane NE Magrey M Miller M Morrison L Rao M Robinson AB Saha S Wolver S Bannuru RR Vaysbrot E Osani M Turgunbaev M Miller AS McAlindon T 2017 American college of rheumatology guideline for the prevention and treatment of glucocorticoid-induced osteoporosis Arthritis Rheumatol 2017 69 1521 1537 10.1002/art.40137 28585373
8. Kanis JA Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report WHO Study Group Osteoporos Int 1994 4 368 381 10.1007/BF01622200
9. Nomura S Kurata Y Tomiyama Y Takubo T Hasegawa M Saigo K Nishikawa M Higasa S Maeda Y Hayashi K Effects of bisphosphonate administration on the bone mass in immune thrombocytopenic purpura patients under treatment with steroids Clin Appl Thromb Hemost 2010 16 622 627 10.1177/1076029609350889 19959489
10. Yamasaki S Kamezaki K Ito Y Horiuchi T Bisphosphonate use for glucocorticoid-induced osteoporosis in elderly patients with immune thrombocytopenia receiving prolonged steroid therapy: a single institute retrospective Study Hematol Rep 2022 14 276 285 10.1016/j.schres.2022.12.008 36135323
11. Frederiksen H Schmidt K The incidence of idiopathic thrombocytopenic purpura in adults increases with age Blood 1999 94 909 913 10.1182/blood.V94.3.909.415k02_909_913 10419881
12. Kohli R Chaturvedi S Epidemiology and clinical manifestations of immune thrombocytopenia Hamostaseologie 2019 39 238 249 10.1055/s-0039-1683416 30868551
13. Neunert C Noroozi N Norman G Buchanan GR Goy J Nazi I Kelton JG Arnold DM Severe bleeding events in adults and children with primary immune thrombocytopenia: a systematic review J Thromb Haemost 2015 13 457 467 10.1111/jth.12813 25495497
14. Godeau B Chevret S Varet B Lefrère F Zini JM Bassompierre F Chèze S Legouffe E Hulin C Grange MJ Fain O Bierling P French ATIP Study Group Intravenous immunoglobulin or high-dose methylprednisolone, with or without oral prednisone, for adults with untreated severe autoimmune thrombocytopenic purpura: a randomised, multicentre trial Lancet 2002 359 23 29 10.1016/S0140-6736(02)07275-6 11809183
15. Mizutani H Furubayashi T Imai Y Kashiwagi H Honda S Take H Kurata Y Yonezawa T Tarui S Ikehara S Mechanisms of corticosteroid action in immune thrombocytopenic purpura (ITP): experimental studies using ITP-prone mice, (NZW x BXSB) F1 Blood 1992 79 942 947 10.1182/blood.V79.4.942.942 1737103
16. Mithoowani S Gregory-Miller K Goy J Miller MC Wang G Noroozi N Kelton JG Arnold DM High-dose dexamethasone compared with prednisone for previously untreated primary immune thrombocytopenia: a systematic review and meta-analysis Lancet Haematol 2016 3 e489 e496 10.1016/S2352-3026(16)30109-0 27658982
17. Cortelazzo S Finazzi G Buelli M Molteni A Viero P Barbui T High risk of severe bleeding in aged patients with chronic idiopathic thrombocytopenic purpura Blood 1991 77 1 31 33 10.1182/blood.V77.1.31.31 1984800
18. Guidry JA George JN Vesely SK Kennison SM Terrell DR Corticosteroid side-effects and risk for bleeding in immune thrombocytopenic purpura: patient and hematologist perspectives Eur J Haematol 2009 83 175 182 10.1111/j.1600-0609.2009.01265.x 19374704
19. Guidry JA Watson S George JN Vesely SK Terrell DR Addendum to corticosteroid side effects and risk for bleeding in immune thrombocytopenic purpura: patient perspectives Eur J Haematol 2009 83 497 498 10.1111/j.1600-0609.2009.01322.x 19614752
20. Cuker A Prak ET Cines DB Can immune thrombocytopenia be cured with medical therapy? Semin Thromb Hemost 2015 41 395 404 10.1055/s-0034-1544001 25793364
21. Bradbury CA Pell J Hill Q Bagot C Cooper N Ingram J Breheny K Kandiyali R Rayment R Evans G Talks K Thomas I Greenwood R Mycophenolate mofetil for first-line treatment of immune thrombocytopenia N Engl J Med 2021 385 885 895 10.1056/NEJMoa2100596 34469646
22. Bolton-Maggs PHB George JN Immune thrombocytopenia treatment N Engl J Med 2021 385 948 950 10.1056/NEJMe2110953 34469652
23. Bylsma LC Fryzek JP Cetin K Callaghan F Bezold C Mehta B Wasser JS Systematic literature review of treatments used for adult immune thrombocytopenia in the second-line setting Am J Hematol 2019 94 118 132 10.1002/ajh.25301 30264861
24. Gonzalez-Porras JR Escalante F Pardal E Sierra M Garcia-Frade LJ Redondo S Arefi M Aguilar C Ortega F de Cabo E Fisac RM Sanz O Esteban C Alberca I Sanchez-Barba M Santos MT Fernandez A Gonzalez-Lopez TJ representing the Grupo de Trombosis y Hemostasia de Castilla y León Safety and efficacy of splenectomy in over 65-yrs-old patients with immune thrombocytopenia Eur J Haematol 2013 91 236 241 10.1111/ejh.12146 23679653
25. Kojouri K Vesely SK Terrell DR George JN Splenectomy for adult patients with idiopathic thrombocytopenic purpura: a systematic review to assess long-term platelet count responses, prediction of response, and surgical complications Blood 2004 104 2623 2634 10.1182/blood-2004-03-1168 15217831
26. Fabris F Tassan T Ramon R Carraro G Randi ML Luzzatto G Moschino P Girolami A Age as the major predictive factor of long-term response to splenectomy in immune thrombocytopenic purpura Br J Haematol 2001 112 637 640 10.1046/j.1365-2141.2001.02615.x 11260065
27. Cragg MS Walshe CA Ivanov AO Glennie MJ The biology of CD20 and its potential as a target for mAb therapy Curr Dir Autoimmun 2005 8 140 174 10.1159/000082102 15564720
28. Chugh S Darvish-Kazem S Lim W Crowther MA Ghanima W Wang G Heddle NM Kelton JG Arnold DM Rituximab plus standard of care for treatment of primary immune thrombocytopenia: a systematic review and meta-analysis Lancet Haematol 2015 2 e75 81 10.1016/S2352-3026(15)00003-4 26687612
29. Ghanima W Khelif A Waage A Michel M Tjønnfjord GE Romdhan NB Kahrs J Darne B Holme PA RITP study group Rituximab as second-line treatment for adult immune thrombocytopenia (the RITP trial): a multicentre, randomised, double-blind, placebo-controlled trial Lancet 2015 385 16531661 10.1016/S0140-6736(14)61495-1
30. Auger S Duny Y Rossi JF Quittet P Rituximab before splenectomy in adults with primary idiopathic thrombocytopenic purpura: a meta-analysis Br J Haematol 2012 158 386 398 10.1111/j.1365-2141.2012.09169.x 22612239
31. Arnold DM Dentali F Crowther MA Meyer RM Cook RJ Sigouin C Fraser GA Lim W Kelton JG Systematic review: efficacy and safety of rituximab for adults with idiopathic thrombocytopenic purpura Ann Intern Med 2007 146 25 33 10.7326/0003-4819-146-1-200701020-00006 17200219
32. Patel VL Mahévas M Lee SY Stasi R Cunningham-Rundles S Godeau B Kanter J Neufeld E Taube T Ramenghi U Shenoy S Ward MJ Mihatov N Patel VL Bierling P Lesser M Cooper N Bussel JB Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia Blood 2012 119 5989 5995 10.1182/blood-2011-11-393975 22566601
33. Nazi I Kelton JG Larché M Snider DP Heddle NM Crowther MA Cook RJ Tinmouth AT Mangel J Arnold DM The effect of rituximab on vaccine responses in patients with immune thrombocytopenia Blood 2013 122 1946 1953 10.1182/blood-2013-04-494096 23851398
34. Khellaf M Charles-Nelson A Fain O Terriou L Viallard JF Cheze S Graveleau J Slama B Audia S Ebbo M Le Guenno G Cliquennois M Salles G Bonmati C Teillet F Galicier L Hot A Lambotte O Lefrère F Sacko S Kengue DK Bierling P Roudot-Thoraval F Michel M Godeau B Safety and efficacy of rituximab in adult immune thrombocytopenia: results from a prospective registry including 248 patients Blood 2014 124 3228 3236 10.1182/blood-2014-06-582346 25293768
35. Crickx E Moulis G Ebbo M Terriou L Briantais A Languille L Limal N Guillet S Michel M Mahevas M Godeau B Safety of anti-SARS-CoV-2 vaccination for patients with immune thrombocytopenia Br J Haematol 2021 195 703 705 10.1111/bjh.17813 34467525
36. Visser C Swinkels M van Werkhoven ED Croles FN Noordzij-Nooteboom HS Eefting M Last-Koopmans SM Idink C Westerweel PE Santbergen B Jobse PA Baboe F RECOVAC-IR ConsortiumTe Boekhorst PAW Leebeek FWG Levin MD Kruip MJHA Jansen AJG COVID-19 vaccination in patients with immune thrombocytopenia Blood Adv 2022 6 1637 1644 10.1182/bloodadvances.2021006379 34941989
37. Carson KR Evens AM Richey EA Habermann TM Focosi D Seymour JF Laubach J Bawn SD Gordon LI Winter JN Furman RR Vose JM Zelenetz AD Mamtani R Raisch DW Dorshimer GW Rosen ST Muro K Gottardi-Littell NR Talley RL Sartor O Green D Major EO Bennett CL Progressive multifocal leukoencephalopathy after rituximab therapy in HIV-negative patients: a report of 57 cases from the research on adverse drug events and reports project Blood 2009 113 4834 4840 10.1182/blood-2008-10-186999 19264918
38. Zeng Y, Duan X, Xu J, Ni X (2011) TPO receptor agonist for chronic idiopathic thrombocytopenic purpura. Cochrane Database Syst Rev CD008235. 10.1002/14651858.CD008235
39. González-López TJ Pascual C Álvarez-Román MT Fernández-Fuertes F Sánchez-González B Caparrós I Jarque I Mingot-Castellano ME Hernández-Rivas JA Martín-Salces M Solán L Beneit P Jiménez R Bernat S Andrade MM Cortés M Cortti MJ Pérez-Crespo S Gómez-Núñez M Olivera PE Pérez-Rus G Martínez-Robles V Alonso R Fernández-Rodríguez A Arratibel MC Perera M Fernández-Miñano C Fuertes-Palacio MA Vázquez-Paganini JA Gutierrez-Jomarrón I Valcarce I de Cabo E Sainz A Fisac R Aguilar C Paz Martínez-Badas M Peñarrubia MJ Calbacho M de Cos C González-Silva M Coria E Alonso A Casaus A Luaña A Galán P Fernández-Canal C Garcia-Frade J González-Porras JR Successful discontinuation of eltrombopag after complete remission in patients with primary immune thrombocytopenia Am J Hematol 2015 90 E40 E43 10.1002/ajh.23900 25400215
40. Bussel JB Kuter DJ Pullarkat V Lyons RM Guo M Nichol JL Safety and efficacy of long-term treatment with romiplostim in thrombocytopenic patients with chronic ITP Blood 2009 113 2161 2171 10.1182/blood-2008-04-150078 18981291
41. Lal LS Said Q Andrade K Cuker A Second-line treatments and outcomes for immune thrombocytopenia: A retrospective study with electronic health records Res Pract Thromb Haemost 2020 4 1131 1140 10.1002/rth2.12423 33134779
42. Rodeghiero F Ruggeri M Is splenectomy still the gold standard for the treatment of chronic ITP? Am J Hematol 2008 83 91 10.1002/ajh.21109 17994550
43. Cooper N Evangelista ML Amadori S Stasi R Should rituximab be used before or after splenectomy in patients with immune thrombocytopenic purpura? Curr Opin Hematol 2007 14 642 646 10.1097/MOH.0b013e3282c8ca50 17898569
44. Brynes RK Orazi A Theodore D Burgess P Bailey CK Thein MM Bakshi KK Evaluation of bone marrow reticulin in patients with chronic immune thrombocytopenia treated with eltrombopag: Data from the EXTEND study Am J Hematol 2015 90 598 601 10.1002/ajh.24011 25801698
45. Zelissen PM Croughs RJ van Rijk PP Raymakers JA Effect of glucocorticoid replacement therapy on bone mineral density in patients with Addison disease Ann Intern Med 1994 120 207 210 10.7326/0003-4819-120-3-199402010-00005 8273983
46 Homik J Suarez-Almazor ME Shea B Cranney A Wells G Tugwell P Calcium and vitamin D for corticosteroid-induced osteoporosis Cochrane Database Syst Rev 2000 1998 CD000952 10.1002/14651858.CD000952 10796394
47. Buckley LM Leib ES Cartularo KS Vacek PM Cooper SM Calcium and vitamin D3 supplementation prevents bone loss in the spine secondary to low-dose corticosteroids in patients with rheumatoid arthritis. A randomized, double-blind, placebo-controlled trial Ann Intern Med 1996 125 961 968 10.7326/0003-4819-125-12-199612150-00004 8967706
48. Boutsen Y Jamart J Esselinckx W Devogelaer JP Primary prevention of glucocorticoid-induced osteoporosis with intravenous pamidronate and calcium: a prospective controlled 1-year study comparing a single infusion, an infusion given once every 3 months, and calcium alone J Bone Miner Res 2001 16 104 112 10.1359/jbmr.2001.16.1.104 11149473
49. Sambrook P Birmingham J Kelly P Kempler S Nguyen T Pocock N Eisman J Prevention of corticosteroid osteoporosis. A comparison of calcium, calcitriol, and calcitonin N Engl J Med 1993 328 1747 1752 10.1056/NEJM199306173282404 7684512
50. Cohen S Levy RM Keller M Boling E Emkey RD Greenwald M Zizic TM Wallach S Sewell KL Lukert BP Axelrod DW Chines AA Risedronate therapy prevents corticosteroid-induced bone loss: a twelve-month, multicenter, randomized, double-blind, placebo-controlled, parallel-group study Arthritis Rheum 1999 42 2309 2318 10.1002/1529-0131(199911)42:11<2309::AID-ANR8>3.0.CO;2-K 10555025
51. Saag KG Emkey R Schnitzer TJ Brown JP Hawkins F Goemaere S Thamsborg G Liberman UA Delmas PD Malice MP Czachur M Daifotis AG Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis N Engl J Med 1998 339 292 299 10.1056/NEJM199807303390502 9682041
52. Richy F Ethgen O Bruyere O Reginster JY Efficacy of alphacalcidol and calcitriol in primary and corticosteroid-induced osteoporosis: a meta-analysis of their effects on bone mineral density and fracture rate Osteoporos Int 2004 15 301 310 10.1007/s00198-003-1570-5 14740153
53. Lukert BP Raisz LG Glucocorticoid-induced osteoporosis: pathogenesis and management Ann Intern Med 1990 112 352 364 10.7326/0003-4819-112-5-352 2407167
54. Okamoto H Shibazaki N Yoshimura T Uzawa T Sugimoto T Association between elcatonin use and cancer risk in Japan: A follow-up study after a randomized, double-blind, placebo-controlled study of once-weekly elcatonin in primary postmenopausal osteoporosis Osteoporos Sarcopenia 2020 6 15 19 10.1016/j.afos.2020.02.001 32226828
55. Suzuki Y Ichikawa Y Saito E Homma M Importance of increased urinary calcium excretion in the development of secondary hyperparathyroidism of patients under glucocorticoid therapy Metabolism 1983 32 151 156 10.1016/0026-0495(83)90221-4 6298567
56. Sambrook PN Kotowicz M Nash P Styles CB Naganathan V Henderson-Briffa KN Eisman JA Nicholson GC Prevention and treatment of glucocorticoid-induced osteoporosis: a comparison of calcitriol, vitamin D plus calcium, and alendronate plus calcium J Bone Miner Re 2003 18 919 924 10.1359/jbmr.2003.18.5.919
57. de Nijs RN Jacobs JW Lems WF Laan RF Algra A Huisman AM Buskens E de Laet CE Oostveen AC Geusens PP Bruyn GA Dijkmans BA Bijlsma JW Investigators STOP Alendronate or alfacalcidol in glucocorticoid-induced osteoporosis N Engl J Med 2006 355 675 684 10.1056/NEJMoa053569 16914703
58. Kanis JA Johansson H Oden A McCloskey EV Guidance for the adjustment of FRAX according to the dose of glucocorticoids Osteoporos Int 2011 22 809 816 10.1007/s00198-010-1524-7 21229233
59. Saag KG Shane E Boonen S Marín F Donley DW Taylor KA Dalsky GP Marcus R Teriparatide or alendronate in glucocorticoid-induced osteoporosis N Engl J Med 2007 357 2028 2039 10.1056/NEJMoa071408 18003959
60 Allen CS Yeung JH Vandermeer B Homik J Bisphosphonates for steroid-induced osteoporosis Cochrane Database Syst Rev 2016 10 CD001347 10.1002/14651858.CD001347.pub2 27706804
61. MacLean C Newberry S Maglione M McMahon M Ranganath V Suttorp M Mojica W Timmer M Alexander A McNamara M Desai SB Zhou A Chen S Carter J Tringale C Valentine D Johnsen B Grossman J Systematic review: comparative effectiveness of treatments to prevent fractures in men and women with low bone density or osteoporosis Ann Intern Med 2008 148 197 213 10.7326/0003-4819-148-3-200802050-00198 18087050
62. Reid IR Wattie DJ Evans MC Stapleton JP Testosterone therapy in glucocorticoid-treated men Arch Intern Med 1996 156 1173 1177 10.1001/archinte.1996.00440100065008 8639011
63. Crawford BA Liu PY Kean MT Bleasel JF Handelsman DJ Randomized placebo-controlled trial of androgen effects on muscle and bone in men requiring long-term systemic glucocorticoid treatment J Clin Endocrinol Metab 2003 88 3167 3176 10.1210/jc.2002-021827 12843161
64 Plotkin LI Weinstein RS Parfitt AM Roberson PK Manolagas SC Bellido T Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin J Clin Invest 1999 104 1363 1374 10.1172/JCI6800 10562298
65. Weinstein RS Chen JR Powers CC Stewart SA Landes RD Bellido T Jilka RL Parfitt AM Manolagas SC Promotion of osteoclast survival and antagonism of bisphosphonate-induced osteoclast apoptosis by glucocorticoids J Clin Invest 2002 109 1041 1048 10.1172/JCI14538 11956241
66. Adachi JD Saag KG Delmas PD Liberman UA Emkey RD Seeman E Lane NE Kaufman JM Poubelle PE Hawkins F Correa-Rotter R Menkes CJ Rodriguez-Portales JA Schnitzer TJ Block JA Wing J McIlwain HH Westhovens R Brown J Melo-Gomes JA Gruber BL Yanover MJ Leite MO Siminoski KG Nevitt MC Sharp JT Malice MP Dumortier T Czachur M Carofano W Daifotis A Two-year effects of alendronate on bone mineral density and vertebral fracture in patients receiving glucocorticoids: a randomized, double-blind, placebo-controlled extension trial Arthritis Rheum 2001 44 202 211 10.1002/1529-0131(200101)44:1<202::AID-ANR27>3.0.CO;2-W 11212161
67. Stoch SA Saag KG Greenwald M Sebba AI Cohen S Verbruggen N Giezek H West J Schnitzer TJ Once-weekly oral alendronate 70 mg in patients with glucocorticoid-induced bone loss: a 12-month randomized, placebo-controlled clinical trial J Rheumatol 2009 36 1705 1714 10.3899/jrheum.081207 19487264
68. Axelsson KF Nilsson AG Wedel H Lundh D Lorentzon M Association between alendronate use and hip fracture risk in older patients using oral prednisolone JAMA 2017 318 146 155 10.1001/jama.2017.8040 28697254
69. Reid DM Hughes RA Laan RF Sacco-Gibson NA Wenderoth DH Adami S Eusebio RA Devogelaer JP Efficacy and safety of daily risedronate in the treatment of corticosteroid-induced osteoporosis in men and women: a randomized trial. European Corticosteroid-Induced Osteoporosis Treatment Study J Bone Miner Res 2000 15 1006 1013 10.1359/jbmr.2000.15.6.1006 10841169
70. Eastell R Devogelaer JP Peel NF Chines AA Bax DE Sacco-Gibson N Nagant de Deuxchaisnes C Russell RG Prevention of bone loss with risedronate in glucocorticoid-treated rheumatoid arthritis patients Osteoporos Int 2000 11 331 337 10.1007/s001980070122 10928223
71. Reid DM Devogelaer JP Saag K Roux C Lau CS Reginster JY Papanastasiou P Ferreira A Hartl F Fashola T Mesenbrink P Sambrook PN HORIZON investigators, Zoledronic acid and risedronate in the prevention and treatment of glucocorticoid-induced osteoporosis (HORIZON): a multicentre, double-blind, double-dummy, randomised controlled trial Lancet 2009 373 1253 1263 10.1016/S0140-6736(09)60250-6 19362675
72. Hermus AR Smals AG Swinkels LM Huysmans DA Pieters GF Sweep CF Corstens FH Kloppenborg PW Bone mineral density and bone turnover before and after surgical cure of Cushing’s syndrome J Clin Endocrinol Metab 1995 80 2859 2565 10.1210/jcem.80.10.7559865 7559865
73. Compston J Cooper A Cooper C Gittoes N Gregson C Harvey N Hope S Kanis JA McCloskey EV Poole KES Reid DM Selby P Thompson F Thurston A Vine N National Osteoporosis Guideline Group (NOGG) UK clinical guideline for the prevention and treatment of osteoporosis Arch Osteoporos 2017 12 43 10.1007/s11657-017-0324-5 28425085
74. Suzuki Y Glucocorticoid and Bone. Updated Japanese guidelines for the management of glucocorticoid-induced osteoporosis Clin Calcium 2014 24 1309 1318 25177003
75. 2022 American College of Rheumatology guideline for the prevention and treatment of glucocorticoid-induced osteoporosis. Published online September 14, 2022. https://www.rheumatology.org/Portals/0/Files/Prevention-Treatment-GIOP-Guideline-Summary.pdf. Accessed 21 Sept 2022
76. Jiang B Feng C Li C Tu C A bibliometric and visualization analysis of glucocorticoid-induced osteoporosis research from 2012 to 2021 Front Endocrinol (Lausanne) 2022 13 961471 10.3389/fendo.2022.961471 35992120
77. Morikawa T Sakuma M Nakamura T Sonoyama T Matsumoto C Takeuchi J Ohta Y Kosaka S Morimoto T Effectiveness of a computerized clinical decision support system for prevention of glucocorticoid-induced osteoporosis Sci Rep 2022 12 14967 10.1038/s41598-022-19079-7 36056121
78. Takahata M Shimizu T Yamada S Yamamoto T Hasegawa T Fujita R Kobayashi H Endo T Koike Y Amizuka N Todoh M Okumura J Kajino T Iwasaki N Bone biopsy findings in patients receiving long-term bisphosphonate therapy for glucocorticoid-induced osteoporosis J Bone Miner Metab 2022 40 613 622 10.1007/s00774-022-01323-9 35333984
|
PMC010xxxxxx/PMC10180672.txt |
==== Front
Rheumatol Int
Rheumatol Int
Rheumatology International
0172-8172
1437-160X
Springer Berlin Heidelberg Berlin/Heidelberg
37173546
5343
10.1007/s00296-023-05343-0
Epidemiology of RMDs
Trends in the incidence of musculoskeletal diseases in Kazakhstan in 2011–2020: an information-analytical study
http://orcid.org/0000-0003-2511-6918
Yessirkepov Marlen 1
http://orcid.org/0000-0002-9651-7295
Bekaryssova Dana bekaryssova.da@gmail.com
1
http://orcid.org/0000-0001-8336-0711
Mutalipova Gulmira 1
http://orcid.org/0000-0002-4933-0153
Narkabulov Aidynbek 2
1 grid.443628.f 0000 0004 1799 358X Department of Biology and Biochemistry, South Kazakhstan Medical Academy, Shymkent, Kazakhstan
2 Arys Central District Hospital, Turkestan Region, Arys, Kazakhstan
12 5 2023
2023
43 8 15411545
21 2 2023
5 5 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
According to the World Health Organization, there is an increase in the incidence of musculoskeletal diseases worldwide. The problem of this group of diseases is that they are associated with the onset of temporary and permanent disability. A number of studies have demonstrated an increase in the incidence of musculoskeletal diseases in the US, Canada, Australia, and European countries. The current informational and analytical study was aimed to reflect on related morbidity trends in Kazakhstan. We analyzed data on the incidence of diseases of the musculoskeletal system in 2011–2020. Ten annual statistical yearbooks of the Ministry of Health of Kazakhstan were used to obtain data. The results showed an increase in the total incidence of musculoskeletal diseases of 304,492 cases between 2011 and 2020. Primary incidence of musculoskeletal disorders in the whole population increased by a factor of 1.5. The incidence rate of musculoskeletal diseases increased in the age group over 18 years and in the 0–14 years’ child group. A comparative analysis of morbidity figures for rural and urban populations was also presented. An increase in the incidence of musculoskeletal diseases in both populations was observed. Finally, comparative data analysis on morbidity across Central Asian countries was provided. This information-analytical study shows that the incidence of musculoskeletal disorders is steadily increasing in Kazakhstan. The scientific community should pay attention to this trend to prevent further increases in the incidence of musculoskeletal disorders.
Keywords
Musculoskeletal diseases
Incidence
Morbidity
Statistical yearbook
Kazakhstan
issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIntroduction
The number of people suffering from musculoskeletal disorders is steadily increasing worldwide [1]. Factors influencing this growth are not only related to the global population growth, but are also associated with increased life expectancy and global spread of rheumatic diseases and injuries [2]. According to the World Health Organization data as of February 8, 2021, musculoskeletal disorders comprise an average of 150 different pathologies [3]. This broad group of clinical conditions and diseases includes osteoarthritis, rheumatoid arthritis, psoriatic arthritis, gouty arthritis, ankylosing spondylitis, systemic lupus erythematosus, osteoporosis, fractures, dislocations, and many other entities [3]. Although musculoskeletal disorders are highly prevalent among the elderly, younger adults are also increasingly affected by the same diseases [4, 5]. The absolute number of subjects with musculoskeletal disorders is predicted to increase annually, particularly in developing countries [6].
Steadily increasing rate of temporary and permanent disabilities is a consequence of the global spread of musculoskeletal disorders [7]. Premature disabilities overburden societies with physical and psychological issues and result in economic hardships for individuals, their families, and societies. There are 1.71 billion people worldwide with musculoskeletal disorders [3]. A large number of them suffer from lumbago syndrome (568 million) [3]. The 2nd largest disease group presents with fractures (436 million) [3]. And the 3rd group is represented by subjects with osteoarthritis (343 million) [3].
Rheumatoid arthritis, one of the main autoimmune rheumatic diseases, affects 14 million people worldwide [8]. Overall, rheumatic diseases are spread across countries. In India, rheumatic diseases affect up to 24% of the population [9]. These diseases are among the most common chronic conditions leading to disability in Australia, Canada, Europe, and the US [10]. Joint pain is the most common reason of specialist referrals. At least 47.8 million people in the US suffer from arthritis, with a predicted increase to 60 million by 2020 [11]. Arthritis affects 8 million people in the UK and 108 million people across the European continent [12]. According to official data in Eastern European countries such as Ukraine, the percentage of rheumatic diseases increased by 40% in 1988–1993 [13]. In Bulgaria as of 2016, the number of patients with rheumatic diseases was 1712.1 per 100,000 population [14]. Musculoskeletal diseases are also supposedly a pressing issue in Kazakhstan. Therefore, the aim of this study was to explore related trends in Kazakhstan. We aimed to present the dynamics of musculoskeletal diseases in Kazakhstan in 2011–2020.
Methods
This study is informational and analytical in nature. For the analysis of epidemiological features of diseases of musculoskeletal system in Kazakhstan, we analyzed 10-year statistical data based on statistical yearbooks of the Ministry of Health of Kazakhstan titled—"Population health of the Republic of Kazakhstan and activity of public health organizations" (2011–2020 years) [15]. Statistics on the activities of health care organisations and health indicators in the Republic of Kazakhstan for each year are presented in each compendium. All yearbooks contain 20 sections each of which reflects numerical data of the activities of health-care organizations and health indicators. The indicators in the sections are divided into public, private, and departmental. All the numerical data in these compilations are generated by the Statistics tool of the Republican State Enterprise for "Republican e-Health Centre". All morbidity indicators for the period 2011–2020 belonging to the category 'musculoskeletal and connective tissue diseases' are included in the inclusion criteria. The exclusion criteria are morbidity rates from other disease categories that are not relevant for the period 2011–2020. Two tables, each with a 10-year summary, are generated to group the information obtained from the ten collections into a separate Word document. All the data are presented in Table 1. Morbidity per 100,000 population by sex and place of residence is reported. Morbidity per 100,000 is divided into age groups in Table 2. All statistical data are presented in absolute and relative numbers.Table 1 Gender- and residence-based distribution of the incidence of musculoskeletal diseases in Kazakhstan in 2011–2020 (per 100,000 inhabitants)
Dynamic in 2011–2020
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
The entire population of the country 1616.0 1603.7 1549.0 1503.4 1631.7 1884.3 2022.2 2117.3 2098.5 2086.5
Female population 1619.4 1661.7 1663.7 1640.8 1807.2 2094.7 2290.1 2408.3 2308.8 2336.9
Urban population 2006.7 1965.5 1905.0 1834.6 1974.3 2231.1 2413.3 2594.6 2531.8 2605.6
Rural population 1146.4 1165.1 1114.5 1073.1 1181.0 1421.4 1497.1 1457.8 1450.2 1342.9
Table 2 Age-related incidence of musculoskeletal diseases in 2011–2020 in Kazakhstan (per 100,000 inhabitants)
Dynamic in 2011–2020
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
0 to 14 years 1002.2 966.6 817.0 757.0 752.9 927.3 1078.4 1165.7 1135.7 978.1
15 to 17 years 2895.2 2893.1 2882.1 2796.7 2787.0 3022.3 3032.3 2952.7 2838.0 2317.7
Above 18 years 1743.5 1749.6 1738.2 1711.8 1908.9 2203.8 2352.9 2468.4 2465.7 2553.6
Results
Morbidity of total population of Kazakhstan has increased almost 1.4-fold from 682,585 to 987,077 in 2011–2020. During the same period, the level of overall morbidity in subjects above 18 years has increased from 563,226 to 861,178. An increase in the overall incidence is also noted in 0–14 age group, with an increase of 16,019 cases by 2020. In contrast, in 15–17 age group, the total incidence decreased over the decade from 45,821 to 36,342 cases. Alongside the increase in general morbidity, there has been an increase in primary morbidity by 122,052 since 2012. The dynamic of general and primary morbidity in the population is shown in Fig. 1.Fig. 1 Dynamics of general and primary incidence of musculoskeletal disorders in the population of Kazakhstan. *Statistical data on diseases of the musculoskeletal system according to statistical yearbooks of the Ministry of Health of Kazakhstan titled—“Population health of the Republic of Kazakhstan and activity of public health organizations” (2011–2020 years). *The graphs were made using Excel
The relative incidence rate per 100,000 has increased from 1,616 to 2,086.5 over the study period. At the same time, morbidity of females has increased from 1,619.4 to 2,336.9 per 100,000 over the same period.
The following incidence data are available for urban population: in 2011, the rate is 2006.7 per 100,000; there is a gradual decrease in the incidence rate from 1965.5 to 1834.6 per 100,000 in 2012–2014; a gradual increase from 1974.3 to 2594.6 per 100,000 in 2015–2018; however, there is a slight decline in the incidence are to 2531.8 in 2018–2019 and a further gradual increase to 2605.6 per 100,000 by 2020.
A fluctuating morbidity trend has been identified among rural subjects: from 2011 to 2012–from 1146.4 to 1165.1 per 100,000, from 2013 to 2014—1073.1 per 100,000, from 2015 to 2017—again an increase to 1497.1 per 100,000, and from 2018 a gradual decrease that reached 1342.9 per 100,000 by 2020. The rural and urban morbidity dynamic is presented in Fig. 2.Fig. 2 Dynamics of incidence of musculoskeletal disorders among rural and urban populations in Kazakhstan (per 100 000 inhabitants) *Statistical data on diseases of the musculoskeletal system according to statistical yearbooks of the Ministry of Health of Kazakhstan titled—“Population health of the Republic of Kazakhstan and activity of public health organizations” (2011–2020 years). *The graphs were made using Excel
In comparative terms, the urban incidence rate is 2006.7 per 100,000 in 2011 and rural morbidity is 1,146.4 per 100,000. After a 10-year interval, the urban incidence rate is already 2,605.6 per 100,000 and rural incidence rate 1,342.9 per 100,000. The distribution of morbidity among the urban and rural subjects is shown in Table 1.
The incidence rate of diseases of musculoskeletal system in the age group above 18 years is as follows: 1743.5 per 100,000 in 2011 and 2553.6 per 100,000 in 2020. In the age group 0–14 years, there has been a decrease in the incidence rate during the study period, from 1,002.2 to 978.1 per 100,000 inhabitants. During the same period, the age group 15–17 showed a similar trend, with the incidence rate declining from 2,895.2 to 2,317.7 per 100,000 inhabitants. Data by age group are presented in Table 2.
Discussion
Musculoskeletal diseases are an urgent issue in Kazakhstan. According to the annual statistical yearbooks titled "On the state of health of the population of the Republic of Kazakhstan and the activities of health care organizations" (Ministry of Health of Kazakhstan), there is an increase in morbidity throughout the country. The predominant majority in the age structure are people older than 18 years. This is particularly important for the whole society whose work activities may be associated with increased strain, triggering—work-related musculoskeletal disorders [16]. Based on our results, the overall morbidity incidence rate in Kazakhstan has increased 1.4 times. The primary morbidity rate for the entire population of the country has risen 1.5-fold. In the 10-year time-span, the incidence rate increased in the age group above 18 years and in the child group 0–14 years, while a reduction in the incidence rate was recorded in the 15–17-year-old group. A decline in the incidence was recorded across the country from 2019 to 2020 at the beginning of the COVID-19 pandemic.
Data from other Central Asian countries were obtained to compare with local statistics. According to the Statistical Collection titled—"Health of the Population of the Republic of Tajikistan. 30 Years of State Independence", Tajikistan, like Kazakhstan, has seen an increase in morbidity. While in 2011, the morbidity rate was 3729 cases, by 2020, it had reached 52,483. In 2019, the incidence rate was higher at 64,417 [17]. A notable morbidity dynamic was observed in Kyrgyzstan. According to the National Statistical Committee of the Kyrgyz Republic, the primary incidence of diseases of the musculoskeletal system and connective tissue decreased from 40,276 to 37,751 cases from 2011 to 2020 [18]. However, 55,000 cases were reported in 2019 [18]. The COVID-19 pandemic and related quarantine are the most likely reasons for a sharp decline in the incidence from 2019 to 2020. Apparently, referrals to doctors had declined in the pandemic.
A decrease in domestic and crime-related injuries, which also account for a proportion of musculoskeletal disorders, can also be a big issue. A reverse increase in the incidence of musculoskeletal diseases from 2019 to 2020 could mean an improvement in the diagnostic capacity of health facilities. The consequence of this increase in musculoskeletal diseases is a steady increase in the rate of temporary and permanent disability among the patients [19]. Premature disability, in addition to physical and psychological damage, causes economic damage, primarily to the patients, their families, and ultimately to the whole health-care system and the state.
The limitations of this study are that the incidence rates for 2021, 2022 and 2023 are not reflected in the study. The authors plan to make a new information and analysis study as soon as the new statistical yearbooks are available.
Conclusion
As this informational-analytical study demonstrates, morbidity incidence throughout the country has been steadily increasing over the study period. Musculoskeletal diseases is a priority issue due to the poorly understood etiopathogenesis and progressive course. The issue of timely diagnosis and complexity of therapeutic tactic confound an increasing level of disabilities in the population. This study results draw the attention to this big issue and encourage the scientific community to act jointly to prevent further increases in the incidence of musculoskeletal diseases.
Author contribution
All authors substantively contributed to the data processing and writing. They agreed to be fully accountable for the integrity of all aspects of the work.
Funding
Authors state no funding involved.
Data availability
All data processed for this study are available from the corresponding author upon reasonable request.
Declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Woolf AD Pfleger B Burden of major musculoskeletal conditions Bull World Health Organ 2003 81 9 646 656 14710506
2. Fejer R Ruhe A What is the prevalence of musculoskeletal problems in the elderly population in developed countries? A systematic critical literature review Chiropr Man Therap 2012 20 1 31 10.1186/2045-709X-20-31 23006836
3. WHO (2021) Musculoskeletal health. https://www.who.int/ru/news-room/fact-sheets/detail/musculoskeletal-conditions. Accessed 08 Feb 2021
4. Gonzalez EB Goodwin JS Duthie EH Katz PR Malone ML Chapter 36 - Musculoskeletal disorders Practice of Geriatrics 2007 4 Philadelphia W.B. Saunders 495 509
5. Azabagic S Spahic R Pranjic N Mulic M Epidemiology of musculoskeletal disorders in primary school children in bosnia and herzegovina Mater Sociomed 2016 28 3 164 167 10.5455/msm.2016.28.164-167 27482154
6. Hartvigsen J Hancock MJ Kongsted A Louw Q Ferreira ML Genevay S What low back pain is and why we need to pay attention Lancet 2018 391 10137 2356 2367 10.1016/S0140-6736(18)30480-X 29573870
7. Badley EM Rasooly I Webster GK Relative importance of musculoskeletal disorders as a cause of chronic health problems, disability, and health care utilization: findings from the 1990 Ontario Health Survey J Rheumatol 1994 21 3 505 514 8006895
8. Cieza A Causey K Kamenov K Hanson SW Chatterji S Vos T Global estimates of the need for rehabilitation based on the Global Burden of Disease study 2019: a systematic analysis for the Global Burden of Disease Study 2019 Lancet 2021 396 10267 2006 2017 10.1016/S0140-6736(20)32340-0 33275908
9. Misra DP Agarwal V Negi VS Rheumatology in India: a Bird's eye view on organization, epidemiology, training programs and publications J Korean Med Sci 2016 31 7 1013 1019 10.3346/jkms.2016.31.7.1013 27365996
10. Dunlop DD Manheim LM Yelin EH Song J Chang RW The costs of arthritis Arthritis Rheum 2003 49 1 101 113 10.1002/art.10913 12579600
11. Centers for Disease Control and Prevention (CDC) Prevalence and activity limitations–United States, 1990 MMWR Morb Mortal Wkly Rep 1994 43 24 433 438 8202076
12. VanItallie TB Gout: epitome of painful arthritis Metabolism 2010 10.1016/j.metabol.2010.07.009 20837194
13. Doskaliuk B Zimba O Yatsyshyn R Kovalenko V Rheumatology in Ukraine Rheumatol Int 2020 40 2 175 182 10.1007/s00296-019-04504-4 31898762
14. Georgiev T Stoilov R Bulgarian rheumatology: science and practice in a cost-constrained environment Rheumatol Int 2019 39 3 417 429 10.1007/s00296-018-4202-2 30413925
15. Compilations of the Ministry of Health of the Republic of Kazakhstan (2011–2020) Health of the population of the Republic of Kazakhstan and the activities of healthcare organizations
16. van der Beek AJ Dennerlein JT Huysmans MA Mathiassen SE Burdorf A van Mechelen W van Dieën JH Frings-Dresen MH Holtermann A Janwantanakul P van der Molen HF Rempel D Straker L Walker-Bone K Coenen P A research framework for the development and implementation of interventions preventing work-related musculoskeletal disorders Scand J Work Environ Health 2017 43 6 526 539 10.5271/sjweh.3671 28945263
17. Statistical compilation “Health of the population of the Republic of Tajikistan. 30 years of state independence”. https://www.stat.tj/en/news/publications/health-care-in-the-republic-of-tajikistan. Accessed 02 Feb 2023
18. National Statistical Committee of the Kyrgyz Republic (2023) Healthcare. Number of morbidity with acute and chronic diseases by main diseases groups. http://www.stat.kg/en/statistics/zdravoohranenie/. Accessed 02 Feb 2023
19. Malik KM Beckerly R Imani F Musculoskeletal disorders a universal source of pain and disability misunderstood and mismanaged: a critical analysis based on the model of care Anesth Pain Med 2018 8 6 e85532 10.5812/aapm.85532 30775292
|
PMC010xxxxxx/PMC10182598.txt |
==== Front
J Am Acad Dermatol
J Am Acad Dermatol
Journal of the American Academy of Dermatology
0190-9622
1097-6787
by the American Academy of Dermatology, Inc.
S0190-9622(23)00829-0
10.1016/j.jaad.2023.05.017
Original Article
Risk of autoimmune skin and connective tissue disorders after mRNA-based COVID-19 vaccination
Ju Hyun Jeong MD, PhD a
Lee Ju Yeong MD b
Han Ju Hee MD, PhD c
Lee Ji Hae MD, PhD a
Bae Jung Min MD, PhD a
Lee Solam MD, PhD b∗
a Department of Dermatology, St. Vincent's Hospital, College of Medicine, The Catholic University of Korea, Suwon, Republic of Korea
b Department of Dermatology, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea
c Department of Dermatology, Seoul St. Mary's Hospital College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
∗ Correspondence to: Solam Lee, MD, PhD, Department of Dermatology, Yonsei University Wonju College of Medicine, 20 Ilsan-ro, Wonju 26426, Republic of Korea.
13 5 2023
13 5 2023
3 5 2023
© 2023 by the American Academy of Dermatology, Inc.
2023
American Academy of Dermatology, Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Data on the association between the development of autoimmune diseases and COVID-19 vaccination are limited.
Objective
To investigate the incidence and risk of autoimmune connective tissue disorders following mRNA-based COVID-19 vaccination.
Methods
This nationwide population-based study was conducted in South Korea. Individuals who received vaccination between September 8, 2020-December 31, 2021, were identified. Historical prepandemic controls were matched for age and sex in 1:1 ratio. The incidence rate and risk of disease outcomes were compared.
Results
A total of 3,838,120 vaccinated individuals and 3,834,804 controls without evidence of COVID-19 were included. The risk of alopecia areata, alopecia totalis, primary cicatricial alopecia, psoriasis, vitiligo, anti-neutrophil cytoplasmic antibody-associated vasculitis, sarcoidosis, Behcet disease, Crohn disease, ulcerative colitis, rheumatoid arthritis, systemic lupus erythematosus, systemic sclerosis, Sjogren syndrome, ankylosing spondylitis, dermato/polymyositis, and bullous pemphigoid was not significantly higher in vaccinated individuals than in controls. The risk was comparable according to age, sex, type of mRNA-based vaccine, and cross-vaccination status.
Limitations
Possible selection bias and residual confounders.
Conclusion
These findings suggest that most autoimmune connective tissue disorders are not associated with a significant increase in risk. However, caution is necessary when interpreting results for rare outcomes due to limited statistical power.
Key words
autoimmune disease
connective tissue disease
COVID-19
epidemiology
mRNA
risk
skin disease
vaccination
Abbreviations used
AA alopecia areata
aHR adjusted hazard ratio
ANCA anti-neutrophil cytoplasmic antibody
CI confidence interval
COVID-19 Coronavirus disease 2019
ICD-10 International Classification of Diseases, tenth revision
NHIS National Health Insurance Service
SARS-CoV-2 severe acute respiratory syndrome associated corona virus 2
==== Body
pmc Capsule Summary
• The risk of autoimmune connective tissue diseases after mRNA-based COVID-19 vaccination is not well elucidated.
• This nationwide study found that most autoimmune connective tissue diseases were not significantly increased after mRNA-based COVID-19 vaccination. However, limited statistical power prevented detection of potential risks for some rare outcomes. Nevertheless, results suggest any existing risk is not large. These findings could aid in the evaluation and management of autoimmune manifestations following vaccination for COVID-19.
Introduction
COVID-19 vaccines have been introduced to reduce the impact of corona virus 2 (SARS-CoV-2) infections worldwide, preventing 90% of hospitalizations and deaths and 40% to 65% of symptomatic illnesses.1 The 2 most common COVID-19 vaccine platforms include mRNA and adenovirus vector vaccines based on whether they deliver the genetic material mRNA or spike protein using adenoviruses as vectors. Multiple clinical trials have shown that mRNA-based COVID-19 vaccines are effective and have a tolerable safety profile2 , 3; however, real-world data regarding the safety of these vaccines are lacking.
As vaccination programs have been conducted worldwide, new-onset autoimmune manifestations following COVID-19 vaccination have been reported in several studies.4, 5, 6, 7 Previous studies have suggested that molecular mimicry between vaccines and their adjuvants with self-antigens can perturb self-tolerance and consequently lead to the production of autoantibodies and autoimmune response.8 However, the association between COVID-19 vaccines and the risk of autoimmune connective tissue disorders remains unclear. Therefore, we investigated the risk of autoimmune connective tissue disorders associated with mRNA-based COVID-19 vaccination.
Methods
Data source
We performed a nationwide population-based cohort study using data from the Korea Disease Control and Prevention Agency-COVID-19-National Health Insurance Service (NHIS) cohort. The COVID-19 vaccination registry is managed by the Korean government and provides information on the date, type, and dose of COVID-19 vaccination for all individuals vaccinated in Korea. Korea has a single health care insurance system that covers over 99% of the entire Korean population; hence, the NHIS database provides comprehensive information on socioeconomic status, inpatient and outpatient care, diagnoses of diseases, procedures, and prescriptions of enrolled patients.9
Study design
Our database included approximately 20% of all individuals in Korea. We first identified all individuals who had received at least 1 dose of the mRNA-based COVID-19 vaccination (BNT162b2, Pfizer-BioNTech; mRNA-1273, Moderna) before December 31, 2021, as the primary cohort. Individuals with any evidence of SARS-CoV-2 infection (confirmed by polymerase chain reaction) before December 31, 2021, were excluded from the study. The vaccination cohort was established by extracting 50% of the individuals from the primary cohort. The date of the first dose of the mRNA-based COVID-19 vaccination served as the study index date for the vaccination cohort. For comparison, as the COVID-19 vaccination was conducted nationwide, utilizing individuals who had never received COVID-19 vaccination as controls would be more likely to have a higher risk of selection bias. Therefore, we used historical controls that were observed over the same period of time, which was shifted back by 1 year. A historical control cohort was established, with the remaining 50% of individuals not selected for the vaccination cohort. To ensure that the 2 cohorts had a similar observational period, we randomly assigned the study index date for the historical controls based on the distribution of the study index date of the vaccination group but with a subtraction of 1 year (365 days). The study population was followed up from the study index date until disease diagnosis, emigration, death, or the end of the study period. Observations for the vaccination group ceased on December 31, 2021, and those for the historical cohorts ceased on December 21, 2020.
Outcomes
The incidence and risk of autoimmune and autoinflammatory disease outcomes were assessed during follow-up in patients without history of each outcome before the study index date. The occurrence of the outcome disease was ascertained based on at least 3 medical visits using the corresponding International Classification of Diseases, tenth revision (ICD-10) codes as diagnostic codes. To validate our cohort and analyses, the positive control outcomes of certain diseases that have been reported to be strongly associated with COVID-19 vaccination and negative outcome controls, which are less likely to be associated with COVID-19 vaccination, were set and examined.10 The outcomes and the corresponding ICD-10 codes are summarized in Supplementary Table I, available via Mendeley at https://doi.org/10.17632/4kzy8v78bm.1.
Covariates control
Demographics, socioeconomic status, and comorbidity profiles of the study population were obtained from the NHIS database. Although both the vaccination and historical cohorts were constructed from the primary cohort, there could be considerable differences in baseline characteristics that may be associated with the occurrence of disease outcomes. Therefore, 18 covariates, including demographic, socioeconomic, and comorbidity factors, were balanced using inverse probability weighting. The predefined covariates are listed in Table I .Table I Demographic and health characteristics of the COVID-19 vaccination cohort and the historical control cohort before and after the inverse probability treatment weighting
Preweighting Postweighting
COVID-19 vaccination (N = 3,838,120) Historical control (N = 3,834,804) SMD COVID-19 vaccination, mean or (%) Historical control, mean or (%) SMD
Age, mean, ± SD, y 45.7 ± 18.7 44.8 ± 18.7 0.054 45.3 45.3 0.000
<20, n (%) 307,952 (8.0) 355,898 (9.3) (8.3) (8.9)
20-39, n (%) 1,212,206 (31.6) 1,233,807 (32.2) (32.4) (31.4)
40-59, n (%) 1,317,711 (34.3) 1,312,149 (34.2) (34.2) (34.4)
>60, n (%) 1,000,251 (26.1) 932,950 (24.3) (25.1) (25.3)
Sex, n (%) 0.000 0.000
Male 1,978,161 (51.5) 1,975,860 (51.5) (51.4) (51.4)
Female 1,859,959 (48.5) 1,858,944 (48.5) (48.6) (48.6)
Insurance type, n (%) 0.000 0.000
Standard 3,686,686 (96.1) 3,683,382 (96.1) (96.0) (96.0)
Medicaid 151,434 (3.9) 151,422 (3.9) (4.0) (4.0)
Income level, n (%) 0.000 0.000
Highest 1,202,331 (31.3) 1,203,161 (31.4) (31.5) (31.6)
Higher 1,066,344 (27.8) 1,063,166 (27.7) (28.0) (27.8)
Lower 931,261 (24.3) 931,833 (24.3) (24.5) (24.4)
Lowest 614,555 (16.0) 613,542 (16.0) (16.0) (16.2)
Location of residence, n (%) 0.000 0.000
Metropolitan area 1,703,487 (44.4) 1,702,755 (44.4) (44.5) (44.5)
Rural area 2,134,633 (55.6) 2,132,049 (55.6) (55.5) (55.5)
Underlying disease, n (%)
Hypertension 873,910 (22.8) 809,426 (21.1) 0.040 (22.0) (22.0) 0.000
Diabetes mellitus 480,764 (12.5) 427,834 (11.2) 0.042 (11.9) (11.9) 0.000
Dyslipidaemia 1,170,293 (30.5) 1,041,841 (27.2) 0.073 (28.9) (28.9) 0.000
Atopic dermatitis 61,573 (1.6) 55,219 (1.4) 0.013 (1.5) (1.5) 0.000
Allergic rhinitis 201,309 (5.2) 192,579 (5.0) 0.010 (5.1) (5.1) 0.000
Asthma 84,739 (2.2) 82,440 (2.2) 0.004 (2.2) (2.2) 0.000
Hypothyroidism 138,433 (3.6) 121,613 (3.2) 0.024 (3.4) (3.4) 0.000
Hyperthyroidism 54,618 (1.4) 49,120 (1.3) 0.012 (1.4) (1.4) 0.000
Hashimoto thyroiditis 23,205 (0.6) 20,708 (0.5) 0.008 (0.6) (0.6) 0.000
Vitamin D deficiency 113,309 (3.0) 84,192 (2.2) 0.048 (2.6) (2.6) 0.000
Hepatitis B 60,342 (1.6) 55,555 (1.5) 0.010 (1.5) (1.5) 0.000
Hepatitis C 6159 (0.2) 8597 (0.2) 0.003 (0.2) (0.2) 0.000
HIV infection 189 (0.0) 163 (0.0) 0.000 (0.0) (0.0) 0.000
COVID-19, Coronavirus 2019 disease; HIV, human immunodeficiency virus; SMD, absolute standardized mean difference.
Statistical analyses
The baseline demographic characteristics of the study population are described as frequencies with percentages or means with standard deviations. The propensity score for individuals was estimated as the probability of belonging to the vaccination cohort based on the 18 aforementioned covariates and was used to calculate the inverse probability weight, which is the probability of belonging to the vaccination cohort divided by 1-the probability of being in the vaccination cohort. The covariate balance before and after the application of weights was assessed using standardized mean differences. We estimated the risk of predefined outcomes in the vaccination cohort versus the historical control cohort. Statistical estimates were derived from multivariate Cox proportional hazards analysis after adjusting for all 18 covariates used for inverse probability weighting. For each analysis, individuals who had already been diagnosed with the target outcome at the index date or before were excluded; hence, the analysis included only at-risk individuals. We then conducted subgroup analyses according to age, sex, type of mRNA vaccination (BNT162b2 or mRNA-1273), and history of non-mRNA vaccination prior to mRNA vaccination (AZD1222, AstraZeneca, or Ad26.COV2.S, Janssen). All statistical analyses were performed using the SAS statistical software (version 9.4; SAS Institute, Cary, NC, USA) and R statistical software (version 3.4.1; R Foundation for Statistical Computing, Vienna, Austria) at a significance level of 5%.
Results
Study population
From the entire database, 7,672,924 individuals who were vaccinated with at least 1 dose of the mRNA-based COVID-19 vaccine and had never been diagnosed with COVID-19 were selected (Fig 1 ). Among them, 3,838,120 and 3,834,804 were assigned to the vaccination and historical cohorts, respectively. Baseline demographics of the study population are summarized in Table I. The COVID-19 vaccination profiles of the vaccination cohort are summarized in Supplementary Table II, available via Mendeley at https://doi.org/10.17632/4kzy8v78bm.1. The assessment of covariate balance after the application of inverse probability weighting suggested that the covariates were well-balanced (Supplementary Fig 1, available via Mendeley at https://doi.org/10.17632/4kzy8v78bm.1). The mean follow-up times for the vaccination cohort and the historical cohorts were 100.7 ± 90.3 and 100.7 ± 88.5 days, respectively.Fig 1 Flow chart of study population selection. A total of 3,838,120 individuals who received mRNA-based COVID-19 vaccination and 3,834,804 individuals in the historical control cohort were selected from the Korean Disease Control and Prevention Agency-COVID-19-National Health Insurance Service cohort. COVID-19, Coronavirus disease 2019.
Positive and negative outcomes
Prior to the main analysis, we investigated the risk of positive and negative control outcomes to validate our cohort and capture overdetection bias (Fig 2 ). Our data showed a considerably increased risk of myocarditis (adjusted hazard ratio [aHR], 76.48; 95% confidence interval (CI), 18.67-313.39), pericarditis (aHR, 6.24; 95% CI, 3.80-10.24), and thrombocytopenia (aHR 1.45; 95% CI, 1.31-1.61) in the vaccination cohort compared with the historical control cohort. Negative control outcomes indicated minimal overdetection bias in the vaccination cohort.Fig 2 Risk of incident autoimmune connective tissue disorders in mRNA-based COVID-19 vaccinated cohort compared with the historical control cohort. The incidence rate was expressed as the number of events per 10,000 person-years. The forest plot depicts aHRs and 95% confidence intervals (CIs) in individuals with mRNA-based COVID-19 vaccination compared to historical controls. The hazard estimates were adjusted for all 18 covariates used for inverse probability of treatment weighting. Individuals who had already been diagnosed with the target outcome on or before the index date were excluded from each analysis. aHR, Adjusted hazard ratio; CI, confidence interval; COVID-19, coronavirus disease 2019.
The COVID-19 vaccination cohort versus the historical control cohort
The incidence rate and risk of predefined autoimmune and autoinflammatory diseases were estimated in the vaccinated and control cohorts (Fig 2). The vaccinated individuals did not show increased risk of alopecia areata (AA) (aHR, 0.98; 95% CI, 0.92-1.05), alopecia totalis (aHR, 0.89; 95% CI, 0.69-1.14), primary cicatricial alopecia (aHR, 0.91; 95% CI, 0.70-1.20), psoriasis (aHR, 0.89; 95% CI, 0.81-0.97), vitiligo (aHR, 0.96; 95% CI, 0.84-1.11), anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (aHR, 1.36; 95% CI, 0.60-3.08), sarcoidosis (aHR, 1.25; 95% CI, 0.70-2.24), Behcet disease (aHR, 1.06; 95% CI, 0.75-1.50), Crohn disease (aHR, 0.99; 95% CI, 0.78-1.27), ulcerative colitis (aHR, 0.88; 95% CI, 0.72-1.07), rheumatoid arthritis (aHR, 0.92; 95% CI, 0.87-0.98), systemic lupus erythematosus (aHR 0.75; 95% CI, 0.58-0.99), systemic sclerosis (aHR, 0.82; 95% CI, 0.46-1.45), Sjogren syndrome (aHR, 0.90; 95% CI, 0.74-1.10), ankylosing spondylitis (aHR, 1.00; 95% CI, 0.85-1.17), dermato/polymyositis (aHR, 1.20; 95% CI, 0.77-1.88), and bullous pemphigoid (aHR, 2.15; 95% CI, 0.82-5.61) compared with control.
Subgroup analyses
We further examined the risk of outcome diseases in subgroups of the mRNA-based COVID-19 vaccinated cohort in comparison with those in the historical cohort, stratified by age, sex, type of COVID-19 vaccine, and history of non-mRNA vaccination prior to mRNA vaccination. The risks of outcome diseases were not different according to subgroups by age (<40 years or ≥40 years) and sex (Fig 3 ). Anti-neutrophil cytoplasmic antibody-associated vasculitis was increased in female who had vaccination (aHR 5.09; 95% CI, 1.09-23.68); however, this observation requires careful interpretation due to the very small number of events. Regarding the type of mRNA vaccination, the risk of all outcome diseases did not increase, regardless of whether BNT162b2 or mRNA-1273 was used (Fig 4 ). Cross-vaccination status, defined as any history of non-mRNA vaccination (AZD1222, AstraZeneca; or Ad26.COV2.S, Janssen) prior to mRNA vaccination, was not independently associated with an increased risk of disease outcomes (Supplementary Fig 2, available via Mendeley at https://doi.org/10.17632/4kzy8v78bm.1).Fig 3 Subgroup analyses of the risk of incident autoimmune and autoinflammatory disease outcomes in COVID-19 cohort compared with the historical control cohort according to age and sex. The forest plot depicts aHRs and 95% confidence intervals (CIs) in individuals with mRNA-based COVID-19 vaccination compared to historical controls. Subgroup analyses stratified by age and sex are also presented. The hazard estimates were adjusted for all 18 covariates used for inverse probability of treatment weighting. Individuals who had already been diagnosed with the target outcome on or before the index date were excluded from each analysis. aHR, Adjusted hazard ratio; CI, confidence interval; COVID-19, coronavirus disease 2019.
Fig 4 Subgroup analyses of the risk of incident autoimmune and autoinflammatory disease outcomes in COVID-19 cohort compared with the historical control cohort according to types of mRNA vaccines. The forest plot depicts aHRs and 95% confidence intervals (CIs) in individuals with mRNA-based COVID-19 vaccination compared to historical controls. Subgroup analyses stratified by type of mRNA-based COVID-19 vaccine (BNT162b2 or mRNA-1273) are shown. The hazard estimates were adjusted for all 18 covariates used for inverse probability of treatment weighting. Individuals who had already been diagnosed with the target outcome on or before the index date were excluded from each analysis. aHR, Adjusted hazard ratio; CI, confidence interval; COVID-19, coronavirus disease 2019.
Discussion
This nationwide population-based study extensively examined the risk of developing autoimmune and autoinflammatory diseases in COVID-19 vaccinated individuals in comparison with a historical control cohort. Using real-world data, we found that mRNA-based COVID-19 vaccination was not associated with a significantly increased risk of AA, vitiligo, psoriasis, sarcoidosis, Behcet disease, Crohn disease, ulcerative colitis, rheumatoid arthritis, systemic lupus erythematosus, systemic sclerosis, ankylosing spondylitis, and dermato/polymyositis; these risks were not increased by age, sex, type of mRNA vaccine, or cross-vaccination status. Albeit statistically not significant, the risk of bullous pemphigoid showed tendency to increase. The female vaccinated group showed an increased risk of ANCA-associated vasculitis, although the number of events was very small.
Clinical trials of mRNA-based COVID-19 vaccines have shown that they are effective and have acceptable safety profiles.11 , 12 Despite the numerous contributions of vaccines to public health, their role in triggering autoimmune diseases has been controversial for decades. Development of postvaccination autoimmune diseases could be explained as molecular mimicry, which is a similarity between specific pathogenic elements contained in the vaccine or vaccine adjuvants and specific human proteins.13 Previous literature reported that the mRNA vaccines stimulate a switch in the immune system to chronic inflammation state through continuous production of particular autoantibodies, including complement products, anti-platelet factor 4, and polyethylene glycols.14 , 15 Moreover, some authors suggested that nucleic acids-based vaccines could induce autoimmune diseases mainly through acting as agonists of toll-like receptor-7/8/9 and stimulate innate immunity.16
With the gradual increase in COVID-19 vaccination, increasing number of studies documented new-onset autoimmune and autoinflammatory diseases including AA, vitiligo, Grave's disease, immune thrombocytopenic purpura, autoimmune hepatic diseases, Guillain-Barre syndrome, rheumatoid arthritis, type-1-diabetes mellitus, and rheumatic diseases following vaccination.17, 18, 19, 20 However, most of these cases were limited to case series and were diagnosed based on the temporal relationship of the administration of vaccination and onset of the disease. Therefore, real-world data with a large sample size are required to confirm the true association between vaccines and the onset of autoimmune diseases and not just the result of an increased number of vaccinated people.
Our study found that the risk of AA, vitiligo, psoriasis, sarcoidosis, Behcet disease, Crohn disease, ulcerative colitis, rheumatoid arthritis, systemic lupus erythematosus, systemic sclerosis, ankylosing spondylitis, and dermato/polymyositis was not significantly higher in vaccinated individuals than in controls. These results suggest that the post-vaccination onset of autoimmune diseases could be overestimated due to the intensive observation of patients and physicians regarding the possible side effects after vaccination. In addition, vaccination could act as one of the triggering environmental factors only in certain populations with genetic susceptibility and not in a healthy population. The findings of our study may address some of the public's excessive concerns about vaccination through a real-world population-based study.
We observed an increasing trend in the risk of ANCA vasculitis in female population after the COVID-19 vaccination. Multiple case reports have described that COVID-19 vaccination can be associated with development of ANCA-associated vasculitis, both relapsing and de novo cases.21, 22, 23 ANCA-associated vasculitis refers to a multi-system inflammation of small vessels characterized by the presence of anti-neutrophil cytoplasmic antibodies, and environmental factors such as vaccination or infection can lead to the loss of immune tolerance, formation of neutrophil extracellular traps and production of ANCAs.24 T-lymphocyte responses shifting towards specific subsets and production of cytokines, such as interferon-β, following vaccination could also trigger ANCA-associated vasculitis.25 Checking serum creatinine or urinalysis may be considered in susceptible populations who present with systemic symptoms after vaccination. Further studies are needed to clarify the relationship between ANCA-associated vasculitis and vaccination.
In addition, albeit not significant, the risk of bullous pemphigoid tended to increase following vaccination. The development of bullous pemphigoid-like disease following COVID-19 vaccination has been reported in several studies, potentially related to off-target immune activation.26 The exact mechanism of whether bullous pemphigoid-like disease occurred coincidentally in individuals with pre-existing subclinical autoreactivity or whether the vaccine itself increases the disease risk remains to be elucidated.26 In our study, the number of events was too small to detect the statistical significance which requires careful interpretation. Therefore, further studies would be needed to investigate the potential association between COVID-19 vaccination and those conditions.
First, the strength of our study is the use of nationwide real-world data from a single insurance system covering more than 99% of the population. Second, we selected a historical control cohort to minimize selection bias and validated the data using positive and negative outcomes to address detection bias. Finally, we performed a comprehensive subgroup analysis based on age, sex, type of mRNA vaccine, and cross-vaccination status.
This study has several limitations. The demographic composition comprised a single ethnicity. Although we extracted data from the same primary cohort, a possible selection bias exists because of the historical cohort study design. We aimed to maximize our sample size by using Korean national population data, but even with this approach, we were unable to achieve sufficient statistical power for some rare outcomes. Our data lacked detailed information on individual factors such as genetic susceptibility or underlying diseases. Lastly, the follow-up period was not long enough to assess the long-term side effects of the mRNA vaccines.
Conclusion
This study comprehensively investigated the incidence and risk of autoimmune and autoinflammatory outcomes following mRNA-based COVID-19 vaccination. Overall, we did not observe evidence of a significantly increased risk of most autoimmune or autoinflammatory diseases in the vaccinated group compared to controls, although some of these conditions had small number of events, which should be interpreted with caution. Nonetheless, our findings suggest that any potential risk is likely to be not large. Sex-stratified analysis revealed an increasing trend of ANCA-associated vasculitis in female-vaccinated individuals. Our data should relieve excessive public concern about vaccinations and not discourage clinicians from prescribing COVID-19 vaccines; however, long-term follow-up is necessary.
Conflicts of interest
None disclosed.
This study used the database of the KDCA and the NHIS for policy and academic research. The research number of this study is KDCA-NHIS-2022-1-496. The KDCA is the Korea Disease Control and Prevention Agency, Republic of Korea. The NHIS is the National Health Insurance Service, Republic of Korea.
Funding sources: This research was supported by a fund from the research program of the Korea Medical Institute and a National Research Foundation (NRF) of Korea grant funded by the Korea government (MSIT) (no. 2017R1A5A2015369).
IRB approval status: This study was approved by the Korean National Institute for Bioethics Policy (NHIS-2022-1-496) and a waiver of informed consent was granted owing to the deidentified data used.
==== Refs
References
1 Hajjo R. Sabbah D.A. Tropsha A. Analyzing the systems biology effects of COVID-19 mRNA vaccines to assess their safety and putative side effects Pathogens 11 7 2022 743 35889989
2 Dagan N. Barda N. Kepten E. BNT162b2 mRNA Covid-19 vaccine in a nationwide mass vaccination setting N Engl J Med 384 2021 1412 1423 33626250
3 Hall V.J. Foulkes S. Saei A. COVID-19 vaccine coverage in health-care workers in England and effectiveness of BNT162b2 mRNA vaccine against infection (SIREN): a prospective, multicentre, cohort study Lancet 397 2021 1725 1735 33901423
4 Akinosoglou K. Tzivaki I. Marangos M. Covid-19 vaccine and autoimmunity: awakening the sleeping dragon Clin Immunol 226 2021 108721
5 Garrido I. Lopes S. Simões M.S. Autoimmune hepatitis after COVID-19 vaccine - more than a coincidence J Autoimmun 125 2021 102741
6 Jara L.J. Vera-Lastra O. Mahroum N. Autoimmune post-COVID vaccine syndromes: does the spectrum of autoimmune/inflammatory syndrome expand? Clin Rheumatol 41 2022 1603 1609 35378658
7 Klok F.A. Pai M. Huisman M.V. Vaccine-induced immune thrombotic thrombocytopenia Lancet Haematol 9 2022 e73 e80 34774202
8 Blank M. Israeli E. Gertel S. Molecular mimicry in autoimmunity and vaccinations Anaya J.M. Shoenfeld Y. Rojas-Villarraga A. Autoimmunity: from bench to bedside [internet]. Bogota (Colombia) 2013 El Rosario University Press Chapter 21
9 Cheol Seong S. Kim Y.Y. Khang Y.H. Data resource profile: the national health information database of the national health insurance Service in South Korea Int J Epidemiol 46 2017 799 800 27794523
10 Xie Y. Xu E. Bowe B. Long-term cardiovascular outcomes of COVID-19 Nat Med 28 2022 583 590 35132265
11 Lamb Y.N. BNT162b2 mRNA COVID-19 vaccine: first approval Drugs 81 2021 495 501 33683637
12 Polack F.P. Thomas S.J. Kitchin N. Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine N Engl J Med 383 2020 2603 2615 33301246
13 Segal Y. Shoenfeld Y. Vaccine-induced autoimmunity: the role of molecular mimicry and immune crossreaction Cell Mol Immunol 15 2018 586 594 29503439
14 Chen Y. Xu Z. Wang P. New-onset autoimmune phenomena post-COVID-19 vaccination Immunology 165 2022 386 401 34957554
15 Vojdani A. Kharrazian D. Potential antigenic cross-reactivity between SARS-CoV-2 and human tissue with a possible link to an increase in autoimmune diseases Clin Immunol 217 2020 108480
16 Teijaro J.R. Farber D.L. COVID-19 vaccines: modes of immune activation and future challenges Nat Rev Immunol 21 2021 195 197 33674759
17 May Lee M. Bertolani M. Pierobon E. Alopecia areata following COVID-19 vaccination: vaccine-induced autoimmunity? Int J Dermatol 61 2022 634 635 35107173
18 Birkett L. Singh P. Mosahebi A. Possible associations between alopecia areata and COVID-19 vaccination and infection Aesthet Surg J 42 2022 Np699 Np702 35724419
19 Vera-Lastra O. Ordinola Navarro A. Cruz Domiguez M.P. Two cases of Graves' disease following SARS-CoV-2 vaccination: an autoimmune/inflammatory syndrome induced by adjuvants Thyroid 31 2021 1436 1439 33858208
20 Furer V. Eviatar T. Zisman D. Immunogenicity and safety of the BNT162b2 mRNA COVID-19 vaccine in adult patients with autoimmune inflammatory rheumatic diseases and in the general population: a multicentre study Ann Rheum Dis 80 2021 1330 1338 34127481
21 El Hasbani G. Uthman I. ANCA-associated vasculitis following the first dose of pfizer-BioNTech COVID-19 vaccine Nephron 147 2 2022 103 107 35850104
22 Mai A.S. Tan E.K. COVID-19 vaccination precipitating de novo ANCA-associated vasculitis: clinical implications Clin Kidney J 15 2022 1010 1011 35498903
23 Feghali E.J. Zafar M. Abid S. De-novo antineutrophil cytoplasmic antibody-associated vasculitis following the mRNA-1273 (Moderna) vaccine for COVID-19 Cureus 13 2021 e19616
24 Kitching A.R. Anders H.J. Basu N. ANCA-associated vasculitis Nat Rev Dis Primers 6 2020 71 32855422
25 Shimagami H. Yamaguchi Y. Kato Y. Marked increase of interferon-β after BNT162b2 mRNA vaccination: a case of polyarthritis with pleurisy BMJ Case Rep 15 3 2022 e246533
26 Tomayko M.M. Damsky W. Fathy R. Subepidermal blistering eruptions, including bullous pemphigoid, following COVID-19 vaccination J Allergy Clin Immunol 148 3 2021 750 751 34275656
|
PMC010xxxxxx/PMC10184972.txt |
==== Front
Int J Environ Res
Int J Environ Res
International Journal of Environmental Research
1735-6865
2008-2304
Springer International Publishing Cham
37213715
530
10.1007/s41742-023-00530-0
Research Paper
How Does Health Uncertainty Impact Greenhouse Gas Emissions in European Union Economies? A Blessing in Disguise
http://orcid.org/0000-0001-8944-8896
Ali Sajid sajidali1136@gmail.com
1
Anser Muhammad Khalid 23
1 grid.411501.0 0000 0001 0228 333X School of Economics, Bahauddin Zakariya University, Multan, Pakistan
2 grid.444934.a 0000 0004 0608 9907 Faculty of Business and Management Sciences, The Superior University, Lahore, Pakistan
3 grid.11142.37 0000 0001 2231 800X Putra Business School, UPM, Seri Kembangan, Malaysia
15 5 2023
2023
17 3 4415 4 2022
7 4 2023
9 4 2023
© University of Tehran 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The global outbreak of COVID-19 caused serious threats to public health and economic growth all around the world, but on the other hand, the betterment of the environment took place. How pandemics’ health uncertainty will affect environmental quality is a crucial matter to address. The paper investigates the asymmetric association between pandemics-related health uncertainty and greenhouse gas emissions (GHG) in the top emitter European Union economies (Italy, Germany, France, Poland, Netherlands, Spain, Czech Republic, Belgium, Romania, and Greece). Employing data from 1996 to 2019, a unique approach called ‘Quantile-on-Quantile’, is adopted to evaluate the influence of various quantiles of the health uncertainty on GHG emissions. According to estimates, health uncertainty enhances environmental quality by minimizing GHG in most of our chosen nations at certain quantiles of data, which makes pandemics a blessing in disguise for environmental quality. Additionally, the estimations indicate that the grades of asymmetry between our variables varies by locality, accentuating the requisite for authorities to give specific consideration while executing health uncertainty and environmental quality policies.
Keywords
Pandemics-related health uncertainty
Greenhouse gas emissions
Quantile-on-Quantile estimation
Environmental quality
issue-copyright-statement© University of Tehran 2023
==== Body
pmcIntroduction
The universal outbreak of COVID-19 endangers not merely human health but similarly impedes economic growth and greenhouse gas emissions (GHG) (Tobias et al. 2020; Jafari et al. 2022). COVID-19 is believed to have a detrimental influence on the international economy for two key reasons. First, the dramatic increase in the epidemic throughout the world has significantly increased the unpredictability of economic growth, causing instability in capital and financial markets. Second, in order to contain the propagation of the pandemic, governments have rigorously limited the people and transportation movement, as well as economic activities (Yazdani et al. 2021; Aghashariatmadari et al. 2022), putting strain on economic activities from both the production and consumption sides. Many economists believe that COVID-19 will have a greater economic impact than the 2008 financial crisis. According to ongoing investigations, pollution rises with economic expansion and falls with economic downturns (Nicolini et al. 2022; Tobias et al. 2020). Because COVID-19 causes an economic slowdown, it also reduces pollution by lowering GHG levels.
COVID-19, in reality, has shocked our global economy, causing devastation worse than World War II (Muhammad et al. 2020; Heidarin and Jafari 2021). Restrictions on travel, border closure, and quarantine designed to flatten the pandemic curve have created legitimate concerns about a long-term economic disaster (Menut et al. 2020). Economic crises are often caused by a shock to either supply or demand. However, pandemics-related health uncertainty (PU) disrupted both components, resulting in a disaster from a global perspective (Yazdani et al. 2021). The government enforced safety regulations that limited the peoples’ access to their professional activities, altering output and, eventually, the economy’s aggregate supply. Furthermore, constraints on free moment caused a reduction in the consumption of commodities, leading to a fall in aggregate demand and a direct impact on the environmental quality (EQ) due to reduced GHG (Muhammad et al. 2020; Ayodeji et al. 2022). However, pandemic breakouts have severe environmental repercussions owing to the increased volume of local and clinical waste, which may be hazardous and can spread infections to other society members if not properly managed (Syed et al. 2022). Cheval et al. (2020) claimed that each environmental repercussion may not be favorable. PU degraded the EQ by boosting the amount of wastage that is unable to be recycled, creating huge amounts of organic waste owing to reduced agricultural and fisheries exports, and making it impossible to manage and monitor natural habitats (Zambrano-Monserrate et al. 2020; Amini et al. 2022).
The investigation of the PU-GHG link is complicated. Is it true that PU boosts EQ in European Union (EU) nations? Is the PU-EQ association shows non-linear behavior? What are the policy implications of PU-caused environmental change? Literature study reveals that these are unsolved challenges, and as far as we are aware, a few experimental works addressed the said concerns. The contribution of the present study to earlier available literature might be classified into many groups: In the beginning, several researches regarding the connection between PU and EQ have been organized during previous years (e.g., Tobias et al. 2020; Yazdani et al. 2021; Nicolini et al. 2022). As far as the authors are aware, no prior research has been done on the connection between PU and GHG in the most polluted EU nations. PU is the cause of several communal changes, but its impact on EQ is unidentified. Understanding how pandemic-induced severe behavioral disruptions influence GHG will give crucial information on its relationship with environmental sustainability. As per our knowledge, it is the pioneer investigation to apply the World Pandemic Uncertainty Index (WPUI) proposed by Ahir et al. (2018) to investigate the impact of PU on GHG in the top polluted EU nations.1
Prior works depend on the panel data to find the PU-GHG link, despite of the reality that few other localities do not have validation regarding this kind of link distinctly. Nevertheless, this research utilized the Quantile-on-Quantile (QQ) tool to offer international yet economy-related awareness of the relationship between PU and GHG. The QQ methodology estimates the time-series dependency of every nation individually. Numerous aspects of the PU-GHG relationship make it challenging to study with conventional econometric techniques. Typical parametric estimated values are receptive to deviations and do not tolerate heterogeneous slopes (Shahbaz et al. 2018). Therefore, evaluating the PU’s influence over EQ demands the use of a persuasive econometric method, like QQ, that is reluctant towards deviations and might tackle slopes heterogeneity (Sharif et al. 2020). Many of the prior works on pandemics-environment nexus have included carbon dioxide emissions (CO2) while ignoring GHG as a proxy for EQ. Only CO2 comprises a small portion of the total anthropogenic impact on the environment (Gu et al. 2021). On the other hand, GHG is made up of a variety of emissions such as CO2, CH4, N2O, and SF6, and it might be a beneficial indication to exchange CO2 in appraising EQ (Gu et al. 2018; Gu et al. 2020). Previous research tested negative, positive or neutral signs of parameters throughout the whole sample data. This research, conversely, entails that unique signs (either negative or positive) might be obtained throughout a spread of quantiles. While the economy is in a depression, the PU’s influence might be unlike as compared to when it is flourishing. Likewise, the effect of the increase in PU ranks on the EQ might diverge from that of lesser PU ranks. The complexity and dynamic nature of the relationship between PU and EQ may increase as the intensity of PU grows. We predict an asymmetrical PU-GHG nexus because dispersed features cause non-linear variations (Yu et al. 2022). We might additionally realize co-movements (causalities) at several segments of the data (at tails, mid, or top). Because the PU-GHG correlation changes, our single-economy tool might offer officials of government and policymakers critical nation-related recommendations to achieve economic, political, and social assessments on the top, lower, and medium ranks of PU and GHG.
The present research considers the EU countries on diverse grounds. First, the economies picked by us experience both the consequences of environmental degradation and pandemics (Nicolini et al. 2022). Second, EU countries’ features are associated according to their mutual social, economic, and political structures. Third, we apply the QQ method, which regresses every economy autonomously to handle the slopes heterogeneity and cross-sectional interdependence, as these concerns may create considerable bias and deformations (Sharif et al. 2020). Regional integration is essential in various economies because it fosters social, economic, and financial progress. As history revealed that a nation-related uncertainties can quickly expand to other economies in the zone, as history has demonstrated. Furthermore, the economic sector is dependent on its neighbours’ economic sectors, besides internal and external shocks (Chang et al. 2022). Health-related uncertainties, financial crises, quick policy changes, and political disputes all lead to regime-switching conduct that introduces asymmetry into the PU-GHG connection (Chang et al. 2022; Hartono et al. 2021). Resultantly, the initial stage of our investigation was to observe every economy exclusively with a view to overcome the aforesaid hurdles. Fourth, even with their close relationship, these economies can recurrently be exceptionally not dependent because everyone holds its own arrangements in the efficacy of PU to alter EQ. The experimental model structure cannot manage country-specific heterogeneity without an econometric method like QQ. We hope that the results derived from the present study will offer a more comprehensive overview of the relation among the above said factors that may be challenging to get by utilizing usual econometric tools. Lastly, the research’s findings would facilitate upcoming studies related to PU-EQ links and its repercussions for different economies.
The remaining components of this paper are planned as follows: A review of earlier empirical investigations is offered in Section “Literature Review”. Section “Data and its Description” of the study examines the data, whereas segment 4 sets out the methodology used in the study. Section “Outcomes and Discussion” contains basic and major outcomes along with a discussion regarding outcomes. Section “Concluding Remarks and policy implications” abridges the study by suggesting certain possible policy ramifications.
Literature Review
As climatic variations are the foremost concern in various places globally, there are various studies on the elements that influence EQ (e.g., Ali et al. 2020; Khalid et al. 2021; Guo et al. 2021). Previous empirical research has overlooked the relevance of PU, which is inextricably linked to GHG and EQ (Chu and Le 2021).
Many studies have found that GHG levels in many economies decreased during pandemics, which might help people breathe cleaner air (Tobias et al. 2020; Yazdani et al. 2021; Nicolini et al. 2022). In a recent study, Muhammad et al. (2020) analyzed the pandemic outbreak-EQ association using data from NASA2 and ESA.3 During COVID-19, the quality of air in Italy, Spain, Wuhan and the USA improved by almost 30%. In the same way, Menut et al. (2020) discovered that the pandemic outbreak had an inverse influence on the levels of PM4 and N2O in Western European nations. Tobias et al. (2020) discovered that black carbon and NO2 were decreased by 50% in Spain throughout the lockdown time, whereas PM10 was lowered to some extent. During the lockdown in Barcelona, however, the amount of O3 jumped by more than 50%. Abdullah et al. (2020) revealed that travel restrictions had a substantial influence on the minimization of PM2.5 in Malaysia. The lockdown reduced air pollution by 30% while restricting mobility by nearly 90%. For Iran, Yazdani et al. (2021) observed the positive influence of the pandemic outbreak on EQ.
Syed et al. (2022) analyzed the heterogeneous influences of geopolitical and economic policy unpredictability on CO2 in BRICS nations. Economic policy uncertainty reduced CO2 at low and medium quantiles, while it surged emissions at upper quantiles. On the other hand, the geopolitical uncertainty increased CO2 at low quartiles and plunged it at mid and high-level quantiles. Wang and Su (2020) indicated that the outburst of COVID-19 improved the environment of China and significantly contributed to global CO2 reduction. Watts and Kommenda (2020) and Myllyvirta (2020) discovered that the level of GHG in China reduced during COVID-19 phase. Additionally, Liu et al. (2020) reported a 7.8% reduction in CO2 owing to the consumption of fossil fuels in 2020 compared to 2019 because of COVID-19. Moreover, Zambrano-Monserrate et al. (2020) observed the indirect influence of travel restrictions and lockdowns regarding EQ in the form of reduced noise pollution, air pollution, and marine pollution. Similarly, Dantas et al. (2020) discovered that nitrogen dioxide, carbon monoxide, and particulate matter decreased significantly during COVID-19, while the level of ozone (O3) increased during this period. Similarly, Nicolini et al. (2022) assessed the effectiveness of social restrictions on CO2 in major European cities. The findings revealed reductions in CO2 levels throughout the national lockdowns. Brzezinski (2021) examined the influence of several pandemics i.e. MERS, SARS, H3N2 (Flu), Ebola, Zika, and H1N1 (Swine Flu) on CO2 for the panel of 174 economies. It was observed that past pandemics minimized the level of CO2 by 3.4–3.7%. In the same way, Zscheischleret al. (2017), Gherheș et al. (2021), Hartono et al. (2021), and Cheval et al. (2020) also observed improved EQ during COVID-19.
In contrast to the research listed above, some economists have observed that pandemics or PU have a negative influence on EQ (Zuo 2020; Cheval et al. 2020; Robert 2020). According to Zambrano-Monserrat et al. 2020, hospitals in Wuhan produced 240 metric tons of wastage daily, as related to 50 tons daily before COVID-19. Similarly, domestic wastage has grown as a result of the growing reliance on home delivery and online shopping (Zambrano-Monserrate et al. 2020). For G7 economies, Chu and Le (2021) observed that pollution increased due to economic policy uncertainty during the period 1986–2016. In China, Zuo (2020) appraised the connection between COVID-19 pandemic and medical wastage. It was revealed that nearly 245 tons of medical waste were created per day that was 600% higher than the average value. Moreover, Robert (2020) and Benson et al. (2021) also observed that COVID-19 decreased EQ.
Finally, the existing literature contains a lot of awareness of the impacts of various pandemics on EQ, such as MERS-Cov, SARS, Covid-19, and Ebola. There has not been single research that looked at the impact of PU on GHG. In these circumstances, our research will reduce the discrepancy in the empirical findings by investigating the above-mentioned link.
Data along with Description
Our data set contains two variables. Pandemics-related health uncertainty (PU) is considered an explanatory variable. The dependent variable of this study is GHG, which serves as a proxy for EQ. We assess the association between PU and GHG for the top 10 emitter EU economies.5 The research period spans from 1996 to 2019, according to the availability of data. The data for GHG is taken from the website of World Development Indicator. The dataset for PU is taken from (https://worlduncertaintyindex.com/) created by Ahir et al. (2018).
The WPUI is used to gauge the impact of PU on GHG. The WPUI fluctuates from the WUI in respect of its conceptual foundation and significance. The WUI examines cumulative uncertainty (social, economic, and political risk), while the WPUI solely considers the health uncertainty caused by pandemics (WPUI 2020; Ahir et al. 2018; WUI 2020). The WPUI keeps track of how frequently the official of the Economist Intelligence Unit (EIU) uses the phrase “uncertainty” in terms of pandemics. The WPUI especially estimates the amount of risk arising from universal pandemics like SARS, Avian flu, Ebola, and COVID-19. Table 1 includes the nomenclature for the acronyms and symbols used in this study.Table 1 The taxonomy of symbol and abbreviation
Symbols or abbreviations Narration Symbols or abbreviations Narration
PU Pandemics-related health uncertainty J-B Jarque–Bera
WUI World uncertainty index ADF Augmented Dickey-Fuller
WPUI World pandemic uncertainty index ρϕ quantile loss function
QQ Quantile-on-Quantile Estimation kt Kiloton
QR Quantile regression μtθ Quantile's error term
QC Quantile cointegration τ τth quantile of greenhouse gas emissions
OLS Ordinary Least Squares Supτ |Vn(τ)| Supremum norm value of parameters (α and γ)
GHG Greenhouse gas emissions h Bandwidth parameter
EQ Environmental quality CO2 Carbon dioxide emissions
Figure 1 indicates the PU pattern from 1996Q1 to 2021Q3. The trend line exhibits how WPUI fluctuates with time, reaching its topmost spot in 2021Q1 due to COVID-19 pandemic.Fig. 1 Pandemics-related Health Uncertainty (1996Q1-2021Q3).
Source: Author’s own estimation on the basis of Ahir (2018) and WPUI (2020). The WPUI denotes to the simple mean of WPUI of 141 economies
Econometric Tool
The present section discusses the econometric technique employed in present paper. We utilize quantile-based cointegration test to look at the variables’ relationships in long-run with each other. We also use the QQ technique for econometric estimation.
Quantile Cointegration (QC) Test
Many usual tests of cointegration make use of constant cointegrating vectors, which might be the reason why correlations between variables are not always seen in the long-run (Yu et al. 2022). To avoid estimate-related bias, Xiao (2009) presented a quantile cointegration (QC) test, which incorporates time-based variations along with the influence of multiple explanatory variable’s quantiles on a dependent variable to study long-run relationships in conditionally distributed data. Since endogeneity affects typical cointegration tests, Xiao (2009) modified them to comply with the rules of Saikkonen (1991) by including fragmentary cointegrating residuals.
If α (τ) represents a fixed vector, then we can express the cointegration model as follows:1 Xi=α+α′Yi+∑k=-skΔYi-k′Πk+vi
and2 QτX(XiMiX,Miy)=β(τ)+α(τ)′Yi+∑k=-ssΔYi-k′Πj+Fv-1(τ),
where β(τ) shows a drift term while α(τ) signifies persistent parameters. Fv-1(τ) shows residual for the conditional data series. The component of the cointegration model might be stated as follows:3 QτX(XiMiX,Miy)=β(τ)+α(τ)′Yi+δ(τ)′Yi2+∑k=-ssΔYi-k′Πk+∑k=-ssΔYi-k2Πk+Fv-1(τ)
For the QC test, H0: α (π) = α is used as the default hypothesis acquired with the help of Eq. 3 to find the cointegration coefficients. V^n(τ)=[α^(τ)-α^] represents the null hypothesis (supermum rule) in our research. SupτVn′(τ) is utilized as a test statistic throughout the whole quantile distribution in our study.
Quantile-on-Quantile (QQ) Method
Due to their non-linear distribution, the Quantile-on-Quantile (QQ) tool is taken as the most suitable evaluation for forming a connection between the two variables. The customary QR merely assesses the mean influence of the independent variable on the numerous dependent variables’ quantiles. QQ tool, presented by Sim and Zhou (2015) is as an addition to the conventional QR method to address its limitations. QQ is an excellent choice for studying several aspects regarding association within explanatory and explained variable (Sim and Zhou 2015). This method combines standard QR with non-parametrical estimations. It appraises the effect of the quantiles of PU on the GHG quantiles in order to fix the interdependence concern. As a consequence, the QQ tool is adopted in the present research to ascertain complexities in the PU-EQ link that may be hard to investigate, picking other commonly utilized econometric techniques, like ordinary OLS or QR.
We use the non-parametrical model containing its simple version that is experimentally supported by Abdullah et al. (2020) and Wang and Su (2020) as follows:4 GHGt=αθ(PUt)+μtθ,
where PUt and GHGt represent pandemics-related health uncertainty and greenhouse gas emissions, respectively, over t time period. θth GHG quantile of the conditional distribution is depicted by θ. Due to a lack of previous information on the PU-GHG association, factor load αθ(.) is not known to us. μtθ denotes quantile error term along the θ quantile.
We evaluate Eq. 4 using local linear regression in the locality of PU as follows:5 αθ(PUt)≈αθ(PUτ)+αθ′(PUτ)(PUt-PUτ)
In present situation, αθʹ is a derivative of αθ (PUt) with respect of PUt that is identified as partial impact. αθ(PUτ) and αθʹ(PUτ) represents the function of θ and τ, individually. So, the altered form of the Eq. 5 might be specified as follows:6 αθ(PUt)≈α0(θ,τ)+α1(θ,τ)(PUt-PUτ)
We derive the following QQ regression by substituting Eq. 4 in Eq. 6:7 GHGt=α0θ,τ+α1θ,τPUt-PUτ(∗)+utθ
The QQ model’s functional version is embodied with the help of Eq. (7), which explains the tie between the θth PU quantile and the τth GHG quantile. The conditional PU quantile is denoted by the term (*). The quantile-based linkage between PU and GHG is defined through parameters α0 and α1 and the said parameters are dually-indexed in θ and τ. α0 and α1 values might alter subject to the PU and GHG quantiles. By integrating their unique distributions, Eq. 7 exhibits the basic pattern of dependency between PU and GHG. As a bivariate QQ tool, it has no control variable except PU, but it outperforms other old time-series techniques. It is capable of forecasting the asymmetric link between PU and GHG at both the minimum and maximum quantiles, giving extra authentic and trustworthy results as compared to other frequently used tools (Yu et al. 2022).
The selection of bandwidth is vital as it contributes to understanding the PU-GHG linkage.8 Minδ0δ1∑t=1nρϕGHGt-δ0-δ1PUt-PUτLMn(PUt)-τh,
where is the quantile loss function, while L (.) shows the Gaussian kernel function. The Gaussian kernel weighted parameters are not directly associated with the difference between the PU distribution function and the value of the PU quantile distribution function. h designates the bandwidth parameter. Small-bandwidth estimates exhibit considerable variation, whereas large-bandwidth estimates are skewed (Sim and Zhou 2015). So, it is crucial to strike the right balance between bias and variation. Hence, we agree to take h = 0.05 (5%) as the bandwidth’s limit, followed by Sharif et al. (2020).
Robustness of the QQ Tool
The QQ model may propose us with typical QR estimations by allowing proper predictions for various quantiles of the PU. Despite the fact that its quantile-based coefficients are merely classified by θ, the QR model may anticipate the influence of θth quantile of PU on GHG. Unlike the method of QR, the QQ assesses the effect of the θth PU quantile on the τth GHG quantile and indexes the quantile constituents by both τ and θ, leading to further segmented data. In consequence, by taking the mean of the QQ parameters with τ yields the QR parameters. The QR regression’s slope coefficients are represented by γ1(θ), and it is applied to analyze the impact of PU on various GHG quantiles as follows:9 γ1(θ)≡α^¯1=1s∑τα^1(θ,τ)
In the present case, s = 19 embodies the number of quantiles, whereas τ = [0.05, 0.10, 0.15,…,0.95] represents the range quantile series. For this study, we might test the soundness of the QQ tool by relating the forecasted QR parameters to the QQ regression τ-averaged parameters.
Outcomes along with Discussion
The present section offers initial and key findings of the study.
Preliminary Findings
Table 2 shows the descriptive analysis of dependent variable (GHG) and independent variable (PU).Table 2 Descriptive analysis of PU and GHG
Variable Mean Max Min Std. Dev J-B Stats ADF Level ADFΔ
Panel A: Greenhouse Gas Emissions (kt of CO2 equivalent)
Germany 921245 1083882 806092 75007 2.05* − 1.55 − 5.72*
Italy 488307 559872 399601 54197 2.09* − 1.84 − 4.14*
France 479638 521880 423350 34386 2.52* − 1.90 − 5.86*
Poland 385624 436365 362365 17628 5.98* − 1.57 − 4.42*
Spain 358122 434660 296730 40164 1.76 − 1.73 − 7.65*
Netherlands 199802 230292 178642 13852 2.95* − 1.68 − 4.48*
Czech Republic 135336 150742 120360 9274 2.67* − 1.09 − 5.72*
Belgium 124528 142471 106460 12452 2.87* − 1.13 − 5.72*
Romania 117832 160310 98570 16176 2.54* − 1.98 − 7.82*
Greece 106934 125640 78500 14761 2.35* − 5.70* − 5.25*
Panel B: Pandemics-related Health Uncertainty (PU)
Germany 2.13 16.21 0.00 4.09 80.11* − 1.47 − 4.34*
Italy 2.81 20.20 0.21 4.16 172.05* − 1.77 − 5.81*
France 2.07 16.05 0.00 3.67 77.86* − 1.69 − 5.05*
Poland 1.80 7.57 0.00 1.80 24.87* − 0.93 − 6.78*
Spain 2.38 17.51 0.16 4.36 79.21* − 1.71 − 4.82**
Netherlands 2.36 9.50 0.00 2.67 5.97* − 3.76* − 5.56**
Czech Republic 1.24 7.97 0.10 2.12 51.42* − 1.38 − 5.87*
Belgium 3.73 28.31 0.11 5.98 136.81* − 1.49 − 4.62*
Romania 2.76 15.60 0.00 3.20 117.22* − 1.97 − 5.81*
Greece 5.34 47.12 0.52 9.91 176.75* − 1.54 − 5.64*
* and ** specify the rank of significance at 1% and 5%, correspondingly
In terms of GHG, Germany has highest pollution, having an average value of 921245 kt changing from 806092 to 1083882 kt of GHG. Italy is ranked on second place, having GHG mean value of 488307 kt ranging from 3996001 to 559872 kt. France and Poland are rated third and fourth place containing mean GHG of 479638 and 385624 kt, respectively. Greece has the highest level of PU, with an average value of 5.34, falling between 0.52 and 47.12. Belgium ranks second with a mean or average PU score of 3.73, ranging from 0.11 to 28.31. Italy is positioned third, followed by Romania, Spain, and the Netherlands.
Excluding Spain, which has normal distribution of GHG, the results of the JB test reveal that PU and GHG have non-normal data distributions in our selected countries. Additionally, our selected countries’ non-normal data distribution lends credibility to the rationality of the QQ tool that is the best fit for this situation (Razzaq et al. 2020). It is shown by ADF6 test that the variables are stable at their first difference in most of the nations. As a consequence, a stationary data series is used, as suggested by Shahbaz et al. (2018), by transforming the variables into their first difference.
The correlation coefficients of PU and GHG are substantially interconnected with each other for entire countries, as displayed in Table 3. The probability values show that the coefficients are significant at 1% level. Italy and Germany have the biggest correlation coefficients (− 0.84), accompanied by the Netherlands (− 0.78), Poland (− 0.76), and Romania (− 0.75). PU and GHG are shown to be inversely associated with each other in all economies except the Czech Republic.Table 3 Correlation between PU and GHG
Country Correlation t statistics p value
Germany − 0.84 − 10.75* 0.00
Italy − 0.84 − 11.27* 0.00
France − 0.68 − 8.61* 0.00
Poland − 0.76 − 5.34* 0.00
Spain − 0.53 − 2.81* 0.00
Netherlands − 0.78 − 20.32* 0.00
Czech Republic 0.73 7.43* 0.00
Belgium − 0.74 − 4.15* 0.00
Romania − 0.75 − 4.08* 0.00
Greece − 0.66 − 7.83* 0.00
‘*’ shows the level of significance at 1%
Major Findings
Table 4 demonstrates the QC results. τ signifies the τth PU quantile. The supremum norm parameters (α and γ) indicate the stability of parameters.Table 4 Result of QC Test (PU and GHG)
Country Coefficients Supτ |Vn(τ)| CR1 CR5 CR10
Germany PU vs. GHG α 8318.22 5282.23 3137.08 2535.30
γ 177.64 105.47 53.83 39.37
Italy PU vs. GHG α 68591.87 58357.30 57318.28 54881.72
γ 2455.67 1499.21 1437.47 1432.10
France PU vs. GHG α 1241.79 939.72 544.09 207.73
γ 786.81 587.92 499.90 379.74
Poland PU vs. GHG α 9381.08 7206.07 5907.72 2625.09
γ 253.65 167.46 129.58 99.17
Spain PU vs. GHG α 8755.58 6711.19 4771.13 1476.89
γ 398.18 202.69 103.07 99.19
Netherlands PU vs. GHG α 539.46 326.35 295.72 239.57
γ 287.90 198.85 126.73 99.38
Czech Republic PU vs. GHG α 6238.61 5680.90 4688.52 3788.70
γ 3272.93 2697.78 2186.70 1855.44
Belgium PU vs. GHG α 7119.36 3494.13 3085.19 2226.36
γ 607.41 302.86 217.17 119.13
Romania PU vs. GHG α 3932.98 3767.24 249.59 203.95
γ 160.79 159.82 48.09 45.77
Greece PU vs. GHG α 1836.73 1541.76 1040.75 999.80
γ 946.75 687.65 491.70 346.85
To compute the t-statistics for QC, a set of 19 quantiles ranging from 0.05 to 0.95 is utilized. The corresponding critical limits for the supremum norms at the 1%, 5%, and 10% significance levels are denoted as CR1, CR5, and CR10
The estimations of the QC test explains that the cointegration or long-run relationship between PU and GHG alters throughout the range of various quantiles. It is proved that α and γ (coefficients) have higher supremum norms as compared to related critical limits (CR1, CR5, and CR10), showing that PU and GHG hold a significant non-linear long-run association in whole economies (Fig. 2).Fig. 2 Quantile-on-Quantile (QQ) estimates of the slope parameter α1 (θ, τ). The slope parameter estimates α1 (σ, τ) are provided along z-axis while the PU quantiles are given along x-axis and GHG quantiles along y-axis
In Fig. 1, slope estimations α1(θ, τ) are depicted to highlight the effect of the θth quantile of PU on the τth quantile of GHG by utilizing various values of θ and τ.
In Germany and Romania, the inverse effect of PU on GHG is powerful. A vigorous and inverse connection between PU and GHG is established between the zones, which joins overall PU quantiles having mid–low to high GHG quantiles (0.40–0.95) in Germany and medium to upper quantiles (0.50–0.95) in Romania. This significant negative bond specifies that the PU raises the EQ by decreasing GHG at rising pollution ranks. Though, a weak inverse PU-GHG association is established within the localities that join overall PU quantiles at the bottom to middle GHG quantiles (0.05–0.50) in Romania and lower to medium–low GHG quantiles (0.05–0.35) in Germany. In Italy, the inverse influence of PU on GHG is dominant. A robust and inverse link is established between PU and GHG across the areas that link the medium–low to high PU quantiles with all quantiles of GHG. This mainly significant inverse correlation indicates that PU increases the EQ by reducing pollution at strong amounts of PU. Though, a positive and strong PU-GHG relationship exists in the localities that join medium–low PU quantiles (0.05–0.25) along with whole GHG quantiles. It defines that PU is a cause of increasing pollution by raising the levels of GHG at the lower ranks of PU. The domination of the inverse link of PU with GHG supports the works of Nicolini et al. (2022) and Menut et al. (2020), who noticed that COVID-19 increased EQ in EU region.
France shows the dominance of inverse ties between PU and GHG. An extremely inverse link between PU and GHG is noticed in the vicinity that integrates entire PU quantiles with bottom to medium–low and lower-mid to top GHG quantiles (0.05–0.35 and 0.45–0.95). This powerful inverse linkage specifies that the PU boosts the EQ by curtailing emissions at both bottom and higher pollution levels, though an inverse and weak PU-GHG bond is assessed in the locations that join overall PU quantiles with low-mid GHG quantiles. In Poland, the inverse influence of PU on GHG is eminent. A vigorous and inverse bond between PU and GHG is observed throughout the regions that join whole PU quantiles with bottom to medium–high GHG quantiles (0.05–0.75). This mainly powerful inverse bond entails that the PU improves the EQ by reducing GHG at upper pollution ranks. Though, a strong and positive PU-GHG link is witnessed in the vicinities that combine overall PU quantiles and mid-high GHG quantiles. Though, this powerful positive link turns into a weak inverse connection in the locations that join all PU quantiles with higher GHG quantiles (0.90–0.95). In Spain, there is a significant and inverse bond between PU and GHG among the areas that unite whole PU quantiles with bottom to middle and upper-middle to top GHG quantiles. This significant and inverse relationship implies as the PU raises the EQ by lessening GHG at both higher and lower ranks of pollution. Though, this linkage becomes powerful positive in the locations that join whole quantiles of PU with upper-middle quantiles of GHG (0.55–0.65). The findings are consistent with the study of Schulte-Fischedick et al. (2021), who observed that COVID-19 increased EQ in European nations.
The Netherlands shows a significant negative link between PU and GHG. There is a strong correlation across all GHG quantiles and the medium to higher PU quantiles (0.50–0.95), and this correlation is negative. According to this incredibly strong negative link, the PU improves environment by reducing the volume of GHG at significant quantiles of PU and GHG. A weak inverse bond exists among the areas that join bottom to medium–low PU quantiles with the lower to medium–low GHG quantiles. Though, this weak inverse tie becomes a positive and powerful among the locations that unite low to lower-middle PU quantiles along with medium to upper GHG quantiles (0.50–0.95). There is supremacy of strong inverse PU-GHG association in Belgium. A strong inverse bond between PU and GHG is observed within the regions that join low to higher-middle PU quantiles (0.05–0.80) with all GHG quantiles. This predominantly powerful inverse association entails as the PU boosts the EQ by reducing the amount of GHG in phases of both bottom and medium–high PU levels. Both weak negative and weak positive bonds prevail among the regions that link top PU quantiles with whole GHG quantiles. The domination of the inverse link of PU with GHG is in corroboration with the investigations of Menut et al. (2020) and Nicolini et al. (2022), who observed that COVID-19 enhanced EQ in EU economies.
In Greece, the powerful and inverse correlation prevails between PU and GHG. A persistent inverse tie is established among the locations that integrate whole PU quantiles with the lower to moderately high GHG quantiles (0.05–0.65). This mainly strong negative connection indicates that the PU increases the EQ by reducing GHG at lowest to higher-middle level of pollution. A mixed PU-GHG nexus exists in the areas that bind all PU quantiles along with high-middle GHG quantiles (0.70–0.85). Additionally, this mixed connection turns into a strong positive link within the localities that joins whole quantiles of PU with upper quantiles of GHG (0.90–0.95). It entails that PU deteriorates the environment by enhancing the amount of GHG during periods of higher pollution.
There is a diverse link between PU and GHG in the Czech Republic. A substantial inverse association between PU and GHG is realized in the locations that join whole PU quantiles with the low to medium GHG quantiles (0.05–0.50). This predominantly powerful inverse connection entails that the PU uplift the EQ by lessening the amount of GHG during the phase when GHG is at low to medium ranks. A strong positive bond holds among the regions that join whole PU quantiles along with high-middle to top GHG quantiles (0.55–0.95), which means that PU considerably raises pollution during upper ranks of GHG. The mixed PU–GHG link is endorsed by the finding of Syed et al. (2022), who discovered that economic policy uncertainty decreased CO2 at low and medium quantiles while it surged CO2 at higher quantiles.
Table 5 identifies the link between numerous PU and GHG quantiles for our selected economies based on Fig. 1. Most of the economies we looked at had a high negative correlation between PU and GHG, indicating that PU enhances EQ. In the Czech Republic, however, there is a mixed linkage between PU and GHG.Table 5 Summarized Findings regarding (Relation b/w Various Quantiles of PU and GHG)
Country Quantiles of PU Quantiles of GHG Association b/w Quantiles Dominant Linkage
Germany All quantiles Mid-low to high quantiles Powerful and inverse Powerful and inverse
All quantiles Lowest to lower-middle quantiles Weak and negative
Italy Medium–low to top quantiles Whole quantiles Powerful and inverse Powerful and inverse
Lower-middle quantiles Whole quantiles Powerful and positive
France Whole quantiles Bottom to lower-middle and lower-middle to top quantiles Strong and inverse Powerful and inverse
Whole quantiles Lower-middle quantiles Weak negative
Poland Whole quantiles Bottom to high-middle quantiles Powerful and inverse Powerful and inverse
Whole quantiles Medium–high quantiles Powerful and positive
Whole quantiles Higher quantiles Weak and negative
Spain Whole quantiles Bottom to medium and higher-middle to top quantiles Powerful and inverse Powerful and inverse
Whole quantiles Higher middle quantiles Strong and positive
Netherlands Middle to higher quantiles Whole quantiles Powerful and inverse Powerful and inverse
Bottom to lower–middle quantiles Low to lower-middle quantiles Weak and inverse
Lower to lower-middle quantiles Middle to higher quantiles Powerful and positive
Czech Republic Whole quantiles Low to medium quantiles Powerful and inverse Mixed relationship
Whole quantiles High-middle to top quantiles Powerful and positive
Belgium Low to medium–high quantiles Whole quantiles Powerful and inverse Powerful and inverse
Top quantiles Whole quantiles Mixed relationship
Romania Whole quantiles Medium to high quantiles Strong and inverse Strong and inverse
Whole quantiles Bottom to medium quantiles Weak and inverse
Greece All quantiles Low to higher-middle quantiles Powerful and inverse Strong and inverse
All quantiles High-middle quantiles Mixed relationship
All quantiles High quantiles Powerful and positive
Verification of the Robustness of the QQ Method
The QQ estimated values may be analyzed to check if they are the same. The QQ technique’s prior findings are supported by Fig. 3. The graphs demonstrate that for every nation under consideration, the average QQ and QR estimated values of the slope coefficients reveal a similar trend.Fig. 3 Testing the Robustness of the QQ Approach by Relating QR and QQ Regression Estimates. The estimates of the typical QR parameters and the averaged QQ parameters are shown against various GHG quantiles
Figure 3 displays the PU and GHG heterogeneity in our sample vicinities. Confirming the extent of the coefficients, the influence of PU on GHG is significantly higher in the Netherlands, Italy, Belgium, and Greece, while substantially lower in Spain and France.
Discussion of Results
In most of the sample economies, the outcomes demonstrate an inverse correlation between PU and GHG. The findings of the study corroborate our hypothesis, as do those of other empirical investigations like Abdullah et al. (2020) and Wang and Su (2020), which imply that pandemics improve EQ. The detection of significant and negative PU coefficients gives credit to past policy comments regarding the Paris agreement (COP21), underlying the significance of minimizing GHG. We can compare our results with preliminary estimations of how the pandemics affect pollution levels, as well as strict government measures imposed in response to these pandemics.6 Global emissions declined by 6.2 percent in 2020, according to Carbon Monitor Programs,7 with substantial nation-specific heterogeneity (Nicolini et al. 2022). Watts & Kommenda (2020), Myllyvirta (2020) Yazdani et al. (2021) also support the findings. The outcomes are moderately coherent with those of Nicolini et al. (2022) and Menut et al. (2020) for the EU, Tobias et al. (2020) for Spain, and similarly Chu and Le (2021) for the G-7 countries, who claim that PU improves EQ. The data further corroborate the notion that Covid-19-related reductions of GHG are driven by a reduction in economic activity, which might be induced by the health disasters themselves or the measures by the government that limit human and industrial actions to fight the epidemic.
We, however, detect a diverse link between PU and GHG in a significant number of quantiles in the Czech Republic, which might be attributable to distinguishing traits such as population, growth trends, business cycles, and technology. This corresponds to the conclusions of Syed et al. (2022), who revealed that economic policy uncertainty decreased CO2 at low and medium quantiles, while it surged CO2 at higher quantiles. Overall, current projections of PU-GHG are relatively comparable to past pandemic estimations. The cumulative GHG reduction associated with COVID-19 is expected to be transitory and will dissipate in the near future, although pandemic-linked uncertainty is expected to have slightly wider exposure than prior pandemics analyzed individually in many prior studies. Our estimates are quite related to those observed for the effect of pandemics on economic performance (Goswami et al. 2021), financial market (Sharif et al. 2020), economic growth (Salisu et al. 2020), energy usage (Li et al. 2022), and investment (Sharma et al. 2020).
On several grounds, the influence of PU fluctuates significantly among quantiles and throughout the sample localities. For instance, high GHG quantiles show a powerful negative bond with PU (for Spain, Germany, Romania, and France). The differences in the PU impact amongst sample economies might be determined by the economic situations of economies picked by us. In the context of population, technology, and potential for economic growth, France and Germany, for instance, stand out among the other chosen economies. In a locality, overlooking this sort of heterogeneity could lead to erroneous findings. PU and GHG slope coefficients differ among vicinities, showing that the PU-Environment relationship is not persistent along discrete high and low data quantiles but rather linked to the frequency and severity of economic shocks as well as the specific economic stage that has an influence on PU.
Concluding Remarks and Policy Implications
We observed the non-linear bond between PU and GHG in the topmost emitter localities of the EU (Germany, Italy, France, Poland, Spain, Netherlands, Czech Republic, Belgium, Romania, and Greece). Using panel data from 1996 to 2019, the ‘QQ’ method is employed that permits researchers to autonomously explore dependency in every economy in terms of offering international but locality-specific evidence on the bond between the variables. Estimates reveal that PU expands EQ by reducing GHG in most of our chosen economies at certain quantiles of data distribution.
Pandemic outbreaks have major implications, which might be utilized to reconsider individual and communal decisions and objectives. Most modern architecture demonstrates how people have adapted to contagious illnesses by changing their workspaces. During the COVID-19 outbreak, the utilization of media and webinars for information dissemination and skills gained significant use. When we increasingly work from home, study and update skills online, and buy essentials on e-commerce sites, we replace conventional physical venues with virtual ones that can be visited from smart/digital devices, resulting in less transportation and hence less GHG. The growing dependence on digital platforms in the built environment might persist for a long time after the pandemic, affecting all aspects of the layout and urban planning. Human civilization is facing a worldwide crisis, maybe the worst of our generation. Many pandemic-related measures will become a part of everyday life, influencing behaviors and routines, and may have a beneficial effect on EQ. The pandemics have exposed our inability to govern our environment and have provided some lessons from this forced experiment. One of the most significant lessons we will learn is the importance of having a network of streets for cycling and walking. Walking has been demonstrated to be ecologically friendly as well as beneficial to humans’ physical and mental betterment as a major form of transportation and physical activity. Streets may need to be modified to meet the needs of multi-modal mobility, resulting in healthier, safer, and greener societies. While public transportation is a good choice for decreasing pollution, it is not appropriate during a pandemic because it may contribute to the transmission of illness among passengers.
National decision-makers and international organizations are being encouraged to take action in order to reverse the growing tendency in worldwide GHG and to break the relationship between growth and GHG. International collaboration is crucial for managing pandemic crises like COVID-19. Depending on their national conditions, various nations actively react to pandemics. Developed and developing nations must exchange information, as well as learn from complementing each other’s capabilities. Countries may pool their pandemic and environmental research abilities and resources to form a robust and worldwide scientific research coordination force able to respond to pandemic issues and boost EQ. To promote long-term growth, governments should emphasize policies with multilateral environmental limitations, as well as R&D investment and severe environmental laws. Nobody can be completely protected from the indirect effects of these disasters in any part of the world since humans share the same planet (Gu et al. 2019). Meanwhile, until quick coordinated action is taken, no country will be spared from the disastrous effects of climate change.
Last, this study has a few shortcomings that would set the stage for potential researchers. Due to the limitations of the QQ model, we have taken just the total GHG while neglecting its various sub-components (like N2O, CO2, CH4, SO2, and CH4). These metrics could be utilized in upcoming research to evaluate how the results vary across different environmental factors. In addition, future studies might see at the impact of different forms of uncertainties on EQ, such as policy uncertainty, trade uncertainty, and overall uncertainty. Another prominent disadvantage of this research is the utilization of the bivariate QQ technique that precludes the addition of other control elements that alter the effect of PU on GHG. As a consequence, in future research, we might improve our model by using multivariate quantile-based approaches (like Quantile ARDL) to well perceive the relationship with extra independent variables.
Author Contributions
SA: Conceptualization; Writing-original draft; Methodology; Software. MKA: Writing-original draft; Writing-review.
Funding
No funding is received for conducting this study.
Data Availability
The data that support the findings of this study are available on request from the corresponding author.
Declarations
Conflict of Interest
The corresponding author declares that there is no conflict of interest with this work on behalf of all coauthors.
1 Germany, Italy, France, Poland, Spain, Netherlands, Czech Republic, Belgium, Romania, and Greece.
2 National Aeronautics and Space Administration.
3 European Space Agency.
4 Particulate Matter.
5 Germany, Italy, France, Poland, Spain, Netherlands, Czech Republic, Belgium, Romania, and Greece.
6 See Menut et al. (2020) and Nicolini et al. (2022).
7 https://carbonmonitor.org/.
==== Refs
References
Abdullah S Napi NNLM Ahmed AN Mansor WNW Mansor AA Ismail M Ramly ZTA Development of multiple linear regression for particulate matter (PM10) forecasting during episodic transboundary haze event in Malaysia Atmosphere 2020 11 3 289 10.3390/atmos11030289
Aghashariatmadari Z Golmohammadian H Shariatmadari Z Mohebbi F Bazrafshan J Satellite-based monitoring of the algal communities of aras dam reservoir: meteorological dependence analysis and the footprint of COVID-19 pandemic lockdown on the eutrophication status Int J Environ Res 2022 16 5 1 18 10.1007/s41742-022-00447-0
Ahir H Bloom N Furceri D The world uncertainty index SSRN 2018 10.2139/ssrn.3275033
Ali S Yusop Z Kaliappan SR Chin L Dynamic common correlated effects of trade openness, FDI, and institutional performance on environmental quality: evidence from OIC countries Environ Sci Pollut Res 2020 27 11671 11682 10.1007/s11356-020-07768-7
Amini Z Malekmohammadi B Jafari H Developing a Wetland management framework based on ecological approach (case study: Anzali international Wetland) J Environ Sci Technol 2022 113 119
Ayodeji OJ Hilliard TA Ramkumar S Particle-size-dependent filtration efficiency, breathability, and flow resistance of face coverings and common household fabrics used for face masks during the COVID-19 pandemic Int J Environ Res 2022 16 1 1 12 10.1007/s41742-021-00390-6
Benson NU Bassey DE Palanisami T COVID pollution: impact of COVID-19 pandemic on global plastic waste footprint Heliyon 2021 7 2 e06343 10.1016/j.heliyon.2021.e06343 33655084
Brzezinski M The impact of past pandemics on CO2 emissions and transition to renewable energy SSRN 2021 10.2139/ssrn.3837444
Chang L Saydaliev HB Meo MS Mohsin M How renewable energy matter for environmental sustainability: evidence from top-10 wind energy consumer countries of European Union Sustain Energy Grids Netw 2022 31 100716 10.1016/j.segan.2022.100716
Cheval S Mihai Adamescu C Georgiadis T Herrnegger M Piticar A Legates DR Observed and potential impacts of the COVID-19 pandemic on the environment Int J Environ Res Public Health 2020 17 11 4140 10.3390/ijerph17114140 32532012
Chu LK Le NTM Environmental quality and the role of economic policy uncertainty, economic complexity, renewable energy, and energy intensity: the case of G7 countries Environ Sci Pollut Res 2021 29 1 17
Dantas G Siciliano B França BB da Silva CM Arbilla G The impact of COVID-19 partial lockdown on the air quality of the city of Rio de Janeiro Braz Sci Total Environ 2020 729 139085 10.1016/j.scitotenv.2020.139085
Gherheș V Cernicova-Buca M Fărcașiu MA Palea A Romanian students’ environment-related routines during COVID-19 home confinement: water, plastic, and paper consumption Int J Environ Res Public Health 2021 18 15 8209 10.3390/ijerph18158209 34360502
Goswami B Mandal R Nath HK Covid-19 pandemic and economic performances of the states in India Econ Anal Policy 2021 69 461 479 10.1016/j.eap.2021.01.001
Gu K Qiao J Lin W Recurrent air quality predictor based on meteorology-and pollution-related factors IEEE Trans Industr Inf 2018 14 9 3946 3955 10.1109/TII.2018.2793950
Gu K Xia Z Qiao J Lin W Deep dual-channel neural network for image-based smoke detection IEEE Trans Multimedia 2019 22 2 311 323 10.1109/TMM.2019.2929009
Gu K Zhang Y Qiao J Ensemble meta-learning for few-shot soot density recognition IEEE Trans Industr Inf 2020 17 3 2261 2270 10.1109/TII.2020.2991208
Gu K Liu H Xia Z Qiao J Lin W Thalmann D PM2.5 monitoring: use information abundance measurement and wide and deep learning IEEE Trans Neural Netw Learn Syst 2021 32 10 4278 4290 10.1109/TNNLS.2021.3105394 34460393
Guo J Zhou Y Ali S Shahzad U Cui L Exploring the role of green innovation and investment in energy for environmental quality: an empirical appraisal from provincial data of China J Environ Manage 2021 292 112779 10.1016/j.jenvman.2021.112779 34022650
Hartono D Yusuf AA Hastuti SH Saputri NK Syaifudin N Effect of COVID-19 on energy consumption and carbon dioxide emissions in indonesia Sustain Prod Consum 2021 28 391 10.1016/j.spc.2021.06.003 37274733
Heidari M Jafari H Challenges of COVID-19 vaccination in Iran: In the fourth wave of pandemic spread Prehosp Disaster Med 2021 36 5 659 660 10.1017/S1049023X21000777 34287113
Jafari H, Mousavi SM, Ghoshouni H, Nemati M, Kelishadrokhi SF (2022) Strengthening the primary health care system in the face of emerging and re-emerging epidemics—provide a native and Applicable model based on best practices of countries. Disaster and Emergency Medicine Journal
Khalid K Usman M Mehdi MA The determinants of environmental quality in the SAARC region: a spatial heterogeneous panel data approach Environ Sci Pollut Res 2021 28 6 6422 6436 10.1007/s11356-020-10896-9
Li Z Ye H Liao N Wang R Qiu Y Wang Y Impact of COVID-19 on electricity energy consumption: a quantitative analysis on electricity Int J Electr Power Energy Syst 2022 140 108084 10.1016/j.ijepes.2022.108084
Liu Z Ciais P Deng Z Lei R Davis SJ Feng S Schellnhuber HJ Near-real-time monitoring of global CO2 emissions reveals the effects of the COVID-19 pandemic Nat Commun 2020 11 5172 10.1038/s41467-020-18922-7 33057164
Menut L Bessagnet B Siour G Mailler S Pennel R Cholakian A Impact of lockdown measures to combat Covid-19 on air quality over western Europe Sc Total Environ 2020 741 140426 10.1016/j.scitotenv.2020.140426 32593893
Muhammad S Long X Salman M COVID-19 pandemic and environmental pollution: a blessing in disguise? Sci Total Environ 2020 728 138820 10.1016/j.scitotenv.2020.138820 32334164
Myllyvirta L (2020) Coronavirus temporarily reduced China’s CO2 emissions by a quarter. Carbon Brief. https://www.carbonbrief.org/analysis-coronavirus-has-temporarily-reduced-chinas-co2-emissions-by-a-quarter. Accessed 30 Jun 2022
Nicolini G Antoniella G Carotenuto F Christen A Ciais P Feigenwinter C Papale D Direct observations of CO2 emission reductions due to COVID-19 lockdown across European urban districts Sci Total Environ 2022 830 154662 10.1016/j.scitotenv.2022.154662 35318060
Razzaq A Sharif A Aziz N Irfan M Jermsittiparsert K Asymmetric link between environmental pollution and COVID-19 in the top ten affected states of US: a novel estimations from quantile-on-quantile approach Environ Res 2020 191 110189 10.1016/j.envres.2020.110189 32919963
Robert A Lessons from New Zealand’s COVID-19 outbreak response Lancet Public Health 2020 5 11 e569 e570 10.1016/S2468-2667(20)30237-1 33065024
Saikkonen P Asymptotically efficient estimation of cointegration regressions Economet Theor 1991 7 1 1 21 10.1017/S0266466600004217
Salisu A, Gupta R, Demirer R (2020) A note on uncertainty due to infectious diseases and output growth of the United States: a mixed-frequency forecasting experiment (Working Paper No. 2020-50).
Schulte-Fischedick M Shan Y Hubacek K Implications of COVID-19 lockdowns on surface passenger mobility and related CO2 emission changes in Europe Appl Energy 2021 300 117396 10.1016/j.apenergy.2021.117396 34305265
Shahbaz M Zakaria M Shahzad SJH Mahalik MK The energy consumption and economic growth nexus in top ten energy-consuming countries: fresh evidence from using the quantile-on-quantile approach Energy Econom 2018 71 282 301 10.1016/j.eneco.2018.02.023
Sharif A Aloui C Yarovaya L COVID-19 pandemic, oil prices, stock market, geopolitical risk and policy uncertainty nexus in the US economy: fresh evidence from the wavelet-based approach Int Rev Financ Anal 2020 70 101496 10.1016/j.irfa.2020.101496
Sharma P Leung TY Kingshott RP Davcik NS Cardinali S Managing uncertainty during a global pandemic: an international business perspective J Bus Res 2020 116 188 192 10.1016/j.jbusres.2020.05.026 32501304
Sim N Zhou H Oil prices, US stock return, and the dependence between their quantiles J Bank Financ 2015 55 1 8 10.1016/j.jbankfin.2015.01.013
Syed QR Bhowmik R Adedoyin FF Alola AA Khalid N Do economic policy uncertainty and geopolitical risk surge CO2 emissions? New insights from panel quantile regression approach Environ Sci Pollut Res 2022 29 19 27845 27861 10.1007/s11356-021-17707-9
Tobias A Carnerero C Reche C Massagué J Via M Minguillón MC Querol X Changes in air quality during the lockdown in Barcelona (Spain) one month into the SARS-CoV-2 epidemic Sci Total Environ 2020 726 138540 10.1016/j.scitotenv.2020.138540 32302810
Wang Q Su M A preliminary assessment of the impact of COVID-19 on environment–a case study of China Sci Total Environ 2020 728 138915 10.1016/j.scitotenv.2020.138915 32348946
Watts J, Kommenda N (2020) Coronavirus pandemic leading to huge drop in air pollution. The Guardian. https://www.theguardian. Accessed 4 Apr 2021
World Pandemic Uncertainty Index (WPUI). 2020. World Pandemic Uncertainty Index (WPUI): Country. Available online: https://worlduncertaintyindex.com/wp-content/uploads/2020/10/WPUI_Data.xlsx. Accessed on 8 Mar 2022
World Uncertainty Index (WUI). 2020. World Uncertainty Index (WUI): Country. Available online: https://worlduncertaintyindex.com/wp-content/uploads/2020/10/WUI_Data.xlsx. Accessed on 8 Mar 2022
Xiao Z Quantile cointegrating regression J Econom 2009 150 2 248 260 10.1016/j.jeconom.2008.12.005
Yazdani M Baboli Z Maleki H Birgani YT Zahiri M Chaharmahal SSH Goudarzi G Contrasting Iran’s air quality improvement during COVID-19 with other global cities J Environ Health Sci Eng 2021 19 2 1801 1806 10.1007/s40201-021-00735-5 34493956
Yu J Tang YM Chau KY Nazar R Ali S Iqbal W Role of solar-based renewable energy in mitigating CO2 emissions: evidence from quantile-on-quantile estimation Renew Energy 2022 182 216 226 10.1016/j.renene.2021.10.002
Zambrano-Monserrate MA Ruano MA Sanchez-Alcalde L Indirect effects of COVID-19 on the environment Sci Total Environ 2020 728 138813 10.1016/j.scitotenv.2020.138813 32334159
Zscheischler J Mahecha MD Avitabile V Calle L Carvalhais N Ciais P Reichstein M Reviews and syntheses: an empirical spatiotemporal description of the global surface–atmosphere carbon fluxes: opportunities and data limitations Biogeosciences 2017 14 15 3685 3703 10.5194/bg-14-3685-2017
Zuo M (2020) Coronavirus leaves China with mountains of medical waste. South China Morning Post
|
PMC010xxxxxx/PMC10185369.txt |
==== Front
Biosens Bioelectron
Biosens Bioelectron
Biosensors & Bioelectronics
0956-5663
1873-4235
Elsevier B.V.
S0956-5663(23)00344-5
10.1016/j.bios.2023.115402
115402
Article
CoLAMP: CRISPR-based one-pot loop-mediated isothermal amplification enables at-home diagnosis of SARS-CoV-2 RNA with nearly eliminated contamination utilizing amplicons depletion strategy
Cao Yumeng a
Lu Xiao a
Lin Haosi a
Rodriguez Serrano Alan Fernando a
Lui Grace C.Y. b
Hsing I-Ming a∗
a Department of Chemical and Biological Engineering, The Hong Kong University of Science and Technology, Hong Kong Special Administrative Region of China
b Department of Medicine & Therapeutics, The Chinese University of Hong Kong, Hong Kong Special Administrative Region of China
∗ Corresponding author.
15 5 2023
15 9 2023
15 5 2023
236 115402115402
2 3 2023
6 5 2023
14 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Rapid point-of-care diagnostics, essential in settings such as airport on-site testing and home-based screening, displayed important implications for infectious disease control during the SARS-CoV-2 outbreak. However, the deployment of simple and sensitive assays in real-life scenarios still faces the concern of aerosol contamination. Here, we report an amplicon-depleting CRISPR-based one-pot loop-mediated isothermal amplification (CoLAMP) assay for point-of-care diagnosis of SARS-CoV-2 RNA. In this work, AapCas12b sgRNA is designed to recognize the activator sequence sited in the loop region of the LAMP product, which is crucial for exponential amplification. By destroying the aerosol-prone amplifiable products at the end of each amplification reaction, our design can significantly reduce the amplicons contamination that causes false positive results in point-of-care diagnostics. For at-home self-testing, we designed a low-cost sample-to-result device for fluorescence-based visual interpretation. As well, a commercial portable electrochemical platform was deployed as a proof-of-concept of ready-to-use point-of-care diagnostic systems. The field deployable CoLAMP assay can detect as low as 0.5 copies/μL of SARS-CoV-2 RNA in clinical nasopharyngeal swab samples within 40 min without the need for specialists for its operation.
Graphical abstract
Image 1
Keywords
Contamination eliminated
LAMP
CRISPR-Based diagnostics
SARS-CoV-2
POCT
==== Body
pmc1 Introduction
During the SARS-CoV-2 pandemic, rapid point-of-care tests (POCTs) have shown important implications in airports and customs on-site testing and home-based screening (Budd et al., 2020) for disease control and self-care. There are POCTs with satisfactory performance available in the market (e.g., Lucira, Cue, Detect, Aptitude), but routine home-based tests without lab settings may still cause amplicons aerosol contamination and false positives to some extent (Mahmoud et al., 2021). Thus, further research for accurate and contamination eliminated POCTs is still needed. AapCas12b-based biosensors are good candidates for simple, rapid, and sensitive detection of viral RNA (Yin et al., 2021) in combination with isothermal nucleic acid amplification methods such as loop-mediated isothermal amplification (LAMP) (Joung et al., 2020). AapCas12b, a type of RNA-guided endonuclease that works at around 60 °C (Joung et al., 2020), can be designed to recognize and cleave specific target DNA sequences, known as ‘cis-cleavage’. Such event unleashes a non-specific endonuclease activity, called ‘trans-cleavage’, on single-stranded DNA (ssDNA) (Chen et al., 2018). The trans-cleavage activity on labelled ssDNA oligonucleotides can be used as a signal amplification strategy to transduce the recognition of a target DNA sequence at sub-attomolar concentrations (Aman et al., 2020; Teng et al., 2019). The high sensitivity and specificity of the AapCas12b protein arises from the complementarity between single guide RNA (sgRNA) and target DNA (activator) downstream to a protospacer adjacent motif (PAM) (Saifuddin et al., 2018).
In the above-mentioned CRISPR-based approaches, some reported POCT biosensors mitigate the amplicons contamination issue via one-pot assay requiring complicated manual operations (Joung et al., 2020). In other studies, to automate a fully-sealed microfluidic close system, solenoid valves, pumps, and manifold components that significantly increase the price of a device are utilized (Chandrasekaran et al., 2022). There is also a conventional approach to prevent amplicons contamination using the deoxyuridine triphosphate (dUTP) to partially substitute deoxythymidine triphosphate (dTTP) and destroy the deoxyuridine (dU) residue-containing amplicons via uracil DNA glycosylase (UDG) prior to the next amplification reaction (Hsieh et al., 2014). However, this method needs additional incubation steps at different temperatures for UDG cleavage and inactivation while incomplete UDG digestion may still allow the generation of amplicon aerosol. Thus far, a one-pot, simple, and high-performance contamination-eliminated or significantly reduced solution has yet to be developed for use at home or in low-resource areas in need of community surveillance testing.
Here, we report CoLAMP (CRISPR-based one-pot detection with loop-mediated isothermal amplification) to detect viral RNA in non-laboratory environments with good performance. CoLAMP leverages the rationally programmed competition kinetics between LAMP and AapCas12b reactions to transduce a visual or electrochemical signal and simultaneously deplete amplifiable DNA molecules to reduce the risk of aerosol contamination. Additionally, the design of the primer set and Cas12b sgRNA used in the CoLAMP test circumvents the undesired trans-cleavage seen in conventional CRISPR-based assays. By balancing the cleavage activity of Cas12b and the amplification efficiency, a limit of detection (LoD) of 0.25 cps/μL was achieved for the detection of SARS-CoV-2 RNA in mock samples. Moreover, CoLAMP showed a sensitivity of 100.0% and a specificity of 100.0% compared to Reverse Transcription q-PCR (RT-qPCR) using clinical samples.
2 Material and methods
The chemicals and instruments used in this work and the methods of the supplementary experiments are found in the supplementary information.
2.1 CoLAMP reaction buffer preparation
Tris-HCl (20 mM), (NH4)2SO4 (10 mM), Tween 20 (0.1%), dNTPs (1.4 mM), MgSO4 (8 mM), Bst 2.0 DNA polymerase (320 units/μL), Rtx Reverse Transcriptase (300 units/μL), Taurine (50 mM), Proteinase K inhibitor (1X), and SYBR green (1X) were mixed, then aliquoted as 30 μL ready-to-use CoLAMP reaction buffer in a tube for storage at −20 °C for up to 3 months.
2.2 AapCas12b cleavage reaction
The assembled AapCas12b-crRNA complex (30 nM) was mixed and incubated with 250 nM ssDNA reporters in TE buffer (pH 8.0) at room temperature for 15 min, then aliquoted as 2.2 μL ready-to-use CRISPR mix in a reaction tube for storage at −20 °C for up to 6 months. For sensitivity enhancement, 3.75 μL of 100% glycerol was added to the CRISPR mix. Before the reaction, CRISPR mix, CoLAMP reaction buffer, 10X primer pool and template were mixed to a final volume of 50 μL.
2.3 Signal generation
2.3.1 Electrochemical detection
Each AapCas12b cleavage reaction solution was pipetted onto the commercial carbon paste electrode surface or into the PDMS testing chip, and differential pulse voltammetry (DPV) measurements (PalmSens4) were obtained from −0.5 V to −0.1 V with the following conditions: potential step of 2.5 mV, scan rate of 25 mV/s, and sampling interval of 10 s. The standard error of the mean (SEM) of the triplicate peak measurements are shown as error bars.
2.3.2 Point-of-care device-based visual fluorescent CoLAMP assay
For device-based CoLAMP, nasopharyngeal swab samples are loaded in the extraction chamber of extraction-reaction-readout 3-in-1 chips where viral RNA is extracted and absorbed on the magnetic beads. The beads are then moved into the reaction chamber separated by mineral oil from the extraction chamber using a magnet for one-tube elution-reaction in 60 °C. An optical filter (525 nm) and a light-emitting diode (450 nm) were used for fluorescence readout with naked eyes or smartphone cameras. Smartphone photos were analysed with ImageJ.
3 Results and discussion
3.1 CoLAMP mechanism and signal readout
Conventional LAMP generates amplicons that are connected via loop structures, which are necessary for the rapid and exponential amplification. The rationale of our design is to leverage the loop structure to trigger a CRISPR-mediated signal amplification and simultaneously destroy the amplifiable units, impeding further amplification. The reaction mechanism of CoLAMP is realized by siting the cis-cleavage site into the loop structure of the LAMP inner primers (Fig. 1 A). Therefore, the CRISPR activator is the LAMP amplicons containing the loop-sequence. Thus, the amplicons cannot be amplified again once they are recognized and cleaved by CRISPR/AapCas12b. Then, AapCas12b collateral cleavage activity will induce the signal transduction by cleaving single-stranded fluorophore-quencher-labelled (ssFQ) or single-stranded methylene blue (MB)-labelled (ssMB) reporters, producing fluorescence or electrochemical signals, respectively. Notably, by siting the activator sequence of AapCas12b on the loop region of amplicons instead of the target sequences, the target strands would not be digested and wasted prior to the amplification as in other one-pot CRISPR/Cas-based assays (Chen et al., 2018; Wang et al., 2020).Fig. 1 (A) The mechanism of the CoLAMP assay (red color region: loop region of dumbbell structure). (B) Illustration of CoLAMP assay using an integrated fluorescence-based point-of-care device. (C) Illustration of CoLAMP assay using the electrochemical test chip. (ssFQ: single-stranded fluorophore-quencher-labelled reporter; ssMB: single-stranded methylene blue (MB)-labelled reporter; RFU: relative fluorescence unit; nA: nanoampere).
Fig. 1
Our amplicon-depleting detection can be done within 40 min from nasopharyngeal swab sampling to close-tube signal readout including a 30-min CoLAMP reaction. The visual fluorescence signal can be seen by the naked eye or captured via a smartphone camera. We also established an immobilization-free electrochemical test chip, which is a feasible alternative for point-of-care applications given the compatibility with commercial electrochemical platforms (Fig. 1B and C).
In this work, one of the challenges is to balance the competition between the polymerase used in LAMP and AapCas12b for the same DNA template. In the following sections, we will further introduce how we balanced this competition through sequence designs and reactions optimization.
3.2 Compatibility of the amplicon-depleting CRISPR-based one-pot detection with linear LAMP primer pool
AapCas12b retains highly specific cis-cleavage activity after LAMP amplification. In our design, the LAMP linear primer pool (P0 in SI.11, 13), alike the primer pool in conventional LAMP (Joung et al., 2020), contains a pair of forward and backward inner primers (FIP and BIP, respectively), forward and backward outer primers (FOP and BOP, respectively) and a sequence before the PAM code in this segment (red in Fig. 2 A), which is crucial for exponential amplification based on the dumbbell structure created in the initial stage. An alternative way to destroy the dumbbell structure is by recognizing the stem region of the linear LAMP primer. However, this alternative could easily result in false-positives from stem sequences in the 5’ end of FIP or BIP (in dark blue colour) (Joung et al., 2020). Thus, in a CoLAMP assay, amplicons are selectively destroyed during the CRISPR activation process by cis-cleavage before their accumulation, which can reach 109-fold of the initial target load in a conventional LAMP reaction.Fig. 2 Compatibility of the amplicon-depleting mechanism with linear LAMP primer pool using conventional linear primer design. (A) Schematic of CoLAMP utilizing a linear primer pool (CoLAMP-P0). Linear primer pool (P0): FIP is forward inner primer, BIP is backward inner primer, FOP is forward outer primer, BOP is backward outer primer. (B) AapCas12b collateral cleavage can be activated by 10 cps/μL mock sample without LAMP amplification. (C) AapCas12b sgRNA optimization. CoLAMP-P0-1 utilizes sgRNA-1 same as the target sequence while CoLAMP-P0-2 utilizes sgRNA-2 complementary to the target sequence. (D) Real-time SYBR green signal of CoLAMP-P0 reaction in comparison with LAMP. (E) CoLAMP-P0 reaction product melting curve characterization compared with conventional LAMP. RFU: relative fluorescence unit; NTC: no template control.
Fig. 2
During the design process, we used 10 cps/μL of SARS-CoV-2 ssRNA in mock samples (Supporting Information 1). First, we assessed if the ssRNA target could activate the AapCas12b without amplification. We observed that the ssRNA target can slowly activate AapCas12b (Fig. 2B), which may lead to target waste at the initial stage of the amplification. Therefore, we designed the sgRNA recognition site to be the same as the target sequence (CoLAMP-P0-1 in Fig. 2C) rather than complementary to it (CoLAMP-P0-2 in Fig. 2C). CoLAMP-P0-1 avoids target waste and solves the issue of competition between the AapCas12b cleavage and the amplification. Thus, higher fluorescence signal was generated after activation (Fig. 2C). Another way to address the target waste issue is to design the activator sequence to include the spacer region of the intermediate “dumbbell” strand (region between the loop structures) complementary to the viral RNA. But such design leads to an activator sequence including 5’ ends of inner primers, causing undesired activation of trans-cleavage (Joung et al., 2020).
For amplicon-depleted detection, it is important to control the length and amount of the LAMP product via CRISPR/Cas activity. Interestingly, CoLAMP-P0-1 can not only avoid target waste but also deplete LAMP amplicons more completely (Fig. 2D, E, and S1, S2). As shown in Fig. S3, our design (green curve) yields fewer dsDNA product than in conventional LAMP (pink curve) with the same primer pool (P0) even when using samples with a high concentration of the target, indicating that the accumulation of amplicons in CoLAMP is decoupled by siting the recognition sequence in the loop region of the dumbbell structure. Through rational design, CRISPR/AapCas12b remains inactive prior to the amplification. Besides LAMP, amplicon-depleted point-of-care diagnostics can be initiated by other isothermal amplification methods such as recombinase polymerase amplification (RPA) by designing the recognition sequence partially on the forward or backward primer and siting a PAM code downstream of the target for CRISPR/Cas recognition (Fig. S4).
3.3 sited in the stem of a hairpin loop LAMP primer facilitates LAMP initiation
To overcome the sequence constraint of having a PAM site on the SARS-CoV-2 target, we used a PAM-containing loop primer (Fig. 3 A). The loop primer and the corresponding sgRNA for Cas12 cleavage were designed in silico (Supporting Information 3). Similar to the linear LAMP primer pool, there are a pair of inner primers, termed forward inner loop and backward inner loop primers (FILP and BILP, respectively), and outer primers (FOP and BOP). The PAM code is located within the stem region of the FILP (Lu et al., 2022), while the sgRNA recognizes the sequence in the loop region of FILP (illustrated in red in Fig. 3A), which is crucial for the dumbbell structure-based exponential amplification.Fig. 3 Compatibility of the amplicon-depleting mechanism with LAMP using a loop-primer design. (A) PAM code sited in the loop-containing LAMP primers. FILP is forward inner loop primer, BILP is backward inner loop primer, FOP is forward outer primer, BOP is backward outer primer. (B) LAMP with loop-primer product characterization by Tm value. (C) CoLAMP primer optimization by AapCas12b collateral cleavage signal after activation by dsDNA LAMP product loop sequence region (recognition of the sequence complementary to the red part in FILP). (D) CoLAMP product characterization by Tm value (100 min incubation to accumulate enough CoLAMP product for the amplicons characterization). (E) Optimization of the primer pool concentration for CRISPR activation (P7* refers to half of concentration of the LAMP Primer Pool 7). (F) Positive control of endpoint Confirm-Cas verification. (G) CoLAMP product sequence verification by endpoint Confirm-Cas enzyme.
Fig. 3
In the design of the sgRNA for looped primers, the sgRNA sequence is the same as that of the loop of FILP to avoid false-positive signals that may arise from CRISPR/Cas activation due to sgRNA targeting unreacted loop primers (Fig. S5). Additionally, we examined the impact of the location of the PAM site on the FILP. We designed three sets of FILPs with the PAM sited on the loop, partially on the loop and on the stem, termed CoLAMP-P5, P6, and P7, respectively (SI.11, 13, Fig. S6). We found that the PAM located in the stem region improved the LAMP amplification (Fig. 3C). This may be due to the unwinding function of CRISPR/Cas towards PAM in the dsDNA target (Anders et al., 2014; Jinek et al., 2014), which facilitates the unwinding of the looped primer and thus the Bst polymerase extension. CoLAMP-P7 also yielded good depletion of amplicons assessed by melting curve analysis (Fig. 3D) and gel electrophoresis (Fig. S6).
Interestingly, higher CRISPR signal was achieved by decreasing the primer pool concentration (Fig. 3E) without significant change in the cleaved amplicons products (Fig. S7). This may result from less AapCas12b occupied in the unwinding of PAM sites in looped primers that are not adjacent to a targeting sequence (Qin et al., 2022). Such PAM-containing strands without target sequences are the unreacted FILP and LAMP non-specific dsDNA products. To check whether products of CoLAMP using looped primers are from specific amplification of target, a Confirm-CRISPR/Cas reaction was designed to recognize the cleaved target-specific amplicons and report it through Cas12 trans-cleavage of fluorophore-quencher probes. In this way, we confirmed that all the cleaved products in a positive CoLAMP detection are target-specific products from the LAMP amplification (Fig. 3F and G). These results demonstrated the specificity of loop primer CoLAMP. Additionally, the loop and target sequences in CoLAMP products were quantified with qPCR when using high-dose samples (Fig. S8).
The PAM-containing loop-primer is also compatible with RPA by introducing an extra sequence (red in Fig. S9) upstream of the forward primer and backward primer for CRISPR/Cas recognition, which are crucial for exponential amplification (Fig. S9).
3.4 CoLAMP analytical performance and sensitivity enhancement
After the design and optimization work, we characterized performance of CoLAMP achieving nearly eliminated contamination using both linear primer and loop primer pools (CoLMAP-P0, LAMP Primer Pool 0 and CoLAMP-P7, LAMP Primer Pool 7 in Supplementary Material SI.11, 13, respectively). Both CoLAMP-P0 and CoLAMP-P7 generated digested products confirmed by gel electrophoresis when using high-dose mock samples (106 cps/mL) representative of at-home self-collection (Mane et al., 2022) (Fig. 4A and D, S10). In carryover tests, positive CoLAMP products as input samples did not induce fluorescent signals in a new CoLAMP reaction, though carryover amplification products that did not induce Cas12 activation were detected by fragment analysis. These amplification products became negligible after 103- and 102-dilutions of the positive products for CoLAMP-P0 and P7 respectively (Fig. 4B, E, S11). The one-pot reaction was able to detect the RNA target in mock samples with a concentration as low as 0.25 copies/μL (Fig. 4 C and F).Fig. 4 Amplicons-depleting AapCas12b-based LAMP reaction performance assessment and sensitivity enhancement. (A). CoLAMP-P0 product characterization by gel electrophoresis. (B). CoLAMP-P0 carryover test by adding positive test product (103 copies/μL input) in new reaction. (C). CoLAMP-P0 performance at the limit of detection (0.25 cps/μL). (D). CoLAMP-P7* product length characterization by gel electrophoresis. (E). CoLAMP-P7* carryover test by adding positive test product (103 copies/μL input) in new reaction. (F). CoLAMP-P7* performance at the limit of detection (0.25 cps/μL). (G). Effect of the glycerol addition on the assay sensitivity. (H). Clinical tests referring to qPCR. (*P < 0.05, **P < 0.01).
Fig. 4
Thus far, the CoLAMP assay with the linear primer design (P0) enables contamination significantly reduced and sensitive performance detection by straightforward application of reported LAMP linear primer sets. Nevertheless, primer selection is needed to maintain high amplification efficiency under the limitation of PAM containing loop region. Conversely, looped primers (P7) enable more design flexibility by introducing the PAM containing hairpin structure. However, the amplification efficiency of P7 may be limited by the unwinding issue as mentioned in 3.3.
The ultimate goal of the proposed CoLAMP is to realize a proof-of-concept amplicon-depleting diagnosis of SARS-CoV-2 outside laboratory settings. This requires the higher signal intensity to enable a simpler and clearer signal readout. Therefore, signal generation enhancement was further achieved with P0 by adding 3.75 μL glycerol, which showed no interference on the amplicon-depleting performance (Fig. 4G, S12, S13). Since glycerol exerted no positive impact on either the LAMP or CRISPR reaction system, the notable increment in detection efficiency of the one-pot system may be due to its physical property to create a temporary two-phase interface, allowing more reaction time for the LAMP reaction (Fig. 4G), which can be stored at 4 °C for 24 h (Fig. S14). As a proof-of-concept, clinical sample testing showed that CoLAMP-P0 had a sensitivity of 100.0% and a specificity of 100.0% compared to qPCR (Table S1, Fig. 4H, S15, S16).
3.5 Extraction-reaction-readout one-pot device design for POCT
Resulting from good signal-on performance in clinical sample assay (Figs. S15 and S16), CoLAMP can be performed in low-cost integrated devices. We developed a fluorescence-based CoLAMP device that enables a 40-min sample-to-result assay. Results can be observed by the naked-eyes or captured by a smartphone camera. Importantly, the operation of the integrated device is highly automated. First, the user needs to power on the device and load the raw sample, after 10-min extraction, they only need to push a magnet bar for RNA injection to start the CoLAMP assay (see the CoLAMP assay demonstration video). In the device, a dark enclosure is designed to block out external light and make sure the background control is clear (Fig. 5 A). The combination of a light-emitting diode (LED 450 nm) and a narrow wavelength filter (525 nm) allows the distinction between positive and negative controls (Fig. 5B).Fig. 5 Device design for the amplicon-depleting diagnosis of SARS-CoV-2 without lab settings. (A). The fluorescence-based readout dark enclosure. (B). Visualization of the reaction chamber under LED and optical filter. (C). The extraction-reaction-readout 3 in 1 disposable chamber with microfluidic channel design. First, RNA is captured by magnetic beads after virus lysis, then transferred with the magnet to the reaction chamber isolated by mineral oil. RNA is eluted by the reaction buffer eventually. (D). The fluorescence intensity standard curve using pure FAM fluorophore in the dark device for visual readout. (E). Mock sample (0.5 and 10 cps/μL) real-time readout characterized by baseline subtracted RGB value. (F). The immobilization-free electrochemical-based detection using methylene blue labelled DNA reporter fragment length identification. (G). Electrochemical signal of varying lengths of single- and double-stranded DNA reporters using CoLAMP-P0 in 0.5 cps/μL mock sample. (H). 1 mm thin test chip enables better signal-on differentiation using CoLAMP-P0 in 0.5 cps/μL mock sample. (***P < 0.001).
Fig. 5
As an automatic platform ready for home use, an extraction-reaction-readout 3 in 1 disposable chamber is used for each test (Fig. 5C, S17). Inside the chamber, magnetic beads are utilized to transfer the ssRNA extracted from the virus extraction buffer to the reaction solution with high efficiency (Lee et al., 2020) (Figs. S18, S19, S20). As designed, the streamlined CoLAMP assay is realized by combining the sample elution with the reaction at 60 °C. The RNA on the magnetic beads is eluted by the reaction buffer. Importantly, the magnetic beads at the bottom of the reaction chamber will not inhibit the fluorescence signal (Fig. S21) because the carboxylate magnetic bead SiO2 surface avoids the excitation wavelength overlap of FAM and the Fe3O4 metal core (Magnan et al., 2013).
To maximize the fluorescence intensity, the LED was positioned at a high incidence angle of 30° from the side of the reaction chamber (Fig. S22). Image acquisition is done by a smartphone camera and the analysis is then performed by image processing software, such as ImageJ, by measuring the green (G) spectrum from the RGB model (Fig. 5D). With the abovementioned protocol a sensitivity of 0.5 cps/μL was achieved, which is comparable to other CRISPR-based POCT platforms (Fig. S22, SI10 Table S3), as well as a reproducible real-time visual readout (Fig. 5E, S23). Notably, the integrated reusable prototype costs less than 15 USD (Table S2).
Additionally, CoLAMP is compatible with different signal transduction mechanisms besides the home-based visual assay, so it can be deployed in an electrochemistry-based testing chip with a commercial platform to benefit further developments of ready-to-use portable biosensors. As such, when we used ssDNA methylene blue (MB) labelled reporters, our detection platform can be a disposable screen-printed carbon electrode test chip (Fig. 5F).
Based on the immobilization-free electrochemical fundamental principle, the large difference in the diffusion coefficient of ssDNA of different lengths will result in large current signal-on changes during differential pulse voltammetry (DPV) scan measurements (Lee Yu et al., 2021). The length difference from double-stranded or single-stranded reporters to cleaved fragments of less than 10 nucleotides is transduced to a signal-on current peak at the methylene blue (MB) potential to distinguish positive from negative samples (Fig. 5G, S24). Moreover, using a disposable testing chip with polydimethylsiloxane (PDMS) achieves around two-fold signal-on performance compared to a bare electrode (Fig. 5H, S25).
4 Conclusion
In this work, we report the development and application of CoLAMP, a CRISPR-based one-pot loop-mediated isothermal amplification assay for on-site diagnostics and home-based screening of infectious pathogens. CoLAMP leverages the cis- and trans-cleavage mechanisms of AapCas12b and rational sequence design to deplete the amplicons after the LAMP reaction and thus reduce the risk of aerosol contamination common in point-of-care scenarios. Specifically, the AapCas12b sgRNA is designed to recognize the loop structure of the LAMP product, which is cleaved and thus cannot be used in further amplification steps. We demonstrated the performance of CoLAMP in the detection of SARS-CoV-2 RNA in mock samples, achieving a limit of detection of 0.5 cps/μL. CoLAMP can be implemented with reported conventional linear LAMP primer sets for straightforward and sensitive detection with insignificant contamination. As well, we introduced a rationally designed loop primer set to circumvent the requirement of a PAM code in the target sequence, which is needed when using linear primers. Future work of the loop primer-based assay can focus on the unwinding problem in the stem region to enhance the sensitivity by changing the structural design of the hairpin primers and siting the PAM region in the stem region. Compared to other methods that rely on specialized hardware design, e.g., fully sealed microfluidic devices, the amplicon-depleting feature of CoLAMP makes it compatible with relatively simple and low-cost devices for on-site deployment. We demonstrated the suitability of CoLAMP to detect SARS-CoV-2RNA with a visual readout in a semi-automated sample-in, the result-out manner in 40 min using a fluorescence-based device. In addition, an electrochemical readout can be achieved by an immobilization-free assay on a disposable testing chip. Further development of the home-based assay can be focused on the storage issue by incorporating the lyophilized enzyme and probes into a nitrocellulose membrane-based formats by lateral flow strip or wearable tapes.
CRediT authorship contribution statement
Yumeng Cao: Methodology, Investigation, Formal analysis, Writing – original draft. Xiao Lu: Investigation, Formal analysis, Writing – review & editing. Haosi Lin: Data curation, Formal analysis, Writing – review & editing. Alan Fernando Rodriguez Serrano: Investigation, Writing – original draft. Grace C.Y. Lui: Collection, Formal analysis. I-Ming Hsing: Conceptualization, Methodology, Supervision, Project administration, Funding acquisition, Writing – review & editing.
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.
Appendix A Supplementary data
The following are the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Multimedia component 2
Multimedia component 2
Data availability
Data will be made available on request.
Acknowledgement
The work was financially supported by the 10.13039/501100003452 Innovation and Technology Commission (MRP/077/20) and Research Grants Council (CRF 6107-20G, 16306519) of the Hong Kong SAR Government of China. Clinical samples were provided by The Prince of Wales Hospital approved by the Chinese University of Hong Kong Clinical Research Ethics Committee (CREC-2021-0188).
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bios.2023.115402.
==== Refs
References
Aman R. Mahas A. Mahfouz M. Nucleic acid detection using CRISPR/Cas biosensing technologies ACS Synth. Biol. 9 6 2020 1226 1233 32159950
Anders C. Niewoehner O. Duerst A. Jinek M. Structural basis of PAM-dependent target DNA recognition by the Cas9 endonuclease Nature (Lond.) 2014 2014 569 573 5137519 513
Budd J. Miller B.S. Manning E.M. Lampos V. Zhuang M. Edelstein M. Rees G. Emery V.C. Stevens M.M. Keegan N. Short M.J. Pillay D. Manley E. Cox I.J. Heymann D. Johnson A.M. McKendry R.A. Digital technologies in the public-health response to COVID-19 Nat. Med. 2020 268 26 2020 1183 1192
Chandrasekaran S.S. Agrawal S. Fanton A. Jangid A.R. Charrez B. Escajeda A.M. Son S. Mcintosh R. Tran H. Bhuiya A. Díaz León Derby M. Switz N.A. Armstrong M. Harris A.R. Prywes N. Lukarska M. Biering S.B. J Smock D.C. Mok A. Knott G.J. Dang Q. Dis E. Dugan E. Kim S. Liu T.Y. Hamilton J.R. Lin-Shiao E. Stahl E.C. Tsuchida C.A. Giannikopoulos P. McElroy M. McDevitt S. Zur A. Sylvain I. Ciling A. Zhu M. Williams C. Baldwin A. Moehle E.A. Kogut K. Eskenazi B. Harris E. Stanley S.A. Lareau L.F. Tan M.X. Fletcher D.A. Doudna J.A. Savage D.F. Hsu P.D. Rapid detection of SARS-CoV-2 RNA in saliva via Cas13 Nat. Biomed. Eng. 68 6 2022 944 956 2022
Chen J.S. Ma E. Harrington L.B. Da Costa M. Tian X. Palefsky J.M. Doudna J.A. CRISPR-Cas12a target binding unleashes indiscriminate single-stranded DNase activity Science 360 2018 436 439 29449511
Hsieh K. Mage P.L. Csordas A.T. Eisenstein M. Soh H.T. Simultaneous elimination of carryover contamination and detection of DNA with uracil-DNA-glycosylase-supplemented loop-mediated isothermal amplification (UDG-LAMP) Chem. Commun. 50 2014 3747 3749
Jinek M. Jiang F. Taylor D.W. Sternberg S.H. Kaya E. Ma E. Anders C. Hauer M. Zhou K. Lin S. Kaplan M. Iavarone A.T. Charpentier E. Nogales E. Doudna J.A. Structures of Cas9 endonucleases reveal RNA-mediated conformational activation Science 80 2014 343
Joung J. Ladha A. Saito M. Kim N.-G. Woolley A.E. Segel M. Barretto R.P.J. Ranu A. Macrae R.K. Faure G. Ioannidi E.I. Krajeski R.N. Bruneau R. Huang M.-L.W. Yu X.G. Li J.Z. Walker B.D. Hung D.T. Greninger A.L. Jerome K.R. Gootenberg J.S. Abudayyeh O.O. Zhang F. Detection of SARS-CoV-2 with SHERLOCK one-pot testing N. Engl. J. Med. NEJMc2026172 2020
Lee H. Oh M.T. Lee Y.J. Son K.H. Choi N. Lee B.C. Lee S.H. Highly-efficient microfluidic ultrasonic transducers assisted gDNA extraction system in whole blood for POCT applications Sensor. Actuator. B Chem. 319 2020 128317
Lee Yu H. Cao Y. Lu X. Hsing I.M. Detection of rare variant alleles using the AsCas12a double-stranded DNA trans-cleavage activity Biosens. Bioelectron. 189 2021 113382
Lu X. Lee Yu H. Lin H. Cao Y. Hsing I.M. Rapid and highly specific detection of communicable pathogens using one-pot loop probe-mediated isothermal amplification (oLAMP) Sensor. Actuator. B Chem. 357 2022 131385
Magnan F. Gagnon J. Fontaine F.G. Boudreau D. Indium@silica core–shell nanoparticles as plasmonic enhancers of molecular luminescence in the UV region Chem. Commun. 49 2013 9299 9301
Mahmoud S.A. Ganesan S. Ibrahim E. Thakre B. Teddy J.G. Raheja P. Zaher W.A. Evaluation of six different rapid methods for nucleic acid detection of SARS-COV-2 virus J. Med. Virol. 93 2021 5538 5543 34002401
Mane A. Jain S. Jain A. Pereira M. Sirsat A. Pathak G. Bhoi V. Bhavsar S. Panda S. Diagnostic performance of oral swab specimen for SARS-CoV-2 detection with rapid point-of-care lateral flow antigen test Sci. Rep. 121 12 2022 1 7 2022
Qin P. Chen P. Deng N. Tan L. Yin B.-C. Ye B.-C. Switching the Activity of CRISPR/Cas12a Using an Allosteric Inhibitory Aptamer for Biosensing 2022
Saifuddin F.A. Rybarski J.R. Russell R. Strohkendl I. Finkelstein I.J. Kinetic basis for DNA target specificity of CRISPR-cas12a Mol. Cell. 71 2018 816 824.e3 30078724
Teng F. Guo L. Cui T. Wang X.G. Xu K. Gao Q. Zhou Q. Li W. CDetection: CRISPR-Cas12b-based DNA detection with sub-attomolar sensitivity and single-base specificity Genome Biol. 20 2019 1 7 30606230
Wang Y. Ke Y. Liu W. Sun Y. Ding X. A one-pot toolbox based on cas12a/crRNA enables rapid foodborne pathogen detection at attomolar level ACS Sens. 5 2020 1427 1435 32337966
Yin L. Man S. Ye S. Liu G. Ma L. CRISPR-Cas based virus detection: recent advances and perspectives Biosens. Bioelectron. 193 2021 113541
|
PMC010xxxxxx/PMC10185489.txt |
==== Front
J Thromb Haemost
J Thromb Haemost
Journal of Thrombosis and Haemostasis
1538-7933
1538-7836
International Society on Thrombosis and Haemostasis. Published by Elsevier Inc.
S1538-7836(23)00415-4
10.1016/j.jtha.2023.04.044
Original Article
Recombinant human DNase-I improves acute respiratory distress syndrome via neutrophil extracellular trap degradation
Jarrahi Abbas 1
Khodadadi Hesam 2
Moore Nicholas S. 1
Lu Yujiao 1
Awad Mohamed E. 2
Salles Evila L. 2
Vaibhav Kumar 1
Baban Babak 23
Dhandapani Krishnan M. 1∗
1 Department of Neurosurgery, Medical College of Georgia, Augusta University, Augusta, Georgia, USA
2 Department of Oral Biology and Diagnostic Sciences, Dental College of Georgia, Augusta University, Augusta, Georgia, USA
3 Department of Surgery, Medical College of Georgia, Augusta University, Augusta, Georgia, USA
∗ Correspondence Krishnan M. Dhandapani, Department of Neurosurgery, Medical College of Georgia, Augusta University, 1120 15th Street, Augusta, GA 30912, USA.
16 5 2023
16 5 2023
20 1 2023
21 4 2023
28 4 2023
© 2023 International Society on Thrombosis and Haemostasis. Published by Elsevier Inc. All rights reserved.
2023
International Society on Thrombosis and Haemostasis
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Respiratory failure is the primary cause of death in patients with COVID-19, whereas coagulopathy is associated with excessive inflammation and multiorgan failure. Neutrophil extracellular traps (NETs) may exacerbate inflammation and provide a scaffold for thrombus formation.
Objectives
The goal of this study was to determine whether degradation of NETs by recombinant human DNase-I (rhDNase), a safe, Food and Drug Administration–approved drug, reduces excessive inflammation, reverses aberrant coagulation, and improves pulmonary perfusion after experimental acute respiratory distress syndrome (ARDS).
Methods
Intranasal poly(I:C), a synthetic double-stranded RNA, was administered to adult mice for 3 consecutive days to simulate a viral infection, and these subjects were randomized to treatment arms, which received either an intravenous placebo or rhDNase. The effects of rhDNase on immune activation, platelet aggregation, and coagulation were assessed in mice and donor human blood.
Results
NETs were observed in bronchoalveolar lavage fluid and within regions of hypoxic lung tissue after experimental ARDS. The administration of rhDNase mitigated peribronchiolar, perivascular, and interstitial inflammation induced by poly(I:C). In parallel, rhDNase degraded NETs, attenuated platelet-NET aggregates, reduced platelet activation, and normalized the clotting time to improve regional perfusion, as observed using gross morphology, histology, and microcomputed tomographic imaging in mice. Similarly, rhDNase reduced NETs and attenuated platelet activation in human blood.
Conclusion
NETs exacerbate inflammation and promote aberrant coagulation by providing a scaffold for aggregated platelets after experimental ARDS. Intravenous administration of rhDNase degrades NETs and attenuates coagulopathy in ARDS, providing a promising translational approach to improve pulmonary structure and function after ARDS.
Keywords
coagulation
immunothrombosis
inflammation
platelets
thrombosis
==== Body
pmc1 Introduction
Acute respiratory distress syndrome (ARDS), a common sequala following pneumonia, nonpulmonary sepsis, aspiration of gastric contents, or severe trauma [1], is characterized by shortness of breath, tachypnea, hypoxemia, respiratory failure, and increased mortality [2,3]. Among the numerous causes, viral pneumonia frequently induces ARDS. Respiratory viruses, such as influenza virus and coronaviruses, are associated with an increased prevalence of ARDS [4], as evidenced by the recent COVID-19 pandemic. Unfortunately, the clinical management of ARDS is complicated by a lack of efficacious therapeutic strategies.
Respiratory failure is the primary cause of death in patients with COVID-19, whereas aberrant coagulopathy associated with excessive inflammation frequently contributes to multiorgan failure and patient deterioration. Patients with SARS-CoV-2 infection exhibit prolonged prothrombin time, increased fibrin degradation products, and disseminated intravascular coagulopathy, which is noted in a majority of COVID-19–associated deaths [5]. Coagulopathy is associated with tissue hypoxia, venous thromboembolism, acute coronary syndrome, and cerebral infarction [6,7], and anticoagulant therapy with low-molecular–weight heparin appeared to be associated with lower mortality in a subpopulation that met sepsis-induced coagulopathy criteria or that with markedly elevated D-dimer levels in a study of 449 patients with severe COVID-19 [5,8]. Coagulation and inflammation are tightly linked processes that exhibit reciprocal cross-talk via a process termed as thromboinflammation [9]. Systemic inflammation increases tissue factor–mediated thrombin production and limits endogenous fibrinolysis to enhance coagulation, whereas increased coagulation perpetuates inflammatory activation. Thus, elucidation of the interplay between the immune system and coagulation may identify efficacious therapeutics to proactively address the complications of SARS-CoV-2 infection.
Mobilization of circulating neutrophils establishes a proinflammatory milieu after hemorrhage- or endotoxemia-induced acute lung injury, providing the first line of host defense against pathogens [10]. Circulating neutrophils may also initiate endothelial injury and fluid leakage into the alveoli, contributing to decreased lung compliance and hypoxemia in patients with ARDS. Elevated neutrophil-to-lymphocyte ratio is an independent risk factor for disease severity and mortality in hospitalized patients with COVID-19, whereas extensive neutrophil infiltration is associated with pulmonary fibrin deposition and vascular lesions [[11], [12], [13]]. Neutrophil depletion prevents endotoxin-induced lung vascular permeability in sheep [14]; however, given the critical role of neutrophils in host defense, global depletion strategies are not a clinically feasible approach to improve outcomes after ARDS. Thus, alternative, targeted approaches are needed to minimize the detrimental effect of neutrophil activation after acute lung injuries.
In addition to engulfing pathogens into phagosomes, activated neutrophils extrude a meshwork of chromatin fibers to form cloud-like neutrophil extracellular traps (NETs). While primarily implicated in extracellular pathogen trapping and host defense, NETs increase endothelial permeability and exacerbate inflammatory activation to potentiate tissue injury during viral pneumonitis and ARDS [15,16] and provide a scaffold for thrombus formation [17,18]. The extent of neutrophil priming and NET formation is correlated with disease severity and mortality in patients with ARDS [19]. Moreover, neutrophil hyperactivation and increased NET formation is correlated with an elevated risk of venous thromboembolism in hospitalized patients with COVID-19 compared with that in healthy controls [19]. Thus, NETs may provide a therapeutic target to proactively reduce tissue injury associated with coagulopathy following SARS-CoV-2 infection.
In the present study, we identified NETs as scaffolds for aggregated platelets, which promoted thrombus formation in an experimental ARDS model. We further showed the therapeutic potential of recombinant human DNase-I (rhDNase; Pulmozyme [dornase-α]), a safe, Food and Drug Administration–approved drug under investigation for the management of COVID-19–induced ARDS [20], to degrade NETs and improve pulmonary structure and function.
2 Methods
2.1 Experimental design
A double-blinded, randomized study design was utilized. Mice were the research subjects in controlled laboratory experiments. Donor human blood was also studied ex vivo. The primary study endpoints were quantification of NETs and platelet activation. Parallel studies were used to measure the pulmonary vascular structure. Mice were randomized to study arms using a random number generator. Power analyses were performed a priori to determine sample sizes using an α value of 0.05 and β value of 0.10. Investigators (A.J., N.S.M., Y.L., M.E.A., H.K., E.L.S.) performed data acquisition of predetermined outcome measures. Following final data acquisition, the mice were decoded and final analyses were performed. No subjects were removed from the final analyses. The Institutional Animal Care and Use Committee at Augusta University approved all animal studies, in compliance with the National Institutes of Health guidelines.
2.2 ARDS model
Adult (10-12 weeks old), mixed-sex C57BL/6J mice (Jackson Laboratories, Stock #000664) were housed in an Association for Assessment and Accreditation of Laboratory Animal Care International-accredited, pathogen-free vivarium. The mice were anesthetized using 3% isoflurane and maintained with 1.5% to 2% isoflurane throughout all procedures. Sham controls received 50 μL of intranasal sterile phosphate-buffered saline (PBS) for 3 consecutive days. Experimental groups received intranasal poly(I:C) (200 μg in 50 μL of sterile PBS; Sigma Aldrich), a synthetic double-stranded RNA that mimics viral infection, for 3 consecutive days, per our group [21,22]. Experimental mice were treated with either an intravenous (i.v.) injection of placebo (PBS) or 5-mg/kg rhDNase via the tail vein on days 2 and 3 after the first poly(I:C) administration. The body temperature was maintained at 37 °C using a small animal temperature controller throughout all procedures. Food and water were provided ad libitum.
2.3 Tissue collection
On experimental day 5, a subset of randomly selected mice from each group was sacrificed, blood was collected via cardiac puncture, and bronchoalveolar lavage fluid (BALF) was collected using standard protocols [23]. Thereafter, the lungs were harvested for analytic flow cytometry. On experimental day 6, the remaining mice from each group were sacrificed for gross and histologic examinations or randomly assigned to undergo microcomputed tomography (microCT) imaging.
2.4 Analytic flow cytometry
Single-cell suspensions were sieved through a 100-μm cell strainer, centrifuged (252× g, 10 minutes), and stained with fluorescent antibodies to quantify neutrophils, macrophages, lymphocytes, and cytokine expression. Cells were stained with anti-Ly6G (Cat#127625; BioLegend), anti-F4/80 (Cat#123116; BioLegend), anti-CD3 (Cat#100214; BioLegend), anti-CD4 (Cat#100432; BioLegend), anti-CD8 (Cat#140416; BioLegend), anti-CD19 (Cat#551001; BD Bioscience), anti-CD45 (Cat#103138; BioLegend), anti-NE (Cat#bs-6982R; Bioss), or anti-MPO (Cat#PA5-16672; Invitrogen). Cells were then fixed, permeabilized, and stained for the intracellular cytokines IL-6 (Cat#504602; BioLegend), TNF-alpha (Cat#506344; BioLegend), IL-2 (Cat#517605; BioLegend), IL-17 (Cat#506907; BioLegend), IL-10 (Cat#505006; BioLegend), and IFN-gamma (Cat#505850; BioLegend) per our laboratory protocol [[24], [25], [26]]. Platelets were stained based on the functional mode of resting or activation. Resting platelets, which retain their discoid shape, were identified using anti-CD41 (mouse: Cat#133932; BioLegend) or anti-CD42b (human: Cat#303903; BioLegend). Activated platelets, which exhibit a more amorphous form, were stained with anti-CD62P (mouse: Cat#148305; BioLegend and human: Cat#304942; BioLegend). Macroaggregates of platelet-NETs were identified using 2 analytical live gates that were set during the acquisition process. While the first gate was based on the morphologic characteristics of platelets (forward scatter/side scatter), the second gate was based on the expression of CD62P antigen on platelets already expressing CD41. These platelets were further analyzed based on the association and attachment level of NET markers (eg, extracellular MPO/Cit-H3/chromatin), providing a quantitative assessment of platelet-NET macroaggregates. Human blood was obtained from mixed-sex, healthy donors and randomized to undergo a 24-hour incubation with a vehicle, 200 μg of poly(I:C), or 200 μg of poly(I:C) + 200 μg of rhDNase. Platelets were mechanophenotyped as either resting or activated, as above. Samples were gated based on forward and side scatter properties as well as marker combinations using the 4-Laser LSR II flow cytometer. All acquired flow cytometry data were analyzed using the FlowJo software (version 10; Becton Dickinson).
2.5 Histology
Mice were transcardially perfused with chilled PBS, followed by 10% neutral buffered formalin. The lungs were removed and 5-μm midcoronal sections were stained with hematoxylin and eosin to visualize tissue structure. Random views were analyzed by an investigators (A.J., Y.L.) using bright-field microscopy. For quantification of hypoxia, hypoxyprobe-1 (Cat #HP1) was administered via the tail vein 1.5 hours prior to sacrifice per our laboratory protocol [24]. Coronal sections were stained with mouse anti-hypoxyprobe-1 antibody overnight, followed by incubation with secondary Alexa Flour-488–tagged IgG (Cat#A21202; Invitrogen). In parallel, neutrophils and NETs were visualized using anti-mouse Ly6G antibody (Cat#127608; BioLegend) and anti-histone H3 (citrulline R26) (Cit-H3) antibody (Cat#ab19847; Abcam) per our group [24]. Optical images were captured by an investigators (A.J., Y.L.) using the Zeiss AxioImager2 microscope. Fluorescence intensity was quantified using color intensity measurement using the Zeiss ZEN blue software (version 3.2) as previously detailed [25].
2.6 MicroCT analysis
Anesthetized mice were transcardially perfused with PBS, formalin, and BriteVu contrast agent, which has a high radiodensity and penetrates to the capillary level. The BriteVu solution was freshly prepared by dissolving 25 g of BriteVu powder in 200 mL of distilled water plus 500 μL of BriteVu Enhancer. Following perfusion, the mice were placed on ice for 2 hours, and then, the lungs were harvested, fixed in 4% paraformaldehyde for 72 hours, and washed with 70% ethanol. The lungs were scanned by an investigators (M.E.A., A.J.) using the Bruker Skyscan 1272 microCT system with scanning parameters of 2k (2452 × 1640) resolution, 9.3-μm image pixel size, 0.5-mm aluminum filter, and 0.4 rotation step. Three-dimensional reconstruction was created using the NRecon software (Bruker Corporation), followed by analysis using the CTan software, which uses high-order rendering algorithms to create volume-rendering and maximum-intensity projection images of complex vascular networks within the lungs. Identical parameters were used for all samples during scanning, reconstruction, and analysis.
2.7 Clotting time
On day 5, 100 μL of whole blood was placed on a glass slide. Blood was gently pricked with a needle tip every 30 seconds until a fibrin thread was observed. The time from blood withdrawal to clot formation was recorded using a stopwatch [27].
2.8 Western blotting
On experimental day 5, a subset of randomly selected mice from each group was sacrificed, and blood collected via cardiac puncture was centrifuged to collect plasma. Plasma protein concentrations were quantified using the Pierce Rapid Gold BCA protein assay kit. Proteins were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Mini-PROTEAN TGX Stain-Free Precast Gel, Bio-Rad) and transferred onto a polyvinylidene difluoride membrane. The membrane was blocked with 5% bovine serum albumin for 1 hour and incubated overnight at 4 °C with mouse anti-DNase-I antibody (1:200; clone B-4, sc-376207; Santa Cruz Biotechnology), followed by 1-hour incubation at room temperature with a goat anti-mouse Alexa Fluor 750-tagged secondary antibody (1:3000; A21037; Life technologies). Total protein loading was visualized and normalized using Ponceau S staining. Blots were visualized using the Bio-Rad ChemiDoc MP imaging system, and densitometry analysis was performed using the ImageJ software, version 1.53.
2.9 Statistical analysis
Data were analyzed using the GraphPad Prism 9 software. Two group comparisons were analyzed using the Student’s t-test. Multigroup comparisons were made using 1-way analysis of variance with adjustments for multiple comparisons, followed by the Tukey post hoc test. Results are expressed as mean ± SD. A p value of <.05 was considered to be statistically significant.
3 Results
3.1 rhDNase reversed proinflammatory activation after experimental ARDS
Intranasal poly(I:C) administration induced a robust innate immune response, characterized by neutrophil activation in the lungs and BALF (Figure 1A–C). These changes were paralleled by a lesser, but statistically significant, increase in the number of circulating neutrophils. Further analysis revealed the presence of NETs in the lungs, BALF, and blood (Figure 1B, C). The administration of rhDNase reduced the number of infiltrated neutrophils in the lungs and BALF following poly(I:C) administration, whereas the number of circulating neutrophils was not statistically different from that in the placebo-treated mice (Figure 1A–C). rhDNase also decreased the number of NETs in the lungs, BALF, and blood (Figure 1B, C). In parallel with these data, intranasal poly(I:C) administration significantly reduced the number of circulating T and B cells in the blood, compared with that in the saline-treated control mice (Supplementary Figure S1), to produce an elevated neutrophil-to-lymphocyte ratio. The administration of rhDNase, which decreased total neutrophils and NETs after poly(I:C), increased the total number of T and B cells in the lungs, BALF, and blood (Supplementary Figure S1). Moreover, poly(I:C) administration increased the expression of IL-6 and IL-17 in the lungs, BALF, and blood (Supplementary Figure S1). The expression of these proinflammatory cytokines was completely reversed by rhDNase administration. Conversely, the anti-inflammatory cytokine IL-10 was suppressed by poly(I:C) administration in the lungs and blood, although no significant changes were observed in BALF (Supplementary Figure S1). Increased plasma expression of DNase-I was observed in poly(I:C) mice treated with i.v. administration of 5 mg/kg of rhDNase compared with that in the sham mice or poly(I:C) mice treated with saline (placebo) (Supplementary Figure S2).Figure 1 Increased neutrophil infiltration and neutrophil extracellular trap (NET) formation in the lung and bronchoalveolar lavage fluid (BALF) after experimental acute respiratory distress syndrome. (A) Representative flow cytometry plots showing activation of CD45+ Ly6G+ neutrophils in the blood, lung tissue, or BALF on experimental day 5. Poly(I:C) mice were treated with either saline (placebo) or 5-mg/kg recombinant human DNase-I (rhDNase) (intravenous). Data are representative of 5 to 7 mice per group. (B) Quantification of poly(I:C)-induced neutrophil activation and NET formation in the lung, BALF, and blood after administration of placebo or rhDNase. Scatterplots represent 5 to 7 mice per group. Data were compared using 1-way analysis of variance, followed by the Tukey post hoc test. ∗p < .05, ∗∗p < .01, ∗∗∗∗p < .0001. (C) Immunohistochemistry of Ly6G+ Cit-H3+ NETs in lung tissue after administration of placebo or rhDNase. 4′,6-diamidino-2-phenylindole was used as a counterstain. Scale bar = 100 μm. Mean fluorescence intensity from 25 random fields from 5 mice per group was quantified and presented as mean ± SD. Data were compared using 1-way analysis of variance, followed by the Tukey post hoc test. ∗p < .05, ∗∗p < .01, ∗∗∗∗p < .0001. DAPI, 4′,6-diamidino-2-phenylindole; FSC, forward scatter; MFI, mean fluorescence intensity; SSC, side scatter.
3.2 rhDNase improves pulmonary structure after administration of intranasal poly(I:C)
Consistent with increased inflammation, gross analysis of the lungs after poly(I:C) administration revealed regions of hyperemia as well as pale tissue indicative of regional hypoperfusion and tissue hypoxia (Figure 2A). The administration of rhDNase reduced the relative occurrence of poly(I:C)-induced gross tissue injury compared with that in the placebo-treated control mice (Figure 2A). Histologic analysis of lung tissue revealed hypercellularity, indicating immune cell infiltration (peribronchiolar, perivascular, and interstitial inflammation) and diffuse alveolar damage after poly(I:C) administration (Figure 2B). These changes induced by poly(I:C) were partially reduced by rhDNase compared with that in the placebo-treated mice (Figure 2B). Immunofluorescence staining showed a distinct spatial and temporal overlap between infiltrated, Ly6G+ neutrophils and hypoxic lung tissue after poly(I:C) administration (Figure 2C). The administration of rhDNase reduced neutrophil infiltration in parallel with decreased pulmonary tissue hypoxia, which returned to sham levels (Figure 2C).Figure 2 Recombinant human DNase-I (rhDNase) reduces neutrophil infiltration and improves lung morphology after experimental acute respiratory distress syndrome. (A) Gross images of lung tissue on experimental day 6. Poly(I:C) mice were treated with intravenous administration of either saline (placebo) or 5-mg/kg rhDNase. Bilateral regions of hyperemia are denoted by black dotted lines. Regions of pale tissue, indicative of hypoperfusion and regional hypoxia, are labeled with blue dotted lines. Lung injury was observed across the lung lobes from placebo-treated mice, with less pronounced effects observed after rhDNase treatment. Scale bar = 5 mm. (B) Histologic assessment of lung tissues. Data are representative of 6 mice per group. Top panel: Note that the increased cellularity observed in the poly(I:C) group, consistent with inflammatory activation, was attenuated following rhDNase treatment. Scale bar = 200 μm. Middle panel (second and third rows): Peribronchiolar and perivascular inflammation was observed in the poly(I:C) group. These changes were reduced with rhDNase treatment. Scale bar = 50 μm. Bottom panel: Interstitial inflammation and diffuse alveolar damage, a feature associated with early stages of acute respiratory distress syndrome, were seen in the poly(I:C) group. rhDNase treatment mitigated interstitial inflammation. Scale bar = 50 μm. (C) Confocal microscope images of poly(I:C)-treated lung tissue on experimental day 6. Ly6G+ infiltrating neutrophils localized in regions of tissue hypoxia, as indicated by increased hypoxyprobe-1 fluorescence. Scale bar = 100 μm. The mean fluorescence intensity was quantified from 25 random fields from 5 mice per group. Scatterplots were compared using 1-way analysis of variance, followed by the Tukey post hoc test. ∗∗∗p < .001, ∗∗∗∗p < .0001. DAPI, 4′,6-diamidino-2-phenylindole; MFI, mean fluorescence intensity.
We next utilized microCT imaging to assess pulmonary function. Using volume-rendering images, we observed perfusion of the contrast dye in large, medium, and small pulmonary vessels of the control mice. Following poly(I:C) administration, the contrast dye was largely confined to larger vessels, with less detection in smaller vessels. Treatment with rhDNase partially reversed the deficits observed after poly(I:C) administration, with some smaller vessels exhibiting perfusion (Figure 3A). A similar pattern was observed using maximum-intensity projection images. In the control mice, the contrast dye was observed in both larger and smaller vessels, whereas poly(I:C) administration revealed a higher concentration of the dye in larger vessels only (Figure 3A). Three-dimensional reconstructions are provided in Supplementary Figures S3–S5. Quantitative analysis revealed no changes in blood vessel lumen thickness or separation between the groups (Figure 3B). In contrast, poly(I:C) administration reduced the vascular volume, capillary network, and number of blood vessels. Treatment with rhDNase completely reversed the effect of poly(I:C), returning values to levels observed in the sham-treated control mice (Figure 3B).Figure 3 Recombinant human DNase-I (rhDNase) improves pulmonary perfusion after experimental acute respiratory distress syndrome. (A) Ex vivo microcomputed tomography imaging of the lungs on experimental day 6. Poly(I:C) mice were treated with intravenous administration of either sterile phosphate-buffered saline or 5-mg/kg rhDNase. Representative images depict volume-rendering imaging and maximal intensity project. For volume-rendering imaging, every voxel is assigned an emission color and opacity, which indicates the intensity of the contrast dye within the vasculature. All lungs were scanned using identical parameters, and the same log-scaled histogram was used to color map the lung vasculature. Red color represents lower-intensity structures (vessels with lower contrast in their lumen), blue color represents higher-intensity structures (vessels with higher contrast in their lumen), and green is the range in between those 2. Dorsal, ventral, side, and transverse views are depicted and show perfusion deficits in the poly(I:C)-treated mice. Following rhDNase treatment, perfusion of smaller-caliber blood vessels was observed. (B) Quantification of vascular volume, capillary network, number of blood vessels (BVs), BV lumen thickness, and BV separation, as assessed using microcomputed tomography. rhDNase improved vascular volume, capillary network, and the number of BVs to sham-injured levels. Quantified data are from 6 mice per group and were analyzed using 1-way analysis of variance, followed by the Tukey post hoc test. ∗p < .05, ∗∗p < .01. MIP, maximum-intensity projection; VRI, volume-rendering imaging.
3.3 rhDNase reduces platelet activation and platelet-neutrophil aggregates after intranasal poly(I:C) administration
We observed disruption of vascular networks and an association between pulmonary hypoxia and infiltrated neutrophils/NETs. NETs provide a scaffold for platelet aggregation and thrombus formation; thus, we next explored whether rhDNase affected platelet activation. Poly(I:C) administration reduced the number of resting platelets and increased the number of activated platelets in the blood, lung tissue, and BALF on day 5 (Figure 4 ). The administration of rhDNase significantly increased the number of resting platelets in all groups following the administration of poly(I:C) while simultaneously suppressing the number of activated platelets in both blood and lung tissue; however, no significant effect of rhDNase on activated platelets was observed in BALF (Figure 4). In line with these data, rhDNase reduced the number of activated platelet-NET aggregates in lung tissue compared with that in the placebo-treated mice (Figure 5 ).Figure 4 Recombinant human DNase-I reduces platelet activation in blood and lung tissue after experimental acute respiratory syndrome. (A) Quantification of platelet activation in blood, lung tissue, or bronchoalveolar lavage fluid on experimental day 5. Poly(I:C) mice were treated with intravenous administration of either sterile phosphate-buffered saline or 5-mg/kg recombinant human DNase-I. Top panels depict flow cytometry plots to differentiate CD41+ (resting) and CD62P+ (activated) platelets. (B) Quantification of platelet activation data from 7 mice per group was analyzed using 1-way analysis of variance, followed by the Tukey post hoc test. ∗p < .05, ∗∗p < .01, ∗∗∗p < .001, ∗∗∗∗p < .0001. BALF, bronchoalveolar lavage fluid; FSC, forward scatter.
Figure 5 Recombinant human DNase-I attenuates activated platelet-neutrophil extracellular trap (NET) macroaggregates in lung tissue after experimental acute respiratory distress syndrome. Association of activated platelets (CD41+ CD62p+) with NETs in lung tissue on experimental day 5, as assessed using flow cytometry. Poly(I:C) mice were treated with intravenous administration of either sterile phosphate-buffered saline or 5-mg/kg recombinant human DNase-I. Quantification of platelet-NET aggregates from 7 mice per group was analyzed using 1-way analysis of variance, followed by the Tukey post hoc test. ∗∗∗∗p < .0001. FSC, forward scatter; SSC, side scatter.
3.4 rhDNase normalizes clotting time after experimental ARDS
Given the increased number of activated platelets following intranasal poly(I:C) administration, we explored whether these changes reduced clotting times in a mixed-sex cohort of mice. A significant reduction in clotting time was observed in whole blood collected from the placebo-treated poly(I:C) mice compared with that in untreated control mice. Furthermore, i.v. administration of rhDNase reversed the effect of poly(I:C), normalizing the clotting time to control levels (Figure 6A). As male sex was associated with a 3-fold higher risk of admission to the intensive care unit and an increased risk of death after SARS-CoV-2 infection [28], we further assessed whether the observed differences in clotting were sex-dependent after poly(I:C) administration. Of note, we found that the effect of poly(I:C) was lost in females compared with that in males, suggesting that male sex was associated with elevated coagulation (Figure 6B). Finally, poly(I:C) treatment of whole blood from healthy human donors increased NETs (Figure 6C), reduced the number of resting platelets, and increased the number of activated platelets (Figure 6D), mirroring the response in mice after experimental ARDS (Figure 4). The addition of rhDNase completely reversed the number of both NETs and activated platelets to control levels, with a simultaneous increase in the number of resting platelets (Figure 6C, D).Figure 6 Recombinant human DNase-I (rhDNase) normalizes the clotting time after experimental acute respiratory distress syndrome. (A) Ex vivo quantification of the clotting time of whole blood collected from mixed-sex mice on experimental day 5. Administration of poly(I:C) reduced the clotting time in placebo-treated mice, whereas rhDNase treatment normalized the clotting time to control levels. Data were compared using 1-way analysis of variance (ANOVA), followed by the Tukey post hoc test. ∗∗p < .01. (B) Sex-dependent changes in clotting time were observed following experimental acute respiratory distress syndrome, with males exhibiting a more robust response to poly(I:C)-induced clotting compared with female mice. Data were compared using 1-way ANOVA, followed by the Tukey post hoc test. ∗∗p < .01, ∗∗∗p < .001. Quantification of (C) neutrophil extracellular traps and (D) platelet activation in healthy donor human blood following poly(I:C) treatment ex vivo. Poly(I:C) increased the number of activated platelets, with a concomitant reduction in resting platelets. Treatment with rhDNase normalized platelet activity to untreated control levels. Data were analyzed using 1-way ANOVA, followed by the Tukey post hoc test. ∗∗p < .01, ∗∗∗p < .001, ∗∗∗∗p < .0001. FSC, forward scatter; NET, neutrophil extracellular trap; SSC, side scatter.
4 Discussion
Throughout the recent pandemic, critically ill patients overwhelmed hospitals and intensive care units worldwide, with many patients requiring mechanical ventilation. While rapid vaccine development reduced the number of hospitalized patients, identification of safe, efficacious therapeutics is needed to limit the deleterious consequences of pulmonary inflammation and/or structural injury following viral infection. In this study, we provide experimental evidence to support the clinical repurposing of rhDNase, a Food and Drug Administration–approved therapy with an excellent patient safety record, to reduce inflammatory activation, degrade NETs, and improve pulmonary structure after experimental ARDS.
Individuals infected with SARS-CoV-2 exhibit early alveolar damage, followed by a robust immune reaction, platelet hyperactivation, hypercoagulation, and microvascular pulmonary thrombosis [29]. A progressive endothelial thromboinflammatory syndrome, termed as “microvascular COVID-19 lung vessels obstructive thromboinflammatory syndrome” was observed in a study of 850 patients with COVID-19 and bilateral pneumonia [29,30]; however, the underlying mechanisms are undefined. Neutrophils derived from patients with COVID-19 exhibited excessive NET formation, whereas NETs contributed to immunothrombosis, vascular occlusion, and poor clinical outcomes in patients with COVID-19 [12,19,31,32]. Impaired clearance of NETs, which occurs in patients with ARDS [33], is clinically associated with acute thrombotic microangiopathies [34], whereas the presence of citrullinated histone H3, a component of NETs in retrieved thrombi, was independently associated with mortality in patients with acute ischemic stroke [35]. In addition, we reported that NET formation was associated with microvascular occlusion and cerebral hypoperfusion after brain injury in mice and humans [24]. Notably, patients with COVID-19 requiring intensive care frequently exhibit ≥1 comorbid conditions, including obesity and hypertension, which exacerbate NET formation [15,18,36]. In this study, we did not observe a change in plasma DNase-I levels following poly(I:C) administration, suggesting that experimental ARDS is associated with elevated NET formation rather than impaired degradation. Thus, targeted degradation of NETs may reduce microthrombus formation, improve tissue perfusion, and enhance gas exchange in patients with COVID-19, while circumventing the associated risks of global neutrophil depletion [12,31].
Platelets contribute to tissue damage in experimental models of acute lung injury, at least in part, via the release of inflammatory mediators and neutrophil activation at sites of endothelial injury [37,38]. Consistent with the aforementioned clinical findings, an increased presence of neutrophils and NETs was observed in the blood, lung tissue, and BALF after poly(I:C) administration, paralleling increases in platelet activation. In particular, we observed a spatial overlap between infiltrated neutrophils and NETs in regions of insufficient perfusion and pulmonary hypoxia after experimental ARDS, consistent with a purported role of NETs in microthrombus formation. In line with our observation of increased activated platelet-NET complexes after experimental ARDS, rolling neutrophils extract large fragments from dying platelets to generate neutrophil macroaggregates, resulting in widespread occlusion of pulmonary arteries, veins, and microvasculature after experimental gut ischemia [39]. Although international case-control studies revealed that elevated levels of platelet activation markers and dysregulated coagulation in BALF were associated with increased mortality in patients with ARDS [40], administration of antiplatelet drugs produced inconsistent improvements in patients with ARDS [37], suggesting a need for alternative therapeutic approaches. That rhDNase reduced platelet-neutrophil complexes in parallel with improved lung perfusion indicates a potential mechanism of action to support therapeutic potential in patients with ARDS.
DNase-I is an endonuclease that catalyzes the hydrolysis of extracellular DNA. Aerosolized rhDNase, a monomeric, 260-amino–acid glycoprotein produced in Chinese hamster ovary cells [41], has been used for >3 decades to reduce high-molecular–weight DNA to smaller fragments to reduce mucus viscosity and improve airway function in patients with cystic fibrosis [42]. Given its clinical safety and efficacy, the use of rhDNase has extended to patients with prolonged mechanical ventilation, chronic sinusitis, emphysema, and pediatric pulmonary diseases [[43], [44], [45], [46]]. Moreover, rhDNase has been shown by our laboratory and others to efficiently degrade NETs, which comprise, in part, extracellular DNA that is expelled from activated neutrophils. While the precise relationship between NETs and pulmonary hypoxia is undetermined, the administration of rhDNase degraded extracellular NETs, decreased lung inflammation, and improved pulmonary perfusion after poly(I:C) treatment. Coupled with data from our laboratory and others who showed that rhDNase improved blood flow and outcomes after experimental stroke and traumatic brain injury [24,47], rhDNase may provide a safe, feasible, and clinically efficacious treatment option to improve lung function during ARDS.
Despite the translational potential of rhDNase, several study limitations warrant consideration. In the present study, i.v. rhDNase administration was utilized to preferentially target circulating NETs following experimental ARDS; however, the short plasma half-life of DNase-I may provide a potential translational limitation [48]. While aerosolized rhDNase may partially overcome this limitation, the recent development of recombinant DNase-I–coated polydopamine-polyethylene glycol nanoparticles may provide longer-acting enzymatic activity, with delayed excretion [49]. Of note, long-acting DNase-I reduced the number of peritoneal neutrophils and reduced mortality after experimental sepsis compared with that in control or DNase-I-treated mice [49]. Thus, comparison of the routes of administration (inhalational vs. i.v.) and different formulations may establish preclinical support prior to clinical translation.
In addition to extracellular DNA, NETs contain histones and granular proteins (eg, MPO and NE) that may directly contribute to tissue damage. While not explored in this study, the addition of antihistone antibodies, polysialic acid (binds histones), or MPO (dihydrolipoic acid) reduced NET-mediated cytotoxicity in A549 human lung adenocarcinoma cells [50]. Additionally, rhDNase reduced thrombus formation via a NET-independent mechanism by hydrolyzing adenosine triphosphate and adenosine diphosphate to adenosine, which in turn inhibited platelet aggregation and neutrophil activation after laser-induced injury [51]. Thus, adjunct therapies that target other aspects of NETs may provide further protection after ARDS. In support of this possibility, a phase 2 multicenter, double-blind, randomized, placebo-controlled trial concluded that i.v. administration of α-1 antitrypsin, a serine protease inhibitor that limits NE activity, was safe and well tolerated and decreased inflammation in patients with moderate-to-severe ARDS secondary to SARS-CoV-2 infection [52]. As such, combination therapy with rhDNase and α-1 antitrypsin may provide maximal protection against the detrimental effects of NETs.
Finally, intranasal poly(I:C) recapitulates many aspects of viral infection, including systemic and local inflammation as well as pulmonary injury; however, this sterile inflammation model may not fully replicate infectious ARDS. Importantly, poly(I:C) is a synthetic analog of double-stranded RNA, which is a molecular pattern associated with viral infections. Along these lines, poly(I:C) binds to TLR3, a pattern recognition receptor that recognizes RNA viruses, including influenza A, coronavirus, and rhinovirus. Future studies will be expanded to infectious disease models to better determine the clinical reach of our findings. An i.v. bolus of rhDNase (5 mg/kg) was chosen based on reports showing maximal NET degradation [24]. Future translational work will also optimize the dose and therapeutic window to maximize efficacy after ARDS. We postulate that rhDNase degrades NETs to improve outcomes; however, we cannot exclude the possibility that degradation of other extracellular DNA contributes to the observed benefits. Extracellular DNA does not typically circulate under physiological conditions, supporting the rationale and safety of repurposing rhDNase for ARDS; however, elevated levels of circulating mitochondrial DNA predicted poor outcomes in hospitalized patients with COVID-19 [53]. As mitochondrial DNA can potentiate inflammatory activation and NET formation via activation of TLR9 [54], rhDNase may reduce pulmonary inflammation and improve function via multiple mechanisms of action. Additionally, medium-term treatment (8 days) with rhDNase reduced NET-dependent thrombosis in a murine breast cancer model, whereas long-term treatment (18 days) reduced overall survival, an effect that was ameliorated by a combination of rhDNase with ertapenem [55]. In our study, short-to-medium-term treatment (3 days) with rhDNase diminished NETs, reduced platelet activation, and limited activated platelet-NETs macroaggregates, which in turn normalized the clotting time after experimental ARDS. The effects of long-term rhDNase treatment require further investigation along with studies of concomitant rhDNase treatment with antibiotics for those developing sepsis. While a reduction in NETs could heighten the risk of infection in critically ill patients, PAD4−/− mice, which cannot form NETs [56], exhibited decreased organ dysfunction and improved survival following hemorrhagic shock and sepsis [57]. Moreover, elevated NETs are positively correlated with the severity of sepsis in pediatric patients, whereas treatment with rhDNase or a PAD4 inhibitor ameliorated sepsis in infant mice, which exhibit significantly more NETs than adult mice [58]. Thus, the suitability of rhDNase in the management of pediatric ARDS requires further exploration. Taken together, our findings suggest that rhDNase is a low-risk, cost-effective, and efficacious candidate for drug repurposing to target NETs in the context of ARDS.
Supplemental Data
Supplemental Figure 1 rhDNase suppresses inflammatory activation after experimental ARDS. (A) Quantification of lymphocyte activation in lung tissue, BALF, or blood at experimental d5. Poly(I:C) mice were treated with intravenous injections of either sterile PBS or 5 mg/kg rhDNase. Data were analyzed using flow cytometry and are representative of n=3-6 mice/group. (B) Quantification of poly(I:C)-induced inflammatory cytokine (IL-6, IL-17, IL-10) expression in lung, BALF, and blood after administration of placebo or rhDNase. Data were analyzed using flow cytometry and scatterplots represent n=6-8 mice/group. For all panels, data were compared by one-way ANOVA followed by Tukey’s post hoc test. ∗p<0.05, ∗∗p<0.01, ∗∗∗p<0.001, ∗∗∗∗p<0.0001.
Supplemental Figure 2 Plasma DNase-I levels are increased at day 5 following rhDNase intravenous administration. (A) Plasma was collected from randomly selected mice from each group at experimental day 5. Western blots were probed for DNase-I. (B) Total protein loading was visualized by Ponceau S staining. (C) DNase-I expression was normalized to total protein. Densitometry analysis revealed the increased expression of plasma DNase-I in poly(I:C) mice treated with intravenous 5mg/kg rhDNase, as compared to sham and poly(I:C) mice treated with saline (placebo). Data are representative of n=3 mice/group and were compared by one-way ANOVA followed by Tukey’s post hoc test. ∗∗p<0.01.
Supplemental Figure 3
Three-dimensional reconstruction of the lung in placebo-treated sham mice.
Supplemental Figure 4
Three-dimensional reconstruction of the lung in placebo-treated mice following intranasal administration of poly(I:C).
Supplemental Figure 5
Three-dimensional reconstruction of the lung in rhDNase-treated mice following intranasal administration of poly(I:C).
Author contributions
A.J., N.S.M., Y.L., K.V., B.B., and K.M.D. conceived the study and analyzed the data. A.J. and K.V. performed acute respiratory distress syndrome modeling. A.J. and M.E.A. performed microcomputed tomography studies. H.K. and E.S.L. performed flow cytometry studies. K.M.D. oversaw the project, had full access to all the data in the study, and takes responsibility for its integrity and the data analysis. All authors provided input regarding experimental design, data analysis, and manuscript preparation.
Declaration of competing interests
There are no competing interests to disclose.
Funding information Grant Number: R01NS110378 (to K.M.D. and B.B.), R01NS117565 (to K.M.D.), and R01NS114560 (to K.V.) from the 10.13039/100000002 National Institutes of Health .
Manuscript handled by: Christophe Dubois
Final decision: Christophe Dubois, 28 April 2023
The online version contains supplementary material available at https://doi.org/10.1016/j.jtha.2023.04.044
==== Refs
References
1 Matthay M.A. Zemans R.L. Zimmerman G.A. Arabi Y.M. Beitler J.R. Mercat A. Herridge M. Randolph A.G. Calfee C.S. Acute respiratory distress syndrome Nat Rev Dis Primers 5 2019 18 30872586
2 Bellani G. Laffey J.G. Pham T. Fan E. Brochard L. Esteban A. Gattinoni L. Van Haren F. Larsson A. McAuley D.F. Ranieri M. Rubenfeld G. Thompson B.T. Wrigge H. Slutsky A.S. Pesenti A. LUNG SAFE Investigators; ESICM Trials Group Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries JAMA 315 2016 788 800 26903337
3 Force A.D.T. Ranieri V.M. Rubenfeld G.D. Thompson B.T. Ferguson N.D. Caldwell E. Fan E. Camporota L. Slutsky A.S. Acute respiratory distress syndrome: the Berlin Definition JAMA 307 2012 2526 2533 22797452
4 Luyt C.E. Combes A. Trouillet J.L. Nieszkowska A. Chastre J. Virus-induced acute respiratory distress syndrome: epidemiology, management and outcome Presse Med 40 2011 e561 –8 22094172
5 Tang N. Bai H. Chen X. Gong J. Li D. Sun Z. Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019 patients with coagulopathy J Thromb Haemost 18 2020 1094 1099 32220112
6 Connors J.M. Levy J.H. COVID-19 and its implications for thrombosis and anticoagulation Blood 135 2020 2033 2040 32339221
7 Gupta N. Zhao Y.Y. Evans C.E. The stimulation of thrombosis by hypoxia Thromb Res 181 2019 77 83 31376606
8 Kollias A. Kyriakoulis K.G. Dimakakos E. Poulakou G. Stergiou G.S. Syrigos K. Thromboembolic risk and anticoagulant therapy in COVID-19 patients: emerging evidence and call for action Br J Haematol 189 2020 846 847 32304577
9 Levi M. van der Poll T. Inflammation and coagulation Crit Care Med 38 2010 S26 S34 20083910
10 Abraham E. Carmody A. Shenkar R. Arcaroli J. Neutrophils as early immunologic effectors in hemorrhage- or endotoxemia-induced acute lung injury Am J Physiol Lung Cell Mol Physiol 279 2000 L1137 L1145 11076804
11 Liu Y. Du X. Chen J. Jin Y. Peng L. Wang H.H.X. Luo M. Chen L. Zhao Y. Neutrophil-to-lymphocyte ratio as an independent risk factor for mortality in hospitalized patients with COVID-19 J Infect 81 2020 e6 e12
12 Barnes B.J. Adrover J.M. Baxter-Stoltzfus A. Borczuk A. Cools-Lartigue J. Crawford J.M. Daßler-Plenker J. Guerci P. Huynh C. Knight J.S. Loda M. Looney M.R. McAllister F. Rayes R. Renaud S. Rousseau S. Salvatore S. Schwartz R.E. Spicer J.D. Yost C.C. Targeting potential drivers of COVID-19: neutrophil extracellular traps J Exp Med 217 2020 e20200652
13 Fox S.E. Akmatbekov A. Harbert J.L. Li G. Quincy Brown J. Vander Heide R.S. Pulmonary and cardiac pathology in African American patients with COVID-19: an autopsy series from New Orleans Lancet Respir Med 8 2020 681 686 32473124
14 Heflin A.C. Brigham K.L. Prevention by granulocyte depletion of increased vascular permeability of sheep lung following endotoxemia J Clin Invest 68 1981 1253 1260 7298850
15 Jiang S. Park D.W. Tadie J.M. Gregoire M. Deshane J. Pittet J.F. Abraham E. Zmijewski J.W. Human resistin promotes neutrophil proinflammatory activation and neutrophil extracellular trap formation and increases severity of acute lung injury J Immunol 192 2014 4795 4803 24719460
16 Narasaraju T. Yang E. Samy R.P. Ng H.H. Poh W.P. Liew A.A. Phoon M.C. van Rooijen N. Chow V.T. Excessive neutrophils and neutrophil extracellular traps contribute to acute lung injury of influenza pneumonitis Am J Pathol 179 2011 199 210 21703402
17 Fuchs T.A. Brill A. Duerschmied D. Schatzberg D. Monestier M. Myers D.D. Jr. Wrobleski S.K. Wakefield T.W. Hartwig J.H. Wagner D.D. Extracellular DNA traps promote thrombosis Proc Natl Acad Sci U S A 107 2010 15880 15885 20798043
18 Sharma S. Hofbauer T.M. Ondracek A.S. Chausheva S. Alimohammadi A. Artner T. Panzenboeck A. Rinderer J. Shafran I. Mangold A. Winker R. Wohlschläger-Krenn E. Moser B. Taghavi S. Klepetko W. Preissner K.T. Lang I.M. Neutrophil extracellular traps promote fibrous vascular occlusions in chronic thrombosis Blood 137 2021 1104 1116 33512471
19 Middleton E.A. He X.Y. Denorme F. Campbell R.A. Ng D. Salvatore S.P. Mostyka M. Baxter-Stoltzfus A. Borczuk A.C. Loda M. Cody M.J. Manne B.K. Portier I. Harris E.S. Petrey A.C. Beswick E.J. Caulin A.F. Iovino A. Abegglen L.M. Weyrich A.S. Neutrophil extracellular traps contribute to immunothrombosis in COVID-19 acute respiratory distress syndrome Blood 136 2020 1169 1179 32597954
20 Earhart A.P. Holliday Z.M. Hofmann H.V. Schrum A.G. Consideration of dornase alfa for the treatment of severe COVID-19 acute respiratory distress syndrome New Microbes New Infect 35 2020 100689
21 Salles E.L. Khodadadi H. Jarrahi A. Ahluwalia M. Paffaro V.A. Jr. Costigliola V. Yu J.C. Hess D.C. Dhandapani K.M. Baban B. Cannabidiol (CBD) modulation of apelin in acute respiratory distress syndrome J Cell Mol Med 24 2020 12869 12872 33058425
22 Khodadadi H. Salles E.L. Jarrahi A. Chibane F. Costigliola V. Yu J.C. Vaibhav K. Hess D.C. Dhandapani K.M. Baban B. Cannabidiol modulates cytokine storm in acute respiratory distress syndrome induced by simulated viral infection using synthetic RNA Cannabis Cannabinoid Res 5 2020 197 201 32923657
23 Sun F. Xiao G. Qu Z. Murine bronchoalveolar lavage Bio Protoc 7 2017 e2287
24 Vaibhav K. Braun M. Alverson K. Khodadadi H. Kutiyanawalla A. Ward A. Banerjee C. Sparks T. Malik A. Rashid M.H. Khan M.B. Waters M.F. Hess D.C. Arbab A.S. Vender J.R. Hoda N. Baban B. Dhandapani K.M. Neutrophil extracellular traps exacerbate neurological deficits after traumatic brain injury Sci Adv 6 2020 eaax8847
25 Braun M. Vaibhav K. Saad N. Fatima S. Brann D.W. Vender J.R. Wang L.P. Hoda M.N. Baban B. Dhandapani K.M. Activation of myeloid TLR4 mediates T lymphocyte polarization after traumatic brain injury J Immunol 198 2017 3615 3626 28341672
26 Vaibhav K. Braun M. Khan M.B. Fatima S. Saad N. Shankar A. Khan Z.T. Harris R.B.S. Yang Q. Huo Y. Arbab A.S. Giri S. Alleyne C.H. Jr. Vender J.R. Hess D.C. Baban B. Hoda M.N. Dhandapani K.M. Remote ischemic post-conditioning promotes hematoma resolution via AMPK-dependent immune regulation J Exp Med 215 2018 2636 2654 30190288
27 Wang Y. Hao J. Gao W. Liu Z. Wu S. Jing S. Study on hemostatic activities of the rhizome of Paris bashanensis Pharm Biol 51 2013 1321 1325 23855262
28 Peckham H. de Gruijter N.M. Raine C. Radziszewska A. Ciurtin C. Wedderburn L.R. Rosser E.C. Webb K. Deakin C.T. Male sex identified by global COVID-19 meta-analysis as a risk factor for death and ITU admission Nat Commun 11 2020 6317 33298944
29 Ciceri F. Beretta L. Scandroglio A.M. Colombo S. Landoni G. Ruggeri A. Peccatori J. D'Angelo A. De Cobelli F. Rovere-Querini P. Tresoldi M. Dagna L. Zangrillo A. Microvascular COVID-19 lung vessels obstructive thromboinflammatory syndrome (MicroCLOTS): an atypical acute respiratory distress syndrome working hypothesis Crit Care Resusc 22 2020 95 97 32294809
30 Renzi S. Landoni G. Zangrillo A. Ciceri F. MicroCLOTS pathophysiology in COVID 19 Korean J Intern Med 2020 10.3904/kjim.2020.336
31 Zuo Y. Yalavarthi S. Shi H. Gockman K. Zuo M. Madison J.A. Blair C. Weber A. Barnes B.J. Egeblad M. Woods R.J. Kanthi Y. Knight J.S. Neutrophil extracellular traps in COVID-19 JCI Insight 5 2020 e138999
32 Leppkes M. Knopf J. Naschberger E. Lindemann A. Singh J. Herrmann I. Stürzl M. Staats L. Mahajan A. Schauer C. Kremer A.N. Völkl S. Amann K. Evert K. Falkeis C. Wehrfritz A. Rieker R.J. Hartmann A. Kremer A.E. Neurath M.F. Vascular occlusion by neutrophil extracellular traps in COVID-19 EBioMedicine 58 2020 102925
33 Grégoire M. Uhel F. Lesouhaitier M. Gacouin A. Guirriec M. Mourcin F. Dumontet E. Chalin A. Samson M. Berthelot L.L. Tissot A. Kerjouan M. Jouneau S. Le Tulzo Y. Tarte K. Zmijewski J.W. Tadié J.M. Impaired efferocytosis and neutrophil extracellular trap clearance by macrophages in ARDS Eur Respir J 52 2018 1702590
34 Jimenez-Alcazar M. Napirei M. Panda R. Köhler E.C. Hovinga J.K. Mannherz H.G. Peine S. Renne T. Lammle B. Fuchs T.A. Impaired DNase1-mediated degradation of neutrophil extracellular traps is associated with acute thrombotic microangiopathies J Thromb Haemost 13 2015 732 742 25418346
35 Laridan E. Denorme F. Desender L. Francois O. Andersson T. Deckmyn H. Vanhoorelbeke K. De Meyer S.F. Neutrophil extracellular traps in ischemic stroke thrombi Ann Neurol 82 2017 223 232 28696508
36 Chrysanthopoulou A. Gkaliagkousi E. Lazaridis A. Arelaki S. Pateinakis P. Ntinopoulou M. Mitsios A. Antoniadou C. Argyriou C. Georgiadis G.S. Papadopoulos V. Giatromanolaki A. Ritis K. Skendros P. Angiotensin II triggers neutrophil extracellular traps release linking thromboinflammation with essential hypertension JCI Insight 6 2021 e148668
37 Yadav H. Kor D.J. Platelets in the pathogenesis of acute respiratory distress syndrome Am J Physiol Lung Cell Mol Physiol 309 2015 L915 L923 26320157
38 Zarbock A. Singbartl K. Ley K. Complete reversal of acid-induced acute lung injury by blocking of platelet-neutrophil aggregation J Clin Invest 116 2006 3211 3219 17143330
39 Yuan Y. Alwis I. Wu M.C.L. Kaplan Z. Ashworth K. Bark D. Jr. Pham A. Mcfadyen J. Schoenwaelder S.M. Josefsson E.C. Kile B.T. Jackson S.P. Neutrophil macroaggregates promote widespread pulmonary thrombosis after gut ischemia Sci Transl Med 9 2017 eaam5861
40 Wang T. Liu Z. Wang Z. Duan M. Li G. Wang S. Li W. Zhu Z. Wei Y. Christiani D.C. Li A. Zhu X. Thrombocytopenia is associated with acute respiratory distress syndrome mortality: an international study PLoS One 9 2014 e94124
41 Shak S. Aerosolized recombinant human DNase I for the treatment of cystic fibrosis Chest 107 1995 65S 70S 7842816
42 Konstan M.W. Ratjen F. Effect of dornase alfa on inflammation and lung function: potential role in the early treatment of cystic fibrosis J Cyst Fibros 11 2012 78 83 22093951
43 Cimmino M. Nardone M. Cavaliere M. Plantulli A. Sepe A. Esposito V. Mazzarella G. Raia V. Dornase alfa as postoperative therapy in cystic fibrosis sinonasal disease Arch Otolaryngol Head Neck Surg 131 2005 1097 1101 16365224
44 Rahman N.M. Maskell N.A. West A. Teoh R. Arnold A. Mackinlay C. Peckham D. Davies C.W. Ali N. Kinnear W. Bentley A. Kahan B.C. Wrightson J.M. Davies H.E. Hooper C.E. Lee Y.C. Hedley E.L. Crosthwaite N. Choo L. Helm E.J. Intrapleural use of tissue plasminogen activator and DNase in pleural infection N Engl J Med 365 2011 518 526 21830966
45 Scala M. Hoy D. Bautista M. Palafoutas J.J. Abubakar K. Pilot study of dornase alfa (Pulmozyme) therapy for acquired ventilator-associated infection in preterm infants Pediatr Pulmonol 52 2017 787 791 28052587
46 Simpson G. Roomes D. Reeves B. Successful treatment of empyema thoracis with human recombinant deoxyribonuclease Thorax 58 2003 365 366 12668806
47 Peña-Martínez C. Durán-Laforet V. García-Culebras A. Ostos F. Hernández-Jiménez M. Bravo-Ferrer I. Pérez-Ruiz A. Ballenilla F. Díaz-Guzmán J. Pradillo J.M. Lizasoain I. Moro M.A. Pharmacological modulation of neutrophil extracellular traps reverses thrombotic stroke tPA (tissue-type plasminogen activator) resistance Stroke 50 2019 3228 3237 31526124
48 Prince W.S. Baker D.L. Dodge A.H. Ahmed A.E. Chestnut R.W. Sinicropi D.V. Pharmacodynamics of recombinant human DNase I in serum Clin Exp Immunol 113 1998 289 296 9717980
49 Lee Y.Y. Park H.H. Park W. Kim H. Jang J.G. Hong K.S. Lee J.Y. Seo H.S. Na D.H. Kim T.H. Choy Y.B. Ahn J.H. Lee W. Park C.G. Long-acting nanoparticulate DNase-1 for effective suppression of SARS-CoV-2-mediated neutrophil activities and cytokine storm Biomaterials 267 2021 120389
50 Saffarzadeh M. Juenemann C. Queisser M.A. Lochnit G. Barreto G. Galuska S.P. Lohmeyer J. Preissner K.T. Neutrophil extracellular traps directly induce epithelial and endothelial cell death: a predominant role of histones PLoS One 7 2012 e32366
51 Carminita E. Crescence L. Brouilly N. Altie A. Panicot-Dubois L. Dubois C. DNAse-dependent, NET-independent pathway of thrombus formation in vivo Proc Natl Acad Sci U S A 118 2021 e2100561118
52 McElvaney O.J. McEvoy N.L. Boland F. McElvaney O.F. Hogan G. Donnelly K. Friel O. Browne E. Fraughen D.D. Murphy M.P. Clarke J. Choileáin O.N. O'Connor E. McGuinness R. Boylan M. Kelly A. Hayden J.C. Collins A.M. Cullen A. Hyland D. A randomized, double-blind, placebo-controlled trial of intravenous alpha-1 antitrypsin for ARDS secondary to COVID-19 Med 3 2022 233 248 35291694
53 Scozzi D. Cano M. Ma L. Zhou D. Zhu J.H. O'Halloran J.A. Goss C. Rauseo A.M. Liu Z. Sahu S.K. Peritore V. Rocco M. Ricci A. Amodeo R. Aimati L. Ibrahim M. Hachem R. Kreisel D. Mudd P.A. Kulkarni H.S. Circulating mitochondrial DNA is an early indicator of severe illness and mortality from COVID-19 JCI Insight 6 2021 e143299
54 Zhang Q. Raoof M. Chen Y. Sumi Y. Sursal T. Junger W. Brohi K. Itagaki K. Hauser C.J. Circulating mitochondrial DAMPs cause inflammatory responses to injury Nature 464 2010 104 107 20203610
55 Varady C.B.S. Oliveira A.C. Monteiro R.Q. Gomes T. Recombinant human DNase I for the treatment of cancer-associated thrombosis: a pre-clinical study Thromb Res 203 2021 131 137 34015562
56 Martinod K. Fuchs T.A. Zitomersky N.L. Wong S.L. Demers M. Gallant M. Wang Y. Wagner D.D. PAD4-deficiency does not affect bacteremia in polymicrobial sepsis and ameliorates endotoxemic shock Blood 125 2015 1948 1956 25624317
57 Biron B.M. Chung C.S. Chen Y. Wilson Z. Fallon E.A. Reichner J.S. Ayala A. PAD4 deficiency leads to decreased organ dysfunction and improved survival in a dual insult model of hemorrhagic shock and sepsis J Immunol 200 2018 1817 1828 29374076
58 Colón D.F. Wanderley C.W. Franchin M. Silva C.M. Hiroki C.H. Castanheira F.V.S. Donate P.B. Lopes A.H. Volpon L.C. Kavaguti S.K. Borges V.F. Speck-Hernandez C.A. Ramalho F. Carlotti A.P. Carmona F. Alves-Filho J.C. Liew F.Y. Cunha F.Q. Neutrophil extracellular traps (NETs) exacerbate severity of infant sepsis Crit Care 23 2019 113 30961634
|
PMC010xxxxxx/PMC10186287.txt |
==== Front
Indian J Gynecol Oncol
Indian J Gynecol Oncol
Indian Journal of Gynecologic Oncology
2363-8397
2363-8400
Springer India New Delhi
37214638
719
10.1007/s40944-023-00719-3
Original Article
The Effect of Fear of COVID-19 on Women's Attitudes toward Cancer Screening and Healthy Lifestyle Behaviors: A Cross-Sectional Study
http://orcid.org/0000-0001-8242-2773
Calpbinici Pelin pelince2@yandex.com
1
https://orcid.org/0000-0001-6985-2971
Uzunkaya Öztoprak Pınar 2
1 grid.449442.b 0000 0004 0386 1930 Department of Obstetrics and Gynecology Nursing, Semra and Vefa Küçük Faculty of Health Sciences, Nevşehir Hacı Bektaş Veli University, Nevşehir, Turkey
2 grid.14442.37 0000 0001 2342 7339 Department of Obstetrics and Gynecology Nursing, Faculty of Nursing, Hacettepe University, Ankara, Turkey
16 5 2023
2023
21 2 4523 1 2023
4 4 2023
30 4 2023
© The Author(s) under exclusive licence to Association of Gynecologic Oncologists of India 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Purpose
The study was conducted to evaluate the effect of fear of COVID-19 on women's attitudes toward cancer screening and healthy lifestyle behaviors.
Method
The study is of descriptive and cross-sectional type. The sample of the study consisted of 221 women living in Turkey. Research data were collected using Introductory Information Form, Attitude Scale for Cancer Screening, The Fear of COVID-19 Scale and Healthy Lifestyle Behaviors Scale II (HLBS-II).
Results
It was found out that 92.3% of the women did not have cancer screening during the pandemic period, 33.0% of the women who did not have it because they were afraid of the contamination, 33.0% thought they were healthy, 13.1% did not have screening tests because they thought that screening tests were not easy and accessible during the pandemic period. While no significant relationship was found between women's attitudes toward cancer screenings and fear of COVID-19 (P > 0.05), a positive significant relationship was found between women's attitudes toward cancer screenings and spiritual growth, health responsibility and interpersonal relations scores, which are sub-dimensions of the HLBS-II scale (P > 0.05). In addition, it was found out that women's fear of COVID-19 affected interpersonal relations and stress management (P < 0.05).
Conclusion
In our study, it was concluded that most of the women did not have cancer screening during the pandemic, and that the fear of COVID-19 affected such healthy lifestyle behaviors as interpersonal relations and stress management.
Keywords
Fear of COVID-19
Cancer screening
Healthy lifestyle behaviors
issue-copyright-statement© Association of Gynecologic Oncologists of India 2023
==== Body
pmcIntroduction
Coronavirus (COVID-19) first emerged in China's Hubei Province in December 2019 and rapidly affected the whole world [1, 2]. The rapid spread of COVID-19, its rapid mutation, the uncertainty of mutations, the lack of definitive treatment and effective prevention methods have affected many aspects of human life and activities [3].
As a disadvantaged gender, women are more affected by and afraid of epidemics. During the COVID-19 pandemic, there have been situations in which women could not meet their own age-related needs such as reproductive health, menopause and old age, as well as the risks they have been exposed to together with the general population, and in order to get those needs met, they have frequently had to apply to hospitals and to get in contact with the health professionals where and for whom the rate of contamination is high [4, 5]. The increasing fear among women about COVID-19 disease may adversely affect their ability to benefit from health care services, for the epidemic has had a tremendous impact on our health systems as much as it has affected our lives. It is possible that the fear of contracting COVID-19 will cause a large number of deaths such as those due to cancer due to diagnostic delays of life-threatening diseases. Studies report that cancer screenings have decreased compared to the pre-pandemic period, and that the number of individuals at high risk for cancer has increased [6, 7]. On the other hand, despite the introduction of various restrictions due to the epidemic, cancer screening continues in accordance with the national cancer screening standards of the countries even during the epidemic. However, the delay of cancer screening by women due to fear of coronavirus transmission raises concerns about the delay of cancer diagnosis [8].
Due to the fear of coronavirus as well as social isolation measures, it has not been possible for physicians, nurses and midwives to reach every woman. Cancer screenings have been suspended as many healthcare professionals have been directed to therapeutic services in the fight against coronavirus. For these reasons, the invitation method cannot be applied in screening, and family physicians have to devote more time to the follow-up of those with COVID and those in close contact with them [4, 5].
In the event of a crisis such as the current pandemic, it is necessary for health professionals to continue preventive health services such as cancer screening, to determine cancer risk factors and to screen women at risk for cancer by taking precautions to prevent infection. Necessary health care and counseling should be provided through applications such as tele-health, remote health monitoring and appointment health screenings [9]. It will thereby be of help in the early diagnosis of female cancers and in the development of healthy lifestyle behaviors.
It has been reported that the genetic risk factor, which plays an important role in the development of cancer, can be brought under control by some changes in the lifestyle. Studies show that individuals with genetic susceptibility can reduce their risk of developing cancer if they acquire healthy lifestyle behaviors [10, 11]. Therefore, healthy lifestyle behaviors gain importance in the prevention of cancers. However, given the enormous impact of COVID-19 on society, little is known about the healthy lifestyle behaviors of women under extremely limited circumstances [12]. On the other hand, it is thought that the pandemic may increase the sedentary lifestyle, have negative effects on health-related life quality and have negative effects on such healthy lifestyle behaviors as nutrition and sleep quality. However, in addition to being more comfortable than daily busy work schedules, staying at home with the family members can also have a positive effect on individuals' healthy lifestyle behaviors [13].
Few studies have been conducted to evaluate the effects of cancer screening and healthy lifestyle behaviors in women during the pandemic [6]. The early diagnosis of cancers and the necessary interventions are important in protecting and improving the health of women and therefore that of the society. Therefore, in this study, it was aimed to find out the attitudes toward cancer screening and healthy lifestyle behaviors of women during the coronavirus (COVID-19) pandemic.
Purpose and Questions of the Research
This research was conducted to evaluate the effect of fear of COVID-19 on the attitudes of women (30–65 years old) toward cancer screening and healthy lifestyle behaviors.What are women's attitudes toward cancer screening during the COVID-19 pandemic?
What are women's healthy lifestyle behaviors during the COVID-19 pandemic?
What is the relationship between women's fear of COVID-19, healthy lifestyle behaviors and attitudes toward cancer screening?
Does the fear of COVID-19 affect women's attitudes toward cancer screening?
Does the fear of COVID-19 affect women's healthy lifestyle behaviors?
Methods
Study Design
This descriptive and cross-sectional research used the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist.
Sample
The population of the study consisted of women aged 30 to 65 across the country. No sample selection was made, and the study was conducted with 221 women who volunteered to participate in the study and filled out the data collection tools completely. In the post-study power analysis (PostHoc) conducted to examine the power of the sample, it was discovered that the sample had a 99% power at a 95% confidence interval with a 0.349 effect size (G * Power 3.1.9.7).
Measures
Research data were collected using Introductory Information Form, Attitude Scale for Cancer Screening, The Fear of COVID-19 Scale (FCV-19S) and Healthy Lifestyle Behaviors Scale II (HLBS-II).
Introductory Information Form
This form, prepared by the researchers by scanning the literature [4, 7, 14–16], consists of 13 questions including socio-demographic and cancer screening features.
The Fear of COVID-19 Scale (FCV-19S)
The validity and reliability of the Fear of COVID-19 Scale developed by Ahorsu et al. [16] and its adaptation into Turkish were performed by Bakiroğlu et al. [18] The scale is a self-assessment, 5-point Likert scale and has seven items. The scale yields a score ranging from 7 to 35. Cronbach's alpha value was determined to be 0.88 in the study conducted in the Turkish sample [15]. Cronbach's alpha value of the scale was found to be 0.92 in our study.
Attitude Scale for Cancer Screening
It was developed by Öztürk-Yıldırım et al. [17] to evaluate individuals' attitudes toward cancer screening. The scale consists of 24 items and is one-dimensional. The scale is 5-point Likert type. The items making up the scale are answered in a range from 5 to 1 as "5: I totally agree, 4: I somewhat agree, 3: I neither agree nor disagree, 2: I somewhat disagree, 1: I strongly disagree." The lowest score that can be obtained from the scale is 24 and the highest is 120. It is interpreted that there is a negative attitude toward cancer screenings as the scores of the participants near to 24, and a positive attitude if near to 120. No specific cut-off point was set for the scale. Cronbach's alpha value of the scale was calculated as 0.95 [17]. In this study, the Cronbach alpha value of the scale was calculated as 0.87.
Healthy Lifestyle Behaviors Scale II (HLBS-II)
Based on Pender's Health Promotion Model and developed by Walker et al. it measures health-promoting behaviors associated with an individual's healthy lifestyle. The scale was revised in 1996 and named as HLBS-II [18]. The Turkish validity and reliability study of the scale was conducted by Bahar et al. in 2008 [17]. The scale consists of 52 items, which are all positive, in a 4-point Likert type (1 (Never), 2 (Sometimes), 3 (Often) and 4 (Regularly)). The lowest score is 52, and the highest is 208. An increase in the scores obtained from the scale indicates that the individual applies the stated health behaviors at a high level, and the scale does not have a cut-off value [14]. It consists of six sub-dimensions, which are Spiritual Growth, Nutrition, Physical Activity, Health Responsibility, Interpersonal Relations and Stress Management. The Cronbach's alpha value of the original scale is 0.94 [18]. In the study conducted for the Turkish sample, the Cronbach alpha values of the subscales were found to be 0.94 [14]. In our study, the Cronbach's alpha value of the scale was found to be 0.96.
Procedure
Before applying them to the women, data collection forms had been applied to 10 women —5 of whom in written and 5 via an online interface— who were not included in the sample group, thus giving the forms their final forms. The data of the research were collected through electronic surveys created through Google Forms between March and July 2021. During the pandemic, where direct contact was reduced as much as possible due to the social distance rules, the participants were invited to the research via social media groups (WhatsApp groups, public forums, Twitter and Facebook accounts). All participants were informed about the study at the beginning of the online survey, and their consent was obtained. No names, Internet Protocol (IP) addresses, or other identifying information were collected; thus, participants’ responses were anonymous, and no personal information was attached to the data. All questions had to be completed before submission. Additionally, survey responses from Google Forms are limited to only one response per person. Thus, people were prevented from filling out a questionnaire more than once.
Ethical Dimension of Research
Written approval (Date: 03.02.2021, Decision No: 2021.02.47) was obtained from the Ethics Committee of the relevant university before conducting the research. Written permission (Date: 06.02.2021; Decision No:2021-02-02T16_28_53) was obtained from the Ministry of Health in order to conduct the research. In addition, all participants were informed about the study at the beginning of the online survey, and their consent was obtained. The study was based on the principles of the Declaration of Helsinki.
Evaluation of the Data
Analysis of the data collected was performed using the Statistical Package for Social Science (SPSS) version 25 software program. Number and percentage were used for categorical measurements, and mean and standard deviation were used for numerical measurements. The conformity of the data to normal distribution was determined by the Kolmogorov–Smirnov and Shapiro–Wilk tests according to the sample size. Because all variables in all the scales were normally distributed, Pearson correlation test was used for the correlation analysis. Multiple linear regression analyses were used to define the predictor variables. A P value of < 0.05 was considered statistically significant in all analyses.
Results
The distribution of some descriptive characteristics of the women is given in Table 1. It was found out that 61.5% of the women were between the ages of 30–40, 67.4% had a university degree or higher, 60.6% had a job, 76.9% lived in the metropolis/city and 61.1% had an income level equal to their income and expenses. It was found out that 90.5% of the women did not use alcohol, 76.5% did not smoke, and 72.9% did not have any chronic disease. It was found out that 26.7% of the women had cancer in their first-degree relatives, 53.4% had cancer in their second-degree relatives, 81.4% had knowledge about cancer, 75.1% feared cancer, 45.2% regarded cancer as fatal, 44.8% thought that the treatment is possible.Table 1 The distribution of some descriptive characteristics of women (n = 221)
Descriptive characteristics n %
Age (years)
30–40 136 61.5
41–50 61 27.6
51–65 24 10.9
Age (Mean ± SD) 39.48 ± 8.17
Educational status
Elementary 33 14.9
High school 39 17.6
University and higher 149 67.4
Employment
Employed 134 60.6
Not employed (Housewife) 87 39.4
Location
Metropolis/City 170 76.9
District/Village 51 23.1
Income level
Income is less than expenses 50 22.6
Income is equal 135 61.1
Income is more than expenses 36 16.3
Alcohol use
Yes 21 9.5
No 200 90.5
Smoking
Yes 52 23.5
No 169 76.5
Chronic disease status
Yes 60 27.1
No 161 72.9
Presence of cancer in first degree relatives
Yes 59 26.7
No 162 73.3
Presence of cancer in second degree relatives
Yes 118 53.4
No 103 46.6
Status of knowledge about cancer
Yes 180 81.4
No 41 18.6
Fear of cancer
Yes 166 75.1
No 55 24.9
Thoughts about cancer
Fatal 100 45.2
Not deadly 3 1.4
Not contagious 4 1.8
Possible to treat 99 44.8
Cannot be cured 15 6.8
Table 2 shows the distribution of the characteristics of the women regarding cancer screening. It was found out that 49.3% of the women had never been screened for cancer before, 31.2% of those who had cancer screening had mammography and 18.6% had Pap-smear. It was found out that 92.3% of the women did not have cancer screening during the pandemic period, 33.0% of the women who did not have it because they were afraid of the contamination, 33.0% thought they were healthy, 13.1% did not have screening tests because they thought that screening tests were not easy and accessible during the pandemic.Table 2 Distribution of women's characteristics regarding cancer screening (n = 221)
Characteristics n %
Previous cancer screening
Never done 109 49.3
Have done at least once 112 50.7
Cancer screening (n = 112)
Mammography 69 31.2
Pap-Smear 41 18.6
Colonoscopy 1 0.5
Occult blood in stool 1 0.5
Status of cancer screening in the pandemic period
Yes 17 7.7
No 204 92.3
Reason for not screening for cancer in the pandemic period * (n = 204)
I was afraid of the contamination coronavirus 73 33.0
I thought I was healthy 73 33.0
I didn't know about cancer screenings 27 12.2
Screening tests were not easy and accessible during the pandemic period 29 13.1
I couldn't find the time 24 10.9
I couldn't leave the house due to curfews 21 9.5
I was afraid of getting bad news 9 4.1
I was ashamed of the healthcare worker / having an examination 3 1.4
I thought cancer screening wouldn't help 3 1.4
*More than one option is marked
The mean score of the women’s fear of COVID-19 was 17.99 ± 7.33, the total mean score of the Healthy Lifestyle Behaviors Scale II was 130.46 ± 29.16, the sub-dimension mean scores of HLBS-II was 25.73 ± 5.73 for Spiritual Growth, 21.51 ± 5.06 for Nutrition, 16.56 ± 5.71 for Physical Activity, 22.34 ± 5.88 for Health Responsibility, 25.38 ± 5.81 for Interpersonal Relations and 18.94 ± 5.13 for Stress Management. The mean score of the Attitude Scale for Cancer Screening was calculated as 66.10 ± 15.71.
No significant relationship was found between women's attitudes toward cancer screening and fear of COVID-19 (P > 0.05) (Table 3). A positive significant relationship was found between women's attitudes toward cancer screenings and spiritual growth, health responsibility and interpersonal relations scores, which are sub-dimensions of the HLBS-II scale (P > 0.05) (Table 4).Table 3 Distribution of total mean scores of women's Fear of COVID-19 Scale (FCV-19S), Healthy Lifestyle Behaviors Scale II (HLBS-II) and its sub-dimensions, and Attitude Scale for Cancer Screening (ASCS)
Scales Mean ± SD Min Max
Fear of COVID-19 Scale (FCV-19S)
FCV-19S 17.99 ± 7.33 7 35
Healthy Lifestyle Behaviors Scale II (HLBS-II)
Spiritual growth 25.73 ± 5.73 9 36
Nutrition 21.51 ± 5.06 9 36
Physical activity 16.56 ± 5.71 8 32
Health responsibility 22.34 ± 5.88 9 36
Interpersonal relations 25.38 ± 5.81 9 36
Stress management 18.94 ± 5.13 8 32
HLBS-II (Total) 130.46 ± 29.16 52 208
Attitude Scale for Cancer Screening (ASCS)
ASCS (Total) 66.10 ± 15.71 24 107
Table 4 The relationship between women's fear of COVID-19, healthy lifestyle behaviors, and attitudes toward cancer screening
Attitude Scale for Cancer Screening (ASCS)
r P
Fear of COVID-19 Scale (FCV-19S) 0.122 0.071
Healthy Lifestyle Behaviors Scale II (HLBS-II)
Spiritual growth 0.238 < 0.001*
Nutrition 0.058 0.393
Physical activity − 0.074 0.272
Health responsibility 0.134 0.047*
Interpersonal relations 0.228 0.001*
Stress management 0.061 0.366
HLBS-II (Total) 0.125 0.063
* P < 0.05, r = Pearson correlation coefficient
Multiple linear regression results to examine the effect of fear of COVID-19 on women's attitudes toward cancer screening and healthy lifestyle behaviors are given in Table 5. The established model was found to be statistically significant (P < 0.05). According to the results of the research, it was determined that the fear of COVID-19 did not have an effect on the attitude toward cancer screening, but it affected the interpersonal relationships and stress management of the women (P < 0.05).Table 5 The effect of women's fear of COVID-19 on their attitudes toward cancer screening and healthy lifestyle behaviors
Independent variable β SE t P value F model P model R2
ASCS 0.087 0.032 1.253 0.212
HLBS-II sub-dimensions
Spiritual growth − 0.248 0.198 − 1.597 0.112
Nutrition 0.179 0.160 1.619 0.107 3.889 0.001* 0.113
Physical activity − 0.134 0.144 − 1.194 0.234
Health responsibility 0.149 0.168 1.105 0.270
Interpersonal relations 0.318 0.185 2.163 0.032*
Stress management − 0.306 0.201 − 2.180 0.030*
ASCS Attitude Scale for Cancer Screening, HLBS-II Healthy Lifestyle Behaviors Scale II, *P < 0.05
Discussion
Cancer is an important public health problem today, and it maintains its current importance. Cancer, which ranks second after cardiovascular diseases in the list of known deaths, causes heavy losses in the workforce and the country's economy due to the disabilities it causes and the high costs of its treatment. Raising awareness about cancer, improving public awareness and cancer screening are among the most effective methods in the fight against cancer. Breast cancer, cervical cancer and colorectal cancer screenings are carried out free of charge by the Ministry of Health within the scope of the National Cancer Screening Program in Turkey. Within the scope of this screening program, breast cancer screening with mammography every two years on women aged 40–69, cervical cancer screening on women aged 30–65 with Papanicolaou smear and HPV test every five years, colorectal cancer screening with a stool occult blood test on women and men aged 50–70 every two years performed [19]. According to the results of the Turkey Household Health Survey conducted in 2017, it was found out that 46.4% of women are aware of the existence of cancer screening tests, two out of five women aged 40 to 69 have never had a mammogram, and nearly half of the women aged 30–65 have never had any cervical cancer screening [20]. COVID-19 pandemic, as in all health services, affected these numbers, which were not good at all, and all population-based screening programs around the world came to a standstill [21]. Many governments have suspended cancer screenings and called for not applying to health institutions except in emergencies. According to the World Health Organization (WHO) data, about half of the 155 countries have postponed their cancer screening programs and reported disruptions in cancer treatment services [22]. In our study, it was found out that 50.7% of women had cancer screening at least once in their lives, and 92.3% of women did not have cancer screening during the pandemic period. In the study of Erdoğan and Akkaya [23], it was found out that due to COVID-19 pandemic, HPV scanning decreased significantly in 2020 compared to previous years. The studies conducted by Peacock et al. [24] in Belgium, by Jacob et al. [25] in Germany and by Skovlund et al. [26] in Denmark all shown that cancer diagnosis in the pandemic has decreased significantly compared to the pre-pandemic period. The results of our study are similar to those around the world. Our study is important in terms of showing the dramatic and negative impact of the restrictions stemming from COVID-19 pandemic on cancer screening rates in the Turkish sample as well.
Factors causing the decrease in cancer screening rates have been reported as the need to direct health personnel and resources to the fight against the pandemic, reduced availability of public transport, the lack of personnel, medicine, diagnosis and technology [22]. In addition, many patients feared exposure to SARS-CoV-2 or overburdening their healthcare and therefore were less likely to seek healthcare for cancer screening and diagnosis. In our study, too, approximately one-third of the women stated that they did not have cancer screening because they were afraid of contracting the Coronavirus. Similar to the results of our study, De Pelsemaeker et al. [27] stated that histological and cytological examinations of colon biopsies, breast biopsies and cervical cytology decreased significantly due to fear of COVID-19, while Cheng et al. [28] stated that half of colonoscopy cancelations during the pandemic were due to fear of infection. Individuals with potential non-specific cancer symptoms faced barriers to consulting a specialist [29], largely due to fear and anxiety about getting infected with COVID-19 in a healthcare setting. Both patients and staff at the hospitals experience fear and anxiety [30]. This fear and anxiety affect women's participation in cancer screening programs and prevents early diagnosis of cancer [31]. Cancer screenings are an opportunity to detect precancerous lesions early and to initiate interventions that prevent or delay disease progression. Failure to detect early symptoms of cancer causes cancers to be diagnosed in later stages [32]. Yong et al. [33] predict that about 5300 additional deaths from breast cancer and 4500 from colorectal cancer will occur in Canada due to delays in cancer diagnosis. Sud et al. [34] stated that a 3- to 6-month delay in cancer screening and surgery in the UK has a significant impact on survival, and a delay in diagnosis and treatment will reduce individuals' life expectancy by 19 to 43%. On the other hand, suspension of cancer screening or cancer prevention programs is expected to aggravate patients' suffering, disease burden, 5-year mortality, economic burden and workload for surgeons and oncologists [35]. Although cancer screening programs in Turkey have not been halted during the pandemic, the result of our study shows that women were hesitant to participate in cancer screening programs. From this perspective, we are of the opinion that raising women's awareness of cancer prevention during the ongoing pandemic is critically important and that health care professionals should refer women to cancer screenings during the pandemic.
Attitudes, defined as a state of emotional and mental readiness that is formed as a result of experiences and has a directive effect on the behavior of the individual against all relevant situations [36], have an important place in cancer screenings [17]. In our study, the mean score of Attitude Scale for Cancer Screening was found to be 66.10 ± 15.71. According to this result, considering that the highest score on the scale is 120, we can say that the attitude of women toward cancer screening is moderate. Attitude has an important power directing the behavior of the individual, and strong attitudes are more reflected in behavior than the weak ones [36]. Therefore, strengthening women's attitudes toward cancer screenings can help women to go to screening regularly.
Mammography and PAP smear test, which are early diagnosis and screening methods in breast and cervical cancer, are among the healthy lifestyle behaviors. Healthy lifestyle behaviors of women affect the knowledge of early diagnosis and the practices in cancer [37]. In our study, the mean score of the Healthy Lifestyle Behaviors Scale II was found to be 130.46 ± 29.16. In a study conducted in a Turkish sample before the pandemic, it was found to be 142.73 ± 26.3 [38], and 126.8 ± 19.2 [39] in another study. According to these results, we can say that the healthy lifestyle behavior levels of women in the pre-pandemic and pandemic period are similar.
In our study, a positive and significant relationship was found between such healthy lifestyle behaviors as spiritual growth, health responsibility, interpersonal relationship and attitudes toward cancer screening. Gözüyeşil et al. [38] found a statistically significant difference between knowing about and performing Breast Self-Exam (BSE) and health responsibility and interpersonal relations, which are two sub-dimensions of the HLBS-II. In the study conducted by Gök-Uğur and Aydın-Avcı [37], a statistically significant difference was found between the status of performing BSE, mammography and smear test and spiritual growth and health responsibility. Im Kim et al. [40] reported that women who regularly practice breast cancer screening tend to exhibit better healthy lifestyle behaviors. It can be seen that our study results are compatible with the data obtained from other studies. These results show that healthy lifestyle behaviors affect early diagnosis practices in women.
The most important nationwide goal during COVID-19 pandemic has been to reduce the spread of the virus. For this purpose, serious restrictions were placed on individual freedom and people were called to stay at home. In this process, the unavoidable increases in the number of positive cases and loss of life led to fear, which is a psychological aspect of COVID-19 [41]. While fear of COVID-19 triggered psychological distress, it also helped to encourage a reduction in risky behaviors [42]. In our study, it was found out that while women's fear of COVID-19 did not affect their attitudes toward cancer screening, healthy lifestyle behaviors affected interpersonal relationships and stress management. Ayandele et al. [42] reported that high levels of fear of COVID-19 are more likely to trigger participation in frequent preventive health behaviors. Our results are consistent with the previous studies [42]. Fear can be effective in triggering certain problem-solving or problem-avoidance behaviors, which can prevent the feared event or situation from happening. In addition to increasing people's alertness to the seriousness of risks, fear may also have a feature that may prevent preventive behaviors [42]. Therefore, it may be wrong to assume that fear will lead individuals to healthy lifestyle behaviors in the pandemic. In this process, individuals should be encouraged to participate in cancer screening programs and healthy lifestyle behaviors by providing them with sufficient information.
Limitations
The current study has some limitations. An internet-based (online) questionnaire was used in the study. This can lead to selection bias and poor generalization. In addition, since the data obtained from the study are cross-sectional, they do not provide long-term results. Individuals may also respond differently to the questionnaire depending on the stage of COVID-19 pandemic. Therefore, in our study, data were collected in a short time to minimize differences and changes in restrictions due to COVID-19. Despite these limitations, it provides important information about the attitudes of women toward cancer screening during the pandemic.
Conclusion
In our study, it was found out that most of the women did not have cancer screening during the pandemic, and that there was a positive and significant relationship between women's attitudes toward cancer screening and spiritual growth, health responsibility, interpersonal relationships, and that fear of COVID-19 affected interpersonal relationships and stress management. According to the results of the research, in order to protect, maintain and improve women's health during the pandemic, it is important to increase the awareness of women about common cancers, especially in primary healthcare centers, to question health behaviors and the variables affecting health behaviors and to prepare, conduct and maintain health education programs regarding all these.
Acknowledgements
The authors wish to thank all the participants to participate in this study.
Authors’ contributions
CP was involved in data curation (equal), formal analysis (equal), investigation (equal), methodology (equal), resources (equal), software (equal), visualization (equal), writing—original draft (equal) and writing-review and editing (equal); ÖPU contributed to conceptualization (equal), methodology (equal), project administration (equal), resources (equal), software (equal), visualization (equal), writing-original draft (equal) and writing—review and editing (equal).
Funding
This study has not received financial support from any official or private institution.
Declarations
Conflict of interest
The authors declared no disclosures or potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report. Published 2020. Accessed March 2, 2022.
2. Wu JT Leung K Leung GM Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study Lancet 2020 395 10225 689 697 10.1016/S0140-6736(20)30260-9 32014114
3. Alpago H Alpago DO Koronavirüs ve sosyoekonomik Sonuçlar IBAD Sosyal Bilimler Dergisi. 2020 8 99 114 10.21733/ibad.716444
4. Suzuki S Psychological status of postpartum women under the COVID-19 pandemic in Japan J Matern Fetal Neonatal Med 2022 35 9 1798 1800 10.1080/14767058.2020.1763949 32423262
5. Chen H Selix N Nosek M Perinatal anxiety and depression during COVID-19 J Nurse Pract 2021 17 1 26 31 10.1016/j.nurpra.2020.09.014 33013242
6. D'Ovidio V Lucidi C Bruno G Lisi D Miglioresi L Bazuro ME Impact of COVID-19 pandemic on colorectal cancer screening program Clin Colorectal Cancer 2021 20 1 e5 e11 10.1016/j.clcc.2020.07.006 32868231
7. London JW Fazio-Eynullayeva E Palchuk MB Sankey P McNair C Effects of the COVID-19 pandemic on cancer-related patient encounters JCO Clin Cancer Inform 2020 4 657 665 10.1200/CCI.20.00068 32716647
8. Xia Y Jin R Zhao J Li W Shen H Risk of COVID-19 for patients with cancer Lancet Oncol 2020 21 4 e180 10.1016/S1470-2045(20)30150-9 32142622
9. Chrzan-Dętkoś M Walczak-Kozłowska T Lipowska M The need for additional mental health support for women in the postpartum period in the times of epidemic crisis BMC Pregnancy Childbirth 2021 21 1 114 10.1186/s12884-021-03544-8 33557768
10. Ramzi NH Chahil JK Lye SH Role of genetic & environment risk factors in the aetiology of colorectal cancer in Malaysia Indian J Med Res 2014 139 6 873 882 25109722
11. Aran V Victorino AP Thuler LC Ferreira CG Colorectal cancer: epidemiology, disease mechanisms and interventions to reduce onset and mortality Clin Colorectal Cancer 2016 15 3 195 203 10.1016/j.clcc.2016.02.008 26964802
12. Mayasari NR Ho DKN Lundy DJ Impacts of the COVID-19 pandemic on food security and diet-related lifestyle behaviors: an analytical study of google trends-based query volumes Nutrients 2020 12 10 3103 10.3390/nu12103103 33053656
13. Wang X Lei SM Le S Bidirectional influence of the COVID-19 pandemic lockdowns on health behaviors and quality of life among Chinese adults Int J Environ Res Public Health 2020 17 15 5575 10.3390/ijerph17155575 32748825
14. Bahar Z Beşer A Gördes N Ersin F Kıssal A Sağlıklıyaşam biçimi davranışları ölçeği II'nin geçerlik ve güvenirlik çalışması Cumhur Üniversitesi Hemşirelik Yüksekokulu Dergisi 2008 12 1 13
15. Bakioğlu F Korkmaz O Ercan H Fear of COVID-19 and positivity: mediating role of intolerance of uncertainty, depression, anxiety, and stress Int J Ment Health Addict 2021 19 6 2369 2382 10.1007/s11469-020-00331-y 32837421
16. Ahorsu DK Lin CY Imani V Saffari M Griffiths MD Pakpour AH The Fear of COVID-19 Scale: development and initial validation [published online ahead of print, 2020 Mar 27] Int J Ment Health Addict. 2020 10.1007/s11469-020-00270-8 32837427
17. Öztürk-Yıldırım EN Uyar M Şahin TK Development of an attitude scale for cancer screening Turk J Oncol 2020 35 4 394 404 10.5505/tjo.2020.2341
18. Walker SN, Hill-Polerecky DM. Psychometric evaluation of the health-promoting lifestyle profile II. Unpublished manuscript, University of Nebraska Medical Center. 1996;13:120–6.
19. Republic of Turkey Ministry of Health. Cancer Screenings. Published 2018. Accessed March 20, 2022. https://hsgm.saglik.gov.tr/tr/kanser-taramalari
20. World Health Organization. National household health survey in Turkey: prevalence of noncommunicable disease risk factors 2017. Published 2018. Accessed March 20, 2022. https://hsgm.saglik.gov.tr/depo/birimler/kronik-hastaliklar-engelli-db/hastaliklar/kalpvedamar/raporlar/WHO-Turkey-Risk-Factors-A4_ENG.08_10_2018.pdf
21. Lozar T Nagvekar R Rohrer C Dube Mandishora RS Ivanus U Fitzpatrick MB Cervical cancer screening postpandemic: self-sampling opportunities to accelerate the elimination of cervical cancer Int J Womens Health 2021 13 841 859 10.2147/IJWH.S288376 34566436
22. World Health Organization news release, COVID-19 significantly impacts health services for non communicable diseases. Published June 2020. Accessed March 19, 2022. https://www.who.int/news/item/01-06-2020-covid-19-significantly-impacts-health-services-for-noncommunicable-diseases
23. Erdoğan P Akkaya F Ulusal Primer HPV Tarama Programının Niğde İlinde Mevsimsellik ve Demografik Eğilimleri: COVID-19 Pandemisinin Etkisi Turk J Public Health 2022 20 1 152 163
24. Peacock HM Tambuyzer T Verdoodt F Decline and incomplete recovery in cancer diagnoses during the COVID-19 pandemic in Belgium: a year-long, population-level analysis ESMO Open. 2021 6 4 100197 10.1016/j.esmoop.2021.100197 34474811
25. Jacob L Loosen SH Kalder M Luedde T Roderburg C Kostev K Impact of the COVID-19 pandemic on cancer diagnoses in general and specialized practices in Germany Cancers (Basel) 2021 13 3 408 10.3390/cancers13030408 33499223
26. Skovlund CW Friis S Dehlendorff C Nilbert MC Mørch LS Hidden morbidities: drop in cancer diagnoses during the COVID-19 pandemic in Denmark Acta Oncol 2021 60 1 20 23 10.1080/0284186X.2020.1858235 33287604
27. de Pelsemaeker MC Guiot Y Vanderveken J Galant C Van Bockstal MR The impact of the COVID-19 pandemic and the associated Belgian Governmental measures on cancer screening Surg Pathol Cytopathol Pathobiol 2021 88 1 46 55 10.1159/000509546
28. Cheng SY Chen CF He HC Impact of COVID-19 pandemic on fecal immunochemical test screening uptake and compliance to diagnostic colonoscopy J Gastroenterol Hepatol 2021 36 6 1614 1619 10.1111/jgh.15325 33217055
29. Dinmohamed AG Visser O Verhoeven RHA Fewer cancer diagnoses during the COVID-19 epidemic in the Netherlands Lancet Oncol 2020 21 6 750 751 10.1016/S1470-2045(20)30265-5 32359403
30. Gong K Xu Z Cai Z Chen Y Wang Z Internet hospitals help prevent and control the epidemic of COVID-19 in China: multicenter user profiling study J Med Internet Res 2020 22 4 e18908 10.2196/18908 32250962
31. Duong TL Lee N Kim Y Kim Y Assessment of the fear of COVID-19 and its impact on lung cancer screening participation among the Korean general population Transl Lung Cancer Res. 2021 10 12 4403 4413 10.21037/tlcr-21-746 35070750
32. Alkatout I Biebl M Momenimovahed Z Has COVID-19 affected cancer screening programs? A Systematic Review Front Oncol 2021 11 675038 10.3389/fonc.2021.675038 34079764
33. Yong JH Mainprize JG Yaffe MJ The impact of episodic screening interruption: COVID-19 and population-based cancer screening in Canada J Med Screen 2021 28 2 100 107 10.1177/0969141320974711 33241760
34. Sud A Jones ME Broggio J Collateral damage: the impact on outcomes from cancer surgery of the COVID-19 pandemic Ann Oncol 2020 31 8 1065 1074 10.1016/j.annonc.2020.05.009 32442581
35. Del Vecchio BG Calabrese E Biancone L Monteleone G Paoluzi OA The impact of COVID-19 pandemic in the colorectal cancer prevention Int J Colorectal Dis 2020 35 10 1951 1954 10.1007/s00384-020-03635-6 32500432
36. Yılmazer A Eroğlu C Meslek Yüksekokulları için Davranış Bilimleri ve Örgütsel Davranış 2018 Ankara Detay Yayıncılık
37. Gök-Uğur H Aydın-Avcı İ Kanser tarama merkezine başvuran kadınların sağlıklı yaşam biçimi davranışlarının erken tanı bilgi ve uygulamalarına etkisi Gümüşhane Üniversitesi Sağlık Bilimleri Dergisi 2015 4 2 244 260
38. Gözüyeşil E Filiz TAŞ Düzgün AA 15–49 yaş arası kadınlarda meme kanseri endişesi ve sağlıklı yaşam biçimi davranışlarını etkileyen faktörler Cukurova Med J 2019 44 4 1215 1225 10.17826/cumj.493360
39. Sonmezer H Cetinkaya F Nacar M Healthy life-style promoting behaviour in Turkish women aged 18–64 Asian Pac J Cancer Prev 2012 13 4 1241 1245 10.7314/apjcp.2012.13.4.1241 22799312
40. Kim JI Oh KO Li CY Min HS Chang ES Song R Breast cancer screening practice and health-promoting behavior among chinese women Asian Nurs Res (Korean Soc Nurs Sci) 2011 5 3 157 163 10.1016/j.anr.2011.09.005 25030364
41. Pakpour AH Griffiths MD The fear of COVID-19 and its role in preventive behaviors J Concurr Disord 2020 2 1 58 63
42. Ayandele O Ramos-Vera CA Iorfa SK Chovwen CO Olapegba PO Exploring the complex pathways between the fear of COVID-19 and preventive health behavior among Nigerians: mediation and moderation analyses Am J Trop Med Hyg 2021 105 3 701 707 10.4269/ajtmh.20-0994 34237014
|
PMC010xxxxxx/PMC10186567.txt |
==== Front
J Adolesc Health
J Adolesc Health
The Journal of Adolescent Health
1054-139X
1879-1972
Society for Adolescent Health and Medicine. Published by Elsevier Inc.
S1054-139X(23)00150-7
10.1016/j.jadohealth.2023.02.040
Original Article
Pandemic-Related Changes in the Prevalence of Early Adolescent Alcohol and Drug Use, 2020–2021: Data From a Multisite Cohort Study
Pelham William E. III Ph.D. a∗
Tapert Susan F. Ph.D. a
Zúñiga María Luisa Ph.D. b
Thompson Wesley K. Ph.D. c
Wade Natasha E. Ph.D. a
Gonzalez Marybel R. Ph.D. a
Patel Herry Ph.D. d
Baker Fiona C. Ph.D. e
Dowling Gayathri J. Ph.D. f
Van Rinsveld Amandine M. Ph.D. g
Baskin-Sommers Arielle Ph.D. h
Kiss Orsolya Ph.D. e
Brown Sandra A. Ph.D. ia
a Department of Psychiatry, University of California San Diego, La Jolla, California
b School of Social Work, College of Health and Human Services, San Diego State University, San Diego, California
c Laureate Institute for Brain Research, Tulsa, Oklahoma
d Department of Psychiatry and Behavioural Neuroscience, McMaster University, Hamilton, Ontario, Canada
e Center for Health Sciences, SRI International, Menlo Park, California
f National Institute on Drug Abuse, Rockville, Maryland
g Graduate School of Education, Stanford University, Palo Alto, California
h Department of Psychology, Yale University, New Haven, Connecticut
i Department of Psychology, University of California San Diego, La Jolla, California
∗ Address correspondence to: William Pelham, Ph.D., Department of Psychiatry, University of California, San Diego, 9500 Gilman Dr, La Jolla, CA 92093.
16 5 2023
16 5 2023
17 6 2022
26 2 2023
© 2023 Society for Adolescent Health and Medicine. Published by Elsevier Inc. All rights reserved.
2023
Society for Adolescent Health and Medicine
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Purpose
Evaluate changes in early adolescent substance use from May 2020 to May 2021 during the coronavirus disease 2019 pandemic using data from a prospective nationwide cohort: the Adolescent Brain Cognitive Development Study.
Methods
In 2018–2019, 9,270 youth aged 11.5–13.0 completed a prepandemic assessment of past-month alcohol and drug use, then up to seven during-pandemic assessments between May 2020 and May 2021. We compared the prevalence of substance use among same-age youth across these eight timepoints.
Results
Pandemic-related decreases in the past-month prevalence of alcohol use were detectable in May 2020, grew larger over time, and remained substantial in May 2021 (0.3% vs. 3.2% prepandemic, p <.001). Pandemic-related increases in inhalant use (p = .04) and prescription drug misuse (p < .001) were detectable in May 2020, shrunk over time, and were smaller but still detectable in May 2021(0.1%-0.2% vs. 0% pre-pandemic). Pandemic-related increases in nicotine use were detectable between May 2020 and March 2021 and no longer significantly different from prepandemic levels in May 2021 (0.5% vs. 0.2% prepandemic, p = .09). There was significant heterogeneity in pandemic-related change in substance use at some timepoints, with increased rates among youth identified as Black or Hispanic or in lower-income families versus stable or decreased rates among youth identified as White or in higher-income families.
Discussion
Among youth ages 11.5–13.0 years old, rates of alcohol use remained dramatically reduced in May 2021 relative to prepandemic and rates of prescription drug misuse and inhalant use remained modestly increased. Differences remained despite the partial restoration of prepandemic life, raising questions about whether youth who spent early adolescence under pandemic conditions may exhibit persistently different patterns of substance use.
Keywords
Adolescence
COVID-19
Alcohol
Cannabis
Nicotine
Drugs
==== Body
pmc Implications and Contribution
Compared to prepandemic, in May 2021, fewer teens in early adolescence used alcohol and more used inhalants or misused prescription drugs. Data indicated persistent and heterogeneous pandemic-related changes in early adolescent substance use, with adverse impacts being largest among Black, Hispanic, or low-income youth.
A handful of studies have examined changes in alcohol and drug use among adolescents during the coronavirus disease 2019 (COVID-19) pandemic, finding unchanged [1] or decreased [[2], [3], [4], [5]] prevalence of alcohol, binge drinking, cannabis use, cigarette, and e-cigarette use during the pandemic. Other studies have found an increased frequency of hospital visits for substance use disorders [6] and deaths from drug overdoses [7] among adolescents. Today, three gaps in the emerging evidence must be filled to guide an effective public health response. The first gap is the lack of extended follow-up and limited temporal resolution when identifying pandemic-related changes, despite the pandemic's evolving nature [8]. Almost all published studies have reported on the pandemic's impact in its first several months, during 2020, leaving the pandemic's subsequent impact unclear. Moreover, all published studies have examined changes at a single timepoint during the pandemic. Documenting how pandemic-related changes in substance use unfolded across different phases of the pandemic and into 2021 could inform expectations of whether changes will persist or remit as the pandemic continues.
The second gap is a limited focus on early adolescence [8], spanning ages 10–13 years. A study of 11–12-year-old youth in May 2020 found that, compared to prepandemic, fewer were using alcohol and more were using nicotine or misusing prescription drugs [9]. The finding of increased use of some drugs contrasts with the evidence reviewed above for older adolescents, perhaps indicating a differential impact of the pandemic in this age range.
The third gap is limited investigation of how the pandemic's effect varies across the population [8]. Two studies have failed to find significant differences by sex [2,3] while another study found larger decreases in substance use among males [5]. Studies of Norwegian [5] and Icelandic [3] teens ages 13–18 years old found larger reductions in alcohol and nicotine product use among older adolescents [3]. No study has investigated potential differences in substance use among racial/ethnic or sexual orientation minority youth. Racial/ethnic minority groups have suffered disproportionate disease burden [10], economic hardship [11], and other stressors [12], and sexual orientation minority groups have suffered disproportionate psychiatric distress and barriers to care [13]—these disproportionate burdens may contribute to disproportionate changes in substance use [14]. Likewise, despite evidence of larger adverse impacts of the pandemic among economically vulnerable populations [15], no study has investigated whether changes in adolescent substance use vary by household income.
The current study evaluated pandemic-related changes in the past-month prevalence of alcohol and drug use using data from a cohort of United States youth serially assessed at seven timepoints during the COVID-19 pandemic: the Adolescent Brain and Cognitive Development (ABCD) Study. We extend a previous analysis [9] of data at a single timepoint in May 2020 to incorporate seven timepoints spanning through May 2021. We hypothesized that previously documented pandemic-related changes in the prevalence of alcohol and drug use in the ABCD Study sample would persist to May 2021. Persistence of changes was found in a previous study of eighth graders spanning 2020–2021 [4]. We hypothesized that any adverse impacts of the pandemic on adolescent substance use would be greater among youth identifying as a racial/ethnic minority, as a sexual orientation minority, and/or living in households with lower income. This hypothesis followed from empirical evidence that these groups have experienced more stressors and barriers to care during the pandemic, which in turn could lead to more substance use [16,17].
Methods
Sample
From 2016–2018, the ABCD Study [18] recruited 11,880 youth aged 9–10 at 21 study sites across the United States. Recruitment occurred primarily through schools, and the sample was intended to reflect the sociodemographics of the United States [19]. At study entry, 48% of youth were female; 52% identified as White, 20% as Hispanic, 15% as Black, 2% as Asian, and 11% as another racial/ethnic identity. 68% of parents were married. Both parents were in the labor force in 49% of families, and no parent was in the labor force in 6% of families. 59% of youth had ≥1 parent with a bachelor's degree. 57% of families had an annual household income above $75,000. The mean household size was 4.7 people. Participants have been followed prospectively since initial recruitment with annual assessments. The most recently completed and publicly released assessment wave was the 2-year follow-up assessment, at which 88% of participants had been retained. To date, 127 participants have withdrawn from the study [20]. All procedures were conducted in accordance with the ethical standards of the 1964 Helsinki Declaration and its later amendments and were approved by an institutional review board.
Longitudinal design and measurement of substance use
Our analyses combine prepandemic data from the ongoing main ABCD Study protocol with during-pandemic data from a rapidly implemented pandemic-focused survey protocol that began in May 2020. Figure 1 depicts how data from each protocol was combined to form the analytic sample.Figure 1 Selection of observations for analysis from within ongoing main ABCD study longitudinal design and subsequently established pandemic-focused survey protocol. Note. Depicts how the observations included in the analysis at each timepoint (bottom row of eight boxes) were derived from the ABCD Study protocols [21].
During-pandemic assessments
Beginning in May 2020, all ABCD Study participants were invited to complete up to seven web-based surveys measuring the impact of the pandemic on them and their families. 38%–48% of eligible participants completed each survey wave. Survey waves were spaced 5–11 weeks apart: wave 1 (May 16, 2020), wave 2 (June 23, 2020), wave 3 (August 4, 2020), wave 4 (October 8, 2020), wave 5 (December 13, 2020), wave 6 (March 2, 2021), and wave 7 (May 17, 2021). Table 1 provides information on the national state of the pandemic at the time of each survey wave (e.g., case rates, percentage of families who were socially distancing).Table 1 Context of the COVID-19 pandemic at each survey wave
Variable Survey wave
May
2020 June
2020 Aug.
2020 Oct.
2020 Dec.
2020 Mar.
2021 May
2021
Number of observations 3,138 3,277 2,952 2,539 2,589 1,841 1,493
Mean age of participants (years) with included data 12.3 12.3 12.3 12.4 12.4 12.4 12.5
Date of initiation of survey dissemination May 16 July 23 Aug. 4 Oct. 8 Dec. 13 Mar. 2 May 17
ABCD data
Percent of youth reporting full-time in-person schoolingab 0.5% 0.7% 1.9% 19.1% 17.6% 29.2% 44.1%
Percent of families who engaged in social distancingb 85% 78% 80% 79% 79% 75% 61%
Percent of families who avoided visiting family or friendsb 59% 43% 45% 42% 60% 44% 23%
Geocoded ABCD datac
Case rates in participants' counties (per 100,000)d 6.1 12.2 18.1 14.7 62.2 - -
Death rates in participants' counties (per 100,000)d 0.37 0.13 0.27 0.18 0.68 - -
Unemployment rates in participants' countiese 12.4% 10.4% 7.9% 6.0% 6.0% - -
U.S. National data
Number of new cases (7-day rolling average)f 24,301 22,058 57,972 46,939 212,859 63,506 30,935
Number of new deaths (7-day rolling average)f 1,418 765 1,144 682 2,616 1,740 548
Percent of persons age ≥ 18 years old who have completed a vaccination seriesg - - - - - - 47.3%
Percentage of persons ages 12–17 years old who have completed a vaccination seriesg - - - - -- 0.1% 7.4%
Percent of employed adults with children working exclusively in-persong 61.1% 65.3% 73.0% 74.4% 74.1% 77.3% 82.2%
These data are provided to give descriptive context about the national state of the pandemic at the time of each survey wave—these data were not analyzed in this manuscript. Number of observations is the number of participants at each survey wave contributing data to regression models. Regression models adjusted for the mean differences in age of participants across survey waves (see Methods). Hyphens in table cell indicate that data was not available. ABCD data and Geocoded ABCD data were weighted to be sociodemographically representative of children in the United States Census (see Methods).
a Reported by youth on the web-based surveys during the COVID-19 pandemic. Values indicate proportion of youth who selected “in-person” as their response to a multiple choice survey item asking, “In the past week, [how] was your schooling taking place?” At May, June, or August 2020 surveys, youth may have been on summer vacation from school.
b Reported by parents on the web-based surveys during the COVID-19 pandemic. Values indicate proportion of parents who checked boxes indicating whether in the past week their family had “engaged in social distancing” or “avoided visiting family or friends outside our immediate family.”
c Geocoded ABCD data have only been publicly released for survey waves 1–5.
d Source: Johns Hopkins Coronavirus Resource Center COVID-19 Data Dashboard [22]. Data at the level of county and day.
e Source: Bureau of Labor Statistics (BLS) Local Area Unemployment Statistics (LAUS) [23]. Data at the level of county and month.
f Source: Center for Disease Control (CDC) COVID Data Tracker [24]. Data at the level of nation and day. The CDC COVID Data Tracker does not include adult vaccination data for all states prior to December 14, 2020.
g Source: Bureau of Labor Statistics (BLS) Current Population Survey (CPS) [25]. Data at the level of nation and month.
At each survey wave, youth reported the number of days in the past month on which they: (1) drank alcohol; (2) smoked cigarettes; (3) used an electronic nicotine delivery system; (4) smoked a cigar/hookah/pipe; (5) used smokeless tobacco/chew/snus; (6) used a cannabis product (flower/concentrate/edible); (7) used prescription drugs in a way not prescribed; or (8) used inhalants. Items were modeled on the prepandemic ABCD Study assessments [26] and the Monitoring the Future Study 2020 interview [27]. The response scale ranged from 0 days to 10 + days; responses were dichotomized into no use versus any use to match the response scale at the assessments completed prepandemic. Responses were collapsed across items (1)–(8) to form an indicator of use of any substance. Responses were collapsed across items (2)–(5) to form a single indicator of nicotine use. Dependent variables included use of each substance category as well as of any substance.
Prepandemic assessment
Youth had been followed for 2–4 years prepandemic. For comparison to during-pandemic assessments, we drew data from an assessment wave (18-month follow-up [26]) that measured youth substance use on the same timescale (past-month use), with comparable item wording, and at an age range overlapping with that assessed at each during-pandemic timepoint. Between February 2018 and March 2019, when the youth were 11–12 years old, all had been invited to complete a phone interview at which they reported on their past-month alcohol and drug use. Youth verified they were in a private setting before the interview began. Items were modeled on the Monitoring the Future Study's [4] questions about monthly use, with updates to wording to address changes in nicotine and cannabis products (e.g., advent of vaping). 93% of participants completed the interview.
Comparisons of longitudinal data
Developmental increases in drinking and drug use are expected during early adolescence. Thus, maturation and pandemic effects will be confounded in longitudinal data: an apparent increase in substance use at measurements before versus during the pandemic, or in earlier versus later phases of the pandemic, could be explained by the maturation of the sample. To control for maturation effects, we used an age-period design [28] that we have used to study pandemic effects on substance use in two previously published studies [9,29]. The age-period design leveraged the fact that the ABCD Study participants span a 4-year range of ages on any given calendar date, given that recruitment spanned 2016 to 2018, and all youth were 9–10 years old at study entry. Given a mixed-age cohort, we can compare the substance use of participants that reach the same age on different calendar dates, either before or during the pandemic (e.g., a 12-year-old assessed in 2019 vs. a 12-year-old assessed in 2021). If we compare same-age youth across timepoints, maturation can no longer explain any differences in the rates of substance use across timepoints. First, we restricted the data as necessary to have observations of youth across a similar age range at every timepoint, spanning 11.5–13.0 years old, ensuring that we could properly adjust for maturation (Figure 1 shows the number of observations excluded at each timepoint) [30]. Next, we adjusted for age-at-observation in all analyses to account for maturation effects.
There were eight timepoints—the prepandemic visit plus the seven during-pandemic web-based surveys—with the following number of participants at each (Figure 1): prepandemic (n = 4,988), May 2020 (n = 3,138), June 2020 (n = 3,277), August 2020 (n = 2,952), October 2020 (n = 2,539), December 2020 (n = 2,059), March 2021 (n = 1,841), May 2021 (n = 1,493). The number of participants decreases at later timepoints because a greater proportion of the sample had aged beyond the age range that could be compared to the prepandemic observations (i.e., is older than 13.0 years). At least one observation was contributed by 9,270 youths, with a mean of 2.4 observations per youth (standard deviation = 1.8, range = [1,8]). Both the prepandemic observations and ≥1 during-pandemic observations were contributed by 1,457 youths. In each during-pandemic wave, 57%–69% of participants were present in the subsequent wave. Table A1 compares the sociodemographic characteristics of participants in the analytic sample at each timepoint. Small differences (<5 percentage points) between timepoints in youth racial/ethnic identity, household income, parent education, and parent marital status of participants were accounted for during analysis via weighting (see Analytic plan below).
Measurement of putative moderators of pandemic-related change in substance use
We tested four putative moderators: youth sex, youth identification as a sexual orientation minority, youth racial/ethnic identification, and household income prepandemic. Youth's sex assigned at birth and racial/ethnic identification were reported by caregivers at study entry. Caregivers answered two questions to assess race/ethnicity: “What race do you consider the child to be? (check all that apply)”, followed by “Do you consider the child Hispanic/Latino/Latina?” Responses across the two items were recoded into a five-level variable designed to map onto United States Census categories—non-Hispanic White, non-Hispanic Black, Hispanic, Asian, or other racial/ethnic identity [31]. Youth identification as a sexual orientation minority was derived from their responses to the question “Are you gay or bisexual?” at each longitudinal assessment within the ABCD Study protocol [32] (ages 9–13 years). Youth replying “yes” or “maybe” at any assessment were included in the sexual orientation minority group [32]—12% of participants. Youth responses of “I don't know” were excluded when determining membership in the sexual orientation minority group [33]. For each participant, prepandemic household income was drawn from the ABCD Study annual assessment completed most recently before March 19, 2020. Caregivers reported annual household gross income from all sources on a 10-point scale listing income ranges. Responses were recoded to the center of the stated range. 30% of families reported annual income of less than $50,000, 28% of $50,000–$100,000, and 43% of more than $100,000.
Analytic plan
Analyses were conducted in R v4.2.1. Data were weighted to address survey nonresponse and improve sample representativeness (see the supplement for details). Response rates were lower for the during-pandemic survey waves (Figure 1), as expected given they occurred during a more chaotic period during family's lives (i.e., the pandemic). Thus, it was important to ensure that differential sample composition across timepoints did not confound our findings. Following recommended practice for survey analysis [34], we estimated the inverse probability of nonresponse weights [35] to ensure that the analytic sample at each timepoint was similar to the full ABCD Study sample on both key risk factors for substance use (family history of alcohol and drug use, diagnosis with an externalizing spectrum disorder, prepandemic history of alcohol or drug use) and sociodemographic characteristics (youth sex and race/ethnicity; family income, structure, and employment; Census region; and household size). Table A2 confirms that after applying the nonresponse weights, the composition of the analytic sample at every longitudinal timepoint was nearly identical to the full ABCD Study sample on both the key risk factors for substance use and the sociodemographic characteristics. Thus, we could safely proceed to compare the rates of substance use across timepoints.
Next, we multiplied the nonresponse weights by preconstructed baseline weights to create product weights that would ensure the analytic sample at each timepoint was representative of those aged 9–10 years old in the United States Census' Bureau's American Community Survey (2011–2016) on the same sociodemographic characteristics listed above [31]. After weighting, the analytic sample at each timepoint had a similar sociodemographic composition to the Census data, differing by < 1 percentage point on all variables.
We fit regressions in the survey package [36] using the logistic link function and clustering observations on study site, family, and youth to account for repeated observations and nonindependence. A regression was fit for each dependent variable. Timepoint was entered as seven dummy variables contrasting each during-pandemic timepoint with the prepandemic timepoint as the reference level. Age-at-observation was included as a covariate to adjust for maturation between timepoints. Parents’ marital status and education were included as covariates known to predict early adolescent substance use [9]. Data were weighted with the product weight.
Tests of moderation were conducted separately for each putative moderator. To reduce the number of tests, we focused on moderating the change in a single dependent variable: past-month use of any substance. We compared models with and without terms for the interaction between the timepoint and the putative moderator using likelihood ratio tests [37]. When that overall test was significant, we applied a multiple testing correction [38] for contrasts at individual timepoints. Data were weighted with the nonresponse weight.
Results
See Table 1 for pandemic-related context when interpreting findings across timepoints. Table 2 reports estimates from regression models. Figures 2 and 3 graph the model-estimated past-month prevalence of alcohol and drug use at each of the eight timepoints. For graphing, the youth age was set at 12.5 years old to capture the mean prevalence of past-month substance use among 12-year-olds. All regression models adjusted for age-at-observation, so the estimates being graphed in Figures 2 and 3 should not be interpreted as individual-level longitudinal trajectories or developmental changes in substance use—rather they reflect differences between same-age youth on different calendar dates.Table 2 Regression models for testing impact of COVID-19 pandemic on prevalence of alcohol and drug use
Term Odds of youth using in the past month:
Any substance Alcohol Nicotine Cannabis Prescription drugsa Inhalantsb
OR Coef. SE p OR Coef. SE p OR Coef. SE p OR Coef. SE p OR Coef. SE p OR Coef. SE p
(Intercept) - −3.57 0.16 <.001 - −3.40 0.12 <.001 - −6.52 0.43 <.001 - −7.60 0.80 <.001 - −21.78 0.33 <.001 - −9.52 1.03 <.001
Age-at-observation 1.7 0.51 0.13 .002 2.1 0.76 0.16 <.001 1.7 0.54 0.27 .07 5.4 1.69 0.78 .054 - 0.16 0.38 .68 1.6 0.46 0.51 .38
COVID:
May 2020 1.2 0.15 0.18 .42 0.4 −0.83 0.20 .002 7.0 1.95 0.46 .001 1.5 0.38 0.80 .65 - 16.38 0.35 <.001 42.4 3.75 1.13 .007
COVID:
June 2020 1.0 −0.00 0.16 .99 0.4 −0.94 0.15 <.001 6.7 1.90 0.43 <.001 1.2 0.14 1.13 .90 - 15.65 0.47 <.001 23.8 3.17 1.24 .03
COVID:
August 2020 1.2 0.20 0.23 .39 0.5 −0.64 0.20 .008 7.4 2.01 0.47 .001 0.8 −0.21 0.93 .83 - 16.16 0.53 <.001 7.8 2.06 1.25 .13
COVID:
October 2020 1.1 0.06 0.21 .76 0.4 −0.83 0.24 .006 6.2 1.83 0.45 .002 2.4 0.86 0.90 .36 - 16.15 0.35 <.001 28.6 3.35 1.24 .02
COVID:
December 2020 0.9 −0.11 0.19 .59 0.3 −1.06 0.32 .006 4.9 1.59 0.44 .004 1.6 0.46 1.01 .66 - 15.99 0.36 <.001 13.7 2.61 1.19 .051
COVID:
March 2021 0.7 −0.36 0.21 .12 0.2 −1.48 0.37 .002 3.6 1.29 0.48 .02 1.0 0.01 0.66 .98 - 15.59 0.52 <.001 20.1 3.00 1.18 .03
COVID:
May 2021 0.5 −0.63 0.27 .04 0.1 −2.53 0.40 <.001 3.5 1.27 0.67 .09 2.8 1.04 1.13 .38 - 15.63 0.71 <.001 18.8 2.93 1.27 .04
Reports six logistic regression models, one for each dependent variable. Coefficients and standard errors are in the log-odds metric. Age at observation was centered at 12.5 years old and scaled in years. The value of 12.5 years was chosen to reflect the average 12-year-old, given that 12-year-olds range in age from 12.0 to 12 .9¯ years. Fixed effects for parent education and marital status are omitted. Each model included 22,287 observations of 9,270 youth. Data were weighted to be sociodemographically representative of children in the United States Census (see Methods).
Coef. = coefficient; OR = odds ratio (exponentiated coefficient); SE = standard error; p = p-value.
a Odds ratios are omitted for the dependent variable of prescription drugs. The model-estimated prevalence of prescription drug misuse pre-COVID was nearly zero (see Figure 2), leading to the very large odds ratios implied by the reported coefficients. As a sensitivity analysis, we re-fit the model for prescription drugs using a linear (vs. logistic) link function and obtained a similar pattern of findings as reported in the table. The coefficient on all timepoint terms was positive; p-values for the coefficients ranged from .007 to .08 at the timepoints from May 2020 to March 2021, and the p-value for the coefficient on the May 2021 timepoint equaled .22.
b As for the dependent variable of prescription drugs, the model-estimated prevalence of inhalant use pre-COVID was nearly zero (see Figure 2), leading to the very large odds ratios listed. As a sensitivity analysis, we refit the model for prescription drugs using a linear (vs. logistic) link function. Using the linear link function, the difference from prepandemic was only statistically significant in May 2020 (p = .04); p values ranged from .18–.61 at the remaining timepoints.
Figure 2 Model-estimated past-month prevalence of use of alcohol and drugs by timepoint. Note. Table 2 reports the corresponding regression models. Prevalences of use were estimated for participants aged 12.5 years old (i.e., the average age of 12-year-olds) at each timepoint. Timepoint was modeled as an eight-level categorical variable. White dots indicate timepoints that are significantly different from pre-COVID levels (p < .05); black dots indicate that timepoints are not. Horizontal, dashed red lines indicate the pre-COVID prevalence, for comparison. Vertical bars indicate asymptotic 95% confidence intervals about the mean. Data were weighted with the product weight (see Analytic plan).
Figure 3 Model-estimated past-month prevalence of substance use by timepoint, by youth racial/ethnic identification, or prepandemic household income. Note. Table A3 reports the corresponding regression models. Prevalences of use were estimated for participants aged 12.5 years old (i.e., the average age of 12-year-olds) at each timepoint. Data were weighted with the nonresponse weight (see Analytic plan). Panels A–D graph estimates for a minority racial/ethnic group (colored lines unique to each panel) against estimates for the reference group of White youth (gray line reproduced across panels). Points with a box around them indicate timepoints at which those identified as the minority racial/ethnic group (Black, Hispanic, Asian, or other racial/ethnic identity) had changed from Pre-COVID to a degree significantly different from the change by those identified as White (p < .05 after adjustment for multiple testing [38]). Among participants identified as Asian, statistically significant contrasts are likely due to near-zero predicted prevalences and the implied near-infinite odds ratio: these contrasts were not statistically significant when tested in linear (vs. logistic) models. Thus, we do not interpret these contrasts in the text of manuscript. Panel E graphs estimates for youth at different levels of household income. Household income was modeled as a continuous variable; graphed are the estimated prevalences at four levels of income ($20,000, $50,000, $100,000, and $200,000). Points with boxes around them indicate timepoints at which household income significantly moderated the change in prevalence of any substance use relative to pre-COVID (p < .05 after adjustment for multiple testing [38]).
Consistent with the age range of participants, the estimated past-month prevalence of use across timepoints was ≤3.2% for alcohol, ≤1.1% for nicotine, and ≤0.4% for other categories. Most endorsements of alcohol or drug use (77%) were for 1–2 days of use in the past month. See Figure 2. The rate of alcohol use was significantly lower than the prepandemic level at all seven during-pandemic timepoints, with the decrease growing larger over time (relative risks [RRs] = 0.1–0.5, ps=<0.001–0.008 across timepoints). The rate of nicotine use was significantly higher than the prepandemic level at the first six timepoints, from May 2020 to March 2021 (RRs = 3.3–7.1, ps = <0.001–0.02) and was no longer significantly different from prepandemic in May 2021 (p = .09). The rate of prescription drug misuse (ps<0.001) and inhalant use (ps = 0.007–0.13) was significantly higher than the prepandemic level at nearly all during-pandemic timepoints, including the final timepoint in May 2021. The rate of cannabis use was not significantly different from the prepandemic level at any of the during-pandemic timepoints (ps = 0.36–0.98). Reflecting offsetting changes across the substance categories, the rate of any substance use did not differ significantly from the prepandemic level at the first six timepoints before being significantly lower (RR = 0.5, p = .04) in May 2021.
Table A3 reports estimates from regression models with interactions. When predicting youth use of any substance, tests of the interaction between timepoint and youth sex (p = .45) and youth identification as a sexual orientation minority (p = .53) were not statistically significant. There was a statistically significant interaction between timepoint and youth racial/ethnic identification (p = .01). See Figure 3, Panels A-D. After adjusting contrasts at individual timepoints for multiple testing, there were five significant contrasts (p < .05). Pre-COVID, adjusting for parent education, marital status, and household income, both Black and Hispanic youth were less likely than White youth to report past-month substance use. While the rate of any substance use fell throughout the pandemic among White youth, it grew during the initial phase of the pandemic among Black and Hispanic youth and returned to near the pre-COVID level by May 2021. Compared to White youth, the degree of change from pre-COVID was significantly greater for Black youth in June 2020 and for Hispanic youth in June 2020, August 2020, March 2021, and May 2021 (ps<0.05).
There was also a significant interaction between timepoint and prepandemic household income (p = .03). See Figure 3, Panel E. After adjusting contrasts at individual timepoints for multiple testing [38], the degree of change from pre-COVID was significantly moderated by household income at four timepoints spanning June 2020 to December 2020 (ps<0.05). Pre-COVID, youth from higher-income families were more likely than those from lower-income families to report past-month use of any substance. This pattern reversed from June 2020 to December 2020, with youth from lower-income families reporting greater rates of any substance use. While rates of use among higher-income families had fallen or were stable compared to prepandemic, rates among lower-income families had risen.
Discussion
We examined pandemic-related changes in the past-month prevalence of alcohol and drug use using data collected from 9,270 youth ages 11.5–13.0 years old at 21 sites across the United States. Between May 2020 and May 2021, adolescents experienced a partial return to life prepandemic: fewer families were engaging in social distancing, more youth were completing schooling in-person, and more parents were working exclusively outside the home (Table 1). Nonetheless, three of the four differences in substance use compared to prepandemic that were detectable in May 2020—fewer youth using alcohol, more youth misusing prescription drugs, and more youth using inhalants—persisted at follow-up in May 2021.
In a nationwide sample of United States eighth graders [4], past-month rates of alcohol, cannabis, and cigarette use were lower in spring 2021 than in 2020, prepandemic. Our findings in slightly younger youth are consistent in that we also found that changes persisted into spring 2021 and that the rate of alcohol use decreased. However, we found increased (vs. decreased) prevalence of nicotine product use and we found unchanged (vs. decreased) prevalence of cannabis use. Neither the sample of eighth graders [4] nor any previously published analysis has investigated the evolution of pandemic-related changes across multiple timepoints, so comparison to previous findings along that dimension is not possible.
The increase in nicotine use, prescription drug misuse, and inhalant use shrank as the pandemic continued beyond the acute phase in May 2020, with the increase in nicotine use no longer being statistically significant in May 2021. This pattern is consistent with the hypothesis that the remaining pandemic-related increases will continue to shrink as adolescents return closer to the structure of their daily lives before the pandemic (e.g., returning from 44% completing schooling in-person in May 2021 to 100% doing so [Table 1]). In contrast, the magnitude of the decrease in the rate of alcohol use grew steadily larger as the pandemic continued beyond the acute phase in May 2020, reaching a minimum rate in May 2021. In May 2021, 12-year-olds were less than one-tenth as likely to report past-month alcohol use as in May 2020. Our analyses do not explain why reductions in alcohol use occurred, but it seems plausible that entering early adolescence before versus during the pandemic could yield different socialization toward alcohol use [39]. For example, many United States youth in the fifth grade in the 2019-2020 school year spent none of the sixth grade and much of the seventh grade not attending any schooling in person [25], likely reducing the capacity of the middle school peer environment to socialize their thoughts, feelings, and behaviors around alcohol use. Likewise, they likely experienced fewer opportunities to drink in social contexts, such as spending time with friends after school or attending social gatherings [40].
We did not find evidence that the pandemic-related change in substance use varied by sex [2,3] or by sexual orientation minority status. We did find evidence of larger adverse impacts of the pandemic (i.e., increases in substance use) at some timepoints among youth who identified as Black or Hispanic and whose families had lower income, which consistent with a conceptualization of the COVID-19 pandemic as a syndemic [14,41], interacting with and exacerbating pre-existing inequities in the health risks and resources. For example, Black, Hispanic, and low-income parents were more likely to be front-line workers working outside the home [11], which may have reduced capacity to monitor youth who were completing their schooling online at home. Likewise, the greater disease burden experienced by Black, Hispanic, and low-income families may have placed these youth at greater risk for maladaptive coping through substance use in response to the hospitalization or death of a family member. Furthermore, national data indicates that Black and Hispanic adults in the United States were more likely to increase drinking during the pandemic [42], so parental drinking may have increased for adolescents in these groups. Future studies should investigate such mechanisms. Black, Hispanic, and low-income adolescents may especially benefit from supports around substance use as they emerge from the pandemic context [43]. However, findings should be viewed with caution-this was a single study, and findings have not yet been replicated. Our pre-COVID finding of a positive association between household income and substance use is not typical of the broader literature [[40], [41], [42]], and multiple moderators were examined.
This study had limitations. First, we examined the prevalence of primarily isolated use occasions (1–2 times per month) among youth in early adolescence (aged 11.5–13.0 years): the effect of the pandemic may differ in older youth who drink or use drugs regularly [2,3,47]. Second, the effect of the pandemic cannot be separated from the effects of concurrent events or unrelated secular trends, which must be considered when interpreting our findings. For example, the increases in substance use among Black youth in summer 2020 could be driven in part by the high-profile killings of Black individuals and associated civic activities [48]. Likewise, adolescents' perceptions of the harms of cannabis use had been decreasing in the years leading up to the pandemic [4], so continued decreases in perceived harm from 2020–2021 could confound any effect of the pandemic on cannabis use. Third, youth completed the prepandemic survey via phone interview and the during-pandemic surveys via the web, potentially introducing differences. However, we found both increase and decrease across substance categories, arguing against mode-of-assessment effects as a sole explanation of the pandemic-related changes. Fourth, alcohol and drug use was self-reported and not validated against toxicology—under-reporting may have occurred [49]. Fifth, past-month rates of cannabis, prescription drug, and inhalant use were very low, so the estimates for these drug classes should be viewed with caution. Finally, we identified the sexual orientation minority group via youth endorsement of being gay or bisexual, a procedure that may not have included all youth who identify as a sexual orientation minority [32] and that did not allow for evaluating potential differences among specific sexual orientations. Thus, the associated findings should be regarded with caution pending replication with more detailed measurements of sexual orientation that are planned for future ABCD assessments [32].
The current study comprises the longest follow-up of adolescent substance use during the COVID-19 pandemic published to date. We leveraged a mixed-age, prospective cohort to rigorously distinguish pandemic-related changes from the expected developmental increases in drinking and drug use. We characterized variability in pandemic-related changes across the population and evaluated changes in multiple drug classes. The large sample size, multisite recruitment, and racial/ethnic diversity all enhance the generalizability of findings. Continued follow-up will be necessary to anticipate the long-term impact of the COVID-19 pandemic on adolescents' alcohol and drug use.
Funding Sources
This research was supported by the 10.13039/100000027 National Institute on Alcohol Abuse and Alcoholism Grant AA030197. Pelham received additional support from the 10.13039/100000026 National Institute on Drug Abuse Grant DA055935. 10.13039/100004686 Thompson received additional support from the National Institute on Mental Health Grant MH128959. The ABCD Study is supported by the National Institutes of Health and additional federal partners under award numbers U01DA041048, U01DA050989, U01DA051016, U01DA041022, U01DA051018, U01DA051037, U01DA050987, U01DA041174, U01DA041106, U01DA041117, U01DA041028, U01DA041134, U01DA050988, U01DA051039, U01DA041156, U01DA041025, U01DA041120, U01DA051038, U01DA041148, U01DA041093, U01DA041089, U24DA041123, and U24DA041147. A full list of supporters is available at https://abcdstudy.org/federal-partners.html. Additional support for this work was made possible from supplements to U24DA041123 and U24DA041147, the 10.13039/100000001 National Science Foundation (NSF 2028680), and Children and Screens: Institute of Digital Media and Child Development Inc. The ABCD data repository grows and changes over time. The ABCD data used in this report came from the ABCD 4.0 data release (DOI: 10.15154/1523041), the ABCD COVID-19 Survey First Data Release (DOI: 10.15154/1520584), and the ABCD COVID-19 Survey Second Data Release (DOI: 10.15154/1522601). DOIs can be found at https://nda.nih.gov/study.html?id=1299, https://nda.nih.gov/study.html?&id=1041, and https://nda.nih.gov/study.html?&id=1225. The code for the analysis is available from the first author upon request. Dr. Gayathri Dowling was substantially involved in all of the cited grants.
Supplementary Data
Table A1-A3
Acknowledgments
Data used in the preparation of this article were obtained from the Adolescent Brain Cognitive Development (ABCD) Study (https://abcdstudy.org), held in the NIMH Data Archive (NDA). This is a multisite, longitudinal study designed to recruit more than 10,000 children aged 9–10 and follow them for 10 years into early adulthood.
Conflicts of interest: The authors have no conflicts of interest to declare.
Disclaimer: A listing of participating sites and a complete listing of the study investigators can be found at https://abcdstudy.org/Consortium_Members.pdf. ABCD consortium investigators designed and implemented the study and/or provided data but did not necessarily participate in the analysis or writing of this report. The views and opinions expressed in this manuscript are those of the authors only and do not necessarily represent the views, official policy or position of the U.S. Department of Health and Human Services or any of its affiliated institutions or agencies.
Supplementary data related to this article can be found at 10.1016/j.jadohealth.2023.02.040.
==== Refs
References
1 Chaffee B.W. Cheng J. Couch E.T. Adolescents’ substance use and physical activity before and during the COVID-19 pandemic JAMA Pediatr 175 2021 715 722 33938922
2 Dumas T.M. Ellis W. Litt D.M. What does adolescent substance use look like during the COVID-19 pandemic? Examining changes in frequency, social contexts, and pandemic-related predictors J Adolesc Health 67 2020 354 361 32693983
3 Thorisdottir I.E. Asgeirsdottir B.B. Kristjansson A.L. Depressive symptoms, mental wellbeing, and substance use among adolescents before and during the COVID-19 pandemic in Iceland: A longitudinal, population-based study Lancet Psychiatry 8 2021 663 672 34090582
4 Monitoring the Future Study. 1975-2021 Data for in-school surveys of 8th, 10th, and 12th Grade Students Available at: http://monitoringthefuture.org/data/21data.htm
5 von Soest T. Kozák M. Rodríguez-Cano R. Adolescents’ psychosocial well-being one year after the outbreak of the COVID-19 pandemic in Norway Nat Hum Behav 6 2022 217 228 35058644
6 Masonbrink A. Middlebrooks L. Gooding H.C. Substance use disorder visits among adolescents at children’s hospitals during COVID-19 J Adolesc Health 70 2022 673 676 35177345
7 Friedman J. Godvin M. Shover C.L. Trends in drug overdose deaths among us adolescents, January 2010 to June 2021 JAMA 327 2022 1398 1400 35412573
8 Layman H.M. Thorisdottir I.E. Halldorsdottir T. Substance use among youth during the COVID-19 pandemic: A systematic review Curr Psychiatry Rep 24 2022 307 324 35476186
9 Pelham W.E. III Tapert S.F. Gonzalez M.R. Early adolescent substance use before and during the COVID-19 pandemic: A longitudinal survey in the ABCD study cohort J Adolesc Health 69 2021 390 397 34452728
10 Mude W. Oguoma V.M. Nyanhanda T. Racial disparities in COVID-19 pandemic cases, hospitalisations, and deaths: A systematic review and meta-analysis J Glob Health 11 2021 05015 34221360
11 Gemelas J. Davison J. Keltner C. Ing S. Inequities in employment by race, ethnicity, and sector during COVID-19 J Racial Ethn Health Disparities 9 2022 350 355 33452573
12 McKnight-Eily L.R. Okoro C.A. Strine T.W. Racial and ethnic disparities in the prevalence of stress and worry, mental health conditions, and increased substance use among adults during the COVID-19 pandemic — United States, April and May 2020 Morb Mortal Wkly Rep 70 2021 162 166
13 Krueger E.A. Barrington-Trimis J.L. Unger J.B. Leventhal A.M. Sexual and gender minority young adult coping disparities during the COVID-19 pandemic J Adolesc Health 69 2021 746 753 34412952
14 Gravlee C.C. Systemic racism, chronic health inequities, and COVID-19: A syndemic in the making? Am J Hum Biol 32 2020 e23482
15 Karmakar M. Lantz P.M. Tipirneni R. Association of social and demographic factors with COVID-19 incidence and death rates in the US JAMA Netw Open 4 2021 e2036462
16 Tai D.B.G. Shah A. Doubeni C.A. The disproportionate impact of COVID-19 on racial and ethnic minorities in the United States Clin Infect Dis 72 2021 703 706 32562416
17 Ruprecht M.M. Wang X. Johnson A.K. Evidence of social and structural COVID-19 disparities by sexual orientation, gender identity, and race/ethnicity in an urban environment J Urban Health 98 2021 27 40 33259027
18 Volkow N.D. Koob G.F. Croyle R.T. The conception of the ABCD study: From substance use to a broad NIH collaboration Dev Cogn Neurosci 32 2018 4 7 29051027
19 Garavan H. Bartsch H. Conway K. Recruiting the ABCD sample: Design considerations and procedures Dev Cogn Neurosci 32 2018 16 22 29703560
20 Feldstein Ewing S.W. Dash G.F. Thompson W.K. Measuring retention within the adolescent brain cognitive development (ABCD) study Dev Cogn Neurosci 54 2022 101081 35152002
21 Pelham W.E. III Tapert S.F. Gonzalez M.R. Parental knowledge/monitoring and adolescent substance use: A causal relationship? Health Psychol 2022 1 11
22 Center for systems Science and Engineering. COVID-19 Dashboard. Johns Hopkins Coronavirus Resource Center.
23 Bureau of Labor Statistics Local Area Unemployment statistics Available at: https://www.bls.gov/lau/
24 Centers for Disease Control and Prevention COVID data Tracker Available at: https://covid.cdc.gov/covid-data-tracker/
25 Bureau of Labor Statistics Labor force statistics from the Current Population Survey: Supplemental data measuring the effects of the coronavirus (COVID-19) pandemic on the labor market Available at: https://www.bls.gov/cps/effects-of-the-coronavirus-covid-19-pandemic.htm
26 Lisdahl K.M. Sher K.J. Conway K.P. Adolescent brain cognitive development (ABCD) study: Overview of substance use assessment methods Dev Cogn Neurosci 32 2018 80 96 29559216
27 Johnston L.D. Miech R.A. O’Malley P.M. Monitoring the future national survey results on drug Use, 1975-2020: Overview, key findings on adolescent drug Use 2021 Institute for Social Research Ann Arbor
28 Fosse E. Winship C. Analyzing age-period-cohort data: A review and critique Annu Rev Sociol 45 2019 467 492
29 Pelham W.E. III Yuksel D. Tapert S.F. Did the acute impact of the COVID-19 pandemic on drinking or nicotine use persist? Evidence from a cohort of emerging adults followed for up to nine years Addict Behav 131 2022 107313 35413486
30 Stuart E.A. Matching methods for causal inference: A review and a look forward Stat Sci 25 2010 1 21 20871802
31 Heeringa S.G. Berglund P.A. A guide for population-based analysis of the Adolescent Brain Cognitive Development (ABCD) study baseline data bioRxiv 2020
32 Potter A.S. Dube S.L. Barrios L.C. Measurement of gender and sexuality in the adolescent brain cognitive development (ABCD) study Dev Cogn Neurosci 4 2022 101057
33 Potter A. Dube S. Allgaier N. Early adolescent gender diversity and mental health in the adolescent brain cognitive development study J Child Psychol Psychiatry 62 2021 171 179 32463952
34 Heeringa S.G. West B.T. Berglund P.A. Applied survey data analysis 2017 CRC Press
35 Seaman S.R. White I.R. Review of inverse probability weighting for dealing with missing data Stat Methods Med Res 22 2013 278 295 21220355
36 Lumley T. Analysis of complex survey samples J Stat Softw 9 2003 1 19
37 Lumley T. Scott A. Tests for regression models fitted to survey data Aust N Z J Stat 56 2014 1 14
38 Benjamini Y. Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing J R Stat Soc Ser B Methodol 57 1995 289 300
39 Donovan J.E. Child and adolescent socialization into substance use Zucker R.A. Brown S.A. The Oxford Handbook of Adolescent Substance Use 2019 Oxford University Press 345 372
40 Calihan J.B. Levy S. Coronavirus disease pandemic and adolescent substance use Curr Opin Pediatr 34 2022 334 340 35836392
41 Bailey Z.D. Krieger N. Structural racism and health inequities in the USA: Evidence and interventions The Lancet 389 2017 1453 1463
42 Kerr W.C. Ye Y. Martinez P. Longitudinal assessment of drinking changes during the pandemic: The 2021 COVID-19 follow-up study to the 2019 to 2020 national alcohol survey Alcohol Clin Exp Res 46 2022 1050 1061 35753040
43 Shim R.S. Starks S.M. COVID-19, structural racism, and mental health inequities: Policy implications for an emerging syndemic Psychiatr Serv 72 2021 1193 1198 33622042
44 Miech R.A. Johnston L.D. Bachman J.G. Monitoring the future: National survey results on drug Use, 1975-2019: Volume I, secondary school Students 2020 Institute for Social Research, University of Michigan
45 Maggs J.L. Staff J. Patrick M.E. Wray-Lake L. Very early drinking: Event history models predicting alcohol use initiation from age 4 to 11 years Addict Behav 89 2019 121 127 30290300
46 Donovan J.E. Molina B.S.G. Childhood risk factors for early-onset drinking J Stud Alcohol Drugs 72 2011 741 751 21906502
47 Miech R. Patrick M.E. Keyes K. Adolescent drug use before and during U.S. national COVID-19 social distancing policies Drug Alcohol Depend 226 2021 108822 34214884
48 Baskin-Sommers A. Simmons C. Conley M. Adolescent civic engagement: Lessons from black lives Matter Proc Natl Acad Sci 118 2021 e2109860118
49 Wade N.E. Tapert S.F. Lisdahl K.M. Substance use onset in high-risk 9–13 year-olds in the ABCD study Neurotoxicol Teratol 91 2022 107090 35341934
|
PMC010xxxxxx/PMC10187525.txt |
==== Front
Nat Rev Endocrinol
Nat Rev Endocrinol
Nature Reviews. Endocrinology
1759-5029
1759-5037
Nature Publishing Group UK London
37193881
822
10.1038/s41574-023-00822-7
Review Article
The G protein-coupled oestrogen receptor GPER in health and disease: an update
http://orcid.org/0000-0001-9190-8302
Prossnitz Eric R. eprossnitz@salud.unm.edu
123
http://orcid.org/0000-0002-8200-4341
Barton Matthias barton@access.uzh.ch
45
1 grid.266832.b 0000 0001 2188 8502 Department of Internal Medicine, Division of Molecular Medicine, University of New Mexico Health Sciences Center, Albuquerque, NM USA
2 grid.266832.b 0000 0001 2188 8502 Center of Biomedical Research Excellence in Autophagy, Inflammation and Metabolism, University of New Mexico Health Sciences Center, Albuquerque, NM USA
3 grid.266832.b 0000 0001 2188 8502 University of New Mexico Comprehensive Cancer Center, University of New Mexico Health Sciences Center, Albuquerque, NM USA
4 grid.7400.3 0000 0004 1937 0650 Molecular Internal Medicine, University of Zürich, Zürich, Switzerland
5 Andreas Grüntzig Foundation, Zürich, Switzerland
16 5 2023
2023
19 7 407424
28 2 2023
© Springer Nature Limited 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Oestrogens and their receptors contribute broadly to physiology and diseases. In premenopausal women, endogenous oestrogens protect against cardiovascular, metabolic and neurological diseases and are involved in hormone-sensitive cancers such as breast cancer. Oestrogens and oestrogen mimetics mediate their effects via the cytosolic and nuclear receptors oestrogen receptor-α (ERα) and oestrogen receptor-β (ERβ) and membrane subpopulations as well as the 7-transmembrane G protein-coupled oestrogen receptor (GPER). GPER, which dates back more than 450 million years in evolution, mediates both rapid signalling and transcriptional regulation. Oestrogen mimetics (such as phytooestrogens and xenooestrogens including endocrine disruptors) and licensed drugs such as selective oestrogen receptor modulators (SERMs) and downregulators (SERDs) also modulate oestrogen receptor activity in both health and disease. Following up on our previous Review of 2011, we herein summarize the progress made in the field of GPER research over the past decade. We will review molecular, cellular and pharmacological aspects of GPER signalling and function, its contribution to physiology, health and disease, and the potential of GPER to serve as a therapeutic target and prognostic indicator of numerous diseases. We also discuss the first clinical trial evaluating a GPER-selective drug and the opportunity of repurposing licensed drugs for the targeting of GPER in clinical medicine.
The 7-transmembrane G protein-coupled receptor GPR30 has been recognized as a G protein-coupled oestrogen receptor (GPER) since 2008. This Review discusses progress in GPER research in physiology and disease and its potential implications for clinical medicine.
Key points
Oestrogens exert multiple activities in physiology, including reproduction, immunity, cardiovascular and endocrine functions, and ageing, as well as in diseases such as hormone-sensitive cancers, arterial hypertension, atherosclerosis and osteoporosis.
Oestrogen signalling mediates both acute (non-genomic) and chronic (transcriptional) effects through cytosolic or nuclear oestrogen receptors ERα and ERβ and membrane subpopulations and the G protein‐coupled oestrogen receptor (GPER), which is a 7-transmembrane protein.
Molecules that activate oestrogen receptors include natural oestrogens, phytooestrogens, mycooestrogens and synthetic compounds, such as selective oestrogen receptor modulators and downregulators and xenooestrogens (also known as endocrine disruptors), activate oestrogen receptors and/or GPER.
Research using Gper-deficient animals, GPER‐selective agonists and antagonists, and non-selective compounds has revealed multiple roles of GPER in physiology and disease, including as a constitutive activator of the reactive oxygen species-producing enzyme NOX1.
GPER holds potential to become a diagnostic, prognostic and therapeutic target in clinical medicine, including the repurposing of licensed drugs targeting GPER and the ongoing first-in-human clinical trial of the GPER-selective agonist G-1.
Subject terms
Medical research
Therapeutics
Endocrine system and metabolic diseases
issue-copyright-statement© Springer Nature Limited 2023
==== Body
pmcIntroduction
Although actions of sex steroid hormones were described more than 2,000 years ago1, the concept of a ‘hormone’ was first introduced in 1910 by Starling2. It has been a hundred years since the chemical structures of oestrogens (and other steroids) were determined3,4 (Box 1). Identification and characterization of oestrogen receptors began in the 1950s by Jensen, Szego and others5–7, leading to the cloning of oestrogen receptor-α (ERα) by Chambon and associates in 1985 (ref. 8) (Box 1). In 1996, Kuiper et al.9 and Mosselman et al.10 cloned and identified oestrogen receptor-β (ERβ) contemporaneously with several reports describing the cloning of the orphan G protein-coupled receptor GPR30 (reviewed in refs. 7,11) (Box 1). GPR30 is a protein that predates the evolutionary divergence of fish and tetrapods more than 450 million years ago12. The discoveries that oestrogen binds to and activates cell signalling via GPR30 (refs. 13–15), establishing it as a transmembrane oestrogen receptor, resulted in its designation as the G protein-coupled oestrogen receptor (GPER) by the International Union of Basic and Clinical Pharmacology in 2008 (refs. 11,16). Following up on our previous article in Nature Reviews Endocrinology11, we now provide an update on the field of GPER research over the past decade. We will discuss advances made in cell signalling, molecular biology, pharmacology and genetics related to GPER. Special emphasis is given to the roles of GPER in pathophysiology and human disease and as a potential diagnostic, prognostic and therapeutic target in numerous and diverse areas of clinical medicine.
Box 1 Timeline of key discoveries in oestrogen and oestrogen receptor research This timeline shows the important milestones in the discovery and study of oestrogen. These include oestrogen chemistry, its receptors, mechanisms of action and pharmacology, with particular emphasis on the recent advances related to the study of GPER functions in health, disease and drug discovery.
1920s 1920s: Isolation and purification of oestrogens3,4
1928: Progynon (a 16α-oestriol glucuronide extract) commercially produced and prescribed to treat amenorrhoea4
1929: Acute vasodilatation in response to oestrogen of tissue transplanted into the eye277
1930s 1930: Ovarian extracts containing oestrogens acutely lower capillary pressure278
1930: Emmenin (16α-oestriol glucuronide extract) commercially produced and prescribed as oestrogen replacement279
1938: Diethylstilbestrol (DES) discovered280
1939: Acute vasodilation by oestrogens shown in humans281
1940s 1941: Urine extract from pregnant mares (Premarin) marketed by Pfizer as an oestrogen replacement282
1941: FDA approves DES for atrophic vaginitis, menopausal symptoms and lactation suppression283
1950s 1950s: Contraceptive pill developed284
1958–1960: First non-steroidal anti-oestrogen ethamoxytiphetol discovered285
1958–1960: Radioactive tracers concentrate in reproductive tissues; the binding sites are called ‘oestrogen receptors’5
1960s 1960s: ICI-46,474 (later named Tamoxifen) developed for use as a contraceptive286
1966–1968: Oestrogen binding characterized in rat uterus287,288
1967–1975: Rapid oestrogen effects on cAMP and intracellular calcium release discovered289,290
1969: Purification of an oestrogen receptor from rat uterus; anti-receptor immunoglobulin abolishes 17β-oestradiol binding6
1970s 1972: Tamoxifen repurposed for breast cancer treatment291
1979: Plasma membrane oestrogen receptors identified292
1980s 1985–1986: Cloning of oestrogen receptor-α (ERα)8
1990s 1996: Cloning of oestrogen receptor-β (ERβ)9,10
1996–1998: Cloning of GPR30 (refs. 293–299)
2000s 2000–2002: The role of GPR30 in mediating rapid 17β-oestradiol signalling discovered13,19
2005: 17β-Oestradiol binding to GPR30 demonstrated14,15
2006–2009: GPR30 activation dilates human arteries and lowers blood pressure; 17β-oestradiol regulates human arterial GPR30 expression; GPR30 expression prevents obesity91,121
2006–2011: First GPR30-selective agonist (G-1) and antagonists (G15 and G36) developed67–69
2008: International Union of Basic and Clinical Pharmacology designates GPR30 as G protein-coupled oestrogen receptor (GPER)16
2009: Role of GPER and efficacy of G-1 treatment in multiple sclerosis shown119,249
2010s 2010: Protective effects of GPER in myocardial reperfusion injury shown127
2011: GPER mediates 17β-oestradiol-stimulated pancreatic β-cell insulin secretion26
2016: GPER regulates NOX1; G36 identified as NOX1 downregulator39,157
2016: Roles of GPER in melanin production and therapeutic effects of G-1 in malignant melanoma shown207,228
2019: Phase I clinical trial of G-1 (LNS8801) for cancer78–80
2019: First ERα-selective and ERβ-selective agonist AB-1 developed77
2020s 2020: Efficacy of G-1 in obesity and diabetes mellitus treatment shown168
Molecular signalling mediated by GPER
G protein-coupled receptors (GPCRs) are 7-transmembrane spanning proteins that conventionally reside at the plasma membrane and signal to heterotrimeric G proteins, among other proteins, upon binding of ligands to their extracellular surface or within their transmembrane helices. GPER is predominantly expressed on intracellular membranes (the endoplasmic reticulum and Golgi apparatus), with little detected at the plasma membrane in many cell types14. While most investigations support this localization, limited expression in the plasma membrane in certain cell types (for example, uterine and renal epithelium), with constitutive internalization, has been reported17. Nuclear localization of GPER has also been observed and was suggested to be required for the GPER-mediated induction of transcription and cell migration18.
GPER signals through multiple G proteins, including Gαs15,19 and Gαi14,20 proteins, as well as via Gβγ signalling13, and possibly Gαq/11 protein21 (Fig. 1). GPER signalling involves, or possibly requires, epidermal growth factor (EGF) receptor transactivation13, a mechanism that, at the time this study was published in 2000, had only recently been discovered22. In addition to adenylyl cyclase19 and ERK1/2, GPER also activates PI3K–Akt signalling, which has been implicated in tumour cell survival23, activation of endothelial nitric oxide synthase (NOS3, also known as eNOS), nitric oxide (NO) formation and, thus, in cGMP-dependent vasodilation24,25 (Fig. 1). GPER also regulates ion channels, including those for calcium26, sodium27 and potassium28, and has been implicated in mTOR signalling and autophagy29.Fig. 1 Cellular signalling pathways activated by ERα, ERβ and GPER.
Non-genomic and genomic signalling pathways are activated by oestrogen and oestrogenic ligands (in yellow) through binding to the three known oestrogen receptors, oestrogen receptor-α (ERα), oestrogen receptor-β (ERβ) and the G protein-coupled oestrogen receptor (GPER). 17β-Oestradiol (E2), selective agonists such as G-1, or selective oestrogen receptor modulators (SERMs) and selective oestrogen receptor downregulators and/or degraders (SERDs) activate GPER (1), which is localized predominantly intracellularly at the endoplasmic reticulum. GPER activates several heterotrimeric G proteins (2), leading to multiple downstream cascades, including cAMP production (3) and activation of PKA (3) and CREB (3). G protein activation also leads to calcium (Ca2+) mobilization from intracellular stores, which activates PKC and leads to activation of plasma membrane calcium channels. GPER activation can also lead to regulation of gene expression via activation of the YAP–TAZ transcription factors via Rho–ROCK signalling (4). Activation of SRC via G proteins can also lead to activation of matrix metalloproteinases (MMPs) (5) that cleave pro-heparin-binding epidermal growth factor (HB-EGF) (5), releasing free HB-EGF. HB-EGF then transactivates the EGF receptor (5), which in turn activates MAPK (ERK1/2), Akt and other pathways. These induce additional, rapid (non-genomic) effects such as activation of the l-arginine–endothelial nitric oxide synthase (NOS3)–NO–cGMP pathway (in combination with mobilization of calcium stores). Akt causes phosphorylation of endothelial NOS3 (6), which releases nitric oxide (NO) and leads to juxtacrine signalling from endothelial to vascular smooth muscle cells (7), and activation of PKG. Activation of MAPK and Akt signalling also causes genomic effects regulating gene transcription such as FOXO3 phosphorylation and degradation (8). In the classic, genomic oestrogen receptor pathway, 17β-oestradiol binds cytosolic and nuclear oestrogen receptors (9), inducing receptor dimerization and binding to the promoters of target genes. Alternatively, activated oestrogen receptors modulate the function of other classes of transcription factors (TF) through protein–protein interactions (10). Subpopulations of membrane-bound oestrogen receptors (mER) are present at the plasma membrane (11). Once activated, these oestrogen receptors interact with adaptor proteins (adaptor) and signalling molecules, such as SRC, which mediate rapid signalling events (for example, PI3K–Akt and MAPK signalling) (11). Oestrogen receptor ERα, potentially following transactivation of EGFR by GPER, is regulated by phosphorylation through kinases (such as MAPK and Akt), resulting in the regulation of gene expression (12). HIF1α, following GPER activation, induces γ-secretase-dependent activation of NOTCH (13) and VEGF signalling (13). Basal expression and/or activity of GPER constitutively induces expression of the NADPH oxidase NOX1 (14).
Transcriptional regulation is often a consequence of rapid signalling, yielding sustained genomic effects (Fig. 1). Rapid signalling pathways initiated by GPER that lead to transcriptional regulation include adenylyl cyclase-generated cAMP-dependent phosphorylation of CREB30 and MITF31 by PKA. GPER inactivates the FOXO3 transcription factor via Akt, promoting cell survival23. GPER-mediated ERK1 and ERK2 activation leads to Elk1-mediated transcription, which upregulates FOS and subsequently CTGF, FGF2 and CYP1B1 production32,33. GPER can either activate or inhibit NF-κB transcriptional activity, depending on the cellular context34,35; GPER also γ-secretase-dependent activation of Notch, resulting in expression of HES1 and SNAIL36. GPER stimulation can activate YAP and TAZ, two homologous transcription coactivators and key effectors of the Hippo tumour suppressor pathway, via Gαq/11, PLCβ–PKC, ERK1/2 and the Rho–ROCK signalling pathways37 (Fig. 1). GPER expression, and therefore function, is also regulated by multiple microRNAs38. Finally, basal expression and activity of GPER constitutively regulate expression and activity of the NADPH oxidase NOX1 (ref. 39) (Fig. 1), a reactive oxygen species (ROS)-producing enzyme implicated in many non-communicable diseases40.
Natural and synthetic ligands of GPER
Oestrogen receptors are activated by a wide range of chemical entities derived from diverse sources, including endogenous oestrogens, phytooestrogens (plant-derived oestrogens), mycooestrogens (fungus-derived oestrogens) and xenooestrogens (synthetic molecules also known as ‘endocrine disruptors’) (Fig. 2). The identification and characterization of oestrogen receptors facilitated the development of targeted drugs, including selective oestrogen receptor modulators (SERMs) and selective oestrogen receptor downregulators (or degraders) (SERDs), some of which were, in fact, already available in the 1960s41 (Box 1). In the following section, we will discuss GPER-targeting steroidal ligands, xenooestrogens, plant-derived and fungus-derived molecules, and synthetic receptor-selective ligands and their activities with respect to GPER (Fig. 2).Fig. 2 Chemical structures of compounds that act as ligands for ERα, ERβ and/or GPER.
Shown are examples of natural steroids, phytooestrogens, xenooestrogens/endocrine disrupting chemicals (EDCs), therapeutic agents and experimental compounds that display varying activities towards oestrogen receptor-α (ERα), oestrogen receptor-β (ERβ) and the G protein-coupled oestrogen receptor (GPER) but are generally non-selective. Also shown are synthetic experimental compounds that exhibit selectivity for ERα and/or ERβ, such as propylpyrazoletriol (PPT), diarylpropionitrile (DPN) and AB-1, or for GPER, such as G-1, G15, G36 and CIMBA. p,p′-DDT, p,p′-dichlorodiphenyltrichloroethane.
Steroid hormones
GPER, at the time still known as the orphan receptor GPR30, was first linked to oestrogen-mediated signalling, in 2000, through the activation of ERK via transactivation of the EGF receptor13 (Box 1). High-affinity, competitive binding of 17β-oestradiol to GPER was first demonstrated in 2005 (refs. 14,15). In contrast to 17β-oestradiol, oestrogens, such as oestrone and oestriol, exhibit poor binding to GPER15. GPER shows no binding to other steroids, such as testosterone, progesterone, aldosterone and cortisol15,42–44, although aldosterone has been shown to be involved in crosstalk between the mineralocorticoid receptor and GPER and between the EGF receptor and GPER43. The catecholoestrogen 2-methoxy-oestradiol45 and the glucuronic acid metabolite 17β-oestradiol-17-d-glucuronide46 act as GPER agonists, whereas another catecholoestrogen, 2-hydroxy-oestradiol, is reported to act as an antagonist47. Dehydroepiandrosterone shows agonistic behaviour towards GPER48,49, whereas its metabolite 7β-hydroxy-epiandrosterone antagonizes GPER-mediated oestrogenic responses50. Most recently, 27-hydroxycholesterol, a cholesterol metabolite implicated in oestrogen receptor-negative breast cancer, was reported to bind and activate GPER, although with relatively low affinity compared with its most important physiological ligand, 17β-oestradiol51.
Xenooestrogens and natural oestrogenic molecules
Xenooestrogens are a large family of chemically stable synthetic molecules with oestrogenic activities often referred to as environmental oestrogens or endocrine-disrupting chemicals (EDCs). They are found in a wide range of consumer products and plastics, and most of them are toxic52. Endocrine-disrupting chemicals can be found in detergents, surfactants, resins, lubricants, plasticizers, fire retardants and pesticides52. Xenooestrogens that bind and/or regulate the activity of GPER (typically acting as agonists) include bisphenol A (BPA), polychlorinated biphenyls (PCBs), diethylstilbestrol (DES), nonylphenol, dichlorodiphenyltrichloroethane (DDT) and dichlorodiphenyltrichloroethylene isomers, kepone, methoxychlor and atrazine (Fig. 2).
Many molecules present in soy or green tea plants also target oestrogen receptors. Such naturally occurring phytooestrogens include flavonoids, isoflavonoids, chalcones, coumestans, stilbenes, lignans, ginsenosides and tetrahydrofurandiols53. Phytooestrogens that bind and/or activate GPER include genistein54, daidzein55, equol56, quercetin57, resveratrol58, oleuropein59, icariin60 and the green tea polyphenol (-)-epicatechin61. The mycooestrogen zearalenone also shows agonism towards GPER54,62.
Discovery of GPER-selective ligands
Owing to the highly conserved nature of the binding sites in ERα and ERβ, the typical affinity difference for oestrogen receptor subtype-specific compounds ranges from ~30-fold to 300-fold63. Oestrogen receptor subtype-biased ligands, such as propylpyrazoletriol (PPT, an ERα-selective agonist) and diarylpropionitrile (DPN, an ERβ-selective agonist) (Fig. 2), have been developed and are widely used64,65. PPT, however, also acts as a GPER agonist66, complicating the interpretation of its use.
The discovery and development of highly GPER-selective ligands were essential to facilitating research into the physiology and pathophysiology related to this receptor. In 2006, compound library screening led to the identification of G-1 (1-(4-(6-bromobenzo[1,3]dioxol-5-yl)-3a,4,5,9b-tetrahydro-3H-cyclopenta[c]quinolin-8-yl)-ethanone), a small molecule that acts as a selective agonist of GPER67 (Fig. 2). The discovery of GPER-selective antagonists G15 and G36 complemented the use of G-1 as an agonist in understanding the roles of GPER in cell biology and physiology68,69. Some reports suggest that the activity of these compounds can vary depending on the system employed70,71. Other reported GPER-selective ligands include the agonists GPERL1 and GPERL2 (ref. 72), a series of indole-thiazole derivates that act as GPER agonists73, the antagonist CIMBA (an acyclic analogue of G36)74, as well as the pan-oestrogen receptor and GPER antagonist MIBE75 (Fig. 2). Proteolysis-targeting chimaeras (PROTACs), which are molecules that induce degradation of specific proteins (via selective recruitment of E3 ubiquitin ligases and target ubiquitination followed by degradation in proteosomes), were developed to target ERα as early as 2005 (ref. 76), with a pair of 17β-oestradiol–proteolysis-targeting chimaeras shown to degrade GPER in addition to ERα in a study published in 2021 (ref. 44). The 2019 discovery of AB-1, an agonist of ERα and ERβ that lacks affinity for GPER, should allow further dissection of the functions of ERα and/or ERβ compared with GPER in cells that express multiple oestrogen receptors77. Of these GPER-targeting ligands, only G-1 has so far entered clinical trials, specifically for use in combination therapy with immune checkpoint inhibitors (ICIs) in cancer. G-1 exhibits a favourable safety profile in these trials, either alone or in combination with pembrolizumab, with encouraging initial antitumour activity observed to date (NCT04130516)78–80.
Roles of GPER in physiology and disease
In the following sections, we will review advances in understanding the functions of GPER in cardiovascular and kidney disease, endocrinology and metabolism, gastrointestinal and liver diseases, immunity and immunology, neurology, and the physiological ageing process. Findings are frequently based on effects due to phenotypes of Gper-deficient mice or the effects of GPER-selective ligands (Fig. 3). Reported phenotypes of multiple differently derived Gper-deficient mice are not entirely consistent, probably due to differences in genetic background and other factors, including age81. The available evidence points to multiple roles of GPER in oestrogen-dependent and oestrogen-independent functions and pathologies, allowing the development of possible diagnostic and therapeutic approaches with regard to GPER.Fig. 3 GPER in health and disease.
The G protein-coupled oestrogen receptor (GPER) regulates many physiological functions (white background) and is involved in multiple pathologies and diseases (pink background) CKD, chronic kidney disease; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; PAH, pulmonary arterial hypertension; VSMC, vascular smooth muscle cell.
Clinical genetics
Sex chromosomes, sex steroids and sex steroid receptors contribute to and determine disease risk and efficacy of pharmacological therapy82,83. In humans, the GPER gene maps to chromosome 7p22.3, a region associated with arterial hypertension in genetic linkage studies84. The GPER single-nucleotide polymorphism rs11544331, which results in a Pro16Leu alteration in the receptor (amino acid substitution of proline 16 to leucine), produces a hypofunctional variant of GPER. The Leu variant is associated with slightly higher blood pressure than the Pro variant in women but not in men, and its allele frequency is two-fold higher in women with hypertension compared with age-matched men85. The inhibitory effect of GPER on pro-inflammatory gene expression in induced pluripotent stem cell-derived endothelial cells is reduced in the GPER Leu variant compared with the Pro variant86. Moreover, GPER activation induces LDL receptor expression, in part by downregulating proprotein convertase subtilisin–kexin type 9 (PCSK9) resulting in increased plasma levels of LDL cholesterol in Pro16Leu variant carriers87. Finally, expression of the Pro16Leu variant of GPER in cancer-associated fibroblasts increases secretion of paracrine factors promoting migration of breast cancer cells88. Together, these genetic observations support potentially important roles for GPER for human diseases.
Cardiovascular and kidney diseases
Endogenous oestrogens in premenopausal women protect against cardiovascular diseases in general, and particularly against arterial hypertension, coronary heart disease (including myocardial infarction) and heart failure11,89,90. GPER is widely expressed in the cardiovascular system in mammals, including the arterial wall and the heart11. In the cardiovascular system, physiological functions of GPER include the regulation of arterial blood pressure, angiogenesis, myocardial contractility and suppression of inflammation11. Activation of GPER results in acute vasodilatation of human, pig, rat and mouse arteries91–93. The underlying mechanisms include direct effects on vascular smooth muscle91,92,94 and activation of the endothelial l-arginine–NOS3–NO–cGMP pathway24,95,96 (Fig. 1). GPER-mediated vasodilatation also involves cAMP-dependent97 and Rho kinase-dependent mechanisms98 as well as inhibiting contractile factors such as endothelial vasoconstrictor prostanoids99 and endothelin-1 (refs. 92,100). GPER-dependent vasodilation is augmented during pregnancy101 and is reduced by ageing39,102,103. Systemic deletion of Gper prevents age-dependent, endothelium-dependent dysfunction, probably due to a reduction in NOX1 abundance39,103; the main effects are summarized in Fig. 3.
Blood pressure and arterial hypertension
Endothelium-derived contracting factors, such as cyclooxygenase-derived vasoconstrictor prostanoids and endothelin 1, are involved in the pathogenesis of arterial hypertension89; their activity is suppressed by constitutive GPER activity and augmented by systemic deletion of Gper99. Similarly, acute (seconds to minutes)91 and chronic treatment (hours to days) with the GPER agonist G-1, via its nitric oxide (NO)-liberating and antioxidant effects24,95, induces vasodilation and lowers blood pressure. Interestingly, deletion of Gper prevents angiotensin II-induced elevations of blood pressure, which are also markedly lowered by the GPER antagonist and NOX1 downregulator G36 (refs. 39,40). These data suggest that either agonist-dependent activation (through increased NO bioactivity) or chronic antagonism of GPER (via NOX1 downregulation) could be suitable for the treatment of different forms of arterial hypertension and related diseases such as atherosclerosis, stroke and chronic kidney disease (CKD).
The GPER agonist G-1 prevents hypertension in intrauterine growth-restricted female rat offspring later in life, suggesting a potential role in embryonic priming of adult hypertension104. Arterial blood pressure in Gper-deficient animals is normal91,105 or slightly reduced compared with animals expressing GPER106. Crosstalk between GPER and endothelin receptors has been described, resulting in natriuretic effects107. Aldosterone, which also has natriuretic effects, has been implicated in the actions of GPER, yet there is no evidence of aldosterone binding to GPER42–44,108,109. Consistent with this, deletion of Gper has no effect on the hypertensive effects induced by aldosterone110; however, GPER does regulate autocrine aldosterone synthesis in the renal medulla111. In addition, crosstalk between the mineralocorticoid receptor and GPER has been reported43. Correspondingly, mineralocorticoid receptor antagonists downregulate the expression of GPER112. Moreover, aldosterone triggers both direct interactions between the mineralocorticoid receptor and GPER involving the EGF receptor, which is abrogated by GPER gene silencing in endothelial and SkBr3 breast cancer cells in vitro43. Such interactions between the mineralocorticoid receptor and GPER might also contribute to aldosterone-mediated regulation of the sodium–chloride cotransporter, which is reduced in male mice lacking Gper113.
Atherosclerosis and coronary artery disease
Atherosclerosis is a chronic systemic inflammatory vascular disease89 and the underlying cause of coronary artery disease (CAD), peripheral artery disease and stroke. The main complications of CAD are myocardial infarction, fatal ventricular arrhythmias following reperfusion injury after infarction, and heart failure89. Natural or surgical menopause accelerates CAD progression and can be alleviated by oestrogen therapy, which activates all three oestrogen receptors89. In mice of both sexes fed either a regular diet or a high-calorie diet rich in fat and sugars, deletion of Gper results in moderate dyslipidaemia114,115. In endothelial cells, oestrogen-mediated activation of GPER attenuates transcytosis of LDL cholesterol into endothelial cells, compatible with an indirect vasculoprotective effect116. G-1 also reduces cardiac lipid accumulation and PPARα expression in surgically postmenopausal rats with type 2 diabetes mellitus (T2DM)117. In human monocytes, which contribute to the earliest stages of atherogenesis118, the anti-inflammatory effects of oestrogen might involve both direct effects via GPER119 as well as crosstalk between ERα and GPER120.
In the arteries of patients with coronary artery disease, GPER expression is sensitive to 17β-oestradiol regulation121. Activation of GPER by G-1 or green tea polyphenols inhibits the growth of coronary vascular smooth muscle cells61,91,118,122–124, a crucial step during atherogenesis. Deletion of Gper increases both perivascular adipose tissue growth and the production of cyclooxygenase-dependent adipose-derived contracting factor (ADCF), suggesting that endogenous GPER activity negatively regulates these processes125. In ovariectomized, that is, surgically postmenopausal, ApoE-deficient mice or in surgically postmenopausal C57BL/6J mice fed a cholate-containing atherogenic diet, G-1 reduces inflammation and atherosclerosis126. G-1 also reduces steady-state mRNA levels of the angiotensin AT1 receptor in ApoE-deficient mice123, a receptor protein that mediates angiotensin II-dependent vasoconstriction, vascular cell growth, inflammation and oxidative stress.
Myocardial disease and heart failure
GPER activation attenuates reperfusion injury following myocardial infarction through pathways involving GSK3β, mitophagy and mechanisms regulating mitochondrial permeability127–129. Arterial hypertension, T2DM and the resulting coronary artery disease and loss of myocardial tissue from myocardial infarction are the most frequent causes of heart failure. While heart failure with reduced ejection fraction (HFrEF) is often due to the loss of contractile tissue following myocardial infarction, heart failure with preserved ejection fraction (HFpEF) is a consequence of diabetes mellitus, arterial hypertension and ageing, all resulting in myocardial fibrosis and stiffening89,90. Patients with HFpEF are primarily perimenopausal or early postmenopausal women, suggesting that the cessation of endogenous oestrogen production contributes to the pathogenesis of HFpEF.
In experimental models of HFrEF, oestrogen therapy can reverse heart failure-induced myocardial fibrosis130. ERα and ERβ, as well as GPER, are all involved in the inhibitory effects of oestrogen on cardiomyocyte proliferation131,132. Interestingly, SERMs and SERDs, which are also GPER agonists, also inhibit cardiomyocyte proliferation133. Hypoxia and/or hypoxaemia, which occur during myocardial ischaemia and heart failure, upregulate GPER134. GPER controls myocardial contractility involving crosstalk between GPER and β1 adrenoceptors135. In a model of ageing-associated HFpEF, systemic deletion of Gper in male mice prevents the development of heart failure and myocardial fibrosis, an effect that is related to downregulation of NOX1 protein expression and associated reduction of NOX1 function39. In vitro studies using Nox1-knock-in experiments in aortic vascular smooth muscle cells from Gper-deficient mice further underscored that constitutive NOX1 expression and activity require GPER expression, which, probably through ligand-independent or basal activity, enables ROS formation, inflammation and myocardial fibrosis39. By contrast, in young female mice, cardiomyocyte-specific deletion of Gper worsens cardiomyocyte function compared with wild-type mice both in vitro and in vivo, which can be partly rescued by inhibiting cardiac NLRP3 inflammatory pathways136. G-1 reduces diastolic dysfunction in experimental HFpEF137 and in rats with hypertensive cardiomyopathy137,138; G-1 treatment also improves cardiac function and reduces cardiac fibrosis in surgically postmenopausal rats139. Taken together, either reducing constitutive NOX1-dependent production of ROS by blocking GPER or increasing NO bioactivity by activating GPER, holds potential for pharmacological intervention in heart failure, possibly in a sex-dependent manner.
Renal physiology and disease
Loss of functional kidney tissue, particularly due to CKD, facilitates the development of arterial hypertension and cardiovascular disease. Similar to cardiovascular diseases, CKD displays sex differences with premenopausal women being largely protected from CKD development compared with age-matched men, implicating a role for oestrogens and oestrogen receptors140. GPER regulates renal artery and intrarenal vascular tone103,141, and its activation increases Ca2+ flux and H+-ATPase activity in renal tubular cells142; GPER also regulates natriuresis107 via crosstalk with endothelin ETA and ETB receptors143. Deletion of Gper counteracts the development of focal segmental glomerulosclerosis (FSGS) and the resulting proteinuria144 and tubulo-interstitial injury caused by inflammation and oxidative stress, by reducing NOX1 upregulation145. Activation of GPER also reduces glomerular mesangial cell proliferation induced by hyperglycaemia in vitro (which is associated with oxidative stress)60, and Gper silencing in these cells markedly reduces NOX1 abundance144. The GPER antagonist and NOX1 downregulator G36 reduces mRNA expression of podocyte injury markers NPHS1 (coding for nephrin), COL4A1 (collagen IV) and WT1 (Wilms-tumour 1) in human podocytes in vitro144. Protective effects of GPER signalling on podocytes have also been demonstrated for treatment with GPER agonists, probably via activation of the l-arginine–NOS–nitric oxide pathway146. In a model of hypertensive nephropathy, GPER activation reduces proteinuria as well as tubular injury but not glomerular injury via pressure-independent mechanisms147,148. Possibly, the stimulating effect of G-1 on tubular epithelial cell proliferation could contribute to this effect149. Protective effects of GPER signalling have also been reported for methotrexate-induced human renal epithelial cell injury in vitro150 and for acute renal endothelial cell injury following renal ischaemia in female mice151.
Pulmonary diseases
Pulmonary arterial hypertension (PAH) is a chronic fibroproliferative disorder of the pulmonary vasculature, ultimately leading to right-heart failure. Four out of five patients are women, suggesting a role for sex chromosomes, sex steroids, or sex steroid receptors. In experimental rat models of PAH, ovariectomy increases mortality152, while 17β-oestradiol (a non-selective oestrogen receptor and GPER agonist)153 or the GPER agonists G-1 (ref. 154) or 2-ME152 partially reduce or even reverse established cardiopulmonary injury. G-1 also improves skeletal muscle function and exercise intolerance in rats with PAH, possibly through normalization of SERCA2a and phospholamban expression154,155. Finally, in experimental hypoxia-induced PAH in rats, blocking GPER using G36 improves cardiac function by lowering right ventricular pressure, probably involving the downregulation of NOX1 (refs. 156,157). Thus, both agonists and antagonists of GPER might aid in the treatment of PAH.
Endocrinology and metabolism
Metabolic homeostasis is differentially regulated in men and women, with the metabolic actions of oestrogens mediated through both ERα158–160 and GPER; discussed later in this section. Premenopausal women exhibit lower incidences of obesity and T2DM compared with age-matched men; these protective effects are lost following menopause, with similar effects seen in rodents. Oestrogen therapy can alleviate weight gain and its associated adverse metabolic effects present in postmenopausal women and in surgically postmenopausal mice161–163.
Obesity and diabetes mellitus
Since the first reports demonstrating roles of endogenous GPER in the regulation of body weight, adipose tissue growth, obesity and insulin function in 2009 (refs. 91,164), studies in mice lacking Gper have found that these mice develop dyslipidaemia and show reduced energy expenditure compared with wild-type mice. These effects are probably responsible for the observed increases in visceral and subcutaneous adipose tissue depots, given that food intake and locomotor activity remain unaffected in Gper-deficient mice114,115,165. Compared with males, female ovary-intact Gper-deficient mice exhibit a lower sensitivity to acute leptin-stimulated food intake and short-term cholecystokinin-stimulated satiety signals165. The expression of thermogenic genes, such as those encoding uncoupling protein 1 (Ucp1) and the β3-adrenergic receptor, is reduced in brown adipose tissue of Gper-deficient mice consistent with the decreased energy expenditure.
17β-Oestradiol treatment protects β-cells from apoptosis and prevents diabetes mellitus in mice166. The severity of diabetes mellitus in mice lacking both ERα and ERβ worsens following surgical menopause167. 17β-Oestradiol supplementation improves glucose homeostasis in these mice, suggesting alternative mechanisms of oestrogen action other than signalling through ERα or ERβ, for example, through GPER167. Indeed, in mice lacking Gper, plasma levels of glucose are increased and these animals exhibit glucose intolerance, defective glucose-stimulated and oestrogen-stimulated insulin secretion, and insulin resistance114,164,165. Insulin secretion in response to both 17β-oestradiol and G-1 in healthy islets is reduced by pharmacological GPER inhibition and is absent in mouse islets lacking Gper26. In a mouse model of streptozotocin-induced diabetes mellitus, deletion of Gper results in greater loss of pancreatic β-cells, reduced pancreatic insulin content and, consequently, abnormally increased plasma levels of glucose compared with wild-type mice167.
GPER as a therapeutic target in obesity and diabetes mellitus
Therapeutic targeting of GPER in glucose homeostasis and lipid metabolism has been studied in models of Western diet-induced obesity in male mice and in models of surgical menopause in female mice, both of which result in obesity and metabolic dysfunction. G-1 treatment over a period of 6–8 weeks reduced overall body weight, adiposity and circulating levels of lipids compared with vehicle-treated mice, without affecting lean mass or bone density, via increased basal energy expenditure168. No changes in either daily food consumption or locomotion were observed in this study although, in surgically postmenopausal obese rats, G-1 treatment acutely and transiently decreased food intake169. G-1 treatment in surgically postmenopausal mice increased the expression of genes involved in mitochondrial biogenesis and fatty acid oxidation in brown and white adipose tissue and in skeletal muscle, while reducing the expression of genes involved in inflammation, hypoxia and angiogenesis168.
In line with previous results69,126, G-1 treatment of surgically postmenopausal obese mice was devoid of the feminizing effects of 17β-oestradiol as indicated by the absence of uterine imbibition168. In addition to weight loss and improved lipid profiles, G-1 also improved glucose homeostasis at the level of glucose and insulin tolerance tests, and reduced fasting blood levels of glucose and insulin168. In postmenopausal rats with streptozotocin-induced diabetes G-1 treatment reduced disease-induced weight loss to a comparable degree as did 17β-oestradiol treatment, and similarly improved glucose homeostasis and lipid profiles compared with vehicle-treated diabetic rats170. While surgically postmenopausal obese mice show improved glucose homeostasis in response to acute or chronic 17β-oestradiol treatment, deletion of Gper abrogates this response, indicating a key role of GPER in 17β-oestradiol-mediated glucose homeostasis in vivo164,165. Moreover, G-1 amplifies glucose-stimulated insulin secretion ex vivo in pancreatic islets obtained from patients with T2DM, while also suppressing glucagon and somatostatin secretion171,172. Thus, selective GPER agonists hold potential for the treatment of obesity and associated diseases such as diabetes mellitus.
Gastrointestinal and liver diseases
Oestrogens modulate multiple gastrointestinal and hepatic functions via their receptors173, including via GPER173. GPER is a cell-specific marker of gastric epithelium chief cells174 and also controls lower oesophageal sphincter tone175, colonic motility and severity of visceral pain176,177. In human Crohn’s disease178, ulcerative colitis179 and irritable bowel syndrome (IBS)180–182, the majority of studies found intestinal GPER expression to be increased compared with healthy individuals. GPER activation reduces inflammation, tissue injury and mortality in a mouse model of Crohn’s disease178 and G-1 reduces colonic crypt cell injury related to reperfusion injury following intestinal ischaemia183. Finally, intestinal inflammation in a mouse model of acute colitis induced by dextran sulfate sodium is reduced by GPER activation, improving intestinal mucosal barrier function184.
GPER regulates liver in zebrafish185 and contributes to oestrogen-dependent proliferation and lipid metabolism in human hepatocytes185,186. In addition, both GPER or ERα protect hepatocytes from fatty degeneration, a predisposing factor propagating non-alcoholic fatty liver disease and steatohepatitis187.
Obesity in premenopausal women is associated with an increased risk of developing gallstones, which are formed via GPER-dependent mechanisms188. Oestrogen-dependent cholesterol crystallization pathways differ markedly between those involving ERα or GPER189, yet deletion of Gper190 or its pharmacological inhibition74 completely prevents gallstone formation in female mice.
Cancer biology and oncology
GPER is expressed in tumours and tumour cells of cancer patients, including the mammary gland191–195, endometrium66,196, ovaries197, prostate198, pancreas199, thyroid200, colon201 and lung202. Increased GPER expression correlates with a worse outcome in breast191–193, endometrial196 and ovarian197 cancer. Although pharmacological activation of GPER can increase proliferation and associated signalling in breast203, endometrial204, thyroid200 and ovarian205 cancer cells, inhibition of proliferation due to GPER signalling has also been reported in breast206, pancreatic199 and melanoma207 cancer cells. With these — sometimes — opposing results in different cell lines, the role of GPER in cancer in vivo appears to be more complex than anticipated. Indeed, in certain forms of cancer, endogenous GPER activity might be protective, possibly through anti-inflammatory effects208.
Breast cancer
Much has been published regarding GPER and breast cancer due to obvious questions arising from the well-documented importance of presence or absence of ER for the efficacy of anti-oestrogen therapies in cancer treatment209. The fact that SERMs, such as tamoxifen14 and raloxifene66, as well as SERDs, such as fulvestrant13, act as GPER agonists to activate growth and survival pathways has led to the suggestion that GPER expression and/or activity could contribute to breast cancer recurrence194. This complex pharmacology has also led to a search for ERα-selective compounds that do not cross-react with GPER77.
Supporting roles for GPER in breast cancer recurrence and metastasis, GPER expression is elevated in metastases of patients with breast cancer compared with matched primary tumours210,211. However, this elevated GPER expression, where assessed, is only observed in women originally treated with tamoxifen211. Aromatase inhibitors are more effective than tamoxifen at inhibiting tumour growth in primary breast tumours that are both ERα-positive and GPER-positive, with this difference in treatment efficacy being absent in primary ERα-positive and GPER-negative breast tumours192. Moreover, aromatase inhibition resulted in better disease-free progression for patients with breast cancer compared with a tamoxifen-based therapy, consistent with a role for GPER in recurrence and metastasis193. Using a genetic mouse model of mammary gland tumorigenesis, systemic Gper deficiency resulted in reduced tumour size and metastasis compared with wild-type mice, consistent with a pro-tumorigenic role for GPER in vivo212.
In vitro, tamoxifen induces proliferation of tamoxifen-resistant MCF-7 cells through a GPER-dependent pathway210,213. This proliferation can be blocked by GPER knockdown or co-treatment with the GPER-selective antagonist G15 (refs. 69,210) as tamoxifen binds to and cross-activates GPER15,66,214. Breast cancer cell survival in the presence of tamoxifen might be mediated by Akt-induced inactivation of the pro-apoptotic transcription factor FOXO3, suggesting a mechanism to enhance eventual tamoxifen resistance23. Tamoxifen-mediated cross-activation of GPER also induces breast cancer cell migration215, potentially via the YAP–TAZ pathway37 (Fig. 1), and increases aromatase expression in tamoxifen-resistant (ERα-positive) cells216. In vivo, GPER also contributes to tamoxifen resistance in MCF-7 cells, with tamoxifen-resistant xenografts derived from MCF-7 cells regaining sensitivity to tamoxifen in female mice upon treatment with a combination of tamoxifen and G15, where neither alone had an effect210. GPER downregulation and G15 treatment also sensitize breast cancer cells to doxorubicin by inhibiting epithelial-to-mesenchymal transition217. Lastly, G-1 (as well as tamoxifen and fulvestrant) increases natural killer cell-mediated growth inhibition of both ERα-negative and ERα-positive breast cancer cells, suggesting a novel role for GPER in cancer therapy218.
Cancer-associated fibroblasts (CAFs) express GPER, with most studies to date employing breast CAFs, which have previously described roles supporting breast tumour progression18,219,220. In breast CAFs, GPER mediates expression of HIF1α and VEGF195 and has been implicated in promoting tumour progression by increasing migration and invasion221–223. Tamoxifen and G-1 induce increased aromatase expression in breast CAFs, resulting in increased oestrogen production219, potentially leading to tamoxifen resistance216.
The tumour microenvironment also contains adipocytes, particularly in adipose-rich tissues such as the breast. Obesity has been clinically established as an important contributor to multiple cancers224. Adipocytes not only express aromatase, resulting in intracrine oestrogen synthesis, but also adipokines and other (pro-inflammatory) cytokines and hormones that can promote tumorigenesis. The actions of GPER in reducing obesity and mitigating metabolic dysfunction168, inflammation194 and chemotherapy-associated cardiotoxicity225 could, in part, reduce the incidence of and improve outcomes in breast cancer and other cancers.
Malignant melanoma
Female patients with malignant melanoma have a better clinical outcome than male patients226, although ICIs, an effective treatment for melanoma, show better therapeutic efficacy in men than in women227. A role for GPER activity in melanoma was first suggested by the observation that GPER (but not ERα) mediates oestrogen-induced melanogenesis (melanocyte differentiation and melanin production)31,228. Treatment of mouse melanoma cells with G-1 or tamoxifen, interestingly, inhibits proliferation in vitro229. Combining ICIs (specifically an anti-PD1 antibody) with G-1 not only reduces tumour growth but also improves survival of melanoma-bearing female mice, far more than either anti-PD1 antibodies or G-1 treatment alone. Combination therapy utilizing immune checkpoint inhibition and G-1 can result in long-term clearance of tumours, indicating immunological memory207, with similar results in pancreatic cancer mouse xenograft models199. This effect is potentially mediated through lowering Myc levels, which results in decreased expression of PDL1 and increased expression of HLA class I in melanoma tumour cells, which together could lead to improved immune recognition of melanoma tumour cells207. In 2019, these results led to the initiation of the first Phase 1 clinical trial of G-1 for the treatment of malignant melanoma (NCT04130516)78.
Other forms of cancer
The type of cancer might determine whether GPER activity promotes or inhibits carcinogenesis and/or metastasis. Pharmacological activation of GPER reduces liver tumorigenesis, at least in part, through inhibiting inflammation and fibrosis208. In mouse models of non-small-cell lung cancer (urethane-induced adenocarcinoma), tumour burden increases following treatment with 17β-oestradiol or G-1, and decreases upon treatment with G15 (ref. 202), possibly with the involvement of NOTCH-dependent pathways230. GPER expression is increased in castration-resistant prostate cancer231, and its activation is associated with sustained cytotoxic ERK activation198. In a prostate cancer mouse xenograft model, chronic treatment with G-1 for several weeks inhibits cancer progression but only following cancer recurrence after castration231, suggesting the potential for GPER-targeted therapies in castration-resistant prostate cancer.
GPER expression and function have also been implicated in gastric epithelial metaplasia and gastric cancer173,174,232,233 as well as in colon cancer173,234. In mouse syngeneic pancreatic cancer xenograft models, G-1, alone or in combination with ICIs improves survival compared with vehicle only or ICIs alone, respectively, resulting in a substantial cure rate199. In line with the beneficial effects of G-1 on pancreatic cancer, tamoxifen, also acting as a GPER agonist, inhibits the recruitment and polarization of tumour-associated macrophages and interferes with myofibroblastic differentiation of pancreatic stellate cells in the tumour microenvironment235. This reduces the cells’ ability to remodel the extracellular matrix and to promote cancer cell invasion235. GPER is highly overexpressed in Waldenström macroglobulinaemia, yet G-1 treatment, both in vitro and in vivo, induces apoptosis of tumour cells, even in the protective bone marrow milieu236. In this study, G-1 treatment improved survival in a murine xenograft model but had no effect on B cells transplanted from healthy donors236.
Immune system and immunology
Regulation of fish granulocyte functions by oestrogens through GPER predates the evolutionary divergence of fish and tetrapods more than 450 million years ago, which indicates that oestrogens are modulators of the immune response and that GPER have played a pivotal role in immunity throughout evolution12. Sex plays an important role in immune responses with oestrogens frequently exerting anti-inflammatory effects, traditionally through ERα and, to a lesser extent, through ERβ237. However, 17β-oestradiol also mediates part of its anti-inflammatory effects through GPER, which is widely expressed in white blood cells, (including neutrophils, eosinophils, monocytes and lymphocytes) as well as in macrophages238.
Regulation of immune cells by GPER
GPER regulates apoptosis in eosinophils239, suggesting a role for GPER in allergic immune responses. Indeed, in a model of allergic pulmonary inflammation, G-1 attenuates airway hyper-responsiveness, reducing bronchoalveolar levels of inflammatory cells and the T helper 2 (TH2) cell cytokines IL-5 and IL-13, while increasing the frequency of splenic regulatory T cells (which produce the anti-inflammatory cytokine IL-10), thus establishing crosstalk between GPER and IL-10 (ref. 240). Moreover, G-1 treatment also promotes the formation of IL-10 in pro-inflammatory TH17 cells241,242. In macrophages, G-1 inhibits the production of lipopolysaccharide-induced cytokines, such as TNF and IL-6 (ref. 119), through the inhibition of NF-κB120, while also downregulating TLR4 expression243. Neutrophils show complex responses to G-1 in vitro, with G-1 treatment causing activation of human neutrophils244 and increased cell death-associated neutrophil extracellular trap formation245. In fish granulocytes, G-1 has multiple effects245, including suppression of ROS production12.
Regulation of inflammation by GPER
Deletion of Gper in mice increases circulating levels of pro-inflammatory cytokines, with a concomitant decrease in adiponectin levels compared with the wild type114,165. In a mouse model of diethylnitrosamine-induced liver cancer, deletion of Gper increases inflammation, fibrosis and tumorigenesis208. Consistent with this, GPER activation reduces expression of fibrosis markers in hepatic stellate cells in vitro, suggesting a possible role for GPER in counteracting liver inflammation and liver cancer208. In a mouse model of atherosclerosis, G-1 treatment reduces the increased number of CD68+ macrophages but not of CD3+ T cells, whereas deletion of Gper has the opposite effect126.
Modulation of GPER activity in immunity, inflammation and infection
In surgically postmenopausal mice with diet-induced obesity, chronic treatment with G-1 reduces levels of TNF, MCP1 and IL-6 as well as the expression of inflammatory genes in multiple metabolic tissues168. GPER may also play a role in inflammatory bowel diseases; in a model of Crohn’s disease, G-1 treatment reduces mortality, improves macroscopic and microscopic injury scores, and lowers C-reactive protein levels173,178. In a mouse model of Staphylococcus aureus skin and soft tissue infection, G-1 reduces dermonecrosis and increases bacterial clearance, indicating a role of GPER for the innate immune system246,247. These effects are more pronounced in females, suggesting a sex-specific response, and are absent in Gper-deficient mice, confirming the selectivity of G-1 for its target GPER247.
Clinical data suggest a sex bias in COVID-19 severity following SARS-CoV-2 infection, with men exhibiting increased hospitalization and mortality compared with women. A role for GPER in this sex bias is suggested based on experimental models of both overexpression of GPER and treatment with G-1, each of which (similar to 17β-oestradiol treatment) leads to reduced SARS-CoV-2 viral load in infected bronchial cells in vitro compared with uninfected cells. These reductions in viral load caused by 17β-oestradiol and G-1 treatment are reversed by treatment with G15 (ref. 248). GPER activation also results in anti-inflammatory immune responses in numerous neurological diseases249–251. Lastly, in a genome-wide CRISPR–Cas9 screen, GPER was identified as a downregulator of type I interferon252. GPER expression during pregnancy is both necessary and sufficient to suppress IFNγ signalling, which is elevated in reproductive and fetal tissues in influenza A virus-infected female mice. During virus-induced maternal inflammation, blocking GPER with G15 delays fetal development and promotes fetal demise compared with vehicle-treated mice252. Thus, GPER expression and activity are required to protect the fetus during maternal infection. Taken together, pharmacological activation of GPER holds promise for the treatment of diseases and conditions that are associated with activation of inflammation (due to infectious pathogens such as bacteria or viruses) and of conditions associated with an abnormal immune response.
Ageing and neurological diseases
Cardiovascular and renal ageing
Physiological ageing is an unmodifiable risk factor for arterial hypertension, myocardial disease and atherosclerotic vascular disease. In addition, vascular ageing is further accelerated by modifiable risk factors, including obesity (which is often associated with hypertension and diabetes) and smoking89. Endogenous Gper expression is associated with suppresion of the age-dependent increases in endothelin ETB receptors, and endothelin-converting enzyme-2 in the heart253. Moreover, Gper deficiency abrogates age-dependent impairment of vasodilatation by interfering with NOX1-dependent ROS formation, specifically by reducing NOX1 expression, which is induced by GPER39,103. Accordingly, Gper deficiency prevents ageing-induced myocardial fibrosis and the associated development of diastolic heart failure (HFpEF) and for the most part prevents angiotensin-induced hypertension39. In addition, Gper deficiency is associated with a supression of development of age-dependent CKD due to FSGS144. The effect of Gper deficiency could be partly recapitulated pharmacologically by reducing NOX1 abundance and the associated production of ROS with G36, the first NOX1 downregulator39. Thus, blocking the GPER–NOX1 axis holds therapeutic opportunities for ageing-associated non-communicable diseases, including arterial hypertension.
Neurological diseases
In premenopausal women, endogenous oestrogens protect against stroke and dementia254. GPER, like ERα and ERβ, regulates arterial tone of the cerebral vasculature255. Antisense oligonucleotide knockdown of Gper in vivo largely abrogates the protective effects of oestrogen on cerebral ischaemia256, whereas activation of GPER with G-1 reduces reperfusion injury following cerebral ischaemia in both male and female mice257,258. This involves inhibition of both apoptosis259 and inflammatory pathways, such as TLR4 (ref. 258), with concomitant activation of anti-inflammatory pathways260. GPER-dependent protective effects have been demonstrated in rodent models of ischaemic261 and haemorrhagic stroke262. G-1-dependent protection from ischaemic stroke is completely abrogated by systemic deletion of Gper, while only partial protection was observed in animals with astrocyte- or neuronal cell-specific Gper deletion261. Activation of GPER by G-1 also attenuates blood–brain barrier injury263 and improves immunoprotection following stroke264. GPER also might play a role in psychiatric disorders such as anxiety265, depression266 and addiction267. Systemic deletion of Gper increases anxiety in rats265; accordingly, activation of GPER by G-1 has anxiolytic and also antidepressant effects in rodents69,266. Finally, deletion of Gper or GPER antagonism enhances morphine analgesia and reduces pain involving µ-type opioid receptors, suggesting the potential of GPER blockade for the treatment of pain, substance addiction, and opioid tolerance268.
Ageing is the main risk factor for Parkinson disease and Alzheimer disease as well as for vascular dementia. Studies in neurotoxic mouse models of Parkinson disease have shown that 17β-oestradiol-dependent, ERα-mediated protective effects on dopaminergic neurons require crosstalk with GPER and that GPER also has independent protective effects against Parkinson disease267. In a mouse model of Parkinson disease, G-1 treatment reduces the release of pro-inflammatory cytokines251 and also mediates part of the neuroprotective effects of IGF1 on dopaminergic neuronal injury269. G-1 treatment also reduces microglial activation and decreases pro-inflammatory cytokine production251. GPER is important for maintaining long-term memory, and G-1 enhances object recognition and long-term memory in male mice270. Accordingly, in a mouse model of Alzheimer disease and after traumatic brain injury in rats, improvements in neuropsychological functions are observed upon G-1 treatment271–273. GPER also mediates the anti-inflammatory effects of genistein in microglia250.
Elevated levels of 17β-oestradiol present in pregnant women are associated with reduced severity of multiple sclerosis274, and 17β-oestradiol supplementation reduces symptom severity and immune infiltration in a mouse model of MS (experimental autoimmune encephalomyelitis) in mice of both sexes275. In this model, female Gper-deficient mice exhibit reduced 17β-oestradiol-mediated protection against multiple sclerosis disease severity and reduced protective effects of 17β-oestradiol on white matter damage compared with wild-type mice119,249,276. Conversely, GPER activation by G-1 reduces multiple sclerosis severity, an effect absent in female Gper-deficient mice. Mechanistically, G-1 reduced inflammatory cytokine production in macrophages and upregulated PD1 to enhance the activity of T regulatory cells249.
Conclusions
Progress made in the past decade in the field of GPER has broadened our understanding of the multiple functions of this receptor at the cell, tissue and organismal level, including in humans. Widely expressed, GPER mediates both rapid and genomic effects in all main organs, being involved in multiple aspects of health and disease (Fig. 3). In addition to oestrogens, many natural and synthetic molecules target GPER, either as selective or combined oestrogen receptor agonists or antagonists. Importantly, clinically approved ERα antagonists, such as the SERMS tamoxifen and raloxifene or the SERD fulvestrant, licensed for the treatment of breast cancer209, show agonistic activity towards GPER13,14,66. Diverse molecules present in plants (such as genistein, daidzein and green tea polyphenols) and EDCs also activate GPER; further study is required to determine how their effects on health or disease involve GPER. Utilizing GPER expression as a diagnostic marker in tissues or in circulating cells provides new opportunities to further characterize pathological conditions at different stages during disease progression or even before diseases develop. Targeting GPER pharmacologically could provide new opportunities to treat diseases for which no or only a few effective therapies exist (such as malignant melanoma and other cancers), including inhibition of the constitutive inducing effect of GPER on NOX1 activity. Clinical studies that should also consider sex, genetics and hormonal status are needed to determine whether utilizing or targeting GPER could improve diagnosis, prognosis, therapy and the clinical course of human diseases and thus overall health82.
Acknowledgements
E.R.P. is supported by grants from the US National Institutes of Health (R01 CA163890 and R01 CA194496), from Dialysis Clinic, Inc., and by the UNM Comprehensive Cancer Center (NIH P30 CA118100) and the Autophagy, Inflammation and Metabolism (AIM) Center of Biomedical Research Excellence (CoBRE, NIH P20 GM121176). M.B. is supported by grants 108 258 and 122 504 from the Swiss National Science Foundation (SNSF).
Author contributions
Both authors contributed equally to all aspects of this manuscript.
Peer review
Peer review information
Nature Reviews Endocrinology thanks Guichun Han and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Competing interests
M.B. and E.R.P. are inventors on U.S. patent Nos. 10,251,870, 10,682,341 and 10,980,785, and E.R.P. is an inventor on U.S. Patent Nos. 10,471,047 and 10,561,648, all for the therapeutic use of compounds targeting GPER (“Method for treating obesity, diabetes, cardiovascular and kidney diseases by regulating GPR30/GPER”). E.R.P. is an inventor on U.S. Patent Nos. 7,875,721 and 8,487,100 for GPER-selective ligands and imaging agents (“Compounds for binding to ERα/β and GPR30, methods of treating disease states and conditions mediated through these receptors and identification thereof”). M.B. has served or serves as a consultant to Abbott, Inc., Abbvie, Inc., Travere, Inc. and Pharmazz, Inc.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Review criteria
Abstracts of all articles published on GPR30 or GPER, published between January 1996 and February 2023, were retrieved from the U.S. National Library of Medicine (PubMed.gov). Articles were assessed for relevance, importance and scientific rigour, with a focus on publication in the past 10 years. The authors apologize to their colleagues whose work could not be included due to space and reference restrictions.
