Background Studies have shown that COVID-19 disproportinately affected individuals based on their ethnic groups and occupation types. This study investigated the extent to which ethnic inequalities in COVID-19 hospitalisation or deaths could be reduced by eliminating differences by occupational risks.

Methods We used a population-based cohort study linking the 2011 Scottish Census to health records. We included all individuals aged 30 - 65 years and living in Scotland on 1 March 2020. The study period was from 1 March 2020 to 17 April 2022. Self-reported ethnic group was taken from the census. We derived occupational risk based on the extent of SARS-CoV-2 exposure using the 3-digit Standard Occupational Classification (SOC2010). We estimated controlled direct effects of ethnicity on COVID-19 hospitalisations or deaths that was not mediated by occupation using inverse probability weighted marginal structural models, adjusting for occupation and several confounders.

Results Of the 2, 064, 982 individuals analysed, 11, 055 (0.5%) had COVID-19 related hospitalisations or deaths. The total effect (TE) of Non White ethnic group had 50% higher risk of COVID-19 hospitalisation or deaths (HR 1.5 (95% CI 1.5 -1.8) compared to White group Compared with White ethnic group, the controlled direct effect (CDE) of Non-White ethnicity on severe COVID-19 was 1.7(1.4 – 2.1) if occupational risk differences were eliminated, equating to an increase of 6.1 % (95%CI -6.2%, - 6.0%) in relative ethnic inequalities. For disaggregated ethnicity, the TE for South Asian 2.0(0.1-2.3), African, Caribbean, or Black 1.3(0.2-2.0) and Other 1.1(0.1-0.9) was higher compared to White Scottish. For the CDE, the risk of severe COVID-19 was elevated for South Asian 1.8 (0.2-3.1), African, Caribbean, or Black 1.4 (0.3-1.1) and Other ethnic group 1.7 (0.2-2.5) compared to White Scottish. Consequently, the relative ethnic inequalities reduced by 11.8% (10.7% - 13.8%) for South Asian, 19.0% (18.0 - 19.9%) for Other White. However, relative ethnic inequalities increased by 16.4% (19.0% - 13.8%) for African, Caribbean, or Black and 59.0% (31.3% - 56.7%) for other ethnic group if occupation risk differences were eliminated.

Conclusion Our findings suggest that ethnicity minority status directly influences severe COVID-19 in Scotland, after elimination of occupational risk differences. In some cases, the elevated risk of severe COVID-19 outcomes was greater after blocking the occupation pathway; this is because some ethnic minority groups were less likely to work in jobs with high-risk exposure