Impact of Wildfire Exposure on Preterm Births Identified in MarketScan Claims-Based Cohort, 2010-2015
On this page:
Wildfire smoke is a growing contributor to ambient particulate matter and may lead to elevated risks of preterm birth. Studies that have investigated the effects of gestational wildfire exposure (WFE) on preterm birth are inconsistent and inconclusive. This study aimed to estimate the effect of near-birth and trimester-based WFE on preterm births using a large-scale claims database.
We linked IBM’s MarketScan Commercial Claims and Encounters database with smoke plume data from NASA MODIS satellites using metropolitan statistical area (MSA) of residence at birth. Children born during 2010-2015 into an MSA in six Western states, whose gestational age (GA) and birthdate were estimable from claims data were included (N=115,469); extremely preterm births (<32 weeks gestation) were excluded. MSA, GA and birthdate were used to estimate average weekly days of WFE for each trimester and two near-birth periods (2 and 4 weeks pre-birth). Preterm birth was defined as live birth at GA between 32 and 37 weeks. Odds ratios were estimated using conditional logistic regression models accounting for clustering MSA and adjusted for infant sex, maternal age (restricted cubic spline), birth season, and average temperature within period of interest.
WFE in the first or second trimester was associated with reduced likelihood of preterm birth (0.92 [95% CI: 0.86, 0.98], and (0.96 [95% CI: 0.90, 1.01], respectively), while WFE in the third trimester was associated with increased risk of preterm birth 1.05 [95% CI: 0.98, 1.12]); the strongest association was observed in the second trimester. There was no evidence of association for the 2- and 4-week periods before birth. Sex-stratified analyses of near-birth and trimester exposures were not different from overall effect estimates.
These claims-based results suggest that the effect of wildfire smoke on preterm birth may vary by trimester of exposure.