==== Refs
References
1. Aristotélēs. Historia Animalium Books 1–10 (transl. and ed. Peck, A. L., Balme, D. M. & Gotthelf, A.) 1–624 (Harvard Univ. Press, 1965).
2. Starling EH Croonian lecture: on the chemical correlation of the functions of the body Lancet 1905 2 579 583
3. Allen E Doisy EA An ovarian hormone: preliminary report on its localization, extraction and partial purification, and action in test animals JAMA 1923 81 819 821 10.1001/jama.1923.02650100027012
4. Butenandt A Über “Progynon”, ein krystallisiertes weibliches Sexualhormon Die Naturwissenschaften 1929 17 879 10.1007/BF01506919
5. Jensen E A conversation with Elwood Jensen. Interview by David D. Moore Annu. Rev. Physiol. 2012 74 1 11 10.1146/annurev-physiol-020911-153327 21888507
6. Soloff MS Szego CM Purification of estradiol receptor from rat uterus and blockade of its estrogen-binding function by specific antibody Biochem. Biophys. Res. Commun. 1969 34 141 147 10.1016/0006-291X(69)90540-3 4178559
7. Barton M Twenty years of the G protein-coupled estrogen receptor GPER: historical and personal perspectives J. Steroid Biochem. Mol. Biol. 2018 176 4 15 10.1016/j.jsbmb.2017.03.021 28347854
8. Walter P Cloning of the human estrogen receptor cDNA Proc. Natl Acad. Sci. USA 1985 82 7889 7893 10.1073/pnas.82.23.7889 3865204
9. Kuiper GG Enmark E Pelto-Huikko M Nilsson S Gustafsson JA Cloning of a novel receptor expressed in rat prostate and ovary Proc. Natl Acad. Sci. USA 1996 93 5925 5930 10.1073/pnas.93.12.5925 8650195
10. Mosselman S Polman J Dijkema R ERβ: identification and characterization of a novel human estrogen receptor FEBS Lett. 1996 392 49 53 10.1016/0014-5793(96)00782-X 8769313
11. Prossnitz ER Barton M The G-protein-coupled estrogen receptor GPER in health and disease Nat. Rev. Endocrinol. 2011 7 715 726 10.1038/nrendo.2011.122 21844907
12. Cabas I Chaves-Pozo E Mulero V Garcia-Ayala A Role of estrogens in fish immunity with special emphasis on GPER1 Dev. Comp. Immunol. 2018 89 102 110 10.1016/j.dci.2018.08.001 30092317
13. Filardo EJ Quinn JA Bland KI Frackelton AR Jr. Estrogen-induced activation of Erk-1 and Erk-2 requires the G protein-coupled receptor homolog, GPR30, and occurs via trans-activation of the epidermal growth factor receptor through release of HB-EGF Mol. Endocrinol. 2000 14 1649 1660 10.1210/mend.14.10.0532 11043579
14. Revankar CM Cimino DF Sklar LA Arterburn JB Prossnitz ER A transmembrane intracellular estrogen receptor mediates rapid cell signaling Science 2005 307 1625 1630 10.1126/science.1106943 15705806
15. Thomas P Pang Y Filardo EJ Dong J Identity of an oestrogen membrane receptor coupled to a G protein in human breast cancer cells Endocrinology 2005 146 624 632 10.1210/en.2004-1064 15539556
16. Alexander SPH Mathie A Peters JA Guide to Receptors and Channels (GRAC), 3rd edn. Br. J. Pharmacol. 2008 153 Suppl. 2 1 209 10.1038/sj.bjp.0707746 17952110
17. Gaudet HM Cheng SB Christensen EM Filardo EJ The G-protein coupled estrogen receptor, GPER: the inside and inside-out story Mol. Cell Endocrinol. 2015 418 207 219 10.1016/j.mce.2015.07.016 26190834
18. Madeo A Maggiolini M Nuclear alternate estrogen receptor GPR30 mediates 17β-estradiol-induced gene expression and migration in breast cancer-associated fibroblasts Cancer Res. 2010 70 6036 6046 10.1158/0008-5472.CAN-10-0408 20551055
19. Filardo EJ Quinn JA Frackelton AR Jr. Bland KI Estrogen action via the G protein-coupled receptor, GPR30: stimulation of adenylyl cyclase and cAMP-mediated attenuation of the epidermal growth factor receptor-to-MAPK signaling axis Mol. Endocrinol. 2002 16 70 84 10.1210/mend.16.1.0758 11773440
20. Lucas TF Royer C Siu ER Lazari MF Porto CS Expression and signaling of G protein-coupled estrogen receptor 1 (GPER) in rat Sertoli cells Biol. Reprod. 2010 83 307 317 10.1095/biolreprod.110.084160 20445128
21. Deng Q GPER/Hippo-YAP signal is involved in bisphenol S induced migration of triple negative breast cancer (TNBC) cells J. Hazard. Mater. 2018 355 1 9 10.1016/j.jhazmat.2018.05.013 29758456
22. Daub H Weiss FU Wallasch C Ullrich A Role of transactivation of the EGF receptor in signalling by G-protein-coupled receptors Nature 1996 379 557 560 10.1038/379557a0 8596637
23. Zekas E Prossnitz ER Estrogen-mediated inactivation of FOXO3a by the G protein-coupled estrogen receptor GPER BMC Cancer 2015 15 702 10.1186/s12885-015-1699-6 26470790
24. Fredette NC Meyer MR Prossnitz ER Role of GPER in estrogen-dependent nitric oxide formation and vasodilation J. Steroid Biochem. Mol. Biol. 2018 176 65 72 10.1016/j.jsbmb.2017.05.006 28529128
25. Peixoto P Aires RD Lemos VS Bissoli NS Santos RLD GPER agonist dilates mesenteric arteries via PI3K-Akt-eNOS and potassium channels in both sexes Life Sci. 2017 183 21 27 10.1016/j.lfs.2017.06.020 28645860
26. Sharma G Prossnitz ER Mechanisms of estradiol-induced insulin secretion by the G protein-coupled estrogen receptor GPR30/GPER in pancreatic β-cells Endocrinology 2011 152 3030 3039 10.1210/en.2011-0091 21673097
27. Greenlee MM Estradiol activates epithelial sodium channels in rat alveolar cells through the G protein-coupled estrogen receptor Am. J. Physiol. Lung Cell Mol. Physiol. 2013 305 L878 889 10.1152/ajplung.00008.2013 24097558
28. Evanson KW Goldsmith JA Ghosh P Delp MD The G protein-coupled estrogen receptor agonist, G-1, attenuates BK channel activation in cerebral arterial smooth muscle cells Pharmacol. Res. Perspect. 2018 6 e00409 10.1002/prp2.409 29938113
29. Yue J Activation of G-protein-coupled receptor 30 protects neurons against excitotoxicity through inhibiting excessive autophagy induced by glutamate ACS Chem. Neurosci. 2019 10 4227 4236 10.1021/acschemneuro.9b00287 31545891
30. Zhang H Mechanisms of estradiol-induced EGF-like factor expression and oocyte maturation via G protein-coupled estrogen receptor Endocrinology 2020 161 bqaa190 10.1210/endocr/bqaa190 33068422
31. Sun M G protein-coupled estrogen receptor enhances melanogenesis via cAMP-protein kinase (PKA) by upregulating microphthalmia-related transcription factor-tyrosinase in melanoma J. Steroid Biochem. Mol. Biol. 2017 165 236 246 10.1016/j.jsbmb.2016.06.012 27378491
32. Pandey DP Estrogenic GPR30 signalling induces proliferation and migration of breast cancer cells through CTGF EMBO J. 2009 28 523 532 10.1038/emboj.2008.304 19153601
33. Cirillo F GPER is involved in the regulation of the estrogen-metabolizing cyp1b1 enzyme in breast cancer Oncotarget 2017 8 106608 106624 10.18632/oncotarget.22541 29290975
34. Zhu P GPER/ERK&AKT/NF-κB pathway is involved in cadmium-induced proliferation, invasion and migration of GPER-positive thyroid cancer cells Mol. Cell Endocrinol. 2017 442 68 80 10.1016/j.mce.2016.12.007 27940299
35. Chen ZJ Activation of GPER suppresses epithelial mesenchymal transition of triple negative breast cancer cells via NF-κB signals Mol. Oncol. 2016 10 775 788 10.1016/j.molonc.2016.01.002 26842883
36. De Francesco EM Maggiolini M Musti AM Crosstalk between Notch, HIF-1α and GPER in breast cancer EMT Int. J. Mol. Sci. 2018 19 2011 10.3390/ijms19072011 29996493
37. Zhou X Estrogen regulates Hippo signaling via GPER in breast cancer J. Clin. Invest. 2015 125 2123 2135 10.1172/JCI79573 25893606
38. Zhu L MicroRNA-2861 and microRNA-5115 regulates myocardial ischemia-reperfusion injury through the GPR30/mTOR signaling pathway by binding to GPR30 J. Cell Physiol. 2020 235 7791 7802 10.1002/jcp.29427 31930508
39. Meyer MR Obligatory role for GPER in cardiovascular aging and disease Sci. Signal. 2016 9 ra105 10.1126/scisignal.aag0240 27803283
40. Barton M Meyer MR Prossnitz ER Nox1 downregulators: a new class of therapeutics Steroids 2019 152 108494 10.1016/j.steroids.2019.108494 31518594
41. Jensen EV Jordan VC The estrogen receptor: a model for molecular medicine Clin. Cancer Res. 2003 9 1980 1989 12796359
42. Cheng SB Anatomical location and redistribution of G protein-coupled estrogen receptor-1 during the estrus cycle in mouse kidney and specific binding to estrogens but not aldosterone Mol. Cell Endocrinol. 2014 382 950 959 10.1016/j.mce.2013.11.005 24239983
43. Rigiracciolo DC GPER is involved in the stimulatory effects of aldosterone in breast cancer cells and breast tumor-derived endothelial cells Oncotarget 2016 7 94 111 10.18632/oncotarget.6475 26646587
44. Lu AS Rouhimoghadam M Arnatt CK Filardo EJ Salem AK Proteolytic targeting chimeras with specificity for plasma membrane and intracellular estrogen receptors Mol. Pharm. 2021 18 1455 1469 10.1021/acs.molpharmaceut.1c00018 33600191
45. Koganti S Snyder R Gumaste U Karamyan VT Thekkumkara T 2-Methoxyestradiol binding of GPR30 down-regulates angiotensin AT(1) receptor Eur. J. Pharmacol. 2014 723 131 140 10.1016/j.ejphar.2013.10.064 24262995
46. Zucchetti AE G-protein-coupled receptor 30/adenylyl cyclase/protein kinase A pathway is involved in estradiol 17ss-D-glucuronide-induced cholestasis Hepatology 2014 59 1016 1029 10.1002/hep.26752 24115158
47. Chourasia TK Pang Y Thomas P The catecholestrogen, 2-hydroxyestradiol-17beta, acts as a G protein-coupled estrogen receptor 1 (GPER/GPR30) antagonist to promote the resumption of meiosis in zebrafish oocytes Biol. Reprod. 2015 92 69 1 13
48. Teng Y Dehydroepiandrosterone activation of G-protein-coupled estrogen receptor rapidly stimulates MicroRNA-21 transcription in human hepatocellular carcinoma cells J. Biol. Chem. 2015 290 15799 15811 10.1074/jbc.M115.641167 25969534
49. Cao J Lu M Yan W Li L Ma H Dehydroepiandrosterone alleviates intestinal inflammatory damage via GPR30-mediated nrf2 activation and NLRP3 inflammasome inhibition in colitis mice Free Radic. Biol. Med. 2021 172 386 402 10.1016/j.freeradbiomed.2021.06.025 34182071
50. Sandra N Ester P Marie-Agnes P Robert M Olivier H The DHEA metabolite 7β-hydroxy-epiandrosterone exerts anti-estrogenic effects on breast cancer cell lines Steroids 2012 77 542 551 10.1016/j.steroids.2012.01.019 22342541
51. Avena P 27-Hydroxycholesterol binds GPER and induces progression of estrogen receptor-negative breast cancer Cancers 2022 14 1521 10.3390/cancers14061521 35326671
52. Diamanti-Kandarakis E Endocrine-disrupting chemicals: an Endocrine Society scientific statement Endocr. Rev. 2009 30 293 342 10.1210/er.2009-0002 19502515
53. Lorand T Vigh E Garai J Hormonal action of plant derived and anthropogenic non-steroidal estrogenic compounds: phytoestrogens and xenoestrogens Curr. Med. Chem. 2010 17 3542 3574 10.2174/092986710792927813 20738246
54. Thomas P Dong J Binding and activation of the seven-transmembrane estrogen receptor GPR30 by environmental estrogens: a potential novel mechanism of endocrine disruption J. Steroid Biochem. Mol. Biol. 2006 102 175 179 10.1016/j.jsbmb.2006.09.017 17088055
55. Zhou C Li P Han M Gao X Daidzein stimulates fatty acid-induced fat deposition in C2C12 myoblast cells via the G protein-coupled receptor 30 pathway Anim. Biotechnol. 2020 33 851 863 10.1080/10495398.2020.1842749 33164657
56. Moriyama M S-equol, a major isoflavone from soybean, inhibits nitric oxide production in lipopolysaccharide-stimulated rat astrocytes partially via the GPR30-mediated pathway Int. J. Inflam. 2018 2018 8496973 29692883
57. Maggiolini M The G protein-coupled receptor GPR30 mediates c-fos up-regulation by 17β-estradiol and phytoestrogens in breast cancer cells J. Biol. Chem. 2004 279 27008 27016 10.1074/jbc.M403588200 15090535
58. Dong WH Chen JC He YL Xu JJ Mei YA Resveratrol inhibits K(v)2.2 currents through the estrogen receptor GPR30-mediated PKC pathway Am. J. Physiol. Cell Physiol. 2013 305 C547 557 10.1152/ajpcell.00146.2013 23804203
59. Chimento A Oleuropein and hydroxytyrosol activate GPER/ GPR30-dependent pathways leading to apoptosis of ER-negative SKBR3 breast cancer cells Mol. Nutr. Food Res. 2013 58 478 489 10.1002/mnfr.201300323 24019118
60. Li YC Ding XS Li HM Zhang C Icariin attenuates high glucose-induced type IV collagen and fibronectin accumulation in glomerular mesangial cells by inhibiting transforming growth factor-β production and signalling through G protein-coupled oestrogen receptor 1 Clin. Exp. Pharmacol. Physiol. 2013 40 635 643 10.1111/1440-1681.12143 23772748
61. Moreno-Ulloa A (-)-Epicatechin stimulates mitochondrial biogenesis and cell growth in C2C12 myotubes via the G-protein coupled estrogen receptor Eur. J. Pharmacol. 2018 822 95 107 10.1016/j.ejphar.2018.01.014 29355558
62. Lo EK Lee JC Turner PC El-Nezami H Low dose of zearalenone elevated colon cancer cell growth through G protein-coupled estrogenic receptor Sci. Rep. 2021 11 7403 10.1038/s41598-021-86788-w 33795755
63. Paterni I Granchi C Katzenellenbogen JA Minutolo F Estrogen receptors alpha (ERα) and beta (ERβ): subtype-selective ligands and clinical potential Steroids 2014 90 13 29 10.1016/j.steroids.2014.06.012 24971815
64. Stauffer SR Pyrazole ligands: structure-affinity/activity relationships and estrogen receptor-α-selective agonists J. Med. Chem. 2000 43 4934 4947 10.1021/jm000170m 11150164
65. Meyers MJ Estrogen receptor-β potency-selective ligands: structure-activity relationship studies of diarylpropionitriles and their acetylene and polar analogues J. Med. Chem. 2001 44 4230 4251 10.1021/jm010254a 11708925
66. Petrie WK G protein-coupled estrogen receptor-selective ligands modulate endometrial tumor growth Obstet. Gynecol. Int. 2013 2013 472720 10.1155/2013/472720 24379833
67. Bologa CG Virtual and biomolecular screening converge on a selective agonist for GPR30 Nat. Chem. Biol. 2006 2 207 212 10.1038/nchembio775 16520733
68. Dennis MK Identification of a GPER/GPR30 antagonist with improved estrogen receptor counterselectivity J. Steroid Biochem. Mol. Biol. 2011 127 358 366 10.1016/j.jsbmb.2011.07.002 21782022
69. Dennis MK In vivo effects of a GPR30 antagonist Nat. Chem. Biol. 2009 5 421 427 10.1038/nchembio.168 19430488
70. Urban N Leonhardt M Schaefer M Multiplex GPCR screen reveals reliably acting agonists and a Gq-phospholipase C coupling mode of GPR30/GPER1 Mol. Pharmacol. 2023 103 48 62 10.1124/molpharm.122.000580 36400433
71. Tutzauer J Ligand-independent G protein-coupled estrogen receptor/G protein-coupled receptor 30 activity: lack of receptor-dependent effects of G-1 and 17β-estradiol Mol. Pharmacol. 2021 100 271 282 10.1124/molpharm.121.000259 34330822
72. Lappano R Two novel GPER agonists induce gene expression changes and growth effects in cancer cells Curr. Cancer Drug Targets 2012 12 531 542 10.2174/156800912800673284 22414008
73. O’Dea A Sondergard C Sweeney P Arnatt CK A series of indole-thiazole derivatives act as GPER agonists and inhibit breast cancer cell growth ACS Med. Chem. Lett. 2018 9 901 906 10.1021/acsmedchemlett.8b00212 30258538
74. DeLeon C A novel GPER antagonist protects against the formation of estrogen-induced cholesterol gallstones in female mice J. Lipid Res. 2020 61 767 777 10.1194/jlr.RA119000592 32127396
75. Lappano R MIBE acts as antagonist ligand of both estrogen receptor α and GPER in breast cancer cells Breast Cancer Res. 2012 14 R12 10.1186/bcr3096 22251451
76. Bargagna-Mohan P Baek SH Lee H Kim K Mohan R Use of PROTACS as molecular probes of angiogenesis Bioorg. Med. Chem. Lett. 2005 15 2724 2727 10.1016/j.bmcl.2005.04.008 15876533
77. Revankar CM A selective ligand for estrogen receptor proteins discriminates rapid and genomic signaling Cell. Chem. Biol. 2019 26 1692 1702 10.1016/j.chembiol.2019.10.009 31706983
78. US National Library of Medicine. ClinicalTrials.gov https://clinicaltrials.gov/ct2/show/NCT04130516 (2019).
79. Muller C Phase 1 trial of a novel, first-in-class G protein-coupled estrogen receptor (GPER) agonist, LNS8801 in patients with advanced or recurrent treatment-refractory solid malignancies J. Clin. Oncol. 2021 39 3084 10.1200/JCO.2021.39.15_suppl.3084
80. Muller C Phase 1b study of the novel first-in-class G protein-coupled estrogen receptor (GPER) agonist, LNS8801 in combination with pembrolizumab in patients with immune checkpoint inhibitor (ICI)-relapsed and refractory solid malignancies and dose escalation update J. Clin. Oncol. 2022 40 2574 10.1200/JCO.2022.40.16_suppl.2574
81. Prossnitz ER Hathaway HJ What have we learned about GPER function in physiology and disease from knockout mice? J. Steroid Biochem. Mol. Biol. 2015 153 114 126 10.1016/j.jsbmb.2015.06.014 26189910
82. Mielke MM Miller VM Improving clinical outcomes through attention to sex and hormones in research Nat. Rev. Endocrinol. 2021 17 625 635 10.1038/s41574-021-00531-z 34316045
83. Mauvais-Jarvis F Sex and gender: modifiers of health, disease, and medicine Lancet 2020 396 565 582 10.1016/S0140-6736(20)31561-0 32828189
84. Lafferty AR A novel genetic locus for low renin hypertension: familial hyperaldosteronism type II maps to chromosome 7 (7p22) J. Med. Genet. 2000 37 831 835 10.1136/jmg.37.11.831 11073536
85. Feldman RD A common hypofunctional genetic variant of GPER is associated with increased blood pressure in women Br. J. Clin. Pharmacol. 2014 78 1441 1452 10.1111/bcp.12471 25039431
86. Fredette NC Malik E Mukhtar ML Prossnitz ER Terada N A hypertension patient derived induced pluripotent stem cell model demonstrates a role for GPER in hypertension risk and development Am. J. Physiol. Cell. Physiol. 2020 319 C825 C838 10.1152/ajpcell.00350.2019 32783656
87. Hussain Y G-protein estrogen receptor as a regulator of low-density lipoprotein cholesterol metabolism: cellular and population genetic studies Arterioscler. Thromb. Vasc. Biol. 2015 35 213 221 10.1161/ATVBAHA.114.304326 25395619
88. Pupo M A genetic polymorphism repurposes the G-protein coupled and membrane-associated estrogen receptor GPER to a transcription factor-like molecule promoting paracrine signaling between stroma and breast carcinoma cells Oncotarget 2017 8 46728 46744 10.18632/oncotarget.18156 28596490
89. Meyer MR Barton M Estrogens and coronary artery disease: new clinical perspectives Adv. Pharmacol. 2016 77 307 360 10.1016/bs.apha.2016.05.003 27451102
90. Barton M Meyer MR Heart failure with preserved ejection fraction in women: new clues to causes and treatment JACC Basic Transl. Sci. 2020 5 296 299 10.1016/j.jacbts.2020.02.001 32215378
91. Haas E Regulatory role of G protein-coupled estrogen receptor for vascular function and obesity Circ. Res. 2009 104 288 291 10.1161/CIRCRESAHA.108.190892 19179659
92. Meyer MR Baretella O Prossnitz ER Barton M Dilation of epicardial coronary arteries by the G protein-coupled estrogen receptor agonists G-1 and ICI 182,780 Pharmacology 2010 86 58 64 10.1159/000315497 20639684
93. Arefin S Vasodilatory effects of the selective GPER agonist G-1 is maximal in arteries of postmenopausal women Maturitas 2014 78 123 130 10.1016/j.maturitas.2014.04.002 24796498
94. Yu X Activation of G protein-coupled estrogen receptor induces endothelium-independent relaxation of coronary artery smooth muscle Am. J. Physiol. Endocrinol. Metab. 2011 301 E882 888 10.1152/ajpendo.00037.2011 21791623
95. Lindsey SH Liu L Chappell MC Vasodilation by GPER in mesenteric arteries involves both endothelial nitric oxide and smooth muscle cAMP signaling Steroids 2014 81 99 102 10.1016/j.steroids.2013.10.017 24246735
96. Pang Y Thomas P Additive effects of low concentrations of estradiol-17β and progesterone on nitric oxide production by human vascular endothelial cells through shared signaling pathways J. Steroid Biochem. Mol. Biol. 2017 165 258 267 10.1016/j.jsbmb.2016.06.014 27369115
97. Yu X Stallone JN Heaps CL Han G The activation of G protein-coupled estrogen receptor induces relaxation via cAMP as well as potentiates contraction via EGFR transactivation in porcine coronary arteries PLoS ONE 2018 13 e0191418 10.1371/journal.pone.0191418 29360846
98. Yu X Activation of G protein-coupled estrogen receptor 1 induces coronary artery relaxation via Epac/Rap1-mediated inhibition of RhoA/Rho kinase pathway in parallel with PKA PLoS ONE 2017 12 e0173085 10.1371/journal.pone.0173085 28278256
99. Meyer MR Fredette NC Barton M Prossnitz ER G protein-coupled estrogen receptor inhibits vascular prostanoid production and activity J. Endocrinol. 2015 227 61 69 10.1530/JOE-15-0257 26303299
100. Meyer MR Field AS Kanagy NL Barton M Prossnitz ER GPER regulates endothelin-dependent vascular tone and intracellular calcium Life Sci. 2012 91 623 627 10.1016/j.lfs.2012.01.007 22326502
101. Tropea T Pregnancy augments G protein estrogen receptor (GPER) induced vasodilation in rat uterine arteries via the nitric oxide–cGMP signaling pathway PLoS ONE 2015 10 e0141997 10.1371/journal.pone.0141997 26536245
102. Gurrala R Alterations in the estrogen receptor profile of cardiovascular tissues during aging Geroscience 2021 43 433 442 10.1007/s11357-021-00331-3 33558965
103. Meyer MR Rosemann T Barton M Prossnitz ER GPER mediates functional endothelial aging in renal arteries Pharmacology 2017 100 188 193 10.1159/000478732 28704834
104. Davis GK Newsome AD Cole AB Ojeda NB Alexander BT Chronic estrogen supplementation prevents the increase in blood pressure in female intrauterine growth-restricted offspring at 12 months of age Hypertension 2019 73 1128 1136 10.1161/HYPERTENSIONAHA.118.12379 30929518
105. Luo P Stress-related arterial hypertension in Gper-deficient rats Acta Physiol. Sin. 2017 69 532 540
106. Waghulde H Attenuation of microbiotal dysbiosis and hypertension in a CRISPR/Cas9 gene ablation rat model of GPER1 Hypertension 2018 72 1125 1132 10.1161/HYPERTENSIONAHA.118.11175 30354811
107. Gohar EY Evidence for G-protein-coupled estrogen receptor as a pronatriuretic factor J. Am. Heart Assoc. 2020 9 e015110 10.1161/JAHA.119.015110 32390531
108. Barton M Meyer MR Nicolaus Copernicus and the rapid vascular responses to aldosterone Trends Endocrinol. Metab. 2015 26 396 398 10.1016/j.tem.2015.05.005 26088671
109. Wehling M Rapid actions of aldosterone revisited: receptors in the limelight J. Steroid Biochem. Mol. Biol. 2018 176 94 98 10.1016/j.jsbmb.2017.01.016 28126566
110. Dinh QN Aldosterone-induced hypertension is sex-dependent, mediated by T cells and sensitive to GPER activation Cardiovasc. Res. 2021 117 960 970 10.1093/cvr/cvaa075 32215568
111. Caroccia B Aldosterone stimulates its biosynthesis via a novel GPER-mediated mechanism J. Clin. Endocrinol. Metab. 2019 104 6316 6324 10.1210/jc.2019-00043 31125081
112. Wang D Wang M Sun P Gao Q Eplerenone inhibits oxidized low-density lipoprotein-induced proliferation and migration of vascular smooth muscle cells by downregulating GPER expression Adv. Clin. Exp. Med. 2021 30 405 412 10.17219/acem/133484 33915037
113. Cheng L Rapid aldosterone-mediated signaling in the DCT increases activity of the thiazide-sensitive NaCl cotransporter J. Am. Soc. Nephrol. 2019 30 1454 1470 10.1681/ASN.2018101025 31253651
114. Sharma G GPER deficiency in male mice results in insulin resistance, dyslipidemia, and a proinflammatory state Endocrinology 2013 154 4136 4145 10.1210/en.2013-1357 23970785
115. Meoli L Sex- and age-dependent effects of Gpr30 genetic deletion on the metabolic and cardiovascular profiles of diet-induced obese mice Gene 2014 540 210 216 10.1016/j.gene.2014.02.036 24582972
116. Ghaffari S Naderi Nabi F Sugiyama MG Lee WL Estrogen inhibits LDL (low-density lipoprotein) transcytosis by human coronary artery endothelial cells via GPER (G-protein-coupled estrogen receptor) and SR-BI (scavenger receptor class B type 1) Arterioscler. Thromb. Vasc. Biol. 2018 38 2283 2294 10.1161/ATVBAHA.118.310792 30354216
117. Jafarynezhad F The G-protein-coupled estrogen receptor agonist prevents cardiac lipid accumulation by stimulating cardiac peroxisome proliferator-activated receptor alpha: a preclinical study in ovariectomized-diabetic rat model Int. J. Endocrinol. Metab. 2022 20 e123560 10.5812/ijem-123560 36407026
118. Locher R Emmanuele L Suter PM Vetter W Barton M Green tea polyphenols inhibit human vascular smooth muscle cell proliferation stimulated by native low-density lipoprotein Eur. J. Pharmacol. 2002 434 1 7 10.1016/S0014-2999(01)01535-7 11755158
119. Blasko E Beneficial role of the GPR30 agonist G-1 in an animal model of multiple sclerosis J. Neuroimmunol. 2009 214 67 77 10.1016/j.jneuroim.2009.06.023 19664827
120. Pelekanou V Estrogen anti-inflammatory activity on human monocytes is mediated through cross-talk between estrogen receptor ERα36 and GPR30/GPER1 J. Leukoc. Biol. 2016 99 333 347 10.1189/jlb.3A0914-430RR 26394816
121. Haas E Differential effects of 17β-estradiol on function and expression of estrogen receptor alpha, estrogen receptor beta, and GPR30 in arteries and veins of patients with atherosclerosis Hypertension 2007 49 1358 1363 10.1161/HYPERTENSIONAHA.107.089995 17452498
122. Li F Activation of GPER induces differentiation and inhibition of coronary artery smooth muscle cell proliferation PLoS ONE 2013 8 e64771 10.1371/journal.pone.0064771 23840305
123. Jehle J G protein-coupled estrogen receptor GPR30 exerts vasoprotective effects in apolipoprotein E-deficient mice Arch. Med. Sci. 2021 10.5114/aoms/127200
124. Moreno-Ulloa A The effects of (-)-epicatechin on endothelial cells involve the G protein-coupled estrogen receptor (GPER) Pharmacol. Res. 2015 100 309 320 10.1016/j.phrs.2015.08.014 26303816
125. Meyer MR Fredette NC Barton M Prossnitz ER Regulation of vascular smooth muscle tone by adipose-derived contracting factor PLoS ONE 2013 8 e79245 10.1371/journal.pone.0079245 24244459
126. Meyer MR G protein-coupled estrogen receptor protects from atherosclerosis Sci. Rep. 2014 4 7564 10.1038/srep07564 25532911
127. Bopassa JC Eghbali M Toro L Stefani E A novel estrogen receptor GPER inhibits mitochondria permeability transition pore opening and protects the heart against ischemia-reperfusion injury Am. J. Physiol. Heart Circ. Physiol. 2010 298 H16 23 10.1152/ajpheart.00588.2009 19880667
128. Kabir ME G protein-coupled estrogen receptor 1 mediates acute estrogen-induced cardioprotection via MEK/ERK/GSK-3β pathway after ischemia/reperfusion PLoS ONE 2015 10 e0135988 10.1371/journal.pone.0135988 26356837
129. Feng Y Madungwe NB da Cruz Junho CV Bopassa JC Activation of G protein-coupled oestrogen receptor 1 at the onset of reperfusion protects the myocardium against ischemia/reperfusion injury by reducing mitochondrial dysfunction and mitophagy Br. J. Pharmacol. 2017 174 4329 4344 10.1111/bph.14033 28906548
130. Iorga A Rescue of pressure overload-induced heart failure by estrogen therapy J. Am. Heart Assoc. 2016 5 e002482 10.1161/JAHA.115.002482 26802104
131. Goncalves GK Neonatal cardiomyocyte hypertrophy induced by endothelin-1 is blocked by estradiol acting on GPER Am. J. Physiol. Cell Physiol. 2018 314 C310 C322 10.1152/ajpcell.00060.2017 29167148
132. Watanabe T 17β-Estradiol inhibits cardiac fibroblast growth through both subtypes of estrogen receptor Biochem. Biophys. Res. Commun. 2003 311 454 459 10.1016/j.bbrc.2003.09.232 14592435
133. Mercier I Mader S Calderone A Tamoxifen and ICI 182,780 negatively influenced cardiac cell growth via an estrogen receptor-independent mechanism Cardiovasc. Res. 2003 59 883 892 10.1016/S0008-6363(03)00517-0 14553828
134. Recchia AG The G protein-coupled receptor 30 is up-regulated by hypoxia-inducible factor-1α (HIF-1α) in breast cancer cells and cardiomyocytes J. Biol. Chem. 2011 286 10773 10782 10.1074/jbc.M110.172247 21266576
135. Whitcomb V Regulation of beta adrenoceptor-mediated myocardial contraction and calcium dynamics by the G protein-coupled estrogen receptor 1 Biochem. Pharmacol. 2020 171 113727 10.1016/j.bcp.2019.113727 31759979
136. Wang H Sun X Hodge HS Ferrario CM Groban L NLRP3 inhibition improves heart function in GPER knockout mice Biochem. Biophys. Res. Commun. 2019 514 998 1003 10.1016/j.bbrc.2019.05.045 31092335
137. Wang H Activation of GPR30 attenuates diastolic dysfunction and left ventricle remodelling in oophorectomized mRen2.Lewis rats Cardiovasc. Res. 2012 94 96 104 10.1093/cvr/cvs090 22328091
138. Jessup JA Lindsey SH Wang H Chappell MC Groban L Attenuation of salt-induced cardiac remodeling and diastolic dysfunction by the GPER agonist G-1 in female mRen2.Lewis rats PLoS ONE 2010 5 e15433 10.1371/journal.pone.0015433 21082029
139. Alencar AK Effect of age, estrogen status, and late-life GPER activation on cardiac structure and function in the Fischer344xBrown Norway female rat J. Gerontol. A Biol. Sci. Med. Sci. 2017 72 152 162 10.1093/gerona/glw045 27006078
140. Bairey Merz CN Sex and the kidneys: current understanding and research opportunities Nat. Rev. Nephrol. 2019 15 776 783 10.1038/s41581-019-0208-6 31586165
141. Chang Y G protein-coupled estrogen receptor activation improves contractile and diastolic functions in rat renal interlobular artery to protect against renal ischemia reperfusion injury Biomed. Pharmacother. 2019 112 108666 10.1016/j.biopha.2019.108666 30784936
142. Hofmeister MV 17β-Estradiol induces nongenomic effects in renal intercalated cells through G protein-coupled estrogen receptor 1 Am. J. Physiol. Ren. Physiol. 2012 302 F358 368 10.1152/ajprenal.00343.2011
143. Gohar EY Pollock DM Functional interaction of endothelin receptors in mediating natriuresis evoked by G protein-coupled estrogen receptor 1 J. Pharmacol. Exp. Ther. 2021 376 98 105 10.1124/jpet.120.000322 33127751
144. Meyer MR GPER is required for age-dependent albuminuria and glomerulosclerosis: evidence for its role in podocyte injury and mesangial Nox1 regulation Hypertension 2018 72 AP261 10.1161/hyp.72.suppl_1.P261
145. Meyer MR Deletion of GPER protects from age-related renovascular dysfunction and tubulo-interstitial injury Hypertension 2015 66 AP607 10.1161/hyp.66.suppl_1.p607
146. Qiao C Ye W Li S Wang H Ding X Icariin modulates mitochondrial function and apoptosis in high glucose-induced glomerular podocytes through G protein-coupled estrogen receptors Mol. Cell. Endocrinol. 2018 473 146 155 10.1016/j.mce.2018.01.014 29373840
147. Lindsey SH Yamaleyeva LM Brosnihan KB Gallagher PE Chappell MC Estrogen receptor GPR30 reduces oxidative stress and proteinuria in the salt-sensitive female mRen2.Lewis rat Hypertension 2011 58 665 671 10.1161/HYPERTENSIONAHA.111.175174 21844484
148. Gohar EY Activation of G protein-coupled estrogen receptor 1 ameliorates proximal tubular injury and proteinuria in Dahl salt-sensitive female rats Am. J. Physiol. Regul. Integr. Comp. Physiol. 2021 320 R297 R306 10.1152/ajpregu.00267.2020 33407017
149. Sanchez DS Estradiol stimulates cell proliferation via classic estrogen receptor-alpha and G protein-coupled estrogen receptor-1 in human renal tubular epithelial cell primary cultures Biochem. Biophys. Res. Commun. 2019 512 170 175 10.1016/j.bbrc.2019.03.056 30879772
150. Kurt AH Bozkus F Uremis N Uremis MM The protective role of G protein-coupled estrogen receptor 1 (GPER-1) on methotrexate-induced nephrotoxicity in human renal epithelium cells Ren. Fail. 2016 38 686 692 10.3109/0886022X.2016.1155398 26981789
151. Hutchens MP Fujiyoshi T Komers R Herson PS Anderson S Estrogen protects renal endothelial barrier function from ischemia-reperfusion in vitro and in vivo Am. J. Physiol. Ren. Physiol. 2012 303 F377 385 10.1152/ajprenal.00354.2011
152. Tofovic SP Zhang X Jackson EK Zhu H Petrusevska G 2-Methoxyestradiol attenuates bleomycin-induced pulmonary hypertension and fibrosis in estrogen-deficient rats Vasc. Pharmacol. 2009 51 190 197 10.1016/j.vph.2009.06.002
153. Umar S Estrogen rescues preexisting severe pulmonary hypertension in rats Am. J. Respir. Crit. Care Med. 2011 184 715 723 10.1164/rccm.201101-0078OC 21700911
154. Alencar AKN Cardioprotection induced by activation of GPER in ovariectomized rats with pulmonary hypertension J. Gerontol. A Biol. Sci. Med. Sci. 2018 73 1158 1166 10.1093/gerona/gly068 29790948
155. Alencar AK Activation of GPER ameliorates experimental pulmonary hypertension in male rats Eur. J. Pharm. Sci. 2017 97 208 217 10.1016/j.ejps.2016.11.009 27836751
156. Ahmadian R GPER contributes to the development of pulmonary hypertension in female rats FASEB J. 2020 34 1
157. Meyer MR Barton M GPER blockers as Nox downregulators: a new drug class to target chronic non-communicable diseases J. Steroid Biochem. Mol. Biol. 2018 176 82 87 10.1016/j.jsbmb.2017.03.019 28343901
158. Mauvais-Jarvis F Clegg DJ Hevener AL The role of estrogens in control of energy balance and glucose homeostasis Endocr. Rev. 2013 34 309 338 10.1210/er.2012-1055 23460719
159. Meyer MR Clegg DJ Prossnitz ER Barton M Obesity, insulin resistance and diabetes: sex differences and role of oestrogen receptors Acta Physiol. 2011 203 259 269 10.1111/j.1748-1716.2010.02237.x
160. Madak-Erdogan Z Design of pathway preferential estrogens that provide beneficial metabolic and vascular effects without stimulating reproductive tissues Sci. Signal. 2016 9 ra53 10.1126/scisignal.aad8170 27221711
161. Gurney EP Nachtigall MJ Nachtigall LE Naftolin F The Women’s Health Initiative Trial and related studies: 10 years later: a clinician’s view J. Steroid Biochem. Mol. Biol. 2014 142 4 11 10.1016/j.jsbmb.2013.10.009 24172877
162. Stubbins RE Holcomb VB Hong J Nunez NP Estrogen modulates abdominal adiposity and protects female mice from obesity and impaired glucose tolerance Eur. J. Nutr. 2012 51 861 870 10.1007/s00394-011-0266-4 22042005
163. Bonds DE The effect of conjugated equine oestrogen on diabetes incidence: the women’s health initiative randomised trial Diabetologia 2006 49 459 468 10.1007/s00125-005-0096-0 16440209
164. Mårtensson UE Deletion of the G protein-coupled receptor 30 impairs glucose tolerance, reduces bone growth, increases blood pressure, and eliminates estradiol-stimulated insulin release in female mice Endocrinology 2009 150 687 698 10.1210/en.2008-0623 18845638
165. Davis KE Sexually dimorphic role of G protein-coupled estrogen receptor (GPER) in modulating energy homeostasis Horm. Behav. 2014 66 196 207 10.1016/j.yhbeh.2014.02.004 24560890
166. Le May C Estrogens protect pancreatic β-cells from apoptosis and prevent insulin-deficient diabetes mellitus in mice Proc. Natl Acad. Sci. USA 2006 103 9232 9237 10.1073/pnas.0602956103 16754860
167. Liu S Importance of extranuclear estrogen receptor-α and membrane G protein-coupled estrogen receptor in pancreatic islet survival Diabetes 2009 58 2292 2302 10.2337/db09-0257 19587358
168. Sharma G Preclinical efficacy of the GPER-selective agonist G-1 in mouse models of obesity and diabetes Sci. Transl Med. 2020 12 eaau5956 10.1126/scitranslmed.aau5956 31996464
169. Butler MJ Hildebrandt RP Eckel LA Selective activation of estrogen receptors, ERα and GPER-1, rapidly decreases food intake in female rats Horm. Behav. 2018 103 54 61 10.1016/j.yhbeh.2018.05.018 29807036
170. Azizian H Khaksari M Asadi Karam G Esmailidehaj M Farhadi Z Cardioprotective and anti-inflammatory effects of G-protein coupled receptor 30 (GPR30) on postmenopausal type 2 diabetic rats Biomed. Pharmacother. 2018 108 153 164 10.1016/j.biopha.2018.09.028 30218860
171. Balhuizen A Kumar R Amisten S Lundquist I Salehi A Activation of G protein-coupled receptor 30 modulates hormone secretion and counteracts cytokine-induced apoptosis in pancreatic islets of female mice Mol. Cell. Endocrinol. 2010 320 16 24 10.1016/j.mce.2010.01.030 20122988
172. Kumar R Balhuizen A Amisten S Lundquist I Salehi A Insulinotropic and antidiabetic effects of 17β-estradiol and the GPR30 agonist G-1 on human pancreatic islets Endocrinology 2011 152 2568 2579 10.1210/en.2010-1361 21521748
173. Jacenik D Beswick EJ Krajewska WM Prossnitz ER G protein-coupled estrogen receptor in colon function, immune regulation and carcinogenesis World J. Gastroenterol. 2019 25 4092 4104 10.3748/wjg.v25.i30.4092 31435166
174. Hata M GPR30-expressing gastric chief cells do not dedifferentiate but are eliminated via PDK-dependent cell competition during development of metaplasia Gastroenterology 2020 158 1650 1666.e15 10.1053/j.gastro.2020.01.046 32032583
175. Tsai CC Estradiol mediates relaxation of porcine lower esophageal sphincter Steroids 2018 136 56 62 10.1016/j.steroids.2018.05.001 29733860
176. Zielinska M G protein-coupled estrogen receptor and estrogen receptor ligands regulate colonic motility and visceral pain Neurogastroenterol. Motil. 2017 29 13025 10.1111/nmo.13025
177. Li Y G protein-coupled estrogen receptor is involved in modulating colonic motor function via nitric oxide release in C57BL/6 female mice Neurogastroenterol. Motil. 2016 28 432 442 10.1111/nmo.12743 26661936
178. Jacenik D G protein-coupled estrogen receptor mediates anti-inflammatory action in Crohn’s disease Sci. Rep. 2019 9 6749 10.1038/s41598-019-43233-3 31043642
179. Wlodarczyk M G protein-coupled receptor 30 (GPR30) expression pattern in inflammatory bowel disease patients suggests its key role in the inflammatory process. A preliminary study J. Gastrointestin. Liver. Dis. 2017 26 29 35 10.15403/jgld.2014.1121.261.gpr 28338111
180. Qin B Expression of G protein-coupled estrogen receptor in irritable bowel syndrome and its clinical significance Int. J. Clin. Exp. Pathol. 2014 7 2238 2246 24966932
181. Shao X Li J Xu F Chen D Liu K Mir-155-mediated deregulation of GPER1 plays an important role in the gender differences related to inflammatory bowel disease Can. J. Infect. Dis. Med. Microbiol. 2020 2020 8811477 10.1155/2020/8811477 33014211
182. Jacenik D Estrogen signaling deregulation related with local immune response modulation in irritable bowel syndrome Mol. Cell. Endocrinol. 2018 471 89 96 10.1016/j.mce.2017.07.036 28774781
183. Chai S Activation of G protein-coupled estrogen receptor protects intestine from ischemia/reperfusion injury in mice by protecting the crypt cell proliferation Clin. Sci. 2019 133 449 464 10.1042/CS20180919
184. Wang Q Activation of the G protein-coupled estrogen receptor prevented the development of acute colitis by protecting the crypt cell J. Pharmacol. Exp. Ther. 2021 376 281 293 10.1124/jpet.120.000216 33318078
185. Chaturantabut S Estrogen activation of G-protein-coupled estrogen receptor 1 regulates phosphoinositide 3-kinase and mTOR signaling to promote liver growth in zebrafish and proliferation of human hepatocytes Gastroenterology 2019 156 1788 1804.e13 10.1053/j.gastro.2019.01.010 30641053
186. Tian L The developmental wnt signaling pathway effector β-catenin/TCF mediates hepatic functions of the sex hormone estradiol in regulating lipid metabolism PLoS Biol. 2019 17 e3000444 10.1371/journal.pbio.3000444 31589598
187. Farruggio S Genistein improves viability, proliferation and mitochondrial function of cardiomyoblasts cultured in physiologic and peroxidative conditions Int. J. Mol. Med. 2019 44 2298 2310 31638174
188. Wang HH Liu M Clegg DJ Portincasa P Wang DQ New insights into the molecular mechanisms underlying effects of estrogen on cholesterol gallstone formation Biochim. Biophys. Acta 2009 1791 1037 1047 10.1016/j.bbalip.2009.06.006 19589396
189. de Bari O Wang TY Liu M Portincasa P Wang DQ Estrogen induces two distinct cholesterol crystallization pathways by activating ERα and GPR30 in female mice J. Lipid Res. 2015 56 1691 1700 10.1194/jlr.M059121 26152119
190. Wang HH Activation of a novel estrogen receptor GPR30 enhances cholesterol cholelithogenesis in female mice Hepatology 2020 72 2077 2089 10.1002/hep.31212 32112420
191. Filardo EJ Distribution of GPR30, a seven membrane-spanning estrogen receptor, in primary breast cancer and its association with clinicopathologic determinants of tumor progression Clin. Cancer Res. 2006 12 6359 6366 10.1158/1078-0432.CCR-06-0860 17085646
192. Ignatov T Treeck O Kalinski T Ortmann O Ignatov A GPER-1 expression is associated with a decreased response rate to primary tamoxifen therapy of breast cancer patients Arch. Gynecol. Obstet. 2020 301 565 571 10.1007/s00404-019-05384-6 31900584
193. Ignatov T G-protein-coupled estrogen receptor GPER-1 expression in hormone receptor-positive breast cancer is associated with poor benefit of tamoxifen Breast Cancer Res. Treat. 2019 174 121 127 10.1007/s10549-018-5064-8 30478785
194. Pepermans RA Sharma G Prossnitz ER G protein-coupled estrogen receptor in cancer and stromal cells: functions and novel therapeutic perspectives Cells 2021 10 672 10.3390/cells10030672 33802978
195. De Francesco EM GPER mediates activation of HIF1α/VEGF signaling by estrogens Cancer Res. 2014 74 4053 4064 10.1158/0008-5472.CAN-13-3590 24894716
196. Smith HO GPR30: a novel indicator of poor survival for endometrial carcinoma Am. J. Obstet. Gynecol. 2007 196 386 e381 e389
197. Smith HO GPR30 predicts poor survival for ovarian cancer Gynecol. Oncol. 2009 114 465 471 10.1016/j.ygyno.2009.05.015 19501895
198. Chan QK Activation of GPR30 inhibits the growth of prostate cancer cells through sustained activation of Erk1/2, c-jun/c-fos-dependent upregulation of p21, and induction of G(2) cell-cycle arrest Cell. Death Differ. 2010 17 1511 1523 10.1038/cdd.2010.20 20203690
199. Natale CA Pharmacologic activation of the G protein-coupled estrogen receptor inhibits pancreatic ductal adenocarcinoma Cell. Mol. Gastroenterol. Hepatol. 2020 10 868 880 10.1016/j.jcmgh.2020.04.016 32376419
200. Vivacqua A 17β-Estradiol, genistein, and 4-hydroxytamoxifen induce the proliferation of thyroid cancer cells through the G protein-coupled receptor GPR30 Mol. Pharmacol. 2006 70 1414 1423 10.1124/mol.106.026344 16835357
201. Gilligan LC Estrogen activation by steroid sulfatase increases colorectal cancer proliferation via GPER J. Clin. Endocrinol. Metab. 2017 102 4435 4447 10.1210/jc.2016-3716 28945888
202. Liu C G-protein-coupled estrogen receptor antagonist G15 decreases estrogen-induced development of non-small cell lung cancer Oncol. Res. 2019 27 283 292 10.3727/096504017X15035795904677 28877783
203. Scaling AL Prossnitz ER Hathaway HJ GPER mediates estrogen-induced signaling and proliferation in human breast epithelial cells and normal and malignant breast Horm. Cancer 2014 5 146 160 10.1007/s12672-014-0174-1 24718936
204. Vivacqua A The G protein-coupled receptor GPR30 mediates the proliferative effects induced by 17β-estradiol and hydroxytamoxifen in endometrial cancer cells Mol. Endocrinol. 2006 20 631 646 10.1210/me.2005-0280 16239258
205. Albanito L G protein-coupled receptor 30 (GPR30) mediates gene expression changes and growth response to 17β-estradiol and selective GPR30 ligand G-1 in ovarian cancer cells Cancer Res. 2007 67 1859 1866 10.1158/0008-5472.CAN-06-2909 17308128
206. Ariazi EA The G protein-coupled receptor GPR30 inhibits proliferation of estrogen receptor-positive breast cancer cells Cancer Res. 2010 70 1184 1194 10.1158/0008-5472.CAN-09-3068 20086172
207. Natale CA Activation of G protein-coupled estrogen receptor signaling inhibits melanoma and improves response to immune checkpoint blockade eLife 2018 7 e31770 10.7554/eLife.31770 29336307
208. Wei T G protein-coupled estrogen receptor deficiency accelerates liver tumorigenesis by enhancing inflammation and fibrosis Cancer Lett. 2016 382 195 202 10.1016/j.canlet.2016.08.012 27594673
209. McDonnell DP Wardell SE Chang CY Norris JD Next-generation endocrine therapies for breast cancer J. Clin. Oncol. 2021 39 1383 1388 10.1200/JCO.20.03565 33705209
210. Mo Z GPR30 as an initiator of tamoxifen resistance in hormone-dependent breast cancer Breast Cancer Res. 2013 15 R114 10.1186/bcr3581 24289103
211. Ignatov A G-protein-coupled estrogen receptor GPR30 and tamoxifen resistance in breast cancer Breast Cancer Res. Treat. 2011 128 457 466 10.1007/s10549-011-1584-1 21607586
212. Marjon NA Hu C Hathaway HJ Prossnitz ER G protein-coupled estrogen receptor regulates mammary tumorigenesis and metastasis Mol. Cancer Res. 2014 12 1644 1654 10.1158/1541-7786.MCR-14-0128-T 25030371
213. Ignatov A Ignatov T Roessner A Costa SD Kalinski T Role of GPR30 in the mechanisms of tamoxifen resistance in breast cancer MCF-7 cells Breast Cancer Res. Treat. 2010 123 87 96 10.1007/s10549-009-0624-6 19911269
214. Prossnitz ER Arterburn JB International Union of Basic and Clinical Pharmacology. XCVII. G protein-coupled estrogen receptor and its pharmacologic modulators Pharmacol. Rev. 2015 67 505 540 10.1124/pr.114.009712 26023144
215. Li Y 4-Hydroxytamoxifen-stimulated processing of cyclin E is mediated via G protein-coupled receptor 30 (GPR30) and accompanied by enhanced migration in MCF-7 breast cancer cells Toxicology 2013 309 61 65 10.1016/j.tox.2013.04.012 23624423
216. Catalano S Tamoxifen through GPER upregulates aromatase expression: a novel mechanism sustaining tamoxifen-resistant breast cancer cell growth Breast Cancer Res. Treat. 2014 146 273 285 10.1007/s10549-014-3017-4 24928526
217. Liu Y G15 sensitizes epithelial breast cancer cells to doxorubicin by preventing epithelial-mesenchymal transition through inhibition of GPR30 Am. J. Transl Res. 2015 7 967 975 26175858
218. Wolfson B Padget MR Schlom J Hodge JW Exploiting off-target effects of estrogen deprivation to sensitize estrogen receptor negative breast cancer to immune killing J. Immunother. Cancer 2021 9 e002258 10.1136/jitc-2020-002258 34244306
219. Luo H GPER-mediated proliferation and estradiol production in breast cancer-associated fibroblasts Endocr. Relat. Cancer 2014 21 355 369 10.1530/ERC-13-0237 24481325
220. Yang K Yao Y Mechanism of GPER promoting proliferation, migration and invasion of triple-negative breast cancer cells through CAF Am. J. Transl Res. 2019 11 5858 5868 31632554
221. Ren J GPER in CAFs regulates hypoxia-driven breast cancer invasion in a CTGF-dependent manner Oncol. Rep. 2015 33 1929 1937 10.3892/or.2015.3779 25647524
222. Santolla MF GPER mediates a feedforward FGF2/FGFR1 paracrine activation coupling CAFs to cancer cells toward breast tumor progression Cells 2019 8 223 10.3390/cells8030223 30866584
223. De Marco P GPER signalling in both cancer-associated fibroblasts and breast cancer cells mediates a feedforward IL1β/IL1R1 response Sci. Rep. 2016 6 24354 10.1038/srep24354 27072893
224. Divella R De Luca R Abbate I Naglieri E Daniele A Obesity and cancer: the role of adipose tissue and adipo-cytokines-induced chronic inflammation J. Cancer 2016 7 2346 2359 10.7150/jca.16884 27994674
225. De Francesco EM Protective role of GPER agonist G-1 on cardiotoxicity induced by doxorubicin J. Cell Physiol. 2017 232 1640 1649 10.1002/jcp.25585 27607345
226. Smalley KS Why do women with melanoma do better than men? eLife 2018 7 e33511 10.7554/eLife.33511 29336304
227. Chakraborty B Inhibition of estrogen signaling in myeloid cells increases tumor immunity in melanoma J. Clin. Invest. 2021 131 e151347 10.1172/JCI151347 34637400
228. Natale CA Sex steroids regulate skin pigmentation through nonclassical membrane-bound receptors eLife 2016 5 e15104 10.7554/eLife.15104 27115344
229. Ribeiro MPC Santos AE Custodio JBA The activation of the G protein-coupled estrogen receptor (GPER) inhibits the proliferation of mouse melanoma K1735-M2 cells Chem. Biol. Interact. 2017 277 176 184 10.1016/j.cbi.2017.09.017 28947257
230. Shen Y Li C Zhou L Huang JA G protein-coupled oestrogen receptor promotes cell growth of non-small cell lung cancer cells via YAP1/QKI/circNOTCH1/m6A methylated NOTCH1 signalling J. Cell. Mol. Med. 2021 25 284 296 10.1111/jcmm.15997 33237585
231. Lam HM Targeting GPR30 with G-1: a new therapeutic target for castration-resistant prostate cancer Endocr. Relat. Cancer 2014 21 903 914 10.1530/ERC-14-0402 25287069
232. Lee SJ G protein-coupled estrogen receptor-1 agonist induces chemotherapeutic effect via ER stress signaling in gastric cancer BMB Rep. 2019 52 647 652 10.5483/BMBRep.2019.52.11.007 31234952
233. Zheng S Screening and survival analysis of hub genes in gastric cancer based on bioinformatics J. Comput. Biol. 2019 26 1316 1325 10.1089/cmb.2019.0119 31233344
234. Bustos V GPER mediates differential effects of estrogen on colon cancer cell proliferation and migration under normoxic and hypoxic conditions Oncotarget 2017 8 84258 84275 10.18632/oncotarget.20653 29137421
235. Cortes E GPER is a mechanoregulator of pancreatic stellate cells and the tumor microenvironment EMBO Rep. 2019 20 e46556 10.15252/embr.201846556 30538117
236. Morelli E Therapeutic activation of G protein-coupled estrogen receptor 1 in Waldenstrom macroglobulinemia Exp. Hematol. Oncol. 2022 11 54 10.1186/s40164-022-00305-x 36096954
237. Klein SL Flanagan KL Sex differences in immune responses Nat. Rev. Immunol. 2016 16 626 638 10.1038/nri.2016.90 27546235
238. Chakraborty B Estrogen receptor signaling in the immune system Endocr. Rev. 2023 44 117 141 10.1210/endrev/bnac017 35709009
239. Tamaki M Expression and functional roles of G-protein-coupled estrogen receptor (GPER) in human eosinophils Immunol. Lett. 2014 160 72 78 10.1016/j.imlet.2014.03.012 24718279
240. Itoga M G-protein-coupled estrogen receptor agonist suppresses airway inflammation in a mouse model of asthma through IL-10 PLoS ONE 2015 10 e0123210 10.1371/journal.pone.0123210 25826377
241. Brunsing RL Owens KS Prossnitz ER The G protein-coupled estrogen receptor (GPER) agonist G-1 expands the regulatory T-cell population under TH17-polarizing conditions J. Immunother. 2013 36 190 196 10.1097/CJI.0b013e31828d8e3b 23502766
242. Brunsing RL Prossnitz ER Induction of interleukin-10 in the T helper type 17 effector population by the G protein coupled estrogen receptor (GPER) agonist G-1 Immunology 2011 134 93 106 10.1111/j.1365-2567.2011.03471.x 21722102
243. Rettew JA McCall SH Marriott I GPR30/GPER-1 mediates rapid decreases in TLR4 expression on murine macrophages Mol. Cell. Endocrinol. 2010 328 87 92 10.1016/j.mce.2010.07.017 20654686
244. Rodenas MC G protein-coupled estrogen receptor 1 regulates human neutrophil functions Biomed. Hub. 2017 2 1 13 10.1159/000454981 31988900
245. Yasuda H 17-β-Estradiol enhances neutrophil extracellular trap formation by interaction with estrogen membrane receptor Arch. Biochem. Biophys. 2019 663 64 70 10.1016/j.abb.2018.12.028 30590021
246. Castleman MJ Innate sex bias of Staphylococcus aureus skin infection is driven by α-hemolysin J. Immunol. 2018 200 657 668 10.4049/jimmunol.1700810 29222165
247. Triplett KD GPER activation protects against epithelial barrier disruption by Staphylococcus aureus α-toxin Sci. Rep. 2019 9 1343 10.1038/s41598-018-37951-3 30718654
248. Costa AJ Overexpression of estrogen receptor GPER1 and G1 treatment reduces SARS-CoV-2 infection in BEAS-2B bronchial cells Mol. Cell. Endocrinol. 2022 558 111775 10.1016/j.mce.2022.111775 36096380
249. Wang C Membrane estrogen receptor regulates experimental autoimmune encephalomyelitis through up-regulation of programmed death 1 J. Immunol. 2009 182 3294 3303 10.4049/jimmunol.0803205 19234228
250. Du ZR G protein-coupled estrogen receptor is involved in the anti-inflammatory effects of genistein in microglia Phytomedicine 2018 43 11 20 10.1016/j.phymed.2018.03.039 29747742
251. Guan J Yang B Fan Y Zhang J GPER agonist G1 attenuates neuroinflammation and dopaminergic neurodegeneration in parkinson disease Neuroimmunomodulation 2017 24 60 66 10.1159/000478908 28810246
252. Harding AT Goff MA Froggatt HM Lim JK Heaton NS GPER1 is required to protect fetal health from maternal inflammation Science 2021 371 271 276 10.1126/science.aba9001 33446553
253. Meyer MR Fredette NC Sharma G Barton M Prossnitz ER GPER is required for the age-dependent upregulation of the myocardial endothelin system Life Sci. 2016 159 61 65 10.1016/j.lfs.2016.02.041 26880534
254. Robison LS Gannon OJ Salinero AE Zuloaga KL Contributions of sex to cerebrovascular function and pathology Brain Res. 2019 1710 43 60 10.1016/j.brainres.2018.12.030 30580011
255. Patkar S Farr TD Cooper E Dowell FJ Carswell HV Differential vasoactive effects of oestrogen, oestrogen receptor agonists and selective oestrogen receptor modulators in rat middle cerebral artery Neurosci. Res. 2011 71 78 84 10.1016/j.neures.2011.05.006 21624404
256. Tang H GPR30 mediates estrogen rapid signaling and neuroprotection Mol. Cell. Endocrinol. 2014 387 52 58 10.1016/j.mce.2014.01.024 24594140
257. Murata T Dietrich HH Xiang C Dacey RG Jr. G protein-coupled estrogen receptor agonist improves cerebral microvascular function after hypoxia/reoxygenation injury in male and female rats Stroke 2013 44 779 785 10.1161/STROKEAHA.112.678177 23362079
258. Zhang Z The novel estrogenic receptor GPR30 alleviates ischemic injury by inhibiting TLR4-mediated microglial inflammation J. Neuroinflammation 2018 15 206 10.1186/s12974-018-1246-x 30001721
259. Han ZW GPER agonist G1 suppresses neuronal apoptosis mediated by endoplasmic reticulum stress after cerebral ischemia/reperfusion injury Neural Regen. Res. 2019 14 1221 1229 10.4103/1673-5374.251571 30804253
260. Bai N G-protein-coupled estrogen receptor activation upregulates interleukin-1 receptor antagonist in the hippocampus after global cerebral ischemia: implications for neuronal self-defense J. Neuroinflammation 2020 17 45 10.1186/s12974-020-1715-x 32007102
261. Wang XS Activation of G protein-coupled receptor 30 protects neurons by regulating autophagy in astrocytes Glia 2020 68 27 43 10.1002/glia.23697 31429156
262. Peng J Activation of GPR30 with G1 attenuates neuronal apoptosis via src/EGFR/stat3 signaling pathway after subarachnoid hemorrhage in male rats Exp. Neurol. 2019 320 113008 10.1016/j.expneurol.2019.113008 31295444
263. Lu D Activation of G protein-coupled estrogen receptor 1 (GPER-1) ameliorates blood-brain barrier permeability after global cerebral ischemia in ovariectomized rats Biochem. Biophys. Res. Commun. 2016 477 209 214 10.1016/j.bbrc.2016.06.044 27311857
264. Zhang B Estradiol and G1 reduce infarct size and improve immunosuppression after experimental stroke J. Immunol. 2010 184 4087 4094 10.4049/jimmunol.0902339 20304826
265. Zheng Y GPER-deficient rats exhibit lower serum corticosterone level and increased anxiety-like behavior Neural Plast. 2020 2020 8866187 10.1155/2020/8866187 32908490
266. Wang J The antidepressant and anxiolytic effect of GPER on translocator protein (TSPO) via protein kinase a (PKA) signaling in menopausal female rats J. Steroid Biochem. Mol. Biol. 2021 207 105807 10.1016/j.jsbmb.2020.105807 33345973
267. Bourque M Morissette M Di Paolo T Neuroprotection in Parkinsonian-treated mice via estrogen receptor alpha activation requires G protein-coupled estrogen receptor 1 Neuropharmacology 2015 95 343 352 10.1016/j.neuropharm.2015.04.006 25892506
268. Jiao Y Molecular identification of bulbospinal on neurons by GPER which drives pain and morphine tolerance J. Clin. Invest. 2023 133 e154588 10.1172/JCI154588 36346677
269. Yuan LJ G protein-coupled estrogen receptor is involved in the neuroprotective effect of IGF-1 against MPTP/MPP+-induced dopaminergic neuronal injury J. Steroid Biochem. Mol. Biol. 2019 192 105384 10.1016/j.jsbmb.2019.105384 31175966
270. de Souza LO The G protein-coupled estrogen receptor (GPER) regulates recognition and aversively-motivated memory in male rats Neurobiol. Learn. Mem. 2021 184 107499 10.1016/j.nlm.2021.107499 34352396
271. Kubota T Matsumoto H Kirino Y Ameliorative effect of membrane-associated estrogen receptor G protein coupled receptor 30 activation on object recognition memory in mouse models of Alzheimer’s disease J. Pharmacol. Sci. 2016 131 219 222 10.1016/j.jphs.2016.06.005 27423484
272. Wang ZF Pan ZY Xu CS Li ZQ Activation of G-protein coupled estrogen receptor 1 improves early-onset cognitive impairment via PI3K/Akt pathway in rats with traumatic brain injury Biochem. Biophys. Res. Commun. 2017 482 948 953 10.1016/j.bbrc.2016.11.138 27908726
273. Amirkhosravi L E2-BSA and G1 exert neuroprotective effects and improve behavioral abnormalities following traumatic brain injury: the role of classic and non-classic estrogen receptors Brain Res. 2021 1750 147168 10.1016/j.brainres.2020.147168 33096091
274. Confavreux C Hutchinson M Hours MM Cortinovis-Tourniaire P Moreau T Rate of pregnancy-related relapse in multiple sclerosis. Pregnancy in multiple sclerosis group N. Engl. J. Med. 1998 339 285 291 10.1056/NEJM199807303390501 9682040
275. Seifert HA Estrogen protects both sexes against EAE by promoting common regulatory cell subtypes independent of endogenous estrogen Metab. Brain Dis. 2017 32 1747 1754 10.1007/s11011-017-0063-8 28689297
276. Subramanian S Miller LM Grafe MR Vandenbark AA Offner H Contribution of GPR30 for 1,25 dihydroxyvitamin D(3) protection in EAE Metab. Brain Dis. 2012 27 29 35 10.1007/s11011-011-9266-6 21994003
277. Markee JE Rhythmic uterine vascular changes Am. J. Physiol. 1932 100 32 39 10.1152/ajplegacy.1932.100.1.32
278. Carloni E L’azione degli estratti ovarici sull’atteggiamento dei capillari e sulla loro pressione, nelle varie fasi della rivoluzione funziozionale utero-ovarica Arch. Ostet. Ginecol. 1930 17 327i
279. Collin JB Browne JS Thomson DL The chemical nature of emmenin Endocrinology 1934 18 71 74 10.1210/endo-18-1-71
280. Dawson RF Robson JM The pharmacological actions of diethylstilboestrol and other oestrogenic and non-oestrogenic substances J. Physiol. 1939 95 420 430 10.1113/jphysiol.1939.sp003739 16995100
281. Reynolds SR Foster FI Peripheral vascular action of estrogen in the human male J. Clin. Invest. 1939 18 649 655 10.1172/JCI101080 16694698
282. Vance DA Premarin: the intriguing history of a controversial drug Int. J. Pharm. Compd. 2007 1 282 286
283. Veurink M Koster M Berg LT The history of DES, lessons to be learned Pharm. World Sci. 2005 27 139 43 10.1007/s11096-005-3663-z 16096877
284. Pincus G Rock J Chang MC Garcia CR Effects of certain 19-nor steroids on reproductive processes and fertility Fed. Proc. 1959 18 1051 1056 14433037
285. Lerner LJ Holthaus FJ Thompson CR A non-steroidal estrogen antagonist 1-(p-2-diethylaminoethoxyphenyl)-1-phenyl-2-p-methoxyphenyl ethanol Endocrinology 1958 63 295 318 10.1210/endo-63-3-295 13574085
286. Jordan VC Tamoxifen: a most unlikely pioneering medicine Nat. Rev. Drug Discov. 2003 2 205 13 10.1038/nrd1031 12612646
287. Toft D Gorski J A receptor molecule for estrogens: isolation from the rat uterus and preliminary characterization Proc. Natl Acad. Sci. USA 1966 55 1574 1581 10.1073/pnas.55.6.1574 5227676
288. Jensen EV A two-step mechanism for the interaction of estradiol with rat uterus Proc. Natl Acad. Sci. USA 1968 59 632 638 10.1073/pnas.59.2.632 5238991
289. Szego CM Davis JS Adenosine 3’,5’-monophosphate in rat uterus: acute elevation by estrogen Proc. Natl Acad. Sci. USA 1967 58 1711 1718 10.1073/pnas.58.4.1711 4295833
290. Pietras RJ Szego CM Endometrial cell calcium and oestrogen action Nature 1975 253 357 359 10.1038/253357a0 1167402
291. Jordan VC Dowse LJ Tamoxifen as an anti-tumour agent: effect on oestrogen binding J. Endocrinol. 1976 68 297 303 10.1677/joe.0.0680297 176295
292. Pietras RJ Szego CM Estrogen receptors in uterine plasma membrane J. Steroid Biochem. 1979 11 1471 1483 10.1016/0022-4731(79)90124-9 513766
293. Owman C Blay P Nilsson C Lolait SJ Cloning of human cDNA encoding a novel heptahelix receptor expressed in Burkitt’s lymphoma and widely distributed in brain and peripheral tissues Biochem. Biophys. Res. Commun. 1996 228 285 292 10.1006/bbrc.1996.1654 8920907
294. Carmeci C Thompson DA Ring HZ Francke U Weigel RJ Identification of a gene (GPR30) with homology to the G-protein-coupled receptor superfamily associated with estrogen receptor expression in breast cancer Genomics 1997 45 607 617 10.1006/geno.1997.4972 9367686
295. Feng Y Gregor P Cloning of a novel member of the G protein-coupled receptor family related to peptide receptors Biochem. Biophys. Res. Commun. 1997 231 651 654 10.1006/bbrc.1997.6161 9070864
296. Takada Y Kato C Kondo S Korenaga R Ando J Cloning of cDNAs encoding G protein-coupled receptor expressed in human endothelial cells exposed to fluid shear stress Biochem. Biophys. Res. Commun. 1997 240 737 741 10.1006/bbrc.1997.7734 9398636
297. Kvingedal AM Smeland EB A novel putative G-protein-coupled receptor expressed in lung, heart and lymphoid tissue FEBS Lett. 1997 407 59 62 10.1016/S0014-5793(97)00278-0 9141481
298. O’Dowd BF Discovery of three novel G-protein-coupled receptor genes Genomics 1998 47 310 313 10.1006/geno.1998.5095 9479505
299. Bonini JA Anderson SM Steiner DF Molecular cloning and tissue expression of a novel orphan G protein-coupled receptor from rat lung Biochem. Biophys. Res. Commun. 1997 234 190 193 10.1006/bbrc.1997.6591 9168987
|
PMC010xxxxxx/PMC10193776.txt |
==== Front
Journal of Building Engineering
2352-7102
2352-7102
Elsevier Ltd.
S2352-7102(23)00986-5
10.1016/j.jobe.2023.106807
106807
Article
Assessing the impact of architectural and behavioral interventions for controlling indoor COVID-19 infection risk: An agent-based approach
Zhang Anxiao ab
Zhen Qi ba∗
Zheng Chi c
Li Jing d
Zheng Yue a
Du Yiming a
Huang Qiong a
Zhang Qi a
a School of Architecture, Tianjin University, Tianjin, China
b Key Laboratory of Ecology and Energy Saving Study of Dense Habitat, Ministry of Education, China
c China Construction First Group Corporation Limited, China
d School of Public Health, Tianjin Medical University, Tianjin, China
∗ Corresponding author. School of Architecture, Tianjin University, No. 92 Weijin Street, Nankai District, Tianjin, 300072, PR China.
18 5 2023
1 9 2023
18 5 2023
74 106807106807
18 2 2023
5 5 2023
8 5 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The COVID-19 pandemic changed our lives, forcing us to reconsider our built environment. In some buildings with high traffic flow, infected individuals release viral particles during movement. The complex interactions between humans, building, and viruses make it difficult to predict indoor infection risk by traditional computational fluid dynamics methods. The paper developed a spatially-explicit agent-based model to simulate indoor respiratory pathogen transmission for buildings with frequent movement of people. The social force model simulating pedestrian movement and a simple forcing method simulating indoor airflow were coupled in an agent-based modeling environment. The impact of architectural and behavioral interventions on the indoor infection risk was then compared by simulating a supermarket case. We found that wearing a mask was the most effective single intervention, with all people wearing masks reducing the percentage of infections to 0.08%. Among the combined interventions, the combination of customer control is the most effective and can reduce the percentage of infections to 0.04%. In addition, the extremely strict combination of all the interventions makes the supermarket free of new infections during its 8-h operation. The approach can help architects, managers, or the government better understand the effect of nonpharmaceutical interventions to reduce the infection risk and improve the level of indoor safety.
Keywords
Agent-based model
Indoor SARS-CoV-2 transmission
Architecture design
Behavioral interventions
Buildings with high traffic flow
Supermarket
==== Body
pmc1 Introduction
Over the past 40 years, the frequency of outbreaks of respiratory infectious diseases has increased significantly [1]. The global coronavirus disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused negative impacts on health and tremendous losses of human lives, with over 632 million confirmed cases and 6.5 million deaths (as of Nov 17, 2022) [2]. More than 90% of infections occurred indoors [3] due to difficulties in maintaining social distancing and limited ventilation [4]. It is reported that the indoor infection risk is 18.7 times higher than outdoors [5]. However, currently many studies on infection risk prediction focus on the city or country scale [6,7], and there are relatively few indoor infection risk prediction tools for various building types and occupant characteristics [8].
Public buildings such as offices, railway stations, supermarkets, hospitals, schools, etc. are the main places where respiratory infectious diseases spread [9]. These buildings here can be divided into two categories: one is the buildings with relatively fixed spatial positions of indoor occupants, such as office buildings, hotels or hospital wards, etc., where occupants remain relatively static for a long time; the other type is the buildings with the frequent internal pedestrian flow, such as airports, subway stations, supermarkets, etc., where the spatial positions of the occupants are constantly changing (Fig. 1 (a) (b)). The former has been studied by many researchers using Computational Fluid Dynamic (CFD) calculation [10] or the tracer gas method [11]. However, for the latter building category, indoor virus transmission simulation is more difficult. First, since the location of the infected individual is constantly changing, it is difficult for the traditional CFD method to handle this kind of dynamic simulation. Second, people's indoor positions are constantly changing due to the influence of environmental factors, such as spatial scale and real-time pedestrian flow. The complexity of the interactions between building, humans, and viruses makes it difficult to predict the infection risk of people in such buildings.Fig. 1 Scenarios of indoor infection risk assessment and the ABM modeling framework and approach.
Fig. 1
Regarding human-building interaction, some researchers have explored the field of thermal performance. Hong et al. [12], Langevin et al. [13] and Zhang et al. [14] studied the effects of people opening/closing windows, turning on/off heaters, and clothing adjustment on building energy consumption and thermal comfort. For indoor virus transmission, it is more important that spatial factors affect people's paths, stay time, gathering conditions, etc. [15]. For instance, when people are looking for the products they want in a supermarket, the destinations and paths will vary depending on the spatial layout. Furthermore, in the process of moving to the destination, crowding may be caused due to the variation of spatial scale. In addition, queuing may occur due to the artificial division of some areas into linear spaces. These are human-building interactions closely related to the spread of diseases in indoor spaces [16].
Human-virus interaction mainly includes two categories: air-based transmission and surface-based transmission. For the former, high-speed camera equipment was used to capture the distribution characteristics of droplets sprayed from the mouth when people breathe, talk, cough, sneeze [17], or sing [18]. Wei et al. [19] established a mathematical model to calculate the short-range airborne transmission of SARS-CoV-2 considering the talking and coughing. For the latter, touching-related parameters and models were also established. Duives et al. [15] described and modeled in detail the process of contamination of surfaces by infected individuals and touching of surface viruses by susceptible individuals associated with SARS-CoV-2 transmission. Zhang et al. [20] established a hand contact model in an office setting to simulate the effectiveness of disinfection and behavioral interventions.
The virus-building interaction mainly includes the flow or decay of viruses caused by environmental conditions such as ventilation, temperature, and humidity. For ventilation, many researchers have used CFD methods to simulate virus flow under different ventilation conditions. Li et al. [21], Vuorinen et al. [22] and Arpino et al. [23] modeled the aerosol transmission in restaurants, retail-store, car cabins, etc. However, this method is limited by the assumption that the infected individual is fixed, which makes it difficult to simulate scenarios where people move frequently. On the other hand, the stability characteristics of viruses in different mediums, temperatures, and humidity have also been explored by medical researchers. Studies have shown that SARS-CoV-2 can remain infective on plastic surfaces for up to 9 days [24], but the survival time on copper surfaces does not exceed 4 h [25]. These studies provide basic valid data for virus-building interaction modeling.
This paper explores the prediction of infection risk of respiratory infectious disease in buildings with high traffic flow. It focuses on the interaction between humans, building, and viruses, and uses the agent-based modeling (ABM) method to couple the three kinds of interactions. Taking a supermarket as a case, we built a spatially-explicit building-scale model, based on which a parametric study was done to explore the impact of different building and behavioral measures on indoor infection risk. Through this analysis, we provide guidance for minimizing SARS-CoV-2 transmission during indoor gatherings.
2 Background
Studies on the evaluation of indoor COVID-19 infection risk of buildings can be divided into several categories, one of which is dominated by CFD simulations. Li et al. [26] performed detailed CFD simulations to simulate the spread of fine exhaled droplets in a restaurant in China to assess the possibility of airborne transmission and to characterize the associated environmental conditions. Vuorinen et al. [22] give various examples on the transport and dilution of aerosols over distances in a supermarket by using CFD simulations. In addition, Motamedi et al. [27] proposed a framework based on CFD method to evaluate the effect of different ventilation strategies on infection probability in an office room.
The other category is based on statistical data and mathematical formulas for indoor infection risk prediction. For example, Peng et al. [28] combine the key factors that control indoor airborne disease transmission including aerosol-generation rate, breathing flow rate, masking, ventilation and aerosol-removal rates, number of occupants, and duration of exposure, and finally proposed two indicators of infection risk, i.e., relative risk parameter and risk parameter.
In addition, some studies base on ABM method have coupled pedestrian models and viral transmission models for indoor infection risk prediction. D'Orazio et al. [29] proposed a probabilistic simulation model based on consolidated proximity and exposure-time-based rules to evaluate the effectiveness of mask wearing, density control and access control solutions for COVID-19 spreading in university buildings. Antczak et al. [30] built an agent-based model for evaluating the effectiveness of social distancing and checkout zone design in supermarkets during COVID-19 breaks. Alvarez & Ford [31] proposed a geospatial 3D agent-based model to explore the effect of face masks, lockdown, and self-isolation on the transmission of COVID-19 in university campuses. Harweg et al. [32] proposed an agent-based simulation of pedestrian dynamics to assess the distance measures to control close contact transmission of COVID-19. Farthing and Lanzas [33] developed an agent-based model for simulating indoor respiratory pathogen transmission in a single room scenario. The efficacy of four interventions including mask use, ventilation, pedestrian movement and social distancing was examined. Moreover, Lee et al. [34] proposed an OccSim system built in C# that generates occupancy behaviours in a 3D model of a building and helps users analyze the potential effect of virus transmission. Air-based transmission and surface-based transmission routes were both included in the model. Ronchi and Lovreglio [35] developed an occupant exposure model “EXPOSED” based on microscopic crowd models to assess the occupant exposure level in confined spaces. Additionally, Duives et al. [15] coupled a microscopic simulation model (Nomad) and a virus spread model (QVEmod) to investigate the spread of SARS-CoV-2 in indoor spaces. A restaurant was studied as a case study, and the effect of ventilation rate and face masks was compared.
Table 1 summarizes the building scenarios, interventions, involved transmission routes, methods of the model and implementation tools of recent studies on indoor infection risk assessment. Although there are many studies on indoor COVID-19 infection risk, there is still room for improvement: (1) Most of the models only cover mostly one or two transmission routes, while models combining multiple transmission routes containing air-based routes and surface-based routes are rare. Architects are difficult to obtain an overall assessment of the risk of infection in a certain spatial context covering all transmission routes. (2) Traditional CFD methods are mostly used to assess the effectiveness of ventilation interventions. It is difficult to apply CFD methods to buildings with frequent movement of occupants, where the source of contamination is not static. Therefore, it is necessary to develop an indoor infection risk assessment method that can integrate multiple transmission routes, and can quickly predict the infection risk in buildings with frequent movement of people to help architects and managers make better decisions. The flexible behavior simulation of the ABM model provides an opportunity to simulate indoor virus transmission in buildings with high traffic flow.Table 1 Typical research on the indoor COVID-19 infection risk assessment model.
Table 1Building scenario Interventions Transmission routes Methods of the model Implementation tools Ref.
Restaurant Ventilation Short-range airborne route, Long-range airborne route CFD method Ansys Fluent [26]
Supermarket Ventilation Short-range airborne route, Long-range airborne route CFD, Monte-Carlo method PALM, OpenFOAM, NS3dLab, Fire Dynamics Simulator [22]
Office room Ventilation Short-range airborne route, Long-range airborne route CFD method Ansys Fluent [27]
Classrooms, subway, supermarket, Stadium etc. Ventilation rate, occupant density, mask efficiency Long-range airborne route Statistics-based method Excel [28]
University building Wearing masks, density control, access control Short-range airborne route Statistics-based method NetLogo [29]
Supermarket Social distancing, Checkout zone design Short-range airborne route Agent-based method NetLogo [30]
University campus Face masks, lockdown, self-isolation Short-range airborne route Agent-based method GAMA [31]
Supermarket Social distancing Short-range airborne route Agent-based method – [32]
Single room Mask use, ventilation, pedestrian movement, social distancing Short-range airborne route, Long-range airborne route Agent-based method NetLogo [33]
Confined space Occupant density Short-range airborne route, Long-range airborne route Agent-based method – [35]
Office building Architecture design, facility and behavior management, Short-range airborne route, Long-range airborne route, Environmental fomite route Agent-based method C# [34]
Restaurant Ventilation rate, face masks Short-range airborne route, Long-range airborne route, Environmental fomite route Agent-based method – [15]
The novelty of this study are as follows: (1) A spatially-explicit agent-based model that integrates building space, human behavior, and virus transmission to determine the efficacy of different types of interventions on the indoor infection risk was proposed. (2) The social force model simulating pedestrian movement and a simple forcing method simulating indoor airflow were coupled in an agent-based modeling environment. This study can provide architects and researchers with more comprehensive and in-depth understanding of the impact of architectural and behavioral interventions on indoor COVID-19 infection risk.
3 Methodology
3.1 Modeling framework for simulating indoor SARS-CoV-2 transmission
The above analysis indicates that an integrated model simulating the human-building-virus interaction is required to model indoor SARS-CoV-2 transmission for buildings with frequent movement of people (Fig. 1 (b)). We have designed an agent-based model, which focuses on simulating three types of interactions during indoor SARS-CoV-2 transmission: human-building interaction, human-virus interaction, and virus-building interaction. The model framework is shown in Fig. 1 (c).
The first step is to collect parametric data for humans, building, and viruses. The input variables consist of two categories. One is the fixed variables in the model, which is a constant, such as the decay rate of a virus. The other is the variables that can be changed, which can be adjusted and optimized to reduce the probability of indoor infection. This paper studies the impact of architectural and behavioral interventions on infection risk, so the adjustable variables are mainly building-related parameters and human-related parameters. The second step is to simulate the interaction of humans, building, and viruses, which includes three aspects. Human-building interaction modeling mimics indoor human activities due to architecture design, which includes walking route choice, crowding, and queuing phenomena. Human-virus interaction modeling refers to changes in virus transmission caused by various human behaviors, including air-based transmission and surface-based transmission. Virus-building interaction modeling includes two types: one is the effect of ventilation on the flow of viruses in the air, while the other is the effect of environmental media, temperature, and humidity on the virus decay rate. The third step is the infection risk assessment. The total exposure of individuals through air-based and surface-based transmission routes is counted, and then the individual infection probability is determined according to the dose-response equation. The last step is to output the predicted results, including the percentage of probable infections, high-risk contagion space, risky behavior, etc. The output data includes real-time situated visualization and timeline-based analysis.
The model was created using the open-source modeling software NetLogo and is available at https://github.com/zax1111/Indoor-COVID-19-Infection-Risk-Assessment-Model. Fig. 1 (d) shows the conceptual diagram of the model. At the beginning of the model is the input panel, which covers both building interventions and behavioral interventions. The architectural interventions rely on the building design block for inputting building plan information and the ventilation design block for inputting ventilation conditions. The behavioral interventions include a customer control block, a checkout control block, and a sanitary control block, which regulate customer shopping behavior, checkout management, and hygiene behavior, respectively. Then comes the virus transmission module, which includes the air-based transmission block and the surface-based transmission block. Finally, there is a risk assessment block, followed by a logging block and a visualization block to record data and plot real-time images.
3.2 Human-building interaction modeling
3.2.1 Route choice
When people go to their destination, the route selection problem will arise. People choose different paths in different architectural scenarios. In a familiar environment, people tend to choose the shortest path [36]. Since it is difficult to determine whether the occupants are familiar with the built environment or not, the shortest route toward one's destination is adopted in this paper. Algorithms for the shortest route include Dijkstra's algorithm [37], best-first search algorithm [38], and A-star algorithm [39], etc. We choose the A-star algorithm which is a combination of the former two.
3.2.2 Crowding phenomenon
Crowding in space means that the distance between people has decreased dramatically, and this has a strong correlation with the transmission of respiratory diseases. Many pedestrian models have been proposed to characterize crowds, among which physics-based approaches are most common. Well-known examples are the fluid-dynamic model [40] and the social force model [41,42]. This paper chooses the social force model of Helbing and Peter [43], which uses driving and repulsive forces to describe the aggregation and dissipation of crowds under the variation of spatial scale. It successfully reproduced the “faster is slower” effect when crowded people push each other to go through a narrow opening. The social force model here mainly includes three types of forces, namely the driving force, the force from other pedestrians, and the force from obstacles (Fig. 2 ). We obtained the surveillance video of the case supermarket, extracted the crowd density and throughput rate at the entrance and compared them with the simulation data in the same spatial range to calibrate and enhance the model parameters. The validation and improvement process are detailed in Appendix S1.Fig. 2 Diagram of social force model to simulate the indoor crowding phenomenon.
Fig. 2
3.2.3 Queuing phenomenon
Queuing is also a common phenomenon in buildings, due to the linear division of space or linear restrictions on crowd behavior. For queuing behavior, we designate a specific queuing zone in the building, where people in the zone cannot move freely but can only move in a certain direction. Furthermore, the distance between the queues can be set to reflect the effectiveness of social distancing measures.
3.3 Human-virus interaction modeling
3.3.1 Air-based transmission
Air-based transmission mainly includes two processes. One is the exhalation process, in which a person spread the virus into the air through breathing, speaking, coughing, etc. While the other is the inhalation process, in which a person inhales the virus in the air through breathing.
To describe this human-virus interaction process in the air, the model space is divided into grid cells. Each grid is 0.5 m × 0.5 m and is assumed to store air and droplets. When an infected person performs contagion-related behaviors, e.g., breathing, speaking, coughing, etc., the number of droplets in specific grids nearby will increase. All droplets are assumed to be ejected at an average human height of 1.7 m [44]. Besides, the model time step Δt is set to 0.25 s. We assume that a COVID-19 patient has the same cough frequency as a common chronic cough patient, with a 19% probability of coughing per minute [45,46], so the cough probability per Δt is 19%/60 × 0.25 = 7.92 × 10−4.
The droplets emitted from the human body vary in size, quantity, and distance in terms of different ejection activities. Although many studies classify “aerosols” (small) and “droplets” (large) by size, in this paper, we refer to all the ejected particles as “droplets” regardless of size for convenience. The droplets were classified into 16 size classes ranging from 3 μm to 750 μm in diameter as adopted from Duguid [33,47]. The distribution of droplet sizes is shown in Fig. 3 . Distributions of size classes during coughing and non-coughing events are based on the experimental findings of Chao et al. [48], which were recorded 60 mm away from people's mouths following these activities. And the air temperature and RH averaged from all experiments were 24.9 °C and 73.9%, respectively. For the droplets’ quantity, the average value of the droplet count is set to 9.7 × 105 droplets/expectoration with a standard deviation of 3.9 × 105 based on the model described in Ref. [33], which is derived from the Skagit County choir COVID-19 outbreak event [49]. These values are approximately equal to 970 (SD = 390) quanta/hr.Fig. 3 Distribution of droplet sizes during expectoration events.
Fig. 3
The spreading distance and angle differ for coughing and non-coughing event. Many studies have used high-speed cameras or CFD methods to collect droplet data under different behaviors, and the spatial range of droplets obtained is generally fan-shaped. We use the spread distance and spread angle index to describe the distribution of droplets in space (Fig. 4 ). Here the log-normal distributions are used to describe the randomness of the two variables [50]. Travel distances for coughing events follow the distribution with a mean of 5 m and a standard deviation of 0.256 m [51]. Travel distances for non-coughing events follow the distribution with a mean of 0.55 m and a standard deviation of 0.068 m based on Das et al.’ finding [52]. They found that the majority of 100 μm droplets will fall 0.55–2.35 m away from the expelling individual, depending on initial velocity. Since the spatial distribution of droplets is similar to a fan, the in-cone function in NetLogo is used. The cone is defined by two inputs, the distance and the angle which may range from 0 to 360 and is centered around the agent's current heading. The spread angle during coughing and non-coughing expectations were 35° and 63.5°, respectively according to Kwon et al. [53] and Gupta et al. [54]. Table 2 summarizes the parameters under coughing and non-coughing event.Fig. 4 Schematic diagram of spread angle and distance during coughing and non-coughing event.
Fig. 4
Table 2 Droplets spreading parameters under coughing and non-coughing event.
Table 2 Spread angle Spread distance (mean) Spread distance (standard deviation) Number(mean) Number(standard deviation)
Non-coughing 63.5° 0.55 m 0.068 m 9.7 × 105 3.9 × 105
Coughing 35° 5 m 0.256 m 9.7 × 105 3.9 × 105
For the inhalation process, if a susceptible person is located in the contaminated cell, the droplets containing the virus will be inhaled at a certain speed, and the virus exposure of the individual is reached. The inhalation rate for simulated individuals was set to 0.023 m3 air/min, equivalent to light physical activity for adults [55,56]. The number of droplets inhaled by an individual is calculated using Eq. (4).(4) Dinhale(t)=DVcellγinhaleΔt
Where D inhale (t) is the number of droplets inhaled by a person during (t-Δt, t), D is the total number of droplets in the unit cell, V cell is the volume of the spatial unit grid, and γ inhale is the inhalation rate of the person.
The total amount of viruses inhaled by a person is the sum of the number of viruses contained in droplets of each size, while the number of viruses per droplet scales with droplet size. We assume that the droplets are spherical [57] and hence the volume can be estimated as shown by the solid line in Fig. 5 . Furthermore, we set ρ virus equals 2.35 × 109 viruses/mL fluid according to the findings of Wölfel et al. [58] and Villers et al. [59]. Finally, the individual's virus exposure is calculated as shown in Eq. (5).(5) Vinhale(t)=∑Dinhale(t)Vdropletsρvirus×(1−M)
Where V inhale (t) is the number of viruses inhaled by a person during (t-Δt, t), V droplets is the volume of droplets, ρ virus is the number of viruses per mL fluid, M denotes the filter efficiency of face masks against droplets. It is noted that here we have added M as a parameter indicating the filtration efficiency of the mask to measure the efficacy of mask interventions.Fig. 5 The volume and terminal speed of droplets of different sizes in the model.
Fig. 5
3.3.2 Surface-based transmission
A person can be exposed to the virus by touching a contaminated surface with their hands and then touching the mucous membranes of the body, which is surface-based transmission. It includes two processes: one is the contamination of environmental surfaces, while the other is the intake of viruses triggered by touching.
For the contamination process, there are two sources of viruses on indoor environmental surfaces. One is the deposition of droplets caused by activities such as speaking or coughing when an infected person is close to an environmental surface. The other is virus transmission due to an infected person touching a surface with their hands. Since the number of viruses transmitted by the hands of the infector is quite limited and difficult to determine, only the former source is considered here. That means the only way for an environmental surface to be contaminated is if it is spoken or coughed on, not through contact in this model.
When an infectious person speaks or coughs toward an object, the droplets may settle to the surface. Here, we assume that when the object is within the spreading distance and spreading angle of the infected person, the droplets in this cell will be evenly deposited on its surface, and the object becomes a "fomite" (Eq. (6)).(6) Fsurface(t)=Dcell(t)
Where F surface denotes the viruses on the environmental surfaces, and D cell denotes the viruses sprayed into the cell in the form of droplets.
When a susceptible person touches a contaminated surface with hands, the virus is transferred to his/her hands. Eq. (7) is used to calculate the number of viruses received by the hands.(7) Fhands(t)=Fsurface(t)TshAhandAsurface
Where F hands denotes the amount of virus that hands acquire from the environment susceptible individuals, and F surface represents the viruses on the environmental surfaces involved in the touch. T sh is the transfer efficiency from the environmental surface to the hands. A hand denotes the contaminated hand surface area. A surface is the average environmental surface area touched per hand contact. Virus transfer efficiency from a nonporous surface to a fingertip has been estimated to be 0.5% per touch per fingertip [56,60]. The authors could not locate published data on the virus transfer efficiency from porous surfaces to fingertips, but the transfer efficiency for bacteria from a porous surface has been estimated to be 0.1% per touch per fingertip [60]. Therefore, for all the indoor surfaces in general we assumed that T sh equals 0.3% [61]. A hand is set to 10 cm2, which is approximately the area of five fingertips [62]. A surface here is set to 50 cm × 50 cm = 2500 cm2, which is the area of a grid cell [63].
In addition, supermarkets have their particularity for surface-based transmission, since the virus is deposited on the surface of the goods while the goods will be continuously taken away due to shopping activity. We roughly assume that each cell representing a shelf is loaded with approximately 10 goods, so each time a customer takes away the goods in the grid, it will take away 1/10 of the virus on the surface of the cell.
For some environmental surfaces that people stay nearby for a relatively long time, such as counters in supermarkets, people will touch them with a certain frequency. Therefore, such high-touch surfaces are assumed to be touched by proximate individuals at a constant rate as shown in Eq. (8).(8) Fhands(t)=Fsurface(t)FshTshAhandAsurfaceΔt
Where F sh is the probability of hand contact with certain high-touch surfaces during (t-Δt, t). For example, in the supermarket case in this paper, we set the probability of people touching the counter surface as 0.2 times/min based on the observation of the supermarket counter [64,65].
Susceptible people usually touch their facial mucous membranes with a certain frequency, so that the viruses enter the body and lead to infection. Similar to contact with high-touch surfaces, we use Eq. (9) with a frequency index to calculate the amount of virus transferred during this process.(9) Eface(t)=Fhand(t)FhfThfAfaceAhandΔt
Where E face denotes the amount of virus that facial mucous membranes acquire from hands, and F hand denotes the viruses on the hands involved in the touch. F hf is the probability of hand contact with facial mucous membranes during (t-Δt, t). T hf is the transfer efficiency from hands to the facial membranes per touch. A face denotes the facial mucosal membranes area touched, and A hand is the contaminated hand surface area. F hf was set to 1.6 × 10−1 per min based on the observation of 26 persons who collectively touched facial mucosal membranes unconsciously 1024 times in 4 h [66,67]. T hf from a fingertip to facial mucosal membranes per touch was set to 0.35 [62,67]. Considering that the touch involves one fingertip of the five fingers on the same hand, A face/A hand was set to 0.2 [67].
3.4 Virus-building interaction modeling
3.4.1 Droplet deposition
Droplets may deposit to the surface of the environment due to the force of gravity after being ejected from the mouth. Droplets have different terminal speeds due to their different sizes and masses. According to Anchordoqui and Eugene [57], we assume that each droplet is spherical to obtain different terminal speeds for different sizes of droplets, which is shown by the dotted line in Fig. 5.
3.4.2 Droplet flow
Some small droplets can float in the air for a long time, which allows the droplets to move throughout the space via forced airflow before the droplets settle to the ground. Here we designed a simple forcing method to simulate indoor airflow (Fig. 6 ). In a mechanically ventilated space, the air will generally flow continuously from the supply vents to return vents. Therefore, special supply vent cells and return vent cells were designed in NetLogo, and the undeposited droplets in other cells will gradually move toward the return vent according to certain rules. For the indoor airflow process, we consider three important indicators: air path, air change rate, and air filtration rate. The mechanism of the simple forcing method is detailed in Appendix S2.Fig. 6 Simple forcing method for simulating indoor airflow by agent-based modeling.
Fig. 6
3.4.3 Droplet diffusion
In addition to flow by ventilation, droplets in the air will also spontaneously diffuse, which is caused by the diffusion properties of the gas itself. We set the droplets to diffuse to 8 surrounding cells with no walls or furniture at a rate of 1.5 × 10−3 m3/min according to the findings of Castillo and Weibel [68].
3.4.4 Virus decay
Viruses gradually decay in the air and on environmental surfaces. Half-life values are often used to describe the stability of viruses in different mediums. For the convenience of calculation, the variation of decay rate over time was simplified as linear regression. We assume that the viruses in the air decay at a rate of 1.27 × 10−4/s in terms of the half-life median of SARS-CoV-2 [25,69].
For the supermarket case we studied, an investigation of the surface material shows that the surfaces of the goods and shelves in the supermarket are mostly plastic, while the surfaces of the counters are made of stainless steel. Therefore, the virus decay rate on store shelves was set to 2.04 × 10−5/s [70], and the virus decay rate on counters was set to 2.47 × 10−5/s [70]. The decay rate here is the rate when the ambient temperature is 21 °C–23 °C and the relative humidity is 40%. Since the indoor temperature and humidity fluctuations are relatively small, the influence of indoor temperature and humidity variations on the decay rate is not considered [70].
3.5 Infection risk assessment
We modeled the relationship between exposure and infection risk using an exponential dose-response relationship [71] (Eq. (10)).(10) p(d)=1−exp(−dk)
Where p(d) is the risk of illness at the dose of d; and parameter k equals the reciprocal of the probability that a single pathogen will initiate the response. The value of k is set to 4.1 × 102 according to Watanabe et al.'s research on SARS Coronavirus [72]. In addition, viruses that enter the body from different mucous membranes are simplified here as being of equal efficacy.
The main parameter settings adopted for the indoor SARS-CoV-2 transmission model are summarized in Table 3 . It is worth noting that there are many variants of SARS-CoV-2 virus, e.g., Delta and Omicron, and these variants may have an impact on parameters such as the amount of virus exhaled by infected patients, the rate of viral decay and the pathogenic dose. For example, Lina et al. proved that more SARS-CoV-2 particles were exhaled by the Omicron patients than the Delta patients [73]. To avoid confounding of the results by variant factors, the original strain (D614) data were used in this paper to better facilitate comparison of the role of architectural and behavioral factors.Table 3 Parameters for the agent-based indoor SARS-CoV-2 transmission model.
Table 3Model parameter Value (s)/Distribution Reference (s)
Human-related parameters
Expectoration height 1.7 m [44]
Interaction intensity (A) 2000 N/m [74]
Interaction range (B) 0.5 m [75]
Relaxation time (Tα) 0.5 s [75]
Expected speed (ve) ve ∼ N (1.29, 0.262) m/s [76]
Pedestrian mass (mi) mi ∼ N (65, 52) kg [77]
Anisotropy index (λi) 0 [77]
Cough frequency 0.19 coughs/min [45,46]
Inhalation rate 0.023 m3 air/min [55,56]
Virus-related parameters
Droplets number (Dn) ln Dn ∼ N (9.7 × 105, (3.9 × 105)2)/expectoration [33,49]
Droplets spread angle-coughing (ωcoughing) 35° [53,54]
Droplets spread angle-not coughing (ωnot-coughing) 63.5° [53,54]
Droplets spread distance-coughing (Dd-coughing) Dd-coughing ∼ N (5, 0.2562) m [51]
Droplets spread distance-not coughing (Dd-not-coughing) Dd-not-coughing ∼ N (0.55, 0.0682) m [52]
Viruses density in droplets (ρvirus) 2.35 × 109 viruses/mL [58,59]
Transfer efficiency from surfaces to hands (Tsh) 3 × 10−3 [56,60,61]
Transfer efficiency from hands to facial membranes (Thf) 0.35 [62]
Hand surface area (Ahand) 10 cm2 [62]
Touch frequency with counter surface (Fsh) 0.2 times/min [64,65]
Touch frequency with target facial membranes (Fhf) 1.6 × 10−1 times/min [66,67]
Viruses' decay rate in the air 1.27 × 10−4/sec [25,69]
Viruses' decay rate on the plastic 2.04 × 10−5/sec [25,70]
Viruses' decay rate on the steel 2.47 × 10−5/sec [25,70]
Infection risk of a single virion (k) 4.1 × 102 [72]
Diffusion rate 1.5 × 10−3 m3/min [68]
Building-related parameters
Size of spatial cell 0.5 m × 0.5 m –
3.6 Case study
3.6.1 The underground supermarket and customer activity at Tianjin University, China
The case is an underground supermarket located in Tianjin University, China, as shown in Fig. 7 . The supermarket has a rectangular shape with an area of 217.5 m2. It has only one entrance and exit and the shelves are distributed in rows. There are five checkouts in the supermarket, but they are not always open. The supermarket is open from 9:00 to 17:00 daily. The supermarket has three air supply vents in the central part, but only one return vent in the corner. Most of the customers are students, who come to buy some daily necessities such as fruit and vegetables. Crowding and queuing often occur in the supermarket according to site investigation.Fig. 7 The underground supermarket at Tianjin University, China.
Fig. 7
Based on the detailed observations of the customers in the supermarket and previous studies [78,79], consumer behavior rules in the supermarket were developed (Fig. 8 ). We assume that each consumer enters the supermarket with a specifically defined shopping list [80]. The consumer first moves to the closest product, picks up the item and adds it to his/her hands or shopping cart, and then walks to the next closest product. Moreover, customers often want to buy something impromptu, or forget about something they want to buy and return to buy it according to some studies [81,82]. To describe the shopping behavior more accurately, we add a recall mechanism when the shopping list ends, i.e., if the consumer forgets to buy something, he/she can add the item to the shopping list and go get it. When all the items on the shopping list are picked up, the consumer will move towards the checkout stations.Fig. 8 The customer's activity flow and checkout selection strategies.
Fig. 8
According to the questionnaire survey of customers, customers mainly have several cashier selection strategies: (1) choosing the queue closest to him, (2) choosing the queue with the least number of customers, (3) choosing the queue with the least number of products in the hands/carts of waiting customers, and (4) choosing randomly. This is consistent with the findings of some studies on queuing behavior in supermarkets [83,84]. A random function was set up and customers will choose one of the above queuing strategies randomly. In addition, if there are too many customers in the current queue, which exceeds the patience of some customers, he/she will choose to leave the supermarket directly without checking out [85]. When queuing, the customer's route is a straight line, and the moving speed is determined by the checkout speed of the cashier. Finally, when payment is made, the customer leaves the supermarket through the exit. Fig. 8 presents the customer's activity flow and checkout selection strategies.
3.6.2 Architectural and behavioral interventions
Among prospective preventions [24,[86], [87], [88]], five types of possible and pragmatic architectural and behavioral interventions were considered: (a) spatial layout, (b) air ventilation, (c) customer control, (d) checkout control, and (e) sanitary measures.(a) Spatial layout
The spatial layout can change the movement trajectory of the crowd and form different gathering or contact situations, resulting in different infection risks. Taking the supermarket as the case, we mainly consider changing the layout of supermarket shelves and entrances/exits, which is an intervention that can be easily realized. Here, under the premise of ensuring the same quantity of goods, we designed three types of shelf layouts (with one exit), as shown in Fig. 9 (a) (b) (c). Supermarket Layout B is based on the special control measures for supermarkets in some areas of China during the epidemic period. According to the Operational Guidelines for Prevention and Control of COVID-19 Epidemic issued by the Chinese government, supermarkets, restaurants et al. may have one-way pedestrian flow restrictions to control indoor infection risk. Literatures such as Ying, F. and O'Clery, N [89]. also studied the effect of supermarket one-way aisle layout on indoor transmission of COVID-19. Therefore, we set up Layout B that restricts people to one-way walking routes to investigate the effect of one-way pedestrian flow layout on the indoor infection risk. On the other hand, Layout C was designed to cope with the increasing variety of supermarket shelf layout designs. The design of supermarket shelves is pushing forward, and some more flexible layouts are starting to emerge [90,91], instead of the traditional constant rows and columns. Therefore, we designed a flexible shelf Layout C to explore the impact of a looser planar organization on the indoor infection risk. In addition, two exits were added to the layout of the base scenario with only one exit to reduce the possibility of congestion as shown in Fig. 9 (d). After checking out customers will choose the nearest exit to leave.(b) Air ventilation
Fig. 9 Scenarios with different spatial layouts.
Fig. 9
Three factors are considered for the ventilation system, including air change rate, air-vent layout, and filter efficiency. For the air change rate, we assumed that the ACH of the supermarket was increased to 10 times/hour, 15 times/hour, and 20 times/hour from the baseline of 5 times/hour. For filter efficiency, we assume that the filter efficiency varies from 40% to 60%, 80%, and 100%. Additionally, three types of the air-vent layout were designed for comparison as shown in Fig. 10 .(c) Customer control
Fig. 10 Scenarios with different air-vent layouts.
Fig. 10
For customer control, four interventions were explored. One intervention is to control the maximum number of customers in the supermarket, including 50 people (base scenario), 30 people, and 20 people. The second is to control the time interval for the flow of people entering the supermarket, which are 5s (base scenario), 10s, and 15s respectively. The third is controlling physical distance when queuing. We assumed a greater physical distance of 1.0 m between customers while queuing to check out than the base scenario (0.5 m).(d) Checkout control
Checkout control refers to adjusting the open percentage of checkout, which are 50% (base scenario), 20%, 80%, and 100% open respectively.(e) Sanitary measures
According to previous reports on mask efficiency, we assumed that face mask-wearing could reduce the emission of viruses by 95% [92]. Furthermore, the use of masks reduced the probability of facial mucous membrane touch per min from 1.6 × 10−1 per min to 5.4 × 10−2 per min, because touches of the eyes accounted for 33% of the facial mucosal membrane touches involving the eyes, nose, and mouth [66]. In addition, if the infected person wears a mask, the spread distance and angle were set to 0. Here we assume three scenarios: the proportion of people wearing masks is 0% (base scenario), 50%, and 100% respectively. Surface cleaning can eliminate viruses on the surface of objects. We assumed that 99.9% of the viruses on the environmental surface were removed through decontamination when being cleaned [24]. According to our survey of several supermarkets, we found that the supermarkets on campus were mostly cleaned in the early morning, lunchtime and dinner time with an interval of approximately 4 h. Therefore, two cleaning scenarios were explored, the base scenario with no cleaning and cleaning every 4 h.
All the architectural and behavioral interventions involved are presented in Table 4 together with their variable values. In addition, the initial number of infections was set to 5%. The simulation period is 8 h (9:00–17:00). Due to the randomness of the simulation, we performed 10 simulations for each scenario to overcome the fluctuation of results due to randomness. A total of 38 × 10 = 380 simulations were performed, and each simulation took about 6 h on a desktop with an i7-CPU 3.60 GHz processor and 8 GB of RAM. The results of each metric are stored as NetLogo “list” data, and each data in the list corresponds to a time t. These data have two destinations: first, they are processed in NetLogo and programmed into “World” interface for real-time monitoring of spatial aggregation status, virus exposure and number of infections etc.; second, the multiple list data including the time list, are combined as a matrix and exported as.csv data, which was then imported into IBM SPSS Statistics software for further statistical analysis.Table 4 Descriptions and values of architectural and behavioral interventions.
Table 4Interventions Variables Values Unit
Spatial layout Shelf layout Layout (base), A, B, and C –
Exit Single exit (base), Multiple exits –
Air ventilation Air change rate 5 (base), 10, 15, 20 times/hour
Filter efficiency 0.4 (base), 0.6, 0.8, 1.0 –
Air-vent layout Layout (base), A, B, and C –
Customer control Maximum customer number 50 (base), 30, 20 –
Enter interval 5 (base), 10, 15 second
Queueing distance 0.5 (base), 1.0 m
Checkout control Checkout open percentage 20%, 50% (base), 80%, 100% –
Sanitary measures Mask percentage 0 (base), 50%, 100% –
Surface cleaning interval No cleaning (base), 4.0 hour
4 Results
4.1 The base scenario
4.1.1 Real-time situated visualization
An interface showing the movement of people and environmental pollution in real time was created. Fig. 11 shows the instantaneous state of a simulation of the base scenario at the fifth minute. Fig. 11 (a) shows the movement paths of all customers in the supermarket in the first 5 min. And Fig. 11 (b) is the heat map of the cumulated path, indicating the number of times people pass by here. It can be seen that the flow of people mainly moves along the three longitudinal aisles formed by the shelves, while the spaces in front of the checkout are the area where people pass the most. Crowding is easy to form in two local spaces (areas “A” and “B”) with high traffic near checkouts, which are relatively high-risk spaces for virus transmission.Fig. 11 The real-time visualization of customers’ status and environmental contamination level of a simulation for the base scenario at the fifth minute. Crowding is easy to form in two local spaces (areas “A” and “B”) with high traffic near checkouts. The degree of air contamination between the infected individual and the return air vent is higher than in other directions, and the shelves near the entrance have been polluted to a high degree.
Fig. 11
Fig. 11 (c) shows the air contamination level of the supermarket at 5 min. Red dots are infectious people who were originally infected by default. It can be seen that air contamination levels are much higher near infectious people. The instantaneous number of viruses in the cell where the infectious person is located can reach 7.36 × 104. The high-concentration virus air will float to the return vent after being exhaled, increasing the air contamination level on the right side of the infectious person. Fig. 11 (d) shows the contamination level of shelves and counter surfaces at 5 min. It indicates that the shelves near the entrance have been polluted to a high degree, which is most likely due to the large number of people passing by here. In other words, this real-time spatially-based visualization can detect high-risk local spaces that are prone to congestion or environmental surfaces prone to be contaminated.
4.1.2 Timeline-based analysis
The timeline-based analysis includes three levels: space-oriented analysis, environment-oriented analysis, and occupant-oriented analysis. Fig. 12 shows the timeline-based analysis of a base scenario simulation.Fig. 12 The 8-h variations of customer exposure and environmental contamination of a simulation for the base scenario of the supermarket.
Fig. 12
The space-oriented analysis includes two indicators: average interpersonal distance and customers in close contact. Average interpersonal distance refers to the average distance between every customer and other customers in the supermarket at a certain moment. It can be seen from Fig. 12 (a) that the average distance between people in the supermarket fluctuates around 12.8 m. Customers in close contact (Fig. 12 (b)) refers to the number of customers in the cone of the infectious people at each tick. It can be seen that the number of close contacts in the supermarket is 1 or 2 most of the time. Rarely do 3 or more close contacts occur at the same time. During the 8 h of simulation, a total of 33624 close contacts occurred.
The environment-oriented analysis includes two indicators: air contamination level and surface contamination level. Fig. 12 (c) shows the average air contamination level of the supermarket, namely the average number of viruses per cell. Results show that the instantaneous contamination level of indoor air can reach up to 3.59 × 104, and the average value is 6.53 × 103. Fig. 12 (d) shows the average surface contamination level variation. The overall upward trend of viruses on environmental surfaces is mainly due to the absence of surface cleaning measures. The decline of surface viruses can only be caused by the removal of goods during customers’ shopping process. The average surface contamination level of the supermarket is 1.90 × 108.
The occupant-oriented analysis includes customers' inhaled viruses, customers' touched viruses, and total exposed viruses as well as changes in the infection status of the supermarket population. Fig. 12 (e) shows the virus exposure of all customers through inhalation at each tick, and the peak value can reach 4.33 × 104. The cumulative respiratory exposure of all customers in the 8-h simulation period is 1.19 × 109. Fig. 12 (f) shows the variation of virus exposure through surface-based transmission, of which the highest value is 1.66 × 104. It can be seen that the virus exposure obtained through surface-based transmission is much smaller than the air-based transmission route, which is consistent with many existing studies [93,94]. This also resulted in the total exposure curve (Fig. 12 (g)) being very similar to the exposure curve via inhalation. Fig. 12 (h) shows the infection status of supermarket customers over time. The supermarket received a total of 5758 people in 8 h. When the initial infection rate was 5%, 29 people were eventually infected by entering the supermarket for shopping.
4.2 Effect of single intervention on customer exposures and infections
4.2.1 Spatial layout
Fig. 13 (a) shows the difference in virus exposure of customers caused by various spatial layout interventions. The mean cumulative virus exposure (Ev‾) of customers for the base scenario is 1.12 × 109. The Ev‾ under shelf layout A, shelf layout C and multiple exits scenarios all decreased, among which the virus exposure of multiple exits was the least, with an average of 7.57 × 108. However, the Ev‾ of shelf layout B increased significantly, with an average of 3.87 × 109.Fig. 13 Customers’ cumulative virus exposure (a) and percentage of probable infections (b) for different spatial layout scenarios.
Fig. 13
The percentage of probable infections (p i) has been chosen as the primary indicator of effectiveness as it is unaffected by the number of customers served. Here p i is defined as a ratio of newly infected and initially healthy customers (Eq. (11)):(11) pi=NinfectednewlyNtotal−N0
Where p i is the percentage of probable infections, N total is the total number of people entering the supermarket, and N 0 is the number of initially infected customers.
Fig. 13 (b) shows the p i distribution for different spatial interventions, which is consistent with the trend of virus exposure. The average p i (pi‾) under the base scenario is 0.41%, while the pi‾ of shelf layout B reaches 1.06%. On the other hand, the pi‾ of multiple exits decreased to 0.31%. The pi‾ of other distributed spatial layouts, namely layout B and layout C, also decreased. The single-aisle design of shelf layout B makes customers walk back and forth in the narrow aisle to find products, resulting in an increase in people's stay time and the number of close contacts. While the scattered layout of shelves and multiple exits can greatly reduce the congestion of people in supermarkets, thereby reducing the number of close contacts and the infection probability. This indicates that unreasonable space design can significantly increase the risk of infection, while proper space measures such as scattered furniture layout and multiple exits can effectively reduce the indoor infection risk.
4.2.2 Air ventilation
Fig. 14 shows the effect of different air ventilation interventions on cumulative virus exposure and infection risk. Trends in virus exposure and infection probability were similar across scenarios. As ACH increases from 5 times/hour to 20 times/hour, Ev‾ gradually decreases from 1.12 × 109 to 1.03 × 109, and pi‾ has also dropped from 0.41% to 0.32%. Furthermore, as the filter efficiency increased from 0.4 to 1.0, the Ev‾ decreased to 9.80 × 108, and pi‾ gradually decreased to 0.33%. As for different air-vent layouts, results show that the Ev‾ and pi‾ of air-vent layout A is slightly higher than that of the base scenario, while air-vent layouts B and C are much higher. The pi‾ of air-vent layout B and air-vent layout C reach 0.62% and 0.53% respectively. The possible reason is that though the number of vents was increased in this experiment, the ACH did not increase. More return vents resulted in indoor air flowing at a slower rate to the return vents, and contaminated air residing longer in the room, leading to more virus exposure. This suggests that we should not arbitrarily increase the number of air vents while the ACH remains the same, which will cause a greater risk of indoor infection.Fig. 14 Customers’ cumulative virus exposure (a) and percentage of probable infections (b) for different air ventilation scenarios.
Fig. 14
We posited that air-vent layout and filter efficiency may have cross-effects, so we added a set of control experiments to simulate three air-vent layouts with a filter efficiency of 1.0. Results show that the changing trends of the three air-vent layouts are consistent when the filter efficiency is 1.0 and 0.4, all of which are pi‾ (layout B) > pi‾ (layout C) > pi‾ (layout A) > pi‾ (layout base) (Fig. 15 ). This shows that increasing filter efficiency only uniformly reduces the E v and p i of different air-vent layouts, and does not change their trend, which means that the cross-effect between the two is very small.Fig. 15 Customers’ cumulative virus exposure (a) and percentage of probable infections (b) for various air-vent layouts under filter efficiency 0.4 and 1.0.
Fig. 15
4.2.3 Customer control
The effect of customer control interventions is shown in Fig. 16 . Results show that as the maximum customer number decreased from 50 (base scenario) to 20, the customer's Ev‾ decreased significantly to 3.76 × 108 and pi‾ is reduced to 0.25%. Furthermore, increasing the entry interval of customers has a more pronounced effect. When the entry interval increased from 5s (base scenario) to 10s, the Ev‾ decreased sharply to 2.85 × 108, and pi‾ decreased to 0.13%. And as the time interval further increased to 15s, the Ev‾ decreased to 1.62 × 108, and pi‾ decreased to 0.10%. Reducing the maximum number of customers and increasing the entry time interval will reduce the real-time number of people in the supermarket, thereby reducing the probability of contact between susceptible people and infectious people, leading to a decrease in infection risk.Fig. 16 Customers’ cumulative virus exposure (a) and percentage of probable infections (b) for different customer control scenarios.
Fig. 16
Furthermore, we were surprised to find that when the queue spacing rose from 0.5 m to 1.0 m, the mean virus exposure rose slightly to 1.18 × 108 and pi‾ increased slightly to 0.45%. This is likely due to the narrower aisle in front of the supermarket checkout in this case. The longer queue spacing led to more traffic congestion at the end of the queue, increasing the number of close contacts. This indicates that the queue spacing should be adjusted according to the actual space conditions and should not overly squeeze the traffic space.
4.2.4 Checkout control
Fig. 17 shows the variation of virus exposure and infection probability due to checkout control. Results show that when the checkout open proportion is reduced to 20%, the Ev‾ increases to 1.29 × 109 and pi‾ increases to 0.44%. When the proportion of open checkout is increased to 80% and 100%, the Ev‾ decreases to 1.11 × 109 and 1.08 × 109, and pi‾ decreases to 0.40% and 0.38%, respectively. It is obvious that the proportion of open percentage of checkout greatly change the infection risk of customers.Fig. 17 Customers’ cumulative virus exposure (a) and percentage of probable infections (b) for different checkout control scenarios.
Fig. 17
4.2.5 Sanitary measures
Sanitary measures include wearing masks and surface cleaning. Fig. 18 shows that when 50% of the population wears a mask, the Ev‾ drops dramatically to 6.76 × 108 and the pi‾ drops to 0.29%. And when 100% of the population wears masks, the Ev‾ drops to an even lower 2.29 × 108 and the pi‾ drops to a very low level of 0.08%. This suggests that wearing masks is a very useful measure to reduce the risk of indoor infection. However, the effect of surface cleaning measures was minimal. There was almost no change in Ev‾ with surface cleaning every 4 h, and pi‾ also remained at the original 0.41%. This indicates that the effect of surface-cleaning measures on the control of indoor infection is very limited. The main reason for this is that exposure via the surface-based transmission route is well below that of airborne transmission, and changes in virus exposure due to cleaning have little effect on the infection probability.Fig. 18 Customers’ cumulative virus exposure (a) and percentage of probable infections (b) for different sanitary measures.
Fig. 18
4.3 Effect of combined interventions on customer exposures and infections
After the parametric study of single interventions, we attempted to combine measures to determine the possible infection reduction potential of each category of measures. The best parameters for each category of interventions were combined as (1) combination of spatial interventions: shelf layout C + multiple exits, (2) combination of ventilation interventions: ACH 20 + filter efficiency 1.0 + Air-vent layout base, (3) combination of customer control interventions: max customer number 20 + enter interval 15s + queueing distance 1.0 m, (4) checkout control interventions: checkout open 100%, (5) combination of sanitary interventions: mask percentage 100% + surface cleaning interval 4 h, and (6) combination of all single interventions as an extremely harsh measure. It is worth noting that these combined designs may have been used in combination during the epidemic period, but that does not mean they are optimal, as there may be interactions between these variables.
Fig. 19 shows the effect of different categories of interventions on customer virus exposure and infection risk. It can be seen that customer behavior control is the most effective category of interventions, with a significant reduction in Ev‾ to 5.11 × 107 and a reduction in pi‾ to 0.04%. The reason lies in the strict combination of the maximum number of customers and the entry interval, which makes the customers finish the shopping process and leave the supermarket quickly in a very short time, largely avoiding the encounter with infectious people in the supermarket.Fig. 19 Customers’ cumulative virus exposure (a) and percentage of probable infections (b) for different combinations of measures.
Fig. 19
Furthermore, sanitary interventions were the second most effective. The combination of mask-wearing and surface cleaning reduces Ev‾ to 1.89 × 108 and pi‾ to 0.06%.
The third is the combination of spatial interventions. The scattered shelf layout and multiple exit design reduce Ev‾ to 7.71 × 108 and pi‾ to 0.26%. For checkout control, 100% open checkout reduces Ev‾ to 1.08 × 109 and pi‾ to 0.38%. The effectiveness of both types of interventions is to reduce the probability of customer encounters or stopping in the space, reduce the probability of close contact, and thus reduce the risk of infection.
The combination of ventilation interventions led to a decrease in Ev‾ to 9.64 × 108 and pi‾ to 0.29%. It is interesting to note that the effect of ventilation measures on viral exposure was not as significant as in several other categories, which differs from the general sense of perception. The reason is that the main transmission routes differ in various spatial scenarios due to different behaviors of people. In such high-traffic but low-stay venues, close contact transmission becomes the dominant transmission route. Therefore, it is more effective to reduce the probability of close contact by changing the spatial layout or controlling the behavior of customers. Ventilation interventions affect mainly aerosols that are suspended in the air for a long time. For some places where people remain relatively stationary for a long time (e.g., offices, wards, etc.), ventilation measures might be more effective when the cumulative inhalation of aerosols by people becomes the main infection factor.
Finally, the combination of "extremely strict" interventions resulted in an Ev‾ of only 3.04 × 106 and a pi‾ of 0. This proves the effectiveness of non-pharmaceutical measures, which may be used for supermarket control in some extreme outbreaks.
5 Discussion and future works
5.1 Application of the agent-based indoor SARS-CoV-2 transmission model
We developed an agent-based model that integrates building space, human behavior, and virus transmission to determine the efficacy of different types of interventions on indoor infection risk. The social force model simulating pedestrian movement and a simple forcing method simulating indoor airflow were coupled. Moreover, a user interface (Appendix S3) was designed so that the user can freely customize space configurations (e.g., changing shelf layout, adding exits) and occupant behaviors (e.g., entering interval, mask-wearing) to see the impact of each design decisions on infection risk. This approach can effectively help architects, supermarket managers, or the government to understand the impact of various measures on infection risk.
The model can be used to detect high-risk spaces, environmental surfaces, or behaviors. On one hand, real-time situated visualization shows spaces in buildings that are prone to crowding or congestion, and these are precisely the spaces at high risk of close contact. The model can show the contamination level of indoor air and surfaces, and diagnose localized spaces or surfaces that are vulnerable to contamination to take countermeasures. On the other hand, timeline-based analysis allows long-term monitoring of space, environment, or customers to determine the possible risks in the time dimension.
A relative comparison of the impacts of different measures is more valuable, which can help determine the effectiveness of the measures and take the most efficient measures. However, it should be noted that in different building scenarios, the impact of different measures is likely to be different. In supermarket buildings with a frequent flow of people, close contact is the main route of virus exposure. In this case, the proximity of people due to measures such as spatial layout and behavioral control becomes the main cause of infection. In some buildings such as offices and wards, people may stay in one spatial location for a long time. In this case, instead, people inhale aerosols for a long time and the accumulated viruses may become the main contributor to infection, and then ventilation will probably be a more important control factor. Therefore, each kind of building and behavior scenario should be analyzed specifically.
5.2 Limitations and future research
This agent-based model lacks validation, which is a problem that many infection prediction models currently face [95]. It is difficult to perform actual experimental validation and obtain data for the spread of the virus indoors. With the increase in medical tests, it may be possible to obtain some field data for validation in the future.
Second, the ventilation involved in the model only includes mechanical ventilation. Natural ventilation is too complicated to mimic in the ABM environment until now. Simulating natural ventilation in an ABM environment is challenging because it is affected by many factors such as wind speed, wind direction, air temperature difference between indoor and outdoor, and building layout. These parameters will have large fluctuations within a day, or even within an hour. Therefore, simulation of natural ventilation requires a high level of detail in modeling the physical environment. Moreover, the vertical airflow is not reflected in this model due to the limitations of the current ABM environment. However, due to the flattening of supermarket spaces with a length/width to height ratio greater than 3:1, horizontal airflow is more important than the vertical direction in this case. Furthermore, due to the initial spraying angle of the coughing action and the gravitational acceleration of droplets, most of the dispersal of the sprayed droplets is below the height of a person, i.e., below the level of 1.7 m height. The height of the supermarket shelf is approximately 1.6 m, which means that most of the droplets will be blocked by the shelf in the horizontal direction. Therefore, for the shelf in the supermarket we mainly focus on its airflow blocking effect in the horizontal direction. Coupled with the need to obtain indoor infection risks more efficiently and quickly to assist architects or managers in the decision-making process, complex vertical airflow was not considered in this study.
Third, the surface-based transmission simulation in the model needs to be supported by more measurement and experimental data. Supermarkets have their special characteristics in surface-based transmission because the virus falls on the goods, and the goods will be taken away continuously as the shopping process goes on. We roughly assume here that each grid is loaded with about 10 copies of goods. Each time the goods of the grid are taken away, 1/10 of the amount of virus of the grid will be taken away. However, the surface area varies depending on the goods, so it is difficult to estimate an accurate value of the surface area. For instance, vegetables and fruits cannot be fully “cleaned”. On the other hand, we also found that one of the major problems in simulating surface-based transmission is the lack of specific behavioral data. For example, what are the touching habits of people when selecting goods, and whether they will make some touching but not buy? These underlying data affect the accuracy of the model. Although some touch behaviors in offices have been studied by researchers such as Zhang et al. [20], this is a relatively uncharted territory that needs to be carefully studied and explored in the future.
In addition, there is a corresponding cost behind each intervention. For example, the best type of intervention in terms of reducing the probability of infection may be to restrict customer behavior, but this is likely to result in lower sales. Increasing ACH will increase operating costs while opening more checkouts will increase labor costs. This involves a decision balance problem, i.e., what decision is the most cost-effective? This model can provide benefits in terms of infection risk reduction for making more scientific decisions.
Finally, there is the problem of high time cost. In this paper, an 8-h supermarket simulation takes up to 6 h on a desktop with an i7-CPU 3.60 GHz processor. On the one hand, the use of a large number of cells to slice the space and store different sizes of droplet variables, which is similar to the CFD gridding simulation method, takes up a lot of memory and computation time. On the other hand, the social force model needs to calculate the driving force and repulsive force of each customer at each tick to arrive at the next position. This is more time-consuming than some algorithms that use fixed points and networks [89]. However, the problem of high time cost can be resolved by the use of hybrid models. For instance, Lutz and Giabbanelli [96] have developed machine leaning regression models for 4 COVID-19 ABMs to assist in fast decision-making. With the continuous improvement of hardware computing power and supervised learning algorithms, time cost will become less of a problem.
6 Conclusion
A spatially-explicit agent-based model for studying indoor respiratory pathogen transmission was proposed and used to demonstrate the potential effectiveness of multiple interventions for reducing SARS-CoV-2 transmission in the case study of a supermarket. The social force model simulating pedestrian movement and a simple forcing method simulating indoor airflow were coupled in the NetLogo modeling environment.
The results of the supermarket case study showed that for a single intervention, wearing a mask was the most effective, with all masks worn by the population reducing the pi‾ to 0.08%. In addition, customer control interventions were also quite effective, where tripling the entry interval reduced the pi‾ to 0.10%. Among the spatial interventions, the design of multiple exits was the most effective, reducing pi‾ to 0.31%. It was also found that some unreasonable space layouts can significantly increase the pi‾ and should be prevented in advance. Among the ventilation interventions, increasing the filter efficiency is the most effective, the pi‾ can be decreased to 0.33% when the filter efficiency is 1.0. Finally, opening all checkouts reduces pi‾ to 0.38%.
For the combination of measures, the customer control combination is the most effective and can reduce pi‾ to 0.04% by controlling the maximum number of customers, the customer entry interval, and the number of the customers shopping. The main reason is that customer control directly and significantly reduces the probability of contact between susceptible people and infected people in the same space. This is even more effective than the combination of sanitary measures, of which the pi‾ is 0.06%. Furthermore, the combination of space interventions through scattered shelf layouts and multiple exit design reduces pi‾ to 0.26%, which is more effective than the combination of ventilation interventions with a pi‾ of 0.29%. This is because close contact becomes the main source of virus exposure in buildings like supermarkets where people move frequently and stay less. Finally, the extremely stringent combination of all interventions can achieve an 8-h infection-free situation in supermarkets. This suggests that the risk of indoor infection can be reduced to a large extent by combining various non-pharmaceutical measures.
The findings of this study have some implications. The results of different architectural and behavioral interventions on indoor virus transmission can help architects, supermarket managers, and the government to better understand and choose epidemic prevention interventions to control indoor infection risk. Furthermore, this model can be easily adapted and applied to other building types and behavioral scenarios, particularly buildings with high traffic flow. User can freely customize space configurations and occupant behaviors to check the impact of decisions on indoor infection risk through the user-defined interface.
Finally, there are some limitations to this study which need to be considered. One of the limitations of this study is the lack of validation due to the difficulty of collecting virus data in the field and tracking the infection status of people. Future studies will use experimental or real-world measurements to further mine indoor viral and human behavior data and validate the model to enhance its accuracy.
CRediT authorship contribution statement
Anxiao Zhang: Conceptualization, Methodology, Software, Visualization, Writing – original draft. Qi Zhen: Data curation, Formal analysis, Writing – review & editing. Chi Zheng: Investigation, Validation. Jing Li: Data curation, Supervision. Yue Zheng: Data curation, Supervision. Yiming Du: Supervision. Qiong Huang: Supervision. Qi Zhang: Funding acquisition.
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.
Appendix A Supplementary data
The following are the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Multimedia component 2
Multimedia component 2
Multimedia component 3
Multimedia component 3
Data availability
Data will be made available on request.
Acknowledgments
This research was supported by the Key Laboratory of Ecology and Energy Saving Study of Dense Habitat, Ministry of Education, China (No. 20220109), the 10.13039/501100001809 National Natural Science Foundation of China (No. 72174138), and the Independent Innovation Fund of Tianjin University (No. 2023XS-0098). We thank the Xuesi Underground Supermarket of Tianjin University for data supporting.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jobe.2023.106807.
==== Refs
References
1 Smith Katherine F. Global rise in human infectious disease outbreaks J. R. Soc. Interface 11 101 2014 20140950
2 World Health Organization WHO Coronavirus Disease (COVID-19) Dashboard 2021 https://covid19.who.int/. Accessed on November 17, 2022
3 Bulfone Tommaso Celeste Outdoor transmission of SARS-CoV-2 and other respiratory viruses: a systematic review J. Infect. Dis. 223 4 2021 550 561 33249484
4 Cadoni Mariano Gaeta Giuseppe How Long Does a Lockdown Need to Be? 2020 11633 arXiv preprint arXiv: 2004
5 Nishiura Hiroshi Closed environments facilitate secondary transmission of coronavirus disease 2019 (COVID-19) medRxiv 2020
6 Gallo Marin Benjamin Predictors of COVID‐19 severity: a literature review Rev. Med. Virol. 31 1 2021 1 10
7 Xiang Yue COVID-19 epidemic prediction and the impact of public health interventions: a review of COVID-19 epidemic models Infect. Dis. Model. 6 2021 324 342 33437897
8 Peng Zhe Practical indicators for risk of airborne transmission in shared indoor environments and their application to COVID-19 outbreaks Environ. Sci. Technol. 56 2 2022 1125 1137 34985868
9 Moon Jinyoung Ryu Byung-Han Transmission Risks of Respiratory Infectious Diseases in Various Confined Spaces: A Meta-Analysis for Future Pandemics 2021 Environmental research 111679 202
10 Qian Hua Spatial distribution of infection risk of SARS transmission in a hospital ward Build. Environ. 44 8 2009 1651 1658
11 Wu Yan Tung Thomas CW. Niu Jian-lei On-site measurement of tracer gas transmission between horizontal adjacent flats in residential building and cross-infection risk assessment Build. Environ. 99 2016 13 21 32288039
12 Hong Tianzhen Occupant behavior models: a critical review of implementation and representation approaches in building performance simulation programs Build. Simulat. 11 1 2018 Springer Berlin Heidelberg
13 Langevin Jared Jin Wen Gurian Patrick L. Simulating the human-building interaction: development and validation of an agent-based model of office occupant behaviors Build. Environ. 88 2015 27 45
14 Zhang Anxiao Agent-based modelling of occupants' clothing and activity behaviour and their impact on thermal comfort in buildings IOP Conference Series: Earth and Environmental Science vol. 329 2019 IOP Publishing 1
15 Duives Dorine The Multi-Dimensional Challenges of Controlling SARS-CoV-2 Transmission in Indoor Spaces: Insights from the Linkage of a Microscopic Pedestrian Simulation and Virus Transmission Models 2021 medRxiv
16 Zhen Qi Overview of the role of spatial factors in indoor SARS-CoV-2 transmission: a space-based framework for assessing the multi-route infection risk Int. J. Environ. Res. Publ. Health 19 17 2022 11007
17 Dbouk Talib Drikakis Dimitris On coughing and airborne droplet transmission to humans Phys. Fluids 32 5 2020 053310
18 Kniesburges Stefan Effects of surgical masks on aerosol dispersion in professional singing J. Expo. Sci. Environ. Epidemiol. 32 5 2022 727 734 34611302
19 Jia Wei Exposure and respiratory infection risk via the short-range airborne route Build. Environ. 219 2022 109166
20 Zhang Nan Real human surface touch behavior based quantitative analysis on infection spread via fomite route in an office Build. Environ. 191 2021 107578
21 Li Yuguo Evidence for Probable Aerosol Transmission of SARS-CoV-2 in a Poorly Ventilated Restaurant 2020 MedRXiv
22 Vuorinen Ville Modelling aerosol transport and virus exposure with numerical simulations in relation to SARS-CoV-2 transmission by inhalation indoors Saf. Sci. 130 2020 104866
23 Arpino F. A Eulerian-Lagrangian approach for the non-isothermal and transient CFD analysis of the aerosol airborne dispersion in a car cabin Build. Environ. 209 2022 108648
24 Kampf Günter Persistence of coronaviruses on inanimate surfaces and their inactivation with biocidal agents J. Hosp. Infect. 104 3 2020 246 251 32035997
25 Van Doremalen Neeltje Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1 N. Engl. J. Med. 382 16 2020 1564 1567 32182409
26 Li Y. Qian H. Hang J. Chen X. Cheng P. Ling H. …Kang M. Probable airborne transmission of SARS-CoV-2 in a poorly ventilated restaurant Build. Environ. 196 2021 107788
27 Motamedi H. Shirzadi M. Tominaga Y. Mirzaei P.A. CFD modeling of airborne pathogen transmission of COVID-19 in confined spaces under different ventilation strategies Sustain. Cities Soc. 76 2022 103397
28 Peng Z. Rojas A.P. Kropff E. Bahnfleth W. Buonanno G. Dancer S.J. …Jimenez J.L. Practical indicators for risk of airborne transmission in shared indoor environments and their application to COVID-19 outbreaks Environ. Sci. Technol. 56 2 2022 1125 1137 34985868
29 D’Orazio, M., Bernardini, G., & Quagliarini, E. A probabilistic model to evaluate the effectiveness of main solutions to COVID-19 spreading in university buildings according to proximity and time-based consolidated criteria. In Building Simulation(Vol. vol. 14, pp. 1795-1809). Tsinghua University Press.
30 Antczak T. Skorupa B. Szurlej M. Weron R. Zabawa J. Simulation modeling of epidemic risk in supermarkets: investigating the impact of social distancing and checkout zone design Computational Science–ICCS 2021: 21st International Conference, Krakow, Poland, June 16–18, 2021, Proceedings, Part I 2021, June Springer International Publishing Cham 26 33
31 Alvarez Castro D. Ford A. 3D agent-based model of pedestrian movements for simulating COVID-19 transmission in university students ISPRS Int. J. Geo-Inf. 10 8 2021 509
32 Harweg T. Bachmann D. Weichert F. Agent-based simulation of pedestrian dynamics for exposure time estimation in epidemic risk assessment J. Public Health 2021 1 8
33 Farthing T.S. Lanzas C. Assessing the efficacy of interventions to control indoor SARS-Cov-2 transmission: an agent-based modeling approach Epidemics 37 2021 100524
34 Lee B. Lee M. Mogk J. Goldstein R. Bibliowicz J. Brudy F. Tessier A. Designing a multi-agent occupant simulation system to support facility planning and analysis for covid-19 Designing Interactive Systems Conference 2021 2021, June 15 30
35 Ronchi E. Lovreglio R. EXPOSED: an occupant exposure model for confined spaces to retrofit crowd models during a pandemic Saf. Sci. 130 2020 104834
36 Guo Zhan Loo Becky PY. Pedestrian environment and route choice: evidence from New York City and Hong Kong J. Transport Geogr. 28 2013 124 136
37 Johnson Donald B. A note on Dijkstra's shortest path algorithm J. ACM 20 3 1973 385 388
38 Dechter Rina Judea Pearl Generalized best-first search strategies and the optimality of A J. ACM 32.3 1985 505 536
39 Duchoň František Path planning with modified a star algorithm for a mobile robot Procedia Eng. 96 2014 59 69
40 Henderson L.F. The statistics of crowd fluids Nature 229 1971 381 383 5284 16059256
41 Helbing Dirk Farkas Illés Vicsek Tamas Simulating dynamical features of escape panic Nature 407 2000 487 490 6803 11028994
42 Hoogendoorn Serge P. Pedestrian flow modeling by adaptive control Transport. Res. Rec. 1878 1 2004 95 103
43 Helbing Dirk Molnar Peter Social force model for pedestrian dynamics Phys.Rev. E51 5 1995 4282
44 Zheng Jinping Zhong Nanshan Normative values of pulmonary function testing in Chinese adults Chin. Med. J. 115 1 2002 50 54
45 Lee Kai K. Four-hour cough frequency monitoring in chronic cough Chest 142 5 2012 1237 1243 22797468
46 Ni X. Ouyang W. Jeong H. Kim J.T. Tzavelis A. Mirzazadeh A. …Rogers J.A. Automated, multiparametric monitoring of respiratory biomarkers and vital signs in clinical and home settings for COVID-19 patients Proc. Natl. Acad. Sci. USA 118 19 2021 e2026610118
47 Duguid J.P. The size and the duration of air-carriage of respiratory droplets and droplet-nuclei Epidemiol. Infect. 6 1946 471 479
48 Chao Christopher Yu Hang Characterization of expiration air jets and droplet size distributions immediately at the mouth opening J. Aerosol. sci.40 2 2009 122 133
49 Miller Shelly L. Transmission of SARS‐CoV‐2 by inhalation of respiratory aerosol in the Skagit Valley Chorale superspreading event Indoor Air 31 2 2021 314 323 32979298
50 Railsback Steven F. Grimm Volker Agent-based and Individual-Based Modeling: a Practical Introduction 2019 Princeton university press
51 Bourouiba Lydia Dehandschoewercker Eline Bush John WM. Violent expiratory events: on coughing and sneezing J. Fluid Mech. 745 2014 537 563
52 Das Santosh K. Transmission of airborne virus through sneezed and coughed droplets Phys. Fluids 32 9 2020 097102
53 Kwon Soon-Bark Study on the initial velocity distribution of exhaled air from coughing and speaking Chemosphere 87 11 2012 1260 1264 22342283
54 Gupta J.K. Lin C.H. Chen Q. Flow dynamics and characterization of a cough Indoor Air 19 6 2009 517 525 19840145
55 Adams W.C. Measurement of Breathing Rate and Volume in Routinely Performed Daily Activities 1993 Final Report ContractA033-205
56 Rheinbaben F.V. Transmission of viruses via contact in ahousehold setting: experiments using bacteriophage φX174 as a model virus J. Hosp. Infect. 46 1 2000 61 66 11023725
57 Anchordoqui Luis A. Chudnovsky Eugene M. A Physicist View of COVID-19 Airborne Infection through Convective Airflow in Indoor Spaces 2020 13689 arXiv preprint arXiv:2003
58 Wölfel Roman Virological assessment of hospitalized patients with COVID-2019 Nature 581.7809 2020 465 469 32235945
59 Villers J. Henriques A. Calarco S. Rognlien M. Mounet N. Devine J. …Keiser O. SARS-CoV-2 Aerosol Transmission in Schools: the Effectiveness of Different Interventions 2021 MedRxiv 2021-08
60 Sattar S.A. Transfer of bacteria from fabrics to hands and other fabrics: development and application of a quantitative method using Staphylococcus aureus as a model J. Appl. Microbiol. 90.6 2001 962 970 11412326
61 Nicas Mark Best Daniel A study quantifying the hand-to-face contact rate and its potential application to predicting respiratory tract infection J. Occup. Environ. Hyg. 5.6 2008 347 352 18357546
62 Nicas Mark Sun Gang An integrated model of infection risk in a health‐care environment Risk Anal. 26.4 2006 1085 1096 16948699
63 Gao Caroline X. Multi-route respiratory infection: when a transmission route may dominate Sci. Total Environ. 752 2021 141856
64 Lei Hao Hand hygiene and surface cleaning should be paired for prevention of fomite transmission Indoor Air 30.1 2020 49 59 31545534
65 Donelan A.K. Chambers D.H. Chambers IV E.D.G.A.R. Godwin S.L. Cates S.C. Consumer poultry handling behavior in the grocery store and in-home storage J. Food Protect. 79 4 2016 582 588
66 Kwok Yen Lee Angela Jan Gralton McLaws Mary-Louise Face touching: a frequent habit that has implications for hand hygiene Am. J. Infect. Control 43.2 2015 112 114 25637115
67 Murakami Michio COVID-19 risk assessment at the opening ceremony of the Tokyo 2020 Olympic Games Microbial Risk Anal. 19 2021 100162
68 Castillo Julian E. Weibel Justin A. A point sink superposition method for predicting droplet interaction effects during vapor-diffusion-driven dropwise condensation in humid air Int. J. Heat Mass Tran. 118 2018 708 719
69 Dabisch P. Schuit M. Herzog A. Beck K. Wood S. Krause M. …Ratnesar-Shumate S. The influence of temperature, humidity, and simulated sunlight on the infectivity of SARS-CoV-2 in aerosols Aerosol. Sci. Technol. 55 2 2021 142 153
70 Aboubakr H.A. Sharafeldin T.A. Goyal S.M. Stability of SARS‐CoV‐2 and other coronaviruses in the environment and on common touch surfaces and the influence of climatic conditions: a review Transbound. Emerg. Dis. 68 2 2021 296 312 32603505
71 Nicas Mark An analytical framework for relating dose, risk, and incidence: an application to occupational tuberculosis infection Risk Anal. 16.4 1996 527 538 8819343
72 Watanabe Toru Development of a dose‐response model for SARS coronavirus Risk Anal.: Int. J. 30.7 2010 1129 1138
73 Lina L. Zhendong G. Sevalie S. Fangfang Z. Dawei Z. Weiwei C. …Zongzheng Z. Comparison of the amount of SARS-CoV-2 exhaled by Delta and Omicron patients J. Infect. 85 5 2022 573 607
74 Tang Ming Jia Hongfei An approach for calibration and validation of the social force pedestrian model Proceedings 2011 International Conference on Transportation, Mechanical, and Electrical Engineering (TMEE) 2011 IEEE
75 Lakoba Taras I. Kaup David J. Finkelstein Neal M. Modifications of the Helbing-Molnar-Farkas-Vicsek social force model for pedestrian evolution Simulation 2005 339 352 81.5
76 Montufar Jeannette Pedestrians' normal walking speed and speed when crossing a street Transport. Res. Rec. 2007 90 97 2002.1
77 Zhu Qiankun Du Yongfeng Simulation of crowd motion based on social force model Proceedings of the 19th CCSSTA 2018 (in Chinese)
78 Terano T. Kishimoto A. Takahashi T. Yamada T. Takahashi M. Agent-based in-store simulator for analyzing customer behaviors in a super-market Knowledge-Based and Intelligent Information and Engineering Systems: 13th International Conference, KES 2009, Santiago, Chile, September 28-30, 2009, Proceedings, Part II vol. 13 2009 Springer Berlin Heidelberg 244 251
79 Bloch P.H. Ridgway N.M. Sherrell D.L. Extending the concept of shopping: an investigation of browsing activity J. Acad. Market. Sci. 17 1989 13 21
80 Block L.G. Morwitz V.G. Shopping lists as an external memory aid for grocery shopping: influences on list writing and list fulfillment J. Consum. Psychol. 8 4 1999 343 375
81 Bellini S. Cardinali M.G. Grandi B. A structural equation model of impulse buying behaviour in grocery retailing J. Retailing Consum. Serv. 36 2017 164 171
82 Bell D.R. Ho T.H. Tang C.S. Determining where to shop: fixed and variable costs of shopping J. Market. Res. 35 3 1998 352 369
83 Zhang Q. Multi-agent based supermarket queuing model and optimization Advances in Multimedia, Software Engineering and Computing Vol. 2: Proceedings of the 2011 MSEC International Conference on Multimedia, Software Engineering and Computing, November 26–27 2012 Springer Berlin Heidelberg Wuhan, China 635 640
84 Wang J. Zhou Y.P. Impact of queue configuration on service time: evidence from a supermarket Manag. Sci. 64 7 2018 3055 3075
85 Maister D.H. The Psychology of Waiting Lines 1984 Harvard Business School Boston 71 78
86 Boyce John M. Didier Pittet Guideline for hand hygiene in health-care settings: recommendations of the healthcare infection control practices advisory committee and the HICPAC/SHEA/APIC/IDSA hand hygiene task force Infect. Control Hosp. Epidemiol. 23.S12 2002 S3 S40 12515399
87 Chu Derek K. Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19: a systematic review and meta-analysis Lancet 395.10242 2020 1973 1987 32497510
88 Johnson D.F. A quantitative assessment of the efficacy of surgical and N95 masks to filter influenza virus in patients with acute influenza infection Clin. Infect. Dis. 49.2 2009 275 277 19522650
89 Ying F. O’Clery N. Modelling COVID-19 transmission in supermarkets using an agent-based model PLoS One 16 4 2021 e0249821
90 https://www.archdaily.com/481321/supermarket-in-athens-klab-architecture.
91 https://www.archdaily.com/880960/ginza-loft-schemata-architects.
92 Rengasamy Samy Filtration performance of FDA-cleared surgical masks J. Int. Soc.Respirat. Protect. 26.3 2009 54
93 Razzini Katia SARS-CoV-2 RNA Detection in the Air and on Surfaces in the COVID-19 Ward of a Hospital in Milan, Italy vol. 742 2020 Science of The Total Environment 140540
94 Lee Sang-Eun Detection of novel coronavirus on the surface of environmental materials contaminated by COVID-19 patients in the Republic of Korea Osong Publ. Health Res.Perspect. 11.3 2020 128
95 Lee Bokyung Designing a multi-agent occupant simulation system to support facility planning and analysis for COVID-19 Designing Interactive Systems Conference 2021 2021
96 Lutz C.B. Giabbanelli P.J. When do we need massive computations to perform detailed COVID‐19 simulations? Adv. Theor. Simulat. 5 2 2022 2100343
|
PMC010xxxxxx/PMC10195650.txt |
==== Front
Nat Rev Endocrinol
Nat Rev Endocrinol
Nature Reviews. Endocrinology
1759-5029
1759-5037
Nature Publishing Group UK London
37202590
846
10.1038/s41574-023-00846-z
Comment
Mentorship in academic medicine: truth is in the eye of the beholder
http://orcid.org/0000-0001-9284-6289
Fleseriu Maria fleseriu@ohsu.edu
123
Lim Dawn Shao Ting 4
1 grid.5288.7 0000 0000 9758 5690 Department of Medicine, Division of Endocrinology, Diabetes and Clinical Nutrition, Oregon Health & Science University, Portland, OR USA
2 grid.5288.7 0000 0000 9758 5690 Department of Neurological Surgery, Oregon Health & Science University, Portland, OR USA
3 grid.5288.7 0000 0000 9758 5690 Pituitary Center, Oregon Health & Science University, Portland, OR USA
4 grid.163555.1 0000 0000 9486 5048 Department of Endocrinology, Singapore General Hospital, Singapore, Singapore
18 5 2023
2023
19 7 373374
© Springer Nature Limited 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The path to becoming a clinical academic researcher is arduous and convoluted, with many hurdles. A good mentor is key to growth and development, not only as one embarks on the journey, but also as a ‘sounding board’ throughout one’s career.
Subject terms
Endocrine system and metabolic diseases
Outcomes research
issue-copyright-statement© Springer Nature Limited 2023
==== Body
pmcA mentor is an empathic leader, teacher, educator, role model and more. The word itself derives from Homer’s Mentor, the wise advisor, guide and teacher to Odysseus and Telemachus. While far from ‘walks and talks’ in Athenian gardens, with disciples ‘absorbing’ and writing on tablets (stone, not the iPad…), the importance of one-on-one human connection cannot be understated. A similar, yet distinct, mentorship style, the ‘apprentice’ model, was key during the Renaissance. Aspects of this model remain today; the creation of a continuous, longitudinal relationship, often formed over many years, with a mentee learning and developing themselves and their work; in other words, residency and fellowship, no?
As I reflect on my journey (medical, clinical and research training in Romania, Europe and the USA), I appreciate having been exposed to many regions and countries, and thus diverse teams and mentors. Embarking on a second career, as an intern in the USA, was humbling and refreshing, highlighting an essential need for good mentors. Today, I am able to reflect on giving back (or passing forward) to colleagues, supporting trainees and fellows, and honouring my mentors’ support and encouragement.
Who is a ‘good’ mentor?
Mentor–mentee relationships involve communication, commitment and clarity1. A good mentor should have a proven track record of academic success and experience, knowledge of working in diverse and inclusive environments, and the ability to support mentees from various experiences and backgrounds, fostering honest and unbiased collaborations1–3. Being a good mentor takes time (a lot) and dedication, and it is important to be realistic about availability, while continuing to build one’s own professional network2. Shared goals as applied to predefined needs, trust and mutual respect, and encouraging self-reflection, are imperative. Scheduled guidance meetings ensure timely support and constructive feedback along the way.
When asked what makes a good mentor, I typically respond, “it depends.” What does the mentee want and what are their needs? Some might be searching for a role model to emulate, while others might seek active guidance. Do they value flexibility, availability and responsiveness, or do they prefer to be set clear, measurable goals? The mentor should be a good listener, be empathetic and strive for continuous improvement4, but it is also important to mesh well with the mentee’s goals and personality and to continuously evaluate the relationship1.
Mentorship can also extend beyond the confines of academia. It is well known that work–life balance is important and can reduce burnout. Prioritizing one’s well-being and time with family and friends has become crucial in sustaining a career; I freely admit to needing some mentorship myself here…
I communicate with trainees and mentees the importance of immersion and investment (personal and work) in themselves. Take what you know, study it again, and ask questions in your clinic, of every patient. This approach is especially important for patients with rare diseases and should serve as a canvas for creative and innovative thinking. What can you do to improve this patient’s life; or what can you improve on for the next patient?
I am passionate about increasing women’s participation in research. The problem is often oversimplified as a lack of supply and demand; representation in science, technology, engineering and mathematics (STEM), while improving, requires more work4–6. We should encourage women to stay in STEM, nominate them as conference speakers, for leadership positions3 and medical advisory boards2, and continue to create an inclusive, diverse and equitable research environment. More than one mentor, at various career stages and sometimes from other specialties, might be advisable. Small-group mentoring, especially linked to professional society meetings, can be very helpful. Sponsorship mentoring is a more recent concept, whereby a sponsor acts as a mentee advocate who provides access to key decision makers, who can in turn assist with professional development7.
I had superb mentors at various stages throughout my career, for whom I am forever grateful; in their honour, I strive to be even better! Mentorship, both formal and informal, is crucial and rewarding.
The mentee perspective
In the words of Yoda, “Always two there are. No more. No less. A master and an apprentice” (Star Wars: Episode I – The Phantom Menace). Examples include Yoda and Luke Skywalker, Professor Dumbledore and Harry Potter, Master Shifu and Po, and modern adaptations of Homer’s Mentor and Telemachus. These examples capture the spirit of mentorship, described as “the process whereby an experienced, highly regarded, empathic person (the mentor), guides the mentee for development and re-examination of their own ideas, learning and personal and professional growth by listening and talking in confidence to the mentee”8. Proven benefits for mentees include higher academic self-reliance with regards to clinical skills, research output and teaching, increased job satisfaction, a sense of empowerment, enhanced networking and career advancement1,4,5,9.
It is challenging for women in academic medicine to navigate the roles of clinician, educator and scholar. The wise counsel of women who have ‘been there and done that’ serves to motivate further clinical excellence and academic endeavours. Mentees report positive outcomes in perceived skills and self-esteem, career satisfaction and personal well-being3.
A mentor should be a teacher, counsellor, guide and role model. Ideally, mentees should self-identify their mentor; someone with a reputation for excellence, who is passionate in teaching and generous in sharing both personal and professional experiences. A good mentor is an advocate; they seek opportunities for and take pleasure in the mentee’s success. They are responsive, readily available for advice and help, yet provide space for maturation. A good mentor inspires. The mentee is assured a safe space in which to flourish with constructive feedback; doubts can be clarified without fear of reproach. Frank discussion and ideas are exchanged, and opinions on professional, academic, research and personal issues can occur10. However, it takes two to tango; a proactive attitude, commitment and respect for time and effort from the mentee are just as crucial to a successful mentor–mentee relationship. Early setting of clear, realistic and measurable goals, self-reflection and openness to feedback are needed1. In the digital era, mentoring knows no geographic boundaries. The COVID-19 pandemic has only heightened the importance of distance mentoring. Furthermore, mentor–mentee relationships evolve as both parties move along their career.
My journey with Dr Fleseriu did, however, start literally with a flight across the Pacific Ocean. Following endocrinology specialist training in Singapore, my interest in pituitary diseases led me to complete a fellowship at Oregon Health & Science University in Portland, USA. I anticipated that training with Dr Fleseriu, an internationally acclaimed expert in pituitary diseases, would be an enriching experience; one that would provide deeper insight into managing patients with neuroendocrine disorders and research on acromegaly and Cushing disease. Initially, when I was doubtful of my chances of securing a place as an international fellow, serendipity intervened! I was able to personally talk with Dr Fleseriu when she flew to Singapore to present her work at a conference.
Throughout the year-long fellowship, Dr Fleseriu provided me with many opportunities to grow, including hands-on management of challenging cases, participation in clinical trials, conducting and presenting research at international conferences and writing manuscripts. She was passionate about sharing her knowledge and experiences and there was ample opportunity to ask questions and clarify doubts. I was challenged to think through problems and offer solutions, which gave me a sense of independence, but a confidence that difficult situations could be tackled together. She was constantly looking for research opportunities for me and the other fellows (and medical students). She had many ideas (perhaps many is an understatement) but allowed us to take ownership of our direction and goals. She either provided us with resources or directed us to others who had them, encouraged networking with other experts in the field and related specialties, and was always there to provide feedback and encouragement (and plenty of laughter).
Over the past few years, after my return to Singapore, Dr Fleseriu has continued to engage with me virtually, providing a multitude of opportunities to continue publishing and to work with other world-renowned academic clinicians in the pituitary field. She encourages me in academic pursuits, yet is mindful of work–life balance and family commitments. She has been a guide, an advocate, a cheerleader and so much more. Her generosity through the years in sharing her time, talent and life will always be an inspiration for me to pass forward.
Competing interests
The authors declare no competing interests.
==== Refs
References
1. Straus SE Johnson MO Marquez C Feldman MD Characteristics of successful and failed mentoring relationships: a qualitative study across two academic health centers Acad. Med. 2013 88 82 89 10.1097/ACM.0b013e31827647a0 23165266
2. Shroff RT Where are all the women in industry advisory boards? J. Clin. Oncol. 2023 41 1659 1663 10.1200/JCO.21.02219 36331246
3. Laver KE A systematic review of interventions to support the careers of women in academic medicine and other disciplines BMJ Open 2018 8 e020380 10.1136/bmjopen-2017-020380 29572397
4. Brizuela V Chebet JJ Thorson A Supporting early-career women researchers: lessons from a global mentorship programme Glob. Health Action 2023 16 2162228 10.1080/16549716.2022.2162228 36705071
5. Sambunjak D Straus SE Marusić A Mentoring in academic medicine: a systematic review J. Am. Med. Assoc. 2006 296 1103 1115 10.1001/jama.296.9.1103
6. Ruzycki SM Fletcher S Earp M Bharwani A Lithgow KC Trends in the proportion of female speakers at medical conferences in the United States and in Canada, 2007 to 2017 JAMA Netw. Open 2019 2 e192103 10.1001/jamanetworkopen.2019.2103 30977853
7. Thompson K Taylor E Inclusive mentorship and sponsorship Hand Clin. 2023 39 43 52 10.1016/j.hcl.2022.08.012 36402525
8. Oxley, J. and Standing Committee on Postgraduate Medical and Dental Education. Supporting Doctors and Dentists at Work: An Enquiry into Mentoring (SCOPME, 1998).
9. Feldman MD Arean PA Marshall SJ Lovett M O’Sullivan P Does mentoring matter: results from a survey of faculty mentees at a large health sciences university Med. Educ. Online 2010 15 5063 10.3402/meo.v15i0.5063
10. Sng JH Mentoring relationships between senior physicians and junior doctors and/or medical students: a thematic review Med. Teach. 2017 39 866 875 10.1080/0142159X.2017.1332360 28562193
|
PMC010xxxxxx/PMC10197031.txt |
==== Front
Eur Geriatr Med
Eur Geriatr Med
European Geriatric Medicine
1878-7649
1878-7657
Springer International Publishing Cham
37204681
790
10.1007/s41999-023-00790-1
Research Paper
Clinical profile of trazodone users in a multisetting older population: data from the Italian GeroCovid Observational study
http://orcid.org/0000-0003-1687-4493
Coin Alessandra alessandra.coin@unipd.it
1
Noale Marianna 2
Gareri Pietro 3
Trevisan Caterina 14
Bellio Andrea 1
Fini Filippo 1
Abbatecola Angela Marie 5
Del Signore Stefania 6
Malara Alba 7
Mossello Enrico 8
Fumagalli Stefano 8
Volpato Stefano 4
Monzani Fabio 9
Bellelli Giuseppe 10
Zia Gianluca 6
Incalzi Raffaele Antonelli 11
The GeroCovid Observational Working GroupAbbatecola Angela Marie
Andrieri Domenico
Antognoli Rachele
Incalzi Raffaele Antonelli
Antonietti Maria Paola
Bagalà Viviana
Bandini Giulia
Bazzano Salvatore
Bellelli Giuseppe
Bellio Andrea
Bellotti Federico
Benvenuti Enrico
Bergamin Marina
Bertolotti Marco
Biagini Carlo Adriano
Bianchetti Angelo
Bianchi Alessandra
Bianchi Mariangela
Bianchi Paola
Biasin Francesca
Bignamini Silvia
Blandini Damiano
Boffelli Stefano
Bontempi Cristiano
Bordignon Alessandra
Bracchitta Luigi Maria
Bugada Maura
Cafariello Carmine
Caleri Veronica
Calsolaro Valeria
Calvani Donatella
Campagna Francesco Antonio
Capasso Andrea
Capurso Sebastiano
Carino Silvia
Carpagnano Elisiana
Carrieri Barbara
Castaldo Viviana
Castelli Manuela
Castellino Manuela
Cavarape Alessandro
Cazzulani Ilaria
Celesti Carilia
Ceolin Chiara
Ceresini Maria Giorgia
Ceretti Arcangelo
Cherubini Antonio
Chizzoli Anita
Ciarrocchi Erika
Cicciomessere Paola
http://orcid.org/0000-0003-1687-4493
Coin Alessandra
Colombo Mauro
Corsi Annalisa
Crispino Antonella
Cucunato Roberta
Custodero Carlo
D’Agostino Federica
D’Errico Maria Maddalena
D’Amico Ferdinando
De Iorio Aurelio
De Marchi Alessandro
Dell’Armi Annalaura
Delmonte Marta
Desideri Giovambattista
Devita Maria
Di Matteo Evelyn
Espinosa Emma
Esposito Luigi
Fazio Chiara
Ferro Christian
Filippini Chiara
Fini Filippo
Fiore Lucia
Fiorillo Serafina
Fontana Caterina
Forte Lina
Montorzi Riccardo Franci
Fumagalli Carlo
Fumagalli Stefano
Gareri Pietro
Gasbarri Pier Paolo
Giordano Antonella
Giuliani Evelina
Granata Roberta
Greco Antonio
Grillo Nadia
Guaita Antonio
Gucciardino Liana
Herbst Andrea
Iarrera Marilena
Ielo Giuseppe
Ippolito Valerio Alex
La Marca Antonella
La Porta Umberto
Lazzari Ilaria
Lelli Diana
Longobucco Yari
Lubian Francesca
Lucarelli Giulia
Lucchini Flaminia
Lucente Daniela
Maestri Lorenzo
Maggio Marcello
Mainquà Paola
Maiotti Mariangela
Malara Alba
Mancini Carlotta
Mancuso Irene
Marelli Eleonora
Marengoni Alessandra
Marescalco Eleonora
Martin Benedetta
Massa Valentina
Matteucci Giulia
Mattioli Irene
Mazza Liliana
Mazzoccoli Carmela
Monacelli Fiammetta
Moneti Paolo
Monzani Fabio
Morellini Federica
Mormile Maria Teresa
Mossello Enrico
Mussi Chiara
Nigro Francesca Maria
Noale Marianna
Okoye Chukwuma
Orio Giuseppe
Osso Sara
Padovan Chiara
Paglia Annalisa
Pelagalli Giulia
Pelizzoni Laura
Perri Agostino
Perticone Maria
Piccardo Giacomo
Picci Alessandro
Pippi Margherita
Provenzano Giuseppe
Pruzzo Matteo
Addamo Francesco Raffaele
Raffaelli Cecilia
Remelli Francesca
Resta Onofrio
Riccardi Antonella
Rinaldi Daniela
Rozzini Renzo
Rubino Matteo
Sabbà Carlo
Sacco Leonardo
Santoliquido Mariateresa
Savino Mariella
Scarso Francesco
Sergi Giuseppe
Serviddio Gaetano
Sgarito Claudia
Sgrò Giovanni
Sidoli Chiara
Sirianni Federica
Solfrizzi Vincenzo
Soli Benedetta
Spaccaferro Debora
Spadea Fausto
Spadoni Laura
Tafaro Laura
Tagliafico Luca
Tedde Andrea
Terziotti Camilla
Testa Giuseppe Dario
Tinti Maria Giulia
Tonarelli Francesco
Tonon Elisabetta
Trevisan Caterina
Ursino Rita
Vella Filomena
Villanova Maria
Vitali Aurora
Volpato Stefano
Zoccarato Francesca
Zotti Sonia
Zurlo Amedeo
1 grid.5608.b 0000 0004 1757 3470 Geriatrics Unit, Azienda Ospedale Università di Padova, Department of Medicine (DIMED), University of Padova, Via N. Giustiniani 2, 35128 Padua, Italy
2 grid.5326.2 0000 0001 1940 4177 Neuroscience Institute, National Research Council (CNR), Aging Branch, Padua, Italy
3 Center for Cognitive Disorders and Dementia-Catanzaro Lido, ASP Catanzaro, Catanzaro, Italy
4 grid.8484.0 0000 0004 1757 2064 Department of Medical Science, University of Ferrara, Ferrara, Italy
5 Azienda Sanitaria Locale (ASL) Alzheimer’s Disease Day Clinics, Frosinone, Italy
6 Bluecompanion Ltd, London, UK
7 ANASTE Humanitas Foundation, Rome, Italy
8 grid.8404.8 0000 0004 1757 2304 Geriatric Intensive Care Unit, Department of Experimental and Clinical Medicine, University of Firenze, Florence, Italy
9 Intermediate Care Unit, Nursing Home Misericordia, Navacchio, Pisa, Italy
10 grid.7563.7 0000 0001 2174 1754 Acute Geriatric Unit, IRCCS Foundation San Gerardo Hospital, School of Medicine and Surgery, University of Milano-Bicocca, Monza, Italy
11 grid.9657.d 0000 0004 1757 5329 Unit of Geriatrics, Department of Medicine, Campus Bio-Medico University and Teaching Hospital, Rome, Italy
19 5 2023
2023
14 3 465476
19 1 2023
19 4 2023
© The Author(s), under exclusive licence to European Geriatric Medicine Society 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Key summary points
Aim
To comparatively assess the clinical profiles of older patients treated with trazodone or other antidepressants in a large dataset from the GeroCovid Observational multiscope and multisetting study.
Findings
10.8% out of 3396 persons included used trazodone and the 8.5% other antidepressants; the use of trazodone was highly prevalent in functionally dependent and comorbid older adults admitted to long-term care facilities or living at home. Conditions associated with trazodone use included depression, dementia and behavioral and psychological symptoms of dementia.
Message
The present data suggest an off-label use of trazodone as a possible therapeutic option in the challenging field of behavioral and psychological disturbances in older adults with dementia.
Background and objectives
Depression is highly prevalent in older adults, especially in those with dementia. Trazodone, an antidepressant, has shown to be effective in older patients with moderate anxiolytic and hypnotic activity; and a common off-label use is rising for managing behavioral and psychological symptoms of dementia (BPSD). The aim of the study is to comparatively assess the clinical profiles of older patients treated with trazodone or other antidepressants.
Methods
This cross-sectional study involved adults aged ≥ 60 years at risk of or affected with COVID-19 enrolled in the GeroCovid Observational study from acute wards, geriatric and dementia-specific outpatient clinics, as well as long-term care facilities (LTCF). Participants were grouped according to the use of trazodone, other antidepressants, or no antidepressant use.
Results
Of the 3396 study participants (mean age 80.6 ± 9.1 years; 57.1% females), 10.8% used trazodone and 8.5% others antidepressants. Individuals treated with trazodone were older, more functionally dependent, and had a higher prevalence of dementia and BPSD than those using other antidepressants or no antidepressant use. Logistic regression analyses found that the presence of BPSD was associated with trazodone use (odds ratio (OR) 28.4, 95% confidence interval (CI) 18-44.7 for the outcome trazodone vs no antidepressants use, among participants without depression; OR 2.17, 95% CI 1.05-4.49 for the outcome trazodone vs no antidepressants use, among participants with depression). A cluster analysis of trazodone use identified three clusters: cluster 1 included mainly women, living at home with assistance, multimorbidity, dementia, BPSD, and depression; cluster 2 included mainly institutionalized women, with disabilities, depression, and dementia; cluster 3 included mostly men, often living at home unassisted, with better mobility performance, fewer chronic diseases, dementia, BPSD, and depression.
Discussion
The use of trazodone was highly prevalent in functionally dependent and comorbid older adults admitted to LTCF or living at home. Clinical conditions associated with its prescription included depression as well as BPSD.
Keywords
Depression
Dementia
Trazodone
BPSD, Profile, Clinical characteristics
issue-copyright-statement© European Geriatric Medicine Society 2023
==== Body
pmcIntroduction
Depression is a highly prevalent, yet commonly underdiagnosed condition in older patients across all settings of care [1–3]. A wide range of depressive traits, from mild to major depression symptoms, can be found in advanced age associated with numerous psychiatric and non-psychiatric conditions [4]. Clinical expression of depression in old age is highly variable. Approximately, one-third of older patients with dementia show depressive symptoms associated with other neuropsychiatric symptoms, while those with depression (without dementia) more frequently have atypical behavioral disorders [2, 5, 6], such as sleep troubles, anorexia, agitation, and confusion [7], compared to younger adults. Moreover, anxiety often coexists in comorbid older persons with depression and prevalence estimates ranging from 23 to 48% [8]. Due to this clinical heterogeneity, the choice of an antidepressant treatment is often guided by the expected activity of the drug on coexistent neuropsychiatric symptoms. For this reason, real-life prescription studies of different antidepressant drugs according to clinical profiles may shed light on the available literature [9].
Among various antidepressants, trazodone qualifies as an eclectic drug showing an effective antidepressant action as well as moderate anxiolytic and hypnotic activities in older patients [10, 11]. Due to this sedative action, trazodone is often used in older people with dementia or delirium to manage agitation, insomnia, and other behavioral and psychological symptoms of dementia (BPSD) [12, 13]. However, available literature on the prescribing patterns of trazodone in different care settings is lacking. Iaboni et al. [14] described an increased use of trazodone and quetiapine over years, paralleling decreased benzodiazepine prescription in older adults with dementia in Canada, in both long-term care facilities (LTCF) and in the community.
In this context, the GeroCovid Observational initiative offered a unique opportunity of exploring the use of trazodone in older individuals from an epidemiological point of view. This observational study was launched by the Italian Society of Gerontology and Geriatrics (SIGG) to assess the direct and indirect impact of the COVID-19 pandemic on the health of older people in different care settings, including acute wards, dementia and geriatric outpatient clinics, home services, and LTCF. In addition to COVID-19-related information, the initiative also included data collection related to chronic diseases and treatments, which allowed for secondary analyses exploring the current management of diverse common conditions.
In this work, we aimed to characterize the prescription patterns of trazodone in comparison with other antidepressants in older people across different settings of care. In particular, we aimed at investigating the versatile use of trazodone in older persons with dementia and BPSD across diverse clinical settings.
Methods
Study design
This is a cross-sectional study using data from GeroCovid Observational, an initiative involving adults aged ≥ 60 years evaluated during the COVID-19 pandemic. GeroCovid Observational is a multicenter, multiscope, and multisetting study designed by the Italian Society of Gerontology and Geriatrics and is structured into the following research settings (details can be found in previous publications [15, 16]: GeroCovid acute care wards (including patients hospitalized for SARS-CoV-2 infection), GeroCovid home and outpatient care (including individuals accessing geriatric outpatient or home care services), GeroCovid dementia–drug monitoring and GeroCovid dementia–psychological health (including outpatients with cognitive impairment), GeroCovid long-term care facilities (LTCF) (including residents with suspected or confirmed SARS-CoV-2 infection), and GeroCovid outcomes (including individuals followed up after a COVID-19-related hospitalization). The study was conducted following the STROBE guidelines. Data registration was performed using a dedicated electronic register designed by Bluecompanion (UK, France) to collect all clinical data from every investigational site across Italy.
The GeroCovid Observational study protocol was approved by the Campus Bio-Medico University Ethical Committee in April 2020. All participating investigational sites further submitted relevant sub-protocols to their competent local ethical committee and institutional review boards, as applicable according to Italian regulations. All investigators accepted to work according to the Good Clinical Practice (GCP) (ICH E6-R2). Written or dematerialized informed consent was obtained from each participant. Alternatively, a written declaration was kept on file by the local investigator, which responded to applicable derogations during the pandemic.
Data collection
Data collection for this study included: demographic characteristics (age, sex), lifestyle (smoking habits, alcohol consumption), mobility (independent walking, walking with a cane or walker, bed-rest condition), social determinants (care setting: nursing home, in-patient, outpatients, home-based; household: lives at home alone and is autonomous, lives at home regularly assisted, lives at home alone with informal caregiver, lives at home with family and is autonomous, lives in a nursing home; social distancing impact: major impact, no or moderate impact), chronic diseases from medical records coded according to Medical Dictionary for Regulatory Activities (MedDRA) [17] (hypertension, cardiovascular diseases, diabetes mellitus, osteoarthritis, chronic obstructive pulmonary disease (COPD), obesity, depression, BPSD, mild cognitive impairment (MCI), dementia, and type of dementia), and functional status with activity of daily livings (ADL) by Katz index and instrumental activity daily livings (IADL) by Lawton and Brody index. Based on the list of drugs chronically used by the study participants, coded according to the ATC, we identified participants treated with trazodone (TRAZ), other antidepressants (AnDep), or no antidepressant treatment (No AnDep). All participants with a sars-cov-2 infection (diagnosed with a real-time PCR) were included in the inpatient care setting, while in the other care settings only 26 participants were tested and resulted negative.
Statistical analysis
The characteristics of the study participants are presented as means ± standard deviation (SD), or median and interquartile range [25th–75th percentile] for quantitative measures, or counts and percentages for categorical variables. Normality of the distributions was evaluated considering the Kolmogorov–Smirnov test. No imputation of missing values was performed.
The study participants were grouped according to the use of antidepressants (TRAZ, AnDep, No AnDep) and compared according to sociodemographic characteristics (age, gender), lifestyle (smoking habits, alcohol habits, physical activity), mobility, social determinants (care setting, household, and social distancing impact), chronic conditions, functional status, and SARS-CoV-2 infection, using the Chi-squared or Fisher exact tests for categorical variables and generalized linear model testing for homoscedasticity through the Levene’s test for quantitative variables. Post hoc analyses with Bonferroni adjustment for multiple comparisons were applied.
A multinominal logistic regression model with outcome groups defined according to antidepressant drugs, TRAZ, AnDep, or No AnDep use, was defined with the following independent variables: age, gender, mobility, household, chronic conditions (hypertension, cardiovascular diseases, stroke, diabetes mellitus, osteoarthrosis, chronic obstructive pulmonary disease, chronic renal failure, obesity, depression, mild cognitive impairment, dementia, and behavioral disorders)a. A block entering of independent variables was applied and results are presented as adjusted odds ratios (OR) and 95% confidence intervals (95% CI). Multiplicative interactions between the variables were also tested in the logistic models.
A cluster analysis based on k-means method was used to classify participants using TRAZ into clusters of coexisting characteristics. K-means algorithm divides the n sample into k clusters so that internal similarity of the clusters is high and the external similarity of the clusters is low. The optimal number of clusters was determined by comparing pseudo-F statistic and cubic clustering criterion for models with different number of clusters. The characteristics of participants in different clusters were compared using Chi-squared, Fisher exact tests or generalized linear model, as appropriate. All statistical tests were two tailed and statistical significance was assumed for p value < 0.05. The analyses were performed using SAS, V.9.4 (SAS Institute, Cary, NC).
Results
Out of the 3396 cases included in the analysis, 367 (10.8%) were using TRAZ and 288 (8.5%) AnDep. As shown in Table 1, the mean age was significantly higher in the TRAZ (84.3 ± 6.6) and AnDep (82.5 ± 7.2) groups than the No AnDep group (80.0 ± 9.4) (F = 44.5, df = 2, p < 0.0001). Over 60% of women were using antidepressants. Mobility deficits were approximately 50% in the entire study group (n = 3170). Patients using TRAZ had significantly lower levels of independent mobility than those taking AnDep or No AnDep treatments (32%, 36%, and 55% can walk independently, respectively) (χ2 = 101.5, df = 6, p < 0.0001).Table 1 Demographic, lifestyle, and social characteristics of the GeroCovid population according to TRAZ, AnDep, and No AnDep groups
All (n = 3396) TRAZ (n = 367) AnDep (n = 288) No AnDep (n = 2741) p value
Demographic characteristics
Age (years) 80.6 ± 9.1 84.3 ± 6.6 82.5 ± 7.2 80.0 ± 9.4 < 0.0001abc
Gender (Female) 1938 (57.1) 251 (68.4) 191 (66.3) 1496 (54.6) < 0.0001ab
Lifestyle
Smoking habits (available for n = 1830) 0.0077a
Current smoker 86 (4.7) 10 (4.9) 8 (4.7) 68 (4.7)
Former smoker 371 (20.3) 26 (12.8) 24 (14.2) 321 (22.0)
Never smoker 1373 (75.0) 168 (82.4) 137 (81.1) 1068 (73.3)
Alcohol consumption (available for n = 1707), yes 158 (9.3) 9 (3.8) 6 (3.6) 143 (11.0) < 0.0001ab
Mobility (available for n = 3170) < 0.0001ab
Walks independently 1615 (51.0) 112 (32.4) 99 (36.4) 1404 (55.0)
Walks with help (cane or walker) 613 (19.3) 95 (27.4) 79 (29.0) 439 (17.2)
Wheelchair (autonomous or pushed) 334 (10.5) 57 (16.5) 43 (15.8) 234 (9.2)
Bed-rest condition 608 (19.2) 82 (23.7) 51 (18.8) 475 (18.6)
Social determinants
Care setting (available for n = 2974) < 0.0001ab
Outpatients 918 (30.9) 114 (33.8) 60 (22.6) 744 (31.4)
Nursing home 559 (18.8) 72 (21.4) 63 (23.7) 424 (17.9)
Inpatient 1113 (37.4) 89 (26.4) 91 (34.2) 933 (39.3)
Home based 384 (12.9) 62 (18.4) 52 (19.6) 270 (11.4)
Household (available for n = 2763) < 0.0001abc
Lives at home alone, autonomous 160 (5.8) 11 (3.2) 11 (4.3) 138 (6.3)
Lives at home alone, regularly assisted 195 (7.1) 25 (7.4) 29 (11.3) 141 (6.5)
Lives at home alone, informal caregiver 528 (19.1) 125 (37.0) 67 (26.3) 336 (15.5)
Lives at home with family, autonomous 1057 (38.2) 52 (15.4) 66 (25.9) 939 (43.3)
Lives in a nursing home 823 (29.8) 125 (37.0) 82 (32.2) 616 (28.4)
Social distancing impact (available for n = 1477) 0.0030a
Major impact 543 (36.8) 64 (27.7) 51 (33.3) 428 (39.2)
No or moderate impact 934 (63.2) 167 (72.3) 102 (66.7) 665 (60.8)
Numbers are mean ± SD, or count (%), as appropriate
TRAZ (participants on trazodone); AnDep (participants taking other antidepressants; No AnDep (participants not taking antidepressants); SD (standard deviation)
Overall p value is calculated among the three groups of antidepressant users/not users
aSignificant post hoc difference for TRAZ vs NoAnDep
bSignificant post hoc difference for AnDep vs NoAnDep
cSignificant post hoc difference for TRAZ vs AnDep
In the entire study population, we found that the prescription of TRAZ was higher in outpatient settings (mainly outpatients followed for cognitive deficits), while any AnDep use was higher in inpatients (χ2 = 47.9, df = 6, p < 0.0001). In the TRAZ group, there was a significantly greater need for patient assistance either at home from informal caregivers or in nursing homes (χ2 = 170.4, df = 8, p < 0.0001).
The prevalence of chronic diseases, the use of antipsychotics, and TRAZ dosages are reported in Table 2. A recorded diagnosis of depression was present in 54% of those using AnDep, in 28% of those using TRAZ, and in 12% of those not treated (No AnDep). (χ2 = 344.7, df = 2, p < 0.0001). More than half of the participants suffering from depression (55%) were not undergoing any pharmacological treatment for depression. In the TRAZ and AnDep groups, approximately 70% had three or more chronic diseases vs. 40% in the No AnDep group (χ2 = 177.1, df = 2, p < 0.0001). In particular, hypertension, cardiovascular diseases, and osteoarthritis were more prevalent in the TRAZ and AnDep groups compared to the No AnDep group (for hypertension, 72.2% and 72.9% vs 58%, respectively, p < 0.0001; for cardiovascular diseases, 48.8% and 47.6% vs 41.2%, p = 0.0048; for osteoarthrosis, 36.5% and 35.5% vs 22.3%, p < 0.0001). Dementia reached 28% prevalence with a higher prevalence in the TRAZ group (67%) than the AnDep (49%) and the No AnDep group (21.6%) (χ2 = 384.3, df = 2, p < 0.0001). Similarly, BPSD and antipsychotic treatments were more prevalent in the TRAZ group: BPSD were present in 34.6% of the individuals in the TRAZ group, 13.9% in the AnDep group, and 3.8% in the No AnDep group (p < 0.0001); antipsychotic treatments were used in 44.7% of individuals in the TRAZ group, 30.9% in the AnDep group, and 10.7% in the No AnDep group (p < 0.0001). The daily dosage of TRAZ was lower than 100 mg in most of the participants (Table 2).Table 2 Health-related, functional, and frailty characteristics of the GeroCovid population according to the TRAZ, AnDep, and No AnDep groups
All (n = 3396) TRAZ (n = 367) AnDep (n = 288) No AnDep (n = 2741) p value
Chronic conditions
Hypertension (available for n = 3236) 1972 (60.9) 262 (72.2) 210 (72.9) 1500 (58.0) < 0.0001ab
Cardiovascular diseases* (available for n = 3219) 1371 (42.6) 177 (48.8) 136 (47.6) 1058 (41.2) 0.0048a
Stroke (available for n = 3212) 305 (9.5) 44 (12.3) 35 (12.2) 226 (8.8) 0.0285
Diabetes mellitus (available for n = 3217) 684 (21.3) 76 (21.1) 61 (21.3) 547 (21.3) 0.9970
Osteoarthrosis (available for n = 3213) 806 (25.1) 131 (36.5) 102 (35.5) 573 (22.3) < 0.0001ab
Chronic obstructive pulmonary disease (available for n = 3211) 426 (13.3) 52 (14.5) 36 (12.5) 338 (13.2) 0.7367
Chronic renal failure (available for n = 3213) 384 (12.0) 49 (13.7) 40 (13.9) 295 (11.5) 0.2691
Obesity (available for n = 3224) 254 (7.9) 25 (7.0) 24 (8.4) 205 (8.0) 0.7689
Depression (available for n = 3213) 564 (17.6) 101 (27.9) 155 (54.0) 308 (12.0) < 0.0001abc
Mild cognitive impairment 140 (4.1) 15 (4.1) 10 (3.5) 115 (4.2) 0.8410
Dementia 978 (28.8) 245 (66.8) 141 (49.0) 592 (21.6) < 0.0001abc
If dementia, specify
Alzheimer’s disease 402 (11.8) 119 (32.4) 62 (21.5) 221 (8.1) < 0.0001abc
Vascular dementia 151 (4.5) 33 (9.0) 14 (4.9) 104 (3.8) < 0.0001a
Dementia with Lewy bodies 18 (0.5) 6 (1.6) 3 (1.0) 9 (0.3) 0.0039a
Frontotemporal dementia 10 (0.3) 3 (0.8) 2 (0.7) 5 (0.2) 0.0311
Mixed dementia 105 (3.1) 24 (6.5) 19 (6.6) 62 (2.3) < 0.0001ab
Other 253 (7.5) 46 (12.5) 40 (13.9) 167 (6.1) < 0.0001ab
Dementia due to Parkinson’s disease 11 (0.3) 2 (0.5) 0 (0.0) 9 (0.3) 0.5098
Dementia due to other medical condition 10 (0.3) 4 (1.1) 0 (0.0) 6 (0.2) 0.0389
Dementia with BPSD 270 (8.0) 127 (34.6) 40 (13.9) 103 (3.8) < 0.0001abc
Comorbidities**, 3+ 1546 (45.5) 245 (66.8) 205 (71.2) 1096 (40.0) < 0.0001ab
Functional status
Activities of daily living, median (Q1, Q3) (available for n = 1668) 4 (1, 6) 2 (1, 4) 3 (1, 5) 4 (1, 6) < 0.0001abc
Instrumental activities of daily living, median (Q1, Q3) (available for n = 1444) 1 (0, 5) 0 (0, 1) 1 (0, 4) 2 (0, 6) < 0.0001abc
Antipsychotic drugs use, n (%) 545 (16.1) 164 (44.7) 89 (30.9) 292 (10.7) < 0.0001abc
Trazodone dose, n (%) – –
> 100 mg/day 19 (7.8)
≤ 100 mg/day 226 (92.2)
Numbers are mean ± SD, median (Q1, Q3) or count (%), as appropriate
TRAZ, participants on trazodone; AnDep, participants assuming other antidepressants; No AnDep, participants not taking antidepressant; SD, standard deviation; BPSD, behavioral and psychological symptoms of dementia
Overall p value is calculated among the three groups of antidepressant users/not users
*Includes atrial fibrillation, peripheral arteriopathy, heart failure, and cardiomyopathy
**Total number of comorbidities, considering hypertension, cardiovascular diseases, stroke, diabetes mellitus, osteoarthrosis, chronic obstructive pulmonary disease, chronic renal failure, obesity, depression, behavior disorders, mild cognitive impairment, and dementia
aSignificant post hoc difference for NoAnDep vs TRAZ
bSignificant post hoc difference for NoAnDep vs AnDep
cSignificant post hoc difference for TRAZ vs AnDep
Clinical variables associated with the use of TRAZ or AnDep in multinomial models are reported in Table 3. Motor disability, depression, and dementia (with or without the presence of BPSD) were associated with an increased likelihood of using TRAZ or AnDep compared to not being treated. The interaction between depression and dementia (with or without BPSD) was statistically significant, meaning that the relationship between dementia and the outcome (use of TRAZ or AnDep or NoAnDep) depends on the presence of depression (χ2 = 23.7, df = 4, p < 0.0001). In fact, the association of dementia (and especially of dementia with BPSD) with the outcome was stronger in participants without depression compared to those with depression. In particular, in participants without depression, the association between TRAZ and dementia with BPSD (OR 28.4, 95% CI 18.8, 44.7; Wald χ2 = 209.1, df = 1, p < 0.0001) was significant, as well as the association between AnDep and dementia with BPSD (OR 5.68, 95% CI 2.94, 10.9; Wald χ2 = 26.7, df = 1, p < 0.0001). Among individuals with depression, an independent association between TRAZ and dementia with or without BPSD was still present, while no association was observed between AnDep and dementia. Stratifying analyses by care setting (Supplementary Table S1), results were substantially confirmed in home-based, nursing home, and outpatient participants, while a significant association between TRAZ and dementia without BPSD and between AnDep and dementia with BPSD among participants without depression was found for inpatients.Table 3 Multinomial logistic regression model with outcome “Trazodone”, “Other antidepressants” or “No antidepressants” use in the GeroCovid population
TRAZ vs no AnDep AnDep vs no AnDep TRAZ vs AnDep
OR 95% CI p value OR 95% CI p value OR 95% CI p value
Age, years 1.01 0.99–1.03 0.4027 1.00 0.98–1.02 0.8798 1.01 0.98–1.04 0.4446
Gender, female vs male 1.15 0.85–1.55 0.3694 1.05 0.76–1.45 0.7689 1.09 0.73–1.63 0.6602
Mobility vs walks independently
Walks with help 1.28 0.87–1.89 0.2047 1.45 0.96–2.17 0.0749 0.89 0.54–1.46 0.6391
Wheelchair 2.00 1.23–3.25 0.0055 1.95 1.16–3.27 0.0120 1.03 0.55–1.91 0.9356
Bed-rest condition 1.34 0.85–2.13 0.2096 1.31 0.79–2.16 0.2969 1.03 0.56–1.89 0.9286
Household vs lives at home alone, autonomous
Lives at home alone, regularly assisted 1.12 0.48–2.61 0.7924 1.36 0.60–3.08 0.4678 0.83 0.28–2.44 0.7304
Lives at home alone, informal caregiver 1.67 0.78–3.55 0.1874 1.14 0.53–2.45 0.7447 1.47 0.54–3.97 0.4520
Lives at home with family, autonomous 0.90 0.43–1.91 0.7901 0.96 0.46–2.00 0.9043 0.95 0.35–2.53 0.9101
Lives in a nursing home 1.67 0.77–3.63 0.1954 1.00 0.46–2.18 0.9901 1.68 0.61–4.66 0.3198
Hypertension 1.20 0.89–1.62 0.2248 1.39 1.00–1.93 0.0474 0.87 0.58–1.29 0.4773
Cardiovascular diseases 1.16 0.88–1.53 0.3009 1.16 0.86–1.57 0.3221 1.00 0.69–1.43 0.9814
Stroke 1.19 0.78–1.80 0.4213 1.16 0.74–1.82 0.5138 1.02 0.60–1.75 0.9343
Diabetes mellitus 0.85 0.61–1.18 0.3332 0.90 0.64–1.28 0.5707 0.94 0.61–1.45 0.7733
Osteoarthrosis 1.14 0.85–1.53 0.3832 1.07 0.78–1.47 0.6837 1.07 0.73–1.56 0.7423
Chronic obstructive pulmonary disease 1.12 0.76–1.65 0.5792 0.89 0.58–1.36 0.5877 1.26 0.75–2.11 0.3880
Chronic renal failure 0.78 0.52–1.16 0.2183 0.99 0.65–1.50 0.9625 0.79 0.47–1.31 0.3594
Obesity 0.86 0.51–1.47 0.5862 1.12 0.67–1.86 0.6681 0.77 0.40–1.49 0.4401
Mild cognitive impairment 1.67 0.87–3.20 0.1207 0.96 0.45–2.03 0.9094 1.75 0.70–4.36 0.2324
Study participants with depression
Dementia with no BPSD vs no dementia 2.29 1.26–4.16 0.0066 1.53 0.92–2.56 0.1037 1.50 0.78–2.89 0.2305
Dementia with BPSD vs no dementia 4.97 2.52–9.80 < 0.0001 1.96 0.99–3.78 0.0560 2.54 1.22–5.29 0.0128
Dementia with BPSD vs dementia with no BPSD 2.17 1.05–4.49 0.0365 1.28 0.63–2.60 0.4988 1.70 0.78–3.68 0.1797
Study participants without depression
Dementia with no BPSD vs no dementia 3.87 2.66–5.62 < 0.0001 3.25 2.12–4.98 < 0.0001 1.19 0.70–2.03 0.5254
Dementia with BPSD vs no dementia 28.4 18.0–44.7 < 0.0001 5.68 2.94–10.9 < 0.0001 5.00 2.48–10.1 < 0.0001
Dementia with BPSD vs dementia with no BPSD 7.34 4.72–11.4 < 0.0001 1.75 0.90–3.38 0.0974 4.20 2.12–8.32 < 0.0001
Block entering of independent variables
OR, odds ratio; 95% CI, 95% confidence interval; TRAZ, participants on trazodone; AnDep, participants taking other antidepressants; No AnDep, participants not taking antidepressants; BPSD, behavioral and psychological symptoms of dementia
The cluster analysis according to TRAZ use characterized three clusters (Supplementary Table 1; Figs. 1, 2). Cluster 1 included mainly older women living at home requiring assistance due to functional limitations (more than 60% moved with help or in a wheelchair), multimorbidity (73%), dementia (82%), and BPSD (62% among those with dementia); one out of four had a diagnosis of depression. Cluster 2 was mainly composed of institutionalized women with disabilities and chronic diseases (in particular, hypertension, diabetes mellitus, COPD, and obesity). Approximately 30% had depression, while 55% had dementia, but only 5% among those with dementia had BPSD. Cluster 3 included younger participants, mostly men, often living at home unassisted, with better mobility performance and fewer chronic diseases; however, 58% had dementia and 36% BPSD, while only 21% had depression.Fig. 1 Depression, dementia, and BPSD prevalence by clusters identified among TRAZ users
Fig. 2 Summary of participants’ characteristics among clusters
In all clusters, the dose of TRAZ was mainly < 100 mg/day and over 40% were using antipsychotics with a slightly higher frequency in cluster 1.
Discussion
In this sample of older participants with, or at risk of, COVID-19, approximately one out of five persons (18.5%) over 60 years of age received an antidepressant, with trazodone being the most prescribed drug (10.8% vs. 8.5% other antidepressants). We also found that the use of trazodone was associated with clinical conditions other than depression. The frequency of antidepressant use in our sample was similar to the available literature [18], although lower in comparison with data from LTCF studies [19, 20]. Moreover, our study showed a higher use of trazodone than those reported in other studies [13, 21]. This may be explained by the fact that our study included multiple care settings with diverse clinical characteristics. The GeroCovid study includes a large data collection from diverse care settings, which extends clinical knowledge from previous population-based cohort or LTCF studies [13, 21].
We found that participants using trazodone were more likely to be functionally dependent and to have several comorbidities and a higher prevalence of cardiovascular diseases or osteoarthritis. Furthermore, 28% of the participants in the TRAZ group had a formal diagnosis of depression (vs. 54% among participants treated with other antidepressants and 12% among participants not receiving antidepressants), while two-thirds of TRAZ group participants suffered from dementia (vs. 49% of those treated with other antidepressants and 22% of those not receiving antidepressant treatment), and one-third of TRAZ group had BPSD (vs. 14% in the AnDep and 4% in the No AnDep groups). Analyses underlined that independent of a depression diagnosis, BPSD was associated with an increased use of trazodone compared to other antidepressants. The association was the strongest among participants without depression, but still statistically significant in those with depression. Indeed, this antidepressant has been shown to be used off-label for insomnia [22, 23] and hold anxiolytic [24, 25] effects, as well as control aggression [26] and other BPSD [27, 28].
The cluster analysis provides some further insights regarding the prescription of trazodone. In particular, clusters 1 and 3 included older community-dwelling adults with a high prevalence of BPSD. However, cluster 3 included mainly those with better general clinical conditions, even if affected by dementia and BPSD and often not needing direct assistance at home. Furthermore, the dosage of trazodone in clusters 1 and cluster 3 was lower than 75 mg/day, a dosage which has been reported for treating BPSD, such as wandering, agitation, delusions [29], and treating psychotic symptoms in depression and dementia [30]. Interestingly, antipsychotics were used in 40–50% of individuals using trazodone. Trazodone has also been reported to be effective in neuroleptic-induced akathisia [31].
Cluster 2 identified institutionalized individuals and inpatients with disabilities and chronic pathologies, among which about one-third had depression and less than half dementia, but very few had BPSD.
Regardless of diagnosed dementia, antidepressants other than trazodone were prescribed in participants with depression. On the other hand, in those with dementia, there was a vast use of antidepressants regardless of depression or BPSD [32]. Dementia is often accompanied by affective disorders, including anxiety and emotional lability, which are widely treated with antidepressants even if clinical evidence regarding their efficacy remains conflicting [33]. Moreover, antidepressants such as sertraline and citalopram have also shown to reduce agitation and psychosis in dementia, which may partially explain their increased use among non-depressed participants with dementia [33]. Nevertheless, cardiovascular side effects might limit their off-label use in frail older individuals without depression. Psychiatric reactions to SSRI use, including anxiety, irritability and, more rarely, mania and psychosis, may also limit their use in older patients [34].
Another interesting finding from our study was that more than half of the participants suffering from depression were not treated with any antidepressant agent, which confirms other reports [35, 36], suggesting that the stigma of using psychiatric drugs in old people has not yet been overcome [37–39].
The strength of the present study relies on the large sample size and on the extensive clinical assessment in a real-life older population from different care settings.
The main limitations are represented by the cross-sectional analysis, the lack of information regarding the purpose of prescription, duration, effectiveness, as well as the tolerability of the prescribed drugs. Diagnoses were collected by medical records and we did not collect data regarding pain.
In conclusion, this study underlined a high prevalence of trazodone prescription in a large, multiset sample of older Italian participants with, or at risk of, COVID-19. The present data suggest an off-label use of this drug at doses ≤ 100 mg/day for the treatment of dementia with BPSD with or without depression. Further studies assessing reason of prescription, effectiveness and tolerability of different antidepressants in frail older adults, with and without dementia are necessary to design. Then, dedicated clinical trials trazodone on behavioral and psychological disturbances in the rapidly growing number of older adults with dementia.
Acknowledgements
We thank Gilda Borselli for her precious support for the organization of the GeroCovid initiative. Co-authors, members of the GeroCovid OBSERVATIONAL Working Group (in alphabetical order): Angela Marie Abbatecola [ASL Frosinone; RSA INI Città Bianca, Veroli (FR)], Domenico Andrieri [RSA Villa Santo Stefano, S. Stefano di Rogliano (CS)], Sara Antenucci [Ambulatorio Psicogeriatrico, Ortona (CH)], Rachele Antognoli (Azienda Ospedaliero Universitaria Pisana; RSA Villa Isabella, Pisa), Raffaele Antonelli Incalzi (Università Campus Bio-Medico, Roma), Maria Paola Antonietti (Ospedale Regionale di Aosta), Viviana Bagalà (Azienda Ospedaliero-Universitaria di Ferrara), Giulia Bandini (USL Toscana Centro, Ospedale San Jacopo, Pistoia), Salvatore Bazzano [ULSS 3 Serenissima, Presidio di Dolo (VE)], Giuseppe Bellelli (Ospedale San Gerardo, Monza), Andrea Bellio (Azienda Ospedale Università di Padova), Federico Bellotti (Azienda Ospedaliero-Universitaria di Ferrara), Enrico Benvenuti [USL Toscana Centro, Ospedale Santa Maria Annunziata, Bagno a Ripoli (FI)], Marina Bergamin (Azienda Ospedaliero-Universitaria di Parma), Marco Bertolotti (Azienda Ospedaliero-Universitaria di Modena), Carlo Adriano Biagini (USL Toscana Centro, Ospedale San Jacopo, Pistoia), Angelo Bianchetti (Istituto Clinico Sant’Anna, Brescia), Alessandra Bianchi [Spedali Civili, Montichiari (BS)], Mariangela Bianchi (Policlinico Sant’Orsola-Malpighi, Bologna), Paola Bianchi (Associazione Nazionale Strutture Territoriali e per la Terza Età, Roma), Francesca Biasin (Azienda Ospedale Università di Padova), Silvia Bignamini (Casa di Cura San Francesco, Bergamo), Damiano Blandini (Policlinico Sant’Orsola-Malpighi, Bologna), Stefano Boffelli (Fondazione Poliambulanza, Brescia), Cristiano Bontempi (Azienda Ospedale Università di Padova), Alessandra Bordignon (Azienda Ospedale Università di Padova), Luigi Maria Bracchitta (ATS Milano), Maura Bugada (Casa di Cura San Francesco, Bergamo), Carmine Cafariello [RSA Villa Sacra Famiglia, IHG, Roma; I RSA Geriatria, IHG, Guidonia (RM); III RSA Geriatria, IHG, Guidonia (RM); RSA Estensiva, IHG, Guidonia (RM); RSA Intensiva, IHG, Guidonia (RM)], Veronica Caleri (USL Toscana Centro, Ospedale San Jacopo, Pistoia), Valeria Calsolaro (Azienda Ospedaliero Universitaria Pisana; RSA Villa Isabella, Pisa), Donatella Calvani (USL Toscana Centro, Presidio Misericordia e Dolce, Prato; USL Toscana Centro, Ospedale Santo Stefano, Prato), Francesco Antonio Campagna [Centro di Riabilitazione San Domenico, Lamezia Terme (CZ)], Andrea Capasso (ASL Napoli 2 Nord), Sebastiano Capurso [RSA Bellosguardo, Civitavecchia (RM)], Silvia Carino [RSA San Domenico, Lamezia Terme (CZ); Centro di Riabilitazione San Domenico, Lamezia Terme (CZ); RSA Villa Elisabetta, Cortale (CZ); Casa Protetta Madonna del Rosario, Lamezia Terme (CZ)], Elisiana Carpagnano (Ospedale Giovanni XXIII Policlinico di Bari), Barbara Carrieri (IRCCS INRCA, Ancona), Viviana Castaldo (Presidio Ospedaliero Universitario Santa Maria della Misericordia, Udine), Manuela Castelli [Istituto Geriatrico Camillo Golgi, Abbiategrasso (MI)], Manuela Castellino [Fatebenefratelli, Presidio Ospedaliero Riabilitativo "Beata Vergine Consolata", San Maurizio Canavese (TO)], Alessandro Cavarape (Presidio Ospedaliero Universitario Santa Maria della Misericordia, Udine), Ilaria Cazzulani (Ospedale San Gerardo, Monza), Carilia Celesti (Policlinico Universitario Campus Bio-Medico, Roma), Chiara Ceolin (Azienda Ospedale Università di Padova), Maria Giorgia Ceresini (Azienda Ospedaliero-Universitaria di Ferrara), Arcangelo Ceretti [Istituto Geriatrico Camillo Golgi, Abbiategrasso (MI)], Antonio Cherubini (IRCCS INRCA, Ancona), Anita Chizzoli (Istituto Clinico Sant’Anna, Brescia), Erika Ciarrocchi (IRCCS INRCA, Ancona), Paola Cicciomessere (Azienda Ospedaliero Universitaria di Foggia), Alessandra Coin (Azienda Ospedale Università di Padova), Mauro Colombo [Istituto Geriatrico Camillo Golgi, Abbiategrasso (MI)], Annalisa Corsi (USL Toscana Centro, Ospedale San Jacopo, Pistoia), Antonella Crispino [RSA Villa Santo Stefano, S. Stefano di Rogliano (CS); RSA Villa Silvia, Altilia Grimaldi (CS)], Roberta Cucunato [RSA Villa Santo Stefano, S. Stefano di Rogliano (CS); RSA Villa Silvia, Altilia Grimaldi (CS)], Carlo Custodero (Ospedale Giovanni XXIII Policlinico di Bari), Federica D’Agostino [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Maria Maddalena D’Errico [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Ferdinando D’Amico [RSA San Giovanni di Dio, Patti (ME); RSA Sant’Angelo di Brolo (ME)], Aurelio De Iorio (Azienda Ospedaliero-Universitaria di Parma), Alessandro De Marchi [Policlinico Sant’Orsola-Malpighi, Bologna), Annalaura Dell’Armi (III RSA Geriatria, IHG, Guidonia (RM)], Marta Delmonte (Azienda Ospedaliero-Universitaria di Ferrara), Giovambattista Desideri [Ospedale di Avezzano (AQ)], Maria Devita (Azienda Ospedale Università di Padova), Evelyn Di Matteo (Policlinico Universitario Campus Bio-Medico, Roma), Emma Espinosa [Azienda Ospedali Riuniti Marche Nord, Fano (PU)], Luigi Esposito [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Chiara Fazio (Azienda Ospedaliero-Universitaria di Parma), Christian Ferro [RSA Sant’Angelo di Brolo (ME)], Chiara Filippini [Spedali Civili, Montichiari (BS)], Filippo Fini (Azienda Ospedale Università di Padova), Lucia Fiore [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Serafina Fiorillo [ASP Vibo Valentia; RSA Madonna delle Grazie, Filadelfia (VV); Casa di Riposo Mons. Francesco Luzzi, Acquaro (VV); Casa di Riposo Villa Betania, Mileto (VV); Casa di Riposo Pietro Rosano, Dasà (VV); Casa di Riposo Serena Diocesi, Mileto (VV); Alloggio per Anziani Villa Amedeo, Francavilla Angitola (VV); Casa Albergo Villa Fabiola, Monterosso Calabro (VV); Casa di Riposo Villa Sara, San Nicola da Crissa (VV); Casa di Riposo Don Mottola, Tropea (VV); Casa di Riposo San Francesco, Soriano Calabro (VV); RSA Anziani, Soriano Calabro (VV); Casa di Riposo Suore Missionarie del Catechismo, Pizzo (VV)], Caterina Fontana (Azienda Ospedaliero-Universitaria di Modena), Lina Forte [Ospedale di Avezzano (AQ)], Riccardo Franci Montorzi (Azienda Ospedaliero Universitaria Careggi, Firenze), Carlo Fumagalli (Azienda Ospedaliero Universitaria Careggi, Firenze), Stefano Fumagalli (Azienda Ospedaliero Universitaria Careggi, Firenze), Pietro Gareri (ASP Catanzaro), Pier Paolo Gasbarri (Associazione Nazionale Strutture Territoriali e per la Terza Età, Roma), Antonella Giordano (Azienda Ospedaliero Universitaria Careggi, Firenze), Evelina Giuliani [USL Toscana Centro, Ospedale Santa Maria Annunziata, Bagno a Ripoli (FI)], Roberta Granata (RSA Villa Sacra Famiglia, IHG, Roma), Antonio Greco [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Nadia Grillo [RSA San Domenico, Lamezia Terme (CZ); Casa di Riposo San Domenico, Lamezia Terme (CZ); RSA Villa Elisabetta, Cortale (CZ)], Antonio Guaita [Istituto Geriatrico Camillo Golgi, Abbiategrasso (MI)], Liana Gucciardino (ASP Agrigento), Andrea Herbst (Azienda Ospedaliero Universitaria Careggi, Firenze), Marilena Iarrera [RSA Sant’Angelo di Brolo (ME)], Giuseppe Ielo (Azienda Ospedaliero-Universitaria di Parma), Valerio Alex Ippolito [Casa Protetta Villa Azzurra, Roseto Capo Spulico (CS)], Antonella La Marca [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Umberto La Porta (Azienda Ospedaliero-Universitaria di Parma), Ilaria Lazzari (Policlinico Sant’Orsola-Malpighi, Bologna), Diana Lelli (Policlinico Universitario Campus Bio-Medico, Roma), Yari Longobucco (Azienda Ospedaliero-Universitaria di Parma), Francesca Lubian (Ospedale di Bolzano), Giulia Lucarelli (Azienda Ospedaliero Universitaria Careggi, Firenze; USL Toscana Centro, Ospedale San Jacopo, Pistoia), Flaminia Lucchini (Azienda Ospedaliero Universitaria Careggi, Firenze), Daniela Lucente [Spedali Civili, Montichiari (BS)], Lorenzo Maestri (Policlinico Sant’Orsola-Malpighi, Bologna), Marcello Maggio (Azienda Ospedaliero-Universitaria di Parma), Paola Mainquà [Azienda Ospedali Riuniti Marche Nord, Fano (PU)], Mariangela Maiotti [Ospedale San Giovanni Battista, Foligno (PG)], Alba Malara [RSA San Domenico, Lamezia Terme (CZ); Casa di Riposo Villa Marinella, Amantea (CS); Casa Protetta Madonna del Rosario, Lamezia Terme (CZ); Casa Protetta Villa Azzurra, Roseto Capo Spulico (CS); Centro di Riabilitazione San Domenico, Lamezia Terme (CZ); RSA Casa Amica, Fossato Serralta (CZ); RSA La Quiete, Castiglione Cosentino (CS); Casa di Riposo San Domenico, Lamezia Terme (CZ); RSA Villa Elisabetta, Cortale (CZ); RSA Villa Santo Stefano, S. Stefano di Rogliano (CS); RSA Villa Silvia, Altilia Grimaldi (CS)], Carlotta Mancini (Azienda Ospedaliero Universitaria Careggi, Firenze), Irene Mancuso [RSA San Giovanni di Dio, Patti (ME)], Eleonora Marelli [Istituto Geriatrico Camillo Golgi, Abbiategrasso (MI)], Alessandra Marengoni [Spedali Civili, Montichiari (BS)], Eleonora Marescalco (Azienda Ospedale Università di Padova), Benedetta Martin [Ospedale di Avezzano (AQ)], Valentina Massa [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Giulia Matteucci (Azienda Ospedaliero-Universitaria di Ferrara), Irene Mattioli (Azienda Ospedaliero-Universitaria di Ferrara), Liliana Mazza (Policlinico Sant’Orsola-Malpighi, Bologna), Carmela Mazzoccoli (Ospedale Giovanni XXIII Policlinico di Bari), Fiammetta Monacelli (IRCCS Policlinico San Martino, Genova), Paolo Moneti (RSA Villa Gisella, Firenze), Fabio Monzani (Azienda Ospedaliero Universitaria Pisana; RSA Villa Isabella, Pisa), Federica Morellini (Azienda Ospedaliero-Universitaria di Modena), Maria Teresa Mormile (ASL Napoli 2 Nord), Enrico Mossello (Azienda Ospedaliero Universitaria Careggi, Firenze), Chiara Mussi (Azienda Ospedaliero-Universitaria di Modena), Francesca Maria Nigro (USL Toscana Centro, Ospedale Santo Stefano, Prato), Marianna Noale (RSA AltaVita, Istituzioni Riunite di Assistenza, Padova), Chukwuma Okoye (Azienda Ospedaliero Universitaria Pisana), Giuseppe Orio (Policlinico Sant’Orsola-Malpighi, Bologna), Sara Osso [RSA La Quiete, Castiglione Cosentino (CS)], Chiara Padovan (Azienda Ospedale Università di Padova), Annalisa Paglia (Azienda Ospedaliero Universitaria di Foggia), Giulia Pelagalli (Azienda Ospedaliero Universitaria Careggi, Firenze), Laura Pelizzoni (Policlinico Sant’Orsola-Malpighi, Bologna), Agostino Perri [RSA La Quiete, Castiglione Cosentino (CS)], Maria Perticone [Casa di Riposo Villa Marinella, Amantea (CS)], Giacomo Piccardo (IRCCS Policlinico San Martino, Genova), Alessandro Picci (Presidio Ospedaliero Universitario Santa Maria della Misericordia, Udine), Margherita Pippi [Ospedale San Giovanni Battista, Foligno (PG)], Giuseppe Provenzano (ASP Agrigento), Matteo Pruzzo (IRCCS Policlinico San Martino, Genova), Francesco Raffaele Addamo [RSA San Giovanni di Dio, Patti (ME)], Cecilia Raffaelli (Azienda Ospedale Università di Padova), Francesca Remelli (Azienda Ospedaliero-Universitaria di Ferrara), Onofrio Resta (Ospedale Giovanni XXIII Policlinico di Bari), Antonella Riccardi (Policlinico Sant’Orsola-Malpighi, Bologna), Daniela Rinaldi (Ospedale di Comunità (Camposampiero), Distretto Alta Padovana, ULSS 6 Euganea, Padova), Renzo Rozzini (Fondazione Poliambulanza, Brescia), Matteo Rubino (IRCCS Policlinico San Martino, Genova), Carlo Sabbà (Ospedale Giovanni XXIII Policlinico di Bari), Leonardo Sacco [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Mariateresa Santoliquido [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Mariella Savino [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Francesco Scarso (Azienda Ospedaliero-Universitaria Sant’Andrea, Roma), Giuseppe Sergi (Azienda Ospedale Università di Padova), Gaetano Serviddio (Azienda Ospedaliero Universitaria di Foggia), Claudia Sgarito (ASP Agrigento), Giovanni Sgrò [RSA Istituto Santa Maria del Soccorso, Serrastretta (CZ); RSA San Vito Hospital, San Vito sullo Jonio (CZ); Casa Protetta Villa Mariolina, Montauro (CZ); Casa Protetta Villa Sant’Elia, Marcellinara (CZ)], Chiara Sidoli (Ospedale San Gerardo, Monza), Federica Sirianni [Casa di Riposo Villa Marinella, Amantea (CS)], Vincenzo Solfrizzi (Ospedale Giovanni XXIII Policlinico di Bari), Benedetta Soli (Azienda Ospedaliero-Universitaria di Modena), Debora Spaccaferro [RSA Estensiva, IHG, Guidonia (RM); RSA Intensiva, IHG, Guidonia (RM)], Fausto Spadea [RSA Casa Amica, Fossato Serralta (CZ)], Laura Spadoni [Ospedale San Giovanni Battista, Foligno (PG)], Laura Tafaro (Azienda Ospedaliero-Universitaria Sant’Andrea, Roma), Luca Tagliafico (IRCCS Policlinico San Martino, Genova), Andrea Tedde (Azienda Ospedaliero-Universitaria di Modena), Camilla Terziotti (Azienda Ospedale Università di Padova), Giuseppe Dario Testa (USL Toscana Centro, Ospedale San Jacopo, Pistoia), Maria Giulia Tinti [Casa Sollievo della Sofferenza, S. Giovanni Rotondo (FG)], Francesco Tonarelli (USL Toscana Centro, Presidio Misericordia e Dolce, Prato), Elisabetta Tonon (USL Toscana Centro, Ospedale San Jacopo, Pistoia), Caterina Trevisan (Ospedale di Comunità (Camposampiero), Distretto Alta Padovana, ULSS 6 Euganea, Padova; Azienda Ospedale Università di Padova), Rita Ursino [I RSA Geriatria, IHG, Guidonia (RM)], Filomena Vella (Azienda Sanitaria Universitaria Giuliano Isontina, Trieste), Maria Villanova (Azienda Ospedale Università di Padova), Aurora Vitali (Azienda Ospedaliero-Universitaria di Ferrara), Stefano Volpato (Azienda Ospedaliero-Universitaria di Ferrara), Francesca Zoccarato (Azienda Ospedale Università di Padova), Sonia Zotti (Policlinico Universitario Campus Bio-Medico, Roma), Amedeo Zurlo (Azienda Ospedaliero-Universitaria di Ferrara).
Author contributions
Conceptualization: AC, MN, RAI. Methodology: AC, MN, CT, PG, AM, EM, SV, SF, SS, GZ. Data Collection: the GeroCovid Observational Working Group. Formal analysis and investigation: MN, FF, AB, CT. Writing—original draft preparation: AC, AB, FF, AMA, PG, CT, EM. Writing—review and editing: AC, AMA, AM, GB, FM, RAI. Resources: RAI. Supervision: AC, RAI.
Funding
This study received unconditioned funding from Angelini Pharma S. p. A. The founder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication. All authors declare no other competing interests.
Declarations
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.
Ethical approval
The GeroCovid Observational study protocol was approved by the Campus Bio-Medico University Ethical Committee in April 2020. All participating investigational sites further submitted relevant sub-protocols to their competent local ethical committee and institutional review boards, as applicable according to Italian regulations. All investigators accepted to work according to the Good Clinical Practice (GCP) (ICH E6-R2).
Informed consent
Written or dematerialized informed consent was obtained from each participant. Alternatively, a written declaration was kept on file by the local investigator, which responded to applicable derogations during the pandemic.
The members of The GeroCovid Observational Working Group are listed in Acknowledgements.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Niculescu I Arora T Iaboni A Screening for depression in older adults with cognitive impairment in the homecare setting: a systematic review Aging Ment Health 2021 25 9 1585 1594 10.1080/13607863.2020.1793899 32677506
2. Ulbricht CM Rothschild AJ Hunnicutt JN Lapane KL Depression and cognitive impairment among newly admitted nursing home residents in the USA: depression and cognitive impairment Int J Geriatr Psychiatry 2017 32 11 1172 1181 10.1002/gps.4723 28544134
3. Pickett YR Raue PJ Bruce ML Late-life depression in home healthcare Aging Health 2012 8 3 273 284 10.2217/ahe.12.28 23483034
4. Steffen A Nübel J Jacobi F Bätzing J Holstiege J Mental and somatic comorbidity of depression: a comprehensive cross-sectional analysis of 202 diagnosis groups using German nationwide ambulatory claims data BMC Psychiatry 2020 20 1 142 10.1186/s12888-020-02546-8 32228541
5. Parrotta I De Mauleon A Abdeljalil AB Depression in people with dementia and caregiver outcomes: results from the European Right Time Place Care Study J Am Med Dir Assoc 2020 21 6 872 878.e1 10.1016/j.jamda.2020.02.023 32307275
6. Henskens M Nauta IM Vrijkotte S Drost KT Milders MV Scherder EJA Mood and behavioral problems are important predictors of quality of life of nursing home residents with moderate to severe dementia: a cross-sectional study PLoS ONE 2019 14 12 e0223704 10.1371/journal.pone.0223704 31860641
7. Birrer RB, Vemuri SP (2004) Depression in later life: a diagnostic and therapeutic challenge 69(10):8
8. Van der Veen D Van Zelst W Schoevers R Comijs H Oude Voshaar R Comorbid anxiety disorders in late-life depression: results of a cohort study Int Psychogeriatr 2015 27 7 1157 1165 10.1017/S1041610214002312 25370017
9. Cipriani A Furukawa TA Salanti G Chaimani A Atkinson LZ Ogawa Y Leucht S Ruhe HG Turner EH Higgins JPT Egger M Takeshima N Hayasaka Y Imai H Shinohara K Tajika A Ioannidis JPA Geddes JR Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis Lancet 2018 391 10128 1357 1366 10.1016/S0140-6736(17)32802-7 29477251
10. Haria M Fitton A McTavish D Trazodone. A review of its pharmacology, therapeutic use in depression and therapeutic potential in other disorders Drugs Aging 1994 4 4 331 355 10.2165/00002512-199404040-00006 8019056
11. Jaffer KY, Chang T, Vanle B, Dang J, Steiner AJ, Loera N, Abdelmesseh M, Danovitch I, Ishak WW (2017) Trazodone for insomnia: a systematic review. Innov Clin Neurosci 14(7–8):24–34
12. Khouzam HR A review of trazodone use in psychiatric and medical conditions Postgrad Med 2017 129 1 140 148 10.1080/00325481.2017.1249265 27744763
13. Macías Saint-Gerons D, Huerta Álvarez C, García Poza P, Montero Corominas D, de la Fuente Honrubia C (2018) Trazodone utilization among the elderly in Spain. A population based study. Rev Psiquiatr Salud Ment 11(4):208–215. 10.1016/j.rpsm.2016.11.003
14. Iaboni A Bronskill SE Reynolds KB Wang X Rochon PA Herrmann N Flint A Changing pattern of sedative use in older adults: a population-based cohort study Drugs Aging 2016 33 7 523 533 10.1007/s40266-016-0380-3 27241038
15. Trevisan C Del Signore S Fumagalli S Gareri P Malara A Mossello E Volpato S Monzani F Coin A Bellelli G Zia G Ranhoff AH Antonelli Incalzi R GeroCovid Working Group Assessing the impact of COVID-19 on the health of geriatric patients: The European GeroCovid Observational Study Eur J Intern Med 2021 87 29 35 10.1016/j.ejim.2021.01.017 33573885
16. Abbatecola AM, Antonelli Incalzi R, Malara A, Palmieri A, Di Lonardo A, Borselli G, Russo M, Noale M, Fumagalli S, Gareri P, Mossello E, Trevisan C, Volpato S, Monzani F, Coin A, Bellelli G, Okoye C, Del Signore S, Zia G, Bottoni E, Cafariello C, Onder G, GeroCovid Observational, GeroCovid Vax Group (2021) Disentangling the impact of COVID-19 infection on clinical outcomes and preventive strategies in older persons: an Italian perspective. J Gerontol Geriatr 2021:1–9. 10.36150/2499-6564-N440
17. MedDRA (2022) Organisation. https://www.meddra.org/about-meddra/organisation. Accessed 8 May 2022
18. Hsu CW Tseng WT Wang LJ Yang YH Kao HY Lin PY Comparative effectiveness of antidepressants on geriatric depression: real-world evidence from a population-based study J Affect Disord 2022 1 296 609 615 10.1016/j.jad.2021.10.009
19. Casey DA Depression in older adults Prim Care Clin Off Pract 2017 44 3 499 510 10.1016/j.pop.2017.04.007
20. Hedna K Fastbom J Jonson M Wilhelmson K Waern M Psychoactive medication use and risk of suicide in long-term care facility residents Int J Geriatr Psychiatry 2022 10.1002/gps.5684 35142388
21. Giovannini S Onder G van der Roest HG Topinkova E Gindin J Cipriani MC Denkinger MD Bernabei R Liperoti R SHELTER Study Investigators Use of antidepressant medications among older adults in European long-term care facilities: a cross-sectional analysis from the SHELTER study BMC Geriatr 2020 20 1 310 10.1186/s12877-020-01730-5 32854659
22. Settimo L Taylor D Evaluating the dose-dependent mechanism of action of trazodone by estimation of occupancies for different brain neurotransmitter targets J Psychopharmacol (Oxf) 2018 32 1 96 104 10.1177/0269881117742101
23. Bossini L Casolaro I Koukouna D Cecchini F Fagiolini A Off-label uses of trazodone: a review Expert Opin Pharmacother 2012 13 12 1707 1717 10.1517/14656566.2012.699523 22712761
24. Odagaki Y Toyoshima R Yamauchi T Trazodone and its active metabolite m-chlorophenylpiperazine as partial agonists at 5-HT1A receptors assessed by [35S]GTPγS binding J Psychopharmacol (Oxf) 2005 19 3 235 241 10.1177/0269881105051526
25. Gareri P Falconi U De Fazio P De Sarro G Conventional and new antidepressant drugs in the elderly Prog Neurobiol 2000 61 4 353 396 10.1016/S0301-0082(99)00050-7 10727780
26. Saletu-Zyhlarz GM Anderer P Arnold O Saletu B Confirmation of the neurophysiologically predicted therapeutic effects of trazodone on its target symptoms depression, anxiety and insomnia by postmarketing clinical studies with a controlled-release formulation in depressed outpatients Neuropsychobiology 2003 48 4 194 208 10.1159/000074638 14673218
27. López-Pousa S Garre-Olmo J Vilalta-Franch J Turon-Estrada A Pericot-Nierga I Trazodone for Alzheimer’s disease: a naturalistic follow-up study Arch Gerontol Geriatr 2008 47 2 207 215 10.1016/j.archger.2007.07.010 17897735
28. Lebert F Stekke W Hasenbroekx C Pasquier F Frontotemporal dementia: a randomised, controlled trial with trazodone Dement Geriatr Cogn Disord 2004 17 4 355 359 10.1159/000077171 15178953
29. Cankurtaran ES Management of behavioral and psychological symptoms of dementia Noro Psikiyatri Arsivi 2014 51 4 303 312 10.5152/npa.2014.7405 28360647
30. Seitz DP, Adunuri N, Gill SS, Gruneir A, Herrmann N, Rochon P (2011) Antidepressants for agitation and psychosis in dementia. Cochrane Dementia and Cognitive Improvement Group (ed) Cochrane database system review. Published online February 16, 2011. 10.1002/14651858.CD008191.pub2
31. Stryjer R Strous RD Bar F Poyurovsky M Weizman A Kotler M Treatment of neuroleptic-induced akathisia with the 5-HT2A antagonist trazodone Clin Neuropharmacol. 2003 26 3 137 141 10.1097/00002826-200305000-00006 12782915
32. Burke AD Goldfarb D Bollam P Khokher S Diagnosing and treating depression in patients with Alzheimer’s disease Neurol Ther 2019 8 2 325 350 10.1007/s40120-019-00148-5 31435870
33. Kales HC Gitlin LN Lyketsos CG Assessment and management of behavioral and psychological symptoms of dementia BMJ 2015 350 h369 h369 10.1136/bmj.h369 25731881
34. Draper B Berman K Tolerability of selective serotonin reuptake inhibitors: issues relevant to the elderly Drugs Aging 2008 25 6 501 519 10.2165/00002512-200825060-00004 18540689
35. Barry LC Abou JJ Simen AA Gill TM Under-treatment of depression in older persons J Affect Disord 2012 136 3 789 796 10.1016/j.jad.2011.09.038 22030136
36. Boehlen FH Freigofas J Herzog W Meid AD Saum KU Schoettker B Brenner H Haefeli WE Wild B Evidence for underuse and overuse of antidepressants in older adults: results of a large population-based study Int J Geriatr Psychiatry 2019 34 4 539 547 10.1002/gps.5047 30623499
37. Drageset J Eide GE Ranhoff AH Anxiety and depression among nursing home residents without cognitive impairment Scand J Caring Sci 2013 27 4 872 881 10.1111/j.1471-6712.2012.01095.x 23072281
38. Ray KD Mittelman MS Music therapy: a nonpharmacological approach to the care of agitation and depressive symptoms for nursing home residents with dementia Dementia 2017 16 6 689 710 10.1177/1471301215613779 26519453
39. Guerrera CS Furneri G Grasso M Caruso G Castellano S Drago F Di Nuovo S Caraci F Antidepressant drugs and physical activity: a possible synergism in the treatment of major depression? Front Psychol 2020 11 857 10.3389/fpsyg.2020.00857 32435223
|
PMC010xxxxxx/PMC10197073.txt |
==== Front
Stat Biosci
Stat Biosci
Statistics in Biosciences
1867-1764
1867-1772
Springer US New York
9370
10.1007/s12561-023-09370-0
Article
Bayesian Index Models for Heterogeneous Treatment Effects on a Binary Outcome
http://orcid.org/0000-0002-8994-9583
Park Hyung G. parkh15@nyu.edu
1
Wu Danni 1
Petkova Eva 1
Tarpey Thaddeus 1
Ogden R. Todd 2
1 grid.137628.9 0000 0004 1936 8753 Division of Biostatistics, Department of Population Health, New York University School of Medicine, New York, NY 10016 USA
2 grid.21729.3f 0000000419368729 Department of Biostatistics, Columbia University, New York, NY 10032 USA
19 5 2023
2023
15 2 397418
16 12 2022
2 3 2023
21 3 2023
© The Author(s) under exclusive licence to International Chinese Statistical Association 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
This paper develops a Bayesian model with a flexible link function connecting a binary treatment response to a linear combination of covariates and a treatment indicator and the interaction between the two. Generalized linear models allowing data-driven link functions are often called “single-index models” and are among popular semi-parametric modeling methods. In this paper, we focus on modeling heterogeneous treatment effects, with the goal of developing a treatment benefit index (TBI) incorporating prior information from historical data. The model makes inference on a composite moderator of treatment effects, summarizing the effect of the predictors within a single variable through a linear projection of the predictors. This treatment benefit index can be useful for stratifying patients according to their predicted treatment benefit levels and can be especially useful for precision health applications. The proposed method is applied to a COVID-19 treatment study.
Keywords
Bayesian single-index models
Heterogeneous treatment effects
Precision medicine
http://dx.doi.org/10.13039/100000025 National Institute of Mental Health 5 R01 MH099003 Park Hyung G. http://dx.doi.org/10.13039/100006108 National Center for Advancing Translational Sciences 3 UL1TR001445-06A1S2 Wu Danni issue-copyright-statement© International Chinese Statistical Association 2023
==== Body
pmcIntroduction
In precision medicine, a critical concern is to characterize individuals’ heterogeneity in treatment responses in order to enable individual-specific treatment decisions to be made [1–3]. Tailoring medical treatments according to individuals’ characteristics requires inferring individual-level treatment effects (ITE) (as opposed to inferring the treatment effects on average across the entire population). Then, developing an individualized treatment rule (ITR) [4–14, see, e.g.,] naturally follows from drawing inferences about the ITE. As an alternative route to developing an ITR, direct optimization of the population average over a class of ITRs [9, 10, 15, 16, see, e.g.,] may be considered.
Substantial developments have been made in the statistical methodology on the ITE estimation [17, see, e.g., review provided in]. Examples within the frequentist paradigm include R-learner [18], kernel-based multi-task learner [19], neural networks [20], tree-based ensembles [21], and semi-parametric regression [22], among many others. In the Bayesian paradigm, one prominent approach is the causal Bayesian additive regression trees (BART) [23–26, see, e.g.,] among others. Although there are many statistical learning methods that exhibit good performance in capturing complex nonlinear relationship for individual treatment effects, in this paper, we focus on the Bayesian estimation of single-index regression models for ITE, as there has been no specific work designed to model the heterogeneous treatment effects using a single-index model in the Bayesian paradigm, albeit the usefulness that arises from the simplicity in its model formulation.
A single-index model [27, 28, see, e.g.,] is one of the most popular semi-parametric models and provides an efficient way of dealing with multivariate nonparametric regression. Single-index models expand the scope of generalized linear models through a flexible data-driven link function. The model summarizes the effect of the predictors within a single variable through a linear projection of the predictors, called the (single) index. In the context of modeling heterogeneous treatment effects, such an index corresponds to a composite moderator of treatment effects, and we will demonstrate that such an index is useful for optimizing treatment decisions and the approach provides a natural way to summarize uncertainties associated with this composite moderator, through its posterior distribution.
There are broadly two lines of research on Bayesian single-index models. One approach employs a spline-based representation of the link [29–33]. The other line of research employs a Gaussian process-type representation of the link [34–38], where the unknown link function is assumed to be a Gaussian process a priori. In this article, we take the former spline-based approach as it allows us to easily incorporate an identifiability constraint on the link function used to model the heterogeneous treatment effect term.
In this paper, we consider a treatment variable A taking a value in {0,1} with the associated randomization probabilities {π0,π1} in the context of randomized clinical trials (RCTs), with the corresponding potential outcomes denoted as
{Y(0),Y(1)}. Depending on A, the observed outcome is Y=(1-A)Y(0)+AY(1), where the outcome Y is assumed to be a member of the exponential family. Specifically, we focus on a binary outcome Y∈{0,1}, where we assume, without loss of generality, that the value of Y=0 is desired so that Y=1 indicates a bad outcome (e.g., death). On the population level, this means that a small value of h(E[Y]) is desired, where h(·) denotes the canonical link of the assumed exponential family distribution. For the binary outcome considered here, h(·) is the logit function.
The proposed single-index approach to modeling heterogeneous treatment effects was motivated by an application to a COVID-19 convalescent plasma (CCP) treatment study [39]. One of the primary goals of this RCT was to guide CCP treatment recommendations by providing an estimate of a differential treatment outcome when a patient is treated with CCP vs. without CCP, where a larger differential in favor of CCP would indicate a more compelling reason for recommending CCP. The estimated index, defined as a linear combination of covariates X, which is part of the heterogeneous treatment effect term of the model, can be used to discover profiles of patients with COVID-19 associated with different benefits from CCP treatment. Specifically, the covariates X∈Rp are observed pretreatment measurements and predictors of {Y(0),Y(1)}. Our goal is to utilize the information in X to develop an ITR that optimizes the value of h(E[Y]) for future patients.
Method
Optimal Individualized Treatment Rules
In this subsection, we define an optimal ITR. The Bayes decision a∗:x↦{0,1} minimizes, over treatment decision (action) a∈{0,1}, the patient-specific posterior expected loss for a patient with pretreatment characteristic X=x. Let us define the loss function for making treatment decision a as follows:1 L(a,θ,x)=h(E[Y(a)|θ,x]),
where θ represents the model parameters characterizing the relationship between the potential treatment outcomes (Y(0),Y(1)) and predictors X. In (1), E[Y(a)|θ,x]=(1-a)E[Y(0)|θ,x]+aE[Y(1)|θ,x] is the expected outcome under a particular treatment assignment a. Let us denote the observed clinical trial data as D consisting of triplets Di={Xi,Ai,Yi} with i∈{1,…,n}, where Xi∈Rp is a set of observed pretreatment covariates and Yi∈{0,1} is an observed outcome for individual i.
Viewing the loss function L(a,θ,x) in (1) as a function of treatment assignment a given a particular x, the optimal Bayes decision a∗(x) will minimize the posterior expected loss given x, i.e.,a∗(x)=argmina∈{0,1}Eθ|D,x[L(a,θ,x)],
where the expectation is taken with respect to the posterior distribution of θ given the observed data D. In particular, if we define the loss contrast Δ(θ,x):=L(a=1,θ,x)-L(a=0,θ,x), then the above optimal Bayes decision a∗(x) is equivalent to2 a∗(x)=I(Eθ|D,x[Δ(θ,x)]<0),
(where I(·) is the indicator function), which we define as the optimal ITR.
We will utilize the following standard causal inference assumptions [40, 41, see, e.g.,]: no unmeasured confoundedness, Y(a)⊥A given X=x,
positivity, 0<P(A=1|X=x)<1, which together imply h(E[Y(a)|θ,X=x])=h(E[Y|A=a,θ,X=x]), and
stable unit treatment value assumption (SUTVA). Under these assumptions, we can write the loss contrast function Δ(θ,x) in (2) as follows: Δ(θ,x)=h{E[Y|A=1,θ,x]}-h{E[Y|A=0,θ,x]}. Therefore, we can construct the optimal Bayes decision (2) based on posterior inference on the canonical parameter h{E[Y|A,θ,X]} of the exponential family response Y. In the following subsection, we will specify our model for the distributions of (Y|A,θ,X) and θ, to estimate the optimal ITR (2).
Model and Prior Specification
Model
Let Y=(Y1,…,Yn)⊤ be a vector of the treatment outcomes, with each Yi independently following a specific exponential family distribution with density3 f(Yi|ηi,ϕ)=expϕ-1[Yiηi-b(ηi)]+c(Yi,ϕ)ηi=Xi⊤m+g(Xi⊤β,Ai),
where the unknown parameters (which we collectively denote as θ) will be estimated in a Bayesian framework. In (3), b(·) and c(·) are known functions specific to the given member of the exponential family, whereas g(·,·) is an unknown flexible function and ϕ>0 is an unknown dispersion parameter (ϕ=1 specializes to a one-parameter exponential family distribution). Throughout we will drop ϕ from (3) because it is fixed at unity in our motivating dataset that has Bernoulli responses.
The canonical parameter η∈R in (3) of the response distribution is related to the treatment decision loss function L(a,θ,x) in (1), through η(θ,x,a)=h(E[Y|θ,x,a])=h(E[Y(a)|θ,x])=L(a,θ,x), under the standard causal inference assumptions.
Within the specification of η in model (3), the first term X⊤m represents the pretreatment covariates’ “main” effect, whereas the second term g(X⊤β,A) is the X-by-A interaction effect. This interaction effect is characterized by an unspecified treatment a-specific smooth function g(u, a) (a=0,1) which is a function of a linear projection u=X⊤β∈R. The projection vector β∈Rp is subject to ‖β‖=1, i.e., restricted to β∈Sp-1, where Sp-1 is the p-1-dimensional unit sphere, and the single-index X⊤β provides a dimension reduction specifically for the X-by-A interaction effect. In (3), for any X and β, we shall impose the following identifiability condition for the component g4 E[g(X⊤β,A)|X]=0,
which separates the component g(X⊤β,A) of interest (the “prescriptive” term representing the heterogeneous treatment effect), from the component X⊤m (the “prognostic” term that does not represent the heterogeneous treatment effect). In general, the covariates, X, represented in the terms g(X⊤β,A) and X⊤m in (3), are not necessarily the same variates, but for notational simplicity, the same notation for these sets of covariates was employed. As another abuse of notation for simplicity, the treatment A’s main effect, which can be represented by β0A∈R for some unknown β0∈R will be estimated, together with the model intercept, as part of the component X⊤m in (3).
Remark 2.1
Model (3) with the identifiability condition (4) is more suitable to conduct a posterior inference for heterogeneous treatment effects than the model with η=X⊤m+g(X⊤β)A, because this particular parametrization (3) is invariant of the choice of coding of A. In the latter model, the choice of the treatment A coding can meaningfully impact posterior inferences because the two effects X⊤m and g(X⊤β)A can alias each another, particularly when the same set X enters into the both terms. Under condition (4), the effect captured by the prognostic term, X⊤m, and the prescriptive term, g(X⊤β,A), can be distinguished regardless of treatment A coding.
For an individual with pretreatment characteristics x, the loss contrast Δ(θ,x) in (2) under model (3) is5 Δ(θ,x)=g(x⊤β,A=1)-g(x⊤β,A=0).
The loss contrast (3) indicates that only the parameters g and β (and not m and ϕ) in model (3) are used to specify the ITR (2), hence g and β correspond to the “signal” parameters of interest. Given definition (2), we will define a “treatment benefit index” (TBI) in terms of a (posterior) probability,6 TBI(x):=P(Δ(θ,x)<0|D)∈[0,1],
that is, the probability of the (active) treatment A=1 providing a greater benefit than the treatment A=0 (for a patient with pretreatment characteristic x), in which the probability is evaluated with respect to the posterior distribution of θ. The optimal Bayes decision a∗(x) in (2) is then represented by I(TBI(x)>0.5). Since a large value of (6) indicates a large expected “benefit” of taking the active treatment A=1 vs. control A=0, the TBI(x) in (6) constructs a “gradient” of the active treatment’s benefit that ranges from 0 to 1,
as a function of patient characteristic x. Furthermore, for each X=x, we can obtain a posterior distribution of the treatment a-specific expected outcome h-1{x⊤m+g(x⊤β,a)} based on the posterior distribution of the parameters θ={m,g,β}.
Representation of the Link Function g
Following [29], we will use a cubic B-spline basis to represent the flexible function g(·,a) of (3).
Using B-splines is appealing because the basis functions are strictly local, as each basis function is only non-zero over the intervals between five adjacent knots [42].
Using splines allows us to easily incorporate the constraint (4) on the function g, as we describe shortly.
For each fixed β, the function g is represented as follows:7 g(β⊤xi,ai)=ψ~(β⊤xi)⊤γ~ai(i=1,…,n)
for some fixed L-dimensional basis ψ~(·)∈RL
(e.g., B-spline basis on evenly spaced knots on a bounded range of {β⊤xi}i=1n) and a set of unknown treatment a-specific spline coefficients γ~a∈RL (a=0,1). In our simulation illustration and in the application, we used a cubic B-spline basis with L=4+[n1/5.5] where [n1/5.5] denotes the integer part of n1/5.5, as recommended by [43].
Given representation (7) for g, the identifiability constraint (4) is implied by the linear constraint:8 π0γ~0+π1γ~1=πγ~=0,
where π=[π0IL;π1IL] with πa=P(A=a) (i.e., the randomization probabilities) is the L×2L matrix (in which IL denotes the L×L identity matrix) and γ~=(γ~0⊤,γ~1⊤)∈R2L is an unknown basis coefficient vector.
To represent (7) in matrix notation, given β, let the n×L matrices D~β,a (a=0,1) denote the evaluation matrices of the basis ψ~(·) on {β⊤xi}i=1n, specific to the treatment A=a (a=0,1), whose ith row is the 1×L vector ψ~(β⊤xi)⊤ if Ai=a and a row of zeros 0L⊤ if Ai≠a. Then, the column-wise concatenation of the design matrices D~β,a (a=0,1), i.e., the n×2L matrix D~β=[D~β,0;D~β,1] defines the model matrix associated with γ~∈R2L. Then, we can represent the function g in (7) evaluated on the sample data, by the length-n vector: g=D~βγ~∈Rn.
The linear constraint (8) on γ~ can be conveniently absorbed into the model matrix D~β by reparametrization, as we describe next. We can find a 2L×L basis matrix Q (that spans the null space of the linear constraint (8)) such that if we set γ~=Qγ for any arbitrary vector γ∈RL, then the resulting vector γ~∈R2L automatically satisfies the constraint (8). Such a basis matrix Q can be constructed by a QR decomposition of the matrix π⊤ in (8). Then representation g=D~βγ~ can be reparametrized, in terms of the unconstrained vector γ∈RL, by replacing D~β with the reparametrized model matrix Dβ=D~βQ, yielding the representation g=Dβγ.
Once we have an inferential procedure on γ, we can also consider inference on the transformed parameter γ~=Qγ, from which we can make inference on the treatment a-specific functions g(β⊤·,a)=ψ~(β⊤·)⊤γ~a (a=0,1).
We note that ψ~(·)∈RL in (7) defines a system of functions specifically chosen to be used as building blocks to represent a (smooth, as implied by penalized splines) link function g(·,a) for each treatment condition a. If a different basis function is used to represent the link function, we may have a different performance. We may need to identify a best suitable basis for the data which depend on the underlying heterogeneous treatment effect (i.e., the underlying log odds ratio function). For example, if there is a reason to expect that the log odds ratio function (5) is a jagged (or, less likely, a cyclic) function over the index x⊤β, then a wavelet [44, e.g.,] (or Fourier [45, e.g.,]) basis might be more suitable. Although the associated basis coefficient vector, γ~, will still be subject to the identifiability constraint (8) for a different basis system, a specifically tailored prior and penalization would be more appropriate, for example, (L1 type) Laplace-Zero prior with some hyperparameter determining the sparsity basis [44].
Prior Specification
How we specify priors for β, m, and γ associated with model (3) is given in this subsection.For the distribution of β∈Sp-1, we will use the von Mises–Fisher with concentration parameter κ>0 and modal parameter β0∈Sp-1, 9 P(β)∝exp(κβ⊤β0),
which is a probability distribution for β on the (p-1) unit sphere in Rp.
We will use m∼N(m0,R), for some vector m0∈Rp and p×p symmetric positive definite matrix R.
Since the domain of the function g in (3) depends on β, the prior on g will depend on β. Conditioning on β, following [29], we will use data-dependent prior for γ∈RL, 10 γ|β∼N(ΣρDβ⊤WZ,Σ0),
where Σρ:=(Dβ⊤WDβ+ρI)-1,
and Σ0:=(Dβ⊤WDβ)-1 which corresponds to a special case of Σρ at ρ=0. The prior (10) is a Zellner’s g-prior that has the same dispersion matrix as a weighted least squares estimator defined based on the vector of adjusted responses Z and the matrix of weights W, which are specified in the next subsection. In the prior (10), ρ≥0 is a hyperparameter which will be selected via an empirical Bayes procedure with the generalized cross-validation (GCV), as in [29]. An advantage of using the prior (10) is that it allows us to analytically integrate γ out of the joint posterior P(β,γ|m,Y), facilitating the Gibbs sampling of β.
Posterior Computation
To conduct posterior inference on (m,β,γ), we will simulate samples from the joint posterior P(m,β,γ|Y) (where we use Y=(y1,…,yn)⊤ to denote the observed treatment outcomes). Since it is difficult to draw samples directly from this joint posterior distribution, we will use a Metropolis-Within-Gibbs algorithm. The Gibbs algorithm will iterate between the following two Steps: Step 1) sample m from P(m|β,γ,Y) and Step 2) sample (β,γ) from P(β,γ|m,Y). Specifically, in Step 2, since the joint conditional posterior P(β,γ|m,Y) does not have a convenient form to directly sample from, we will employ a Metropolis–Hastings step.
Conditional Posteriors
Derivation of (m|β,γ,Y). For fixed β and γ, we will quadratically approximate the log likelihood function of m at its mode, which we denote by mˇ. To find the mode mˇ, we will use a Fisher scoring, iteratively updating the center of the quadratic approximation. For fixed β and γ, at the convergence of the Fisher scoring, we define the adjusted response vector Zˇ:=(zˇ1,…,zˇn)⊤∈Rn where zˇi:=h′(μˇi)(yi-μˇi)+ηˇi, in which ηˇi=mˇ⊤xi+ψ(β⊤xi)⊤γ and μˇi=h-1(ηˇi), and the n×n weight matrix Wˇ=diag(wˇi), where wˇi=1/{(h′(μˇi))2V(μˇi)}, in which V(μˇi)=μˇi(1-μˇi). Given each β and γ, the negative log likelihood of m is approximately represented in terms of a weighted least squares (WLS) objective function (up to a constant of proportionality), ∑i=1nwˇi(zˇi-m⊤xi)2=(Zˇ-Xm)⊤Wˇ(Zˇ-Xm).
Given the prior m∼N(m0,R) and the above approximated negative log likelihood, the conditional posterior for m is given by 11 P(m|β,γ,Y)=N(R-1+X⊤WˇX)-1(R-1m0+X⊤WˇZˇ),(R-1+X⊤WˇX)-1.
Derivation of (β,γ|m,Y). Given the joint conditional posterior P(β,γ|m,Y)=P(β|m,Y)P(γ|β,m,Y), we will first sample β from P(β|m,Y) and then γ from P(γ|β,m,Y). Specifically, following [29], we will use a Metropolis–Hastings algorithm to sample β from p(β|m,Y). However, this approach employed in [29] cannot be directly applied to our settings, due to the non-Gaussian likelihood. Thus, we will perform a quadratic approximation of the negative log likelihood function of γ at its mode, which we denote by γˇ. To find γˇ, as in Step 1, we will conduct a Fisher scoring. For each fixed β and m, this quadratic approximation at the convergence of the Fisher scoring is summarized in the form of the WLS objective function (up to a constant of proportionality), 12 ∑i=1nwi(zi-ψ(β⊤xi)⊤γ)2=(Z-Dβγ)⊤W(Z-Dβγ),
as a function of γ, in which Z:=(z1,…,zn)⊤∈Rn is the adjusted response vector with zi:=h′(μ^i)(yi-μ^i)+η^i obtained at the convergence, where η^i=m⊤xi+ψ(β⊤xi)⊤γˇ and μ^i=h-1(η^i) and W=diag(wi) is the n×n weight matrix with wi=1/{(h′(μ^i))2V(μ^i)}. Given the quadratic approximation (12), we can write the joint conditional posterior P(β,γ|m,Y): 13 P(β,γ|m,Y)=P(Y|γ,β,m)P(γ|β,m)P(β)∝exp{-12(Z-Dβγ)⊤W(Z-Dβγ)}P(γ|β,m)exp(κβ⊤β0).
Given the conditional prior P(γ|β,m) in (10), we can write the terms involving γ in (13) as follows: 14 ∝exp-12(Z-Dβγ)⊤W(Z-Dβγ)+(γ-ΣρDβ⊤WZ)⊤Σ0-1(γ-ΣρDβ⊤WZ)∝exp-12(γ-Σ0Dβ⊤WZ)⊤Σ0-1(γ-Σ0Dβ⊤WZ)+(γ-ΣρDβ⊤WZ)⊤Σ0-1(γ-ΣρDβ⊤WZ)=exp-122γ⊤Σ0-1γ-2γ⊤(I+Σ0-1Σρ)Dβ⊤WZ+Z⊤WZ+Z⊤W⊤DβΣρΣ0-1ΣρDβ⊤WZ=exp-122γ⊤Σ0-1γ-2γ⊤(I+Σ0-1Σρ)Dβ⊤WZ+S1(β),
where S1(β)=Z⊤WZ+Z⊤W⊤DβΣρΣ0-1ΣρDβ⊤WZ.
Specifically, given (14), we can analytically integrate γ out of the joint conditional P(β,γ|m,Y) in (13), which yields 15 P(β|m,Y)=∫P(β,γ|m,Y)dγ∝∫exp-122γ⊤Σ0-1γ-2γ⊤(I+Σ0-1Σρ)Dβ⊤WZ+S1(β)1|Σ0|1/2exp(κβ⊤β0)dγ∝φγ~(I+Σ0-1Σρ)Dβ⊤WZexp-12S1(β)exp(κβ⊤β0),
where φγ~(I+Σ0-1Σρ)Dβ⊤WZ is the moment generating function (MGF) of the variate γ~∼N(0,Σ0/2) evaluated at (I+Σ0-1Σρ)Dβ⊤WZ. The familiar closed-form expression of the Gaussian MGF allows us to write the last line of (15) as follows: 16 P(β|m,Y)∝exp14Z⊤W⊤DβΛDβ⊤WZexp-12S1(β)exp(κβ⊤β0),
where Λ=(I+ΣρΣ0-1)Σ0(I+ΣρΣ0-1). The expression (16) provides a closed form for the approximated P(β|m,Y) up to a constant of proportionality, which we will use to conduct a random walk Metropolis Markov chain Monte Carlo (MCMC) algorithm. The MCMC algorithm to sample (β|m,Y) based on (16) is described in the next subsection. Given each m and β, we can sample γ from P(γ|β,m,Y), 17 P(γ|β,m,Y)=N(12Σ0(I+Σ0-1Σρ)Dβ⊤WZ,12Σ0),
derived from expression (14).
MCMC Algorithm for the Posterior Sampling
In this subsection, we provide a detailed sampling scheme based on the conditional posterior derived in the previous subsection. First, we initialize the chain with the maximum likelihood estimates of the parameters (m,β,γ)
of model (3) with representation (7) for the link g, where the tuning parameter ρ≥0 is optimized through the generalized cross-validation (GCV) criterion. We will then cycle through the following steps. Sample m from P(m|β,γ,Y) in (11) given (β,γ).
Sample β from P(β|m,Y) in (16) given m, using the Metropolis algorithm. Specifically, given the current state βcur for β of the chain, a new value βnew is accepted with the acceptance probability min{1,r}, where the Metropolis ratio r is given by r=P(βnew|m,Y)P(βcur|m,Y),
using the conditional posterior (16) given m. Here, we provide some more details on this Metropolis procedure. The proposal distribution for βnew was taken to be von Mises–Fisher with concentration parameter κ to be λprop>0 and direction parameter given by the current value βcur. In the simulation example in the next section, we used λprop=1000, which yielded the acceptance probability of around 0.3∼0.7 for proposal βnew, and the sampler appeared to explore the state space for β adequately (examining the traceplots of typical MCMC chains did not show any peculiarity). We used the R package movMF [46] to generate random samples of β∈Sp-1 from von Mises–Fisher distributions.
For the prior distribution of β in (9), we can choose κ≥0 (typically in the range of 0≤κ<700 [29]), where κ=0 corresponds to an uninformative prior, depending on the degree of confidence in the prior direction β0.
ρ≥0 in (10) is another unknown that controls the smoothness of the data-driven function g, which is crucial to avoid overfitting g. This will be selected via an empirical Bayes procedure using the GCV criterion, at each MCMC update.
Sample γ from P(γ|β,m,Y) in (17) given (β,m).
To obtain the estimated expected response given a new X=x and a treatment condition a∈{0,1}, we take the posterior mean of the expected response h-1η=h-1m⊤x+ψ~(β⊤x)γ~a, based on the posterior sampler output. In particular, we construct a treatment decision rule using the posterior distribution of ψ~(β⊤x)(γ~1-γ~0). Specifically, we will use the posterior probability Pψ~(β⊤x)(γ~1-γ~0)<0|D as the TBI(x), which we will utilize to obtain a decision rule a∗(x)=I(TBI(x)>0.5), using the probability threshold of 0.5.
Simulation Illustration
In this section, via a set of simulation experiments, we compare the performance of the proposed Bayesian single-index approach to modeling heterogeneous treatment effect with an approach that relies on a Bayesian linear model. We replicate the experiment 100 times with sample sizes n=500,1000,2000. We use a cubic B-spline for representing g as in Sect. 2.2.2. The simulation code is available at https://github.com/syhyunpark/bayesSIMML.
Simulation Setting
We independently generated the treatment indicators Ai∈{0,1} from Bernoulli distribution with P(Ai=1)=0.5 and the vector of covariates Xi∈Rp from the mean zero multivariate normal distribution with compound symmetry correlation (=0.2) and the unit variances. Given (Xi,Ai), we generated Yi∼Bernoulli(P(Yi=1)), where logit(P(Yi=1))=m(Xi)+g(Xi,Ai), with the following specifications of the functions m (either a “nonlinear” or “linear” X main effect) and g (either a “nonlinear” or “linear” A-by-X interaction effect):18 mXi=1/2sinπ/2mTXi``nonlinearXmain effect"π/8mTXi``linearXmain effect"gXi,Ai=2exp-βTXi-0.52-0.6Ai-0.5``nonlinearX-by-Ainteraction effect"0.6βTXiAi-0.5``linearX-by-Ainteraction effect"
In (18), we set m=(1,2,3,4,0,…,0)⊤∈Rp and β=(1,0.5,0.25,0.125,0,…,0)⊤∈Rp, where each vector was normalized to have unit norm. We considered the cases with p∈{5,10}. Throughout the paper, we took m∼N(0,52I) and an uninformative prior for β, i.e., set κ=0 in (9). As a comparison method, we used the following logistic linear model: logit(P(Yi=1))=α+m⊤Xi+(β0+β⊤Xi)(Ai-0.5),
with the prior distributions, α∼t(df=3) with location=0 and scale=8, β0∼N(0,52), and m,β∼N(0,52I).
As an evaluation metric, we first consider the expected deviance measured on an independently generated test set (of size n~=104) to assess the accuracy of the models in predicting the new data, as defined by19 Deviance(y~,θ)=-2n~∑i=1n~log1T∑t=1TP(y~i|θt)
(note that it is scaled by the testing set size n~), where T(=2000) is the length of the Markov chain (after a burn-in of 2000) and θt is the tth set of sampled parameter values within the Markov chain obtained from the training set. Accordingly, 1T∑t=1TP(y~i|θt) in (19) corresponds to the average posterior probability of y~i, and the quantity (19) corresponds to a Bayesian version of the deviance, evaluated on the test set (of size n~). Smaller values of (19) are better, indicating greater average accuracy of the predictive model.
Furthermore, we report the expected outcome under the treatment regime a∗, i.e., E[Y(a∗(X))] (which is called the “value” of the regime a∗) that is Monte Carlo approximated by n~-1∑i=1n~y~i(a∗(x~i)) based on the test set of size n~, where the ITR a∗(x)=I(TBI(x)>0.5) (see (6) for the definition of the TBI) which is trained based on the training set. We also report the proportion of correct decision (PCD), which is the proportion of the cases such that a∗(x~i) (i=1,…,n~) match with the correct optimal treatment assignment under the true data generation model.
Simulation Results
Figure 1 below displays the results from the simulation experiments when we vary n∈{500,1000,2000}, p∈{5,10}, and the form of the interaction effect component g∈{``nonlinear'',``linear''}, for the linear X main effect (i.e., m(X)=π8m⊤X) case. In Table 2 Appendix, as MCMC convergence diagnostic, we report the Gelman–Rubin potential scale reduction factor (PSRF) [47] computed for each scenario using the method of [48], which provides some assurance that the sampler has performed reasonably. For the nonlinear A-by-X interaction effect scenarios (i.e., the gray panels in Fig. 1), the proposed index model that utilizes the flexible link function g clearly outperforms the logistic linear model which assumes a restricted linear model on the interaction term, with respect to the all three criteria (the deviance, PCD and the expected outcome). When there is no nonlinearity in the A-by-X interaction effect term in the underlying model (i.e., the white panels in Fig. 1), not surprisingly the logistic linear model outperforms the index model. However, the contrast in the performance between the two models is relatively small, compared to that under the nonlinear A-by-X interaction. This suggests that, in the absence of prior knowledge about the form of the A-by-X interaction effect, the more flexible index model that accommodates nonlinear treatment effect modifications (i.e., the nonlinear g term) can be a useful alternative to the linear model approach.Fig. 1 Results for the linear X main effect (m=``linear'') case, with varying g∈{``nonlinear'',``linear''}, p∈{5,10}, and n∈{500,1000,2000}, comparing the performance of the proposed index model (red) with that of the logistic linear regression model (blue), with respect to the deviance (the first row; a smaller deviance is desired), the proportion of correct decisions (PCD) (the second row; a larger PCD is desired), and the “value” (the expected outcomes under ITRs) (the third row; a smaller value is desired)
Fig. 2 Results for the nonlinear X main effect (m=``nonlinear'') case, with varying g∈{``nonlinear'',``linear''}, p∈{5,10}, and n∈{500,1000,2000}, comparing the performance of the proposed index model (red) with that of the logistic linear regression model (blue), with respect to the deviance (the first row; a smaller deviance is desired), the proportion of correct decisions (PCD) (the second row; a larger PCD is desired), and the “value” (the expected outcomes under ITRs) (the third row; a smaller value is desired)
The results under the nonlinear X main effect case, i.e., m(X)=12sinπ2m⊤Xi, are given in Fig. 2 and are quite similar to those from the linear X main effect case (i.e., the results present in Fig. 1), except that the expected deviance of the models (both the index model and the logistic linear model), which is a generalization of the mean-squared error that assesses the overall predictive performance, slightly increased due to the X main effect model misspecification. However, optimal ITRs are derived based only on the model’s A-by-X interaction effect terms. With respect to the PCD and the expected outcome under ITRs, the results under the linear X main effect and those under the nonlinear X main effect were close to each other, indicating that these models were quite robust to misspecification of the X main effect with respect to the ITR estimation performance. We also note that in both Figs. 1 and 2 there was relatively large variability in the ITR estimation performance for the index model, especially when n=500. On the other hand, the variability in the overall predictive performance metric (deviance) was comparable for the two approaches. This reflects the tendency that estimation of optimal ITRs is generally more challenging than making outcome predictions, as the difference (5) between two predictions has to be computed and its expected error will be invariably larger than for a single prediction.
Application
In this section, we illustrate the proposed model on data from a COVID-19 convalescent plasma (CCP) study [39], a meta-analysis of pooled individual patient data from 8 randomized clinical trials. One of the goals of this study was to guide CCP treatment recommendations by providing an estimate of a differential treatment outcome when a patient is treated with CCP vs. without CCP [49]. A larger differential in favor of CCP would indicate a more compelling reason for recommending CCP. In this context, we aim to use profiles of patients with COVID-19 associated with different benefits from CCP treatment, to optimize treatment decisions.
The study included 2369 hospitalized adults, not receiving mechanical ventilation at randomization, enrolled from April 2020 through March 2021. We used only complete cases for this analysis. A total of 2287 patients were included, with a mean age of 60.3 (SD 15.2) years and 815(35.6%) women. One of the study’s primary outcomes was the binary variable indicating mechanical ventilation or death (hence Y=1 indicates a bad outcome, whose probability we want to minimize) at day 14 post-treatment, where 336 out of 2287 patients (14.7%) experienced the Y=1 case at day 14. The patients were randomized to be treated with either CCP (A=1) (1190 patients, 52%) or control (A=0) (i.e., standard of care; 1097 patients, 48%). Pretreatment patient characteristics were collected at baseline. As in [49], in our application, the baseline variables used to model the covariates “main” effect, i.e., the component associated with the coefficient m in model (3), included age, sex, baseline symptom conditions, age-by-baseline symptom conditions interaction, blood type, the indicators for history of diabetes, pulmonary and cardiovascular disease, and days since the symptoms onset. We also included the RCT-specific intercepts and the patients’ enrollment quarters as part of the covariates “main” effect component.
Since our goal in this analysis is to investigate the differential treatment effect explained by the baseline variables X, we will focus on reporting the estimation results of the heterogeneous treatment effect (HTE) term g(X⊤β,A) in model (3) and the corresponding treatment effect contrast Δ(x,θ) in (5). The patient characteristics X included in the HTE term, along with the sample proportions, are given in the first column of Table 1. The posterior mean of the index coefficients β=(β1,…,β7)⊤, along with the corresponding 95% posterior credible intervals (CrI), are provided in the second column of Table 1. By examining the posterior CrI, the patient’s symptoms severity (oxygen support status) at baseline, blood type, a history of cardiovascular disease, and a history of diabetes appear to be important predictors of HTE, as the CrI of the coefficients associated with these variables do not include 0.
In the first panel of Fig. 3, we display the exponentiated individualized treatment effect, exp(Δ(x,θ)), as a function of the single-index x⊤β. Specifically, the horizontal axis is the posterior mean of x⊤β (a point estimate), where the “observed” posterior mean values xi⊤β (i=1,…,n) (n=2287) computed from the posterior sample mean of β are represented by the small blue ticks along the axis. The uncertainty in the estimation of the coefficient β (as well as that of the coefficient m) is also accounted for in the credible bands in Fig. 3. For the sake of interpretability, we exponentiated the HTE estimate Δ(x,θ)=g(x⊤β,A=1)-g(x⊤β,A=0), so that the vertical axis in the panel represents the odds ratio (CCP vs. control) for a bad outcome (mechanical ventilation or death). An odds ratio of less than 1 indicates a superior CCP efficacy over the control treatment. As most of the observed values xi⊤β of the single-index fall below the line representing the odds ratio of 1, most of the patients are expected to benefit from CCP treatment, except those with the xi⊤β values greater than 0.45, in which the corresponding expected individualized odds ratios are greater than 1 (about 28% of the observed patients). The U-shaped nonlinear relationship between the expected odds ratio and the single index of the model suggests that the use of the flexible link function g in (3) is to be preferred over a more restricted linear model for this HTE modeling.Table 1 Pretreatment patient characteristics X and the corresponding estimated index coefficients β (and 95% CrI)
Pretreatment characteristic xj (sample prevalence, %) Index coefficient βj [95% CrI]
Oxygen support by mask or nasal prongs∗ (1/0) (63%) 0.68 [0.50, 0.80]
Oxygen support by high flow∗ (1/0) (18%) 0.47 [0.16, 0.61]
Age (dichotomized, ≥67) (1/0) (35%) -0.13 [-0.46,0.04]
Blood type (A or AB vs. O or B) (1/0) (37%) -0.31 [-0.49, -0.16]
Cardiovascular disease (1/0) (42%) -0.24 [-0.65,-0.06]
Diabetes (1/0) (34%) -0.26 [-0.52, -0.08]
Pulmonary disease (1/0) (12%) 0.05 [-0.16,0.22]
∗ The reference level: hospitalized but no oxygen therapy required
Fig. 3 The left panel displays the exponentiated version of the estimated individualized treatment effect and the posterior mean of Δ(x)=g(x⊤β,A=1)-g(x⊤β,A=0) in (5) (solid curve), along with the corresponding upper and lower 95% credible interval (CrI) (dashed curves), as a function of the posterior mean of x⊤β. The right panel also displays the expected odds ratio (CCP vs. control) (solid curve) and the corresponding 95% CrI (dashed curves), but the horizontal axis is now the treatment benefit index (TBI) (6), P(exp(Δ(x))<1|D), where exp(Δ(x)) represents the odds ratio, and the TBI probability is evaluated with respect to the posterior distribution of the parameters in Δ. The TBI provides a gradient of benefit that ranges from 0 to 1, with a higher value of the TBI indicating a greater benefit from the CCP treatment, compared to control. The observed values for the quantities on the horizontal axes are represented by the small blue ticks
Although the first panel of Fig. 3 displays an information about the relationship between the (exponentiated) individualized treatment effect exp(Δ(x)) (i.e., the individualized odds ratio) and the posterior mean of the single index x⊤β, this relationship is non-monotonic, which makes it difficult to construct a “gradient” of the treatment benefit of A=1 vs. A=0, as a function of the patient characteristics x. Thus, in the second panel of Fig. 3, we additionally display the individualized odds ratio exp(Δ(x)), as a function of TBI(x) defined in (6), i.e., TBI(x)=P(exp(Δ(x))<1|D), where the probability is evaluated with respect to the posterior distribution of the parameters involving Δ. As a probability, TBI(x) ranges from 0 to 1, where larger values are associated with larger CCP benefit. For example, patients with a large value of TBI(x) (i.e., TBI scores near 1) were expected to experience large, clinically meaningful benefits from CCP.
The second panel of Fig. 3 displays a monotonically decreasing trend of the expected odds ratio (an increasing CCP benefit), as the TBI score increases from 0 to 1. Some portions of the expected odds ratio as well as the corresponding 95% CrI exceed 1 for very small TBI values (near 0), suggesting the possibility of harm from CCP as the TBI approaches 0, whereas the TBI values close to 1 indicate a substantial benefit from the CCP treatment over the control treatment. We can use the TBI scores to stratify patients according to their predicted treatment benefit levels, by setting the treatment decision rule a^∗(x)=I(TBI(x)>0.5)∈{0,1}.
To evaluate the performance of the treatment decision rule a^∗(x), we randomly split the data with a ratio of 2:1 into a training set and a testing set (of size n~), replicated 100 times, each time obtaining a^∗(x) based on the training set, and the corresponding “value” V(a^∗)=E[Y(a^∗(X))] by an inverse probability weighted estimator [50] V^(a^∗)=∑i=1n~YiI(Ai=a^∗(Xi))∑i=1n~I(Ai=a^∗(Xi)) computed based on the testing set (of size n~).
For comparison, we also include two naive rules: treating all patients with Control (“All Control”) and treating all patients with CCP (“All CCP”), each regardless of the individual patients’ characteristics x, in addition to the decision rules based on the Bayesian linear logistic model which was compared with the proposed index model in Sect. 3. The resulting boxplots obtained from the 100 random splits are illustrated in Fig. 4.Fig. 4 Boxplots of “value,” obtained from 100 randomly split testing sets. A smaller “value” is desirable
The results in Fig. 4 demonstrate that the index model and the logistic linear regression perform at a similar level for this dataset, while showing a clear advantage over the näive rules of giving everyone CCP or giving everyone the control treatment: the averaged proportion of patients with the undesirable outcome (i.e., Y=1) was considerably less for the two regression approaches than the two näive rules. This suggests that accounting for patient characteristics can help optimizing treatment decisions. Although some of the nonlinearities in the association between the treatment effect and patient characteristics x is captured by the model-implied odds ratio displayed in the first panel of Fig. 3, for this dataset, the simpler linear model appears to perform nearly as well as the index model due to its model parsimony. However, as demonstrated in Sect. 3, the more flexible index model may be preferable to the linear model, as it allows for discovering some key nonlinearities in modeling heterogeneous treatment effects.
Discussion
The idea in the Bayesian estimation approach of [29] was to treat the link function g as another unknown and approximate it by a linear combination of B-spline basis functions. In this article, to estimate heterogeneous treatment effect using a flexible link function, we used the adjusted responses and weights associated with the iteratively re-weighted least squares (IWLS) algorithm in the quadratic approximation of the log likelihood, for each MCMC sampler. The approximation under the IWLS framework and the specific prior choice (10) allows us to analytically integrate γ out of the approximated posterior (13), which simplifies the sampling procedure for β. Although the sampling was done using approximated conditional posteriors, this approach appears to work reasonably well.
This paper focused on the context of a randomized clinical trial where the treatment Ai∈{0,1} is randomized independently of pretreatment characteristics Xi. However, the method can be potentially extended to the case where the treatment assignment depends on Xi. To estimate individual treatment effects with observational or non-fully randomized data, we can take a “propensity method” [17, 40, see, e.g.,] upon taking an appropriate reparametrization of the proposed model, which we describe below for a more general context of k treatment conditions, in which the treatment Ai takes a value a∈{1,…,k} with probability (i.e., propensity score) P(Ai=a|Xi)=πa(Xi) (a=1,…,k). Let a=1 be the reference (control) treatment.
For each fixed β, the condition (4) implies E[g(Xi⊤β,Ai)|Xi]=∑a=1kg(Xi⊤β,a)πa(Xi)=0 or equivalently, g(Xi⊤β,a=1)=-∑a=1kg(Xi⊤β,a)πa(Xi)π1(Xi). Given this representation for g(Xi⊤β,a=1), we can reparametrize the canonical parameter of model (3), that is, ηi(Xi,Ai)=m⊤Xi+∑a=1kI(Ai=a)g(Xi⊤β,a), by20 ηi(Xi,Ai)=m⊤Xi+∑a=2kI(Ai=a)g(Xi⊤β,a)-I(Ai=1)∑a=2kg(Xi⊤β,a)πa(Xi)π1(Xi)=m⊤Xi+∑a=2kg(Xi⊤β,a)wa(Ai,Xi),
where wa(Ai,Xi)=I(Ai=a)-πa(Xi)π1(Xi)I(Ai=1). This parametrization is an unconstrained formulation of model (3) without the constraint (4), where the propensity score πa(Xi) (a=1,…,k) is incorporated through the subject i- and treatment a-specific weight wa(Ai,Xi) in the formulation. In model (20), the interaction term g∗(Xi,Ai):=∑a=2kg(Xi⊤β,a)wa(Ai,Xi) still satisfies the condition of the form (4), since E[wa(Ai,Xi)|Xi]=0, indicating that E[g∗(Xi,Ai))|Xi]=0.
In the estimation, as in Sect. 2.2.2 for the binary treatment condition Ai∈{0,1}, we can proceed as follows for the general k treatment conditions with Ai∈{1,…,k} and also in the context of an observational or a non-fully randomized study. We can define the n×L(k-1) design matrix Dβ=(Dβ,2;…;Dβ,k), where each element (the n×L matrix) Dβ,a (a=2,…,k) that is specific to each treatment condition A=a (a=2,…,k) denotes the evaluation matrix of the basis ψ~(·) on {β⊤Xi}i=1n multiplied by the subject i and treatment condition a-specific weight wia=wa(Ai,Xi)(the weight is defined on the second line of (20)), so that its ith row corresponds to the 1×L vector, wiaψ~(β⊤Xi)⊤, with the weight wia incorporating the pre-estimated treatment propensity score πa(X) (a=1,…,k). The spline coefficient vector γ∈RL(k-1) associated with the design matrix Dβ can be introduced, which yields the representation g=Dβγ∈Rn as in Sect. 2.2.2, and the same estimation procedure of Sect. 2.2.3 and 2.3 can be employed to conduct posterior inference.
Future work will extend the model to accommodate multiple treatment outcomes to allow for borrowing of strength between the available outcomes in modeling heterogeneous treatment effects.
Appendix
As MCMC convergence diagnostic for the simulation example in Sect 3, we report the Gelman–Rubin potential scale reduction factor (PSRF) computed for each scenario (Table 2) .Table 2 Summary of the Gelman–Rubin (GR) diagnostic over 100 simulation replications for each scenario. The GR potential scale reduction factor (PSRF) was computed for each element of the components β, γ, and m. For each of these components, we computed the proportion of the individual elements’ PSRF statistics being less than a certain threshold (1.03, 1.05, 1.07) by averaging the PSRF statistics across the corresponding component’s elements and across the 100 simulation replications (for example, for the component β∈Rp, the proportion was computed based on 100∗p PSRF statistics). The results for β indicate that almost all PSRF statistics were less than 1.07, while those of γ and m were consistently smaller than 1.03 for almost all the scenarios, providing some assurance that the sampler has performed reasonably
Shape of the component m Shape of the component g p n Pr(GR < 1.03) Pr(GR < 1.05) Pr(GR < 1.07)
β γ m β γ m β γ m
Linear Nonlinear 5 500 0.856 1.000 1.000 0.970 1.000 1.000 0.998 1.000 1.000
1000 0.908 0.999 1.000 0.986 1.000 1.000 0.998 1.000 1.000
2000 0.916 1.000 1.000 0.984 1.000 1.000 1.000 1.000 1.000
10 500 0.602 0.997 1.000 0.950 1.000 1.000 1.000 1.000 1.000
1000 0.744 1.000 1.000 0.981 1.000 1.000 1.000 1.000 1.000
2000 0.803 1.000 1.000 0.988 1.000 1.000 1.000 1.000 1.000
Linear 5 500 0.844 1.000 1.000 0.984 1.000 1.000 0.998 1.000 1.000
1000 0.902 1.000 1.000 0.984 1.000 1.000 1.000 1.000 1.000
2000 0.930 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
10 500 0.485 0.999 1.000 0.925 1.000 1.000 0.999 1.000 1.000
1000 0.689 0.997 1.000 0.968 0.999 1.000 1.000 1.000 1.000
2000 0.758 1.000 1.000 0.977 1.000 1.000 1.000 1.000 1.000
Nonlinear Nonlinear 5 500 0.950 1.000 1.000 0.998 1.000 1.000 1.000 1.000 1.000
1000 0.970 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
2000 0.956 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
10 500 0.750 0.996 1.000 0.991 1.000 1.000 1.000 1.000 1.000
1000 0.888 1.000 1.000 0.998 1.000 1.000 1.000 1.000 1.000
2000 0.933 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
Linear 5 500 0.922 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
1000 0.978 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
2000 0.968 1.000 1.000 0.996 1.000 1.000 1.000 1.000 1.000
10 500 0.734 0.999 1.000 0.986 1.000 1.000 1.000 1.000 1.000
1000 0.884 1.000 1.000 0.998 1.000 1.000 1.000 1.000 1.000
2000 0.936 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
Acknowledgements
The first author thanks Ian McKeague of Columbia University for his insightful suggestions at the outset of this work. This work was supported by the National Institute of Mental Health (NIH Grant No. 5 R01 MH099003) and the National Center for Advancing Translational Sciences (Grant No. 3 UL1TR001445-06A1S2).
==== Refs
References
1. Murphy SA Optimal dynamic treatment regimes J R Stat Soc B 2003 65 2 331 355 10.1111/1467-9868.00389
2. Robins J Optimal Structural Nested Models for Optimal Sequential Decisions 2004 New York Springer
3. Collins Francis S Varmus Harold A new initative on precision medicine N Engl J Med 2015 372 9 793 795 10.1056/NEJMp1500523 25635347
4. Qian M Murphy SA Performance guarantees for individualized treatment rules Ann Stat 2011 39 2 1180 1210 10.1214/10-AOS864 21666835
5. Lu W Zhang H Zeng D Variable selection for optimal treatment decision Stat Methods Med Res 2011 22 493 504 10.1177/0962280211428383 22116341
6. Tian L Alizadeh A Gentles A Tibshrani R A simple method for estimating interactions between a treatment and a large number of covariates J Am Stat Assoc 2014 109 508 1517 1532 10.1080/01621459.2014.951443 25729117
7. Shi C Song R Lu W Robust learning for optimal treatment decision with np-dimensionality Electron J Stat 2016 10 2894 2921 10.1214/16-EJS1178 28781717
8. Jeng X Lu W Peng H High-dimensional inference for personalized treatment decision Electron J Stat 2018 12 2074 2089 10.1214/18-EJS1439 30416643
9. Zhao Y Zeng D Rush AJ Kosorok MR Estimating individualized treatment rules using outcome weighted learning J Am Stat Assoc 2012 107 1106 1118 10.1080/01621459.2012.695674 23630406
10. Zhao Y Zheng D Laber EB Kosorok MR New statistical learning methods for estimating optimal dynamic treatment regimes J Am Stat Assoc 2015 110 583 598 10.1080/01621459.2014.937488 26236062
11. Song R Kosorok M Zeng D Zhao Y Laber EB Yuan M On sparse representation for optimal individualized treatment selection with penalized outcome weighted learning Stat 2015 4 59 68 10.1002/sta4.78 25883393
12. Laber EB Zhao Y Tree-based methods for individualized treatment regimes Biometrika 2015 102 501 514 10.1093/biomet/asv028 26893526
13. Laber EB Staicu A Functional feature construction for individualized treatment regimes J Am Stat Assoc 2018 113 1219 1227 10.1080/01621459.2017.1321545
14. Zhao Y Laber E Ning Y Saha S Sands B Efficient augmentation and relaxation learning for individualized treatment rules using observational data J Mach Learn Res 2019 20 1 23
15. Liu Y Wang Y Kosorok MR Zhao Y Zeng D Augmented outcome-weighted learning for estimating optimal dynamic treatment regimens Stat Med 2018 37 3776 3788 10.1002/sim.7844 29873099
16. Xie S Tarpey T Petkova E Ogden RT Multiple domain and multiple kernel outcome- weighted learning for estimating individualized treatment regimes J Comput Graph Stat 2022 31 4 1 18 10.1080/10618600.2022.2067552
17. Caron A Baio G Manolopoulou I Estimating individual treatment effects using non-parametric regression models: a review J R Stat Soc A 2022 185 1115 1149 10.1111/rssa.12824
18. Nie X Wager S Quasi-oracle estimation of heterogeneous treatment effects Biometrika 2021 108 299 319 10.1093/biomet/asaa076
19. Alaa AM, Schaar M van der (2017) Bayesian inference of individualized treatment effects using multi-task gaussian processes. 31st Conference on Neural Information Processing Systems (NIPS 2017), pp 3427–3435
20. Shalit U Johansson FD Sontag D Estimating individual treatment effect: generalization bounds and algorithms Proc 34th Int Conf Mach Learn 2017 70 3076 3085
21. Wager Stefan Athey Susan Estimation and inference of heterogeneous treatment effects using random forests J Am Stat Assoc 2018 113 1228 1242 10.1080/01621459.2017.1319839
22. Guo W Zhou X Ma S Estimation of optimal individualized treatment rules using a covariate-specific treatment effect curve with high-dimensional covariates J Am Stat Assoc 2022 116 533 309 321 10.1080/01621459.2020.1865167
23. Richard Hahn P Murray Jared S Carvalho Carlos M Bayesian regression tree models for causal inference: Regularization, confounding, and heterogeneous effects Bayesian Analy 2020 15 3 965 1056
24. Klausch T, Ven Peter van de, Brug Tim van de, van de Wiel Mark A, Berkhof Johannes (2018) Estimating Bayesian optimal treatment regimes for dichotomous outcomes using observational data.
25. Murray T Yuan Y Thall PF A Bayesian machine learning approach for optimizing dynamic treatment regimes J Am Stat Assoc 2018 113 523 1255 1267 10.1080/01621459.2017.1340887 30739965
26. Logan BR Sparapani R McCulloch RE Laud PW Decision making and uncertainty quantification for individualized treatments using Bayesian additive regression trees Stat Methods Med Res 2019 28 4 1079 1093 10.1177/0962280217746191 29254443
27. David Brillinger R (1982) A generalized linear model with "Gaussian” regressor variables In A Festschrift for Erich L. Lehman (Edited by P. J. Bickel, K. A. Doksum and J. L. Hodges). Wadsworth, New York
28. Stoker TM Consistent estimation of scaled coefficients Econometrica 1986 54 1461 1481 10.2307/1914309
29. Antoniadis A Gregoire G McKeague I Bayesian estimation in single-index models Stat Sin 2004 14 1147 1164
30. Poon W Wang H Bayesian analysis of generalized partially linear single-index models Comput Stat Data Anal 2013 68 251 261 10.1016/j.csda.2013.07.018
31. Poon W Wang H Multivariate partially linear single-index models: Bayesian analysis J Nonparametric Stat 2014 26 4 755 768 10.1080/10485252.2014.965706
32. Wang H Bayesian estimation and variable selection for single index models Comput Stat Data Anal 2009 53 2617 2627 10.1016/j.csda.2008.12.010
33. Wang H A Bayesian multivariate partially linear single-index probit model for ordinal responses J Stat Comput Simul 2018 88 1616 1636 10.1080/00949655.2018.1442469
34. Choi T Shi J Wang B A gaussian process regression approach to a single-index model J Nonparametric Stat 2011 23 21 36 10.1080/10485251003768019
35. Gramacy RB Lian H Gaussian process single-index models as emulators for computer experiments Technometrics 2012 54 1 30 41 10.1080/00401706.2012.650527
36. Yu Y Zou Z Wang S Meyer R Bayesian nonparametric modelling of the link function in the single-index model using a Bernstein-Dirichlet process prior J Stat Comput Simul 2019 89 3290 3312 10.1080/00949655.2019.1663191
37. Dhara Kumaresh Lipsitz Stuart Pati Debdeep Sinha Debajyoti A new Bayesian single index model with or without covariates missing at random Bayesian Anal 2020 15 3 759 780 10.1214/19-BA1170 33692872
38. Liu C, Liang H (2022) Bayesian analysis in single-index quantile regression with missing observation. Communications in Statistics—Theory and Methods
39. Troxel Andrea B, Petkova Eva , Goldfeld Keith, Liu Mengling, Tarpey Thaddeus, Wu Yinxiang, Wu Danni, Agarwal Anup, Avendaño-Solá Cristina, Bainbridge Emma, Bar Katherine J, Devos Timothy, Duarte Rafael F, Gharbharan Arvind, Hsue Priscilla Y, Kumar Gunjan, Luetkemeyer Annie F, Meyfroidt Geert, Nicola André M, Mukherjee Aparna, Ortigoza Mila B, Pirofski Liise-anne, Rijnders Bart J A, Rokx Casper , Sancho-Lopez Arantxa, Shaw Pamela, Tebas Pablo, Yoon Hyun Ah, Grudzen Corita, Hochman Judith, Antman Elliott M. Association of convalescent plasma treatment with clinical status in patients hospitalized with COVID-19: a meta-analysis. JAMA Netw Open, 5(1): e2147331–e2147331, 2022. 10.1001/jamanetworkopen.2021.47331.
40. Imbens GW Rubin DB Causal inference in statistics, social, and biomedical sciences 2015 Cambridge University Press
41. Rubin DB Causal inference using potential outcomes: Design, modeling, decisions J Am Stat Assoc 2005 100 469 322 331 10.1198/016214504000001880
42. Eilers Paul Marx Brian Flexible smoothing with B-splines and penalties Stat Sci 1996 11 2 89 121 10.1214/ss/1038425655
43. Wang Li Yang Lijian Spline estimation of single-index models Stat Sin 2009 19 765 783
44. Wand MP Ormerod JT Penalized wavelets: embedding wavelets into semiparametric regression Electron J Stat 2011 5 1654 1717 10.1214/11-EJS652
45. Lenk PJ Bayesian inference for semiparametric regression using a Fourier representation J R Stat Soc B 2002 61 863 879 10.1111/1467-9868.00207
46. Hornik K, Grun B (2022) movMF: mixtures of Von Mises-Fisher distributions. R package version 0.2-7
47. Gelman A Rubin DB Inference from iterative simulation using multiple sequences Stat Sci 1992 7 4 457 472 10.1214/ss/1177011136
48. Vats Dootika Knudson Christina Revisiting the Gelman-Rubin diagnostic Stat Sci 2021 36 4 518 529 10.1214/20-STS812
49. Park Hyung, Tarpey Thaddeus, Liu Mengling, Goldfeld Keith, Wu Yinxiang, Wu Danni, Li Yi, Zhang Jinchun, Ganguly Dipyaman, Ray Yogiraj, Paul Shekhar Ranjan, Bhattacharya Prasun, Belov Artur, Huang Yin, Villa Carlos, Forshee Richard, Verdun Nicole C, Yoon Hyun ah, Agarwal Anup, Simonovich Ventura Alejandro, Scibona Paula, Burgos Pratx Leandro, Belloso Waldo, Avendaño-Solá Cristina, Bar Katharine J, Duarte Rafael F, Hsue Priscilla Y, Luetkemeyer Anne F, Meyfroidt Geert, Nicola André M, Mukherjee Aparna, Ortigoza Mila B, Pirofski Liise-anne, Rijnders Bart J A, Troxel Andrea, Antman Elliott M, Petkova Eva. Development and Validation of a Treatment Benefit Index to Identify Hospitalized Patients With COVID-19 Who May Benefit From Convalescent Plasma. JAMA Network Open, 5(1): e2147375–e2147375, 2022. ISSN 2574-3805. 10.1001/jamanetworkopen.2021.47375
50. Murphy SA A generalization error for Q-learning J Mach Learn 2005 6 1073 1097
|
PMC010xxxxxx/PMC10197436.txt |
==== Front
Knowl Based Syst
Knowl Based Syst
Knowledge-Based Systems
0950-7051
1872-7409
The Author(s). Published by Elsevier B.V.
S0950-7051(23)00392-1
10.1016/j.knosys.2023.110642
110642
Article
Towards COVID-19 fake news detection using transformer-based models
Alghamdi Jawaher ab⁎
Lin Yuqing a
Luo Suhuai a
a School of Information and Physical Sciences, The University of Newcastle, Newcastle, Australia
b Department of Computer Science, King Khalid University, Abha, Saudi Arabia
⁎ Corresponding author at: School of Information and Physical Sciences, The University of Newcastle, Newcastle, Australia.
19 5 2023
15 8 2023
19 5 2023
274 110642110642
10 11 2022
17 4 2023
13 5 2023
© 2023 The Author(s)
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The COVID-19 pandemic has resulted in a surge of fake news, creating public health risks. However, developing an effective way to detect such news is challenging, especially when published news involves mixing true and false information. Detecting COVID-19 fake news has become a critical task in the field of natural language processing (NLP). This paper explores the effectiveness of several machine learning algorithms and fine-tuning pre-trained transformer-based models, including Bidirectional Encoder Representations from Transformers (BERT) and COVID-Twitter-BERT (CT-BERT), for COVID-19 fake news detection. We evaluate the performance of different downstream neural network structures, such as CNN and BiGRU layers, added on top of BERT and CT-BERT with frozen or unfrozen parameters. Our experiments on a real-world COVID-19 fake news dataset demonstrate that incorporating BiGRU on top of the CT-BERT model achieves outstanding performance, with a state-of-the-art F1 score of 98%. These results have significant implications for mitigating the spread of COVID-19 misinformation and highlight the potential of advanced machine learning models for fake news detection.
Keywords
COVID-19
Fake news
Misinformation
Pre-trained transformer models
Social media
==== Body
pmc1 Introduction
COVID-19 was declared a Public Health Emergency of International Concern by the World Health Organization (WHO) on 30 January 2020 [1]. Due to the physical restrictions imposed by governments to reduce the impact of COVID-19, people tend to rely more on social media as the primary source of communication. A recent study found that average user activity on social media has increased by 25% due to the global lockdown [2]. Increasing concerns related to COVID-19 have prompted people to seek and share information about the pandemic on social media [3]. However, the dark side of the coin is fake tweets dissemination that spreads fear and panic about such a pandemic. The dissemination of such falsified tweets has led to various adverse consequences, including vaccination hesitancy [4], changes in health behaviour intentions [5], also some falsified information propagated about chloroquine’s effectiveness has led to an increase of cases of chloroquine drug overdose [6]. In less than two months, the International Fact-Checking Network (IFCN) at the Poynter Institute found over 3500 false claims related to COVID-19 [7].
The coronavirus-related misinformation may have led to more than 800 deaths worldwide in the first three months of 2020.1 Social media platforms differ from traditional news outlets in that they have a tendency to spread false information based on some characteristics more than the latter. Several recent WHO reports describe the spread of misinformation related to COVID-19 as an Infodemic which is defined as “an overabundance of information, both online and offline. It includes deliberate attempts to disseminate wrong information to undermine the public health response and advance alternative agendas of groups or individuals”.2
The massive amount of user-generated content becomes prohibitively cumbersome to manually process due to the large volume of data. Therefore, it is imperative to develop automated fake news detection systems that detect fake content effectively.
However, detecting fake news on social media becomes even more challenging when considering the poor quality of user-generated content, complex semantics of natural language and the high dimensionality of textual data, especially when malicious entities can frequently manipulate and change their writing style to mimic trustworthy content [8].
As a key element of the detection approaches, researchers have proposed different ways to interpret the meaning of a word by representing it as an embedding vector. For learning word embeddings from large word corpora, neural network-based methods (e.g., Word2Vec [9]) and count-based models (e.g., GloVe [10]) are commonly used. The downside of these embedding models is that they are context-free, i.e., context is neglected, and a static embedding for the words is generated regardless of their contexts. Thus, a model that can capture deep semantic and contextualised word embeddings is required for more fine-grained detection performance. Lately, the concept of attention has received more and more attention, and the natural language processing (NLP) community is starting to approach a paradigm shift, developing a set of models that not only improve accuracy but also address the problem of lacking labelled data, which has been a well-known problem in the NLP research.
Automatic detection of fake news is a non-trivial task, given that existing [deep] machine learning models (prior to the advent of almost ubiquitous transformer models) are impotent towards providing a deeper semantic understanding of input text. To respond to this, the NLP research has made great strides by introducing the transformer architecture. In addition, pre-trained language models (PLMs) that have been trained on massive unlabelled corpora are a current trend for text classification tasks. Such models have made a great breakthrough in many NLP tasks where the PLMs can be easily fine-tuned on many different NLP tasks. The main idea is to extract the pre-trained neural network layers from the language model (LM) and stack new neural network layers on top of them to adapt for the downstream task [11].
While deep learning (DL) approaches allow for capturing more salient and relevant information, transformer-based approaches have the power to encode deeper semantic and contextualised information about a given input text. To take advantage of the former and the latter, we propose different architectures using different neural-based structures on top of the PLMs. This paper explores the potential of various PLMs such as CT-BERT [12], RoBERTa [13], and BERT-based classifiers for detecting COVID-19 fake news. We introduce different downstream neural network structures on top of BERT [14] and CT-BERT (fine-tuning strategies) to examine how effective different downstream neural network structures, added on top of BERT and CT-BERT architectures, are in improving COVID-19 fake news detection.
Our experiments have demonstrated (see Section 3.5.3) that different pre-trained transformer-based models perform differently under different strategies. For example, using the pre-trained CT-BERT with a Bidirectional Gated Recurrent Units (BiGRU) layer on top of it has shown to work the best among all other proposed strategies and official baseline models. Generally speaking, extending the PLMs using more expressive powered architectures such as CNN and BiGRU shows promising results; of course, the potential issue of these architectures is the excessive number of parameters used, which slows down the training process. Also, we have demonstrated that fine-tuning the PLMs produce better results than feature-based approaches by a clear margin.
The main contributions of this work may be resumed as:
1. To obtain the best performance, we have explored novel transferring transformer-based methodologies based on different downstream neural network structures.
2. Exploring the effects of fine-tuning approach on the proposed transferring methodologies to support the stated hypothesis that fine-tuning approach has good potential for further improving the performance.
3. Extensive experiments presented using different classical and advanced machine learning models, and the performance differences between these widely used models and the proposed models are compared and analysed.
This paper is structured as follows. Section 2 reviews previous studies on the topic. Section 3 describes the methodology and covers the analysis of the results. Section 4 provides the discussion and insights. Finally, Section 5 concludes the paper.
2 Related work
This section presents a brief overview of the related work directly relevant to Constraint@AAAI2021 COVID19 [15], with a particular emphasis on transformer-based approaches that utilise this dataset. It also highlights a selection of prior work that uses classical and advanced machine learning-based fake news detection approaches. There are different approaches that have been employed to detect fake news, ranging from metaheuristic-based methods to traditional machine learning (ML) methods and the-state-of-the-art transformer-based models.
2.1 Fake news detection
2.1.1 Classical and advanced machine learning approaches
Metaheuristic-based methods have been utilised as an intriguing solution search strategy for detecting fake news in online social networks and media. For example, a study by [16] proposed a novel model for detecting fake news that uses optimisation methods. However, metaheuristic-based methods need more time and computational resources to explore the solution space compared to traditional ML approaches. Statistical ML methods such as Logistic Regression (LR), Support Vector Machines (SVMs), K-Nearest Neighbour (K-NN), Naive Bayes (NB), Decision Trees (DT), Random Forest (RF), Gradient Boost (GB), and XGBoost (XGB) have traditionally been used in text classification. The goal of the Constraint@AAAI2021 COVID19 fake news detection challenge [17] is to develop a model capable of distinguishing between real and fake news related to COVID-19. As part of an ensemble constructed in [18], the team used bidirectional Long Short-Term Memory (BiLSTM), SVMs, LR, NB, and NB combined with LR, achieving a 0.94 F1 score.
Fake COVID-19-related news was examined by [19], in which the authors obtained data from 150 users by extracting information from their social media accounts, such as Twitter, email, mobile, WhatsApp, and Facebook, for a period span from March 2020 to June 2020. They removed information irrelevant to the COVID-19 data and incomplete news during the pre-processing phase. The classification was performed using K-NN, where the results showed the best prediction scores for June with a 0.91 F1 score and the worst ones for March with a 0.79 F1 score. Compared in [15] are four ML baseline models, namely, DT, LR, GB and SVMs, to detect COVID-19-related fake news with the best performance of 93.46% F1-score with SVMs. In [20], the authors applied classical ML algorithms using several linguistic features, such as n-grams, readability, emotional tone and punctuation. Their experimental results found a linear SVM to be the best performing model with a weighted average F1 score of 95.19% on test data.
In recent years, researchers have increasingly focused on deep neural networks (DNN)-based models, such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and BiLSTM that combine multiple DNN configurations. The technique of learning how to transfer knowledge is a concept in ML known as transfer learning, which stores and applies the knowledge gained from performing a specific task to another problem. Learning this way is useful when it comes to training and evaluating models with relatively small amounts of data. In the area of NLP, transfer learning is achieved by creating a set of pre-trained embedding models trained on a massive amount of text. Several neural-based models have been proposed to model the COVID-19 fake news detection problem using a context-independent pre-trained word embedding layer [15], [18], [20]. As these basic models are context-free, they are impotent towards capturing deep contextualised trends in the input text. Therefore, researchers developed attention-based models that can provide context-aware word embeddings pre-trained on large-scale datasets, which are paramount for the success of most NLP tasks. The main objective of this paper is to investigate the use of advanced ML models and transfer learning in evaluating the credibility of news content related to COVID-19.
2.1.2 Transfer learning
A text mining model is a method for extracting useful information and knowledge from unstructured text [21]. Text mining models have become much more sophisticated with the advancements in DL techniques used in NLP. As of recently, with the advent of transformer-based structures, PLMs have become mainstream for downstream text classification [14]. For example, major advances have been driven by the use of PLMs, such as ELMO [22], GPT [23] or BERT. BERT and RoBERTa, as the most commonly utilised PLMs, were trained on exceptionally large corpora. The success of such approaches raises the question of how to use such models for downstream text classification tasks? Over the PLMs, task-specific layers are added for each downstream task, and then the new model is trained with only those layers from scratch [12], [13], [14] in a supervised manner; see Fig. 1.
More specifically, these models exhibit a two-step learning approach. They learn pre-trained language representations by analysing much text in a self-supervised fashion. This process is commonly called pre-training. Then these pre-trained language representations can be applied to downstream NLP tasks by selecting either of two approaches: feature-based and fine-tuning. The former uses pre-trained representations and includes them as additional features for learning a given task. The latter introduces minimal task-specific parameters, and all pre-trained parameters are fine-tuned on the downstream tasks. The advantage of such transfer learning is that the deep context-aware word representations can be learned from large unannotated text corpora in self-supervised pre-training; this is especially useful when learning a domain-specific language with insufficient labelled data.Fig. 1 PLMs fine-tuning.
Besides the fact that surface-level features cannot effectively capture semantical patterns in text, the lack of a sufficient amount of data constitutes a bottleneck for the advanced ML models. Thus, to address this, we exploit the power of BERT and its variations in building robust fake news predictive models. Relatively little research has been done to detect fake news using the recent pre-trained transformer-based models. The few observational studies that have been done using such models, despite the use of different methodologies and different scenarios, have shown promising results.
One recent example of this is a study conducted by Aggarwal et al. [24] showed that BERT, even with minimal text pre-processing, provided better performance compared to that of LSTM and gradient boosted tree models. Similarly, Jwa et al. [25] achieved a high F1 score of 0.746 for fake news detection using BERT on the FNC dataset by analysing the relationship between the headline and body text of news. Baruah et al. [26] also utilised BERT for the classification task of automatically detecting fake news spreaders, achieving an accuracy of 0.690. However, BERT is computationally expensive due to its millions of parameters (e.g., BERTBASE has 110 million parameters while BERTLARGE has 340 million parameters) [14]. DistilBERT [27], a variation of BERT, reduces its size by 40% while retaining 97% of its language understanding abilities, resulting in faster training (60% faster). Another robust BERT model, RoBERTa [13], was developed using a larger dataset, larger batches, and more iterations.
In [28], the authors applied a pre-trained transformer model, so-called XLNet, combined with Latent Dirichlet Allocation (LDA) by integrating contextualised representations generated from the former with topical distributions produced by the latter. Their model achieved an F1 score of 0.967 on the Constraint@AAAI2021-COVID19 fake news dataset. Using a combination of existing contextual representations, such as BERT, and knowledge graph-based representations, a study by [29] achieved an accuracy of 95.70% and an F1 score of 95.69% on the same dataset. In the same vine, a fine-tuned transformer-based ensemble model has been proposed by [30]. The proposed model achieved an F1 score of 97.9% using the same dataset. The ensemble transformer model differs from the previously mentioned models in that it can combine the advantages of multiple transformer models, leading to an enhanced overall performance. This could explain its achievement of state-of-the-art performance. A compact overview of the ML models proposed by related work is shown in Table 1.
To this end, classical ML algorithms are easy to comprehend and perform well on small datasets, but they (i) require complex feature engineering and (ii) fail to capture substantial semantical contextual knowledge for a specific input text. To overcome this, advanced ML techniques such as CNNs and RNN-based methods are well suited for complicated classification problems, powered by a massive amount of data, and can learn more complicated (latent) features. However, even though CNNs have proven effective in extracting local features, they typically struggle with capturing long-term contextual dependencies. In contrast, RNN-based methods perform sub-optimally in handling such dependencies and are not a good candidate for capturing local features. As such, a combination of these two architectures may be able to overcome some of their inherent limitations. Plus, adding this unified architecture on top of a transformer-based model such as BERT (more specifically, the variant trained on massive COVID-19 tweets) would give the model far more expressive power by allowing it to learn deep and meaningful insights for a given input text. As stated in the introduction and different from these studies mentioned above, we aim to exploit the power of the advanced ML models in combination with the deep contextualised PLMs by testing the effectiveness of different downstream neural network architectures added on top of different transformer-based models for COVID-19 fake news detection. A comparison of the applied approach with the state-of-the-art on COVID-19 dataset is shown in Table 2.Fig. 2 PLMs + downstream neural network structures.
Table 1 A compact overview of ML models applied by the related work.
Ref Model Advantages Disadvantages
[18] Ensemble LR+SVMs+NB+BiLSTM The ensemble method combines the strengths of multiple models and helps to reduce the impact of individual weaknesses and can improve the overall accuracy of the model. Ensemble may be more computationally intensive and require more resources than a single model.
[19] KNN KNN is a simple and intuitive algorithm that is easy to understand and implement. KNN is a lazy learner, which means it requires a significant amount of time to make predictions compared to other algorithms.
[15], [20] SVM SVMs can effectively model complex decision boundaries and have been shown to perform well on high-dimensional datasets. SVMs can be computationally expensive, especially on large datasets.
[29] BERT BERT can handle complex language structures, including context-dependent meanings of words, idiomatic expressions, and long-range dependencies. BERT requires a large amount of computational resources for training and inference, making it difficult to deploy on resource-constrained systems.
[28] XLNet XLNet can capture bidirectional context more effectively than BERT, allowing it to better handle tasks that require understanding of long-term dependencies. XLNet requires significant computational resources and may not be suitable for all applications.
[30] Transformer ensemble Transformer ensemble models can combine the strengths of multiple pre-trained transformer models to achieve better performance on downstream tasks. Ensemble models can be computationally expensive and require large amounts of memory, making them difficult to deploy on resource-constrained devices.
Table 2 A comparison of the best performing applied approach with the state-of-the-art on COVID-19 dataset.
Ref. Model F1 score (%) Features used Adding downstream neural based structures on top of transformers
[15] SVMs 93.46% TF-IDF N/A
[18] Ensemble LR+SVMs+NB+BiLSTM 94% TF-IDF N/A
[20] SVMs 95.19% N-grams, readability, emotional tone and punctuation N/A
[29] BERT 95.69% Context-aware and knowledge graph representations No
[28] XLNet 95.70% Context-aware representations and topical features No
[30] Transformer ensemble 97.9% Context-aware representations No
Ours [5] CT-BERT+BiGRU 98.54% Context-aware representations Yes
3 Methodology
3.1 Problem statement
Suppose we have a set T={t1,t2,..,tL} of L labelled news (i.e., tweets) either fake or real, where L is the total number of news. The task is to learn fake news detection function f(T)→yˆ, such that it maximises prediction accuracy. Here, we cast the problem as a binary classification where a piece of information could be fake (yˆ = 0) or real (yˆ = 1).
We investigate and compare the effectiveness of a set of different PLMs using different (simple and sophisticated) downstream neural network structures. In particular, we extend the pre-trained BERT and CT-BERT LMs by adding various downstream neural network architectures (See Fig. 2). We test the effectiveness of such architectures using both frozen (feature-based approach) and unfrozen weights (fine-tuning approach where all model parameters are jointly trained on a given supervised task).
Using BERT coupled with various advanced ML models, we fine-tune these variants to predict whether the information is fake or not by adding a classification (output) layer on top of each variant with a sigmoid function. The outcome of the activation of the output layer is what will be shown as the model’s prediction. The workflow of the proposed models can be seen in Algorithm 1.
3.2 The proposed models
• BERTBASE [14] : This model, developed by Google AI, has proven to be a powerful tool for text classification [31], [32]. BERT is a multi-layer bidirectional transformer encoder trained on English Wikipedia and Book Corpus containing 2500M and 800M tokens. BERTBASE uses the transformer’s encoder with 12 layers, 12 self-attention heads, and 110 million parameters. In our work, we used the uncased version of BERT, which is considered as a baseline model for BERTBASE model set. Input sequences of a maximum length of 128 are fed into BERT, and based on some analysis, 768-d vector representation is produced for each token. These vectors carry meaningful information about the context of each token. This model uses the corresponding [CLS] token’s representation (a vector of size 768 representing the entire sequence) as input to the output layer. We also experiment with BERTLARGE, which has 24 layers, 16 attention heads, and 340 million parameters.
• BERTBASE+CNN: In this architecture, we extend the BERT model by adding a CNN layer for fine-tuning. First, the representations of the last transformer encoder (sequence output) are used as input to a Conv1D with 128 filters, each with a kernel size of 5, activated with the ReLU function. This is followed by a max-pooling layer to reduce the feature maps. Finally, the resultant feature map is flattened, and the output is passed to the output layer.
• BERTBASE+(Bi)LSTM: Similar to the above architecture, all representations of the latest transformer encoder are used as input to a single (bidirectional) LSTM layer with 128 units, followed by an output layer.
• BERTBASE+(Bi)GRU: The sequence output generated from the latest transformer encoder is used as an input to a (bidirectional) GRU layer with 128 units. The resultant hidden state is then fed into an output layer.
• BERTBASE+CNN-BiLSTM: The sequence output generated from the latest transformer encoder is used as an input to a hybrid model consisting of a single CNN layer followed by a max-pooling layer. The output is then encoded using a BiLSTM layer.
• BERTBASE+CNN-BiGRU: Similar to the previous model, with BiLSTM layer was replaced with a BiGRU one.
• BERTBASE+mCNN: This model is defined with three input channels for processing different n-grams of the input text. Each of these channels consists of three layers: convolution that extracts different word n-gram features and a kernel size set to 4, 6, 8 g to read at once; a max-pooling layer to enable the extraction of the most salient features from each feature map; and a flatten layer to reduce the three-dimensional output into a two-dimensional one. Then the extracted features from the three channels are concatenated into a single vector, and the output is passed to a classification layer.
• BERTBASE+mCNN-BiLSTM: We extend the previous model by passing the resulted single vector of the mCNN model to a BiLSTM layer. Apart from that, the configuration is identical to BERTBASE+mCNN.
• RoBERTa [13] : Stands for the Robustly optimised BERT approach introduced by Facebook. It is simply retraining of BERT with improved training methodology (i) by removing the Next Sentence Prediction task from the pre-training process, (ii) RoBERTa was trained over ten times more data, and (iii) by introducing dynamic masking using larger batch sizes so that the masked token changes during the training rather than static masking pattern used in BERT. Thus, RoBERTa introduces a different pre-training approach to BERT. We experiment with the two variations: RoBERTaBASE and RoBERTaLARGE.
• CT-BERT [12] : Stands for COVIDTwitter-BERT is a transformer-based model, pre-trained on a large corpus of Twitter posts (160M tweets) on the topic of COVID-19 collected from January 12 to April 16, 2020. Thus, the pre-trained CT-BERT has the same domain as the COVID-19 dataset used in this work; thus, we expect CT-BERT to provide better results than the general pre-trained BERT model. We extended the model with the same downstream neural structures used on top of BERT.
3.3 Experimental setup
All the work for the experiments was carried out using Intel Core i5 2.3 GHz, 8 GB RAM system running macOS. We implemented the transformer-based models using Tensorflow Hub—Tensorflow official model repository. Scikit-learn is used to implement classical ML algorithms, while Tensorflow and Keras libraries are used to implement the advanced ML models. This study used the common performance metrics to evaluate the performance of the proposed models, namely, accuracy, precision, recall, and F1 score.
3.4 Dataset
A collection of COVID-19-related social media posts, comments and news, classified as real or fake, based on their truthfulness. The dataset [15] is collected from various social media platforms, such as Twitter and YouTube. The challenge organisers collected 10,700 social media posts and news articles about COVID-19 in the form of an annotated dataset in English. As the dataset has been separated in advance by the task organisers into training, validation, and testing sets, we opted to evaluate our models using the original split.
Fig. 3 depicted the corpus’s word cloud representation of real and fake news, respectively. From the figure, it can be deduced that the most frequent words apart from “covid19” are “people”, “India”, “pandemic”, “vaccine”, “risk”, “hospital”, “government” and so on. The statistics of the dataset are shown in Table 3.Fig. 3 Wordcloud of (a) real news and (b) fake news.
Table 3 The statistics of COVID-19 dataset.
# Candidate news 6420
# True news 3360
# Fake news 3060
3.4.1 Preprocessing
Preprocessing of input text includes tokenising given text using the model’s tokeniser to generate input ids, input masks, and input type ids for further processing. By design, BERT (and its variations) take a sequence of tokens with a maximum length of 512 and produce a representation of the sequence in a 768-dimensional vector for BERTBASE and 1024-dimensional vector for BERTLARGE. Thus, the text must be padded or truncated to ensure that all sequences have similar lengths. In this case, all sequences have been truncated to a length of 128.3
Since user-generated content is often noisy and ambiguous, preprocessing the data is important before feeding it to the models. While removing the emojis/emoticons is common based on the assumption that it reduces the noise in data, this assumption does not always hold. As users tend to use emojis to express their emotions, the emojis may provide deep insights into a text (sentiment words), such as sentiment and emotions; thus, it is considered beneficial to retain them in such a way by converting them into text – using the Python library emoji – in our work.
The models are trained with a batch size of 4 for 3 epochs as we found it to be best for all models based on trial and error and as part of future experiments, we will investigate the selection of hyperparameter values. The sigmoid function is used in the output layer to reduce the error during training, while binary cross-entropy is used to calculate the loss during backpropagation. In our final configuration, we use Adam optimiser with a learning rate of 2×10−5 for BERT and its variations while 1×10−5 for training CT-BERT and RoBERTa models. Again, based on a trial-and-error examination, these values performed the best. Furthermore, these models were trained with minimal preprocessing since transformer-based models come with their own tokenisers and can handle punctuation and lowercase text.
3.5 Results
3.5.1 Classical and advanced ML approaches
We have conducted extensive experiments using classical and advanced ML models on the COVID-19 dataset. The former includes LR, SVMs, NB, RF, and XGB with two types of feature extractor methods, namely, CV and TF-IDF. The latter use GloVe with a 100-d vector to encode input text, which includes CNN, BiLSTM, BiGRU and a hybrid model of CNN with BiLSTM and BiGRU. Table 4 displays the summary of the results obtained. The table shows that the recurrent-based models [BiLSTM and BiGRU] reported significantly higher detection scores than other algorithms. This is followed by CNN and LR classifiers with an F1 score of 0.9463 and 0.9424, respectively. Based on the analysis, statistical methods such as LR and SVMs yield better results than the bagging and boosting techniques such as RF and XGB. Moreover, the findings demonstrate that using CV as a feature extractor is better than TF-IDF for the classical ML methods and fine-tuning the embedding layer during training is better than the static approach for the advanced ML models. We can see how much dynamic embedding layer contributes to the overall performance. Although we applied the same maximum length limit as in transformer-based models (i.e., 128), we can observe the power of the advanced ML models in capturing useful patterns.
It is well known that the size of the data has an impact on the performance of such approaches. Traditional ML algorithms, on the other hand, are less affected by the size of data. As such, advanced ML models typically outperform other approaches when there is a large amount of data. This can be attributed to the outstanding capability of such algorithms in automatically extracting a wide range of informative features. According to [33], supervised DL models will usually perform well at about 5000 examples per class, so such models may not be suitable for scenarios involving few labelled examples. Interestingly, other studies, such as [34], have shown that when one has only 100–1000 labelled examples per class, BERT is a more effective technique for document classification than other traditional ML approaches, owing to the benefits of pre-training and fine-tuning. Transfer learning is a technique that allows us to take advantage of the benefits of DL while using fewer amounts of data.
Table 4 Performance comparison (%) of different classical and advanced ML models on the COVID-19 dataset.
Type Method Acc. (%) Pre. (%) Rec. (%) F1 (%)
Baselines ML (CV) LR 0.9402 0.9348 0.9501 0.9424
SVMs 0.9313 0.9304 0.9379 0.9341
NB 0.9248 0.9420 0.9166 0.9291
RF 0.9178 0.9143 0.9275 0.9209
XGB 0.8874 0.8920 0.8928 0.8924
Baselines ML (TFIDF) LR 0.9276 0.9437 0.9199 0.9317
SVMs 0.9393 0.9473 0.9373 0.9423
NB 0.9037 0.9652 0.8662 0.9130
RF 0.9187 0.9250 0.9201 0.9225
XGB 0.8841 0.8946 0.8852 0.8899
Baselines ML (GloVe, static) CNN 0.8766 0.9100 0.8482 0.8780
BiLSTM 0.9313 0.9140 0.9589 0.9359
BiGRU 0.9294 0.9001 0.9732 0.9352
CNN-BiLSTM 0.8715 0.8132 0.9795 0.8886
CNN-BiGRU 0.9107 0.8894 0.9473 0.9174
Baselines ML (GloVe, dynamic) CNN 0.9449 0.9647 0.9286 0.9463
BiLSTM 0.9514 0.9504 0.9571 0.9537
BiGRU 0.9523 0.9465 0.9634 0.9549
CNN-BiLSTM 0.9393 0.9396 0.9446 0.9421
CNN-BiGRU 0.9332 0.9469 0.9241 0.9354
3.5.2 Transformer-based models
In this section, by comparing our proposed models to other official baselines and other counterpart algorithms, we try to answer the following questions through the experiments:
1. EQ1. Can CT-BERT trained on a corpus of COVID-19 tweets achieve high detection performance compared to the other PLMs trained on a generic language?
2. EQ2. Do the sophisticated downstream neural network structures, compared to the superficial output dense layer, improve the model’s performance?
3. EQ3. Does the fine-tuning approach improve the model’s performance?
3.5.3 Fake news detection performance
Since CT-BERT has the same domain as the used COVID-19 dataset, we expect it to provide better results than other general PLMs. To answer EQ1, we first compare the proposed architectures with the official baselines. The effectiveness of different downstream neural network approaches is evaluated. The results are depicted in Table 5, with the best performance scores highlighted in bold. Based on the results, we have the following observations:
• The experimental results show that CT-BERT+BiGRU outperforms all models, which is not so surprising, given that CT-BERT, again, was pre-trained on a large corpus of Twitter posts on the topic of COVID-19. This seems to confirm our hypothesis that a model based on CT-BERT would perform better than other transformer-based approaches. However, a natural question arises: is this finding simply relate to the task of pre-training? Or there are other factors at play? Answering this question will form the basis of our future work.
• Moreover, GRU, as an improved version of LSTM, with bidirectionality, is capable of capturing informative features, leading to better detection performance. As such, it can be clearly seen from Table 5 that CT-BERT coupled with BiGRU achieved the state-of-the-art results, outperforming the official baselines and other algorithms.
• Additionally, to answer the question EQ2, we found that complex downstream neural structures added on top of PLMs perform better than simply adding a single output dense layer, i.e., CT-BERT + BiGRU > CT-BERT. This also applies to the BERT model where BERTBASE + BiGRU > BERTBASE. This indicates that extending PLMs with more complex structures has the potential to capture more deep contextual information. This, of course, complicates the model conceptually and computationally, yet adding downstream neural structures is shown to be effective.
• Moreover, it seems that the base versions have better accuracy than their larger counterparts, which indicates that adding more model parameters increases the risk of performance degradation. However, this assumption does not always hold since (some) models with a smaller number of parameter values may produce lower accuracy compared to their larger counterparts. For example, we found recurrent neural-based models to outperform CNN-based models, see Table 4. Research [22] found that RNN is able to learn contextual representations in such a way by inspecting representations at different hidden layers.
• Furthermore, the results showed that BiGRU holds potential in learning CT-BERT features better than other algorithms, despite the little difference in their resulted F1 scores.
• Comparing LSTM and GRU, at this level of analysis, it seems there is no clear evidence that one consistently overperforms the other. However, GRU is shown to perform well and be more efficient given a small amount of training data. Furthermore, using bidirectionality shows outstanding performance since it captures information both from left and right, leading to learning more useful contextualised representations.
• Comparing the classical and advanced ML approaches with transformer-based methods, the latter outperforms the former with a clear margin. Indeed, in the case of Twitter (user-generated content often contains misspellings, noise, and abbreviations), one of the main advantages of BERT (and its variations) is the use of sub-tokens instead of a fixed per-word token; it is, thus, ideal for use with such a data [35] compared to the off-the-shelf context-independent word embeddings.
Table 5 Performance comparison (%) of different PLMs with different downstream neural network architectures.
Models Metrics
Acc. (%) Pre. (%) Rec. (%) F1 (%)
SVMs+LR+NB+biLSTM [18] N/A N/A N/A 0.94
SVMs [20] 0.9570 0.9571 0.9570 0.9570
SNN(LM+KG) [29] 0.9570 0.9533 0.9652 0.9569
XLNet+LDA [28] N/A 0.968 0.967 0.967
SVMs [15] 0.9332 0.9333 0.9332 0.9332
Ensemble transformers [30] 0.9799 0.9799 0.9799 0.9799
BERTBASE 0.9617 0.9474 0.9812 0.9640
BERTLARGE 0.9673 0.9589 0.9795 0.9691
BERTBASE+CNN 0.9743 0.9776 0.9732 0.9754
BERTBASE+LSTM 0.9710 0.9673 0.9777 0.9725
BERTBASE+BiLSTM 0.9780 0.9752 0.9830 0.9791
BERTBASE+CNN-BiLSTM 0.9771 0.9735 0.9830 0.9782
BERTBASE+GRU 0.9654 0.9548 0.9804 0.9674
BERTBASE+BiGRU 0.9808 0.9737 0.9902 0.9819
BERTBASE+CNN-BiGRU 0.9664 0.9448 0.9938 0.9687
BERTBASE+mCNN 0.9678 0.9574 0.9821 0.9696
BERTBASE+mCNN-BiLSTM 0.9729 0.9601 0.9893 0.9745
DistilBERT 0.9617 0.9617 0.9652 0.9635
CT-BERT 0.9757 0.9692 0.9848 0.9770
CT-BERT+CNN 0.9822 0.9830 0.9830 0.9830
CT-BERT+LSTM 0.9762 0.9785 0.9759 0.9772
CT-BERT+GRU 0.9762 0.9652 0.9902 0.9775
CT-BERT+BiLSTM 0.9724 0.9538 0.9955 0.9742
CT-BERT+BiGRU 0.9846 0.9797 0.9911 0.9854
CT-BERT+CNN-BiLSTM 0.9645 0.9394 0.9964 0.9671
CT-BERT+CNN-BiGRU 0.9804 0.9695 0.9938 0.9815
CT-BERT+mCNN 0.9799 0.9795 0.9821 0.9808
CT-BERT+mCNN-BiLSTM 0.9799 0.9804 0.9812 0.9808
RoBERTaBASE 0.9668 0.9541 0.9839 0.9688
RoBERTaLARGE 0.9565 0.9581 0.9589 0.9585
3.5.4 Fine-tuning or not
To answer EQ3, here, we study how fine-tuning impacts performance; we experimented with two approaches (i) we evaluate the role of the PLMs when they are frozen during the training phase (i.e., feature extractor) and (ii) we assess the impact of fine-tuning (unfreezing all parameters). As a feature extractor, all the layers of the pre-trained transformer-based model are frozen during the fine-tuning phase, and extra downstream neural layers, including an output sigmoid layer, can be added on top of the model and trained from scratch. That is, only the weights of the added layers will be updated during model training.
In the second approach, the entire pre-trained model has trained on our domain-specific dataset, and the output is fed into downstream neural layers, followed by a sigmoid layer. The error is then backpropagated throughout the architecture, and the model’s pre-trained weights are updated based on the new dataset. In other words, the fine-tuning phase begins with the model parameters obtained from the pre-training, and all of the parameters are fine-tuned during that phase. To prove the robustness of the fine-tuning approach, we assess the impact of the two approaches, i.e. fine-tuning or not, using (some of) the proposed methodologies. Table 6 illustrates the comparative results between these two approaches. The experimental results clearly show the effectiveness of the fine-tuning approach. We further assess the influence of using downstream neural structures with and without fine-tuning. The comparative analysis is shown in Fig. 4, Fig. 5. With the feature extractor approach, it is found that extending the pre-trained transformer-based models with sophisticated neural network structures provides better accuracy than extending such models with a simple output layer. To illustrate, as can be seen in Fig. 4, Fig. 5, training the CNN-BiGRU model with CT-BERT as a feature extractor model yields (almost) equivalent results with slightly small differences in accuracy when compared to the fine-tuned CT-BERT+BiGRU model. This further amplifies the previous observation that downstream neural network structures are able to extract useful information effectively. In addition, fine-tuning the CT-BERT+output layer provides better performance than it does with the non-fine-tuning approach. Thus, we suspect fine-tuning is more critical as fine-tuned versions of DL outperform the frozen ones by a considerable margin.
Fig. 4 Comparison (%) performance using Accuracy curves for models.
Fig. 5 Comparison (%) performance using Loss curves for models.
Table 6 The results of the three best performing models with and w/o fine-tuning.
Models Metrics
Acc. Prec. Rec. F1
CT-BERTa 0.6364 0.6311 0.7348 0.6790
CT-BERT 0.9757 0.9692 0.9848 0.9770
BERTBASE+BiLSTMa 0.9332 0.9208 0.9545 0.9373
BERTBASE+BiLSTM 0.9780 0.9752 0.9830 0.9791
BERTBASE+BiGRUa 0.9285 0.9290 0.9348 0.9319
BERTBASE+BiGRU 0.9808 0.9737 0.9902 0.9819
CT-BERT+BiGRUa 0.9435 0.9222 0.9741 0.9475
CT-BERT+BiGRU 0.9846 0.9797 0.9911 0.9854
a Indicates models with frozen weights.
4 Discussion and insights
The study investigated the effectiveness of various downstream neural network approaches using transformer-based models for COVID-19 detection. While previous research has shown the effectiveness of transformer-based models in detecting COVID-19-related news, limited research has explored how to combine these models with different downstream neural structures for COVID-19 detection. Thus, this study extends the existing research on transformer-based models by examining their effectiveness in the context of COVID-19 detection with different downstream neural structures. The study’s findings are consistent with previous research on the effectiveness of transformer-based models, such as the ensemble transformer model, for COVID-19 detection. The CT-BERT+BiGRU model outperformed all other models, demonstrating the effectiveness of combining advanced ML models with PLMs in capturing context and generating informative representations for downstream tasks. However, the study’s findings contrast with some previous research that suggested larger models with more parameters perform better. The study found that the base versions of the models had better accuracy than their larger counterparts, suggesting that adding more model parameters may lead to performance degradation. In addition, the study found that recurrent neural-based models outperformed CNN-based models, which contradicts some previous research that suggested CNN-based models are more effective in natural language processing tasks.
We have the following observations.
• Our hypothesis is valid. This is observed from the outstanding performance of CT-BERT+BiGRU compared to the other baselines.
• Adding different downstream neural structures on top of transformer-based models is effective. This is observed through the superior performance of e.g., CT-BERT+BiGRU, compared to the original CT-BERT model.
• Fine-tuning approach is effective. This is observed from the promising results generated by fine-tuning the proposed models. This shows how effective the fine-tuning approach compared to the off-the-shelf approaches.
Although we have very interesting results in terms of recall, the precision of the model shows the portion of false detection we have. To better understand this phenomenon, we analyse the errors of the best-performing model. We investigate the confusion matrix resulting from the CT-BERT+BiGRU model shown in Fig. 6. It is evident that the model can separate fake from real content properly. Only ten samples belonging to the real class are misclassified as fake, and 23 of the fake samples are misclassified as real. Thus, almost 0.47% of real samples are misclassified as fake, while 1.07% of fake samples are classified as real. We provided some examples of misclassified samples in Table 7. We observed that the model is confused when classifying claims involving words like “vaccine”.Table 7 A sample of misclassified classes obtained by CT-BERT+BiGRU model.
Tweet Actual Pred
A common question: are coronavirus cases going up because we are testing so many more people?
A: Certainly not in Florida where testing slowed down 3%
while new cases grew 88% over the last week. Real Fake
FDA official says if a COVID vaccine is approved before
it is ready – he’s outta there. Real Fake
Oxford coronavirus vaccine is safe and induces strong
immune response early trial results suggest. Fake Real
Fig. 6 Confusion Matrix of CT-BERT + BiGRU.
4.1 Limitations
The results presented in this study provide valuable insights into the effectiveness of various downstream neural network approaches for COVID-19 detection using transformer-based models. Nonetheless, future work could address some potential limitations. For example, the study did not explore the impact of different hyperparameters and optimisation techniques on model performance. Moreover, the study utilised a relatively small COVID-19 dataset, potentially limiting the generalisability of the findings. Further research could investigate the impact of larger datasets and the transferability of the proposed models to related tasks. In addition, examining the interpretability of the models could provide valuable insights into their underlying mechanisms and improve transparency and trustworthiness. Finally, the study found that CT-BERT+BiGRU outperforms other models, which is an interesting finding. However, it remains unclear whether this result is solely due to the pre-training task or if other factors are at play. Therefore, it is necessary to conduct further analysis to identify the specific factors that contribute to the proposed model’s superior performance. Future research could investigate the impact of different pre-training tasks and datasets to determine the robustness of the proposed models.
5 Conclusions and future research
Our study explores the effectiveness of transformer-based models for COVID-19 fake news detection, presenting novel and effective approaches. Our findings indicate that transformer- based models outperform both traditional and advanced ML baselines for detecting COVID-19 fake news. Fine-tuned CT-BERT with BiGRU achieved state-of-the-art performance with an F1 score of 98.5%, highlighting the importance of fine-tuning and the potential of incorporating complex downstream neural structures. However, to fully understand the limitations of transfer learning and the behaviour of the detectors on different datasets with varying domains, further research is needed. This includes investigating the impact of hyperparameters and optimisation techniques, evaluating the robustness of the models against biased data, and improving interpretability. Overall, our study demonstrates the potential of PLMs for COVID-19-related fake news detection and has important implications for the development of more accurate fake news detection models.
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.
Data availability
I have shared the link to the data.
1 https://www.bbc.com/news/world-53755067.
2 https://www.who.int/news/item/23-09-2020-managing-the-covid-19-infodemic-promoting-healthy-behaviours-and-mitigating-the-harm-from-misinformation-and-disinformation.
3 We found that simple truncation worked well where we consider only the first 128 tokens while ignoring the rest.
==== Refs
References
1 Organization W.H. Statement on the second meeting of the international health regulations (2005) emergency committee regarding the outbreak of novel coronavirus (2019-nCoV) 2022 https://www.who.int/news-room/detail/30-01-2020-statement-on-the-second-meeting-of-the-international-health-regulations-(2005)-emergency-committee-regarding-the-outbreak-of-novel-coronavirus-(2019-ncov), (Accessed June 20, 2022)
2 Yang K.-C. Torres-Lugo C. Menczer F. Prevalence of low-credibility information on twitter during the covid-19 outbreak 2020 arXiv preprint arXiv:2004.14484
3 Huynh T.L. The COVID-19 risk perception: A survey on socioeconomics and media attention Econ. Bull. 40 1 2020 758 764
4 Islam M.S. Kamal A.-H.M. Kabir A. Southern D.L. Khan S.H. Hasan S.M.M. Sarkar T. Sharmin S. Das S. Roy T. Harun M.G.D. Chughtai A.A. Homaira N. Seale H. COVID-19 vaccine rumors and conspiracy theories: The need for cognitive inoculation against misinformation to improve vaccine adherence PLOS ONE 16 5 2021 1 17 10.1371/journal.pone.0251605
5 Greene C.M. Murphy G. Quantifying the effects of fake news on behavior: Evidence from a study of COVID-19 misinformation J. Exp. Psychol.: Appl. 2021
6 Busari S. Adebayo B. Nigeria records chloroquine poisoning after trump endorses it for coronavirus treatment 2020
7 Poynter S. Fighting the infodemic: The coronavirusfacts alliance 2020 https://www.poynter.org/coronavirusfactsalliance/
8 Alghamdi J. Lin Y. Luo S. Modeling fake news detection using BERT-CNN-BiLSTM architecture 2022 IEEE 5th International Conference on Multimedia Information Processing and Retrieval (MIPR) 2022 354 357 10.1109/MIPR54900.2022.00069
9 T. Mikolov, W.-t. Yih, G. Zweig, Linguistic regularities in continuous space word representations, in: Proceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, 2013, pp. 746–751.
10 J. Pennington, R. Socher, C.D. Manning, Glove: Global vectors for word representation, in: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2014, pp. 1532–1543.
11 Gururangan S. Marasović A. Swayamdipta S. Lo K. Beltagy I. Downey D. Smith N.A. Don’t stop pretraining: Adapt language models to domains and tasks Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics 2020 Association for Computational Linguistics 8342 8360 10.18653/v1/2020.acl-main.740 Online, https://aclanthology.org/2020.acl-main.740
12 Müller M. Salathé M. Kummervold P.E. Covid-twitter-bert: A natural language processing model to analyse covid-19 content on twitter 2020 arXiv preprint arXiv:2005.07503
13 Liu Y. Ott M. Goyal N. Du J. Joshi M. Chen D. Levy O. Lewis M. Zettlemoyer L. Stoyanov V. Roberta: A robustly optimized BERT pretraining approach 2019 ArXiv abs/1907.11692
14 Devlin J. Chang M.-W. Lee K. Toutanova K. BERT: Pre-training of deep bidirectional transformers for language understanding 2019 ArXiv abs/1810.04805
15 Patwa P. Sharma S. Pykl S. Guptha V. Kumari G. Akhtar M.S. Ekbal A. Das A. Chakraborty T. Fighting an infodemic: Covid-19 fake news dataset International Workshop on Combating Online Hostile Posts in Regional Languages During Emergency Situation 2021 Springer 21 29
16 Ozbay F.A. Alatas B. Adaptive salp swarm optimization algorithms with inertia weights for novel fake news detection model in online social media Multimedia Tools Appl. 80 26–27 2021 34333 34357 10.1007/s11042-021-11006-8
17 Patwa P. Bhardwaj M. Guptha V. Kumari G. Sharma S. Pykl S. Das A. Ekbal A. Akhtar M.S. Chakraborty T. Overview of constraint 2021 shared tasks: Detecting english covid-19 fake news and hindi hostile posts International Workshop on Combating Online Hostile Posts in Regional Languages During Emergency Situation 2021 Springer 42 53
18 Shushkevich E. Cardiff J. Tudublin team at constraint@aaai2021 - COVID19 fake news detection 2021 ArXiv abs/2101.05701
19 Bandyopadhyay S. Dutta S. The analysis of fake news in social medias for four months during lockdown in COVID-19-a study: biostatistical analysis of COVID-19 Xeno J. Biomed. Sci. 1 1 2020 1 6
20 Felber T. Constraint 2021: Machine learning models for COVID-19 fake news detection shared task 2021 arXiv preprint arXiv:2101.03717
21 Hotho A. Nürnberger A. Paaß G. A brief survey of text mining Ldv Forum, Vol. 20 2005 Citeseer 19 62
22 Peters M.E. Neumann M. Iyyer M. Gardner M. Clark C. Lee K. Zettlemoyer L. Deep contextualized word representations 2018
23 Radford A. Narasimhan K. Salimans T. Sutskever I. Improving language understanding by generative pre-training 2018
24 Aggarwal A. Chauhan A. Kumar D. Mittal M. Verma S. Classification of fake news by fine-tuning deep bidirectional transformers based language model EAI Endorsed Trans. Scalable Inf. Syst. 7 27 2020
25 Jwa H. Oh D. Park K. Kang J.M. Lim H. exbake: Automatic fake news detection model based on bidirectional encoder representations from transformers (bert) Appl. Sci. 9 19 2019 4062
26 Baruah A. Das K.A. Barbhuiya F.A. Dey K. Automatic detection of fake news spreaders using BERT CLEF (Working Notes) 2020
27 Sanh V. Debut L. Chaumond J. Wolf T. Distilbert, a distilled version of BERT: smaller, faster, cheaper and lighter 2019 ArXiv abs/1910.01108
28 Gautam A. V V. Masud S. Fake news detection system using xlnet model with topic distributions: Constraint@aaai2021 shared task 2021 10.48550/ARXIV.2101.11425 https://arxiv.org/abs/2101.11425
29 Koloski B. Perdih T.S. Robnik-Šikonja M. Pollak S. Škrlj B. Knowledge graph informed fake news classification via heterogeneous representation ensembles Neurocomputing 2022
30 Shifath S.M.S.-U.-R. Khan M.F. Islam M.S. A transformer based approach for fighting COVID-19 fake news 2021 10.48550/ARXIV.2101.12027 https://arxiv.org/abs/2101.12027
31 Gao Z. Feng A. Song X. Wu X. Target-dependent sentiment classification with BERT Ieee Access 7 2019 154290 154299
32 A. Nikolov, V. Radivchev, Nikolov-radivchev at semeval-2019 task 6: Offensive tweet classification with bert and ensembles, in: Proceedings of the 13th International Workshop on Semantic Evaluation, 2019, pp. 691–695.
33 Goodfellow I. Bengio Y. Courville A. Deep Learning 2016 MIT Press
34 Usherwood P. Smit S. Low-shot classification: A comparison of classical and deep transfer machine learning approaches 2019 ArXiv abs/1907.07543
35 Horne L. Matti M. Pourjafar P. Wang Z. GRUBERT: A GRU-based method to fuse BERT hidden layers for Twitter sentiment analysis Proceedings of the 1st Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 10th International Joint Conference on Natural Language Processing: Student Research Workshop 2020 Association for Computational Linguistics Suzhou, China 130 138 https://aclanthology.org/2020.aacl-srw.19
|
PMC010xxxxxx/PMC10198026.txt |
==== Front
Curr Nutr Rep
Curr Nutr Rep
Current Nutrition Reports
2161-3311
Springer US New York
37204636
472
10.1007/s13668-023-00472-1
Review
Dairy Milk Protein–Derived Bioactive Peptides: Avengers Against Metabolic Syndrome
Koirala Pankaj 1
Dahal Merina 2
Rai Sampurna 1
Dhakal Milan 1
http://orcid.org/0000-0003-2732-4534
Nirmal Nilesh Prakash nilesh.nir@mahidol.ac.th
1
Maqsood Sajid 3
Al-Asmari Fahad 4
Buranasompob Athisaya 1
1 grid.10223.32 0000 0004 1937 0490 Institute of Nutrition, Mahidol University, 999 Phutthamonthon 4 Road, Salaya, Nakhon Pathom, 73170 Thailand
2 grid.80817.36 0000 0001 2114 6728 Department of Nutrition and Dietetics, Central Campus of Technology, Tribhuvan University, Kirtipur, Nepal
3 grid.43519.3a 0000 0001 2193 6666 Department of Food Science, College of Food and Agriculture, United Arab Emirates University, Al Ain, 15551 United Arab Emirates
4 grid.412140.2 0000 0004 1755 9687 Department of Food Science and Nutrition, College of Agriculture and Food Sciences, King Faisal University, Al-Hofuf, P. O. Box 400, Al-Ahsa, 31982 Saudi Arabia
19 5 2023
2023
12 2 308326
10 4 2023
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Purpose of Review
Metabolic syndrome is continuously increasing among the world’s populations. Metabolic syndrome is a medical condition in which individuals suffer from high blood pressure, high blood glucose levels, and obesity. The in vitro and in vivo bioactivities of dairy milk protein–derived peptides (MPDP) have proven their potential as an excellent natural alternative to the current medical treatment for metabolic syndrome. In this context, the review discussed the major protein source of dairy milk and provides current knowledge on the novel and integrated approach to MPDP production. A detailed comprehensive discussion is provided on the current state of knowledge regarding the in vitro and in vivo bioactivities of MPDP against metabolic syndrome. In addition, the most important aspect of digestive stability, allergenicity, and further directions for MPDP application is provided.
Recent Findings
The major proteins found in milk are casein and whey, while a minor portion of serum albumin and transferrin are reported. Upon gastrointestinal digestion or enzymatic hydrolysis, these proteins produce peptides with various biological activities including antioxidative, antiinflammatory, antihypertensive, antidiabetic, and antihypercholesterolemic, which could help in ameliorating metabolic syndrome.
Summary
Bioactive MPDP has the potential to curtail metabolic syndrome and potentially act as a safe replacement for chemical drugs with fewer side effects.
Supplementary Information
The online version contains supplementary material available at 10.1007/s13668-023-00472-1.
Keywords
Dairy milk
Protein
Peptide
Bioactivities
Metabolic syndrome
Health benefits
issue-copyright-statement© Springer Science+Business Media, LLC, part of Springer Nature 2023
==== Body
pmcIntroduction
Bioactive peptides have been gaining recognition on both nutraceutical and pharmacological fronts due to their high therapeutic potential and compatibility without adverse interactions with their drug counterparts. The bioactive peptides are characteristically short peptide chains of 2 to 20 amino acids having less than 6000 Da molar mass [1, 2]. They are obtained from the proteolytic clipping of specially sequenced amino acid chains catalyzed by endogenous and exogenous enzymes from plant, animal, and/or microbial sources during food processing [3]. In vitro, the desired peptides may be fractionated and tracked during complex proteolytic processes by using peptidome tools [4]. These peptides can be derived from both animal and plant proteins, whose functionality and potency are determined by C- and N-terminal amino acids, sequence, and affinity after proteolysis from their derivative proteins [1]. For their pharmacological and therapeutic application, the most potent peptide sequence is synthesized in a pure form and subjected to confirmation tests for their specific biological activities. These activities include anticancer, antioxidant, antiinflammatory, antimicrobial, antihypertensive, hypoglycemic, and immunoregulatory actions [1]. There are extensive ongoing investigations into the effects of bioactive peptides on metabolic syndrome, particularly because the modern sedentary lifestyle is a prime cause of metabolic syndrome, which increases the chance of developing cardiovascular disease fivefold. Therefore, bioactive peptides are rapidly gaining recognition for their diverse biological properties [5]. Grundy [6] explains that metabolic syndrome is a proinflammatory and prothrombotic condition characterized by hypertension, hyperglycemia, atherogenic dyslipidemia, and insulin resistance that can be treated by adopting a healthy lifestyle. Bioactive peptides play a vital role in modulating cholecystokinin (a hormone-stimulating pancreatic enzyme secretion) receptor expression, glucose uptake, insulin signaling regulation, adipocyte differentiation, and even mimicking the insulin hormone itself. Bioactive peptides have a potential future in modern medicine because of these functional bioactivities [7•].
Major proteins in milk are casein (αS1- and αS2-caseins, β-casein, and κ-casein) and whey (α-lactalbumin, β-lactoglobulin, lactoferrin, and glycol-macropeptide), while serum albumin and transferrin are minor proteins. Milk proteins subject to systematic hydrolysis form amino acid sequences having activities analogous to hormones [8] and are thus an excellent source of bioactive peptides. Milk proteins may either be enzymatically hydrolyzed or microbially fermented in in vitro conditions followed by the application of favorable separation and purification techniques. Various chromatography techniques are used for the fractionation, purification, and identification of the bioactive peptides in the hydrolysate [9••].
Bovine and buffalo milk are commercially abundant kinds of dairy milk, and their peptides obtained by hydrolysis have promising health benefits, including antioxidative, antiinflammation, antihypertensive, antidiabetic, antihypercholesterolemic, and anticarcinogenic [10, 11]. These peptides are often categorized as excellent natural substances that could be used as alternatives to current treatments for metabolic syndrome with fewer side effects. In vivo research on milk-derived peptides has been conducted for two decades, which has reported their potential effectiveness to treat metabolic syndrome by their effects on multiple body systems such as digestive, endocrine, immune, circulatory, and cardiovascular systems [11]. Thus, dairy MPDPs, solely or in combination, possess multifunctional properties that can be used to control or treat metabolic syndrome as medicinal alternatives.
Even though numerous studies have emphasized the production and identification of bioactive MPDPs, very few have evaluated their in vivo conditions. This has led to a huge gap in the demonstrated efficacy of in in vitro conditions and in vivo assessment and applications. This review provides a comprehensive overview of the potential effects of dairy (bovine and/or buffalo) milk–derived peptides on metabolic syndrome as well as their mechanisms of action. The search for scientific literature published since 2010 was conducted using Scopus, PubMed, and Google Scholar search engines. Relevant articles examining the effects of bovine and/or buffalo milk–derived peptides on antioxidants, hypertension, diabetes, inflammation, cardiovascular disease, and lipidemia are included in the review. The following sections review the composition of milk proteins; peptide production methods; the mechanism of action of bovine and/or buffalo peptides against various diseases associated with metabolic syndrome; peptide digestibility; and allergenicity.
Dairy Milk Proteins
The major milk proteins are casein micelle and whey, while there is a certain residual protein in the milk fat globule membrane. Milk proteins constitute around 4% of the total milk components, and casein and whey comprise 80% and 20% of total protein, respectively, where caseins are categorized into α-casein, β-casein, and κ-casein and whey into β-lactoglobulin and α-lactalbumin mainly [12] (Fig. 1). Proteins such as lactoferrin, fatty acid binding protein, lactoferrin, prolactin, and folate binding proteins are also found in milk and are considered minor proteins. Positive effects of casein and whey have been found in the digestive, nervous, hormonal, circulatory, immune, and cardiovascular systems, along with other body functions [13, 14]. Additionally, milk proteins form the basis for a wide range of peptides with outstanding functional and immunological activities [8, 15].Fig. 1 Overall constituent of dairy milk protein and peptide with its potential health protective activity
Casein
Casein protein contains all nine essential amino acids, with an unusually high leucine fraction. Cow and buffalo milks contain 108 mg/g and 90 mg/g of casein respectively [16]. Due to the presence of proline and cystine fractions, casein lacks the disulfide bridge with α-helix structures. The casein protein is also known as casein phosphoproteins because it contains 0.7–0.9% phosphorous, is hydrophobic, and has a strong affinity for calcium binding, which makes them water-insoluble [14]. The αs1-casein, αs2-casein, β-casein, γ-casein, and κ-casein fractions in the buffalo milk are 14.4–18, 2.2–2.8, 12.5–15.8, 1.6, and 4.3–5.4 g/kg, respectively. These casein fractions in bovine milk are 12–15, 3–4, 9–11, 1–2, and 3–4 g/kg, respectively. Compared to bovine milk, buffalo milk has a 40% higher concentration of αs1-casein and a 35% higher concentration of β-casein fractions. In contrast, buffalo milk has lower fractions of αs2-casein and κ-casein than bovine milk, at just 6.3 and 12% respectively. Because casein from both species has homogeneous phosphoserine clusters, the proteolysis of buffalo milk casein (i.e., αs1, αs2, and β) is similar to that of bovine milk casein [12]. NAVPITPTL, YQEPVLGPVR, YFYPQL, and VLPVPQK are some buffalo milk casein–derived peptides, and HLPGRG and QNVLPLH are derived from the casein of bovine milk either through hydrolysis or some other proteolysis process [17–21].
Whey Protein
Whey proteins are regarded as beneficial proteins from a nutritional standpoint because of their abundance in branched-chain amino acids and bioactive proteins such as α-lactalbumin, β-lactoglobulin, lactoferrin, immunoglobulins, and serum albumin [22]. Whey protein consumption has a wide range of beneficial effects on metabolic health, including the reduction of metabolic syndrome, hyperlipidemia, atherosclerosis, and hypertension [23, 24]. Proteomics analysis has revealed that buffalo whey has a potent hypotensive and immune-enhancing component, and is one of the main sources of immunoreactive components, indicative of good health-promoting functions [25]. The major whey proteins in milk are alpha-lactalbumin and beta-lactoglobulin. Concentrations of α-lactalbumin and β-lactoglobulin in buffalo milk are 1.4 and 3.9 g/kg, respectively, while those of cow milk are 2–4 and 1–5 g/kg [26]. Alpha-lactalbumin, beta-lactoglobulin, and lactoferrin contain 129, 162, and 690 amino acids, respectively, and alpha-lactalbumin has a comparatively high proportion of cysteine, lysine, tryptophan, and branched-chain amino acids [14]. IQKVAGTW and SVDGKEDLIW are some examples of buffalo milk whey–derived peptides; LDQWLCEKL, ELKDLKGY, and ILDKVGINY are derived from bovine whey [27–30]. Whey proteins can thus be used to produce bioactive peptides with various health benefits [11].
Novel and Integrated Approach for the Production of Milk Protein–Derived Peptide (MPDP)
Apart from conventional methods (enzymatic hydrolysis, fermentation, and in vitro digestion) mentioned in the supplementary file for producing bioactive peptides, novel technologies, such as high hydrostatic pressure, ultrasounds, microwave-assisted extractions, ohmic heating, pulsed electric fields, and subcritical water hydrolysis (Fig. 2), are being recently explored to achieve efficient proteolysis of the parent proteins without compromising their functionality and bioactivity [31]. These novel technologies, when coupled with microbial fermentation or enzymatic hydrolysis to produce bioactive peptides, can increase their yield bioactivity and reduce production cost and time as compared to conventional approaches [32, 33]. For instance, ultrasonic waves produce cavitation bubbles and generate tremendous energy due to oscillations that enhance the production of bioactive peptides during enzymatic hydrolysis. Peptides produced from milk proteins by pretreatment with ultrasonic waves (single frequency 28 kHz, power density 20 W/L) followed by enzymatic hydrolysis showed higher angiotensin I-converting enzyme (ACE) inhibition rate than non-ultrasound treated peptides [34].Fig. 2 The schematic diagram showing bioactive peptide production from cow and buffalo milk
High-pressure processing, on the other hand, is the most recent approach to cleaving peptides and assisting hydrolysis by proteolytic enzymes [35, 36••]. High hydrostatic pressure involves the application of isostatic pressures ranging from 100 to 1000 MPa with or without thermal treatment. Combined high pressure and heat cause protein denaturation which exposes hidden peptide sequences and increases the number of sites for proteolytic activity. High hydrostatic pressure–assisted enzymatic hydrolysis of β-lactoglobulin from cow milk increases the yield of bioactive peptides with antioxidant and antiinflammatory properties [31]. Likewise, evidence-based research on the application of microwave pulses of electric fields and ohmic heating has improved the release of bioactive peptides by unfolding and denaturation of peptide sequences [37]. Furthermore, previous studies have suggested an increase in the degree of hydrolysis of whey protein isolate when hydrolyzed with subcritical water [38]. However, there is still a dearth of studies relating to the application of these novel technologies to extract peptides from cow and buffalo milk proteins.
In addition to the intervention of green technology for generating bioactive peptides, in silico approaches also play a crucial role in predicting the bioactivities and binding affinity and mechanism of the peptides released from a parent protein [9••]. In silico approaches are also useful for overcoming the hurdle of gastrointestinal transit by simulating the conditions, molecular docking, protein structure, protein–ligand interaction, and quantitative structure–activity relationship models. Such an approach provides some insight into the fate of these peptides during their transit through the gastrointestinal tract. Even though the in silico approach has enormous potential for understanding virtual protein and peptide phenomena in humans, in vitro and in vivo testing remain the only valid experimental approaches for peptide identification and understandings their biological activities [39].
Milk Protein–Derived Peptide Effects on Metabolic Syndrome
In the last decade, several functional peptides derived from cow and buffalo milk have been identified and investigated for both their in vitro and in vivo biological activities. These peptides, produced during the processing of milk proteins, serve as a source of inhibition for several metabolic syndrome-related enzymes and pathways, as indicated in Tables 1 and 2.Table 1 Peptides from cow or buffalo milk’s in vitro activities in the treatment of the metabolic syndrome and associated diseases
Species Isolated and identified peptide Techniques used for synthesis and isolation of MPDP Assay methods Biological activity Reference
Bovine and buffalo milk PYPQ, YFYPE, EMPFK, PQSV; RELEE(f18–20 β-CN), TVA(f163–165 κ-CN), MADNKQ(f69–74 αs1-CN), EQL Enzymatic hydrolysis (trypsin, alcalase); Ultrafiltration (UF)/ RP-HPLC/ LC–MS/MS Antioxidant ABTS•+ radical scavenging [42, 43]
Buffalo milk casein YQEPVLGPVR
YFPYQL
LLY
- Antioxidant/antiinflammatory ORAC/ABTS•+/Caco-2 cells/Swiss Albino mice [18–20]
Buffalo milk YPSG, HPFA, KFQ; FPGPIPK, IPPK, IVPN, QPPQ Enzymatic hydrolysis: (papain, pepsin, or trypsin) Antioxidant/antihypertensive ACE inhibitory activity [44]
Commercial milk AGWNIPM, YLGYLEQLLR Fermentation (3 different strains of Lactobacillus) Antioxidant/immunomodulatory activity ABTS•+ radical inhibition, modulation of the pro- and antiinflammatory cytokines and NO release [45]
Commercial milk (bovine milk) YLGYLEQLLR (αS1-casein), VKEAMAPK (β-casein), YIPIQYVLSR (κ-casein) Fermentation (3 different strains of Lactobacillus)/ LC–MS/MS Antioxidant DPPH/ABTS•+ radical scavenging [46]
Bovine milk VPYPQR, ARHPHPHLSFM,
RHPHPHLSFM
Fermentation milk (four synthetic peptides) Antioxidant ABTS•+/DPPH radical scavenging; lipid peroxidation inhibition; activated Keap1-Nrf2 signaling pathway [47, 48••]
Buffalo/bovine milk β-casein VLPVPQK (PEP) - Antioxidant; antiinflammatory; antihypertensive ABTS•+ radical scavenging; Nrf2 inhibitor (Keap1); reduced LDH activity, lipid peroxidation and intracellular ROS production, oxidative stress in fibroblast cells; ACE inhibitory activity [50, 51, 59, 63]
Casein YQLD, FSDIPNPIGSEN, FSDIPNPIGSE, YFYP Hydrolysis by two microbial proteases, protein SD-NY10 and protease A “Amano” 2SD Antioxidant Exhibited different antioxidant activity by activating the Keap1-Nrf2 signaling pathway in oxidative damaged HepG2 cell model [114]
Bovine/buffalo milk IPP
VPP
- Antioxidant; antiinflammatory; antidiabetic; antihypertensive Inactivate inflammatory signaling pathway; DPP-IV inhibition [56–58, 72, 86]
Buffalo milk casein YQEPVLGPVR - Antiinflammatory Alteration in cytokines and macrophages [18]
Bovine milk casein MKP - Antihypertensive ACE-inhibitory activity [65]
Bovine milk casein hydrolysate EKVNELSKαs1-casein, NMAINPSKENLCSTFCKαs2-casein – Antihypertensive ACE-inhibition [67, 68]
β- and αs1-casein YPFPGPIPN, HLPLP, AYFYPEL Human jejunal digests and SGD Antihypertensive ACE-inhibition [69]
Bovine milk whey EVLNENLLRF Fermentation (Pediococcus acidilactici SDL1414) Antihypertensive ACE inhibition [73]
Bovine milk Lys-Ala-Ala-Leu-Ser-Gly-Met; Lys-Pro-Ala-Gly-Asp-Phe; Lys-Lys-Ala-Ala-Met-Ala-Met; Leu-Asp-His-Val-Pro-Gly-Gly-Ala-Arg Fermentation (Lactobacillus strains) Antihypertensive ACE inhibition [74]
Bovine whey α-lactalbumin LDQWLCEKLf(115–123) Enzymatic hydrolysis (trypsin) Antidiabetic DPP-IV inhibition [29]
Bovine α-lactalbumin hydrolysates ELKDLKGY, ILDKVGINY Enzymatic hydrolysis (Alcalase) Antidiabetic DPP-IV inhibitory activity [27]
β-Lactoglobulin LKPTPEGDL; LKPTPEGDLEIL Enzymatic hydrolysis (Pepsin) Antidiabetic DPP-IV inhibition [84]
Commercial milk (αS1, αS2-CN) - Enzymatic hydrolysis (Dregea sinensis protease) Antidiabetic α-glucosidase inhibition [83]
Bovine casein HLPGRG, QNVLPLH, PLMLP; MFE; GPAHCLL, ACGP Enzymatic hydrolysis (alcalase and pronase E) Antidiabetic Inhibition of DPP-IV; α-glucosidase and α-amylase [21]
Commercial milk YPSYGL, HPHPHLSFMAIPP, SLPQNIPPL Fermentation (two strains of Lactococcus) Antihypertensive/antidiabetic Dual function (ACE and DPP-IV) inhibition [88]
Table 2 In vivo/ex vivo bioactivities of a peptide derived from bovine/buffalo milk in the treatment of metabolic syndrome and related diseases
Species Peptide Methods and study models Biological activity References
Bovine milk LLY Antioxidant and antiinflammatory in mice-model Reduce the activities of antioxidative enzymes; cytokines modulation [19]
Buffalo milk lactoferrin SVDGKEDLIW Mice-model study Improve antioxidant enzymes (SOD and GSH-PX) activity [30]
Dairy peptide PGPIPN Antioxidant/antiinflammatory/antihypercholesteremic in AALI mice-model Lower MDA, improve GSH-Px and SOD activity; antiinflammatory markers [53]
Buffalo milk casein VLPVPQK Mice-model Suppress weight gain and lipid peroxidation; enhance antioxidation [51]
- VLPVPQKAV, RYPSYGLN Lipopolysaccharide (LPS)-stimulated rodents Pro- and antiinflammatory markers alteration [61]
Buffalo milk casein NAVPITPTL Antiinflammatory activity in ovariectomized (OVX) rats Cytokines modulation [62]
Buffalo casein YFYPQL Antiinflammatory activity in cultured mice splenocytes Pro- and antiinflammatory markers modulation [20]
Commercial Milk KFWGK Antihypertensive in spontaneously hypertensive rats (SHRs) CCK-dependent vasorelaxation [75]
Bovine casein YQKFPQYLQY (YQK) Antihypertensive in Wistar rats and SHRs Lowering SBP [76]
Bovine casein MKP Antihypertensive in spontaneously hypertensive rats (SHRs) Lowering SBP [65]
Bovine casein 90RYLGY94 143AYFYPEL149 Antihypertensive in spontaneously hypertensive rats (SHRs) Lowering SBP [77]
Bovine/buffalo milk IPP; VPP Antihypertensive in spontaneously hypertensive rats (SHRs) Decreased CASNA following mean arterial pressure reduction [78]
Bovine/buffalo milk IPP; VPP Antihypertensive in human subject BP-lowering effects of lacto-tripeptide [79]
Bovine milk LIVTQTMKG Antidiabetes in mice-model Protective effect on β cells of alloxan-induced type-1 diabetic mice [91]
Cow milk beta casein VPYPQ (f 193–197) Antidiabetes in mice Reduced postprandial blood glucose levels in a dose-dependent manner [86]
Milk whey ALPM, LWM Antihyperuricemia in Sprague–Dawley rats xanthine oxidase inhibition [96]
Antioxidant Activity
In Vitro
Studies have shown that metabolic syndrome is associated with oxidative stress, a proinflammatory state, and intracellular redox imbalance as a result of an increase in reactive oxygen species (ROS) formation, which results in mitochondrial dysfunction, protein accumulation, lipid oxidation, and ROS-related impairment [40, 41]. Novel buffalo casein peptides (RELEE, TVA, MEDNKQ, EQL) and bovine peptides (PYPQ, YFYPE, EMPFPK, PQSV) were identified in milk samples hydrolyzed by the two enzymes, trypsin and alcalase. In vitro, peptides RELEE, TVA, and EQL have demonstrated ROS binding potential in 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS•+) radical [42, 43]. Furthermore, the investigation of buffalo casein–derived peptides YQEPVLGPVR, YFPQL, and LLY displayed extraordinary free radical scavenging capacity [18–20]. According to Abdel-Hamid et al. [44], peptides with Tyr/His residues and Pro and/or Phe residues were found to have high antioxidant activity in papain hydrolyzed buffalo milk. The hydrolyzed buffalo milk fractions included a variety of peptides with antioxidant activity, some of which have been reported in previous studies, and some peptides (YPSG, KFQ, and HPFA) were anticipated to contribute to the antioxidant activity [44]. In addition, research by Srivastava et al. [45] outlined that fermented milk–derived peptides AGWNIPM and YLGYLEQLLR possess higher antioxidant activity with AGWNIPM possessing the highest ABTS•+ radical inhibition (73.45%) followed by YLGYLEQLLR (64.46%). Also, peptides YLGYLEQLLR, VKEAMAPK, and YIPIQYVLSR derived from the milk fermentation by different strains of lactobacillus (L. brevis CGMCC15954, L. plantarum A3, and L. reuteri WQ-Y1) displayed potent 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical scavenging activity [46].
Besides the in vitro antioxidant inhibitory assay, understanding the mechanism of a peptide in cell lines is crucial. The antioxidant activity in the cell medium is mediated by activation of the Keap1-nuclear factor erythroid 2–related factor 2 (Nrf2) signaling pathway [18, 20, 47]. Peptides derived from milk proteins can shield Caco-2 cells from peroxide-induced oxidative stress (lipid peroxidation) by activating the Keap1-Nrf2 pathway [47]. Antioxidant activities of bovine milk–derived tripeptide peptide (LLY) were explored in Caco-2 cells, which revealed peptide-induced antioxidative properties by inhibition of intracellular ROS production, mitigation of malondialdehyde (MDA), and protein carbonyls production; augmented catalase activity had little impact on glutathione peroxidase and superoxide dismutase (SOD). Also, relative expression of genes Nrf2 and Keap1 and Nrf2 nuclear translocation by LLY peptide were allied with antioxidative signaling [19]. Evidence of transportation of milk casein–derived peptides YFYPQL and YQEPVLGPVR across the epithelial membrane was reported wherein their antioxidant activity was achieved by activating the Nrf-2 stress signaling pathway [18, 20]. Tonolo et al. [48••] showed that fermentation-derived milk peptides VPYPQR, ARHPHPHLSFM, and RHPHPHLSFM exhibited ABTS•+ and DPPH radical scavenging activity and an inhibition of TbOOH-stimulated lipid peroxidation in a cell line by the activation of the Keap1/Nrf2 pathway. The peptide VLPVPQK (PEP), which was released from buffalo milk β-casein, increases cellular protection by the relative expression of a gene in the Nrf2 inhibitory pathway and reduces oxidative stress in fibroblast cells [49, 50]. Studies on the buffalo casein–derived peptide VLPVPQK also suggest that it can reduce lactate dehydrogenase (LDH) activity and intracellular ROS production. Even at low doses, VLPVPQK showed positive antioxidant effects, including a decrease in glutathione levels and changes in SOD and catalase activity brought on by H2O2 [49, 51]. Overall, recent studies have shown increasing implementation of cell line-based models for investigating the antioxidant properties of bioactive peptides.
In Vivo (Mouse/Human)
The antioxidant properties of milk-derived peptides are multifaceted and partake in multiple pathways, including damage cell recovery and oxidative enzyme modulation or inhibition. They can also play a role in chelating metal ions, controlling the production and removal of ROS by enzyme modulation glutathione peroxidase and SOD, and maintaining cellular integrity [52]. To have an effective antioxidant capacity, peptides should comprise amino acids such as tryptophan, tyrosine, histidine, and proline, and display a hydrophobic character. In addition, two hydrophobic amino acids, leucine or valine, influence peptide antioxidant and lipid peroxidation capacity. The sequence, configuration, structure, and molecular weight of amino acids also have an impact on these processes. In an in vivo study, tripeptide (LLY) unveiled remarkable antioxidative potential against ethanol-induced oxidative stressed mice by increasing glutathione and reducing the activities of MDA. In addition, tripeptide administration revealed reduced activity of glutathione peroxidase and catalase regardless of the dosage, but higher peptide doses (1 mg/kg BW/day) were only sufficient to lower the activity of SOD [19].
In a recent study, after 6 weeks of oral administration of the peptide SVDGKEDLIW derived from buffalo milk lactoferrin, there was a significant improvement in the activity of antioxidant enzymes SOD and GSH-PX in several organs and systems, including the liver, heart, brain, and blood. Additionally, it has also been reported that the levels of MDA in blood and heart tissue were also lowered by the peptide [30]. In another study, a similar observation was reported wherein acute alcoholic liver-injured (AALI) mice were treated with bovine milk–derived peptide PGPIPN via oral gavage to evaluate the role of dairy peptides in preventing and reducing AALI [53]. The study suggested that PGPIPN reduced alcoholic hepatocyte damage and oxidative stress in mouse liver tissues, including a lowering of MDA levels and increased GSH-Px and SOD activity [53]. An investigation of peptide VLPVPQK from buffalo milk casein administration in the mice model was carried out by Mada et al. [51], which indicated a suppression of excessive body weight gain and lipid peroxidation, and an enhanced antioxidative status. The in vivo studies investigating the antioxidative properties of dairy protein–derived peptides are limited, hence more emphasis on carrying out confirmative studies on the biological activities of peptides in the in vivo model should be carried out.
Antiinflammatory
In Vitro
Numerous aspects of metabolic syndrome–related oxidative stress and inflammation have been identified, including circulating inflammatory biomarkers like fibrinogen, C-reactive protein, serum amyloid A, macrophage/monocyte, neutrophil, cytokines, immune cells, and adipose tissue abnormalities [54]. Obesity-induced inflammation is characterized by an elevation of cytokines and dysregulation of adipokines like interleukin (IL)-6, IL-8, IL-1β, interferon-gamma (IFN-γ), tumor necrosis factor-ɑ (TNF-ɑ), and transforming growth factor (TGF-β) [55]. Antihypertensive peptides such as VPP and IPP derived from bovine milk have varying inhibitory effects on inflammatory pathways and potent impacts on the migration, proliferation, and mitigation of inflammatory factors in vascular smooth muscle cells [56, 57]. To elucidate, mitogen-activated protein kinases and tyrosine kinases are the main intracellular kinases that are activated when the AT1R is activated by Ang II, which results in excessive proliferation, an inflammatory response, and oxidative stress in vascular smooth muscle cells. When IPP and VPP were evaluated for their effects on AT1R expression in vascular smooth muscle cells, the results showed no consistent effect on AT1R protein levels, indicating that the protective effect was exerted downstream of AT1R. Tripeptide VPP reduced the activation of inflammatory signaling kinases such as NF-κB pathways, which can trigger inflammation and cell proliferation [58]. Furthermore, possible antiinflammatory effects were observed through cytokine (IL-10, IL-1β, and IL-8) and nitric oxide (NO) production, indicating that they may cause proinflammatory activity through IL-8 or IL-6 production only if an inflammatory stimulus was already present [56, 57].
