Background: Most evidences regarding ambient PM2.5 or PM10 (particulate matter of median aerodynamic diameter ≤2.5 μm or ≤10 μm) and preterm birth (PTB) come from western countries which has relatively low PM pollution exposure, and the results are still inconsistent. This study aims to examine whether exposure to high concentrations of PM2.5 or PM10 was associated with PTB (<37 weeks) and near term birth (37–38 weeks). Method: We established a birth cohort with 1,280,524 singleton pregnancies who delivered from Dec 1st, 2013 to Nov 30th, 2014 and matched their home address to PM2.5 and PM10 concentrations which were predicted with machine learning methods based satellite remote sensing, meteorological and land use information. Cox proportional hazard regression models were used to analyze the associations between PTB and exposure of PM2.5 or PM10, after controlling for individual level covariates. Results: Exposure to PM2.5 or PM10 during pregnancy increases the risk of PTB and near term birth [e.g., Hazard ratios: 1.09 (95% CI: 1.09, 1.10), 1.08 (95% CI: 1.07, 1.08), 1.01 (95% CI: 1.01, 1.02), and 1.09 (95% CI: 1.08, 1.10) for each 10 μg/m3 increase in PM2.5 for the 1st, 2nd, 3rd trimester and over the entire pregnancy, respectively]. The effects appeared to be stronger among women who come from rural areas, worked as farmers, were overweight before conception, whose mate was smoking during pregnancy, and conceived in autumn. Conclusion: This study provides clear evidence that exposure to PM2.5 or PM10 during pregnancy increases the risk of PTB and near term birth. Public policies regarding improvement of air quality would produce great health benefit by reducing the burden of preterm birth.
- Birth cohort
- Preterm birth
- Satellite-based PM and PM predictions
ASJC Scopus subject areas
- Environmental Science(all)