Perfluoroalkyl substances and sex hormones in postmenopausal women: NHANES 2013-2016

Background: Although an alteration in sex hormones has been linked to perfluoroalkyl substances (PFAS) in premenopausal women and girls, whether such associations exist in postmenopausal women remains uncertain. Objects: To examine the associations between serum PFAS concentrations and sex hormone levels in postmenopausal women. Methods: Data from the National Health and Nutrition Examination Survey (NHANES) 2013-2016 waves were used. A total of 706 postmenopausal women with information on serum PFAS [perfluorohexane sulfonic acid (PFHxS), pefluorodecanoic acid (PFDA); perfluorononanoic acid (PFNA); linear perfluorooctanoate (n-PFOA); linear perfluorooctane sulfonate (n-PFOS); monomethyl branched isomers of PFOS (Sm-PFOS)], sex hormones indicators [e.g., total testosterone (TT), estradiol (E2) and sex hormone binding globulin (SHBG)] as well as selected covariates were included. An indicator of circulating free testosterone (FT), and ratio of TT to E-2 (TT/E-2) were generated. Multiple linear regression accounting for the primary sampling unit, strata, and environmental sampling weights of PFAS was used for association analyses. Effect modification by obesity and type of menopause was explored via stratified analyses as well as the testing of interaction terms. Principal component analysis (PCA) and Bayesian kernel machine regression (BKMR) were conducted to assess these relationships in a multiple PFAS exposure setting. Results: After adjusting for potential confounders, total perfluorooctanoate (TPFOA: n-PFOA + Sb-PFOA) and total perfluorooctane sulfonate (TPFOS: n-PFOS + Sm-PFOS), and their linear and branched isomers were positively associated with two androgen indicators (i.e., TT and FT). PCA results revealed that the principal component (PC) composed of n-PFOA was positively associated with ln (TT) [beta = 0.09, 95% confidential interval (CI): 0.02, 0.16; per ln-ng/mL increase in exposure], and ln (FT) (beta = 0.12, 95% CI: 0.05, 0.2) in overweight/ obese [body mass index (BMI) = 25 kg/m(2)] women, but not in those with BMI < 25 kg/m(2). Additionally, among overweight/obese women, PFHxS was positively associated with androgens and negatively with ln (SHBG) (beta = -0.06, 95% CI: 0.12, 0.01). The PC composed of Sm-PFOS, n-PFOS, and PFHxS was positively associated with ln (TT) levels among overweight/obese women. Results from BKMR also confirmed the findings on n-PFOA and PFHxS. Conclusions: Our study indicates that n-PFOA and PFHxS were positively associated with levels of several androgen indicators in postmenopausal women, particularly among overweight/obese ones. Given the higher risk of cardiometabolic diseases associated with elevated levels of androgens in postmenopausal women, future studies are needed to explore the potential underlying mechanisms.