In another study, the effects of VLPVPQK derived from buffalo casein on inflammation were examined, wherein immune cells, including monocytes, mast cells, and leucocytes, were observed at the site of peptide administration. Additionally, the proinflammatory mediator TNF-α, produced by fibroblasts and endothelial cells, was lowered on treatment with various concentrations of the peptide VLPVPQK [49, 59]. Moreover, decapeptide YQEPVLGPVR from buffalo casein inhibited the growth of murine splenocytes, decreased levels of proinflammatory cytokines (Interferon-γ), while increasing antiinflammatory cytokines TGF-β and Interleukin-10, and increasing macrophage phagocytosis [18].
In Vivo/Ex Vivo
Currently, the exploration of the antiinflammation effects of peptides is being extensively studied in animal model systems. The amino acid type in the amino-terminus and carboxyl-terminus of peptides plays a pivotal role in cytokine regulation caused by the peptides. This finding demonstrates that the dominant molecule recognized by receptors on lymphocytes and macrophages is the amino acid from the R group in each terminus of peptides. Research on immunomodulating effects of hexapeptide PGPIPN on acute alcoholic liver injured (AALI) mice demonstrated prevention and reduction of AALI, mitigation of alcoholic hepatocyte damage, and a dose-dependent alleviation of hepatocyte oxidative stress and endoplasmic reticulum stress by regulating IL-1β, TNF- α, and IL-6 expression [53, 60]. Furthermore, the elevation of serum alanine transaminase and aspartate aminotransferase levels reduced inflammation and hepatocyte damage [53, 60]. Research on the administration of peptides derived from milk fermented by a specific strain of L. fermentum on lipopolysaccharide (LPS)-stimulated mice demonstrated that peptides VLPVPQKAV and RYPSYGLN have a possible antiinflammation effect after 3 weeks of treatment. At the same time, the formation of proinflammatory cytokines IL-6 and TNF-α decreased, while production of the antiinflammatory cytokine IL-10 increased [61] upon treatment with these peptides.
Recent research has investigated changes in proinflammatory and antiinflammatory factors upon oral ingestion of peptide NAVPITPTL derived from buffalo milk casein in ovariectomized (OVX) rats. Interestingly, after administration of NAVPITPTL at 100 µg/kg for 8 weeks, there was a significant increase in serum TGF-β levels and a decrease in IL-6 and TNF-α levels [62]. Likewise, when hexapeptide YFYPQL derived from buffalo casein was incubated in cultured mice splenocytes (ex vivo), production of the proinflammatory cytokine IFN-γ, which is a crucial component of the metabolic inflammation circuit, was reduced, and an inhibition of proliferation and phagocytosis of peritoneal macrophages was observed. Moreover, activation of the antioxidative enzyme and an increment in antiinflammatory cytokine-IL-10 was detected, thus a combination of antioxidant and antiinflammatory properties was noted [20]. Tripeptide LLY has also been shown to have an antiinflammatory response in an ex vivo environment by modulating several cytokines (IL-10, IFN-γ, and TGF-β) and associated pathways with improvement in peritoneal macrophage phagocytosis [19]. Likewise, when treated with aged rat skin fibroblasts, the bovine milk β-casein–derived peptide VLPVPQK increased cell migration by decreasing IL-6 and TNF-α, nuclear transmigration of the dormant Nrf2, and cell proliferation. The peptide reverses the growth arrest in aged fibroblast cells by lessening the activities of caspase-9 and -3, retaining nuclear integrity, and downregulating NF-κB/p38MAP kinase signaling by means of decreasing phosphorylated p38MAP kinases in the cytoplasm and by Nrf2 activation [63, 64].
Antihypertensive Activity
In Vitro
In both in vivo and in vitro model systems, the mechanisms responsible for hypertension rely on blood pressure–lowering effects via renin inhibition, nitric oxide (NO)-mediated vasodilation, ACE inhibition, and increased antioxidant response. The different MPDPs are shown to inhibit ACE through various inhibitory mechanisms such as competitive, non-competitive, or a mixed-type of enzyme inhibition. Abdel-Hamid [44] discovered three novel peptides with active ACE inhibition activity (FPGPIPKβ-CN, QPPQβ-CN, IVPNαS1-CN, and IPPKκ-CN) derived from hydrolyzed buffalo milk protein, where each has a terminal hydrophobic amino acid and IC50 values ranging from 9 to 49 μgmL−1. Other ACE inhibitory bioactive peptides (YPVEPFT, GPFPIIV, YPFPGPIPK, YPFPGPIPN, LPVPQ, and DMPIQ) reported by Abdel-Hamid et al. [44] were also identified from bovine milk protein. Likewise, peptide Met-Lys-Pro (MKP), as a fraction of AMKPW, was isolated from bovine milk casein and reported to be a potent ACE inhibitor (IC50: 0.43 μM) [65]. The mechanism behind antihypertension and ACE inhibition can be understood to be due to the functional ability of the renin–angiotensin–aldosterone system (RAAS). Conversion of angiotensinogen by renin to form angiotensin I (Ang I) activates the RAAS system, which then proceeds to the conversion of angiotensin I (Ang I) to angiotensin II (Ang II) by ACE (EC 3.4.15.1). Vasoconstriction is caused by Ang II binding to angiotensin II receptor type 1 in smooth muscle cells of blood vessels [66]. Thus, inhibition of ACE is the plausible therapeutic target for hypertension. Peptides EKVNELSK and NMAINPSKENLCSTFCK were two novel ACE inhibitor peptides derived from bovine casein hydrolysate (αs1-casein and αs2-casein, respectively), with IC50 values of 5.998 mM and 129.07 μM, respectively [67, 68].
In addition, antihypertensive peptides derived from β- and αs1-casein (YPFPGPIPN, HLPLP, and AYFYPEL) were found to be present in human jejunal digests after oral intake and upon simulated gastric digestion (SGD) of dairy casein and whey milk protein powders [69]. Exploration of partially hydrolyzed whey protein yielded the four potential ACE inhibitory peptides, PQVSTPTL, MPGP, PMHIR, and PPLT, with no allergenicity or toxicity and IC50 values of 86, 179, 90, and 168 μM, respectively [70]. Simulated gastrointestinal digestion of cow and buffalo milk protein subsequently resulted in the release of an ACE inhibition peptide. The novel peptide VLPVPQK obtained from simulated gastrointestinal digestion of buffalo milk casein had the highest ACE inhibitory activity and the strongest bond with hACE [71]. IPP (Ile-Pro-Pro) and VPP (Val-Pro-Pro), on the other hand, isolated solely or in conjugation with other amino acids, exhibited antihypertensive effects via RAAS modulation, upregulated endothelial nitric oxide, proinflammatory cytokine expression, and monocyte activity, and/or reduced oxidative stress in vascular smooth muscle and endothelial dysfunction. Furthermore, VPP has been shown to suppress Ang II–induced cell proliferation, oxidative stress, and inflammation [56, 58, 72].
Besides enzymatic hydrolysis–derived milk peptides, numerous studies have been conducted on the antihypertensive peptides formed by proteolytic cleavage during the fermentation of milk. Microbe-specific proteolytic cleavage is common during fermentation, where hydrophobic peptides show higher ACE inhibitory activity. For instance, EVLNENLLRF, a previously known ACE inhibitor, was present in bovine milk whey fermented by Pediococcus acidilactici SDL1414, which demonstrated strong ACE inhibitory activity (IC50: 19.78 µg/ml) [73]. Similarly, Lactobacillus helveticus KLDS.31 and Lactobacillus casei KLDS.105 fermentation of bovine milk resulted in the release of four ACE inhibitory peptides (Lys-Ala-Ala-Leu-Ser-Gly-Met, Lys-Pro-Ala-Gly-Asp-Phe, Lys-Lys-Ala-Ala-Met-Ala-Met, and Leu-Asp-His-Val-Pro-Gly-Gly-Ala-Arg) whose IC50 value ranged from 77.45 to 201 μM [74].
In Vivo
Several animal studies have been conducted to investigate potential metabolic modifications after ingestion of peptides (Table 2), but it is crucial to note that these studies do not replicate the anticipated benefit of oral administration of peptides and bio-active compounds in humans. The mechanisms governing physiology may vary depending on the species, e.g., human or mouse. Nevertheless, before human consumption investigations begin, animal studies of peptides are necessary to understand their potential mechanisms, bioactive modifications, speculated potency, bioavailability, and toxicity. In vivo experiments studied over the past decades provide invaluable information for the human application of bovine or buffalo milk–derived peptides as an antihypertensive agent. The milk-derived pentapeptide KFWGK, for instance, demonstrated a potent and long-lasting antihypertensive effect via cholecystokinin (CCK)-dependent vasorelaxation. KFWGK also decreased blood pressure with a minimum effective dose of 5 g/kg when administered orally to spontaneously hypertensive rats (SHR) with advanced hypertension [75]. An interesting study on Wistar rats and SHRs assessed the effects of single and repeated doses of oral administration of the antihypertensive peptide-YQKFPQYLQY (YQK), obtained by pepsin and trypsin hydrolyzed bovine casein. Intriguingly, 11-week-old male SHRs, weighing 260 g on average, who had received the three different doses of 1, 3, or 9 mg/kg body weight of the peptide YQK (IC50: 11.1 M) displayed a significant reduction in systolic blood pressure (SBP) after 5 h of administration of all three doses. The maximum reduction in the single dose of 1, 3, or 9 mg YQK/kg body weight was 17.5, 26.7, and 40 mmHg, respectively. In a repeated dose experiment, SBP declined considerably following the first dose, which after 4 h dropped by a maximum of 36.8 mmHg. When the second dose was administered after 4 h of the first dose, SBP remained stable. At 8 h following the initial YQK dose, SBP increased once more, and SBP changed by 18.2 mmHg at 12 h, yet still lower than it had been before the first oral administration [76], thus demonstrating an effective antihypertensive effect.
To explain the antihypertensive action of bovine casein–derived peptide in vivo, peptide MKP was orally administered in single and repeated doses to SHRs. Upon oral ingestion of peptide MKP, brief SBP reduction was noted, which occurred in a dose-dependent manner. The ingestion of 1 mg/kg MKP resulted in a drop of SBP to 158.8, 152.2, 158.2, and 166.0 mmHg at 2, 4, 6, and 8 h, respectively. Daily repeated administration of MKP at 10 mg/kg, on the other hand, demonstrated that prolonged oral administration subsequently lowered blood pressure to 163.3 mmHg as compared to 171.7 mmHg in controls [65]. Similarly, the peptides 90RYLGY94 and 143AYFYPEL149 derived from bovine casein produced a significant reduction in the SBP of the animal model, with a maximum decrease after administration of 90RYLGY94 at 6 h, 23.8 mmHg, and 143AYFYPEL149 at 4 h, 21.1 mmHg [77]. Inhibition of cutaneous arterial sympathetic nerve activity (CASNA) contributes to lowering blood pressure by regulating peripheral artery and arteriole constriction, which is intrinsically related to peripheral vascular resistance. Possibly via the afferent vagus nerve, IPP and VPP gastric administration significantly decreased CASNA following mean arterial pressure reduction. Additionally, it has been demonstrated that IPP and VPP inhibit renal sympathetic nerve activity, whose stimulation promotes the release of renin and sodium reabsorption, both of which are mechanisms that can lead to hypertension [78].
In addition to animal studies, a double-blind randomized, placebo-controlled cross-over human study to evaluate the effect of milk tripeptide on blood pressure and vascular renal function in prehypertensive Japanese subjects, IPP/VPP supplementation reduced SBP to 127 mmHg in comparison to the placebo group of 130 mmHg on 7-day mean tele-monitored BP [79]. Although the detailed mechanism governing the blood pressure–lowering effects of lacto-tripeptide supplementation remains unclear, a plausible mechanism might involve the inhibition of an enzyme involved in the RAAS system and/or an increase in the release of vasodilatory peptides like bradykinins. The improvement of arterial stiffness and endothelial function by lacto-tripeptide supplementation has been reported in a study by Cicero et al. [80], while Tomiyama et al. [79] demonstrated no effect on either of these parameters in prehypertensive subjects. Overall, sufficient in vitro and significant in vivo studies on antihypertensive dairy–derived bioactive peptides have been carried out since the last decade, which have demonstrated their antihypertensive effects.
Antidiabetic
In Vitro
Diabetes mellitus, a hyperglycemic condition, is caused by insulin resistance or the inability to produce insulin, and appropriate ways of maintaining blood glucose homeostasis are critically needed. Dipeptidyl peptidase IV (DPP-IV) inhibits the activity of incretin hormones, glucagon-like peptide (GLP), and glucagon-like peptide-1 (GLP-1), which signal insulin secretion from pancreatic beta cells, resulting in hyperglycemia; DPP-IV inhibitors, on the other hand, provides an opportunity to control blood sugar by blocking DPP-IV action. Additionally, inhibition of α-glucosidase and α-amylase, two enzymes involved in complex carbohydrate digestion, is considered to be an efficient approach to lower blood sugar levels [81]. It is anticipated that α-amylase competitively interacts with peptide sequences having hydrophobic amino acids Leu, Met, and Pro at the terminal. Furthermore, α-glucosidase is inhibited by the hydrophobic amino acids Met, Pro, Phe, and Leu, while DPP-IV inhibition occurs with both hydrophobic Ala, Gly, Leu, Pro, Met, and Trp amino acids and by hydrophilic Gln, His, Arg, and Ser amino acids at peptide termini [21]. Enzymatic hydrolysis of milk and milk protein from cow and buffalo liberate peptides with diabetes marker inhibition potential [82, 83]. Several novel peptides with strong inhibition of DPP-IV, alpha-glucosidase, and alpha-amylase have been identified, such as LDQWLCEKL [29], LKPTPEGDL, and LKPTPEGDLEIL [84], ELKDLKGY, ILDKVGINY, and KILDK [27, 85•], and RNAVPITPTLNR, TKVIPYVRYL, YLGYLEQLLR, and FALPQYLK [83].
A novel DPP-IV inhibitory peptide LDQWLCEKL (IC50: 131 μM) was isolated from bovine whey α-lactalbumin (f 115–123) upon trypsin hydrolysis, indicating that this peptide is effective as a preventive or adjuvant therapy for type 2 diabetes management [29]. The majority of peptides with proline or alanine at the N-terminus possess DPP-IV inhibition, such as milk-derived peptide VPYPQ, Diprotin A (IPI), and Diprotin B (VPL), which have either one or two proline residues and that are considered strong competitive inhibitors of DPP-IV [86, 87]. However, for LDQWLCEKL, the hydrophobic amino acid Leu was located at the penultimate position, and it is hypothesized that it interacted with the hydrophobic pocket found at the active site of DPP-IV [29]. Similarly, bovine α-lactalbumin hydrolysate–derived peptides, ELKDLKGY and ILDKVGINY, were obtained by alcalase hydrolysis and demonstrated the DPP-IV inhibitory activity [27]. Interestingly, when TNF-stimulated 3T3-L1 adipocytes interacted with peptide KILDK, it sufficiently minimized insulin resistance in the adipocytes by suppressing JNK phosphorylation (Thr183/Tyr185) and it inhibited proinflammatory gene expression by impeding NF-κB signaling [85•]. Previously, pepsin digested β-lactoglobulin were reported to be enriched with the most potent un-competitive DPP-IV inhibiting fragments, LKPTPEGDL and LKPTPEGDLEIL, with IC50 values of the fractions of 45 μM and 57 μM, respectively [84].
Apart from DPP-IV inhibitory peptides, potent alpha-glucosidase inhibitory peptides have been identified from different fractions of milk hydrolyzed by Dregea sinensis protease. Four novel peptides from αS1, αS2-casein exhibited promising α-glucosidase inhibition. These peptides are anticipated to occupy the active sites of α-glucosidase—potentially Arg428, Arg387, Arg801, Arg727, Arg799, and Trp710—by the formation of hydrogen bonds, thereby circumventing the complexation formation and glycosylation of α-glucosidase with the substrate [83]. Likewise, upon hydrolysis of bovine caseins with alcalase and pronase E at different times led to the production of prospective antidiabetic peptides (HLPGRG, QNVLPLH, PLMLP, MFE, GPAHCLL, and ACGP) capable of inhibiting three diabetic-related enzymes (DPP-IV, α-glucosidase, and α-amylase) [21]. Milk fermentation with the two strains of Lactococcus released YPSYGL, HPHPHLSFMAIPP, and SLPQNIPPL peptides with the dual function of ACE and DPP-IV inhibition [88].
Lastly, two well-known DPP-IV inhibitors, IPP and VPP, demonstrated effective involvement through enhancement of insulin signals, antiinflammation via NF-κB pathway under TNF stimulation, and prospective contribution to insulin resistance prevention. Furthermore, IPP and VPP possess insulin-sensitizing effects that are independent of insulin receptors present in adipocytes. Upon the administration of VPP, glucose transporter-4 (GLUT-4) expression was enhanced in adipocytes, which restored the absorption of glucose in TNF-treated adipocytes [72, 89].
In Vivo
Diabetes is a common complication of metabolic syndrome, which is characterized by hyperinsulinemia, hyperglycemia, insulin resistance, and multisystem inflammation as a result of impaired glucose-insulin metabolism. With the hypothesis that milk peptides are protective against GLP-1 degradation and attenuate DPP-IV activity in vivo, data indicated that whey protein is taken orally elevated plasma GLP-1 levels, which was effective without influencing the DPP-IV activity [90].
Studies in animal models demonstrated ingestion of bovine and buffalo milk–derived peptides ameliorates pancreatic cell damage, improves oxidative stress, and decreases blood glucose. Lacto-ghrestatin (LGP9), a bovine milk–derived peptide with the sequence LIVTQTMKG, has been shown to protect β cells of alloxan-induced type-1 diabetic mice. On treatment with LGP9, alloxan-induced type-1 diabetic mice with hyperglycemia showed a reduction in blood glucose levels and glycated serum proteins (GSP). At the same time, LGP9 treatment increased glucose transporter-2 expression, protected injured β cells by suppressing apoptosis, rescued Ki67 immunoreactivity through IRS2/PI3K/Akt signaling, increased the phosphorylation of FOXO1, and upregulated PDX-1 expression, which resulted in increased insulin secretion [91]. Another peptide, VPYPQ, derived from cow milk beta-casein fraction (fraction number 193–197) was administered orally to 6-week-old mice and found to decrease the postprandial blood glucose levels in a dose-dependent manner, with 90 mol/kg BW being effective and 45 mol/kg BW being ineffective in an oral glucose tolerance test (OGTT) [86]. Thus, these in vivo studies signify that the peptide from cow and/or buffalo milk possesses potent antidiabetic activity and immensely supports the prevention of the symptoms and diseases associated with metabolic syndrome in vitro and in in vivo animal models. But still, there is a dearth of information that directly demonstrates that dairy milk–derived bioactive peptide ingestion attenuates DPP-IV activity in humans, and thus would positively affect type-2 diabetes.
Antihypercholesteremic
In Vitro
Pancreatic lipase and cholesterol esterase are the two lipolytic enzymes responsible for the hydrolysis and digestion of fat and cholesterol esters. When these enzymes are inhibited, intestinal fat and cholesterol absorption are suppressed, which instigates a slow but steady decline in body weight. Peptides from different sources, including cow and buffalo, exhibit inhibitory actions on pancreatic lipase and esterase. The antihypercholesteremic activity and its associated enzyme inhibition by bovine and buffalo milk peptides are shown in Table 1. Mudgil et al. [92•] investigated camel and cow casein–derived peptides with potential inhibitory activity of key lipid-digesting enzymes. The results showed that cow milk casein hydrolyzed by enzyme alcalase and pronase-E increased inhibition of PL. Peptides generated through enzymatic hydrolysis or upon simulated gastric digestion (SGD) from cow casein hydrolysates MMML, FDML, and HLPGRG were prospective pancreatic lipase inhibitors, while peptide LP showed potent cholesterol-esterase inhibition. It has been reported that the enzyme-specific inhibition of two vital hypercholesteremic enzymes, cholesterol-esterase and pancreatic lipase, aids in the prevention of hypercholesterolemia and obesity by inhibiting the uptake of fatty acids, thus limiting the deposition of fatty acids in the body. The lipid regulating function of bovine alpha-lactalbumin and its peptide in HepG2 cell lines was explained by a cell viability assay, triglyceride levels, and peroxisome proliferator–activated receptor-α (PPAR-α) levels in hepatic cells. Upon simulated gastrointestinal digestion (SGID) of bovine alpha-lactalbumin, the peptide sequences GINY and DQW were obtained. These two peptides increased PPARα levels, activated the PPARα signaling pathway, increased expression of β-oxidation-related genes CPT-1a, PPARα, and ACOX1, and decreased expression of lipogenesis-related genes SCD-1, ACC1, and FASN, improving lipid metabolism and decreasing lipid accumulation. Hence, α-lactalbumin and peptides from lactalbumin show potential sources to ameliorate obesity [93].
Exploration of the role of casein-hydrolyzed peptide in the initiation of trans-intestinal cholesterol excretion (TICE) and the hyperlipidemic effect of the peptide and cell lines treated with isolated peptide were studied. Bovine casein hydrolysate–derived peptides SQSKVLPVPQK and HPHPHLSF induced TICE via regulation of the liver X receptor-α (LXR-α) signaling pathway, and enhanced expression of ABCG5 was recorded. Induction of ABCG5 expression in the intestine could contribute to elevated fecal cholesterol excretion. Moreover, these peptides induced fibroblast growth factor 15/19 (FGF15/19) exudation from enterocytes, which diminished hepatic bile acid synthesis involved in adjusting hepato-biliary cholesterol, thereby aiding in the maintenance of cholesterol homeostasis [94]. Consistently, bioactive peptides LQPE, VAPFPE, TDVEN, and VLPVPQ from milk casein hydrolyzed with neutrase facilitated cholesterol-lowering activity by diminishing cholesterol micellar solubility and absorption. The phenomenon of reduced mRNA expression of acetyl-CoA-acetyltransferase-2 and microsomal triacylglycerols (MTP) in proximal intestinal cells was shown to influence the cholesterol-related proteins and enzymes’ expression, which affects cholesterol absorption [95].
Lastly, xanthine oxidase (XO) is presumed to be the source of reactive oxygen species that cause atherosclerosis and cholesterol crystals. The XO inhibitory assay is crucial to understand uric acid biosynthesis. Peptides ALPM and LWM interactions with XO result in stable complexes and inhibit XO. The IC50 values of these peptides to inhibit XO were 7.23 and 5.01 mM, respectively, with both contributing via non-competitive routes [96]. These studies carried out to date did indicate that specific peptide sequences can play a role in lowering cholesterol and fatty acid absorption via different modes of action; however, these studies are not sufficient to draw a strong conclusion, even when considering the effect demonstrated in the in vitro model systems.
In Vivo
Hyperlipidemia, a condition of excess blood lipids or fats, is one of the speculated risk factors of metabolic syndrome and is closely associated with obesity, atherosclerosis, and thrombosis. Although the exact pathophysiology of the metabolic syndrome is unknown, evidence-based experiments have proven that controlling excessive cholesterol prevents mortality. Therefore, maintaining blood cholesterol levels is pivotal and could be beneficial both physically and medically. Several studies have shown that peptides from different sources, including milk, exhibit cholesterol-lowering effects by signaling inhibitory pathways, gene expression, and/or attenuating lipid absorption (Table 2). Hence, the antilipidemic roles of bovine casein–derived peptides generated via pepsin and trypsin were investigated against a high-cholesterol diet-induced hyperlipidemic mouse model. The results demonstrated diminished serum cholesterol, suppression of hepatic CYP7A1 and CYP8B1 expression, and inhibited hepatic bile acid synthesis in the treated group. Meanwhile, there was an increased level of fecal cholesterol and serum fibroblast growth factor [94]. Exploration of the effect of hexapeptide PGPIPN (0.04, 0.4, and 4.0 mg/kg) in AALI mice revealed that PGPIPN had a significant reduction in the serum and liver TG and TC levels. Interestingly, the low-dose (0.04 mg/kg) ingestion of peptide PGPIPN sufficiently decreased TG (0.5 µmol/mL) and TC (0.8 µmol/mL) levels. Thus, it was speculated that PGPIPN attenuates alcohol-induced liver damage by regulating lipid metabolism [53].
Hyperuricemia (HUA), on the other hand, is a metabolic disease that is closely associated with metabolic syndrome. The enzyme involved in the metabolic pathway of purine is a xanthine oxidase (XO), which catalyzes the oxidation of hypoxanthine and xanthine to form uric acid and liberate superoxide anions (O2–), hydrogen peroxide (H2O2), and ROS. Inhibition of the enzyme XO is necessary to prevent uric acid and lipid crystal formation [97]. An experiment in the potassium oxonate-induced HUA rat model displayed that the milk-derived peptides are the prospective XO inhibitors. Their serum uric acid levels and XO activity were considerably lowered by ALPM and LWM interventions (the latter more so), particularly in comparison to the model group, even though the uric acid–lowering effect was not as strong as that of the commercial drug allopurinol [96]. These studies conducted in vivo for investigating the antihyperlipidemia effect are again not sufficient to draw solid conclusions. Therefore, more robust studies involving peptides with different physiochemical properties on various metabolic targets playing roles in hyperlipidemia should be carried out.
Digestive Stability of Dairy Milk–Derived Peptides
Throughout the digestive tract, protein breakdown, modification, and digestion take place, while the main organ for protein and peptide absorption is the small intestine. Peptides, either newly formed or surviving from gastric and intestine enzyme hydrolysis, are transported into the bloodstream. On ingestion of the readily hydrolyzed peptide, it encounters the brush border membrane peptidase before passing through the intestinal epithelium and being absorbed. The brush border membrane peptidase hydrolyzed the ingested peptide further, potentially altering its bioactivity and functional properties [98]. Thus, protecting or preserving the functional molecular characteristics of the peptide through the digestive system is crucial for their bioactivity [76].
To study the digestibility and stability of the peptides, mostly three gastrointestinal proteases, pepsin, chymotrypsin, and trypsin, are employed to mimic the human gastrointestinal tract under simulated gastrointestinal digestion conditions of food. This helps to study the retention of functional activities of peptides when orally administered in a human sample study. However, in the Caco-2 cell digestive stability study, the cell is subjected to several membrane peptidases, such as dipeptidyl peptidase IV, endopeptidases, aminopeptidase, enteropeptidases, and aminopeptidase [99, 100]. According to a study by Xia et al. [101], ACE inhibitory activity declined throughout the various stages of simulated in vitro digestion but was still high after trypsin treatment, which gives plausible insight that the Gly-Ala (GA) dipeptide may have some in vivo stability. Also, when the peptide was fed to SHR rats at a dose of 15 mg/kg, it showed a consistent decrease in blood pressure with a drop of 17.48 mmHg. Another study by Xue et al. [76] demonstrated that on the administration of peptide YQKFPQYLQY (YOK) at different pHs (3–9), ACE activity was still stable, and when subjected to in vitro digestion by digestive enzyme pepsin (1–2 h) and trypsin (4 h), the ACE activity of peptide remained unaffected. Casein hydrolysate–derived peptide VLPVPQK was hydrolyzed by cellular peptidases before efflux, resulting in the new peptide VLPVPQ on the apical surface, which rapidly reached the basolateral chamber [98]. Similarly, peptidases on the surface of Caco-2 cells hydrolyze five milk protein–derived peptides LPYPY, LKPTPEGDL, IPIQY, WR, and IPI, with 8 to 30% of the peptides hydrolyzed after 2 h [99]. Tripeptide transport across the Caco-2 cell layer, however, revealed that it was transported from the apical to the basal chamber in the trans well at a concentration of 6.9 μg/ml in its intact form without being hydrolyzed [19].
For the effective and potent delivery of the functional peptide, different multifactorial and complex approaches are under scrutiny. Delivery of the bovine lactoferrin in conjunction with the solid lipid particles and biopolymer-encrypted liposomes showed improved stability, the solid lipid particles being found to be the primary medium for oral delivery of the bovine milk lactoferrin [102]. Certain emerging technologies such as encapsulation, double emulsion with and without Pickering, liposomes, niosomes, or enteric-coated capsules are being prepared for the safe and optimal delivery of these peptides (Fig. 3) [103, 104]. A comparative study of IPP peptide-loaded niosomes and liposomes, for instance, provides a clear insight into a more stable and effective delivery. During long-term preservation, a niosomal composite–loaded functional beverage displayed better palatability, biological activity, and physicochemical characteristics than a liposomal one [103]. When compared to nonencapsulated peptide, an optimized double emulsion loaded with DPP-IV inhibitory peptide decrypted from hydrolyzed α-lactalbumin of Gir cow milk revealed approximately four times better functionality, i.e., DPP-IV inhibition activity [104].Fig. 3 Digestive stability with various approaches to enhance stability of the milk-derived peptides (description: free form indicates the peptide solely without nano-carriers for delivery, whereas carrier-based indicates the peptide incorporated in the nano-carriers such as nano emulsion and liposomes)
Allergenicity
Food allergy is an IgE-mediated phenomenon that causes a clinical condition or a combination of complications in the respiratory tract (edema in the larynx, rhinorrhea, sneezing, and wheezing), gastrointestinal tract (nausea, vomiting, diarrhea, and abdominal pain), cutaneous region (urtication and angioedema), and cardiovascular (hypotension and tachycardia). Most of the common foods that have allergenicity are peanuts, shellfish, wheat (gluten), eggs, milk, and soybeans [105]. Individuals with impaired digestion are vulnerable to developing food allergies after ingestion; however, allergic reactions can occur through sensitization via other routes, such as skin contact and inhalation (respiratory tract) [106•].
During milk protein hydrolysis, apart from health-beneficial physiologically active peptides, the speculation of cytotoxic or allergenic peptides is very much under consideration, which has placed a clinical safety assessment as a mandatory step. Despite its beneficial property, the mass commercialization of the bioactive peptide is still not up to par with its potential, which could be linked to the infant allergenicity profile of the various peptides resulting from different means of proteolysis [107]. The discrete heterogeneity of bioactive peptides resulting from the hydrolysis is huge, and assessing the beneficial bioactive peptides among all the resulting peptides is complex. Identification of the beneficial bioactive peptides and bioactive peptides with potential allergenicity is a crucial turning point in the commercializing of milk peptides [14]. However, complete and reliable allergenicity profiling of the resultant peptide is not possible, as the study of allergenicity in a clinical setting is unethical; thus, assessments are limited to rodents and cell lines. This only gives the proximal estimation, which does not fully correlate with human physiology, leaving a gap from theoretical benefit to actual implications arising after human consumption [108]. To bridge the gap results from various assessment approaches, such as in vitro, in vivo, and in silico must be performed and integrated to overcome the absence of human clinical trials [9••, 109]. The safety evaluation for peptide-like functional foods still does not have a robust guideline, hence in vivo trials on rodents are considered to be the standard [110]. Because avoiding allergen-containing foods is the only known intervention, safety evaluation for possible allergenicity should be prudent [105]. In the case of milk-derived food products, the prevalence of cow milk allergy is 2–3%; however, enzyme-induced partial hydrolyzation might diminish its allergenicity, as the resultant hydrolysate has an immunoregulatory effect. Based on the degree of hydrolyzation, completely hydrolyzed proteins could be administered to treat the allergic reaction. Similarly, partially hydrolyzed proteins could be administered to prevent allergic reactions [111].
Future Perspectives and Conclusion
The worldwide annual production of milk climbed by an average of 1.6% (838 million tons) in 2018, with India’s production leading the way at 3% (174 million tons). However, India is the least milk-exporting country. Major exporters such as New Zealand, the European Union, and the USA had production increased by 3.2%, 0.8%, and 1.1%, respectively. The projection of annual milk production worldwide was postulated to grow at 1.8% (1060 million metric tons) by 2031. Despite the COVID-19 pandemic, global milk production is still trending up to meet the anticipated growth. However, regardless of the production curve, dairy consumption is declining due to the emergence of plant-based dairy substitutes, which may have negative effects on dairy demand [112], resulting in surplus milk. Hence, commercialized production of the milk protein–derived peptide would be an effective way of relieving the economic burden for the producers.
A possible best way of utilizing milk-derived bioactive peptides would either be the optimization of the yogurt fermentation for increased production of functional peptides or fortification, which might also be an option to supply peptides as per the demands of consumers’ clinical conditions [113]. But, this could only be envisioned if the novel production technologies work in conjunction with the computational (in silico) approach for discovering resultant bioactive peptides, as in silico analysis utilizes bioinformatics to virtually simulate the occurrence of biological systems [35, 39].
Even though the production and purification of milk peptides are challenging, there is growing interest in these bioactive peptides due to the wide range of functionality of milk-derived peptides. As articulated in this review, these peptides have the potential to minimize the effects of metabolic syndrome. In vitro or in animal model systems, indications are that the majority of milk peptides would or have a significant effect on alleviating symptoms of various diseases associated with metabolic syndrome; however, there are still only a handful of human studies published. Given this, extensive research into the promising peptides in human health is required before commercializing milk peptide–based antihypertensive, antidiabetic, and antihypercholesteremic products. Although some peptides are being studied in human trials, determining the mechanisms underlying their physiological effects remains difficult. The major challenge in peptide research is to investigate and get reliable data on the fate of peptides when consumed or ingested and on the changes occurring during their transit through a gastrointestinal phase of digestion. Synthetic peptides of known sequence can be validated for their biological properties in different systems (in vitro, cell lines, ex vivo); however, what effect will transit of the peptide through the human gastric and intestinal phase of digestion have on the bioactive property is a crucial challenge. Research to address this challenge is very important for designing effective peptide therapies for the treatment of various metabolic syndromes.
In conclusion, MPDPs appear to be potential therapeutics, nutraceuticals, and natural medicines in the pharmaceutical industry for those with metabolic syndrome. Furthermore, research is needed to investigate the optimal production and isolation approaches of MPDPs. Additionally, designing stable and efficient delivery systems to be utilized in peptide therapy with enhanced bioavailability is also equally important, which calls for novel research in this area.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 28 KB)
Abbreviations
AALI Acute alcoholic liver-injured
ABTS• + 2,2′-Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)
ACE Angiotensin-converting enzyme
DPPH 2,2-Diphenyl-1-picrylhydrazyl
DPP-IV Dipeptidyl peptidase IV
GLP Glucagon-like peptide
HUA Hyperuricemia
IFN-γ Interferon-gamma
IL Interleukin
MDA Malondialdehyde
MPDP Milk protein–derived peptides
NF-κB Nuclear factor-κB
NO Nitric oxide
Nrf2 Nuclear factor erythroid 2–related factor 2
RAAS Renin-angiotensin-aldosterone system
ROS Reactive oxygen species
SBP Systolic blood pressure
SGD Simulated gastric digestion
SHR Spontaneously hypertensive rats
SOD Superoxide dismutase
TGF Transforming growth factor
TNF-ɑ Tumor necrosis factor-ɑ
XO Xanthine oxidase
Acknowledgements
The authors would like to thank Dr. Sabri Bromage for the English language editing and fluency of the article.
Author Contribution
The author contributions were as follows: P. K., M. D., S. R., and M. D. search and collected literature, compiled data, and drafted the manuscript; N. P. N. conceived the work, analyzed, revised, and supervised the manuscript; S. M., F. A., and A. B. critically evaluated the review for important intellectual content. All authors agreed to their accountable contributions, read, and approved the final manuscript.
Declarations
Conflict of Interest
The authors report no conflict of interest.
Statement of Significance
This review summarizes our current state of knowledge regarding dairy MPDP and their bioactivities against metabolic syndrome. Literature suggests that various MPDPs and their bioactivities can be produced from milk proteins. These identified MPDPs have the potential to reduce metabolic syndrome symptomology via various enzymatic and biochemical pathways. Designing an effective peptide therapy is of prime importance for the application of MPDPs in metabolic syndrome.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Papers of particular interest, published recently have been highlighted as: • Of importance •• Of outstanding importance
1. Balandrán-Quintana RR, Mendoza-Wilson AM, Ramos-Clamont Montfort G, et al. Peptides and Proteins, in Food Bioactives and Health, C.M. Galanakis, Editor. Springer International Publishing: Cham. 2021;79–117.
2. Lafarga T Sánchez-Zurano A Villaró S Industrial production of spirulina as a protein source for bioactive peptide generation Trends Food Sci Technol 2021 116 176 185 10.1016/j.tifs.2021.07.018
3. Toldrá F Reig M Aristoy M-C Generation of bioactive peptides during food processing Food Chem 2018 267 395 404 10.1016/j.foodchem.2017.06.119 29934183
4. Mora L Gallego M Toldrá F New approaches based on comparative proteomics for the assessment of food quality Curr Opin Food Sci 2018 22 22 27 10.1016/j.cofs.2018.01.005
5. Jahandideh F Wu J A review on mechanisms of action of bioactive peptides against glucose intolerance and insulin resistance Food Sci Human Wellness 2022 11 6 1441 1454 10.1016/j.fshw.2022.06.001
6. Grundy SM Metabolic syndrome update Trends Cardiovasc Med 2016 26 4 364 373 10.1016/j.tcm.2015.10.004 26654259
7. • Samtiya M, Samtiya S, Badgujar PC, et al. Health-promoting and therapeutic attributes of milk-derived bioactive peptides. Nutrients. 2022;14(15):3001. The review elucidated the overall bioactivity of the milk-derived bioactive peptide; however, it was not specific to dairy milk–derived peptide and metabolic syndrome.
8. Sultan S Huma N Butt MS Therapeutic potential of dairy bioactive peptides: a contemporary perspective Crit Rev Food Sci Nutr 2018 58 1 105 115 10.1080/10408398.2015.1136590 26852912
9. •• Du Z, Li Y. Review and perspective on bioactive peptides: a roadmap for research, development, and future opportunities. J Agric Food Res. 2022;100353. The review has illustrated the methods for isolation, identification, and purification of bioactive peptides from different sources. Also, it articulated the application of those peptides to human health.
10. Bielecka M Cichosz G Czeczot H Antioxidant, antimicrobial and anticarcinogenic activities of bovine milk proteins and their hydrolysates-a review Int Dairy J 2022 127 105208 10.1016/j.idairyj.2021.105208
11. Nongonierma AB FitzGerald RJ The scientific evidence for the role of milk protein-derived bioactive peptides in humans: a review Journal of Functional Foods 2015 17 640 656 10.1016/j.jff.2015.06.021
12. Khetra Y Meena G Arora S Buffalo milk and its products: composition, nutrition and benefits biotechnological applications in buffalo research 2022 Springer 121 143
13. Kadam B Ambadkar R Rathod K Health benefits of whey: a brief review International Journal of Livestock Research 2018 8 5 31 49 10.5455/ijlr.20170411022323
14. Auestad N, Layman DK. Dairy bioactive proteins and peptides: a narrative review. Nutr Rev. 2021;79(Supplement_2):36–47.
15. Marcone S Belton O Fitzgerald DJ Milk-derived bioactive peptides and their health promoting effects: a potential role in atherosclerosis Br J Clin Pharmacol 2017 83 1 152 162 10.1111/bcp.13002 27151091
16. Rafiq S Huma N Pasha I Chemical composition, nitrogen fractions and amino acids profile of milk from different animal species Asian Australas J Anim Sci 2016 29 7 1022 1028 10.5713/ajas.15.0452 26954163
17. Reddi S Kumar N Vij R Akt drives buffalo casein-derived novel peptide-mediated osteoblast differentiation J Nutr Biochem 2016 38 134 144 10.1016/j.jnutbio.2016.08.003 27736733
18. Sowmya K Bhat MI Bajaj RK Buffalo milk casein derived decapeptide (YQEPVLGPVR) having bifunctional anti-inflammatory and antioxidative features under cellular milieu Int J Pept Res Ther 2019 25 2 623 633 10.1007/s10989-018-9708-7
19. Sowmya K Mala D Bhat MI Bio-accessible milk casein derived tripeptide (LLY) mediates overlapping anti-inflammatory and anti-oxidative effects under cellular (Caco-2) and in vivo milieu J Nutr Biochem 2018 62 167 180 10.1016/j.jnutbio.2018.09.002 30300836
20. Sowmya K Bhat MI Bajaj R Antioxidative and anti-inflammatory potential with trans-epithelial transport of a buffalo casein-derived hexapeptide (YFYPQL) Food Biosci 2019 28 151 163 10.1016/j.fbio.2019.02.003
21. Mudgil P Kamal H Kilari BP Simulated gastrointestinal digestion of camel and bovine casein hydrolysates: identification and characterization of novel anti-diabetic bioactive peptides Food Chem 2021 353 129374 10.1016/j.foodchem.2021.129374 33740505
22. Master PBZ Macedo RCO Effects of dietary supplementation in sport and exercise: a review of evidence on milk proteins and amino acids Crit Rev Food Sci Nutr 2021 61 7 1225 1239 10.1080/10408398.2020.1756216 32363897
23. Zhang Z Zhang R Qin Z-Z Effects of chronic whey protein supplementation on atherosclerosis in ApoE−/− mice J Nutr Sci Vitaminol 2018 64 2 143 150 10.3177/jnsv.64.143 29710032
24. Wang, K., Z. Fu, X. Li, et al., Whey protein hydrolysate alleviated atherosclerosis and hepatic steatosis by regulating lipid metabolism in apoE-/-mice fed a Western diet. Food Research International, 2022: p. 111419.
25. Zhang F Pu J Gu Z DIA proteomics reveals hypotensive and immune-enhancing constituents in buffalo whey from different altitudes Int J Biol Macromol 2020 164 4146 4154 10.1016/j.ijbiomac.2020.08.213 32882282
26. Chauhan, M.S. and N. Selokar, Biotechnological Applications in Buffalo Research. 2022, Springer.
27. Gao J Gong H Mao X Dipeptidyl peptidase-IV inhibitory activity and related molecular mechanism of bovine α-lactalbumin-derived peptides Molecules 2020 25 13 3009 10.3390/molecules25133009 32630113
28. Ashok NR Aparna HS Empirical and bioinformatic characterization of buffalo (Bubalus bubalis) colostrum whey peptides & their angiotensin I-converting enzyme inhibition Food Chem 2017 228 582 594 10.1016/j.foodchem.2017.02.032 28317767
29. Jia C-L Hussain N Ujiroghene OJ Generation and characterization of dipeptidyl peptidase-IV inhibitory peptides from trypsin-hydrolyzed α-lactalbumin-rich whey proteins Food Chem 2020 318 126333 10.1016/j.foodchem.2020.126333 32151919
30. Yang P Abel-Hamid M Romieh E Effect of peptides synthesized from lactoferrin of buffalo milk on oxidative stress in kunming mice J Anim Plant Sci 2020 30 65 71
31. Bamdad F Bark S Kwon CH Anti-inflammatory and antioxidant properties of peptides released from β-lactoglobulin by high hydrostatic pressure-assisted enzymatic hydrolysis Molecules 2017 22 6 949 10.3390/molecules22060949 28590420
32. Ketnawa S Wickramathilaka M Liceaga AM Changes on antioxidant activity of microwave-treated protein hydrolysates after simulated gastrointestinal digestion: purification and identification Food Chem 2018 254 36 46 10.1016/j.foodchem.2018.01.133 29548465
33. Dong XH, Li J, Jiang GX, et al., Effects of combined high pressure and enzymatic treatments on physicochemical and antioxidant properties of peanut proteins. Food Sci Nutr. 2019;7(4):1417–1425.
34. Cui P Yang X Liang Q Ultrasound-assisted preparation of ACE inhibitory peptide from milk protein and establishment of its in-situ real-time infrared monitoring model Ultrason Sonochem 2020 62 104859 10.1016/j.ultsonch.2019.104859 31810869
35. Murtaza MA Irfan S Hafiz I Conventional and novel technologies in the production of dairy bioactive peptides Front Nutr 2022 9 780151 10.3389/fnut.2022.780151 35694165
36. Ulug SK Jahandideh F Wu J Novel technologies for the production of bioactive peptides Trends Food Sci Technol 2021 108 27 39 10.1016/j.tifs.2020.12.002
37. Urbizo-Reyes U, San Martin-González MF, Garcia-Bravo J, et al. Physicochemical characteristics of chia seed (Salvia hispanica) protein hydrolysates produced using ultrasonication followed by microwave-assisted hydrolysis. Food hydrocolloids. 2019;97:105187.
38. Rivas-Vela CI Amaya-Llano SL Castaño-Tostado E Protein hydrolysis by subcritical water: a new perspective on obtaining bioactive peptides Molecules 2021 26 21 6655 10.3390/molecules26216655 34771063
39. FitzGerald RJ, Cermeño M, Khalesi M, et al. Application of in silico approaches for the generation of milk protein-derived bioactive peptides. J Funct Foods. 2020;64:103636.
40. Vona R, Gambardella L, Cittadini C, et al. Biomarkers of oxidative stress in metabolic syndrome and associated diseases. Oxidative Med Cell Longevity. 2019;2019.
41. Monserrat-Mesquida M Quetglas-Llabrés M Capó X Metabolic syndrome is associated with oxidative stress and proinflammatory state Antioxidants 2020 9 3 236 10.3390/antiox9030236 32178436
42. Shazly AB, He Z, Abd El-Aziz M, et al. Fractionation and identification of novel antioxidant peptides from buffalo and bovine casein hydrolysates. Food Chem. 2017;232:753–762.
43. Shazly AB Mu H Liu Z Release of antioxidant peptides from buffalo and bovine caseins: influence of proteases on antioxidant capacities Food Chem 2019 274 261 267 10.1016/j.foodchem.2018.08.137 30372937
44. Abdel-Hamid M Otte J De Gobba C Angiotensin I-converting enzyme inhibitory activity and antioxidant capacity of bioactive peptides derived from enzymatic hydrolysis of buffalo milk proteins Int Dairy J 2017 66 91 98 10.1016/j.idairyj.2016.11.006
45. Srivastava U Nataraj BH Kumari M Antioxidant and immunomodulatory potency of Lacticaseibacillus rhamnosus NCDC24 fermented milk-derived peptides: a computationally guided in-vitro and ex-vivo investigation Peptides 2022 155 170843 10.1016/j.peptides.2022.170843 35878657
46. Cui L Yang G Lu S Antioxidant peptides derived from hydrolyzed milk proteins by Lactobacillus strains: a BIOPEP-UWM database-based analysis Food Res Int 2022 156 111339 10.1016/j.foodres.2022.111339 35651087
47. Tonolo F Folda A Cesaro L Milk-derived bioactive peptides exhibit antioxidant activity through the Keap1-Nrf2 signaling pathway Journal of Functional Foods 2020 64 103696 10.1016/j.jff.2019.103696
48. Tonolo F Fiorese F Moretto L Identification of new peptides from fermented milk showing antioxidant properties: mechanism of action Antioxidants 2020 9 2 117 10.3390/antiox9020117 32013158
49. Devi S Kumar N Kapila S Buffalo casein derived peptide can alleviates H2O2 induced cellular damage and necrosis in fibroblast cells Exp Toxicol Pathol 2017 69 7 485 495 10.1016/j.etp.2017.04.009 28483348
50. Kumar N Devi S Mada SB Anti-apoptotic effect of buffalo milk casein derived bioactive peptide by directing Nrf2 regulation in starving fibroblasts Food Biosci 2020 35 100566 10.1016/j.fbio.2020.100566
51. Mada SB Reddi S Kumar N Antioxidative peptide from milk exhibits antiosteopenic effects through inhibition of oxidative damage and bone-resorbing cytokines in ovariectomized rats Nutrition 2017 43 21 31 10.1016/j.nut.2017.06.010 28935141
52. Wen C Zhang J Zhang H Plant protein-derived antioxidant peptides: isolation, identification, mechanism of action and application in food systems: a review Trends Food Sci Technol 2020 105 308 322 10.1016/j.tifs.2020.09.019
53. Xu Q Xi H Chen X Milk-derived hexapeptide PGPIPN prevents and attenuates acute alcoholic liver injury in mice by reducing endoplasmic reticulum stress Int J Mol Med 2020 46 3 1107 1117 10.3892/ijmm.2020.4643 32705158
54. Gu Y Lambert JD Modulation of metabolic syndrome-related inflammation by cocoa Mol Nutr Food Res 2013 57 6 948 961 10.1002/mnfr.201200837 23637048
55. Reddy P Lent-Schochet D Ramakrishnan N Metabolic syndrome is an inflammatory disorder: a conspiracy between adipose tissue and phagocytes Clin Chim Acta 2019 496 35 44 10.1016/j.cca.2019.06.019 31229566
56. Song T Lv M Zhang L The protective effects of tripeptides VPP and IPP against small extracellular vesicles from angiotensin II-induced vascular smooth muscle cells mediating endothelial dysfunction in human umbilical vein endothelial cells J Agric Food Chem 2020 68 47 13730 13741 10.1021/acs.jafc.0c05698 33180478
57. Adams C Sawh F Green-Johnson J Characterization of casein-derived peptide bioactivity: differential effects on angiotensin-converting enzyme inhibition and cytokine and nitric oxide production J Dairy Sci 2020 103 7 5805 5815 10.3168/jds.2019-17976 32448573
58. Chakrabarti S Liao W Davidge ST Milk-derived tripeptides IPP (Ile-Pro-Pro) and VPP (Val-Pro-Pro) differentially modulate angiotensin II effects on vascular smooth muscle cells Journal of Functional Foods 2017 30 151 158 10.1016/j.jff.2016.12.022
59. Devi S Pal GK Kapila R C-terminal sequence deletion effect on antioxidative characteristics of VLPVPQK bioactive peptide from buffalo milk casein LWT 2020 119 108816 10.1016/j.lwt.2019.108816
60. Gu F, Xi H, Ruan X, et al. Protective effect of immunomodulating peptide (PGPIPN) derived from beta-casomorphin in bovine milk on acute alcohol-induced liver injury. Chinese Pharmacol Bullet. 2018;712–716.
61. Reyes-Díaz A Mata-Haro V Hernández J Milk fermented by specific lactobacillus strains regulates the serum levels of IL-6, TNF-α and IL-10 cytokines in a LPS-stimulated murine model Nutrients 2018 10 6 691 10.3390/nu10060691 29844254
62. Reddi S Mada SB Kumar N Antiosteopenic effect of buffalo milk casein-derived peptide (NAVPITPTL) in ovariectomized rats Int J Pept Res Ther 2019 25 3 1147 1158 10.1007/s10989-018-9763-0
63. Le TT Poulsen NA Kristiansen GH Quantitative LC-MS/MS analysis of high-value milk proteins in Danish Holstein cows Heliyon 2020 6 9 e04620 10.1016/j.heliyon.2020.e04620 32995587
64. Kumar N Reddi S Devi S Nrf2 dependent antiaging effect of milk-derived bioactive peptide in old fibroblasts J Cell Biochem 2019 120 6 9677 9691 10.1002/jcb.28246 30592315
65. Yamada A Sakurai T Ochi D Antihypertensive effect of the bovine casein-derived peptide Met-Lys-Pro Food Chem 2015 172 441 446 10.1016/j.foodchem.2014.09.098 25442576
66. Hsu C-N Tain Y-L Targeting the renin–angiotensin–aldosterone system to prevent hypertension and kidney disease of developmental origins Int J Mol Sci 2021 22 5 2298 10.3390/ijms22052298 33669059
67. Tu M Liu H Zhang R Analysis and evaluation of the inhibitory mechanism of a novel angiotensin-I-converting enzyme inhibitory peptide derived from casein hydrolysate J Agric Food Chem 2018 66 16 4139 4144 10.1021/acs.jafc.8b00732 29637780
68. Tu M Wang C Chen C Identification of a novel ACE-inhibitory peptide from casein and evaluation of the inhibitory mechanisms Food Chem 2018 256 98 104 10.1016/j.foodchem.2018.02.107 29606478
69. Sanchón J, Fernández-Tomé S, Miralles B, et al. Protein degradation and peptide release from milk proteins in human jejunum. Comparison with in vitro gastrointestinal simulation. Food Chem. 2018;239:486–494.
70. Li X Feng C Hong H Novel ACE inhibitory peptides derived from whey protein hydrolysates: identification and molecular docking analysis Food Biosci 2022 48 101737 10.1016/j.fbio.2022.101737
71. Shanmugam VP Kapila S Kemgang TS Isolation and characterization of angiotensin converting enzyme inhibitory peptide from buffalo casein Int J Pept Res Ther 2021 27 2 1481 1491 10.1007/s10989-021-10185-0
72. Chakrabarti S Jahandideh F Davidge ST Milk-derived tripeptides IPP (Ile-Pro-Pro) and VPP (Val-Pro-Pro) enhance insulin sensitivity and prevent insulin resistance in 3T3-F442A preadipocytes J Agric Food Chem 2018 66 39 10179 10187 10.1021/acs.jafc.8b02051 30160110
73. Daliri EB-M Lee BH Park B-J Antihypertensive peptides from whey proteins fermented by lactic acid bacteria Food science and biotechnology 2018 27 6 1781 1789 10.1007/s10068-018-0423-0 30483443
74. Li J, Zhao J, Wang X, et al. Novel angiotensin-converting enzyme-inhibitory peptides from fermented bovine milk started by Lactobacillus helveticus KLDS. 31 and Lactobacillus casei KLDS. 105: purification, identification, and interaction mechanisms. Front Microbiol. 2019;10:2643.
75. Koyama D Sasai M Matsumura S A milk-derived pentapeptide reduces blood pressure in advanced hypertension in a CCK system-dependent manner Food Funct 2020 11 11 9489 9494 10.1039/D0FO01122C 33064120
76. Xue L Wang X Hu Z Identification and characterization of an angiotensin-converting enzyme inhibitory peptide derived from bovine casein Peptides 2018 99 161 168 10.1016/j.peptides.2017.09.021 28987277
77. Sánchez-Rivera L, Ferreira Santos P, Sevilla MA, et al. Implication of opioid receptors in the antihypertensive effect of a bovine casein hydrolysate and αs1-casein-derived peptides. J Agric Food Chem. 2019;68(7):1877–1883.
78. Miyazaki H Nakamura T Ohki K Effects of the bioactive peptides Ile-Pro-Pro and Val-Pro-Pro upon autonomic neurotransmission and blood pressure in spontaneously hypertensive rats Auton Neurosci 2017 208 88 92 10.1016/j.autneu.2017.09.017 28988803
79. Tomiyama H Fujii M Shiina K Effects of lactotripeptide supplementation on tele-monitored home blood pressure and on vascular and renal function in prehypertension-randomized, double-blind, placebo-controlled, cross-over study- Circulation Reports 2019 1 10 438 444 10.1253/circrep.CR-19-0061 33693081
80. Cicero AF Colletti A Rosticci M Effect of lactotripeptides (Isoleucine–Proline–Proline/Valine–Proline–Proline) on blood pressure and arterial stiffness changes in subjects with suboptimal blood pressure control and metabolic syndrome: a double-blind, randomized, crossover clinical trial Metab Syndr Relat Disord 2016 14 3 161 166 10.1089/met.2015.0093 26683986
81. Antony P Vijayan R Bioactive peptides as potential nutraceuticals for diabetes therapy: a comprehensive review Int J Mol Sci 2021 22 16 9059 10.3390/ijms22169059 34445765
82. Kumar A, Kumar MH S, CS R, et al. Dipeptidyl peptidase‐IV inhibitory potential of alpha‐lactalbumin extracted from milk of Gir cows: a Bos indicus species. Int J Dairy Technol. 2022.
83. Zhao Q Wei G Li K Identification and molecular docking of novel α-glucosidase inhibitory peptides from hydrolysates of Binglangjiang buffalo casein LWT 2022 156 113062 10.1016/j.lwt.2021.113062
84. Lacroix IM Li-Chan EC Isolation and characterization of peptides with dipeptidyl peptidase-IV inhibitory activity from pepsin-treated bovine whey proteins Peptides 2014 54 39 48 10.1016/j.peptides.2014.01.002 24440459
85. Gao J Guo K Du M Bovine α-lactalbumin-derived peptides attenuate TNF-α-induced insulin resistance and inflammation in 3T3-L1 adipocytes through inhibiting JNK and NF-κB signaling Food Funct 2022 13 4 2323 2335 10.1039/D1FO01217G 35142310
86. Zheng L Xu Q Lin L In vitro metabolic stability of a casein-derived dipeptidyl peptidase-IV (DPP-IV) inhibitory peptide VPYPQ and its controlled release from casein by enzymatic hydrolysis J Agric Food Chem 2019 67 38 10604 10613 10.1021/acs.jafc.9b03164 31466448
87. Nongonierma AB FitzGerald RJ An in silico model to predict the potential of dietary proteins as sources of dipeptidyl peptidase IV (DPP-IV) inhibitory peptides Food Chem 2014 165 489 498 10.1016/j.foodchem.2014.05.090 25038703
88. Rendon-Rosales MA Torres-Llanez MJ Mazorra-Manzano MA In vitro and in silico evaluation of multifunctional properties of bioactive synthetic peptides identified in milk fermented with Lactococcus lactis NRRL B-50571 and NRRL B-50572 LWT 2022 154 112581 10.1016/j.lwt.2021.112581
89. Iwasa M Takezoe S Kitaura N A milk casein hydrolysate-derived peptide enhances glucose uptake through the AMP-activated protein kinase signalling pathway in skeletal muscle cells Exp Physiol 2021 106 2 496 505 10.1113/EP088770 33369793
90. Shimizu Y Hara H Hira T Glucagon-like peptide-1 response to whey protein is less diminished by dipeptidyl peptidase-4 in comparison with responses to dextrin, a lipid and casein in rats Br J Nutr 2021 125 4 398 407 10.1017/S0007114520002834 32713353
91. Huang R, Lu Y, Xie Z, et al. A bovine milk-derived peptide ameliorates alloxan-injured pancreatic β cells through IRS2/PI3K/Akt signaling. Life Sci. 2022;120907.
92. Mudgil P Baba WN Kamal H A comparative investigation into novel cholesterol esterase and pancreatic lipase inhibitory peptides from cow and camel casein hydrolysates generated upon enzymatic hydrolysis and in-vitro digestion Food Chem 2022 367 130661 10.1016/j.foodchem.2021.130661 34348197
93. Chen H, Qi X, Guan K, et al. Peptides released from bovine α-lactalbumin by simulated digestion alleviated free fatty acids-induced lipid accumulation in HepG2 cells. J Funct Foods. 2021;85:104618.
94. Lee S Youn B Hypolipidemic roles of casein-derived peptides by regulation of trans-intestinal cholesterol excretion and bile acid synthesis Nutrients 2020 12 10 3058 10.3390/nu12103058 33036208
95. Jiang X Pan D Zhang T Novel milk casein–derived peptides decrease cholesterol micellar solubility and cholesterol intestinal absorption in Caco-2 cells J Dairy Sci 2020 103 5 3924 3936 10.3168/jds.2019-17586 32113776
96. Qi X Chen H Guan K Novel xanthine oxidase inhibitory peptides derived from whey protein: identification, in vitro inhibition mechanism and in vivo activity validation Bioorg Chem 2022 128 106097 10.1016/j.bioorg.2022.106097 35985156
97. Aziz N, Jamil RT. Biochemistry, xanthine oxidase. 2019.
98. Vij R Reddi S Kapila S Transepithelial transport of milk derived bioactive peptide VLPVPQK Food Chem 2016 190 681 688 10.1016/j.foodchem.2015.05.121 26213026
99. Lacroix IM Chen X-M Kitts DD Investigation into the bioavailability of milk protein-derived peptides with dipeptidyl-peptidase IV inhibitory activity using Caco-2 cell monolayers Food Funct 2017 8 2 701 709 10.1039/C6FO01411A 28098291
100. Tianrui Z, Bingtong L, Ling Y, et al. ACE inhibitory activity in vitro and antihypertensive effect in vivo of LSGYGP and its transepithelial transport by Caco-2 cell monolayer. J Funct Foods. 2019;61:103488.
101. Xia Y Yu J Xu W Purification and characterization of angiotensin-I-converting enzyme inhibitory peptides isolated from whey proteins of milk fermented with Lactobacillus plantarum QS670 J Dairy Sci 2020 103 6 4919 4928 10.3168/jds.2019-17594 32229119
102. Yao X Bunt C Cornish J Oral delivery of bovine lactoferrin using pectin-and chitosan-modified liposomes and solid lipid particles: improvement of stability of lactoferrin Chem Biol Drug Des 2015 86 4 466 475 10.1111/cbdd.12509 25581616
103. Rezvani M Hesari J Peighambardoust SH Potential application of nanovesicles (niosomes and liposomes) for fortification of functional beverages with Isoleucine-Proline-Proline: a comparative study with central composite design approach Food Chem 2019 293 368 377 10.1016/j.foodchem.2019.05.015 31151624
104. Patange SR Sabikhi L Shelke PA Encapsulation of dipeptidyl peptidase-IV inhibitory peptides from alpha-lactalbumin extracted from milk of Gir cows–a Bos indicus species Int J Dairy Technol 2022 75 3 575 587 10.1111/1471-0307.12882
105. Patel BY, Volcheck GW. Food allergy: common causes, diagnosis, and treatment. In Mayo Clinic Proceedings. 2015. Elsevier.
106. • Caira S, Pinto G, Picariello G, et al. In vivo absorptomics: identification of bovine milk-derived peptides in human plasma after milk intake. Food Chemistry, 2022;385:132663. The paper eloquently described the identification procedure of the milk-derived peptides absorbed and circulating in the plasma of the human blood after ingestion.
107. Shivanna SK Nataraj BH Revisiting therapeutic and toxicological fingerprints of milk-derived bioactive peptides: an overview Food Biosci 2020 38 100771 10.1016/j.fbio.2020.100771
108. Beltrán-Barrientos LM García HS Torres-Llanez MJ Safety of milk-derived bioactive peptides Int J Dairy Technol 2017 70 1 16 22 10.1111/1471-0307.12338
109. Mendes O Chen J Gowda R Safety assessment of herbal food ingredients and nutraceuticals Nutraceutical and Functional Food Regulations in the United States and around the World 2019 Elsevier 107 120
110. Pradhan D Mallappa RH Grover S Comprehensive approaches for assessing the safety of probiotic bacteria Food Control 2020 108 106872 10.1016/j.foodcont.2019.106872
111. Kiewiet MBG van Esch BCAM Garssen J Partially hydrolyzed whey proteins prevent clinical symptoms in a cow’s milk allergy mouse model and enhance regulatory T and B cell frequencies Mol Nutr Food Res 2017 61 11 1700340 10.1002/mnfr.201700340
112. FAO. Dairy market review - overview of global dairy market developments in 2020. 2021; Available from: https://www.fao.org/markets-and-trade/publications/detail/en/c/1437811/.
113. Ali MA Kamal MM Rahman MH Functional dairy products as a source of bioactive peptides and probiotics: current trends and future prospectives J Food Sci Technol 2022 59 4 1263 1279 10.1007/s13197-021-05091-8 35250052
114. Zhao X, Cui YG, Bai SS, et al. Antioxidant activity of novel casein-derived peptides with microbial proteases as characterized via Keap1-Nrf2 pathway in HepG2 cells. 2021.
|
PMC010xxxxxx/PMC10198384.txt |
==== Front
J Infect
J Infect
The Journal of Infection
0163-4453
1532-2742
The British Infection Association. Published by Elsevier Ltd.
S0163-4453(23)00258-X
10.1016/j.jinf.2023.04.022
Letter to the Editor
Causal relationships between short-chain fatty acids and L-isoleucine biosynthesis and susceptibility and severity of COVID-19: Evidence from Mendelian randomization
Lv Wanqiang abc
He Jia abc
Shao Jinjin abc
Chen Yunxiang abc
Xia Lijuan abc
Zhang Lijiang abc⁎
a Center of Safety Evaluation and Research, Hangzhou Medical College, Hangzhou, Zhejiang, China
b Key Laboratory of Drug Safety Evaluation and Research of Zhejiang Province, Hangzhou Medical College, Hangzhou, Zhejiang, China
c Engineering Research Center of Novel Vaccine of Zhejiang Province, Hangzhou Medical College, Hangzhou, Zhejiang, China
⁎ Correspondence to: Center of Safety Evaluation and Research, Hangzhou Medical College, China.
5 5 2023
7 2023
5 5 2023
87 1 e16e18
26 4 2023
© 2023 The British Infection Association. Published by Elsevier Ltd. All rights reserved.
2023
The British Infection Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcRecently, we read a paper by Qiu and colleagues with interest, who reported potential causal candidates for COVID-19 in 123 blood metabolites by Mendelian randomization (MR).1 In Journal of Infection and other journals, the association between altered gut microbiota composition and Coronavirus disease 2019 (COVID-19) has been broadly discussed, whereas such association was based on observational studies without clear cause or consequence effect established, likely to be potentially biased by confounders. We extremely appreciated the association between blood metabolites and COVID-19 by using MR, we attempted to explore whether the changes to short-chain fatty acids (SCFAs) and L-isoleucine biosynthesis directly affect COVID-19 susceptibility and severity by MR analysis in the present study, where gut bacteria involved in the metabolism of SCFAs and L-isoleucine were depleted and correlated with disease severity.2
MR is a novel and promising approach utilizing genetic variants as instrumental variables (IVs) to evaluate the genetic associations between two phenotypes which are generally unconfounded ( Fig. 1A), indicating it can be the promising methodology to infer potential causality between the gut microbiome and COVID-19. MR analysis was conducted using the TwoSampleMR R package, which is created and provided by MR-Base (www.mrbase.org/).3 The main MR result was calculated by using the inverse variance weighted (IVW) method based on a random-effects model. MR simple median, weighted median, and MR-Egger methods were applied for sensitivity analyses, and we further examined whether some SNPs could influence the results independently via leave-one-out analysis. Examination of pleiotropic effects was employed by MR-Egger regression analysis. Heterogeneity between estimates from contributing studies was tested using Cochran's Q test in the IVW method. Several features related to SCFAs and L-isoleucine biosynthesis were selected as the exposures, including gut production of the SCFA butyrate, fecal propionate levels (gut production or absorption of the SCFA propionate), and plasma L-isoleucine levels. For microbial related exposures, SNPs that were significantly associated (p < 1 ×10−5) with any of these features were used as IVs in the GWAS results of the gut microbiome from the independent population of 18,340 individuals.4, 5 This approach to increase the number of IVs has been demonstrated by previous studies to be a viable methodology in microbiome MR.4, 5 For the IVs of L-isoleucine, SNPs that were significantly associated (p < 5 ×10−8) with plasma L-isoleucine were selected in the GWAS results from the independent population of 24,925 individuals. Detailed information of each exposure-specific single nucleotide polymorphism selected for MR analyses was presented in Supplementary Table 1. For the outcomes, our study was based on the GWAS results of 14,134 COVID-19 cases and 1,284,876 controls, conducted by the Host Genetics Initiative consortium with severe COVID-19 disease for the following outcomes: SARS-CoV-2 infection, hospitalized COVID-19, critically ill COVID-19.6 Fig. 1 Causal relationships between short-chain fatty acids and L-isoleucine biosynthesis and COVID-19. A: Schematic overview of Mendelian randomization design. B: Forest plot represented the effect of per 1-SD increase in gut microbial synthesis of the SCFA butyrate abundance on susceptibility and severity of COVID-19.
Fig. 1
We used an online tool (https://sb452.shinyapps.io/power/) to calculate power, our study was sufficiently powered (>80%) for MR analyses based on the sample size of study population and the variance explained (r 2) for the features. Detailed information is presented in Supplemental Table 2. The results of all tested features were presented in Supplementary Table 3. Our results indicated that host-genetic-driven increase in gut production of the SCFA butyrate was correlated with a lower risk of SARS-CoV-2 infection (odds ratio (OR) = 0.96, 95% confidence interval (CI) 0.923–0.999, P = 0.044), and hospitalized COVID-19 (OR = 0.968, 95%CI 0.944–0.993, P = 0.013) based on the IVW analysis (Fig. 1B). A trend of lower risk of SARS-CoV-2 infection (OR = 0.980, 95%CI 0.859–1.116, P = 0.757), hospitalized COVID-19 (OR = 0.941, 95%CI 0.808–1.097, P = 0.439), and critically ill COVID-19 (OR = 0.969, 95%CI 0.858–1.094, P = 0.609) was also observed in patients with increase in fecal propionate whereas without statistical significance. However, a trend of higher risk of SARS-CoV-2 infection (OR = 1.123, 95%CI 0.982–1.284, P = 0.090), hospitalized COVID-19 (OR = 1.163, 95%CI 0.888–1.525, P = 0.273), and a trend of lower risk of critically ill COVID-19 (OR = 0.858, 95%CI 0.655–1.125, P = 0.269) was observed in patients with increase in plasma L-isoleucine. These observed associations were directionally consistent with MR sensitivity analyses, including weighted median, simple median, and MR-Egger methods (Supplementary Table 3). However, MR-Egger causal estimates between fecal propionate, plasma L-isoleucine and COVID-19 diverged from the IVW method, it may had been due to poor precision.7 Overall, these estimates were similar in terms of direction and magnitude, and they were unlikely to have happened by chance alone. Neither the pleiotropy nor the heterogeneity was observed (Supplementary Tables 4 and 5). Leave-one-SNP-out sensitivity analysis showed that none of the SNPs used is driving the association (Supplementary Table 6).
To our knowledge, we first found that host-genetic-driven increase in gut production of SCFA butyrate was associated with lower risk of SARS-CoV-2 infection and hospitalized COVID-19 using observational data from large-scale GWASs, whereas the causal estimates of fecal propionate, plasma L-isoleucine on COVID-19 may be due to insufficient statistical power, which requires further investigation. Our results may help inform clinical decision-making and the development of novel intervention for preventing COVID-19. Common chronic diseases, like sarcopenia and type 2 diabetes mellitus, which also displayed impaired capacity for SCFA,8, 9 may have been underestimated to be high-risk factors of COVID-19. Early detecting patients susceptible to developing critical illness is vital and may help optimize the usage of restricted medical resources under the pandemic of COVID-19. This could be due to the limited power of suitable IVs available for MR analyses. On the other hand, the GWAS result of plasma SCFAs levels was not available in published cohorts, further studies are warranted to explore the relationship between plasma SCFAs levels and COVID-19.
To conclude, our study provided evidence for a causal effect of biosynthesis of SCFA butyrate on COVID-19 susceptibility. Further investigation is warranted to explore SCFAs as a potential intervention approach for preventing individuals with susceptibility of COVID-19.
CRediT authorship contribution statement
Wanqiang Lv: Formal analysis, Writing – original draft, Writing – review & editing. Jia He: Formal analysis, Writing – review & editing. Jinjin Shao: Writing – review & editing. Yunxiang Chen: Writing – review & editing. Lijuan Xia: Writing – review & editing. Lijiang Zhang: Project conceiving, designing and initiating, Writing – review & editing.
Declaration of Competing Interest
None declared.
Appendix A Supplementary material
Supplementary material
.
Acknowledgments
This research was funded by National Key R&D Program of China (2018YFA0903200), Basic Research Project of Hangzhou Medical College (KYYB202201), Key R&D Program of Zhejiang Province (2021C03077), Zhejiang Medical and Health Science and Technology Plan (WKJ-ZJ-2203), and the Central Leading Local Science and Technology Development Fund Project (2023ZY1019). We thank the COVID-19 Host Genetics Initiative and other consortium for making GWAS summary statistics publicly available. The summary GWASs data used in our study could be downloaded in the MR-Base platform for researchers (www.mrbase.org/).
Appendix A Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jinf.2023.04.022.
==== Refs
References
1 Qiu S. Wang D. Zhang Y. Hu Y. Mendelian randomization reveals potential causal candidates for COVID-19 in 123 blood metabolites J Infect 84 2 2022 248 288
2 Zhang F. Wan Y. Zuo T. Yeoh Y.K. Liu Q. Zhang L. Prolonged impairment of short-chain fatty acid and L-isoleucine biosynthesis in gut microbiome in patients with COVID-19 Gastroenterology 162 2 2022 548 561 e4 34687739
3 Hemani G. Zheng J. Elsworth B. Wade K.H. Haberland V. Baird D. The MR-Base platform supports systematic causal inference across the human phenome eLife 7 2018
4 Kurilshikov A. Medina-Gomez C. Bacigalupe R. Radjabzadeh D. Wang J. Demirkan A. Large-scale association analyses identify host factors influencing human gut microbiome composition Nat Genet 53 2 2021 156 165 33462485
5 Sanna S. van Zuydam N.R. Mahajan A. Kurilshikov A. Vich Vila A. Vosa U. Causal relationships among the gut microbiome, short-chain fatty acids and metabolic diseases Nat Genet 51 4 2019 600 605 30778224
6 Initiative C.-H.G. The COVID-19 host genetics initiative, a global initiative to elucidate the role of host genetic factors in susceptibility and severity of the SARS-CoV-2 virus pandemic Eur J Hum Genet 28 6 2020 715 718 32404885
7 Burgess S. Thompson S.G. Interpreting findings from Mendelian randomization using the MR-Egger method Eur J Epidemiol 32 5 2017 377 389 28527048
8 Chen X. Wu Q. Gao X. Wang H. Zhu J. Xia G. Gut microbial dysbiosis associated with type 2 diabetes aggravates acute ischemic stroke mSystems 6 6 2021 e0130421
9 Lv W.Q. Lin X. Shen H. Liu H.M. Qiu X. Li B.Y. Human gut microbiome impacts skeletal muscle mass via gut microbial synthesis of the short-chain fatty acid butyrate among healthy menopausal women J Cachexia Sarcopenia Muscle 12 6 2021 1860 1870 34472211
|
PMC010xxxxxx/PMC10198790.txt |
==== Front
Stat Biosci
Stat Biosci
Statistics in Biosciences
1867-1764
1867-1772
Springer US New York
9372
10.1007/s12561-023-09372-y
Article
Analysis of the Cox Model with Longitudinal Covariates with Measurement Errors and Partly Interval Censored Failure Times, with Application to an AIDS Clinical Trial
Sun Yanqing 1
http://orcid.org/0000-0002-9499-9040
Zhou Qingning qzhou8@uncc.edu
1
Gilbert Peter B. 2
1 grid.266859.6 0000 0000 8598 2218 Department of Mathematics and Statistics, University of North Carolina at Charlotte, Charlotte, NC USA
2 grid.270240.3 0000 0001 2180 1622 Vaccine and Infectious Disease and Public Health Sciences Divisions, Fred Hutchinson Cancer Center, Seattle, WA USA
20 5 2023
2023
15 2 430454
13 5 2022
18 4 2023
27 4 2023
© The Author(s) under exclusive licence to International Chinese Statistical Association 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Time-dependent covariates are often measured intermittently and with measurement errors. Motivated by the AIDS Clinical Trials Group (ACTG) 175 trial, this paper develops statistical inferences for the Cox model for partly interval censored failure times and longitudinal covariates with measurement errors. The conditional score methods developed for the Cox model with measurement errors and right censored data are no longer applicable to interval censored data. Assuming an additive measurement error model for a longitudinal covariate, we propose a nonparametric maximum likelihood estimation approach by deriving the measurement error induced hazard model that shows the attenuating effect of using the plug-in estimate for the true underlying longitudinal covariate. An EM algorithm is devised to facilitate maximum likelihood estimation that accounts for the partly interval censored failure times. The proposed methods can accommodate different numbers of replicates for different individuals and at different times. Simulation studies show that the proposed methods perform well with satisfactory finite-sample performances and that the naive methods ignoring measurement error or using the plug-in estimate can yield large biases. A hypothesis testing procedure for the measurement error model is proposed. The proposed methods are applied to the ACTG 175 trial to assess the associations of treatment arm and time-dependent CD4 cell count on the composite clinical endpoint of AIDS or death.
Supplementary Information
The online version contains supplementary material available at 10.1007/s12561-023-09372-y.
Keywords
AIDS clinical trial
Cox model
Longitudinal covariates
Measurement errors
Partly interval censored data
http://dx.doi.org/10.13039/100000001 National Science Foundation DMS1916170 DMS1915829 Sun Yanqing Zhou Qingning National Institutes of HealthR37AI054165 issue-copyright-statement© International Chinese Statistical Association 2023
==== Body
pmcIntroduction
This work is motivated by the AIDS Clinical Trials Group (ACTG) 175 trial that compared four antiretroviral regimens in study participants living with HIV-1 [1]. CD4 cell count has long been considered as an important prognostic biomarker for disease progression. The participants had CD4 cell count measured every 12 weeks and were followed for occurrence of the composite clinical endpoint of AIDS or death. One of the study objectives was to assess the associations of treatments and time-dependent CD4 cell count with the clinical endpoint of AIDS or death. One complication is that CD4 cell count is measured intermittently and with measurement errors. Naive approaches that ignore the measurement errors or replace them with their estimated values can lead to biased estimation [2, 3]. The other challenge is that the time to the clinical endpoint is partly interval censored in which the time to death is subject to right censoring while the time to AIDS is interval censored between two visit dates. Statistical modeling with interval censored data has been well studied. There has also been extensive study of measurement errors in covariates for right censored data. However, to the best of our knowledge, no methodology exists for achieving valid statistical inference for the proportional hazards model with both partly interval censored failure times and longitudinal covariates subject to measurement error.
Many authors have studied regression analysis of interval censored failure time data under the Cox proportional hazards model [4–8]. Most of the existing work focused on time-independent covariates. Recently, Zeng et al. [9] considered maximum likelihood estimation for a class of semiparametric transformation models that includes the proportional hazards model and allows for time-dependent covariates. On the other hand, research on partly interval censored failure time data is fairly limited. Due to the presence of exact failure times, partly interval censored data requires a different treatment than interval censored data. To our knowledge, only two papers considered the proportional hazards model for partly interval censored data. In particular, Kim [10] studied maximum likelihood estimation for the proportional hazards model with time-independent covariates, while Zhou et al. [11] developed an EM algorithm for nonparametric maximum likelihood estimation for a class of semiparametric transformation models as in Zeng et al. [9] that allows for time-dependent covariates for partly interval censored data. However, these two papers did not consider covariate measurement errors.
There is extensive literature on statistical methods for right censored failure time data when some covariates are subject to measurement error. It is well known that standard estimation procedures yield biased estimation if measurement error is not taken into account. For the proportional hazards model, Prentice [2] showed that the naive approach using the observed covariate values with measurement errors in place of the underlying covariate values in the partial likelihood may result in substantial estimation bias, and proposed a modified partial likelihood method that required estimating the conditional expectation of the hazard of each individual at each failure time. Nakamura [12] proposed the corrected partial score method that yielded approximately unbiased estimates. Hughes [13] investigated regression dilution bias in the presence of covariate measurement errors. Buzas [14] removed the condition of normal errors while assuming that the moment generating function of the error distribution is known. Huang and Wang [15] proposed a nonparametric-correction approach for the Cox proportional hazards model. Hu and Lin [16] extended the work of Nakamura [12] and Buzas [14] to obtain a class of consistent estimators when the true covariate is ascertained on a randomly selected validation set. Tsiatis and Davidian [3] proposed the conditional score estimator, which was further studied by Song et al. [17, 18]. Yi and Lawless [19] employed the corrected score methods of Nakamura [12] assuming a piecewise constant form of the baseline hazard function. Fu and Gilbert [20] extended the conditional score approach to accommodate missing values of the longitudinal covariates following a two-phase sampling design. Tsiatis and Davidian [21] overviewed joint modeling of longitudinal covariates and time-to-event data.
All of the aforementioned works considered right censored data. Although both measurement error problems and interval censored data have been well studied, the literature on statistical methods for interval censored failure time data with covariate measurement error is rather limited. Song and Ma [22] proposed a multiple imputation method for the Cox model with time-independent covariates to impute the time-to-event that falls within an interval and then analyzed the imputed data sets by the conditional score approach for right censored data. Mandal et al. [23] applied multiple imputation to handle both covariates with measurement error and interval censored failure time data under the linear transformation model. The imputed data were then analyzed using the method of Chen et al. [24]. Wen and Chen [25] proposed a conditional score approach for the proportional odds model with interval censoring and covariate measurement error using the working independence strategy. All of these approaches were developed assuming time-independent covariates with measurement errors. The multiple imputation approach depends on the imputation models and can only be approximate. The conditional score methods proposed by Tsiatis and Davidian [3] for right censored data and studied by many others [17, 18, 20] are no longer applicable for interval censored data.
In this article, we develop an estimation method for the Cox proportional hazards model for partly interval censored failure times and longitudinal covariates measured with error. Assuming an additive measurement error model for a longitudinal covariate, we propose a nonparametric maximum likelihood estimation approach by deriving the measurement error induced hazard model that shows the attenuating effect of ignoring measurement errors. An EM algorithm is devised to facilitate maximum likelihood estimation that accounts for the partly interval censored failure times. Simulation studies show that the proposed methods perform well with satisfactory finite-sample performances and that the naive methods ignoring measurement error or using the plug-in estimate can yield large biases. The simulation studies also show the attenuating bias of using the plug-in estimate for the true underlying longitudinal covariate. While the commonly used additive measurement error model for a time-independent covariate can be checked and often holds well in practice, use of a measurement error model for time-varying covariates requires more care. Additive random effects models with known time-dependent basis functions are commonly used, but misspecification may lead to bias. Although statistical models of longitudinal covariates measured with error have been studied by many authors (e.g., papers noted above), few methods are available to evaluate their goodness-of-fit. In this article, we also propose a diagnostic testing procedure for the measurement error model of longitudinal covariates.
The rest of this article is organized as follows. Section 2 introduces the data structure, models and model assumptions. Section 3.1 derives the measurement error induced hazard model. Section 3.2 presents a nonparametric maximum likelihood estimation approach. An EM algorithm is devised to facilitate maximum likelihood estimation that accounts for the partly interval censored failure times. Section 3.3 derives the variance estimator based on the profile likelihood that accounts for variation in the parameter estimation for the measurement error model. A test procedure for the measurement error model of longitudinal covariates is given in Section 4. The finite-sample performance of the proposed methods is examined through simulation studies in Section 5. The proposed methods are applied to the ACTG 175 trial data in Section 6. Some concluding remarks are given in Section 7.
Preliminaries
Suppose Ti is the failure time of interest with the end of follow-up time τ. Let Zi be the d×1 vector of time-independent covariates that includes baseline covariates and treatment assignment, and Xi(t) the time-dependent covariate of interest. Let X¯i(t)={Xi(u),0≤u≤t} denote the history of Xi(·) up to time t. We assume that the conditional hazard function of Ti given X¯i(t) and Zi only depends on Zi and the current value Xi(t). Let λ(t|X¯i(t),Zi) be the conditional hazard function of Ti given X¯i(t) and Zi. We consider the proportional hazards model1 λ(t|X¯i(t),Zi)=λ(t)exp{βXi(t)+γTZi},
for 0≤t≤τ, where λ(t) is an unspecified baseline function, and β and γ are 1- and d-dimensional vectors of parameters, respectively. We investigate model (1) under partly interval censored failure time data and when the time-dependent covariate Xi(t) is subject to measurement error.
Partly interval censored failure time data include observations of failure times that are precisely observed, and failure times that are left, interval and/or right censored. Let Δi indicate whether the failure time Ti is exactly observed, i.e., Δi=1 if Ti is exactly observed and 0 otherwise. If Δi=0, let (Li,Ri] denote the smallest observed interval that brackets Ti, where Li≥0 is the last monitoring time at which failure has not occurred and Ri≥0 is the first monitoring time at which failure has occurred. Let Ri=∞ if failure has not occurred by the last monitoring time. Thus, if Li=0, Ti is left censored; if Ri=∞, Ti is right censored; if 0<Li<Ri<∞, Ti is interval censored. The partly interval censored failure time data for individual i can be represented by {(Δi,ΔiTi,(1-Δi)Li,(1-Δi)Ri}. The notations ΔiTi, (1-Δi)Li and (1-Δi)Ri mean that we observe Ti if Δi=1 and observe (Li,Ri] if Δi=0.
In the ACTG 175 study, the failure time of interest is the time to compositie endpoint of AIDS or death, whichever occurs first. For individual i, if death has occurred before AIDS, then we observe the exact death time Ti and Δi=1; if AIDS has occurred prior to death, then we observe a time interval (Li,Ri] that brackets the AIDS onset time Ti and Δi=0.
Linear mixed effects models are commonly used to model longitudinal covariates measured with errors [3, 17, 20]. Suppose that Xi(t) is measured at times vi1<⋯<vi,Mi before τ with errors and there are Bij repeated measurements or replicates of Xi(vij), where we let Bij=1 if there are no replicates. Let Wi,b(vij) denote the bth measurement of Xi(·) at time vij, j=1,…,Mi, b=1,…,Bij. We consider the linear mixed effects model for longitudinal covariates with measurement errors:2 Wi,b(vij)=Xi(vij)+eij,b=θiTf(vij)+eij,b,
where f(vij) is an r×1 vector of known design functions, θi is an r×1 vector of unobserved random effects, and eij,b is the measurement error at time vij. We assume θi=ϑ+νi, where ϑ is a vector of fixed parameters and νi (i=1,…,n) are independent and identically distributed (iid) N(0, G) with G being a r×r nonnegative definite matrix. We also assume that eij,b (j=1,…,Mi, b=1,…,Bij) are iid N(0,σ2) independent of νi. Thus, the unknown parameters for the measurement error model are θW=(ϑ,G,σ2). Also, note that the design function f(·) is usually chosen as a vector of basis functions, such as polynomials. In our simulation study and real data analysis below, we consider f(t)=(1,t) or (1,t,t2).
Define Wij=(Wi,1(vij),…,Wi,Bij(vij)) and eij=(eij,1,…,eij,Bij). Let v~i=(vi1,…,vi,Mi)T, W~i=(Wi1,…,Wi,Mi)T and e~i=(ei1,…,ei,Mi)T. The observed data consist of a random sample of n iid observations{Δi,ΔiTi,(1-Δi)Li,(1-Δi)Ri,Zi,v~i,W~i},i=1,…,n.
We will employ individual-specific estimation of the longitudinal covariate Xi(t) via model (2). It does not require repeated measurements at each measurement time vij, as long as the number of longitudinal measurements over time is sufficient to estimate θi, i.e., Mi≥r. The proposed estimation method allows Bij=1 for all i, j. However, the repeated measurements reduce the standard error in estimating θi and thus in estimating Xi(t), which results in increased efficiency in estimating β for model (1).
Estimation of the Cox Model with Partly Interval Censored Failure Times and Longitudinal Covariates with Measurement Errors
In this section, we propose a method for estimation of the Cox model (1). In Sect. 3.1, we derive the measurement error induced hazard model under the additive measurement error model for longitudinal covariates. In Sect. 3.2, we design an EM algorithm for the nonparametric maximum likelihood estimation of the measurement error induced hazard model based on partly interval censored failure times. A variance estimation procedure is proposed in Sect. 3.3.
Measurement Error Induced Hazard Model
The true longitudinal covariate Xi(t) is not observed. We obtain an individual-specific estimate X^i(t) of Xi(t) using the ordinary least squares method based on the observed data (v~i,W~i) and propose an approach by deriving the conditional hazard function of Ti at time t conditional on Zi and X^i(t). Only the longitudinal covariates in the past can be meaningfully used to model current or future risk of failure. For example, in assessing the association of time-dependent CD4 cell count with the composite clinical endpoint of AIDS or death in the ACTG 175 trial, only the CD4 count measurements before AIDS or death are meaningfully associated with the endpoint. Therefore, we estimate Xi(t) based on the data before t to preserve the predictability [3, 20].
Let Mi(t) denote the index of the last measurement time before t such that vi,Mi(t)<t≤vi,Mi(t)+1. Since θi is r-dimensional, at least r longitudinal measurements from individual i before t are required, i.e., Mi(t)≥r. Let v~i(t)=(vi1,…,vi,Mi(t))T, W~i(t)=(Wi1,…,Wi,Mi(t))T and e~i(t)=(ei1,…,ei,Mi(t))T. Under model (2), W~i(t)=F~i(t)θi+e~i(t), where F~i(t)=Bi(t)f~i(t)T with Bi(t)=diag(1Bi1,…,1Bi,Mi(t)), 1m is a m×1-vector of ones, and f~i(t)=(f(vi1),…,f(vi,Mi(t))). Hence the ordinary least squares estimator of θi based on (v~i(t),W~i(t)) for individual i equals3 θ^i(t)={F~iT(t)F~i(t)}-1F~iT(t)W~i(t).
It is easy to see that F~iT(t)F~i(t)=∑j=1Mi(t)Bijfi(vij)fiT(vij) and F~iT(t)W~i(t)=∑j=1Mi(t)fi(vij)∑b=1BijWi,b(vij).
We estimate θi based on the observations from subject i without pulling information from other individuals because only the past history of subject i can forecast his/her risk of failure. The longitudinal covariate Xi(t) is estimated by X^i(t)=fT(t)θ^i(t) based on the observed error-prone covariate information for individual i up to time t. This allows us to derive the measurement error induced hazard model conditional on the observed information from individual i’s past.
Since θ^i(t)=θi+{F~iT(t)F~i(t)}-1F~iT(t)e~i(t), we haveX^i(t)=Xi(t)+fT(t){F~iT(t)F~i(t)}-1F~iT(t)e~i(t).
The two terms Xi(t) and e~i(t) are independent under model (2). Then conditional on (θi,v~i(t)), X^i(t) is normally distributed with mean Xi(t) and variance di(t,σ2)=σ2fT(t){F~iT(t)F~i(t)}-1f(t). An estimator of σ2 can be constructed using the residuals:4 σ^2=n-1∑i=1nMi-1∑j=1MiBij-1∑b=1Bij(Wi,b(vij)-X^i(vij))2.
Next, we derive the induced hazard model of Ti conditional on Zi and X^i(t) under the measurement error model (2). The estimator X^i(t) is based on the observed information before t, and thus is predictable for the risk of failure at t. Define the counting process increment dNi(t)=I(t≤Ti<t+dt,vir≤t) and the at-risk process Yi(t)=I(Ti≥t,vir≤t). That is, dNi(t)=1 if the failure time occurs at time t and after the rth longitudinal measurement.
Our approach is motivated by the conditional score method of [3]. They first derived the conditional likelihood of {dNi∗(t),X^i(t)} given (θi,Zi,v~i(t),Yi(t)=1), where Ni∗(t) is the counting process for the right censored data. They then noted that the conditional likelihood given Yi(t)=1, Qi(t,β,σ2)=X^i(t)+di(t,σ2)βdNi∗(t), is a complete sufficient statistic for θi, and thus, conditional on Qi(t,β,σ2), removes the dependence of the conditional distribution on the random effects θi. [3, 20] derived the conditional intensity process by conditioning on Qi(t,β,σ2), which turns out to be a Cox model with Zi and Qi(t,β,σ2) as the independent variables. We note that their papers did not derive the intensity model because of dNi∗(t) involved in Qi(t,β,σ2). Further, their approaches do not work in the current setting because the counting process framework can not be utilized for interval censored or partly interval censored data.
We pursue a different approach by deriving the hazard model for Ti conditional on Zi,v~i(t), Yi(t)=1 and X^i(t). Let Ft be the filtration generated by {Ni(s),Yi(s),Zi,Xi(s),X^i(s),v~i(s), W~i(s)}, 0≤s≤t. Then X^i(t) and σi,rel2(t) are both predictable with respect to Ft. The following proposition presents the conditional hazard function of Ti at time t given (X^i(t),Zi,v~i(t), vir≤t).
Proposition 1
Under Conditions (A1)-(A3) given in the Appendix,5 λ∗(t|X^i(t),Zi,v~i(t))=λ(t)exp{βωi(t)X^i(t)+γTZi+Oi(β,t,θW)},fort≥vir,
where ωi(t)=1-σi,rel2(t), Oi(β,t,θW)=βσi,rel2(t)[ϑTf(t)+12fT(t)Gf(t)β], σi,rel2(t)=di(t,σ2)/(fT(t)Gf(t)+di(t,σ2)), and θW=(ϑ,G,σ2).
The proof of Proposition 1 is given in Web Appendix A. We refer to model (5) as the measurement error induced hazard model. This approach based on the induced hazard model can be easily extended to handle multivariate Xi(t). The parameter σi,rel2(t) measures the percentage of the measurement error variation over the total variation in Wi,b(vij) under model (2). The factor ωi(t)=1-σi,rel2(t) is termed as the reliability ratio [26] representing the attenuating effect of use of the estimated covariate X^i(t). If there is no measurement error, i.e., σ2=0, then ωi(t)=1, Oi(β,t,θW)=0 and X^i(t)=Wi(t)=Xi(t). If the measurement times v~i do not vary with i, then σi,rel2(t) does not depend on i.
Since only the longitudinal measurements in the past can be meaningfully used to model current or future risk of failure, the subject-specific estimates, X^i(t), are based on the measurements Wij before τ if the failure event has not occurred by the end of study time, before Ti if the failure time is observed (Δi=1), and before Li that is right before the failure time Ti if Δi=0.
Estimation of Measurement Error Induced Hazard Model with Partly Interval Censored Data
Next, we derive an estimator of the induced hazard model based on partly interval censored data. The observed data from a random sample of n study participants consist of {(Δi,ΔiTi,(1-Δi)Li,(1-Δi)RiI(Ri<∞),Zi,v~i,W~i}, i=1,…,n. Recently, [11] developed maximum likelihood estimation for semiparametric transformation models with partly interval censored data. The method extended the EM algorithm approach of [9] for interval censored data to partly interval censored data. We adopt this approach to estimate the measurement error induced hazard model (5) with partly interval censored data.
Under model (5), the conditional survival function of Ti given Ti≥vir equals exp(-∫virtλ∗(x|X^i(x), Zi,v~i(x))dx). Let Λ0(t)=∫0tλ(s)ds. Note that θW in model (5) can be estimated based on model (2) such that we treat it as known for now. Let hi(t,β,γ)=βωi(t)X^i(t)+γTZi+Oi(β,t,θW). The observed data likelihood function for (β,γ,Λ) under model (5) is Ln(β,γ,Λ;θW)=6 =∏i=1n{[Λ′(Ti)exp{hi(Ti,β,γ)}]I(vir≤Ti)exp(-∫virTiexp{hi(t,β,γ)}dΛ(t))}Δi{exp(-∫virLiexp{hi(t,β,γ)}dΛ(t))-exp(-∫virRiexp{hi(t,β,γ)}dΛ(t))}1-Δi.
Because the likelihood (6) can become arbitrarily large within the class of absolutely continuous functions Λ(·), the nonparametric maximum likelihood estimator (NPMLE) is often obtained on a restricted space. Following this typical approach, e.g., [9], we regard Λ(t) as a step function with nonnegative jumps at observed Ti and at the endpoints of the intervals (Li,Ri], i=1,…,n. Let 0=t0<t1<⋯<tm be the ordered unique values of the set {(ΔiTi,(1-Δi)Li,(1-Δi)RiI(Ri<∞)):i=1,…,n}.
Let λk be the jump size of the estimator for Λ(t) at tk for k=1,…,m and let λ0=0. Let hi(tik,β,γ)=βωikX^ik+γTZi+Oi(β,tk,θW), where X^ik=X^i(tk) and ωik=ωi(tk). With Λ(t) a step function with jumps λk at tk, k=1,…,m, the likelihood (6) becomes Ln(β,γ,Λ;θW)=7 =∏i=1n{[Λ{Ti}exp{hi(Ti,β,γ)}]I(vir≤Ti)exp(-∑tk≤TiI(vir≤tk)λkexp{hi(tik,β,γ)})}Δi{exp(-∑tk≤LiI(vir≤tk)λkexp{hi(tik,β,γ)})[1-exp(-∑Li<tk≤RiI(vir≤tk)λkexp{hi(tik,β,γ)})]I(Ri<∞)}1-Δi,
where Λ{Ti} denotes the jump size of Λ(t) at Ti.
We consider an EM algorithm to maximize Ln(β,γ,Λ;θW). Let ηik be independent Poisson random variables with means μik=λkexp{hi(tik,β,γ)}, i=1,…,n, k=1,…,m. Following [11], for i=1,…,n, we defineAi=Δi∑tk<TiI(vir≤tk)ηik,Bi=Δi∑tk=TiI(vir≤tk)ηik,Ci=(1-Δi)∑tk≤LiI(vir≤tk)ηik,Di=(1-Δi)I(Ri<∞)∑Li<tk≤RiI(vir≤tk)ηik.
Let X^i·={X^ik,k=1,…,m}. The likelihood of the observed data given by (v~i,Ti,X^i·,Zi, Ai=0,Bi=1) for Δi=1 and (v~i,Li,Ri,X^i·,Zi,Ci=0,Di>0) for Δi=0 under model (5) isLn∗=∏i=1n{P(Ai=0,Bi=1)}Δi{P(Ci=0,Di>0)}1-Δi.
Note that P(Ai=0,Bi=1) equals the term in the likelihood (7) corresponding to Δi=1, and P(Ci=0,Di>0) equals the term in the likelihood (7) corresponding to Δi=0. Hence, Ln∗ equals the observed likelihood (7), which takes the form8 Ln(β,γ,Λ;θW)=∏i=1n∏tk<TiP(ηik=0)I(vir≤tk)∏tk=TiP(ηik=1)I(vir≤tk)Δi∏tk≤LiP(ηik=0)I(vir≤tk)1-∏Li<tk≤RiP(ηik=0)I(vir≤tk)I(Ri<∞)1-Δi.
We maximize the likelihood (8) through an EM algorithm by treating ηik as missing data. Let Ri∗=ΔiTi+(1-Δi){LiI(Ri=∞)+RiI(Ri<∞)}. Let 1ik∗=I(vir≤tk≤Ri∗). The complete-data log likelihood is given by9 Cln(β,γ,Λ;θW)=∑i=1n∑k=1m1ik∗[ηiklog(μik)-log(ηik!)-μik].
Taking derivatives of (9), we obtain the score functions10 ∂Cln(β,γ,Λ;θW)∂(β,γ)=∑i=1n∑k=1m1ik∗Zik∗[ηik-λkexp{βωikX^ik+γTZi+Oi(β,tk,θW)}],
11 ∂Cln(β,γ,Λ;θW)∂λk=∑i=1n1ik∗ηikλk-exp{βωikX^ik+γTZi+Oi(β,tk,θW)},
for k=1,…,m, where Zik∗=((ωikX^ik+O˙i(β,tk,θW))T,ZiT)T and O˙i(β,tk,θW) is the derivative of Oi(β,tk,θW) with respect to β.
In the M-step, we calculate λk based on the score (11):12 λk=∑i=1n1ik∗E^(ηik)∑i=1n1ik∗exp{βωikX^ik+γTZi+Oi(β,tk,θW)},
for k=1,…,m, where E^(ηik) denotes the posterior mean given the observed data. We then plug (12) into (10) and solve the score equations for β and γ:13 ∑i=1n∑k=1m1ik∗E^(ηik)Zik∗-∑j=1n1jk∗exp{γTZj+βωjkS^jk+Oj(β,tk,θW)}Zjk∗∑j=1n1jk∗exp{γTZj+βωjkS^jk+Oj(β,tk,θW)}=0.
The M-step estimators of λk (k=1,…,m) and (β,γ) are obtained from (12) and (13).
In the E-step, we calculate the posterior mean E^(ηik) of ηik conditional on the observed data (v~i,Ti,X^i·,Zi,Ai=0,Bi=1) for Δi=1 and (v~i,Li,Ri,X^i·,Zi,Ci=0,Di>0) for Δi=0. For Δi=1, E^(ηik)=0 for vir<tk<Ti and E^(ηik)=1 for vir<tk=Ti. For Δi=0, E^(ηik)=E(ηik|v~i,Li,Ri,X^i·,Zi,Ci=0,Di>0). It follows that E^(ηik)=0 for vir≤tk≤Li, and14 E^(ηik)=E(ηik|v~i,Li,Ri,X^i·,Zi,Ci=0,Di>0)=λkexp{βωikX^ik+γTZi+Oi(β,tk,θW)}1-exp{-∑Li<tk≤Ri1ik∗λkexp{βωikX^ik+γTZi+Oi(β,tk,θW)}},
for vir≤tk and Li<tk≤Ri with Ri<∞.
The estimators of (λk,k=1,…,m) and (β,γ) are obtained by iterating between the E and M steps until convergence, which are denoted by (λ^k,k=1,…,m) and (β^,γ^). We estimate Λ(·) by Λ^(·), which is the step function with jump size λ^k at tk, k=1,…,m. This EM procedure assumes that the measurement error model parameters θW are known. In practice, they are usually unknown. These parameters can be estimated by existing methods for estimating a linear mixed effects model. In the numerical studies, we obtain the maximum likelihood estimates θ^W using the lmer function in the R package lme4 [27]. The aforementioned EM procedure is then carried out by replacing θW with θ^W. Therefore, (β^,γ^,Λ^(·)) is a plug-in estimator that maximizes logLn(β,γ,Λ;θ^W) for (β,γ)∈B and Λ∈C, where B is a known compact set in Rd+1 and C is the set of step functions with nonnegative jumps at tk, k=1,…,m.
The following theorem summarizes the asymptotic properties of the estimators (β^,γ^,Λ^(·)). The proof is outlined in Web Appendix A.
Theorem 1
Under Conditions (A1)-(A3) and (B1)-(B4) given in the Appendix, (β^,γ^,Λ^(t)) converges almost surely to (β,γ,Λ(t)) uniformly in t∈[ζ,τ], and n(β^-β,γ^-γ,Λ^(t)-Λ(t)) converges in distribution to a mean zero Gaussian process for t∈[ζ,τ].
Variance Estimation
The proposed estimator (β^,γ^) for model (1) is the profile likelihood estimator by profiling out the baseline Λ with the plugging in of θ^W for θW. Define the profile log likelihoodpln(β∗;θW)=maxΛ∈ClogLn(β,γ,Λ;θW),
where β∗=(β,γT)T, Ln(β,γ,Λ;θW) is given in (7) and C is the set of step functions with nonnegative jumps at tk, k=1,…,m. Then β^∗=argmaxβ∗∈B pln(β∗;θ^W). When θW is known, the profile likelihood approach can be used to estimate the covariance matrix of β^ [28]. With the plug-in estimator θ^W, the estimator of the variance of β^∗=(β^,γ^T)T needs to account for the variation of θ^W.
Let U(β∗;θ^W)=∂∂β∗pln(β∗;θ^W). Then U(β^∗;θ^W)=0. By (1) in the proof of Theorem 1 in the Web Appendix A, we have15 β^∗-β∗=-(∂U(β∗;θW)∂β∗)-1[U(β∗;θW)+∂U(β∗;θW)∂θW(θ^W-θW)]+op(n-1/2).
Under the measurement error model (2), the estimator θ^W admits the approximation θ^W-θW=J-1∑i=1nξi+op(n-1/2), where ξi are iid random vectors with mean zero and J is a positive definite matrix. Under Conditions (A1)-(A3) given in the Appendix, U(β∗;θW) and θ^W-θW are uncorrelated. Therefore, the two summands in (15) are asymptotically independent.
The covariance matrix of β^∗ equals16 Cov(β^∗)=(∂U(β∗;θW)∂β∗)-1+(∂U(β∗;θW)∂β∗)-1∂U(β∗;θW)∂θWCov(θ^W)(∂U(β∗;θW)∂θW)T(∂U(β∗;θW)∂β∗)-1+op(n-1).
Thus Cov(β^∗) can be consistently estimated by replacing β∗ with β^∗, θW with θ^W and Cov(θ^W) with its estimator Cov^(θ^W). The details of derivations for the variance estimation are given in Web Appendix A.
The (j,k)th element of matrix ∂U(β^∗;θ^W)∂β∗ is estimated bypln(β^∗;θ^W)-pln(β^∗+hnek;θ^W)-pln(β^∗+hnej;θ^W)+pln(β^∗+hnek+hnej;θ^W)hn2,
where ej and ek are the jth and kth canonical vector in Rd+1, respectively, and hn is at the order of n-1/2.
Similarly, the (j,k)th element of matrix ∂U(β∗;θW)∂θW is estimated bypln(β^∗;θ^W)-pln(β^∗+hnek;θ^W)-pln(β^∗;θ^W+hnuj)+pln(β^∗+hnek;θ^W+hnuj)hn2,
where ek is the kth canonical vector in Rd+1 and uj is the jth canonical vector in Rq with q the dimension of θW.
The R package merDeriv developed by [29] for generalized linear mixed models can be used to estimate Cov(θ^W) [27].
To calculate pln(β∗;θW), we apply the proposed EM algorithm with β∗ and θW held fixed. For any given values of β∗ and θW, the procedure iterates between (12) for λk and (14) for E^(ηik). For fast convergence, one can take the estimate λ^k of the jump size of of the cumulative baseline function Λ(·) for model (5) as the initial value. The step size hn in calculating the second order differences can be taken as hn=Cn-1/2, where C is a constant that can be calibrated depending on data applications. Although there has been no existing study examining the optimal choice of hn, our simulation studies show that hn=5n-1/2 works well.
We summarize the steps for implementing the proposed method as follows: Obtain the the maximum likelihood estimates θ^W of the parameters θW=(ϑ,G,σ2) under the measurement error model (2).
Calculate the estimated longitudinal covariates X^i(t)=fT(t)θ^i(t), where θ^i(t) is the least squares estimator of θi given by (3), i=1,…,n.
Estimate the parameters (β,γ,Λ) in the measurement error induced hazard model (5) using the EM algorithm described in Section 3.2, where θW is replaced by θ^W.
Estimate the covariance matrix of β^∗=(β^,γ^) using Cov(β^∗) given by (16).
A Diagnostic Testing Procedure for the Measurement Error Model
This section presents a diagnostic procedure to examine validity of the measurement error model (2). An invalid model can introduce additional bias and diminish the benefits of dealing with the measurement errors. The proposed test procedure provides a formal procedure to check for the model assumptions for the longitudinal covariate.
For each individual i, let e^ij=Wij-f(vij)Tθ^i1BijT and eij=Wij-θiTf(vij)1BijT, where θ^i=θ^i(τ). The regression residual process is defined as ϵ^i(vij)=e^ij1Bij. Let σ^ be the estimator of σ given in (4) under model (2). We introduce the following weighted residual process for individual i,HiMi(t)=∑j=1Mi(t)Bij-1/2ϵ^i(vij)σ^(1-BijfT(vij){F~iT(τ)F~i(τ)}-1f(vij))1/2.
In the following we construct the test based on the differences of the weighted residual processes. Let 0=τ0<τ1<τ2<⋯<τK≤τ be the grid points on [0,τ]. We set Mi(0)=0 and HiMi(0)=0. Define Dik=HiMi(τk)-HiMi(τk-1) for 1≤k≤K, and Hn=∑i=1n(Di1,Di2,…,DiK)T. We propose the test statistic17 Q=HnTΣH-1Hn,
where ΣH is the covariance matrix of Hn. The diagonal of ΣH includes∑i=1nVar(Dik)=∑i=1n(Mi(τk)-Mi(τk-1))-2∑i=1n∑Mi(τk-1)<l<m≤Mi(τk)Ψi,lm,
for 1≤k≤K, and the off-diagonal elements of ΣH are given by∑i=1nCov(Dij,Dik)=-∑i=1n∑Mi(τj-1)<l≤Mi(τj)∑Mi(τk-1)<m≤Mi(τk)Ψi,lm,
for j≠k. HereΨi,jk=Bij-1/2Bik-1/2fT(vij){F~iT(τ)F~i(τ)}-1f(vik)(1-BijfT(vij){F~iT(τ)F~i(τ)}-1f(vij))1/2(1-BikfT(vik){F~iT(τ)F~i(τ)}-1f(vik))1/2.
Theorem 2
Under model (2), the test statistic Q has a chi-square distribution with K degrees of freedom.
By Theorem 2, the test rejects model (2) at significance level α if Q>χK,1-α2. The proof of Theorem 2 is given in the Web Appendix A.
It is easy to show that the test statistic Q has an asymptotic chi-square distribution with K degrees of freedom as long as the random effects νi and the measurement errors eij,b are iid with mean zero and finite variances. The proposed test provides a method to test the form of the within-individual patterns defined by the basis function f(t). It does not test the normality assumptions of θi and the errors eij. Many existing tests such as the Kolmogorov-Smirnov test, Shapiro-Wilk test, and Anderson-Darling test can be used to test for normality. Testing of the normality assumption of θi can be conducted based on {θ^i(τ),i=1,…,n}, while testing of the normality assumption of eij can be conducted based on {e^ij,j=1,…,Mi,i=1,…,n}. Diagnostic tools such as Q-Q plots can be used to compliment the formal test procedures for real data applications.
The proposed test is not overly sensitive to the choice of K. We suggest 3≤K≤8 and that the grid points 0=τ0<τ1<τ2<⋯<τK≤τ be evenly spaced in [0,τ]. Our simulation results show that the test performs well.
Simulation Studies
We evaluate the proposed method via simulation studies. Let n be the sample size. For i=1,…,n, the failure time Ti is generated from the proportional hazards model18 λ(t|X¯i(t),Zi)=λ(t)exp{βXi(t)+γZi},
where λ(t)=1/(2+t), β=0.5, γ=-log(2), Zi∼Ber(0.3), and Xi(t) has the form Xi(t)=(ν0+b0i)+(ν1+b1i)t, with ν0=1, ν1=0.5, (b0i,b1i)∼N(0,G), and G=[0.02,-0.01;-0.01,0.02]. Let Unif(0, a) denote a uniform random variable on (0, a) and Ber(p) a Bernoulli random variable with success probability p. We simulate the measurement times (vi1,…,vi,Mi) for Xi(·) as follows. We first generate the measurement times as the cumulative sums of independent Unif(0, 0.2) random variates until τ/6 is reached, and then keep adding up independent Unif(0, 0.4) random variates until τ. We simulate partly interval censored data for individual i as follows. We first generate the number of monitoring times Ki∼Ber(0.8)+1. If Ki=1, we generate one monitoring time Ui1∼Unif(0,τ/2); define (Li,Ri]=(0,Ui1] if Ti≤Ui1 and (Li,Ri]=(Ui1,∞) if Ti>Ui1. If Ki=2, we generate two monitoring times Ui1∼Unif(0,τ/2) and Ui2∼min{0.1+Ui1+Unif(0,3τ/4),τ}; define (Li,Ri]=(0,Ui1] if Ti≤Ui1, (Li,Ri]=(Ui1,Ui2] if Ui1<Ti≤Ui2, and (Li,Ri]=(Ui2,∞) if Ti>Ui2. If Ri=∞, we set Δi=0; if Ri<∞, we generate Δi∼Ber(p) with p=0.25 or 0.75. If Δi=1, the failure time Ti is exactly observed. The length of study is taken to be τ=3 yielding about 40% right censoring. The error-prone measurements Wi,b(vij) are generated from the model19 Wi,b(vij)=Xi(vij)+eij,b,b=1,…,Bij,
where Xi(vij)=(ν0+b0i)+(ν1+b1i)vij is specified above, eij,b∼N(0,σ2) with σ=0.1 or 0.2 and the number of repeated measurements of Xi(vij) is Bij=B=1 or 3 for all i, j.
We compare four methods: (i) the proposed method; (ii) the ideal method using true X(t) which is not available in practice; (iii) the naive method that ignores measurement error and uses W(t) directly, where W(t) at any time t is evaluated via last value carried forward from the longitudinal measurements (the average is used if there are replicates for W(t)); (iv) the naive method using X^(t) by simply replacing X(t) with X^(t) in the proportional hazards model. For the variance estimation based on the profile likelihood method, we take hn=5n-1/2. The results are similar with other choices of hn, such as n-1/2 and 10n-1/2, which is also noted in [9]. We consider the sample size n=400 and 600. The estimation results for (β,γ) based on 500 simulations are presented in Table 1 for B=1 and in Table S1 of Web Appendix B for B=3, where Bias is the average point estimate minus the true parameter value, SSD is the sample standard deviation of point estimates, ESE is the average of estimated standard errors and CP is the coverage proportion of the 95% confidence interval.Table 1 Simulation results for (β,γ) under models (18) and (19) when there are no repeated measurements of Xi(t), i.e., B=1. The random effects (b0i,b1i)∼N(0,G) with G=[0.02,-0.01;-0.01,0.02]. Each entry is based on 500 replicates
p=0.25
β=0.5 γ=-log(2)
n σ Method Bias SSD ESE CP Bias SSD ESE CP
400 0.1 Proposed 0.045 0.554 0.565 0.962 -0.017 0.166 0.154 0.922
X(t) 0.010 0.443 0.453 0.970 -0.016 0.162 0.152 0.930
W(t) -0.072 0.417 0.424 0.958 -0.015 0.162 0.152 0.928
X^(t) -0.356 0.248 0.189 0.433 -0.016 0.167 0.155 0.918
0.2 Proposed 0.077 0.703 0.719 0.958 -0.017 0.166 0.154 0.920
X(t) 0.010 0.443 0.453 0.970 -0.016 0.162 0.152 0.930
W(t) -0.216 0.342 0.347 0.908 -0.014 0.161 0.152 0.930
X^(t) -0.440 0.150 0.112 0.150 -0.014 0.168 0.154 0.914
600 0.1 Proposed 0.040 0.455 0.452 0.950 -0.001 0.130 0.126 0.944
X(t) -0.010 0.375 0.365 0.946 0.001 0.129 0.124 0.938
W(t) -0.082 0.352 0.341 0.936 0.001 0.129 0.124 0.932
X^(t) -0.393 0.183 0.117 0.245 0.002 0.131 0.126 0.940
0.2 Proposed 0.073 0.569 0.572 0.956 -0.000 0.130 0.126 0.944
X(t) -0.010 0.375 0.365 0.946 0.001 0.129 0.124 0.938
W(t) -0.215 0.290 0.280 0.866 0.002 0.128 0.124 0.932
X^(t) -0.457 0.110 0.067 0.053 0.002 0.130 0.126 0.942
p=0.75
β=0.5 γ=-log(2)
n σ Method Bias SSD ESE CP Bias SSD ESE CP
400 0.1 Proposed 0.022 0.535 0.537 0.954 -0.017 0.167 0.156 0.934
X(t) 0.017 0.430 0.437 0.960 -0.014 0.159 0.150 0.936
W(t) -0.128 0.380 0.377 0.942 -0.013 0.159 0.150 0.938
X^(t) -0.405 0.197 0.125 0.267 -0.015 0.166 0.156 0.933
0.2 Proposed 0.030 0.652 0.662 0.948 -0.016 0.167 0.156 0.934
X(t) 0.017 0.430 0.437 0.960 -0.014 0.159 0.150 0.936
W(t) -0.296 0.289 0.282 0.804 -0.012 0.158 0.150 0.938
X^(t) -0.466 0.111 0.068 0.049 -0.014 0.167 0.157 0.932
600 0.1 Proposed -0.004 0.423 0.432 0.952 0.002 0.132 0.128 0.938
X(t) -0.011 0.359 0.354 0.950 0.003 0.126 0.123 0.944
W(t) -0.146 0.310 0.306 0.912 0.004 0.126 0.123 0.944
X^(t) -0.437 0.136 0.081 0.107 0.004 0.131 0.128 0.936
0.2 Proposed 0.004 0.517 0.532 0.958 0.003 0.132 0.128 0.944
X(t) -0.011 0.359 0.354 0.950 0.003 0.126 0.123 0.944
W(t) -0.304 0.236 0.229 0.725 0.005 0.125 0.123 0.940
X^(t) -0.477 0.075 0.042 0.013 0.007 0.130 0.128 0.940
We can see from Tables 1 and S1 that (i) for all scenarios considered, the proposed method yields unbiased estimates with reasonable estimated standard errors and coverage proportions; (ii) the sample standard deviation of the proposed estimator of β decreases when the degree of measurement error represented by σ decreases and when the sample size n, the number of repeated measurements B of X(t), or the proportion of exact observations p increases; (iii) as expected, the ideal method that uses true X(t) is more efficient than the proposed method and the efficiency gain increases with the degree of measurement error given by σ; (iv) the naive method that ignores measurement error and uses W(t) directly yields acceptable results when σ=0.1, but has large bias when σ=0.2; particularly, it tends to underestimate β; (v) the naive method that replaces X(t) with X^(t) in the proportional hazards model gives severely biased estimates of β for all scenarios considered (and it should be noted that this method underestimates β due to the attenuating effect); and (vi) all methods perform well for the estimation of γ.
We have also investigated the computational cost of the proposed method. For the simulation setup n=400, p=0.75, σ=0.1 and B=1, it takes about 409 seconds (132 for parameter estimation and 277 for variance estimation) to implement the proposed method on a MacBook Pro (3.1GHz Quad-Core Intel Core i7).
To evaluate the robustness of our method to the normality assumption on random effects in the measurement error model, we generate the random effects b0i and b1i from Unif(-0.25,0.25) independently while keeping the other settings the same as in Tables 1 and S1. The results are presented in Table 2 for B=1 and Table S2 of Web Appendix B for B=3. One can see that the proposed method performs well in such situations.Table 2 Simulation results for (β,γ) under models (18) and (19) when there are no repeated measurements of Xi(t), i.e., B=1. The random effects b0i and b1i are independent Unif(-0.25,0.25). Each entry is based on 500 replicates
p=0.25
β=0.5 γ=-log(2)
n σ Method Bias SSD ESE CP Bias SSD ESE CP
400 0.1 Proposed 0.055 0.393 0.392 0.936 -0.007 0.168 0.154 0.938
X(t) 0.015 0.340 0.341 0.938 -0.007 0.164 0.152 0.930
W(t) -0.021 0.324 0.335 0.960 -0.007 0.164 0.152 0.930
X^(t) -0.309 0.230 0.164 0.453 -0.006 0.165 0.154 0.944
0.2 Proposed 0.093 0.472 0.470 0.936 -0.008 0.168 0.154 0.942
X(t) 0.015 0.340 0.341 0.938 -0.007 0.164 0.152 0.930
W(t) -0.131 0.284 0.294 0.944 -0.006 0.164 0.152 0.934
X^(t) -0.413 0.150 0.108 0.178 -0.006 0.166 0.154 0.944
600 0.1 Proposed 0.045 0.296 0.316 0.958 -0.014 0.135 0.126 0.932
X(t) 0.011 0.261 0.276 0.970 -0.013 0.135 0.124 0.938
W(t) -0.030 0.245 0.271 0.968 -0.012 0.135 0.124 0.936
X^(t) -0.352 0.173 0.108 0.260 -0.011 0.133 0.126 0.929
0.2 Proposed 0.079 0.354 0.378 0.958 -0.013 0.135 0.126 0.930
X(t) 0.011 0.261 0.276 0.970 -0.013 0.135 0.124 0.938
W(t) -0.142 0.210 0.237 0.942 -0.011 0.134 0.124 0.932
X^(t) -0.432 0.106 0.063 0.049 -0.009 0.132 0.126 0.929
p=0.75
β=0.5 γ=-log(2)
n σ Method Bias SSD ESE CP Bias SSD ESE CP
400 0.1 Proposed 0.031 0.381 0.380 0.940 -0.004 0.170 0.156 0.936
X(t) 0.017 0.327 0.334 0.952 -0.004 0.161 0.150 0.936
W(t) -0.074 0.291 0.311 0.962 -0.004 0.161 0.150 0.940
X^(t) -0.356 0.199 0.121 0.318 -0.000 0.169 0.156 0.933
0.2 Proposed 0.038 0.446 0.443 0.940 -0.004 0.170 0.156 0.936
X(t) 0.017 0.327 0.334 0.952 -0.004 0.161 0.150 0.936
W(t) -0.224 0.234 0.250 0.888 -0.003 0.161 0.150 0.936
X^(t) -0.441 0.116 0.069 0.065 0.000 0.169 0.156 0.928
600 0.1 Proposed 0.018 0.288 0.307 0.964 -0.012 0.135 0.128 0.936
X(t) 0.008 0.257 0.271 0.960 -0.011 0.135 0.123 0.932
W(t) -0.080 0.231 0.252 0.954 -0.010 0.134 0.123 0.936
X^(t) -0.398 0.142 0.075 0.136 -0.009 0.133 0.128 0.944
0.2 Proposed 0.028 0.338 0.358 0.956 -0.011 0.136 0.128 0.938
X(t) 0.009 0.257 0.271 0.960 -0.011 0.135 0.123 0.932
W(t) -0.227 0.185 0.203 0.830 -0.009 0.134 0.123 0.932
X^(t) -0.461 0.077 0.041 0.006 -0.009 0.133 0.128 0.943
In addition, we obtain the estimate of the baseline hazard function λ(t) using kernel smoothing with the Gaussian kernel and bandwidth 0.1. Figures S1∼S4 in Web Appendix B plot the estimated baseline hazard functions based on the simulation results of Tables 1, 2, S1 and S2, respectively. One can see that the proposed method and the ideal method yield unbiased estimates of the baseline hazard function λ(t) except for t close to 0, while the naive methods yield biased estimates.
We also conduct a simulation study to examine the empirical sizes and powers of the proposed test for Xi(vij)=(ϑ+νi)Tf(vij) under the measurement error model (2), for i=1,…,n, j=1,…,Mi and b=1,…,B, where f(t) is an r×1 vector of basis functions, ϑ is a vector of fixed parameters, and νi(i=1,…,n) are iid N(0, G). We set the null model to be f(t)=(1,t) and generate data from the following four models: I: f(t)=(1,t), ϑ=(1,0.5) and G=[0.02,-0.01;-0.01,0.02]
II: f(t)=(1,t,t2), ϑ=(1,0.5,0.01) and G=[0.02,-0.01,0;-0.01,0.02,0;0,0,0.02]
III: f(t)=(1,t,t2), ϑ=(1,0.5,0.02) and G=[0.02,-0.01,0;-0.01,0.02,0;0,0,0.02]
IV: f(t)=(1,t,t2), ϑ=(1,0.5,0.02) and G=[0.02,-0.01,-0.01;-0.01,0.02,-0.01;-0.01, -0.01,0.02]
Here for easy presentation G is the covariance matrix with rows separated by semicolons.
We simulate the measurement times (vi1,…,vi,Mi) for X(·) as in Tables 1 and S1. Specifically, we first generate the measurement times as the cumulative sums of Unif(0, 0.2) until τ/6 is reached, and then keep adding up Unif(0, 0.4) variates until τ. We set the length of follow-up to be τ=3 and take equally spaced grid points in [0,τ] with K=1,3,5,7. We conduct the test and compute the p-value for each simulated dataset. We calculate the empirical size for Model I and the power for Model II, III and IV as the proportion of p-values ≤0.05 for 500 simulated datasets. The results are presented in Table 3 for B=1 and in Table S3 of the Web Appendix B for B=3. The empirical sizes under Model I are around the 0.05 nominal level for all cases and remain the same for different values of σ. The powers under Model II, III and IV are fairly high given the small effect sizes 0.01 and 0.02. The power increases with the number of repeated measurements B, the sample size n and the effect size. Moreover, the power decreases when σ increases, and seems to be similar when using different numbers of grid points K in the test.Table 3 Simulation results for the proposed test of the measurement error model (2) with Xi(t)=(ϑ+νi)Tf(t) at significance level 0.05, when there are no repeated measurements of Xi(t), i.e., B=1. Each entry is based on 500 replicates
Model I (size) Model II (power)
n σ K=1 K=3 K=5 K=7 K=1 K=3 K=5 K=7
200 0.05 0.040 0.062 0.054 0.050 0.746 0.676 0.668 0.666
0.1 0.040 0.062 0.054 0.050 0.682 0.608 0.582 0.572
0.2 0.040 0.062 0.054 0.050 0.520 0.432 0.390 0.368
400 0.05 0.050 0.048 0.066 0.042 0.832 0.764 0.756 0.760
0.1 0.050 0.048 0.066 0.042 0.778 0.718 0.698 0.680
0.2 0.050 0.048 0.066 0.042 0.638 0.568 0.508 0.488
Model III (power) Model IV (power)
n σ K=1 K=3 K=5 K=7 K=1 K=3 K=5 K=7
200 0.05 0.930 0.890 0.878 0.876 0.932 0.902 0.902 0.904
0.1 0.904 0.866 0.838 0.844 0.904 0.880 0.858 0.870
0.2 0.798 0.746 0.686 0.648 0.822 0.740 0.688 0.664
400 0.05 0.984 0.974 0.974 0.964 0.984 0.982 0.976 0.974
0.1 0.976 0.962 0.956 0.946 0.976 0.968 0.960 0.962
0.2 0.938 0.898 0.890 0.868 0.950 0.906 0.898 0.884
Application to ACTG 175
We apply the proposed method to the ACTG 175 trial, a randomized, double-blind phase II/III trial of antiretroviral regimens in persons living with HIV infection with CD4 cell count from 200 to 500 per cubic millimeter [1]. Between December 1991 and October 1992, 2467 individuals were recruited and followed until November 1994. Among these, 1396 participants received antiretroviral therapy (ART) prior to the study while 1061 participants were ART naive. The objective of the trial was to compare the effectiveness of four antiretroviral regimens (zidovudine only, zidovudine + didanosine, zidovudine + zalcitabine, and didanosine only) in preventing disease progression to AIDS or death. An important prognostic biomarker for progression to the clinical endpoint is CD4 cell count per cubic millimiter of blood [30]. All ACTG 175 trial participants had CD4 cell count measured every 12 weeks starting at Week 8, and were followed for occurrence of the composite clinical endpoint of AIDS or death. The median number of measurement times is 12 with interquartile range (IQR) [8,14], and its histogram is given by Figure S8 in the Web Appendix.
The original analysis of [1] found zidovudine alone to be inferior to the other three therapies. Following [17] and [20], we consider two treatment groups, zidovudine alone and the combination of the other three therapies. We demonstrate the utility of the proposed method by investigating associations of treatment arm and time-dependent trajectory log10(CD4) with the composite clinical endpoint of AIDS or death. Let Z be the treatment indicator (TRT) with value 0 for zidovudine alone and 1 for other three regimens. Let X(·) be the error-prone time-dependent covariate log10(CD4) (measured without replicates, i.e., B=1) and T be the time from enrollment to AIDS or death, whichever occurred first. We assume that the conditional hazard function of T given X¯(t) and Z follows the proportional hazards model (1),λ(t|X¯(t),Z)=λ(t)exp{βX(t)+γZ}=λ(t)exp{βlog10(CD4)+γTRT},
where the regression coefficients β and γ can be interpreted as log hazard ratios and represent the association of time to AIDS or death with log10(CD4) and TRT, respectively. Also, we assume that the measurement error model for X(t) is (2) with the quadratic basis function f(t)=(1,t,t2). Our analysis includes 1396 participants who received ART prior to the study. There were 215 composite endpoint cases (15.4%) with 167 AIDS events and 48 deaths. The time to the AIDS onset is interval censored while the time to death prior to AIDS is observed or right-censored. The observed data consist of exact, interval- and right-censored event times.
The true CD4 cell count values X(t) are generally not attainable. The observed CD4 cell count is measured intermittently and is an error-prone time-dependent covariate. The naive approaches often replace X(t) in model (1) with the observed W(t) by last value carried forward that ignores the measurement errors or with a model-based estimate X^(t) without modifying the hazard model for the induced error. The former naive approach – termed the “naive approach using W(t)” – imputes the CD4 values at each failure time using “last value carried forward” that substitutes the unavailable CD4 cell count with the last observed value prior to the failure time. The latter naive approach – termed the “naive approach using X^(t)” – replaces CD4 cell count values with the estimated X^(t) in model (1) using the measurement error model (2) based on each individual’s longitudinal profile prior to time t.
Fitting the quadratic measurement error model (2), we obtain average values of the individual-specific estimates of coefficients of (2.515,-0.035,-0.047). The plot of the average fitted individual-specific curves along with the plot of the observed log10(CD4) for 50 randomly selected individuals shows the downward trend in log10(CD4) (Figure S5). The results of analysis using the proposed method and the two naive methods are summarized in Table 4, where Est is the estimates of the regression parameters and SE is the estimated standard errors. The estimated regression coefficient (the standard error) of log10(CD4) using the proposed method is -2.472 (0.116), whose absolute value is much larger than that of the estimated regression coefficient (the standard error) -0.240 (0.051) obtained using the naive approach using X^(t) but slightly less than that of the estimated regression coefficient (the standard error) -2.668 (0.119) using the naive approach using W(t). The estimated regression coefficient of log10(CD4) is the log hazards ratio for every unit increase in log10(CD4) under the Cox model (equivalently every 10-fold increase in CD4 cell count with units number of cells per cubic milimeter of blood, cells/mm3) and represents the association of log10(CD4) with the failure time. All methods suggest that lower values of log10(CD4) are significantly associated with higher risk of AIDS or death. The naive approach using X^(t) yields an estimated association closer to zero (-0.240) as compared to the proposed method (-2.472) partly due to the attenuating effect ωi(t) from the measurement errors. The naive approach using W(t) yields a slightly stronger inverse association (-2.668) as compared to the proposed method because it tends to carry forward a too-large value of CD4 cell count. While there is no significant treatment effect after adjusting for log10(CD4) with the proposed method and the naive approach using W(t), the naive approach using X^(t) shows a significant treatment effect with p-value 0.034. Our study further confirms that the naive approaches can lead to biased estimates of the associations of interest for the variables measured in errors as well as biased estimates of treatment effects [2, 3, 31].
Figure 1 plots the estimated survival functions at four different combinations of covariates: two values of Z (0 or 1) and two curves of log10(CD4) (25th or 75th percentile of the estimated log10(CD4) at each time point). The plots show that the naive approach using W(t) overestimates the survival probabilities and the naive approach using X^(t) substantially underestimates the survival probabilities. The discrepancy is very large when X(t) is the 75th percentile of log10(CD4). Plots of the estimated baseline hazard functions for the three methods considered are given in Figure S7 in Web Appendix B. It can be seen that the naive approach using W(t) overestimates the baseline hazard function, while the naive approach using X^(t) highly underestimates the baseline hazard function.Table 4 Analysis results for ACTG 175
Proposed Method Naive Using W(t) Naive Using X^(t)
Covariates Est SE p-value Est SE p-value Est SE p-value
log10(CD4) -2.472 0.116 <0.001 -2.668 0.119 <0.001 -0.240 0.051 <0.001
Treatment -0.119 0.160 0.458 0.056 0.155 0.718 -0.338 0.159 0.034
Fig. 1 Plots of the estimated survival functions at four different combinations of covariates. For example, ‘X25+Z0’ corresponds to the covariates combination with X(t) being the 25th percentile of log10(CD4) and for zidovudine alone, and ‘X25+Z1’ corresponds to the covariates combination with X(t) being the 25th percentile of log10(CD4) and for the other three treatment arms pooled
To examine appropriateness of the quadratic measurement error model (2), we conduct the model checking procedure. In particular, we consider the quadratic measurement error model corresponding to f(t)=(1,t,t2) in (2). We set the grid points to be the (0,25,50,75,100)th quantiles of the follow-up times. The quadratic measurement error model yields a p-value of 0.298 suggesting that the quadratic model fits reasonably well to the data. As a comparison, we also test fitness of the linear measurement error model with f(t)=(1,t). The test yields a p-value close to zero. In addition, we note that the log-likelihood value at the final estimates of the proposed method under the linear measurement error model is -932.56, while the log-likelihood value under the quadratic model is -856.54. The analysis supports that the quadratic measurement error model fits the data better than the linear model.
We further examine the fit of the quadratic measurement error model via graphical tools. The residual plots for the quadratic measurement error model, including the normal Q-Q plot and histogram, are presented in Figure S6 of Web Appendix B. The normal Q-Q plot suggests that the sample quantiles of the standardized residuals are close to the theoretical ones except for slight deviations at the two tails, while the histogram of the standardized residuals looks like a standard normal density curve except having slightly shorter tails. Furthermore, to evaluate the normality of random effects, we obtain the least-squares estimates of individual-specific coefficients in the quadratic measurement error model. The normal Q-Q plot and histogram of these estimated coefficients are given in Figure S6. These plots look satisfactory in general except for in the tails. The p-values from the Kolmogorov-Smirnov tests for normality are 0.0006 for the errors and <0.0001 for the random effects of the quadratic measurement error model. The small p-values reflect lack of fit in the tail areas of the distributions though the normality assumptions seem reasonable overall based on the diagnostic plots. The very small p-values can also be a result of the large sample size. Nevertheless, as shown in the simulation studies, the proposed method seems to be robust to the normality assumption of the random effects.
Concluding Remarks
This article develops an estimation method for the Cox model based on partly interval censored failure time and a longitudinal covariate with measurement errors. The research is motivated by the ACTG 175 trial to understand the association of longitudinal CD4 cell count on the hazard of the composite clinical endpoint of AIDS or death, where the time to the composite endpoint is partly interval censored and the recorded values of CD4 cell count are error-prone measures of the unattainable true values. The proposed measurement error induced hazard approach is intuitively appealing and easy to interpret. The EM-algorithm is proposed to implement the maximum likelihood estimation with partly interval censored data. The developed method has broad applications. For example, COVID-19 vaccine efficacy trials will study longitudinal antibody biomarkers over time as correlates of the study endpoint acquisition of SARS-CoV-2 infection. This endpoint is a composite endpoint with the same structure as the AIDS/death composite endpoint, defined as the first event of asymptomatic SARS-CoV-2 infection measured by seroconversion from a blood sample (interval censored) and symptomatic virologically confirmed SARS-CoV-2 infection that is symptom-triggered and hence measured exactly [32].
This paper assumes that the measurement errors are independent identically distributed. In practice, the measurement errors may be correlated or the measurement error variance is heterogeneous over time. In this case, the likelihood method can be used to estimate model (2) by assuming a certain covariance structure for the measurement errors. This is an interesting scenario that is worth investigation in a future project. We have regarded Xi(t) as a scalar. The method can be extended to multivariate longitudinal covariates. As with many works in the joint modeling framework, a limitation of the proposed method is the normality assumption of the random effects in the measurement error model for Xi(t). A simulation study conducted to examine the robustness of the proposed method shows that the estimation bias remains small. A formal test procedure is proposed to examine validity of the measurement error model.
Supplementary Information
Web Appendices A and B, referenced in Sections 3, 5 and 6, are included in the Supplementary Information and are available with this paper on the journal website.
Supplementary Information
Below is the link to the electronic supplementary material.Supplemenatry File1 (PDF 1401 kb)
Appendix: Technical Details
In this appendix, we present the regularity conditions needed for Propositions 1 and Theorem 1.
Let U~i=(Ui1,…,Ui,Ki) denote the monitoring times for the failure event for individual i, where 0=Ui0<Ui1<⋯<Ui,Ki<Ui,Ki+1=∞. The monitoring times are the mechanism used to generate interval censored failure times [Li,Ri], where Li=max{Uik:Ti>Uik,k=0,…,Ki} and Ri=min{Uik:Ti≤Uik,k=1,…,Ki+1} with Ui0=0 and Ui,Ki+1=∞. We assume Conditions (A1)-(A3) below that require noninformative monitoring/measurement times and a nondifferential measurement error mechanism for the time-dependent covariates. The monitoring times, measurement times and measurement errors are noninformative given the information already provided by Zi and θi, i.e., Ti is independent of (Δi,U~i,v~i,e~i) given (Zi,θi).
Measurement error e~i is independent of (Ti,Δi,U~i,v~i,Zi).
Xi(t), 0≤t≤τ, is a left continuous process.
More discussion on the assumptions of noninformative observation times and a nondifferential measurement error mechanism can be found in [21, 26].
The estimation of the induced hazard model (5) for partly interval censored data follows the EM procedure developed by [11]. In addition to Conditions (A1)-(A3), we assume the following regularity conditions similar to [11]. The true value of (β,γ) lies in the interior of a known compact set B in Rd+1, and the true value of Λ(·) is continuously differentiable with positive derivatives in [ζ,τ], where [ζ,τ] is the union of the supports of {ΔiTi,(1-Δi)Li,(1-Δi)Ri}.
The vector of the basis functions f(t) is left continuous and with bounded total variation over [ζ,τ].
If h(t)+βωi(t)X^i(t)+γTZi+Oi(β,t,θW)=0 for all t∈[ζ,τ] with a positive probability, then h(t)=0 for t∈[ζ,τ] and (β,γ)=0.
0<P(Δi=0)<1, P(Li=τ,Ri=∞|Δi=0,X¯i(τ),Zi)≥c1 and P(Ri-Li>c2|Δi=0,X¯i(τ),Zi)=1 for some positive constants c1 and c2. The conditional density of (Li,Ri) given (X¯i(τ),Zi), denoted by g(u, v), has continuous second-order partial derivatives with respect to u and v when v-u>c2 and are continuously differentiable with respect to (X¯i(τ),Zi).
Acknowledgements
The authors thank the reviewers for their constructive comments that have improved the paper. This research was partially supported by NIAID NIH award R37AI054165. Dr. Sun’s research was also partially supported by National Science Foundation grant DMS1915829 and the Reassignment of Duties fund provided by the University of North Carolina at Charlotte. Dr. Zhou’s research was partially supported by National Science Foundation grant DMS1916170. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Declarations
Conflict of interest
The authors declare that there is no conflict of interest.
==== Refs
References
1. Hammer SM Katzenstein DA Hughes MD Gundacker H Schooley RT Haubrich RH Henry WK Lederman MM Phair JP Niu M A trial comparing nucleoside monotherapy with combination therapy in HIV-infected adults with CD4 cell counts from 200 to 500 per cubic millimeter N Engl J Med 1996 335 1081 1090 10.1056/NEJM199610103351501 8813038
2. Prentice RL Covariate measurement errors and parameter estimation in a failure time regression model Biometrika 1982 69 331 342 10.1093/biomet/69.2.331
3. Tsiatis AA Davidian M A semiparametric estimator for the proportional hazards model with longitudinal covariates measured with error Biometrika 2001 88 447 458 10.1093/biomet/88.2.447
4. Finkelstein DM A proportional hazards model for interval-censored failure time data Biometrics 1986 42 845 854 10.2307/2530698 3814726
5. Huang J Efficient estimation for the proportional hazards model with interval censoring Ann Stat 1996 24 540 568 10.1214/aos/1032894452
6. Huang J, Wellner JA (1997) Interval censored survival data: a review of recent progress. In proceedings of the first seattle symposium in biostatistics pp. 123–169 Springer
7. Zhang Y Hua L Huang J A spline-based semiparametric maximum likelihood estimation method for the Cox model with interval-censored data Scand Stat Theory Appl 2010 37 338 354 10.1111/j.1467-9469.2009.00680.x
8. Wang L McMahan CS Hudgens MG Qureshi ZP A flexible, computationally efficient method for fitting the proportional hazards model to interval-censored data Biometrics 2016 72 222 231 10.1111/biom.12389 26393917
9. Zeng D Mao L Lin D Maximum likelihood estimation for semiparametric transformation models with interval-censored data Biometrika 2016 103 253 271 10.1093/biomet/asw013 27279656
10. Kim JS Maximum likelihood estimation for the proportional hazards model with partly interval-censored data J R Stat Soc Series B 2003 65 489 502 10.1111/1467-9868.00398
11. Zhou Q Sun Y Gilbert PB Semiparametric regression analysis of partly interval-censored failure time data with application to an AIDS clinical trial Stat Med 2021 40 4376 4394 10.1002/sim.9035 34080723
12. Nakamura T Proportional hazards model with covariates subject to measurement error Biometrics 1992 48 829 838 10.2307/2532348 1420844
13. Hughes M Regression dilution in the proportional hazards model Biometrics 1993 49 1056 1066 10.2307/2532247 8117900
14. Buzas J Unbiased scores in proportional hazards regression with covariate measurement error J Stat Plan Inference 1998 67 247 257 10.1016/S0378-3758(97)00108-0
15. Huang Y Wang CY Cox regression with accurate covariates unascertainable: a nonparametric-correction approach J Am Stat Assoc 2000 95 1209 1219 10.1080/01621459.2000.10474321
16. Hu C Lin DY Cox regression with covariate measurement error Scand Stat Theory Appl 2002 29 637 655 10.1111/1467-9469.00310
17. Song X Davidian M Tsiatis AA An estimator for the proportional hazards model with multiple longitudinal covariates measured with error Biostatistics 2002 3 4 511 528 10.1093/biostatistics/3.4.511 12933595
18. Song X Huang Y On corrected score approach for proportional hazards model with covariate measurement error Biometrics 2005 61 702 714 10.1111/j.1541-0420.2005.00349.x 16135021
19. Yi G Lawless J A corrected likelihood method for the proportional hazards model with covariates subject to measurement error J Stat Plan Inference 2007 137 1816 1828 10.1016/j.jspi.2006.04.007
20. Fu R Gilbert PB Joint modeling of longitudinal and survival data with the cox model and two-phase sampling Lifetime Data Anal 2017 23 136 159 10.1007/s10985-016-9364-1 27007859
21. Tsiatis AA Davidian M Joint modeling of longitudinal and time-to-event data: an overview Stat Sin 2004 14 809 834
22. Song X Ma S Multiple augmentation for interval-censored data with measurement error Stat Med 2008 27 16 3178 3190 10.1002/sim.3156 18040984
23. Mandal S Wang S Sinha S Analysis of linear transformation models with covariate measurement error and interval censoring Stat Med 2019 38 4642 4655 10.1002/sim.8323 31347177
24. Chen K Jin Z Ying Z Semiparametric analysis of transformation models with censored data Biometrika 2002 89 659 668 10.1093/biomet/89.3.659
25. Wen CC Chen YH Functional inference for interval-censored data in proportional odds model with covariate measurement error Stat Sin 2014 24 1301 1317
26. Yi G Statistical Analysis with Measurement Error or Misclassification 2017 New York Springer
27. Bates D Mächler M Bolker B Walker S Fitting linear mixed-effects models using lme4 J Stat Softw 2015 67 1 48 10.18637/jss.v067.i01
28. Murphy SA Van der Vaart AW On profile likelihood J Am Stat Assoc 2000 95 449 465 10.1080/01621459.2000.10474219
29. Wang T Merkle EC merDeriv: derivative computations for linear mixed effects models with application to robust standard errors J Stat Softw 2018 87 1 1 16
30. Mellors JW Munoz A Giorgi JV Margolick JB Tassoni CJ Gupta P Kingsley LA Todd JA Saah AJ Detels R Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection Ann Intern Med 1997 126 12 946 954 10.7326/0003-4819-126-12-199706150-00003 9182471
31. Dafni UG Tsiatis AA Evaluating surrogate markers of clinical outcome when measured with error Biometrics 1998 54 4 1445 1462 10.2307/2533670 9883544
32. Mehrotra DV Janes HE Fleming TR Annunziato PW Neuzil KM Carpp LN Benkeser D Brown ER Carone M Cho I Clinical endpoints for evaluating efficacy in covid-19 vaccine trials Ann Intern Med 2020 22 M20 6169
|
PMC010xxxxxx/PMC10199272.txt |
==== Front
Arch Virol
Arch Virol
Archives of Virology
0304-8608
1432-8798
Springer Vienna Vienna
37209311
5755
10.1007/s00705-023-05755-0
Original Article
Epidemiological characteristics of patients from fever clinics during the COVID-19 epidemic in 2022 in Shanghai, China
Zhang Yuanjing 1
Wang Jianrong 1
Xie Ying 1
Cao Xinghao 1
Huang Huili 1
Liu Qingyang 2
http://orcid.org/0000-0001-9115-9647
Hang Xiaofeng hangxfdoc@smmu.edu.cn
1
Wang Junxue docd1@sina.com
1
1 grid.73113.37 0000 0004 0369 1660 Department of Infectious Diseases, Second Affiliated Hospital of Naval Medical University, Shanghai, 200003 China
2 grid.73113.37 0000 0004 0369 1660 Department of Radiology, Second Affiliated Hospital of Naval Medical University, Shanghai, 200003 China
Handling Editor: Pablo Pineyro .
20 5 2023
2023
168 6 16429 9 2022
5 3 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
An outbreak of COVID-19 in Shanghai, China, in March 2022 was caused by the Omicron variant. The epidemic lasted for more than 3 months, and the cumulative number of infected people reached 626,000. We investigated the impact of clinical factors on disease outcomes in patients with COVID-19. Using a case-control study design, we examined cases from fever clinics with confirmed Omicron variant infection, analyzed their population and laboratory diagnostic characteristics, and provided theoretical support for subsequent epidemic prevention and control. Logistic regression was used to identify factors associated with infection with the Omicron variant. The results of this study show that the COVID-19 vaccine can protect against infection with the Omicron variant, and more than 50% of infected people had not been vaccinated. Compared with the epidemic in Wuhan 2 years ago, most of the patients in the hospital in the Shanghai epidemic had underlying diseases (P = 0.006). A comparison of patients infected with the Omicron variant in Shanghai and patients with other respiratory tract infections showed no significant difference in the levels of neutrophils, lymphocytes, eosinophils, white blood cells, hemoglobin, or platelets (P > 0.05). People over 60 years old and those with underlying diseases were at risk for pneumonia (OR = 14.62 (5.49-38.92), P < 0.001; OR = 5.29 (2.58-10.85), P < 0.001, respectively), but vaccination was a protective factor (OR = 0.24 (0.12-0.49), P < 0.001). In summary, vaccination has a potential effect on infection with Omicron variant strains and provides protection against pneumonia. The severity of illness caused by the Omicron variant in 2022 was significantly lower than that of the original SARS-CoV-2 variant from two years previously.
General Program of Huoshenshan Hospital zhuanyuan [2020]52 Wang Junxue issue-copyright-statement© Springer-Verlag GmbH Austria, part of Springer Nature 2023
==== Body
pmcIntroduction
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron variant was first detected in South Africa on November 9, 2021 [1]. It then rapidly spread to the United States, Europe, and other regions and was declared by WHO to be the fifth "variant of concern" [2]. The Omicron variant spread widely in March 2022 in Shanghai, China [3], with its first local case detected March 2, 2022, and the cumulative number of SARS-CoV-2 infections reaching more than 626,000 [4, 5]. The Omicron epidemic began earlier in Shanghai than in 71 other cities in 21 provinces in China, so it is of great significance to analyze the characteristics of this epidemic in Shanghai. Omicron lineage BA.2 was primarily responsible for the coronavirus outbreak in Shanghai in the first half of 2022, which had a severe disease rate of 0.27% and a case fatality rate of 0.09% [6]. Despite vaccination, the elderly are at significant risk of progression to severe disease or death due to infection with Omicron. At present, residents of the Chinese mainland have three inactivated vaccine COVID-19 vaccines available (China's biological vaccine, Wuhan biological vaccine, and Beijing CoronaVac), as well as an adenovirus-vectored vaccine (Ad5-nCoV). Although the overall vaccination coverage of the population of Shanghai exceeds 90%, only 62% of the 5.8 million people over 60 years old have been vaccinated, and only 38% of the population has received a booster vaccine [7]. Despite the lower virulence of the Omicron BA.2 sublineage, the risk of progression to severe disease or death after infection remains high [8].
Earlier clinical studies on wild strains of SARS-CoV-2 showed that laboratory indicators of infection, such as lymphocyte count, T lymphocyte count, liver enzyme index, myocardial enzyme index, and D-dimer, are useful for predicting whether patients develop severe disease [9, 10]. The Omicron variant strain has been reported to have a shorter incubation period, higher infectivity, and pathogenicity and to produce a higher viral load than the original strain [11]. The local outbreak in Shanghai occurred after widespread vaccination of the population against COVID-19.
We compared the characteristics of patients in Shanghai fever clinics and patients in Wuhan in 2020. The general profiles of patients diagnosed with Omicron infection and those hospitalized due to influenza during the same period were compared. The susceptibility factors for pneumonia caused by Omicron variant strains were analyzed, and the protective effect of COVID-19 vaccination on patients infected with Omicron variant strains was assessed. This study will provide a reference for prevention and control of COVID-19.
Materials and methods
General information
COVID-19-positive patients attending the infection department of Shanghai Changzheng Hospital between 1 March 2022 and 30 May 2022 were included in this study. A total of 316 target cases were included in the initial stage. After screening by the inclusion and exclusion criteria, 144 patients were finally included in the study group (SH-C group), and 144 patients with similar clinical symptoms who visited a doctor due to other upper respiratory tract infections, such as influenza virus, adenovirus, or respiratory syncytial virus, were included as a control group (SH-N group) (Fig. 1).Fig. 1 Study flow chart
In addition, we retrospectively collected information on 173 positive cases of COVID-19 in Wuhan from February 1, 2020, to May 30, 2020, as a historical control (WH-C group). This retrospective study, supported by the General Program of Huoshenshan Hospital, complies with the Declaration of Helsinki, and all patients gave informed consent.
Diagnostic criteria
The diagnostic criteria for positive cases of COVID-19 in this study are based on the "COVID-19 Pneumonia Diagnosis and Treatment Program (Trial Version 9)" issued by the General Office of the National Health Commission [12]. The clinical classification of COVID-19-positive cases is also based on the above scheme, and the research subjects were divided into mild and severe groups.
Collection of information
The hospitalization history was collected, and general data (including epidemiological history, gender, age, underlying diseases), clinical symptoms, laboratory indicators, diagnosis, treatment, and outcomes were retrospectively analyzed and grouped by type for comparison. The flow chart for patient enrollment is shown in Fig. 1.
Statistical methods
Data were processed using SPSS 24.0. Enumeration data were expressed as cases (%), and the χ2 test or Fisher's exact test was used. Measurement data that conformed to a normal distribution were expressed as the mean ± standard deviation (x ± s), and were analyzed by t-test; those that did not exhibit a normal distribution were expressed as the median and interquartile interval (M, median; P25, 25 percent, P75, 75 percent), and the Mann-Whitney U test was used. Logistic regression was used to analyze the influencing factors of outcome variables, and the odds ratio (OR) represents the association strength. P < 0.05 indicates statistical significance.
Results
General information about the enrolled patients
A total of 461 subjects were enrolled in this study and divided into three groups: the 2022 Shanghai COVID-19 group (SH-C), the 2022 Shanghai non-COVID-19 group (SH-N), and the 2020 Wuhan COVID-19 group (WH-C). The general information about the three groups of subjects is shown in Table 1. Among the population included in this study, the WH-C group had the highest mean age (60.4 years), and SH-N had the lowest mean age (42.7 years). There was no significant difference in the male-female ratio among the three groups. The etiology of the three groups was significantly different, namely Omicron variant, respiratory infection, and the original SARS-CoV-2 strain. The rate of vaccination against COVID-19 in the SH-C group was significantly lower than in the SH-N group (49.65% vs. 72.73%, P < 0.001).Table 1 General information about the patients in the study
SH-C, n = 144 SH-N, n =144 WH-C, n = 173
Age, y 57.5 ± 22.4 42.7 ± 18.6 60.4 ± 12.4
Gender, F/M 81/63 77/67 72/101
Cause of disease Omicron Other upper respiratory tract infections SARS-CoV-2
Vaccination ratea 49.65% 72.73% 0
SH-C Shanghai COVID-19 group, SH-N non-COVID-19 group (other respiratory infections), WH-C Wuhan COVID-19 group (2020)
aPercentage who had received at least one vaccine dose
Differences between the Shanghai epidemic and the Wuhan epidemic
The COVID-19 outbreak in Shanghai was fundamentally different from the outbreak in Wuhan that occurred two years previously. Among the hospitalized patients, 34.0% of the patients in the SH-C group were less than 50 years old, which was a significantly higher percentage than that in WH-C group. There was no statistical difference in the percentage of each sex among the people with COVID-19 during the different periods. The mean body temperature of the SH-C group was 37.5℃, whereas in the WH-C group it was 38.0°C, and this difference was statistically significant (P < 0.001). In addition, the mean leukocyte count and incidence of pneumonia and symptoms were higher in the SH-C group than in the WH-C group (P < 0.05) (Table 2). The overall prevalence of underlying disease in the SH-C group was also higher than that in the WH-C group (P = 0.006). Ranked from high to low by prevalence, these diseases included hypertension, cerebrovascular disease, diabetes, heart disease, post-operative illness, liver disease, kidney disease, and autoimmune disease. The prevalence of heart disease was statistically different between the two groups (P = 0.032). However, no statistical differences were found among the groups for the other underlying diseases due to their low prevalence rates (P > 0.05) (Table 2). Underlying diseases often increase the impact of COVID-19 on the health of the patient.Table 2 Comparison of hospital admissions of patients with COVID-19 in Shanghai and Wuhan
Factor SH-C, n = 144 (%) WH-C, n = 173 (%) P value
Age <0.001
< 50 49 (34.0) 29 (16.8)
50-60 19 (13.2) 45 (26.0)
≥ 60 76 (52.8) 99 (57.2)
Gender 0.702
Female 81 (56.3) 101 (58.4)
Male 63 (43.8) 72 (41.6)
Body temperature (℃) 37.5 ± 1.0 38.0 ± 0.9
Leukocytes 6.92 ± 2.85 6.02 ± 2.88
Pneumonia < 0.001
No 81 (56.3) 148 (85.5)
Yes 63 (43.8) 25 (14.5)
Symptoms < 0.001
Mild 61 (42.4) 104 (60.1)
Severe 83 (57.6) 69 (39.9)
Underlying illness 0.006
Hypertension 31 (21.5) 26 (15.0) 0.134
Cerebrovascular 16 (11.1) 12 (6.9) 0.192
Diabetes 14 (9.7) 13 (7.5) 0.483
Heart disease 12 (8.3) 5 (2.9) 0.032
Post-operative illness 6 (4.1) 7 (4.0) -
Liver disease 4 (2.7) 5 (2.9) -
Kidney disease 4 (2.7) 2 (1.1) -
Autoimmune disease 1 (0.7) 0 (0) -
SH-C patients with COVID-19 in Shanghai in 2022, WH-C patients with COVID-19 in Wuhan in 2020
Differences between COVID-19 and other respiratory infections
Comparative analysis showed that there was no significant difference in body temperature between patients with COVID-19-related pneumonia and those with other respiratory tract infections in Shanghai (P = 0.099). On average, patients with COVID-19 in Shanghai had lower white blood cell counts (WBCs) and lower C-reactive protein (CRP) levels than those with other respiratory tract infections, and the difference was statistically significant (P < 0.05). There was no significant difference in neutrophil rate (N%), lymphocyte rate (L%), eosinophils, red blood cell (RBC) count, hemoglobin (HB), or platelet (PLT) count between patients with COVID-19 and those with other respiratory tract infections (P > 0.05). However, monocytes were higher in patients with COVID-19 than in patients with other respiratory tract infections (P < 0.001) (Fig. 2).Fig. 2 Comparative analysis of patients with COVID-19-related pneumonia and respiratory tract infections in Shanghai. WBC white blood cell count, N% neutrophil rate, L% lymphocyte rate, RBC red blood cell count, HB hemoglobin, PLT platelet count, CRP C-reactive protein
Analysis of risk factors for pneumonia caused by Omicron
To identify risk factors for COVID-19-related pneumonia caused by Omicron, we use the internal control method. Among the 144 patients with COVID-19, 63 had pneumonia and 81 did not. Univariate analysis showed that age over 60 years was a risk factor for COVID-19-related pneumonia (OR = 14.62 (5.49-38.92), P < 0.001). Vaccination was a protective factor (OR = 0.24 (0.12-0.49), P < 0.001), and people with underlying diseases were more likely to develop pneumonia (OR = 5.29 (2.58-10.85), P < 0.001). No significant difference in body temperature was observed between pneumonia patients and non-pneumonia patients (P > 0.05). However, in terms of WBC and CRP indicators, the abnormal rate was significantly higher in pneumonia patients than in the control group (P < 0.05) (Table 3).Table 3 Comparison of factors related to pneumonia caused by the epidemic of COVID-19 in Shanghai in 2022
Factor Non-pneumonia, n = 81 Pneumonia, n = 63 Univariate analysis OR (95% CI) P value
Age
< 50 43 6 1
50-60 13 6 3.31 (0.91-12.02) 0.069
≥ 60 25 51 14.62 (5.49-38.92) <0.001
Gender
Female 49 32 1
Male 32 31 1.48 (0.76-2.88) 0.768
Vaccine
No 28 42 1
Yes 52 19 0.24 (0.12-0.49) <0.001
Underlying illness
No 55 18 1
Yes 26 45 5.29 (2.58-10.85) <0.001
Body temperature
Normal 49 31
Abnormal 32 32 - 0.176*
WBC
Normal 71 48
Abnormal 5 15 - 0.004*
Monocytes
Normal 32 34
Abnormal 44 29 - 0.163*
CRP
Normal 54 28
Abnormal 22 35 - 0.001*
*Chi-square test results; normal range: body temperature, < 37.3℃;, White blood cell count (WBC), 3.5-9.5 × 109/L; monocytes, 3-10%;, C reactive protein (CRP), 0-10 mg/L
Discussion
The Omicron variant was first discovered in South Africa in November 2021 and soon spread around the world [13]. On June 2, 2022, the Shanghai Municipal Government Information Office reported that gene sequencing had confirmed that the Shanghai epidemic was caused mainly by the Omicron BA.2 and BA.2.2 variants [5, 14].
There was a larger proportion of the patients treated in the hospital during the Shanghai epidemic who had underlying diseases than in Wuhan two years previously. Most patients who come to the fever clinic have fever or other symptoms of respiratory tract infection, whereas many individuals with COVID-19 with no obvious symptoms and are identified by screening and sent to mobile hospital units.
In the comparative analysis of patients with COVID-19-related pneumonia and other respiratory tract infections in Shanghai, we found that there was no significant difference in body temperature between patients with COVID-19-related pneumonia and those with other respiratory tract infections. This shows that although the Omicron variant spreads rapidly and has a wide range, its symptoms are similar to those of other respiratory tract infections. Among people with COVID-19, age over 60 years old and having underlying diseases are risk factors for the occurrence of pneumonia, but vaccination against the Omicron variant is an important protective factor.
According to China's "Zero COVID-19" policy, active surveillance, epidemiological investigation, and timely nucleic acid testing can help to facilitate the rapid detection of new cases, minimize the scope of screening, and stop transmission [15]. Furthermore, enhanced health management of the elderly, especially those with underlying diseases, may help to reduce the incidence of severe and critical pneumonia [16].
Conclusions
The outbreak of local cases in Shanghai was caused by the Omicron variant, and vaccination with the COVID-19 vaccine provides protection against the development of severe disease after Omicron infection. Since this study did not include investigation of dynamic changes in laboratory parameters of patients during the treatment process, this needs to be addressed in future follow-up and research work.
Acknowledgements
This work was funded by the General Program of Huoshenshan Hospital (zhuanyuan [2020]52).
Author contributions
Huilan Tu, Sheng Tu, and Shiqi Gao wrote the paper; Huilan Tu made the figures and tables; Anwen Shao and Jifang Sheng revised the paper.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Declarations
Conflict of interest
The authors state that there was no conflict of interest in the preparation of this study.
Ethical approval
The study protocol conformed to the ethical guidelines of the 2013 Helsinki Declaration and was approved by the Institutional Ethical Review Board of Second Military Medical University. Each study participant provided written informed consent.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yuanjing Zhang, Jianrong Wang and Ying Xie have contributed equally to this work.
==== Refs
References
1. He X Hong W Pan X Lu G Wei X SARS-CoV-2 Omicron variant: Characteristics and prevention MedComm (2020) 2021 2 4 838 845 10.1002/mco2.110 34957469
2. Cascella M, Rajnik M, Aleem A, Dulebohn SC, Di Napoli R (2022) Features, evaluation, and treatment of coronavirus (COVID-19). In: StatPearls [Internet]. StatPearls Publishing, Treasure Island
3. Cai J Deng X Yang J Sun K Liu H Chen Z Modeling transmission of SARS-CoV-2 Omicron in China Nat Med 2022 28 7 1468 1475 10.1038/s41591-022-01855-7 35537471
4. Huang L Adjusted control rate closely associated with the epidemiologic evolution of the recent COVID-19 wave in Shanghai, with 94.3% of all new cases being asymptomatic on first diagnosis J Infect 2022 10.1016/j.jinf.2022.07.010 36539043
5. Chen Z, Deng X, Fang L, Sun K, Wu Y, Che T et al (2022) Epidemiological characteristics and transmission dynamics of the outbreak caused by the SARS-CoV-2 Omicron variant in Shanghai, China: a descriptive study. medRxiv [Preprint]. 2022. 10.1101/2022.06.11.22276273
6. Chen X Yan X Sun K Zheng N Sun R Zhou J Estimation of disease burden and clinical severity of COVID-19 caused by Omicron BA.2 in Shanghai, February–June 2022 Emerg Microbes Infect 2022 11 1 2800 2807 10.1080/22221751.2022.2128435 36205530
7. Zhang X Zhang W Chen S Shanghai's life-saving efforts against the current omicron wave of the COVID-19 pandemic Lancet 2022 399 10340 2011 2012 10.1016/S0140-6736(22)00838-8 35533708
8. Cheung PH Chan CP Jin DY Lessons learned from the fifth wave of COVID-19 in Hong Kong in early 2022 Emerg Microbes Infect 2022 11 1 1072 1078 10.1080/22221751.2022.2060137 35348429
9. Ochani R Asad A Yasmin F Shaikh S Khalid H Batra S COVID-19 pandemic: from origins to outcomes. A comprehensive review of viral pathogenesis, clinical manifestations, diagnostic evaluation, and management Infez Med 2021 29 1 20 36 33664170
10. Ai J Zhang H Zhang Y Lin K Zhang Y Wu J Omicron variant showed lower neutralizing sensitivity than other SARS-CoV-2 variants to immune sera elicited by vaccines after boost Emerg Microbes Infect 2022 11 1 337 343 10.1080/22221751.2021.2022440 34935594
11. Wu J Wei Y Shen F Zhu S Lu Y Tian X Vaccination is associated with shorter time to target cycle threshold value in patients with SARS-CoV-2 omicron variant Front Cell Infect Microbiol 2022 12 943407 10.3389/fcimb.2022.943407 35873164
12. National Health Commission. COVID-19 diagnosis and treatment plan (trial version 7) [BE/OL]. http://www.nhc.gov.a/yzygj/s7653p/202003/46c9294a7dfe4cef80dc7f5912eb1989.shtml,2020-03-03
13. Aleem A, Akbar Samad AB, Slenker AK (2022) Emerging variants of SARS-CoV-2 and novel therapeutics against coronavirus (COVID-19). In: StatPearls [Internet]. StatPearls Publishing, Treasure Island
14. Guo Y Han J Zhang Y He J Yu W Zhang X SARS-CoV-2 omicron variant: epidemiological features, biological characteristics, and clinical significance Front Immunol 2022 13 877101 10.3389/fimmu.2022.877101 35572518
15. Shi Y Jiang HL Yang MX Dong LJ Chen Y Zhou YB The precision of epidemiological investigation of COVID-19 transmission in Shanghai, China Infect Dis Poverty 2021 10 1 58 10.1186/s40249-021-00849-w 33947468
16. Yu X Sun X Cui P Pan H Lin S Han R Epidemiological and clinical characteristics of 333 confirmed cases with coronavirus disease 2019 in Shanghai, China Transbound Emerg Dis 2020 67 4 1697 1707 10.1111/tbed.13604 32351037
|
PMC010xxxxxx/PMC10199488.txt |
==== Front
Epidemics
Epidemics
Epidemics
1755-4365
1878-0067
The Authors. Published by Elsevier B.V.
S1755-4365(23)00024-5
10.1016/j.epidem.2023.100688
100688
Article
Influence of setting-dependent contacts and protective behaviours on asymptomatic SARS-CoV-2 infection amongst members of a UK university
Fairbanks Emma L. ab
Bolton Kirsty J. b⁎
Jia Ru c
Figueredo Grazziela P. d
Knight Holly c
Vedhara Kavita c
a School of Veterinary Medicine and Science, University of Nottingham, United Kingdom
b School of Mathematical Sciences, University of Nottingham, United Kingdom
c School of Medicine, University of Nottingham, United Kingdom
d School of Computer Science, University of Nottingham, United Kingdom
⁎ Corresponding author.
20 5 2023
6 2023
20 5 2023
43 100688100688
7 11 2021
26 3 2023
12 5 2023
© 2023 The Authors
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
We survey 62 users of a university asymptomatic SARS-CoV-2 testing service on details of their activities, protective behaviours and contacts in the 7 days prior to receiving a positive or negative SARS-CoV-2 PCR test result in the period October 2020–March 2021. The resulting data set is novel in capturing very detailed social contact history linked to asymptomatic disease status during a period of significant restriction on social activities. We use this data to explore 3 questions: (i) Did participation in university activities enhance infection risk? (ii) How do contact definitions rank in their ability to explain test outcome during periods of social restrictions? (iii) Do patterns in the protective behaviours help explain discrepancies between the explanatory performance of different contact measures? We classify activities into settings and use Bayesian logistic regression to model test outcome, computing posterior model probabilities to compare the performance of models adopting different contact definitions. Associations between protective behaviours, participant characteristics and setting are explored at the level of individual activities using multiple correspondence analysis (MCA). We find that participation in air travel or non-university work activities was associated with a positive asymptomatic SARS-CoV-2 PCR test, in contrast to participation in research and teaching settings. Intriguingly, logistic regression models with binary measures of contact in a setting performed better than more traditional contact numbers or person contact hours (PCH). The MCA indicates that patterns of protective behaviours vary between setting, in a manner which may help explain the preference for any participation as a contact measure. We conclude that linked PCR testing and social contact data can in principle be used to test the utility of contact definitions, and the investigation of contact definitions in larger linked studies is warranted to ensure contact data can capture environmental and social factors influencing transmission risk.
Keywords
SARS-CoV-2
Asymptomatic infection
Universities
Contact patterns
Transmission risk
Protective behaviours
Mask wearing
Social distancing
Hand washing
==== Body
pmc1 Introduction
In the 2020/2021 academic year 37% of 18-year-olds in the UK were offered a higher education place (Bolton, 2021) and altogether approximately 2.5 million students are registered in higher education in the UK across over 160 providers (Higher Education Statistics Agency, 2021). Universities provide much of this education; many comprising of order tens of thousands students and staff (Higher Education Statistics Agency, 2021) with highly connected communities through teaching, research, leisure and residential networks. Prior to the emergence of SARS-CoV-2, there was limited data available on the contact networks of university members, nonetheless preliminary modelling studies flagged universities as settings of potential high risk for SARS-CoV-2 transmission (Hill et al., 2021, Brooks-Pollock et al., 2021). Despite mitigations to reduce transmission risk, many universities in the UK (UCU, 2020) experienced outbreaks of SARS-CoV-2 at the beginning of the 2020/2021 academic year (UCU, 2020), some of which may have amplified infection rates in their local community (Enright et al., 2021). Students in halls of residence were noted to be at higher risk of experiencing SARS-CoV-2 infection (Children’s Task and Finish Group, 2021), however insight into the risk associated with other activities undertaken by university members is limited.
Contact diary studies have proved useful for measuring contact rates to parameterise epidemic models in structured populations (Mossong et al., 2008). However, the most relevant contact definition remains uncertain. For a pathogen with potential for aerosolised and fomite transmission, such as SARS-CoV-2 (Klompas et al., 2020), contacts need not be close or conversational contacts, as typically measured in many contact surveys (e.g. Mossong et al., 2008). Furthermore, contact networks may be modified by the adoption of protective behaviours such as mask wearing, social distancing and hand washing (Golding et al., 2023). It is thus of interest to examine the role of protective behaviours in concert with a broad definition of contact when surveying the potential transmission risk of a particular activity. For diseases with high rates of asymptomatic infection, including SARS-CoV-2 (Oran, 2021), it is difficult to examine the utility of contact measures for estimating transmission risk using commonly collected epidemiological data streams based on symptomatic status.
In this study we were motivated to understand the SARS-CoV-2 infection risk of university activities for staff and students in October 2020 to March 20201. To this end we link asymptomatic SARS-CoV-2 testing data with quantitative data on social interactions while on and off campus in the week preceding an asymptomatic SARS-CoV-2 PCR test result. Given the restrictions on social mixing in place during the study period, and the uncertainty around the role of these in mitigating transmission risk, our contact survey records information about the protective behaviours adopted in each activity. We consider the utility of different contact measures, and the potential role of protective behaviour, in explaining the setting-specific infection risk in our participants. Our analysis is structured as follows. In Section 3.1 we provide summary statistics for our data. We then explore the ability of different individual contact definitions, that variably account for the duration, number of contacts, and presence of extra-household members in each setting to explain asymptomatic SARS-CoV-2 test outcome amongst participating staff and students (Section 3.2.1). We present the inferred associations between setting-specific contact and infection risk in Section 3.2.2. To investigate why interactions in particular settings may present enhanced infection risk we pool activities across individuals and consider correlations between protective behaviours, setting type and participant characteristics (Section 3.3).
2 Methods
2.1 Data collection and curation
Participant enrolment was based on a convenience sample of university staff and students returning SARS-CoV-2 positive saliva samples via the Nottingham Asymptomatic Testing (NATS) service. For each participant testing positive, we randomly invited a consenting NATS user with the same university role (staff or student) who had returned a negative test result. Ethical approval was obtained via the University of Nottingham Faculty of Medicine and Health Sciences ethics board (FMHS 96-0920). Participants provided informed consent. A copy of the survey is available upon request.
We divide participants into three groups; those who tested positive (positive), tested negative and had never had a positive test (negative) and those who tested negative but when surveyed had previously tested positive (previous). Unless stated otherwise the analysis is performed on the individuals in the groups positive and negative, minimising any bias due to the impact of SARS-CoV-2 immunity or assumed immunity on susceptibility and behaviour.
Participants were asked to recall information about social interactions and protective behaviours in each activity outside their home undertaken 7 days preceding the receipt of an asymptomatic PCR SARS-CoV-2 test result using a structured interview, offered in person or online. This interview was developed by two psychologists (KV & HK) and piloted prior to use. If prompting was needed, participants are encouraged to check their calendars or social media feed. Protective behaviours include whether the participant wore a mask, socially distanced (over 2 m away from possible contacts) and cleaned (washed or hand sanitised) their hands before and/or after each activity. Activities are assigned one of twelve settings. These are; abroad/aeroplane, campus other, exercise, hospitality, non-university work, non-private travel, other, research, retail, social, teaching and testing (see A1, Supplementary Information). Survey questions were motivated to capture adherence to pre-July 2021 guidelines for COVID-secure workplaces (Department for Business, Energy & Industrial Strategy and Department for Digital, Culture, Media & Sport, 2020).
Participants are prompted to recall each transition to a new activity and estimate the number of people present (0, 2–5, 5–10, 10–20, 20–50, 50–100, 100+) and duration of the activity. Note that we use the term contact in the broad sense of others present in the same setting, unlike many social contact surveys that assume contacts involve touch or conversation (e.g. Mossong et al., 2008). To capture social contact behaviour that could plausibly result in transmission via combinations of droplets, aerosol and/or fomites, we consider seven contact definitions when constructing setting-specific contact measures over the 7 day survey period: participation in the setting, the number of distinct activities, total contacts (the sum of the mid-points of estimated contacts during each activity), total duration of activities, and person-contact-hours (PCH) calculated as the summed product of the midpoint of the estimated contacts and the duration (in hours) of each activity, the total contacts not including the participant’s household members and PCH not including the participant’s household members. We conservatively set a maximum contact number of 100 for the 100+ option when computing contact measures. We assume that the number of possible household contacts is equivalent to the participant’s household size. Therefore, if the number of household contacts is not given for an activity we calculate the non-household contacts and PCH as the difference between the maximum possible household contacts and reported contacts, providing a lower limit on the non-household contacts.
2.2 Setting-specific contact measures associated with asymptomatic SARS-CoV-2 PCR test result
We use a logistic regression model to regress asymptomatic SARS-CoV-2 test result on contact measures. Due to separation issues in the data (all participants who visited the aeroplane/abroad or non-university work setting tested positive) a Bayesian logistic regression is performed in Stan (2019) with a logit link function. As recommended by Gelman et al. (2008) priors for the logistic regression coefficients are assumed to follow independent Cauchy distributions centred at 0 and with scale parameter 10 for the constant term and 2.5 for all other coefficients (Gelman et al., 2008). Prior to fitting the data the binary input (whether a setting was visited) was transformed to have mean 0 and the numeric inputs (all others) are scaled to have mean 0 and standard deviation 0.5 (Gelman et al., 2008).
We assess the significance of regression covariates using Bayes factors. Since the model posteriors are sensitive to the prior a support interval (SI) is computed for the coefficient for each setting-specific contact measure (Wagenmakers et al., 2020). SIs provide information regarding the change in the credibility of values from the prior to the posterior indicating which values of a parameter gain support. Here we present values receiving ‘moderate support’, with a Bayes factor larger than 3, using the bayestestR package (Makowski et al., 2019). A leave-one-out error analysis is performed on the bounds of the SIs.
We estimate the marginal likelihoods and posterior model probabilities (PMPs) for the logistical model for each set of setting-specific contact measures generated by a contact definition using the bridgesampling package (Gronau et al., 2020). The PMPs are rescaled to sum to 1 across models considered. To examine the support for each contact definition we compute PMPs, averaging over 10 repetitions of the bridge sampling procedure to obtain an empirical estimate of the estimation uncertainty. For the model associated the largest PMP we present posterior predictive checks for the positive and negative groups, and generate an out of sample prediction for test outcome in the previous group.
Fig. 1 The proportion of participants who took part in an activity, as well as the mean number of activities, duration spent, contacts, non-household contacts, PCH and non-household PCH in each setting, by test result.
2.3 Protective behaviours
To examine whether protective behaviours performed during an activity are influenced by the setting or environment we consider the relationship between these behaviours and the university role, gender, age and SARS-CoV-2 test result. As protective behaviours vary between activities even for individuals in the same setting, we pool individual data for this analysis. Each activity is described by nine properties; (i) age of the participant, (ii) gender of the participant, (iii) role (UG, PG, or employed), (iv) test result, (v) setting, (vi) environment (outdoors, ventilated indoors or unventilated indoors), (vii) did the participant wear a mask, (viii) did the participant socially distance at all times and (ix) did the participant use hand sanitiser or wash their hands before and/or after? To assess room ventilation participants were given the example of a room with doors and/or windows open being ventilated.
We perform a multiple correspondence analysis (MCA) in RStudio Team (2020) using the FactoMineR package (Husson et al., 2008) to examine the relationship between these properties. Since responses were not given for all activity properties, the missMDA package was used to estimate the number of dimensions for the MCA by leave-one-out cross-validation and impute missing values by cross-validation. Age is denoted a quantitative supplementary variable and gender, role, test result and setting as qualitative supplementary variables. The MCA is performed on the remaining ‘active’ variables (environment, mask wearing, social distancing and hand washing). Coordinates for the supplementary variables are predicted using the information from this MCA. The dimdesc function is used to determine which categorical variables best describe each dimension and whether age (the continuous variable) is correlated to each dimension. For the quantitative variable, age, correlation coefficients are calculated. For the categorical variables, a univariate ANOVA model is performed for each variable and dimension. An F-test examines whether each variable influences the dimension.
Fisher’s exact test is performed on each pair of variables to determine whether they are significantly linked, with p-values corrected for multiple comparisons using the Benjamini–Hochberg method (Benjamini and Hochberg, 1995). Results are compared to the Bayesian logistic regression and used to verify consistency of the MCA analysis.
3 Results
3.1 Descriptive statistics
Participants are predominantly students between 18 and 30, however staff up to 60 years old are represented. The majority of participants completed the survey online (50, 23 positive and 27 negative), the remaining 12 (9 positive, 3 negative) completed the interviews in person. In all we have data on 447 distinct activities from the 62 participants. There were 20 participants in the positive group, 29 participants in the negative group and 13 in the previous group. The test result dates of the positive group are skewed towards the early period of NATS operation, reflecting the epidemic of self-reported PCR-confirmed SARS-CoV-2 infection within the University (UCU, 2020) (see Figure A1(a), Supplementary Information).
The mean and standard deviation of each setting-specific contact measure is provided in Table A1 (Supplementary Information). Retail settings were visited by the largest proportion of participants and had the highest mean non-household contacts, with exercise the most frequently reported activity. Research settings had the highest mean activity duration and mean non-household PCH, with teaching the highest mean PCH. Fig. 1 shows the proportion of individuals who participated, mean number of activities, mean total contacts, mean total duration, mean total PCH, mean non-household contacts and mean-non-household PCH for each setting by test result. Of interest, the mean number of contacts across all activities was highest in the negative group, but mean PCH and non-household PCH was higher in the positive group. A Kruskal–Wallis test showed no significant differences between the positive, negative and previous groups for total distinct types of activity (χ2 = 0.17, df = 2, p = 0.92), number of activities (χ2= 4.93, df = 2, p = 0.08), number of contacts (χ2= 0.76, df = 2, p = 0.69), duration of activities (χ2 = 1.91, df = 2, p = 0.38) or PCH (χ2 = 0.44, df = 2, p = 0.80). There was however a significant difference between the non-household contacts (χ2 = 9.71, df = 2, p = 0.008) and non-household PCH (χ2 = 6.78, df = 2, p = 0.03). The mean household size of participants who tested positive was 2.9 (sd = 2.6), whereas the mean household size of participants who tested negative was 3.1 (sd = 2.0).
Protective behaviours reportedly practised in activities by setting are summarised in Table 1. The percentage of participants who provided answers for whether they wore a mask or socially distanced was the same for all activities except teaching. In teaching settings the participant was only asked about socially distancing. However, for 78% of teaching activities students stated that they wore a mask in the “additional comments” free text field. At the time of this study mask wearing was compulsory (unless exempt) during teaching activities at the university.
Table 1 The percentage of activities in each setting where a protective behaviour was performed. Data given are percentage who wore a mask, socially distanced (SD) at all times and when they washed their hands and the percentage of activities which these questions were answered (Ans.). Here individuals who washed their hands both are included in the percentage of people who washed their hands before and after.
Setting Respiratory protection Hand cleaning
Mask SD Ans. No Before After Both Ans.
Abroad/Aeroplane (12) 0 0 75 92 8 8 8 100
Campus other (19) 74 74 100 6 94 94 94 95
Exercise (72) 41 79 99 9 57 91 57 97
Hospitality (13) 15 23 100 15 77 54 46 100
Non-private travel (26) 96 31 100 54 46 46 46 100
Non-university work (7) 100 14 100 100 100 100 100 100
Other (30) 53 70 100 17 60 80 57 100
Research (63) 89 62 97 2 89 98 80 98
Retail (51) 95 39 86 2 63 94 59 100
Social (23) 5 45 87 26 57 74 57 100
Teaching (27) – 54 96 0 100 96 96 96
Testing (19) 82 82 89 5 89 95 89 100
3.2 Contact measures associated with asymptomatic SARS-CoV-2 PCR test result
3.2.1 Comparison of models with different contact definitions
Table A2 (Supplementary Information) shows the median, minimum and maximum marginal likelihood estimates PMP for each contact definition. The narrow range of the marginal likelihood estimates indicates that the estimation uncertainty is small. The model adopting participation as a contact definition had the most support (PMP = 0.38), followed number of activities (PMP = 0.23) and the non-household PCH (PMP = 0.15). The duration of activities and PCH received less support (0.09 and 0.08, respectively). The models with the least support were the number of contacts and non-household contacts in each setting (PMP = 0.03).
3.2.2 Associations between setting-specific contact measures and test outcome
Fig. 2 gives the coefficient ranges within the SI for each covariate and contact measure, indicating values that received moderate support. We observe that the model with the most support, using whether participants entered each setting as the contact measure, has SIs which are all positive for the covariates aeroplane/abroad and non-university work and all negative for the covariates research and teaching. Reassuringly, more generally, the trends in distributions of the SIs appear similar for the five models with the most support (Section 3.2.1), however the distribution for the two models for the least support (contact definitions contacts and non-household contacts) are sometimes at odds with the SI for the other models. We discuss the consistency of SIs across contact definition models further below.
The SIs for household size, campus other, hospitality, retail and testing straddled zero, or were non-existent, for all contact definition models. For settings that had a strictly positive or negative SIs for at least one contact definition model, none of the models contradicted each other (i.e. with strictly positive SIs for one model and strictly negative SIs in another). When considering settings that are inferred as significant in the best-performing model, aeroplane/abroad and teaching had median and SIs with consistent sign across models, but moderate support was not found for any values for aeroplane/abroad in some models. In contrast the other significant settings, non-university work and research, had SIs that straddled zero in some of the poorer performing models. However, we note that these SIs barely traversed zero (the moderate support for non-university work had a lower bound of −0.03 and research had an upper bound of 0.02).
For the five best-performing models coefficients in the SI for exercise is skewed towards negative values, and these are strictly negative for the models adopting the number of activities and the duration as the contact measure. Similarly, for all models non-private travel SI distributions appear to mostly contain negative values, but SIs are strictly negative for models with non-household PCH and PCH as the contact measure. The covariate “Other” has an all negative SI for contacts, however this model had poor support. The models adopting the duration or PCH as the contact measure yield an all positive SI for social activities. For the best five performing models coefficients in the SI for social are skewed towards positive values.
It is clearly of interest to understand the extent to which SIs are driven by the activities of individual participants, and we explore this in a leave-one-out analysis (A4.2, Supplementary Information). In brief, the model with the largest PMP (based on the participation contact definition) displayed stable SIs across the leave-one-out deletions, but models with low PMPs have large standard deviations in the bounds for the SIs, and signs of the SI bounds are not always consistent with those obtained in the main analysis. Further model checks based on the Fisher’s analysis and the posterior predictive checks for the group previously testing positive are provided in §A4 (Supplementary Information).
Fig. 2 Median (circles) of the covariate coefficients for each contact definition. Support intervals (SIs) give ranges of parameters with a Bayes factor larger than or equal to 3, interpreted as moderate support (bars). ∗ = contact definitions include only external non-household contacts. Solid bars represent SIs where all supported values have the same sign. Note that as the Bayes factor depends on the prior, the marginal posterior median does not always lie in the SI.
3.3 Protective behaviours
Altogether, four dimensions are required to explain 70.9% of the variance in the MCA analysis (Figure A4, Supplementary Information). Fig. 3 shows how the active and qualitative variables relate to the first two dimensions of the MCA, and provides evidence that patterns of protective behaviour differ between settings. Significant components for the first four dimensions are provided in Table A5 (Supplementary Information) and results from the Fisher’s exact tests in Table A6 (Supplementary Information).
Fisher’s tests showed positive test results were positively correlated with no hand washing and females, and were negatively correlated with mask wearing, washing hands after and maintaining social distancing at all times. Explanations for the association of test positivity with gender in our sample are considered in §A5 (Supplementary Information).Fig. 3 Graph visualising the coordinates of each variable categories in dimensions 1 and 2. The distance between any points gives a measure of their similarity. Supplementary variables are shown in green and active variables are shown in red. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Of the activities on campus, teaching and retail are associated with wearing a mask and washing hands both before and after. PGs activities are positively correlated with indoor unventilated environments, social distancing and hand washing. Being classified as staff is negatively correlated with hand washing.
Outside of campus-based activities we find that retail was associated with mask wearing and washing hands before and after an activity. Masks are unlikely to be worn in social settings and outdoors. The MCA analysis shows an association between the abroad/aeroplane setting and not cleaning hands before or after travel. The Fisher’s analysis indicates a negative correlation of air travel with wearing a mask and social distancing. Participants in the study that were staff were more likely to participate in travel activities.
4 Discussion
We have presented linked asymptomatic SARS-CoV-2 testing and social contact data for a UK university collected from October 2020 to March 2021, after the initial surge in infections at the beginning of the 2020/2021 academic year (UCU, 2020). During the study period, restrictions on mixing outside of your household/bubble and educational activities were in place, with particularly stringent rules about social contact with those outside of your household in place during tier 3 restrictions (30 October 2020–5 November 2020), and the second (5 November 2020–2 December 2020) and third (6 January 2021–8 March 2021) national lockdowns (UK Government, 2020, UK Government, 2021). Strict social distancing measures were also in place on campus throughout the study period. Teaching was undertaken in a hybrid (online/in-person) manner to accommodate social distancing. During the third national lockdown only students on a limited selection of courses (Medicine & Dentistry, Health & Social Care including Nursing & Midwifery, Physiotherapy and Veterinary Science, Education, and Social Work) were permitted to travel to campus without exemption. Research staff were asked to work from home whenever possible (University of Nottingham, 2021).
Within this context and for the cohort studied, participating in air travel (and holiday activities) and non-university work increases the probability of a positive asymptomatic SARS-CoV-2 PCR test result. This is consistent with evidence that workers with public facing roles are at higher risk of infection (see P.H.E. Transmission Group, 2020 and references therein, Thomas et al., 2021). Students or staff in part-time work in other settings could mediate spillover between the University and community. Encouraging vaccination of these groups could be particularly important for mitigating the risk of university outbreaks in periods of restrictions.
Participating in research and teaching activities at the university was associated with a lower risk of a positive asymptomatic SARS-CoV-2 PCR test result, however no association was found with participation in other activities on campus. Although our data does not permit estimate of the risk attributable to any activity or setting, our results are consistent with teaching and research activities having lower infection risk than other activities noted by participants.
Participation in a setting was the best-performing contact measure, followed by counts of the number of activities in each setting and the non-household PCH (each for 7 days preceding test result). Out of sample model predictions for the group previously testing positive yielded an expected test positivity between that for the negative and positive groups across all contact definitions.
Indications that non-household PCH and duration of activities provide a better model of SARS-CoV-2 PCR test status than contacts or PCH provides tentative evidence for the role of contact duration in infection risk, also reported by Thompson et al. (2021). There are several potential reasons for the difference in model performance between contact definitions, and in particular the superior performance of participation in a setting rather than the definitions capturing contact numbers. Infected people can remain PCR positive for up to 3 weeks following exposure (Sethuraman et al., 2020) depending on the sensitivity of the saliva assay (Teo et al., 2021) and therefore we cannot guarantee we have surveyed participant behaviour during the period of exposure. Additionally, contacts in each activity were assumed to extend for the duration of an activity. We have not collected data on repeated contacts, and it is possible that the participation contact definition is preferred because of this. Contact definitions that capture the proximity and duration of every contact in each activity may perform better. While collecting more detailed contact data was considered prohibitive from a recall perspective in our retrospective survey, similar contact diary studies have been piloted (Bolton et al., 2012), and it would be of interest to link these to repeated asymptomatic test outcome data for SARS-CoV-2 or other respiratory viruses in larger studies with greater statistical power.
We adjusted for participant household size in our regression, but did not find a significant effect. Other studies examining the risk associated with household size have been mixed, with an analysis of setting-dependent transmission risk not identifying household size as significant (Thompson et al., 2021), but recent SARS-CoV-2 prevalence higher in larger households (ONS, 2021). Analysis of contact patterns in another UK university suggested that extra-household contacts were higher amongst those living in smaller households (Nixon et al., 2021), and it is plausible that such an effect could offset the enhanced risk of importation into larger student households. Outbreaks in halls of residence may have been more strongly influenced by hall than household size (Enright et al., 2021) and it is possible that there are other risks associated with residential contacts in this setting that we have not captured.
Positive asymptomatic SARS-CoV-2 PCR test results were negatively associated with mask wearing, social distancing and hand washing, as reported in another case-control study of asymptomatically infected contacts of SARS-CoV-2 cases (Doung-Ngern et al., 2020). Our MCA highlighted differential adoption of protective measures between settings and suggests that protective behaviours can be different in university and non-university activities. Teaching and research setting may be lower risk (despite similar or larger mean contact measures across contact definitions) as they were associated with mask wearing and hand washing. Although mask wearing and hand washing was practised uniformly in non-University work settings, this was less likely to be in a ventilated space with complete adherence to social distancing than other settings. Other studies suggest that adoption of protective behaviours is also determined by psychological factors (Zickfeld et al., 2020, Faasse and Newby, 2020, Wise et al., 2020), which could explain some of the variability between participant behaviour. Measuring the prevalence of micro-distancing behaviour as well as macro-distancing behaviour now widely captured by mobility patterns has shown utility in estimating the effective reproduction number in low-prevalence settings (see, e.g., Golding et al., 2023) and may also aid in parameterising agent-based simulation of transmission (Kerr et al., 2021).
Our results come with a number of important caveats. Our sample size is relatively small, was chosen based on consenting positive cases, and may not be representative of all users of the NATS or the wider University population. A greater proportion of positive cases in our study were from periods with lower levels of restriction which could generate time-varying confounding in our analysis. A larger sample would likely allow for adjustment for this and other potential confounders, and potentially provide statistical power to include setting-specific protective behaviours in the regression model. As discussed elsewhere (Royal Society SET-C, 2020), the opportunities for contact and transmission depend on community prevalence and the social restrictions in place. Data for this study was collected over a period during which there were strong (albeit changing) restrictions on permitted social and travel activities, which may partly explain the absence of a significant effect of social interactions on risk of obtaining a positive test result as reported in other studies (Hobbs et al., 2020). Similarly, occupancy on campus was low during the study with much teaching online, and we expect the relative risk of activities in different settings will change depending on how university and national policies, and the behavioural response to these, evolve. Furthermore, participants were surveyed at a time when the circulating SARS-CoV-2 strain was either phenotypically akin to the original Wuhan strain, or the alpha variant of concern, and it is plausible that different patterns of risk would be observed for delta or other variants with different infectiousness profiles.
Although the structured interview adopted aims to optimise recall of social contact behaviour, the limitations of recalling such details accurately are well documented (e.g. Garry et al., 2021). Participants who received positive test results could be experiencing stress/anxiety that may influence their ability to recall events (Garry et al., 2021). Others have suggested that recall bias could act in the opposite direction, with SARS-CoV-2 positive participants more likely to recall possible contact events (Delgado-Rodríguez and Llorca, 2004). The significant delays between test result and survey (Fig. A2, Supplementary Information) may also influence recall ability (Hipp et al., 2020). Our findings relate to a highly educated population, a characteristic that has been associated with adopting protective behaviours (Vally, 2020). Despite likely ready access to PPE and other resources enabling protective behaviour, protective behaviours were not uniformly reported amongst participants. For all of these reasons – but in particular our small sample and potential for recall bias – we prefer our work to be considered a proof of concept study; demonstrating the types of questions about contact measure, settings and infection risk that can be addressed with linked testing and detailed contact data. Future work in this area may require the development of real-time data streams efficiently capturing details of contact and protective behaviours, that can be embedded within community and/or strategically targeted surveillance of respiratory pathogens.
CRediT authorship contribution statement
Emma L. Fairbanks: Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualisation, Writing – original draft, Writing – review & editing. Kirsty J. Bolton: Conceptualization, Funding acquisition, Investigation, Methodology, Software, Visualisation, Supervision, Writing – original draft, Writing – review & editing. Ru Jia: Conceptualization, Investigation, Methodology, Project administration, Writing – review & editing. Grazziela P. Figueredo: Conceptualization, Funding acquisition, Data analysis, Writing – review & editing. Holly Knight: Conceptualization, Methodology, Project administration, Writing – review & editing. Kavita Vedhara: Conceptualization, Funding acquisition, Investigation, Methodology, Resources, Software, Supervision, Writing – review & editing.
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.
Appendix A Supplementary data
The following is the Supplementary material related to this article. MMC S1
The following is the Supplementary Material related to our article.
Acknowledgements
The authors thank the Nottingham Asymptomatic Testing Service and its users.
Financial support
This work was funded by the 10.13039/501100000837 University of Nottingham . EF acknowledges support via the Nottingham BBSRC Doctoral Training Partnership (grant number BB/M008770/1). KB acknowledges support from a University of Nottingham Anne McLaren Fellowship .
Appendix A Supplementary material related to this article can be found online at https://doi.org/10.1016/j.epidem.2023.100688.
==== Refs
References
Benjamini Y. Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing J. R. Stat. Soc. Series B. Stat. Methodol. 57 1 1995 289 300
Bolton P. Higher education student numbers 2021 (Accessed: 2021-07-29) URL: https://researchbriefings.files.parliament.uk/documents/CBP-7857/CBP-7857.pdf
Bolton K.J. McCaw J.M. Forbes K. Nathan P. Robins G. Pattison P. Nolan T. McVernon J. Influence of contact definitions in assessment of the relative importance of social settings in disease transmission risk PLoS One 7 2 2012 e30893
Brooks-Pollock E. Christensen H. Trickey A. Hemani G. Nixon E. Thomas A. Turner K. Finn A. Hickman M. Relton C. Danon L. High COVID-19 transmission potential associated with re-opening universities can be mitigated with layered interventions Nature Commun. 12 2021 5017 34404780
Children’s Task and Finish Group Paper on higher education settings 2021 (Accessed: 2021-07-29) URL: https://assets.publishing.service.gov.uk/government/uploads/system/upload/attachment_data/file/963387/S1103_Children_s_Task_and_finish_Group_Paper_on_Higher_Education_Settings_.pdf
Delgado-Rodríguez M. Llorca J. Bias J. Epidemiol. Community Health 58 8 2004 635 641 15252064
Department for Business, Energy & Industrial Strategy and Department for Digital, Culture, Media & Sport Working safely during coronavirus (COVID-19) 2020 (Accessed: 2020-11-30) URL: https://www.gov.uk/guidance/working-safely-during-coronavirus-covid-19
Doung-Ngern P. Suphanchaimat R. Panjangampatthana A. Janekrongtham C. Ruampoom D. Daochaeng N. Eungkanit N. Pisitpayat N. Srisong N. Yasopa O. Plernprom P. Promduangsi P. Kumphon P. Suangtho P. Watakulsin P. Chaiya S. Kripattanapong S. Chantian T. Bloss E. Namwat C. Limmathurotsakul D. Case-control study of use of personal protective measures and risk for SARS-CoV 2 infection, Thailand Emerg. Infect. Diseases 26 11 2020 2607 32931726
Enright J. Hill E.M. Stage H.B. Bolton K.J. Nixon E.J. Fairbanks E.L. Tang M.L. Brooks-Pollock E. Dyson L. Budd C.J. Sars-cov-2 infection in uk university students: lessons from september–december 2020 and modelling insights for future student return Royal Soc. Open Sci. 8 8 2021 210310
Faasse K. Newby J. Public perceptions of COVID-19 in Australia: perceived risk, knowledge, health-protective behaviors, and vaccine intentions Front. Psychol. 11 2020 551004
Garry M. Hope L. Zajac R. Verrall A.J. Robertson J.M. Contact tracing: A memory task with consequences for public health Perspect Psychol Sci 16 1 2021 175 187 33301692
Gelman A. Jakulin A. Pittau M.G. Su Y.-S. A weakly informative default prior distribution for logistic and other regression models Ann. Appl. Stat. 2 4 2008 1360 1383
Golding N. Price D.J. Ryan G. McVernon J. McCaw J.M. Shearer F.M. A modelling approach to estimate the transmissibility of sars-cov-2 during periods of high, low, and zero case incidence eLife 12 2023 e78089 36661303
Gronau Q.F. Singmann H. Forster J.J. Wagenmakers E.-J. The JASP Team J.M. Guo J. Gabry J. Goodrich B. Mulder K. de Valpine P. bridgesampling: Bridge sampling for marginal likelihoods and Bayes factors 2020 URL: https://github.com/quentingronau/bridgesampling
Higher Education Statistics Agency Open data and official statistics 2021 (Accessed: 2021-07-29) URL: https://www.hesa.ac.uk/data-and-analysis
Hill E.M. Atkins B.D. Keeling M.J. Tildesley M.J. Dyson L. Modelling SARS-CoV-2 transmission in a UK university setting Epidemics 36 2021 100476
Hipp L. Bünning M. Munnes S. Sauermann A. Problems and pitfalls of retrospective survey questions in COVID-19 studies Surv. Res. Methods 14 2 2020 109 1145
Hobbs C.V. Martin L.M. Kim S.S. Kirmse B.M. Haynie L. McGraw S. Byers P. Taylor K.G. Patel M.M. Flannery B. CDC COVID-19 Response Team P. Factors associated with positive SARS-CoV-2 test results in outpatient health facilities and emergency departments among children and adolescents aged < 18 years—Mississippi, September–November 2020 MMWR Morb Mortal Wkly Rep. 69 50 2020 1925 1929 33332298
Husson F. Josse J. Le S. Mazet J. FactoMineR: Multivariate exploratory data analysis and data mining 2008 URL: http://http://factominer.free.fr/
Kerr C.C. Stuart R.M. Mistry D. Abeysuriya R.G. Rosenfeld K. Hart G.R. Núñez R.C. Cohen J.A. Selvaraj P. Hagedorn B. George L. Jastrzebski M. S I.A. Fowler G. Palmer A. Delport D. Scott N. Kelly S.L. Bennette C.S. Wagner B.G. Chang S.T. Oron A.P. Wenger E.A. Panovska-Griffiths J. Famulare M. Klein D.J. Covasim: an agent-based model of COVID-19 dynamics and interventions PLOS Comput. Biol. 17 7 2021 e1009149
Klompas M. Baker M.A. Rhee C. Airborne transmission of SARS-CoV-2: Theoretical considerations and available evidence JAMA 324 2020 441 442 32749495
Makowski D. Lüdecke D. Ben-Shachar M.S. Patil I. Wilson M.D. Bürkner P.-C. Mahr T. Singmann H. Gronau Q.F. Crawley S. bayestestR: Understand and describe Bayesian models and posterior distributions 2019 URL: https://easystats.github.io/bayestestR/
Mossong J. Hens N. Jit M. Beutels P. Auranen K. Mikolajczyk R. Massari M. Salmaso S. Tomba G.S. Wallinga J. Heijne J. Sadkowska-Todys M. Rosinska M. Edmunds W.J. Social contacts and mixing patterns relevant to the spread of infectious diseases PLoS Med. 5 3 2008 e74
Nixon E. Trickey A. Christensen H. Finn A. Thomas A. Relton C. Montgomery C. Hemani G. Metz J. Walker J.G. Turner K. Kwiatkowska R. Sauchelli S. Danon L. Brooks-Pollock E. Contacts and behaviours of university students during the COVID-19 pandemic at the start of the 2020/2021 academic year Sci. Rep. 11 1 2021 11728 34083593
ONS Coronavirus (COVID-19) Infection Survey technical article: analysis of populations in the UK by risk of testing positive for COVID-19, September 2021 2021 URL: https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/articles/coronaviruscovid19infectionsurveytechnicalarticle/analysisofpopulationsintheukbyriskoftestingpositiveforcovid19september2021
Oran D.P. Topol EJ the proportion of SARS-CoV-2 infections that are asymptomatic: A systematic review Ann. Intern. Med 174 5 2021 655 662 33481642
P.H.E. Transmission Group PHE: Factors contributing to risk of SARS-CoV-2 transmission in various settings, 26 november 2020 2020 URL: https://www.gov.uk/government/publications/phe-factors-contributing-to-risk-of-sars-cov2-transmission-in-various-settings-26-november-2020
Royal Society SET-C SARS-CoV-2: Where do people acquire infection and ‘who infects whom’? 2020 URL: https://royalsociety.org/-/media/policy/projects/set-c/set-c-transmission-paper.pdf
RStudio Team Rstudio: Integrated Development Environment for R 2020 RStudio, PBC. Boston, MA URL: http://www.rstudio.com/
Sethuraman N. Jeremiah S.S. Ryo A. Interpreting diagnostic tests for SARS-CoV-2 JAMA 323 22 2020 2249 2251 32374370
Stan N. Stan user’s guide version 2.27 2019 URL: https://mc-stan.org/docs/2_27/stan-users-guide/index.html
Teo A.K.J. Choudhury Y. Tan I.B. Cher C.Y. Chew S.H. Wan Z.Y. Cheng L.T.E. Oon L.L.E. Tan M.H. Chan K.S. Hsu L.Y. Saliva is more sensitive than nasopharyngeal or nasal swabs for diagnosis of asymptomatic and mild COVID-19 infection Sci. Rep. 11 1 2021 3134 33542443
Thomas D.R. Fina L.H. Adamson J.P. Sawyer C. Jones A. Nnoaham K. Barrasa A. Shankar A.G. Williams C.J. Social, demographic and behavioural determinants of SARS-CoV-2 infection: A case-control study carried out during mass community testing of asymptomatic individuals in South Wales, December 2020 2021 MedRxiv URL: https://www.medrxiv.org/content/early/2021/04/09/2021.04.06.21253465
Thompson H.A. Mousa A. Dighe A. Fu H. Arnedo-Pena A. Barrett P. Bellido-Blasco J. Bi Q. Caputi A. Chaw L. Maria L.D. Hoffmann M. Mahapure K. Ng K. Raghuram J. Singh G. Soman B. Soriano V. Valent F. Vimercati L. Wee L.E. Wong J. Ghani A.C. Ferguson N.M. Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) setting-specific transmission rates: a systematic review and meta-analysis Clin. Infect. Dis. 73 2021 e754 e764 33560412
UCU Monitoring cases of Covid-19 in UK higher and further education 2020 (Accessed: 2020-11-15) URL: https://www.ucu.org.uk/covid-dashboard
UK Government The health protection (coronavirus, restrictions) (England) regulations 2020 2020 (Accessed: 2021-05-22) URL: https://www.legislation.gov.uk/uksi/2020/350/regulation/6/2020-03-26
UK Government The health protection (coronavirus, restrictions) (England) regulations 2021 2021 (Accessed: 2021-05-22) URL: https://www.legislation.gov.uk/uksi/2021/364/contents/made
University of Nottingham Latest updates and news 2021 URL: https://www.nottingham.ac.uk/coronavirus/communications/archive-of-communications.aspx
Vally Z. Public perceptions, anxiety and the perceived efficacy of health-protective behaviours to mitigate the spread of the SARS-Cov-2/COVID-19 pandemic Public Health 187 2020 67 73 32927291
Wagenmakers E.-J. Gronau Q.F. Dablander F. Etz A. The support interval Erkenntnis 2020
Wise T. Zbozinek T.D. Michelini G. Hagan C.C. Mobbs D. Changes in risk perception and self-reported protective behaviour during the first week of the COVID-19 pandemic in the United States R. Soc. Open Sci. 7 9 2020 200742
Zickfeld J.H. Schubert T.W. Herting A.K. Grahe J. Faasse K. Correlates of health-protective behavior during the initial days of the COVID-19 outbreak in Norway Front. Psychol. 11 2020 564083
|
PMC010xxxxxx/PMC10199497.txt |
==== Front
Omega
Omega
Omega
0305-0483
0305-0483
Elsevier Ltd.
S0305-0483(23)00062-2
10.1016/j.omega.2023.102898
102898
Article
Equitable and effective vaccine access considering vaccine hesitancy and capacity constraints
Sengul Orgut Irem a⁎
Freeman Nickolas a
Lewis Dwight b
Parton Jason a
a Department of Information Systems, Statistics, and Management Science, The University of Alabama, 361 Stadium Dr, Tuscaloosa, AL 35487, United States
b Department of Management, The University of Alabama, 361 Stadium Dr, Tuscaloosa, AL 35487, United States
⁎ Corresponding author.
20 5 2023
10 2023
20 5 2023
120 102898102898
12 9 2022
15 5 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The COVID-19 pandemic continues to have an unprecedented impact on people’s lives and the economy worldwide. Vaccines are the strongest evidence-based defense against the spread of the disease. The release of COVID-19 vaccines to the general public created policy challenges associated with how to best allocate vaccines among different sub-regions. In the United States, after vaccines became widely available for all eligible adults, policymakers faced objectives such as (i) achieving an equitable allocation to reduce populations’ travel times to get vaccinated and (ii) effectively allocating vaccine doses to minimize waste and unmet need. This problem was further exacerbated by the underlying factors of population vaccine hesitancy and sub-regions’ varying capacity levels to administer vaccines to eligible and willing populations. Although simple to implement, commonly used pro rata policies do not capture the complexities of this problem. We propose two alternatives to simple pro rata policies. The first alternative is based on a Mixed-Integer Linear Programming Model that minimizes the maximum travel duration of patients and aims to achieve an equitable and effective allocation of vaccines to sub-regions while considering capacity and vaccine hesitancy. A second alternative is a heuristic approach that may be more palatable for policymakers who (i) are not familiar with mathematical modeling, (ii) are reluctant to use black-box models, and (iii) prefer algorithms that are easy to understand and implement. We demonstrate the results of our model through a case study based on real data from the state of Alabama and show that substantial improvements in travel time-based equity are achievable through capacity improvements in a small subset of counties. We perform additional computational experiments that compare the proposed methods in terms of several metrics and demonstrate the promising performance of our model and proposed heuristic. We find that while our mathematical model can achieve equitable and effective vaccine allocation, the proposed heuristic performs better if the goal is to minimize average travel duration. Finally, we explore two model extensions that aim to (i) lower vaccine hesitancy by allocating vaccines, and (ii) prioritize vaccine access for certain high-risk sub-populations.
Keywords
COVID-19
Vaccine allocation
Travel time equity
Effectiveness
Vaccine hesitancy
Capacity
==== Body
pmc1 Introduction
Coronavirus disease 2019 (COVID-19) is a respiratory disease that caused an unprecedented global pandemic by modern day standards [1], with well-documented societal effects in both scholarly literature and news media outlets. The disease spreads between people and was first identified in Wuhan, the capital of the Hubei province in China, in December 2019 [2]. On March 12, 2020, the World Health Organization (WHO) declared the COVID-19 outbreak a pandemic due to the rapid increase in the observed cases and deaths from COVID-19 globally. Although prior global efforts have successfully suppressed novel respiratory viruses (e.g., Middle East respiratory syndrome and Severe acute respiratory syndrome 2002), COVID-19 has been difficult to contain due to its high transmissibility and ability to mutate quickly [3], [4], [5]. As of September 2022, the number of confirmed COVID-19 cases exceeded 608 million worldwide, resulting in over 6.5 million deaths, out of which over one million have been in the United States [6].
In addition to health impacts, the COVID-19 pandemic has also negatively impacted the U.S. domestic and global economy. Early in the pandemic, the U.S. economy saw one of the worst declines in history, causing the unemployment rate to reach 4.4% in March 2020, which is the largest over-the-month rate increase since January 1975 [7]. The International Money Fund (IMF) estimates that the final cost of the pandemic on the global economy will exceed $12.5 trillion by 2024 [8].
Although the true causal relationship between public health policy interventions (i.e., shutdowns, social distancing, and mask mandates) and COVID-19 transmission rates is unclear, findings suggest that such efforts slowed the spread and reduced the incidence of COVID-19-related morbidity and mortality. However, vaccination is recognized as the most effective way to curb the spread of COVID-19 [9]; thus, vaccine uptake is key for the long-term prevention and control of the pandemic [10].
In the U.S., following initial development and clinical trials, vaccines must receive approval or authorization from the Food and Drug Administration (FDA) before being administered to eligible populations [11]. On December 2020, the FDA issued an Emergency Use Authorization (EUA) for the Pfizer-BioNTech™ and Moderna™ COVID-19 Vaccines. Later, on February 2021, the FDA issued an EUA for the Janssen™ COVID-19 Vaccine [12]. Based on guidelines from the Centers for Disease Control and Prevention (CDC), government officials dispersed vaccines to the public using a phased approach [13]. The three phases were: (i) Limited doses available (vaccines available to healthcare personnel, essential workers, people with high-risk medical conditions or aged 65 and above), (ii) Large number of doses available (sufficient supply to vaccinate all the eligible population), and (iii) Continued vaccination. This paper considers the second and third phases of vaccination when sufficient doses exist to vaccinate the eligible population and focuses on the challenges associated with vaccine distribution decisions during those phases.
Despite the ramp-up of vaccine production and distribution, state and local governments faced several challenges when choosing strategies to equitably and effectively allocate vaccines to sub-regions, e.g. counties, in their respective administrative areas, e.g. a state. Although there are varying definitions of equity in literature, it is a general consensus that public policies must promote fairness to be acceptable [14]. The CDC defines vaccine equity as situations “when all people who are eligible have fair and just access to COVID-19 vaccination” [11]. Therefore, an equitable vaccine allocation refers to the scenario where all individuals in a community have equal access to the vaccine, and no individual is at a disadvantage, which may be due to many factors, such as their location and socio-economic status. In addition to altruistic reasons, achieving an equitable allocation of vaccines may also help reduce the severity of the outbreaks [15]. On the other hand, effectiveness refers to “the degree to which a resource allocation causes needs to be met, and is also intended to reflect the extent to which unintended negative impacts of an allocation are avoided” [16]. Following these definitions, our objective is to allocate available vaccines among sub-regions to achieve travel time-based equity (minimize the maximum travel duration of individuals across sub-regions) while ensuring that the entire willing and eligible population has access to a vaccine dose, limiting the number of unused vaccines, and considering vaccine hesitancy and capacities. We use a travel-time based metric since vaccination uptake has been shown to depend on availability and convenience, which are functions of travel time to receive the vaccine, especially once sufficient doses are available to vaccinate the entire eligible population (CDC’s second and third phases of vaccination). Thus, in our primary models, we do not consider allocating scarce vaccines among sub-populations based on health status and risk groups, as those problems are associated with the first phase of vaccine distribution.
Multiple factors affect the equitable and effective allocation of the COVID-19 vaccine. Vaccination hesitancy is a challenge to the nation's ability to achieve herd immunity [17], as some populations appear more reluctant to get vaccinated [18]. Vaccine hesitancy is defined as the “delay in acceptance or refusal of vaccines despite availability of vaccine services.” Based on the WHO’s “3Cs” model, vaccine hesitancy may be impacted by a person’s Confidence (lack of trust in the vaccine or the systems associated with vaccine administration), Complacency (perception of the disease risk to be low), and Convenience for receiving the vaccine [19]. Among these factors, convenience refers to the physical availability, affordability and willingness to pay, geographical accessibility, ability to understand (language and health literacy), and appeal of immunization services. Convenience is recognized as a significant factor affecting vaccine hesitancy [19], [20]. Since convenience is directly related to the closeness of a person’s home or workplace to vaccination sites, it is also a function of their transportation capabilities, which we focus on in the current study.
The second factor affecting vaccine distribution is vaccination capacity, which we broadly define as a sub-region’s total capability to administer available vaccines to eligible individuals. A sub-region’s vaccination capacity depends not only on the number and size of vaccination sites but also on medical staffing volume and availability during times convenient for prospective patients. The BioNTech/Pfizer™ and Moderna™ mRNA vaccines require storage below −60∘C and −15∘C, respectively, which creates additional challenges related to capacity requirements [21]. The capacity for vaccine administration varies among different locations and may not always be sufficient to support the demand in a given region, whereas other regions may have vaccine capacity in excess of their demand. In fact, when the vaccines first became widely available in the U.S., the CDC developed a website [22] to help the public find locations in their region where vaccines were available, sometimes leading to people traveling to farther locations to receive their vaccines. This problem is exacerbated by the fact that many households in the U.S. do not have access to a personal vehicle and may live in areas without public transport. According to the U.S. Census Bureau, in 2019, 8.6% of U.S. households did not have access to a personal vehicle [23], which makes traveling to a distant location challenging, creating a barrier for vaccine administration.
Varying capacity levels and hesitancy rates among the sub-regions make achieving equitable and effective vaccine distribution challenging. As an example, Fig. 1 shows vaccination facility locations (left), travel time to the closest vaccination facility (middle), and facilities per 10,000 people (right) in Alabama during February 2021 when vaccines initially became available for the entire adult population. The color-gradient maps show that vaccine access varies significantly among the different census tracts. The middle map shows that the travel time to the closest vaccination facility varies from less than one minute to over 40 minutes. Although our paper is motivated by the allocation of the COVID-19 vaccines, similar challenges exist for other vaccines, especially those that are regularly administered, such as the influenza vaccine.Fig. 1 COVID-19 Vaccine locations (left), travel time to closest vaccination facility (middle), and facilities per 10,000 people (right) in Alabama in February 2021.
Fig. 1
Problem description and research questions In this paper, we address the problem of equitable and effective allocation of vaccines to population areas while considering varying vaccine hesitancy rates and vaccination capacities. We consider the decision maker to be healthcare policymakers at the region level, e.g., state, and consider the vaccine allocation decisions to sub-regions, e.g., counties. Vaccine hesitancy and capacity restrictions complicate this problem. In the U.S., pro rata policies were used for allocating vaccines to the states, and usually from states to counties. Although pro rata policies have proven to be effective when sufficient doses are not available and during the early days of a pandemic/epidemic [24], these policies resulted in inequities in access when all adults were eligible to receive a vaccine [25]. These inequities may be attributed to the fact that pro rata policies do not consider complex factors such as capacities and hesitancy levels which may result in some areas receiving doses more than they can administer. For example, consider a vaccine allocation strategy that simply allocates the available vaccines in proportion to the population in each county, a policy that is similar to the pro rata policy that many state governments used even after sufficient doses were available to vaccinate all adults nationwide [26], [27]. This would be an equitable distribution as each person would have access to a vaccine dose; however, this is not an effective solution due to two reasons: (i) If a county’s vaccine hesitancy rate is high, vaccines that are allocated to that county would remain unused and would occupy unnecessary refrigeration space, and (ii) If a county’s vaccination capacity is low, they may have to waste the extra amount of doses that they cannot store or administer. In contrast, consider a vaccination strategy such that all the vaccines are allocated to areas with high vaccination capacity and low hesitancy rates. This would be an effective solution as the risk of vaccine spoilage would be low, yet, it is an inequitable solution as vaccines may be allocated to more urban areas, and the people residing in rural areas would have to travel long distances to receive their vaccines.
This paper considers four research questions:1. How should public health officials allocate available vaccines to their service areas to ensure equitable and effective allocation while considering vaccination capacity and vaccine hesitancy? To address this question, we develop a Mixed-Integer Linear Programming Model that we refer to as the Equitable and Effective Vaccine Allocation Model (EEVAM). EEVAM determines the optimal allocation of vaccines to sub-regions of a larger region while considering vaccination capacity levels at local facilities and hesitancy among prospective patients. A case study based on the state of Alabama demonstrates the utility of the model for informing policy decisions, and additional experiments demonstrate how allocations change as problem parameters vary.
2. How do solutions from simple heuristic approaches and our model compare in terms of the travel time and effectiveness levels they achieve? To address this question, we first develop an interpretable and easy-to-implement heuristic for allocating vaccines, which we refer to as the Vaccine Allocation Heuristic (VAH). We also consider a Pro Rata Heuristic (PRH) for vaccine allocation, which mimics the approach used by many healthcare administrators in the U.S. Our computational experiments compare the results from these heuristics and EEVAM in terms of the maximum per-person travel duration, average per-person travel duration, the percentage of unused vaccines, and the percentage of unmet need. We also explore how each method’s performance varies under different parameter settings.
3. How can public health officials improve vaccine access in their service areas? Our computational experiments and case study show how model parameters such as vaccine capacity affect the levels of equity and effectiveness achieved in service areas and generate managerial insights regarding ways to improve vaccine access.
4. What are the practical implications of (i) reducing vaccine hesitancy through strategically allocating vaccines to sub-regions, and (ii) prioritizing minimizing the travel efforts of more vulnerable high-risk sub-populations? To address these questions, we propose two model extensions and examine the impact of these approaches on vaccine allocation decisions and performance metrics.
The rest of this paper is structured as follows. We deliver a literature review in Section 2. We present model formulations and related solution approaches in Sections 3 and 4, respectively. Our case study is also presented as a subsection in Section 3 to demonstrate the utility of our developed model. In Section 5, we present the results of a series of computational experiments. Section 6 proposes two extensions to our model and examines the impact of those approaches on several performance metrics. We conclude the paper with some key findings and future directions in Section 7.
2 Literature review
Our research is most closely related to two areas of literature: (i) Vaccine Supply Chain and Vaccine Allocation, and (ii) Equity in Transportation.
2.1 Vaccine supply chain and allocation
Vaccine supply chains differ from traditional supply chains due to many aspects, including the multiple and sometimes conflicting objectives of the decision makers, uncertainties associated with supply and demand (which is a function of hesitancy), and the decentralized structure of the supply chain (no single decision maker). We refer the reader to Duijzer et al. [28], Lemmens et al. [29], and Blasioli et al. [30] for extensive reviews of the vaccine supply chain and logistics literature. Duijzer et al. [28] classify this literature into four categories: (1) product (the kind of vaccine), (2) production (quantities and schedules), (3) allocation (of vaccines among regions or populations), and (4) distribution (of vaccines to end-users). Among these categories, our study falls under “allocation” which may be among geographical regions or age groups. They state that although allocation among multiple regions has been studied in the medical/epidemiological literature, it is a relatively unexplored area in OR/OM literature. Further, they highlight that vaccine hesitancy has hardly been incorporated into any papers in the OR/OM field and state that papers considering this aspect as a part of allocation decisions are needed.
Some studies in the OR/OM field consider the vaccine allocation problem among geographically different regions with a cost minimization objective. The costs to be minimized include the cost of lost productivity and medical services [31], the operational cost associated with opening and operating vaccination sites, and the travel distance of vaccination recipients [32], the total distribution cost of the vaccination campaign and the total expected number of deaths among the population [33], the total economic, environmental and social costs associated with the COVID-19 vaccine supply chain at an international and domestic level considering the uncertainty in vaccine availability [34], the cost associated with storage and transportation of vaccines, fleet and staff costs and indirect costs of wasted doses [35], and the global cost in a system of multiple interacting countries [36]. These papers do not consider travel time equity in vaccine allocation as a part of their modeling structure or incorporate vaccine hesitancy.
Other studies focus on effectiveness metrics related to minimizing the number of infections or minimizing vaccine doses required to contain outbreaks. Yarmand et al. [37] propose a two-phased vaccination approach motivated by the seasonal influenza vaccine and formulate a 2-stage stochastic model for allocating vaccines among the different regions (e.g., counties) under uncertain epidemic containment. They include transportation costs but do not explicitly incorporate the distances individuals need to travel to get vaccinated. Araz et al. [24] compare four vaccine allocation strategies: (1) pro rata, (2) sequential by population size, (3) sequential by the estimated order of pandemic peaks, and (4) reverse sequential by the estimated order of pandemic peaks. They find that prioritizing counties expected to experience the latest epidemic waves provides the best outcomes in reducing the overall attach rate; however, the pro rata policy is effective if the waiting period for receiving vaccines is also considered. Teytelman and Larson [38] consider the sequential decision making problem of allocating vaccine doses to geographical areas as doses become available and compare different heuristic approaches to achieve the minimum expected number of infections. They show that using an adaptive strategy that considers the current status of flu in each region outperforms the pro rata policy. They incorporate equity considerations by placing limits on how many vaccines each state can get; however, do not consider travel times. Rey et al. [39] consider a set of global agents (e.g., countries) and use a reinforcement learning approach to learn stochastic vaccine efficiency rates over time and allocate vaccines to populations to minimize the expected number of susceptible individuals. Enayati and Özaltın [40] study the optimal allocation of influenza vaccines in a heterogeneous population consisting of multiple subgroups based on geographic regions and age groups. Their mathematical model incorporates disease transmission dynamics and minimizes the total vaccine doses required to effectively extinguish an emerging outbreak in its early stages. They incorporate the equitable allocation of vaccines across regions through a constraint but do not consider transportation access or capacity. The critical node problem, which aims to find the nodes whose elimination results in the minimum pair-wise connectivity among the remaining nodes, is commonly used to decide on which areas to vaccinate and has been studied by several researchers [41], [42], [43], [44], [45]. This approach aims to minimize the transmission of disease and hence overall infections and does not consider equity in vaccine access.
Several papers have incorporated fairness metrics in vaccine allocation decisions. Rastegar et al. [46] develop an inventory-location mixed-integer linear programming model for equitable influenza vaccine distribution in developing countries subject to a budget constraint. Transportation cost to be minimized is a function of the distance between the distribution center and a demand point and is not a part of the fairness objective, which maximizes the minimum fill rate (number of vaccines allocated over demand) for each population group in each province and time period. Tavana et al. [47] extend the model by Rastegar et al. [46] by incorporating COVID-19 vaccine-specific factors such as a multi-product framework, varying refrigeration storage levels, and waiting times of the vaccines between order and delivery. Wang et al. [48] incorporate equity through a constraint that limits the difference between the maximum and minimum fill rate (the ratio of the total number of scheduled vaccinations to the population at each demand area). Their two-stage robust optimization model considers supply uncertainty and aims to maximize social and economic benefits. Similar to Rastegar et al. [46], they also consider an access cost that is based on the distance between a health facility and a demand site but is not a part of the equity constraint. Bertsimas et al. [49] consider the strategic problem of determining the optimal locations of mass COVID-19 vaccination sites in the U.S. and the tactical problem of the subsequent allocation of vaccines to each site. They incorporate equity by ensuring that: (i) the fraction of vaccination sites opened at each state does not deviate too much from its population share, (ii) the vaccine distribution across sites resembles a uniform distribution, and (iii) the fraction of vaccines allocated to each state does not exceed their population share too much. They incorporate disease dynamics into a mathematical model that simultaneously minimizes the pandemic’s death toll, the number of exposed people, and the distance between vaccination sites and population centers (counties). Like our paper, they also use a pro rata policy as a benchmark. Emanuel et al. [50] consider the equitable distribution of vaccinations on a global scale and argue that equitable vaccine distribution should be built on three fundamental values: 1) benefiting people and limiting harm, 2) prioritizing the disadvantaged, and 3) equal moral concern. Jarumaneeroj et al. [51] develop a model that combines the COVID-19 transmission dynamics with the vaccine allocation problem to minimize the total number of infected individuals. They assess four vaccination strategies: (i) each region receiving the same amount of vaccines regardless of their demographics and COVID-19 situations, (ii) the epicenter-based strategy that prioritizes the allocation of vaccines to provinces with a greater number of confirmed cases, (iii) the capacity-based strategy that allocates vaccines to the provinces based on their respective vaccine administration capacity, and (iv) the epicenter-based strategy that allocates twice the number of procured vaccines to provinces with non-zero cases based on the initial number of confirmed cases. Our proposed heuristic in this paper differs from their capacity-based strategy in that we incorporate the need and hesitancy levels in addition to capacity. Although these papers consider an equity metric, none of them consider transportation equity (in terms of travel durations or distances of beneficiaries) or vaccine hesitancy.
Some studies have considered the problem of prioritization in vaccine allocation based on factors such as population risk levels and infection rate. These considerations are important in the initial phases when a vaccine is introduced and supply is scarce. Balcik et al. [52] develop a model that accounts for priority groups, vaccine types, and the capacity of regions to achieve equitable coverage levels for different locations and population subgroups. Anahideh et al. [53] develop a mathematical model that aims to achieve a vaccine allocation strategy that is fair both geographically and also based on people’s demographic or socioeconomic backgrounds. Fadaki et al. [54] propose a multi-period mathematical model that incorporates the individual’s level of exposure risk and susceptibility ratings. Sahinyazan and Araz [55] explore the association between several metrics associated with food insecurity and COVID-19 related health outcomes in the US through multiple regression analysis. They also develop a COVID-19 vulnerability score, which officials can use to determine equitable prioritization schemes for vaccine allocation. They show that a pro rata vaccine allocation approach may aggravate health disparities in some areas. Our study does not consider vaccine scarcity but focuses on Phases 2 and 3 of CDC’s vaccination approach [13] when sufficient vaccine doses exist to vaccinate the entire eligible and willing population. Further, these studies do not consider the impacts of transportation access or vaccination hesitancy.
The vaccine allocation and distribution problems can be considered similar to the disaster relief operations, which have been studied extensively in literature [56], [57], [58]. Researchers in this area focused on problems such as the allocation of relief supplies to the disaster sites [59], [60], [61], supplier selection [62], last-mile distribution of supplies and routing [63], [64], [65], [66], and facility location and network design [67], [68]. Although some of these studies also consider equity and effectiveness as objectives as we do in this paper [59], [61], [63], [65], [66], [69], some inherent differences separate our problem from disaster relief chains. First, vaccine supply chains are more widespread in the sense that they have to consider all eligible individuals as beneficiaries, whereas disaster relief chains focus on the areas impacted by the disasters. Second, our definition of equity in this paper is related to transportation equity, i.e. allocations that minimize the maximum travel duration that a person needs to travel to get vaccinated. However, disaster relief operations usually aim to distribute relief supplies directly to the affected populations. Finally, aspects such as hesitancy are usually not relevant to disaster relief operations.
Our paper builds on these existing studies and differentiates from them through (i) considering transportation (travel time-based) equity and formulating it as an objective, (ii) incorporating underlying factors in vaccine distribution such as the vaccine hesitancy of populations and capacity restrictions of different areas, and (iii) developing an easy-to-implement heuristic that provides an alternative to the naive pro rata policy that is used by many policymakers when the goal is to minimize average travel duration.
2.2 Equity in transportation
Fairness and equity are abstract socio-political and subjective concepts, which makes it difficult to determine an equity measure that applies to all problems [70], highlighting the need to select the equity measure based on the structure of the problem at hand [16], [71], [72], [73]. In this paper, we focus on transportation inequities and their impact on vaccine allocation decisions. In the OR/OM literature, equity regarding transportation has mainly been considered in terms of facility location problems, although a few studies have explored resource allocation.
First, we examine facility location problems that consider equitable transportation access. In their seminal work, Marsh and Schilling [72] give an overview of common equity measures with a focus on facility siting decisions. They define equity such that each affected group receives its fair share of the total effect. Studies in this context consider application areas such as the location of Emergency Medical Service (EMS) centers [16], [74], [75], [76], [77] and other public facilities [78], [79] to ensure equitable coverage of the service area, and the design of disaster relief networks for equitable service [69], [80], [81], [82]. Our paper differs from these studies as we do not consider the strategic problem of facility location and focus on the tactical problem of allocating resources to existing capacitated facilities to achieve equitable travel times by beneficiaries.
Few studies in the OR/OM literature focus on the resource allocation problem for achieving transportation equity. Meng and Yang [83] use a bilevel programming approach to allocate capacity increase among existing roads under a budget constraint. They incorporate equity as a constraint such that network users benefit equally from a capacity increase in terms of their average origin-destination travel costs. Noyan et al. [69] consider a last-mile disaster relief network and develop a model that determines distribution center locations and capacities, their assigned demand to serve, and supply allocations. They ensure that a minimum threshold of post-disaster mobility accessibility level is met by each location. They also propose a hybrid approach that combines a pro rata policy with a policy that sets a limit on the maximum shortage amount to address equity in supply distribution. McCoy and Lee [84] are motivated by a nonprofit organization in sub-Saharan Africa that provides reliable transportation to healthcare workers to improve access to remote villages, hence, making healthcare access more equitable. They develop a mathematical model based on a family of fairness functions that consider different modes of transportation and seeks to achieve equitable and effective policies for allocating additional transportation capacity via development programs. De Boeck et al. [85] combine spatial modeling with discrete-event simulation to explore the effect of transportation disruptions (e.g., road closures due to flooding) on equitable vaccine access and explore methods to mitigate these disruptions through capacity increases or changing the vaccine replenishment frequency at the vaccination sites. To our knowledge, ours is the first paper that proposes a travel time-based equity metric in the objective function for a resource allocation problem. Thus, our paper addresses a gap in transportation literature by considering a resource allocation problem for distributing available supplies to existing, capacitated facilities to minimize the maximum travel duration by the population to reach those facilities.
3 Model formulation
We consider the problem faced by a regional decision maker, e.g., a state public health official, when deciding how to allocate vaccine doses among sub-regions under their administration, which is illustrated in Fig. 2 . The decision maker aims to achieve an equitable allocation of vaccines such that the maximum travel time of individuals residing across different sub-regions is minimized. The average travel time considers different modes of transportation, i.e. personal vehicle and public transportation. Although we focus on the scenario where sufficient vaccine doses are available to satisfy the region’s demand, several challenges make achieving an equitable and effective distribution among sub-regions difficult. First, sub-regions may have varying levels of vaccination capacity, which depends on vaccination facilities and personnel available. Therefore, sub-regions may not be able to receive or administer enough vaccines to satisfy their demand due to their limited capacity. In such cases, the model aims to allocate vaccines to neighboring sub-regions such that the maximum per-person travel time is minimized. Individuals who would like to get vaccinated can then use tools such as the Vaccine Finder tool developed by CDC [22] to find vaccines that are closest to their area. Second, since vaccines are valuable resources, wasting a vaccine dose is undesirable. For this reason, we limit the percentage of unused vaccines in each sub-region in our model. Third, our model ensures that all people who want to be vaccinated can do so either in the sub-region they reside in or by traveling to a neighboring sub-region. Finally, our model can accommodate varying levels of vaccine hesitancy across sub-regions when making allocation decisions. We now present the notation for the Equitable and Effective Vaccine Allocation Model (EEVAM).Fig. 2 Vaccine allocation problem.
Fig. 2
Notation
Sets and indices:
J: set of sub-regions, J={i,j:1,2,…|J|}
Parameters:
Pj: Vaccine-eligible population in sub-region j, j∈J
hj: Vaccine hesitancy rate in sub-region j, 0≤hj≤1j∈J
tijPV: Travel duration (minutes) between sub-regions i and j by personal vehicle, i,j∈J
tijPT: Travel duration (minutes) between sub-regions i and j by public transportation, i,j∈J
δijPV: Proportion of people traveling from i to j estimated to use personal vehicle, i,j∈J
δijPT: Proportion of people traveling from i to j estimated to use public transportation,
δijPV+δijPT=1, i,j∈J
Cj: Vaccination capacity (in number of doses) in sub-region j, i,j∈J
ξ: Maximum limit on the percentage of unused vaccine doses in each sub-region, 0≤ξ≤1
A: Number of vaccine doses available for allocation
Decision variables:
Vj: Number of vaccine doses to allocate to sub-region j, j∈J
Xij: Number of people who will travel from sub-region i to j to receive their vaccine, i,j∈J
E: The maximum average travel duration (minutes) for people residing in a sub-region, across the entire region
Below are our modeling assumptions.Assumption 1 We do not differentiate between vaccine types and we model each vaccine as a single dose.
Assumption 1 states that although there may be different vaccine types available with different administration schemes, we do not differentiate between them and consider them a single dose. For COVID-19 vaccines, some vaccines are administered as a single dose whereas others as two doses. This assumption is only limiting in the sense that the model does not capture whether a person must make single or multiple trips to become fully vaccinated. We discuss this limitation in the Conclusions section.Assumption 2 We consider a single-period allocation problem.
Assumption 2 states that EEVAM models the vaccine allocation problem as a single-period model and does not consider the dynamics of the changing population. The model can initially be run for the target population to be vaccinated over a certain time period, e.g. vaccinating one million people over a month. Later, as the data evolves, the model can be run on a rolling-horizon basis using the updated model parameters.Assumption 3 There are sufficient vaccine doses for the eligible and willing population to receive the vaccine, i.e. A≥∑j∈J⌈Pj(1−hj)⌉.
Our paper focuses on the case where there are sufficient doses for everyone who is eligible and willing to receive the vaccine. As stated in Section 1, this refers to CDC’s Phase 2 and 3 of vaccination. Although determining eligible populations, and vaccine distribution under scarcity are also interesting problems, they are outside the scope of this paper. Our case study and computational experiments assume that A=∑j∈J⌈Pj(1−hj)⌉. Considering the case where supply exceeds demand would not provide any additional insights as vaccines that are not needed would either not be distributed, or would be distributed but not administered at sub-regions.Assumption 4 The total capacity in the service region is sufficient to meet the demand, i.e. ∑j∈JCj≥∑j∈J⌈Pj(1−hj)⌉.
Assumption 4 states that the total capacity in the entire region is sufficient to serve the total demand. This does not mean that a sub-region’s capacity must be greater than its demand. This assumption is necessary to obtain a feasible solution to the problem. If there is not sufficient capacity in the entire region, public health officials should either investigate ways to increase state-wide capacity, or limit the current eligible or target population. In Section 5, we examine the impact of varying the sub-region capacities on equity.
We formulate EEVAM as a Mixed-Integer Linear Programming Model as follows:
EEVAM: (1) minE
(2) Subjectto:Averagetravelduration:∑j∈J(tijPVδijPVXij+tijPTδijPTXij)Pi(1−hi)≤E∀i∈J
(3) Peopletobevaccinated:∑j∈JXij≥Pi(1−hi)∀i∈J
(4) Sub−regionsupply:∑i∈JXij≤Vj∀j∈J
(5) Unusedvaccinecap:Vj−∑i∈JXij≤ξVj∀j∈J
(6) Capacityconstraint:Vj≤Cj∀j∈J
(7) Regionsupplyconstraint:∑j∈JVj≤A
(8) Vj,Xij∈Z≥0∀i,j∈J
(9) E∈R≥0.
The objective function (1) together with constraint (2) minimizes the maximum average per-person travel duration across the sub-regions to achieve equitable access to vaccine supplies. This objective seeks an equitable distribution of vaccine supplies such that no sub-region is at a disadvantage in terms of the average duration that its population needs to travel to get vaccinated. Note that (2), together with the objective function can be reformulated as(10) minmaxi∈J∑j∈J(tijPVδiPVXij+tijPTδiPTXij)Pi(1−hi).
In this formulation, tijPVδiPVXij represents the total travel time for the people residing in sub-region i who are estimated to use personal vehicle to travel to sub-region j to receive their vaccinations. Similarly, tijPTδiPTXij represents the total travel time for the people who are estimated to use public transport for traveling between i and j. We sum this value across all regions to be traveled to (j) to receive the total travel duration (by personal vehicle or public transport) for all people residing in sub-region i to receive their vaccine. We divide this value by the total demand in sub-region i to receive the average travel duration per person. We linearize (10) by formulating it as the objective function (1), together with constraints (2).
Constraint (3) requires that the number of people residing in a sub-region i and traveling to any sub-region j to get vaccinated should equal the population of sub-region i that needs to be vaccinated. Note that Xii represents the number of people who live and receive their vaccination in sub-region i and that the population of sub-region i, Pi is multiplied by one minus the hesitancy rate of sub-region i, hi, to obtain demand. Constraint (4) states that the number of people receiving vaccinations in sub-region j cannot exceed the number of vaccine doses allocated to sub-region j. Constraint (5) limits the number of unused, and potentially wasted, vaccines in each sub-region j to be no more than ξ (a percentage) of the vaccine doses allocated to that sub-region. This constraint is especially relevant when the number of available vaccine doses exceeds the total need in the region. In those situations, the decision maker may want to limit the number of unused vaccines at each sub-region since it is easier for the state warehouses to store unused vaccines than smaller locations at sub-regions. Constraint (6) limits the amount of doses allocated to sub-region j to the vaccination capacity in sub-region j. Constraint (7) states that the total number of allocated vaccines cannot exceed the total available doses in the entire region. Note that, by Assumption 3, there are sufficient vaccine doses for the eligible population in the region. Constraint (8) requires the decision variables Vj and Xij to be non-negative integers and constraint (9) requires E to be a non-negative continuous variable.
3.1 Case study: vaccine allocation in Alabama
To demonstrate the utility of the developed model, we first solve EEVAM using real-life data in the state of Alabama and then present an application that uses a slightly modified version of the model to identify counties where allocating additional vaccine capacity is most beneficial from the perspective of equity. The decision to perform this analysis at the county level is based on: i) the availability of data and ii) the tendency for state governments to make decisions at the county level rather than more granular regional delineations, e.g., city or census tract.
Data For our case study, we consider the allocation of vaccine doses by state public health officials to the 67 counties in Alabama. We use the U.S. Census county-level estimates for the population above 18 years of age for the vaccine-eligible population [86] and published estimations of vaccine hesitancy rates by county [87]. Since no direct data exists for the vaccination capacities in each county, we consider the daily number of COVID-19 vaccines administered in each county during the first six months of 2021 when vaccines became available for the general public [88]. We take the 90th percentile of these daily vaccination counts for each county and use this as an estimate of the county’s daily vaccination capacity. We then multiply this daily capacity by 180 to achieve an estimate of vaccination capacity over six months, which is the duration we consider for the entire willing population to receive their vaccines. Alabama is a rural state, and public transportation is mainly used in the Birmingham metropolitan area (Jefferson county, where Birmingham is located, is the county with the highest population). Bureau of Transportation Statistics [89] estimates that only 0.42% of the population used public transportation in Alabama in 2019, corresponding to 508,723 people. By using this number, we calculate the percentage of the population expected to use public transport for reaching vaccination facilities in Jefferson County to be approximately 3%, i.e. δiiPT=0.03 for i = Jefferson county; and δjkPT=0 for all other counties j and k. We estimate the personal vehicle travel duration between the population-weighted centroids of each county using a proprietary GIS network dataset [90], and a customized algorithm from ESRI© using the General Transit Feed Specification dataset for transit-related travels [91]. Both travel modes were modeled with the impedance of travel time (i.e., minutes) with average traffic patterns witnessed on weekdays between the hours of 8:00AM and 5:00PM. Finally, we set ξ=0.1 and A=∑j∈J⌈Pj(1−hj)⌉ assuming that there are sufficient vaccine doses for the entire eligible and willing population.
Base case scenario We solve EEVAM to obtain the vaccine allocations for a base case scenario, which assumes that the number of vaccines allocated to the state is exactly equal to the number of willing residents. Fig. 3 shows the vaccine allocations (left) and average travel times (right) for each county.Fig. 3 Case study - base case vaccine allocations and expected travel times (1: Jefferson county).
Fig. 3
Taken together, the allocation and travel time results highlight areas where capacity limits equitable access. We notice that highly populated areas, such as Jefferson county (labeled “1”), receive a high number of vaccine allocations and still have a high average travel time, due to capacity restrictions. In the next section, we demonstrate how our model may be employed to investigate methods for improving the equity of vaccine allocations.
Allocating additional capacity We consider a setting where decision makers can increase capacity in a subset of counties to improve equity. Specifically, we assume that there are two additional decisions: 1) which counties should receive supplemental capacity, and 2) how much capacity should each of the selected counties receive. Accommodating these decisions does require some slight modifications to EEVAM. In addition to the notation described above, we define two additional parameters and two additional decision variables. We let M denote the number of counties eligible for additional vaccine capacity allocations, and K denote the maximum amount of additional vaccine capacity to allocate. We also define Yj as the additional vaccine capacity to allocate to sub-region j and Zj as a binary decision variable which equals one if additional vaccine capacity is allocated to sub-region j, and zero otherwise, for all j∈J. The formulation for the revised model, which we refer to as EEVAM-Case, follows.
EEVAM-Case: (11) minE
(12) Subjectto:Averagetravelduration:∑j∈J(tijPVδijPVXij+tijPTδijPTXij)Pi(1−hi)≤E∀i∈J
(13) Peopletobevaccinated:∑j∈JXij≥Pi(1−hi)∀i∈J
(14) Sub−regionsupplyconstraint:∑i∈JXij≤Vj∀j∈J
(15) Unusedvaccinecap:Vj−∑i∈JXij≤ξVj∀j∈J
(16) Capacityconstraint:Vj≤Cj+Yj∀j∈J
(17) Statesupplyconstraint:∑j∈JVj≤A
(18) Numberofcapacityincreases:∑j∈JZj≤M
(19) Sub−regioncapacityincrease:Yj≤KZj∀j∈J
(20) Totalcapacityincrease:∑j∈JYj≤K
(21) Vj,Xij,Yj∈Z≥0∀i,j∈J
(22) Zj∈{0,1}∀j∈J
(23) E∈R≥0.
We now describe any constraints added or modified in the formulation. Constraint set (16) is modified to ensure that the vaccines allocated to a county do no exceed the available capacity, which includes any additional capacity allocations specified by the Yj variables. Constraint set (18) ensures that the number of counties that receive additional capacity is less than or equal to the limit specified by the parameter M. Constraint set (19) ensures that additional capacity can only be allocated to the selected counties, i.e., Zj=0→Yj=0. Constraint set (20) ensures that the additional capacity allocated among all selected counties is less than or equal to the upper bound specified by the parameter K. Finally, constraint sets (21) and (22) are updated to include the Yj and Zj variables, respectively.
We use a factorial experiment to demonstrate how our modified model can be used to understand the impact of supplemental capacity on equity. Table 1 describes the factors and levels used in the experiment.Table 1 Experimental design - case study.
Table 1Parameter Value(s)
# Counties receiving additional vaccine capacity 0, 1, 2, 4, 8, 16, 32, 64
Additional vaccine capacity to allocate (as a % of the number of willing residents in the state) 0.01, 0.02, 0.04, 0.08, 0.16, 0.32, 0.64, 1.28, 2.56
Fig. 4 uses a heatmap to show the percent improvement in equity, relative to the base case, as the experimental factors vary. The number of counties receiving additional vaccine capacity increases from bottom to top along the vertical axis and the total amount of additional vaccine capacity to allocate increases from left to right along the horizontal axis. Note that cases where the number of counties eligible for increases equals zero essentially correspond to the base case.Fig. 4 Case study experiment summary.
Fig. 4
From the standpoint of a decision maker, the results shown in Fig. 4 suggest that a small subset of counties limit improvements in equity. To see this, note that the most drastic improvements in equity occur as we move left to right in the heatmap as opposed to what we observe by moving from bottom to top. In fact, we can achieve a 50% improvement in equity by supplementing the vaccine capacity for two counties with an amount equaling 0.16% of the number of willing residents in the state. Increasing the capacity allocations to 2.56% of the number of willing residents and allocating the additional capacity across nearly all of the 67 counties in Alabama only yields a further improvement of 38%. Fig. 5 depicts how the additional capacity is allocated geographically for a subset of the experimental instances.Fig. 5 Case study experiment summary maps (the color-gradient represents the number of additional vaccines allocated as a percentage of that county’s population and reaches its maximum intensity at 20%. This value was chosen to best illustrate the solution differences.).
Fig. 5
This representation of the results shows that initial efforts to improve equity would be focused on small subsets of counties. For example, note that if the additional capacity to allocate is low, additional capacity is entirely allocated to a single county, regardless of the number of eligible counties. As the available capacity increases, additional counties are supplemented. As a final observation, it is important to note that the prescribed allocations are not additive in general. For example, consider a setting where policymakers are considering allocating additional capacity equivalent to 0.64% of the willing population and have chosen to allocate this capacity among four counties (second row, fourth column). Suppose that sometime after executing the described solution, decision makers want to perform a similar increase to bring the total additional capacity allocated to a level equivalent to 1.28% of the willing population. Considering the same four-county allocation, the proposed solution, if this increase had been made initially (second row, fifth column), actually reduces the allocation to one of the affected counties, which may not be possible.
Although the assumption of zero travel times within a region is strong given that we consider a county-level analysis, the previously described case shows that our model offers an effective way for decision makers to strategically consider various allocation strategies and choose a course of action that best meets their current and anticipated future situation. In the next section, we prove some structural properties of the model and present two heuristic solution approaches that may be employed in circumstances where personnel does not have the mathematical programming expertise necessary to formulate and solve the model using commercially available solvers.
4 Structural properties and heuristic approaches
Although EEVAM developed in Section 3 provides equitable and effective vaccine allocation policies, it requires the person implementing it to have knowledge of optimization modeling and access to a capable solver. However, government policymakers generally prefer policies that are easy to understand and implement. In practice, most states in the U.S. currently use a pro rata allocation scheme, which is based on a simple county-level eligible population formula [27]. In this section, we present the structural properties of EEVAM that we later use in a heuristic for allocating vaccines to sub-regions as an alternative to the pro rata policy. We also present a formal definition of the pro rata algorithm to use as a benchmark.Lemma 1 Consider the scenario where there is infinite capacity such that constraint(6)is removed from EEVAM. Since the travel time is minimized if all people in a sub-region can get vaccinated in the same sub-region that they reside in, i.e.min(i,j)∈JtijPV=tiiPVandmin(i,j)∈JtijPT=tiiPT, the optimal solution for this case would be as follows:(24) Vi=⌈Pi(1−hi)⌉∀i∈J,
(25) Xij={Vifori=jand(i,j)∈J0fori≠jand(i,j)∈J.
The proof for Lemma 1 is provided in Appendix A. Lemma 1 shows that if there are no sub-region capacity restrictions, the optimal solution is to allocate vaccines to each sub-region based on the exact number of people who are eligible and willing to get vaccinated. In reality, different sub-regions have different capacities for vaccination, and their capacity may be greater, or less than their demand. Corollary 1 follows directly from Lemma 1 and states that if a sub-region i has sufficient capacity to vaccinate its population, the number of vaccines allocated to that sub-region should be at least ⌈Pi(1−hi)⌉.Corollary 1 If, for a sub-regionj,Cj≥⌈Pi(1−hi)⌉, then we should haveCj≥Vj≥⌈Pi(1−hi)⌉.
The next section presents the Vaccine Allocation Heuristic (VAH), which uses these results.
4.1 Vaccine Allocation Heuristic (VAH)
Appendix B presents pseudocode for the Vaccine Allocation Heuristic (VAH). VAH starts by initializing sets JO and JI to be equal to the set of all sub-regions J and the empty set, respectively. Set JI, which we refer to as the insufficient capacity set, represents sub-regions that have insufficient capacity to satisfy their needs completely. This set is initially empty and will be filled in subsequent steps. Set JO, which we refer to as the ordered set, represents the sub-regions ordered in increasing order of their capacity Ci over need Pi(1−hi) ratios. This ratio represents a sub-region’s ability to satisfy its area’s need. Specifically, a smaller value that is less than one means that a sub-region does not have sufficient capacity to satisfy their need, and people residing in that region will need to travel to other sub-regions to receive vaccines. We also initialize the “remaining” capacity vector CR to be equal to the capacity vector C. As vaccines are allocated to sub-regions, we update their capacity in this vector to represent the remaining available capacity.
We start by checking whether a sub-region i has insufficient capacity to satisfy their need, i.e. Ci<⌈Pi(1−hi)⌉. If this is true, following from Corollary 1, we allocate the region’s entire capacity to satisfy a portion of the need in that sub-region, and update the sub-region’s remaining capacity CiR to be equal to zero. We add this sub-region to the set of counties with insufficient capacity, JI. If this is not true, i.e., there is sufficient capacity to satisfy need, we allocate enough vaccines to satisfy the need of the sub-region and update the remaining capacity to be the initial capacity minus the need of the sub-region. Thus, at this point, any sub-region with sufficient capacity to satisfy their need has been allocated vaccines. We then consider the sub-regions in set JI, namely those without sufficient capacity and allocate vaccines and assignments of people to sub-regions.
Our next step is to decide on the assignments of people in the sub-regions JI to other sub-regions to get vaccinated and allocate vaccines accordingly. We start with the sub-region i that has the smallest capacity over need ratio since this sub-region has the largest percent gap between its capacity and demand. For this sub-region i, we determine the set Jidist that contains all the sub-regions in increasing order of their travel duration from sub-region i. By definition, the first element of set Jidist would be i since the within-sub-region travel duration would be minimum. The second element would be the index of the sub-region that is the second closest (in terms of travel duration) to sub-region i, and so on. Then, while there is still an unassigned need in sub-region i, we go through the set Jidist and allocate vaccines to the sub-regions with at least one unit of remaining capacity. After allocation, we update the sub-region’s remaining capacity and make assignments of people in sub-region i to the regions with remaining capacity. If a sub-region in set Jidist has sufficient capacity to satisfy the current unsatisfied need in sub-region i, we make allocations and move on to the next sub-region in set JI. If not, we make allocations and move on to the next sub-region in set Jidist to make further allocations. Once all allocations are completed, we calculate E, the maximum average travel duration across the sub-regions, and end the heuristic.
By Lemma 1 and Corollary 1, VAH is expected to produce optimal solutions for EEVAM if each sub-region has sufficient capacity to satisfy their need. In this case, a trivial solution is to allocate vaccines to sub-regions such that the allocated amount equals the sub-region’s need. Such a solution allows each person to get vaccinated in the sub-region that they reside in, which minimizes travel time. In reality, as we have seen from the Case Study in Section 3.1, sub-regions have varying capacities and may not always have sufficient capacities to satisfy their need. In these situations, VAH will produce sub-optimal solutions for EEVAM due to its myopic nature of allocating vaccines sequentially. Because of this myopic behavior, VAH is expected to provide solutions with good average per-person travel times, but not necessarily good maximum per-person travel times. Therefore decision makers should take this into consideration when using VAH. Next, we describe a heuristic that is used by many state governments in the U.S. for allocating vaccines to their counties to use as a benchmark.
4.2 Pro Rata Heuristic (PRH) for vaccine allocation
Although state and local governments use many different algorithms for allocating vaccines in their respective regions, a commonly used algorithm allocates vaccines based on the eligible populations in the sub-regions. We call this algorithm the Pro Rata Heuristic (PRH). The federal government also used this algorithm to allocate vaccines in the U.S.: the algorithm “divides the total amount of vaccine available each week among the 50 states - as well as U.S. territories and a few big cities like New York - based on the number of people over 18 in each place” [26]. States also used similar algorithms for distributing vaccines in their regions [27]. In this section, we formally define PRH. Later, we will compare the results from EEVAM and VAH to PRH in terms of travel times and the level of effectiveness achieved.
PRH distributes all the available doses among the sub-regions based on the proportion of the total eligible population in the entire region who reside in a sub-region. Since this allocation scheme does not take the sub-regions’ capacities into consideration, if a sub-region’s capacity is below their pro rata allocation, the doses are either not distributed to that region, or wasted. Therefore, we define the following equation for allocating vaccines for PRH:(26) Vi=min(⌊Ci⌋,⌊PiA∑j∈JPj⌋),∀i∈J.
To compare the results from PRH to EEVAM and VAH, we also need to obtain values for the decision variables Xij, i.e. the number of people who travel from sub-region i to j to receive their vaccine, for all i,j∈J. To do this, after we make the vaccine allocations based on Eq. (26), we run an optimization model that determines the optimal assignments of people to travel between sub-regions. This approach assumes that once vaccines are allocated to the sub-regions, the individuals who would like to get vaccinated have complete knowledge of the vaccines available in the different sub-regions and act optimally in terms of minimizing the maximum per-person travel duration. The Mixed Integer Linear Program, Pro Rata Vaccine Allocation Model (PRVAM) is presented in Appendix C. In Section 5, we perform extensive experiments to compare the three introduced methodologies for allocating vaccines: EEVAM, VAH, and PRH.
5 Computational experiments
In this section, we perform experiments to (i) compare the results from EEVAM, VAH, and PRH in terms of the levels of equity and effectiveness achieved considering different metrics, (ii) explore how the optimal solution changes for different layout structures and problem settings, and (iii) investigate the differences in how urban (densely-populated) versus rural (sparsely-populated) areas are impacted based on varying problem settings. In order to understand how solutions change for different types of regions, we generate three 200×200 layouts consisting of densely-populated and sparsely-populated areas. We also vary the number of sub-regions |J| and the capacity levels. Fig. 6 illustrates the three layouts and Table 2 provides the experimental design settings. The coordinates for the centroids of sub-regions for each layout, represented by (xj,yj) for entity j, are generated randomly and are also presented in Table 2. We assume that the population-weighted centroids and vaccination location centroids are the same for each geographic entity, i.e., tii=0 for all i∈J. We calculate the distance between each sub-region centroid pair using Euclidean distances and use these distances as a proxy for travel time.Fig. 6 Layouts for computational experiments.
Fig. 6
Table 2 Experimental design.
Table 2Parameter Levels Area Settings
|J| All All 50,100,150,200
(xj,yj) Layout 1 Densely-populated xj∼U(75,125), yj∼U(75,125)
Sparsely-populated xj∼U(0,150), yj∼U(0,150);
if xj>75 and yj>75:xj=xj+50,yj=yj+50
Layout 2 Densely-populated xj∼U(50,150), yj∼U(50,150)
Sparsely-populated xj∼U(0,100),yj∼U(0,100);
if xj>50 and yj>50:xj=xj+50,yj=yj+50
Layout 3 Densely-populated xj∼U(0,50),yj∼U(0,200);
if xj>25,xj=xj+150
Sparsely-populated xj∼U(25,175),yj∼U(0,200)
Pj All Densely-populated U(500,2000)
All Sparsely-populated U(100,1000)
hj All All U(0.05,0.4)
Cj Nominal All ∑j∈J⌈Pj(1−hj)⌉|J|U(0.8,1.4)
Higher on Dense Densely-populated ∑j∈J⌈Pj(1−hj)⌉|J|U(1.1,1.4)
Sparsely-populated ∑j∈J⌈Pj(1−hj)⌉|J|U(0.8,1.1);
Lower on Dense Densely-populated ∑j∈J⌈Pj(1−hj)⌉|J|U(0.8,1.1)
Sparsely-populated ∑j∈J⌈Pj(1−hj)⌉|J|U(1.1,1.4);
ξ All All 0.1
A All All ∑j∈J⌈Pj(1−hj)⌉
Layout 1 represents a case where a single small densely-populated area exists in the center of the entire region. In Layout 2, the densely-populated area is again at the center but larger. Finally, in Layout 3, there are two densely populated areas on the right and left edges of the region. Note that there are multiple sub-regions within each densely- and sparsely-populated area. If we consider that the entire region represents a state, we can think about sub-regions as counties or census blocks. We use these layouts instead of a real-life geographical area in our experiments for two main reasons: (i) using these layouts allows us to test extreme situations, such as comparing when a very densely-populated area (i.e. an urban area) is surrounded by a sparsely-populated area (i.e. a rural area), versus two urban areas being situated at the outer edges of a mostly rural area and (ii) it allows us to have more control in the experimental design and be able to fairly compare the individual impact of the layouts while controlling for factors such as the number of sub-regions, population quantities and capacities for different layouts. We assume that half of |J| sub-regions are located in the densely-populated area (shaded regions in Fig. 6) and the other half in sparsely-populated area (solid white regions in Fig. 6) for each layout. Our goal is to use Layouts 1 and 2 to compare results when the densely-populated area is in the center but larger (Layout 2) versus smaller (Layout 1). Further, Layouts 2 and 3 have the same area size for both densely-populated and sparsely-populated areas so our goal is to use those layouts to see if there are differences between a case when the dense area is at the center versus split into two sides. The population at each sub-region also follows a different discrete uniform distribution based on whether it is located in the densely-populated, or sparsely-populated area (higher population at the dense area). Note that the capacities are generated using uniform distributions such that Assumption 4 is satisfied, i.e., the total capacity in the entire region is sufficient to satisfy the total need. We consider three levels for capacity: (i) the nominal case where the distribution used to generate the capacities is the same for all sub-regions, (ii) the “higher on dense” case where sub-regions in the dense area have higher capacities than the sub-regions in the sparse area, and (iii) the “lower on dense” case where sub-regions on the dense area have lower capacities than the sub-regions on the sparse area. Considering three layouts, four levels for J, and three levels for capacities Cj results in 36 test settings. We generate 10 parameter instances for each random parameter, and 10 layout instances for each layout setting, resulting in a total of 100 instances for each problem test setting. All analyses were done on a 64-bit Intel Core i7 machine with 3.6 GHz processor and 64 GB memory. Computation times were negligible for all models and heuristics (maximum run time below one minute per instance) and hence, are not reported.
5.1 Comparisons of EEVAM and the heuristic methods
We compare the solutions from EEVAM, VAH, and PRH in terms of several metrics. We start by defining the metrics that we use for performance assessment. Specifically, we use two travel time-related metrics (Emax and Eavg) and two effectiveness-related metrics (% Unmet Need and % Unused Vaccines). Let Ei represent the average travel time for a person residing in sub-region i to get vaccinated, i.e.,(27) Ei=∑j∈J(tijPVδijPVXij+tijPTδijPTXij)Pi(1−hi)∀i∈J.
We define the travel time-related metrics Emax and Eavg as:(28) Emax=maxi∈JEi,
and(29) Eavg=∑i∈JEi|J|.
The metrics Emax and Eavg capture the maximum per-person travel duration and the average per-person travel duration across all the sub-regions respectively.
We define the effectiveness-related metrics as follows where [x]+ refers to max(0,x):(30) %UnmetNeed(region)=∑j∈J[Pj(1−hj)−Vj]+∑j∈JPj(1−hj).
Note that we can also define the % Unmet Need for a single sub-region i as follows:(31) %UnmetNeedi=[Pi(1−hi)−Vi]+Pi(1−hi)∀i∈J.
Similarly, we can make the following definitions:(32) %UnusedVaccines(region)=∑j∈J[Vj−Pj(1−hj)]+∑j∈JVj,and
(33) %UnusedVaccinesi=[Vi−Pi(1−hi)]+Vi∀i∈J.
Note that Emax is essentially equivalent to the objective function of EEVAM and hence represents the level of equity achieved in the region. Thus, we will also refer to Emax as the equity metric. Therefore, when we compare VAH and EEVAM in terms of this performance measure, we expect to see EmaxEEVAM≤EmaxVAH. However, a similar expectation cannot be stated for Eavg as this measure considers the average across all the sub-regions. In fact, we expect VAH to perform better for Eavg than Emax due to its myopic nature of making allocations sequentially. On the other hand, PRH is expected to perform very well for Emax since the vaccine allocation scheme is entirely based on need in each sub-region. However, since PRH does not consider capacities and vaccine hesitancy, we would expect it to have a much worse level of effectiveness in terms of a high percentage of unmet need.
Results from EEVAM, VAH and PRH for varying number of sub-regions,|J|Fig. 7 illustrates the results of metrics (a) Eavg, (b) Emax, (c) % Unmet Need, and (d) % Unused Vaccines from EEVAM (blue circle), VAH (orange square) and PRH (green asterisk) for varying levels of |J|. The error bars represent the 95% confidence intervals for each parameter setting, which are narrow for all parameter settings. In these instances, we use the Nominal Capacity levels and consider all layouts. We will later explore the impact of capacity and layouts on solutions.Fig. 7 The impact of |J| on (a) Eavg, (b) Emax, (c) % unmet need, and (d) % unused vaccines for EEVAM, VAH, and PRH.
Fig. 7
VAH provides the best Eavg, %UnmetNeed (tied with EEVAM) and %UnusedVaccines (tied with PRH) values for all considered levels of |J| and for Nominal Capacity. This is expected as VAH follows an approach of allocating vaccines based on need first (similar to PRH, but also considering vaccine hesitancy), and then based on a myopic approach of focusing on one sub-region at a time, starting with the sub-region with the lowest capacity over need ratio. As a result, most people are assigned to closer sub-regions. However, as expected due to the myopic nature of the heuristic, the people residing in sub-regions that still have insufficient supply but have higher capacity over need ratios may have to be assigned to areas that are farther away, which causes a high Emax value, as can be seen from Fig. 7(b). PRH provides the lowest levels of Emax since by definition it only allocates vaccines based on need and does not consider capacity or hesitancy. However, this achievement in equity comes at a cost of effectiveness as based on PRH’s allocation, many vaccines end up not being allocated to the region as capacities are not considered. As a result, on average, over 10% of the need is not met. In contrast, EEVAM provides the best results for %UnmetNeed (tied with VAH) and meets the need in the entire region. The level of Emax achieved is slightly worse than PRH. The %UnusedVaccines is highest for EEVAM; however, this is because we set ψ=0.1 in our models and therefore this outcome can be controlled by the decision makers.
From Fig. 7(a), we see that as the number of sub-regions |J| increases, all methods result in better Eavg values which is expected as we would have more locations where vaccines can be allocated while keeping the size of the region the same. Fig. 7(b) shows the same trend for Emax with the exception of VAH’s behavior. VAH results in a higher Emax value for |J|=100 than for |J|=50, which is unexpected. We believe this is due to the myopic nature of VAH and the generation of some regions that are harder to reach when we increase |J| to 100. No significant change is observed for %UnmetNeed and %UnusedVaccines for any of the methods considered as these measures are represented as percentages and are not impacted by the number of nodes.
Results from EEVAM, VAH, and PRH for varying capacity levelsFig. 8 illustrates the same metrics as Fig. 7 but for varying levels of capacity. Here we set |J|=100 as our base case and consider all layouts. We will focus on observations that are different than the observations made previously. First, we see from Fig. 8(a) and (b) that the travel time-related metrics improve for all methods as more capacity is allocated to denser regions. However, the change in EEVAM and VAH are more drastic than the change for PRH. Indeed, we see that when the capacity is high on the dense area, EEVAM outperforms PRH for Eavg and almost achieves the same level of Emax. On the other hand, the %UnmetNeed levels of PRH improve drastically as capacity on dense areas is increased, as can be seen from Fig. 8c. This is intuitive as in an extreme case where the capacities of all regions are greater than their need, we would expect PRH to meet all the need and perform optimally. Finally, since the %UnusedVaccines is not a function of capacity, but of need and hesitancy, the levels are not impacted by the change in capacity levels.Fig. 8 The impact of capacity on (a) Eavg, (b) Emax, (c) % unmet need, and (d) % unused vaccines for EEVAM, VAH, and PRH.
Fig. 8
Results from EEVAM, VAH and PRH for varying layout settings Finally, Fig. 9 illustrates the same metrics as Figs. 7 and 8 but for different Layouts. We set |J|=100 and use the Nominal Capacity levels. As a reminder, Layouts 1 and 2 use similar structures of a single dense region in the center but Layout 2 has a larger dense region than Layout 1. Layouts 2 and 3 have the same total area of dense and sparse regions but Layout 3 has two dense areas separated into the farther edges whereas Layout 2 has a single dense area. Examining Fig. 9(a) and (b), we see that EEVAM gives almost the same Eavg and Emax values for Layouts 1 and 2. Therefore, we can say that the model is robust to changes in the size of the dense region based on our experiments. However, it provides much higher levels of both metrics for Layout 3. Therefore, EEVAM performs worse when multiple dense regions are farther away from each other. We believe this can be explained as follows: since the dense areas have more capacity limitations, it becomes harder to find a location to allocate people in dense regions to when the capacity is decentralized. When dense regions are centralized, this allows for the pooling of capacity and makes it easier for people to find vaccines. PRH also shows a similar pattern to EEVAM.Fig. 9 The impact of layout settings on (a) Eavg, (b) Emax, (c) % unmet need, and (d) % unused vaccines for EEVAM, VAH, and PRH.
Fig. 9
In contrast, we can see from Fig. 9(a) that VAH, which provides the best levels of Eavg, provides better results for this metric for Layouts 2 and 3 as compared to Layout 1. We believe this is because this heuristic has a myopic nature and when the dense area is small, has difficulty allocating vaccines to some nodes with insufficient capacity that are located in the dense region. Finally, no significant change is observed for the effectiveness measures as the layouts use the same level of need and hesitancy.
5.2 Comparison of the results from densely- and sparsely-populated regions for model EEVAM
In this section, we use EEVAM’s results to examine how the densely-populated and sparsely-populated areas are impacted by varying levels of capacity and layout settings. We set |J|=100. Fig. 10 presents the box plots for the Eavg metric for dense (blue) and sparse (orange) areas for varying levels of capacity and each layout. First, we notice that dense areas always have higher levels of Eavg. The Eavg levels improve as we allocate higher capacity to dense areas. This is in parallel with what occurred in real life, as when the vaccines initially became available, many people living in urban areas had to travel to rural areas to receive their vaccines. We also saw a similar pattern in our Case Study in Section 3.1. The differences between urban and rural areas for Emax were not significant; therefore, we do not discuss them here. This result is expected as in a min-max model the maximum level of the metric would be observed in several locations with limited capacity.Fig. 10 Comparison of average Eavg values for dense and sparse regions for varying layout settings and capacity levels.
Fig. 10
Fig. 11 presents similar box plots for the effectiveness measures. Notice that in Section 5.1, we used an aggregate measure for the %UnmetNeed and %UnusedVaccines metrics for the entire region (Eqs. (30) and (32)). When considered at the aggregate level, we found that %UnmetNeed(region)=0 for EEVAM for all instances considered. That is, enough vaccines were allocated to satisfy the need of the entire region. In this section, we calculate the %UnmetNeed and %UnusedVaccines metrics separately for the dense and sparse regions as follows:(34) %UnmetNeed(dense)=∑j∈JD[Pj(1−hj)−Vj]+∑j∈JDPj(1−hj),
(35) %UnmetNeed(sparse)=∑j∈JS[Pj(1−hj)−Vj]+∑j∈JSPj(1−hj),
where JD and JS refer to the sub-regions in densely-populated and sparsely-populated areas respectively. Similarly, we let(36) %UnusedVaccines(dense)=∑j∈JD[Vj−Pj(1−hj)]+∑j∈JDVj,
(37) %UnusedVaccines(sparse)=∑j∈JS[Vj−Pj(1−hj)]+∑j∈JSVj.
Fig. 11 Comparison of the effectiveness metrics for dense and sparse regions for varying capacity levels.
Fig. 11
Fig. 11 (a) shows that the percentage of unmet need is higher in the dense areas compared to sparse areas for all capacity levels considered. However, the difference decreases as the capacity of dense areas increases. We would expect this gap to decrease as more capacity is allocated to the areas with the highest need. Also, we observe the opposite behavior for the percentage of unused vaccines on Fig. 11(b). Because the capacity in the densely populated areas is too small to satisfy all the need, the vaccines get allocated to the sparsely-populated regions, causing a surplus of vaccines in those areas.
Finally, Fig. 12 (a)–(i) illustrate the maps for the considered regions for each layout where each sub-region is color-coded based on a metric. Fig. 12(a)–(c) show the value of the Ei metric for each sub-region, (d)–(f) show the values of %UnmetNeed (Eq. (31)) and (g)–(i) highlight the values of %UnusedVaccines (Eq. (33)) for each sub-region. The color gradient is based on the magnitude of the considered metric and a gradient legend is provided for each figure. We select a single layout seed (instance) to be able to visualize the coordinates and take the average of the results for each sub-region over ten parameter seed results. We make some interesting observations. First, from Fig. 12(a)–(c), we see that the Ei values decrease as we move away from the dense regions. Although we had already seen that the Ei values are higher for dense areas, this observation shows that even if a sub-region is located in a sparse area, its proximity to the dense area may cause them to have lower access to vaccines. Therefore, not all sub-regions in sparse areas are equal in terms of their equity levels. Examining Fig. 12(d)–(i), we see that a similar relationship is not observed for the effectiveness metrics. These metrics show similar levels for the sub-regions within the densely-populated and sparsely-populated areas and highlight a large percentage of unmet need in dense areas and a large percentage of unused vaccines in sparse areas.Fig. 12 Ei, %unmet need, and %unused vaccines values for all sub-regions for all layout settings.
Fig. 12
6 Model extensions
In this section, we examine two ways our EEVAM model can be extended to address special considerations that impact vaccine access and disease progression. Specifically, in Section 6.1, we examine the improvements in vaccine hesitancy that can potentially be achieved through increasing vaccine access. In Section 6.2, we consider a special case where, although sufficient vaccines are available for the entire population, decision makers wish to prioritize transportation access for specific subgroups that have a higher risk of severe disease.
6.1 Lowering vaccine hesitancy by improving access
In Section 1, we discussed that one of the factors that are known to impact vaccine hesitancy is convenience, which is related to transportation access. In this section, we explore an alternative to EEVAM where we model vaccine hesitancy as a decision variable that is a direct function of the number of vaccines allocated to each region. The research questions that we explore in this section are: (i) Assuming a linear relationship between vaccine hesitancy and vaccine access, how much improvement can we achieve in reducing the number of vaccine-hesitant people? (ii) Does this improvement vary by vaccination capacity and population density? (iii) What is the impact of improving vaccine hesitancy on transportation equity, the unmet need, and the unused vaccines in the region?
We assume that the vaccine hesitancy level in a sub-region j, hj, decays linearly as a function of the number of vaccines allocated to sub-region j, Vj. Letting h^j represent the nominal hesitancy levels based on our experimental design (see Table 2), we assume a function of the form hj(Vj)=h^j−CVj where C is a constant. However, since we do not expect hesitancy levels to reach zero in reality, we assume that the lowest vaccine hesitancy levels are achieved when Vj=Pj, that is when a vaccine is available for every person residing in that sub-region. Although we could not find any studies in the literature that show the mathematical relationship between the number of vaccines available and hesitancy, Siegler et al. [92] examine the association between initial vaccine hesitancy and subsequent vaccination and find that approximately 32% of individuals who initially reported hesitancy to the COVID-19 vaccine eventually received the vaccine. Thus, we use this number as our minimum possible value and use it to estimate C. The resulting function for each county j becomes hj=h^j−0.68h^jPjVj, which is illustrated in Fig. 13 .Fig. 13 Hesitancy as a function of allocated vaccines.
Fig. 13
We make several modifications on EEVAM to address this problem. First, setting the hesitancy rate hj to be a decision variable makes the mathematical program nonlinear due to constraint (2). To address this, we define the maximum average travel duration E as a parameter that we vary through our experiments. We also modify the objective function to minimize the total number of vaccine hesitant people in the region through the allocation of vaccines. Thus, objective function (1) is replaced with the following:(38) min∑j∈JPjhj
Lastly, we define the variable hesitancy rate through the addition of the following constraints:(39) hj≥h^j−0.68h^jPjVj∀j∈J
(40) hj≥0.32h^j∀j∈J
(41) hj≤h^j∀j∈J
We refer to this modified model as Vaccine Hesitancy Model (VHM). The complete formulation for VHM is presented in Appendix D.
6.1.1 Results
We run VHM considering the experimental design settings as defined in Table 2 only considering Layout 1. For the travel duration limit E (which is now a parameter), we use the settings {0,5,10,15,20,25,30}. However, some parameter settings lead to infeasible solutions which we discuss below:• All solutions are infeasible for E=0. In this case, we are enforcing that each person should get vaccinated in the sub-region that they reside in. However, this results in infeasible solutions due to capacity limitations.
• For |J|=50, all solutions are infeasible for E<30, however for |J|∈{100,150,200}, we obtain optimal solutions for E≥10. This shows that if the population density is very low, for example in a case where the entire region is rural, it is difficult to achieve average travel durations that are low. In this case, if there are only 50 sub-regions, the model cannot find any solutions that achieve a maximum average travel duration less than 30 min.
Based on these results, we only consider the solutions for E≥10 and |J|≥100. We define a new metric called the % Improvement in Hesitancy which is calculated as ∑j∈JPjh^j−∑j∈JPjhj∑j∈JPjh^j. Fig. 14 (a) examines the relationship between the % improvement in hesitancy and the transportation duration limit, E, for varying capacity levels. We find that the improvement in hesitancy either increases or stays the same as we increase E. In other words, if the travel duration limit is restrictive, this limits the model’s ability to achieve higher improvements in hesitancy as people are not allowed to travel to longer distances. However, all the lines becoming flat after a certain value of E shows that solutions can be achieved without having to increase E drastically (E=15 appears to provide maximum improvement for all capacity settings). Further, the improvements increase as more capacity is allocated to population-dense areas where more people reside. This result can be attributed to more people residing in dense areas, which leads to a greater potential for improving hesitancy. Fig. 14(b) shows that the improvement in hesitancy is most sensitive to capacity levels when the population density is low (for lower |J|). This is because when there are a lower number of sub-regions, there are less opportunities for people to get vaccinated (or they have to travel to farther distances), and hence, allocating capacity to areas where there is great need (dense areas) has a greater impact on reducing the hesitancy levels.Fig. 14 % Improvement in hesitancy versus (a) transportation duration limit, E, and (b) capacity settings.
Fig. 14
Fig. 15 examines the relationship between the percent improvement in hesitancy and (a) % Unmet Need and (b) % Unused Vaccines in the region. We see that there is a negative relationship between the reduction in hesitancy and both the percentage of unmet need in the region (correlation coefficient, r=−0.6) and the percentage of unused vaccines (r=−0.39) in the region. This is a promising result because it shows that through this approach of trying to reduce vaccine hesitancy by allocating more vaccines, we are not only able to reduce hesitancy levels, but we are also able to meet more demand and also waste fewer vaccines.Fig. 15 % Improvement in hesitancy versus (a) % unmet need, and (b) % unused vaccines.
Fig. 15
6.2 Consideration of population risk groups
Although in this paper we focus on CDC’s Phases 2 and 3 of vaccination, when sufficient doses are available to vaccinate the entire population, decision makers may still want to prioritize some high-risk sub-populations for ensuring vaccine access. In this section, we propose a modified model which prioritizes minimizing the average travel duration by the high-risk (HR) sub-populations over low-risk (LR) sub-populations by using a weight. The research questions that we explore in this section are: (i) If we consider a policy where HR sub-population is prioritized over LR sub-population for vaccine access, how much improvement can we achieve on the HR sub-population’s travel duration? (ii) What would be the impact of such a policy on LR sub-population’s travel duration, the unmet need, and the unused vaccines in the region? We now describe the changes in notation and formulation that we make to EEVAM.
The population in each sub-region is now separated into two groups: the HR sub-population at sub-region j (PjHR) and the LR sub-population at sub-region j (PjLR). Further, we also distinguish between the number of people to travel from i to j for each risk group by letting XijHR (XijLR) represent the number of HR (LR) people who travel from i to j. The average travel duration constraint (2) in EEVAM is replaced with the following set of constraints, where ψ≥1 represents the weight associated with the average travel duration for HR group:(42) ψ∑j∈J(tijPVδijPVXijHR+tijPTδijPTXijHR)PiHR(1−hi)≤E∀i∈J
(43) ∑j∈J(tijPVδijPVXijLR+tijPTδijPTXijLR)PiLR(1−hi)≤E∀i∈J
Constraint (42) limits the average travel duration for a HR person to be below E/ψ whereas constraint (43) limits the average travel duration for a LR person to be below E. To understand this, consider a case where the optimal value of E has been found to be 60 min by the model. Let us also assume that ψ=2. This would enforce the average travel duration for a LR person to get vaccinated to be at most 60 min at each sub-region; however the average travel duration for a HR person to get vaccinated to be at most 60/ψ=30 min so the HR sub-population would be prioritized. Constraints (3), (4), and (5) on EEVAM are also replaced with the following:(44) ∑j∈JXijHR≥PiHR(1−hi)∀i∈J
(45) ∑j∈JXijLR≥PiLR(1−hi)∀i∈J
(46) ∑i∈JXijHR+∑i∈JXijLR≤Vj∀j∈J
(47) Vj−(∑i∈JXijHR+∑i∈JXijLR)≤ξVj∀j∈J
Constraints (44) and (45) ensure that the number of HR and LR people residing in i and traveling to any sub-region j to get vaccinated should be equal to the respective sub-populations of sub-region i to be vaccinated. Constraints (46) and (47) have the same meanings as EEVAM. We refer to this modified model as Population Risk Groups Model (PRGM) and present its complete formulation in Appendix E.
6.2.1 Results
In this context, we use CDC’s definition of high-risk sub-populations for COVID-19 based on age and comorbidities [93] to determine the high-risk and low-risk sub-population counts in each sub-region. We let α represent the HR sub-population rate such that PjHR=⌈αPj⌉ and PjLR=Pj−PjLR. It is estimated that 37.6% of adults in the United States are at high risk for COVID-19 [94] therefore we set α=0.376 and obtain the sub-population counts for HR and LR groups for each sub-region. We use the experimental design settings defined in Table 2 and limit our analysis to Layout 1, Nominal Capacity, and |J|=100. For the HR sub-population travel time weight, ψ, we use the settings {1,1.5,2,2.5,3}. Fig. 16 illustrates the change in (a)Emax, and (b)Eavg for varying ψ. The blue circle-marker line represents the average results for the HR sub-population, the orange square-marker line represents the average results for the LR sub-population, and the green dashed line represents the average results for the entire population.Fig. 16 The impact of ψ on (a) Emax, and (b) Eavg, for HR, LR and the entire sub-population.
Fig. 16
The maximum travel duration Emax and the average travel duration Eavg show similar patterns. For both cases, the travel durations for HR sub-populations decrease drastically as ψ is increased, as the travel durations for LR sub-populations increase. As expected, the travel durations for the entire population are not impacted. The good news is that the change in LR sub-populations’ travel durations is not as drastic as HR sub-populations. For example, setting ψ=3, we see that the maximum travel duration for HR sub-population decreases from 10.8 min to 4.8 min (a 55% decrease), whereas the maximum travel duration for the LR sub-population increases from 10.8 min to 14.4 min (33% increase). So, this indicates that we can achieve considerable improvements in the travel time for HR sub-populations while not sacrificing LR sub-populations completely. We would like to note that changing ψ has no impact on the % unmet need and % unused vaccines metrics.
7 Conclusions
The COVID-19 pandemic has caused an unprecedented impact worldwide. Although vaccines were shown to be the strongest defense for curbing the spread of the disease, they also highlighted challenges associated with vaccine supply chains and allocation decisions. We studied the problem of allocating vaccine supplies among different sub-regions in order to achieve transportation equity in terms of the travel durations of patients. We also considered the objectives of effectiveness with the goal of minimizing the unused vaccine doses while satisfying the need, and underlying factors, i.e. vaccination capacities of the different sub-regions and vaccine hesitancy levels. We first developed a Mixed-Integer Linear Program to formulate this problem and demonstrated an application of our model as a case study based on real-life data from the state of Alabama. Although our developed model captures the objectives and underlying constraints, policymakers may prefer algorithms that are easier to implement and understand, e.g., the pro rata policy that was used by many state health departments to allocate COVID-19 vaccines. As an alternative to the naive pro rata policy, we also developed a heuristic that considers the vaccination capacity and vaccine hesitancy levels in the different sub-regions. We then performed extensive experiments to compare results for the proposed mathematical model, the developed heuristic, and the pro rata allocation policy based on several metrics.
We would like to highlight several policy-related insights from our results. First, our case study shows that allocating additional capacity in a small subset of sub-regions leads to drastic improvements in the overall levels of equity achieved throughout the region. Therefore, policymakers may use our proposed models to make strategic decisions about how to allocate additional vaccination capacity among the sub-regions. Second, if policymakers prefer using simple and easy-to-implement algorithms, they should choose VAH over PRH if they want to achieve a policy that achieves lower travel durations on average while meeting the need and not wasting vaccine supplies. However, if policymakers would like to prioritize equity and make sure no sub-region is at a disadvantage, they should use EEVAM. This model meets all the need and provides the flexibility to control the level of unused vaccines in a sub-region. Finally, the capacities of the sub-regions highly impact the level of equity achieved, and urban areas with high demand may not have sufficient capacity to satisfy their need causing people to travel to rural areas to get vaccinated. In this case, although a sub-region may be considered a rural area, its proximity to urban area(s) impacts its level of vaccine access. Rural sub-regions that are closer to urban areas are more likely to have lower levels of vaccine access due to the high need and relatively low capacity in those areas. We also consider two extensions to our model for addressing some specific goals and priorities of policymakers. First, we introduce a modified model that aims to reduce vaccine hesitancy by strategically allocating vaccines to sub-regions while controlling the allowed maximum per-person travel duration. We find that this approach is especially beneficial in population-dense areas, however, capacity is an important factor as low capacity hinders the ability to reduce hesitancy. We also find that policymakers can use this approach to reduce vaccine hesitancy, while also reducing the unmet need and unused vaccines. Second, we consider a case where policymakers would like to prioritize certain sub-populations that are at risk for severe disease. Assuming a 37.6% high-risk population rate, we run a slightly modified version of our model and find that we can achieve considerable improvements for the travel duration of high-risk sub-populations without completely sacrificing the travel duration of low-risk sub-populations.
Our study has several limitations. First, we assume that each person must travel once to get vaccinated. Some of the COVID-19 vaccines consist of multiple doses so an extension of our model would be to incorporate different types of vaccines and the corresponding number of trips that people would have to make to get vaccinated. Second, our models assume that all the underlying parameters are deterministic and known. In reality, vaccine hesitancy levels are stochastic and dynamic. Further, the capacities of the different sub-regions may also vary over time as new vaccination sites are opened, or existing ones are closed. A possible extension of our model is to develop a stochastic model that considers the uncertainty in capacity and hesitancy levels. Third, our models do not consider the possibility of random shocks in vaccine supply due to issues such as shipment delays, contamination, or obsolescence. Thus, the design of allocation methods or schemes that are robust against such shocks is a needed area for future study.
Finally, although our study was motivated by the vaccine allocation decisions during the COVID-19 pandemic, our models can be used in different settings where the equitable and effective allocation of resources is considered under varying capacity and demand levels. Some possible examples of these are the allocation of healthcare staff to rural areas, the distribution of medications, such as AIDS medicines, in countries with high prevalence rates, and the allocation of education resources, e.g. funds, supplies, teachers, or scholarships, to the different districts.
CRediT authorship contribution statement
Irem Sengul Orgut: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Visualization, Supervision, Writing – original draft, Writing – review & editing. Nickolas Freeman: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. Dwight Lewis: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. Jason Parton: Conceptualization, Methodology, Data curation, Writing – original draft, Writing – review & editing.
Declaration of Competing Interest
Authors declare that they have no conflict of interest.
Appendix A Proof of Lemma 1
Proof Proof of Lemma 1
We first show that the solution provided in Eqs. (24) and (25) are feasible for EEVAM when Constraint (6) is removed. Then we show that (24) and (25) are indeed the optimal solutions.
Since E does not have an upper bound, but only has a lower bound of zero, and Xij≥0 by (24) and (25), constraint (2) is satisfied.
Inserting (24) and (25) into (3), we obtain(48) ∑j∈JXij=Xii=Vi=⌈Pi(1−hi)⌉≥Pi(1−hi)∀i∈J
so, constraint (3) is satisfied.
Next, inserting (24) into (4), we get(49) ∑i∈JXij=Xjj=Vj≤Vj∀j∈J
so, constraint (4) is also satisfied.
Inserting (25) into (5), we get(50) Vj−∑i∈JXij=Vj−Xjj=Vj−Vj=0≤ξVj
hence, constraint (5) is also satisfied.
Constraint (6) is skipped due to condition of the lemma.
Inserting (24) into (7), we obtain(51) ∑j∈JVj=∑j∈J⌈Pj(1−hj)⌉≤A
by Assumption 1.
Finally, constraints (8) and (9) are satisfied by definition.
Therefore, the solutions given by Eqs. (24) and (25) are feasible for EEVAM when Constraint (6) is removed. Next, we prove optimality.
We provide a proof by contradiction. We will use the superscript “L” to refer to the solution proposed in Lemma 1 and “A” to refer to an alternative solution. First, let us calculate the left-hand side of constraint (2) based on (24) for a given sub-region i, and refer to this as EiL.(52) EiL=∑j∈J(tijPVδiPVXijL+tijPTδiPTXijL)Pi(1−hi)=tiiPVδiPV⌈Pi(1−hi)⌉+tiiPTδiPT⌈Pi(1−hi)⌉Pi(1−hi)=⌈Pi(1−hi)⌉(tiiPVδiPV+tiiPTδiPT)Pi(1−hi)
We would like to make several observations: (i) Since E≥EiL for all i by constraint (2), any increase in EiL for any sub-region i will cause the objective function E to increase. Therefore, the only way to achieve a lower objective function value is by reducing the left-hand side of constraint (2) for all sub-regions i. (ii) The value of the number of vaccines allocated to a sub-region i, Vi, does not have a direct impact on the objective function and only impacts the objective function indirectly through constraining Xij for all i and j. Now, consider an alternative solution as the following(53) XiiA=⌈Pi(1−hi)⌉−ϵ
where ϵ is a positive integer value. In this case, in order for the solution to be feasible, we must have(54) ∑i∈J,i≠jXijA≥ϵ
to satisfy constraint (3). For the sake of simplicity, since we are trying to lower the left-hand side of Eq. (2) as much as possible, let us set ∑i∈J,i≠jXijA=ϵ. Further, without loss of generality, we are going to set XijA=ϵ for the sub-region j that is closest to sub-region i and set XikA=0 for other sub-regions k∈J.
We now write the left-hand side of constraint (2) for this alternative solution as the following:(55) EiA=∑j∈J(tijPVδiPVXijA+tijPTδiPTXijA)Pi(1−hi)=tiiPVδiPV(⌈Pi(1−hi)⌉−ϵ)+tijPVδiPVϵ+tiiPTδiPT(⌈Pi(1−hi)⌉−ϵ)+tijPTδiPTϵPi(1−hi)=⌈Pi(1−hi)⌉(tiiPVδiPV+tiiPTδiPT)Pi(1−hi)+ϵ(−tiiPVδiPV+tijPVδiPV−tiiPTδiPT+tijPTδiPT)Pi(1−hi)=⌈Pi(1−hi)⌉(tiiPVδiPV+tiiPTδiPT)Pi(1−hi)+ϵ(δiPV(−tiiPV+tijPV)+δiPT(−tiiPT+tijPT))Pi(1−hi).
We would have EiA≤EiL if and only if the second term in Eq. (55) would be negative, i.e. ϵ(−tiiPVδiPV+tijPVδiPV−tiiPTδiPT+tijPTδiPT)Pi(1−hi)<0. Since by definition traveling to a different sub-region would on average take longer than getting vaccinated in the same sub-region that a person resides in, we have tiiPV<tijPV and tiiPT<tijPT for i≠j. Therefore, −tiiPV+tijPV>0 and −tiiPT+tijPT>0. Since we have ϵ>0, δiPV≥0, and δiPT≥0, this indicates that EiA>EiL, which is a contradiction. Therefore the solution given by Lemma 1 is optimal under the given conditions. □
Appendix B Vaccine Allocation Heuristic pseudocode
The psudocode for the Vaccine Allocation Heuristic is presented below as Algorithm 1 :Algorithm 1 Vaccine allocation heuristic.
Algorithm 1
Appendix C Pro Rata Vaccine Allocation Model (PRVAM)
We introduce a Mixed Integer Linear Program, Pro Rata Vaccine Allocation Model (PRVAM) for obtaining optimal assignments of people to sub-regions based on the vaccine allocations by the Pro Rata Heuristic. We use the same notation introduced in Section 3, except that vaccine allocations Vj are now considered to be deterministic parameters obtained via (26) prior to running the optimization model, rather than decision variables.
PRVAM: (56) minE
(57) Subjectto:Averagetravelduration:∑j∈J(tijPVδijPVXij+tijPTδijPTXij)Pi(1−hi)≤E∀i∈J
(58) Peopletobevaccinated:∑j∈JXij≥⌊Pi(1−hi)∑j∈JPj(1−hj)∑j∈JVj⌋∀i∈J
(59) Sub−regionsupplyconstraint:∑i∈JXij≤Vj∀j∈J
(60) Xij∈Z≥0∀i,j∈J
(61) E∈R≥0.
PRVAM determines assignments of people to the sub-regions in order to minimize the maximum average travel duration across the sub-regions, which is the same objective of EEVAM. The main difference in comparison to EEVAM is that the vaccine allocation decisions are made prior to running the model, therefore the Vj are deterministic parameters. Constraint (58) ensures that people who are willing to get vaccinated receive assignments. However, since PRH does not consider capacity restrictions, there may not be sufficient doses allocated to the region to satisfy the entire need and we make sure that the satisfied need is equitable across service regions, i.e., the satisfied need for sub-region j is not less than the pro rata allocation of vaccines based on each sub-region’s need. Constraint (59) limits the total assignment to sub-region j to be no greater than the number of vaccines allocated to that sub-region. Finally, constraint (60) requires the decision variables Xij to be non-negative integers, and constraint (61) requires E to be a non-negative continuous variable.
Appendix D The formulation for Vaccine Hesitancy Model (VHM)
We first define the notation for VHM:
Notation
Sets and indices:
J: set of sub-regions, J={i,j:1,2,…|J|}
Parameters:
E: The maximum average travel duration (minutes) for people residing in a sub-region, across the entire region
Pj: Vaccine-eligible population in sub-region j, j∈J
hj^: Nominal vaccine hesitancy rate in sub-region j, 0≤hj^≤1j∈J
tijPV: Travel duration (minutes) between sub-regions i and j by personal vehicle, i,j∈J
tijPT: Travel duration (minutes) between sub-regions i and j by public transportation, i,j∈J
δijPV: Proportion of people traveling from i to j estimated to use personal vehicle, i,j∈J
δijPT: Proportion of people traveling from i to j estimated to use public transportation,
δijPV+δijPT=1, i,j∈J
Cj: Vaccination capacity (in number of doses) in sub-region j, i,j∈J
ξ: Maximum limit on the percentage of unused vaccine doses in each sub-region, 0≤ξ≤1
A: Number of vaccine doses available for allocation
Decision variables:
Vj: Number of vaccine doses to allocate to sub-region j, j∈J
hj: Vaccine hesitancy rate in sub-region j, j∈J
Xij: Number of people who will travel from sub-region i to j to receive their vaccine, i,j∈J
We present the formulation for VHM as follows:
VHM: (62) min∑j∈JPjhj
(63) Subjectto:Averagetravelduration:∑j∈J(tijPVδijPVXij+tijPTδijPTXij)−Pi(1−hi)E≤0∀i∈J
(64) Peopletobevaccinated:∑j∈JXij≥Pi(1−hi)∀i∈J
(65) Sub−regionsupply:∑i∈JXij≤Vj∀j∈J
(66) Unusedvaccinecap:Vj−∑i∈JXij≤ξVj∀j∈J
(67) Capacityconstraint:Vj≤Cj∀j∈J
(68) Regionsupplyconstraint:∑j∈JVj≤A
(69) Variablehesitancyrate:hj≥h^j−0.68h^jPjVj∀j∈J
(70) hj≥0.32h^j∀j∈J
(71) hj≤h^j∀j∈J
(72) Vj,Xij∈Z≥0∀i,j∈J
The objective function of VHM minimizes the total number of vaccine hesitant people in the population. Constraint (63) is the equity constraint and ensures that the average travel duration per each able and willing person is less than a pre-determined constant, E. Note that in this formulation, E is a parameter that the decision maker controls. Constraints (64)–(68) have the same meaning as in the original model EEVAM. Constraints (69)–(71) define the upper and lower limits on the achievable vaccine hesitancy rates at each region j. Finally, constraints (72) are nonnegativity constraints.
Appendix E The formulation for Population Risk Groups Model (PRGM)
We first define the notation for PRGM:
Notation
Sets and indices:
J: set of sub-regions, J={i,j:1,2,…|J|}
Parameters:
PjHR: Vaccine-eligible population that is at high risk for severe disease in sub-region j, j∈J
PjLR: Vaccine-eligible population that is at low risk for severe disease in sub-region j, j∈J
hj: Vaccine hesitancy rate in sub-region j, 0≤hj≤1j∈J
tijPV: Travel duration (minutes) between sub-regions i and j by personal vehicle, i,j∈J
tijPT: Travel duration (minutes) between sub-regions i and j by public transportation, i,j∈J
δijPV: Proportion of people traveling from i to j estimated to use personal vehicle, i,j∈J
δijPT: Proportion of people traveling from i to j estimated to use public transportation,
δijPV+δijPT=1, i,j∈J
Cj: Vaccination capacity (in number of doses) in sub-region j, i,j∈J
ξ: Maximum limit on the percentage of unused vaccine doses in each sub-region, 0≤ξ≤1
A: Number of vaccine doses available for allocation
ψ: Weight assigned to the average travel duration for high-risk sub-population, ψ≥1
Decision variables:
Vj: Number of vaccine doses to allocate to sub-region j, j∈J
XijHR: Number of high-risk people who will travel from sub-region i to j to receive their vaccine, i,j∈J
XijLR: Number of low-risk people who will travel from sub-region i to j to receive their vaccine, i,j∈J
E: The maximum average travel duration (minutes) for people residing in a sub-region, across the entire region
We present the formulation for PRGM as follows:
PRGM: (73) minE
(74) Subjectto:Averagetravelduration:ψ∑j∈J(tijPVδijPVXijHR+tijPTδijPTXijHR)PiHR(1−hi)≤E∀i∈J
(75) ∑j∈J(tijPVδijPVXijLR+tijPTδijPTXijLR)PiLR(1−hi)≤E∀i∈J
(76) HRpeopletobevaccinated:∑j∈JXijHR≥PiHR(1−hi)∀i∈J
(77) LRpeopletobevaccinated:∑j∈JXijLR≥PiLR(1−hi)∀i∈J
(78) Sub−regionsupply:∑i∈JXijHR+∑i∈JXijLR≤Vj∀j∈J
(79) Unusedvaccinecap:Vj−(∑i∈JXijHR+∑i∈JXijLR)≤ξVj∀j∈J
(80) Capacityconstraint:Vj≤Cj∀j∈J
(81) Regionsupplyconstraint:∑j∈JVj≤A
(82) Vj,XijHR,XijLR∈Z≥0∀i,j∈J
(83) E∈R≥0.
The objective function minimizes the maximum average per-person travel duration, E. Constraint (74) limits the average travel duration for an HR person to be below E/ψ whereas constraint (75) limits the average travel duration for a LR person to be below E. Constraints (76) and (77) ensure that the total number of HR and LR people residing in i and traveling to any sub-region j to get vaccinated should be equal to the respective sub-populations of sub-region i that needs to be vaccinated. Constraints (78)–(83) have the same meaning as in the original model EEVAM.
Data availability
Data will be made available on request.
Area - Supply Chain Management. This manuscript was processed by Associate Editor Pazour.
==== Refs
References
1 Yan Z. Unprecedented pandemic, unprecedented shift, and unprecedented opportunity Hum Behav Emerg Technol 2 2 2020 110 32427197
2 Hui D.S. Azhar E.I. Madani T.A. Ntoumi F. Kock R. Dar O. The continuing 2019-nCoV epidemic threat of novel coronaviruses to global health—the latest 2019 novel coronavirus outbreak in Wuhan, China Int J Infect Dis 91 2020 264 266 31953166
3 Madabhavi I. Sarkar M. Kadakol N. COVID-19: a review Monaldi Arch Chest Dis 90 2 2020 248 258
4 Khan M. Adil S.F. Alkhathlan H.Z. Tahir M.N. Saif S. Khan M. COVID-19: a global challenge with old history, epidemiology and progress so far Molecules 26 1 2021 39
5 Wang R. Hozumi Y. Yin C. Wei G.-W. Mutations on COVID-19 diagnostic targets Genomics 112 6 2020 5204 5213 32966857
6 Johns Hopkins University. COVID-19 dashboard. 2022. https://coronavirus.jhu.edu/map.html, Last accessed on 2022-01-28.
7 Bureau of Labor Statistics. The employment situation—December 2021. 2021. https://www.bls.gov/news.release/pdf/empsit.pdf, Last accessed on 2022-01-28.
8 Reuters. IMF sees cost of COVID pandemic rising beyond $12.5 trillion estimate. 2022. https://www.reuters.com/business/imf-sees-cost-covid-pandemic-rising-beyond-125-trillion-estimate-2022-01-20, Last accessed on 2022-01-28.
9 CDC. COVID-19: how to protect yourself & others. 2022a. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/prevention.html, Last accessed on 2022-01-29.
10 Hou Z. Tong Y. Du F. Lu L. Zhao S. Yu K. Assessing COVID-19 vaccine hesitancy, confidence, and public engagement: a global social listening study J Med Internet Res 23 6 2021 e27632 34061757
11 CDC. COVID-19 vaccine equity for racial and ethnic minority groups. 2021a. https://www.cdc.gov/coronavirus/2019-ncov/community/health-equity/vaccine-equity.html, Last accessed on 2022-01-28.
12 United States Food and Drug Administration. COVID-19 frequently asked questions. 2022. https://www.fda.gov/emergency-preparedness-and-response/coronavirus-disease-2019-covid-19/covid-19-frequently-asked-questions#:∼:text=On%20December%2011%2C%202020,)%20of%20a%20vaccine., Last accessed on 2022-01-29.
13 Kathleen Dooling, MD MPH. Phase 1 allocation COVID-19 vaccine: work group considerations. 2020. https://www.cdc.gov/vaccines/acip/meetings/downloads/slides-2020-09/COVID-07-Dooling.pdf, Last accessed on 2022-01-29.
14 Mitchell S. Andersson N. Ansari N.M. Omer K. Soberanis J.L. Cockcroft A. Equity and vaccine uptake: a cross-sectional study of measles vaccination in Lasbela district, Pakistan BMC Int Health Hum Rights 9 1 2009 1 10 19192307
15 Lee B.Y. Brown S.T. Bailey R.R. Zimmerman R.K. Potter M.A. McGlone S.M. The benefits to all of ensuring equal and timely access to influenza vaccines in poor communities Health Aff 30 6 2011 1141 1150
16 Leclerc P.D. McLay L.A. Mayorga M.E. Modeling equity for allocating public resources Community-based operations research 2012 Springer 97 118
17 Rosenbaum L. Escaping catch-22overcoming COVID vaccine hesitancy N Engl J Med 384 14 2021 1367 1371 33577150
18 Khubchandani J. Sharma S. Price J.H. Wiblishauser M.J. Sharma M. Webb F.J. COVID-19 vaccination hesitancy in the United States: a rapid national assessment J Community Health 46 2 2021 270 277 33389421
19 MacDonald N.E. Vaccine hesitancy: definition, scope and determinants Vaccine 33 34 2015 4161 4164 25896383
20 González-Block M.A. Arroyo-Laguna J. Rodríguez-Zea B. Pelcastre-Villafuerte B.E. Gutiérrez-Calderón E. Díaz-Portillo S.P. Puentes-Rosas E. Sarti E. The importance of confidence, complacency, and convenience for influenza vaccination among key risk groups in large urban areas of Peru Hum Vaccines Immunother. 17 2 2021 465 474
21 CDC. Developing COVID-19 vaccines. 2021b. https://www.cdc.gov/coronavirus/2019-ncov/vaccines/distributing/steps-ensure-safety.html, Last accessed on 2022-01-28.
22 CDC. Find COVID19 vaccines. 2022b. https://www.vaccines.gov/search/, Last accessed on 2022-08-29.
23 United States Census Bureau. Why we ask questions about vehicles available. 2022. https://www.census.gov/acs/www/about/why-we-ask-each-question/vehicles/, Last accessed on 2022-01-28.
24 Araz O.M. Galvani A. Meyers L.A. Geographic prioritization of distributing pandemic influenza vaccines Health Care Manag Sci 15 3 2012 175 187 22618029
25 Sharfstein, et al. Hogan can’t explain away Maryland’s vaccine inequities. He needs a plan to fix them instead. 2021. https://www.washingtonpost.com/opinions/2021/02/27/hogan-cant-explain-away-marylands-vaccine-inequities-he-needs-plan-fix-them-instead/, Last accessed on 2023-02-27.
26 Singer N. Where do vaccine doses go, and who gets them? The algorithms decide 2021 The New York Times
https://www.nytimes.com/2021/02/07/technology/vaccine-algorithms.html
27 Looker R. County health departments seek more vaccines through state allocation process Natl Assoc Count 2021
https://www.naco.org/articles/county-health-departments-seek-more-vaccines-through-state-allocation-process
28 Duijzer L.E. Van Jaarsveld W. Dekker R. Literature review: the vaccine supply chain Eur J Oper Res 268 1 2018 174 192
29 Lemmens S. Decouttere C. Vandaele N. Bernuzzi M. A review of integrated supply chain network design models: key issues for vaccine supply chains Chem Eng Res Des 109 2016 366 384
30 Blasioli E. Mansouri B. Tamvada S.S. Hassini E. Vaccine allocation and distribution: a review with a focus on quantitative methodologies and application to equity, hesitancy, and COVID-19 pandemic Operations research forum vol. 4 2023 Springer 27
31 Uribe-Sánchez A. Savachkin A. Santana A. Prieto-Santa D. Das T.K. A predictive decision-aid methodology for dynamic mitigation of influenza pandemics OR Spectrum 33 3 2011 751 786
32 Tang L. Li Y. Bai D. Liu T. Coelho L.C. Bi-objective optimization for a multi-period COVID-19 vaccination planning problem Omega 110 2022 102617 35185262
33 Mohammadi M. Dehghan M. Pirayesh A. Dolgui A. Bi-objective optimization of a stochastic resilient vaccine distribution network in the context of the COVID-19 pandemic Omega 113 2022 102725 35915776
34 Gilani H. Sahebi H. A data-driven robust optimization model by cutting hyperplanes on vaccine access uncertainty in COVID-19 vaccine supply chain Omega 110 2022 102637 35291647
35 Georgiadis G.P. Georgiadis M.C. Optimal planning of the COVID-19 vaccine supply chain Vaccine 39 37 2021 5302 5312 34373118
36 Domingo D. Biber S. Dian G. Dorey P. Haddad K. Heslop P. Holm I. Liu Y. Pang R. Trans-national equitable strategies of vaccine distribution during the COVID-19 pandemic Math Ind Rep 2021
37 Yarmand H. Ivy J.S. Denton B. Lloyd A.L. Optimal two-phase vaccine allocation to geographically different regions under uncertainty Eur J Oper Res 233 1 2014 208 219
38 Teytelman A. Larson R.C. Multiregional dynamic vaccine allocation during an influenza epidemic Serv Sci 5 3 2013 197 215
39 Rey D. Hammad A.W. Saberi M. Vaccine allocation policy optimization and budget sharing mechanism using reinforcement learning Omega 115 2023 102783
40 Enayati S. Özaltın O.Y. Optimal influenza vaccine distribution with equity Eur J Oper Res 283 2 2020 714 725
41 Arulselvan A. Commander C.W. Elefteriadou L. Pardalos P.M. Detecting critical nodes in sparse graphs Comput Oper Res 36 7 2009 2193 2200
42 Ventresca M. Global search algorithms using a combinatorial unranking-based problem representation for the critical node detection problem Comput Oper Res 39 11 2012 2763 2775
43 Ventresca M. Aleman D. A randomized algorithm with local search for containment of pandemic disease spread Comput Oper Res 48 2014 11 19
44 Ventresca M. Aleman D. A derandomized approximation algorithm for the critical node detection problem Comput Oper Res 43 2014 261 270
45 Nandi A.K. Medal H.R. Methods for removing links in a network to minimize the spread of infections Comput Oper Res 69 2016 10 24
46 Rastegar M. Tavana M. Meraj A. Mina H. An inventory-location optimization model for equitable influenza vaccine distribution in developing countries during the COVID-19 pandemic Vaccine 39 3 2021 495 504 33342632
47 Tavana M. Govindan K. Nasr A.K. Heidary M.S. Mina H. A mathematical programming approach for equitable COVID-19 vaccine distribution in developing countries Ann Oper Res 2021 1 34
48 Wang X. Jiang R. Qi M. A robust optimization problem for drone-based equitable pandemic vaccine distribution with uncertain supply Omega 119 2023 102872 37020741
49 Bertsimas D. Digalakis V. Jr Jacquillat A. Li M.L. Previero A. Where to locate COVID-19 mass vaccination facilities? Naval Res Logist (NRL) 69 2 2022 179 200
50 Emanuel E.J. Persad G. Kern A. Buchanan A. Fabre C. Halliday D. Heath J. Herzog L. Leland R.J. Lemango E.T. An ethical framework for global vaccine allocation Science 369 6509 2020 1309 1312 32883884
51 Jarumaneeroj P. Dusadeerungsikul P.O. Chotivanich T. Nopsopon T. Pongpirul K. An epidemiology-based model for the operational allocation of COVID-19 vaccines: a case study of Thailand Comput Ind Eng 2022 108031 35228772
52 Balcik B. Yucesoy E. Akca B. Karakaya S. Gevsek A.A. Baharmand H. A mathematical model for equitable in-country COVID-19 vaccine allocation Int J Prod Res 60 24 2022 1 25
53 Anahideh H. Kang L. Nezami N. Fair and diverse allocation of scarce resources Socio-Econ Plann Sci 80 2022 101193
54 Fadaki M. Abareshi A. Far S.M. Lee P.T.-W. Multi-period vaccine allocation model in a pandemic: a case study of COVID-19 in Australia Transp Res Part E 161 2022 102689
55 Sahinyazan F.G. Araz O.M. An alternative vaccine prioritization approach in response to COVID-19 pandemic J Humanitarian Logist Supply Chain Manag 12 4 2022 532 553
56 Altay N. Green W.G. III Or/ms research in disaster operations management Eur J Oper Res 175 1 2006 475 493
57 Behl A. Dutta P. Humanitarian supply chain management: a thematic literature review and future directions of research Ann Oper Res 283 1 2019 1001 1044
58 Gupta S. Starr M.K. Farahani R.Z. Matinrad N. Disaster management from a POM perspective: mapping a new domain Prod Oper Manag 25 10 2016 1611 1637
59 Altay N. Capability-based resource allocation for effective disaster response IMA J Manag Math 24 2 2013 253 266
60 Ozbay K. Ozguven E.E. Stochastic humanitarian inventory control model for disaster planning Transp Res Rec 2022 1 2007 63 75
61 Yu L. Zhang C. Yang H. Miao L. Novel methods for resource allocation in humanitarian logistics considering human suffering Comput Ind Eng 119 2018 1 20
62 Balcik B. Ak D. Supplier selection for framework agreements in humanitarian relief Prod Oper Manag 23 6 2014 1028 1041
63 Balcik B. Beamon B.M. Smilowitz K. Last mile distribution in humanitarian relief J Intell Transp Syst 12 2 2008 51 63
64 Ekici A. Özener O.O. Inventory routing for the last mile delivery of humanitarian relief supplies OR Spectrum 42 3 2020 621 660
65 Huang K. Rafiei R. Equitable last mile distribution in emergency response Comput Ind Eng 127 2019 887 900
66 Vitoriano B.n. Ortuño M.T. Tirado G. Montero J. A multi-criteria optimization model for humanitarian aid distribution J Global Optim 51 2 2011 189 208
67 Balcik B. Silvestri S. Rancourt M.-E. Laporte G. Collaborative prepositioning network design for regional disaster response Prod Oper Manag 28 10 2019 2431 2455
68 Dalal J. Üster H. Combining worst case and average case considerations in an integrated emergency response network design problem Transp Sci 52 1 2018 171 188
69 Noyan N. Balcik B. Atakan S. A stochastic optimization model for designing last mile relief networks Transp Sci 50 3 2016 1092 1113
70 Ogryczak W. Multicriteria models for fair resource allocation Control Cybern 36 2 2007 303 332
71 Amartya S. Amartya S.M. Foster J.E. Amartya S. Foster J.E. On economic inequality 1997 Oxford University Press
72 Marsh M.T. Schilling D.A. Equity measurement in facility location analysis: areview and framework Eur J Oper Res 74 1 1994 1 17
73 Balcik B. Iravani S. Smilowitz K. Multi-vehicle sequential resource allocation for a nonprofit distribution system IIE Trans 46 12 2014 1279 1297
74 Felder S. Brinkmann H. Spatial allocation of emergency medical services: minimising the death rate or providing equal access? Reg Sci Urban Econ 32 1 2002 27 45
75 Chanta S. Mayorga M.E. Kurz M.E. McLay L.A. The minimum p-envy location problem: a new model for equitable distribution of emergency resources IIE Trans Healthc Syst Eng 1 2 2011 101 115
76 Khodaparasti S. Maleki H.R. Bruni M.E. Jahedi S. Beraldi P. Conforti D. Balancing efficiency and equity in location-allocation models with an application to strategic EMS design Optim Lett 10 5 2016 1053 1070
77 Bélanger V. Ruiz A. Soriano P. Recent optimization models and trends in location, relocation, and dispatching of emergency medical vehicles Eur J Oper Res 272 1 2019 1 23
78 Berman O. Kaplan E.H. Equity maximizing facility location schemes Transp Sci 24 2 1990 137 144
79 Talen E. Anselin L. Assessing spatial equity: an evaluation of measures of accessibility to public playgrounds Environ Plann A 30 4 1998 595 613
80 Hong X. Lejeune M.A. Noyan N. Stochastic network design for disaster preparedness IIE Trans 47 4 2015 329 357
81 Muggy L. Stamm J.L.H. Dynamic, robust models to quantify the impact of decentralization in post-disaster health care facility location decisions Oper Res Health Care 12 2017 43 59
82 Liu K. Zhang H. Zhang Z.-H. The efficiency, equity and effectiveness of location strategies in humanitarian logistics: a robust chance-constrained approach Transp Res Part E 156 2021 102521
83 Meng Q. Yang H. Benefit distribution and equity in road network design Transp Res Part B 36 1 2002 19 35
84 McCoy J.H. Lee H.L. Using fairness models to improve equity in health delivery fleet management Prod Oper Manag 23 6 2014 965 977
85 De Boeck K. Decouttere C. Jónasson J.O. Vandaele N. Vaccine supply chains in resource-limited settings: mitigating the impact of rainy season disruptions Eur J Oper Res 301 1 2022 300 317
86 United States Census Bureau. County population by characteristics: 2010–2019. 2021. https://www.census.gov/data/tables/time-series/demo/popest/2010s-counties-detail.html, Last accessed on 2022-03-27.
87 Office of the Assistant Secretary for Planning and Evaluation (ASPE). Vaccine hesitancy for COVID-19: State, county, and local estimates. 2021. https://aspe.hhs.gov/reports/vaccine-hesitancy-covid-19-state-county-local-estimates, Last accessed on 2022-03-27.
88 CDC. COVID-19 vaccinations in the United States, county. 2022c. https://data.cdc.gov/Vaccinations/COVID-19-Vaccinations-in-the-United-States-County/8xkx-amqh/data, Last accessed on 2022-03-27.
89 Bureau of Transportation Statistics. Commute mode. 2020. https://www.bts.gov/browse-statistical-products-and-data/state-transportation-statistics/commute-mode, Last accessed on 2022-03-27.
90 ESRI H. Esri streetmap premium 2021 ESRI Redlands, CA
91 ESRI. Public transit analysis in arcgis. 2021b. https://esri.github.io/public-transit-tools/.
92 Siegler A.J. Luisi N. Hall E.W. Bradley H. Sanchez T. Lopman B.A. Sullivan P.S. Trajectory of COVID-19 vaccine hesitancy over time and association of initial vaccine hesitancy with subsequent vaccination JAMA Netw Open 4 9 2021 e2126882 34559232
93 CDC. Underlying medical conditions associated with higher risk for severe COVID-19: Information for healthcare professionals. 2023. https://www.cdc.gov/coronavirus/2019-ncov/hcp/clinical-care/underlyingconditions.html, Last accessed on 2023-02-02.
94 Kaiser Family Foundation. How many adults are at risk of serious illness if infected with coronavirus? Updated data. 2020. https://www.kff.org/coronavirus-covid-19/issue-brief/how-many-adults-are-at-risk-of-serious-illness-if-infected-with-coronavirus/, Last accessed on 2023-02-02.
|
PMC010xxxxxx/PMC10200368.txt |
==== Front
Vaccine
Vaccine
Vaccine
0264-410X
1873-2518
Elsevier Ltd.
S0264-410X(23)00298-0
10.1016/j.vaccine.2023.03.026
Article
How to reduce vaccination hesitancy? The relevance of evidence and its communicator☆
Eger Jens a
Kaplan Lennart C. bcd
Sternberg Henrike e⁎
a DEval, German Institute for Development Evaluation, Bonn, Germany
b Georg-August University, Göttingen, Germany
c German Institute of Development and Sustainability, Bonn, Germany
d Kiel Institute for the World Economy, Kiel, Germany
e School of Social Sciences and Technology, Technical University of Munich, München, Germany
⁎ Corresponding author at: School of Sciences and Technology, Technical University of Munich, Richard-Wagner-Straße 1, 80333 München, Germany.
22 5 2023
19 6 2023
22 5 2023
41 27 39643975
14 11 2022
1 3 2023
9 3 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Even though the immediate urgency of the COVID-19 pandemic seems to have passed, many countries did not reach the vaccination rates they initially aimed for. The stagnation in vaccine uptake during the height of the pandemic presented policy makers with a challenge that remains unresolved and is paramount for future pandemics and other crises: How to convince the (often not insubstantial) unvaccinated proportion of the population of the benefits of a vaccination? Designing more successful communication strategies, both in retrospect and looking ahead, requires a differentiated understanding of the concerns of those that remain unvaccinated. Guided by the elaboration likelihood model, this paper has two objectives: First, it explores by means of a latent class analysis how unvaccinated individuals might be characterized by their attitudes towards COVID-19 vaccination. Second, we investigate to what extent (i) varying types of evidence (none/anecdotal/statistical) can be employed by (ii) different types of communicators (scientists/politicians) to improve vaccination intentions across these subgroups. To address these questions, we conducted an original online survey experiment among 2145 unvaccinated respondents from Germany where a substantial population share remains unvaccinated. The results suggest three different subgroups, which differ regarding their openness towards a COVID-19 vaccination: Vaccination opponents (N = 1184), sceptics (N = 572) and those in principle receptive (N = 389) to be vaccinated. On average, neither the provision of statistical nor anecdotal evidence increased the persuasiveness of information regarding the efficacy of a COVID-19 vaccine. However, scientists were, on average, more persuasive than politicians (relatively increase vaccination intentions by 0.184 standard deviations). With respect to heterogeneous treatment effects among the three subgroups, vaccination opponents seem largely unreachable, while sceptics value information by scientists, particularly if supported by anecdotal evidence (relatively increases intentions by 0.45 standard deviations). Receptives seem much more responsive to statistical evidence from politicians (relatively increases intentions by 0.38 standard deviations).
Keywords
Vaccination hesitancy
COVID-19
Elaboration likelihood model
Latent class analysis
Persuasive messaging
Evidence provision
==== Body
pmc1 Introduction
Despite sufficient vaccination capacity and evidence on vaccine efficacy [1], many countries around the world did not overcome the challenge of reaching vaccination rates sufficiently high to achieve community immunity during the peak of the COVID-19 pandemic. In autumn 2021 (during our data collection), Germany, among several other European countries, was hit strongly by the fourth wave of the pandemic and found itself struggling to find the right policies amidst the increased contagiousness of the Delta and Omicron variants. In response to the stagnating national vaccination rate, the government had introduced increasing restrictions on unvaccinated citizens. Nevertheless, at the time of writing (February 2023), the national vaccination rate remains at just above 75 % (i.e., those having received at least two doses [2]). To achieve community immunity, estimates of required vaccination rates range up to 90 %, underlining the importance of effective immunization campaigns [3]. Such a campaign, however, can only be created, if the needs and considerations of the unvaccinated, which have been dominated by concerns about vaccine safety and efficacy, are sufficiently understood [4], [5], [6]. Against this background, our paper engages in a differentiated classification of the unvaccinated population and builds on sociopsychological theory to assess how evidence regarding vaccine efficacy being presented through different communicators might increase vaccination willingness.
Existing research suggests that the judgment of information largely depends on the perception of trustworthiness and credibility of its communicator [e.g., 7]. Particularly for vaccinations, recent empirical evidence confirms the importance of a trustworthy communicator when it comes to the decision to get vaccinated [8], [9]. To increase their trustworthiness, communicators may support their information with evidence [10], [11]. An extensive literature on persuasion processes addresses the effects of information and information attributes on attitude change [see 12, for a review].
According to the dual-process theory of the Elaboration-Likelihood Model (ELM) [13], the processing of persuasive messages can occur through two distinct pathways. The first, known as the “peripheral route,” runs through quick cognitive shortcuts, while the second, called the “central route,” builds on more elaborate cognitive processing and deliberate reasoning [e.g., 14]. Thus, different types of evidence may align better with a certain processing path. Specifically, a distinction can be made between two forms of evidence: statistical and anecdotal. Previous research has shown that when subjects do not engage in deep elaboration, anecdotal evidence is more convincing [15], [16] since it is more descriptive and easier to process via the peripheral route [17]. In contrast, statistical evidence is likely to be particularly compelling, if the respective population considers the subject to be prominent and engages via the central route.
Given heterogeneities in the population and their attitudes towards different communicators, we propose that communicators might be able to optimize the effectiveness of their message by choosing the evidence type that best complements their own credibility and aligns with the perspective of the recipient of the information. We examine these theoretical expectations within a unique sample of unvaccinated German citizens (N = 2145) who participated in an online survey experiment specifically targeting this critical share of the population.
The contribution of this paper is threefold. First, we engage in a latent class analysis to characterize unvaccinated individuals based on their attitudes towards the COVID-19 vaccination. In doing so, we conceptually rely on the 5C-scale of vaccination hesitancy [18]. Second, to determine the most effective strategies for increasing vaccination uptake, we combine our classification system with the theoretical framework of the ELM. Specifically, we experimentally examine how different types of evidence (none, statistical, anecdotal) and communicators (politicians, scientists) can be targeted to the respective subgroups that we have identified. We argue that treatment effects might vary significantly across subgroups if either the motivation or the ability in information processing also varies between them. To the best of our knowledge, this is the first study that considers the potential complementarity between communicator and evidence to increase messaging effectiveness and assesses treatment effect heterogeneity across different subgroups of unvaccinated respondents.1
The remainder of the paper is structured as follows: Section 2 briefly presents the utilized materials and methods, introduces the empirical strategy, and outlines the experimental approach. Subsequently, Section 3 presents the results of the latent class analysis and our experimental findings. Section 4 concludes with a brief discussion of the main findings and outlines avenues for future research.
2 Materials and methods
2.1 Setting and sampling
We conducted an online survey with a sample of 2,1452 unvaccinated individuals from Germany between August 20 and September 16, 2021. Respondents were recruited from a German online access panel maintained by the survey company Bilendi & respondi, through which participants received an URL via email and could choose to participate in the survey using their own digital device. Individuals were eligible to participate in the study if they were at least 18 years old and had not yet been vaccinated against COVID-19 (not yet received the first dose). The survey covered socioeconomic characteristics, measures regarding respondents’ elaboration likelihood to engage with information about COVID-19 vaccines, and their intentions to get vaccinated. Respondents received ‘mingle points’ (worth roughly 1 Euro) for participating in the survey, which they could redeem as cash, vouchers, or donations.3
2.2 Empirical strategy
2.2.1 Latent class analysis
According to the ELM, behavioural intentions are affected by one’s motivation and ability, e.g., education [19], to engage with available information. Thus, to assess differences in the responsiveness to communicators and the effectiveness of evidence on COVID-19 vaccination intentions, we first classified and characterized different groups of unvaccinated respondents based on their attitudinal patterns towards COVID-19 vaccinations. For this purpose, we conducted a latent class analysis (LCA). This analysis was an explorative component of the information experiment as outlined in the Pre-Analysis-Plan.4
The purpose of LCA is to condense numerous observed ordinal variables (reflective indicators) in order to assign probabilities of belonging to a smaller set of underlying, latent classes [20]. By reducing dimensionality, LCA facilitates subgroup analyses [21] which makes it particularly useful to assess heterogeneous effects in experiments [22]. As reflective indicators, we considered the extended COVID-19-adapted 5C scale by Betsch et al. [18]. Using three survey questions for each dimension, the scale captures five different aspects of vaccination intentions (namely: Confidence, Complacency, Collective responsibility, Constraints, and Calculation; see Section 2.3 and Table A15 for details), intended to proxy the motivation component of the elaboration likelihood of respondents. The LCA method assumes that the underlying latent class membership (i.e., here classes of vaccination hesitancy) induces differential response patterns in the reflective indicators (i.e., here the 5C scale questions). Thus, in the LCA, the 5C scale indicators are the dependent variables and the categorical latent class variable is the independent variable, as illustrated in Fig. 1 below. Based on this, we estimated a generalized structural equation model by means of an ordered logistic regression using maximum likelihood estimation.Fig. 1 LCA model: Identification of classes of vaccination hesitancy.
2.2.2 Survey experiment
The second component of the empirical strategy comprised a survey experiment testing different information treatments about the benefits of a COVID-19 vaccination to identify persuasive communication strategies. Specifically, we first informed all participants about the current COVID-19 incidence in Germany (at the time of the survey) and described a hypothetical scenario in which a new COVID-19 vaccine had been developed and approved. In a second step, respondents were informed that this newly developed and approved vaccine is highly effective in reducing hospitalization following a COVID-19 infection. This second step was randomly varied regarding two components: (i) the profession of the communicator of this information and (ii) the type of evidence employed by the communicator. In our case, anecdotal evidence referred to a visit to an intensive care unit, while statistical evidence made reference to a clinical study on the efficacy of the new hypothetical COVID-19 vaccine. Appendix A provides the exact wording of the experimental treatments.
We estimated ordinary least squares (OLS) and ordered logit models to test the main hypotheses whether evidence increases vaccination willingness and whether heterogeneous messaging and respondent characteristics moderate this effect.(1) VIi=α1+β1Evii×Comi+β2Comi+β3Evii+β4X′+∊i
VIi in Equation 1 refers to the vaccination intentions of respondent i, Comi, denotes whether a politician or scientist communicated the information treatment, Evii, refers to the type of provided evidence (none/anecdotal/statistical), X' refers to a vector of socioeconomic control variables, and ∊i is the error term.5
We also consider how the different subgroups of unvaccinated respondents (i.e., the identified classes of vaccination hesitancy) react to the differential messaging. For this purpose, we introduce interaction terms with LCAi to Eq. (1):(2) VIi=α1+β1Evii×Comi+β2Comi+β3Evii+β4X′+β5Evii×LCAi+β6Comi×LCAi+β7Evii×Comi×LCAi+β8LCAi+∊i
We later employ t-tests to assess if treatment effects across senders and evidence types differ by our assigned classes (LCAi). Since there is uncertainty in the class assignment when using LCA, we employed a multiple imputation approach to address this (see Appendix A for further details in this regard).
2.3 Data and outcome variables
Beyond the information about treatment group assignment, the survey contained questions about the main dependent variable of interest: the intention to get vaccinated against COVID. We moreover collected information on additional explanatory variables, including demographic and socioeconomic characteristics of respondents. Question wording and summary statistics for the primary and secondary outcome variables as well as for selected control variables are provided in Table 1 below. A full list of all survey items is shown in Tables A1 and A15 in Appendix B.Table 1 Survey items, Question wording and Summary statistics.
Variable Item/Question Scale Summary Stat.
Primary outcome variable:
Vaccination willingness How would you decide if you had the opportunity to get vaccinated against COVID-19 with this vaccine next week? Scale from (1) I would definitely not get vaccinated to (7) I would definitely get vaccinated Mean: 2.69
SD: 1.62
Secondary outcome variables (mediation analysis): Please think about the new vaccine that was the subject of some of the previous questions. Now we would like to know how you assess the information provided by Ms. Sommer on this new vaccine. Please indicate to what extent you agree or disagree with the following statements: Scale from (1) Strongly disagree to (7) Strongly agree
Credibility The information is credible. Mean: 3.11
SD: 1.62
Relevance The information is relevant for my decision to get vaccinated. Mean: 2.80
SD: 1.79
Credible Source Ms. Sommer is credible as the source of information. Mean: 2.89
SD: 1.59
Controls
Age In which year were you born? (transferred to age) Mean: 41.41
SD: 11.82
Gender What gender do you identify with? (1) Female (1) 63.8 %
(2) Male (2) 36.0 %
(3) Diverse (3) 0.1 %
(4) Prefer not to say (4) 0 %
Education What is your highest level of education (educational qualification)? (1) No school-leaving qualification (1) 0.5 %
(2) Elementary or secondary school leaving certificate without completed apprenticeship (2) 3.0 %
(3) Elementary or secondary school leaving certificate with completed apprenticeship (3) 10.9 %
(4) Secondary school leaving certificate, Realschulabschluss (4) 35.6 %
(5) Advanced technical college certificate (5) 8.5 %
(6)Abitur(general higher education entrance qualification) (6) 17.4 %
(7) University of applied sciences or university degree (Bachelor, Master, Magister, Diplom or Staatsexamen) (7) 22.1 %
(8) Doctorate/PhD (8) 1.3 %
(9) Other degree: (9) 0.5 %
5C Vaccination attitudes Scale from (1) Strongly
disagree to (7) Strongly agree
The COVID-19 vaccinations are effective. (conf 1) Mean: 3.09
SD: 1.78
I have full confidence in the safety of COVID-19 vaccinations. (conf 2) Mean: 2.29
SD: 1.74
As far as COVID-19 vaccinations are concerned, I trust that government authorities will always decide in the best interest of the general public. (conf 3) Mean: 2.56
SD: 1.80
My immune system is so strong, it also protects me from contracting COVID-19. (comp 1) Mean: 4.27
SD: 1.88
Vaccination against COVID-19 is superfluous, since diseases against which one can be vaccinated are generally rare. (comp 2) Mean: 3.23
SD: 1.82
COVID-19 is not so bad that I need to be vaccinated against it. (comp 3) Mean: 4.17
SD: 2.03
It is costly for me to get vaccinated against COVID-19. (const 1) Mean: 2.50
SD: 1.81
My discomfort at doctor’s appointments keeps me from getting vaccinated against COVID-19. (const 2) Mean: 2.47
SD: 1.84
Everyday stress keeps me from getting vaccinated against COVID-19. (cons 3) Mean: 2.18
SD: 1.71
I think very carefully about whether it makes sense for me to be vaccinated against COVID-19. (calc 1) Mean: 5.98
SD: 1.56
A full understanding of the issue of COVID-19 vaccination is important to me before I get vaccinated. (calc 2) Mean: 5.67
SD: 1.58
When I think about getting vaccinated against COVID-19, I weigh the benefits and risks to make the best possible decision. (calc 3) Mean: 5.83
SD: 1.57
If everyone is vaccinated against COVID-19, I don’t need to get vaccinated too. (core 1, reverse coded) Mean: 3.68
SD: 2.04
I get vaccinated against COVID-19 because I can protect people with a weak immune system. (core 2) Mean: 2.84
SD: 2.00
Vaccination is a community measure to prevent the spread of COVID-19. (core 3) Mean: 3.38
SD: 2.04
2.3.1 Vaccination intention
The main outcome variable is respondents’ intention to get vaccinated against COVID-19, measured on a 7-point Likert scale. We elicited vaccination intentions by asking respondents the following question: “How would you decide, if you had the opportunity to get vaccinated against COVID-19 with this vaccine next week?”. Responses were measured on a scale from (1) “I would definitely not get vaccinated” to (7) “I would definitely get vaccinated.”. 6, 7
A broad psychological literature - largely based on the Theory of Planned Behavior [23] provides evidence that intentions are a relevant predictor of actual behaviour, for instance with respect to other health behaviours during the COVID-19 pandemic [24]. Nonetheless, we are aware that reported intended and actual vaccination decisions might differ [25], [26], [27]. Hence, we try to address this concern as follows: First, by sampling respondents with real-world hesitancy, we select the relevant target group for messaging campaigns (i.e., the approx. 25 % of the German population that indeed had not been vaccinated in September 2021 and is largely still unvaccinated today). Second, we include in our tested messaging recent information on the actual COVID-19 infection numbers at the time of the data collection. Third, we used messaging as employed by politicians/scientists [e.g., 28]. We deliberately did not refer to any of the existing COVID-19 vaccines in the experiment to avoid potential difficulties resulting from vaccine-specific (negative or positive) associations (e.g., with AstraZeneca), which would most likely systematically bias reactions to the employed messages in unintended ways. The indicated efficacy of the hypothetical new vaccine, however, matches the efficacy of the preferred vaccines which were used in Germany at that time (BioNTech and Moderna). Finally, the approval of a new vaccine was not an unrealistic prospect at the time of the survey, since new vaccines were still under development (which is even currently still the case, e.g., for variant-specific vaccines). We recognize that it would be valuable to conduct future research to measure the impact of messaging campaigns on actual vaccination decisions.
2.3.2 COVID-19 vaccination attitudes and elaboration likelihood
As outlined above, we employed the 5C scale by Betsch et al. [18] as reflective indicators for the LCA. We adjusted this scale, initially designed for general vaccination attitudes, to fit a COVID-19 specific application based on Betsch et al. [9]. The scale aims to elicit five central aspects of attitudes towards vaccination, namely (i) Confidence in the COVID-19 vaccines and their endorsers (this captures both safety and effectiveness concerns of the vaccines as well as confidence in the entities producing, administering, and encouraging vaccinations), (ii) Complacency (not perceiving the virus as a serious risk), (iii) Collective responsibility (willingness to protect others), (iv) Constraints (structural and psychological barriers), and (v) Calculation (extensive information searching for weighing costs and benefits) [18]. We employed the full extended scale, which consists of three items for each aspect (see Table A15 in Appendix B for a list of the 15 questions and the exact wording). Moreover, the survey collected information to validate whether the LCA class assignment, in fact, captures the elaboration likelihood with which respondents take up the different information treatments. Specifically, we elicited trust in politicians and scientists (a proxy for motivation to engage with the provided information) and the perceived value and understanding of anecdotal and statistical evidence (a proxy for ability to engage with the provided information), all measured on a 7-point Likert scale. Besides building on the previously validated survey items in the respective cited works above, we piloted (soft-launched) our survey with a sample of 110 respondents to validate our outcomes and covariates.
3 Results
3.1 Latent class analysis: Not one, but many publics
The results of the LCA suggest the existence of three underlying classes of vaccination hesitancy in our sample.8 The classes are characterized and distinguishable by differential response patterns to the COVID-19 adjusted 5C scale (see Fig. 2 below and Tables A2-A5 in Appendix B).9 Given these differing response patterns to the 5C scale, we expect members of the three classes to have coherently varying attitudes towards getting vaccinated against COVID-19 and have therefore labelled the three classes as opponents, sceptics and receptives. The following paragraphs explain these differences by describing response patterns to the 5C scale (see Fig. 2) and stated intentions to get vaccinated against COVID-19 across classes (employed as an accuracy check of these labels, see Table 2 ). We further contrast socioeconomic characteristics across classes.Fig. 2 Classes of vaccination hesitancy: Identification via the 5C scale. Notes: Results refer to mean values of the 5C scale for vaccination attitudes, separately for each class as identified by the LCA. The mean values shown here are those of the initial 5C scale, measured on a 7-point Likert scale. See Tables A3-A5 in Appendix B for the condensed scale. Values of the 15 5C items (three items for each aspect) were averaged for each of the five aspects of vaccination hesitancy that we aim to capture. See Figure A1 in Appendix B for the same graphic with all 15 items. In order to better illustrate class differences regarding the 5C vaccination hesitancy scale, the graph employs a definite class assignment where respondents were assigned to the class with the maximum predicted probability.
Table 2 Respondent characteristics by class membership.
Variable (1)
Opponents (2)
Sceptics (3)
Receptives T-test Difference
(N=1,184) (N=572) (N=389) (1)-(2) (1)-(3) (2)-(3)
Willingness to get vaccinated 1.732 3.428 4.495 −1.696*** −2.763*** −1.068***
(0.036) (0.061) (0.102)
Female 0.658 0.592 0.652 0.066*** 0.006 −0.060*
(0.014) (0.021) (0.024)
18-34 yrs. 0.266 0.404 0.365 −0.138*** −0.099*** 0.039
(0.013) (0.021) (0.024)
35-54 yrs. 0.527 0.484 0.460 0.043* 0.067** 0.024
(0.015) (0.021) (0.025)
55 yrs. and above 0.207 0.112 0.175 0.095*** 0.032 −0.063***
(0.012) (0.013) (0.019)
Primary education 0.151 0.142 0.132 0.009 0.019 0.010
(0.010) (0.015) (0.017)
Secondary education 0.375 0.347 0.323 0.027 0.051* 0.024
(0.014) (0.020) (0.024)
Tertiary education 0.474 0.510 0.545 −0.036 −0.071** −0.034
(0.014) (0.021) (0.025)
Residence in new federal states 0.272 0.243 0.198 0.029 0.074*** 0.045
(0.013) (0.018) (0.020)
Intention to vote AFD 0.352 0.219 0.091 0.133*** 0.261*** 0.128***
(0.013) (0.015) (0.012)
Trust in scientists 3.716 4.229 5.532 −0.513*** −1.816*** −1.303***
(0.046) (0.057) (0.064)
Trust in politicians 1.613 2.962 3.653 −1.348*** −2.040*** −0.691***
(0.029) (0.061) (0.085)
Value/Understanding of anecdotal evidence 3.708 3.458 3.126 0.250*** 0.582*** 0.332***
(0.028) (0.036) (0.055)
Value/Understanding of statistical evidence 3.260 3.503 3.972 −0.244*** −0.712*** −0.468***
(0.033) (0.035) (0.047)
F-test of joint significance (F-stat) 52.297*** 97.001*** 25.024***
F-test, number of observations 1,756 1,573 961
Notes: Summary statistics contain means and standard deviations for continuous variables and percentages for categorical variables. All summary statistics are calculated based on the full sample.
Notes: Variables are binary indicators, except for Willingness to get vaccinated, Trust in scientists, Trust in politicians and Value/Understanding of anecdotal/statistical evidence, which were measured on a 1-7 point Likert scale, with higher values indicating higher willingness to get vaccinated/trust/self-rated ability, respectively. Willingness to get vaccinated refers to the willingness to get vaccinated with the hypothetical new COVID-19 vaccine introduced in the survey experiment (i.e., it is the outcome variable of interest in the empirical analysis of the survey experiment for evaluating treatments effects of evidence and communicator). See Table A9 in the Appendix for class characteristics regarding the Willingness to get vaccinated with existing vaccines. Class assignment is definite and defined according to each respondents’ highest predicted class probability. The resulting class assignment is captured in a categorical variable with three categories. ***, **, and * indicate significance at the 1, 5, and 10 percent critical level.
The first class, vaccination opponents, had the largest prevalence in our sample (55.14 %). This class was characterized by relatively low confidence levels in vaccine safety and efficacy, the system that delivers them, and the (motivations of the) actors deciding on the need for vaccines (Mean = 1.55). Similarly, vaccination opponents expressed high levels of complacency (perceived invulnerability towards the COVID-19 virus) (Mean = 5.52) and low levels of collective responsibility (Mean = 2.52), which may make them more susceptible to ‘vaccination free-riding’. Finally, the aspect of calculation (i.e., in terms of careful searching for information to weigh infection and vaccination risks and benefits) was a highly important factor (Mean = 6.19), whereas practical constraints or inconveniences such as geographical accessibility did, on average, not seem to present a substantial barrier to vaccination in this identified subgroup (Mean = 1.88). In light of these characteristics, the class of opponents seems to have opposing attitudes towards a COVID-19 vaccination.
The second class with contrary characteristics, vaccination receptives, had the smallest prevalence in our sample (18.24 %). While levels of (i) vaccine confidence and (ii) collective responsibility in getting vaccinated were on average relatively high (Confidence: Mean = 4.80; Collective responsibility: Mean = 5.83), the degree of (iii) complacency was relatively low compared to the values in the class of vaccination opponents (Mean = 2.00). Practical constraints to getting vaccinated were slightly more important among vaccination receptives (Mean = 2.41), while calculation aspects were slightly less important than they were among vaccination opponents (Mean = 5.66). In light of these characteristics, the class of receptives, in principle, seems to support getting a COVID-19 vaccination, but their careful weighing of the risks of a newly developed vaccine compared to the consequences of infection may keep them from doing so.
The third class, vaccination sceptics, had a prevalence of approximately one-quarter of our sample (26.63 %). In this class, respondents attributed a similar, moderate relevance to all the investigated reasons against or in favor of a COVID-19 vaccination (Confidence Mean = 3.44; Collective responsibility: Mean = 4.00; Complacency: Mean = 3.87; Constraints: Mean = 3.42). The only exception was calculative considerations of the vaccination decision, which presented the most dominant factor in this class (Mean = 5.20).10 In light of these characteristics, the class of sceptics seems to have serious doubts towards a COVID-19 vaccination but does not oppose it as strongly as the class of opponents.
In line with the above characterization, we find that the three classes differed linearly in their intentions to get vaccinated with the hypothetical COVID-19 vaccine introduced in the survey experiment (see Table 2 below). Reported willingness was on average lowest among opponents (Mean (SD) = 1.73(1.23)) and highest among the class of vaccination receptives (Mean (SD) = 4.50 (2.00)). Sceptics, coherently, indicated moderate intentions to get vaccinated (Mean (SD) = 3.43(1.46)).11 Reassuringly, this pattern also holds for respondents’ willingness to get vaccinated with the existing developed and widely known COVID-19 vaccines, underlining the credibility of our outcome variable with respect to having chosen a hypothetical vaccine (see Table A9 in Appendix B).
Regarding demographic and socioeconomic characteristics, the class of opponents differed from the other two classes, e.g., with respect to age, gender, educational attainment, and federal state of residence (see Table 2). Specifically, opponents were, on average, (i) older than receptives and sceptics, (ii) more likely to be female compared to sceptics, and (iii) less educated and were more likely to reside in Germany’s new Eastern federal states than members of the receptives class. However, the classes differed much more clearly in terms of their intentions and beliefs than in terms of socioeconomic characteristics: Looking at respondents’ (i) reported voting intentions in the next national election,12 (ii) their trust in scientists and (iii) politicians, as well as their perceived value and understanding of (iv) anecdotal and (v) statistical evidence, there seems to be a near-linear pattern across the three classes (see Table 2): The more supportive a class’ attitude is towards a COVID-vaccination (i.e., receptives > sceptics > opponents), the less often respondents of this class indicated their intention to vote for the AFD party in the next national election (a right-wing party in the German parliament, which opposed pandemic-related restrictions), the higher was their trust in scientists and politicians, the more they valued statistical evidence, and the less they valued anecdotal evidence.
Our findings partly resemble those by Rieger et al. (2022), who surveyed a representative sample of both vaccinated and unvaccinated German respondents and also find that women, right-wing voters and respondents with less trust in the political system were more likely to oppose vaccination. Yet, our results contrast Rieger et al. (2022) in that older and less educated respondents and respondents living in the new federal states are also more likely to oppose vaccination [29].13
3.2 Survey experiment: Average effects of communicator and evidence type
We briefly report the average treatment effects of the tested communication strategies in the entire sample and then examine a potential heterogeneity in their effectiveness in terms of communicator and evidence type across the identified classes. In all of the results reported here, the dependent variable is respondents’ willingness to get vaccinated with the new hypothetical COVID-19 vaccine.14, 15
Table 3 below presents the estimation results for average treatment effects: First, the coefficient of scientists as communicators is statistically significant at the 5 % level and positive in both estimations (Columns (1) and (2)), suggesting that scientists were on average more persuasive as communicators than politicians. Employing scientists as the communicator relatively increased the reported willingness to get vaccinated, on average, by approximately 0.184 standard deviations.Table 3 Treatment effects of evidence type and communicator on vaccination intentions.
(1) (2) (3) (4) (5) (6)
Politician Reference category
Scientist 0.184∗∗ 0.185∗∗
(2.32) (2.34)
No Evidence Reference category
Anecdotal Evidence 0.137 0.153
(1.41) (1.57)
Statistical Evidence 0.0354 0.0477
(0.36) (0.49)
No Evidence Reference category
x Politicians
Anecdotal Evidence 0.204 0.209
x Politicians (1.49) (1.53)
Statistical Evidence 0.143 0.159
x Politicians (1.04) (1.16)
No Evidence 0.300∗∗ 0.297∗∗
x Scientists (2.18) (2.16)
Anecdotal Evidence 0.372∗∗∗ 0.395∗∗∗
x Scientists (2.70) (2.88)
Statistical Evidence 0.228∗ 0.234∗
x Scientists (1.66) (1.70)
Ex post t-tests (t-values):
No Evidence Scientists vs. Anecdotal Evidence Scientists 0.28 0.50
No Evidence Scientists vs. Statistical Evidence Scientists 0.27 0.21
Anecdotal Evidence Scientists vs. Statistical Evidence Scientists 1.09 1.37
Socioeconomic Controls No Yes No Yes No Yes
Observations 2142 2141 2142 2141 2142 2141
Notes: The table shows standardized regression coefficients. Estimations in Columns (2), (4), and (6) include controls for age, gender, level of education, state of residency, and level of income. t statistics in parentheses.
∗, ∗∗, ∗∗∗ indicate significance at the 1, 5, and 10 percent critical level
Second, the statistically insignificant coefficients of both evidence types in Columns (3) and (4) suggest that neither the provision of statistical nor anecdotal evidence, on average, increased the persuasiveness of information about the efficacy of a COVID-19 vaccine.
Third, Columns (5) and (6) and the t-tests at the bottom of the table report the results for the interaction between types of evidence and each communicator. The t-tests on the point estimates reveal that there are no significant differences between no evidence (reference category), anecdotal evidence, and statistical evidence – neither for politicians nor for scientists as communicators. Therefore, when informing the average unvaccinated public about the efficacy of COVID-19 vaccines, it does not seem to matter what form of evidence communicators use, and whether they use any at all.16
3.3 Survey experiment: Heterogeneity by classes of vaccination hesitancy
We now turn to possibly differing effects of the explored treatments due to respondents’ differential elaboration likelihood, as reflected in the identified classes of vaccination opponents, sceptics, and receptives. The results of this exercise are presented in Fig. 3, Fig. 4 below, and we discuss them in turn for each class separately.Fig. 3 Heterogeneity by Class: Separate Treatment Effects of Evidence Type and Communicator on Vaccination Intentions. Notes: The figure shows the mean estimation coefficient and 95 % confidence intervals of 1,000 simulations for each of the three classes of vaccination hesitancy. Class assignment for each respondent is based on the class membership probability, which is derived from the LCA. The left column shows the treatment effects of the communicator, with the reference category ”Politician”. The right column shows treatment effects for evidence type with the reference category ”no evidence”. Estimations include controls for age, gender, education level, state of residency, and income level. Detailed estimation results are available in Table A11, A13, and A14 in Appendix B.
Fig. 4 Heterogeneity by Class: Interacted Treatment Effects of Communicator-Evidence Combinations on Vaccination Intentions. Notes: The figure shows mean treatment effects and 95 % confidence intervals of 1,000 simulations for each of the three classes of vaccination hesitancy. The reference category is “No Evidence Politicians”. Class assignment for each respondent is based on the class membership probability, which is derived from the LCA. Estimations include controls for age, gender, education level, state of residency, and income level. Detailed estimation results are available in Table A11, A13, and A14 in Appendix B.
3.3.1 Opponents
For vaccination opponents, the results in Fig. 3 show that neither the communicator nor the provision and type of evidence seem to be crucial for the persuasiveness of the information about COVID-19 vaccine efficacy. In accordance with this, the results in Fig. 4 suggest no clear communication strategy for politicians and scientists to target vaccination opponents: for both communicators, there are no significant differences between the evidence types.
3.3.2 Sceptics
For sceptics, the results are more unequivocal. First, concerning the communicator, column 2 row 2 of Fig. 3 shows that the coefficient for scientists is 0.2 standard deviations higher than for politicians. This difference is statistically significant at the 5 % level (average p-value17 : 0.023). This comparatively strong effect suggests that it may be the subgroup of sceptics that drives the effect found for the average population.
Second, we observe that, for sceptics, the provision of evidence about the efficacy of the COVID-19 vaccine does seem to matter. Compared to the no evidence condition, both anecdotal evidence and statistical evidence have a positive and statistically significant effect at the 1 % level (average p-value: 0.006 for anecdotal evidence and 0.019 for statistical evidence). However, sceptics seem not to differentiate between the evidence type since we found no statistically significant difference between anecdotal and statistical evidence.
Third, in terms of the explored interaction between communicator and evidence type, the results in Fig. 4 below reveal that scientists in particular can enhance the persuasiveness of their conveyed information by providing additional evidence (average p-value anecdotal evidence: 0.012; average p-value statistical evidence: 0.043). While the provision of anecdotal evidence seems to be the most promising in this regard, the effects are not significantly different from statistical evidence (average p-value: 0.574).
3.3.3 Receptives
For the group with the highest willingness to get vaccinated, the results reveal no statistically significant differences between communicators and the evidence types. Interestingly, however, the results in Fig. 4 suggest that politicians providing statistical evidence offer the most promising communication strategy (average p-value: 0.010; reference group are politicians without any evidence). This is surprising since, on average, neither politicians nor statistical evidence taken separately had a statistically significant effect on respondents’ reported vaccination intentions.
In sum, the above findings reveal notably different patterns across the three identified classes of vaccination hesitancy and point towards potentially promising communication strategies, especially for sceptics and receptives. These insights had not been visible by just exploring average effects. Moreover, the statistically significant subgroup effects are substantially larger in magnitude, ranging from 0.38 (politicians addressing receptives based on statistical evidence) to 0.45 standard deviations (scientists addressing sceptics based on anecdotal evidence). This evidence suggests that the different groups of vaccination hesitant respondents indeed express different elaboration likelihoods depending on the communication strategies explored here.18
4 Discussion and conclusion
Despite soaring case numbers and a sufficient supply of vaccine doses at their disposal, policy makers in several countries struggled with convincing their unvaccinated population of the benefits of a COVID-19 vaccination. Previous immunization campaigns have included various strategies to increase vaccination rates, ranging from mandatory vaccination, structured appointment planning and reminders, to carefully designed information campaigns [31], [32]. Evidence on the effectiveness of the different immunization campaigns is, therefore, urgently needed to inform political decision making. Given that mandatory vaccinations were highly contested politically [29], we investigated how to persuasively communicate information about the efficacy of a COVID-19 vaccination to the unvaccinated parts of the population.
Previous research suggests that the effectiveness of information campaigns depends crucially on the communicator (e.g., the media, politicians, religious leaders), the content of the message, and the characteristics of the targeted population (e.g., scepticism, social attitudes) [8], [33], [34]. Our latent class analysis points towards the existence of three distinct subgroups within the sample of unvaccinated respondents, who can be differentiated in terms of their views and motivations regarding a COVID-19 vaccination: opponents, sceptics, receptives.
While vaccination opponents seem rather difficult to target, our findings suggest potentially fruitful combinations of communicators and evidence types for the other two subgroups. Specifically, we found that anecdotal evidence provided by scientists is a promising communication strategy to encourage vaccination intentions in the subgroup of sceptics.
Similarly, statistical evidence presented by politicians fortified intentions within receptives. For the case of sceptics, one potential explanation for the somewhat unexpected result could be that scientists are already perceived as highly credible by this subgroup. While anecdotal evidence may decrease the perceived distance towards the public, the additional credibility gains from statistical evidence are limited. Thus, providing communicator-evidence combinations that are less present in the public debate could prove particularly effective.
These insights suggest that, in the short term, receptives and sceptics are the most promising target groups for German vaccination campaigns. Yet, in the medium term, opponents need not be forgotten. While mandatory vaccinations [35], [36] may appear as the only strategy to target strict vaccination opponents, politicians and researchers are advised to focus on ways how to rebuild trust and address beliefs in misinformation within this population group, not only in Germany [37], [38], [39], [40]. The inconsistency in vaccine related communication in Germany has led to a loss of trust in political and scientific decision-makers [41]. It is therefore important to rebuild this trust through evidence-based communication. The way we understand and perceive the credibility of a source significantly impacts our processing of messages and can also significantly affect related behaviours [42], [43]. Using evidence to validate relevant and reliable information can therefore also be vital to build trust and credibility in the vaccines themselves and their safety.
We acknowledge the following limitations of our study: First, the choice to employ an outcome variable that measures intentions to get vaccinated with a hypothetical vaccine is likely both a limitation and strength of our study. On the one hand, the hypothetical vaccine may be more vague and less informative to respondents. But, on the other hand, using existing vaccines like Astra Zeneca would have come with serious repercussions in terms of pre-defined opinions about them and would have likely biased the results of the messages tested in the survey experiment. The same logic applies to the hypothetical communicators of the messages being “a scientist” or “a politician”, deliberately not mentioning specific individuals already well-known to participants in the German pandemic context. Second, we acknowledge that our outcome variable merely measures intentions to get vaccinated as opposed to actual vaccination decisions (though evidence with respect to the Theory of Planned Behavior suggests that stated intentions and observed behaviour are closely related [23], [24]). Third, researchers and policy makers should, naturally, interpret our findings with caution in terms of external validity and the potential discrepancies between the measured effects of employing a message once in an online survey compared to the effects of employing messages repeatedly in real-world situations. External validity concerns also include the sample being recruited from an incentivized online panel population. This population may differ somewhat from the representative German population in respects other than the included quotas, though methodological research on this issue has so far not convincingly confirmed these concerns [44], [45], [46].
In terms of avenues for future research, further studies may want to examine the perceived relevance of the provided information more closely - a factor that might increase the persuasiveness of information campaigns. Specifically, in this paper, we examined an information treatment about COVID-19 hospitalization risk, but information and evidence on the infection probability or the risk of long-COVID might have a higher relevance [47].
In sum, our study employed sociopsychological theory to challenge the view of the existence of a single homogeneous group of unvaccinated citizens. By drawing on a large sample of unvaccinated citizens and combining latent class analysis with experimental methods, we encourage decision-makers to carefully consider heterogeneities in the effectiveness of their communication strategies, especially regarding their communicator and employed evidence type.
Funding
We gratefully acknowledge funding from the Dr. Hans-Riegel Foundation and the Chair of Development Economics (Andreas Fuchs) at Göttingen University. The funding sources had no involvement in the study design, collection, analysis and interpretation of data, in the writing of the report, and in the decision to submit the article for publication.
Ethical approval
The ethical review board of Göttingen University reviewed the study prior to implementation (Ethikkommission, date 15/07/2021).
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.
Appendix A Supplementary data
The following are the Supplementary data to this article:Supplementary data 1
Data availability
Data will be made available on request.
☆ The authors wish to thank Renate Hartwig and Jan Priebe for comprehensive comments on the experimental design. Moreover, the manuscript benefited from comments by Philip Ernst, Andreas Fuchs, Annabell Kaplan, Janina Steinert, Christian Thiele, Participants of the Göttingen Development Group Seminar and the Dr. Hans-Riegel COVID-19 Conference. Erin Flannagan and Atanas V. Spasov assisted with diligent proofreading.
1 For an overview of our proposed hypotheses, please see https://osf.io/vhjeg/. For specifications of pre-analysis plan deviations, see Appendix A.
2 Please note that the actual number of observations might vary slightly in the results tables below due to missing values in socio-economic control variables. The number of missings is, however, very low (see below).
3 We acknowledge that monetary incentives may have severe effects on the study results, if the financial incentives do influence the participation and/or response behaviour. The panel provider, Bilendi & respondi, places great importance on controlling the quality of the panellists through regular checks in order to assure high quality participants. This starts with the registration process and is continued during the lifetime of a panellist as well as in every survey participation. When registering for the panel, an automatic verification checks duplicate registrations with the same email address. During the registration process, the digital fingerprint highlights participants with identical IP addresses and browser configurations and automatically renders these respondents inactive. Furthermore, a double opt-in email with an invitation to complete the start questionnaire is sent to every new member straight after registration. Within the framework of Bilendi & respondi’s quality strategy, the company does not use routing or river sampling in their controlling measures. The panellists are invited directly by email to Bilendi & respondi’s client`s questionnaire without being diverted to a router in between. This applies also to Bilendi & respondi’s partners if the company needs additional purchase. Furthermore, Bilendi & respondi always clearly communicates the origin and proportion of external panellists to their clients.
4 For the experimental protocol and the survey, please refer to the supplementary materials and the pre-analysis plan at https://osf.io/vhjeg/.
5 To account for multiple hypothesis testing, we apply also sharpened q-values based on Stata code by Anderson [22] (see Tables A11 and A13 in Appendix B).
6 For the exact wording of the other survey items, please see Table A15 in Appendix B.
7 At the time of the data collection (mid-August-mid-September 2021), COVID-19 vaccines were still scarce in Germany, but the initially high demand for vaccine shots had largely passed. Approx. 61 % of the population were already vaccinated twice and not yet eligible for the booster vaccination (Impfdashboard, Bundesministerium für Gesundheit, 2023). During this time, the centralized online system and practitioners were no longer overwhelmed and citizens could mostly schedule their appointments for a vaccine without complications. While it was rather unlikely to book an appointment within a timeframe as short as one week (though it did occasionally happen that appointments were cancelled and citizens were offered their appointment on short notice), we chose this wording of our outcome variable as to more adequately elicit participants’ vaccination intentions (i) independent of extreme scarcity considerations and (ii) to provide a more concrete time frame for respondents when making the decision. In doing so, we hoped to make the question more precise and viable for respondents and minimize potential biasing effects of its inherent hypothetical nature.
8 We ran models with one to four and partly five underlying classes, but opted for a final model with three classes due to a combination of interpretability, reductions in the goodness of fit improvements, and model convergence problems. In the LCA, where the model’s fit can also be assessed for a single model using the likelihood-ratio test of the fitted model versus the saturated model (G2 statistic), we fail to reject the null hypothesis that our model fits just as well as the saturated model.
9 While we initially measured agreement with the items of the 5C scale on a 7-point Likert scale, we condensed the scale into three categories due to skewed and non-normally distributed data, not allowing the model to converge. The three condensed categories combined the two extreme points of the initial Likert scale as well as the three moderate points (initial values 1 or 2 => condensed value 1; values 3, 4 or 5 => value 2; values 6 or 7 => value 3). The resulting ordinal variables with three categories were then used as categorical indicators for the LCA. The initial 7-point Likert scale and the condensed scale are, on average, across all 15 items highly correlated by a value of approximately 0.95.
10 Due to the condensed indicators employed in the LCA, we additionally conducted a latent profile analysis (LPA) as a robustness check, which employs the initial 7-point Likert scale as continuous indicators of the latent classes. Class prevalences and characteristics (i.e., marginal class probabilities and class means) were very similar to in the LCA and are presented in Tables A6-A8 in Appendix B.
11 While we argue for a differentiated assessment across the identified subgroups within the unvaccinated population, we also consider average effects within the entire sample of unvaccinated respondents (Mean: 2.69; SD: 1.84).
12 The next national elections in Germany were held a couple of weeks after the survey was fielded.
13 Notably, while we focused on the unvaccinated population, Rieger et al. (2022) focused on the general public i.e., vaccinated and unvaccinated individuals. Aside from the different samples, the different outcome variables in terms of vaccination intentions (individual outcome) in our study and attitudes towards mandatory vaccinations (collective outcome) in Rieger et al. (2022) lack comparability and thus prevent strong conclusions [r7].
14 All models are estimated using OLS regressions. To account for the ordered scale of the dependent variable, we report results of ordered logit estimations in Appendix (see Table A10).
15 After the respondents received the treatment and were asked about their willingness to get vaccinated, we conducted an attention test in which we asked the participants what profession –the communicating person in the treatment has. 54 % of the respondents answered this question correctly. Table A12 in Appendix B presents the regression results if only the respondents who answered the question correctly are considered.
16 In the appendix, we also present the main results, when accounting for multiple hypotheses testing via sharpened q-values [30].
17 We report average p-values as our estimations are based on 1,000 simulations of Equation 1 with a probabilistic assignment of respondents to the respective classes.
18 Appendix A and Figures A2-A4 contain additional analyses that analyse whether the reported effects are due to class differences in (i) perceptions of the credibility of the information, (ii) the relevance of the information for one’s reported intention to get vaccinated, and (iii) the credibility of the information source.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.vaccine.2023.03.026.
==== Refs
References
1 Kim D. Lee Y.J. Vaccination strategies and transmission of COVID-19: Evidence across advanced countries J Health Econ 82 2022 102589
2 Robert-Koch-Institut (RKI). Digitales Impfquotenmonitoring zur COVID-19-Impfung. [Online; accessed 13-February-2022]; 2022. url: https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Daten/Impfquoten-Tab.html.
