Abstract
Many studies have found associations between particulate matter having an aerodynamic diameter of ≤2.5 μm (PM2.5) and adult mortality. Comparatively few studies evaluated particles and infant mortality, although infants and children are particularly vulnerable to pollution. Moreover, existing studies mostly focused on short-term exposure to larger particles. We investigated PM2.5 exposure during pregnancy and lifetime and postneonatal infant mortality. The study included 465,682 births with 385 deaths in Massachusetts (2001-2007). Exposures were estimated from PM2.5-prediction models based on satellite imagery. We applied extended Cox proportional hazards modeling with time-dependent covariates to total, respiratory, and sudden infant death syndrome mortality. Exposure was calculated from birth to death (or end of eligibility for outcome, at age 1 year) and pregnancy (gestation and each trimester). Models adjusted for sex, birth weight, gestational length, season of birth, temperature, relative humidity, and maternal characteristics. Hazard ratios for total, respiratory, and sudden infant death syndrome mortality per-interquartile-range increase (1.3 μg/m3) in lifetime PM2.5 exposure were 2.66 (95% confidence interval (CI): 2.11, 3.36), 3.14 (95% CI: 2.39, 4.13), and 2.50 (95% CI: 1.56, 4.00), respectively. We did not observe a statistically significant relationship between gestational exposure and mortality. Our findings provide supportive evidence that lifetime exposure to PM2.5 increases risk of infant mortality.