The use of improved exposure factors in the interpretation of fine particulate matter epidemiological results

Multi-city population-based epidemiological studies have consistently reported a significant association between ambient fine particulate matter (PM2.5) concentrations and daily mortality. However, in these studies heterogeneity between-community effect estimates is often observed but not thoroughly examined, leaving much of the difference in the effects of individual communities inadequately explained. In this study, we evaluated whether community-specific exposure factors play a role in explaining heterogeneity in the associations between ambient PM2.5 concentrations and several causes of mortality in 27 US communities from 1997 to 2002 as reported by Franklin et al. (J Expo Sci Environ Epidemiol 3:279-287, 2007). Using publically available databases, we created factors capturing home ventilation characteristics and commuting patterns. These factors include the normalized leakage, annual and seasonal temperatures, and in-vehicle commuting distances and time. In-vehicle commuting distance and time, and annual, spring, and fall temperatures were significant negative effect modifiers of the relationship between PM2.5 exposure and respiratory and non-accidental mortality. Additionally, cardiovascular mortality PM2.5 effect estimates were negatively modified by in-vehicle commuting distances. We concluded that future multi-community studies of particle health effects should consider these and other determinants of personal-ambient exposure relationships during the epidemiological analysis.