3 Anderson R.M. Vegvari C. Truscott J. Collyer B.S. Challenges in creating Herd immunity to SARS-CoV-2 infection by mass vaccination Lancet 396 10263 2020 1614 1616 33159850
4 Karlsson L.C. Soveri A. Lewandowsky S. Karlsson L. Karlsson H. Nolvi S. Fearing the disease or the vaccine: The case of COVID-19 Pers Individ Differ 172 2021 110590
5 Paul E. Steptoe A. Fancourt D. Attitudes towards vaccines and intention to vaccinate against COVID-19: Implications for public health communications Lancet Regional Health-Europe 1 2021 100012
6 Courbage C. Peter R. On the effect of uncertainty on personal vaccination decisions Health Econ 30 11 2021 2937 2942 34346125
7 Hewgill MA, Miller GR. Source credibility and response to fear∙arousing communications; 1965.
8 Argote P. Barham E. Daly S.Z. Gerez J.E. Marshall J. Pocasangre O. The shot, the message, and the messenger: COVID-19 vaccine acceptance in Latin America NPJ Vaccines 6 1 2021 118 34593822
9 Betsch C, et al. German COVID-19 Snapshot Monitoring (COSMO)-Welle 36: Zweite Fokuserhebung zu Impfungen; 2020.
10 O'Keefe D.J. Justification explicitness and persuasive effect: A meta-analytic review of the effects of varying support articulation in persuasive messages Argumentation and Advocacy 35 2 1998 61 75
11 Reinard J.C. The persuasive effects of testimonial assertion evidence 1998 Advances through meta-analysis Persuasion 69 86
12 O'Keefe D.J. Persuasion: Theory and research 2015 Sage Publications
13 Chaiken S. Trope Y. Dual-process theories in social psychology 1999 Guilford Press
14 Strack F, Deutsch R. The duality of everyday life: dual-process and dual system models in social psychology. In APA handbook of personality and social psychology, Volume 1: Attitudes and social cognition. American Psychological Association; 2015. pp. 891-927.
15 Betsch C. Ulshöfer C. Renkewitz F. Betsch T. The influence of narrative v. statistical information on perceiving vaccination risks Med Decis Making 31 5 2011 742 753 21447730
16 Haase N. Betsch C. Renkewitz F. Source credibility and the biasing effect of narrative information on the perception of vaccination risks J Health Commun 20 8 2015 920 929 26065492
17 Hoeken H. Anecdotal, statistical, and causal evidence: Their perceived and actual persuasiveness Argumentation 15 2001 425 437
18 Betsch C. Schmid P. Heinemeier D. Korn L. Holtmann C. Böhm R. Beyond confidence: Development of a measure assessing the 5C psychological antecedents of vaccination PLoS ONE 13 12 2018 e0208601 30532274
19 Maurer J. Who has a clue to preventing the flu? Unravelling supply and demand effects on the take-up of influenza vaccinations J Health Econ 28 3 2009 704 717 19278744
20 Hagenaars J.A. McCutcheon A.L. Applied latent class analysis 2002 Cambridge University Press
21 Lanza S.T. Rhoades B.L. Latent class analysis: an alternative perspective on subgroup analysis in prevention and treatment Prev Sci 14 2 2013 157 168 21318625
22 Lancsar E. Ride J. Black N. Burgess L. Peeters A. Social acceptability of standard and behavioral economic inspired policies designed to reduce and prevent obesity Health Econ 31 1 2022 197 214 34716628
23 Ajzen I. From intentions to actions: A theory of planned behavior 1985 Springer Berlin Heidelberg 11 39
24 Wollast R. Schmitz M. Bigot A. Luminet O. The theory of planned behavior during the COVID-19 pandemic: A comparison of health behaviors between Belgian and French residents PLoS ONE 16 11 2021 e0258320 34735473
25 Campos-Mercade P. Meier A.N. Schneider F.H. Meier S. Pope D. Wengström E. Monetary incentives increase COVID-19 vaccinations Science 374 6569 2021 879 882 34618594
26 Dai H. Saccardo S. Han M.A. Roh L. Raja N. Vangala S. Behavioural nudges increase COVID-19 vaccinations Nature 597 7876 2021 404 409 34340242
27 Krawczyk A.L. Perez S. Lau E. Holcroft C.A. Amsel R. Knäuper B. Human papillomavirus vaccination intentions and uptake in college women Health Psychol 31 5 2012 685 22268713
28 Jens Spahn, Interview with “Die Welt”. “Die Impfung schützt – aber nicht zu 100 Prozent – vor der Infektion”. [Online; accessed 6-June-2021]; 2021. url: https://www.welt.de/politik/deutschland/article230681991/Jens-Spahn-Es-geht-um- Grundrechte-da-gibt-s-nichts-zu-relativieren.html.
29 Rieger T. Schmidt-Petri C. Schröder C. Attitudes Toward Mandatory COVID-19 Vaccination in Germany: A Representative Analysis of Data From the Socio-Economic Panel for the Year 2021 Deutsches Ärzteblatt International 119 19 2022 335 35384836
30 Anderson M.L. Multiple inference and gender differences in the effects of early intervention: A reevaluation of the Abecedarian, Perry Preschool, and Early Training Projects J Am Stat Assoc 103 484 2008 1481 1495
31 Trucchi C, Costantino C, Restivo V, Bertoncello C, Fortunato F, Tafuri S et al. Immunization campaigns and strategies against human papillomavirus in Italy: the results of a survey to regional and local health units representatives. BioMed Research International; 2019.
32 Geoffard P.Y. Philipson T. Disease eradication: private versus public vaccination Am Econ Rev 87 1 1997 222 230
33 Jin Q. Raza S.H. Yousaf M. Zaman U. Siang J.M.L.D. Can communication strategies combat COVID-19 vaccine hesitancy with trade-off between public service messages and public skepticism? Experimental Evidence from Pakistan. Vaccines 9 7 2021 757 34358173
34 Betsch C. Böhm R. Korn L. Holtmann C. On the benefits of explaining herd immunity in vaccine advocacy Nat Hum Behav 1 3 2017 0056
35 Graeber D. Schmidt-Petri C. Schröder C. Attitudes on voluntary and mandatory vaccination against COVID-19: Evidence from Germany PLoS ONE 16 5 2021 e0248372 33970933
36 Hirani J.L.J. Inattention or reluctance? Parental responses to vaccination reminder letters J Health Econ 76 2021 102439
37 Lazarus J.V. Ratzan S.C. Palayew A. Gostin L.O. Larson H.J. Rabin K. A global survey of potential acceptance of a COVID-19 vaccine Nat Med 27 2 2021 225 228 33082575
38 Schernhammer E, Weitzer J, Laubichler MD, Birmann BM, Bertau M, Zenk L et al. Correlates of COVID-19 vaccine hesitancy in Austria: trust and the government. J Public Health, 2022; 44(1): e106-e116.
39 Stoler J. Klofstad C.A. Enders A.M. Uscinski J.E. Sociopolitical and psychological correlates of COVID-19 vaccine hesitancy in the United States during summer 2021 Soc Sci Med 306 2022 115112
40 Strupat C. Shigute Z. Bedi A.S. Rieger M. Willingness to take COVID-19 vaccination in low-income countries: evidence from Ethiopia PLoS ONE 17 3 2022 e0264633 35239711
41 Busemeyer M.R. The welfare state in really hard times: Political trust and satisfaction with the German healthcare system during the COVID-19 pandemic J Eur Soc Policy 32 4 2022 393 406
42 Pornpitakpan C. The persuasiveness of source credibility: A critical review of five decades' evidence J Appl Soc Psychol 34 2 2004 243 281
43 Stiff, James B, Paul A. Mongeau. Persuasive communication. Guilford Publications; 2016.
44 Revilla M. Saris W. A comparison of the quality of questions in a face-to-face and a web survey International J Public Opinion Res 25 2 2013 242 253
45 Struminskaya B. Kaczmirek L. Schaurer I. Bandilla W. Assessing representativeness of a probability‐based online panel in Germany Online Panel Research: Data Quality Perspective, A 2014 61 85
46 Revilla M. Saris W. Loewe G. Ochoa C. Can a non-probabilistic online Panel achieve question quality similar to that of the European Social Survey? Int J Mark Res 57 3 2015 395 412
47 Griffin B. Conner M. Norman P. Applying an extended protection motivation theory to predict Covid-19 vaccination intentions and uptake in 50–64 year olds in the UK Soc Sci Med 298 2022 114819
|
PMC010xxxxxx/PMC10200718.txt |
==== Front
Soc Sci Med
Soc Sci Med
Social Science & Medicine (1982)
0277-9536
1873-5347
Elsevier Ltd.
S0277-9536(23)00336-2
10.1016/j.socscimed.2023.115979
115979
Article
From pandemic to Plandemic: Examining the amplification and attenuation of COVID-19 misinformation on social media
Lee Edmund W.J. a
Bao Huanyu a∗
Wang Yixi b
Lim Yi Torng c
a Wee Kim Wee School of Communication and Information, Nanyang Technological University, Singapore
b School of Journalism and Communication, Renmin University of China, Beijing, China
c School of Social Sciences, Nanyang Technological University, Singapore
∗ Corresponding author.
22 5 2023
7 2023
22 5 2023
328 115979115979
8 12 2022
16 5 2023
19 5 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
This study examines the proliferation of COVID-19 misinformation through Plandemic—a pseudo-documentary of COVID-19 conspiracy theories—on social media and examines how factors such as (a) themes of misinformation, (b) types of misinformation, (c) sources of misinformation, (d) emotions of misinformation, and (e) fact-checking labels amplify or attenuate online misinformation during the early days of the pandemic. Using CrowdTangle, a Facebook API, we collected a total of 5732 publicly available Facebook pages posts containing Plandemic-related keywords from January 1 to December 19, 2020. A random sample of 600 posts was subsequently coded, and the data were analyzed using negative binomial regression to examine factors associated with amplification and attenuation. Overall, the extended an extended Social Amplification of Risk Framework (SARF) provided a theoretical lens to understand why certain misinformation was amplified, while others were attenuated. As for posts with misinformation, results showed that themes related to private firms, treatment and prevention of virus transmission, diagnosis and health impacts, virus origins, and social impact were more likely to be amplified. While the different types of misinformation (manipulated, fabricated, or satire) and emotions were not associated with amplification, the type of fact-check labels did influence the virality of misinformation. Specifically, posts that were flagged as false by Facebook were more likely to be amplified, while the virality of posts flagged as containing partially false information was attenuated. Theoretical and practical implications were discussed.
Keywords
Misinformation
Social amplification of risk
Pandemic
COVID-19
Social media
Big data
Handling Editor: Medical Sociology Office
==== Body
pmcThe COVID-19 pandemic is arguably the first pandemic where communication technology and social media were employed on a large scale to relay critical information in updating, connecting, and ensuring the safety of many people (World Health Organization, 2020). However, social media may be both a bane and a boon. Its far-reaching ability to amplify and spread information may also result in an unintentional or intentional spread of misinformation. This is echoed by the World Health Organization (WHO), which has coined the term “infodemic” to refer to the proliferation of excessive information, including misinformation (Department of Global Communications, 2020). Misinformation is generally defined as false or inaccurate information, regardless of the original intent of the information (Scheufele and Krause, 2019; Wardle, 2019). During the COVID-19 pandemic, a well-known case of misinformation circulated on social media platforms, suggesting that 5G technology was responsible for the creation or increased transmission of the virus. This misleading idea gained considerable attention, becoming a popular discussion on Twitter in the UK. Numerous videos and articles supporting the false link were widely disseminated across various social media channels. The ramifications of this misinformation were serious, as people in Birmingham and Merseyside, UK, set fire to 5G towers as a result of these misguided fears. In a particularly concerning event, a telecommunications mast at Nightingale hospital in Birmingham was targeted and damaged, potentially hindering the hospital's capacity to function effectively during a crucial period (Ahmed et al., 2020). The infodemic inhibits clear public health communication efforts in relaying accurate public health information to manage the pandemic, undermining efforts to bring the pandemic under control. Therefore, given how widespread social media is used in the relay of information for the COVID-19 pandemic, tackling the infodemic on social media has become increasingly pressing and critical to also manage the ongoing health crisis plaguing the world.
To address this challenge, our study employs the Social Amplification of Risk Framework (SARF) as a guiding theoretical lens (Kasperson et al., 1988). By extending the SARF, we aim to provide a more comprehensive understanding of why certain types of misinformation gain traction on social media platforms, while others do not. In order to achieve this, our investigation will encompass a holistic examination of factors associated with risk amplification. These factors include: (a) themes of misinformation, (b) types of misinformation, (c) sources of misinformation, (d) emotions of misinformation, and (e) presence of fact-checking labels by social media platforms. By thoroughly examining these factors, this study aims to offer valuable insights into the mechanisms behind the viral spread of misinformation on the social media platforms and inform future efforts to mitigate its harmful consequences.
1 Context of study – Plandemic posts on social media
Compared to previous pandemics, COVID-19 is unique as the interconnected brought about by social media fuels the spread of different types of misinformation surrounding it (Brennen et al., 2020). Prominent types of misinformation include skepticism toward health information and policies put forth by public authorities, controversies surrounding its prevention, treatment, and symptoms, as well as conspiracy theories hinging on anti-Chinese rhetoric and racism (Brennen et al., 2020; Pei and Mehta, 2020). At the height of the pandemic in 2020, one of the most prominent forms of misinformation circulating on social media was the pseudo-documentary Plandemic. This 30-min viral video featured discredited research scientist Dr. Judy Mikovits, who put forth a series of falsehoods surrounding COVID-19, such as wearing masks would activate the COVID-19 virus, flu vaccines would increase the chances of getting the virus, and arguing that there was rampant corruption within the U.S. public health service. At its peak, it reached over 8 million views on major social media channels of YouTube, Facebook, Twitter, and Instagram (Frenkel and Alba, 2020). Research has shown that active Facebook groups peddling conspiracy theories and far-right groups such as QAnon further enabled the viral spread of Plandemic on social media (Gallagher, 2020). Given the significant impact of Plandemic and the role of Facebook in its dissemination, this study will focus on examining the misinformation present in the Plandemic case and how these posts disserminated on Facebook. As one of the main platforms for information search today, Facebook has evolved from a platform for sharing personal information to one where people share and recommend a variety of information, including news (Olmstead et al., 2011). Furthermore, since Facebook posts do not have strict word limits (compared to Twitter's 140-character limit), it offers the possibility to gather more valuable information from posts. By analyzing the discourse surrounding Plandemic on Facebook, this research aims to provide insights into the factors that contribute to the amplification or attenuation of COVID-19-related misinformation.
2 Literature review
2.1 Social amplification of risk framework
SARF offers a theoretical lens for health communication scholars to understand why and how misinformation related Plandemic becomes amplified or attenuated in some instances. SARF postulates that risk perception is socially constructed, influenced by the public response to risk hazards, social experience of risk, and perceived consequences (Kasperson et al., 1988; Renn et al., 1992; Pidgeon et al., 2002). Factors such as information sources, channels, frequency and volume of media coverage, dramatization of media content, ambiguity, and the degree of dispute among experts all contribute to shaping these perceptions and attitudes (Kasperson et al., 1988). SARF comprises two components: the first stage focuses on the transfer of risk information, while the second stage addresses societal response mechanisms, including the response or ripple effect based on risk perception (Kasperson et al., 1988, 2003). Since its introduction, SARF has been applied to various types of risk, including genetically modified foods (Frewer et al., 2002), Bovine Spongiform Encephalopathy (Lewis and Tyshenko, 2009), and the oral contraceptive pill scare (Barnett and Breakwell, 2003).
Initially developed and extended within traditional media settings, SARF has more recently been employed to understand social media's role in amplifying environmental and health risks, such as cancer risk (Strekalova, 2017), haze-related risk (Chong and Choy, 2018), Zika (Wirz et al., 2018) and dengue fever (Ng et al., 2018). Social media has changed the media environment originally envisioned by SARF (Fellenor et al., 2018), as it has increased the complexity of the risk amplification process (Chong and Choy., 2018; Fellenor et al., 2018) and has even become more powerful than traditional media in amplifying risk (Ng et al., 2018). Social media serves as a multidimensional source of information, channel, and social station for various behaviors, such as user interactions in the form of likes, comments, and reshares, which can influence the feedback and iteration of misinformation on these platforms (Zhang and Cozma, 2022).
In classic communication theory, “amplification” is defined as the intensification or attenuation of transmitted signals, resulting in the original signal having information added or removed before being passed on (Kasperson et al., 1988). In the context of risk events, amplification refers to the risk information that passes through communicators such as mass media, ultimately leading to public reaction (Ng et al., 2018). For example, the more interactions a post receives on social media, the higher its engagement. Comments, likes, and shares are often considered three dimensions of measuring social media engagement in research (Brubaker and Wilson, 2018; Kim and Yang, 2017; Jiang and Beaudoin, 2016). Prior research has partially argued that the act of amplification is a result of liking, commenting, and sharing as various extents to which information amplification occurs online (Strekalova, 2017). Liking a Facebook post results in a social signal of value to other users (Gittelman et al., 2015), while commenters act as “amplification stations for select information topics” (Strekalova, 2017). Based on previous studies, we conceptualize amplification as the total interactions of social media posts, where interactions include reactions, comments, and shares.
During the pandemic, many studies have shown that social media are misinformation amplification stations - users are exposed to and spread COVID-19 misinformation on social media platforms (Lee et al., 2021; Zhang and Cozma, 2022; Zhou et al., 2021). For example, public engagement, emotions, and information seeking on social media have been found to be strongly associated with the social amplification of risk (Zhang and Cozma., 2022). In addition, the types of misinformation during COVID-19 has been associated with their transmission during health crises (Zhou et al., 2021). While these studies have contributed significantly to our understanding of the amplification of misinformation during COVID-19, we cannot yet determine which specific content and forms of misinformation are more likely to be amplified and which are not. Given that SARF is a conceptual framework for examining social risk amplification and “initiating research as a guide to produce results beyond the scope of traditional frameworks” (Renn, 1991, p. 321), this study proposes a whole range of message-related factors within SARF, including (a) themes of misinformation, (b) types of misinformation, (c) sources of misinformation, (d) emotions of misinformation, and (e) fact-checking labels, to examine how they were associated with amplification or attenuation of misinformation.
2.2 Themes of misinformation
Prior research has showcased that “information behavior and audience engagement is topic dependent” (Strekalova, 2017), implying a connection between salient themes in social media posts and information amplification. Strekalova (2017) argued that awareness of and interest in a particular topic were important antecedents of social media user engagement. For example, risk-related messages tend to receive more amplification through user engagement compared to non-risk messages on Facebook (Strekalova, 2017). A study on social media in China also found that health warnings, advice and help-seeking misinformation significantly increased the spread of COVID-19 misinformation (Zhou et al., 2021). In such cases, certain themes of misinformation could be more widely disseminated and have more adverse effects due to the power of social media. During the COVID-19 pandemic, misinformation with a scientific veneer has been more strongly associated with a decline in vaccination intentions (Loomba et al., 2021). As a result, it is critical to situate our analysis of misinformation within its specific social milieu, where we believe that themes more likely to be amplified may also be indicative of broader social issues. Therefore, this study asks the following two research questions (RQ):RQ1 What are the different themes of misinformation in Plandemic-related content?
RQ2 What is the relationship between the themes of misinformation and the amplification or attenuation of Plandemic posts on Facebook?
2.3 Types of misinformation
Prior research on misinformation has primarily focused on broader macro-level variables, such as social networks and information ecologies (DiFonzo et al., 2013; Scheufele and Krause, 2019), or more content-based variables, such as the thematic categories of misinformation (Chen et al., 2018). Tandoc et al. (2018) proposed six types of fake news based on two dimensions: factuality and level of deception. These types include news satire, news parody, fabrication, manipulation, advertising, and propaganda. A preliminary analysis of COVID-19 misinformation revealed that misinformation containing elements of truth is more susceptible to amplification (Brennen et al., 2020). Building on these prior studies, we developed a classfication system for misinformation related to Plandemic on Facebook. We divided the misinformation into five types: satire or parody, manipulated content, fabricated content, both manipulated and fabricated content, and imposter content. Therefore, in terms of types of misinformation on social media, this study posed the following research question:RQ3 What is the relationship between the types of misinformation and the amplification or attenuation of Plandemic posts on Facebook?
2.4 Sources of misinformation
The realm of research on the sources of misinformation and the direction of amplification presents mixed findings. The sources of misinformation can vary widely due to the extensive range of actors involved in its dissemination, from prominent public figures like politicians to collateral influences (i.e., scientists, universities, science journalists, and readers of science news that may unintentionally spread misinformation among non-expert audiences), or organized and active groups of individuals that “unit [ing] their purported knowledge and political actions” (Hochschild and Einstein, 2015; Scheufele and Krause, 2019). Thus, sources of misinformation may move in a top-down manner, diffusing from publicly prominent figures, or from bottom-up actors (Brennen et al., 2020). In the context of the Plandemic misinformation movement, it is notable that the original video stemmed from a relatively outlandish and previously unheard-of source. We argue that the amplification of misinformation could be attributed to prominent public figures spreading misinformation via their Facebook posts. In this regard, this research sought to explore:RQ4 What is the relationship between the sources of misinformation and the amplification or attenuation of Plandemic posts on Facebook?
2.5 Emotions of misinformation
The psychosocial role of emotions influencing misinformation amplification has been extensively studied, with findings generally suggesting that information with a higher emotional impact is more likely to be amplified (Milkman and Berger, 2014; Strekalova, 2017; Vosoughi et al., 2018). Emotions, especially negative emotions such as anger and fear, can be predictors of risk amplification in social media (Scheufele and Krause, 2019; Wirz et al., 2018). Zhang et al. (2018) found that high-arousal negative emotions (e.g., anger, fear) were more influential than low-arousal emotions (e.g., shame, guilt) in influencing people's post-crisis social media engagement intentions. A study in the United States found that in the early days of COVID-19, when the public primarily relied on social media for information, blame and anger had a significant impact on the amplification of risk information (Zhang et al., 2021). In addition, blame is an important factor in risk amplification beyond the initial risk or risk event. For instance, public blame sentiment toward the United States government executive triggered a second peak on Twitter about the Zika risk amplification process (Wirz et al., 2018). However, in a study on risk amplification of H7N9, Zhang et al. (2017) found that positive emotions accelerated the spread of outbreak information on social media more than neutral emotions. Based on prior research, we believe that posts containing different types of emotions will have a strong relationship with total interaction. Therefore, this study raises the following question:RQ5 What is the relationship between the emotions of misinformation and the amplification or attenuation of Plandemic posts on Facebook?
2.6 Fact-checking labels
Social media sites, such as Facebook, are designed to amplify information by enabling the sharing and discussing of content within various established social networks, which may inadvertently facilitate the spread of misinformation with relative ease (Messing and Westwood, 2014; Scheufele and Krause, 2019). Algorithmic sorting of content on Facebook's feed prioritizes users' friends and family members, potentially contributing misinformation amplification (Isaac, 2018; Mosseri, 2018; Mozur, 2018). In recent years, due to public pushback against big tech companies as hotbeds of misinformation, social media platforms have taken steps to address and slow the spread of misinformation. One such strategy is to build in fact-check labels. For example, Facebook implemented a third-party fact-checking system within the app, which allows independent fact-checkers to identify misinformation. Once flagged, these posts may be removed, experience reduced visibility, or receive labels indicating the specific type of misinformation they contain (Facebook, 2020). In the context of COVID-19, Facebook has taken unprecedented fact-checking action on fake content shared online by users (Clarke, 2021). While we believe that misinformation posts without any fact-checking labels by Facebook would be more likely to be amplified, research in recent years has yielded mixed results regarding the efficacy of fact-checking tags in slowing the spread of misinformation. Several studies have shown that fact-checking labels was effective (Bode and Vraga, 2015; Zhang et al., 2021; Clayton et al., 2020), but some articles have demonstrated a limited impact (Geeng et al., 2020). As such, we asked:RQ6 What is the relationship between the presence of fact-checking labels and the amplification or attenuation of Plandemic posts on Facebook?
3 Methods
3.1 Data collection
To determine appropriate search terms for our study, we referred to Google Trends. Google search terms were used as a proxy to gauge the popularity and search volume of specific keywords or phrases associated with the #Plandemic movement on Facebook. The most prominent Plandemic-related keywords identified through Google Trends were “dr Judy Mikovits” and “Plandemic”. Therefore, our study employed the following Boolean search terms to extract Facebook data: “dr judy mikovits” OR “Plandemic” AND “Covid-19 OR covid19 OR coronavirus OR coronavirus”. We then used CrowdTangle, a public insight tool developed by Meta, to extract public posts available on Facebook. The time frame of January 1, 2020, to December 19, 2020, was used to capture the trends surrounding the #Plandemic movement in order to get as many relevant posts as possible. In total, we collected 5732 posts from public Facebook pages. The data we obtained from CrowdTangle encompasses the account, username, post creation time, post link, post content, as well as the number of likes, comments, and total interactions on the post. To assess the amplification or attenuation of Plandemic posts in this study, we used total interaction as a measurement. Interaction in CrowdTangle encompasses the sum of reactions, comments, and shares for each post, serving as an indicator of the post's engagement and reach within the Facebook community (Miles, 2022). For example, if one person sees a post three times and likes it, and comments on it, the number of interactions is 2 (Miles, 2022).
3.2 Content analysis
To qualitatively code the content of the selected posts, we randomly sampled 10% of the initial volume of Facebook posts (approximately 600 posts) for analysis. Our content analysis consisted of three steps. First, to address RQ1, two authors conducted a thematic analysis of the selected posts using inductive coding techniques. The two authors independently used inductive open coding to identify themes of Plandemic posts. Subsequently, they scrutinized the outcomes stemming from the coding process and, working autonomously, extrapolated broader thematic categories. To ensure the rigor and reliability of the findings, a third author was enlisted to meticulously review, deliberate upon, and ultimately finalize the consolidated list of eight overarching thematic categories.
Second, to address RQ2 to RQ6 and provide a multidimensional understanding of the misinformation present, we drafted a codebook for coding based on Brennen et al. (2020) and Lee et al. (2021), which included (1) misinformation, (2) themes of misinformation, (3) types of misinformation, (4) sources of misinformation, (5) emotions of misinformation, (6) fact-checking by Facebook. In determining whether a Facebook post is misinformation, we adopted a multi-step approach recommended by the World Economic Forum (Broom, 2020). First, cross-check the information against reputable sources and assess the reliability of the post's source, such as the WHO (World Health Organization, 2022), the U.S. Centers for Disease Control and Prevention (Centers for Disease Control and Prevention, 2023), and Google fact checking tools (Google, n.d). Next, look for fact-checking labels on the post and analyze its content for emotional language or provocative elements. Additionally, verify the authenticity of any images or videos using reverse image search tools, and check the date of the information to ensure its relevance. Finally, consult fact-checking websites to verify the accuracy of claims. Posts categorized as non-misinformation were subsequently excluded from further coding categories (2) to (6). Further elaboration of the categories for each coding category and their respective definitions is summarized in Table 1. While a post could fit into multiple thematic and emotional categories, we coded the post into the most dominant category, which means that a post has only one most compatible thematic and emotional category. Here, source of misinformation is the page on which the post appears, not the original source of the message (i.e., who created the misinformation originally).
Finally, we engaged two independent coders who were thoroughly trained in the coding process. To ascertain intercoder reliability, we conducted two distinct rounds of coding. In the initial round, both coders independently examined the top 10% of posts (60 posts), addressing any discrepancies in their coding through discussion and reaching consensus to guarantee a shared understanding. Subsequently, in the second round, the coders independently revisited and recoded all of the posts. To evaluate intercoder reliability, we utilized Cohen's Kappa, which is appropriate for the nominal level of measurement and the participation of two coders. We achieved high intercoder reliability scores across several categories, including Misinformation (Cohen's Kappa = .81), Themes of Misinformation (Cohen's Kappa = .66), Type of Misinformation (Cohen's Kappa = .70), Source of Misinformation (Cohen's Kappa = .84), Emotions of Misinformation (Cohen's Kappa = .71), and Fact-checking by Facebook (Cohen's Kappa = .81). A comprehensive presentation of the intercoder reliability scores can be found in Table 2.
3.3 Negative binomial regression
Addressing RQ2-6 on locating key factors affecting the amplification of COVID-19 Plandemic misinformation, negative binomial regression was carried out. This model was chosen to account for the over-dispersed nature of the count-based data collected (Date, 2019). We used this regression model to examine the relationship between total interaction for misinformation, configured as the dependent variable, and the other factors, configured as independent variables, such as the themes of misinformation, types of misinformation, sources of misinformation, emotions of misinformation, and presence of fact-checking by Facebook.
4 Results
4.1 Descriptive statistics
Out of the 600 coded posts, 333 (55.50%) posts were coded as misinformation, while the remaining posts consisted of critiques and corrections of Plandemic. RQ1 asked different themes of Plandemic misinformation related to Covid-19. The results eight themes of misinformation: Public authority personnel, action or policy (n = 91, 27.33%), treatment and prevention of virus transmission (n = 73, 21.91%), Origins of the virus (n = 48, 14.41%), Social impact (n = 39, 11.71%), Virus information (n = 34, 10.21%), Diagnosis and health impacts (n = 26, 7.81%), Private firms (n = 18, 5.41%), and Economic impact (n = 4, 1.2%).
Regarding the types of misinformation, most posts were classified as manipulated content (n = 143, 42.94%), followed by a combination of both manipulated and fabricated content (n = 116, 34.83%), fabricated content (n = 59, 17.72%), satire or parody (n = 15, 4.50%), and no imposter contect. Manipulated content included original content or information manipulated to form misinformation, whereas fabricated content comprised content that entirely false and misleading information, intentionally designed to deceive and cause harm. Thus, the distortion of existing or accurate content was a prevalent feature of the Plandemic misinformation identified in our study. Our analysis regarding the source of misinformation identified three distinct categories. The first category consists of prominent persons or sources, such as politicians, celebrities, experts, or news sources, contributing to 95 posts (28.52%). The second category encompasses non-prominent persons or sources, accounting for 156 posts (46.85%). Finally, the third category includes instances where the source was removed or missing, which represented 82 posts (24.62%).
Our results of the emotions expressed in the posts revealed a diverse range of sentiments, with the majority being neutral (n = 160, 48.05%) or expressing blame and anger (n = 68, 20.42%). Notably, there were no posts expressing hope or caring. A smaller percentage of posts conveyed fear and anxiety (n = 17, 5.00%), while an even fewer number displayed happiness, joy, or celebration (n-5, 1.50%). Lastly, 83 posts (25.00%) did not have discernible emotions. Regarding the fact-checking labels in the posts, 114 (34.23%) had no labels present. For flagged posts, 48 (14.41%) were marked as false information, meaning they had no basis in fact. Furthermore, 112 (33.63%) were flagged as partly false information due to containing some factual inaccuracies. A smaller proportion, 36 (10.81%) posts, were flagged as altered, meaning media content was edited in a misleading manner beyond adjustments for clarity. Only 17 (5.10%) posts were flagged as missing context, indicating that additional information was needed to prevent misleading interpretations. A total of 6 (1.80%) posts were set to private or deleted, rendering the content inaccessible to the coders.
4.2 Factors affecting amplification or attenuation of misinformation
In the various categories of theme, emotions, type of misinformation, source of misinformation, and type of fact-checking by Facebook, RQ2-6 attempted to examine which factors in these different categories would be associated with the amplification or attenuation of misinformation. The negative binomial regression results were presented in Table 3.
RQ2 examined the relationship between the themes of minisinformation and the amplification or attenuation of Plandemic posts on Facebook. Compared to the reference category of public authority personnel, action or policy, statistically significant results showed that the themes were positively associated with total interactions included: treatment and prevention of virus transmission (β = .51, p < .01), social impact (β = 0.46, p < .01), diagnosis and health impacts of the virus (β = 0.43, p < .01), virus origins (β = 0.41, p < .05), and private firms (β = 0.28, p < .05). Therefore, misinformation involving themes directly related to public health responses to COVID-19 in terms of treatment, prevention, health impacts, and origins were more likely to be amplified. Conversely, the categories of virus information and economic impact were not statistically significant.
RQ3 examined the relationship between the types of misinformation and the amplification or attenuation of Plandemic posts on Facebook. The findings indicated that compared to satire or parody content, the categories of both manipulated and fabricated content, manipulated content and fabricated content did not yield statistically significant results for analysis, while none of the posts were categorized as imposter content.
RQ4 examined the relationship between the sources of misinformation and the amplification or attenuation of Plandemic posts on Facebook. The results showed that none of the categories analyzed yielded statistically significant results. In addressing RQ5, which examined the relationship between emotions of misinformation and the amplification or attenuation of Plandemic posts on Facebook, the findings showed that using neutral emotions as the reference group, only posts with emotions not included in the other categories demonstrated a moderately negative relationship with total interactions (β = −.43, p < .05). The categories of blame/anger, fear/anxiety, and happiness/joy/celebration did not yield any statistically significant results, whereas none of the posts were coded as hope/caring.
Lastly, RQ6 examined the relationship between Facebook's fact-checking labeling and the amplification or attenuation of Plandemic posts on Facebook. Using the category of unflagged posts as the reference, the results showed that posts flagged as false information (β = .39, p < .01) was positively associated with the total interaction, while flagged as partly false information (β = −0.31, p < .01) had a negative relationship with the total interactions.
5 Discussion
The spread of the pseudo-documentary Plandemic on social media is an example of how misinformation could run rampant during the COVID-19 pandemic. Drawing upon and extending the social amplification of risk framework, we examined the key factors associated with amplification or attenuation of misinformation, including themes, types, sources, emotions, and the presence of fact-check labels. Our research yielded two key findings in dissecting the various factors involved in the process of misinformation amplification. First, we established Facebook users as social amplification stations within the SARF framework, examining how the perception of risk varies and encompasses both the original risk event of the pandemic and the misinformation surrounding it. We identified Facebook as a social channel within the SARF framework. Second, we examined the significance of social channel design in risk and misinformation amplification, exploring how Facebook's fact-checking labeling impacts user engagement and the amplification of misinformation. Consistent with prior research on SARF and social media, we found that social media users act as social amplification stations, with their level of interaction with information reflecting the amplification or attenuation of such information online (Strekalova, 2017).
The most frequently occurring themes in Plandemic-related misinformation include public authority personnel, action or policy, followed by treatment and prevention of virus transmission. This may suggest a mistrust of political and scientific institutions within society. In the context of misinformation, the perception of risk becomes notably more varied and expansive. Beyond the original risk event of the COVID-19 pandemic, the perception of risk now also includes misinformation surrounding the pandemic. Due to the pandemic's evolving nature, risk perceptions are not limited to the virus but also extend to public health safety measures prescribed to curtail its spread. This is highlighted by the dominant themes of amplified misinformation. User engagement is also topic-dependent, influencing the salient information circulated online (Strekalova, 2017). Misinformation with themes relating to public health responses to the COVID-19 virus typically exhibits a strong effect size, signaling a strong relationship between these misinformation topics and user engagement and subsequent amplification. Additionally, beyond immediate health impacts or measures, perceived risk in the form of misinformation on social issues and private firms further underscores the diverse set of risk perceptions that expand beyond the initial risk event. These posts seem to reflect the ripple effects of the initial misinformation that was amplified, reinforcing the complex and multidimensional character of misinformation, which may be amplified and attenuated in various ways.
Additionally, the most common misinformation type tends to be either manipulated content or a combination of both manipulated and fabricated content, implying that misinformation often contains elements of truth. This observation aligns with previous research (Brennen et al., 2020), and suggests that the distortion of existing or true content may potentially further complicate fact-checking efforts. Interestingly, our findings show that misinformation not containing emotions of blame/anger, hope/caring, fear/anxiety, or happiness/joy/celebration was less likely to be amplified compared to neutral posts. In relation to SARF, this may suggest that emotions play a prominent role in the amplification of misinformation.
The SARF framework also posits that feedback and iteration processes are involved during the amplification and attenuation stages of information dissermination. It is crucial to note that although we use total post interactions to measure the amplification and reduction of misinformation, the total number of interactions is not a direct proxy for the amplification of misinformation. Instead, total interactions are an indicator of a post's engagement and reach within the social media community. In our research, these feedback processes manifest in the interaction between the social channel platform (Facebook) and the audience social amplification stations, as analyzed through the relationship between total interactions and the presence of Facebook's fact-checking labels. Interestingly, when Facebook applies a “partly false” label, the post is less likely to be amplified, suggesting that information amplification of such posts is also reduced. This highlights the importance of acknowledging the role of social media platforms in amplifying misinformation, particularly in terms of how feedback processes between the platform and audiences may be shaped by the platform's design.
However, posts flagged as false information by Facebook were more likely to be amplified than misinformation that was not flagged. This seems inconsistent with Facebook's fact-checking program, which claims that once fact-checking labels are administered by independent fact-checkers, the platform will also take action to limit the distribution and amplification of such posts (Facebook, 2020). Our findings might suggest that the fact-checking program may not be adequately comprehensive in capturing the entire scope of misinformation posts. Although measures are taken to curb misinformation, the reactive nature of the fact-checking program results in a gap between the misinformation amplified and misinformation removed or barred from amplification. Moreover, some studies have found that the effects of fact-checking labels are limited. For example, Oeldorf-Hirsch et al. (2020) suggested that fact-check labels might not have a beneficial effect on credibility perceptions. Meanwhile, fact-checking labels failed to work when used for fear-arousing misinformation, which may be due to the boomerang effect (Lee et al., 2021). The boomerang effect refers to the reaction by an audience that is opposite to the intended response of persuasion messages (Cho and Salmon, 2007; Hart, 2014). This unintended effect often occurs in health and science communication campaigns, such as during COVID-19, where people's compliance with social distancing interventions is not absolute, even when strictly enforced (Balog-Way and McComas, 2020). Additionally, the valence of these interactions, whether affirmative or adverse, may play a critical role in shaping the dissemination patterns of posts. Consequently, an alternative explanation might be that posts flagged as false information by Facebook could provoke heightened negative emotions (e.g., anger, fear), thereby contributing to an increase in “amplification” (Han et al., 2020). When posts are flagged as misinformation, users may opt not to share the content but respond emotively instead, such as by attaching anger reactions or posting irate comments. Given our reliance on CrowdTangle's computation of total post interactions, it is imperative to recognize that negative emotions are also encompassed within the scope of “amplification.” Nevertheless, in actuality, these adverse emotions might signify a boycott or repudiation of misinformation, as opposed to endorsing it (Eberl et al., 2020).
Recognizing the complex social process of negotiating the perception of risk, SARF proves to be notably relevant in dissecting how misinformation may be amplified through social media. While prior research has acknowledged the role of social media users as amplification stations (Strekalova, 2017; Wirz et al., 2018), our study attempts to capture a more comprehensive analysis of how users interact with Facebook as a social channel. The design of the algorithm and implementation of fact-checking labels within the infrastructure of the social channel also impact misinformation amplification. Thus, we propose that the analysis of SARF in social media should also encompass an examination of how the social channel interacts with its users and how this, in turn, shapes information amplification. Given the prolonged and ever-evolving nature of the COVID-19 pandemic, another theoretical implication of using SARF to analyze misinformation is to recognize how risk perception of the original risk event may subsequently expand to include other topics related to the initial risk event, creating ripple effects beyond a singular risk event. This observation is consistent with prior research (Wirz et al., 2018).
5.1 Practical implications for tackling the infodemic
The information plague, characterized by misinformation and the prolific nature of the information-swamped public, calls for a multi-pronged solution. Firstly, this study underscores necessity of understanding the micro-level perspective on how Facebook users leverage the platform to amplify their risk perception. Health communication practitioners should be attentive to the role of social media users in disseminating and amplifying misinformation, taking measures to prevent users from exacerbating misinformation on social media. For example, implementing flags and alert pop-ups for misinformation can be beneficial. When people seek information about COVID-19, social media platforms can direct them to more reliable sources—such as the WHO or official health agencies—for accurate information (Zarocostas, 2020). Recognizing the significance of social media users, it is crucial to address misinformation through user interaction. A long-term approach involves enhancing users' ability to identify misinformation. Our research indicates that fact-checking is not always effective, and improving social media users’ information literacy can fundamentally address the problem of misinformation dissemination.
Secondly, our result showed that explicitly labeling posts as false may trigger a boomerang effect—people may be resistant. As such, fact-check labels should be applied to misinformation by alerting users to the potential for misinformation, which seems to reduce amplification. Gisondi et al. (2022) suggested that social media companies could label the blurred lines between factual news and falsehoods about COVID-19 for users through better oversight of their platforms. Social media platforms should not solely focus on identifying every instance of misinformation but instead need to reconsider their algorithms to mitigate the spread of COVID-19 misinformation, taking into account users’ psychological factors.
Thirdly, it is essential to expand the reach of experts and official accounts, with expert or institutional accounts possessing a substantial following on social media providing COVID-19 content expertise. Given the severe negative consequences of disinformation, our study emphasizes the need to address this situation by concentrating on other characteristics of information, such as misinformation themes, types, emotions, and fact-checking labels. Health organizations can debunk circulating misinformation by proactively high-quality information on social media that stresses facts without waiting for direct sharing in their streams. This approach expands the opportunity to observe corrections as they occur (Vraga and Bode, 2021). Most importantly, managing the proliferation of misinformation and infodemic necessitates deeper collaboration among social media companies, doctors, scientists, and traditional media.
6 Limitations and future research
There are several limitations to this research. One limitation of this study is the potential for biases regarding misinformation due to the possibility of completely removed posts not being mentioned in the analysis. The data collected was limited to publicly available Facebook data, thus it would not contain private, deleted, or removed posts. While efforts were made to include as many relevant posts as possible, it is possible that some posts containing misinformation were removed before they could be captured. Due to the retrospective method of collecting data, this analysis may not truly encapsulate the full extent of how misinformation is amplified on various social media platforms. Additionally, due to the lack of access to the time order in which posts are fact-checked, this research cannot compare the extent of information amplification for misinformation labelled and not labelled by Facebook. As such, the findings of this study should be interpreted with caution and further research should be conducted to explore the prevalence of misinformation on this topic. It is important to note that the limitations of this study do not negate the valuable insights gained from the analysis of the included posts.
Also, accounting for the statistically insignificant data, there was an inability to establish the relationship between the three categories of emotional frames of misinformation, type of misinformation, and source of misinformation with the total interactions of the posts. For the category of emotional frames, the only statistically significant category was on posts that did not fall into the other established emotional categories. This could be due to the subjective perception of emotion, and future research could examine the varied spectrum of emotions present in misinformation, and how it impacts misinformation amplification. Similarly, for the type of misinformation, the lack of a statistically significant relationship between the type of misinformation and total interactions garnered may indicate a need to probe further into the configuration of the different types of misinformation. This could be carried out in the form of a more in-depth textual analysis of the type of content that makes up misinformation. To better analyze how the source of misinformation impacts misinformation amplification, future studies could use network analysis to better understand the type of network of misinformation amplification, such as in terms of its modularity or density.
7 Conclusion
The COVID-19 pandemic has underscored the crucial role of social media in mobilizing and orchestrating public health responses. The rapid dissemination and far-reaching impact of misinformation, if left unchecked, can prove detrimental, undermining communication efforts by global public health authorities. Our study demonstrates that, when addressing misinformation, it is vital to adopt a comprehensive theoretical perspective informed by our extended social amplification of risk framework. This approach necessitates paying close attention to the nature of conversations surrounding COVID-19 and related policies online, as well as the types and categories of misinformation. By doing so, we can develop effective communication strategies to curb the onslaught of misinformation, as not all fact-checking labels yield successful outcomes.
Appendix Table 1 Coding categories of misinformation.
Table 1Coding Categories Definition N (%) Example Tweet
Misinformation
Misinformation False or inaccurate information, regardless of the original intent of the information 333 (55.50%) “I am reposting this video because people continue to report that the previous uploads prematurely end. I believe what we presented here on April 11th is very important, if I am permitted to say, especially in the light of the recent “Plandemic” video featuring Dr. Judy Mikovits, with whom I'm very familiar beyond her recent videos. I have her books. I must point out that it was weeks before this video and weeks before Fox News and now other mainstream media began reporting it that we exposed this University of North Carolina - Wuhan Laboratory - Dr. Fauci background to this novel coronavirus on April 11th. IJS. www.drwesley.online
Information Factual information or non-misinformation 267 (44.50%) “Pandemic or #plandemic? The latest viral internet video is a conspiracy theory documentary chock full of misleading claims about the origin of #coronavirus, flu vaccines, and wearing a mask. The video is clearly produced by professionals and looks ‚Äúnice.‚Äù But we know that there‚Äôs always more than meets the eye.
In this video, we talk about some claims hand-picked from PolitiFact’s roundup of misleading claims from the documentary.
This video was produced in partnership with PolitiFact, you can check out their excellent reporting on the ‘Plandemic’ video here:
Fact-checking ‘Plandemic’: A documentary full of false conspiracy theories about the coronavirus https://www.politifact.com/article/2020/may/08/fact-checking-plandemic-documentary-full-false-con/
Also check out our fact-check referenced in the video (Can this malaria drug (chloroquine) also cure COVID-19?) uploaded previously to IGTV in our Coronavirus series.
MediaWise is a nonprofit media literacy project of The Poynter Institute. Poynter is also home to PolitiFact.”
Themes of Misinformation
Public authority personnel, action, or policy State policies, actions, communications and recommendations, or pertaining public authority personnel such as politicians 91 (27.33%) “Watch this highly informative video by fellow physician Dr. Carrie Madej, explaining the 3 major components of the Moderna vaccine and the implications. This is a must view video. If this video doesn't open the eyes of those who don't get it, who I refer to affectionately as ""sheeple"", nothing will!
I've previously brought up how if we allow the mandatory vaccination rhetoric to proceed, humans will no longer be human. And I've talked about Moderna and the implications of their RNA vaccine which has never been commercially produced before. But Dr. Madej does an outstanding job going into greater detail. Watch this, and share it with everyone.
But Dr. Madej is wrong about one thing. She says it ""may"" cause genetic modification in our genome. Remember, the function of RNA is to repair and re-write the DNA. She's being overly generous with her words … it WILL cause a genetic change which will continue to re-write and ""repair"", ie, change our DNA.
In addition, the PLANDemic full length movie came out yesterday. If you haven't watched it, watch it now on Bitchute or on London Real.
#covid19 #covidconspiracy #corona #patriots #patriot #qanon #DrButtar #AHEADMAP #AdvancedMedicine #AdvancedMedicineConference #IADFW #Fitness #Longevity #livelonger #livehealthy #power #facts #knowledge #truth #empowerment #livefree #livefreeordie #populationcontrol #whistleblower #soldiers #wechangetheworld #healthfreedom #health #freedom #medicalfreedom”
Private firms Descriptions of commercial firms or multinational corporations, and their impact or influence pertaining to the virus 18 (5.41%) “Ben Swann takes a look at the highly unusual timeline by which Moderna Therapeutics is developing its C0-vId 19 virus vaccine. Now, 4 scientists with the NIH claim they hold partial patent rights on that vaccine and stand to make up to $150,000 per year. Meanwhile, as Moderna's stock price continues to soar, 5 top executives have sold off $89 million dollars worth of shares, even as the company continues to bypass standard vaccine protocols in the development of its C0-vId v@ ccine”
#plandemic #scamdemic #truthseekers #covid19 #truthbomb #spiritualrevolution #pizzagate #fifthdimension #exitthematrix #spiritualawakening #newearth #freethinker #lawofone #mkultra #chemtrails #follow #greatawakening #thetruthwillsetyoufree #instagram #truth #newworldorder #5dconsciousness #wakeup #governmentcorruption #higherfrequency #truths #truthhurts #truthbetold"
Treatment and prevention of virus transmission Descriptions of possible cures for the virus, vaccine development, availability, or measures to prevent virus transmission, such as social distancing, quarantine, mask-wearing, etc 73 (21.91%) “It's so easy nowadays to tell who is a brainwashed idiot and who isn't. Just check if they're wearing a mask! Don't waste a minute of your time trying to talk some sense into the few cubic centimeters of brain cells inside their thick heads. They've been completely and irreversibly brainwashed thanks to the years they spent in indoctrination centers (schools in doublespeak) and countless hours of exposure to corporate media's powerful propaganda machinery. It's an insult to the sheep to call these brainwashed idiots sheep. At least sheep are true to their nature as command-following animals. The sheeple, however, lost their nature and dignity and the ability to think critically. They're simply obedient trend-following fearful morons. They do what they're told. And they're always terrified of something like the TV conditions them to be. Fear controls them, which is precisely the same reason why people hold onto their religions and why religion will never go away. Because the fear of death and the unknown is the most ancient fear of all while religions are in the business of selling afterlife fantasies. And now wearing a mask has become a religious ritual for these fearful idiots.
#DoYourPartStayApart #N95Masks #MaskOfShame #NoMaskNoService #ItsForYourSafety #ForYourSafety #BeSafe #StaySafe #ProblemReactionSolution #ShockDoctrine #TheShockDoctrine #PlannedPandemic #Plandemic #ID 2020 #TrustTheHealthExperts #TrustTheExperts #MedicalMartialLaw #MedicalMartialLaw 2020 #Event201 #NWO #NewWorldOrder #OneWorldGovernment #OWG #HouseArrest #HouseArrestForEveryone #COVID19 #Coronavirus #COVID19Lockdown
#melbourne”
Diagnosis and health impacts Symptoms of the virus, reports of Covid-19 cases, or negative health implications, including both physical and mental health 26 (7.81%) “Yes I did check and did verify that this is true. You can as well. Government ordered hospitals many weeks ago to stop performing elective surgeries to make way for the projected numbers of coronavirus patients. So they did. And in so doing, they cut off their revenue streams. So Congress passed legislation giving hospitals billions of dollars to treat coronavirus patients. Conflict of interest? Yikes. Yes! One (many are Dr's saying) Dr. Said: “When I'm writing up my death report I'm being pressured to add COVID. Why is that? Why are we being pressured to add COVID? To maybe increase the numbers, and make it look a little bit worse than it is. We're being pressured in-house to add COVID to the diagnostic list when we think it has nothing to do with the actual cause of death. We are seeing this across the US! This is why our numbers are much higher than other countries! The actual cause of death was not COVID, but it's being reported as one of the diseases processes. … COVID didn't kill them, 25 years of tobacco use killed.” Does it get any clearer than that? Seriously, America. The only reason America is still in shutdown mode is political and fear tactics! FYI, at the ventilators to the list and hospitals get $39,000+!! For the rest of hardworking, freedom-loving America — it's time to reel in the radically unconstitutional! If you're going to dance on someone's constitutional rights, you better have a good reason! Sheltering in place decreases your immune system. This is immunology — microbiology 101. This is the basis of what we've known for years: When you take human beings and you say, ‘Go into your house, clean all your counters, Lysol them down’ … what does it do to our immune system? You know in your heart what the answer is.”
Virus origins Descriptions or claims about the source of the virus 48 (14.41%) “Really Think Corona Came From Bats?
##drjudymikovits #coronavirus #batvirus #i#infectedanimals #drbaker #howlongcoronahasbeenaround #beenaroundforcenturies
Corona didn't come from bats? Dr. Judy Mikovits explains! For the full interview click link: https://youtu.be/R0Tu8XYpQQ0
For more censored content please subscribe to my email list: https://www.drstevenbaker.com
To buy Dr. Judy's book ""The Plague Of Corruption"" click link: https://www.amazon.com/Plague-Corruption-Restoring-Promise-Science-ebook/dp/B07S5H6T4Q/ref=sr_1_1?crid=1EF9M7PX6DLV3&dchild=1&keywords=the+plague+of+corruption+book&qid=1588288768&sprefix=the+plague+o%2Caps%2C206&sr=8-1 ¿Corona no vino de los murciélagos? ¡La Dra. Judy Mikovits explica! Para el enlace completo de clic de la entrevista: https://youtu.be/R0Tu8XYpQQ0
Para obtener más contenido censurado, suscríbete a mi lista de correo electrónico: https://www.drstevenbaker.com/spanish
Para comprar el libro de la Dra. Judy ""La plaga de la corrupción"" haga clic en el enlace: https://www.amazon.com/Plague-Corruption-Restoring-Promise-Science-ebook/dp/B07S5H6T4Q/ref=sr_1_1?crid=1EF9M7PX6DLV3&dchild=1&keywords=the+plague+of+corruption+book&qid=1588288768&sprefix=the+plague+o%2Caps%2C206&sr=8-1
Virus information Descriptions of the characteristics of the virus, or how the virus is transmitted 34 (10.21%) “This interview is ideal for everybody on any level who wants to learn the truth about the Covid-19 PLANDEMIC, viruses, germ theory and how viruses are NOT IN FACT CONTAGIOUS. This is a scientific fact that has been kept from people in order to control them for centuries and Dr. Kaufman also discussed how TOGETHER we can break this spell.
.
This interview is from early March, the title card from April. It has been banned all over the world and it has been buried on this channel for months so I am reposting it in it's entirety because of how important it is. You can now also find it and other illuminating videos at DavidIcke.com under: VIRUS Video Package for London Real Viewers.
.
I simply cannot recommend this interview enough. If you watch only one video on this channel, please make it this one and give it at least 10–15 min. I assure you it gets better and better as the interview goes on. You'll also find a lot more videos and posts like these if you dig back on this channel a ways.
.
Also I've spliced in a little extra background on Dr. Kaufman from a different interview for added background.
. (Update:There is a new Andy Kaufman Video available now on londonreal.TV/Kaufman which has more updated stats than this does, but this one still remains the best in my opinion)
.
#DrAndrewKaufman #protest #wedonotconsent #notmyvirus #Icantbreath #blacklivesmatter
#endthelockdown #flattenthecurve
#WeAreTheNewsNow #wwg1wga #londonrealarmy #maga #unbearables #candiceowens #nonewnormal #digitalsoldiers #tuckercarlson #fuckBillGates #firefauci #Vaccines #billgatesisevil #BIGPHARMA #DrButtar #virusesareNOTcontagious #vaccines #DrJudyMikovitz #hoax #falseflag #PlandemicMovie
Follow my backup @trueearther2″
Economic impact Descriptions or claims about the economic impact of COVID-19, such as job losses, business closure, or economic downturn 4 (1.2%) “Cigarettes kill over 10 million people a year!!! Over 1 million people Die from second hand smoke!!! Yes they die from the choice of others
Alcohol is the cause of 5.3% of all deaths worldwide yes 5.3%
Alcohol and drug addiction costs the US economy 600billion yearly!! The worldwide figure
If you dont know the damage junk food is having on your body you are ignorant as hell!!
Cancer diabetes heart disease the list goes on!!
Hundreds of thousands of people die each year from pharmaceutical drug over dose and medical malpractice is the 3rd leading cause of death!!! But you guys keep wearing your mask sanitising your hands and believing the government want to save your life!! Repost @sharing_love_and_health
Why are the government pushing a vaccination and not healthy lifestyle???
The money is in the medicine
The government and the pharmaceutical industries worst nightmare is a healthy world
They need you to be sick
Vaccinations are a billion dollar business!! In just 6 years they went from profits of 30billion to 60billion
If the covid19 vaccination is mandatory triple that figure if not more
You are more likely to get sick from a vaccination than you are to die from the virus!!!! Tyrannical rule has always been here and now we need to stand against that!! You are a human being with the right to your own body and you should never be forced to take anything against your will EVER #saynotobillgates
#fuckbillgates #plandemic #health #healthyliving #cigarette #cancer #covidhoax #covid19 #virusscam #wakeupworld #vaccination #money
#nonewnormal #wakeup #research #question #agenda21 #event20 #arrestbillgates #sheeple #timforchange #itsbuisness #nothealthy #crimesagainsthumanity #getup #standup #standupforyourrights #tyranny”
Social impact Descriptions or claims about the social impact of COVID-19, such as domestic violence, inequality, or discrimination 39 (11.71%) “Lockdown Supporters should go tell domestic violence victims and child abuse victims as well as children and families who have no food, no electricity and can not pay their mortgage or rent just how much the lockdown made them feel safer … https://www.nytimes.com/2020/04/06/world/coronavirus-domestic-violence.html
#NoNewNormal #SayNoToBillGates #plandemic #scamdemic #fuckyourvaccine #planneddemic #scamdemic #constitution #thisisamerica #thisisntchina #truth #humanrights #freedom #usa #globalfearenterprises #takeamericaback #cannabiscommunity #cannabis #rebel #freedomisntfree #veterans #openamerica #coronavirus #endthelockdown #plandemic #childabuse #domesticviolenceawareness”
Types of Misinformation
Satire or parody No explicit intention to cause harm but has the potential to be misleading. 15 (4.50%) “Below are health experts you can search for across any and all platforms that will explain it all very clearly to you so you can stop living in mindless fear. You'll also find many of their clips on this channel if you dig.
.
Dr. Andy Kaufman
Dr. Sebi
Dr. Shiva Ayyadurai
Dr JohnBergman
Dr. Stefan Lanka
Tom Barnett
James True
Antoine Béchamp
Spacebusters - bitchute only.
.
.
#DrSebi #virusesareNOTcontagious #yourbeingliedto #wakeup
#unbearables #notmyvirus #wakeup #SocialDistancing #filmyourhospital #fearisthevirus #healthandwellness #feartactics #digitalsoldiers #coronavirus #corona #covid #plandemic #unbearables
#coronavirusoutbreak #mindfulness
#pandemic 2020 #coronaviruspandemic
#coronavirus #covid ##yoga #stopthespread #healthyliving #wearethenewsnow #citizenjournalist”
Manipulated content Content which includes original content or information that has been manipulated to form misinformation, such as: Misuse of facts or statistics, genuine content is shared with false contextual information, or genuine information or imagery is manipulated to deceive, e.g. deepfakes 143 (42.94%) “Reposted from @svisionfamily2 Coronavirus has always came from a strain of the cold family, till political influence & msm turned into a plandemic false narrative.
From false positive testing in the 80.33%
https://pubmed.ncbi.nlm.nih.gov/32133832/
Results: When the infection rate of the close contacts and the sensitivity and specificity of reported results were taken as the point estimates, the positive predictive value of the active screening was only 19.67%, in contrast, the false-positive rate of positive results was 80.33%.
Oxygen deprivation; hypercapnia, or breathing too much carbon dioxide, is a threat too your health as well.
Covid-19 virus particle size averages 125 nm (0.125 μm); the range is 0.06 μm to 0.14 μm; one needs an electron microscope to see a covid-19 virus particle.
Recommendations about masks can easily get confusing, because all masks are not made equal. The N95 mask effectively prevents viral spread. These masks, when properly fitted, seal closely to the face and filter out 95% of particles 0.3 μm or larger. But N95 masks are in serious shortage even for medical professionals. - #regrann - #regrann”
Fabricated content Content is made up and false; designed to deceive and do harm 59 (17.72%) “Dr. Judy Mikovits says 50 Million will die in the United States from Covid Vaccine, Dr. Sherry Tenpenny agrees, listen to what they even say about Bill Gates.
#lexit #blexit #covid_19 #freedom #usa #america #novaccine #deepstate #democratsdestroyamerica #latinos #hispanics #billgatesisevil #maga #kaga #nyc #wakeup #openyoureyes”
Both manipulated and fabricated content Content features a mix of fabricated and manipulated content. 116 (34.83%) “This interview is ideal for everybody on any level and will reveal the true nature of Covid-19, viruses, germ theory, how viruses are NOT contagious, the plandemic and how TOGETHER we break this spell they are using to control us.
.
This interview is from March and been banned all over the world so I have been searching for it for weeks. You can find it and other wonderful videos now at DavidIcke.com under: VIRUS Video Package for London Real Viewers.
.
I simply cannot recommend this interview enough. If you watch only one video on this channel, please make it this one.
.
.#DrAndrewKaufman
#viruseareNOTcontagious
#protests #wedonotconsent #reopenamerica #reopennyc #reopencalifornia #tyranny #filmyourhospital #notmyvirus
#givemelibertyorgivemedeath #endthelockdown #flattenthecurve
#WeAreTheNewsNow #opencalifornia #openamerica #operationgridlock #londonrealarmy #maga #constitution #candiceowens #openamericanow #yoga #endmedicaltyranny #nonewnormal #digitalsoldiers #tuckercarlson #fuckBillGates #firefauci”
Imposter content Impersonation of genuine sources, e.g. news outlets or government agencies, e.g. links with the misspelling of organization, design of webpages or graphics that closely resembles the designs of the genuine source 0 /
Sources of Misinformation
Prominent person/source Prominent sources may include politicians, celebrities, well-known experts, popular figures, or news sources 95 (28.52%) Account: Trump 2020
Non-prominent person/source Posts published by non-prominent person/source 156 (46.85%) Account: Conscious_god
Source was removed/missing Unable to find the source of the post 82 (24.62%) Account: Empty
Emotions of Misinformation
Neutral Post simply state news and information without expressing positive or negative emotions. 160 (48.05%) “Watch @doctor.mike fact check that viral‚ Plandemic‚ conspiracy theory video
#conspiracytheory #plandemic #debunked #nurseproud #nurselife #nursesrock #nursestrong #nurses #nursesofinstagram #rnlife #Stayhome #nursesofinstagram #nurselife
#coronavirus #coronavirusnurses #nursesonthefrontlines #nurseproud #nursesonthefrontline
#socialdistancing
#thenewnormal #covidnurses #nursestrong”
Blame/anger Post attacks a person or a group or accusation towards a person or a group. The post might express indignation that such a pandemic could happen. 68 (20.42%) “Trump is not going to save you and Q is for Quarantine - GROW UP! Superhero's are for children and it's waaay past time you put away your childish things. While your spending your time looking at WikiLeaks, virtue signaling your various political views and debating celebrity affiliations, your freedoms are being conquered and decimated and they will never return. Put down your playthings and pay attention to what's going on in the Here and Now. Your precious Trump is PRO VACCINE for a virus that doesn't even exist! (Mic drop) Please stop focusing on sewers and mysterious children and how pedophiles are being arrested while 100 times the amount have been released from prison. You are a subset of some of the smartest people out there and we need all hands on deck right now fighting real issues, not imaginary ones.
If Q or Trump are actually on our side they would agree 100% with this post! In fact, they literally have said dozens of times it's up to US.
.
There is no more red versus blue or Q versus the Clinton's there's only US versus them. We've all been played!.
.
Watch Rose/Icke 3 at londonreal.tv or your uninformed.
.
#qanon #qtards #trusttheplan #q #coronavirus #lockdown #truthseekers #covid19 #plandemic #protests #wedonotconsent #reopenamerica #tyranny #notmyvirus #operationgridlock #smallbusiness
#givemelibertyorgivemedeath #endthelockdown #lockdown #flattenthecurve #WeAreTheNewsNow #maga #knowyourrights #tuckercarlson #endmedicaltyranny #nonewnormal #digitalsoldiers"
Hope/caring Post proving social support, offering sympathy for victims, friends, families or others Post might include thoughts or prayers for the victims 0 /
Fear/anxiety Post about death and uncertainty of the future/economy that will reflect fear/anxiety 17 (5.00%) If you are a parent and you are actually WILLING to send your kid to indoctrination camps (aka School) and ALLOWING this form of brainwashing to occur, let me remind how god awful of a parent you are, and how you are contributing to the demoralization of a new generation, that will be enslaved by a Future Dictatorship.
When in your lifetime have you ever been programmed in school with children books to always keep your face covered and hide your identity? Never … C0vid is nothing but a Trojan horse to slowly establish a Dystopian One World Dictatorship and strip away your Identity, to be nothing but a bot in their system, that knows how to take orders, fall in line, and move or stay when told. ü§ñ If a Contagion actually existed, everyone would be dead, since nobody is constantly disinfecting door handles, car doors, store products, etc, or disposing their masks in Biohazard bins. Better yet, what about the people who work jobs that come in contact with hundreds of people a day and have no health issues?
It appalling how parents actually tolerate this nonsense, and do nothing about it. This PL4NDEMIC just shows how soft and weak willed everyone actually is, and how everyone nowadays just goes along to get along, regardless of what future consequences it may hold. Just remember, you can hide from reality all you want, but you cant run from the consequences of hiding from reality.
I even see parents screaming, Save The Children‚but sending them to school with masks lol ….
Follow these shadowbanned truth pages:
@iamnadiasavage
@red_pill_spill
@erich_alphabet
@lepetersworld
@iam.reverence
@psyopsurvivor
@holistic.nomad2
@taniatheherbalist
@at.lasthemav
@operation.wake.humanity
#plandemic #awakening #sheeple #nwo #newworldorder #society #flatearth #truth #vaccines #owo #spirituality #God #illuminati #government #media #bigpharma #world #quarantine #lockdown #psyop #5G #covid19 #coronavirus #wakeup #spiritualawakening #oneworldorder #health #wellness #fakenews #healthcare
Happiness/joy/celebration Post reflects happiness and joy despite the pandemic, and/or illustrate people coming together to celebrate certain milestones despite COVID-19 5 (1.50%) “In these unprecedented times of global upheaval, we are being given an opportunity to heal the broken aspects of ourselves and the world around us by bringing loving awareness to it. But first we must be willing to see it, to face it, in order to love it into the wholeness we all desire and deserve. It is a time to be brave and true and trust in and act out of the goodness that exists within each and everyone of us.
Thank you my dear friend, @MikkiWillis and your Elevate team for your devotion to truth, and for your willingness to stand up for the rights of all people to see what exists and to choose the course of their own lives and the lives of their loved ones.
Thank you to the millions of individual truth seekers and your willingness to prevail through the censorship, to look up and see, to dig up all possibilities and share with one another your findings. For you are paving the way for the rights of ALL people to choose, even if differently than you.
Thank you @therealbrianrose for having the courage to, in the face of such aggressive technological bullying, create a platform dedicated to the sharing of ideas, knowledge, and experience, leaving judgement to that of the many who comprise your audience.
and Thank you my followers, fans, friends and family for your willingness to LOVE along side me, with me, through me or even in spite of me. This single choosing will make us whole.
#unity #love #covid19 #health #freedom #plandemic 2020 #coronavirus @askrsb"
None of the above No emotion could be discerned 83 (25.00%) “This documentary is a reupload.
#plandemic #woke #awakening #”
Fact-checking by Facebook
Nothing was flagged No labels present 114 (34.23%) /
Flagged as false information Content that has no basis in fact 48 (14.41%) Image 1
Flagged as partly false information Content that has some factual inaccuracies 112 (33.63%) Image 2
Flagged as altered Media content edited beyond adjustments for clarity in a misleading manner 36 (10.81%) Image 3
Flagged as missing context Content that may mislead without additional context. 17 (5.10%) Image 4
Post was set to private, or was deleted Content not accessible by coders 6 (1.80%) /
Table 2 Intercoder reliability scores.
Table 2Category Cohen Kappa
Misinformation .81
Themes of Misinformation .66
Types of Misinformation .70
Sources of Misinformation .84
Emotions of Misinformation .71
Fact-checking by Facebook .81
Table 3 Negative binominal regression results.
Table 3 Beta 95% Confidence Interval
Lower Limit Upper Limit
Themes (Reference = Public authority personnel, action or policy)
Private firms .28* .05 .50
Treatment and prevention of virus transmission .51*** .24 .77
Diagnosis and health impacts .43*** .18 .67
Virus origins .41* .12 .71
Virus information −.02 −.27 .23
Economic impact −.21 −.42 .01
Social impact .46*** .19 .72
Types (Reference = Satire or parody)
Manipulated content −.28 −.82 .26
Fabricated content −.09 −.56 .38
Both manipulated and fabricated content −.52 −1.05 .01
Imposter content – – –
Emotions (Reference = Neutral)
Blame/anger .15 −.08 .39
Hope/caring – – –
Fear/anxiety .04 −.18 .25
Happiness/Joy/Celebration −.08 −.29 .13
None of the above −.43*** −.66 −.20
Sources (Reference = Originated from prominent person or source)
Non-prominent person/source −.18 −.52 .16
Source was removed/missing −.15 −.50 .20
Fact-checking by Facebook (Reference = Nothing was flagged)
Flagged as false information by Facebook .39** .16 .63
Flagged as partly false information −.31** −.53 −.10
Flagged as altered −.21 −.42 .01
Flagged as missing context −.17 −.38 .04
Post was set to private, or was deleted −.10 −.31 .12
Note. *: p < .05; **: p < .01, ***: p < .001.
Data availability
Data will be made available on request.
Acknowledgements
.
This work was supported by 10.13039/501100001475 Nanyang Technological University (grant number Start Up Grant: 020154-00001).
==== Refs
References
Ahmed W. “Covid-19 and the 5G conspiracy theory: Social Network Analysis of Twitter data,” Journal of Medical Internet Research 22 5 2020 e19458 10.2196/19458
Balog-Way D.H. McComas K.A. COVID-19: reflections on trust, tradeoffs, and preparedness J. Risk Res. 23 7–8 2020 838 848 10.1080/13669877.2020.1758192
Barnett J. Breakwell G.M. The social amplification of risk and the hazard sequence: the October 1995 oral contra- ceptive pill scare. Health, Risk & Society 5 3 2003 301 313 10.1080/13698570310001606996
Bode L. Vraga E.K. In related news, that was wrong: the correction of misinformation through related stories functionality in social media J. Commun. 65 4 2015 619 638 10.1111/jcom.12166
Brennen J.S. Simon F.M. Howard P.N. Nielsen R.K. Types, Sources, and Claims of COVID-19 Misinformation [Doctoral Dissertation 2020 University of Oxford]
Broom D. Are you an infodemiologist? here are 7 tips on how to spot false information about the pandemic. World Economic Forum https://www.weforum.org/agenda/2020/11/tackling-covid-19-misinformation-who/ 2020
Brubaker P.J. Wilson C. Let's give them something to talk about: global brands' use of visual content to drive engagement and build relationships Publ. Relat. Rev. 44 3 2018 342 352 10.1016/j.pubrev.2018.04.010
Centers for Disease Control and Prevention Myths and facts about covid-19 vaccines 2023 https://www.cdc.gov/coronavirus/2019-ncov/vaccines/facts.html
Chen L. Wang X. Peng T.-Q. Nature and diffusion of gynecologic cancer–related misinformation on social media: analysis of tweets J. Med. Internet Res. 20 10 2018 e11515 10.2196/11515
Cho H. Salmon C.T. Unintended effects of health communication campaigns J. Commun. 57 2 2007 293 317 10.1111/j.1460-2466.2007.00344.x
Chong M. Choy M. The social amplification of haze-related risks on the internet Health Commun. 33 1 2018 14 21 10.1080/10410236.2016.1242031 27901352
Clarke L. Covid-19: who fact checks health and science on Facebook? BMJ, n1170 2021 10.1136/bmj.n1170
Clayton K. Blair S. Busam J.A. Forstner S. Glance J. Green G. Nyhan B. Real solutions for fake news. Measuring the effectiveness of general warnings and fact-check tags in reducing belief in false stories on social media Polit. Behav. 42 4 2020 1073 1095 10.1007/s11109-019-09533-0
Date S. Negative binomial regression: a step by step guide https://towardsdatascience.com/negative-binomial-regression-f99031bb25b4 2019
Department Of. Global Communications U.N. tackles 'infodemic’ of misinformation and cybercrime in COVID-19 crisis https://www.un.org/en/un-coronavirus-communications-team/un-tackling-%E2%80%98infodemic%E2%80%99-misinformation-and-cybercrime-covid-19 2020
DiFonzo N. Bourgeois M.J. Suls J. Homan C. Stupak N. Brooks B.P. Ross D.S. Bordia P. Rumor clustering, consensus, and polarization: dynamic social impact and self-organization of hearsay J. Exp. Soc. Psychol. 49 3 2013 378 399 10.1016/j.jesp.2012.12.010
Eberl J.-M. “What's in a post? how sentiment and issue salience affect users’ emotional reactions on Facebook,” Journal of Information Technology & Politics 17 1 2020 48 65 10.1080/19331681.2019.1710318
Facebook How our fact-checking program works. Facebook https://www.facebook.com/journalismproject/programs/third-party-fact-checking/how-it-works 2020
Fellenor J. Barnett J. Potter C. Urquhart J. Mumford J.D. Quine C.P. The social amplification of risk on Twitter: the case of ash dieback disease in the United Kingdom J. Risk Res. 21 10 2018 1163 1183 10.1080/13669877.2017.1281339
Frenkel S. Alba D. How the ‘plandemic’ movie and its falsehoods spread widely online. The New York times https://www.nytimes.com/2020/05/20/technology/Plandemic-movie-youtube-facebook-coronavirus.html 2020
Frewer Lynn J. Miles Susan Marsh Roy “The Media and Genetically Modified Foods: Evidence in Support of Social Amplification of Risk.” Risk Analysis 22 4 2002 701 711 10.1111/0272-4332.00062 12224744
Gallagher E. Facebook groups and YouTube enabled viral spread of ‘plandemic’ misinformation https://onezero.medium.com/facebook-groups-and-youtube-enabled-viral-spread-of-Plandemic-misinformation-f1a279335e8c 2020
Geeng C. Francisco T. West J. Roesner F. Social media COVID-19 misinformation interventions viewed positively, but have limited impact. arXiv preprint arXiv:2012.11055 10.48550/arXiv.2012.11055 2020
Gisondi M.A. Barber R. Faust J.S. Raja A. Strehlow M.C. Westafer L.M. Gottlieb M. A deadly infodemic: social media and the power of COVID-19 misinformation [editorial] J. Med. Internet Res. 24 2 2022 e35552 10.2196/35552
Gittelman S. Lange V. Crawford Gotway A C. Okoro C.A. Lieb E. Dhingra S.S. Trimarchi E. A new source of data for public health surveillance: Facebook likes J. Med. Internet Res. 17 4 2015 e98 10.2196/jmir.3970 25895907
Google (n.d.) Fact check tools Google. https://toolbox.google.com/factcheck/explorer/search/COVID%20OR%20Coronavirus;hl=en.
Han J. Cha M. Lee W. Anger contributes to the spread of covid-19 misinformation Harvard Kennedy School Misinformation Review 1 3 2020 10.37016/mr-2020-39
Hart S. "Boomerang effects in risk communication." Arvai J. Rivers L. Effective Risk Communication 2014 Routledge Oxon, U.K 304 332
Hochschild J.L. Einstein K.L. Do facts matter. Information and misinformation in American politics: do facts matter Polit. Sci. Q. 130 4 2015 585 624 10.1002/polq.12398
Isaac M. Facebook overhauls news feed to focus on what friends and family share. The New York times https://www.seattletimes.com/business/facebook-overhauls-news-feed-to-focus-on-what-friends-and-family-share/ 2018
Jiang S. Beaudoin C.E. Smoking prevention in China: a content analysis of an anti-smoking social media campaign J. Health Commun. 21 7 2016 755 764 10.1080/10810730.2016.1157653 27232655
Kasperson R.E. Renn O. Slovic P. Brown H.S. Emel J. Goble R. Kasperson J.X. Ratick S. The social amplification of risk: a conceptual framework Risk Anal. 8 2 1988 177 187 10.1111/j.1539-6924.1988.tb01168.x
Kasperson J.X. Kasperson R.E. Pidgeon N. Slovic P. The social amplification of risk: assessing fifteen years of research and theory Pidgeon N. Kasperson R.E. Slovic P. The Social Amplification of Risk 2003 Cambridge University Press Cambridge 13 46 10.1017/CBO9780511550461.002
Kim C. Yang S.U. Like, comment, and share on Facebook: how each behavior differs from the other Publ. Relat. Rev. 43 2 2017 441 449 10.1016/j.pubrev.2017.02.006
Lee J. Choi J. Britt R.K. Social media as risk-attenuation and misinformation-amplification station: how social media interaction affects misperceptions about COVID-19 Health Commun. 2021 1 11 10.1080/10410236.2021.1996920
Lewis R.E. Tyshenko M.G. The impact of social amplification and attenuation of risk and the public reaction to mad cow disease in Canada Risk Anal.: Int. J. 29 5 2009 714 728 10.1111/j.1539-6924.2008.01188.x
Loomba S. de Figueiredo A. Piatek S.J. de Graaf K. Larson H.J. Measuring the impact of COVID-19 vaccine misinformation on vaccination intent in the U.K. and USA Nat. Human Behav. 5 3 2021 337 348 10.1038/s41562-021-01056-1 33547453
Messing S. Westwood S.J. Selective exposure in the age of social media: endorsements Trump partisan source affiliation when selecting news online Commun. Res. 41 8 2014 1042 1063 10.1177/0093650212466406
Miles C. FAQ: Interactions Per Post 2022 https://help.crowdtangle.com/en/articles/4003674-faq-interactions-per-post
Milkman K.L. Berger J. The science of sharing and the sharing of science Proceedings of the National Academy of Sciences - PNAS 4 111 2014 13642 13649 10.1073/pnas.1317511111
Mosseri A. Helping ensure news on Facebook is from trusted sources 2018 https://about.fb.com/news/2018/01/trusted-sources/
Mozur S.F.N.C.P. Some Countries, Facebook's Fiddling Has Magnified Fake News. The New York Times 2018 https://www.nytimes.com/2018/01/14/technology/facebook-news-feed-changes.html
Ng Y.J. Yang J. Vishwanath A. To fear or not to fear. Applying the social amplification of risk framework on two environmental health risks in Singapore J. Risk Res. 21 2018 1487 1501 10.1080/13669877.2017.1313762
Oeldorf-Hirsch A. Schmierbach M. Appelman A. Boyle M.P. The ineffectiveness of fact-checking labels on news memes and articles Mass Commun. Soc. 23 5 2020 682 704 10.1080/15205436.2020.1733613
Olmstead K. Mitchell A. Rosenstiel T. Facebook is becoming increasingly important https://www.pewresearch.org/journalism/2011/05/09/facebook-becoming-increasingly-important/ 2011
Pei X. Mehta D. # Coronavirus or# Chinesevirus?!: understanding the negative sentiment reflected in Tweets with racist hashtags across the development of COVID-19. arXiv preprint arXiv:2005.08224 10.48550/arXiv.2005.08224 2020
Pidgeon N. Kasperson R.E. Slovic P. The Social Amplification of Risk 2002 Cambridge Univ. Press Cambridge, UK
Renn O. Risk communication and the social amplification of risk Communicating Risks to the Public 1991 Springer Dordrecht 287 324
Renn O. Burns W.J. Kasperson J.X. Kasperson R.E. Slovic P. The social amplification of risk: theoretical foundations and empirical applications J. Soc. Issues 48 4 1992 137 160 10.1111/j.1540-4560.1992.tb01949.x
Scheufele D.A. Krause N.M. Science audiences, misinformation, and fake news. (Essay) Proceedings of the National Academy of Sciences of the United States 116 16 2019 7662 10.1073/pnas.1805871115
Strekalova Y.A. Health risk information engagement and amplification on social media: news about an emerging pandemic on Facebook Health Educ. Behav. 44 2 2017 332 339 10.1177/1090198116660310 27413028
Tandoc E.C. Jr. Lim Z.W. Ling R. Defining “fake news” A typology of scholarly definitions Digital journalism 6 2 2018 137 153 10.1080/21670811.2017.1360143
Vosoughi S. Roy D. Aral S. The spread of true and false news online Science 359 6380 2018 1146 1151 10.1126/science.aap9559 29590045
Vraga E.K. Bode L. Addressing COVID-19 misinformation on social media preemptively and responsively Emerg. Infect. Dis. 27 2 2021 396 33395379
Wardle C. First draft's essential guide to understanding information disorder https://firstdraftnews.org/wpcontent/uploads/2019/10/Information_Disorder_Digital_AW.pdf?x76701 2019
Wirz C.D. Xenos M.A. Brossard D. Scheufele D. Chung J.H. Massarani L. Rethinking social amplification of risk: social media in Zika in three languages Risk Anal. 38 12 2018 2599 2624 10.1111/risa.13228 30408201
World Health Organisation Managing the COVID-19 infodemic: promoting healthy behaviors and mitigating the harm from misinformation and disinformation 2020 https://www.who.int/nefirstws/item/23-09-2020-managing-the-covid-19-infodemic-promoting-healthy-behaviours-and-mitigating-the-harm-from-misinformation-and-disinformation
World Health Organization Covid-19 mythbusters 2022 https://www.who.int/emergencies/diseases/novel-coronavirus-2019/advice-for-public/myth-busters
Zarocostas J. How to fight an infodemic The lancet 395 10225 2020 676 10.1016/S0140-6736(20)30461-X
Zhang X.A. Cozma R. Risk sharing on Twitter: social amplification and attenuation of risk in the early stages of the COVID-19 pandemic Comput. Hum. Behav. 126 106983 2022 10.1016/j.chb.2021.106983
Zhang L. Xu L. Zhang W. Social media as amplification station: factors that influence the speed of online public response to health emergencies Asian J. Commun. 27 3 2017 322 338 10.1080/01292986.2017.1290124
Zhang X. Borden J. Kim S. Understanding publics' post-crisis social media engagement behaviors: an examination of antecedents and mediators Telematics Inf. 35 8 2018 2133 2146 10.1016/j.tele.2018.07.014
Zhang J. Featherstone J.D. Calabrese C. Wojcieszak M. Effects of fact-checking social media vaccine misinformation on attitudes toward vaccines Prev. Med. 145 2021 106408 10.1016/j.ypmed.2020.106408
Zhou C. Xiu H. Wang Y. Yu X. Characterizing the dissemination of misinformation on social media in health emergencies: an empirical study based on COVID-19 Inf. Process. Manag. 58 4 2021 102554 10.1016/j.ipm.2021.102554
|
PMC010xxxxxx/PMC10201310.txt |
==== Front
Med Intensiva (Engl Ed)
Med Intensiva (Engl Ed)
Medicina Intensiva
2173-5727
Published by Elsevier España, S.L.U.
S2173-5727(23)00067-X
10.1016/j.medine.2023.05.009
Scientific Letter
Clinical experience of prophylactic enoxaparin dosage adjustment guided by AntiXa factor levels in critical care patients with COVID-19-induced pneumonia: observational study☆
Experiencia del ajuste de dosificación de enoxaparina profiláctica dirigida con niveles de factor AntiXa en pacientes críticos con neumonía COVID-19: estudio observacionalBermúdez-Ruiz María del Carmen a
Vilar Sánchez Irene a
Aparicio Pérez Clara b
Carmona Flores Rosario a
Rodríguez-Gómez Jorge ac⁎
de la Fuente-Martos Carmen ac
a Servicio de Medicina Intensiva, Hospital Universitario Reina Sofía, Córdoba, Spain
b Servicio de Hematología, Hospital Universitario Reina Sofía, Córdoba, Spain
c Instituto de Investigación Maimones de Investigación Biomédica de Córdoba (IMIBIC), Córdoba, Spain
⁎ Corresponding author.
22 5 2023
22 5 2023
© 2023 Published by Elsevier España, S.L.U.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
==== Body
pmcTo the Editor,
The risk of thrombotic events increases in patients with SARS-CoV-2 (COVID-1)-induced pneumonia.1, 2 Clinical practice guidelines and scientific societies simply do not agree on their recommendations for the management and prevention of these events. Therefore, the Spanish Society of Intensive and Critical Care Medicine and Coronary Units (SEMICYUC) has proposed adjusting the dosage of low molecular weight heparin (LMWH) by obtaining the Anti-Xa factor levels.3, 4 While they justified this recommendation due to the high risk of thrombotic/hemorrhagic events and the rate of under/overdose (23% and 46%), the level of evidence granted was C-III (weak support for the recommendation based on expert opinions or descriptive studies). The utility of this strategy to reduce these complications, and to determine the appropriate levels, is controversial.3, 5, 6, 7 The objectives of our study were to assess the adjustment of prophylactic enoxaparin by Anti-Xa factor, the levels obtained, and the presence of possible risk factors associated with overdose in patients admitted to the ICU with COVID-19-induced pneumonia.
We conducted an observational and retrospective study at the ICU of a tertiary care level hospital from July 2020 through February 2021. Consecutive critically ill patients with COVID-19-induced pneumonia were evaluated in whom the prophylactic enoxaparin dose was directed by the Anti-Xa factor levels. The inclusion criteria were: (1) patients with COVID-19-induced pneumonia (presence of radiological infiltrates and positive PCR test for SARS-CoV-2 in respiratory samples); (2) need for respiratory support (high-flow oxygen therapy, non-invasive support, or invasive mechanical ventilation); (3) adjustment of prophylactic enoxaparin dosage directed by Anti-Xa factor. Cases with an initial therapeutic indication were excluded. In those patients who modified their regimen from prophylactic to therapeutic during admission, the doses and levels were both included, as well as thrombotic/hemorrhagic complications during the prophylaxis period.
The protocol of our center for initial enoxaparin doses consisted of (1) standard dose of 40 mg/24 h; (2) dose of 60 mg/24 h in patients with body mass index (BMI) > 30 or elevated acute phase reactants (D-dimer levels >1500 ng/mL and C-reactive protein >150 mg/L); (3) dose of 80 mg/24 h if both risk factors were present (BMI + elevated reactants). In case of kidney disease, dosage was reduced based on creatinine clearance. Afterwards, the dose was adjusted based on the Anti-Xa levels. In addition, given the high risk of thrombosis reported, target range from 0.30 IU/mL to 0.59 IU/mL were agreed upon by a multidisciplinary team, which are similar target range compared to those from former studies.1, 6, 7 Measurements were taken in a sort of treatment balance state of enoxaparin (48 h–72 h without dose changes) and in a peak phase, 3 h–5 h after the administration. They were periodically repeated every 2–3 days depending on the patients’ clinical situation and response to treatment.
Variables associated with demographics, comorbidities, severity at the ICU admission, lab test results, treatment, and need for support were all collected. The outcomes "thrombotic events" at the ICU stay (deep vein thrombosis and/or acute pulmonary thromboembolism diagnosed through imaging modalities requested at the doctor’s discretion), "hemorrhagic events" (severe: critical location, reduction of Hb by 2 g/dL or transfusion of 2 units of red blood cells), and mortality at the ICU stay were also recorded. Regarding the statistical analysis, the categorical variables were expressed as absolute value and percentage and compared using the chi-square test or Fisher's exact test. The continuous ones were expressed as median while 25th-75th percentiles were compared using the Student t test, or the Mann-Whitney U test when the variables did not have a normal distribution. The factors associated with dependent variables "Anti-Xa levels ≥0.60 IU/mL" (values above the target range) and "thrombotic events" were analyzed through multivariable logistic regression studies. Variables with P values <.2 or of greater clinical interest were explored. The cut-off value for continuous variables was adjusted using the Youden index method.
A total of 160 critically ill patients with COVID-19-induced pneumonia on enoxaparin prophylaxis were included in the study, and a total of 589 Anti-Xa factor determinations were performed. The patients’ median age was 63 years being obesity the most common comorbidity of all (67%, [n = 107/160]), and showing rates of creatinine clearance <30 mL/min at admission of 6% (10/160). The median of D-dimer values at admission were 1242 ng/mL (p25−75, 704−3765) with 58% (94/160) requiring invasive mechanical ventilation. The mortality rate at the ICU stay was 28% (45/160) with rates of diagnosed thrombotic events and severe bleeding of 9% (15/160) and 4% (7/160), respectively (Table 1 ).Table 1 Clinical characteristics of the patients.
Table 1N 160
Demographic data
Age (years), median (p25−75) 63 (54−69)
Sex, male, n (%) 107 (66.9)
Past medical history
No comorbidities, n (%) 41 (25.6)
Obesity (BMI ≥ 30), n (%) 108 (67.5)
Weight 82.5 (76.5−100)
Hypertension, n (%) 92 (57.5)
Diabetes mellitus, n (%) 43 (26.9)
Smoker, n (%) 4 (2.5)
Chronic heart failure, n (%) 14 (8.8)
Chronic respiratory failure, n (%) 29 (18.1)
Chronic kidney disease, n (%) 7 (4.4)
Previous thromboembolic disease, n (%) 4 (2.5)
Past medical history of cancer, n (%) 14 (8.8)
Solid organ transplant, n (%) 1 (0.6)
Severity at admission
APACHE II score, median (p25−75) 9.5 (7−13)
SOFA score, median (p25−75) 4 (3−4)
Lab test results
Platelets × 103/uL, median (p25−75) 285 (224−350)
Creatinine clearance, mL/min, median (p25−75) 103 (70−125)
Creatinine clearance <30 mL/min, median (p25−75) 10 (6.3)
D-dimer levels, ng/mL, median (p25−75) 1242 (704−3765)
C-reactive protein, mg/L, median (p25−75) 58 (24−127)
Treatment
High-flow oxygen therapy, n (%) 138 (86.3)
Invasive mechanical ventilation, n (%) 94 (58.8)
Prone position, n (%) 77 (48.1)
Vasoactive drugs, n (%) 74 (46.3)
Renal replacement therapy, n (%) 18 (11.3)
ECMO, n (%) 6 (3.8)
Boluses of corticoids, n (%) 109 (68.1)
Results
Hemorrhagic events, n (%) 17 (10.6)
Severe bleeding, n (%) 7 (4.4)
Thrombotic events, n (%) 15 (9.4)
Pulmonary thromboembolism, n (%) 12 (7.5)
Mortality at the ICU admission, n (%) 45 (28.1)
APACHE, Acute Physiology and Chronic Health Evaluation; BMI, body mass index; COVID-19, coronavirus 2019; ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; SOFA, Sequential Organ Failure Assessment.
Table 2 shows the doses of enoxaparin received and the Anti-Xa factor values obtained at the first determination and at the ICU stay. When the first determination was performed, the median dose of enoxaparin administered was 60 mg/24 h (p25−75, 40−60). In relation to these doses, the first determination of Anti-Xa factor levels had a median of 0.28 IU/mL (0.34−0.50). A total of 53% of the patients (85/160) were within the target range of 0.30 IU/mL to 0.59 IU/mL.Table 2 Doses of prophylactic enoxaparin and AntiXa factor levels in critical patients.
Table 2 Overall cohort Creatinine clearance BMI
<30 mL/min ≥30 mL/min P <30 ≥30 p
N 160 (100) 10 (6.2) 150 (93.7) 52 (32.5) 108 (67.5)
First determination at the ICU admission
Doses of enoxaparin (mg/24 h), median (p25−75) 60 (40−60) 40 (20−40) 60 (40−60) <.001 60 (40−60) 60 (40−60) .14
Patients with doses of enoxaparine > 40 mg/24 h, n (%) 155 (96.6) 5 (50) 150 (100) <.001 50 (96.2) 105 (97.2) .71
First determination of antiXa levels, median (p25−75) 0.38 (0.24−0.50) 0.18 (0.12−0.26) 0.4 (0.26−0.50) <.001 0.40 (0.27−0.51) 0.34 (0.23−0.49) .20
First determination of antiXa levels (IU/mL), n (%): .01 .47
AntiXa < 0.10 2 (1.3) 0 (0) 2 (1.3) 1 0 (0) 2 (1.9)
AntiXa 0.10−0.29 53 (33.1) 8 (80) 45 (30) .02 14 (26.9) 39 (36.1)
AntiXa 0.30−0.59 86 (53.8) 2 (20) 84 (56) .04 31 (59.6) 55 (50.9)
AntiXa 0.60−0.69 11 (6.9) 0 (0) 11 (7.3) 1 4 (7.7) 7 (6.5)
AntiXa ≥ 0.70 8 (5.0) 0 (0) 8 (5.3) 1 3 (5.8) 5 (4.6)
N 160 (100) 22 (13.7) 138 (86.2) 52 (32.5) 108 (67.5)
At the ICU admission
Mean dose of enoxaparin (mg/24 h), median (p25−75) 60 (53−73) 60 (40−66) 60 (53−73) .10 60 (40−70) 60 (55−80) .01
Mean levels of AntiXa, median (p25−75) 0.48 (0.39−0.59) 0.48 (0.35−0.57) 0.48 (0.39−0.60) .80 0.47 (0.39−0.57) 0.49 (0.39−0.6) .90
Mean levels of AntiXa (IU/mL), n (%): .48 .93
AntiXa < 0.10 2 (1.3) 1 (4.5) 1 (0.7) 1 (1.9) 1 (0.9)
AntiXa 0.10−0.29 18 (11.3) 3 (13.6) 15 (10.9) 6 (11.5) 12 (11.1)
AntiXa 0.30−0.59 100 (62.5) 13 (59.1) 87 (63) 33 (63.5) 67 (62)
AntiXa 0.60−0.69 29 (18.1) 3 (13.6) 26 (18.8) 7 (13.5) 22 (20.4)
AntiXa ≥ 0.70 11 (6.9) 2 (9.1) 9 (6.5) 5 (9.6) 6 (5.6)
Number of determination of the AntiXa levels/patient, median (p25−75) 3 (2−5) 4 (3−9) 3 (2−5) .02 3 (1.5−4) 2 (3−5.5) .31
BMI, body mass index; COVID-19, coronavirus 2019.
During the rest of the stay, the dose of prophylactic enoxaparin was adjusted based on the Anti-Xa factor levels within the target range of 0.30 IU/mL to 0.59 IU/mL. The mean Anti-Xa factor level was 0.48 IU/mL (0.39−0.59). A total of 62% (100/160) of the patients fell within the target range (0.30 IU/mL to 0.59 IU/mL) while 25% (24/160) within ranges ≥0.60 IU/mL. Enoxaparin doses > 60 mg/24 h (OR, 4.57; 95%CI, 3.17–6.60; P < .001) and C-reactive protein levels < 175 mg/dL (OR, 2.30; 95%CI, 1.28–4.11; P = .002) were independently associated with an increased risk of obtaining Anti-Xa factor values ≥0.60 IU/mL (above the target range) in the multivariate analysis (Supplementary Table S1).The median adjusted enoxaparin dose at the stay was 60 (53−73) mg/24 h, which is significantly higher in patients with BMI > 30 of 60 (55−80) mg/24 h vs. 60 (40−70) mg/24 h (P = .01). No variables independently associated with the development of thrombotic events were seen (Supplementary Table S2).
Currently, new studies on the management and prevention of thrombosis in patients with COVID-19 are being published like the INSPIRATION, REMAP-CUP, ATTACC, ACTIV-4a, etc. However, key aspects such as the appropriate dose of LMWH and possible optimization through Anti-Xa factor levels2, 3 have not yet been clarified. In our experience, we saw that although steering the dose of prophylactic enoxaparin resulted in 62% of patients having mean Anti-Xa factor levels within the target range, 25% showed higher ranges (≥0.60 IU/mL). Higher doses of heparin (enoxaparin >60 mg/24 h) and lower C-reactive protein levels (<175 mg/dL) were independently associated with a higher risk of overdose. Bösch J et al. found that elevated C-reactive protein values can play a role in resistance to LMWH activity, which could explain why patients from our study with lower values were more exposed to overdose.8 Although levels above the target range can potentially pose a greater risk of complications, only 6% had levels ≥0.70 IU/mL, and the rate of major bleeding was 4%, not particularly higher compared to that described in the medical literature (2%–6%).6, 7, 9
In this study, the dose of enoxaparin administered at the ICU stay was 60 mg/24 h (p25−57, 50−73 mg). These doses are similar to those reported in former studies (median, 60 mg/24 h, 50 mg/24 h to 80 mg/24 h) that used monitoring with similar target levels (0.4 IU/mL to 0.5 IU/mL) 6. Mohamed A et al. found a high rate of overdose (48%) when intermediate doses (0.5 mg/kg/12 h) are administered—which is one of the strategies recommended in the medical literature available—and the Anti-Xa levels are monitored.3, 7 The risk of overdose could have been higher if this regimen had been used without monitoring given the high rate of obesity (67%) and elevated levels of acute phase reactants found in our cohort.2, 3, 7
Finally, the objective of steering the dose of enoxaparin based on the Anti-Xa factor is to reduce the risk of thrombotic events. We should mention that, in our study, we found rates of 9%, which is the lower border limit of what has been reported by the medical literature available (9%–26%). However, this analysis is limited because we did not conduct a systematic search for thrombotic events meaning that some events may have been underdiagnosed.2, 3
We are aware of the limitations of the study as it is observational and retrospective in nature and includes a non-systematic search for thrombotic events. Also, the size of its sample is limited. In addition, the target range proposed, although similar to that from former studies, is controversial with significant variability in publications (eg, surgical patients, 0.1 IU/mL to 0.3 IU/mL vs other COVID-19 studies, 0.3 IU/mL to 0.7 IU/mL).5, 10 Still, we believe that this clinical experience can be useful regarding the design of future studies.
In conclusion, in this study of critically ill patients with COVID-19-induced pneumonia, the Anti-Xa factor levels obtained at the ICU stay allowed us to adjust doses of prophylactic enoxaparin. Nonetheless, there is a high risk of overdose especially in patients with higher enoxaparin doses and lower C-reactive protein levels.
Ethics of this scientific publication
This study has been approved by Hospital Universitario Reina Sofía Ethics Committee (Code 010408). The need for informed consents was not deemed necessary given the observational and retrospective nature of the study.
Funding
None whatsoever.
Conflicts of interest
None reported.
Appendix A Supplementary data
The following are Supplementary data to this article:
Acknowledgements
the authors wish to thank Prof. Manuel Rodríguez Peralvarez and Dr. Rafael León López for their support in the statistical analysis of data.
☆ Please cite this article as: Bermúdez-Ruiz MC, Vilar Sánchez I, Aparicio Pérez C, Carmona Flores R, Rodríguez-Gómez J, de la Fuente-Martos C. Experiencia del ajuste de dosificación de enoxaparina profiláctica dirigida con niveles de factor AntiXa en pacientes críticos con neumonía COVID-19: estudio observacional, Med Intensiva. 2023. https://doi.org/10.1016/j.medin.2023.04.012https://doi.org/10.1016/j.medin.2023.04.012
Appendix A Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.medin.2023.04.012.
==== Refs
References
1 Wichmann D. Sperhake J.-P. Lütgehetmann M. Steurer S. Edler C. Heinemann A. Autopsy findings and venous thromboembolism in patients with COVID-19 Ann Intern Med. 173 2020 268 277 32374815
2 Vincent J.L. Levi M. Hunt B.J. Prevention and management of thrombosis in hospitalised patients with COVID-19 pneumonia Lancet Respir Med. 10 2022 214 220 34838161
3 Flaczyk A. Rosovsky R.P. Reed C.T. Bankhead-Kendall B.K. Bittner E.A. Chang M.G. Comparison of published guidelines for management of coagulopathy and thrombosis in critically ill patients with COVID-19: implications for clinical practice and future investigations Critical Care. 24 2020 559 572 32938471
4 Vidal-Cortés P. Díaz Santos E. Aguilar Alonso E. Amezaga Menéndez R. Ballesteros M.A. Bodí M.A. Recomendaciones para el manejo de los pacientes críticos COVID-19 en las Unidades de Cuidados Intensivos Med Intensiva. 46 2022 81 89
5 Witt D.M. Nieuwlaat R. Clark N.P. Ansell J. Holbrook A. Skov J. American Society of Hematology 2018 guidelines for management of venous thromboembolism: optimal management of anticoagulation therapy Blood Adv 2 2018 3257 3291 30482765
6 Kofteridis D.P. Ioannou P. Kondili E. Chamilos G. Filippatos T.D. Personalized prophylactic anticoagulation in hospitalized patients with Covid-19 - The role of anti-Xa monitoring Clin Microbiol Infect. 27 2021 1188 1189 33933565
7 Mohamed A. Shemanski S.M. O Saad M. Ploetz J. Haines M.M. Schlachter A.B. Anti-Xa directed thromboprophylaxis in critically ill patients with coronavirus disease 2019 Clin Appl Thromb Hemost. 28 2022 1 9
8 Bösch J. Rugg C. Schäfer V. Lichtenberger P. Staier N. Treichl B. Low-molecular-weight heparin resistance and its viscoelastic assessment in critically ill COVID-19 patients Semin Thromb Hemost. 48 2022 850 857 36174602
9 Trunfio M. Salvador E. Cabodi D. Marinaro L. Alcantarini C. Gaviraghi A. e-COVID Study group Anti-Xa monitoring improves low-molecular-weight heparin effectiveness in patients with SARS-CoV-2 infection Thromb Res. 196 2020 432 434 33049598
10 Cook D. Crowther M. Meade M. Deep venous thrombosis in medical-surgical critically Ill patients: prevalence, incidence, and risk factors Crit Care Med. 33 2005 1565 1571 16003063
|
PMC010xxxxxx/PMC10201327.txt |
==== Front
Semin Immunol
Semin Immunol
Seminars in Immunology
1044-5323
1096-3618
The Authors. Published by Elsevier Ltd.
S1044-5323(23)00069-6
10.1016/j.smim.2023.101778
101778
Article
Systems analysis of human innate immunity in COVID-19
Müller Sophie abc
Schultze Joachim L. acd⁎
a Systems Medicine, Deutsches Zentrum für Neurodegenerative Erkrankungen (DZNE) e.V., Bonn, Germany
b Department of Microbiology and Immunology, The University of Melbourne at the Peter Doherty Institute for Infection and Immunity, Melbourne, Victoria, Australia
c Genomics & Immunoregulation, Life and Medical Sciences (LIMES) Institute, University of Bonn, Bonn, Germany
d PRECISE Platform for Single Cell Genomics and Epigenomics, DZNE and University of Bonn, Bonn, Germany
⁎ Corresponding author at: Systems Medicine, Deutsches Zentrum für Neurodegenerative Erkrankungen (DZNE) e.V., Bonn, Germany.
22 5 2023
7 2023
22 5 2023
68 101778101778
2 2 2023
13 5 2023
13 5 2023
© 2023 The Authors
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Recent developments in sequencing technologies, the computer and data sciences, as well as increasingly high-throughput immunological measurements have made it possible to derive holistic views on pathophysiological processes of disease and treatment effects directly in humans. We and others have illustrated that incredibly predictive data for immune cell function can be generated by single cell multi-omics (SCMO) technologies and that these technologies are perfectly suited to dissect pathophysiological processes in a new disease such as COVID-19, triggered by SARS-CoV-2 infection. Systems level interrogation not only revealed the different disease endotypes, highlighted the differential dynamics in context of disease severity, and pointed towards global immune deviation across the different arms of the immune system, but was already instrumental to better define long COVID phenotypes, suggest promising biomarkers for disease and therapy outcome predictions and explains treatment responses for the widely used corticosteroids. As we identified SCMO to be the most informative technologies in the vest to better understand COVID-19, we propose to routinely include such single cell level analysis in all future clinical trials and cohorts addressing diseases with an immunological component.
Keywords
Systems immunology
Single cell multi-omics technologies
COVID-19
Innate immune system
Monocytes
Neutrophils
NK cells
Long COVID-19
==== Body
pmc1 Introduction
The numerous waves of SARS-CoV-2 virus variants [1] have caused a rather broad spectrum of disease courses of COVID-19 [2], [3], [4]. The spectrum of symptoms induced by different SARS-CoV-2 strains varies [4], but for all virus variants disease courses range from mainly asymptomatic to very severe and critical courses in a smaller subset of patients [4], [5]. It became obvious that the interaction between the human immune system and the virus is a major driver of viral evolution and in contrast to other coronaviruses SARS-CoV-2 is still evolving very fast [6]. Patients with immune deficiencies cannot clear the virus very efficiently, which can prolong viremia and accelerate viral evolution [7], [8], clearly indicating that the functional status of the immune system including its innate arm is a major determinant of disease course and severity. Furthermore, comorbidities including obesity, diabetes and cardiovascular diseases associated with chronic inflammation are major risk factors for severe courses [9], [10] further supporting the notion that the immune system is a decisive component when it comes to disease course and severity. Genetic studies have identified a small number of mutations related to autoimmune or inflammatory diseases [11]. Furthermore, patients with mutations in TLR7 are at elevated risk for severe disease courses [12], [13]. Similarly, mutations in interferon (IFN) pathways downstream of pattern recognition receptor signaling also point towards an important role of the innate immune system in defining disease courses and severity [14].
The time course of COVID-19 is best described by an initial phase of viremia followed by the second inflammatory response phase, which turns into viral sepsis in severe and critical cases [15]. In case of recovery, the second phase is followed by a third phase of convalescence with a return to immune homeostasis. While the dynamics of the disease phases is varying widely between patients, sophisticated time course analysis of high-content data suggest that convalescence is associated with a universal cellular and molecular program [16]. A smaller percentage of patients are not recovering and either experience prolonged clinical symptoms or even develop new symptoms after a period without any disease symptoms [4], [17], [18]. This clinical syndrome also termed long COVID has been linked to chronic inflammation, but also to the development of functional autoantibodies and even prolonged viral titers have been suggested as pathophysiological mechanisms [19], [20]. In contrast to acute COVID-19, the role of the immune system and its individual cellular components including innate immune cells in long COVID are still not sufficiently understood. However, we are convinced that systems immunology approaches as we and others applied them to better understand the role of the immune system in acute COVID-19 [21], [22], [23], [24], [25] will also be most instrumental to better understand the role of the immune system in long COVID.
Here, we focus on knowledge gained from studies interrogating the immune response to COVID-19 with technologies supporting systems level analysis and description of the dynamic processes occurring during the different phases of the infection. We focus on the surprisingly strong heterogeneity of activation of cellular components of the innate immune system not only at the entry site of the virus but also during systemic inflammation. Further, we summarize what is known so far about processes during convalescence and lay out further necessary steps to better understand the heterogeneity of the pathophysiology during long COVID. We will highlight some surprising innate immune functions of a subset of T cells during SARS-CoV-2 infections and the potential link to endothelial damage. Finally, we will give an outlook on how we envision to further develop systems immunology approaches into a response system for emerging infectious diseases.
2 Systems level interrogation of the human immune response to COVID-19
Prior to the pandemic, a large consortium of European scientists had worked for over two years to design a roadmap toward improving European healthcare through cell-based interceptive medicine [26]. This roadmap includes three major areas to be developed for clinical applications: 1. Single-cell multi-omics (SCMO), 2. Artificial intelligence / machine learning (AI/ML), and 3. Patient-derived experimental disease models. All three areas are major ‘ingredients’ of the big data-driven circle of systems immunology we previously suggested [27] and recently extended to SCMO applications [28]. At the beginning of the pandemic, we conceptualized how these technologies could be utilized to better understand the pathophysiology of the SARS-CoV-2 infection and to determine mechanisms that might explain the heterogeneous dynamics and disease trajectories of COVID-19 [15]. Looking at the major breakthroughs for a better understanding of COVID-19, it can be argued that genome-wide methods on the single cell level, in particular single cell RNA-seq technologies, have contributed most to a fast understanding of disease trajectories, cell types involved, key pathways altered and important genes dysregulated as a consequence of the infection [15], [29]. However, it is also important to mention that scRNA-seq data very successfully guided follow-up studies including functional assessment of immune cells, phenotyping of immune deviations using targeted single cell technologies such as flow cytometry-based assays. Further, it became rather clear that systems level interrogation at scale requires clinical validation, which we introduced early on in the pandemic [21], illustrating that data integration - as it is currently being pursued for example in the Human Cell Atlas for generating healthy tissue maps - might be complemented or replaced by classical approaches from clinical research, namely clinical validation trials independently validating and confirming findings from early studies. Collectively, our experience during the first three years of the pandemic clearly point towards the enormous advantage of genome-wide single cell technologies in accelerating the identification of the major disease-specific and -associated changes in any given organ system and that targeted approaches are better suited for follow up or validation and confirmation studies.
As a consequence, we propose for any future threat by (re-)emerging infectious and also non-infectious diseases to apply SCMO - best combined with AI/ML applications - early on during the exploratory phase complemented by targeted validation studies ( Fig. 1), which then will also allow to develop patient-derived experimental disease models, which could further accelerate the identification and development of therapeutic targets, an area that requires more attention in the future.Fig. 1 General setup in clinical cohorts and validation cohorts to study COVID-19 and Long COVID by high-resolution single cell multi-omics technologies.
Fig. 1
3 Acute innate immune response to COVID-19
Here, we summarize findings that have been mainly obtained by studies using systems immunology technologies, in particular single cell omics approaches.
3.1 Local immune response to SARS-CoV-2
The respiratory tract is the major entry route for SARS-CoV-2, where it binds to ACE2 receptors on nasal and oral epithelial cells, mainly goblet secretory cells [30], [31], [32], [33]. A first line of defense against SARS-CoV-2 infection are IgA antibodies in sputum and saliva [34]. If local cells are productively infected by SARS-CoV-2, tissue-resident immune cells are activated in the upper respiratory tract, which in turn leads to activation and recruitment of circulating immune cells to the site of infection through inflammatory mediators [15]. This process is performed in close interaction of epithelial cells and tissue-resident immune cells, which sense the virus and initiate a defense response. By production of chemoattractants, innate and adaptive immune cells are recruited from the circulation to support the ongoing immune response. Following viral clearance, the resolution of the inflammatory response is initiated to inhibit hyperactivation of the immune system, which can induce exacerbated tissue damage.
Several single cell studies examining bronchoalveolar lavage fluid (BALF) and nasopharyngeal swabs from COVID-19 patients have yielded insight into the role of the innate immune response to SARS-CoV-2. A single cell study of nasopharyngeal swabs from COVID-19 patients identified increased frequencies of IFN-responsive ciliated epithelial cells in mild and moderate disease contributing to the antiviral response in the upper respiratory tract [33]. Another single cell study including samples from the upper (nasopharyngeal/pharyngeal swabs) and lower respiratory tract (BALF) during the first wave of the pandemic showed a stronger inflammatory profile for alveolar tissue-resident macrophages in the lower airways [35]. Furthermore, in critically ill patients, there were more ligand-receptor interactions predicted between epithelial cells and immune cells [35]. Together with the high pro-inflammatory profile of the interacting immune cells such as monocyte-derived macrophages and cytotoxic T cells, the prediction of a large number of ligand-receptor interactions between immune cells and epithelial cells is likely indicative for a higher number of highly activated infiltrating cells into the lung and more inflammatory tissue damage in critical compared to moderate disease [35]. Whether this difference is still valid for later SARS-CoV-2 variants such as omicron characterized by reduced involvement of the lower respiratory tract [36] requires further evaluation. An early finding was the identification of ACE2- tissue resident and monocyte-derived alveolar macrophages, which harbored SARS-CoV-2 transcripts likely from uptake of escaped virus particles through phagocytosis [37], [38]. SARS-CoV-2 harboring myeloid cells showed a distinct transcriptional program including upregulation of cytokine and chemokine transcripts such as CCL4, CCL20, CXCL10, CXCL11, IL1B [37].
The release of chemoattractants by innate immune cells is likely responsible for the influx of additional immune cells to the lung, which is reflected by elevated levels of activated monocytes, monocyte-derived macrophages and neutrophils in BALF derived from COVID-19 patients [37], [38]. In contrast, the number of tissue-resident alveolar macrophages is drastically reduced in severe disease [38], [39]. While the remaining alveolar macrophage pool is altered in its functionality including reduced antigen presentation capacity [39], depending on disease severity and phase of the infection, the infiltrating immune cells have been described as hyperinflammatory [40]. Moreover, lung infiltrating monocyte-derived macrophages acquire a pro-fibrotic phenotype in severe disease [23].
In addition to BALF, local immune activation was also assessed by spatial single cell transcriptomics in post-mortem lung tissues. Among several studies, we highlight two of the larger studies [41], [42]. In the first study, 116,314 nuclei from lungs of 19 COVID-19 infected lungs and 7 control lungs contained highly activated myeloid cells including monocytes, monocyte-derived macrophages and alveolar macrophages in the lungs of deceased COVID-19 patients [41]. In the second study, lung, heart, liver, and kidney tissues were studied by scRNA-seq [42]. Similar to early studies in BALF [37], [38], SARS-CoV-2 RNAs were primarily detected in endothelial cells and myeloid phagocytes, which was associated with higher expression of viral response genes, chemokines, and cytokines [42]. Collectively, SARS-CoV-2 infection induces a profound local innate immune response, however, the magnitude and quality of this response in context of different viral strains, disease severity and phase, including prolonged immune deviations in long COVID are neither comprehensively characterized on the single cell level nor completely understood.
3.2 Heterogeneous deviations in the blood myeloid cell compartment
Analysis of the peripheral blood gives insight into the systemic effects of SARS-CoV-2 infection. Several studies examining peripheral blood from COVID-19 patients using SCO revealed complex immune response regulations during SARS-CoV-2 infection [43]. Despite different sampling times, study design (cross sectional versus longitudinal), SCO technologies used and heterogeneity of important clinical metadata (e.g. sex, age, comorbidities, ongoing medication including immunosuppressive drugs such as corticosteroids), several immune deviations were reported independently by several studies strongly supporting these changes to be hallmarks of COVID-19 [44]. For example, concerning cell type frequencies it was shown that circulating non-classical CD16+ monocytes were reduced in moderate and even more so in severe disease [21], [45], [46], [47]. At the same time, SCO revealed additional transcriptional states within the myeloid cell compartment, which differed between disease phase and severity. While monocyte states identified in mild to moderate disease were characterized by expression of genes related to a productive antiviral immune response such as CD83 or IFN-stimulated genes (ISGs), dysregulated molecular phenotypes were revealed in patients with severe disease, for example monocyte states characterized by reduced expression of HLA-DR genes [21], [46], [47], [48]. HLA-DR genes encoding for MHC molecules are crucial for antigen presentation and the induction of an adaptive immune response. A recent study combining murine data from a murine virus infection and single cell RNA seq data of human PBMCs from COVID-19 patients, found strong differences in the functional capacity of monocytes and dendritic cells to productively present antigens to CD8 + T cells depending on the disease severity. A transcriptional program, taking place in antigen-presenting cells induced by type-1 IFNs and CD4 + T cell help, was found to be required to effectively activate anti-viral CD8 + T cells. Importantly, DCs and monocytes from mild but not severe COVID-19 patients have high chromatin accessibility and transcription of these genes, which correlates with the CD8 + T cell response in those patients [49]. Focusing on monocytes in severe COVID-19, cell states could be further subdivided into HLA-DRlow monocytes appearing early during infection, characterized by expression of CD163 and alarmins S100A4/8/9/12, while later during severe infection elevated CD163 expression was lost [21], [46]. Despite some early discrepancies concerning the role of the interferon system in COVID-19 [50], in monocytes expression levels of ISGs are highest at early time points, which is even more prominent in severe COVID-19, and levels consistently decrease over time, linking the IFN response inversely to disease severity and phase [21], [51]. This transient expression of ISGs correlated with plasma IFN-α levels [48]. However, the regulation of the IFN response is even more complex as in a subset of critically ill COVID-19 patients type-1 IFN-specific autoantibodies were found, which are associated with an impaired type-1 IFN response in myeloid populations including monocytes [52]. An interesting observation was that those monocytes displaying the highest type-1 IFN response also expressed ligands and receptors for interaction with platelets [25]. When directly comparing to monocytes in severe influenza infection, a monocyte cell state in severe COVID-19 patients was associated with higher TNF and IL-1b driven responses together with an enrichment of a type-1 IFN gene signature [53], which was often related to hyperinflammation. Since these patients were also characterized by harboring HLA-DRlow tolerizing or suppressed monocytes, we would suggest that the myeloid compartment in severe COVID-19 is better described as being a major part of a viral sepsis phenotype with COVID-19 related immune deviations.
3.3 Elevation and transcriptional reprogramming of neutrophils in COVID-19
Many single cell studies of peripheral blood are based on the isolation of peripheral blood mononuclear cells (PBMC), which misses out the large majority of granulocytes including neutrophils. However, in diseases such as COVID-19, where neutrophils have been shown to be significantly elevated and potential drivers of the disease pathology [21], [50], [54] these studies might have missed important pathophysiology of the disease. Bulk RNA sequencing of whole blood samples already indicated that the blood neutrophil compartment in severe COVID-19 is characterized by signatures reminiscent of pre-/immature neutrophils and simultaneous inflammatory and suppressive features, arguing for a complex dysregulation best captured by SCO technologies [55]. Indeed, studying the complete white blood cell compartment in COVID-19 patients by scRNA-seq led to the discovery of an enrichment of neutrophil precursors and immature neutrophils pointing towards emergency myelopoiesis, which was not present in patients with flu-like illness or healthy controls [21]. Generation of single cell transcriptomes of the bone marrow compartment from patients with severe COVID-19 validated this hypothesis illustrating accumulation of granulocyte-monocyte precursors (GMPs) and an upregulation of transcription factors determining the granulocyte lineage [56]. Accumulation of immature Arginase 1 (ARG1 +) neutrophils in peripheral blood was also confirmed for critically ill COVID-19 patients and bacterial induced acute respiratory distress syndrome (ARDS) patients that were admitted to intensive care units but not in healthy individuals [57]. Not only the presence of immature neutrophils but also significant alterations in the mature neutrophil pool characterized COVID-19. In severe COVID-19 a peculiar neutrophil state occurs, which simultaneously expresses interferon response signature genes and genes associated with immunosuppression, e.g. CD274 (PD-L1) or ARG1 [21], [45], [47], [57]. Immunosuppressive neutrophils have been previously reported in the context of systemic inflammation, e.g. in septic shock patients and after in vivo stimulation with LPS, to inhibit lymphocyte proliferation [58], [59]. In severe COVID-19 a similar mechanism might take place [54]. However, he precise effect of these dysregulated neutrophil states on cell-cell interaction and the disease outcome requires further investigation [60]. Additionally, neutrophils with high expression of alarmins S100A8 and S100A9 were detected in the peripheral blood and BALF from severe COVID-19 patients [21], [47]. CyTOF analysis confirmed the suppressive phenotype showing increased expression of PD-L1 and CD62L downregulation on mature and immature neutrophils from severe COVID-19 patients [21].
Taken together, neutrophils are not only increased in frequencies in COVID-19, but show dramatic transcriptional alterations leading to novel cell states (activated/ immunosuppressive) but also reprogramming of known differentiation-associated neutrophil states, particular precursor and immature neutrophils, reminiscent of emergency hematopoiesis. The assessment of the neutrophil compartment by SCO technologies is probably a prime example of the power of the application of high-resolution technologies to complete tissues including blood, when it comes to a fast, comprehensive understanding of pathophysiological processes. This then also allows to prioritize downstream analysis and potentially target identification for biomarker and therapies.
3.4 Transcriptional alterations in the megakaryocyte compartment
Recent findings strongly support innate-like functions of megakaryocytes and platelets [61], [62]. A large enough longitudinal single cell multi-omics study identified an increase of megakaryocytes with an activated phenotype and a strong type-1 IFN signature in the peripheral blood of COVID-19 patients [63], which was later confirmed and extended to the identification of megakaryocyte progenitors [25]. The expression of megakaryocyte genes in the early hematopoietic stem cell-derived progenitor cells in peripheral blood and bone marrow from COVID-19 points towards emergency megakaryopoiesis, a phenomena also seen in other infections to replenish the peripheral platelet consumption during acute infection, both in patients and murine models [25], [56], [64]. Megakaryocytes expanded in COVID-19 patients have increased expression of PKM2, encoding a pyruvate kinase, which points towards higher pyruvate metabolism and glycolysis in this cell compartment during COVID-19 [63]. There is some indication that platelets, the product of megakaryocytes, promote thromboinflammation in SARS-CoV-2 pneumonia [65], but much more work is necessary to fully appreciate the role of molecular reprogramming of the megakaryocyte pool in COVID-19.
3.5 Persistently dysfunctional NK cells in severe COVID-19
For effective control of many viral infections cytokine production and cell-mediated cytotoxicity by NK cells is essential [66]. Under physiological conditions, cytokine-producing CD56bright and cytotoxic CD56dim NK cells are the most prevalent subsets [67]. During acute COVID-19, there is a decrease in circulating NK cells, especially in severe COVID-19 patients [22], [68]. NK cells present in the circulation have clear signs of activation as assessed by the surface expression of CD69 and CD38, which is increased in moderate and severe COVID-19 patients [45]. Longitudinal characterization of NK cells from acute COVID-19 patients by sc-RNA-seq identified high expression of IFN stimulated genes and a prolonged expression of genes involved in IFNa signaling in severe COVID-19, while TNF-related genes were upregulated in early moderate disease [22]. scRNA-seq data also predicted functional impairment (virus control, cytokine production and cell-mediated cytotoxicity) in severe COVID-19, which was supported by functional flow cytometry assays [22]. A subsequent study confirmed NK cell dysfunctionality and linked these to an unbalanced TGF-b response during the early phases of the infection, which altered the expression of genes related to cell-mediated cytotoxicity, granule exocytosis and cell-cell adhesion [68]. How more recent single cell transcriptome studies, describing hyper-activated NK cells and NK cell-platelet aggregates [69] or the appearance of a memory-like NK cell subset [70], are related to earlier studies requires further investigation.
3.6 Innate immune function of T cells during COVID-19
While the importance of T cells in antigen-specific immune responses required for viral control and clearance is unquestioned [71], single cell transcriptomics was critical in identifying innate immune functionality of T cells in COVID-19 [24]. Here, a subset of CD4+ T cells, CD8+ T cells as well as gamma-delta T cells all expressed CD16 on their surface. While natural killer T (NKT) cells, an unconventional T cell population recognizing antigen in context of CD1d molecules, are also known to express CD16 [72], they make out only a small proportion of the activated CD16 + T cells in COVID-19 patients [24]. These CD16+ T cells were significantly elevated in severe COVID-19 in the circulation and the lung, but absent in influenza infection, and exerted high TCR-independent cytotoxic potential by immune-complex-mediated degranulation. CD16 expression on T cells was induced by the complement protein C3a and both elevated plasma levels of C3a and higher frequencies of CD16+ T cells were associated with fatal outcome. The expression of CD16 on T cells was confirmed in a recent study also demonstrating that CD16+ cells are activated via soluble multimeric immune complexes, which are present in ∼80% of severe and critically ill COVID-19 patients [73].
Collectively, acute COVID-19 is characterized by reprogramming of all major innate immune compartments even inducing innate immune-like functionality in T cells and certain immune deviations within the innate immune system are directly linked to more severe or even fatal outcome of COVID-19, further underlining the importance to not only understand the adaptive arm of the anti-viral immune response, but also changes in the innate immune system.
4 Innate immune responses during convalescence and long COVID-19
Systems immunology approaches have also started to be applied to better understand recovery from acute COVID-19 and the development of long COVID [4], [17], [18], [19], [20], a heterogeneous clinical syndrome, which comes with different definitions and names and is still not well-defined (for more information see Table 1). For simplicity we use the term long COVID from here on. Gaining insight into the kinetics of recovery from COVID-19 is important to better understand deviations from the recovery process as seen in long COVID. The dynamics of recovery are rather patient specific and difficult to grasp when using chronological time. Applying computational modeling on longitudinally sampled blood transcriptomes from severe COVID-19 patients revealed common disease regression dynamics with reduction in neutrophils being the best predictor for convalescence, followed by an early rise in T cell activation and differentiation as well as the rebalancing between NF-κB and IFN signaling [16]. Normalization of cell numbers as assessed by SCMO was reported for circulating monocytes by an increase in non-classical monocytes and a reduction in classical monocyte frequency during convalescence [74], [75]. While decreased HLA-DR expression is a hallmark of monocytes in severe COVID-19, in convalescent individuals HLA-DR lo S100A hi monocytes are absent and replaced by a monocyte subset with high antigen presentation capacity (high expression of HLA-DQA and HLA-DPA) [75]. However, in convalescent omicron patients, circulating immune cell type frequencies (e.g. activated monocyte subsets and megakaryocytes) as well as plasma cytokines and chemokines (e.g. CXCL10, CCL4, IL-9), remained altered around 42 days post-acute infection [76]. Patients fully clinically recovered from mild-to-moderate COVID-19 showed immune alterations even four months post-acute COVID-19, e.g. elevated pro-inflammatory plasma cytokines (IFN-β, IFN-λ1, IFN-γ, CXCL9, CXCL10, IL-8, soluble TIM-3) [77].Table 1 Definitions of Long COVID-19.
Table 1Name Symptoms/ clinical manifestation Time after COVID-19 COVID-19 severity Reference
Post COVID-19 condition/syndrome fatigue, shortness of breath, cognitive dysfunction or other symptoms affecting everyday functioning > 3 months, lasting at least 2 months mild-severe WHO, 2021 https://www.who.int/publications/i/item/WHO-2019-nCoV-Post_COVID-19_condition-Clinical_case_definition-2021.1
fatigue, exertion intolerance, partly meeting the criteria for ME/CFS 4–15 months mild-moderate [91]
fatigue, exertion intolerance, partly meeting the criteria for ME/CFS 7–19 months mild-moderate [86]
respiratory symptoms, pulmonary abnormalities 3–6 months severe [94]
Mainly cognitive and neurological symptoms > 3 months na [84]
Post-acute sequelae of COVID-19 (PASC) one or more of the following: constitutional, respiratory, cardiopulmonary, gastrointestinal, genitourinary, reologic, rash, musculoskeletal symptoms, trouble sleeping > 3 months mild-severe [78]
respiratory, gastrointestinal, neurological symptoms, anosmia/dysgeusia 2–3 months mild-severe [20]
pulmonary and extrapulmonary symptoms 2–3 months severe [96]
fatigue, respiratory, digestive, neurological, skin and mucous membrane, circulation symptoms > 1 months mild [76]
different symptoms including fatigue, dyspnea, respiratory, body aches > 1 months mainly mild-moderate [79]
respiratory symptoms, interstitial lung changes 3–12 months mild-severe [95]
pulmonary and extrapulmonary symptoms 2–3 months severe [96]
According to WHO case definition > 6 months mild-severe (mainly hospitalized) [97]
Long COVID fatigue, dyspena, chest pain > 4 months mild-severe [77]
breathing difficulties/breathlessness, fatigue/malaise, chest/throat pain, anxiety/depression, headache, myalgia, other pain, cognitive, abdominal symptoms 3–6 months mild-severe [71]
33 symptoms based on WHO case definition for post COVID-19 condition > 4 months mild [18]
with and without cognitive symptoms (e.g. "brain fog") na mainly mild [93]
Physical symptoms (e.g. arthralgia/myalgia, fatigue, cough), exertion intolerance 12 months severe [80]
e.g. fatigue, neurological, cardiopulmonary, gastrointestinal symptoms, trouble sleeping ∼4 months mild-severe Peluso et al., 2022
Fatigue, cough, shortness of breath 3–12 months mild-severe [85]
One or more symptoms of the following e.g. pulmonary, systemic, neurological or psychiatric symptoms 12 months mild and moderate/severe [83]
Post-acute COVID-19 syndrome different symptoms including fatigue, memory issues, loss of smell, abdominal pain ∼7 months mainly mild [81]
In contrast, in patients experiencing long COVID, IFN-β and IFN-λ1 remained elevated even 8 months after acute infection. Other studies described other sets of plasma cytokines (TNF, IFNγ, IL-10, IL-1β, IL-6, IL-12, IL-17) to be increased in patients with long COVID [78], [79], [80] further supporting the notion that long COVID is a rather heterogeneous syndrome with different endotypes or heterogeneous causes including persistent inflammation, viral persistence, autoimmune phenomena or a combination thereof [19]. Support for persistent inflammation as a cause of long COVID comes from studies describing elevated frequencies of circulating activated monocytes and pDCs [77]. Interestingly, a small observational study in SARS-CoV-2 infected patients with intestinal bowel disease (IBD) suggested viral persistence in gut mucosa derived from long COVID patients [81], a finding that certainly requires further investigation. An early single cell multi-omics study providing longitudinal data from acute infection until 2–3 months after infection suggested that risk factors for the development of long COVID include high level SARS-CoV-2 viremia, type 2 diabetes, Epstein-Barr virus reactivation during acute disease and the presence of autoantibodies [20]. Further validation of these intriguing findings are required to better understand whether the described patient subgroups are indeed endotypes of this heterogeneous syndrome. Another recent study links Epstein-Barr virus reactivation with a higher risk for the development of long COVID and especially fatigue and neurological symptoms [82]. Additionally, multiple studies have reported the association of autoantibodies with long COVID symptoms [83], [84], [85].
Major long COVID-associated symptoms resemble those of myalgic encephalomyelitis/ chronic fatigue syndrome (ME/CFS) [86]. ME/CFS is a complex, chronic disease, which is characterized by cognitive impairment, fatigue, post-exertional malaise but can also include chronic muscle pain and cardiovascular complaints [87], [88]. The development of the syndrome is implicated with a preceded infection similar to long COVID [89], [90]. Since precise biomarkers are lacking, the diagnosis of ME/CFS is difficult, mostly based on the clinical phenotype and the exclusion of other diseases. The molecular and cellular mechanisms underlying ME/CFS are not well understood. Signs of endothelial dysfunction were reported in long COVID patients with a ME/CFS like clinical phenotype, for example demonstrated by higher endothelin-1 plasma concentrations [91]. Given the broad overlap of symptoms between ME/CFS and long COVID similar pathophysiological mechanisms might occur, which could be identified by high-resolution technologies such as SCMO.
Other important clinical manifestations of long COVID are neurological alterations such as cognitive impairment and concentration deficits. Neurological symptoms have been described for both, acute COVID-19 and long COVID [92]. From a murine SARS-CoV-2 infection model inducing mild to moderate respiratory COVID-19 disease, we know that microglia/macrophage activation, depletion of oligodendrocytes, decreased myelination, and signs of impaired hippocampal neurogenesis are hallmarks of a prolonged disease phenotype persisting for more than 7 weeks post-acute infection [93]. At the same time, elevated cytokine levels including CCL11 in cerebrospinal fluid and plasma were associated with reduced neurogenesis and cognitive impairments. Elevated plasma levels of CCL11 were also identified in patients with lasting neurological symptoms, suggesting that a similar mechanism takes place in humans and mice [93]. Persistent neurological inflammation might be one mechanism explaining neurological symptoms in long COVID. A major challenge of future research will be the identification of risk factors predicting neurological symptoms, which occur only in a subset of long COVID patients. Worldwide efforts including the European consortium NeuroCOV (https://www.neurocov.eu/) are set up to define those molecular drivers of SARS-CoV-2-induced neuropathology that cause the manifestation of neurological symptoms and complications.
Most likely a different endophenotype of long COVID is characterized by prolonged pulmonary dysfunction, which seems to occur more frequently in patients with severe COVID-19, particularly after prolonged ICU treatment [94], [95], [96]. In these patients pulmonary symptoms persist for a prolonged period of time and are associated with alterations of the lung interstitium as assessed by computed tomography imaging [95]. Furthermore, plasma proteomics of these patients revealed cellular and molecular changes with enriched plasma IL17C levels and elevated numbers of neutrophils, which were characterized by increased expression of chemokines and neutrophil protease myeloperoxidase. Whether these changes are an effect of prolonged hospitalization and ICU treatment or specific for a protracted course of acute severe COVID-19 remains to be determined. There are indications that the risk to develop long COVID symptoms is linked to the severity of the acute disease. One study reported fewer mucosal CD8 + β7Integrin+ T cells and higher SARS-CoV-2 specific IgA levels in the circulation of long COVID patients with severe acute disease compared to milder acute disease [97]. Future studies need to further address, how other cellular and humoral characteristics of long COVID are connected to the severity of the acute disease.
Collectively, the heterogeneous syndrome long COVID following acute COVID-19 infections in a smaller subset of patients requires further research to better define patients at risk, determine endophenotypes, unravel different molecular mechanisms causing the myriad of clinical symptoms as a prerequisite for the development of patient-tailored therapy regimens. Large comparative therapy trials combined with biomarker platforms as they have been initiated for example in Germany (National Clinical Study Group for long COVID, https://longcoviddeutschland.org/nksg/) will be instrumental to reach these goals.
5 Conclusions and future perspectives
Systems level analysis of the human immune response to SARS-CoV-2 infection uncovered the major cells, pathways and genes involved in both, a productive immune response to the virus leading to protective immunity and even more so the immune deviations, particularly seen in those patients with severe and critical disease trajectories. Systems approaches have the great advantage that all major components of the immune system are measured simultaneously, which allows to quickly determine a hierarchy of events at any given time of the infection and to prioritize mechanisms, pathways and genes for biomarker development but also to identify potential targets for treating immune deviation. While widespread clinical application of single-cell multi-omics was certainly in its infancies shortly prior to the start of the pandemic, this devastating world-wide event also triggered an accelerated use of these technologies and many hurdles that were anticipated to take years until solutions would be found, were overcome in much shorter time and leading to additional innovations. For us, one such innovation is the introduction of validation studies, when it comes to the clinical use of SCMO. Rather than integration of data from different sites, which is currently the preferred method for tissue atlasing, we are convinced that data obtained at different clinical sites should be treated as individual clinical trials or cohorts and used to validate findings from the other sites [21], [22], [23], [24], [63] (Fig. 1). One major result of SCMO-based systems approaches was the identification of disease stage- and severity-associated changes within the innate immune system in response to the SARS-CoV-2 infection [21]. Very surprising was the identification of innate functionality by cytotoxic CD16+ T cells appearing within the major T cell compartments, namely CD4+, CD8+ and gamma-delta T cells [24]. Such findings are a particular strength of systems approaches since classical approaches, for example a classical gating strategy applied in flow cytometry, would probably have used CD16 expression to distinguish T cells from CD16+ NK cells. In contrast, SCMO independently performed in two cohorts unequivocally established this T cell functionality. Further, functional validation experiments later confirmed the SCMO-predicted function.
Based on these exciting developments in clinical applications of SCMO-based technologies, two major directions require further attention. First, it needs to be determined if SCMO is only applicable during the exploratory phase and requires clinical validation of the initial findings via easier and more clinically applicable technologies. Or, if SCMO technologies can be further scaled and provided as much more cost-efficient technical solutions, with costs currently being one major hurdle for widespread clinical application. Second, even if these technologies would become widely clinically applicable, for a faster response towards (re-) emerging pandemic threats, we need to develop standardized and connected worldwide systems that would allow a fast and coordinated response when it comes to determine disease hallmarks and pathogen-specific immune responses and/or immune deviations. We have recently developed such a system, termed Swarm Learning [98] and are currently in the process to provide proof-of-principle that this AI- and block chain-based system could be used to build world-wide networks of hubs that are capable of measuring immune responses with highest resolution [99]. One big advantage of Swarm Learning is that there is no need to exchange data. Swarm Learning only exchanges insights, which makes working together across institutions and jurisdictions much easier.
Acknowledgements
The work was supported by the 10.13039/501100001659 German Research Foundation (DFG, GRK2168 – 272482170, Excellence Cluster ImmunoSensation2, project number 390873048. Figures were created with BioRender.com.
==== Refs
References
1 Mahilkar S. Agrawal S. Chaudhary S. Parikh S. Sonkar S.C. Verma D.K. SARS-CoV-2 variants: Impact on biological and clinical outcome Front Med (Lausanne) 9 2022 995960 10.3389/fmed.2022.995960
2 Berlin D.A. Gulick R.M. Martinez F.J. Severe covid-19 N. Engl. J. Med 383 2020 2451 2460 10.1056/NEJMcp2009575 32412710
3 Gandhi R.T. Lynch J.B. Del Rio C. Mild or moderate Covid-19 N. Engl. J. Med 383 2020 1757 1766 10.1056/NEJMcp2009249 32329974
4 Thygesen J.H. Tomlinson C. Hollings S. Mizani M.A. Handy A. Akbari A. COVID-19 trajectories among 57 million adults in England: a cohort study using electronic health records Lancet Digit Health 4 2022 e542 e557 10.1016/S2589-7500(22)00091-7 35690576
5 Gupta A. Madhavan M.V. Sehgal K. Nair N. Mahajan S. Sehrawat T.S. Extrapulmonary manifestations of COVID-19 Nat. Med 26 2020 1017 1032 10.1038/s41591-020-0968-3 32651579
6 Turakhia Y. Thornlow B. Hinrichs A. McBroome J. Ayala N. Ye C. Pandemic-scale phylogenomics reveals the SARS-CoV-2 recombination landscape Nature 609 2022 994 997 10.1038/s41586-022-05189-9 35952714
7 Kemp S.A. Collier D.A. Datir R.P. Ferreira I.A.T.M. Gayed S. Jahun A. SARS-CoV-2 evolution during treatment of chronic infection Nature 592 2021 277 282 10.1038/s41586-021-03291-y 33545711
8 Markarian N.M. Galli G. Patel D. Hemmings M. Nagpal P. Berghuis A.M. Identifying markers of emerging SARS-CoV-2 variants in patients with secondary immunodeficiency Front Microbiol 13 2022 933983 10.3389/fmicb.2022.933983
9 Morens D.M. Fauci A.S. Emerging pandemic diseases: how we got to COVID-19 Cell 182 2020 1077 1092 10.1016/j.cell.2020.08.021 32846157
10 Thakur B. Dubey P. Benitez J. Torres J.P. Reddy S. Shokar N. A systematic review and meta-analysis of geographic differences in comorbidities and associated severity and mortality among individuals with COVID-19 Sci. Rep. 11 2021 8562 10.1038/s41598-021-88130-w 33879826
11 Host Genetics Initiative COVID-19 Mapping the human genetic architecture of COVID-19 Nature 600 2021 472 477 10.1038/s41586-021-03767-x 34237774
12 van der Made C.I. Simons A. Schuurs-Hoeijmakers J. van den Heuvel G. Mantere T. Kersten S. Presence of genetic variants among young men with severe COVID-19 JAMA 324 2020 663 673 10.1001/jama.2020.13719 32706371
13 Butler-Laporte G. Povysil G. Kosmicki J.A. Cirulli E.T. Drivas T. Furini S. Exome-wide association study to identify rare variants influencing COVID-19 outcomes: Results from the Host Genetics Initiative PLoS Genet 18 2022 e1010367 10.1371/journal.pgen.1010367
14 Zhang Q. Bastard P. COVID Human Genetic Effort Cobat A. Casanova J.-L. Human genetic and immunological determinants of critical COVID-19 pneumonia Nature 603 2022 587 598 10.1038/s41586-022-04447-0 35090163
15 Schultze J.L. Aschenbrenner A.C. COVID-19 and the human innate immune system Cell 184 2021 1671 1692 10.1016/j.cell.2021.02.029 33743212
16 Frishberg A. Kooistra E. Nuesch-Germano M. Pecht T. Milman N. Reusch N. Mature neutrophils and a NF-κB-to-IFN transition determine the unifying disease recovery dynamics in COVID-19 Cell Rep. Med 3 2022 100652 10.1016/j.xcrm.2022.100652
17 Taquet M. Dercon Q. Luciano S. Geddes J.R. Husain M. Harrison P.J. Incidence, co-occurrence, and evolution of long-COVID features: A 6-month retrospective cohort study of 273,618 survivors of COVID-19 PLoS Med 18 2021 e1003773 10.1371/journal.pmed.1003773
18 Subramanian A. Nirantharakumar K. Hughes S. Myles P. Williams T. Gokhale K.M. Symptoms and risk factors for long COVID in non-hospitalized adults Nat. Med 28 2022 1706 1714 10.1038/s41591-022-01909-w 35879616
19 Merad M. Blish C.A. Sallusto F. Iwasaki A. The immunology and immunopathology of COVID-19 Science 375 2022 1122 1127 10.1126/science.abm8108 35271343
20 Su Y. Yuan D. Chen D.G. Ng R.H. Wang K. Choi J. Multiple early factors anticipate post-acute COVID-19 sequelae Cell 185 881–895 2022 e20 10.1016/j.cell.2022.01.014
21 Schulte-Schrepping J. Reusch N. Paclik D. Baßler K. Schlickeiser S. Zhang B. Severe COVID-19 is marked by a dysregulated myeloid cell compartment Cell 182 1419–1440 2020 e23 10.1016/j.cell.2020.08.001
22 Krämer B. Knoll R. Bonaguro L. ToVinh M. Raabe J. Astaburuaga-García R. Early IFN-α signatures and persistent dysfunction are distinguishing features of NK cells in severe COVID-19 Immunity 54 2650–2669 2021 e14 10.1016/j.immuni.2021.09.002
23 Wendisch D. Dietrich O. Mari T. von Stillfried S. Ibarra I.L. Mittermaier M. SARS-CoV-2 infection triggers profibrotic macrophage responses and lung fibrosis Cell 184 6243–6261 2021 e27 10.1016/j.cell.2021.11.033
24 Georg P. Astaburuaga-García R. Bonaguro L. Brumhard S. Michalick L. Lippert L.J. Complement activation induces excessive T cell cytotoxicity in severe COVID-19 Cell 185 493–512 2022 e25 10.1016/j.cell.2021.12.040
25 Stephenson E. Reynolds G. Botting R.A. Calero-Nieto F.J. Morgan M.D. Tuong Z.K. Single-cell multi-omics analysis of the immune response in COVID-19 Nat. Med 2021 10.1038/s41591-021-01329-2
26 Rajewsky N. Almouzni G. Gorski S.A. Aerts S. Amit I. Bertero M.G. LifeTime and improving European healthcare through cell-based interceptive medicine Nature 587 2020 377 386 10.1038/s41586-020-2715-9 32894860
27 Schultze J.L. Teaching “big data” analysis to young immunologists Nat. Immunol. 16 2015 902 905 10.1038/ni.3250 26287590
28 Bonaguro L. Schulte-Schrepping J. Ulas T. Aschenbrenner A.C. Beyer M. Schultze J.L. A guide to systems-level immunomics Nat. Immunol. 23 2022 1412 1423 10.1038/s41590-022-01309-9 36138185
29 Rood J.E. Maartens A. Hupalowska A. Teichmann S.A. Regev A. Impact of the Human Cell Atlas on medicine Nat. Med 28 2022 2486 2496 10.1038/s41591-022-02104-7 36482102
30 Ziegler C.G.K. Allon S.J. Nyquist S.K. Mbano I.M. Miao V.N. Tzouanas C.N. SARS-CoV-2 Receptor ACE2 is an interferon-stimulated gene in human airway epithelial cells and is detected in specific cell subsets across tissues Cell 181 1016–1035 2020 e19 10.1016/j.cell.2020.04.035
31 Sungnak W. Huang N. Bécavin C. Berg M. Queen R. Litvinukova M. SARS-CoV-2 entry factors are highly expressed in nasal epithelial cells together with innate immune genes Nat. Med 26 2020 681 687 10.1038/s41591-020-0868-6 32327758
32 Huang N. Pérez P. Kato T. Mikami Y. Okuda K. Gilmore R.C. SARS-CoV-2 infection of the oral cavity and saliva Nat. Med 27 2021 892 903 10.1038/s41591-021-01296-8 33767405
33 Ziegler C.G.K. Miao V.N. Owings A.H. Navia A.W. Tang Y. Bromley J.D. Impaired local intrinsic immunity to SARS-CoV-2 infection in severe COVID-19 Cell 184 4713–4733 2021 e22 10.1016/j.cell.2021.07.023
34 Sterlin D. Mathian A. Miyara M. Mohr A. Anna F. Claër L. IgA dominates the early neutralizing antibody response to SARS-CoV-2 Sci. Transl. Med 2021 13 10.1126/scitranslmed.abd2223
35 Chua R.L. Lukassen S. Trump S. Hennig B.P. Wendisch D. Pott F. COVID-19 severity correlates with airway epithelium-immune cell interactions identified by single-cell analysis Nat. Biotechnol. 38 2020 970 979 10.1038/s41587-020-0602-4 32591762
36 Meng B. Abdullahi A. Ferreira I.A.T.M. Goonawardane N. Saito A. Kimura I. Altered TMPRSS2 usage by SARS-CoV-2 Omicron impacts infectivity and fusogenicity Nature 603 2022 706 714 10.1038/s41586-022-04474-x 35104837
37 Grant R.A. Morales-Nebreda L. Markov N.S. Swaminathan S. Querrey M. Guzman E.R. Circuits between infected macrophages and T cells in SARS-CoV-2 pneumonia Nature 590 2021 635 641 10.1038/s41586-020-03148-w 33429418
38 Liao M. Liu Y. Yuan J. Wen Y. Xu G. Zhao J. Single-cell landscape of bronchoalveolar immune cells in patients with COVID-19 Nat. Med 26 2020 842 844 10.1038/s41591-020-0901-9 32398875
39 Chen S.T. Park M.D. Del Valle D.M. Buckup M. Tabachnikova A. Thompson R.C. A shift in lung macrophage composition is associated with COVID-19 severity and recovery Sci. Transl. Med 14 2022 eabn5168 10.1126/scitranslmed.abn5168 36103512
40 Wauters E. Van Mol P. Garg A.D. Jansen S. Van Herck Y. Vanderbeke L. Discriminating mild from critical COVID-19 by innate and adaptive immune single-cell profiling of bronchoalveolar lavages Cell Res 0 2021 1 19 10.1038/s41422-020-00455-9
41 Melms J.C. Biermann J. Huang H. Wang Y. Nair A. Tagore S. A molecular single-cell lung atlas of lethal COVID-19 Nature 595 2021 114 119 10.1038/s41586-021-03569-1 33915568
42 Delorey T.M. Ziegler C.G.K. Heimberg G. Normand R. Yang Y. Segerstolpe Å. COVID-19 tissue atlases reveal SARS-CoV-2 pathology and cellular targets Nature 595 2021 107 113 10.1038/s41586-021-03570-8 33915569
43 Tian Y. Carpp L.N. Miller H.E.R. Zager M. Newell E.W. Gottardo R. Single-cell immunology of SARS-CoV-2 infection Nat. Biotechnol. 40 2022 30 41 10.1038/s41587-021-01131-y 34931002
44 Knoll R. Schultze J.L. Schulte-Schrepping J. Monocytes and macrophages in COVID-19 Front Immunol. 12 2021 720109 10.3389/fimmu.2021.720109
45 Wilk A.J. Lee M.J. Wei B. Parks B. Pi R. Martínez-Colón G.J. Multi-omic profiling reveals widespread dysregulation of innate immunity and hematopoiesis in COVID-19 J. Exp. Med 2021 218 10.1084/jem.20210582
46 Su Y. Chen D. Yuan D. Lausted C. Choi J. Dai C.L. Multi-omics resolves a sharp disease-state shift between mild and moderate COVID-19 Cell 183 1479–1495 2020 e20 10.1016/j.cell.2020.10.037
47 Silvin A. Chapuis N. Dunsmore G. Goubet A.-G. Dubuisson A. Derosa L. Elevated calprotectin and abnormal myeloid cell subsets discriminate severe from mild COVID-19 Cell 182 1401–1418 2020 e18 10.1016/j.cell.2020.08.002
48 Arunachalam P.S. Wimmers F. Mok C.K.P. Perera R.A.P.M. Scott M. Hagan T. Systems biological assessment of immunity to mild versus severe COVID-19 infection in humans Science 369 2020 1210 1220 10.1126/science.abc6261 32788292
49 Gressier E., Schulte-Schrepping J., Petrov L., Brumhard S., Stubbemann P., Hiller A., et al. CD4+ T cell calibration of APC optimizesantiviral CD8+ T cell immunity. Nat Immunol n.d.;in press.
50 Hadjadj J. Yatim N. Barnabei L. Corneau A. Boussier J. Smith N. Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients Science 369 2020 718 724 10.1126/science.abc6027 32661059
51 Combes A.J. Courau T. Kuhn N.F. Hu K.H. Ray A. Chen W.S. Global absence and targeting of protective immune states in severe COVID-19 Nature 591 2021 124 130 10.1038/s41586-021-03234-7 33494096
52 van der Wijst M.G.P. Vazquez S.E. Hartoularos G.C. Bastard P. Grant T. Bueno R. Type I interferon autoantibodies are associated with systemic immune alterations in patients with COVID-19 Sci. Transl. Med 13 2021 eabh2624 10.1126/scitranslmed.abh2624 34429372
53 Lee J.S. Park S. Jeong H.W. Ahn J.Y. Choi S.J. Lee H. Immunophenotyping of COVID-19 and influenza highlights the role of type I interferons in development of severe COVID-19 Sci. Immunol. 2020 5 10.1126/sciimmunol.abd1554
54 Reusch N. De Domenico E. Bonaguro L. Schulte-Schrepping J. Baßler K. Schultze J.L. Neutrophils in COVID-19 Front Immunol. 12 2021 652470 10.3389/fimmu.2021.652470
55 Aschenbrenner A.C. Mouktaroudi M. Krämer B. Oestreich M. Antonakos N. Nuesch-Germano M. Disease severity-specific neutrophil signatures in blood transcriptomes stratify COVID-19 patients Genome Med 13 2021 7 10.1186/s13073-020-00823-5 33441124
56 Wang X. Wen Y. Xie X. Liu Y. Tan X. Cai Q. Dysregulated hematopoiesis in bone marrow marks severe COVID-19 Cell Disco 7 2021 60 10.1038/s41421-021-00296-9
57 Sinha S. Rosin N.L. Arora R. Labit E. Jaffer A. Cao L. Dexamethasone modulates immature neutrophils and interferon programming in severe COVID-19 Nat. Med 28 2022 201 211 10.1038/s41591-021-01576-3 34782790
58 de Kleijn S. Langereis J.D. Leentjens J. Kox M. Netea M.G. Koenderman L. IFN-γ-stimulated neutrophils suppress lymphocyte proliferation through expression of PD-L1 PLoS One 8 2013 e72249 10.1371/journal.pone.0072249
59 Darcy C.J. Minigo G. Piera K.A. Davis J.S. McNeil Y.R. Chen Y. Neutrophils with myeloid derived suppressor function deplete arginine and constrain T cell function in septic shock patients Crit. Care 18 2014 R163 10.1186/cc14003 25084831
60 Xu J. He B. Carver K. Vanheyningen D. Parkin B. Garmire L.X. Heterogeneity of neutrophils and inflammatory responses in patients with COVID-19 and healthy controls Front Immunol. 13 2022 970287 10.3389/fimmu.2022.970287
61 Ribeiro L.S. Migliari Branco L. Franklin B.S. Regulation of innate immune responses by platelets Front Immunol. 10 2019 1320 10.3389/fimmu.2019.01320 31244858
62 Koupenova M. Livada A.C. Morrell C.N. Platelet and megakaryocyte roles in innate and adaptive immunity Circ. Res 130 2022 288 308 10.1161/CIRCRESAHA.121.319821 35050690
63 Bernardes J.P. Mishra N. Tran F. Bahmer T. Best L. Blase J.I. Longitudinal multi-omics analyses identify responses of megakaryocytes, erythroid cells, and plasmablasts as hallmarks of severe COVID-19 Immunity 53 1296–1314 2020 e9 10.1016/j.immuni.2020.11.017
64 Haas S. Hansson J. Klimmeck D. Loeffler D. Velten L. Uckelmann H. Inflammation-induced emergency megakaryopoiesis driven by hematopoietic stem cell-like megakaryocyte progenitors Cell Stem Cell 17 2015 422 434 10.1016/j.stem.2015.07.007 26299573
65 Taus F. Salvagno G. Canè S. Fava C. Mazzaferri F. Carrara E. Platelets promote thromboinflammation in SARS-CoV-2 pneumonia Arterioscler. Thromb. Vasc. Biol. 40 2020 2975 2989 10.1161/ATVBAHA.120.315175 33052054
66 Björkström N.K. Strunz B. Ljunggren H.-G. Natural killer cells in antiviral immunity Nat. Rev. Immunol. 22 2022 112 123 10.1038/s41577-021-00558-3 34117484
67 Björkström N.K. Ljunggren H.-G. Michaëlsson J. Emerging insights into natural killer cells in human peripheral tissues Nat. Rev. Immunol. 16 2016 310 320 10.1038/nri.2016.34 27121652
68 Witkowski M. Tizian C. Ferreira-Gomes M. Niemeyer D. Jones T.C. Heinrich F. Untimely TGFβ responses in COVID-19 limit antiviral functions of NK cells Nature 600 2021 295 301 10.1038/s41586-021-04142-6 34695836
69 Malengier-Devlies B. Filtjens J. Ahmadzadeh K. Boeckx B. Vandenhaute J. De Visscher A. Severe COVID-19 patients display hyper-activated NK cells and NK cell-platelet aggregates Front Immunol. 13 2022 861251 10.3389/fimmu.2022.861251
70 Guo C. Wu M. Huang B. Zhao R. Jin L. Fu B. Single-cell transcriptomics reveal a unique memory-like NK cell subset that accumulates with ageing and correlates with disease severity in COVID-19 Genome Med 14 2022 46 10.1186/s13073-022-01049-3 35501841
71 Sette A. Crotty S. Adaptive immunity to SARS-CoV-2 and COVID-19 Cell 184 2021 861 880 10.1016/j.cell.2021.01.007 33497610
72 Krovi S.H. Gapin L. Invariant natural killer T cell subsets-more than just developmental intermediates Front Immunol. 9 2018 1393 10.3389/fimmu.2018.01393 29973936
73 Ankerhold J. Giese S. Kolb P. Maul-Pavicic A. Voll R.E. Göppert N. Circulating multimeric immune complexes contribute to immunopathology in COVID-19 Nat. Commun. 13 2022 5654 10.1038/s41467-022-32867-z 36163132
74 Rajamanickam A. Kumar N.P. Pandiarajan A.N. Selvaraj N. Munisankar S. Renji R.M. Dynamic alterations in monocyte numbers, subset frequencies and activation markers in acute and convalescent COVID-19 individuals Sci. Rep. 2021
75 Zhang B. Zhang Z. Koeken V.A.C.M. Kumar S. Aillaud M. Tsay H.-C. Altered and allele-specific open chromatin landscape reveals epigenetic and genetic regulators of innate immunity in COVID-19 Cell Genom. 2022 100232 10.1016/j.xgen.2022.100232
76 Li Z. Chen X. Dan J. Hu T. Hu Y. Liu S. Innate immune imprints in SARS-CoV-2 Omicron variant infection convalescents Signal Transduct. Target Ther. 7 2022 377 10.1038/s41392-022-01237-y 36379915
77 Phetsouphanh C. Darley D.R. Wilson D.B. Howe A. Munier C.M.L. Patel S.K. Immunological dysfunction persists for 8 months following initial mild-to-moderate SARS-CoV-2 infection Nat. Immunol. 23 2022 210 216 10.1038/s41590-021-01113-x 35027728
78 Peluso M.J. Lu S. Tang A.F. Durstenfeld M.S. Ho H.-E. Goldberg S.A. Markers of immune activation and inflammation in individuals with postacute sequelae of severe acute respiratory syndrome coronavirus 2 infection J. Infect. Dis. 224 2021 1839 1848 10.1093/infdis/jiab490 34677601
79 Schultheiß C. Willscher E. Paschold L. Gottschick C. Klee B. Henkes S.-S. The IL-1β, IL-6, and TNF cytokine triad is associated with post-acute sequelae of COVID-19 Cell Rep. Med 3 2022 100663 10.1016/j.xcrm.2022.100663
80 Bellan M. Apostolo D. Albè A. Crevola M. Errica N. Ratano G. Determinants of long COVID among adults hospitalized for SARS-CoV-2 infection: A prospective cohort study Front Immunol. 13 2022 1038227 10.3389/fimmu.2022.1038227 36601115
81 Zollner A. Koch R. Jukic A. Pfister A. Meyer M. Rössler A. Postacute COVID-19 is characterized by gut viral antigen persistence in inflammatory bowel diseases Gastroenterology 163 495–506 2022 e8 10.1053/j.gastro.2022.04.037
82 Peluso M.J. Deveau T.-M. Munter S.E. Ryder D. Buck A. Beck-Engeser G. Chronic viral coinfections differentially affect the likelihood of developing long COVID J. Clin. Invest 2023 133 10.1172/JCI163669
83 Muri J. Cecchinato V. Cavalli A. Shanbhag A.A. Matkovic M. Biggiogero M. Autoantibodies against chemokines post-SARS-CoV-2 infection correlate with disease course Nat. Immunol. 2023 10.1038/s41590-023-01445-w
84 Franke C. Boesl F. Goereci Y. Gerhard A. Schweitzer F. Schroeder M. Association of cerebrospinal fluid brain-binding autoantibodies with cognitive impairment in post-COVID-19 syndrome Brain Behav. Immun. 109 2023 139 143 10.1016/j.bbi.2023.01.006 36657623
85 Son K. Jamil R. Chowdhury A. Mukherjee M. Venegas C. Miyasaki K. Circulating anti-nuclear autoantibodies in COVID-19 survivors predict long COVID symptoms Eur. Respir. J. 2023 61 10.1183/13993003.00970-2022
86 Kedor C. Freitag H. Meyer-Arndt L. Wittke K. Hanitsch L.G. Zoller T. A prospective observational study of post-COVID-19 chronic fatigue syndrome following the first pandemic wave in Germany and biomarkers associated with symptom severity Nat. Commun. 13 2022 5104 10.1038/s41467-022-32507-6 36042189
87 Sotzny F. Blanco J. Capelli E. Castro-Marrero J. Steiner S. Murovska M. Myalgic encephalomyelitis/chronic fatigue syndrome - evidence for an autoimmune disease Autoimmun. Rev. 17 2018 601 609 10.1016/j.autrev.2018.01.009 29635081
88 Wirth K.J. Scheibenbogen C. Paul F. An attempt to explain the neurological symptoms of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome J. Transl. Med 19 2021 471 10.1186/s12967-021-03143-3 34809664
89 Komaroff A.L. Lipkin W.I. Insights from myalgic encephalomyelitis/chronic fatigue syndrome may help unravel the pathogenesis of postacute COVID-19 syndrome Trends Mol. Med 2021 10.1016/j.molmed.2021.06.002
90 Wostyn P. COVID-19 and chronic fatigue syndrome: Is the worst yet to come Med Hypotheses 146 2021 110469 10.1016/j.mehy.2020.110469
91 Haffke M. Freitag H. Rudolf G. Seifert M. Doehner W. Scherbakov N. Endothelial dysfunction and altered endothelial biomarkers in patients with post-COVID-19 syndrome and chronic fatigue syndrome (ME/CFS J. Transl. Med 20 2022 138 10.1186/s12967-022-03346-2 35317812
92 Balcom E.F. Nath A. Power C. Acute and chronic neurological disorders in COVID-19: potential mechanisms of disease Brain 144 2021 3576 3588 10.1093/brain/awab302 34398188
93 Fernández-Castañeda A. Lu P. Geraghty A.C. Song E. Lee M.-H. Wood J. Mild respiratory COVID can cause multi-lineage neural cell and myelin dysregulation Cell 185 2452–2468 2022 e16 10.1016/j.cell.2022.06.008
94 Vijayakumar B. Boustani K. Ogger P.P. Papadaki A. Tonkin J. Orton C.M. Immuno-proteomic profiling reveals aberrant immune cell regulation in the airways of individuals with ongoing post-COVID-19 respiratory disease Immunity 55 542–556 2022 e5 10.1016/j.immuni.2022.01.017
95 George P.M. Reed A. Desai S.R. Devaraj A. Faiez T.S. Laverty S. A persistent neutrophil-associated immune signature characterizes post-COVID-19 pulmonary sequelae Sci. Transl. Med 14 2022 eabo5795 10.1126/scitranslmed.abo5795 36383686
96 Cheon I.S. Li C. Son Y.M. Goplen N.P. Wu Y. Cassmann T. Immune signatures underlying post-acute COVID-19 lung sequelae Sci. Immunol. 6 2021 eabk1741 10.1126/sciimmunol.abk1741 34591653
97 Santa Cruz A. Mendes-Frias A. Azarias-da-Silva M. André S. Oliveira A.I. Pires O. Post-acute sequelae of COVID-19 is characterized by diminished peripheral CD8+β7 integrin+ T cells and anti-SARS-CoV-2 IgA response Nat. Commun. 14 2023 1772 10.1038/s41467-023-37368-1 36997530
98 Warnat-Herresthal S. Schultze H. Shastry K.L. Manamohan S. Mukherjee S. Garg V. Swarm Learning for decentralized and confidential clinical machine learning Nature 594 2021 265 270 10.1038/s41586-021-03583-3 34040261
99 Schultze J.L. Büttner M. Becker M. Swarm immunology: harnessing blockchain technology and artificial intelligence in human immunology Nat. Rev. Immunol. 22 2022 401 403 10.1038/s41577-022-00740-1 35624333
|
PMC010xxxxxx/PMC10201330.txt |
==== Front
Transplant Proc
Transplant Proc
Transplantation Proceedings
0041-1345
1873-2623
Elsevier Inc.
S0041-1345(23)00325-1
10.1016/j.transproceed.2023.05.009
Article
Clinical Course, Nosocomial, and Opportunistic Infections Among Kidney Transplant Recipients with COVID-19: A Retrospective Single Center Study
Escalante Elias Jatem ⁎
Rodríguez Jorge González
Salas Jacqueline Del Carpio
Castañeda Zaira
Conde María Luisa Martín
Servicio de Nefrología, Hospital Universitari Arnau de Vilanova, Lleida, Catalonia, Spain
⁎ Address correspondence to Elias Jatem Escalante, MD, Servicio de Nefrología, Av Alcalde Rovira Roure, 80, 25198 Lleida, Spain. Tel: +034)973248100.
22 5 2023
22 5 2023
© 2023 Elsevier Inc. All rights reserved.
2023
Elsevier Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
We report the results of an observational study, analyzing the clinical course of kidney transplant patients hospitalized for COVID-19 and comparing it with a control to determine if outcomes, nosocomial, and opportunistic infections were different between groups.
Methods
An observational, retrospective, case-control, single-center study, including a group of kidney transplant adults diagnosed with COVID-19, from March 2020 to April 2022. Transplant patients hospitalized for COVID-19 comprised the cases. The control group consisted of non-transplanted adults, without immunosuppressive treatment, hospitalized for COVID-19, and matched by age, sex, and month at diagnosis of COVID-19. Study variables were collected, including demographic/clinical, epidemiologic, clinical/biological at diagnosis, evolutive, and outcome variables.
Results
Fifty-eight kidney transplant recipients were included. Thirty required hospital admission. Ninety controls were included. Transplant recipients had a higher frequency of intensive care unit (ICU) admission, ventilatory support, and death. The relative risk for death was 2.45. When adjusted by baseline estimated glomerular filtration rate (eGFR) and comorbidity, only the risk for opportunistic infection remained high. Variables independently associated with death were dyslipidemia, eGFR at admission, MULBSTA score, and ventilatory support. Pneumonia by Klebsiella oxytoca was the most frequent nosocomial infection. Pulmonary aspergillosis was the most frequent opportunistic infection overall. Pneumocystosis and cytomegalovirus colitis were more frequent among transplant patients. The relative risk for opportunistic infection in this group was 1.88. Baseline eGFR, serum interleukin 6 level, and coinfection were independently associated with it.
Conclusions
Evolutive course of COVID-19 requiring hospitalization in renal transplant recipients was primarily determined by comorbidity and baseline kidney function. At equal comorbidity and renal function, there were no differences in mortality, ICU admission, nosocomial infection, and hospital stay. However, the risk for opportunistic infection remained high.
==== Body
pmcCOVID-19, caused by SARS-CoV-2, was officially declared a pandemic by the World Health Organization in 2020. Despite the development of vaccines and novel antiviral therapies, it has not yet been eradicated [1]. Considering the viral mutation rate with the appearance of new variants with variable profiles of contagiousness, virulence, and resistance to active immunization and antiviral treatment, there is uncertainty regarding its duration and evolution.
Older patients and those affected by multiple comorbidities are at a higher risk of infection-related death [2,3], especially in immunocompromised individuals and those affected by malignant neoplasms [4].
The mortality rate of COVID-19 in kidney transplant recipients has been reported to be 20% to 40% [5], [6], [7], [8], a comparatively greater rate than that reported in the general population (10%-15%) [9], [10], [11]. This difference is mainly explained by the immunocompromised status of these patients; it is considered advisable to temporarily lower immunosuppression therapy intensity in high-risk transplant patients who contract the infection. However, this may result in a higher risk for rejection [1].
In relation to the clinical behavior of COVID-19 in kidney transplant patients, diverse studies have been published. In general, a series of baseline epidemiologic and clinical factors are associated with a higher risk of death, intensive care unit (ICU) admission, and ventilatory support; among them are age, basal comorbidity, being a recipient of a graft from a cadaveric donor [12], [13], [14], [15], [16], time since transplant ≤6 months, respiratory symptoms, pneumonia, and acute kidney injury (AKI) [17], [18], [19], [20], [21]. An analysis from the Spanish Registry of Kidney Transplants published in 2021 [21] found that patients older than 65 years who were recipients of kidney grafts in the 6 months before SARS-CoV-2 infection experienced worse survival and mortality than other transplant patients. Most studies, in this regard, focus on transplant populations only; studies with control groups conformed of non-transplant patients are less numerous. Among the latter, Aziz et al, in a preliminary report published in 2020, found that kidney transplant recipients had relative risks (RRs) for AKI and ICU admission of 15.4% and 34.1%, respectively, compared with RRs of 13.3% and 3.3% in non-transplant patients (P < .001) [11]. Current available data come mainly from registries from the first year of the pandemic when active immunization was not widely available, and certain antiviral therapies were not backed up by solid clinical evidence. Data concerning the rates of nosocomial infections (NI) and opportunistic infections (OI) in this context and in these patients is relatively limited. Both NI and OI are important factors affecting prognosis and hospital stay in the general population.
We report the results of an observational single-center study designed with the following objectives: (1) to analyze the clinical course of kidney transplant patients with diagnosis of COVID-19, those requiring hospitalization, and compare it with a control group of non-transplant patients; and (2) to determine if mortality, ICU admission, need for ventilatory support, NI, OI, hospital, and ICU stay differed between groups and which variables (clinical, epidemiologic, biochemical) were associated with these outcomes.
Patients and Methods
This is an observational, retrospective, case-control, single-center study, including a group of kidney transplant adults with diagnosis of COVID-19 based on suggestive symptomatology and confirmed through SARS-CoV-2 determination by polymerase chain reaction (either from a blood sample or nasopharyngeal swab) who were attended in Hospital Universitari Arnau de Vilanova (HUAV), Lleida, Spain, from March 2020 to April 2022. In this group, patients requiring hospitalization for COVID-19 comprised the case group. A control group was also included, consisting of adults who were non-recipients of solid or hematopoietic organs, without active immunosuppressive treatment (for any cause) in the last 6 months before inclusion in the study, with diagnoses of COVID-19 requiring hospital admission; this group was matched with the cases by age, sex, and month at diagnosis of COVID-19 in a proportion of 1:3.
In both groups, the digital clinical records from HUAV and Primary Health Service were reviewed. The study variables were collected, including demographic-clinical (including Charlson´s comorbidity index), epidemiologic, clinical-biological at diagnosis (including Horowitz index and MULBSTA score), evolutive, and outcomes (coinfection at hospital admission, treatment prescribed for COVID-19, AKI at admission, AKI during follow-up, shock, ICU admission, ventilatory support, renal replacement therapy, frequency of NI and OI, death, cause of death, serum creatinine, and estimated glomerular filtration rate at the end of follow-up and sequels). In transplant patients, de novo positivity for donor-specific antibodies after discharge was recorded.
Coinfection at diagnosis was defined as an infectious disease caused by any pathogenic agent different from SARS-CoV-2 confirmed or suspected at the time of COVID-19 diagnosis.
Nosocomial infection was defined as any infection suspected or confirmed during hospitalization that was not present, even in its incubation stages, at the time of admission.
Opportunistic infection was defined as any infectious disease caused by pathogens that usually do not produce clinically significant infections in immunocompetent individuals.
The AKI definitions used were those established by the Kidney Disease: Improving Global Outcomes guidelines, with severity staging done according to the Acute Kidney Injury Network [22].
The estimated glomerular filtration rate (eGFR) was calculated by the Chronic Kidney Disease Epidemiology Collaboration 2009 formula [23].
Statistical Analysis
Categorical variables were expressed in percentages, and continuous variables were expressed as mean and SD for normal distribution variables or as median and IQR. Statistical significance of the differences when comparing categorical variables was calculated by the Chi-square or the Fisher test. When comparing categorical with continuous variables, the t test or Mann-Whitney U test was used. Multivariable analysis was performed using a step-by-step regression logistics model. Survival analysis was performed with Kaplan-Meier curves, and the log-rank test was used to determine statistical significance. The statistical programs used were SPSS version 20.0 (IBM SPSS, Inc, Armonk, NY, United States) and MedCalc (MedCalc Software, Ltd, Ostend, West Flanders, Belguim).
The study was reviewed and approved by the bioethical committee of the center.
Results
From March 2020 to March 2022, 58 kidney transplant recipients were diagnosed with COVID-19. Most of the patients had received only one graft (94.8%). At diagnosis, 6 patients had active neoplasms (2 cases of breast cancer, 2 of skin cancer, 1 case of lung cancer, and 1 of a GIST), of which 2 were on active oncological treatment. Thirty-three patients, at diagnosis, had received at least 1 dose of a vaccine against SARS-CoV-2, of which, 39.7% had received 3 doses. Of the 58 patients with COVID-19, 30 (50.84%) required hospital admission (Fig 1 ).Fig 1 NI, nosocomial infection; OI, opportunistic infections.
Fig 1
In a proportion of 1:3 matched by age, sex, and month of diagnosis, 90 controls were included.
Among the cases (transplant recipients), 10 (33.3%) had concomitant infection at admission, 10 (33.33%) required ICU admission, 6 (20%) developed NI, 8 (26.6%) OI, and 9 (30%) died. In the control group, 23 (25.8%) had coinfection at admission, 9 (10.2%) were admitted to ICU, 8 (9%) developed NI, 3 (3.4%) OI, and 11 (12.6%) died (Fig 1).
Fig 2 represents the frequencies for both diagnosis of and hospitalization for COVID-19 in kidney transplant populations between March 2020 and March 2022. Peaks of hospitalization frequency for COVID-19 were observed between epidemiologic months 6 to 8 (June and July 2020), 11 to 12 (November and December 2020), and 24 to 25 (December 2021 to January 2022) of the pandemic.Fig 2 New cases and hospitalizations by COVID-19 in kidney transplant population per month. Month 1 corresponds to January 2020. Log Rank = 1.22, P = .25.
Fig 2
At diagnosis, all transplant patients received tacrolimus as an immunosuppressive maintenance treatment. In all patients, antiproliferative agents were suspended (mammalian target of rapamycin [mTOR] inhibitors or mycophenolic acid analogs (MAA)), tacrolimus dose was adjusted to target levels of 3 to 5 ng/mL, and prednisone was increased to 10 to 20 mg/d. In those patients who required dexamethasone treatment, prednisone was suspended.
Kidney transplant patients with COVID-19 admitted to the hospital had less transplant age, worse baseline kidney function parameters, received higher doses of prednisone as maintenance therapy, less frequency of mTOR inhibitors prescription, worse kidney function parameters at diagnosis, higher reactive C protein levels, MULBSTA score, and lower Horowitz index and serum hemoglobin. This group of patients experienced longer times from symptom onset to diagnosis (TbStDx) (Table 1 ).Table 1 Demographics, Clinical Variables, and Hospitalization by COVID-19 in Transplant Patients
Table 1 Hospitalized
n = 30 Non-hospitalized
n = 28 P Value
Age 61.93 ± 14.11 56.57 ± 14.2 .15
Sex n (%)
Male 18 (60) 21 (75) .21
Female 12 (40) 7 (25)
Transplant age (mo) 50 (187-10) 68 (333-9) .021†
Baseline creatinine (mg/dL) 1.5 (3.2-0.96) 1.15 (15-0.6) .003†
Baseline eGFR (mL/min) 44.33 ± 16.48 67.72 ± 24.63 < .001†
Hypertension n (%) 28 (93.3) 24 (85.7) .41
Active smoking n (%) 2 (6.7) 2 (50) .66
Dyslipidemia n (%) 22 (73.3) 26 (92.9) .051
Diabetes mellitus n (%) 18 (60) 10 (35.7) .056
Obesity n (%) 4 (13.3) 5 (17.9) .45
Ischemic heart disease n (%) 6 (20) 4 (14.3) .41
Peripheral vascular disease n (%) 5 (16.7) 6 (21.4) .44
Cerebral vascular disease n (%) 2 (6.7) 5 (17.9) .18
COPD n (%) 3 (10) 0 (0) .13
Active neoplasm 4 (13.3) 2 (7.1) .36
Vaccine doses received n (%)
None 14 (46.7) 11 (39.3) .33
1 1 (3.3) 0 (0)
2 6 (20) 3 (10.7)
3 9 (30) 14 (50)
Anti-SARS-CoV-2 titer at dx (BAU/mL) 0 (2081.2-0) 0 (2084-0) .004†
Type of vaccine
ARNm Pfizer 4 (13.3) 2 (7.1) .45
ARNm Moderna 12 (40) 15 (53.6)
Immunosuppressive treatment at diagnosis n (%)
Prednisone 30 (100) 24 (85.7) .048†
AMMF 26 (87.7) 19 (67.9) .08
Anti-calcineurinics 30 (100) 28 (100) N/A
mTOR inhibitors 2 (6.7) 8 (28.6) .03†
Prednisone dose (mg/d) 5 (5-5) 5 (25-0) .15
TbStDx (d)* 4 (10-0) 0 (4-0) < .001†
Charlson comorbidity index 4 (11-2) 4 (11-2) .023†
Creatinine at dx (mg/dL) 1.79 (1.34-0.8) 1.21 (2.73-0.75) .004†
eGFR at dx (mL/min) 39.14 ± 20.27 63.58 ± 24.7 .001†
IL-6 (pg/mL) 39.85 (1682-2.6) 10.2 (33.4-3.1) .028†
Reactive C protein (mg/l) 79.25 (306-1.3) 4.8 (21.5-1) < .001†
Leukocyte count (cells*10³/mm³) 7.55 ± 4.82 7.01 ± 2.73 .65
Hemoglobin (gr/dL) 12.48 ± 2.27 14.45 ± 2.01 .003†
Platelet count (cells*10³/mm³) 197.66 ± 127.31 198.35 ± 58.66 .98
Lactate dehydrogenase (U/L) 632 ± 227.76 412.33 ± 118.62 .056
D-dimer (ng/mL) 237.5 (3638-100) 208 (788-130) .47
MULBSTA score 11 (19-4) 4 (8-0) < .001†
Horowitz index 274.2 ± 109.51 396 ± 39.59 .04†
Logistic Regression Model.
B E.T. P Value Exp(B) IC 95%
Transplant age (mo) –0.04 0.02 .046 0.96 0.92-0.99
TbStDx (d) 1.5 0.63 .018 4.52 1.29-15.83
Baseline eGFR (mL/min) –0.071 0.033 .029 0.93 0.87-0.993
R2 Nagelkerke: 0.79.
MAA, mycophenolic acid analogs…; B, unstandarddized regression weight COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; dx, diagnosis; E.T Standard Error; CI, confidence interval; IL, interleukin; mTOR, mammalian target of rapamycin; TbStDx, time between symptom onset to diagnosis.
⁎ TbStDx.
† P < .05.
The circulating anti-SARS-CoV-2 IgG antibody titer among transplant patients was lower (Table 1).
Upon multivariate analysis, the variables independently associated with hospital admission were transplant age, TbStDx, and the eGFR at diagnosis (Table 1).
Table 2 summarizes the clinical, epidemiologic, biochemical, and evolutive features of both cases and controls. Cases presented significantly higher cardiovascular comorbidity and Charlson index scores and lower baseline kidney function. Transplant patients had lower anti-SARS-CoV-2 IgG antibody titer and TbStDx with predominating respiratory symptomatology. At admission, transplant patients had lower eGFR, lower platelet count, and greater AKI frequency, regardless of its severity. These patients received tocilizumab and remdesevir more frequently.Table 2 Demographics, Clinical Features, and Outcomes Between Transplant and Non-transplant Recipients Hospitalized for COVID-19
Table 2 Cases
n = 30 Controls
n = 90 P Value
Age 61.93 ± 14.11 63.59 ± 13.79 .57
Sex n (%)
Male 18 (60) 52 (57.8) .83
Female 12 (40) 38 (42.2)
Baseline creatinine (mg/dL) 1.5 (3.2-0.96) 0.84 (8.9-0.35) .001†
Baseline eGFR (mL/min) 44.33 ± 16.48 78.35 ± 25.89 < .001†
Hypertension n (%) 28 (93.3) 50 (56.2) < .001†
Active smoking n (%) 2 (6.7) 6 (6.7) .64
Dyslipidemia n (%) 22 (73.3) 36 (40) .002†
Diabetes mellitus n (%) 18 (60) 21 (23.3) < .001†
Obesity n (%) 4 (13.3) 19 (21.1) .34
Ischemic heart disease n (%) 6 (20) 9 (10) .15
Peripheral vascular disease n (%) 5 (16.7) 13 (14.4) .76
Cerebral vascular disease n (%) 2 (6.7) 11 (12.2) .39
COPD n (%) 3 (10) 16 (17.8) .31
Type of vaccine n (%)
ARNm Pfizer 4 (13.3) 23 (25.55) .004†
ARNm Moderna 12 (40) 10 (11.11)
Jcovden (Janssen) 0 (0) 3 (3.33)
ChAdOx1 (AstraZeneca) 0 (0) 1 (1.1)
Vaccine doses received n (%)
None 14 (46.7) 55 (61.1) .50
1 1 (3.3) 5 (5.6)
2 6 (20) 11 (12.2)
3 9 (30) 18 (20)
4 0 (0) 1 (1.1)
Anti-SARS-CoV-2 titer at dx (BAU/mL) 0 (2081.2-0) 218 (2081-0) .01†
TbStDx (d)* 4 (10-0) 5 (45-0) .04†
ICU admission n (%) 10 (33.3) 9 (10.2) .003†
Charlson comorbidity index 4 (11-2) 2 (10-0) .001†
Creatinine at dx (mg/dL) 1.79 (9.83-0.84) 0.86 (8.2 -0.31) .001†
eGFR at dx (mL/min) 39.14 ± 20.27 74.31 ± 29.56 .02†
IL-6 (pg/mL) 39.85 (1682-2.6) 29.6 (744.7-1) .266
Reactive C protein (mg/l) 102.99 ± 78.82 107.77 ± 91.3 .93
Leukocyte count (cells*10³/mm³) 6.88 (27.79-2.99) 7.54 (27.25-2.76) .15
Hemoglobin (gr/dL) 12.48 ± 2.27 12.95 ± 1.96 .27
Platelet count (cells*10³/mm³) 170 (777-50) 220.5 (724-74) .008†
Lactate dehydrogenase (U/L) 632 ± 227.76 619.58 ± 299.41 .84
D-dimer (ng/mL) 237.5 (3638-100) 272 (46768-80) .55
MULBSTA score 10.76± 3.74 9.21 ± 4.17 .08
Horowitz index 274.20 ± 109.51 271.07 ± 93.97 .23
AKI at admission n (%) 12 (40) 16 (18) .014†
AKIN n (%)
1 6 (42.9) 10 (62.5) .35
2 4 (28.6) 3 (18.8)
3 2 (14.3) 3 (18.8)
Shock n (%) 6 (20) 5 (5.7) .02†
Ventilatory support/O2 therapy n (%) 15 28
High flow nasal cannula O2 6 (20) 24 (27) < .001†
Non-invasive mechanical ventilation 3 (10) 4 (4.5)
Invasive mechanical ventilation 6 (20) 0 (0)
AKI during follow-up n (%) 8 (26.7) 7 (7.9) .007†
AKIN during follow-up n (%)
1 1 (10) 4 (57.1) .16
2 3 (30) 1 (14.3)
3 4 (40) 2 (28.6)
CVVHD/HD n (%) 3 (10) 2 (2.3) .07
Concomitant infection n (%) 10 (33.3) 23 (25.8) .42
Respiratory 5 (50) 16 (69.5) .15
Urinary 5 (50) 3 (13)
Skin 0 (0) 1 (4.4)
Abdominal 0 (0) 2 (8.7)
Joint 0 (0) 1 (4.4)
COVID-19 pharmacologic therapy n (%)
Tocilizumab 15 (50) 24 (28.2) .03†
N° of doses 1.13 ± 0.5 0.87 ± 0.5 .1
Dexamethasone 27 (90) 66 (75) .06
Initial dose (md/d) 6 (15-0) 7.2 (8-0) .004†
Remdesivir 11 (36.7) 10 (11.4) .002†
Hydroxychloroquine 3 (10) 3 (3.4) .15
Nosocomial infection n (%) 6 (20) 8 (9) .10
Respiratory 4 (66.6) 4 (50) .2
Septicemia 2 (33.4) 2 (25)
Catheter related septicemia 0 (0) 1 (12.5)
Skin 0 (0) 1 (12.5)
Opportunistic infection n (%) 8 (26.7) 3 (3.4) .001†
Pulmonary aspergillosis 3 (37.5) 2 (66.6) .001†
Candidiasis 1 (12.5) 1 (33.4)
Invasive CMV 2 (25) 0 (0)
Pneumocystosis 2 (25) 0 (0)
Final creatinine (mg/dL) 1.4 (4.8-0.56) 0.77 (8-0.31) .001†
Final eGFR (mL/min) 50.6 ± 22.48 78.85 ± 29.73 < .001†
Hospitalization time (d) 11.5 (61-6) 7 (9-1) .001†
ICU time (d) 17.7 ± 9.84 14.4 ± 17.76 .61
Exitus n (%) 9 (30) 11 (12.6) .02†
Sequels n (%) 2 (9.5) 7 (9) .61
AKI, acute kidney injury; AKIN, Acute Kidney Injury Network; CMV, cytomegalovirus; COPD, chronic obstructive pulmonary disease; CVVHD, continuous veno-venous hemodialysis; eGFR, estimated glomerular filtration rate; HD, hemodialysis; ICU, intensive care unit; IL, interleukin; TbStDx, time between symptom onset to diagnosis.
⁎ TbStDx.
† P < .05.
Compared with the controls, transplant recipients had worse outcomes, with higher frequencies of AKI during follow-up, shock, ICU admission, and the need for invasive and non-invasive ventilatory support. There were no significant differences in ICU stay, but transplant patients had longer hospital stays. Opportunistic infection and mortality were greater. The RR for death in transplant patients was 2.45 (confidence interval 95%; 1.12-5.34).
At the end of the follow-up, 28 (93.3%) cases featured functional grafts. None developed de novo donor-specific antibodies.
Seeing that comorbidity and baseline eGFR were worse among the cases, we matched the cases and controls by baseline eGFR, Charlson index, frequency of high blood pressure, diabetes, and dyslipidemia in a proportion of 1:1 to obtain 20 matched pairs (Table 3 ).Table 3 Demographics, Clinical Features, and Outcomes Between Transplant and Non-transplant Recipients Hospitalized for COVID-19 after Adjustment for Baseline eGFR and Comorbidities
Table 3 Cases
n = 20 Controls
n = 20 P Value
Transplant age (mo) 63 ± 47.85 - -
TbStDx (d)* 4.3 ± 3.01 5.47 ± 3.93 .17
ICU admission n (%) 6 (30) 2 (10) .11
Creatinine at dx (mg/dL) 2.19 (9.83-0.84 1.41 (8.9-0.5) .38
eGFR at dx (mL/min) 44.5 ± 21.63 44.59 ± 23.69 .88
IL-6 (pg/mL) 91.05 (1682-2.6) 33.4 (210-4.4) .38
Reactive C protein (mg/l) 96.7 ± 80.12 163.06 ± 109.56 .03†
Leukocyte count (cells*10³/mm³) 8.35 ± 5.54 10.86 ± 6.14 .18
Hemoglobin (gr/dL) 12.96 ± 2.27 12.64 ± 2.12 .89
Platelet count (cells*10³/mm³) 212.3 ± 153.02 229.5 ± 95.93 .67
Lactate deshidrogenase (U/L) 620.65 ± 254.45 605.28 ± 412.41 .12
D-dimer (ng/mL) 237 (2726-100) 352 (46768-130) .06
MULBSTA score 9.9 ± 3.66 10.23 ± 4.22 .38
Horowitz index 287.84 ± 100.33 269.38 ± 106.56 .62
AKI at admission n (%) 7 (35) 10 (50) .33
Shock n (%) 3 (15) 2 (10) .63
Ventilatory support/O2 therapy n (%) 10 7
High flow nasal cannula O2 4 (40) 6 (85.71) .18
Non-invasive mechanical ventilation 3 (30) 1 (14.28)
Invasive mechanical ventilation 3 (30) 0 (0)
AKI during follow-up n (%) 5 (25) 6 (30) .5
CVVHD/HD n (%) 1 (5) 2 (10) .54
Concomitant infection n (%) 6 (30) 6 (30) .63
COVID-19 pharmacologic therapy n(%)
Tocilizumab 10 (50) 3 (15) .02†
Dexamethasone 17 (85) 13 (65) .14
Initial dose (md/d) 6 (15-0) 6 (7.5-0) .61
Remdesivir 8 (40) 3 (15) .07
Hydroxychloroquine 3 (15) 1 (5) .29
Nosocomial infection n (%) 4 (20) 3 (15) .67
Opportunistic infection n (%) 5 (25) 1 (5) .04†
Final creatinine (mg/dL) 1.4 (4.8-0.7) 1.2 (8-0.7) .37
Final eGFR (mL/min) 53.35 ± 23.83 52.47 ± 29.8 .3
Hospitalization time (d) 13.4 ± 7.01 11.65 ± 11.43 .05
Exitus n (%) 6 (30) 7 (35) .73
Sequels n (%) 1 (7.1) 0 (0) .32
AKI, acute kidney injury; CVVHD, continuous veno-venous hemodialysis; eGFR, estimated glomerular filtration rate; HD, hemodialysis; ICU, intensive care unit; IL, interleukin; TbStDx, time between symptom onset to diagnosis.
⁎ TbStDx.
† P < .05.
When adjusted by these variables, at admission, transplant patients presented higher levels of interleukin (IL) 6, lower levels of D-dimer, and higher frequencies of tocilizumab prescription. Opportunistic infection frequency continued to be significantly higher. No differences were observed in mortality, ICU admission, hospital stay, or NI (Table 3). The relative risk for death was 0.89 (CI 95%; 0.44-1.77), losing its statistical significance.
Mortality
Figure 3 shows the survival curves for the cases vs controls without adjustment by comorbidity and baseline eGFR and after adjustment. Because hospital/ICU stays were not different between groups (time to death), even without adjustment, the differences were not statistically significant (Figure 3A; Log Rank: 1.22; P = .25).Fig 3 Survival between transplant and non-transplant recipients hospitalized for COVID-19. (A) Kaplan-Meier survival curve in unadjusted group. (B) Survival curve after adjustment for baseline estimated glomerular filtration rate, comorbidities, and Charlson score. Tx, transplant. Log Rank = 0.00, P = .99.
Fig 3
Figure 4 shows the causes of death among the study groups. There were no significant differences between the groups. Overall, the most frequent cause of death was due to SARS-CoV-2 infection or NI/OI.Fig 4 Causes of death in the study groups. Tx, transplant. P = .075.
Fig 4
Variables associated with death in all the study groups are shown in Table 4 . Age, hypertension, dyslipidemia, diabetes, brain vascular disease, Charlson index, and baseline eGFR were associated with higher mortality. Patients with lower eGFR, AKI, greater LDH, D-dimer, MULBSTA score, and lower Horowitz index had higher mortality at admission. The presence of coinfection at admission, NI, OI, ICU admission, invasive and non-invasive ventilatory support, septic shock, and the need for renal replacement therapy were associated significantly with death (Table 4). On multivariate analysis, variables independently associated with death were dyslipidemia, eGFR at admission, higher MULBSTA score, and invasive and non-invasive ventilatory support (Table 4).Table 4 Demographics, Clinical Variables, and Death in the Study Groups
Table 4 Exitus
n = 20 Non-exitus
n = 100 P Value
Age 70.15 ± 14.16 61.96 ± 13.47 .016†
Sex n (%)
Male 10 (50) 59 (59) .47
Female 10 (50) 41 (41)
Group n (%)
Tx patients 9 (45) 21 (21) .029†
Non-Tx patients 11 (55) 79 (79)
Transplant age (mo) 48 (187-19) 51 (164-10) .92
Baseline creatinine (mg/dL) 1.4 (2.5-0.49) 0.88 (8.9-0.35) .003†
Baseline eGFR (mL/min) 49.89 ± 23.35 73.74 ± 27.33 < .001†
Hypertension n (%) 18 (90) 59 (59) .014†
Active smoking n (%) 0 (0) 6 (6) .25
Dyslipidemia n (%) 17 (85) 41 (41) < .001†
Diabetes mellitus n (%) 11 (55) 28 (28) .024†
Obesity n (%) 4 (20) 18 (18) .88
Ischemic heart disease n (%) 4 (20) 11 (11) .29
Peripheral vascular disease n (%) 2 (10) 16 (16) .46
Cerebral vascular disease n (%) 6 (30) 6 (6) .001†
COPD n (%) 3 (15) 16 (16) .86
Active neoplasm 3 (15) 9 (9.3) .44
Vaccine doses received n (%)
None 9 (45) 58 (58) .6
1 1 (5) 5 (5)
2 5 (25) 12 (12)
3 5 (25) 21 (21)
4 0 (0) 1 (1)
Anti-SARS-CoV-2 titer at dx (BAU/mL) 0 0 (2081.2-0) .12
TbStDx (d)* 3.5 (10-0) 5 (45-0) .04†
Charlson comorbidity index 5 (8-0) 3 (11-0) .001†
Creatinine at dx (mg/dL) 2.1 ± 1.18 2.26 ± 8.34 .08
eGFR at dx (mL/min) 37.95 ± 22.91 70.48 ± 29.93 < .001†
IL-6 (pg/mL) 174.9 (1682-40.8) 23.2 (1004.6-1) .002†
Reactive C Protein (mg/l) 150.01 ± 99.23 96.79 ± 83.29 .013
Leukocyte count (cells*10³/mm³) 7.8 (27.25-3.27) 7.05 (27.79-2.76) .32
Hemoglobin (gr/dL) 12.17 ± 2.46 12.95 ± 4.46 .12
Platelet count (cells*10³/mm³) 196.5 (381-112) 202 (777-50) .88
Lactate dehydrogenase (U/L) 820.21 ± 332.26 590.9 ± 257.05 0.004†
D-dimer (ng/mL) 346.5 (46768-148) 251 (6574-80) .008†
MULBSTA score 13.35 ± 2.37 8.94 ± 4 < .001†
Horowitz index 202.62 ± 115.06 285.41 ± 89.6 .002†
AKI at admission n (%) 13 (65) 15 (15) < .001†
AKIN n (%)
1 4 (30.8) 12 (80) .013†
2 4 (30.8) 3 (20)
3 5 (38.4) 0 (0)
ICU admission 9 (45) 10 (10) .001†
Shock n (%) 8 (40) 3 (3) < .001†
Ventilatory support/O2 therapy n (%) 16 26
High flow nasal cannula O2 9 (56.2) 20 (76.9) < .001†
Non-invasive mechanical ventilation 3 (18.8) 4 (15.4)
Invasive mechanical ventilation 4 (25) 2 (7.7)
AKI during follow-up n (%) 9 (45) 6 (6) < .001†
AKIN during follow-up n (%)
1 1 (11.1) 4 (70) .02†
2 2 (22.2) 2 (30)
3 6 (66.7) 0 (0)
CVVHD/HD n (%) 3 (15) 2 (2) .01†
Concomitant infection n (%) 11 (55) 22 (22) .003†
Respiratory 5 (45.45) 17 (77.37) .005†
Urinary 4 (36.37) 3 (13.63)
Skin 0 (0) 1 (4.5)
Abdominal 1 (9.09) 1 (4.5)
Joint 1 (9.09) 0 (0)
COVID-19 pharmacologic therapy n (%)
Tocilizumab 9 (45) 29 (29) .29
N° of doses 1 ± 0.63 0.94 ± 0.48 .56
Dexamethasone 15 (75) 77 (77) .66
Initial dose (md/d) 6 (15-0) 6 (8-0) .96
Remdesivir 3 (15) 18 (18) .7
Hydroxychloroquine 2 (10) 4 (4) .27
Nosocomial infection n (%) 5 (25) 9 (9) .04†
Opportunistic infection n (%) 6 (30) 5 (5) .001†
Final creatinine (mg/dL) 1.87 (4.8-0.56) 0.8 (5.8-0.31) .003†
Final eGFR (mL/min) 42.44 ± 27.74 77.06 ± 27.91 < .001†
Hospitalization time (d) 12 (34-1) 7 (9-1) .02†
ICU time (d) 17.75 ± 8.44 14.92 ± 17.13 .27
Functioning graft at death n (%) 7 (77.8) 21 (100) .08
Logistic Regression Model.
B E.T. P Value Exp(B) IC 95%
Dyslipidemia –4.77 2.18 .029 0.008 0.00-0.61
eGFR at dx (mL/min) –0.051 0.025 .041 0.95 0.9-0.99
MULBSTA 0.49 0.2 .016 1.63 1.09-2.43
Ventilatory support 1.32 0.64 .041 3.75 1.05-13.34
R2 Nagelkerke: 0.71.
AKI, acute kidney injury; AKIN, Acute Kidney Injury Network; B, …; COPD, chronic obstructive pulmonary disease; CVVHD; dx, diagnosis; eGFR, estimated glomerular filtration rate; E.T stantdard Error, …; HD, hemodialysis; ICU, intensive care unit; CI, confidence interval; IL, interleukin; TbStDx, time between symptom onset to diagnosis; Tx, transplant.
⁎ TbStDx.
† P < .05.
Among transplant patients with coinfection at admission, a cause for infection was identified in 6 of 10 cases; in the control group, 6 of 9 cases included identified causes for infection. No significant differences were observed between groups (Table 5 ). Urinary and respiratory tract infections were the most frequent coinfections at admission (Table 2); the pathogens isolated in most cases were Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa.Table 5 Etiology of Coinfection at COVID-19 Diagnosis and Transplant Versus Non-transplant Patients
Table 5Pathogen TX patients n (%) Non-Tx patients n (%) P Value
Mycobacterium tuberculosis 0 (0) 1 (16.66) .06
Escherichia coli 2 (33.34) 0 (0)
Haemophylus influenzae 0 (0) 1 (16.66)
Klebsiella pneumoniae 1 (16.66) 0 (0)
Pseudomonas aeruginosa 1 (16.66) 1 (16.66)
Staphylococcus aureus 1 (16.66) 1 (16.66)
Streptococcus pyogenes 0 (0) 1 (16.66)
Herpes simplex virus 0 (0) 1 (16.66)
Legionella pneumophila 1 (16.66) 0 (0)
Tx, transplant.
Nosocomial Infection
The nosocomial infection diagnosed with the most frequency was pneumonia (Table 2), and the most frequent etiologic agent was Klebsiella oxytoca.
Table 6 shows the variables significantly associated with NI. Among patients with NI, the baseline eGFR was lower, they had more frequency of coinfection, higher levels of IL-6, AKI, MULBSTA score, ICU admission, ventilatory support, mortality, and hospital stay. The RR for NI in transplant patients was 1.66 CI 95%; 0.6-4.5).Table 6 Variables Significantly Associated with NIs in the Study Groups
Table 6 NI
n = 14 Non-NI
n = 106 P Value
Baseline creatinine (mg/dL) 1.5 (8.9-0.6) 0.95 (8-0.35) .03*
Baseline eGFR (mL/min) 56.79 ± 27.1 71.77 ± 27.88 .04*
IL-6 (pg/mL) 89.3 (1682-3.2) 24.2 (1004.6-1) .01*
D-dimer 412 (46768-168) 251.5 (19313-80) .01*
MULBSTA score 12.84 ± 3.97 9.21 ± 3.93 .002*
ICU admission 9 (64.3) 10 (9.43) < .001*
Shock n (%) 6 (42.9) 5 (4.71) < .001*
Ventilatory support/O2 therapy n (%)
High flow nasal cannula O2 4 (28.6) 26 (76.47) < .001*
Non-invasive mechanical ventilation 0 (0) 7 (20.59)
Invasive mechanical ventilation 5 (35.7) 1 (2.94)
AKI during follow-up n (%) 8 (57.1) 7 (6.7) < .001*
CVVHD/HD n (%) 3 (21.4) 2 (1.9) .001*
Concomitant infection n (%) 8 (57.1) 25 (23.58) .009*
Final creatinine (mg/dL) 2.02 ± 2.22 1.15 ± 0.96 0.01*
Hospitalization time (d) 25 (90-6) 7 (35-1) < .001*
ICU time (d) 25.44 ± 16.11 8.36 ± 4.94 .004*
AKI, acute kidney injury; CVVHD, continuous veno-venous hemodialysis; eGFR, estimated glomerular filtration rate; HD, hemodialysis; ICU, intensive care unit; IL, interleukin; NI, nosocomial infection.
⁎ P < .05.
Transplant recipients with NI received significantly higher doses of MAA (953.33 ± 114.31 vs 691.43 ± 307.08 mg in 24 hours; P = .003).
Among transplant patients who developed NI, 5 of 6 microbiological isolations were documented; among controls, 2 of 8 were documented. Among transplant recipients, there was a significantly higher frequency of infection by Legionella pneumophila, P. aeruginosa, Stenotrophomona maltophilia, and Staphylococcus epidermidis (Table 7 ).Table 7 Etiology of Nosocomial Infection and Transplant Versus Non-transplant Patients
Table 7Pathogen Cases n (%) Controls n (%) P Value
Klebsiella oxytoca 1 (20) 1 (50) .04*
Stenotrophomona maltophilia 1 (20) 0 (0)
Legionella pneumophila 1 (20) 0 (0)
Pseudomonas aeruginosa 1 (20) 0 (0)
Staphylococcus aureus 0 (0) 1 (50)
Streptococcus epidermidis 1 (20) 0 (0)
⁎ P < .05.
Opportunistic Infection
Table 8 summarizes the variables associated with OI during the study period. Patients with OI were more likely to be diabetic; as in those with NI, their evolutive course was worse, with a higher frequency of ICU admission, shock, renal replacement therapy, mortality, and hospital stay. Upon multivariate analysis, baseline eGFR, serum IL-6 level, and the presence of coinfection were independently associated with OI (Table 8).Table 8 Variables Significantly Associated with OIs in the Study Groups
Table 8 OI
n = 11 Non-OI
n = 109 P Value
Group n (%)
Tx patients 8 (72.7) 22 (20.18) < .001*
Non-Tx patients 3 (27.3) 87 (79.81)
Baseline eGFR (mL/min) 42.06 ± 15.65 72.86 ± 27.55 < .008*
Diabetes mellitus n (%) 8 (72.7) 31 (28.44) .003*
eGFR at dx (mL/min) 36.28 ± 19.87 68.24 ± 30.94 .01*
IL-6 (pg/mL) 135.4 (1682-25) 23.3 (744.7-1) .001*
MULBSTA score 13.3 ± 3.88 9.27 ± 3.94 .003*
AKI at admission n (%) 6 (54.5) 22 (20.18) .01*
ICU admission 6 (54.5) 13 (11.92) < .001*
Shock n (%) 6 (54.5) 5 (4.58) < .001*
Ventilatory support/O2 therapy n (%)
High-flow nasal cannula O2 4 (50) 26 (74.28) .001*
Non-invasive mechanical ventilation 0 (0) 7 (20)
Invasive mechanical ventilation 4 (50) 2 (5.72)
AKI during follow-up n (%) 6 (54.5) 9 (8.25) < .001*
CVVHD/HD n (%) 2 (18.2) 3 (2.75) .01*
Concomitant infection n (%) 6 (54.3) 27 (24.77) .03*
COVID-19 pharmacologic therapy n (%)
Remdesevir 5 (45.5) 16 (15) .01*
Final eGFR (mL/min) 45.35 ± 28.45 74.2 ± 29.59 .003*
Hospitalization time (d) 19 (61-7) 7 (90-1) .001*
ICU time (d) 25 ± 7.92 12.21 ± 14.58 .02*
Logistic Regression Model.
B E.T. P Value Exp(B) IC 95%
Basal eGFR (mL/min) –0.097 0.042 .02 0.9 0.83-0.98
IL-6 (pg/mL) 0.005 0.002 .021 1 1-1.01
Coinfection at admission –3.08 1.34 .02 0.046 0.003-0.63
R2 Nagelkerke: 0.70.
AKI, acute kidney injury; B,unstandarddized regression weight; COPD, chronic obstructive pulmonary disease; CVVHD; eGFR, estimated glomerular filtration rate; E.T., standard Error; HD, hemodialysis; ICU, intensive care unit; CI, confidence interval; IL, interleukin; OI, opportunistic infection; Tx, transplant.
⁎ P < .05.
Overall, the most frequent OI was pulmonary aspergillosis, mainly by Aspergillus niger. Among transplant recipients, infection by Pneumocystis jirovecii and cytomegalovirus (CMV) was more frequent (Table 2). No significant differences were observed in transplant age and maintenance immunosuppressive therapy.
The RR for OI in transplant recipients, without adjustment for baseline eGFR and comorbidity, was 8 (IC 95%; 2.26-28.22). After adjusting for the aforementioned factors, transplant patients still had significant RR for OI, 1.88 (IC 95%: 1.12-3.17). In this group, the mortality for pulmonary aspergillosis was 100%.
Discussion
The results of this study showed that the clinical course and evolution of kidney transplant patients affected by COVID-19 and requiring hospital admission are generally worse, with higher mortality and ICU admission and ventilatory support than non-transplant patients. After adjustment for comorbidity and baseline eGFR, these differences lost their statistical significance, except for OI (Table 3). At equivalent magnitudes of comorbidity and kidney function, there were no differences between transplant and non-transplant patients. These findings are in accordance with those of Chavarot et al [24], who found no differences in patient survival and the composite survival/ICU admission between groups after adjusting for comorbidities and renal function. However, in practice, organ transplant recipients usually have substantial comorbidity that traces back to the pretransplant stage. This comorbidity is mainly determined by chronic kidney disease and its cause (which frequently includes high blood pressure, diabetes, and dyslipidemia). In addition, immunosuppressive therapy prescribed to prevent rejection, besides its effects on the immune response, is associated with adverse metabolic effects, with a higher risk for new-onset hypertension, diabetes, and dyslipidemia [25], [26], [27] and, consequently, greater cardiovascular morbidity.
Among kidney transplant recipients, the condition that is undoubtedly associated with better outcomes does not require hospitalization for COVID-19. In this case, immunosuppressive treatment has detrimental effects on the immunogenicity of the vaccines against SARS-CoV-2 [28]. In our study, patients with a higher IgG antibody titer against SARS-CoV-2 were hospitalized less frequently. Transplant patients who were not admitted received a lower prednisone dose and had older transplant age and, consequently, less intense immunosuppression. In addition, baseline eGFR was better, which is associated with a more robust antibody response after immunization [29], and they were receiving mTOR inhibitors more frequently (Table 1). The latter have been correlated with better antibody responses after vaccination for a diversity of infective agents [30]; furthermore, they are prescribed for controlling recurrent chronic opportunistic infections, such as CMV and BK virus [31], [32], [33]. This explains our findings on the multivariate analysis (Table 1).
Dyslipidemia, lower eGFR at admission, higher MULBSTA score, and the need for ventilatory support were independently associated with death (Table 4).
Patients treated for SARS-CoV-2 infection are at risk for developing OI because of the anti-inflammatory therapy used to treat the cytokine storm that characterizes the severe forms of the infection. Corticoid treatment and IL-6 blockade, despite improving the course of the infection by reducing inflammation and tissue damage, impair immune responses against other pathogens, such as mycobacteria, fungi, and protozoa [34,35]. Pulmonary aspergillosis, rhino-cerebral mucormycosis, candidemia, and other infectious diseases typically found in immunocompromised patients have been reported in COVID-19 patients without these conditions [35,36]. Nutritional status, comorbidity, and corticoid and tocilizumab dosage are associated with a higher risk for OI. In our study, when the groups were analyzed as a whole, pulmonary aspergillosis was the most frequent OI. The variables independently associated with OI were high levels of IL-6, coinfection, and lower eGFR (Table 8). Higher IL-6 levels were associated with greater frequency of tocilizumab prescription; however, when comparing patients who developed OI with those who did not, no significant differences were observed in the doses of dexamethasone and tocilizumab (Table 8). When the analysis was done, separating the patients into transplant and non-transplant groups, kidney recipients with OI received tocilizumab more frequently (Table 3). Higher IL-6, besides indicating a more intense inflammatory response against SARS-CoV-2, was associated with coinfection at diagnosis (predominantly bacterial), indicating a greater frequency of antibiotic therapies of variable spectrums. Wide-spectrum antibiotic therapy is a predisposing factor for OI, especially fungus. In our study, after adjusting for baseline kidney function and comorbidity, transplant patients with COVID-19 were still 1.88 times more likely to develop OI (IC 95%; 1.12-3.17). Maintenance immunosuppressive treatment is the predisposing factor, especially corticoids and anticalcineurinics. This explains the pulmonary aspergillosis (with high mortality) observed in this group and the greater frequency of pneumocystosis and invasive intestinal CMV infection (Table 2).
The present study has its limitations, mainly its retrospective design and the sample size. The virus variant responsible for each infection episode was unavailable for all cases. The treatment protocols for COVID-19 were not homogenous and varied depending on the availability of clinical evidence. Another limitation was not including non-invasive CMV and BK virus infection in the analysis. Given the study´s retrospective design, measurements of polymerase chain reaction for CMV in blood were not performed systematically; thus, only those cases of invasive infection were included. The same may be said about BK virus infection.
In conclusion, the evolutive course of SARS-CoV-2 infection requiring hospitalization in renal transplant recipients is primarily determined by comorbidity and baseline renal function. At equal comorbidity and eGFR, there were no differences in mortality, ICU admission, NI, or hospital stay between transplant recipients and controls. The risk for OI among transplant patients was significantly greater, with a higher frequency of pneumocystosis and CMV colitis. Baseline eGFR, higher levels of serum IL-6, and coinfection at admission were independently associated with OI. In general, the prognosis in kidney transplant recipients hospitalized for COVID-19 is comparatively bad; the principal measure to effectively impact these outcomes is to prevent hospital admission, either by improving the antibody response to vaccination or by early diagnosis and treatment (monoclonal antibodies; antivirals, such as remdesevir and molnupiravir) in an ambulatory dimension.
Declaration of Competing Interest
This study protocol was reviewed and approved by the Comitè d´Ètica d´Investigació I Calidad of Hospital Universitari Arnau de Vilanova.
Data availability
Data will be made available on request.
==== Refs
References
1 Udomkarnjananun S Kerr SJ Townamchai N Susantitaphong P Tulvatana W Praditpornsilpa K Mortality risk factors of COVID-19 infection in kidney transplantation recipients: a systematic review and meta-analysis of cohorts and clinical registries Sci Rep 11 2021 20073 34625642
2 Grasselli G Greco M Zanella A Albano G Antonelli M Bellani G Risk factors associated with mortality among patients with COVID-19 in intensive care units in Lombardy, Italy JAMA Intern Med 180 2020 1345 1355 32667669
3 Zhou F Yu T Du R Fan G Liu Y Liu Z Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study Lancet 395 2020 1054 1062 32171076
4 Fung M Babik JM. COVID-19 in immunocompromised hosts: what we know so far Clin Infect Dis 2 2020 340 350
5 Mahalingasivam V Craik A Tomlinson LA Ge L Hou L Wang Q A systematic review of COVID-19 and kidney transplantation Kidney Int Rep 6 2021 24 25 33163708
6 Moosavi SA Mashhadiagha A Motazedian N Hashemazar A Hoveidaei AH Bolignano D. COVID-19 clinical manifestations and treatment strategies among solid-organ recipients: a systematic review of cases Transpl Infect Dis 22 2020 e13427 32779820
7 Nacif LS Zanini LY Waisberg DR Pinheiro RS Galvão F Andraus W COVID-19 in solid organ transplantation patients: a systematic review Clinics (Sao Paulo) 75 2020 e1983 32520225
8 Oltean M Søfteland JM Bagge J Ekelund J Felldin M Schult A Covid-19 in kidney transplant recipients: a systematic review of the case series available three months into the pandemic Infect Dis (London) 52 2020 830 837 32657186
9 Asch DA Scheils NE Islam MN Chen Y Werner RM Buresh J Variation in US hospital mortality rates for patients admitted with COVID-19 during the frst 6 months of the pandemic JAMA Intern Med 181 2021 471 478 33351068
10 Piroth L Cottenet J Mariet A-S Bonniaud P Blot M Tubert-Bitter P Comparison of the characteristics, morbidity, and mortality of COVID-19 and seasonal infuenza: a nationwide, population-based retrospective cohort study Lancet Respir Med 9 2020 251 259 33341155
11 Aziz F Mandelbrot D Singh T Parajuli S Garg N Mohamed M Early report on published outcomes in kidney transplant recipients compared to nontransplant patients infected with coronavirus disease 2019 Transplant Proc 52 2020 2659 2662 32753243
12 Naderi GH Mehraban D Kazemeyni SM Darvishi M Latif AH. Living or deceased donor kidney transplantation: a comparison of results and survival rates among Iranian patients Transplant Proc 41 2009 2772 2774 19765431
13 Nemati E Einollahi B Lesan Pezeshki M Porfarziani V Fattahi MR Does kidney transplantation with deceased or living donor affect graft survival Nephrourol Mon 6 2014 e12182 25695017
14 Nicholson ML Metcalfe MS White SA Waller JR Doughman TM Horsburgh T A comparison of the results of renal transplantation from non-heart-beating, conventional cadaveric, and living donors Kidney Int 58 2000 2585 2591 11115095
15 Pratschke J Dragun D Hauser IA Horn S Mueller TF Schemmer P Immunological risk assessment: the key to individualized immunosuppression after kidney transplantation Transplant Rev (Orlando) 30 2016 77 84 26965071
16 Wu WK Famure O Li Y Kim SJ. Delayed graft function and the risk of acute rejection in the modern era of kidney transplantation Kidney Int 88 2015 851 858 26108067
17 Cravedi P Mothi SS Azzi Y Haverly M Farouk SS Pérez-Sáez M COVID-19 and kidney transplantation: results from the TANGO international transplant consortium Am J Transplant 20 2020 3140 3148 32649791
18 Caillard S Anglicheau D Matignon M Durrbach A Greze C Frimat L An initial report from the French SOT COVID registry suggests high mortality due to COVID-19 in recipients of kidney transplants Kidney Int 98 2020 1549 1558 32853631
19 Hilbrands LB. COVID-19-related mortality in kidney transplant and dialysis patients: results of the ERACODA collaboration Nephrol Dial Transplant 35 2020 1973 1983 33151337
20 Kute VB Bhalla AK Guleria S Ray DS Bahadur MM Shingare A Clinical profle and outcome of COVID-19 in 250 kidney transplant recipients: a multicenter cohort study from India Transplantation 105 2021 851 860 33350674
21 Villanego F Mazuecos A Pérez-Flores IM Moresco F Andrés A Jiménez-Martín C Predictors of severe COVID-19 in kidney transplant recipients in the diferent epidemic waves: analysis of the Spanish Registry Am J Transplant 21 2021 2573 2582 33756067
22 Mehta RL Kellum JA Shah SV Molitoris BA Ronco C Warnock DG Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury Crit Care 11 2007 R31 17331245
23 Levey AS Stevens LA Schmid CH Zhang YL Castro AF 3rd Feldman HI A new equation to estimate glomerular filtration rate Ann Intern Med 150 2009 604 612 19414839
24 Chavarot N Gueguen J Bonnet G Jdidou M Trimaillee A Burger C COVID-19 severity in kidney transplant recipients is similar to nontransplant patients with similar comorbidities Am J Transplant 21 2021 1285 1294 33252201
25 European FK506 Multicentre Liver Study Group Randomised trial comparing tacrolimus (FK506) and cyclosporin in prevention of liver allograft rejection Lancet 344 1994 423 428 7520105
26 Pham PT Pham PM Pham SV Pham PA Pham PC. New onset diabetes after transplantation (NODAT): an overview Diabetes Metab Syndr Obes 4 2011 175 186 21760734
27 Nankivell BJ PʼNg CH OʼConnell PJ Chapman JR. Calcineurin inhibitor nephrotoxicity through the lens of longitudinal histology: comparison of cyclosporine and tacrolimus eras Transplantation 100 2016 1723 1731 27306529
28 Babel N Hugo C Westhoff TH. Vaccination in patients with kidney failure: lessons from COVID-19 Nat Rev Nephrol 18 2022 708 723 35999285
29 Haddiya I. Current knowledge of vaccinations in chronic kidney disease patients Int J Nephrol Renovas Dis 13 2020 179 185
30 Netti GS Infante B Troise D Mercuri S Panico M Spadaccino F mTOR inhibitors improve both humoral and cellular response to SARS-CoV-2 messenger RNA BNT16b2 vaccine in kidney transplant recipients Am J Transplant 22 2022 1475 1482 35038362
31 Jacobi J Prignitz A Büttner M Korn K Weidemann A Hilgers KF BK viremia and polyomavirus nephropathy in 352 kidney transplants; risk factors and potential role of mTOR inhibition BMC Nephrol 14 2013 207 24088187
32 Tedesco-Silva H Felipe C Ferreira A Cristelli M Oliveira N Sandes-Freitas T Reduced incidence of cytomegalovirus infection in kidney transplant recipients receiving everolimus and reduced tacrolimus doses Am J Transplant 15 2015 2655 2664 25988935
33 Aguiar PV Campistol JM Diekmann F. Safety of mTOR inhibitors in adult solid organ transplantation Expert Opin Drug Saf 15 2016 303 319 26667069
34 Scherger S Henao-Martínez A Franco-Paredes C Shapiro L. Rethinking interleukin-6 blockade for treatment of COVID-19 Med Hypotheses 144 2020 110053
35 Abdoli A Falahi S Kenarkoohi A. COVID-19-associated opportunistic infections: a snapshot on the current reports Clin Exp Med 22 2022 327 346 34424451
36 Kurra N Woodard PI Gandrakota N Gandhi H Polisetty SR Ang SP Opportunistic infections in COVID-19: a systematic review and meta-analysis Cureus 14 2022 e23687 35505698
|
PMC010xxxxxx/PMC10202354.txt |
==== Front
Arch Virol
Arch Virol
Archives of Virology
0304-8608
1432-8798
Springer Vienna Vienna
37217624
5798
10.1007/s00705-023-05798-3
Original Article
Coinfection with porcine epidemic diarrhea virus and Clostridium perfringens type A enhances disease severity in weaned pigs
Lee Duri 1
Jang Guehwan 1
Min Kyeng-Cheol 2
Lee Inn Hong 2
Won Hokeun 2
Yoon In-Joong 2
Kang Sang Chul 3
http://orcid.org/0000-0002-5930-5461
Lee Changhee changhee@gnu.ac.kr
14
1 grid.256681.e 0000 0001 0661 1492 College of Veterinary Medicine and Virus Vaccine Research Center, Gyeongsang National University, 52828 Jinju, Republic of Korea
2 ChoongAng Vaccine Laboratories, 34055 Daejeon, Republic of Korea
3 Optipham Inc, 28158 Cheongju, Republic of Korea
4 grid.256681.e 0000 0001 0661 1492 College of Veterinary Medicine, Gyeongsang National University, 52828 Jinju, Republic of Korea
Communicated by Diego G. Diel
22 5 2023
2023
168 6 16630 1 2023
25 4 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Clostridium perfringens is a constituent of the normal gut microbiome in pigs; however, it can potentially cause pre- and post-weaning diarrhea. Nevertheless, the importance of this bacterium as a primary pathogen of diarrhea in piglets needs to be better understood, and the epidemiology of C. perfringens in Korean pig populations is unknown. To study the prevalence and typing of C. perfringens, 203 fecal samples were collected from diarrheal piglets on 61 swine farms during 2021–2022 and examined for the presence of C. perfringens and enteric viruses, including porcine epidemic diarrhea virus (PEDV). We determined that the most frequently identified type of C. perfringens was C. perfringens type A (CPA; 64/203, 31.5%). Among the CPA infections, single infections with CPA (30/64, 46.9%) and coinfections with CPA and PEDV (29/64, 45.3%) were the most common in diarrheal samples. Furthermore, we conducted animal experiments to investigate the clinical outcome of single infections and coinfections with highly pathogenic (HP)-PEDV and CPA in weaned piglets. The pigs infected with HP-PEDV or CPA alone showed mild or no diarrhea, and none of them died. However, animals that were co-inoculated with HP-PEDV and CPA showed more-severe diarrheal signs than those of the singly infected pigs. Additionally, CPA promoted PEDV replication in coinfected piglets, with high viral titers in the feces. A histopathological examination revealed more-severe villous atrophy in the small intestine of coinfected pigs than in singly infected pigs. This indicates a synergistic effect of PEDV and CPA coinfection on clinical disease in weaned piglets.
http://dx.doi.org/10.13039/501100014189 Korea Institute of Planning and Evaluation for Technology in Food, Agriculture and Forestry 321018-1 Lee Changhee issue-copyright-statement© Springer-Verlag GmbH Austria, part of Springer Nature 2023
==== Body
pmcIntroduction
Clostridium perfringens is a ubiquitous Gram-positive spore-forming anaerobic bacterium that has been isolated from diverse environments, including swine ecosystems [1]. C. perfringens isolates from pigs are divided into five toxinotypes (A–E) based on their ability to produce a combination of six toxins: α- (cpa), β1- (cpb), β2- (cpb2), entero- (cpe), ε- (etx), and ι-toxins (iap) [2]. Although all types of C. perfringens can generally be found in the gastrointestinal tract of piglets, only C. perfringens types A (CPA) and C (CPC) are considered pathogenic in nursing pigs [3]. In particular, CPA is the most frequently detected bacterial pathogen in porcine neonatal diarrhea [3, 4].
Porcine epidemic diarrhea virus (PEDV) is a swine enteric coronavirus belonging to the subgenus Pedacovirus of the genus Alphacoronavirus within the family Coronaviridae of the order Nidovirales [5]. PEDV is highly contagious and can be fatal for neonatal piglets within their first week of life, with morbidity and mortality up to 100%. However, it causes milder clinical disease with lower mortality in older pigs [6, 7]. In 2013, a highly pathogenic (HP) PEDV strain emerged in the United States. Subsequently, it spread to other pig-raising countries in America, Asia, and Europe, posing a significant health threat to the global pig population [6, 8]. Since then, PEDV strains have been divided into two genotypes, each with two subgenotypes: the low-pathogenic genotype 1 (historical G1a and recombinant G1b) and the HP genotype 2 (local epidemic G2a and pandemic G2b) [6, 7].
Since the 2013–2014 pandemic, the HP-G2b strain of PEDV has evolved and become endemic to South Korea, resulting in year-round small- to large-scale outbreaks nationwide [6, 7]. In endemic circumstances, PEDV persists in weaning or growing-finishing barns, where the virus circulates, causing mild post-weaning diarrhea with low mortality rates [6]. PEDV infection in weaned pigs leads to decreased growth performance, including lower average daily gain (ADG) and average daily feed intake (ADFI) [9]. Furthermore, the severity of clinical disease in nursing and weaning piglets may be exacerbated by coinfections with other viral and bacterial enteropathogens [10]. Since the incidence and type of C. perfringens associated with pre- and post-weaning diarrhea in South Korea remain undetermined, the aim of the present study was to investigate the prevalence and typing of C. perfringens in diarrheic samples collected from domestic swine farms. We evaluated the pathogenic outcomes of single and sequential coinfections with PEDV (HP-G2b) and CPA in weaned pigs.
Materials and methods
Collection of clinical samples
In total, 61 commercial farrow-to-finish swine farms located in South Korea with a reported history of diarrheic disease were sampled to identify C. perfringens. Fecal samples (n = 203) of affected piglets (2–4 weeks of age) showing pre- or post-weaning diarrhea collected from January 2021 through January 2022 were submitted for laboratory diagnosis. Fecal samples were diluted with PBS to make 10% (wt/vol) suspensions [11–13], which were vortexed and then centrifuged for 10 min at 4,500 × g in a Hanil FLETA5 centrifuge (Incheon, South Korea). The clarified supernatants were initially subjected to conventional PCR for the detection and typing of C. perfringens as described previously [2, 3]. In brief, cpa-gene-based PCR was performed to detect C. perfringens using diarrheic fecal samples, and C. perfringens isolates were further assigned to one of the five toxinotypes based on the combination of toxin genes detected by PCR (e.g., CPA was diagnosed by testing for the cpa and cpb2 genes, while CPC was diagnosed by testing for the cpa and cpb genes). To determine if any other diarrhea-causing viral pathogens were present in the clinical samples, we performed virus-specific RT-PCR analysis for PEDV, transmissible gastroenteritis virus (TGEV), porcine deltacoronavirus (PDCoV), and porcine rotaviruses (PRV) as described previously [14–16]. PEDV-positive samples were further subjected to nucleotide sequence analysis for genotyping as described elsewhere [12, 15, 16].
Pig infection experiments and clinical examination
The animal studies described here were approved by the Institutional Animal Care and Use Committee (IACUC) of the Choong Ang Vaccine Laboratory (CAVAC) and conducted at the CAVAC Animal Facility as described previously [11, 13, 17, 18]. A total of 16 weaned piglets (19 days old at the start of the experiment) were obtained from commercial crossbred sows (Great Yorkshire × Dutch Landrace) at a conventional breeding farm with a good health record and without a previous herd history of a PEDV outbreak or PEDV vaccination. All pigs were also tested to confirm that they were not infected with any porcine enteric viruses, including PEDV, TGEV, PDCoV, and PRV [14–16]. Pigs were assigned to four experimental groups and housed in four separate rooms: the PEDV (HP-G2b KNU-141112 strain)-inoculated group (n = 4), the CPA (CAVAC strain carrying the cpa and cpb2 genes encoding the α- and β2-toxins)-inoculated group (n = 4), the PEDV/CPA co-inoculated group (n = 4), and the control group (n = 4). The animals were fed commercial milk replacer (3–4 times daily) and had ad libitum access to water for the duration of the study [11, 13, 17, 18]. At 21 days of age (after a 2-day acclimation period), the pigs were inoculated orally with KNU-141112 at a dose of 104.0 50% tissue culture infective doses (TCID50) per pig [17] or CPA at a dose of 1.1 × 109.0 colony-forming units (CFU) per pig [19]. Pigs in the PEDV and CPA coinfection group were first inoculated orally with PEDV and then with CPA at the same dose as described above. The dose of PEDV was equivalent to 1,000 times the median pig diarrhea dose (PDD50) of the KNU-141112 strain in neonatal piglets [11, 17, 20]. The sham-inoculated piglets were inoculated with cell culture medium as a control. All piglets were observed for 7 days post-inoculation (dpi), with daily recording of clinical signs (vomiting and diarrhea) and mortality [11, 13, 17, 18]. Stool samples from the pigs in all groups were collected prior to inoculation and thereafter daily using 16-inch cotton-tipped swabs and subjected to quantitative real-time RT-PCR (rRT-PCR) to measure PEDV fecal shedding titers as described previously [13, 15, 17, 18]. The KNU-141112 strain with a known infectivity titer was serially diluted tenfold to generate a standard curve in each PCR plate. The virus concentrations (genome copies/ml) in the samples were calculated based on this standard curve. The mean cycle threshold (Ct) values were calculated for the PCR-positive samples, and the mean virus titers were calculated for all pigs within each group [13, 15, 17, 18]. In addition, the area under the curve (AUC) of viral shedding in each group was calculated using the log transformed values of the viral loads from 1 to 7 dpi [21]. The clinical significance score (CSS) was determined using the following scoring criteria for diarrheal severity based on the fecal consistency at 7 dpi as described previously [13, 17, 18]: 0, normal and no diarrhea (mean Ct values > 35); 1, mild and fluidic feces; 2, moderate mucous to watery diarrhea; 3, severe watery and projectile diarrhea (mean Ct values < 20); 4, death. The animals were weighed daily to measure their ADG. Any pigs that died during the study were necropsied, while the surviving pigs from the inoculation and control groups were euthanized at 7 dpi for post-mortem examination [11, 13, 17, 18].
Histopathology and immunohistochemistry of the small intestine
At necropsy, intestinal tissues and other major organs were examined macroscopically. Small intestinal tissue (duodenum, proximal jejunum, and ileum) specimens (< 3 mm thick) were collected from each piglet, fixed in 10% formalin for 24 h at room temperature, and embedded in paraffin according to standard laboratory procedures as described previously [11, 13, 17, 18]. The formalin-fixed paraffin-embedded tissue samples were cut into 4- to 5-µm-thick sections on a microtome (Leica, Wetzlar, Germany), floated in a 40°C water bath containing distilled water, and transferred to glass slides. The tissue was then deparaffinized in xylene for 5 min and rehydrated in decreasing concentrations of ethanol (100%, 95%, 90%, 80%, and 70%, respectively) for 3 min each. The deparaffinized intestinal tissue sections were stained with hematoxylin and eosin (H&E; Sigma-Aldrich, St. Louis, MO) for histopathology or subjected to immunohistochemistry (IHC) for the detection of PEDV antigens using a PEDV nucleocapsid (N)-specific monoclonal antibody (MAb; CAVAC, Daejeon, South Korea) as described previously [11, 13, 17, 18]. The severity of villus atrophy was also quantified by measuring the ratio between villous height and crypt depth (VH:CD) throughout the H&E-stained small intestinal sections, and the mean ratio of VH:CD in each small intestine segment was calculated as described previously [18, 22].
Statistical analysis
All values are expressed as the standard deviation of the mean (SDM). Statistical analysis was conducted using the GraphPad Prism 7 software package (GraphPad Software, San Diego, CA). P-values less than 0.05 were considered statistically significant.
Results
Prevalence of PEDV or/and CPA in the swine samples
A total of 203 fecal samples from 61 swine farms (3.3 samples on average) were examined bacteriologically for C. perfringens as well as for enteric viruses, including PEDV, TGEV, PDCoV, and PRV. On seven farms (11.5%), none of the pathogens investigated in this study were detected, but 129 fecal samples (63.5%) from 49 farms (80.3%) tested positive for PEDV. All of the PEDV isolates that were detected were further investigated by nucleotide sequencing and found to belong to the HP-G2b strain. These data indicated that most swine farms tested in this study were PEDV endemic. In 65 (32.0%) of the samples from 25 farms (41%), C. perfringens was identified. Subsequent typing of the C. perfringens-positive samples revealed that 64 samples (98.5%) from 25 farms (100%) contained CPA, while one sample (1.5%) from one farm (4%) was positive for CPC. Molecular toxin typing revealed that 82.8% of the CPA isolates (53/64) carried the coding genes for the α- (cpa) and β2- (cpb2) toxins, suggesting virulence potential. The remaining CPA-positive samples (11/64, 17.2%) contained only the cpa toxin gene. No sample or farm was positive for other toxinotypes. Among the CPA-positive samples, single CPA infections (30/64, 46.9%) and CPA/PEDV coinfections (29/64, 45.3%) were almost equally prevalent, and 29 out of 64 CPA-positive samples from 17 out of 25 CPA-positive farms were also positive for PEDV. CPA/PDCoV (2/58) and CPA/PRV (3/58) coinfections were detected on one and two of the CPA-positive farms, respectively. One CPC-positive fecal sample tested positive for PEDV (CPC/PEDV coinfection). An overview of the combination of pathogens that were detected in 203 diarrheal samples from 61 farms and toxin typing of 65 CPA isolates is shown in Fig. 1. The predominant combinations of pathogens detected in pre- and post-weaning diarrheic samples were PEDV (99/203, 48.8%), CPA (30/203, 14.8%), and CPA/PEDV (29/203, 14.3%), and the most relevant major toxin gene profile in the CPA isolates was the combination of cpa (α-toxin) and cpb2 (β2-toxin) genes.
Fig. 1 The combination of pathogens detected in pre- or post-weaning diarrheic samples (A, n = 204) from commercial swine farms (B, n = 61) in South Korea, 2020–2022. The circular graphs in each panel show the percentage of pigs (A) and farms (B) with at least one CPA isolate encoding cpa and/or cpb2 toxin genes.
Clinical outcomes of coinfection with PEDV and CPA in weaned pigs
Based on the CPA prevalence study, it was interesting to note that CPA/PEDV coinfection appeared to result in higher disease severity than a single infection with either pathogen. To determine the clinical consequences of the interaction between the co-circulating PEDV and CPA strains, we investigated the effects of coinfection with HP-PEDV and CPA in weaned piglets. Sixteen pigs were divided into four groups of four animals each and challenged orally with HP-PEDV (group 1), CPA (group 2), or HP-PEDV/CPA (group 3), and the remaining four animals in the control group were inoculated with cell culture medium. Clinical signs were recorded three times daily, and fecal swabs were collected before inoculation and daily after inoculation throughout the experiment. All animals were vigorous and clinically normal, showing normal stool consistency, and their fecal samples were negative during PEDV RNA for the 2-day acclimation period.
The negative-control piglets demonstrated normal clinical manifestations throughout the study. As shown in Fig. 2A, the piglets infected with PEDV KNU-141112 showed mild diarrhea at 2 dpi (mean CSS = 1.0) and then recovered at 3 or 4 dpi. PEDV-associated mortality was absent in the HP-PEDV group. Similarly, except one pig in group 2 that exhibited moderate watery diarrhea at 2 dpi, the animals inoculated with CPA alone displayed mild diarrhea at 1 or 2 dpi (mean CSS = 0.25 and 0.75, respectively) without mortality and recovered at 5 or 7 dpi. However, the piglets from the HP-PEDV/CPA coinfection group showed moderate to severe watery diarrheal symptoms at 2 dpi (mean CSS = 2.25) but remained alive until the end of the study. The diarrhea of two piglets in group 3 lasted until 7 dpi, while the remaining animals recovered at 7 dpi (Fig. 2A).
Fig. 2 The clinical outcomes of single infections and coinfections with PEDV and CPA in weaned piglets. (A) The clinical significance scores (CSS) for each group. The CSS of individual pigs from each group was measured as described in Materials and methods. (B) The fecal viral RNA shedding profile of PEDV in each group. The PEDV RNA titers (log10 genomic copies/ml) in rectal swaps at the indicated sampling time points were determined using rRT-PCR. The mean values for each group at each time point are presented, and error bars indicate the SDM. The P-values were calculated by comparing the data between the PEDV/CPA coinfection and single-infection groups. *, P = 0.001–0.05; **, P < 0.001
Most inoculated piglets (3/4) in group 1 tested positive for PEDV by 2 dpi with a mean titer of 102.9 genomic copies/ml. All of the animals shed large amounts of PEDV in their feces, with mean titers ranging from 104.8 to 106.4 genomic copies/ml (Fig. 2B). Although fecal shedding of PEDV was observed in all piglets of the HP-PEDV/CPA coinfection group by 2 dpi, these pigs shed much higher amounts of PEDV, with a mean titer of 106.9 genomic copies/ml, than the piglets inoculated with PEDV alone. The titer then increased gradually, ranging from 107.1 to 107.7 genomic copies/ml at 5 dpi, and declined thereafter. Overall, a significantly larger amount of fecal shedding at 2 and 4 dpi was observed in the HP-PEDV/CPA coinfected animals (group 3) than in those in group 1, with a maximum of a 4-log elevation compared to group 1 upon coinfection (Fig. 2B). We also found that the mean viral AUC in the HP-PEDV/CPA coinfection group was greater than that in the HP-PEDV single-infection group (35.5 ± 0.8 vs. 27.5 ± 1.8 log10 copies/ml × days). The feces of the piglets in the CPA-inoculated and control groups were negative for PEDV RNA during the trial.
In addition, the starting and final body weights of individual piglets from each group were measured at 0 and 7 dpi during the experimental period (Fig. 3A). The pigs inoculated with CPA (group 2) weighed 2.3% less and those inoculated with HP-PEDV/CPA (group 3) weighed 7.3% less at the end of the study than the control pigs (group 4), which were not significantly different in weight from the HP-PEDV-inoculated animals (group 1). The ADG per group was calculated from 0 to 7 dpi, and the overall ADG in the challenged pigs from groups 1–3 was compared to that in the unchallenged animals (group 4) (Fig. 3B). Single infection and coinfection with CPA resulted in a significant decline in ADG compared with the control group. The HP-PEDV/CPA-coinfected pigs (group 3) had the greatest decrease in ADG (mean ADG of -0.525 kg) compared to the control pigs (mean ADG of + 0.65 kg). There was no significant variation in ADG between the HP-PEDV-infected and control groups.
Fig. 3 The mean body weight (A) and average daily weight gain (ADG) (B) in weaned piglets from each group after single infection and coinfection with PEDV and CPA through 7 dpi. The mean body weights and ADG between the PEDV/CPA coinfection and single-infection groups were compared. The error bars represent the mean ± SDM. *, P = 0.001–0.05; **, P < 0.001
All animals in the inoculated and control groups were euthanized for postmortem evaluation at 7 dpi (Fig. 4). No macroscopic or histologic intestinal alterations were recorded in the sham-inoculated piglets (Fig. 4D). Although all animals in the single-infection and coinfection groups displayed no gross macroscopic lesions at necropsy, they had microscopic intestinal changes demonstrating jejunal villous atrophy (Fig. 4A–C). IHC staining detected PEDV N proteins predominantly in the cytoplasm of epithelial cells in atrophied jejunal villi of the piglets infected with HP-PEDV alone or coinfected with HP-PEDV and CPA (groups 1 and 3; Fig. 4A and C). PEDV antigen was absent in the small intestines of all of the animals infected with CPA alone (group 2; Fig. 4B) or inoculated with cell culture medium (group 4; Fig. 4B). PEDV RNA titers were also determined in the intestinal tissue samples collected at necropsy. There were no statistical differences in the PEDV titers in the small intestinal segments between the HP-PEDV single-infection and HP-PEDV/CPA coinfection groups (Fig. 5A). However, the mean (± SDM) VH:CD ratios of the jejunum differed significantly among the groups (Fig. 5B). The lowest ratio was observed in the HP-PEDV/CPA coinfection group (3.12 ± 0.39), medium to low ratios were observed in the CPA (3.67 ± 0.92) and HP-PEDV single-infection (4.36 ± 0.89) groups, and the highest ratio was observed in the uninoculated control group (5.82 ± 0.86).
Fig. 4 Macroscopic and microscopic small intestine lesions in piglets after inoculation with PEDV (group 1), CPA (group 2), or PEDV/CPA (group 3), and in sham control animals (group 4). Small intestines of individual piglets from each group were examined for gross lesions. Representative necropsy images are presented at the top of each panel. Hematoxylin- and eosin-stained tissue sections of the proximal jejunum from representative piglets in each group are shown at the middle of each panel (100× magnification, scale bar = 100 µm). IHC analysis results showing PEDV antigens in jejunal tissue sections from representative pigs in each group are presented at the bottom of each panel (200× magnification, scale bar = 50 µm). Immunostaining of the PEDV N proteins (brown staining) was detected in the epithelial cells of the proximal jejunum in piglets inoculated with PEDV alone and in those co-inoculated with CPA and PEDV. No PEDV antigen was identified in the small intestine of the CPA single-inoculated and mock-inoculated piglets.
Fig. 5 The viral load and villous height:crypt depth (VH:CD) ratios in the small intestinal tissues of piglets from each group. (A) PEDV RNA loads (log10 genomic copies/ml) in each intestinal tissue collected at necropsy were determined using rRT-PCR. There were no significant differences between the groups. (B) Five villi and crypts of each small intestine section were measured. The mean VH:CD ratios of individual small intestine sections in each group are presented, and the error bars indicate the SDM. *, P = 0.001–0.05; **, P < 0.001
Discussion
Porcine diarrheal disease is a leading cause of neonatal mortality and reduced weight gain, resulting in significant economic losses. The primary enteric pathogens include bacteria such as Escherichia coli and C. perfringens as well as coronaviruses and rotaviruses. In South Korea, PEDV is the most widespread agent of diarrhea, with high neonatal mortality that has devastated domestic pig-producing farms for the last three decades. Furthermore, endemic PEDV causes post-weaning diarrhea in older pigs, which shed the virus continuously and can act as a source of virus circulation and recurrence on affected farms, increasing the risk of viral transmission to contiguous farms. Although PEDV infection causes less-severe disease and fewer deaths in weaned pigs than in nursing piglets, coinfection with other bacterial or viral pathogens may enhance viral virulence, thereby worsening the disease severity and increasing morbidity and mortality in weaned pigs. Furthermore, PEDV-infected weaned pigs have reduced ADG and ADFI, leading to a slow growth rate [9]. In this regard, the production of post-weaning pigs with single PEDV infections or coinfections that do not reach a suitable slaughter weight at a certain age is not favorable for the swine industry. C. perfringens is a part of the pig intestinal microbiome and is considered an enteric bacterial pathogen in pre- and post-weaning piglets [3, 23, 24]. However, whether C. perfringens is involved in diarrheal disease on domestic pig farms is still unclear. To provide insights into the epidemiological status of C. perfringens in South Korea, the present study was performed to investigate the prevalence and types of C. perfringens circulating in pig populations.
We found a relatively high prevalence (32.0%) of C. perfringens in diarrheal samples collected from 41% of the pig farms that were sampled. The C. perfringens isolates in 64 out of 65 samples belonged to the CPA toxinotype. More than 80% of the CPA isolates that contain the cpa (α-toxin) and cpb2 (β2-toxin) genes can be virulent or potentially virulent. Consistent with recent studies [3, 23, 24], our data show that the toxin-producing CPA was the most frequently detected type of C. perfringens in pre- and post-weaning diarrhea cases. By contrast, CPC is less commonly associated with neonatal diarrhea, although it causes fatal necrohemorrhagic enteritis [4]. We identified only one CPC case, suggesting that it plays a minor role in diarrhea. Testing for bacterial and viral pathogens on 54 farms revealed simultaneous infections with at least two pathogens (e.g., CPA and PEDV) in nearly 40% of the farms, indicating that diarrhea in weaning piglets infrequently has only a single causative agent. It is generally acknowledged that CPA can cause diarrhea, but is often a part of a multifactorial disease, which must be considered when implementing prophylactic and therapeutic measures [25, 26]. Interestingly, we found that 50% of the CPA-positive samples tested positive for PEDV, showing CPA/PEDV coinfection to be the second most frequently detected combination of pathogens in diarrheic pigs. Although its role in the pathogenesis of diarrhea is not well understood, CPA can increase disease severity and aggravate growth retardation in PEDV-infected pigs. To corroborate these results, we tested the effect of coinfection on diarrhea in weaned pigs and found a synergistic pathogenic relationship between CPA and PEDV.
In the present study, 21-day-old weaned piglets were inoculated with HP-PEDV KNU-141112 or/and CPA, and the clinical manifestations were observed and evaluated daily. Our previous studies showed the high enteropathogenicity of KNU-141112, with 100% mortality in 5-day-old neonatal piglets [17, 18]. Considering the strong virulence of KNU-141112 and its high lethality in neonatal piglets, 21-day-old piglets were used in this study to reflect the authentic clinical features of weaned piglets caused by HP-PEDV infection. The weaned pigs infected with HP-PEDV KNU-141112 showed either no diarrhea or mild diarrhea with no deaths, and the disease was self-limiting, confirming the age-dependent resistance to PEDV [6]. The clinical presentation of CPA, including creamy or mucoid diarrhea, has been described in neonatal piglets [19, 27]. In this study, singly infected weaned pigs exhibited limited diarrheal symptoms that lasted 4–5 days, but the coinfected animals in the HP-PEDV/CPA group showed more-severe disease, mainly characterized by a longer duration of diarrhea and a higher level of PEDV shedding when compared to the single CPA- or HP-PEDV-inoculated groups. In addition, the average body weight of the piglets in the coinfection group was significantly lower than in the single-infection group, indicating that CPA has a detrimental effect on the growth performance of pigs that are coinfected with PEDV. Together, these data suggest a confirmed synergistic pathogenic effect of PEDV and CPA coinfection.
The HP-PEDV single-infection pigs had no macroscopic lesions typical of PEDV at necropsy. This observation could be due to the rapid turnover of enterocytes and the high proliferation rate and numbers of crypt stem cells in weaned pigs, which are essential for efficient digestion and adsorption of milk or water to prevent severe dehydration and contribute to recovery from disease in older piglets upon PEDV infection [10]. Likewise, no gross lesions indicative of CPA were observed in the CPA single-infection pigs at necropsy, as shown previously [19, 27]. Consequently, there were no significant differences in macroscopic lesions among the experimental groups. Although CPA infection results in villous tip necrosis associated with heavy colonization of bacillary bacteria in close contact with injured enterocytes, which could be used for CPA diagnosis, such microscopic findings are rarely detected, and lesions are frequently not noticed [19]. We did not observe CPA-associated microscopic changes, except for jejunal villous atrophy, in the CPA single-infection and HP-PEDV/CPA coinfection groups under our experimental conditions. In addition, PEDV viral RNA titers were similar in the duodenum, jejunum, and ileum of the HP-PEDV/CPA-coinoculated pigs and PEDV single-infection pigs. Consistent with these results, IHC showed that the amount of PEDV antigen in the jejunum was similar between the HP-PEDV singly infected and HP-PEDV/CPA coinfected piglets. However, the VH:CD ratio, measuring the severity of villous atrophy, was lower in the jejunum of the HP-PEDV/CPA co-inoculated pigs than in HP-PEDV or CPA singly infected pigs. Moreover, our animal experiments showed that the PEDV titers in feces were significantly higher in the HP-PEDV/CPA-co-inoculated pigs than in the pigs with PEDV infection alone. Therefore, we can deduce that CPA promoted infection and replication of PEDV in the co-inoculated weaned pigs.
In conclusion, this is the first report of the prevalence of CPA associated with piglet diarrhea in South Korea and the disease outcomes of coinfection with HP-PEDV and CPA in weaned piglets. Our PCR-based survey indicated that HP-G2b PEDV was the predominant pathogen detected in pre- and post-weaning diarrheic samples, reflecting the current situation in which the virus continues to affect domestic pig farms endemically. Second, the prevalence of CPA single infection and coinfection with HP-G2b PEDV was similar in weaning pigs from diarrheal farms. Although HP-PEDV or CPA alone caused no or limited diarrheal diseases in weaned piglets, coinfection with HP-PEDV and CPA significantly elevated the disease severity in older piglets. The synergistic pathogenic consequences are associated with enhanced replication of PEDV in the presence of CPA. Given that coinfection with enteric pathogens is associated with post-weaning diarrhea in South Korea, where PEDV is endemic, the monitoring and prophylaxis or therapy of CPA in PEDV-affected farms should lessen the impact of disease caused by PEDV and CPA. Our data will advance our understanding of the synergistic pathogenic mechanisms triggered by porcine enteric pathogens and help control PEDV on infected or endemic farms.
Acknowledgements
This research was supported by Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, and Forestry (IPET) through the Animal Disease Management Technology Development Program, funded by the Ministry of Agriculture, Food, and Rural Affairs (MAFRA) (321018-1).
Declarations
Ethical declarations
All animal procedures were carried out in accordance with the guidelines established by the Institutional Animal Care and Use Committee.
Conflict of interest
The authors declare that they have no conflict of interest.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Chan G Farzan A Soltes G Nicholson VM Pei Y Friendship R Prescott JF The epidemiology of Clostridium perfringens type A on Ontario swine farms, with special reference to cpb2-positive isolates BMC Vet Res 2012 8 156 10.1186/1746-6148-8-156 22947389
2. Baums CG Schotte U Amtsberg G Goethe R Diagnostic multiplex PCR for toxin genotyping of Clostridium perfringens isolates Vet Microbiol 2004 100 11 16 10.1016/S0378-1135(03)00126-3 15135508
3. Mertens N Theuß T Köchling M Dohmann K Lillie-Jaschniski K Pathogens detected in 205 German farms with porcine neonatal diarrhea in 2017 Vet Sci 2022 9 44 10.3390/vetsci9020044 35202297
4. Mesonero-Escuredo S Strutzberg-Minder K Casanovas C Segalés J Viral and bacterial investigations on the aetiology of recurrent pig neonatal diarrhoea cases in Spain Porcine Health Manag 2018 4 5 10.1186/s40813-018-0083-8 29632701
5. Schoch CL Ciufo S Domrachev M Hotton CL Kannan S Khovanskaya R Leipe D Mcveigh R O'Neill K Robbertse B Sharma S Soussov V Sullivan JP Sun L Turner S Karsch-Mizrachi I NCBI Taxonomy: a comprehensive update on curation, resources and tools Database 2020 2020 baaa062 10.1093/database/baaa062 32761142
6. Lee C Porcine epidemic diarrhea virus: an emerging and re-emerging epizootic swine virus Virol J 2015 12 193 10.1186/s12985-015-0421-2 26689811
7. Lee C Zakaryan H Porcine epidemic diarrhoea virus Porcine viruses: from pathogenesis to strategies for control 2019 Norfolk, UK Caister Academic Press 107 134
8. Choudhury B Dastjerdi A Doyle N Frossard JP Steinbach F From the field to the lab - An European view on the global spread of PEDV Virus Res 2016 226 40 49 10.1016/j.virusres.2016.09.003 27637348
9. Curry SM Schwartz KJ Yoon KJ Gabler NK Burrough ER Effects of porcine epidemic diarrhea virus infection on nursery pig intestinal function and barrier integrity Vet Microbiol 2017 211 58 66 10.1016/j.vetmic.2017.09.021 29102122
10. Jung K Saif LJ Porcine epidemic diarrhea virus infection: Etiology, epidemiology, pathogenesis and immunoprophylaxis Vet J 2015 204 134 143 10.1016/j.tvjl.2015.02.017 25841898
11. Lee S Kim Y Lee C Isolation and characterization of a Korean porcine epidemic diarrhea virus strain KNU-141112 Virus Res 2015 208 215 224 10.1016/j.virusres.2015.07.010 26196056
12. Lee S Lee C Genomic and antigenic characterization of porcine epidemic diarrhoea virus strains isolated from South Korea, 2017 Transbound Emerg Dis 2018 65 949 956 10.1111/tbed.12904 29770590
13. Lee S Lee DU Noh YH Lee SC Choi HW Yang HS Seol JH Mun SH Kang WM Yoo H Lee C Molecular characteristics and pathogenic assessment of porcine epidemic diarrhoea virus isolates from the 2018 endemic outbreaks on Jeju Island, South Korea Transbound Emerg Dis 2019 66 1894 1909 10.1111/tbed.13219 31055885
14. Jang G Lee KK Kim SH Lee C Prevalence, complete genome sequencing and phylogenetic analysis of porcine deltacoronavirus in South Korea, 2014–2016 Transbound Emerg Dis 2017 64 1364 1370 10.1111/tbed.12690 28758347
15. Jang G Park J Lee C Successful eradication of porcine epidemic diarrhea in an enzootically infected farm: a two-year follow-up study Pathogens 2021 10 830 10.3390/pathogens10070830 34357980
16. Park J Lee C Emergence and evolution of novel G2b-like porcine epidemic diarrhea virus inter-subgroup G1b recombinants Arch Virol 2020 165 2471 2478 10.1007/s00705-020-04767-4 32772174
17. Lee S Son KY Noh YH Lee SC Choi HW Yoon IJ Lee C Genetic characteristics, pathogenicity, and immunogenicity associated with cell adaptation of a virulent genotype 2b porcine epidemic diarrhea virus Vet Microbiol 2017 207 248 258 10.1016/j.vetmic.2017.06.019 28757031
18. Jang G Won H Lee DU Noh YH Lee SC Choi HW Yoon IJ Lee YJ Sang Yoo H Lee C Assessment of the safety and efficacy of an attenuated live vaccine based on highly virulent genotype 2b porcine epidemic diarrhea virus in nursing piglets Vet Microbiol 2019 231 120 128 10.1016/j.vetmic.2019.03.009 30955799
19. Wang X Ren W Nie Y Cheng L Tan W Wang C Wei L Zhang R Yan G A novel watery diarrhoea caused by the co-infection of neonatal piglets with Clostridium perfringens type A and Escherichia coli (K88, 987P) Vet J 2013 197 812 816 10.1016/j.tvjl.2013.05.023 23846029
20. Baek PS Choi HW Lee S Yoon IJ Lee YJ Lee DS Lee S Lee C Efficacy of an inactivated genotype 2b porcine epidemic diarrhea virus vaccine in neonatal piglets Vet Immunol Immunopathol 2016 174 45 49 10.1016/j.vetimm.2016.04.009 27185262
21. Wang B Russell ML Fonseca K Earn DJD Horsman G Van Caeseele P Chokani K Vooght M Babiuk L Walter SD Loeb M Predictors of influenza a molecular viral shedding in Hutterite communities Influenza Other Respir Viruses 2017 11 254 262 10.1111/irv.12448 28207989
22. Jung K Kim J Ha Y Choi C Chae C The effects of transplacental porcine circovirus type 2 infection on porcine epidemic diarrhoea virus-induced enteritis in preweaning piglets Vet J 2006 171 445 450 10.1016/j.tvjl.2005.02.016 16624710
23. Dors A Czyżewska-Dors E Wasyl D Pomorska-Mól M Prevalence and factors associated with the occurrence of bacterial enteropathogens in suckling piglets in farrow-to-finish herds Vet Rec 2016 179 598 10.1136/vr.103811 27655436
24. Vidal A Martín-Valls GE Tello M Mateu E Martín M Darwich L Prevalence of enteric pathogens in diarrheic and non-diarrheic samples from pig farms with neonatal diarrhea in the North East of Spain Vet Microbiol 2019 237 108419 10.1016/j.vetmic.2019.108419 31585655
25. Gould LH Limbago B Clostridium difficile in food and domestic animals: a new foodborne pathogen? Clin Infect Dis 2010 51 577 582 10.1086/655692 20642351
26. Springer S Finzel J Florian V Schoepe H Woitow G Selbitz HJ Occurrence and control of the Clostridium perfringens type A associated diarrhea of the suckling pigs with special consideration of the immunoprophylaxis Tierarztl Prax Ausg G Grosstiere Nutztiere 2012 40 375 382 23242147
27. Johannsen U Arnold P Köhler B Selbitz HJ Untersuchungen zur experimentellen Clostridium perfringens-Typ-AEnterotoxämie der Saugferkel MH Vet Med 1993 48 129 136
|
PMC010xxxxxx/PMC10202897.txt |
==== Front
Telematics and Informatics Reports
2772-5030
2772-5030
The Authors. Published by Elsevier B.V.
S2772-5030(23)00027-0
10.1016/j.teler.2023.100067
100067
Article
Investigating autistic traits, social phobia, fear of COVID-19, and internet use disorder variables in the context of videoconference fatigue
Zhang Yingying a
Yao Shuxia b
Sindermann Cornelia ac
Rozgonjuk Dmitri adg
Zhou Menghan b
Riedl René ef
Montag Christian ab⁎
a Department of Molecular Psychology, Institute of Psychology and Education, Ulm University, Ulm, Germany
b The Clinical Hospital of Chengdu Brain Science Institute, Ministry of Education Key Lab for Neuroinformation, University of Electronic Science and Technology of China, Chengdu, China
c Interchange Forum for Reflecting on Intelligent Systems, University of Stuttgart, Stuttgart, Germany
d Institute of Mathematics and Statistics, University of Tartu, Tartu, Estonia
e Digital Business Institute, University of Applied Sciences Upper Austria, Steyr, Austria
f Institute of Business Informatics-Information Engineering, Johannes Kepler University Linz, Linz, Austria
g Institute of Computer Science, University of Tartu, Tartu, Estonia
⁎ Corresponding author.
23 5 2023
9 2023
23 5 2023
11 100067100067
22 12 2022
22 4 2023
20 5 2023
© 2023 The Authors. Published by Elsevier B.V.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
In response to the coronavirus disease 2019 (COVID-19) pandemic, many individuals turned to synchronous online video communication technologies as a substitute for real-world face-to-face interactions. Evidence indicates that some users of such technologies show symptoms of exhaustion and fatigue during and after videoconferences (VCs) – this phenomenon is referred to as Videoconference Fatigue (VC fatigue). Research characterizing the possible vulnerability factors for VC fatigue is still scarce and considered to be in its early stage. Contributing to closing this gap in the existing literature is the motivation for the present study. Survey data was collected from 311 German-speaking participants to explore the relationships of VC fatigue with several psychological factors including autistic traits, social phobia, Fear of COVID-19, tendencies towards Internet Use Disorders (IUD tendencies), and Fear of Missing Out (FoMO, trait and state variables). Results showed that VC fatigue was significantly positively correlated with all of these psychological factors except state-FoMO, and corss-sectional mediation analyses provided further evidence for the positive association between autistic traits and VC fatigue. Specifically, the relationship between autistic traits and VC fatigue was mediated by Fear of COVID-19 and IUD tendencies rather than social phobia, with the latter being a preregistered hypothesis. This study adds to the literature by revealing several possible vulnerability factors associated with VC fatigue. In essence, the present work sheds light on the complex association between autistic traits and VC fatigue. We discuss the implications of our study as well as its limitations and potential avenues for future research.
Keywords
Autistic Traits
Social Phobia
Fear of COVID-19
Internet Use Disorder
Fear of Missing Out
Videoconference Fatigue
Zoom Fatigue
==== Body
pmc1 Introduction
The coronavirus disease 2019 (COVID-19), a rapidly emerging zoonotic infectious disease, has led to repeated lockdowns with social and physical distancing around the world [1]. Consequently, online video communication via videoconferencing (VC) platforms (e.g., Zoom, Skype, Teams, Webex and Facetime) became a substitute for real-world face-to-face interactions [2,3]. The use of VC platforms has increased drastically since the onset and spread of the COVID-19 pandemic, and the trend towards VC usage continues [4]. As people spend more time on VC platforms, users are increasingly concerned about being affected by the newly emerging phenomenon referred to as VC fatigue (also known as Zoom Fatigue or virtual meeting fatigue; we prefer a more general term in the form of VC fatigue) [5,6]. VC fatigue, firstly described by Degges-White in April 2020 [7], refers to the symptoms of exhaustion, burnout and social disconnection [8]. Additionally, the associations between VC fatigue and mental health symptoms such as depression and anxiety has also been investigated [9,10]. With the rise of VC platform usage and concerns related to VC fatigue, a growing number of studies have been conducted to explore the causes and factors contributing to VC fatigue [11], [12], [13]. However, research regarding VC fatigue is scarce and thus is considered to be in its early stage. It is important to understand how VC technologies can be safely used while respecting physical and mental health. This fact is substantiated by research on the potentially insidious nature of inappropriate use of VC platforms. For instance, in a short correspondence even a new diagnosis called Chronic Zoom (video meeting) Syndrome was proposed [14]. Against this background, it is critical to explore and understand variables that potentially contribute to VC fatigue. In this context, a newly developed 4D conceptual model of VC fatigue is of interest. It states that four key causal dimensions are of relevance to understand VC fatigue: (1) personal factors, (2) organizational factors, (3) technological factors and (4) environmental factors [12]. Personal factors mainly include general individual factors (sociodemographic variables such as gender, age, ethnicity; personality traits, and cognitive traits) and VC-specific individual factors (mental/physical health and fitness, and stress management skills) [12]. Personal factors were introduced to better understand why people differed in their VC fatigue from the perspective of individual differences. Besides, previous studies indicated that cognitive factors were likely to contribute to VC fatigue, because an increased ability to virtually multitask might threaten attentional capacity [[15], [16], [17], [18], [19]]. This said, due to reduced non-verbal cues being transmitted via VC (e.g., body language and facial expressions), a greater need to concentrate, and the new experience of the very close proximity of facial images [15], [16], [17], [18], using VC platforms could be more psychologically demanding than real-world face-to-face interactions [19]. Considering these new demands, we deem it to be of relevance to focus on psychological factors causing VC fatigue that help to establish healthy VC use. However, to our knowledge, only socio-demographic variables and personality traits (Big Five of Personality) have been investigated in the context of VC fatigue. In essence, the results of these studies indicated that women, younger, and more neurotic individuals were more prone to suffer from higher levels of VC fatigue [13,20]. The present work aims to go beyond these variables by exploring further psychological factors that might be related to VC fatigue.
1.1 Literature backgrounds
1.1.1 The relationships between autistic traits, social phobia and VC fatigue
Autistic traits are continuously distributed in the general population, with the most severe end of the continuum being associated with clinical recognition of autism spectrum disorder (ASD) [21], [22], [23]. Deficits in social interaction and communication, as well as restricted, repetitive patterns of behavior, are emphasized by the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) as symptoms of ASD [24]. Such characteristics can be conceptualized as being distributed along a spectrum of impairments ranging from mild to very severe levels in ASD [25] and are also found in a subclinical population with autistic traits [21,26,27]. Moreover, some endogenous or co-occurring characteristics associated with ASD, such as lower social cognition and lower cognitive flexibility, have also been found among individuals with higher autistic traits in the general population [28]. We assumed the psychological dimension of autistic traits to be of relevance to understand individual differences in VC fatigue. This is because, during VC, users have to draw upon their cognitive processing capabilities to manage information from multimodal channels (e.g., audio, video, images, and/or text, as well as non-verbal cues). For individuals with higher autistic traits while participating in VCs, such cognitive (social) processing demands might be more challenging and could result in experiencing higher levels of cognitive overload, which has already been linked to VC fatigue [13]. Furthermore, in light of the literature, it is of interest that the personality domain of higher trait neuroticism (lower emotional stability) was shown to be related to higher VC fatigue [19] and higher neuroticism was also associated with higher autistic traits [29]. Therefore, it is likely that higher autistic traits might be linked to higher levels of VC fatigue. Following from this, we hypothesized that autistic traits might be a possible vulnerability factor related to VC fatigue (see the preregistration Open Science Framework (OSF), https://osf.io/wr7sb).
In recent years, it has been hypothesized that individuals with higher autistic traits may be at a greater risk of developing psychiatric conditions [30], [31], [32]. In the context of our work, individuals with higher autistic traits may be in particular more vulnerable to the threat of the COVID-19 pandemic. In fact, a recent study has demonstrated that autistic traits predicted increased negative emotions such as anxiety, anger and fear responses to COVID-19 [33,34]. Among anxiety disorders, social phobia is defined as a persistent fear of social situations that involve interactions with others. These interactions relate, in particular, to those with unfamiliar people or to being put into situations with possible scrutiny by others [24]. The cognitive model of social phobia suggests that individuals with higher social phobia tend to pay more attention to themselves than to others in social situations [35]. For example, they focus on ensuring that they are not embarrassed and what they are going to say next, rather than listening to what other people are saying [35]. Consequently, during a VC individuals with higher social phobia might pay more attention to irrelevant information, which slows the speed of information processing [36]. Therefore, people with higher social phobia likely need to invest more cognitive effort in processing (social) information from multimodal channels during VCs, which might result in higher VC fatigue. Because, as mentioned above, autistic traits predicted increased anxiety responses to COVID-19 [33], we hypothesized that social phobia may mediate the relationship between autistic traits and VC fatigue (see preregistration OSF: https://osf.io/wr7sb).
1.1.2 The relationships between autistic traits, Fear of COVID-19, Internet Use Disorder tendencies and VC fatigue
The unprecedented nature of COVID-19 (such as its high transmission, infectivity, and mortality) and uncertainties arising from the pandemic have affected various aspects of health and well-being. In particular, Fear of COVID-19 became a central construct to understand psychological responses to the pandemic [37,38]. Of interest, individuals with higher autistic traits might be more prone to experience Fear of COVID-19 during the pandemic because it was more difficult for such individuals to adjust to the myriad COVID-induced changes in everyday life [39]. It has also been reported that autistic traits may go along with a higher general fear response to the COVID-19 pandemic, which was assessed using the Positive and Negative Affect Schedule [33,40]. Adding to this, a growing body of research showed positive associations between Fear of COVID-19 and burnout (a syndrome of feelings including emotional exhaustion, depersonalization, and low personal accomplishment [41], [42], [43], [44]). For instance, Çalışkan and Kargın (2021) found that Fear of COVID-19 was significantly associated with the level of burnout and occupational fatigue (mental and physical toll of workplace fatigue) [42]. Therefore, we expected that individuals with higher autistic traits would show higher Fear of COVID-19, and be more likely to experience higher VC fatigue, accordingly.
According to the Compensatory Internet Use Theory shedding light on Internet Use Disorder (IUD) [45], Internet use as a means to alleviate negative emotions might have paved the way for the development of IUD [46,47]. Excessive Internet users are also likely to be more exposed to rapidly spreading misinformation and unfounded fears via social media. It is not surprising that recent studies observed a positive relationship between Fear of COVID-19 and IUD tendencies [48,49]. As the present study focuses on VC fatigue, we also highlight recent works that established links between autistic traits and IUD tendencies [50], [51], [52], and between IUD and fatigue including physical and mental health symptoms [53,54]. Therefore, we investigated these vulnerability factors in one study and explored the possible relationships between these vulnerability factors (autistic traits, Fear of COVID-19, and IUD tendencies) and VC fatigue as the exploratory nature of the study.
1.2.3 The relationships between Fear of Missing Oout and VC fatigue
IUD and problematic technology use have repeatedly been shown to correlate with Fear of Missing Out (FoMO) [55,56]. FoMO is defined as a ‘a pervasive apprehension that others might be having rewarding experiences from which one is absent’ ( [57], p.1841). During the pandemic, with physical distancing at home, individuals likely continued to experience FoMO due to the difficulties in catching up with real-time social media content, others’ posts and videos, and newly released movies and series [58]. As a consequence, FoMO might threaten the well-being by causing psychologically and health issues including sleep deprivation, loss of focus, and lower motivation to fulfill daily responsibilities [58]. Furthermore, a recent study reported that FoMO was positively significantly associated with both communication and information overload, with two variables inherently related to the concept of fatigue [59]. Expanding these findings to VC usage, we wanted to explore that FoMO might be positively correlated with the experience of VC fatigue.
1.3 Aims and hypotheses
The present study investigated a series of possible vulnerability factors for VC fatigue. For all variables, we expected positive associations with VC fatigue (except for the relation of FoMO with VC fatigue, which was investigated in an exploratory way; see preregistration OSF, https://osf.io/wr7sb). In our preregistration, we also hypothesized that social phobia might be a mediator in the relationship between autistic traits and VC fatigue. Against the background of the discussed literature, we further explored possible relationships between autistic traits and VC fatigue, and whether this relationship would be mediated via Fear of COVID-19 and IUD tendencies as exploratory parts of the study.
2 Materials and methods
2.1 Procedure
The present research project including (some of the) hypotheses, study design, sample size, and analysis plan was preregistered (OSF, https://osf.io/wr7sb). Data will be made available on the OSF website after the paper has been accepted for publication.
This cross-sectional study was implemented on the SurveyCoder platform (www.ckannen.com) in German language and data collection took place between May 2021 and March 2022. Data collection was promoted by students, via social media and other media appearances by the researchers. The inclusion criteria for participation were as follows: (1) Internet access, (2) a minimum age of 18 years, (3) German-speaking, (4) being a VC platforms user (e.g., Zoom, Skype, Teams, Webex or Facetime) in areas such as work, education, healthcare or social meetings, and (5) providing informed electronic consent prior to taking part in the study. This research project was approved by the local ethics committee at Ulm University, Ulm, Germany.
2.2 Participants
A total of N = 315 participants provided data for the present study. However, n = 1 subject was excluded because of being younger than 18 years and thus not being eligible for study participation. Moreover, n = 3 subjects were excluded due to invalid data (i.e., invalid consent, not using VC tools, or careless response to the online survey). Ultimately, the effective sample comprised N = 311 subjects (Age: 29.40 ± 11.22, 18 – 74 years; gender: 112 men, 197 women, 2 diverse). Of note, the here-analyzed participants have been included already in another study investigating the interplay between aspects of VC, personality traits, depression and burnout (being separately preregistered)[20]. Participants’ information including nationality, educational levels and VC tools used are presented in Table 1 . Among the effective samples, Zoom (87.5%) was the most popular VC tool used.Table 1 Participants’ information including nationality, educational level, and VC platforms used.
Table 1 Frequency % Frequency %
Nationality VC tools used
Germany 171 55.0% Zoom 272 87.5%
Austria 86 27.7% Microsoft-Teams 198 63.7%
Others1 54 17.3% Skype 151 48.6%
Education level CiscoWebEx 138 44.4%
Secondary school leaving certificate 16 5.1% Facetime 89 28.6%
A-level/high school diploma 115 37.0% BBB 91 29.3%
University of applied sciences degree 55 17.7% Jitsi 44 14.1%
University degree 116 37.3% GoTo 38 12.2%
Others2 9 2.9% Others3 72 23.2%
Notes.
1 participants from other countries and participants who had missing data in this item.
2 including the education levels “no graduation” and “vocational baccalaureate diploma” (Of note: the option “streamed secondary education for lesser able students” (German: “Hauptschule”) was not chosen by any participant).
3 other kinds of VC platforms not listed above-mentioned; participants were allowed to choose more than one kind of VC tool (i.e., VC tools were not mutually exclusive).
2.3 Questionnaires
2.3.1 The Zoom Exhaustion & Fatigue Scale
For the present research, the German version of the Zoom Exhaustion & Fatigue scale [20] (ZEF, original in English by [60]) was adopted to determine VC fatigue levels. The ZEF scale has 15 items with the following answer format: 13 items: from “1 = not at all” to “5 = extremely”; 2 items: from “1 = never” to “5 = always”. The total ZEF scores can range from 15 to 75, and higher scores indicate higher VC fatigue levels. In addition, the ZEF scale was built on five dimensions, including general fatigue, visual fatigue, social fatigue, motivational fatigue, and emotional fatigue, which form a higher-order factor of VC fatigue [60]. Therefore, the 15-item scores are summed to form one total index of VC fatigue. The computed Cronbach's alphα of the total ZEF scores was 0.926, and the alphαs for the ZEF subscales were as follows in the present sample: general, α = 0.617; visual, α = 0.715; social, α = 0.732; motivational, α = 0.801; emotional, α = 0.715.
2.3.2 A short form of the Autism Spectrum Quotient
Individual differences in the levels of autistic traits were assessed by a short form of the Autism Spectrum Quotient (AQ-k), which was created by Freitag et al. (2017) containing 33 items [61], which were derived from an original 50-item version of the AQ scale [21]. The AQ-k is a reliable and valid self-assessment instrument for individuals with normal intelligence and a minimum age of 16 years and cannot be used for the diagnosis of ASD [61]. Thus, this tool is suitable to measure individual differences in a subclinical level of autistic traits in the general population. Each item is answered via a 4-point Likert scale (from “1 = definitely agree” to “4 = definitely disagree”). The response to each item is coded using a binary system (0/1), where an endorsement of the autistic trait is scored as 1, while other responses are scored as 0 [21]. Total AQ-k scores can range from 0 to 33 with higher scores reflecting the presence of more severe symptoms of autistic traits. Internal consistency of the AQ-k was acceptable (α = 0.742).
2.3.3 The Mini-Social Phobia Inventory
The Mini-Social Phobia Inventory (mini-SPIN), a reliable and valid instrument [62], was used to assess individual differences in social phobia. The questionnaire with 3 items was administered with a 5-point scale (from “0 = not at all” to “4 = extremely”). A very short form of the mini-SPIN questionnaire with sufficient psychometric properties is hardly time-consuming during the assessment, thus more efficient for our survey. Total scores potentially range between 0 and 12. Internal consistency of the Mini-SPIN was acceptable (α = 0.776).
2.3.4 The Fear of COVID-19 Scale
The Fear of COVID-19 Scale (FCV-19S) with 7 items in the German version as put forward in a recent work [63] was used to assess individual differences in fearing COVID-19. The items of the FCV-19S were answered on a 5-point Likert scale (from “1 = strongly disagree” to “5 = strongly agree”). Total scores range between 7 and 35, and a higher score indicates a higher Fear of COVID-19. The internal consistency estimating for the FCV-19S was acceptable (α = 0.767).
2.3.5 The short Internet Addiction Test
In the present work, the German version of the Short Internet Addiction Test (s-IAT) as proposed by Stodt et al. (2018) was used [64]. The s-IAT consists of 12 items, and answers are based on a 5-point response scale (from “1 = rarely” to “5 = very often”). Total scores of the s-IAT can vary between 12 and 60 points. Higher scores indicate more severe problems because of one's own Internet use. The s-IAT revealed a good internal consistency (α = 0.867).
2.3.6 The Fear of Missing Out scale
Finally, the Fear of Missing Out scale (FoMO) developed by Wegmann et al. (2017), assessing individual differences in FoMO, was used in the current study [65]. It comprises 12 items to be answered on a 5-point Likert response scale (from “1 = totally disagree” to “5 = totally agree”), with two dimensions called trait-FoMO and state-FoMO. Trait-FoMO refers to a stable individual characteristic in everyday life, whereas state-FoMO refers to FoMO in the realm of online content and interactions with others [56]. Scores of the trait dimension range between 5 and 25 points, and the state dimension ranges between 5 and 35 points, with higher scores indicating more severe of trait-/state-FoMO. The internal consistency coefficients of these two dimensions were acceptable (α = 0.783 (trait)/0.772 (state)).
2.4 Statistical analyses
All variables were analyzed using the SPSS software (version 26.0; IBM Corp., Armonk, NY, USA). Measures of normality (i.e., skewness and kurtosis) were initially computed. According to guidelines by Kim H. Y. (2013) [66], only the s-AQ scores and FCV-19S scores showed skewness and kurtosis exceeding ± 2 in the total sample, which implies a violation of the normality assumption (details see Table 1). However, as the deviations were of minor nature, parametric tests were used for all statistic calculations (note that results did not change meaningfully depending on whether parametric or non-parametric tests were implemented).
First, gender differences in variables of interest were investigated using independentsamples t-tests, and potential associations with age were computed via Pearson correlations to check if these socio-demographic factors needed to be controlled in the main analyses. Second, we computed Zero-order Pearson correlations to determine the associations among these variables of interest, along with age; p < 0.05 was considered statistically significant (although we had directed hypotheses, we stay with a conservative p < 0.05 level).
Finally, mediation models (in part depending on our hypotheses) were conducted using the PROCESS macro [67], and socio-demographic factors (age and gender) were included as covariates in all models. In the first model, autistic traits were entered as the predictor, social phobia as the mediator, and VC fatigue or its dimensions (measured by the total ZEF scores or ZEF subscales) as the outcome variable (PROCESS template model 4; see Fig. 1 A). This model was preregistered. In the second model, the predictor and outcome variables were the same as those in the first model, but the mediators were replaced by Fear of COVID-19 and IUD tendencies (PROCESS template model 6, see Fig. 1B). To test the mediation effects, 5000 bootstrap resamples and 95% bias-corrected confidence intervals (CI) were applied to construct indirect paths. If the 95% CI did not include 0, the indirect effect was considered statistically significant [68].Fig. 1 caption is as follows: Fig. 1. The mediation models. (A) represents the single mediator model. a, b, and c' are direct effects, whereas c is the total effect (direct effect + indirect effect); (B) represents the multiple mediators model. a1, a2, b1, b2, d1, and c' are direct effects, whereas c is the total effect (direct effect + indirect effect).
Fig. 1
3 Results
3.1 Descriptive statistics and gender differences among study variables
Skewness and kurtosis and means and standard deviations for all key variables are reported in Table 2 . Results revealed that subjects on average reported medium scores on autistic traits, social phobia, Fear of COVID-19, IUD tendencies, trait-/state-FoMO, and VC fatigue when taking into account the possible range of scores on each scale. In addition, gender differences were observed: women showed significantly higher levels of VC fatigue (p < 0.001), as well as all facets (general fatigue, p = 0.005; visual fatigue, p = 0.027; social fatigue, p < 0.001; motivational fatigue, p = 0.002; emotional fatigue, p = 0.006), Fear of COVID-19 (p < 0.001), and trait-/state-FoMO (trait, p = 0.022; state, p = 0.004) than men (see Table 2).Table 2 Distribution parameters, means and standard deviations for the variables of interest for the total sample and separated by gender in addition to t-test results on gender differences.
Table 2Variables skewness kurtosis Total
(n = 311) Men
(n = 112) Women
(n = 197) t p
Age 1.45 –- 1.57 29.40 (11.22) 31.36 (11.79) 28.38 (10.78) 2.26 0.025
ZEF scores
total (VC fatigue) 0.40 −0.01 36.65 (11.61) 33.58 (12.12) 38.27 (10.95) −3.48 < 0.001
general fatigue 0.22 −0.28 7.88 (2.62) 7.31 (2.75) 8.18 (2.49) −2.83 0.005
visual fatigue 0.51 −0.17 7.09 (2.63) 6.64 (2.87) 7.32 (2.43) −2.22 0.027
social fatigue 0.68 −0.07 6.64 (2.78) 5.77 (2.61) 7.11 (2.77) −4.17 < 0.001
motivational fatigue 0.30 −0.46 7.51 (2.68) 6.87 (2.85) 7.85 (2.49) −3.16 0.002
emotional fatigue 0.41 −0.17 7.53 (2.55) 6.99 (2.56) 7.82 (2.50) −2.75 0.006
AQ-k (autistic traits) 0.88 1.00 10.07 (4.50) 10.23 (4.65) 9.91 (4.31) 0.61 0.545
Mini-SPIN (social phobia) 0.06 −0.99 8.62 (3.21) 7.71 (3.23) 9.10 (3.10) −3.71 < 0.001
FCV-19S (Fear of COVID-19) 0.93 0.78 12.15 (4.11) 11.03 (3.70) 12.81 (4.21) −3.73 < 0.001
s-IAT (IUD tendencies) 0.51 −0.17 25.93 (7.65) 25.05 (8.09) 26.31 (7.32) −1.39 0.164
trait-FoMO 0.19 −0.59 14.06 (4.35) 13.31 (4.28) 14.49 (4.35) −2.31 0.022
state-FoMO 0.56 0.09 14.88 (4.98) 13.78 (4.68) 15.49 (5.07) −2.94 0.004
Notes. ZEF = the Zoom Exhaustion & Fatigue scale; AQ-k = short form of Autism Spectrum Quotient; Mini-SPIN = Mini-Social Phobia Inventory; FCV-19S = Fear of COVID-19 Scale; s-IAT = short Internet Addiction Test; FoMO = Fear of Missing Out Scale. 2 subjects’ data with diverse gender identities were not included in the calculation for gender differences.
3.2 Correlation analyses
The correlations between age and all variables of interest were computed and the results indicated that age was negatively significantly correlated with all variables except autistic traits (p = 0.507) and Fear of COVID-19 (p = 0.205). Therefore, partial correlations controlling for age among all variables of interest were calculated. Table 3 presents the partial correlations between VC fatigue and other variables of interest controlling for age (without any correction for multiple testing). Zero-order Pearson's correlations revealed that VC fatigue exhibited significant positive correlations with all variables of interest (autistic traits (p = 0.013), social phobia (p = 0.002), Fear of COVID-19 (p < 0.001), IUD tendencies (p < 0.001), and trait-FoMO (p < 0.001)) except state-FoMO (p = 0.057). After applying a Bonferroni correction for multiple testing (alpha = 0.05/21 = 0.0024; divided by 21 because 21 correlations were calculated), the correlations between VC fatigue and other variables of interest remained significant, except for autistic traits. In addition, partly in line with our expectations and hypotheses, even after multiple testing corrections, autistic traits correlated significantly and positively with social phobia (p < 0.001), IUD tendencies (p = 0.001), and Fear of COVID-19 correlated significantly and positively with IUD tendencies (p < 0.001). We additionally calculated correlations between VC fatigue facets and variables of interest. Those results are presented in Supplementary Table 1. Results indicated that VC fatigue facets were also significantly and positively correlated with all other variables of interest.Table 3 Zero-order Pearson correlations among variables of interest.
Table 3 1 2 3 4 5 6 7
1 total ZEF
(VC fatigue)
2 AQ-k
(autistic traits) 0.14*
3 Mini-SPIN
(social phobia) 0.18⁎⁎ 0.44⁎⁎⁎
4 FCV-19S
(Fear of COVID-19) 0.25⁎⁎⁎ 0.15⁎⁎ 0.18⁎⁎
5 s-IAT
(IUD tendencies) 0.35⁎⁎⁎ 0.19⁎⁎⁎ 0.18⁎⁎ 0.19⁎⁎⁎
6 trait-FoMO 0.30⁎⁎⁎ 0.14* 0.32⁎⁎⁎ 0.28⁎⁎⁎ 0.38⁎⁎⁎
7 state-FoMO 0.11 −0.04 0.15* 0.15⁎⁎ 0.39⁎⁎⁎ 0.41⁎⁎⁎
Notes. All correlations were controlled for age.
⁎ , p < 0.05.
⁎⁎ , p < 0.01.
⁎⁎⁎ , p < 0.001. ZEF = the Zoom Exhaustion & Fatigue scale, AQ-k = short form of Autism Spectrum Quotient, Mini-SPIN = Mini-Social Phobia Inventory, FCV-19S = Fear of COVID-19 Scale, s-IAT = short Internet Addiction Test, FoMO = Fear of Missing Out scale. Of the significant (* p <0.05) correlations reported in this table, the correlations between VC fatigue and autistic traits (p = 0.013), autistic traits and Fear of COVID-19 (p = 0.007), autistic traits and trait-FoMO (p = 0.017), social phobia and state-FoMO (p = 0.010), and Fear of COVID-19 and state-FoMO (p = 0.009) did not remain significant after manually applying a Bonferroni correction for multiple testing (alpha = 0.05/21 = 0.0024).
3.3 Direct and indirect effects of autistic traits on VC fatigue via social phobia
The first mediation model tested how autistic traits were related to VC fatigue via the possible mediator social phobia. The mediation effect of social phobia on the relationship between autistic traits and VC fatigue (measured by the total ZEF scores) was assessed using PROCESS template model 4 (see Fig. 1A), and socio-demographic variables (age and gender) were included as covariates. As presented in Table 4 , the total effect of autistic traits on VC fatigue was significant (c = 0.41, p = 0.006), providing further evidence for autistic traits being a (coross-sectional) predictor for VC fatigue. The effect of autistic traits on social phobia was significant, indicating that higher autistic traits were associated with higher social phobia (a = 0.31, p < 0.001). The effect of social phobia on VC fatigue (b path) was not significant (b = 0.35, p = 0.137), and the 95% CI of the indirect effect (a*b = 0.11, 95% CI [−0.034, 0.259]) included 0, indicating that social phobia did not significantly mediate the relationship between autistic traits and VC fatigue.Table 4 Mediating effect of social phobia on the association between autistic traits and VC fatigue.
Table 4 β SE t p 95% CI
lower upper
Direct effect
a 0.31 0.04 8.96 < 0.001 0.245 0.384
b 0.35 0.24 1.49 0.137 −0.112 0.815
c’ 0.29 0.16 1.80 0.073 −0.027 0.612
Indirect effect
a*b 0.11 0.07 – – −0.034 0.259
Total effect
c 0.41 0.14 2.78 0.006 0.118 0.688
Notes. N = 309 (two participants with diverse gender identities were not included), controlled for age and gender. CI = confidence intervals. Bootstrap confidence intervals were constructed using 5000 resamples and the 95% bias-corrected confidence intervals.
Moreover, mediation models with social phobia as the mediator between autistic traits and VC fatigue facets were calculated and are presented in Supplementary Table 2. Results indicated that the effect of autistic traits on visual fatigue was fully mediated by social phobia (c’ = 0.05, p = 0.172, a*b = 0.04, 95% CI [0.006, 0.076]), the effect of autistic traits on emotional fatigue was significant (c’ = 0.08, p = 0.024), while there was not a significant mediation effect by social phobia (a*b = 0.03, 95% CI [−0.005, 0.060]). Finally, the remaining mediation models with general, social and motivational fatigue as outcome variables showed that the mediation effect via social phobia did not exist.
3.4 Direct and indirect effects of autistic traits on VC fatigue via Fear of COVID-19 and IUD tendencies
The second mediation model examined the indirect effects via the mediators Fear of COVID-19 and IUD tendencies. Fig. 1B depicts each pathway between autistic traits and VC fatigue with multiple mediators and results are presented in Table 5 . The effects of autistic traits on Fear of COVID-19 and IUD tendencies were both significant, indicating that higher autistic traits were associated with higher Fear of COVID-19 (a1 = 0.16, p = 0.002) and higher tendencies towards IUD (a2 = 0.25, p = 0.007). Higher Fear of COVID-19 was associated with higher tendencies towards IUD (d1 = 0.30, p = 0.003). The indirect effect of the two-mediator sequential pattern was significant, as indicated by the fact that the 95% CI did not include 0 (path 3, a1*d1*b2 = 0.02, 95% CI [0.005, 0.050]), providing evidence that autistic traits exerted a significant indirect effect on VC fatigue via Fear of COVID-19 and IUD tendencies, accounting for a total effect of 5.0%. Besides, the mediation model simultaneously tested two alternative single-mediator pathways separately. First, it tested whether the effect of autistic traits on VC fatigue was mediated by Fear of COVID-19 alone. The indirect effect was significant (path 1, a1*b1 = 0.07, 95% CI [0.012, 0.148]), accounting for 17.5% of the total effect. Second, the model tested whether the effect of autistic traits on VC fatigue was mediated by IUD tendencies alone. The indirect effect was again significant (path 2, a2*b2 = 0.12, 95% CI [0.028, 0.237]), which accounted for a total effect of 30.0%. Taken together, the independent mediating mechanisms of Fear of COVID-19 and IUD tendencies and a chain-mediating mechanism from Fear of COVID-19 to IUD tendencies in the relationship between autistic traits and VC fatigue were significant.Table 5 Multiple mediation effects of Fear of COVID-19 and IUD tendencies in the association between autistic traits and VC fatigue.
Table 5 β SE t p 95% CI
lower upper
Direct effect
a1 0.16 0.05 3.08 0.002 0.057 0.260
a2 0.25 0.09 2.73 0.007 0.069 0.423
d1 0.30 0.10 3.05 0.003 0.107 0.494
b1 0.45 0.15 2.92 0.004 0.146 0.748
b2 0.48 0.09 5.44 < 0.001 0.305 0.650
c’ 0.19 0.14 1.38 0.170 −0.083 0.466
Indirect effect
path 1: a1*b1 0.07 0.03 – – 0.012 0.148
path 2: a2*b2 0.12 0.05 – – 0.028 0.237
path 3: a1*d1*b2 0.02 0.01 – – 0.005 0.050
Total effect
c 0.40 0.14 2.78 0.006 0.118 0.688
Notes. N = 309 (two participants with diverse gender identidies were not included), controlled for age and gender. CI = Confidence intervals. Bootstrap confidence intervals were constructed using 5000 resamples and the 95% bias-corrected confidence intervals.
In the same way with the PROCESS template model 6, mediation models with Fear of COVID-19 and IUD tendencies being potential mediators between autistic traits and VC fatigue facets were also calculated and are presented in Supplementary Table 3. Similarly significant multiple mediation models were found when visual fatigue or motivational fatigue was entered as the outcome variable, indicating that the relationship between autistic traits and visual fatigue or motivational fatigue was fully mediated by Fear of COVID-19 and IUD tendencies (visual, c’ = 0.05, p = 0.139, c = 0.09, p = 0.007; motivational c’ = 0.04, p = 0.198, c = 0.09, p = 0.006). Moreover, results revealed that the relationship between autistic traits and emotional fatigue was partially mediated by Fear of COVID-19 and IUD tendencies (c’ = 0.07, p = 0.021, c = 0.11, p = 0.001). There were no multiple mediation effects on general or social fatigue dimensions.
4 Discussion
In the present study, we examined factors of potential relevance for the emergence of VC fatigue. In line with the literature, women suffered more from VC fatigue than men, and younger participants were more prone to experience VC fatigue [13,60] (see also the paper by Montag et al. (2022) with an overlapping dataset with the present work making this observation [20]). Zero-order correlation analyses demonstrated that VC fatigue correlated with all variables of main interest for the present work (i.e., autistic traits, social phobia, Fear of COVID-19, IUD tendencies, and trait-FoMO) except state-FoMO. Mediation models provided evidence for the positive association between autistic traits and VC fatigue, and this relationship was mediated by Fear of COVID-19 and IUD tendencies. In the following section, we discuss our findings in more detail and compare them with our preregistration and findings observed in previous literature (when available).
As hypothesized, we found that individuals with higher autistic traits reported more VC fatigue. A similar result was found in a recent study in which greater autistic characteristics predicted personal burnout among non-autistic students [69]. Possible explanations were the aforementioned cognitive and executive dysfunctions and sensory hypersensitivity in individuals with higher autistic traits [70,71]. All in all, there is still too little evidence linking autistic traits to fatigue, in particular to VC fatigue in the general population. Therefore, more work is needed in this area. Nonetheless, research on ASD patients related to the use of VC technologies has been conducted [72], [73], [74]. For instance, online healthcare conducted via VCs in order to provide remote clinical healthcare services has shown to be an accessible option for ASD treatment. However, it is also apparent that barriers to telehealth delivery for ASD were experienced by the majority, such as challenges associated with sensory sensitivities, body awareness, processing speed, and nonverbal communication [73]. Aside from this, research investigating difficulties in remote learning (classroom community online) in children with ASD suggested that lower levels of engagement, as well as increased levels of distractibility and VC fatigue, were often reported by autistic children during the pandemic [75]. In this context, we mention that also the term “autistic burnout” has been proposed [76], [77], [78], [79]. Common features of autistic burnout include impaired cognitive function, the loss of previously acquired skills (i.e., speech or self-care), loss of focus and concentration, social and sensory withdrawal and a marked increase in observable autistic traits [77,78]. Unique characteristics of ASD (e.g., social interaction difficulties and sensory sensitivities) appear to drive autistic burnout [76]. As mentioned above, such characteristics can be conceptualized as being distributed along a spectrum of impairments ranging from mild to very severe levels in ASD [25] and are also found within autistic traits [21,26,27]. We expected that individuals with higher autistic traits may more easily show ‘autistic burnout’ symptoms, which might be a possible explanation for the relationship between autistic traits and VC fatigue in the general population. However, more research is needed to confirm this notion. Taken together, individuals with higher autistic traits or ASD are likely to experience more difficulties during online communication and are also more likely to be fatigued by VCs.
Anti-COVID-19-pandemic measures (e.g., self-isolation and social distancing) forced the public to make rapid changes in daily living habits. In this context, it is important to mention that autistic traits are associated with lower tolerance of uncertainty and unexpected changes in everyday life, making it more difficult for individuals with high autistic traits to effectively manage pandemic-related changes [39]. As a consequence, individuals with higher autistic traits are likely to experience more negative emotions and particularly more pandemic related anxiety and fear [33]. At least in parts consistent with our expectations, we observed in our mediation model that autistic traits predicted increased social phobia in our present study (note that we are aware of the cross-sectional nature of our study). This said, we also expected in our preregistration that the association between autistic traits and VC fatigue might be mediated by social phobia, but this was not the case in the present study and we also highlighted the relatively small effect size of the association between autistic traits and VC fatigue (r = 0.14). Taken together, our results expand the understanding of the relationship between autistic traits and VC fatigue in the general population and provide an explanatory mechanism.
With the exploratory nature of the study (not preregistered at the OSF website), independent mediating mechanisms of both Fear of COVID-19 and IUD tendencies, as well as a chain-mediating mechanism via Fear of COVID-19 and IUD tendencies in the relationship between autistic traits and VC fatigue were identified. First, the results of the present study not only revealed that individuals with higher autistic traits reported increased Fear of COVID-19, but also that the link between autistic traits and VC fatigue was mediated by Fear of COVID-19. This might be due to cognitive and executive dysfunctions of autistic traits [28], as well as due to increased burnout and fatigue tendencies as a result of fear COVID-19 [42]. These results are, at least in part, further supported by a recent study showing that the level of fear experienced by individuals with higher autistic traits was significantly more pronounced compared to individuals with lower autistic traits during the COVID-19 pandemic [34]. Second, we observed a positive association between autistic traits and IUD tendencies [50], [51], [52], and the relationship between autistic traits and VC fatigue was mediated by IUD tendencies. Third, a chain-mediating mechanism via Fear of COVID-19 and IUD tendencies in the relationship between autistic traits and VC fatigue was identified in the present work. Although existing research has not yet explicitly linked these vulnerability factors with the experience of VC fatigue, it might be plausible that autistic traits mark the beginning of the causal chain-mediation mechanism as these traits are known to be rather stable [80,81]. As previously pointed out, individuals with higher autistic traits are more prone to experience higher levels of negative emotionality, such as fear, especially during the COVID-19 pandemic [33], and also show higher tendencies towards IUD [50], [51], [52]. It is noteworthy that the positive relationship between Fear of COVID-19 and IUD tendencies was also observed in the multiple mediation model, which is consistent with the growing body of research [49,82,83] and might be explained by the idea that individuals being stressed out due to the pandemic might seek distraction online, which – with excessive use patterns – can result in IUD tendencies. Hence, we inferred from the present data that the experience of higher VC fatigue might occur, particularly among individuals with higher autistic traits, which likely accompanies higher Fear of COVID-19 and higher IUD tendencies. However, causal conclusions based on our cross-sectional data were not possible. These associations observed here need to be replicated and also be investigated in longitudinal studies, as well as in studies with experimental manipulation and effect measurements.
In the present study, we found that trait-FoMO was more robustly associated with VC fatigue than state-FoMO. Trait-FoMO has been described as a general FoMO in the context of any social experiences (not exclusively linked to the online area), whereas state-FoMO represents specific cognition that develops during online activities, especially Internet communication [65,84]. Our observations of the positive relationship between trait-FoMO and VC fatigue rather than state-FoMO are, at first glance, presumably surprising, as state-FoMO touches upon FoMO in the online area, and videoconferences also belong to the area of online communication. A possible explanation is that several of the state-FoMO items assess FoMO in the context of social networks, yet, and videoconferences do not belong to this category.
It is essential to keep the limitations of the present study in mind when interpreting the results. First, given the well-known biases when participants answer to self-report inventories, further studies should be conducted using more objective methodologies. Second, the investigated relationships between possible vulnerability factors and VC fatigue have been only studied in a German-speaking setting. It is unclear whether these results could generalize to other cultural or geographic areas. Again, we emphasize that the multiple mediation model has not been preregistered, hence these findings are explorative and need to be revisited in future works. Finally, the present work is of cross-sectional nature and cannot disentangle cause and effect.
5 Conclusion and future research directions
In conclusion, this study reports several novel findings regarding the relationships between autistic traits, social phobia, Fear of COVID-19, trait/state-FoMO, IUD tendencies, and VC fatigue. Our results offer insights into a series of possible vulnerability factors being more or less linked to VC fatigue, improving these factors should reduce the susceptibility to VC fatigue. Considering that videoconferences may continue to be an integral aspect of our future life, both in professional and private contexts, and that research on VC fatigue is still in its early stage, more research is needed to understand the nature, causes, and consequences of VC fatigue. For example, further research can explore additional potential factors which might be related to VC fatigue such as depression might be a potential factor for VC fatigue. Depressiveness often includes tiredness and a lack of energy [24], as a consequence, individuals who feel more depressed may be more exhausted and fatigued. Additionally, the influence of VC fatigue on mental health (e.g., loneliness and depression) has also been investigated: Higher levels of VC fatigue correlated with higher levels of loneliness and depression [85]. However, the complex mechanisms underlying the relationship between VC fatigue and mental health remains unclear. There may be individual differences to find a sense of connection and belonging during VC interactions, which may, in turn, increase both VC fatigue and loneliness. Accordingly, increased loneliness may lead to depressive symptoms. Additional and longitudinal research is necessary to understand the machine between VC fatigue and mental health.
Funding
This research did not receive any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.
Author contributions
Christian Montag and YingYing Zhang designed the present study. YingYing Zhang conducted the statistical analysis and wrote the first version of the manuscript. Menghan Zhou double checked the statistical analysis. Christian Montag, Shuxia Yao, Cornelia Sindermann, René Riedl and Dmitri Rozgonjuk critically revised the manuscript and approved the final version of the manuscript for submission.
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Appendix Supplementary materials
Image, application 1
Data availability
Data will be available on the OSF website: https://osf.io/wr7sb.
Acknowledgements
We are grateful to all subjects who participated in the study. Dr. Cornelia Sindermann acknowledges the support of the Ministerium für Wissenschaft, Forschung und Kunst Baden-Württemberg (MWK, Ministry of Science, Research and the Arts Baden-Württemberg under Az. 33–7533–9–19/54/5) in Künstliche Intelligenz & Gesellschaft: Reflecting Intelligent Systems for Diversity, Demography and Democracy (IRIS3D) and the support by the Interchange Forum for Reflecting on Intelligent Systems (IRIS) at the University of Stuttgart. The authors would also like to thank the editors and the reviewers for their constructive feedback.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.teler.2023.100067.
==== Refs
References
1 Nussbaumer-Streit B. Quarantine alone or in combination with other public health measures to control COVID-19: a rapid review Cochrane Database Systematic Rev 9 2020
2 Lowenthal P. Thinking beyond Zoom: using asynchronous video to maintain connection and engagement during the COVID-19 pandemic J. Technol. Teacher Educ 28 2 2020 383 391
3 Bloom N. Davis S.J. Zhestkova Y. Covid-19 shifted patent applications toward technologies that support working from home AEA Papers and Proceedings 111 2021 263 266
4 Tudor, C.J.E., The impact of the COVID-19 pandemic on the global web and video conferencing SaaS market. 2022. 11(16): p. 2633.
5 Parker, S. Are you a zoom Zombie. in How to Take Control of Your Virtual Meetings. Available online: https://www.transformativeworkdesign.com/post/are-you-a-zoom-zombie-how-to-take-control-of-your-virtual-meetings (accessed on 3 February 2022). 2020.
6 Wiederhold, B.K., Connecting Through Technology During the Coronavirus Disease 2019 pandemic: Avoiding “Zoom Fatigue”. 2020, Mary Ann Liebert, Inc., publishers 140 Huguenot Street, 3rd Floor New …. p. 437–438.
7 Degges-White S. Zoom Fatigue: don't let video meetings zap your energy Psychol. Today 2020 Available online https://www.psychologytoday.com/us/blog/lifetime-connections/202004/zoom-fatigue-dont-let-video-meetings-zap-your-energy
8 Lebni J.Y. A study of internet addiction and its effects on mental health: a study based on Iranian University Students J. Educ. Health Promot. 9 2020
9 Sarres, C.P., et al., The effects of telework satisfaction and Zoom fatigue on mental health: a pilot study. 2021.
10 Oducado, R.M.F., et al., Zoom fatigue and mental well-being among school executives. 2022. 27(4): p. 393–394.
11 Li B.J. Yee A.Z.H. Understanding videoconference fatigue: a systematic review of dimensions, antecedents and theories Internet Res. 2022 ahead-of-print(ahead-of-print)
12 Döring N. Videoconference fatigue: a conceptual analysis Int. J. Environ. Res. Public Health 19 4 2022 2061 35206250
13 Fauville G. Nonverbal Mechanisms Predict Zoom Fatigue and Explain Why Women Experience Higher Levels than Men Available at SSRN 3820035 2021
14 Anderson K. Looi J.C. Chronic Zoom Syndrome: emergence of an insidious and debilitating mental health disorder during COVID-19 Austr. Psychiatry 28 6 2020 669
15 Lee J.J.P.T. A neuropsychological exploration of Zoom fatigue Psychiatric Times 37 11 2020 38 39
16 Riedl R. On the stress potential of videoconferencing: definition and root causes of Zoom fatigue Electron. Mark 32 1 2022 153 177 35600914
17 Bailenson J.N. Nonverbal overload: a theoretical argument for the causes of Zoom fatigue Technol. Mind. Behav. 2 1 2021
18 Oducado R.M.F. Dequilla M. Villaruz J.F. Factors predicting videoconferencing fatigue among higher education faculty Educ. Inf. Technol. (Dordr.) 27 7 2022 9713 9724 35399786
19 Williams N. Working through COVID-19:‘Zoom'gloom and ‘Zoom'fatigue Occup Med (Chic Ill) 71 3 2021 164-164
20 Montag, C., et al., On the associations between videoconference fatigue, burnout and depression including personality associations. J. Affect. Disord. Rep., 100409.
21 Baron-Cohen S. The autism-spectrum quotient (AQ): evidence from asperger syndrome/high-functioning autism, malesand females, scientists and mathematicians J Autism. Dev. Disord. 31 1 2001 5 17 11439754
22 Lundström S. Autism spectrum disorders and autisticlike traits: similar etiology in the extreme end and the normal variation Arch. Gen. Psychiatry 69 1 2012 46 52 22213788
23 Ronald A. Phenotypic and Genetic Overlap Between Autistic Traits at the Extremes of the General Population J. Am. Acad. Child Adolesc. Psychiatry 45 10 2006 1206 1214 17003666
24 Edition F. Diagnostic and statistical manual of mental disorders Am. Psychiatric Assoc. 21 21 2013 591 643
25 Volkmar F.R. Autism and pervasive developmental disorders J. Child Psychol. Psychiatry 45 1 2004 135 170 14959806
26 Jobe L.E. White S.W. Loneliness, social relationships, and a broader autism phenotype in college students Pers. Individ. Dif. 42 8 2007 1479 1489
27 Sasson N.J. Nowlin R.B. Pinkham A.E. Social cognition, social skill, and the broad autism phenotype Autism 17 6 2013 655 667 22987889
28 Gökçen E. Sub-threshold autism traits: the role of trait emotional intelligence and cognitive flexibility Brit. J. Psychol. 105 2 2014 187 199 24754807
29 Lodi-Smith J. Meta-analysis of Big Five personality traits in autism spectrum disorder Autism 23 3 2019 556 565 29676605
30 Lundström S. Autistic-like traits and their association with mental health problems in two nationwide twin cohorts of children and adults Psychol. Med. 41 11 2011 2423 2433 21426604
31 Matsuo J. Autistic-like traits in adult patients with mood disorders and schizophrenia PLoS ONE 10 4 2015 e0122711
32 Martinez A.P. Robust association between autistic traits and psychotic-like experiences in the adult general population: epidemiological study from the 2007 Adult Psychiatric Morbidity Survey and replication with the 2014 APMS Psychol. Med. 51 15 2021 2707 2713 32441234
33 Zhao X. Autistic traits and negative emotions in the general population during COVID-19: mediating roles of the behavioural immune system and COVID-19 risk perception Psychiatry Res. 300 2021 113918
34 Zhao X. Autistic traits and gender modulate emotion changes before and during the COVID-19 pandemic Curr. Psychol. 2021 1 11
35 Clark D.M. Wells A. A cognitive model. Social phobia: diagnosis Assess. Treat. 69 1995 1025
36 Yeung R.C. Fernandes M.A. Social anxiety enhances recognition of task-irrelevant threat words Acta Psychol. (Amst.) 194 2019 69 76 30779989
37 Asmundson G.J. Taylor S. Coronaphobia: fear and the 2019-nCoV outbreak J. Anxiety Disord. 70 2020 102196
38 Panda P.K. Psychological and behavioral impact of lockdown and quarantine measures for COVID-19 pandemic on children, adolescents and caregivers: a systematic review and Meta-analysis J. Trop. Pediatr. 67 1 2021
39 Oomen D. Nijhof A.D. Wiersema J.R. The psychological impact of the COVID-19 pandemic on adults with autism: a survey study across three countries Mol. Autism. 12 1 2021 1 21 33436060
40 Zhao X. Autistic traits and prosocial behaviour in the general population: test of the mediating effects of trait empathy and state empathic concern J. Autism. Dev. Disord. 49 10 2019 3925 3938 30203310
41 Maslach C. Maslach Burnout Inventory: Manual Fourth Edition 2018 Mind Garden Inc Menlo Park, CA
42 Çalışkan E. Kargın M. The relationship between coronavirus 19 fear and occupational fatigue, burnout and recovery levels in healthcare professionals in Turkey Arch. Environ. Occup. Health 2021 1 10 32048551
43 Afulani, P.A., et al., Inadequate preparedness for response to COVID-19 is associated with stress and burnout among healthcare workers in Ghana. 2021. 16(4): p. e0250294.
44 Yıldırım, M. and F.J.D.S. Solmaz, COVID-19 burnout, COVID-19 stress and resilience: initial psychometric properties of COVID-19 Burnout Scale. 2022. 46(3): p. 524–532.
45 Kardefelt-Winther D. A conceptual and methodological critique of internet addiction research: towards a model of compensatory internet use Comput. Hum. Behav. 31 2014 351 354
46 Shadzi M.R. Salehi A. Vardanjani H.M. Problematic internet use, mental health, and sleep quality among medical students: a path-analytic model Indian J. Psychol. Med. 42 2 2020 128 135 32346253
47 Priego-Parra B.A. Anxiety, depression, attitudes, and internet addiction during the initial phase of the 2019 coronavirus disease (COVID-19) epidemic: a cross-sectional study in México medRxiv 2020 2020.05.10.20095844
48 Lee J.J. Shin S.H. Associations between fear of COVID-19, depression, and internet addiction in south korean adults Healthcare (Basel) 10 5 2022
49 Servidio R. Fear of COVID-19, depression, anxiety, and their association with Internet addiction disorder in a sample of Italian students J. Affect. Disord. Rep. 4 2021 100097
50 Zhang Y. Individual differences in tendencies toward internet use disorder, internet literacy and their link to autistic traits in both China and Germany Front. Psychiatry 2021 12
51 Finkenauer C. Brief report: examining the link between autistic traits and compulsive Internet use in a non-clinical sample J. Autism. Dev. Disord. 42 10 2012 2252 2256 22350338
52 Zhang Y. Yao S. Schmitt H. Molecular genetic associations between a prominent serotonin transporter gene polymorphism (5-HTTLPR/rs25531) and individual differences in tendencies toward autistic traits and generalized internet use disorder in China and Germany Brain. Behav. 12 10 2022 e2747 36106519
53 Bener A. Internet addiction, fatigue, and sleep problems among adolescent students: a large-scale study Int. J. Ment. Health Addict. 17 2019 959 969
54 Bener A.J.E.P. Lifestyle factors and internet addiction among school children Int. J. Community Fam. Med. 1 118 2016 2
55 Gil F. Chamarro A. Oberst U. PO-14: addiction to online social networks: a question of" Fear of Missing Out"? J. Behav. Addict. 4 S1 2015 51 52
56 Elhai J.D. Yang H. Montag C. Fear of missing out (FOMO): overview, theoretical underpinnings, and literature review on relations with severity of negative affectivity and problematic technology use Brazil. J. Psychiatry 43 2020 203 209
57 Przybylski A.K. Motivational, emotional, and behavioral correlates of fear of missing out Comput. Hum. Behav. 29 4 2013 1841 1848
58 Hayran C. Anik L. Well-being and fear of missing out (FOMO) on digital content in the time of COVID-19: a correlational analysis among university students Int. J. Environ. Res. Public Health 18 4 2021 1974 33670639
59 Whelan E. Islam A.N. Brooks S. Applying the SOBC paradigm to explain how social media overload affects academic performance Comput. Educ. 143 2020 103692
60 Fauville G. Zoom exhaustion & fatigue scale Comput. Hum. Behav. Rep. 4 2021 100119
61 Freitag C.M. Evaluation der deutschen version des Autismus-Spektrum-Quotienten (AQ)-die Kurzversion AQ-k Zeitschrift für klinische Psychologie und Psychotherapie 36 4 2007 280 289
62 Connor K.M. Psychometric properties of the Social Phobia Inventory (SPIN): new self-rating scale Brit. J. Psychiatry 176 4 2000 379 386 10827888
63 Sanwald S. Primary emotions as predictors for fear of COVID-19 in former inpatients with Major Depressive Disorder and healthy control participants BMC Psychiatry 22 1 2022 1 15 34983462
64 Stodt B. Investigating the effect of personality, internet literacy, and use expectancies in internet-use disorder: a comparative study between China and Germany Int. J. Environ. Res. Public Health 15 4 2018 579 29570663
65 Wegmann E. Online-specific fear of missing out and Internet-use expectancies contribute to symptoms of Internet-communication disorder Addict. Behav. Rep. 5 2017 33 42 29450225
66 Kim H.-Y. Statistical notes for clinical researchers: assessing normal distribution (2) using skewness and kurtosis Restor. Dent. Endod. 38 1 2013 52 54 23495371
67 Hayes, A.F., Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. 2017: Guilford publications.
68 Shrout P.E. Bolger N. Mediation in experimental and nonexperimental studies: new procedures and recommendations Psychol. Methods 7 4 2002 422 12530702
69 Cage E. McManemy E. Burnt out and dropping out: a comparison of the experiences of autistic and non-autistic students during the COVID-19 pandemic Front. Psychol. 12 2021 792945
70 Caminha R.C. Lampreia C. Findings on sensory deficits in autism: implications for understanding the disorder Psychol. Neurosci. 5 2012 231 237
71 Demetriou E.A. Autism spectrum disorders: a meta-analysis of executive function Mol. Psychiatry 23 5 2018 1198 1204 28439105
72 de Nocker Y.L. Toolan C.K. Using telehealth to provide interventions for children with ASD: a systematic review Rev. J. Autism Dev. Disord. 2021 1 31
73 Bundy R. The impact of early stages of COVID-19 on the mental health of autistic adults in the United Kingdom: a longitudinal mixed-methods study Autism 2022 13623613211065543
74 Cassidy S.A. An expert discussion on autism in the COVID-19 pandemic Autism Adulthood 2 2 2020 106 117 36601566
75 Mottern, E. and L. Berger, Victory for all: overcoming the difficulties of remote learning in children with autism. 2022.
76 Higgins, J.M., et al., Defining autistic burnout through experts by lived experience: grounded delphi method investigating# AutBurnout. 2021.
77 Mantzalas, J., A.L. Richdale, and C.J.A.R. Dissanayake, A conceptual model of risk and protective factors for autistic burnout. 2022. 15(6): p. 976–987.
78 Mantzalas J. What is autistic burnout? A thematic analysis of posts on two online platforms Autism Adulthood 4 1 2022 52 65 36605565
79 Raymaker D.M. Having all of your internal resources exhausted beyond measure and being left with no clean-up crew”: defining autistic burnout Autism. Adulthood 2 2 2020 132 143 32851204
80 Robinson E.B. Stability of autistic traits in the general population: further evidence for a continuum of impairment J. Am. Acad. Child Adolescent Psychiatry 50 4 2011 376 384
81 Whitehouse A.J. Hickey M. Ronald A.J.P.O. Are autistic traits in the general population stable across development? PLoS ONE 6 8 2011 e23029 21829684
82 Montag C. Investigating links between fear of COVID-19, neuroticism, social networks use disorder, and smartphone use disorder tendencies Front. Psychol. 2021 12
83 Kayis A.R. Fear of COVID-19, loneliness, smartphone addiction, and mental wellbeing among the Turkish general population: a serial mediation model Behav. Inf. Technol. 2021 1 13
84 Balta S. Neuroticism, trait fear of missing out, and phubbing: the mediating role of state fear of missing out and problematic Instagram use Int. J. Ment. Health Addict. 18 3 2020 628 639
85 Elbogen E.B. A National study of zoom fatigue and mental health during the COVID-19 pandemic: implications for future remote work Cyberpsychol. Behav. Soc. Netw. 25 7 2022 409 415 35648039
|
PMC010xxxxxx/PMC10204006.txt |
==== Front
Nat Rev Endocrinol
Nat Rev Endocrinol
Nature Reviews. Endocrinology
1759-5029
1759-5037
Nature Publishing Group UK London
37221400
851
10.1038/s41574-023-00851-2
News & Views
How can social jetlag affect health?
http://orcid.org/0000-0003-2939-0332
Roenneberg Till Roenneberg@LMU.de
12
1 grid.5252.0 0000 0004 1936 973X Institute for Occupational, Social and Environmental Medicine, LMU Munich, Munich, Germany
2 grid.5252.0 0000 0004 1936 973X Institute for Medical Psychology, LMU Munich, Munich, Germany
23 5 2023
2023
19 7 383384
© Springer Nature Limited 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Our lives are governed by three clocks: the social clock that organizes our lives with others (local time), the biological clock that controls our physiology (circadian time) and the sun clock that defines natural light and darkness. The more misaligned these clocks are, the higher our odds of developing certain diseases. ‘Social jetlag’ quantifies the difference between local and circadian time.
Subject terms
Epidemiology
Type 2 diabetes
issue-copyright-statement© Springer Nature Limited 2023
==== Body
pmcRefers to Bouman, E. J., et al. Social jet lag and (changes in) glycemic and metabolic control in people with type 2 diabetes. Obesity 31, 945–954 (2023).
The term ‘social jetlag’ was inspired by a study showing that people who slept at different times during the working week than the weekends had increased tobacco, caffeine and alcohol consumption compared with people who had similar sleeping patterns throughout the week1. Social jetlag quantifies the difference in mid-sleep times on nights before work (or school) days and those before work-free days. It assumes that people live more according to the social clock (local time) during the working week and more according to their biological clock (circadian clock) on work-free days.
The circadian clock has a genetic basis, which contributes to inter-individual differences; that is, different ‘chronotypes’, from early ‘larks’ to late ‘owls’. Owls (who sleep and wake later on work-free days than on workdays) would be expected to have a greater degree of social jetlag than larks (who sleep and wake at approximately the same times on both day types). Being a late chronotype is a strong predictor for having more social jetlag, and as a person’s chronotype changes with age2, so does the amount of social jetlag they experience (Fig. 1). Coining the term ‘social jetlag’ apparently gave a name to a state that many people had experienced: within a month after the paper1 appeared, the term went from zero to 100,000 counts in search engines.Fig. 1 Average social jetlag across a lifetime.
Social jetlag, which can be calculated using the Munich ChronoType Questionnaire (MCTQ)3, is defined as the difference between the mid-point of sleep on workdays and on work-free days. Data are taken from the MCTQ database (n = 198,370) and were first published in Roenneberg et al. (2012)8. Except for the age groups of 10 and 11 years, standard errors of the mean are smaller than the symbols. The blue area indicates where social jetlag is smaller than 20 min, which appears to be the tolerable limit for people6. The average age ± standard deviation of the study population in Bouman et al.4 is indicated by the dashed grey box.
The phenomenon social jetlag is as young as industrialisation, which has drastically changed our exposure to light and darkness, the predominant synchronising signals for circadian clocks. By working inside, we decrease daytime light exposure up to a thousand-fold, and by using artificial light, we eliminate dark exposure (except for when we sleep). Under these conditions, most circadian clocks, except for those of people with extremely early chronotypes, had to become later to ensure stable synchronisation to the 24-h Earth rotation. Because the demands of the social clock (for example, work and/or school times) are early relative to the ‘industrialised’ circadian clocks, over 80% of the population (represented in the MCTQ (Munich ChronotType Questionnaire) database; see legend to Fig. 1) use an alarm clock and accumulate a sleep debt over the course of the working week. Weekend sleep is characterized by both sleeping at more appropriate times for the individual biological clock and by catching up on lost sleep.
Many epidemiological studies have shown strong associations between social jetlag and the prevalence of diverse pathologies, ranging from depression and cardiovascular risks to metabolic dysfunction (reviewed in ref. 3). The most recent epidemiological study, by Bouman and colleagues4, reports apparently contradicting results that offer explanations on how social jetlag and health might be connected. The authors assessed social jetlag and metabolic and glycaemic outcomes in approximately 1,000 people over the age of 60 with type 2 diabetes mellitus (T2DM), using both a cross-sectional and longitudinal design (re-testing after 2 years). A covariate of the study was whether participants were currently employed or retired. While the cross-sectional results in employed people corroborate previous associations between increased social jetlag and increased metabolic dysfunction5, the relationship seemed to be the reverse in retired people. The longitudinal assessment results do not reach statistical significance. The authors postulate that this finding could be because the study population included people with well-controlled T2DM. However, these apparent contradictions might be clarified when put into a larger context of social jetlag.
“While enforced social jetlag disrupts health, voluntary sleep extension on weekends might protect it”
The study population in Bouman et al.4 (shown in the dashed grey box in Fig. 1) has lower social jetlag than younger people (that is, those <60 years old). This known age trajectory is even present within the age range of this population: participants with the greatest social jetlag are the youngest (62 years) and those with the lowest are the oldest (72 years). Above the age of 60, social jetlag drops sharply, most probably due to retirement, and quickly approaches 20 min: when people can reduce social jetlag, as occurred during the COVID-19 lockdowns, they do so only if the pre-lockdown social jetlag was >20 min6 (shown in the blue area in Fig. 1). Individuals who are late chronotypes at the age of 60 years most probably were even later chronotypes for most of their lives. The health effect of social jetlag probably has both acute and chronic components. The immediate consequences of interrupted and insufficient sleep would contribute to the acute effects, while the strain on metabolism by being active and eating at the wrong biological times would accumulate as the chronic effects of social jetlag. The acute social jetlag status (the effects that people report at the time of the study) would then be a marker for the social jetlag load they accumulated over decades, which in turn is a predictor for pathology, especially in older populations. This distinction would also explain why cross-sectional associations between social jetlag and metabolic dysfunctions are statistically significant while the prospective associations after 1 and 2 years are not: at ages >60 years, most of the social jetlag load has already occurred.
The question remains, why does social jetlag in this >60 years age group have a detrimental association with health in those who still work but become protective after retirement. Social jetlag is defined as a difference in sleep timing between workdays and work-free days. It can persist after retirement for three reasons: first, people continue to get up with alarm clocks during the working week but not on weekends; second, people sleep in during the working week and get up earlier on weekends (producing a negative social jetlag); or third, people don’t use alarm clocks throughout the week but allow themselves extra sleep on weekends (beyond their usual biological waking time). While Bouman et al. address the first two possibilities in suggesting that social jetlag in retirement “could be an indication of maintaining an active social life”, the third possibility simply points to the protective role of sleep. A study has shown that we usually underestimate our sleep need7: when people were instructed to stay in bed in the dark for many hours a day, all participants slept longer than their habitual sleep duration, even if they claimed to get enough sleep in their everyday life. Thus, social jetlag occurs both when we are forced to live against our body clock and when we allow ourselves to sleep in on weekends. While enforced social jetlag disrupts health, voluntary sleep extension on weekends might protect it. We know that longer sleep is protective against metabolic dysfunction even on free days8 and that life-expectancy is reduced by not sleeping in on weekends9.
“A reduction of enforced social jetlag should therefore be central to strategies to prevent disease”
The many epidemiological studies investigating social jetlag suggest that the higher its accumulation, the higher the prevalence and the earlier the onset of clinical symptoms for many different pathologies beyond metabolic dysfunction. This effect is similar to the accumulating effects of sleep loss on health10. A reduction of enforced social jetlag should therefore be central to strategies to prevent disease. Candidates for such a prevention could be more flexible work schedules, later school start times for adolescents or eliminating daylight saving time.
Competing interests
The author declares no competing interests.
==== Refs
References
1. Wittmann M Dinich J Merrow M Roenneberg T Social jetlag: misalignment of biological and social time Chronobiol. Int. 2006 23 497 509 10.1080/07420520500545979 16687322
2. Roenneberg T A marker for the end of adolescence Curr. Biol. 2004 14 R1038 R1039 10.1016/j.cub.2004.11.039 15620633
3. Roenneberg T Pilz LK Zerbini G Winnebeck EC Chronotype and social jetlag: a (self-) critical review Biology 2019 8 54 10.3390/biology8030054 31336976
4. Bouman EJ Social jet lag and (changes in) glycemic and metabolic control in people with type 2 diabetes Obesity 2023 31 945 954 10.1002/oby.23730 36855048
5. Parsons MJ Social jetlag, obesity and metabolic disorder: investigation in a cohort study Int. J. Obes. 2015 39 842 848 10.1038/ijo.2014.201
6. Korman M COVID-19-mandated social restrictions unveil the impact of social time pressure on sleep and body clock Sci. Rep. 2020 10 22225 10.1038/s41598-020-79299-7 33335241
7. Klerman EB Dijk D-J Age-related reduction in the maximal capacity for sleep–implications for insomnia Curr. Biol. 2008 18 1118 1123 10.1016/j.cub.2008.06.047 18656358
8. Roenneberg T Allebrandt KV Merrow M Vetter C Social jetlag and obesity Curr. Biol. 2012 22 939 943 10.1016/j.cub.2012.03.038 22578422
9. Åkerstedt T Sleep duration and mortality-does weekend sleep matter? J. Sleep Res. 2019 28 e12712 10.1111/jsr.12712 29790200
10. Ferini-Strambi L Sleep disorders and increased risk of dementia Eur. J. Neurol. 2022 29 3484 3485 10.1111/ene.15562 36094745
|
PMC010xxxxxx/PMC10204340.txt |
==== Front
Biosens Bioelectron
Biosens Bioelectron
Biosensors & Bioelectronics
0956-5663
1873-4235
Elsevier B.V.
S0956-5663(23)00363-9
10.1016/j.bios.2023.115421
115421
Article
Localized surface plasmon resonance biosensor chip surface modification and signal amplifications toward rapid and sensitive detection of COVID-19 infections
Hao Xingkai a
St-Pierre Jean-Philippe a
Zou Shan b
Cao Xudong a∗
a Department of Chemical and Biological Engineering, University of Ottawa, 161 Louis Pasteur, Ottawa, Ontario, K1N 6N5, Canada
b Metrology Research Centre, National Research Council Canada, 100 Sussex Drive, Ottawa, Ontario, K1A 0R6, Canada
∗ Corresponding author.
23 5 2023
15 9 2023
23 5 2023
236 115421115421
16 2 2023
1 5 2023
22 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
We developed a multi-pronged approach to enhance the detection sensitivity of localized surface plasmon resonance (LSPR) sensor chips to detect SARS-CoV-2. To this end, poly(amidoamine) dendrimers were immobilized onto the surface of LSPR sensor chips to serve as templates to further conjugate aptamers specific for SARS-CoV-2. The immobilized dendrimers were shown to reduce surface nonspecific adsorptions and increase capturing ligand density on the sensor chips, thereby improving detection sensitivity. To characterize the detection sensitivity of the surface-modified sensor chips, SARS-CoV-2 spike protein receptor-binding domain was detected using LSPR sensor chips with different surface modifications. The results showed that the dendrimer-aptamer modified LSPR sensor chip exhibited a limit of detection (LOD) of 21.9 pM, a sensitivity that was 9 times and 152 times more sensitive than the traditional aptamer- or antibody-based LSPR sensor chips, respectively. In addition, detection sensitivity was further improved by combining rolling circle amplification product and gold nanoparticles to further amplify the detection signals by increasing both the target mass and plasmonic coupling effects. Using pseudo SARS-CoV-2 viral particles as detection targets, we demonstrated that this combined signal intensification approach further enhanced the detection sensitivity by 10 folds with a remarkable LOD of 148 vp/mL, making it one of the most sensitive SARS-CoV-2 detection assays reported to date. These results highlight the potential of a novel LSPR-based detection platform for sensitive and rapid detection of COVID-19 infections, as well as other viral infections and point-of-care applications.
Keywords
Aptamers
Dendrimers
SARS-CoV-2 detection
Rolling circle amplification
Nonfouling
Surface modifications
==== Body
pmc1 Introduction
The current COVID-19 pandemic has caused millions of deaths and affected economies around the world, highlighting the need for improved technological preparedness for the next pandemic. During the COVID-19 pandemic, many biosensors (Qiu et al., 2020; Raziq et al., 2021) have been developed for the rapid detection of COVID-19 infections; however, these rapid detection biosensors suffer from various drawbacks, such as low detection sensitivities or complicated sample pre-treatments (Cennamo et al., 2021a; Tian et al., 2020), severely limiting their application as reliable alternatives to polymerase chain reaction (PCR) tests.
Localized surface plasmon resonance (LSPR) is an optical phenomenon generated by collective electron charge oscillations in metallic nanoparticles excited by a light source (Schasfoort, 2017). These charge oscillations create a highly localized evanescent field extending from the surface of the nanoparticles, which is highly sensitive to minute local variations, such as surface mass changes due to molecular binding (Schasfoort, 2017). As such, LSPR-based biosensors have been widely used in many real-time and label-free detection applications (Cennamo et al., 2021b; Lewis et al., 2021). In fact, several LSPR biosensors have already been developed to detect COVID-19-related biological targets, including nucleic acids, antibodies, proteins, and whole viral particles (Huang et al., 2021; Lewis et al., 2021; Qiu et al., 2020). While these LSPR-based methods have demonstrated good detection performances, there is still a need to further improve the sensitivity of these LSPR biosensors in order to approach that achieved by PCR-based methods (Samson et al., 2020).
To improve the detection sensitivity of LSPR biosensors, two critically important considerations – non-fouling properties of the sensor chip surfaces and high surface-ligand densities on the sensor chip surfaces – must be simultaneously targeted and integrated into the sensor chip designs (Hao et al. 2023a, 2023b). For example, it is crucial for the sensor chip surfaces to minimize nonspecific adsorptions, thereby decreasing the background noises and ultimately enhancing the signal-to-noise ratio and thus detection sensitivity (Hao et al. 2023a, 2023b). Furthermore, sensor chip surfaces with increased detection ligand density could lead to a higher probability of target capture and stronger binding avidity (Wang et al., 2022), thus resulting in higher detection sensitivity.
Aptamers are a special type of single-stranded DNA or RNA oligonucleotides -- selected through a cyclic process known as systematic evolution of ligands by exponential enrichment (SELEX) from DNA/RNA libraries of random sequences -- that can recognize specific targets, such as proteins, toxins, and cells (Tuerk and Gold, 1990). Recently, Song et al. selected an aptamer specifically targeting the SARS-CoV-2 spike protein receptor-binding domain (SRBD) with a high binding affinity (Kd = 5.8 nM) and specificity (Song et al., 2020). This aptamer has been widely used in different biosensors for SARS-CoV-2 detections (Amouzadeh Tabrizi and Acedo, 2022; Jiang et al., 2022; Xue et al., 2022), and it is also used in the current study.
Poly(amidoamine) (PAMAM) dendrimers are hyper-branched polymers that can be synthesized to feature a range of different surface functional groups, including -OH, -COOH, and –NH2. Because of their unique structure, PAMAM dendrimers have been used as multi-handled templates via which capturing ligands, such as aptamers and antibodies, can be conjugated onto capturing surfaces to allow for enhanced target capturing probability and binding avidity for improved detection performance (Wang et al., 2022). In addition, immobilization of PAMAM dendrimers onto the detection surfaces can render the resulting surface nonfouling to minimize background noises (Hao et al. 2023a, 2023b).
Rolling circle amplification (RCA) reactions can generate long single-stranded DNA molecules with hundreds or thousands of tandem repeating units from a circular template, the nucleotide sequence of which can be designed to either capture specific targets, amplify signals, or do both (He et al., 2014). With these repeating units, the RCA products can be used to hybridize with a large amount of ssDNA functionalized gold nanoparticles (i.e., ssDNA-AuNPs) to form RCA-gold nanoparticle complexes (RCA-AuNPs) that are known to intensify LSPR detection signals due to both significantly improved target mass and plasmonic coupling effect (He et al., 2014; Kim et al., 2022; Xiang et al., 2013). For example, the RCA-AuNPs have been applied to conjugate with surface-captured targets on a surface plasmon resonance (SPR) sensor chip to improve detection signals (He et al., 2014).
In this study, we report our most recent efforts to enhance the detection sensitivity of regular LSPR sensor chips – prepared using a widely employed gold nanoislands/poly (allylamine hydrochloride) (PAH)/poly(sodium 4-styrenesulfonate) (PSS) method (Lu et al., 2021) – by implementing a multi-pronged design that aims to both reduce background noises and enhance detection signals. To reduce the background noises, both the exposed glass surfaces and gold nanoislands of the sensor chip are modified with PAMAM dendrimers to render them nonfouling and prevent non-target molecules and particles from adsorbing to the sensor chip surfaces. To enhance the detection signals, the following strategies are explored: a) an increased density of capturing aptamers specific for the targets are immobilized onto the sensor chip detection surfaces via the multi-handled PAMAM dendrimers, resulting in higher capturing probability and binding avidity for target recognition and capture (i.e., first-layer amplification); and b) RCA-AuNPs complexes that incorporate both a large number of AuNPs and tandem repeating aptamers specific for the targets are employed to bind to surface captured targets (i.e., SARS-CoV-2 pseudo viral particles) in a sandwich detection format to improve detection signals (i.e., second-layer amplification) by both (1) increasing the target mass (from both RCA products and AuNPs) and (2) improving plasmonic coupling effect (from the AuNPs) (He et al., 2014; Kim et al., 2022). The performances of these PAMAM dendrimer-aptamer modified sensor chips are evaluated using both the SARS-CoV-2 SRBD and SARS-CoV-2 pseudo viral particles. Our results show that the modified sensor chip is 152 times more sensitive than the antibody-immobilized sensor chip (traditional method) in detecting SRBD samples (i.e., due to first-layer amplification). Application of RCA-AuNPs signal amplification results in an additional 10-time improvement in the detection signal intensity when detecting SARS-CoV-2 pseudo viral particles (i.e., due to second-layer amplification). With this two-layered signal amplification approach, we report a LOD of 148 viral particles per milliliter (vp/mL), one of the highest sensitivities reported in whole viral particle detections amongst all detection platforms (Huang et al., 2021; Nguyen et al., 2016; Park et al., 2014; Yang et al., 2022), making it sufficiently sensitive for the diagnosis of early COVID-19 infections (Huang et al., 2021).
2 Material and methods
2.1 Overall design
As illustrated in Fig. 1 , LSPR sensor chip surfaces are modified sequentially with generation 3.5 carboxylated PAMAM dendrimers (i.e., G3.5-COOH) and generation 4 aminated PAMAM dendrimers (i.e., G4-NH2) to render the resulting surfaces nonfouling. The immobilized G4 dendrimers act as multi-handled templates to allow for subsequent conjugation of numerous copies of aptamers specific for binding of SARS-CoV-2 SRBD (Song et al., 2020) (first-layer amplification). The resulting sensor chips are used to detect COVID-19 targets (i.e., either SARS-CoV-2 SRBD or SARS-CoV-2 pseudo viral particles presenting spike proteins on the viral particle surfaces). For the case of SARS-CoV-2 pseudo viral particle detection, RCA-AuNPs are also employed to bind with the surface-captured viral particles and to further amplify the detection signal by increasing the target mass and plasmonic coupling effect (second-layer amplification).Fig. 1 Schematic illustration showing the general approach to prepare (G3.5+G4)-aptamer modified LSPR sensor chips for sensitive detection of the SARS-CoV-2 SRBD and SARS-CoV-2 pseudo viral particles. Note that the second-layer amplification is only possible with SARS-CoV-2 pseudo viral particles as the detection targets upon which the detection sandwich format can be built.
Fig. 1
2.2 Materials
Detailed information on chemicals and pseudo-SARS-CoV-2 viral particles for this study are provided in S1.1 in Supporting Information (SI); gold nanoislands (diameter = 100 nm)/PAH/PSS LSPR sensor chips with carboxyl functionalization on gold nanoislands were prepared as detailed in S1.2 (SI). Details of DNA sequences used for this study are listed below:
Capturing aptamer (i.e., aptamer specific for binding of SARS-CoV-2 SRBD) (Song et al., 2020):
NH2, 12C spacer-CAG CAC CGA CCT TGT GCT TTG GGA GTG CTG GTC CAA GGG CGT TAA TGG ACA.
RCA primer:
GGA CAT TTT TTT TTT TTT TTT CAG CA.
RCA padlock probe:
Phos-AAA AAA AAT GTC CAT TAA CGC CCT TGG ACC AGC ACT CCC AAA GCA CAA GGT CGG TGC TGA AAA AAA A-3
ssDNA probe:
Thio, 6C spacer-AAA AAA AAA AAA AAA A.
Note: The underlined portion in the RCA padlock probe sequence is complementary to the sequence of capturing aptamer in order to produce tandem repeating aptamer by the RCA reaction. The ssDNA probe is complementary to the sequence of the RCA product using a circular template, a conjugation product of RCA primer and padlock probe.
2.3 LSPR measurements
LSPR response graphs were measured using a Nicoya OpenSPR instrument (Kitchener, ON). The response graphs provide real-time tracking of the shift of peak absorbance wavelength of light absorbed by the gold nanoislands on the LSPR sensor chip as an LSPR signal in response units (RU) units (y-axis) vs. time (x-axis). Specifically, 1 RU equals to a 1 p.m. shift of the peak absorbance wavelength.
2.4 Surface modifications using PAMAM dendrimers and aptamers
Surface modification of the LSPR sensor chips followed the general steps outlined in Fig. 1. Since the exposed glass areas of the original LSPR sensor chip presented -NH2 functional groups while the gold nanoislands presented -COOH groups (S2.1, SI), the sensor chip surface was first reacted with G3.5-COOH using N-hydroxysuccinimide/N-ethyl-N′-(3-(dimethylamino)propyl) carbodiimide (NHS/EDC) chemistry to immobilize G3.5-COOH on the glass areas and to convert the carboxyl surface functional groups on both the glass areas and the gold nanoislands to NHS esters (Hermanson, 2013). Specifically, the LSPR sensor chips were loaded into the Nicoya OpenSPR instrument to form a microfluidic channel based LSPR biosensor, after which a G3.5-COOH immobilization solution, including 50 μM G3.5-COOH, 0.1 M NHS, and 0.1 M EDC in PBS (pH 7.4), was injected into the channel to react with the LSPR sensor chip surfaces at a flow rate of 10 μL/min for 10 min. Subsequently, the resulting sensor chip was carefully rinsed with PBS (pH 7.4) at a flow rate of 10 μL/min until the signal baseline was stabilized.
Subsequently, G4-NH2 was immediately immobilized onto the surfaces of the LSPR sensor chip via the NHS-esters obtained during the previous step. Briefly, a 100 μM G4-NH2 in PBS (pH 7.4) was injected into the channel to react with the G3.5 modified LSPR sensor chip at a flow rate of 10 μL/min for 10 min, after which the G4 immobilized chip was rinsed with PBS (pH 7.4) at a flow rate of 10 μL/min until the signal baseline was stabilized. In this modification step, the G4-NH2 molecules were immobilized onto both the gold nanoislands and the immobilized G3.5 molecules. It is important that G4-NH2 immobilization should be conducted immediately after the G3.5-COOH modification, as NHS-esters are known to be readily hydrolyzed (Hermanson, 2013). The immobilized G4-NH2 molecules were used as multi-handled templates to provide multiple binding sites (i.e., -NH2 groups) in the next step to further conjugate aptamers.
Finally, to conjugate amino-capped aptamers (i.e., NH2-aptamer, specific for SRBD) onto the sensor chip surfaces via the immobilized G4-NH2 templates, a widely used bis(sulfosuccinimidyl)suberate (BS3) crosslinker was employed (Staros, 1982). As the BS3 crosslinker contains homobifunctional sulfo-NHS-esters that can efficiently conjugate molecules containing primary amino groups (Staros, 1982), the aptamer-NH2 can be readily conjugated with the surface immobilized G4-NH2 molecules acting as templates. In a typical reaction, an aptamer conjugation solution containing 200 μM aptamers and 20 mM BS3 in PBS (pH 7.4) was injected into the (G3.5+G4) modified LSPR sensor chip at a flow rate of 5 μL/min for 20 min to conjugate aptamers onto the G3.5+G4 templated surface. Subsequently, an ethanolamine-HCL solution (1 M, pH 8.5) was injected into the modified LSPR sensor chip at a flow rate of 20 μL/min for 5 min to deactivate the reaction, after which the sensor chip was carefully rinsed with PBS (pH 7.4) at a flow rate of 30 μL/min until the signal baseline was stabilized.
2.5 Preparation of RCA-AuNPs
To prepare RCA-AuNPs, ssDNA-AuNPs and RCA products were first individually synthesized. Specifically, to obtain ssDNA-AuNPs, a well-established protocol that involved the reduction of HAuCl4 by trisodium citrate was followed (Frens, 1973); subsequently, the obtained AuNPs were used to conjugate ssDNA probes as demonstrated in a previous study (Hurst et al., 2006). To synthesize the RCA product, a previously published RCA reaction protocol was followed with minor modifications (Jiang et al., 2019; Li et al., 2021). Specifically, 100 μL of 1 μM circular template was first prepared using T4 DNA ligase to ligate hybridization RCA primer and padlock probe. Subsequently, the prepared circular template solution was mixed with an RCA reaction mixture, including 10 μL of phi29 DNA polymerase (10 U/μL), 40 μL of dNTP (10 mM), 100 μL of 10 × phi29 polymerase buffer, and 750 μL of nuclease-free water, to react at 37 °C for 1 h to obtain RCA products. The RCA reaction was stopped by heating at 65 °C for 10 min to inactivate the phi29 DNA polymerase.
To synthesize the RCA-AuNPs complex, RCA products obtained from the previous step were mixed with 1 mL of ssDNA-AuNPs, after which the reaction mixture was treated in 95 °C for 10 min, quickly cooled in an ice bath for 1 min, and subsequently incubated at 37 °C for 30 min to allow hybridization between RCA products and ssDNA-AuNPs conjugates. Finally, the resulting solution was centrifuged for 20 min at 8000×g, and the ssDNA-AuNPs complex was collected at the bottom of the centrifuge tube. The RCA-AuNPs complex was resuspended in PBS (pH 7.4) and washed three times.
2.6 Evaluation of detection performances using SRBD samples
To evaluate the detection performances of the (G3.5+G4)-aptamer modified LSPR sensor chips, the SARS-CoV-2 SRBD was employed. Briefly, SRBD samples of different concentrations were separately injected into the (G3.5+G4)-aptamer modified LSPR sensor chip at a flow rate of 30 μL/min for 3.3 min, after which the sensor chip was regenerated using a regeneration buffer (i.e., glycine-HCL (10 mM, pH 2.0) at a flow rate of 150 μL/min for 0.67 min so that the sensor chip could be regenerated (Zhao et al., 2014).
2.7 Detection of SARS-CoV-2 pseudo viral particles with RCA-AuNPs signal amplifications
To evaluate the real-world sample detection performances of the newly developed sensor chips, pseudo SARS-CoV-2 viral particles were used as detection targets; in addition, the detection signals were subsequently intensified using an in situ amplification step based on RCA-AuNPs complexes. Specifically, different concentrations of pseudo viral particle suspensions were separately injected into the (G3.5+G4)-aptamer modified LSPR sensor chip at a flow rate of 30 μL/min for 3.3 min, after which the sensor chip was carefully rinsed using PBS (pH 7.4) at a flow rate of 30 μL/min until the signal baseline was stabilized. Subsequently, the detection signal was amplified in situ based on RCA-AuNPs complexes using either a one-step or two-step amplification format. For the one-step amplification format, the RCA-AuNPs complexes were directly injected into the sensor chip at a flow rate of 30 μL/min for 3.3 min. For the two-step amplification format, RCA products were first injected into the sensor chip at a flow rate of 30 μL/min for 3.3 min, after which ssDNA-AuNPs were injected into the sensor chip again at a flow rate of 30 μL/min for another 3.3 min. To regenerate the sensor chip, the LSPR sensor chip channel was washed using a regeneration buffer at a flow rate of 150 μL/min for 0.67 min to release the captured viral particles (Zhao et al., 2014). Caution: all experiments involving SARS-CoV-2 pseudo viral particles were carried out in a certified Risk Group II (RG II, Public Health Agency of Canada) laboratory.
3 Results and discussion
3.1 Characterization
Characterization results are detailed in SI, including characterization of carboxyl functionalized LSPR sensor chips (S2.1), (G3.5+G4)-aptamer surface modifications (S2.2), AuNPs and ssDNA-AuNPs conjugates (S2.3), and RCA products (S2.4).
3.2 Detection performance of (G3.5+G4)-aptamer sensor chip
The ability of sensor chip surfaces to prevent nonspecific adsorption of molecules is a critical consideration for enhancing LSPR sensor sensitivity (Lichtenberg et al., 2019); therefore, the nonfouling property of the modified sensor chips was investigated by evaluating the detection background noises generated by incubation with 1 mg/mL bovine serum albumin (BSA) as a nonspecific molecule. Specifically, four different sensor chips with different surface modification configurations (see Legend of Fig. 2 ), namely (1) the original sensor chips functionalized directly with aptamers (i.e., gold-aptamer, black curve), (2) the sensor chips with G3.5 modification on glass areas in-between the gold nanoislands and functionalized with aptamers (i.e., G3.5-aptamer, red curve), (3) the sensor chips with G4 modification on the gold nanoislands and functionalized with aptamers (i.e., G4-aptamer, blue curve), and (4) the sensor chips with both G3.5 and G4 modifications and functionalized with aptamers (i.e., (G3.5+G4)-aptamer, green curve) were tested. As shown in Fig. 2A, gold-aptamer modified (i.e., no dendrimers) sensor chips showed the highest background noises, indicating the highest amount of nonspecific surface adsorption; in contrast, the (G3.5+G4)-aptamer modified sensor chips exhibited the lowest background noises, suggesting excellent nonfouling properties of the modified surfaces. As expected, both the G3.5- and G4-aptamer modified sensor chips demonstrated significantly improved nonfouling properties compared to the gold-aptamer modified surfaces (i.e., no dendrimer immobilization on the sensor chip surface), as the background noises of these sensor chips were substantially lower than those of the gold-aptamer modified. This observation suggests that nonspecific adsorption occurs on both gold nanoislands and areas between the nanoislands and that the combined immobilization of G3.5 and G4 molecules can effectively prevent nonspecific adsorptions in both areas. It was noted that the background noise of the G3.5-aptamer surface was higher than that of the G4-aptamer surface (300 vs. 100 RU). This is most likely because the nonspecific adsorptions on/near the gold nanoislands (i.e., within the electromagnetic fields) contributed more to affecting the LSPR signals (Unser et al., 2015), further suggesting the importance of G4 immobilization on the surface of gold nanoislands. This significant improvement in nonfouling properties of the sensor chip surfaces -- as a result of functionalizing the sensor chip surface with a layer of dendrimers using combined (G3.5+G4) surface modifications -- results in significantly reduced background noises, thus potentially improving the detection performance of the LSPR sensor chips by enhancing the signal-to-noise ratio (Hao et al. 2023a, 2023b). It is worthwhile mentioning that the excellent nonfouling properties of the dendrimer-modified sensor chip eliminate the need for both blocking agents and reference channels specifically designed to reduce and compensate for surface nonspecific bindings in the LSPR assays, respectively (Lewis et al., 2021).Fig. 2 LSPR sensorgraphs of different tests to evaluate performances of different sensor chip surface modifications. A) Detection background noises of different sensor chips as evaluated in a nonspecific adsorption experiment using BSA (1 mg/mL). B) Comparison of the amount of aptamers immobilized on the surface of gold nanoislands of different sensor chips. C) Comparisons of the relative amount of aptamers immobilized on the surfaces of different sensor chips as indicated by the fluorescence intensity of the surface-immobilized fluorescently labeled aptamers; the dotted line indicates fluorescence intensity of original sensor chips (i.e.,144.2 ± 8.2 a.u. (n = 3)); and error bars indicate standard deviation, n = 3, the p values (*<0.05, **<0.01) were calculated by a Student’s t-test, 2 tails. D) Detection signals of SRBD (377.36 nM) on different modified surfaces. Note: all except the (G3.5+G4)-aptamer in Fig. 2D were blocked by BSA before SRBD testing. Legends signify surfaces with different modification strategies, including gold-aptamer (black curve), G3.5-aptamer (red), G4-aptamer (blue), and (G3.5+G4)-aptamer (green) for all tests.
Fig. 2
As the number of surface immobilized detection ligands (i.e., aptamers) is one of the most critical considerations affecting the detection sensitivity of biosensors (Hao et al. 2023a, 2023b), we studied the relative amount of aptamers immobilized on the surface of gold nanoislands of different modified sensor chips (i.e., gold-aptamer, G3.5-aptamer, G4-aptamer, and (G3.5+G4)-aptamer surfaces) by directly evaluating the LSPR sensorgraph signal changes before and after aptamer immobilizations. As shown in Fig. 2B, the gold-aptamer surface showed the lowest signal, suggesting the lowest amount of immobilized aptamers, likely due to a limited number of aptamer binding sites on the gold-nanoislands. In contrast, the (G3.5+G4)-aptamer sensor chip surface exhibited the highest aptamer immobilization signal, approximately 10 times higher than that of the gold-aptamer surface. This significant increase of immobilized aptamers amount was most likely caused by the presence of multi-handled G4 dendrimer templates (i.e., 64 binding sites per G4 molecule) via which the aptamers were immobilized on the surface of gold nanoislands. In addition, fluorescently labeled (i.e., cy3) aptamers were used to conjugate to the four differently modified sensor chips in order to compare the relative amount of the aptamer amount on the complete sensor chip surfaces by comparing the fluorescence intensities of each sensor chip. As shown in Fig. 2C, the fluorescence intensity of the (G3.5+G4)-aptamer surface was significantly higher than any other surfaces investigated (p<0.01). Therefore, it can be concluded that the (G3.5+G4) surface-modification configuration allows for the largest amount of aptamers to be tethered on the sensor chip capturing surfaces in this study.
Furthermore, to evaluate the performance of the four modified sensor chips (i.e., gold-aptamer, G3.5-aptamer, G4-aptamer, and (G3.5+G4)-aptamer surfaces), SRBD detection was carried out. For this study, the gold-aptamer, G3.5-aptamer, and G4-aptamer sensors were blocked with BSA prior to SRBD detection, while this blocking step was not performed for (G3.5+G4)-aptamer sensor chips. As shown in Fig. 2D, the detection signals from all 4 sensor chips showed patterns and trends similar to that of the gold nanoisland surface-immobilized aptamer signals seen in Fig. 2B. Specifically, the (G3.5+G4)-aptamer sensor chip showed the highest detection signal of 4490 RU, while detection signals from other sensor chips were 735 RU (gold-aptamer surface), 1080 RU (G3.5-aptamer surface), and 3900 RU (G4-aptamer surface), respectively. These results strongly indicate that the (G3.5+G4) surface modification on the sensor chip is the most effective design in conjugating a higher density of aptamers to the capturing surface, which results in an improved detection signal in comparison with other sensor chip modification approaches investigated in this study.
To further investigate the detection performances of the (G3.5+G4)-aptamer sensor chips, the sensor chips were used to detect different concentrations of SRBD samples. The same samples were also assayed using gold-aptamer, gold-antibody (i.e., antibodies against SRBD), and (G3.5+G4)-antibody modified sensor chips for comparisons. As shown in Fig. 3 , the (G3.5+G4)-aptamer modified sensor chips showed the highest detection signal at any given SRBD concentration assayed amongst all sensor chips investigated, including gold-aptamer, gold-antibody, and (G3.5+G4)-antibody modified surfaces. In addition, it is worth noting that the (G3.5+G4)-aptamer modified sensor chips demonstrated a much broader detection range (0.04 nM to 377.36 nM) than the gold-aptamer modified sensor chips (0.19 nM to 60.38 nM). This significantly increased detection range exhibited by the (G3.5+G4)-aptamer modified sensor chips was most likely caused by the presence of a greater amount of capturing aptamers on the (G3.5+G4) modified surface than on the non-modified sensor chips, thereby providing more target capturing capacity by allowing both more binding sites for target capturing and stronger binding avidity (Wang et al., 2022). Furthermore, the slope of the linear region of the response curve for the (G3.5+G4)-aptamer modified sensor chips (k = 367.56) was significantly greater (p<0.05) than those for the gold-aptamer (k = 48.71), gold-antibody (k = 3.58) and (G3.5+G4)-antibody (k = 5.77) modified sensor chips, suggesting a much higher detection sensitivity by the (G3.5+G4)-aptamer modified sensor chips (Hao et al. 2023a, 2023b).Fig. 3 Detection performances of different modified sensor chips over a wide range of SRBD concentrations; insets show linear regions of the response curves, LOD, and signal of negative controls; error bars indicate standard deviations, n = 3. Note: in obtaining the response curve, all detection parameters were optimized (S3, SI), and the gold-aptamer and gold-antibody sensor chips were properly blocked using BSA; the LSPR sensorgraphs for creating these concentration-response curves are shown in S6 (SI).
Fig. 3
The limit of detection, defined as the lowest target concentration to provide a signal at least three standard deviations greater than the signal from a negative control (Hao et al. 2023a, 2023b), was also calculated for each type of sensor chip. The results showed that the LOD of the (G3.5+G4)-aptamer modified sensor chip was 21.9 pM, which was around 9-time more sensitive than that of the gold-aptamer (205.2 pM) and 152-time more sensitive than that of the gold-antibody (3330.0 pM) modified sensor chip, respectively. In addition, a comparison between the (G3.5+G4)-antibody and the (G3.5+G4)-aptamer modified sensor chips reveals that the LOD of the (G3.5+G4)-antibody sensor chip was much higher (i.e., 1390.0 pM vs. 21.9 pM). This difference is likely due to a significantly lower amount of antibodies conjugated onto the sensor chip surfaces via the G3.5+G4 templates than aptamers, resulting in less capturing ligands on the sensor detection surfaces. Indeed, aptamers are much smaller molecules than antibodies (1 nm vs. 10 nm) (Hao et al., 2023b); this significant differences in sizes would allow more copies of aptamers to be conjugated to the sensor chip surfaces via the G3.5+G4 templates (size 4.5 nm) than their antibodies counterparts. In fact, we estimated the number of aptamers vs. antibodies on the (G3.5+G4) modified surfaces as reported elsewhere (De Feijter et al., 1978; Unser et al., 2015) (see details in S5, SI). The result showed that the number of immobilized surface aptamers was 16 to 44 times greater than that of the immobilized antibodies.
3.3 Detection of SARS-CoV-2 pseudo viral particles with RCA-AuNPs signal amplification (second-layer amplification)
To further study the detection performance of the (G3.5+G4)-aptamer sensor chips, we tested SARS-CoV-2 pseudo viral particles as detection targets. In comparison with the SRBD, the SARS-CoV-2 viral particles feature multiple spike proteins (i.e., 26) on their surfaces (Yao et al., 2020), thus allowing for detection signals to be further enhanced using a sandwich detection format (Huang et al., 2021). Since detection signals can be amplified using either a two-step amplification format (i.e., conjugation of captured viral particles with RCA products that are subsequently hybridized with ssDNA-AuNPs) or a one-step amplification format (i.e., conjugation of captured viral particles with RCA-AuNPs complex), both amplification strategies were tested and the amplified signals from the two methods were compared. As shown in Fig. 4 A, the sensorgraph of the two-step format (i.e., the black curve) showed a weak signal enhancement after the first step of signal amplification that involved RCA addition, with the signal intensified only by a factor of 1.7 (310 vs. 175 RU, see the black curve), an observation that was consistent with a previous study (He et al., 2014). However, the detection signal was significantly enhanced by approximately 6 times (1700 vs. 310 RU) after subsequent introduction of ssDNA-AuNPs. This is likely because a large number of ssDNA-AuNPs probes were hybridized with the RCA products, resulting in a significant enhancement of both the target mass and the plasmonic coupling effect (He et al., 2014; Kim et al., 2022). It should be noted that prolonging the RCA reaction time – thus the detection mass -- to better increase the mass of targets did not work well (data not shown), as longer RCA reaction times resulted in visible white magnesium pyrophosphate precipitates that blocked the LSPR channels (Mori et al., 2001). In comparison, the one-step amplification format (see the red curve, Fig. 4A) exhibited no significant difference (p>0.05) in the final amplified signal when compared with the two-step amplification format (see the black curve). However, since the one-step amplification format required fewer injection steps (i.e., 2 vs. 3 injections) and a shorter run time (500s vs. 800s), it was selected as the optimal signal intensification format for subsequent studies unless otherwise indicated.Fig. 4 Detection of SARS-CoV-2 pseudo viral particles with RCA-AuNPs signal amplification. A) LSPR sensorgarphs of one- and two-step amplification formats for signal amplification of SARS-CoV-2 pseudo viral particles (107 vp/mL); and B) the LSPR sensorgraph of detection of SARS-CoV-2 pseudo viral particles with RCA-AuNPs signal amplification; inset shows the relationship between the signals and the target concentrations as well as signal of the negative control; error bars indicate standard deviations, n = 3.
Fig. 4
To evaluate the detection performances of pseudo SARS-CoV-2 viral particles with RCA-AuNPs signal amplification, different concentrations of pseudo SARS-CoV-2 viral particles were detected using the (G3.5+G4)-aptamer modified sensor chips, followed by the signal amplification step using RCA-AuNPs. As shown in Fig. 4B, the sensorgraph presented low detection signals at any virus concentration before signal amplification was employed; in contrast, the detection signals were increased approximately 10-fold after the RCA-AuNPs complex was applied, indicating successful (and significant) detection signal intensification. Moreover, the LOD was calculated to be 148 vp/mL, one of the best sensitivities reported in whole viral particle detections amongst all detection platforms (Huang et al., 2021; Nguyen et al., 2016; Park et al., 2014; Yang et al., 2022). It should be noted that the typical SARS-CoV-2 viral concentration from nasopharyngeal and saliva swabbed samples is 104-1010 vp/mL (Huang et al., 2021), suggesting that the currently reported sensor chip modification and signal amplification approach can be readily used for early infection diagnostics to sensitively detect the SARS-CoV-2 virus. Moreover, the current approach directly detects whole viral particles; therefore, no sample pre-treatment would be required and the whole detection process can be done in 500 s, more efficient than any existing methods that require laborious sample preparations (Qiu et al., 2020; Yan et al., 2020). In addition, as shown in S4 (SI), the modified sensor chip can be regenerated multiple times using a regeneration solution (i.e., glycine-HCL (10 mM, pH 2.0)), significantly saving the overall detection time and cost.
3.4 Detection specificity and influence of biological matrices
To study the specificity of (G3.5+G4)-aptamer sensor chip for detecting the SRBD samples, the target (i.e., SARS-CoV-2 SRBD) and non-targets (i.e., SARS-CoV SRBD and Middle East Respiratory Syndrome (MERS)-CoV SRBD) were tested. As shown in Fig. 5 A (blue bars), signals for the target (i.e., SARS-CoV-2 SRBD) were significantly higher than those of non-targets (i.e., SARS-CoV SRBD and MERS-CoV SRBD), suggesting excellent detection specificity of the sensor chip for SARS-CoV-2 SRBD. To further study the specificity for pseudo viral particles detections, the SARS-CoV-2 pseudo viral particles (i.e., targets) and negative pseudo viral particles with no spike protein on the surface (i.e., non-targets) were detected using the (G3.5+G4)-aptamer sensor chip followed by signal amplification with RCA-AuNPs. As shown in Fig. 5A (yellow bars), the negative pseudo viral particles (i.e., non-target) showed a low signal value at 23 RU compared to SARS-CoV-2 pseudo viral particles (i.e., target) with a signal value at 970 RU, indicating an excellent detection specificity for SARS-CoV-2 pseudo viral particles. Moreover, the weak signal (7 RU) of the control sample (i.e., Control = no viral particles) indicates minimal nonspecific interactions between the detection surface and the RCA-AuNPs. It is interesting to note that in comparison with a previously reported study in which RCA-AuNPs was also used to amplify SPR sensor chip detection signals (He et al., 2014), the background noises detected in the current study were much lower, further demonstrating the advantages of nonfouling properties of PAMAM dendrimers on the sensor chip detection surface design that also combines signal amplifications.Fig. 5 Detection specificity and influence of sample matrices. A) The detection specificity of the (G3.5+G4)-aptamer modified LSPR sensor chip; the targets are SARS-CoV-2 SRBD and SARS-CoV-2 pseudo viral particles, and nonspecific targets are SARS-CoV SRBD, MERS-CoV SRBD, and negative pseudo viral particles; all SRBD samples (blue bars) were tested at 12 nM in PBS (pH 7.4), and all pseudo viral particle samples (yellow bars) were tested at 105 vp/mL in PBS (pH 7.4). B) Influence of sample matrices on detection performances; the SARS-CoV-2 SRBD were tested at 12 nM, and the SARS-CoV-2 pseudo viral particles were tested at 105 vp/mL. Error bars are standard deviations, n = 3.
Fig. 5
To investigate device performance under conditions that mimic real-world conditions, SARS-CoV-2 SRBD and SARS-CoV-2 pseudo viral particles were analyzed in PBS (pH 7.4), artificial saliva (1% v/v), and BSA (40 μg/mL) solutions (Lewis et al., 2021), as biological matrices are known to influence detection performances (Masson, 2020). As shown in Fig. 5B, in comparison with samples analyzed in PBS (pH 7.4), there was no significant (p>0.05) difference when either SRBD or pseudo viral particle samples were analyzed in saliva and BSA solutions. This is likely due to the nonfouling property of the modified detection surface, suggesting that the sensor chips proposed in this study were sufficiently robust to reliably detect samples in complex biomatrices.
Table 1 compares the detection sensitivities achieved in the current report with those reported in recent literature. As can be seen in the table that the LODs achieved in this study for both targets (i.e., SRBD and pseudovirus) were lower than most reported results. It is interesting to note that although the LOD for SRBD detection in this study was higher than the LOD reported by Yang et al., (2022) (i.e., 21.9 pM vs. 0.83 pM), the LOD in this study for SARS-CoV-2 pseudovirus detections was lower (i.e., 148 vp/mL vs. 391 vp/mL), potentially highlighting the added advantage of the RCA-AuNPs strategy for multiple signal amplification. In addition, unlike other conventional LSPR sensor chips designed for binding kinetic studies that aim for intermediate surface ligand densities to quickly reach binding equilibrium and to avoid mass transfer effects (Schasfoort, 2017), the current surface modification approach tries to maximize the surface ligand density to enhance detection sensitivities. As a result, the surface modification approach in combination with signal intensification strategies discussed in this study promise an interesting new direction for further exploration in order to expand the range of LSPR applications from what is conventionally only an analytical platform to a detection platform.Table 1 Sensor-based LSPR system for detection of SARS-CoV-2 spike proteins and viral particles.
Table 1Biosensor technique Target Immobilized ligands Sample matrix LOD Ref.
Polymer receptor functionalized plasmonic optical fibers Spike protein subunit 1 Molecularly imprinted polymer receptor Protein in buffer 58 nM (Cennamo et al., 2021a)
Aptamer-PEG immobilized gold nanofilm-based optical fiber SRBD Aptamer against SRBD Targets in diluted human serum 37 nM (Cennamo et al., 2021b)
Streptavidin-aptamer immobilized LSPR sensor chip Spike protein subunit 1 Aptamer against SRBD S1 protein diluted in buffer 0.26 nM (Lewis et al., 2021)
Antibody immobilized LSPR sensor chip with AuNPs signal amplification method SARS-CoV-2 pseudovirus Antibody against SRBD Pseudovirus in buffer 370 vp/mL (Huang et al., 2021)
ACE2 protein immobilized silver nanotriangle array SRBD and Coronavirus ACE2 protein Buffer and Saliva 0.83 pM for SRBD, and 391 vp/mL for Coronavirus (Yang et al., 2022)
Dendrimer-aptamer immobilized LSPR sensor chip SRBD and SARS-CoV-2 pseudovirus Aptamer against SRBD Targets diluted in BSA and saliva 21.9 pM for SRBD, and 148 vp/mL for pseudovirus This study
4 Conclusions
This study aims to improve the detection sensitivity of LSPR sensor chips to detect SARS-CoV-2 viral targets using a multi-pronged approach to both intensify detection signals and reduce background noises. To intensify the detection signals, three mechanisms have been employed: 1) surface immobilized PAMAM dendrimers as templates to introduce a large number of aptamers to the capturing sensor chip surfaces to enhance capturing probability and binding avidity for target recognition and capturing; 2) application of RCA to improve the target mass on the detection surface; and 3) use of AuNPs to increase both target mass and plasmonic coupling effect. To reduce the background noises, surface modification with two distinct PAMAM dendrimers that bind specifically to the different chip structures has been used to render the sensor chip surfaces nonfouling. Our results demonstrated that the modified sensor chips had a LOD of 21.9 pM for SRBD detection, approximately 152 times lower than that of the traditional antibody-based sensor chips. Our results also showed that the application of the RCA-AuNPs as the second-layer signal amplification further intensified detection signals by approximately 10 times and that the overall LOD was 148 vp/mL when used to detect pseudo SARS-CoV-2 viral particles, a sensitivity sufficiently high to suggest a potential application in early detection of COVID-19 infections. In addition, it is perceivable that the multi-pronged approach discussed in the current study can be readily adapted for other point-of-care applications based on already well-established LSPR/SPR detection platforms.
