Maternal air pollution exposure increases the risk of preterm birth: Evidence from the meta-analysis of cohort studies.

Preterm birth (PTB), a major public health impact, has been shown to be associated with prenatal air pollution exposure, but the results are still inconsistent. This meta-analysis was performed to quantitatively evaluate the correlation between maternal air pollutant exposure and PTB, and provide evidence of higher grade to help improving the pregnancy outcomes. Databases including Web of Science and PubMed were searched to retrieve eligible studies published up to October 2020. The quality of the articles was assessed by the Newcastle-Ottawa Quality Score (NOS), after which the pooled estimate of the effect was calculated. The robustness of the joint estimates was confirmed by sensitivity analysis of excluded studies one by one, and the sources of heterogeneity were discussed by stratification analysis. Egger's and Begg's tests were performed to examine publication bias. Sixty studies that met the eligible criteria were finally included in this study. The findings showed combined relative risks of 1.032-1.070 for PTB, 0.859-1.081 for moderate PTB, 1.119-1.194 for very PTB and 1.128-1.259 for extremely PTB when mothers were exposed to PM2.5, PM10, NO2, O3, SO2, CO and NOx during pregnancy, while the sensitive windows varied for different air pollutants. Notably, PM2.5 exposure in only the 2nd trimester, NO2 exposure in only the 3rd trimester, and O3 exposure in all three trimesters were positively associated with PTB, while NO2 exposure in the 1st trimester was negatively associated with PTB. In addition, exposure of PM2.5 and PM10 in the 2nd trimester was positively associated with moderate PTB, and in the 1st and 2nd trimesters were positively associated with very PTB. These findings demonstrated that PM2.5, PM10, O3, NO2 were associated with PTB (including moderate PTB, very PTB, and/or extremely PTB), while NOx was not, and the relationship between CO and SO2 and PTB was not stable.