The Effect of Maternal PM2.5 Exposure on the Risk Pre-Term Births: Results from Project ELEFANT

摘要

Background/Aims: Exposure to fine particulate matter of <2.5mm (PM2.5) during pregnancy is associated with adverse birth outcomes, including miscarriage and birth defects. There is a pressing public health need to better understand the impact of such exposures. Here, we analysed the effects of pre-pregnancy and in utero PM2.5 exposure upon the risk of pre-term births by its timing and intensity. Methods: We used data from 10,306 live births within Project ELEFANT, a multi-cohort study based in Tianjin, China. Data on pre-term births (n=220) was recorded by obstetricians. Hourly air pollution measured by local ambient monitoring stations was averaged for the exposure period and used to calculate the number of days individuals were exposed to PM2.5 levels over 50, 100,150, 200, 250 and 300μg/m3 by month/trimester. We used generalized additive models to represent the non-linear relationship between outcome and spatio-temporal varying PM2.5 exposure. Models were adjusted for maternal age, education & occupation, neonate gender, region, paternal smoking, gravidity, parity, and mean temperature & dew point of the exposure period. Results: Risk of pre-term birth increased with higher frequency of extreme PM2.5 exposure in the final month of pregnancy. Relative risks (RR) ranged from 1.04 (95% CI: 1.00-1.07) with increased exposure to >50μg/m3 PM2.5 to 1.53 (1.32-1.79) with exposures to >250μg/m3. Higher risk was also observed with increased exposure to >250μg/m3 in the second- (1.19,1.02-1.40) and third-last months of pregnancy (1.47, 1.30-1.67). Further, the RR of pre-term birth was 1.67 (1.40-1.99), 1.65 (1.32-2.07) and 1.57(1.27-1.94) with increased exposure to >250μg/m3 PM2.5 in the first-, second-and third-months before pregnancy respectively. Conclusions: Our findings reveal that the level and timing of PM2.5 exposure are critical in establishing the risk of pre-term birth. Further work is required to elucidate their associations with birthweight and birth defects.