CRediT authorship contribution statement
Xingkai Hao: Conceptualization, Methodology, Validation, Formal analysis, Writing – original draft, Revision. Jean-Philippe St-Pierre: Writing – review & editing, Revision, Resources. Shan Zou: Investigation, Resources. Xudong Cao: Writing – review & editing, Revision, Supervision, Resources.
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.
Appendix A Supplementary data
The following is the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Data availability
Data will be made available on request.
Acknowledgements
This work was supported by the 10.13039/501100000038 Natural Sciences and Engineering Research Council of Canada (NSERC) Alliance COVID-19 Grants (ALLRP 550114-2020) and 10.13039/501100000038 NSERC Discovery Grants (RGPIN-2018-06370) to XC. XH is supported by a scholarship from 10.13039/501100004543 China Scholarship Council . We also thank Mr. Oltion Kodra from Energy, Mining and Environment Research Centre, National Research Council Canada, for the technical assistance with XPS analysis.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bios.2023.115421.
==== Refs
References
Amouzadeh Tabrizi M. Acedo P. Appl. Surf. Sci. 598 2022 153867
Cennamo N. D'Agostino G. Perri C. Arcadio F. Chiaretti G. Parisio E.M. Camarlinghi G. Vettori C. Di Marzo F. Cennamo R. Sensors 21 5 2021 1681 33804378
Cennamo N. Pasquardini L. Arcadio F. Lunelli L. Vanzetti L. Carafa V. Altucci L. Zeni L. Talanta 233 2021 122532
De Feijter J. Benjamins d.J. Veer F. Biopolymers: Original Research on Biomolecules 17 7 1978 1759 1772
Frens G. Nat. Phys. Sci. (Lond.) 241 105 1973 20 22
Hao X. Yang X. Zou S. Cao X. Appl. Surf. Sci. 610 2023 155603
Hao X. Yang X. Zou S. Cao X. Colloids Surf. B Biointerfaces 221 2023 113003
He P. Liu L. Qiao W. Zhang S. Chem. Commun. 50 12 2014 1481 1484
Hermanson G.T. Chapter 5 - homobifunctional crosslinkers Hermanson G.T. Bioconjugate Techniques third ed. 2013 Academic Press Boston 275 298
Huang L. Ding L. Zhou J. Chen S. Chen F. Zhao C. Xu J. Hu W. Ji J. Xu H. Biosens. Bioelectron. 171 2021 112685
Hurst S.J. Lytton-Jean A.K.R. Mirkin C.A. Anal. Chem. 78 24 2006 8313 8318 17165821
Jiang F. Xiao Z. Wang T. Wang J. Bie L. Saleh L. Frey K. Zhang L. Wang J. Chem. Commun. 58 52 2022 7285 7288
Jiang Y. Li S. Qiu Z. Le T. Zou S. Cao X. J. Food Saf. 39 5 2019 e12671
Kim H.-M. Kim H.-J. Park J.-H. Lee S.-K. Anal. Chim. Acta 1213 2022 339960
Lewis T. Giroux E. Jovic M. Martic-Milne S. Analyst 146 23 2021 7207 7217 34724009
Li S. Jiang Y. Yang X. Lin M. Dan H. Zou S. Cao X. Anal. Chim. Acta 1150 2021 338229
Lichtenberg J.Y. Ling Y. Kim S. Sensors 19 11 2019 2488 31159167
Lu M. Peng W. Lin M. Wang F. Zhang Y. Nanomaterials 11 3 2021 616 33804516
Masson J.-F. ACS Sens. 5 11 2020 3290 3292 33233896
Mori Y. Nagamine K. Tomita N. Notomi T. Biochem. Biophys. Res. Commun. 289 1 2001 150 154 11708792
Nguyen V.T. Seo H.B. Kim B.C. Kim S.K. Song C.S. Gu M.B. Biosens. Bioelectron. 86 2016 293 300 27387259
Park J.-W. Jin Lee S. Choi E.-J. Kim J. Song J.-Y. Bock Gu M. Biosens. Bioelectron. 51 2014 324 329 23994614
Qiu G. Gai Z. Tao Y. Schmitt J. Kullak-Ublick G.A. Wang J. ACS Nano 14 5 2020 5268 5277 32281785
Raziq A. Kidakova A. Boroznjak R. Reut J. Öpik A. Syritski V. Biosens. Bioelectron. 178 2021 113029
Samson R. Navale G.R. Dharne M.S. 3 Biotech 10 9 2020 385
Schasfoort R.B.M. Introduction to surface plasmon resonance Schasfoort R.B.M. Handbook of Surface Plasmon Resonance 2017 The Royal Society of Chemistry 1 26
Song Y. Song J. Wei X. Huang M. Sun M. Zhu L. Lin B. Shen H. Zhu Z. Yang C. Anal. Chem. 92 14 2020 9895 9900 32551560
Staros J.V. Biochemistry 21 17 1982 3950 3955 7126526
Tian B. Gao F. Fock J. Dufva M. Hansen M.F. Biosens. Bioelectron. 165 2020 112356
Tuerk C. Gold L. Science 249 4968 1990 505 510 2200121
Unser S. Bruzas I. He J. Sagle L. Sensors 15 7 2015 15684 15716 26147727
Wang Z. Yang X. Lee N.Z. Cao X. Micromachines 13 3 2022 436 35334728
Xiang Y. Deng K. Xia H. Yao C. Chen Q. Zhang L. Liu Z. Fu W. Biosens. Bioelectron. 49 2013 442 449 23811476
Xue J. Li Y. Liu J. Zhang Z. Yu R. Huang Y. Li C. Chen A. Qiu J. Talanta 248 2022 123605
Yan C. Cui J. Huang L. Du B. Chen L. Xue G. Li S. Zhang W. Zhao L. Sun Y. Clin. Microbiol. Infection 26 6 2020 773 779
Yang Y. Murray J. Haverstick J. Tripp R.A. Zhao Y. Sensor. Actuator. B Chem. 359 2022 131604
Yao H. Song Y. Chen Y. Wu N. Xu J. Sun C. Zhang J. Weng T. Zhang Z. Wu Z. Cheng L. Shi D. Lu X. Lei J. Crispin M. Shi Y. Li L. Li S. Cell 183 3 2020 730 738 e713 32979942
Zhao Q. Duan R. Yuan J. Quan Y. Yang H. Xi M. Int. J. Nanomed. 9 2014 1097
|
PMC010xxxxxx/PMC10205138.txt |
==== Front
Eur J Med Chem
Eur J Med Chem
European Journal of Medicinal Chemistry
0223-5234
1768-3254
Elsevier Masson SAS.
S0223-5234(23)00453-1
10.1016/j.ejmech.2023.115487
115487
Research Paper
Discovery of quinazolin-4-one-based non-covalent inhibitors targeting the severe acute respiratory syndrome coronavirus 2 main protease (SARS-CoV-2 Mpro)
Zhang Kuojun a1
Wang Tianyu a1
Li Maotian c1
Liu Mu d
Tang He a
Wang Lin a
Ye Ke a
Yang Jiamei a
Jiang Sheng a
Xiao Yibei a
Xie Youhua d∗∗∗
Lu Meiling c∗∗
Zhang Xiangyu b∗
a Department of Medicinal Chemistry, School of Pharmacy, China Pharmaceutical University, Nanjing, 210009, China
b Department of Biomedical Engineering, School of Engineering, China Pharmaceutical University, Nanjing, 210009, China
c Department of Pharmacology, School of Life Science and Technology, China Pharmaceutical University, Nanjing, 210009, China
d Key Laboratory of Medical Molecular Virology (MOE/NHC/CAMS), Shanghai Institute of Infectious Diseases and Biosecurity, School of Basic Medical Sciences, Fudan University, 200032, Shanghai, China
∗ Corresponding author.
∗∗ Corresponding author.
∗∗∗ Corresponding author.
1 Kuojun Zhang, Tianyu Wang and Maotian Li contributed equally to this work.
24 5 2023
5 9 2023
24 5 2023
257 115487115487
28 2 2023
30 4 2023
13 5 2023
© 2023 Elsevier Masson SAS. All rights reserved.
2023
Elsevier Masson SAS
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The COVID-19 pandemic caused by SARS-CoV-2 continues to pose a great threat to public health while various vaccines are available worldwide. Main protease (Mpro) has been validated as an effective anti-COVID-19 drug target. Using medicinal chemistry and rational drug design strategies, we identified a quinazolin-4-one series of nonpeptidic, noncovalent SARS-CoV-2 Mpro inhibitors based on baicalein, 5,6,7-trihydroxy-2-phenyl-4H-chromen-4-one. In particular, compound C7 exhibits superior inhibitory activity against SARS-CoV-2 Mpro relative to baicalein (IC50 = 0.085 ± 0.006 and 0.966 ± 0.065 μM, respectively), as well as improved physicochemical and drug metabolism and pharmacokinetics (DMPK) properties. In addition, C7 inhibits viral replication in SARS-CoV-2-infected Vero E6 cells more effectively than baicalein (EC50 = 1.10 ± 0.12 and 5.15 ± 1.64 μM, respectively) with low cytotoxicity (CC50 ＞ 50 μM). An X-ray co-crystal structure reveals a non-covalent mechanism of action, and a noncanonical binding mode not observed by baicalein. These results suggest that C7 represents a promising lead for development of more effective SARS-CoV-2 Mpro inhibitors and anti-COVID-19 drugs.
Graphical abstract
Image 1
Keywords
SARS-CoV-2 Mpro
Noncovalent Mpro inhibitors
Antiviral drugs
Antiviral activity
Handling Editor: Dr. Z Liu
==== Body
pmc1 Introduction
The Coronavirus Disease 2019 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has heavily impacted the global economy and threatened public health. According to the World Health Organization (WHO) report, there have been more than 645 million confirmed cases of COVID-19 worldwide, including more than 6.6 million deaths, as of 13 December 2022 [1]. Since the outbreak of COVID-19, enormous global efforts from academic and pharmaceutical companies have been made to discover preventive and therapeutic strategies to fight against this life-threatening disease [2,3]. To date, more than 20 effective SARS-CoV-2 vaccines have been made available globally, and play an important preventive role in controlling the COVID-19 pandemic. Nevertheless, a large number of people have not yet been vaccinated due to personal unwillingness or limited medicinal conditions; on the other hand, increasing numbers of breakthrough infections have been observed in convalescent and/or vaccinated populations, mainly due to decline in protective efficacy of vaccines with time since vaccination, or compromised effectiveness against the emerging omicron variants and subvariants [[4], [5], [6], [7], [8]]. Accordingly, the public are still at high risk of being infected, and effective antiviral drugs against SARS-CoV-2 and its emerging variants are still highly needed.
A large number of anti-COVID-19 drug targets have been reported, with SARS-CoV-2 main protease [Mpro; also known as 3C-like (3CL) protease or nonstructural protein 5 (nsp5)] capturing much attention. Mpro has been validated as an effective target for development of orally available small-molecule antiviral drugs [9,10]. Mpro cleaves viral polyproteins at 11 distinct sites to release functional non-structural proteins (nsps) that are essential for viral replication. Inhibition of Mpro is therefore able to block the viral replication and shut down the viral life cycle. Moreover, Mpro features a unique substrate specificity of glutamine (Gln) at the P1 position and no homologous human proteases have been known, and hence Mpro inhibitors are likely to cause no side-effects by interfering with host proteases. In contrast, Mpro are highly conserved among various coronaviruses, such as SARS-CoV-2, SARS-CoV and MERS, making it a promising target for development of broad-spectrum coronavirus antivirals. Notably, Mpro inhibitors should be effective against all the emerging variants of SARS-CoV-2, since the mutations occurring in spike proteins of variant strains cannot affect Mpro, and Mpro itself was shown to have an extremely low mutation rate in the emerging SARS-CoV-2 variants [[10], [11], [12]]. To date, a large number of SARS-CoV-2 Mpro inhibitors have been reported, and are mainly divided into two types: peptide-like covalent inhibitors represented by nirmatrelvir (PF-07321332, 1) [13], PF-00835231 (2) [14], PF-07304814 (3) [15], GC-376 (4) [16], MI-09 (5), MI-30 (6) [17] and compounds (7–9) [18,19], as well as nonpeptidic, noncovalent inhibitors represented by Ensitrelvir (S-217622, 10) [11], ML-188 (11), compound 12 [20], HL-3-68 (13) [21], baicalein (14) [22] and compounds 15–17 [[23], [24], [25]] (Fig. 1 ). Potential off-target effects, low membrane permeability and poor metabolic stability pose challenges for the extensive clinical application of peptidelike covalent inhibitors, notwithstanding the approval of Paxlovid (a combination of PF-07321332 and ritonavir) by the US Food and Drug Administration (FDA). The nonpeptidic, noncovalent Mpro inhibitor seems to be a more attractive modality. Ensitrelvir (S-217622) developed by Shionogi & Co., Ltd. showed robust antiviral potency against SARS-CoV-2 and its variants, excellent selectivity for Mpro over host proteases, as well as an outstanding DMPK profile in preclinical models [11]. Ensitrelvir (S-217622) exhibited promising single-agent antiviral potency and safety profiles in phase II/III clinical trials [26]. More recently, it was approved for clinical application by the Pharmaceuticals and Medical Device Agency (PMDA). But effective nonpeptidic, non-covalent Mpro inhibitors are very limited, and it is still necessary to search for more non-covalent Mpro inhibitors with diverse chemical scaffolds and improved properties.Fig. 1 Chemical structures of representative peptide-like covalent, and nonpeptidic, noncovalent SARS-CoV-2 Mpro inhibitors.
Fig. 1
Here, we describe our discovery of quinazolin-4-one-based nonpeptidic, non-covalent inhibitors of SARS-CoV-2 Mpro that are derived from baicalein (14). Baicalein was the first reported nonpeptidic, noncovalent inhibitor of SARS-CoV-2 Mpro discovered by Shanghai Institute of Materia Medica, Chinese Academy of Sciences [22], and its good biochemical and cellular antiviral activity captured our interest. Nevertheless, the potency of baicalein requires further structural optimization and extensive structure activity relationship (SAR) studies. Moreover, the drug metabolism and pharmacokinetics (DMPK) properties and target specificity were largely unknown. In order to fill in gaps in this area and explore the potential of this series as a lead for the development of anti-COVID-19 drugs, we firstly used a scaffold hopping strategy turning baicalein's chromen-4-one core to the alternative privileged scaffold, quinazolin-4-one. In this way, we discovered the first quinazolin-4-one-based SARS-CoV-2 Mpro inhibitor, and further structural optimizations allowed us to produce quinazolin-4-one-based SARS-CoV-2 Mpro inhibitors that are superior to baicalein in terms of biochemical potency, cellular antiviral activity and DMPK profile. An X-ray cocrystal structure of D8 in SARS-CoV-2 Mpro revealed a ligand-induced conformation change, which allows the sec-butyl moiety of D8 to occupy a newly formed binding site between the canonical S1′ and S2 subpockets, suggesting a binding mode to Mpro of quinazolin-4-one-based SARS-CoV-2 Mpro inhibitors that is different from that of baicalein.
2 Results and discussion
2.1 Design of the quinazolin-4-one class of SARS-CoV-2 Mpro inhibitors
The active site of Mpro is composed of five subpockets, S4–S1′, which can accommodate substrate and inhibitor groups at positions P4–P1′, with a Cys145-His41 catalytic dyad. As summarized by researchers from Shionogi Pharmaceutical in a recent publication, the pharmacophore based on the known Mpro inhibitors generally includes: (i) a hydrogen acceptor that interacts with the side-chain NH of His163 in the S1 subpocket, (ii) a hydrogen acceptor that forms a hydrogen bond with the main-chain NH of Glu166, and (iii) a lipophilic group in the S2 subpocket [11]. As shown in Fig. 2 A, the cocrystal structure of baicalein (14) in SARS-CoV-2 Mpro (PDB code: 6M2N [22]) showed that baicalein follows this pharmacophore model. Specially, three phenolic hydroxyl groups in baicalein form a hydrogen-bond network with the side chains of Ser144/His163 and main chains of Leu141/Gly143 directly or indirectly through water molecules. The carbonyl group at C4 position forms a critical hydrogen bond with the Glu166. Accordingly, three phenolic hydroxyl groups and C4 carbonyl group are necessary for the potency, and therefore were retained in our design. The free phenyl ring (C ring) occupies the S2 subpocket with extensive hydrophobic interactions. The core chromen-4-one forms many key interactions contributing to baicalein's binding affinity with protein: (i) the A ring forming S–π and NH2–π with the catalytic Cys145 and Asn142, respectively; (ii) the B ring π–π stacking with catalytic His41, and forming hydrophobic interactions with Met165. In addition, the chromen-4-one motif acts as a scaffold to assemble these necessary components. We envisioned that other aromatic heterocycles with similar structure could replace the chromen-4-one structure in baicalein. Because quinazolin-4-one and quinolin-4-one have been often seen in many natural products, biologically active compounds and clinically used drugs, together with well-established efficient synthesis and functionalization methodologies of these heterocycles, we firstly used a scaffold hopping strategy to replace the core chromen-4-one in baicalein with quinazolin-4-one or quinolin-4-one, respectively, in order to discover structurally novel non-covalent Mpro inhibitors. The resulting quinolin-4-one-based compound 18 had a complete loss of Mpro enzymatic inhibitory activity, while quinazolin-4-one-based compound 19 exhibited comparable inhibitory activity with that of baicalein. Therefore, we proceeded with 19 as a lead compound. A step-by-step optimization strategy and SAR studies of substituent groups at the C2 and N3 positions of quinazolin-4-one were conducted (Fig. 2B). Finally, the optimal substituents at the C2 and N3 positions were integrated, leading to a series of potent SARS-CoV-2 Mpro inhibitors (Fig. 2B).Fig. 2 Rational design of non-covalent SARS-CoV-2 Mpro inhibitors. (A) Binding pocket of baicalein in SARS-CoV-2 Mpro (PDB code 6M2N). (B) Detailed binding mode of baicalein complexed with Mpro. The protein is shown in gray cartoon, baicalein in blue sticks, and the selected residues in yellow sticks. Hydrogen bonds are indicated as gray dashes, and water molecule as red sphere. S–π, NH2–π and π–π stacks are indicated as green dashes. (C) Step-by-step optimization strategy of non-covalent SARS-CoV-2 Mpro inhibitors starting from baicalein.
Fig. 2
2.2 Chemistry
As shown in Scheme 1 , the quinolin-4-one derivative 18 was obtained by the base-mediated cyclization of N-(ketoaryl)amide (the Camps cyclization) according to the published literature [27]. Condensation of compound 20 and benzoyl chloride afforded benzamide (21). The Friedel-Crafts acylation of 21 using acetyl chloride in the presence of SnCl4 gave the cyclization precursor (22) [28], which was then cyclized at 110 °C in the presence of KOH to generate compound 23 [27], and this was followed by BBr3-mediated demethylation to yield the desired product (18).Scheme 1 Synthesis of compound 18. Reagents and conditions: (a) benzoyl chloride, TEA, THF, 0 °C to rt, 2 h, 91%; (b) acetyl chloride, SnCl4, anhydrous DCM, 0 °C to rt, 5 h, 45%; (c) KOH, anhydrous 1,4-dioxane, reflux, 4 h, 35%; (d) BBr3/DCM, DCM, −10 °C to rt, 24 h, 75%.
Scheme 1
There are many efficient ways to construct quinazolin-4-one derivatives. We selected the appropriate synthetic method for each target compound based on the availability of raw materials. Compounds A1–A6, A8–A12, A14, A16, A18, A19, A21–A27 and 19 were obtained in two steps: (i) the reaction of key intermediate 24 [29] with various amidine hydrochlorides 25a–25w that are commercially available or easily synthesized in efficient mild copper-catalyzed conditions to yield compounds 26a–26w [30]; (ii) BBr3-mediated demethylation of compounds 26a–26w (Scheme 2 ). A one-pot I2-mediated oxidative cyclization of o-anthranilamide 27 [31] and various commercially available aldehydes 28a–28v [32] and subsequent demethylation were used to synthesize compounds A7, A13, A15, A17, A20 and compounds B1–B17 (Scheme 3 ). As depicted in Scheme 4 , the N 3 substituted quinazolin-4-one derivatives C1–C15 were prepared by condensation of the prepared o-anthranilamides 30a–30o with benzaldehyde under similar reaction conditions with B series compounds followed by demethylation. Compounds D6–D9 were obtained by the oxidative cyclization approach that was used to access compounds in the B and C series in a good yield (Scheme 7), but compounds D1–D5 and D10–D12 could not be synthesized under the same conditions. D1–D4, D10 and D11 were successfully achieved by cyclization of N-acylanthranilic acids and amines in the presence of PCl3 in a good yield (Scheme 5 ) [33]. However, N 3-alkyl substituted compounds D5 and D12 could not be prepared with the synthetic method shown in Scheme 5, and these were obtained by a one-pot, two step cyclization: (i) the reaction of compound 36 with the prepared 2-phenylpropanoyl chloride or commercially available phenylacetyl chloride in the presence of triphenyl phosphite (TPP) and pyridine, and (ii) the substitution with isobutylamine [34], followed by demethylation (Scheme 6 ).Scheme 2 Synthesis of compounds A1–A6, A8–A12, A14, A16, A18, A19, A21–A27 and 19. Reagents and conditions: (a) CuI, CsCO3, DMF, rt, overnight, 55%–90%; (b) BBr3/DCM, DCM, −10 °C to rt, 24–36 h, 35%–64%.
Scheme 2
Scheme 3 Synthesis of compounds A7, A13, A15, A17, A20, B1–B17. Reagents and conditions: (a) I2, EtOH, reflux, 3–5 h, 40%–80%; (b) BBr3/DCM, DCM, −10 °C to rt, 24–36 h, 40%–63%.
Scheme 3
Scheme 4 Synthesis of compounds C1–C15. Reagents and conditions: (a) I2, EtOH, reflux, 3–5 h, 45%–87%; (b) BBr3/DCM, DCM, −10 °C to rt, 24–36 h, 35%–57%.
Scheme 4
Scheme 5 Synthesis of compounds D1–D4, D10 and D11. Reagents and conditions: (a) PCl3, 50 °C, 4–12 h, 70%–82%; (b) BBr3/DCM, DCM, −10 °C to rt, 24–36 h, 45%–63%.
Scheme 5
Scheme 6 Synthesis of compounds D5 and D12. Reagents and conditions: (a) For 37a: (i) 2-phenylpropanoic acid (38), (COCl)2, DCM, cat. DMF, 0 °C to rt, 2 h; (ii) triphenyl phosphite (TPP), Py, 0 °C–70 °C, 2 h; iii) isobutylamine (40), 70 °C, overnight, 40%; (b) For 37b: (i) phenylacetyl chloride (39), TPP, Py, 0 °C–70 °C, 2 h; (ii) isobutylamine (40), 70 °C, overnight, 45%; (c) BBr3/DCM, DCM, −10 °C, 24 h, 50%–60%.
Scheme 6
Scheme 7 Synthesis of compounds D6–D9. Reagents and conditions: (a) I2, EtOH, reflux, 5 h, 72%–88%; (b) BBr3/DCM, DCM, −10 °C to rt, 24 h, 55%–56%.
Scheme 7
2.3 SAR studies of the quinazolin-4-one class of SARS-CoV-2 Mpro inhibitors
All target compounds were preliminarily evaluated for inhibitory activity against SARS-CoV-2 Mpro in fluorescence resonance energy transfer (FRET)-based enzymatic assays. As shown in Table 1 , quinolin-4-one-based compound 18 had a complete loss of enzymatic inhibitory activity against SARS-CoV-2 Mpro. To our delight, quinazolin-4-one-based compound 19 was only slightly less potent than baicalein (IC50 = 1.372 ± 0.047 μM and 0.966 ± 0.065 μM, respectively). With the promising potency for compound 19 in hand, we firstly performed the SAR analysis of the substituents on the free phenyl group (C ring) in 19. Generally, the positions on the phenyl ring led to this order of potency: 2′-position > 3′-position > 4′-position (Table 1). In particular, compounds with fluorine (F), trifluoromethyl (CF3), trifluoromethoxy (OCF3) or tert-butyl at the C2′ position exhibited much more potency than the corresponding compounds with 3′ and/or 4′ substituents (A1 vs A2 and A3, A13 vs A14, A15 vs A16, A21 vs A22). The significant influence of the substituent position on activity can also be concluded from the difluorine-substituted compounds: 2′,3′-difluoro (A25, IC50 = 1.130 ± 0.031 μM) > 2′,4′-difluoro (A26, IC50 = 1.541 ± 0.042 μM) > 3′,4′-difluoro (A27, IC50 = 2.716 ± 0.051 μM).Table 1 SAR exploration of the substituent groups on the C ring of quinazolin-4-one a.
Table 1Compd. R1 IC50 (μM)
A1 2′-F 1.443 ± 0.060
A2 3′-F 5.132 ± 0.094
A3 4′-F 12.760 ± 0.067
A4 2′-Cl 0.435 ± 0.041
A5 3′-Cl 6.960 ± 0.076
A6 4′-Cl 7.706 ± 0.055
A7 2′-Br 0.554 ± 0.041
A8 3′-Br 8.017 ± 0.033
A9 4′-Br 9.797 ± 0.051
A10 2′-methyl 0.365 ± 0.033
A11 3′-methyl 2.277 ± 0.029
A12 4′-methyl 1.867 ± 0.034
A13 2′-CF3 1.481 ± 0.047
A14 3′-CF3 10.53 ± 0.050
A15 2′-OCF3 1.798 ± 0.061
A16 4′-OCF3 ＞20
A17 2′-NO2 10.588 ± 0.040
A18 3′-NO2 11.169 ± 0.034
A19 4′-NO2 ＞20
A20 3′-isopropyl 7.635 ± 0.058
A21 3′-tert-butyl 8.423 ± 0.031
A22 4′-tert-butyl ＞20
A23 2′-OH 18.285 ± 0.040
A24 4′-OH ＞20
A25 2′,3′-di-F 1.130 ± 0.031
A26 2′,4′-di-F 1.541 ± 0.042
A27 3,4′-di-F 2.716 ± 0.051
19 – 1.372 ± 0.047
baicalein (14) – 0.966 ± 0.065
a Baicalein was used as the positive control. Inhibitory activity against SARS-CoV-2 Mprowas determined with the FRET protease activity assay. Values represent a mean ± SD of at least three independent experiments.
The volume, polarity and electronegativity of the substituent groups also affect the potency significantly. Introduction of a bulky isopropyl (A20) and tert-butyl (A21 and A22), or a polar hydroxyl (A23 and A24) and a nitro group (A17–A19) resulted in a sharp decline in potency. It seemed that less bulky substituents with moderate-to-low electronegativity benefit the inhibitory activity against Mpro. A10 with 2′-methyl and A4 with 2′-chlorine were the two most active compounds among the A series, with IC50 values of 0.365 ± 0.033 μM and 0.435 ± 0.04 μM, respectively. Adding a bromine (Br) atom at the C2′ position, whose van der Waals radius is slightly larger than that of chlorine (Cl) atom and whose electronegativity is lower than that of Cl, led to compound A7, which showed a slightly decreased potency relative to A4 (A7, IC50 = 0.554 ± 0.041 μM), but was more active than 19 and baicalein. The introduction of more electronegative substituents at the C2′ position led to a poorer potency against Mpro, for example, F (IC50 = 1.443 ± 0.060 μM) > CF3 (IC50 = 1.481 ± 0.047 μM) > nitro group (IC50 = 10.588 ± 0.040 μM). Compound A7 containing an electron-donating trifluoromethoxy (OCF3) group was slightly less potent than compound A13 with an electron-withdrawing CF3 (IC50 = 1.798 ± 0.061 μM for A7 vs 1.481 ± 0.047 μM for A13), which might be attributed to the larger volume of OCF3 than that of CF3, and the volume effect affected the potency more significantly. Together, these results suggested that the position, steric size, polarity and electronic property of substituents on the C ring integratedly affect the inhibitory activity against SARS-CoV-2 Mpro.
Next, we replaced the C ring with different types of groups in order to diversify the structures of quinazolin-4-one based Mpro inhibitors and search for more desirable substituents at the C2 position. As shown in the Table 2 , the replacement of phenyl (the C ring) with naphthyl and heteroaromatic rings, including thienyl, 3-pyridyl, 4-pyridyl and pyrazolyl, led to compounds B1–B5, and resulted in a remarkable decrease in potency against Mpro. Different cycloalkyl replacements afforded compounds B6–B8. Substituting phenyl with cyclopentyl (B7, IC50 = 0.539 ± 0.061 μM) led to improved potency against Mpro over that of 19 (IC50 = 1.372 ± 0.047 μM) and baicalein (14) (IC50 = 0.966 ± 0.065 μM). B8 with a cyclohexyl (IC50 = 1.370 ± 0.140 μM) was equipotent with 19, while B6 with cyclopropyl (IC50 = 5.485 ± 0.791 μM) remarkably decreased the potency. The phenyl was replaced by different alkyl groups leading to compounds B9–B12. The introduction of isopropyl (B9) and tert-butyl (B10) resulted in a sharp drop of the potency. In contrast, B11 with sec-butyl (IC50 = 0.385 ± 0.024 μM) and B12 with a bulkier tert-amyl (IC50 = 0.970 ± 0.075 μM) were more potent than or comparable to 19, respectively. Subsequently, we explored the effects of styryl and benzyl groups at the C2 position on the inhibitory activity against Mpro. Compound B13 with a rigid styryl (IC50 = 2.874 ± 0.030 μM) led to some loss of potency against Mpro compared to 19. The relatively flexible benzyl group in compound 14 improved the IC50 value to 0.327 ± 0.052 μM. Small groups, such as methyl or ethyl, when introduced to the benzyl in 14 could further increase the activity, and the resulting compounds 15 and 16 exhibited the most potent inhibitory activity against Mpro among B series, with IC50 values of 0.174 ± 0.038 μM and 0.210 ± 0.028 μM, respectively. Addition of a larger isopropyl group (compound 17) led to a slight loss of the potency of 14, with an IC50 value of 0.390 ± 0.048 μM.Table 2 SAR exploration of substituent groups at the C2 position of quinazolin-4-onea.
Table 2Compd R2 IC50 (μM)
B1 Image 3 ＞20
B2 Image 4 ＞20
B3 Image 5 15.19 ± 0.0405
B4 Image 6 ＞20
B5 Image 7 3.565 ± 0.295
B6 Image 8 5.485 ± 0.791
B7 Image 9 0.539 ± 0.061
B8 Image 10 1.370 ± 0.140
B9 Image 11 4.943 ± 0.504
B10 Image 12 4.086 ± 0.647
B11 Image 13 0.385 ± 0.024
B12 Image 14 0.970 ± 0.075
B13 Image 15 2.874 ± 0.030
B14 Image 16 0.327 ± 0.052
B15 Image 17 0.174 ± 0.038
B16 Image 18 0.210 ± 0.028
B17 Image 19 0.390 ± 0.048
19 Image 20 1.372 ± 0.047
baicalein (14) – 0.966 ± 0.065
a Baicalein was used as the positive control. Inhibitory activity against SARS-CoV-2 Mprowas determined by using the FRET protease activity assay. Values represent a mean ± SD of at least three independent experiments.
We tentatively kept R2 as phenyl and explored the SAR at the N3 position (Table 3 ). Generally, C series compounds with different substituents at the N-3 position of quinazolin-4-one exhibited improved inhibitory potency against Mpro compared to compound 19 and baicalein (14), except for compounds C2 with the polar hydroxyethyl (IC50 = 1.365 ± 0.062 μM), C3 with a bulky tert-butyl (0.949 ± 0.077 μM) and C14 with a biphenylyl (IC50 = 1.476 ± 0.117 μM). For cycloalkyl replacements, cyclopentyl was the optimal substituent group, and the corresponding compound C5 showed significantly increased potency relative to 19, with an IC50 value of 0.124 ± 0.014 μM (Table 3, Fig. S1). C1 with a less bulky sec-butyl substituent showed potency as good as C5 (Table 3, Fig. S1). The introduction of phenyl leading to compound C7 improved the IC50 value to 0.085 ± 0.006 μM (Table 3, Fig. S1). Inspired by this result, different substituted phenyl groups were introduced to the N3 position. F, Br and hydroxyl substituents affording compounds C8–C11 retained the most of the potency of C7, showing IC50 values of about 0.2 μM. Compound C12 containing a 3′-methyl-4′-fluorophenyl substituent was equipotent with C7 (IC50 = 0.117 ± 0.016 μM, Table 3, Fig. S1). In addition, adding a 3-pyridyl or 4-pyrazolyl at the N3 position resulted in compounds C13 and C14, which had good inhibitory activity against Mpro (IC50 = 0.212 ± 0.024 and 0.390 ± 0.003 μM, respectively).Table 3 SAR exploration of substituent groups at the N3 position of quinazolin-4-onea.
Table 3Compd R3 IC50 (μM)
C1 Image 22 0.124 ± 0.018
C2 Image 23 1.365 ± 0.062
C3 Image 24 0.949 ± 0.077
C4 Image 25 0.290 ± 0.028
C5 Image 26 0.124 ± 0.016
C6 Image 27 0.274 ± 0.022
C7 Image 28 0.083 ± 0.006
C8 Image 29 0.205 ± 0.033
C9 Image 30 0.236 ± 0.018
C10 Image 31 0.271 ± 0.018
C11 Image 32 0.207 ± 0.024
C12 Image 33 0.117 ± 0.016
C13 Image 34 1.476 ± 0.117
C14 Image 35 0.212 ± 0.024
C15 Image 36 0.390 ± 0.003
19 Image 37 1.372 ± 0.047
baicalein (14) – 0.966 ± 0.065
a Baicalein was used as the positive control. Inhibitory activity against SARS-CoV-2 Mprowas determined by using the FRET protease activity assay. Values represent a mean ± SD of at least three independent experiments.
D series compounds were derived from the combination of the desirable substitutions at the C2 and N3 positions that we have identified (Table 4 ). Unexpectedly, the combination of 2-methyl-benzyl that is the optimal substitution at the C2 position with phenyl and sec-butyl groups that were the favorable substituents at the N3 position resulted in decreased activity compared to the corresponding C series compounds. In particular, compounds D2 (IC50 = 1.005 ± 0.136 μM) and D5 (IC50 = 0.692 ± 0.119 μM) showed a sharp decrease in potency relative to the corresponding compounds C12 (IC50 = 0.117 ± 0.016 μM) and C1 (IC50 = 0.124 ± 0.018 μM). The possible reason for the decreased inhibitory activity was that C2-2-methyl-benzyl clashed with N3-phenyl or N3-sec-butyl, leading one or both of them to be unable to smoothly enter into the corresponding Mpro protein subpockets. N3-sec-butyl series compounds D6–D9 did not improve the potency compared to C7, with IC50 values ranging from 0.100 to 0.284 μM. The combination of C2-benzyl with N3-phenyl resulted in compound D10 being equipotent with C7 (IC50 = 0.103 ± 0.014 μM), while the combination of C2-benzyl with 3-methyl-4-fluorobenzyl and sec-butyl at the N3 position led to some loss of the potency relative to C7.Table 4 SAR exploration of substituent groups at C-2 and N-3 positions of quinazolin-4-onea.
Table 4Compd R4 R5 IC50 (μM)
D1 Image 39 Image 40 0.477 ± 0.078
D2 Image 41 Image 42 1.005 ± 0.136
D3 Image 43 Image 44 0.290 ± 0.030
D4 Image 45 Image 46 0.386 ± 0.017
D5 Image 47 Image 48 0.692 ± 0.119
D6 Image 49 Image 50 0.107 ± 0.023
D7 Image 51 Image 52 0.239 ± 0.028
D8 Image 53 Image 54 0.100 ± 0.012
D9 Image 55 Image 56 0.284 ± 0.028
D10 Image 57 Image 58 0.131 ± 0.022
D11 Image 59 Image 60 0.502 ± 0.038
D12 Image 61 Image 62 0.466 ± 0.040
C7 Image 63 Image 64 0.085 ± 0.006
a C7 was used as the positive control. Inhibitory activity against SARS-CoV-2 Mprowas determined with the FRET protease activity assay. Values represent a mean ± SD of at least three independent experiments.
2.4 Target validation and selectivity
To examine if the inhibitory activity against Mpro is associated with covalent modification of cysteine residues, we compared the inhibition of C5, C7, D6, D8 and D10 in the presence or absence of the reducing agent dithiothreitol (DTT) (Table S1). Addition of DTT guarantees accessible cysteine residues of a protein at their reduced state and will indicate whether an inhibitor acts by a false mechanism of redox-cycling [24,35]. There were no significant differences in the IC50 values in the presence of DTT and in the absence of DTT.
One of the major limitations encountered by cysteine protease inhibitors is the target selectivity. Compound C7 was selected as an example of our target scaffold to be profiled with the selectivity over several common human proteases, including caspase 2, cathepsin L, thrombin, cathepsin B and cathepsin D. Compound C7 showed no apparent inhibitory activity against these host proteins at a concentration of 10 μM, suggesting that C7 had a good target specificity against Mpro over host proteases (Table S2).
2.5 In vitro DMPK profiling
As shown in Table S3, compounds C7 (Papp AB = 9.67 × 10−6 cm/s, Efflux radio (ER) = 0.44), D6 (Papp AB = 7.45 × 10−6 cm/s, ER = 0.40), and D8 (Papp AB = 7.83 × 10−6 cm/s, ER = 0.41) exhibited improved Madin-Darby canine kidney (MDCK) membrane permeability relative to baicalein (Papp AB = 6.07 × 10−6 cm/s, ER = 0.34). The kinetic solubility in phosphate buffer solution (PBS, pH = 7.4) of C7 and D6 was measured as 179.35 and 155.98 μM, respectively, which were superior to baicalein (112.18 μM). C7, D6, D8 and baicalein exhibited reasonable plasma protein binding (PPB) in human plasma (fraction unbound fu = 0.70%, 2.61%, 3.17% and 0.93%, respectively). Baicalein exhibited high intrinsic clearance in human liver microsome (HLM), with clearance rate (CLint) of 333.05 μL/min/mg protein and half-time (T1/2) of 4.16 min. C7 and D6 showed improved HLM stability relative to baicalein (CLint = 108.68 and 68.34 μL/min/mg protein, respectively; T1/2 = 12.75 and 20.28 min). The two compounds showed relatively high HLM metabolic stability in the presence of the cofactor NADPH alone, indicative of mainly undergoing phase II metabolism. With the presence of three phenolic hydroxyl groups in the structures, the above results were not unexpected. Despite the improvement relative to baicalein, the metabolic stability of this quinazolin-4-one series inhibitors requires further optimization.
2.6 Cellular cytotoxicity and antiviral activity
The antiviral activity of compounds C7, D6 and D8 was evaluated in SARS-CoV-2-infected Vero E6 cells. Prior to antiviral assays, the cytotoxicity of selected Mpro inhibitors against Vero E6 cells was evaluated by using cell counting kit-8 (CCK-8) assays. The test compounds exhibited no significant cytotoxicity at the highest tested concentration (50% cytotoxicity concentration, CC50 > 50 μM, Table 5 and Fig. S2). The antiviral activity was determined in a yield reduction assay that assesses the inhibitory activity of the test compounds on the viral replication using quantitative real-time polymerase chain reaction (qRT-PCR), with baicalein as the positive control. As shown in Table 5 and Fig. S2, compounds C7 (EC50 = 1.10 ± 0.12 μM), D6 (EC50 = 2.87 ± 1.43 μM) and D8 (EC50 = 2.11 ± 1.12 μM) showed better antiviral activity in Vero E6 cells than baicalein (EC50 = 5.15 ± 1.64 μM).Table 5 Antiviral activity and cytotoxicity of selected compounds in Vero E6 cells.
Table 5Compd. EC50 (μM)a CC50 (μM)
C7 1.10 ± 0.12 ＞50
D6 2.87 ± 1.43 ＞50
D8 2.11 ± 1.16 ＞50
baicalein 5.15 ± 2.46 ＞50
a Inhibitory effect on viral replication induced by SARS-CoV-2 infection in Vero E6 cells (RT-qPCR assay). Values are expressed as mean ± SD from three independent experiments.
2.7 X-ray crystal structure of D8 in complex with SARS-CoV-2 Mpro
To further examine the binding mode of quinazolin-4-one inhibitors in SARS-CoV-2 Mpro, we resolved an X-ray co-crystal structure of D8 complexed with Mpro at a resolution of 2.2 Å (PDB code 8I4S, Fig. 3 A–3C). The X-ray data and refinement statistics are depicted in Table S4. The trihydroxyphenyl moiety of D8 sits in a similar position as in the baicalein/Mpro complex (Fig. 3B and E). Three phenolic hydroxyl groups of D8 form critical hydrogen bond interactions with the main chains of Gly143/Ser144/Gly145, and also form hydrogen bond interactions with the side chain of His163 via a buried water molecule (Fig. 3B). Distinct from the case of baicalein, the C4 carbonyl oxygen atom of D8 was a little aside from Glu166, and forms a hydrogen bond with the backbone NH of Glu166 with the aid of a buried water molecule (Fig. 3B). Unexpectedly, the 3′-methyl-4′-fluorophenyl at the N3 position occupies the S2 pocket, forming hydrophobic interactions with Gln189 and Met49, while the sec-butyl at the C2 position projects into a newly formed binding site that was observed in CCF0058981 analogue (12) complexed with Mpro (PDB code 7LMF), termed the S2c pocket by Han et al. (Fig. 3D).20 When D8 bound to Mpro, the flexible side chains of some amino acid residues of the protein, particularly Met49 and Gln189, exhibited a ligand-induced conformation change relative to most other inhibitor/Mpro structures (e.g., baicalein, nirmatrelvir and ensitrelvir), further rearranging the binding surface of subpockets S4 and S2 (Fig. 3D and S3). This allows the sec-butyl to occupy the S2c pocket, forming favorable hydrophobic interactions with Cys44, Thr45 and Ser46. As with baicalein, the phenyl ring of D8 with three hydroxyl groups is sandwiched between Cys145 and Asn142 by forming S–π and NH2-π with Cys145 and Asn142, respectively (Fig. 3B). In addition, the middle ring π–π stacks with catalytic His41, also contributing to the binding affinity of D8 with protein (Fig. 3B).Fig. 3 (A) Overview of the co-crystal structure of D8 bound to SARS-CoV-2 Mpro (PDB code 8I4S). The protein is shown in cartoon, and domains I, II and III are colored light orange, magenta and violet, respectively. D8 is shown as spheres with carbons in cyan. (B) Detailed binding mode of compound D8 in complex with SARS-CoV-2 Mpro. The protein is shown in gray cartoon, D8 in cyan sticks, and the selected residues in yellow sticks. Hydrogen bonds are indicated as gray dashes, and the water molecule as a red sphere. S–π, NH2–π and π–π stacks are indicated as green dashes. (C) Fo−Fc density map (contoured at 3.00σ) around D8 (cyan mesh). (D) Overlay of D8/SARS-CoV-2 Mpro (PDB code 8I4S, cyan) with 12/SARS-CoV-2 Mpro (PDB code 7LMF20, pink) complexes, highlighting the residues Gln189 and Met49. (E) Overlay of D8/SARS-CoV-2 Mpro (cyan) with baicalein/SARS-CoV-2 Mpro (PDB code 6M2N22, blue) complexes, highlighting the residues Gln189 and Met49.
Fig. 3
Although highly conserved among variants of SARS-CoV-2, some mutated amino acid residues in Mpro have been observed, such as K90R in Beta strain B.1.351, K90R and A193V in Beta B.1.351.2, L205V in Zeta P.2, and P132H in Omicron B.1.529 [36]. However, it is worth noting that these amino residues are more than 10 Å away from the binding site of D8 (Fig. S4), suggesting that D8 should be have inhibitory effects against Mpro from SARS-CoV-2 variants. The assays assessing the inhibitory activity of the inhibitors against Mpro from SARS-CoV-2 Omicron variant are currently in progress.
3 Conclusions
Mpro has been validated as an effective target for development of orally available small molecule anti-COVID19 drugs. In this study, we sought to use medicinal chemistry and rational drug design approaches to structurally modify the first reported nonpeptidic, noncovalent SARS-CoV-2 Mpro inhibitor, baicalein. These efforts led to a series of quinazolin-4-one-derived noncovalent inhibitors with nanomolar potencies against SARS-CoV-2 Mpro. In particular, an optimized compound, C7 exhibited superior inhibitory potency against Mpro relative to baicalein and is endowed with improved physicochemical and DMPK properties. Significantly, C7 also showed more potent antiviral activity than baicalein (EC50 = 1.1 and 5.15 μM, respectively) in SARS-CoV-2-infected Vero E6 cells. Moreover, C7 exhibited relatively high selectivity over a panel of human proteases (IC50 ＞ 10 μM) and low cytotoxicity against Vero E6 cells (CC50 ＞ 50 μM). The co-crystal structure of another potent inhibitor D8 complexed with Mpro showed that the inhibitor noncovalently binds to the active site of Mpro, and occupies a newly formed S2c pocket that is not observed with baicalein and most other Mpro inhibitors, and can be further exploited for inhibitor design. Meanwhile, the S1 and S4 subpocket remain largely unoccupied by D8 and its analogues, leaving room for further improvement. While having improvements, the inhibitors still lack sufficient properties necessary to be profiled for in vivo antiviral efficacy in SARS-CoV-2-infected animal models. Further structural optimizations of this series of inhibitors are ongoing, and are focused on improving the DMPK properties, as well as further improving biochemical and cellular potencies. In addition, profiling the potency against Mpro from SARS-CoV-2 Omicron variant and other coronavirus is underway in our laboratory. Collectively, compound C7 represents a promising lead for further development of more effective Mpro inhibitors and antiviral drugs against SARS-CoV-2 infection.
4 Experimental
4.1 Chemistry
Reagents and solvents from commercial sources were used without further purification. The progress of all reactions was monitored by TLC using EtOAc/petroleum ether (PE) or dichloromethane (DCM)/MeOH as the solvent system, and spots were visualized by irradiation with UV light (254 nm) or by staining with phosphomolybdic acid. Flash chromatography was performed using silica gel (200−300 mesh). 1H NMR and 13C NMR spectra were recorded on a Bruker Avance ARX-400, a Bruker Avance ARX-500 or a Bruker Avance ARX-600. Chemical shifts δ are reported in ppm, and multiplicity of signals are denoted as: br = broad, s = singlet, d = doublet, t = triplet, q = quartet and m = multiplet. The low resolution ESI-MS was recorded on Shimadzu GCMS-2010 instruments and the high resolution mass spectra (HRMS) on a Water Q-Tofmicro mass spectrometer. Anhydrous DCM and N,N-dimethylformamide (DMF) were freshly distilled from calcium hydride. Anhydrous tetrahydrofuran (THF) was freshly distilled over sodium using benzophenone as the indicator. All other solvents were reagent grade. All moisture sensitive reactions were carried out in flame dried flasks under an argon atmosphere. The chemical purity of the target compounds was analyzed by high performance liquid chromatography (HPLC) on an InertSustain C18 column (4.6 mm × 250 mm, 5 μm) under gradient 60–100% MeOH in water (with 0.1% TFA in each mobile phase), with a flow rate of 1.0 mL/min and peak detection at 254 nm. All the target compounds showed purity greater than 95%. The synthesis and compound information of the intermediates can be found in the Supporting Information.
4.1.1 5,6,7-Trihydroxy-2-phenylquinolin-4(1H)-one(18)
Step1: To the stirred solution of compound 20 (1.83 g, 10 mmol) in dry THF (25 mL) was added TEA (2 mL, 15 mmol) and benzoyl chloride (1.8 mL, 15 mmol) at 0 °C. Then the reaction mixture was stirred at room temperature (rt) for 2 h. The resulting mixture was concentrated and partitioned between water and DCM. The combined organic layers were washed with brine, dried over anhydrous Na2SO4, filtered, and concentrated under the reduced pressure. The resulting residue triturated with n-hexane, filtered, washed with n-hexane, dried in vacuum to afford 21 as a white solid (2.61g, 91%), which was used directly in the next step without further purification.
Step2: To the stirred solution of compound 21 in anhydrous DCM (30 mL) was successively dropwise added acetyl chloride (0.78 mL, 11 mmol) and SnCl4 (2 mL, 20 mmol) at 0 °C. The resulting reaction mixture was stirred at 0 °C for 4 h and then moved to room temperature for 1 h, after which the reaction mixture was poured into ice water. The aqueous portion was extracted with DCM (3 × 20 mL), and the combined organics were washed with brine, dried over anhydrous Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by silica column chromatography and eluted with petroleum ether: EtOAc (20:1) to give compound 22 as a white solid (1.48 g, 45%). 1H NMR (400 MHz, CDCl3) δ 12.37 (s, 1H), 8.42 (s, 1H), 8.05–7.99 (m, 2H), 7.59–7.47 (m, 3H), 4.00 (s, 6H), 3.85 (s, 3H), 2.67 (s, 3H). m/z (ESI-MS): 330.2 [M + H ]+.
Step3: To the stirred solution of compound 22 (1.32 g, 4.01 mmol) in anhydrous 1,4-dioxane (40 mL) was added NaOH (480 mg, 12 mmol), and the reaction mixture was heated to reflux under a nitrogen atmosphere for 4 h. Then, the reaction mixture was cooled to room temperature, and the solvent was removed under reduced pressure. Next, the residue was treated with water and n-hexane and the resulting mixture was sonicated for approximately 2 min. The resulting suspension was adjusted to pH∼7 with 1 M HCl and filtered. The precipitate obtained was washed with n-hexane and dried under vacuum to afford compound 23 as a white solid (435 mg, 35%). 1H NMR (400 MHz, CDCl3) δ 8.07–8.01 (m, 2H), 7.96 (s, 1H), 7.54–7.45 (m, 3H), 7.02 (s, 1H), 4.21 (s, 3H), 4.05 (s, 3H), 3.95 (s, 3H). m/z (ESI-MS): 312.2 [M + H ]+.
Step4: To the stirred suspension of compound 23 (200 mg, 0.64 mmol) in anhydrous DCM (1.5 mL) was dropwise added BBr3/DCM (9.6 mL, 9.6 mmol, 1 M) at −10 °C under a nitrogen atmosphere. The resulting reaction mixture was stirred at −10 °C overnight and then moved to room temperature for 12 h. Then, the mixture was moved to −10 °C again, and was quenched by slowly adding ice methanol, after which the solvent was removed under reduced pressure. The resulting residue was triturated with water, filtered, washed with water and DCM, dried in vacuum to afford 18 as a pale yellow solid (130 mg, 75%). 1H NMR (400 MHz, DMSO‑d 6) δ 14.02 (s, 1H), 11.26 (s, 1H), 9.31 (s, 1H), 8.04–7.95 (m, 2H), 7.67–7.57 (m, 3H), 7.32 (s, 1H), 7.03 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.7, 156.1, 152.6, 139.2, 134.4, 132.2, 131.8, 131.5, 129.1, 128.5, 105.3, 101.1, 94.6. HRMS (ESI): m/z calcd for C15H10NO4 [M − H]–: 268.0615, found 268.0610. HPLC analysis: tR = 10.216 min, 96.3%.
4.1.2 General procedure a for the preparation of A1–A6, A8 – A12, A14, A16, A18, A19, A21 – A27 and 19
Compound 24 (337 mg, 1 mmol) was dissolved in anhydrous DMF (6 mL), and compounds 25a–25w (1.5 mmol), CsCO3 (652 mg, 2 mmol) and CuI (39 mg, 0.2 mmol) were added. The resulting mixture was stirred at room temperature under nitrogen atmosphere overnight. After completion of the reaction, excess saturated aqueous NH4Cl was added and filtered. The precipitate obtained was washed with a large amount of water, dried in vacuum to afford trimethoxyquinazolinones 26a–26w.
To the stirred suspension of the obtained trimethoxyquinazolinone (0.5 mmol) in anhydrous DCM (1.5 mL) was dropwise added BBr3/DCM (5 mL, 6 mmol, 1 M) at −10 °C under a nitrogen atmosphere. The resulting reaction mixture was stirred at −10 °C overnight and then moved to room temperature for 12–24 h. Then, the mixture was moved to −10 °C again, and was quenched by slowly adding ice methanol, after which the solvent was removed under the reduced pressure. The resulting residue was triturated with water, filtered, washed with water and DCM, dried in vacuum to afford target compounds A1–A6, A8–A12, A14, A16, A18, A19, A21–A27 and 19.
4.1.3 5,6,7-Trihydroxy-2-phenylquinazolin-4(3H)-one (19)
The product was obtained as a offwhite solid (91 mg), yield 67%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.46 (s, 1H), 11.76 (s, 1H), 10.33 (s, 1H), 8.91 (s, 1H), 8.10 (d, J = 6.7 Hz, 3H), 7.59–7.49 (m, 3H), 6.65 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.4, 154.0, 149.4, 146.8, 142.1, 132.8, 131.1, 131.1, 128.6, 127.5, 103.2, 100.6. HRMS (ESI): m/z calcd for C14H11N2O4 [M + H]+: 271.0713, found 271.0717. HPLC analysis: tR = 6.765min, 96.9%.
4.1.4 2-(2-Fluorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A1)
According to general procedure A, the product was obtained as a white solid (65 mg), yield 45%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.48 (s, 1H), 11.71 (s, 1H), 10.33 (s, 1H), 8.93 (s, 1H), 7.72 (t, J = 7.4 Hz, 1H), 7.63–7.56 (m, 1H), 7.41–7.31 (m, 2H), 6.63 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.6, 159.5 (d, J = 250.2 Hz), 153.9, 146.8 (d, J = 6.8 Hz), 142.0, 132.6 (d, J = 8.5 Hz), 131.3, 131.0, 124.6 (d, J = 3.7 Hz), 122.2 (d, J = 13.1 Hz), 116.1 (d, J = 21.3 Hz), 103.2, 100.6. HRMS (ESI): m/z calcd for C14H10FN2O4 [M + H]+: 289.0619, found 289.0623. HPLC analysis: tR = 6.825 min, 95.3%.
4.1.5 2-(3-Fluorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A2)
According to general procedure A, the product was obtained as a white solid (68 mg), yield 47%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.51 (br, 1H), 11.77 (s, 1H), 10.29 (s, 1H), 8.97 (s, 1H), 8.04–7.87 (m, 1H), 7.63–7.52 (m, 1H), 7.41 (td, J = 8.5, 2.6 Hz, 1H), 6.67 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.6, 162.1 (d, J = 243.5 Hz), 154.0, 148.4, 146.8, 141.4, 135.0 (d, J = 8.1 Hz), 131.4, 130.7 (d, J = 8.4 Hz), 123.7 (d, J = 2.7 Hz), 118.0 (d, J = 21.0 Hz), 114.3 (d, J = 23.8 Hz), 103.0, 101.0. HRMS (ESI): m/z calcd for C14H10FN2O4 [M + H]+: 289.0619, found 289.0622. HPLC analysis: tR = 7.435 min, 95.5%.
4.1.6 2-(4-Fluorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A3)
According to general procedure A, the product was obtained as a white solid (56 mg), yield 39%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.80 (s, 1H), 10.33 (s, 1H), 8.89 (br, 1H), 8.25–8.08 (m, 2H), 7.44–7.29 (m, 2H), 6.64 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.7, 163.8 (d, J = 247.5 Hz), 153.9, 148.9, 146.8, 141.5, 131.0, 130.1 (d, J = 8.9 Hz), 129.2 (d, J = 2.9 Hz), 115.6 (d, J = 21.9 Hz), 102.7, 100.4. HRMS (ESI): m/z calcd for C14H10FN2O4 [M + H]+: 289.0619, found 289.0617. HPLC analysis: tR = 8.683 min, 98.9%.
4.1.7 2-(2-Chlorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one(A4)
According to general procedure A, the product was obtained as a white solid (90 mg), yield 59%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.56 (s, 1H), 11.73 (s, 1H), 10.35 (s, 1H), 8.93 (s, 1H), 7.65–7.52 (m, 3H), 7.48 (td, J = 7.4, 1.5 Hz, 1H), 6.60 (s, 1H). 13C NMR (150 MHz, DMSO‑d 6) δ 165.5, 153.8, 149.1, 146.8, 141.7, 133.6, 131.6, 131.5, 131.2, 131.0, 129. 6, 127.2, 102.9, 100.7. HRMS (ESI): m/z calcd for C14H10ClN2O4 [M + H]+: 305.0324, found 305.0323. HPLC analysis: tR = 6.920 min, 95.2%.
4.1.8 2-(3-Chlorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A5)
According to general procedure A, the product was obtained as a white solid (94 mg), yield 62%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.50 (br, 1H), 11.73 (s, 1H), 10.35 (br, 1H), 8.98 (br, 1H), 8.16 (t, J = 1.9 Hz, 1H), 8.07 (d, J = 7.9 Hz, 1H), 7.69–7.48 (m, 2H), 6.67 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.5, 153.9, 148.2, 146.8, 141.7, 134.8, 133.4, 131.4, 130.8, 130.5, 127.3, 126.2, 103.4, 100.7. HRMS (ESI): m/z calcd for C14H10ClN2O4 [M + H]+: 305.0324, found 305.0323. HPLC analysis: tR = 12.106 min, 96.2%.
4.1.9 2-(4-Chlorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A6)
According to general procedure A, the product was obtained as a white solid (97 mg), yield 64%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.51 (s, 1H), 11.72 (s, 1H), 10.36 (s, 1H), 8.94 (s, 1H), 8.15–8.09 (m, 2H), 7.62–7.56 (m, 2H), 6.64 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.3, 154.0, 148.4, 146.8, 141.9, 135.9, 131.6, 131.3, 129.3, 128.7, 103.3, 100.6. HRMS (ESI): m/z calcd for C14H10ClN2O4 [M + H]+: 305.0324, found 305.0317. HPLC analysis: tR = 9.457 min, 98.0%.
4.1.10 2-(3-Bromophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A8)
According to general procedure A, the product was obtained as a white solid (104 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.54 (br, 1H), 11.78 (s, 1H), 10.37 (s, 1H), 8.96 (s, 1H), 8.29 (s, 1H), 8.11 (d, J = 7.9 Hz, 1H), 7.76 (d, J = 7.8 Hz, 1H), 7.49 (t, J = 7.9 Hz, 1H), 6.66 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.6, 153.9, 148.3, 146.8, 141.3, 134.9, 133.7, 131.3, 130.7, 130.1, 126.5, 121.9, 102.9, 100.6. HRMS (ESI): m/z calcd for C14H10 79BrN2O4 [M + H]+: 348.9819, found 348.9823. HPLC analysis: tR = 10.440 min, 95.4%.
4.1.11 2-(4-Bromophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A9)
According to general procedure A, the product was obtained as a white solid (108 mg), yield 62%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.51 (s, 1H), 11.77 (s, 1H), 10.35 (s, 1H), 8.94 (s, 1H), 8.09–8.02 (m, 2H), 7.780–7.71 (m, 2H), 6.65 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.6, 153.9, 148.8, 146. 8, 141.5, 131.9, 131.6, 131.2, 129.5, 124.8, 102.8, 100.6. HRMS (ESI): m/z calcd for C14H8 79BrN2O4 [M − H]–: 346.9673, found 346.9666. HPLC analysis: tR = 11.866 min, 96.6%.
4.1.12 5,6,7-Trihydroxy-2-(o-tolyl)quinazolin-4(3H)-one (A10)
According to general procedure A, the product was obtained as a white solid (50 mg), yield 35%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.37 (s, 1H), 11.78 (s, 1H), 10.28 (s, 1H), 8.87 (s, 1H), 7.47–7.38 (m, 2H), 7.35–7.28 (m, 2H), 6.58 (s, 1H), 2.35 (s, 3H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.3, 153.8, 151.7, 146.8, 141.3, 136.2, 134.0, 131.0, 130.5, 129.8, 129.2, 125.7, 102.8, 100.5, 19.6. HRMS (ESI): m/z calcd for C15H11N2O4 [M − H] –: 283.0724, found 283.0718. HPLC analysis: tR = 9.143 min, 99.2%.
4.1.13 5,6,7-Trihydroxy-2-(m-tolyl)quinazolin-4(3H)-one (A11)
According to general procedure A, the product was obtained as a white solid (57 mg), yield 40%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.84 (br, 1H), 10.36 (s, 1H), 8.90 (br, 1H), 7.93 (s, 1H), 7.88 (d, J = 7.5 Hz, 1H), 7.45–7.34 (m, 2H), 6.65 (s, 1H), 2.39 (s, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.8, 153.9, 150.1, 146.8, 141.5, 137.9, 132.5, 131.8, 131.0, 128.5, 128.1, 124.7, 102.6, 100.5, 21.0. HRMS (ESI): m/z calcd for C15H11N2O4 [M − H] –: 283.0724, found 283.0722. HPLC analysis: tR = 7.412 min, 95.1%.
4.1.14 5,6,7-Trihydroxy-2-(p-tolyl)quinazolin-4(3H)-one (A12)
According to general procedure A, the product was obtained as a white solid (61 mg), yield 43%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.38 (br, 1H), 11.82 (s, 1H), 10.31 (s, 1H), 8.87 (s, 1H), 8.05–7.98 (m, 2H), 7.36–7.30 (m, 2H), 6.62 (s, 1H), 2.38 (s, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.8, 154.4, 150.1, 147.3, 141.6, 131.3, 130.3, 129.6, 127.9, 102.7, 100.9, 21.42. HRMS (ESI): m/z calcd for C15H11N2O4 [M − H] –: 283.0724, found 283.0717. HPLC analysis: tR = 5.061 min, 97.0%.
4.1.15 5,6,7-Trihydroxy-2-(3-(trifluoromethyl)phenyl)quinazolin-4(3H)-one (A14)
According to general procedure A, the product was obtained as a white solid (68 mg), yield 40%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.68 (s, 1H), 11.72 (s, 1H), 10.38 (s, 1H), 8.99 (s, 1H), 8.46 (s, 1H), 8.41 (d, J = 8.1 Hz, 1H), 7.93 (d, J = 7.7 Hz, 1H), 7.77 (t, J = 7.8 Hz, 1H), 6.68 (s, 1H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.3, 154.0, 148.0, 146.7, 141.8, 133.7, 131.53, 131.45, 129.8, 129.4 (q, J = 32.0 Hz), 127.5, 124.2, 124.0 (q, J = 272.3 Hz), 103.4, 100.7. HRMS (ESI): m/z calcd for C15H8F3N2O4 [M − H] –: 337.0442, found 337.0436. HPLC analysis: tR = 8.083 min, 97.2%.
4.1.16 5,6,7-Trihydroxy-2-(4-(trifluoromethoxy)phenyl)quinazolin-4(3H)-one (A16)
According to general procedure A, the product was obtained as a yellow solid (76 mg), yield 43%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.78 (br, 1H), 10.35 (br, 1H), 8.27–8.18 (m, 2H), 7.56–7.49 (m, 2H), 6.66 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.5, 158.2, 153.9, 150.2, 148.6, 146.8, 141.4, 131.8, 131.2, 129.8, 128.3, 120.8, 120.0 (q, J = 257.4 Hz), 102.7, 100.5. HRMS (ESI): m/z calcd for C15H8F3N2O5 [M − H] –: 353.0391, found 353.0381. HPLC analysis: tR = 8.742 min, 96.3%.
4.1.17 5,6,7-Trihydroxy-2-(3-nitrophenyl)quinazolin-4(3H)-one (A18)
According to general procedure A, the product was obtained as a yellow solid (71 mg), yield 45%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.76 (s, 1H), 11.71 (s, 1H), 10.42 (s, 1H), 9.02 (s, 1H), 8.94 (s, 1H), 8.53 (d, J = 7.8 Hz, 1H), 8.38 (dd, J = 8.2, 2.2 Hz, 1H), 7.81 (t, J = 8.0 Hz, 1H), 6.69 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 167.0, 154.4, 148.4, 147.8, 147.2, 141.9, 134.7, 134.2, 132.0, 130.7, 125.9, 122.8, 103.9, 101.2. HRMS (ESI): m/z calcd for C14H8N3O6 [M − H] –: 314.0419, found 314.0412. HPLC analysis: tR = 11.254 min, 97.5%.
4.1.18 5,6,7-Trihydroxy-2-(4-nitrophenyl)quinazolin-4(3H)-one (A19)
According to general procedure A, the product was obtained as a yellow solid (61 mg), yield 39%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.74 (s, 1H), 11.70 (s, 1H), 10.44 (s, 1H), 9.07 (s, 1H), 8.41–8.30 (s, 4H), 6.70 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.1, 153.9, 148.7, 147.6, 146.7, 141.6, 138.5, 131.8, 128.9, 123.6, 103.6, 100.7. HRMS (ESI): m/z calcd for C14H8N3O6 [M − H] –: 314.0419, found 314.0412. HPLC analysis: tR = 8.392 min, 97.7%.
4.1.19 2-(3-(tert-Butyl)phenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A21)
According to general procedure A, the product was obtained as a white solid (68 mg), yield 42%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.53 (br, 1H), 11.84 (s, 1H), 10.37 (s, 1H), 8.89 (s, 1H), 8.10 (s, 1H), 7.92 (d, J = 8.0 Hz, 1H), 7.58 (d, J = 7.9 Hz, 1H), 7.45 (t, J = 7.8 Hz, 1H), 6.66 (s, 1H), 1.34 (s, 9H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.9, 153.9, 151.1, 150.0, 146.8, 141.4, 132.4, 131.0, 128.4, 128.1, 125.0, 124.3, 102.5, 100.5, 34.8, 31.1.
HRMS (ESI): m/z calcd for C18H17N2O4 [M − H] –: 325.1194, found 325.1189. HPLC analysis: tR = 12.850 min, 95.2%.
4.1.20 2-(4-(tert-Butyl)phenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A22)
According to general procedure A, the product was obtained as a white solid (67 mg), yield 41%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.38 (s, 1H), 11.80 (s, 1H), 10.29 (s, 1H), 8.85 (s, 1H), 8.10–8.01 (m, 2H), 7.56–7.51 (m, 2H), 6.63 (s, 1H), 1.32 (s, 9H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.3, 153.9, 149.6, 146.8, 142.2, 130. 9, 130.0, 128.6, 127.3, 125.4, 102.9, 100.5, 34.7, 30.9. HRMS (ESI): m/z calcd for C18H17N2O4 [M − H] –: 325.1194, found 325.1185. HPLC analysis: tR = 10.744 min, 97.0%.
4.1.21 5,6,7-Trihydroxy-2-(2-hydroxyphenyl)quinazolin-4(3H)-one (A23)
According to general procedure A, the product was obtained as a white solid (50 mg), yield 35%. 1H NMR (400 MHz, DMSO‑d 6) δ 13.71 (br, 1H), 12.42 (br, 1H), 11.60 (s, 1H), 10.48 (s, 1H), 9.03 (s, 1H), 8.15 (dd, J = 8.1, 1.6 Hz, 1H), 7.47–7.37 (m, 1H), 7.01–6.92 (m, 2H), 6.63 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.4, 159.8, 154.2, 151.1, 147.0, 139.0, 133.4, 131.4, 127.4, 118.8, 117.8, 113.8, 101.8, 100.4. HRMS (ESI): m/z calcd for C14H9N2O5 [M − H] –: 285.0517, found 285.0509. HPLC analysis: tR = 9.255 min, 95.4%.
4.1.22 5,6,7-Trihydroxy-2-(4-hydroxyphenyl)quinazolin-4(3H)-one (A24)
According to general procedure A, the product was obtained as a white solid (54 mg), yield 38%. 1H NMR (300 MHz, DMSO‑d 6) δ 12.27 (br, 1H), 11.85 (s, 1H), 10.27 (s, 1H), 10.13 (s, 1H), 8.81 (s, 1H), 8.06–7.93 (m, 2H), 6.93–6.82 (m, 2H), 6.58 (s, 1H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.4, 160.3, 154.0, 149.2, 146.8, 142.4, 130.5, 129.3, 123.3, 115.3, 102.5, 100.2. HRMS (ESI): m/z calcd for C14H9N2O5 [M − H] –: 285.0517, found 285.0514. HPLC analysis: tR = 10.262 min, 96.1%.
4.1.23 2-(2,3-Difluorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A25)
According to general procedure A, the product was obtained as a white solid (77 mg), yield 50%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.60 (s, 1H), 11.67 (s, 1H), 10.40 (s, 1H), 9.01 (s, 1H), 7.69–7.60 (m, 1H), 7.58–7.51 (m, 1H), 7.40–7.32 (m, 1H), 6.64 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.0, 154.35, 151.17 (d, J = 12.5 Hz), 149.17 (dd, J = 13.2, 9.5 Hz), 147.20, 146.07 (d, J = 3.5 Hz), 142.26, 132.04, 126.60 (d, J = 3.5 Hz), 125.88–125.24 (m), 124.68 (d, J = 9.6 Hz), 119.93 (d, J = 17.3 Hz), 103.9, 101.1. HRMS (ESI): m/z calcd for C14H7F2N2O4 [M − H] –: 305.0379, found 305.0369. HPLC analysis: tR = 8.571 min, 96.1%.
4.1.24 2-(2,4-Difluorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A26)
According to general procedure A, the product was obtained as a white solid (84 mg), yield 55%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.52 (s, 1H), 11.69 (s, 1H), 10.36 (s, 1H), 8.97 (s, 1H), 7.83–7.76 (m, 1H), 7.50–7.42 (m, 1H), 7.25 (td, J = 8.5, 2.5 Hz, 1H), 6.62 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.6, 163.5 (dd, J = 248.8, 11.3Hz), 160.0 (J = 253.2, 12.6 Hz), 153.9, 146.7, 146.0, 141.9, 132.6 (dd, J = 10.6, 3.7 Hz), 131.4, 119.0 (dd, J = 13.1, 3.5 Hz), 111.9 (dd, J = 22.1, 3.1 Hz), 104.6 (t, J = 26.0 Hz), 103.2, 100.6. HRMS (ESI): m/z calcd for C14H7F2N2O4 [M − H] –: 305.0379, found 305.0369. HPLC analysis: tR = 7.082 min, 95.4%.
4.1.25 2-(3,4-Difluorophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A27)
According to general procedure A, the product was obtained as a white solid (77 mg), yield 50%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.53 (s, 1H), 11.69 (s, 1H), 10.38 (s, 1H), 8.97 (s, 1H), 8.20–8.13 (m, 1H), 8.04–7.97 (m, 1H), 7.67–7.57 (m, 1H), 6.65 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.6 (d, J = 141.3 Hz), 153.9, 152.2 (d, J = 13.4 Hz), 150.2 (t, J = 12.7 Hz), 148.3 (d, J = 12.9 Hz), 147.3, 146.7, 141.7, 131.32, 130.2 (dd, J = 6.2, 3.4 Hz), 124.84, 117.79 (d, J = 17.7 Hz), 116.76 (d, J = 19.3 Hz), 103.3, 100.5. HRMS (ESI): m/z calcd for C14H7F2N2O4 [M − H] –: 305.0379, found 305.0371. HPLC analysis: tR = 7.863 min, 95.4%.
4.1.26 General procedure B for the preparation of compounds A7, A13, A15, A17, A20, B1–B17, C1–C15 and D6–D9
To the stirred solution of 27 or 30a–30p (1 mmol) in anhydrous EtOH (10 mL) was added I2 (1.2 mmol) and the corresponding aldehyde (1.2 mmol). The resulting reaction mixture was heated to reflux and stirred for 3–5 h. After completion of the reaction, the reaction mixture was cooled to room temperature and was quenched with excess 5% aqueous Na2S2O3. The aqueous portion was extracted with DCM (3 × 10 mL), and the combined organic layers were washed with brine, dried over anhydrous Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by silica column chromatography (DCM/MeOH = 300:1 to 150:1) to give the corresponding quinazolinones 29a–29v, 32a–32o, and 41a–41d.
According to general procedure A, the demethylation of the above quinazolinones afforded the target compounds A7, A13, A15, A17, A20, B1–B17, C1–C15 and D6–D9.
4.1.27 2-(2-Bromophenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (A7)
According to general procedure B, the product was obtained as a white solid (99 mg), yield 57%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.55 (s, 1H), 11.76 (s, 1H), 10.36 (s, 1H), 8.94 (s, 1H), 7.76 (dd, J = 7.7, 1.5 Hz, 1H), 7.60 (dd, J = 7.4, 2.1 Hz, 1H), 7.56–7.42 (m, 2H), 6.60 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.5, 153.8, 150.2, 146.8, 141.8, 135.7, 132.6, 131.6, 131.3, 130.9, 127.6, 121.1, 103.2, 100.7. HRMS (ESI): m/z calcd for C14H10 79BrN2O4 [M + H]+: 348.9819, found 348.9822. HPLC analysis: tR = 9.626 min, 98.2%.
4.1.28 5,6,7-Trihydroxy-2-(2-(trifluoromethyl)phenyl)quinazolin-4(3H)-one (A13)
According to general procedure B, the product was obtained as a yellow solid (71 mg), yield 42%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.62 (s, 1H), 11.70 (s, 1H), 10.35 (s, 1H), 8.94 (s, 1H), 7.94–7.85 (m, 1H), 7.84–7.69 (m, 3H), 6.57 (s, 1H). 13C NMR (150 MHz, DMSO‑d 6) δ 165.5, 153.9, 149.3, 147.0, 141.6, 132.9, 132.4, 131.3, 130.8, 130.4, 127.09 (q, J = 30.8 Hz), 126.41 (q, J = 4.7 Hz), 123.76 (q, J = 273 Hz), 103.0, 100.6. HRMS (ESI): m/z calcd for C15H8F3N2O4 [M − H] –: 337.0442, found 337.0431. HPLC analysis: tR = 7.223 min, 96.4%.
4.1.29 5,6,7-Trihydroxy-2-(2-(trifluoromethoxy)phenyl)quinazolin-4(3H)-one (A15)
According to general procedure B, the product was obtained as a white solid (71 mg), yield 40%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.58 (s, 1H), 11.70 (s, 1H), 10.36 (s, 1H), 8.97 (s, 1H), 7.76 (dd, J = 7.6, 1.8 Hz, 1H), 7.68 (td, J = 7.8, 1.8 Hz, 1H), 7.58–7.50 (m, 2H), 6.62 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.6, 153.9, 147.2, 146.7, 145.7, 141.9, 132.1, 131.4, 128.0, 127.6, 121.5, 120.0 (q, J = 257.4 Hz), 103.3, 100.5. HRMS (ESI): m/z calcd for C15H8F3N2O5 [M − H] –: 353.0391, found 353.0377. HPLC analysis: tR = 9.658 min, 95.0%.
4.1.30 5,6,7-Trihydroxy-2-(2-nitrophenyl)quinazolin-4(3H)-one (A17)
According to general procedure B, the product was obtained as an orange solid (68 mg), yield 43%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.76 (s, 1H), 11.68 (s, 1H), 10.37 (s, 1H), 8.97 (s, 1H), 8.20–8.16 (m, 1H), 7.94–7.76 (m, 3H), 6.54 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.7, 153.9, 148.4, 147.6, 146.8, 141.9, 133.8, 131.6, 131.5, 131.4, 129.0, 124.5, 103.2, 100.6. HRMS (ESI): m/z calcd for C14H8N3O6 [M − H] –: 314.0419, found 314.0411. HPLC analysis: tR = 6.699 min, 95.1%.
4.1.31 5,6,7-Trihydroxy-2-(3-isopropylphenyl)quinazolin-4(3H)-one (A20)
According to general procedure B, the product was obtained as a white solid (72 mg), yield 46%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.46 (s, 1H), 11.83 (s, 1H), 10.30 (s, 1H), 8.87 (s, 1H), 7.98 (s, 1H), 7.94–7.89 (m, 1H), 7.44 (d, J = 4.8 Hz, 2H), 6.64 (s, 1H), 2.99 (hept, J = 6.8 Hz, 1H), 1.27 (s, 3H), 1.26 (s, 3H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.6, 153.9, 150.0, 148.8, 146.8, 141.8, 132.6, 131.0, 129.2, 128.6, 125.4, 125.2, 102.8, 100.5, 33.5, 23.8. HRMS (ESI): m/z calcd for C17H15N2O4 [M − H] –: 311.1037, found 311.1033. HPLC analysis: tR = 13.162 min, 97.7%.
4.1.32 5,6,7-Trihydroxy-2-(thiophen-2-yl)quinazolin-4(3H)-one (B1)
According to general procedure B, the product was obtained as a offwhite solid (83 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.59 (br, 1H), 11.75 (s, 1H), 10.33 (s, 1H), 8.90 (br, 1H), 8.16–8.11 (m, 1H), 7.85–7.78 (m, 1H), 7.24–7.17 (m, 1H), 6.54 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.3, 154.0, 147.00, 145.40, 141.5, 137.5, 131.5, 131.0, 128.7, 128.5, 102.4, 100.4. HRMS (ESI): m/z calcd for C12H7N2O4S [M − H] –: 275.0123, found 275.0132. HPLC analysis: tR = 7.024 min, 96.7%.
4.1.33 5,6,7-Trihydroxy-2-(pyridin-2-yl)quinazolin-4(3H)-one (B2)
According to general procedure B, the product was obtained as an orange solid (61 mg), yield 45%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.92 (s, 1H), 10.45 (s, 1H), 9.04 (s, 1H), 8.73 (d, J = 4.6 Hz, 1H), 8.36 (d, J = 7.8 Hz, 1H), 8.04 (t, J = 7.8 Hz, 1H), 7.62 (t, J = 6.1 Hz, 1H), 6.74 (s, 1H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.1, 153.8, 148.9, 148.8, 148.0, 147.0, 140.7, 137.9, 131.6, 126.3, 122.0, 102.8, 101.3. HRMS (ESI): m/z calcd for C13H8N3O4 [M − H] –: 270.0520, found 270.0520. HPLC analysis: tR = 5.840 min, 95.0%.
4.1.34 5,6,7-Trihydroxy-2-(pyridin-4-yl)quinazolin-4(3H)-one (B3)
According to general procedure B, the product was obtained as an yellow solid (68 mg), yield 50%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.68 (s, 1H), 11.73 (s, 1H), 10.44 (s, 1H), 9.07 (s, 1H), 8.82–8.73 (m, 2H), 8.12–8.05 (m, 2H), 6.70 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.9, 153.9, 149.9, 149.8, 147.5, 146.7, 146.0, 141.3, 140.2, 131.8, 121.5, 103.6, 101.0. HRMS (ESI): m/z calcd for C13H8N3O4 [M − H] –: 270.0520, found 270.0519. HPLC analysis: tR = 5.971 min, 96.7%.
4.1.35 2-(1,3-Dimethyl-1H-pyrazol-5-yl)-5,6,7-trihydroxyquinazolin-4(3H)-one (B4)
According to general procedure B, the product was obtained as a yellow solid (75 mg), yield 52%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.34 (br, 1H), 11.73 (s, 1H), 10.38 (s, 1H), 8.97 (br, 1H), 6.89 (s, 1H), 6.62 (s, 1H), 4.10 (s, 3H), 2.18 (s, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.2, 153.9, 146.9, 145.7, 142.4, 141.1, 134.7, 131.4, 107.8, 103.0, 100.4, 39.6, 13.2. HRMS (ESI): m/z calcd for C13H11N4O4 [M − H] –: 287.0786, found 287.0777. HPLC analysis: tR = 7.618 min, 96.5%.
4.1.36 5,6,7-Trihydroxy-2-(naphthalen-2-yl)quinazolin-4(3H)-one (B5)
According to general procedure B, the product was obtained as a white solid (99 mg), yield 62%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.84 (s, 1H), 10.37 (s, 1H), 8.73 (s, 1H), 8.23 (dd, J = 8.6, 1.9 Hz, 1H), 8.08–7.97 (m, 3H), 7.67–7.56 (m, 2H), 6.70 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.6, 154.0, 150.0, 146.9, 141.2, 134.0, 132.3, 131.3, 129.7, 128.9, 128.2, 127.9, 127.8, 127.7, 127.0, 124.4, 102.5, 100.6. HRMS (ESI): m/z calcd for C18H11N2O4 [M − H] –: 319.0724, found 319.0721. HPLC analysis: tR = 8.708 min, 97.4%.
4.1.37 2-Cyclopropyl-5,6,7-trihydroxyquinazolin-4(3H)-one (B6)
According to general procedure B, the product was obtained as a white solid (49 mg), yield 42%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.34 (s, 1H), 11.71 (s, 1H), 10.15 (s, 1H), 8.65 (s, 1H), 6.35 (s, 1H), 1.92–1.84 (m, 1H), 1.06–0.93 (m, 4H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.0, 156.3, 153.8, 146.9, 141.8, 129.8, 102.0, 100.2, 13.4, 9.2. HRMS (ESI): m/z calcd for C11H9N2O4 [M − H] –: 233.0568, found 233.0560. HPLC analysis: tR = 5.988 min, 95.3%.
4.1.38 2-Cyclopentyl-5,6,7-trihydroxyquinazolin-4(3H)-one (B7)
According to general procedure B, the product was obtained as a white solid (49 mg), yield 48%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.06 (s, 1H), 11.75 (s, 1H), 10.19 (s, 1H), 8.71 (s, 1H), 6.46 (s, 1H), 3.00–2.90 (m, 1H), 1.99–1.89 (m, 2H), 1.88–1.77 (m, 2H), 1.76–1.65 (m, 2H), 1.63–1.52 (m, 2H). 13C NMR (150 MHz, DMSO‑d 6) δ 165.9, 157.4, 153.7, 146.8, 142.2, 130.2, 102.6, 100.4, 43.7, 30.8, 25.3. HRMS (ESI): m/z calcd for C13H13N2O4 [M − H] –: 261.0881, found 261.0875. HPLC analysis: tR = 8.081min, 95.7%.
4.1.39 2-Cyclohexyl-5,6,7-trihydroxyquinazolin-4(3H)-one (B8)
According to general procedure B, the product was obtained as a white solid (49 mg), yield 53%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.01 (s, 1H), 11.75 (s, 1H), 10.19 (s, 1H), 8.71 (s, 1H), 6.47 (s, 1H), 2.53–2.45 (m, 1H), 1.91–1.82 (m, 2H), 1.80–1.72 (m, 2H), 1.66 (d, J = 10.6 Hz, 1H), 1.58–1.45 (m, 2H), 1.35–1.13 (m, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.8, 157.6, 153.7, 146.8, 142.3, 130.3, 102.1, 100.4, 42.6, 30.2, 25.5, 25.3. HRMS (ESI): m/z calcd for C14H15N2O4 [M − H] –: 275.1037, found 275.1034. HPLC analysis: tR = 9.657 min, 96.3%.
4.1.40 5,6,7-Trihydroxy-2-isopropylquinazolin-4(3H)-one (B9)
According to general procedure B, the product was obtained as a white solid (71 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.03 (s, 1H), 11.73 (s, 1H), 10.16 (s, 1H), 8.69 (s, 1H), 6.48 (s, 1H), 2.81 (hept, J = 6.8 Hz, 1H), 1.21 (d, J = 6.9 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.1, 158.7, 153.7, 146.9, 142.0, 130.3, 102.4, 100.4, 33.1, 20.4. HRMS (ESI): m/z calcd for C11H11N2O4 [M − H] –: 235. 0724, found 235.0721. HPLC analysis: tR = 5.456 min, 96.1%.
4.1.41 2-(tert-Butyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (B10)
According to general procedure B, the product was obtained as a white solid (70 mg), yield 56%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.79 (s, 1H), 10.22 (s, 1H), 8.75 (s, 1H), 6.51 (s, 1H), 1.29 (s, 9H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.5, 160.1, 153.8, 146.7, 141.4, 130.5, 102.5, 100.1, 37.1, 27.8. HRMS (ESI): m/z calcd for C12H13N2O4 [M − H] –: 249.0881, found 249.0873. HPLC analysis: tR = 7.203 min, 95.2%.
4.1.42 5,6,7-Trihydroxy-2-isobutylquinazolin-4(3H)-one (B11)
According to general procedure B, the product was obtained as a white solid (58 mg), yield 56%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.08 (s, 1H), 11.70 (s, 1H), 10.21 (s, 1H), 8.74 (s, 1H), 6.48 (s, 1H), 2.39 (d, J = 7.3 Hz, 2H), 2.11 (hept, J = 6.7 Hz, 1H), 0.90 (d, J = 6.6 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.1, 153.7, 146.8, 142.1, 130.3, 102.4, 100.4, 43.1, 27.0, 22.1. HRMS (ESI): m/z calcd for C12H13N2O4 [M − H] –: 249.0881, found 249.0879. HPLC analysis: tR = 9.580 min, 98.8%.
4.1.43 5,6,7-Trihydroxy-2-neopentylquinazolin-4(3H)-one (B12)
According to general procedure B, the product was obtained as a white solid (79 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.99 (s, 1H), 11.71 (s, 1H), 10.21 (s, 1H), 8.74 (s, 1H), 6.49 (s, 1H), 2.41 (s, 2H), 0.97 (s, 9H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.7, 153.7, 152.4, 146.8, 142.1, 130.3, 102.2, 100.1, 47.0, 31.9, 29.4. HRMS (ESI): m/z calcd for C13H15N2O4 [M − H] –: 263.1037, found 263.1030. HPLC analysis: tR = 7.704 min, 95.1%.
4.1.44 (E)-5,6,7-Trihydroxy-2-styrylquinazolin-4(3H)-one (B13)
According to general procedure B, the product was obtained as a white solid (93 mg), yield 63%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.27 (s, 1H), 11.78 (s, 1H), 10.28 (s, 1H), 8.89 (s, 1H), 7.84 (d, J = 16.2 Hz, 1H), 7.69–7.60 (m, 2H), 7.51–7.37 (m, 3H), 6.93 (d, J = 16.2 Hz, 1H), 6.58 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.6, 153.8, 148.4, 146.8, 142.1, 137.2, 135.1, 130.9, 129.5, 129.0, 127.5, 121.0, 102.8, 100.6. HRMS (ESI): m/z calcd for C16H11N2O4 [M − H] –: 295.0724, found 295.0720. HPLC analysis: tR = 9.694 min, 95.3%.
4.1.45 2-Benzyl-5,6,7-trihydroxyquinazolin-4(3H)-one (B14)
According to general procedure B, the product was obtained as a white solid (85 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.34 (s, 1H), 11.68 (s, 1H), 10.23 (s, 1H), 8.77 (s, 1H), 7.38–7.28 (m, 4H), 7.27–7.21 (m, 1H), 6.48 (s, 1H), 3.86 (s, 2H). 13C NMR (150 MHz, DMSO‑d 6) δ 166.1, 153.9, 153.6, 146.8, 141.1, 136.4, 130.7, 128.8, 128.5, 126.9, 101.6, 100.2, 40.3. HRMS (ESI): m/z calcd for C15H11N2O4 [M − H] –: 283.0724, found 283.0716. HPLC analysis: tR = 8.088 min, 95.0%.
4.1.46 5,6,7-Trihydroxy-2-(1-phenylethyl)quinazolin-4(3H)-one (B15)
According to general procedure B, the product was obtained as a white solid (85 mg), yield 57%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.16 (s, 1H), 11.65 (s, 1H), 10.25 (s, 1H), 8.76 (s, 1H), 7.40–7.27 (m, 4H), 7.26–7.18 (m, 1H), 6.54 (s, 1H), 4.03 (q, J = 7.0 Hz, 1H), 1.55 (d, J = 7.0 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.8, 155.9, 153.8, 146.8, 142.4, 142.0, 130.5, 128.5, 127.4, 126.8, 102.8, 100.4, 43.6, 19.3. HRMS (ESI): m/z calcd for C16H13N2O4 [M − H] –: 297. 0881, found 297. 0871. HPLC analysis: tR = 7.318 min, 97.7%.
4.1.47 5,6,7-Trihydroxy-2-(1-phenylpropyl)quinazolin-4(3H)-one (B16)
According to general procedure B, the product was obtained as a white solid (92 mg), yield 59%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.69 (br, 1H), 10.27 (br, 1H), 7.42–7.35 (m, 2H), 7.34–7.29 (m, 2H), 7.26–7.20 (m, 1H), 6.55 (s, 1H), 3.74 (t, J = 7.7 Hz, 1H), 2.27–2.15 (m, 1H), 1.96–1.84 (m, 1H), 0.83 (t, J = 7.3 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.1, 156.2, 154.5, 153.9, 146.8, 140.7, 130.7, 128.5, 127.9, 127.1, 101.8100.4, 51.2, 26.3, 12.2. HRMS (ESI): m/z calcd for C17H15N2O4 [M − H] –: 311.1037, found 311.1028. HPLC analysis: tR = 7.867 min, 95.1%.
4.1.48 5,6,7-Trihydroxy-2-(2-methyl-1-phenylpropyl)quinazolin-4(3H)-one (B17)
According to general procedure B, the product was obtained as a white solid (103 mg), yield 63%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.23 (s, 1H), 11.61 (s, 1H), 10.24 (s, 1H), 8.76 (s, 1H), 7.44 (d, J = 7.1 Hz, 2H), 7.31 (t, J = 7.5 Hz, 2H), 7.25–7.18 (m, 1H), 6.55 (s, 1H), 3.40 (d, J = 11.3 Hz, 1H), 2.67–2.53 (m, 1H), 0.93 (d, J = 6.4 Hz, 3H), 0.69 (d, J = 6.6 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.7, 155.4, 153.8, 146.7, 142.3, 140.3, 130.4, 128.33, 128.28, 127.0, 102.7, 100.2, 57.8, 30.9, 21.2, 20.5. HRMS (ESI): m/z calcd for C18H18N2O4Na [M + Na]+: 349. 1159, found 349. 1158. HPLC analysis: tR = 9.494 min, 96.0%.
4.1.49 5,6,7-Trihydroxy-3-isobutyl-2-phenylquinazolin-4(3H)-one (C1)
According to general procedure B, the product was obtained as a light yellow solid (77 mg), yield 47%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.83 (s, 1H), 10.37 (s, 1H), 8.95 (s, 1H), 7.61–7.56 (m, 2H), 7.55–7.47 (m, 3H), 6.57 (s, 1H), 3.83 (d, J = 7.4 Hz, 2H), 1.73 (hept, J = 6.9 Hz, 1H), 0.61 (d, J = 6.7 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.2, 153.9, 153.0, 146.4, 140.3, 135.2, 131.4, 129.5, 128.5, 128.3, 102.8, 100.4, 50.6, 27.3, 19.7. HRMS (ESI): m/z calcd for C18H19N2O4 [M + H] +: 327.1339, found 327.1342. HPLC analysis: tR = 14.369 min, 96.7%.
4.1.50 5,6,7-Trihydroxy-3-(2-hydroxyethyl)-2-phenylquinazolin-4(3H)-one (C2)
According to general procedure B, the product was obtained as a white solid (71 mg), yield 45%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.80 (s, 1H), 10.33 (s, 1H), 8.93 (s, 1H), 7.62–7.55 (m, 2H), 7.54–7.47 (m, 3H), 6.55 (s, 1H), 4.83 (t, J = 5.7 Hz, 1H), 3.94 (t, J = 6.2 Hz, 2H), 3.48 (q, J = 5.9 Hz, 2H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.1, 153.9, 153.4, 146.4, 140.5, 135.4, 131.3, 129.3, 128.6, 128.3, 102.6, 100.7, 57.8, 46.9. HRMS (ESI): m/z calcd for C16H15N2O5 [M + H] +: 315. 0976, found 315. 0973. HPLC analysis: tR = 11.691 min, 98.1%.
4.1.51 N-(tert-butyl)-2-(5,6,7-trihydroxy-4-oxo-2-phenylquinazolin-3(4H)-yl)acetamide (C3)
According to general procedure B, the product was obtained as a white solid (82 mg), yield 43%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.57 (s, 1H), 10.39 (s, 1H), 8.96 (br, 1H), 7.67 (s, 1H), 7.58–7.43 (m, 5H), 6.58 (s, 1H), 4.36 (s, 2H), 1.17 (s, 9H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.4, 164.7, 154.0, 153.1, 146.3, 140.4, 134.8, 131.4, 129.6, 128.3, 128.1, 102.8, 100.2, 50.3, 47.1, 28.3. HRMS (ESI): m/z calcd for C20H21N3O5Na [M + Na] +: 406. 1373, found 406.1370. HPLC analysis: tR = 12.294 min, 96.8%.
4.1.52 3-Cyclopropyl-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one (C4)
According to general procedure B, the product was obtained as a white solid (71 mg), yield 46%.1H NMR (400 MHz, DMSO‑d 6) δ 11.88 (s, 1H), 10.26 (s, 1H), 8.88 (s, 1H), 7.84–7.66 (m, 2H), 7.56–7.41 (m, 3H), 6.55 (s, 1H), 3.17 (tt, J = 7.3, 4.1 Hz, 1H), 0.87–0.65 (m, 2H), 0.57–0.30 (m, 2H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.7, 153.9, 153.6, 146.4, 140.1, 136.0, 131.2, 129.3, 128.4, 127.9, 102.7, 100.4, 29.0, 10.7. HRMS (ESI): m/z calcd for C17H15N2O4 [M + H] +: 311. 1026, found 311. 1023. HPLC analysis: tR = 12.671 min, 98.4%.
4.1.53 3-Cyclopentyl-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one (C5)
According to general procedure B, the product was obtained as a white solid (85 mg), yield 50%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.90 (s, 1H), 10.30 (s, 1H), 8.91 (s, 1H), 7.70–7.38 (m, 5H), 6.53 (s, 1H), 4.35–4.25 (m, 1H), 2.34–2.21 (m, 2H), 1.96–1.82 (m, 2H), 1.79–1.65 (m, 2H), 1.46–1.34 (m, 2H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.4, 153.8, 153.7, 146.5, 140.1, 135.9, 131.3, 129.5, 128.6, 127.8, 102.4, 101.4, 61.0, 28.7, 25.4. HRMS (ESI): m/z calcd for C19H19N2O4 [M + H] +: 339.1339, found 339.1338. HPLC analysis: tR = 11.188 min, 96.5%.
4.1.54 3-Cyclohexyl-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one (C6)
According to general procedure B, the product was obtained as a white solid (97 mg), yield 55%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.98 (s, 1H), 10.30 (s, 1H), 8.91 (s, 1H), 7.60–7.48 (m, 5H), 6.52 (s, 1H), 3.71 (t, J = 12.1 Hz, 1H), 2.58–2.43 (m, 2H), 1.81–1.62 (m, 4H), 1.54–1.42 (m, 1H), 1.16–0.98 (m, 1H), 0.91–0.73 (m, 2H). 13C NMR (125 MHz, DMSO‑d 6) δ 166.4, 154.2, 153.9, 147.0, 140.5, 136.4, 131.7, 129.9, 129.0, 127.9, 102.9, 101.9, 62.0, 28.8, 26.3, 25.1. HRMS (ESI): m/z calcd for C20H21N2O4 [M + H] +: 353.1496, found 353.1503. HPLC analysis: tR = 8.000 min, 98.4%.
4.1.55 5,6,7-Trihydroxy-2,3-diphenylquinazolin-4(3H)-one (C7)
According to general procedure B, the product was obtained as a white solid (99 mg), yield 57%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.60 (s, 1H), 10.42 (s, 1H), 9.00 (s, 1H), 7.37–7.15 (m, 10H), 6.67 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.0, 154.1, 152.1, 146.8, 140.7, 136.9, 135.3, 131.5, 129.6, 129.0, 128.8, 128.6, 128.3, 127.5, 103.2, 100.4. HRMS (ESI): m/z calcd for C20H15N2O4 [M + H] +: 347.1026, found 347.1025. HPLC analysis: tR = 8.939 min, 98.9%.
4.1.56 3-(2-Fluorophenyl)-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one (C8)
According to general procedure B, the product was obtained as a white solid (73 mg), yield 40%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.32 (s, 1H), 10.50 (s, 1H), 9.07 (s, 1H), 7.51 (td, J = 7.8, 1.7 Hz, 1H), 7.43–7.20 (m, 7H), 7.16 (td, J = 7.7, 1.4 Hz, 1H), 6.70 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 164.3, 157.0 (d, J = 248.1 Hz), 154.3, 151.7, 146.8, 140.4, 134.5, 131.7 (d, J = 24.6 Hz) 131.2 (d, J = 8.2 Hz), 129.3, 128.3, 127.7, 124.7 (d, J = 3.4 Hz), 124.5, 124.4, 115.8 (d, J = 19.6 Hz), 103.6, 99.9. HRMS (ESI): m/z calcd for C20H14FN2O4 [M + H] +: 365.0932, found 365.0931. HPLC analysis: tR = 7.934 min, 98.3%.
4.1.57 3-(3-Fluorophenyl)-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one (C9)
According to general procedure B, the product was obtained as a white solid (73 mg), yield 40%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.50 (s, 1H), 10.44 (s, 1H), 9.03 (s, 1H), 7.45–7.07 (m, 9H), 6.67 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 164.8, 161.6 (d, J = 244.4 Hz), 154.1, 151.8, 146.8, 140.5, 138.4 (d, J = 10.6 Hz), 135.1, 131.6, 130.1 (d, J = 9.0 Hz), 128.9, 127.6, 126.1 (d, J = 3.1 Hz), 117.2 (d, J = 23.7 Hz), 115.4 (d, J = 20.8 Hz), 103.3, 100.3. HRMS (ESI): m/z calcd for C20H14FN2O4 [M + H] +: 365.0932, found 365.0930. HPLC analysis: tR = 7.986 min, 96.7%.
4.1.58 5,6,7-Trihydroxy-3-(4-hydroxyphenyl)-2-phenylquinazolin-4(3H)-one (C10)
According to general procedure B, the product was obtained as a white solid (94 mg), yield 52%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.69 (s, 1H), 10.37 (s, 1H), 9.62 (s, 1H), 8.97 (s, 1H), 7.34–7.29 (m, 2H), 7.27–7.19 (m, 3H), 7.10–7.05 (m, 2H), 6.65–6.60 (m, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.3, 157.0, 154.0, 152.6, 146.8, 140.7, 135.6, 131.4, 130.5, 129.0, 128.7, 127.9, 127.5, 115.1, 103.1, 100.5. HRMS (ESI): m/z calcd for C20H15N2O5 [M + H] +: 363.0976, found 363.0984. HPLC analysis: tR = 7.195 min, 98.2%.
4.1.59 3-(4-Bromo-2-fluorophenyl)-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one(C11)
According to general procedure B, the product was obtained as a white solid (93 mg), yield 42%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.23 (s, 1H), 10.57 (s, 1H), 9.13 (br, 1H), 7.66 (dd, J = 9.4, 2.1 Hz, 1H), 7.53 (t, J = 8.2 Hz, 1H), 7.42 (dd, J = 8.6, 2.1 Hz, 1H), 7.39–7.25 (m, 5H), 6.70 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 164.1, 156.9 (d, J = 253.0 Hz), 154.4, 151.4, 146.8, 140.3, 134.3, 133.2, 131.9, 129.5, 128.3, 128.0 (d, J = 3.4 Hz), 127.9, 124.2 (d, J = 13.1 Hz), 122.7 (d, J = 9.0 Hz), 119.4 (d, J = 23.2 Hz), 103.7, 99.8. HRMS (ESI): m/z calcd for C20H11 79BrFN2O4 [M − H]–: 440.9892, found 440.9878. HPLC analysis: tR = 8.670 min, 97.1%.
4.1.60 3-(4-Fluoro-3-methylphenyl)-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one (C12)
According to general procedure B, the product was obtained as a white solid (108 mg), yield 57%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.57 (s, 1H), 10.42 (br, 1H), 9.01 (br, 1H), 7.38–7.29 (m, 3H), 7.31–7.13 (m, 4H), 7.06 (t, J = 9.1 Hz, 1H), 6.66 (s, 1H), 2.12 (d, J = 1.9 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.0, 159.8 (d, J = 244.8 Hz), 154.1, 152.0, 146.8, 140.5, 135.2, 132.69 (d, J = 3.4 Hz), 132.66, 132.61, 131.5, 128.9, 128.8 (d, J = 9.2 Hz), 127.5, 124.5 (d, J = 18.8 Hz), 115.0 (d, J = 23.7 Hz), 103.2, 100.3, 14.0 (d, J = 3.1 Hz). HRMS (ESI): m/z calcd for C21H16FN2O4 [M + H] +: 379. 1089, found 379.1089. HPLC analysis: tR = 8.177 min, 96.6%.
4.1.61 3-([1,1′-biphenyl]-4-yl)-5,6,7-trihydroxy-2-phenylquinazolin-4(3H)-one (C13)
According to general procedure B, the product was obtained as a white solid (74 mg), yield 35%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.61 (s, 1H), 10.44 (s, 1H), 9.02 (s, 1H), 7.67–7.60 (m, 4H), 7.50–7.33 (m, 7H), 7.28–7.19 (m, 3H), 6.68 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.1, 154.1, 152.0, 146.8, 140.7, 139.7, 138.8, 136.2, 135.3, 131.6, 130.1, 129.1, 129.0, 128.9, 127.8, 127.6, 126.7, 126.7, 126.6, 103.3, 100.4. HRMS (ESI): m/z calcd for C26H19N2O4 [M + H] +: 423. 1340, found 423. 1345. HPLC analysis: tR = 10.271 min, 97.1%.
4.1.62 5,6,7-Trihydroxy-2-phenyl-3-(pyridin-3-yl)quinazolin-4(3H)-one (C14)
According to general procedure B, the product was obtained as a white solid (74 mg), yield 50%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.42 (s, 1H), 10.49 (s, 1H), 9.06 (s, 1H), 8.52 (d, J = 2.4 Hz, 1H), 8.43 (dd, J = 4.8, 1.5 Hz, 1H), 7.88–7.83 (m, 1H), 7.44–7.30 (m, 3H), 7.29–7.19 (m, 3H), 6.69 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 164.9, 154.2, 151.7, 150.0, 149.0, 146.8, 140.6, 137.2, 134.9, 133.9, 131.7, 129.1, 129.0, 127.7, 123.5, 103.4, 100.2. HRMS (ESI): m/z calcd for C19H14N3O4 [M + H] +: 348.0979, found 348.0974. HPLC analysis: tR = 6.881min, 97.9%.
4.1.63 5,6,7-Trihydroxy-2-phenyl-3-(1H-pyrazol-4-yl)quinazolin-4(3H)-one (C15)
According to general procedure B, the product was obtained as a white solid (82 mg), yield 49%. 1H NMR (400 MHz, DMSO‑d 6) δ 12.86 (br, 1H), 11.62 (s, 1H), 10.44 (s, 1H), 9.00 (s, 1H), 7.69 (br, 1H), 7.42–7.35 (m, 2H), 7.32–7.23 (m, 3H), 6.64 (s, 1H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.1, 154.1, 152.9, 146.8, 140.4, 135.4, 131.5, 128.9, 128.9, 127.6, 117.8, 103.2, 100.2. HRMS (ESI): m/z calcd for C17H11N4O4 [M − H] –: 335.0786, found 335.0779. HPLC analysis: tR = 8.686 min, 95.4%.
4.1.64 5,6,7-Trihydroxy-2-isobutyl-3-phenylquinazolin-4(3H)-one (D6)
According to general procedure B, the product was obtained as a white solid (98 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.53 (s, 1H), 10.27 (s, 1H), 8.82 (s, 1H), 7.62–7.48 (m, 3H), 7.47–7.39 (m, 2H), 6.57 (s, 1H), 2.16 (d, J = 6.9 Hz, 2H), 2.08–1.96 (m, 1H), 0.78 (d, J = 6.6 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.1, 153.9, 152.7, 146.7, 140.6, 136.4, 130.8, 129.4, 129.0, 128.9, 102.6, 100.1, 43.2, 25.8, 22.2. HRMS (ESI): m/z calcd for C18H19N2O4 [M + H]+: 327.1339, found 327.1345. HPLC analysis: tR = 8.174 min, 99.8%.
4.1.65 3-(2-Fluorophenyl)-5,6,7-trihydroxy-2-isobutylquinazolin-4(3H)-one (D7)
According to general procedure B, the product was obtained as a white solid (107 mg), yield 62%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.26 (s, 1H), 10.43 (br, 1H), 7.71–7.58 (m, 2H), 7.54–7.46 (m, 1H), 7.42 (t, J = 7.4 Hz, 1H), 6.60 (s, 1H), 2.29–2.12 (m, 2H), 1.99 (hept, J = 6.7 Hz, 1H), 0.80 (dd, J = 8.2, 6.6 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 163.9, 157.3 (d, J = 248.5 Hz), 154.4, 153.1, 146.7, 145.9, 139.3, 132.0 (d, J = 7.9 Hz), 131.3 (d, J = 43.4 Hz), 125.5 (d, J = 3.6 Hz), 123.4 (d, J = 13.4 Hz), 116.5 (d, J = 19.4 Hz), 102.3, 99.6, 42.7, 25.9, 22.1 (d, J = 7.7 Hz). HRMS (ESI): m/z calcd for C18H17FN2O4 [M − H]–: 343.1100, found 343.1090. HPLC analysis: tR = 9.123 min, 96.3%.
4.1.66 3-(4-Fluoro-3-methylphenyl)-5,6,7-trihydroxy-2-isobutylquinazolin-4(3H)-one (D8)
According to general procedure B, the product was obtained as a white solid (107 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.51 (s, 1H), 10.31 (s, 1H), 8.84 (s, 1H), 7.42–7.37 (m, 1H), 7.34–7.29 (m, 2H), 6.56 (s, 1H), 2.28 (d, J = 2.1 Hz, 3H), 2.17 (d, J = 6.3 Hz, 2H), 2.11–1.98 (m, 1H), 0.81 (dd, J = 6.6, 1.9 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.1, 160.5 (d, J = 245.1 Hz), 154.0, 152.7, 146.7, 140.6, 132.2 (d, J = 3.2 Hz), 132.0 (d, J = 5.8 Hz), 130.9, 128.3 (d, J = 8.8 Hz), 125.7 (d, J = 18.7 Hz), 115.9 (d, J = 23.7 Hz), 102.6, 100.1, 43.1, 25.8, 22.3 (d, J = 5.4 Hz), 14.1 (d, J = 2.9 Hz). HRMS (ESI): m/z calcd for C19H18FN2O4 [M − H]–: 357.1256, found 357.1252. HPLC analysis: tR = 10.674 min, 95.2%.
4.1.67 3-(4-Bromo-2-fluorophenyl)-5,6,7-trihydroxy-2-isobutylquinazolin-4(3H)-one (D9)
According to general procedure B, the product was obtained as a white solid (116 mg), yield 55%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.17 (s, 1H), 10.43 (br, 1H), 7.94–7.87 (m, 1H), 7.70–7.61 (m, 2H), 6.60 (s, 1H), 2.29–2.11 (m, 2H), 2.01 (hept, J = 6.6 Hz, 1H), 0.82 (dd, J = 7.8, 6.6 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 164.1, 157.2 (d, J = 253.0 Hz), 154.3, 152.0, 146.7, 140.2, 132.6, 131.3, 128.8 (d, J = 3.5 Hz), 123.3, 123.2 (d, J = 5.4 Hz), 120.1(d, J = 23.2 Hz), 102.9, 99.5, 42.8, 25.7, 22.1 (d, J = 2.7 Hz). HRMS (ESI): m/z calcd for C18H15 79BrFN2O4 [M − H]–: 421.0205, found 421.0200. HPLC analysis: tR = 10.166 min, 96.6%.
4.1.68 General procedure C for the preparation of D1–D4, D10 and D11
To a solution of compound 33a or 33b (1 mmol) in anhydrous acetonitrile (8 mL) was added the corresponding anilines (1.2 mmol) and PCl3 (0.18 mL, 2 mmol) at 0 °C, and the resulting suspension was heated to 50 °C and stirred for 4–12 h. Then, the reaction was moved to 0 °C and was quenched by adding 1 N aqueous HCl solution. The aqueous portion was extracted with DCM (3 × 20 mL), and the combined organics were washed with 10% aqueous NaHCO3 solution and brine, dried over anhydrous Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by silica column chromatography (petroleum ether/EtOAc = 8:1 to 6:1) to give the corresponding quinazolinones 35a–35f.
According to general procedure A, the demethylation of the above quinazolinones afforded the target compounds D1–D4, D10 and D11.
4.1.69 5,6,7-Trihydroxy-3-phenyl-2-(1-phenylethyl)quinazolin-4(3H)-one (D1)
According to general procedure C, the product was obtained as a white solid (118 mg), yield 63%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.49 (s, 1H), 10.39 (s, 1H), 8.92 (s, 1H), 7.57 (d, J = 4.5 Hz, 2H), 7.49–7.40 (m, 1H), 7.23–7.12 (m, 4H), 6.80 (dd, J = 6.6, 2.9 Hz, 2H), 6.68 (s, 1H), 6.54–6.48 (m, 1H), 3.81 (q, J = 6.8 Hz, 1H), 1.46 (d, J = 6.9 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.2, 155.0, 154.0, 146.7, 142.3, 140.3, 135.8, 131.2, 129.6, 129.4, 128.9, 128.7, 128.4, 127.1, 126.6, 103.1, 100.3, 43.2, 21.8.
HRMS (ESI): m/z calcd for C22H19N2O4 [M + H]+: 375. 1339, found 375. 1345. HPLC analysis: tR = 9.099 min, 99.1%.
4.1.70 3-(4-Fluoro-3-methylphenyl)-5,6,7-trihydroxy-2-(1-phenylethyl)quinazolin-4(3H)-one (D2)
According to general procedure C, the product was obtained as a white solid (106 mg), yield 52%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.46 (s, 1H), 10.40 (s, 1H), 8.92 (s, 1H), 7.54–7.48 (m, 0.4 H), 7.49–7.39 (m, 0.6 H), 7.32 (t, J = 9.0 Hz, 0.6 H), 7.23–7.11 (m, 3H), 6.94 (t, J = 9.1 Hz, 0.4H), 6.88–6.81 (m, 1H), 6.80–6.72 (m, 1H), 6.68 (d, J = 4.5 Hz, 1H), 6.36–6.28 (m, 0.4 H), 6.20–6.14 (m, 0.6H), 3.89–3.74 (m, 1H), 2.31 (s, 1.2 H), 1.91 (s, 1.9 H), 1.45 (dd, J = 6.9, 3.1 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.2, 161.6 (d, J = 5.0 Hz), 159.2 (d, J = 5.0 Hz), 154.9 (d, J = 8.7 Hz), 154.1 (d, J = 1.9 Hz), 146.7, 142.6, 142.4, 140.3 (d, J = 1.2 Hz), 133.1 (d, J = 6.0 Hz), 131.7 (d, J = 5.7 Hz), 131.5 (d, J = 3.3 Hz), 131.4 (d, J = 3.2 Hz), 131.2, 128.9 (d, J = 8.9 Hz), 128.4 (d, J = 4.7 Hz), 128.0 (d, J = 8.8 Hz), 127.1 (d, J = 6.9 Hz), 126.6 (d, J = 7.8 Hz), 125.5 (d, J = 18.7 Hz), 124.7 (d, J = 18.9 Hz), 115.7 (d, J = 23.9 Hz), 115.0 (d, J = 23.6 Hz), 103.2, 103.1, 100.2, 43.7, 43.3, 22.0, 21.9, 14.2 (d, J = 2.9 Hz), 13.8 (d, J = 3.1 Hz). HRMS (ESI): m/z calcd for C23H20FN2O4 [M + H]+: 407.1402, found 407.1395. HPLC analysis: tR = 13.337 min, 96.5%.
4.1.71 3-(2-Fluorophenyl)-5,6,7-trihydroxy-2-(1-phenylethyl)quinazolin-4(3H)-one (D3)
According to general procedure C, the product was obtained as a white solid (110 mg), yield 56%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.28 (s, 1H), 10.45 (s, 1H), 8.96 (s, 1H), 7.81–7.71 (m, 1H), 7.58–7.44 (m, 1H), 7.43–7.35 (m, 1H), 7.26–7.03 (m, 3H), 7.03–6.93 (m, 1H), 6.72 (s, 1H), 6.71–6.60 (m, 2H), 3.95 (q, J = 6.7 Hz, 1H), 1.47 (d, J = 6.8 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 164.4, 157.7 (d, J = 250.6 Hz), 154.5, 154.3, 146.7, 141.3, 140.2, 131.6 (d, J = 8.0 Hz), 131.5, 130.6, 128.5, 128.3, 126.8, 126.7, 125.0 (d, J = 3.5 Hz), 123.3 (d, J = 13.4 Hz), 116.1, 115.9, 103.4, 99.7, 43.5, 21.9. HRMS (ESI): m/z calcd for C22H18FN2O4 [M + H]+: 393. 1245, found 393. 1251. HPLC analysis: tR = 8.657 min, 96.6%.
4.1.72 3-(4-Bromo-2-fluorophenyl)-5,6,7-trihydroxy-2-(1-phenylethyl)quinazolin-4(3H)-one (D4)
According to general procedure C, the product was obtained as a white solid (141 mg), yield 60%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.18 (s, 1H), 10.48 (s, 1H), 8.98 (s, 1H), 7.77 (t, J = 8.2 Hz, 1H), 7.63 (dd, J = 8.5, 2.1 Hz, 1H), 7.32 (dd, J = 9.3, 2.1 Hz, 1H), 7.17–7.09 (m, 3H), 6.74–6.68 (m, 3H), 3.95 (q, J = 6.7 Hz, 1H), 1.47 (d, J = 6.8 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 164.1, 158.6, 156.6, 154.3, 154.1, 146.7, 141.2, 140.0, 132.1, 131.5, 128.4, 128.3 (d, J = 3.5 Hz), 126.8, 126.7, 123.1–122.9 (m), 119.3 (d, J = 23.2 Hz), 103.5, 99.6, 43.6, 21.9. HRMS (ESI): m/z calcd for C22H17 79BrFN2O4 [M + H]+: 471.0350, found 471.0352. HPLC analysis: tR = 13.244 min, 99.4%.
4.1.73 2-Benzyl-5,6,7-trihydroxy-3-phenylquinazolin-4(3H)-one (D10)
According to general procedure C, the product was obtained as a white solid (112 mg), yield 62%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.48 (s, 1H), 10.33 (s, 1H), 8.88 (s, 1H), 7.49–7.38 (m, 3H), 7.23–7.14 (m, 5H), 6.90–6.82 (m, 2H), 6.59 (s, 1H), 3.75 (s, 2H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.1, 154.0, 152.2, 146.7, 140.5, 136.0, 135.8, 131.2, 129.1, 129.0, 128.5, 128.2, 126.6, 102.8, 100.3, 41.3. HRMS (ESI): m/z calcd for C21H17N2O4 [M + H]+: 361.1183, found 361.1190. HPLC analysis: tR = 8.737 min, 95.8%.
4.1.74 2-Benzyl-3-(4-fluoro-3-methylphenyl)-5,6,7-trihydroxyquinazolin-4(3H)-one (D11)
According to general procedure C, the product was obtained as a white solid (88 mg), yield 45%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.44 (s, 1H), 10.35 (s, 1H), 8.89 (s, 1H), 7.23–7.16 (m, 4H), 7.14–7.08 (m, 1H), 6.94 (dd, J = 7.0, 2.6 Hz, 1H), 6.88–6.83 (m, 2H), 6.60 (s, 1H), 3.77 (q, J = 15.5 Hz, 2H), 2.12 (d, J = 1.8 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.6, 160.9 (d, J = 245.1 Hz), 154.5, 152.8, 147.2, 140.9, 136.2, 132.8 (d, J = 5.8 Hz), 132.1 (d, J = 3.2 Hz), 131.7, 128.9, 128.7 (d, J = 9.0 Hz), 128.6, 127.0, 125.6 (d, J = 18.8 Hz), 116.0 (d, J = 23.9 Hz), 103.3, 100.7, 42.0, 14.4 (d, J = 2.9 Hz). HRMS (ESI): m/z calcd for C25H18FN2O4 [M + H]+: 393.1245, found 393.1249. HPLC analysis: tR = 12.068 min, 96.2%.
4.1.75 General procedure D for the preparation of D5 and D12
To the stirred solution of the prepared 2-phenylpropanoyl chloride or commercially available phenylacetyl chloride (1 mmol) in pyridine (5 mL) was added TPP (816 mg, 2.5 mmol) under the N2 atmosphere at 0 °C, and the reaction mixture was heated to 70 °C and stirred for 2 h. Then the reaction mixture was cooled to room temperature, and was treated with isobutylamine (40) (0.12 mL, 1.2 mmol). Then the resulting mixture was heated to 70 °C and stirred overnight. After completion, the reaction mixture was cooled to room temperature, and the solvent was removed under reduced pressure to obtain the residue. The residue was then diluted with DCM (50 mL), washed with 3 N aqueous HCl solution (1 × 20 mL) and brine (3 × 20 mL), and dried anhydrous Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by silica column chromatography (petroleum ether: EtOAc = 6:1) to give the corresponding quinazolinones 37a, 37b.
According to general procedure A, the demethylation of the above quinazolinones afforded the target compounds D5, D12.
4.1.76 5,6,7-Trihydroxy-3-isobutyl-2-(1-phenylethyl)quinazolin-4(3H)-one (D5)
According to general procedure D, the product was obtained as a white solid (97 mg), yield 55%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.75 (s, 1H), 10.30 (s, 1H), 8.83 (s, 1H), 7.36–7.20 (m, 5H), 6.61 (s, 1H), 3.80 (d, J = 15.5 Hz, 1H), 3.73 (d, J = 15.5 Hz, 1H), 4.02–3.93 (m, 1H), 3.50 (dd, J = 14.0, 7.1 Hz, 1H), 2.04 (hept, J = 6.9 Hz, 1H), 1.56 (d, J = 6.7 Hz, 3H), 0.90 (d, J = 6.7 Hz, 3H), 0.82 (d, J = 6.6 Hz, 3H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.3, 155.1, 153.9, 146.3, 142.7, 139.9, 131.1, 128.9, 127.1, 126.9, 102.6, 100.0, 48.0, 42.3, 28.1, 22.5, 19.7, 19.6. HRMS (ESI): m/z calcd for C20H23N2O4 [M + H]+: 355.1652, found 355.1655. HPLC analysis: tR = 11.131 min, 96.2%.
4.1.77 2-Benzyl-5,6,7-trihydroxy-3-isobutylquinazolin-4(3H)-one (D12)
According to general procedure D, the product was obtained as a white solid (95 mg), yield 56%. 1H NMR (400 MHz, DMSO‑d 6) δ 11.74 (s, 1H), 10.30 (s, 1H), 8.84 (s, 1H), 7.36–7.30 (m, 2H), 7.28–7.22 (m, 3H), 6.52 (s, 1H), 4.18 (s, 2H), 3.77 (d, J = 7.5 Hz, 2H), 2.08 (hept, J = 7.0 Hz, 1H), 0.86 (d, J = 6.6 Hz, 6H). 13C NMR (125 MHz, DMSO‑d 6) δ 165.3, 153.9, 152.6, 146.3, 140.3, 136.2, 131.0, 128.7, 128.4, 126.8, 102.3, 100.1, 49.0, 40.6, 27.8, 19.7. HRMS (ESI): m/z calcd for C19H21N2O4 [M + H]+: 341.1496, found 341.1495. HPLC analysis: tR = 11.644 min, 97.8%.
4.2 SARS-CoV-2 Mpro protein expression and purification
The cDNA of SARS-CoV-2 Mpro (GeneBank: YP_009725301.1) was synthesized and cloned into pET-28a (+) vector with codon optimization by Sangon (Sangon Biotech, Shanghai, China). The SARS-CoV-2 Mpro construct contains protease autoprocessing site of Mpro (SAVLQ⬇SGFRK, arrow indicating the cleavage site) and PreScission site (SGVTFQ⬇GP, arrow indicating the cleavage site) in the N terminus and C terminus, respectively. The pET-28a(+)-SARS-CoV-2 M pro plasmid was then transformed into competent E.coli BL21 (DE3) cells and a single colony was picked, which was inoculated in 5 ml LB medium supplemented with 50 g/mL kanamycin and grown at 37 °C to an optical density at 600 nm of 0.8, and then induced using 0.5 mM isopropyl-β-d-thiogalactoside (IPTG). Induced cultures were grown at 25 °C for an additional 16 h. Cells were harvested via centrifugation at 10,000 rpm for 10 min and the precipitate were resuspended in a buffer (20 mM HEPES, pH 7.5, 500 mM NaCl, 20 mM imidazole) and homogenized with an ultrahigh-pressure cell disrupter at 4 °C. The insoluble material was removed by centrifugation at 18,000 rpm for 60 min. The protein was eluted by elution buffer (20 mM HEPES, pH 7.5, 500 mM NaCl, 300 mM imidazole) after incubation with Ni-resin and washed by buffer (20 mM HEPES, pH 7.5, 500 mM NaCl, 20 mM imidazole). After auto-cleaved and cleaved by PreScission protease at 4 °C overnight, native Mpro was generated. In order to obtain homogeneous and pure protein suitable for crystallization and enzymatic inhibition assays, the further purification using Q column was performed with A buffer (20 mM HEPES, pH 7.5, 50 mM NaCl) and B buffer (20 mM HEPES, pH 7.5, 1 M NaCl). Protein sample was concentrated into 10 mg/mL and stored at −80 °C.
4.3 SARS-CoV-2 Mpro enzymatic assays
The enzyme activity of SARS-CoV-2 Mpro was determined using a FRET assay, which was performed as previously reported [37]. The FRET assay uses fluorogenic peptide Dabcyl-KLSAVLQSGFRKM-Edans-NH2 as the substrate, which was custom synthesized and obtained from GL Biochem Ltd (Shanghai, China). In brief, protease SARSCov-2 Mpro (600 nM at a final concentration) was mixed with compounds at the indicated concentration in 50 μL of assay buffer (20 mM Tris-HCl, pH 7.5, 10 mM EDTA, 150 mM NaCl) and incubated at room temperature for 30 min. Then, the reaction was initiated by adding 50 μL of peptide substrate to the mixture (20 μM at a final concentration) and monitored at 340 nM (excitation)/490 nM (emission) with a Tecan's Spark multimode reader. The Vmax of reactions with compounds added at the indicated concentrations compared to the solvent control was calculated and used to generate IC50 curves. For each compound, the half-maximal inhibitory concentration (IC50) values against SARS-CoV-2 Mpro were measured at 10 different concentrations, and three independent experiments were performed. The data were analyzed using GraphPad Prism 8.4.3 software.
4.4 Cellular antiviral assays
A clinical isolate of SARS-CoV-2 (GenBank: MT121215.1) was propagated in Vero E6 cells and the viral titer was determined as 50% tissue culture infectious dose (TCID50) per milliliter (mL) by CPE quantification. All the infection experiments were performed at biosafety level-3 (BLS-3) as previously reported [38]. Pre-seeded Vero E6 cells (5 × 104 cells/well) were incubated with test compounds at the indicated concentrations at 37 °C in 5% CO2 for 1 h. Then, 100 μL of the cell supernatant was removed, followed by addition of the prepared virus-drug mixture (100 μL) to the final virus titer of 100 TCID50/well. After infection for 1 h, the supernatant medium was discarded, and the cells were washed twice with PBS and further cultured with maintenance medium at 37 °C in 5% CO2 for 48 h. At 48 h postinfection, the cell supernatant was collected and lysed using TRIzol LS Reagent (Invitrogen, Carlsbad, USA). Subsequently, RNA was extracted using a PureLink™ RNA Mini Kit (Thermo Fisher, Waltham, USA). The viral RNA copy number was quantified by RT-qPCR using Verso SYBR Green 1-Step qRT-PCR Kit (Thermo Fisher, Waltham, USA) on CFX96™ Real-Time PCR System (Bio-Rad, Hercules, CA) according to the manufacturer's instructions. The PCR primers targeting the N gene (nt608-706) of SARS-CoV-2 are as follows:
5′-GGGGAACTTCTCCTGCTAGAAT-3’;
5′-CAGACATTTTGCTCTCAAGCTG-3’.
4.5 SARS-CoV-2 Mpro protein crystallization and structure determination
8 mg/mL and 12 mg/mL SARS-CoV-2 Mpro in buffer (20 mM Tris-HCl, pH 7.5, 10 mM EDTA, 150 mM NaCl) were incubated with 2 mM compound D8 at 4 °C for 18 h. Then, the mixture was centrifuged at 12,500 rpm for 10 min to remove the precipitate, and the supernatant was used for crystallization. All crystallization experiments were set up at 293 K using the hanging drop approach where 1 μL of each precipitant solution was mixed with 1 μL of protein solution. Crystals were observed after a few days of incubation in 12% PEG 20,000, 0.1 M MES monohydrate (pH 6.5), which were next immersed in the crystallization solution containing 15% glycerol, and were frozen in liquid nitrogen before diffraction. The diffraction data sets were collected at the Shanghai Synchrotron Radiation Facility (SSRF, Shanghai, China) on beamline BL18U1. Using the molecular-replacement method, the structure was resolved with the Phenix program Phaser-MR [39,40], and the electron density could be clearly observed in the protein catalytic site. The eLBOW and LigandFit program in Phenix was used to fit the ligand D8 [41]. Next, adjustment of the structural model was performed in Coot, and structure refinement in Phenix was conducted with NCS restrains [42,43]. Finally, the MolProbity program in Phenix was used to check the quality of the resulting structure [44]. Data collection and structural refinement statistics are depicted in Table S3.
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.
Abbreviations
CCK-8 cell counting kit-8
COVID-19 Coronavirus Disease 2019
DTT dithiothreitol
DMPK drug metabolism and pharmacokinetics
FRET fluorescence resonance energy transfer
HLM human liver microsome
IPTG isopropyl-β-d-thiogalactoside
Mpro main protease
nsp non-structural protein
PBS phosphate buffer solution
PPB plasma protein binding
SAR structure activity relationship
SARS-CoV-2 severe acute respiratory syndrome coronavirus 2
TPP triphenyl phosphite
Appendix A Supplementary data
The following are the Supplementary data to this article:Multimedia component 1
Multimedia component 1
Multimedia component 2
Multimedia component 2
Supporting data to this article can be found online.
Data availability
Data will be made available on request.
Acknowledgements
We thank Biosafety Level 3 Laboratory, Fudan University for providing the experiment platform of cellular antiviral assay. This work was supported by grants from the 10.13039/501100001809 National Natural Science Foundation of China (No. 82161138005, 22107117 and 22001267), 10.13039/501100002858 China Postdoctoral Science Foundation (No. 2021M693516), and 10.13039/501100002949 Postdoctoral Research Program of Jiangsu Province (No. 2021K218B).
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.ejmech.2023.115487.
==== Refs
References
1 WHO, WHO Coronavirus Disease (COVID-19) Dashboard. https://covid19.who.int/(accessed date: 2022.December.14).
2 Ledford H. Hundreds of COVID trials could provide a deluge of new drugs Nature 603 2022 25 27 35233098
3 Li Y. Tenchov R. Smoot J. Liu C. Watkins S. Zhou Q.A. A comprehensive review of the global efforts on COVID-19 vaccine development ACS Cent. Sci. 7 2021 512 533 34056083
4 Lipsitch M. Krammer F. Regev-Yochay G. Lustig Y. Balicer R.D. SARS-CoV-2 breakthrough infections in vaccinated individuals: measurement, causes and impact Nat. Rev. Immunol. 22 2022 57 65 34876702
5 Feikin D.R. Higdon M.M. Abu-Raddad L.J. Andrews N. Araos R. Goldberg Y. Groome M.J. Huppert A. O'Brien K.L. Smith P.G. Wilder-Smith A. Zeger S. Deloria Knoll M. Patel M.K. Duration of effectiveness of vaccines against SARS-CoV-2 infection and COVID-19 disease: results of a systematic review and meta-regression Lancet (London, England) 399 2022 924 944 35202601
6 Ren S.Y. Wang W.B. Gao R.D. Zhou A.M. Omicron variant (B.1.1.529): infectivity, vaccine breakthrough, and antibody resistance J. Chem. Inf. Model. 10 2022 1 11
7 Cele S. Jackson L. Khoury D.S. Khan K. Moyo-Gwete T. Tegally H. San J.E. Cromer D. Scheepers C. Amoako D.G. Karim F. Bernstein M. Lustig G. Archary D. Smith M. Ganga Y. Jule Z. Reedoy K. Hwa S.H. Giandhari J. Blackburn J.M. Gosnell B.I. Abdool Karim S.S. Hanekom W. von Gottberg A. Bhiman J.N. Lessells R.J. Moosa M.S. Davenport M.P. de Oliveira T. Moore P.L. Sigal A. Omicron extensively but incompletely escapes Pfizer BNT162b2 neutralization Nature 602 2022 654 656.8 35016196
8 .
9 Cannalire R. Cerchia C. Beccari A.R. Di Leva F.S. Summa V. Targeting SARS-CoV-2 proteases and polymerase for COVID-19 treatment: state of the art and future opportunities J. Med. Chem. 65 2022 2716 2746 33186044
10 Gao K. Wang R. Chen J. Tepe J.J. Huang F. Wei G.W. Perspectives on SARS-CoV-2 main protease inhibitors J. Med. Chem. 64 2021 16922 16955 34798775
11 Unoh Y. Uehara S. Nakahara K. Nobori H. Yamatsu Y. Yamamoto S. Maruyama Y. Taoda Y. Kasamatsu K. Suto T. Kouki K. Nakahashi A. Kawashima S. Sanaki T. Toba S. Uemura K. Mizutare T. Ando S. Sasaki M. Orba Y. Sawa H. Sato A. Sato T. Kato T. Tachibana Y. Discovery of S-217622, a noncovalent oral SARS-CoV-2 3CL protease inhibitor clinical candidate for treating COVID-19 J. Med. Chem. 65 2022 6499 6512 35352927
12 Drayman N. DeMarco J.K. Jones K.A. Azizi S.A. Froggatt H.M. Tan K. Maltseva N.I. Chen S. Nicolaescu V. Dvorkin S. Furlong K. Kathayat R.S. Firpo M.R. Mastrodomenico V. Bruce E.A. Schmidt M.M. Jedrzejczak R. Muñoz-Alía M. Schuster B. Nair V. Han K.Y. O'Brien A. Tomatsidou A. Meyer B. Vignuzzi M. Missiakas D. Botten J.W. Brooke C.B. Lee H. Baker S.C. Mounce B.C. Heaton N.S. Severson W.E. Palmer K.E. Dickinson B.C. Joachimiak A. Randall G. Tay S. Masitinib is a broad coronavirus 3CL inhibitor that blocks replication of SARS-CoV-2 Science (New York, N.Y.) 373 2021 931 936 34285133
13 Owen D.R. Allerton C.M.N. Anderson A.S. Aschenbrenner L. Avery M. Berritt S. Boras B. Cardin R.D. Carlo A. Coffman K.J. Dantonio A. Di L. Eng H. Ferre R. Gajiwala K.S. Gibson S.A. Greasley S.E. Hurst B.L. Kadar E.P. Kalgutkar A.S. Lee J.C. Lee J. Liu W. Mason S.W. Noell S. Novak J.J. Obach R.S. Ogilvie K. Patel N.C. Pettersson M. Rai D.K. Reese M.R. Sammons M.F. Sathish J.G. Singh R.S.P. Steppan C.M. Stewart A.E. Tuttle J.B. Updyke L. Verhoest P.R. Wei L. Yang Q. Zhu Y. An oral SARS-CoV-2 M(pro) inhibitor clinical candidate for the treatment of COVID-19 Science (New York, N.Y.) 374 2021 1586 1593 34726479
14 Hoffman R.L. Kania R.S. Brothers M.A. Davies J.F. Ferre R.A. Gajiwala K.S. He M. Hogan R.J. Kozminski K. Li L.Y. Lockner J.W. Lou J. Marra M.T. Mitchell L.J. Jr. Murray B.W. Nieman J.A. Noell S. Planken S.P. Rowe T. Ryan K. Smith G.J. 3rd Solowiej J.E. Steppan C.M. Taggart B. Discovery of Ketone-based bovalent inhibitors of coronavirus 3CL proteases for the potential therapeutic treatment of COVID-19 J. Med. Chem. 63 2020 12725 12747 33054210
15 Boras B. Jones R.M. Anson B.J. Arenson D. Aschenbrenner L. Bakowski M.A. Beutler N. Binder J. Chen E. Eng H. Hammond H. Hammond J. Haupt R.E. Hoffman R. Kadar E.P. Kania R. Kimoto E. Kirkpatrick M.G. Lanyon L. Lendy E.K. Lillis J.R. Logue J. Luthra S.A. Ma C. Mason S.W. McGrath M.E. Noell S. Obach R.S. Mn O.B. O'Connor R. Ogilvie K. Owen D. Pettersson M. Reese M.R. Rogers T.F. Rosales R. Rossulek M.I. Sathish J.G. Shirai N. Steppan C. Ticehurst M. Updyke L.W. Weston S. Zhu Y. White K.M. García-Sastre A. Wang J. Chatterjee A.K. Mesecar A.D. Frieman M.B. Anderson A.S. Allerton C. Preclinical characterization of an intravenous coronavirus 3CL protease inhibitor for the potential treatment of COVID19 Nat. Commun. 12 2021 6055 34663813
16 Ma C. Sacco M.D. Hurst B. Townsend J.A. Hu Y. Szeto T. Zhang X. Tarbet B. Marty M.T. Chen Y. Wang J. Boceprevir G.C.- 376, and calpain inhibitors II, XII inhibit SARS-CoV-2 viral replication by targeting the viral main protease Cell Res. 30 2020 678 692 32541865
17 Qiao J. Li Y.S. Zeng R. Liu F.L. Luo R.H. Huang C. Wang Y.F. Zhang J. Quan B. Shen C. Mao X. Liu X. Sun W. Yang W. Ni X. Wang K. Xu L. Duan Z.L. Zou Q.C. Zhang H.L. Qu W. Long Y.H. Li M.H. Yang R.C. Liu X. You J. Zhou Y. Yao R. Li W.P. Liu J.M. Chen P. Liu Y. Lin G.F. Yang X. Zou J. Li L. Hu Y. Lu G.W. Li W.M. Wei Y.Q. Zheng Y.T. Lei J. Yang S. SARS-CoV-2 M(pro) inhibitors with antiviral activity in a transgenic mouse model Science (New York, N.Y.) 371 2021 1374 1378 33602867
18 Dai W. Zhang B. Jiang X.M. Su H. Li J. Zhao Y. Xie X. Jin Z. Peng J. Liu F. Li C. Li Y. Bai F. Wang H. Cheng X. Cen X. Hu S. Yang X. Wang J. Liu X. Xiao G. Jiang H. Rao Z. Zhang L.K. Xu Y. Yang H. Liu H. Structure-based design of antiviral drug candidates targeting the SARS-CoV-2 main protease Science (New York, N.Y.) 368 2020 1331 1335 32321856
19 Zhang L. Lin D. Sun X. Curth U. Drosten C. Sauerhering L. Becker S. Rox K. Hilgenfeld R. Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors Science (New York, N.Y.) 368 2020 409 412 32198291
20 Han S.H. Goins C.M. Arya T. Shin W.J. Maw J. Hooper A. Sonawane D.P. Porter M.R. Bannister B.E. Crouch R.D. Lindsey A.A. Lakatos G. Martinez S.R. Alvarado J. Akers W.S. Wang N.S. Jung J.U. Macdonald J.D. Stauffer S.R. Structure-based optimization of ML300-derived, noncovalent inhibitors targeting the severe acute respiratory syndrome coronavirus 3CL protease (SARS-CoV-2 3CL(pro)) J. Med. Chem. 65 2022 2880 2904 34347470
21 Kneller D.W. Li H. Galanie S. Phillips G. Labbé A. Weiss K.L. Zhang Q. Arnould M.A. Clyde A. Ma H. Ramanathan A. Jonsson C.B. Head M.S. Coates L. Louis J.M. Bonnesen P.V. Kovalevsky A. Structural, electronic, and electrostatic determinants for inhibitor binding to subsites S1 and S2 in SARS-CoV-2 main protease J. Med. Chem. 64 2021 17366 17383 34705466
22 Su H.X. Yao S. Zhao W.F. Li M.J. Liu J. Shang W.J. Xie H. Ke C.Q. Hu H.C. Gao M.N. Yu K.Q. Liu H. Shen J.S. Tang W. Zhang L.K. Xiao G.F. Ni L. Wang D.W. Zuo J.P. Jiang H.L. Bai F. Wu Y. Ye Y. Xu Y.C. Anti-SARS-CoV-2 activities in vitro of Shuanghuanglian preparations and bioactive ingredients Acta Pharmacol. Sin. 41 2020 1167 1177 32737471
23 Kitamura N. Sacco M.D. Ma C. Hu Y. Townsend J.A. Meng X. Zhang F. Zhang X. Ba M. Szeto T. Kukuljac A. Marty M.T. Schultz D. Cherry S. Xiang Y. Chen Y. Wang J. Expedited approach toward the rational design of noncovalent SARS-CoV-2 main protease inhibitors J. Med. Chem. 65 2022 2848 2865 33891389
24 Luttens A. Gullberg H. Abdurakhmanov E. Vo D.D. Akaberi D. Talibov V.O. Nekhotiaeva N. Vangeel L. De Jonghe S. Jochmans D. Krambrich J. Tas A. Lundgren B. Gravenfors Y. Craig A.J. Atilaw Y. Sandström A. Moodie L.W.K. Lundkvist Å. van Hemert M.J. Neyts J. Lennerstrand J. Kihlberg J. Sandberg K. Danielson U.H. Carlsson J. Ultralarge virtual screening identifies SARS-CoV-2 main protease inhibitors with broad-spectrum activity against coronaviruses J. Am. Chem. Soc. 144 2022 2905 2920 35142215
25 Zhang C.H. Stone E.A. Deshmukh M. Ippolito J.A. Ghahremanpour M.M. Tirado-Rives J. Spasov K.A. Zhang S. Takeo Y. Kudalkar S.N. Liang Z. Isaacs F. Lindenbach B. Miller S.J. Anderson K.S. Jorgensen W.L. Potent noncovalent inhibitors of the main protease of SARS-CoV-2 from molecular sculpting of the drug perampanel guided by free energy perturbation calculations ACS Cent. Sci. 7 2021 467 475 33786375
26 Mukae H. Yotsuyanagi H. Ohmagari N. Doi Y. Imamura T. Sonoyama T. Fukuhara T. Ichihashi G. Sanaki T. Baba K. Takeda Y. Tsuge Y. Uehara T. A randomized phase 2/3 study of Ensitrelvir, a novel oral SARS-CoV-2 3C-Like protease inhibitor, in Japanese patients with mild-to-moderate COVID-19 or asymptomatic SARS-CoV-2 infection: results of the phase 2a part Antimicrob. Agents Chemother. 66 2022 e0069722
27 Jones C.P. Anderson K.W. Buchwald S.L. Sequential Cu-catalyzed amidation-base-mediated camps cyclization: a two-step synthesis of 2-aryl-4-quinolones from o-halophenones J. Org. Chem. 72 2007 7968 7973 17850097
28 Hadjeri M. Peiller E.L. Beney C. Deka N. Lawson M.A. Dumontet C. Boumendjel A. Antimitotic activity of 5-hydroxy-7-methoxy-2-phenyl-4-quinolones J. Med. Chem. 47 2004 4964 4970 15369401
29 Carreras J. Gopakumar G. Gu L. Gimeno A. Linowski P. Petuškova J. Thiel W. Alcarazo M. Polycationic ligands in gold catalysis: synthesis and applications of extremely π-acidic catalysts J. Am. Chem. Soc. 135 2013 18815 18823 24308636
30 Liu X. Fu H. Jiang Y. Zhao Y. A simple and efficient approach to quinazolinones under mild copper-catalyzed conditions Angew. Chem., Int. Ed. Engl. 48 2009 348 351 19053126
31 Zhang Y. Jin L. Xiang H. Wu J. Wang P. Hu D. Xue W. Yang S. Synthesis and anticancer activities of 5,6,7-trimethoxy-N-phenyl(ethyl)-4-aminoquinazoline derivatives Eur. J. Med. Chem. 66 2013 335 344 23811258
32 Tian X. Song L. Li E. Wang Q. Yu W. Chang J. Metal-free one-pot synthesis of 1,3-diazaheterocyclic compounds via I2-mediated oxidative C–N bond formation RSC Adv. 5 2015 62194 62201
33 Xue S. McKenna J. Shieh W.C. Repic O. A facile synthesis of C2,N3-disubstituted-4-quinazolone J. Org. Chem. 69 2004 6474 6477 15357612
34 Liu K. Li D. Zheng W. Shi M. Chen Y. Tang M. Yang T. Zhao M. Deng D. Zhang C. Liu J. Yuan X. Yang Z. Chen L. Discovery, optimization, and evaluation of quinazolinone derivatives with novel linkers as orally efficacious phosphoinositide-3-kinase delta inhibitors for treatment of inflammatory diseases J. Med. Chem. 64 2021 8951 8970 34138567
35 Dahlin J.L. Nissink J.W. Strasser J.M. Francis S. Higgins L. Zhou H. Zhang Z. Walters M.A. PAINS in the assay: chemical mechanisms of assay interference and promiscuous enzymatic inhibition observed during a sulfhydryl-scavenging HTS J. Med. Chem. 58 2015 2091 2113 25634295
36 Quan B.X. Shuai H. Xia A.J. Hou Y. Zeng R. Liu X.L. Lin G.F. Qiao J.X. Li W.P. Wang F.L. Wang K. Zhou R.J. Yuen T.T. Chen M.X. Yoon C. Wu M. Zhang S.Y. Huang C. Wang Y.F. Yang W. Tian C. Li W.M. Wei Y.Q. Yuen K.Y. Chan J.F. Lei J. Chu H. Yang S. An orally available M(pro) inhibitor is effective against wild-type SARS-CoV-2 and variants including Omicron Nat. Microbiol. 7 2022 716 725 35477751
37 Yu W. Zhao Y. Ye H. Wu N. Liao Y. Chen N. Li Z. Wan N. Hao H. Yan H. Xiao Y. Lai M. Structure-based design of a dual-targeted covalent inhibitor against papain-like and main proteases of SARS-CoV-2 J. Med. Chem. 65 2022 16252 16267 36503248
38 Ni Y. Liao J. Qian Z. Wu C. Zhang X. Zhang J. Xie Y. Jiang S. Synthesis and evaluation of enantiomers of hydroxychloroquine against SARS-CoV-2 in vitro Bioorg. Med. Chem. 53 2022 116523
39 McCoy A.J. Grosse-Kunstleve R.W. Adams P.D. Winn M.D. Storoni L.C. Read R.J. Phaser crystallographic software J. Appl. Crystallogr. 40 2007 658 674 19461840
40 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. McCoy A.J. Moriarty N.W. Oeffner R. Read R.J. Richardson D.C. Richardson J.S. Terwilliger T.C. Zwart P.H. PHENIX: a comprehensive Python-based system for macromolecular structure solution Acta Crystallogr. D Biol. Crystallogr. 66 2010 213 221 20124702
41 Moriarty N.W. Grosse-Kunstleve R.W. Adams P.D. Electronic ligand Builder and Optimization Workbench (eLBOW): a tool for ligand coordinate and restraint generation Acta Crystallogr. D Biol. Crystallogr. 65 2009 1074 1080 19770504
42 Afonine P.V. Grosse-Kunstleve R.W. Echols N. Headd J.J. Moriarty N.W. Mustyakimov M. Terwilliger T.C. Urzhumtsev A. Zwart P.H. Adams P.D. Towards automated crystallographic structure refinement with phenix refine. Acta Crystallogr. D Biol. Crystallogr. 68 2012 352 367 22505256
43 Emsley P. Lohkamp B. Scott W.G. Cowtan K. Features and development of Coot Acta Crystallogr. D Biol. Crystallogr. 66 2010 486 501 20383002
44 Williams C.J. Headd J.J. Moriarty N.W. Prisant M.G. Videau L.L. Deis L.N. Verma V. Keedy D.A. Hintze B.J. Chen V.B. Jain S. Lewis S.M. Arendall W.B. 3rd Snoeyink J. Adams P.D. Lovell S.C. Richardson J.S. Richardson D.C. MolProbity: more and better reference data for improved all-atom structure validation Protein Sci. 27 2018 293 315 29067766
|
PMC010xxxxxx/PMC10205643.txt |
==== Front
Pediatr Neonatol
Pediatr Neonatol
Pediatrics and Neonatology
1875-9572
2212-1692
Taiwan Pediatric Association. Published by Elsevier Taiwan LLC.
S1875-9572(23)00094-3
10.1016/j.pedneo.2023.01.005
Original Article
BNT162b2 immunization-related myocarditis in adolescents and consequent hospitalization: Report from a medical center
Yen Chen-Wei af
Lee Jung a
Chang Ya-Ting d
Lee En-Pei e
Wu Chang-Teng a
Chang Yi-Jung abc∗
a Division of Pediatric General Medicine, Department of Pediatrics, Chang Gung Memorial Hospital, Taoyuan, Taiwan
b Department of Pediatrics, Chang Gung Memorial Hospital, Chang Gung University College of Medicine, Taoyuan, Taiwan
c Molecular Infectious Disease Research Center, Chang Gung Memorial Hospital, Taoyuan, Taiwan
d Division of Pediatric Cardiology, Lin-Kou Chang Gung Memorial Hospital, Taoyuan, Taiwan
e Division of Pediatric Critical Care Medicine, Department of Pediatrics, Chang Gung Memorial Hospital at Linko, Gweishan, Taoyuan, Taiwan
f Department of Pediatric Nephrology, Lin-Kou Chang Gung Memorial Hospital, Taoyuan, Taiwan
∗ Corresponding author. Division of Pediatric General Medicine, Department of Pediatrics, Chang Gung Memorial Hospital, Chang Gung University, No.5, Fuxing St., Guishan Dist., Taoyuan City 333, Taiwan.
24 5 2023
24 5 2023
16 8 2022
22 11 2022
18 1 2023
© 2023 Taiwan Pediatric Association. Published by Elsevier Taiwan LLC.
2023
Taiwan Pediatric Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
To investigate Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine (BNT162b2) immunization-related myocarditis and describe the risk factors for consequent hospitalization in the pediatric intensive care unit (PICU) in children between 12 and 18 years.
Methods
Children and adolescents 12 years of age and older who presented with discomfort after BNT162b2 immunization (BNTI) and visited pediatric emergency room (PER) at Chang Gung Memorial Hospital from September 22, 2021 to March 21, 2022, were included for analysis.
Results
681 children presented with discomfort after BNTI and visited our PER. The mean age was 15.1 ± 1.7 years. Three hundred and ninety-four (57.9%) and 287 (42.1%) events were after 1st and 2nd dose, respectively. 58.4% (n = 398) were male. The most common complaints were chest pain (46.7%) and chest tightness (27.0%). The median (interquartile range [IQR]) interval of discomfort after BNTI was 3.0 (1.0−12.0) days. BNTI-related pericarditis, myocarditis and myopericarditis were diagnosed in 15 (2.2%), 12 (1.8%) and 2 (0.3%) patients, respectively. Eleven (1.6%) needed hospitalization in PICU. The median (IQR) hospital stay was 4.0 (3.0−6.0) days. There was no mortality. More patients were diagnosed myocarditis (p = 0.004) after 2nd dose BNTI. PICU admission occurred more commonly after 2nd dose BNTI (p = 0.007). Risk factors associated with hospitalization in PICU were abnormal EKG findings (p = 0.047) and abnormal serum troponin levels (p = 0.003) at PER.
Conclusion
Myocarditis in children aged 12−18 years occurred more commonly following 2nd dose BNTI. Most cases were of mild or intermediate severity without death. Factors predicting BNTI-related myocarditis and consequent hospitalization in PICU were abnormal EKG findings and abnormal serum troponin levels at PER in this study.
Key Words
adolescent
myocarditis
pediatric emergency room
pediatric intensive care unit
Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine
==== Body
pmc1 Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), also known as coronavirus disease 2019 (COVID-19), has spread across the globe since 2019, causing many people infection and death.1, 2, 3, 4 In the U.S.A., the American Academy of Pediatrics (AAP) recorded almost 13 million children that tested positive for COVID-19 since the onset of the pandemic according to available state reports.5 AAP claimed any COVID-19 vaccine authorized through Emergency Use Authorization or approved by the US Food and Drug Administration and appropriate by age and health status could be vaccinated according to Centers for Disease Control and Prevention guidelines for children and adolescents in protecting individuals and populations against infectious diseases.6
Though heart complications are not common, occurrence of pericarditis and/or myocarditis in adolescents and young adults after mRNA COVID-19 vaccinations was sporadically reported from April 2021.7, 8, 9, 10, 11, 12
The benefits of COVID-19 vaccination outweigh the reported known and potential risks in the current pandemic.9 , 13 In Taiwan, the Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine (BNT162b2) was approved and utilized in pediatric patients aged 12−18 years old from September 2021.14 Increasing cases of myocarditis were reported as BNT162b2 vaccinations were administered. Here, we aim to investigate factors associated with BNT162b2 immunization (BNTI)-related myocarditis and consequent hospitalization in a pediatric intensive care unit (PICU) in adolescents between 12 and 18 years.
2 Materials and methods
2.1 Participants and study design
This retrospective observational study was conducted at pediatric emergency departments (PERs) in Chang Gung Medical Hospital (CGMH) in Northern Taiwan from September 22, 2021, to March 21, 2022. Medical records for all pediatric patients aged 12−18 years old visiting our PERs were taken in detail. This study was approved by CGMH institutional review board (No. 202200782B0). All methods were carried out and followed in accordance with the approved guidelines and regulations for medical research.
In Taiwan, the Pfizer-BioNTech was approved on August 3, 2021. It was utilized for all children and adolescents 12 years of age and older without contraindications from September 22, 2021, at two doses per person.14 According to the policy from Advisory Committee on Immunization Practices (ACIP) and the Ministry of Health and Welfare in Taiwan, the interval between the doses was at least 12 weeks in adolescents.15 In our enrolled participants, those who suffered from uncomfortable symptoms and signs after BNTI were divided into Group A (after 1st dose BNTI) and Group B (after 2nd dose BNTI). Demographic information was obtained as follows: age, gender, interval of discomfort after BNTI, systolic and diastolic blood pressure (SBP and DBP) at PER, number of patients with abnormal Electrocardiogram (EKG) findings at PER, serum laboratories findings at PER, number of patients diagnosed with pericarditis, myocarditis and myopericarditis,16, 17, 18, 19 number of patients admitted and discharged home, and number of surviving patients. Abnormal EKG findings at PER indicated that previously healthy patients without any heart disease showed ST-segment elevation, incomplete right bundle branch block, T wave inversion, tachy- and bradyarrhythmia, and first-degree AV block in our study population.16, 17, 18, 19
The diagnosis and the definition of COVID-19 vaccine-associated acute pericarditis and acute myocarditis were determined according to the ACIP about COVID-19 vaccine safety updates on June 23, 2021,16, 17, 18, 19 as shown in the Supplementary Figure.
Pediatric patients hospitalized due to BNTI-related acute pericarditis, myocarditis, and myopericarditis16, 17, 18, 19 were divided into Group C (admitted to ward) and Group D (admitted to PICU). The outcome comparison between these two groups was conducted; the difference of ejection fraction on heart echocardiogram, COVID-19 polymerase chain reaction result and hospital length of stay (LOS) were also recorded.
We compared the difference between junior high school age patients (aged 12−15 years) and high school age patients (aged 16−18 years) on clinical characteristics after 1st or 2nd dose of BNTI and hospitalization in the ward or PICU.
2.2 Statistical analysis
Descriptive statistics are presented to represent specific data (e.g., demographics). Univariate summaries are provided for continuous variables (e.g., mean ± standard deviation [SD] for age, SBP and DBP at PER; median and interquartile range [IQR] for serum troponin, CK-MB, NT-ProBNP, brain natriuretic peptide [BNP], C-reactive protein [CRP] level). In contrast, frequencies and percentages summarize categorical variables (e.g., gender, abnormal EKG findings at PER, number of patients who survived). All analyses were performed using SPSS ver. 26.0 (IBM Corp., Armonk, NY, USA), and a p-value <0.05 was taken to indicate statistical significance. The Student's t-test and the χ2 test or Fisher's exact test was used for continuous and categorical variables, respectively. Multivariate logistic regression analyses were performed to determine the predictive factors for pediatric patients who needed admission to PICU. Variables were kept in the final model if the p-value was <0.05.
3 Results
In all, 681 pediatric patients aged 12−18 years old were enrolled in our study. Among them, three hundred ninety-four (57.9%) patients had discomfort after 1st dose of BNTI (Group A) and 287 (42.1%) after 2nd dose of BNTI (Group B). Their chief discomfort complaints were chest pain, chest tightness, palpitation, tachycardia, shortness of breath, dyspnea, dizziness, weakness, abdominal pain, and fever; the proportion of each symptom is shown in Fig. 1 . Fig. 1 also presents the detailed percentage after 1st and 2nd dose of BNTI.Figure 1 Chief discomfort complaints in adolescent visited our PERs after BNTI PERs: pediatric emergency rooms; BNTI: Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine immunization; SOB: shortness of breath, Ohers Ψ: including out-of-hospital cardiac arrest, numbness, tremor, skin rash, back pain, diarrhea, general sore pain, loss of consciousness, vomiting, flank pain, bradycardia and myalgia.
Figure 1
The mean age was 15.1 ± 1.7 years. Three hundred ninety-six (58.1%) were aged 12−15 years and 285 (48.9%) were aged 16−18 years. Over half (58.4%, n = 398) were male children. The median (IQR) interval of discomfort after BNTI was 3.0 (1.0−12.0) days. The mean SBP and DBP at PER were 127.4 ± 18.1 and 74.9 ± 12.1 mmHg, respectively. Eighteen (2.6%) patients had abnormal EKG findings at PER, including 9 with ST-segment elevation, 4 with incomplete right bundle branch block, 1 with T wave inversion, 1 with tachyarrhythmia, 2 with bradyarrhythmia and 1 with first-degree AV block. The serum laboratories findings at PER were as follows: mean white blood cell (WBC) count (1000/uL) 7.9 ± 4.5; mean hemoglobulin (Hb) (g/dL) 14.4 ± 7.0; mean platelet (PLT) count (1000/uL) 276.4 ± 63.9; median (IQR) troponin I level (ng/mL) 0.0 (0.0−0.0); median (IQR) troponin T level (ng/L) 3.1 (1.5−4.7); median (IQR) CK-MB level (ng/mL) 1.1 (0.8−1.6); median (IQR) NT-ProBNP level (pg/mL) 18.2 (10.9−31.8); median (IQR) BNP level (pg/mL) 5.1 (2.5−8.8); mean serum creatinine (SCr) level (mg/dL) 0.7 ± 0.3; and median (IQR) CRP level (mg/L) 0.9 (0.4−0.7). BNTI-related pericarditis, myocarditis and myopericarditis were diagnosed in 15 (2.2%), 12 (1.8%) and 2 (0.3%) patients, respectively. Six hundred forty-seven (95%) patients were discharged home and 34 (5.0%) needed admission, including 23 (3.4%) admitted to ward and 11 (1.6%) admitted to PICU. All patients survived (Table 1 ).Table 1 Demographics of the patients visited our PERs with discomfort after 1st (Group A) or 2nd (Group B) dose BNTI.
Table 1 Total (n = 681, 100%) Group A (n = 394, 57.9%) Group B (n = 287, 42.1%) p value
Age (years old) 15.1 ± 1.7 14.8 ± 1.8 15.4 ± 1.6 <0.001∗
Junior high school age (12−15 yrs), n (%) 396 (58.1%) 242 (61.4%) 154 (53.7%) 0.043∗
High school age (16−18 yrs), n (%) 285 (48.9%) 152 (38.6%) 133 (46.3%) 0.043∗
Male gender, n (%) 398 (58.4%) 212 (53.8%) 186 (64.8%) 0.004∗
Discomfort after BNTI (days), M (IQR) 3.0 (1.0–12.0) 4.0 (1.6–14.0) 2.0 (1.0–7.0) <0.001∗
Systolic blood pressure (mmHg) at PER 127.4 ± 18.1 125.8 ± 18.8 129.6 ± 17.0 0.007∗
Diastolic blood pressure (mmHg) at PER 74.9 ± 12.1 74.6 ± 12.3 75.4 ± 11.8 0.350
Abnormal EKG findings at PERa, n (%) 18 (2.6%) 9 (2.3%) 9 (3.1%) 0.495
Serum laboratories findings at PER
White blood cell count (1000/uL) 7.9 ± 4.5 8.0 ± 5.0 7.7 ± 3.6 0.410
Hemoglobulin (g/dL) 14.4 ± 7.0 14.2 ± 5.8 14.7 ± 8.4 0.413
Platelet count (1000/uL) 276.4 ± 63.9 284.4 ± 63.1 264.7 ± 63.4 <0.001∗
Troponin I (ng/mL), M (IQR) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.780
Troponin T (ng/L), M (IQR) 3.1 (1.5–4.7) 3.4 (1.5–4.7) 1.5 (1.5–4.6) 0.118
CK-MB (ng/mL), M (IQR) 1.1 (0.8–1.6) 1.0 (0.8–1.5) 1.1 (0.8–1.6) 0.640
NT-ProBNP (pg/mL), M (IQR) 18.2 (10.9–31.8) 18.4 (11.3–35.5) 18.1 (10.8–28.7) 0.234
BNP (pg/mL), M (IQR) 5.1 (2.5–8.8) 5.7 (2.5–10.1) 2.5 (2.5–6.9) 0.291
Serum Creatinine (mg/dL) 0.7 ± 0.3 0.7 ± 0.2 0.6 ± 0.1 0.801
C-reactive protein (mg/L), M (IQR) 0.9 (0.4–3.9) 0.5 (0.3–1.8) 2.9 (0.6–8.4) 0.001∗
Diagnosis
Pericarditis, n (%) 15 (2.2%) 11 (2.8%) 4 (1.4%) 0.220
Myocarditis, n (%) 12 (1.8%) 2 (0.5%) 10 (3.5%) 0.004∗
Myopericarditisb, n (%) 2 (0.3%) 0 (0%) 2 (0.7%) 0.097
Admission, n (%) 34 (5.0%) 15 (3.8%) 19 (6.6%) 0.096
Ward, n (%) 23 (3.4%) 13 (3.3%) 10 (3.5%) 0.895
PICU, n (%) 11 (1.6%) 2 (0.5%) 9 (3.1%) 0.007∗
Discharge home, n (%) 647 (95.0%) 379 (96.2%) 268 (93.4%) 0.096
Survival, n (%) 681 (100%) 394 (100%) 287 (100%) NA
PER: pediatric emergency room; BNTI: Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine immunization; yrs.: years old; M: median; IQR: interquartile range; EKG: Electrocardiogram; BNP = brain natriuretic peptide; PICU: pediatric intensive care unit.
NA: not applicable.
∗p < 0.05.
a Including: ST-segment elevation, incomplete right bundle branch block, T wave inversion, tachy- and bradyarrhythmia, and first-degree AV block.
b Myopericarditis: according to Centers for Disease Control and Prevention Case Definitions for COVID-19 Vaccine-Associated acute myocarditis and pericarditis, this term may be used for patients who meet criteria for both myocarditis and pericarditis.16, 17, 18, 19
In Table 1, Group B seems to be older (p < 0.001), with more male patients (p = 0.004), shorter interval of discomfort after vaccination (p < 0.001), lower platelet count level (p < 0.001), higher CRP level (p = 0.001), more patients diagnosed with myocarditis (p = 0.004) and in need of PICU admission (p = 0.007). Group A only showed lower SBP at PER (p = 0.007). There was no statistical difference regarding DBP at PER, number of patients with abnormal EKG findings at PER, WBC count level, Hb level, troponin I level, troponin T level, CK-MB level, NT-ProBNP level, BNP level, SCr level, number of patients diagnosed with pericarditis and myopericarditis, number of patients in need of ward admission and number of patients discharged home.
In Table 2 , patients admitted to PICU (Group D) were associated with 2nd dose BNTI (p = 0.024), lower SBP at PER (p = 0.011), lower DBP at PER (p = 0.002), more frequent of abnormal EKG findings at PER (p = 0.041), higher troponin I level (p < 0.001), higher CK-MB level (p = 0.031), higher CRP level (p = 0.030), more patients diagnosed with myocarditis (p = 0.006) and longer hospital LOS (p < 0.001). Patients admitted to ward (Group C) were associated with 1st dose BNTI (p = 0.024) and more pericarditis (p < 0.001) being diagnosed. There was no statistical difference regarding age, gender, triage classification, interval of discomfort after BNTI, patient's body height and body weight, WBC count level, Hb level, PLT level, prothrombin time, activated partial thromboplastin time, D-dimer level, troponin T level, NT-ProBNP level, BNP level, SCr level, lactate level, number of patients diagnosed with myopericarditis, or ejection fraction on heart echocardiogram.Table 2 Comparison of adolescent admitted to ward (Group C) or PICU (Group D) due to BNTI-related myocarditis.
Table 2 Total (n = 29, 100%) Group C (n = 18, 62.1%) Group D (n = 11, 37.9%) p value
Age (years old) 15.6 ± 1.5 15.3 ± 1.5 16.2 ± 1.5 0.128
Junior high school age (12−15 yrs), n (%) 12 (41.4%) 9 (50%) 3 (27.3%) 0.243
High school age (16−18 yrs), n (%) 17 (58.6%) 9 (50%) 8 (72.7%) 0.243
Male gender, n (%) 22 (75.9%) 14 (77.8%) 8 (72.7%) 0.768
Triage at PER
Classification I, n (%) 1 (3.4%) 0 (0%) 1 (9.1%) 0.206
Classification II, n (%) 4 (13.8%) 1 (5.6%) 3 (27.3%) 0.107
Classification III, n (%) 24 (82.8%) 17 (94.4%) 7 (63.6%) 0.295
Discomfort after BNTI (days), M (IQR) 3.0 (2.0–15.0) 4.0 (2.3–15.0) 3.0 (2.5–10.5) 0.970
After 1st dose BNTI, n (%) 13 (44.8%) 11 (61.1%) 2 (18.2%) 0.024∗
After 2nd dose BNTI, n (%) 16 (55.2%) 7 (38.9%) 9 (81.8%) 0.024∗
Body Height (cm) 163.9 ± 6.8 164.6 ± 6.1 162.7 ± 8.0 0.479
Body Weight (kg) 57.0 ± 13.7 58.6 ± 15.8 54.2 ± 9.4 0.414
Systolic blood pressure (mmHg) at PER 123.8 ± 30.0 134.7 ± 16.3 116.6 ± 17.3 0.011∗
Diastolic blood pressure (mmHg) at PER 74.3 ± 19.9 82.8 ± 10.2 66.5 ± 14.6 0.002∗
Abnormal EKG findings at PERa, n (%) 14 (41.4%) 6 (33.3%) 8 (72.7%) 0.041∗
Serum laboratories findings at PER
White blood cell count (1000/uL) 8.4 ± 2.2 8.3 ± 2.1 8.6 ± 2.4 0.674
Hemoglobulin (g/dL) 14.3 ± 1.9 14.5 ± 1.7 14.0 ± 2.2 0.448
Platelet count (1000/uL) 249.1 ± 52.6 252.8 ± 44.9 243.2 ± 65.1 0.642
Prothrombin time (s) 13.5 ± 3.7 12.5 ± 0.4 14.1 ± 4.6 0.414
aPTT (s) 30.7 ± 6.9 28.0 ± 0.7 31.9 ± 8.1 0.307
D-dimer (1000 ng/mL), M (IQR) 0.38 (0.23–0.64) 0.44 (0.38–0.63) 0.33 (0.20–0.95) 0.468
Troponin I (ng/mL), M (IQR) 0.2 (0.0–2.4) 0.0 (0.0–0.1) 4.9 (2.0–8.0) <0.001∗
Troponin T (ng/L), M (IQR) 5.6 (4.3–106.0) 4.9 (4.5–5.2) 106.0 (55.2–139.5) 0.252
CK-MB (ng/mL), M (IQR) 2.6 (1.3–16.0) 1.6 (1.2–2.7) 16.0 (8.4–47.0) 0.031∗
NT-ProBNP (pg/mL), M(IQR) 38.3 (15.4–122.7) 27.0 (14.3–87.5) 98.5 (61.6–159.4) 0.943
BNP (pg/mL), M (IQR) 9.0 (5.5–19.3) 6.0 (5.5–7.5) 9.0 (5.0–19.5) 0.468
Serum Creatinine (mg/dL) 0.7 ± 0.3 0.7 ± 0.1 0.8 ± 0.5 0.627
C-reactive protein (mg/L), M (IQR) 4.9 (1.0–19.2) 3.0 (0.5–10.4) 16.3 (7.2–33.4) 0.030∗
Lactate (mg/dL), M (IQR) 10.5 (9.2–11.5) 10.7 (10.6–13.5) 9.7 (9.1–11.4) 0.688
Diagnosis
Pericarditis, n (%) 15 (51.7%) 14 (77.8%) 1 (9.1%) <0.001∗
Myocarditis, n (%) 12 (41.4%) 4 (22.2%) 8 (72.7%) 0.006∗
Myopericarditisb, n (%) 2 (6.9%) 0 (0%) 2 (18.2%) 0.064
EF on cardiac echocardiogram (%) 72.5 ± 10.6 75.3 ± 6.1 68.0 ± 14.6 0.069
COVID-19 PCR positive (NP), n (%) 0 (0%) 0 (0%) 0 (0%) NA
Hospital LOS, M (IQR) (days) 4.0 (3.0–6.0) 3.0 (2.3–4.0) 7.0 (5.5–8.5) <0.001∗
PICU: pediatric intensive care unit; BNTI: Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine immunization; yrs.: years old; M: median; IQR: interquartile range; PER: pediatric emergency room; aPTT: Activated Partial Thromboplastin Time.
EKG: Electrocardiogram; BNP: brain natriuretic peptide; EF: Ejection fraction; COVID-19: coronavirus disease 2019.
NP: Nasopharyngeal; PCR: polymerase chain reaction; LOS: length of stay; NA: not applicable.
∗p < 0.05.
a Including: ST-segment elevation, incomplete right bundle branch block, T wave inversion, tachy- and bradyarrhythmia and first-degree AV block.
b Myopericarditis: according to Centers for Disease Control and Prevention Case Definitions for COVID-19 Vaccine-Associated acute myocarditis and pericarditis, this term may be used for patients who meet criteria for both myocarditis and pericarditis.16, 17, 18, 19
A multivariate analysis of predictive factors for patients needing hospitalization in PICU due to BNTI-related pericarditis and myocarditis was conducted (Table 3 ) and indicated that abnormal EKG findings at PER (95% confidence interval [CI]: 0.036−0.976, p = 0.047) and abnormal serum troponin level at PER (95% CI: 0.002−0.296, p = 0.003) were statistically significant. After 1st or 2nd dose of BNTI, abnormal SBP and DBP at PER, abnormal serum CK-MB level at PER, and abnormal serum CRP level at PER were statistically significant in univariate analyses. Still, they were not retained in the final model of multivariate analyses.Table 3 Multivariable analyses of predictive factors for patients needing hospitalization in PICU due to BNTI-related myocarditis.
Table 3Parameters 95% CI p value
After 1st or 2nd dose BNTI 0.028–1.219 0.079
Abnormal systolic blood pressure (mmHg) at PER 0.093–2.622 0.407
Abnormal diastolic blood pressure (mmHg) at PER 0.077–2.221 0.303
Abnormal EKG findings at PERa 0.036–0.976 0.047∗
Abnormal serum troponin level at PER 0.002–0.296 0.003∗
Abnormal serum CK-MB level at PER 0.029–7.211 0.577
Abnormal serum C-reactive protein level at PER 0.052–4.063 0.486
PICU: pediatric intensive care unit; BNTI: Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine immunization.
CI: confidence interval; PER: pediatric emergency room; EKG: Electrocardiogram.
∗p < 0.05.
a Including: ST-segment elevation, incomplete right bundle branch block, T wave inversion, tachy- and bradyarrhythmia and first-degree AV block.
4 Discussion
BNTI-related myocarditis in adolescents and subsequent hospitalization in PICU was a public safety concern and this study was the first original related observation report with novel finding in Taiwan. We observed that most chief discomfort complaints were chest pain, chest tightness or palpitation after 1st dose or 2nd dose of BNTI. More acute myocarditis and myopericarditis occurred in healthy adolescents after 2nd dose of BNTI; this result was similar to other reports from America, Europe, and Singapore.7 , 8 , 20, 21, 22, 23, 24 Therefore, we considered that BNTI-related myocarditis might not have racial differences. More young male adolescents suffered from myocarditis after 2nd BNTI (myocarditis in 9 males and 1 female; pericarditis in 3 males and 1 female; myopericarditis in 2 male and 0 female patients) in our study, which was compatible with other studies in America, Europe, Korea, Hong Kong and Singapore.8, 9, 10 , 20, 21, 22, 23, 24, 25, 26 All our patients with BNTI-related myocarditis had favorable outcomes without any mortality, so BNTI may be safe in most children and adolescents 12 years of age and older, which was the same as in other international reports.8 , 23, 24, 25, 26, 27, 28, 29
All our 681 enrolled participants were collected from PER, including 18 and 11 patients hospitalized to ward and PICU, respectively. The admission rate might have the possibility of mild overestimation because not all patients with BNTI-related uncomfortable symptoms would visit PER; rather, they might be brought to pediatric outpatient department for milder complaints, where EKG and cardiac echocardiogram would be performed by the cardiology specialists. However, the epidemic prevention policy in our hospital during the pandemic of COVID-19 was that all patients needing admission must be referred to emergency room to complete COVID-19 polymerase chain reaction in order to achieve segregation of patients and flow control measures. Hence, admission numbers are be credible because those in need of ward or ICU hospitalization would be transferred to PER.30
In our study, the initial evaluation of BNTI-related myocarditis usually included measurement of serum WBC count, Hb level, PLT count, troponin level (troponin T or troponin I), CK-MB level, NT-ProBNP level, BNP level CRP level, SCr level, chest X rays and EKG, just like the initial evaluation for those under suspicion of typical myocarditis8 , 31 , 32 at PER. There was only one fly in the ointment: some of our PERs check serum troponin I level, but others check serum troponin T level. Supportive care, medications for uncomfortable symptoms relief, and closely monitoring, including clinical manifestations and serum laboratories data, were mainstay strategies for those with mild or intermediate severity after BNTI8 , 31 , 32; intensive care would be arranged according to patients' clinical condition. Cardiac echocardiogram and cardiac magnetic resonance imaging (MRI) can also be used for the diagnostic and consequent prognosis evaluation of BNTI-related myocarditis.8 , 12 , 27 , 33 , 34 Previous studies mentioned that cardiac MRI might be a potential differentiator in children and adolescents with multisystem inflammatory syndrome rather than myocarditis.35 Cardiac MRI was performed in only four of our patients with BNTI-related myocarditis because their clinical symptoms were much more severe and persistent. The results were all compatible with recent myocarditis.
Though no obvious myocarditis was detected at PER initially in patients presenting with discomfort after 1st dose BNTI, some were still admitted to the ward (Group C) for observation and further examination on the following grounds: 1) the uncomfortable symptoms influenced their daily routine; and 2) physicians had no similar experience before and feared the poor progression of patients' clinical condition in reports regarding pericarditis and/or myocarditis occurrence in adolescents and young adults after BNTI.7, 8, 9, 10, 11, 12 These reasons might explain why Group C seems to be significantly related to 1st dose BNTI, rather than 2nd dose. Several possible mechanisms for mRNA-related myocarditis had been proposed, but these have not yet been elucidated clearly.11 , 12 With increasing experience and research, the most hypothesized pathogenesis might be likely mediated through an autoimmune mechanism (autoantibody-mediated), causing increased prevalence of myocarditis after 2nd dose of mRNA COVID-19 vaccines,11 , 12 which could account for why Group D was more significantly associated with 2nd dose BNTI.
Most of our patients admitted to ward or PICU were mild or intermediate severity without death. They were mainly healthy without history of hospitalization; only 6 patients had underlying diseases or had been hospitalized because of asthma with acute exacerbation, acute gastroenteritis caused by Norovirus, acute epididymitis and renal biopsy for Papillorenal syndrome with renal hypodysplasia. One case encountered serious and critical status due to fulminant myocarditis. She was a case of out-of-hospital cardiac arrest status under the cardiopulmonary resuscitation process in the ambulance during transfer. She was sent to our PICU for further intensive care management. Ventricular tachycardia with poor biventricular function and acute kidney injury were also noted subsequently. Veno-arterial extracorporeal membrane oxygenation for cardiogenic shock and continuous veno-venous hemofiltration program for acute kidney injury with unstable hemodynamic status were both been conducted for life support. She survived with mild residual sequelae due to intensive care and she gradually improved after following a tailored rehabilitation program.
Five patients had discomfort and were hospitalized in the ward post-BNTI, but they were not diagnosed with pericarditis or myocarditis. They were pneumothorax (n = 3), asthma with acute exacerbation (n = 1) and right hemisphere infarction related left side limbs weakness (n = 1). The serum laboratory findings in patient with right hemisphere infarction were all within the normal range. Still, she had the underlying disease of precursor B cell acute lymphoblastic leukemia in remission status. We could not be sure whether her right hemisphere infarction was associated with BNTI or her underlying disease.
Few previous reports presented the predictive factors for patients needing hospitalization in PICU compared to those in the ward because of COVID-19 vaccine or BNTI-related myocarditis. Multivariate analysis showed that abnormal EKG findings and serum troponin levels at PER could be two possible predictive factors. This could aid physicians at PERs in early detection of patients at risk of critical conditions requiring intensive care.
According to the recommendations from the American Heart Association and the American College of Cardiology guidelines,36 those who have suffered from myocarditis need to follow several precautions: 1) before returning to competitive sports, a resting cardiac echocardiogram and EKG should be performed 3–6 months after initial illness; 2) only when the ventricular systolic function returns to the normal range, and serum markers of myocardial injury have normalized, with absence of clinically relevant arrhythmias, can they then resume training and competition; and 3) those with probable or definite myocarditis should not participate in competitive sports while active inflammation is present.36 Those who have had pericarditis must not participate in competitive sports during the acute phase regardless of its pathogenesis and they can only return to full activity when there is complete absence of pericardial effusion and serum markers of inflammation are within normal range.36
The long-term impact on children and adolescents of 12 years of age and older after COVID-19 mRNA vaccine-related myocarditis remains unknown. Further follow-up and surveillance of these patients will be required to assess health, cardiac function, and complications; and there may be an interval of at least 3–6 months after the initial illness before they can return to usual, competitive activity according to the American Heart Association and the American College of Cardiology guidelines.36
4.1 Limitations
This study had some limitations. First, it is possible that patients with mild or self-limited discomfort might not be reported, so our result may be underestimated. Second, as our cases came only from a single tertiary care hospital, the number of patients with BNTI-related myocarditis was relatively small. Additional multicenter studies of larger cohorts are required to identify the accurate incidence rate of myocarditis and to clarify the predictive factors for patients needing hospitalization in PICU compared to those in ward from PER. Third, there might exist a selection bias of total enrolled cases. We could not collect all adolescents visiting our hospital with discomfort after BNTI because some of them with milder complaints might have visited the pediatric outpatient department for EKG and cardiac echocardiogram examination instead of PER. Fourth, cardiac MRI was not applied in all of our patients with BNTI-related myocarditis. Standardized protocols may be considered for cardiac MRI acquisition. Last but not least, the information on long-term follow-up and surveillance for patients' cardiac function, serum markers of cardiac enzyme, and outcome after returning to usual competitive activity is lacking, and further evaluations may be needed.
5 Conclusions
BNTI-related myocarditis seems to be more common after 2nd dose of BNTI in patients aged 12−18 years according to our observational study. Most cases of pericarditis, myocarditis, or myopericarditis were of mild or intermediate severity without death, so that BNT162b2 may be safe in most children and adolescents 12 years of age and older. Abnormal EKG findings and abnormal serum troponin level at PER were noted to be two predictive factors associated with subsequent hospitalization in PICU in adolescents with BNTI-related myocarditis. However, this risk should still be considered in the context of the benefits from BNTI.
Availability of data and materials
The datasets analyzed during the current study are not publicly available because the data are part of the patients' medical record and are treated as confidential. A completely de-identified version of the data is available from the corresponding author on reasonable request, following approval of the institutional review board.
Ethics approval and consent to participate
All methods were carried out in accordance with the relevant guidelines and regulations in accordance with the Declaration of Helsinki. The need to obtain informed consent from participants was waived because this is a purely retrospective study that does not affect patient care. This waiver was approved by Ethics Committee on Human Studies at Chang Gung Memorial Hospital, in Taiwan, R.O.C. (202200782B0). The study was approved by the Ethics Committee on Human Studies at Chang Gung Memorial Hospital, in Taiwan, R.O.C.
Declaration of competing interest
All authors declare that they have no competing of interest.
Abbreviations
COVID-19 coronavirus disease 2019
BNT162b2 Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine
BNTI Pfizer-BioNTech 162b2 mRNA COVID-19 vaccine immunization
PICU pediatric intensive care unit
PER pediatric emergency room
Appendix A Supplementary data
The following is the supplementary data to this article.Multimedia component 1
Multimedia component 1
Acknowledgements
The authors would like to thank the statistician at Chang-Gung Memorial Hospital for assistance with the statistical analysis.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.pedneo.2023.01.005.
==== Refs
References
1 Calcaterra G. Mehta J.L. de Gregorio C. Butera G. Neroni P. Fanos V. COVID 19 vaccine for adolescents. concern about myocarditis and pericarditis Pediatr Rep 13 2021 530 533 34564344
2 Chen Y.C. Yang H.P. Li H.C. Huang P.Y. Chen C.L. Chiu C.H. Features and transmission dynamics of SARS-CoV-2 superspreading events in Taiwan: implications for effective and sustainable community-centered control Pediatr Neonatol 62 2021 437 440 34120866
3 Hong H. Wang Y. Chung H.T. Chen C.J. Clinical characteristics of novel coronavirus disease 2019 (COVID-19) in newborns, infants and children Pediatr Neonatol 61 2020 131 132 32199864
4 Lien Y.L. Li S.C. Kao K.L. Chen C.C. Liang J.S. Clinical manifestations of Taiwanese pediatric patients with COVID-19 infection: a preliminary report of a tertiary center in Northern Taiwan Pediatr Neonatol 63 2022 301 303 35219641
5 Children and COVID-19: state-level data report Available at: https://www.aap.org/en/pages/2019-novel-coronavirus-covid-19-infections/children-and-covid-19-state-level-data-report
6 Committee on Infectious Diseases COVID-19 vaccines in children and adolescents Pediatrics 149 2022 e2021054332
7 Marshall M. Ferguson ID Lewis P Jaggi P Gagliardo C Collins JS Symptomatic acute myocarditis in 7 adolescents after Pfizer- BioNTech COVID-19 vaccination Pediatrics 148 2021 e2021052478
8 Oster M.E. Shay D.K. Su J.R. Gee J. Creech C.B. Broder K.R. Myocarditis cases reported after mRNA-based COVID-19 vaccination in the US from December 2020 to August 2021 JAMA 327 2022 331 340 35076665
9 Kwan M.Y.W. Chua G.T. Chow C.B. Tsao S.S.L. To K.K.W. Yuen K.Y. mRNA COVID vaccine and myocarditis in adolescents Hong Kong Med J 27 2021 326 327 34393110
10 Gargano J.W. Wallace M Hadler SC Langley G Su JR Oster ME Use of mRNA COVID-19 vaccine after reports of myocarditis among vaccine recipients: update from the Advisory Committee on Immunization Practices—United States, June 2021 MMWR Morb Mortal Wkly Rep 70 2021 977 982 34237049
11 Tsilingiris D. Vallianou N.G. Karampela I. Liu J. Dalamaga M. Potential implications of lipid nanoparticles in the pathogenesis of myocarditis associated with the use of mRNA vaccines against SARS-CoV-2 Metabol Open 13 2022 100159 34938983
12 Isaak A. Feisst A. Luetkens J.A. Myocarditis following COVID-19 vaccination Radiology 301 2021 E378 E379 34342500
13 Oliver DMWaDS COVID-19 mRNA vaccines in adolescents and young adults: benefit-risk discussion ACIP meeting 2021 Available at: https://www.cdc.gov/vaccines/acip/meetings/downloads/slides-2021-06/05-COVID-Wallace-508.pdf
14 Taiwan Centers for Disease Control COVID-19 vaccination programs in Taiwan Available at: https://www.cdc.gov.tw/File/Get/-fdNpaJWeGlZQaN4EGiPyw
15 The suggestion of second dose of BioNTech COVID-19 vaccination in adolescent aged 12– 17 years Available at: https://www.mohw.gov.tw/cp-5022-64182-1.html
16 Bozkurt B. Kamat I. Hotez P.J. Myocarditis with COVID-19 mRNA vaccines Circulation 144 2021 471 484 34281357
17 Tom Shimabukuro M. COVID-19 vaccine safety updates ACIP meeting 2021 Available at: https://www.cdc.gov/vaccines/acip/meetings/downloads/slides-2021-06/03-COVID-Shimabukuro-508.pdf
18 Truong D.T. Dionne A. Muniz J.C. McHugh K.E. Portman M.A. Lambert L.M. Clinically suspected myocarditis temporally related to COVID-19 vaccination in adolescents and young adults: suspected myocarditis after COVID-19 vaccination Circulation 145 2022 345 356 34865500
19 Oster M. Overview of myocarditis and pericarditis ACIP meeting 2021 Available at: https://www.cdc.gov/vaccines/acip/meetings/downloads/slides-2021-06/02-COVID-Oster-508.pdf
20 Centers for Disease Control and Prevention Myocarditis and pericarditis after mRNA COVID-19 vaccination Available at: https://www.cdc.gov/coronavirus/2019-ncov/vaccines/safety/myocarditis.html
21 Centers for Disease Control and Prevention Clinical considerations: myocarditis and pericarditis after receipt of mRNA COVID-19 vaccines among adolescents and young adults Available at: https://www.cdc.gov/vaccines/covid-19/clinical-considerations/myocarditis.html
22 Foltran D. Delmas C. Flumian C. De Paoli P. Salvo F. Gautier S. Drici M.D. Myocarditis and pericarditis in adolescents after first and second doses of mRNA COVID-19 vaccines Eur Heart J Qual Care Clin Outcomes 8 2022 99 103 34849667
23 Lee A.S.Y. Balakrishnan I.D.D. Khoo C.Y. Ng C.T. Loh J.K.X. Chan L.L. Myocarditis following COVID-19 vaccination: a systematic review (October 2020–October 2021) Heart Lung Circ 31 2022 757 765 35227610
24 Ilonze O.J. Guglin M.E. Myocarditis following COVID-19 vaccination in adolescents and adults: a cumulative experience of 2021 Heart Fail Rev 27 2022 2033 2043 35449353
25 Park H. Yun K.W. Kim K.R. Song S.H. Ahn B. Kim D.R. Epidemiology and clinical features of myocarditis/pericarditis before the introduction of mRNA COVID-19 vaccine in Korean children: a multicenter study J Korean Med Sci 36 2021 e232 34402230
26 June Choe Y. Yi S. Hwang I. Kim J. Park Y.J. Cho E. Safety and effectiveness of BNT162b2 mRNA Covid-19 vaccine in adolescents Vaccine 40 2022 691 694 35012777
27 Chelala L. Jeudy J. Hossain R. Rosenthal G. Pietris N. White C.S. Cardiac MRI findings of myocarditis after COVID-19 mRNA vaccination in adolescents AJR Am J Roentgenol 218 2022 651 657 34704459
28 Barda N. Dagan N. Ben-Shlomo Y. Kepten E. Waxman J. Ohana R. Safety of the BNT162b2 mRNA Covid-19 vaccine in a nationwide setting N Engl J Med 385 2021 1078 1090 34432976
29 Gurdasani D. Bhatt S. Costello A. Denaxas S. Flaxman S. Greenhalgh T. Vaccinating adolescents against SARS-CoV-2 in England: a risk-benefit analysis J R Soc Med 114 2021 513 524 34723680
30 The management policy during COVID-19 in Chang Gung memorial hospital Available at: https://webapp.cgmh.org.tw/csr/chapter.aspx?id_seq=E1AJ8AG001
31 Leslie T. Cooper J. Myocarditis N Engl J Med 360 2009 1526 1538 19357408
32 Park J. Brekke D.R. Bratincsak A. Self-limited myocarditis presenting with chest pain and ST segment elevation in adolescents after vaccination with the BNT162b2 mRNA vaccine Cardiol Young 32 2022 146 149 34180390
33 Sachdeva S. Song X. Dham N. Heath D.M. DeBiasi R.L. Analysis of clinical parameters and cardiac magnetic resonance imaging as predictors of outcome in pediatric myocarditis Am J Cardiol 115 2015 499 504 25554534
34 Ferreira V.M. Schulz-Menger J. Holmvang G. Kramer C.M. Carbone I. Sechtem U. Cardiovascular magnetic resonance in nonischemic myocardial inflammation: expert recommendations J Am Coll Cardiol 72 2018 3158 3176 30545455
35 Blondiaux E. Parisot P. Redheuil A. Tzaroukian L. Levy Y. Sileo C. Cardiac MRI of children with multisystem inflammatory syndrome associated with COVID-19 Radiology 297 2020 E283 E288 32515676
36 Maron B.J. Udelson J.E. Bonow R.O. Nishimura R.A. Ackerman M.J. Estes N.A. 3rd Eligibility and disqualification recommendations for competitive athletes with cardiovascular abnormalities: task force 3: hypertrophic cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy and other cardiomyopathies, and myocarditis: a scientific statement from the American Heart Association and American College of Cardiology J Am Coll Cardiol 66 2015 2362 2371 26542657
|
PMC010xxxxxx/PMC10205647.txt |
==== Front
Psychiatry Res
Psychiatry Res
Psychiatry Research
0165-1781
1872-7123
Elsevier B.V.
S0165-1781(23)00212-3
10.1016/j.psychres.2023.115262
115262
Article
Psychosocial predictors of trajectories of mental health distress during the COVID-19 pandemic: A four-wave panel study
Lo Coco Gianluca a⁎
Salerno Laura a
Albano Gaia a
Pazzagli Chiara b
Lagetto Gloria c
Mancinelli Elisa de
Freda Maria Francesca f
Bassi Giulia de
Giordano Cecilia a
Gullo Salvatore a
Di Blasi Maria a
a Department of Psychology, Educational Science and Human Movement, University of Palermo, Palermo, Italy
b Department of Philosophy, Social Sciences and Education, University of Perugia, 06123 Perugia, Italy
c Department of History, Society and Human Studies Studium 2000, University of Salento, Edificio 5, Via di Valesio, 24–73100, Lecce, Italy
d Department of Developmental and Socialization Psychology, University of Padova, Via Venezia 8–35132, Padova, Italy
e Digital Health Lab, Centre for Digital Health and Wellbeing, Fondazione Bruno Kessler, Via Sommarive 18–38123, Trento, Italy
f Department of Humanistic Studies, University of Naples Federico II, Naples, Italy
⁎ Corresponding author at: Department of Psychology, Educational Science and Human Movement, University of Palermo, Palermo, Italy, Viale delle Scienze, Edificio 15, Palermo, 90128.
24 5 2023
8 2023
24 5 2023
326 115262115262
12 1 2023
21 5 2023
22 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Previous research suggested that during the COVID-19 pandemic, mental distress did not affect all people equally. This longitudinal study aims to examine joint trajectories of depressive, anxiety, and stress symptoms in a sample of Italian adults during the pandemic, and to identify psychosocial predictors of distress states. We analyzed four-wave panel data from 3,931 adults who had received assessments of depressive, anxiety and stress symptoms between April 2020 and May 2021. Trajectories of individual psychological distress were identified by Latent Class Growth Analysis (LCGA) with parallel processes, and multinomial regression models were conducted to identify baseline predictors. Parallel process LCGA identified three joint trajectory classes for depression, anxiety and stress symptoms. Most individuals (54%) showed a resilient trajectory. However, two subgroups showed vulnerable joint trajectories for depression, anxiety and stress. Expressive suppression, intolerance to uncertainty, and fear of COVID-19 were risk characteristics associated with vulnerable trajectories for mental health distress. Moreover, vulnerability to mental health distress was higher in females, younger age groups and those unemployed during the first lockdown. Findings support the fact that group heterogeneity could be detected in the trajectories of mental health distress during the pandemic and it may help to identify subgroups at risk of worsening states.
Keywords
Anxiety
COVID-19
Depression
Emotion regulation
Intolerance of uncertainty
Parallel-process latent class growth analysis
==== Body
pmc1 Introduction
It has been well-established that the COVID-19 pandemic is having a negative effect on mental health, and there is a need for research to address how to effectively reduce the psychosocial burden among vulnerable groups (COVID-19 Mental Disorders Collaborators, 2021). Several meta-analyses have evidenced that the mental health consequences of COVID-19 are high across countries and gender (Cénat et al., 2021; Robinson et al., 2022; Wu et al., 2021), with a higher prevalence of depression, anxiety, PTSD and insomnia when compared to the general population under normal circumstances (Daly et al., 2022; Kunzler et al., 2021; Prati and Mancini, 2021). Previous meta-analyses of longitudinal studies showed that the overall increase in mental health symptoms was most pronounced during the first months of the pandemic (when measures of shelter-in-place and lockdown had been adopted), before decreasing by mid-2020 (Cénat et al., 2022; Richter et al., 2021; Robinson et al., 2022; Salanti et al., 2022). However, the increase in mental distress did not affect all people equally, with some subgroups showing marked increases. Some prior longitudinal studies identified different mental health distress trajectories during the pandemic (Bendau et al., 2022; Fancourt et al., 2021; Fioravanti et al., 2022; Liang et al., 2022). For example, Pierce et al. (2021) showed that the mental health of most UK adults remained resilient between April and October 2020, whereas around one in nine individuals had deteriorating mental health. In a French cohort study (Lu et al., 2022), most individuals exhibited trajectories with a relatively low level of anxiety and depressive symptoms, whereas younger individuals and females at large were found to be more vulnerable as regards their mental health. Indeed, female gender was associated with a higher prevalence of anxiety and depression (Cénat et al., 2022). Thus, to sum up, prior research suggested heterogeneity in the psychological responses to the COVID-19 outbreak, and there was a preliminary effort to identify classes of individuals displaying non-resilient mental health trajectories (Ahrens et al., 2021; Shevlin et al., 2021). However, most previous studies did not monitor long-lasting fluctuations in mental health distress, from data reported and collected in 2020. Moreover, research examining psychological predictors of mental health distress trajectories is still limited (Fancourt et al., 2021; McPherson et al., 2021).
In the current study, we will examine intolerance of uncertainty (IU), emotion regulation (ER), and fear of the COVID-19 pandemic as psychological predictors of mental health distress trajectories. The COVID-19 pandemic was an unprecedented event and represented a special challenge for individuals with a low capacity to tolerate uncertainty (Rettie and Daniels, 2021). To date, there is initial evidence that IU could predict mental health problems during the pandemic (Reizer et al., 2021; Shevlin et al., 2021). Individuals with IU often experience difficulties in regulating emotions (Sahib et al., 2023). ER is defined as the process in which individuals manage their emotional experience by using regulation strategies (Aldao et al., 2010; Gross, 2003). ER strategies can be grouped into either adaptive or maladaptive strategies (i.e., if they regulate emotions effectively or if they do not) (Gross, 2014). An example of ER maladaptive strategy is linked to expressive suppression, i.e., when individuals restrain unwanted emotional expressions; on the other hand, an example of ER adaptive strategy includes reappraisal, i.e., when individuals develop positive interpretations of the situation after an initial, negative appraisal (Aldao et al., 2010). To date, only a few studies have examined the role of ER processes on mental health during the COVID-19 pandemic, suggesting that increased emotional suppression was associated with poorer psychological health, whereas cognitive reappraisal to regulate emotions was associated with greater resilience (Cardi et al., 2021; Low et al., 2021; Xu et al., 2020). However, none of the previous studies investigated whether maladaptive ER strategies increased the likelihood of associations with vulnerable mental health trajectories. The present longitudinal study aims to identify (1) empirical trajectories of mental health distress (i.e., depression, anxiety and stress) over time by analyzing panel data from four waves of a national sample of Italian adults collected between April 2020 and May 2021; and (2) to identify which demographic and psychological factors were associated with the different longitudinal profiles. Consistently with prior longitudinal research (Bendau et al., 2022; McPherson et al., 2021; Shevlin et al., 2021), it was predicted that trajectories reflecting worsening mental health distress over four-time points would be associated with demographics such as female gender, and younger age, and psychological variables such as high IU, ER maladaptive strategies and fear of the COVID-19 pandemic.
2 Methods
2.1 Participants and procedures
Data for this study comes from a large-scale national project on the mental health correlates of the COVID-19 pandemic (Di Blasi et al., 2021), which involved a general adult population sample from Italy. Inclusion criteria included: (1) age 18 years or older, (2) resident in Italy at the time the survey was completed, and (3) having sufficient language skills to complete the survey. Participants were assessed repeatedly in up to four waves during the COVID-19 pandemic (i.e., T1 = 7th- 24th April 2020; T2 = 18th-31st May 2020; T3 = 26th June-8th July 2020; T4 = 24th April-11th May 2021). A total of 3864 individuals participated at T1, 1174 individuals at T2, 714 individuals at T3, and 731 individuals at T4. Twenty-five (0.6%) participants were excluded because they were not residents in Italy, and sixteen participants (0.4%) were excluded because of age < 18 years. Since we kept missing data points when matching the data for all four waves, the analytical sample included 3931 participants (n = 3823 at T1; n = 1162 at T2; n = 709 at T3; n = 726 at T4). Two-hundred and ninety-nine participants (7.6%) present complete data on all four waves; 710 participants (18.1%) present data on at least three waves; and 1480 participants (37.6%) present data on at least two waves. Table 1 reports participants’ socio-demographic and health-related characteristics. A detailed description of the national Covid-related restrictions as well as the number of cases, deaths and recovery at the moment of the four assessment points are reported in Figure S1 (Supplementary materials). Participants were recruited via social media platforms and were assessed using an online survey. All participants provided written informed consent before inclusion in the study. The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the University of [blinded for review].Table 1 Participants’ socio-demographic and health-related characteristics.
Table 1 Participants (n = 3931)
Age, M (SD) 36.55 (14.76)
Gender, n (%)
Females 2802 (71.3)
Males 1021 (26.0)
missing 108 (2.7)
Educational level, n (%)
8–13 years of education 1675 (42.6)
degree/post-degree 2148 (54.6)
missing 108 (2.7)
Employment status, n (%)
Employed 2059 (52.4)
Unemployed 1764 (44.9)
missing 108 (2.7)
Personal COVID-19 infection at T1, n (%) yes 15 (0.4)
Relatives COVID-19 infection at T1, n (%) yes 780 (19.8)
Chronic diseases, n (%) yes 279 (7.1)
Disabilities, n (%) yes 100 (2.5)
Note: T1 = 7th-24th April 2020.
2.2 Measures
2.2.1 Demographic and health-related information
Participants reported their age, gender, educational level, and employment status (categorized as employed vs unemployed based on whether or not having a regular income). Moreover, data about health-related characteristics (i.e., personal and relatives COVID-19 infection, presence of chronic diseases or disabilities) were collected.
2.2.2 Depressive symptoms, anxiety, and stress
The Depression, Anxiety, and Stress Scale (DASS-21; Lovibond and Lovibond, 1995; Bottesi et al., 2015) was used to assess psychological distress. The DASS is a 21-item measure that yields three subscales (7 items each): depression, anxiety, and stress. Participants rate items using a 4-point Likert Scale ranging from 0 (did not apply to me at all) to 3 (applied to me very much or most of the time). The DASS-21 total score had good to excellent internal consistency in the present study (Cronbach's alpha across T1-T4: depression = range 0.895 - 0.919; anxiety = range 0.874 - 0.900; stress = range 0.916 - 0.933).
2.2.3 Intolerance of uncertainty
The Intolerance of Uncertainty Scale-Revised (IUS-R; Carleton et al., 2007; Lauriola et al., 2016) was used to assess IU. The IUS-R consists of 12 items, rated on a five-point Likert scale ranging from 1 (not at all characteristic of me) to 5 (entirely characteristic of me). Only the IUS-R total score was used in this study, with good internal consistency (T1 Cronbach's alpha = 0.881).
2.2.4 Emotion regulation
The Emotion Regulation Questionnaire (ERQ; Gross and John, 2003; Balzarotti et al., 2010) was used to measure individuals’ tendency to regulate their emotions. The ERQ consists of ten items which yield two subscales: Cognitive Reappraisal (6 items) and Expressive Suppression (4 items). Participants rate items using a 7-point Likert-type scale, ranging from 1 (strongly disagree) to 7 (strongly agree). In the current study, the ERQ showed good internal consistency (T1 Cronbach's alphas = 0.871 and 0.607, for Cognitive Reappraisal and Expressive Suppression subscale, respectively; mean inter-item correlation for Expressive Suppression subscale = 0.278).
2.2.5 Fear of the COVID-19 pandemic
The Fear of COVID-19 Scale (Ahorsu et al., 2022; Soraci et al., 2020) was used to measure the individual's fear of the COVID-19 pandemic. This scale includes 7 items, rated on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). In the current study, this scale showed good internal consistency (T1 Cronbach's alpha = 0.877).
2.3 Plan of data analysis
As a preliminary step, data were screened for missing values. Little's MCAR test revealed that the missing values were not missing completely at random (χ2 = 214.785, p = .022). No significant differences were found in demographics (i.e., sex, employment status, presence of chronic diseases or disabilities, and personal COVID-19 infection), nor in depressive and stress symptoms at T1 among participants with complete data in all waves and those with missing data. Significant differences were only found for participants’ educational level (p < .001), COVID-19 infection of relatives at T1 (p < .05) and anxiety symptoms at T1 (M±SD = 3.85±4.57 and 3.07±4.03 for participants with and without missing data, respectively; t = 3.179, p < .01). The missing data were handled using the full information maximum likelihood (FIML) method, which has been shown to perform better than data deletion-based methods in reducing bias in longitudinal studies, also with high rates of missing data (Lee and Shi, 2021) and is frequently used in latent trajectory studies (van De Schoot et al., 2017). The normality of the continuous variables was checked by using skewness and kurtosis. All continuous variables had a normal distribution (|Sk| < 2 and |Ku| < 7; Hair et al., 2010). The internal consistency of the scales was computed by Cronbach's α and the mean inter-item correlation (only for the ERQ-Expressive Suppression subscale). Mean inter-item correlations between 0.15 and 0.50 indicate adequate internal consistency (Clark and Watson, 1995). Descriptive statistics were computed for demographics and variables of interest.
Joint trajectories (or co-development) of depressive, anxiety, and stress symptoms were created by using a three process parallel Latent Class Growth Analysis (LCGA; Berlin et al., 2014), following the Guidelines for Reporting on Latent Trajectory Studies (GRoLTS; van de Schoot et al., 2017; more details are reported in Supplementary data, Table S1). Parallel-process LCGA is a data-driven technique that extends the typical univariate LCGA to a parallel-process approach, allowing for the consideration of multiple growth trajectories simultaneously through a small number of classes. Trajectory classes are operationalized as groups of individuals who approximately follow the same developmental trajectory (Andruff et al., 2009). Models with 1 to 5 classes were fitted. To decide the optimum number of classes, the following were used: the Akaike Information Criteria (AIC), the Bayesian Information Criteria (BIC), the sample size adjusted BIC (ssaBIC), the size of the smallest class size (> 5%), entropy, the Lo-Mendel-Rubin Likelihood Ratio Test (LMR-LRT). In addition, the choice of best-fitting solutions was based on theoretical coherence (i.e., substantive interpretability of the trajectory classes and identification of trajectories without overfitting) and explanatory relevance. The nature of classes was examined based on initial levels (i.e., intercept) and changes (i.e., linear slope and quadratic terms) in depression, anxiety and stress symptoms. We used the clinical severity cut-offs for the DASS-21 subscales (Lovibond and Lovibond, 1995) to label the different classes. The syntax file of the selected model is reported in Table S2 (Supplementary data).
Latent class membership was regressed on baseline socio-demographic characteristics (i.e., age, gender, educational level, and employment status), health-related characteristics (i.e., personal and relatives’ COVID-19 infection, presence of chronic diseases or disabilities) and psychological factors (i.e., IU, ER, fear of COVID-19 pandemic) in order to identify risk factors associated with trajectories of change for psychological distress (i.e., depressive, anxiety and stress symptoms). More specifically, classes were compared using the chi-square test for categorical variables and ANOVA for continuous variables. Furthermore, multinomial regression models were conducted, whilst adjusted odds ratios (ORs) and 95% confidence intervals (Cis) were calculated in order to examine factors associated with trajectories of change for psychological distress (i.e., depressive, anxiety, and stress symptoms). Only variables with a p-value of < 0.10 on univariate analysis were subjected to multivariable analysis and non-significant variables in the multivariable model were removed by a backward stepwise approach.
Finally, an additional sensitivity analysis was conducted to test whether the trajectory modeling would hold when removing the 36.7% of participants who only have one assessment point. Analyses were conducted in SPSS v. 22 and Mplus v. 7.0.
3 Results
3.1 Joint longitudinal trajectories of depressive, anxiety and stress symptoms
A three-class model (Table S3) was selected for a combination of factors, including (1) the smallest class in the four-class model, which was close to 5%, and thus it may indicate model overfitting, (2) the entropy was lower in the four-class model than in the three-class model, and (3) the classes identified as clinically distinct. As shown in Fig. 1 and Table S4, class 1 (labelled as “Moderate-chronic class”; n = 505, 13%) had moderate depression levels (i.e., range 14–20), moderate anxiety levels (i.e., range 10–14), and mild stress levels (i.e., 15–19). Baseline levels of depression, anxiety and stress are higher than the other classes and remain stable over time (both linear and quadratic slopes were not significant; linear slopes: p = .674, p = .441, and p = .348; quadratic slopes: p = .856, p = .603, and p = .267, for depression, anxiety and stress, respectively; Table 2 ). Class 2 (labelled as “Normal-increasing”; n = 2110, 54%) had normal depression, anxiety and stress levels. Baseline scores were in the normal range (furthermore, only a low percentage of subjects in this class exceed the cut-off of mild problems: 4.7%, 2.1%, and 0.8% for depression, anxiety and stress, respectively) and showed a significant decrease from T1 to T3, but a subsequent significant increase at T4 (all linear and quadratic slopes were significant at p < .001; Table 2). Class 3 (labelled as “Mild-vulnerable”; n = 1316, 33%) had mild depression levels (i.e., range 10–13), but normal levels of anxiety (i.e., range 0–7) and stress (i.e., range 0–14). In this class, the number of subjects who exceed the mild problematic cut-off is higher than in class 2 (i.e., 54.5%, 33.6%, and 45.1% for depression, anxiety and stress, respectively). Depression and anxiety levels showed a significant decrease from T1 to T3, but a subsequent significant increase at T4, remaining in the mild range (linear and quadratic slopes were significant at p < .05 and p < .01 for depression and anxiety, respectively; Table 2). Stress levels remain stable over time (both linear and quadratic slopes were not significant: p = .117 and p = .099, respectively; Table 2).Fig. 1 Plot showing the mean trajectories with 95% CIs of the three-class model for depressive, anxiety and stress symptoms.
Fig. 1
Table 2 Intercepts, liner slopes, quadratic terms and standard errors (in parentheses) of the Latent Class Longitudinal Trajectory Groups for depression, anxiety and stress symptoms.
Table 2Class Parameter Depressive symptoms Anxiety symptoms Stress symptoms
Class 1 (n = 505) Intercept 14.316 (0.281)⁎⁎⁎ 12.445 (0.462)⁎⁎⁎ 16.941 (0.194)⁎⁎⁎
Linear −0.278 (0.660) −0.677 (0.879) .465 (0.495)
Quadratic .043 (0.240) .162 (0.311) −0.197 (0.178)
Class 2 (n = 2110) Intercept 3.004 (0.123)⁎⁎⁎ 1.152 (0.051)⁎⁎⁎ 4.926 (0.163)⁎⁎⁎
Linear −0.690 (0.163)⁎⁎⁎ −0.540 (0.098)⁎⁎⁎ −0.860 (0.211)⁎⁎⁎
Quadratic .268 (0.063)⁎⁎⁎ .231 (0.042)⁎⁎⁎ .379 (0.080)⁎⁎⁎
Class 3 (n = 1316) Intercept 9.283 (0.312)⁎⁎⁎ 4.645 (0.244)⁎⁎⁎ 12.134 (0.277)⁎⁎⁎
Linear −0.777 (0.358)* −0.798 (0.285)⁎⁎ −0.537 (0.343)
Quadratic .287 (0.131)* .354 (0.107)⁎⁎ .203 (0.123)
Note: * p < .05; ** p < .01; *** p < .001.
3.2 Predicting joint class membership
Factors associated with joint trajectories of depressive, anxiety and stress symptoms are presented in Tables S5-S6 (Supplementary Materials) and Fig. 2 . Univariate analyses (Table S5) showed an association between age, sex, educational level, employment status, IU, ERQ-cognitive reappraisal, ERQ-expressive suppression, fear of the COVID-19 pandemic, and joint trajectories of depressive, anxiety and stress symptoms. Data about the multivariate regression model (Fig. 2) showed that the odds against older and male participants, as well as those with higher scores on cognitive reappraisal, were lower with regard to having a problematic joint trajectory of depressive, anxiety and stress symptoms (i.e., Moderate-chronic or Mild-vulnerable classes). Moreover, for unemployed participants, as well as those with greater IU, expressive suppression and fear of the COVID-19 pandemic, the odds for registering a problematic joint trajectory of depressive, anxiety and stress symptoms (i.e., Moderate-chronic and Mild-vulnerable classes) were higher (see Table S6 - Supplementary Material – for more information).Fig. 2 Multivariable analysis including variables significantly associated with trajectories of depressive, anxiety and stress symptoms. Note: * p < .05; ** p < .01; *** p < .001.
Fig. 2
3.3 Sensitivity analysis
Sensitivity checks were carried out by running the analyses with a subsample of 1480 participants with two+ time points (Supplementary Material). Results about the joint longitudinal trajectories of depressive, anxiety and stress symptoms as well as the factors associated with class membership remained largely unchanged (see Tables S7-S11).
4 Discussion
The current longitudinal study examined joint trajectories of anxiety, depression and stress from April 2020 to May 2021 during the COVID-19 pandemic. The analysis identified a three-class model for relationships between depression, anxiety and stress outcomes, with mental health trajectories able to identify participants in relation to a stability or to a worsening of symptoms over time. Results show that only one subgroup (13% of participants) reported significant moderate levels of depression and anxiety, and a mild level of stress, which remained stable across time. However, the majority of the sample (i.e., 54% of participants) exhibited a resilient mental health trajectory characterized by minimal changes in depression, anxiety and stress. Overall, our findings support the fact that group heterogeneity could be detected in the trajectories of mental health distress during the COVID-19 pandemic and it may help to identify subgroups at risk of chronic distress. Previous studies explored longitudinal trajectories of mental health symptoms in 2020 (Fancourt et al., 2021; Pierce et al., 2021; Shevlin et al., 2021) and suggested a negative impact of the pandemic on some vulnerable subgroups (Liang et al., 2022; Lu et al., 2022; McPherson et al., 2021). A previous longitudinal study with an adult Italian population showed that after the first lockdown, mental health symptoms decreased slightly but, in conjunction with the newly imposed restrictions, they rose, due to the second wave of the pandemic (Fioravanti et al., 2022). Our findings add that more than one year after the national lockdown, around one third of the sample exhibited mild levels of depression and anxiety, which increased in 2021, and around 13% of individuals belonged to a class displaying a moderate level of depression and anxiety. Based on the cut-off points of DASS severity (Lovibond and Lovibond, 1995), these two class trajectories remained in the moderate to mild range. Taken together, these findings seem in line with previous meta-analytic evidence showing a decline in mental health during the onset of the COVID-19 pandemic and a slight decrease in symptoms after the ease of social restrictions (Richter et al., 2021; Robinson et al., 2022; Salanti et al., 2022). Our findings are also in line with those reported in a study with a Chinese population (Chen et al., 2022), which found a three-class solution for depression and anxiety, with similar trajectories of resilience, chronicity and mild, declining symptoms in 2020. However, the current findings suggest that a vulnerable subgroup reported stable and moderate mental health symptoms in mid 2021, which might well reflect an ongoing struggle against the pandemic. This finding differs from those reported in a study with a Polish population (Gambin et al., 2022), which also found a small, worsening class, the members of which experienced an increase in anxiety and depression symptoms in mid-2020 and a slight decrease during the second lockdown in April 2021. It is likely that these different classes may reflect specific reactions to the restrictions due to the pandemic waves. Although the national lockdown in Italy ended in early May 2020, the Italian government activated area-specific differential restrictions for COVID-19 prevention after the spread of the second wave of the pandemic in October 2020. Thus, our data may suggest that some vulnerable individuals showed sustained mental health distress in 2021, given the ongoing social restriction measures due to the high number of cases and hospitalizations due to COVID-19 infection. However, in the present study we were unable to examine the role of participants’ adherence to restriction measures, because they differed between Regions and the assessment spanned a period in which these rules changed. Our finding seems also to be in line with those reported by a large community study in Germany (Bendau et al., 2022), which reported some peaks in symptoms during the second and third pandemic waves, in times of increased infection rates. Taken together, these findings suggest that research is certainly needed to examine different trajectories of mental health distress over the course of the pandemic, by also considering the impact of differentiated restrictions in different countries, as well as the occurrence of the immune-escape virus variant Omicron and the uptake of vaccination.
Regarding the psychological and sociodemographic characteristics associated with vulnerable mental health trajectories, the results of the multivariate regression models showed that IU, expressive suppression, and fear of COVID-19 pandemic at baseline were risk characteristics associated with vulnerable joint trajectories for anxiety, depression and stress. To date, the role of IU (Reizer et al., 2021; Shevlin et al., 2021) and maladaptive ER strategies (Chen et al., 2022; Groarke et al., 2021; Low et al., 2021) in predicting mental health distress during the pandemic has received initial support. However, whereas these previous reports tracked mental health during the first year of the pandemic, we found that both these psychological risk factors could allow one to predict individual belonging to vulnerable trajectories for depression and anxiety one year after the end of national lockdown in Italy. Of note, cognitive reappraisal (i.e., modifying the cognitive meaning attributed to a threat) was associated with a lower likelihood that participants would show vulnerable class membership for anxiety, depression, and stress. This finding suggests that an ability to effectively regulate one's emotions may play a role in alleviating the negative mental effects of the pandemic and improve one's resilience to threat of pandemic, in accordance with previous research reports from Italy (Cardi et al., 2021; Preti et al., 2021). The findings of the current study also suggest that high feelings of fear of COVID-19 pandemic during the lockdown were associated with belonging to vulnerable classes as regards depression, anxiety and stress, thus supporting the role of these negative feelings in mental health distress during the pandemic (Alimoradi et al., 2022). Regarding sociodemographic variables, our results showed that vulnerability to depression, anxiety and stress was higher in females, younger age groups and those unemployed during the first lockdown. These results may reflect a gendered and age-related response to the pandemic, which were evidenced by prior reviews (Cénat et al., 2022; Robinson et al., 2022). However, in the current study females and young adults were overrepresented and this pattern of results may constitute a sample artifact.
The current findings have relevant implications for mental health policy makers. Given the ongoing spread of the pandemic and to prepare for future pandemics, policy makers need to ensure that an adequate mental health service provision might be targeted at the vulnerable groups of people reporting enduring patterns of distress. Maladaptive expressive suppression and IU seem to be pronounced risk factors and should get particular attention in therapeutic or preventive interventions. Emotion regulation training aimed at improving reappraisal, in order to tackle psychological stress and protect one from adopting risky behavior, might be especially important in this context. Finally, our results showed that many inequalities in mental health (such as inequalities by age, gender or unemployment) did remain and these vulnerable groups have remained at risk. Thus, it is essential to find ways of supporting vulnerable groups throughout the pandemic.
This study has some limitations. Given the non-probability convenience sampling, participants are not unconditionally representative of the general population in Italy. Secondly, online-based recruitment might amplify a selection bias, as previously outlined in mental health research during the pandemic (Richter et al., 2021). Moreover, high-frequency online data collection in the context of COVID-19 can lead to a loss of participants with poorer mental health, resulting in biased trends of deterioration. Heterogeneity revealed by the LCGA could indicate other time-dependent effects not caught by the model, e.g., localized spikes in infections or local/regional restrictions. Finally, we lacked reliable data on previous mental health diagnosis which can be an important predictor of worsening mental health during the pandemic (Pierce et al., 2021).
In summary, the current study showed that a substantial group of individuals in the general population has been unaffected by mental health distress during the pandemic. However, a chronic and stable mental distress trajectory can be isolated, and psychological characteristics such as IU, ER expressive suppression, and high fear of the COVID-19 pandemic emerged as risk factors for sustained mental health problems during the pandemic. Future research should focus on further clarifying what psychosocial factors might play a key role in heightening or buffering psychological distress in the population (Kunzler et al., 2021) in order to explain why worsening mental distress did not affect all people equally. Furthermore, given the ongoing struggle with the pandemic, future research will need to investigate the impact of new SARS-CoV-2 variants as well as prolonged economic difficulties on mental health distress in the general population.
Author statement
G.L.C.: Conceptualization, Investigation, Supervision, Writing – Original draft preparation. L.S.: Methodology, Formal analysis, Writing – Original draft preparation. G.A.: Data Curation, Writing - Review & Editing C.P.: Data Curation, Writing – Review & Editing. G.L.: Data Curation, Writing – Review & Editing. E.M.: Data Curation, Writing – Review & Editing M.F.F.: Data Curation, Writing – Review & Editing. G.B.: Data Curation, Writing – Review & Editing. C.G.: Writing – Review & Editing. S.G.: Methodology, Formal analysis, Writing – Review & Editing. M.D.B.: Conceptualization, Supervision, Project administration. All authors approved the submitted paper.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of Competing Interest
None of the authors report any conflicts of interest with this work.
Appendix Supplementary materials
Image, application 1
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.psychres.2023.115262.
==== Refs
References
Ahorsu D.K. Lin C.Y. Imani V. Saffari M. Griffiths M.D. Pakpour A.H. The fear of COVID-19 scale: development and initial validation Int. J. Ment. Health Addict. 20 2022 1537 1545 10.1007/s11469-020-00270-8 32226353
Ahrens K.F. Neumann R.J. Kollmann B. Plichta M.M. Lieb K. Tüscher O. Reif A. Differential impact of COVID-related lockdown on mental health in Germany World Psychiatry 20 2021 140 141 10.1002/wps.20830 33432755
Alimoradi Z. Ohayon M.M. Griffiths M.D. Lin C.Y. Pakpour A.H. Fear of COVID-19 and its association with mental health-related factors: systematic review and meta-analysis BJPsych Open 8 2022 e73 10.1192/bjo.2022.26 35307051
Aldao A. Nolen-Hoeksema S. Schweizer S. Emotion-regulation strategies across psychopathology: a meta-analytic review Clin. Psychol. Rev. 30 2 2010 217 237 10.1016/j.cpr.2009.11.004 20015584
Andruff H. Carraro N. Thompson A. Gaudreau P. Benoît L. Latent class growth modelling: a tutorial Tutor Quant. Methods Psychol. 5 2009 11 24 10.20982/tqmp.05.1.p011
Balzarotti S. John O.P. Gross J.J. An Italian adaptation of the Emotion Regulation Questionnaire Eur. J. Psychol. Assess. 26 2010 61 67 10.1027/1015-5759/a000009
Bendau A. Asselmann E. Plag J. Petzold M.B. Ströhle A. 1.5 years pandemic - Psychological burden over the course of the COVID-19 pandemic in Germany: a nine-wave longitudinal community study J. Affect. Disord. 319 2022 381 387 10.1016/j.jad.2022.09.105 36162668
Berlin K.S. Parra G.R. Williams N.A. An introduction to latent variable mixture modeling (part 2): longitudinal latent class growth analysis and growth mixture models J. Pediatr. Psychol. 39 2014 188 203 10.1093/jpepsy/jst085 24277770
Bottesi G. Ghisi M. Altoè G. Conforti E. Melli G. Sica C. The Italian version of the Depression Anxiety Stress Scales-21: factor structure and psychometric properties on community and clinical samples Compr. Psychiatry 60 2015 170 181 10.1016/j.comppsych.2015.04.005 25933937
Cardi V. Albano G. Gentili C. Sudulich L. The impact of emotion regulation and mental health difficulties on health behaviours during COVID19 J. Psychiatr. Res. 143 2021 409 415 10.1016/j.jpsychires.2021.10.001 34655950
Carleton R.N. Norton M.A. Asmundson G.J. Fearing the unknown: a short version of the Intolerance of Uncertainty Scale J. Anxiety Disord. 21 2007 105 117 10.1016/j.janxdis.2006.03.014 16647833
Cénat J.M. Blais-Rochette C. Kokou-Kpolou C.K. Noorishad P.G. Mukunzi J.N. McIntee S.E. Dalexis R.D. Goulet M.A. Labelle P.R. Prevalence of symptoms of depression, anxiety, insomnia, posttraumatic stress disorder, and psychological distress among populations affected by the COVID-19 pandemic: a systematic review and meta-analysis Psychiatry Res. 295 2021 113599 10.1016/j.psychres.2020.113599
Cénat J.M. Farahi S.M.M.M. Dalexis R.D. Darius W.P. Bekarkhanechi F.M. Poisson H. Broussard C. Ukwu G. Auguste E. Nguyen D.D. Sehabi G. Furyk S.E. Gedeon A.P. Onesi O. El Aouame A.M. Khodabocus S.N. Shah M.S. Labelle P.R. The global evolution of mental health problems during the COVID-19 pandemic: a systematic review and meta-analysis of longitudinal studies J. Affect. Disord. 315 2022 70 95 10.1016/j.jad.2022.07.011 35842064
Chen S. Bi K. Sun P. Bonanno G.A. Psychopathology and resilience following strict COVID-19 lockdowns in Hubei, China: examining person- and context-level predictors for longitudinal trajectories Am. Psychol. 77 2 2022 262 275 10.1037/amp0000958 35143231
Clark L.A. Watson D. Constructing validity: basic issues in objective scale development Psychol. Assess. 7 1995 309 319 10.1037/1040-3590.7.3.309
COVID-19 Mental Disorders Collaborators Global prevalence and burden of depressive and anxiety disorders in 204 countries and territories in 2020 due to the COVID-19 pandemic Lancet 398 2021 1700 1712 10.1016/S0140-6736(21)02143-7 34634250
Daly M. Sutin A.R. Robinson E. Longitudinal changes in mental health and the COVID-19 pandemic: evidence from the UK Household Longitudinal Study Psychol. Med. 52 2022 2549 2558 10.1017/S0033291720004432 33183370
Di Blasi M. Gullo S. Mancinelli E. Freda M.F. Esposito G. Gelo O.C.G. Lagetto G. Giordano C. Mazzeschi C. Pazzagli C. Salcuni S. Lo Coco G. Psychological distress associated with the COVID-19 lockdown: a two-wave network analysis J. Affect. Disord. 284 2021 18 26 10.1016/j.jad.2021.02.016 33582428
Fancourt D. Steptoe A. Bu F. Trajectories of anxiety and depressive symptoms during enforced isolation due to COVID-19 in England: a longitudinal observational study Lancet Psychiatry 8 2021 141 149 10.1016/S2215-0366(20)30482-X 33308420
Fioravanti G. Bocci Benucci S. Prostamo A. Banchi V. Casale S. Effects of the COVID-19 pandemic on psychological health in a sample of Italian adults: a three-wave longitudinal study Psychiatry Res. 315 2022 114705 10.1016/j.psychres.2022.114705
Gambin M. Oleksy T. Sękowski M. Wnuk A. Woźniak-Prus M. Kmita G. Holas P. Pisula E. Łojek E. Hansen K. Gorgol J. Kubicka K. Huflejt-Łukasik M. Cudo A. Łyś A.E. Szczepaniak A. Bonanno G.A. Pandemic trajectories of depressive and anxiety symptoms and their predictors: five-wave study during the COVID-19 pandemic in Poland Psychol. Med. 2022 1 3 10.1017/S0033291721005420 Advance online publication
Groarke J.M. McGlinchey E. McKenna-Plumley P.E. Berry E. Graham-Wisener L. Armour C. Examining temporal interactions between loneliness and depressive symptoms and the mediating role of emotion regulation difficulties among UK residents during the COVID-19 lockdown: longitudinal results from the COVID-19 psychological wellbeing study J. Affect. Disord. 285 2021 1 9 10.1016/j.jad.2021.02.033 33610876
Gross J.J. John O.P. Individual differences in two emotion regulation processes: implications for affect, relationships, and well-being J. Pers. Soc. Psychol. 85 2003 348 362 10.1037/0022-3514.85.2.348 12916575
Gross J.J. Handbook of Emotion Regulation 2nd ed. 2014 Guilford New York, NY
Hair J. Black W.C. Babin B.J. Anderson R.E. Multivariate Data Analysis 7th ed. 2010 Pearson Education International Upper Saddle River, New Jersey
Kunzler A.M. Röthke N. Günthner L. Stoffers-Winterling J. Tüscher O. Coenen M. Rehfuess E. Schwarzer G. Binder H. Schmucker C. Meerpohl J.J. Lieb K. Mental burden and its risk and protective factors during the early phase of the SARS-CoV-2 pandemic: systematic review and meta-analyses Glob. Health 17 2021 34 10.1186/s12992-021-00670-y
Lauriola M. Mosca O. Carleton R.N. Hierarchical factor structure of the Intolerance of Uncertainty Scale short form (IUS-12) in the Italian version TPM Test Psychom. Methodol Appl. Psychol. 23 3 2016 377 394 10.4473/TPM23.3.8
Lee T. Shi D. A comparison of full information maximum likelihood and multiple imputation in structural equation modeling with missing data Psychol. Methods 26 4 2021 466 485 10.1037/met0000381 33507765
Liang K. Huang L. Qu D. Bu H. Chi X. Self-compassion predicted joint trajectories of depression and anxiety symptoms during the COVID-19 pandemic: a five-wave longitudinal study on Chinese college students J. Affect. Disord. 319 2022 589 597 10.1016/j.jad.2022.09.078 36155236
Lovibond P.F. Lovibond S.H. The structure of negative emotional states: comparison of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and Anxiety Inventories Behav. Res. Ther. 33 1995 335 343 10.1016/0005-7967(94)00075-u 7726811
Low R.S.T. Overall N.C. Chang V.T. Henderson A.M.E. Sibley C.G. Emotion regulation and psychological and physical health during a nationwide COVID-19 lockdown Emotion 21 2021 1671 1690 10.1037/emo0001046 34843308
Lu L. Contrand B. Dupuy M. Ramiz L. Sztal-Kutas C. Lagarde E. Mental and physical health among the French population before and during the first and second COVID-19 lockdowns: latent class trajectory analyses using longitudinal data J. Affect. Disord. 309 2022 95 104 10.1016/j.jad.2022.04.095 35452759
McPherson K.E. McAloney-Kocaman K. McGlinchey E. Faeth P. Armour C. Longitudinal analysis of the UK COVID-19 Psychological Wellbeing Study: trajectories of anxiety, depression and COVID-19-related stress symptomology Psychiatry Res. 304 2021 114138 10.1016/j.psychres.2021.114138
Pierce M. McManus S. Hope H. Hotopf M. Ford T. Hatch S.L. John A. Kontopantelis E. Webb R.T. Wessely S. Abel K.M. Mental health responses to the COVID-19 pandemic: a latent class trajectory analysis using longitudinal UK data Lancet Psychiatry 8 2021 610 619 10.1016/S2215-0366(21)00151-6 33965057
Prati G. Mancini A.D. The psychological impact of COVID-19 pandemic lockdowns: a review and meta-analysis of longitudinal studies and natural experiments Psychol. Med. 51 2021 201 211 10.1017/S0033291721000015 33436130
Preti E. Pierro R.D. Perego G. Bottini M. Casini E. Ierardi E. Madeddu F. Mazzetti M. Riva Crugnola C. Taranto P. Mattei V.D. Short-term psychological consequences of the COVID-19 pandemic: results of the first wave of an ecological daily study in the Italian population Psychiatry Res. 305 2021 114206 10.1016/j.psychres.2021.114206
Reizer A. Geffen L. Koslowsky M. Life under the COVID-19 lockdown: on the relationship between intolerance of uncertainty and psychological distress Psychol. Trauma 13 2021 432 437 10.1037/tra0001012 33539162
Rettie H. Daniels J. Coping and tolerance of uncertainty: predictors and mediators of mental health during the COVID-19 pandemic Am. Psychol. 76 2021 427 437 10.1037/amp0000710 32744841
Richter D. Riedel-Heller S. Zürcher S.J. Mental health problems in the general population during and after the first lockdown phase due to the SARS-Cov-2 pandemic: rapid review of multi-wave studies Epidemiol. Psychiatr. Sci. 30 2021 e27 10.1017/S2045796021000160 33685551
Robinson E. Sutin A.R. Daly M. Jones A. A systematic review and meta-analysis of longitudinal cohort studies comparing mental health before versus during the COVID-19 pandemic in 2020 J. Affect. Disord. 296 2022 567 576 10.1016/j.jad.2021.09.098 34600966
Sahib A. Chen J. Cárdenas D. Calear A.L. Intolerance of uncertainty and emotion regulation: a meta-analytic and systematic review Clin. Psychol. Rev. 101 2023 102270 10.1016/j.cpr.2023.102270
Salanti G. Peter N. Tonia T. Holloway A. White I.R. Darwish L. Low N. Egger M. Haas A.D. Fazel S. Kessler R.C. Herrman H. Kieling C. De Quervain D.J.F. Vigod S.N. Patel V. Li T. Cuijpers P. Cipriani A. Furukawa T.A. MHCOVID Crowd Investigators The impact of the COVID-19 pandemic and associated control measures on the mental health of the general population: a systematic review and dose-response meta-analysis Ann. Intern. Med. 175 2022 1560 1571 10.7326/M22-1507 36252247
Soraci P. Ferrari A. Abbiati F.A. Del Fante E. De Pace R. Urso A. Griffiths M.D. Validation and psychometric evaluation of the Italian version of the fear of COVID-19 scale Int. J. Ment. Health Addict. 20 2020 1913 1922 10.1007/s11469-020-00277-1 32372892
Shevlin M. Butter S. McBride O. Murphy J. Gibson-Miller J. Hartman T.K. Levita L. Mason L. Martinez A.P. McKay R. Stocks T.V.A. Bennett K. Hyland P. Bentall R.P. Refuting the myth of a 'tsunami' of mental ill-health in populations affected by COVID-19: evidence that response to the pandemic is heterogeneous, not homogeneous Psychol. Med. 2021 1 9 10.1017/S0033291721001665
van De Schoot R. Sijbrandij M. Winter S.D. Depaoli S. Vermunt J.K. The GRoLTS-checklist: guidelines for reporting on latent trajectory studies Struct. Equ. Modeling 24 3 2017 451 467 10.1080/10705511.2016.1247646
Wu T. Jia X. Shi H. Niu J. Yin X. Xie J. Wang X. Prevalence of mental health problems during the COVID-19 pandemic: a systematic review and meta-analysis J. Affect. Disord. 281 2021 91 98 10.1016/j.jad.2020.11.117 33310451
Xu C. Xu Y. Xu S. Zhang Q. Liu X. Shao Y. Xu X. Peng L. Li M. Cognitive reappraisal and the association between perceived stress and anxiety symptoms in COVID-19 isolated people Front. Psychiatry 11 2020 858 10.3389/fpsyt.2020.00858 32982809
|
PMC010xxxxxx/PMC10205648.txt |
==== Front
Soc Sci Med
Soc Sci Med
Social Science & Medicine (1982)
0277-9536
1873-5347
Pergamon
S0277-9536(23)00330-1
10.1016/j.socscimed.2023.115973
115973
Article
Enhanced unemployment benefits, mental health, and substance use among low-income households during the COVID-19 pandemic
Jeong Soyun ∗
Fox Ashley M.
Department of Public Administration and Policy, Rockefeller College of Public Affairs & Policy, University at Albany, State University of New York, 1400 Washington Avenue, Albany, NY, 12222, USA
∗ Corresponding author.
24 5 2023
7 2023
24 5 2023
328 115973115973
17 7 2022
17 3 2023
15 5 2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objective
To buffer the economic impacts of the pandemic-induced economic downturns, the U.S. government passed major economic stimulus bills that provided cash payments to affected citizens and a large boost to unemployment benefits. We ask what impact these enhanced safety-net policies have had on mental health and stress-induced substance use among low-income Americans, especially enhanced unemployment insurance (UI) benefits, which constituted a large economic transfer to those eligible.
Methods
Using individual fixed effects analysis of a panel of nearly 900 low-income Americans since the start of the pandemic from the Understanding America Survey, we examine how receipt of enhanced unemployment benefits has impacted the mental health burden and substance use behaviors of low-income Americans. We additionally examine the buffering effect of a set of other safety-net measures (Stimulus, Medicaid, SNAP, TANF, housing assistance, EITC, WIC, and CHIP).
Results
We found that job loss, regardless of benefit receipt, was associated with increased stress and decreased average substance use, driven by reduced smoking when compared with those were employed. Yet, when factoring in UI receipt we see that receiving UI was associated with reduced stress, but no impact on depression or substance use. In contrast, those who did not receive UI experienced greater stress compared with those who were employed. Overall, we found that people who remained employed used substances more than people who were unemployed regardless of UI receipt with the exception of drinking.
Conclusions
We conclude that enhanced unemployment offset some of the negative mental health effects of the pandemic and did not increase routine substance use among the unemployed.
Keywords
Unemployment
Unemployment benefits
COVID-19
Mental health
Substance use
Handling editor: Blair T. Johnson
==== Body
pmc1 Introduction
The COVID-19 pandemic generated massive unemployment worldwide as more than 255 million full-time jobs were lost in 2020 (ILO, 2021). Previous research has found that job loss during difficult times yields many adverse consequences including negative mental and physical health effects (Henkel, 2011). Among the many negative impacts of job loss, we concentrate on its effects on mental health and substance use. Even prior to COVID-19, the US was experiencing a fatal drug use epidemic that some have tied to prolonged economic disenfranchisement, resulting in “deaths of despair” that have driven down life expectancy in certain US population groups (Case and Deaton, 2017; Ruhm, 2018). Case and Deaton (2017) define “deaths of despair” as drug overdose, alcohol-related illness and suicide that “… come with prolonged economic distress” (Deaton 2017, p. 3). The economic impacts of COVID-19 accrued particularly harshly to economically vulnerable households already living at the margins, sometimes referred to as the “precariat” (Standing, 2014) and has the potential to have worsened mental health and driven up substance use among economically vulnerable groups.
In response to the economic downturns, the federal government directed an extra $600 a week in jobless benefits to out-of-work Americans under the CARES Act while also suspending some of the usual administrative burden associated with receipt of Unemployment Insurance (UI) including mandated work requirements, and waiting periods to receive benefits. Claimants were also allowed to receive benefits for longer: the traditional 26-week maximum in most states was extended by 13–20 weeks. A parallel Pandemic Unemployment Program has been able to offer benefits to those not typically eligible for UI, such as gig economy workers (Karpman and Acs, 2020). Additionally, two stimulus checks, pandemic SNAP, maintenance of effort requirements under Medicaid and heightened access to other safety-nets have served to offset some of the wider effects of job loss and the economic downturn. Moreover, federal eviction moratoriums were in effect beginning from March 2020 through December 2020 to protect people from losing their residence and to reduce financial burdens. Parolin et al. (2021) estimates that as a consequence of the American Rescue Plan, the U.S. poverty rate fell to 8.5% from 12.8% in 2018, the lowest figure on record. The Census Bureau estimates that enhanced Unemployment Insurance benefits on its own lowered the overall poverty rate by 1.4 percentage points in 2020 decreasing poverty across all racial groups and all age groups (Chen and Shrider, 2021). Given this large and unprecedented economic transfer to low-wage workers, we ask to what extent these added unemployment benefits reduced unemployment-induced mental health and substance use by comparing low-income individuals who lost their jobs but did not receive enhanced UI and those who lost their jobs but did receive enhanced UI.
2 Background
Unemployment and mental health. Numerous studies have explored the relationship between unemployment and mental health and revealed that job loss has a negative impact on psychological well-being of the unemployed (Murphy and Athanasou, 1999; Paul and Moser, 2009; Drydakis, 2015; Lee et al., 2021). Several mechanisms can explain this association. First, employment termination creates economic deprivation which makes it difficult to obtain sufficient food, adequate housing, and other necessities. It thereby increases the sense of insecurity and stress (Janlert and Hammarström, 2009). Second, unemployment also causes people to experience lower self-esteem, isolation in society, and therefore a lack of social support (Mathers and Schofield, 2012). This can lead to higher levels of depression and anxiety among the unemployed. Men, blue-collar workers and individuals who have been unemployed for longer durations are particularly affected by the mental health effects of unemployment (Paul and Moser, 2009).
Other studies have examined how government programs can mitigate the adverse impact of unemployment on mental health. Regarding receipt of unemployment insurance specifically, studies have found that benefit receipt could reduce the risks of having symptoms of depression and anxiety (Rodriguez et al., 2001;; Berkowitz and Basu, 2021), suicide rates (Cylus et al., 2014), and opioid overdose mortality ( Wu and Evangelist, 2022). In Australia, a temporary income support to the unemployed lowered reported financial stress and mental distress (Botha et al., 2022). O'Campo et al. (2015) carried out a meta-analysis of 237 international studies on the relationship between unemployment and mental health and suggested two major findings. First, generous UI can improve mental health of the unemployed by offering financial security. However, generous UI might not fully amend the mental health of the unemployed because the increased economic security still does not completely offset the negative psychosocial impacts of job loss such as loss of the status and self-confidence.
Unemployment and substance use. Studies from past economic downturns have shown that job loss is likely to contribute to increased substance use as a coping strategy in response to economic stress (Lee et al., 2015). However, current research on the association between unemployment and substance use is mixed and unclear (Azagba et al., 2021; Lee et al., 2015; Compton et al., 2014). Overall, the prevalence of substance use, alcohol abuse and alcohol dependence are generally higher among unemployed people than among those who are employed (Henkel, 2011). However, fewer studies have employed experimental or quasi-experimental designs that are capable of teasing out the direction of causality. Economic downturns provide an opportunity to examine whether sudden and unexpected job loss among the usually more job secure population increases psychological distress, and in turn, drug use. A recent systematic review finds that drug use increases in times of recession because unemployment increases psychological distress, which increases drug use (Nagelhout et al., 2017). Few studies to date have specifically investigated the relationship between unemployment insurance benefit receipt and substance use.
The impact of unemployment during COVID-19. Given the known association between recessions and increases in substance use via psychological distress, there is good reason to believe that the pandemic induced unemployment may have increased substance abuse. First, the job loss experienced during the pandemic has been far deeper than previous downturns with an estimated 15% percent of jobs being lost in the few months subsequent to COVID-19 cases being identified in the U.S., compared with a previous maximum of 6% during the Great Recession (Groshen, 2020). Job losses were also more abrupt compared with past downturns. Moreover, the stay-at-home orders and anxiety from the pandemic itself is likely to have contributed to greater substance use as well as having compounding effects on people with substance use disorders. There is already suggestive evidence of increases in mental health burden, suicides and substance use during the pandemic (Gruber et al., 2021). During the first 3 weeks of lockdowns/stay-at-home restrictions, stress reactions were elevated relative to prior population with estimates ranging from 27 to 32% for depression, 30–46% for anxiety disorders, 15–18% for acute/post-traumatic stress, 25% for insomnia, and 18% for suicidal ideation. These prevalence estimates were 1.5–1.7 times higher for those who reported job loss due to COVID-19 restrictions than those who did not (Russell et al., 2020). Though preliminary evidence suggests that suicide did not increase above previous levels, firearm suicide accounted for 24,245 deaths in 2020 with a rate of 8.1 per 100,000 persons aged ≥10 years, which maintained the record high levels after steady increases leveled off in 2018 (Kegler et al., 2022). There is some evidence that overdose death rates have increased-more than 107,000 people in the U.S. died from drug overdoses in 2021, which is roughly a 15% increase from 2020 and the greatest number in a single calendar year (CDC, 2022). Although the deaths are not confined to the unemployed, this implies that substance use has escalated during the pandemic and may be compounded by job loss.
To an even greater extent than previous economic downturns, during the COVID-19 pandemic, the US government has adopted a series of counter-cyclical “stimulus” policies aimed at reducing the economic stress experienced by Americans and bolstering the economy against recession, which may serve to offset some of the increased stress associated with pandemic measures and the economic downturn (Cooney and Shaefer, 2021). Over the course of the pandemic the US has committed a total of $4.6 trillion towards COVID-19 relief and stimulus policies compared with $787 billion in deficit spending under the American Recovery and Reinvestment Act, during the Great Recession of 2008 (USAspending.gov, 2022; Congressional Budget Office, 2015). These policies have the potential to buffer against the economic stresses that contribute to mental health burden and substance use. According to the Department of Labor statistics, 58 million people in the US made initial claims for unemployment insurance benefits from April to December in 2020 (Department of Labor, 2022). A previous study found a negative association between UI receipt and smoking (Fu and Liu, 2019); however, another study found that receiving UI increases alcohol consumption (Lantis and Teahan, 2018). Previous studies have been limited in their ability to detect effects due to the relatively modest changes in employment and few studies have explicitly examined UI as a buffering social policy.
In contrast with previous emergencies, federal and state governments have endeavored to protect those who lose their job during the pandemic by expanding UI benefits and alleviating eligibility criteria and other requirements to an unprecedented degree during the pandemic. In three states, the generous enhanced unemployment benefits raised low-wage workers salaries well above the average weekly wage for those earning the federal minimum wage (Goodkind, 2021). In addition to enhanced UI, low-income workers, whether unemployed or not, may be eligible for a variety of other safety-net measures enacted during the pandemic. It is estimated that enhanced unemployment insurance and other forms of COVID-19 emergency funding adds up to 90% or more of the average weekly wage for those earning the federal minimum in nine states (Goodkind, 2021).
While the receipt of UI has the potential to offset economic stress, not everyone that might benefit from UI benefits was able to access them and those who did often faced a stressful application process. Long wait times and difficulty getting through to Unemployment Offices across the country have been well documented to have hindered access for many (Zipperer and Gould, 2020). In addition to these administrative burdens, for workers interacting with the UI system for the first time, the jargon and complexity involved in understanding whether they are eligible, how their benefits are calculated, and how much they can expect to get can be daunting, which may discourage uptake even among those eligible. Further, for those accessing traditional unemployment insurance, the result may be disappointing. Benefit levels for traditional UI are quite low, generally designed to replace no more than half of people's former wages in order to avoid disincentivizing returning to work.
Moreover, many essential workers remained in their minimum wage jobs and likely received less than they would have if they were eligible for enhanced unemployment benefits. The fact that many people were making more by not working than those who were working during the pandemic creates a particularly unique point of comparison between those who are eligible for enhanced UI and those who were not. It is also possible that receipt of UI during this stressful period of the pandemic could facilitate access to substances in ways that might exacerbate substance use as unemployed people use substances to cope with the stresses of stay-at-home orders and lack of employment. On the other hand, the economic security afforded by enhanced UI may reduce stress and reduce coping through substance-use.
Given conflicting theoretical and empirical findings on the relationship among job loss, unemployment/safety-net benefit receipt, mental health and substance use, we propose the following research questions: 1) First, is job loss during the pandemic associated with an increase in mental health and substance use burden? 2) Second, does receipt of unemployment insurance “buffer” the unemployed against negative mental health and substance use outcomes? 3) Third, what effect did participation in other safety-nets have on mental health and substance use burden? The first question is to clarify prior ambiguous research results about the effects of recessions on health and provide new evidence in the context of the current pandemic. The second and third question aims to evaluate the effectiveness of economic security policies, which have been generously implemented during the COVID-19 pandemic. To focus on those who were likely to benefit the most from economic security policies, we confine our analysis to low-income households whose annual income is less than $30,000 and most hard hit by the economic effects of the pandemic.
3 Methods
Data. We used data from the Understanding America Study (UAS) collected by the University of Southern California. During COVID-19, UAS launched a special survey, ‘Understanding Coronavirus in America.’ Since March 2020, UAS has accumulated information on the attitudes and behaviors of respondents around the pandemic including whether they have been working and what safety-net programs they have been accessing. UAS provides panel data, meaning that a subset of the same individuals has been followed over time and surveyed on a bi-weekly basis. The UAS is a nationally representative panel of American households that are randomly recruited from United States Postal Service delivery sequence files. Eligible participants are all adults aged 18 or older. The survey sends a pre-notification letter (both in English and Spanish) to randomly draw prospective participants and asks if they want to join UAS panel. The response rate of surveys ranges from 74% to 96% across waves. Apart from the nationally representative sample, the UAS also collects Los Angeles county sample which is a subset of the national data. We dropped the Los Angeles county sample for the analysis and only considered nationally representative sample.
For this study, we analyzed responses across the 19 waves of data collection from March to December 2020, a time when both lock-down measures and pandemic unemployment benefits were at their peak. This amounted to 3000 respondents in each wave. However, we limited our analysis to low-income households earning $30,000 a year or less in order to capture individuals for whom receipt UI and other safety-nets may be particularly impactful due to its very generous wage replacement and their likely income eligibility for means-tested programs. When we only consider low-income individuals, we have approximately 900 respondents across each of the 19 waves. The total number of person-wave observations is 16,986. Below we consider treatment of missing data and describe how we formed a balanced panel.
Treatment of Missing Data. We conducted missing data imputation to create a balanced panel to improve estimation and accuracy of results (Engels and Diehr, 2003; Donders et al., 2006). There are two types of missingness in panel data: within-wave and whole-wave missingness (Young and Johnson, 2015). The former one occurs when respondents did not answer specific questions or when questions are only asked in few waves. The latter missingness occurs when respondents did not participate in certain waves (e.g., attrition or study dropout). We considered both types of missingness for the imputation. When it comes to within-wave imputation, we imputed some of our key variables. First, UI/welfare receipt and control variables (stress, food insecurity, isolation/quarantine, and social connection) are missing in Wave 1 because these questions were not asked in the survey whereas key questions about employment status, depression, the use of cannabis, recreational drugs, and smoking were asked. For the missing variables, we replaced the missing values with the answers from the following wave (Wave 2). Since both waves are surveyed in the early stage of the pandemic, we assume that respondents’ status in Wave 2 (collected in April) can be a proxy for Wave 1 (collected in March). For example, people who received the benefit in wave 2 are likely to receive it in Wave 1 since it implies that they have a high chance of being unemployed even before the pandemic hit. Likewise, we can infer that people who said they did not receive the benefits in wave 2 are likely to not have received them in wave 1. Although some people experienced layoffs during Wave 1, it might take more than a month to receive benefits.
Second, welfare receipt questions are not asked in Wave 9 to all participants. We replaced the missing values using information from the previous and the following waves. If the answers are the same in these two waves, we input the consistent value. The rationale behind this imputation decision is that we link observations across waves in order to exploit as much as information as possible (Christelis, 2011). In case the information from the two waves is different, we coded them as they did not receive the benefits. Lastly, the cigarette and vaping variables are missing in Waves 1–3, but we did not impute them because we lacked information to predict the initial status (more information described in supplemental materials). In a similar vein, we imputed missing waves by using the previous and subsequent waves.
Dependent variables. For our main outcome variable, we examined two mental health indicators (depression and stress) and five substance use categories in the dataset (cannabis, recreational drugs, drinking, cigarettes, and vaping). Mental health was measured using two validated scales. First, to measure depression, the survey used the Patient Health Questionnaire (PHQ-4). The PHQ-4 includes four questions about mental illness (anxiety, worrying, depression, and little interest) and ask respondents whether they experienced them past fourteen days with the following four answer choices: ‘0 = nearly every day’ ‘1 = more than half the days’ ‘2 = several days’ and ‘3 = not at all’. We aggregated the four categories and created anxiety/depression variable which ranges from 0 to 12. Second, the Perceived Stress Scale (PSS-4) measures an individual's perceived stress levels with four items (confidence in handling personal problems, unable to control, things are going your way, and difficulties are piled up). The questionnaires ask how often respondents feel stressed and the answer choices include: ‘1 = never’ ‘2 = almost never’ ‘3 = sometimes’ ‘4 = fairly often’ and ‘5 = very often.’ We reverse coded the ‘confidence’ and ‘things are going your way’ items to make their direction consistent with the other two items. We aggregated the four categories and created a perceived stress variable which ranges from 0 to 16 in line with prior studies (Cohen et al., 1983). A higher number represents increased depression and stress.
Regarding substance use measures, respondents were asked to report the number of days that they consumed each substance item over the last seven days. The original question is, “Out of the past seven days, what is your best estimate of the number of days that you did each of the following activities?” and the answers are ranged from 0 to 7 days. Respondents were asked about five substances including cannabis, recreational drugs, drinking, cigarettes, and vaping. In regard to recreational drugs, respondents were asked to provide an estimate of how many days that they used recreational drugs other than alcohol and cannabis products. This is a general frequency measure that is widely used in surveys that measure drug use (EMCDDA, 2002). However, this question did not provide specific drug types (e.g., cocaine, opioid or other synthetic drugs) so we are unable to disaggregate these specific types of substances from this overall measure. We combined cigarettes and vaping variables and created a smoking variable. We also generated an average substance use category that aggregated the five different substances to show an overall measure of substance use.
Overall, we considered seven outcome variables – depression, stress, average substance use, cannabis, recreational drugs, drinking, and smoking and tested each category of substances as different outcomes in separate models. We treated all dependent variables as continuous.
Independent variables. The first explanatory variable is the respondent's employment status. At each wave assessed every two weeks, respondents reported their current employment status at the time they were surveyed (employed, unemployed-layoff, unemployed-looking for work, on sick or other types of leave, retired and not in labor force). People who reported being retired, on leave, or not in labor force were excluded from the sample. We treated people who experienced reduced work hours as employed as UI benefits may not be available to those whose hours have been reduced. The second explanatory variable was receipt of unemployment benefits. The survey asks respondents whether they have received unemployment insurance benefits in the past fourteen days.
In order to examine the moderating effect of UI receipt on mental health and substance use outcomes, we constructed a group variable to classify people into three different categories of respondents: 1) 1 = Employed individuals who did not receive UI benefits (57% of observations); 2) 2 = Unemployed individuals who received UI benefits (7% of observations); and 3) 3 = Unemployed individuals who did not receive UI benefits (35% of observations). The groups are not permanent for individuals and thus they can belong to different groups at different time points. Thiry-percent of the respondents experienced at least one shift in their employment status either from employed to unemployed or vice versa (see supplementary material; appendix 5). The employed group was used as the reference group in the analysis. A small number of respondents reported being employed and receiving UI (less than 2% of total observations), potentially due to delays in UI receipt. These individuals were dropped from the analysis.
Intervening and Control variables. As the analysis employs individual fixed effects (see “Analyses” below), we did not include time invariant demographic characteristics such as gender, race, ethnicity, education, or age. In individual fixed-effects analysis, an individual operates as their own control. While our primary focus was on the potential buffering effects of UI given the large income boost this constituted for eligible low-income workers, we were also interested in the effects of other safety-nets in potentially offsetting pandemic induced mental health burden and substance use. We therefore included a set of additional safety-net measures in our models (Medicaid, economic stimulus check, TANF, SNAP, WIC, CHIP, EITC, housing assistance, and having health insurance) that a low-income household might have benefited from other than UI. If respondents received those programs, the variable is coded as 1 and 0 otherwise.
In an additional set of models, we treat mental health (depression and stress), isolation/quarantine (due to COVID-19 exposure), social connection and economic/food insecurity as additional explanatory/control variables. As we hypothesize that individuals experiencing depression or stress over their economic situation may contribute to increased use of substances, we examine the independent contribution of these factors to substance use. We examine to what extent entering mental health/stress into the model ‘soaks up’ some of the explanatory relationships between the UI receipt and substance use. Lastly, we also controlled for isolation/quarantine and social connection with friends or family as these factors may independently contribute to mental health and substance use outcomes separate from receipt of safety-nets/economic insecurity and were experienced to different degrees across individuals.
Analyses. We ran two-way fixed-effects models (time and individual) to estimate the association between unemployment benefit receipt and mental health/substance use focusing on variations in within-individuals. We adjusted the analysis with clustered standard errors by using states as clusters to consider varying state circumstances during the pandemic. The fixed-effects model is useful in that the model controls time-invariant characteristics in individuals such as race, gender, age, and education (Williams, 2015). Two-way fixed effects with a policy variable (in this case, UI receipt among those who are unemployed) produces results that are similar to a difference-in-difference approach (De Chaisemartin & D'Haultfoeuille, 2022). We present analyses in a stepwise manner using seven models to assess the impact of unemployment and UI on mental health and substance use. We generate margins plots to examine the interactive effects of UI receipt across these groups. All analysis was conducted using Stata version 16.
Robustness checks. We ran a number of sensitivity analyses and robustness checks to ensure that our results are sound (see supplemental materials). This included running random-effects models rather than fixed effects and running analysis without imputation of missing variables. Results were largely consistent with our main results presented in the paper. We also attempted to dichotomize the dependent variables and run logistic regression. However, due to the relatively low variation in mental health and substance use behaviors over time when dichotomized, fixed effects models were not possible with binary outcomes or ordered logit models.
4 Results
Descriptive statistics.Fig. 1, Fig. 2 show the mean of the main outcome variables across waves. We observe a large increase in mental health and substance use burden between March and May of 2020 constituting the first 3 months after the first case of COVID-19 was identified in the US. The average mental health and substance use level off after the high spike in the early pandemic. While the stress measure was not available in the first wave and the first three waves for smoking are missing, we can assume that these patterns are similar to other outcome measures and constitute an increase from prior trends.Fig. 1 Average mental health of the sample during 2020
Notes: (1) Wave 1 is missing for the stress variable. (2) A higher score for Depression and Stress constitutes worse mental health outcomes.
Fig. 1
Fig. 2 Average substance use behaviors of the sample during 2020
Notes: (1) Waves 1–3 are missing for the smoking variable. (2) Substance use represents the mean number of days used substances past week.
Fig. 2
Fig. 3, Fig. 4 present the proportion of respondents who experience elevated depressive symptoms, stress and substance use. For PHQ-4 anxiety and depression measurement, a score of 3 or greater is considered as an empirical cut off for detecting potential depression that needs mental health referral. The cutoff point of PSS-4 is a score of 6. Among our final sample (16,986 observations), 36.82% of the observations are likely to have higher depressive symptoms and 55.99% of higher stress. When it comes to substance use, respondents exhibit different consumption patterns. Only 6% of respondents said they used recreational drugs at least one day during the past week. On the contrary, 32.2% of respondents consumed alcohol and 31.4% smoked at least one day a week. 18.5% of respondents reported that they used cannabis products. On average, 55% of respondents consumed at least one type of substances 1 or more days.Fig. 3 Proportion of respondents experienced depressive symptoms and stress. Notes: (1) Depression variable has a scale of 0–12. A score of 3 or greater is considered as an empirical cut off for detecting potential depression. (2) Stress variable has a scale of 0–16. A score of 6 or greater is considered as an empirical cut off for detecting potential stress.
Fig. 3
Fig. 4 Proportion of respondents experienced substance use
Notes: (1) Average substance use is an aggregated measure of the five difference substances. (2) All substance use variables have a scale of 0–7.
Fig. 4
Fig. 5 demonstrates the sharp increase in people reporting that they were laid off beginning in March 2020 and the decline in individuals who are employed. The unemployed or laid off population before the pandemic was around 7% rising to almost 57% of the respondents by April before gradually decreasing to around 40% over time. While employment recovers over time, it never returns to pre-pandemic levels. Fig. 6 illustrates the trends in unemployment benefit receipt during the pandemic. On average, 30% of unemployed people received UI benefits while the remaining 70% did not report receiving benefits.Fig. 5 Trends of unemployment before and during the COVID-19. Notes: We merged pre-COVID data from Understanding America Study database to demonstrate pre-COVID unemployment status.
Fig. 5
Fig. 6 Trends of unemployment insurance (UI) benefit receipt
Notes: The take-up rate was calculated by dividing the number of recipients by the number of total unemployed respondents. The left Y-axis shows the number of the unemployed and UI recipients, and the right Y-axis shows the take-up rate.
Fig. 6
Fixed-Effects Analysis. To answer the first research question about the association of job loss on outcomes, we examined how job loss solely affects mental health and substance use adjusting for UI receipt. We found that people who were unemployed or experienced a lay off during the pandemic have higher levels of stress than people who were consistently employed while there is no significant association with depressive symptoms (see Table 1). When it comes to substance use, the unemployed appeared to spend less time using substances than their employed counterparts and this is largely due to reduced smoking. The unemployed smoked 0.132 days (95% CI: −0.25, −0.05) less than the employed. Other substances such as cannabis, recreational drugs, and alcohol did not have meaningful relationship with unemployment. Those receiving UI experienced lower stress, but UI receipt increases smoking. Including other safety-nets in the model diminished the effect size of relationship between job loss and mental health outcomes as well as for UI receipt, but did not fully mediate the relationship.Table 1 Study results (Unemployment and unemployment insurance (UI) benefit as independent predictors) – Fixed-effects model responding to Research question 1.
Table 1
VARIABLES Unemployment and UI as predictors With all controls
(1) (2) (3) (4) (5) (6) (7) (1) (2) (3) (4) (5) (6) (7)
Depression Stress Average
Substance Cannabis Drugs Drinking Smoking Depression Stress Average
Substance Cannabis Drugs Drinking Smoking
Unemployed 0.278 0.424** −0.070* −0.060 −0.035 −0.012 −0.132*** 0.185 0.337* −0.076** −0.050 −0.072 −0.040 −0.132***
(0.203) (0.162) (0.038) (0.075) (0.074) (0.109) (0.044) (0.192) (0.183) (0.035) (0.069) (0.066) (0.105) (0.045)
UI benefits −0.362 −0.845*** 0.071 0.013 −0.021 0.112 0.130** −0.301 −0.785*** 0.076 0.013 0.013 0.138 0.126**
(0.259) (0.172) (0.058) (0.122) (0.074) (0.153) (0.054) (0.243) (0.166) (0.057) (0.126) (0.073) (0.152) (0.053)
Medicaid 0.133 −0.055 0.012 0.062 0.099 −0.032 −0.022
(0.116) (0.149) (0.034) (0.054) (0.064) (0.052) (0.064)
Economic stimulus funds 0.143 0.083 −0.027 0.003 −0.011 −0.047 −0.016
(0.096) (0.089) (0.030) (0.033) (0.049) (0.044) (0.026)
TANF 1.033** −0.064 0.071 0.211 −0.291* 0.113 0.246
(0.429) (0.489) (0.109) (0.274) (0.153) (0.271) (0.163)
SNAP 0.341* −0.014 −0.030 −0.055 0.018 0.035 −0.054
(0.175) (0.142) (0.059) (0.086) (0.084) (0.063) (0.076)
EITC 0.295 −0.261 −0.001 −0.102 0.272 −0.076 −0.059
(0.309) (0.322) (0.054) (0.119) (0.216) (0.154) (0.086)
WIC 0.189 −0.211 −0.228** −0.240 −0.158 −0.215 −0.277**
(0.273) (0.254) (0.109) (0.149) (0.196) (0.136) (0.110)
CHIP −0.052 −0.156 0.115 0.307** 0.089 0.312* −0.015
(0.273) (0.373) (0.082) (0.130) (0.139) (0.181) (0.117)
Housing −0.292 −0.002 0.079 −0.015 0.101 0.334*** −0.026
(0.365) (0.230) (0.071) (0.202) (0.080) (0.117) (0.094)
Health insurance 0.228 0.064 −0.105* −0.003 −0.095 −0.050 −0.092
(0.177) (0.172) (0.056) (0.107) (0.124) (0.100) (0.065)
Stress −0.002 0.006 0.005 −0.002 0.002
(0.005) (0.007) (0.009) (0.009) (0.008)
Depression 0.022** 0.022*** 0.030* 0.044*** 0.014
(0.008) (0.008) (0.015) (0.013) (0.011)
Food insecurity 0.710*** 0.357* 0.002 −0.028 0.039 0.078 −0.053
(0.139) (0.189) (0.041) (0.077) (0.083) (0.064) (0.048)
Isolation 0.325 −0.176 −0.008 −0.019 −0.113 −0.030 0.098
(0.265) (0.172) (0.054) (0.095) (0.110) (0.068) (0.068)
Social connection 0.018 −0.015 0.041*** 0.049*** 0.035*** 0.037*** 0.043***
(0.022) (0.018) (0.007) (0.013) (0.008) (0.012) (0.009)
Economic insecurity 0.007*** 0.008*** 0.000 −0.001 0.001 0.002** −0.000
(0.002) (0.002) (0.001) (0.001) (0.001) (0.001) (0.001)
Eviction protection 0.458* −0.086 0.095 0.267 0.053 0.066 0.085
(0.268) (0.223) (0.080) (0.174) (0.145) (0.164) (0.069)
Observations 16,986 16,986 16,644 16,986 16,986 16,986 16,644 16,986 16,986 16,644 16,986 16,986 16,986 16,644
R2 0.010 0.011 0.007 0.001 0.004 0.007 0.003 0.038 0.022 0.041 0.019 0.021 0.026 0.027
Number of individuals 894 894 876 894 894 894 876 894 894 876 894 894 894 876
Note: (1) Robust standard errors in parentheses; ***p < 0.01, **p < 0.05, *p < 0.1 (2) Time fixed effects are included in the model but did not show in the table.
Table 2 shows the main results of two-way fixed-effects model with clustered standard errors on the interactive effect of job loss and UI receipt, responding to the second research question. The first model was run with just a group variable for UI receipt compared with those who remained employed. The second model includes the other safety-net programs and the third model includes additional controls including food/economic insecurity, isolation/quarantine, mental health and social connection.Table 2 Study results (Interacting effects of unemployment and UI)– Fixed-effects model responding to Research question 2.
Table 2Model 1: No controls Model 2: Safety-nets Model 3: All controls
Variables (1)
Depression (2)
Stress (3)
Ave
Sub use (4)
Cannabis (5)
Drugs (6)
Drinking (7)
Smoking (1)
Depression (2)
Stress (3)
Ave
Sub use (4)
Cannabis (5)
Drugs (6)
Drinking (7)
Smoking (1)
Depression (2)
Stress (3)
Ave
Sub use (4)
Cannabis (5)
Drugs (6)
Drinking (7)
Smoking
Employed (no UI) (ref) Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref Ref
Unemployed#UI −0.084 −0.421** 0.001 −0.047 −0.056 0.100 −0.002 −0.111 −0.420** −0.008 −0.058 −0.068 0.096 −0.011 −0.117 −0.448** 0.000 −0.037 −0.057 0.099 −0.007
(0.258) (0.199) (0.057) (0.096) (0.069) (0.162) (0.060) (0.254) (0.201) (0.052) (0.092) (0.070) (0.159) (0.055) (0.238) (0.220) (0.056) (0.100) (0.077) (0.152) (0.059)
Unemployed#No UI 0.278 0.424** −0.070* −0.060 −0.035 −0.012 −0.132*** 0.254 0.432** −0.071* −0.061 −0.050 −0.004 −0.130*** 0.185 0.336* −0.076** −0.049 −0.069 −0.038 −0.133***
(0.203) (0.162) (0.038) (0.075) (0.074) (0.109) (0.044) (0.196) (0.161) (0.037) (0.075) (0.065) (0.106) (0.042) (0.191) (0.182) (0.035) (0.070) (0.065) (0.105) (0.045)
Medicaid 0.123 −0.077 0.020 0.071 0.102 −0.027 −0.012 0.130 −0.055 0.012 0.061 0.100 −0.032 −0.023
(0.118) (0.146) (0.034) (0.058) (0.066) (0.053) (0.063) (0.116) (0.149) (0.034) (0.055) (0.063) (0.052) (0.063)
Stimulus
Check 0.145 0.076 −0.017 0.016 −0.002 −0.036 −0.006 0.146 0.082 −0.026 0.005 −0.010 −0.047 −0.016
(0.101) (0.091) (0.031) (0.035) (0.049) (0.045) (0.027) (0.096) (0.089) (0.030) (0.033) (0.049) (0.045) (0.026)
TANF 0.989** −0.109 0.091 0.242 −0.267* 0.143 0.256 1.033** −0.063 0.070 0.210 −0.293* 0.112 0.247
(0.455) (0.503) (0.105) (0.272) (0.154) (0.285) (0.158) (0.424) (0.490) (0.109) (0.276) (0.153) (0.272) (0.162)
SNAP 0.288 −0.037 −0.026 −0.055 0.022 0.041 −0.048 0.329* −0.012 −0.032 −0.062 0.019 0.035 −0.057
(0.175) (0.144) (0.062) (0.088) (0.084) (0.065) (0.080) (0.178) (0.142) (0.060) (0.088) (0.084) (0.064) (0.078)
EITC 0.362 −0.228 0.014 −0.094 0.287 −0.044 −0.047 0.303 −0.265 0.002 −0.096 0.276 −0.073 −0.060
(0.300) (0.310) (0.058) (0.120) (0.227) (0.150) (0.086) (0.309) (0.323) (0.054) (0.118) (0.217) (0.154) (0.086)
WIC 0.240 −0.160 −0.228** −0.242 −0.149 −0.199 −0.290** 0.208 −0.214 −0.224** −0.229 −0.158 −0.213 −0.273**
(0.266) (0.264) (0.111) (0.154) (0.198) (0.138) (0.111) (0.271) (0.254) (0.110) (0.153) (0.198) (0.136) (0.109)
CHIP −0.046 −0.161 0.124 0.320** 0.094 0.319* −0.005 −0.043 −0.157 0.116 0.312** 0.088 0.313* −0.012
(0.280) (0.376) (0.079) (0.132) (0.133) (0.183) (0.118) (0.276) (0.373) (0.082) (0.130) (0.138) (0.182) (0.117)
Housing Assistance −0.250 0.039 0.069 −0.034 0.093 0.331** −0.031 −0.284 −0.006 0.081 −0.009 0.107 0.338*** −0.027
(0.376) (0.218) (0.071) (0.204) (0.087) (0.130) (0.081) (0.368) (0.231) (0.071) (0.202) (0.081) (0.117) (0.091)
Health
Insurance 0.238 0.078 −0.101 −0.001 −0.086 −0.039 −0.093 0.226 0.065 −0.106* −0.005 −0.096 −0.051 −0.092
(0.177) (0.172) (0.061) (0.111) (0.126) (0.106) (0.068) (0.177) (0.172) (0.056) (0.107) (0.124) (0.100) (0.065)
Stress −0.002 0.006 0.006 −0.002 0.001
(0.005) (0.007) (0.009) (0.009) (0.008)
Depression 0.023*** 0.023*** 0.029* 0.044*** 0.014
(0.008) (0.007) (0.015) (0.013) (0.011)
Food
Insecurity 0.711*** 0.356* 0.002 −0.027 0.040 0.079 −0.054
(0.139) (0.189) (0.041) (0.077) (0.084) (0.064) (0.048)
Isolation 0.328 −0.177 −0.007 −0.017 −0.112 −0.029 0.098
(0.262) (0.172) (0.054) (0.093) (0.109) (0.068) (0.068)
Social
Connection 0.018 −0.015 0.041*** 0.049*** 0.034*** 0.037*** 0.043***
(0.022) (0.018) (0.007) (0.013) (0.008) (0.011) (0.009)
Economic
Insecurity 0.007*** 0.008*** 0.000 −0.001 0.001 0.002** −0.000
(0.002) (0.002) (0.001) (0.001) (0.001) (0.001) (0.001)
Eviction Protection 0.246* −0.102 0.073* 0.175* 0.144** 0.099 −0.021
(0.139) (0.128) (0.041) (0.094) (0.068) (0.080) (0.037)
Observations 16,986 16,986 16,644 16,986 16,986 16,986 16,644 16,986 16,986 16,644 16,986 16,986 16,986 16,644 16,986 16,986 16,644 16,986 16,986 16,986 16,644
R2 (within) 0.010 0.011 0.007 0.001 0.004 0.007 0.003 0.018 0.012 0.013 0.005 0.009 0.011 0.010 0.037 0.023 0.041 0.018 0.022 0.026 0.026
Number of individuals 894 894 876 894 894 894 876 894 894 876 894 894 894 876 894 894 876 894 894 894 876
Note: (1) Robust standard errors in parentheses; ***p < 0.01, **p < 0.05, *p < 0.1 (2) Time fixed effects are included in the model but did not show in the table.
We find that, across the models, people who lost their jobs and received unemployment benefits experienced reduced stress compared with people who remained employed. Conversely, people who lost their job but did not receive UI benefits experienced heightened stress in equal and opposite proportion. Specifically, individuals who experienced a job loss and received UI benefits during the pandemic reported having lower stress symptoms by 0.448 points on the stress scale compared with individuals who maintained their employment status. Individuals who lost their job but did not have UI benefits displayed more frequent stress than the employed group by 0.336 points (β = 0.336; 95% CI: −0.03, 0.70; P < 0.01) (see Table 2, Model 3).
Pertaining to substance use outcomes, only smoking showed a statistically significant effect among the unemployed who did not receive UI benefits. This group reported smoking less than those who had a job by 0.133 days (95% CI: −0.22, −0.04). Unemployed people who received UI experienced no change in substance use behaviors compared with the employed group.
The margins plots in Fig. 7, Fig. 8 illustrate these findings visually and allow us to have more straightforward comparison across groups, including those who remained employed. The margins plot shows mental health and substance use of three group categories; employed, unemployed who received the UI benefit, and unemployed who did not receive the UI benefit. Although not statistically significant, the plots show that UI receipt is associated with reduced depression compared with both those who remained employed and those who lost their jobs but did not receive benefits. We found a similar pattern for stress with statistically significant effects.Fig. 7 Margins plot (1) – mental health outcomes
Notes: (1) In the X-axis, three groups are displayed for comparison of marginal effects. (2) The depression variable has a scale of 0–12 and the stress variable has scale of 0–16.
Fig. 7
Fig. 8 Margins plot (2) – substance use outcomes
Notes: (1) In the X-axis, three groups are displayed for comparison of marginal effects. (2) All substance use variables range from 0 to 7.
Fig. 8
Fig. 8 displays the three groups’ substance use behaviors. As we mentioned, we found that smoking is the only outcome that has statistically significant effects among all substance use measures. However, the margins plot reveals that there are only minimal differences among groups in terms of substance consumption. When we count the marginal differences between UI recipients and non-recipients, they were an additional 0.07 days of average substance use, 0.01 in cannabis, 0.05 in recreational drug, 0.10 in drinking, and 0.01 in smoking.
In the third research question, we asked whether receipt of other safety-net programs were associated with improved mental health and reduced substance use and we found mixed results. TANF receipt was associated increased depressive symptoms over time compared with non-beneficiaries. Medicaid enrollment was associated with spending more time using cannabis and recreational drugs whereas participating in the WIC program was associated with reduced smoking. Receiving housing assistance was associated to increased spending in recreational drugs and drinking. Finally, possessing any health insurance reduced average substance use (see Table 2, Model 3).
5 Discussion
In this study, we asked three research questions: First, we asked whether job loss was associated with an increase in mental health and substance use burden during the pandemic. Second, we examined whether the receipt of unemployment insurance was associated with improved mental health outcomes and reduced use of substances among the unemployed compared with low-income workers who remained employed during the pandemic. Third, we examined the impact of other safety-nets receipt and economic/food insecurity on mental health and substance use outcomes.
We found that job loss, regardless of benefit receipt, was associated with increased stress and decreased average substance use, driven by reduced smoking when compared with those were employed. However, when factoring in UI receipt we see that receiving UI was associated with reduced stress, whereas those who did not receive UI experienced greater stress compared with those who were employed. Depression symptoms did not differ significantly across the groups. We also did not find that UI reduced substance use among the unemployed. In fact, we found some suggestive evidence that drinking was modestly higher among unemployed people who received UI and that unemployed individuals who did not receive UI smoked modestly less than those who were employed. Overall, we found that people who remained employed used substances more than people who were unemployed regardless of UI receipt with the exception of drinking.
Although existing literature have found UI benefit can help reducing depression (Rodriguez et al., 2001; Berkowitz and Basu, 2021), our study only confirms that UI benefit cuts down on stress levels. Clinical depression is more severe than stress and stress is oftentimes considered as a trigger of depression (Hammen, 2005). Specifically, chronic stress can develop depression over a longer term (Tafet and Bernardini, 2003). Given that depression is more intense and long-lasting, UI benefit might be insufficient to fully buffer against serious mental health issues and cannot address the status loss associated with job loss (O'Campo et al., 2015).
Previous research has found that unemployment increases the use of substances (Lee et al., 2015), but less is known about how receiving UI benefits affect the level of substance use. Previous studies have also shown a potentially bidirectional relationship between UI benefits and substance use whereby receiving UI can contribute to more substance use by providing more financial resources or receiving UI can reduce substance use by mitigating stress (Fu and Liu, 2019; Lantis and Teahan, 2018). While we expected UI could reduce the use of substances by alleviating the financial stress of job loss (Cylus and Avendano, 2017), it is also possible that having more economic resources during a stressful period can facilitate substance use due to enhanced purchasing power (Delva et al., 2000; Baigi et al., 2008). Evans & Popova (2017) find that alcohol and tobacco can be considered normal goods, in which added income can raise the consumption of the goods. We found that receipt of certain safety-nets (TANF, housing assistance and Medicaid) was associated with higher depressive symptoms and substance use over time whereas receiving WIC decreased smoking. Reduced smoking among WIC recipients may be due to smoking cessation programs or increased efforts to quit during pregnancy rather than its economic buffering effects.
Overall, the effect sizes were more modest than we might have expected given the large temporary increase in income afforded to those who received enhanced unemployment benefits and enhanced stress early in the pandemic. The results are also surprising given reports suggesting that drug overdose death escalated during the pandemic (Panchal et al., 2021; CDC, 2020) and more frequent drug use among people who use substances (Busse et al., 2021). It could be that although drug overdose deaths were exacerbated among habitual drug users, it might not be hold for occasional substance users. A large majority of our sample reported less than daily substance use across all categories of substances. Lower substance use could also be a product of the social aspects of substance use. The social distancing measures adopted at the outset of the COVID-19 pandemic inhibited social gatherings, potentially reducing the use of cannabis and other drugs often consumed in recreational settings (Barratt and Aldridge, 2020). Nevertheless, we did find that UI receipt was associated with significant reductions in stress among the unemployed and non-receipt of UI with heightened stress compared with those who remained employed. We also did not find evidence that receiving UI significantly increased substance use due to heightened disposable income as some may have feared. Substance use among low-income workers overall was not concerningly high though we did detect spikes in depression/stress and substance use at the outset of the pandemic.
Overall, the descriptive statistics revealed concerning levels of depression and stress in this economically vulnerable population with nearly 40% of the sample experiencing potential depression and nearly 60% reporting elevated stress levels. We also observed a large increase in adverse mental health and substance use between March and May of 2020 during the early phase of the pandemic. These trends correspond with the nationwide strict lockdown initiated in April 2020. Previous studies have shown that early stay-at-home orders were associated with heightened mental health problems, with the strongest effects in certain subgroups such as women and women in couples with children (Adams-Prassl et al., 2022; Butterworth et al., 2022). Also, there is empirical evidence that negative mental health effects was the most common triggers for increased substance use in the early stages of the pandemic (Roberts et al., 2021). The subsequent declines are likely due to people developing better understanding of COVID-19 and coping skills over time.
Limitations. It is possible that our models are limited in their ability to detect how changes in employment status and UI receipt affected outcomes for several reasons. First, as Fig. 1, Fig. 2 demonstrate, there was a fairly universal increase in depression, stress and substance use at the beginning of the pandemic, which dissipated more as the pandemic continued. It is possible that the Fixed-Effects models results are washed out by that early effect whereas receipt of safety-net policies took longer to occur. It may be difficult to isolate the impact of enhanced safety-net policies on individuals as we are not accounting for overall household receipt of safety-nets. In other words, we cannot capture whether someone in the household other than the individual being interviewed received enhanced UI or other safety-nets. Thus, it is possible that more low-income individuals indirectly benefited from these programs than are being captured in our analysis. Second, there is low within-individual transition in employment status and UI benefit receipt (Appendix 5): 70% of respondents did not experience any changes in employment status and 84% did not show transitions in UI benefit receipt. As fixed-effects models draw their power from within-individual changes, there should be enough within-individual variations to pick up significant effects. Although we have some variations, the small proportion of individuals experiencing transitions might not fully capture the actual effects of UI benefit receipt and thus we need to careful about generalizing the main results. However, the random effects model which utilize both within and between-individual changes also show consistent results that confirm our findings.
We were also limited by the presence of missing data. We have missing observations in the earlier waves which can be critical given the impacts of COVID-19 immediately affected individuals. Although we imputed the missing values to build a complete dataset, we cannot rule out potential bias in estimates. We may have also been underpowered to detect modest size effects. The take-up rate of UI was concerningly low (30% of workers who lost their jobs or were unemployed). Strengths of the study include panel data allowing for individual fixed-effects and individual-level data on benefit receipt.
6 Conclusions
Federal and state governments have endeavored to protect workers who abruptly lost their jobs during the pandemic by offering generous UI benefits; our findings suggest that these efforts partially succeeded. We found that job loss among low-income workers was associated with increased stress levels during the pandemic, but that UI receipt buffered these effects. However, the effects on substance use were more mixed as we found no difference in substance use between those who were employed during the pandemic and those who were unemployed but received UI, but reduced daily smoking rates in those who did not receive benefits. However, even this large transfer might not be enough to cushion all the adverse impacts of the pandemic that affect various aspects of people's lives in the longer term. Overall, we find that depression and substance use spiked early in the pandemic and dissipated as time went on. Our data also showed that only 30% of the unemployed population had received the UI benefits on average in 2020. In future research, we want to investigate why UI benefits did not reach out to the rest 70% of the unemployed and how it might affect our results. Although the social benefits were available for eligible citizens, if they could not access them because of the complex administrative process and difficulty in learning about the program, it might have caused additional stress and anxiety to people.
Credit Author Statement
Soyun Jeong: Formal analysis, Data curation, Methodology, Visualization, Writing – original draft, Reviewing and editing, Ashley M. Fox: Conceptualization, Formal analysis, Methodology, Writing – original draft, Reviewing and editing
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.
Appendix A Supplementary data
The following is the Supplementary data to this article.Multimedia component 1
Multimedia component 1
Data availability
Data will be made available on request.
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.socscimed.2023.115973.
==== Refs
References
Adams-Prassl A. Boneva T. Golin M. Rauh C. The impact of the coronavirus lockdown on mental health: evidence from the United States Econ. Pol. 37 109 2022 139 155
Azagba S. Shan L. Qeadan F. Wolfson M. Unemployment rate, opioids misuse and other substance abuse: quasi-experimental evidence from treatment admissions data BMC Psychiatr. 21 1 2021 1 9
Baigi A. Lindgren E.C. Starrin B. Bergh H. In the shadow of the welfare society ill-health and symptoms, psychological exposure and lifestyle habits among social security recipients: a national survey study Biopsychosoc. Med. 2 1 2008 1 7 18194547
Barratt M.J. Aldridge J. No magic pocket: buying and selling on drug cryptomarkets in response to the COVID-19 pandemic and social restrictions Int. J. Drug Pol. 83 2020 102894
Berkowitz S.A. Basu S. Unemployment insurance, health-related social needs, health care access, and mental health during the COVID-19 pandemic JAMA internal medicine 181 5 2021 699 702 33252615
Botha F. Butterworth P. Wilkins R. Protecting mental health during periods of financial stress: Evidence from the Australian Coronavirus Supplement income support payment Social Science & Medicine 306 2022 115158 35751987
Busse H. Buck C. Stock C. Zeeb H. Pischke C.R. Fialho P.M.M. …Helmer S.M. Engagement in health risk behaviours before and during the COVID-19 pandemic in German university students: results of a cross-sectional study International Journal of Environmental Research and Public Health 18 4 2021 1410 33546344
Butterworth P. Schurer S. Trinh T.A. Vera-Toscano E. Wooden M. Effect of lockdown on mental health in Australia: evidence from a natural experiment analysing a longitudinal probability sample survey Lancet Public Health 7 5 2022 e427 e436 35461593
Case A. Deaton A. Mortality and morbidity in the 21st century Brookings Pap. Econ. Activ. 2017 397 2017
CDC Overdose Deaths Accelerating during COVID-19 2020 Centers for Disease Control and Prevention https://www.cdc.gov/media/releases/2020/p1218-overdose-deaths-covid-19.html
CDC Provisional Drug Overdose Death Counts 2022 Centers for Disease Control and Prevention https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm?mod=ANLink
Chen F. Shrider E.M. Expanded Unemployment Insurance Benefits during Pandemic Lowered Poverty Rates across All Racial Groups 2021 U.S. Census Bureau https://www.census.gov/library/stories/2021/09/did-unemployment-insurance-lower-official-poverty-rates-in-2020.html
Christelis, D., 2011. Imputation of missing data in waves 1 and 2 of SHARE. Available at SSRN: https://ssrn.com/abstract=1788248.
Cohen S. Kamarck T. Mermelstein R. A global measure of perceived stress J. Health Soc. Behav. 1983 385 396 6668417
Congressional Budget Office Estimated Impact of the American Recovery and Reinvestment Act on Employment and Economic Output in 2014 2015 https://www.cbo.gov/publication/49958
Cooney P. Shaefer H.L. Material Hardship and Mental Health Following the COVID-19 Relief Bill and American Rescue Plan Act 2021 Poverty Solutions, University of Michigan Ann Arbor, Michigan
Compton W.M. Gfroerer J. Conway K.P. Finger M.S. Unemployment and substance outcomes in the United States 2002–2010 Drug Alcohol Depend. 142 2014 350 353 25042761
Cylus J. Avendano M. Receiving unemployment benefits may have positive effects on the health of the unemployed Health Affairs 36 2 2017 289 296 28167718
Cylus J. Glymour M.M. Avendano M. Do generous unemployment benefit programs reduce suicide rates? A state fixed-effect analysis covering 1968–2008 American journal of epidemiology 180 1 2014 45 52 24939978
De Chaisemartin C. D'Haultfoeuille X. Two-way Fixed Effects and Differences-In-Differences with Heterogeneous Treatment Effects: A Survey (No. W29691) 2022 National Bureau of Economic Research
Delva J. Neumark Y.D. Furr C.D. Anthony J.C. Drug use among welfare recipients in the United States Am. J. Drug Alcohol Abuse 26 2 2000 335 342 10852365
Department of Labor Employment & Training Administration. Unemployment Insurance Data. Claims Data Reports 2022 [Web archive] Retrieved May 29 from https://oui.doleta.gov/unemploy/DataDashboard.asp
Donders A.R.T. Van Der Heijden G.J. Stijnen T. Moons K.G. A gentle introduction to imputation of missing values J. Clin. Epidemiol. 59 10 2006 1087 1091 16980149
Drydakis N. The effect of unemployment on self-reported health and mental health in Greece from 2008 to 2013: a longitudinal study before and during the financial crisis Soc. Sci. Med. 128 2015 43 51 25589031
Engels J.M. Diehr P. Imputation of missing longitudinal data: a comparison of methods J. Clin. Epidemiol. 56 10 2003 968 976 14568628
European Monitoring Centre for Drugs and Drug Addiction Handbook for Surveys on Drug Use Among the General Population 2002 EMCDDA project CT
Evans D.K. Popova A. Cash transfers and temptation goods Economic Development and Cultural Change 65 2 2017 189 221
Fu W. Liu F. Unemployment insurance and cigarette smoking J. Health Econ. 63 2019 34 51 30453224
Groshen E. How Job Losses during the COVID-19 Recession Compare to Past Recessions COVID-19 2020 Cornell ILR school: https://www.ilr.cornell.edu/work-and-coronavirus/work-and-jobs/how-job-losses-during-covid-19-recession-compare-past-recessions
Goodkind N. America has an employment problem—and two distinct paths ahead [Web archive] Retrieved May 28, 2022 from Fortune 2021 https://fortune.com/2021/06/01/us-employment-problem-covid-19-minimum-wage-unemployment-benefits/#:∼:text=America%20has%20an%20employment%20problem%E2%80%94and%20two%20distinct%20paths%20ahead&text=The%20immediate%20crisis%20of%20COVID,will%20reopen%20on%20June%2015
Gruber J. Prinstein M.J. Clark L.A. Rottenberg J. Abramowitz J.S. Albano A.M. Aldao A. Borelli J.L. Chung T. Davila J. Forbes E.E. Gee D.G. Hall G. Hallion L.S. Hinshaw S.P. Hofmann S.G. Hollon S.D. Joormann J. Kazdin A.E. Klein D.N. Mental health and clinical psychological science in the time of COVID-19: challenges, opportunities, and a call to action Am. Psychol. 76 3 2021 409 426 10.1037/amp0000707 32772538
Hammen C. Stress and depression.Annual Review of Clinical Psychology 1 1 2005 293 319 2005
Henkel D. Unemployment and substance use: a review of the literature (1990-2010) Curr. Drug Abuse Rev. 4 1 2011 4 27 21466502
ILO ILO Monitor: COVID-19 and the World of Work. Seventh edition. Updated Estimates and Analysis 2021 International Labour Organization
Janlert U. Hammarström A. Which theory is best? Explanatory models of the relationship between unemployment and health BMC Publ. Health 9 2009 1 9 10.1186/1471-2458-9-235
Karpman M. Acs G. Unemployment Insurance and Economic Impact Payments Associated with Reduced Hardship Following CARES Act 2020 Urban Institute Washington, DC
Kegler S.R. Simon T.R. Zwald M.L. Chen M.S. Mercy J.A. Jones C.M. Vital signs: changes in firearm homicide and suicide rates—United States, 2019–2020 MMWR (Morb. Mortal. Wkly. Rep.) 71 19 2022 656 35550497
Lantis R. Teahan B. The effect of unemployment insurance on alcohol use and abuse following job loss Econ. Hum. Biol. 30 2018 92 103 30015294
Lee J.O. Hill K.G. Hartigan L.A. Boden J.M. Guttmannova K. Kosterman R. Unemployment and substance use problems among young adults: does childhood low socioeconomic status exacerbate the effect? Soc. Sci. Med. 143 2015 36 44 26342911
Lee J.O. Kapteyn A. Clomax A. Jin H. Estimating influences of unemployment and underemployment on mental health during the COVID-19 pandemic: who suffers the most? Publ. Health 201 2021 48 54
Mathers C.D. Schofield D.J. The health consequences of unemployment: the evidence Colin Read. Res. Q. 26 1 2012 174 207 10.1680/udap.2010.163
Murphy G.C. Athanasou J.A. The effect of unemployment on mental health J. Occup. Organ. Psychol. 72 1 1999 83 99
Nagelhout G.E. Hummel K. Goeij M.C.M. Vries H. Kanerd E. Lemmens P. How economic recessions and unemployment affect illegal drug use: a systematic realist literature review Int. J. Drug Pol. 44 2017 69 83
O'Campo P. Molnar A. Ng E. Renahy E. Mitchell C. Shankardass K. Social welfare matters: a realist review of when, how, and why unemployment insurance impacts poverty and health Soc. Sci. Med. 132 2015 88 94 25795992
Paul K.I. Moser K. Unemployment impairs mental health: meta-analyses J. Vocat. Behav. 74 3 2009 264 282
Panchal N. Garfield R. Cox C. Artiga S. Substance Use Issues Are Worsening Alongside Access to Care 2021 Kaiser Family Foundation https://www.kff.org/policy-watch/substance-use-issues-are-worsening-alongside-access-to-care/?utm_campaign=KFF-2021-Coronavirus&utm_medium=email&_hsmi=2&_hsenc=p2ANqtz-_fXOUy6u2QfT3EZYUtgoL0okyhWUCRSj0_99pyR46ajhctK43K6HXx7ni_vmHh6KwL5cq__aWfEKi_reep3TvsOAaiiA&utm_content=2&utm_source=hs_email
Parolin Z. Collyer S. Curran M. Wimer C. The Potential Poverty Reduction Effect of the American Rescue Plan (No. 20411). Center on Poverty and Social Policy 2021 Columbia University
Roberts A. Rogers J. Mason R. Siriwardena A.N. Hogue T. Whitley G.A. Law G.R. Alcohol and other substance use during the COVID-19 pandemic: a systematic review Drug Alcohol Depend. 229 2021 109150
Rodriguez E. Frongillo E.A. Chandra P. Do social programmes contribute to mental well-being? The long-term impact of unemployment on depression in the United States International Journal of Epidemiology 30 1 2001 163 170 11171879
Ruhm C.J. Deaths of Despair or Drug Problems? (No. W24188) 2018 National Bureau of Economic Research
Russell B.S. Hutchison M. Tambling R. Tomkunas A.J. Horton A.L. Initial challenges of caregiving during COVID-19: caregiver burden, mental health, and the parent–child relationship Child Psychiatr. Hum. Dev. 51 5 2020 671 682
Standing G. The precariat Contexts 13 4 2014 10 12 10.1177/1536504214558209
Tafet G.E. Bernardini R. Psychoneuroendocrinological links between chronic stress and depression Prog. Neuro Psychopharmacol. Biol. Psychiatr. 27 6 2003 893 903
Usaspending The Federal Response to COVID-19 2022 [Web archive] Retrieved May 28, 2022 from https://www.usaspending.gov/disaster/covid-19?publicLaw=all
Williams R. Panel Data: Very Brief Overview 2015 University of Notre Dame
Wu P. Evangelist M. Unemployment insurance and opioid overdose mortality in the United States Demography 59 2 2022 485 509 35212718
Young R. Johnson D.R. Handling missing values in longitudinal panel data with multiple imputation J. Marriage Fam. 77 1 2015 277 294 26113748
Zipperer B. Gould E. Unemployment Filing Failures. Working Economics Blog 2020 https://www.epi.org/blog/unemployment-filing-failures-new-survey-confirms-that-millions-of-jobless-were-unable-to-file-an-unemployment-insurance-claim/
|
PMC010xxxxxx/PMC10207836.txt |
==== Front
Pediatr Neonatol
Pediatr Neonatol
Pediatrics and Neonatology
1875-9572
2212-1692
Taiwan Pediatric Association. Published by Elsevier Taiwan LLC.
S1875-9572(23)00093-1
10.1016/j.pedneo.2023.02.007
Original Article
Association between children's home-schooling and parental psychological distress during the COVID-19 pandemic in Taiwan: Risk and protective factors in a multilevel approach
Wang Chin-Wan a
Lu Hsin-Hui bc∗
Liang Jao-Shwann de
Chen Duan-Rung f
Chen Chia-Chun de∗∗
a Department of Social Work, Tunghai University, Taichung, Taiwan
b Department of Psychology, College of Medicine, Chung Shan Medical University, Taichung, Taiwan
c Clinical Psychological Room, Chung Shan Medical University Hospital, Taichung, Taiwan
d Department of Pediatrics, Far Eastern Memorial Hospital, New Taipei City, Taiwan
e Department of Nursing, Asia Eastern University of Science and Technology, New Taipei City, Taiwan
f Institute of Health Behaviors and Community Sciences, College of Public Health, National Taiwan University, Taipei, Taiwan
∗ Corresponding author. Department of Psychology, Chung Shan Medical University, No. 110, Sec. 1, Jianguo N. Rd., Taichung City, 40201, Taiwan.
∗∗ Corresponding author. Department of Pediatrics, Far Eastern Memorial Hospital, No. 21, Section 2, Nanya S. Road, Banqiao District, New Taipei City, 22060, Taiwan.
24 5 2023
24 5 2023
4 9 2022
7 1 2023
1 2 2023
© 2023 Taiwan Pediatric Association. Published by Elsevier Taiwan LLC.
2023
Taiwan Pediatric Association
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
The role home-schooling of children in parental mental health during the COVID-19 pandemic in Taiwan remains unknown. This study aimed to assess the association between parental psychological distress and home-schooling in a socio-ecological context during the peak of the first wave of the COVID-19 pandemic in Taiwan.
Methods
This was a prospective cohort study. In total, 902 parents (father: n = 206, mother: n = 696) who home-schooled children under 18 years of age were recruited by purposive sampling from 17 cities in Taiwan. Data were collected between 19 July and 30 September 2021 through a survey. Multilevel regression models were used to examine the association between parents’ psychological distress and home-schooling considering the characteristics at the person and city levels.
Results
Parental psychological distress was positively associated with difficulty in setting up electronic devices and increased disputes between parents and children, and it was negatively associated with time management and increased time spent bonding with their children during home-schooling (Ps < 0.05). Parents who had a child with health conditions, lived in an extended family, worked from home, lived during the Level 3 alert level, and lived with a median/sporadic level of the COVID-19 community spread by city also reported greater psychological distress (Ps < 0.05). However, parents who had greater household family support reported less psychological distress (P < .05).
Conclusions
Clinicians and policy makers must carefully consider parental mental health while home-schooling during the COVID-19 pandemic in a broader socio-ecological context. A focus is advised on the home-schooling experiences of parents and other risk and protective factors for parental psychological distress at the person and city levels, especially for those with children who require medical interventions and have a medical condition.
Key Words
COVID-19
hierarchical linear model
lockdown
parental mental health
remote learning
Abbreviations
COVID-19 Coronavirus Disease 2019
BSRS-5 Brief Symptom Rating Scale 5
WFH Work from home
RWO Regular work in the office
==== Body
pmc1 Introduction
Over the past 3 years, the Coronavirus disease 2019 (COVID-19) pandemic has changed family lives worldwide. In mid-May of 2021, the number of COVID-19 cases in Taiwan jumped abruptly and deaths started to rise steadily. The government put a level-3 emergency alert in place; and parents and children experienced the full shutdown of their country from 15 May to 26 July 2021. This led to a sudden shift from traditional classrooms (in-person instruction) to home-schooling (remote learning) in response to the school shutdowns, which resulted in a completely different learning experience for children and parents.1 , 2
Parents may pass their psychological distress on to their children and practice inappropriate parenting behaviours, which could later contribute to the development of mental problems in these children.3, 4, 5, 6 The association between parental psychological distress and home-schooling during the COVID-19 in Taiwan is unclear. The current findings could help clinicians to assess the mental health of parents who are home-schooling children during additional waves of the COVID-19 pandemic as well as future pandemics.
Home-schooling places considerable demands on parents to, for example, set up electronic devices and ensure adequate internet connectivity. Furthermore, such online set-ups can be accidently disconnected, and parents are responsible for immediate repairs. Additionally, parents have to manage the supervision of their children's learning progress alongside their professional, personal, and parenting roles. Only 14% of primary school students can complete remote school activities without assistance7; therefore, parents need to be involved in their children's home-schooling.
By contrast, the impact of home-schooling is not universally negative; current evidence suggests that parents’ experiences of home-schooling are mixed.8 , 9 Some parents report positive aspects of home-schooling during school closures, including less time spent commuting and more time spent bonding with their children, which can increase family closeness. Additionally, the home is perceived as a better learning environment for some children.8 Therefore, the effectiveness of parent-child interactions or parental time management may increase during home-schooling.
Bronfenbrenner's socio-ecological model10 provides a comprehensive framework to examine psychological distress during home-schooling by considering the risks and protective factors for parental mental health. These factors are individual (e.g., age, father/mother, and educational level of the parent), interpersonal (social support, child factors, and family factors), organizational (types of work), and community-related (e.g., COVID-19 alert level).11 , 12 Furthermore, the recorded level of community spread differed across cities in Taiwan during the COVID-19 epidemic, and previous studies had found that an individual's psychological distress is associated with the level of COVID-19 community spread in their respective city.13 , 14 Parental psychological distress during home-schooling is thus associated with multiple factors at the person-as well as the city level.
Given the ongoing impact of COVID-19, there is a clear need to comprehensively quantify the parental psychological distress caused by home-schooling. This study therefore aimed to (1) assess the association between parental distress and negative and positive experiences of home-schooling during the peak of the first wave of the COVID-19 pandemic in Taiwan; and (2) identify other risk and protective factors against parental psychological distress in a broader socio-ecological context.
2 Methods
2.1 Sampling procedures and participants
This was a prospective cohort study with an anonymous questionnaire survey. In total, 992 parents (fathers: n = 206, 22.84%; mothers: n = 696, 77.16%) who were home-schooling children aged ≤18 years were recruited through convenience sampling from 17 cities in Taiwan. Ninety participants (9.07%) were excluded due to failures on three attention check questions. Therefore, 902 parents completed the questionnaire and were included in our analyses; 661 were interviewed online via the survey management software SurveyCake, and 241 filled in the paper-and-pencil version of the questionnaire. The first page of the questionnaire provided information on the purpose of the research and the consent statement, to which the participants had to agree. Completion of the questionnaire took approximately 15–20 mins. Data were collected between 19 July and 30 September 2021.
2.2 Measures
2.2.1 Parental psychological distress
Parental psychological distress was assessed using the Brief Symptom Rating Scale 5 (BSRS-5).15 This five–item scale has been shown to have excellent validity and reliability in Taiwan and accounts for psychological symptoms such as depressive feelings and low mood.16 , 17 Each item was scored on a Likert scale from 0 to 5 (0 = “never,” 1 = “mild,” 2 = “moderate,” 3 = “severe,” and 4 = “very severe”). The sum of these scores ranged from 0 to 20. Cronbach's α for BSRS-5 was 0.88. The BSRS-5 comprises the following five dimensions: (1) Anxiety (feeling tense of “keyed up”), (2) Depression (feeling blue), (3) Hostility (feeling easily annoyed or irritated), (4) Interpersonal Sensitivity (feeling inferior to others), and (5) sleep difficulties. According to Lee's findings,15 the BSRS-5 is an efficient tool for the screening of suicidal ideation-prone psychiatric inpatients, general medical patients, and community residents. Understanding the differentiated symptom domains for each group and the relationship between them can help health care professionals in their preventative programs and clinical treatment. A total score on the BSRS-5 above 14 may indicate severe mood disorders. Scores between 10 and 14 may indicate moderate mood disorders and scores between 6 and 9 may indicate mild mood disorders.
2.3 General negative and positive experiences of home-schooling
We surveyed the experiences of home-schooling, which have been identified as an important factor for parental mental health during the COVID-19 pandemic.8 , 12 , 18 , 19 The negative experiences index was evaluated using the following two items: “The dispute between parents and children increased” and “Dealing with children's home-schooling devices or emergencies took longer and was harder than I expected.” The positive experiences index was evaluated using the following two items: “The time for parents to chat with children increased” and “Time management is more efficient during home-schooling than during classroom learning.” Each item was scored on a Likert-type scale from 1 to 4 (1 = “disagree,” 2 = “slightly disagree,” 3 = “agree,” and 4 = “strongly agree”). The scores of each index ranged from 2 to 8.
2.4 Covariates
2.4.1 City-level variable
2.4.1.1 COVID-19 community level by city
Taiwan was categorized according to three levels of the spread of COVID-19 during the pandemic. The red zone was defined as having high or massive levels of COVID-19 infections, i.e., Taipei and New Taipei City. The yellow zone cities with median or sporadic levels of COVID-19 infections, namely Taoyuan, Miaoli, Taichung, Changhua, Kaohsiung, and Pingtung. The green zone was defined as having low levels of COVID-19 infections, with few suspected or confirmed cases admitted to hospitals; it covered Keelung, Hsinchu, Yunlin, Chiayi, Tainan, Hualien, Yilan, Nantou, and the offshore islands.
2.5 Person-level variables
2.5.1 Individual
(1) Age, reported by parents; (2) Parents, self-reported: “Father” coded as “0” and “mother” as “1”; and (3) Educational level, was classified as “junior high school and below,” “senior high school,” and “university and above.”
2.5.2 Interpersonal
(1) Social support. (a) Perceived household family support was assessed with four items: whether the respondent felt close to their family (emotional support); whether their family helped them when they were in need (instrumental support); whether they provided them with advice when they were in need (informational support); and whether they appreciated their thoughts and behaviors (appraisal support). (b) Perceived support by relatives, (c) perceived friend support, and (d) perceived neighbor support were similarly evaluated with four items. Each item was scored on a Likert-type scale from 1 to 4 (1 = “disagree,” 2 = “slightly disagree,” 3 = “agree,” and 4 = “strongly agree”). Cronbach's α for the household family, the relatives, the friends, and the neighbors support items were 0.86, 0.90, 0.90, and 0.93, respectively. (2) Child factors. (a) Number of children being home-schooled, reported by parents. (b) Children who required medical interventions, coded yes = 1 vs. no = 0. (c) Children with medical diagnosis, including neurodevelopmental disorders (e.g., autism spectrum disorders or developmental delay), neurological conditions (e.g., epilepsy), sensory impairment or other paediatrics diseases, coded yes = 1 vs. no = 0. (3) Family factors. (a) Immigrant families were coded as yes = 1 vs. no = 0. (b) Type of family structure was categorized into nuclear family, extended family, and others.
2.5.3 Organizational
Type of work during home-schooling was categorized into work from home (WFH), alternate working to the office, and regular work in the office (RWO).
2.5.4 Community
The COVID-19 alert levels that were in place during this study were levels 2 and 3. The level 3 COVID-19 alert was in place from 19 to 26 July 2021, and the level 2 alert from 27 July to 30 September 2021.
2.6 Data analysis
The characteristics of the sample are presented in Table 1 . Most parents were university graduates and above (73.95%), with those who only completed junior high school and below representing the smallest group of respondents (1.77%). Few respondents were from immigrant families (3.99%). Most parents lived in a nuclear family (72.28%) and only approximately 20% in an extended family. Furthermore, 32.26% of the children required medical interventions during home-schooling and 10.31% had a diagnosis of neurodevelopmental or neurological disorders. Approximately 30% of parents maintained RWO; however, 41.46% reported WFH, and 24.94% worked alternately. As for the COVID-19 alert level, 43.79% of respondents filled in the questionnaire during a level 3 phase. Most parents lived in a red zone during their reported home-schooling phase (79.38%), while those in a green zone represented the smallest group (7.87%).Table 1 Characteristics of the study sample (N = 902).
Table 1Characteristics Mean SD n %
Parental psychological distress 9.81 3.84
Experiences of home-schooling
Negative experiences index 5.06 1.45
Positive experiences index 5.79 1.24
City-level variable
COVID-19 community level by city
Red zone 716 79.38
Yellow zone 115 12.75
Green zone 71 7.87
Person-level variable
Individual
Age 42.48 5.73
Parents
Father 206 22.84
Mother 696 77.16
Education level
Junior and below 16 1.77
Senior 219 24.28
University and above 667 73.95
Interpersonal
Social supports
Household 3.19 0.52
Relatives 2.19 1.25
Friends 2.32 1.28
Neighbors 1.53 1.09
Child factors
Number of children in home-schooling 1.56 0.62
Child having medical intervention 291 32.26
Child with medical diagnosis 93 10.31
Family factors
Immigrant family 36 3.99
Types of family structure
Nuclear family 652 72.28
Extended family 180 19.96
Other family 70 7.76
Characteristics Mean SD n %
Organizational: Types of work
WFH 374 41.46
Alternately 225 24.94
RWO 303 33.59
Community: During COVID-19 alert level
Level 3 395 43.79
Level 2 507 56.21
WFH, work from home; RWO, regular work in the office.
Multilevel regression models (MLMs) were used to investigate the association between parental psychological distress and home-schooling, and they facilitated the examination of connections at each level (person- and city-level) and the amount of variation taken into account at each level.20 All continuous variables were grand-centered. We used HLM version 7.03 (Scientific Software International Inc.) to apply the MLMs and calculate a population-averaged model with robust standard errors. The data had a multilevel structure, with persons (level 1) nested within cities (level 2), and each multilevel regression model thus consisted of two hierarchical levels. The level 1 model included two experience indices of home-schooling (negative and positive) and individual, interpersonal, organizational, and community-related factors. The level 2 model included COVID-19 community spread by city. Let Yij = parental psychological distress, our dependent variable taken on the ith parent associated with the jth city. The level 1 equation of the full model was as follows:Yij = β0j + β1jNegativeij + β2jPositiveij + β3jAgeij + β4jMotherij(Ref=Father) + β5jSeniorij(Ref=Junior and below)+ β6jUniversity and aboveij(Ref=Junior and below)+ β7jHouseholdij + β8jRelativesij + β9jFriendsij + β10jNeighborsij + β11jNumber of children in home-schoolingij + β12jChild having medical interventionij(Ref=None) + β13jChildren with medical diagnosisij(Ref=None) + β14jImmigrant familyij(Ref=None) + β15jExtended familyij(Ref=Nuclear family) + β16jOther familyij(Ref=Nuclear family) + β17jWFHij(Ref=RWO) + β18jAlternatelyij(Ref=RWO) + β19jLevel 3ij(Ref=Level 2) + γij
The level 2 equation for the full model is as follows:β0j = γ00 + γ01*Yellow zonej(Ref=Red zone) + γ02∗Green zonej(Ref=Red zone) + μ0j
3 Results
3.1 Parental psychological distress and home-schooling
The intra-class correlation coefficient (ICC), reflecting the variance across the 17 cities (level 2), was 4.12% for parental psychological distress. This indicates that the multilevel modeling approach was appropriate for our analysis.21
Table 2 presents the results of Models 1, 2, and 3. Model 1 shows that parental psychological distress was negatively associated with positive experiences index, e.g., effectiveness (P < .001) and positively associated with negative experiences index, e.g., emergency interventions (P < .001). In Model 2, with added variables at the personal level, parental psychological distress was also negatively associated with positive experiences index (P = .037) and positively with negative experiences index (P < .001). The final full model (Model 3) indicated a further significant effect on positive (P = .039) and negative (P < .001) experiences after adding the COVID-19 community spread variable.Table 2 Summary of MLM for parental psychological distress (N = 902).
Table 2 Model 1 Model 2 Model 3
Fixed effect Coeff 95% CI P Coeff 95% CI P Coeff 95% CI P
Intercept 10.19 9.62–10.77 <0.001 8.12 5.33–10.91 <0.001 7.50 4.77–10.23 <0.001
Experiences ofhome-schooling
Negative experiences index 0.89 0.77–0.01 <0.001 0.74 0.59–0.90 <0.001 0.74 0.58–0.89 <0.001
Positive experiences index −0.45 −0.57 to −0.33 <0.001 −0.20 −0.39 to −0.01 0.037 −0.20 −0.38 to −0.01 0.039
City-level variable
COVID-19 community level by city
Yellow zone(Ref=Red zone) 1.61 0.95–2.27 <0.001
Green zone(Ref=Red zone) 0.17 −0.67 – 1.00 0.702
Person-level variable
Individual (Parents)
Age −0.03 −0.07 – 0.01 0.157 −0.03 −0.07 – 0.01 0.113
Parents: other(Ref = Father) 0.29 −0.27 – 0.84 0.309 0.27 −0.28 – 0.82 0.342
Educational level
Senior(Ref = Junior and below) 0.40 −2.31 – 3.12 0.772 0.48 −2.22–0.18 0.730
University and above(Ref = Junior and below) 0.88 −1.82 – 3.58 0.522 0.92 −1.76–3.61 0.501
Interpersonal
Social supports
Household −1.63 −2.09 to −1.17 <0.001 −1.62 −2.08 to −1.17 <0.001
Relatives −0.11 −0.31 – 0.09 0.273 −0.13 −0.32 – 0.07 0.209
Model 1 Model 2 Model 3
Fixed effect Coeff 95% CI P Coeff 95% CI P Coeff 95% CI P
Social supports
Friends −0.08 −0.28 – 0.09 0.414 −0.06 −0.26 to −0.14 0.566
Neighbors −0.17 −0.38 – 0.05 0.127 −0.17 −0.38 – 0.04 0.119
Children factors
Number of children in home-schooling −0.09 −0.46 – 0.28 0.630 −0.10 −0.46 – 0.27 0.605
Child having medical intervention (Ref = None) 0.57 0.10–1.05 0.018 0.59 0.12–1.05 0.015
Child with medical diagnosis (Ref = None) 0.75 0.02–1.49 0.045 0.83 0.10–1.56 0.026
Family factors
Immigrant family(Ref=None) −0.88 −2.01 – 0.24 0.123 −0.88 −2.00 – 0.23 0.122
Types of family structure
Extended family(Ref=Nuclear family) 0.64 0.09–1.20 0.024 0.62 0.06–1.17 0.029
Other family(Ref=Nuclear family) 0.46 −0.38–1.30 0.282 0.40 −0.43 – 1.24 0.345
Organizational: Types of work
WFH(Ref=RWO) 0.94 0.42–1.47 <0.001 0.96 0.44–1.48 <0.001
Alternately(Ref=RWO) 0.38 −0.20 – 0.96 0.199 0.38 −0.20 – 0.96 0.195
Community: During COVID-19 alert level
Level 3 (Ref=Level2) 0.47 0.02–0.92 0.041 0.45 0.01–0.90 0.047
Random effect Variance Component P Variance Component P Variance Component P
Intercept (μ) 0.53 <0.001 0.44 <0.001 0.00005 >0.500
Individual-level (γ) 12.31 10.85 10.81
WFH, work from home; RWO, regular work in the office.
Additionally, Model 3 indicated other risks and protective factors. On the interpersonal level, parental psychological distress was negatively associated with household social support (P < .001). Moreover, parental psychological distress was higher in those with children who required medical interventions (P = .015) or had neurodevelopmental or neurological disorders (P = .026). Additionally, compared with parents living in a nuclear family, those living in an extended family were more likely to report psychological distress during home-schooling (P = .029). Our organizational-level analysis showed that compared with parents who maintained their RWO, WFH parents were more likely to report psychological distress (P < .001). The community level analysis revealed that parents reported greater psychological distress during level 3 than level 2 alert periods (P = .047). Finally, parents in the yellow zone reported greater psychological distress than those in the red zone (P < .001).
4 Discussion
To the best of our knowledge, this is the first study to examine the association between parental psychological distress and home-schooling experiences during the COVID-19 pandemic in Taiwan. Our main findings are that parents who reported having to deal with urgent emergency interventions or experiencing low effectiveness during home-schooling exhibited more psychological distress. The risk factors for parental mental health during home-schooling due to the COVID-19 pandemic were (1) having a child with health conditions, (2) living in an extended family, (3) working from home, (4) home-schooling during a level 3 alert level, and (5) living in an environment with a median/sporadic level of community spread. In contrast, having a supportive household family was a protective factor for parental mental health.
4.1 Parental psychological distress associated with experience of home-schooling
4.1.1 Negative experiences
Parents who had to manage emergency interventions during remote learning reported greater psychological distress. This pandemic was sudden and unexpected; therefore, most parents were unprepared for this quick shift from traditional learning to home-schooling. Previous research indicates that parental depression and stress due to home-schooling are significantly positively associated with parents' perceived failure to provide at-home education.22 Parents reported that dealing with equipment or devices for home-schooling was difficult. Furthermore, the transition to remote learning required parents to teach their children knowledge, design individualized content, and constantly assess their children's emotional needs. Therefore, they became proxy educators and supervisors of the remote learning of their children.2 These added extra responsibilities and stressors to parents' usual tasks during school closures, and they were required to perform multiple, and sometimes conflicting, roles.23 In addition, some parents had to work longer hours each day to meet home-management and home-schooling obligations, which can affect sleep and reduce time for leisure activities.1 , 24
4.1.2 Positive experiences
The psychosocial impact of home-schooling during the COVID-19 pandemic on parents’ mental health was not only negative. Parents reported that home-schooling increased the time they had to chat with their children, and that eliminating or reducing physical commute times aided their time management, which may have increased closeness and understanding among family members. Some research has reported that the pandemic provided an opportunity to cultivate positive qualities in parent-child interactions, including “appreciation,” “developing tolerance and understanding,” and “learning to cope and develop patience”.8 The effectiveness of parent-child interactions and time arrangement increase during home-schooling. However, the precise mechanisms underlying positive effects of home-schooling need further exploration.
4.2 Parental psychological distress associated with other factors
4.2.1 Protective factors
We found that social support from the household family was the strongest protective factor against psychological distress among parents during home-schooling. Calear et al.1 found that parents who reported higher perceived social support from their child's school tended to exhibit lower levels of psychological distress; and these together with our findings are consistent with the general literature on mental health that links social support to better mental health outcomes.25 The mediating effect of social support on the association of perceived stress and mental health outcomes has long been recognized.26 , 27 Social support can increase an individual's resilience and support their adaptive psychological capacities.28
4.2.2 Risk factors
Parents who had a child with health conditions, lived in an extended family, worked from home during the level 3 alert phase, and lived in a city/area with a median/sporadic level of COVID-19 community spread reported greater psychological distress. Previous research has highlighted the potential risk factors for parental psychological distress during the COVID-19 pandemic, which are in line with the factors we identify here, that is, having a child with health conditions,29 , 30 living in an extended family,31 working from home,23 and experiencing a level 3 alert phase.32
It is also worth noting that, compared with parents living in a nuclear family, parents living in an extended family had higher psychological distress. Why do grandparents not reduce parental psychological distress? It may be that many grandparents may fall under the “high-risk” category for COVID-19-related illnesses, or even death, as a result of their age or the presence of underlying health conditions.24 As such, grandparents cannot provide social support to the family, and parents’ duties to care for grandparents would increase significantly during the COVID-19 pandemic. Many parents did not rely on support from grandparents or other family members for childcare or help with parenting-related activities.
Additionally, the psychological distress reported by parents living in cities with only sporadic Covid-19 spread (such as Taoyuan) was higher than those living in cities with massive spread (such as New Taipei City), which was most likely due to the “uncertainty” in cities with only sporadic spread and to the constant fear of infection. Therefore, it was important to determine the associations between parental mental health and home-schooling in a larger socio-ecological context.
Lastly, there were possible associations between parental psychological distress and living with extended family who had Covid-19, immunodeficiency, metabolic syndrome, cardiovascular diseases, or other potential infection risks. While living with extended family may not necessarily be considered bad for mental health, it does increase the chance of elderly having pressure regarding potential infections in the household. While this study did not examine those particular variables, it will certainly be worth exploring further in future research.
4.3 Strengths and limitations
This study has several strengths: (1) the survey was conducted during the peak and slowdown of the first wave of the COVID-19 pandemic in Taiwan. (2) Data were collected during school closures rather than relying on retrospective reports. (3) This was a large and diverse sample of parents who had children ≤18 years of age with and without health conditions. (4) We examined the associations of parental psychological distress and home-schooling with a wide range of socio-ecological factors, including individual, interpersonal, organization, and community-related factors. (5) The collected sample shows strong representativeness of the Taiwanese population regarding families with a wide range of children's ages (under aged 18), as well as the data gathered from 17 of 22 cities in Taiwan. Additionally, we assessed different levels of the spread of COVID-19 across cities, which relates to the perceived risk of infection in the community more than to the official alert level.22 These strengths increase the validity of the findings of this study.
The following limitations should, however, be considered: (1) we relied on the parents' reports of their current psychological distress rather than on longitudinal data. Therefore, we do not know whether parents were psychologically distressed before the COVID-19 pandemic. Additionally, we were unable to assess the parental psychopathology that contributed to the negative outcomes of home-schooling. (2) Many standardized and validated measures were not used to assess parents' and children's experiences of home-schooling in this study, since we chose to not substantially increase the length of the survey; we assumed keeping the questionnaire short would increase the response rate. Furthermore, the items used in this study were designed based on previous COVID-19 research on the effects of home-schooling on parents and family functioning. (3) Data were collected only once rather than in repeated measures. Therefore, the observed association between parental psychological distress and home-schooling does not necessarily indicate a cause-effect relationship. (4) Additionally, the period for data collection was brief; however, we did observe an association between home-schooling and parental distress even within this short time. Further studies are needed to confirm any long-term impact of home-schooling on parental distress. (5) Children with a medical diagnosis were not clearly differentiated in this study. As children with a medical diagnosis account for only 93, we could not differentiate different diagnosis groups in the analysis. However, children with neurological disorders, neurodevelopmental disorders, or sensory impairment may have different effects on parental psychological distress. This is an important issue to be examined in the future study.
5 Conclusions
In Taiwan, school closures have been used to reduce the spread of COVID-19. In response, parents and children had to adapt to new modes of remote learning.33 The present results clearly show that parental mental health is negatively associated with having to deal with emergencies during home-schooling of a child. However, home-schooling does not only have negative impacts on parental mental health. Parents also reported positive experiences of more parent-child interactions and better time management during home-schooling. Another interesting finding of this study is that social support from the household family is an important protective factor against parental psychological distress and may reduce the negative impact of other risk factors. The risk factors for parental mental health during home-schooling are having a child with health conditions, living in an extended family, working from home, and a median/sporadic level of COVID-19 community spread. Therefore, clinicians and policy makers must carefully consider the mental health condition of parents during home-schooling in a broader socio-ecological context during later phases of the COVID-19 pandemic or future natural disasters (e.g., fires, floods, and earthquakes). When the schools are closed and the needs of home-schooling increase, it would be crucial for the government to aid in the form of remote devices and settings, home-based parent-child activities, and advocacy of informal household social support helping strengthen family functions, especially for those with children who require medical interventions and have a medical condition.
Ethical approval
Ethics committee approval was not required for this anonymous questionnaire study. Furthermore, the study was performed in accordance with the ethical standards as laid down in the Declaration of Helsinki of 1964 and its subsequent amendments or comparable ethical standards.
Availability of data and materials
The datasets used or analyzed during the current study are available from the corresponding author, Hsin-Hui Lu, on reasonable request.
Funding
Editing of this article was funded with a grant from the National Science and Technology Council, Taiwan (10.13039/501100004663 MOST 108-2410-H-040-010-MY3; 10.13039/501100004663 MOST 111-2410-H-040-003-MY4).
Declaration of competing interest
The authors declare that they have no competing interests.
Acknowledgments
The authors would like to thank the families and children who participated in this study. We also thank Wai-Fan Chan and the other research study staff.
==== Refs
References
1 Calear A.L. McCallum S. Morse A.R. Banfield M. Gulliver A. Cherbuin N. Psychosocial impacts of home-schooling on parents and caregivers during the COVID-19 pandemic BMC Public Health 22 2022 119 35039044
2 Davis C.R. Grooms J. Ortega A. Rubalcaba J.A.A. Vargas E. Distance learning and parental mental health during COVID-19 Educ Res 50 2021 61 64
3 Babore A. Trumello C. Lombardi L. Candelori C. Chirumbolo A. Cattelino E Mothers’ and children’s mental health during the COVID-19 pandemic lockdown: the mediating role of parenting stress Child Psychiatry Hum Dev 54 2023 134 146 34417927
4 Demaria F. Vicari S. COVID-19 quarantine: psychological impact and support for children and parents Ital J Pediatr 47 2021 58 33750452
5 Roos L.E. Salisbury M. Penner-Goeke L. Cameron E.E. Protudjer J.L.P. Giuliano R. Supporting families to protect child health: parenting quality and household needs during the COVID-19 pandemic PLoS One 16 2021 e0251720
6 Westrupp E.M. Bennett C. Berkowitz T. Youssef G.J. Toumbourou J.W. Tucker R. Child, parent, and family mental health and functioning in Australia during COVID-19: comparison to pre-pandemic data Eur Child Adolesc Psychiatry 32 2023 317 330 34417875
7 Lau E.Y.H. Lee K. Parents’ views on young children’s distance learning and screen time during COVID-19 class suspension in Hong Kong Early Educ Dev 32 2021 863 880
8 Evans S. Mikocka-Walus A. Klas A. Olive L. Sciberras E Karantzas G. From “It has stopped our lives” to “Spending more time together has strengthened bonds”: the varied experiences of Australian families during COVID-19 Front Psychol 11 2020 588667 33192922
9 Morse A.R. Banfield M. Batterham P.J. Gulliver A. McCallum S. Cherbuin N. What could we do differently next time? Australian parents’ experiences of the short-term and long-term impacts of home schooling during the COVID-19 pandemic BMC Public Health 22 2022 80 35027020
10 Eriksson M. Ghazinour M. Hammarström A. Different uses of Bronfenbrenner’s ecological theory in public mental health research: what is their value for guiding public mental health policy and practice? Soc Theor Health 16 2018 414 433
11 McRae C.S. Overall N.C. Henderson A.M.E. Low R.S.T. Chang V.T. Parents’ distress and poor parenting during a COVID-19 lockdown: the buffering effects of partner support and cooperative coparenting Dev Psychol 57 2021 1623 1632 34807685
12 Zhuang X. Lau Y.Y. Chan W.M.H. Lee B.S.C. Wong D.F.K. Risk and resilience of vulnerable families in Hong Kong under the impact of COVID-19: an ecological resilience perspective Soc Psychiatry Psychiatr Epidemiol 56 2021 2311 2322 34100978
13 Giordano F. Cipolla A. Ungar M. Building resilience for healthcare professionals working in an Italian red zone during the COVID-19 outbreak: a pilot study Stress Health 38 2022 234 248 34312986
14 Tshering K. Dema K. Psychological impact during COVID-19 pandemic: a web-based cross-sectional study among students studying at College of Science and Technology (CST), Phuentsholing, Bhutan PLoS One 17 2022 e0263999
15 Lung F.W Lee M.B. The five-item Brief-Symptom Rating Scale as a suicide ideation screening instrument for psychiatric inpatients and community residents BMC Psychiatry 8 2008 53 18597675
16 Chen H.C. Wu C.H. Lee Y.J. Liao S.C. Lee M.B. Validity of the five-item Brief Symptom Rating Scale among subjects admitted for general health screening J Formos Med Assoc 104 2005 824 829 16496062
17 Lee M.B. Liao S.C. Lee Y.J. Wu C.H. Tseng M.C Gau S.F. Development and verification of validity and reliability of a short screening instrument to identify psychiatric morbidity J Formos Med Assoc 102 2003 687 694 14691593
18 Morse A.R. Banfield M. Batterham P.J. Gulliver A. McCallum S. Cherbuin N. What could we do differently next time? Australian parents’ experiences of the short-term and long-term impacts of home schooling during the COVID-19 pandemic BMC Public Health 22 2022 80 35027020
19 Thorell L.B. Skoglund C. de la Peña A.G. Baeyens D. Fuermaier A.B.M. Groom M.J. Parental experiences of homeschooling during the COVID-19 pandemic: differences between seven European countries and between children with and without mental health conditions Eur Child Adolesc Psychiatry 31 2022 649 661 33415470
20 Snijders T.A. Bosker R.J. Multilevel analysis: an introduction to basic and advanced multilevel modeling 2011 Sage
21 Hox J.J. Moerbeek M. Van de Schoot R. Multilevel analysis: techniques and applications 2017 Routledge
22 Lee S.J. Ward K.P. Chang O.D. Downing K.M. Parenting activities and the transition to home-based education during the COVID-19 pandemic Child Youth Serv Rev 122 2021 105585
23 Garbe A. Ogurlu U. Logan N. Cook P. COVID-19 and remote learning: experiences of parents with children during the pandemic Am J Qual Res 4 2020 45 65
24 Griffith A.K. Parental burnout and child maltreatment during the COVID-19 pandemic J Fam Violence 37 2022 725 731 32836736
25 Gariépy G. Honkaniemi H. Quesnel-Vallée A. Social support and protection from depression: systematic review of current findings in Western countries Br J Psychiatry 209 2016 284 293 27445355
26 Ye Z. Yang X. Zeng C. Wang Y Shen Z Li X Resilience, social support, and coping as mediators between COVID-19-related stressful experiences and acute stress disorder among college students in China Appl Psychol Health Well Being 12 2020 1074 1094 32666713
27 Kalaitzaki A. Tsouvelas G. Koukouli S. Social capital, social support and perceived stress in college students: the role of resilience and life satisfaction Stress Health 37 2021 454 465 33206451
28 Southwick S.M. Sippel L. Krystal J. Charney D. Mayes L. Pietrzak R. Why are some individuals more resilient than others: the role of social support World Psychiatry 15 2016 77 79 26833614
29 Wu Q. Xu Y. Parenting stress and risk of child maltreatment during the COVID-19 pandemic: a family stress theory-informed perspective Dev Child Welfare 2 2020 180 196
30 Lawson M. Piel M.H. Simon M. Child maltreatment during the COVID-19 pandemic: consequences of parental job loss on psychological and physical abuse towards children Child Abuse Negl 110 2020 104709
31 Brown S.M. Doom J.R. Lechuga-Peña S. Watamura S.E. Koppels T. Stress and parenting during the global COVID-19 pandemic Child Abuse Negl 110 2020
32 Eales L. Ferguson G.M. Gillespie S. Smoyer S. Carlson S.M. Family resilience and psychological distress in the COVID-19 pandemic: a mixed methods study Dev Psychol 57 2021 1563 1581 34807681
33 Gadermann A.C. Thomson K.C. Richardson C.G. Gagné M. McAuliffe C. Hirani S Examining the impacts of the COVID-19 pandemic on family mental health in Canada: findings from a national cross-sectional study BMJ Open 11 2021 e042871
|
PMC010xxxxxx/PMC10207842.txt |
==== Front
Spectrochim Acta A Mol Biomol Spectrosc
Spectrochim Acta A Mol Biomol Spectrosc
Spectrochimica Acta. Part A, Molecular and Biomolecular Spectroscopy
1386-1425
1873-3557
Elsevier B.V.
S1386-1425(23)00596-6
10.1016/j.saa.2023.122911
122911
Article
Sustainable Stability-Indicating spectra manipulations for the concurrent quantification of a novel Anti-COVID-19 drug and its active Metabolite: Green profile assessment
Gamal Fawzy Michael
Kamel Ebraam B. ⁎
Pharmaceutical Chemistry Department, Faculty of Pharmacy, Egyptian Russian University, Badr City, Cairo 11829, Egypt
⁎ Corresponding author.
24 5 2023
5 11 2023
24 5 2023
300 122911122911
14 3 2023
11 5 2023
23 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Graphical abstract
Millions of individuals have lost their lives and changed their routines as a direct consequence of exposure to the coronavirus (Covid-19). Molnupiravir (MOL) is an orally bioavailable tiny molecule antiviral prodrug that is effective for curing the coronavirus that produces serious acute respiratory disorder (SARS-CoV-2). Fully green-assessed stability-indicating simple spectrophotometric methods have been developed and fully validated as per ICH criteria. The potential impact of degradation products of drug components on the safety and efficacy of a medication's shelf life is likely to be negligible. The field of pharmaceutical analysis necessitates various stability testing under different conditions. The conduct of such inquiries affords the prospect of predicting the most probable routes of degradation and ascertaining the inherent stability characteristics of the active drugs. Consequently, a surge in demand arose for the creation of an analytical methodology that could consistently measure the degradation products and/or impurities that may be present in pharmaceuticals. Herein, five smart and simple spectrophotometric data manipulation techniques have been produced for the concurrent estimation of MOL and its active metabolite as its possible acid degradation product namely; N-hydroxycytidine (NHC). Structure confirmation of NHC build-up through IR, MS and NMR analyses. All current techniques verified linearity ranging from 10 to 150 μg/ml and 10–60 μg/ml for MOL and NHC, respectively. The limit of quantitation (LOQ) values were in the range of 4.21–9.59 μg/ml, while the limit of detection (LOD) values were ranging from 1.38 − 3.16 μg/ml. The current methods were evaluated in terms of greenness by four assessing methods and confirmed to be green. The significant novelty of these methods depends on their being the first environmentally soundness stability-indicating spectrophotometric approaches for the concurrent estimation of MOL and its active metabolite, NHC. Also, the preparation of purified NHC delivers significant cost savings, instead of purchasing an expensive ingredient. These smart methods were utilized for analyzing the pharmaceutical dosage form which may be of great benefit to the pharmaceutical market.
Keywords
Coronavirus
Green
Molnupiravir
N-hydroxycytidine
Stability-indicating
UV- spectrophotometry
==== Body
pmc1 Introduction
There have been an estimated 270 million recorded incidents of the coronavirus disease pandemic of 2019 (COVID-19), driven by acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Around 5.2 million deaths have been attributed to the pandemic. [1]. Numerous vaccinations with proven efficacy in minimizing hospitalization and mortalities have been introduced; even yet, the outcomes of vaccination coverage remain unsatisfactory [2], [3]. Antiviral medications are required to moderate the risk of COVID-19 development. Given the importance of initiating treatment as soon as feasible following the start of symptoms [4], [5], [6], these medications should ideally be accessible to patients and simple to implement [7], [8].
Molnupiravir (MOL), (Fig. 1 (a)) is a tiny molecule ribonucleoside prodrug of N-hydroxycytidine (NHC), (Fig. 1(b)), with antiviral effectiveness against SARS-CoV-2 and other RNA viruses with low susceptibility to resistance formation [9], [10], [11], [12], [13], [14]. NHC flows systemically following oral treatment of MOL and undergoes an intracellular phosphorylation process to form NHC triphosphate. During viral replication, the viral RNA polymerase adds NHC triphosphate to the viral RNA, resulting in the viral polymerase being misdirected into adding guanosine or adenosine. The viral genome becomes littered with deleterious errors, eventually leaving the virus noninfectious and unable to replicate [15].Fig. 1 Chemical structures of MOL (a) and NHC (b).
There may be minimal effects on the safety and effectiveness of a medicine's shelf life from degradation products of drug components. Pharmaceutical analysis needs considerable stability tests under a wide range of situations. These investigations provide the opportunity to predict the most likely degradation pathways and to ensure the active substance's inherent stability features. Because of this, there was an increase in the need for developing an analytical technique that could reliably separate and quantify degradation products and/or impurities in pharmaceuticals [16], [17], [18], [19], [20]. Therefore, in the present investigation, MOL and its active metabolite NHC, which is a possible acid degradation product, were determined simultaneously.
In recent years, there has been a change in emphasis towards analytical techniques that better conform to the requirements of eco-friendly analytical chemistry [21].
As far as we know, no established spectrophotometric approaches for the concurrent estimation of MOL and NHC have been described. Whereas few chromatographic techniques exist for analyzing MOL with NHC, either for simultaneous quantification or to indicate stability [22], [23], [24]. Furthermore, by searching the literature, some spectrophotometric [25] and chromatographic methods [25], [26] for the quantification of MOL with some other antiviral drugs have been published.
Thus, we attempted to establish environmentally soundness spectrophotometric techniques for the concurrent estimation of MOL and its active metabolite NHC in raw materials, laboratory-synthetic combinations, and for the detection of MOL in its pharmaceutical dose form (Molunvac® Capsules).
As a consequence, our work has a unique novelty and superiority due to its invention, simplicity, time-saving, and cost-saving advantages.
2 Experimental
2.1 Instrument
An ACER-compatible personal computer was utilized with a UV/visible spectrophotometer run utilizing Jasco's spectrum management software. The reference and test solutions' absorption spectra were acquired in 1.0 cm quartz cells between 200 and 400 nm.
Schimadzu FTIR spebtrophotometer 8400, Agilant triple quadrupole mass detector 6420 and Bruker Ascend 400/R (1 H: 400 MHZ, 130 and DEPT-135: 100 MHZ) NMR spectrophotometer were used for detecting NHC formation.
2.2 Materials, solvents and reagents
MOL has been generously provided by Amoun Pharmaceutical Company, Cairo, Egypt, with a purity of 99.80 %. Molunvac® Capsules (Batch No. 15GG407X, labelled to have 200 mg MOL) were manufactured and gifted by Amoun pharmaceutical company, Cairo, Egypt. Ethanol (HPLC grade, Sigma, Germany) was utilized. NHC (acid degradation product) was prepared by utilizing hydrochloric acid and sodium hydroxide (Sigma, Germany).
2.3 Preparation of N-hydroxycytidine (NHC) (acid degradation product)
An accurately weighed amount of MOL (100 mg) was dissolved in 25 ml 1 N ethanolic HCl and the solution was refluxed for 1 h at 100 °C. To ensure complete degradation, 0.2 ml of the solution was further diluted by ethanol and applied on a TLC plate next to the MOL spot (intact drug) utilizing the developing system; chloroform: methanol: ammonia (2:4:0.5, v/v/v). Two spots were discovered after inspecting the TLC plate, one for MOL drug (Rf = 0.75) and the other for NHC (Rf = 0.5). After cooling and neutralization of the solution with 1 N sodium hydroxide, it was evaporated to dryness and purified by dissolving in a hot ethanol solution, followed by filtration and evaporation to dryness. Utilizing IR, MS and NMR spectroscopy, the acid degradation product was identified and its structure was elucidated.
2.4 Stock solutions
2.4.1 Standard solutions
To prepare the standard stock solution, MOL was dissolved in ethanol at a concentration of (1 mg/ml). A precise dilution of the prepared standard stock solutions is made with ethanol and then diluted and standard working solution (100 μg/ml) of MOL was produced. The standard solutions were freshly prepared and stable for at least 4 weeks in the refrigerator.
2.4.2 Acid degradation product solutions
Acid degradation product solution (NHC) (1 mg/ml) was established by accurately weighting 100 mg of NHC into a 100 ml volumetric flask in ethanol. Standard working solutions of MOL and NHC (100 µg/ml) were generated by diluting adequate aliquots of the established standard stock solutions with ethanol.
2.5 Preparation of laboratory-synthetic combinations
In two separate sets of volumetric flasks with 10 ml volume, accurate aliquots in the ranges equivalent to (10–100 µg) of MOL and (10–32 µg) of NHC were precisely transferred from the corresponding working solutions and completed to volume with ethanol and mixed well.
2.6 Procedures
2.6.1 Linearity ranges and establishment of calibration graphs
In order to achieve final concentrations between (10–150 µg/ml) MOL and (10–60 µg/ml) NHC, aliquots from the standard working solutions of MOL (100 μg/ml) and NHC (100 μg/ml) were moved separately into two sets of volumetric flasks with 10 ml volume, utilizing ethanol to complete the volume. The absorption spectra of the produced solutions for both drugs between 200 and 400 nm utilizing ethanol as a blank were measured and stored.
2.6.1.1 Dual wavelength spectrophotometric method (DW)
The calibration graph was made by estimating the difference between 264.5 nm and 299 nm, and then plotting the resulting data against different concentrations of MOL. For the determination of NHC, a difference between absorbance values at 214.9 nm and 250 nm was measured from NHC spectra and plotted against NHC concentrations. Regression equations were derived from both graphs.
2.6.1.2 Fourier self-Deconvolution spectrophotometric method (FSD)
The stored spectra were deconvoluted utilizing the FSD technique integrated into the Jasco spectrum software. The signals of the deconvoluted spectra of MOL at 217 nm and NHC at 216 nm were displayed against their concentrations. Each related regression equation was then utilized to predict either the quantities of MOL and NHC in their laboratory-synthetic combinations or MOL in Molunvac® capsules.
2.6.1.3 First derivative spectrophotometric method (1D)
The previously stored absorption spectra of the solutions were uploaded and the first derivative spectra were manipulated versus ethanol as blank. The signal was estimated at 247 and 262 nm for MOL and NHC, respectively. Then, two calibration graphs linking the signsls of MOL (at 247 nm) or NHC (at 262 nm) and the corresponding drug concentrations were created. Finally, the regression equations were developed.
2.6.1.4 Ratio difference spectrophotometric method (RD)
The stored absorption spectra of the prepared solutions were divided by the spectrum of 20 µg/ml of NHC and 40 µg/ml MOL (as divisors) for the determination of MOL and NHC, respectively. Signals of the ratio spectra were recorded at 242 and 210 nm (for MOL) and 295 and 250 nm (for NHC). Calibration graphs were constructed, and the regression equations were calculated, correlating the signal difference of the ratio spectra (ΔP242-210) and (ΔP295-250) to the concentrations corresponding to MOL and NHC, respectively.
2.6.1.5 First derivative of ratio spectra spectrophotometric method (1DD)
The ratio spectra of MOL and NHC utilized for “Ratio difference method” were manipulated to their first order utilizing Δλ = 7 employing a scaling factor of 10. The amplitudes were measured at 249 nm and 310 nm for MOL and NHC, respectively. The concentrations of drugs and regression equations were then calculated after the calibration graphs had been made by correlating the amplitudes at 249 nm and 310 nm for MOL and NHC, respectively.
2.6.2 Analysis of laboratory-synthetic combinations
The previously scanned spectra of the laboratory-synthetic combinations were treated as mentioned under “Linearity and construction of calibration graphs” for all methods. The concentrations of each drug were estimated by the previously developed regression equations.
2.6.3 Analysis of Molnuvac® capsules
Ten Molnuvac® capsules were accurately weighed. The powder content of the ten capsules was emptied and mixed well. A part of the powder weighing 100 mg MOL was weighed and transferred into flasks with 100 ml volume, 30 ml ethanol was utilized as diluent and sonication of the mixture persisted for 30 min. Utilizing ethanol, the volume reached to 100 ml mark and was mixed well. A dry funnel and dry filter paper were employed for the filtration process, with the first few milliliters being discarded. Sample stock solutions of concentration equivalent to 1 mg/ml of MOL were obtained. To produce sample working solutions of MOL (100 μg/ml), an appropriate dilution of the prepared sample stock solutions (1 mg/ml) was carried out.
In a set of 10 ml volumetric flasks, various aliquots were moved and diluted utilizing ethanol. The steps outlined under the section “Linearity and construction of calibration graphs” were applied. The methods' validity was confirmed by utilizing the standard addition approach. The mentioned drug concentrations were obtained utilizing the appropriate regression equations.
3 Results and discussion
3.1 Identification of the acid degradation product of MOL
MOL was refluxed with 1 N ethanolic HCl for one hour at 100 °C to achieve complete degradation with a high yield of the produced NHC. The refluxed solution was then neutralized and evaporated. Then, the acid degradation product was obtained by extraction and purification in hot ethanol followed by evaporation to dryness. TLC was utilized to identify the obtained product, and a new spot of NHC (Rf = 0.5) was detected, which is different from that of the intact drug (Rf = 0.75). NHC IR spectrum (the acid degradation product), (Fig. 2 (a)), illustrated the disappearance of CO stretching bands at 1753.93 cm−1, which was related to isobutyl ester and present in the IR spectrum of the intact MOL, (Fig. 2(b)). Moreover, mass spectroscopy, M+ at 258.21 which is identical to the molecular weight of NHC, (Fig. 2(c)). Also, 1H NMR spectra of MOL, (Fig. 2(d)), reveals the appearance of two singlet signal of the methyl protons of isobutyl ester at 3.36 – 3.39 ppm, respectively. While 1H NMR spectra of NHC (Fig. 2(e)) show the disappearance of these proton signals. On the other hand. 13C NMR Spectra of MOL show the presence of two signals at 19.22 – 33.63 ppm attributed to aliphatic carbon of methyl groups in isobutyl ester (Fig. 2(f)), which disappears in NHC 13C NMR spectra, (Fig. 2(g)). Additionally, two carbonyl signals are observed in the MOL spectra (Fig. 2(f)) at 149.88 and 176.42 ppm which corresponds to the carbonyl groups of the pyrimidine ring and isobutyl ester, respectively. While the carbonyl signal from the isobutyl ester disappears in the NHC 13C spectrum, but the carbonyl signal from the pyrimidine ring, which develops at 161.67 ppm, remains present, (Fig. 2(g)). The conclusions of the IR, MS and NMR spectral analyses support that a purified NHC was obtained. This delivers significant cost savings, instead of purchasing an expensive ingredient. Furthermore, the produced NHC with MOL and other antiviral medicines could be utilized in further research related to COVID-19 management.Fig. 2 IR spectrum of NHC (a), MOL (b), MS spectrum of NHC (c), 1H NMR spectrum of MOL (d), NHC (e) 13C NMR spectrum of MOL (f) and NHC (g).
3.2 Methods optimization and validation
The absorption spectrum of MOLwas overlaid with NHC absorption spectrum and exhibited severe overlapping in the range 220–320 nm, with no zero-crossing of any drug, (Fig. 3 ). The primary objective of the proposed techniques is to manipulate spectra with severe interference easily by employing easy and flexible spectrophotometric approaches excelling the time-consuming and expensive chromatographic techniques. The techniques described have been validated, then applied to the evaluation of pharmaceutical dosage form.Fig. 3 Absorption spectra of MOL (40 μg/ml) (____) and NHC (25 μg/ml) (−−−) utilizing ethanol as a blank.
3.2.1 Methods optimization
3.2.1.1 Dual Wavelength spectrophotometric method (DW).
The premise behind the dual wavelength method is the detection of a non-negligible absorbance difference between two points when the absorbance difference for any other overlapping component in the mixture is zero [27]. So, for the estimation of MOL in a mixture with NHC, numerous pairs of wavelengths at which the difference in absorbance values of NHC is equal to zero were examined. By estimating the difference in absorbance values at 299 and 264.5 nm, the accepted linearity of MOL was obtained, (Fig. 4 (a)). The difference in absorbance values at 299 and 264.5 nm, increases with the increase in MOL concentrations without contribution from NHC.Fig. 4 Absorption spectra of (a) MOL (10–150 µg/ml) and (b) NHC (10–60 µg/ml) utilizing ethanol as a blank.
On the other hand, NHC was determined without interruption from MOL utilizing the difference in absorbance values between 250 and 214.9 nm, which increases with the increase of NHC concentrations (Fig. 4(b)).
3.2.1.2 Fourier self-Deconvolution spectrophotometric method (FSD)
Fourier self-deconvolution (FSD) is a widespread band-narrowing mathematical method, allowing for the resolution of overlapped bands as an instance of the overlapped spectra of MOL and NHC. The FSD method is utilized to precisely distinguish the peak positions of each band in the measured spectra of MOL and NHC mixture which is composed of multiple spectra having bands with the same bandwidth.
Both the degree to which the deconvolved bands are narrowed (the effective narrowing) and the degree to which the signal-to-noise ratio is degraded as a result of FSD are highly dependent on the filter function utilized [28]. The band form of the deconvolved bands is also heavily influenced by the filter function. The stored absorption spectra of MOL and NHC can be resolved by utilizing FSD function in the Jasco spectra software. The produced deconvoluted spectra were utilized to determine MOL at 217 nm (Fig. 5 a) and NHC at 216 nm (Fig. 5b) in presence of NHC and MOL (zero-crossing point), respectively utilizing full width at half maximum (FWHM) value of expected Lorenz waveform = 60. Different FWHM values were tested until well-resolved spectra were produced with good and reproducible recoveries. FWHM value is an important factor that has to be studied as the program cannot be applied to resolve overlapped measured spectra having peak bands with different widths. As a result, the half-width of spectra bands before and after the deconvolution process will be taken into consideration and these are represented by FWHM values [29], [30].Fig. 5 Deconvoluted spectra of (a) MOL at 217 nm with a zero-crossing point for NHC (−−−) and (b) NHC at 216 nm with a zero-crossing point for MOL (−−−).
3.2.1.3 First derivative spectrophotometric method (1D)
In this method, absorption spectra of both MOL and NHC were manipulated to get the first derivative spectra utilizing Δ λ = 5 nm with a “scaling factor” of 10. For MOL estimation, 247 nm was chosen at which NHC was at a zero-crossing point (Fig. 6 a). So, MOL calibration graph was established at 247 nm. For NHC estimation, the measured amplitudes at 262 nm (a zero-crossing point of MOL) were constructed versus the corresponding concentration of NHC. The obtained regression equations were utilized to determine MOL and NHC concentration in laboratory-synthetic combinations (Fig. 6b).Fig. 6 First derivative absorption spectra of (a) MOL in the range (10–150 µg/ml) showing no contribution of NHC (−−−) at 247 nm and (b) NHC in the range (10–60 µg/ml) at 262 nm (zero-crossing of MOL(−−−)).
3.2.1.4 Ratio difference spectrophotometric method (RD)
The ratio difference spectrophotometric technique was utilized for resolving the overlapped spectra of MOL and NHC binary mixture without needing previous separation. The ratio difference approach [31], [32], was relied on the principle that the difference in amplitudes between two chosen points on the ratio spectra of MOL and NHC binary mixture increases with the increase of the concentration of the component of interest, irrespective of any competing components. For a mixture of MOL and NHC, MOL can be estimated by dividing the spectrum of the laboratory-prepared mixture by a known concentration of NHC (20 µg/ml) as divisor (NHC′). The division will yield a new graph (MOL/NHC′ + NHC/NHC‘ = MOL/NHC′ + constant).
The amplitudes were subtracted (P1-P2) at two chosen points λ1 = 242 nm and λ2 = 210 nm on the produced ratio graph, (Fig. 7 a), the constant NHC/NHC′ will be canceled. And so, the other drug NHC interference was canceled as follows:P242-P210=MOLNHC′242+constant-MOLNHC′210+constant=MOLNHC′242-MOLNHC′210
Fig. 7 Ratio spectra of (a) MOL utilizing NHC‘ (25 μg/ml) as a divisor and (b) NHC utilizing MOL‘ (20 μg/ml) as a divisor illustrating the two chosen points (242 nm and 210 nm) and (295 nm and 250 nm), respectively.
Then, utilizing the regression equation reflecting the linear relationship between the amplitude differences of the ratio spectra at the two chosen wavelengths against its corresponding concentration, the concentration of MOL was determined.
For the determination of NHC, the previous steps were repeated by utilizing a known concentration of MOL (40 µg/ml) as divisor (MOL′), (Fig. 7b), and illustrated by the following equation:P295-P250=NHCMOL′295+constant-NHCMOL`250+constant=NHCMOL′295-NHCMOL′250
When ratio difference method was applied, the noise and interference constants were eliminated and the spectral interference was canceled [32].
The primary functions that impact the ratio spectrum shape like of the choice of the divisor concentration and the smoothing parameter were well tested. Varying concentrations of MOL‘ (40, 50 and 60 μg/ml) and NHC‘ (10, 20 and 30 μg/ml) were tested as a divisor. It was found that the concentration of 40 μg/ml of MOL′, 20 μg/ml of NHC′ and smoothing factor 10 gave the least noise, smoother ratio spectra and better sensitivity.
3.2.1.5 The first derivative of ratio spectra spectrophotometric method (1DD)
Derivative orders were studied for MOL and NHC simultaneous determination. The first-order derivative was found to be the most satisfactory concerning selectivity and accuracy [20]. The influence of divisor concentration and smoothing factor was investigated. A divisor concentration of 40 μg/ml of MOL′ and 20 μg/ml of NHC′ gave the best results, concerning sensitivity and repeatability. Due to the extent of the noise levels on the ratio spectra, a smoothing function (10) was utilized and Δλ = 7 was found to be suitable, in terms of signal-to-noise ratio and the spectra showed good resolution. Recording the 1DD amplitudes at 249 nm and 310 nm allowed sensitive and reproducible determination of MOL and NHC, respectively, (Fig. 8 a and Fig. 8b).Fig. 8 First derivative of the ratio spectra for MOL at 249 nm (a) and NHC at 310 nm (b) utilizing NHC' (20 μg/ml and MOL' (40 μg/ml) as a divisor.
3.2.2 Methods validation
ICH guidelines and criteria were followed upon developing the current approaches [33]. (Table 1 ), introduces the all the regression results. The methods' selectivity was presented by analysis of MOL and NHC quantitatively in their laboratory-synthetic combinations, as shown in (Table 2 ). The applied approaches were also utilized for the quantification of MOL in Molunvac® capsules with no influence from the inactive ingredients present as evidenced by the results of the standard addition technique, (Table 3 ).Table 1 Regression and validation parameters of the current spectrophotometric approaches for the concurrent estimation of MOL and NHC.
Drug MOL NHC
Validation parameters DW FSD 1D RD 1DD DW FSD 1D RD 1DD
Linearity Range (µg/ml) 10–150 10–60
Regression coefficient (r) 0.9999 0.9998 0.9998 0.9996 0.9996 0.9993 0.9997 0.9998 0.9995 0.9995
Slope 0.0013 0.0235 0.0004 0.0645 0.0060 0.0069 0.0561 0.0003 0.0642 0.0035
Intercept −0.0009 0.0108 −0.00004 0.0244 −0.0239 −0.0023 −0.0140 0.000008 0.0207 −0.00043
LOD (µg/ml) 2.06 2.56 2.58 3.16 3.11 2.31 1.38 2.72 1.84 1.99
LOQ(µg/ml) 6.26 7.77 7.83 9.59 9.44 7.00 4.21 8.24 5.59 6.05
Accuracy (Meana ± SD) 101.34 ± 0.52 100.74 ± 0.78 100.55 ± 0.63 101.48 ± 0.87 99.87 ± 1.56 98.06 ± 1.33 99.28 ± 1.24 99.52 ± 0.85 100.44 ± 1.31 100.66 ± 0.91
Intra-day precision (RSD %)b 0.62 0.37 0.77 1.59 0.88 0.47 0.77 0.73 0.58 0.49
Inter-day precision (RSD %)b 0.96 0.86 1.05 1.72 1.37 0.84 1.34 0.97 1.49 1.29
a Average of nine estimations.
b RSD% of nine estimations.
Table 2 Concurrent estimation of MOL and NHC in their laboratory-synthetic combinations applying the current spectrophotometric approaches.
Conc. Taken (µg/mL) % Recovery a
MOL NHC MOL NHC
DW FSD 1D RD 1DD DW FSD 1D RD 1DD
10 10 101.43 99.28 100.76 101.43 101.43 99.65 100.66 100.54 100.65 101.58
20 16 100.07 100.32 101.84 100.21 98.54 100.34 101.23 100.98 100.88 99.23
40 24 99.41 99.44 99.64 99.84 100.19 100.87 99.76 99.81 101.65 100.73
80 32 100.88 98.55 101.18 100.98 101.44 100.88 99.63 101.54 99.22 101.22
100 10 101.21 99.64 100.93 99.42 99.89 99.56 99.72 99.78 100.11 100.45
Mean ± SD 101.04 ± 0.23 99.09 ± 0.77 101.05 ± 0.17 100.20 ± 1.10 100.66 ± 1.09 100.26 ± 0.63 100.20 ± 0.71 100.53 ± 0.75 100.50 ± 0.90 100.64 ± 0.90
aAverage of three determinations.
Table 3 Applyiny the current spectrophotometric approaches for analyzing MOL in Molunvac® capsules and application of the standard addition technique.
Conc. Taken
(μg/ml) % Recoverya
MOL
DW FSD 1D RD 1DD
Added Tablet Added Tablet Added Tablet Added Tablet Added Tablet Added Tablet
25 100.66 99.45 100.54 100.33 100.84
50 50 99.92 99.76 100.98 100.54 101.66 98.44 99.24 99.23 100.65 99.11
75 100.43 101.24 101.28 101.55 100.85
Mean 99.92 100.28 100.98 100.41 101.66 100.08 99.24 100.73 100.65 100.26
SD 0.49 0.46 0.97 0.90 1.06 1.47 1.14 1.16 0.78 1.00
a Average of three determinations.
3.3 Statistical analysis
A statistical comparison between the outcomes of the suggested approaches and the reference one [24] is shown in Table 4 . In this case, the resulting t and F values were less than the theoretical ones, showing that the established and reported procedures were equally accurate and precise. Additionally, one-way ANOVA was utilized to compare these methods. As the p-value is larger than 0.05, (Table 5 ), the suggested approaches are not significantly different from the reference procedure. These results demonstrated that the established methods could quantitatively determine MOL with good precision and considerable accuracy.Table 4 Statistical data interpretation and comparison of the results produced by establishing the current spectrophotometric approaches and the reference method for the estimation of MOL in the pharmaceutical dosage form.
MOL
Form Parameters Reference
method ** DW FSD 1D RD 1DD
Pharmaceutical dosage form Mean 101.88 99.22 100.98 101.66 99.24 100.65
SD 1.735 0.49 0.97 1.06 1.14 0.78
Variance 3.010 0.240 0.941 1.232 1.300 0.608
t-test* --- 0.28 1.11 0.27 0.32 1.58
F-ratio* --- 0.80 3.20 2.68 2.32 4.95
* The theoretical values of t and F at P = 0.05 are (2.23) and (5.05), respectively where n = 6.
**Reference HPLC method [24]: For the determination of MOL consisting of C18 column was utilized, the mobile phase consists of acetonitrile: water (20:80, v/v), UV determination of MOL has estimated at 240 nm and the flow rate was 0.5 ml/min.
Table 5 One-way ANOVA testing for the different current and reference methods utilized for the estimation of MOL in the pharmaceutical dosage form.
Source of variation Degree of freedom Sum of squares Mean square F- value p-value
MOL Between experiments 5 6.738 1.348 1.554* 0.246
Within experiments 12 10.403 0.867
Total 17 17.140
*The theoretical value of F at P = 0.05.
3.4 Assessment of greenness of the analytical methods
Green chemistry is a generally accepted idea in chemical laboratories. To precisely evaluate the environmental impact of chemical processes, specialized evaluation tools are required. For the assessment of greenness, various methodologies can be utilized. The greenness of analytical methods was validated utilizing four main parameters: quantities and hazards associated with chemical utilization, significant energy consumption, occupational hazards, and waste output. The greenness of the proposed method was assessed via the National Environmental Methods Index (NEMI), analytical Greenness calculator (AGREE) green analytical procedure index (GAPI), and analytical eco-scale.
The first system is principally based on the National Environmental Methods Index (NEMI) [34], where the greenness is designated by a pictogram divided into four quadrants. The NEMI is a qualitative technique in which pictograms indicate whether hazardous or corrosive reagents are applied or whether the process generates considerable volumes of waste.
The four quadrants of the suggested spectroscopic methods (Fig. 9 a), will be shaded in green because the solvent (ethanol) utilized in the developed procedures was neither PBT (persistent, bio-accumulative, and toxic) nor hazardous. Besides that, the pH varies between 2 and 12, and the waste generated is less than 50 g.Fig. 9 Assessment of greenness of the current spectrophotometric approaches utilizing NEMI method (a), AGREE calculator (b) and GAPI index (c).
The AGREE® program forms the backbone of the second system [35], [36], [37]. The 12 elements of Green Analytical Chemistry (GAC) are the basis for a clock-shaped graph provided by AGREE. The analytical procedure's compatibility with the green analytical chemistry concept is graded on a colour scale (red–yellow–green), where each subdivision represents one principle. The AGREE graph's central area has both a colour representing the overall rating and a numeric rating between 0 and 1 that represents how strongly the rating was assigned to each category. The calculator details the impact on the environment of many factors as reagent toxicity, waste generation, energy consumption, labour intensity, and degree of automation. The aim was to reduce harmful effects on the environment and make production methods more sustainable. A final score (0.89), reflecting the eco-friendliness of the approach, was calculated by assigning various values to the parameters utilized in the study as given in (Fig. 9b).
The green analytical procedure index, GAPI [36], [37], [38] was also utilized to evaluate green assessment. GAPI is made up of fifteen separate areas (five pentacle shapes) for estimating and evaluating the environmental impact of each step of any analytical process, beginning with preparation of sample and ending with instrumentation and waste quantity and treatment. All of these parameters are displayed on the pentagram and are colored green, yellow, or red to indicate a minimum, medium or high impact on the environment, respectively.
Although the developed spectroscopic procedures included inline preparation of sample, no storage or transfer was necessary. Also, the sample was prepared directly with no extraction steps utilizing green solvents and without any additional treatments. Sections 4 and 9, were colored yellow because normal conditions of storage were employed and the number of solvents utilized was 10–100 ml, respectively, as shown in (Fig. 9c).
In addition, the yellow color of section 5 denotes the quantification of the utilized techniques and the sample preparation (filter) utilized for the pharmaceutical dosage form. Section 14 was colored yellow because the amount of waste was greater than 10 ml (Fig. 9c). Yet, as shown in (Fig. 9c), the overall GAPI results revealed the greenness of the approaches.
The analytical Eco-Scale [36], [37], [39], [40] was also utilized to evaluate the greenness of the current spectrophotometric approaches for the concurrent analysis of MOL and NHC. This tool is a current comprehensive scale that calculates the penalty points (PPs) of any process's analytical parameters such as the nature and quantity of reagents, occupational hazards, the energy required, and waste generated. The calculated penalty points will be subtracted from 100 (the ideal green score). At higher scores, the analytical process is more cost-effective and environmentally friendly (near 100), (Table 6 ). Because of the minimal amount of waste and usage of less hazardous chemicals, the PPs score in this study was assessed to be eight and the analytical eco-scale is 92, demonstrating outstanding green behavior as shown in (Table 6).Table 6 Greenness assessment of the proposed spectrophotometric methods according to the analytical Eco-scale method.
Parameters Penalty points
Ethanol 2
Instrument
Energy consumptiona 0
Occupational hazard 0
Waste (1–10 ml, no treatment) 6
Total penalty points 8
Analytical EcoScale total scoreb,c 92
Comment Excellent green analysis
aA score of 0 is given for UV/is spectrophotometer; the energy utilized is ≤ 0.1 kWh per sample.
bAnalytical Eco-Scale total score = 100- total penalty point.
cIf the score is greater than 75, it shows excellent green analysis.
If the score is greater than 50, it shows an acceptable green analysis.
If the score is less than 50, it shows inadequate green analysis.
In conclusion, it was determined that the systems metrics assessment tools utilized were supplementary to one another for the greenness profile with the established methods, and so they can be utilized for the routine analysis of the chosen drugs with a good impact on the environment, lesser hazardous reagents, and a minimal risk of toxic effects [38].
4 Conclusion
The current research introduces simple, sensitive, precise, low-cost and environmentally friendly spectrophotometric approaches for the concurrent estimation of MOL and its active metabolite NHC as a possible acid degradation product. MOL is a prodrug for NHC, which is the most recently approved FDA drug for COVID-19 treatment. The described spectrophotometric methods can be utilized as a revolving tool for the overlapped spectra of MOL and NHC or any complex multi-mixtures. The eco-friendliness of the current methods was evaluated utilizing NEMI method, AGREE calculator, GAPI index and penalty eco-scale system.
CRediT authorship contribution statement
Michael Gamal Fawzy: Conceptualization, Methodology, Software, Visualization, Investigation, Supervision, Data curation, Validation, Writing – original draft, Writing – review & editing. Ebraam B. Kamel: Data curation, Validation, Conceptualization, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing.
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.
Data availability
Data will be made available on request.
==== Refs
References
1 2021 (https://covid19.who.int. opens in new tab) WHO coronavirus (COVID-19) dashboard. Geneva: World Health Organization, No Title.
2 C.E. Wagner, C.M. Saad-Roy, S.E. Morris, R.E. Baker, M.J. Mina, J. Farrar, E.C. Holmes, O.G. Pybus, A.L. Graham, E.J. Emanuel, Vaccine nationalism and the dynamics and control of SARS-CoV-2, Science (80-.). 373 (2021) eabj7364.
