Early-pregnancy per- and polyfluoroalkyl substances (PFAS), vitamin status, and risk of gestational diabetes mellitus: a prospective birth cohort study

Per- and polyfluoroalkyl substances (PFAS) are ubiquitous environmental contaminants with endocrine-disrupting properties that may impair glucose homeostasis during pregnancy. Whether maternal vitamin D and folate status can attenuate PFAS-related metabolic risk remains unclear. We assessed associations between early-pregnancy PFAS exposure and gestational diabetes mellitus (GDM) risk, and examined whether maternal 25-hydroxyvitamin D [25(OH)D] and 5-methyltetrahydrofolate (5-MTHF) modify these associations at both individual and mixture levels. This prospective cohort included 420 pregnant women from Ma’anshan, China. Fourteen PFAS, 25(OH)D, and 5-MTHF were quantified in early-pregnancy serum. GDM was diagnosed by the 75-g oral glucose tolerance test at 24–28 gestational weeks. Modified Poisson regression estimated risk ratios for individual PFAS. Effect modification was evaluated using interaction terms and stratified analyses. Vitamin D and folate status-specific weighted quantile sum (WQS) regression with 100 repeated holdouts assessed modification at the mixture level. Seventy women (16.7%) developed GDM. Per doubling of serum concentration, increased GDM risk was observed for PFNA, PFDA, PFOS, PFOA, and 6:2 Cl-PFESA. Vitamin D status appeared to modify these associations, with significant positive associations largely confined to vitamin D–insufficient women at both the individual (PFOA, PFNA, PFDA, and PFOS) and mixture levels (insufficient: mean OR = 1.10, 95% CI: 1.01–1.24; sufficient: mean OR = 1.01, 95% CI: 0.81–1.17). However, formal tests of effect modification provided only limited support, in that significant interaction terms were observed for selected individual PFAS (PFHpA and PFBS, p for interaction < 0.10), whereas the WQS × vitamin D interaction term was not significant (mean OR = 0.85, 95% CI: 0.61–1.26). Several individual PFAS were associated with increased GDM risk among women within low folate group, but folate-stratified WQS regression did not reveal significant between-group differences. Early-pregnancy exposure to multiple legacy and emerging PFAS was associated with elevated GDM risk. Significant associations were largely confined to women with insufficient vitamin D, with a weaker pattern for folate, but formal interaction tests provided only limited support; these exploratory findings warrant confirmation in adequately powered studies before vitamin D optimization can be considered a strategy to mitigate PFAS-related metabolic risk.