key: cord-0307983-3ofuf398
authors: Santa-Ramirez, H. A.; Wisniak, A.; Pullen, N.; Zaballa, M. E.; Pennacchio, F.; Lorthe, E.; Dumont, R.; Baysson, H.; Guessous, I.; Stringhini, S.
title: Socio-economic determinants of SARS-CoV-2 infection: results from a population-based serosurvey in Geneva, Switzerland
date: 2022-05-10
journal: nan
DOI: 10.1101/2022.05.09.22274838
sha: 93f226a99f33456a38fd3a46b5ebb5618fff93ef
doc_id: 307983
cord_uid: 3ofuf398

Background SARS-CoV-2 infection and its health consequences have disproportionally affected disadvantaged socio-economic groups globally. This study aimed to analyze the association between socio-economic conditions and having developed anti-SARS-CoV-2 antibodies in a population-based sample in the canton of Geneva, Switzerland. Methods Data was obtained from a population-based serosurvey of adults in Geneva and their household members, between November and December, 2020, towards the end of the second pandemic wave in the canton. Participants were tested for anti-SARS-CoV-2 antibodies. Socio-economic conditions representing different dimensions were self-reported. Mixed effects logistic regressions were conducted for each predictor to test its association with seropositive status as the main outcome. Results 2,889 adults completed the study questionnaire and were included in the final analysis. Retired participants and those living in suburban areas had lower odds of a seropositive result when compared to employed participants (OR 0.42, 95% CI - 0.20 - 0.87) and those living in urban areas (OR 0.67, 95% CI - 0.46 - 0.97), respectively. People facing financial hardship for less than a year had higher odds of a seropositive result compared to those who had never faced them (OR 2.23, 95% CI - 1.01 - 4.95). Educational level, occupational position and household income were not associated with being seropositive, nor were ethnicity or country of birth. Discussion While traditional measures of socio-economic position did not seem to be related to the risk of being infected in this sample, this study sheds lights on the importance of examining the broader social determinants of health when evaluating the differential impact of the pandemic within the population.

Background 20 SARS-CoV-2 infection and its health consequences have disproportionally affected disadvantaged 21 socio-economic groups globally. This study aimed to analyze the association between socio-economic 22

conditions and having developed anti-SARS-CoV-2 antibodies in a population-based sample in the 23 canton of Geneva, Switzerland. 24 Methods 25

Data was obtained from a population-based serosurvey of adults in Geneva and their household 26 members, between November and December, 2020, towards the end of the second pandemic wave in 27

the canton. Participants were tested for anti-SARS-CoV-2 antibodies. Socio-economic conditions 28

representing different dimensions were self-reported. Mixed effects logistic regressions were 29 conducted for each predictor to test its association with seropositive status as the main outcome. 30

Results 31 2,889 adults completed the study questionnaire and were included in the final analysis. Retired 32 participants and those living in suburban areas had lower odds of a seropositive result when compared 33

to employed participants (OR 0.42, 95% CI -0.20 -0.87) and those living in urban areas (OR 0.67, 34 95% CI -0.46 -0.97), respectively. People facing financial hardship for less than a year had higher 35

odds of a seropositive result compared to those who had never faced them (OR 2.23, 95% CI -1.01 -36 4.95). Educational level, occupational position and household income were not associated with being 37 seropositive, nor were ethnicity or country of birth. 38

Since the beginning of the COVID-19 pandemic, studies have shown that SARS-COV-2 infection and 46

its health-related consequences have disproportionally affected disadvantaged socio-economic groups 47

(1-3). Disadvantaged populations accumulate several vulnerabilities to infection, such as poor living 48

conditions, higher job instability, fewer job opportunities, poorer social benefits and lower financial 49 security (4,5), household crowding and possible impairments of their immune status due, among others, 50

to work-related and financial stress (6). This may lead to a higher need of continued work outside the 51 home, particularly for essential workers. Socioeconomically disadvantaged populations are also known 52

to have a higher burden of chronic diseases and reduced access to healthcare (7), both risk factors for 53 COVID-19 severity (8). In New York City, underprivileged neighborhoods, neighborhoods with higher 54 household density, and those with higher proportions of black and immigrant populations were more 55 likely to have a positive COVID-19 test result (9). An analysis of data reported to the Swiss Federal 56

