Environmental exposure to per- and polyfluoroalkyl substances mixture and male reproductive hormones

• Published in: Environment international Vol. 152; p. 106496
• Main Authors: Luo, Kai, Liu, Xiaotu, Nian, Min, Wang, Yuqing, Qiu, Jin, Yu, Hao, Chen, Xiangfeng, Zhang, Jun
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Ltd 01.07.2021; Elsevier
• Subjects: Adult; Alkanesulfonic Acids; Bayes Theorem; Cross-Sectional Studies; Environmental Exposure; Environmental Pollutants - toxicity; Estradiol; Fluorocarbons - toxicity; Humans; Insulin-like factor 3 (INSL3); Male; Mixture effects; Per- and polyfluoroalkyl substances; Sulfonic Acids; Young adult men; Insulin-like factor 3 (INSL3); Mixture effects; Per- and polyfluoroalkyl substances; Estradiol; Young adult men
• ISSN: 0160-4120; 1873-6750
• DOI: 10.1016/j.envint.2021.106496

•PFAS mixture was inversely associated with E2 and E2/TT in young men, with PFuDA being the major contributor.•Exposure to PFTrDA and PFDoA were associated with reduced INSL3 levels.•No consistent associations were found between branched PFAS isomers and F-53B and male reproductive hormones. Previous epidemiological studies on the relationship between per- and polyfluoroalkyl substances (PFAS) exposure and male reproductive hormones were mainly limited to a few legacy PFAS and ignored the possible mixture effects. To assess the associations of PFAS mixture, branched isomers and emerging alternatives of PFAS with male reproductive hormones. A total of 902 men (mean age: 31.3 years) were recruited in this cross-sectional study. We quantified 24 targeted PFAS, including 7 branched PFOS isomers, 2 branched PFOA isomers and 2 components of F-53B, in blood plasma. Five reproductive hormones, including total testosterone (TT), estradiol (E2), follicular stimulating hormone (FSH), luteinizing hormone (LH) and insulin like factor 3 (INSL3), and sex hormone binding globulin (SHBG) were measured in serum. Associations were first assessed by confounder-adjusted multiple linear regression while correcting for multiple comparisons. Bayesian kernel machine regression (BKMR) and adaptive elastic net (AENET) were further used to assess mixture effects and the adjusted exposure response (ER) relationship of individual PFAS. After adjusting for confounders, we found that PFAS mixture was significantly and inversely associated with E2 and E2/TT, with perfluoro-n-undecanoic acid (PFuDA) being the major contributor. Although the associations between PFAS mixture and other hormones were non-significant, certain individual PFAS presented significant associations. Notably, perfluoro-n-tridecanoic acid (PFTrDA) and perfluoro-n-dodecanoic acid (PFDoA) were found to be significantly and inversely associated with INSL3, a unique indicator of Leydig cells function. Meanwhile, significant positive associations were found between perfluorobutane sulfonic acid (PFBS) and FSH and between PFuDA and LH. But the associations with branched isomers or F-53B were sporadic and inconsistent. Our findings provided the evidence that PFAS mixture may reduce E2 level, and certain PFAS (i.e., PFTrDA and PFDoA) may have negative effects on Leydig cells function among young men. Additional studies are much needed to confirm our results and elucidate potential mechanisms.