Chronic Ambient Air Pollution and Acute COVID-19 Disease Severity or Death among Confirmed Cases in Southern California

Michael Jerrett, MICHAEL KLEEMAN, Yiting Li, Yusheng Zhao, Anikender Kumar, Jason Su, Claudia Nau, Deborah Young, Rebecca Butler, Christina Batteate, Richard Burnett, University of California, Davis

     Abstract Number: 278
     Working Group: Aerosol Science of Infectious Diseases: What We Have Learned and Still Need to Know about Transmission, Prevention, and the One Health Concept

Abstract
A growing body of evidence links ambient air pollution to COVID-19 incidence, severity, and death, but many studies have important limitations: (i) limited ability to control for important confounders, (ii) relatively simplistic exposure assessments, (iii) limited exploration of other environmental variables such as temperature and humidity. It is desirable to address these limitations to fully understand whether air pollution exposure worsens the prognosis for patients infected with COVID-19.

Here we analyzed air pollution effects on COVID-19 outcomes using a retrospective cohort of 316,224 patients who were diagnosed with, or tested positive for COVID-19 during 06/01/2020 up to 1/30/2021. The cohort was created using electronic health records data (EHR) from Kaiser Permanente Southern California (KPSC). Chronic air pollution exposure fields were developed across the study region using chemical transport models (CTMs) and Land Use Regression Models (LURMs) with a spatial resolution of 1 km. The exposure fields included PM2.5 mass, PM0.1 mass, and several physio-chemical constituents of PM2.5, source tracers as well as ozone and NO2. We used Cox proportional hazards models to study relationships between air pollution and COVID-19 outcomes with adjustment for potential individual and neighborhood variables. All models were stratified at baseline for age, sex and race-ethnicity. We controlled for potential non-independence at the census tract level. We conducted several subgroup analyses and examined the dose-response curves for the pollutants with significant effects.

Results indicate significant air pollution effects on the risk of death for PM2.5 mass, PM0.1 mass, and several of the particle species or source tracers, including PM2.5 nitrates, PM2.5 elemental carbon, PM2.5 on-road diesel, and PM2.5 on-road gasoline. Effects for PM2.5 mass were only slightly reduced in the multipollutant models. Lower temperatures and lower humidity were associated with higher risks for COVID-19 death. Both of these variables were selected as confounders in all air pollution models. Based on these results, we conclude that there is a high likelihood that chronic ambient air pollution exposures contribute to a higher risk of death within 90 days of COVID-19 infection. Such findings may lead to a re-evaluation of whether current air pollution standards need to be tightened as another measure to protect people from the adverse effects of COVID-19 and from future pandemics.