Residential Segregation and Transportation-Related Air Pollution Exposure Disparities in United States Urban Areas, 2000-2010

Abstract

Background: Residential segregation has been proposed as a factor contributing to environmental health disparities by race-ethnicity in the United States (US). This study investigates relationships between residential segregation and transportation-related air pollution (TRAP) exposure disparities across US urban areas and over time.Methods: We combined estimates of a TRAP (nitrogen dioxide [NO2] annual average concentrations from a national regression model) with Census demographic data for urban areas in the contiguous United States (n= 481) in 2000 and 2010. We estimated two aspects of differences between race-ethnicity minorities and non-Hispanic whites: disparities in NO2 air pollution exposure and segregation in residential location (dissimilarity index). We constructed linear regression models relating those two aspects (racial disparities in exposure; residential racial segregation), controlling for racial disparities in urbanicity, racial economic disparities, racial diversity, population, pollution, and region.Results: After controlling for other urban characteristics, residential segregation is positively associated (p<0.001) with racial disparity in NO2 exposure in 2000 and in 2010, and change in residential segregation over time is positively associated (p<0.001) with change in racial disparity in NO2 exposure (adjusted R2: 0.65 [2000], 0.66 [2010], 0.38 [change: 2000-2010]). Increasing segregation across the interquartile range (holding all other variables constant at mean value) is associated with a 30% (0.2 ppb) increase in racial disparity in NO2 exposure in 2000 and a 40% (0.2 ppb) increase in 2010.Conclusions: Higher levels of racial segregation were associated with larger racial disparities in TRAP exposure in 2000 and in 2010. Larger reductions in segregation over time (2000 to 2010) were associated with larger reductions in racial disparities in TRAP exposure.