Association of perfluoroalkyl and polyfluoroalkyl substances with polyendocrine metabolic ovarian syndrome and endocrine abnormalities: A case-control study.
BACKGROUND Polyendocrine metabolic ovarian syndrome (PMOS) is a systemic endocrine and metabolic disorder that has been increasingly linked to exposure to environmental endocrine-disrupting chemicals. Although perfluoroalkyl and polyfluoroalkyl substances (PFASs) have been implicated as potential contributors, the combined health effects of PFASs mixtures and the potential pathways linking PFAS exposure to endocrine abnormalities remain incompletely understood. METHODS In this hospital-based case-control study, 109 women with PMOS and 96 healthy controls were recruited. Serum concentrations of 19 PFASs were measured using ultra-high-performance liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS), and 11 PFASs congeners with detection frequencies exceeding 80% were included in subsequent analyses. Bayesian logistic regression and dose response analyses were used to evaluate the associations of individual PFASs with PMOS risk and endocrine biomarkers, while generalized additive models were used to characterize potential nonlinear exposure response relationships. The combined effects of PFASs mixtures were evaluated using quantile g-computation (Qgcomp) and weighted quantile sum (WQS) regression. Propensity score overlap weighting was used in sensitivity analyses to assess the robustness of the findings with respect to baseline differences in age and BMI. Exploratory mediation analyses were conducted using circulating hormone ratios related to steroidogenesis as surrogate markers to evaluate potential statistical indirect effect signals. RESULTS Eleven PFASs were included in the statistical analyses: PFBA, PFOA, PFBS, PFPeS, PFTrDA, PFHpS, PFOS, PFDS, PFHxS, PFNA, and 6:2 Cl-PFESA. PFDS, PFBA, PFHxS, and 6:2 Cl-PFESA were associated with higher PMOS risk and elevated total testosterone (TT) levels, whereas PFPeS, PFTrDA, and PFDS were associated with a lower FSH/LH ratio. Mixture analyses consistently showed that combined PFASs exposure was associated with higher PMOS risk, elevated TT levels, and a lower FSH/LH ratio. PFDS showed the largest relative contribution to the mixture associations with PMOS risk and TT levels, whereas PFTrDA showed the largest relative contribution to the association with a lower FSH/LH ratio. Sensitivity analyses using overlap weighting were broadly concordant with the primary analyses. The overall association between PFASs mixtures and PMOS remained statistically significant after overlap weighting (β = 0.57, 95% CI: 0.44-0.69). In addition, the inverse associations of PFTrDA, PFPeS, and PFDS with the FSH/LH ratio and the positive associations of PFDS and 6:2 Cl-PFESA with TT remained statistically significant. Exploratory mediation analysis identified a statistically significant indirect effect involving the circulating DHT/TT ratio, which was used as an exploratory surrogate marker of androgen conversion related to SRD5A. This indirect effect accounted for 25.9% of the association between PFDS and PMOS. CONCLUSIONS PFASs mixture exposure was associated with higher PMOS risk, elevated TT levels, and a lower FSH/LH ratio. PFDS and PFTrDA showed relatively large contributions to the observed mixture associations. Exploratory mediation analysis further identified an indirect effect involving the circulating DHT/TT ratio in the association between PFDS and PMOS; however, the biological relevance of this finding requires further mechanistic validation.