Cardiopulmonary Mortality and Fine Particulate Air Pollution by Species and Source in a National US Cohort

The purpose of this study was to estimatecardiopulmonary mortality associations for long-term exposure toPM2.5species and sources (i.e., components) within the U.S. NationalHealth Interview Survey cohort. Exposures were estimated through achemical transport model for six species (i.e., elemental carbon (EC),primary organic aerosols (POA), secondary organic aerosols (SOA),sulfate (SO4), ammonium (NH4), nitrate (NO3)) andfive sources ofPM2.5(i.e., vehicles, electricity-generating units (EGU), non-EGUindustrial sources, biogenic sources (bio),"other"sources). In single-pollutant models, we found positive, significant (p< 0.05) mortalityassociations for all components, except POA. After adjusting forremaining PM2.5(total PM2.5minus component), we found significantmortality associations for EC (hazard ratio (HR) = 1.36; 95% CI[1.12, 1.64]), SOA (HR = 1.11; 95% CI [1.05, 1.17]), and vehicle sources (HR = 1.06; 95% CI [1.03, 1.10]). HRs for EC, SOA, andvehicle sources were significantly larger in comparison to those for remaining PM2.5(per unit mu g/m3). Ourfindings suggest thatcardiopulmonary mortality associations vary by species and source, with evidence that EC, SOA, and vehicle sources are importantcontributors to the PM2.5mortality relationship. With further validation, thesefindings could facilitate targeted pollution regulationsthat more efficiently reduce air pollution mortality.