(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Income differences in COVID-19 incidence and severity in Finland among people with foreign and native background: A population-based cohort study of individuals nested within households [1]

['Sanni Saarinen', 'Population Research Unit', 'Faculty Of Social Sciences', 'University Of Helsinki', 'Helsinki', 'Heta Moustgaard', 'Helsinki Institute For Social Sciences', 'Humanities', 'Hanna Remes', 'Riikka Sallinen']

Date: 2022-08 

We used total population data covering over 4 million individuals nested within households to assess the associations of household income with COVID-19 incidence and severity, and to quantify the clustering of COVID-19 in working-age households. In line with prior evidence, we found that individuals living in low-income households had higher risk of both COVID-19 incidence and severity—however, this was largely driven by the foreign background of household members. In fact, a separate analysis revealed a strong income gradient among individuals with foreign background only, and no income association at all among those with native background. The odds of severe illness among the infected were likewise highest among those with lowest income, but this association was also strongly driven by other risk factors: comorbidities, personal occupational status, and foreign background. Both incidence and severity were strongly clustered in households: Around 77% of the variation in incidence and 20% of the variation in severity were attributable to differences between households.

Comparison with other studies

To our knowledge, this is the first study to assess income and other sociodemographic risk factors for COVID-19 incidence and severity using both household and individual-level data. Incorporating the household level is a major contribution, because, as we show, not only infections, but also severe illness is strongly clustered within households.

Our general finding of a higher COVID-19 incidence in low-income households is in accordance with previous results associating incidence with area-level measures of low education, deprivation [1,6], and low mean income by district [2,3]. Our study differed from many others in that we were able to control for a set of other important individual- and household-level risk factors, and indeed we found that having foreign background was a major driver of the income differences, while other risk factors for higher COVID-19 incidence made no difference for the income association. Furthermore, we showed that the income gradient was only present among people with foreign background and nonexistent among those with a Finnish background. New to the existing literature, this finding has no direct point of reference from previous studies. While foreign background and ethnicity have been linked to higher COVID-19 incidence and severity in many previous studies [18], to our knowledge, only 1 prior study has assessed the role of foreign background or ethnicity in the association between income and COVID-19 outcomes [1]. Their results contrast with ours: In this study based on UK Biobank data, controlling for ethnicity and country of birth explained little of the higher incidence in socioeconomically deprived areas [1]. The differing results may relate to the selective sample and area-based measurement of socioeconomic exposures in the UK Biobank data, compared with our total population data with household-based measures.

Our results suggest that, overall, there is no independent association between household income and COVID-19 incidence. Low income was related to higher COVID-19 incidence only among people with foreign background, and this association was independent of household size and the work and school exposures that we measured. The combination of foreign background and low income is likely to capture vulnerabilities related to race, ethnic minority, and refugee background which may influence infection risk through material, social, and behavioural mechanisms. Prior studies from the United States indicate that people with low income have fewer material and social resources to protect themselves from COVID-19 infection [27]. Social mechanisms may relate to lower health literacy, language barriers, racism, and structural discrimination faced by people with foreign background and low socioeconomic position [28–30]. Higher susceptibility to COVID-19 may also relate to a lowered immune response due to higher stress levels [14,31]. Finally, behavioural factors such as possible broader social interactions among people with foreign background [29] may further add to the clustering of risk factors among those with foreign background and low income. Due to the epidemic nature of COVID-19, the extent of social interactions could be of particular importance if outbreaks are concentrated in schools and neighborhoods having more people with low socioeconomic position and foreign background. The lack of neighborhood-level controls is a limitation of our study.

A major contribution of our study is that we were able to assess the risk factors for case severity in the total infected population. Previous studies on severe COVID-19 have either limited the study population to those already hospitalized or seeking health care [10,11,19], or analysed the general population irrespective of infection status [7–9]. The former target a group that is already selected on case severity whereas the latter conflate the risk factors of incidence and severity. A study on the general population of Sweden, for example, found that low income and educational level predicted increased COVID-19 mortality [7], but these estimates may reflect socioeconomic differences in either incidence or case fatality, or both. Moreover, few previous studies have been able to control for important confounders. A Swiss study reported higher case severity and fatality rates among the infected in neighbourhoods with a low socioeconomic index based on 2,000 census data, but it did not adjust for individual risk factors other than age and sex [12]. In contrast, our adjusted results indicate that household income is not independently associated with case severity. In fact, the higher risk of severe COVID-19 in low-income households was strongly driven by personal occupational status and foreign background. Personal occupational status is likely to capture health-related confounding by controlling for being on early-age pension or having an unknown personal occupational status. Foreign background, in turn, may capture ethnic differences in the risk for severe COVID-19, which could relate to a complex set of factors such as racism and structural discrimination, or barriers in access to care that we were not able to measure [18,32]. Another reason for our results may relate to selection bias caused by differential testing [33]. If people with foreign background tend to test less for mild COVID-19 symptoms, this will lead to a disproportional share of severe cases among those identified as infected in this group. Further studies should investigate the mechanisms producing social and ethnic differences in case severity, preferably in samples where infections are identified by screening rather than by self-selective testing.

We found very high infection clustering within households, corroborating previous evidence that the risk of COVID-19 infection is higher within households than in other social contexts [21,22]. Likewise, our finding that large low-income households had the highest incidence implies an accumulation of risk factors in specific types of household, and supports the earlier observation on the disadvantages of household crowding [5]. Our results further suggest that this accumulation mainly occurs among the population with foreign background. Our study is the first to assess the household clustering of case severity. The correlation in the likelihood of severe illness between household members was around 20%, which the measured individual-level risk factors failed to explain. However, unmeasured risk factors such as general health, obesity, smoking, and other health behaviours may cluster in households and offer some explanation of why several members of some households tend to have severe COVID-19. Shared genetic vulnerability is unlikely to explain much of the household clustering of severity because the vast majority of multigenerational households in our data consist of parents and their underage children, and severe COVID-19 is rare among the young.

[END]
---
[1] Url: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004038

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/