Meta-analysis of fine particulate matter exposure during pregnancy and birth weight: Exploring sources of heterogeneity.

Background: Several meta-analyses assessed the relationship between exposure to PM with aerodynamic diameter ≤ 2.5 μm (PM _2.5 ) during pregnancy and birth weight (BW), but results were inconsistent and substantial unexplained heterogeneity was reported. We aimed to investigate the above association and to explore sources of heterogeneity across studies.
Methods: We systematically reviewed the current worldwide evidence examining the association between PM _2.5 and BW. The review protocol was registered on the PROSPERO website (CRD42020188996) and followed PRISMA guidelines. We extracted association measures for BW and low birth weight (LBW, BW < 2500 g) from each study to evaluate pooled summary measures and to explore sources of between-study heterogeneity.
Findings: Of the 2677 articles identified, 84 met the inclusion criteria (~42 M births). Our random effects meta-analyses revealed substantial heterogeneity among included studies (I ²  = 98.4 % and I ²  = 77.7 %, for BW and LBW respectively). For LBW, the heterogeneity decreased (I ²  = 59.7 %) after excluding four outlying studies, with a pooled odds ratio 1.07 (95 % confidence interval, CI: 1.05, 1.09) per a 10-μg/m ³ increase in mean PM _2.5 exposure over the entire pregnancy. Further subgroup analysis revealed geographic heterogeneity with higher association in Europe (1.34, (1.16, 1.55)) compared to Asia (1.06, (1.03, 1.10)) and US (1.07, (1.04, 1.10)).
Conclusion: The association between PM _2.5 and birth weight varied depending on several factors. The sources of heterogeneity between studies included modifiers such as study region and period. Hence, it is advisable not to pool summary measures of PM _2.5 -BW associations and that policy would be informed by local evidence.
Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(Copyright © 2024 Elsevier B.V. All rights reserved.)