Racial/ethnic disparities in nationwide PM2.5 exposure: perils of assuming a linear relationship

Abstract

BACKGROUND AND AIM Studies of racial/ethnic disparities in air pollution exposure assume an equal magnitude of disparity across the demographic distribution, which may under- or over-estimate the true disparity. Further, historic practices such as racist zoning policies could plausibly drive non-linear relationships between racial/ethnic composition and air pollution exposure. We aimed to conduct a nationwide analysis in urban areas allowing for non-linearity in the relationship between racial/ethnic composition and ambient particulate matter less than 2.5 micrometers (PM₂.₅), and test whether racial segregation modified this relationship. METHODS We obtained tract-level demographic information from the 2010 Census and computed the multi-group dissimilarity index, a metropolitan-area level measure of racial segregation. Ambient PM₂.₅ data was obtained from a publicly available source. We evaluated the relationship between percent non-Hispanic Black and percent non-Hispanic White residents and ambient PM₂.₅ using natural splines. RESULTS In 58,029 U.S. urban census tracts in 2010, we observed first a very steep and then a plateauing relationship between percent non-Hispanic Black residents and ambient PM₂.₅ concentration. Moving from the 25th percentile (1%) to the median (5%) of non-Hispanic Black residents was associated with a 0.58 μg/m³ (95% CI: 0.56, 0.60) increase in PM₂.₅, while moving from the median to the 75th percentile (16% non-Hispanic Black) was associated with a 0.005 μg/m³ (95% CI: -0.012, 0.020) increase. Reciprocally, in predominantly White census tracts, increases in the proportion of White residents was associated with a rapid decrease in ambient PM₂.₅. Segregation modified these relationships in a non-monotonic manner. CONCLUSIONS Our results suggest larger than previously reported Black-White disparities in exposure to PM₂.₅, especially in communities with a modest Black population, which make up the majority of urban census tracts in the US. KEYWORDS Racial/ethnic disparities; Air pollution; Particulate Matter; Environmental justice