3 Nguyen K.H. Nguyen K. Corlin L. Allen J.D. Chung M. Changes in COVID-19 vaccination receipt and intention to vaccinate by socioeconomic characteristics and geographic area, United States, January 6–March 29 Ann. Med. 53 2021 2021 1419 1428 34482788
4 J.R. Arribas, S. Bhagani, S.M. Lobo, I. Khaertynova, L. Mateu, R. Fishchuk, W.Y. Park, K. Hussein, S.W. Kim, J. Ghosn, M.L. Brown, Y. Zhang, W. Gao, C. Assaid, J.A. Grobler, J. Strizki, M. Vesnesky, A. Paschke, J.R. Butterton, C. De Anda, Randomized Trial of Molnupiravir or Placebo in Patients Hospitalized with Covid-19, NEJM Evid. 1 (2022) EVIDoa2100044. https://doi.org/10.1056/EVIDoa2100044.
5 Hurt A.C. Wheatley A.K. Neutralizing antibody therapeutics for COVID-19 Virutilizes. 13 2021 628
6 A. Gupta, Y. Gonzalez-Rojas, E. Juarez, M. Crespo Casal, J. Moya, D.R. Falci, E. Sarkis, J. Solis, H. Zheng, N. Scott, A.L. Cathcart, C.M. Hebner, J. Sager, E. Mogalian, C. Tipple, A. Peppercorn, E. Alexander, P.S. Pang, A. Free, C. Brinson, M. Aldinger, A.E. Shapiro, Early Treatment for Covid-19 with SARS-CoV-2 Neutralizing Antibody Sotrovimab, N. Engl. J. Med. 385 (2021) 1941–1950. https://doi.org/10.1056/NEJMoa2107934.
7 Fischer W. Eron J.J. Holman W. Cohen M.S. Fang L. Szewczyk L.J. Sheahan T.P. Baric R. Mollan K.R. Wolfe C.R. Duke E.R. Azizad M.M. Borroto-Esoda K. Wohl D.A. Loftis A.J. Alabanza P. Lipansky F. Painter W.P. Molnupiravir, an Oral Antiviral Treatment for COVID-19., MedRxiv Prepr Serv. Heal. Sci. 2021 10.1101/2021.06.17.21258639
8 Cohen M.S. Wohl D.A. Fischer W.A. Smith D.M. Eron J.J. Outpatient Treatment of Severe Acute Respiratory Syndrome Coronavirus 2 Infection to Prevent Coronavirus Disease 2019 Progression Clin. Infect. Dis. 73 2021 2019 1717 1721 10.1093/cid/ciab494
9 Cox R.M. Wolf J.D. Plemper R.K. Therapeutically administered ribonucleoside analogue MK-4482/EIDD-2801 blocks SARS-CoV-2 transmission in ferrets Nat. Microbiol. 6 2021 11 18 33273742
10 Wahl A. Gralinski L.E. Johnson C.E. Yao W. Kovarova M. Dinnon K.H. Liu H. Madden V.J. Krzystek H.M. De C. White K.K. Gully K. Schäfer A. Zaman T. Leist S.R. Grant P.O. Bluemling G.R. Kolykhalov A.A. Natchus M.G. Askin F.B. Painter G. Browne E.P. Jones C.D. Pickles R.J. Baric R.S. Garcia J.V. SARS-CoV-2 infection is effectively treated and prevented by EIDD-2801 Nature 591 2021 451 457 10.1038/s41586-021-03312-w 33561864
11 Abdelnabi R. Foo C.S. De Jonghe S. Maes P. Weynand B. Neyts J. Molnupiravir inhibits replication of the emerging SARS-CoV-2 variants of concern in a hamster infection model J. Infect. Dis. 224 2021 749 753 34244768
12 Agostini M.L. Pruijssers A.J. Chappell J.D. Gribble J. Lu X. Andres E.L. Bluemling G.R. Lockwood M.A. Sheahan T.P. Sims A.C. Natchus M.G. Saindane M. Kolykhalov A.A. Painter G.R. Baric R.S. Denison M.R. Small-Molecule Antiviral β- d - N 4 -Hydroxycytidine Inhibits a Proofreading-Intact Coronavirus with a High Genetic Barrier to Resistance J. Virol. 93 2019 e01348 e01419 10.1128/jvi.01348-19 31578288
13 Urakova N. Kuznetsova V. Crossman D.K. Sokratian A. Guthrie D.B. Kolykhalov A.A. Lockwood M.A. Natchus M.G. Crowley M.R. Painter G.R. β-d-N 4-hydroxycytidine is a potent anti-alphavirus compound that induces a high level of mutations in the viral genome J. Virol. 92 2018 e01965 e02017 29167335
14 J. Grobler, J. Strizki, N. Murgolo, W. Gao, Y. Cao, Y. Zhang, J. Du, M. Nair, Y. Huang, Y. Luo, 543. Molnupiravir Maintains Antiviral Activity Against SARS-CoV-2 Variants In Vitro and in Early Clinical Studies, in: Open Forum Infect. Dis., Oxford University Press US, 2021: pp. S373–S373.
15 Kabinger F. Stiller C. Schmitzová J. Dienemann C. Kokic G. Hillen H.S. Höbartner C. Cramer P. Mechanism of molnupiravir-induced SARS-CoV-2 mutagenesis Nat. Struct. Mol. Biol. 28 2021 740 746 34381216
16 ICH Q1A(R2), International Conference on Harmonization (ICH). Guidance for industry: Q1A(R2) Stability testing of new drug substances and products, Ich Harmon. Tripart. Guidel. 4 (2003) 24.
17 Bahgat E.A. Hashem H. Saleh H. Kamel E.B. Eissa M.S. HPLC-DAD technique for the quantification of a recently approved anti-diabetic triple combination along with two toxic official impurities: Toxicity confirmation aided by molecular docking application BMC Chem. 17 2023 1 10 10.1186/s13065-023-00927-0 36726157
18 Fayed A.S. Hegazy M.A. Kamel E.B. Eissa M.S. Smart Mathematical Manipulation οf Spectral Signals: Stability Indicating, fοr the Estimation of Sοlifenacin Succinate: Anti-Muscarinic Drug, in Existence οf Its Acid Degradatiοn Product J. AOAC Int. 105 2022 323 331 34190982
19 E.B. Kamel, Two green chromatographic methods for the quantification of tamsulosin and solifenacin along with four of their impurities, J. Sep. Sci. 45 (2022) 1305–1316. https://analyticalsciencejournals.onlinelibrary.wiley.com/doi/abs/10.1002/jssc.202100980.
20 Bahgat E.A. Hashem H. Saleh H. Kamel E.B. Eissa M.S. Stability-indicating HPLC-DAD and TLC-densitometry methods for the quantification of bupivacaine and meloxicam in their co-formulated mixture Microchem. J. 190 2023 108683 10.1016/j.microc.2023.108683
21 Bahgat E.A. Hashem H. Saleh H. Kamel E.B. Eissa M.S. Green micelle and complex inclusion enhance synchronous spectrofluorimetric quantification of a novel analgesic combination: Tramadol and celecoxib in tablet dosage form and spiked human plasma Luminescence 38 2023 389 400 10.1002/bio.4458 36754851
22 Amara A. Penchala S.D. Else L. Hale C. FitzGerald R. Walker L. Lyons R. Fletcher T. Khoo S. The development and validation of a novel LC-MS/MS method for the simultaneous quantification of Molnupiravir and its metabolite ß-d-N4-hydroxycytidine in human plasma and saliva J. Pharm. Biomed. Anal. 206 2021 114356 10.1016/j.jpba.2021.114356
23 Kumar G.R. Babu B.S. Rao R.R. Vardhan V.M. Abbaraju V.D.N.K. Method Development and Validation of a Specific Stability Indicating RP-HPLC Method for Molnupiravir API J. Pharm. Res. Int. 2021 3026 3035 10.9734/jpri/2021/v33i60B34973
24 T. Reçber, S.S. Timur, S. Erdoğan Kablan, F. Yalçın, T.C. Karabulut, R. Neslihan Gürsoy, H. Eroğlu, S. Kır, E. Nemutlu, A stability indicating RP-HPLC method for determination of the COVID-19 drug molnupiravir applied utilizing nanoformulations in permeability studies, J. Pharm. Biomed. Anal. 214 (2022) 114693. https://doi.org/10.1016/j.jpba.2022.114693.
25 Sharaf Y.A. El Deeb S. Ibrahim A.E. Al-Harrasi A. Sayed R.A. Two Green Micellar HPLC and Mathematically Assisted UV Spectroscopic Methods for the Simultaneous Determination of Molnupiravir and Favipiravir as a Novel Combined COVID-19 Antiviral Regimen Molecules 27 2022 2330 10.3390/molecules27072330 35408729
26 Saraya R.E. El Deeb S. Salman B.I. Ibrahim A.E. Highly sensitive high-performance thin-layer chromatography method for the simultaneous determination of molnupiravir, favipiravir, and ritonavir in pure forms and pharmaceutical formulations J. Sep. Sci. 2022 10.1002/jssc.202200178
27 Eissa M.S. Fayed A.S. Hegazy M.A. Kamel E.B. Spectrophotometry analysis for simultaneous determination of the new antiviral drug combination: Daclatasvir/sofosbuvir in their pure form and pharmaceutical preparation Res. J. Pharm. Technol. 13 2020 5939 5946 10.5958/0974-360X.2020.01037.9
28 Kaupplnen J.K. Moffatt D.J. Mantsch H.H. Cameron D.G. Fourier Transforms in the Computation of Self-Deconvoluted and First-Order Derivative Spectra of Overlapped Band Contours Anal. Chem. 53 1981 1454 1457 10.1021/ac00232a034
29 Lórenz-Fonfría V.A. Padrós E. The role and selection of the filter function in Fourier self-deconvolution revisited Appl. Spectrosc. 63 2009 791 799 10.1366/000370209788701161 19589217
30 Jones R.N. Shimokoshi K. Some Observations on the Resolution Enhancement of Spectral Data by the Method of Self-Deconvolution Appl. Spectrosc. 37 1983 59 67 10.1366/0003702834634271
31 Lotfy H.M. Mohamed D. Mowaka S. A comparative study of smart spectrophotometric methods for simultaneous determination of sitagliptin phosphate and metformin hydrochloride in their binary mixture Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 149 2015 441 451 10.1016/j.saa.2015.04.076
32 Moussa B.A. Mahroutilize M.A. Fawzy M.G. Different resolution techniques for management of overlapped spectra: Application for the determination of novel co-formulated hypoglycemic drugs in their combined pharmaceutical dosage form Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 205 2018 235 242 10.1016/j.saa.2018.06.072
33 International Conference on Harmonization (ICH), Q2B: validation of analytical procedures: methodology, 62, US FDA, Federal Register, (2005).
34 Saleh S.S. Lotfy H.M. Tiris G. Erk N. Rostom Y. Analytical tools for greenness assessment of chromatographic approaches: Application to pharmaceutical combinations of Indapamide, Perindopril and Amlodipine Microchem. J. 159 2020 105557 10.1016/j.microc.2020.105557
35 Pena-Pereira F. Wojnowski W. Tobiszewski M. AGREE—Analytical GREEnness metric approach and software Anal. Chem. 92 2020 10076 10082 32538619
36 Fawzy M.G. Hassan W.E. Mostafa A.A. Sayed R.A. Different approaches for the assessment of greenness of spectrophotometric methodologies utilized for resolving the spectral overlap of newly approved binary hypoglycemic pharmaceutical mixture Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 272 2022 120998 10.1016/j.saa.2022.120998
37 Gamal Fawzy M. Saleh H. Reda A. Bahgat E.A. A green spectrophotometric method for the simultaneous determination of nasal binary mixture utilized in respiratory diseases: Applying isosbestic point and chemometric approaches as a resolving tool, greenness evaluation Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 2022 121585 10.1016/j.saa.2022.121585
38 Płotka-Wasylka J. A new tool for the evaluation of the analytical procedure: Green Analytical Procedure Index Talanta 181 2018 204 209 10.1016/j.talanta.2018.01.013 29426502
39 Gałuszka A. Migaszewski Z.M. Konieczka P. Namieśnik J. Analytical Eco-Scale for assessing the greenness of analytical procedures TrAC Trends Anal. Chem. 37 2012 61 72
40 Mahroutilize M.A. Lamie N.T. Experimental design methodology for optimization and robustness determination in ion pair RP-HPLC method development: Application for the simultaneous determination of metformin hydrochloride, alogliptin benzoate and repaglinide in tablets Microchem. J. 147 2019 691 706 10.1016/j.microc.2019.03.038
|
PMC010xxxxxx/PMC10207859.txt |
==== Front
J Clin Virol
J Clin Virol
Journal of Clinical Virology
1386-6532
1873-5967
The Authors. Published by Elsevier B.V.
S1386-6532(23)00119-1
10.1016/j.jcv.2023.105496
105496
Article
Direct comparison of clinical diagnostic sensitivity of saliva from buccal swabs versus combined oro-/nasopharyngeal swabs in the detection of SARS-CoV-2 B.1.1.529 Omicron
Puyskens Andreas a
Michel Janine a
Stoliaroff-Pepin Anna b
Bayram Fatimanur a
Sesver Akin a
Wichmann Ole b
Harder Thomas b
Schaade Lars a
Nitsche Andreas a
Peine Caroline b⁎
a Highly Pathogenic Viruses, Centre for Biological Threats and Special Pathogens, WHO Reference Laboratory for SARS-CoV-2 and WHO Collaborating Centre for Emerging Infections and Biological Threats, Robert Koch Institute, Berlin, Germany
b Department for Infectious Disease Epidemiology, Immunization Unit, Robert Koch Institute, Berlin, Germany
⁎ Corresponding author at: Robert Koch-Institute, Seestr. 10, 13353 Berlin, Germany.
24 5 2023
8 2023
24 5 2023
165 105496105496
3 3 2023
22 5 2023
© 2023 The Authors. Published by Elsevier B.V.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background/Purpose
While current guidelines recommend the use of respiratory tract specimens for the direct detection of SARS-CoV-2 infection, saliva has recently been suggested as preferred sample type for the sensitive detection of SARS-CoV-2 B.1.1.529 (Omicron). By comparing saliva collected using buccal swabs and oro-/nasopharyngeal swabs from patients hospitalized due to COVID-19, we aimed at identifying potential differences in virus detection sensitivity between these sample types.
Methods
We compare the clinical diagnostic sensitivity of paired buccal swabs and combined oro-/nasopharyngeal swabs from hospitalized, symptomatic COVID-19 patients collected at median six days after symptom onset by real-time polymerase chain reaction (PCR) and antigen test.
Results
Of the tested SARS-CoV-2 positive sample pairs, 55.8% were identified as SARS-CoV-2 Omicron BA.1 and 44.2% as Omicron BA.2. Real-time PCR from buccal swabs generated significantly higher quantification cycle (Cq) values compared to those from matched combined oro-/nasopharyngeal swabs and resulted in an increased number of false-negative PCR results. Reduced diagnostic sensitivity of buccal swabs by real-time PCR was observed already at day one after symptom onset. Similarly, antigen test detection rates were reduced in buccal swabs compared to combined oro-/nasopharyngeal swabs.
Conclusion
Our results suggest reduced clinical diagnostic sensitivity of saliva collected using buccal swabs when compared to combined oro-/nasopharyngeal swabs in the detection of SARS-CoV-2 Omicron in symptomatic individuals.
Graphical abstract
Image, graphical abstract
Keywords
SARS-CoV-2
B.1.1.529 (Omicron)
Clinical diagnostic sensitivity
Saliva
Buccal swab sampling
==== Body
pmc1 Background
Respiratory tract specimens, such as those collected by oro-/nasopharyngeal swabs, are currently recommended for the direct detection of SARS-CoV-2 by real-time polymerase chain reaction (PCR) and most antigen tests [1], [2], [3]. In November 2021, a new SARS-CoV-2 variant B.1.1.529 (Omicron) emerged and spread rapidly around the globe [4]. Several studies have suggested an improved sensitivity of saliva over upper respiratory tract specimens in the detection of SARS-CoV-2 Omicron and other variants by real-time PCR [5], [6], [7], [8]. Saliva could offer an appealing alternative to oro- and/or nasopharyngeal swabs as sample collection is considered less invasive and could potentially be easily performed by caretakers and patients themselves [9]. Hence, the aim of this study was to assess the diagnostic performance of saliva collected using buccal swabs versus oro-/nasopharyngeal swab samples in the detection of the SARS-CoV-2 Omicron variant. To do so, we compared the clinical diagnostic sensitivity of matched buccal and combined oro-/nasopharyngeal swabs collected from hospitalized, symptomatic individuals by real-time PCR and antigen test.
2 Methods
2.1 Study design and sample collection
Clinical specimens were collected as part of the COViK study conducted by the Robert Koch Institute in collaboration with the Paul-Ehrlich-Institut [[10], [11]]. Thirteen hospitals across Germany served as study sites. Samples were collected between January and March 2022. Sampling was performed on symptomatic individuals on day six (median) after symptom onset. Trained study nurses performed sampling, using swabs of identical design (eSwab™, COPAN Diagnostics, Murrieta, CA, USA) for both buccal and combined oro-/nasopharyngeal sample collection. A total of 107 matched sample pairs consisting of one buccal and one combined oro-/nasopharyngeal swab were collected. Collection of saliva using buccal swabs was performed immediately before collection of the combined oro-/nasopharyngeal swab. Prior to buccal swabbing participants were asked to think of their favorite food for approximately 0.5–1.0 minute to stimulate saliva flow. Buccal swab samples were collected by streaking both the left and right lower inner cheek for 30 seconds each while applying light pressure to the swab and rotating it around its own axis allowing for full saturation of the swab tip with saliva. Subsequently, a fresh swab was used for oropharyngeal sampling directly followed by nasopharyngeal sampling using the same swab. After sample collection swabs were transferred to their respective collection tubes containing transport medium. Matched buccal and combined oro-/nasopharyngeal swabs were shipped together at 2–8 °C and were stored at 4 °C upon arrival, ensuring identical transport and storage conditions for matched samples. Time from sampling to result were 2 days (median), while nearly half of all samples (47.93%) required only 1 day from sampling to result.
2.2 RNA extraction and real-time PCR analysis
To extract viral RNA from samples, 140 µl of swab-containing transport medium were manually inactivated using AVL+Ethanol and extracted using the QIAamp Viral RNA Mini kit (QIAGEN, Hilden, Germany) according to manufacturer's instructions. SARS-CoV-2 RNA was detected by real-time PCR using two separate assays, each targeting a distinct SARS-CoV-2 genomic region (E-gene and Orf1ab) as has been described previously [12]. For the identification of SARS-CoV-2 variants, PCR positive samples were further analyzed using variant-specific PCR assays and/or next-generation sequencing [13].
2.3 Antigen testing
For antigen detection, the Panbio™ COVID-19 Ag Rapid Test Device (Abbott Rapid Diagnostics, Jena, Germany) was used. For testing, 50 µl of the native swab-containing transport medium were transferred directly to the test-specific extraction buffer and subsequent testing was performed according to manufacturer's instructions. Results were analyzed independently by two trained laboratory technicians. If results were not in agreement a third person analyzed the test and the result in favor was noted. All antigen tests included in this study showed a visible control line.
3 Results
First, we compared clinical diagnostic sensitivities for the detection of SARS-CoV-2 Omicron from matched buccal and combined oro-/nasopharyngeal swabs by real-time PCR. In total, 107 matched sample pairs were collected at median six days after symptom onset from previously confirmed SARS-CoV-2 positive hospitalized, symptomatic individuals. Of those, 11 sample pairs (10.28%) tested PCR negative for SARS-CoV-2 in both buccal and oro-/nasopharyngeal swabs. Of the positive samples, all were identified as the SARS-CoV-2 Omicron variant (55.8% BA.1 and 44.2% BA.2). Only two oro-/nasopharyngeal swabs (1.87%) tested PCR negative while the matched buccal swabs tested PCR positive (Fig. 1 A). In contrast, 17 buccal swabs (15.89%) tested PCR negative, while the matched oro-/nasopharyngeal swabs tested PCR positive, resulting in a higher number of false-negative real-time PCR results for buccal swabs in comparison to combined oro-/nasopharyngeal swabs (Fig. 1A). Comparing only those sample pairs that tested PCR positive for both buccal and oro-/nasopharyngeal swab, real-time PCR from buccal swabs resulted in significantly higher Cq values compared to matching oro-/nasopharyngeal swabs with a difference in means for E-gene of 7.36 Cq (CI 6.23 to 8.5) and Orf1ab of 7.2 Cq (CI 6.1 to 8.3) (Fig. 1B). Overall, lower Cq values in buccal swabs were observed for only 7 (E-gene) and 8 (Orf1ab) sample pairs. Notably, reduced performance of buccal swabs was observed for both SARS-CoV-2 Omicron BA.1 and BA.2 (Supplementary Fig. 1). Higher Cq values in buccal swabs were detected as early as day one to two after symptom onset (Fig. 2 ). We also tested detection performance of matched buccal and combined oro-/nasopharyngeal swabs by antigen test. While the positive detection rate by antigen test for combined oro-/nasopharyngeal swab samples was 58.44% (45/77 samples), positive detection rate for buccal swab samples was only 3.9% (3/77 samples) (Fig. 3 ).Fig. 1 Comparison of SARS-CoV-2 B1.1.529 Omicron detection performance of matched buccal and combined oro-/nasopharyngeal swabs by real-time PCR. A Comparison of cycle threshold (Cq) values for two distinct genomic regions of SARS-CoV-2 (E-gene and Orf1ab) of matched buccal and combined oro-/nasopharyngeal swabs (ONS) by quantitative real-time PCR (RKI/ZBS1 SARS-CoV-2 protocol, Michel et al. Virol J (2021) 18:110); n = 107. B Estimation plot of SARS-CoV-2 positive sample pairs with Cq values ≤45; n = 77. Line at mean with 95% Confidence Interval; Paired t-test, p **** <0,0001.
Fig 1
Fig. 2 Comparison of real-time PCR results from buccal and combined oro-/nasopharyngeal swabs at different time periods after symptom onset. Comparison of cycle threshold (Cq) values for two distinct genomic regions of SARS-CoV-2 (E-gene and Orf1ab) of matched buccal and combined oro-/nasopharyngeal swabs (ONS) by quantitative real-time PCR (RKI/ZBS1 SARS-CoV-2 protocol, Michel et al. Virol J (2021) 18:110) at different days post (d.p.) symptom onset. Data displayed as Tukey box plot.
Fig 2
Fig. 3 Comparison of SARS-CoV-2 B.1.1.529 Omicron detection performance of matched buccal and combined oro-/nasopharyngeal swabs by antigen test. PCR positive buccal and combined oro-/nasopharyngeal (ONS) sample pairs (n = 77) were tested using the Panbio™ COVID-19 Ag Rapid Test Device. Shown are the number of positive and negative antigen test results. All tests showed a visible control line.
Fig 3
4 Discussion
In this study, we observed reduced clinical diagnostic sensitivity of saliva collected using buccal swabs in comparison to matched combined oro-/nasopharyngeal swabs in the detection of SARS-CoV-2 Omicron (BA.1 and BA.2). Several studies on the sensitivity of saliva versus respiratory tract specimens for the detection of SARS-CoV-2, including the Omicron variant, have been conducted, leading to mixed and in parts contradictory results [[5], [6], [7], [8],[14], [15], [16]]. In this study, samples were collected from hospitalized, symptomatic individuals who had previously been confirmed to be SARS-CoV-2 positive, resulting in sample collection at median six days after initial symptom onset. We observed that around 10% of initially PCR positive individuals were negative by the time of the second PCR testing, probably due to the relatively late time of sampling. Despite the majority of samples being collected at late stages of infection, higher Cq values in buccal swabs were detected already from day one to two after symptom onset in this limited data set. In a recent study, Lai et al. compared saliva and nasal swabs from close contacts of COVID-19 cases over time and found that, in those contacts who became infected, saliva samples showed higher viral loads compared to those in nasal swabs from three days prior to symptom onset to two days after symptom onset [17]. In contrast, two days after symptom onset there was a trend towards improved sensitivity with nasal swabs compared to saliva, indicating the importance of time of sampling for subsequent specimen sensitivity [17]. Furthermore, we applied buccal swabbing to collect saliva using swabs of identical design for both the collection of saliva and oro-/nasopharyngeal specimens. Using identical swabs enabled direct comparison between specimen types by ensuring identical conditions for transport and handling during all downstream manipulations, including RNA extraction. While the swabs used in this study are suitable for versatile applications, they are not specifically designed for buccal swabbing, which ultimately could impact saliva sensitivity. We did not assess other types of swabs and saliva sampling methods, such as drooling, spitting or sampling from specific salivary glands or other locations, which might further impact subsequent saliva sensitivity. Overall, factors such as the time of sampling and specific sampling methods are likely to play a critical role in the diagnostic sensitivity of saliva and might explain some of the differences found across studies.
In addition to real-time PCR, we also performed antigen testing using the Panbio™ COVID-19 antigen rapid test, which resulted in substantially reduced detection rates among buccal swab samples in comparison to combined oro-/nasopharyngeal swab samples. While reduced performance of this specific antigen test for the detection of SARS-CoV-2 Omicron is not predicted by the manufacturer due to the use of the nucleocapsid (N) protein as target antigen [18], it has previously been shown that the use of throat and saliva samples with the Panbio™ COVID-19 antigen rapid test led to poorer sensitivity compared to nasopharyngeal swab samples [19]. Although all swab samples in this study were subject to prior dilution in transport medium, it is not clear whether the reduced performance of buccal swab samples is due to a reduced concentration of N protein in buccal saliva or whether saliva is a suboptimal sample type for use in the Panbio™ COVID-19 antigen rapid test.
At the time of study, SARS-CoV-2 Omicron BA.1 and BA.2 were the dominant variants present in Germany [20], which is also reflected in our sample set. A study using ex vivo infections of different tissues found that SARS-CoV-2 Omicron BA.2 displayed increased replication competence in human nasal and bronchial tissues compared to Omicron BA.1 as well as the original SARS-CoV-2 wild-type strain and the Delta variant [21]. It remains to be elucidated how recently emerged and currently dominant variants of SARS-CoV-2 might affect diagnostic sensitivities of different specimen types.
Taken together, despite the reduced invasiveness and ease of sampling, the use of saliva collected by buccal swabs displays substantially reduced sensitivity in comparison to combined oro-/nasopharyngeal swab specimens for the detection of SARS-CoV-2 Omicron. This further highlights the importance to carefully consider time and context of sampling for choosing the optimal specimen type for diagnostics.
Ethical statement
The study obtained ethical approval by the Berliner Ärztekammer (Berlin Chamber of Physicians, Eth 20/40).
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.
Appendix Supplementary materials
Image, application 1
Acknowledgements
The authors thank all study nurses for the valuable contribution, namely Sawsanh Al-Ogaidi, Nancy Beetz, Belgin Esen, Rola Khalife, Katja Lange, Luise Mauer, Antje Micheel, Marlies Schmidt, Yvonne Weis, Franziska Weiser and Aysete Yencilek. The authors are grateful to Ursula Erikli for copyediting. This work was funded by the German Federal Ministry of Health.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jcv.2023.105496.
==== Refs
References
1 RKI. Coronavirus SARS-CoV-2 - Hinweise zur Testung von Patientinnen und Patienten auf SARS-CoV-2, (n.d.). https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Vorl_Testung_nCoV.html;jsessionid=3C0BAB0295736A1BE9A8524E9F42A5A6.internet071?nn=13490888#doc13490982bodyText1 (accessed December 6, 2022).
2 Interim Guidelines for Clinical Specimens for COVID-19 | CDC, (n.d.). https://www.cdc.gov/coronavirus/2019-ncov/lab/guidelines-clinical-specimens.html (accessed December 6, 2022).
3 Laboratory testing for 2019 novel coronavirus (2019-nCoV) in suspected human cases, (n.d.). https://www.who.int/publications/i/item/10665-331501 (accessed December 6, 2022).
4 Hadfield J. Megill C. Bell S.M. Huddleston J. Potter B. Callender C. Sagulenko P. Bedford T. Neher R.A. NextStrain: real-time tracking of pathogen evolution Bioinformatics 34 2018 4121 4123 10.1093/bioinformatics/bty407 29790939
5 Marais G. Hsiao N. Iranzadeh A. Doolabh D. Enoch A. Chu C. Williamson C. Brink A. Hardie D. Saliva swabs are the preferred sample for Omicron detection MedRxiv 2021 10.1101/2021.12.22.21268246 2021.12.22.21268246
6 Cornette M. Decaesteker B. Martens G.A. Vandecandelaere P. Jonckheere S. From Delta to Omicron SARS-CoV-2 variant: switch to saliva sampling for higher detection rate J. Clin. Virol. Plus. 2 2022 100090 10.1016/j.jcvp.2022.100090
7 Jo S.J. Kim J. Kim S. Kim J. Kwon H. Kwon Y. Kim H. Kim H.H. Lee H. Kim S.W. Yeo C.D. Lee S.H. Lee J. Simple saliva sample collection for the detection of SARS-CoV-2 variants compared to nasopharyngeal swab sample Arch. Pathol. Lab. Med. 2022 146 10.5858/arpa.2022-0298-sa
8 Migueres M. Mansuy J.M. Vasseur S. Claverie N. Lougarre C. Soulier F. Trémeaux P. Izopet J. Omicron wave SARS-CoV-2 diagnosis: evaluation of saliva, anterior nasal, and nasopharyngeal swab samples Microbiol. Spectr. 2022 10.1128/spectrum.02521-22
9 Bellagambi F.G. Lomonaco T. Salvo P. Vivaldi F. Hangouët M. Ghimenti S. Biagini D. Di Francesco F. Fuoco R. Errachid A. Saliva sampling: methods and devices. An overview TrAC - Trends Anal. Chem. 124 2020 115781 10.1016/j.trac.2019.115781
10 Stoliaroff-Pepin A. Peine C. Herath T. Lachmann J. Perriat D. Dörre A. Nitsche A. Michel J. Grossegesse M. Hofmann N. Rinner T. Kohl C. Brinkmann A. Meyer T. Dorner B.G. Stern D. Treindl F. Hein S. Werel L. Hildt E. Gläser S. Schühlen H. Isner C. Peric A. Ghouzi A. Reichardt A. Janneck M. Lock G. Schaade L. Wichmann O. Harder T. Effectiveness of vaccines in preventing hospitalization due to COVID-19: a multicenter hospital-based case-control study, June 2021 to January 2022, Germany Vaccine 41 2 2023 290 293 10.1016/j.vaccine.2022.11.065 Epub 2022 Dec 2. PMID: 36509640; PMCID: PMC9715487 36509640
11 Stoliaroff-Pepin A. Peine C. Herath T. Lachmann J. Hellenbrand W. Perriat D. Dörre A. Nitsche A. Michel J. Grossegesse M. Hofmann N. Rinner T. Kohl C. Brinkmann A. Meyer T. Stern D. Treindl F. Dorner B.G. Hein S. Werel L. Hildt E. Gläser S. Schühlen H. Isner C. Peric A. Ghouzi A. Reichardt A. Janneck M. Lock G. Huster D. Grünewald T. Schaade L. Wichmann O. Harder T. Vaccine effectiveness against severe COVID-19 during the Omicron wave in Germany: results from the COViK study Infection 2023 1 10 10.1007/s15010-023-02012-z Epub ahead of print. PMID: 36913112; PMCID: PMC10009838
12 Michel J. Neumann M. Krause E. Rinner T. Muzeniek T. Grossegesse M. Hille G. Schwarz F. Puyskens A. Förster S. Biere B. Bourquain D. Domingo C. Brinkmann A. Schaade L. Schrick L. Nitsche A. Resource-efficient internally controlled in-house real-time PCR detection of SARS-CoV-2 Virol. J. 18 2021 10.1186/s12985-021-01559-3
13 Brinkmann A. Ulm S.L. Uddin S. Förster S. Seifert D. Oehme R. Corty M. Schaade L. Michel J. Nitsche A. AmpliCoV: rapid whole-genome sequencing using multiplex PCR amplification and real-time oxford nanopore MinION sequencing enables rapid variant identification of SARS-CoV-2 Front. Microbiol. 12 2021 1 9 10.3389/fmicb.2021.651151
14 Uršič T. Kogoj R. Šikonja J. Roškarič D. Jevšnik Virant M. Bogovič P. Petrovec M. Performance of nasopharyngeal swab and saliva in detecting Delta and Omicron SARS-CoV-2 variants J. Med. Virol. 94 2022 4704 4711 10.1002/jmv.27898 35642439
15 Czumbel L.M. Kiss S. Farkas N. Mandel I. Hegyi A. Nagy Á. Lohinai Z. Szakács Z. Hegyi P. Steward M.C. Varga G. Saliva as a candidate for COVID-19 diagnostic testing: a meta-analysis Front. Med. 7 2020 465 10.3389/fmed.2020.00465
16 Jamal A.J. Mozafarihashjin M. Coomes E. Powis J. Li A.X. Paterson A. Anceva-Sami S. Barati S. Crowl G. Faheem A. Farooqi L. Khan S. Prost K. Poutanen S. Taylor M. Yip L. Zhong X.Z. McGeer A.J. Mubareka S. Coleman B.L. Chen D. Farshait N. Gold W. Kandel C.E. Katz K. Kozak R. Mazzulli T. Muller M. Opavsky A. Ostrowski M. Plevneshi A. Rau N. Ricciuto D. Richardson D. Rose D. Sales V. Walmsley S. Sensitivity of nasopharyngeal swabs and saliva for the detection of severe acute respiratory syndrome Coronavirus 2 Clin. Infect. Dis. 72 2021 1064 1066 10.1093/cid/ciaa848 32584972
17 Lai J. German J. Hong F. Tai S.-H.S. McPhaul K.M. Milton D.K. Comparison of saliva and midturbinate swabs for detection of SARS-CoV-2 Microbiol. Spectr. 2022 10.1128/spectrum.00128-22
18 Evaluating Omicron and other COVID variants to ensure test effectiveness | Abbott Newsroom, (n.d.). https://www.abbott.com/corpnewsroom/diagnostics-testing/monitoring-covid-variants-to-ensure-test-effectiveness.html (accessed December 13, 2022).
19 Stokes W. Berenger B.M. Portnoy D. Scott B. Szelewicki J. Singh T. Venner A.A. Turnbull L.A. Pabbaraju K. Shokoples S. Wong A.A. Gill K. Guttridge T. Proctor D. Hu J. Tipples G. Clinical performance of the Abbott Panbio with nasopharyngeal, throat, and saliva swabs among symptomatic individuals with COVID-19 Eur. J. Clin. Microbiol. Infect. Dis. 40 2021 1721 1726 10.1007/s10096-021-04202-9 33742322
20 RKI - Coronavirus SARS-CoV-2 - Wochenbericht vom 24.3.2022, (n.d.). https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Situationsberichte/Wochenbericht/Wochenbericht_2022-03-24.html (accessed December 5, 2022).
21 Hui K.P.Y. Ng K.C. Ho J.C.W. Yeung H.W. Ching R.H.H. Gu H. Chung J.C.K. Chow V.L.Y. Sit K.Y. Hsin M.K.Y. Au T.W.K. Poon L.L.M. Peiris M. Nicholls J.M. Chan M.C.W. Replication of SARS-CoV-2 Omicron BA.2 variant in ex vivo cultures of the human upper and lower respiratory tract EBioMedicine 2022 83 10.1016/j.ebiom.2022.104232
|
PMC010xxxxxx/PMC10207860.txt |
==== Front
Diabetes Res Clin Pract
Diabetes Res Clin Pract
Diabetes Research and Clinical Practice
0168-8227
1872-8227
Elsevier B.V.
S0168-8227(23)00494-1
10.1016/j.diabres.2023.110731
110731
Article
Global acceptance of COVID-19 vaccine among persons with diabetes: A systematic review and meta-analysis
Ekpor Emmanuel ab⁎
Akyirem Samuel c
a School of Nursing and Midwifery, University of Ghana, Legon, Accra, Ghana
b Christian Health Association of Ghana, Accra, Ghana
c Yale School of Nursing, Yale University, New Haven, CT, USA
⁎ Corresponding author at: School of Nursing and Midwifery, University of Ghana, Legon, Accra, Ghana.
24 5 2023
7 2023
24 5 2023
201 110731110731
22 4 2023
18 5 2023
21 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Aim
This review aimed to estimate the level of acceptance of COVID-19 vaccine among persons with diabetes.
Methods
A systematic search was conducted on PubMed, MEDLINE, Embase, and CINAHL to identify relevant studies for this review. A random-effects meta-analysis was performed to generate an overall estimate of vaccine acceptance. The I2 statistic was used to quantify the degree of variation across studies, and subgroup analysis was conducted to identify the sources of heterogeneity. The review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA).
Results
This review included 18 studies involving 11,292 diabetes patients. The pooled prevalence of COVID-19 vaccine acceptance among persons with diabetes was 76.1% (95% CI: 66.7%–83.5%). The pooled prevalence across the continent ranged from 68.9% (95% CI: 47.8%–84.3%) in Asia to 82.1% (95% CI: 80.2%–83.8%) in Europe. Barriers to vaccine acceptance included misinformation, lack of information, mistrust, health concerns, and external influences.
Conclusion
The barriers to vaccine acceptance identified in this review, could inform the formulation of health policies and public health interventions that are specifically tailored to address the needs of persons with diabetes.
Keywords
Diabetes
COVID-19 vaccine
Vaccine acceptance
Systematic review
Meta-analysis
==== Body
pmc1 Introduction
The COVID-19 pandemic has had a devastating global impact and has led to unprecedented disruptions to healthcare systems, economies, and daily life. COVID-19 is said to have no boundaries and thus affects all individuals irrespective of their composition. However, the clinical spectrum of the disease disproportionately impacts persons with chronic disease, including diabetes [1]. Diabetes has been identified as a significant risk factor for contracting COVID-19 and is associated with higher rates of hospitalization in intensive care units [2], [3]. Additionally, individuals with diabetes are nearly two times more likely to experience COVID-19-related mortality compared to those without the condition [3].
The COVID-19 pandemic has prompted a robust response from the healthcare community, including the development and deployment of vaccines. These vaccines have demonstrated the potential to effectively curb the transmission of the virus and mitigate its severe consequences [4]. However, the success of vaccination efforts is contingent upon the willingness and preparedness of the population to accept and receive vaccines [5]. This is particularly crucial for vulnerable populations, such as persons with diabetes. Nonetheless, there have been reports of vaccine hesitancy among some diabetes patients [6], [7], despite their prioritization in early vaccine distribution programs [8]. This is partly driven by the belief that COVID-19 is not dangerous to diabetes patients’ health and that vaccination does not reduce the risk of infection [9]. Furthermore, safety and efficacy concerns associated with COVID-19 vaccines have been identified as critical predictors of vaccine acceptance [10].
Vaccine hesitancy is a complex issue that impedes efforts against vaccine-preventable diseases [11]. Considering the serious ramifications of COVID-19 on diabetes patients, vaccination is an essential aspect of diabetes management and a vital approach for mitigating the impact of the pandemic on this patient population. This underscores the importance of investigating the rate of vaccine acceptance among persons with diabetes. However, to date, there has been a lack of a comprehensive review synthesizing the acceptance of COVID-19 vaccines among individuals with diabetes. The extent of COVID-19 vaccine acceptance among diabetes patients remains unclear, as individual studies on this topic have reported varying acceptance rates. Therefore, this review aimed to estimate the level of acceptance of COVID-19 vaccines among persons with diabetes by synthesizing the results of relevant studies across the globe. The barriers to COVID-19 vaccine acceptance were also assessed.
2 Materials and methods
Prior to this study, an extensive search of electronic medical databases was done to explore the available literature and to identify evidence gaps. Subsequently, a protocol for this review was developed and registered on PROSPERO (CRD42022371963). This review adhered to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines.
2.1 Inclusion and exclusion criteria
The present study adopted the CoCoPop framework (condition, context, and population) to establish the inclusion criteria. This framework was used given its wide recognition in systematic review assessing prevalence or incidence data [12], [13]. As such, the inclusion criteria were comprised of primary studies that quantitatively assessed COVID-19 vaccine acceptance (willingness or hesitancy) among persons with diabetes. No restrictions were placed on the study context, provided that the research was conducted in a real-world setting. In addition, only peer-reviewed articles published in the English language were considered for inclusion.
Priority was given to studies that focused primarily on diabetes patients. However, studies conducted on the general population were included if the COVID-19 vaccine acceptance rate was extractable for diabetes patients who were part of the study. On the other hand, review articles, preprints, case reports, conference abstracts, posters, and letters to the editor were excluded. In cases where duplicate studies were identified, wherein a single study was reported in multiple publications, only one study was selected based on its relevance to the outcome of interest and its high methodological quality.
2.2 Outcome definition
The principal outcome of interest of this study was vaccine acceptance among persons with diabetes, operationalized as the proportion of participants who indicated their willingness to receive the COVID-19 vaccine or were not hesitant to take the vaccine. As a secondary objective, barriers to vaccine acceptance were defined as reasons for diabetes patients’ unwillingness or hesitancy to receive the COVID-19 vaccine.
2.3 Search strategy
A comprehensive systematic search was conducted on November 20, 2022 to identify relevant literature pertaining to the topic of interest. The search was executed on four major electronic databases: PubMed, MEDLINE, Embase, and Cumulative Index to Nursing and Allied Health Literature (CINAHL). To ensure exhaustiveness, the search was supplemented with additional searches on specialized databases dedicated to COVID-19 research, such as LitCovid and the World Health Organization's COVID-19 research database. A manual examination of reference lists from eligible articles was also conducted to ensure the inclusion of all relevant studies.
The search was updated on February 01, 2023 to include recently indexed studies, with the aim of ensuring the comprehensiveness of the results. The initial search was limited to articles published between the years 2020 and 2022 to reflect the most recent and up-to-date information.
The search strategy was informed by the population and outcome of interest and was developed based on a combination of key terms including “COVID-19 vaccine,” “hesitancy or uptake,” and “diabetes.” Boolean combinations (AND, OR, NOT) of these terms, along with database-specific index terms, were utilized to optimize the search results. The full details of the search strategy are presented in the Supplementary file.
2.4 Screening and study selection
Duplicates from retrieved articles were removed using EndNote 20. The remaining articles were then uploaded to Rayyan [14] for title and abstract screening based on the predefined eligibility criteria. The full text of selected studies was retrieved and assessed to ascertain if they met the requirements for inclusion in this review.
2.5 Data extraction
A standardized data extraction form in Excel format was designed to retrieve important information from the included studies. The extracted data included: first author’s name, year of publication, country, study design, type of diabetes, and data on the COVID-19 vaccination acceptance. Both reviewers independently extracted the data.
2.6 Quality assessment
This review comprised of studies that provided data on the prevalence or proportion of COVID-19 vaccine uptake or hesitancy among diabetes patients. Hence, the Joanna Briggs Institute (JBI) critical appraisal checklist for studies reporting prevalence data was used to assess the methodological quality of the included studies [15]. The appraisal tool addresses 9 critical questions and has response options such as “Yes”, “No”, “Unclear” and “Not applicable”. In the absence of a defined cut-off point for the quality score of studies, we denote studies with 50% or more “Yes” across the quality assessment parameters as low risk. A comprehensive description of the quality assessment tool is presented in the Supplementary file.
2.7 Statistical analysis
The meta-analysis was performed based on a random-effects model. This model was used given the considerable variability across the included studies. Furthermore, we computed the pooled prevalence of COVID-19 vaccine acceptance using a generalized linear mixed model with the logit transformation, as recommended by Warton and Hui [16]. The 95% confidence interval (CI) for the proportion was calculated using the Clopper-Pearson interval and results from the analysis presented in forest plots. We quantified the proportion of variability due to heterogeneity across studies using the I2 statistic, with values of 25%, 50%, and 75% indicating low, moderate, and high levels of heterogeneity, respectively [17]. Subgroup analysis was done for gender, continent, data collection method, sample size, and diabetes participants as the primary focus of the study. The presence of publication bias was evaluated using funnel plots and was statistically explored using the test proposed by Peters et al [18]. The meta-package in R statistical software was used for the statistical analysis.
3 Results
3.1 Search results
Our systematic search yielded a total of 2499 studies, which comprised of 2464 records from four databases (PubMed, MEDLINE, Embase, and CINAHL) and 35 records from others sources. After duplicate records were removed and 2351 articles had been screened for their titles and abstracts, the full text of 62 articles were assessed. Finally, 18 articles met our inclusion criteria for this study. A summary of the steps involved in the screening process and reasons for exclusion of articles after full-text review are provided in Fig. 1 .Fig. 1 PRISMA flow chart summarizing the article selection process.
3.2 Characteristics of included studies
All included studies consisted of cross-sectional designs. However, 1 study used a mixed-method approach to evaluate the responses of the participants [19]. Majority of the studies were published in 2022 (n = 12). The study involved 11,292 diabetes participants, the majority of which were females (54.1%) as per studies that reported both gender proportions. Apart from Osuagwu et al’s study which recruited participants from multiple countries in Sub-Saharan Africa, all the remaining studies involved a single country. Grouped under the continent of study, Asia dominated with 8 studies [6], [9], [20], [21], [22], [23], [24], [25], followed by North America (n = 5) [26], [27], [28], [29], [30], Europe (n = 2) [7], [31], Africa (n = 2) [19], [32], and Australia (n = 1) [33]. A summary of the characteristics of the included studies is provided in Table 1 .Table 1 Characteristics of the included studies.
First author (Year) Country Type of diabetes Study Design Diabetes sample Female
proportion Vaccine acceptance Quality assessment score Quality of study
Wang (2022) China Both CS 483 47.8 210 7 High
Lu (2022) China T2DM CS 170 51.8 131 5 High
Tsai (2022) USA T2DM CS 1400 NR 1134 5 High
Day (2022) Australia NI CS 842 44.9 742 7 High
Osuagwu (2022) SSA NI MM 73 34.2 48 5 High
Asadi-Pooya
(2022) Iran NI CS 127 NR 108 5 High
Kolobov (2022) Israel Both CS 308 61.2 127 5 High
Mesele (2022) Ethiopia Both CS 386 30.6 319 7 High
Velario (2022) Canada NI CS 193 NR 114 7 High
Kanyangarara (2022) USA NI CS 1210 NR 1135 5 High
Syed (2021) Malaysia NI CS 97 NR 73 8 High
Czeisler (2021) USA NI CS 760 59.1 597 5 High
Guaraldi (2021) Italy T2DM CS 1176 73.1 967 5 High
Aldossari (2021) Saudi Arabia Both CS 709 59.5 257 7 High
Scoccimarro
(2021) Italy Both CS 502 60.2 410 5 High
Waite (2021) Canada NI CS 744 NR 589 6 High
Abedin (2021) Bangladesh NI CS 488 NR 331 7 High
Okobo (2021) Japan NI CS 1628 NR 1531 8 High
CS: Cross-sectional, MM: Mixed method, NI: Not indicated, NR: Not reported, T2DM: Type 2 diabetes mellitus.
3.3 Quality of included studies
The quality assessment of the included studies identified that all studies had a score of 50% and above with a mean score of 67.3%. The authors (EE and SA) involved in the quality assessment of the included studies agreed on almost 85% of the scores awarded. Disagreements were discussed and consensus was attained. Results of the quality assessment are presented in Table 1.
3.4 Meta-analysis of COVID-19 vaccine acceptance
The proportion of persons with diabetes who accepted the COVID-19 vaccine ranged from 36.2% [6] to 94.0% [25]. The meta-analysis showed that the pooled prevalence of COVID-19 vaccine acceptance was 76.1% (95% CI: 66.7 – 83.5). There was a significantly high heterogeneity among the studies (I2 = 99%, p < 0.01) as shown in Fig. 2 .Fig. 2 COVID-19 vaccine acceptance among persons with diabetes.
Results from the subgroup analysis are presented in Table 2 . Subgroup analysis per continent of the study revealed regional variability in vaccine acceptance, with the highest acceptance reported in Europe 82.1% (95% CI: 80.2 – 83.8) and the lowest in Asia 68.9% (95% CI: 47.8 – 84.3). In terms of gender-based comparisons, the acceptance of COVID-19 vaccine was higher in male diabetes patients 73.2% (95% CI: 54.3 – 86.3) compared to their female counterparts 59.1% (95% CI: 40.2 – 75.8). Moreover, studies with a sample size > 500, conducted online and primarily focused on non-diabetes participants, had a high proportion of COVID-19 vaccine acceptance. No heterogeneity was identified in studies from Europe. However, the result was insignificant (I2 = 0, p = 0.79).Table 2 Subgroups analysis of COVID-19 vaccine acceptance.
Variables No. of studies Proportion
(95% CI) I2 p-value
Sample size <500
>500 9
9 67.7 (55.9–77.7)
82.6 (71.3–90.0) 97%
99% P < 0.01
P < 0.01
Gender Male
Female 6
6 73.2 (54.3–86.3)
59.1 (40.2–75.8) 96%
96% P < 0.01
P < 0.01
Continent Asia
North America
Europe
Africa 8
5
2
2 68.9 (47.8–84.3)
80.8 (70.8–88.0)
82.1 (80.2–83.8)
75.6 (56.0–88.3) 99%
98%
0%
90% P < 0.01
P < 0.01
P = 0.79
P < 0.01
Data collection method In person
Online 6
12 71.7 (56.8–83.0)
78.0 (66.2–86.6) 98%
99% P < 0.01
P < 0.01
Diabetes as primary focus Yes
No 9
9 67.5 (51.9–80.0)
83.0 (74.5–89.0) 99%
98% P < 0.01
P < 0.01
3.5 Publication bias
The results of the funnel plot analysis, presented in Fig. 3 , revealed an asymmetrical distribution of the studies, suggesting the potential presence of publication bias. In contrast, the results from the statistical tests for publication bias, as assessed through Peters' test, indicated no evidence of publication bias (p-value = 0.3406). The divergent results between the funnel plot analysis and the statistical test may be a consequence of the high heterogeneity of the studies included in the analysis, rather than publication bias, as noted by Sterne et al [34].Fig. 3 Funnel plot for COVID-19 vaccine acceptance among persons with diabetes.
3.6 Barriers to vaccine acceptance
Reasons behind diabetes patients’ unwillingness to accept COVID-19 vaccine were extracted from 5 studies. In all, nineteen reasons were cited by diabetes patients as barriers for their unwillingness to accept the COVID-19 vaccine. These reasons emerged under 5 themes as summarized in Table 3 . Reasons for vaccine unwillingness stemmed predominately from health or safety concerns in relation to COVID-19 vaccination [6], [9], [19], [20], [22]. These concerns were centered on the perceived negative experience with the vaccine, such as side effects from the vaccine, adverse reactions, allergic reactions, glucose variation, and the possibility of vaccines leading to other diseases.Table 3 Barriers to COVID-19 vaccine acceptance.
Themes Reasons for vaccine hesitancy
Misinformation • Belief that the SARS-CoV-2 vaccine is unsafe
• Belief that COVID-19 is not dangerous to diabetes patients’ health
• Belief that vaccination does not reduce the risk of infection
• Different conspiracy theories
Lack of Information • No current reports of follow-up after vaccination of diabetes patients
• Relatively fast production
• Not many trials done
• Uncertainties about the genetic component
Mistrust • Not trust pharmaceutical manufacturing companies
• Not trust the countries where the vaccines were produced
• Not trust the manufacturing process of the vaccine.
Health Concerns • Fear of side effects from the vaccine
• Fear of other adverse reactions after vaccination
• Glucose variation
• Possibility of leading to other diseases
• Afraid of allergic reaction
External Influence • COVID-19 vaccination was not recommended by their physicians
• Families suggested not to receive vaccination
• Advice from religious leaders
4 Discussion
4.1 Summary of the results
The aim of this study was to determine the acceptance rate of the COVID-19 vaccine among individuals with diabetes. To achieve this objective, a comprehensive literature search was conducted to identify relevant studies, and a meta-analysis was performed to synthesize the prevalence data. The findings of our meta-analysis revealed that individuals with diabetes have a combined COVID-19 vaccine acceptance rate of 76.1% (95% CI: 66.7– 83.5). This high acceptance rate could be attributed to diabetes patients' heightened awareness of their susceptibility to the COVID-19 disease [35] which may have led to a better understanding of the benefits of vaccination in preventing severe illness and hospitalization. Our findings align with empirical studies that demonstrate the conscientiousness of diabetes patients regarding preventive measures against COVID-19, such as proper hand hygiene, wearing of face masks, and maintaining social distancing [36], [37]. Moreover, the earlier prioritization of individuals with diabetes in the COVID-19 vaccine rollout [8] may have played a role in increasing awareness and education about the importance of vaccination in this population, which in turn may have boosted their willingness to accept the vaccine. Compared to the findings of meta-analysis of the general population, the acceptance rate of the COVID-19 vaccine among individuals with diabetes is higher, as demonstrated in our results [11], [38], [39], [40]. It is also worth noting that the results of this study demonstrate a higher acceptance rate of COVID-19 vaccine among individuals with diabetes as compared to other high-risk populations, such as cancer patients (59%) [41] and patients with chronic disease in general (65%) [42].
Our subgroup analysis, stratified according to continent, has revealed notable variations in the acceptance rates of COVID-19 vaccines. Specifically, we observed the highest acceptance rates in Europe, followed by America, Africa, and the lowest in Asia. This finding is consistent with previous meta-analyses that examined vaccine acceptance rates among individuals with chronic diseases, which similarly reported higher acceptance rates in Europe and America compared to Asia [42]. The high vaccine acceptance rate in Europe and America is attributed to the implementation of policies such as the early prioritization of persons with chronic disease which is likely to have increased awareness about the importance of vaccination [42]. It is worth nothing that other factors may also account for the variation of the vaccine willingness across continents. Additionally, our analysis of gender differences in vaccine acceptance rates revealed that male diabetes patients had a higher acceptance rate compared to their female counterparts. This finding aligns with the outcomes of previous meta-analyses conducted on the general population, which indicated that males are more inclined to accept COVID-19 vaccines [11], [38], [40]. Research has shown that males are more susceptible to contracting COVID-19, more prone to experiencing severe symptoms, and have a poorer prognosis [43], [44]. This heightened risk may serve as a motivating factor for males to actively seek vaccination. Moreover, it has been observed that males possess a greater level of confidence in the safety of the COVID-19 vaccine [22], which could also contribute to the observed higher acceptance rates. It is important to acknowledge that, apart from gender, other sociodemographic factors may also influence COVID-19 vaccine acceptance rates. However, due to limited information on these variables from the included studies, their analysis was not conducted. Future research should therefore explore the influence of these sociodemographic factors on COVID-19 vaccine acceptance rates to gain a more thorough understanding of the subject matter.
Barriers associated with COVID-19 vaccine acceptance emerged under 5 themes, namely, misinformation, lack of information, mistrust, health concerns, and external influences. Several studies on vaccine hesitancy have consistently identified these themes [45], [46]. The dissemination of false information via digital media platforms presents a major challenge, particularly in the realm of misinformation. Addressing this barrier requires a robust fact-checking and counter-messaging strategy to provide accurate information to the public [47]. It is also crucial for the scientific and healthcare communities to work in unison to dispel myths and provide reliable information, thereby building trust and encouraging vaccine uptake, particularly among high-risk populations such as individuals with diabetes. Addressing health-related concerns about COVID-19 vaccines is of utmost importance and requires healthcare providers to furnish credible and evidence-based information on vaccine safety and efficacy, in addition to closely monitoring and reporting any adverse events. Given that adverse events associated with COVID-19 vaccines are mostly mild and resolvable [48], healthcare providers must help patients to understand this concept and stay informed.
4.2 Strengths and limitations
This review provides a comprehensive analysis of the acceptance of COVID-19 vaccination among individuals diagnosed with diabetes, offering valuable insights into the current status of vaccine uptake within this population. The primary strength of this review lies in its novelty, as it establishes a benchmark for future investigations in this field. Additionally, the literature search employed specialized databases dedicated to COVID-19 research, ensuring the inclusion of all relevant studies on this topic. Despite these strengths, the review does have some limitations that must be acknowledged. Firstly, the studies included in the analysis exhibit a significant level of heterogeneity, which may challenge the validity of definitive conclusions regarding vaccine acceptance among individuals with diabetes. Another limitation lies in the fact that we did not assess the impact of different vaccine types on acceptance rates among diabetes patients, which could have influenced the outcomes. Furthermore, the exclusion of non-English language articles may introduce a potential bias by omitting perspectives from diverse cultural and linguistic groups, thereby limiting the generalizability of the findings. In light of these limitations, it is crucial to interpret the results of this review with caution.
5 Conclusion
In conclusion, the results of this study suggest that a majority of diabetes patients have accepted the COVID-19 vaccine. However, given the higher risk of severe COVID-19 among individuals with diabetes, it is important for healthcare providers to continue promoting the benefits of vaccination and addressing any concerns to increase vaccine uptake. The barriers that have been identified can be leveraged to formulate health policies and public health interventions that are specifically tailored to address the needs of persons with diabetes.
Funding
No funding was received to undertake this project.
Authors’ Contributions
The study was a collaborative effort between both authors. E.E conceptualized the study and designed the methods. Both authors (S.A and E.E) performed the study selection, data extraction, quality assessment, analysis, and wrote the manuscript. Both authors also critically reviewed the manuscript. The authors read and approved the final manuscript.
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.
Appendix A Supplementary data
The following are the Supplementary data to this article:Supplementary Data 1
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.diabres.2023.110731.
==== Refs
References
1 Sen S. Chakraborty R. Kalita P. Pathak M.P. Diabetes mellitus and COVID-19: Understanding the association in light of current evidence World J Clin Cases 9 28 2021 8327 8339 34754842
2 Wolff D. Nee S. Hickey N.S. Marschollek M. Risk factors for Covid-19 severity and fatality: a structured literature review Infection 49 1 2021 15 28 32860214
3 Kastora S. Patel M. Carter B. Delibegovic M. Myint P.K. Impact of diabetes on COVID-19 mortality and hospital outcomes from a global perspective: An umbrella systematic review and meta-analysis Endocrinol Diabetes Metab 5 3 2022 e00338
4 Moghadas S.M. Vilches T.N. Zhang K. The Impact of Vaccination on Coronavirus Disease 2019 (COVID-19) Outbreaks in the United States Clin Infect Dis 73 12 2021 2257 2264 33515252
5 Kaine G. Wright V. Greenhalgh S. Predicting willingness to be vaccinated for Covid-19: Evidence from New Zealand PLoS One 17 4 2022 e0266485
6 Aldossari K.K. Alharbi M.B. Alkahtani S.M. Alrowaily T.Z. Alshaikhi A.M. Twair A.A. COVID-19 vaccine hesitancy among patients with diabetes in Saudi Arabia Diabetes Metab Syndr 15 5 2021 102271
7 Guaraldi F. Montalti M. Di Valerio Z. Rate and Predictors of Hesitancy toward SARS-CoV-2 Vaccine among Type 2 Diabetic Patients: Results from an Italian Survey Vaccines (Basel) 9 5 2021 460 34064486
8 Nguyen K.H. Srivastav A. Razzaghi H. COVID-19 Vaccination Intent, Perceptions, and Reasons for Not Vaccinating Among Groups Prioritized for Early Vaccination - United States, September and December 2020 MMWR Morb Mortal Wkly Rep 70 6 2021 217 222 33571174
9 Wang Y. Duan L. Li M. COVID-19 Vaccine Hesitancy and Associated Factors among Diabetes Patients: A Cross-Sectional Survey in Changzhi, Shanxi, China Vaccines (Basel) 10 1 2022 129 35062790
10 Patwary M.M. Alam M.A. Bardhan M. COVID-19 Vaccine Acceptance among Low- and Lower-Middle-Income Countries: A Rapid Systematic Review and Meta-Analysis Vaccines (Basel) 10 3 2022 427 35335059
11 Dubé E. Vivion M. MacDonald N.E. Vaccine hesitancy, vaccine refusal and the anti-vaccine movement: influence, impact and implications Expert Rev Vaccines 14 1 2015 99 117 25373435
12 Munn Z. Stern C. Aromataris E. Lockwood C. Jordan Z. What kind of systematic review should I conduct? A proposed typology and guidance for systematic reviewers in the medical and health sciences BMC Med Res Methodol 18 1 2018 5 29316881
13 Ekpor E. Akyirem S. Adade D.P. Prevalence and associated factors of overweight and obesity among persons with type 2 diabetes in Africa: a systematic review and meta-analysis Ann Med 55 1 2023 696 713 36821504
14 Ouzzani M. Hammady H. Fedorowicz Z. Elmagarmid A. Rayyan-a web and mobile app for systematic reviews Syst Rev 5 1 2016 210 27919275
15 Munn Z. Moola S. Lisy K. Riitano D. Tufanaru C. Methodological guidance for systematic reviews of observational epidemiological studies reporting prevalence and cumulative incidence data Int J Evid Based Healthc 13 3 2015 147 153 26317388
16 Warton D.I. Hui F.K. The arcsine is asinine: the analysis of proportions in ecology Ecology 92 1 2011 3 10 21560670
17 Higgins J.P. Thompson S.G. Quantifying heterogeneity in a meta-analysis Stat Med 21 11 2002 1539 1558 12111919
18 Peters J.L. Sutton A.J. Jones D.R. Abrams K.R. Rushton L. Comparison of two methods to detect publication bias in meta-analysis JAMA 295 6 2006 676 680 16467236
19 Osuagwu U.L. Langsi R. Ovenseri-Ogbomo G. Analysis of Perception, Reasons, and Motivations for COVID-19 Vaccination in People with Diabetes across Sub-Saharan Africa: A Mixed-Method Approach Int J Environ Res Public Health 19 13 2022 7875 35805551
20 Lu D. Gao Y. Qi X. Li A. Zhang J. The COVID-19 vaccination hesitancy among Chinese individuals with diabetes and the impact on glycemic control of vaccination: a questionnaire study BMC Endocr Disord 22 1 2022 329 36550448
21 Asadi-Pooya A.A. Barzegar Z. Sadeghian S. Nezafat A. Shahisavandi M. Nabavizadeh S.A. COVID-19 Vaccine Hesitancy Among Patients With Epilepsy or Other Chronic Conditions Disaster Med Public Health Prep 16 5 2022 1848 1850 34629142
22 Kolobov T. Djuraev S. Promislow S. Tamir O. Determinants of COVID-19 vaccine acceptance among adults with diabetes and in the general population in Israel: A cross-sectional study Diabetes Res Clin Pract 189 2022 109959
23 Syed Alwi S.A.R. Rafidah E. Zurraini A. Juslina O. Brohi I.B. Lukas S. A survey on COVID-19 vaccine acceptance and concern among Malaysians BMC Public Health 21 1 2021 1129 34118897
24 Abedin M. Islam M.A. Rahman F.N. Willingness to vaccinate against COVID-19 among Bangladeshi adults: Understanding the strategies to optimize vaccination coverage PLoS One 16 4 2021 e0250495
25 Okubo R. Yoshioka T. Ohfuji S. Matsuo T. Tabuchi T. COVID-19 Vaccine Hesitancy and Its Associated Factors in Japan Vaccines 9 6 2021 662 34204465
26 Tsai R. Hervey J. Hoffman K. COVID-19 Vaccine Hesitancy and Acceptance Among Individuals With Cancer, Autoimmune Diseases, or Other Serious Comorbid Conditions: Cross-sectional, Internet-Based Survey JMIR Public Health Surveill 8 1 2022 e29872
27 Valerio V. Rampakakis E. Zanos T.P. High Frequency of COVID-19 Vaccine Hesitancy among Canadians Immunized for Influenza: A Cross-Sectional Survey Vaccines (Basel) 10 9 2022 1514 36146592
28 Kanyangarara M. McAbee L. Daguise V.G. Nolan M.S. Factors Associated with COVID-19 Vaccine Intentions among South Carolina Residents Vaccines (Basel) 10 6 2022 942 35746550
29 Czeisler M.É. Barrett C.E. Siegel K.R. Health Care Access and Use Among Adults with Diabetes During the COVID-19 Pandemic - United States, February-March 2021 MMWR Morb Mortal Wkly Rep 70 46 2021 1597 1602 34793416
30 Waite N.M. Pereira J.A. Houle S.K.D. Gilca V. Andrew M.K. COVID-19's Impact on Willingness to Be Vaccinated against Influenza and COVID-19 during the 2020/2021 Season: Results from an Online Survey of Canadian Adults 50 Years and Older Vaccines (Basel) 9 4 2021 346 33916364
31 Scoccimarro D. Panichi L. Ragghianti B. Silverii A. Mannucci E. Monami M. Sars-CoV2 vaccine hesitancy in Italy: A survey on subjects with diabetes Nutr Metab Cardiovasc Dis 31 11 2021 3243 3246 34629250
32 Mesele H. Shiferaw M. Tunta A. Seid A. Kassahun W. Willingness to Receive COVID-19 Vaccination Among Adult Diabetes Patients in Woldia Comprehensive Specialized Hospital, North Ethiopia; A Cross-Sectional Study Patient Prefer Adherence 16 2022 2451 2459 36081925
33 Day D. Grech L. Nguyen M. Serious Underlying Medical Conditions and COVID-19 Vaccine Hesitancy: A Large Cross-Sectional Analysis from Australia Vaccines (Basel) 10 6 2022 851 35746458
34 Sterne J.A. Sutton A.J. Ioannidis J.P. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials BMJ 343 2011 d4002
35 Leong D.P. Banerjee A. Yusuf S. COVID-19 Vaccination Prioritization on the Basis of Cardiovascular Risk Factors and Number Needed to Vaccinate to Prevent Death Can J Cardiol 37 7 2021 1112 1116 33933606
36 Alsaleh F.M. Elzain M. Alsairafi Z.K. Naser A.Y. Perceived Knowledge, Attitude, and Practices (KAP) and Fear toward COVID-19 among Patients with Diabetes Attending Primary Healthcare Centers in Kuwait Int J Environ Res Public Health 20 3 2023 2369 36767736
37 Pal R. Yadav U. Grover S. Saboo B. Verma A. Bhadada S.K. Knowledge, attitudes and practices towards COVID-19 among young adults with Type 1 Diabetes Mellitus amid the nationwide lockdown in India: A cross-sectional survey Diabetes Res Clin Pract 166 2020 108344
38 Norhayati M.N. Che Yusof R. Azman Y.M. Systematic Review and Meta-Analysis of COVID-19 Vaccination Acceptance Front Med (Lausanne) 8 2022 783982
39 Wang Q. Hu S. Du F. Mapping global acceptance and uptake of COVID-19 vaccination: A systematic review and meta-analysis Commun Med (Lond) 2 2022 113 36101704
40 Wang Q. Yang L. Jin H. Lin L. Vaccination against COVID-19: A systematic review and meta-analysis of acceptability and its predictors Prev Med 150 2021 106694
41 Prabani K.I.P. Weerasekara I. Damayanthi H.D.W.T. COVID-19 vaccine acceptance and hesitancy among patients with cancer: a systematic review and meta-analysis Public Health 212 2022 66 75 36244261
42 Zhao Y, Du J, Li Z, et al. It Is Time to Improve the Acceptance of COVID-19 Vaccines Among People with Chronic Diseases: A Systematic Review and Meta-analysis [published online ahead of print, 2023 Jan 19]. J Med Virol. 2023;10.1002/jmv.28509.
43 Jin J.M. Bai P. He W. Gender Differences in Patients With COVID-19: Focus on Severity and Mortality Front Public Health 8 2020 152 32411652
44 Ortolan A. Lorenzin M. Felicetti M. Doria A. Ramonda R. Does gender influence clinical expression and disease outcomes in COVID-19? A systematic review and meta-analysis Int J Infect Dis 99 2020 496 504 32800858
45 Fieselmann J. Annac K. Erdsiek F. Yilmaz-Aslan Y. Brzoska P. What are the reasons for refusing a COVID-19 vaccine? A qualitative analysis of social media in Germany BMC Public Health 22 1 2022 846 35484619
46 Shearn C. Krockow E.M. Reasons for COVID-19 vaccine hesitancy in ethnic minority groups: A systematic review and thematic synthesis of initial attitudes in qualitative research SSM Qual Res Health 3 2023 100210
47 Germani F. Pattison A.B. Reinfelde M. WHO and digital agencies: how to effectively tackle COVID-19 misinformation online BMJ Glob Health 7 8 2022 e009483
48 Majumder M.A.A. Singh K. Johnson W.M.S. Evaluation of Short-Term Side Effects Following the First Dose of COVID-19 Vaccines Among Physicians and Dentists: A Cross-Sectional Study from India J Multidiscip Healthc 16 2023 161 174 36700174
|
PMC010xxxxxx/PMC10208261.txt |
==== Front
Epilepsy Behav
Epilepsy Behav
Epilepsy & Behavior
1525-5050
1525-5069
Elsevier Inc.
S1525-5050(23)00181-6
10.1016/j.yebeh.2023.109262
109262
Letter to the Editor
COVID-19 vaccination status and related process of care outcomes
Kleebayoon Amnuay ⁎
Samraong, Cambodia
Wiwanitkit Viroj
Chandigarh University, Punjab, India
Joesph Ayobabalola University, Ikeji-Arakeji, Nigeria
⁎ Corresponding author.
24 5 2023
7 2023
24 5 2023
144 109262109262
30 4 2023
8 5 2023
11 5 2023
© 2023 Elsevier Inc. All rights reserved.
2023
Elsevier Inc.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Keywords
Covid
Vaccine
Effect
==== Body
pmcDear Editor,
We would like to share ideas on the publication “COVID-19 vaccination status and related process of care outcomes among U.S. adults with Active Epilepsy-National Health Interview Survey, United States, 2021 [1].” According to Kobau et al., in 2021, the receipt of one COVID-19 vaccine among U.S. people with current epilepsy was comparable to that of those without a history of epilepsy [1]. According to Kobau et al., persons aged 18–44 years with current epilepsy were considerably less likely to have reported receiving two COVID-19 immunizations than their counterparts without a history of epilepsy [1]. Adults with active epilepsy had similar experiences and outcomes regarding COVID-19 testing and accessing health treatment throughout the COVID-19 pandemic, according to Kobau et al. [1]. According to Kobau et al., this study provides baseline estimates of select COVID-19 outcomes among U.S. people with active epilepsy to guide interventions and future research [1].
Regarding the vaccine hesitancy among the cases with underlying epilepsy, the most important concern is the reliability of a vaccine and in this regard, two health agencies, the WHO and the FDA, are the most trusted organizations to approve a vaccine against COVID-19 [2].
Efforts to broaden vaccination acceptability deserve praise. Concerns are voiced each time a new COVID-19 vaccination is created and made accessible to the general population. Residents may experience anxiety if they learn that a harmful effect is present. How well the COVID-19 vaccination is received can be significantly influenced by community trust in the public health system [3]. Patients with underlying illnesses, such as epilepsy, are given priority when it comes to vaccine allocation in our environment in Indochina. The COVID-19 vaccination is given to registered epilepsy cases during a follow-up appointment for their neurological condition [4]. Knowledge about immunizations may occasionally cause local vaccination resistance. The public's waning confidence in regional public health management systems must be taken into account when analyzing resistance in any setting [3]. The underlying local COVID-19 vaccination regulations have a substantial impact on the rate of immunization uptake. The mandatory vaccination policy is in place in many developing nations in Asia, and it may help to boost adoption but it might also be related to stress and hesitancy in some specific populations [4]. The rollout of the vaccine has also been flexible in our particular setting, Cambodia, shifting from a strategy of prioritizing risk groups and crucial workers to one of expanding the campaign from population centers to rural areas and gradually expanding the target age group, which includes patients with underlying neurological diseases [5]. Cambodia has attained 95% primary series coverage as a consequence of the government's strong commitment and adaptable reaction [5].
Additionally, as environmental factors and the timing of the COVID-19 epidemic change, so does the pattern of resistance [6]. Therefore, in the context of the study, the history of the COVID-19 outbreak must be covered. Even if the pandemic's circumstances altered, it does not seem likely that the vaccination's acceptance rate would change. The identified pattern of reluctance may be useful for future research.
Funding
The authors have no funds and ask for waiving any charge from the journal.
Ethics approval and consent to participate
Not applicable.
Consent for publication
Agree.
Authors' contributions
AK 50% ideas, writing, analyzing, approval for submission.
VW 50% ideas, supervision, approval for submission.
CRediT authorship contribution statement
Viroj Wiwanitkit: Supervision.
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.
Acknowledgements
Not applicable.
==== Refs
References
1 Kobau R. Luncheon C. Pastula D.M. Greenlund K.J. COVID-19 vaccination status and related process of care outcomes among U.S. adults with active epilepsy-National Health Interview Survey, United States, 2021 Epilepsy Behav 143 2023 109223 10.1016/j.yebeh.2023.109223 Online ahead of print 37119577
2 Ungmunpuntipantip R. Wiwanitkit V. COVID-19 vaccination hesitancy Recenti Prog Med 112 9 2021 596 34392326
3 Asadi-Pooya A.A. Sahraian A. Badv R.S. Sahraian M.A. Physicians' opinions on the necessity of COVID-19 vaccination in patients with epilepsy Epileptic Disord 23 3 2021 485 489 34057409
4 Mahasing C. Yasopa O. Sansilapin C. Rattanathumsakul T. Thammawijaya P. Suphanchaimat R. Investigation of a Cluster of Immunization Stress-Related Reactions after Coronavirus Disease 2019 (COVID-19) Vaccination, Thailand, 2021 Vaccines (Basel) 10 3 2022 441 35335073
5 Nozaki I. Hachiya M. Ikeda C. COVID-19 vaccination program in Cambodia: Achievements and remaining challenges Glob Health Med 5 2 2023 92 98 37128223
6 Xiao J. Cheung J.K. Wu P. Ni M.Y. Cowling B.J. Liao Q. Temporal changes in factors associated with COVID-19 vaccine hesitancy and uptake among adults in Hong Kong: serial cross-sectional surveys Lancet Reg Health West Pac 23 2022 100441 35359914
|
PMC010xxxxxx/PMC10208263.txt |
==== Front
Vaccine
Vaccine
Vaccine
0264-410X
1873-2518
Elsevier Science
S0264-410X(23)00613-8
10.1016/j.vaccine.2023.05.054
Short Communication
COVID-19 vaccine safety inquiries to the centers for disease control and prevention immunization safety office
Miller Elaine R. a⁎
Moro Pedro L. a
Shimabukuro Tom T. a
Carlock Grace a
Davis Shaeyla N. b
Freeborn Emma M. a
Roberts Amy L. a
Gee Julianne a
Taylor Allan W. c
Gallego Ruth a
Suragh Tiffany a
Su John R. a
a Immunization Safety Office, Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention, Atlanta, GA, United States
b Deputy Director for Infectious Disease, Centers for Disease Control and Prevention, Atlanta, GA, United States
c Global Health Center, Centers for Disease Control and Prevention, Atlanta, GA, United States
⁎ Corresponding author.
24 5 2023
19 6 2023
24 5 2023
41 27 39603963
29 3 2023
10 5 2023
18 5 2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Following the authorization and recommendations for use of the U.S. COVID-19 vaccines, the Centers for Disease Control and Prevention (CDC)’s Immunization Safety Office (ISO) responded to inquiries and questions from public health officials, healthcare providers, and the general public on COVID-19 vaccine safety.
Methods
We describe COVID-19 vaccine safety inquiries, by topic, received and addressed by ISO from December 1, 2020–August 31, 2022.
Results
Of the 1978 COVID-19 vaccine-related inquiries received, 1655 specifically involved vaccine safety topics. The most frequently asked-about topics included deaths following vaccination, myocarditis, pregnancy, and reproductive health outcomes, understanding or interpreting data from the Vaccine Adverse Event Reporting System (VAERS), and thrombosis with thrombocytopenia syndrome.
Conclusions
Inquiries about vaccine safety generally reflect issues that receive media attention. ISO will continue to monitor vaccine safety inquiries and provide accurate and timely information to healthcare providers, public health officials, and the general public.
Keywords
COVID-19 vaccination
Vaccine safety
Inquiry response
Abbreviations
ISO Immunization Safety Office
VAERS Vaccine Adverse Event Reporting System
AE Adverse Event
TTS Thrombosis with thrombocytopenia syndrome
==== Body
pmc1 Introduction
The Centers for Disease Control and Prevention (CDC)’s Immunization Safety Office (ISO) responds to questions about vaccine safety topics from public health officials, healthcare providers, and the general public [1]. Following the authorization and recommendations for use of U.S. COVID-19 vaccines [2], ISO staff responded to questions and concerns about the safety of these vaccines and requests for data and guidance related to vaccine safety issues. Here, we summarize the most frequent inquiry topics ISO addressed related to U.S. COVID-19 vaccines to understand perceived safety concerns and help guide efforts to address those concerns.
2 Methods
ISO receives inquiries on vaccine safety topics from healthcare providers, public health officials, and the general public through multiple channels. Most inquiries are received through the CDC-INFO [3] (a telephone and email system that responds to general questions for CDC) and CDC’s National Immunization Program (NIP-INFO) (an email service that responds to general questions about immunization). When inquiries to CDC-INFO and NIP-INFO require vaccine safety expertise beyond what CDC-INFO or NIP-INFO provide (e.g., a request for a search of an ISO data system or an explanation of a technical issue), they are triaged to the ISO vaccine safety inquiry response program. Additionally, ISO staff receive inquiries directly, primarily by email. The ISO vaccine safety inquiry response program includes nurses, physicians, epidemiologists, and other health scientists with expertise in vaccine safety. Inquiries received by ISO and subsequent responses are entered into a Research Electronic Data Capture (REDCap) electronic database hosted at CDC and put into use for ISO inquiries beginning December 1, 2020 [4]. For some data fields in REDCap such as topic of inquiry and role of inquirer, we used dropdown menus for data entry.
Vaccine safety inquires received by ISO were analyzed descriptively using R 4.1.1 [5] and Microsoft Power BI data visualization software [6]. We summarized data from the most frequently asked topics about COVID-19 vaccine safety that ISO staff responded to from December 1, 2020–August 31, 2022.
3 Results
From December 1, 2020–August 31, 2022, ISO received 2096 inquiries. Of those, 118 were about vaccines other than COVID-19; 323 were related to COVID-19 vaccine but were not about safety, and the remaining 1655 were about COVID-19 vaccine safety (Table 1 ). The volume of inquiries was greater during the analytic period (i.e., during the COVID-19 vaccination program) compared to the pre-COVID-19 period – approximately 105 inquiries per month versus 30 per month, respectively. The most common COVID-19 vaccine safety topics were about deaths following vaccination (160, 10 %); myocarditis and related topics (153, 9 %); pregnancy and reproductive health outcomes (123, 7 %); understanding or interpreting data from the Vaccine Adverse Event Reporting System (VAERS), a safety monitoring system co-managed by CDC and the U.S. Food and Drug Administration (FDA) [7] (111, 7 %); and thrombosis with thrombocytopenia syndrome (TTS) and related questions about blood clotting (95, 6 %); the remaining inquiries covered a wide range of topics (Table 1). Another 323 (16 %) inquiries were not directly about vaccine safety, including how to report an adverse event (AE) to VAERS [7] or how to access publicly available VAERS data. Fig. 1 shows a timeline of the most common inquiries and important events in the U.S. COVID-19 vaccination program.Table 1 Topics of COVID-19 vaccine safety inquiries received and addressed by the CDC Immunization Safety Office (N = 1655 total inquiries), December 2020–August 2022.
Topics of inquiries N (%)
Deaths 160 (10)
Myocarditis and related topics 153 (9)
Pregnancy and reproductive health outcomes 123 (7)
Vaccine Adverse Event Reporting System (VAERS) data interpretation 111 (7)
Thrombosis with thrombocytopenia (TTS) and related clot issues 95 (6)
Vaccine Administration Error 35 (2)
Anaphylaxis and other allergic reactions 32 (2)
Guillain-Barré syndrome (GBS) 30 (2)
Tinnitus 28 (2)
Thrombocytopenia (includes Idiopathic) 18 (1)
Rash 16 (1)
Hearing loss 14 (1)
Bell’s Palsy 13 (1)
Neurological disorders (unspecified) 12 (1)
Irregular Menses 12 (1)
Shingles or zoster 11 (1)
Other topics 792 (48)
Fig. 1 Main inquiry topics received by the CDC Immunization Safety Office, December 2020–August 2022 (N = 642).
Deaths following COVID-19 vaccination was the most frequent inquiry topic (Table 1, Fig. 1), and the topic most frequently asked about by health department staff. Questions on death were asked regularly throughout the period of analysis (Fig. 1). Requestors asked if any deaths were caused by the vaccine, the causes of death among reports submitted to VAERS, and how CDC evaluates deaths after vaccination. Also, family members of deceased persons informed CDC of their family member’s death after COVID-19 vaccination (regardless of the cause of death) and asked if vaccination could have been a cause.
Myocarditis, a known causally associated adverse event after mRNA COVID-19 vaccination [8], and related topics (pericarditis and myopericarditis) was the second most frequent topic (Table 1, Fig. 1), and the topic most frequently asked about by healthcare providers. Examples included requests for case counts of myocarditis following vaccination and the risk of myocarditis after vaccination for different age groups. Inquirers also asked how CDC conducted its benefit-risk assessment for mRNA COVID-19 vaccination, given the risk of vaccine-associated myocarditis. Additionally, ISO received requests from the general public for medical advice on treating symptoms related to myocarditis.
The safety of COVID-19 vaccines during pregnancy and reproductive health outcomes following vaccination was the third most frequent topic of inquiries (Table 1, Fig. 1) and the most frequent topic asked about by the public. People asked about the safety of COVID-19 vaccines for pregnant women and the potential adverse effects of COVID-19 vaccine on developing fetuses, and the risk of miscarriage.
Interpretation of VAERS data was the fourth most frequent topic (Table 1, Fig. 1). Some inquiries conveyed a misunderstanding or misinterpretation of VAERS data. A common misconception was that reports to VAERS of an AE after COVID-19 vaccination proved the vaccine caused the AE. Inquiries included requests for information on what actions CDC and FDA take when VAERS identifies a potential safety problem and how CDC and FDA use VAERS data to evaluate if a safety problem with a vaccine genuinely exists. Inquiries also asked how well VAERS captures the total number of AEs occurring after COVID-19 vaccination.
Thrombosis with thrombocytopenia syndrome (TTS), a rare adverse event involving blood clots in unusual locations with low platelets, is causally associated with the Janssen COVID-19 vaccine but not with COVID-19 mRNA vaccines [9]. TTS was the fifth most frequent topic of inquiries (Table 1, Fig. 1), including requests for the number of reports, ages of patients, and how to report potential cases to the CDC.
4 Discussion
From December 1, 2020–August 31, 2022, ISO received 1655 inquiries about COVID-19 vaccine safety (Table 1). The most common inquiries involved deaths following vaccination, myocarditis, pregnancy, and reproductive health outcomes, understanding or interpreting data from VAERS, and TTS and related blood clotting topics. Inquiry topics tended to reflect safety issues recently communicated by CDC or FDA or following public meetings by federal advisory committees, such as CDC’s Advisory Committee on Immunization Practices (ACIP) [8], [9], [10]. The three and a half fold increase in the volume of inquiries during the analytic period (i.e., during the COVID-19 vaccination program) compared to the pre-COVID-19 period likely reflects both the volume of COVID-19 vaccines administered (as of this writing, over 675 million doses in the United States [11]) and heightened public awareness of the pandemic vaccination program. During the analytic period, questions to ISO on vaccines other than COVID-19 were relatively rare (less than 6 % of total inquiries).
People receiving COVID-19 vaccines are at no greater risk of death from non-COVID causes than unvaccinated people and are less likely to die from COVID-19 disease [10], [12]. Nonetheless, several factors might have contributed to public interest in deaths after COVID-19 vaccination, especially early in the U.S. vaccination program. Early reports of death following COVID-19 vaccination involved residents of long-term care facilities (LTCFs) [13], leading to public speculation about a potential causal association between vaccination and deaths in this population. At the start of the national COVID-19 vaccination program, vaccination was prioritized based on risk for severe disease and public health needs (e.g., vaccinating direct care providers). Accordingly, residents of LTCFs were among the first to be offered the vaccine. The next groups prioritized for vaccination included people aged ≥ 75 years, followed by people aged 65–74 years, and people aged 16–64 years with high-risk medical conditions [14]. Age-related mortality and increased mortality due to medical comorbidities could be anticipated among these populations. Indeed, a review of available autopsy reports, death certificates, and medical records was consistent with expected causes of death for people in these age and risk groups [12], [13]. Additionally, FDA Emergency Use Authorizations for COVID-19 vaccines and CDC COVID-19 Vaccination Provider Enrollment Agreements required healthcare providers who administer vaccines to report deaths following COVID-19 vaccination to VAERS regardless of the cause or circumstances surrounding death [15]. Importantly, this requirement does not apply to other vaccines [16]. These mandatory reporting requirements, combined with heightened awareness of the national COVID-19 vaccination program, likely resulted in substantial stimulated reporting of deaths after COVID-19 vaccination and disproportionality in deaths reported after COVID-19 vaccines compared to non-COVID-19 vaccines, both currently and historically. Lastly, it is not uncommon for people to incorrectly interpret a close temporal association between a vaccination and an AE as, by itself, an indication of a causal association. When death was the AE occurring soon after COVID-19 vaccination, this tendency might have been heightened, contributing to the increased concern.
The highest volume of inquiries on myocarditis coincided with anecdotal reports and reports to VAERS of myocarditis after mRNA COVID-19 vaccination beginning in the spring of 2021 (Fig. 1), followed by vaccine safety surveillance findings from VAERS and from CDC’s Vaccine Safety Datalink (VSD) [which uses electronic medical record data from integrated healthcare organizations] [17]. These data were presented at the June 2021 ACIP meeting [18] showing that observed reporting rates of myocarditis following mRNA COVID-19 vaccination exceeded anticipated background incidence rates. Reporting rates were highest among males in their late teens and early 20 s following dose 2 of the primary vaccination series. These data suggested a potentially increased myocarditis risk associated with mRNA COVID-19 vaccination. During the June 2021 and August 2021 ACIP meetings [18], CDC presented data demonstrating that the benefits of COVID-19 vaccinations outweighed the risks of myocarditis among people of all ages, including people ages 16–29 years [8]. Subsequent research found that the risk of myocarditis was higher following SARS-CoV-2 infection compared to the risk of myocarditis after mRNA COVID-19 vaccination, further reinforcing the favorable benefit-risk balance [19].
When COVID-19 vaccines were first authorized for use, CDC and ACIP, along with the American College of Obstetricians and Gynecologists and the American Academy of Pediatrics, issued guidance that COVID-19 vaccines should not be withheld from pregnant women. On September 29, 2021, a CDC Health Advisory on Pregnancy strongly recommended COVID-19 vaccination for pregnant women based on data showing increased risk of severe illness and death in pregnant women with SARS-CoV-2 infection, and increased risk of premature birth, stillbirth, and SARS-CoV-2 infection in their newborn infants [20]. Questions on pregnancy sent to ISO declined substantially shortly thereafter (Fig. 1).
VAERS, a spontaneous (passive surveillance) reporting system, is the nation’s early warning program for vaccine safety and has traditionally provided the quickest safety data for new vaccines [7]. VAERS data are publicly available to anyone with computer and internet access. In responding to questions about VAERS, ISO explained its strengths and limitations in every response on this topic. ISO also explained that VAERS is very important in detecting unusual or unexpected reporting patterns of AEs that might indicate potential safety concerns. However, CDC uses other, more robust data systems to evaluate risk, assess causality, and further characterize potential associations between vaccination and AEs. Knowing which vaccine safety topics are important to the public and other partners can help guide CDC efforts to better communicate relevant information. For example, CDC learned from inquiries that VAERS data were being misinterpreted as definitive evidence that vaccines caused the reported AEs; ISO subsequently provided information on the CDC’s VAERS public data website emphasizing that data from VAERS generally cannot determine causality.
Shortly after the authorization and recommendation for Janssen COVID-19 vaccine in February 2021, ISO received reports of a rare but serious and life-threatening AE, later known as TTS. Patients with TTS following Janssen vaccination typically had symptom onset within two weeks of vaccination and were most commonly women of reproductive age. Questions about TTS peaked during April 2021 (Fig. 1), likely reflecting widespread media coverage of the condition after the CDC and FDA temporarily paused use of Janssen’s COVID-19 vaccine on April 13, 2021, while the issue was further assessed [9], [21]. Given concerns about TTS and a rare neurological condition (Guillain-Barré syndrome) after vaccination with Janssen’s COVID-19 vaccine, ACIP preferentially recommended mRNA COVID-19 vaccines over Janssen’s COVID-19 vaccine in December 2021 [21].
5 Limitations
Our analysis of inquiries received by ISO provides insight into concerns from healthcare providers, public health officials, and the general public. However, routine vaccine safety questions are often answered by other parts of CDC [1]; ISO typically receives the most challenging and technically complex inquiries, which might not be truly representative of the full range of safety concerns about which people contact CDC. Also, ISO received many broad inquiries (e.g., “How safe is the COVID vaccine?”) that did not fit into a defined category.
6 Conclusions
Inquiries about vaccine safety generally reflect and track safety issues that public health and regulatory authorities are actively investigating and communicating. Presentation of vaccine safety information at federal advisory committee meetings is covered by the media and appears to stimulate topical inquiries to CDC. ISO will continue to monitor vaccine safety inquiries and provide accurate and timely information to healthcare providers, public health officials, and the general public.
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.
Data availability
Data will be made available on request.
==== Refs
References
1 Miller E. Batten B. Hampton L. Campbell S.R. Gao J. Iskander J. Tracking vaccine-safety inquiries to detect signals and monitor public concerns Pediatrics 127 Suppl 1 2011 S87 S91 21502241
2 FDA. COVID-19 Vaccine. The FDA has regulatory processes in place to facilitate the development of COVID-19 vaccines that meet the FDA's rigorous scientific standards. U.S. Food & Drug Administration; 2022.
3 CDC. CDC-INFO. Atlanta, GA: Centers for Disease Control and Prevention; 2022.
4 Harris P.A. Taylor R. Minor B.L. Elliott V. Fernandez M. O'Neal L. The REDCap consortium: Building an international community of software platform partners J Biomed Inform 95 2019 103208
5 Foundation TR. The R Project for Statistical Computing; 2022.
6 Microsoft. Microsoft Power BI; 2022.
7 Shimabukuro T.T. Nguyen M. Martin D. DeStefano F. Safety monitoring in the Vaccine Adverse Event Reporting System (VAERS) Vaccine 33 2015 4398 4405 26209838
8 Gargano J.W. Wallace M. Hadler S.C. Langley G. Su J.R. Oster M.E. Use of mRNA COVID-19 vaccine after reports of myocarditis among vaccine recipients: update from the advisory committee on immunization practices - United States, June 2021 MMWR Morb Mortal Wkly Rep 70 2021 977 982 34237049
9 See I. Lale A. Marquez P. Streiff M.B. Wheeler A.P. Tepper N.K. Case series of thrombosis with thrombocytopenia syndrome after COVID-19 vaccination-United States, December 2020 to August 2021 Ann Intern Med 175 2022 513 522 35038274
10 Xu S. Huang R. Sy L.S. Hong V. Glenn S.C. Ryan D.S. A safety study evaluating non-COVID-19 mortality risk following COVID-19 vaccination Vaccine 41 2023 844 854 36564276
11 Centers for Disease Control and Prevention. COVID Data Tracker. Centers for Disease Control and Prevention; 2022.
12 Day B, Menschik D, Thompson D, Jankosky C, Su J, Moro P, et al. Reporting rates for VAERS death reports following COVID-19 vaccination, December 14, 2020-November 17, 2021. Pharmacoepidemiol Drug Saf. 2023 Feb 22. Epub ahead of print. PMID: 36813704.
13 Rosenblum H.G. Gee J. Liu R. Marquez P.L. Zhang B. Strid P. Safety of mRNA vaccines administered during the initial 6 months of the US COVID-19 vaccination programme: an observational study of reports to the Vaccine Adverse Event Reporting System and v-safe Lancet Infect Dis 22 2022 802 812 35271805
14 Dooling K. Marin M. Wallace M. McClung N. Chamberland M. Lee G.M. The advisory committee on immunization practices' updated interim recommendation for allocation of COVID-19 vaccine – United States, December 2020 MMWR Morb Mortal Wkly Rep 69 2021 1657 1660 33382671
15 U.S. Department of Health & Human Services, Centers for Disease Control and Prevention, National Center for Immunization and Respiratory Diseases. CDC COVID-19 Vaccination Program Provider Requirements and Support; 2022.
16 Health Resources and Services Administration. VAERS Table of Reportable Events Following Vaccination; 2017.
17 McNeil M.M. Gee J. Weintraub E.S. Belongia E.A. Lee G.M. Glanz J.M. The Vaccine Safety Datalink: successes and challenges monitoring vaccine safety Vaccine 32 2014 5390 5398 25108215
18 Centers for Disease Control and Prevention. ACIP Meeting Information.
19 Block J.P. Boehmer T.K. Forrest C.B. Carton T.W. Lee G.M. Ajani U.A. Cardiac complications after SARS-CoV-2 infection and mRNA COVID-19 vaccination - PCORnet, United States, January 2021-January 2022 MMWR Morb Mortal Wkly Rep 71 2022 517 523 35389977
20 U.S. Department of Health & Human Services, Centers for Disease Control and Prevention. CDC Statement on Pregnancy Health Advisory; 2021.
21 Oliver S.E. Wallace M. See I. Mbaeyi S. Godfrey M. Hadler S.C. Use of the Janssen (Johnson & Johnson) COVID-19 vaccine: Updated interim recommendations from the advisory committee on immunization practices - United States, December 2021 MMWR Morb Mortal Wkly Rep 71 2022 90 95 35051137
|
PMC010xxxxxx/PMC10208557.txt |
==== Front
Rheumatol Int
Rheumatol Int
Rheumatology International
0172-8172
1437-160X
Springer Berlin Heidelberg Berlin/Heidelberg
37226016
5346
10.1007/s00296-023-05346-x
Review
Systemic lupus erythematosus: latest insight into etiopathogenesis
http://orcid.org/0000-0001-8953-2617
Akhil Akhil 1
http://orcid.org/0000-0002-2827-5685
Bansal Rohit 1
http://orcid.org/0000-0001-8134-9463
Anupam Kumari 2
http://orcid.org/0000-0003-4668-5255
Tandon Ankit 3
http://orcid.org/0000-0001-7468-4423
Bhatnagar Archana bhatnagar.archana@gmail.com
1
1 grid.261674.0 0000 0001 2174 5640 Department of Biochemistry, BMS-Block II, South Campus, Panjab University, Chandigarh, 160014 India
2 grid.262962.b 0000 0004 1936 9342 Department of Pathology, Saint Louis University, St. Louis, MO 63103 USA
3 grid.415131.3 0000 0004 1767 2903 Department of Endocrinology, PGIMER, Chandigarh, 160012 India
24 5 2023
2023
43 8 13811393
19 1 2023
15 5 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Systemic lupus erythematosus (SLE) is a complex autoimmune disorder of unknown etiology. Multifactorial interaction among various susceptible factors such as environmental, hormonal, and genetic factors makes it more heterogeneous and complex. Genetic and epigenetic modifications have been realized to regulate the immunobiology of lupus through environmental modifications such as diet and nutrition. Although these interactions may vary from population to population, the understanding of these risk factors can enhance the perception of the mechanistic basis of lupus etiology. To recognize the recent advances in lupus, an electronic search was conducted among search engines such as Google Scholar and PubMed, where we found about 30.4% publications of total studies related to genetics and epigenetics, 33.5% publications related to immunobiology and 34% related to environmental factors. These outcomes suggested that management of diet and lifestyle have a direct relationship with the severity of lupus that influence via modulating the complex interaction among genetics and immunobiology. The present review emphasizes the knowledge about the multifactorial interactions between various susceptible factors based on recent advances that will further update the understanding of mechanisms involved in disease pathoetiology. Knowledge of these mechanisms will further assist in the creation of novel diagnostic and therapeutic options.
Keywords
Lupus
Multifactorial interactions
Epigenetics
Immunobiology
Environmental factors
Etiology
http://dx.doi.org/10.13039/501100001501 University Grants Commission 191620089214 Akhil Akhil http://dx.doi.org/10.13039/501100001407 Department of Biotechnology, Ministry of Science and Technology, India BT/INF/22/SP41295/2020 Bansal Rohit issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIntroduction
Systemic lupus erythematosus (SLE), a rare multifactorial disease, has shown an upward trend in prevalence. A potentially severe autoimmune systemic disease characterized by the production of autoantibodies to components of the cell nucleus results in a diverse group of clinical manifestations. While the exact pathology of SLE is still unknown, patients depict an inflammatory milieu, deposition of immune complexes (ICs) in various organs, and vasculopathy. Clinical heterogeneity of SLE suggested that various susceptible factors including genetic, epigenetic, environmental, infections, and hormones modulate the disease pathology [1]. Multiple genes, the interaction of sex hormones along with defective immune regulatory mechanisms [2], including impaired clearance of apoptotic cell debris and immune complex deposition, are the important contributors to the development of SLE [3]. Various epigenetic modifications such as methylation, acetylation, and small RNA have also been found to modulate the disease pathology; however, these modifications can vary individually, and thus personalized approach is required to elaborate the role of these mechanisms among lupus patients [4]. Altered immunometabolism observed in various immune cells had a positive correlation with cellular differentiation and lupus severity. Aberrant cell signaling mechanisms have been reported to trigger abnormalities in cell differentiation and over-activation of immune cells thereby enhancing the autoantibody generation [5]. Environmental triggers such as chemical/physical factors, dietary factors, and infectious agents probably contribute to the initiation of SLE disease. The interplay of these factors (as shown in Fig. 1) is associated with disease heterogeneity and complexity; therefore, the understanding of these pathological factors can assist in understanding this dreadful disease. In this review, the focus is on the recent advances in information relevant to factors that can contribute to SLE susceptibility.Fig. 1 The complex interactions between the various susceptible factors implicated in lupus etiology. Implications of genetics known to directly regulate the disease manifestations or indirectly through epigenetic modifications induced by environmental factors. These indirect modifications can also regulate the immunobiology and mitochondrial health which combines to affect the immunometabolism, an important aspect in lupus pathology. Diet is equally important as it can regulate the disease severity and hence assist in disease management
Search strategy
An electronic search was conducted among various databases such as google scholar/PubMed and articles published from January 2019 to September 2022 were retrieved and shortlisted based on the search words; “Susceptible or pathological factor” and “SLE or lupus etiology”. A further search was conducted based on the keywords such as genetics, epigenetics, immunobiology, and environmental factors in lupus pathology. Studies that included general clinical manifestations in lupus were excluded. Out of 128 articles analyzed, 70.4% were research articles and clinical reports, while 29.5% were review articles and 47% were published in 2021–2022.
SLE immunobiology
The pathophysiology of SLE is complex involving the interplay between factors and cells of innate/adaptive immunity in the microenvironment and triggers an immune response (Fig. 2). A very important innate trigger is neutrophil, which is deregulated in SLE due to high reactive oxygen species (ROS) levels and get triggered by enhanced IL-18 receptors [6]. Neutrophil extracellular traps (NETs) are enhanced and are known to be associated with enhanced IgG2 production from B cells in lupus [7]. Neutrophil degranulation is known to increase the secretion of pro-inflammatory cytokines (IFN-ϒ) that can fan the flames of lupus etiology [8]. Innate immune receptors such as the Toll-like receptors (TLRs) have the unique capacity to recognize pathogen-associated molecular patterns (PAMPs) while inducing immune system activation by linking the innate and adaptive immunity responses. The TLRs 1, 2, 4, 5, and 6 for bacterial or fungal PAMPs are located on the cell surface and TLRs 3, 7, 8, and 9 for nucleic acid (single-stranded/double-stranded RNA or DNA) are located on the endosomal membrane [9]. TLRs 7 and 9 have been strongly implicated in SLE pathogenesis since they mediate IFNα production by plasmacytoid dendritic cells (pDCs) upon their activation by circulating immune complexes containing self-nucleic acid components [10]. TLR-7 can regulate the extra-follicular B cells response in the germinal center which may enhance autoantibody generation, while TLR-9 can limit the TLR-7 stimulatory activity, demonstrating its protective function in lupus etiology [11]. Platelets have functional TLR-7, which is responsible for their activation, leading to platelet–leukocyte aggregates possibly involved in priming the immune system upon viral attack [12]. Other TLRs such as TLR-2 and 4 were also found to be highly expressed in the saliva of SLE patients, however, their low expression was demonstrated with the presence of chronic periodontitis in lupus patients [13].Fig. 2 Immunobiology of systemic lupus erythematosus. Interaction between dysregulated innate and adaptive immune system leads to production of inflammatory cytokines and autoantibody. Over-activation of innate system further interacts with the adaptive immune system leads to over-activation of various immune cells. Self-antigen presentation by dendritic cells leads to the activation of T cells that further activates the autoreactive B cells and secretes the autoantibodies. FcR Immunoglobulin Fc receptors, TLR Toll-like receptor, BAFF B cell activating factors, IL Interleukin
The dendritic cells play an important role in linking the innate and adaptive immune systems and serve as both a “break” and “engine” for lupus pathology [14]. It has been shown that monocytes from healthy individuals may be differentiated into myeloid dendritic cells (mDCs) in the presence of serum from a lupus patient [15]. These mDCs can also have the ability to phagocytize nuclear material and present antigens to naive CD4+ T cells, which can later activate the cytotoxic activity of CD8+ T cells and thus support the proliferation and differentiation of B cells [16]. Autoreactive IgE, such as anti-dsDNA IgE, was found to be elevated in SLE patients and can also elicit an increase in IFNα production by binding the Fc epsilon RI (FcεRI) of plasmacytoid dendritic cells (pDCs) [17]. The binding of IgE to pDCs also enhances the follicular T cell expansion and reduces the Treg population, which increased the inflammatory condition in lupus patients [18]. IFN-α can also activate the IL-1 receptor-associated kinase that further induces apoptosis in Treg cells from SLE patients [19]. Tregs have potent anti-inflammatory activity, therefore their apoptosis will further enhance the immune response in lupus patients. In a recent study, P-Selectins was also demonstrated as a suppressor of Treg cell’s function in SLE pathogenesis [20]. The suppressive function of PD-1+ Treg cells was found to be impaired, which leads to the over-activation of T and B cells in lupus patients [21]. Over-activation of T cells has led to its exhaustion, which can be further correlated with tolerance mechanisms such as prolonged remission in lupus patients [22]. The type-I interferon has been demonstrated to affect the metabolic fitness of CD8+ T cells, which may increase their death and lupus severity [23]. The microenvironment in systemic lupus patients favors the generation of double-negative T cells (CD4−, CD8−, TCRαβ+), which facilitates the secretion of IL-17 and increases disease severity [24]. IL-17 can further expand the proliferation of Th17 and autoreactive T and B cells which can worsen the lupus symptoms. The autoreactive T cells such as nuclear antigen-specific CD4+ T cells are elevated in lupus patients and were further correlated with kidney manifestations such as lupus nephritis [25]. Increased expression of hypoxia-inducing factor-1α (HIF-1α) was found to be associated with an expansion of Th17 cells (Fig. 3), depicting that metabolic alteration can also regulate the lupus etiology [26]. CD4+ follicular T cells were also found to be associated with B cell maturation and generation of autoantibodies in lupus patients [27]. Absent in melanoma 2 (AIM2) expression is elevated in B cells from lupus patients and was further correlated with Blimp expression and autoantibody generation [28]. The failure of an immune regulatory system such as Treg cells, and the over-activation of plasma cells are well-known mechanisms in lupus etiology. Recent advances also elaborated on the mechanisms associated with immunoregulation failure. Thus, the understanding of these mechanisms will provide the base for the immune regulations that can act as targets for various therapeutics (Table 1).Fig. 3 Recently elucidated mechanisms and their interactions involved in systemic lupus erythematosus (SLE) modification. Regulation of innate response and enhanced pro-inflammatory cytokines secretion through polymorphism have been recently identified in SLE. Role of mitochondrial dysfunction and associated oxidative stress in metabolic alterations and their impact on immune cell activity and differentiation have been majorly focused in recent studies on SLE. Novel signaling mechanisms have also been elaborated related to metabolic fitness of T cell and their importance in SLE immunobiology. Implications of inflammatory cytokines rich microenvironment on immune cell differentiation specifically, on T cells has been recently focused in SLE related studies. SNP Single nucleotide polymorphism, TLR toll- like receptor, HIF-1α Hypoxia inducing factor 1α, PER2 Period circadian protein homolog 2, DNMT3B DNA-methyltransferase 3 beta, VDAC voltage- dependent anion channel, EZH2 Enhancer of zeste homolog 2, PKC protein kinase C, STAT6 signal transducer and activator of transcription 6, GATA3 GATA binding protein 3, pDC plasmacytoid dendritic cells, ITAM Immunoreceptor tyrosine-based activation motif, ZAP-70 Zeta chain associated protein kinase 70 mTORC1 mammalian target of rapamycin complex 1, IRAK interleukin-1 receptor-associated kinases, IFNα interferon α, Ras-MPK Ras-mitogen-activated protein kinase, Cxcl10 C-X-C motif chemokine ligand 10, ISGs interferon stimulated genes, CD3 cluster of differentiation 3
Table 1 Function of immune cells in lupus pathogenesis
Immune cell Receptor/therapeutic target Function in lupus pathogenesis References
Innate immunity
Neutrophils IL-18R Enhance Netosis [6]
TLR Secretion of IFN gamma [8]
Plasmacytoid Dendritic cells TLR-7, 9 Mediate IFN alpha production [10]
IgE Enhances the follicular T cell expansion and reduces the Treg population [18]
Myeloid Dendritic cells T cells Phagocytize nuclear material and present antigens to naive CD4+ T cells [16]
Platelets TLR-7 Platelet–leukocyte aggregates possibly involved in priming the immune system [12]
Adaptive immunity
Treg cells T and B cells Over-activation of T and B cells [21]
T cells T cell receptor Secretion of IL-17 and proliferation of Th17 [24]
CD4+ follicular T cells B cell B cell maturation and autoantibodies generation [27]
B cells ProBDNF/p75NTR Autoantibody generation [39]
Aberrant signaling in SLE
Cellular hyperactivity and hyperresponsiveness associated with deregulated signaling pathways in T and B lymphocytes of SLE suggest detailed analysis for a better understanding of signaling events to pave the path for better management and prevention of this complex disease. T cell receptor (TCR) is a heterodimer, consisting of the TCRα and TCRβ, which recognize antigenic peptides presented by MHC on antigen-presenting cells. CD3 proteins (δ, ε, γ, and ζ) have also been assembled with TCR. CD3ζ contains three immunoreceptors tyrosine-based activation motif (ITAM) domains, and the phosphorylation of ITAM by Src kinase recruits the spleen tyrosine kinase (Syk) family kinase ζ-associated protein kinase 70 (ZAP-70), resulting in the activation of ZAP-70 (Fig. 3) [9]. This results in altered calcium flux, activation of protein kinase C (PKC), and recruitment of Ras guanine-releasing protein 1 leading to activation of the Ras-mitogen protein kinase pathway. The reduced expression level of CD3ζ protein contributes to the aberrant signaling phenotype of SLE T cells [29]. Dual specific phosphatase, a regulator of mitogen-activated protein kinase (MAPK), was also over-expressed (Fig. 3) in T cells from SLE patients [30]. Protein phosphatase 2A (PP2A) regulatory subunit (PPP2R2A) has been demonstrated to regulate the Th1 and Th17 differentiation, but not of Tregs, and their deficiency may trigger autoimmune-like conditions [31]. STAT6-GATA3 signaling axis (Fig. 3) acts as double edge sword in SLE pathology as it enhances the Treg cell differentiation but also leads to the expansion of CD8+ T cells that can secrete IL-13 and IFN-ϒ [32]. Enhanced type-I interferon has been demonstrated in both lupus models and patients where it increases the STAT4 expression which leads to secretion of follicular CD4+ T cells dependent cytokines and is associated with autoantibody generation [33]. An increased level of the mechanistic target of rapamycin (mTOR) (Fig. 3) has been found to have an important role in memory CD8+ cell regulation and maintenance via glyco-metabolism [34]. B cells from SLE expressed high Ca2+ in response to B cell receptor (BCR) stimulation associated with increased tyrosine phosphorylation. Defective B cell signaling recruits inhibitory phosphatase SH2 domain-containing inositol 5′-phosphatase (SHIP) via the inhibitory FcγRIIb receptor [35]. FOXM1, a transcriptional factor of the cell cycle, was found to be highly increased in plasmablast, naïve, and memory B cells, thus, expanding their number in lupus patients [36]. A decrease in Lyn and Syk and an increase in ZAP-70 expression were found in B cells from active SLE patients [37]. Recently, Leptin has been demonstrated to induce B cell dysfunction via activating the JAK/STAT3/5 and ERK1/2 pathways in patients with systemic lupus erythematosus [38]. In a recent study, the higher expression of brain-derived neurotrophic factor precursor and its high-affinity pan-75 receptor (ProBDNF/p75NTR) has been observed in B cells from lupus patients and was positively correlated with disease severity and autoantibody generation [39]. Therefore, the evaluation of these signaling cascades involved in complex immunobiology provides a molecular mechanism that can be used as a diagnostic biomarker or novel targets for various therapies.
Cytokine network in SLE
Cytokines have an important role in the immunobiology of lupus which connects the innate and adaptive immune systems in many ways. The secretion of cytokines from one cell will further modulate the activation and differentiation of other immune cells resulting in the secretion of subsequent cytokines. Cytokines which are comprised of chemotactic activity, known as chemokines, also play an important role in recruiting various immune cells. Different levels of cytokines may regulate the inflammatory and anti-inflammatory immune response, although, in lupus, cytokines associated with inflammatory response were found to be enhanced which triggered the over-activated immune response. The increase in various pro-inflammatory cytokines such as CXCL10, IL-6, IL-2, and interferon-ϒ production in SLE patients results in reinforcement of Th1 differentiation and naïve T cell proliferation leading to IgG production from B cells [40]. Exacerbated production of CXCL10, IL-6, IL-2, and interferon-ϒ (Fig. 3) was associated with a memory-like phenotype in CD4+ polarized B cells towards pro-inflammatory B cells in SLE patients [41]. High mobility group box 1 protein (HMGB1) and type-I IFN are the key molecules that promote the autoreactivity process in SLE and both can be used as the biomarkers for detection of disease severity [42]. Other important cytokines such as high levels of IL-8, macrophage inflammatory protein (MIP) 1α, and MIP1β were found to be associated with high disease severity in SLE patients [43]. IL-25 mRNA serum level was also found to be elevated in SLE patients and correlated with disease severity [44]. Decreased concentration of IL-35 and IL-35+ Breg cells was observed in SLE patients which describes their protective role in SLE pathogenesis [45]. Higher expression of IL-18 was also observed in SLE patients and can be used as a potential biomarker for disease severity [46]. The serum IL-21 level was found to be positively correlated with lupus nephritis activity and can be used potentially as the biomarker [47]. Another cytokine IL-33, which has been demonstrated as the important regulator of TLR-4 was found to be elevated in lupus patients in comparison to healthy controls [48]. As cytokines have an important role in regulating the innate and adaptive immune response, various targeting strategies have been adopted to improve disease outcomes.
The complex immunobiology of lupus includes the over-immune response triggers due to the overactive immune cells towards self-antigens. Various factors including the pro-inflammatory cytokines, chemokines, apoptosis, over-activated APCs, and failure of phagocytotic activity have been well defined to be involved in lupus etiology. These factors can vary from population to population and besides these factors, other important aspects such as genetics, epigenetics, and environmental factors can also be an important part of lupus pathology and can modulate the immunobiology and disease severity.
Genetic factors
Besides the overstimulation of the immune system towards the self-antigens via innate and adaptive immune networks, various other susceptible factors such as genetics can also regulate immunobiology and etiology of lupus. A strong familial aggregation has been observed in SLE along with higher frequency among first-degree relatives, and a higher chance of developing SLE in siblings of the patient shows the polygenic inheritance of the disease. Polygenic inheritance pattern shows the complexity and heterogeneity nature of the disease, which will require an advance and hassle-free analysis method to elucidate the mechanisms behind disease etiology. Genome-wide association studies (GWAS) have been performed to identify the susceptible loci associated with SLE heterogeneity. A novel gene, ILRUN was recently reported by gene-wide association analysis whose expression was significantly low in SLE patients as compared to healthy control [49]. The concurrence of SLE in identical twins is approximately 25–50% and that in dizygotic twins is around 5% [50]. Population studies reveal the positive haplotype associated with HLA genes in SLE (HLA-DR3; DR9; DR15; DQA1*0101), and with lupus nephritis (LN) (DQA1*0101; DR3; DR15), while the possible protective haplotypes noted are HLA-DR4, DR11, DR14 [51]. These population-based genetics studies can enhance the understanding of personalized approaches to genetic susceptibility in lupus pathology that will further assist in solving the complexity of lupus on an individual basis. The non-HLA genes were also found to be of great importance as the single-nucleotide polymorphism (SNP) in genes such as PTPN2, STAT4, RUNX, and SLC have been positively correlated with SLE susceptibility specifically in Belarusian women [52]. TLRs are overstimulated on innate cells from SLE patients and their activation has been linked with SNP (Fig. 3), where specifically the polymorphism in TLR 5 and 9 were found to be associated with a higher risk of nephritis and SLE in the Egyptian population [53]. Polymorphism in DNMT3B (Fig. 3), an important enzyme for DNA methylation, has been found to have a positive correlation with disease severity and with co-existing periodontitis in the Brazilian population, depicting the interplay between genetics and epigenetic factors and lupus pathology [54]. Polymorphism associated with circadian rhythms can also play an important role in disease, where the SNP in the Period 2 gene (PER2), a circadian clock regulatory gene, was also found to be positively associated with the clinical manifestation of SLE pathogenesis [55]. Furthermore, analysis of SNP in various cytokines including IL-1β, IL-17, and IL-10 (Fig. 3) was found to be related to disease etiology in SLE patients of Indian origin [56, 57].
Monogenic lupus
Beyond the complex genetics’ perturbations, rare mutations in genes can also modulate the complexity of lupus pathology in the Mendelian inheritance pattern termed “monogenic lupus”. Recent examples of monogenic lupus such as the homozygous mutation in the DNASE1L3 gene were found to be associated with urticarial skin lesions, recurrent hemoptysis, and renal involvement in pediatric lupus patients [58]. Deleterious mutations in lipopolysaccharide-responsive beige-like anchor (LRBA) genes lead to its deficiency which was found to be associated with juvenile lupus [59]. Mutation (p.His198Gln) in the C1QTNF4 gene was found to be potentially involved in SLE risk in the Iranian population [60]. Current evidence in the genetic susceptibility of lupus shows that genes can also potentially regulates the dysregulated immunological mechanisms associated with disease manifestation that will be further modified through various epigenetic phenomenon. Therefore, the genetic analysis may lead to the identification of novel susceptible targets which are associated with the complex network of lupus and will be further targeted with potent therapeutics. Besides genetics, epigenetic events have been shown to play an important role in the pathophysiology of SLE.
Epigenetic factors
Epigenetics is an important phenomenon that can regulate gene expression in a stable, sometimes heritable fashion. Epigenetics mechanisms such as DNA methylation, post-translational histone modification, and micro RNAs proved to have an important role in SLE (Fig. 4). Alterations such as DNA hypomethylation in both T and B cells were found to be associated with disease pathology [61]. Although major epigenetic investigations have been carried out in T cells, mechanisms such as DNA hypomethylation in T cells were found to be positively correlated with disease activity [62]. Sex-based comparison of methylation pattern in CD4+ T cells shows the dysregulated apoptosis and pro-inflammatory effect in males that were associated with epigenetic over-activation of the Rho family GTPase pathway which enhances the Th17 cell differentiation [63]. Enhancers of zeste homolog 2 (EZH2), a histone methyl transferase and part of polycomb repressive complex 2 (PRC2) which regulates the DNA methyl transferase (DNMT) in several ways were found to be highly expressed in CD4+ T cells from lupus patients [64]. Increased expression of EZH2 can be mediated by higher glycolysis and mechanistic target of rapamycin (mTORC1) activation in lupus T cells leading to metabolic alterations which shows the interplay between immunometabolism and epigenetic modification [65]. Decreased expression of 3-hydroxybutyrate dehydrogenase type 2 (BDH2) has been reported in CD4+ T cells of SLE patients which results in overstimulation of ten-eleven translocation (TET) protein leading to hypomethylation of DNA and activation of autoreactivity related genes [66]. Hypomethylation of the CD-70 promoter region was observed in T cells of juvenile SLE patients and was found to be positively correlated with disease activity [67]. Site-specific hypomethylation of the CD40-ligand gene (CD40L) in CD4+ T cells in SLE patients was also associated with disease activity and severity [68]. Contrary to hypomethylation, the hypermethylated 5-methyl cytosine (m5c) region of the gene has also been observed in CD4+ T cells of SLE patients, which were found to be significantly involved in the modulation of the immune system, cytokines secretion, and various inflammatory mechanisms [69]. Hypermethylation of the FOXP3 gene was also investigated in T cells from children with SLE and demonstrated as one of the causes of increased immune response as FOXP3 acts as an important transcription factor for Treg cell differentiation [70]. Hypermethylation of genes in autoreactive peripheral blood mononuclear cells (PBMCs) from SLE patients can be inhibited via MEK/ERK signaling by co-culture them with mesenchymal stem cells which demonstrated the therapeutic potential of mesenchymal stem in SLE [71]. Methylation in the promoter region of cytotoxic T lymphocytes associated antigen-4 gene (CTLA4) in CD8+ T cells was found to be positively correlated with SLE pathogenesis as CTLA4 has an important role in inhibitory immune response [72]. Oxidative stress in lupus is well determined which can also enhance DNA methylation defects that can result in angiotensin-converting enzyme 2 (ACE2) hypermethylation, which makes lupus patients more prone to COVID-19 infection [73]. Expression of Growth arrest-specific 5 (GAS5), a long non-coding RNA (LncRNAs) that controls cell response and apoptosis, was observed to be low in CD4+ T cells which were further correlated with higher disease activity in lupus patients [74]. N4-acetylcysteine, a novel mRNA modification may have translational efficiency in CD4+ T cells which can be associated with inflammation and critical immune response in SLE patients [75]. Besides the methylation-based mechanisms, micro-RNA such as miR-152-3p was highly expressed in CD4+ T cells and found to be associated with double-stranded DNA and IgG antibodies [76]. miR-183-5p expression levels were found significantly high in patients and were positively correlated with SLEDAI score and anti-dsDNA antibody; thus, this micro-RNA can be a promising biomarker of SLE [77].Fig. 4 Epigenetics implications modifies the various mechanisms involved in lupus pathology. Oxidative stress and diet have been recently shown as the important triggers of the epigenetic events demonstrates the importance of environmental factors in disease etiology. Methylation patterns have been found potentially involved in modulation of immunobiology via regulating the T reg and B cell activity. Hypermethylation of the anti-inflammatory sites and hypomethylation of pro-inflammatory genes have been demonstrated in recent studies. Epigenetic events can potentially modulate the lupus outcomes and can be utilized for diagnosis as biomarkers. SNP single-nucleotide polymorphism, BAFF B-cell activating factor, IFI 44L Interferon-inducible 44 like, PBMC Peripheral blood mononuclear cells, FOXP3 forkhead box P3, BDH2 3-hydroxybutyrate dehydrogenase 2, TET ten–eleven translocation, ROS reactive oxygen species
B cells in both pediatric and adult SLE patients have a significant reduction in epigenetic modification on the inactive X chromosome and aberrant X-linked gene expression can underlie the mechanism of female bias of SLE and abnormal autoantibody production [78]. EZH2, a histone methyl transferase, was highly expressed in germinal center (GC) B cells and involved in SLE pathogenesis through enhanced autoantibody generation [79]. MiR-29a, a micro-RNA, can also affect autoantibody secretion in B cells by modulating the Crk-like protein (CKL), thereby contributing to SLE pathogenesis [80]. Inadequate expression of miR-1246 has been found in B cells from lupus patients and was positively correlated with p53 deficiency [81]. DNA methylation patterns may vary during different stages of B cell development. Global DNA methylation in B cells is decreased, but hypermethylation in a few specific genes may be associated with SLE pathogenesis [82]. Hypomethylation of HERS-1 in B cells was found to be associated with SLE disease activity [83]. Hypomethylation of various cytokine genes associated with B cell activation and differentiation, such as interferon-induced protein 44 like (IFI44L) and B cells activation factor (BAFF) in PBMCs was found to be correlated with increased autoantibody secretion from B cells in SLE patients [84]. Global histone modification analysis reveals that H3 and H4 in B cells (Fig. 3) from lupus patients were hypo-acetylated and positively correlated with autoantibody generation and SLEDAI score [85]. Epigenetic alteration in lupus was found to be associated with different cellular mechanisms to guide their nuclear and cytoplasmic factors to regulate the different transcription/translational processes [4]. Epigenetic processes in immune cells regulate their function and bridge the gap between genomics and environmental factors in the etiology of SLE.
Environmental factors
Physical/chemical triggers
Various environmental factors lead to the ensuing of SLE in genetically predisposed individuals towards SLE. These environmental factors include lifestyle factors (cigarette smoking and alcohol drinking) occupational exposure, oral contraceptives, dietary causes, pollution, viral infections especially Epstein–Barr virus, etc. These environmental stimuli may lead to epigenetic changes as they can cause the inhibition of DNA methyl transferases which further leads to hypomethylation of DNA specifically in CD4+ T cells from genetically predisposed individuals. Hypomethylation also leads to oxidative stress, which further activates the signaling pathway mediated by protein kinase C which finally decreases the level of the extracellular signal-regulated kinase (ERK). It is noticeable that the expression of ERK protein was found to be decreased in CD4+ T cells of SLE patients [86]. Exposure to UV radiation is known to worsen pre-existing SLE but its role in the development of SLE is still unclear. However, UV exposure was also known to play an important role in the production of the active form of vitamin D. Vitamin D once converted to 1α,25(OH)2D3 form may prove to be immunosuppressive, hence reducing the risk of SLE [87]. A study on the susceptible subgroup of individuals (relatives of a patient with SLE), suggested an important clue that vitamin D-deficient individuals were more prone to lupus development [88]. Although the cause and consequence relationship between vitamin D and SLE is not specified yet, large-scale studies with controlled experiments are needed to determine the role of UV exposure and SLE incidence.
Occupational exposure to chemical agents is an imperative factor in the development of SLE [89]. Silica exposure and incidence of SLE have been reported in both urban and rural areas. Similarly, silicates such as asbestos have been reported to be associated with the production of anti-nuclear antibodies and proteinuria along with an elevated risk of RA [90]. Lupus-prone NZ-2410 mice exposed to silica have been demonstrated to increase circulating immune complex, autoantibodies, renal deposits of C3, and proteinuria [91].
SLE has been linked to air pollution via a preliminary genome-wide assessment of DNA methylation study which indicates that SLE patients residing near highways have hypomethylated ubiquitin gene (UBE2U gene) encoding enzyme that is involved in ubiquitination and DNA repair [92]. However, more in-depth studies are needed to confirm the above correlation. Epidemiologic studies have established that exposures to petroleum distillates, trichloroethylene, and organochlorines are linked to the intensity of symptoms in lupus patients [93]. Another study of mercury-exposed gold miners showed a higher level of ANA in comparison to miners in diamond and emerald mines with no mercury exposure [94].
Cigarette smoking leads to the inhalation of toxic substances such as tars, nicotine, carbon monoxide, polycyclic aromatic hydrocarbons, and free radicals. Smoking leads to oxidative stress which causes demethylation of DNA and increased expression of inflammatory genes leading to lupus-like manifestations [95]. Various epidemiological studies have confirmed that cigarette smoking is strongly correlated with the risk of SLE incidence [96]. Both cigarette smoking and hypoxia can elevate oxidative stress which has multifactorial effects to induce autoimmunity such as the generation of autoreactive T cells and autoantibodies, inhibition of Treg activity, and enhanced expression of pro-inflammatory mediators [97]. These toxic components bring about damage to biomolecules such as proteins and DNA resulting in genetic aberrations causing gene activation which may be involved in SLE development via an imbalance of antioxidant mechanism and oxidative stress. In a population-based cohort study, it has been demonstrated that nitrogen oxides (NOX) in polluted air and drinking water may lead to significant predilections of SLE, especially for patients with renal involvement [98]. Contrary to the conventional notion, alcohol consumption has been shown to relieve the inflammatory milieu in SLE, diminishing the response to immunogens and decreasing pro-inflammatory cytokines [99].
Infections
SLE patients are generally susceptible to major infections that might be due to profound immunological disturbances [100]. Infections cause changes in normal immune regulation and stimulate immune pathways mediated by molecular mimicry. Adults and children with SLE have higher rates of seropositivity for Epstein–Barr virus (EBV) than other individuals [101]. A possible mechanism involved the complex formation between viral RNA and single-strand binding protein (SSB) which stimulates the TLR3 receptor to induce TNF-α. Furthermore, molecular mimicry between EBV and SLE antigens is another mechanism involved in this process [102]. A study identified a molecular mimicry between SARS-CoV-2 antigen and human molecular chaperons that may induce autoimmunity against the endothelial cells [103]. The mimicry between the symptoms of COVID-19 infections and SLE flares can be of clinical interest to find out the incidence of COVID in SLE patients [104]. Although the incidence of COVID-19 infection in SLE patients was slightly low and the SLE patients with major organ involvement were found to have asymptomatic COVID-19 manifestation that might be due to the high dose of corticosteroids and immunosuppressive treatment regime among SLE patients [105].
Gut microbiota
Recent developments in research have indicated a correlation between gut microbiota and SLE disease activity. In lupus, the ratio of Firmicutes to Bacteroidetes is lower and several other genera are found in abundance [106, 107]. Katz et al. have reported that there is a decrease in the number of Lactobacillaceae and an increase in Lachnospiraceae in patients with SLE [108]. A study in young lupus-prone mice showed clear depletion of lactobacilli and upsurges in Lachnospiraceae in comparison to age-matched healthy controls [109]. However dietary intervention with retinoic acid could restore the number of lactobacilli associated with improvement in symptoms [110]. Furthermore, an increase in Ruminococcus gnavus of the Lachnospiraceae family led to raised serum sCD14 and elevated levels of fecal secretory IgA and calprotectin in female SLE patients [111]. In SLE patients, a leaky gut may elevate the serum endotoxin lipopolysaccharide (LPS) level which is suggestive of chronic microbial translocation that can contribute to the pathogenesis of SLE [112]. Similarly, in lupus-prone NZBxW/F1 mice, complexes of bacterial amyloid and DNA have been shown to stimulate autoimmune responses such as the production of type-I IFN and autoantibodies [113]. Although the definite role of gut symbiotic or pathogenic microbes in SLE is yet to be clarified, the pristane-induced lupus model demonstrated the beneficial role of Lactobacillus probiotics is linked to the reduction of Th1, Th17, and cytotoxic T lymphocytes [114]. The cross-talk between the host and commensal microbiome in addition to infectious bacteria, viruses, and parasites can also impact the presentation of autoimmune diseases. Implications of alteration in the microbiome because of environmental mediators are imperative discoveries and need to be evaluated carefully for their role, especially when establishing a cause-and-effect relationship.
Diet
Intake of excessive carbohydrates has been projected as a risk factor that can worsen the clinical manifestations of autoimmune diseases like rheumatoid arthritis and SLE [115]. Obesity is a well-known mediator of low-grade inflammation mediated by the initiation of several pathways associated with inflammatory cytokine expression such as TNF-α and IL-6 [116, 117]. This favors a continuous inflammatory response, partly contributing to co-morbidities seen in SLE patients [115]. It has also been demonstrated that mice fed with a high-fat diet can have elevated levels of oxidative stress and inflammation leading to autoimmunity [118]. Undoubtedly, SLE patients are at high risk of developing metabolic syndrome, insulin resistance, and type 2 diabetes mellitus leading to a higher risk of cardiovascular co-morbidities and also a major cause of premature death in SLE patients [119]. In SLE, obesity has been linked with higher disease activity and cumulative organ damage [120]. The influence of obesity on gene expression has also been positively correlated with disease severity [121]. Hence, clinical interventions involving meditation and exercise can ease lupus symptoms. Conclusively, SLE patients should maintain a balanced diet avoiding excess calories in addition to following an active lifestyle with daily exercise.
Restriction in protein intake especially in patients with lupus nephropathy has been taken as a beneficial approach to reduce the burden on the kidney. The renal function in SLE patients could be improved by a moderate protein intake of 0.6 g/kg/day [122]. A high salt diet can also accelerate the development of lupus through the regulation of dendritic cell activity via MAPK and STAT1 signaling pathways [123]. Higher dietary sodium and lower dietary potassium intake were significantly correlated with C-reactive protein (CRP), which may lead to an increase in disease severity [124]. A diet rich in eicosapentaenoic acid (EPA) was found to ameliorate the lupus nephritis manifestations including immune complex accumulation and autoantibody generation in the kidney [125].
Vitamins also have an important role to play in the pathogenesis of SLE as vitamins A, B C, D, and E are known to improve clinical manifestations of SLE [126]. Vitamin C has been shown to reduce oxidative stress and inflammation and decrease levels of antibodies against dsDNA, IgG, etc.; and intake of 1 g of vitamin C per day is recommended [122]. A cross-sectional study on 280 patients with SLE demonstrated that the Mediterranean diet can exert a beneficial effect on disease activity and cardiovascular risk [127].
Conclusion
Various complicated and multifactorial aspects have been implicated in SLE pathogenesis. Multiple genes and various epigenetic modifications confer susceptibility to the development of this complex disease. Mitochondrial dysfunction has time and again been demonstrated in SLE, and it is one of the important parameters. The epigenetic modification of various lupus-associated genes has also been implicated in metabolic alterations through direct or indirect mechanisms that can further be modified by environmental factors such as diet. Desirable healthy gut microbiota can be kept in the best form through diet management and can help in regulating homeostasis. These efforts may prevent the adverse effects of various therapies and improve the mental and physical health of SLE patients. Defective immune regulation such as clearance of apoptotic cells and accumulation of antigens, cytokine imbalance, loss of self-tolerance, excess T cells, and immune complex infiltration in various organs are major role players in SLE. Novel approaches towards the signaling defects in T and B cells have been made which reflect the complex nature of the pleiotropic effect in SLE. These signaling abnormalities offer both hope and challenges in the direction of therapeutic interventions. Additionally, analysis of various targets and classification of patients based on microbiota composition may contribute to the development of more personalized strategies in SLE treatment.
Authors contributions
Concept design: AA, RB, and KA. Data collection: AA, AT, and KA. Data analysis and interpretation: AA and RB. Drafting manuscript: AB and AT. Revising manuscript: AA and AB. All authors take full responsibility for the integrity and accuracy of all aspects of the work.
Funding
The authors thank “University Grant Commission, New Delhi” (Grant no. 191620089214) and “Department of Biotechnology (DBT-BUILDER), New Delhi” (Grant no. BT/INF/22/SP41295/2020) for providing fellowship to Mr. Akhil Akhil and Mr. Rohit Bansal, respectively.
Declarations
Conflict of interest
The authors declare no conflict of interest.
Human and animal participants
Not required for this study.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Alaanzy MT Alsaffar JM Bari AA Evaluate the correlation between antioxidant capacity and interferon Γ level with the disease activity of Sle patients in Iraqi Woman Indian J Public Health Res Dev 2020 11 1 1278 1282 10.37506/v11/i1/2020/ijphrd/194018
2. Terahara N Flavonoids in foods: a review Nat Prod Commun 2015 10 3 19345781501000334
3. Mok C Lau C Pathogenesis of systemic lupus erythematosus J Clin Pathol 2003 56 7 481 490 10.1136/jcp.56.7.481 12835292
4. Adams DE Shao W-H Epigenetic alterations in immune cells of systemic Lupus erythematosus and therapeutic implications Cells 2022 11 3 506 10.3390/cells11030506 35159315
5. Katsuyama T Tsokos GC Moulton VR Aberrant T cell signaling and subsets in systemic lupus erythematosus Front Immunol 2018 9 1088 10.3389/fimmu.2018.01088 29868033
6. Ma J Elevated interleukin-18 receptor accessory protein mediates enhancement in reactive oxygen species production in neutrophils of systemic lupus erythematosus patients Cells 2021 10 5 964 10.3390/cells10050964 33919154
7. Bertelli R Neutrophil extracellular traps in systemic lupus erythematosus stimulate IgG2 production from B lymphocytes Front Med 2021 8 635436 10.3389/fmed.2021.635436
8. Liu Y Kaplan MJ Neutrophil dysregulation in the pathogenesis of systemic lupus erythematosus Rheum Dis Clin 2021 47 3 317 333 10.1016/j.rdc.2021.04.002
9. Pan L Immunological pathogenesis and treatment of systemic lupus erythematosus World J Pediatr 2020 16 1 19 30 10.1007/s12519-019-00229-3 30796732
10. Nasser N Kurban M Abbas O Plasmacytoid dendritic cells and type I interferons in flares of systemic lupus erythematosus triggered by COVID-19 Rheumatol Int 2021 41 5 1019 1020 10.1007/s00296-021-04825-3 33666725
11. Fillatreau S Manfroi B Dörner T Toll-like receptor signalling in B cells during systemic lupus erythematosus Nat Rev Rheumatol 2021 17 2 98 108 10.1038/s41584-020-00544-4 33339987
12. Brilland B Scherlinger M Khoryati L Goret J Duffau P Lazaro E Charrier M Guillotin V Richez C Blanco P Platelets and IgE: shaping the innate immune response in systemic lupus erythematosus Clin Rev Allerg Immunol 2019 58 1 19
13. Marques CP Expression of Toll-like receptors 2 and 4 in the saliva of patients with systemic lupus erythematosus and chronic periodontitis Clin Rheumatol 2021 40 7 2727 2734 10.1007/s10067-020-05560-z 33570702
14. Liu J Zhang X Cao X Dendritic cells in systemic lupus erythematosus: from pathogenesis to therapeutic applications J Autoimmun 2022 132 102856 10.1016/j.jaut.2022.102856 35773101
15. Charrier M The role of dendritic cells in systemic lupus erythematosus Systemic Lupus Erythematosus 2021 Amsterdam Elsevier 143 150
16. Sim TM Type I interferons in systemic lupus erythematosus: a journey from bench to bedside Int J Mol Sci 2022 23 5 2505 10.3390/ijms23052505 35269647
17. Brilland B Platelets and IgE: shaping the innate immune response in systemic lupus erythematosus Clin Rev Allergy Immunol 2020 58 2 194 212 10.1007/s12016-019-08744-x 31254159
18. Palomares O Regulatory T cells and immunoglobulin E: a new therapeutic link for autoimmunity? Allergy 2022 77 3293 3308 10.1111/all.15449 35852798
19. Li M Interferon-α activates interleukin-1 receptor-associated kinase 1 to induce regulatory T-cell apoptosis in patients with systemic lupus erythematosus J Dermatol 2021 48 8 1172 1185 10.1111/1346-8138.15899 33882150
20. Scherlinger M Selectins impair regulatory T cell function and contribute to systemic lupus erythematosus pathogenesis Sci Transl Med 2021 13 600 eabi4994 10.1126/scitranslmed.abi4994 34193612
21. Kurata I Impaired function of PD-1+ follicular regulatory T cells in systemic lupus erythematosus Clin Exp Immunol 2021 206 1 28 35 10.1111/cei.13643 34240405
22. Lima G Exhausted T cells in systemic lupus erythematosus patients in long-standing remission Clin Exp Immunol 2021 204 3 285 295 10.1111/cei.13577 33475152
23. Buang N Type I interferons affect the metabolic fitness of CD8+ T cells from patients with systemic lupus erythematosus Nat Commun 2021 12 1 1 15 10.1038/s41467-021-22312-y 33397941
24. Li H Systemic lupus erythematosus favors the generation of IL-17 producing double negative T cells Nat Commun 2020 11 1 1 12 31911652
25. Abdirama D Nuclear antigen–reactive CD4+ T cells expand in active systemic lupus erythematosus, produce effector cytokines, and invade the kidneys Kidney Int 2021 99 1 238 246 10.1016/j.kint.2020.05.051 32592813
26. Liao H-J Increased HIF-1α expression in T cells and associated with enhanced Th17 pathway in systemic lupus erythematosus J Formos Med Assoc 2022 121 2446 2456 10.1016/j.jfma.2022.05.003 35599104
27. Nakayamada S Tanaka Y Clinical relevance of T follicular helper cells in systemic lupus erythematosus Expert Rev Clin Immunol 2021 17 10 1143 1150 10.1080/1744666X.2021.1976146 34469695
28. Yang M AIM2 deficiency in B cells ameliorates systemic lupus erythematosus by regulating Blimp-1–Bcl-6 axis-mediated B-cell differentiation Signal Transduct Target Ther 2021 6 1 1 11 33384407
29. Chen P-M Tsokos GC T cell abnormalities in the pathogenesis of systemic lupus erythematosus: an update Curr Rheumatol Rep 2021 23 2 1 9 10.1007/s11926-020-00978-5
30. Chuang H-C Tan T-H MAP4K family kinases and DUSP family phosphatases in T-cell signaling and systemic lupus erythematosus Cells 2019 8 11 1433 10.3390/cells8111433 31766293
31. Pan W The regulatory subunit PPP2R2A of PP2A enhances Th1 and Th17 differentiation through activation of the GEF-H1/RhoA/ROCK signaling pathway J Immunol 2021 206 8 1719 1728 10.4049/jimmunol.2001266 33762326
32. Kato H Perl A Double-edged sword: interleukin-2 promotes T regulatory cell differentiation but also expands interleukin-13-and interferon-γ-producing CD8+ T Cells via STAT6-GATA-3 axis in systemic lupus erythematosus Front Immunol 2021 12 635531 10.3389/fimmu.2021.635531 33763079
33. Dong X Type I interferon–activated STAT4 regulation of follicular helper T cell–dependent cytokine and immunoglobulin production in Lupus Arthritis Rheumatol 2021 73 3 478 489 10.1002/art.41532 33512094
34. Cai X mTOR participates in the formation, maintenance, and function of memory CD8+ T cells regulated by glycometabolism Biochem Pharmacol 2022 204 115197 10.1016/j.bcp.2022.115197 35926651
35. Yap DY Chan TM B cell abnormalities in systemic lupus erythematosus and lupus nephritis—role in pathogenesis and effect of immunosuppressive treatments Int J Mol Sci 2019 20 24 6231 10.3390/ijms20246231 31835612
36. Akita K Interferon α enhances B cell activation associated with FOXM1 induction: potential novel therapeutic strategy for targeting the plasmablasts of systemic lupus erythematosus Front Immunol 2020 11 498703 10.3389/fimmu.2020.498703 33633721
37. Vásquez A Altered recruitment of Lyn, Syk and ZAP-70 into lipid rafts of activated B cells in systemic lupus erythematosus Cell Signal 2019 58 9 19 10.1016/j.cellsig.2019.03.003 30840855
38. Chen H Leptin accelerates B cell dysfunctions via activating JAK/STAT3/5 and ERK1/2 pathways in patients with systemic lupus erythematosus Clin Exp Rheumatol 2022 40 11 2125 2132 35084324
39. Shen W-Y Up-regulation of proBDNF/p75NTR signaling in antibody-secreting cells drives systemic lupus erythematosus Sci Adv 2022 8 3 eabj2797 10.1126/sciadv.abj2797 35044824
40. Simon Q et al (2021) A cytokine network profile delineates a common Th1/Be1 pro-inflammatory group of patients in four systemic autoimmune diseases. Arthritis Rheumatol 73. 10.1002/art.41697
41. Simon Q A pathogenic cytokine network is associated with pro-inflammatory B cells in systemic lupus erythematosus patients SSRN J 2019 10.2139/ssrn.3420376
42. Tanaka A Serum high-mobility group box 1 is correlated with interferon-α and may predict disease activity in patients with systemic lupus erythematosus Lupus 2019 28 9 1120 1127 10.1177/0961203319862865 31299881
43. Park J Cytokine clusters as potential diagnostic markers of disease activity and renal involvement in systemic lupus erythematosus J Int Med Res 2020 48 6 0300060520926882 10.1177/0300060520926882 32489126
44. Li Y Interleukin-25 is upregulated in patients with systemic lupus erythematosus and ameliorates murine lupus by inhibiting inflammatory cytokine production Int Immunopharmacol 2019 74 105680 10.1016/j.intimp.2019.105680 31200339
45. Ye Z The plasma interleukin (IL)-35 level and frequency of circulating IL-35+ regulatory B cells are decreased in a cohort of Chinese patients with new-onset systemic lupus erythematosus Sci Rep 2019 9 1 1 12 10.1038/s41598-019-49748-z 30626917
46. Elshal A Serum interleukin-18 as a novel biomarker for disease activity of systemic lupus erythematosus patients J Egypt Soc Parasitol 2019 49 1 145 152 10.21608/jesp.2019.68297
47. Shater H The potential use of serum interleukin-21 as biomarker for lupus nephritis activity compared to cytokines of the tumor necrosis factor (TNF) family Lupus 2022 31 1 55 64 10.1177/09612033211063794 34978958
48. Li Y Potential role of interleukin-33 in systemic lupus erythematosus by regulating toll like receptor 4 Eur J Inflamm 2022 20 1721727X221094455 10.1177/1721727X221094455
49. Ding X Gene-based association analysis identified a novel gene associated with systemic lupus erythematosus Ann Hum Genet 2021 85 6 213 220 10.1111/ahg.12439 34145571
50. Putra Gofur NR Putri Gofur AR Soesilaningtyas RPGR Kahdina M Systemic lupus erythematsous: a review article Int J Surg Case Rep Images 2021 2 1 1
51. Lever E Alves MR Isenberg DA Towards precision medicine in systemic lupus erythematosus Pharmacogen Personal Med 2020 13 39
52. Dostanko N ab0010 association of some non-hla gene polymorphisms with susceptibility to systemic lupus erythematosus in women in belarusian population 2022 London BMJ Publishing Group Ltd. 1140 1141
53. Gharbia OM Toll-like receptor 5 and Toll-like receptor 9 single nucleotide polymorphisms and risk of systemic lupus erythematosus and nephritis in Egyptian patients Egypt Rheumatol Rehabilit 2021 48 1 1 10
54. Dias LNDS DNMT3B (rs2424913) polymorphism is associated with systemic lupus erythematosus alone and with co-existing periodontitis in a Brazilian population J Appl Oral Sci 2022 10.1590/1678-7757-2021-0567 35507987
55. Dan Y-L Association of PER2 gene single nucleotide polymorphisms with genetic susceptibility to systemic lupus erythematosus Lupus 2021 30 5 734 740 10.1177/0961203321989794 33497301
56. Umare V Cytokine genes multi-locus analysis reveals synergistic influence on genetic susceptibility in Indian SLE–A multifactor-dimensionality reduction approach Cytokine 2020 135 155240 10.1016/j.cyto.2020.155240 32795905
57. Padhi S Interleukin 17A rs2275913 polymorphism is associated with susceptibility to systemic lupus erythematosus: a meta and trial sequential analysis Lupus 2022 31 6 674 683 10.1177/09612033221090172 35353646
58. Ekinci RMK Monogenic lupus due to DNASE1L3 deficiency in a pediatric patient with urticarial rash, hypocomplementemia, pulmonary hemorrhage, and immune-complex glomerulonephritis Eur J Med Genet 2021 64 9 104262 10.1016/j.ejmg.2021.104262 34161863
59. Liphaus BL LRBA deficiency: a new genetic cause of monogenic lupus Ann Rheum Dis 2020 79 3 427 428 10.1136/annrheumdis-2019-216410 31852671
60. Pakzad B C1QTNF4 gene p His198Gln mutation is correlated with early-onset systemic lupus erythematosus in Iranian patients Int J Rheum Dis 2020 23 11 1594 1598 10.1111/1756-185X.13981 33009720
61. Hedrich CM Epigenetics Systemic lupus erythematosus 2021 Amsterdam Elsevier 277 292
62. Zhang Y Impaired DNA methylation and its mechanisms in CD4+ T cells of systemic lupus erythematosus J Autoimmun 2013 41 92 99 10.1016/j.jaut.2013.01.005 23340289
63. He Z Comprehensive analysis of epigenetic modifications and immune-cell infiltration in tissues from patients with systemic lupus erythematosus Epigenomics 2022 14 2 81 100 10.2217/epi-2021-0318 34913398
64. Tsou PS EZH2 modulates the DNA methylome and controls T cell adhesion through junctional adhesion molecule A in lupus patients Arthritis Rheumatol 2018 70 1 98 108 10.1002/art.40338 28973837
65. Zheng X Tsou P-S Sawalha AH Increased expression of EZH2 is mediated by higher glycolysis and mTORC1 activation in lupus CD4+ T cells Immunometabolism 2020 10.20900/immunometab20200013 32395334
66. Zhao M Downregulation of BDH2 modulates iron homeostasis and promotes DNA demethylation in CD4+ T cells of systemic lupus erythematosus Clin Immunol 2018 187 113 121 10.1016/j.clim.2017.11.002 29113828
67. Keshavarz-Fathi M DNA methylation of CD70 promoter in juvenile systemic lupus erythematosus Fetal Pediatr Pathol 2022 41 1 58 67 10.1080/15513815.2020.1764681 32427516
68. Vordenbäumen S Associations of site-specific CD4+-T-cell hypomethylation within CD40-ligand promotor and enhancer regions with disease activity of women with systemic lupus erythematosus Lupus 2021 30 1 45 51 10.1177/0961203320965690 33081589
69. Guo G Disease activity-associated alteration of mRNA m5 C methylation in CD4+ T cells of systemic lupus erythematosus Front Cell Dev Biol 2020 8 430 10.3389/fcell.2020.00430 32582707
70. Hanaei S The status of FOXP3 gene methylation in pediatric systemic lupus erythematosus Allergol Immunopathol 2020 48 4 332 338 10.1016/j.aller.2020.03.014
71. Xiong H Mesenchymal stem cells activate the MEK/ERK signaling pathway and enhance DNA methylation via DNMT1 in PBMC from systemic lupus erythematosus BioMed Res Int 2020 2020 1
72. NosratZehi S Promoter methylation and expression status of cytotoxic T-lymphocyte-associated antigen-4 gene in patients with lupus J Epigenetics 2020 2 1 33 39
73. Sawalha AH Epigenetic dysregulation of ACE2 and interferon-regulated genes might suggest increased COVID-19 susceptibility and severity in lupus patients Clin Immunol 2020 215 108410 10.1016/j.clim.2020.108410 32276140
74. Liu Q LncRNA GAS5 suppresses CD4+ T cell activation by upregulating E4BP4 via inhibiting miR-92a-3p in systemic lupus erythematosus Immunol Lett 2020 2021 41 47 10.1016/j.imlet.2020.08.001
75. Guo G Epitranscriptomic N4-acetylcytidine profiling in CD4+ T cells of systemic lupus erythematosus Front Cell Dev Biol 2020 8 842 10.3389/fcell.2020.00842 32984334
76. Tao B Regulation of Toll-like receptor-mediated inflammatory response by microRNA-152-3p-mediated demethylation of MyD88 in systemic lupus erythematosus Inflamm Res 2021 70 3 285 296 10.1007/s00011-020-01433-y 33507312
77. Zhou S Zhang J Luan P Ma Z Dang J Zhu H Huo Z miR-183–5p is a potential molecular marker of systemic lupus erythematosus J Immunol Res 2021 2021 1 11
78. Pyfrom S The dynamic epigenetic regulation of the inactive X chromosome in healthy human B cells is dysregulated in lupus patients Proc Natl Acad Sci 2021 10.1073/pnas.2024624118 34103397
79. Zhen Y Ezh2-mediated epigenetic modification is required for allogeneic T cell-induced lupus disease Arthritis Res Ther 2020 22 1 10 10.1186/s13075-020-02225-9 31898524
80. Shi X Downregulated miR-29a promotes B cell overactivation by upregulating Crk-like protein in systemic lupus erythematosus Mol Med Rep 2020 22 2 841 849 10.3892/mmr.2020.11166 32467986
81. Zhang Q Deficiency of p53 causes the inadequate expression of miR-1246 in B cells of systemic lupus erythematosus J Immunol 2022 209 8 1492 1498 10.4049/jimmunol.2200307 36165173
82. Scharer CD Epigenetic programming underpins B cell dysfunction in human SLE Nat Immunol 2019 20 8 1071 1082 10.1038/s41590-019-0419-9 31263277
83. Fali T DNA methylation modulates HRES1/p28 expression in B cells from patients with Lupus Autoimmunity 2014 47 4 265 271 10.3109/08916934.2013.826207 24117194
84. Fan H Gender differences of B cell signature related to estrogen-induced IFI44L/BAFF in systemic lupus erythematosus Immunol Lett 2017 181 71 78 10.1016/j.imlet.2016.12.002 27923569
85. Gautam P Sharma A Bhatnagar A Global histone modification analysis reveals hypoacetylated H3 and H4 histones in B Cells from systemic lupus erythematosus patients Immunol Lett 2021 240 41 45 10.1016/j.imlet.2021.09.007 34614442
86. Luo M The regulatory effect of ERK1 pathway in the DNA hypomethylation of MRL/lpr Mice Int J Immunol 2021 9 3 41 10.11648/j.iji.20210903.11
87. Ao T Kikuta J Ishii M The effects of vitamin D on immune system and inflammatory diseases Biomolecules 2021 11 11 1624 10.3390/biom11111624 34827621
88. Hayashi K Real-world data on vitamin D supplementation and its impacts in systemic lupus erythematosus: cross-sectional analysis of a lupus registry of nationwide institutions (LUNA) PLoS One 2022 17 6 e0270569 10.1371/journal.pone.0270569 35767524
89. Ehrlich R Silica—a multisystem hazard Int J Epidemiol 2021 50 4 1226 1228 10.1093/ije/dyab020 33570131
90. Boudigaard SH Occupational exposure to respirable crystalline silica and risk of autoimmune rheumatic diseases: a nationwide cohort study Int J Epidemiol 2021 50 4 1213 1226 10.1093/ije/dyaa287 33462590
91. Brown J Silica accelerated systemic autoimmune disease in lupus-prone New Zealand mixed mice Clin Exp Immunol 2003 131 3 415 421 10.1046/j.1365-2249.2003.02094.x 12605693
92. Gilcrease GW Is air pollution affecting the disease activity in patients with systemic lupus erythematosus? State of the art and a systematic literature review Eur J Rheumatol 2020 7 1 31 10.5152/eurjrheum.2019.19141 32023206
93. Kilburn KH Warshaw RH Prevalence of symptoms of systemic lupus erythematosus (SLE) and of fluorescent antinuclear antibodies associated with chronic exposure to trichloroethylene and other chemicals in well water Environ Res 1992 57 1 1 9 10.1016/S0013-9351(05)80014-3 1740091
94. Cooper GS Occupational and environmental exposures and risk of systemic lupus erythematosus: silica, sunlight, solvents Rheumatology 2010 49 11 2172 2180 10.1093/rheumatology/keq214 20675707
95. Perricone C Smoke and autoimmunity: the fire behind the disease Autoimmun Rev 2016 15 4 354 374 10.1016/j.autrev.2016.01.001 26772647
96. Speyer CB Costenbader KH Cigarette smoking and the pathogenesis of systemic lupus erythematosus Expert Rev Clin Immunol 2018 14 6 481 487 10.1080/1744666X.2018.1473035 29724134
97. Hussain MS Tripathi V Smoking under hypoxic conditions: a potent environmental risk factor for inflammatory and autoimmune diseases Mil Med Res 2018 5 1 1 14 29502531
98. Chen J The relationship of polluted air and drinking water sources with the prevalence of systemic lupus erythematosus: a provincial population-based study Sci Rep 2021 11 1 1 11 33414495
99. Wang J Association between alcohol intake and the risk of systemic lupus erythematosus: a systematic review and meta-analysis Lupus 2021 30 5 725 733 10.1177/0961203321991918 33557684
100. Wang H Major infections in newly diagnosed systemic lupus erythematosus: an inception cohort study Lupus Sci Med 2022 9 1 e000725 10.1136/lupus-2022-000725 36041813
101. James JA Systemic lupus erythematosus in adults is associated with previous Epstein-Barr virus exposure Arthritis Rheum 2001 44 5 1122 1126 10.1002/1529-0131(200105)44:5<1122::AID-ANR193>3.0.CO;2-D 11352244
102. Poole BD Epstein-Barr virus and molecular mimicry in systemic lupus erythematosus Autoimmunity 2006 39 1 63 70 10.1080/08916930500484849 16455583
103. Anand P SARS-CoV-2 strategically mimics proteolytic activation of human ENaC Elife 2020 9 e58603 10.7554/eLife.58603 32452762
104. Fu X-L COVID-19 in patients with systemic lupus erythematosus: a systematic review Lupus 2022 31 6 684 696 10.1177/09612033221093502 35382637
105. Schioppo T Clinical and peculiar immunological manifestations of SARS-CoV-2 infection in systemic lupus erythematosus patients Rheumatology 2022 61 5 1928 1935 10.1093/rheumatology/keab611 34352079
106. de Oliveira GL Probiotic applications in autoimmune diseases Probiotics-current knowledge and future prospects 2018 Houston InTech 69 89
107. Luo XM Gut microbiota in human systemic lupus erythematosus and a mouse model of lupus Appl Environ Microbiol 2018 84 4 e02288 e2317 10.1128/AEM.02288-17 29196292
108. Neuman H Koren O The gut microbiota: a possible factor influencing systemic lupus erythematosus Curr Opin Rheumatol 2017 29 4 374 377 10.1097/BOR.0000000000000395 28376066
109. Zhang H Dynamics of gut microbiota in autoimmune lupus Appl Environ Microbiol 2014 80 24 7551 7560 10.1128/AEM.02676-14 25261516
110. Abdelhamid L Luo XM Retinoic acid, leaky gut, and autoimmune diseases Nutrients 2018 10 8 1016 10.3390/nu10081016 30081517
111. Azzouz D Lupus nephritis is linked to disease-activity associated expansions and immunity to a gut commensal Ann Rheum Dis 2019 78 7 947 956 10.1136/annrheumdis-2018-214856 30782585
112. Ogunrinde E A link between plasma microbial translocation, microbiome, and autoantibody development in first-degree relatives of systemic lupus erythematosus patients Arthritis Rheumatol 2019 71 11 1858 1868 10.1002/art.40935 31106972
113. Tursi SA Bacterial amyloid curli acts as a carrier for DNA to elicit an autoimmune response via TLR2 and TLR9 PLoS Pathog 2017 13 4 e1006315 10.1371/journal.ppat.1006315 28410407
114. Mardani F In vivo study: Th1–Th17 reduction in pristane-induced systemic lupus erythematosus mice after treatment with tolerogenic Lactobacillus probiotics J Cell Physiol 2019 234 1 642 649 10.1002/jcp.26819
115. dos Santos FDMM Excess weight and associated risk factors in patients with systemic lupus erythematosus Rheumatol Int 2013 33 3 681 688 10.1007/s00296-012-2402-8 22527136
116. Kono M The impact of obesity and a high-fat diet on clinical and immunological features in systemic lupus erythematosus Nutrients 2021 13 2 504 10.3390/nu13020504 33557015
117. Kono M The impact of obesity and a high-fat diet on clinical and immunological features in systemic lupus erythematosus Nutrients 2021 13 504 10.3390/nu13020504 33557015
118. Yang Y Spirulina lipids alleviate oxidative stress and inflammation in mice fed a high-fat and high-sucrose diet Mar Drugs 2020 18 3 148 10.3390/md18030148 32143330
119. Liu Y Kaplan MJ Cardiovascular disease in systemic lupus erythematosus: an update Curr Opin Rheumatol 2018 30 5 441 448 10.1097/BOR.0000000000000528 29870498
120. Kang J-H Obesity increases the incidence of new-onset lupus nephritis and organ damage during follow-up in patients with systemic lupus erythematosus Lupus 2020 29 6 578 586 10.1177/0961203320913616 32208798
121. La Cava A The influence of diet and obesity on gene expression in SLE Genes 2019 10 5 405 10.3390/genes10050405 31137916
122. Constantin MM Significance and impact of dietary factors on systemic lupus erythematosus pathogenesis Exp Ther Med 2019 17 2 1085 1090 30679978
123. Xiao ZX High salt diet accelerates the progression of murine lupus through dendritic cells via the p38 MAPK and STAT1 signaling pathways Signal Transduct Target Ther 2020 5 1 1 13 32296011
124. Correa-Rodríguez M Dietary sodium, potassium, and sodium to potassium ratio in patients with systemic lupus erythematosus Biol Res Nurs 2022 24 10998004211065491 10.1177/10998004211065491
125. Kobayashi A Dietary supplementation with eicosapentaenoic acid inhibits plasma cell differentiation and attenuates lupus autoimmunity Front Immunol 2021 10.3389/fimmu.2021.650856 35003105
126. Islam MA Immunomodulatory effects of diet and nutrients in systemic lupus erythematosus (SLE): a systematic review Front Immunol 2020 11 1477 10.3389/fimmu.2020.01477 32793202
127. Pocovi-Gerardino G Beneficial effect of Mediterranean diet on disease activity and cardiovascular risk in systemic lupus erythematosus patients: a cross-sectional study Rheumatology 2021 60 1 160 169 10.1093/rheumatology/keaa210 32594173
|
PMC010xxxxxx/PMC10209935.txt |
==== Front
Nexus Netw J
Nexus Netw J
Nexus Network Journal
1590-5896
1522-4600
Springer International Publishing Cham
740
10.1007/s00004-023-00740-1
Conference Report
Bridges Aalto 2022
http://orcid.org/0000-0002-6384-9330
Hammack Richard rhammack@vcu.edu
Richard Hammack
is a professor of mathematics at Virginia Commonwealth University in Richmond, Virginia. He works mainly in the areas of graph theory and combinatorics, but also has a keen interest in mathematical visualization and arts. He is author of a research monograph and several textbooks.
grid.224260.0 0000 0004 0458 8737 Dept. of Mathematics and Applied Mathematics, Virginia Commonwealth University, Richmond, VA 23284-2014 USA
25 5 2023
15
20 5 2023
© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2023, corrected publication 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The annual international Bridges Conference is the preeminent meeting on the connections between mathematics and art, music, architecture, education and culture. Bridges 2022 was held at Aalto University and Helsinki University in Espoo and Helsinki, Finland. This five-day conference included invited and contributed talks, workshops, a juried art exhibition, musical performances, a short film festival, a poetry reading and family day.
Keywords
Bridges conference
Math
Art
National museum of Finland
http://dx.doi.org/10.13039/100000893 Simons Foundation 523748 Hammack Richard
==== Body
pmcThe 25th annual Bridges conference (referred to here as Bridges Aalto) was held August 1-5, 2022 in beautiful Espoo and Helsinki, Finland. Jointly sponsored by Aalto University, the University of Helsinki and the University of the Arts Helsinki, the conference took place at Aalto University, with public lectures at Helsinki University, and a family day at the National Museum of Finland. Aalto University professor Kirsi Peletonen skillfully served as local coordinator for the conference. In addition to tireless attention to the day-to-day logistics, she arranged a reception at the Helsinki City Hall and family day at the national museum, venues that gave the conference a nice sense of local flavor and demonstrated the importance of the conference to the region.
This in-person event followed two consecutive virtual on-line conferences during the years of the Covid-19 pandemic. Despite the continued risks of pandemic travel, Bridges Aalto attracted 220 in-person attendees from 29 countries, plus an additional 20 persons who registered to access the livestreamed talks. (Normally, the bridges conferences attract 300 to 350 participants from about 30 countries.) Bridges Aalto was the second Bridges conference to be held in Finland, the first being Bridges 2016 at the University of Jyväskylä.
The bulk of the conference was held at the Otaniemi campus of Aalto University in Espoo, Finland, just west of Helsinki. Aalto University was formed in 2010, as a merger of the Helsinki School of Economics, the University of Art and Design Helsinki and the Helsinki University of Technology. The university is named in honor of Finnish architect and designer Alvar Aalto (1898-1975) who designed much of the campus. Indeed, Aalto designed the Aalto University Undergraduate Centre (originally the Helsinki University of Technology main building), which was the main conference venue. See Fig. 1.
Fig. 1 A Bridges lecture in the main auditorium of the Aalto University Undergraduate Centre, which was designed by Alvar Aalto and completed in 1964. (Photo by Henry Segerman.)
Fig. 2 The Bridges 2022 mathematical art exhibit. (Photo by Bruce Torrence.)
The conference opened there on the morning of August 1, with two of a total of six plenary talks. The Reza Sarhangi Memorial Lecture was given by Daina Taimina. Her address, What I Learned in 25 Years of Crocheting Hyperbolic Planes was a richly illustrated reflection on one of the main themes of her life’s work, creating mathematical models in the fiber arts. This was followed by writer and artist Rudy Rucker. Rucker, considered the founder of the cyberpunk movement in science fiction, entertained the audience with quirky musings on the interconnections between mathematics, computer science, painting and writing.
The first three afternoons of the conference featured paper presentations and workshops. In all, there were 33 regular papers (five given remotely), 50 short papers (12 given remotely) and six workshops, which provided active hands-on experiences. One highlight was Eve Torrence’s workshop using origami to build colored topological surfaces that illustrate map coloring theorems.
In addition to the 17 regular and short papers presented remotely, all plenary lectures were livestreamed. Electronic copies of each paper are archived on the Bridges website. (See the link at the end of this article.)
The Exhibition of Mathematical Art also opened on the first day, and remained up for the next four days. This year’s exhibit was curated by Robert Fathauer and Bruce Torrence, with local arrangements by Taneli Luotoniemi and Markus Holste. A primary component of each Bridges meeting, the art exhibit featured works by 91 artists from around the world. See Figs. 2 and 3. A complete catalog can be found at the Bridges website.
Fig. 3 Knots on n×n×n Rubik’s cube, by David Plaxco. An artwork in the Bridges art exhibition, his piece is a collection of 15 Rubik’s cubes with non-standard solutions, showing all knots through 7 crossings. (Photo by Bruce Torrence.)
The first evening concluded with a lavish and lively reception at the beautiful and historic Helsinki City Hall, a quick metro ride from the Otaniemi campus.
The second day opened with back-to-back plenary talks by wife-and-husband origami artists Miri Golan and Paul Jackson. Golan discussed the Origamertria Project, which incorporates paper folding into mathematics lessons. Each lesson uses a series of folds to explain concepts while leading to a final origami creation for the children to enjoy. The digital platform has been carefully developed to be easy for teachers to use in lessons and to provide excellent pedagogy. Jackson, a world-renowned paper engineer and origami artist, explained the development of his mirror symmetry collages and summarized the artistic decisions he made over many years to develop this series of works.
After an afternoon of contributed paper sessions and workshops, the day concluded with a reception for Bridges artists.
The morning of the third day saw two more plenary talks. Jessica Wynne discussed the making of her most recent book, Do Not Erase, a collection of photographs of mathematicians’ chalkboards. Jessica traveled the world to photograph well-known mathematicians and capture the great variety and beauty of their work on chalkboards. She entertained the audience with anecdotes from her many interviews and photo shoots. Next, Andrew Witt’s lecture Formulations: Practices of Architectural Mathematics explored various aspects of the role of mathematics in architectural design, with fascinating case studies from his design office and recent book. Witt suggests that a fusion of design and scientific processes can open new avenues in design. One fascinating example of this is his use of mathematics and computing to analyze and categorize reclaimed construction materials, and reassemble them in new designs.
The third evening featured Formal Music Night, a Bridges tradition that typically features performances by local musicians. Aalto University graduates Petteri Mäkiniemi and Tuomas Ahvas’ performance, Electronic Soundscapes, was haunting and minimalist, using layers of sound to create a mellow sonic tapestry.
All events on the fourth day were open to the public, beginning with the morning’s plenary addresses at Helsinki University. The first scheduled talk was The Art of Inverse Problems, by Helsinki University professor Samuli Siltanen. Although he was unable to attend his presentation because of illness, the audience was nonetheless transfixed by several of his highly entertaining and informative YouTube videos on inverse problems in areas such as tomography, image processing and ancient navigational practices. This was followed by an amusing and engaging performance by stand-up comedian/mathematician Matt Parker. Parker is the best-selling author of Humble Pi: When Math Goes Wrong in the Real World and Things to Make and Do in the Fourth Dimension.
The afternoon was Family Day, a Bridges tradition that is open to the general public and features hands-on workshops for all ages, a short film festival and a poetry reading. This year’s family day was held at Helsinki’s National Museum of Finland, which offered free admission to all Bridges participants. Completed in 1910, the museum building itself is a stunning edifice designed to reflect Finland’s medieval churches and castles. Conference goers divided their time between Bridges events and museum exhibits.
The poetry reading, coordinated by Sara Glaz, included readings or pre-recorded performances of poetry with links to mathematics by some 19 poets.
The short film festival was organized and curated by Bianca Violet, and showcased 11 films selected on the basis of such factors as mathematical content, aesthetic appeal and craftsmanship.
The 2023 Bridges Conference will be held at Dalhousie University in Halifax, Nova Scotia. More details can be found at the Bridges website, www.bridgesmathart.org. The website also has links to all papers, films and artworks from the 2022 Aalto conference.
The original online version of this article was revised to correct the name for the coordinator of the Bridges art exhibit and add the names of the actual organizers.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
|
PMC010xxxxxx/PMC10213295.txt |
==== Front
J Formos Med Assoc
J Formos Med Assoc
Journal of the Formosan Medical Association
0929-6646
0929-6646
Formosan Medical Association. Published by Elsevier Taiwan LLC.
S0929-6646(23)00196-1
10.1016/j.jfma.2023.05.027
Article
Epidemiological characteristics of the three waves of COVID-19 epidemic in Taiwan during April 2022 to March 2023
Chen Yi-Hsuan
Cheuh Yu-Neng
Chen Chiu-Mei
Kuo Hung-Wei ∗
Epidemic Intelligence Center, Taiwan Centers for Disease Control, Taipei, Taiwan
∗ Corresponding author.
26 5 2023
26 5 2023
2 5 2023
13 5 2023
23 5 2023
© 2023 Formosan Medical Association. Published by Elsevier Taiwan LLC.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Background
Taiwan experienced a relatively low incidence of COVID-19 before 2022. However, from April 2022 to March 2023, the country was struck by a nationwide outbreak that occurred in three waves. Despite the considerable magnitude of the epidemic, the epidemiological characteristics of this outbreak have yet to be clearly understood.
Methods
This was a nationwide, population-based, retrospective cohort study. We recruited patients who had been confirmed as domestically-acquired COVID-19 patients from April 17, 2022, to March 19, 2023. The three epidemic waves were analyzed in terms of numbers of cases, cumulative incidence, numbers of COVID-19-related deaths, mortality, gender, age, residence, SARS-CoV-2 variant sub-lineages, and reinfection status.
Results
The numbers of COVID-19 patients (cumulative incidence per million population) were 4,819,625 (207,165.3) in the first wave, 3,587,558 (154,206.5) in the second wave, and 1,746,698 (75,079.5) in the third wave, showing a progressive decline. The numbers of COVID-19-related deaths and mortalities also decreased throughout the three waves. The coverage of vaccination was observed to increase over time.
Conclusion
During the three waves of COVID-19 epidemic, the numbers of cases and deaths gradually declined, while the vaccine coverage increased. It may be appropriate to consider easing restrictions and returning to normality. However, continued monitoring of the epidemiological situation and tracking the emergence of new variants are crucial to prevent the possibility of another epidemic.
Keywords
COVID-19
Epidemiology
Population-based
Taiwan
Abbreviations
COVID-19 Severe Pneumonia with Novel Pathogens
SARS-CoV-2 Severe Acute Respiratory Syndrome Coronavirus 2
Taiwan CDC Taiwan Centers for Disease Control
CECC Central Epidemic Command Center
NIDRS National Infectious Disease Reporting System
==== Body
pmcIntroduction
Since the emergence of Coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the global pandemic has resulted in more than 760 million reported cases and 6.9 million deaths.1
At the end of 2019, Taiwan was alerted to an atypical pneumonia outbreak with unknown pathogen in Wuhan, China, and responded promptly. On January 15, 2020, the government declared “Severe Pneumonia with Novel Pathogens” (COVID-19) as a “Category Five” notifiable communicable disease.2 All patients confirmed with COVID-19 infection must be reported to Taiwan Centers for Disease Control (Taiwan CDC) within 24 hours. On January 20, Taiwan activated the Central Epidemic Command Center (CECC) to coordinate the national pandemic response efforts.3
During January 2020 to April 2021, through a combination of strict border quarantine, test and isolation, as well as non-pharmaceutical interventions, there were only 90 confirmed cases of domestically-acquired COVID-19 out of 208,351 individuals got tested.4 , 5 However, beginning in May 2021, it was the first community outbreak occurred in the Taipei region, caused by the Alpha variant, the daily number of cases increased rapidly, exceeding 100 within a week.6 To strengthen control measures for this outbreak, the CECC raised the COVID-19 warning to Level 3 nationwide on May 19.7 With the aid of public compliance, active case finding, and contact tracing, the outbreak was successfully contained within three months without lockdowns.6 , 8 The warning level was subsequently downgraded to Level 2 on July 26, 2022.9 Thereafter, although sporadic clusters or events were linked to the Delta variant, they did not lead to large-scale outbreaks.10
In late 2021, a newly emerged variant of SARS-CoV-2, known as Omicron, was disseminated globally and was found to possess increased transmissibility.11 It was introduced into Taiwan and led to a rapidly and extensively outbreak throughout the country in April 2022. The epidemic progressed through three waves, each dominated by different sub-lineages of the Omicron variant. Till early 2023, the epidemic was gradually decline and reached a state of stability.
Prior to 2022, Taiwan had experienced a relatively limited incidence of COVID-19. However, from April 2022 to March 2023, the country was struck by a nationwide outbreak that occurred in three distinct waves. Despite the scale and scope of this epidemic, the underlying epidemiological characteristics remained unclear. Gaining a better understanding of the patterns and trajectory of this outbreak could be beneficial in preparing for future epidemics.
Methods
Study design and case definition
This was a nationwide, population-based, retrospective cohort study. A COVID-19 patient was defined as a person with laboratory confirmation of SARS-CoV-2 infection by reverse transcription polymerase chain reaction (RT-PCR) or with positive rapid antigen test confirmed by a medical doctor (implemented after May 26, 2022).12 We recruited COVID-19 patients confirmed from April 17, 2022 to March 19, 2023, and excluded imported cases. In this study, death patients referred specifically to patients whose cause of death was attributed to COVID-19. COVID-19-related death was determined when the death certificate listed COVID-19 as an immediate cause or underlying cause of death. Additionally, any deaths in which COVID-19 was reported as a contributing factor on the death certificate were further reviewed by an expert committee.13 , 14 All patients were followed up till April 14, 2023.
Three waves of epidemic
To mitigate the weekend effect, we aggregated the data into weekly intervals to present the epidemic trends. The initiation of each epidemic wave was determined by the week with the number of new cases exceeded 10,000 patients or the subsequent week when the number of new cases reached its lowest point. We divided the domestic epidemic into three waves: the first wave was from April 17, 2022 to August 12, 2022 (Week 16-Week 32); the second wave was from August 13, 2022 to December 10, 2022 (Week 33-Week 49); the third wave was from December 11, 2022 to March 19, 2023 (2022 Week 50–2023 Week 11). After March 20, 2023, CECC declared the notification criterium for COVID-19 was adjusted for only patients with associated complications.15
COVID-19 reinfection
The definition of COVID-19 reinfection was according to CECC announcement.16, 17 On July 1, 2022, the CECC declared a definition of reinfection as a COIVD-19 patient experienced another positive RT-PCR/rapid antigen test along with worsen symptoms within 3 months of their initial infection and confirmed by a medical profession; or a patient with another positive RT-PCR/rapid antigen test more than three months.16 On October 3, 2022, it was revised that the time interval of two positive results was shorten to 14 days, and no need to evaluate the symptoms by a medical profession.17 Proportion of reinfection was calculated as the number of reinfection patients divided by the number of confirm patients per week. Notably, the statistical count of COVID-19 cases was based on the number of infections. If one patient was confirmed infection for twice or more, he/she would be counted as two or more cases in the database.
Data sources
All patients confirmed with COVID-19 infection must be reported to Taiwan CDC via the National Infectious Disease Reporting System (NIDRS).18 We retrieved the epidemiological data of COVID-19 confirmed cases from NIDRS database, including: notification report number, identity number (ID), birthday, age, gender, residence, confirm date, and date of death. Residence was categorized into six regions, and the corresponding counties were shown in Supplementary Figure 1.19
Among confirmed cases, clinical specimens from patients who got positive results for RT-PCR testing were selected and sent to the National Reference Laboratory of Taiwan CDC for virus genome sequencing analysis to determine the lineage of SARS-CoV-2 variant strain.20 The selection priority was given to certain categories of patients, such as imported cases, cluster cases, breakthrough infections, reinfections, and patients with unknown sources of infection.20 We accessed the resulting database to explore the trends of variant strains.
The COVID-19 vaccination database was established by the National Immunization Information System (NIIS). All vaccination against COVID-19 of individuals was recorded in the database. To determine the vaccination coverage over the entire population, the number of individuals who received the vaccine, regardless of their citizenship status or infection status, was divided by the population size of Taiwanese citizens. However, the population count did not include non-citizens, meaning that the coverage percentage could potentially exceed 100%. We retrieved the weekly reports of COVID-19 vaccination statistics published by Taiwan CDC.21
Statistical analyses
The data analysis was mainly descriptive. Continuous variables were presented as medians and interquartile ranges (IQRs). Categorical variables were reported as counts, percentages and incidences. Incidence was calculated as the number of patients per week divided by the population size at the end of 2022, per million population.22 Cumulative incidence was calculated as the number of patients during a specific period of time (wave) divided by the population size. Data cleaning, tabulation and figure plotting were conducted using R (version 4.2.2). The geographical illustration of incidence by county was analyzed using QGIS (version 3.10.12) software.
Ethical approval
This study was approved by the Institutional Review Board of Taiwan CDC (IRB No.112103), with a waiver of informed consent.
Results
There were 10,202,525 patients confirmed with COVID-19 infection during April 17, 2022 to March 19, 2023 in Taiwan. After excluding for 48,644 imported cases, we recruited 10,153,881 domestically-acquired COVID-19 patients in this study.
Throughout the three waves of COVID-19 epidemic, there was 43.6% of the population confirmed to have been infected with COVID-19 (Table 1 ). In the first wave, there were 4,819,625 patients, corresponding to a cumulative incidence rate of 207,165.3 cases per million population during a time interval of 118 days. This wave reached its peak after 6 weeks, with an average of 309,006 cases per week during the escalation period. The second wave contained 3,587,558 patients with the cumulative incidence of 154,206.5 cases per million population during a time interval of 120 days. It took 8 weeks to reach its peak, with average 246,362 cases per week. The third wave included 1,746,698 patients with the cumulative incidence of 75,079.5 cases per million population during a time interval of 98 days. It took 4 weeks to reach its peak, with average 145,164 cases per week (Table 1 and Fig. 1 A). It was observed that the total number of cases, cumulative incidence rates, and average weekly cases during the growth periods all exhibited a gradual decline across the three waves. Notably, within the third wave, an additional peak was observed following the Chinese New Year holiday at the end of January 2023 (Fig. 1A). This spike was likely due to factors such as increased family gatherings or temporary closures of outpatient services during the holiday. By the end of the follow-up, there were 18,295 patients with COVID-19-related death and the mortality was 786.4 per million population. Across the three waves of epidemic, the numbers of COVID-19-related deaths (mortality per million population) were 8,978 (385.9), 5,235 (225.0), and 4,082 (175.5), respectively. It was observed that the numbers of deaths and mortalities during the three waves had declined (Fig. 1B).Table 1 Epidemiological characteristics of the three waves of COVID-19 epidemic.
Table 1 First waveApril 17, 2022–Aug 12, 2022 Second waveAug 13, 2022–Dec 10, 2022 Third waveDec 11, 2022–Mar 19, 2023 TotalApril 17, 2022–Mar 19, 2023
Cases % Cumulative incidence (per million population) Cases % Cumulative incidence (per million population) Cases % Cumulative incidence (per million population) Cases % Cumulative incidence (per million population)
N 4,819,625 207,165.3 3,587,558 154,206.5 1,746,698 75,079.5 10,153,881 436,451.2
Gender
Male 2,280,934 47.3 198,357.0 1,649,403 46.0 143,437.1 793,215 45.4 68,980.4 4,723,552 46.5 410,774.5
Female 2,537,794 52.7 215,697.9 1,937,550 54.0 164,680.6 952,970 54.6 80,997.0 5,428,314 53.5 461,375.4
Unknown 897 0.02 NA 605 0.02 NA 513 0.03 NA 2015 0.02 NA
Age group (years)
Median (IQR) 37 (29) 37 (31) 38 (29) 37 (29)
0–4 236,813 4.9 291,737.6 132,200 3.7 162,861.4 64,681 3.7 79,682.6 433,694 4.3 534,281.6
5–11 404,590 8.4 278,552.3 298,012 8.3 205,175.4 117,542 6.7 80,925.4 820,144 8.1 564,653.1
12–17 237,479 4.9 203,327.0 283,110 7.9 242,395.8 103,240 5.9 88,393.0 623,829 6.1 534,115.7
18–49 2,581,831 53.6 246,390.9 1,877,136 52.3 179,140.0 938,933 53.8 89,604.8 5,397,900 53.2 515,135.7
50–64 818,330 17.0 155,337.6 587,553 16.4 111,530.9 296,797 17.0 56,338.8 1,702,680 16.8 323,207.2
65–74 344,270 7.1 132,134.9 260,618 7.3 100,028.2 138,432 7.9 53,131.8 743,320 7.3 285,294.9
75 + 196,312 4.1 132,612.0 148,929 4.2 100,604.0 87,073 5.0 58,819.2 432,314 4.3 292,035.2
Residence
Taipei region 1,697,505 35.2 228,095.0 1,231,568 34.3 165,486.7 585,586 33.5 78,685.6 3,514,659 34.6 472,267.2
Northern region 861,819 17.9 223,874.0 658,905 18.4 171,163.2 344,008 19.7 89,362.7 1,864,732 18.4 484,399.8
Central region 878,136 18.2 193,452.2 698,424 19.5 153,861.8 321,137 18.4 70,746.0 1,897,697 18.7 418,060.0
Southern region 545,768 11.3 166,994.9 432,689 12.1 132,394.8 216,293 12.4 66,181.7 1,194,750 11.8 365,571.4
Kaohsiung-Pingtung region 715,158 14.8 196,793.0 488,380 13.6 134,389.5 250,588 14.3 68,955.3 1,454,126 14.3 400,137.8
Eastern region 121,239 2.5 228,131.7 77,592 2.2 146,002.5 29,086 1.7 54,730.2 227,917 2.2 428,864.4
Reinfection
Infected two times 2,879 0.1 123.8 82,704 2.3 3,554.9 158,762 9.1 6,824.2 244,345 2.4 10,502.8
Infected more than two times 2 0.00 0.1 469 0.01 20.2 3,018 0.2 129.7 3,489 0.03 150.0
Figure 1 (A) Number of COVID-19 patients by confirm date; (B) Number of COVID-19-related deaths by death date.
Fig.ure 1
In terms of geographic distribution, Taipei and the Northern regions had the highest cumulative incidence rates across three waves. Although the cumulative incidence in the Eastern region was relatively high during the first wave, it subsequently decreased to levels below the nationwide average in the following periods. Each region or county was observed to experience a decline in cumulative incidence across the three waves (Table 1 and Fig. 2 ).Figure 2 COVID-19 cumulative incidence rate of county during the three waves of COVID-19 epidemic.
Figure 2
Of the three waves examined, female patients outnumbered male patients (Table 1). However, the incidence rate of COVID-19 varied across age groups and fluctuated across waves. In the first wave, the highest incidence was observed among toddlers and children (aged 0–4 and 5–11). During the second wave, it shifted to children and teenagers (aged 5–11 and 12–17), while in the third wave, it was observed among teenagers and young adults (aged 12–17 and 18–49) (Table 1 and Fig. 3 A). Notably, while the incidence in elderly patients was consistently lower than other populations, they are a high-risk group for developing sever symptoms or experiencing mortality.23 , 24 It is crucial for the elderly population to receive full vaccination and adhere to recommended precautions.Figure 3 (A) COVID-19 incidence rate stratified by age group; (B) Proportion trend of SARS-CoV-2 variants; (C) Number and proportion trend of COVID-19 reinfection patients. (A) Due to the similar incidence rates among aged 65 to 75 and aged above 75 groups (Table 1), these two groups were combined into a single category labeled “65+” in Fig. 3A.
Figure 3
During the April 2022 to March 2023 epidemic, domestic cases that were sequenced and analyzed were all found to be of the Omicron variant and its sub-lineages. In the first wave, the Omicron BA.2 dominated, accounting for 93.9% of cases. In the second wave, the Omicron BA.5 was predominant, representing 74.4% of cases. In the third wave, the Omicron BA.2.75 accounted for the majority of cases at 52.7%, followed by BA.5 at 29.4% and BQ.1 at 11.3%, and other variants such as BF.7 and XBB each accounted for less than 3% (Fig. 3B).
The definition of reinfection was announced on July 1, 2022 (Week 26).16 Due to the limited scope of the epidemic prior to 2022, the number of patients experiencing reinfection in the first wave was 2,881 (0.1%). In the second wave, there was an increased number of reinfections, which amounted to 83,173 (2.3%). Subsequently, the cases surged to 161,780 (9.3%) during the third wave. Among all of the reinfection patients, 1.4% of them contracted the virus more than twice (Table 1 and Fig. 3C).
The vaccination coverage was a nationwide statistic that represented the proportion of individuals who have received the COVID-19 vaccine, regardless of whether they were confirmed COVID-19 cases. Given the diversity of vaccination regimens, brands, target populations, and immunization schedules, we demonstrated a general timeline of vaccination campaign for each group and collected vaccination coverage data at the onset of each wave. For toddlers (aged 6 months to 4 years), the vaccine campaign was initiated on July 21, 2022, (Week 29) following a recommendation that primary series with an interval of at least 4–8 weeks by the Advisory Committee on Immunization Practices (ACIP) at the Ministry of Health and Welfare, Taipei, Taiwan.25 Coverage of the first dose among them increased during the second wave and that of second dose raised in the third wave. For children (aged 5–11), the vaccine campaign started on May 25, 2022 (Week 21) following a recommendation that two-dose primary series and boosters with an interval of 5 months.25 For teenagers (aged 12–17) and individuals above aged 18, the campaign for booster dose began on May 25, 2022 (Week 21) and January 7, 2022 (Week 1), respectively. It was observed that vaccination coverage was increasing in each group across three waves of the epidemic, with the highest coverage for each vaccine dose exhibiting among individuals aged 18–49 years. Among the elderly (aged 65+ and aged 75+), the booster coverage reached 70% in the third wave (Table 2 ).Table 2 Coverage of vaccination against COVID-19 among population, stratified by age group.
Table 2Age group First wave (Apr 17 - Aug 12, 2022) Vaccination statistics as of Apr 18, 2022 Second wave (Aug 13 - Dec 10, 2022)Vaccination statistics as of Aug 15, 2022 Third wave (Dec 11, 2022–March 19, 2023)Vaccination statistics as of Dec 5, 2022a
First dose Second dose Booster First dose Second dose Booster First dose Second dose Booster
6m-4y – – – 17.2% – – 45.2% 20.5% –
5-11 – – – 76.8% 46.6% – 85.1% 63.3% 0.4%
12-17 88.9% 80.8% 0.2% 94.7% 84.4% 58.0% 95.0% 87.9% 64.4%
18-49 100.3% 94.2% 63.5% 102.8% 98.9% 84.0% 102.9% 99.3% 86.7%
50-64 88.5% 84.3% 65.5% 91.4% 88.3% 78.4% 91.6% 88.7% 80.6%
65+ 84.5% 79.9% 66.4% 87.5% 84.0% 74.9% 88.4% 85.1% 77.8%
75+ 78.2% 72.1% 57.6% 82.0% 77.1% 66.5% 83.5% 78.8% 70.4%
∗Blank fields indicate that the COVID-19 vaccine was not yet available for that particular group or that there was insufficient time between doses (see the last paragraph of Results for details).
a The vaccination statistics was modified after Dec 5, 2022, with not including non-citizens. Therefore, the data regarding the third wave was obtained on December 5, 2022.
Discussion
This was a nationwide cohort study describing the epidemic characteristics and trends across three waves of COVID-19 epidemic during April 2022 to March 2023 in Taiwan. It was observed that the numbers of COVID-19 cases and deaths were progressively decline, while the vaccination coverage was increasing.
The incidence of COVID-19 varied among different age groups, geographic regions and waves. This variation may be attributed to multiple and complicated factors, including transmissibility of the virus, vaccine coverage, social interactions, occupations, underlying health conditions, living environments, prevention and control policies, and other unknown factors.26, 27, 28, 29 Due to the presence of several unmeasured variables in this observational study, we were unable to establish the association between these factors and incidence rates. Nonetheless, we could observe certain trends and patterns in the epidemic situation. Generally, we found that counties/regions with high population density tend to have a higher cumulative incidence of COVID-19. For toddlers (aged 0–4) and children (aged 5–11) the incidence rates were the highest in the first wave, possibly be due to a lack of vaccine protection at that period. However, as vaccine coverage increased, the incidence decreased in subsequent waves.
The situation of patients obtaining self-testing or reporting their positive results was unclear in this study. Especially in the third wave, as the epidemic was declining and control measures were being relaxed with putting more emphasis on self-health management, patients with mild or asymptomatic cases may not have sought testing as frequently. Nevertheless, the accessibility for patients with severe symptoms to seek medical care and the criteria of evaluating COVID-19-related mortality remained unchanged.
The number and proportion of patients experiencing reinfection had increased during the second wave and especially in the third wave. Cases of patients being infected for more than twice were not rare either. It's noteworthy that most of these reinfection patients were infected SARS-CoV-2 Omicron twice but with different sub-lineages within several months. The risk factors and severity of them require further investigation.
The prevention and control policies implemented to combat against COVID-19 were constantly being modified to respond to the evolving epidemiological situation in Taiwan. As the epidemic gradually decreased, the government sought to strike a balance between preserving public health and promoting economic development, while simultaneously trying to return to normality.30 Regarding the face mask mandate, it remained unchanged from level 3 alert in 2021 to July 2022 (amid the first wave), requiring individuals to wear masks at all times when going out, whether indoors or outdoors.31 As the epidemic decreased, the regulations were gradually relaxed, beginning with outdoor activities (December 2022, amid the second wave) and later extending to indoor activities as well as in schools (February and March 2023, amid the third wave).32, 33, 34 As for isolation and quarantine measures, the duration of the period was progressively shortened, with a greater emphasis on self-health management.35 Prior to the COVID-19 outbreak in 2022, confirmed patients were required to be isolated and received medical care in hospitals.36 However, as the number of patients surged in May 2022, the policy was revised to allow adults with no or mild symptoms to self-isolate at home for 7 days, followed by another 7 days for self-health management (amidst the first wave).37 Subsequently, in November 2022, the isolation period was reduced to 5 days, with the length of the self-health management period based on the results of at-home rapid tests (amid the second wave).38 After March 20, 2023, COVID-19 was adjusted to be notifiable only for patients with associated complications, patients with no or mild symptom were no need to report or be isolated.15 On the other hand, the provision of antiretroviral therapies and vaccination campaigns against COVID-19, both were government-funded, continued and even expanded. Confirmed COVID-19 patients with risk factors for developing severe conditions were eligible for free antiretroviral treatments.39 In March 2023, the CECC launched “2023 National COVID-19 Vaccination Campaign” with its main theme being “Vaccine Plus One. Easing Restriction Safely” to enhance individual protection against COVID-19.40
There were some limitations in this study: First, the number and proportion of people who had positive results on at-home rapid antigen tests but not seeking medical confirmation were hard to be estimated. It needs further studies applying mathematical modeling analysis or seroprevalence surveys to explore this issue. Second, the sample of patients who had positive PCR specimens was not selected randomly for genome sequencing analysis. Instead, priority was given to certain categories of patients, such as imported cases, cluster cases, breakthrough infections, reinfections, and patients with unknown sources of infection.20 Therefore, the representativeness of the entire population was uncertain. However, this approach allowed us to monitor the transmission of new variants within the community. Third, for this study, we did not have the accessibility to obtain data on the hospitalization rates and vaccination statuses of individual patients. As a result, we focused our analysis on describing the epidemiological characteristics of COVID-19 across three waves of the epidemic. While this constraint may have limited the scope of our findings, we considered that our descriptive analysis provided valuable insights into the overall patterns and trends of the disease. Nonetheless, future research with more comprehensive data could help to provide a more detailed understanding of the impact of COVID-19, including the effectiveness of vaccination efforts and the severity of the disease.
Throughout the three waves of the COVID-19 epidemic in Taiwan, there was a progressive decrease in numbers of patients and deaths, coinciding with an increase in the number of people receiving vaccinations. It may be appropriate to consider easing restrictions and measures as well as returning to normality. However, it is crucial to continue monitoring the epidemiological situation and tracking the emergence of new variants to prevent the possibility of another epidemic.
Funding
This study was supported by Taiwan Centers for Disease Control (Taipei, Taiwan).
Declaration of competing interest
The authors have no conflicts of interest relevant to this article
Appendix A Supplementary data
The following is the Supplementary data to this article.Multimedia component 1
Multimedia component 1
Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.jfma.2023.05.027.
==== Refs
References
1 World Health Organization Coronavirus (COVID-19) dashboard Available from: https://covid19.who.int/
2 Ministry of Health and Welfare, Taiwan (R.O.C.) Formally classify “severe pneumonia with Novel pathogens” as a category 5 notifiable disease 2020 Available from: https://covid19.mohw.gov.tw/en/cp-4868-53709-206.html
3 Centers for Disease Control Ministry of health and Welfare, Taiwan (R.O.C). Taiwan CDC announces activation of central epidemic Command center (CECC) for severe special infectious pneumonia 2020 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/vmv22PiH7-k3K-yh6FkmKw?typeid=158
4 Central Epidemic Command Center, Taiwan (R.O.C.) CECC confirms 1 more indigenous case; family member of Case #1102 found to have COVID-19 2021 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/e7azey1e9OJj5XnFupQ1iA?typeid=158
5 Chen Y.H. Fang C.T. Combined interventions to suppress R0 and border quarantine to contain COVID-19 in Taiwan J Formos Med Assoc 120 2 2021 903 905 32855037
6 Akhmetzhanov A. Cheng H.-Y. Linton N. Ponce L. Jian S.-W. Lin H.-H. Transmission dynamics and effectiveness of control measures during COVID-19 surge, Taiwan, april–august 2021 Emerg Infect Dis J 28 10 2022 2051
7 Central Epidemic Command Center, Taiwan (R.O.C.) CECC raises epidemic warning to Level 3 nationwide from May 19 to May 28; strengthened measures and restrictions introduced across Taiwan to reduce community transmission 2021 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/VN_6yeoBTKhRKoSy2d0hJQ?typeid=158
8 Liu C.J. Yang C.Y. Chen Y.L. Wang S.S. Chu C.M. Hsieh M.H. Prompt successful response to a COVID-19 outbreak: performance of community-based rapid screening station J Formos Med Assoc 121 11 2022 2356 2359 35115199
9 Central Epidemic Command Center, Taiwan (R.O.C.) CECC to lower epidemic alert level to Level 2 from July 27 to August 9 2022 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/C-9A7siqyLWsca5Qb6npkA?typeid=158
10 Centers for Disease Control Ministry of health and Welfare, Taiwan (R.O.C). Taiwan national infectious disease statistics system Available from: https://nidss.cdc.gov.tw/nndss/disease?id=19CoV
11 Yuan S. Ye Z.W. Liang R. Tang K. Zhang A.J. Lu G. Pathogenicity, transmissibility, and fitness of SARS-CoV-2 Omicron in Syrian hamsters Science 377 6604 2022 428 433 35737809
12 Central Epidemic Command Center, Taiwan (R.O.C.) CECC revises case definition of COVID-19; people considered as confirmed cases if they test positive with rapid tests and have their test results confirmed by medical personnel 2022 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/4FN9QHzlxKuOtnGTHVWjfA?typeid=158
13 Cheng H.Y. Chen C.M. Cheuh Y.N. Liu Y.L. Jian S.W. Hsu C.B. Interim epidemiological analysis of the SARS-CoV-2 Omicron wave in Taiwan, january–june 2022 Taiwan Epidemiol Bullet 38 24 2022 366 371
14 Taiwan standard certificate of death (English version) Available from: https://dep.mohw.gov.tw/DOS/cp-5070-61793-113.html
15 Central Epidemic Command Center, Taiwan (R.O.C.) Effective March 20, mild COVID-19 cases exempt from reporting, isolation and should seek medical care immediately if they experience severe symptoms/warning signs 2023 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/jzLbo6nkOffFtRAOm9IfUg?typeid=158
16 Central Epidemic Command Center, Taiwan (R.O.C.) CECC draws up definition and principles for handling reinfection in confirmed cases 2022 Available from: https://covid19.mohw.gov.tw/en/cp-4868-71129-206.html
17 Central Epidemic Command Center, Taiwan (R.O.C.) CECC draws revises definition and principles for handling reinfection in confirmed cases 2022 Available from: https://covid19.mohw.gov.tw/en/cp-4868-72299-206.html
18 Centers for Disease Control Ministry of health and welfare, Taiwan (R.O.C). National infectious disease reporting system (NIDRS) Available from: https://www.cdc.gov.tw/En/Category/MPage/g2B1VbUeuu_gEgdYmNiiKw
19 Centers for Disease Control Ministry of health and welfare, Taiwan (R.O.C). Organizations and responsibilities Available from: https://www.cdc.gov.tw/CdcOrganization/Index/cBX61rWwT5TKpS7BbMzKag
20 Yang J.R. Kuo C.Y. Lin Y.T. Lin Y.C. Chen H.J. Huang H.Y. Laboratory surveillance of COVID-19 variants in Taiwan Taiwan Epidemiol Bullet 38 15 2022 174 184
21 Central Epidemic Command Center, Taiwan (R.O.C) COVID-19 vaccine statistics Available from: https://www.cdc.gov.tw/Category/Page/9jFXNbCe-sFK9EImRRi2Og
22 Department of Household Registration Ministry of the interior, Taiwan (R.O.C.). Demographic statistics 2022 Available from: https://www.ris.gov.tw/app/portal/346
23 Karla Romero S. David R. Gabriela P.-H. Stefanie S. Albert N. Andreas S. The isolated effect of age on the risk of COVID-19 severe outcomes: a systematic review with meta-analysis BMJ Glob Health 6 12 2021 e006434
24 Beaney T. Neves A.L. Alboksmaty A. Ashrafian H. Flott K. Fowler A. Trends and associated factors for Covid-19 hospitalisation and fatality risk in 2.3 million adults in England Nat Commun 13 1 2022 2356 35487905
25 Central Epidemic Command Center, Taiwan (R.O.C.) ACIP's COVID-19 vaccine recommendations for children aged 6 months to 5 years 2022 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/2S3JYo5E-LScO5SWnD1VOg?typeid=158
26 UK Health Security Agency COVID-19: epidemiology, virology and clinical features 2022 Available from: https://www.gov.uk/government/publications/wuhan-novel-coronavirus-background-information/wuhan-novel-coronavirus-epidemiology-virology-and-clinical-features
27 McIntosh Kenneth COVID-19: epidemiology, virology, and prevention (UpToDate) 2023 Available from: https://www.uptodate.com/contents/covid-19-epidemiology-virology-and-prevention
28 Rahmani K. Shavaleh R. Forouhi M. Disfani H.F. Kamandi M. Oskooi R.K. The effectiveness of COVID-19 vaccines in reducing the incidence, hospitalization, and mortality from COVID-19: a systematic review and meta-analysis Front Public Health 10 2022 873596 36091533
29 Gholami M. Fawad I. Shadan S. Rowaiee R. Ghanem H. Hassan Khamis A. COVID-19 and healthcare workers: a systematic review and meta-analysis Int J Infect Dis 104 2021 335 346 33444754
30 Taipei Times Taiwan launches new pandemic response strategy 2022 Available from: https://www.taipeitimes.com/News/front/archives/2022/04/08/2003776214
31 Central Epidemic Command Center, Taiwan (R.O.C.) From April 1-30, current mask mandate and other epidemic prevention measures to remain in place 2022 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/QGqFFcQy-nato7_MrRX39Q?typeid=158]
32 Central Epidemic Command Center, Taiwan (R.O.C.) Taiwan to ease current mask mandate and related epidemic prevention measures starting December 1 2022 [Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/VRVYABkMZ3OLkKDMQk1RFQ?typeid=158
33 Central Epidemic Command Center, Taiwan (R.O.C.) CECC announces plans to relax indoor mask rules; eased rules scheduled to take effect on Feb. 20 2023 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/eTaCQ3IdcCfeAwa0HfDErQ?typeid=158
34 Ministry of Education, Taiwan (R.O.C.) Effect March 6, ease the indoor mask-wearing requirements in all levels of schools 2023 Available from: https://cpd.moe.gov.tw/page_two.php?id=35978
35 Central Epidemic Command Center, Taiwan (R.O.C.) Effective November 7, CECC to adjust control measures for confirmed COVID-19 cases and contacts and cancel some prevention measures in community 2022 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/JKAvijGZ3KGo-KMHWM72sg?typeid=158
36 Central Epidemic Command Center, Taiwan (R.O.C.) Measures and criteria for management and isolation of confirmed cases of COVID-19 2022 Available from: https://www.cdc.gov.tw/Uploads/44805672-7776-455e-897e-9c875ac43fce.pdf
37 Central Epidemic Command Center, Taiwan (R.O.C.) Principles for admission and treatment of mild and severe COVID-19 cases 2022 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/0bypzhk24jp7-_hNkNTZsw?typeid=158
38 Central Epidemic Command Center, Taiwan (R.O.C.) Effective November 14, CECC to adopt “5+n” days of isolation/self-health management policy for non-severe cases receiving home care 2022
39 Central Epidemic Command Center, Taiwan (R.O.C.) CECC announces evaluation process for receiving oral antiviral drugs 2022 Available from: https://www.cdc.gov.tw/En/Bulletin/Detail/_TfxyFQh0tO0-GiOeMyoxA?typeid=158
40 Central Epidemic Command Center, Taiwan (R.O.C) National COVID-19 vaccination campaign" with its main theme being "vaccine Plus one. Easing restriction safely" will be launched on March 6 2023 Available from: https://www.cdc.gov.tw/En/Category/ListContent/tov1jahKUv8RGSbvmzLwFg?uaid=qVtP4HeCQ0z2sHzTbFWhfQ
|
PMC010xxxxxx/PMC10213572.txt |
==== Front
Sustain Water Resour Manag
Sustain Water Resour Manag
Sustainable Water Resources Management
2363-5037
2363-5045
Springer International Publishing Cham
37273915
868
10.1007/s40899-023-00868-5
Original Article
A review of research trends on the usage of photocatalysis for wastewater treatment: bibliometric analysis
Mohammed Abdussamad Mukhtar 12
http://orcid.org/0000-0003-4537-8421
Aziz Farhana farhana@petroleum.utm.my
34
Mohtar Safia Syazana 4
Mhamad Shakhawan Ahmad 15
Ahmadu Bello 6
Nasir Mustapha Usman 2
Muhammad Khuzaifa Yahuza 7
Aziz Madzlan madzlan@utm.my
13
1 grid.410877.d 0000 0001 2296 1505 Faculty of Science, Universiti Teknologi Malaysia, 81310 Johor Bahru, Johor Malaysia
2 grid.442527.2 0000 0000 9365 7327 Department of Chemistry, Yobe State University, Damaturu, Yobe State Nigeria
3 grid.410877.d 0000 0001 2296 1505 Advanced Membrane Technology Research Centre (AMTEC), Universiti Teknologi Malaysia, 81310 Johor Bahru, Johor Malaysia
4 grid.410877.d 0000 0001 2296 1505 Faculty of Chemical and Energy Engineering, Universiti Teknologi Malaysia, 81310 Johor Bahru, Johor Malaysia
5 grid.440843.f Department of Chemistry, Faculty of Education, University of Sulaimani, Sulaimani, Kurdistan Iraq
6 grid.462640.2 0000 0001 2219 5564 Academy Library, Nigerian Defence Academy, Kaduna, Kaduna State Nigeria
7 grid.411092.f 0000 0001 0510 6371 Department of Chemistry, Abubakar Tafawa Balewa University, Bauchi, Bauchi State Nigeria
26 5 2023
2023
9 3 8814 4 2022
16 5 2023
© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Photocatalysis is seen as a viable alternative to treating water pollution, due to its flexibility, low cost, and ability to use visible light which is a plentiful and free energy source. Hence, determining the topics of interest and widening collaboration networks will go a long way in improving research in this field. In this study, we aimed to analyze the global research trends on the usage of photocatalysis for wastewater treatment using bibliometric analysis, centered on the outputs of publications, co-authorships, countries of affiliation, and author’s keyword co-occurrences. Bibliometric analysis is a review method that is well-known and more conversant to Social Science. Employing it in Physical Science, which is rarely seen, will provide an avenue and yet another method of determining common research topics as well as the potential opportunities and future research in the field. A potential hybrid review paper of great importance to future research in the area will be produced. A total of 1373 articles published within 27 years between 1993 and 2020 were extracted from the Scopus database. In the beginning, less attention was given to the said topic, because after the oldest article was published in 1993, there was no record of other publications until after 5 years (1998). However, from 2002 there was a growing interest in research in that field, with a cumulative increase every year to date, except for a few years with fewer publications. Meanwhile, the number of publications has risen significantly from 2017 to 2020, with an increase of more than 70 publications every year; this is expected to increase rapidly in the coming years. Recently researchers are focusing on developing efficient photocatalysts for contaminants of emerging concern, like pharmaceutical and refinery wastewater, however, the usage of conducting polymers to produce nanocomposite which was found to be very effective is still lagged in wastewater treatment, as such it will be a good area of future research on effective photocatalysts for wastewater treatment.
Supplementary Information
The online version contains supplementary material available at 10.1007/s40899-023-00868-5.
Keywords
Bibliometric analysis
Conducting polymers
Photocatalysis
VOSviewer
Wastewater
Malaysia Ministry of Higher Education FRGS/1/2018/STG07/UTM/02/3 Aziz Farhana http://dx.doi.org/10.13039/501100005417 Universiti Teknologi Malaysia Q.J130000.2451.08G35 Aziz Farhana issue-copyright-statement© Springer Nature Switzerland AG 2023
==== Body
pmcIntroduction
The industrial revolution brought about a wide range of problems including water pollution, leading to a significant impact on the environment and living things. Direct discharge of industrial effluents into waterways makes them unfit for consumption, as the water may receive nonbiodegradable and undesirable chemicals from the effluents which proved to be hazardous, and death by pollution-related diseases is increasing day by day (Pawar and Lee 2015; Chowdhary et al. 2020; Bruce and Limin 2021). Large effluents discharged by various industries into waterways are extremely poisonous. This leads to the contamination of surface and groundwater as it contains a high concentration of heavy metals and other harmful organic compounds, these compounds are believed to be carcinogenic, mutagenic, and in some cases teratogenic to living things (Hussain and Wahab 2018; Hitam and Jalil 2020). Coagulation, flocculation, sedimentation, filtration, and disinfection are ancient water remediation methods, but they are slow and ineffective. Apart from being unsustainable with the environment, some of these traditional approaches often necessitate a large amount of space, resulting in chemical waste and, in some cases, failing to eliminate a sufficient amount of hazardous contaminants, resulting in the production of secondary harmful products (Mani and Bharagava 2016; Hitam and Jalil 2020).
Advanced oxidation processes (AOPs) are among the new facile water treatment methods developed, they have gained prominence as a result of their effectiveness and ability to degrade pollutants in water via a redox reaction (Hodges et al. 2018; Khan et al. 2020; Santos et al. 2022). AOPs are effective for a broad spectrum of pollutants because of their robust nonselective oxidability; and the generation of nonhazardous byproducts like CO2, H2O, and other small inorganic ions (Yang et al. 2020). The AOPs are based on the generation and use of hydroxyl radicals (•OH), due to their high reduction potential (2.80 V vs. Normal hydrogen electrode), they can degrade a wide range of organic pollutants, including the most stable ones (Asghar et al. 2015; Li et al. 2021). Ultraviolet (UV) photolysis, hydrogen peroxide photo-Fenton, photo-ozonation, and heterogeneous photocatalysis are the main classes of AOPs. However, the benefit of using sunlight, which is a free source of energy, has increased the popularity of heterogeneous photocatalysis (Ghernaout and Elboughdiri 2020; Zhang et al. 2020; Mishra et al. 2022).
Photocatalysis as one of the techniques of AOPs has emerged as an effective way to speed up the degradation and removal of a wide variety of organic pollutants in contaminated water (Chi et al. 2020; Ma et al. 2022). Apart from its utilization of sunlight which is free and abundant in nature, photocatalysis is considered a low or nonwaste generation method, using highly reactive chemical species with high oxidation capacity to break complicated structures (Tomar et al. 2020; Choudhury 2021). Fujishima and Honda pioneered photocatalysis research in 1972 when they used TiO2 electrodes in water splitting (Fujishima, A., and Honda 1972). Because photocatalysis can fully mineralize contaminants at low temperatures and pressures, it has gained tremendous popularity in the treatment of polluted gaseous and liquid wastes (Hitam and Jalil 2020; Sharma et al. 2020; Akiyama et al. 2022).
Bibliometric analysis is used to analyze and determine past and present research trends in a particular field of knowledge centered on academic repositories’ research outcomes (Md Khudzari et al. 2018; Macías-Quiroga et al. 2020; Anuar et al. 2021; Tan et al. 2021). Scopus is considered the most extensive collection of peer-reviewed articles on a wide range of topics, as such, it is mostly used in bibliometric analysis to cover more topics that may not be available in other databases (Md Khudzari et al. 2018; Garrido-Cardenas et al. 2020; Cascajares et al. 2021). Despite the increasing interest in the field of wastewater treatment using photocatalysis, there are limited studies on the analysis of scientific publications done on the said subject matter. Both Garrido-Cardenas et al. 2020 and Macías-Quiroga et al. 2020 reported the research trends on the use of advanced oxidation processes for wastewater treatment in which photocatalysis was part of the discussion, however, the former presented the trend between the period of 1980 to 2018 while the latter reported between 1990 to 2018 with emphasis on the Ibero-American region.
Meanwhile, Singh and Borthakur 2018 reported a bibliometric and comparative analysis between the biodegradation and photocatalytic degradation of organic pollutants during the period between 1997 and 2017 (20 years), they stress the importance and capability of photocatalysis as one of the most promising techniques for the degradation of a variety of organic liquid and gaseous contaminants. In this review, the bibliometric analysis was performed to determine the emerging trends of scientific publications regarding photocatalytic wastewater treatment globally; this study provides insight into the current research on photocatalysis and common research topics as well as the potential opportunities and future research in the field.
Materials and methods
Description of the study area
Unlike review papers which usually evaluate a topic's most recent trends, problems, and future research directions, the study of bibliometric analysis is a deterministic approach for evaluating global research directions in particular fields based on the literature found in academic databases (Arsad et al. 2022). Bibliometric analysis is a review method that is well-known and more conversant to Social Science. Employing it in Photocatalysis will provide an avenue and yet another method of determining common research topics as well as the future direction in the field. The Scopus database was used to retrieve the data on January 27, 2021. The study’s central theme was research papers that included photocatalysis* “OR photocatalyst*” wastewater treatment in the title and abstract. The first publication was in 1993, and the latest was in 2020. For the quest, the question text was: TITLE-ABS (“photocatalysis*” OR photocatalyst* “wastewater treatment”) AND (EXCLUDE (PUBYEAR, 2021)) AND (LIMIT-TO (LANGUAGE, “English”)) AND (LIMIT-TO (SRCTYPE, “j”)), it gives 1493 documents. Additional phrases were added to the question string to exclude the review articles from our study, resulting in a total of 208 possible review articles being screened. After screening them we found 120 articles that are actual review articles, to exclude those review papers from the next search results, Scopus paper identifiers (EIDs) of the articles were taken and added to the next search query.
The central query theme's search results were analyzed using the following criteria: year, source, author, affiliation, country/region, subject area, and document type. The ranking was based on the bibliometric metrics such as total publications, total citations, CiteScore, and h-index. In addition, subthemes were also added to explore the output of various photocatalytic materials used in wastewater treatment. The major photocatalytic materials included are heterojunction, semiconductors, polymer, composite, and conducting polymer. The search text for each material was searched independently, using a search text containing precise terms depending on the intended material. For example: (TITLE-ABS (“photocatalysis*” OR “photocatalyst*” “wastewater treatment” AND heterojunction). The search results for the subthemes were examined based on the number of publications published per year. The process of data mining is presented in Fig. 1. Likewise, the table containing the search text used in the data extraction from the database can be seen in Table S1 of supplementary materials.Fig. 1 Data mining flowchart for core and subthemes
Bibliometric maps
The downloaded data of 1373 articles were exported to VOSviewer software which is used in creating and visualizing bibliometric maps. The software used was version 1.6.7 originating from the Centre for Science and Technology Studies, Leiden University, Netherlands. Maps created using the software include countries and author keywords. A link is a connection or relationship that exists between any two objects; each link has an integer that reflects its power, the higher the integer the stronger the link. In the analysis of co-authorship, the number of publications co-authored by those countries is expressed in the strength of the links between them, whereas the total link strength indicates the total strength of a country's co-authorship links with other countries. Likewise, in co-occurrence analysis, the number of publications in which two keywords appear together is indicated by the link strength between author keywords.
Analysis of co-authorship
We included 61 countries in the analysis of co-authorship. The affiliated countries were clustered into 6 continents: Africa, Asia, Australia, Europe, North America, and South America.
Analysis of co-occurrence
Analyses of the co-occurrence of author keywords involve 1000 keywords from 1373 articles. Single synonyms and relatable wordings were skimmed, counted, and re-labeled before importing the keywords into the software. The lowest occurrences of a keyword to be assessed were set to 1 in the software. The average publication, year, number of occurrences, and link strength of the keywords were all viewed in the software.
Photocatalysis applications (subthemes of photocatalysis)
The frequency of research outputs between keyword co-occurrences and total publication of the subthemes were analyzed. The subthemes used are heterojunction, semiconductors, polymer, composites, and conducting polymers. The number of publications on these subthemes was analysed, respectively, to find the potential gap in the usage of photocatalysis in wastewater treatment.
Results and discussion
Development in research interest and publication output
Industries produce a large amount of wastewater into waterways, leading to environmental contamination (Sharma et al. 2022). Regular and conventional water treatment methods are not fast, efficient, and cost-effective (Pincheira et al. 2021). Among the new and easy water treatment methods, photocatalysis is seen as an alternative, due to its effectiveness, the capability to degrade contaminants through a redox reaction, and most importantly, the usage of sunlight which is an abundant and free source of energy (Wang et al. 2022). That is why growing research interest and publications on the treatment of wastewater were focused on photocatalysis. From Fig. 2, it can be seen that within 27 years, a total of 1373 research articles have been published on the usage of photocatalysis in wastewater treatment. In the beginning, less attention was given to the said topic, because after the oldest article was published in 1993, there was no record of other publications until after 5 years (1998). However, from 2002, there was a growing interest in research in that field, with a cumulative increase every year to date, except for a few years with fewer publications like 2004, 2005, 2007, 2009, and 2011. It is important to know that there is a massive increase in the number of publications from 2017 to 2020, with an increase of more than 70 publications every year. As a result, the annual publication is forecasted to grow, as researchers are now focusing on visible-light-driven photocatalysis because of the abundance of sunlight in nature. The first known photocatalysts were wide bandgap semiconductors (Bitaraf and Amoozadeh 2021). Despite their popularity and effectiveness, they have significant drawbacks of absorption only at UV region, hence limiting their photocatalytic activity (Yang et al. 2022). Besides, it is already known that solar energy comprises only 3–5% UV, whereas 43% of solar energy comprises visible light, therefore a significant amount of solar radiation is lost (Yang et al. 2022). Owing to that factor growing interest and publications were also focused on the visible-light-driven narrow bandgap semiconductors that absorb at visible region.Fig. 2 Annual and cumulative numbers of research articles on photo
Interestingly, in 2020, despite the effect of the covid-19 pandemic that cripples research activities almost everywhere in the world, 331 research articles were published, and this is another indicator that there is a growing interest in the research in this area. A quick analysis of most of the research papers published within these years revealed that most of the articles are closed access, meaning that an intended researcher or user has to pay before accessing those valuable research outputs. Out of all the research articles published as of 2020, only 17% were published as open access, as such, it is advisable to publish more in open access journals to increase the number of citations, attracts more readerships, and wide visibility of the research ideas in the field of wastewater treatment using photocatalysis.
Productive journals
Table 1 showed the results of the top 20 most productive journals, most of the productive journals are published by Elsevier. Chemical Engineering Journal comes first in the list with 63 total publications and the highest total citations of 2616. It has a relatively high CiteScore (15.2), as such, it is not surprising when it has the highest total publications, because usually, CiteScore influenced the decision of researchers in finding journals that suit their new findings in research (Md Khudzari et al. 2018). However, the only journal published by the Royal Society of Chemistry (RSC Advances) is second among the list with 37 total publications and total citations of 727. Journal of Hazardous Materials published by Elsevier is third in the list of most productive journals and also third in the list of journals with the highest total citations (1943), it has a CiteScore of 13.1 and 35 total publications respectively.Table 1 Top 20 most productive journals on the usage of photocatalysis for wastewater treatment and their most cited articles
S/N Journal TP TC Citescore (2019) The most cited article (reference) Times cited Publisher
1 Chemical Engineering Journal 63 2616 15.2 “Construction of iodine vacancy-rich BiOI/Ag@AgI Z-scheme heterojunction photocatalysts for visible-light-driven tetracycline degradation: Transformation pathways and mechanism insight” (Yang et al. 2018) 222 Elsevier
2 RSC Advances 37 727 6.5 “NiO nanostructures: Synthesis, characterization and photocatalyst application in dye wastewater treatment” (Motahari et al. 2014) 158 Royal Society of Chemistry
3 Journal of Hazardous Materials 35 1943 13.1 “Preparation and photocatalytic property of a novel dumbbell-shaped ZnO microcrystal photocatalyst” (Sun et al. 2009) 190 Elsevier
4 Applied Catalysis B: Environmental 34 2100 25.3 “Construction of high-dispersed Ag/Fe3O4/g-C3N4 photocatalyst by selective photo-deposition and improved photocatalytic activity” (Zhu et al. 2016) 246 Elsevier
5 Environmental Science and Pollution Research 31 242 4.9 “Performance evaluation and application of surface-molecular-imprinted polymer-modified TiO2 nanotubes for the removal of estrogenic chemicals from secondary effluents” (Zhang et al. 2013a) 28 Springer Nature
6 Applied Surface Science 28 781 8.7 “Preparation of photocatalytic Fe2O3-TiO2 coatings in one step by metal organic chemical vapor deposition” (Zhang and Lei 2008) 113 Elsevier
7 Journal of Environmental Chemical Engineering 26 278 6.7 “Magnetically recoverable graphitic carbon nitride and NiFe2O4 based magnetic photocatalyst for degradation of oxytetracycline antibiotic in simulated wastewater under solar light” (Sudhaik et al. 2018) 75 Elsevier
8 Journal of Photochemistry and Photobiology A: Chemistry 26 631 5.2 “Treatment of paper pulp and paper mill wastewater by coagulation − flocculation followed by heterogeneous photocatalysis” (Rodrigues et al. 2008) 183 Elsevier
9 Chemosphere 25 817 8.8 “Solar light induced and TiO2 assisted degradation of textile dye reactive blue 4” (Neppolian et al. 2002) 389 Elsevier
10 Journal of Colloid and Interface Science 24 859 11.0 “Ultralong Cu(OH)2 and CuO nanowire bundles: PEG200-directed crystal growth for enhanced photocatalytic performance” (Li et al. 2010) 88 Elsevier
11 Separation and Purification Technology 24 433 8.3 “Optimisation of an annular photoreactor process for degradation of Congo Red using a newly synthesized titania impregnated kaolinite nano-photocatalyst” (Chong et al. 2009) 78 Elsevier
12 Catalysis Today 22 672 9.5 “Solar photocatalysis as a tertiary treatment to remove emerging pollutants from wastewater treatment plant effluents” (Bernabeu et al. 2011) 132 Elsevier
13 Desalination and Water Treatment 22 129 2.7 “Removal of polycyclic aromatic hydrocarbons from municipal sludge using UV light” (Salihoglu et al. 2012) 27 Desalination Publications
14 Journal of Alloys and Compounds 21 376 7.6 “Uniformly distributed anatase TiO2 nanoparticles on graphene: Synthesis, characterization, and photocatalytic application” (Bai et al. 2014) 60 Elsevier
15 Water Research 21 1386 14.5 “Application of the colloidal stability of TiO2 particles for recovery and reuse in solar photocatalysis” (Fernández-Ibáñez et al. 2003) 195 Elsevier
16 Journal of Cleaner Production 20 415 10.9 “Fabrication of fluorine doped graphene and SmVO4 based dispersed and adsorptive photocatalyst for abatement of phenolic compounds from water and bacterial disinfection” (Shandilya et al. 2018) 79 Elsevier
17 Ceramics International 17 355 6.1 “Morphology controlled hydrothermal synthesis and photocatalytic properties of ZnFe2O4 nanostructures” (Dhiman et al. 2016) 51 Elsevier
18 Journal of Materials Science 16 249 6.2 “An investigation on synthesis and photocatalytic activity of polyaniline sensitized nanocrystalline TiO2 composites” (Min et al. 2007) 99 Springer Nature
19 Water Science and Technology 15 118 2.9 “Dye photo-enhancement of TiO2-photocatalyzed degradation of organic pollutants: The organobromine herbicide bromacil” (Feigelson et al. 2000) 25 IWA Publishing
20 ACS Sustainable Chemistry and Engineering 14 782 9.7 “Self-Assembly Reduced Graphene Oxide Nanosheet Hydrogel Fabrication by Anchorage of Chitosan/Silver and Its Potential Efficient Application toward Dye Degradation for Wastewater Treatments” (Jiao et al. 2015) 111 American Chemical Society
Interestingly, Applied Catalysis B: Environmental also published by Elsevier comes forth in the list of most productive journals with 34 total publications, however, it has the highest CiteScore (25.3) among all the 20 journals and falls second in the list of journals with the highest total citation (2100). This revealed that a journal may have a large number of total publications, but not a large number of total citations because of the issue of accessibility. This can be seen in Water Research Journal published by Elsevier, the journal has only 21 total publications and falls at 15th among the most productive journals, but surprisingly it received 1386 total citations making it 4th among the journal with the highest citations.
ACS Sustainable Chemistry and Engineering journal published by the American Chemical Society with a 9.7 CiteScore is 20th on the list. It has total publications and total citations of 14 and 782 respectively. CiteScore is a metric used in tracking journal performance based on the citation data available in the Scopus database. According to the CiteScore 2019 report, almost all the journals have a CiteScore of more than 5, except for 3 journals; Environmental Science and Pollution Research (4.9), Desalination and Water Treatment (2.7), and Water Science and Technology (2.9).
Leading countries, top institutions, and collaborations
Figure 3 depicts the top 20 most productive countries in the field of photocatalysis for wastewater treatment. Almost 38% of the total publications globally regarding this field are contributed by China, with 516 total publications. India the second-most populous country in the world was also the second-most productive country in the world, having a total publication of 86. Regarding the most productive Institution in China, Jiangsu University can be considered as the most productive with 25 total publications as can be seen in Table 2. However, our result revealed that there are 88 total publications affiliated with the Ministry of Education China, and 39 total publications affiliated with the Chinese Academy of Sciences. Ministry of Education China cannot be regarded as one institution because it comprises so many academic and nonacademic institutions, likewise, since the Chinese Academy of Sciences has 124 branches, comparisons will be hard and prejudiced (Md Khudzari et al. 2018).Fig. 3 Bibliometric map based on co-authorships catalysis for wastewater treatment from 1990 to 2020
Table 2 Topmost productive countries with the institutions having highest publications
Rank Country TP Affiliation
1 China 516 Jiangsu University (25)
2 India 86 Anna University (5)
3 Iran 55 Tarbiat Modares University (9)
4 Spain 33 Universidad Rey Juan Carlos (7)
5 Brazil 33 Universidade Estadual de Maringa (6)
6 Malaysia 28 Universiti Teknologi Malaysia (8)
7 South Korea 25 Inha University, Incheon (4)
8 Taiwan 16 National Taiwan University (3)
9 Italy 15 Università di Salerno (8)
10 United States 14 University of South Florida, Tampa (2)
11 Canada 12 University of Calgary (2)
12 Australia 10 University of Technology Sydney (3)
13 Germany 10 Universität Bayreuth (2)
14 Egypt 9 American University in Cairo (3)
15 Pakistan 7 University of Agriculture, Faisalabad (2)
16 Hong kong 7 Hong Kong University of Science and Technology (4)
17 Portugal 6 Universidade do Porto (3)
18 France 6 CNRS Centre National de la Recherche Scientifique (2)
19 Saudi Arabia 3 King Saud University College of Science (1)
20 United kingdom 2 Loughborough University (1)
In India, however, the most productive institution is Anna University with 5 total publications, the number might appear very small when compared with the total publications in the whole country, which may not be unrelated to the large numbers of academic research institutions in India. Figure 3 shows the relationship between countries. In VOSviewer, the closer countries are situated close to each other the stronger their interaction, and the stronger the link between two countries, the ticker the line connecting them. The highest number of countries per region comes from Asia (25), followed by Europe (17), Africa (8), South America (5), North America (4), and Australia (2).
Meanwhile, co-authorship results showed that China is the most affiliated country linked to 28 countries/territories with 101 times co-authorship. China was followed by Spain (18 links, 34 co-authorships), India (14 links, 32 co-authorships), Malaysia (11 links, 14 co-authorships), the United States (10 links, 43 co-authorships), the UK (10 links, 19 co-authorships), and the rest. China being the most affiliated country may not come as a surprise, because of the rate at which they provide postgraduate scholarships globally through their various academic institutions, and most importantly through the Chinese Academy of Science and Ministry of Education China. This contributes largely to the availability of diverse research partners and the high rate of foreign postgraduates/researchers, which in turn led to a substantial increase in their international collaboration.
Highly cited articles
Citation count is a measure of the impact of academic works, it is generally referring to how many times a journal article, book, or author is cited by other journal articles, books, or researchers. Table 3 presents the 20 most highly cited articles on the usage of photocatalysis for wastewater treatment. The 1st in the list was an article written by Xuming Zhang and coworkers in 2013, even though it was published 8 years ago, it has a total citation (TC) of 785, out of which global citation (GC) was 783, and local citation (LC) of 2. This is not surprising because plasmonic photocatalysis has recently initiated rapid progress in enhancing photocatalytic activity under visible light, leading to the opportunity of using sunlight for energy generation and environmental wastewater treatment. Another factor that may contribute to such high number of citations is its accessibility, it was pointed out earlier that open-access articles usually receive much more citation, because they are always available for free, unlike close access article that has to be paid before accessibility.Table 3 list of the highly cited papers in wastewater treatment using photocatalysis
S/N Title Access Ref Year LC GC TC Journal
1 “Plasmonic photocatalysis” Open (Zhang et al. 2013b) 2013 2 783 785 Reports on progress in physics
2 “Advanced oxidation processes for wastewater treatment: Formation of hydroxyl radical and application” Close (Wang and Xu 2012) 2012 7 629 636 Critical reviews in environmental science and technology
3 “Study of Au/Au3+ − TiO2 photocatalysts toward visible photooxidation for water and wastewater treatment” Close (Li and Li 2001) 2001 2 625 627 Environmental science and technology
4 “Solar light induced and TiO2 assisted degradation of textile dye reactive blue 4” Close (Neppolian et al. 2002) 2002 5 384 389 Chemosphere
5 “An Investigation of TiO2 Photocatalysis for the Treatment of Water Contaminated with Metals and Organic Chemicals” Close (Prairie et al. 1993) 1993 11 315 326 Environmental science and technology
6 “BiVO4/CeO2 nanocomposites with high visible-light-induced photocatalytic activity” Close (Wetchakun et al. 2012) 2012 6 301 307 ACS applied materials and interfaces
7 “Construction of high-dispersed Ag/Fe3O4/g-C3N4 photocatalyst by selective photo-deposition and improved photocatalytic activity” Close (Zhu et al. 2016) 2016 17 229 246 Applied catalysis B: environmental
8 “Preparation of Au-BiVO4 heterogeneous nanostructures as highly efficient visible-light photocatalysts” Close (Cao et al. 2012) 2012 9 214 223 ACS applied materials and interfaces
9 “Construction of iodine vacancy-rich BiOI/Ag@AgI Z-scheme heterojunction photocatalysts for visible-light-driven tetracycline degradation: Transformation pathways and mechanism insight” Close (Yang et al. 2018) 2018 43 179 222 Chemical engineering journal
10 “New perspectives for Advanced Oxidation Processes” Open (Dewil et al. 2017) 2017 14 193 207 Journal of environmental management
11 “Application of the colloidal stability of TiO2 particles for recovery and reuse in solar photocatalysis” Close (Fernández-Ibáñez et al. 2003) 2003 7 188 195 Water research
12 “Preparation and photocatalytic property of a novel dumbbell-shaped ZnO microcrystal photocatalyst” Close (Sun et al. 2009) 2009 3 187 190 Journal of hazardous materials
13 “Treatment of petroleum refinery sourwater by advanced oxidation processes” Close (Coelho et al. 2006) 2006 4 185 189 Journal of hazardous materials
14 “Treatment of paper pulp and paper mill wastewater by coagulation-flocculation followed by heterogeneous photocatalysis” Close (Rodrigues et al. 2008) 2008 0 183 183 Journal of photochemistry and photobiology A: chemistry
15 “Application of solar AOPs and ozonation for elimination of micropollutants in municipal wastewater treatment plant effluents” Close (Prieto-Rodríguez et al. 2013) 2013 10 170 180 Water research
16 “Engineering of solar photocatalytic collectors” Close (Rodríguez et al. 2004) 2004 13 163 176 Solar energy
17 “A TiO2/AC composite photocatalyst with high activity and easy separation prepared by a hydrothermal method” Close (Liu et al. 2007) 2007 10 161 171 Journal of hazardous materials
18 “Removing pharmaceuticals and endocrine-disrupting compounds from wastewater by photocatalysis” Close (Dalrymple et al. 2007) 2007 3 167 170 Journal of chemical technology and biotechnology
19 “Heterogenous photocatalytic degradation kinetics and detoxification of an urban wastewater treatment plant effluent contaminated with pharmaceuticals” Close (Rizzo et al. 2009) 2009 8 161 169 Water research
20 “Novel Bi2S3/Bi2O2CO3 heterojunction photocatalysts with enhanced visible light responsive activity and wastewater treatment” Close (Liang et al. 2014) 2014 10 156 166 Journal of materials chemistry A
The 2nd on the list is an article by Jian Long Wang and Le Jin Xu in 2012, it receives 636 TC, out of which 629 is for GC, and 7 for LC. However, it is a close-access article that may reduce its number of citations to some extent. The 3rd in the list is a close access article, it is a research article that presents an interesting method of tuning large bandgap photocatalyst (TiO2) to absorb visible light and reduce its rate of recombination, by employing noble metal (gold). This article was written by X. Z. Li and F. B. Li has attracted a wide range of readers, it receives 627 TC, out of which 625 are GC and 2 are LC.
The rest of the highly cited articles all have a TC range of 380–166, with a range of local citations of 43–0. Interestingly, the article that falls 14th on the list is the only one that has 0 LC. This is interesting because the authors (Angela Claudia Rodrigues and coworkers), tried to bring a hybrid idea of using photocatalysis along with biological treatment (coagulation-flocculation) in the treatment of paper mill wastewater. This has generated a lot of interest and opened yet another novel idea in the treatment of wastewater; as such it generates a lot of GC. It is important to state that out of the 20 most highly cited articles, only 2 are open access, we, therefore, recommend more publications on open access to aid visibility and increase the rate of citations. Likewise, it is worthy of note that review articles receive much more citations than research articles, because of the brainstorming and critical analysis done in review articles leading to the identification of various research gaps and potential areas in need of attention.
Leading authors
Table 3 presents the top 20 most productive authors in wastewater treatment using photocatalysis, the authors are affiliated with 4 countries namely China (11 authors), Portugal (4 authors), Italy (3 authors), and Spain (2 authors). The first publications range from 2001 to 2017. S. Malato from Spain top the list with 18 total publications from 2001, as well as 1218 total citations and 81 h-index. Liugi Rizzo from Italy is the second most productive author, he has a total of 11 publications since 2007, an h-index of 38, and 585 total citations. Guangming Zeng is ranked 3rd on the list, however, he also has 11 total publications but unlike Liugi Rizzo his first year of publication was 2016. Guangming Zeng with an h-index of 136 and 946 total citations is from Hunan University Changsha, China. Looking at the rate of his citations and h-index he may likely top the list of the most productive authors in near future (Table 4).Table 4 List of the most productive authors in photocatalysis for wastewater treatment
S/N Author Scopus Author ID Year of 1st publication TP h-index TC Current affiliation Country
1 Malato, S 57,207,915,948 2001 18 81 1218 CIEMAT-Plataforma Solar de Almería, Almeria Spain
2 Rizzo, Liugi 9,044,416,100 2007 11 38 585 Università di Salerno, Salerno Italy
3 Zeng, Guangming 55,454,449,900 2016 11 136 946 Hunan University, Changsha China
4 Li, Yi 55,881,885,600 2012 9 34 215 Hohai University, Nanjing China
5 Vaiano, Vincenzo 8,432,129,900 2015 9 33 317 Università di Salerno, Salerno Italy
6 Zhang, Wenlog 55,063,813,600 2012 9 22 215 Hohai University, Nanjing China
7 Zhu, Zhi 56,427,214,900 2016 9 19 497 Jiangsu University, Zhenjiang China
8 Oller, Isabel 8,415,190,600 2009 8 40 480 CIEMAT-Plataforma Solar de Almería, Almeria Spain
9 Sacco, Olga 55,502,359,200 2015 8 21 300 Università di Salerno, Salerno Italy
10 Vilar, V.J.P 10,540,195,800 2009 8 45 131 Universidade do Porto, Porto Portugal
11 Yan, Yongsheng 57,217,677,633 2016 8 38 406 Jiangsu University, Zhenjiang China
12 Boaventura, Rui Alfredo Rocha R 6,701,822,293 2009 7 57 118 Universidade do Porto, Porto Portugal
13 Huo, Pengwei 24,366,451,100 2017 7 40 188 Jiangsu University, Zhenjiang China
14 Li, Shijie 56,257,988,500 2017 7 23 273 Zhejiang Ocean University, Zhoushan China
15 Liu, Jianshe 15,755,883,100 2012 7 42 311 Donghua University, Shanghai China
16 Silva, Adrian M.T 56,329,177,700 2013 7 52 313 Universidade do Porto, Porto Portugal
17 Chen, Fei 55,619,290,134 2016 6 30 366 Hunan University, Changsha China
18 Dong, Hongjun 55,543,111,300 2014 6 35 459 Jiangsu University, Zhenjiang China
19 Faria, Joaquim Luis 7,006,045,981 2013 6 55 301 Universidade do Porto, Porto Portugal
20 Jiang, Wei 56,931,753,200 2017 6 19 214 Zhejiang Ocean University, Zhoushan China
The authors that fall 4th–7th in the list all have 9 total publications that started from 2012 to 2016, three of the authors namely Yi Li, Wenlog Zhang, and Zhi Zhu are all from China, while the remaining one (Vincenzo Vaiano) is from Italy. The remaining authors all have at least 6 total publications ranging from 2009 to 2017. It is important to know that about 4 productive authors on the list come from Jiangsu University, Zhenjiang (China), and such can be termed as the most productive institution on the list.
Author keywords
A total of 1000 keywords were identified, with 801 (80.1%) being used once, 304 (30.4%) being used twice, and 199 (19.9%) being used three times. After re-labeling single synonyms words and relatable wordings, 785 keywords had at least one occurrence and the largest set of connected items for the mapping in VOSviewer.
Terminology and concept
‘Photocatalysis’ was found to be the most frequently used keyword as can be seen in Fig. 4, it has 458 occurrences and 249 links to other keywords. ‘TiO2 particles’ is the second most encountered keyword with 204 occurrences and linked to 249 other keywords. The third most encountered keyword is ‘wastewater treatment’ with 203 occurrences and 236 links to other keywords. The rest are ‘photocatalytic activity’ (164 occurrences, 214 links), ‘photocatalysts’ (118 occurrences, 158 links), ‘visible light irradiation’ (109 occurrences, 88 links), ‘nanomaterials’ (53 occurrences, 88 links), ‘ZnO particles’ (50 occurrences, 80 links), ‘Advanced oxidation process’ (49 occurrences, 87 links), ‘adsorption’ (46 occurrences, 74 links), and others.Fig. 4 Bibliometric map based on co-occurrence of author keywords
The prevalence of TiO2 and ZnO particles in the most encountered keywords might not be surprising, because TiO2 and ZnO were among the first photocatalyst known, as such their properties and usage were studied extensively in various photocatalytic applications. To date, there is so much research still on TiO2 and ZnO photocatalysts, most importantly in using them to produce nanocomposites with other narrow bandgap semiconductors. Hence, the prevalence of nanomaterials is also one of the most encountered keywords. Nanomaterials are usually employed in photocatalysis because of their ability to provide a very large surface area suitable for adsorption during photocatalytic reactions leading to enhanced photocatalytic activity. Likewise, recent research on photocatalysis is focusing on visible-light-driven photocatalysts, because of the abundance of visible light in nature, unlike ultraviolet light which is limited. So it is evident that almost all the 10 most frequently encountered keywords listed above are related to one another.
Topics of interest
As one of the successful strategies for producing efficient photocatalysts, the ‘Heterojunction’ keyword has 32 occurrences, signifying its frequent usage. Likewise ‘Z-scheme heterojunction’ has 21 occurrences because a recent research on heterojunction is focused on producing Z-scheme-based photocatalysts. Meanwhile, ‘Graphitic carbon nitrate’ appeared in 31 occurrences, it is one of the efficient materials currently under investigation for various photocatalytic reactions; as such, there is growing interest in it.
‘Semiconductor photocatalysis’ was used 13 times (13 occurrences), among the various methods of synthesis of photocatalysts sol-gel methods were frequently mentioned (18 occurrences) and linked to numerous keywords like ‘TiO2 particles, ‘ZnO particles, ‘calcination temperature’, etc. The sol-gel method is one of the efficient methods of synthesis of nanomaterials, it is largely employed in the synthesis of most photocatalysts because of the simplicity and flexibility of the process.
Among the various sources of wastewater, pharmaceutical wastewater appeared the most with 26 occurrences, followed by organic dye wastewater (20 occurrences), textile wastewater (13 occurrences), dyes wastewater (9 occurrences), contaminants of emerging concern (6 occurrences), industrial wastewater (5 occurrences), dairy wastewater (4 occurrences), ‘sulfate wastewater’ (1 occurrence), and refinery wastewater (1 occurrence). So, cumulatively it can be seen that dyes are the main contributor to the menace of wastewater generation because organic dye wastewater and textile wastewater, and some portion of pharmaceutical wastewater all constitute a large proportion of dyes.
Among the various types of dyes, methylene blue generated a lot of concern as the keyword ‘methylene blue photodegradation’ has 43 occurrences and is linked to 64 keywords. Methylene blue is a cationic dye known to be resistant to so many remediation techniques, but using photocatalysis, its mineralization can largely be achieved. Other dyes that appeared the most are ‘Rhodamine B’ (25 occurrences), ‘methyl orange’ (13 occurrences), and ‘congo red’ (9 occurrences), while ‘Acid orange 7’ (2 occurrences) and ‘Synazol yellow dye’ (1 occurrence) appear a few times.
Other sources of wastewater that appeared most are ‘Bisphenol A’ (11 occurrences), ‘Endocrine disruptors’ (11 occurrences), ‘carbamazepine’ (7 occurrences), ‘inorganic pollutants’ (3 occurrences), ‘arsenic’ (2 occurrences), and ‘4-chlorophenol’ (2 occurrences). The keyword ‘Degradation; has also been repeated 41 times and linked to about 61 keywords like ‘reusability’, ‘disinfection’, ‘tetracycline degradation’, and others. The complete degradation of and mineralization of wastewater is desirable, due to the shortage of fresh water, as such, emphasis is given to photocatalysts that can eventually mineralize the wastewater completely, which will enhance the reusability of the water.
Distribution of photocatalysis publications based on the major subthemes
The connections between the outputs of the subtheme search and central theme search were revealed in Fig. 5. Composites with a search phrase: (“photocatalysis*” OR “photocatalyst*” “wastewater treatment” AND composite*) was the most prominent subtheme with a total of 332 articles. This was followed by Semiconductor (132 articles), Heterojunction (94 articles), and conducting polymers (3 articles). The keywords containing semiconductor appeared in ‘TiO2 semiconductor (1 occurrence), ‘semiconductor photocatalysis’ (14 occurrences), ‘alloyed semiconductor’ (1 occurrence), and ‘semiconducting mineral’ (1 occurrence). TiO2 and ZnO were the first set of semiconductor photocatalysts found; as such research in photocatalysis including semiconductors was reported a long time ago, which is why there are relatively higher publications on semiconductors within these years. Earlier semiconductor photocatalysts have wide-bandgap as such their usage is limited to an ultraviolet region of the electromagnetic spectrum.Fig. 5 Research trends of the selected major strategies used in photocatalysis for wastewater treatment
Likewise, after the advent of photocatalysis, it was understood that single semiconductor photocatalysts are not efficient enough, because of the rapid rate of recombination of electrons and holes after excitation, which subsequently leads to the inactivation of the catalyst. This led to the search for various strategies for improving the activity of photocatalysts, since the effective separation of charges is believed to reduce the rate of recombination and improve the photocatalytic activity, composites are therefore produced with other materials, in order to improve the electron and holes separation as well as tuning the bandgap of the photocatalysts to be active under visible light, which is much abundant than ultraviolet light. That is why there is high research interest in composites evidenced by the highest number of publications within those years.
Heterojunctions are also one of the most promising techniques for the preparation of efficient photocatalysts, due to their feasibility and effectiveness for the spatial separation of electron and hole pairs. Heterojunction refers to the interface between two different semiconductors with unequal band structures, which can result in band alignments. Three types of heterojunctions were identified based on the band structure, there are those with a straddling gap (type 1), staggered gap (type 2), and broken gap (type 3) (Low et al. 2017). The keyword ‘Heterojunction’ appeared 32 times and linked to 52 keywords. The research on heterojunction has increased drastically from 2015 to date as can be seen in Fig. 5; this is projected to increase in the coming years because of increased interest in Z-scheme heterojunction, which is far more efficient than other types of heterojunction. Z-scheme heterojunction, however, has 21 occurrences and is linked to 28 other keywords.
As pointed out earlier there is also a growing interest in the usage of graphitic carbon nitrate, which is a polymeric material, the keyword ‘graphic carbon nitrate” has 31 occurrences and is linked to numerous keywords. This is an indicator that polymers are receiving attention recently, probably due to their resiliency; they can be used in doping, protonation, and molecular functionalization. Polymers such as the graphic carbon nitrate were reported to show enhanced photocatalytic activity in the decomposition of water and degradation of various pollutants.
Conducting polymers are a set of polymeric materials having the properties of inorganic semiconductors, as they provide a similar band structure to most metal oxide semiconductors. They are very stable and easy to synthesize, and their great charge carriers’ mobility, as well as compatibility, qualifies them to be used as efficient photocatalysts. The members of this interesting class of polymers include polyaniline (PANI), poly(3,4-ethylenedioxythiophene)(PEDOT), polypyrrole (PPy), and polythiophene (PTh). As can be seen from Fig. 5, their usage in wastewater treatment is very limited, interestingly in the few publications reported on them, they were found to be very effective, as such we strongly recommend their exploration for wastewater treatment. Besides, it was reported that the composites of metal oxide and conducting polymers are more efficient than that of metal oxides or conductive polymers alone (Janáky et al. 2012).
Limitations of the study
The search for the main theme was restricted to “photocatalysis*” OR photocatalyst* “wastewater treatment” from the titles and abstracts, as such the search result will be limited to the application of photocatalysis in wastewater treatment and might not cover all photocatalysis-related articles in Scopus. Although photocatalysis can be used in so many processes like hydrogen generation and other water-splitting reactions but its application in wastewater treatment is increasing tremendously, because of the flexibility and simplicity of the processes. As such, it is recommended that future research analyze the applications of photocatalysis in other processes, and also comparison can be made using the outputs of the available databases like Scopus and Web of Sciences.
Conclusion
This study provides insight into the research trends of the use of photocatalysis in wastewater treatment, based on bibliometric analysis which is rarely seen in Physical Science, using 1373 publications extracted from the database of Scopus. The result revealed that from the year 2002 there was a growing interest in research in that field, with a cumulative increase every year to date. A massive increase in the number of publications was discovered from 2017 to 2020, with an increase of more than 70 publications every year; this is expected to increase rapidly in the coming years. Almost 38% of the total publications globally on this topic were contributed by China, followed by India and others. The majority of the most productive Institutions also come from China, with Jiangsu University as the single most productive institution. China besides its large number of publications, has strong collaborations globally, this will give potential countries and regions ideas for further collaborations to widen their spectrum of research. The most current research is aimed at resolving photocatalyst bottlenecks, such as fast photogenerated electron–hole recombination, limited visible-light response ability, and low specific surface area. The development of composite photocatalyst has gotten a lot of attention, and conducting polymers are playing an important role because of their unique properties of good physicochemical stability and an appealing electronic structure combined with a medium band gap. As a result, their importance as a potential agent for enhancing and modifying the properties of conventional photocatalyst is acknowledged. Different research areas that received increased attention recently were discussed, and hot topics that require urgent consideration like the usage of conducting polymers as a component of nanocomposite photocatalysts were also outlined. This study produced a kind of hybrid review paper that can provide a synopsis of the way forward regarding the usage of photocatalysis in various wastewater treatments, as well as potential opportunities, future research, and collaborations in the field.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 17 KB)
Acknowledgements
This work was supported by the Malaysia Ministry of Higher Education under Fundamental Research Grant Scheme [FRGS/1/2018/STG07/UTM/02/3] and the Universiti Teknologi Malaysia under Collaborative Research Grant [grant number Q.J130000.2451.08G35].
Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Declarations
Conflict of 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.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Akiyama K Nojima S Ito Y Synthesis of a gold-inserted iron disilicide and rutile titanium dioxide heterojunction photocatalyst via the vapor-liquid-solid method and its water-splitting reaction ACS Omega 2022 10.1021/acsomega.2c04360
Anuar A Marwan NF Smith J (2021) Bibliometric analysis of immigration and environmental degradation: evidence from past decades Environ Sci Pollut Res 2021 299 29 13729 13741 10.1007/S11356-021-16470-1
Arsad AZ Hannan MA Al-Shetwi AQ Hydrogen energy storage integrated hybrid renewable energy systems: a review analysis for future research directions Int J Hydrogen Energy 2022 47 17285 17312 10.1016/J.IJHYDENE.2022.03.208
Asghar A Raman AAA Daud WMAW Advanced oxidation processes for in-situ production of hydrogen peroxide/hydroxyl radical for textile wastewater treatment: a review J Clean Prod 2015 87 826 838 10.1016/j.jclepro.2014.09.010
Bai X Zhang X Hua Z Uniformly distributed anatase TiO2 nanoparticles on graphene: synthesis, characterization, and photocatalytic application J Alloys Compd 2014 599 10 18 10.1016/j.jallcom.2014.02.049
Bernabeu A Vercher RF Santos-Juanes L Solar photocatalysis as a tertiary treatment to remove emerging pollutants from wastewater treatment plant effluents Catal Today 2011 161 235 240 10.1016/j.cattod.2010.09.025
Bitaraf M Amoozadeh A The first report of covalently grafted semiconductors; n-TiO2-P25@ECH@WO3 as a new, efficient, robust and visible-light-responsive photocatalyst J Chem Technol Biotechnol 2021 96 963 970 10.1002/JCTB.6605
Bruce M Limin M Recent advances on water pollution research in Africa: a critical review Int J Sci Adv 2021 2 3 96 375 10.51542/ijscia.v2i3.23
Cao SW Yin Z Barber J Preparation of Au-BiVO4 heterogeneous nanostructures as highly efficient visible-light photocatalysts ACS Appl Mater Interfaces 2012 4 418 423 10.1021/am201481b 22141400
Cascajares M Alcayde A Salmerón-Manzano E Manzano-Agugliaro F The bibliometric literature on scopus and wos: The medicine and environmental sciences categories as case of study Int J Environ Res Public Health 2021 18 5851 10.3390/ijerph18115851 34072479
Chi Y Xu S Li M Effective blockage of chloride ion quenching and chlorinated by-product generation in photocatalytic wastewater treatment J Hazard Mater 2020 396 122670 10.1016/j.jhazmat.2020.122670 32353733
Chong MN Lei S Jin B Optimisation of an annular photoreactor process for degradation of Congo Red using a newly synthesized titania impregnated kaolinite nano-photocatalyst Sep Purif Technol 2009 67 355 363 10.1016/j.seppur.2009.04.001
Choudhury AKR Technologies for the management of wastewater generated in wet processing Waste Manag Fash Text Ind 2021 10.1016/B978-0-12-818758-6.00005-3
Chowdhary P, Bharagava RN, Mishra S, Khan N (2020) Role of Industries in Water Scarcity and Its Adverse Effects on Environment and Human Health. In: Environmental Concerns and Sustainable Development. Springer. Singapore, pp 235–256.
Coelho A Castro AV Dezotti M Sant Anna GL Treatment of petroleum refinery sourwater by advanced oxidation processes J Hazard Mater 2006 137 178 184 10.1016/j.jhazmat.2006.01.051 16530949
da Santos DHS Xiao Y Chaukura N Regeneration of dye-saturated activated carbon through advanced oxidative processes: a review Heliyon 2022 8 e10205 10.1016/j.heliyon.2022.e10205 36033294
Dalrymple OK Yeh DH Trotz MA Removing pharmaceuticals and endocrine-disrupting compounds from wastewater by photocatalysis J Chem Technol Biotechnol 2007 82 121 134 10.1002/jctb.1657
Dewil R Mantzavinos D Poulios I Rodrigo MA New perspectives for advanced oxidation processes J Environ Manage 2017 195 93 99 10.1016/j.jenvman.2017.04.010 28456288
Dhiman M Sharma R Kumar V Singhal S Morphology controlled hydrothermal synthesis and photocatalytic properties of ZnFe2O4 nanostructures Ceram Int 2016 42 12594 12605 10.1016/j.ceramint.2016.04.115
Feigelson L Muszkat L Bir L Muszkat KA Dye photo-enhancement of TiO2-photocatalyzed degradation of organic pollutants: the organobromine herbicide bromacil. In: Water Science and Technology Int Water Assoc 2000 42 1 275 279 10.2166/wst.2000.0325
Fernández-Ibáñez P Blanco J Malato S De Las Nieves FJ Application of the colloidal stability of TiO2 particles for recovery and reuse in solar photocatalysis Water Res 2003 37 3180 3188 10.1016/S0043-1354(03)00157-X 14509705
Fujishima A Honda K Electrochemical photolysis of water one and two-dimensional structure of Poly (L-Alanine ) shown by specific heat measurements at low Nature 1972 238 37 38 10.1038/238037a0 12635268
Garrido-Cardenas JA Esteban-García B Agüera A Wastewater treatment by advanced oxidation process and their worldwide research trends Int J Environ Res Public Health 2020 17 170 10.3390/ijerph17010170
Ghernaout D Elboughdiri N Advanced oxidation processes for wastewater treatment: facts and future trends Oalib 2020 07 1 15 10.4236/oalib.1106139
Hitam CNC Jalil AA A review on exploration of Fe2O3 photocatalyst towards degradation of dyes and organic contaminants J Environ Manage 2020 258 110050 10.1016/j.jenvman.2019.110050 31929077
Hodges BC Cates EL Kim JH Challenges and prospects of advanced oxidation water treatment processes using catalytic nanomaterials Nat Nanotechnol 2018 13 642 650 10.1038/s41565-018-0216-x 30082806
Hussain T Wahab A A critical review of the current water conservation practices in textile wet processing J Clean Prod 2018 198 806 819 10.1016/j.jclepro.2018.07.051
Janáky C De Tacconi NR Chanmanee W Rajeshwar K Bringing conjugated polymers and oxide nanoarchitectures into intimate contact: light-induced electrodeposition of polypyrrole and polyaniline on nanoporous WO3 or TiO2 nanotube array J Phys Chem C 2012 116 19145 19155 10.1021/jp305181h
Jiao T Zhao H Zhou J Self-assembly reduced graphene oxide nanosheet hydrogel fabrication by anchorage of chitosan/silver and its potential efficient application toward dye degradation for wastewater treatments ACS Sustain Chem Eng 2015 3 3130 3139 10.1021/acssuschemeng.5b00695
Khan AH Khan NA Ahmed S Application of advanced oxidation processes followed by different treatment technologies for hospital wastewater treatment J Clean Prod 2020 269 269 10.1016/j.jclepro.2020.122411
Li XZ Li FB Study of Au/Au3+-TiO2 photocatalysts toward visible photooxidation for water and wastewater treatment Environ Sci Technol 2001 35 2381 2387 10.1021/es001752w 11414049
Li Y Yang XY Rooke J Ultralong Cu(OH)2 and CuO nanowire bundles: PEG200-directed crystal growth for enhanced photocatalytic performance J Colloid Interface Sci 2010 348 303 312 10.1016/j.jcis.2010.04.052 20546764
Li Y Dong H Li L Recent advances in waste water treatment through transition metal sulfides-based advanced oxidation processes Water Res 2021 192 116850 10.1016/J.WATRES.2021.116850 33513467
Liang N Zai J Xu M Novel Bi2S3/Bi2O2CO3 heterojunction photocatalysts with enhanced visible light responsive activity and wastewater treatment J Mater Chem A 2014 2 4208 4216 10.1039/c3ta13931j
Liu SX Chen XY Chen X A TiO2/AC composite photocatalyst with high activity and easy separation prepared by a hydrothermal method J Hazard Mater 2007 143 257 263 10.1016/j.jhazmat.2006.09.026 17049160
Low J Yu J Jaroniec M Heterojunction Photocatalysts Adv Mater 2017 29 1 20 10.1002/adma.201601694
Ma J Miao TJ Tang J Charge carrier dynamics and reaction intermediates in heterogeneous photocatalysis by time-resolved spectroscopies Chem Soc Rev 2022 51 5777 5794 10.1039/d1cs01164b 35770623
Macías-Quiroga IF Henao-Aguirre PA Marín-Flórez A Bibliometric analysis of advanced oxidation processes (AOPs) in wastewater treatment: global and Ibero-American research trends Environ Sci Pollut Res 2020 10.1007/s11356-020-11333-7
Mani S, Bharagava RN (2016) Exposure to crystal violet, its toxic, genotoxic and carcinogenic effects on environment and its degradation and detoxification for environmental safety. In: Reviews of Environmental Contamination and Toxicology. Springer. New York LLC, pp 71–104.
Md Khudzari J Kurian J Tartakovsky B Raghavan GSV Bibliometric analysis of global research trends on microbial fuel cells using Scopus database Biochem Eng J 2018 136 51 60 10.1016/j.bej.2018.05.002
Min S Wang F Han Y An investigation on synthesis and photocatalytic activity of polyaniline sensitized nanocrystalline TiO2 composites J Mater Sci 2007 42 9966 9972 10.1007/s10853-007-2074-z
Mishra R Bera S Chatterjee R A review on Z/S–scheme heterojunction for photocatalytic applications based on metal halide perovskite materials Appl Surf Sci Adv 2022 9 100241 10.1016/j.apsadv.2022.100241
Motahari F Mozdianfard MR Soofivand F Salavati-Niasari M NiO nanostructures: synthesis, characterization and photocatalyst application in dye wastewater treatment RSC Adv 2014 4 27654 27660 10.1039/c4ra02697g
Neppolian B Choi HC Sakthivel S Solar light induced and TiO2 assisted degradation of textile dye reactive blue 4 Chemosphere 2002 46 1173 1181 10.1016/S0045-6535(01)00284-3 11951983
Pawar RC Lee CS Basics of photocatalysis Heterog Nanocomp Photocataly Water Purif 2015 10.1016/b978-0-323-39310-2.00001-1
Pincheira M Vecchio M Giaffreda R Kanhere SS Cost-effective IoT devices as trustworthy data sources for a blockchain-based water management system in precision agriculture Comput Electron Agric 2021 180 105889 10.1016/J.COMPAG.2020.105889
Prairie MR Evans LR Stange BM Martinez SL An investigation of TiO2 photocatalysis for the treatment of water contaminated with metals and organic chemicals Environ Sci Technol 1993 27 1776 1782 10.1021/es00046a003
Prieto-Rodríguez L Oller I Klamerth N Application of solar AOPs and ozonation for elimination of micropollutants in municipal wastewater treatment plant effluents Water Res 2013 47 1521 1528 10.1016/j.watres.2012.11.002 23332646
Rizzo L Meric S Guida M Heterogenous photocatalytic degradation kinetics and detoxification of an urban wastewater treatment plant effluent contaminated with pharmaceuticals Water Res 2009 43 4070 4078 10.1016/j.watres.2009.06.046 19596131
Rodrigues AC Boroski M Shimada NS Treatment of paper pulp and paper mill wastewater by coagulation-flocculation followed by heterogeneous photocatalysis J Photochem Photobiol A Chem 2008 194 1 10 10.1016/j.jphotochem.2007.07.007
Rodríguez SM Gálvez JB Rubio MIM Engineering of solar photocatalytic collectors Sol Energy 2004 77 513 524 10.1016/j.solener.2004.03.020
Salihoglu NK Karaca G Salihoglu G Tasdemir Y Removal of polycyclic aromatic hydrocarbons from municipal sludge using UV light Desalin Water Treat 2012 44 324 333 10.1080/19443994.2012.691689
Shandilya P Mittal D Soni M Fabrication of fluorine doped graphene and SmVO4 based dispersed and adsorptive photocatalyst for abatement of phenolic compounds from water and bacterial disinfection J Clean Prod 2018 203 386 399 10.1016/j.jclepro.2018.08.271
Sharma RK Arora B Dutta S Gawande MB Photo-oxidation technologies for advanced water treatment 2020 Cham Springer 221 255
Sharma P Iqbal HMN Chandra R Evaluation of pollution parameters and toxic elements in wastewater of pulp and paper industries in India: a case study Case Stud Chem Environ Eng 2022 5 100163 10.1016/j.cscee.2021.100163
Singh P Borthakur A A review on biodegradation and photocatalytic degradation of organic pollutants: a bibliometric and comparative analysis J Clean Prod 2018 196 1669 1680 10.1016/j.jclepro.2018.05.289
Sudhaik A Raizada P Shandilya P Singh P Magnetically recoverable graphitic carbon nitride and NiFe2O4 based magnetic photocatalyst for degradation of oxytetracycline antibiotic in simulated wastewater under solar light J Environ Chem Eng 2018 6 3874 3883 10.1016/j.jece.2018.05.039
Sun JH Dong SY Wang YK Sun SP Preparation and photocatalytic property of a novel dumbbell-shaped ZnO microcrystal photocatalyst J Hazard Mater 2009 172 1520 1526 10.1016/j.jhazmat.2009.08.022 19735975
Tan H Li J He M Global evolution of research on green energy and environmental technologies: a bibliometric study J Environ Manage 2021 297 113382 10.1016/J.JENVMAN.2021.113382 34332345
Tomar R, Abdala AA, Chaudhary RG, Singh NB (2020) Photocatalytic degradation of dyes by nanomaterials. In: Materials Today: Proceedings. Elsevier Ltd, pp 967–973. 10.1016/j.matpr.2020.04.144
Wang JL Xu LJ Advanced oxidation processes for wastewater treatment: formation of hydroxyl radical and application Crit Rev Environ Sci Technol 2012 42 251 325 10.1080/10643389.2010.507698
Wang H Li X Zhao X A review on heterogeneous photocatalysis for environmental remediation: from semiconductors to modification strategies Chinese J Catal 2022 43 178 214 10.1016/S1872-2067(21)63910-4
Wetchakun N Chaiwichain S Inceesungvorn B BiVO4/CeO2 nanocomposites with high visible-light-induced photocatalytic activity ACS Appl Mater Interfaces 2012 4 3718 3723 10.1021/am300812n 22746549
Yang Y Zeng Z Zhang C Construction of iodine vacancy-rich BiOI/Ag@AgI Z-scheme heterojunction photocatalysts for visible-light-driven tetracycline degradation: transformation pathways and mechanism insight Chem Eng J 2018 349 808 821 10.1016/j.cej.2018.05.093
Yang Y Li X Zhou C Recent advances in application of graphitic carbon nitride-based catalysts for degrading organic contaminants in water through advanced oxidation processes beyond photocatalysis: a critical review Water Res 2020 184 116200 10.1016/j.watres.2020.116200 32712506
Yang H Yang B Chen W Yang J Preparation and photocatalytic activities of TiO2-based composite catalysts Catalysts 2022 12 10 1263 10.3390/catal12101263
Zhang X Lei L Preparation of photocatalytic Fe2O3-TiO2 coatings in one step by metal organic chemical vapor deposition Appl Surf Sci 2008 254 2406 2412 10.1016/j.apsusc.2007.09.067
Zhang W Li Y Wang Q Performance evaluation and application of surface-molecular-imprinted polymer-modified TiO2 nanotubes for the removal of estrogenic chemicals from secondary effluents Environ Sci Pollut Res 2013 20 1431 1440 10.1007/s11356-012-0983-0
Zhang X Chen YL Liu RS Tsai DP Plasmonic photocatalysis Reports Prog Phys 2013 76 046401 10.1088/0034-4885/76/4/046401
Zhang J Hu W Cao S Piao L Recent progress for hydrogen production by photocatalytic natural or simulated seawater splitting Nano Res 2020 13 2313 2322 10.1007/s12274-020-2880-z
Zhu Z Lu Z Wang D Construction of high-dispersed Ag/Fe3O4/g-C3N4 photocatalyst by selective photo-deposition and improved photocatalytic activity Appl Catal B Environ 2016 182 115 122 10.1016/j.apcatb.2015.09.029
|
PMC010xxxxxx/PMC10213596.txt |
==== Front
Ann Hematol
Ann Hematol
Annals of Hematology
0939-5555
1432-0584
Springer Berlin Heidelberg Berlin/Heidelberg
37233774
5284
10.1007/s00277-023-05284-5
Original Article
Contributions of bone marrow monocytes/macrophages in myeloproliferative neoplasms with JAK2V617F mutation
Fan Wenjuan 1
Cao Weijie 1
Shi Jianxiang 2
Gao Fengcai 1
Wang Meng 1
Xu Linping 3
Wang Fang 1
Li Yingmei 1
Guo Rong 1
Bian Zhilei 145
http://orcid.org/0000-0001-5550-508X
Li Wei zlyylw3028@zzu.edu.cn
145
Jiang Zhongxing jiangzx@zzu.edu.cn
145
Ma Wang doctormawang@126.com
6
1 grid.412633.1 0000 0004 1799 0733 Department of Hematology, the First Affiliated Hospital of Zhengzhou University, Zhengzhou, 450052 Henan China
2 grid.207374.5 0000 0001 2189 3846 BGI College & Henan Institute of Medical and Pharmaceutical Sciences in Academy of Medical Science, Zhengzhou University, Zhengzhou, 450052 Henan China
3 grid.414008.9 0000 0004 1799 4638 Department of Research and Foreign Affairs, The Affiliated Cancer Hospital of Zhengzhou University & Henan Cancer Hospital, Zhengzhou, 450008 China
4 grid.207374.5 0000 0001 2189 3846 The Academy of Medical Science, College of Medical, Zhengzhou University, Zhengzhou, 450052 Henan China
5 Department of Hematology, Henan Provincial Hematology Hospital, Zhengzhou, 450000 Henan China
6 grid.412633.1 0000 0004 1799 0733 Department of Oncology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, 450008 Henan China
26 5 2023
2023
102 7 17451759
18 1 2023
17 5 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The classic BCR-ABL1-negative myeloproliferative neoplasm (MPN) is a highly heterogeneous hematologic tumor that includes three subtypes, namely polycythemia vera (PV), essential thrombocytosis (ET), and primary myelofibrosis (PMF). Despite having the same JAK2V617F mutation, the clinical manifestations of these three subtypes of MPN differ significantly, which suggests that the bone marrow (BM) immune microenvironment may also play an important role. In recent years, several studies have shown that peripheral blood monocytes play an important role in promoting MPN. However, to date, the role of BM monocytes/macrophages in MPN and their transcriptomic alterations remain incompletely understood. The purpose of this study was to clarify the role of BM monocytes/macrophages in MPN patients with the JAK2V617F mutation. MPN patients with the JAK2V617F mutation were enrolled in this study. We investigated the roles of monocytes/macrophages in the BM of MPN patients, using flow cytometry, monocyte/macrophage enrichment sorting, cytospins and Giemsa-Wright staining, and RNA-seq. Pearson correlation coefficient analysis was also used to detect the correlation between BM monocytes/macrophages and the MPN phenotype. In the present study, the proportion of CD163+ monocytes/macrophages increased significantly in all three subtypes of MPN. Interestingly, the percentages of CD163+ monocytes/macrophages are positively correlated with HGB in PV patients and PLT in ET patients. In contrast, the percentages of CD163+ monocytes/macrophages are negatively correlated with HGB and PLT in PMF patients. It was also found that CD14+CD16+ monocytes/macrophages increased and correlated with MPN clinical phenotypes. RNA-seq analyses demonstrated that the transcriptional expressions of monocytes/macrophages in MPN patients are relatively distinct. Gene expression profiles of BM monocytes/macrophages suggest a specialized function in support of megakaryopoiesis in ET patients. In contrast, BM monocytes/macrophages yielded a heterogeneous status in the support or inhibition of erythropoiesis. Significantly, BM monocytes/macrophages shaped an inflammatory microenvironment, which, in turn, promotes myelofibrosis. Thus, we characterized the roles of increased monocytes/macrophages in the occurrence and progression of MPNs. Our findings of the comprehensive transcriptomic characterization of BM monocytes/macrophages provide important resources to serve as a basis for future studies and future targets for the treatment of MPN patients.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00277-023-05284-5.
Keywords
Myeloproliferative neoplasm
Monocytes/macrophages
JAK2V617F mutation
RNA-seq
Microenvironment
http://dx.doi.org/10.13039/501100006407 Natural Science Foundation of Henan Province 222300420567 Jiang Zhongxing Natural Science Foundation of China82270149 82270141 32100698 82170211 Li Wei China Postdoctoral Science Foundation 2022T150592 2021M692930 Li Wei Key research and development and promotion of special projects in Henan Province222102310204 Wang Fang Henan Province Medical Science and Technology Research ProjectSBGJ202102146 LHGJ20220304 LHGJ20220305 Jiang Zhongxing Outstanding Youth Fund of Henan Province222300420068 Bian Zhilei issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIntroduction
Myeloproliferative neoplasms (MPNs) comprise a group of chronic and heterogenous hematological malignancies that exhibit terminal myeloid cell expansion, including erythrocytes, thrombocytes, and leukocytes [1–3]. The three main Philadelphia chromosome–negative MPN subentities, polycythemia vera (PV), essential thrombocytosis (ET), and primary myelofibrosis (PMF), are characterized by chronic activation of the JAK-STAT pathway, resulting from transforming mutations in JAK2, calreticulin, or MPL genes [3–8]. Genetic analysis of patients with MPNs revealed that more than 95% of patients with PV and ∼50% of patients with ET or PMF carry an activating point mutation in the JAK2 gene (JAK2V617F) [9–13]. It is necessary to note that, despite the presence of the same JAK2V617F mutation, the clinical presentation of patients with different types of MPN varies considerably, which indicates that other “driver factors” also play an important role in MPNs.
The bone marrow (BM) niche is a complex and dynamic structure composed of a multitude of cell types which functionally create an interactive network that facilitates hematopoietic stem cell (HSC) development and maintenance [14–17]. Monocytes/macrophages are the key component of the BM microenvironment, which have been reported to support the development of erythroid cells [18–20], HSCs [21, 22], and megakaryocytes [23] in normal conditions. It has also been reported that macrophages promote erythroid cell [18, 24] and platelet production [23] in mouse models. In addition, Marina Dorigatti Borges et al. found that PB monocytes expressed several higher molecules, including cell adhesion, iron metabolism, and those capable of identifying and engulfing damaged and senescent erythrocytes, which suggests that monocytes may be more likely to attach erythroid cells and could, therefore, contribute to form erythroblastic islands (EBIs) if differentiated from macrophages [25]. In contrast, macrophages, which have been shown to induce the proliferation of myofibroblasts via vitamin D receptor signaling [26]], have been found to be increased in BM biopsies of PMF patients [27]. These studies demonstrate that monocytes/macrophages play distinct roles in MPNs; however, the roles of BM monocytes/macrophages in human MPN patients remain unclear.
Although experiments and clinical studies have revealed that the JAK2V617F mutation is associated with MPNs, the transcriptional consequences of the JAK2V617F mutation in different cellular components of the BM have not yet been fully elucidated. Using ScRNA-seq, Tong et al. report that HSCs with the JAK2V617F mutation in ET show strong megakaryocyte (Mk) lineage priming, are more sensitive to IFN signaling than to Mk differentiation, and exhibit distinct signatures upon treatment [3]. Van Egeren et al. found the same results of JAK2V617F mutant HSCs [28]. In addition, they also found that monocytes with JAK2V617F have a pro-inflammatory intermediate monocyte phenotype, and JAK2V617F monocytes express SLAMF7, which is associated with fibrosis in MPNs [28]. Maekawa et al. also report that SLAMF7high monocytes increased in the PB of patients with MF in correlation with the JAK2V617F mutation [4], and anti-SLAMF7 antibodies suppressed monocyte-derived fibrocyte differentiation and could be a potent therapeutic agent for MF [4]. Therefore, an understanding of the mechanism that drives activated monocytes/macrophages in MPN will guide strategies to target monocytes/macrophages in this disease.
In this study, we aimed to investigate the role of monocytes/macrophages in MPN patients with the JAK2V617F mutations. We performed flow cytometry to examine the expression of CD163, CD14, and CD16 in monocytes/macrophages from JAK2V617F mutant MPNs and present RNA-seq analysis of BM monocytes/macrophages of four healthy controls (HCs, also known as normal bone marrow, NBM), four PV patients, three ET patients, and five PMF patients. The analysis in our study shows that either CD163 or combined CD14 and CD16 are significantly elevated in MPN and correlates closely with the clinical phenotype of MPNs. RNA-seq analysis of BM monocytes/macrophages of four HCs, four PV patients, three ET patients, and five PMF patients demonstrated that different subtypes of MPN have relatively specific BM immune microenvironments. And the comprehensive transcriptomic characterization of BM monocytes/macrophages provides important resources to serve as a basis for future studies and future targets for the treatment of MPN patients.
Materials and methods
Patients and bone marrow samples
Eighty-five PV patients, 131 ET patients, and 42 PMF patients with the JAK2V617F mutation who were treated at the First Affiliated Hospital of Zhengzhou University between January 14, 2017, and March 05, 2021, were enrolled in this study for the analysis of peripheral blood routine and BM aspiration biopsy. An amount of 3–5 ml BM was aspirated from four HCs, four PV patients, three ET patients and five PMF patients. This study was approved by the Ethics Committee at the First Affiliated Hospital of Zhengzhou University. All methods and procedures associated with this study were conducted in accordance with the Good Clinical Practice guidelines and the ethical principles of the Declaration of Helsinki, as well as the local laws.
Preparation of single cell mononuclear cells and enrichment of CD163+ monocytes/macrophages
The BM of MPN patients were collected and processed for single-cell preparation for flow cytometry and enrichment of CD163+ monocytes/macrophages. The single cell suspensions were prepared as previously described [18, 29]. Then, BM CD163+ monocytes/macrophages from HCs and MPN patients were enriched with the CD163 MicroBead Kit (Miltenyi, Cat#:130–124-420), according to the manufacturer's instructions. BM cells were first incubated with CD163-biotin for 15 min on ice in the dark. The cells were then washed with PBS plus 2% FBS and 2 mM EDTA, centrifuged at 1200 rpm for 10 min, and then incubated with anti-biotin microbeads for 15 min on ice in the dark. The cells were again washed with PBS plus 2% FBS and 2 mM EDTA, again centrifuged at 300 g for 10 min, and then resuspended with 2-mL PBS plus 2% FBS and 2 mM EDTA. Finally, the BM CD163+ monocytes/macrophages were enriched with Quadro MACS, according to the manufacturer’s instructions.
Flow cytometry staining and analyses
Flow cytometry staining and analyses were performed as previous described [18, 29–31]. Briefly, 5 × 106 mononuclear cells (MNCs) were blocked with 50-μL human Fc receptor blocking (dilution 1:100) for 15 min at 4 °C, then stained with APC-CD163, Percp-CD14, and PE-CY7-CD16 for 30 min on ice in the dark. After staining, the cells were washed once with PBS plus 0.5% BSA and 2 mM EDTA. DAPI was used to gate out dead cells. Then the cells were resuspended with PBS plus 0.5% BSA and 2 mM EDTA, and run on a BD Air III (BD bioscience). Flow Jo software (BD) was used to analyze the data.
RNA-sequencing and analysis
RNA-seq was prepared and analyzed as previously described [29, 31–34]. RNA was extracted from the monocytes/macrophages of the control and MPN patients. Approximately 100 ng of total RNA was used as input for the cDNA library preparation, which was preformed using an Illumina TruSeq kit followed by sequencing with an Illumina HiSeq 4000 platform (Beijing Genomics Institute, BGI, China). Gene read counts for Gencode hg19 version 31 protein coding genes were generated using Kallisto for the analysis of the RNA-seq data. Differential gene expressions were examined with DESeq2, using a Wald test with a 0.05 adjusted p value cutoff and independent filtering to remove genes with lower expression levels, as previously described [29, 31–34]. The KEGG signaling pathway was performed as previously described [29, 31–34].
Cytospins and Giemsa-Wright staining
Cytospins and Giemsa-Wright staining of monocytes/macrophages from the control and MPN patients were performed as described in our previous papers [18, 29, 31].
Statistics
GraphPad Prism 9.0 software (GraphPad Software, Inc.) was used to perform the statistical analysis. All the experiments were repeated at least three times. All data were reported as mean ± SEM. Comparisons between different groups were performed by the student’s t test. Pearson correlation coefficient analysis was used to detect the correlations between monocytes/macrophages CD163 and CD14 combined with CD16 with the MPN phenotype. P < 0.05 was selected to indicate a statistically significant difference.
Results
The proportion of BM CD163+ or CD14+CD16+ monocytes/macrophages increased in MPN and correlated with different MPN subtypes
To evaluate the hypothesis that CD163+ monocytes/macrophages play significant roles in MPN, we collected the BM samples of MPN patients with the JAK2V617F mutation. We first confirmed that phenotype of distinct MPN using blood routine and BM biopsies (Supplementary Fig. 1A and B). Then, we examined the percentage of BM CD163+ monocytes/macrophages in the subtypes of MPN. Figure 1A demonstrates the representative flow cytometry image of CD163 in PV, ET, PMF, and HCs. A quantitative analysis indicated that the percentage of BM CD163+ monocytes/macrophages significantly increase in MPNs (Fig. 1B). Furthermore, the BM CD163+ monocytes/macrophages populations in HC and in MPNs were morphologically different (Fig. 1C). Next, we analyzed the correlation of the BM CD163+ monocyte/macrophage percentage with clinical phenotypes of MPNs. Figure 1D indicates that the BM CD163+ monocyte/macrophage percentage is positively correlated with the HGB levels in PV patients. Figure 1E shows that BM CD163+ monocyte/macrophage percentage is positively correlated with PLT in ET patients. In contrast, BM CD163+ monocyte/macrophage percentage is negatively associated with HGB in PMF patients (Fig. 1F). CD14+CD16+ monocytes/macrophages display an inflammatory function [30, 35, 36]. We next performed flow cytometry to analyze the CD14+CD16+ subset in CD163+ monocytes/macrophages in MPNs. Similarly, the percentage of the CD14+CD16+ subpopulation was also found to increase in PV, ET and PMF as compared to HCs (Supplementary Fig. 2A and B). The BM CD14+CD16+ subpopulation percentage is positively correlated with HGB in PV patients (Supplementary Fig. 2C), but is negatively correlated with HGB in PMF patients (Supplementary Fig. 2E), while there is no correlation between the CD14+CD16+ subpopulation percentage and PLT in ET patients (Supplementary Fig. 2D). Collectively, these data suggest that the BM monocyte/macrophage percentage correlates with MPN phenotypes.Fig. 1 The detection and correlation analysis between the percentage of BM CD163+ monocytes/macrophages and MPN phenotype. A Representative plot of CD163 versus FSC-H of DAPI− BM monocytes/macrophages in HC, PV, ET, and PMF. B Quantitative analysis of CD163 proportion in HC (N = 6), PV (N = 10), ET (N = 10), and PMF (N = 8). C Representative cytospin images of BM monocytes/macrophages in HC, PV, ET, and PMF. D The correlation between RBC, HGB and the percentage of BM CD163+ monocytes/macrophages in PV patients (N = 10). E The correlation between PLT and the percentage of BM CD163+ monocytes/macrophages in ET patients (N = 10). F The correlation between WBC, RBC, HGB, PLT, and the percentage of BM CD163+ monocytes/macrophages in PMF patients (N = 8)
Significant transcriptional differences of monocytes/macrophages in PV patients compared to NBM
We then performed an RNA-seq analysis of BM monocytes/macrophages between treatment-naive PV patients and HCs to characterize the underlying molecular changes of the monocytes/macrophages. Principal component analysis (Fig. 2A) and hierarchical clustering analysis (Supplementary Fig. 3A) indicated that the monocytes/macrophages of treatment-naive PV patients clustered distinctly from those of HCs. A heatmap of the differential expression of genes between the two groups is shown in Fig. 2B. A total of 2,236 genes are differentially expressed, of which 1070 are upregulated and 1166 are downregulated in treatment-naive PV patients versus HCs (Supplementary Fig. 3B). All the differentially expressed genes are listed in Supplementary Table 1. Finally, the upregulated pathways of PV include cytokine–cytokine receptor interaction, cytokine and cytokine receptor, the Rap1 signaling pathway, the cGMP-PKG signaling pathway, platelet activation, and complement and coagulation cascades (Supplementary Fig. 3C). The downregulated pathways of PV include ribosome, cell cycle, oxidative phosphorylation, and the pathways of neurodegeneration-multiple diseases (Supplementary Fig. 3D).Fig. 2 Comparison of RNA-seq data of monocytes/macrophages between PV patients and healthy controls. A Principal component analysis between PV patients (N = 4) and healthy controls (N = 4). B Heatmap of the differentially expressed genes between PV patients (N = 4) and healthy controls (N = 4). C The FPKM value of cytokines and chemokines, CCL5, CXCL5, CXCL9, CXCL10, VEGF-C, and IL27 in PV and NBM. D The FPKM value of FCGR3A, CCR2, TLR1, TLR6, TLR8, and TLR9 in PV and NBM. E The FPKM value of C1QA, C1QB, and C2 in PV and NBM. F The FPKM value of APOL1, APOL2, APOL3, and APOL4 in PV and NBM. G The FPKM value of SLC25A16, SLC46A1, and SLC6A12 in PV and NBM. H The FPKM value of ITGA2, ITGA7, ITGB3, and ITGB5 in PV and NBM. I The FPKM value of NR1H3, CEBPA, and PPARGC1B in PV and NBM
Our research, as well as that of other groups, indicates that the expression levels of genes encoding proteins that are known to be important for the macrophage function of supporting erythropoiesis include adhesion molecules, molecules for nucleus engulfment and digestion, iron recycling molecules, and growth factors [18–21, 24, 29, 37–42]. Fig. 2C demonstrates the increased cytokines and chemokines, including CCL5, CXCL5, CXCL9, CXCL10, and VEGF-C. Monocytes/macrophages in treatment-naive PV patients also increased some of the receptors expressed by inflammatory monocytes/macrophages [5, 40, 43–45], including FCGR3A, CCR2, TLR1, TLR6, TLR7, TLR8, and TLR9 (Fig. 2D). C1QA, C1QB, and C2 are phagocytosis facilitators [46], which also increased in treatment-naive PV patients (Fig. 2E). APOL families play an important role in lipid metabolism and transport [47]. Significantly, the expression of APOL1, APOL2, APOL3, and APOL4 increased in treatment-naive PV patients (Fig. 2F). Solute carrier families play significant roles in ions exchange and homeostasis [48]. Interestingly, the expression of SLC25A16, SLC25A30, SLC46A1, and SLC6A12 also increased in treatment-naive PV patients (Fig. 2G). Integrins are known to participate in cell adhesion as well as cell-surface-mediated signaling [49]. We found that the expression of ITGA2, ITGA7, ITGB3, and ITGB5 increased significantly in treatment-naive PV patients (Fig. 2H). Gene expression is regulated by transcription factors. Finally, we demonstrated that the expression of the transcription factors NR1H3, CEBPA, and PPARGC1B increased significantly in treatment-naive PV patients (Fig. 2I).
Monocytes/macrophages promote ET via several distinct mechanisms compared to NBM
Using the same approach, we examined transcriptional alterations in monocytes/macrophages in the BM microenvironment of ET patients. Principal component analysis (Fig. 3A) and hierarchical clustering analysis (Supplementary Fig. 4A) indicated that treatment-naive ET patients cluster differently from HCs. A heatmap of the differential expression of genes is shown in Fig. 3B. A total of 2161 genes are differentially expressed, of which 1040 are upregulated and 1121 are downregulated in treatment-naive ET patients versus HCs (Supplementary Fig. 4B). All the differentially expressed genes are listed in Supplementary Table 2. The KEGG pathways analysis demonstrated that the upregulated pathways of ET include platelet activation, cell adhesion molecules, ECM-receptor interaction, focal adhesion, and the regulation of actin cytoskeleton (Supplementary Fig. 4C). The downregulated pathways of ET include viral protein interaction with cytokine and cytokine receptors, oxidative phosphorylation, thermogenesis, the NF-kappa B signaling pathway, the TNF signaling pathway, and the IL-17 signaling pathway (Supplementary Fig. 4D).Fig. 3 Comparison of RNA-seq data of monocytes/macrophages between ET patients and healthy controls. A Principal component analysis between ET patients (N = 3) and healthy controls (N = 4). B Heatmap of the differentially expressed genes between ET patients (N = 3) and healthy controls (N = 4). C The FPKM value of CCL5, PDGFA, PDGFB, VWF, FGF13, EGF, and IGF2 in ET and NBM. D The FPKM value of ITGA2, ITGA2B, ITGA6, and ITGA9 in ET and NBM. E The FPKM value of SLC40A1 in ET and NBM. F The FPKM value of MMP9 and MMP25 in ET and NBM. G The FPKM value of FOS, ARG1 and CEBPA in ET and NBM
Emerging evidence suggests that effective megakaryopoiesis depends on the homeostasis of the BM macrophage microenvironment [23, 50]. Fig. 3C demonstrates that the expression of CCL5, PDGFA, PDGFB, VWF, FGF13, EGF, and IGF2 increased significantly in treatment-naive ET patients. Integrins are known to mediate the adhesion of platelets and other cell types to the extracellular matrix to maintain the homeostasis of megakaryopoiesis [51]. Consistent with this, we found that the expression of ITGA2, ITGA2B, ITGA6, and ITGA9 increased significantly (Fig. 3D). SLC40A1 is involved in iron export from duodenal epithelial cells or macrophages, which is important for erythropoiesis [18]. Megakaryopoiesis also requires iron export macrophages. Significantly, the expression of SLC40A1 increased in treatment-naive ET patients (Fig. 3E). The matrix metalloproteinase (MMP) family is involved in the breakdown of extracellular matrix in normal physiological processes, such as embryonic development, reproduction, tissue remodeling, and megakaryocytes migration [52–54]. Fig. 3F indicates that the expression of MMP9 and MMP25 increased significantly. Studies have indicated that M2 macrophages, but not M1 macrophages, are significant for megakaryopoiesis [23]. Consistent with this, the expression of the M2 macrophage marker ARG1 increased significantly (Fig. 3G). Two transcription factors, FOS and CEBPA expressions, were also found to be expressed in treatment-naive ET patients (Fig. 3G).
Monocytes/macrophages of PMF distinct heterogeneous expression of M1 and M2 markers compared to NBM
We examined the transcriptional alterations in monocytes/macrophages in the BM microenvironment of PMF patients. Principal component analysis (Fig. 4A) and hierarchical clustering analysis indicated that treatment-naïve clustering in PMF patients differ from clustering in HCs (Supplementary Fig. 5A). A heatmap of the differential expression of genes is shown in Fig. 4B. A total of 2252 genes are differentially expressed, of which 1347 are upregulated and 905 are downregulated in treatment-naive ET patients versus HCs (Supplementary Fig. 5B). All the differentially expressed genes are listed in Supplementary Table 3. The upregulated pathway of PMF includes cytokine–cytokine receptor interaction, platelet activation, a JAK-STAT signaling pathway, complement and coagulation cascades, and Fc gamma R-mediated phagocytosis (Supplementary Fig. 5C). However, the downregulated pathways only include the coronavirus disease COVID-19 and ribosome (Supplementary Fig. 5D).Fig. 4 Comparison of RNA-seq data of monocytes/macrophages between PMF patients and healthy controls. A Principal component analysis between PMF patients (N = 5) and healthy controls (N = 4). B Heatmap of the differentially expressed genes between PMF patients (N = 5) and healthy controls (N = 4). C The FPKM value of FCGR1A, FCGR1B, FCGR1C, FCGR3A, FCGR3B, HLA-DOA, HLA-DPA1, and HLA-DPB1 in PMF and NBM. D The FPKM value of STAT1, IFITM1, IFITM2, IFITM3, IFIT1, IFIT2, IFIT3, and IFIT27 in PMF and NBM. E The FPKM value of CCL5, GAS6, CCL25, CXCL9, CXCL10, CXCL12, VEGF-C, EGF, VWF, PDGFA, PDGFB, and IL27 in PMF and NBM. F The FPKM value of SIGLEC1, VCAM1, AXL, CD5L, and SLAMF7 in PMF and NBM. G The FPKM value of SPIC in PMF and NBM
It has been shown that macrophages increase in PMF [27]; however, transcriptional alterations in monocytes/macrophages in the BM microenvironment of PMF patients have not been performed. Figure 4C demonstrates that the expressions of FCGR1A, FCGR1B, FCGR1C, FCGR3A, FCGR3B, HLA-DOA, HLA-DPA1, and HLA-DPB1 increased significantly. Figure 4D indicates that many interferon-inducible genes, including STAT1, IFITM1, IFITM2, IFITM3, IFIT1, IFIT2, IFIT3, and IFIT27 increased significantly. We analyzed the production of cytokines and chemokines next and found that the expressions of CCL5, GAS6, CCL25, CXCL9, CXCL10, CXCL12, VEGF-C, EGF, VWF, PDGFA, PDGFB, and IL27 increased significantly (Fig. 4E). SIGLEC1, VCAM1, AXL, CD5L, and SLAMF7 were also found to be significantly increased (Fig. 4F), and, finally, we found that the transcription factor SPIC was also significantly increased (Fig. 4G).
Monocytes/macrophages in PMF, PV, and ET have relatively specific transcriptional profiles
We have shown that monocytes/macrophages in PMF, PV, and ET demonstrate distinct transcriptional expressed genes compared to NBM. To investigate the distinct transcriptional profiles of monocytes/macrophages in PMF, PV, and ET, we performed a two-by-two comparison of the transcriptional expression of monocytes/macrophages in the BM microenvironment of these three different types of MPN. Significantly, we still obtained similar results using this analysis (Supplementary 6A-F, Supplementary Tables 4, 5, and 6, Fig. 5A–H). Taken together, monocytes/macrophages may play relatively specific role in distinct MPNs.Fig. 5 Comparison of differential expression genes of monocytes/macrophages between PV, ET, and PMF patients. A The FPKM value of CCL18, CCL25, CXCL5, CXCL10, GAS6, IL15, IL27, IL18BP, and S100A12 in PV, ET and PMF. B The FPKM value of NR1H3, KLF1, HIF1α, MAFB, and ARG1 in PV, ET and PMF. C The FPKM value of STAT1, IRF1, IFITM3, IFI27, and AIM2 in PV, ET, and PMF. D The FPKM value of SLC2A5, SLC31A2, SLC40A1, and SLC6A12 in PV, ET and PMF. E The FPKM value of APOL1, APOL2, APOL3, and APOL4 in PV, ET and PMF. F The FPKM value of DNASE2, ICAM1, and MARCO in PV, ET and PMF. G The FPKM value of MMP8 and MMP9 in PV, ET, and PMF. H The FPKM value of HLA-DPB1, HLA-DQA1, HLA-DRB1, HLA-DMB, HLA-DRA, HLA-A, HLA-DMA and CD74 in PV, ET, and PMF
Discussion
The classical view is that malignant clonal expansion of hematopoietic stem/progenitor cells is the central driver of MPN development, but the same JAK2V617F mutation allows MPN to manifest in three clinical phenotypes, namely PV, ET, and PMF [9–12]. Studies have illustrated the complexity of MPN pathogenesis and the heterogeneity of clinical manifestations well, and there is increasing evidence that the BM microenvironments may play important roles [55–57]. Medical interventions, immune dysregulation, and inflammation have all been shown to alter selective pressures for mutant clones in the BM. Transcriptome analysis of PBMCs has shown that MPN patients are significantly enriched for genes related to inflammation, immune response, and oxidative stress [58]. Immune infiltration analysis as well as mathematical modeling studies have also suggested the presence of complex immune remodeling in the MPN microenvironment [59, 60]. It has also been shown that there is significant immune activation in the microenvironments of MPN patients, including decreased levels of anti-inflammatory cytokines, increased levels of pro-inflammatory cytokines, as well as a significant Th1/Th2 immune bias [61]. Despite these studies, however, the role of monocytes/macrophages in the BM microenvironment in MPN is still unclear. In the present study, we found that CD163+ monocytes/macrophages in the BM of MPN increased significantly compared to HCs. Interestingly, there was significant variability in the morphology of monocytes/macrophages in different subtypes of MPN, suggesting that monocytes/macrophages may have relatively specific functions in the different subtypes of MPN. Correlation analyses showed a positive correlation between monocyte/macrophage counts and elevated hemoglobin in patients with PV. Monocyte/macrophage counts in ET patients were positively correlated with elevated platelets. In contrast, the monocyte/macrophage count in PMF patients was positively correlated with a decrease in hemoglobin. Overall, our study suggests that the BM monocyte/macrophage may associate to the development of MPN.
CD14+CD16+ monocytes/macrophages are thought to be involved in a variety of inflammation-related diseases in humans [62–65]. Bassan et al. reported that the peripheral blood CD14+CD16+ monocytes were significantly elevated in different MPN subtypes [66]. Importantly, the JAK2V617F mutation was significantly associated with the percentage of CD14+CD16+ monocytes. Borges et al. demonstrated that the peripheral blood CD14+CD16+ non-classical monocytes significantly expressed CD169, CD206, CD163, SIRPα, and CD106, which are involved in erythropoiesis, in PV [25]. Despite these findings, studies have mainly focused on peripheral blood inflammatory monocytes in MPN. In contrast, alterations in the immune microenvironment of the BM may be more influential in the development and progression of MPN. However, the role of CD14+CD16+ inflammatory monocytes/macrophages in the BM microenvironment in MPN is still unclear. In this study, we found that the percentage of CD14+CD16+ inflammatory monocytes/macrophages in BM increases in the three subtypes of MPN. Interestingly, the percentage of BM CD14+CD16+ inflammatory monocytes/macrophages was positively associated with hemoglobin in PV and platelets in ET. In contrast, the percentage of BM CD14+CD16+ inflammatory monocytes/macrophages was negatively associated with hemoglobin in PMF. In conclusion, our study demonstrates for the first time that inflammatory monocytes/macrophages are elevated in different subtypes of MPN and further confirms that the BM microenvironment of MPN presents an inflammatory microenvironmental state.
Macrophages can promote erythropoiesis through a number of different mechanisms [18–20, 24, 29, 39, 40]. The removal of macrophages from PV mice resulted in a significant decrease in erythrocytes and hemoglobin, suggesting that macrophages are also involved in the development of PV [67]. In the present study, we performed RNA-seq of CD163+ monocytes/macrophages from PV and NBM. Our study found that several cytokines and chemokines, such as CCL5, CXCL5, CXCL9, CXCL10, VEGF-C and IL27, increased significantly compared to NBM. Fang et al. demonstrated that VEGF-C is essential for the mobilization, maturation, and enucleation of primitive erythroblasts, and VEGF-C deletion compromises liver colonization by erythro-myeloid progenitors and subsequent macrophage/erythroid expansion [68]. The same group also indicates that VEGF-C deletion in endothelial or LepR+ cells compromises the bone marrow perivascular niche and hematopoietic stem cell maintenance and exogenously administered VEGF-C improves hematopoietic recovery after irradiation by accelerating endothelial and LepR+ cell regeneration [69]. Therefore, VEGF-C is significant for erythropoiesis, and our study extends the understanding of the role of monocytes/macrophages’ VEGF-C in the development of PV. Our study indicates that GM-CSF causes erythrocytopenia by affecting the formation of EBI [29]. IL27 suppresses the GM-CSF expression [70]. Therefore, IL27 may be a novel cytokine that can promote the production of red blood cells, and further studies should be performed to confirm this. Furthermore, it would also be worth examining the roles of other cytokines or chemokines that are enhanced in the development of PV. The immune microenvironment of PV patients is an inflammatory microenvironment, and an inflammatory microenvironment in BM inhibits erythropoiesis [29, 71] and induces extramedullary erythropoiesis [40, 72]. This is also illustrated by our data on the elevated expression of TLRs and FCGR3A in PV. CCR2 is significant for the homing of HSCs/HPCs to sites of inflammation [44], and CCR2+ monocyte-derived macrophages expand the murine stress erythropoietic niche during recovery from anemia [40]. In conjunction with our finding of an elevated monocyte/macrophage CCR2 expression in the BM microenvironment of PV patients, we hypothesize that PV patients may also have an inflammatory microenvironment that recruits CCR2+ monocytes into the BM microenvironment to differentiate into EBI macrophages that promote erythropoiesis. This should be verified in the future in a mouse model of PV. C1QA [73], C1QB [74], and C2 [74] are significant for phagocytosis. APOL1, APOL2, APOL3, and APOL4 are significant for lipid metabolism [47, 75]. SLC25A16, SLC46A1, and SLC6A12 are significant for ion transport, and ITGA2, ITGA7, ITGB3, and ITGB5 are significant for cell adhesion. These genes may be involved in the occurrence and progression of PV; however, more research should be conducted to demonstrate this. Furthermore, the increased expression of transcription factors in PV included Nr1H3, CEBPA, and PPARGC1B. The exact role of these genes in PV is unclear, and more studies should be conducted to confirm the role of these genes in PV.
Huang’s group report that TGF-β released by M2 macrophages may facilitate megakaryopoiesis through the upregulation of the JAK2/STAT5 and MAPK/ERK pathways in megakaryocytes [23]. In the present study, the expression of ARG1, a specific marker for M2 macrophages, was significantly enhanced in ET. The finding in our study confirms that macrophages in ET patients incline more to the M2 type. We also found that several cytokines and chemokines expressed by monocytes/macrophages were significantly enhanced in ET patients. Previous studies have indicated that CCL5 enhances megakaryocyte differentiation and development [76–79]. We found in our study that CCL5 expression by monocytes/macrophages was significantly elevated in ET, which further demonstrates that monocytes/macrophages can promote platelet production by secreting CCL5. Previous studies have also found that EGF transcript levels and cytokine levels from platelets were significantly elevated in ET [80, 81]. Our study extends the origin of EGF to monocytes/macrophages in ET. MMP9 is involved in platelet formation and the migration of megakaryocytes [52, 53, 82]. We found that monocyte/macrophage expression of MMP9 was significantly elevated in ET patients, which revealed a novel mechanism of platelet production promotion by monocytes/macrophages. Other cytokines, such as PDGFA, PDGFB, VWF, FGF13, and IGF2; cell adhesion molecules, such as ITGA2, ITGA2B, ITGA6, and ITGA9; the iron exporter gene, SLC40A1, matrix metalloproteinase family gene; MMP25 and transcription factors FOS, and CEBPA have also been found to be increased in ET patients. However, the exact roles of these genes expressed by monocytes/macrophages in ET are unclear, and more studies should be conducted to confirm the role of these genes in ET.
Several studies on mouse models have been used to explain the mechanism of myelofibrosis development caused by megakaryocytes [83, 84], stromal cells [85, 86], monocytes [4, 87], and macrophages [26, 88]. However, the transcriptional expression of monocytes/macrophages in PMF patients is still not completely clear. In the present study, we found that the expression of FCGR1A, FCGR1B, FCGR1C, FCGR3A, FCGR3B, HLA-DOA, HLA-DPA1, and HLA-DPB1; the interferon-inducible genes STAT1, IFITM1, IFITM2, IFITM3, IFIT1, IFIT2, and IFIT3; and IFI27 by monocytes/macrophages increased significantly in PMF. These genes expressed by monocytes/macrophages correlate with an inflammatory microenvironment. Our study further confirms that PMF bone marrow microenvironment is an inflammatory microenvironment. Studies have indicated that PDGF-A and -B, as well as PDGF receptor α (PDGFRα) and PDGF receptor β (PDGFRβ) expression is increased in the BM of PMF patients [89]. Decker et al. identified Lepr+ stromal lineage cells as the origin of myofibroblasts in PMF. The PDGFRA pathway in BM Lepr+ cells lead to the expansion of these cells and extramedullary hematopoiesis, and this suggests that targeting PDGFRA signaling could be an effective way to treat BM fibrosis [86]. Kramer et al. have demonstrated that PDGFRβ and PDGF-B protein expression in overt fibrotic BM and PDGFRβ–PDGF-B interaction were significantly increased [90]. In the present study, we found that the expression of PDGFA and PDGFB derived from monocytes/macrophages increased significantly. Mouse models should also confirm the roles of PDGFA and PDGFB derived from monocytes/macrophages in PMF. The expression of CXCL12 by perivascular mesenchymal stromal cells (MSCs) plays an essential role in HSC maintenance [91–93]. Wang et al. report that the splenic microenvironment in MF is characterized by increased levels of intact, functional CXCL12, which contributes to the localization of MF CD34+ cells to the spleen and the establishment of extramedullary hematopoiesis [94]. In the present study, we found that CXCL12 expression by monocytes/macrophages significantly increase in PMF patients. Therefore, we hypothesize that elevated CXCL12 expression may play a compensatory hematopoietic role in the PMF BM hematopoietic microenvironment, and further studies should be conducted to confirm the role of CXCL12 expressed by BM monocytes/macrophages in PMF mouse models. The expression of CCL5 and CXCL9 were also found to be increased in PMF mouse models [95]. Our study confirms that the expression of CCL5 and CXCL9 from monocytes/macrophages also increased in PMF patients. Beitzen-Heineke et al. found that AXL is abundantly activated in MPN cells and that its ligand growth arrest-specific gene 6 is upregulated in MPN patients [96]. The findings from our present study, which involved the expression of AXL and GAS6 from monocytes/macrophages in PMF patients, are consistent with their findings. Xu et al. report that the soluble vascular cell adhesion molecule-1 (sVCAM-1) increases dramatically in myelofibrosis [97]. Increased VCAM1 expression may contribute to the retention of hematopoietic stem cells/PCs in the spleen and play a role in the initiation of extramedullary hematopoiesis in the early stages of MF [98, 99]. Our study found that VCAM1 expressed by monocytes/macrophages in PMF increases significantly. Further studies should be conducted to examine the roles of monocytes/macrophages VCAM1 in PMF models. Maekawa et al. found that SLAMF7high monocytes increase in the PB of patients with MF, in correlation with the JAK2V617F mutation and anti-SLAMF7 antibodies suppressed monocyte-derived fibrocyte differentiation and could be a potent therapeutic agent for MF [4]. In this study, we found that the expression of SLAMF7 in monocytes/macrophages increase significantly in the BM of PMF patients with the JAK2V617F mutation and confirmed that SLAMF7 could be a new therapeutic target for PMF.
In normal conditions, the expression of SPIC by macrophages is significant for iron metabolism [100, 101]. Our previous study also demonstrates that EBI macrophages, but not non-EBI macrophages, express SPIC, which provides iron for erythropoiesis [18]. Inflammatory disorders and infections are associated with cytopenias, including anemia and thrombocytopenia [43]. In a mouse model of macrophage activation syndrome, Akilesh et al. demonstrated that TLR7 or TLR9 signaling in monocytes causes these cells to differentiate into inflammatory hematophagocytes (iHPCs), which highly expressed SPIC [43], and the iHPCs showed higher phagocytic uptake of RBCs [43]. In combination, the increased expression of SPIC in PMF may share similar function to inflammatory disorders. Further studies are also required to examine the exact role of SPIC in PMF models.
Studies have shown that PV is more easily converted into PMF than ET and that the BM microenvironment may play an important role in this process [102, 103]. We compared the transcriptional differences between the monocytes/macrophages in PV, ET, and PMF. Consistent with the easier conversion of PV to PMF, the transcriptional profile of monocytes/macrophages in PV is more similar to those of PMF relative to ET. The differences in the monocyte/macrophage expression profiles in PMF compared to ET patients were both different and identical to those in PMF and in healthy controls. These genes included CCL25, CXCL10, GAS6, IL27, NR1H3, MAFB, STAT1, IFITM3, IFI27, SLC31A2, SLC6A12, APOL1, APOL2, APOL3, APOL4, ICAM1, MARCO, HLA-DPA1, HLA-DPB1, HLA-DQA1, HLA-DRB1, HLA-DMB, HLA-DRA, HLA-A, HLA-DMA, and CD74. According to our study, it seems that many genes of monocytes/macrophages could be involved in PV, ET, and PMF and play similar or specific roles in these three MPN subtypes. Further studies should be performed to confirm the roles of these genes in distinct types of MPN.
However, this study has limitations. The way that differentially expressed genes regulate the development of distinct MPNs should be studied further. Single-cell sequencing of monocytes/macrophages in different subtypes of MPN should also be performed to investigate the similarities and differences in monocytes/macrophages in different subtypes of MPN in-depth. And independent studies are required to confirm the findings of this study. Nevertheless, we have characterized that BM CD163 or CD14 in combination with CD16 expression in monocytes/macrophages increase and may correlate with the clinical phenotypes of MPN. Transcriptome sequencing of monocytes/macrophages in PV, ET, and PMF also showed relatively specific transcriptional profiles of monocytes/macrophages in these three types of MPN, which suggests that monocyte/macrophage-mediated immune remodeling of different subtypes of MPN may also plays a role in MPN.
Supplementary Information
Below is the link to the electronic supplementary material. Supplementary Fig. 1 The clinical phenotype of MPNs. (A) Blood routine test of MPNs (85 of PV patients, 131 of ET patients and 42 of PMF patients), including RBC, HGB, HCT, Reti, PLT and WBC. (B) The representative HE staining of post-bone marrow biopsies of different types of MPN. (PNG 785 kb)
High resolution image (TIF 5709 kb)
Supplementary Fig. 2 The detection of CD14+CD16+subset in CD163+ monocytes/macrophages in MPN and correlation analyses between the proportion of BM CD14+CD16+ monocytes/macrophages and MPN phenotype. (A) Representative plot of CD14 versusCD16 of DAPI-CD163+BM monocytes/macrophages in HC, PV, ET and PMF. (B) Quantitative analysis of CD14+CD16+ proportion in HC(N=6), PV(N=10), ET(N=10) and PMF(N=8). (C) The correlation between RBC, HGB and the percentage of BM CD14+CD16+ monocytes/macrophages in PV patients(N=10). (D) The correlation between PLT and the percentage of BM CD14+CD16+ monocytes/macrophages in ET patients(N=10). (E) The correlation between BC, RBC, HGB, PLT and the percentage of BM CD14+CD16+ monocytes/macrophages in PMF patients(N=8). (PNG 216 kb)
High resolution image (TIF 7051 kb)
Supplementary Fig. 3 Comparison of differential expression genes and KEGG pathways of BM monocytes/macrophages between PV patients and healthy controls. (A) Hierarchical clustering analysis between PV patients (N=4) and healthy controls (N=4).(B) Numbers of unregulated and downregulated genes between PV patients (N=4) and healthy controls (N=4). (C) The upregulated KEGG pathway in PV patients compared to NBM. (D) The downregulated KEGG pathway in PV patients compared to NBM. (PNG 217 kb)
High resolution image (TIF 10315 kb)
Supplementary Fig. 4 Comparison of differential expression genes and KEGG pathways of BM monocytes/macrophages between ET patients and healthy controls. (A) Hierarchical clustering analysis between ET patients (N=3) and healthy controls (N=4).(B) Numbers of unregulated and downregulated genes between ET patients (N=3) and healthy controls (N=4). (C) The upregulated KEGG pathway in ET patients compared to NBM. (D) The downregulated KEGG pathway in ET patients compared to NBM. (PNG 202 kb)
High resolution image (TIF 10514 kb)
Supplementary Fig. 5 Comparison of differential expression genes and KEGG pathways of BM monocytes/macrophages between PMF patients and healthy controls. (A) Hierarchical clustering analysis between PMF patients (N=5) and healthy controls (N=4).(B) Numbers of unregulated and downregulated genes between PMF patients (N=5) and healthy controls (N=4).(C) The upregulated KEGG pathway in PMF patients compared to NBM. (D) The downregulated KEGG pathway in PMF patients compared to NBM. (PNG 228 kb)
High resolution image (TIF 10512 kb)
Supplementary Fig. 6 Comparison of differential expression genes of monocytes/macrophages between PV, ET and PMF patients. (A) Principal component analysis between PMF patients (N=5) and PV patients (N=4). (B) Heatmap of the differentially expressed genes between PMF patients (N=5) and PV patients (N=4). (C) Principal component analysis between PMF patients (N=5) and ET patients (N=3). (D) Heatmap of the differentially expressed genes between PMF patients (N=5) and ET patients (N=3). (E) Principal component analysis between PV patients (N=4) and ET patients (N=3). (F) Heatmap of the differentially expressed genes between PV patients (N=4) and ET patients (N=3). (PNG 147 kb)
High resolution image (TIF 5030 kb)
Supplementary Table 1 The upregulated and downregulated genes in PV compared to NBM. (CSV 686 kb)
Supplementary Table 2 The upregulated and downregulated genes in ET compared to NBM. (CSV 627 kb)
Supplementary Table 3 The upregulated and downregulated genes in PMF compared to NBM. (CSV 725 kb)
Supplementary Table 4 The upregulated and downregulated genes in PMF compared to PV. (CSV 72 kb)
Supplementary Table 5 The upregulated and downregulated genes in PMF compared to ET. (CSV 71 kb)
Supplementary Table 6 The upregulated and downregulated genes in PV compared to ET. (CSV 236 kb)
Acknowledgements
We thank all our authors listed in this manuscript.
Author contribution
WF, WC, and FG designed experiments, performed experiments, and analyzed the data. JS analyzed the RNA-seq data. MW, LX, FW, YL, RG, and ZB designed and supervised the study and edited the manuscript. WL drafted manuscript and edited the manuscript. ZJ and WM revised and edited the manuscript. All authors approved the final manuscript.
Funding
This work was supported by the Natural Science Foundation of China (82270149, 82270141, 32100698, 82170211), China Postdoctoral Science Foundation (2022T150592, 2021M692930) and Young Postdoctoral Innovators in Henan Province (WL). Key research and development and promotion of special projects in Henan Province (222102310204). Henan Province Medical Science and Technology Research Project (SBGJ202102146, LHGJ20220304, LHGJ20220305). Natural Science Foundation of Henan Province (222300420567). Outstanding Youth Fund of Henan Province (222300420068). Funding for Scientific Research and Innovation Team of The First Affiliated Hospital of Zhengzhou University.
Availability of data and materials
All data generated and materials in the study are included in the present article and supplementary data.
Declarations
Ethics approval
This study was reviewed and approved by The First Affiliated Hospital of Zhengzhou University in Zhengzhou, Henan Province, China (2021-KY-0575–002).
Consent for publication
Informed consent was obtained from all patients for being included in the study.
Conflict of interest
The authors declare no competing interests.
Abbreviations
ET Essential thrombocytosis
HCs Healthy controls
HSC Hematopoietic stem cells
iHPCs Inflammatory hematophagocytes
MSCs mesenchymal stromal cells
Mk megakaryocyte
MPNs myeloproliferative neoplasms
NBM normal bone marrow
PV polycythemia vera
PMF primary myelofibrosis
sVCAM-1 soluble vascular cell adhesion molecule-1
Wenjuan Fan, Weijie Cao, Jianxiang Shi, and Fengcai Gao contribute the same to this paper.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Tefferi A Myeloproliferative neoplasms: a decade of discoveries and treatment advances Am J Hematol 2016 91 1 50 58 10.1002/ajh.24221 26492355
2. Dameshek W Some speculations on the myeloproliferative syndromes Blood 1951 6 4 372 375 10.1182/blood.V6.4.372.372 14820991
3. Tong J Sun T Ma S Hematopoietic stem cell heterogeneity is linked to the initiation and therapeutic response of myeloproliferative neoplasms Cell Stem Cell 2021 28 3 502 513.e6 10.1016/j.stem.2021.01.018 33621485
4. Maekawa T Kato S Kawamura T Increased SLAMF7high monocytes in myelofibrosis patients harboring JAK2V617F provide a therapeutic target of elotuzumab Blood 2019 134 10 814 825 10.1182/blood.2019000051 31270105
5. Lai HY Brooks SA Craver BM Defective negative regulation of Toll-like receptor signaling leads to excessive TNF-α in myeloproliferative neoplasm Blood Adv 2019 3 2 122 131 10.1182/bloodadvances.2018026450 30647074
6. Guy A Poisson J James C Pathogenesis of cardiovascular events in BCR-ABL1-negative myeloproliferative neoplasms Leukemia 2021 35 4 935 955 10.1038/s41375-021-01170-z 33658660
7. Skoda RC Duek A Grisouard J Pathogenesis of myeloproliferative neoplasms Exp Hematol 2015 43 8 599 608 10.1016/j.exphem.2015.06.007 26209551
8. Pikman Y Lee BH Mercher T MPLW515L is a novel somatic activating mutation in myelofibrosis with myeloid metaplasia PLoS Med 2006 3 7 e270 10.1371/journal.pmed.0030270 16834459
9. Baxter EJ Scott LM Campbell PJ Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders Lancet (London, England) 2005 365 9464 1054 1061 10.1016/S0140-6736(05)71142-9 15781101
10. Kralovics R Passamonti F Buser AS A gain-of-function mutation of JAK2 in myeloproliferative disorders N Engl J Med 2005 352 17 1779 1790 10.1056/NEJMoa051113 15858187
11. Levine RL Wadleigh M Cools J Activating mutation in the tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis Cancer Cell 2005 7 4 387 397 10.1016/j.ccr.2005.03.023 15837627
12. James C Ugo V Le Couédic JP A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera Nature 2005 434 7037 1144 1148 10.1038/nature03546 15793561
13. Luque Paz D Ashcroft P Skoda RC Myeloproliferative neoplasms: the long wait for JAK2-mutant clone expansion Cell Stem Cell 2021 28 3 359 361 10.1016/j.stem.2021.02.018 33667352
14. Morrison SJ Scadden DT The bone marrow niche for haematopoietic stem cells Nature 2014 505 7483 327 334 10.1038/nature12984 24429631
15. Kunisaki Y Frenette PS The secrets of the bone marrow niche: enigmatic niche brings challenge for HSC expansion Nat Med 2012 18 6 864 865 10.1038/nm.2825 22673997
16. Méndez-Ferrer S Michurina TV Ferraro F Mesenchymal and haematopoietic stem cells form a unique bone marrow niche Nature 2010 466 7308 829 834 10.1038/nature09262 20703299
17. Li W Wang F Guo R Bian Z Song Y Targeting macrophages in hematological malignancies: recent advances and future directions J Hematol Oncol 2022 15 1 110 10.1186/s13045-022-01328-x 35978372
18. Li W Wang Y Zhao H Identification and transcriptome analysis of erythroblastic island macrophages Blood 2019 134 5 480 491 10.1182/blood.2019000430 31101625
19. Li W Wang Y Chen L An X Erythroblast island macrophages: recent discovery and future perspectives Blood Sci 2019 1 1 61 64 10.1097/BS9.0000000000000017 35402789
20. Li W Guo R Song Y Jiang Z Erythroblastic island macrophages shape normal erythropoiesis and drive associated disorders in erythroid hematopoietic diseases Front Cell Dev Biol 2021 8 1858 10.3389/fcell.2020.613885
21. Chow A Lucas D Hidalgo A Bone marrow CD169+ macrophages promote the retention of hematopoietic stem and progenitor cells in the mesenchymal stem cell niche J Exp Med 2011 208 2 261 271 10.1084/jem.20101688 21282381
22. Naik S Larsen SB Cowley CJ Fuchs E Two to tango: dialog between immunity and stem cells in health and disease Cell 2018 175 4 908 920 10.1016/j.cell.2018.08.071 30388451
23. Zhao H-Y Zhang Y-Y Xing T M2 macrophages, but not M1 macrophages, support megakaryopoiesis by upregulating PI3K-AKT pathway activity Signal Transduct Target Ther 2021 6 1 234 10.1038/s41392-021-00627-y 34140465
24. Chen Y Xiang J Qian F Epo receptor signaling in macrophages alters the splenic niche to promote erythroid differentiation Blood 2020 136 2 235 246 10.1182/blood.2019003480 32350523
25. Borges MD Sesti-Costa R de Albuquerque DM Monocytes from patients with polycythemia vera express molecules related to stress erythropoiesis and have increased erythrocyte phagocytosis Blood 2021 138 1466 10.1182/blood-2021-152694
26. Wakahashi K Minagawa K Kawano Y Vitamin D receptor-mediated skewed differentiation of macrophages initiates myelofibrosis and subsequent osteosclerosis Blood 2019 133 15 1619 1629 10.1182/blood-2018-09-876615 30718230
27. Molitor DCA Boor P Buness A Macrophage frequency in the bone marrow correlates with morphologic subtype of myeloproliferative neoplasm Ann Hematol 2021 100 1 97 104 10.1007/s00277-020-04304-y 33104881
28. Van Egeren D Kamaz B Liu S Transcriptional differences between JAK2-V617F and wild-type bone marrow cells in patients with myeloproliferative neoplasms Exp Hematol. 2021 107 14 19 10.1016/j.exphem.2021.12.364 34921959
29. Cao W Fan W Wang F GM-CSF impairs erythropoiesis by disrupting erythroblastic island formation via macrophages J Transl Med 2022 20 1 11 10.1186/s12967-021-03214-5 34980171
30. Xu L Yan M Long J Identification of macrophage correlated biomarkers to predict the prognosis in patients with intrahepatic cholangiocarcinoma Front Oncol 2022 12 967982 10.3389/fonc.2022.967982 36158683
31. Wang Y Li W Schulz VP Impairment of human terminal erythroid differentiation by histone deacetylase 5 deficiency Blood 2021 138 17 1615 1627 10.1182/blood.2020007401 34036344
32. Xu L Yan M Li W Matrix metalloproteinase 1 is a poor prognostic biomarker for patients with hepatocellular carcinoma 2022
33. Yang H Yan M Li W Xu L SIRPα and PD1 expression on tumor-associated macrophage predict prognosis of intrahepatic cholangiocarcinoma J Transl Med 2022 20 1 140 10.1186/s12967-022-03342-6 35317832
34. Li W Li T Sun C Identification and prognostic analysis of biomarkers to predict the progression of pancreatic cancer patients Mol Med 2022 28 1 43 10.1186/s10020-022-00467-8 35428170
35. Schinke C Poos AM Bauer M Characterizing the role of the immune microenvironment in multiple myeloma progression at a single-cell level Blood Adv 2022 6 22 5873 5883 10.1182/bloodadvances.2022007217 35977111
36. Shi H He H Cui L Transcriptomic landscape of circulating mononuclear phagocytes in Langerhans cell histiocytosis at the single-cell level Blood 2021 138 14 1237 1248 10.1182/blood.2020009064 34132762
37. Chasis JA Mohandas N Erythroblastic islands: niches for erythropoiesis Blood 2008 112 3 470 478 10.1182/blood-2008-03-077883 18650462
38. Fabriek BO Polfliet MMJ Vloet RPM The macrophage CD163 surface glycoprotein is an erythroblast adhesion receptor Blood 2007 109 12 5223 5229 10.1182/blood-2006-08-036467 17353345
39. Wei Q Boulais PE Zhang D Maea expressed by macrophages, but not erythroblasts, maintains postnatal murine bone marrow erythroblastic islands Blood 2019 133 11 1222 1232 10.1182/blood-2018-11-888180 30674470
40. Liao C SandeepPrabhu K Paulson RF Monocyte-derived macrophages expand the murine stress erythropoietic niche during the recovery from anemia Blood 2018 132 24 2580 2593 10.1182/blood-2018-06-856831 30322871
41. Kusakabe M Hasegawa K Hamada M c-Maf plays a crucial role for the definitive erythropoiesis that accompanies erythroblastic island formation in the fetal liver Blood 2011 118 5 1374 1385 10.1182/blood-2010-08-300400 21628412
42. Chow A Huggins M Ahmed J CD169 + macrophages provide a niche promoting erythropoiesis under homeostasis and stress Nat Med 2013 19 4 429 436 10.1038/nm.3057 23502962
43. Akilesh H M Buechler M B Duggan J M Chronic TLR7 and TLR9 signaling drives anemia via differentiation of specialized hemophagocytes Science 2019 363 6423 eaao5213 10.1126/science.aao5213 30630901
44. Si Y Tsou C-L Croft K Charo IF CCR2 mediates hematopoietic stem and progenitor cell trafficking to sites of inflammation in mice J Clin Invest 2010 120 4 1192 1203 10.1172/JCI40310 20234092
45. Ziegler-Heitbrock L The CD14+ CD16+ blood monocytes: their role in infection and inflammation J Leukoc Biol 2007 81 3 584 592 10.1189/jlb.0806510 17135573
46. Galvan MD Greenlee-Wacker MC Bohlson SS C1q and phagocytosis: the perfect complement to a good meal J Leukoc Biol 2012 92 3 489 497 10.1189/jlb.0212099 22715140
47. Duchateau PN Pullinger CR Cho MH Eng C Kane JP Apolipoprotein L gene family: tissue-specific expression, splicing, promoter regions; discovery of a new gene J Lipid Res 2001 42 4 620 630 10.1016/S0022-2275(20)31171-8 11290834
48. Wu Z Xu J Liang C Emerging roles of the solute carrier family in pancreatic cancer Clin Transl Med 2021 11 3 e356 10.1002/ctm2.356 33783998
49. Laskowska J Szczepanek J Tretyn A Styczynski J Identification of the genomic rearrangements associated with the ex vivo resistance to anthracyclines in childhood acute leukemias Blood 2015 126 23 4997 10.1182/blood.V126.23.4997.4997
50. Hou M Regulation of megakaryopoiesis by bone marrow macrophage polarization Blood Sci 2021 03 04 149 150 10.1097/BS9.0000000000000090
51. Yang X Chitalia SV Matsuura S Ravid K Integrins and their role in megakaryocyte development and function Exp Hematol 2022 106 31 39 10.1016/j.exphem.2021.11.007 34910941
52. Lane WJ Dias S Hattori K Stromal-derived factor 1–induced megakaryocyte migration and platelet production is dependent on matrix metalloproteinases Blood 2000 96 13 4152 4159 10.1182/blood.V96.13.4152 11110686
53. Cecchetti L Tolley ND Michetti N Megakaryocytes differentially sort mRNAs for matrix metalloproteinases and their inhibitors into platelets: a mechanism for regulating synthetic events Blood 2011 118 7 1903 1911 10.1182/blood-2010-12-324517 21628401
54. Page-McCaw A Serano J Santé JM Rubin GM Drosophila matrix metalloproteinases are required for tissue remodeling, but not embryonic development Dev Cell 2003 4 1 95 106 10.1016/S1534-5807(02)00400-8 12530966
55. Schepers K Pietras EM Reynaud D Myeloproliferative neoplasia remodels the endosteal bone marrow niche into a self-reinforcing leukemic niche Cell Stem Cell 2013 13 3 285 299 10.1016/j.stem.2013.06.009 23850243
56. Koschmieder S Chatain N Role of inflammation in the biology of myeloproliferative neoplasms Blood Rev 2020 42 100711 10.1016/j.blre.2020.100711 32505517
57. Curto-Garcia N Harrison C McLornan DP Bone marrow niche dysregulation in myeloproliferative neoplasms Haematologica 2020 105 5 1189 1200 10.3324/haematol.2019.243121 32241851
58 Hasselbalch HC Thomassen M Riley CH Whole blood transcriptional profiling reveals deregulation of oxidative and antioxidative defence genes in myelofibrosis and related neoplasms. Potential implications of downregulation of Nrf2 for genomic instability and disease progression PLoS One 2014 9 11 e112786 10.1371/journal.pone.0112786 25397683
59. Andersen M Sajid Z Pedersen RK Mathematical modelling as a proof of concept for MPNs as a human inflammation model for cancer development PLoS ONE 2017 12 8 e0183620 10.1371/journal.pone.0183620 28859112
60. Mendez Luque LF Blackmon AL Ramanathan G Fleischman AG Key role of inflammation in myeloproliferative neoplasms: instigator of disease initiation, progression and symptoms Curr Hematol Malig Rep 2019 14 3 145 153 10.1007/s11899-019-00508-w 31119475
61. Sun T Ju M Dai X Multilevel defects in the hematopoietic niche in essential thrombocythemia Haematologica 2020 105 3 661 673 10.3324/haematol.2018.213686 31289202
62. Devisscher L Verhelst X Colle I Van Vlierberghe H Geerts A The role of macrophages in obesity-driven chronic liver disease J Leukoc Biol 2016 99 5 693 698 10.1189/jlb.5RU0116-016R 26936934
63. Hénaut L Candellier A Boudot C New insights into the roles of monocytes/macrophages in cardiovascular calcification associated with chronic kidney disease Toxins (Basel). 2019 11 9 529 10.3390/toxins11090529 31547340
64. Zhao W Beers DR Hooten KG Characterization of gene expression phenotype in amyotrophic lateral sclerosis monocytes JAMA Neurol 2017 74 6 677 685 10.1001/jamaneurol.2017.0357 28437540
65. Tu J Huang W Zhang W Mei J Zhu C A Tale of two immune cells in rheumatoid arthritis: the crosstalk between macrophages and T cells in the synovium Front Immunol 2021 12 655477 10.3389/fimmu.2021.655477 34220809
66. Bassan VL Barretto GD de Almeida FC Philadelphia-negative myeloproliferative neoplasms display alterations in monocyte subpopulations frequency and immunophenotype Med Oncol 2022 39 12 223 10.1007/s12032-022-01825-6 36175590
67. Ramos P Casu C Gardenghi S Macrophages support pathological erythropoiesis in polycythemia vera and β-thalassemia Nat Med 2013 19 4 437 445 10.1038/nm.3126 23502961
68. Fang S Nurmi H Heinolainen K Critical requirement of VEGF-C in transition to fetal erythropoiesis Blood 2016 128 5 710 720 10.1182/blood-2015-12-687970 27343251
69. Fang S Chen S Nurmi H VEGF-C protects the integrity of the bone marrow perivascular niche in mice Blood 2020 136 16 1871 1883 10.1182/blood.2020005699 32842144
70. Young A Linehan E Hams E Cutting edge: suppression of GM-CSF expression in murine and human T cells by IL-27 J Immunol 2012 189 5 2079 2083 10.4049/jimmunol.1200131 22837488
71. Mei Y Zhao B Basiorka AA Age-related inflammatory bone marrow microenvironment induces ineffective erythropoiesis mimicking del(5q) MDS Leukemia 2018 32 4 1023 1033 10.1038/leu.2017.326 29263441
72. Müller P Baldauf CK Haage TR Anti-inflammatory treatment in MPN: targeting TNFR1 and TNFR2 in JAK2-V617F–induced disease Blood Adv 2021 5 23 5349 5359 10.1182/bloodadvances.2021004438 34592754
73. Hodrea J Majai G Doró Z The glucocorticoid dexamethasone programs human dendritic cells for enhanced phagocytosis of apoptotic neutrophils and inflammatory response J Leukoc Biol 2012 91 1 127 136 10.1189/jlb.0511243 22028334
74. Mukundan L Odegaard JI Morel CR PPAR-δ senses and orchestrates clearance of apoptotic cells to promote tolerance Nat Med 2009 15 11 1266 1272 10.1038/nm.2048 19838202
75. Monajemi H Fontijn RD Pannekoek H Horrevoets AJG The apolipoprotein L gene cluster has emerged recently in evolution and is expressed in human vascular tissue Genomics 2002 79 4 539 546 10.1006/geno.2002.6729 11944986
76. Machlus KR Johnson KE Kulenthirarajan R CCL5 Released from platelets increases megakaryocyte maturation and proplatelet formation Blood 2015 126 23 1162 10.1182/blood.V126.23.1162.1162
77. Machlus KR Johnson KE Kulenthirarajan R CCL5 derived from platelets increases megakaryocyte proplatelet formation Blood 2016 127 7 921 926 10.1182/blood-2015-05-644583 26647394
78. Noetzli L Parvin S Battinelli E Machlus K CCL5 Enhances megakaryocyte differentiation and development Blood 2018 132 522 10.1182/blood-2018-99-117518
79. Stone AP Nascimento TF Barrachina MN The bone marrow niche from the inside out: how megakaryocytes are shaped by and shape hematopoiesis Blood 2022 139 4 483 491 10.1182/blood.2021012827 34587234
80. Collinson RJ Mazza-Parton A Fuller KA Gene expression of CXCL1 (GRO-α) and EGF by platelets in myeloproliferative neoplasms HemaSphere 2020 4 6 e490 10.1097/HS9.0000000000000490 33134870
81. Øbro NF Grinfeld J Belmonte M Longitudinal cytokine profiling identifies GRO-α and EGF as potential biomarkers of disease progression in essential thrombocythemia HemaSphere 2020 4 3 e371 10.1097/HS9.0000000000000371 32647796
82. Mannello F Medda V Differential expression of MMP-2 and MMP-9 activity in megakaryocytes and platelets Blood 2011 118 24 6470 6471 10.1182/blood-2011-07-366195 22161853
83. Varricchio L, Iancu-Rubin C, Upadhyaya B et al (2021) TGF-β1 protein trap AVID200 beneficially affects hematopoiesis and bone marrow fibrosis in myelofibrosis. JCI insight. 6(18)
84. Arciprete F Zingariello M Mazzarini M The immature morphology of the megakaryocytes present in the bone marrow of patients with myelofibrosis reflects changes in the frequency of functionally distinctive subpopulations Blood 2022 140 Supplement 1 12148 12149 10.1182/blood-2022-157526
85. Martinaud C Desterke C Konopacki J Osteogenic potential of mesenchymal stromal cells contributes to primary myelofibrosis Cancer Res 2015 75 22 4753 4765 10.1158/0008-5472.CAN-14-3696 26404004
86. Decker M Martinez-Morentin L Wang G Leptin-receptor-expressing bone marrow stromal cells are myofibroblasts in primary myelofibrosis Nat Cell Biol 2017 19 6 677 688 10.1038/ncb3530 28481328
87. Barone M Catani L Ricci F The role of circulating monocytes and JAK inhibition in the infectious-driven inflammatory response of myelofibrosis Oncoimmunology 2020 9 1 1782575 10.1080/2162402X.2020.1782575 32923146
88. Shinoda D Nakajima-Takagi Y Oshima M Insufficiency of non-canonical PRC1 synergizes with JAK2V617F in the development of myelofibrosis Leukemia 2022 36 2 452 463 10.1038/s41375-021-01402-2 34497325
89. Lev PR Salim JP Kornblihtt LI PDGF-A, PDGF-B, TGFβ, and bFGF mRNA levels in patients with essential thrombocythemia treated with anagrelide Am J Hematol 2005 78 2 155 157 10.1002/ajh.20268 15682418
90. Kramer F Dernedde J Mezheyeuski A Platelet-derived growth factor receptor β activation and regulation in murine myelofibrosis Haematologica 2020 105 8 2083 2094 10.3324/haematol.2019.226332 31672904
91. Ding L Saunders TL Enikolopov G Morrison SJ Endothelial and perivascular cells maintain haematopoietic stem cells Nature 2012 481 7382 457 462 10.1038/nature10783 22281595
92. Greenbaum A Hsu Y-MS Day RB CXCL12 in early mesenchymal progenitors is required for haematopoietic stem-cell maintenance Nature 2013 495 7440 227 230 10.1038/nature11926 23434756
93. Ding L Morrison SJ Haematopoietic stem cells and early lymphoid progenitors occupy distinct bone marrow niches Nature 2013 495 7440 231 235 10.1038/nature11885 23434755
94. Wang X Cho SY Hu CS C-X-C motif chemokine 12 influences the development of extramedullary hematopoiesis in the spleens of myelofibrosis patients Exp Hematol 2015 43 2 100 109.e1 10.1016/j.exphem.2014.10.013 25461253
95. Kleppe M Kwak M Koppikar P JAK–STAT pathway activation in malignant and nonmalignant cells contributes to MPN pathogenesis and therapeutic response Cancer Discov 2015 5 3 316 331 10.1158/2159-8290.CD-14-0736 25572172
96. Beitzen-Heineke A Berenbrok N Waizenegger J AXL inhibition represents a novel therapeutic approach in BCR-ABL negative myeloproliferative neoplasms HemaSphere 2021 5 9 e630 10.1097/HS9.0000000000000630 34396051
97. Xu M Bruno E Chao J Constitutive mobilization of CD34+ cells into the peripheral blood in idiopathic myelofibrosis may be due to the action of a number of proteases Blood 2005 105 11 4508 4515 10.1182/blood-2004-08-3238 15705794
98. Dutta P Hoyer FF Grigoryeva LS Macrophages retain hematopoietic stem cells in the spleen via VCAM-1 J Exp Med 2015 212 4 497 512 10.1084/jem.20141642 25800955
99. Qiu J Salama ME Hu CS The characteristics of vessel lining cells in normal spleens and their role in the pathobiology of myelofibrosis Blood Adv 2018 2 10 1130 1145 10.1182/bloodadvances.2017015073 29776986
100. Kohyama M Ise W Edelson BT Role for Spi-C in the development of red pulp macrophages and splenic iron homeostasis Nature 2009 457 7227 318 321 10.1038/nature07472 19037245
101. Haldar M Kohyama M So AYL Heme-mediated SPI-C induction promotes monocyte differentiation into iron-recycling macrophages Cell 2014 156 6 1223 1234 10.1016/j.cell.2014.01.069 24630724
102. Tefferi A, Guglielmelli P, Larson DR et al (2014) Long-term survival and blast transformation in molecularly annotated essential thrombocythemia, polycythemia vera, and myelofibrosis. Blood 124(16):2507–13; quiz 2615
103. Silver RT Life, genes, and death in Ph- MPNs Blood 2014 124 16 2471 2472 10.1182/blood-2014-08-590364 25323682
|
PMC010xxxxxx/PMC10219672.txt |
==== Front
Kidney Int Rep
Kidney Int Rep
Kidney International Reports
2468-0249
International Society of Nephrology. Published by Elsevier Inc.
S2468-0249(23)01315-3
10.1016/j.ekir.2023.05.020
Research Letter
Serum Neutralization of Omicron XBB.1.5 in Kidney Transplant Recipients After Bivalent mRNA Booster Vaccination
Pedersen Rune M. 1
Bang Line L. 1
Holm Dorte K. 2
Madsen Lone W. 34
Johansen Isik S. 3
Jensen Thøger G. 1
Justesen Ulrik S. 1
Bistrup Claus 5
Andersen Thomas E. 1∗
1 Department of Clinical Microbiology, Odense University Hospital and Research Unit for Clinical Microbiology, University of Southern Denmark, Odense, Denmark
2 Department of Clinical Immunology, Odense University Hospital and Research Unit for Clinical Immunology, University of Southern Denmark, Odense, Denmark
3 Department of Infectious Diseases, Odense University Hospital and Research Unit for Infectious Diseases, University of Southern Denmark, Odense, Denmark
4 Unit for Infectious Diseases, Department of Medicine, Sygehus Lillebælt, Kolding, Denmark
5 Department of Nephrology, Odense University Hospital and the Nephrology Research Unit, University of Southern Denmark, Odense, Denmark
∗ Correspondence: Thomas E. Andersen, J. B. Winsløws Vej 21, 2nd Floor, 5000 Odense C, Denmark.
26 5 2023
26 5 2023
4 5 2023
22 5 2023
© 2023 International Society of Nephrology. Published by Elsevier Inc.
2023
International Society of Nephrology
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Keywords
bivalent
COVID-19
kidney transplant recipients
vaccine
XBB.1.5
==== Body
pmcImmunosuppression is critical to prevent rejection episodes in kidney transplant recipients (KTRs) and entails a reduced resistance to infections and efficacy of vaccines. The reduced efficacy of vaccines has challenged COVID-19 vaccination strategies for KTRs.1 This patient group displays a lower-than-normal antibody response against the COVID-19 vaccines, which has left a large proportion with low levels of spike antibodies even after several booster doses.2, 3, 4, 5 In addition, SARS-CoV-2 has shown an exceptional ability to adapt to the original vaccines, which has gradually rendered vaccine-induced antibodies less neutralizing.5 Together, this has generated the need for updated COVID-19 vaccines with restored efficacy against newer SARS-CoV-2 subvariants, in particular the Omicron variant of concern, which has dominated the pandemic since late 2021. During 2022, Pfizer-BioNTech adapted their original BNT162b2 mRNA COVID-19 vaccine to target the Omicron BA.1 and later the BA.4/BA.5 subvariants, in addition to the Wuhan-1 index strain. The resulting bivalent mRNA vaccines were approved during autumn 2022 and recommended to vulnerable patients including KTRs. Since then, the more transmissible Omicron subvariant XBB.1.5 has rapidly replaced the BA.1/BA.4/BA5 subvariants, in particular in the United States where XBB.1.5 prevalence is 88% at the time of writing (April 10, 2023).6 Consequently, this has raised concerns about whether the updated vaccines still provide the intended protection against COVID-19 in this patient group. To answer this question, we analyzed the capacity of KTRs to neutralize authentic Omicron XBB.1.5 after receiving the bivalent vaccine from Pfizer-BioNTech.
Sera from 46 KTRs collected 1 month after receiving a BNT162b2 bivalent mRNA vaccination from Pfizer-BioNTech as a fifth dose were tested for neutralization of SARS-CoV-2 Omicron XBB.1.5. In addition, serum-neutralization against ancestral SARS-CoV-2, BA.1, BA.5, and XBB.1.5 were measured in a subgroup of representative KTRs (n = 21) before and after receiving the bivalent vaccine. For reference, sera from healthy individuals (n = 26) and KTRs (n = 25) were tested against XBB.1.5 one month after the third BNT162b2 monovalent vaccination. All participants were nucleocapsid IgG antibody seronegative, indicating not previously infected. All groups were matched on the basis of time from vaccination to serum collection (median 33 days, interquartile range 30–39 days). The KTRs were matched according to age (median 67 years, interquartile range 59–73 years) whereas the healthy controls were younger (median 48 years, interquartile range 38–57 years). The neutralization capacity was measured using a microneutralization assay as recently described.7 In this assay, cultured Vero E6 cells are challenged with authentic SARS-CoV-2 isolates in 2-fold serially diluted serum. The highest dilution protecting more than 90% of the cells from virus-induced cytopathic effects is designated the ED90 titer. A dilution of 10 was used as a threshold of a neutralizing response, because a titer of 8.8 in microneutralization assays has been suggested to indicate real-world protection against Omicron.8 The SARS-CoV-2 strains were clinical isolates identified by whole-genome sequencing. The samples were analyzed for spike antibodies using the Liaison TrimericS IgG Quantitative immunoassay (Diasorin, Saluggia, Italy). For details on patient groups, SARS-CoV-2 strains, immunoassays, and statistics, see Supplementary Material. A flow chart showing selection criteria for KTRs is shown in Supplementary Figure S1.
Clinical characteristics of the KTRs are shown in Table 1 and the spike antibody levels in Figure 1 a. Overall, bivalent vaccination increased the fraction of KTRs with above-threshold neutralizing capacity against XBB.1.5 from 52% (11/21) to 76% (35/46) (P = 0.09, Fisher Exact test) and increased the geometric mean titer (GMT) from 9.1 (95% confidence interval [CI] 6.6–12.5) to 22.2 (95% CI 15.4–32.1) (P = 0.0025, Mann-Whitney test). This was significantly more than the responding fractions and the GMT of both KTRs (4%, GMT = 5.1) and healthy controls (35%, GMT = 6.9) after the third monovalent booster (P < 0.0001 and 0.0002, respectively, Mann-Whitney test) (Figure 1b). Spike antibody levels after bivalent vaccination (geometric mean 3654 binding antibody units/ml, 95% CI 1956–6825 binding antibody units/ml) correlated with the XBB.1.5 neutralization capacity (r = 0.80, P < 0.0001, Spearman’s correlation, Figure 1a and b). After bivalent vaccination the serum-GMT for the KTR subgroup (n = 21), were for ancestral SARS-CoV-2: 154.8 (95% CI 75.1–318.9); for BA.1: 77.4 (95% CI 35.0–171.4), for BA.5: 52.1 (95% CI 25.6–105.8), and for XBB.1.5: 17.5 (95% CI 9.7–31.8). These differences were statistically significant (P < 0.0001, Friedmans test). However, the increases in GMT induced against these subvariants following bivalent vaccination were not significantly different (GMT increase x1.8-x2.6, P = 0.05, Friedmans test, Figure 1c). Neither induction nor maintenance immunosuppression were associated with the bivalent vaccine response (P = 0.852 and P = 0.122, respectively, Fisher exact test).Table 1 Characteristics of kidney transplant recipients
Demographic characteristics Monovalent (3rd dose)a Bivalent (5th dose)b
Numbers (%) 25 (35) 46 (65)
Age Y (IQR) 65 (57–74) 67 (59–72)
Female (%) 11 (44) 19 (41)
TX characteristics
Time from TX Y (IQR) 6 (3–12) 9 (6–16)
TX number
First TX (%) 20 (80) 37 (80)
Second TX (%) 5 (20) 8 (17)
Third TX (%) 0 (0) 1 (2)
Deceased donor (%) 17 (68) 34 (74)
Induction
Rituximab (%) 0 (0) 1 (2)
anti-CD25 (%) 13 (52) 35 (76)
anti-CD25 + Rituxmab (%) 2 (8) 3 (7)
Thymoglobuline (%) 6 (24) 3 (7)
Thymoglobuline + Rituxmab (%)
Unknown 3 (12)
1 (4) 2 (4)
2 (4)
Maintenance
Tacrolimus (%) 19 (76) 36 (78)
Ciclosporin A (%) 4 (16) 9 (20)
MMF (%) 24 (96) 39 (85)
Azathioprine (%) 1 (4) 6 (13)
Steroids (%) 5 (20) 8 (17)
Renal function
eGFR ml/min (IQR) 41 (25–62) 45 (35–68)
Underlying disease
cNon-immune disease (%) 10 (40) 25 (54)
dImmune disease (%) 9 (36) 13 (28)
Diabetes mellitus (%) 2 (8) 4 (9)
Unknown (%) 4 (16) 4 (9)
eGFR, estimated glomerular filtration rate; IQR, interquartile range; MMF, mycophenolate mofetil; N/A, not applicable; TX, transplant; Y, years.
a KTRs who received 3 vaccine doses with the monovalent BNT162b2 (Pfizer-BioNTech).
b KTRs who received 4 monovalent BNT162b2 vaccine doses followed by 1 bivalent BNT162b2 (Pfizer-BioNTech) mRNA vaccination.
c Immune disease designate diseases such as glomerulonefritis, systemic lupus, ANCA associated vasculitis etc.
d Nonimmune disease designate diseases such as cystic kidney diseases, Alports disease, urinary outlet obstruction etc.
Figure 1 Antispike antibody levels and neutralization capacity in kidney transplant recipients against SARS-CoV-2 Omicron subvariants after Pfizer-BioNTech bivalent mRNA vaccination. (a) Anti-SARS-CoV-2 spike specific IgG antibody levels in kidney transplant recipients (KTRs) and healthy controls after the third Pfizer-BioNTech monovalent mRNA vaccine booster (Post 3rd vac) and before and after Pfizer-BioNTech bivalent mRNA vaccination (Pre 5th vac and Post 5th vac, respectively). Bars indicate geometric mean and error bars indicate 95% confidence interval. The red dotted line indicates the threshold of seropositivity as provided by the manufacturer. The black dotted line indicates lower level of detection. Controls: healthy controls. The 4 groups were initially compared using 1-way ANOVA followed by one-to-one comparison using the unpaired t-test. The antibody levels in healthy controls and in KTRs after the third vaccination were extracted from our previous study.5 (b) Virus neutralization titers of authentic XBB.1.5 measured as effective dilution 90% (ED90) using sera from KTRs and healthy controls at the indicated levels of vaccination. Bars indicate geometric mean titers (titers <10 counted as 5 and ≥1280 as 1280) and error bars indicate 95% confidence interval. The red dotted line indicates the neutralizing threshold titer of 1:8.8. Controls: healthy controls. The 4 groups were initially compared using the Kruskal-Wallis test followed by one-to-one comparison using the Mann-Whitney test. (c) ED90 titers of KTR sera (n = 21) against ancestral SARS-CoV-2, BA.1, BA.5, and XBB.1.5 Omicron subvariants. The bars indicate geometric mean titers and error bars indicate 95% confidence interval. Changes in neutralization capacity in individual patients from before to after the bivalent vaccination are indicated with gray lines. The 4 groups were initially compared using the Friedman test followed by one-to-one comparison using the Wilcoxon test. The red dotted line indicates the neutralizing threshold titer of 1:8.8. Fold differences between groups are indicated together with levels of significance: ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001. The percentage of samples with antibody levels (a) or neutralizing titers (b and c) above their respective thresholds are indicated above each column. BAU, binding antibody units.
We recently found that the Pfizer-BioNTech bivalent mRNA vaccine largely failed to protect immune-compromised cancer patients against XBB.1.5.7 Similarly, a considerable fraction of the KTRs analyzed here remained below neutralization threshold after receiving a bivalent vaccine. Still, compared to younger healthy individuals and KTRs after 3 monovalent BNT162b2 doses, the bivalent boosted KTRs were overall better protected against XBB.1.5. This is encouraging because KTRs throughout the pandemic have struggled with evoking a protective humoral immune response after vaccination as compared to healthy controls.1, 2, 3, 4, 5 Moreover, our data show that bivalent vaccination, as intended, elicits a neutralizing humoral response across both new and old SARS-CoV-2 variants in the KTRs. The study is limited by the small sample size, groups are not fully comparable with respect to treatment history, no KTR group vaccinated with a fourth or fifth dose of the monovalent vaccine were included, and an ED90 threshold value indicating real-world protection against XBB.1.5 is not yet available. Overall, our data indicate that KTRs benefit from bivalent vaccination, but the most immune deficient may require additional boosters to reach a neutralizing humoral response.
Disclosure
All the authors declared no competing interests.
Supplementary Material
Supplementary File (PDF)
Patient Groups.
SARS-CoV-2 strains, Immunoassays, Statistics.
Figure S1. Flow chart showing selection criteria for KTRs.
Acknowledgments
We thank the staff at the Department of Clinical Microbiology, Odense University Hospital for the handling of patient samples.
Data Availability Statement
The data underlying this article will be shared on reasonable request to the corresponding author.
Supplementary File (PDF)
Patient Groups.
SARS-CoV-2 strains, Immunoassays, Statistics.
Figure S1. Flow chart showing selection criteria for KTRs.
==== Refs
References
1 Prendecki M. Willicombe M. SARS-CoV-2 vaccine strategies in kidney transplant recipients Lancet Infect Dis 23 2023 263 264 10.1016/S1473-3099(22)00666-1 36354033
2 Pedersen R.M. Bang L.L. Tornby D.S. The SARS-CoV-2-neutralizing capacity of kidney transplant recipients 4 weeks after receiving a second dose of the BNT162b2 vaccine Kidney Int 100 2021 1129 1131 10.1016/j.kint.2021.09.006 34547366
3 Jurdi A.A. Gassen R.B. Borges T.J. Suboptimal antibody response against SARS-CoV-2 Omicron variant after third dose of mRNA vaccine in kidney transplant recipients Kidney Int 101 2022 1282 1286 10.1016/j.kint.2022.04.009 35429496
4 Caillard S. Thaunat O. Benotmane I. Masset C. Blancho G. Antibody response to a fourth messenger RNA COVID-19 vaccine dose in kidney transplant recipients: a case series Ann Intern Med 175 2022 455 456 10.7326/L21-0598 35007148
5 Pedersen R.M. Bang L.L. Tornby D.S. Serum neutralization of Omicron BA.5, BA.2 and BA.1 in Triple Vaccinated Kidney Transplant Recipients Kidney Int Rep 8 2022 667 671 10.1016/j.ekir.2022.12.004 36532715
6 Hasell J. Mathieu E. Beltekian D. A cross-country database of COVID-19 testing Sci Data 7 2020 345 10.1038/s41597-020-00688-8 33033256
7 Ehmsen S. Pedersen R.M. Bang L.L. BQ.1.1, XBB.1, and XBB.1.5 neutralization after bivalent mRNA COVID-19 booster in patients with cancer Cancer Cell 41 2023 649 650 10.1016/j.ccell.2023.02.003 36804967
8 Cheng S.M.S. Mok C.K.P. Leung Y.W.Y. Neutralizing antibodies against the SARS-CoV-2 Omicron variant BA.1 following homologous and heterologous CoronaVac or BNT162b2 vaccination Nat Med 28 2022 486 489 10.1038/s41591-022-01704-7 35051989
|
PMC010xxxxxx/PMC10225284.txt |
==== Front
Int Rev Educ
Int Rev Educ
International Review of Education. Internationale Zeitschrift Fur Erziehungswissenschaft. Revue Internationale De Pedagogie
0020-8566
1573-0638
Springer Netherlands Dordrecht
10002
10.1007/s11159-023-10002-4
Original Paper
An exploratory study to understand faculty members’ perceptions and challenges in online teaching
http://orcid.org/0000-0001-7866-1675
Mulla Tausif tausif@westford.org.uk
1Tausif Mulla
is a senior lecturer & course leader at Westford University College in the UAE. He has a postgraduate diploma in business management and a master’s degree in business administration. Currently, he is pursuing his PhD from Aligarh Muslim University, India. He teaches general management to undergraduates and master’s level students. His areas of research include consumer behaviour towards entertainment and improving quality of teaching and learning in online education.
http://orcid.org/0000-0002-2753-3136
Munir Sufia sufia@westford.org.uk
2Sufia Munir
is Assistant Dean at Westford University College in the UAE. She has a bachelor’s degree in fashion and apparel design and a master’s degree in business administration. Currently, she is pursuing her PhD from the University of Salford, UK. She teaches general management and fashion design courses. Her areas of research interest are “Sustainability in Business” and “Teaching and Learning in the Virtual World”.
http://orcid.org/0009-0004-7946-3266
Mohan Vivek vivek@westford.org.uk
3Vivek Mohan
is the Dean at Exceed College Westford Education Group, he oversees student and academic affairs for executive education and has several years of experience in leading, designing and delivering programmes in the higher education sector. His research interests include Expatriate performance, Transmedia storytelling, Sustainability, and online learning and teaching in higher education.
1 Faculty of Business Management, Westford University College, Sharjah, UAE
2 Faculty of Business Management & Fashion Design, Westford University College, Sharjah, UAE
3 Westford Education Group, Exeed College, Sharjah, UAE
29 5 2023
2023
69 1-2 7399
28 4 2023
© UNESCO Institute for Lifelong Learning and Springer Nature B.V. 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
The government of the United Arab Emirates (UAE) considers technology to be one of the main pillars of its vision for moving towards a knowledge-based society. Due to several factors such as globalisation, demand for information technology infrastructure and COVID-19 lockdowns, e-learning has become a popular method of delivery across higher education institutions in the UAE. In a first step, the authors of this article conducted a systematic review of existing literature (49 items published between 1999 and 2020). They found that the existing literature on online learning predominantly focuses on student-specific challenges, while there is still a dearth of published work covering faculty members’ specific challenges in facilitating online learning in the UAE. The second part of this exploratory study drew on stakeholders’ reflections of several years of designing and delivering online courses, analysing faculty members’ perspectives on online teaching and learning in the UAE. The authors present their qualitative research, which involved open-ended semi-structured interviews with 15 faculty members, followed by a thematic analysis of their responses using NVivo 12 pro software. The most critical themes which emerged were learners’ expectations, culture, perception, pedagogy and technology. The article also reveals how these topics contribute to the various strategies for seamless adoption and delivery of online education in the UAE.
Résumé
Une étude exploratoire pour comprendre les perceptions et les défis des enseignants dans l'enseignement en ligne – Le gouvernement des Émirats Arabes Unis (EAU) considère la technologie comme l'un des principaux piliers de son projet d'évolution vers une société axée sur la connaissance. En raison de plusieurs facteurs tels que la mondialisation, la demande d'infrastructures pour les technologies de l’information et les confinements dus au COVID-19, l'apprentissage en ligne est devenu une méthode d'enseignement populaire dans les établissements d'enseignement supérieur aux EAU. Dans un premier temps, les auteurs de cet article ont procédé à une revue systématique de la littérature existante (49 articles publiés entre 1999 et 2020). Ils ont constaté que la littérature existante sur l'apprentissage en ligne se concentre principalement sur les défis propres aux étudiants, alors qu'il y a encore peu de travaux couvrant les défis spécifiques aux membres du corps enseignant pour faciliter l'apprentissage en ligne aux EAU. La deuxième partie de cette étude exploratoire s'est appuyée sur les réflexions de différents acteurs sur plusieurs années de conception et d'enseignement de cours en ligne, analysant les perspectives des membres du corps enseignant sur l'enseignement et l'apprentissage en ligne aux EAU. Les auteurs présentent leur recherche qualitative, qui comprend des entretiens semi-structurés ouverts avec 15 membres du corps enseignant, suivis d'une analyse thématique de leurs réponses à l'aide du logiciel NVivo 12 pro. Les thèmes les plus importants qui ont émergé ont été les attentes des apprenants, la culture, la perception, la pédagogie et la technologie. L'article révèle également comment ces sujets contribuent aux diverses stratégies pour une adoption et une diffusion harmonieuses de l'enseignement en ligne aux EAU.
Keywords
Online education
Online teaching
E-learning
Online learning
Online teaching staff
Teacher-specific challenges
issue-copyright-statement© UNESCO Institute for Lifelong Learning 2023
==== Body
pmcIntroduction
Based on the current trends in online education, it is apparent that one of the reasons driving most higher education institutions (HEIs) to offer online programmes is to ensure that they stay competitive. Online courses have seen a rapid increase in enrolments, and HEIs have shown keen interest in providing quality education to satisfy the demand. According to Bilquis Ferdousi (2016), the economic downturn [2007–2009] prompted higher interest in online education, since it was recognised as a more affordable means of learning. Also, the ongoing situation presented by COVID-19 has revealed that the importance of online learning is likely to increase in the future as the world tries to keep going during phases of extreme disruption. According to figures released by the United Nations Educational, Scientific and Cultural Organization (UNESCO n.d.), around 1.3 billion learners around the world could not attend a school or university as of 23 March 2020 due to the lockdowns caused by the spread of COVID-19 (McCarthy 2020, referring to UNESCO n.d.). The pace of these closures and the rapid push towards online learning further emphasised its importance. Brittany Hunt and Beth Oyarzun mention that “every faculty member is going to be delivering education online, and every student is going to be receiving education online” (Hunt and Oyarzun 2020, p. 10).
Since 2009, the United Arab Emirates (UAE) has witnessed incredible growth in the number of internet users which has been possible due to uninterrupted online access and connectivity. This in turn has enabled learners to pursue higher education which was previously out of reach due to visa requirements, exorbitant fees and work commitments. The surge in demand for online courses has raised questions about the quality of education and highlighted the challenges of maintaining the same standards as on-campus delivery (Elison-Bowers et al. 2010). Although many researchers believe that online and on-site courses are equally effective in terms of knowledge provision and acquisition, the tools and strategies implemented in imparting the knowledge and facilitating learning need to be tailored to the mode of instruction (Jacobs 2013). Faculty members teaching courses online face several challenges which include managing diverse content suitable and relevant for the larger student community, mastering technological skills, managing time, and being innovative in delivering the course (Burchum et al. 2007). They have to be extremely aware of learners’ backgrounds to be able to provide an “inclusive classroom environment” (Hunt and Oyarzun 2020, p. 3). In an online satisfaction survey conducted by Elaine Strachota (2003), “learner–content interaction” ranked first as a determinant of student satisfaction, followed by “learner–instructor and learner–technology interaction” (Cole et al. 2014, p. 113; referring to Strachota 2003). These findings emphasise the integral role of faculty members in achieving student satisfaction in online courses.
Purpose of our study
Internationally, there is a paucity of literature on challenges specifically encountered by faculty members, since the majority of available articles are conceptual papers and literature reviews with no data to substantiate the arguments presented. Other studies have a narrow focus, e.g. on challenges related to interpersonal interaction (Mehall 2020), workload (Gregory and Lodge 2015), institutional support (Orr et al. 2009), and design and delivery of online learning (Tham and Werner 2005). Some studies have focused on student-specific challenges in online education (Childs et al. 2005), but have not considered faculty members’ perspective.
This is no different in the UAE, where recent studies on e-learning focus primarily on learners; their perceptions (Awofeso and Bamidele 2017; Vrazalic et al. 2010), interactions (Abulibdeh and Syed Hassan 2011) and satisfaction (Al-hawari and Mouakket 2010; Sher 2009). Other studies are based on identifying factors that impact the adoption of instructional technologies by faculty members teaching computer information sciences (Daouk and Aldalaien 2019) and Islamic teachers’ perception of integrating ICT into teaching in public schools in the UAE (Al-Gumaei et al. 2019). Considering that business is one of the most popular disciplines in higher education in the UAE (UAE MoE 2019; Reynolds and Rizvi 2019), and is offered extensively via the online mode, it is important to understand the challenges encountered by university teaching staff delivering business management courses to students from around the globe. The study we are presenting here intends to bridge this knowledge gap by considering the perspectives of faculty members delivering online courses in the UAE. Understanding their situation will help other academics and practitioners who are facing similar challenges.
Research questions
What are the content-specific challenges faculty members encounter in online delivery?
What are the challenges faculty members encounter in interacting with learners in an online class?
What are the institutional challenges faculty members encounter that impact the delivery of online classes?
Literature review
According to a report entitled “Where to invest now in GCC private education”1 released by the Boston Consulting Group (Hoteit et al. 2018), the UAE’s education market was expected to grow from USD 4.4 billion in 2017 to USD 7.1 billion by 2023. Because of emerging digital technologies in recent years, the higher education sector has witnessed significant changes in teaching and learning practices in the UAE, but as yet few HEIs have joined the bandwagon to provide flexible online learning options to both on-campus and distance learners. In the past, some researchers did find e-learning to be a better option than face-to-face learning, provided the pedagogy of the courses was of high quality and appealed to online learners (Islam et al. 2015). However, recent studies have shown that the unprecedented growth in e-learning has brought significant challenges to educational stakeholders, mainly to teaching staff, since online education delivery has a learning curve (Ferdousi 2016). Coupled with high demand, a new spectrum of challenges has emerged for faculty members, including workload increase. It is evident from the literature on the topic that online teaching required teaching staff to invest “a minimum of 14% more time than traditional instruction, most of which was spent presenting instructional content” (Ruth 2018, p. 15, quoting Tomei 2006).
Achieving student satisfaction is another challenge for teaching staff. According to Rustam Haydarov et al. (2013), in higher education, considerable research efforts are being made to address the attrition and retention rates that are associated with institutional performance. In many countries, public reputation and government funding for educational institutions are directly linked to their ability to retain students. While researchers have highlighted the challenges of student retention and satisfaction in online education over the years, not much has been discussed about faculty members’ struggles with successful online delivery.
Higher education administrators are under tremendous pressure to keep up momentum with an ever-changing e-learning environment, as online classes become more popular and accessible. With the rapid increase in online learning, quality has come under scrutiny. Commissioned by the National Education Association (NEA), the Institute for Higher Education Policy (IHEP) in the United States published a report which provided quality guidance for distance learning institutes (IHEP 2000, 2014). The requirements are divided into seven separate categories: (a) institutional support; (b) course development; (c) teaching/learning; (d) course structure; (e) assistance to students; (f) instructional assistance; and (g) assessment and review (Tham and Werner 2005, referring to IHEP 2000, 2014). In terms of workload, online teaching has been reported to take twice the time in comparison to traditional instruction, making it complex and challenging for faculty members, potentially leading to burnout. Administrators of higher education institutes (HEIs) who try to replicate traditional on-campus teaching methods in the e-learning environment struggle to get the same output. Often, traditional training approaches do not readily translate into e-learning (Elison-Bowers et al. 2010). This can lead to frustration and low satisfaction among both teaching staff and students. According to Joel Hartman et al. (2000), faculty members’ satisfaction and effective student learning are highly correlated.
