Associations of per- and polyfluoroalkyl substances and alternatives with reproductive hormones in women of childbearing age

Abstract

Background: Experimental studies suggested that per- and polyfluoroalkyl substances (PFAS) may have endocrine-disrupting effects. However, the epidemiological evidence on the associations of PFAS with female reproductive hormones is sparse and limited to perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA).

Objective: To evaluate effects of legacy and emerging PFAS alternatives on female reproductive hormones.

Methods: A total of 433 reproductive-aged females were recruited from 2014 to 2016. Information on age, age at menarche, gravity, menstrual cycle, BMI, education, and income was obtained from medical records and questionnaires. Serum samples were collected for reproductive hormones, and plasma samples for PFAS measurement by ultraperformance liquid chromatography - tandem mass spectrometer (UPLC-MS/MS). Multiple linear regression and quantile g-computation (q-gcomp) were used to examine the associations of individual PFAS and their mixture with reproductive hormones.

Results: Multiple linear regression analysis showed significant effects of certain PFAS on total testosterone (TT): adjusted estimate (β) for perfluoroheptanoic acid (PFHpA) was 0.57 (95% CI: 0.18, 0.97). Moreover, a positive association was detected between PFAS mixture and TT in the q-gcomp model. Higher concentrations of 6:2 chlorinated polyfluorinated ether sulfonic acid (6:2 Cl-PFESA) were associated with significantly lower prolactin level (β = -0.07, 95% CI: -0.14, -0.001).

Conclusion: Our study found that exposure to PFAS alternatives was associated with altered levels of reproductive hormones in women of childbearing age.

Keywords: Cross-sectional study; Female reproductive hormones; Mixture analyses; PFAS.