Environmental exposure to perfluoroalkyl substances in early pregnancy, maternal glucose homeostasis and the risk of gestational diabetes: A prospective cohort study.

Abstract

Humans are widely exposed to environmental perfluoroalkyl substances (PFAS), which may affect glucose homeostasis. However, research linking PFAS exposure to glucose homeostasis during pregnancy is limited and the results were inconsistent. We aimed to investigate the association between PFAS exposure and glucose homeostasis in pregnancy in a large prospective cohort. A total of 2747 pregnant women who participated in the Shanghai Birth Cohort, had blood samples in early pregnancy and completed a 75g oral glucose tolerance test (OGTT) at 24-28 gestational weeks were included. 10 PFAS were determined by high-performance liquid chromatography/tandem mass spectrometry (HPLC/MS-MS) in the plasma samples in early pregnancy. Logistic regression was used to explore the associations between PFAS concentrations and gestational diabetes mellitus (GDM), while multiple linear regression was used to model the associations between PFAS and OGTT fasting, 1-h and 2-h glucose levels. Potential confounders were adjusted. Bayesian kernel machine regression (BKMR) and a quantile-based g-computation approach (qgcomp) were employed to explore the joint and independent effects of PFAS on glucose homeostasis. The incidence of GDM was 11.8%. One log-unit increment in plasma concentrations in early pregnancy was associated with an increased risk of GDM for perfluorobutane sulfonate (PFBS) (adjusted odd ratio (aOR)=1.23, 95% confidence interval (95% CI): 1.05, 1.44) and perfluoroheptanoic acid (PFHpA) (aOR=1.25, 95% CI: 1.07, 1.46). Perfluorooctane sulfonic acid (PFOS), perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), perfluorohexanesulfonate (PFHxS) and PFHpA were positively correlated with 1-h and 2-h glucose levels. Results of the mixed exposure model showed that the joint effects of PFAS were significantly associated with abnormal glucose homeostasis; In the BKMR model, PFAS mixture exposure was positively associated with the GDM incidence, 1-h and 2-h glucose levels and negatively correlated with FBG level. A similar trend could be observed in qgcomp and the positive correlation between PFAS and 2-h glucose level was significant (β=0.12, 95% CI: 0.04, 0.20). PFOS, PFNA and PFHpA may be the main contributors after controlling for other PFAS congeners. PFOS was significantly correlated with GDM incidence and 2-h glucose level, and PFHpA was significantly associated with FBG and 2-h glucose levels. The above associations were more prominent among women with a normal prepregnant BMI. Environmental exposure to PFAS may affect glucose homeostasis in pregnancy and increase the risk of GDM, especially in normal weight women.