The role of interaction in online learning
Interaction is critical in online learning. It helps create a sense of community and encourages participation. Interaction also allows instructors to get to know their students and understand their needs. According to Terry Anderson (2008), there are six key components of online learning:Student–student interaction is one of the key components of online learning environments. According to modern constructivist and connectivist theorists, peer-to-peer interaction is essential to researching and developing multiple perspectives. Collaborative learning and student-led teams promote “reciprocal teaching” and build communities of learners. Moreover, research indicates that peer-to-peer interaction is key to effective learning, as it provides an opportunity for learners to discuss and formulate ideas and reflect on their thoughts. Since traditional means of interaction such as meeting in the library and on campus are not possible in an online setting, it becomes the responsibility of faculty members to facilitate peer-to-peer learning, for example by designing quizzes, discussion boards and group activities (Chandra and Palvia 2021). This implies that faculty members must be trained to use information and communication technologies (ICT), since the requirements in an online class are quite different from an on-site one (Lai et al. 2019).
The importance of student–content interaction in formal education and online learning cannot be overstated. The opportunities offered by the internet include interactive content that responds to student behaviour and allows for customising content to meet the needs of each learner.
Student–teacher interaction is supported in a variety of ways through online learning, including text, audio and video communication. Since online courses are less teacher-centric than traditional classroom sessions, learners are more likely to become committed to and engaged in their learning.
Teacher–content interaction examines content created by the teacher, such as learning objects, units of study, complete courses and associated learning activities. Interaction between teachers and course content facilitates the monitoring, planning and updating of course content resources and activities.
Teacher–teacher interaction enhances professional development and allows teachers to support each other. As a result of these interactions, teachers have the opportunity to gain knowledge and discover new things in their subject area as well as within their scholarly community.
Content–content interaction is a new mode of educational interaction which involves content continually updating as it interacts with other automated information sources.
To create solid online learning contexts, these six modes of interaction, along with a learning environment which is conducive to learning, are essential. They serve as the basis for Anderson’s e-learning model. However, as mentioned earlier, despite the growth in e-learning, the current literature lacks a critical perspective on the challenges encountered by teaching staff.
Based on our systematic review of existing literature,2 we identify the challenges encountered by faculty members in online delivery concerning (a) interaction with and among students; (b) content; and (c) institutional support.
Faculty member–student interaction challenges
Traditionally, faculty members are considered to be the “sage on the stage” (King 1993, p. 30) who primarily sets the educational goals and provides most educational content. But in the online learning environment, the faculty member is often described as a “guide on the side” (ibid.). Papia Bawa refers to instructors “unable to keep up or understand the language of the digital native community” as “digital immigrants” (Bawa 2016, p. 6; citing Prensky 2001).
Defining student expectations
With the exponential growth in e-learning, a unique set of challenges need to be addressed. What is needed first of all is a clear definition of instructor performance in the online teaching environment. Faculty members and learners must be prepared for technical problems in the online classroom, and instructors should communicate expectations and policies with the learners before the commencement of the course (Elison-Bowers et al. 2010; Stoffregen et al. 2015). Before or during the beginning of the course, learners may have queries such as will faculty members be available on weekends or after college hours? How soon can the learner expect an answer? Setting these standards will help reduce the number of repetitive e-mails and telephone calls and increase student satisfaction.
Multiple roles played by faculty members
In recent years, universities and colleges have tried to extend the scope of work and career paths for their teaching staff. Today, faculty members are expected to take up roles that go beyond teaching, in both qualitative and quantitative terms. While teaching itself is multi-faceted, the instructor is not only the facilitator but also expected to be the instructional designer, subject-matter expert and course manager (Conceição 2006; Zhen et al. 2008). Furthermore, it is common today for a faculty member to be promoted to “team leader” or “head of learning and development” or other similar designations. These added roles and responsibilities have an impact on the quality of the delivery (Andersson and Gronlund 2009).
Building rapport with online learners
In a typical classroom setting, to build a successful learning environment, a faculty member uses their understanding of the audience, and learners’ reactions are monitored by body language evaluation, verbal responses and eye contact, which is not possible in an online class (McLendon and Albion 2000). For this reason, some institutions have promoted blended learning to personalise the course and establish deeper relationships with the learners. In the virtual environment, a fun, coherent and relaxed learning atmosphere should be built using digital communication tools, but accomplishing this is not easy, since the medium does not detect non-verbal messages, making it challenging for faculty members to build a rapport with the learners.
Cultural factors
A complete understanding of culture, and particularly its value and impact in online courses, is a complex undertaking. Within most learning communities, there is a dominant culture that affects all the other components. In academic settings, culture is often taken for granted by the instructors as well as the management. Researchers like Lejla Vrazalic et al. (2010) discuss linguistic and cultural factors linked to the usage of online learning tools for international students. They observe thatlinguistic factors ten[d] to have more impact on the participants’ actual use of online resources while cultural factors ha[ve] greater influence on their wider educational experience (ibid., p. 3; quoting Hughes 2005).
Moreover, while faculty members who are native speakers do not have to worry about the language of instruction, they may have to “deal with a different teaching/learning culture” with “different expectations of teacher role models and status” (Beaven et al. 2010, p. 16). An important part of this discussion is the cultural background of the learners. Culture does not only affect learning, motivation and satisfaction in a course, but also has an impact on the overall classroom experience. Every individual has a different learning style and expectations, which needs to be considered in e-learning (Ali et al. 2018).
Lack of social interaction among students
A recent study in the Chronicle for Higher Education (Blumenstyk 2019) indicates that dropout rates for distance learners recorded by institutions range from 20% to 50%. However, online course dropout rates are often 10% to 20% higher in distance offerings as compared to on-site courses (Kataeva and DeYoung 2018; Sahin and Shelley 2008). Research indicates that interaction between the faculty member and learners is important to student success and retention (Ali et al. 2018). Learners have reported feelings of isolation, lack of self-direction, and eventually a decrease in motivation levels. According to Alfred Rovai and Mervyn Wighting (2005), social integration of learners and the involvement of faculty members in online courses affect students’ overall experience of learning.
Grading of learner’s work
In a study conducted in the early days of online learning, Robert Sellani and William Harrington (2002) found that online teaching was more labour-intensive than traditional delivery, since faculty members need more time to grade assignments and respond to queries. The challenge for online educators is to design efficient systems for obtaining, monitoring, grading, reporting and returning assignment work (Conceição 2006; McLendon and Albion 2000). Robert Taylor (2002) identified challenges such as obtaining instant feedback in asynchronous classes, planning and scheduling an online class based on the learner’s time zone, managing in-class participation with more than twelve learners, distraction among learners, and technological support. In addition to the above-stated challenges, pressing challenges faculty members find themselves facing also include handling students’ inquiries, teacher–learner interaction, peer-to-peer interaction, responsive teaching, language barriers, lack of training on e-learning, responding to queries on online discussion boards, and grading assignment submissions are (Ali et al. 2018; Conceição 2006; McLendon and Albion 2000; Hebert 2007; Panda and Mishra 2007).
Teacher–content challenges while developing and delivering
Faculty members not only have a tight time frame for completing student assessment evaluations, but are also expected to design a course in the language of digital natives that would meet the needs of a diverse student body with a wide spectrum of experience and technical expertise (Rosenjack et al. 2007; Al-hawari and Mouakket 2010). These challenges can be categorised as follows:
Online pedagogy
The current literature seems to accept that online education is different from on-site face-to-face education, and requires the creation of an adapted pedagogy. Creativity is required from instructors to create a course in an online learning context that keeps learners engaged. To monitor or improve students’ intellectual skills, faculty members operating in virtual classrooms are required to use relevant tracking tools such as Google Jamboard, Kahoot and Survey Monkey (Baran et al. 2011; Elison-Bowers et al. 2010; Tham and Werner 2005). Although pedagogy is important in leading a deeper learning process, effective learning in virtual classrooms is driven by collaboration among students and instructors. According to Mohammad Al Gamdi and Ahmad Samarji, a “lack of instructional design support for e-learning” (Al Gamdi and Samarji 2016, p. 26) is one of the barriers faculty members find themselves facing.
Quality of content
It is important to mention that the consistency of a learning process is not what an e-learning provider offers a learner, but rather a collaborative effort between the learner and the learning environment. Hence the quality has to do with inspiring and motivating the student. According to Dawn Birch and Bruce Burnett, student engagement varies in an online learning environment depending on the quality and amount of course content, e.g. “providing manageable ‘chunks’ of information” (Birch and Burnett 2009, p. 127). Alec Sithole et al. (2019) and Justin Ortagus and Luke Stedrak (2013) suggest six quality parameters in e-learning: tutor support, cooperation and communication in the course, use of technology, cost-value expectation, information transparency and course structure.
Maintaining the content
The time required to develop technical skills, incorporate technology and manage the curriculum is a major area of concern for academics. The risks of slipping into obsolete material are high due to the abundance of free online content. A variety of factors can hamper or prevent content updates in courses, which include an absence of dedicated resources (budget, time, expertise), lack of administrative will, and protectionism against existing curricula on the part of curriculum developers (Hai Jew 2010). E-learning content developers, therefore, find it challenging to determine how to organise a curriculum that best fits learners’ requirements but also allows for structural flexibility for future updates in the course. Another challenge for faculty members in maintaining the content is the failure to customise/adapt content to local culture, language and religious beliefs.
Adapting to learning styles and culture
Cheryl Holly et al. (2008) argue that the challenge for online teaching staff involves recognising and appreciating the learning style of remote students. In recent times, Nurul Islam et al. (2015) identified challenges related to learning styles and culture, pedagogical e-learning, technology, technical training and time management. The same set of challenges was also identified by other researchers (Babie et al. 2016). In the last decade, lack of training on e-learning and inadequate professional development helping teachers to understand learner styles emerged as the most common challenges encountered by faculty members in online delivery.
Other challenges identified by researchers were related to faculty members’ ability to cater to students’ different learning styles, time management, immediate response to queries, and technical support. With e-learning continuing to spread, educators were found to struggle with developing pedagogy-based content for modules and catering to the different learning styles of the students as they lacked the necessary skills (Islam et al. 2015).
Faculty members’ challenges and institutional support
In the last two decades, online education has become increasingly popular in the higher education sector, and most institutions agree that this form of learning will be vital to the future of education. As more and more universities move to online delivery, the pressure for change in teaching and learning practices has risen. Faculty members have expressed concern regarding the adequacy of institutional support, the transition in interpersonal relationships, and the effectiveness of online teaching and learning (Bower 2001). Although the development of an engaging distance learning course involves a significant commitment in terms of time and energy for an educator, many institutional administrators consider moving a course from the traditional classroom into an online format as part of the regular workload (ibid.).
Workload
Elaine Allen and Jeff Seaman (2014) conducted an exhaustive survey of 10,000 faculty members at 69 institutions. In this study, 85% of their respondents said developing an online course takes more effort as compared to face-to-face teaching. Julian Betts (1998) emphasises that incorporating and implementing e-learning technologies has an impact on the workload of academics (Birch and Burnett 2009; Chou and Tsai 2002; Hiltz and Turoff 2005; Mashile and Pretorius 2003; Shea et al. 2002; Tham and Werner 2005; Yang and Cornelious 2005). Other studies (DiBiase 2000; Hislop and Ellis 2004; Visser and Molin 2022; Ali and Leeds 2009) have addressed the issue of faculty members’ workload by comparing prior regular classroom teaching with online teaching. Several researchers have claimed workload issues are the greatest hurdle in online education adoption, as educators consider the workload to be higher than that of conventional courses (Bolliger and Wasilik 2009; Al Gamdi and Samarji 2016; Maguire 2005; Mihhailova 2006). Also, the lack of incentives to balance additional workload (Cook et al. 2009; Gregory and Lodge 2015) made faculty members feel that e-learning was forced upon them rather than being a natural part of institutional operations (Nichols 2008). David DiBiase notes that the effectiveness and efficiency of an online course are directly related to “the amount, and the quality, of the instructional design and development effort that produced it” (DiBiase 2000, p. 19). Faculty members who are situated at the root of this growing demand and under pressure to teach online are forced to reconsider their underlying assumptions about teaching and learning, and the roles they take on as educators.
Institutional pressures on faculty members
In online education, lower staffing costs and greater flexibility in scheduling compared to traditional teaching have appealed to institutional management across the board. A 1998–1999 national faculty member study conducted by the Higher Education Research Institute (HERI) of the University of California, Los Angeles (UCLA) found that two-thirds of college and university faculty members considered it difficult to keep up with information technology (IT). In the study, faculty members rated IT above research/publication demands, teaching load, and tenure/promotion as significant stressors (Ardichvili 2008; Bower 2001; Childs et al. 2005; Holly et al. 2008). Besides, faculty members have also raised concerns over the pressure to maximise profits at the expense of educational quality (Herbert 2006; Kebritchi et al. 2017).
Lack of administrative support
Emory McLendon and Peter Cronk (1999) and Arold Visser and Magdalena Molin (2022) suggest that the content development, delivery time, and effort may partially depend on the level of institutional support. Similar challenges are echoed by Beverley Bower (2001); referring to a survey carried out by the National Education Association (NEA), she notes that while institutional incentives known to encourage teaching staff to get involved are workload adjustments, release time and monetary support for advancement, the study (NEA 2000) found this form of support to diminish with the adoption of e-learning. Likewise, other research found insufficient support from top management (Alebaikan and Troudi 2010; Ali et al. 2018; Bolliger and Wasilik 2009; Willging and Johnson 2009), lack of professional development programmes (Ali et al. 2018; Al Gamdi and Samarji 2016), lack of formal e-learning policies (Panda and Mishra 2007), and unavailability of quality hardware and software to be some of the challenges teaching staff find themselves facing due to lack of administrative support.
Remuneration, job retention and lack of training
The current literature illustrates that university staff remuneration is no different for online, blended, or face-to-face on-site classes. At most institutions, permanent and tenured faculty members receive a salary, not compensation for each course. Stephen Ruth (2018) presents a contrasting view on the remuneration and job retention of faculty members based on an article by George Schell (2004). Ruth explains that in a traditional face-to-face setting, faculty members’ appraisal is “based on three fundamental criteria – teaching, publications/research, and service” (Ruth 2018, p. 14), putting extra pressure on the performance of staff having to accommodate online teaching in their workload.
Apart from remuneration and job retention, researchers have pointed to a lack of training as a big concern. For HEIs, the focus is more on designing and introducing online courses quickly to maximise enrolment, rather than building a pool of well-trained instructors to enhance the standard of delivery. Online teaching quality is closely tied to an institution's ability to address obstacles faculty members encounter in the development and instruction of online courses. Such hurdles include (a) remuneration and time; (b) change in the organisation; and (c) technological skills, support and infrastructure (Lloyd et al. 2012; Orr et al. 2009). Doris Bolliger and Oksana Wasilik (2009) found inadequate compensation and inequitable reward systems for promotion to be key factors influencing faculty members’ satisfaction with online teaching and learning in higher education.
Lack of training and professional development
Jennifer Herman (2012) identifies various professional development programmes available for online teaching staff, namely self-teaching, peer mentoring, collaborative course design and synchronous online training. Referring to Herman’s study of 821 institutions (ibid.), Sharla Berry notes that “only 53% offered synchronous training and 32% had formal mentoring programs” for online teaching staff (Berry 2019, p. 123). However, Leslie Pagliari et al. (2009) have a different perspective of faculty member training. They observe that faculty members do not participate in online training when it is available; the reasons being limited scope and the absence of resources for active learning. Similarly, Yoon Hi Sung et al. found that while technical information was helpful in professional development, the training was deemed useless, as it was not linked to their specific teaching needs (Sung et al. 2018).
Technology
Many faculty members do not have a sufficiently fast internet connection for online courses, which results in poor student satisfaction. Existing literature highlights a lack of network stability, difficulties with hardware and latency concerns. According to Birch and Burnett (2009), the cost of innovation and software combined with limited monetary resources creates an obstacle both for institutions and academics in adopting and integrating educational technology. The failure to make use of education technology is often due to the lack of specialised technical assistance (Al-Adwan 2020; Awofeso and Bamidele 2017; Daouk and Aldalaien 2019; Mehall 2020; Regmi and Jones 2020). A study conducted by the University of California, Los Angeles (Stolzenberg et al. 2019) revealed that faculty members found keeping abreast with IT more stressful compared to research obligations.
Job security
According to Mark Nichols (2008), challenges encountered by faculty members teaching e-learning courses are time commitment, workload concerns, IT support and lack of sufficient staff development. Other challenges are lack of: time, incentive, cooperation, building relationships, compensation, a reward system, and the standard of teaching in a virtual setting (Angelino et al. 2007; Bolliger and Wasilik 2009; Mihhailova 2006). To these identified challenges, Micki Washburn et al. add that the most significant obstacle faced by faculty members is job security. They assert that from the outset, it was evident that the goal of utilising information technology to improve education was at risk of being overshadowed by less admirable objectives such as profit-making, cost-cutting, and decreasing the reliance on full-time professors (Washburn et al. 2021).
Furthermore, Sithole et al. (2019) agree with Khe Foon Hew and Wing Sum Cheung’s (2014) views about job security and add that pedagogy, large class sizes, academic dishonesty, lack of connection with students, too many e-mails and lack of student self-discipline are the pressing challenges. Similarly, presenting a staff development initiative in New Zealand, Cathy Gunn and Mary Panko (1998) address the expectation that besides already challenging tasks, such as research and obtaining higher-level training, academics adopt technologies like e-learning, which adds to their existing workload. A vast number of educators have still not subscribed to the idea of online teaching as a full-time medium of instruction. Given the high growth rate of online instruction in higher education and the scenario of limited research on managing online teaching in the UAE, it is imperative to delineate the expectations of online teaching and examine the related challenges.
We complement our systematic review of existing literature (49 items published between 1999 and 2020)3 with qualitative research. This involved open-ended semi-structured interviews with 15 teaching staff in the UAE, followed by a thematic analysis of their responses.
Methodology for the qualitative part of our study
The research design we selected for this part of our study was qualitative in nature, using interviews. Sharan Merriam (2019) argues that qualitative methods are often critical in understanding how participants make meaning of the problem being studied. Jack Fraenkel and Norman Wallen (1993) state that a deeper understanding of any phenomenon could be provided by qualitative research. This prompted our choice of a qualitative approach to better understand the issues related to online teaching in the context of higher education in the UAE beyond just determining cause and effect.
We conducted in-depth semi-structured interviews to gain a deeper understanding of each respondent’s articulations and to capture their experiences first hand. We reckoned that an open-ended style of inquiry would help us to obtain more voluntarily shared opinions and to avoid the potential bias from restricting responses to the researcher’s fixed categories on the challenges and issues of online teaching. We obtained ethical clearance from Westford Education Group, and informed written consent from all participants.
Sample selection and data collection
Our key informant group for this study was faculty members who were involved in the planning, design and delivery of online business management courses in the higher education sector in the UAE for at least five years. We used a purposive sampling approach which involves the selection of informants based on an important characteristic under study. Moreover, as noted by Eileen Gambrill (1991), in a purposive sampling approach, the researchers determine the ideal sample based on their knowledge of the population and invariably the aims of the research itself. Respondents (N=15) were contacted in advance to arrange a convenient time for the semi-structured interviews, which averaged approximately 45–60 minutes in length and were conducted in English.
To explore the validity of the results, we applied member checking, ensuring that results were returned to participants to verify that their statements were rendered accurately and in line with their experience. Linda Birt et al. (2016) confirm that high-quality qualitative research is built on the trustworthiness and reliability of its results, and member checking is frequently mentioned as one of the validation methods. Each interview was video-recorded with the participant’s permission and later transcribed by us, the researchers. We then proofread the transcripts and compared them with the recordings to pick up any discrepancies.
In the analysis below, we have included selected verbatim descriptions to provide context to the discussion. We also maintained field notes where we noted pauses, repetitions and tonality that supported a well-rounded analysis. Although we used NVivo4 for data analysis, we listened to the recordings of the interview together to ensure that an accurate interpretation of the meanings had been noted. Hence the reliability was ensured by referring to the stability of responses. Their interpretation was an ongoing process in this study and was not relegated to the end of data collection – whenever we had doubts, we confirmed our understanding right away with the participants during the interview. This allowed our final analysis “to rest on more secure ground” (Kvale 2011, p. 156).
Table 1 outlines the semi-structured interview questions which we developed based on existing studies on learner issues, content issues and instructor issues. The first section involves background questions, followed by questions based on student-specific challenges (Section 2), content-specific challenges (Section 3), and institutional challenges teaching staff were facing (Section 4).Table 1 Semi-structured interview questions
Section 1
Background questions
Are you an industry professional teaching online or a full-time academic?
How long have you been teaching online in higher education in the UAE?
Section 2
Learner issues
Based on: -
(Li and Irby 2008; Washburn et al. 2021)
What do you think are the expectations of online learners in higher education in the UAE?
Do you think learners have the required technical skills to engage in an online class? If yes, please explain. If no, please explain.
Do you think cultural differences impact the ability to learn online in higher education in the UAE?
Do you think building an educator–learner relationship via a digital platform is difficult in online education?
Section 3
Content issues
Based on: -
(Herman 2012)
Can you elaborate on issues concerning content development for online courses in higher education in the UAE?
What do you think are the challenges in the use of multimedia for content creation and delivery of online courses in higher education in the UAE?
How do you gauge the level of success of the existing content development techniques in online teaching in higher education in the UAE?
What learning materials/resources would you like to use but do not have at your disposal?
Section 4
Instructor issues
Based on: -
(Anderson 2008)
What are the challenges you face in motivating and engaging online learners?
Do you feel the transition from face-to-face to online teaching can impact an instructor’s ability to deliver successful online courses in higher education in the UAE? If yes, please explain. If not, please explain why not.
What is your take on communication barriers which exist in online teaching in the higher education sector in the UAE?
Do you face any challenges in supporting students with special needs in an online environment?
Do you think the different teaching styles can impact the delivery of online teaching in comparison with on-site delivery? If yes, please elaborate.
Have you undergone any professional development to enhance in-class engagement while teaching online? If yes, can you explain how it helped you with online teaching?
Do you think the classroom capacity hinders the quality of online education? If yes, explain why, if not please explain why not.
Do you think the compensation offered to an online faculty member is on a par with on-campus faculty members? If yes, explain why, if no, please explain why not.
Would you like to provide any other information?
Do you have any questions about the interview?
Further background questions
In a study by Greg Guest et al. (2006), the authors explain that saturation was achieved after as few as 12 interviews even though the total number of participants involved in the study was 60. Clark Moustakas (1994) argues that a small population—in our case, 15 participants—can justify the dissemination of new knowledge by producing patterns and identifying relationships. We limited our own analysis to 12 interviews, since a point of saturation was attained after 12 interviews and no new information could be gained.
Discussion and thematic analysis
Challenges to online teaching should be explored in detail, and as indicated by Virginia Braun and Victoria Clarke (2013), thematic analysis can help investigate the latent meanings, assumptions and ideas that lie beneath what is explicitly stated. Hence we chose a thematic method of analysis since it is not tied to any particular theoretical perspective as such, making it a very flexible method as argued by Moira Maguire and Brid Delahunt (2017). We used NVivo 12 pro to perform the thematic analysis for this research, a software which ensures easy and efficient coding which makes data retrieval easier. As suggested by AlYahmady Hamed Hilal and Saleh Said Alabri (2013), NVivo helps reshape and reorganise coding and node structure quickly. Through our thematic analysis, we identified five emerging themes: (1) learners’ expectations; (2) culture; (3) lack of incentives for faculty members to engage in online teaching; (4) pedagogy; and (5) technology.
Emerging themes
Learners’ expectations
"They want to make sure that everything is crystal clear and provided to them, that is a very important part so that when they are taking a module or any kind of an online course, they are clear about the what, why, and how. They also want to make sure that it’s interactive because it is substituting the face-to-face on-campus mode of teaching." [Respondent 2]
"The issue is that most of the students do not know each other, have never met face to face, and have very little common context; therefore they hesitate in interacting in the class. Some students are more open than others in sharing their views and getting the faculty [member]’s attention while others feel alienated." [Respondent 7]
This theme refers to learners’ expectations of online learning in higher education in the UAE. Our analysis revealed that online learners believe the clarity of the topic discussed in the class is essential. There is also a general expectation that the same level of support that is provided in on-site classes will also be provided in the online mode of delivery. While faculty members can facilitate learning through interactive group discussions and other class activities in on-site classes, the same is extremely challenging in the online mode owing to aspects such as varied internet speed, knowledge of interactive technologies, transparency, cultural differences, and students’ motivation to engage in group activities online. In an online class, socialising with peers is limited, which in turn acts as a barrier to interactive class participation and knowledge sharing. Besides, most students in the UAE are from diverse backgrounds as the majority of the population are expatriates.5 Other students undertaking the course from their respective countries base their expectations on their experience of undertaking courses in their home country, thereby making the management of expectations a difficult task for faculty members. Participants also identified the lack of trust-building opportunities and transparency in online classes as a key barrier to effective delivery. This theme and its analysis are also relevant to earlier studies on teacher–student interaction (Bawa 2016; Prensky 2001) where e-learners expected the instructors to understand their expectations and provide a very high level of learning support.
Culture
"See, cultural differences will be there in a country like UAE. After all, we are supposed to be a melting pot of different cultures, and different countries. We have more than a hundred nationalities, and of course, many of them come as our students with different needs and expectations." [Respondent 7]
"Cultural differences have an impact because people from different cultures log in for a session and it also impacts the ability to learn because the culture is what defines the way you approach things and it is true with online learning as well." [Respondent 10]
“Culture” was revealed as another significant theme through our analysis, since culture in a way defines the approach to online learning in the context of the UAE, a melting pot of a variety of cultures. The theme contributes to studies such as those of Lejla Vrazalic et al. (2010) and Somnan Ali et al. (2018) where linguistic and cultural factors were linked to the usage of online learning and learning styles and expectations. The majority of the students enrolled in the UAE universities where our participants teach are non-English speakers, however, all the courses delivered are from UK and Australian universities and therefore in English. The assignments, lectures and class activities are primarily Western in their approach, but the students undertaking the study are mostly African and Asian, with very different learning orientations. Also, differences exist between African and Asian students in terms of ethnic and national orientation. These cultural differences have an impact on their learning, since some have an individualist approach towards their learning while others expect learning to be more collectivist.
Lack of incentives for faculty members to engage in online teaching
"They have high expectations about the quality of content in online learning. They believe that online teaching is very easy. And it is at your doorstep, which is underestimating your teaching workload. It is the method of delivering a session that has changed, not just the material which is to be taught." [Respondent 12]
"I mean if you are asking about compensation I feel now because of the advent of online learning with the open universities and reputed universities such as Harvard and Cambridge, a lot of these universities are flourishing with their online learning and platforms. Although overall, the remuneration of faculty [members] has increased, the additional workload and responsibilities are not being taken seriously." [Respondent 10]
The emergence of this theme from our analysis of participants’ responses points towards the lack of adequate compensation offered to online teaching staff. It mainly adds to the existing study by Bolliger and Wasilik (2009) which identified inadequate compensation as a key problem area for online teaching staff. Our participants believed that their salary did not take into consideration the differences in teaching and the additional workload that comes with a new mode of delivery. All our participants expressed their view that the content delivered in a traditional on-site class is unsuitable for the online mode, therefore, they found themselves having to spend considerable time and effort in aligning the materials to the needs and requirements of the online students. Moreover, learning how to use new technological tools to deliver classes was mentioned as an additional pressure without any monetary incentives to motivate continuous learning and development.
Pedagogy
"I have realised that scaffolding6 here is a major concern and needs to be critically planned as part of pedagogy." [Respondent 4]
"So, I think the use of illustrations matters a lot. And I think one of the ways to do it is to strongly understand your audience in an online mode." [Respondent 2]
The emergence of the “pedagogy” theme ties in with the existing models on pedagogy and the relevance of pedagogy in online learning discussed by Terry Anderson (2008) and Judith Harris et al. (2009). Our findings reveal that to achieve effective teaching and learning, “pedagogy” should be prioritised ahead of technology in online learning. Several other factors such as the use of graphic illustrations and critical feedback were also identified by our participants among the pedagogical elements of online learning in higher education in the UAE. Furthermore, the process of scaffolding students into the online learning environment was revealed as another major challenge in pedagogy in online learning in higher education in the UAE.
Technology
"Multimedia has evolved as a part of technology evolution. So, I feel in terms of multimedia the main challenge that one would be having is the resources to make all electronic devices compatible with the LMS [learning management system]." [Respondent 9]
"I think the internet itself is a big concern due to bandwidth issues. Let me give you an example of technological limitations – when we teach, we use our video and audio but the students respond to us via chat, the communication can be challenging." [Respondent 4]
The “technology” theme contributes immensely to the existing knowledge on the use of technology in online learning in higher education in the UAE. In the context of the UAE, this theme ties in with Sithole et al.’s (2019) quality parameters and also reveals faculty members’ challenges due to students’ increasing use of smartphones and tablets as important technology components in higher education in the UAE. The theme reaffirms that even though such components help in collaboration and facilitation, there still exists a gap in terms of compatibility with learning management systems.
Conclusion
Following global trends in online education, most HEIs in the UAE now also offer online programmes. To be competitive in the industry, institutions have shown concern about high attrition and low retention of students as well as challenges online learners are facing, but there is still a dearth of studies focusing on the challenges arising for teaching staff. Our own study explored the challenges encountered by faculty members teaching online business management courses in the UAE. By undertaking a review of e-learning literature published between 1999 and 2020 and conducting in-depth semi-structured interviews with 15 faculty members, we identified teacher-specific challenges and grouped them into five themes: (1) learners’ expectations; (2) culture; (3) lack of incentives for faculty members to engage in online teaching; (4) pedagogy; and (5) technology, thereby extending the work of Anderson (2008). Understanding these identified themes will help academic institutions improve delivery of their online programmes. All the challenges encountered by faculty members in the UAE seem to be related to each other and need to be addressed coherently for a seamless adoption and delivery of online education.
Future research
This study explored and examined the challenges encountered by faculty members who delivered business management courses online. There are some important caveats to the study that deserve to be mentioned. Its findings are limited since the study primarily focused only on teacher-specific challenges in online education and did not explore challenges related to students or institutions in online education. Besides, it would be of interest to consider reviews of faculty members from other academic disciplines apart from business management. Furthermore, our work focused on the UAE and therefore the conclusions drawn cannot represent the situation in other countries. Future research might undertake a quantitative study with a larger sample to generalise the results. Our current findings may lead to future research in developing strategies to overcome the challenges encountered by faculty members that are student-specific, content-specific, and dependent on institutional support. The results from the present study will help key stakeholders better understand the challenges encountered by faculty members in the e-learning environment and come up with strategies that act as catalysts in improving the learning and delivery in online education.
Acknowledgements
We wish to thank the faculty members who participated in the study conducted by us. We appreciate the insights that have informed the study and helped us and the readers of this study in understanding the challenges in teaching online courses in higher education. All the authors confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this study that could have influenced its outcome. We wish to confirm that the data being reported are accurate, that due consideration has been given to the protection of intellectual property associated with this work, and that there are no impediments to publication. We wish to confirm that the software we used for our data analysis is NVivo 12, and that the NVivo coding and analysis are available and can be produced for verification any time upon request. Finally, we wish to confirm that all ethical considerations have been taken into account in this study.
1 GCC stands for Gulf Cooperation Council, also known as the Cooperation Council for the Arab States of the Gulf. It is an alliance of six countries, established in 1981. Its member states are Bahrain, Kuwait, Oman, Qatar, Saudi Arabia and the United Arab Emirates. For more information, visit GCC’s official website at https://www.gcc-sg.org/en-us/AboutGCC/Pages/StartingPointsAndGoals.aspx [accessed 12 April 2023].
2 A summary table of the 49 items included in our systematic literature review (providing author name[s], publication year, sample size, methodology, research location, findings, and limitations and scope of future research) is available at https://docs.google.com/document/d/1u7JuiJ5Fl6-M_m4j1eA7derOYkSBum-T/edit?usp=sharing&ouid=114482687801984676698&rtpof=true&sd=true.
3 These items are marked with asterisks in the reference section.
4 NVivo is a software designed for qualitative data analysis.
5 According to Statista, “The United Arab Emirates has an estimated population of 10.54 million … Today just over a million of the residents in the United Arab Emirates are nationals, the majority of residents are expats and foreign workers. [The m]ajority of foreigner[s] in the United Arab Emirates originate from South Asia with Indian nationals in the lead” (Puri-Mirza 2022).
6 According to Adam Shaw, “scaffolding is an instructional method that progressively moves students toward greater independence and understanding during the learning process. Similar to how builders require scaffolding during construction to access new heights, instructional scaffolding helps students navigate coursework and accomplish tasks they otherwise might not have been able to” (Shaw 2019, online).
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
The 49 items which were included in our systematic literature review are marked with an asterisk
Abulibdeh ES Syed Hassan SS E-learning interactions, information technology self efficacy and student achievement at the University of Sharjah, UAE Australasian Journal of Educational Technology 2011 10.14742/ajet.926
Al-Adwan AS Investigating the drivers and barriers to MOOCs adoption: The perspective of TAM Education and Information Technologies 2020 25 6 5771 5795 10.1007/s10639-020-10250-z
Alebaikan R Troudi S Blended learning in Saudi universities: Challenges and perspectives Research in Learning Technology 2010 18 1 49 59 10.1080/09687761003657614
Al Gamdi MA Samarji A Perceived barriers towards e-learning by faculty members at a recently established university in Saudi Arabia International Journal of Information and Education Technology 2016 6 1 23 28 10.7763/IJIET.2016.V6.652
Al-Gumaei GS Alzouebi K Abdelaziz HA The Islamic studies teachers’ perception of integrating ICT into the teaching and learning in the UAE public schools: Challenges, opportunities and practices International Journal of Technology Diffusion 2019 10 2 69 82 10.4018/IJTD.2019040104
Al-hawari M Mouakket S The influence of technology acceptance model (TAM) factors on students' e-satisfaction and e-retention within the context of UAE e-learning Education, Business and Society: Contemporary Middle Eastern Issues 2010 3 4 299 314 10.1108/17537981011089596
*Ali, R. & Leeds (2009). The impact of face-to-face orientation on online retention: A pilot study. Online Journal of Distance Learning Administration, 12(4) [online]. Retrieved 15 April 2023 from https://ojdla.com/archive/winter124/ali124.pdf
Ali S Uppal MA Gulliver SR A conceptual framework highlighting e-learning implementation barriers Information Technology & People 2018 31 1 156 180 10.1108/ITP-10-2016-0246
Allen IE Seaman J Grade change: Tracking online education in the United States 2014 Oakland CA Babson Survey Research Group and Quahog Research Group
Anderson T The theory and practice of online learning 2008 Edmonton AB Athabasca University Press
Andersson A Gronlund A A conceptual framework for e-learning in developing countries: A critical review of research challenges The Electronic Journal on Information Systems in Developing Countries 2009 38 1 1 16 10.1002/j.1681-4835.2009.tb00271.x
Angelino, L. M., Williams, F. K., & Natvig, D. (2007). Strategies to engage online students and reduce attrition rates. Journal of Educators Online, 4(2) [online]. https://eric.ed.gov/?id=EJ907749
Ardichvili A Learning and knowledge sharing in virtual communities of practice: Motivators, barriers, and enablers Advances in Developing Human Resources 2008 10 4 541 554 10.1177/1523422308319536
Awofeso N Bamidele M Learners’ perceptions of instructor feedback in e- learning courses – findings from HBMSU, United Arab Emirates Application and Theory of Computer Technology 2017 2 3 35 47 10.22496/atct.v2i3.74
*Babie, S., Čičin-Šain, M., & Bubaš, G. (2016). A study of factors influencing higher education teachers’ intention to use E-learning in hybrid environments. In P. Biljanovic, Z. Butkovic, K. Skala, … & D. Lukman (Eds), Proceedings of the 39th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO) (pp. 998–1003). Opatija/Rijeka: Croatian Society for Information and Communication Technology, Electronics and Microelectronics. 10.1109/MIPRO.2016.7522285
Baran E Correia A-P Thompson A Transforming online teaching practice: critical analysis of the literature on the roles and competencies of online teachers Distance Education 2011 32 3 421 439 10.1080/01587919.2011.610293
Bawa P Retention in online courses: Exploring issues and solutions – a literature review SAGE Open 2016 10.1177/2158244015621777
Beaven T Emke M Ernest P Rutherford AG Hampel R Stanojevic MM Stickler U Needs and challenges for online language teachers – the ECML project DOTS Teaching English with Technology 2010 10 2 5 20
Berry S Professional development for online faculty: Instructors’ perspectives on cultivating technical, pedagogical and content knowledge in a distance program Journal of Computing in Higher Education 2019 31 1 121 136 10.1007/s12528-018-9194-0
Betts, J. R. (1998). The impact of educational standards on the level and distribution of earnings. The American Economic Review, 88(1), 266–275. Retrieved from http://www.jstor.org/stable/116829
Birch D Burnett B Bringing academics on board: Encouraging institution-wide diffusion of e-learning environments Australasian Journal of Educational Technology 2009 10.14742/ajet.1184
Birt L Scott S Cavers D Campbell C Walter F Member checking: A tool to enhance trustworthiness or merely a nod to validation? Qualitative Health Research 2016 26 13 1802 1811 10.1177/1049732316654870 27340178
Blumenstyk, G. (2019). Online education: An overlooked lever of education policy. The Chonicle of Higher Education, 11 June [online article]. Retrieved 18 July 2020 from https://www.chronicle.com/article/Online-Education-An/246464
Bolliger DU Wasilik O Factors influencing faculty satisfaction with online teaching and learning in higher education Distance Education 2009 30 1 103 116 10.1080/01587910902845949
*Bower, B. L. (2001). Distance education: Facing the faculty challenge. Online Journal of Distance Learning Administration, 4(2), 1–6. Retrieved 11 April 2023 from https://ojdla.com/archive/summer42/bower42.pdf
Braun V Clarke V Using thematic analysis in psychology Qualitative Research in Psychology 2006 3 2 77 101 10.1191/1478088706qp063oa
*Burchum, J., Russell, C. K., Likes, W., Adymy, C., Britt, T., Driscoll, C., … Cowan, P. (2007). Confronting challenges in online teaching: The WebQuest solution. MERLOT Journal of Online Learning and Teaching, 3(1), 40–57. Retrieved 11 April 2023 from https://jolt.merlot.org/vol3no1/burchum.pdf
Chandra S Palvia S Online education next wave: peer to peer learning Journal of Information Technology Case and Application Research 2021 23 3 157 172 10.1080/15228053.2021.1980848
Childs S Blenkinsopp E Hall A Walton G Effective e-learning for health professionals and students – barriers and their solutions: A systematic review of the literature – findings from the HeXL project Health Information and Libraries Journal 2005 22 2 20 32 10.1111/j.1470-3327.2005.00614.x
Chou C Tsai C-C Developing web-based curricula: Issues and challenges Journal of Curriculum Studies 2002 34 6 623 636 10.1080/00220270210141909
Cole MT Shelley DJ Swartz LB Online instruction, e-learning, and student satisfaction: A three-year study The International Review of Research in Open and Distributed Learning 2014 10.19173/irrodl.v15i6.1748
Conceição SCO Faculty lived experiences in the online environment Adult Education Quarterly 2006 57 1 26 45 10.1177/1059601106292247
Cook RG Ley K Crawford C Warner A Motivators and inhibitors for university faculty in distance and e-learning British Journal of Educational Technology 2009 40 1 149 163 10.1111/j.1467-8535.2008.00845.x
Daouk L Aldalaien M The usage of e-learning instructional technologies in higher education institutions in the United Arab Emirates (UAE) The Turkish Online Journal of Educational Technology 2019 18 3 97 109
DiBiase D Is distance teaching more work or less work? The American Journal of Distance Education 2000 14 3 6 20 10.1080/08923640009527061
Elison-Bowers P Henderson K Sand J Osgood L Resolving instructor challenges in the online classroom The International Journal of Learning 2010 17 1 339 346 10.18848/1447-9494/CGP/v17i01/46845
Ferdousi, B. (2016). Addressing student retention and persistence issue in online classes. In Proceedings of the 2016 American Society for Engineering Education (ASEE) North Central Section (NCS) Conference [online]. Mt. Pleasant, MI: Central Michigan University/ASEE NCS. Retrieved 11 April 2023 from https://asee-ncs.org/wp-content/uploads/2021/12/proceedings/2016/faculty_regular_papers/2016_ASEE_NCS_paper_49.pdf
Fraenkel JR Wallen NE How to design and evaluate research in education 1993 London McGraw-Hill Education
Gambrill E Research methods for social work by Allen Rubin and Earl Babbie [book review] Social Work Research & Abstracts 1991 27 4 36 37 10.1093/swra/27.4.36-b
Gregory MS-J Lodge JM Academic workload: The silent barrier to the implementation of technology-enhanced learning strategies in higher education Distance Education 2015 36 2 210 230 10.1080/01587919.2015.1055056
Guest G Bunce A Johnson L How many interviews are enough? An experiment with data saturation and variability Field Methods 2006 18 1 59 82 10.1177/1525822X05279903
Gunn C Panko M Latchem C Lockwood F Staff development within a flexible and open learning course Staff development in open and flexible learning 1998 London Routledge
Hai Jew, S. (2010). An instructional design approach to updating an online course curriculum. Educause Review, 15 December [online article]. Retrieved 15 July 2020 from https://er.educause.edu/articles/2010/12/an-instructional-design-approach-to-updating-an-online-course-curriculum
Harris J Mishra P Koehler M Teachers’ technological pedagogical content knowledge and learning activity types Journal of Research on Technology in Education 2009 41 4 393 416 10.1080/15391523.2009.10782536
Hartman J Dziuban C Moskal P Faculty satisfaction in ALNs: A dependent or independent variable Journal of Asynchronous Learning Networks 2000 4 3 155 179 10.24059/olj.v4i3.1892
Haydarov R Moxley V Anderson D Counting chickens before they are hatched: An examination of student retention, graduation, attrition, and dropout measurement validity in an online master’s environment Journal of College Student Retention: Research, Theory & Practice 2013 14 4 429 449 10.2190/CS.14.4.a
Hebert DG Five challenges and solutions in online music teacher education Research and Issues in Music Education 2007 5 1 1 10
Herbert, M. (2006). Staying the course: A study in online student satisfaction and retention. Online Journal of Distance Learning Administration, 9(4), 300–317. Retrieved 11 April 2023 from https://ojdla.com/archive/winter94/herbert94.pdf
Herman JH Faculty development programs: The frequency and variety of professional development programs available to online instructors Online Learning 2012 10.24059/olj.v16i5.282
Hew KF Cheung WS Students’ and instructors’ use of massive open online courses (MOOCs): Motivations and challenges Educational Research Review 2014 12 45 58 10.1016/j.edurev.2014.05.001
Hilal, A. H., & Alabri, S. S. (2013). Using NVivo for data analysis in qualitative research. International Interdisciplinary Journal of Education, 2(2), 181–186. Retrieved 11 April 2023 from http://www.iijoe.org/v2/IIJOE_06_02_02_2013.pdf
Hiltz SR Turoff M The evolution of online learning and the revolution in higher education Communications of the ACM 2005 48 10 59 64 10.1145/1089107.1089139
Hislop GW Ellis HJC A study of faculty effort in online teaching The Internet and Higher Education 2004 7 1 15 31 10.1016/j.iheduc.2003.10.001
Holly C Legg TJ Mueller D Adelman DS Online teaching: Challenges for a new faculty role Journal of Professional Nursing 2008 24 4 254 258 10.1016/j.profnurs.2007.07.003 18662662
Hoteit, L., El Hachem, M., Patrick Erker, C., & Farah, S. (2018). Where to invest now in GCC private education [online paper]. Boston, MA: The Boston Consulting Group. Retrieved 11 April 2023 from https://web-assets.bcg.com/img-src/BCG-Where-Invest-Now-GCC-Private-Education-v4_tcm9-190607.pdf
Hughes H Actions and reactions: Exploring international students’ use of online information resources Australian academic and research libraries 2005 36 4 169 179 10.1080/00048623.2005.10755308
Hunt BD Oyarzun B Online learning perspectives of Native American students Journal of Educational Technology Systems 2020 48 3 321 334 10.1177/0047239519867921
IHEP (Institute for Higher Education Policy). (2000). Quality on the line: Benchmarks for success in internet-based distance education. Washington, DC: IHEP. Retrieved 12 April 2023 from https://www.ihep.org/wp-content/uploads/2014/05/uploads_docs_pubs_qualityontheline.pdf
IHEP. (2014). Quality on the line [webnews, 15 May]. Washington, DC: IHEP. Retrieved 24 April 2023 from https://www.ihep.org/press/quality-on-the-line/
Islam N Beer M Slack F E-learning challenges faced by academics in higher education: A literature review Journal of Education and Training Studies 2015 3 5 102 112 10.11114/jets.v3i5.947
Jacobs, P. (2013). The challenges of online courses for the instructor. Research In Higher Education Journal, 21 [online]. Retrieved 11 April 2023 from https://www.aabri.com/manuscripts/131555.pdf
Kataeva Z DeYoung AJ Faculty challenges and barriers for research and publication in Tajik higher education European Education 2018 50 3 249 265 10.1080/10564934.2017.1408417
Kebritchi M Lipschuetz A Santiague L Issues and challenges for teaching successful online courses in higher education: A literature review Journal of Educational Technology Systems 2017 46 1 4 29 10.1177/0047239516661713
King A From sage on the stage to guide on the side College Teaching 1993 41 1 30 35 10.1080/87567555.1993.9926781
Kvale S Doing interviews 2011 London SAGE
Lai C-H Lin H-W Lin R-M Tho PD Effect of peer interaction among online learning community on learning engagement and achievement International Journal of Distance Education Technologies 2019 17 1 66 77 10.4018/IJDET.2019010105
Li, C.S. & Irby, B. (2008). An Overview of online education: Attractiveness, benefits, challenges, concerns and recommendations. College Student Journal, 42(2), 449–458. Retrieved 26 April 2023 from https://www.learntechlib.org/p/103183/
*Lloyd, S. A., Byrne, M. M., & McCoy, T. S. (2012). Faculty – perceived barriers of online education. MERLOT Journal of Online Learning and Teaching, 8(1), 1–12. Retrieved 11 April 2023 from https://jolt.merlot.org/vol8no1/lloyd_0312.pdf
*Maguire, L. L. (2005). Literature review – faculty participation in online distance education: Barriers and motivators. Online Journal of Distance Learning Administration, 8(1) [online]. Retrieved 11 April 2023 from https://ojdla.com/archive/spring81/maguire81.pdf
Maguire, M., & Delahunt, B. (2017). Doing a thematic analysis: A practical, step-by-step guide for learning and teaching scholars. All Ireland Journal of Higher Education, 9(3), 3351–33514. Retrieved 11 April 2023 from https://ojs.aishe.org/index.php/aishe-j/article/view/335/553
Mashile EO Pretorius FJ Challenges of online education in a developing country South African Journal of Higher Education 2003 17 1 132 139 10.4314/sajhe.v17i1.25202
McCarthy, N. (2020). COVID-19’s staggering impact on global education. World Economic Forum, 26 March [webnews item]. Retrieved 20 April 2020 from https://www.weforum.org/agenda/2020/03/infographic-covid19-coronavirus-impact-global-education-health-schools/
McLendon, E., & Albion, P. (2000). Rethinking academic practices: meeting some challenges of online delivery. In B. Harper & R. Oliver (Eds), Proceedings of the 2000 Apple University Consortium Conference [online]. Sydney: Apple Computer Australia. Retrieved 15 April 2020 from https://eprints.usq.edu.au/6998
*McLendon, E., & Cronk, P. (1999). Rethinking academic management practices: a case of meeting new challenges in online delivery. Journal of Distance Learning Assessment, 2(1). Retrieved 11 April 2023 from https://ojdla.com/archive/spring21/mclendon21.pdf
*Mehall, S. (2020). Purposeful interpersonal interaction in online learning: What is it and how is it measured? Online Learning, 24(1), 182–204. https://eric.ed.gov/?id=EJ1249281
Merriam, S. B. (2019). Qualitative research: Designing, implementing, and publishing a study. In V. Wang (Ed.), Scholarly publishing and research methods across disciplines (pp. 1–18). Hershey, PA: IGI Global. 10.4018/978-1-5225-7730-0.ch001
Mihhailova G E-learning as internationalization strategy in higher education: Lecturer’s and student’s perspective Baltic Journal of Management 2006 1 3 270 284 10.1108/17465260610690926
Moustakas, C. (1994). Phenomenological research methods. Thousand Oaks, CA: SAGE.
NEA (National Education Association) (2000). A survey of traditional and distance learning higher education members. Washington, DC: National Education Association. Retrieved 12 April 2023 from http://docushare3.dcc.edu/docushare/dsweb/Get/Version-2234/NEAdlstudy.pdf
Nichols M Institutional perspectives: The challenges of e-learning diffusion British Journal of Educational Technology 2008 39 4 598 609 10.1111/j.1467-8535.2007.00761.x
Orr R Williams MR Pennington K Institutional Efforts to Support Faculty in Online Teaching Innovative Higher Education 2009 34 4 257 268 10.1007/s10755-009-9111-6
Ortagus JC Stedrak LJ Online education and contingent faculty: An exploratory analysis of issues and challenges for higher education administrators (perspectives on online education) Educational Considerations 2013 40 3 31 33 10.4148/0146-9282.1096
Pagliari, L., Batts, D., & McFadden, C. (2009). Desired versus actual training for online instructors in community colleges. Online Journal of Distance Learning Administration, 12(4) [online]. Retrieved 11 April 2023 from https://ojdla.com/archive/winter124/pagliari124.pdf
Panda S Mishra S E-learning in a mega open university: Faculty attitude, barriers and motivators Educational Media International 2007 44 4 323 338 10.1080/09523980701680854
Prensky M Digital natives, digital immigrants, Part 1 On the Horizon 2001 9 5 1 6 10.1108/10748120110424816
Puri-Mirza, A. (2022). Demographics in UAE: Statistics & facts. Statista, 9 Dec 2022 [online resource]. Retrieved 17 April 2023 from https://www.statista.com/topics/10338/demographics-in-uae/?#topicOverview
*Regmi, K., & Jones, L. (2020). A systematic review of the factors – enablers and barriers – affecting e-learning in health sciences education. BMC Medical Education, 20, Art. no. 91. 10.1186/s12909-020-02007-6
Reynolds, R., & Rizvi, A. (2019). Dubai university figures show nearly half study business degrees, as science lags behind. The National, 30 May [online article] Retrieved 28 May 2021 from https://www.thenationalnews.com/uae/education/dubai-university-figures-show-nearly-half-study-business-degrees-as-science-lags-behind-1.867847
Rosenjack, J., Russell, C., Likes, W., Adymy, C., Britt, & Teresa. (2007). Confronting challenges in online teaching: The WebQuest solution. MERLOT Journal of Online Learning and Teaching, 3(1), 40–57. Retrieved 11 April 2023 from https://jolt.merlot.org/vol3no1/burchum.pdf
Rovai AP Wighting MJ Feelings of alienation and community among higher education students in a virtual classroom The Internet and Higher Education 2005 8 2 97 110 10.1016/j.iheduc.2005.03.001
Ruth S Faculty opposition to online learning: Challenges and opportunities International Journal of Technology in Teaching and Learning 2018 14 1 12 23 10.37120/ijttl.2018.14.1.02
Sahin I Shelley M Considering students’ perceptions: The distance education student satisfaction model Journal of Educational Technology & Society 2008 11 3 216 223 10.2307/jeductechsoci.11.3.216
Schell G Universities marginalize online courses: Why should faculty members develop online courses if the effort may be detrimental to their promotion or tenure? Communications of the Association for Computing Machinery 2004 47 7 53 56 10.1145/1005817.1005821
Sellani RJ Harrington W Addressing administrator/faculty conflict in an academic online environment The Internet and Higher Education 2002 5 2 131 145 10.1016/s1096-7516(02)00090-8
Shaw, A. (2019). Scaffolding learning in the online classroom. Wiley University Services, 5 September [online resource]. Retrieved 17 April 2023 from https://ctl.wiley.com/scaffolding-learning-in-the-online-classroom/
Shea TP Sherer PD Kristensen EW Harnessing the potential of online faculty development: Challenges and opportunities To Improve the Academy 2002 20 1 162 182 10.1002/j.2334-4822.2002.tb00580.x
Sher A Assessing the relationship of student-instructor and student-student interaction to student learning and satisfaction in web-based online learning environment Journal of Interactive Online Learning 2009 8 2 102 120
Sithole A Mupinga DM Kibirige JS Manyanga F Bucklein BK Expectations, challenges and suggestions for faculty teaching online courses in higher education International Journal of Online Pedagogy and Course Design 2019 9 1 62 77 10.4018/IJOPCD.2019010105
Stoffregen J Pawlowski JM Pirkkalainen H A barrier framework for open e-learning in public administrations Computers in Human Behavior 2015 51 674 684 10.1016/j.chb.2014.12.024
Stolzenberg, E. B., Eagan, M. K., Zimmerman, H. B., Berdan Lozano, J., Cesar-Davis, N. M., Aragon, M. C., & Rios-Aguilar, C. (2019). Undergraduate teaching faculty: The HERI Faculty Survey 2016–2017. Los Angeles, CA: Higher Education Research Institute, University of California Los Angeles (UCLA). Retrieved 26 April 2023 from https://heri.ucla.edu/monographs/HERI-FAC2017-monograph.pdf
Strachota, E.M. (2003). Student satisfaction in online courses: An analysis of the impact of learner–content, learner–instructor, learner–learner and learner–teacher interaction. PhD thesis, The University of Wisconsin–Milwaukee. Dissertation Abstracts International, 64(8), 2746.
Sung YH Kang EY Lee W-N Why do we indulge? Exploring motivations for binge watching Journal of Broadcasting & Electronic Media 2018 62 3 408 426 10.1080/08838151.2018.1451851
Taylor RW Pros and cons of online learning: A faculty perspective Journal of European Industrial Training 2002 26 1 24 37 10.1108/03090590210415876
Tham CM Werner JM Designing and evaluating e-learning in higher education: A review and recommendations Journal of Leadership & Organizational Studies 2005 11 2 15 25 10.1177/107179190501100203
Tomei L The impact of online teaching of faculty load: Computing the ideal class size for online courses Journal of Technology and Teacher Education 2006 14 3 531 541 10.4018/IJOPCD.2019070101
UAE MoE (United Arab Emirates Ministry of Education). (2019). Higher education [online resource]. Retrieved 28 May 2021 from https://www.moe.gov.ae/en/EServices/Pages/ServiceCatalog.aspx?Servicelevel=higher%20education
UNESCO (United Nations Educational, Scientific and Cultural Organization). (n.d.). COVID-19 recovery [dedicated webpage]. Retrieved 12 April 2023 from https://www.unesco.org/en/covid-19/education-response
Visser, A., & Molin, M. (2022). Where should I buy my sofa? A qualitative study on Swedish digital natives in the omni-channel furniture retailing environment. Master’s dissertation, Jönköping International Business School. Retrieved 15 August 2022 from https://www.diva-portal.org/smash/record.jsf?pid=diva2:1667001
*Vrazalic, L., MacGregor, R., Behl, D., & Fitzgerald, J. (2010). E-learning barriers in the United Arab Emirates: Preliminary results from an empirical investigation. IBIMA Business Review Journal, 2010(2010), Art. no. 458727. 10.5171/2010.458727
Washburn M Crutchfield J Roper DO Smith D Padilla Y Changes to MSW [Masters of Social Work] faculty workload resulting from COVID-19: An issue of equity Journal of Social Work Education 2021 57 sup1 S209 S223 10.1080/10437797.2021.1934209
*Willging, P. A., & Johnson, S. D. (2009). Factors that influence students’ decision to dropout of online courses. Journal of Asynchronous Learning Networks, 13(4), 115–127. https://eric.ed.gov/?id=EJ862360
*Yang, Y., & Cornelious, L. F. (2005). Preparing instructors for quality online instruction. Online Journal of Distance Learning Administration, 8(1) [online]. Retrieved 11 April 2023 from https://ojdla.com/archive/spring81/yang81.pdf
*Zhen, Y., Garthwait, A., & Pratt, P. (2008). Factors affecting faculty members’ decision to teach or not to teach online in higher education. Online Journal of Distance Learning Administration, 11(3) [online]. Retrieved 11 April 2023 from https://ojdla.com/archive/fall113/zhen113.pdf
|
PMC010xxxxxx/PMC10225761.txt |
==== Front
Int Rev Educ
Int Rev Educ
International Review of Education. Internationale Zeitschrift Fur Erziehungswissenschaft. Revue Internationale De Pedagogie
0020-8566
1573-0638
Springer Netherlands Dordrecht
10004
10.1007/s11159-023-10004-2
Original Paper
Interventions to improve refugee children’s access to education and quality learning: A scoping review of existing impact evaluations
http://orcid.org/0000-0001-5099-7759
Palik Júlia julpal@prio.org
Júlia Palik
is a Senior Researcher at the Peace Research Institute Oslo (PRIO). Her research focuses on the relationship between disarmament and conflict recurrence, and education and conflict. She has published on refugee education in Malaysia, on national regulatory restrictions on refugee rights to formal education, and on rebel-provided education in Yemen. Her work has been published in journals such as International Peacekeeping, the International Journal of Educational Development, and Comparative Social Research. Palik has provided consultancies for Save the Children, the Food and Agriculture Organization (FAO) and the United Nations Development Programme (UNDP). Between 2019 and 2022, she led the 6-week long Peace Research Course in Oslo.
http://orcid.org/0000-0002-5521-5610
Østby Gudrun gudrun@prio.org
Gudrun Østby
is a Research Professor at the Peace Research Institute Oslo (PRIO) and Deputy Editor of the Journal of Peace Research. Her research interests include the link between conflict, health, education and development aid effectiveness. Her work has appeared in journals such Demography, the International Studies Quarterly, the Journal of Conflict Resolution, the Journal of Peace Research, the Population and Development Review, the Review of Educational Research, and World Development. Østby has done consultancies for the United Nations Educational, Scientific and Cultural Organization (UNESCO), the United States Agency for International Development (USAID) and the World Bank, among others, on issues related to conflict, education, gender and horizontal inequalities. She is a member of the Inter-Agency Network for Education in Emergencies (INEE) Gender Task Team.
grid.425244.1 0000 0001 1088 4063 Peace Research Institute Oslo, Oslo, Norway
29 5 2023
2023
69 1-2 227247
12 5 2023
© UNESCO Institute for Lifelong Learning and Springer Nature B.V. 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Refugee children face numerous challenges in accessing quality education. In the past years, the number of interventions aiming to address these challenges has grown substantially. What is still scarce, however, is systematic evidence on what works to improve refugee children’s enrolment and learning. The authors of this article set out to find what robust quantitative evidence exists regarding interventions that seek to improve access to education and quality learning for refugee children. They conducted a first scoping review of quantitative peer-reviewed articles which evaluate the effect of specific interventions which aimed to improve access to education and/or quality learning for refugee children. While their literature search for the time-period 1990–2021 resulted in 1,873 articles, only eight of these fit the authors’ selection criteria. This low number indicates that there is a general lack of robust evidence as to what works to improve quality learning for refugee children. What the authors’ mapping of the research evidence does suggest is that cash transfer programmes can increase school attendance and that learning outcomes, such as second-language acquisition, can be improved through physical education, early childhood development programmes, or online game-based solutions. Other interventions, such as drama workshops, appear to have had zero effect on second-language acquisition. The authors conclude their article by addressing the limitations and implications of this body of interventions for future research.
Résumé
Interventions visant à améliorer l’accès des enfants réfugiés à l’éducation et la qualité de l’apprentissage : un examen approfondi des évaluations d’impact existantes – Les enfants réfugiés sont confrontés à de nombreuses difficultés pour accéder à une éducation de qualité. Ces dernières années, le nombre d’interventions visant à relever ces défis a considérablement augmenté. Cependant, les preuves systématiques de ce qui fonctionne pour améliorer la scolarisation et l’apprentissage des enfants réfugiés sont encore rares. Les auteures de cet article se sont mises en quête de preuves quantitatives solides concernant les interventions qui cherchent à améliorer l’accès à l’éducation et la qualité de l’apprentissage pour les enfants réfugiés. Elles ont procédé à une première analyse des articles quantitatifs examinés par des pairs qui évaluent l’effet d’interventions spécifiques visant à améliorer l’accès à l’éducation et/ou la qualité de l’apprentissage pour les enfants réfugiés. Bien que leur recherche documentaire pour la période 1990–2021 ait donné lieu à 1 873 articles, seuls huit d’entre eux correspondaient aux critères de sélection des auteures. Ce faible nombre témoigne d’une insuffisance généralisée de preuves solides sur ce qui fonctionne pour améliorer la qualité de l’apprentissage pour les enfants réfugiés. Ce que la cartographie des résultats de recherche des auteures suggère, c’est que les programmes de transferts monétaires peuvent augmenter la fréquentation scolaire et que les résultats d’apprentissage, tels que l’acquisition d’une seconde langue, peuvent être améliorés grâce à l’éducation physique, aux programmes de développement de la petite enfance ou aux solutions axées sur les jeux en ligne. D’autres interventions, telles que les ateliers de théâtre, semblent n’avoir eu aucun effet sur l’acquisition d’une seconde langue. Les auteures concluent leur article en abordant les limites et les implications de cet ensemble d’interventions pour la recherche future.
Keywords
Refugee education
Scoping review
Access to education
Learning quality, quality education
Fourth sustainable development goal (SDG 4)
Research Council of Norway287047 Østby Gudrun issue-copyright-statement© UNESCO Institute for Lifelong Learning 2023
==== Body
pmcIntroduction
In 2020, 42 per cent of the 82.4 million forcibly displaced people worldwide were children, and almost half of all refugee children were out of school (UNHCR 2020, p. 6, 2021, p. 6). Despite a growing number of education interventions targeting refugee learners over the past years, access and quality remain significant challenges. Access to education remains uneven across regions, between camp and urban settings, between different operations within the same country, and between boys and girls (Dryden-Peterson 2011, p. 32). Refugee children, especially girls, lack access to post-primary education (UNHCR 2021). Access is partly conditioned by the national legislation of host states (Dryden-Peterson 2011). While some countries, such as Uganda, allow refugees to access public education, others, like Bangladesh, only offer informal education for non-registered refugees (Dupuy et al. 2022). Furthermore, many refugee children are living in learning poverty, i.e. despite being in school, they are unable to read or to understand a simple text by the age of 10, or do not achieve basic numeracy skills (Piper et al. 2020; Saavedra and Bousquet 2020). These challenges significantly endanger the achievement of the fourth Sustainable Development Goal (SDG 4), which aims to “ensure inclusive and equitable quality education and promote lifelong learning opportunities for all” by 2030 (UN 2015, p. 19). Building up for some time already, the education crisis has only been amplified by the COVID-19 pandemic (World Bank 2022).
The past few years have seen an upward trend in humanitarian funding for education (Dupuy et al. 2019), while a marked increase in different education interventions and non-governmental organisations (NGOs) is accompanied by increasing pressure from funders to provide empirical evidence regarding the impact of their interventions. This state of affairs prompted our overarching research question, What robust quantitative evidence exists regarding interventions that improve access to education and quality learning for refugee children? Briefly, our scoping review of the existing literature, presented in more detail below, revealed that there appears to be little rigorous quantitative research testing the impact of specific interventions to improve education for refugees. Most studies appear to be observational, with a staggeringly low number of empirically robust quantitative studies. We reviewed 1,873 peer-reviewed journal articles published in the period 1990–2021, only eight of which were rigorous quantitative studies that examined the impact of various educational interventions for refugee children.
While literature reviews on refugee education have been conducted before (e.g. Hamilton and Moore 2004), these have either focused exclusively on high-income countries (McBrien 2005) or have examined only higher education (Ramsay and Baker 2019; Streitwieser et al. 2019). Other reviews have analysed interventions in emergency settings more broadly (i.e. beyond situations of forced displacement) and included both qualitative and quantitative studies (Burde et al. 2015, 2017, 2019). In our own scoping review presented here, we zoomed in on education for refugees employing a strict methodological criterion to only include studies which provide robust quantitative evidence regarding the causal impact of various programmes. Our rationale for doing so is that experimental studies can be powerful scientific tools to generate evidence to inform policies. Randomised controlled trials (RCTs) are considered to be the “gold standard” of impact analysis, providing information on the causal impact of specific interventions since participants are assigned randomly to a treatment or a control group, and these groups are followed across time and compared (Riddell and Niño-Zarazúa 2016).
Quasi-experimental or experimental research
While conducting quasi-experimental or experimental research in situations of forced displacement is challenging in terms of security-related, ethical and practical reasons, the very low number of solid empirical quantitative studies on refugee education is alarming. In the worst case, while donors increase their funding for refugee education interventions, they might risk financing, in the absence of robust impact evaluations, programmes which have little or no effect on improving refugee children’s access to education or quality learning. Robust evidence from impact evaluations is therefore critical to establishing that the effects found can confidently be attributed to the intervention and not to other factors. Such systematic evidence is crucial for donors so that they fund programmes that actually work, and which in turn can improve the lives of refugee children.
The scarce robust experimental and quasi-experimental evidence we did find to be available focuses mostly on programmes implemented in high-income countries and in school settings. This does not, however, reflect the reality on the ground, since most refugees reside in low- and middle-income countries and many of them can access education only in camps. Furthermore, we do not know enough about the long-term impact of interventions, or how the same programme might have different impacts on newly arrived refugee children compared to settled refugee children. The interventions reviewed, however, do seem to point to several promising results: language acquisition – one of the most important barriers – can be improved by language learning-enhanced physical education classes, enrolment in early childhood education programmes, and through online, game-based solutions. Attendance in child and youth learning centres (CYLCs) improved literacy and numeracy skills for both girls and boys, and cash transfer programmes increased attendance rates. However, given the low and heterogeneous number of studies, it is hard to say to what extent the findings may be applicable across different countries and contexts.
Rigorous qualitative and descriptive studies
Importantly, quasi-experimental and experimental studies also have their limitations. The validity of the results from such studies is intrinsically tied to the assumptions the research is based on and the strength of those assumptions. Furthermore, these methods are not able to unpack the mechanisms underlying the overall treatment effects (i.e. how changes in the dependent variable(s) occur). Rigorous qualitative and descriptive studies can and should therefore play an important role in filling these gaps (Gopalan et al. 2020, pp. 232–233). Qualitative methods are able to provide a deeper understanding of the context and how individuals experience and perceive various interventions. Focus groups can be appropriate qualitative tools to uncover, discuss and understand commonly held perceptions and attitudes; individual interviews open up researchers’ insights into privately held opinions or experiences (e.g. Busch et al. 2021; Hattar-Pollara 2019); and in participatory research, education communities work together to better understand and respond to problems in their day-to-day work, thereby contributing to programme quality and sustainability (e.g. Cappella et al. 2011).
In the next section of this article, we provide an overview of the barriers associated with access to education and quality learning for refugees, and potential interventions that can address them. This is followed by a description of our search method and data collection. After presenting the articles we selected for review, we offer a discussion of our findings. We conclude our article with some implications for policy and needs for future research.
Refugee education: barriers to access and quality learning
Refugees face several challenges when it comes to their access to education. These barriers vary between camp and urban settings and between girls and boys. In the case of urban settings, host states might have restrictive regulatory frameworks regarding the inclusion of refugees in national education systems (Dupuy et al. 2022). Refugee children often lack necessary identification (ID) documents and certificates, making enrolment virtually impossible. Insufficient infrastructure, crowded classrooms, and the lack of appropriate hygiene facilities can prevent children, especially girls, from going to school. Distance from schools and safety issues related to travel are also major concerns, again, especially for girls. Even if education is free, indirect costs of schooling, such as uniforms and school material, are often impossible for refugee families to finance. Since refugee children often do not speak the language of the host country, linguistic challenges remain one of the most important barriers in enrolment or placing children in age-appropriate grades. When refugee children are unable to speak the language of instruction, their academic progress can be derailed, potentially leading to dropout. There might be no education opportunities beyond the primary level, and the lack of parents’ income often means that they require their child to work instead of attending school. Even those refugee students who were able to enrol in public schools might face discrimination and violence at the school, leading to high dropout rates. Furthermore, certain belief systems, such as child marriage, or the prioritisation of boys’ educations over girls, are major reasons for dropout (Burde et al. 2015; Mendenhall et al. 2017; UNHCR 2009a, pp. 9–11).
The most typical ways of measuring quality of education focus either on inputs (e.g. student–teacher ratios, percentage of qualified teachers, the extent to which certificates are recognised) or on learning outcomes (i.e. academic performance) (Dryden-Peterson 2011, pp. 31–32). However, quality can also be measured in terms of “process” (including participation), “perceptions” or “impact” (e.g. occupational opportunities) (Millán et al. 2019). Delivering quality education for refugee children can be challenging because schools often lack the necessary infrastructure and teaching materials, or there is a high student–teacher ratio, moreover, there are often only a few female teachers. Available teachers are often not well trained to consider the specific needs of refugee children, or they do not speak the refugee children’s first language. Discrimination and language barriers not only impact access, but also the quality of education (Burde et al. 2015).
Typical interventions aimed at increasing refugees’ access to education include information campaigns, accelerated education programmes (AEPs), technology-based solutions for distance learning, cash transfer programmes, opening an afternoon shift in public schools, the provision of school meals, uniforms and school materials, safe transportation, intense language education or bilingual curricula (Burde et al. 2015; UNHCR et al. 2019). When it comes to addressing barriers to quality learning, potential interventions include involvement of teachers from the refugee community, lowering the student–teacher ratio, grouping of students according to their academic performance, provision of teaching materials in multiple languages, the establishment of community centres for out-of-school children, parent-teacher associations (PTAs), and technology-based solutions (Burde et al. 2015; UNHCR 2009b).
While there is a plethora of specific education interventions, there is considerable variation when it comes to whether these programmes were evaluated or not, and if evaluated, what method was used. For example, an interview-based evaluation of an Australian school support programme called Ucan21 found that the programme raised awareness amongst school staff of the specific educational needs of refugee students which led to an increase in the use of interpreters and introduced additional support through the enrolment period (Block et al. 2013, p. 1350). While qualitative evaluations are important, our review focused on quasi-experimental and experimental studies because these research designs can allow for the determination of causality and are able to isolate the key factors influencing the outcome of a particular intervention.
Literature review: methodology
The main aim of our scoping review was to identify what robust quantitative evidence exists on the effectiveness of interventions aiming to improve access to and quality of education for refugee children. Our search strategy was guided by the following two research questions:What interventions can increase refugee children’s school attendance and enrolment rates?
What interventions can impact positively the quality of education that refugee children receive?
To identify studies which evaluate the impact of interventions on education access and quality learning for refugee children, we searched in three major electronic databases (Web of Science [WoS], JSTOR and Google Scholar) and then employed a pearl growing search strategy. In the pearl growing process we reviewed the bibliography of identified articles to supplement the number of candidates selected for the final review. We stopped running our search in September 2021. We used Boolean operators, truncation (*), and searched for the keywords and their combinations, shown in Table 1.Table 1 Keywords included in the literature search
Target group Outcome Intervention
Refuge* Educat* Program*
IDPs Access to education Impact
Forced displacement Enrol* Intervention
Asylum-seeker* Attainment Randomised control trial
Migrant Quality learning Experiment*
Learning quality Quasi-experiment*
Quality education
Literacy
Numeracy
Academic outcome*
Learning outcome*
School*
IDP = internally displaced person
We applied the following inclusion criteria:English language, peer-reviewed article published between 1990 and March 2021
Interventions targeted at refugees, forced migrants, and internally displaced persons (IDPs) below the age of 18 years
Evaluations of specific interventions which addressed refugee children’s access to education and/or learning quality
Studies that were using a quantitative methodology, quasi-experimental or experimental research design, preferably randomised control trials
Interventions delivered in schools, refugee camps or community settings
It is important to note that among extant rigorous quantitative studies, only a few examine the effectiveness of access and quality of refugee education interventions, while there are a number focusing on psychosocial interventions. We excluded studies which examined interventions aiming to improve the psychosocial well-being of refugee children because several systematic literature reviews focusing on such interventions have already been conducted (e.g. Frounfelker et al. 2020; Simenec and Reid 2020; Sullivan and Simonson 2016; Tyrer and Fazel 2014).
Figure 1 shows our quorum flowchart. In total, by September 2021, we identified 1,873 articles through our abstract search method. First, we read the abstracts and categorised articles as 0 = excluded, 1 = included, 99 = to be decided. Both of us coded the articles independently. We compared our findings and discussed cases when we coded differently. Inter-coder reliability was high, and discrepancies were resolved by discussion. In the first round we screened the abstracts of the 1,873 articles and selected 39 potential articles for full review. Of the excluded studies, most did not evaluate an intervention; did not sample refugees; or did not apply a quantitative research strategy. Next, we reviewed the entire text of each of the 39 selected articles, only five of which proved to meet our strict methodological inclusion criteria.Fig. 1 Quorum flowchart
Due to the low number of studies resulting from this last step, we applied two additional search strategies to our shortlist of 39 articles. First, we conceded some flexibility regarding the robustness of the study. For example, studies were accepted if they had a control and intervention group even if the participants were not randomly assigned to these groups or if the study lacked a control a group. Second, we applied a snowball technique whereby we reviewed the bibliographies of the selected five articles to identify potential additional works. This step resulted in the inclusion of three more articles, increasing the total number of articles that met our inclusion criteria to eight.
A possible reason for the relatively low number of articles is that research on vulnerable groups such as refugees is fraught with practical and ethical challenges. Practical challenges entail difficulties in accessing refugee camps, overcoming language barriers, and the generally high costs associated with conducting RCTs. When it comes to ethical barriers associated with RCTs, randomisation can be an issue. For example, when testing an intervention, one has to rely on sufficiently large numbers of children and youth with similar levels of needs/vulnerability within the population while also acknowledging that it is not possible to grant universal access to a specific intervention. Other ethical challenges relate to the risk of re-traumatisation, e.g. during survey interviews, which can (at least in part) be overcome through complementary qualitative approaches (Gaywood et al. 2020; Habib 2019).
The eight articles that were included in our shortlist for the final review are listed in Table 2. We collected information on the following variables for each article: spatio-temporal scope of intervention and target population; type of intervention; outcome of interest; research design; and main findings. Due to the variety of methodologies, sample sizes and intervention types, we provide qualitative discussion of these variables.Table 2 Quantitative evidence on what works to improve access to education and quality of learning for refugee children
Study Target population; spatio-temporal scope of intervention Outcome of interest Type of intervention Research design N Main Finding(s)
Busch et al. (2021) Newly arrived refugees aged 3–7 in Germany, 5 months of programme attendance Language, cognition, motor skills, socio-emotional behaviour Early-Childhood Development programme (ECD) Cross-sectional, longitudinal, and quasi-experimental between-group design 207 (152 refugee students attending preschool ECD + 55 refugee students (not previously enrolled in ECDs) attending first grade Longer duration of ECD attendance predicted better German language skills.
de Hoop et al. (2019) Syrian refugees aged 5–14 in Lebanon, 4 months School enrolment and attendance Cash transfer programme Geographical regression discontinuity design 1,440 households with 1,784 children aged 5–9 and 1,647 children aged 10–14 School enrolment and afternoon shift enrolment did not increase through the programme. Annual education expenditure per child rose by about USD 74 (at the baseline it averaged USD 62), probability of commuting to school by bus increased by 14 percentage points. The number of days of school attendance rose by 0.6 days a week among children enrolled in the afternoon shift schools.
Krüger (2018) Refugees aged 6–11 in Germany, 6 lessons Domain-specific vocabulary, listening comprehension and use of local prepositions Language-enriched physical education (PE) Pre- and post-test design 61 (31 treatment + 30 control) Domain-specific vocabulary and listening comprehension scores were higher for the intervention group. No effect on the use of local prepositions (e.g. “upon”, “behind”).
Meloche et al. (2020) Refugee students aged 11–14 in the United States, 2 academic years Academic outcomes, non-academic outcomes, college readiness Community school (CS) and refugee centre services (including 13 different services; for the full list see p. 5 of the study) Quasi-experimental design 3,426 (1,045 treatment + 2,381 control group) Attendance: No differences between refugee and comparison by school groups. In 8th grade, both groups (refugee and non-refugee) had significantly higher average out-of-school suspensions than students in the intervention school refugee group. Academic outcomes: Regardless of refugee status, CS school students outperformed comparison school students in all content areas.
Metzler et al. (2021) Somali refugees aged 6–17 in a camp in Ethiopia, follow-up data collected between 3 and 6 months Literacy, numeracy, psychosocial well-being, protection needs Child and youth learning centres (CYLC) Pre- and post-test design Baseline assessment: 925 (587 children aged 6–11 + 338 youth aged 12–17; follow-up: 693 (437 aged 6–11 + 256 aged 12–17) For younger children and youth alike, the mean literacy and numeracy scores rose as a result of the programme.
Rousseau et al. (2014) Secondary school refugee students in special classes in Canada, 12 weeks Mental health and academic outcomes in French and maths School-based theatre intervention Cluster randomised trial design 477 (theatre intervention: 157+ tutorship intervention: 180+ control: 140) Maths and French grades: no significant difference between the 3 groups
Rousseau et al. (2007) Newly arrived refugees aged 12–18, attending integration classes in Canada, 9 weeks Emotional and behavioural symptoms, self-esteem, school performance Drama therapy programme Pre- and post-test design 123 (treatment: 66 + control: 57) Maths: significant improvement in mathematics for the experimental group. French: No significant improvement.
Sirin et al. (2018) Syrian refugee children aged 9–14 in Turkey, 4 weeks Turkish language skills, cognitive skills, coding, mental health Online, game-based intervention (curriculum included Cerego, Alien game, Code.org, and Minecraft. For details, see p. 11 of the study) Pre and post-test design 147 (treatment: 75 + control: 72) Students in the intervention group had significantly higher language scores than children in the control group.
Results
Although our search included articles published between 1990 and 2021, the eight studies that met our inclusion criteria were all published in or after 2007. The interventions took place in the United States, Canada, Lebanon, Germany, Ethiopia and Turkey. Only one study was conducted in a camp (in Ethiopia: Metzler et al. 2021), the others examined interventions implemented in public or community school settings.
Two of the eight studies examined interventions which were aimed at increasing access to education (de Hoop et al. 2019; Meloche et al. 2020), whereas the remaining six interventions were aimed at improving the quality of learning focusing either on second-language acquisition or on various academic outcomes (Busch et al. 2021; Krüger 2018; Metzler et al. 2021; Rousseau et al. 2007, 2014; Sirin et al. 2018). Sample sizes ranged from 62 (Krüger 2018) to 3,426 (Meloche et al. 2020). Some interventions examined multiple different outcomes such as academic outcomes and psychosocial well-being (Busch et al. 2021; Metzler et al. 2021; Rousseau et al. 2007, 2014), or both academic and non-academic outcomes (Meloche et al. 2020). The eight studies each examined different types of interventions: community school practices including 13 special services, cash transfer, language learning-enhanced physical education, enrolment in child and youth learning centres (CYLCs), an online game-based intervention, enrolment in an early childhood education programme, and two drama therapy workshops. Four out of the eight studies reported gender effects (de Hoop et al. 2019; Metzler et al. 2021; Rousseau et al. 2007, 2014). As for methodological sophistication, only one study applied a randomised control trial methodology (Rousseau et al. 2014); three applied a quasi-experimental design (Busch et al. 2021; de Hoop et al. 2019; Meloche et al. 2020); and four applied a pre- and post-test design (Krüger 2018; Metzler et al. 2021; Rousseau et al. 2007; Sirin et al. 2018). The target populations of all eight studies were refugee children aged between 3 and 18 years.
Interventions aimed at increasing refugees’ access to education
We identified only two articles which examined interventions aimed at improving refugees’ access to education (de Hoop et al. 2019; Meloche et al. 2020). In terms of evaluating enrolment, this low number is in itself worrying, given the high number of refugees who are deprived of education or drop out of school.
The two articles addressed different types of barriers to access, and both applied a quasi-experimental research design. Alysha Meloche et al. (2020) conducted their study in the United States, where they examined the different impacts of a school that implemented both community school (CS) and refugee centre programming compared to two other schools from the same district in terms of “academic, non-academic and college readiness outcomes of urban immigrant and refugee youth”, as they state in the title of their article (ibid.). The CS school’s refugee programme entailed 13 special services, such as citizenship classes, extended language support, mental health workshops and trauma-sensitive schooling, amongst others. The study further examined whether various academic and non-academic outcomes of English Language Learners (ELLs), a proxy for refugee status, differed from those of non-ELL students across the two school types (CS school and two comparison schools).
In this section, we report only on the results pertaining to non-academic outcomes, i.e. attendance rates and behavioural incidence (measured as out-of-school suspension rates). Meloche et al. (ibid.) found no statistical differences in attendance rates by ELL status or school type (each subgroup maintained an attendance rate of above 90 per cent over the period the study covered). The relatively low rates of dropouts however could not be causally connected to the CS practice given that there was a district-wide policy implemented to reduce absenteeism. The authors found that ELL students had more behavioural incidents than non-ELL students, however this was not present in the CS school. In fact, ELL students in the CS school were at the lowest risk of dropout, which is likely to be the result of the school counselling and trauma-sensitive training in place.
In the other study in this category, Jacobus de Hoop et al. (2019) evaluated an intervention addressing two critical barriers to access: the cost of transportation and foregone income because children are at school instead of working. The cash transfer programme (No lost Generation)2 targeted Syrian displaced children in Lebanon. The intervention provided cash for the benefit of children enrolled in afternoon shifts at public primary schools and was designed to cover costs of commuting to school and to compensate the household for income foregone because children were at school instead of working. The authors applied a geographical discontinuity design to compare children in pilot governorates with children in neighbouring governorates and found no evidence that the cash transfer programme led to an increase in enrolment rates. However, children who received the programme benefits experienced increased household expenditures on education, and an increase in the probability of commuting to school by bus. Most importantly, children enrolled in afternoon shifts in the pilot areas spent approximately 20 per cent more time in school relative to children in the comparison areas. The results were similar across different age groups and between boys and girls.
Interventions aimed at improving the quality of education for refugees
We identified seven studies which examined interventions aimed at improving the quality of education for refugee children (Busch et al. 2021; Krüger 2018; Meloche et al. 2020; Metzler et al. 2021; Rousseau et al. 2007, 2014; Sirin et al. 2018). These included one experimental study (Rousseau et al. 2014), two quasi-experimental investigations (Busch et al. 2021; Meloche et al. 2020), and the remaining four used pre-and post-test design (Krüger 2018; Metzler et al. 2021; Rousseau et al. 2007; Sirin et al. 2018). Quality was measured in terms of learning outcomes, including literacy, numeracy and language skills.
One cluster randomised control trial examined the impact of a 12-week drama theatre and tutorship intervention programme in Canada on the maths and French grades of secondary school refugee students who were placed in special classes due to behavioural problems (Rousseau et al. 2014). The study found that for both maths and French grades, there was no significant difference in either the theatre or the tutorship intervention groups compared to the control group. It is interesting to note that an earlier study in our sample (Rousseau et al. 2007) found that a similar drama theatre intervention resulted in a significant improvement in mathematics for the experimental group compared to the control group, but no significant improvement was reported in either group with regard to French results. The main difference between the two studies was that the one conducted in 2007 focused on newly arrived refugees, while the 2014 randomised control trial examined refugee children who were first- and second-generation immigrants presenting with emotional and behavioural problems. This is an important difference between the two target populations since it shows that intervention effectiveness is likely to be conditioned by refugees’ length of stay in their host community.
One of the two quasi-experimental studies was the one by Meloche et al. (2020), already featured in the previous section above, which examined the different impacts on refugee students’ academic and non-academic outcomes by a school implementing community school (CS) and refugee centre programming compared to two other schools from the same district (ibid.). The study further examined whether various academic and non-academic outcomes of English Language Learners (ELLs), a proxy for refugee status, differed from those of non-ELL students across the two school types (CS school and two comparison schools). The authors found significant differences at all grade levels on overall grade point average (GPA) in each area (English, maths, science, social studies) at the end of year grades (p < .001). In sixth grade, non-ELL students attending the CS school outperformed all students in the comparison school in all measures. In seventh and eighth grades, both non-ELL and ELL students at the CS school scored significantly higher than non-ELL and ELL students in the comparison school across all measures. These results demonstrate that CS practices can benefit both refugee and non-refugee students’ learning outcomes.
The second of the two quasi-experimental studies we identified focused on the impact of a physical education (PE) intervention on second-language acquisition (Krüger 2018) in Germany. The intervention group received language-enriched PE sessions, while the control group did not receive any treatment. The authors tested refugee students’ domain-specific vocabulary learning, listening comprehension and use of prepositions. The intervention group showed better performance in the domain-specific vocabulary learning and listening comprehension tests, but no significant effect was found in case of the use of local prepositions (such as “upon”, “behind” etc.).
Another study which focused on second-language acquisition examined the impact of an online and game-based learning intervention (Cerego) for Syrian refugee children on Turkish language acquisition (Sirin et al. 2018).3 In this particular study, learning through Cerego involved using 20 themed sets, each with 10–15 Turkish words and visual images. The authors found that students in the intervention group had significantly higher language scores than children in the control group.
The third study on language acquisition was carried out in Germany and examined the impact of an early childhood development programme (ECD) on newly arrived refugee children’s German language skills (Busch et al. 2021). The cross-sectional and longitudinal analyses found that attendance in the ECD was positively linked to improvements in German language skills, making ECD an important stage for second-language acquisition which can eventually facilitate a smoother transition to school.
The only study in our sample which was conducted in a camp setting examined the impact of child-and-youth-friendly spaces (CYLCs) at Buramino Camp in Dollo Ado, Ethiopia on educational, psychosocial and protection outcomes (Metzler et al. 2021). The study tested the impact of CYLC on literacy and numeracy skills of children enrolled in the programme between 3 and 6 months after the baseline data collection. The study found that younger children who attended CYLCs had major improvements in literacy and numeracy, although gains in numeracy were higher for boys than for girls. Older children who attended CYLCs recorded even bigger increases in both literacy and numeracy than younger ones, and older boys showed significantly greater improvements than girls both in literacy and numeracy.
Discussion
Despite being few in number, the studies we reviewed show that it is indeed possible to carry out rigorous quantitative evaluations in the context of forced displacement. Amongst the studies we reviewed, we found that cash transfer programmes can be effective in increasing the time refugee children spend in school by providing the financial means for families to send their kids to school. However, there was no evidence that cash transfers attracted a higher number of refugee children to school. Hence to increase enrolment, other types of intervention, such as targeted information campaigns and home visits by teachers, need to be evaluated.
There are some promising results regarding interventions to improve second-language acquisition, a major barrier both to access and quality learning. Evidence from our reviewed articles suggests that enrolment in early childhood development programmes can be appropriate for developing second-language skills and that in certain contexts, online, game-based solutions can also contribute to language acquisition. Child and youth learning centres in camp settings can improve refugee children’s numeracy and literacy skills, although more research is needed to find out how to improve these skills specifically for girls. Lastly, there was some evidence that theatre interventions can positively impact newly arrived refugees’ maths grades.
Despite these positive findings, the quantitative studies reviewed do suffer from some limitations regarding methodology, temporal aspects, target population and scope of analysis. Also, there is a lack of studies carried out in the Global South. Below, we discuss these limitations.
Limitations
Methodological limitations
The few studies that fit our inclusion criteria nonetheless exhibited some methodological limitations. The sample sizes were often small (Krüger 2018), there were often no control groups, only pre- and post-treatment evaluations (Metzler et al. 2021), and participants were often not randomly allocated to intervention and control groups (Busch et al. 2021).
Temporal limitations
Most studies evaluated the intervention shortly after the termination of the programme; hence we know little about the long-term effects. Furthermore, evaluations rarely considered the length of displacement which is however likely to influence the intervention’s impact (Rousseau et al. 2007, 2014).
Limitations regarding the target population
Amongst the eight articles we reviewed in detail, four reported gender-specific findings (de Hoop et al. 2019; Metzler et al. 2021; Rousseau et al. 2007, 2014). Results were often the same for boys and girls, but in one study, boys performed better than girls in both literacy and numeracy scores (boys also had higher baseline scores). This might have been caused by enrolment differences before displacement, gender differences in education engagement and variation in caregivers focus on girls’ learning (Metzler et al. 2021). This finding highlights the need for developing programmes specifically designed for helping refugee girls access quality education. There is also a need to conduct more research on internally displaced children. We did not find any study that addressed the wider environment of refugee children, i.e., teacher, parents and peers, even though these actors have been identified as important in quality learning (Burde et al. 2015).
Lack of comparison across intervention contexts
Studies were conducted in high-income countries (United States, Canada, Germany), one in an upper-middle-income country (Turkey), one in a lower-middle-income country (Lebanon), and only one study was conducted in a camp setting (Ethiopia). We did not find any study that compared the effectiveness of a particular intervention between low- and middle-income and high-income countries, hence we know little about the transferability of interventions to other contexts. Furthermore, there is a notable absence of studies which examine the situation of refugees displaced due to natural disasters; this gap has already been highlighted in earlier literature. Knowing what works regarding disaster risk reduction (DRR) education, or interventions that support children after disasters is urgently needed (Burde et al. 2015, 2017).
Lack of focus on individual project components
Studies usually examine interventions as a whole and rarely focus on specific components, making it difficult to understand which part of the intervention exactly influences educational outcomes. For example, the community and refugee school practice in the United States examined by Meloche et al. (2020) consisted of 13 special service components such as counselling, parent–teacher associations and cultural training, among others, but the research design did not envisage examining these components individually. This lack of disaggregation might mean that there were multiple pathways through which access to education or quality learning was impacted (equifinality) but we were unable to identify which project component was influential to what degree.
Conclusion
Even though millions of refugee children are in dire need of education, our systematic mapping of the literature indicates that there are only very few empirically rigorous quantitative studies that provide robust evidence on “what works” in improving these children’s access to education and quality learning. The eight studies which met our selection criteria suggest some promising results (e.g., cash transfers can increase educational attainment). However, since the studies are so few and varied with respect to geographical location, type of intervention and sample size, only limited conclusions can be drawn about the interventions’ general effectiveness and scalability.
Documenting the lack of such impact evaluations is an important finding in its own right, and one that demonstrates there is a significant need for future RCTs to provide information on the causal, direct impact of specific interventions. However, quantitative studies, when resources allow, should be combined with qualitative methods. Evaluations which combine these two research methods are the best suited to provide a comprehensive assessment of the effectiveness of a particular intervention.
Based on our scoping review, we have identified some areas for future research. First, as displacement has become increasingly protracted, in contexts where it is possible, it is important to compare a particular intervention’s impact across newly arrived refugees, refugees who have been staying in the host country/community for a longer period, and national/local citizens. Second, there is a need for more disaggregated data by sex, age groups, ethnicity, race, class, religious faith and disabilities. Knowing how the intersection of two or more of these attributes can promote or hinder access to education or quality learning is critical. Finally, it is crucial that interventions introduced during the COVID-19 pandemic-related school closures are evaluated. Remote or digital learning has often been unavailable for refugee children due to the lack of connectivity and necessary equipment. The increasing digital divide between different groups of learners can exacerbate existing barriers to access and quality.
Our systematic scoping review is an important step towards understanding the current refugee education intervention landscape. When it comes to the evaluation of interventions, we see two important structural challenges. First, while there is an increasing number of education interventions, most of them remain unevaluated. Our mapping of existing impact evaluations on what works to improve access to education and quality of learning for refugee children demonstrates that available rigorous quantitative evidence is scarce and often inconclusive. This is problematic because the cost-effectiveness or the overall performance of these programmes often remains unknown. Second, while the importance of multistakeholder engagement – researchers, practitioners and policymakers – is often highlighted (e.g. Siarova and van der Graaf 2022) in the field of refugee education, in reality significant gaps remain in interactions between these various actors. This in turn negatively impacts the development of sound programme design, and the collection and evaluation of empirical evidence on what works, when, and under what conditions in improving refugees’ access to and quality of education. Providing regular rigorous systematic reviews on refugee education interventions and fostering close collaboration between humanitarian and development actors, researchers and governments is crucial in improving refugee children’s lives and achieving SDG 4’s stipulated inclusive and equitable quality education for all by 2030.
Acknowledgements
Funding: This research was funded by the Research Council of Norway, grant # 287047. Authors’ contributions: Julia Palik performed the literature search, the coding, the data analysis, and drafted the article. The literature review was initiated by Gudrun Østby. Østby performed data checking, supplementary coding, and critically revised the manuscript. Both authors read and approved the final manuscript.
1 “Ucan2 is an innovative intervention programme … aimed at supporting mental health and wellbeing and improving settlement outcomes for young people with refugee backgrounds. The Ucan2 intervention addresses the multiple and interlinked causes of social exclusion and targets young people between the ages of 16 and 24” (Block et al. 2013, pp. 74–75). For more information, visit https://foundationhouse.org.au/specialised-programs/ucan2/ [accessed 1 May 2023].
2 “[T]he No Lost Generation Programme (NLG) … [known] locally as Min Ila (‘from to’) … [is an] initiative of the government of Lebanon, the United Nations Children’s Fund (UNICEF), and the World Food Programme (WFP)” (de Hoop et al. 2019, p. 107). For more information, visit https://www.nolostgeneration.org/ [accessed 1 May 2023].
3 “Cerego … is an adaptive learning engine that allows users to quickly create sets of items, such as vocabulary, which can then be presented to learners using a spaced repetition paradigm” (Sirin et al., p. 11). For more information, visit https://www.cerego.com/ [accessed 1 May 2023].
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
Block K Warr D Gibbs L Riggs E Addressing ethical and methodological challenges in research with refugee-background young people: Reflections from the field Journal of Refugee Studies 2013 26 1 69 87 10.1093/jrs/fes002
Burde, D., Guven, O., Kelcey, J., Lahmann, H., & Al-Abbadi, K. (2015). What works to promote children’s educational access, quality of learning, and wellbeing in crisis-affected contexts. DFID Education Rigorous Literature Review series. London: United Kingdom Department for International Development (DFID). Retrieved 28 April 2023 from https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/470773/Education-emergencies-rigorous-review2.pdf
Burde D Kapit A Wahl RL Guven O Skarpeteig MI Education in emergencies: A review of theory and research Review of Educational Research 2017 87 3 619 658 10.3102/0034654316671594
Burde, D., Lahmann, H., & Thompson, N. (2019). Education in emergencies: “What works” revisited. Education and Conflict Review, 2, 81–88. Retrieved 28 April 2023 from https://discovery.ucl.ac.uk/id/eprint/10081593/1/Burde_Article_14_Burde.pdf
Busch J Cabrera N Ialuna F Buchmüller T Leyendecker B Refugee children’s early development during attendance of specialized preschool programs and transition into first grade in Germany Early Education and Development 2021 33 8 1304 1325 10.1080/10409289.2021.1970427
Cappella E Jackson DR Bilal C Hamre BK Soulé C Bridging mental health and education in urban elementary schools: Participatory research to inform intervention development School Psychology Review 2011 40 4 486 508 10.1080/02796015.2011.12087526
de Hoop J Morey M Seidenfeld D No lost generation: Supporting the school participation of displaced Syrian children in Lebanon The Journal of Development Studies 2019 55 sup1 107 127 10.1080/00220388.2019.1687875
Dryden-Peterson, S. (2011). Refugee education: A global review. Geneva: UNHCR. Retrieved 28 April 2023 from https://www.unhcr.org/research/evalreports/4fe317589/refugee-education-global-review-sarah-dryden-peterson-november-2011.html
Dupuy, K., Palik, J., & Østby, G. (2019). Walk the talk? Review of Financing for Education in Emergencies, 2015–2018. Oslo: Save the Children Norway. Retrieved 28 April 2023 from https://resourcecentre.savethechildren.net/document/walk-talk-review-financing-education-emergencies-2015-2018/
Dupuy, K., Palik, J., & Østby, G. (2022). No right to read: National regulatory restrictions on refugee rights to formal education in low- and middle-income host countries. International Journal of Educational Development, 88, Art. no. 102537. 10.1016/j.ijedudev.2021.102537
Frounfelker RL Miconi D Farrar J Brooks MA Rousseau C Betancourt TS Mental health of refugee children and youth: Epidemiology, interventions, and future directions Annual Review of Public Health 2020 41 159 176 10.1146/annurev-publhealth-040119-094230
Gaywood D Bertram T Pascal C Involving refugee children in research: Emerging ethical and positioning issues European Early Childhood Education Research Journal 2020 28 1 149 162 10.1080/1350293X.2020.1707369
Gopalan M Rosinger K Ahn JB Use of quasi-experimental research designs in education research: Growth, promise, and challenges Review of Research in Education 2020 44 1 218 243 10.3102/0091732X20903302
Habib, R. R. (2019). Ethical, methodological, and contextual challenges in research in conflict settings: The case of Syrian refugee children in Lebanon. Conflict and Health, 13(1), Art. no. 29. 10.1186/s13031-019-0215-z
Hamilton R Moore D Educational interventions for refugee children: Theoretical perspectives and implementing best practice 2004 New York Routledge
Hattar-Pollara M Barriers to education of syrian refugee girls in Jordan: Gender-based threats and challenges Journal of Nursing Scholarship 2019 51 3 241 251 10.1111/jnu.12480 30977586
Krüger, M. (2018). Second language acquisition effects of a primary physical education intervention. A Pilot study with young refugees. PLoS ONE, 13(9), Art. no. e0203664. 10.1371/journal.pone.0203664
McBrien JL Educational needs and barriers for refugee students in the United States: A review of the literature Review of Educational Research 2005 75 3 329 364 10.3102/00346543075003329
Meloche, A., Provinzano, K., Sondergeld, T., & Moy, M. (2020). Examining academic, non-academic, and college readiness outcomes of urban immigrant and refugee youth (IRY) in full-service community schools. Children and Youth Services Review, 119, Art. no. 105489. 10.1016/j.childyouth.2020.105489
Mendenhall M Russell SG Buckner E Urban refugee education: Strengthening policies and practices for access, quality and inclusion 2017 New York, NY Teachers College, Columbia University
Metzler J Jonfa M Savage K Ager A Educational, psychosocial, and protection outcomes of child- and youth-focused programming with Somali refugees in Dollo Ado, Ethiopia Disasters 2021 45 1 67 85 10.1111/disa.12392 31322750
Millán TM Barham T Macours K Maluccio JA Stampini M Long-term impacts of conditional cash transfers: Review of the evidence The World Bank Research Observer 2019 34 1 119 159 10.1093/wbro/lky005
Piper B Dryden-Peterson S Chopra V Reddick C Oyanga A Are refugee children learning? Early grade literacy in a refugee camp in Kenya Journal on Education in Emergencies 2020 5 2 71 107 10.33682/f1wr-yk6y
Ramsay G Baker S Higher education and students from refugee backgrounds: A Meta-scoping study Refugee Survey Quarterly 2019 38 1 55 82 10.1093/rsq/hdy018
Riddell A Niño-Zarazúa M The effectiveness of foreign aid to education International Journal of Educational Development 2016 48 23 36 10.1016/j.ijedudev.2015.11.013
Rousseau, C., Beauregard, C., Daignault, K., Petrakos, H., Thombs, B. D., Steele, R., Vasiliadis, H.-M., & Hechtman, L. (2014). A cluster randomized-controlled trial of a classroom-based drama workshop program to improve mental health outcomes among immigrant and refugee youth in special classes. PLoS ONE, 9(8), Art. no. e104704. 10.1371/journal.pone.0104704
Rousseau C Benoit M Gauthier M-F Lacroix L Alain N Viger Rojas M Moran A Bourassa D Classroom drama therapy program for immigrant and refugee adolescents: A pilot study Clinical Child Psychology and Psychiatry 2007 12 3 451 465 10.1177/1359104507078477 17953131
Saavedra, J., & Bousquet, F. (2020). Committing to learning for every child, despite conflict and crisis. World Bank Blogs, 26 October. [blog post]. Retrieved 28 April 2023 from https://blogs.worldbank.org/dev4peace/committing-learning-every-child-despite-conflict-and-crisis
Siarova, H., & van der Graaf, L. (2022). Multi-stakeholder approach for better integration of refugee students: Stakeholder engagement in the practice-research-policy transfer in refugee education policy. OECD Education Working Papers, no. 265. Paris: OECD Publishing. 10.1787/82b390fb-en
Simenec TS Reid BM Refugee children and interventions for depression: A review of current interventions and implications of the ecological context Trauma, Violence, & Abuse 2020 23 3 877 890 10.1177/1524838020979844
Sirin S Plass J Homer B Vatanartiran S Tsai T Digital game-based education for Syrian refugee children: Project Hope Vulnerable Children and Youth Studies 2018 13 1 7 18 10.1080/17450128.2017.1412551
Streitwieser B Loo B Ohorodnik M Jeong J Access for refugees into higher education: A review of interventions in North America and Europe Journal of Studies in International Education 2019 23 4 473 496 10.1177/1028315318813201
Sullivan AL Simonson GR A systematic review of school-based social-emotional interventions for refugee and war-traumatized youth Review of Educational Research 2016 86 2 503 530 10.3102/0034654315609419
Tyrer RA Fazel M School and community-based interventions for refugee and asylum seeking children: A systematic review PLoS ONE 2014 9 2 e89359 10.1371/journal.pone.0089359 24586715
UN (United Nations) (2015). Transforming our world: The 2030 agenda for sustainable development. New York: UN. Retrieved 1 May 2023 from https://sdgs.un.org/sites/default/files/publications/21252030%20Agenda%20for%20Sustainable%20Development%20web.pdf
UNHCR (United Nations Refugee Agency) (2009a). Refugee education in urban settings: Case studies from Nairobi, Kampala, Amman, Damascus. Geneva: Operational Solutions and Transition Section (OSTS), Division for Programme Support and Management (DPSM), UNHCR. Retrieved 28 April 2023 from https://www.unhcr.org/us/media/refugee-education-urban-settings-case-studies-nairobi-kampala-amman-damascus
UNHCR (2009b). Designing appropriate interventions in urban settings: Health, education, livelihoods, and registration for urban refugees and returnees. Geneva: UNHCR, Division of Programme Support and Management (DPSM). Retrieved 28 April 2023 from https://www.unhcr.org/media/designing-appropriate-interventions-urban-settings-health-education-livelihoods-and
UNHCR (2020). Global trends in forced displacement – 2020. Copenhagen: Statistics and Demographics Section, UNHCR Global Data Service. Retrieved 28 April 2023 from https://www.unhcr.org/statistics/unhcrstats/60b638e37/global-trends-forced-displacement-2020.html
UNHCR (2021). Staying the course: The challenges facing refugee education. UNHCR Education Report 2021. Geneva: UNHCR. Retrieved 28 April 2023 from https://www.unhcr.org/publications/education/612f85d64/unhcr-education-report-2021-staying-course-challenges-facing-refugee-education.html
UNHCR, UNICEF, & IOM (2019). Access to education for refugee and migrant children in Europe [Factsheet]. Geneva: UNHCR, United Nations Children’s Fund (UNICEF) & International Organization for Migration (IOM). Retrieved 28 April 2023 from https://www.unhcr.org/neu/wp-content/uploads/sites/15/2019/09/Access-to-education-europe-19.pdf
World Bank (2022). The state of global learning poverty: 2022 update. Washington, DC: World Bank, the United Nations Educational, Scientific and Cultural Organization (UNESCO), the United Nations Children’s Fund (UNICEF), the Foreign, Commonwealth and Development Office (FCDO), the United States Agency for International Development (USAID), and the Bill & Melinda Gates Foundation (BMGF). Retrieved 5 May 2023 from https://thedocs.worldbank.org/en/doc/e52f55322528903b27f1b7e61238e416-0200022022/original/Learning-poverty-report-2022-06-21-final-V7-0-conferenceEdition.pdf
|
PMC010xxxxxx/PMC10225770.txt |
==== Front
Ann Hematol
Ann Hematol
Annals of Hematology
0939-5555
1432-0584
Springer Berlin Heidelberg Berlin/Heidelberg
37246975
5131
10.1007/s00277-023-05131-7
Original Article
Infectious complications following CAR-t cell therapy for B cell non-Hodgkin lymphoma: a single-center experience and review of the literature
http://orcid.org/0000-0003-4741-7885
Mercadal Santiago smercadal@iconcologia.net
12
http://orcid.org/0000-0001-5486-5710
Gomez Carlos A. 3
Lee Catherine J 1
Couriel Daniel R 12
1 grid.223827.e 0000 0001 2193 0096 Transplant and Cellular Therapy Program, Huntsman Cancer Institute, University of Utah, Salt Lake City, Utah USA
2 grid.223827.e 0000 0001 2193 0096 Bone Marrow Transplant/Cellular Therapy and Regenerative Medicine, University of Utah, 2000 Cir of Hope Dr Ste 1950, Salt Lake City, Utah 84112 USA
3 grid.266813.8 0000 0001 0666 4105 Department of Internal Medicine, Division of Infectious Diseases, University of Nebraska Medical Center, Omaha, Nebraska USA
29 5 2023
2023
102 7 18371843
21 11 2022
6 2 2023
© This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Chimeric antigen receptor T-cell (CAR-T) therapy targeting CD19 has significantly improved outcomes in the treatment of refractory or relapsed (R/R) B-cell non-Hodgkin lymphoma (NHL). Several risk factors including CAR-T cell-related toxicities and their treatments often lead to infectious complications (ICs); however, the pattern and timeline is not well established. We evaluated ICs in 48 patients with R/R B-cell NHL following CAR-T cell therapy at our institution. Overall, 15 patients experienced 22 infection events. Eight infections (4 bacterial, 3 viral and 1 fungal) occurred within the first 30 days and 14 infections (7 bacterial, 6 viral, 1 fungal) between days 31 to 180 following CAR-T infusion. Most infections were mild-to-moderate and fifteen infections involved the respiratory tract. Two patients developed mild-to-moderate COVID-19 infection and one patient a cytomegalovirus reactivation after CAR-T infusion. Two patients developed IFIs: one case each of fatal disseminated candidiasis and invasive pulmonary aspergillosis at day 16 and 77, respectively. Patients with more than 4 prior antitumor regimens and patient’s ≥ 65 years had a higher infection rate. Infections in patients with relapsed/refractory B-cell NHL are common after CAR-T despite the use of infection prophylaxis. Age ≥ 65 years and having > 4 prior antitumor treatments were identified as risk factors for infection. Fungal infections carried significant impact in morbidity and mortality, suggesting a role for increase fungal surveillance and/or anti-mold prophylaxis following high-dose steroids and tocilizumab. Four of ten patients developed an antibody response following two doses of SARS-CoV-2 mRNA vaccine.
Keywords
Infectious complications
CAR-T
B-cell NHL
issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIntroduction
Chimeric antigen receptor T-cell (CAR-T) therapy is an effective treatment for relapsed and refractory (R/R) b-cell non-Hodgkin lymphoma (NHL) [1-4]. Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are well-known complications of CAR-T therapy; however, patients are at risk for significant infections as well. Patients receiving CAR-T cell therapy are typically immunosuppressed due to prior treatments leading to prolonged cytopenias and hypogammaglobinemia followed by exposure to additional lymphodepleting chemotherapy. CRS and ICANS exhibit non-specific clinical manifestations of a hyperinflammatory state, and their differential diagnosis includes, among others, sepsis. The use of high-dose corticosteroids and IL-6 blockers to treat severe CRS/ICANS not only intensifies the risk of infection but can also mask infection-related symptoms such as fever, making the diagnosis of infection more challenging [5]. Severe infections (i.e., grade ≥ 3) have been reported in each of the CAR-T registration studies, with the incidence of infectious complications (ICs) ranging from 8 to 25% [1, 4]. Bacterial complications (e.g., bacteremia, C. difficile gastroenteritis, pneumonia), viral (e.g., respiratory viruses, cytomegalovirus, varicella zoster virus), and fungal infections (e.g., candidemia and invasive molds) have been described, along with a higher burden of ICs occurring within 30 days following CAR-T cell infusion [5]. While data from clinical trials have contributed to the understanding of infection risk, they often lack the granularity provided by single-center studies where the local epidemiology, patients’ characteristics, and treatment practices affect the pattern of ICs related to CAR-T cell therapy.
Herein, we report a single-center retrospective study of IC following CAR-T cell therapy in 48 adult patients with R/R B-cell NHL.
Methods
Patients and data collection
Adult patients (≥ 18 years) with R/R CD19 B-cell malignancies treated with CAR-T cell therapy at the Huntsman Cancer Institute between May 2018 and March 2022 were included. Patients’ electronic medical records were reviewed for patient and disease characteristics, including demographics, laboratory data, ICs early (0–30 days) and late (31–180 days), antimicrobial prophylaxis and treatment courses, prior therapies received, bridging therapy, disease status before CAR-T infusion, CAR-T indications, CAR-T infusion, CAR-T-related toxicities, ICU admission, and death events. Also, pre-infusion (age, sex, type of disease, ECOG, median prior treatments, lymphodepletion, type of CAR-T, Ferritin, lactic acid dehydrogenase, C-reactive protein, and IgG levels) and post-infusion factors (CRS, ICANS, use of steroids, and use of tocilizumab) that could be associated with infections were collected. COVID-19 infections, along with COVID-19 vaccination status and subsequent testing for anti-spike IgG, were captured following the availability of mRNA vaccines in December 2020. Antigen-specific immunoglobulin G (IgG) antibody titers were assessed using fresh plasma samples and direct enzyme-linked immunosorbent assay after 4 weeks receiving at least two COVID-19 vaccines. The Huntsman Cancer Center Institutional Review Board approved this retrospective analysis. Informed consent was obtained from all patients for being included in the study. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964, as revised in 2008.
Lymphodepletion chemotherapy and adoptive transfer of CD19-directed CAR-T cells
All patients received lymphodepleting chemotherapy consisting of cyclophosphamide 300–500 mg/m2 and fludarabine 30 mg/m2 on days − 5, − 4, and − 3, followed by single infusion of axicabtagene (n = 20), tisagenlecleucel (n = 13), lisocabtagene (n = 9), and brexucabtagene (n = 6).
Infectious prophylaxis
All patients received standard infectious prophylaxis consisting of levofloxacin and fluconazole from the day of cell infusion until absolute neutrophil count (ANC) recovery above 500 cells/mm3. Acyclovir and anti-Pneumocystis jirovecii (PJP) prophylaxis were given for a minimum of 6 months or until CD4+ lymphocyte count > 200 cells/mm3.
Definition of infection
Bacterial infections were categorized as a bacteremia or a site-specific infection. Repeated positive cultures for the same organism were considered a second event if they occurred ≥ 21 days after the initial event and interval cultures were negative. Site-specific infections were defined by a positive culture of a normally sterile site or by culture and clinical or radiographic evidence of infection in a nonsterile site. Viral infections were defined as upper or lower respiratory tract infections based on the presence of compatible symptoms and a positive viral molecular test, including detection for SARS-CoV-2. Cytomegalovirus (CMV) viremia was defined as any detection of CMV DNA in peripheral blood at any level using our institutional real-time CMV PCR assay independent of the present of compatible CMV viremia symptoms. Invasive fungal infections (IFIs) were classified as proven or probable based on the EORTC/MSG 2019 revised criteria for invasive fungal disease [6]. Infection severity was graded according to CTCAE v5.0 specific to the event of interest. Severe infections (grade ≥ 3) were those that required IV antimicrobial therapy and/or hospitalization or were associated with life-threatening symptoms or invasive interventions, in keeping with previously reported categorizations.
Statistical analysis
Descriptive statistics were used to summarize the clinical data. Categorical data were compared using the chi-square test, whereas, for continuous variables, non-parametric tests were used. Differences among the subgroups of patients were compared by using the chi-square test (two-tailed), Student’s t-test, or non-parametric tests when necessary. Prognostic factors significant (p-value less than 0.05) in the univariate analysis were included in multivariate analyses. Multivariate logistic regression analysis was used to test the association between infection occurrence and putative risk factors.
Results
General characteristics of the series (N = 48)
The cohort included patients with large B-cell lymphoma (LBCL) (n = 38), mantle cell lymphoma (n = 5), transformed LBCL (n = 4), and follicular lymphoma (n = 1). Demographic and clinical characteristics are shown in Table 1. Thirty-eight out of 48 patients (79%) developed any grade of CRS, being 21% of them grade 3 or higher. Twenty-one (44%) developed any grade of ICANs, being 38% grade 3 or higher. The main clinical and biological characteristics are shown in Table 1. Responses after CAR-T infusion at day 30 and relapses are shown in Table 2.Table 1 Demographic, clinical and biological characteristics of CAR-T cell patients
Characteristics CAR-T cell patients (N = 48)
Median age (range) 62.5 (26–85)
Sex (male/female) 40/8
Type of disease:
Large B-Cell Lymphoma (LBCL) 38
Mantle cell lymphoma 5
Transformed LBCL 4
Follicular lymphoma 1
ECOG 0-1 (%) 95
Median prior treatments 3 (3–10)
Prior autologous transplant 8 (17)
Flu/Cy* lymphodepletion (LD) (%) 100
Type of CAR-T:
Axicabtagene 20
Tisagenlecleucel 13
Lisocabtagene 9
Brexucabtagene 6
LDH > 250 U/L prior LD (%) 35
CRP > 0.8 mg/dL prior LD (%) 58
Ferritin > 500 ng/mL prior LD (%) 38
IgG < 400 mg/dL prior LD (%) 48
High tumor burden (%) 21
CRS grade 3–4 (%) 17
ICANS grade 3–4 (%) 17
Use of corticosteroids (%) 48
Use of tocilizumab (%) 54
ECOG Eastern Cooperative Oncology Group, LD lymphodepletion, LDH lactic acid dehydrogenase, CRP C-reactive protein, CRS cytokine release syndrome, ICANS immune effector cell-associated neurotoxicity syndrome
*Flu/Cy fludarabine and cyclophosphamide
Table 2 Prior treatments, response and relapse characteristics of CAR-T cell patients
CAR-T cell patients (N = 48)
Prior treatments before CAR-T
R-CHOP/R-CHOP like (%) 94
Fludarabine/bendamustine-based regimen (%) 19
Cisplatin-based regimen (%) 50
Gemcitabine-based regimen (%) 45
Tyrosine kinase inhibitors (%) 31
Lenalidomide (%) 5
Polatuzumab (%) 7
Radiotherapy (%) 24
Response day 30 after CAR-T
Complete remission (CR) 27
Partial remission 6
Progression disease 15
Relapse (patients in CR) 4
Median time relapsing patients after CAR-T (months) 3
Infectious complications
Thirty-seven (77%) patients of all series developed neutropenic fever after CAR-T infusion. Overall, fifteen patients (31%) experienced 22 infection events. Eight infections (36%) (4 bacterial, 3 viral and 1 fungal) occurred within the first 30 days of CAR-T infusion, and 14 infections (64%) (7 bacterial, 6 viral, 1 fungal) between days 31 to 180 following CAR-T infusion (Table 3). Most infections were mild-to-moderate in severity (87%) and did not require intravenous antibiotic therapy or hospital admission. Fifteen infections (68 %) involved the respiratory tract (upper respiratory infection: 10, pneumonia: 5). In general, bacterial and viral infections were detected at a median of 19 and 33 days following CAR-T infusion, respectively. Two patients developed mild-to-moderate COVID-19 infection and one patient had CMV reactivation at days 73, 82, and 49 after CAR-T infusion, respectively. Two patients (3.9%) developed IFIs: one case each of fatal disseminated candidiasis from a fluconazole-resistant Candida glabrata and invasive pulmonary aspergillosis (Aspergillus fumigatus complex) that occurred at day 16 and 77, respectively. Both cases required before the use of high-dose corticosteroids and tocilizumab to treat CRS toxicity grades 2 and 3, respectively. All bacterial, viral, and fungal infections and their pathogens are described in Table 3. Also, we developed a univariate and multivariate analysis with some of the prognostic factors previously reported in the literature and risk of infection as is showed in Table 4. In our series, patients with more than 4 prior antitumor regimens and patient’s ≥ 65 years had a higher infection rate, being age ≥ 65 years the most important factor for infection in the multivariate analysis (see Table 4).Table 3 Microbiological description of infection events (n = 22) and pathogens
Infection type < 30 days
Post-CAR-T 30–180 days
Post-CAR-T
Bacterial 4 7
Skin soft tissue 2
Line associated 1
Pneumonia 1 3
Intra-abdominal 1
Otitis/sinusitis 2
UTI 1
Viral 3 6
Rhinovirus URI 2 2
Influenza/rhinovirus LRTI (mixed) 1
CMV viremia 1
COVID-19 pneumonia 3
Funga 1 1
Invasive candidiasis 1
Invasive pulmonary aspergillosis 1
Total infection episodes 8 14
LRTI lower respiratory tract infection, URI upper respiratory tract infection, UTI urinary tract infection
Table 4 Univariate and multivariate analysis for prognostic factors and risk of infection
Univariate (p) Multivariate (p)
Age (< 65 y vs. ≥ 65 y) 0.001 0.009
Sex (M vs. F) 0.210
Type of disease (DLBCL vs. others) 0.871
ECOG (0–1 vs. ≥ 2) 0.168
Median prior treatments (≤ 4 vs. > 4) 0.026 0.140
Prior autologous transplant (yes vs. no) 0.495
Type of CAR-T (axicabtagene vs. tisagenlecleucel vs. lisocabtagene vs. brexucabtagene) 0.381
LDH prior LD (normal vs. high) 0.654
CRP prior LD (normal vs. high) 0.173
Ferritin prior LD (normal vs. high) 0.913
IgG prior LD (normal vs. high) 0.203
High tumor burden (< 6 cm vs. ≥ 6 cm) 0.536
CRS grade 3–4 (yes vs. no) 0.676
ICANS grade 3–4 (yes vs. no) 0.676
Use of corticosteroids (yes vs. no) 0.613
Use of tocilizumab (yes vs. no) 0.584
M male, F female, DLBCL diffuse large b-cell lymphoma, ECOG Eastern Cooperative Oncology Group, LDH lactic acid dehydrogenase, LD lymphodepletion, CRP C-reactive protein, CRS cytokine release syndrome, ICANS immune effector cell-associated neurotoxicity syndrome
COVID-19 vaccination
We evaluated humoral response to two-dose SARS-CoV-2 mRNA vaccines (BNT162b2 or mRNA-1273) in ten patients who were vaccinated after 90 days of CAR-T cell infusion when COVID-19 vaccine was available in December 2020. In this series, four patients received also tixagevimab plus cilgavimab after CAR-T infusion. At a median of 48 days following the second dose, an anti-Spike IgG was detectable in four (40%) of them.
Conclusions
In this single-center study, we report a high proportion of infectious complications occurring predominantly within the first 30 days following CAR-T cell infusion, with a modest predominance of viral over bacterial infections. Infections of the respiratory tract were the most common; most ICs were mild-to-moderate in severity, not requiring hospitalization or IV antibiotic therapy. Moreover, our uni/multivariate analysis showed age ≥ 65 years and ≥ 4 prior antitumor treatments as the main risk factors for infection.
Our results are in line with previously reported studies. Hill et al. [7] analyzed infectious complications in 133 adult patients in a cohort study including ALL (n = 47), CLL (n = 24), and NHL (n = 62) receiving CAR-T in a phase 1/2 study. In this series, 24% of patients experienced any infection, including 5% fungal and 4% of fatal events. Risk factors identified in this series were ≥ 4 prior treatments, a diagnosis of ALL and receiving a higher dose of CAR-T cells (2 × 107cells/kg). Strati et al. [8] analyzed 31 patients with relapsed/refractory LBCL who underwent CAR-T cell treatment with axicabtagene and were included in the clinical trials ZUMA-1 (NCT02348216) and ZUMA-9 (NCT03153462). Among all 31 patients, 71 infectious events of any grade were reported (42% grade 3–4 infection). Most common infections were viral and 6% of patients developed a fungal infection. Logue et al. [9] conducted a retrospective, single-center study of 85 patients with relapsed/refractory LBCL treated with axicabtagene. In the first 30 days, 36.5% patients presented an infection event, with 12.9% of them requiring IV antibiotics or hospitalization. CRS, ICANS, use of tocilizumab, or steroids and bridging therapy were risk factors for ICs in this series. After day 30, 44.3% of patients had any infection requiring hospitalization or IV antibiotics. Infection was a contributor to death in 3 cases (3.5%) in this series. Cordeiro et al. [10] analyzed a cohort of 86 patients with relapsed/refractory ALL, NHL, and CLL treated with CAR-T infusion included on a phase 1/2 clinical trial. After day 30th, 61% of patients developed any infection, with 71% respiratory tract involvement. Bacterial was the most frequent (60%), 31% viral and 9% fungal etiology. Moreover, 20% of patients in this series required hospitalization.
Baird et al. [11] evaluated hematologic recovery, immune reconstitution, and also infectious complications in 41 patients with NHL treated with axicabtagene ciloleucel. In the first 28 days following infusion, 46.3% patients had an infection, with the majority being mild-to-moderate in severity (68.4%). The most common etiology was viral respiratory tract infections (21.1%). Receipt of corticosteroids was the only factor that predicted risk of infection in a multivariate analysis.
Wudhikarn et al. [12] analyzed 60 patients with LBCL treated with CD19 CAR-T cells and a total of 101 infectious events were observed, being 75% of them mild-to-moderate and bacteria the most common causative pathogens. Thirty-seven percent of them are within the first 30 days. In a multivariate analyses, the use of systemic corticosteroids for the management of CRS or ICANS was associated with an increased risk of infections.
Lastly, a study by Mikkilineni et al. [13] focused on infections occurring within the first 30 days of treatment in 162 CAR-T cell patients who received CAR-T antigen targets (CD19, CD22, D2, and BCMA). The proportion of infectious complications in the first 30 days was 32.7%; greater lines of chemotherapy and a recent infection within 100 days of CAR-T cell infusion were associated with higher risk of infection.
Despite high degree of immunosuppression and prolonged neutropenia, fungal infections have remained infrequent in patients undergoing CAR-T cell therapy. Depending on the study, the incidence of IFI has ranged between 1 and 15%, and most of these infections occur as breakthrough to antifungal prophylaxis [14]. In our series, two patients (3.9%) developed IFIs: one fatal case of disseminated Candidemia due to fluconazole-resistant Candida glabrata (breakthrough fluconazole prophylaxis) and one case of invasive pulmonary aspergillosis. Of note, both patients had received immunosuppression augmentation with tocilizumab and high-dose steroids for high-grade CRS. Several reports have described molds other than Aspergillus, complicating CAR-T cell therapy [15]. Targeted anti-mold prophylaxis has been proposed for individuals with a history of past-fungal infection, severe and prolonged neutropenia (> 3 weeks), previous allogeneic HSCT, and in patients receiving high-dose corticosteroids5. In low-risk patients and/or at treatments centers with low IFI incidence, a preemptive strategy based on biomarkers and imaging screening could be adopted. As the option for CAR-T cell therapy move upstream in the treatment line for several hematological malignancies, the landscape for IFI complications will continue evolving in the upcoming years. Future studies are needed to elucidate specific risk factors for IFI and define the population who benefit the most of anti-mold prophylaxis.
In our case series, the spectrum of clinical syndromes associated with bacterial and viral infections was similar to those previously reported in the literature, with predominance of viral respiratory infections (including COVID-19) and nosocomial infections (see Table 2).
Our study has several limitations. First, it is restricted to a single center and included only patients with relapse/refractory CD19 B-cell NHL. As such, the overall results might not be applicable to centers with different antimicrobial prophylaxis practices or centers that offer CAR-T cell for other than CD19 B-cell NHL. Second, our median time to follow-up was almost 8 months post-CAR-T cell infusion. Hence, infection complications that occurred late in the post-CAR-T cell period might have not been entirely captured in our review. Notably, our study did not link laboratory markers for B-cell aplasia, B-cell dysfunction, or reconstitution of the T-cell compartment with tangible clinical outcomes for infection. Other factors such as CAR-T-cell-related neutropenia, reactivation of latent viral infections (e.g., CMV, HHV-6), and the need for stem-cell reinfusion in refractive cytopenias might play a role in late-onset infection following CAR-T infusion and should be assessed in prospective studies. Lastly, in our series, serological testing to assess immune response to SARS-CoV-2 vaccination was performed inconsistently and at different time points following CAR-T cell infusion. For instance, only 21% of patients had anti-spike IgG titers available to assess immunogenicity following two doses of SARS-CoV-2 mRNA vaccine. Several studies have described suboptimal vaccination responses in patients with hematological malignancies, including CAR-T cell recipients [16-19]. Other markers of vaccine-immune response (e.g., T-cell response) and the effect of vaccine boosters over the augmentation of anti-spike IgG titers were not addressed in this study and may play a substantial role in COVID-19 disease prevention following CAR-T cell therapy. Moreover, the ”real-world” impact of prophylactic strategies using anti-SARS-CoV-2 long-lasting monoclonal antibodies (e.g., tixagevimab plus cilgavimab) remains unknown and deserves further research in this population.
In summary, the reporting of infectious complications from CAR-T cell therapy’s clinical trials has been inconsistent, often lacking details about the nature, timing and course of common infections. In that context, single-center case series like ours contribute to understanding the local epidemiology and to guide antimicrobial prophylaxis strategies. Data from patients’ registries are welcome to delineate national and center-specific CAR-T cell-associated infection rates and define future research targets. In the upcoming years, several factors will continue reshaping the risk for infection following CAR-T cell therapy. The use of CAR-T cells with novel antigen targets (e.g., CAR-T targets NK, anti-BCMA, CD22, disialonganglioside [GD2]) along with the expansion of CAR-T cell indications and its use earlier in the treatment course will inevitably alter the infection risk framework. Research focused on infectious complications CAR-T cells will hopefully provide guidance on adequate standards and extension of antimicrobial prophylaxis in CAR-T cell patients.
Acknowledgements
The authors thank the staff and faculty of the Marrow Transplantation Program at Huntsman Cancer Institute and University of Utah for tireless work caring for the patients involved in this study.
Author contributions
S. M. and C. A. G. designed and performed research, analyzed data and wrote the manuscript. D. C. and C. L. reviewed that data and edited the manuscript; all authors approved the final manuscript prior to submission.
Declarations
Ethical approval
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.
Informed consent
Informed consent was obtained from all individual participants included in the study.
Conflict of interest
C. J. L. has received honoraria for educational activities and/or consultancy and/or participation in advisory boards from Jazz, Incyte, Fresensius Kabi, BMS, Kite, Kadmon, CareDx; has received research funding from Incyte; and serves on a trial steering committee for Incyte. S. M., C. A. G., and D. R. C. declare no competing interests.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Schuster SJ Bishop MR Tam CS Waller EK Borchmann P McGuirket JP Tam CS Waller EK Borchmann P McGuirk JP Jäger U Jaglowski S Andreadis C Westin JR Fleury I Bachanova V Foley SR Ho PJ Mielke S Magenau JM Holte H Pantano S Pacaud LB Awasthi R Chu J Anak O Salles G Maziarz RT JULIET Investigators Tisagenlecleucel in adult relapsed or refractory diffuse large B-cell lymphoma N Engl J Med 2019 380 1 45 56 10.1056/NEJMoa1804980 30501490
2. Neelapu SS Locke FL Bartlett LNL Lekakis J Miklos DB Jacobson CA Braunschweig I Oluwole OO Siddiqi T Lin Y Timmerman JM Stiff PJ Friedberg JW Flinn IW Goy A Hill BT Smith MR Deol A Farooq U McSweeney P Munoz J Avivi I Castro JE Westin JR Chavez JC Ghobadi A Komanduri KV Levy R Jacobsen ED Witzig TE Reagan P Bot A Rossi J Navale L Jiang Y Aycock J Elias M Chang D Wiezorek J Go WY Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell lymphoma N Engl J Med 2017 377 26 2531 2554 10.1056/NEJMoa1707447 29226797
3. Abramson JS Palomba ML Gordon LI Lunning MA Wang M Arnason J Mehta A Purev E Maloney DG Andreadis C Sehgal A Solomon SR Ghosh N Albertson TM Garcia J Kostic A Mallaney M Ogasawara K Newhall K Kim Y Li D Siddiqi T Lisocabtagene maraleucel for patients with relapsed or refractory large B-cell lymphomas (TRANSCEND NHL 001): a multicentre seamless design study Lancet 2020 396 10254 839 852 10.1016/S0140-6736(20)31366-0 32888407
4. Wang M Munoz J Goy A Locke FL Jacobson CA Hill BT Timmerman JM Holmes H Jaglowski S Flinn IW McSweeney PA Miklos DB Pagel JM Kersten MJ Milpied N Fung H Topp MS Houot R Beitinjaneh A Peng W Zheng L Rossi JM Jain RK Rao AV Reagan PM KTE-X19 CAR T-cell therapy in relapsed or refractory mantle-cell lymphoma N Engl J Med 2020 382 14 1331 1342 10.1056/NEJMoa1914347 32242358
5. Hill JA Seo SK How I prevent infections in patients receiving CD19-targeted chimeric antigen receptor T cells for B-cell malignancies Blood 2020 136 8 925 935 10.1182/blood.2019004000 32582924
6. Donnelly JP Chen SC Kauffman CA Steinbach WJ Baddley JW Verweij PE Clancy CJ Wingard JR Lockhart SR Groll AH Sorrell TC Bassetti M Akan H Alexander BD Andes D Azoulay E Bialek R Bradsher RW Bretagne S Calandra T Caliendo AM Castagnola E Cruciani M Cuenca-Estrella M Decker CF Desai SR Fisher B Harrison T Heussel CP Jensen HE Kibbler CC Kontoyiannis DP Kullberg BJ Lagrou K Lamoth F Lehrnbecher T Loeffler J Lortholary O Maertens J Marchetti O Marr KA Masur H Meis JF Morrisey CO Nucci M Ostrosky-Zeichner L Pagano L Patterson TF Perfect JR Racil Z Roilides E Ruhnke M Prokop CS Shoham S Slavin MA Stevens DA Thompson GR Vazquez JA Viscoli C Walsh TJ Warris A Wheat LJ White PL Zaoutis TE Pappas PG Revision and update of the consensus definitions of invasive fungal disease from the European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education and Research Consortium Clin Infect Dis. 2020 71 6 1367 1376 10.1093/cid/ciz1008 31802125
7. Hill JA Li D Hay KA Green ML Cherian S Chen X Riddell SR Maloney DG Boeckh M Turtle CJ Infectious complications of CD19-targeted chimeric antigen receptor–modified T-cell immunotherapy Blood. 2018 131 1 121 130 10.1182/blood-2017-07-793760 29038338
8. Strati P Varma A Adkins S Nastoupil LJ Westin J Hagemeisteret FB Fowler NH Lee HJ Fayad LE Samaniego F Ahmed S Chen Y Horowitz S Arafat S Johncy S Kebriaei P Mulanovich VE Heredia EA Neelapu SS Hematopoietic recovery and immune reconstitution after axicabtagene ciloleucel in patients with large B-cell lymphoma Haematologica 2021 106 10 2667 2672 10.3324/haematol.2020.254045 32732355
9. Logue JM Zucchetti E Bachmeier CA Krivenko GS Larson V Ninh D Grillo G Cao B Kim J Chavez JC Baluch A Khimani F Lazaryan A Nishihori T Liu HD Pinilla-Ibarz J Shah BD Faramand R Coghill AE Davila ML Dholaria BR Jain MD Locke FL Immune reconstitution and associated infections following axicabtagene ciloleucel in relapsed or refractory large B-cell lymphoma Haematologica 2021 106 4 978 986 10.3324/haematol.2019.238634 32327504
10. Cordeiro A Bezerra ED Hirayama AV Hill JA Wu QV Voutsinas J Sorror ML Turtle CJ Maloney DG Bar M Late events after treatment with CD19-targeted chimeric antigen receptor modified T cells Biol Blood Marrow Transplant 2020 26 26 33 10.1016/j.bbmt.2019.08.003 31419568
11. Baird JH Epstein DJ Tamaresis JS Ehlinger Z Spiegel JY Craig J Claire GK Frank MJ Muffly L Shiraz P Meyer E Arai S Brown JW Johnston L Lowsky R Negrin RS Rezvani AR Weng WK Latchford T Sahaf B Mackall CL Miklos DB Sidana S Immune reconstitution and infectious complications following axicabtagene ciloleucel therapy for large B-cell lymphoma Blood Adv. 2021 5 1 143 155 10.1182/bloodadvances.2020002732 33570626
12. Wudhikarn K Palomba ML Pennisi M Garcia-Recio M Flynn JR Devlin SM Afuye A Silverberg ML Maloy MA Shah GL Scordo M Dahi PB Sauter CS Batlevi CL Santomasso BD Mead E Seo SK Perales MA Infection during the first year in patients treated with CD19 CAR T cells for diffuse large B cell lymphoma Blood Cancer J 2020 10 8 79 10.1038/s41408-020-00346-7 32759935
13. Mikkilineni L Yates B Steinberg SM Shahani SA Molina JC Palmore T Lee DW Kaplan RN Mackall CL Fry TJ Gea-Banacloche J Jerussi T Nussenblatt V Kochenderfer JN Shah NN Infectious complications of CAR T-cell therapy across novel antigen targets in the first 30 days Review Blood Adv 2021 5 23 5312 5322 10.1182/bloodadvances.2021004896 34619768
14. Stewart AG Henden AS Infectious complications of CAR T-cell therapy: a clinical update Review Ther Adv Infect Dis 2021 24 8 20 10.1177/20499361211036773
15. Haidar G Dorritie K Farah R Bogdanovich T Nguyen MH Palash SP Invasive mold infections after chimeric antigen receptor-modified T-cell therapy: a case series, review of the literature, and implications for prophylaxis Clin Infect Dis 2020 71 3 672 676 10.1093/cid/ciz1127 31756246
16. Monin L Laing AG Muñoz-Ruiz M Del Molino I Alaguthurai T Hayday D-VC TS GC Seow J Abdul-Jawad S Kamdar S Harvey-Jones E Graham R Cooper J Khan M Vidler J Kakkassery H Sinha S Davis R Dupont L Quijorna IF O’Brien-Gore C Ling Lee P Eum J Poole MC Joseph M Davies D Wu Y Swampillai A North BV Montes A Harries M Rigg A Spicer J Malim MH Fields P Patten P Di Rosa F Papa S Tree T Doores KJ Hayday AC Irshad S Safety and immunogenicity of one versus two doses of the COVID-19 vaccine BNT162b2 for patients with cancer: interim analysis of a prospective observational study Lancet Oncol 2021 22 6 765 778 10.1016/S1470-2045(21)00213-8 33930323
17. Ram R Freund T Halperin T Ben-Ami R Amit O Bar-On Y Beyar-Katz O Eilaty N Gold R Kay S Glait-Santar C Hagin D Immunogenicity of a third dose of the BNT162b2 mRNA COVID-19 vaccine in patients with impaired B cell reconstitution after cellular therapy-a single center prospective cohort study Transplant Cell Ther 2022 28 5 278.e1 278.e4 10.1016/j.jtct.2022.02.012 35182795
18. Hill JA Krantz EM Hay KA Dasgupta S Stevens-Ayers T Bender RA Bar M Maalouf J Cherian S Chen X Pepper G Riddell SR Maloney DG Boeckh MJ Turtle CJ Durable preservation of antiviral antibodies after CD19-directed chimeric antigen receptor T-cell immunotherapy Blood Adv 2019 3 22 3590 3601 10.1182/bloodadvances.2019000717 31743392
19. Bhoj VG Arhontoulis D Wertheim G Capobianchi J Callahan CA Ellebrecht CT Obstfeld AE Lacey SF Melenhorst JJ Nazimuddin F Hwang WT Maude SL Wasik MA Bagg A Schuster S Feldman MD Porter DL Grupp SA June CH Milone MC Persistence of long-lived plasma cells and humoral immunity in individuals responding to CD19-directed CAR T-cell therapy Blood 2016 128 3 360 370 10.1182/bloodadvances.2019000717 27166358
|
PMC010xxxxxx/PMC10226288.txt |
==== Front
Eval Program Plann
Eval Program Plann
Evaluation and Program Planning
0149-7189
1873-7870
Elsevier Ltd.
S0149-7189(23)00104-0
10.1016/j.evalprogplan.2023.102327
102327
Article
Performance of two educational approaches in increasing knowledge of high-school students about COVID-19 during the first wave of pandemic
Janbani Zahra a
Osmani Freshteh bc⁎1
a Master of surgical technology,Faculty member, Faculty of Paramedical Sciences, Birjand University of Medical Sciences, Birjand, Iran
b Infectious diseases Research center, Birjand University of Medical Sciences, Birjand, Iran
c Assistant Professor,Department of Epidimiology and Biostatistics, Faculty of Health, Birjand University of Medical Science, Birjand, Iran
⁎ Corresponding author at: Infectious Diseases Research Center, Birjand University of Medical Sciences, Birjand, Iran.
1 ORCID: https://orcid.org/0000-0002-6112-7131
29 5 2023
10 2023
29 5 2023
100 102327102327
6 9 2021
10 5 2023
26 5 2023
© 2023 Elsevier Ltd. All rights reserved.
2023
Elsevier Ltd
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objectives
The coronavirus disease (COVID-19) pandemic has greatly altered peoples’ daily lives. Teachers and students were found quite unprepared for the emergence of the first COVID-19 wave. So, improving the knowledge of students about COVID-19 is an important issue.
Methods
In this study, 240 high students attended. Two interventions with the same contents, but in different ways, were delivered. A structured questionnaire was utilized to collect data on demographic information, and information about the behavioral intention toward COVID-19 before and after the educational interventions as well as a control group that received no educational intervention.
Results
students in all arms had similar baseline knowledge of COVID-19. The results of the post-analysis showed the efficiency of educational techniques in increasing students' knowledge about COVID-19. So the audio-visual training method performed significantly better than the visual training method (p = 0.03). Both approaches achieved better scores than the control group (P < 0.001).
Conclusion
During the outbreak of COVID-19, multimedia-based learning is a more effective educational approach and can improve the learning outcomes related to COVID-19 and achieve learning goals without close contact than written materials.
Keywords
Adolescent health
Covid-19
Coronavirus
Intervention study
Audio-visual media
Knowledge
==== Body
pmc1 Introduction
The new 2019 Coronavirus epidemic is more extensive in comparison to previous human Coronaviruses, indicating the extremely high transmission power of this virus (Farnoosh, Alishiri, Zijoud, Dorostkar & Farahani, 2020). Knowing to create healthy behaviors is the first necessary element for changing the belief and behavior of people (Tian, Rong, Nian & He, 2020). So, one of the proper procedures for decreasing the spread and mortality rate of Covid-19 disease is increasing people's awareness about the ways of transmission and prevention of the disease (Hewitt et al., 2020). However, social networks, have considerably increased the propagation of information about this disease. But, this platform may publish and extend false or fake information. Valid and accurate information is essential to help individuals to cope with this pandemic (Farnoosh et al., 2020). On the other hand, schools have been central in the debate about COVID-19. On the one hand, many have argued that they should be kept open, given their importance to youngsters and the future of the country, and the effort many countries have made in establishing protocols to keep them safe (Alfano, Ercolano, & Cicatiello, 2021). Lockdown policies have had a positive impact on the pandemic, and they have been able to reduce the number of COVID-19 cases in the countries that implemented them.
Iran is one of the youngest countries (Rahaei, Mirzaei Alavijeh, Soltanei, Bakhshi, & Shadkam, 2012) and adolescence is a critical period for health promotion. (Mirzaii & Olfati, 2015). On the other hand, students are considered the vulnerable group (Jamshidian, Hasanpour, & Najafi, 2017). Also, teaching can cause learning in the learner, and learning is a process for acquiring awareness and skills that can lead to improving their ability in making the right decisions to change to preventive healthy behaviors (Stanhope & Lancaster, 2004). Also, due to the COVID-19 pandemic, online learning has been adopted in all stages of education. One of the available methods of teaching is using educational pamphlets (Ewles & Simnett, 2003). Among all the approaches used in health education, none of them was as penetrating as printed educational materials (Barnes, Neiger, Hanks, Lindman, & Trockel, 2000). Also, video tutorial is created indirectly and have advantages such as durability of information, simplicity of use, and cost-effectiveness (Asselin & Cullen, 2001). Now, social media has been able to impressively change the social system of different countries (Soleimani Pour, 2012). Also, using social media is growing rapidly in most fields of healthcare and education (Lee & Lee, 2010). Research has proven that advances in information technology positively affect the approach of high school students on learning in contemporary educational environments. The effectiveness of online education has shown several advantages such as easy access to experts and exposure to educational environments. There are also several disadvantages of online education, such as: computer compatibility, or technical issues (Alfano & Ercolano, 2020). This pandemic is leading to social distancing policies worldwide (Bayham & Fenichel, 2020). While the first policy is aimed at strengthening the capacity of the health system to deal with the effects of COVID-19, the latter aims to ease the burden on the health system by reducing the probability of contracting the virus for all citizens (Alfano, 2022). School closures are some of the highest-profile social distancing measures used to slow the spread of an infectious disease. Many countries in Asia and Europe have instituted a nationwide school closure. These closures prevent contact among children and reduce cases (Bayham & Fenichel, 2020). Social media provide information about the pandemic to adolescents and may influence prevention behaviors. Owing to the need for strategies for dealing with COVID-19, schools can be considered as a center to raise adolescence awareness. Hence, this study aimed to assess the performance of two educational interventions about knowledge of protective strategies of COVID-19 among high school students in 2020 and determine students’ behavior change on the pandemic as a randomized controlled trial study.
2 Methods
2.1 Participants
All procedures were approved by the lead author’s university institutional review board of Birjand University of medical science, Iran before the recruitment of the first participant. This interventional study was implemented in two months June and July 2020 at one of the high schools in Birjand. Two-stage random cluster sampling was applied to include all eligible students who need to have internet access. Totally, 240 students in three groups of 80 were randomized in a simple, non-stratified randomization scheme as a control group and two interventional groups.
2.2 Study setting
The audio-visual group was taught instructions on how to access the online designed video, which addressed all the characteristics of COVID-19, transmission ways, preventive measures, and guidelines during home quarantine. The same educational content was prepared as written but delivered in a pamphlet. The control group received no educational intervention. Participants were instructed to view the material at least once during the study and were allowed to view the material as many as desired.
2.3 Assessment of knowledge
At the beginning of the survey, baseline knowledge regarding covid-19 was evaluated with a standard questionnaire, consisting of several questions. It is worth mentioning that demographic and socioeconomic information including (age, family income level, grade, parent's education, and the most used source of information) were collected once during the study.
Different parts of the questionnaire included knowledge assessed by questions containing characteristics of COVID-19, signs and symptoms, ways of virus transmission, and conditions that require home quarantine. So, each correct answer was given one point. The same questionnaire was distributed at the end of the intervention after the educational content was removed from the reach of the students.
The scoring of questions was 1 for the correct answer and 0 for an incorrect or "I don’t know" answer. The total score of students’ knowledge was expressed as the percentage of correct answers over a range of 0–100 %.
Some items of the questionnaire such as satisfaction measurement items ranked by a 5-point Likert scale, ranging from 1 = completely disagree, to 5 = agree; then the total score of these questions was expressed as Mean±SD.
All survey questions were optional and took participants approximately 7 min to complete.
Primary endpoints for the study included participants’ improvement in knowledge. Improvement in knowledge was defined as the improvement in knowledge between the baseline and the end of the study. The secondary endpoint was satisfaction with the educational materials.
2.4 Data analysis
Paired t-test and analysis of variance (ANOVA) were used to analyze continuous variables. The chi-square test was used for examining the homogeneity of demographic categorical variables in groups. For all statistical tests, p < 0.05 was considered statistically significant.
3 Results
3.1 Demographic characteristics of participants
Totally, from the four high schools, 240 students were included in the analysis. Overall, all students were female, with an average age of 16.4 ± 1.4 years (SD) ranging from 13 to 18 years. The majority of students (27.5 %), were in the 11th grade of education level. Among students, 37 % stated that they have a history of disease among their family members. Also, according to the answers given, the most used source of information receiving about covid-19 was social networks (35 %), after that, radio and TV (30 %), friends and acquaintances (21.1 %), and websites (14.9 %) were other using sources respectively. The mother's level of education of the majority of respondents (49 %) was secondary education. A preliminary examination of the groups showed that there were no significant differences in baseline characteristics including baseline knowledge and demographic variables between the groups (p.value>0.05), therefore, the study groups were homogenous.( Table 1).Table 1 Comparison of demographic factors between the study groups.
Table 1Variable Written group audio-visual group control group p value
History of disease in their family No 32(40) 44(55) 45(56.25) 0.81
Yes 48(60) 36(45) 35(43.75)
Source of information about covid-19 Radio and TV 24(30) 28(35) 21(26.25)
Website 12(15) 10(12.5) 14(17.5)
Social networks 26(32.5) 23(28.75) 29(36.25) 0.21
Friends and acquaintances 18(22.5) 19(23.75) 16(20)
University education 39(48.7) 32(40) 37(46.2)
Mother education Secondary education 36(45) 41(51.2) 39(48.7) 0.36
Read and write 5(6) 7(8) 4(5)
3.2 Assessment of performance of educational interventions
The results of the ANOVA analysis showed that there were no statistically significant differences between the three studied groups in all dimensions of knowledge evaluated before intervention. In contrast, after the intervention implementation, significant differences were observed between the control group against the two intervention groups in all of the questionnaire dimensions. Among written and audio-visual interventions, the obtained results showed more effectiveness in the audio-visual group based on increasing knowledge, but this improvement wasn't statistically significant according to Tukey’s test result (p.value=0.08). The mean knowledge scores were 84.31 ± 11.31, 76.41 ± 8.43, and 48.1 ± 15.41 in written, audio-visual interventions, and control groups respectively with a significant difference (p.value<0.05) ( Table 2).Table 2 Comparison of COVID-19 Knowledge before and after the interventions in the studied group (n = 240).
Table 2Area of Knowledge Before intervention After intervention
Pamphlet
(n = 80) Video
(n = 80) Control
(n = 80) Pamphlet
(n = 80) Video
(n = 80) Control
(n = 80)
Definition Correct 22.2 % 20.18 % 26.18 % 72.2 % 78.28 % 29.1 %
I don't know 32.3 % 40.3 % 37.9 % 12.3 % 13.3 % 31.9 %
Incorrect 43.5 % 30.6 % 35.9 % 15.5 % 8.46 % 30 %
Significance test/p value Χ2 = 12.18, p.value= 0.52 Χ2 = 43.24, p.value= 0.003
Signs and symptoms Mean±SD 3.2 ± 1.6 4.1 ± 2.9 3.8 ± 0.1 12.4 ± 3.6 14.1 ± 5.9 4.8 ± 3.1
Significance test/p value F= 6.21, p.value= 0.27 F= 68.19, p.value< 0001
Infection control
measures for COVID19
patients Mean±SD 4.4 ± 2.6 5.2 ± 1.3 4.9 ± 2.3 10.5 ± 3.6 9.2 ± 2.9 4.6 ± 7.3
Significance test/p value F= 5.14, p.value= 0.17 F= 46.15, p.value= 0.036
Mode of transmission Mean±SD 8.7 ± 1.4 7.2 ± 2.9 8.3 ± 1.2 17.3 ± 4.2 15.2 ± 3.1 8.3 ± 1.2
Significance test/p value F= 6.21, p.value= 0.27 F= 73.31, p.value< 0001
Nutrition during Corona Mean±SD 10.2 ± 7.7 9.2 ± 4.6 10.6 ± 4.2 21.2 ± 8.3 18.4 ± 2.9 11.6 ± 2.3
Significance test/p value F= 14.21, p.value= 0.81 F= 56.21, p.value= 0.016
How to use protective equipment Mean±SD 16.2 ± 3.6 16.3 ± 5.2 15.7 ± 3.6 29.5 ± 4.5 27.3 ± 3.6 13.7 ± 9.4
Significance test/p value F= 7.91, p.value= 0.67 F= 65.21, p.value= 0.013
High risk groups Mean±SD 7.56 ± 2.3 8.5 ± 3.6 8.15 ± 5.1 19.8 ± 7.3 16.5 ± 6.2 7.18 ± 8.1
Significance test/p value F= 12. 1, p.value= 0.43 F= 46.32, p.value< 0.001
Preventive methods Mean±SD 11.2 ± 3.3 12.3 ± 1.7 11.4 ± 6.6 21.7 ± 8.3 18.3 ± 6.4 12.4 ± 3.9
Significance test/p value F= 9.24, p.value= 0.33 F= 32.12, p.value= 0.021
Total awareness Mean±SD 43.98 ± 15.18 44 ± 13.98 47 ± 15.53 76.41 ± 8.43 84.31 ± 11.31 48.1 ± 15.41
Significance test/p value F= 98.34, p.value= 0.53 F= 164.14, p.value< 0001
3.3 Students' satisfaction regarding appeal with used educational tools
The mean satisfaction score in the audio-visual group regarding appeal with their material was 8.6 ± 0.8, significantly higher than the written pamphlet group (7.1 ± 1.4) (p < 0.001). Also, at the end of the study, all educational content was delivered to all students, and the results of the survey showed that students rated the efficiency of the educational video higher than educational written content significantly (p < 0.001).
3.4 Relationship between students' COVID-19 knowledge and sociodemographic characteristics
There was a statistically significant difference between the mean score of students' knowledge based on their assessed sociodemographic characteristics in this study(P ≤ 0.05), except for the family history of coronavirus variable (P = 0.52).
4 Discussion
The Covid-19 pandemic significantly disrupted the lives of people. As well as improving the quality of life, training programs can reduce the incidence of the disease, mortality, and cost of treatment. Due to the lack of proper planning and inadequate educational programs to increase public awareness of disease prevention, the incidence rate of infectious diseases such as COVID-19 is high. In this regard and to reduce the mortality rate education should be tried to increase public awareness. So, to address these issues, we designed two educational tools aimed to increase high school students' knowledge of covid-19 and affect behavioral changes in students.
This study investigated the effectiveness of the performance of these designed educational tools in improving high school students' knowledge about Covid19. It is worth mentioning that a high proportion of all students in three arms initially had little knowledge about preventive measures for COVID-19.
Although early research has suggested that adolescents may be less susceptible to severe symptoms from COVID-19, there is concern that adolescents who do not perceive themselves as high-risk or do not believe that the virus is of serious consequence may be less likely to engage in social distancing and disease prevention behaviors (Neidhöfer & Neidhöfer, 2020; Shao, 2020).
The results of this study showed that both designed educational tools could improve students' knowledge. However, the results showed more effectiveness in the audio-visual group than in a pamphlet, showing the superiority of the audio-visual over the written one. In line with the results of the present study, another study showed that educational videos could increase knowledge significantly; so, well-designed videos can help to increase knowledge about HPV infection (Ali, Ping, Prajapati, Padmapriya, & Nazer, 2017). In another study, the effectiveness of the pamphlet educational tool in promoting students’ knowledge and attitudes about AIDS was confirmed (Bastami, Zareban, Beiranvand, & Vahedi, 2012). The results of another study expressed that female high school students’ awareness about AIDS had increased after training with educational videos (Shahid et al., 2020). Several studies showed that audio-visual materials were effective in training about HPV (Krawczyk et al., 2012; Huang et al., 2020; Guan et al., 2020).
Zhong et al. showed that educational programs are a great tool to enhance public awareness of COVID-19. Also, their results expressed that the right awareness leads to appropriate preventive practices (Zhong et al., 2020).
The strengths of educational audio-visual tools can be explained by the fact that they were appropriate to the age of adolescents with simple and understandable content. Hence, it can play an efficient role in conveying information to high-school students and can be used to increase their knowledge about preventive measures for covid-19.
Dissemination of information about the prevention of COVID-19 among people by using informative/educational materials such as pamphlets may be a valuable tool as an adjunct to healthcare service protocols and control programs. Pamphlets are generally valuable informative tools, useful for disease control programs; however, they should be used within the context of a continuing educational process. Pamphlets can be used to generate effective changes and strengthen control, health, and educational measures. Pamphlets are more affordable since they do not require access to computers or software.
Pamphlets are informative published subjects containing training information. Designing and developing educational pamphlets is very simple and low-cost and can be widely distributed among people.
The result of Krawczyk et al. showed that written and video interventions were effective in educating about HPV and increasing young adults’ vaccination intentions.
Furthermore, the result showed that most students received news and information about COVID-19 from social networks.
Also, a significant association between the total score of students' knowledge and their mother's education level was observed, so that, the students with university education mothers had higher knowledge than other students. It seems that these students had more self-efficacy (Mahmoodpoor, Valadkhani, Ozayi, & Asayesh, 2017).
Moreover, the obtained results showed that by increasing the grade, students’ knowledge scores increased.
According to the results reported by Karimi et al., with increasing age and grade, students’ awareness about AIDS increased.
One limitation of the study was despite our best efforts at randomly assigning students, nearly, 20 % of them in the video were older than age 14 compared with those in two other groups.
5 Conclusion
The findings of this study expressed that education by tools such as audio-visual and written materials could increase students’ knowledge about the symptoms and preventive measures of covid-19. Also, it can be concluded that online audio-visual educational tools may become an innovative and efficient way to deliver education requiring long-term health behaviors. Audio-visual educational tools can be widely distributed and easily available has the potential to improve behavioral outcomes over current traditional methods. Also, the school-age is one of the best ages to teach the symptoms and ways of transmission and prevention of this disease. It is better appropriate education interventions be used as a way to deal with COVID-19. Also, this study suggests adolescents are obtaining COVID-19 knowledge from different sources, including social media. Increases in screen time and reduced physical activity may impact long-term health among adolescents.
Ethics approval
All procedures in this study were approved by the ethics board committee of Birjand University of Medical Sciences, reference number: IR.BUMS.REC.1399.185.
Funding
None.
CRediT authorship contribution statement
ZJ and FO designed the experiments. FO collected data and performed the statistical analyses and wrote the results section. ZJ interpreted the results.FO wrote the initial manuscript. ZJ critically reviewed and modified the manuscript. Both authors approved the final manuscript.
Declaration of Competing Interest
The authors of this article declare they have no conflicts of interest.
Freshteh Osmani Born in 1989/06/25 in Birjand. Parents were Mohsen and Fatemeh. Studied Biostatistics Phd in tarbiat Modares University. Overcame studied statistics in Ferdowsi University and MSc of Biostatistics in tarbiat Modares University. Worked as Assisstant professor in Biostatistics. Personal traits were Quiet /Curious /creative. Always Researching and doing new ideas. Never improvidence. Known for Active with new idea.
Appendix 1 : Knowledge questionnaire
1. The corona is a disease transmitted from animals to humans and the routes of transmission, animal reservoirs, ways to prevent not specified.
2. This disease publish through respiratory droplets of coughs who have Covid-19 from people who have no symptoms, is very high.
3. The incubation period means the time of infection with the virus and the onset of symptoms.
4. The maximum estimate for this course is 1–10 days for Covid-19.
5. The most effective way to prevent Covid-19 is to disinfect your hands with soap and water or alcohol-based disinfectant solutions.
6. Hand washing time to prevent corona is 10 s
7. To prevent Covid-19, only the use of a mask is sufficient.
8. After applying the mask, make sure that there is no gap between the mask and the face.
9. Hands should be disinfected as soon as the front of the mask is touched.
10. To remove the mask, take the front of the mask and take it out and throw it in the closed bucket.
11. Regular rinsing of the nose with saline solutions will prevent you from getting the Corona virus.
12. Maximum consumption of one serving of fruit as a snack during the day or before meals is effective in preventing corona.
13. Antibiotics can be used to prevent and treat the corona virus.
14. To prevent from getting corona, seasonal fruits contain antioxidants like pomegranate, orange and grapefruit can be effective.
15. During the corona, it is recommended to eat vegetables such as carrots, squash and spinach, beet leaves and lettuce leaves, which contain vitamin A.
16. To prevent the corona, taking 2–3 large meals a day is recommended.
17. To remove the second glove from the hand, dip the fingers of the right hand into the outside of the glove of the left hand and pull it out.
18. When sneezing or coughing can only be in front of your mouth with a tissue and throw the tissue in a closed area.
19. The most common symptoms of Corona, are dry cough, nasal congestion.
20. COVID-19, is a heavy small virus and therefore cannot remain suspended in the air.
Acknowledgment
We would like to thank all the study subjects for their participation and also thank the dentistry clinical research development unit, Birjand University of Medical Science, Birjand, Iran for their consultation.
Code availability
Not applicable.
Consent to participate
Participation in the study was completely voluntary.
Consent for publication
Not applicable.
Human Subjects Approval Statement
All procedures in this study were approved by the ethics board committee of Birjand University of Medical Sciences, reference number: IR.BUMS.REC.1399.185. It was also registered in the Iranian Registry of Clinical Trials ( IRCT20150405021601N2).
==== Refs
References
Alfano V. The effects of school closures on COVID-19: A cross-country panel analysis Applied Health Economics and Health Policy 20 2 2022 223 233 34890025
Alfano V. Ercolano S. The efficacy of lockdown against COVID-19: A cross-country panel analysis Applied Health Economics and Health Policy 18 2020 509 517 32495067
Alfano V. Ercolano S. Cicatiello L. School openings and the COVID-19 outbreak in Italy. A provincial-level analysis using the synthetic control method. Health Policy 125 9 2021 1200 1207 34247845
Ali A. Ping N. Prajapati S. Padmapriya M. Nazer Z. Development and validation of ‘Educational Pamphlet’in prevention of Human Papilloma Virus (HPV) infection among age eligible adults for HPV vaccination in Kedah State, Malaysia MedCrave Online Journal of Bioequivalence & Bioavailability 4 1 2017 216 221
Asselin M.E. Cullen H.A. What you need to know about the new BLS guidelines-Get a jump on recent changes to basic life support protocols with this summary Nursing 2001 2001 48 50
Barnes M.D. Neiger B.L. Hanks W.A. Lindman N.T. Trockel M. Print material purchasing decisions in health education practice Journal of Health Education 31 3 2000 128 137
Bastami, F., E. Zareban, A. Beiranvand and S. Vahedi (2012). Effect of educational pamphlet on knowledge and attitude of non-medical students about aids in university of Zabol in 2008.
Bayham J. Fenichel E.P. Impact of school closures for COVID-19 on the US health-care workforce and net mortality: A modelling study The Lancet Public Health 5 5 2020 e271 e278 32251626
Ewles, L. and I. Simnett (2003). Changing policy and practice. Promoting Health: A Practical Guide, 5th ed., Elsevier, London: 314–332.
Farnoosh G. Alishiri G. Zijoud S.H. Dorostkar R. Farahani A.J. Understanding the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease (COVID-19) based on available evidence-a narrative review Journal of Military Medicine 22 1 2020 1 11
Guan W.-j Ni Z.-y Hu Y. Liang W.-h Ou C.-q He J.-x Liu L. Shan H. Lei C.-l Hui D.S. Clinical characteristics of coronavirus disease 2019 in China New England Journal of Medicine 382 18 2020 1708 1720 32109013
Hewitt J. Carter B. Vilches-Moraga A. Quinn T.J. Braude P. Verduri A. …Guaraldi G. The effect of frailty on survival in patients with COVID-19 (COPE): a multicentre, European, observational cohort study The Lancet Public Health 5 8 2020 e444 e451 32619408
Huang C. Wang Y. Li X. Ren L. Zhao J. Hu Y. Zhang L. Fan G. Xu J. Gu X. Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China The lancet 395 10223 2020 497 506
Jamshidian Z. Hasanpour M. Najafi M. The effect of sleep hygiene education on the knowledge level of parents of elementary school students Health Prom Manage 1 4 2017 61 68
Krawczyk A. Lau E. Perez S. Delisle V. Amsel R. Rosberger Z. How to inform: Comparing written and video education interventions to increase human papillomavirus knowledge and vaccination intentions in young adults Journal of American College Health 60 4 2012 316 322 22559091
Lee J. Lee H. The computer-mediated communication network: Exploring the linkage between the online community and social capital New Media & Society 12 5 2010 711 727
Mahmoodpoor A. Valadkhani M. Ozayi N. Asayesh M. The relation of educational level and maternal employment on self-efficacy in Tehran sixth grade elementary students, 2014 Community Health Journal 9 1 2017 1 9
Mirzaii K. Olfati F. Explaining the educational needs of adolescents about reproductive health from the perspective of teachers The Journal of Qazvin University of Medical Sciences 73 2 2015 68 76
Neidhöfer G. Neidhöfer C. The effectiveness of school closures and other pre-lockdown COVID-19 mitigation strategies in Argentina, Italy, and South Korea ZEW-Centre for European Economic Research Discussion Paper 2020 20 034
Rahaei Z. Mirzaei Alavijeh M. Soltanei T. Bakhshi Z. Shadkam V. The survey of knowledge and attitude of university students on reproductive health in the city of Yazd in 2010 Tolooebehdasht 10 3 2012 139 149
Shahid Z. Kalayanamitra R. McClafferty B. Kepko D. Ramgobin D. Patel R. Aggarwal C.S. Vunnam R. Sahu N. Bhatt D. COVID‐19 and older adults: What we know Journal of the American Geriatrics Society 68 5 2020 926 929 32255507
Shao, P. (2020). Impact of city and residential unit lockdowns on prevention and control of COVID-19. MedRxiv: 2020.2003. 2013.20035253.
Soleimani Pour R. Social networks, opportunities and threats Rahavard Noorwegen 14 2012 9 31
Stanhope, M. and J. Lancaster (2004). Community & public health nursing, Mosby St. Louis.
Tian Y. Rong L. Nian W. He Y. Gastrointestinal features in COVID‐19 and the possibility of faecal transmission Alimentary pharmacology & therapeutics 51 9 2020 843 851 32222988
Zhong B.-L. Luo W. Li H.-M. Zhang Q.-Q. Liu X.-G. Li W.-T. Li Y. Knowledge, attitudes, and practices towards COVID-19 among Chinese residents during the rapid rise period of the COVID-19 outbreak: a quick online cross-sectional survey International Journal of Biological Sciences 16 10 2020 1745 32226294
|
PMC010xxxxxx/PMC10227545.txt |
==== Front
sppfe
PFE
Policy Futures in Education
1478-2103
SAGE Publications Sage UK: London, England
10.1177_14782103231178644
10.1177/14782103231178644
Special Issue: Higher Education Policy and Management in the Post-Pandemic
Online teaching during the COVID-19 pandemic: Vietnamese language teachers’ emotions, regulation strategies and institutional policy and management
https://orcid.org/0000-0001-9979-1321
Phan Anh Ngoc Quynh
Faculty of Education and Social Work, The University of Auckland, Auckland 1023, New Zealand and Centre for the Study of Higher Education, University of Kent, Canterbury, United Kingdom
https://orcid.org/0000-0003-1341-2680
Pham Linh Thi Thuy †
Faculty of English Language Teacher Education, University of Languages and International Studies , Vietnam National University Hanoi, Hanoi, Vietnam
Anh Ngoc Quynh Phan, Faculty of Education and Social Work, The University of Auckland, 74 Epsom Avenue, Auckland 1023, New Zealand. Email: anh.phan@auckland.ac.nz
† deceased.
5 2023
30 5 2023
30 5 2023
21 4 405422
© The Author(s) 2023
2023
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Teaching is often described as one of the most emotion-laden professions. In times of the COVID-19 pandemic, the conversion to online teaching has triggered new emotional experiences of teachers that not many studies have taken into account. Studying emotion from a poststructuralist lens, this study examines the emotional experiences of 10 language teachers in a university in Vietnam and their responses to the new teaching platforms. Analysis of the in-depth semi-structured interviews shows that the pedagogically and technologically distinctive features of online teaching aroused unique challenges for and emotions of the teachers, both positive and negative. Also, the teachers reported a number of strategies to cope with the new situation which we term as in-the-moment and out-of-class emotion regulation. The study highlights the need for acknowledgment and support for teachers in terms of resources, policy and management of institutions in the “new normal situation,” while displaying teachers’ self-reliance and emotional self-regulation. The article calls for attention to teachers emotion as an integral dimension of the profession, regardless of the physical or virtual setting of the classroom.
Teachers emotions
online teaching
emotion regulation
institutional policy
COVID-19
Vietnam
VNU University of Languages and International Studies N.21.06 typesetterts10
==== Body
pmcIntroduction
Teaching is inherently emotional work (Schutz, 2014), which has been accentuated by the strain that COVID-19 has placed on educators and students since its advent in early 2020. Responses to the COVID-19 pandemic have created a vast array of unique challenges for teachers due to the lack of previous experiences, preparation, and expertise of both teachers and institutions (Adedoyin and Soykan, 2020). In this new unprecedented situation, studies show that a high percentage of teachers demonstrated anxiety, depression and stress symptoms (Santamaría et al., 2021), although increased efficacy in classroom management and increased sense of accomplishment among teachers have also been reported (Sokal et al., 2020). From previous research, suggested implications usually focus on technological literacy and utilization. However, as a study by Trust et al. (2016) postulates, teachers also have other social, affective, and identity needs to meet, including the needs to express emotions and receive emotional support.
The importance of teacher recognition of their emotions, especially in crisis contexts, has been emphasized to improve their own well-being, and in turn, enhance students’ well-being (O’Toole and Friesen, 2016). Teachers’ emotional responses are correlated with coping strategies, leading to substantial levels of positive (happiness, resilience, and growth) and negative reactions (stress, anxiety, anger, sadness, and loneliness) (MacIntyre et al., 2020). However, little is known about the emotion management of teachers when switching to the new online mode of teaching. The ways teachers cope with the stress and challenges, as well as the institutional support for teachers of while teaching online in the COVID-19 era, remain unexplored while their journeys to becoming accustomed to the new environment of emergency online teaching is well worth exploring to further enhance the quality of education.
This study aims to address the above-mentioned gaps in the current scholarship by investigating the emotional experiences of 10 Vietnamese language teachers in response to the COVID-19 situation, their strategies to regulate their emotions, and the support they received from their institution both emotionally and professionally. The findings of this study hopefully will empower teachers during periods of uncertainty and facilitate policy implementation to support their teaching and overall well-being.
Literature review
Teachers emotions and emotion regulation
Emotions are an inseparable part of the teaching profession, and teachers experience various emotions in different classroom situations. Emotions arguably play an important role in a teacher’s capacity to thrive, not just survive in their professional life (Mansfield et al., 2012). Despite little agreement upon what constitutes an emotion, it is a must to theorize emotions to be able to study about them (Benesch, 2017).
There is a strong need to explore emotions in renewed theoretical and methodological perspectives that move beyond the dichotomous thinking that separates realms of emotion into private and public, and genuine or fake emotional expressions. In teacher’s emotion studies, researchers (Benesch, 2017; Zembylas, 2007) have suggested the use of poststructuralist perspectives as the most promising to the study of emotion, although the poststructural lens has not been used much in education in teacher emotion. The poststructuralist approach criticizes entrenched binary oppositions (such as mind/body, nature/culture, rationality/emotion) because these binary oppositions support a hierarchy or economy of value that operates by privileging one term over another. Instead, poststructuralist perspectives take social, cultural, and political factors into account to study emotion without overlooking the interpersonal components of emotions (Abu-Lughod and Lutz, 1990; Rosaldo, 1984; Weedon, 1987). It acknowledges the constitutive effects of emotions as “discursive practices” (Abu-Lughod and Lutz, 1990), which means that the words used to describe emotions are themselves “actions or ideological practices” that serve specific purposes to create and negotiate reality (Zembylas, 2005: 937). Following this line of argument, Zembylas (2005) contends that “power, agency and resistance are at the center of exploring the role of emotion and identity in teaching” (936). In other words, emotion is interwoven with issues of power and resistance in teaching, meaning that power relations and the role of culture and ideology (Zembylas, 2003a, 2003b) should also be considered when studying teachers emotion.
Power relations are inherent in emotion and shape the expression of emotions by allowing some emotions too be expressed while others being prohibited. Poststructuralist studies on teacher emotions examine the role of culture, power, and ideology in creating emotion discourses. Simultaneously, teachers’ participation in this process of emotion discourse creation through adopting or resisting these discourses is also investigated (Zembylas, 2002a). The analysis of teachers’ emotions does not cease at mere descriptions but focuses on what can be done to help teachers with these complex everyday emotional transactions. To borrow Zembylas’ (2005) words, poststructuralist ideas of emotion can help educators better analyze the complexities of “emotional rules” (Zembylas, 2002b) and explore the role of emotional practices in teaching. For these reasons, this study follows the poststructuralist approach to teachers’ emotions, which is considered “valuable” as a new direction for emotion study in education (Zembylas, 2005: 22).
Emotion regulation in the workplace has officially entered the mainstream of research since the seminal work of Hochschild (1979, 1983). She defined emotion regulation as the effort to manage emotions by changing the quality or intensity of an emotion or feeling. The regulation of emotion often relies on ‘feeling rules,’ a set of guidelines on what individuals should feel in specific situations. These guidelines are shared among a community, although they are often latent (Hochschild, 1979, 1983). Emotional rules refer to any means that govern both feelings and communications of emotions in desired directions. They reflect power relations, and are both the means and results of human differences in emotion expression and communication (Zembylas, 2002b) which may take place through people assessing which emotions are appropriate or inappropriate, deviant or normal. In an educational context, emotional rules regulate teachers’ language and embodiment of emotions. For instance, Zembylas (2002b: 201) asserts that generally, teachers need to “control emotions of anger, anxiety, and vulnerability, and express empathy, calmness, and kindness.”
Prior studies confirm that emotional authenticity greatly matters to teachers’ health and well-being (Keller et al., 2014; Philipp and Schüpbach, 2010). However, it is shown that teachers practise emotion regulation because they believe it makes them more effective in managing, disciplining, and maintaining their relationships with students (Sutton et al., 2009). They also tend to communicate their positive emotions and use a variety of emotion regulation strategies to reduce negative ones.
Against the COVID-19 context: The shift to online teaching worldwide and in Vietnam
Due to the unrelenting COVID-19 pandemic, many countries have had to resort to nationwide lockdowns and temporary school and campus closures to avoid the rapid infection of the novel virus, leading to the shift to online education. Initial actions of universities were to transfer contents to online learning environments with synchronous classrooms, without catering to the pedagogical strategies of online teaching (Crawford et al., 2020). The lack of proper preparation for online teaching and learning is a major barrier for both teachers and students (Rahimi and Martin, 2020). Although teachers generally believe the new mode of teaching and learning bring new experiences and opportunities, adapted pedagogical content knowledge for online learning is definitely needed. For instance, making digital learning experiences engaging requires teachers to have distinctive knowledge and skills (Rapanta et al., 2020).
In Vietnam in particular, schools were constantly closed and reopened, depending on the times of the outbreaks. However, in the third wave of the COVID-19 outbreak since May 2021 until currently when this paper is being written, Vietnamese students of all levels in many cities and provinces have had to stay at home and switched to online learning instead. Since the beginning of the COVID-19 pandemic, the Ministry of Education and Training (MOET) has made a stipulation of “suspending schools without stopping learning.” MOET has also been working to develop guidelines on online teaching and learning for teachers and parents. The long-term plan is to improve teachers’ skills to use technology to achieve practical, effective, and quality education.
Although virtual classrooms have been a sound replacement for traditional physical classrooms for the past year, with the exceptions of Pham and Phan (2021, 2022) or Foreman-Brown et al. (2022), there is still scant research on teachers’ experiences in the migration process from offline to online teaching, their emotions emerging out of these experiences and the institutional support they have had. This study is among the first to investigate the emotional experiences of Vietnamese language teachers in response to the COVID-19 situation, their strategies to regulate their emotions, and the support they received from their institution both emotionally and professionally. This study seeks to answer the following research questions:1. What were Vietnamese language teachers’ emotional experiences when teaching online in the COVID-19?
2. What strategies did the teachers use to regulate their emotions when teaching online in the COVID-19?
3. What types of support did the teachers receive or expect to receive when converting to online teaching in times of the pandemic?
In what follows, we are going to elaborate on the methodology used to conduct this study before presenting our findings and discussing the findings in relation to the literature and policy implications. We then conclude the paper with a brief summary of the research, its contribution to the field, and suggestions for further study in the future.
Methodology
The paper draws on qualitative data from a project we carried out in May 2021 on teacher emotions and pedagogical practices in online teaching during COVID-19, and a follow-up study conducted in August and September 2021. The preliminary result of the project has been published (see more in Pham and Phan, 2021). This paper employed a phenomenological approach to illuminate Vietnamese language teachers’ lived experiences within the new online mode of teaching. This approach helps to gather deep information and perceptions from the perspective of the research participants, which emphasizes the importance of personal experience and interpretation (Lester, 1999). We, therefore, used this powerful tool to gain insights into the teachers’ emotional experiences. The interpretive dimension of phenomenological research also enables phenomenological research to be used as the basis for practical theory, and to inform, support or challenge policy and action (Lester, 1999).
Ten English teachers in a university in Vietnam were recruited by purposive sampling (Merriam, 2009), which is appropriate as our goal was to select participants “in a strategic way” so that those sampled are relevant to the research objectives of the study (Bryman, 2012: 418). The university in this study was a public, research-intensive higher education institution that offered programmes on language education, linguistics, international studies and related social sciences and humanities. We particularly chose English language teachers as the population of the study because being a foreign language teacher triggers unique emotional characters of the profession that come from various factors including self-doubts about one’s own language ability; emotional anxieties of learners; mixed level of proficiency among learners; threats to sense of self and identity; teaching workload; or working conditions (Gkonou and Miller, 2017; Gkonou et al., 2020). The participants were recruited by purposive sampling through both authors’ personal network. This sampling method allowed us to identify and select information-rich cases for the most effective use of limited resources in the context of COVID-19 (Patton, 2002). We attempted to diversify the participants in terms of years of teaching experience, ranging from six to over 25 years. We used semi-structured interviews which require a limited number of questions prepared in advance and follow-up questions on site (Rubin and Rubin, 2012). The interviews were conducted via online tools due to the COVID-19 social distancing protocol in Vietnam when this study was undertaken.
Before the interview started, the participants were reminded of the purpose of the study. We also obtained their verbal informed consent after informing them that the interview would be recorded, and their personal information would remain confidential. The participants were encouraged to talk about what they felt comfortable with and they could stop the interview whenever they wanted without any repercussions. The interview questions were centered around the first reactions of the participants towards the sudden conversion to online teaching, the ensuing challenges and their learning to master the new teaching platforms, their strategies to cope with or enhance their emotional responses (negative and positive respectively), and the institutional support they received. The interviews were conducted in Vietnamese, recorded, transcribed and translated by both authors. Pseudonyms were given to each participant to protect their anonymity and confidentiality. Each interview lasted from 30 min to 90 min. Table 1 sketches the demographic information of the teachers in this study.Table 1. Demographic information of the participants.
Name of participants (Pseudonyms) Gender Teaching experience
Thao Female 10 years
Duong Female 11 years
Ha Female Over 25 years
Hoa Female Over 20 years
Thu Female 11 years
Nhung Female 11 years
Lan Female 10 years
Mai Female 11 years
An Female 8 years
Minh Male 6 years
We undertook a reflexive thematic analysis of the transcript (Braun and Clarke, 2019), utilizing a collaborative coding approach to identify the shared patterns of meanings in the teachers’ emotional experiences and their emotional regulation strategies. Both authors independently coded all the interview transcripts before discussing our mutual understanding of the coded themes. We began the coding process by writing memos on the margin of the transcriptions which included 35 pages in total. The first round of data exploration focused on summarizing the meanings of each paragraph in each transcription. The second round placed extra emphasis on searching for new contents that were left out. Our coding process was driven by theoretical bearings of the study (the poststructuralist approach to teachers emotions with attention to power, resistance, agency, discourses), with a particular focus on the data that deductively aligned with the concepts of teachers emotions (positive and negative emotions). For example, we paid attention to the teachers' management of their negative emotions, or enhancement of their positive emotions, and their assessment of the classroom situations to either display their emotions or not. Meanwhile, our coding process also involved a inductive approach, which means that the themes were strongly linked to the data and emerged from it. For example, as we coded the data of teachers’ strategies, we noted their strategies were used within and outside their online sessions contexts. The coding resulted in three broad categories, which are: teachers’ emotions, teachers’ emotion regulation, and institutional policy and management. Examples of this process can be seen in Table 2. These categories were then used as the foundation of the analysis, which is presented in the following sections.Table 2. The coding table.
Broad categories Themes Definition of category Example
Teachers emotions Initial stage of online teaching Descriptions of teachers’ emotional experience in their journey of switching to online teaching I was quite confused at first when I needed to teach online. I didn’t really know what and how to initiate an online class because I hadn’t tried it before….I felt a lack of interaction between the teacher and the students….When I was required to use technological-assisted devices for online learning for the first time, I was not so confident because I haven’t tried it before. (Minh)
Subsequent stage of online teaching I started to enjoy online teaching. I know how to design an online lesson. The students were also more familiar with online learning….I felt happier and more comfortable. (An)
Teachers’ emotion regulation In-the-moment regulation strategies Different strategies the teachers used to regulate their emotions I turned my camera off sometimes, and tried to use my voice and other audio-visual aids to teach….I dressed more comfortably and had more flexible movements. (Lan)
Out-of-classroom strategies I shared with my husband and close colleagues who were also going through the same experiences like me when teaching online. They understood what I had to cope with. (Hoa)
Institutional policy and management Professional development support Teachers’ opinions and descriptions of the policy and support of the institution in the “new normal” situation of online teaching There were workshops to provide teachers with technological training on how to use online platforms and software for teaching. (Minh)
Emotional support There has never been a formal programme that specifically supports teachers’ well-being (Thao)
The institution has started to pay attention to teachers’ emotions. Although I was not quite interested in the mindfulness programme organized by the institution, I was glad they started to think about us. (Mai)
In this study, member checking, peer review, and researcher’s reflexivity (Merriam, 2002) are the main measures used to build trustworthiness of the data and the data analysis process. After transcribing the interviews and finishing the first round of analysis, we contacted the participants to check if our understanding of their narratives were appropriate or not, which served the purpose of member checking. In addition, we discussed together on “the congruency of emerging findings with the raw data, and tentative interpretations” so that we ourselves could examine the rationality and validity (Merriam, 2002: 31). As the second author was an English language teacher herself, her insider view helped us beware of the complexity in teachers’ emotion in response to the shift to online teaching. Her insider positionality also granted us easier access to the participants and hence, built up researchers-participants trust. However, we were conscious of our outsider positionality, reinforced by the first author. The cross-checking of understanding between the two authors was to assure that we would not rely on our assumptions about the participants’ experience while analyzing the data but kept a critical stance towards the participants’ narratives. The peer review process while analyzing data allowed both authors to keep a balanced insider and outsider positioning.
Findings
Teachers’ emotional experiences when teaching online
There were a wide range of emotional experiences that the teachers underwent on the migration of their traditional teaching to online platforms. Summarized in Table 3 are the emotions that topped the list of emotional experiences that the teachers reported. We grouped the emotions into two categories, negative emotions which emerged when the teachers were suddenly forced to switch to online teaching, and positive emotions which were felt when the teachers were more familiar with the new mode of delivery. These categories will help to explore the development of the full spectrum of teachers’ emotions in the migration from traditional physical classroom education to digital/distance education.Table 3. Teachers emotions.
Emotions Emotion-laden reasons
Negative emotions: switching to online teaching Confused Teachers’ unfamiliarity and lack of digital literacy
Worried Students’ low level of engagement
Exhausted Workload and longtime online exposure
Positive emotions: getting acquainted to online teaching More confident and adaptive Teachers’ capability to use technological infrastructure
Excited More stimulating classroom environment and increased students’ participation
Empathetic Teachers’ connection and empathy to students
Negative emotions: switching to online teaching with confusion, worry, and exhaustion
When COVID-19 hit Vietnam, the teachers understood clearly that education digitalization was the only choice to keep them safe and to keep their teaching job ongoing. However, the overnight switch did evoke negative emotional responses among the teachers, and some of them were particularly opposed to online teaching as the mode of instruction was changed.
In the beginning, the teachers’ confusion and worry stemmed mostly from their unfamiliarity and lack of experience with technological instructional tools and platforms. The teachers described themselves as “inexperienced,” “nervous,” or “not ready” for remote instruction delivery. Minh, for instance, said that he did not even know “how to start an online session.” At the start of the conversion to online teaching, An and Minh were both clumsy at technological maneuver when they could not “break out rooms” or “share screens on Zoom.” Nhung, Mai and Thu also reported that the first online sessions did not go smoothly because they were learning to use the new teaching tools. In this regard, all of a sudden, the teachers were being deskilled, turning from being experienced in a physical classroom setting into inexperienced teachers in a virtual one (Downing and Dyment, 2013; Pham and Phan, 2021).
In addition to the inexperience in using online tools, the teachers reported significant increase in their workload. The reason is the online platforms were also novel to students, which doubled the hardships for the participants since they had to both educate themselves to use technology effectively and instruct their students to use technology in their learning. The teachers described the class organization on those platforms as “messy,” requiring significant efforts from both themselves and the students. They had to prepare the lessons more carefully and had different back-up plans in cases of Internet connectivity breakdowns or their students’ failure to use technological devices. This created extra work for the teachers and urged them to quickly develop technology literacy. Duong, for instance, had to spend a whole day having one lesson digitalized. Another example was Nhung who reported that it took her 1 year to restructure and redesign a course because converting the face-to-face lessons to virtual platforms required multiple skills. For Nhung, it was a very challenging task. The significantly increased workload led to the teachers’ serious concerns about their health because long synchronous online classes left them exhausted. All participants mentioned physical health issues such as eye-straining, backache, headache, sore throat and fatigue, which made them less enjoy this transition and instead caused them more stress. Some of the teachers mentioned “mood swings,'' “bad-tempered,” and “aggressive” as the direct results of their deteriorating physical health.
The teachers further mentioned a worry over the quality of online teaching and learning because it, in Thao’s explanation, required more student autonomy and teacher’s effort to maintain their students’ engagement. Minh similarly raised his concern over the interactions between himself and students. “Sometimes, the atmosphere was a little bit depressing, and students were not as excited as they were before,” Minh said. This situation in turn led to increasing teacher talking time, which again contributed to health problems of the teachers. An’s account testified to this point.I hold too high expectations in myself and in my students. I had hoped the lesson would be interesting, I talked a lot and hoped to increase students’ interactions. But then I realized online teaching was never similar to offline teaching [...] Things just went sideways.
The low level of students’ engagement in a virtual classroom undoubtedly put stress on the emotional fabric of the teachers’ everyday practices (Pham and Phan, 2021). Furthermore, the students’ invisibility and inaudibility due to camera and microphone turn-off or breakdown in class brought “sadness” to the teachers. There were times when the teachers asked questions and there were no responses from the students, which caused disruption and demotivation to themselves and the whole class. The participants all admitted that it was frustrating to not be able to get a good sense of students’ comprehension, confusion, and general well-being through their facial expressions and demeanors as much as in face-to-face classrooms.
Positive emotions: getting acquainted to online teaching with confidence, excitement and empathy
The teachers gradually developed a wide range of positive emotional experiences towards the new mode of teaching. Most of them reported their excitement and confidence in using novel teaching platforms. They expressed their increasing confidence and capability in using technological infrastructure. They were also excited because of the new mode of teacher-student interaction. Specifically, while in traditional classes, it might be difficult for the students to express opinions on personal topics or sensitive issues, in online classes they were more willing to join the discussion because they could voice their opinions anonymously. Lan, for instance, recalled the first lesson of “Your childhood” she had at the beginning of the semester. By using Padlet (a real-time collaborative web platform on which users can upload, organize, and share content to virtual bulletin boards), there was a stream of answers, which actually boosted the environment and motivation of the whole class. Thao, another teacher, acknowledged that the Chatbox function on Zoom through which students could send teachers questions privately without disturbing others made them feel less embarrassed. Such features aided the teachers’ teaching performance and their students’ learning process, which they cared most about. The improvement in utilizing technology in teaching and students’ better responses to the shift in their learning platforms empowered the teachers and allowed them to start enjoying their new teaching practices. Mai, for instance, explicitly said she was “happy” and “satisfied” with her new online teaching routines. Moreover, the teachers showed their satisfaction when their students had a space to be more open to them.
In addition, the teachers grew more empathetic to the shift in their students’ learning process. The switch to online teaching not only encouraged the teachers to use innovation and creativity in their lessons but also highlighted the importance of teacher-student relationship. Mai cared for the students’ emotional well-being and tried to create more engaging and meaningful lessons:I felt sorry for my students. They were in a situation like my son’s as they could not go to school and meet their friends and teachers. That’s why I wanted to do more for them, I wanted my lessons to be more helpful and interesting for them so they could sustain their learning despite the ongoing challenges. There was one time when in the middle of a session, one of my students asked to be absent for half an hour so he could quickly get a COVID test per the requirement in his hometown. He came back after exactly 30 minutes and was very active, eager to participate in all activities. I felt so moved. How could I not care for my students? I told myself to try better because of their efforts and attitudes.
The teachers understood that their students had their own difficulties when studying online. During the interviews, many teachers emphasized that they tried to do their best to keep the lessons engaging and to ensure students’ engagement as much as possible so that the differences between online and offline learning would not affect their students’ learning results. They were more available to support students by answering emails and staying on Zoom after class to foster collaboration and connection with the students.
The range of emotions displayed by the teachers allows us to understand their agency and resistance to the difficulties brought about by the new situation of online teaching, increased workload, and higher health risks. Their emotions were influenced by discourses such as “teaching and learning quality,” or “student engagement.” They worried because of the possible reduced-quality teaching session. They felt exhausted while trying to maintain students’ interest when no real interactions happened in class. In this regard, the teachers’ emotional experiences were discursively constituted. While grappling with all the ensuing changes, they learned to negotiate to lower their expectations of a stimulating online lesson and students’ engagement in order to regain their confidence and excitement. In so doing, the teachers reframed their emotions in the service of the “new normal” happening to their life and work so that their professional values and efforts could be maintained. Simultaneously, the online platforms also worked as an emotional outlet for students, which partly enhanced teacher-student relationship because the power distance between teachers and students might not have allowed students to display their positive emotions to teachers previously.
Teachers emotion regulation strategies
Emotionally speaking, teaching online and engaging in remote teaching during COVID-19 presented divergent challenges to the teachers. In order to smooth the teaching performances on online platforms and to keep the differences between online and offline classrooms at the minimal level, emotion regulation was an indispensable mechanism through which the teachers could maintain a stimulating and engaging virtual learning environment. This section details the strategies that took place in various emotional expressions at different moments in the participants’ online teaching practices during the COVID-19 situation.
As the teachers invested emotionally in their profession and in every lesson, the switch to online delivery required them to regulate their emotions accordingly and to be strategic in order to be less emotionally vulnerable in the new situation. Summarized in Table 4 are different strategies that the teachers in this study used to regulate their emotions to achieve the teaching goals and alleviate the pressure both they and their students were put under. We grouped their strategies into two main categories: in-the-moment emotion regulation, indicating the strategies that the teachers used during the teaching sessions, and out-of-classroom emotion regulation, meaning that the strategies that the teachers used after their teaching sessions ended.Table 4. Teachers’ emotion regulation strategies.
Strategies Articulation of strategies
In-the-moment emotion regulation Using humor Making jokes during online sessions to alleviate pressure for both teachers and students
Bodily enhancement Dressing up and making up
Displaying negative emotions Letting students aware of teachers’ negative emotions
Using encouragement and compliments Giving students compliments to cheer-up the class atmosphere and teacher’s emotion
Taking breaks Inserting more breaks in the teaching sessions to avoid exhaustion and negative emotions
Out-of-classroom emotion regulation Accepting reality and using positive reframing Acknowledging the reality and necessity of online teaching, seeing the situation from a more affirming light (as a new learning process and opportunity for transformations)
Seeking emotional support Seeking comfort and understanding from others (students and colleagues), verbalizing unpleasant feelings with family members
Seeking instrumental support Seeking support and sharing from colleagues regarding pedagogical issues
Practising self-care Meditating, not putting too much pressure on themselves, and eating healthy food
Body movement Moving around the house (going upstairs downstairs, going out after the lessons to let the steam off)
From the teachers’ in-the-moment emotion regulation strategies, we could observe both genuine and regulated emotional expressions of the teachers, both negative and positive emotions. It should be noted that the genuine emotions of the teachers could be displayed or hidden, but not faked. In other words, the teachers did not mask their negative emotions by pretending they were having positive emotions, but made efforts to hide negativity so that the students would not be affected. In an Asian culture context like Vietnam, teachers’ hiding or suppressing negative emotions is frequently observed in prior research (Matsumoto, 1989). Like most teachers, the participants in this study saw negative emotion displays as a threat to their students’ learning atmosphere and enthusiasm, which meant they failed at their job. They chose to consider their negative emotions “nothing and then continued teaching,” to borrow An’s words. Minh, another participant, took the liberty of breaking the lessons into smaller sections to avoid stress for both himself and his students, by which he would feel less tired and re-control his emotions to prevent negativity. Even when the teachers chose to let the students know they were having negative emotions, it was not for the sake of finding an outlet for their dissatisfaction but for the sake of alerting their students about their responsibility for their own learning while everything was at stake. As Duong emphasized, “teachers are human beings, so I genuinely let my emotions be known to the students so that they know how to adjust their manners and attitude towards their learning.”
Other participants like Ha, Lan, and Mai used various strategies to create a stimulating atmosphere in their sessions, which would then make them feel happier during their teaching. While Mai used humor and gave constant encouragement to her students, Ha used mindful empowering questions such as “What are you most grateful for?” “What makes you happy today?” “Do you feel stressed?” at the beginning of every class as an ice-breaking activity, which helped to “create a stimulating class environment and foster connection between teacher and students.” Using another strategy, Hoa and Lan lift their emotions up in embodied ways by dressing up and using make-up. Hoa explained, “I dressed smartly as I often did in my face-to-face classes so that I had the feeling of going to work and meeting the students.”
When they chose emotional non-expressions while in class, teachers sought other outlets for their suppressed emotions, which we put under the category of out-of-classroom emotion regulation strategies. It means that the teachers seemed to “internalize the emotional rule that negative emotion was an individual problem and that should be taken care during one’s own private space” (Zembylas, 2005: 943). Their coping strategies included looking for mutual sharing from colleagues and sympathy from family members. Having a husband with the same profession as a university teacher, Hoa could find heartfelt sympathy from him, and talking with colleagues also helped her feel less alone because they were going through the same experiences. Similarly, Thao sought understanding from her family members, especially her husband, to help her manage new family organizations as a result of the work-from-home protocol and act as listeners to her suppressed emotions. Most teachers also practised self-care such as meditating or eating more healthy food to enhance their physical health and to “feel better,” in Hoa’s words. But most importantly, all of the teachers considered accepting reality as the main strategy in order to manage their emotions. They chose to see the switch to online teaching from a more positive perspective, considering this process as a chance of professional development and a learning process. Duong added that “distance learning offers a new opportunity for teachers because now we can even have extra tutoring online classes in which students from other provinces or cities can join, which is a new source of income for us.”
In light of the heavy emotional demands in the teaching profession, it is not surprising that the teachers themselves found the need to find spaces to discuss emotions and at the same time were careful to display their emotions. On the one hand, they showed their subscription to the discourse of professionalism in teaching by avoiding emotional exhibition in class, which is commonly found in literature (Zembylas, 2005). On the other hand, their emotional self-regulation strategies highlighted their agency in order to not suppress their own emotions which might lead to stress and mental health issues, but at the same time to not badly reflect on their teaching practices. The analysis shows evidence of teachers’ resistance to the unwritten rules of teachers trying to be emotionally professional, meaning hiding their negative emotions (Matsumoto, 1989). They wanted their negative emotions to be known by students so that changes in students’ behaviors and attitudes would happen. This also points out the power relations between teachers and students whereby teachers used their negative emotions to signal their dissatisfaction and expectation of students’ improvement. Furthermore, both the teachers in-the-moment emotion regulation and out-of-classroom emotion regulation reveal their self-negotiation and “self-control” in teaching (Zembylas, 2005).
Institutional policy and management on teachers’ emotions and teaching practices
When asked the participants to provide their comments on their institutional policy and management that supported teachers both emotionally and professionally in the context of the “new normal” situation, most participants agreed that the response of the university during the conversion to online learning was “timely,” “prompt,” and “flexible” although there were signs of overuse technology that disturbed their personal life. However, in terms of teachers emotions, the support was described as “insufficient,” “unspecific,” “short-term,” and “untimely,” leading to their reliance on self-regulation and self-care.
Due to the shift to online platforms for teaching and learning activities, the institution provided technical support for teachers’ professional development. The immediate adoption of online teaching at the beginning of the pandemic showed the institution’s proactivity and its priority of the safety of students and staff. In the beginning, as Duong reported, any online platforms (Zoom, Skype, and Google Meeting) could be used to make sure students’ learning continuation. Afterwards, according to the teachers, only Zoom was used to ensure consistency. There were a number of workshops and training sessions provided by the institution for the teachers to enhance their skills to effectively incorporate technology (Flipgrid, Padlet, Whiteboard) into their classroom. However, according to the participants, the workshops were mostly introductory training that dealt with technical issues, while what they wanted to be trained about was pedagogical skills of delivering online instruction. Moreover, the teachers struggled with student assessment and evaluation because they were not provided with appropriate tools to conduct online testing that ensured academic honesty.
Some teachers expressed their dislike and discomfort as they raised their concern over the overuse of online platforms for all professional activities in their institution. Thao was the most provocative in voicing her fear for the abuse of online tools. She held strong criticism towards the way her professional commitments consumed her personal life, which she pointed to the unorganized management at different levels of management in her institution.There were even meetings that started at 8:30 or 9 o’clock in the evening and lasted until 10 or 11p.m., which drove me crazy. Some of the meetings yielded no results and were merely time-consuming. Social networking sites now act as official platforms of professional communication in my institution. This has become more and more common, which I think is unprofessional because these channels should only be used for personal purposes. Now they are channels to discuss work matters, which makes me incredibly uncomfortable because people think it’s convenient. It’s way too convenient. I keep receiving instant messages day and night.
Thao showed discomfort towards the new professional practices in her university, which she dubbed as “unprofessional.” The excerpt from Thao’s interview revealed how much power the institution had on the teachers through the overuse of communication (social networking sites as platforms for professional communication), and temporal disciplining (day and night). Despite the dislike, Thao did not mention any specific resisting actions towards this issue, and neither did other participants including Nhung who also complained about virtual meeting after eight o’clock in the evening. The resistance at this stage was more in attitudes, but it could be developed into actions once the teachers found their personal life and family time was negatively influenced.
In terms of teachers’ emotions and well-being in general, the teachers thought the support from their institution was inadequate and too general, although there were initiatives from the managerial level. Several teachers mentioned the mindfulness talks and workshops that they could register and a short professional development course of “Becoming inspirational educators” that was mandatory for all teachers. Thu acknowledged that these two programmes were important in helping the teachers learn to regulate their emotions and “hold a positive emotion towards the teaching profession and our life in general.” Mai shared a similar viewpoint, adding that mindfulness activities were gradually added into recent workshops and meetings, indicating that teachers’ emotion was now more recognized in her workplace. The participants, however, emphasized self-regulation as the most important factor in their own well-being. Although they wanted and expected the institution to develop agendas on teachers’ well-being and emotion, they acknowledged the importance of self-care. Thao explained that:It is not easy to implement a program to support teachers’ emotions, despite its necessity, because every teacher’s individual life is different. Each has their own issues. That is not to mention some teachers don’t feel comfortable sharing with others about their personal life. What if a teacher complains about something and then is reported to the school manager, or judged by others? I don’t think teachers here will be willing to voice their emotions. Maybe in a small group, like what we are doing now. Sometimes when we have a group meeting, we may spend an hour and a half talking about our lives, sharing our concerns, or making complaints about our job.
Here the teachers relied on social bonding as a source of support for their burnout and an outlet for emotion display with less fear of judgment. Thu and Mai had a similar view towards the emotional care they received from their institution. “There was support, but just to cheer-up the teachers, which was not adequate,” Mai commented. Nhung and An said the support in terms of teachers’ emotion was “short-term,” and “lack of a long-term vision with consistency.” However, they did not expect the institution to go beyond what they did in terms of emotional care, but wanted to receive more technical training provision for teachers to be more proficient in utilizing online pedagogical and assessment resources. Again, the teachers relied more on themselves to practise self-care and regulate their emotions, not only because “there is not such a thing as institutional emotional support for teachers” according to Thao’s claim, but also because when there was an initiative, it was not to get down to the actual issue of teachers’ emotional experiences, but to encourage teachers, superficially, to continue their teaching performance.
The lack of a systematic and comprehensive support scheme from the institution forced the teachers to “fight their own fight,” meaning that they had to learn to resist implicit emotional rules of professional teaching in their own ways. It seems that the best resistance was to create spaces for exciting emotional connections among students and teachers, and for self-satisfaction of teachers themselves. In agreement with Boler (1999), we contend that resistances function both as defenses against vulnerability in this particular situation of the pandemic, and as assertions of power in the face of impositions of switching to online teaching while maintaining education quality and student engagement.
Discussion and implications
The article offers an exploration of the emotional experiences of Vietnam language teachers at the onset of the global pandemic when online teaching came as both a rescue for teaching-learning activities and a burden for teachers. Under these conditions, teacher emotions were shaped and performed. From the poststructuralist approach, the article discovers how the emotional rules (for instance hiding emotions in class) and discursive practices (teacher as parent, teaching for quality) implicitly and explicitly imposed on teachers when choosing the expression of certain emotions and the disciplining of others. While COVID-19 has forced teachers to face perhaps the most jarring changes in their professions, it has also brought heartening educational transformations. Specifically, teacher-student interactions reduced, teachers’ care increased. Even when the teachers reported negative emotions, these negative emotions (anxiety, worry or confusion) stemmed from their care for students, which reflected the caring ethics of the teaching profession, particularly when teachers were aware of the precarity and instability their students were experiencing. The analysis on teachers’ emotion expression and regulation unpacks how they attempted to exhibit positive emotions and avoid negative emotions because they cared about their students’ well-being and emotions. They likened their students’ online learning situation to their children’s, which showed how they saw their identity as a teacher somewhat similar to that as a parent.
Teacher emotions are not merely the effects of outside structures which in this case refer to the overnight conversion to online teaching and inadequate institutional support and infrastructure. Teacher emotions are “performative,” meaning that the ways teachers understand, experience, perform, and talk about emotions are highly related to their understanding of culture, power, and themselves. Although there was not any explicit rules written in the institution policies that laid out rules of emotional display and expression for teachers, it needs to be emphasized that the teachers internalized the values of how a teacher should express emotions in class and what emotions deemed appropriate and inappropriate thanks to their own experiences as students and their learning from their former teachers. By describing events (the COVID-19 pandemic), objects (online platforms, technological resources), persons (teachers, students, institution manager) and their relationships (teacher-student interactions, tensions in virtual classrooms), the article highlights the ways emotions were experienced and expressed in relation to the teacher–self (individual reality), the others (social interactions) and the university culture in general (sociopolitical context). While the teachers were confused, anxious and unsure about the overnight digitalization of education, they were also excited and satisfied with the new educational tools and teaching platforms. Although the COVID-19 advent was a good start for developing an institutional policy on teachers’ well-being and teachers’ emotion in particular, the teachers still relied more on themselves to regulate emotions: either enhancing positive emotions or controlling negative emotions. They continued to believe that practising self-care and emotional self-regulation was most salient in their teaching profession. The teachers’ emotions, in the context of the pandemic, were put under the strain of not getting themselves disappointed, at the same time not letting their students down.
The article provides some insights into the issue of teacher emotions and implications for institutional policy. First, the teachers' emotional experiences highlight the fact that the advent of the global pandemic has altered the way teachers work and hence has brought their challenges to the mix: health concerns, social and physical distancing, stress, increased childcare and family care (Pham and Phan, 2021). However, the conversion to online education seemed to be prioritized over other issues, given that the teachers had other commitments and their personal lives were equally affected by the crisis. Their experiences can testify the point made by MacIntyre et al. (2020) that the long-term consequences of COVID-19 for language teachers and teaching are unknown. In that sense, institutions should take these factors of teachers’ personal life commitments into account so that teachers will not be drained from the overwhelming situation both at work and at home. We concur with Starkey et al. (2021) that policies in education should be oriented towards the teachers’ professional learning within the transformation of educational models rather than the mere provision of technological resources. Furthermore, given the consequences of laborious online teaching, there should be efforts from policymakers to reduce the duration of online classes for students and online work hours for teachers to minimize health problems amid the pandemic COVID-19 (Ganne et al., 2021). Therefore, a comprehensive support scheme in terms of psychological, technological, methodological, and professional development for teachers is desirable to both ensure effective online education and minimize negative impacts of the pandemic on the educational process. We call for attention to teachers’ emotions as an integral dimension of the profession, regardless of the physical or virtual setting of the classroom.
Conclusion
While other studies about teaching and learning during the COVID-19 crisis often put the way classrooms “going virtual” in the front of the inquiry, this paper chose to explore teachers’ emotions when they had to quickly shift their teaching to new digital platforms. Specifically, the study charted the emotional experiences of 10 English language teachers in a university in Vietnam when they were required to convert their traditional face-to-face teaching to online synchronous classrooms. The study highlighted the teachers’ strategies to regulate their emotions and avoid a health break-down, both in and out-of-class.
We acknowledge that there are several limitations in this study. First, the sample size is small, hence, the findings cannot be represented for and generalized to a wider population of Vietnamese English language teachers at the tertiary level. Instead, the findings of this study provide specific insight within the scope of this research and can be understood as one among the first attempts to study the cohort of foreign language teachers in Vietnam in the pandemic situation (see more in Pham and Phan, 2021, 2022), which can invite future studies to expand the discussion initiated in this paper. Second, the accuracy of quotes can be influenced by the translation from Vietnamese to English since translation does not always reflect the nuances of what was said in the original language. In this regard, there could be a risk of inaccuracy in expressing the teachers’ emotions in written English. However, the authors were convinced that the essential message of the informants was reflected in these pages. Furthermore, although we did attempt to include both female and male teachers, the uneven representation of the participants’ gender in this study did not allow us to provide any conclusion regarding the differences in emotional expressions and regulation between male and female teachers. This point can be a suggestion for future inquiry about teachers’ emotions in online teaching.
Linh Thi Thuy Pham was an English lecturer in the Faculty of English Language Teacher Education (FELTE), University of Languages and International Studies (ULIS, VNU) from 2011. She earned her Master of Education degree in Australia in 2016. She taught English skills and research methodology courses at her university for students as prospect English teachers in Vietnam.
Anh Ngoc Quynh Phan has completed her PhD in Education at The University of Auckland, New Zealand. Her work has been published in Journal of Gender Studies; Globalisation, Education, and Societies; Asia Pacific Journal of Education; Studies in Continuting Education, etc. She is joining Centre for the Study of Higher Education at the University of Kent, United Kingdom, as a Lecturer in Higher Education.
ORCID iDs
Anh Ngoc Quynh Phan https://orcid.org/0000-0001-9979-1321
Linh Thi Thuy Pham https://orcid.org/0000-0003-1341-2680
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the University of Languages and International Studies (ULIS, VNU) [Project Number N.21.06].
Data availability: The data that support the findings of this study are available from the corresponding author, ANQP, upon reasonable request.
==== Refs
References
Abu-Lughod L Lutz C (1990) Introduction: emotion, discourse, and the politics of everyday life. In: Lutz C Abu-Lughod L (eds), Language and the Politics of Emotion. Cambridge University Press, pp. 1–23.
Adedoyin OB Soykan E (2020) Covid-19 pandemic and online learning: the challenges and opportunities. Interactive Learning Environments 31 .
Benesch S (2017) Emotions and English Language Teaching. England: Routledge.
Braun V Clarke V (2019) Reflecting on reflexive thematic analysis. Qualitative Research in Sport, Exercise and Health 11 (4 ): 589–597.
Boler M (1999) Feeling Power: Emotions and Education. New York: Routledge.
Bryman A (2012) Social Research Methods. 4th edition. Oxford: Oxford University Press.
Crawford J Butler-Henderson K Rudolph J , et al. (2020) COVID-19: 20 countries' higher education intra-period digital pedagogy responses. Journal of Applied Teaching and Learning 3 (1 ): 1–20.
Downing JJ Dyment JE (2013) Teacher educators’ readiness, preparation and perceptions of preparing pre-service teachers in a fully online environment: an exploratory study. The Teacher Educator 48 (2 ): 96–109.
Foreman-Brown G Fitzpatrick E Twyford K (2022) Reimagining teacher identity in the post-covid-19 university: becoming digitally savvy, reflective in practice, collaborative, and relational. Educational and Developmental Psychologist. Ahead-of-print.
Ganne P Najeeb S Chaitanya G , et al. (2021) Digital eye strain epidemic amid COVID-19 pandemic - a cross-sectional survey. Ophthalmic Epidemiology 28 (4 ): 285–292.33369521
Gkonou C Dewaele J-M King J (2020) The Emotional Rollercoaster of Language Teaching. Bristol, UK: Multilingual Matters.
Gkonou C Miller ER (2017) Caring and emotional labour: language teachers’ engagement with anxious learners in private language school classrooms. Language Teaching Research 23 (3 ): 372–387. DOI: 10.1177/1362168817728739
Hochschild AR (1979) Emotion work, feeling rules, and social structure. American Journal of Sociology 85 (3 ): 551–575.
Hochschild AR (1983) The Managed Heart : Commercialization of Human Feeling. California, USA: University of California Press.
Keller MM Chang ML Becker E , et al. (2014) Teachers’ emotional experiences and exhaustion as predictors of emotional labor in the classroom: an experience sampling study. Frontiers in Psychology 5 : 1442.25566124
Lester S (1999) An Introduction to Phenomenological Research. Taunton UK: Stan Lester Developments. https://www.rgs.org/CMSPages/GetFile.aspx?nodeguid=7ad9b8d4-6a93-4269-94d2-585983364b51&lang=en-GB
MacIntyre PD Gregersen T Mercer S (2020) Language teachers’ coping strategies during the covid-19 conversion to online teaching: correlations with stress, wellbeing and negative emotions. System 94 .
Mansfield CF Beltman S Price A , et al. (2012) “Don’t sweat the small stuff:” Understanding teacher resilience at the chalkface. Teaching and Teacher Education 28 (3 ): 357–367.
Matsumoto D (1989) Cultural influences on the perception of emotion. Journal of Cross-Cultural Psychology 20 (1 ): 92–105.
Merriam SB (2002) Qualitative Research in Practice. San Francisco: Jossey-Bass.
Merriam SB (2009) Qualitative Research: A Guide to Design and Implementation. San Francisco, CA: Jossey-Bass.
O’Toole VM Friesen MD (2016) Teachers as first responders in tragedy: the role of emotion in teacher adjustment eighteen months post-earth- quake. Teaching and Teacher Education 59 : 57–67.
Patton MQ (2002) Qualitative Research and Evaluation Methods. 3rd edition. Thousand Oaks, CA: Sage Publications.
Pham LTT Phan ANQ (2021) “Let’s accept it”: Vietnamese university language teachers’ emotion in online synchronous teaching in response to COVID-19. Educational and Developmental Psychologist 40 : 115–124. Ahead-of-print: 1-10.
Pham LTT Phan ANQ (2022) Whilst COVID-19: the educational migration to online platforms and lessons learned. The Clearing House: A Journal of Educational Strategies, Issues and Ideas 95 : 159–165. Ahead-of-print: 1-5.
Philipp A Schüpbach H (2010) Longitudinal effects of emotional labor on emotional exhaustion and dedication of teachers. Journal of Occupational Health Psychology 15 (4 ): 194–504.
Rahimi N Martin NL (2020) Challenges and Strategies for Online Teaching in Information Technology and Other Computing Programs. In Proceedings of the 21st annual conference of the special interest group in information technology education, SIGITE, 7–8 October 2020. 218–222.
Rapanta C Botturi L Goodyear P , et al. (2020) Online university teaching during and after the covid-19 crisis: refocusing teacher presence and learning activity. Postdigital Science and Education 2 (3 ): 923–945.
Rosaldo MZ (1984) Toward an anthropology of self and feeling. In: Shweder RA LeVine RA (eds), Culture Theory: Essays on Mind, Self, and Emotion. Cambridge: University Press, pp. 137–158.
Rubin HJ Rubin IS (2012) Qualitative Interviewing: The Art of Hearing Data. Thousand Oaks, CA: Sage.
Santamaría MD Mondragon NI Santxo NB , et al. (2021) Teacher stress, anxiety and depression at the beginning of the academic year during the COVID-19 pandemic. Global Mental Health 8 : e14.34192001
Schutz PA (2014) Inquiry on teachers’ emotion. Educational Psychologist 49 (1 ): 1–12.
Sokal L Trudel LE Babb J (2020) Canadian teachers’ attitudes toward change, efficacy, and burnout during the COVID-19 pandemic. International Journal of Educational Research Open 1 : 100016.
Starkey L Shonfeld M Prestridge S , et al. (2021) Special issue: covid-19 and the role of technology and pedagogy on school education during a pandemic. Technology, Pedagogy and Education 30 (1 ): 1–5.
Sutton RE Mudrey-Camino R Knight CC (2009) Teachers’ emotion regulation and classroom management. Theory into Practice 48 (2 ): 130–137.
Trust T Krutka DG Carpenter JP (2016) “Together we are better”: professional learning networks for teachers. Computers and Education 102 : 15–34.
Weedon C (1987) Feminist Practice and Poststructuralist Theory. Oxford, UK: B. Blackwell.
Zembylas M (2002a) Constructing genealogies of teachers’ emotions in science teaching. Journal of Research in Science Teaching 39 : 79–103.
Zembylas M (2002b) “Structures of feeling” in curriculum and teaching: theorizing the emotional rules. Educational Theory 52 (2 ): 187–208.
Zembylas M (2003a) Caring for teacher emotion: reflections on teacher self-development. Studies in Philosophy and Education 22 : 103–125.
Zembylas M (2003b) Interrogating ‘‘teacher identity’’: emotion, resistance, and self-formation. Educational Theory 53 : 107–127.
Zembylas M (2005) Discursive practices, genealogies, and emotional rules: a poststructuralist view on emotion and identity in teaching. Teaching and Teacher Education 21 : 935–948.
Zembylas M (2007) Theory and methodology in researching emotions in education. International Journal of Research and Method in Education 30 (1 ): 57–72.
|
PMC010xxxxxx/PMC10229198.txt |
==== Front
Public Health Pract (Oxf)
Public Health Pract (Oxf)
Public Health in Practice
2666-5352
The Authors. Published by Elsevier Ltd on behalf of The Royal Society for Public Health.
S2666-5352(23)00039-3
10.1016/j.puhip.2023.100393
100393
Article
Impact of nasal photodisinfection on SARS-CoV-2 infection in an industrial workplace
Rusk Richard a∗
Hodge Judy b
a Department of Emergency Medicine, Rady Faculty of Health Sciences, University of Manitoba, Winnipeg, Manitoba, Canada
b Katrime Integrated Health, Winnipeg, Manitoba, Canada
∗ Corresponding author.
30 5 2023
12 2023
30 5 2023
6 100393100393
4 1 2023
25 4 2023
2 5 2023
© 2023 The Authors
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Objectives
We aimed to evaluate a quality improvement initiative designed to control SARS-CoV-2 (COVID) using the large-scale deployment of antimicrobial photodisinfection therapy (aPDT) for nasal decolonization in a Canadian industrial workplace (a food processing plant).
Study design
Using a retrospective chart review of treatment questionnaires, linked to COVID laboratory testing results, a quality improvement assessment was analyzed to determine treatment effectiveness and safety.
Methods
This voluntary aPDT intervention involved the administration of a light-sensitive liquid to the nose followed by nonthermal red-light irradiation on a weekly basis. Employees in food processing industries are at increased risk for COVID infection due to the nature of their work environments. In an effort to mitigate the transmission and consequences of the disease among such workers and the community at large, aPDT was added to a well-established bundle of pre-existing pandemic safety measures (e.g., mask-wearing, testing, contact tracing, workplace-engineered barriers, increased paid sick leave).
Results
From December 2020 to May 2021, we found high interest in and compliance with aPDT treatment, along with a statistically significant lower PCR test positivity rate in the study population in comparison to the case rates for the local Canadian province. Treatment safety monitoring and outcomes of the aPDT program demonstrated no serious adverse events.
Conclusions
This study suggests nasal photodisinfection provides safe and effective COVID viral suppression when deployed across the majority of workers in an industrial workplace setting.
Keywords
COVID-19
Occupational safety
Antimicrobial photodisinfection therapy
Essential workers
Nasal decolonization
Infection control
Broad spectrum antimicrobial
==== Body
pmc1 Introduction
SARS-CoV-2 (COVID-19) outbreaks are known to cause significant adverse outcomes, including worker absenteeism, supply chain interruptions, and acute and long-term human illnesses, hospitalizations, and deaths [1]. Beginning in 2020, medical providers and employers throughout the world rapidly implemented enhanced safety protocols in response to the SARS-CoV-2 pandemic, such as personal protective equipment (PPE), social distancing, workplace-engineered barriers, and improved handwashing [2]. Despite these interventions, the impact of the first waves of the pandemic dramatically threatened human and animal health when COVID-19 disproportionately affected essential employees within the food processing (e.g., poultry, pork, beef) industry [3,4]. Staff working in food processing facilities remain at increased risk for SARS-CoV-2 transmission based on the duration and type of work interactions with other employees, such as being in close contact on processing lines, in cold temperatures, and within enclosed work environments during multi-hour shifts [5].
Since the onset of the pandemic, meat processing plants have reported a disproportionately large number of employees contracting COVID-19,2 requiring these plants to shut down operations. Furthermore, the presence of slaughtering plants in a community has been found to be associated with 400–600 additional COVID-19 cases per hundred thousand from the baseline rate, and an increase in the death rate by 7–10 deaths per hundred thousand, or 37 to 50% over the baseline rate [6]. However, it has been documented that transmission in the workplace decreases as safety interventions are implemented [1,3]. Based on this and similar research studies, several Canadian workplaces added nasal photodisinfection to their protection protocols as an emergency response to the pandemic and in consideration of the documented safety and efficacy of the treatment [7].
Nasal photodisinfection, known more specifically as antimicrobial photodisinfection therapy (aPDT), has been investigated as an adjunct to these interventions to further mitigate the potential impact of COVID-19 outbreaks within the workplace and surrounding communities, including the potential impact of asymptomatic transmission [8]. aPDT involves the application of a topical photosensitizer inside the nose, which has a positive charge that preferentially binds to negatively-charged microorganisms. Two small nasal cones connected to a light source are then inserted into the nares (nostrils) to activate the photosensitizer using a specific wavelength of red light. During illumination, the excited photosensitizer reacts with nearby oxygen which generates reactive oxygen species (ROS) that destroy a broad spectrum of microbes including bacteria, viruses, and fungi. This treatment is painless and takes approximately 5 min to administer.
aPDT has been used in Canadian hospitals since 2011 [7] and has proven effective in the destruction of SARS-CoV-2 at the genomic (RNA) level [[9], [10], [11], [12]], as well as against spike proteins and receptor binding domains, both without adversely affecting human cells. A recent clinical trial using aPDT therapy on SARS-CoV-2 positive patients yielded a 90% reduction in infectivity of these cases [13]. In hospital presurgical deployment studies, aPDT intranasal therapy has been shown to result in dramatically-improved surgical outcomes (e.g., reduction of surgical site infections) [7]. Unlike traditional antimicrobials, aPDT does not lead to the development of resistance in targeted pathogens.
In this study, we evaluated a novel, large-scale deployment of nasal photodisinfection (Steriwave™ ND, Ondine Biomedical Inc., Vancouver, Canada) beginning in July 2020. Nasal photodisinfection technology was administered to employees in Canadian workplace settings in addition to other SARS-CoV-2 safety measures recommended by the U.S. Centers for Disease Control and Prevention (CDC) that had been implemented since the start of the pandemic [5]. This manuscript documents a retrospective, industrial workplace quality assurance program analysis that received Canadian ethics review board approval (HS25575 H2022:220).
2 Methods
Beginning July 16, 2020, aPDT was offered to a group of approximately 1500 employees operating in a single food processing plant in western Canada. Enrollment was voluntary and all participants were provided with a written consent document outlining the protocol. Signed consent forms were obtained from each individual and were stored securely within an electronic medical record (QHR Technologies Inc., Accuro® Electronic Medical Software). Processing plant management instituted stringent protocols to ensure employees were properly informed and safeguarded during the consent process, including education, question and answer sessions, pamphlet distribution, access to clinical assessments, understanding of their ability to discontinue participation at any time, and ongoing informed consent measures. Volunteers were given an incentive to encourage compliance with a bundle of infection prevention efforts, including aPDT, that consisted of fifty Canadian dollars per week. Employees were treated with aPDT weekly (on the same day and at approximately the same time each week) in consideration of clinical and operational guidance. All treatments were free of charge and completed during paid work time. Accessibility to the aPDT treatments was widespread in order to maximize ethical access; adherence levels were used to evaluate efficacy, the importance of compliance across the workforce, and the role of incentives in increasing participation. Target employee compliance with aPDT was 90% by September 15, 2020; the plant reached 75% adherence as of October 29, 2020. Prior to and running concurrently with this intervention, the plant also proactively implemented multiple safety measures that became the standard within the food processing industry [2,5]. For example, beginning in March 2020, increased paid sick leave, additional outdoor break rooms, third-party cleaning teams to disinfect high-touch surfaces three times per shift (after each break), engineered barriers, testing, social distancing during breaks, and pre-shift temperature and health screening were implemented. Furthermore, during the same period, participants were encouraged to continue to maintain high compliance with all CDC-recommended safety measures through the use of reminders, internal education videos, and trained staff available to answer any questions or concerns.
Data collected included health screenings for symptoms of active SARS-CoV-2 infection, SARS-CoV-2 test results of suspected cases, aPDT treatment frequency, adverse events/side effects from aPDT, and employee satisfaction measures. Employee satisfaction results are outside of the scope of this manuscript; however, the treatment was shown to be well received by the employee volunteers (J. Hodge, personal communication, 2021). Furthermore, participant questionnaires were completed before and one week after every aPDT session to monitor for any adverse events during treatment or within the 24-h period after treatment. Data were stored in an accredited, secure electronic medical record (QHR Technologies Inc., Accuro ® Electronic Medical Software) and within an employee database using unique identifiers to maintain anonymity and protect personal data.
3 Statistical analysis
SARS-CoV-2 polymerase chain reaction (PCR) test positivity data from individuals experiencing symptoms (e.g., fever, cough) who worked at the food processing plant were considered for this analysis. For privacy protection, individuals' treatment and testing data were linked using unique identifiers that replaced the need for sensitive patient information within the dataset. COVID-19 test positivity rates were calculated using the total number of positive tests (numerator) compared to the total number of positive and negative tests (denominator). Any tests with inconclusive results (n = 19) were not included in the calculation of the positivity rate. Exact binomial 95% confidence intervals (CIs) for test positivity were determined, and the statistical significance of the test positivity rates in the food plant were calculated using a Fisher's Exact Test. Also considered in this analysis were PCR test positivity rates for the same time periods from the entire Canadian province in which the plant was located. These data were downloaded from the public access database on April 1, 2022 [14]. All analyses were retrospective and were conducted in SAS version 9.4 and Microsoft Excel.
4 Results
Table 1 demonstrates the COVID-19 PCR test positivity rates among the food processing plant workers from December 16, 2020 to May 1, 2021 as compared to PCR test positivity rates for the entire province for the same date range.Table 1 Comparing test positivity rates between workplace and overall province (Dec 16, 2020 to May 1, 2021).
Table 1Workplace (n = 558) Province (n = 273,538) P value
n (%) 95% CI n (%) 95% CI
3 (0.5) 0.1–1.6 17,473 (6.4) 6.3–6.5 <0.0001
The total number of PCR tests in the workplace totaled 558; the total number of PCR tests in the province totaled 273,538. Among the plant workers, there were three positive tests, representing 0.5% 95% CI [0.1, 1.6] of the total tests conducted in the time period. In the province as a whole, there were 17,473 positive tests, representing 6.4% 95% CI [6.3, 6.5] of the total tests conducted in the time period. A statistically significant difference (p < 0.0001) in the test positivity rate for the food processing plant workers (0.5%) versus the overall province (6.4%) was identified. This study period was chosen to capture data from points at which full aPDT compliance (∼75%) was reached within the workplace population, the second wave of the pandemic was ongoing but the third wave of the pandemic had not yet occurred, and employees had access to on-site PCR testing with results available within 24 h from a private, fully-accredited diagnostic laboratory.
Table 2, Table 3 illustrate all adverse events (side effects) in the workplace population reported by surveys conducted during treatment in real-time and one week prior (retrospectively) for the period within 24 h after aPDT. These data were collected from December 16, 2020, to May 1, 2021. Side effects were stratified by those that were expected and unexpected.Table 2 Survey responses of side effects reported during and within 24 h of treatment with antimicrobial photodisinfection therapy.
Table 2 During Treatment (n = 21,459), n (%) Within 24 Hours (n = 21,261), n (%) Total, n (%)
Survey Responses
Reported side effectsa 86 (0.4) 738 (3.5)
No side effects 21,373 (99.6) 20,523 (96.5)
Total 21,459 (100.0) 21,261 (100.0)
Side Effects Reported
Expected 61 (0.3) 712 (3.5) 773 (3.6)
Unexpected 30 (0.1) 248 (1.2) 278 (1.3)
Total 91 (0.4) 960 (4.7) 1051 (4.9)
a Reported side effects (86, 738) are less than total side effects reported (91, 960) as some respondents indicated having multiple effects from one treatment.
Table 3 Survey responses of expected versus unexpected side effects reported during and within 24 h of treatment with antimicrobial photodisinfection therapy.
Table 3 During Treatment (n = 21,459), n (%) Within 24 Hours (n = 21,261), n (%) Total, n (%)
Expected Side Effects
Runny nose 14 (0.1) 333 (1.6) 347 (1.6)
Sneeze 17 (.1) 234 (1.1) 251 (1.2)
Nose irritationa 8 (0.0) 24 (0.1) 32 (0.1)
Throat irritation 2 (0.0) 5 (0.0) 7 (0.0)
Itchy nose 15 (0.1) 53 (0.3) 68 (0.3)
Nasal congestion 39 (0.2) 39 (0.2)
Itchy throat 4 (0.0) 17 (0.1) 21 (0.1)
Odd smell/taste 6 (0.0) 6 (0.0)
Warm feeling 1 (0.0) 1 (0.0) 2 (0.0)
Total 61 (0.3) 712 (3.5) 773 (3.6)
Unexpected Side Effects
Headache 1 (0.0) 101 (0.5) 102 (0.5)
Dry nose 15 (0.1) 78 (0.4) 93 (0.4)
Dry throat 4 (0.0) 29 (0.1) 33 (0.2)
Nose bleed 9 (0.0) 16 (0.1) 25 (0.1)
Dizzy 4 (0.0) 4 (0.0)
Red/blood-tinged mucus 1 (0.0) 9 (0.0) 10 (0.0)
Furunculosis 4 (0.0) 4 (0.0)
Uncomfortable 1 (0.0) 1 (0.0)
Increased acne 2 (0.0) 2 (0.0)
Anxiety 1 (0.0) 1 (0.0)
Shortness of breath 1 (0.0) 1 (0.0)
Stomachache 1 (0.0) 1 (0.0)
Triggered sinusitis 1 (0.0) 1 (0.0)
Other - unspecified 1 (0.0) 2 (0.0) 3 (0.0)
Total 30 (0.1) 248 (1.2) 278 (1.3)
a Nose irritation includes three instances in which respondent indicated “redness in the nose” within 24 h after treatment.
The majority of surveys indicated employees experienced no adverse response during treatment (99.6%) or within 24 h after treatment (96.5%). Of the total surveys (n = 21,459) collected during aPDT treatment, 86 (0.4%) indicated some reaction. Of the responses collected retrospectively after treatment (21,261), 738 (3.5%) indicated some mild reaction. The most common, expected treatment-related side effects were runny nose, sneeze, and itchy nose. The most common, unexpected side effects were headache, dry nose, and dry throat. No severe adverse reactions from aPDT were reported and no treatments were discontinued due to adverse events. Expected and unexpected treatment-related side effect characterizations are based on approximately ten years of historical Canadian use of aPDT in other healthcare settings.
5 Discussion
This is the first published study investigating aPDT coupled with an industry-standard bundle of interventions to evaluate COVID-19 test positivity rates of employees in a large commercial food processing operation. Among participants, treatment was well-received and voluntary compliance was high. The results in Table 1 indicate a statistically-significant decrease in test positivity rates when compared to those in the greater Canadian province. This preliminary study suggests that deploying aPDT in a commercial environment with high compliance among workers could decrease SARS-CoV-2 test positivity rates, decrease the incidence of disease among workers during times of high transmission, and potentially decrease the spread of disease in the larger community. Furthermore, aPDT has been shown to be effective against all SARS-CoV-2 variants as well as other viruses [[9], [10], [11], [12], [13],15]; therefore, the technique may prove to be of significant public health benefit as the COVID-19 pandemic evolves alongside future waves of viral illnesses (e.g., influenza). While certain factors of transmission may be modified to reduce the risk of COVID-19, food processing plant employees are subject to close contact for extended periods of time, creating a high-risk environment. As a result, precautions that may be effective in a community setting could be less impactful in food processing plants, which may necessitate additional safety measures [[16], [17], [18]]. Based on the method of SARS-CoV-2 transmission and the nose being a primary point of entry [15], routine nasal decolonization with aPDT may be an effective method to enhance standard SARS-CoV-2 safety measures such as high-quality masks. In this analysis, we evaluated aPDT that was deployed on an emergency basis in an attempt to administer a safe and efficacious intervention [7] to a vulnerable population (essential workers) at increased risk of death and disease who were presented with no approved alternatives before vaccinations were available.
Although no single intervention can adequately protect workers from COVID-19 [18], by offering — and incentivizing — a bundle of interventions, including novel aPDT treatment, employers may be able to better protect their employees, continue the production and distribution of food products production without disruption, and help safeguard the community at large. This kind of mitigation effort could have broader implications to other employee demographics, industries, and countries.
6 Limitations
This study includes certain limitations related to the retrospective design. Also, conclusions were drawn from data at a single industrial plant in Canada, and thus are not generalizable to the larger commercial footprint across Canada. The sample size was relatively small and should be increased in future studies across larger populations. While the present study demonstrated no significant relationship of covariates, such as age, gender, ethnicity, or treatment number to outcomes, potential confounding factors should be addressed in future studies. Furthermore, while industry standard SARS CoV-2 safety measures were established prior to the introduction of aPDT, the impact of the additional nasal photodisinfection can only be associated with the outcomes. Also, while other food processing plants that did not deploy aPDT continued to experience outbreaks, a direct causal relationship associated with the addition of aPDT was not definitively concluded. Lastly, employees were aware they were being observed and were incentivized to adhere to all safety measures which could have increased compliance and led to the Hawthorne effect (participant observation awareness).
7 Conclusion
Results of this study indicate intranasal aPDT added to a bundle of workplace safety interventions potentially suppressed the incidence of COVID-19 cases in an industrial food processing setting during the height of the SARS-CoV-2 (Delta variant) pandemic. No significant adverse events were detected among 21,261 questionnaires administered after treatment, and the process was well-tolerated with high (75%) terminal compliance. Additionally, this intervention may have had broader implications for the surrounding community by mitigating SARS-CoV-2 test positivity among plant workers when protocol compliance was maintained. It is widely documented that food processing industry essential workers are disproportionately affected by COVID-19 infection [6,16,17]. Thus, an important implication from the findings in this study is that enterprise-to-community transmission may have been reduced, preventing acute and long-term illness, disability, and death in plant employees, their families, and the broader community. This type of mitigation effort could have broader implications to other employee demographics, industries, and countries and further studies are warranted.
Author contributions
This was a collaborative effort between the two authors, with contributions for all aspects being shared: Conceptualization, Rusk. and Hodge.; methodology, Rusk. and Hodge.; software, Rusk.; validation, Rusk. and Hodge.; formal analysis, Rusk.; resources, Rusk. and Hodge.; data curation, Rusk and Hodge.; writing—original draft preparation, Rusk. and Hodge.; writing—review and editing, Rusk. and Hodge.; visualization, Rusk. and Hodge.; supervision, Rusk.; project administration, Hodge. All authors have read and agreed to the published version of the manuscript.
Funding
Funding was internal from personal financing by the authors, through the contribution of their time and collaboration. Hodge is a salaried employee of the food processing company.
Statement of ethical approval
Institutional Review Board approval for this study was received from the University of Manitoba.
Informed consent statement
Written informed consent was obtained from employees receiving Steriwave™ aPDT treatment and from employees who were tested by PCR for SARS-CoV-2 through company-provided on-site COVID-19 testing.
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.
Acknowledgements
We are grateful to Katrina Harper, Cali Cerami, and Jason Hickok for their valuable assistance in preparing the manuscript.
==== Refs
References
1 Waltenburg M.A. Rose C.E. Victoroff T. Butterfield M. Dillaha J.A. Heinzerling A. Coronavirus disease among workers in food processing, food manufacturing, and agriculture workplaces Emerg. Infect. Dis. 27 1 2021 243 249 10.3201/eid2701.203821 33075274
2 Dyal J.W. Grant M.P. Broadwater K. Bjork A. Waltenburg M.A. Gibbins J.D. COVID-19 among workers in meat and poultry processing facilities - 19 states, April 2020 MMWR Morb. Mortal. Wkly. Rep. 69 18 2020 10.15585/mmwr.mm6918e3
3 VanderWaal K. Black L. Hodge J. Bedada A. Dee S. Modeling transmission dynamics and effectiveness of worker screening programs for SARS-CoV-2 in pork processing plants PLoS One 16 9 2021 e0249143 10.1371/journal.pone.0249143
4 Marchant-Forde J.N. Boyle L.A. COVID-19 effects on Livestock production: a one welfare issue. [published correction appears in front vet sci. 2020 nov 19;7:625372] Front. Vet. Sci. 7 2020 585787 10.3389/fvets.2020.585787
5 Waltenburg M.A. Victoroff T. Rose C.E. Butterfield M. Jervis R.H. Fedak K.M. Update: COVID-19 among workers in meat and poultry processing facilities ― United States, april–may 2020 MMWR Morb. Mortal. Wkly. Rep. 69 27 2020 887 892 10.15585/mmwr.mm6927e2 32644986
6 Taylor C.A. Boulos C. Almond D. Livestock plants and COVID-19 transmission Proc. Natl. Acad. Sci. U.S.A. 117 50 2020 31706 31715 10.1073/PNAS.2010115117 33214147
7 Bryce E. Wong T. Forrester L. Masri B. Jeske D. Barr K. Nasal photodisinfection and chlorhexidine wipes decrease surgical site infections: a historical control study and propensity analysis J. Hosp. Infect. 88 2 2014 89 95 10.1016/j.jhin.2014.06.017 [published correction appears in J Hosp Infect. 2015 Sep;91(1):93] 25171975
8 Honein M.A. Christie A. Rose D.A. Brooks J.T. Meaney-Delman D. Cohn A. Summary of guidance for public health strategies to address high levels of community transmission of SARS-CoV-2 and related deaths, december 2020 MMWR Morb. Mortal. Wkly. Rep. 69 49 2020 1860 1867 10.15585/mmwr.mm6949e2 33301434
9 Aroso R.T. Schaberle A.F. Arnaut L.G. Pereira M.M. Photodynamic disinfection and its role in controlling infectious diseases Photochem. Photobiol. Sci. 20 11 2021 1497 1545 10.1007/s43630-021-00102-1 34705261
10 Lobo C.S. Rodrigues-Santos P. Pereira D. Núñez J. Trêpa J.C.D. Sousa D.L. Photodynamic disinfection of SARS-CoV-2 clinical samples using a methylene blue formulation Photochem. Photobiol. Sci. 21 6 2022 1101 1109 10.1007/s43630-022-00202-6 35304729
11 Kolenda J. Argemi J. Case series: nasal antimicrobial photodisinfection (APDT) as treatment protocol for asymptomatic and early stage COVID-19 patients Open Forum Infect Dis., Proceedings of IDWeek 2021 S354 10.1093/ofid/ofab466.703 Virtual, 29 September-3 October 2021; Paper 504
12 Romo C, Cross C, Anderson R, Loebel N. Antimicrobial photodynamic disinfection of SARS-COV-2 in a thin-film intranasal model. In Antimicrob Resist Infect Control, Abstracts from the 6th International Conference on Prevention and Infection Control (ICPIC 2021), Geneva, Switzerland, 14-17 September 2021; Paper 104..
13 Pires L. Wilson B.C. Bremner R. Lang A. Larouche J. McDonald R. Translational feasibility and efficacy of nasal photodynamic disinfection of SARS-CoV-2 Sci. Rep. 12 1 2022 14438 10.1038/s41598-022-18513-0
14 COVID-19 and Seasonal Influenza Provincial Respiratory Surveillance Report: Epidemiological Week 52 2020 (December 20 - December 26) https://www.gov.mb.ca/health/publichealth/surveillance/covid-19/2020/week_52/index.html
15 Fernandez-Montero A. Zuaznabar J. Pina-Sanchez M. Maestro S. Martin-Navarro L. Muñoz-Rodríguez N. Photodynamic nasal SARS-CoV-2 decolonization shortens infectivity and influences specific T-Cell responses Front. Cell. Infect. Microbiol. 13 2023 1110467 10.3389/fcimb.2023.1110467
16 Ramos A.K. Lowe A.E. Herstein J.J. Schwedhelm S. Dineen K.K. Lowe J.J. Invisible No more: the impact of COVID-19 on essential food production workers J. Agromed. 25 4 2020 378 382 10.1080/1059924X.2020.1814925
17 Herstein J.J. Degarege A. Stover D. Austin C. Schwedhelm M.M. Lawler J.V. Characteristics of SARS-CoV-2 transmission among meat processing workers in Nebraska, USA, and effectiveness of risk mitigation measures Emerg. Infect. Dis. 27 4 2021 1032 1038 10.3201/eid2704.204800 33591249
18 Fabreau G.E. Holdbrook L. Peters C.E. Ronksley P.E. Attaran A. McBrien K. Vaccines alone will not prevent COVID-19 outbreaks among migrant workers—the example of meat processing plants Clin. Microbiol. Infect. 28 6 2022 773 778 10.1016/j.cmi.2022.02.004 35189335
|
PMC010xxxxxx/PMC10229203.txt |
==== Front
Data Inf Manag
Data Inf Manag
Data and Information Management
2543-9251
The Authors. Published by Elsevier Ltd on behalf of School of Information Management Wuhan University.
S2543-9251(23)00017-7
10.1016/j.dim.2023.100043
100043
Article
Understanding the role of social media usage and health self-efficacy in the processing of COVID-19 rumors: A SOR perspective☆
Zhang Xiaofei a
Liu Yixuan a
Qin Ziru a
Ye Zilin a
Meng Fanbo b∗
a Business School, Nankai University, Tianjin, China
b School of Business, Jiangnan University, Wuxi, China
∗ Corresponding author. School of Business, Jiangnan University, Wuxi, 214122, China.
30 5 2023
6 2023
30 5 2023
7 2 100043100043
4 11 2022
4 4 2023
21 5 2023
© 2023 The Authors
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Apart from the direct health and behavioral influence of the COVID-19 pandemic itself, COVID-19 rumors as an infodemic enormously amplified public anxiety and cause serious outcomes. Although factors influencing such rumors propagation have been widely studied by previous studies, the role of spatial factors (e.g., proximity to the pandemic) on individuals’ response regarding COVID-19 rumors remain largely unexplored. Accordingly, this study, drawing on the stimulus-organism-response (SOR) framework, examined how proximity to the pandemic (stimulus) influences anxiety (organism), which in turn determines rumor beliefs and rumor outcomes (response). Further, the contingent role of social media usage and health self-efficacy were tested. The research model was tested using 1246 samples via an online survey during the COVID-19 pandemic in China. The results indicate that: (1)The proximity closer the public is to the pandemic, the higher their perceived anxiety; (2) Anxiety increases rumor beliefs, which is further positively associated rumor outcomes; (3) When the level of social media usage is high, the relationship between proximity to the pandemic and anxiety is strengthened; (4) When the level of health self-efficacy is high, the effect of anxiety on rumor beliefs is strengthened and the effect of rumor beliefs on rumor outcomes is also strengthened. This study provides a better understanding of the underlying mechanism of the propagation of COVID-19 rumors from a SOR perspective. Additionally, this paper is one of the first that proposes and empirically verifies the contingent role of social media usage and health self-efficacy on the SOR framework. The findings of study can assist the pandemic prevention department in to efficiently manage rumors with the aim of alleviating public anxiety and avoiding negative outcomes cause by rumors.
Keywords
Rumor belief
Rumor outcome
SOR framework
Social media usage
Health self-efficacy
==== Body
pmc1 Introduction
At the initial stage of the COVID-19 pandemic outbreak, public anxiety was significantly aroused owing to increasing numbers of infections and areas infected, thus leading to a series of unusual purchasing behaviors (e.g., hoarding toilet papers, food, and masks) and unscientific preventive behaviors (e.g., abuse of virus-related medicine and alcohol) among the public (Jiang et al., 2020; Laato et al., 2020). These irrational response not only distorted the functioning of local society but also induced physical and mental suffering in residents when they were facing the COVID-19 (Gao et al., 2020; Tasnim et al., 2020). Some research indicated that these negative outcomes such as tremendous economic losses and lengthened psychological distance were partially caused by rumors and misinformation circulating on social media (Tasnim et al., 2020; Zhu, 2021), which posed a serious threat to the COVID-19 response. This is because, under information heterogeneity and polarization, individuals who seeking COVID-19 information are difficult to judge the authenticity of the information in social media (Huang et al., 2021). Besides, individuals without pandemic-relevant knowledge may feel anxiety which in turn influence their decision-making (Wang et al., 2020). Therefore, to diminish the negative effects of rumors and to make an appropriate response to the pandemic, it is imperative to understand how COVID-19 related factors trigger individuals’ anxiety, thus forming trusting beliefs towards rumors and behaving irrationally.
Although previous studies on rumors have extended our understanding of the antecedents that cause individuals' trusting beliefs and irrational behaviors, some significant knowledge gaps remain. First, many scholars have indicated that factors regarding the content of a rumor (e.g., the source reliability of rumor, information without clear source, and argument volume) and rumor receivers' psychological factors (e.g., personal involvement, anxiety, and uncertainty) are critical in causing rumor belief and subsequent behaviors during a crisis (DiFonzo & Bordia, 2007; Pezzo & Beckstead, 2006; Wang et al., 2018). Besides, in the context of the COVID-19, individuals' perceptions of spatial distance to the center of the pandemic may also influence their rumor beliefs. This is because individuals who lived in the risk center area will perceive a higher level of risk and anxiety (Burns & Slovic, 2012), which makes them more prone to trust pandemic-related rumors. Specifically, individuals living closer to the center of the pandemic outbreak may respond to rumors differently from those living farther from the pandemic. However, we have little knowledge regarding how spatial factors (e.g., proximity to the pandemic) influence rumor receivers' rumor beliefs by influencing their anxiety during the COVID-19. Therefore, to better understand the influence mechanism of rumors, we propose the first research question: How do individuals’ perceptions of proximity to the pandemic influence their anxiety, which in turn determines their responses such as rumor beliefs and rumor outcomes?
Second, during this special period, social media to some extent enlarges the public's anxiety by providing an ideal platform for the anxious citizen to access pandemic-relevant rumors and propagating them (Luo et al., 2022; Yu et al., 2021), thus sharpening the public's rumor beliefs and causing rumor outcomes such as a series of irrational behaviors. Previous studies regarding rumors mainly have focused on the dark side of social media, for example, its ability to spread rumors and increase the public's anxiety during the COVID-19 outbreak (Gao et al., 2020; Nekovee et al., 2007). Further, platform characteristics and information factors of social media have also been found to induce rumor beliefs and behavioral outcomes (e.g., cyberchondria and verified information sharing) (Laato et al., 2020; Oh et al., 2018). However, there is also a bright side to social media usage in such situations because of its capacity to provide social support to alleviate the effects of stressful situations (Arora et al., 2007; Oh, Lauckner, et al., 2013). Therefore, another possibility exists—that social media can play a positive role by enabling individuals to evaluate rumors efficiently and avoid unnecessary behavioral outcomes during the COVID-19 pandemic. However, the role of social media usage in shaping individuals' anxiety and response towards the COVID-19 rumors has been significantly under-explored in the current literature. Accordingly, the second research question is as follows: How can social media usage shape the influence mechanism of rumors concerning the COVID-19?
Third, individual differences are closely associated with the ability to discern the validity of rumors circulating on social media (Pennycook et al., 2020). Specifically, individuals with greater cognitive reflection and science knowledge have a stronger ability to discern misinformation posted on social media regarding COVID-19. This finding is consistent with the social cognitive theory proposed by (Bandura, 2006), which indicated that one's self-efficacy refers to her/his competence in coping with tasks or stressors. During the outbreak of the COVID-19, individuals who have high health self-efficacy can effectively engage in preventive behaviors in the face of pandemic and have better mental health (Yıldırım & Güler, 2020). On the other hand, rather than promoting behaviors, self-efficacy not only leads to overconfidence but also lowers their performance (Moores & Chang, 2009), thus shaping rumor beliefs regarding the pandemic much easier. Therefore, we propose that the effect of health self-efficacy on discerning rumors regarding the COVID-19 pandemic may counterintuitively backfire. However, little is known about the role of health self-efficacy in influencing anxiety and discerning rumors. Accordingly, the third question of this study is as follows: Are the effects of individuals' anxiety on their rumor beliefs contingent upon their health self-efficacy?
To answer the research questions provided above, this paper draws upon the stimulus-organism-response (SOR) framework (Mehrabian & Russell, 1974) to conceptualize proximity to the COVID-19 pandemic, anxiety, and rumor beliefs, respectively. Then, a research model with seven hypotheses is tested using data from a large-scale survey conducted during the COVID-19 in China. This study makes multiple contributions to the literature concerning the SOR model and rumors. First, this study fills the research gap in the SOR literature by firstly investigating role of the spatial factor (proximity to the pandemic) on individuals’ anxiety which shaping their beliefs and subsequent behaviors towards the COVID-19 rumors. Second, our study extends the rumor literature by examining the contingent effects of social media usage and health self-efficacy on the processing of the COVID-19 rumors. In practice, the epidemic prevention department and relevant organizations in epidemic place can utilize the findings of this study to efficiently manage rumors, therefore alleviating public anxiety and avoiding practicing unusual behaviors caused by rumors.
2 Literature review
2.1 Rumors and social media
In the social psychology literature, a rumor refers to a story or a statement in general circulation that is lacking confirmation or certainty regarding facts (Allport & Postman, 1947). DiFonzo and Bordia (2007) provided a more detailed and comprehensive definition of rumors relating to three aspects: the context, the content, and the social function. They noted that rumors are unsubstantiated claims that are widespread in vague, dangerous, or potentially threatening situations. In particular, previous research on crisis informatics has shown that social media typically has the advantage of being able to improvise, spread, and disseminate information more easily, faster, and wider than does mainstream media (Guo & Zhang, 2020). With the rise of social media platforms such as Twitter, Facebook, Reddit, Weibo, and WeChat, the spread of rumors has transcended national boundaries and has increased significantly in terms of speed and audience (Yang et al., 2020). Although some scholars proposed some methods for recognition of rumor stances in online social media (Luo et al., 2020), social media usage during the pandemic period still increase individuals’ anxiety and cause severe mental health problems and cyberchondria (Gao et al., 2020; Laato et al., 2020; Oh & Lee, 2019). The rumor problem has presented a severe challenge to the effective use and scientific management of social media, and even has a significant impact on the real world (Liu et al., 2019).
However, regarding the relationship between social media and rumors, some contradictory findings are revealed (Sahni & Sharma, 2020). According to their review, although social media has brought some problems during COVID-19, it also promotes the well-being of individuals and public health when it is used wisely and prudently. For example, social media can hinder the spreading of rumors by providing and transmitting truthful facts from health experts (Sahni & Sharma, 2020). Antheunis et al. (2013) pointed out that patients’ use of social media contributes to their ability to obtain health-related social support and improve their health management. Table 1 presents the relevant research on rumors in relation to social media.Table 1 Relevant research on rumors in relation to social media.
Table 1Aspects Subjects Contents References
Diffusion mechanisms Rumor spread Perceived importance positively affects rumor propagation. Tanaka et al. (2012) and Oh and Lee (2019)
Anxiety, personal involvement, and ambiguity of information sources positively influence rumor propagation. Oh, Agrawal, and Rao (2013)
Public emotions such as anger, fear, sadness, and happiness positively influence rumor spread in the context of COVID-19. Dong et al. (2020)
Rumor trust Both personal involvement and rumor fear positively affect rumor trust, while the presence of counter-rumors negatively influences rumor trust. Chua and Banerjee (2018)
Rumor characteristics Rumor spreaders A lower ratio of following-to-follower is more likely to spark rumors. Bodaghi and Oliveira (2020)
Rumor content Various narrative frameworks increase rumors during the COVID-19 pandemic. Shahsavari et al. (2020)
Outcomes Social level Rumor spreading may cause panic buying during the COVID-19 pandemic. Arafat et al. (2020)
Health level Rumor leads to poor physical and mental health outcomes during the COVID-19 pandemic. Tasnim et al. (2020)
Governance Social media The true facts provided by health professionals through social media can prevent the spread of rumors. Sahni and Sharma (2020)
Social media feeds rumors. Yang et al. (2020)
Although some studies on rumors have been conducted during the COVID-19 pandemic, two main deficiencies are apparent in the current studies. First, scholars have not kept abreast of the rampant rumors circulating on social media in the context of the COVID-19. In general, the existing literature has not systematically examined the role of social media on antecedents and consequences of rumors regarding the pandemic. Second, previous studies mainly focused on the negative role of social media in the process of rumor propagation, with little attention paid to the positive role. The role of social media may reverse in discerning the truth against the rumors, which has a significant influence on controlling the COVID-19 pandemic. Based on the above discussion, this paper aims to elucidate the influencing mechanism of rumors during COVID-19 and introduce social media as a significant contingent factor. Next, we will introduce the SOR model that manifests the decision process regarding rumors and another contingent factor, health self-efficacy.
2.2 Stimulus-organism-response model
Mehrabian and Russell (1974) devised the famous SOR model in the field of behavioral psychology to explain and predict the effects of different environmental stimuli on human cognition, emotion, and behavior. Mehrabian's SOR model is a modification and optimization of Woodworth's stimulus-response (S–R) model (Woodworth & Marquis, 2014), but adding the “O” variable and thereby focusing on the internal consciousness of humans and other organisms. The model assumes that different external stimuli have different effects on the internal state of human beings, and then it determines the decision-making behavior of human beings based on human internal cognitive, and emotional factors (Mehrabian & Russell, 1974).
The SOR model not only has been widely studied in the research fields of marketing and e-commerce to understand consumer behavior (Ettis, 2017; Namkung & Jang, 2010) but also is applicable for exploring the antecedents and consequences of rumors propagation in the context of the pandemic. Böhm and Pfister (2005) found that in the process of spreading rumors, the wider the rumor spreads, the more likely will it present to people the illusion that the epidemic situation is not under control. In addition, they indicated that the more the illusion is, cause the higher risk in relation to the epidemic people may perceive, which will, in turn, cause confusion and uncertainty regarding whether the epidemic will affect them personally, leading to anxiety and other negative emotions.
During the COVID-19, the influence mechanism of rumor can also follow the rule of the SOR model. In this study, proximity to the COVID-19 pandemic regarded as an external stimulus can arouse individuals’ anxiety, thus leading to a series of psychological and physiological reactions (e.g., rumor beliefs and outcomes). Further, in the context of the COVID-19 pandemic, information spreading has two unique features: (1) social media has played a significant role in rumor spreading and witnessed many unusual behaviors (Tasnim et al., 2020); and (2) the virus affects human health that users with different levels of self-efficacy on their health may react to the COVID-19 rumors differently. Thus, to better fit the human decision during the COVID-19 pandemic and gain a better understanding of COVID-19 rumors, this study incorporates social media usage and health self-efficacy as contingent factors into the SOR model.
2.3 Health self-efficacy
Self-efficacy refers to “people's beliefs about their capabilities to produce designated levels of performance that exercise influence over events that affect their lives” (Bandura, 1977). Accordingly, in this paper, health self-efficacy is defined as people's beliefs about their capabilities to keep healthy during the COVID-19 pandemic (Lee et al., 2008). When faced with some fuzzy information or unverified news regarding the pandemic, individuals' discerning abilities significantly depend on their health self-efficacy. Therefore, health self-efficacy is a valid and legible instrument for assessing how people avoid becoming infected by, recognize the symptoms of, and home-manage COVID-19 (Hernández-Padilla et al., 2020).
Health self-efficacy has been investigated as a significant factor influencing individuals' behaviors and emotions during the pandemic. For example, individuals with a higher level of health self-efficacy are more willing to engage in preventive health measures such as complying with government-advised preventive measures and taking vaccines (Bults et al., 2011; Liao et al., 2011). Further, high self-efficacy can increase individuals' elaboration of news and knowledge about swine flu (Lo et al., 2013). On the other hand, health self-efficacy can be served as an efficient means of alleviating individuals’ negative emotions such as anxiety regarding the COVID-19 (Petzold et al., 2020). However, some scholars indicated that high self-efficacy may negatively influence subsequent performance (Vancouver et al. 2001, 2002). Similarly, individuals with a high level of self-efficacy became overconfident in themselves, thus leading to poor performance (Moores & Chang, 2009). These findings can be explained by the perceptual control theory (Powers, 2005), which illustrated that individuals with high self-efficacy may become overconfident when they perceived a relatively small difference between the perceived external state and an inner desired state. Thus, when facing rumors regarding COVID-19, individuals with high health self-efficacy may have a perception that they have a high discerning ability, thus readily placing trust in rumors without seeking further confirmation. Although current literature has investigated the positive role of health self-efficacy during the COIVD-19, its role in the influence mechanism of rumors has rarely been empirically investigated. By testing the moderating role of health self-efficacy, this study fills the gap in the literature regarding the role of health self-efficacy on rumor propagation during the period of the COVID-19 pandemic.
3 Research model and hypotheses
Drawing on the SOR model (Mehrabian & Russell, 1974), our research model (see Fig. 1 ) is to explain the influence mechanism of rumors regarding the COVID-19 pandemic. Specifically, this framework suggests that proximity to the pandemic influences rumor receivers’ anxiety, which intervenes in their rumor beliefs and rumors outcomes. Moreover, the moderating effects of social media usage, as well as health self-efficacy on the SOR framework, are tested.Fig. 1 Stimulus-organism-response model.
Fig. 1
According to the basic tenet of the SOR framework (Mehrabian & Russell, 1974), in the context of this study, the stimulus (S) is defined as an external stimulus that is the proximity to the pandemic, while the organism (O) refers to the emotional response of individuals which refers to individuals' anxiety, which is a common emotional response to disasters (Bergeron & Sanchez, 2005; Kim & Kim, 2017). Jacoby and Jacob (2002) indicated that an external stimulus will trigger organism to feel emotions. Previous scholars have pointed out that panic caused by risk events produces a ripple effect, which means that when individuals are located in the risk center area, their perception level of risk and anxiety is highest; the farther away people are from the center, the lower the risk (Slovic, 1987) and negative emotion they perceive (Burns & Slovic, 2012). For example, the closer people are to nuclear pollution, the more worried they are about the impact of nuclear leakage on their lives, resulting in stronger negative emotions and evasive behavior (Semenova et al., 2019; Suzuki et al., 2018). In the case of our study, rumor receivers who are living nearby the place of COVID-19 outbreak will perceive a higher level of anxiety. In other words, proximity to the pandemic as a stimulus can increase rumor receivers’ anxiety. This is because people perceive they have a high probability of being affected by COVID-19 if there is an outbreak nearby. Therefore, based on the above discussion, we propose the following hypothesis.H1 Proximity to a pandemic is positively associated with anxiety.
In the SOR framework, the response (R) refers to the organism's behavioral decision-making, which integrates external environmental stimulus and internal psychological attitude, including avoidance and approach behavior (Mehrabian & Russell, 1974). Accordingly, response in this study refers to rumor receivers' rumor beliefs aroused by anxiety during the COVID-19. Since anxiety is particularly likely to prevent people from forming an objective evaluation of circumstances (Hannabuss, 2008), an individual is more likely to trust a rumor which is congruent with her/his emotional state (Na et al., 2018). Besides, anxiety serves as an important link between disasters and rumor propagation (Allport & Postman, 1947; Askarizadeh et al., 2019; Oh, Agrawal, & Rao, 2013). Under the pandemic health crisis, pandemic diseases aroused a high level of anxiety (Bergeron & Sanchez, 2005), thus leading anxious individuals to form their beliefs that believing in pandemic-relevant rumors. Consequently, the higher level of anxiety a rumor receiver perceives, the more likely she or he is to trust rumors regarding the COVID-19. Based on the above arguments, we hypothesize that.H2 Anxiety positively affects rumor beliefs.
Based on the Theory of Reasoned Action (Fishbein & Ajzen, 1975), individuals' beliefs affect their attitudes toward a behavior, thus leading to the intention to perform this behavior. In our study, rumor beliefs and rumor outcomes are conceptualized as trusting beliefs regarding COVID-19 related rumors and irrational behavioral intentions respectively. When individuals perceive rumors highly trustworthy in an anxious situation, their usual behaviors will be significantly distorted. For example, previous studies indicated that individuals’ perceived relevance and importance of a rumor is positively associated with their trusting beliefs, thus verifying and sharing rumors (Chua & Banerjee, 2018; Oh & Lee, 2019). In other words, when people perceive that rumors regarding the pandemic are trustworthy, they are more likely to perpetuate the rumors and take relevant actions such as purchasing unnecessary preventative items. Therefore, we propose the following hypothesis.H3 Rumor beliefs positively affect rumor outcomes.
Social media as an emerging interactive communication channel empowers the public to actively search for information. Therefore, through using social media, the public is no longer information recipients of mass media but turns to be active information seekers who wish to receive information instantly (Stephens & Malone, 2009). During the COVID-19, compared to traditional information sources such as mass media, print media, and online official websites, information on social media is easier to arouse the public's information anxiety (Soroya et al., 2021). This is because news regarding the COVID-19 pandemic on social media leads people to understand that the pandemic is highly contagious and deadly (Lee et al., 2020). As such, the public may take more time using social media to search for relevant information when they realize they are getting closer to the center of a pandemic. Accordingly, the positive association between proximity to pandemics and anxiety can be strengthened by the frequent use of social media. Therefore, individuals who are close to the pandemic may perceive a higher level of anxiety when they use social media frequently to search for relevant information about the COVID-19. Based on the above arguments, we propose the following assumption.H4 Social media usage strengthens the relationship between proximity to the pandemic and anxiety.
When people are in a state of anxiety, they are more likely to trust rumors because anxiety prevents their judgment ability (Hannabuss, 2008). However, anxiety also increases people's demand for information during a pandemic. For example, public anxiety leads to an increase in information seeking during the outbreak of influenza H1N1(Tausczik et al., 2012). This is because anxious individuals prefer to collect more information to ease emotional pressure (Rosnow 1991). In this vein, anxious individuals will increase their information seeking on social media during the COVID-19. By doing so, they acquire much professional and social support from social media (Arora et al., 2007; Oh, Lauckner, et al., 2013), which may help ease the public's anxiety. Accordingly, under the condition of a high level of social media usage, individuals' anxious mood may be eased, and they will be more cautious about forming trusting beliefs regarding a rumor. In other words, the association between anxiety and rumor beliefs can be weakened by using social media for social support. Based on the above arguments, we propose the following assumption.H5 Social media usage weakens the relationship between anxiety and rumor beliefs.
Health self-efficacy refers to individuals’ beliefs that they are capable to manage their health (Lee et al., 2008). Although some scholars indicated that health self-efficacy plays a significant role in engaging in preventive actions regarding a pandemic (Bults et al., 2011; Liao et al., 2011) and reducing negative emotions (Petzold et al., 2020), individuals with a high level of health self-efficacy may become over-confident in their capabilities of managing their health during the COVID-19 (Moores & Chang, 2009). This is because that people with strong self-efficacy tend to believe in themselves sufficiently to extricate themselves from a threatening situation (Bandura, 1977). In such a case, when individuals have a higher level of health self-efficacy, anxious ones are likely to prevent themselves from forming an objective judgment, which has led to a higher level of rumor belief during the COVID-19 pandemic. In other words, anxious individuals with a high level of health self-efficacy may be more firmly believe in their subjective judgments regarding rumors of COVID-19. Accordingly, the positive effect of anxiety on rumor beliefs will be strengthened by health self-efficacy. On this basis, we hypothesize that.H6 Health self-efficacy strengthens the positive relationship between anxiety and rumor beliefs.
When faced with a threat, people with high-level self-efficacy tend to engage in higher task intention and less procrastination, which means they are more likely to accept challenges, execute available strategies, and persist regardless of setbacks and risks (Haycock et al., 1998; Wieber et al., 2010). In the realm of healthcare, those with a high level of health self-efficacy are more prone to engage in health-promoting and health-impairing behaviors (Bandura, 1986). During the COVID-19, health self-efficacy has been found to have positive effects on a series of preventative behaviors such as self-isolation (Farooq et al., 2020). When faced with rumors regarding the COVID-19, people with stronger health self-efficacy have greater intention to perpetuate rumors they accept and are more confident in their capabilities to complete the task of protecting themselves, for example, executing an irrational purchase. Under the condition of a high level of health self-efficacy, individuals may have a stronger willingness to turn their rumor beliefs into outcomes. In this vein, the association between rumor beliefs and rumor outcomes is strengthened by health self-efficacy. Therefore, we put forward the following hypothesis.H7 Health self-efficacy strengthens the positive relationship between rumor beliefs and rumor outcomes.
4 Methodology
4.1 Measurements
An online cross-sectional survey was conducted to test the proposed research model and hypotheses. We evaluated all the measures based on a five-point Likert scale ranging from 1 to 5, representing “strongly disagree” to “strongly agree.” Some of the constructs were adapted from a pre-designed research instrument from prior studies. The measures for the degree of anxiety were adapted from Elhai et al. (2020), while the measures for social media usage and health self-efficacy were adapted from Oh, Lauckner, et al. (2013). In addition, the measures for rumor outcomes were adapted from Laato et al. (2020). Furthermore, some of the measures were modified to fit our research context. Proximity to the pandemic was measured by the number of newly confirmed COVID-19 cases in areas where samples lived in. The measures of rumor beliefs were self-defined for our research purpose. All the measures are shown in Appendix.
4.2 Data collection
To collect the data for this study, we chose February 2020 to distribute this survey online for three reasons: (1) our research topic is about the mechanism of proximity to the pandemic, anxiety, beliefs, and outcomes of COVID-19 rumors on social media; (2) in this February, China was suffering heavy levels of morbidity and mortality caused by COVID-19, whereas pandemic issues in other global regions remained relatively low (Qiu et al., 2020); and (3) under the circumstance of the pandemic, it was safer and more convenient to survey our participants online. Therefore, our samples for the online survey were representative to a certain degree. Further, to guarantee the quality of the questionnaire, we firstly translated the English version of the questionnaire into Chinese and then asked three postgraduate students to check the translation accuracy. A pretest with 50 participants has been conducted before formally distributed to make sure the readability and easy-to-follow of each question. Our questionnaire received 4628 responses in total. In the data cleaning process, we removed incomplete questionnaires and invalid questionnaires in which the answers were the same and obviously contradictory. 1246 responses were eventually valid after the quality control checks, with a response rate of 27%. The demographic information is provided in Table 2 .Table 2 Samples’ demographics.
Table 2Category Sub-category Response Percentage
Gender male 618 49.56%
female 629 50.44%
Age ＜18 61 4.89%
18–25 502 40.26%
26–30 294 23.58%
31–40 239 19.17%
41–50 97 7.78%
51–60 47 3.77%
＞60 7 0.56%
Education Primary school 23 1.84%
Middle school 135 10.83%
College and technical secondary school 297 23.82%
Bachelor's degree 578 46.35%
Master's degree 160 12.83%
Doctor's degree 54 4.33%
Area Urban areas 751 60.22%
Rural area 349 27.99%
City suburb 124 9.94%
Other 23 1.84%
Occupation Full-time students 425 34.08%
Production personnel 66 5.29%
Salesperson 94 7.54%
Marketing personnel/public relations practitioner 51 4.09%
Customer service 37 2.97%
Administrative/back office 79 6.34%
Human resources 27 2.17%
Financial personnel/auditor 32 2.57%
Civilian/clerical 41 3.29%
Technical/research and development personnel 68 5.45%
Management personnel 37 2.97%
Teacher 66 5.29%
Consultant 7 0.56%
Professionals (e.g., accountants, lawyers, architects, medical staff, journalists) 47 3.77%
Other 170 13.63%
4.3 Data analyses
Structural equation modeling (SEM) consists of covariance-based SEM and variance-based SEM (e.g., partial least squares-SEM) (Cepeda-Carrion et al., 2019). The partial least squares SEM (PLS-SEM) technique is used to analyze our research model due to the following two advantages. First, compared to covariance-based SEM, PLS-SEM is more suitable for an exploratory study for theory development and the structural model consists of one or more formative constructs (Gefen et al., 2011; Hair et al., 2019; Khan et al., 2019). Second, since multiple regression and ANOVA (analysis of variance) may underestimate the interaction effect, PLS-SEM is a better choice to test the interaction effect in the model (Chin et al., 2003). Below, both the measurement model and the structural model are assessed to test the appropriateness of this research model and related hypotheses.
4.3.1 Measurement model
The reliability and validity of the measurement model were tested to indicate the goodness of fit. The Cronbach's α value of each variable was above 0.70, and all other values were greater than 0.80, except health self-efficacy (as shown in Table 3 ). Composite reliabilities were above 0.85 (as shown in Table 5 ), indicating that the research data had high reliability (Chin et al., 2003; Urbach & Ahlemann, 2010). Further, the loadings of each item were above 0.65 (as shown in Table 4; values in bold) and the average variances extracted (AVEs) were all greater than 0.60 (as shown in Table 5), which indicates that convergent validity was good (Chin et al., 2003). For discriminant validity, the items of each construct loaded greater on themselves than on other constructs (as shown in Table 4) and the square root of the AVE of each variable (as shown in Table 5) was greater than the correlations with the other variable, thus indicating acceptable discriminant validity (Chin et al., 2003). The results are shown in Table 3, Table 4, Table 5 Table 3 Reliability analysis.
Table 3Category Construct Items Cronbach's α
Independent variable Proximity to pandemic 1 1.00
Anxiety 4 0.93
Rumor belief 4 0.80
Moderator variable Social media usage 3 0.85
Health self-efficacy 2 0.72
Dependent variable Rumor outcomes 3 0.89
Table 4 Cross loadings.
Table 4 PTP ANX RB RO SMU HSE
PTP 1.0000 0.1646 0.1846 0.1412 −0.0820 −0.1002
ANX1 0.1540 0.9245 0.2291 0.4816 0.1072 −0.0461
ANX2 0.1553 0.9411 0.2444 0.4917 0.0941 −0.0487
ANX3 0.1514 0.9378 0.2520 0.5080 0.0988 −0.0813
ANX4 0.1371 0.8274 0.2563 0.4558 0.1100 −0.0747
RB1 0.1154 0.1468 0.6571 0.2633 0.0145 −0.0538
RB2 0.1703 0.2617 0.8661 0.3915 −0.0567 −0.1656
RB3 0.1532 0.1862 0.7638 0.2658 −0.0272 −0.0992
RB4 0.1422 0.2398 0.8625 0.3493 −0.0324 −0.1403
RO1 0.1412 0.4730 0.3989 0.9300 −0.0163 −0.1316
RO2 0.1301 0.4678 0.3692 0.9286 −0.0174 −0.1287
RO3 0.1117 0.5146 0.3415 0.8610 0.0297 −0.1277
SMU1 −0.1000 0.0690 −0.0605 −0.0652 0.8598 0.3116
SMU2 −0.1062 0.0429 −0.0710 −0.0549 0.8448 0.2970
SMU3 −0.0381 0.1434 0.0027 0.0620 0.8997 0.2378
HSE1 −0.1002 −0.0671 −0.0416 −0.0593 0.2077 0.7232
HSE2 −0.0893 −0.0624 −0.1707 −0.1528 0.3102 0.9769
Table 5 Discriminant validity.
Table 5 AVE CR PTP ANX RB RO SMU HSE
PTP 1.000 1.000 1.000
ANX 0.826 0.950 0.165 0.682
RB 0.627 0.869 0.185 0.271 0.394
RO 0.823 0.933 0.141 0.534 0.408 0.677
SMU 0.754 0.902 −0.082 0.113 −0.037 −0.002 0.569
HSE 0.739 0.847 −0.100 −0.069 −0.153 −0.143 0.312 0.546
AVE: average variance extracted; CR: composite reliability; PTP: proximity to the pandemic; ANX: anxiety; RB: rumor beliefs; RO: rumor outcomes; SMU: social media usage; HSE: health self-efficacy.
The diagonal values are the square roots of AVEs.
4.3.2 Structural model
Based on the hypotheses, the structural model was tested in two stages. During the first stage, we examined the direct effects of the model. The relationship between proximity to the pandemic and anxiety was found to be positive and significant (β = 0.043; p < 0.001). The results further reveal that anxiety had a positive effect on rumor belief (β = 0.261; p < 0.001). And the effect of rumor beliefs on rumor outcomes was positive and significant (β = 0.393; p < 0.001). Thus, hypotheses H1–H3 are all supported.
In the second stage, we tested the moderating role of social media usage and health self-efficacy. The results show that social media usage had a positive effect on the relationship between proximity to the pandemic and anxiety (β = 0.092, p < 0.01), while having no significant effect on the relationship between anxiety and rumor belief (β = −0.045, p > 0.10). Meanwhile, health self-efficacy had a positive effect both on the relationship between anxiety and rumor belief (β = 0.085, p < 0.01) and the relationship between rumor belief and rumor outcome (β = 0.112, p < 0.01). Thus, H4, H6, and H7 are supported, but not H5. The results of model estimations are presented in Fig. 2 .Fig. 2 Smart PLS results.
Fig. 2
5 Discussion and implications
5.1 Key findings
This study investigated the influence mechanism of COVID-19 rumors drawing upon the SOR framework and the contingent roles of social media usage and health self-efficacy in this mechanism. Several main findings emerge from this study. First, proximity to the pandemic as a stimulus is positively associated with anxiety as an organism, which has a positive effect on rumor beliefs as a response. This finding is consistent with the basic tenet of the SOR framework (Mehrabian & Russell, 1974). During COVID-19, the closer rumor receivers are to the pandemic, the more likely they are to feel anxious and the higher possibility they will choose to trust rumors regarding COVID-19.
Second, social media usage strengthens the relationship between proximity to the pandemic and anxiety but has no significant effect on the relationship between anxiety and rumor beliefs. This implies that in the initial stage of the COVID-19 pandemic, under a high level of social media usage, individuals living near the pandemic will become more anxious. This finding is supported by previous research (Farooq et al., 2020) which indicates that more frequent social media exposure can lead to cyberchondria. However, regarding the unsupported hypothesis 5, one explanation is grounded in the fact that frequent exposure to social media during a global health crisis strengthened the effects of information overload and information anxiety, which in turn leads to information avoidance (Soroya et al., 2021; Yavetz et al., 2022). In such a situation, anxious individuals will not continue using social media for seeking information before they are forming rumor beliefs. Therefore, social media usage has a non-significant effect on the association between anxiety and rumor beliefs.
Third, health self-efficacy strengthens (1) the positive relationship between anxiety and rumor beliefs, and (2) the positive relationship between rumor beliefs and rumor outcomes. This finding implies that under the condition of the high level of health self-efficacy, anxious rumor receivers are more likely to form their rumor beliefs, thus putting their beliefs into practice. As shown by Moores and Chang (2009), this result confirmed that self-efficacy is positively associated with overconfidence in individuals’ judge ability, thus indicating the negative effect of health self-efficacy on performance.
5.2 Theoretical implications
This research yields several contributions. First, this paper not only expands the scope of use of the SOR model but also promotes the systematization of rumor research by linking several scattered variables (e.g., social media usage and health self-efficacy) into an integrated framework. In the previous literature, the SOR framework has been mostly applied in the field of offline and online consumer behavior (Gatautis et al., 2016), and there was a gap in the field of rumor exploration (Pal et al., 2019). The pandemic environment (proximity to the pandemic) is an external stimulus that guides the public into a specific state (anxiety), who take related action (rumor beliefs and rumor outcomes), which is consistent with the mechanism of SOR relations applied in marketing (Eroglu et al., 2001).
Second, this paper, to the best of our knowledge, is one of the first that proposes and empirically verifies both the bright and dark side of social media on the influence mechanism of rumors during COVID-19. Previous studies have mainly focused on the dark side of social media in spreading rumors, increasing public anxiety, causing rumor beliefs and rumor outcomes during the COVID-19 pandemic (Gao et al., 2020; Laato et al., 2020), however, few studies have systematically examined the bright role of social media on the influence mechanism of rumors regarding COVID-19. This study addresses this research gap by investigating the contingent role of social media on (1) the association between stimulus and organism, and (2) the association between organism and response. Although some scholars indicated that social media can provide useful social support to people to mitigate the negative effects of stressful situations on health (Arora et al., 2007; Oh, Lauckner, et al., 2013), the results show that the negative role of social media still dominates arousal of anxiety and beliefs in pandemic rumors during the COVID-19.
Third, this study highlights the significant role of health self-efficacy in the influence mechanism of rumors. Previous studies have pointed out that health self-efficacy is positively related to individuals’ health preventive measures (Bults et al., 2011; Liao et al., 2011; Tang & Wong, 2003) and mental health during COVID-19 (Yıldırım & Güler, 2020). However, the negative role of health self-efficacy has been significantly ignored. These results indicate that a high level of health self-efficacy may make individuals being blindly confident in their abilities to manage their health, in turn easily forming rumor beliefs and rumor outcomes. This study enriches the rumor literature by highlighting the negative role of health self-efficacy in the spreading of rumors regarding a pandemic.
5.3 Practical implications
This empirical study provides some significant implications in practice. First, the results show that proximity to the pandemic will increase rumor receivers’ anxiety, which in turn shapes their rumor beliefs.
The pandemic prevention department should pay more attention to the emotional states of individuals living nearby the pandemic center and take measures to alleviate their anxiety and to avoid forming rumor beliefs.
Second, our results indicate that social media usage positively moderates the positive relationship between proximity to the pandemic and anxiety. The Cyber Security Institute needs to pay more attention to the dark side of social media on rumor propagation and publish some regulations to effectively manage information and news on social media to relieve public anxiety during the pandemic.
Third, our study shows that health self-efficacy strengthens the link between anxiety, rumor beliefs, and outcomes. Health self-efficacy, to some extent, can create negative effects on the influence mechanism of rumors regarding a pandemic. Therefore, forming correct health self-efficacy under scientific and medical guidance is also vital for rumor control. The pandemic prevention department should take advantage of both mass media and social media to educate the public to form the right health self-efficacy, thus taking accurate measures to cope with the COVID-19.
5.4 Limitations and future directions
This study has some limitations. First, this research was conducted in China. Therefore, the generalizability of the results in other countries should be questioned. Second, we have only collected data through an online survey, lacking samples of infrequent Internet users. In future research, samples are expected to be enriched by collecting more offline data. According to the work of Zhao et al. (2021), future research should collect the data set from real social media to build a rumors detection model. Finally, rumor propagation is the result of the combined action of event importance and information fuzziness (Allport & Postman, 1947; Rosnow & Ralph, 1991). Other factors such as interpersonal relationships and social influence could also possibly affect rumor belief (Brock et al., 2012; Kelman, 1958; Neal & Chartrand, 2011), which can serve as directions for future research.
6 Conclusion
The purpose of this study is to understand the influence mechanism of rumors regarding the COVID-19 to alleviate public anxiety and reduce negative outcomes caused by rumors. Therefore, this study draws upon the SOR framework to explore how proximity to the pandemic influences anxiety, which in turn affects rumor beliefs and rumor outcomes. Further, the contingent role of social media usage and health self-efficacy on the SOR framework was examined. Based on an online survey, the research model and seven hypotheses were tested. The results indicated that proximity to the pandemic is positively associated with anxiety, which leads to rumor beliefs and rumor outcomes. In addition, social media usage strengthened the effects of proximity to the pandemic on anxiety. Health self-efficacy positively moderated the relationships (1) between anxiety and rumor beliefs, and (2) between rumor beliefs and rumor outcomes. This study enriches the literature on the SOR framework and rumor by highlighting the influence mechanism of rumors regarding the COVID-19. Practically, the findings of this study will assist the pandemic prevention department to efficiently manage rumor propagation to relieve public anxiety and alleviate negative outcomes cause by rumors during the COVID-19.
Conflict of interest
We have no conflicts of interest to declare.
Xiaofei Zhang is an Associate Professor in Business School, Nankai University, Tianjin, China. He earned his PhDs from The Hong Kong Polytechnic University and Harbin Institute of Technology. His research interest is in the area of Healthcare IT, Human-Computer Interaction, and Affective Response. His research has appeared in Tourism Management, European Journal of Information Systems, Information & Management, International Journal of Production Economics, and others
Yixuan Liu is a Ph.D student in Faculty of Business, the Hong Kong Polytechnic University. Her research interest is in the area of Human-Computer Interaction and Online Health Communities
Ziru Qin is an under-graduate student in Business School, Nankai University, Tianjin, China. His research interest is in the area of eHealth and information behaviors
Zilin Ye is an under-graduate student in Business School, Nankai University, Tianjin, China. Her research interest is in the area of Online Health Communities and information behaviors
Fanbo Meng is an Associate Professor in the School of Business, Jiangnan University, China. He earned his Ph.D. degree at the joint Ph.D. program of the Hong Kong Polytechnic University and Harbin Institute of Technology. His research interest is in the area of eHealth and Consumer Behavior. His research has appeared in International Journal of Production Economics, Information Procession & Management, Electronic Commerce with Research and Applications, and others
Appendix Measurement Items
Rumor Outcomes: (Laato et al., 2020)
RO1. Purchase hygiene products such as face masks and/or hand wash or sanitizers to protect me because of rumors regarding COVID-19.
RO2. Stock up food and/or other necessities because of rumors regarding COVID-19.
RO3. My life became chaotic because of rumors regarding COVID-19.
Rumor Beliefs: (Self-developed)
Have you ever believed the following rumors?RB1. Chinese patent medicine Shuanghuanglian oral solution was found to inhibit COVID-19 by Shanghai Institute of Materia Medica and Wuhan Institute of Virus.
RB2. Drinking strong Chinese wine can resist COVID-19.
RB3. Research by Zhong Nanshan's team shows that smokers have a lower rate of a viral infection than non-smokers.
RB4. Drinking isatidis root and smoked vinegar can prevent COVID-19.
Proximity to the Pandemic:(Self-developed)
PTP. Are there any confirmed cases and suspected cases of COVID-19 within 3 km of your home?
Anxiety: (Elhai et al., 2020)
ANX1. How often have you felt restless and find it difficult to stay calm because of the COVID-19 outbreak?
ANX2. How often have you got insomnia and felt upset because of the COVID-19 outbreak?
ANX3. How often have you felt worried and nervous all day because of the COVID-19 outbreak?
ANX4. How often have you suspected you were infected with the virus because of the coronavirus outbreak?
Social Media Usage:(Oh, Lauckner, et al., 2013)
SMU1. I usually use online social media to acquire COVID-19 related information in the outbreaks.
SMU2. I often use online social media to get information from people who have knowledge about the COVID-19.
SMU3. If I have a problem with the COVID-19, I usually seek advice from online social media.
Heath Self-efficacy: (Oh, Lauckner, et al., 2013)
HSE1: I am confident that I keep my health in the COVID-19 outbreaks.
HSE2: I have set clear goals for not being infected by the COVID-19.
Acknowledgement
This study was partially funded by the National Natural Science of China (71901127 and 72001094) and Young Elite Scientists Sponsorship Program by Tianjin (TJSQNTJ-2020-12).
☆ This study was partially funded by the National Natural Science of China (72001094, 72271131, and 20&ZD142) and Fundamental Research Funds for the Central Universities (63192406).
==== Refs
References
Allport G.W. Postman L. The psychology of rumor 1947 Henry Holt and Company New York
Antheunis M.L. Tates K. Nieboer T.E. Patients' and health professionals' use of social media in health care: Motives, barriers and expectations Patient Education and Counseling 92 3 2013 426 431 23899831
Arafat S.Y. Kar S.K. Menon V. Kaliamoorthy C. Mukherjee S. Alradie-Mohamed A. Sharma P. Marthoenis M. Kabir R. Panic buying: An insight from the content analysis of media reports during COVID-19 pandemic Neurology, Psychiatry and Brain Research 37 2020 100 103 32834528
Arora N.K. Finney Rutten L.J. Gustafson D.H. Moser R. Hawkins R.P. Perceived helpfulness and impact of social support provided by family, friends, and health care providers to women newly diagnosed with breast cancer Psycho-Oncology 16 5 2007 474 486 16986172
Askarizadeh M. Ladani B.T. Manshaei M.H. An evolutionary game model for analysis of rumor propagation and control in social networks Physica A: Statistical Mechanics and its Applications 523 2019 21 39
Bandura A. Self-Efficacy: Toward a unifying theory of behavioral change Advances in Behaviour Research and Therapy 1 4 1977 139 161
Bandura A. Social foundations of thought and action: A social cognitive theory 1986 Prentice-Hall, Inc Englewood Cliffs, NJ, US
Bandura A. Guide for constructing self-efficacy scales Self-efficacy beliefs of adolescents 5 1 2006 307 337
Bergeron S.L. Sanchez A.L. Media effects on students during sars outbreak Emerging Infectious Diseases 11 5 2005 732 15890131
Bodaghi A. Oliveira J. The characteristics of rumor spreaders on Twitter: A quantitative analysis on real data Computer Communications 160 2020 674 687
Böhm G. Pfister H.-R. Consequences, morality, and time in environmental risk evaluation Journal of Risk Research 8 6 2005 461 479
Brock B. Peter K. Matthew H. Joonha P. Koval Feeling bad about being sad: The role of social expectancies in amplifying negative mood Emotion 12 1 2012 69 80 21787076
Bults M. Beaujean D.J. Zwart O.D. Kok G. Empelen P.V. Steenbergen J.E.V. Richardus J.H. Voeten H.A. Perceived risk, anxiety, and behavioural responses of the general public during the early phase of the influenza a (H1N1) pandemic in The Netherlands: Results of three consecutive online surveys BMC Public Health 11 1 2011 1 13 21199570
Burns W.J. Slovic P. Risk perception and behaviors: Anticipating and responding to crises Risk Analysis 32 4 2012 579 582 22500649
Cepeda-Carrion G. Cegarra-Navarro J.-G. Cillo V. Tips to use partial least squares structural equation modelling (PLS-SEM) in knowledge management Journal of Knowledge Management 23 1 2019 67 89
Chin W.W. Marcolin B.L. Newsted P.R. A partial least squares latent variable modeling approach for measuring interaction effects: Results from a Monte Carlo simulation study and an electronic-mail emotion/adoption study Information Systems Research 14 2 2003 189 217
Chua A.Y. Banerjee S. Intentions to trust and share online health rumors: An experiment with medical professionals Computers in Human Behavior 87 2018 1 9
DiFonzo N. Bordia P. Rumor psychology: Social and organizational approaches 2007 American Psychological Association
Dong W. Tao J. Xia X. Ye L. Xu H. Jiang P. Liu Y. Public emotions and rumors spread during the COVID-19 epidemic in China: Web-based correlation study Journal of Medical Internet Research 22 11 2020 e21933 33112757
Elhai J.D. Yang H. McKay D. Asmundson G.J. COVID-19 anxiety symptoms associated with problematic smartphone use severity in Chinese adults Journal of Affective Disorders 274 2020 576 582 32663990
Eroglu S.A. Machleit K.A. Davis L.M. Atmospheric qualities of online retailing: A conceptual model and implications Journal of Business Research 54 2 2001 177 184
Ettis S.A. Examining the relationships between online store atmospheric color, flow experience and consumer behavior Journal of Retailing and Consumer Services 37 2 2017 43 55
Farooq A. Laato S. Islam A.K.M.N. Impact of online information on self-isolation intention during the covid-19 pandemic: Cross-sectional study Journal of Medical Internet Research 22 5 2020 e19128
Fishbein M. Ajzen I. Belief, attitude, intention and behavior: An introduction to theory and research Philosophy and Rhetoric 10 2 1975 130 132
Gao J. Zheng P. Jia Y. Chen H. Dai J. Mental health problems and social media exposure during COVID-19 outbreak Social Science Research Network Electronic Journal 15 4 2020 231924
Gatautis R. Vitkauskaite E. Gadeikiene A. Piligrimiene Z. Gamification as a mean of driving online consumer behaviour: Sor model perspective Engineering Economics 27 1 2016 90 97
Gefen D. Straub D.W. Rigdon E.E. An update and extension to sem guidelines for admnistrative and social science research MIS Quarterly 35 2 2011 (iii-xiv)
Guo L. Zhang Y. Information flow within and across online media platforms: an agenda-setting analysis of rumor diffusion on news websites, Weibo, and WeChat in China Journalism Studies 21 15 2020 2176 2195
Hair J.F. Risher J.J. Sarstedt M. Ringle C.M. When to use and how to report the results of pls-sem European Business Review 31 1 2019 2 24
Hannabuss S. Information and emotion: The emergent affective paradigm in information behaviour research and Theory20087edited by diane nahl and dania bilal. Information and emotion: The emergent affective paradigm in information behaviour research and theory Library Review 57 8 2008 644 646
Haycock L.A. McCarthy P. Skay C.L. Procrastination in college students: The role of self-efficacy and anxiety Journal of Counseling and Development 76 3 1998 317 324
Hernández-Padilla J.M. Granero-Molina J. Ruiz-Fernández M.D. Dobarrio-Sanz I. López-Rodríguez M.M. Fernández-Medina I.M. Correa-Casado M. Fernández-Sola C. Design and psychometric analysis of the covid-19 prevention, recognition and home-management self-efficacy scale International Journal of Environmental Research and Public Health 17 13 2020 4653 32605250
Huang K. Hao X. Guo M. Deng J. Li L. A study of Chinese college students' COVID-19-related information needs and seeking behavior Aslib Journal of Information Management 73 5 2021 679 698
Jacoby Jacob Stimulus-organism-response reconsidered: An evolutionary step in modeling (consumer) behavior Journal of Consumer Psychology 12 1 2002 51 57
Jiang Z. Tieqiao iang H. Zhong X. Shichao H. Chen Expert consensus on the prevention and treatment of substance use and addictive behaviour-related disorders during the COVID-19 pandemic General Psychiatry 33 4 2020 224 229
Kelman H.C. Compliance, identification, and internalization: Three processes of attitude change Journal of Conflict Resolution 2 1 1958 51 60
Khan G.F. Sarstedt M. Shiau W.-L. Hair J.F. Ringle C.M. Fritze M.P. Methodological research on partial least squares structural equation modeling (PLS-SEM) Internet Research 29 3 2019 407 429
Kim S. Kim S. Impact of the fukushima nuclear accident on belief in rumors: The role of risk perception and communication Sustainability 9 12 2017 2188
Laato S. Islam A.N. Farooq A. Dhir A. Unusual purchasing behavior during the early stages of the COVID-19 pandemic: The stimulus-organism-response approach Journal of Retailing and Consumer Services 57 2020 102224
Lee S.Y. Hwang H. Hawkins R. Pingree S. Interplay of negative emotion and health self-efficacy on the use of health information and its outcomes Communication Research 35 3 2008 358 381
Lee J.J. Kang K.A. Wang M.P. Sheng Shin S. Associations between covid-19 misinformation exposure and belief with covid-19 knowledge and preventive behaviors: Cross-sectional online study Journal of Medical Internet Research 20 11 2020 22205
Liao Q. Cowling B.J. Lam W.W.T. Fielding R. Factors affecting intention to receive and self-reported receipt of 2009 pandemic (H1N1) vaccine in Hong Kong: A longitudinal study PLoS One 6 3 2011 17713
Liu Y. Jin X. Shen H. Towards early identification of online rumors based on long short-term memory networks Information Processing & Management 56 4 2019 1457 1467
Lo V.H. Wei R. Su H. Self-efficacy, information-processing strategies, and acquisition of health knowledge Asian Journal of Communication 23 1 2013 54 67
Luo Y. Ma J. Yeo C.K. Exploiting user network topology and comment semantic for accurate rumour stance recognition on social media Journal of Information Science 48 5 2020 660 675
Luo Y. Ma J. Yeo C.K. Identification of rumour stances by considering network topology and social media comments Journal of Information Science 48 1 2022 118 130
Mehrabian A. Russell J.A. An approach to environmental psychology 1974 MIT Press Cambridge
Moores T.T. Chang J.C.-J. Self-efficacy, overconfidence, and the negative effect on subsequent performance: A field study Information & Management 46 2 2009 69 76
Na K. Garrett R.K. Slater M.D. Rumor acceptance during public health crises: Testing the emotional congruence hypothesis Journal of Health Communication 23 8 2018 791 799 30325720
Namkung Y. Jang S.C.S. Effects of perceived service fairness on emotions, and behavioral intentions in restaurants European Journal of Marketing 44 9–10 2010 1233 1259
Neal D.T. Chartrand T.L. Embodied emotion perception amplifying and dampening facial feedback modulates emotion perception accuracy Social Psychological and Personality Science 2 6 2011 673 678
Nekovee M. Moreno Y. Bianconi G. Marsili M. Theory of rumour spreading in complex social networks Physica A: Statistical Mechanics and Its Applications 374 1 2007 457 470
Oh O. Agrawal M. Rao H.R. Community intelligence and social media services: A rumor theoretic analysis of tweets during social crises MIS Quarterly 37 2 2013 407 426
Oh O. Gupta P. Agrawal M. Rao H.R. ICT mediated rumor beliefs and resulting user actions during a community crisis Vol. 35 2018 Government Information Quarterly 243 258 2
Oh H.J. Lauckner C. Boehmer J. Fewins-Bliss R. Li K. Facebooking for health: An examination into the solicitation and effects of health-related social support on social networking sites Computers in Human Behavior 29 5 2013 2072 2080
Oh H.J. Lee H. When do people verify and share health rumors on social media? The effects of message importance, health anxiety, and health literacy Journal of Health Communication 24 11 2019 837 847 31609678
Pal A. Chua A.Y.K. Goh H.L. How do uers respond to online rumor rebuttals? Computers in Human Behavior 106 5 2019 106243
Pennycook G. Mcphetres J. Zhang Y. Lu J.G. Rand D.G. Fighting COVID-19 misinformation on social media: Experimental evidence for a scalable accuracy-nudge intervention Psychological Science 31 7 2020 770 780 32603243
Petzold M.B. Bendau A. Plag J. Pyrkosch L. Strhle A. Risk, resilience, psychological distress, and anxiety at the beginning of the covid pandemic in Germany Brain Behavior 10 9 2020 e01745
Pezzo M.V. Beckstead J.W. A multilevel analysis of rumor transmission: Effects of anxiety and belief in two field experiments Basic and Applied Social Psychology 28 1 2006 91 100
Powers W.T. Behavior: The control of perception 2005
Qiu J. Shen B. Zhao M. Wang Z. Xu Y. A nationwide survey of psychological distress among Chinese people in the covid-19 epidemic: Implications and policy recommendations General Psychiatry 33 2 2020 100213
Rosnow Ralph L. Inside rumor: A personal journey American Psychologist 46 5 1991 484 496
Sahni H. Sharma H. Role of social media during the COVID-19 pandemic: Beneficial, destructive, or reconstructive? International Journal of Academic Medicine 6 2 2020 70 75
Semenova Y. Pivina L. Manatova A. Bjørklund G. Glushkova N. Belikhina T. Dauletyarova M. Zhunussova T. Mental distress in the rural kazakhstani population exposed and non-exposed to radiation from the semipalatinsk nuclear test site Journal of Environmental Radioactivity 33 2 2019 39 47
Shahsavari S. Holur P. Wang T. Tangherlini T.R. Roychowdhury V. Conspiracy in the time of corona: automatic detection of emerging COVID-19 conspiracy theories in social media and the news Journal of computational social science 3 2 2020 279 317 pp. 33134595
Slovic P. Perception of risk Science 236 4799 1987 280 285 3563507
Soroya S.H. Farooq A. Mahmood K. Isoaho J. Zara S.-e. From information seeking to information avoidance: Understanding the health information behavior during a global health crisis Information Processing & Management 58 2 2021 102440
Stephens K.K. Malone P.C. If the organizations won't give us information…: The use of multiple new media for crisis technical translation and dialogue Journal of Public Relations Research 21 2 2009 229 239
Suzuki S. Murakami M. Nishikiori T. Harada S. Annual changes in the fukushima residents' views on the safety of water and air environments and their associations with the perception of radiation risks Journal of Radiation Research 59 2 2018 31 39
Tanaka Y. Sakamoto Y. Matsuka T. Transmission of rumor and criticism in Twitter after the Great Japan Earthquake Annual meeting of the cognitive science society (p. 2387) 2012, August
Tang C.S.K. Wong C.Y. An outbreak of the severe acute respiratory syndrome: Predictors of health behaviors and effect of community prevention measures in Hong Kong, China American Journal of Public Health 93 11 2003 1887 1888 14600058
Tasnim S. Hossain M.M. Mazumder H. Impact of rumors and misinformation on covid-19 in social media Journal of Preventive Medicine Public Health 53 3 2020 171 174 32498140
Tausczik Y. Faasse K. Pennebaker J.W. Petrie K.J. "Public anxiety and information seeking following the H1n1 outbreak: Blogs, newspaper articles, and wikipedia visits," Health Communication 27 2 2012 179 185 21827326
Urbach N. Ahlemann F. Structural equation modeling in information Systems research using partial least squares Journal of Information Technology Theory and Application 11 2 2010 5 40
Vancouver J.B. Thompson C.M. Tischner E.C. Putka D.J. Two studies examining the negative effect of self-efficacy on performance Journal of Applied Psychology 87 3 2002 506 12090608
Vancouver J.B. Thompson C.M. Williams A.A. The changing signs in the relationships among self-efficacy, personal goals, and performance Journal of Applied Psychology 86 4 2001 605 11519645
Wang P. Fan W. Li Q. The alleviation effect of knowledge integration on anxiety in parental vaccine decision-making supported by online Communities Journal of Knowledge Management 25 6 2020 1575 1593
Wang Q. Yang X. Xi W. Effects of group arguments on rumor belief and transmission in online Communities: An information cascade and group polarization perspective Information & Management 55 4 2018 441 449
Wieber F. Odenthal G. Gollwitzer P. Self-efficacy feelings moderate implementation intention effects Self and Identity 9 2 2010 177 194
Woodworth R.S. Marquis D.G. Psychology (psychology revivals): A study of mental life 2014 Psychology Press London
Yang X. Lyu Y. Tian T. Liu Y. Liu Y. Zhang X. Rumor detection on social media with graph structured adversarial learning Proceedings of the twenty-ninth international joint conference on artificial intelligence: Ijcai 2020
Yavetz G. Aharony N. Sofer Y.Y. Information needs and seeking behaviors of Israeli citizens during the COVID-19 outbreak Aslib Journal of Information Management 2022 10.1108/AJIM-06-2021-0167
Yıldırım M. Güler A. COVID-19 severity, self-efficacy, knowledge, preventive behaviors, and mental health in Turkey Death Studies 46 4 2020 979 986 32673183
Yu Z. Lu S. Wang D. Li Z. Modeling and analysis of rumor propagation in social networks Information Sciences 580 2021 857 873
Zhao Y. Da J. Yan J. Detecting health misinformation in online health Communities: Incorporating behavioral features into machine learning based approaches Information Processing & Management 58 1 2021 102390
Zhu H. Rumour detection based on deep hybrid structural and sequential representation networks Journal of Information Science 2021 10.1177/01655515211056579
|
PMC010xxxxxx/PMC10229389.txt |
==== Front
Meas Tech
Measurement Techniques
0543-1972
1573-8906
Springer US New York
2192
10.1007/s11018-023-02192-y
Medical and Biological Measurements
Metrological Support of Nucleic Acid Sequence Analysis
Melkova O. N. melkova@vniims.ru
1
Kulyabina E. V. kuliabina@vniims.ru
1
Fomina S. Yu. fomina@vniro.ru
2
Volkov A. A. alexavolkov@gmail.com
2
1 grid.494345.b Russian Research Institute for Metrological Service, Moscow, Russia
2 grid.465444.3 0000 0000 9551 539X Russian Federal Research Institute of Fisheries and Oceanography (VNIRO), Moscow, Russia
31 5 2023
2023
66 1 7680
1 8 2022
© Springer Science+Business Media, LLC, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Metrological support for the process of molecular genetic identification is proposed. An establishment of a nucleotide sequence is the main objective of the DNA/RNA analysis. Such nucleotide sequence provides key information about the role and function of nucleic acids in living systems, as well as the origin of biological objects during molecular genetic identification. The molecular genetic identification is based on the modern methods of studying DNA sequences, which are inherited by the living organisms from generation to generation. In some cases, comparison of DNA sequences is required to establish the authenticity of biological objects, as well as determine biological affiliation or kinship with other biological objects. The problems of metrological support of molecular genetic identification by Sanger DNA sequencing using fluorescent-labeled reaction terminators and capillary electrophoresis are considered. Methods and means of metrological support of the DNA/RNA analysis were developed, which include the determination of a nucleotide as a nucleic acid monomer, the use of a nucleotide sequence of nucleic acids used in certified methods and procedures, the use of a genetic analyzer as a measurement tool along with related auxiliary equipment, as well as the use of a standard sample of a fragment of human mitochondrial DNA has been developed and approved. A Nanofor 05 domestic genome analyzer was tested. The following two procedures were developed and certified: for measuring the nucleotide sequence of the section of the control region of mitochondrial DNA of fish (sturgeon and paddlefish families) by Sanger sequencing, and for measuring the nucleotide sequence of the 5' region of the COI gene of the mitochondrial DNA of aquatic biological resources by Sanger sequencing. Both measurement procedures are based on the use of fluorescent-labeled reaction terminators and capillary electrophoresis. The results are important for use in the DNA studies in the field of biological sciences, food industry, biotechnology, and DNA forensics.
Keywords
units of measurements
nucleotide sequence
DNA sequencing
metrological support
reference sample
measurement procedures
molecular genetic identification
issue-copyright-statement© The Founders of the Journal and the Editorial Board 2023
==== Body
pmcIntroduction. After the discovery of the primary and secondary structures of deoxyribonucleic acid (DNA) in the second half of the 20th century. a whole new direction has emerged in the field of biochemical research, related to the development and enhancement of the methods for determining DNA nucleotide sequence. Currently, the application of the DNA sequence analysis is continuously expanding every year. This includes analyzing biomaterial fragments during forensic medical examination, establishing paternity, determining the origin of food products and much more. This molecular genetic analysis compares the DNA nucleotide sequences in the analyzed sample with known sequences, published and stored in various annotated databases. Currently, international databases of nucleotide sequences in various fields of DNA and ribonucleic acid (RNA) research are available, the largest of which is the NCBI database (USA)1 [1].
The COVID-19 pandemic has significantly expanded the scope of such research. To date, scientists around the world have sequenced tens of thousands of genomes of various strains of coronavirus in order to exchange data on the basis of the international open platform GISAID,2 created in 2008 after the outbreak of bird flu (H5N1) to collect and update information about viral genomes. The accumulated amount of information allows researchers from diff erent countries to determine the origin of COVID-19 outbreaks in their countries and trace the spread of infection in certain regions.
Metrological support of the DNA analysis in the Russian Federation is significantly behind compared to the global pace of development of the tools and methods of such support. In Russia, such developments are not systematic, and their volume is not sufficient. Part of the reason is a clear lack of funding in this area of research. The units of measurement used in laboratory medicine are not included in the list (even with amendments) from Decree No. 8793 of the Government of the Russian Federation dated October 31, 2009.
In this paper, an example is provided on how to use the estimate of the nominal property values [2] of a specific sequence of a fragment of a single-stranded nucleic acid (DNA/RNA) or double-stranded DNA. The size of a fragment of a specific sequence of a single-stranded nucleic acid (DNA/RNA) or a double-stranded DNA is a countable value with a unit of "one" [3], and is expressed using a number of nucleotides. In addition, such qualitative property as a variation of the DNA/RNA nucleotide sequence within a specific genomic section is also used.
The objective of the study is to develop the methods and means of metrological support of the DNA analysis for the All-Russian Research Institute of Fisheries and Oceanography.
Statement of the problem. The authors proposed to define nucleotide sequence as Seq = (x; N), where x is the consecutive number of a nucleotide in the sequence, which represents a natural number from 1 to X (X is the size of a fragment of a specific sequence of a single-stranded nucleic acid (DNA/RNA) or a double-stranded DNA); N is the nucleotide designation, which can assume any of the following four values: A, G, C, T(U); Seq is the nucleotide order, which determines their specific sequence in the DNA/ RNA composition.
Previously, at the Russian Research Institute for Metrological Service (VNIIMS), a type of the reference sample for calibration and metrological support of a single available genome analyzer (CS FLX) was tested and approved. The reference sample represents a fragment of the DNA sequence of plasmid pUC18 (size – 271 nucleotides) [4]. Later, a type of the reference sample of a fragment of the DNA sequence of plasmid pUC18 (size – 717 nucleotides) was tested and approved [5]. However, in 2016, CS FLX analyzers were discontinued, and the period of validity of the certificates for state standard reference samples (SRS) of the plasmid ended.
One organization that widely uses a DNA analysis in everyday work is the All-Russian Research Institute of Fisheries and Oceanography (VNIRO). This is the head institute of the fishing industry (managed by the Federal Agency for Fishery), which coordinates the implementation of plans and programs of fishery studies. VNIRO and its branches have dozens of well-equipped laboratories throughout the country along with the practical experience of working with genetic material.
VNIRO has developed methods for identifying the species of fish, commercial invertebrates and their products, determining species of sturgeon, conducting molecular genetic certification of sturgeon stock, determining the population of sturgeon breeders and young fish, as well as tracing the origin of the products to a previously certified stock of breeders.
In 2019, the VNIRO researchers reached out to the VNIIMS for the purpose of developing metrological support of such methods within the scope of the work on reviewing the state standard for the identification of sturgeon caviar4 (previously, only morphological identification methods were used with the standard).
Such types of analyzes are also conducted in the field of forensics.5 Thus, such measurements are related to the field of state regulation in terms of the performance of court orders, as well as instructions of the prosecutor's office and state authorities.
Genetic identification of fish is done by Sanger sequencing using fluorescent-labeled reaction terminators and capillary electrophoresis, and includes the following two main stages: polymerase chain reaction (PCR) and sequencing reaction with subsequent analysis of a fluorescent-labeled single-strand DNA.
The measurement method includes the establishment of the sequence of DNA nucleotides in the analyzed sample, which is recorded as a sequence of letters encoding four possible nitrogenous DNA bases (A, T, G, C) (direct and/or reverse sequence) and the measurement of the number of nucleotide pairs.
Determination of the DNA nucleotide sequence includes the following steps:isolation of DNA from a sample of biological origin and its purification;
amplification of a certain DNA section (by PCR);
visual detection of the product of amplification (PCP product) in agarose gel (Fig. 1);
purification of the PCR product from the remaining components of the amplification reaction;
sequencing reaction of the amplification product using fluorescent-labeled terminators according to Sanger; and
electrophoretic separation of the products of sequencing reaction by capillary electrophoresis with the detection of a fluorescent signal and concurrent computer-aided recording of a four-colored electropherogram, followed by the conversion of the fluorescence peaks of the DNA sequencing product into a letter code, describing the nucleotide sequence in the analyzed section of the DNA of the studied sample (Fig. 2).
Fig. 1. Agarose gel electrophoresis stained by ethidium bromide to visualize the PCR products in ultraviolet light: M, molecular weight marker; 1–8, PCR product.
Fig. 2. Detection of a fluorescence signal with simultaneous computer-aided recording of a four-colored electrophoretogram.
Due to the lack of domestic reference DNA samples, it was decided to use a reference sample of the DNA nucleotide sequence (SRM 2392-I), previously tested by NIST (USA), as a basis for ensuring traceability of the method. This sample contains a human mitochondrial DNA (size – 16,569 base pairs) isolated from the HL-60 cell culture [6].
When creating the SRM 2391-I sequence for amplifying DNA of the HL-60 cells, multiplex PCR was used along its entire length with 58 pairs of primers. The PCR products were sequenced using an ABI 310 automated genetic analyzer (Applied Biosystems, Inc., USA). Sequences of the typical PCR products of the final HL-60 DNA included into the SRM 2392-I sequence were retested to ensure sequencing accuracy. An inter-laboratory evaluation of the amplification, sequencing, and analysis of the HL-60 matrix data was performed by four laboratories, including NIST. The sequences obtained in all laboratories were identical. The SRM 2392-I certificate [6] also contains a comparison with the Cambridge Reference Sequence (CRS) [7, 8].
Based on the NIST SRM 2392-I fragment, a reference sample of the human mitochondrial DNA fragment was developed, approved, and (based on the test results) entered into the Federal Information Fund for Ensuring the Uniformity of Measurements (FIF EUM) under No. 11607-2020. The DNA sequence is identical to the human mitochondrial DNA section of the culture of HL-60 cell line with a DNA fragment size of 1.794 nucleotides (site 5999-7792).
When reading the section of the reference sample using a domestic genetic analyzer (Nanofor 05) based on capillary electrophoresis, a reading accuracy of 98.5% was achieved under conditions of repeatability and reproducibility based on the results from various laboratories [9].
Since the nucleotide sequence relates to qualitative properties (and the approval of the reference samples based on the qualitative properties is not yet regulated), such sequence could not be used as a certification characteristic when approving the reference sample. It was decided to rate the mass fractions of nucleotides (in percent) and express the mass concentration in units [ng/μl]. Since the mass fractions were calculated by analyzing the nucleotide sequence, the latter has been included in the description of the type of the reference sample as an additional characteristic.
Study results. Metrological characteristics of the reference sample are shown in Table 1. The relative error margins of the certified value of the reference sample for all certified characteristics (mass fractions) were 0.5% at a confidence probability of P = 0.95.TABLE 1. Mass fractions of nucleotides
Mass fraction Certified value of the reference sample, %
Adenine (A) 27.55
Guanine (G) 22.45
Cytosine (C) 22.45
Thymine (T) 27.55
According to the FIF EUM6 data, the range of permissible certified values of mass concentration of the human mitochondrial DNA fragment of the culture of HL-60 cell line (section 5999–7792) constitutes 75–125 ng/mm3 with the relative error margins of ±1% at P = 0.95.
Testing of the domestic genome analyzer (Nanofor 05) was conducted at the Design Automation Institute of the Russian Academy of Sciences (DAI RAS) (St. Petersburg) and FIF EUM (No. 81576-21).
The following two measurement procedures have been developed and certified:a procedure for measuring the nucleotide sequence of a section of the control region of the mitochondrial DNA in fish (sturgeon and paddlefish families) by Sanger sequencing using fluorescent-labeled reaction terminators and capillary electrophoresis; and
a procedure for measuring the nucleotide sequence of the 5' region of the COI gene of the mitochondrial DNA of aquatic biological resources by Sanger sequencing using fluorescent-labeled reaction terminators and capillary electrophoresis.
The developed procedures were tested, certified, and entered into the FIF EUM under No. FR.1.31.2020.38274; FR.1.39.2021.40174.
The metrological characteristics of the procedures are provided in Table 2 (SD – standard deviation).TABLE 2. Metrological characteristics of the measurement procedures FR.1.31.2020.38274, FR.1.39.2021.40174
Characteristics Value
Measurement range of the DNA nucleotide sequence size, nucleotide pairs 50–750
Accuracy parameter – relative error of reading DNA nucleotide sequence (P = 0.95), % 1.8
Repeatability parameter (relative SD of repeatability), σr, % 0.5
Reproducibility parameter (relative SD of reproducibility), σR, % 0.6
The procedures establish the DNA sequencing techniques according to Sanger with the use of fluorescent-labeled reaction terminators and capillary electrophoresis. These procedures also cover the sequencing of DNA sites of diff erent origins with the chosen amplification conditions, provided a specific amplicon is obtained without additional bands. A modification of the qualitative composition of primers for DNA amplification is permitted along with the expansion of the application field in order to develop and improve the techniques of sequencing certain sections of genomic DNA, including the ones that are diff erent from the specified region of mitochondrial DNA of fish and invertebrates. The above procedures can be used to evaluate the compliance with the requirements of the Technical Regulations of the Eurasian Economic Union (TR EAEU 040/2016) "On the Safety of Fish and Fish Products," Sections II and III.
Metrological support of the results of DNA sequencing is applicable to any obtained DNA sequence and is independent of the sequencing methods and/or generation of utilized sequencing techniques [10], since the result of any DNA sequencing is the same type of data — a sequence of letters encoding certain nucleotides.
The procedures can be used in the following fields: research, forensics, laboratory diagnostics, measures of state control (supervision).
Conclusion. According to the study results, molecular genetic identification by Sanger sequencing using fluorescent labeled reaction terminators and capillary electrophoresis can be used to make informed decisions regarding fish species, determine their habitat, and genome variability. The results of the analysis are traceable, and their accuracy has been determined. The study results can also be used in virology, forensics, the food industry.
Conflict of interest. The author declares no conflict of interest.
1 NCBI website: https://www.ncbi.nlm.nih.gov (access date: 09/01/2022).
2 GISAID website: https://www.gisaid.org/hcov19-variants/ (access date: 09/01/2022).
3 Decree No. 879 of the Government of the Russian Federation "On Approval of the Regulation on Units of Measurement allowed for use in the Russian Federation" dated October 31, 2009.
4 GOST 30812-2002. Raw materials and food products. Method for the identifi cation of sturgeon caviar.
5 Criminal Code of the Russian Federation No. 63-FZ dated 06/13/1996 (as amended on 12/29/2022). Article 258.1: Illegal acquisition and trafficking of highly valued wild animals and aquatic biological resources representing species entered in the Red Book of the Russian Federation and/or protected by the international agreements of the Russian Federation.
6 FIF EUM website: https://fgis.gost.ru/fundmetrology/registry/19 (access date: 12/06/2022).
Translated from Metrologiya, No. 1, pp. 70–74, January, 2023. Original article was submitted January 27, 2022.
==== Refs
References
1. Ratnasingham S Hebert PDN Molecular Ecology Notes 2007 7 3 355 364 10.1111/j.1471-8286.2007.01678.x 18784790
2. Nordin G Dybkaer R Forsum U Fuentes-Arderiu X Pontet F Pure and Applied Chemistry 2018 90 5 913 935 10.1515/pac-2011-0613
3. International system of units SI, 9th ed. (2019), available at: https://www.vniim.ru/files/SI-2019.pdf (access date: 08/18/2022).
4. Golubev SS Kononogov SA Ravin NV Skryabin KG Meas. Tech. 2012 54 1408 1411 10.1007/s11018-012-9903-4
5. Golubev SS Kononogov SA Kudeyarov YuA Mardanov AV Nikolaeva PYu Ravin NV Skryabin KG Meas. Tech. 2012 55 345 350 10.1007/s11018-012-9962-6
6. NIST Certificate of Analysis, Standard Reference Material 2392-I, available at: https://www-s.nist.gov/m-srmors/certificates/2392.pdf (access date: 08/18/2022).
7. Levin BC Holland KA Hancock DK Coble M Parsons TJ Kienker LJ Williams DW Jones MP Richie KL Mitochondrion 2003 2 6 387 400 10.1016/S1567-7249(03)00010-2 16120335
8. S. Anderson, A. Bankier, B. Barrell, et al., Nature, No. 290, 457–465, 10.1038/290457a0.
9. Volkov AA Volkov IA Plugov AG Kulyabina EV Melkova ON Lavrov GS Bocharova DV Alekseev Ya. I Meas. Tech. 2021 64 64 69 10.1007/s11018-021-01897-2
10. A. G. Borodinov, V. V. Manoilov, I. V. Zarutsky, A. I. Petrov, and V. E. Kurochkin, Nauchnoe Priborostroenie, 30, No. 4 (2020), 10.18358/np-30-4-i320.
|
PMC010xxxxxx/PMC10230310.txt |
==== Front
Clin Nurs Res
Clin Nurs Res
CNR
spcnr
Clinical Nursing Research
1054-7738
1552-3799
SAGE Publications Sage CA: Los Angeles, CA
37254308
10.1177/10547738231177481
10.1177_10547738231177481
Research Articles
Experiences of Patients With Type 2 Diabetes Who Recovered From COVID-19 in the Pandemic Period: A Qualitative Study
https://orcid.org/0000-0001-7706-4748
Çetinkaya Özdemir Serap PhD 1
https://orcid.org/0000-0002-7793-2311
Eren Merve Gulbahar MSc 1
https://orcid.org/0000-0002-1658-6515
Sert Havva PhD 1
Öztürk Fatma Can RN 2
1 Department of Internal Medicine Nursing, Faculty of Health Science, Sakarya University, Turkey
2 Sakarya Sadika Sabanci State Hospital, Turkey
Serap Çetinkaya Özdemir, Department of Internal Medicine Nursing, Faculty of Health Science, Sakarya University, Sakarya, 54050, Turkey. Email: serapc@sakarya.edu.tr
30 5 2023
7 2023
30 5 2023
32 6 983991
© The Author(s) 2023
2023
SAGE Publications
This article is made available via the PMC Open Access Subset for unrestricted re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the COVID-19 pandemic or until permissions are revoked in writing. Upon expiration of these permissions, PMC is granted a perpetual license to make this article available via PMC and Europe PMC, consistent with existing copyright protections.
Coronavirus disease 2019 (COVID-19) infection can induce acute and chronic complications by affecting the self-management behaviors of individuals with diabetes. The objective of this study is to examine the physical, psychosocial health, and self-management experiences of type 2 diabetes patients who have recovered from COVID-19, 1 year after the infection. The study adopted a qualitative research design, specifically content analysis. In all, 14 patients with type 2 diabetes who presented to the diabetes outpatient clinic were interviewed by teleconferencing, which lasted approximately 25 to 30 minutes. The Standards for Reporting Qualitative Research guidelines were used. Based on the participants’ responses, four main themes were determined: obstacles in activities of daily living, feeling of psychosocial problems, changes in health and treatment management, and patient self-management practices. Amid the pandemic, diabetes nurses should strive to recognize the issues that diabetes patients encounter. To assist patients, telemedicine should be leveraged, and evidence-based practices must be developed.
COVID-19
type 2 diabetes
qualitative research
pandemic
self-management
typesetterts1
==== Body
pmcIntroduction
Although the pathophysiology of the novel coronavirus disease 2019 (COVID-19), which is an unprecedented global health problem, has not been completely understood yet, severe and fatal cases are mainly reported in elderly individuals and patients with comorbidities, such as cardiovascular diseases, hypertension, obesity, type 2 diabetes, and chronic pulmonary or renal diseases (Chen et al., 2020; Huang et al., 2020; Mahluji et al., 2021; Wang et al., 2020). Studies in the early COVID-19 pandemic period showed that 26.8% of patients with COVID-19 had diabetes, and their risk of COVID-19-associated mortality was approximately 50.0% higher (Mukona & Zvinavashe, 2020; Muniyappa & Gubbi, 2020). A meta-analysis revealed a twofold increase in disease severity and intensive care requirement among patients with type 2 diabetes (Gupta et al., 2020).
The American Diabetes Association (ADA) offered general recommendations for patients with diabetes and COVID-19 during the pandemic (ADA, 2020). These recommendations have included higher fluid intake to prevent dehydration, preserving a glycemic balance close to individualized target values, making accurate and regular blood glucose measurements at home, monitoring blood glucose levels to avoid hypoglycemic attacks and ketoacidosis, increasing physical activity, regulating diet, and practicing hand hygiene (Bornstein et al., 2020; Grabia et al., 2020; Singh et al., 2020). Because of social distancing and lockdown measures during the COVID-19 pandemic, managing diabetes and the negative effects of these factors on healthy lifestyle behaviors became more difficult (Jia et al., 2021; Singh et al., 2020). During the pandemic, patients with type 2 diabetes had reduced physical activity, dietary habit changes, routine clinical follow-up interruptions, and difficulty accessing oral antidiabetic drugs and insulin (Ruiz-Roso et al., 2020; Verma et al., 2020). Consequently, patients with type 2 diabetes may experience continuous hyperglycemic periods or hypoglycemic attacks (Ruiz-Roso et al., 2020). In addition, to cope with increased psychological stress due to uncertainty and social isolation during the pandemic, patients tend to excessively consume products rich in simple carbohydrates, which improve their mood and reduce stress (Graneheim & Lundman, 2004; Önmez et al., 2020). Consequently, the risk of obesity and complications from COVID-19 has increased. A study in China emphasized that the glycemic control of older patients with type 2 diabetes worsened during the pandemic, and their fasting blood glucose levels increased. Jia et al. (2021) reported an increase of 3.68% in the HbA1c levels of diabetic patients within 45 days of the pandemic. In the study examining the effects of lockdown measures in patients with type 2 diabetes in Turkey during the pandemic, Önmez et al. (2020) determined an increase in individuals’ consumption of carbohydrate-rich foods due to increased stress levels and reduced physical activity. Therefore, the weight, waist circumference, and fasting and postprandial glucose levels of diabetic patients also increased.
In the literature review, some studies have investigated the diabetes management routines of diabetic patients who had or had not been infected with COVID-19 during the pandemic (Jia et al., 2021; Önmez et al., 2020; Ruiz-Roso et al., 2020; Verma et al., 2020). However, only a limited number of studies focused on the problems experienced by diabetic patients after recovering from COVID-19, a current problem in Turkey and other countries. This study aims to use the experiences of patients with type 2 diabetes who have recovered from COVID-19 to inform healthcare professionals’ treatment and care processes. Therefore, the objective of this study is to examine the experiences of type 2 diabetes patients after recuperating from COVID-19.
Method
Study Design and Settings
This study employed a qualitative research design that employed content analysis. This approach involves analyzing collected data to make sound conclusions. Content analysis is a systematic and unbiased research method that enables researchers to obtain credible findings from written, visual, or verbal data and to characterize and quantify specific phenomena (DowneWamboldt, 1992; Graneheim & Lundman, 2004). Patients with type 2 diabetes who presented to the diabetes outpatient clinic and had recovered from COVID-19 were interviewed. Their situation and diabetes management-related experiences were questioned. The Standards for Reporting Qualitative Research guidelines were used. The authors of this study used a qualitative methodology, which included content analysis as outlined in the works of Graneheim and Lundman (2004).
Participants
The purposive sampling method was employed in this study. The study participants were individuals who were at least 18 years of age and had been diagnosed with type 2 diabetes for a minimum of 3 years, which is the typical adherence duration to diabetes treatment. Participants who had recovered from COVID-19 1 year after being infected were included in the study, with the diagnosis of diabetes for at least 3 years serving as the inclusion criterion. The severity and prognosis of COVID-19 symptoms may impact patients’ experiences, depending on the recovery period. Therefore, participants in this study were individuals with type 2 diabetes who had recovered from COVID-19. Patients with communication problems, psychiatric illnesses, refusal to participate in the study, or those who had not contracted COVID-19 were excluded. The interviews were held by teleconferencing. Data saturation was reached at 14 participants. During data collection, the researchers prepared a descriptive information form in line with the relevant literature to collect sociodemographic information and a semi-structured interview form. Data were collected between August 2021 and January 2022.
Qualitative Data Collection
Two experts on qualitative research methods evaluated the interview guide, and the questions were tested in two pilot interviews. The interview guide was finalized after revision based on expert opinions and pilot implementation. After the pilot implementation, the data were collected using the descriptive information form and semi-structured questions designed to obtain information on the experiences of our study participants. The descriptive information form included questions on the participant’s age, sex, body mass index (BMI), marital status, education status, medication usage status, and chronic disease status. The semi-structured interview form included questions on participants’ experiences after recovering from COVID-19 and their diabetes management-related experiences (Table 1). Before starting the interviews, the researchers provided the participants with information about the study. Each interview, which lasted approximately 25 to 30 minutes, was participated simultaneously by the patient and two researchers. All data, including the copied interview, were saved on password-protected computer files.
Table 1. Interview Questions.
No. Question
1 What did you experience in terms of your physical health after recovering from COVID-19?
2 What did you experience in terms of your mental/psychological health after recovering from COVID-19?
3 What did you experience in the social sense after recovering from COVID-19?
4 What changed in your diabetes management after recovering from COVID-19?
Note. COVID-19 = coronavirus disease 2019.
Ethical Considerations
This study adhered to the principles outlined in the Declaration of Helsinki and was approved by the Institutional Review Board of Sakarya University (Approval no. E-71522473–050.01.04–45425–400). Prior to the study, the principal researchers provided participants with a detailed explanation of the study’s objectives and methods, assured them of the confidentiality measures in place, and obtained verbal informed consent from each participant.
Data Analysis
The techniques of Graneheim and Lundman (2004) for extracting themes from qualitative data were used for data analysis. To have an idea about the general content of the text, the entire text of the interviews was read by the researchers a few times. The researchers independently created codes, and these codes were rechecked based on the original transcripts. The researchers gathered to determine both main themes and sub-themes. The data were re-evaluated by a third researcher when two researchers did not agree on the codes or themes.
Trustworthiness of the Study
Data trustworthiness was assessed using the criteria of credibility, dependability, conformability, and transferability, as outlined by Guba and Lincoln (1994). To ensure credibility, a third-party expert was consulted to review the themes, sub-themes, and codes developed from the data collected. To ensure dependability, a third researcher reviewed the research process. All research data, including audio recordings and transcripts, were properly stored to ensure confirmability and allow for future validation of the findings. In addition, two diabetes patients who had recovered from COVID-19 but did not participate in the study confirmed that the themes identified were consistent with their experiences, thereby bolstering the transferability of the study’s findings.
Results
These findings included data from people who had type 2 diabetes and had recovered from COVID-19 after a year.
Participant Characteristics
The mean age and BMI of the 14 study participants were 55.14 and 29.93, respectively. In addition, 57.1% were women, 78.6% were married, and 50.0% were primary school graduates. Meanwhile, 50.0% used oral antidiabetic drugs, and 50.0% used insulin and oral antidiabetic drugs. In addition to diabetes, 50.0% of the participants had another chronic disease. Table 1 displays the patients’ demographic information.
Thematic Results
This study determined four themes: “obstacles in activities of daily living,” “feelings of psychosocial problems,” “changes in health and treatment management,” and “patient self-management practices.” Table 2 shows the themes and sub-themes.
Table 2. Themes and Sub-themes.
Themes Sub-themes
Obstacles in activities of daily living Fatigue
Sleepiness
Feeling of psychosocial problems Irritability
Fear
Less socializing
Changes in health and treatment management Blood glucose fluctuations
Elevated blood pressure
Changes in the treatment plan
Patient self-management practices Adherence to diet
Adherence to physical activity
CAM practices
Note. CAM = complementary and alternative medicine.
Theme 1: Obstacles in Activities of Daily Living
Daily activities are all important things that people do to keep their lives going. Obstacles in their daily living activities may disrupt individuals’ roles and functions in life. Most of the participants encountered various obstacles in performing their activities of daily living, and the main obstacles were fatigue and sleepiness.
Fatigue
Diabetes patients experience fatigue, which causes them to miss work. Most of the participants expressed that they felt fatigued. They especially pointed out that their daily functions were disrupted due to fatigue, and they could not perform their daily activities sufficiently:I used to have a barn. I [then] became unable to enter the barn. I sold my animals after the disease because I get really tired and cannot do my chores. (P12, F, 34 years old)
I feel intense physical fatigue in my body. I cannot go on walks. (P1, F, 59 years old)
Sleepiness
COVID-19 has disrupted the sleep patterns of diabetic patients. Even after recovering from the COVID-19 infection, the majority of participants stated that they needed sleep during the day:After I had COVID-19, I was getting too sleepy. That is, sleeping makes me feel dazed. I was feeling drowsy even after having a little bit of food. This started to disturb me. When I slept, when no one woke me up, I would sleep for hours. Such that my eyelashes would ache when I woke up. (P11, F, 47 years old)
I sometimes doze off and cannot do my daily chores. (P12, F, 34 years old)
Theme 2: Feeling of Psychosocial Problems
Patients with diabetes who have had COVID-19 experienced psychosocial difficulties due to their fear of being re-infected with the disease and developing diabetes-related complications as a result of COVID-19. Therefore, the participants experienced irritability, fear, and less frequent socialization.
Irritability
The fluctuating blood glucose levels made them irritable. Many participants stated they developed an irritable mood after the COVID-19 infection:I was not an irritable person before. After having had COVID-19, I get angry. (P14, F, 35 years old)
My mental state deteriorated after COVID-19; I became a nervous person. (P1, F, 59 years old)
Mentally speaking, I shouted at my children a lot; I was livid. Even my husband would say, “Why are you shouting at me for no reason?” (P11, F, 47 years old)
Fear
The majority of participants expressed concern about chronic diabetes complications because they believed COVID-19 worsened diabetes prognosis:My blood sugar increased because of COVID-19. I fear my high blood sugar levels will harm an organ in my body. (P4, M, 42 years old)
Since my COVID-19 infection, the most concerning issue for me about my diabetes has been kidney damage. [They say] high sugar levels ruin the kidneys. (P11, F, 47 years old)
Less socialization
Many participants socialized less frequently after the COVID-19 infection:I do not want to go out; I do not want to walk around. I’m not as social as before. (P1, F, 59 years old)
I do not want to go anywhere. I do not want anyone to visit me either. So, after I recovered from the disease, I withdrew myself by a bit. (P14, F, 35 years old)
Theme 3: Changes in Health and Treatment Management
COVID-19 has a negative impact on several body systems. Individuals with chronic diseases, such as diabetes, are especially vulnerable. The participants reported experiencing blood glucose fluctuations and elevated blood pressure after the COVID-19 infection. To manage these health changes, their physicians supervised the changes in their treatment plans.
Blood Glucose Fluctuations
Most of the participants experienced fluctuations in their blood sugar levels after the COVID-19 infection:I’m diabetic. After COVID-19, my blood sugar increases one day and decreases the other day. (P2, M, 73 years old)
After I recovered from COVID-19, my blood sugar has gone up and down. Its balance is disrupted. I presented to the doctor then. My state of health was very well before my COVID-19 infection. (P5, M, 57 years old)
Elevated Blood Pressure
Many patients experienced elevated blood pressure after COVID-19 infection, and antihypertensive drug treatment was started with their physician’s recommendations:My blood pressure increased after having COVID-19. I started blood pressure medication. My blood pressure became regular after starting to use these drugs. (P13, F, 50 years old)
My blood pressure went up after COVID-19. (P1, F, 59 years old)
Changes in the Treatment Plan
Most participants stated that their treatment plans changed due to blood glucose fluctuations after the COVID-19 infection, and their physicians added oral antidiabetic drugs or insulin:I used to take only insulin before COVID-19. Then, I started using antidiabetic pills in addition to insulin. (P1, F, 59 years old)
After my second COVID-19 vaccine, my blood sugar rose to 400–500. I went to the doctor, and treatment with insulin shots was started. (P8, F, 57 years old)
Theme 4: Patient Self-Management Practices
Self-management practices for patients have been developed to mitigate the long-term damage caused by COVID-19 in the body and to prevent blood sugar fluctuations in individuals with diabetes. The sub-themes were determined as adherence to diet, physical activity, and complementary and alternative medicine (CAM) practices.
Adherence to Diet
Many participants stated that they had not been paying sufficient attention to their diet before the COVID-19 infection. However, they adhered to their diet after the COVID-19 infection:For my diet, I avoid sugary things; I used to eat foods like potatoes before. Now, I eat them less frequently. I try to be careful. (P12, F, 34 years old)
I have removed sugar from my diet; I eat very little bread. I pay attention to the recommendations of the nurse. (P11, F, 47 years old)
Adherence to Physical Activity
The participants reported that they started to spend time on physical activities and tried to maintain them:After the disease [COVID-19], I go on walks for my diabetes every day. I maintain a normal pace because I get shortness of breath. It does not exceed half an hour. (P8, F, 57 years old)
[I realized] I had not been taking care of myself until now. Thankfully, I do now. I have workout equipment at home; I exercise. When I go out, I walk. When I visit downtown, I come back home on foot. (P14, F, 35 years old)
CAM Practices
Most participants started using herbal products to reduce diabetes-related complications and alleviate the harmful effects of COVID-19 on their general health. Specifically, they began to drink green tea, chamomile tea, rose hip tea, and olive leaf tea:After [having had] COVID-19, I used my pills as an oral medication for diabetes. I drink rose hip tea because it is good for diabetes patients. The rose hip I prepare is natural. (P3, F, 75 years old)
I drink olive leaf tea. My dietician said it lowers blood sugar. (P14, F, 35 years old)
The participants reported drinking green tea, chamomile tea, rose hip tea, and olive leaf tea after being diagnosed with COVID-19 to recover easily and prevent potential future infections:We do not use any product other than vinegar. Just natural vinegar. . . It has been beneficial. For instance, COVID-19 made me feel tight in my chest, and I had shortness of breath. Vinegar would make me relax a lot; that is, it would make the disease [COVID-19] easier on our body and breathing. This vinegar really made a difference. (P4, M, 42 years old)
They say, pickle juice is highly effective against COVID-19. As I already like it a lot, I’ve drunk a lot of it. (P5, M, 57 years old)
Discussion
Based on patient’s interviews, four main themes were determined: obstacles in activities of daily living, feeling of psychosocial problems, changes in health and treatment management, and patient self-management practices.
Although the prevalence of post-viral fatigue after COVID-19 infection is high, it is usually independent of patients’ respiratory function and exercise capacity (Mittal et al., 2021). Although the pathogenesis of post-viral fatigue is not entirely known, lack of physical activity, malnourishment, hypoxia, and fibrosis caused by COVID-19 pneumonia can lead to muscle fatigue (Mittal et al., 2021; Townsend et al., 2020). Prolonged post-viral fatigue may negatively affect individuals’ activities of daily living by causing chronic fatigue syndrome (Townsend et al., 2020). In our study, fatigue was the most significant obstacle to activities of daily living, and the participants reported that fatigue led to reduced activities of daily living and physical activity. The sub-theme of fatigue in this study was compatible with other reports in the literature.
A study, including patients with type 1 and 2 diabetes, showed that 17.0% and 45.0% of the participants slept longer than 8 hours daily before and during the pandemic, respectively (Grabia et al., 2020). Moreover, a case series showed that increased sleepiness in diabetic patients was associated with hyperglycemia symptoms after COVID-19 infection (Meza et al., 2020). Regarding our study participants, an imbalance in blood glucose levels and the lockdown measures may have caused a reorganization in their circadian rhythm, and reduced socialization and occupational restrictions may have affected their habits regarding sleeping and staying awake, thereby increasing their sleepiness states.
Individuals with chronic diseases are susceptible to fear and worry during a disease outbreak, such as the COVID-19 pandemic (Al-Rahimi et al., 2021; Shi et al., 2020). Our study participants revealed that they experienced increased irritability and fear after the COVID-19 infection. Usually, the general state of fear associated with social isolation and the COVID-19 pandemic is accompanied by fear regarding diabetes complications. Al-Rahimi et al. (2021) reported that the rates of health-related fear and anxiety in individuals with chronic diseases during the COVID-19 pandemic were higher than in healthy individuals. Thus, we expected our study participants to experience fear, considering that the long-term effects of COVID-19 and its impact on existing chronic diseases are unknown, and its mortality risk is relatively high.
This study showed that the participants who had recovered from COVID-19 did not want to go outside or be visited and socialized less frequently. Previous studies have examined the social effects of the lockdown process, but no study has investigated social life after recovery from COVID-19 (Santini et al., 2020; Shi et al., 2020). Our participant’s increased awareness after COVID-19 infection, fears of COVID-19 re-infection, thoughts of diabetes prognoses worsening due to COVID-19, and acquisition of habits of spending time at home may have led them to socialize less frequently.
Elevated blood glucose levels in diabetic patients suppress their immune response and adversely affect their pulmonary function by exacerbating the respiratory dysfunction caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID-19 (Muniyappa & Gubbi, 2020). Furthermore, significant fluctuations in blood glucose levels can lead to diabetes-related complications after recovery from COVID-19 in patients who are not metabolically stable. In diabetic patients, SARS-CoV-2 infection can trigger stress and increase the imbalance in blood glucose concentrations by increasing the secretion of hyperglycemic hormones, such as catecholamines and glucocorticoids (Mahrooz et al., 2021). In our study, most participants experienced blood glucose fluctuations after recovering from COVID-19. Some had to start antihypertensive treatment due to persistent elevation of their blood pressure. Another significant result was the requirement of most of our study participants to start insulin or oral antidiabetic therapy in addition to current insulin treatment, which changes their treatment plans after recovery from COVID-19. Similarly, another study revealed that patients with type 2 diabetes needed to start insulin treatment, especially after severe COVID-19 infection (Mahrooz et al., 2021). A study in the United States reported that managing diabetes during the pandemic became more difficult. Approximately 24.7% of the participants had high blood glucose levels, and 12.6% had blood glucose level fluctuations. Montefusco et al. (2021) found chronic hyperglycemia in approximately 35.0% of patients who were followed up for 6 months after recovering from COVID-19. Guo et al. (2020) stated that the generalized stress caused by COVID-19 infection might trigger cardiovascular diseases, including hypertension. In a 12-year follow-up study of patients who recovered from SARS infection, diabetic individuals were more likely to develop cardiovascular diseases, such as hyperlipidemia, hypertension, and abnormal glucose metabolisms, than healthy volunteers (Wu et al., 2017). Although long-term outcomes for patients who have recovered from COVID-19 are still not entirely known, healthcare professionals should be aware of the importance of constant monitoring of the long-term sequelae of this disease.
One of the main elements of diabetes self-management is complying with lifestyle changes optimally (Grabowski et al., 2021). Our study found that the participants’ diet and physical activity were affected positively after recovering from COVID-19 compared with the pre-pandemic period. In the literature, rather than investigating the diabetes management routines of diabetic patients after recovering from COVID-19, studies have examined the effects of lockdown measures on self-management behaviors, such as diet, exercise, and weight management (Grabowski et al., 2021; Xue et al., 2020). During the COVID-19 pandemic, strategies that do not require much space or equipment, such as home exercises (e.g., sit-ups, Pilates, yoga) and television-based/online exercise classes, have become more well known. In addition, the importance of walking and running outdoors has increased (Ruiz-Roso et al., 2020; Xue et al., 2020). The increase in our participants’ physical activity-related awareness levels and walking durations after recovering from COVID-19 was an important finding regarding diabetes management. Similarly, another study reported that most patients (62.0%) exercised during the pandemic, and exercise durations increased by 25.0% (Ruiz-Roso et al., 2020). Similarly, other studies have shown that most diabetic patients maintain their diabetes management routines, such as following a regular diet and exercising regularly during the pandemic (Khader et al., 2020; Ruiz-Roso et al., 2020). In addition, these results have demonstrated that diabetic patients can modify their diet and physical activity routines based on their daily lives during a lockdown or change these positively.
Research has indicated that the immune status of patients is a critical factor in COVID-19 infection, and herbal treatments, particularly those with recognized immunomodulatory properties, may be utilized to prevent COVID-19 (Nugraha et al., 2020; Zhang & Liu, 2020). Similarly, our study participants consumed plant-based products, such as vinegar, pickle juice, onions, ginger, and honey, to quickly recover from COVID-19 and prevent re-infection. Furthermore, they consumed green tea, chamomile tea, rose hip tea, and olive leaf tea to supplement their diabetes management after recovering from COVID-19. In the literature, diabetic patients have also frequently used CAM methods before the pandemic (Kaynak & Polat, 2017). In addition, during the spread of COVID-19, the demand of individuals for basic herbal products increased due to the commercial presentation of these products by pharmaceutical firms and the product-related information in advertisements in the media as an agent to prevent COVID-19 (Alyami et al., 2020). Similarly, a study in Saudi Arabia revealed that the participants (22.1%) used herbal products during the pandemic because they believed these effectively prevented COVID-19 infection. Meditation, prayer, bloodletting, acupuncture, massage, honey, and medicinal plants, such as olives, chamomile, and Nigella sativa seeds, were among the CAM methods preferred for prevention and relaxation (Alyami et al., 2020). Propolis has been widely used in Brazil and China because it is believed to inhibit SARS-CoV-2 from infiltrating host cells (Paudyal et al., 2022). Although plant-based supplements may be effective in preventing COVID-19, patients should be informed that healthcare professionals’ advice should be sought before taking such supplements.
Strength and Limitations
To the best of our knowledge, this is the first qualitative in-depth study to investigate the problems and patient experiences of individuals with type 2 diabetes after recovering from COVID-19. However, our study has some limitations. Because of the ongoing COVID-19 pandemic, the interviews could not be held in person as precautions were still in place. Because the interviews were held by teleconferencing, this may have affected our access to and potential participation of other individuals. Hence, the results cannot be generalized. Moreover, this study recruited participants from a single diabetes outpatient clinic from one hospital in Sakarya, which could be a limitation.
Conclusion
The physical, psychological, and social health of the patients with type 2 diabetes in our study who had recovered from COVID-19 was adversely affected. Because the patients were aware of their health problems in general, they integrated self-management practices, which they had difficulty adhering to before the pandemic, into their daily lives more. During this period, they mainly started using herbal products, which are among CAM practices.
Implications for Nursing Practice
Clinical and in-home follow-up protocols should be created during epidemic disease periods by identifying the problems experienced by diabetic patients with a multidisciplinary team approach under the leadership of diabetes nurses. To assist patients, healthcare professionals should leverage telemedicine and develop evidence-based practices. Furthermore, patients must be informed that herbal supplements should not be taken without the advice of a healthcare professional.
Supplemental Material
sj-docx-1-cnr-10.1177_10547738231177481 – Supplemental material for Experiences of Patients With Type 2 Diabetes Who Recovered From COVID-19 in the Pandemic Period: A Qualitative Study
Click here for additional data file.
Supplemental material, sj-docx-1-cnr-10.1177_10547738231177481 for Experiences of Patients With Type 2 Diabetes Who Recovered From COVID-19 in the Pandemic Period: A Qualitative Study by Serap Çetinkaya Özdemir, Merve Gulbahar Eren, Havva Sert and Fatma Can Öztürk in Clinical Nursing Research
The authors would like to thank all the patients with type 2 diabetes who participated in the study.
Author Biographies
Serap Çetinkaya Özdemir, PhD, is a Research Assistant at the Faculty of Health Science at University of Sakarya, Turkey.
Merve Gulbahar Eren, MSc, is a Research Assistant at the Faculty of Health Science at University of Sakarya, Turkey.
Havva Sert, PhD, is an Associate Professor at the Faculty of Health Science at University of Sakarya, Turkey.
Fatma Can Öztürk, is a Registered Nurse at Sakarya Sadika Sabanci State Hospital, Turkey.
Author Contributions: SÇO, MGE, HS, and FCO contributed to the study design. SÇO, MGE, and HS implemented the research, including data collection, and analysis of the results. FCO aided in interpreting the results. SÇO, MGE, HS, and FCO led the writing of the manuscript, in consultation with HS who all provided feedback.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethical Considerations: This study was performed in line with the principles of the Declaration of Helsinki. This study was approved by the Institutional Review Board of the Sakarya University (Approval no. E-71522473-050.01.04–45425–400). Before the study, the primary researchers explained the study’s objectives and methods to the participants, promised to maintain confidentiality, and obtained verbal informed consent from all participants.
ORCID iDs: Serap Çetinkaya Özdemir https://orcid.org/0000-0001-7706-4748
Merve Gulbahar Eren https://orcid.org/0000-0002-7793-2311
Havva Sert https://orcid.org/0000-0002-1658-6515
Data Availability Statement: The data that support the findings of this study are available from the corresponding author upon reasonable request.
Supplemental Material: Supplemental material for this article is available online.
==== Refs
References
Al-Rahimi J. S. Nass N. M. Hassoubah S. A. Wazqar D. Y. Alamoudi S. A. (2021). Levels and predictors of fear and health anxiety during the current outbreak of COVID-19 in immunocompromised and chronic disease patients in Saudi Arabia: A cross-sectional correlational study. PLoS One, 16 (4 ), e0250554. 10.1371/journal.pone.0250554
Alyami H. S. Orabi M. A. A. Aldhabbah F. M. Alturki H. N. Aburas W. I. Alfayez A. I. Alharbi A. S. Almasuood R. A. Alsuhaibani N. A. (2020). Knowledge about COVID-19 and beliefs about and use of herbal products during the COVID-19 pandemic: A cross-sectional study in Saudi Arabia. Saudi Pharmaceutical Journal, 28 (11 ), 1326–1332. 10.1016/j.jsps.2020.08.023 32904846
American Diabetes Association (ADA). (2020). COVID-19: If you do get sick, know what to do. https://www.diabetes.org/blog/coronavirus-COVID-19-know-what-to-do
Bornstein S. R. Rubino F. Khunti K. Mingrone G. Hopkins D. Birkenfeld A. L. Boehm B. Amiel S. Holt R. I. Skyler J. S. DeVries J. H. Renard E. Eckel R. H. Zimmet P. Alberti K. G. Vidal J. Geloneze B. Chan J. C. Ji L. Ludwig B. (2020). Practical recommendations for the management of diabetes in patients with COVID-19. The Lancet Diabetes and Endocrinology, 8 (6 ), 546–550. 10.1016/S2213-8587(20)30152-2 32334646
Chen N. Zhou M. Dong X. Qu J. Gong F. Han Y. Qiu Y. Wang J. Liu Y. Wei Y. Xia J. Yu T. Zhang X. Zhang L. (2020). Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: A descriptive study. The Lancet, 395 (10223 ), 507–513. 10.1016/S0140-6736(20)30211-7
Downe-Wamboldt B. (1992). Content analysis: Method, applications, and issues. Health Care for Women International, 13 , 313–321.1399871
Grabia M. Markiewicz-żukowska R. Puścion-Jakubik A. Bielecka J. Nowakowski P. Gromkowska-Kępka K. Mielcarek K. Socha K. (2020). The nutritional and health effects of the COVID-19 pandemic on patients with diabetes mellitus. Nutrients, 12 (10 ), 3013. 10.3390/nu12103013 33008059
Grabowski D. Overgaard M. Meldgaard J. Johansen L. B. Willaing I. (2021). Disrupted self-management and adaption to new diabetes routines: A qualitative study of how people with diabetes managed their illness during the COVID-19 lockdown. Diabetology, 2 , 1–15. 10.3390/diabetology
Graneheim U. H. Lundman B. (2004). Qualitative content analysis in nursing research: Concepts, procedures and measures to achieve trustworthiness. Nurse Education Today, 24 (2 ), 105–112. 10.1016/j.nedt.2003.10.001 14769454
Guba E. G. Lincoln Y. S. (1994). Competing paradigms in qualitative research. In: Denzin N. K. Lincoln Y. S. (Eds.), Handbook of Qualitative Research (pp. 105–117). Sage Publications, Inc.
Guo T. Fan Y. Chen M. Wu X. Zhang L. He T. Wang H. Wan J. Wang X. Lu Z. (2020). Cardiovascular implications of fatal outcomes of patients with coronavirus disease 2019 (COVID-19). JAMA Cardiology, 5 (7 ), 811–818. 10.1001/jamacardio.2020.1017 32219356
Gupta R. Ghosh A. Singh A. K. Misra A. (2020). Clinical considerations for patients with diabetes in times of COVID-19 epidemic. Diabetes and Metabolic Syndrome: Clinical Research and Reviews, 14 (3 ), 211–212. 10.1016/j.dsx.2020.03.002
Huang C. Wang Y. Li X. Ren L. Zhao J. Hu Y. Zhang L. Fan G. Xu J. Gu X. Cheng Z. Yu T. Xia J. Wei Y. Wu W. Xie X. Yin W. Li H. Liu M. Cao B. (2020). Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. The Lancet, 395 , 497–506. 10.1016/S0140-6736(20)30183-5
Jia P. Zhang L. Yu W. Yu B. Liu M. Zhang D. Yang S. (2021). Impact of COVID-19 lockdown on activity patterns and weight status among youths in China: The COVID-19 impact on lifestyle change survey (COINLICS). International Journal of Obesity, 45 (3 ), 695–699. 10.1038/s41366-020-00710-4 33277588
Kaynak İ. Polat Ü . (2017). Diabetes Mellitus’lu hastaların tamamlayıcı ve alternatif tedavileri kullanma durumları ve diyabet tutumları ile ilişkisi. Genel Tıp Dergisi, 27 (2 ), 56–64.
Khader M. A. Jabeen T. Namoju R. (2020). A cross sectional study reveals severe disruption in glycemic control in people with diabetes during and after lockdown in India. Diabetes and Metabolic Syndrome: Clinical Research and Reviews, 14 (6 ), 1579–1584. 10.1016/j.dsx.2020.08.011
Mahluji S. Jalili M. Ostadrahimi A. Hallajzadeh J. Ebrahimzadeh-Attari V. Saghafi-Asl M. (2021). Nutritional management of diabetes mellitus during the pandemic of COVID-19: A comprehensive narrative review. Journal of Diabetes & Metabolic Disorders, 20 , 963–972. 10.1007/s40200-021-00784-5 33842400
Mahrooz A. Muscogiuri G. Buzzetti R. Maddaloni E. (2021). The complex combination of COVID-19 and diabetes: Pleiotropic changes in glucose metabolism. Endocrine, 72 (2 ), 317–325. 10.1007/s12020-021-02729-7 33886062
Meza J. L. Triana A. de Avila I. del Rio-Pertuz G. Viasus D. (2020). Diabetic ketoacidosis precipitated by COVID-19 in patients without respiratory symptoms: Case reports. Cureus, 12 (8 ), e10031. 10.7759/cureus.10031
Mittal J. Ghosh A. Bhatt S. P. Anoop S. Ansari I. A. Misra A. (2021). High prevalence of post COVID-19 fatigue in patients with type 2 diabetes: A case-control study. Diabetes and Metabolic Syndrome: Clinical Research and Reviews, 15 (6 ), 102302. 10.1016/j.dsx.2021.102302
Montefusco L. Ben Nasr M. D’Addio F. Loretelli C. Rossi A. Pastore I. Daniele G. Abdelsalam A. Maestroni A. Dell’Acqua M. Ippolito E. Assi E. Usuelli V. Seelam A. J. Fiorina R. M. Chebat E. Morpurgo P. Lunati M. E. Bolla A. M. Fiorina P. (2021). Acute and long-term disruption of glycometabolic control after SARS-CoV-2 infection. Nature Metabolism, 3 (6 ), 774–785. 10.1038/s42255-021-00407-6
Mukona D. M. Zvinavashe M. (2020). Self-management of diabetes mellitus during the Covid-19 pandemic: Recommendations for a resource limited setting. Diabetes and Metabolic Syndrome: Clinical Research and Reviews, 14 (6 ), 1575–1578. 10.1016/j.dsx.2020.08.022
Muniyappa R. Gubbi S. (2020). COVID-19 pandemic, coronaviruses, and diabetes mellitus. American Journal of Physiology-Endocrinology and Metabolism, 318 , E736–E741. 10.1152/ajpendo.00124
Nugraha R. V. Ridwansyah H. Ghozali M. Khairani A. F. Atik N. (2020). Traditional herbal medicine candidates as complementary treatments for COVID-19: A review of their mechanisms, pros and cons. Evidence-based Complementary and Alternative Medicine, 2020 , 1–12. 10.1155/2020/2560645
Önmez A. Gamsızkan Z. Özdemir Ş. Kesikbaş E. Gökosmanoğlu F. Torun S. Cinemre H. (2020). The effect of COVID-19 lockdown on glycemic control in patients with type 2 diabetes mellitus in Turkey. Diabetes and Metabolic Syndrome: Clinical Research and Reviews, 14 (6 ), 1963–1966. 10.1016/j.dsx.2020.10.007
Paudyal V. Cadogan C. Fialová D. Henman M. C. Hazen A. Okuyan B. Lutters M. Stewart D. (2022). Complementary and alternative medicines use in COVID-19: A global perspective on practice, policy and research. Research in Social and Administrative Pharmacy, 18 (8 ), 2524–2528. 10.1016/j.sapharm.2020.11.017 33992585
Ruiz-Roso M. B. Knott-Torcal C. Matilla-Escalante D. C. Garcimartín A. Sampedro-Nuñez M. A. Dávalos A. Marazuela M. (2020). Covid-19 lockdown and changes of the dietary pattern and physical activity habits in a cohort of patients with type 2 diabetes mellitus. Nutrients, 12 (8 ), 1–16. 10.3390/nu12082327
Santini Z. I. Nielsen L. Hinrichsen C. Meilstrup C. Madsen K. R. Koushede V. Sanitari Sant Joan Déu de P. Ziggi Ivan Santini D. Ivan Santini Z. Jose P. E. York Cornwell E. Koyanagi A. Nielsen L. Hinrichsen C. Meilstrup C. Madsen K. R. Koushede V. (2020). Social disconnectedness, perceived isolation, and symptoms of depression and anxiety among older Americans (NSHAP): A longitudinal mediation analysis. Articles Lancet Public Health, 5 , e62–e70.
Shi C. Zhu H. Liu J. Zhou J. Tang W. (2020). Barriers to self-management of type 2 diabetes during covid-19 medical isolation: A qualitative study. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy, 13 , 3713–3725. 10.2147/DMSO.S268481 33116721
Singh A. K. Gupta R. Ghosh A. Misra A. (2020). Diabetes in COVID-19: Prevalence, pathophysiology, prognosis and practical considerations. Diabetes and Metabolic Syndrome: Clinical Research and Reviews, 14 (4 ), 303–310. 10.1016/j.dsx.2020.04.004
Townsend L. Dyer A. H. Jones K. Dunne J. Mooney A. Gaffney F. O’Connor L. Leavy D. O’Brien K. Dowds J. Sugrue J. A. Hopkins D. Martin-Loeches I. Ni Cheallaigh C. Nadarajan P. McLaughlin A. M. Bourke N. M. Bergin C. O’Farrelly C. Conlon N. (2020). Persistent fatigue following SARS-CoV-2 infection is common and independent of severity of initial infection. PLoS One, 15 (11 ), e0240784. 10.1371/journal.pone.0240784
Verma A. Rajput R. Verma S. Balania V. K. B. Jangra B. (2020). Impact of lockdown in COVID 19 on glycemic control in patients with type 1 diabetes mellitus. Diabetes and Metabolic Syndrome: Clinical Research and Reviews, 14 (5 ), 1213–1216. 10.1016/j.dsx.2020.07.016
Wang D. Hu B. Hu C. Zhu F. Liu X. Zhang J. Wang B. Xiang H. Cheng Z. Xiong Y. Zhao Y. Li Y. Wang X. Peng Z. (2020). Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China. Journal of the American Medical Association, 323 (11 ), 1061–1069. 10.1001/jama.2020.1585 32031570
Wu Q. Zhou L. Sun X. Yan Z. Hu C. Wu J. Xu L. Li X. Liu H. Yin P. Li K. Zhao J. Li Y. Wang X. Li Y. Zhang Q. Xu G. Chen H. (2017). Altered lipid metabolism in recovered SARS patients twelve years after infection. Scientific Reports, 7 (1 ), 9110. 10.1038/s41598-017-09536-z 28831119
Xue T. Li Q. Zhang Q. Lin W. Wen J. Li L. Chen G. (2020). Blood glucose levels in elderly subjects with type 2 diabetes during COVID-19 outbreak: A retrospective study in a single center. Medrxiv.
Zhang L. Liu Y. (2020). Potential interventions for novel coronavirus in China: A systematic review. Journal of Medical Virology, 92 (5 ), 479–490. 10.1002/jmv.25707 32052466
|
PMC010xxxxxx/PMC10232721.txt |
==== Front
Int J Biol Macromol
Int J Biol Macromol
International Journal of Biological Macromolecules
0141-8130
1879-0003
Elsevier B.V.
S0141-8130(23)02080-9
10.1016/j.ijbiomac.2023.125186
125186
Review
Lateral flow immunoassays for antigens, antibodies and haptens detection
Li Ge a
Li Qingmei b
Wang Xun c
Liu Xiao d
Zhang Yuhang c
Li Rui b
Guo Junqing b⁎
Zhang Gaiping abce⁎⁎
a College of Veterinary Medicine, Northwest Agriculture and Forestry University, Yangling 712100, China
b Key Laboratory of Animal Immunology, Henan Academy of Agricultural Sciences, Zhengzhou 450002, China
c College of Veterinary Medicine, Henan Agricultural University, Zhengzhou 450002, China
d Henan Medical College, Zhengzhou 451191, China
e Jiangsu Co-Innovation Center for the Prevention and Control of Important Animal Infectious Disease and Zoonoses, Yangzhou University, Yangzhou 225009, China
⁎ Corresponding author.
⁎⁎ Correspondence to: G. Zhang, Key Laboratory of Animal Immunology of the Ministry of Agriculture, Henan Provincial Key Laboratory of Animal Immunology, Henan Academy of Agricultural Sciences, Zhengzhou, Henan, China.
1 6 2023
1 7 2023
1 6 2023
242 125186125186
2 3 2023
8 5 2023
30 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Lateral flow immunoassay (LFIA) is widely used as a rapid point-of-care testing (POCT) technique in food safety, veterinary and clinical detection on account of the accessible, fast and low-cost characteristics. After the outbreak of the coronavirus disease 2019 (COVID-19), different types of LFIAs have attracted considerable interest because of their ability of providing immediate diagnosis directly to users, thereby effectively controlling the outbreak. Based on the introduction of the principles and key components of LFIAs, this review focuses on the major detection formats of LFIAs for antigens, antibodies and haptens. With the rapid innovation of detection technologies, new trends of novel labels, multiplex and digital assays are increasingly integrated with LFIAs. Therefore, this review will also introduce the development of new trends of LFIAs as well as its future perspectives.
Keywords
Lateral flow immunoassays
Antigens
Antibodies
Haptens
Detection format
New trends
==== Body
pmc1 Introduction
Lateral flow immunoassays (LFIAs) develop on the basis of monoclonal antibody (mAb) technologies, immunochromatography technologies, new materials and labeling technologies [1,2]. LFIAs can realize qualitative and semi-quantitative detection of various analytes such as antigens, antibodies and haptens without professional skills and expensive instruments [[3], [4], [5], [6]]. The LFIA is one of the ideal immune rapid detection technologies, which is widely used in rapid detection of hormones [7], pathogenic microorganisms [8,9], veterinary drugs [8], pesticides [10], biotoxins and other targets [11,12]. LFIAs show broad application prospects in on-site real-time detection, which are especially suitable for hospitals, veterinary clinics, farms, dairies and other fields [[13], [14], [15]].
LFIAs were derived from the latex agglutination test established by Plotz and Singer [16], it is in the same period of time as radio-immunoassay and enzyme immunoassay. The early development and application of LFIA was the determination of human chorionic gonadotropin in pregnant women urine and serum/plasma in the late 1980s. LFIAs achieved the long-sought standard of “assurance” in diagnostic technology (affordable, sensitive, specific, user-friendly, fast and robust, no equipment required), which significantly promoted the development of immune diagnostic technology [17].
According to the different detection formats, LFIAs can be divided into direct detection and competitive detection. The direct detection can be used for the detection of antigens (such as human or animal proteins and pathogenic microorganism proteins, etc.) and antibodies (such as immunoglobulins). It is mostly applied in the early diagnosis of human and animal diseases, biochemical analysis and antibody titer monitoring. The competitive detection is a competitive inhibitory immunological binding reaction and mainly used for the detection of small molecule compounds with few antigenic sites or only a single antigenic site. These small molecule compounds are usually reactogenic but not immunogenic, they are also called haptens. Therefore, competitive detection is of great convenience in testing residues of pesticide, veterinary drug or mycotoxin.
With the rapid innovation of diagnostics, extensive efforts have been invested into carrying out novel labels to improve the sensitivity, multiplex detection to realize simultaneous detection of multiple targets and digital detection to realize visual reading and quantitative detection [18,19]. Hence, new trends of new labels, multiplex assays and digital assays are introduced. Finally, the future perspectives of LFIAs are also discussed.
2 Principles and components of the LFIAs
The test strip is the commonest form of LFIA, it consists of four basic structures: sample pad, conjugate pad, nitrocellulose (NC) membrane and absorption pad [1,20], which are stacked on a support plate in sequence from the test end to the handle end (Fig. 1 ). The sample pad absorbs the sample solution and makes it flow laterally to the conjugate pad by capillary force. The conjugate pad is labeled with bioactive materials (such as colloidal gold-labeled antibodies), which can bind to the target from sample solution to generate immunological complexes. The NC membrane intercepts labeled immune complexes, and visually displays the result. Two or more different biologically active materials (such as antigens or antibodies) are immobilized to form a “test line” (T line) and a “control line” (C line) on the NC membrane. The absorption pad absorbs the sample solution flowing through the strip, which maintains the pressure difference between the two ends and promotes more sample solution to laterally flow on the NC membrane.Fig. 1 Schematic diagram of a conventional LFIA structure.
Fig. 1
3 LFIAs for antigens detection
Sandwich LFIA based on the principle of two different antibodies bind to antigen simultaneously is mainly detecting antigen with multiple antigenic sites, such as pathogenic bacteria, viruses and proteins. Most natural antigen molecules have complex structures with multiple antigenic epitopes on the surface. They can bind to different antibodies simultaneously and be used as detection targets. There are three LFIA formats for antigen detection.
3.1 MAb labeling-polyclonal antibody capture format
The mAb labeling-polyclonal antibody (pAb) capture format is to prepare the conjugate pad by labeling the mAb that specifically recognizes the analyte with nanomaterial. PAb works as a capture antibody for immunological complexes, anti-mouse IgG antibody or Staphylococcal protein A (SPA) has the ability to bind immunoglobulin and work as the control antibody. They are coated on the NC membrane as the T line and the C line, respectively (Fig. 2a). During detection, the analyte in the sample is combined with the labeled antibody to form the immunological complexes, which are specifically recognized by the capture antibody, presenting a colored band at the T line. The excess labeled antibody and part of the immunological complexes are recognized by the control antibody, forming the C line, which gives a positive result. In addition, the color intensity is correlated with antigen concentration.Fig. 2 Antigen detection format diagram of LFIAs. a: mAb labeling-pAb capture format; b: mAb labeling-mAb capture format; c: combined mAb labeling-mAb capure format.
