Is it you or your neighborhood: what matters more when it comes to air pollution exposure?

Abstract

BACKGROUND AND AIM: Studies have examined individual and neighborhood social factors and air pollution independently; few have investigated which is more strongly associated with air pollution exposure. We leverage individual harmonized data from large observational cohorts to cross-sectionally interrogate this question. METHODS: Data came from three US cohorts at time of enrollment (1989-2007): Cardiovascular Health Study, Reasons for Geographic and Racial Differences in Stroke, and Multi-Ethnic Study of Atherosclerosis (ntotal = 39,860). Individual exposures include race and ethnicity and income (ISES). Neighborhood exposures include SES (NSES) and divergence, a spatial multi-race measure of racial residential segregation (RRs), both derived from 1990-2010 Census data and linked to participants’ residence. Annual averaged fine particulate matter (PM2.5 ) and nitrogen dioxide (NO2) were estimated from spatiotemporal models. Random intercept multilevel models estimated and compared associations between z-scores of individual and neighborhood social factors and unit changes in air pollution concentrations, adjusting for individual sex, age, cohort, and urbanicity. RESULTS: The mean (SD) air pollution exposure was 12.9 µg/m3 (3.55) for PM2.5 and 13.3 ppb (8.72) for NO2. We observed lower air pollution exposure among those with higher ISES (βISES-PM2.5= -0.05, 95% CI: -0.07, -0.03) and higher exposure among racial and ethnic minority participants. In fully adjusted models including individual and neighborhood-level variables, lower NSES and higher RRS were associated with higher air pollution (βNSES-PM2.5= -0.56, 95% CI: -0.61, -0.51; βRRS-PM2.5= 0.98, 95% CI: 0.83, 1.13 ), but ISES-PM2.5 and Black race-PM2.5 associations become null (βISES-PM2.5= -0.00, 95% CI: -0.02, 0.02; βBlack race-PM2.5= 0.03, 95% CI: -0.03, 0.9). CONCLUSION: Findings suggest the neighborhood social environment may be more correlated with air pollution burden than individual-level factors. Careful consideration should be made when considering how to analytically account for social factors in air pollution-health associations. KEYWORDS: air pollution, social environment, socioeconomic status, racial residential segregation