You are here:
Do factors related to combustion-based sources explain heterogeneity in PM-mortality associations across the United States?
Alman, B., K. Rappazzo, J. Reyes, AND L. Neas. Do factors related to combustion-based sources explain heterogeneity in PM-mortality associations across the United States? International Society for Environmental Epidemiology Annual Conference, Rome, ITALY, September 01 - 04, 2016.
This study uses rate ratios generated from the Multi City Multi Pollutant study (1999-2005) for 313 core-based statistical areas (CBSA) and their metropolitan divisions (MD) to examine combustion-based sources of heterogeneity.
Introduction: Spatial heterogeneity of effect estimates in associations between PM2.5 and total non-accidental mortality (TNA) in the United States (US), is an issue in epidemiology. This study uses rate ratios generated from the Multi-City/Multi-Pollutant study (1999-2005) for 313 core-based statistical areas (CBSA) and their metropolitan divisions (MD) to examine combustion-based sources of heterogeneity.Methods: For CBSA/MDs, area-specific log rate ratios (betas) were derived from a model adjusting for time, an interaction with age-group, day of week, and natural splines of current temperature, current dew point, and unconstrained temperature at lags 1, 2, and 3. We assessed the heterogeneity in the betas by linear regression with inverse variance weights, using average NO2, SO2, and CO, which may act as a combustion source proxy, and these pollutants’ correlations with PM2.5. Results: We found that weighted mean PM2.5 association (0.96 percent increase in total non-accidental mortality for a 10 µg/m3 increment in PM2.5) increased by 0.26 (95% confidence interval 0.08 , 0.44) for an interquartile change (0.2) in the correlation of SO2 and PM2.5., but betas showed less dependence on the annual averages of SO2 or NO2. Spline analyses suggest departures from linearity, particularly in a model that examined correlations between PM2.5 and CO.Conclusions: We conclude that correlations between SO2 and PM2.5 as an indicator of combustion sources explains some heterogeneity in PM2.5’s association with mortality in the US. PM2.5 and TNA associations may be modified by proximity to combustion sources. Future research will include point source and roadway data to further elucidate this relationship. This an abstract of a presentation and does not necessarily reflect the policies of the US Environmental Protection Agency.