Fig. 2
An immunochromatographic strip for the detection of avian avulavirus 1 (Newcastle disease virus, NDV) based on one anti-NDV mAb and one pAb was established by Li et al. in Key Laboratory of Animal Immunology (KLAI) of Henan Academy of Agricultural Sciences (HAAS) [21]. The strip can detect NDV with a detection limit of 104.9 EID50 viruses/0.1 mL in the NDV infected sample, whose performance is as good as hemagglutinin (HA) test (Fig. 3a). This strip can detect NDV in infected tissue as early as 36 h, before clinical symptoms and gross anatomical lesions.Fig. 3 LFIA for antigens detection developed in KLAI of HAAS. a: The LFIA for NDV antigen detection [21]; b: The LFIA for SARS-CoV-2 antigen detection [22]; c: The LFIA for H7 subtype avain influenza virus detection [23]; d: The LFIA for H3 subtype influenza virus detection with high sensitivity [24].
Fig. 3
3.2 MAb labeling-mAb capture format
The mAb recognizes specific epitope of antigen coated on the NC membrane can effectively improve the specificity of the detection. Meanwhile, the use of mAbs can overcome the bias between batches of pAbs, thus it is suitable for standardized production of antigen test strips (Fig. 2b). Li et al. had established an immunochromatographic strip for the detection of the spike protein of SARS-CoV-2 [22]. This strip can detect SARS-CoV-2 spike protein in subunit vaccine with a detection limit of 62.5 ng/mL within 15–30 min (Fig. 3b). This strip monitors vaccine quality by detecting the antigen content of spike protein, providing technical support for the early diagnosis of the outbreak of COVID-19.
Li et al. [23] had developed an immunochromatographic strip based on two mAbs against HA protein for rapid detection of H7 subtype avain influenza viruses. The detection limit of the strip was 2.4 log10EID50/0.1 mL for chicken swab samples, which provided an effective approach for the rapid and early detection of H7 subtype avain influenza viruses (Fig. 3c). Liu et al. [24] has developed a strip for the optical determination of influenza virus H3 subtype. The strip utilized gold nanoparticles (AuNP)-coated polystyrene latex microspheres (PS) based on two mAbs against HA protein of H3. It showed a detection limit of 0.016 hemagglutination unit, assisting early determination of influenza virus infection (Fig. 3d).
3.3 Combined mAb labeling-mAb capture format
For pathogenic microorganisms with large antigenic epitopes, the variation of epitopes may lead to evasion of pathogenic microorganisms by the above two detection formats. The combination mAbs that recognize different antigenic sites are labeled with colloidal gold. Its efficiency of epitope recognition greatly improves the sensitivity of the test (Fig. 2c).
4 LFIAs for antibodies detection
Indirect LFIA is used to capture antibodies such as IgG, IgM and IgA. IgM antibodies are commonly used as early detection of infection; IgG antibodies are for the monitoring of immune antibody titers or the detection of differential diagnosis marker antibodies; IgA antibodies are applied to evaluation of mucosal immunity. There are four formats of LFIAs for antibodies detection.
4.1 Secondary antibody labeling-antigen capture format
The conjugate pad is prepared by labeling anti-species IgG antibody or bacterial immunoglobulin binding protein A/G with nanomaterial. The specific antigen works as a capture antigen and protein A/G or IgG antibody works as a control antibody, which are coated on the NC membrane as T line and C line, respectively (Fig. 4a). The labeled antibody can bind to all immunoglobulins in blood or serum samples, non-specific antibodies with immunoglobulins significantly affect the generation of specific immunological complexes, thus affecting the results. In particular, low sensitivity to trace specific antibody often cause false negative results. A gold nanoparticle strip discriminating Foot-and-Mouth disease virus (FMDV) vaccinated animals from infected animals was established [25]. SPA conjugated with colloidal gold nanoparticles were used as a probe. Two proteins of FMDV were coated as test line 1 and test line 2 (T1 and T2) and the goat anti-pig antibody IgG was coated as a control line, as shown in Fig. 5a. The strip showed the specificity of T1 and T2 were 95.17 % and 100 %, which was in accordance with commercial ELISA kits. This strip enables monitoring titers of O-type FMDV antibody on site.Fig. 4 Antibodies detection format diagram of LFIAs. a: Secondary antibody labeling-antigen capture format; b: Antigen labeling-secondary antibody capture format; c: Antigen sandwich format; d: Antibody blocking format.
Fig. 4
Fig. 5 LFIAs for antibodies detection. a: The LFIA for FMDV antibody detection; b: The LFIA for African swine fever virus (ASFV) antibody detection; c and d: LFIAs for Classical swine fever virus (CSFV) antibody detection; e: The LFIA for Porcine reproductive and respiratory syndrome virus (PRRSV) antibody detection; f: The LFIA for Pseudorabies virus (PRV) antibody detection.
Fig. 5
4.2 Antigen labeling-secondary antibody capture format
Nanomaterial labeled antigen as a probe can interact with analyte in the sample to form immunological complexes, which captured by anti-species IgG antibody or protein A/G on T line, and then react with analyte-specific antibody on C line (Fig. 4b). Antigens as probe reduce the interference of non-specific immunoglobulins and improve the sensitivity. Non-specific antibodies have chances to bind to the secondary antibody or protein A/G, which affects the generation efficiency of specific complexes. The affinity of the capture antibody to immunoglobulin is much higher than that of labeled antibody (about 10 times), thus reducing the non-specific bindings. Meanwhile, the diluted serum with less interference of non-specific antibody boost sensitivity. Varieties of strips aiming at detection of diseases, particularly swine diseases have been established. (Table 1 and Fig. 5).Table 1 Antibody detection strips for different diseases.
Table 1Diseases Antigens Expression systems of antigens Test line Control line Detection limit References
African swine fever virus P72 protein HEK 293 SPA Anti-p72 pAbs 1: 409,600 [26]
Classical swine fever virus E2 protein Bac-to-Bac baculovirus SPA Rabbit anti-E2 pAbs 1: 102,400 [27]
Classical swine fever virus E2 protein Transgenic rice endosperm SPA Anti IgG 1:128,000 [28]
Porcine reproductive and respiratory syndrome virus Nsp7 protein Escherichia coli SPA Anti-Nsp7 antibodies 1: 3200 [29]
Pseudorabies virus GB protein Escherichia coli SPA Swine anti-PRV antibody IgG 96 % compared with commercial ELISA [30]
4.3 Antigen sandwich format
Most of the antibodies are multivalent antibodies (IgG, IgE and serum IgA are bivalent, and IgM is tenvalent), which can form antigen-specific antibody-labeled antigen complexes at T line. Nanomaterial labeled one antigen as a probe can interact with analyte in the sample to form immunological complexes, which captured by the other antigen on T line, and then reacted with antigen-specific antibody on C line (Fig. 4c). The color intensity is positively correlated with the antibody level. The double-antigen sandwich format eliminates the interference of non-specific immunoglobulins on the formation of labeled antigen-antibody complexes. When the tested serum contains a small amount of target antibody, the excess labeled antigen may block the binding site of the specific antibody resulting in false negative results.
4.4 Antibody blocking format
Antibody blocking format enables detection of specific antigenic epitopes or differential diagnosis of labeled antibodies through high-affinity mAbs. The blocking test strip established on the base of neutralizing mAbs. It detects the neutralizing antibody and evaluates the titer of the neutralizing antibody. Also, the antibody blocking format can effectively exclude the interference of non-specific antibodies. Nanomaterial-labeled antigens or neutralizing antibodies are key components of this test. PAb or mAb plays a role as blocking antibody immobilized on T line, and anti-mouse IgG antibody is coated on C line (Fig. 4d). An optimal concentration of antigen is critical to antibody diluent. When the sample does not contain analyte-specific antibodies, the antigens in the diluent bind to labeled antibodies from the conjugate pad and then react with capture antibody on T line, and those of rest intercept on control line, representing a negative result. When the sample contains the analyte-specific antibodies, antigen-antibody binding occurs in the first step, and terminates the following reaction, only C line displays, indicating a positive result. The presence of specific antibodies can significantly inhibit or completely block the formation of antigen-labeled antibody complexes, which cannot be effectively captured by the detection antibodies at the test blot, leading to the T line significantly weakened or completely disappeared. The labeled antibody, however, can be capture by the control antibody at C line, which represents a positive result. The color intensity of the detection line is negatively correlated with the antibody level.
Ma [31] has developed an immunochromatographic strip for the detection of NDV antibody, which can distinguish vaccine-immunized animals from naturally infected animals within 10 min. The recombinant HN or F protein was adsorpted in an antigen pad or pre-incubation with serum. Then the anti-HN or anti-F mAb was labeled with colloidal gold as the detector. A chicken anti-NDV pAb and SPA were used as the T and C line, respectively. The colored C line without T line indicates a positive result, while both colored T and C lines indicates a negative result.
5 LFIAs for haptens detection
The haptens detection test strip mainly detects small molecule compounds of single antigenic epitope. The small molecule compounds are non-immunogenic or poorly immunogenic with small molecular weights. The hapten test strips are widely applied to the determination of antibiotics, pesticides, veterinary drugs, toxins, banned additives, drugs, hormones and heavy metals, etc. There are two formats of LFIAs for haptens detection.
5.1 Antibody labeling format
In the antibody labeling format, the conjugate pad is prepared by labeling the mAb that specifically recognizes the target antigen with nanomaterial. The target antigen coupled to a carrier protein (artificial antigen) is used as the capture antigen on T line, and the anti-species IgG antibody coated as C line (Fig. 6a). Sample without target leads to conjugated antibody captured by antigen on T line and then C line, indicating a negative result. Sample with target can be specifically recognized by conjugated antibody, thereby blocking the binding of the conjugated antibody to the capture antigen on T line. The presence of the target antigens interfere with blocking of antibody captured or completely blocked, significantly reducing the color appearing on T line. The unblocked labeled-antibody is recognized by C line, indicating a positive result. There were a series of test strips for the detection of haptens developed in KLAI of HAAS (Table 2 ).Fig. 6 Haptens detection formats diagram of LFIAs. a: Antibody labeling format; b: Antigen labeling format.
Fig. 6
Table 2 Examples of haptens detection strips developed in the past five years.
Table 2Detection targets Samples IC50 values Detection limits Developed time References
Xylazine Milk 0.590 ng/mL 0.100 ng/mL 2022 [32]
Imidocarb Milk 0.400 ng/mL 0.078 ng/mL 2021 [33]
Diminazene Milk 5.200 ng/mL 1.200 ng/mL 2020 [34]
Bacitracin zinc Milk 3.160 ng/mL 0.820 ng/mL 2020 [35]
Nitroxynil Milk 5.716 ng/mL 1.146 ng/mL 2020 [36]
Danofloxacin Milk 0.513 ng/mL 0.092 ng/mL 2019 [37]
Antibiotics Milk 0.040 ng/mL 0.216 pg/mL 2018 [38]
Oseltamivir phosphate Egg and chicken meat 2.56 and 2.63 μg/kg 0.43 and 0.42 μg/kg 2018 [39]
5.2 Antigen labeling format
In the antigen labeling format, the conjugate pad is prepared by labeling the target antigen (artificial antigen) with nanomaterial. The mAb specific for the target antigen is used as the capture antibody on T line, and the anti-carrier protein antibody immobilized as C line (Fig. 6b). Sample without target leads to conjugated antigen captured by antibody on T line and then C line, indicating a negative result. Sample with target can compete with the conjugated antigen to bind the detection antibody on T line making the antibody unable to be captured or completely blocked, reducing the color appearing on T line. The unblocked conjugated-antigen is captured by C line, which represents a positive result.
6 New trends of LFIAs
6.1 LFIAs for hyper sensitivity
With the continuous development of material technologies, more and more nanoparticles have been developed and applied to LFIAs to improve the sensitivity of conventional test strip. Nanomaterials have shown the great potential in the labeling of antigens and antibodies owing to the special structure level, strong adsorption capacity, good orientation performance, biocompatibility and structural compatibility [40]. Nanomaterials have been used in developing LFIA for various kind, including gold nanoparticles (GNPs), carbon nanoparticles (CNPs), colloidal selenium nanoparticles (SNPs), quantum dots (QDs), fluorescent microspheres and magnetic nanoparticles (MNPs). The advantages and disadvantages of different nanoparticles used in LFIA are shown in Table 3 .Table 3 Advantages and disadvantages of different nanoparticles used in LFIAs.
Table 3Nanoparticles Advantages Disadvantages References
Gold nanoparticles Easy to prepare and functionalize, good biocompatibility, low cost, easy-to-read results, detectable with the naked eye Low sensitivity, requires reader for quantification [[41], [42], [43]]
Carbon nanoparticles High signal-to-noise ratio, stable, functional, non-toxic Qualitative or semi-quantitative, background interference [44,45]
Colloidal selenium nanoparticles Rust-colored, easy to prepare and handle Difficult to discern the results with the naked eye [46]
Quantum dots High optical stability, wide absorption spectrum, narrow emission band, strong stability and high sensitivity Easily quenched of fluorescence, toxic, high fluorescence background, require a fluorescence reader for quantification [[47], [48], [49], [50], [51]]
Up conversion nanoparticles Long fluorescence lifetime, easy functionalization, stable, tunable emission color, no background fluorescence interference, high sensitivity require a fluorescence reader for quantification [[52], [53], [54]]
Magnetic nanoparticles Easy surface functionalization, magnetic properties, long signal duration and low background noise Require a magnetic readeror magnetic field sensor [55,56]
GNPs are in a stable colloidal state formed by electrostatic action and the polymerization of chloroauric acid under the reduction of trisodium citrate into gold particles of a specific size. GNP has a diameter of 1–150 nm and is purple color. Colloidal gold is widely used as LFIA-labeled probe on account of the characteristics of bright color, easy preparation, good biocompatibility, high stability, chemical traceability, and fine optical properties [57]. GNPs are a class of colored particle markers that can realize qualitative or semi-quantitative detection, with particle sizes ranging from 10 to 100 nm. The high stability, easy preparation, non-toxicity, easy conjugation and no activation characteristics of GNPs make them widely used in LFIAs [58]. Colloidal selenium is a nanoscale selenium particle with surface effect and small size effect obtained by adjusting the reaction conditions. It is widely used in LFIAs with characteristics of low cost, simple labeling with proteins, uneasy to coagulate after labeling, and easy adjustment of particle size range. Qds are spherical or quasi-spherical fluorescent semiconductor nanocrystals with a particle size ranging from 2 to 20 nm. QDs are considered to be the most potential biomarkers in LFIA, because of excellent fluorescence properties such as strong photochemical stability, long fluorescence lifetime, narrow and symmetrical emission spectrum [59]. Up conversion nanoparticles (UCPs) are fluorescent substances consisting of a host matrix, absorbers and emitters. UCPs are newly developed fluorescent probes that excited by low-energy near-infrared light and emits high-energy visible light. UCPs are ideal fluorescent nanomaterials with little damage to biological tissues and strong penetrating ability [60]. MNPs refer to nano-sized magnetic materials (represented by Fe3O4), which show satisfactory biocompatibility and magnetic orientation. MNPs are easy to prepare to couple with antibodies coated on immunomagnetic beads [61].
6.2 LFIAs for high-throughput detection
Typical LFIAs are suitable for point-of-care tests due to their portability, simplicity, cost-effectiveness, and rapid detection of target biomarkers [62]. However, detecting a single biomarker in typical LFIAs is not conducive to high-throughput diagnosis. Consequently, multi-biomarkers detection of LFIAs have been extensively studied in recent years [63]. Three typical structures are detection of multi-targets in a single-strip, a dual-strip with multi-targets, and microarray, integration of lateral flow with microanalysis. (Fig. 7 ) [64].Fig. 7 Schematic of three multiplex LFIAs.
Fig. 7
The popular strategy for the LFIA multiplexing is to design several T lines or dots on the immunochromatographic strip by using gold nanoparticles (GNPs), quantum dots, or colored/fluorescent microspheres as carriers. Fluorescent nanoparticles require additional equipment for excitation, thus colored nanoparticles visible to naked eyes have been more extensively researched [[65], [66], [67], [68]]. Several individual strips can be combined into a special strip for samples to be collected only once and distributed to each strip in parallel flow. The advantages of this method rely on the large multiplexing capability and the absence of mutual interference between analytes that occur independently on a single strip [69]. Since a single test strip or a multi-test strip limited by quantitative capability and diagnostic validity, the combination of lateral flow and protein chip technology has been developed to achieve rapid and accurate detection of diagnosis of infectious diseases [70].
6.3 LFIAs for digital detection
Conventional LFIAs are convenient yet unable to achieve quantitative detection, and the bias in visual readings limit the application of LFIAs [71]. Thus, LFIAs with integrated optical readers to convert visual signals into more precise quantitative results is an acceptable approach. Instrument-based optical detection and electrochemical detection in conjunction with LFIAs are discussed.
Combining optical readers with LFIA make results to be easily obtained, especially in fields of genotyping and healthcare [72]. The imaging system accommodates a wider range of labeling systems with various color or optical properties, providing a more versatile instrument platform for LFIA. It is performed in conjunction with LFIAs by placing the test strip in a controlled lighting environment, where a complementary metal-oxide-semiconductor camera or high-density charge-coupled device captures images of the test and control lines [73]. Hundreds of millions of mobile phone users worldwide make mobile an attractive distributed platform for the future diagnostic market. The mobile phone camera based on CMOS imaging sensor can provide high image quality. Optical reader added up to LFIA will greatly improve its efficiency and convenience.
7 Outlooks
LFIAs based on three conventional systems of antigen, antibody and hapten have flourished since the establishment. They are extensively available in diagnosis, prognosis, screening, and monitoring of diseases in veterinary industry and human healthcare in hospital wards, clinics, health centers and self-diagnosis. IgM or IgG antibody test strips known to world since the outbreak of the COVID-19 in 2019. In order to monitor the SARS-CoV-2 and achieve effective control of the epidemic, antigen test strips based on antigen-antibody responses have been entrusted with detecting throat swabs or nasal swabs. However, a series of mutant viruses with stronger transmission of COVID have been derived, including the Alpha, Beta, Delta and Omicron variants.
The demands for quantitative LFIAs increase as the LFIA market expands [74]. New trends of novel labels, multiplex and digital assays are increasingly integrated with LFIAs with the rapid innovation of diagnostic technologies. Through the reader, nanoparticle tags captured in the test area are excited by external physical stimuli, such as lasers, electrical potentials or magnetic fields, resulting in amplified signals [75]. LFIAs are the backbone of rapid point-of-care diagnostics, with the potential to enable early case management and change in the epidemiology of infectious diseases. Emergent requirements from users for multiplexed systems capable of detecting multiple biomarkers simultaneously due to the complexity, multiple symptoms, and multiple infectious states of human disease. Imaging systems are also used for quantitative analysis of LFIAs. Cho I [76] used confocal Raman imaging combined with silver-enhanced technology to detect influenza B virus with sensitivity 1000 times higher than that of ordinary method. Mobile telephony (cell phone) healthcare is an emerging field as a next-generation point-of-care diagnostic platform.
8 Conclusions
The unique and remarkable properties of LFIAs have contributed to the detection of disease biomarkers and infectious agents in the fields of drugs, food, agriculture and environmental safety through three major detection systems. The principles of LFIAs have remained for decades, while the updates continue to occur. The sensitivity and reproducibility of LFIAs are promoted through novel labels and digitized devices, multiplex. Moreover, LFIAs effectively performed outside the laboratory with continued refinement. They offer the point-of-care diagnostics for developing countries both on site and in clinical settings.
Funding
This work was supported by the Science and Technology Development Project of Henan Province (212102110091, 222102110453), and Science and Technology Innovation Team of Henan Academy of Agricultural Sciences (2023TD03).
Ethics approval
This article does not contain any studies with human participants or animals performed by any of the authors.
CRediT authorship contribution statement
Ge Li: Writing – original draft. Qingmei Li: Supervision. Xun Wang: Data curation. Xiao Liu: Visualization, Investigation. Yuhang Zhang: Writing – review & editing. Rui Li: Supervision. Junqing Guo: Conceptualization, Methodology, Software. Gaiping Zhang: Writing – review & editing.
Declaration of competing interest
The authors declare no competing interests.
Data availability
Datasets that support the current study are available from the corresponding author, [RS], upon reasonable request.
==== Refs
References
1 Zhang G. Guo J. Wang X. Immunochromatographic lateral flow strip tests Methods Mol. Biol. 504 2009 169 183 10.1007/978-1-60327-569-9_12 19159098
2 Di Nardo F. Chiarello M. Cavalera S. Ten years of lateral flow immunoassay technique applications: trends, challenges and future perspectives Sensors (Basel) 21 15 2021 5185 10.3390/s21155185 34372422
3 Park J.H. Park E.K. Cho Y.K. Normalizing the optical signal enables robust assays with lateral flow biosensors Acs Omega 7 21 2022 17723 17731 10.1021/acsomega.2c00793 35664567
4 Ali Z. Sanchez E. Tehseen M. Bio-SCAN: a CRISPR/dCas9-based lateral flow assay for rapid, specific, and sensitive detection of SARS-CoV-2 ACS Synth. Biol. 11 1 2022 406 419 10.1021/acssynbio.1c00499 34939798
5 Karachaliou C.E. Koukouvinos G. Goustouridis D. Recent developments in the field of optical immunosensors focusing on a label-free, white light reflectance spectroscopy-based immunosensing platform Sensors (Basel) 22 14 2022 5114 10.3390/s22145114 35890791
6 Singh H. Singh S. Bhardwaj S.K. Development of carbon quantum dot-based lateral flow immunoassay for sensitive detection of aflatoxin M1 in milk Food Chem. 393 2022 133374 10.1016/j.foodchem.2022.133374
7 Shyu R.H. Shyu H.F. Liu H.W. Colloidal gold-based immunochromatographic assay for detection of ricin Toxicon. 40 3 2002 255 258 10.1016/s0041-0101(01)00193-3 11711121
8 Nielsen K. Yu W.L. Kelly L. Validation and field assessment of a rapid lateral flow assay for detection of bovine antibody to Anaplasma marginale J. Immunoass. Immunochem. 30 3 2009 313 321 10.1080/15321810903084749
9 Morales-Narvaez E. Naghdi T. Zor E. Photoluminescent lateral-flow immunoassay revealed by graphene oxide: highly sensitive paper-based pathogen detection Anal. Chem. 87 16 2015 8573 8577 26205473
10 Mei Z. Qu W. Deng Y. One-step signal amplified lateral flow strip biosensor for ultrasensitive and on-site detection of bisphenol A (BPA) in aqueous samples Biosens. Bioelectron. 49 2013 457 461 10.1016/j.bios.2013.06.006 23811478
11 Ngom B. Guo Y. Wang X. Development and application of lateral flow test strip technology for detection of infectious agents and chemical contaminants: a review Anal. Bioanal. Chem. 397 3 2010 1113 1135 20422164
12 Kuang H. Xing C. Hao C. Rapid and highly sensitive detection of lead ions in drinking water based on a strip immunosensor Sensors (Basel) 13 4 2013 4214 4224 23539028
13 Alhabbab R.Y. Lateral flow immunoassays for detecting viral infectious antigens and antibodies Micromachines (Basel). 13 11 2022 1901 10.3390/mi13111901 36363922
14 Moshe M. Daunt A. Flower B. SARS-CoV-2 lateral flow assays for possible use in national covid-19 seroprevalence surveys (react 2): diagnostic accuracy study Bmj. 372 2021 n423 10.1136/bmj.n423
15 Vashist S.K. Trends in multiplex immunoassays for in vitro diagnostics and point-of-care testing Diagnostics (Basel) 11 9 2021 1630 10.3390/diagnostics11091630 34573972
16 Plotz C.M. Singer J.M. The latex fixation test. I. Application to the serologic diagnosis of rheumatoid arthritis Am. J. Med. 21 6 1956 888 892 10.1016/0002-9343(56)90103-6 13372565
17 Boehringer H.R. O’Farrell B.J. Lateral flow assays in infectious disease diagnosis Clin. Chem. 68 1 2021 52 58 10.1093/clinchem/hvab194 34969115
18 Wang Y. He Z. Ablimit P. Development of multiplex cross displacement amplification combined with lateral flow biosensor assay for detection of virulent shigella sonnei Front. Cell. Infect. Microbiol. 12 2022 1012105 10.3389/fcimb.2022.1012105 36339345
19 Bishop J.D. Hsieh H.V. Gasperino D.J. Sensitivity enhancement in lateral flow assays: a systems perspective Lab Chip 19 15 2019 2486 2499 10.1039/c9lc00104b 31251312
20 Ching K.H. Lateral flow immunoassay Methods Mol. Biol. 1318 2015 127 137 10.1007/978-1-4939-2742-5_13 26160571
21 Li Q. Wang L. Sun Y. Evaluation of an immunochromatographic strip for detection of avian avulavirus 1 (Newcastle disease virus) J. Vet. Diagn. Investig. 31 3 2019 475 480 10.1177/1040638719837320 30973087
22 Li G. Wang A. Chen Y. Development of a colloidal gold-based Immunochromatographic strip for rapid detection of severe acute respiratory syndrome coronavirus 2 spike protein Front. Immunol. 12 2021 635677 10.3389/fimmu.2021.635677
23 Li G. Wang X. Li Q. Development of an immunochromatographic strip for rapid detection of H7 subtype avian influenza viruses Virol. J. 18 1 2021 68 10.1186/s12985-021-01537-9 33827632
24 Liu X. Yang J. Li Q. A strip test for the optical determination of influenza virus H3 subtype using gold nanoparticle coated polystyrene latex microspheres Mikrochim Acta. 187 5 2020 306 10.1007/s00604-020-04255-1 32356232
25 Yang S. Sun Y. Yang J. A gold nanoparticle strip for simultaneously evaluating FMDV immunized antibody level and discriminating FMDV vaccinated animals from infected animals RSC Adv. 9 52 2019 30164 30170 10.1039/c9ra04810c 35530212
26 Geng R. Sun Y. Li R. Development of a p72 trimer-based colloidal gold strip for detection of antibodies against African swine fever virus Appl. Microbiol. Biotechnol. 106 7 2022 2703 2714 10.1007/s00253-022-11851-z 35291024
27 Bai Y. Jia R. Wei Q. Development and application of a high-sensitivity immunochromatographic test strip for detecting classical swine fever virus antibodies Transbound. Emerg. Dis. 69 4 2022 e788 e798 10.1111/tbed.14367 34724351
28 Xu Q. Sun Y. Yang J. An improved immunochromatographic strip based on plant-derived E2 for detection of antibodies against classical swine fever virus Microbiol. Spectr. 2022 e105022 10.1128/spectrum.01050-22
29 Li H. Yang J. Bao D. Development of an immunochromatographic strip for detection of antibodies against porcine reproductive and respiratory syndrome virus J. Vet. Sci. 18 3 2017 307 316 10.4142/jvs.2017.18.3.307
30 Li X. Sun Y. Yang S. Development of an immunochromatographic strip for antibody detection of pseudorabies virus in swine J. Vet. Diagn. Investig. 27 6 2015 739 742 10.1007/s00604-020-04255-1 26472745
31 Ma F. Zhang E. Li Q. A plant-produced recombinant fusion protein-based Newcastle disease subunit vaccine and rapid differential diagnosis platform Vaccines (Basel) 8 1 2020 122 10.3390/vaccines8010122 32182813
32 Chen L. Hu X. Sun Y. An ultrasensitive monoclonal antibody-based lateral flow immunoassay for the rapid detection of xylazine in milk Food Chem. 383 2022 132293 10.1016/j.foodchem.2022.132293
33 Chen L. Hu X. Sun Y. Immunochromatographic assay based on high-affine monoclonal antibody for the detection of imidocarb in milk J. Food Sci. 86 8 2021 3413 3421 10.1111/1750-3841.15831 34268739
34 Chen L. Sun Y. Hu X. Colloidal gold-based immunochromatographic strip assay for the rapid detection of diminazene in milk Food Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 37 10 2020 1667 1677 10.1080/19440049.2020.1778185 32723010
35 Na G. Hu X. Yang J. Colloidal gold-based immunochromatographic strip assay for the rapid detection of bacitracin zinc in milk Food Chem. 327 2020 126879 10.1016/j.foodchem.2020.126879
36 Na G. Hu X. Yang J. A rapid colloidal gold-based immunochromatographic strip assay for monitoring nitroxynil in milk J. Sci. Food Agric. 100 5 2020 1860 1866 10.1002/jsfa.10074 31602650
37 Yang X. Wang Y. Yang J. An immunochromatographic lateral flow strip test for the rapid detection of danofloxacin in milk Food Anal. Methods 12 11 2019 2430 2437 10.1007/s12161-019-01601-9
38 Shi Q. Huang J. Sun Y. Utilization of a lateral flow colloidal gold immunoassay strip based on surface-enhanced Raman spectroscopy for ultrasensitive detection of antibiotics in milk Spectrochim. Acta A Mol. Biomol. Spectrosc. 197 2018 107 113 10.1016/j.saa.2017.11.045 29195715
39 Yang X. Yang J. Wang Y. A lateral flow Immunochromato-graphic strip test for rapid detection of oseltamivir phosphate in egg and chicken meat Sci. Rep. 8 1 2018 16680 10.1038/s41598-018-35080-5 30420605
40 Xue Z. Zhang Y. Yu W. Recent advances in aflatoxin B1 detection based on nanotechnology and nanomaterials-a review Anal. Chim. Acta 1069 2019 1 27 10.1016/j.aca.2019.04.032 31084735
41 Zhou S. Hu J. Chen X. Hydrazide-assisted directional antibody conjugation of gold nanoparticles to enhance immunochromatographic assay Anal. Chim. Acta 1168 2021 338623 10.1016/j.aca.2021.338623
42 Zhang L. Mazouzi Y. Salmain M. Antibody-gold nanoparticle bioconjugates for biosensors: synthesis, characterization and selected applications Biosens. Bioelectron. 165 2020 112370 10.1016/j.bios.2020.112370
43 Okyem S. Awotunde O. Ogunlusi T. High-affinity points of interaction on antibody allow synthesis of stable and highly functional antibody-gold nanoparticle conjugates Bioconjug. Chem. 32 8 2021 1753 1762 10.1021/acs.bioconjchem.1c00261 34228917
44 Xu L.D. Du F.L. Zhu J. Luminous silica colloids with carbon dot incorporation for sensitive immunochromatographic assay of Zika virus Analyst. 146 2 2021 706 713 10.1039/d0an02017f 33216074
45 Xu L.D. Zhu J. Ding S.N. Immunoassay of SARS-CoV-2 nucleocapsid proteins using novel red emission-enhanced carbon dot-based silica spheres Analyst. 146 16 2021 5055 5060 10.1039/d1an01010g 34282816
46 Wang Z. Zhou Q. Guo Y. Rapid detection of ractopamine and salbutamol in swine urine by immunochromatography based on selenium nanoparticles Int. J. Nanomedicine 16 2021 2059 2070 10.2147/IJN.S292648 33727813
47 Zhou J. Yang Q. Liang C. Detection of ochratoxin a by quantum dots-based fluorescent immunochromatographic assay Anal. Bioanal. Chem. 413 1 2021 183 192 10.1007/s00216-020-02990-1 33064163
48 Zhou J. Qian W. Yang Q. Analysis of virginiamycin M1 in swine feed, muscle and liver samples by quantum dots-based fluorescent immunochromatographic assay Food Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 39 8 2022 1390 1400 10.1080/19440049.2022.2081366 35679322
49 Niu Y. Zhang G. Zhou J. Differential diagnosis of the infection caused by wild-type or CD2v-deleted ASFV strains by quantum dots-based immunochromatographic assay Lett. Appl. Microbiol. 74 6 2022 1001 1007 10.1111/lam.13691 35255156
50 Chen P. Zhou M. Chen X. Quantum dot bead-based competitive immunochromatographic assay for enterotoxin aureus a detection in pasteurized milk J. Dairy Sci. 105 6 2022 4938 4945 10.3168/jds.2021-21568 35465994
51 Jahangir M.A. Gilani S.J. Muheem A. Quantum dots: next generation of smart nano-systems Pharm. Nanotechnol. 7 3 2019 234 245 10.2174/2211738507666190429113906 31486752
52 Lu X. Chen Y. Zou R. Novel immunochromatographic strip assay based on up-conversion nanoparticles for sensitive detection of imidacloprid in agricultural and environmental samples Environ. Sci. Pollut. Res. Int. 28 35 2021 49268 49277 10.1007/s11356-021-14143-7 33931813
53 Chen Y. Lu H. Shi H. An improved up-conversion nanoparticles-based immunochromatographic assay for rapid detection of zearalenone in cereals Food Chem. 412 2023 135555 10.1016/j.foodchem.2023.135555
54 Zhang S. Sun Y. Sun Y. Semiquantitative immunochromatographic colorimetric biosensor for the detection of dexamethasone based on up-conversion fluorescent nanoparticles Mikrochim Acta 187 8 2020 447 10.1007/s00604-020-04418-0 32676892
55 Yin H.Y. Li Y.T. Tsai W.C. An immunochromatographic assay utilizing magnetic nanoparticles to detect major peanut allergen Ara h 1 in processed foods Food Chem. 375 2022 131844 10.1016/j.foodchem.2021.131844
56 Lai X. Zhang G. Zeng L. Synthesis of PDA-mediated magnetic bimetallic Nanozyme and its application in immunochromatographic assay ACS Appl. Mater. Interfaces 13 1 2021 1413 1423 10.1021/acsami.0c17957 33346647
57 Goudarzi S. Ahmadi A. Farhadi M. Development of a new immunochromatographic assay using gold nanoparticles for screening of IgA deficiency Iran J. Allergy Asthma Immunol. 14 1 2015 105 112 doi:ijaai.tums.ac.ir/index.php/ijaai/article/view/405 25530146
58 Zhang X. Yu X. Wen K. Multiplex lateral flow immunoassays based on amorphous carbon nanoparticles for detecting three fusarium mycotoxins in maize J. Agric. Food Chem. 65 36 2017 8063 8071 10.1021/acs.jafc.7b02827 28825819
59 Chen H. Zhang X. Jin Z. Differential diagnosis of PRV-infected versus vaccinated pigs using a novel EuNPs-virus antigen probe-based blocking fluorescent lateral flow immunoassay Biosens. Bioelectron. 155 2020 112101 10.1016/j.bios.2020.112101
60 Lian Y. Ding L.J. Zhang W. Synthesis of highly stable cyanine-dye-doped silica nanoparticle for biological applications Methods Appl. Fluoresc. 6 3 2018 34002 10.1088/2050-6120/aab930
61 Huang X. Aguilar Z.P. Xu H. Membrane-based lateral flow immunochromatographic strip with nanoparticles as reporters for detection: a review Biosens. Bioelectron. 75 2016 166 180 10.1016/j.bios.2015.08.032 26318786
62 Huang L. Tian S. Zhao W. Multiplexed detection of biomarkers in lateral-flow immunoassays Analyst. 145 8 2020 2828 2840 10.1039/c9an02485a 32219225
63 Byzova N.A. Urusov A.E. Zherdev A.V. Multiplex highly sensitive immunochromatographic assay based on the use of nonprocessed antisera Anal. Bioanal. Chem. 410 7 2018 1903 1910 10.1007/s00216-018-0853-9 29353430
64 Kim H. Chung D.R. Kang M. A new point-of-care test for the diagnosis of infectious diseases based on multiplex lateral flow immunoassays Analyst. 144 8 2019 2460 2466 10.1039/c8an02295j 30849145
65 Peng J. Wang Y. Liu L. Multiplex lateral flow immunoassay for five antibiotics detection based on gold nanoparticle aggregations RSC Adv. 6 10 2016 7798 7805 10.1039/C5RA22583C
66 Wang Q. Liu Y. Wang M. A multiplex immunochromatographic test using gold nanoparticles for the rapid and simultaneous detection of four nitrofuran metabolites in fish samples Anal. Bioanal. Chem. 410 1 2018 223 233 10.1007/s00216-017-0714-y 29085985
67 Sun Y. Yang J. Yang S. Development of an immunochromatographic lateral flow strip for the simultaneous detection of aminoglycoside residues in milk RSC Adv. 8 17 2018 9580 9586 10.1039/c8ra01116h 35541843
68 Liu B. Gong H. Wang Y. A gold immunochromatographic assay for simultaneous detection of parathion and triazophos in agricultural products Anal. Methods 10 4 2018 422 428 10.1039/C7AY02481A
69 Carrio A. Sampedro C. Sanchez-Lopez J.L. Automated low-cost smartphone-based lateral flow saliva test reader for drugs-of-abuse detection Sensors (Basel) 15 11 2015 29569 29593 10.3390/s151129569 26610513
70 Chinnasamy T. Segerink L.I. Nystrand M. A lateral flow paper microarray for rapid allergy point of care diagnostics Analyst. 139 10 2014 2348 2354 10.1039/c3an01806g 24690935
71 Mak W.C. Beni V. Turner A. Lateral-flow technology: from visual to instrumental Trends Anal. Chem. 79 2016 297 305 10.1016/j.trac.2015.10.017
72 Chan C.P. Mak W.C. Cheung K.Y. Evidence-based point-of-care diagnostics: current status and emerging technologies Annu Rev Anal Chem (Palo Alto, Calif) 6 2013 191 211 10.1146/annurev-anchem-062012-092641 23527548
73 Tahir M.A. Dina N.E. Cheng H. Surface-enhanced Raman spectroscopy for bioanalysis and diagnosis Nanoscale 13 27 2021 11593 11634 10.1039/d1nr00708d 34231627
74 Kim J. Mohamed M. Zagorovsky K. State of diagnosing infectious pathogens using colloidal nanomaterials Biomaterials 146 2017 97 114 10.1016/j.biomaterials.2017.08.013 28898761
75 Xu Y. Liu Y. Wu Y. Fluorescent probe-based lateral flow assay for multiplex nucleic acid detection Anal. Chem. 86 12 2014 5611 5614 10.1021/ac5010458 24892496
76 Gozdzialski L. Rowley A. Borden S.A. Rapid and accurate etizolam detection using surface-enhanced Raman spectroscopy for community drug checking Int. J. Drug Policy 102 2022 103611 10.1016/j.drugpo.2022.103611
|
PMC010xxxxxx/PMC10232722.txt |
==== Front
Int J Biol Macromol
Int J Biol Macromol
International Journal of Biological Macromolecules
0141-8130
1879-0003
Elsevier B.V.
S0141-8130(23)02047-0
10.1016/j.ijbiomac.2023.125153
125153
Review
How helpful were molecular dynamics simulations in shaping our understanding of SARS-CoV-2 spike protein dynamics?
Abduljalil Jameel M. ab
Elghareib Ahmed M. c
Samir Ahmed c
Ezat Ahmed A. c
Elfiky Abdo A. c⁎
a Department of Biological Sciences, Faculty of Applied Sciences, Thamar University, Dhamar, Yemen
b Department of Botany and Microbiology, College of Science, Cairo University, Giza, Egypt
c Department of Biophysics, Faculty of Science, Cairo University, Giza, Egypt
⁎ Corresponding author.
1 6 2023
1 7 2023
1 6 2023
242 125153125153
7 2 2023
22 3 2023
27 5 2023
© 2023 Elsevier B.V. All rights reserved.
2023
Elsevier B.V.
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
The SARS-CoV-2 spike protein (S) represents an important viral component that is required for successful viral infection in humans owing to its essential role in recognition of and entry to host cells. The spike is also an appealing target for drug designers who develop vaccines and antivirals. This article is important as it summarizes how molecular simulations successfully shaped our understanding of spike conformational behavior and its role in viral infection. MD simulations found that the higher affinity of SARS-CoV-2-S to ACE2 is linked to its unique residues that add extra electrostatic and van der Waal interactions in comparison to the SARS-CoV S. This illustrates the spread potential of the pandemic SARS-CoV-2 relative to the epidemic SARS-CoV. Different mutations at the S-ACE2 interface, which is believed to increase the transmission of the new variants, affected the behavior and binding interactions in different simulations. The contributions of glycans to the opening of S were revealed via simulations. The immune evasion of S was linked to the spatial distribution of glycans. This help the virus to escape the immune system recognition. This article is important as it summarizes how molecular simulations successfully shaped our understanding of spike conformational behavior and its role in viral infection. This will pave the way to us preparing for the next pandemic as the computational tools are tailored to help fight new challenges.
Keywords
SARS-CoV-2
Spike
Glycosylation
ACE2
Structural bioinformatics
MD simulation
==== Body
pmc1 Introduction
Molecular simulations became a robust method for studying molecular systems [1]. Furthermore, biomolecular simulations solved different human-related health problems as the algorithms and hardware developed rapidly during the last two decades [1,2]. For example, the human immunodeficiency virus (HIV) and hepatitis C virus (HCV) were successful case studies that heavily relied on computer simulations to find potential inhibitors [[3], [4], [5], [6], [7]]. The pandemic SARS-CoV-2 represents the best recent example of utilizing molecular simulations to investigate human-health-related problems saving time, effort, money, and life [8]. Many studies were conducted on SARS-CoV-2 that use computational tools to find possible drug targets, repurpose old medicines or natural compounds previously suggested/tested against different viral proteins, and develop a vaccine against the virus [[9], [10], [11], [12], [13]].
Severe acute respiratory syndrome coronavirus (SARS-CoV) (2003) and SARS-CoV-2 (2019) are two pathogenic viruses of the β-coronaviruses (β-CoVs) lineage. CoVs are spherical-shaped, enveloped viruses with their genetic material as a positive-sense single-stranded RNA [14]. The S protein is the main component responsible for viral-host binding and fusion to the cell membrane (Fig. 1 ). Structurally, S is a homo-trimeric protein that decorates the SARS-CoV-2 surface and binds to the host cell receptor, angiotensin-converting enzyme 2 (ACE2), to facilitate viral entry [15]. Each monomer of the spike protein consists of a signal peptide (residues: 1–13) at the N-terminal domain (NTD), S1 subunit (14–685), and S2 subunit (686–1273). The subunits S1 and S2 are responsible for receptor binding and cell fusion, respectively [15,16]. S1 is composed of two domains, a receptor binding domain (RBD) and NTD.Fig. 1 The structure of SARS-CoV-2 spike protein in its trimeric form (PDB ID: 6VYB) [36]. The three monomers in the left panel are shown in the surface representation of different colors (green, magenta, and cyan). The carbohydrate moieties are shown in colored sticks. The right panel shows the top view (upper) and side view (bottom) of the cartoon representation of the same structure. The receptor binding domains (RBD) are shown with only one in the up configuration (chain B) and two in the down state. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 1
The binding of the virus to the human cell receptor, ACE2, is initiated by interactions between RBD and ACE2 (Fig. 2 ) [17]. The ACE2 peptidase domain (ACE2-PD) contains seven N-linked and one O-linked glycans at N53, N90, N103, N322, N432, N546, N690, and S155 positions [18]. On the other hand, the RBD of S contains the receptor-binding motif (RBM), a crucial functional segment as it contributes directly to the attachment to ACE2 and has physiochemical properties that affect the binding affinity [19]. The RBM consists of amino acids in the range T425:Q492 in SARS-CoV and S438:Q506 in SARS-CoV-2, with 16 and 17 residues for SARS-CoV and SARS-CoV-2 [20]. These motifs were reported to be within 4 Å of ACE2 [18]. It is well known that S of other coronaviruses undergoes a series of conformational changes during binding to ACE2, such as the slipping of S1 and refolding of S2 [21,22].Fig. 2 The structure of the complex between SARS-CoV-2 spike RBD (green) and the human receptor angiotensin-converting enzyme 2 (ACE2) (cyan). The left panel shows the cartoon representation, while the right panel shows the surface representation of the complex. The carbohydrate moieties are shown in magenta on both panels. The dashed-red circle indicates the cyclic region C480-C488. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2
The SARS-CoV-2 spike 3D structure was modeled, and its binding was predicted against different host-cell receptors [[23], [24], [25]]. Later on, in 2020, several solved structures of spike were deposited in the Protein Data Bank (PDB IDs: 6VXX, 7CAB, 6UL7, 6VYB, etc.), which shows that at least one monomer of SARS-CoV-2 spike RBD is in its “up” state [[26], [27], [28]]. Due to its pivotal role in the infection process, several studies targeted S with small molecules in an attempt to prevent the infection by blocking S-ACE2 binding [29,30].
Adem et al. studied the binding affinity of caffeic acid derivatives against SARS-CoV-2 proteins, including the spike. They used combined docking and dynamics simulations in their analysis [31]. Abosheasha et al. studied the potential use of antiplatelets against the viral spike and main protease using computational methods [32]. Basal et al. studied the Chaga medicinal mushroom Inonotus obliquus (Agaricomycetes) terpenoids against the spike using molecular docking [33]. Sarfraz et al. reviewed the studied polyphenol compounds that show potential therapeutic efficacy against SAR-CoV and SARS-CoV-2 proteins, including the spike protein. They reported Naringenin, Epigallocatechin gallate (EGCG), and Curcumin as potential interest for the SARS-CoV-2 spike [30]. Singh et al. studied plant-based bioactive compounds as inhibitors of the spike using computational methods. They found that Dicaffeoylquinic acid interacts strongly with the spike RBD at the interface with ACE2 [34]. Computational methods are also used in studying spike mutations and their interaction with ACE2 [35]. These methods include molecular docking, molecular dynamics simulation, and MM/GBSA approach.
In this review article, we will mine several studies that mainly utilized molecular simulation (at least 100 ns) to study the dynamics of the SARS-CoV-2 spike during the last two years to discuss how molecular simulations successfully elucidate the protein's behavior and function. Mutation effects, glycosylation types, and immune evasion have all been investigated by MD simulations as well as by experimental protocols. The contribution of mutations at S of different variants is also discussed based on the MD simulation findings. The current study sheds light on the effectiveness of using computational tools to fight against the COVID-19 pandemic, which can be improved to prepare for the next pandemic.
2 SARS-CoV versus SARS-CoV-2
2.1 SARS-CoV-2 versus SARS-CoV in binding ACE2
The trimeric S glycoprotein (∼150 kDa) comprises two subunits, S1 and S2, responsible for receptor binding and viral membrane fusion with host-cell membrane, respectively [37]. The work of Yan et al. revealed a similarity percentage of almost 80 % between the receptor binding domains (RBD) of SARS-CoV and SARS-CoV-2 upon binding to ACE2, with some differences in their structures at the interface [38]. Afterward, Zhang et al. presented a structure-based sequence alignment between SARS-CoV and SARS-CoV-2 RBDs and their differences in residues, with a shared sequence identity between RBDs and RBMs of both of them, is about 73.2 % and 50 % respectively [39,40]. The RBD structure of SARS-CoV-2 is more open and flexible and has a larger solvent-accessible surface area than SARS-CoV RBD. Several molecular dynamics simulations accompanied by experimental studies have been performed to understand the binding affinity differences between the two complexes. Viral infectivity and spread rates of viruses and how evolving mutations of SARS-CoV-2 variants affect its binding affinity to ACE2 have been studied [24,[41], [42], [43], [44]].
Yan et al. conducted one of these studies, which reported Cryo-electron microscopy (CryoEM) structures of the full-length human ACE2 bound with the RBD of the S protein of SARS-CoV and SARS-CoV-2 individually (PDB ID: 2AJF & 6 M17, respectively) [38]. Superimposition of the RBD in the two complexes showed a similarity with a root mean square deviation (RMSD) of 0.68 Å. There were many sequence variations at the interface with ACE2 (summarized in Table 1 ). They concluded that some of these variations might strengthen the binding affinity between SARS-CoV-2-RBD and ACE2, but others may decrease its binding affinity to ACE2 concerning SARS-CoV-RBD/ACE2 interaction. The alteration from Val404 to Lys317 may result in stronger interaction due to the formation of a salt bridge between Lys317 and Asp30 of ACE2. Moreover, due to the change of Leu472 in SARS-CoV-RBD to Phe486 in SARS-CoV-2-RBD, more hydrophobic interactions with Met82 of ACE2 have been generated which enhances the overall interaction by van der Waals forces [45]. On the other hand, the change from Arg426 to Asn439 has reduced the overall interaction by removing the salt bridge with Asp329 of ACE2 [38].Table 1 The structural and conformational interaction changes between SARS-CoV-RBD and SARS-CoV2-RBD upon the association interface with ACE2, according to Yan et al. 2020.
Table 1Changing position SARS-CoV-RBD SARS-CoV-2-RBD
N terminus of α1 Arg426 Asn439
Tyr484 Gln498
Thr487 Asn501
In the middle Val404 Lys417
Tyr442 Leu455
Leu443 Phe456
Phe460 Tyr473
Asn479 Gln493
C terminus of α1 Leu472 Phe486
Similar to most viral entry scenarios, S interacts with ACE2 via non-covalent interactions where residues vary widely in terms of their contributions to the attraction or repulsion of the partner protein [46]. Rodriguez used a fragment-based quantum-chemical method to evaluate different residues' attraction and repulsion contributions at ACE2 and S surfaces. This method depends on dividing the interacting surfaces of RBD into four-residue fragments called quartets. Two fragments on ACE2 (E37, N330, K353, Q42) and (E329, N330, K353, G354) are essential in binding to viral proteins as they promote intermolecular attraction for SARS-CoV-2 and SARS-CoV, respectively. The former quartet was also found to attract S of SARS-CoV, while (D30, K31, N33, and H34) residues show strong, attractive interactions toward SARS-CoV-2 and weak repulsive ones against SARS-CoV RBD [46].
2.2 Mutations impact on spike binding to ACE2
Mutations' effects on S binding to ACE2 was a hot topic of research that adopted different computational approaches, such as computational alanine mutagenesis, thermodynamic integration, and deep learning-based statistical models [[40], [41], [42], [43], [44]]. Monte Carlo and MD simulations have investigated the binding interactions between mutant SARS-CoV-2 (or SARS-CoV) and ACE2 [47]. Based on the CryoEM structure of S-ACE2 (PDB ID: 6M17), several mutations were constructed (V404K, R426N, Y442L, L443F, F460Y, L472F, N479Q, D480S, Y484Q, and T487N) and the electrostatic potential of each mutant structure was calculated using Adaptive Poisson−Boltzmann Solver (APBS). Enhanced electrostatic interactions were a major factor for the increased binding affinities of SARS-CoV-2 S compared to the spike of SARS-CoV [47]. Much of these electrostatic interactions come from the salt bridges between S and ACE2; R426:E329 in SARS-CoV and K417:D30 in SARS-CoV-2 by a factor of 1.54. The results also showed that these mutations of SARS-CoV-2 had caused sophisticated structural changes that further enhanced the electrostatic and van der Waals energies. Much of these electrostatic interactions come from the salt bridges between S and ACE2; R426:E329 in SARS-CoV and K417:D30 in SARS-CoV-2 [47].
The observations of stronger SARS-CoV-2 S binding to ACE2 are supported by the findings of Nguyen and associates who combined coarse-grained and all-atom steered molecular dynamics (SMD) simulations to evaluate the interaction pattern and KD values [48]. Coarse-grained simulations showed the KD value of the SARS-CoV-2 RBD - ACE2 complex three times less than that of the SARS-CoV RBD-ACE2. The SMD simulations revealed that a higher rupture force and nonequilibrium work (Wneq) are required to unbind SARS-CoV-2 RBD from ACE2 compared to that SARS-CoV. The calculated binding energies between the two complexes revealed that electrostatic interaction dominates the van der Waals interaction. These results are consistent with previous studies that inferred a significant contribution of electrostatic interactions to SARS-CoV-2/ACE2 binding [49].
Indeed, computational alanine scanning and the molecular mechanics-generalized Born surface area (MM/GBSA) method of ACE2 complex with either SARS-CoV RBD and SARS-CoV-2 RBD were employed to assess binding affinities and hotspot residues. The calculations also revealed that Q24, K31, H34, Y41, Y83, and K353 as the most important hotspot residues [42]. Hotspot residues are defined as amino acids that contribute with an increase in the difference between ∆G upon alanine mutatgenesis and ∆G of the wild type (∆∆Gtotal), with a minimum cutoff of −1.5 kcal/mol [50]. The calculated binding free energies showed that SARS-CoV-2 had higher affinity than SARS-CoV (−36.2 ± 1.1 versus −33.6 ± 1.6 kcal/mol, respectively) [49]. This increase in binding energy is likely due to more polar residues at the interface region of SARS-CoV-2 RBD (∼ 45 % more than in SARS-CoV) as reported by Delgado and colleagues [40]. Interestingly, a single mutation could account for a substantial difference in binding energy. For instance, an increase in the binding free energy by −4.3 kcal/mol was found upon the introduction of the N501Y mutation at SARS-CoV-2 (Fig. 3 ) [49]. This was explained by the enhanced van der Waals interaction energy between the Y501 of the mutant SARS-CoV-2 and ACE2. In addition, the thermodynamic integration method revealed that the total binding energy was increased by −3.1 kcal/mol due to this mutation. Experimental validation of these computational approaches was done by Wrapp et al., who showed that SARS-CoV-2 has an association affinity toward ACE2 10–20 times higher than SARS-CoV [26]. Furthermore, Walls et al. reported that SARS-CoV-2 has a dissociation constant (KD) against ACE2 about four times less than SARS-CoV [42]. It should be mentioned that the weaker binding affinity of SARS-CoV RBD may be attributed to its higher flexibility in comparison to SARS-CoV-2 RBD as evidenced by higher root-mean-square of fluctuations (RMSF) experienced by SARS-CoV RBD [48].Fig. 3 Important mutaations on the S of SARS-CoV-2. Based on the full-length model made publicly available by Casalino et al. [60] under Creative Commons licenses.
Fig. 3
Furthermore, Chen et al. have conducted a new deep learning method called “TopNetTree” to analyze the existing RBD mutations' impacts on the spike protein's binding free energies toward the ACE2. They systematically screened 3686 possible future mutations on all 194 residues of the RBD and then classified them according to the possibility of existence. In addition, the infectivity of SARS-CoV-2 mutations was analyzed and compared to SARS-CoV. They found that most of the mutations had small positive changes in the binding free energy, while just a small number had significant changes. In addition, by combining sequence alignment, binding free energy analysis, and probability estimation, they have found that 452, 489, 500, 501, and 505 residues of SARS-CoV-2 RBM has a high probability of mutating into more infectious strains [51].
2.3 Binding of new variants to ACE2
As in other RNA viruses, SARS-CoV-2 underwent several mutations, and variants are being reported from different regions globally. The new variants (alpha, beta, gamma, delta, and omicron) exhibit higher infectivity, disease severity, and mortality rates [[52], [53], [54], [55], [56], [57], [58]]. Several mutations in these variants affect the RBD binding to the human receptor ACE2. The mutations can be in the RBD or the NTD of the spike protein. Using computational methods, this section inspects why some variants bind more strongly to ACE2 than others.
Three mutations, N439K, S477N, and T478K, were reported to have higher binding affinity and infectivity rates (Fig. 3). Using MD simulation, protein-protein docking, and binding free energy calculations, the change in the binding interactions between SARS-CoV-2 RDB and ACE2 receptor was investigated for these mutations. First, mutants and wild-type S were docked with ACE2 receptor using HADDOCK and HDOCK algorithms. The binding affinities were calculated using MM/GBSA approach. The results showed that N439K, S477N, and T478K variants have a higher binding affinity toward the ACE2 with higher numbers of salt bridges, hydrogen bonds, and non-bonded contacts. The increased binding affinities of the mutants were attributed to better van der Waals, electrostatic, and non-polar solvation energies. The MM/GBSA calculation revealed a two-fold increase in the binding energies of the variants compared to that of the wild-type. The total binding energies of the complexes were −31.86 kcal/mol for the wildtype-ACE2 complex, −69.82 kcal/mol for S477N mutation, −69.64 kcal/mol for T478K mutation, and −67.85 kcal/mol for N439K mutation [59].
N501Y mutation had gained special interest due to its potential role in the transmission of new variants [[61], [62], [63]]. Socher et al. studied the wild-type's conformational stability, linear interaction energies, and mutated B.1.1.7 (N501Y) RBD-ACE2 complexes. MD simulation of the trimeric spike protein and RBD-ACE2 complex showed reduced flexibility for R78, L249, and T250 after the deletion of amino acids 69, 70, and 144 of the NTD of the B.1.1.7 variant. The C-terminal of the S2 cleavage site and the fusion peptide showed an increase in flexibility with an overall increase of the whole structure size of up to 3 Å compared to the wild-type. Furthermore, an increase in flexibility of the fusion peptide was related to salt bridge rearrangements made by the D614G mutation in B.1.1.7. The results also showed that insertion of N501Y mutation excluded E498 from the RBD–ACE2 interface and reduced its electrostatic interaction toward ACE2. This could explain the significant decrease in the linear interaction energy and binding affinity for B.1.1.7 variant toward ACE2 [61]. However, this mutation increases the van der Waals interaction of K356.
In contrast to the previous finding, Luan et al. investigated how the N501Y mutation within the RBD of the spike protein of SARS-CoV-2 could enhance the binding affinity toward the ACE2 receptor via MD simulations and free energy perturbation method. The simulated ACE2- RBD interface revealed how this mutation could cause conformational changes affecting the interface's binding energy. Simulations were carried out for both the RBD and ACE2-RBD complex. It is worth mentioning that the production run in this work was carried out with only residues far away from the RBD constrained to prevent rotation of the complex from the water box. They concluded that the naturally occurring N501Y mutation could enhance RBD binding affinity toward ACE2 and lead to avoidance of antibody neutralization. It was assumed that Y501 could favor the closed form of the spike protein to evade antibodies before interacting with ACE2. Moreover, RBD Y501 could form hydrophobic π-π stacking interactions with Y41 and hydrophobic interactions with the side chain of K353 of ACE2 [62]. These findings are supported by cell surface-binding assay, single-molecule force microscopy, kinetics study, and molecular dynamics simulations that found N501Y mutation to trigger a stronger interaction of SARS-CoV-2 RBD to ACE2. In addition, the MD simulations of the complex indicated that the N501Y mutation provides additional π-π and π-cation interactions that raise the binding affinity [63].
Additionally, Ou et al. studied the most dominant mutations of SARS-CoV-2 and focused on how the V367F mutation enhances the binding affinity toward the ACE2. The combined sequence alignment, MD simulations, and binding free energy calculation using MM/PBSA proved that V367F mutation decreased the ΔG energy to about 25 % compared to the wild-type spike. Also, the calculated KD of the V367F mutant was estimated to be 0.11 nM, two times lower than that of the wild-type spike. This indicates an increase in the affinity of V367F mutated RBD to the ACE2. This can be attributed to the enhanced structural stability of the RBD beta-sheet scaffold. In addition, this finding was further verified using pseudotyped virus assays, surface plasmon resonance, and receptor-ligand binding enzyme-linked immunosorbent assay [64].
Laffeber et al. have provided experimental evidence about how some SARS-CoV-2 variants bind more strongly to ACE2 than the wild type [65]. They concluded that B.1.1.7 variant (which has N501Y mutation) has an RBD with a binding affinity toward ACE2 seven times stronger than the wild type. On the other hand, the B.1.351 variant (which contains N501Y, E484K, and K417N mutations in the RBD) binds three times stronger to ACE2 than the wild type but two times weaker than B.1.1.7 variant. This is because the E484K mutation enhances the binding affinity toward ACE2 only slightly, while K417N reduces it. They also found that E484K/N501Y variant binds even stronger than B.1.1.7 variant.
3 Conformational dynamics, accessibility, and energetics of S
During viral attachment to cellular receptors, confirmational changes are experienced by at least one partner to form a stable and efficient binding. To investigate the dynamics events in S-ACE2 binding, Ping and associates performed MD simulations for 100 ns on two structures (PDB IDs): 2AJF [66] and 6M0J [67,68]. At the time of the study, the ACE2-bound SARS-CoV-2 trimer was not available yet, so they used four available structures for the ACE2-bound SARS-CoV trimers (PDB ID: 6ACG, 6ACJ, 6ACK, and 6CS2), in which all of them have one RBD up with different angles ranging from 54.8° to 84.6°, as templates to build SARS-CoV-2 trimer. This RBD angle was defined by residues D405–V622–V991. The results revealed that the SARS-CoV-2 sustained a higher binding affinity to ACE2 than the SARS-CoV, regardless of the RBD angle within the specified range.
By decomposing the ΔG to the contributing residues, they found that SARS-CoV-2 had three more residues interacting with ACE2 in comparison to SARS-CoV with a ΔG ≤ −1.0 kcal/mol, namely Y449, Q493, and G496. The contribution of Q493 was found to be −2.64 kcal/mol of the overall ΔG, while the corresponding residue in SARS-CoV N479 has no energy contribution. This shed light on the vital role of the residue Q493 variations on the binding affinity to ACE2. They mutated the interface of SARS-CoV with ACE2 (18 residues) to be like the corresponding residues of SARS-CoV-2 to examine the importance of residue variations between SARS-CoV-2 and SARS-CoV. The mutated SARS-CoV had a much stronger binding affinity than the wild-type SARS-CoV and nearly the same binding affinity as SARS-CoV-2. This result means that we can quantitatively attribute the stronger binding affinity of SARS-CoV-2 against ACE2 to these residue variations.
3.1 Spike conformational dynamics
Cryo-EM studies of S overexpressed in Z cells have identified two states; pre-fusion and post-fusion. The majority of S (97 %) were found in the pre-fusion state that has three isoforms (closed RBD (31 %), one RBD open (55 %), and two RBD open (14 %)) [69]. The pre-fusion and post-fusion isoforms both are flexible over the viral membrane due to their sparsity, giving a tilt angle between 0o and 90o. Hence, antibodies can access the stalk region that lacks high glycosylation [70].
As time passes, most of the published work focuses on the contribution of the RBD to the infection process. However, other studies showed that targeting the NTD in both MERS-CoV and SARS-CoV-2 with monoclonal antibodies 7D10 and 4A8, respectively, gave a neutralizing effect without blocking the RBD-ACE2 binding. This gave more insights into the possible role of NTD in the RBD functional conformational changes, thus might help control viral infection.
Li et al. proposed other geometrical metrics (angle and distances) to describe the movement of NTD and RBD during the simulation. These metrics were; RBD angle (θ r), the tilting poses between RBD and any other conformation; RBD distance (d r), the distance between the RBD and the center of the spike protein; and NTD distance (d n), the distance between the NTD and the center of the spike protein [16]. Four different states of S protein have been revealed with different orientations of RBDs and varying binding affinities to the ACE2 receptor. First, a closed state in which all the three RBDs in the down direction, so they cannot access the ACE2 (inactive), a partially open state with one RBD flipped upward, a semi-open state with two RBD flipped upward, and finally, the open state in which all the three RBDs flipped upward, and hence are accessible to ACE2. To reveal the functional role of the NTD, each state was simulated for 1 μs using accelerated MD with five repeats to sample the relative motion between the NTD and the RBDs. Their findings suggested that the upward RBD is not favored and tends to incline to reach the more stable favored downward orientation. The NTD tends to either interact with the RBD as a wedge to prevent this motion or detach from the RBD, allowing its transition between the upward and downward positions.
Based on the work of Peng et al., the role of the entire spike protein in determining the binding affinity was highlighted [71]. Depending on MD simulation and binding energy calculations, they studied the conformational distribution of the two spike proteins of SARS-CoV-2 and SARS-CoV, the energy barriers between the two conformations of ACE2-accessible and inaccessible, and the ACE2 accessibility and binding strength toward the S-proteins at different RBD angles. They have confirmed tha t SARS-CoV-2 RBD binds more strongly to ACE2 than SARS-CoV RBD because of the higher electrostatic interactions. With four different angles of RBD of SARS-CoV-2, which were built according to ACE2-bound SARS-CoV-S trimmers retrieved from the solved structure, they found that they all have binding affinity for ACE2 higher than that of SARS-CoV regardless of the RBD angle. In addition, they have computed the contribution of each residue of RBD to the ΔG energy. They have concluded that a SARS-CoV structure with 18 specific residue mutations is approaching the binding affinity of SARS-CoV-2 to ACE2. According to the conformational change of the pathway of SARS-CoV-2-S trimer between the down and up states, 52.2° is considered the smallest accessible RBD angle of SARS-CoV-2-S to ACE2, and with increasing the angle, the binding interaction becomes stronger. However, they found that SARS-CoV-2-S has less accessible conformations to ACE2 than SARS-CoV-S, and the inaccessible conformations cannot easily shift to the accessible state as SARS-CoV-S does. Therefore, the total binding affinity of the entire SARS-CoV-2-S protein is similar to or even lower than SARS-CoV-S toward ACE2. This is referred to as the less accessible conformations of SARS-CoV-2-S and the more difficult shift from inaccessible to accessible conformations in the solution [71].
Using insect cells as expression systems, Toelzer et al. experimentally proved that the closed state is predominant and is stabilized by the binding to linoleic acid (LA), an essential fatty acid [72]. The interaction affinity of LA against S trimer and S RBD alone was computationally assessed through repeated MD simulations. The results showed a stable LA-S trimer association. Additionally, the affinity of LA to the S trimer was higher than its binding to the RBD alone. Tian and Tao used 10 μs MD simulations to study the transition between the two S1 subunits, closed and partially open states. The RMSD values in the two states are 5.9 Å and 10.6 Å, respectively. They clarified that the closed state is relatively stable compared to the partially open state, which is dynamic, and undergoes conformational changes after one μs but stabilizes on reaching the sixth microsecond. Their analyses provided more insights into the closed-open transition probability, and a complete pathway was identified between the closed and open states [73].
3.2 The role of C480-C488 in spike dynamics
Grishin et al. discussed the role of disulfide bonds (S—S) in the stability and function of Spike protein [74]. SARS-CoV-2 S protein has 16 disulfide bridges at specific sites, which play an essential role in the stability of the protein. RBD has four S—S bonds; C480-C488 is in the loop that binds to the ACE2 receptor, while the remaining three bonds, C379-C432, C336-C361, and C391-C525, are found on the opposite side of the RBD domain with no contribution in ACE2 interaction. The MD simulations showed that the flexibility of the RBD increases when one S—S (C480-C488) or all of the disulfide bonds are reduced. This was done by simulating three systems (i) RBD that has no reduced S—S bond, (ii) RBD(−1SS) with only C480-C488 reduced, and (iii) RBD(−4SS) with all the S—S bonds reduced.
Firstly, RMSD values of RBD (−1SS) and RBD(−4SS) changed by three folds (1.0–2.5 vs 3.5–6.0 Å) in the RBD (Fig. 4 ). Secondly, RMSF values showed that the regions adjacent to the S—S bonds have higher flexibility. One of the crucial findings was that the ACE2 binding loop remains relatively stable (closed) as long as the C480-C488 is present. In the absence of C480-C488, a series of structural transitions between open states were identified and none of them was capable of ACE2 binding. By increasing the simulation temperature, efficient conformational sampling can be achieved, as they elevated the temperature to 77 °C and allowed the three systems to be simulated for two μs each. They reported an opening of the ACE2-binding loop in all three trajectories. This happened at the beginning of the simulation in both RBD(−1 SS) and RBD(−4 SS), but it happened after one μs in the non-reduced RBD as the loop kept its conformation along half of the trajectory. This confirmed the effect of S—S bonds on the stability of the RBD and its affinity toward ACE2. Experimental techniques confirmed these results by using reducing agents that break the S—S bonds. As a result, the secondary structures are affected, and the melting temperature is reduced from 52 °C to as low as 36–39 °C.Fig. 4 Snapshots of the RBD ACE2 binding loop at the surface conformation, residues 454–492, and their RMSD along six 37 °C molecular dynamics runs. (A) RMSD of Cα Coordinates (nm) against time (ps). RBD with 4 S-S bonds, RBD lacking C480-C488 bond – RBD (−1 SS), and RBD with all four bonds reduced – RBD (−4 SS) are colored blue, green, and red, respectively. For each of them, two curves correspond to two repeats of the simulations. (B) to the left is a ribbon representation of the Spike RBD – ACE2 complex crystal structure focusing on the loop. We showed here only a part of the structure, in olive, the Spike RBD domain is colored with wheat-colored ACE2. Cyan carbon and yellow sulfur atoms in a wire representation of The C480-C488 S—S bond. To the right are the MD simulations snapshots, which present the different ACE2 – binding loop conformations and are labeled accordingly. Copyrights were acquired from the publisher with license number 5503040177103. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 4
Glucose-regulated protein 78 (GRP78) was reported before to be another host-cell receptor that can bind to the SARS-CoV-2 spike [25,75]. This was predicted by protein-protein docking and later confirmed experimentally [76,77]. The recognition site on the spike that was associated with GRP78 substrate binding domain β (SBDβ) was the C480-C488. The recognition of GRP78 to SARS-CoV-2 was predicted to be not specific for that strain of human coronaviruses (HCoVs) but also for other members of HCoVs [78]. Additionally, this region of the host-cell chaperone (SBDβ) was targeted with natural compounds to inhibit viral recognition by the cell-surface protein GRP78 [[79], [80], [81], [82]]. Additionally, the C480-C488 region of the spike of SARS-CoV-2 was also targeted by small molecule inhibitors [29,[31], [32], [33]].
The use of MD simulations to understand SARS-CoV-2 continued to unveil more insights about spike conformational dynamics. Williams et al. studied the role of the flexible loop (namely loop3) at the interface of spike RBD against ACE2 using MD simulations [83]. The RBD binding interface is usually comprised of four loops that tend to be flexible in both unbound and bound states; Loops 1 (438–450), 2 (455–470), 3 (471–491), and 4 (495–508). It is known that loops 3 and 4 are the most flexible regions in the RBD, while the three residues in loop 3 (F486, N487, and Y489) are known to have a stabilizing effect on the binding to ACE2. RMSD analysis following four μs-long molecular dynamics simulations showed that the RBD remains in a stable equilibrium conformation along the trajectories, with an average RMSD value of 1.39 Å. In addition, increased flexibility in residues 369–373 and the Loop 3 region from ∼471 to 491 was observed, which are part of the large ACE2 binding interface. Loop 3 is centralized around the conserved disulfide bond C480-C488. Four structures containing naturally occurring mutants, namely T478I, S477N, V483A, and G476S, were subjected to two μs-long MD simulations for each mutated structure. The backbone RMSD analysis revealed that they all remain stable relative to their starting structure on average with RMSD values: G476S: 1.65 Å; S477N: 1.48 Å; T478I: 1.46 Å; and V483A: 1.44 Å. Also, the per-residue RMSF confirmed the same, leaving us with one remark that loop3 flexibility is resilient to single mutations.
The S2 subunit of the spike is composed of a hydrophobic fusion peptide, two heptad repeats, a transmembrane region, and a cytoplasmic C-terminal tail (1242–1273). Despite the extensive work on solving spike protein 3d structures, these structures mainly showed the S1 subunit, with S2 not included or missing electron density at the C-terminal tail. However, being the least studied region of S protein, the C-terminal tail contains a conserved ER retrieval signal (KKXX), making it essential to study. In addition to its sub-cellular localization role, deletions in the cytoplasmic C-terminal tail affected the infectivity of the virus [84,85]. Despite the microseconds-long MD simulations at different temperatures, the C-terminal tail of S remained disordered [86]. This unstructured nature was also reported by circular dichroism spectroscopy.
4 Role of glycans on S and ACE2
4.1 Nature of SARS-CoV-2 S glycosylation
Glycosylation is a post-translational modification that confers additional capabilities on the modified protein. Viruses evolved to exploit the cellular machinery to add glycans to their proteins, facilitating many roles during the viral replication cycle [87,88]. Indeed, events of viral entry are initiated by molecular recognition and specific interactions between viral and host proteins (glycoproteins) [88]. Moreover, the glycosylation of viral envelope proteins plays a pivotal role in successful infection, and virion integrity since the folding and trafficking of the viral protein are affected by glycosylation. For instance, tunicamycin-mediated inhibition of spike glycosylation in SARS-CoV-2 resulted in spike-deficient virions due to the disruption of proper protein folding [42]. In various enveloped viruses, glycosylated S proteins evade the host's innate and adaptive immune responses by shielding the amino acids at the immunogenic epitopes from molecular recognition by the host's immune system [69,88].
The glycosylation of proteins is classified into two groups, depending on the linked sugar moiety to the amino acids: N-glycans and O-glycans. The former is established by covalent bonding to the amide nitrogen atom of an asparagine residue, while the O-glycosylation takes place on an oxygen atom at the side chain of serine or threonine residues. N-glycosylation sites are usually found in a sequence of NXT/S, where X is any amino acid except proline [87,88].
SARS-CoV-2 is not an exception. Its trimeric S protein has been found to harbor numerous glycosylation sequons, and 40 % of the protein is shielded by glycans [89,90]. On each protomer of the SARS-CoV-2 S protein, 22 N-glycosylation sites have been predicted, and 17 have already been confirmed experimentally (Fig. 5 ) [91,92]. Additional O-glycosylation sites were also found on the Receptor Binding Domain (RBD) [91,92]. It is worth mentioning that the nature of glycosylation of the S1 subunit varies by the host cell type and growth conditions adopted during the experimentation [91,93,94]. For instance, tandem mass spectrometry studies revealed that S1 domains of spike expressed in human embryonic kidney (HEK) 293 cells were predominantly glycosylated by the complex type N-glycans [93,94]. Another study that used the same expression system reported different findings where high-mannose glycans were the main N-glycans type [91]. However, spikes expressed in HEK293 cells have high-mannose and complex types as the dominant types, while the hybrid glycans were only found in a small proportion [90]. In the baculovirus-infected insect system, however, most glycans were high-mannose type, while hybrid glycans constituted only a small proportion [93].Fig. 5 Surface rendering of the glycosylated S protein (PDB id 6VYB). The spike is in open conformation where the RBD of chain B (colored in light brown) is up, and the RBD of chain C (white) is in the down conformation. Glycans are shown in blue. The upper panel is a cutaway view of the top region showing the N-glycans at positions N234 (green), N165 (cyan), and N343 (purple). Reproduced from [36] under Creative Commons licenses. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 5
4.2 Structural roles of glycans
The spike protein needs to undergo moderate structural changes in its RBD to bind to its cellular receptor ACE2. Several MD simulation studies have suggested pivotal roles for spike glycans to mediate such structural changes. Microsecond-long MD simulations of the pre-fusion full-length glycosylated S protein elucidated N-glycans' contributions in modulating the spike's conformational dynamics [60]. During the conformational change of the RBD from the closed state “down” to the open state “up”, one or more of the RBDs emerge up to the exterior of the spike, leaving a vacant pocket that leads to the interior of the trimeric spike [95]. If water molecules fill the exposed pocket after RBD opening, the structural stability of S is believed to be markedly disrupted [96]. The N-glycans attached to the neighboring ‘down’ protomer at N165, N234 (on the NTD), and N343 (on the closed RBD) play the most critical roles regarding the structural stability of S protein in its “up” conformation [36,60,97]. Upon RBD opening, the large oligomannose at N234 has been found to fill up the exposed pocket and maintain the integrity of S via multiple hydrogen bonding with the N-terminal domain (NTD), the central helix, and the open conformation of RBD [36,60]. In the closed state, the N165 glycan is located above the RBD, while the glycan of N234 is located underneath.
Replacement of large glycans (Man7 & Man9) on N234 with a paucimannose (a Man3 short glycan) resulted in an unstable RBD opening and ∼ 60 % reduction in binding to ACE2 as a response to inefficient interactions, specifically with the disordered loop of the receptor-binding motif (RBM) [36,60]. N-glycan at N165 was also found either to fill the same pocket (filled by N234) or to make extensive interactions by acting as a bridge between NTD and the opened RBD [60]. These observations were supported by additional MD simulations of non-glycosylated mutants (N165A and N234A) in which the opened RBD explored a large conformational space compared to the glycosylated wild-type structure [60]. Principal component and angle analyses of the RBD movement in the mentioned mutants suggested a regulatory role of N-glycans in opening and closing the structure. The absence of glycans at N165 and N234 destabilized the open conformation by permitting its disordered motions. Furthermore, another MD simulation study that adopted a weighted ensemble path-sampling strategy identified N343 glycan as well as three residues (D405, R408, and D427) as major players in the RBD opening process [98]. N343 glycan was also found to initiate spike transition from closed to open conformation by acting as a gate that pushes the RBD from the ‘down’ to the ‘up’ state via intercalation (within 3.5 Å) between F456, R457, Y489, and F490 on the RBM [98]. Alanine mutagenesis of N165 and N234 or N343 prior to biolayer interferometry analysis showed a significant decrease (50–90 %) in the binding response of the mutated spike to ACE2 [60]. Additional evidence for glycans' roles in the spike structural stability was obtained through micro-second MD simulations of de-glycosylated S1 constructs (heads only) that showed an early reversion of open conformation to its closed form during the first 100 ns [97].
4.3 Glycans affect receptor binding
ACE2 is also heavily glycosylated, mostly by complex-type sugars, especially near the region interacting with the RBD of SARS-CoV-2 S glycoprotein [99]. The spike-ACE2 binding interaction has been studied computationally and experimentally, focusing on the glycans' contribution [93,97,99,100]. As mentioned previously, the stability of the RBD by glycans at N165 and N234 allows efficient binding to ACE2. Experimental evidence of such suggestions was obtained via biolayer interferometry that detected a significant reduction in binding response to ACE2 when glycans at N234, N156, or N343 were abolished by alanine mutations [60,98]. This reduction was attributed to the RBD conformational shift toward the closed state based on the increased population of spikes with the closed RBD state. These observations from MD simulation studies explain, at the atomic level, the roles of glycans in the RBD opening and the inefficient entry of virions lacking N-glycans to human cell models [101].
The nature and extent of glycans' contributions to the binding energy are not entirely clear. It is currently believed that glycans play partial roles in binding the RBD to human ACE2. The MD simulation-based comparisons of glycosylated wild-type and omicron variants in terms of binding energies have shown comparable results [102,103]. However, real-time surface plasmon resonance assays showed different dissociation constant KD values (range: 2.05–23.9 nM) for S1-ACE2 obtained from different expression systems (different glycosylation patterns), including Baculovirus-infected insect, Chinese hamster ovarian, and HEK 293 cells [93].
Steered MD simulations and biolayer interferometry experiments revealed that S glycans flexibility allows H-bonds to be formed and relaxed in a catch-slip behavior in which a Hydrogen bond is broken, and another takes its place at a larger distance [104]. These H-bonds are formed between the spike's glycans and ACE2 amino acids (glycan-protein) and/or S glycans and ACE2 glycans (glycan-glycan) [104]. On the other side, the glycan at N322 on ACE2 strengthens the binding of RBD to the receptor by −19.12 to −47.80 kcal/mol [105]. This strong binding is attributed to the architecture of the N322 conserved pocket on the RBD, which is composed of a hydrophobic core surrounded by polar and charged residues [105]. These findings explain the genetic data regarding the failure of coronaviruses to bind to receptors lacking N322 glycosylation. Additionally, N90 glycan interacts tightly with the glycans of the RBD (−9.56 to −35.85 kcal/mol), especially with the high-mannose type on N322 [105].
Strikingly, glycans near the receptor-binding regions may negatively affect the binding process. The de-glycosylated complexes showed stronger binding affinities in plasmon resonance assays as well as in MM/PBSA free energy calculations after 100–200 ns MD simulations [93]. These variations in binding affinity were linked to the observed steric hindrance between the glycosylated N-terminal regions of S1 domains and the ACE2 [93]. Moreover, high molecular size N-glycans caused more steric clashes with the cellular receptor. Furthermore, the Coulomb repulsions were also found between negatively charged residues on ACE2 and the sialic acid moieties on complex glycans [93]. Nonetheless, the glycans on coronaviruses are fewer than seen in other viruses, which is an advantage to the virus for efficient receptor binding [88].
Acharya et al. used MD simulations and binding affinity calculations to reveal the role of ACE2 glycans in the binding of SARS-CoV and SARS-CoV-2. They found that glycans at N357 and N330 positions of SARS-CoV RBD block the interaction between ACE2 glycan at N322 and SARS-CoV RBD. In contrast, the absence of a glycan at N357 of SARS-CoV RBD (N370 in SARS-CoV-2) enhances its binding to N322 glycan on ACE2 via exposing additional sites for interaction and increased stability of the RBD open conformation. Therefore, the absence of the glycan at N370 in the SARS-CoV-2 RBD may be an evolutionary step toward stronger binding to ACE2. Additionally, the N53 glycan on ACE2 plays a significant role in stabilizing the homodimer interface of ACE2. On the contrary, mutations of SARS-CoV-2 may reduce ACE2 binding affinity according to the position of the mutation. For example, the spike mutations N439 & G504 but not G404 & N437 are found in the S RBD–ACE2 interface, and both decrease ACE2 binding [106].
4.4 Immune evasion and antigenicity
The glycans on the spike are believed to shield protein epitopes from the effector immune cells and molecules; however, it is unclear how effective this shielding is on immune evasion [87,88]. Antibodies against different spike epitopes are readily elicited, and the currently effective vaccines are based on S [107]. Computational probing of different epitopes on a spike has been utilized to evaluate the effect of glycosylation on antibody binding possibilities (Fig. 6 ) [89,97]. Different structures of glycosylated spikes with small glycans (paucimannose, M3) were subjected to calculations of accessible surface area (ASA). The results showed that 69 % of the protein surface is accessible for antibodies. However, others have found complex glycans to shield approximately 44 % of the spike (56 % accessible) [89]. The glycans at the stalk region of the spike are more prevalent than in the S1 subunit and provide a more shielding effect. It is worth mentioning that most glycans on the SARS-CoV-2 spike are highly flexible, with a wide conformational space covering most of the spike's surface [97]. Indeed, the superimposition of RBD-antibody complexes on glycosylated spike proteins has shown only a few cases of potential clashes and steric hindrances [97].Fig. 6 The protein surface is colored according to antibody accessibility. The RBD region in the open conformation is circled in blue. Spaces occupied by glycans are shown as moss surface from MD simulations. The glycans are shown in ball-and-stick representation: M9 (green), M5 (dark yellow), hybrid (orange), complex (pink). Reproduced from [42] under Creative Commons licenses. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 6
Four types of RBD-antibodies (B38, CR3022, H11-D4, and S309) were used to interrogate the efficiency of glycans on fully-glycosylated models to occlude epitopes on RBD [97]. The epitopes of B38 and CR3022 are irrelevant to the glycans. B38 can only bind if the open state of the RBD is attained; otherwise, it is precluded by the neighboring RBDs. However, CR3022 requires all RBDs on the spike to be in the “up” conformation. In contrast, the H11-D4 binds easily to the open RBD; however, in the closed state, the residue, N165, has weak steric clashes with the antibody but can be overcome, as seen in the experimentally solved structure (PBD ID: 6Z43). Four glycans are located around the epitope of S309; two (N331 and N343) are on the same RBD, while the other (N122 and N165) are attached to the nearby NTD. Only N343 and N122 were found to experience minor and severe steric clashes with the S309 antibody, respectively [97]. Furthermore, the N343 glycan shielding decreases during the RBD transition to the ‘up’ state, whereas the shielding role of glycans at N165 and N234 is sustained during the opening process [98]. Such observations support the belief that the glycans are not only for shielding purposes but also participate in structural and functional processes.
For different reasons, these results of antigenicity evaluation should be received with caution. First, sphere-based probing of antibody accessible surface area is expected to overestimate the contributions of glycans shielding of S protein from immune recognition as antigen-binding sites on the antibody are not necessarily spherical. Protruded regions (i.e., antigen-binding sites on the antibody) might be able to access deep epitopes. Second, the effects of glycosylation heterogeneity on lectins' abilities to recognize the spike are still unknown (see Fig. 7 ).Fig. 7 Accessibility of neutralizing antibody epitopes. (A) Distribution of glycans when S head structures in multiple snapshots are aligned as mosses (left). Four different epitopes targeted by neutralizing antibodies are shown in different colors (right). (B) H11-D4 epitope (green) in the RBD (left: open, right: closed), N165 glycan (red) on the neighboring NTD, and N343 glycan (violet) on the neighboring RBD. Reproduced from [97] under creative common license. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 7
Lastly, the glycans could be recognized as “self” since human enzymes synthesized and modified them in human host cells. Nonetheless, glycan-dependent neutralizing antibodies have been identified in SARS-CoV-2 as parts of epitopes [107] as well as in human immunodeficiency virus [108].
Based on these findings, computational methods, including MD simulation, MM-GBSA, and protein-protein docking, can intensely study viral protein dynamics. This can help us understand molecular recognition and its changes due to mutations to the viral proteins. Fabricating a vaccine or designing inhibitors to that recognition can help the host cell from being a shelter for new infectious virions. Furthermore, it could be helpful in future pandemics, which could be more ferocious to humans than COVID-19.
5 Conclusion
This review article summarizes the effort spent on simulating the SARS-CoV-2 spike by many groups worldwide and how it shapes our understanding of the spike's role in viral infectivity. The receptor binding domain of the spike is the region of interest that recognizes host-cell receptors and facilitates viral entry. Loop dynamics affect the recognition by the host-cell receptors, where mutation may have a great impact. The carbohydrate moieties that decorate the spike is crucial for host-cell recognition and conformational transitions. Additionally, mutations in the RBD affect ACE2 and GRP78 binding and hence, viral recognition. The new strains have different patterns of increasing or decreasing viral recognition by affecting the spike RBD binding affinities against the host-cell receptors. Experimental data support the simulation findings throughout the lifespan of the pandemic. This paves the way for finding tangible ways to fight against future pandemics.
CRediT authorship contribution statement
J.A., A.Elg., and A.S. wrote the first draft. A.Ez. and A.Elf. revised the manuscript. All authors approve the final form.
Ethics approval and consent to participate.
Not applicable.
Consent for publication
All authors approve the final form and approve the submission.
Funding
Not applicable.
Declaration of competing interest
All the authors affirm no conflict of interest in this work.
Data availability
Not applicable.
Acknowledgments
Not applicable.
==== Refs
References
1 Leach A. Molecular Modelling: Principles and Applications 2nd edition 2001 Prentice Hall
2 Schlick T. Molecular Modeling and Simulation: An Interdisciplinary Guide 2010 Springer
3 Dey D. Biswas P. Paul P. Mahmud S. Ema T.I. Khan A.A. Natural flavonoids effectively block the CD81 receptor of hepatocytes and inhibit HCV infection: a computational drug development approach Mol. Divers. 2022 Online ahead of print
4 Beerenwinkel N. Sing T. Lengauer T. Rahnenfuhrer J. Roomp K. Savenkov I. Computational methods for the design of effective therapies against drug resistant HIV strains Bioinformatics. 21 21 2005 3943-50
5 Elfiky A.A. Novel guanosine derivatives as anti-hcv ns5b polymerase: a QSAR and molecular docking study Medicinal Chemistry 15 2 2019 130 137 (Shariqah (United Arab Emirates)) 30324891
6 Elfiky A.A. Elshemey W.M. Gawad W.A. 2′-Methylguanosine prodrug (IDX-184), Phosphoramidate prodrug (Sofosbuvir), Diisobutyryl prodrug (R7128) are better than their parent nucleotides and ribavirin in hepatitis C virus inhibition: A molecular modeling study J. Comput. Theor. Nanosci. 12 3 2015 376 386
7 Ezat A.A. Elshemey W.M. A comparative study of the efficiency of HCV NS3/4A protease drugs against different HCV genotypes using in silico approaches Life Sci. 217 2019 176 184 30528183
8 Hufsky F. Lamkiewicz K. Almeida A. Aouacheria A. Arighi C. Bateman A. Computational strategies to combat COVID-19: useful tools to accelerate SARS-CoV-2 and coronavirus research Brief. Bioinform. 22 2 2021 642 663 33147627
9 Shehroz M. Zaheer T. Hussain T. Computer-aided drug design against spike glycoprotein of SARS-CoV-2 to aid COVID-19 treatment Heliyon. 6 10 2020 e05278
10 Elfiky A.A. Anti-HCV, nucleotide inhibitors, repurposing against COVID-19 Life Sci. 248 2020 117477
11 Ogunyemi O.M. Gyebi G.A. Elfiky A.A. Afolabi S.O. Ogunro O.B. Adegunloye A.P. Alkaloids and flavonoids from African phytochemicals as potential inhibitors of SARS-Cov-2 RNA-dependent RNA polymerase: an in silico perspective Antivir Chem Chemother. 28 2020 2040206620984076
12 Abraham Peele K. Srihansa T. Krupanidhi S. Ayyagari V.S. Venkateswarulu T.C. Design of multi-epitope vaccine candidate against SARS-CoV-2: a in-silico study J. Biomol. Struct. Dyn. 39 10 2021 3793 3801 32419646
13 Elshemey W.M. Elfiky A.A. Ibrahim I.M. Elgohary A.M. Interference of Chaga mushroom terpenoids with the attachment of SARS-CoV-2; in silico perspective Comput. Biol. Med. 145 2022 105478
14 Chakraborty H. Bhattacharjya S. Mechanistic insights of host cell fusion of SARS-CoV-1 and SARS-CoV-2 from atomic resolution structure and membrane dynamics Biophys. Chem. 265 2020 106438
15 Huang Y. Yang C. Xu X.F. Xu W. Liu S.W. Structural and functional properties of SARS-CoV-2 spike protein: potential antivirus drug development for COVID-19 Acta Pharmacol. Sin. 41 9 2020 1141 1149 32747721
16 Li Y. Wang T. Zhang J. Shao B. Gong H. Wang Y. Exploring the regulatory function of the N-terminal domain of SARS-CoV-2 spike protein through molecular dynamics simulation Advanced Theory and Simulations. 4 10 2021 2100152 34901736
17 Malik Y.A. Properties of coronavirus and SARS-CoV-2 The Malaysian journal of pathology. 42 1 2020 3 11 32342926
18 Acharya A. Lynch D.L. Pavlova A. Pang Y.T. Gumbart J.C. ACE2 glycans preferentially interact with SARS-CoV-2 over SARS-CoV Chem. Commun. (Camb.) 57 48 2021 5949 5952 34019602
19 Li W. Zhang C. Sui J. Kuhn J.H. Moore M.J. Luo S. Receptor and viral determinants of SARS-coronavirus adaptation to human ACE2 EMBO J. 24 8 2005 1634 1643 15791205
20 Belouzard S. Millet J.K. Licitra B.N. Whittaker G.R. Mechanisms of coronavirus cell entry mediated by the viral spike protein Viruses. 4 6 2012 1011 1033 22816037
21 Zhang J. Xiao T. Cai Y. Chen B. Structure of SARS-CoV-2 spike protein Curr Opin Virol. 50 2021 173 182 34534731
22 Nassar A. Ibrahim I.M. Amin F.G. Magdy M. Elgharib A.M. Azzam E.B. A review of human Coronaviruses’ receptors: the host-cell targets for the crown bearing viruses Molecules. 26 21 2021 6455 34770863
23 Ibrahim I.M. Abdelmalek D.H. Elshahat M.E. Elfiky A.A. COVID-19 spike-host cell receptor GRP78 binding site prediction J. Inf. Secur. 80 5 2020 554 562
24 Cai Y. Zhang J. Xiao T. Peng H. Sterling S.M. Walsh R.M. Jr. Distinct conformational states of SARS-CoV-2 spike protein Science. 369 6511 2020 1586 1592 32694201
25 Ibrahim I.M. Abdelmalek D.H. Elshahat M.E. Elfiky A.A. COVID-19 spike-host cell receptor GRP78 binding site prediction J. Infect. 80 5 2020 554 562 32169481
26 Wrapp D. Wang N. Corbett K.S. Goldsmith J.A. Hsieh C.L. Abiona O. Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation Science. 367 6483 2020 1260 1263 32075877
27 Henderson R. Edwards R.J. Mansouri K. Janowska K. Stalls V. Gobeil S.M.C. Controlling the SARS-CoV-2 spike glycoprotein conformation Nat. Struct. Mol. Biol. 27 10 2020 925 933 32699321
28 Chi X. Yan R. Zhang J. Zhang G. Zhang Y. Hao M. A neutralizing human antibody binds to the N-terminal domain of the spike protein of SARS-CoV-2 Science. 369 6504 2020 650 655 32571838
29 Hamdi A.H. S, A Elfiky A. Chitosan, from crayfish wastes, as a possible therapeutic option against COVID-19; an in-silico perspective Egyptian Journal of Aquatic Biology and Fisheries. 26 2 2022 429 441
30 Sarfraz I. Rasul A. Adem Ş. Ucak I. Sarfraz A. Bukhari S.A. Chapter 8 - dietary polyphenols as therapeutic agents to combat COVID-19 Egbuna C. Coronavirus Drug Discovery 2022 Elsevier 203 215
31 Adem S. Eyupoglu V. Sarfraz I. Rasul A. Zahoor A.F. Ali M. Caffeic acid derivatives (CAFDs) as inhibitors of SARS-CoV-2: CAFDs-based functional foods as a potential alternative approach to combat COVID-19 Phytomedicine. 85 2021 153310
32 Abosheasha M.A. El-Gowily A.H. Elfiky A.A. Potential antiviral properties of antiplatelet agents against SARS-CoV-2 infection: an in silico perspective J. Thromb. Thrombolysis 53 2 2022 273 281 34510337
33 Basal W.T. Elfiky A. Eid J. Chaga medicinal mushroom Inonotus obliquus (Agaricomycetes) Terpenoids may interfere with SARS-CoV-2 spike protein recognition of the host cell: A molecular docking study Int J Med Mushrooms. 23 3 2021 1 14
34 Singh R. Bhardwaj V.K. Sharma J. Kumar D. Purohit R. Identification of potential plant bioactive as SARS-CoV-2 spike protein and human ACE2 fusion inhibitors Comput. Biol. Med. 136 2021 104631
35 Aljindan R.Y. Al-Subaie A.M. Al-Ohali A.I. Kumar D.T. Doss C.G. Kamaraj B. Investigation of nonsynonymous mutations in the spike protein of SARS-CoV-2 and its interaction with the ACE2 receptor by molecular docking and MM/GBSA approach Comput. Biol. Med. 135 2021 104654
36 Harbison A.M. Fogarty C.A. Phung T.K. Satheesan A. Schulz B.L. Fadda E. Fine-tuning the spike: role of the nature and topology of the glycan shield in the structure and dynamics of the SARS-CoV-2 S Chem. Sci. 13 2 2022 386 395 35126971
37 Nguyen H.L. Lan P.D. Thai N.Q. Nissley D.A. O’Brien E.P. Li M.S. Does SARS-CoV-2 bind to human ACE2 more strongly than does SARS-CoV? J. Phys. Chem. B 124 34 2020 7336 7347 32790406
38 Yan R. Zhang Y. Li Y. Xia L. Guo Y. Zhou Q. Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2 Science. 367 6485 2020 1444 1448 32132184
39 Zhang Z.B. Xia Y.L. Shen J.X. Du W.W. Fu Y.X. Liu S.Q. Mechanistic origin of different binding affinities of SARS-CoV and SARS-CoV-2 spike RBDs to human ACE2 Cells. 11 8 2022
40 Delgado J.M. Duro N. Rogers D.M. Tkatchenko A. Pandit S.A. Varma S. Molecular basis for higher affinity of SARS-CoV-2 spike RBD for human ACE2 receptor Proteins. 89 9 2021 1134 1144 33864655
41 Zhang J. Xiao T. Cai Y. Chen B. Structure of SARS-CoV-2 spike protein Current Opinion in Virology. 50 2021 173 182 34534731
42 Walls A.C. Park Y.J. Tortorici M.A. Wall A. McGuire A.T. Veesler D. Structure, function, and antigenicity of the SARS-CoV-2 spike glycoprotein Cell. 181 2 2020 281-92 e6
43 Henderson R. Edwards R.J. Mansouri K. Janowska K. Stalls V. Gobeil S.M.C. Controlling the SARS-CoV-2 spike glycoprotein conformation Nat. Struct. Mol. Biol. 27 10 2020 925 933 32699321
44 Juraszek J. Rutten L. Blokland S. Bouchier P. Voorzaat R. Ritschel T. Stabilizing the closed SARS-CoV-2 spike trimer Nat. Commun. 12 1 2021 244 33431842
45 Lopez-Cortes G.I. Palacios-Perez M. Zamudio G.S. Velediaz H.F. Ortega E. Jose M.V. Neutral evolution test of the spike protein of SARS-CoV-2 and its implications in the binding to ACE2 Sci. Rep. 11 1 2021 18847 34552110
46 Rodriguez J.H. Attractive and repulsive residue fragments at the interface of SARS-CoV-2 and hACE2 Sci. Rep. 11 1 2021 12567 34131210
47 Amin M. Sorour M.K. Kasry A. Comparing the binding interactions in the receptor binding domains of SARS-CoV-2 and SARS-CoV The journal of physical chemistry letters. 11 12 2020 4897 4900 32478523
48 Nguyen H.L. Lan P.D. Thai N.Q. Nissley D.A. O’Brien E.P. Li M.S. Does SARS-CoV-2 bind to human ACE2 more strongly than does SARS-CoV? J. Phys. Chem. B 124 34 2020 7336 7347 32790406
49 Li Z. Zhang J.Z.H. Quantitative analysis of ACE2 binding to coronavirus spike proteins: SARS-CoV-2 vs. SARS-CoV and RaTG13 Physical Chemistry Chemical Physics : PCCP. 23 25 2021 13926 13933 34137759
50 Simoes I.C. Costa I.P. Coimbra J.T. Ramos M.J. Fernandes P.A. New parameters for higher accuracy in the computation of binding free energy differences upon alanine scanning mutagenesis on protein-protein interfaces J. Chem. Inf. Model. 57 1 2017 60 72 27936711
51 Chen J. Wang R. Wang M. Wei G.W. Mutations strengthened SARS-CoV-2 infectivity J. Mol. Biol. 432 19 2020 5212 5226 32710986
52 Krause P.R. Fleming T.R. Longini I.M. Peto R. Briand S. Heymann D.L. SARS-CoV-2 variants and vaccines N. Engl. J. Med. 385 2 2021 179 186 34161052
53 Elfiky A.A. Ibrahim I.M. Host-cell recognition through Cs-GRP78 is enhanced in the new omicron variant of SARS-CoV-2, in silico structural point of view J. Inf. Secur. 84 5 2022 722 746
54 Harvey W.T. Carabelli A.M. Jackson B. Gupta R.K. Thomson E.C. Harrison E.M. SARS-CoV-2 variants, spike mutations and immune escape Nat Rev Microbiol. 19 7 2021 409 424 34075212
55 Winger A. Caspari T. The spike of concern—the novel variants of SARS-CoV-2 Viruses. 13 6 2021 1002 34071984
56 Xie X. Liu Y. Liu J. Zhang X. Zou J. Fontes-Garfias C.R. Neutralization of SARS-CoV-2 spike 69/70 deletion, E484K and N501Y variants by BNT162b2 vaccine-elicited sera Nat. Med. 27 4 2021 620 621 33558724
57 Meng B. Kemp S.A. Papa G. Datir R. Ferreira I. Marelli S. Recurrent emergence of SARS-CoV-2 spike deletion H69/V70 and its role in the alpha variant B.1.1.7 Cell Rep. 35 13 2021 109292
58 Wang R. Chen J. Gao K. Wei G.W. Vaccine-escape and fast-growing mutations in the United Kingdom, the United States, Singapore, Spain, India, and other COVID-19-devastated countries Genomics. 113 4 2021 2158 2170 34004284
59 Suleman M. Yousafi Q. Ali J. Ali S.S. Hussain Z. Ali S. Bioinformatics analysis of the differences in the binding profile of the wild-type and mutants of the SARS-CoV-2 spike protein variants with the ACE2 receptor Comput. Biol. Med. 138 2021 104936
60 Casalino L. Gaieb Z. Goldsmith J.A. Hjorth C.K. Dommer A.C. Harbison A.M. Beyond shielding: the roles of Glycans in the SARS-CoV-2 spike protein ACS Cent Sci. 6 10 2020 1722 1734 33140034
61 Socher E. Conrad M. Heger L. Paulsen F. Sticht H. Zunke F. Mutations in the B.1.1.7 SARS-CoV-2 spike protein reduce receptor-binding affinity and induce a flexible link to the fusion peptide Biomedicines. 9 5 2021
62 Luan B. Wang H. Huynh T. Enhanced binding of the N501Y-mutated SARS-CoV-2 spike protein to the human ACE2 receptor: insights from molecular dynamics simulations FEBS Lett. 595 10 2021 1454 1461 33728680
63 Tian F. Tong B. Sun L. Shi S. Zheng B. Wang Z. N501Y mutation of spike protein in SARS-CoV-2 strengthens its binding to receptor ACE2 Elife. 10 2021 e69091
64 Ou J. Zhou Z. Dai R. Zhang J. Zhao S. Wu X. V367F mutation in SARS-CoV-2 spike RBD emerging during the early transmission phase enhances viral infectivity through increased human ACE2 receptor binding affinity J. Virol. 95 16 2021 e0061721
65 Laffeber C. de Koning K. Kanaar R. Lebbink J.H.G. Experimental evidence for enhanced receptor binding by rapidly spreading SARS-CoV-2 variants J. Mol. Biol. 433 15 2021 167058
66 Li F. Li W. Farzan M. Harrison S.C. Structure of SARS coronavirus spike receptor-binding domain complexed with receptor Science. 309 5742 2005 1864 1868 16166518
67 Lan J. Ge J. Yu J. Shan S. Zhou H. Fan S. Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor Nature. 581 7807 2020 215 220 32225176
68 Peng C. Zhu Z. Shi Y. Wang X. Mu K. Yang Y. Computational insights into the conformational accessibility and binding strength of SARS-CoV-2 spike protein to human angiotensin-converting enzyme 2 The journal of physical chemistry letters. 11 24 2020 10482 10488 33274945
69 Ismail A.M. Elfiky A.A. SARS-CoV-2 spike behavior in situ: a Cryo-EM images for a better understanding of the COVID-19 pandemic Signal Transduct Target Ther. 5 1 2020 252 33127886
70 Ke Z. Oton J. Qu K. Cortese M. Zila V. McKeane L. Structures and distributions of SARS-CoV-2 spike proteins on intact virions Nature. 588 7838 2020 498 502 32805734
71 Peng C. Zhu Z. Shi Y. Wang X. Mu K. Yang Y. Computational insights into the conformational accessibility and binding strength of SARS-CoV-2 spike protein to human angiotensin-converting enzyme 2 The journal of physical chemistry letters. 11 24 2020 10482 10488 33274945
72 Toelzer C. Gupta K. Yadav S.K.N. Borucu U. Davidson A.D. Kavanagh Williamson M. Free fatty acid binding pocket in the locked structure of SARS-CoV-2 spike protein Science. 370 6517 2020 725 730 32958580
73 Tian H. Tao P. Deciphering the protein motion of S1 subunit in SARS-CoV-2 spike glycoprotein through integrated computational methods J. Biomol. Struct. Dyn. 39 17 2021 6705 6712 32746720
74 Grishin A.M. Dolgova N.V. Landreth S. Fisette O. Pickering I.J. George G.N. Disulfide bonds play a critical role in the structure and function of the receptor-binding domain of the SARS-CoV-2 spike antigen J. Mol. Biol. 434 2 2022 167357
75 Elfiky A.A. Ibrahim I.M. Amin F.G. Ismail A.M. Elshemey W.M. COVID-19 and cell stress Rezaei N. Coronavirus Disease - COVID-19 2021 Springer International Publishing Cham 169 178
76 Carlos A.J. Ha D.P. Yeh D.W. Van Krieken R. Tseng C.C. Zhang P. The chaperone GRP78 is a host auxiliary factor for SARS-CoV-2 and GRP78 depleting antibody blocks viral entry and infection J. Biol. Chem. 296 2021 100759
77 Shin W.J. Ha D.P. Machida K. Lee A.S. The stress-inducible ER chaperone GRP78/BiP is upregulated during SARS-CoV-2 infection and acts as a pro-viral protein Nat. Commun. 13 1 2022 6551 36376289
78 Elfiky A.A. SARS-CoV-2 spike-heat shock protein A5 (GRP78) recognition may be related to the immersed human coronaviruses Front. Pharmacol. 11 2020 577467
79 Elfiky A.A. Natural products may interfere with SARS-CoV-2 attachment to the host cell J. Biomol. Struct. Dyn. 39 9 2021 3194 3203 32340551
80 Elgohary A.M. Elfiky A.A. Barakat K. GRP78: A possible relationship of COVID-19 and the mucormycosis; in silico perspective Comput. Biol. Med. 139 2021 104956
81 Palmeira A. Sousa E. Köseler A. Sabirli R. Gören T. Türkçüer İ. Preliminary virtual screening studies to identify GRP78 inhibitors which may interfere with SARS-CoV-2 infection Pharmaceuticals. 13 6 2020 132 32630514
82 Abu-Mahfouz A. Ali M. Elfiky A. Anti-breast cancer drugs targeting cell-surface glucose-regulated protein 78: a drug repositioning in silico study J. Biomol. Struct. Dyn. 1-15 2022
83 Williams J.K. Wang B. Sam A. Hoop C.L. Case D.A. Baum J. Molecular dynamics analysis of a flexible loop at the binding interface of the SARS-CoV-2 spike protein receptor-binding domain Proteins. 90 5 2022 1044 1053 34375467
84 Giri R. Bhardwaj T. Shegane M. Gehi B.R. Kumar P. Gadhave K. Understanding COVID-19 via comparative analysis of dark proteomes of SARS-CoV-2, human SARS and bat SARS-like coronaviruses Cellular and Molecular Life Sciences : CMLS. 78 4 2021 1655 1688 32712910
85 Dieterle M.E. Haslwanter D. Bortz R.H. 3rd Wirchnianski A.S. Lasso G. Vergnolle O. A replication-competent vesicular stomatitis virus for studies of SARS-CoV-2 spike-mediated cell entry and its inhibition Cell Host Microbe 28 3 2020 486 96 e6 32738193
86 Kumar P. Bhardwaj T. Garg N. Giri R. Microsecond simulations and CD spectroscopy reveals the intrinsically disordered nature of SARS-CoV-2 spike-C-terminal cytoplasmic tail (residues 1242-1273) in isolation Virology. 566 2022 42 55 34864296
87 Vigerust D.J. Shepherd V.L. Virus glycosylation: role in virulence and immune interactions Trends Microbiol. 15 5 2007 211 218 17398101
88 Watanabe Y. Bowden T.A. Wilson I.A. Crispin M. Exploitation of glycosylation in enveloped virus pathobiology Biochimica et Biophysica Acta General Subjects. 1863 10 2019 1480 1497 31121217
89 Grant O.C. Montgomery D. Ito K. Woods R.J. Analysis of the SARS-CoV-2 spike protein glycan shield reveals implications for immune recognition Sci. Rep. 10 1 2020 14991 32929138
90 Watanabe Y. Allen J.D. Wrapp D. McLellan J.S. Crispin M. Site-specific glycan analysis of the SARS-CoV-2 spike Science. 369 6501 2020 330 333 32366695
91 Shajahan A. Supekar N.T. Gleinich A.S. Azadi P. Deducing the N- and O-glycosylation profile of the spike protein of novel coronavirus SARS-CoV-2 Glycobiology. 30 12 2020 981 988 32363391
92 Sanda M. Ahn J. Kozlik P. Goldman R. Analysis of site and structure specific core fucosylation in liver cirrhosis using exoglycosidase-assisted data-independent LC-MS/MS Sci. Rep. 11 1 2021 23273 34857845
93 Huang N. Perez P. Kato T. Mikami Y. Okuda K. Gilmore R.C. SARS-CoV-2 infection of the oral cavity and saliva Nat. Med. 27 5 2021 892 903 33767405
94 Zhang L. Richards A. Barrasa M.I. Hughes S.H. Young R.A. Jaenisch R. Reverse-transcribed SARS-CoV-2 RNA can integrate into the genome of cultured human cells and can be expressed in patient-derived tissues Proc. Natl. Acad. Sci. U. S. A. 118 21 2021 e2105968118
95 Walls A.C. Park Y.J. Tortorici M.A. Wall A. McGuire A.T. Veesler D. Structure, function, and antigenicity of the SARS-CoV-2 spike glycoprotein Cell. 181 2 2020 281-92.e6
96 Wrobel A.G. Benton D.J. Xu P. Roustan C. Martin S.R. Rosenthal P.B. SARS-CoV-2 and bat RaTG13 spike glycoprotein structures inform on virus evolution and furin-cleavage effects Nat. Struct. Mol. Biol. 27 8 2020 763 767 32647346
97 Choi A. Koch M. Wu K. Chu L. Ma L. Hill A. Safety and immunogenicity of SARS-CoV-2 variant mRNA vaccine boosters in healthy adults: an interim analysis Nat. Med. 27 11 2021 2025 2031 34526698
98 Sztain T. Ahn S.H. Bogetti A.T. Casalino L. Goldsmith J.A. Seitz E. A glycan gate controls opening of the SARS-CoV-2 spike protein Nat. Chem. 13 10 2021 963 968 34413500
99 Gong S.Y. Chatterjee D. Richard J. Prevost J. Tauzin A. Gasser R. Contribution of single mutations to selected SARS-CoV-2 emerging variants spike antigenicity Virology. 563 2021 134 145 34536797
100 Mehdipour A.R. Hummer G. Dual nature of human ACE2 glycosylation in binding to SARS-CoV-2 spike Proc. Natl. Acad. Sci. U. S. A. 118 19 2021 e2100425118
101 Yang L. Liu S. Liu J. Zhang Z. Wan X. Huang B. COVID-19: immunopathogenesis and Immunotherapeutics Signal Transduct Target Ther. 5 1 2020 128 32712629
102 Nguyen L.H. Joshi A.D. Drew D.A. Merino J. Ma W. Lo C.H. Self-reported COVID-19 vaccine hesitancy and uptake among participants from different racial and ethnic groups in the United States and United Kingdom Nat. Commun. 13 1 2022 636 35105869
103 Coderc de Lacam E.G. Blazhynska M. Chen H. Gumbart J.C. Chipot C. When the dust has settled: calculation of binding affinities from first principles for SARS-CoV-2 variants with quantitative accuracy J. Chem. Theory Comput. 18 10 2022 5890 5900 36108303
104 Huang J. Liu J. Tian R. Liu K. Zhuang P. Sherman H.T. Correction: Huang et al. a next generation sequencing-based protocol for screening of variants of concern in autism spectrum disorder Cells. 11 20 2022 3276
105 Mehdipour A.R. Hummer G. Dual nature of human ACE2 glycosylation in binding to SARS-CoV-2 spike Proc. Natl. Acad. Sci. 118 19 2021 e2100425118
106 Acharya A. Lynch D.L. Pavlova A. Pang Y.T. Gumbart J.C. ACE2 glycans preferentially interact with SARS-CoV-2 over SARS-CoV Chem. Commun. 57 48 2021 5949 5952
107 Pinto D. Park Y.J. Beltramello M. Walls A.C. Tortorici M.A. Bianchi S. Cross-neutralization of SARS-CoV-2 by a human monoclonal SARS-CoV antibody Nature. 583 7815 2020 290 295 32422645
108 Blazkova J. Refsland E.W. Clarridge K.E. Shi V. Justement J.S. Huiting E.D. Glycan-dependent HIV-specific neutralizing antibodies bind to cells of uninfected individuals J. Clin. Invest. 129 11 2019 4832 4837 31589168
|
PMC010xxxxxx/PMC10233171.txt |
==== Front
Vis Comput
Vis Comput
The Visual Computer
0178-2789
1432-2315
Springer Berlin Heidelberg Berlin/Heidelberg
2905
10.1007/s00371-023-02905-y
Preface
Preface (Vol 39. Issue 6, June 2023)
Magnenat-Thalmann Nadia thalmann@miralab.ch
grid.8591.5 0000 0001 2322 4988 MIRALab-CUI, University of Geneva, Battelle, Building A, 7, Route de Drize, 1227 Carouge, Geneva, Switzerland
1 6 2023
2023
39 6 22052206
15 5 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2023
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIn this issue, we publish two special sections from two different special issues calls.
The first special issue call contains fifteen (15) best papers from the conference ACIAT202. Some of them are additionally selected from an open call. The theme of this special section that is published here is “Multi-Modality Feature Learning for Visual Understanding”.
The lead guest editor of this special section is Professor Zhigang Tu, from Wuhan University in China.
Other invited editors are Dr Zou Ren from Snap Inc., USA, Dr Jun Liu, from Singapore University of Technology and Design, Dr Yang Cong, from Chinese Academy in China, and Dr Lei Zhang, from Chongqing University in China.The first paper authored by Fei Liao et al. is “A systematic review on application of deep learning in digestive system image processing”. This paper is a survey paper.
The second paper authored by Fei Zhiqin Zhu et al. is “X-Net: a dual encoding–decoding method in medical image segmentation”.
The third paper authored by Fei Qi Zhu et al. is “Pairwise feature-based generative adversarial network for incomplete multi-modal Alzheimer’s disease diagnosis”.
The fourth paper authored by Fei S. Vaidya et al. is “Fingerprint-based robust medical image watermarking in hybrid transform”.
The fifth paper authored by Fei Bo Li et al. is “Object feature selection under high-dimension and few-shot data based on three-way decision”.
The sixth paper authored by Fei Linsen Xu et al. is “A convolution-transformer dual branch network for head-pose and occlusion facial expression recognition”.
The seventh paper authored by Fei Zhiqiang Ran et al. is “SAUNet + + : an automatic segmentation model of COVID-19 lesions from CT slices”.
The eighth paper authored by Fei Chao Wang et al. is “Flow-pose Net: an effective two-stream network for fall detection”.
The ninth paper authored by Fei Fan Gao et al. is “A novel infrared and visible image fusion method based on multi-level saliency integration”.
The tenth paper authored by Fei Lina Zhao et al. is “Point cloud sampling method based on offset-attention and mutual supervision”.
The eleventh paper authored by Fei J. Ranjani et al. is “A paced multi-stage block-wise approach for object detection in thermal images”.
The twelfth paper authored by Fei Yen-Wei Chen et al. is “IDH mutation status prediction by a radiomics associated modality attention network”.
The thirteenth paper authored by Fei Dhiraj Dhane et al. is “Artificial intelligence-assisted cervical dysplasia detection using papanicolaou smear images”.
The fourteenth paper authored by Fei Yafeng Zheng et al. is “VisOJ: real-time visual learning analytics dashboard for online programming judge”.
The fifteenth paper authored by Fei Deepak Gaur et al. is “A new approach to simulate the dynamic behavior of particulate matter using a canny edge detector embedded PIV algorithm”.
The second section contains five (5) papers following a call for papers on “Deep Learning for 3D Segmentation”. This special section has been led by Dr Mingqiang Wei, from Nanjing University in China. Other guest editors are:
Dr Yulan Guo from National University of Defense Technology, China, Dr Xuejin Chen from University of Science and Technology of China, China, Dr Yiping Chen from Xiamen University, China, Dr Weiming Wang from Hong Kong Metropolitan University, Hong Kong, Dr Paul L. Rosin from Cardiff University, the UK, and Dr Yong-Jin Liu from Tsinghua University, China.The first paper authored by Xuefeng Yan et al. is “PV-RCNN + + : semantical point-voxel feature interaction for 3D object detection”.
The second paper authored by Li Zhang et al. is “A class of nonstationary interproximate subdivision algorithm for interpolating feature data points”.
The third paper authored by Xiaolei Chen et al. is “Lightweight head pose estimation without keypoints based on multi-scale lightweight neural network”.
The fourth paper authored by Ningzhong Liu et al. is “GADA-SegNet: gated attentive domain adaptation network for semantic segmentation of LiDAR point clouds”.
The fifth paper authored by Dingkun Zhu et al. is “ImGeo-VoteNet: image and geometry co-supported VoteNet for RGB-D object detection”.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
|
PMC010xxxxxx/PMC10234691.txt |
==== Front
Clin Res Hepatol Gastroenterol
Clin Res Hepatol Gastroenterol
Clinics and Research in Hepatology and Gastroenterology
2210-7401
2210-741X
Elsevier Masson SAS.
S2210-7401(23)00075-X
10.1016/j.clinre.2023.102150
102150
Letter to the Editor
Long-term antibody response to inactivated SARS-CoV-2 vaccination in patients with chronic liver disease: A multicenter study
Yang Yinuo a1
Li Xuemei ab1
Zhao Xinya c
Liu Yiqing d
Zhao Tong e
Zhu Qiang a⁎
a Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Ji'nan, 250021, China
b Department of Gastroenterology, Heze Municipal Hospital, Heze, 274099, China
c Department of Radiology, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Ji'nan, 250021, China
d Department of Clinical Laboratory, Shandong Provincial Hospital Affiliated to Shandong First Medical University, Ji'nan, 250021, China
e Qilu Hospital of Shandong University, Ji'nan, 250012, China
⁎ Corresponding author at: Department of Gastroenterology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Ji'nan, 250021, China
1 These authors contributed equally to this work.
2 6 2023
8 2023
2 6 2023
47 7 102150102150
© 2023 Elsevier Masson SAS. All rights reserved.
2023
Elsevier Masson SAS
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
Patients with chronic liver disease (CLD) are at a greater risk of severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) infection. This study investigated the antibody response to inactivated SARS-CoV-2 vaccination in a long-term prospective cohort of CLD patients. The seropositivity rates and antibody concentrations of anti-SARS-CoV-2 NAbs were similar among patients with different severity of CLD 6 months after the third vaccination. In addition, older CLD patients appeared to have lower antibody responses. These data might be helpful to inform vaccine decisions for patients with chronic liver disease.
Keywords
Chronic liver disease
SARS-CoV-2
Vaccine
Antibody response
Abbreviation
CLD chronic liver disease
SARS-CoV-2 severe acute respiratory syndrome coronavirus 2
Non-ACLD nonadvanced CLD
CACLD compensated advanced CLD
DACLD decompensated advanced CLD
NAb neutralizing antibody
S/CO signal-to-cutoff ratio
IQR interquartile range
==== Body
pmcIntroduction
Patients with chronic liver disease (CLD) are at a greater risk of severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) infection, resulting in an increased risk of morbidity and mortality following infection [1,2]. SARS-CoV-2 vaccination is the mainstay of effective ways to protect individuals from SARS-CoV-2 infection and decrease the overall mortality rate [3]. CLD patients have multiple mechanisms of immune dysfunction which might lead to a lower short-term antibody response to the SARS-CoV-2 vaccine [3], and antibody levels gradually decrease over time after receipt of SARS-CoV-2 vaccines [4]. However, the long-term antibody response induced by the inactivated SARS-CoV-2 vaccine in CLD patients has not yet been investigated, and the understanding of the sustainability of SARS-CoV-2 vaccines in CLD patients is urgently needed. In the current work, we aimed to investigate the antibody response to inactivated SARS-CoV-2 vaccination in a long-term prospective cohort of CLD patients.
Materials and methods
In this prospective multicenter study, patients with chronic liver disease and healthy controls were enrolled from Shandong Provincial Hospital, Heze Municipal Hospital, and Qilu Hospital of Shandong University. Demographic and clinical data were collected from the electronic medical record. All participants received the initial two doses of inactivated SARS-CoV-2 vaccines (CoronaVac or BBIBP-CorV) between June 2021 to September 2021. The third dose of inactivated SARS-CoV-2 vaccine was administered 196 days (interquartile range [IQR], 188–211 days) after the second dose. Participants who were pregnant, less than 18 years old, with malignant tumors or other major diseases, with previous COVID-19 infection, and with a history of receiving systemic immunosuppressants or systemic immunoglobulins were excluded. The participants were monitored for SARS-CoV-2 infection by polymerase chain reaction. According to Fibrosis-4 score and previous history or current history of hepatic decompensation, patients with different severity of CLD were divided into three groups: nonadvanced CLD (non-ACLD), compensated advanced CLD (CACLD), or decompensated advanced CLD (DACLD) [5]. Serum SARS-CoV-2 neutralizing antibody (NAb) and anti-spike IgG antibody were measured at 6 months (±30 days) after the third dose of vaccination using chemiluminescence immunoassay (Maccura Biotechnology Co., Ltd., China). NAb concentrations above 6.00 AU/mL and anti-spike IgG levels above 1.00 S/CO were considered positive. The study was approved by the ethics committees of the participating centers and all patients provided written informed consent before the study procedures.
Results
Between June 2021 to September 2021, four hundred and twenty-one participants were recruited. During the follow-up period, 46 individuals were lost to follow-up and 8 individuals died (3 died of hepatic failure, 2 died of cardiovascular causes, 1 died of variceal bleeding, 1 died of peritonitis, and 1 died of multisystem organ failure). No participant tested positive for SARS-CoV-2 infection by polymerase chain reaction during the follow-up period. The remaining 261 CLD patients and 106 healthy controls were included in this study (Supplementary Table 1). In CLD patients, 149 (57.1%) had non-ACLD, 79 (30.3%) had CACLD and 33 (12.6%) had DACLD. These CLD patients included 175 men (67.0%) and 86 women (33.0%). The most common etiology of CLD was hepatitis B virus infection (67.8%), followed by non-alcoholic fatty liver disease (16.9%). The time from the second SARS-CoV-2 vaccination to blood collection was 384.0 days (IQR, 369.5–392.0 days). The time from the third SARS-CoV-2 vaccination to blood collection was comparable among patients with different severity of CLD and healthy controls (p = 0.731).
The seropositive rates of NAb were 79.2% (84 of 106) in healthy control group, 74.5% (111 of 149) in non-ACLD group, 73.4% (58 of 79) in CACLD group, and 66.7% (22 of 33) in DACLD group, respectively, with no significant difference (p = 0.50) (Fig. 1A ). The seropositive rates of anti-spike IgG were 82.1% (87 of 106) in healthy control group, 79.2% (118 of 149) in non-ACLD group, 75.9% (60 of 79) in CACLD group, and 72.7% (24 of 33) in DACLD group, respectively, also with no significant difference (p = 0.61) (Fig. 1B). After adjusting for age and BMI, the differences remained non-significant (both p > 0.05).Fig. 1 Long-term antibody response to the inactivated SARS-CoV-2 vaccine. (A, B) Seropositive rates of NAb and anti-spike IgG among patients with different severity of CLD and healthy controls. (C, D) NAb concentrations and anti-spike IgG levels among patients with different severity of CLD and healthy controls. (E, F) Subgroup analyses of NAb concentrations in CLD patients. Circles indicate individual antibody responses. NAb concentrations above 6.00 AU/mL and anti-spike IgG levels above 1.00 S/CO ratio were considered positive.
CACLD, compensated advanced CLD; CLD, chronic liver disease; DACLD, decompensated advanced CLD; NAb, neutralizing antibody; Non-ACLD, nonadvanced CLD.
Fig 1
The NAb concentrations were 16.44 AU/mL (IQR, 6.72–48.04 AU/mL) in healthy control group, 14.69 AU/mL (IQR, 5.98–36.97 AU/mL) in non-ACLD group, 11.74 AU/mL (IQR, 4.32–26.51 AU/mL) in CACLD group, and 10.68 AU/mL (IQR, 3.54–31.93 AU/mL) in DACLD group, respectively (Fig. 1C). The anti-spike IgG levels were 3.70 S/CO (IQR, 1.50–6.84) in healthy control group, 2.99 S/CO (IQR, 1.19–6.22) in non-ACLD group, 2.22 S/CO (IQR, 1.03–5.53) in CACLD group, and 2.24 S/CO (IQR, 0.43–5.26) in DACLD group, respectively (Fig. 1D). Neither the NAb concentrations nor the anti-spike IgG levels were significantly different among the four groups (p = 0.25 and p = 0.11, respectively).
In subgroup analyses, older CLD patients (age > 40 years) appeared to have a lower level of NAb concentration as compared to younger patients (11.17 AU/mL [IQR, 4.70–27.47 AU/mL] vs. 21.54 AU/mL [IQR, 7.39–53.14 AU/mL], p = 0.003) (Fig. 1E). Male CLD patients showed a lower level of NAb concentration than female patients (11.96 AU/mL [4.77–28.71 AU/mL] vs. 18.85 AU/mL [IQR, 6.97–43.50 AU/mL], p = 0.04) (Fig. 1F).
Univariate and multivariate logistic regression analyses were performed to select risk factors associated with NAb negativity in CLD patients. After taking the time from the third vaccination to measurement into consideration, age (OR, 1.03; 95% CI, 1.00–1.05; p = 0.025) was identified as an independent risk factor for NAb negativity, while the severity of CLD was not a risk factor (Supplementary Table 2). Similar results were also observed for anti-spike IgG negativity in CLD patients (Supplementary Table 3).
Discussion
This is the first study assessing the long-term antibody response in CLD patients after three doses of inactivated SARS-CoV-2 vaccination. We found that there was no notable difference in antibody response 6 months after the completion of inactivated SARS-CoV-2 vaccination among patients with different severity of CLD and healthy controls. Yet, a previous study has demonstrated that CLD patients had lower antibody response to SARS-CoV-2 vaccination than healthy controls shortly after the second vaccine dose [3]. A similar phenomenon was observed in multiple sclerosis patients treated with cladribine tablets [6]. The possible explanation might be that the third vaccine dose may overcome the negative effect of immune dysfunction in CLD patients [7,8].
This study has several limitations. First, the sample size in our study may not be large enough. Larger cohort of CLD patients with different etiologies need to be further observed. Second, we had no access to data on antibodies for each month, so the dynamic changes in antibodies could not be monitored. Third, we had no data on the vaccine efficacy for preventing laboratory-confirmed COVID-19 in CLD patients. We will continue to investigate the SARS-CoV-2 vaccine efficacy in these patients.
In conclusion, long-term antibody response to inactivated SARS-CoV-2 vaccination was similar among patients with different severity of CLD and healthy controls. These data might be helpful to inform vaccine decisions for patients with chronic liver disease, especially for patients with hepatitis B.
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.
Appendix Supplementary materials
Image, application 1
Funding
This work was supported by the National Natural Science Foundation of China (No. 82160124).
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.clinre.2023.102150.
==== Refs
References
1 Ge J. Pletcher M.J. Lai J.C. Outcomes of SARS-CoV-2 infection in patients with chronic liver disease and cirrhosis: a national COVID cohort collaborative study Gastroenterology 161 5 2021 1487 1501 10.1053/j.gastro.2021.07.010 e5 34284037
2 Marjot T. Webb G.J. ASt Barritt Moon A.M. Stamataki Z. Wong V.W. COVID-19 and liver disease: mechanistic and clinical perspectives Nat Rev Gastroenterol Hepatol 18 5 2021 348 364 10.1038/s41575-021-00426-4 33692570
3 Ai J. Wang J. Liu D. Xiang H. Guo Y. Lv J. Safety and immunogenicity of SARS-CoV-2 vaccines in patients with chronic liver diseases (CHESS-NMCID 2101): a multicenter study Clin Gastroenterol Hepatol 20 7 2022 1516 1524 10.1016/j.cgh.2021.12.022 e2 34942370
4 Silva M.F.S. Pinto A. de Oliveira F.C.E. Caetano L.F. Araújo F.M.C. Fonseca M.H.G. Antibody response 6 months after the booster dose of Pfizer in previous recipients of CoronaVac J Med Virol 2022 10.1002/jmv.28169
5 Bastati N. Beer L. Mandorfer M. Poetter-Lang S. Tamandl D. Bican Y. Does the functional liver imaging score derived from gadoxetic acid-enhanced MRI predict outcomes in chronic liver disease? Radiology 294 1 2020 98 107 10.1148/radiol.2019190734 31743083
6 Brill L. Rechtman A. Shifrin A. Rozenberg A. Afanasiev S. Zveik O. Longitudinal humoral response in MS patients treated with cladribine tablets after receiving the second and third doses of SARS-CoV-2 mRNA vaccine Mult Scler Relat Disord 63 2022 103863 10.1016/j.msard.2022.103863
7 John B.V. Ferreira R.D. Doshi A. Kaplan D.E. Taddei T.H. Spector S.A. Third dose of COVID-19 mRNA vaccine appears to overcome vaccine hyporesponsiveness in patients with cirrhosis J Hepatol 77 5 2022 1349 1358 10.1016/j.jhep.2022.07.036 36181987
8 Wand O. Einbinder Y. Nacasch N. Halperin T. Erez D. Grupper A. Kinetics of the humoral response 1-year following vaccination with BNT162b2 SARS-CoV-2 vaccine among maintenance hemodialysis patients J Nephrol 2022 10.1007/s40620-022-01377-y
|
PMC010xxxxxx/PMC10234840.txt |
==== Front
Eur Econ Rev
Eur Econ Rev
European Economic Review
0014-2921
0014-2921
International Monetary fund. Published by Elsevier B.V.
S0014-2921(23)00128-9
10.1016/j.euroecorev.2023.104499
104499
Article
Policy Packages and Policy Space: Lessons from COVID-19
Bergant Katharina
Forbes Kristin ⁎
1 International Monetary Fund
2 MIT-Sloan School of Management, NBER and CEPR
⁎ Correspondence
2 6 2023
2 6 2023
10449930 10 2022
29 4 2023
15 5 2023
© 2023 International Monetary fund. Published by Elsevier B.V. All rights reserved.
2023
Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active.
This paper uses the onset of COVID-19 to examine how countries construct their policy packages in response to a severe negative shock. We use several new datasets to track the use of a large variety of policy tools: announced fiscal stimulus (both above- and below-the-line), monetary policy (through interest rates, asset purchases, liquidity support and swap lines), foreign currency intervention, adjustments to macroprudential regulations (including the countercyclical capital buffer) and changes in capital controls (on inflows and outflows). The results suggest that pre-existing policy space was usually more important than other country characteristics and the extent of “stress” (in economic, financial, and health measures) in determining how a country responded to COVID-19. The notable exception is for fiscal stimulus, for which existing policy space did not act as a significant constraint in advanced economies. This is a sharp contrast to results for earlier episodes—although advanced economies with higher debt levels may have been constrained in how they provided stimulus (with more below-the-line commitments). Moreover, the use of (and space available) for each policy tool usually did not affect a country's use of other policies. This suggests that countries are not coordinating their tools optimally in an integrated framework, especially when policy space is limited for certain tools.
Keywords
COVID-19
pandemic
policy space
fiscal
monetary
macroprudential
capital controls
==== Body
pmcI Introduction
When1 an economy is under stress—whether from slower growth, rising unemployment, a financial crisis or pandemic—policymakers respond by choosing from a variety of tools. Traditional economic responses often include some combination of: fiscal policy, monetary policy, foreign currency (FX) intervention, and adjustments to macroprudential regulations and/or capital controls. During a pandemic, policymakers may also choose to restrict activity and adopt other measures to restrain the spread of the disease. This paper tests what determines a country's policy package, including whether the choice of tools is constrained by the available “policy space”2 . If pre-existing policy space affects a country's ability to announce and use beneficial tools during periods of stress, countries may need to place greater weight on adjusting policy sooner to create space (such as by reducing debt levels, raising interest rates, building foreign exchange reserves, and/or adjusting macroprudential policy and capital controls).
The results in this paper suggest that pre-existing policy space is usually more important than other country characteristics and the extent of “stress” (in economic, financial, and health measures) in determining how a country responds to a shock. More specifically, policy space was the most important determinant of the extent to which countries used FX intervention, lowered interest rates, and loosened macroprudential policy to support their economies during the initial phase of COVID-19. Policy space affected not only the magnitude by which specific tools were adjusted, but also the form of adjustment—such as whether monetary easing was pursued more through reductions in interest rates or asset purchases, or whether fiscal stimulus was pursued more through on-budget measures or “below-the-line” measures (such as loans, equity, and credit guarantees). Policy space was generally not important, however, in determining the magnitude of fiscal stimulus announced in advanced economies during the early stage of the pandemic (especially for the extent of below-the-line fiscal policy). This finding contrasts with research focusing on other periods, which generally finds that fiscal policy space significantly constrains a country's ability to respond to negative shocks (Romer and Romer, 2018; Romer and Romer, 2019; Jordá et al., 2016). The results also suggest that the space available to use one type of policy tool usually did not affect a country's decision to use other policies during COVID-19. For example, the amount of space a country had for conventional monetary policy (i.e., lowering interest rates) had no significant impact on the size of its fiscal response to COVID-19, the amount of any FX intervention, or its adjustments to macroprudential regulations. This finding is in contrast to standard theoretical models, which suggest that certain policies could partially substitute for others whose use is constrained.3
This paper begins by analyzing the policy responses to COVID-19 in the first half of 2020. This is a unique case study as the pandemic resulted from an external shock that did not reflect domestic policies or imbalances, allowing for a cleaner identification of how policy space affects policy responses than during most negative shocks (which often reflect domestic imbalances and policy choices). The severity of the shock also motivated large and multifaceted policy responses, as well as the creation of detailed cross-country data sets tracking these responses. We focus on six sets of policy tools: announced fiscal stimulus (measured in aggregate, above-the-line, and below-the-line); monetary stimulus (through policy interest rates, asset purchases, liquidity support, and swap lines); FX intervention (including the decision to intervene and corresponding magnitude); macroprudential regulation (defined broadly or focusing on the countercyclical capital buffer (CCyB)); capital controls aimed at reducing net capital outflows (on either gross inflows or outflows) and various “containment” measures targeting the spread of the virus.
We document the prevalence and magnitudes of these different policy responses to COVID-19 in advanced economies and emerging markets, drawing heavily on the IMF's Policy Tracker, IMF's Fiscal Monitor, and Oxford's Coronavirus Government Response Tracker. Almost all countries announced a large fiscal stimulus, averaging about 11% of GDP, and split evenly (on average) between above-the-line measures (i.e., on-budget increases in spending and foregone revenue) and below-the-line measures (i.e., loans, equity infusions and credit guarantees).4 Countries also provided monetary stimulus through a range of tools; most central banks lowered their policy interest rates (albeit by a relatively small 1.7pp on average) and over 85% of the sample provided some type of liquidity support for banks. In addition, 43% used quantitative easing (including 61% of advanced economies and 33% of emerging markets) and 42% used swap lines. Another widely used policy response was easing macroprudential regulations—with 72% of advanced economies and 61% of emerging markets reporting some loosening in regulations (including reductions in the CCyB in about 30% of the sample). Using FX reserves to support the exchange rate was employed in just over half of the emerging markets, but only in three advanced economies. The one policy that was not widely adjusted in any group of countries during this period was capital controls—with only four countries reducing controls on capital inflows and two tightening controls on capital outflows. In addition to these standard economic policy responses to negative shocks, all countries adopted containment measures to address the health aspects of the pandemic. Emerging markets adopted stricter health and containment measures on average, including mobility restrictions, restrictions of public events, and testing and tracing regimes.
One striking result from this analysis of the policy responses to COVID-19 is the substantial variation in how different countries responded. Taking the example of fiscal policy, although all countries announced some fiscal stimulus, the size of the stimulus ranged from 1% to 37% of GDP, and the share of stimulus that was above-the-line ranged from only 3% to 100%. For monetary policy, most countries lowered their policy interest rates, but the average reduction of 166bp includes one country that lowered its rate by 2277bp and another that raised by 25bp. Of the emerging markets that used FX reserves to intervene in currency markets, some used large amounts of reserves to mitigate depreciation pressures (with the largest loss equal to 8.3% of GDP), while others accumulated reserves to mitigate appreciation pressures (with the largest gain reaching 3.6% of GDP).5 The variation for advanced economies was even larger, with the change in FX reserves (relative to GDP) ranging from a loss of 5.7% to a gain of 12.6%. What explains this variation in policy responses to the COVID-19 pandemic?
To better understand this variation, this paper focuses on three sets of factors determining a country's policy response to COVID-19: policy space (for the given tool as well as for other policy tools), the extent of economic, financial, and health stress during the early stages of the pandemic, and other country characteristics. We find that the extent of “health stress” (i.e., the reported number of COVID-19 cases) is a significant determinant of the extent of health and containment measures, and the extent of “financial stress” can impact whether countries report using FX intervention. In contrast, the extent of “economic stress” (as measured by the change in forecast GDP growth for the current year), is not significantly correlated with any policies—including the extent of fiscal or monetary stimulus. In a few cases, certain country characteristics are also significantly correlated with the use of some policies, such as countries with stronger institutional quality being more likely to provide stimulus through above-the-line measures and less likely to use FX intervention.
The most important and consistently significant determinant of the use, form, and magnitude of most policy tools, however, is the extent of “policy space” for the given tool. Countries with a higher policy interest rate before the pandemic lowered this policy rate by more and relied less on other forms of monetary stimulus, such as asset purchases and liquidity provision to banks. Countries with a tighter macroprudential stance (or higher CCyB) were more likely to ease macroprudential regulation (and lowered the CCyB by more). Countries with a larger reserve stockpile (relative to GDP) were more likely to use FX intervention (although not always in the expected direction). In sharp contrast, advanced economies with less fiscal policy space (as measured by debt-to-GDP ratios or other standard metrics) were not constrained in their announcements of fiscal stimulus, especially in their use of below-the-line fiscal policy. This differs from results in previous work showing that fiscal space is a significant constraint on the fiscal response to periods of stress and crises (e.g., Romer and Romer, 2018 and 2019; Jordà et al., 2016). Emerging markets with higher pre-existing debt levels, however, were more constrained in their use of fiscal policy in response to COVID-19, especially in their use of below-the-line measures.
In addition to these results on the importance of “own-policy space” in the use of most policy tools, another key set of results is that “other-policy space” (i.e., for other policy tools) was usually not significant in determining the use of individual tools in response to COVID-19. While most countries adjusted a range of tools simultaneously (with an average adjustment of 6.8 tools in our sample), there seemed to be little coordination between the use of these tools6 —despite recent arguments and economic models suggesting that the ability to use other tools should factor into policy choice (Basu et al., 2020). For example, countries with less space to lower interest rates did not use fiscal policy more aggressively, and countries with higher debt levels did not use any form of monetary policy more aggressively. This suggests that countries are not following the predictions of standard economic models suggesting that they should rely more on fiscal stimulus when monetary policy is constrained and/or when interest rates are low, and that they should rely more on monetary stimulus when debt levels are high (e.g., Aizenman et al., 2019; Auerbach and Gorodnichenko, 2017; Bartsch et al., 2020). In the same spirit, the loosening of macroprudential tools (especially the CCyB) only depended on how high countries set their buffers before the pandemic, but not on the monetary stance or the level of capital flow measures, and a more stringent macroprudential stance did not influence monetary policy actions (as suggested in Aizenman et al., 2020; Bergant et al., 2020). Some countries even used policies that seemed to work in opposite directions (such as lowering interest rates while intervening to appreciate the currency). This series of results suggests that countries decide on the use of each tool independently and are not coordinating the use of the tools or utilizing them in an integrated framework. More specifically, when certain tools are constrained, countries do not appear to be more likely to use other tools that could act as substitutes. Similarly, when certain tools are more effective when used in conjunction with other tools, countries do not appear to be more likely to use them as complements. This apparent lack of coordination is even true for tools that are often implemented by the same organization.7
This paper's findings are subject to one important caveat; our analysis focuses on the determinants of different policy responses and construction of a policy package in immediate response to a shock, but does not analyze the efficacy of these responses, including how the use of other policies could affect this efficacy or whether the policies were used optimally.8 For example, the literature on fiscal multipliers (summarized in Ramey, 2019) finds that fiscal stimulus is less effective in countries with higher debt levels. Another set of papers argues that fiscal policy is more effective when interest rates are low and the output gap is larger (Bouakez et al., 2017; Eggertsson, 2011; Woodford. 2011; Drautzburg and Uhlig, 2015). An extensive literature also analyzes the role of automatic stabilizers, which have shown to be effective in alleviating economic stress in a timely, targeted, and temporary manner (Maravalla and Rawdanowicz, 2020). Conceptually, more social insurance against unexpected shocks can be optimal when the level of economic activity is more responsive to social spending during negative shocks (McKay and Reis, 2020), when policy rates are low (Blanchard and Summers, 2020), or when debt levels are already high (Bi et al., 2016). For countries with larger automatic stabilizers, the optimal discretionary response to Covid-19 may have been smaller (Heinemann, 2022 and Bouabhdallah et al., 2020). We do not consider these interactions, many of which will take time to fully assess (such as the lagged impact on borrowing costs and productivity). Instead, this analysis focuses on what factors affected which policies were announced and adopted in the early stages of COVID-19, as well as the magnitude and form by which each policy was initially implemented.
To conclude, the results in this paper suggest that for most policy responses to shocks, policy space is an important determinant of not only whether a country uses a tool, but the form and extent by which it adjusts that tool. More specifically, countries that raised policy interest rates, tightened macroprudential policy, and accumulated FX reserves before 2020 were more able to adjust these respective instruments to support their economies when COVID-19 spread. Countries that had more space to provide monetary stimulus through the “conventional” tool of reducing policy interest rates relied less on other forms of monetary policy, such as by enacting smaller asset purchase programs and being less likely to provide liquidity to banks. This suggests that as countries recover from negative shocks, they should place some weight on unwinding and tightening these different tools when appropriate, so that they will have the ability to use these tools to respond to shocks in the future. Finally, the noteworthy exception to this key result on the importance of policy space is for the size of fiscal stimulus announced in the first half of 2020 (as well as for realized spending over the first 1 ½ years of the pandemic). It is unclear if the reduced constraint of fiscal space during the initial response to COVID-19 was temporary and related to unique aspects of the pandemic or a longer lasting phenomenon. This is an important topic for future research.
The remainder of the paper is as follows. Section II summarizes several streams of related literature. Section III describes policy responses to COVID-19—including new data sources and patterns across countries. Section IV analyzes the factors determining the use of individual policy tools during the pandemic—including the methodology, baseline results, sensitivity tests, and a closer look at the results for fiscal policy. Section V extends this analysis to incorporate the joint use of and interactions between different policies and the policy space available for multiple tools. Section VI concludes.
II Related Literature
The analysis in this paper draws on four related veins of literature: on policy responses to shocks; on the role of policy space (which primarily focuses on fiscal policy); on the interaction between policy space and the use of different policy tools; and on policy responses to COVID-19.
The literature on policy responses to shocks is extensive, although most papers only consider a subset of the policy tools analyzed in this paper and often focus on the multiplier effects of individual policies rather than the choices between different policy tools.9 Most closely related to this paper, Aizenman and Jinjarak (2011) examines the wide variation in fiscal and exchange rate responses to the 2008-2009 crisis and shows that countries with greater trade openness had smaller fiscal stimulus and larger depreciations—as predicted in a neo-Keynesian open-economy model. Aizenman et al. (2019) focuses on different fiscal policy responses and includes an excellent summary of this literature, including the role of fiscal space. A branch of this literature focuses on the responses of emerging markets to periods of sharp capital outflows (such as Forbes and Klein, 2015) or large capital inflows (such as Ghosh et al., 2017). These papers are similar to this analysis in incorporating a larger set of policy responses (including exchange rate intervention, currency adjustments, capital controls and macroprudential policy, in additional to monetary and fiscal policy), but generally do not incorporate the role of policy space or include advanced economies. Ongoing work at the IMF on the Integrated Policy Framework also focuses on emerging markets and ties together much of this literature by modelling how country characteristics determine the optimal combination of policy responses to a range of different shocks.10
A second (and closely related) strand of literature focuses on the constraints from prior policy actions and policy space. More specifically, as interest rates fell to near zero in many countries in the 2010s, there was increased attention to the space available for monetary policy to adjust to shocks and the potential for unconventional tools to provide stimulus if traditional tools were constrained (Bernanke, 2020). Another branch of this literature focuses on how fiscal space can constrain the use of fiscal policy. Ghosh et al. (2013) and Kose et al. (2022) discuss different approaches for defining fiscal space, and Auerbach and Gorodnichenko (2017) provides an excellent review of the literature, including an analysis of the interaction between fiscal stimulus and fiscal space at different stages of the business cycle. Aizenman and Jinjarak (2011) finds that countries with more fiscal space, as measured by the inverse of the average tax-years it would take to repay the public debt, responded to the 2008-2009 crisis with larger fiscal stimulus. Romer and Romer (2018, 2019) consider longer time horizons and show that countries with more fiscal and monetary policy space (measured by debt to GDP and if interest rates are above zero, respectively) have significantly better economic performance after periods of stress, partly because monetary and fiscal policy can be used more aggressively to support the economy. Romer and Romer (2019) argues that this constraint from fiscal space occurs partly because of the impact on market access, and partly through policymaker decisions (such as the need to abide by EU or IMF conditionality rules). These conclusions agree with Jordà et al. (2016), which analyzes a longer period to show that countries with lower debt ratios respond to crises with more aggressive fiscal stimulus (through financial rescues as well as conventional tax cuts and spending increases), leading to smaller output losses. The conclusion from this literature is that maintaining fiscal space during normal times can be a valuable insurance that allows for stronger responses to financial crises and recessions.
A third focus of this literature has been how constraints on the use of one policy tool can affect not only the use of that specific tool, but also the selection of other policy tools. This interaction of space and tools received increased attention as countries struggled to raise interest rates and reduce debt burdens after the 2008-2009 crisis. More specifically, several papers highlight the increased role for countercyclical fiscal policy when interest rates are near zero (Bouakez et al., 2017; Eggertsson, 2011; Woodford, 2011; Drautzburg and Uhlig, 2015, Bernanke, 2020; and Furman and Summers, 2020). Related research also shows how monetary policy that affects borrowing costs can affect fiscal space and therefore a country's ability to use fiscal stimulus (Aizenman et al., 2019; Auerbach and Gorodnichenko, 2017). Bartsch et al. (2020) provides an overview of issues around the optimal mix of countercyclical fiscal and monetary policy, highlighting how the tradeoffs change when policy rates are at their effective lower bound. This analysis also discusses the institutional constraints in attaining the optimal fiscal-monetary policy mix. Aizenman et al. (2017) and Bergant at al. (2020) show that a tighter macroprudential stance enables countries to use a more independent monetary policy when hit by global financial shocks. This literature, however, generally focuses on the interaction of two individual policy actions, but ignores the range of other policy tools that are included in this paper. The one notable exception is the IMF's recent work on the Integrated Policy Framework, which focuses on the interactions of various policy tools for emerging markets under certain conditions (IMF, 2020b).
Finally, a very recent and rapidly growing literature examines policy responses to the COVID-19 pandemic.11 A few prominent examples showing the range of this research include: English et al., (2022), discusses the different monetary and macroprudential responses around the world; Wieland (2022) assesses the fiscal response in the Euro area; Hong and Lucas (2023) explores the size and impact of credit and liquidity policies; Kirti et al. (2022a) provides detailed information on a wide set of policies adopted by over 70 countries during 2020; Bigio et al. (2020) models the advantages of lump-sum transfers versus credit policy; Auerbach et al. (2021) models how different fiscal policies interact with inequality; Jordà and Nechio (2022) analyzes the impact of fiscal transfers on inflation; Altavilla et al (2020a) evaluates the impact of various policies on bank lending conditions; Gourinchas et al. (2020), focuses on how different policies impact business failures; Eichenbaum et al. (2021) models the efficacy of containment policies; and Guerrieri et al. (2022) examines the efficacy of various fiscal and monetary policies.12
Benmelech and Tzur-Ilan (2020) is the paper closest to the first section of this paper. It explores the fiscal and monetary responses to COVID-19 and is the only other paper to date (to our knowledge) to incorporate some analysis of the role of policy space. It finds that high-income countries announced larger fiscal policies and did not appear to be constrained by high debt-to-GDP ratios (as we also find for advanced economies, but in contrast to most of the literature examining earlier shocks). Benmelech and Tzur-Ilan (2020) also finds that countries with low interest rates before 2020 lowered their interest rates by less and were more likely to use unconventional monetary policy tools. In contrast to our results, they also find evidence that countries with lower interest rates were more likely to relax macroprudential regulations and enact larger fiscal stimulus (primarily through government guarantees). While Benmelech and Tzur-Ilan (2020) focuses primarily on the factors affecting the use of fiscal and monetary policy, our paper analyzes the determinants of a broader set of tools (FX intervention, capital controls, macroprudential policies and containment measures), as well as the role of different measures of stress (financial, economic, and health). This allows a more comprehensive assessment of the focus of this paper: how the policy space for each tool and different forms of stress affected the choice of (and interaction between) a greater range of policy responses.
III Policy Choices During the COVID-19 Pandemic: The Data
As the severity of COVID-19 became apparent and financial markets reacted sharply, governments around the world evaluated how best to support their economies and minimize the damage to health, employment and incomes. This section documents the use of six sets of policies used during the early stage of the pandemic. The severe global nature of the shock prompted the creation of several new data sets, particularly the IMF's Policy Tracker, which provide a wealth of detailed, cross-country information on policy choices during this period.13
To assess how countries responded to COVID-19, we focus on six sets of responses: announced fiscal policy, monetary policy, FX intervention, macroprudential policy, capital controls, and “containment” measures. For the first five policies, we concentrate on actions aimed at providing stimulus, easing monetary conditions, and/or stabilizing the economy, such as any fiscal stimulus, monetary stimulus, using FX reserves to stabilize exchange rate movements, loosening macroprudential regulations and alleviating pressures from net capital outflows. For “containment” measures, we focus on economic and health policies aimed at containing the spread of the virus, namely restrictions on activity and test-and-tracing requirements. We include as large a sample as possible for each policy response, with the resulting dataset covering up to 75 countries. Appendix Table 1 provides more detail on the data discussed in this section and used throughout this paper.
A first, and widely used, response to the pandemic was fiscal policy. To measure the fiscal response, we use the announced change in the 2020 fiscal balance in response to COVID-19 (as a share of 2019 GDP), as measured in June 2020 relative to end-2019.14 This measures the additional fiscal support relative to what was planned at end-2019 and can be broken into above-the-line commitments (additional spending and foregone revenue) and below-the-line commitments (loans, equity injections, asset purchases, debt assumption and contingent liabilities). These fiscal measures only include discretionary measures and do not incorporate any support through automatic stabilizers or revenue losses from slower growth. It is worth highlighting that these measures are the announced fiscal support—even if the program was not fully utilized or drawn down (as occurred in a number of countries, especially for programs involving credit and liquidity support).15 Figure 1 shows the fiscal stimulus for 40 countries using these measures, with more detailed statistics in Table 1 . Most countries announced a large stimulus, with an average size of 11% of GDP across the sample. Seventeen countries provided stimulus over 10% of GDP and five provided stimulus of at least 20% of GDP. On average, countries split this stimulus almost evenly between above-the-line and below-the-line measures. There is a large variance in each of these measures, however, with the overall stimulus ranging from only 1% of GDP (for Mexico) to 37% of GDP (for Germany), and the share of stimulus that is above-the-line ranging from only 3% (for Turkey) to 100% (for Georgia).Figure 1 Fiscal Response to COVID-19 as % of GDP. Notes: Fiscal intervention in 2020q1 and 2020q2 in response to COVID-19 as % of 2019 GDP. Fiscal intervention is the announced fiscal support relative to what was planned at end-2019 and is broken into two components: additional spending and foregone revenue (also referred to as above-the-line) and loans, equity and credit guarantees (also referred to as below-the line). These fiscal measures only include discretionary measures and not any support through automatic stabilizers or revenue losses corresponding to slower growth. Source: Based on data from the IMF's Fiscal Monitor Database of Country Fiscal Measures in Response to the COVID-19 Pandemic.
Figure 1
Table 1 Policy Responses: Summary Statistics
Table 1Policy Measure Unit Mean Median St. Dev. Min Max Sample
Full Sample
Total Fiscal Measures % of GDP 10.90 8.82 8.41 0.92 37.21 40
of this "Above the Line" % of total fiscal spending 49.55 49.12 24.18 2.55 100.00 40
Change in Monetary Policy Rate Percentage points -1.66 -1.00 3.28 -22.77 0.25 53
Shadow Rate where available Percentage points -1.24 -1.97 1.68 -2.99 1.72 8
Central Bank Asset Purchases % of GDP 2.31 0.00 4.40 0.00 22.40 55
Net FX Purchases % of GDP -0.16 0.00 3.04 -12.60 8.34 57
Change in CCyB Percentage points -0.27 0.00 0.59 -2.50 0.00 70
Health & Containment measures Index 56.62 56.25 16.25 24.31 87.50 73
Advanced Economies
Total Fiscal Measures % of GDP 15.93 12.97 8.87 7.41 37.21 20
of this "Above the Line" % of total fiscal spending 43.01 36.18 23.84 9.35 83.41 20
Change in Monetary Policy Rate Percentage points -0.74 -0.65 0.71 -2.00 0.25 19
Shadow Rate where available Percentage points -1.24 -1.97 1.68 -2.99 1.72 8
Central Bank Asset Purchases % of GDP 4.61 2.80 6.54 0.00 22.40 19
Net FX Purchases % of GDP -1.59 -0.07 4.09 -12.60 5.65 19
Change in CCyB Percentage points -0.46 0.00 0.75 -2.50 0.00 35
Health & Containment measures Index 48.90 47.23 14.00 24.31 72.22 36
Emerging Markets
Total Fiscal Measures % of GDP 5.87 4.72 3.63 0.92 14.21 20
of this "Above the Line" % of total fiscal spending 56.10 58.97 23.27 2.55 100.00 20
Change in Monetary Policy Rate Percentage points -2.17 -1.00 3.99 -22.77 0.25 34
Shadow Rate where available Percentage points . . . . . .
Central Bank Asset Purchases % of GDP 1.09 0.00 1.87 0.00 5.80 36
Net FX Purchases % of GDP 0.55 0.20 2.07 -3.58 8.34 38
Change in CCyB Percentage points -0.07 0.00 0.25 -1.00 0.00 35
Health & Containment measures Index 64.13 67.36 14.82 29.17 87.50 37
Notes: Table reports magnitudes of announced policy responses over 2020q1-2020q2. Statistics for each group only include countries that have the ability to adopt each set of policies, e.g., individual countries in the Euro area cannot pursue monetary policy or FX intervention, but can adopt other policies. The Euro area is included as a "country" that can pursue monetary and FX policy, but not other policies.
Sources: Fiscal policies are from the IMF Fiscal Monitor. Changes in the policy rate are from Haver and the shadow rates are from Krippner (2015). Data on Central Bank Asset Purchases are from Central Bank websites and Fratto et al. (2021). FX purchases are from Adler et al. (2021). Data on the CCyB are from the BIS and ESRB. Data on the Health and Containment measures are from Oxford. See Appendix Table 1 and notes to Figure 1, Figure 2, Figure 3, Figure 4, Figure 5 for additional information.
A second, and also widely used, policy response was monetary policy—both “conventional” changes in policy interest rates and several forms of “unconventional” policy. To measure the “conventional” response, we focus on the change from 2019Q4 through 2020Q2 for two measures: (1) the central bank's main policy interest rate (from Haver) and (2) the shadow interest rate for countries at their lower bound (from Leo Krippner's website16 ) combined with the policy interest rate in countries for which a shadow rate is not available. To measure monetary stimulus through “unconventional tools” we use four different measures: the amount of asset purchases scaled by GDP (also referred to as quantitative easing or QE)17 ; a dummy for the announcement of new asset purchases; a dummy if the country injected liquidity into its banking system; and a dummy if the country activated a swap line. These dummy variables are constructed by scrapping information from the IMF's Policy Tracker.18 For all of these policy tools, we exclude countries that do not have an independent monetary policy (e.g., members of the Euro area), but include the Euro area as an independent entity.
Figure 2 graphs changes in the policy interest rates (including shadow rates when applicable), with more detailed information in Table 1. Most countries lowered their main policy rate, with the exceptions of Sweden and Kazakhstan (which raised their rates 25 basis points), and six other countries which had no changes. The magnitude of most reductions in policy rates was small, with 41 of the 52 countries lowering their rates between 0 and 4 percentage points (pp). The exceptions were Pakistan and Ukraine, which had much larger reductions of over 5pp.19 For the eight countries with data on shadow rates, those rates declined by an average of 1.2pp, about twice as much as for policy rates in the same countries. The right side of Figure 3 (with supporting data in Appendix Table 2) provides more information on the “unconventional” monetary responses. It reports the share of countries that implemented the three forms of unconventional monetary policy based on the dummy variable indicators, broken into advanced economies (AEs) and emerging markets (EMs). Over 80% of both AEs and EMs implemented liquidity support for banks, and AEs also made widespread use of asset purchases (61%) and swap lines (71%). Although fewer EMs used asset purchases and swap lines, the 33% of EMs adopting asset purchases is noteworthy as most EMs had not previously used asset purchases and all had policy interest rates above zero (including ten with interest rates above 1%). Table 1 provides more information on these asset purchase programs. The average size over the first six months of 2020 was 2.3% of GDP, and purchase programs were larger in AEs (4.6% of GDP) than EMs (1.1% of GDP). The use of asset purchases during the pandemic by many countries that had interest rates above zero suggests that QE was no longer treated as a policy that could only be implemented after all of the policy space available to lower interest rates had been exhausted.Figure 2 Interest Rate Response to COVID-19. Notes: Blue is the change in the central bank policy interest rate from end-2019 through 2020q2. Red is the change in the shadow interest rate over the same period for countries at their lower bound (Australia, Canada, Euro area, Japan, New Zealand, Switzerland, UK, and US). Argentina is excluded as its 22 percentage point reduction in the policy interest rate is so large it distorts the axis for other countries. Source: Policy interest rates from Haver and shadow interest rates from Leo Krippner's website, based on calculations in Krippner (2015).
Figure 2
Figure 3 Changes in Macroprudential Policy, FX Intervention, Capital Controls and Unconventional Monetary Policy in Response to COVID-19. Notes: Share of advanced economies (AEs) and emerging markets (EMs) that used each policy during 2021q1-2020q2. “Macropru” and “CCyB” report any easing in macroprudential regulations and the CCyB, respectively. “CFM (inf)” and “CFM (outfl)” report any easing of inflow controls or tightening of outflow controls. “FX Int” reports any use of FX reserves for exchange rate intervention. “QE”, “Liqu Banks” and “Swap Lines” report any use of unconventional monetary policy in the form of asset purchases, liquidity provision to banks, or swap lines, respectively. See Appendix Table 2 for details on magnitudes. Source: Based on scrapped data from the IMF's Policy Tracker. See Appendix Table 1 for more information on variable definitions.
Figure 3
A third policy response is FX intervention aimed at moderating sharp currency movements. We use two measures of FX intervention: net changes in FX reserves over 2020q1 and q2 (from Adler et al., 2021) as a percent of 2019 GDP and a dummy variable equal to one if a country reports adjusting FX reserves in the IMF's Policy Tracker. It is important to note that these measures could capture different aspects of FX intervention. The first measure reports intervention through all sales and purchases of reserve assets, including not only spot operations, but also derivatives and other transactions aimed at affecting the exchange rate by altering the central bank's foreign currency position. This proxy was created by Adler et al. (2021) and is a substantial improvement over the traditional approach of just using changes in a country's flow of international reserves as reported in the Balance of Payments.20 The second measure is a dummy indicating whether the country reports conducting any FX intervention, which should include any purchases or sales of reserves as well as any activity through derivative markets. This measure does not capture the direction of the intervention (e.g., whether the country increased or decreased reserve holdings) and is self-reported, so may not agree with the Adler et al. (2021) measure. We continue to exclude individual members of the Euro area (which do not use FX reserves with the intent of affecting their currency) but include the Euro area as a single entity.
Figure 4 (with summary statistics in Table 1) shows the magnitudes of changes in FX reserves according to Adler et al. (2021), and Figure 3 (with summary statistics in Appendix Table 2) shows the share of the sample reporting any FX intervention according to the IMF Tracker. Although only three AEs report using any FX intervention during the first half of 2020 (Iceland, Israel, and Switzerland), the majority of EMs reported engaging in some type of FX intervention (58% of this sample). A comparison with the Adler et al. (2021) data on FX intervention, however, suggests that FX intervention was more widespread than that reported in the IMF Policy Tracker, and the directions and magnitudes of this intervention varied substantially across countries. More specifically, of the 56 countries in the Adler et al. (2021) dataset, 25 sold FX reserves and 24 bought reserves. Some countries that increased FX reserves were traditional safe-haven economies (such as Switzerland, which increased reserves by 10.8% of GDP), while others were EMs that are not obvious safe-haven economies (such as Colombia, Peru, Russia, Thailand, and Uruguay). These countries could have been increasing FX reserves to improve competitiveness in response to a depreciated dollar or to build reserve buffers.Figure 4 FX Intervention during COVID-19 as % of GDP. Notes: Amount of FX intervention over 2020q1 and 2020q2 as % of 2019 GDP. A positive number indicates reserve accumulation. FX intervention is a proxy that includes not only spot transactions, but also derivatives transactions and other central bank operations that alter the central bank's foreign currency position with the purpose of affecting the exchange rate. Source: Based on data from Adler et al. (2021).
Figure 4
A fourth policy response is to adjust macroprudential regulations. We focus on any loosening in macroprudential policy aimed at supporting lending and access to credit.21 More specifically, we measure changes in macroprudential policy using three variables: 1) a dummy if the country reports any loosening in macroprudential policy in the IMF Policy Tracker; 2) a dummy if the country reports adjusting its counter-cyclical capital buffer (CCyB) in the IMF Policy Tracker; and 3) the magnitude of changes in the CCyB.22 The CCyB is the one macroprudential regulation with a magnitude that is comparable across countries and reported on a timely basis, although it is only used in a subset of economies. Figure 3 shows the share of countries reporting any adjustment in macroprudential policy or the CCyB, with more detailed statistics in Appendix Table 2. A large proportion of countries report adjusting macroprudential policy (72% of AEs and 61% of EMs). Just the CCyB was loosened in 44% of AEs, but only 16% of EMs. Table 1 provides additional information on the magnitude of changes in the CCyB; the mean loosening was 27bp, although this includes many countries that did not have a CCyB in place to adjust. Of the 16 countries that adjusted their CCyBs, the size of adjustment ranged from 25bp in Germany to 250bp in Sweden.
A fifth policy is adjustments to capital controls. We focus on two types of capital flow measures (CFMs) aimed at reducing net capital outflows and the corresponding pressure for currency depreciation. More specifically, we use dummy variables to capture if countries reduced controls on capital inflows or increased controls on capital outflows.23 Both measures are based on country responses to the IMF Policy Tracker. Since these measures are self-reported, they may understate the use of capital controls, as some countries may not report adjustments to controls or use a different terminology in order to avoid any perceived negative stigma from the use of these measures. Figure 3 (with additional information in Appendix Table 2) shows that very few countries report making these adjustments to capital controls—with only 8% of EMs reducing controls on capital inflows (Peru, India, and China) and 5% tightening controls on capital outflows (Turkey and Argentina). The only AE that reported changing its capital flow measures is Korea, which adjusted controls on capital inflows.
The final policy response included in this paper is steps aimed at containing the spread of COVID-19 through targeted health measures, mobility restrictions, and testing, tracing, and vaccine policies. The focus of “containment” measures is different than the other policy tools that are used to respond to a variety of economic shocks or periods of financial stress. Nonetheless, it is useful to include these policies as they were important for stabilizing economies and their use may have interacted with other policy choices and the extent of policy space (a focus of this paper). For example, if countries had less space to support incomes and employment through fiscal or monetary stimulus, they may have felt more urgency to take steps to contain the spread of the disease, or they may have been more reticent to restrict economic activity as people would have less support on which to survive. To measure these containment and health policies, we use Oxford's Coronavirus Government Response Tracker (OxCGRT).24 We focus on the “Containment and Health measure”, which includes school closings, workplace closings, cancellation of public events, restrictions on gatherings, stay-at-home requirements, restrictions on international movement, public information campaigns, testing policy, contact tracing, facial coverings, and vaccination policies. Figure 5 graphs this index through the end of 2020Q2 for AEs and EMs. All countries adopted health and containment policies, with an average index of 56.6. This average, however, reflects a large variance in responses. The weakest response was in Estonia (24.3) and strongest was in Colombia (87.5). More generally, EMs introduced stricter measures than AEs in the first half of 2020.Figure 5 Health and Containment Measures in response to COVID-19. Notes: Indices measured through 2020q2. This Health and Containment index combines different lockdown restrictions, testing policies, contact tracing, and vaccination policies. A higher value indicates more stringent health and containment measures.54Source: Based on data from the Oxford's Coronavirus Government Response Tracker (OxCGRT) compiled by the University of Oxford and available at: Coronavirus Government Response Tracker | Blavatnik School of Government (ox.ac.uk)
Figure 5
Appendix Table 3 provides a summary of which countries used each of the tools discussed above.25 Green indicates that the tool was used to provide stimulus and red indicates that the tool was not used. Yellow denotes that the tool was used—but not in the direction typically associated with stimulus or easing financial conditions (e.g., raising interest rates or accumulating FX reserves) and white indicates that no data was available or the tool is not available for that country.26 This table and the series of figures and tables discussed above (Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Table 1, and Appendix Table 2) confirm that countries used an array of different policy tools to respond to COVID-19, with substantial variation in which policies each country selected. Moreover, even for countries that chose to use the same type of tool (such as fiscal policy), there was substantial variation in not only the extent to which they employed each tool, but how it was implemented. For example, for countries intervening in FX markets, some used reserves to slow currency depreciations (in green), while others added to their reserve stockpiles to moderate currency appreciations (in yellow). For countries using monetary stimulus, some only shifted to unconventional tools (such as asset purchase programs) after lowering interest rates, such that their policy interest rates were at their lower bounds, while others used unconventional tools actively even when their policy interest rates were well above zero. What explains this substantial cross-country variation in the choice of tools, intensity by which each tool was used, and form by which each tool was implemented during COVID-19?
IV Factors Determining Policy Choice during the COVID-19 Pandemic
This section analyzes what factors determined the use of each of the policy tools discussed in Section III: pre-existing policy space, the extent of stress in financial, economic and health measures, and other country characteristics. The onset of COVID-19 is a useful case study as the pandemic was an exogenous shock and did not reflect prior policy choices or economic imbalances, thereby providing cleaner identification of the factors driving policy responses. This is also a useful case study as COVID-19 was a global shock that affected all countries simultaneously (at least in terms of the realization of the shock, if not the actual spread of the disease), prompting reactions during the same time period and facilitating cross-country comparisons.
A Methodology and Variables
What determined these different policy responses to the COVID-19 pandemic? For policy responses for which there are quantitative indicators, what determined the size of the response? And for policies that can be delivered in different forms (such as the type of fiscal or monetary stimulus), what determined the specific tools utilized? To answer these questions, this section estimates the use of each policy tool (PT) for each country i as function of three sets of variables: initial policy space (PS), country-specific stress (ST), and other country characteristics (CC):(1) PTi,t=α+β·PSi,t−1+γ·STi,t+δ·CCi,t−1+εi,t.
The policy tool and stress variables are measured during the initial phase of the pandemic (with t defined as 2020q1-2020q2, unless noted otherwise above) and the policy space and other country characteristics variables measured before the pandemic (with t-1 defined as year-end 2019, or the latest date before that if end-2019 is not available). Equation (1) is estimated using OLS when policy tool is a continuous variable, or as a probit when PT is a dummy variable. All regressions include robust standard errors.
In each regression, policy space is measured using an indicator that corresponds to the policy tool on the left-hand side. More specifically, for regressions predicting fiscal stimulus, we follow Romer and Romer (2019) and measure policy space as general government gross debt to GDP.27 For regressions predicting the use of monetary stimulus, policy space is measured using the level of the policy interest rate, as well as the shadow rate – where available – as a robustness test. For regressions predicting the use of FX intervention, policy space is measured as the level of FX reserves as a percent of GDP.28 For regressions predicting the use of macroprudential tools, policy space is measured as an index of three popular macroprudential tools (the level of the CCyB, the level of the LTV ratio, and an index of FX regulations).29 For regressions predicting changes in the CCyB, policy space is measured by the initial level of the CCyB. For regressions predicting adjustments in controls on capital inflows or outflows, policy space is measured using an index of controls on inflows or outflows, respectively, from Fernandez et al. (2016).30 Finally, for regressions predicting the use of containment tools, there is a less obvious measure of policy space, so we use the output gap to capture the stage of the business cycle and thereby whether the economy was starting from a relatively stronger position to absorb any containment in activity. Each of these measures is written so that a positive value indicates more policy space (i.e., lower debt ratios, higher interest rates, higher FX reserves, tighter macroprudential regulation or capital controls, and a smaller output gap).
The second set of variables, measuring country-specific stress (ST), are the same for the regressions predicting the use of each policy tool. More specifically, we focus on capturing the “stress” from COVID-19 in terms of financial markets, economic activity, and health outcomes. We measure financial stress based on changes and percent changes from end-2019 to the date of “peak stress” for each country in the first half of 2020 for sovereign CDS spreads (5-year, US$) from Bloomberg, and if this is not available, from the EMBI+ bond index.31 We measure economic stress as the change in each country's forecast 2020 real GDP growth between January and June, according to the IMF's World Economic Outlook updates.32 We measure health stress as the number of reported cases of COVID-19 as a share of the population, as reported in Oxford's Coronavirus Government Response Tracker (OxCGRT). In each case, a higher value indicates more stress (i.e., greater increase in financial market spreads, greater reduction in forecast GDP growth, or greater incidence of COVID cases).
The final set of variables (CCi) controls for other country characteristics before the spread of COVID-19. Given the limited degrees of freedom in this cross-section analysis, we only include four controls in our baseline for each policy tool. We include a dummy variable equal to one for countries with a fixed exchange rate (based on the classification in Ilzetzki et al., 2019 33 ) and another dummy for emerging markets (based on IMF definitions). We also include a broad measure of institutional quality (the ICRG index) from the Worldwide Governance Indicators and a measure of trade openness (exports plus imports as a share of GDP, from the IMF). For sensitivity tests, we include a range of other control variables, such as the Chinn-Ito measure of financial openness (Chinn and Ito, 2006), changes in credit ratings (based on Fitch ratings), GDP per capita, and exposure to commodity prices.34 These changes in the controls for country characteristics have no impact on the key results discussed below. Appendix Table 1 provides additional information on each of the variables.
B Baseline Results
Table 2 reports results for estimates of equation (1), with separate panels for different groups of policy tools for which there are sufficient degrees of freedom to produce meaningful estimates.35 In each table, we begin with the simple correlation between the policy tool and corresponding policy space, and then add a control for financial stress, then the other two stress variables, and then the full set of control variables.36 On the far right for each tool, we also add an interaction between the policy space variable and the EM dummy in order to assess if policy space is more or less important in emerging markets. Also, since the countries/entities in each set of regressions can vary due to data availability and the number of entities that have the ability to use each tool, Appendix Table 4 shows the countries/entities included in the baseline regressions for each tool.37 Many of the regressions have a high adjusted-R2, reaching 41% for the size of fiscal stimulus, about 60% for adjustments in policy interest rates, and 87% for use of the CCyB. This suggests that these simple cross-country regressions can explain a meaningful share of the variation in policy responses to the COVID-19 pandemic.38 Table 2 Regression Results: Policy Responses as a Function of Policy Space, Stress and Other Country Characteristics
Table 2PANEL A: ANNOUNCED FISCAL STIMULUS
Fiscal Stimulus / GDP Above-the-Line Fiscal Stimulus/GDP Below-the-Line Fiscal Stimulus/GDP
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Policy Space Variables
Policy space -0.114*** -0.120*** -0.118*** -0.0868*** -0.101*** -0.0252* -0.0269* -0.0268** -0.0193 -0.0193 -0.0884*** -0.0921*** -0.0889*** -0.0689*** -0.0808***
(0.0200) (0.0188) (0.0194) (0.0246) (0.0196) (0.0136) (0.0135) (0.0131) (0.0149) (0.0161) (0.0227) (0.0234) (0.0207) (0.0229) (0.0196)
Policy space 0.130*** -0.000101 0.107***
* EM dummy (0.0429) (0.0236) (0.0383)
Stress Variables
Financial -0.863*** -0.854*** -0.169 0.184 -0.271** -0.240** 0.0512 0.0510 -0.589*** -0.611*** -0.322 -0.0319
(0.202) (0.221) (0.297) (0.277) (0.113) (0.108) (0.0935) (0.113) (0.180) (0.184) (0.276) (0.256)
Economic 0.217 0.392 0.566 -0.160 -0.0487 -0.0488 0.480 0.254 0.398
(0.349) (0.419) (0.378) (0.207) (0.221) (0.225) (0.321) (0.330) (0.306)
Health 0.0327 -0.0512 -0.203 0.254 0.116 0.116 -0.228 -0.108 -0.232
(0.239) (0.294) (0.273) (0.204) (0.211) (0.212) (0.229) (0.290) (0.278)
Other Country Characteristics
Fixed ER -0.209 -0.745 -2.256 -2.256 2.646 2.205
Dummy (2.598) (2.608) (1.522) (1.548) (2.113) (2.090)
Institutional 0.300 0.365 0.280** 0.280** -0.118 -0.0642
Quality (0.344) (0.313) (0.105) (0.105) (0.295) (0.273)
Trade -1.460 -1.369 0.713 0.713 -2.323 -2.248
Openness (1.577) (1.477) (1.322) (1.343) (1.853) (1.749)
EM dummy -4.179 2.314 0.547 0.543 -5.840* -0.500
(3.511) (3.602) (1.630) (2.288) (3.384) (3.339)
Observations 40 40 40 39 39 41 41 41 40 40 40 40 40 39 39
Adj. R-squared 0.296 0.351 0.319 0.388 0.414 0.051 0.061 0.052 0.246 0.220 0.254 0.284 0.278 0.308 0.331
PANEL B: MONETARY STIMULUS (POLICY RATES AND ASSET PURCHASES)
Change in Policy Interest Rates Asset Purchases (dummy) Asset Purchases (% of GDP)
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)
Policy Space Variables
Policy space 0.307*** 0.344*** 0.366*** 0.348*** 0.604*** -0.140* -0.135 -0.129 -0.161* 0.177 -0.410*** -0.370*** -0.312*** -0.252 0.279
(0.0795) (0.0875) (0.0862) (0.102) (0.103) (0.0754) (0.0838) (0.0857) (0.0937) (0.280) (0.131) (0.119) (0.109) (0.179) (0.902)
Policy space -0.276 -0.371 -0.573
* EM dummy (0.164) (0.311) (0.916)
Stress Variables
Financial -0.244 -0.282* -0.0756 -0.0409 -0.0594 -0.0744 -0.248 -0.207 -0.403 -0.513 -0.754 -0.682
(0.154) (0.155) (0.191) (0.200) (0.174) (0.176) (0.269) (0.271) (0.334) (0.371) (0.526) (0.526)
Economic 0.120* 0.115 0.107 0.0448 0.0379 0.0263 0.335 0.386 0.370
(0.0703) (0.0689) (0.0730) (0.0826) (0.0867) (0.0910) (0.257) (0.244) (0.251)
Health 0.00703 0.0181 0.0180 0.00499 -0.0336 -0.0277 -0.00504 -0.146 -0.146
(0.0402) (0.0522) (0.0512) (0.0633) (0.0676) (0.0673) (0.178) (0.229) (0.234)
Other Country Characteristics
Fixed ER 0.478 0.451 -0.596 -0.636 -2.654* -2.709**
Dummy (0.393) (0.391) (0.409) (0.421) (1.321) (1.331)
Institutional -0.0489 -0.0546 -0.00595 -0.0140 -0.0690 -0.0807
Quality (0.0523) (0.0533) (0.0601) (0.0603) (0.165) (0.169)
Trade -0.0841 -0.0828 -0.996** -0.974** -1.332** -1.329**
openness (0.205) (0.204) (0.484) (0.457) (0.602) (0.619)
EM dummy -0.959 -0.717 -0.171 0.160 -3.035 -2.534
(0.644) (0.640) (0.696) (0.778) (2.061) (2.209)
Observations 52 49 49 48 48 51 49 49 48 48 52 49 49 48 48
Adj. R-squared 0.520 0.542 0.565 0.591 0.595 0.0844 0.0954 0.101 0.215 0.231 0.078 0.079 0.071 0.224 0.210
PANEL C: LIQUIDITY SUPPORT, SWAPS AND FX INTERVENTION
Liquidity to Banks Swaps FX Intervention Dummy FX Intervention / GDP
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
Policy Space Variables
Policy space -0.0475 -4.402** 0.0670 -0.606 -0.000338 0.000723 0.000902 0.0408*** 0.0438*** -0.0708** -0.0702** -0.0708** -0.0535* -0.0542*
(0.0678) (2.157) (0.0866) (0.376) (0.00757) (0.00778) (0.00773) (0.0123) (0.0116) (0.0322) (0.0334) (0.0319) (0.0273) (0.0282)
Policy space 4.489** 0.734* -0.0242 0.0105
* EM dummy (2.173) (0.389) (0.0336) (0.0460)
Stress Variables
Financial 0.0332 -0.190 -0.127 -0.224 0.00905 0.00493 -0.0824** -0.0967** 0.0478 0.0396 -0.0458 -0.0399
(0.360) (0.351) (0.313) (0.322) (0.0425) (0.0433) (0.0351) (0.0386) (0.0518) (0.0484) (0.0442) (0.0513)
Economic 0.0197 0.0271 -0.108 -0.0956 0.0664 0.0769 0.0855 0.103 0.135 0.128
(0.0875) (0.106) (0.0890) (0.0893) (0.0769) (0.0870) (0.0904) (0.116) (0.115) (0.122)
Health -0.0492 -0.0374 0.0601 0.0850 -0.00378 0.0811 0.0711 -0.208 -0.164 -0.159
(0.0706) (0.0624) (0.0959) (0.0985) (0.0610) (0.0858) (0.0880) (0.147) (0.134) (0.136)
Other Country Characteristics
Fixed ER -0.608 -0.160 -0.166 0.0542 0.121 0.152 -0.721 -0.727
Dummy (0.437) (0.464) (0.481) (0.498) (0.513) (0.518) (0.687) (0.694)
Institutional -0.0137 0.0685 0.0607 0.0889 -0.139** -0.133** -0.184* -0.189*
Quality (0.0481) (0.0912) (0.0635) (0.0657) (0.0606) (0.0628) (0.104) (0.101)
Trade -0.206 -1.590* 0.124 0.0880 -1.459*** -1.472*** 0.260 0.267
openness (0.364) (0.847) (0.334) (0.334) (0.493) (0.492) (1.481) (1.487)
EM dummy -0.0207 -8.501** -0.507 -1.210 0.177 0.797 -0.173 -0.457
(0.932) (4.136) (0.766) (0.869) (0.768) (1.266) (1.171) (1.518)
Observations 48 48 45 45 55 53 53 52 52 55 53 53 52 52
Adj. R-squared 0.0639 0.337 0.194 0.240 2.63e-05 0.000731 0.0120 0.336 0.344 0.274 0.290 0.303 0.390 0.375
PANEL D: MACROPRUDENTIAL POLICY AND CAPITAL CONTROLS
Loosen Macroprudential Regulation (dummy) Loosen CCyB (pp change) Loosen Controls on Capital Inflows
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (1) (2) (3) (4) (5)
Policy Space Variables
Policy space 6.899*** 5.909*** 5.677*** 6.256*** 10.08* 0.677*** 0.675*** 0.674*** 0.654*** 0.775*** 1.047 1.049 1.220 0.813 0.369
(1.989) (1.941) (1.921) (2.050) (5.780) (0.0959) (0.0980) (0.0939) (0.0984) (0.0887) (0.953) (1.021) (0.966) (0.881) (0.973)
Policy space -5.591 -0.390*** 0.584
* EM dummy (6.353) (0.145) (1.611)
Stress Variables
Financial -0.0907 -0.0953 -0.112 -0.135 -0.00198 -0.00213 0.00204 -0.00157 -0.0702 -0.0787 -0.589* -0.620
(0.0553) (0.0605) (0.0761) (0.0847) (0.00176) (0.00170) (0.00361) (0.00241) (0.133) (0.108) (0.349) (0.379)
Economic -0.0184 -0.0632 -0.0781 -0.00327 -0.0118 -0.00901 -0.000818 0.0564 0.0594
(0.0676) (0.0705) (0.0764) (0.0136) (0.0152) (0.0123) (0.127) (0.138) (0.140)
Health -0.0292 -0.00541 -0.00627 0.0262* 0.0255* 0.0194 0.0438 -0.0630 -0.0570
(0.0615) (0.0619) (0.0631) (0.0156) (0.0141) (0.0122) (0.0611) (0.0611) (0.0586)
Other Country Characteristics
Fixed ER 0.166 0.223 0.0717 0.0371 0.138 0.148
Dummy (0.410) (0.428) (0.0785) (0.0611) (0.457) (0.466)
Institutional -0.0713* -0.0666 0.00580 0.00504 0.165** 0.163**
Quality (0.0432) (0.0448) (0.00702) (0.00748) (0.0703) (0.0691)
Trade 0.0565 -0.0898 -0.0998 -0.112 -3.766** -3.721***
openness (0.435) (0.525) (0.0611) (0.0699) (1.463) (1.389)
EM dummy -1.057** -0.114 -0.121 -0.00177 1.932** 1.773**
(0.536) (1.030) (0.0873) (0.0592) (0.895) (0.841)
Observations 73 69 69 68 68 70 65 65 64 64 62 61 61 60 60
Adj. R-squared 0.213 0.226 0.230 0.279 0.292 0.798 0.797 0.804 0.812 0.867 0.0579 0.0622 0.0694 0.349 0.351
PANEL E: CONTAINMENT POLICIES
Containment and Health Index
(1) (2) (3) (4) (5)
Policy Space Variables
Policy space -3.279** -2.503* -2.372* -1.096 -0.581
(1.361) (1.387) (1.339) (1.065) (1.385)
Policy space -0.991
* EM dummy (2.068)
Stress Variables
Financial 0.962** 0.984* 0.280 0.262
(0.471) (0.505) (0.243) (0.238)
Economic -0.415 0.157 0.110
(0.696) (0.580) (0.574)
Health 1.403** 1.528** 1.467**
(0.534) (0.687) (0.699)
Other Country Characteristics
Fixed ER -6.441* -6.541*
Dummy (3.718) (3.730)
Institutional -0.359 -0.352
Quality (0.318) (0.314)
Trade -4.224 -4.008
openness (2.745) (2.849)
EM dummy 10.31** 10.72***
(4.021) (3.901)
Observations 68 64 64 63 63
Adj. R-squared 0.084 0.128 0.155 0.423 0.414
Notes: Regressions predicting the policy response listed at the top as a function of policy space, stress, and other country characteristics. Regressions are estimated using OLS for quantitative measures of policy responses and a probit for dummy variable measures of policy responses. See text, Table 1 and Appendix Table 1 for details on variable definitions. Regressions for each policy response only include countries that have the ability to adopt each policy tool, e.g., individual countries in the Euro area cannot use monetary policy or FX intervention, but can adopt other policies. The Euro area is included as a "country" that can pursue monetary and FX policy, but not other policies. All regressions include a constant (not reported) and robust standard errors. *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.
The coefficient estimates show that policy space is the most consistently significant determinant of the use of most policy tools (albeit not always in the expected direction). As expected, countries with higher interest rates before the pandemic reduced interest rates by significantly more (panel B). Also not surprising, countries with a tighter macroprudential stance before the pandemic, or with a higher CCyB ratio, were more likely to loosen macroprudential regulation and lower the CCyB, respectively (panel D). Countries with more FX reserves (relative to GDP) before 2020 were significantly more likely to report using some form of FX intervention (panel C), although the estimates that incorporate the direction and magnitude of intervention (instead of just self-reporting some form of intervention) suggest that countries with greater reserve holdings tended to increase (instead of decrease) reserves in response to the pandemic. This could reflect a number of influences: (1) some countries experienced sharp appreciation pressures after March and intervened by accumulating reserves to slow this appreciation; (2) countries with larger “war chests” of reserves may not have needed to use as many reserves to support their exchange rates as they had greater credibility or stronger fundamentals; and (3) countries which were more conservative in accumulating reserves may also have been more conservative in using them during the pandemic. Although a less precise measure of policy space, the stage of the economic cycle before the pandemic did not appear to constrain the ability of governments to enact containment measures to restrain the spread of the virus, and countries with weaker economies before the pandemic may have enacted stronger containment policies (panel E).
The most surprising result for the estimated coefficients on policy space in Table 2 are the negative and significant coefficients on fiscal space (panel A). Taken at face value, this would imply that countries with more fiscal space before the pandemic (i.e., lower debt ratios) announced significantly less fiscal stimulus.39 This is the opposite result than in previous research analyzing pre-COVID-19 samples (Romer and Romer, 2018, 2019; and Jordà et al., 2016), and agrees with the results in Benmelech et al. (2020). The estimates that also include fiscal space interacted with the EM dummy (column 5), however, suggest that this lack of constraint of fiscal space on the fiscal response to COVID-19 may only apply to AEs. When the coefficient on the interaction with the EM dummy is combined with the coefficient for policy space, the combined relationship for fiscal stimulus is positive. This indicates that EMs with smaller debt ratios before COVID-19 had a larger announced fiscal response during the initial phase of the pandemic.
The estimated coefficients on policy space also provide information on whether the space available for a given broad category of policies (e.g., fiscal or monetary) affected how the policy was implemented. For fiscal policy, the estimates (on the right of Table 2, panel A) show that the negative relationship between fiscal space and the fiscal response to COVID-19 primarily occurs through below-the-line stimulus for AEs. In other words, AEs with higher debt levels (less space) before the pandemic announced significantly more stimulus in the form of loans, equity and credit guarantees, but did not announce significantly more stimulus through traditional on-budget spending increases and revenue losses. This suggests that even if countries with high debt levels were not constrained in the size of the stimulus they could announce in response to COVID-19, they may have been constrained in the amount of stimulus they could offer on-budget, and therefore were more likely to construct the stimulus in ways that would not as directly contribute to reported debt burdens. This shifting of stimulus spending to “below-the-line”, however, only applied to AEs. Column 15 suggests that EMs with higher debt levels announced significantly less—instead of more—below-the-line stimulus. In other words, EMs with higher debt levels appear to have been more constrained than AEs in announcing stimulus through below-the line measures, but neither set of countries was significantly constrained in their announcement of above-the-line measures.40
For monetary policy, there is also some evidence that the space available contributed to how the policy was implemented. Countries with higher policy interest rates before the pandemic were not only more likely to lower policy interest rates, but less likely to enact certain forms of “unconventional” monetary easing (Table 2, panel B). More specifically, countries with more space to lower interest rates were less likely to enact any form of asset purchases (columns 6-10) and announced smaller asset purchase programs (columns 11-15). Panel C shows that countries with more space to lower interest rates were also less likely to enact a program providing liquidity to banks. The significance of many of these estimated relationships between monetary policy space and the use of unconventional monetary policy tools varies across specifications,41 but when the baseline is adjusted to take into account the simultaneous use of different policy tools and the policy space available for the full set of tools, the relationships are more consistently significant (shown in Section V). Combining these estimates, the results are consistent with models of optimal policy design at the zero lower bound (e.g., Orphanides and Wieland, 2000), which show that countries with less space to provide monetary stimulus through reductions in policy interest rates are more likely to use unconventional monetary tools, and if they do use tools such as asset purchase programs, should do so in smaller magnitudes.
Next, shifting from the coefficients on policy space to those on the stress variables, greater financial stress (but not economic or health stress) is correlated with less fiscal stimulus (in aggregate as well as for above- and below-the-line measures) in specifications that do not control for other country characteristics, but becomes insignificant when adding these additional controls. There is some evidence that greater financial stress is correlated with a lower probability of using FX intervention in either direction, but there is no evidence it is significantly correlated with the size of FX intervention, which may reflect a hesitancy to self-report intervention for countries experiencing substantial financial stress. The extent of economic stress and health stress are not significantly correlated with any of the policy responses, except for a strong correlation between health stress and the use of health and containment policies. This is not surprising—countries with more reported COVID-19 cases tended to impose stricter mobility restrictions and have more extensive testing and tracking regimes.
Finally, shifting to the country characteristics other than policy space in Table 2, some variables are consistently significant. Countries with stronger institutional quality did more of their fiscal stimulus above-the-line, were significantly less likely to use FX intervention, and more likely to loosen controls on capital inflows. Countries with more trade openness were less likely to use asset purchase programs or loosen controls on capital inflows. Emerging markets were significantly more likely to loosen controls on capital inflows, as well as to enact health and containment measures, but were otherwise not significantly different than advanced economies in the use of other policies (at the 5% significance level)—except for the relationship with policy space (as discussed above). None of these country characteristics were as consistently important across the different policy tools, however, as found for policy space. We also performed a series of sensitivity tests with additional control variables (the Chinn-Ito measure of financial openness, changes in credit ratings (based on Fitch ratings), GDP per capita, and exposure to commodity prices), with no meaningful impact on the key results.
C A Closer Look: Did Fiscal Space Matter during COVID-19?
In order to further explore these results, and especially the finding that the aggregate fiscal stimulus that was announced in response to COVID-19 was not significantly constrained by fiscal space in advanced economies, we estimate an extensive series of sensitivity tests. We focus on four sets of tests of the relationship between fiscal space and the fiscal response to COVID-19: (1) sensitivity to the measure of fiscal space; (2) sensitivity to outliers and sample composition; (3) sensitivity to the measure of fiscal stimulus (focusing on realized spending over a longer period instead of the initial, announced stimulus); and (4) sensitivity to the choice of other controls. A subset of these results is reported in Tables 3 and 4 .Table 3 Sensitivity Tests: Announced Fiscal Stimulus
Table 3 Different Measures of Fiscal Space Impact of Outliers and Sample Exclude Japan
Gross Debt/GDP Cyclically-adj primary bal. Gross Debt/Tax Base Fiscal Balance/Tax Base Alternate Stimulus Alt Stimulus + Base Sample Exclude Japan Exclude Debt/GDP>100% Gross Debt/GDP Below- the-line
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Policy Space Variables
Policy space -0.0997*** 0.521 -0.0159*** 0.113 -0.0597 -0.105*** -0.0868 -0.00648 -0.0841 -0.0671
(0.0195) (0.677) (0.00236) (0.220) (0.0383) (0.0231) (0.0635) (0.0687) (0.0626) (0.0673)
Policy space 0.127*** -0.230 0.0283*** -0.0960 0.0787 0.128** 0.114 0.0392 0.110 0.0917
* EM dummy (0.0436) (0.675) (0.00801) (0.198) (0.0512) (0.0527) (0.0783) (0.0739) (0.0767) (0.0781)
Stress Variables
Financial 0.180 0.195 0.210 0.147 0.0578 0.153 0.150 0.111 0.144 -0.0638
(0.276) (0.379) (0.265) (0.384) (0.264) (0.319) (0.293) (0.274) (0.289) (0.254)
Economic 0.563 0.433 0.762** 0.472 0.0900 0.737* 0.582 0.603 0.582 0.413
(0.382) (0.435) (0.340) (0.413) (0.329) (0.423) (0.357) (0.364) (0.360) (0.288)
Health -0.201 -0.0423 -0.285 -0.0498 0.201 -0.126 -0.154 -0.0562 -0.148 -0.187
(0.274) (0.299) (0.278) (0.308) (0.288) (0.313) (0.329) (0.277) (0.329) (0.294)
Other Country Characteristics
Fixed ER -0.749 -0.930 -0.515 -1.005 -2.398 -0.922 -0.633 -1.276 -0.631 2.310
dummy (2.643) (2.612) (2.533) (2.479) (2.250) (2.934) (2.617) (2.454) (2.650) (2.065)
Institutional 0.361 0.306 0.285 0.241 0.377* 0.393 0.314 0.230 0.305 -0.113
quality (0.314) (0.399) (0.291) (0.416) (0.197) (0.337) (0.375) (0.344) (0.374) (0.312)
Trade -1.396 -0.118 -2.389 0.177 -1.604 -1.663 -1.107 -1.230 -1.104 -2.001
openness (1.484) (2.476) (1.517) (3.149) (1.695) (1.644) (2.090) (2.785) (2.114) (2.294)
EM dummy 2.112 -7.206 1.003 -7.646 -0.0900 1.372 0.948 -4.007 0.655 -1.788
(3.624) (4.972) (3.460) (4.669) (3.558) (3.933) (6.060) (5.950) (5.956) (5.630)
Observations 39 38 39 39 65 39 38 35 38 38
Adj. R-squared 0.412 0.186 0.419 0.200
0.287 0.409 0.270 0.152
0.269 0.209
Notes: Regressions of the total, announced fiscal stimulus in response to COVID-19 over the first six months of 2020, using the baseline specification from Table 2, panel A, column 5, except as noted. Columns (1)-(4) use alternate measures of fiscal space from Kose et al. (2017): gross government debt to GDP; the cyclically-adjusted primary fiscal balance; gross government debt as a percent of average tax revenues; and the fiscal balance as a percent of tax revenues. Columns (5) and (6) measure fiscal stimulus using data from the IMF Policy Tracker, which is self-reported stimulus in 2020q1-q2 as a share of 2020 GDP, with the full set of countries for which this variable is available in column (5) and then limited to the smaller sample in the baseline analysis in column (6). Columns (7), (9) and (10) exclude Japan, and column (8) excludes all countries with a net debt/GDP ratio >100%. Column (10) uses the below-the-line portion of the fiscal balance. All regressions include a constant (not reported) and are estimated with robust standard errors. *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Table 4 Sensitivity Tests: Fiscal Stimulus Through 2021
Table 4 Baseline: Initial Announced Stimulus Extend Fiscal Response (2020q1-2021q3)
(2020 q1-q2) Total Stimulus
Total Stimulus Above-the-Line (ATL) Below-the-Line (BTL) Total Stimulus Above-the-Line (ATL) Below-the-Line (BTL) Exclude Japan Larger Sample Larger Sample ex. Japan
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Policy Space Variables
Policy space -0.101*** -0.0193 -0.0808*** -0.130*** -0.0378* -0.0916*** -0.127* -0.104*** -0.0750*
(0.0196) (0.0161) (0.0196) (0.0197) (0.0195) (0.0278) (0.0665) (0.0291) (0.0387)
Policy space 0.130*** -0.000101 0.107*** 0.206*** 0.0781** 0.117** 0.202** 0.140*** 0.110**
* EM dummy (0.0429) (0.0236) (0.0383) (0.0539) (0.0334) (0.0464) (0.0904) (0.0467) (0.0519)
Stress Variables
Financial 0.184 0.0510 -0.0319 0.0623 0.0443 -0.0609 0.0547 0.0618 0.0146
(0.277) (0.113) (0.256) (0.272) (0.143) (0.272) (0.298) (0.291) (0.294)
Economic 0.566 -0.0488 0.398 0.718* 0.0642 0.550 0.721* 0.361 0.467
(0.378) (0.225) (0.306) (0.366) (0.250) (0.347) (0.359) (0.374) (0.348)
Health -0.203 0.116 -0.232 -0.132 0.211 -0.302 -0.121 -0.0644 0.0608
(0.273) (0.212) (0.278) (0.316) (0.278) (0.349) (0.370) (0.257) (0.243)
Other Country Characteristics
Fixed ER -0.745 -2.256 2.205 -1.356 -3.517** 2.492 -1.331 -2.855 -2.411
Dummy (2.608) (1.548) (2.090) (2.494) (1.685) (2.221) (2.475) (2.122) (2.046)
Institutional 0.365 0.280** -0.0642 0.0803 0.125 -0.116 0.0689 0.360* 0.300
Quality (0.313) (0.105) (0.273) (0.246) (0.129) (0.294) (0.318) (0.206) (0.201)
Trade -1.369 0.713 -2.248 -2.371 -0.277 -2.174 -2.313 -0.562 -0.536
openness (1.477) (1.343) (1.749) (1.814) (1.685) (2.115) (2.395) (1.749) (1.646)
EM dummy 2.314 0.543 -0.500 -0.315 -0.272 -1.126 -0.620 1.267 -0.804
(3.602) (2.288) (3.339) (3.995) (2.904) (3.999) (6.715) (3.133) (3.263)
Observations 39 40 39 39 40 39 38 63 62
Adj. R-squared 0.414 0.220 0.331 0.585 0.367 0.375 0.481 0.478 0.386
Notes: Columns (1) - (3) repeat the baseline results from Table 2, panel A, for the announced fiscal stimulus in response to the initial phase of Covid in 2021q1-q2. Columns (4) - (9) repeat these baseline results, except replace announced fiscal stimulus as of 2020q2 with the realized fiscal stimulus as of October 2021 (i.e., 2020q1-2021q3). All columns focus on total fiscal stimulus, except columns (2), (3), (5), and (6), which break this into above-the-line or below-the-line spending. Columns (1) through (7) use the sample available for the baseline regressions that focus on the immediate response to Covid-19, while columns (8) and (9) use the larger set of countries for which the Oct. 2021 data is available. Columns (7) and (9) exclude Japan. All regressions include a constant (not reported) and are estimated with robust standard errors. *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.
First, we test if the results are sensitive to how fiscal space is measured. Although the debt to GDP ratio is the most common measure used in the literature, several papers argue that other statistics better capture the concept of fiscal space. For example, measures of the fiscal balance (especially if adjusted for the stage of the business cycle) could more accurately capture any continuing imbalances than debt measures (which reflect past imbalances). Also, scaling any measure of fiscal space by tax revenues could better capture a country's ability to repay. Ghosh et al. (2013) and Kose et al. (2022) provide excellent summaries of this discussion. To explore if the measurement of fiscal space affects our key results, we use several alternatives that are available for most of our sample: gross (instead of net) debt as a percent of GDP; the fiscal balance as a percent of GDP; the primary fiscal balance as a percent of GDP; the cyclically-adjusted, primary fiscal balance as a percent of GDP; gross government debt as a percent of average tax revenues; and the fiscal balance as a percent of average tax revenues.42 A selection of these results is reported in columns 1-4 of Table 3, with each column including an interaction between the EM dummy and fiscal space (as well as an EM dummy) to allow for different relationships between the new measures of fiscal space and the fiscal response for these two groups of countries.
The signs and significance of the coefficients on fiscal space in Table 3 vary meaningfully across measures. The coefficients are usually negative and significant when some form of debt is incorporated in the numerator of the fiscal space variable (as found in our baseline), but become insignificant (and often positive), when some form of the fiscal balance is used instead.
Second, we examine the impact of outliers and sample composition. We begin by replacing our current measure of the fiscal response to COVID-19 (from the IMF's Fiscal Monitor) with another measure reported in the IMF's Fiscal Tracker (referred to as “Alternative Stimulus” in the table).43 This has the advantage of expanding the sample size (from 39 to 65), but has the disadvantage that the fiscal response is self-reported and expressed relative to 2020 GDP (which could introduce endogeneity as 2020 GDP was affected by the size of the fiscal response in 2020). Column 5 of Table 3 shows that the resulting coefficient on fiscal space is about half as large and becomes insignificant when using this new measure for the larger sample of countries. To test if this reflects the change in the measure of fiscal space or the sample, column 6 uses the new measure but restricts the sample to the smaller group of countries used in the baseline analysis in Table 2 (panel A). The coefficient on fiscal space becomes significant and very close to the corresponding estimate in the baseline (-0.105 as compared to -0.101 in the baseline). This suggests that sample composition is the key driver of the change in significance, and in a larger sample of countries, the relationship between fiscal space and the announced fiscal response to COVID-19 may be insignificant.
As an additional test for the impact of sample selection, we exclude outliers in our baseline sample (instead of trying to expand the sample with imperfect data). More specifically, we exclude outliers with very large debt to GDP ratios: just Japan, or the four countries with debt/GDP>100%. Results are reported in columns 7 and 8 in Table 3. Once again, the significant negative relationship between fiscal space and the announced fiscal response to COVID-19 disappears. In fact, the negative and significant coefficient on fiscal space appears to be driven by one outlier: Japan. Japan had a very high debt ratio before the pandemic (237% of GDP, compared with the sample average of 61%) and responded to the pandemic with an announced fiscal stimulus (relative to GDP) about three times greater than the sample average. To further highlight the role of this outlier, we replicate other results reported earlier that found a significant negative relationship between fiscal space and the fiscal response to the pandemic, but now exclude Japan. In each case the coefficient on fiscal space becomes insignificant, although it usually remains negative. Column (9) and (10) report a sample of these results to highlight the role of this one outlier.
Third, we explore the implications of focusing on the realized fiscal stimulus adopted over a longer period instead of the stimulus initially announced at the start of the pandemic. As discussed in Section III, this paper focuses on the fiscal response announced through 2020q2—thereby focusing on the initial, announced response to COVID-19. Focusing on this short window has several advantages: it captures the immediate ability of countries to respond to a large, negative shock; it clearly identifies spending in response to COVID-19 (as opposed to other spending priorities); and it avoids having to make a judgement about when COVID-related spending stops as the pandemic evolved differently in different countries. One potential disadvantage of this measure, however, is that some of the announced fiscal responses were never realized—such as large credit guarantees or liquidity programs that were not drawn down.44 While these unrealized commitments may have been an important part of a country's response and helped stabilize economies, it is worth exploring whether an alternative measure of fiscal stimulus that focuses on realized spending in response to the pandemic (instead of initial announcements) had a similar, weak relationship with fiscal space.
In order to measure the realized fiscal stimulus in response to COVID-19, we use data from the IMF Fiscal Tracker from October 2021. This database reports the fiscal measures governments announced or adopted in selected economies in response to the COVID-19 pandemic from January 2020 through September 27, 2021. It includes COVID-19 related measures “for implementation in 2020, 2021, and beyond”, and we will refer to this data below as the “extended fiscal response”. This measure could still include fiscal allotments for credit guarantees and loan programs that were not entirely drawn down, but should better capture the actual spending in place about 1 ½ years after the pandemic began and before vaccines were available. This measure of the extended fiscal response as of 2022q3 is 94% correlated with our baseline measure of the announced fiscal response in 2020q1-q2. Appendix Figure 1 graphs the two measures by country. For most countries, the extended response is larger than the initial, announced response. This is not surprising; as COVID-19 continued to effect economies in late 2020 and throughout 2021 (e.g. through the Delta variant), countries announced additional stimulus. In most cases, this additional stimulus was larger than any unspent initial commitments that were wound down by that time.Appendix Figure 1 Initially Announced vs. Realized Fiscal Responses to COVID-19 as % of GDP. Notes: The graph shows fiscal intervention in response to COVID-19 as % of 2019 GDP using two different measures. “Announced until June 2020” is the fiscal response initially announced in 2020q1-q2 as shown in Figure 1. “Realized until October 2021” is the fiscal response governments announced or adopted in response to the COVID-19 pandemic from January 2020 through September 27, 2021. This includes COVID-19 related measures “for implementation in 2020, 2021, and beyond.” Both fiscal measures only include discretionary measures and not any support through automatic stabilizers or revenue losses corresponding to slower growth. Sources: “Announced” fiscal response is from the IMF's Fiscal Monitor Database of Country Fiscal Measures in Response to the COVID-19 Pandemic Database. “Realized” fiscal response is from the IMF's Fiscal Policy Tracker. The different vintages of the data can be found here: Fiscal Monitor Database of Country Fiscal Measures in Response to the COVID-19 Pandemic.
Appendix Figure 1
Next, we repeat our baseline regressions using this measure of the extended fiscal response instead of the initial, announced stimulus. Results are shown in Table 4. The first three columns repeat the baseline results for comparison (for total fiscal stimulus, ATL and BTL stimulus as a percent of GDP) and columns (4) through (6) repeat results using the same sample of countries with the extended response measure of stimulus. The key results reported above are unchanged, although the negative coefficients on fiscal space are somewhat larger for the new measures—suggesting there was even less constraint of fiscal space on realized fiscal stimulus over this longer period. Given the sensitivity of results to sample selection (as highlighted above), we repeat the regressions for the extended fiscal response, except exclude Japan, include a larger sample available in October 2021, and then with the larger sample excluding Japan (columns (7)-(9), respectively). The coefficient on fiscal space continues to be negative, suggesting that fiscal space did not constrain realized spending in response to COVID-19 in advanced economies over a longer period. In fact, even after excluding Japan, the coefficient is marginally significant at the 10% level. Fiscal space, however, continued to act as some constraint for emerging markets (as found in the baseline regressions)—and the more consistently significant positive coefficients on the variable for the interaction of the EM dummy (with policy space) may suggest that EMs were more constrained in their fiscal response over the extended period than in their initial fiscal announcements in the middle of 2020.
As a final set of sensitivity tests, we estimate each of the baseline regressions with additional controls (while maintaining the controls for policy space, stress, and other country characteristics). More specifically, we estimate sensitivity tests with control variables that are widely available and therefore do not change the sample size: (1) a dummy if the country had an IMF program at any point in 2020 (which includes five countries with existing programs, plus seven countries with programs started during the pandemic); (2) a variable measuring country sensitivity to commodity prices;45 (3) country credit ratings from Fitch46 ; and (4) nominal per capita GDP (in US$).47 For some of these tests, we are not able to replicate all of the baseline results as there are insufficient degrees of freedom (especially for the regressions in which the policy tool is a dummy).48 For all the regressions with sufficient degrees of freedom, however, these additional variables do not change the main results and are usually insignificant. The only additional control that is significant in more than one of these tests is the dummy indicating if the country had an IMF program. This coefficient is positively correlated with a country reducing the policy interest rate, reducing the CCyB, putting controls on capital inflows, and enacting stricter health containment measures.
To conclude, this series of robustness tests suggests that the significant negative relationship between fiscal space and the announced fiscal response to COVID-19 found in the baseline for AEs is not consistently significant and is particularly sensitive to outliers and the sample. When Japan is dropped from the sample, the relationship between fiscal space and the fiscal response to COVID-19 is no longer significant and negative (although it is still often negative for advanced economies). While these estimates suggest that having more fiscal space did not correspond to a significantly greater announced fiscal response to COVID-19, there is also little evidence that having less fiscal space acted as a significant constraint in how countries responded to the pandemic—especially in AEs. This lack of constraint of fiscal space on the fiscal response to the pandemic in AEs continues to be true when focusing on the realized fiscal response over the 1 ½ years through October 2021 (instead of just the initial, announced response in the first half of 2020), although EMs may have been somewhat more constrained over this longer period of time. All in all, however, this minimal constraint from fiscal space is a sharply different result than in earlier research, which generally finds that fiscal space was a significant constraint on the fiscal response to negative shocks before the COVID-19 pandemic (e.g., Romer and Romer, 2018, 2019; and Jordà et al. 2016) and during the 2008 Global Financial Crisis (Aizenman and Jinjarak, 2011).
Has something changed in the relationship between fiscal space and a country's ability to enact fiscal stimulus? This is an important topic for future work, but it is worth briefly considering several possible hypotheses. First, the exogenous nature of the COVID-19 pandemic may have reduced concerns about a country's ability (or willingness) to repay additional debt, as it did not reflect prior policy mistakes or domestic imbalances. Second, market participants may have expected that most of the negative impact of the pandemic would be relatively short-lived—which is the standard situation when a large and temporary stimulus to smooth incomes is the optimal policy response to act as a bridge and reduce scarring (IMF, 2020a). Third, the low interest rate environment in 2020 (and expectations for policy interest rates to remain low for an extended period, especially in AEs), would have increased countries’ debt capacity through the decrease in expected debt service costs. Fourth, the nature of the health shock, for which fiscal policy was the most effective tool to save lives, address the inequities from the pandemic, and help economies recover, may have reduced concerns about large stimulus packages. Fifth, norms about the risks from large debt burdens may have changed for a number of reasons: from country experiences (with countries such as Japan carrying debt at levels previously believed to be unsustainable); increased expectations that central banks could hold more debt in the future; or increased concerns about “secular stagnation” that merited more front-loaded fiscal stimulus (Eggertson et al., 2016). Finally, and closely related, the easing of rules and requirements that had previously constrained fiscal policy in high debt countries—such as IMF programs and EU treaties—may have reduced constraints in these countries (e.g., Romer and Romer, 2019).
Several of these reasons why fiscal space may not have constrained fiscal policy during the COVID-19 pandemic, however, suggest that this relationship may not persist or apply in future situations. For example, if a country's next negative shock is seen as related to domestic policies or expected to be longer lasting, financial markets may become more concerned about an increase in debt. Or, if several countries default on their debt, investors could quickly demand higher interest rates in others and cause self-fulfilling debt spirals (Aguiar et al., 2017). Fiscal space could also act as more of a constraint on fiscal responses if borrowing costs increase, such as if inflation picks up and inflation-targeting central banks tighten monetary policy more aggressively than expected.
V Interactions between Policy Choices during the COVID-19 Pandemic
The estimates in the last section show the factors that are correlated with a country's use of each policy individually, ignoring any possible interactions between different policy choices. As discussed in Section II, however, countries could use certain policies as substitutes or complements to other policies, such that the decision to use a specific policy could depend on the use of (or space to use) others. More specifically, countries that use one tool actively (such as fiscal stimulus), might have less need to use other tools to provide stimulus. Or, if a country does not have the policy space to use a preferred tool (such as being unable to lower the policy interest rate if it is already at the lower bound), it could be forced to resort to using other tools to provide stimulus that are less attractive for other reasons (such as reducing macroprudential requirements that could undermine the resilience of the financial system). Similarly, if a country does not have FX reserves available to defend the currency against depreciation pressures, it could rely on tools such as adjusting macroprudential policy or capital controls. The decision to use certain policies could also be affected by the interaction with other policies, such as if reducing interest rates increased the fiscal multiplier and thereby reduced concerns about debt sustainability, thereby making fiscal policy more attractive. For all of these reasons, the policy space available to use one tool may also affect a country's decision to use other tools. This section explores these possible interactions between different policy choices, a number of which have been modelled in the theoretical literature (discussed in Section II).
To begin, Appendix Table 3 documents the joint use of different policies. This table combines the information on individual policy tools from Section III to show simultaneously which tools were used (to any extent) by each country in the sample. The table is color coded as follows: white is no data; green indicates the tool was used to provide stimulus or ease financial conditions; red indicates the tool was not used; and yellow indicates the tool was used in a direction not usually associated with stimulus or easing of financial conditions (e.g., raising interest rates or accumulating FX reserves to dampen a currency appreciation). The right side of the table reports how many of tools were used by each country—summing the number of tools providing stimulus as well as the number used in any direction. These totals include how many categories of tools were used by each country (e.g., only counting one for monetary policy even if several types of monetary tools were adjusted) as well as how many individual tools were used. The latter allows for two forms of fiscal policy (both above- and below-the-line), four forms of monetary policy (the policy rate, asset purchases, liquidity provision to banks, and swap lines), two forms of macroprudential policy (overall and just the CCyB), and two forms of capital controls to ease net outflow pressures (on inflows and outflows).
Evaluating these actions simultaneously provides more detail on the patterns observed in Section III; most countries used a range of policies in response to COVID-19. All countries in the sample used at least two categories of tools: fiscal stimulus and containment measures. On average (and at the sample median), countries used tools from four of the six categories to stimulate their economies, with thirteen countries using tools from five of the categories, and two using tools in all of the categories (China and Turkey). When including countries using a tool in any direction (and not just to provide stimulus or ease financial conditions), the joint use of tools was even larger—mainly because over half of the countries using FX intervention accumulated reserves (instead of the usual response to a shock of spending reserves to slow currency depreciation). More specifically, 28 countries used tools from five of the six broad categories and four countries used all six types of tools in some direction (India and South Korea, in addition to China and Turkey).
Even more impressive is the range of policies activated when focusing on individual tools instead of broad policy categories. When allowing for the different forms of each policy category, countries averaged 6.4 individual tools to provide stimulus or ease financial conditions during the first six months of 2020, or 6.8 tools in any direction. The numbers would be even higher if the data on below-the-line stimulus was more widely available. (It is missing for just over one-third of the sample.) Most impressive was the multifaceted use of monetary policy; about 60% of the countries adjusted at least three of the four monetary policy tools. Most of these countries that used multiple monetary tools also used other tools. For example, of the countries using at least three forms of monetary policy, all of them (for which data is available) also provided fiscal stimulus, 84% used FX intervention (in some direction), and 65% adjusted macroprudential policy (as measured by the index).
Also noteworthy is the incidence of countries that did not simultaneously use popular policies to provide stimulus—or that simultaneously used tools pushing in different directions. For example, although all countries used some form of monetary policy to provide stimulus and/or ease financial conditions, about one-third of these countries did NOT ease macroprudential regulations49 , a policy which would also be expected to further ease financial conditions. Also, about one-third of the countries providing some form of monetary stimulus simultaneously purchased FX reserves to slow the appreciation of their currencies. At the same time, of the 25 countries using FX intervention to slow the depreciation of their currencies, only two (China and Turkey) tightened controls on capital outflows or eased controls on capital inflows—two policies which would also be expected to slow currency depreciation. In fact, three countries which reported using capital controls (India, Peru and South Korea), used them in a direction that would work against that of their FX interventions. More specifically, each of these three countries reduced controls on capital inflows (which would lead to net inflows and appreciation pressures) while simultaneously increasing FX reserves (which would depreciate the exchange rate).
In order to more formally analyze these interactions between the use of different policies during COVID-19, we begin by testing if the use of each policy is affected by the policy space available for other policies (which was shown to be a key determinant of policy use in Section IV). We repeat the baseline regression in equation 1 for each of the policy tools with the full set of control variables (including the three stress variables and other country characteristics), but also include the amount of policy space for each of the other four categories of tools (“other-policy space”), as well as continuing to include a control for the policy space available for the corresponding dependent variable (“own-policy space”).50 We do not report results for the use of capital controls because the limited use of this policy prevents meaningful estimation, but we can control for the policy space to use capital controls. We also exclude Japan from the analysis given the impact of this one outlier on the role of fiscal space (as shown in Section IV.C and Table 3).
The results are reported in Table 5 , with the coefficients on own-policy space that correspond to the analysis in Tables 2 and 3 in grey. The results on own-policy space agree with the baseline results, and the new coefficients suggest that other-policy space is generally insignificant (at the 5 percent level). For example, the size of fiscal stimulus (or just the size of below-the-line stimulus) is not significantly affected by the policy space available for any other tools—including the level of the policy interest rate, level of FX reserves, macroprudential stance, level of capital controls, or even the pre-COVID output gap. The use of macroprudential policy is also not significantly affected by the space to adjust interest rates, and the use of FX intervention and macroprudential regulation is not affected by the space to ease capital controls. These generally insignificant results for other-policy space are a striking contrast to the significant results for own-policy space. This suggests that countries do not rely more on fiscal stimulus or macroprudential easing to support the economy when they are constrained in their ability to provide monetary stimulus through reducing policy rates (as suggested in other papers).51 Countries with a tighter macroprudential stance or more stringent capital controls did not make greater use of monetary policy tools (as suggested in Aizenman et al., 2020). The results also provide little evidence that macroprudential policy, foreign exchange intervention, and adjustments to capital controls are used as substitutes, even when the use of one policy is constrained, as suggested in the IMF's Integrated Policy Framework (IMF, 2020b). All in all, the use of individual policies generally does not appear to reflect the space available to use other policies, or the level at which other tools/regulations were set prior to COVID-19. The use of different types of policy tools does not appear to be well coordinated.Table 5 Regressions Results: Policy Responses as a Function of Other Policy Space
Table 5 Fiscal Monetary FX Macropru Oxford
Stimulus / GDP Below-the-Line ∆ interest rate QEdummy QE as %GDP ∆ Reserves /GDP Dummy CCyB Contain- ment
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Policy Space Variables
Fiscal -0.0473 -0.0440 -0.000653 -0.00348 0.0135 0.0122 0.00349 0.000636 0.0215
(0.0614) (0.0589) (0.00243) (0.00701) (0.0250) (0.0160) (0.00652) (0.000894) (0.0758)
Monetary 0.0249 0.123 0.387*** -0.276** -0.420* -0.119 0.105 -0.00159 -0.535
(0.501) (0.340) (0.0974) (0.115) (0.210) (0.141) (0.0729) (0.0109) (0.539)
FX -0.0214 -0.00819 -0.0100 -0.0660*** -0.0807** -0.0505* 0.0181* -0.00201* -3.33e-05
(0.0819) (0.0682) (0.00780) (0.0239) (0.0369) (0.0250) (0.0108) (0.00111) (0.0667)
Macroprudential -2.160 1.636 -2.537* -1.946 -2.212 5.493 7.901*** 0.642*** 14.64
(11.39) (9.670) (1.332) (2.936) (7.371) (3.314) (2.789) (0.135) (12.58)
Capital Controls -1.046 -2.703 0.731 1.040 1.903 0.337 0.879 0.175 3.525
(5.501) (5.184) (0.638) (0.784) (1.691) (1.353) (1.182) (0.137) (7.302)
Output Gap -0.345 -0.354 -0.102 0.111 0.492 -0.273 -0.421** 0.0341 -0.114
(0.787) (0.823) (0.154) (0.178) (0.349) (0.271) (0.197) (0.0237) (1.202)
Stress Variables
Financial -0.355 -0.903 -0.351 -0.278 -0.112 0.379 -0.565** 0.0332 3.006
(2.010) (1.451) (0.244) (0.352) (0.673) (0.514) (0.288) (0.0392) (2.073)
Economic 0.528 0.450 0.0838 0.0283 0.376 0.0690 -0.0644 0.00146 0.378
(0.419) (0.373) (0.0686) (0.0928) (0.287) (0.116) (0.0931) (0.0174) (0.676)
Health -0.0705 -0.142 0.0309 0.00941 -0.119 -0.119 -0.00488 0.0268 1.489*
(0.319) (0.334) (0.0396) (0.0673) (0.268) (0.128) (0.0761) (0.0162) (0.743)
Other Country Characteristics
Fixed ER 0.280 3.274 0.558 -0.129 -2.268* -1.068 0.481 0.0143 -6.615
dummy (2.941) (2.480) (0.513) (0.501) (1.241) (0.855) (0.557) (0.0902) (3.966)
Institutional 0.224 -0.180 -0.0145 -0.0694 -0.109 -0.235* -0.0654 0.00712 -0.680
quality (0.425) (0.350) (0.0531) (0.0919) (0.220) (0.136) (0.0710) (0.0105) (0.421)
Trade -0.0974 -1.881 0.332 0.437 1.293 -0.339 0.0544 -0.0269 -2.965
openness (3.129) (2.879) (0.316) (0.719) (1.076) (1.400) (0.606) (0.0496) (3.284)
EM dummy -4.570 -5.361 -0.903 -1.154 -4.647* -1.007 -1.758* -0.205 5.273
(4.219) (3.806) (0.693) (0.872) (2.406) (1.260) (0.957) (0.128) (5.525)
Observations 37 37 42 42 42 43 56 52 56
Adjusted R-squared 0.108 0.062 0.674 0.389 0.184 0.413 0.386 0.830 0.368
Notes: Regressions predicting the use of each policy response to COVID-19 as a function of policy space for the given policy (shaded) and the space available for other policies. Japan is excluded from the fiscal regressions. See notes to Table 1 and Appendix Table 1 for variable definitions. Policy space for each category listed at the left is measured as: fiscal is the net debt/GDP ratio; monetary is the policy interest rate; FX is the ratio of FX reserves to GDP; macroprudential is the level of the macroprudential index of the CCyB, LTV ratio and FX intensity; capital controls is the index of the intensity of controls on inflows and outflows; and output gap is the output gap at end-2019. All regressions include a constant (not reported) and are estimated with robust standard errors. *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.
The one area where there may be more coordination across tools based on the policy space available, however, is in the choice of which type of tool to use to provide monetary stimulus. Columns 4 and 5 of Table 5 show that countries with higher policy interest rates before COVID-19 were significantly less likely to announce new asset purchases and adopted smaller quantitative easing programs.52 In other words, countries with more space to use the “conventional” monetary tool of lowering policy interest rates were less likely to resort to the “unconventional” tool of asset purchases. For central banks that would prefer to adjust monetary policy through adjustments to policy rates (and not asset purchases), this is an important reason to raise interest rates to create this policy space when feasible.
As a final analysis of joint policy decisions, we extend this framework but estimate the use of multiple policy tools using a Seemingly Unrelated Regression (SUR) model. This adjusts for the correlation in the contemporaneous errors for each of the equations predicting each policy choice, but has the disadvantage of limiting the sample size to countries with data for each policy response. For policy tools, we focus on the five quantitative measures (instead of the dummy variables) for each of the main policies used in response to COVID-19 in order to have sufficient degrees of freedom to estimate the equations jointly. These five policy tools are: the announced size of the fiscal stimulus, the reduction in the policy interest rate, the reduction in FX reserves relative to GDP (according to the Adler et al., 2021 measure), the reduction in the CCyB, and the change in the health and containment index (from Oxford). We continue to control for the same set of policy space variables used in the baseline analysis (including for capital controls) and the three measures of stress and controls for other country characteristics.
The results from estimating these five policy choices simultaneously are reported in Table 6 . They support the key results from when each policy choice is estimated individually (Table 2, Table 3, Table 4) and when controls for other-policy space are included in Table 5. Given the sensitivity of the results for fiscal space to the inclusion of Japan, the left side of the table reports results for the largest sample possible, and the right side of the table excludes Japan. Although the results are very similar across the two sides of the table, this comparison again highlights the role of this outlier. When Japan is included in the sample, there is a significant, negative correlation between policy space and the announced size of fiscal stimulus. When Japan is excluded, the coefficient becomes insignificant and positive, although there still appears to be no significant constraint of fiscal space on the size of the announced fiscal response to COVID-19. All of the other coefficients agree with the earlier estimates and support the important role of policy space for the use of most policy tools (other than fiscal policy). More specifically, countries with higher policy rates and a higher CCyB before the pandemic then lowered interest rates and the CCyB more aggressively in response to COVID-19. Countries with higher FX reserve ratios intervened in FX markets by more, albeit building reserves on average instead of depleting them. The policy space for the other tools usually had an insignificant effect, continuing to suggest little coordination between policy responses.Table 6 Regression Results: Policy Responses as a Function of Other Policy Space Using SUR
Table 6 Full Sample with all Tools (excludes Euro Area) Also Excludes Japan
Fiscal Stimulus Monetary Policy Rate FX Interven-tion Macropru: CCyB Health: Containment Fiscal Stimulus Monetary Policy Rate FX Interven-tion Macropru: CCyB Health: Containment
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Policy Space Variables
Fiscal -0.0740*** -0.000164 0.0124 0.00133 0.0123 0.0113 0.00144 0.0437** 0.00641* -0.250***
(0.0243) (0.00213) (0.0123) (0.00201) (0.0885) (0.0314) (0.00464) (0.0178) (0.00338) (0.0846)
Monetary -0.284 0.378*** 0.0251 -0.0374 -0.638 -0.229 0.376*** -0.0145 -0.0398 -0.306
(0.583) (0.0796) (0.154) (0.0302) (0.782) (0.492) (0.0789) (0.151) (0.0303) (0.622)
FX 0.0223 -0.0129** -0.0484** -0.00269 0.0249 -0.0221 -0.0142* -0.0739*** -0.00712* 0.239**
(0.0491) (0.00638) (0.0202) (0.00369) (0.123) (0.0455) (0.00769) (0.0175) (0.00431) (0.116)
Macroprudential 3.433 -2.862** 5.775** 2.902*** 22.97 0.776 -2.964** 3.784 2.575*** 39.67**
(8.746) (1.127) (2.583) (0.647) (16.54) (7.759) (1.240) (2.508) (0.707) (17.26)
Capital Controls -5.507** 0.747 -0.0529 0.0817 3.238 0.104 0.821 1.396 0.346 -8.920
(2.451) (0.509) (1.185) (0.265) (7.154) (2.496) (0.574) (1.117) (0.308) (5.812)
Output Gap -0.180 -0.0975 -0.185 0.0855* -0.783 -0.328 -0.0928 -0.0938 0.0959* -1.550
(0.498) (0.122) (0.224) (0.0498) (1.305) (0.364) (0.123) (0.201) (0.0501) (1.086)
Stress Variables
Financial 1.067 -0.342* 0.940** 0.0391 1.941 -1.373 -0.355* 0.702 -0.00549 3.937**
(1.756) (0.194) (0.426) (0.125) (2.447) (1.632) (0.196) (0.525) (0.123) (1.921)
Economic 0.397 0.0930 0.0845 0.0572* 0.647 0.635** 0.0948 0.119 0.0642** 0.355
(0.332) (0.0568) (0.0958) (0.0305) (0.603) (0.297) (0.0578) (0.0901) (0.0314) (0.479)
Health -0.184 0.0275 -0.130 0.0289 1.492** 0.0918 0.0319 -0.0450 0.0429* 0.782**
(0.264) (0.0338) (0.110) (0.0247) (0.735) (0.228) (0.0347) (0.0749) (0.0254) (0.381)
Other Country Characteristics
Fixed ER -1.798 0.534 -1.221* 0.184 -7.137* -1.700 0.544 -1.039 0.214 -8.670***
dummy (2.179) (0.411) (0.691) (0.182) (3.889) (1.807) (0.410) (0.639) (0.186) (3.353)
Institutional 0.0965 -0.0212 -0.167 -0.0199 -0.752* -0.0434 -0.0232 -0.208* -0.0213 -0.411
quality (0.383) (0.0444) (0.122) (0.0225) (0.444) (0.292) (0.0448) (0.111) (0.0237) (0.374)
Trade 0.510 0.451* -0.149 -0.126 -4.546 2.240 0.502 0.847 0.0441 -12.90***
openness (1.900) (0.261) (1.220) (0.170) (5.312) (2.080) (0.326) (0.970) (0.169) (3.986)
EM dummy -2.601 -1.072* -1.039 -0.641** 7.367 -5.216** -1.108* -1.749 -0.726** 13.32**
(2.958) (0.580) (1.110) (0.311) (6.731) (2.306) (0.589) (1.070) (0.297) (6.384)
Observations 41 41 41 41 41 40 40 40 40 40
Notes: Regressions predicting the policy response listed at the top column as function of policy space, stress measures, and other country characteristics. Each set of five policy responses are estimated jointly using Seemingly Unrelated Regression (SUR). Fiscal stimulus is the stimulus relative to GDP; Monetary Policy Rate is the change in the policy interest rate; FX intervention is the change in FX reserves relative to GDP from Adler et al. (2021); Macropru: CCyB is the change in the CCyB and Health: Containment is the change in the Oxford Containment measure. All dependent variables are for 2020 Q1-Q2. Columns 6-10 exclude Japan. All regressions include a constant (not reported) and robust standard errors. *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Appendix Table 1 Data Sources and Definitions
Appendix Table 1 Variable Measure Source Link
Covid-19 Policies (from 01/01/2020 until 06/30/2020 unless stated otherwise) Fiscal (incl. split ATL and BTL) Fiscal interventions until 06/12/2020 in percent of GDP Fiscal Monitor Database of Country Fiscal Measures in Response to COVID-19 and IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Fiscal-Policies-Database-in-Response-to-COVID-19 and https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
Monetary - change policy rate Change Haver; official CB website (Costa Rica); BIS (China) not public
Monetary - change shadow rate Change Krippner (2018) https://www.ljkmfa.com/test-test/international-ssrs/
Monetary QE Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
Monetary QE/Asset purchases Net Purchases in percent of GDP For AE: CB websites; for EMs: Fratto et al. (2021); https://www.imf.org/en/Publications/WP/Issues/2021/01/22/Unconventional-Monetary-Policies-in-Emerging-Markets-and-Frontier-Countries-50013
Monetary - Liquidity Provision Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
Monetary - Swap Line Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
Macroprudential Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
CCyB Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
CCyB Change ESRB and BIS; national sources (Morrocco & Kazahstan) https://www.bis.org/bcbs/ccyb/ and https://www.esrb.europa.eu/national_policy/ccb/html/index.en.html
FX Intervention Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
FX Intervention Net Purchases in percent of GDP Adler et al. (2021) https://www.imf.org/en/Publications/WP/Issues/2021/02/19/Foreign-Exchange-Intervention-A-Dataset-of-Public-Data-and-Proxies-50017
Capital Controls on Inflows Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
Capital Controls on Outflows Dummy IMF Policy Tracker for Responses to Covid-19 (6/30/2020 vintage) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19
Containment and Health Policies Change Oxford's Coronavirus Government Response Tracker https://www.bsg.ox.ac.uk/research/research-projects/covid-19-government-response-tracker
Policy Space Variables (latest available observation before 2020) Fiscal Space Government debt in percent of GDP IMF Financial Monitor Database https://data.imf.org/?sk=4be0c9cb-272a-4667-8892-34b582b21ba6
Monetary Space - policy rate Policy Rate Haver; official CB website (Costa Rica); BIS (China) not public
Monetary Space - shadow rate Shadow Rate Krippner (2018) https://www.ljkmfa.com/test-test/international-ssrs/
CCyB space CCyB ESRB and BIS complemented with official national sources https://www.bis.org/bcbs/ccyb/ and https://www.esrb.europa.eu/national_policy/ccb/html/index.en.html
Macroprudential Space Index composed of (i) FX Restriction, (ii) LTV ratio, and (iii) CCyB (all equally weighted) Alam et al. (2019) for (i) and (ii); ESRB and BIS for (iii) https://www.imf.org/en/Publications/WP/Issues/2019/03/22/Digging-Deeper-Evidence-on-the-Effects-of-Macroprudential-Policies-from-a-New-Database-46658
FX Reserves Space Stock of FX Reserves in percent of GDP IMF BoP Database, ECB website for euro area https://data.imf.org/?sk=7A51304B-6426-40C0-83DD-CA473CA1FD52 https://www.ecb.europa.eu/stats/balance_of_payments_and_external/international_reserves/templates/html/201912eur.en.html
Capital Controls Inflows Space Last value of Index before the pandemic Fernandez et al. (2016) http://www.columbia.edu/∼mu2166/fkrsu/
Capital Controls Outflows Space Last value of Index before the pandemic Fernandez et al. (2016) http://www.columbia.edu/∼mu2166/fkrsu/
Output Gap Real GDP - potential real output World Economic Outlook https://www.imf.org/en/Publications/SPROLLs/world-economic-outlook-databases#sort=%40imfdate%20descending
Stress during COVID-19 Financial Stress Change from Pre-Covid value (12/31/2019) to Peak (from 1.1.-06/30/2020) CDS from Bloomberg and if not available: EMBIG index from JPMorgan not public
Real Stress Change in GDP forecast for 2020 from Jan 20 - Jun 20) World Economic Outlook (in between updates for Jan and Jun 2020) not public
Health Stress Reported cases from 01/01/2020 until 06/30/2020 Oxford's Coronavirus Government Response Tracker https://www.bsg.ox.ac.uk/research/research-projects/covid-19-government-response-tracker
Country Characteristics (latest available observation before 2020) EM Dummy World Economic Outlook https://www.imf.org/en/Publications/SPROLLs/world-economic-outlook-databases#sort=%40imfdate%20descending
Fixed ER Dummy Ilzetzki et al. (2019) https://www.ilzetzki.com/irr-data
Institutional Quality Composite Index The PRS Group not public
Trade Openness (Imports+Exports)/GDP IMF BoP Database https://data.imf.org/?sk=7A51304B-6426-40C0-83DD-CA473CA1FD52
Note: Most of the data collected in October and November 2020, unless published later. FX Reserves and GDP data was updated in July 2022 as these variables are sensitive to revisions. Please note that some of these databases, e.g. World Economic Outlook, are live databases and data revisions can occur.
Appendix Table 2 Share of Countries Reporting Use of Each Policy
Appendix Table 2 Share of Total
Policy Type Instrument Full Sample AE EM Observations
Monetary Asset Purchases/QE 42.6% 61.1% 33.3% 54
Liquidity to Banks 85.7% 83.3% 86.8% 56
Swap Line activated 41.5% 70.6% 27.8% 53
External FX Intervention 44.6% 16.7% 57.9% 56
CFM on inflows 5.4% 2.8% 7.9% 74
CFM on outflows 2.7% 0.0% 5.3% 74
Macroprudential Overall Index 66.2% 72.2% 60.5% 74
CCyB 29.7% 44.4% 15.8% 74
Notes: Share of countries that report using each policy during 2020q1 and 2021q2 according to a 0/1 dummy variable. Statistics for each group only include countries that have the ability to adopt each set of policies, e.g., individual countries in the Euro area cannot pursue monetary policy or FX intervention, but can adopt other policies. The Euro area is included as a "country" that can pursue monetary and FX policy, but not other policies. CFM is capital flow measures and CCyB is the countercyclical capital buffer. AE is advanced economies and EM is emerging markets, according to IMF definitions.
Sources: Reported use of each policy is from the IMF Policy Tracker.
Appendix Table 3 : Joint Use of Policies
Appendix Table 3Image, table 9
Appendix Table 4 : Countries/Entities included in Each Regression
Appendix Table 4Image, table 10
VI Conclusions
This series of results suggests that policy space is an important determinant of how a country responds to a shock—and especially the policy space for the given tool (albeit less so for the policy space available for other tools). More specifically, having more policy space is a significant determinant of a country's ability to: provide monetary stimulus by lowering interest rates; engage in FX intervention to support the exchange rate; and ease macroprudential buffers (including the CCyB) to support lending and access to credit. These results are not surprising. For countries with very low interest rates, it is more difficult to lower interest rates further (even when measured using the shadow rate). For countries that had not previously tightened or even used macroprudential buffers (such as raising a CCyB above zero), it is difficult to lower these buffers to provide support. For countries that had not accumulated FX reserves, it is more challenging to use FX intervention in any direction.53 There is also evidence that countries with more space to adjust monetary policy through the “conventional” tool of reducing interest rates are significantly less likely to use “unconventional” forms of monetary policy (such as asset purchase programs and liquidity provision to banks). This suggests that the traditional hierarchy of central bank tools (of first using interest rates to provide monetary stimulus, and then shifting to asset purchases when interest rates are near zero) was still a consideration during the COVID-19 pandemic, even though some countries with policy rates above their lower bounds also adopted unconventional monetary responses.
More noteworthy are the results on the policy tool that did not appear to be significantly constrained by policy space: announcements of fiscal stimulus. Advanced economies with higher debt to GDP ratios were not significantly constrained in their ability to announce large fiscal stimulus packages, and emerging markets only appeared to be constrained in some specifications. Advanced economies with higher debt burdens, however, did announce more of this fiscal stimulus through below-the-line policies (such as credit guarantees and loan programs). This suggests that countries with less fiscal space had a stronger impetus to moderate further increases in debt by keeping more of the stimulus off-balance sheet. Moreover, the total size of the fiscal stimulus announced in response to COVID-19 not only appeared to be unaffected by a country's debt ratios, but also appeared to be unaffected by any other variables. More specifically, the magnitude of a country's announced fiscal stimulus in the first half of 2020 appears to be unrelated to its policy responses via other tools, to its policy space available for other tools, to its output gap before the pandemic, to its degree of financial market stress, to its contraction in GDP growth, and even to the number of COVID-19 cases.
These results that a country's fiscal space did not seem to constrain its aggregate fiscal stimulus announced in response to the pandemic in advanced economies, and that a country's announced fiscal response seemed unrelated to many standard economic and financial variables, suggests that these relationships changed relative to earlier financial crises and recessions (e.g., Romer and Romer, 2018, 2019; and Jordà et al. 2016). A better understanding of what caused these changes is an important topic for future work, and Section IV.C discusses several possible explanations. Many of these explanations, however, suggest that the apparent lack of a relationship between fiscal space and the announced size of the fiscal stimulus during COVID-19 in advanced economies may not persist and should not be counted on in future situations.
Finally, fiscal stimulus was not the only policy pursued largely independent of other policy choices, as well as independent of the space available to use other policy tools. Adjustments to monetary policy were largely independent of the space available to provide fiscal stimulus, to ease macroprudential regulations, to intervene in FX markets, and to modify capital controls. Adjustments to macroprudential policy were largely independent of the space to loosen monetary policy, intervene in FX markets, and to modify capital controls. In fact, in some cases different policies seemed to be used in directions that would counteract each other, such as some countries lowering interest rates while using reserves to appreciate the exchange rate. A number of papers have modelled how different policy tools should be used as substitutes and/or complements to other tools—a degree of coordination that did not seem to exist during COVID-19. This suggests that there could be substantial room to improve the policy responses to future shocks by better incorporating the interactions between policy choices, including the constraints from the space available to use different tools.
Uncited References
(Bouabdallah et al., 2020, Chen and Friedrich, 2021, Eggertsson et al., 2016, Fernández et al., 2016, Kose et al., 2012, Maravalle and Rawdanowicz, 2020, McKay and Reis, 2021, www.macromodelbase.com., Volker 2022, Altavilla et al., 2020b)
Appendix Supplementary materials
Image, application 1
Image, application 2
Data Availability
We are able to share all of the variables related to policy responses and pre-existing policy space used in the analysis. We are not able to share two confidential control variables.
54 More specifically, the measure includes school closings, workplace closings, cancellation of public events, restrictions on gatherings, stay-at-home requirements, restrictions on international movement, international travel controls, public information campaigns, testing policy, contact tracing, facial coverings, and vaccination policies.
1 The views expressed in this paper are those of the authors and do not necessarily represent the views of the IMF, its Executive Board, IMF management, or any other institution with which the authors are affiliated. The authors can be contacted at: kbergant@imf.org and kjforbes@mit.edu. We thank Gian Maria Milesi-Ferretti, Volker Wieland, and all participants of the IMF's Macro-Financial Research Conference 2021 and the CEPR conference on “The COVID-shock and the new macroeconomic landscape” for helpful comments and discussions. Further thanks to JPMorgan for sharing their data on EMBI spreads.
2 This paper assesses why countries adopted different policies. An important question for future research is the efficacy of these policies, including whether they were used optimally. For an example of this type of analysis, see Wieland (2022), which assesses the impact of the fiscal response to COVID-19 in the Euro area.
3 See Aizenman et al. (2017), Adrian et al. (2020), Basu et al. (2020), Bergant et al. (2020), and Mano and Sgherri (2020).
4 As discussed in Section III, these measures include the support that was announced, even if not fully utilized.
5 As discussed in more detail below, this is based on data from Adler et al. (2021), which only includes reserve sales/accumulation intended for FX intervention.
6 It is important to highlight that using more (or stronger) policies is not always desirable, and using many policies at once is not necessarily a sign of policy coordination. We consider the use of policies as coordinated if the use (or space to use) one policy affects the use of another.
7 For example, central banks are often responsible for, or play a key role in, implementing both monetary and macroprudential policy in many countries.
8 For an evaluation of how the fiscal response to COVID-19 affected debt sustainability, growth and labor market institutions in the Euro area, see Wieland (2022). For an evaluation of the impact of fiscal transfers on inflation and wages, see Jordà and Nechio (2022).
9 For an overview of the literature on fiscal multipliers, including the role of country characteristics and policy space, see Ramey (2019).
10 See IMF website for a list of related papers.
11 For an excellent set of papers analyzing effects of the pandemic and policy responses, see the CEPR/EC/EER conference on “The COVID-shock and the New Macroeconomic Landscape: Taking Stock and Looking Ahead,” in Brussels, 6-7 October 2022. In addition, Kirti et al. (2022b) analyze the impact of fiscal, monetary, and prudential policies during the COVID-19 pandemic on bank lending.
12 Also see the Macroeconomic Model Database at https://www.macromodelbase.com/ for an archive of over 150 structural macroeconomic models that can be used to assess the impact of policy responses. For information on this database, see Wieland et al. (2012) and Wieland et al. (2016).
13 In some cases, the data in the IMF's Policy Tracker differs from other sources. In these cases, and to be consistent across countries, we rely on the data in the IMF's Policy Tracker unless noted explicitly in the text.
14 From the Fiscal Monitor Database of Country Fiscal Measures in Response to the COVID-19 Pandemic, with data through June 12. Also see IMF (2020a). We have also collected data on above-the-line fiscal stimulus as a share of GDP from the IMF's Policy Tracker. Our key results reported below are unchanged with this alternate measure when holding the sample constant, but Section IV.C discusses how some results can change based on the sample composition.
15 For example, Wieland (2022) shows that only a small percentage of the announced fiscal support for businesses was drawn down in certain Euro area countries. Hong and Lucas (2023) shows that only a portion of the announced credit and liquidity support was used in their sample of seven advanced economies.
16 Available at: International SSR estimates (ljkmfa.com). For more information on the calculation of these shadow rates, see Krippner (2015). We include shadow rates for eight entities: Australia, Canada, Euro area, Japan, New Zealand, Switzerland, UK, and US.
17 The magnitude of asset purchases is from Fratto et al. (2021).
18 Scrapping is the process by which data from a website is extracted, collected, and exported in a format that is more useful for the analytical user. More specifically, the IMF data on policy responses is only reported individually by country and not easily comparable across countries. The “scrapped” data extracts the key information from each country report and compiles the data in a way that is directly comparable across countries. The updated data is available at: Policy Responses to COVID19 (imf.org), of which we used the publication of 6/30/2020 in order to capture the “initial” response to the pandemic.
19 Argentina reduced its policy rate by 22pp over this window but is not shown on the graph or included in the statistics or analysis below as it can distort the graph and affect some of the empirical estimates.
20 Simply using changes in FX reserves as reported in the Balance of Payments (a method often employed in the literature) also captures changes in FX reserves unrelated to exchange rate management, including large movements in countries that do not actively intervene in foreign exchange markets.
21 Altavilla et al. (2020a) shows that macroprudential regulation was crucial for supporting bank lending during the early months of the pandemic.
22 Data for changes in the CCyB are from the BIS (www.bis.org/bcbs/ccyb/) and ESRB (www.esrb.europa.eu/national_policy/ccb/html/index.en.html) and then cross-checked with Chen and Friedrich (2020). Several countries report a loosening in counter-cyclical buffers in the IMF Policy Tracker, but are not reported as loosening in the BIS and ESRB data. We check these examples with country specific sources. In most cases, this reflects countries which reduced some buffer on selected institutions, but not a macroprudential CCyB on the entire banking system. For example, the Netherlands reduced a CCyB for selected SIFIs, with different changes for different institutions. In these cases, we do not adjust the raw data. The only exceptions are for two countries not included in the BIS and ESRB data: Morocco (which lowered its CCyB from 2.5% to 2.0%) and Kazakhstan (which lowered its CCyB by 1pp for all institutions, starting from a higher level for SIFIs).
23 Although members of the Euro area are restricted in their ability to use capital controls with respect to other Euro area countries, they can enact controls in certain circumstances and with respect to non-Euro area countries, so we include countries in this region as individual entities.
24 Compiled by the University of Oxford and available at: Coronavirus Government Response Tracker | Blavatnik School of Government (ox.ac.uk)
25 This table uses the measure of fiscal policy from the IMF Policy Tracker, instead of the IMF Fiscal Monitor (which is the focus of Figure 1, Table 1 and the discussion above) in order to include information for a larger set of countries. As discussed in more detail in the sensitivity analysis in Section IV.C, the IMF Policy Tracker includes a larger sample of countries, but the size of the stimulus is only available relative to 2020 GDP, which generates concerns about endogeneity and could bias empirical analysis. For the color-coding in Appendix Table 3, however, we can augment data from the Fiscal Monitor with this flawed data from the IMF Policy Tracker without affecting any key results, as the table just shows the direction of any fiscal stimulus and not the magnitude.
26 For example, the tool of adjusting interest rates is not available for individual countries that are the member of a currency union and/or that are dollarized.
27 There are other measures for fiscal space proposed in the literature, a number of which we explore in the sensitivity analysis. We focus on debt to GDP ratios as Romer and Romer (2019) point out that these variables are a useful measure of policy space as they are slow moving and less cyclically sensitive (as compared to measures such as budget balances or financing costs). They also capture past policy decisions and “more long-run features of a country's policymaking process”.
28 For the Euro area, policy space for FX intervention is Eurosystem FX reserve holdings (relative to Euro area GDP).
29 The index is constructed following the methodology in Bergant and Forbes (2023) and Chari et al. (2022) in order to more precisely measure the intensity of macroprudential policy while including a range of policies targeting key vulnerabilities (for banks, the housing market, and FX exposures). The index combines the two quantitative measures of specific macroprudential policies that are comparable across countries (the CCyB and LTV ratio) with a constructed FX index. Each of the three components is written so that a higher value is a more stringent policy and then scaled so that each component receives equal weight. The FX index is constructed using data in Alam et al. (2019), updated in Oct. 2020, as the cumulated change in FX regulations since 2000 (as done in Bergant et al., 2020 and Forbes, 2021). Data on the CCyB is discussed above, and data on the LTV ratio is from Alam et al. (2019), updated in Oct. 2020.
30 Updated as of June 2019, with data through 2017. We use the 2017 value as a pre-COVID-19 level.
31 We combine both the change and percent change in order to better compare stress across countries with different starting points. Focusing only on percent changes for countries with very low CDS/spreads can overstate the degree of stress.
32 Measured as the change from the Jan 2020 forecast (which was prepared in Dec 2019) through the June 2020 forecast.
33 The data ends in 2016, and we assume the exchange rate regime did not change through 2019. We define a fixed exchange rate regime using the “coarse classifications” and define all countries as fixed if they have a moving band that is narrower than or equal to +/- 2% (classification #11) or anything more restrictive.
34 Exposure to commodity prices is measured as the volatility in the commodity terms-of-trade index from 2008-2018, based on the data in Gruss and Kebhaj (2019).
35 For example, there is not sufficient variation to estimate equation (1) for reduced controls on capital outflows.
36 We abridge the results reported for liquidity provision to banks and FX swaps as there are no meaningful changes in the additional specifications, and this allows us to combine these results with those for FX intervention in one panel.
37 More specifically, see Section 3 for a discussion of data limitations for some variables and the different entities that have the ability to use each tool (e.g., the ECB sets monetary policy in the Euro area instead of individual countries). The list of included countries/entities in Appendix Table 4 is from the baseline regression for each tool, which includes the stress variables, all of the control variables, and the EM dummy interaction.
38 The tables and discussion below focus on results when the interest rate is measured using the policy interest rate instead of the shadow rate. Key results are unchanged when using the shadow rate—so we do not report or discuss the later set of results to save space.
39 Section IV.C explores these results in more detail. It shows that the significant negative coefficient on fiscal space is affected by sample composition (and especially if Japan is included in the analysis). Adjustments to the sample and measurement of fiscal space often render the coefficient on fiscal space insignificant, but it rarely becomes positive and does not become positive and significant as found in past work. See the next section for more details.
40 As discussed in Section III, this captures fiscal stimulus that was announced, even if not fully utilized.
41 This may reflect the high correlation between these country characteristics and the level of the policy interest rate (i.e., monetary policy space), as countries with stronger institutions and flexible exchange rates had lower policy rates before the pandemic.
42 All new measures of fiscal space are from Kose et al. (2022) and the corresponding data set.
43 The sample of countries for the regressions of fiscal policy in the baseline analysis are constrained by the data available in the IMF's Fiscal Monitor. See Section III for a discussion of this data.
44 See the discussion in Section II and Wieland (2022).
45 Calculated as the volatility in the commodity terms-of-trade index, with the index capturing reliance on commodity exports or imports as reported in Gruss and Kebhaj (2019).
46 Based on Fitch Ratings converted to a numerical scale, with a higher number indicating a stronger credit rating.
47 We also estimate a series of tests to explore if the relationship between fiscal space and the fiscal response to COVID-19 is affected by the financial market response, as suggested in Romer and Romer (2019). More specifically, we exclude the control for financial stress or interact this with fiscal space. In these extensions, the coefficient on fiscal space remains negative and significant, and the additional interaction with financial stress is insignificant.
48 For example, we also estimate sensitivity tests with two additional variables that have been highlighted in other papers: the size of the financial sector and real credit growth. These reduce the sample size and make meaningful estimation impossible for a number of policy tools.
49 Granted, our index of macroprudential policy only includes adjustments in the LTV ratio, CCyB, and FX-related measures, so could miss adjustments in macroprudential regulation that are not included in these categories.
50 We measure policy space in 2019 for each category of tools using: the ratio of debt to GDP for fiscal policy; the policy interest rate for monetary policy; the ratio of FX reserves to GDP for FX intervention; the index of the macroprudential policy stance for macroprudential policy; the index of controls on inflows and outflows for capital controls; and the output gap for containment measures.
51 For example, see Eggertsson (2011), Woodford (2011), Drautzburg and Uhlig (2015), Bouakez et al. (2017), Bernanke (2020), and Furman and Summers (2020).
52 This is similar to the comparable coefficient estimates in Table 2, except these relationships between monetary policy space and the “unconventional” monetary policy tools are now consistently significant.
53 For countries that do not often intervene in FX markets, it can be more difficult to accumulate reserves (as well as to use reserve stockpiles) as the institutional framework and expertise is not as well developed.
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.euroecorev.2023.104499.
==== Refs
REFERENCES
Adler Gustavo Chang Kyun Suk Mano Rui Shao Yuting Foreign Exchange Intervention: A Dataset of Public Data and Proxies IMF Working Paper WP/21/47 2021
Adrian Tobias Erceg Christopher Linde Jesper Zabczyk Pawel Zhou Jianping A Quantitative Model for the Integrated Policy Framework IMF Working Paper 2020 20/122
Aguiar Mark Chatterjee Satyajit Cole Harold Stangebye Zachary Self-Fulfilling Debt Crises, Revisited: The Art of the Desperate Deal NBER Working Paper No. 23312 2017
Aizenman Joshua Jinjarak Yothin The Fiscal Stimulus of 2008-2009: Trade Openness, Fiscal Space and Exchange Rate Adjustment NBER International Seminar on Macroeconomics 8 1 2011 301 342 pgs
Aizenman Joshua Chinn Menzie Ito Hiro Financial Spillovers and Macroprudential Policies Open Economies Review 31 3 2020 529 563
Aizenman Joshua Jinjarak Yothin Nguyen Hien Thi Kim Park Donghyun Fiscal Space and Government Spending and Tax-Rate Cyclicality Patterns: A Cross-Country Comparison, 1960-2016 Journal of Macroeconomics 60 2019 229 252
Alam Zohair Alter Adrian Eiseman Jesse Gelos Gaston Kang Heedon Narita Machiko Nier Erlend Wang Naixi Digging Deeper—Evidence on the Effects of Macroprudential Policies from a New Database IMF Working Paper 2019 No. 19/66
Altavilla Carlo Barbiero Francesca Boucinha Miguel Burlon Lorenzo The Great Lockdown: Pandemic Response Policies and Bank Lending Conditions CEPR Discussion Paper No 15298 2020
Altavilla Carlo Laeven Luc Peydro Jose Luis Monetary and Macroprudential Policy Complementarities: Evidence from European Credit Registries CEPR Discussion Paper No 15539 2020
Auerbach Alan Gorodnichenko Yuriy Fiscal Stimulus and Fiscal Sustainability.” in Fostering a Dynamic Global Economy Proceedings of the Kansas City Federal Reserve Banks annual symposium at Jackson Hole, Wyoming 2017
Auerbach Alan Gorodnichenko Yuriy Murphy Daniel Inequality, Fiscal Policy and Covid19 Restrictions in a Demand-determined Economy European Economic Review 137 2021 103810
Bartsch Elga Bénassy-Quéré Agnès Corsetti Giancarlo Debrun Xavier How Can Monetary and Fiscal Policies Work or Fail Together? Geneva Reports on the World Economy 23 2020
Basu Suman Boz Emine Gopinath Gita Roch Francisco Unsal Filiz A Conceptual Model for the Integrated Policy Framework IMF Working Paper No. 20/121 2020
Benmelech Efraim Tzur-Ilan Nitzan The Determinants of Fiscal and Monetary Policies During the Covid-19 Crisis NBER Working Paper 2020 27461
Bergant, Katharina and Kristin Forbes. (2023). “Macroprudential Policy During COVID-19: The Role of Policy Space,” in Claudio Borio, Edward Robinson, and Hyun Song Shin, eds., Macro-Financial Stability in a Globalised World: Lessons from International Experience. Chapter 4. Forthcoming.
Bergant Katharina Grigoli Francesco Hansen Niels-Jakob Sandri Damiano Dampening Global Financial Shocks: Can Macroprudential Regulation Help (More than Capital Controls)? IMF Working Paper 20/106 2020
Bernanke Ben. The New Tools of Monetary Policy American Economic Review 110 4 2020 943 983
Bi Huixin Shen Wenyi Shu-Chun Yang Debt-Dependent Effects of Fiscal Expansions European Economic Review 88 2016 142 157
Bigio Saki Zhang Mengbo Zilberman Eduardo Transfers vs Credit Policy: Macroeconomic Policy Trade-offs During Covid-19 NBER Working Paper 27118 2020
Blanchard Olivier J. Summers Lawrence H. Automatic Stabilizers in a Low-Rate Environment AEA Papers and Proceedings 110 2020 125 130
Bouakez Hafedh Guillard Michel Roulleau-Pasdeloup Jordan Public Investment, Time to Build, and the Zero Lower Bound Review of Economic Dynamics 23 2017 60 79
Bouabdallah Othman Checherita-Westphal Cristina Freier Maximilian Muggenthaler Philip Müller Georg Nerlich Carolin Sławińska Kamila Automatic fiscal stabilisers in the euro area and the COVID-19 crisis Economic Bulletin Articles 6 2020 European Central Bank
Chari Anusha Dilts-Stedman Karlye Forbes Kristin Spillovers at the Extremes: The Macroprudential Stance and Vulnerability to the Global Financial Cycle Journal of International Economics 136 2022 1103582
Chen, David and Christian Friedrich. (2021). “The Countercyclical Capital Buffer and International Bank Lending: Evidence from Canada.” Bank of Canada Staff Working Paper 2021-61.
Chinn Menzie Ito Hiro What Matters for Financial Development? Capital Controls, Institutions and Interactions Journal of Development Economics 81 2006 163 192
Drautzburg Thorsten Uhlig Harald Fiscal stimulus and distortionary taxation Review of Economic Dynamics 18 4 2015 894 920
Eggertsson Gauti. What Fiscal Policy is Effective at Zero Interest Rates? NBER Macroeconomics Annual 25 1 2011 59 112
Eggertsson Gauti Mehrotra Neil Summers Lawrence Secular Stagnation in the Open Economy American Economic Review 106 5 2016 503 507
Eichenbaum Martin Rebelo Sergio Trabandt Mathias The Macroeconomics of Epidemics The Review of Financial Studies 34 11 2021 5149 5187
English, William, Kristin Forbes and Angel Ubide. (2022). Monetary Policy and Central Banking in the Covid Era. E-book published by Centre for Economic Policy Research (June).
Fernández Andrés Klein Michael Rebucci Alessandro Schindler Martin Uribe Martín Capital Control Measures: A New Dataset IMF Economic Review 64 3 2016 548 574
Forbes Kristin. The International Aspects of Macroprudential Policy Annual Review of Economics 13 August 2021 203 228
Forbes Kristin Klein Michael Pick Your Poison: The Choices and Consequences of Policy Responses to Crises IMF Economic Review 63 1 2015 197 237
Fratto Chiara Vannier Breandon Harnoys Mircheva Borislava David de Padua, Helen Poirson Ward Unconventional Monetary Policies in Emerging Markets and Frontier Countries IMF Working Paper 2021/014 2021
Furman Jason Summers Lawrence A Reconsideration of Fiscal Policy in the Era of Low Interest Rates Brookings discussion draft 2020 Link
Ghosh Atish Kim Jun Mendoza Enrique Ostry Jonathan Qureshi Mahvesh Fiscal Fatigue, Fiscal Space and Debt Sustainability in Advanced Economics The Economic Journal 123 Feb 2013 F4 F30
Ghosh Atish Ostry Jonathan Qureshi Mahvesh Managing the Tide: How do Emerging Markets Respond to Capital Flows? IMF Working Paper 2017 17/69
Gourinchas Pierre-Olivier Kalemli-Ozcan Sebnem Penciakova Veronika Sander Nick COVID-19 and SME Failures NBER Working Paper 27877 2020
Guerrieri Veronica Lorenzoni Guido Straub Ludwig Werning Ivan Macroeconomic Implications of COVID-19: Can Negative Supply Shocks Cause Demand Shortages? American Economic Review 112 5 2022 1437 1474
Gruss Bertrand Kebhaj Suhaib Commodity Terms of Trade: A New Database International Monetary Fund Working Paper 19/21 2019
Heinemann Friedrich. Moving From Broad to Targeted Pandemic Fiscal Support EconPol Policy Report No 37, ifo Institute - Leibniz Institute for Economic Research at the University of Munich 2022
Hong Gee Hee Lucas Deborah COVID Credit Policies around the World: Size, Scope, Costs and Consequences Brookings Papers on Economic Activity 2023 presented March 30-31, 2023
Ilzetzki Ethan Reinhart Carmen Rogoff Kenneth Exchange Arrangements Entering the Twenty-first Century: Which Anchor will Hold? The Quarterly Journal of Economics 134 2 2019 599 646
International Monetary Fund. (2020a). “Fiscal Monitor October 2020: Policies for the Recovery.” Link
International Monetary Fund Toward an Integrated Policy Framework IMF Policy Paper No. 2020/046. Link 2020
Jordà Òscar Nechio Fernanda The Inflation-Wage Feedback after the Pandemic: International Evidence Federal Reserve Bank of San Francisco Working Paper 2022-17 2022
Jordà Òscar Schularick Moritz Taylor Alan Sovereigns versus Banks: Credit, Crises, and Consequences Journal of the European Economic Association 14 1 2016 45 79
Kirti, Divya, Yang Liu, Soledad Martinez Peria, Prachi Mishra, and Jan Strasky. (2022a). “Tracking Economic and Financial Policies during COVID-19: An Announcement-Level Database.” IMF Working Paper No. 22/114.
Kirti, Divya, Soledad Martinez Peria, Prachi Mishra, and Jan Strasky. (2022b). “What Policy Combinations Worked? The Effect of Policy Packages on Bank Lending during COVID-19.” IMF Working Paper, forthcoming.
Kose M.Ayhan Kurlat Sergio Ohnsorge Franziska Sugawara Naotaka A Cross-Country Database of Fiscal Space Journal of International Money and Finance 128 2012 102682
Krippner, Leo. (2015). “Zero Lower Bound Term Structure Modeling: A Practitioner's Guide.” Palgrave-Macmillan.
Mano Rui Sgherri Silvia One Shock, Many Policy Responses IMF Working Paper 20/010 2020
Maravalle Alessandro Rawdanowicz Lukasz How effective are automatic stabilizers in OECD Countries? OECD Economics Department Working Paper No. 1635 2020
McKay Alisdair Reis Ricardo Optimal Automatic Stabilizers Review of Economic Studies 88 2021 2375 2406
Orphanides Athanasios Wieland Volker Efficient Monetary Policy Design near Price Stability Journal of the Japanese and International Economies 14 4 2000 327 365
Ramey Valerie. Ten Years after the Financial Crisis: What Have We Learned from the Renaissance in Fiscal Research? Journal of Economic Perspectives 33 2 2019 89 114
Romer Christina Romer David Fiscal Space and the Aftermath of Financial Crises: How it Matters and Why Brookings Papers on Econ. Activity 2019 239 331 Spring, pgs
Romer Christina Romer David Why Some Times Are Different: Macroeconomic Policy and the Aftermath of Financial Crises Economica 85 337 2018 1 40
www.macromodelbase.com., Volker. (2022). Overview of How Major Economies Have Responded to the Covid-19 Pandemic. Study requested by the ECON committee of the European Parliament. Available at: https://www.europarl.europa.eu/RegData/etudes/STUD/2022/689450/IPOL_STU(2022)689450_EN.pdf
Wieland, Volker, Elena Afanasyeva, Jinhyuk Yoo and Meguy Kuete. (2016). New Methods for Macro-Financial Model Comparison and Policy Analysis. In John Taylor and Harald Uhlig(eds.), Handbook of Macroeconomics: Vol 2. North Holland/Elsevier, chp. 15.
Wieland Volker Cwik Tobias Müller Gernot Sebastian Schmidt and Maik Wolters. A New Comparative Approach to Macroeconomic Modeling and Policy Analysis Journal of Economic Behavior & Organization 83 3 2012 523 541
Woodford Michael. Simple Analytics of the Government Expenditure Multiplier American Economic Journal: Macroeconomics 3 1 2011 1 35
|
PMC010xxxxxx/PMC10238249.txt |
==== Front
Arch Virol
Arch Virol
Archives of Virology
0304-8608
1432-8798
Springer Vienna Vienna
37269384
5803
10.1007/s00705-023-05803-9
Original Article
Increase in rotavirus prevalence with the emergence of genotype G9P[8] in replacement of genotype G12P[6] in Sabah, Malaysia
Amit Lia Natasha 1
John Jecelyn Leaslie 2
Mori Daisuke 1
Chin Abraham Zefong 3
Mosiun Andau Konodan 4
http://orcid.org/0000-0002-1869-3701
Ahmed Kamruddin ahmed@ums.edu.my
12
1 grid.265727.3 0000 0001 0417 0814 Department of Pathology and Microbiology, Faculty of Medicine and Health Sciences, Universiti Malaysia Sabah, Kota Kinabalu, Sabah Malaysia
2 grid.265727.3 0000 0001 0417 0814 Borneo Medical and Health Research Centre, Faculty of Medicine and Health Sciences, Universiti Malaysia Sabah, Kota Kinabalu, Sabah Malaysia
3 grid.265727.3 0000 0001 0417 0814 Department of Community and Family Medicine, Faculty of Medicine and Health Sciences, Universiti Malaysia Sabah, Kota Kinabalu, Sabah Malaysia
4 grid.415759.b 0000 0001 0690 5255 Kunak District Health Office, Ministry of Health Malaysia, Kunak, Sabah Malaysia
Handling Editor: Tim Skern.
3 6 2023
2023
168 6 17330 1 2023
19 4 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Rotaviruses are major causative agents of acute diarrhea in children under 5 years of age in Malaysia. However, a rotavirus vaccine has not been included in the national vaccination program. To date, only two studies have been carried out in the state of Sabah, Malaysia, although children in this state are at risk of diarrheal diseases. Previous studies showed that 16%–17% of cases of diarrhea were caused by rotaviruses and that equine-like G3 rotavirus strains are predominant. Because the prevalence of rotaviruses and their genotype distribution vary over time, this study was conducted at four government healthcare facilities from September 2019 through February 2020. Our study revealed that the proportion of rotavirus diarrhea increased significantly to 37.2% (51/137) after the emergence of the G9P[8] genotype in replacement of the G12P[8] genotype. Although equine-like G3P[8] strains remain the predominant rotaviruses circulating among children, the Sabahan G9P[8] strain belonged to lineage VI and was phylogenetically related to strains from other countries. A comparison of the Sabahan G9 strains with the G9 vaccine strains used in the RotaSiil and Rotavac vaccines revealed several mismatches in neutralizing epitopes, indicating that these vaccines might not be effective in Sabahan children. However, a vaccine trial may be necessary to understand the precise effects of vaccination.
http://dx.doi.org/10.13039/501100022424 Ministry of Scientific and Technological Development, Higher Education and Information Society FRGS/1/2017/SKK11/UMS/01/2 Ahmed Kamruddin issue-copyright-statement© Springer-Verlag GmbH Austria, part of Springer Nature 2023
==== Body
pmcIntroduction
Globally, rotaviruses continue to be the major causative agents of diarrhea in children under 5 years of age [1], with > 95% of the mortality associated with this condition occurring in sub-Saharan Africa, South Asia, and Southeast Asia [2–5]. Rotaviruses accounted for 40% of all diarrhoea cases in Southeast Asia from 1990 to 2017 [1]. In Malaysia, the burden of rotaviruses remains unclear, as there is a considerable gap in surveillance studies, with the exception of the 2002–2010 period [6]. This is notable for Sarawak and Sabah, which are two states of Malaysian Borneo where indigenous people of different ethnicities reside. Although, indigenous children in Malaysian Borneo are more vulnerable to acute gastroenteritis than those in other localities [7] in Sabah, only two studies have been carried out, during 2005–2006 and 2018–2019, which showed that rotaviruses are responsible for 16%–17% of all diarrhea cases [8, 9].
Rotaviruses are double-stranded RNA viruses whose genome is divided into 11 segments. Gene segments 1, 2, 3, 4, 6, and 9 encode the structural proteins VP1, VP2, VP3, VP4, VP6, and VP7, respectively, whereas gene segments 5, 8, 7, 10, and 11 encode the non-structural proteins NSP1, NSP2, NSP3, NSP4, and NSP5/6, respectively [10]. Trypsin treatment results in the specific cleavage of VP4 to form VP8* and VP5*, representing the amino- and carboxyl-terminal regions of the protein, respectively. The viral genome is enclosed by an icosahedral capsid consisting of an inner core, an intermediate capsid, and an outer capsid. The VP4 and VP7 proteins are the main components of the outer capsid and contain epitopes that induce neutralizing antibodies and define the P and G genotypes, respectively, of the dual nomenclature. To date, 42 G and 58 P genotypes have been discovered in humans and animals [https://rega.kuleuven.be/cev/viralmetagenomics/virus-classification/newgenotypes (accessed October 12, 2022)]. Due to the segmented nature of their genome, rotaviruses can be characterized according to electropherotype patterns based on the differences in the relative migration rates of genomic segments in polyacrylamide gel electrophoresis (PAGE), and this is useful for detection of strain diversification. Moreover, also because of the segmented nature of the rotavirus genome, gene reassortment occurs between human and animal rotaviruses, resulting in rotavirus strain diversity [11]. Point mutations in genes encoding outer capsid proteins are another evolutionary mechanism that can be used to classify rotaviruses into lineages, and some of these result in the emergence of antibody-escape mutants [12].
Four oral live-attenuated commercial vaccines, i.e., Rotarix, RotaTeq, Rotavac, and RotaSiil, have been prequalified by the World Health Organization. Rotarix (GlaxoSmithKline Biologicals SA, Rixensart, Belgium) is a monovalent vaccine consisting of the G1P[8] genotype, whereas RotaTeq (Merck & Co., Inc., West Point, PA, USA) is a pentavalent vaccine that includes the G1–G4 and P[8] genotypes. India has introduced its own vaccines, i.e., Rotavac (Bharat Biotech Telangana, India), a monovalent vaccine carrying the G9P[11] genotype, and RotaSiil (Serum Institute of India, Pune, India), which is a pentavalent vaccine containing the G1, G2, G3, G4, and G9 genotypes [13]. Despite a decrease in the rotavirus burden since the implementation of vaccination, most of the developing countries in Asia do not include rotavirus vaccination in their national vaccination program [1]. Malaysia is not an exception compared with other Asian countries, but Rotarix and RotaTeq are available in the private sector [9].
The predominant rotavirus types worldwide affecting human populations are G1P[8], G3P[8], G4P[8], G9P[8], and G2P[4], which have been identified in Malaysia, albeit with varying prevalence [6]. The predominant circulating genotypes change and vary according to location, which is very important, as vaccine development is based on the predominant circulating strains. Because the genotypes change, continued surveillance is important to evaluate the effectiveness of the rotavirus vaccines used in the national vaccination program. The genotype distribution of rotavirus in Sabah was first assessed in 2018–2019 [9]. In the present follow-up study (2019–2020), we detected an increase of rotavirus infection, which was associated with the emergence of G9 strains over a short time span.
Materials and methods
Ethical approval
Ethical approval was obtained from the Medical Research Ethics Committee, Ministry of Health, Malaysia, for Menggatal Health Clinic and Telipok Health Clinic (NMRR-16-2245-32787), Sabah Women and Children Hospital (NMRR-19-3925-52370), and Kunak District Hospital (NMRR-20-1324-55178). All procedures were performed in accordance with relevant guidelines and regulations. Informed consent was obtained from the legal guardians of the children participating in this study.
Collection of stool samples and patients’ information
The study period was from September 2019 through February 2020. Stool samples were collected from children aged ≤5 years who attended the above-mentioned healthcare facilities with watery diarrhea with or without vomiting and fever. Watery diarrhea was defined as passing watery stools at least three times during a 24-h period. Patient information such as gender, age in months, and ethnicity was recorded. Stool samples were collected at healthcare facilities in stool-collection bottles, diluted in phosphate-buffered saline (PBS) to make 10% suspensions, transported under cold-chain conditions to Universiti Malaysia Sabah, and stored at -80°C until use.
Rotavirus detection and genomic extraction of RNA
Rotavirus was detected using a commercial enzyme immunoassay kit according to the manufacturer’s instructions (Rotaclone, Meridien Diagnostics Inc., Cincinnati, USA). Rotavirus genomic RNA was extracted from rotavirus-positive samples using a QIAamp Viral RNA Mini Kit according to the manufacturer’s instructions (QIAGEN GmbH, Hilden, Germany). For electropherotyping, genomic RNA was extracted using the phenol:chloroform:isoamyl alcohol method [9].
Determination of G and P genotypes by RT-PCR
For rotavirus VP7 and VP4 gene amplification, reverse transcription (RT-PCR) was performed using an AccessQuick RT-PCR (Promega Corporation, Madison, WI, USA) according to the instructions [14]. The VP7 gene was amplified using the primers Beg9 and End9 [15], and the VP4 gene was amplified using the primers Con2 and Con3 [16]. For G and P genotyping, multiplex PCR was performed using a master mix (Promega Corporation, Madison, WI, USA). Primers for the detection of the G1, G2, G3, G4, G8, G9, G10, and G12 genotypes were used as described previously [15, 18]. For P genotyping, primers for the detection of P[4], P[6], P[8], P[9], P[10], and P[11] were used [17, 18]. The nucleotide sequences of the VP7 (1,062 bp) and VP8* (876 bp) regions of the VP4 gene were determined using a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). The sequencing products were analyzed using an ABI Prism 3100 Genetic Analyzer (Applied Biosystems).
Phylogenetic analysis
The VP7 and VP8* regions of the VP4 gene were sequenced as described previously [14]. Similar sequences were identified using BLAST (www.ncbi.nlm.nih.gov/blast), multiple sequence alignment was performed using ClustalW, and phylogenetic trees were constructed by the maximum-likelihood method using MEGA X software [19]. Bootstrap analysis of 1,000 replicates was used to investigate the significance of the branching of the constructed trees. The nucleotide sequence identity values for individual genes were calculated using online software (www.bioinformatics.org).
Determination of electropherotypes
Electropherotypes were determined by polyacrylamide gel electrophoresis (PAGE) of the extracted dsRNA of rotavirus-positive samples [9]. In brief, 5 µl of extracted dsRNA was mixed with 5 µl of loading buffer, loaded onto a 10% polyacrylamide gel, and subjected to electrophoresis for 16 h at a constant current of 8 mA. The gel was then stained with ethidium bromide to visualize the migration pattern of the genomic RNA. Electropherotype numbers were assigned arbitrarily based on differences in the migration patterns within at least one of the four groups of segments, i.e., segments 1 to 4, 5 and 6, 7 to 9, or 10 and 11.
Comparison of the VP7 antigenic epitopes of circulating rotaviruses with those of vaccine strains
To compare the antigenic epitopes on the VP7 protein of the G9 rotaviruses circulating in Sabah with those of the Rotavac (G9P[11]) and RotaSiil (G9P[5]) vaccine strains, multiple sequence alignments of the amino acid sequences of VP7 were made using the ClustalW plug-in in MEGA X software [19].
Statistical analysis
To assess whether children with G9 vs. those with non-G9 genotypes differed regarding age at presentation, an independent two-tailed t-test was performed, using Social Science Statistics software (https://www.socscistatistics.com/tests/studentttest/default.aspx). A P-value less than 0.05 was considered statistically significant.
Results
Rotavirus prevalence and age distribution
A total of 137 watery stool samples were collected from Sabah Women and Children Hospital (78.1%; 107/137), Kunak District Hospital (13.9%; 19/137), Menggatal Health Clinic (5.8%; 8/137), and Telipok Health Clinic (2.2%; 3/137). The male-to-female ratio was 1.4:1 (80 males, 57 females). Children aged 12–23 months had the highest incidence of acute diarrhea (29.9%; 41/137), followed by children aged 6–11 months (28.5%; 39/137). The median age of patients with acute diarrhea was 15 months (range, 1–64 months).
Fifty-one (37.2%) of the 137 samples were positive for rotavirus. Most infections were reported at Sabah Women and Children Hospital (84.3%; 43/51), followed by Kunak District Hospital (9.8%; 5/51) and Menggatal Health Clinic (5.9%; 3/51). The male-to-female ratio was 1.4:1 (30 males, 21 females). Children aged 12–23 months had the highest incidence of rotavirus infection (33.3%; 17/51).
Genotype distribution
Of the 51 rotavirus-positive samples, 42 (82.3%) underwent successful G and P genotyping. The predominant G genotype was G3 (n = 21), followed by G9 (n = 14), G1 (n = 3), G2 (n = 1), and an undetermined Gx genotype (n = 3). “Undetermined” indicates that no primary PCR product was available. The P[8] genotype was detected in 38 rotavirus-positive samples. In four samples, the P genotype could not be determined, and these were designated P[x]. Based on the combination of the G and P genotypes, the predominant combined genotype was G3P[8] (n = 19; 45.2%), followed by G9P[8] (n = 14; 33.3%), G1P[8] (n = 2; 4.8%), GxP[8] (n = 3; 7.1%), G3P[x] (n = 2; 4.8%), G1P[x] (n = 1; 2.4%), and G2P[x] (n = 1; 2.4%) (Table 1).Table 1 The frequency of distribution of rotavirus genotypes identified in the stool of patients with rotavirus diarrhoea
Genotype Number (percentage)
G3P[8] 19 (45.2%)
G9P[8] 14 (33.3%)
G1P[8] 2 (4.8%)
GxP[8] 3 (7.1%)
G3P[x] 2 (4.8%)
G1P[x] 1 (2.4%)
G2P[x] 1 (2.4%)
Total 42 (100%)
The age difference between the children who were found to have the G9 genotype (14 children; mean age = 22.4 months) and those with a non-G9 genotype (26 children; mean age = 16.9 months) was not statistically significant (P = 0.111981).
Electropherotypes
In electropherotype analysis, all 11 segments were visible in 30 out of 51 (58.8%) samples tested. One short and four long electropherotype patterns were identified (Fig. 1). The short electropherotype S1 was found in 13 isolates. Among the long electropherotypes, the L1 group included 13 isolates, the L2 and L4 groups each included one isolate, and the L3 group included two isolates. The G3P[8] genotype isolates included both short (S1) and long (L2) electropherotypes. All G9P[8]-positive samples exhibited long electropherotype patterns (L1). The G genotype could not be determined for the L3 and L4 electropherotypes.Fig. 1 Electropherotypes of rotaviruses identified in Sabah. In total, four electropherotypes were identified, four long (L1–L4) and one short (S1) electropherotype patterns were identified. L1 and L2 were detected in genotype G9 and G3 strains, respectively. L3 and L4 were detected in untypable strains. S1 was identified in G3 strains. Images of the original gels are presented in Supplementary Figure S1
Phylogenetic analysis
The VP7 and the VP8* portion of the VP4 gene could be sequenced in 36 and 34 isolates, respectively. Phylogenetic analysis based VP7 genes of the G1 genotype, Sabahan G1P[8] belonged to lineage Ia [20] and clustered with a significant bootstrap value with strains from Sabah detected in a previous study [6] and with strains from Indonesia detected in 2017 and 2018 (Fig. 2). It is worth mentioning that most of the G1 strains detected in the previous study from Sabah belonged to lineage II. The G2 isolate belonged to lineage VI and was related to strains from Italy, Russia, Bangladesh, Australia, and Vietnam from 2005 to 2011 (Fig. 3).Fig. 2 Phylogenetic tree based on nucleotide sequences of the VP7 gene of G1 strains. The human rotavirus strain L116 (genotype G3) was used as an outgroup. Numbers at nodes represent the bootstrap value, and values higher than 70% are shown. The scale bar shows the genetic distance, expressed as nucleotide substitutions per site. Strains from the present study (indicated by a filled circle) belong to lineage Ic and are clustered with strains detected in Sabah in a previous study. The GenBank accession number is shown before each strain name
Fig. 3 Phylogenetic tree based on the VP7 gene of G2 strains. The human rotavirus strain L116 (genotype G3) was used as an outgroup. Numbers at nodes represent the bootstrap value, and values higher than 70% are shown. The scale bar shows the genetic distance, expressed as nucleotide substitutions per site. The strain identified in this study (indicated by a filled circle) belongs to lineage IV. The GenBank accession number is shown before each strain name
Four G3P[8] strains belonged to lineage III and formed an independent cluster with significant bootstrap values (Fig. 4). These strains were related to strains from China and those detected previously in Sabah. The current Sabahan strains shared 99.1–99.5% nucleotide sequence identity with each other and shared 96.7–98.3% nucleotide sequence identity with previously detected Sabahan G3P[8] strains. An additional 15 G3P[8] strains belonged to a cluster formed by the previously detected equine-like G3P[8] from Sabah and belonged to lineage I. The current Sabahan strains shared 97.5–100% nucleotide sequence identity with each other and shared 96.6–99.8% nucleotide sequence identity with previously detected Sabahan equine-like G3P[8] strains.Fig. 4 Phylogenetic tree based on the VP7 gene of G3 strains. The human rotavirus strain Wa (genotype G1) was used as an outgroup. Numbers at nodes represent the bootstrap value, and values higher than 70% are shown. The scale bar shows the genetic distance, expressed as nucleotide substitutions per site. The strains identified in this study belong to lineages I and III (indicated by a filled circle). The GenBank accession number is shown before each strain name
All G9P[8] strains identified in this study belonged to lineage VI and clustered together with strains from China (Hubei, Liaoning, and Shandong), Japan, Russia, Vietnam, Thailand, South Africa, and the USA (2013–2019) (Fig. 5).Fig. 5 Phylogenetic tree based on the VP7 gene of G9 strains. The human rotavirus strain DS-1 (genotype G2) was used as an outgroup. Numbers at nodes represent the bootstrap value, and values higher than 70% are shown. The scale bar shows the genetic distance, expressed as nucleotide substitutions per site. Strains from the present study (indicated by a filled circle) belong to lineage VI. The GenBank accession number is shown before each strain name
The VP4 genes of all Sabahan G3P[8], G1P[8], GxP[8], and G9P[8], strains belonged to lineage III and formed three clusters in the phylogenetic tree (Fig. 6). The VP4 gene of seven equine-like G3P[8], two GxP[8], one G1P[8], and all G9P[8] strains clustered with equine-like G3 strains from our previous study, with a significant bootstrap value [9]. They belonged to sublineage III.1. Another cluster contained the VP4 genes of two equine-like G3 strains and one G1[P8] strain from the present study and two G1P[8] strains from our previous study [9]. In the third cluster, the VP4 genes of three G3 and one equine-like G3 strains grouped together. The last two clusters belonged to sublineage III.4.Fig. 6 Phylogenetic tree based on the VP4 gene of P[8] strains. The human rotavirus strain 366 (genotype P[4]) was used as an outgroup. Numbers at nodes represent the bootstrap value, and values higher than 70% are shown. The scale bar shows the genetic distance expressed as nucleotide substitutions per site. All strains from the present study (indicated by a filled circle) belong to lineage III (sublineages III.1 and III.4). The GenBank accession number is shown before each strain name
Comparison of VP7 antigenic epitopes of Sabahan G9 strains with Rotavac and RotaSiil vaccine strains
We compared the amino acid sequences of the VP7 proteins of Sabahan G9 strains with those of the Rotavac and RotaSiil vaccine strains to identify possible epitopes of vaccine-escape strains (Fig. 7). Among the 29 amino acid residues comprising the VP7 antigenic epitopes, four residues differed from Rotavac, and three differed from RotaSiil [21]. In comparison to Rotavac, two substitutions (I87T and G100N) occurred in the 7-1a epitope, whereas no substitutions were detected in the 7-1b epitope. Moreover, two substitutions (N145D and N221S) were detected in the 7-2 epitope. In comparison to RotaSiil, two substitutions (A87T and D100N) occurred in the 7-1a epitope, one substitution (T242N) occurred in the 7-1b epitope, and no substitutions were detected in the 7-2 epitope.Fig. 7 Amino acid sequence alignment of the VP7 proteins of the vaccine strains Rotavac and RotaSiil with Sabahan G9 strains reported in this study. Antigenic epitopes are indicated above the residue numbers. Residues that differed from Rotavac are highlighted in green. Residues that differed from RotaSiil are highlighted in blue. Residues that differed from both Rotovac and RotaSiil are highlighted in yellow. Amino acid substitutions that have been shown to be associated with escape from neutralization by monoclonal antibodies are indicated by an asterisk (*)
Discussion
Fluctuations in the distribution of the predominant rotavirus genotypes have been reported worldwide [18, 22–26]. These fluctuations may be driven by herd immunity, and the emergence of a particular genotype may reflect the presence of a sufficiently large population of susceptible children [22]. One of the significant findings of this study was that, within 5 months of the end of the previous study [9], the prevalence of rotavirus infection had increased by about 256%. In this study, we found that the incidence of rotavirus infection was highest in children 12–23 months of age, as has been reported previously [27–29]. However, in contrast to other studies [30, 31] in which older children were more likely to be infected with G9 strains, in our study, the average age of Sabahan children infected with G9 strains was not significantly different from that of children infected with more-common strains. It is therefore likely that these children had no immunity against the newly circulating G9 strains and, when exposed, were infected.
In the present study, G3P[8] was the predominant circulating genotype among Sabahan children, followed by G9, G1, and G2. Equine-like G3P[8] possibly adapted well to the child population of Sabah, and this is supported by phylogenetic analysis showing clustering of previous and current strains. A similar observation was made for G1P[8], although this is not a predominant strain in Sabah; in contrast, previous strains adapted well and continued to infect children, as found in the present study. The most noteworthy finding was the emergence of G9 strains replacing the G12 strain that was in circulation previously (Fig. 8); moreover, G9 strains might be responsible for the abrupt rise in rotavirus infections in children compared with the previous study [9].Fig. 8 Rotavirus G and P genotypes found in a previous study by Amit et al., 2021 [9] (a) and in the current study (b). In both studies, G3P[8] was the predominant genotype. The present study shows the emergence of G9 strains replacing the G12 strain that was in circulation previously
Several publications have reported the emergence of the G9 genotype in other countries during approximately the same period. A study performed in 2018 in the West Nusa Tenggara region of neighboring Indonesia detected the presence of G9P[8] strains; however, these strains were absent in South Sumatra and West Papua [32], indicating a localized emergence of this genotype. Furthermore, a study carried out in 2017–2018 in China showed G9P[8] to be the predominant strain [33]. In Argentina, G9P[4] was in circulation in 2017, but not in 2018 [34]. In Thailand, G9P[8] was predominant during 2018 but decreased significantly in 2019, when G3 became predominant. Phylogenetically, Sabahan G9 was more closely related to strains circulating in three provinces of China than to those in neighboring Indonesia or Thailand, suggesting that transmission is occurring between China and Sabah state, perhaps because this state is a popular destination for Chinese tourists.
G9 was first identified in 1987 in the USA [35], and it continued to compete with other genotypes until it become the fifth globally predominant rotavirus [36]. Several studies performed in other countries have shown that G9 initially had a low prevalence but then rapidly increased to become the predominant genotype [37–39]. In Malaysia, G9P[8] rotavirus was detected in the early 2000s in Johor, Kuala Lumpur [40], and Sarawak [41, 42]. In Sarawak, rotavirus surveillance studies were conducted from 2001 to 2007, with G9[P8] being the predominant genotype from 2001 to 2003 and G1P[8] being predominant in 2001 and 2007 [41, 42]. The nucleotide sequences of these G9 strains from Malaysia are not available in public databases; therefore, they could not be compared with the G9 strains detected in the present study. Furthermore, for the last 10 years, except in one study [9], no genotyping of rotavirus strains has been performed in Malaysia; therefore, the dynamics of G9 distribution remain unknown. Thus, a regular surveillance system should be established to track the genotype distribution, which will be helpful for planning vaccination policy. Although the VP7 and VP4 genes of Sabahan G9 strains clustered together phylogenetically, these strains harbored four different long electropherotype patterns, indicating diversity within the G9 genotype.
The most prevalent G9 lineages worldwide are lineages III and VI [43]. Geographically, lineage VI is widespread in Asia and was identified in 2011–2019 in Tokyo (2017–2018) [44], Vietnam (2016–2018) [45], Wuhan (2019) [46], and Beijing (2011–2013). Lineages I, II, IV, and V were detected only in samples from humans, whereas lineages III and VI are epidemiologically linked to porcine and human samples [47]. Because all Sabahan G9P[8] strains belonged to lineage VI, additional studies using whole-genome sequence analysis are required to obtain a full picture of these strains regarding their origin, evolution, and reassortment. The only G2 strain detected was phylogenetically related to strains circulating in 2005–2011 in Italy, Russia, Bangladesh, Australia, and Vietnam. Why such an old strain was circulating in Sabah is puzzling and warrants further investigation. Our results indicate that the dynamics of rotavirus genotype changes in Sabah were very rapid. Other studies have shown that fluctuations in the rotavirus genotype distribution occurred continuously over time and according to location.
To shed light on vaccine effectiveness, we compared the antigenic epitopes of our G9 strains with those of the Rotavac and RotaSiil vaccine strains and identified several substitutions. There were two defined antigenic epitopes, 7-1 (7-1a and 7-1b) and 7-2 in VP7, where amino acid substitution could reduce the effectiveness of rotavirus vaccines. Epitopes 7-1a, 7-1b, and 7-2 include the earlier-described antigenic regions A (87–101) and D (291); C (208–221), E (189–190), and F (233–242); and B (142–152), respectively [48]. In this study, we found that the Rotavac and RotaSiil vaccines might have reduced effectiveness against Sabahan G9 rotaviruses, because all of them exhibited substitutions at residues that are associated with neutralization antibody escape (positions 87, 100, 145, 221 and 242) [48]. The substitution N145D results in the loss of a glycosylation site at amino acid residues 145–147 within the 7-2 epitope, making the virus resistant to neutralization [48]. The substitution N221S in this epitope can increase resistance to neutralization by tenfold [49]. The Rotavac and RotaSiil vaccines are currently in use only in India and Palestine (Rotavac only) [50].
The impact of amino acid substitutions within antigenic epitopes on vaccine effectiveness is difficult to predict, as the effectiveness of vaccines in a region can be influenced by multiple factors, ranging from concurrent enteric infections, malnutrition, immune status, healthcare access, and the vaccine coverage rate within the population [51]. Therefore, a vaccine trial may be necessary to evaluate the effectiveness of vaccination. A limitation of this study was that we had to discontinue the study earlier than scheduled because of the introduction of a movement control order (MCO) by Malaysian authorities as of March 2020 to control the spread of COVID-19. Moreover, the hospitals that participated in our study were then designated as COVID-19 hospitals, and sample collection was no longer possible. Nevertheless, the results of this study can be used to evaluate the after-effects of the MCO on rotavirus infections, considering that significant decreases in rotavirus infection were reported in Japan [52], Germany [53], and Korea [54] during the COVID-19 pandemic.
Conclusion
In this study, we found that the emergence of four different electropherotype patterns of genotype G9 in replacement of G12 was associated with a higher prevalence of rotavirus diarrhoea in children in Sabah, Malaysian Borneo; however, equine-like G3P[8] rotavirus continued to be the predominant strain in circulation.
Acknowledgements
The authors would like to thank the Director General of Health Malaysia for permission to publish this article.
Author contributions
KA: conceptualized the study, analyzed data, acquired funds, performed project administration, supervised, reviewed, edited, and wrote the final manuscript. LNA wrote the original draft, curated data, analyzed data, performed investigations, performed laboratory procedures, reviewed, edited, and wrote the final manuscript. JLJ curated data, analyzed data, performed investigations, performed laboratory procedures, reviewed, edited, and wrote the final manuscript. DM analyzed data, performed investigations, performed laboratory procedures, supervised, reviewed, edited, and wrote the final manuscript. AZC analyzed data, performed investigations, performed project administration, reviewed, edited, and wrote the final manuscript. AKM analyzed data, performed investigations, performed project administration, reviewed, edited, and wrote the final manuscript.
Funding
This study was funded by the Ministry of High Education under the Fundamental Research Grant Scheme (FRGS/1/2017/SKK11/UMS/01/2) (URL: https://mygrants.gov.my) to KA.
Data availability
The datasets generated and/or analysed during the current study are available in the DNA Data Bank of Japan/European Molecular Biology Laboratory/GenBank repository: OM928419, OM928420, OM928421, OM928422, OM928423, OM928424, OM928425, OM928426, OM928427, OM928428, OM928429, OM928430, OM928431, OM928432, OM928433, OM928434, OM928435, OM928436, OM928437, OM928438, OM928439, OM928440, OM928441, OM928442, OM928443, OM928444, OM928445, OM928446, OM928447, OM928448, OM928449, OM928450, OM928451, OM928452, OM928453, OM928454, OM928455, OM928456, OM928457, OM928458, OM928459, OM928460, OM928461, OM928462, OM928463, OM928464, OM928465, OM928466,OM928467, OM928468, OM928469, OM928470, OM928471, OM928472, ON922858, ON922858, ON922859, ON922860, ON922861, ON922862, ON922863, ON922864, ON922865, ON922866, ON922867, OP004804, OP004805, OP004806, OP004807, OP004808, OP004809.
Declarations
Conflict of interest
All authors declare that they have no conflict of interest.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Lestari FB Vongpunsawad S Wanlapakorn N Poovorawan Y Rotavirus infection in children in Southeast Asia 2008–2018: disease burden, genotype distribution, seasonality, and vaccination J Biomed Sci 2020 27 1 66 10.1186/s12929-020-00649-8 32438911
2. Troeger C Khalil IA Rao PC Cao S Blacker BF Ahmed T Rotavirus vaccination and the global burden of rotavirus diarrhea among children younger than 5 years JAMA Pediatr 2018 172 10 958 965 10.1001/jamapediatrics.2018.1960 30105384
3. Omatola CA Olaniran AO Rotaviruses: from pathogenesis to disease control-a critical review Viruses 2022 14 5 875 10.3390/v14050875 35632617
4. Burnett E Parashar UD Tate JE Rotavirus infection, illness, and vaccine performance in malnourished children: a review of the literature Pediatr Infect Dis J 2021 40 10 930 936 10.1097/INF.0000000000003206 34117200
5. Varghese T Kang G Steele AD Understanding rotavirus vaccine efficacy and effectiveness in countries with high child mortality Vaccines (Basel) 2022 10 3 346 10.3390/vaccines10030346 35334978
6. Lee WS Lim BT Chai PF Kirkwood CD Lee JK Rotavirus genotypes in Malaysia and universal rotavirus vaccination Hum Vaccin Immunother 2012 8 10 1401 1406 10.4161/hv.21577 23022710
7. Hsu VP Abdul Rahman HB Wong SL Ibrahim LH Yusoff AF Chan LG Estimates of the burden of rotavirus disease in Malaysia J Infect Dis 2005 192 Suppl 1 S80 S86 10.1086/431494 16088810
8. Goh CT Cheah PK Soo TL Lee WS The epidemiology and burden of childhood rotavirus infection in a tertiary hospital in Sabah, Malaysia Med J Malaysia 2009 64 2 146 149 20058575
9. Amit LN Mori D John JL Chin AZ Mosiun AK Jeffree MS Emergence of equine-like G3 strains as the dominant rotavirus among children under five with diarrhea in Sabah, Malaysia during 2018–2019 PLoS ONE 2021 16 7 e0254784 10.1371/journal.pone.0254784 34320003
10. Desselberger U Rotaviruses Virus Res 2014 190 75 96 10.1016/j.virusres.2014.06.016 25016036
11. Yahiro T Takaki M Chandrasena TGAN Rajindrajith S Iha H Ahmed K Human-porcine reassortant rotavirus generated by multiple reassortment events in a Sri Lankan child with diarrhea Infect Genet Evol 2018 65 170 186 10.1016/j.meegid.2018.07.014 30055329
12. Luchs A TimenetskyMdo C Group A rotavirus gastroenteritis: post-vaccine era, genotypes and zoonotic transmission Einstein (Sao Paulo) 2016 14 2 278 287 10.1590/S1679-45082016RB3582 27462899
13. Skansberg A Sauer M Tan M Santosham M Jennings MC Product review of the rotavirus vaccines ROTASIIL, ROTAVAC, and Rotavin-M1 Hum Vaccin Immunother 2021 17 4 1223 1234 10.1080/21645515.2020.1804245 33121329
14. Mitui MT Chandrasena TN Chan PK Rajindrajith S Nelson EA Leung TF Inaccurate identification of rotavirus genotype G9 as genotype G3 strains due to primer mismatch Virol J 2012 9 144 10.1186/1743-422X-9-144 22862784
15. Gouvea V Glass RI Woods P Taniguchi K Clark HF Forrester B Polymerase chain reaction amplification and typing of rotavirus nucleic acid from stool specimens J Clin Microbiol 1990 28 2 276 282 10.1128/jcm.28.2.276-282.1990 2155916
16. Gentsch JR Glass RI Woods P Gouvea V Gorziglia M Flores J Identification of group A rotavirus gene 4 types by polymerase chain reaction J Clin Microbiol 1992 30 6 1365 1373 10.1128/jcm.30.6.1365-1373.1992 1320625
17. Gunasena S Nakagomi O Isegawa Y Kaga E Nakagomi T Steele AD Relative frequency of VP4 gene alleles among human rotaviruses recovered over a 10-year period (1982–1991) from Japanese children with diarrhea J Clin Microbiol 1993 31 8 2195 2197 10.1128/jcm.31.8.2195-2197.1993 8396591
18. Reju S Srikanth P Selvarajan S Thomas RK Barani R Amboiram P A shift in circulating rotaviral genotypes among hospitalized neonates Sci Rep 2022 12 1 2842 10.1038/s41598-022-06506-y 35181717
19. Kumar S Stecher G Li M Knyaz C Tamura K MEGA X: molecular evolutionary genetics analysis across computing platforms Mol Biol Evol 2018 35 6 1547 1549 10.1093/molbev/msy096 29722887
20. Pradhan GN Chitambar SD Full genomic analysis of G1p[8] rotavirus strains recovered from rotavirus vaccinated and non-vaccinated children hospitalized for acute gastroenteritis in Pune, Western India J Med Virol 2018 90 4 772 778 10.1002/jmv.25007 29244210
21. Raju B Parikh RP Vetter VV Kolhapure S Epidemiology of rotavirus gastroenteritis and need of high rotavirus vaccine coverage with early completion of vaccination schedule for protection against rotavirus diarrhea in India: a narrative review Indian J Public Health 2019 63 3 243 10.4103/ijph.ijph_307_18 31552856
22. Mladenova Z Korsun N Geonova T Iturriza-Gómara M Rotavirus Study Group Molecular epidemiology of rotaviruses in Bulgaria: annual shift of the predominant genotype Eur J Clin Microbiol Infect Dis 2010 29 5 555 562 10.1007/s10096-010-0895-1 20217165
23. Manjate F João ED Chirinda P Garrine M Vubil D Nobela N Molecular epidemiology of rotavirus strains in symptomatic and asymptomatic children in Manhiça district, Southern Mozambique 2008–2019 Viruses 2022 14 1 134 10.3390/v14010134 35062336
24. Okitsu S Khamrin P Hikita T Thongprachum A Pham NTK Hoque SA Changing distribution of rotavirus A genotypes circulating in Japanese children with acute gastroenteritis in outpatient clinic, 2014–2020 J Infect Public Health 2022 15 7 816 825 10.1016/j.jiph.2022.06.009 35759807
25. Lu L Zhong H Jia R Su L Xu M Cao L Prevalence and genotypes distribution of group A rotavirus among outpatient children under 5 years with acute diarrhea in Shanghai, China, 2012–2018 BMC Gastroenterol 2022 22 1 217 10.1186/s12876-022-02288-9 35505284
26. Esona MD Gautam R Katz E Jaime J Ward ML Wikswo ME Comparative genomic analysis of genogroup 1 and genogroup 2 rotaviruses circulating in seven US cities, 2014–2016 Virus Evol 2021 7 1 veab023 10.1093/ve/veab023 34522389
27. Nirwati H Hakim MS Aminah S Dwija IBNP Pan Q Aman AT Identification of rotavirus strains causing diarrhoea in children under five years of age in Yogyakarta, Indonesia Malays J Med Sci 2017 24 2 68 77 28894406
28. Zaraket R Salami A Bahmad M El Roz A Khalaf B Ghssein G Prevalence, risk factors, and clinical characteristics of rotavirus and adenovirus among Lebanese hospitalized children with acute gastroenteritis Heliyon 2020 6 6 e04248 10.1016/j.heliyon.2020.e04248 32613122
29. Tian Y Chughtai AA Gao Z Yan H Chen Y Liu B Prevalence and genotypes of group A rotavirus among outpatient children under five years old with diarrhea in Beijing, China, 2011–2016 BMC Infect Dis 2018 18 1 497 10.1186/s12879-018-3411-3 30285635
30. Nordgren J Bonkoungou IJ Nitiema LW Sharma S Ouermi D Simpore J Rotavirus in diarrheal children in rural Burkina Faso: high prevalence of genotype G6P[6] Infect Genet Evol 2012 12 8 1892 1898 10.1016/j.meegid.2012.08.014 22964045
31. Cubitt WD Steele AD Iturriza M Characterisation of rotaviruses from children treated at a London hospital during 1996: emergence of strains G9P2A[6] and G3P2A[6] J Med Virol 2000 61 1 150 154 10.1002/(SICI)1096-9071(200005)61:1<150::AID-JMV24>3.0.CO;2-W 10745248
32. Wahyuni RM Utsumi T Dinana Z Yamani LN Juniastuti WIS Prevalence and distribution of rotavirus genotypes among children with acute gastroenteritis in areas other than Java Island, Indonesia, 2016–2018 Front Microbiol 2021 12 672837 10.3389/fmicb.2021.672837 34025628
33. Dong S Huang D Wang Z Zhang G Zhang F Sai L Clinical and molecular epidemiological characterization of rotavirus infections in children under five years old in Shandong province, China Arch Virol 2021 166 9 2479 2486 10.1007/s00705-021-05161-4 34218319
34. Degiuseppe JI Stupka JA Argentinean Rotavirus Surveillance Network Emergence of unusual rotavirus G9P[4] and G8P[8] strains during post vaccination surveillance in Argentina, 2017–2018 Infect Genet Evol 2021 93 104940 10.1016/j.meegid.2021.104940 34033946
35. Clark HF Hoshino Y Bell LM Groff J Hess G Bachman P Rotavirus isolate WI61 representing a presumptive new human serotype J Clin Microbiol 1987 25 1757 1762 10.1128/jcm.25.9.1757-1762.1987 2443534
36. Wu FT Bányai K Jiang B Liu LT Marton S Huang YC Novel G9 rotavirus strains co-circulate in children and pigs, Taiwan Sci Rep 2017 7 40731 10.1038/srep40731 28098174
37. Bozdayi G Altay A Yahiro T Ahmed S Meral M Dogan B Re-emergence of genotype G9 during a five-and-a-half-year period in Turkish children with rotavirus diarrhea Arch Virol 2016 161 10 2879 2884 10.1007/s00705-016-2986-5 27444180
38. Sashina TA Morozova OV Epifanova NV Novikova NA Predominance of new G9P[8] rotaviruses closely related to Turkish strains in Nizhny Novgorod (Russia) Arch Virol 2017 162 8 2387 2392 10.1007/s00705-017-3364-7 28429133
39. Kaplon J Grangier N Pillet S Minoui-Tran A Vabret A Wilhelm N Predominance of G9P[8] rotavirus strains throughout France, 2014–2017 Clin Microbiol Infect 2018 24 6 660.e1 660.e4 10.1016/j.cmi.2017.10.009 29066401
40. Zuridah H Kirkwood CD Bishop RF Bogdanovic-Sakran N Yap KL Molecular characterization and epidemiology of rotavirus isolates obtained from children with diarrhoea in Malaysia Med J Malaysia 2009 64 3 193 196 20527266
41. Hung LC Wong SL Chan LG Rosli R Ng AN Bresee JS Epidemiology and strain characterization of rotavirus diarrhea in Malaysia Int J Infect Dis 2006 10 6 470 474 10.1016/j.ijid.2006.05.008 17046306
42. Sum MS Perera D Ramji N Elie F Cardosa MJ Molecular characterization of rotavirus A associated with outbreaks of acute gastroenteritis in Sarawak in 2001 and 2007 Int J Infect Dis 2008 12 1 e91 e92 10.1016/j.ijid.2008.05.229
43. Phan TG Okitsu S Maneekarn N Ushijima H Genetic heterogeneity, evolution and recombination in emerging G9 rotaviruses Infect Genet Evol 2007 7 5 656 663 10.1016/j.meegid.2007.05.001 17544926
44. Fujii Y Oda M Somura Y Shinkai T Molecular characteristics of novel mono-reassortant G9P[8] rotavirus A strains possessing the NSP4 gene of the E2 genotype detected in Tokyo, Japan Jpn J Infect Dis 2020 73 1 26 35 10.7883/yoken.JJID.2019.211 31564695
45. Hoa-Tran TN Nakagomi T Vu HM Nguyen TTT Takemura T Hasebe F Detection of three independently-generated DS-1-like G9P[8] reassortant rotavirus A strains during the G9P[8] dominance in Vietnam, 2016–2018 Infect Genet Evol 2020 80 104194 10.1016/j.meegid.2020.104194 31931261
46. Zhou X Wang YH Pang BB Chen N Kobayashi N Surveillance of human rotavirus in Wuhan, China (2011–2019): predominance of G9P[8] and emergence of G12 Pathogens 2020 9 10 810 10.3390/pathogens9100810 33023203
47. Teodoroff TA Tsunemitsu H Okamoto K Katsuda K Kohmoto M Kawashima K Predominance of porcine rotavirus G9 in Japanese piglets with diarrhea: close relationship of their VP7 genes with those of recent human G9 strains J Clin Microbiol 2005 43 3 1377 1384 10.1128/JCM.43.3.1377-1384.2005 15750112
48. Gupta S Tiku VR Gauhar M Khatoon K Ray P Genetic diversity of G9 rotavirus strains circulating among diarrheic children in North India: a comparison with 116E rotavirus vaccine strain Vaccine 2021 39 646 651 10.1016/j.vaccine.2020.12.037 33386177
49. Dyall-Smith ML Lazdins I Tregear GW Holmes IH Location of the major antigenic sites involved in rotavirus serotype-specific neutralization Proc Natl Acad Sci USA 1986 83 10 3465 3468 10.1073/pnas.83.10.3465 2422651
50. Burke RM Tate JE Kirkwood CD Steele AD Parashar UD Current and new rotavirus vaccines Curr Opin Infect Dis 2019 32 5 435 444 10.1097/QCO.0000000000000572 31305493
51. Patel M Shane AL Parashar UD Jiang B Gentsch JR Glass RI Oral rotavirus vaccines: how well will they work where they are needed most? J Infect Dis 2009 200 Suppl 1 01 S39 48 10.1086/605035 19817613
52. Fukuda Y Tsugawa T Nagaoka Y Ishii A Nawa T Togashi A Surveillance in hospitalized children with infectious diseases in Japan: pre- and post-coronavirus disease 2019 J Infect Chemother 2021 27 11 1639 1647 10.1016/j.jiac.2021.07.024 34389224
53. Ullrich A Schranz M Rexroth U Hamouda O Schaade L Diercke M Impact of the COVID-19 pandemic and associated non-pharmaceutical interventions on other notifiable infectious diseases in Germany: an analysis of national surveillance data during week 1–2016—week 32–2020 Lancet Reg Health Eur 2021 6 100103 10.1016/j.lanepe.2021.100103 34557831
54. Ahn SY Park JY Lim IS Chae SA Yun SW Lee NM Changes in the occurrence of gastrointestinal infections after COVID-19 in Korea J Korean Med Sci 2021 36 24 e180 10.3346/jkms.2021.36.e180 34155841
|
PMC010xxxxxx/PMC10239223.txt |
==== Front
Ann Hematol
Ann Hematol
Annals of Hematology
0939-5555
1432-0584
Springer Berlin Heidelberg Berlin/Heidelberg
37269388
5294
10.1007/s00277-023-05294-3
Original Article
Sickle cell disease and acute leukemia: one case report and an extensive review
http://orcid.org/0000-0002-6785-7709
Cannas Giovanna giovanna.cannas@chu-lyon.fr
12
Poutrel Solène 12
Heiblig Maël 3
Labussière Hélène 3
Larcher Marie-Virginie 3
Thomas Xavier 2
Hot Arnaud 12
1 grid.412180.e 0000 0001 2198 4166 Internal Medicine, Hospices Civils de Lyon, Edouard Herriot Hospital, 5, place d’Arsonval, Lyon cedex 03, 69437 Lyon, France
2 grid.412180.e 0000 0001 2198 4166 Constitutive reference center: Major sickle cell syndromes, thalassemias and other rare pathologies of red blood cell and erythropoiesis, Edouard Herriot Hospital, Lyon, France
3 grid.411430.3 0000 0001 0288 2594 Hematology, Hospices Civils de Lyon, Lyon-Sud Hospital, Pierre-Bénite, France
3 6 2023
2023
102 7 16571667
20 3 2023
22 5 2023
© The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2023. Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic.
Population-based studies and case reports suggest that there may be an increased risk of acute leukemia associated with sickle cell disease (SCD). Following the description of a new case report, an extensive review of the literature identified 51 previously described cases. Most cases study showed myelodysplastic features confirmed, when available, by genetic markers such as chromosome 5 and/or chromosome 7 abnormalities and TP53 gene mutations. The increased risk of leukemogenesis is certainly multifactorial and related to the pathophysiologic mechanisms of the clinical manifestations of SCD. Chronic hemolysis and secondary hemochromatosis may cause increased chronic inflammation, resulting in persistent marrow stress, which could potentially compromise the genomic stability of the hematopoietic stem cells generating genomic damage and somatic mutations over the course of SCD and its treatment, resulting in a clone that led to acute myeloid leukemia.
Keywords
Sickle cell disease
Acute myeloid leukemia
Treatment
Prognosis
issue-copyright-statement© Springer-Verlag GmbH Germany, part of Springer Nature 2023
==== Body
pmcIntroduction
Sickle cell disease (SCD) corresponds to an autosomal recessive hemoglobinopathy in which structurally abnormal hemoglobin (HbS) leads to chronic hemolytic anemia and to a variety of severe clinical manifestations. The disorder is caused by a point mutation. A single DNA base change leads to substitution of valine for glutamic acid at the sixth position on β globin chain. Patients with homozygous hemoglobin (SS) often present with severe symptoms, while those with a heterozygous mutant allele (SA) demonstrate minimal clinical symptoms. The combination of hemoglobin S with another type of β subunit gene mutation, such as hemoglobin C or β thalassemia, forms a compound heterozygous hemoglobinopathy (SC or Sβ0). With the exception of the compound Sβ0, the heterozygous genotypes are usually less clinically severe than hemoglobin SS [1].
Since Herrick’s description of SCD in 1910 [2], a wide variety of malignancies, including hematological neoplasms, have been reported in both children and adults with SCD. However, the exact incidence of malignancy has not been accurately determined due to a lack of long-term follow-up. The first description of SCD coexisting with acute leukemia has been reported by Goldin et al. in 1953 in a 38-year-old black man with SCD and acute myeloid leukemia (AML) [3]. Since then, the occurrence of acute leukemia has been reported in several cases of patients with SCD.
We reported here a new case of SCD patient who developed AML and reviewed extensively the literature in order to better understand the relationship between the two diseases. This review leads to the hypothesis of a mechanism involving multifactorial causes through the pathophysiologic mechanisms of the clinical manifestations of SCD.
Patients and methods
Case selection
A sole case of acute leukemia in the setting of SCD was retrieved from the pathology database of the “Centre de Référence Constitutif des pathologies du globule rouge et de l’érythropoïèse” in Lyon (France), including a pool of more than 600 adults and children with SCD. The diagnosis of leukemia was confirmed according to the World Health Organization classification [4]. Informed consent for reporting this case was obtained from this patient in accordance with the declaration of Helsinki. Clinical history and laboratory data, including flow cytometry analysis, cytogenetics, and molecular biology, were collected as well as data regarding SCD history.
Literature data sources
The PubMed database was searched on October 2022 for case reports previously published involving both SCD and acute leukemia. The relevant keywords used were: “sickle cell disease” or “sickle cell anemia”, combined with “acute leukemia”, or “myelodysplastic syndrome” (MDS). Fifty-one previously published cases were identified since 1972, and the relevant data regarding acute leukemia and SCD were collected and analyzed (Tables 1 and 2). These cases did not included those only mentioned in the epidemiologic reports from California and the United Kingdom [5, 6].Table 1 Acute leukemia characteristics and outcome of the 51 patients identified in the literature
Reference Pt/Age/Gender Type of AL Treatment Outcome
(Cause of death)
Jackson (1972) [7] #1/6/F ALL Chemo CR
OS: 17 months
(viremia)
Samal (1979) [8] #2/7/F AML None OS: 4 days
Clinicopathologic conference (1982) [9] #3/27/F MDS/AML4 None OS: 3 days
(ARDS)
Johnson (1984) [10] #4/8/F AML2 HSCT OS: 16+ months
Bigner (1986) [11] #5/4/F ALL null
(del9p13)
NA NA
Stricker (1986) [12] #6/43/M MDS/AML1
(− 3, t13;17, t3;5, 5q − , − 7, + 8)
Chemo OS: 1 month
(hemorrhage)
Njoku (1988) [13] #7/22/M ALL Chemo CR
OS: 10 months
(disease progression)
Sotomayer (1999) [14] #8/14/M ALL (CD10+, CD19+, CD22+, DR+, TdT+) Chemo CR
OS: 2.5+ years
De Montalembert (1999) [15] #9/10/F Ph+ ALL Chemo CR
OS: 12+ months
Rauch (1999) [16] #10/27/F MDS/AML NA NA
Wilson (2000) [17] #11/42/F MDS/AML
(− 5, − 7, del17)
Chemo OS: 13 months
Al-Jam’a (2002) [18] #12/25/F AML1
(Normal karyotype)
Chemo CR
OS: NA
(aspergillosis)
Schultz (2003) [19] #13/14/F ALL NA NA (Alive)
#14/5/NA ALL NA NA (Alive)
#15/7/NA ALL NA NA (Alive)
#16/8/NA AML NA NA (Alive)
#17/8/NA ALL NA NA (Alive)
#18/17/NA ALL NA NA (Alive)
#19/61/NA ALL NA NA (Dead)
#20/20/NA AML NA NA (Dead)
Ferster (2003) [20] #21/21/F AML3v ATRA + Chemo CR
Taylor (2011) [21] #22/33/M MDS/AML6
(Abn5q, del7q, − 15, − 22, − Y, mar5)
Chemo
Allo HSCT
CR
Relapse at 4 months
OS: 9 months
Baz (2012) [22] #23/41/M MDS/AML
(Abn5, del7, − 17)
Chemo OS: 3 months
(sepsis)
Zemenides (2014) [23] #24/55/M MDS/AML
(5q − , 7q − , del17p)
NA NA
Aumont (2015) [24] #25/49/M MDS/AML6
(del17p, del5q, monosomy 20)
BM fibrosis
Chemo OS: 3 weeks
(CNS involvement)
Chauhan (2018) [25] #26/25/F AML3 NA NA
#27/19/M AML2 NA NA
Janakiram (2018) [26] #28/31/F MDS/AML
(5q − , add5p, − 7, t2;5, TP53+, NRas+)
Azacitidine OS: 12 months
(sepsis)
Li (2019) [27] #29/59/F MDS (del4, 5q − , 7q − , − 15, − 16, TP53+) Decitabine OS: 2 months
(progression to AML)
#30/27/M MDS/AML (11q23, + 3, + 19, + 21, KMT2A+) ChemoAllo HSCT OS: 7 months
#31/37/F MDS (del1, del5, t3;6, − 17, + 3, TP53+) Lenalidomide + prednisone OS: 5+ months
#32/34/M MDS (7q22, del20, − 2, Inv9) Matched sibling HSCT OS: 21+ months
Eapen (2019)* [28] #33/19/NA AML NA NA
#34/37/NA MDS NA NA
#35/32/NA AML NA NA
#36/37/NA MDS NA NA
Regan (2019) [29] #37/26/F MDS/AML
(5q − , + 8, del17, TP53 deletion)
Chemo OS: 4 months
Aworanti (2020) [30] #38/15/M AL mixed lineage None Death before any treatment
#39/21/F ALL None Discharged at day5CR after 2 lines
#40/15/M AML4 Chemo Discharged after CRDeath 4 weeks after
#41/3/M AML None Discharged after diagnosis
#42/15/F AML5 Chemo OS: 2 months(sepsis)
Yadav (2020) [31] #43/29/F AML6
(5q −)
Chemo OS: few months
(AML progression)
Ghannam (2020) [32] #44/39/M MDS/AML7
(complex cytogenetics, TP53+, BM fibrosis)
Decitabine
Azacitidine
OS: 12 months
(pulmonary hypertension)
#45/39/M MDS/AML(complex cytogenetics, TP53+) Haplo HSCT OS: 7 months(intracranial hemorrhage)
#46/49/F MDS/AML(7q − , BM fibrosis) NA NA
Chellapandian (2020) [33] #47/14/F AML CNS+
(FLT3-ITD+)
Chemo + sorafenib
Haplo HSCT
CR
OS: 8+ months
Hsieh (2020) [34] #48/42/M MDS/AML
(− 7, 19p abnormality, RUNX1+, KRAS+, PTPN11+)
Azacitidine
Decitabine
Chemo
Haplo HSCT
CR after Haplo
OS: 6+ months
Ahmed (2021) [35] #49/19/M Ph+ ALL Chemo + imatinib OS: 6 months
(meningoencephalitis)
Goyal (2022) [36] #50/31/F AML0
(− 7, 11p − , WT1+, RUNX1+, PTPN11+)
Chemo
Haplo HSCT
CR (MRD+)
OS: 12 months
(AML progression)
Flevari (2022) [37] #51/40/M MDS
(complex cytogenetics, 5q‒, 3p, 7p, ‒16, ‒7, ‒18)
None OS: 3 months
(severe cytopenia)
Our case report #52/27/F MDS/AML
(− 3, t5;7, − 7, del12, − 22, TP53+)
Vyxeos
MEC
HSCT
CR
MRD− after HSCT
OS: 12+ months
Abbreviations: Abn, abnormality; AL, acute leukemia; ALL, acute lymphoblastic leukemia; Allo, allogeneic; AML, acute myeloid leukemia; ARDS, Acute respiratory distress syndrome; ATRA, all-trans retinoic acid; BM, bone marrow; Chemo, intensive chemotherapy; CNS, central nervous system; CR, complete remission; F, female; Haplo, haplo-identical; HSCT, hematopoietic stem cell transplantation; NA, not available; M, male; MEC, chemotherapy combining mitoxantrone, etoposide, and cytarabine; MDS, myelodysplastic syndrome; MRD, measurable residual disease; OS, overall survival; Ph+, Philadelphia chromosome-positive; Pt, patient number
*This reference is based on registry data. The patients may therefore overlap with others reported in the table
Table 2 SCD characteristics of the 51 patients identified in the literature
Reference Pt/Diagnosis/Origin Age at diagnosis Treatment Clinical features
Jackson (1972) [7] #1/SS/Afr.Am NA No HU NA
Samal (1979) [8] #2/SS/NA Infancy Transfusions NA
Clinicopathologic conference (1982) [9] #3/SS/NA NA Transfusions Hemosiderosis
Johnson (1984) [10] #4/SS/Afr.Am Infancy NA NA
Bigner (1986) [11] #5/SS/NA At birth NA NA
Stricker (1986) [12] #6/SC/Afr.Am NA NA Aseptic necrosis humeral head
Njoku (1988) [13] #7/SS/Nigerian NA NA NA
Sotomayer (1999) [14] #8/SS/Afr.Am Infancy No HU VOC
De Montalembert (1999) [15] #9/SS/NA Infancy HU (1.5 m) VOC
(3 to 7/year)
Rauch (1999) [16] #10/SS/NA NA HU (8y) VOC
Wilson (2000) [17] #11/SS/NA NA HU (6y) NA
Al-Jam’a (2002) [18] #12/SS/Saudi Arabian NA HU (2y) VOC
(6/year)
Hepatitis C
Schultz (2003) [19] #13/SS/NA Infancy HU (3 m) NA
#14/SS/NA Infancy No HU NA
#15/SS/NA Infancy No HU NA
#16/SS/NA Infancy No HU
HSCT
NA
#17/SS/NA Infancy No HU NA
#18/SS/NA NA No HU NA
#19/SS/NA NA No HU NA
#20/SS/NA NA No HU NA
Ferster (2003) [20] #21/SS/NA NA HU (8y) VOC
Osteonecrosis
ACS
Taylor (2011) [21] #22/SS/Afr.Am NA HU (5y)
Transfusions
VOC
Priapism
ACS
Baz (2012) [22] #23/SS/Afr.Am 21 Exchange transfusions
HU (15y)
VOC
(14 to 3/year)
Zemenides (2014) [23] #24/SS/Jamaican NA No HU Pulmonary hypertension
Aumont (2015) [24] #25/SS/NA NA HU (14y)
Transfusions
VOC
Hip necrosis
Retinopathy
Infections
Ischemic stroke
Cholelithiasis
Iron overload
Chauhan (2018)[25] #26/SS/Indian NA Transfusions
HU
NA
#27/SS/Indian NA HU NA
Janakiram (2018) [26] #28/SS/Afr.Am Childhood HU (5y)
Haplo HSCT (8 m)
VOC
Li (2019) [27] #29/SC/NA NA HU
Exchange transfusions
HHV8
#30/SS/NA NA Exchange transfusions VOCMyocardial infarctionHIV+
#31/SS/NA Infancy Exchange transfusions VOC
#32/Sβ0/NA NA Exchange transfusionsHU (9y)Matched HSCT (7y) VOCPriapismArterial anevrysmIntracranial bleeding
Eapen (2019)* [28] #33/NA/NA
#34/NA/NA
#35NA/NA
#36/NA/NA
NA
NA
NA
NA
Haplo HSCT (3.6y)
Haplo HSCT (9 m)
Haplo HSCT (1y)
Matched sibling HSCT (2.6y)
NA
NA
NA
NA
Regan (2019) [29] #37/SS/Afr.Am Childhood Transfusion/Exchange
HU (2y)
VOC
Pulmonary fibrosis
Pneumonia
Hips necrosis
Peritonitis
Aworanti (2020) [30] #38/SS/Nigerian
#39/SS/Nigerian
#40/SC/Nigerian
#41/SC/Nigerian
#42/SS/Nigerian
2 years
4 years
Childhood
et al. diagnosis
NA
No HU
Transfusion
No HU
Transfusion
No HU
Transfusion
None
No HU
None
VOC (1/year)
VOC (1/2 years)
None
NA
Yadav (2020) [31] #43/SS/NA NA HU (5y) VOC
Ghannam (2020)[32] #44/SS/NA
#45/SS/NA
#46/SS/NA
NA
NA
NA
HU
Haplo HSCT (2y)
HU
Sibling HSCT (2.5y)
HU
Haplo HSCT (5y)
Stroke
VOC
CRI
VOC
Diastolic dysfunction
ESRD
Pulmonary hypertension
Chellapandian (2020) [33] #47/Sβ0/Haitian At birth HU (9y) VOC
Hsieh (2020) [34] #48/SS/NA NA HU (8y)
Gene therapy (LentiGlobin) (3y)
VOC
Iron overload
Leg ulcers
Hypertension
Gallbladder disease
Ahmed (2021) [35] #49/SS/Nigerian At 1 year Transfusions
No HU
VOC
(> 4/year)
Goyal (2022) [36] #50/SS/NA NA HU (6y)
Gene therapy (LentiGlobin) (5.5y)
VOC
Hip necrosis
Deep-vein thrombosis
Flevari (2022) [37] #51/SS/NA NA HU (17y)
Exchange transfusions
VOC
Priapism
Pulmonary hypertension
Our case report #52/Sβ0/African Childhood HU (7y)
Exchange transfusions
VOC
ACS
Cholelithiasis
Retinopathy
COVID-19
Abbreviations: ACS, acute chest syndrome; Afr.Am, African-American; AL, acute leukemia; CRI, chronic renal insufficiency; ESRD, end-stage renal disease; Haplo, haplo-identical; HIV, human immunodeficiency virus; HSCT, hematopoietic stem cell transplantation; HU, hydroxyurea; LentiGlobin, Gene therapy consisting of autologous hematopoietic stem and progenitor cells transduced with the BB305 lentiviral vector encoding the βA−T87Q-globin gene designed to produced anti-sickling hemoglobin (HbAY87Q); m, months; NA, not available; Pt; patient number; VOC, vaso-occlusive crisis; y, years
*This reference is based on registry data. The patients may therefore overlap with others reported in the table
Results
Case report
A 27-year-old woman of African origin with known SCD (Sβ0), previously complicated by recurrent severe vaso-occlusive crisis (VOC) and acute chest syndrome despite hydroxyurea (HU) therapy (1000 mg/day for 7 years) and regular exchange transfusions, presented on May 2022 a progressive bicytopenia with anemia to 50 g/dL (based-hemoglobin level under compliant treatment with HU around 75 g/dL) and thrombocytopenia to 50 × 109/L, leading to HU discontinuation. A suspicion of MDS was confirmed by a first bone marrow sample showing a hypercellular marrow with signs of dyserythropoiesis with demonstration of ring sideroblasts on a Perls’stain, dysmegacaryopoiesis, and dysgranulopoiesis, but no leukemic cells. The patient was referred to the Hematology Department and a repeat bone marrow aspirate, performed sequentially showed a progressive blast increase up to 25% leading to the diagnosis of AML-MRC (myelodysplastic related changes). The immunophenotypic profile was CD34+/− CD38+/− CD123+/−, CD13+/− CD33+/− CD117++, HLADR+/−, CD36+/− CD71++, CD7− CD19− CD56−, MPO−. Seventeen percent of myeloblasts expressed a multipotent progenitor-like leukemia stem cell (LSC) profile CD34+ CD38− CD90++/− CD45RA+/− CLL1/TIM3/CD97+/−. Cytogenetic analysis showed a complex karyotype: 44–46, XX, der(1)t(1;12)(q31;q15)add(1)(p11), − 3, der(5)t(5;7)(q13;q31), − 7, del(12)9q21), der(15)t(?3;15)(q21;p13), add(16)9p13), − 22, + 3-6mars[cp18]/46, XX [2]. Molecular study by next-generation sequencing (NGS) identified the presence of a TP53 mutation c.1024C > T with a variant allele frequency (VAF) of 0.64. The patient received induction chemotherapy with Vyxeos (daunorubicin/cytarabine) at a dose of 44 mg/m2 on days 1, 3, and 5. On day 22, peripheral blood showed 18% blasts signing remnant leukemia. Salvage chemotherapy combined mitoxantrone 6 mg/m2/day, etoposide 80 mg/m2/day, and intermediate-dose cytarabine 1 g/m2/day from day 1 to day 6. Salvage chemotherapy was complicated by infections including inguinal cellulitis requiring large spectrum antibiotics and white blood cell infusion therapy, pericarditis, and posterior reversible encephalopathy syndrome (PRES) leading to a transitory hospitalization in intensive care unit. Cytological remission was achieved, but measurable residual disease (MRD) remained positive at 0.09% based on leukemia associated immunophenotype (LAIP)/LSC. Allogeneic phenol-identical hematopoietic stem cell transplantation (HSCT) with one mismatch was performed on January 2022 based on thiotepa-busulfan-fludarabine (TBF) conditioning regimen followed by post-transplant cyclophosphamide and everolimus for graft-versus-host prophylaxis. The hospitalization was complicated by a septic shock (Klebsiella pneumonia) and by invasive pulmonary aspergillosis and severe hepato-splenic candidosis (Candida glabrata). Bone marrow evaluation at one month and two months post-transplant confirmed the cytological remission with MRD negativity assessed by multi-parameter flow cytometry and total donor chimerism.
Review of the literature
Fifty-two cases of acute leukemia in SCD patients (including our case report) were identified in the literature since 1972 (Tables 1 and 2). Among patients with available data, male/female sex ratio was 0.45. Median age was 23.5 years (range: 3 – 61 years). Thirteen patients (25%) had acute lymphoblastic leukemia (ALL), one patient an undifferentiated acute leukemia, and 38 patients (73%) a myeloid neoplasm, including 16 AML, 6 MDS and 16 MDS/AML. Among the 26 patients studied for genetic markers, two patients with ALL showed a Philadelphia chromosome (Ph +) (#9, #49), one patient with AML had a normal karyotype (#12), two had acute promyelocytic leukemia (#21, #26), and 20 patients with MDS and/or AML displayed unfavorable cytogenetics [− 5, − 7, del(17), 11q23, chromosome 3 abnormality] and/or molecular abnormalities of poor prognosis [TP53, KMT2A, RAS, RUNX1, PTPN11] (#6, #11, #22–25, #28-#32, #37, #43–46, #48, #50, #51, #52). Most of the patients (84%) displayed a SS homozygous hemoglobin, while only 9% were SC and 7% Sβ0. Data were available in 38 patients regarding the potential use of long-term SCD therapy with HU: 16 (43%) did not receive any HU, while 22 (57%) received HU prior to acute leukemia diagnosis (median duration of treatment: 6.5 years; range: 0.05 – 17 years). Ten patients underwent allogeneic HSCT, after conditioning regimen including alkylating agents and/or total body irradiation, as treatment of SCD. Four patients were allografted from a matched sibling donor (#16, #32, #36, #45) and six patients from a haploidentical donor (#28, #33–35, #44, #46) (Table 2). The median time between HSCT and acute leukemia diagnosis was 2.5 years (range: 0.26 – 7 years). Two cases of AML developed in SCD patients who had been treated by gene therapy with LentiGlobin [34, 36], which required myelo-ablation with an alkylating agent.
Overall acute leukemias in SCD patients were of dismal outcome with overall survival (OS) ranging from few days to 2.5+ years (median in patients with available data: 7 months).
Discussion
Historically, the development of malignancy in children and adults patients with SCD has been documented by several small series [7, 12, 38]. On the basis of a single institution study, the cancer incidence in SCD patients has been estimated to be 1.74 cases per 1,000 patient-years [39]. Malignancies mainly included hematological neoplasias, especially acute leukemias. In the 1970s, Jackson reported, among 58 black children treated for acute leukemia, 4 ALL and 3 AML with sickle cell trait, and one ALL with homozygous HbS [7]. In a low-income country, the association of acute leukemia with SCD was even reported in 8.6% of cases [30]. Actually, the risk of hematologic malignancies is 2 to 11 times as high as that in the general population. This was established by three recent epidemiology reports [5, 6, 19]. The first study used a standardized incidence ratio (SIR) to compare individuals with SCD to the general population. One hundred and fifteen on 6423 SCD individuals were diagnosed with cancer, with a total of 6 AML cases (SIR, 3.59; 95% confidence interval: 1.32–7.82) and 3 cases of ALL (1.83; 0.38–5.35) [5]. In the second study, 8 cases of AML on 7512 individuals with SCD were reported (11.05; 3.86–30.17). Among hematological malignancies, the risks remained elevated for all conditions studied, except for lymphoid leukemia [6]. The third study identified 52 cases of cancer in 49 patients among 16,613 individuals with SCD, 40% of cases occurring in children [19]. The most frequent malignancy was acute leukemia (8 cases).
The vast majority of SCD patients receive conservative therapy. In this setting, HU has greatly improved the survival of SCD patients in developed countries, due to its efficacy in preventing VOC via an inhibitory effect on HbS polymerization by increasing the synthesis of fetal hemoglobin, and an improvement of blood flow in the microcirculation through the expression or activity of several adhesion molecules on red blood and endothelial cells [40, 41]. Three randomized placebo controlled trials have demonstrated the efficacy of HU in SCD, with an excellent safety profile and up to a 40% reduction in mortality after 9 years of follow-up [42–44]. HU is an inhibitor of DNA synthesis that may theoretically lead to an accumulation of acquired DNA mutations and eventually leukemic transformation. Whether acute leukemia in SCD patients with long-term exposure to HU is a co-incidental or related to therapy has been a major issue debated in many reports. The leukemogenic risk could theoretically increase with the duration of drug exposure. The index of DNA damage in peripheral blood leukocytes from HU-treated patients with SCD was demonstrated higher than in controls and was confirmed influenced by the duration and the dose of HU treatment, and by the HbS genotype [45, 46]. The leukemic risk of HU has never been confirmed in patients with chronic myeloproliferative diseases [47, 48], and no increased risks of malignancy were reported in large series of SCD patients [49–51]. Among 278 SCD children receiving long-term treatment with HU, only one developed acute leukemia [15, 52, 53]. If one study in pediatric SCD patients treated with HU showed that genotoxicity increased with HU administration [54], it was demonstrated that individuals may have different susceptibilities to HU, and that this occurred in a patient population that may already have an elevated risk for malignancy evaluated at baseline by a greater Damage Index [55]. Overall the genotoxicity results clearly demonstrate that HU does not directly bind DNA and is not mutagenic [56]. In vitro, HU can result in the accumulation of somatic mutations and chromosomal damages due to interference with DNA repair, but the number of acquired mutations did not increase in patients with long-term exposure to the drug [57]. On another hand, HU therapy can alleviate the risk of chronic hemolysis by increasing the fetal hemoglobin content in the blood, and potentially reduce the accompanying marrow stress in these patients. The recent prospective observational study ESCORT-HU (NCT02516579), which evaluated the long-term safety and effectiveness of HU in SCD patients across several European centers, confirmed the benefit-to-risk ratio of HU in children and adults [58]. Only one incident hematological malignancy was reported.
In contrast to life-long supportive care measures, HSCT offers a curative option but may be followed by various severe complications. It is therefore being reserved to patients who are refractory to conventional therapy. The 5-year OS ranges from 91 to 95% in children who underwent HLA-identical HSCT after myeloablative conditioning, while disease-free survival, rate of rejection, and incidence of chronic graft versus host disease are approximately 82%, 8%, and 12%, respectively [59, 60]. Nine percent of patients died of complications related to transplantation [59]. Peripheral blood stem cells, which would come from AA or AS donors, have also been proposed as a source of stem cells for allogeneic HSCT. The results after related (5-year OS: 97%) and unrelated (2-year EFS: 90%) donor umbilical cord transplantation (UCT) have also been encouraging [61]. However recent results were more discouraging showing a high incidence of graft rejection (50% to 62%) after unrelated UCT [62], although updated data using a reduced intensity conditioning combining HU, alemtuzumab, fludarabine, thiotepa, and melphalan were more impressive [63]. Haploidentical-related donor transplantations are under study. It is becoming a viable alternative curative option for SCD, extending the availability of HSCT as a treatment option to eligible SCD patients. Overall survival was high (91%) in all studies included in a recent meta-analysis [64]. One study has suggested that HSCT for SCD does not increase the risk of developing acute leukemia, compared with patients who have not undergone SCT [60]. However, transplanted patients are generally exposed to alkylating agents and ionizing radiation as part of a conditioning regimen, and intervals, found in the literature, between the procedure and the diagnosis of leukemia are falling in the range of latency reported in other diseases. Furthermore, therapy-related MDS/AML is a well known event after autologous transplantation for lymphomas, with cumulative risks as high as 15%.
Trials in gene therapy are under way and also offer great promise. However, the largest lentiviral vector-mediated β-globin replacement gene therapy trial in SCD reported two cases of adult patients diagnosed with AML [34, 36, 65, 66]. These two cases shared similar cytogenetic and molecular abnormalities with monosomy 7 and RUNX1 and PTPN11 mutations, which were not found in patients pre-conditioning bone marrow samples. The first case was considered to be related to busulfan conditioning [34]. The second case showed vector present in leukemia blast cells, which suggests that blast cells originated from a transduced hematopoietic stem cell and not from residual host cells exposed to busulfan [36]. However, several lines of evidence showed that the development of this case of AML most likely occurred independently of insertional oncogenesis [36].
If a coincidental event between acute leukemia and SCD could be evocated in several cases from the literature that resemble de novo acute leukemia, an increased risk for acute leukemia is suggested by the significant underlying MDS features of most reported cases compared to leukemic patients from the general population of the same age. Extensive literature review demonstrates at least 18 patients with presence of complex structural rearrangements involving complete or partial loss of chromosome 5 and/or chromosome 7 and/or 17p deletions. TP53 gene mutations have also been shown frequently implicated [67]. Furthermore, several cases were classified as AML6 or AML7 and/or presented bone marrow fibrosis. Those facts are not in favor of a simple coincidence between the occurrence of acute leukemia and SCD, but are generally considered as a marker of secondary leukemia.
The exact underlying connection between acute leukemia and SCD is not clearly understood. Beside therapy for SCD, other potential cancer risk factors might exist for SCD patients and have been discussed in a recent published commentary [68]. Red blood cell transfusions can lead to increased iron levels and non-specific immunomodulation that could increase the risk of malignancy. However, heavily transfused patients with thalassemia only show a few cases of cancer [69]. Chronic inflammation implies the potential involvement of inflammasomes in SCD pathogenesis [70]. Chronic organ damage with inflammation could also cause cellular damage with subsequent malignant transformation. The pro-tumorigenic role of inflammasomes is associated with promoting cell proliferation, inhibition of apoptosis, and an immunosuppressive effect on the immune cells. Constant hematopoietic hyperplasia, stimulated by a hemolysis-induced cytokine storm, may increase the risk of somatic mutations, resulting in transformation of myeloid precursors [71]. Other factors associated with the increased risk include increased risk of infections, and increased bone marrow turnover, which form the pathophysiologic mechanisms of the clinical manifestations of SCD [5, 72]. The accumulation of multiple genetic abnormalities over years, due to a high degree of proliferative activity of bone marrow cells, may be responsible of the increased risk of cancer.
After myeloablation, the bone marrow niche undergoes extensive proliferation of hematopoietic stem cells, generating proliferative stress that may lead to mutations as part of the normal engraftment process [73]. After HSCT, myeloid malignancy was only seen within patients who did not engraft [28, 32]. In case of graft failure, the need for more replication cycles is required to repopulate the bone marrow, increasing the probability of acquiring a mutation that could lead to AML. TP53 mutations were detectable in blood before transplantation and increase until therapy-related myeloid malignancy diagnosis [32]. The progression of baseline high-risk TP53 clonal abnormalities into AML in patients with SCD has been reported after unsuccessful allogeneic HSCT. It has been previously demonstrated that the TP53 mutated clones specially expanded after chemotherapy exposure [74]. Because of erythropoietic stress and systemic inflammation, SCD patients may have been predisposed to developing clonal hematopoiesis. As these clones may be more resistant to radiation and/or chemotherapy, it has been suggested that they may preferentially expand after a failed transplant, leading to the myeloid malignancy detected after graft rejection.
TP53 is the most commonly mutated gene in therapy-related MDS/AML. Low folic acid, associated with an increased risk for leukemia, can make cells vulnerable to mutagenesis and can affect the genetic and epigenetic integrity of TP53 [75]. TP53 plays a central role in regulating cellular responses to genotoxic stress, and loss of TP53 provides a selective advantage for neoplastic growth [76]. The specific TP53 mutation has been shown to be present at low frequencies (0.003–0.7%) in blood leucocytes in some cases 3–6 years prior to the development of therapy-related MDS/AML and prior any chemotherapy [74]. TP53 mutations have also identified in small populations of peripheral blood cells of healthy chemotherapy-naïve elderly individuals. Chromosomal aberrations were demonstrated in some SCD patients with no evidence of hematological disease [27]. Furthermore, murine bone marrow chimeras containing wild type and TP53+/− hematopoietic stem/progenitor cells preferentially expanded after exposure to chemotherapy [74]. These data suggest that TP53 mutations precede the development of AML and the acquisition of other mutations, such as TET2, NUP98, or RUNX1.
Despite limitations coming from the retrospective nature of our study involving missing data and biases related to cases reported over an extended period, our review of the literature tend to suggest that chronic hemolysis, increased iron levels, and increased bone marrow turnover, which form the pathophysiologic mechanisms of the clinical manifestations of SCD are mainly responsible for a situation in which cells are undergoing constant hematopoietic hyperplasia, leading to the increased risk of acute leukemia by inducing genomic damage and somatic mutations [77]. The effects of SCD on progenitor cells have not been fully determined [78]. SCD may promote accelerated aging of hematopoietic cells and oncogenic somatic mutations [79]. Further studies are needed to identify risk factors for developing acute leukemia by pre-screening individuals with SCD. Next-generation DNA sequencing can be used to detect expanded peripheral blood progeny of a mutant clone and clonal hematopoisis of indeterminate potential (CHIP), which is a risk factor for subsequent hematologic malignancy [80]. Recent large studies have tried to address clonal hematopoiesis in SCD [81, 82]. Despite different conclusions related to the technique used, the control cohort chosen, and the value of VAF defined for considering clonal hematopoiesis, a small percentage of cases were identified as having somatic variants of TP53, DNMT3A, ASXL1, and/or TET2.
In conclusion, several cases of MDS/AML have been reported in SCD leading to the hypothesis that SCD may lead to the development of hematopoietic malignancies, even in the absence of disease-modifying treatments. The increased risk of leukemogenesis is certainly multifactorial and related to the pathophysiologic mechanisms of the clinical manifestations of SCD, which may promote accelerated aging of hematopoiesis. A prevalence of clonal hematopoiesis in SCD patients should demonstrate a higher risk than in the general population.
Author contributions
GC treated the patient, interpreted the data, and wrote the manuscript; SP, MH, HL, and MVL treated the patient; XT collected and interpreted the data, and wrote the manuscript; and AH reviewed the manuscript. All authors gave final approval of the version to be submitted.
Funding
None.
Data availability
The data that support the findings of this study are available on request from the corresponding author.
Declarations
Ethical approval
All procedures performed in the study were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments.
Informed consent
Informed consent was obtained from the individual included in the study.
Conflicts of interests
The authors do not have any competing financial interest in relation with the work described.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References
1. Piel FB Steinberg MH Rees DC Sickle cell disease N Engl J Med 2017 376 1561 1573 10.1056/NEJMra1510865 28423290
2. Herrick JD Peculiar elongated and sickle-shaped red blood corpuscles in a case of severe anemia Arch Intern Med 1910 6 517 521 10.1001/archinte.1910.00050330050003
3. Goldin AG Kelty KC Beard MF Sickle cell anemia terminating in acute myeloblastic leukemia Ann Intern Med 1953 39 920 928 10.7326/0003-4819-39-4-920 13092745
4. Arber DA Orazi A Hasserjian R The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia Blood 2016 127 2391 2405 10.1182/blood-2016-03-643544 27069254
5. Brunson A Keegan THM Bang H Increased risk of leukemia among sickle cell disease patients in California Blood 2017 130 1597 1599 10.1182/blood-2017-05-783233 28830890
6. Seminog OO Ogunlaja OI Yeates D Risk of individual malignant neoplasms in patients with sickle cell disease: English National Record Linkage Study J R Soc Med 2016 109 303 309 10.1177/0141076816651037 27325377
7. Jackson RE Short BJ Frequency and prognosis of coexisting sickle cell disease and acute leukemia in children Clin Pediatr 1972 183 183 185 10.1177/000992287201100313
8. Samal GC Sickle cell anemia with acute myeloid leukemia – (a case report) Indian Pediatr 1979 16 453 454 292624
9. Conference C Left elbow pain and death in a young woman with sickle-cell anemia Am J Med 1982 73 268 282 10.1016/0002-9343(82)90189-9 6956237
10. Johnson FL Look AT Gockerman J Bone marrow transplantation in a patient with sickle-cell anemia N Engl J Med 1984 311 780 783 10.1056/NEJM198409203111207 6382010
11. Bigner SH Friedman HS Kinney TR 9p− in a girl with acute lymphocytic leukemia and sickle cell disease Cancer Genet Cytogenet 1986 21 267 269 10.1016/0165-4608(86)90007-5 3456265
12. Stricker RB Linker CA Crowley TJ Embury SH Hematology malignancy in sickle cell disease: report of four cases and review of the literature Am J Hematol 1986 21 223 230 10.1002/ajh.2830210212 3455791
13. Njoku OS Johnson SB Kulkarni AG Mba EC Acute lymphoblastic leukaemia in a Nigerian adult with sickle cell anaemia Centr Afr J Med 1988 34 158 160
14. Sotomayer EA Glasser L Acute lymphoblastic leukemia in sickle cell disease Arch Pathol Lab Med 1999 123 744 745
15. De Montalembert M Bégué P Bernaudin F Preliminary report of a toxicity study of hydroxyurea in sickle cell disease: French Study Group on Sickle Cell Disease Arch Dis Child 1999 81 437 439 10.1136/adc.81.5.437 10519721
16. Rauch A Borromeo M Ghafoor A Leukemogenesis of hydroxyurea in the treatment of sickle cell anemia Blood 1999 94 suppl.1 415a
17. Wilson S Acute leukemia in a patient with sickle cell anemia treated with hydroxyurea Ann Intern Med 2000 133 925 926 10.7326/0003-4819-133-11-200012050-00029 11103075
18. Al-Jam’a AH Al-Dabbous IA Al-Khatti AA Esan FG Are we underestimating the leukemogenic risk of hydroxyurea Saudi Med J 2002 23 1411 1413 12506308
19. Schultz WH Ware RE Malignancy in patients with sickle cell disease Am J Hematol 2003 74 249 253 10.1002/ajh.10427 14635205
20. Ferster A Sariban E Meuleman N Malignancies in sickle cell disease patients treated with hydroxyurea Br J Haematol 2003 123 368 369 10.1046/j.1365-2141.2003.04614.x 14531924
21. Taylor JG Darari DS Maric I Therapy-related acute myelogenous leukemia in a hydroxyurea-treated patient with sickle cell anemia Ann Intern Med 2011 155 722 724 10.7326/0003-4819-155-10-201111150-00024 22084346
22. Baz W Najfeld V Yotsuya M Development of myelodysplastic syndrome and acute myeloid leukemia 15 years after hydroxyurea use in a patient with sickle cell anemia Clin Med Insights Oncol 2012 6 149 152 10.4137/CMO.S8810 22550403
23. Zemenides S Erblich T Luqmani A Bain BJ Peripheral blood features of acute myeloid leukemia with myelodysplasia-related changes developing in a patient with sickle cell anemia Am J Hematol 2014 89 1010 10.1002/ajh.23768 24852219
24. Aumont C Driss F Lazure T Myelodysplastic syndrome with clonal cytogenetic abnormalities followed by fatal erythroid leukemia after 14 years of exposure to hydroxyurea for sickle cell anemia Am J Hematol 2015 90 E131 E132 10.1002/ajh.24010 25801602
25. Chauhan S Swain SK Sahu MC Incidence of hematological malignancies in sickle cell patients from an Indian tertiary care teaching hospital Asian J Pharm Clin Res 2018 11 205 209 10.22159/ajpcr.2018.v11i3.22790
26. Janakiram M Verma A Wang Y Accerated leukemic transformation after haplo-identical transplantation for hydroxyurea-treated sickle cell disease Leuk Lymphoma 2018 59 241 244 10.1080/10428194.2017.1324158 28587497
27. Li Y Maule J Neff JL Myeloid neoplasms in the setting of sickle cell disease: an intrinsic association with the underlying condition rather than a coincidence; report of 4 cases and review of the literature Modern Pathol 2019 32 1712 1726 10.1038/s41379-019-0325-6
28. Eapen M Brazauskas R Walters MC Impact of donor type and conditioning regimen intensity on allogeneic transplantation outcomes in patients with sickle cell disease: a retrospective, cohort study Lancet Haematol 2019 6 e585 e596 10.1016/S2352-3026(19)30154-1 31495699
29. Regan S Yang X Finnberg NK Occurrence of acute myeloid leukemia in hydroxyurea-treated sickle cell disease patient Cancer Biol Ther 2019 20 1389 1397 10.1080/15384047.2019.1647055 31423878
30. Aworanti OW Fasola FA Kotila TR Acute leukemia in sickle cell disease patients in a tertiary health facility in Nigeria: a case series Afri Health Sci 2020 20 1304 1312 10.4314/ahs.v20i3.36
31. Yadav DK Paul T Alhamar M Pure erythroid leukemia in a sickle cell patient treated with hydroxyurea Case Rep Oncol 2020 13 857 862 10.1159/000508361 32884531
32. Ghannam JY Xu X Maric I Baseline TP53 mutations in adults with SCD developing myeloid malignancy following hematopoietic cell transplantation Blood 2020 135 1185 1188 32062672
33. Chellapandian D Nicholson CL Haploidentical bone marrow transplantation in a patient with sickle cell disease and acute myeloid leukemia Pediatr Transplant 2020 24 e13641 10.1111/petr.13641 31880407
34. Hsieh MM Bonner M Pierciey FJ Myelodysplastic syndrome unrelated to lentiviral vector in a patient treated with gene therapy for sickle cell disease Blood Adv 2020 4 2058 2063 10.1182/bloodadvances.2019001330 32396618
35. Ahmed IO Ochogwu LO Owojuyigbe TO Philadelphia chromosome-positive B-cell acute lymphoblastic leukemia with e1a3 BCR-ABL1 transcript in a Nigerian with sickle cell anemia: a case report J Med Case Reports 2021 15 504 10.1186/s13256-021-03060-5
36. Goyal S Tisdale J Schmidt M Acute myeloid leukemia after gene therapy for sickle cell disease N Engl J Med 2022 386 138 147 10.1056/NEJMoa2109167 34898140
37. Flevari P Voskaridou E Galacteros F Case report of myelodysplastic syndrome in a sickle-cell disease patient treated with hydroxyurea and literature review Biomedicines 2022 10 3201 10.3390/biomedicines10123201 36551957
38. Paydas S Sickle cell anemia and hematological neoplasias Leuk Lymphoma 2002 43 1431 1434 10.1080/1042819022386833 12389625
39. Dawkins FW Kim KS Squires RS Cancer incidence rate and mortality rate in sickle cell disease patients at Howard University Hospital: 1986–1995 Am J Hematol 1997 55 188 192 10.1002/(SICI)1096-8652(199707)55:4<188::AID-AJH4>3.0.CO;2-O 9257878
40. Green NS Barral S Emerging science of hydroxyurea therapy for pediatric sickle cell disease Pediatr Res 2014 75 196 204 10.1038/pr.2013.227 24252885
41. Steinberg MH Barton F Castro O Effect of hydroxyurea on mortalityand morbidity in adult sickle cell anemia: risks and benefits up to 9 years of treatment JAMA 2003 289 1645 1651 10.1001/jama.289.13.1645 12672732
42. Charache S Terrin ML Moore RD Effect of hydroxyurea on the frequency of painful crises in sickle cell anemia. Investigators of the Multicenter Study of Hydroxyurea in Sickle Cell Anemia N Engl J Med 1995 332 1317 1322 10.1056/NEJM199505183322001 7715639
43. Ferster A Vermylen C Cornu G Hydroxyurea for treatment of severe sickle cell anemia: a pediatric clinical trial Blood 1996 88 1960 1964 10.1182/blood.V88.6.1960.bloodjournal8861960 8822914
44. Wang WC Ware RE Miller ST Hydroxycarbamide in very young children with sickle-cell anaemia: a multicentre, randomized, controlled trial (BABY HUG) Lancet 2011 377 1663 1672 10.1016/S0140-6736(11)60355-3 21571150
45. Da Silva Rocha LB Dias Elias DB Barbosa MC DNA damage in leukocytes of sickle cell anemia patients is associated with hydroxyurea therapy and with HBB*S haplotype Mutat Res 2012 749 48 52 10.1016/j.mrgentox.2012.08.003 22918118
46. Maia Filho PA Pereira JF Almeida Filho TPD Is chronic use of hydroxyurea safe for patients with sickle cell anemia? An account of genotoxicity and mutagenicity Environ Mol Mutagen 2019 60 302 304 10.1002/em.22260 30525240
47. Kiladjian JJ Chevret S Dosquet C Treatment of polycythemia vera with hydroxyurea and pipobroman: final results of a randomized trial initiated in 1980 J Clin Oncol 2011 29 3907 3913 10.1200/JCO.2011.36.0792 21911721
48. Lanzkron S Strouse JJ Wilson R Systematic review: hydroxyurea for the treatment of adults with sickle cell diseases Ann Intern Med 2008 148 939 955 10.7326/0003-4819-148-12-200806170-00221 18458272
49. Steinberg MH McCarthy WF Castro O The risks and benefits of long-term use of hydroxyurea in sickle cell anemia: a 17.5 year follow-up Am J Hematol 2010 85 403 408 20513116
50. Voskaridou E Christoulas D Bilalis A The effect of prolonged administration of hydroxyurea on morbidity and mortality in adult patients with sickle cell syndromes: results of a 17-year, single-center trial (LaSHS) Blood 2010 115 2354 2363 10.1182/blood-2009-05-221333 19903897
51. Castro O Nouraie M Oneal P Hydroxycarbamide treatment in sickle cell disease: estimates of possible leukaemia risk and of hospitalization survival benefit Br J Haematol 2014 167 687 691 10.1111/bjh.13093 25146244
52. Kinney TR Helms RW O’Branski EE Safety of hydroxyurea in children with sickle cell anemia: results of the HUG-KIDS study, a phase I/II trial Blood 1999 94 1550 1554 10477679
53. Ferster A Tahriri P Vermylen C Five years of experience with hydroxyurea in children and young adults with sickle cell disease Blood 2001 97 3628 3632 10.1182/blood.V97.11.3628 11369660
54. Flanagan JM Howard TA Mortier N Assessment of genotoxicity associated with hydroxyurea therapy in children with sickle cell anemia Mutat Res 2010 698 38 42 10.1016/j.mrgentox.2010.03.001 20230905
55. Rodriguez A Duez P Dedeken L Hydroxyurea (hydroxycarbamine) genotoxicity in pediatric patients with sickle cell disease Pediatr Blood Cancer 2018 65 e27022 10.1002/pbc.27022 29512872
56. Ware RE Dertinger SD Absence of hydroxyurea-induced mutational effects supports higher utilization for the treatment of sickle cell anaemia Br J Haematol 2021 194 252 266 10.1111/bjh.17323 33570176
57. Hanft V Fruchtman S Pickens C Acquired DNA mutations associated with in vivo hydroxyurea exposure Blood 2000 95 3589 3593 10.1182/blood.V95.11.3589 10828048
58. De Montelembert M Voskaridou E Oevermann L Real-life experience with hydroxyurea in patients with sickle cell disease: Results from the prospective ESCORT-HU cohort study Am J Hematol 2021 96 1223 1231 10.1002/ajh.26286 34224583
59. Vermylen C Hematopoietic stem cell transplantation in sickle cell disease Blood Rev 2003 17 163 166 10.1016/S0268-960X(03)00006-7 12818226
60. Gluckman E Cappelli B Bernaudin F Sickle cell disease: an international survey of results of HLA-identical sibling hematopoietic stem cell transplantation Blood 2017 129 1548 1556 10.1182/blood-2016-10-745711 27965196
61. Locatelli F Rocha V Reed W Related umbilical cord blood transplant in patients with thalassemia and sickle cell disease Blood 2003 101 2137 2143 10.1182/blood-2002-07-2090 12424197
62. Kamani NR Walters MC Carter S Unrelated donor cord blood transplantation for children with severe sickle cell disease: results of one cohort from the phase II study from the Blood and Marrow Transplant Clinical Trials Network (BMT CTN) Biol Blood Marrow Transplant 2012 18 1265 1272 10.1016/j.bbmt.2012.01.019 22343376
63. Abraham A Cluster A Jacobsohn D Unrelated umbilical cord blood transplantation for sickle cell disease following reduced-intensity conditioning: results of a phase I trial Biol Blood Marrow Transplant 2017 23 1587 1592 10.1016/j.bbmt.2017.05.027 28578010
64. Aydin M Dovern E Leeflang MMG Haploidentical allogeneic stem cell transplantation in sickle cell disease: a systematic review and meta-analysis Transplant Cell Ther 2021 27 1004.e1 1004.e8 10.1016/j.jtct.2021.09.009 34537420
65. Jones RJ DeBaun MR Leukemia after gene therapy for sickle cell disease: insertional mutagenesis, busulfan, both, or neither Blood 2021 138 942 947 10.1182/blood.2021011488 34115136
66. Kanter J Walters MC Krishnamurti L Biologic and clinical efficacy of LentiGlobin for sickle cell disease N Engl J Med 2022 386 617 628 10.1056/NEJMoa2117175 34898139
67. Christiansen DH Andersen MK Pedersen-Bjergaard J Mutations with loss of heterozygosity of p53 are common in therapy-related myelodysplasia and acute myeloid leukemia after exposure to alkylating agents and significantly associated with deletion or loss of 5q, a complex karyotype, and a poor prognosis J Clin Oncol 2001 19 1405 1413 10.1200/JCO.2001.19.5.1405 11230485
68. Ribeil JA (2022) Primary myelofibrosis in untreated sickle cell disease: Are adult patients at higher risk for developing hematological myeloid neoplasms? 97:4–6
69. Forni GL Gianesin B Musallam KM Overall and complication-free survival in a large cohort of patients with β-thalassemia major followed over 50 years Am J Hematol 2023 98 381 387 10.1002/ajh.26798 36588408
70. Tomasik J Basak GW Inflammasomes – New contributors to blood diseases Int J Mol Sci 2022 23 8129 10.3390/ijms23158129 35897704
71. Alves PM Martins PRJ Dias FDL Sensitivity to cisplatin-induced mutations and elevated chromosomal aberrations in lymphocytes from sickle cell disease patients Clin Exp Med 2008 8 31 35 10.1007/s10238-008-0153-3 18385938
72. Crusz SM Balkwill FR Inflammation and cancer: advances and new agents Nat Rev Clin Oncol 2015 12 584 596 10.1038/nrclinonc.2015.105 26122183
73. Nawas MT Schetelig J Damm F The clinical implications of clonal hematopoiesis in hematopoietic cell transplantation Blood Rev 2021 46 100744 10.1016/j.blre.2020.100744 32896435
74. Wong TN Ramsingh G Young AL Role of TP53 mutations in the origin and evolution of therapy-related acute myeloid leukaemia Nature 2015 518 552 555 10.1038/nature13968 25487151
75. Kim YI Pogribny IP Basnakian AG Folate deficiency in rats induces DNA strand breaks and hypomethylation within the p53 tumor suppressor gene Am J Clin Nutr 1997 65 46 52 10.1093/ajcn/65.1.46 8988912
76. Hollstein M Sidransky D Vogelstein B Harris CC P53 mutations in human cancers Science 1991 253 49 53 10.1126/science.1905840 1905840
77. Gondek LP Sheedan VA Fitzhugh CD Clonal hematopoiesis and the risk of hematologic malignancies after curative therapies for sickle cell disease J Clin Med 2022 11 3160 10.3390/jcm11113160 35683547
78. Tolu SS Wang K Yan Z Characterization of hematopoiesis in sickle cell disease by prospective isolation of stem and progenitor cells Cells 2020 9 E2159 10.3390/cells9102159
79. Platt OS Sickle cell anemia as an inflammatory disease J Clin Invest 2000 106 337 338 10.1172/JCI10726 10930436
80. Snetsinger B Ferrone CK Rauh MJ Targeted, amplicon-based, next-generation sequencing to detect age-related clonal hematopoiesis Methods Mol Biol 2019 2045 167 180 10.1007/7651_2019_216 30941718
81. Pincez T Lee SSK Ilboudo Y Clonal hematopoiesis in sickle cell disease Blood 2021 138 2148 2152 10.1182/blood.2021011121 34521115
82. Liggett LA Cato LD Weinstock JS Clonal hematopoiesis in sickle cell disease J Clin Investig 2022 132 e156060 10.1172/JCI156060 34990411
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.