Effect of exposure to per- and polyfluoroalkyl substances in early pregnancy on blood glucose levels

Given the limited evidence on per- and polyfluoroalkyl substance (PFAS) exposure and glucose metabolism in pregnancy, this study evaluated the associations of first-trimester exposure to individual and multiple PFAS with mid-pregnancy blood glucose levels and gestational diabetes mellitus (GDM). This study recruited 314 pregnant women from the Changsha Maternal and Child Health Hospital, Hunan Province, China. Serum samples were collected at 11–13 weeks of gestation, and the concentrations of 25 PFAS compounds were measured. At 24–28 weeks of gestation, all participants underwent a 75 g oral glucose tolerance test. Multivariate logistic regression was applied to estimate associations between individual PFAS and GDM, whereas multivariate linear regression was utilized to assess their associations with fasting blood glucose (FBG), as well as 1-h and 2-h post-load glucose levels. Associations were stratified by maternal age and pre-pregnancy body mass index, and mixed effects were estimated using Bayesian kernel machine regression (BKMR). In the multivariate logistic regression analysis, compared with the lowest tertile, early-pregnancy exposure to the second tertile of perfluorooctanoic acid (PFOA; OR = 1.97, 95% CI : 1.04, 3.73), and the highest tertile of perfluoroheptane sulfonic acid (PFHpS; OR = 2.17, 95% CI : 1.15, 4.10) was positively associated with the risk of GDM. In the multivariate linear regression analysis, the second tertile of PFOA exposure was positively associated with 1-h glucose levels ( β = 0.63, 95% CI : 0.19, 1.08). The second ( β = 0.46, 95% CI : 0.06, 0.86) and highest ( β = 0.44, 95% CI : 0.03, 0.85) tertiles of PFOA, as well as the highest tertile of PFHpS ( β = 0.71, 95% CI : 0.31, 1.12), were significantly associated with increased 2-h glucose levels. In contrast, inverse associations or null associations were observed for certain other PFAS. Exposure to certain PFAS was associated with an increased risk of GDM and glucose intolerance, particularly in women of normal weight and women aged ≥ 30 years. The BKMR model results indicated that the PFAS mixture was negatively associated with FBG. Specifically, when the natural logarithm of the PFAS mixture increased from the median to the 60th percentile, FBG levels significantly decreased ( β = –0.03, 95% CI : –0.05, –0.01), with N-ethylperfluoro-1-octanesulfonamide (N-EtFOSA-M) contributing most to this association. Early pregnancy PFAS exposure was associated with both elevated and reduced glucose levels and GDM risk, varying by PFAS type and maternal characteristics. These findings highlight the complexity of PFAS effects on gestational glucose metabolism.