Maternal per- and polyfluoroalkyl substances exposure and the co-occurrence of gestational diabetes mellitus and preeclampsia/eclampsia: a nested case–control study

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Objective Few studies have investigated the relationship between maternal exposure to per- and polyfluoroalkyl substances (PFAS) and the co-occurrence of gestational diabetes mellitus (GDM) and hypertensive disorders of pregnancy, particularly preeclampsia/eclampsia (PE/E). The present study aimed to explore whether PFAS exposure is associated with the co-occurrence of GDM and PE/E. Methods This study enrolled 333 pregnant women diagnosed with both GDM and PE/E, and 360 healthy controls. First-trimester plasma concentrations of 18 PFAS were determined by ultra-high performance liquid chromatography-tandem triple quadrupole mass spectrometry (UPLC-MS/MS). The associations of individual and combined maternal PFAS exposure with the odds of co-occurrence were assessed by restricted cubic spline (RCS) analysis, multivariable logistic regression, and Bayesian kernel machine regression (BKMR). Results The concentrations of perfluoroheptanoic acid (PFHpA), perfluorobutanesulfonic acid (PFBS), and perfluorotridecanoic acid (PFTrDA) differed significantly between the two groups; however, after false discovery rate (FDR) correction, only PFHpA and PFTrDA remained significantly different. RCS analysis showed significant associations of PFHpA and PFBS with the co-occurrence ( P overall = 0.026 and 0.014, respectively), without nonlinearity ( P non-linear = 0.619 and 0.206, respectively); However, the overall associations were no longer significant after FDR correction. In multivariable logistic regression models, after FDR correction, each log10-unit increment in PFHpA and PFBS was significantly associated with increased odds of the co-occurrence, with adjusted odds ratios (aORs) of 2.439 (95% CI: 1.269–4.689) and 1.688 (95% CI: 1.161–2.453), respectively. For perfluoroheptane sulfonic acid (PFHpS), a positive association was also observed (aOR = 1.712, 95% CI: 1.051–2.788); however, the association was not significant after FDR correction. BKMR analysis revealed a positive association between PFAS mixture exposure and the co-occurrence, with the association strengthening progressively at higher mixture concentrations. Among the individual PFAS, PFHpA, PFBS and PFHpS exhibited positive trends, whereas PFTrDA displayed inverse trends. Conclusion PFHpA, PFBS, and the 10 PFAS mixture were all positively associated with increased odds of co-occurrence of GDM and PE/E.