Modeling disparities in ambient air pollution exposure and residential air exchange rates across Massachusetts using publicly-available data

摘要

Housing is an important social determinant of health. Individual housing characteristics can modify outdoor ambient air pollution infiltration into the home environment through air exchange rate (AER). Due to time and labor-intensive methods needed to measure AER, few studies have characterized AER distributions and the implications for exposure disparities across large geographic areas. Using publicly available housing data and regression models associating AER with housing characteristics, we estimated AER for all Massachusetts residential parcels. We then conducted a disparities analysis of temporally- and spatially-resolved ambient PM_(2.5) exposures, and residential AERs at an address-level. Across the state, mean ambient PM_(2.5)(μg/m~3) concentrations in 2010 were 6.7 (range: 0.01-85.1). Median AERs (h1) with closed windows for winter and summer were 0.44 (IQR: 0.27-.65) and 0.26 (IQR: 0.16-.38), respectively. Overall, duplex and triplex homes, small apartment buildings and large apartment buildings had 2.$32.8, and 2.0 times higher median AERs than single family homes, respectively. Housing parcels in the 90th percentile distribution of both AER and PM_(2.5) (i.e. the leakiest homes in areas of highest ambient air pollution) - versus the 10th percentile - were located in neighborhoods with higher proportions of ethnic/racial minorities (Hispanic 19.2% vs 2.1%), households with an annual income of less than $20,000 (26.0% vs. 7.7%) and low educational attainment populations (23.4% vs. 5.9% with less than a high school degree). We demonstrated a novel application of empirical AER models with high-resolution data. This approach can be applied in epidemiological studies to develop potential exposure modifiers, or to characterize exposure inequalities that are not solely based on ambient concentrations. This work additionally highlights the importance of considering both neighborhood- and housing-level factors as drivers of inequitable ambient air pollutant exposure.