BMJ Open 11 (9) e048952 [2021-08-31; online 2021-08-31]
To evaluate the role of language proficiency and institutional awareness in explaining excess COVID-19 mortality among immigrants. Cohort study with follow-up between 12 March 2020 and 23 February 2021. Swedish register-based study on all residents in Sweden. 3 963 356 Swedish residents in co-residential unions who were 30 years of age or older and alive on 12 March 2020 and living in Sweden in December 2019. Cox regression models were conducted to assess the association between different constellations of immigrant-native couples (proxy for language proficiency and institutional awareness) and COVID-19 mortality and all other causes of deaths (2019 and 2020). Models were adjusted for relevant confounders. Compared with Swedish-Swedish couples (1.18 deaths per thousand person-years), both immigrants partnered with another immigrant and a native showed excess mortality for COVID-19 (HR 1.43; 95% CI 1.29 to 1.58 and HR 1.24; 95% CI 1.10 to 1.40, respectively), which translates to 1.37 and 1.28 deaths per thousand person-years. Moreover, similar results are found for natives partnered with an immigrant (HR 1.15; 95% CI 1.02 to 1.29), which translates to 1.29 deaths per thousand person-years. Further analysis shows that immigrants from both high-income and low-income and middle-income countries (LMIC) experience excess mortality also when partnered with a Swede. However, having a Swedish-born partner is only partially protective against COVID-19 mortality among immigrants from LMIC origins. Language barriers and/or poor institutional awareness are not major drivers for the excess mortality from COVID-19 among immigrants. Rather, our study provides suggestive evidence that excess mortality among immigrants is explained by differential exposure to the virus.