Longer- and Shorter-term Air Pollution Exposure Associated with COVID-19 Severity and Mortality: A Large Cohort Study in Southern California

Abstract

BACKGROUND AND AIM: A growing number of ecological studies have suggested that air pollution contributes significantly to the COVID-19 incidence and mortality. To address the limitations of ecological analysis, this study aims to investigate the associations of longer- and shorter-term air pollution exposures with COVID-19 severity and mortality and explore effect modification by social-demographics using individual-level data from a large multi-ethnic cohort. METHODS: The retrospective cohort includes 75,010 individuals (mean age 42.5 years, 54% female, 66% Hispanic) diagnosed with COVID-19 at Kaiser Permanente Southern California between 3/1/2020-8/31/2020. Ambient air pollutant (PM2.5, NO2 and O3) exposures prior to the COVID-19 diagnosis date were estimated based on residential address history using inverse distance-squared weighted averages of central monitor data. Longer-term exposure was defined as 1-year and 4-year averages; shorter-term exposure was defined as 1-week and 1-month averages prior to COVID-19 diagnosis. Primary outcomes were COVID-19 severity defined as COVID-19-related hospitalizations, intensive respiratory support, intensive care unit admissions within 30 days, mortality within 60 days after COVID-19 diagnosis. RESULTS:Longer-term PM2.5 and shorter-term NO2 exposures were significantly associated with COVID-19 severity and mortality in multi-pollutant models adjusted for confounders. One standard deviation (SD) increase in 1-year PM2.5 exposure was associated with a 24-38% (p.0001) increased odds for severity and a 17% (p.01) increased hazard for mortality. Prior 4-year PM2.5 exposure was also significantly associated with COVID-19 severity. One SD increase in 1-month NO2 exposure was associated with 12-24% (p.0001) increased odds for severity and 14% (p.05) increased hazard for mortality. Prior 1-week NO2 exposure was also significantly associated with COVID-19 outcomes. No significant interactions with sex or ethnicity were observed; however, significant interaction with 1-month PM2.5 exposure was observed for larger associations in individuals age 65+ years. CONCLUSIONS:Data from this large cohort showed that ambient air pollutants are important factors affecting COVID-19 severity and mortality. KEYWORDS: PM2.5, NO2, COVID-19, short-term, long-term, Air Pollution