Office of Public Health (SFOPH) during the first year of the pandemic revealed that people living in 57 neighborhoods with a low socioeconomic position index were less likely to get tested, but had a higher 58

proportion of positive SARS-CoV-2 RT-PCR and antigen test results and were more likely to be 59 hospitalized or die compared to people living in socioeconomically advantaged areas (10). Another 60 study has also shown persistence of SARS-CoV-2 clusters in more disadvantaged neighborhoods, 61 when analyzing RT-PCR positive test results (11). Several studies revealing social inequalities related 62

to COVID-19 have been based on confirmed RT-PCR test results, therefore missing a large part of the 63 population who did not undergo testing (12,13). Socio-economic conditions may also influence the 64 probability of getting tested when presenting with symptoms of COVID-19. A better picture of the 65 distribution of the infection in the population is achieved with serological surveys as they yield more 66 accurate estimations of the real number of infections including mild and asymptomatic cases (14). 67

Further, most studies rely on area-based indicators of socioeconomic status, thereby not allowing a 68 more precise characterization of factors associated with SARS-CoV-2 infection. 69

Previous work by our research team showed associations between employment status and 70 seropositivity during the first wave of the epidemic in the canton of Geneva, with retirees having lower 71

odds of a seropositive result, and found no association with education, occupational status and 72 neighborhood median income (15). A serological survey conducted among essential workers in Geneva 73 after the first epidemic wave showed significant variation in seroprevalence across occupations (16). 74

Nevertheless, other features that might influence serological status could not be assessed in those 75 studies, such as ethnicity, individual income, country of birth and living and residential conditions. 76

Although the canton of Geneva never followed a strict lockdown, there were some differences between 77 the first and second waves, with the relaxation of certain measures such as re-opening of primary 78 schools, as well as shops and establishments, and allowing larger social gatherings. During the second 79

wave, a more strict use of facemasks was mandated and tests were made available free of charge to 80 any person with symptoms. 81 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 10, 2022. ; https://doi.org/10.1101/2022.05.09.22274838 doi: medRxiv preprint Understanding the influence of socioeconomic conditions on the probability of being infected with 82 SARS-CoV-2 is crucial for the implementation of equity-driven public health measures both to contain 83 the spread of the virus during the pandemic phase and to structure the public health response in the 84 post-pandemic phase. This study aimed to analyze the association between socioeconomic conditions 85 and having developed anti-SARS-CoV-2 antibodies during the second COVID-19 wave (October-86

December 2020) in a representative sample of the population in the canton of Geneva. 87

Methods 88

The study sample included adults aged 18 years and older, who were randomly selected from a previous 89 population-based serosurvey conducted in the canton of Geneva in spring 2020, and from population 90 registries of the canton. Household members of recruited participants were also invited to take part in 91 the study. Details of the selection process have been described previously (17 Results 119 A total of 2,986 adults participated in the study and had a blood sample taken, of which 2,889 120 completed the study questionnaire and were included in the final analysis. The mean (SD) age of 121 participants was 47.8 (15.4) years, and 55% were women. Education, occupation and income were not 122 associated with being seropositive in the overall sample (Table 1 and Annex II. Supplementary  123 material). Looking at other socioeconomic indicators, associations were found with employment status, 124 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 10, 2022. In this population-based serological study, we found associations between financial hardship, 138 employment status, residential area and the odds of having developed anti-SARS-CoV-2 antibodies. A 139 higher household density tended to be associated with increased odds of a seropositive result. However, 140

other socioeconomic conditions such as educational level, occupational position and income were not 141 associated with serological status, nor were ethnicity and country of birth. 142

Our study shows lower odds of a SARS-CoV-2 seropositive result for the retired population when 143 compared with the employed one, possibly due to the fact of being considered at higher risk of severe 144

forms of COVID-19, potentially leading them to reduce social contact and increase the use of 145 preventive measures. This result is consistent with previous findings from the first seroprevalence study 146

in Geneva (15) and findings from a report in the UK for the age group comprising the retired population 147 (20). 148

We also found a protective effect of the residential area for people living in suburban areas compared 149 to urban areas, which could be explained by increased use of private transportation and lower 150 population density. While this may also be the case in rural areas, higher commuting times and a 151 potentially lower sense of danger posed by the infection in these areas may explain the lack of 152 significant difference in seropositivity between rural and urban areas. It has been suggested that a lower 153 population density outside the urban areas might have contributed to lower incidence at the beginning 154 of the pandemic in some regions in Europe (21) and some studies have shown lower seroprevalence in 155 municipalities of less than 100.000 inhabitants (22). Further work is needed to uncover the potential 156 mechanisms explaining the association of the residential area with a seropositive result in the 157 population of Geneva, as considering the small size of the canton, the difference between urban and 158 suburban areas is not clearly established and the distribution of SARS-CoV-2 infections might not 159 follow a similar pattern as the one found in other places. 160

