Association of exposure to fine particulate matter wave over the preconception and pregnancy periods with adverse birth outcomes: Results from the project ELEFANT

Background: No study has explored the effects of sustained maternal exposure to high-level ambient fine particulate matter (PM2.5) within a short period, i.e., PM2.5 wave, on adverse birth outcomes, though increasing epidemiological studies demonstrated that exposure to single days of high ambient PM2.5 could increase risks of adverse birth outcomes. In this study, we aim to evaluate associations of maternal PM2.5 wave exposure around pregnancy with preterm birth (PTB), small for gestational age (SGA), and large for gestational age (LGA). Methods: Totally 10,916 singleton pregnant women from all 16 districts in Tianjin, China, and their followed-up birth outcomes were included in this study. We defined PM2.5 wave as at least 2 consecutive days with daily average PM2.5 concentration exceeding 75 mu g/m(3), and 90th, 92.5th, 95th, 97.5th, 99th percentiles of PM2.5 distribution during the study period in Tianjin, respectively. Cox proportional hazard model was applied to evaluate the durational effects of PM2.5 wave during each exposure window on PTB, SGA, and LGA after adjusting for potential confounders. Results: Exposure to PM2.5 wave over the preconception and pregnancy periods was associated with increased risks of adverse birth outcomes. For PTB, the strongest association was found during the first trimester when PM2.5 wave was defined as at least 4 consecutive days with daily average PM2.5 concentration >90th (HR, 10.46; 95% CI, 6.23-17.54); and for SGA (HR, 6.23; 95% CI, 3.34-11.64) and LGA (HR, 4.70; 95% CI, 3.35-6.59), the strongest associations both were found when PM2.5 wave was defined as at least 2 consecutive days with daily average PM2.5 concentration >99th. Additionally, the risks of adverse birth outcomes generally increased at higher PM2.5 thresholds or longer durations of PM2.5 wave. Conclusion: Prolonged exposure to high-level PM2.5 over preconception and pregnancy periods was associated with increasing risks of PTB, SGA and LGA.