Associations between monthly fine particulate matter and ozone exposure, and postneonatal infant mortality in the United States, 2005-2017.

Using NCHS birth and death records, we estimated associations between monthly mean fine particulate matter (PM2.5) or ozone (O3) concentrations during the postneonatal period (29-364 days) and postneonatal infant mortality (IM) among infants born between 2005 and 2017 in United States (U.S.) counties with >250,000 residents. Air quality data were derived from the U.S. EPA's EQUATES project, which fuses multi-year CMAQv5.3.2 model outputs with monitor measurements, spatially weighted to each county, and linked to birth and death records. We estimated incidence rate ratios and 95% confidence intervals using a hierarchical overdispersed Poisson regression model with random-effects. In stage one, climate-zone-specific models were adjusted for temperature and year. We then pooled estimates in stage two for a national estimate. There were 35,665,533 livebirths and 71,690 postneonatal deaths across 261 counties. Congenital anomalies (19.8%) and SIDS (16.8%) were leading causes. A 5 μg/m3 increase in monthly PM2.5 exposure during the postneonatal period was associated with increased postneonatal IM for respiratory causes (1.06 [1.03, 1.10]) and SIDS (1.56 [1.30, 1.87]). A 10 ppb increase in O3 exposure during the postneonatal period was associated with increased postneonatal IM for SIDS (1.31 [1.11, 1.55]). Associations between PM2.5 and all-cause postneonatal IM were strongest in warm temperate zones, the Southwest, male infants, and infants born to younger and less educated mothers. For O3, associations were highest for all-cause postneonatal IM in the West and infants born to younger and less educated mothers. Mean monthly PM2.5 or O3 exposure is associated with increased respiratory and SIDS-related postneonatal IM.