Age-specific seasonal associations between acute exposure to PM2.5 sources and cardiorespiratory hospital admissions in California

Abstract Numerous studies have explored the relationships between short-term exposure to fine particulate matter (PM2.5) and morbidity. However, few studies have investigated which PM2.5 sources and constituents contribute to the health associations, and even fewer studies are available which explored age or seasonal effect modification for the associations between PM2.5 sources and health.. We explored age-specific associations between short-term exposure to PM2.5 chemical constituents and its sources, and hospital admissions in California. We linked hospital admission data (n = 1,679,094) with PM2.5 chemical constituents and source apportionmentdata for eight sites in California for the period of 2002–2009. Site-specific source apportionmentwas conducted using Positive Matrix Factorization, and five PM2.5 sources were commonly identified in most sites (biomass burning, soil, secondary ammonium nitrate, secondary ammonium sulfate, and vehicular emissions). Age-stratified Poisson time-series regression was conducted for each site, and the health risk estimates were combined to generate overall age-specific associations with cardiovascular- and respiratory-related hospital admissions. We further conducted seasonal interaction models to assess seasonal effect modification. An interquartile range increase in PM2.5 vehicular emissions was associated with increased risk of cardiovascular-related hospital admission at lag 0 (1.32% [95% confidence interval (CI): 0.16, 2.49]) for elderly people (≥65 years old). Exposure to PM2.5 vehicular emissions increased the risk of respiratory-related hospitalizations at lag 2 (3.58% [95% CI: 0.90, 6.33]) for children (0–18 years old). Risk estimates of PM2.5 total mass, vehicular emissions, and its related constituents (e.g., iron) for respiratory admissions were higher in the warm season among children. Heterogeneous seasonal estimates were not observed for other age groups. Our results suggest that short-term exposures to several PM2.5 sources and their related constituents are more harmful than exposures to other pollutants, particularly for children in summer. Identifying toxic sources is important for developing effective interventions and protecting susceptible populations.