Association of PM2.5 and black carbon with preterm birth, mediated by gestational diabetes mellitus: genetic- and cohort-based approaches

Exposure to PM 2.5 is associated with an increased risk of preterm birth (PTB), yet the key constituents and underlying pathways remain unclear. Gestational diabetes mellitus (GDM), which is also linked to PM 2.5 , may mediate this association. This study aims to investigate the associations between PM 2.5 , GDM, and PTB using the Mendelian randomization (MR) as an initial exploratory tool to assess potential causal associations, and subsequently a large-scale cohort study to provide pregnancy-specific estimates and mediation analyses, with a particular attention on PM 2.5 pollution events (> World Health Organization's interim target III for daily PM 2.5 ) and constituents. In European MR analysis, univariable models suggested causal associations between GDM and PTB, as well as between PM 2.5 and GDM; however, multivariable models indicated that PM 2.5 absorbance, rather than PM 2.5 itself, was positively associated with PTB. In East Asian MR analysis, no significant association was observed between PM 2.5 and GDM, but PM 2.5 absorbance exhibited a strong positive association with GDM. In cohort study, PM 2.5 events were associated with both GDM and PTB when combined with high-BC concentrations, with BC showing the strongest independent effect on both outcomes among major PM 2.5 constituents. Notably, there was no significant association between low-BC PM 2.5 events and PTB, similar to that only in high-BC regions did PM 2.5 events remain associated with GDM and PTB. Mediation analyses in MR and cohort study suggest that GDM mediated the PM 2.5 effect on PTB. Differences in key PM 2.5 constituent(s) could help explain the causal association between PM 2.5 , GDM, and PTB. Women with GDM require more attention to prevent the transmission of PM 2.5 -related adverse effects from mother to infant.