There seemed to be a trend in the association between duration of financial difficulties and the odds of 161 seropositivity, with people facing financial hardship for less than one year having the highest odds of 162 a seropositive result compared to those who reported never facing financial hardship. This could 163 potentially be explained by the development of coping mechanisms in individuals being used to 164 financial difficulties, while those with unexpected economic hardship may need more time to adapt to 165 their new circumstances, putting them at higher risk of SARS-CoV-2 exposure as they cannot afford 166

to miss work or need to look for economic alternatives. A consistent association of financial hardship 167 due to COVID-19 with health behavior risk changes has been shown in a sample of women in the U.S., 168 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 10, 2022. showing conflicting results (20,25-27). This may be due to differences in survey design and 178 measurement across studies. The association between education and seropositivity may be confounded 179 by increased SARS-CoV-2 exposure in certain professions requiring tertiary education, such as in the 180 health-related field. Future analyses should take into account professional exposure to SARS-CoV-2. 181

Heterogeneity in the socioeconomic circumstances in different countries, as well as diverging policies 182

for pandemic management, may also explain some of the conflicting results. 183

Strengths of this study include the relatively large sample size and comprehensive information related 184 to different social and economic circumstances at the individual level as well as objective information 185

about individual health such as the serological status. Our study also has some limitations. A selection 186 bias should not be disregarded, with people with higher health concerns being more prone to 187 participate, and those most socioeconomically disadvantaged less likely to be included, limiting the 188 generalizability of our results. In addition, the population that was hospitalized at the time of the study 189 or that died because of COVID-19 could not be included in the study, therefore potentially masking 190 the association between socio-economic conditions and SARS-CoV-2 seropositivity for severe cases. 191

As other studies have documented, the severity of the disease might be higher in socioeconomically 192 disadvantaged groups (4,28). 193

The COVID-19 pandemic has disproportionately affected socially vulnerable populations globally. 194

However the impact of socio-economic determinants can vary widely depending on geographical, 195 political and cultural contexts (29-31). In our study we have found associations of employment status, 196 financial hardship and residential area with the natural development of anti SARS-CoV-2 antibodies 197 during the second wave of the pandemic (before the roll-out of the vaccination campaign in 198 Switzerland); but not with other socioeconomic conditions. Our results highlight the importance of 199 examining the broader social determinants of health when evaluating the differential impact of the 200 pandemic within the population. A better understanding of the structural determinants shaping the 201 inequitable distribution of COVID-19 among the population is imperative for tailoring public health 202

interventions, such as vaccine prioritization and public health campaigns, and for setting up supportive 203 mechanisms for vulnerable population groups. 204

The authors declare that the research was conducted in the absence of any commercial or financial 206 relationships that could be construed as a potential conflict of interest. Norwegian University of Science and Technology (NTNU). 236

We thank all the participants, without whom this study would not have been possible 238 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 10, 2022. Seroprevalence of anti-SARS-CoV-2 IgG antibodies, risk factors for infection and associated 286 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 10, 2022. ; https://doi.org/10.1101/2022.05.09.22274838 doi: medRxiv preprint The supplementary material for this article can be found in the additional document 365 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 10, 2022. ; https://doi.org/10.1101/2022.05.09.22274838 doi: medRxiv preprint

a population-based study: Scand J Public Health

Large 290 variation in anti-SARS-CoV-2 antibody prevalence among essential workers in Geneva

Seroprevalence of anti-SARS-CoV-2 antibodies after the second pandemic peak

Male 296 sex identified by global COVID-19 meta-analysis as a risk factor for death and ITU admission

SARS-CoV-2 antibody 299 prevalence in England following the first peak of the pandemic

Demographic and territorial characteristics of COVID-302 19 cases and excess mortality in the European Union during the first wave

305 Prevalence of SARS-CoV-2 in Spain (ENE-COVID): a nationwide, population-based 306 seroepidemiological study. The Lancet

Financial hardship 308 and health risk behavior during COVID-19 in a large US national sample of women. SSM -309 Popul Health

Seroprevalence and correlates of SARS-CoV-2 neutralizing antibodies from a population-based 312 study in Bonn

Estimates and 314 Determinants of SARS-Cov-2 Seroprevalence and Infection Fatality Ratio Using Latent Class 315 Analysis: The Population-Based Tirschenreuth Study in the Hardest-Hit German County in 316 Spring 2020. Viruses

Socioeconomic 318 inequalities in the prevalence and perceived dangerousness of SARS-CoV-2 infections in two 319 early German hotspots: findings from a seroepidemiological study

Global seroprevalence of 322 SARS-CoV-2 antibodies: A systematic review and meta-analysis

Hospitalization and Mortality among Black 325 Patients and White Patients with Covid-19

Socio-economic status and COVID-19-related cases 327 and fatalities. Public Health

Revealing the Unequal Burden of COVID-19 by Income, Race/Ethnicity, 329 and Household Crowding: US County Versus Zip Code Analyses

):e2021892. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity