ASSOCIATIONS BETWEEN PER- AND POLYFLUOROALKYL SUBSTANCES (PFAS) EXPOSURE AND SEMEN QUALITY PARAMETERS AMONG MEN ATTENDING A FERTILITY CLINIC

Per- and polyfluoroalkyl substances (PFAS) are persistent chemicals that have grown concerningly abundant in the environment (half-lives range from 3.5-8.5 years). While previous studies have explored the relationship between PFAS exposure and semen quality parameters, inconsistent relationships have been reported. This study included 274 males with 687 repeated semen quality measurements from a prospective cohort of males attending a fertility clinic in Massachusetts, USA between 2005-2019. Six PFAS analytes (LOD>85%) were measured in serum samples at baseline and semen quality parameters were determined in samples collected at each visit corresponding to fertility treatment cycles. We applied linear mixed effects models to investigate differences in semen quality parameters per doubling of PFAS concentrations, adjusting for age, body mass index (BMI), race, smoking status, education, and days since last ejaculate. Participants had a median age of 36.5 years (interquartile range [IQR]: 7.5), a BMI of 26.9 kg/m2 (IQR: 6.0), and tended to be white (85.4%), never smokers (64.6%) with graduate degrees (53.8%). For every doubling of perfluorooctanesulfonic acid (PFOS), perfluorooctanoic acid (PFOA), and perfluorodecanoic acid (PFDA) concentration, total sperm count (million) decreased by 20.6 (95% CI: -39.5, 1.7), 20.1 (95% CI: -43.8, 3.5), and 16.4 (95% CI: -37.8, 5.1), respectively. Similarly, sperm concentration (mill/mL) was reduced by 9.9 (95% CI: -18.4, -1.4), 4.9 (95% CI: -15.6, 5.9), and 7.1 (95% CI: -16.7, 2.6) for every doubling of PFOS, PFOA, and PFDA concentration, respectively. For every doubling of PFOA and perfluorohexanesulfonic acid (PFHxS) concentration, semen volume (mL) decreased by 0.1 (95% CI: -0.4, 0.1) and 0.1 (95% CI: -0.3, 0.03), respectively. In this study with repeated semen samples, men with higher PFAS exposure had lower semen volume, count, and concentration. However, larger studies that utilize mixture models are needed to confirm these results and determine their clinical significance.