Abstract
BACKGROUND: To explore the syndemic nature of the COVID-19 pandemic by identifying which subpopulations in the United States (US) suffered the greatest losses in life expectancy (LE) in 2020 and 2021, and to which extent these losses can be attributed to COVID-19 and 'other' causes of death. METHODS: We analysed individual death records for 2018-2021 from the National Vital Statistics System and population counts from the American Community Survey. Life table and continuous change decomposition analyses were used to quantify cause-/.specific contributions to changes in LE over time in population subgroups defined by sex, educational attainment, and race/ethnicity. RESULTS: From 2019 to 2020, educational differences in LE (high minus low education) increased substantially by 5.0 and 2.6 years in Hispanic men and women, respectively, with increases of one to two years among Black and White adults. Nearly all losses in LE among high-education Hispanic and White groups were due to COVID-19, while among low-education White and Black groups, COVID-19 accounted for 40%-47% of the total losses in LE. Changes in LE were much smaller during 2020-2021. CONCLUSIONS: COVID-19 widened preexisting inequalities in LE in the US, both via direct mortality and through syndemic interactions with other diseases and health conditions. The underlying social, political, economic, and environmental factors driving the clustering and interaction of diseases among the most disadvantaged Americans need to be addressed.