Local-Scale Health Impact Assessment of Ambient Black Carbon

Abstract

We conducted a health impact assessment to estimate the attributable risk of neighborhood-level ambient fine particulate matter black carbon (BC) exposure on mortality in an urban county in the United States. Using local-scale land use regression models of annual average BC and health effect associations reported in the literature, we estimated the potential impact of three scenarios on total non-accidental deaths and cardiopulmonary deaths by census tract. Scenario 1 estimated impacts assuming a reduction in median-level census tract BC to zero; Scenario 2 estimated the impacts of a reduction from median-level census tract BC concentration to the lowest estimated value for any census tract in the study area; Scenario 3 estimated health effects of removing on-road sources of BC. Annual average non-accidental mortality by census tract ranged from 1.5/1,000 – 41.9/1,000; cardiopulmonary mortality by tract ranged from 0.6/1,000 – 17.2/1,000. Scenario 1 would result in a reduction in total non-accidental deaths by 3.4% to 7.5% across census tracts; the reduction for cardiopulmonary mortality ranged from 1.27% to 2.73%. Scenario 2 resulted in a reduction of total non-accidental deaths by 0% to 4.21% across the census tracts; for cardiopulmonary deaths, the reduction ranged from 0 to 1.55%. Scenario 3 would reduce total non-accidental deaths by 0.10% to 2.20% across the census tracts; for cardiopulmonary deaths, the reduction ranged from 0.04% to 0.82% across census tracts. Median-level census tract BC was significantly and positively correlated with percent of census tract residents with less than a high school education (ρ=0.35, p<0.0001) and percent of Black residents (ρ=0.37, p<0.0001) and negatively associated with percent children under 14 (ρ=-0.24, p<0.0001). Though absolute annual numbers of deaths by census tract was small, local-scale HIA highlight inequalities in exposure and health-related endpoints that may inform community-based policy and planning.