Association of urinary metabolites of non-persistent pesticides with serum sex hormones among the US females: NHANES 2013-2014

Background: Evidence indicated the possibility of non-persistent pesticides disrupting the homeostasis of sex hormones. However, few studies have focused on this relationship in females. We aimed to explore the relationship between non-persistent pesticide exposure and sex hormones among the US females from the National Health and Nutrition Examination Survey 2013-2014.Methods: A total of 790 females, including girls (6-11 years), female adolescents (12-19 years), and adult females (>19 years), were enrolled in this study. Age stratified associations of individual non-persistent pesticide metabolites and their mixtures with sex hormones were analyzed by weighted multiple linear regression and Bayesian kernel machine regression (BKMR) using spot urinary non-persistent pesticide measurement, including 2,4-dichlorophenoxyacetic acid (2,4-D), 3,5,6-trichloropyridinol (TCPY), para-nitrophenol (PNP) and 3-phenoxybenzoic acid (3-PBA), and three serum sex hormones [total testosterone (TT), estradiol (E2) and sex hormone binding globulin (SHBG)].Results: In girls, weighted multivariate linear regression indicated that both 2,4-D and PNP were negatively associated with TT, and TCPY was inversely associated with SHBG. In female adolescents, TCPY was negatively associated with TT and E2, and 3-PBA was negatively associated with SHBG; positive associations were detected both in 2,4-D with SHBG, and in PNP with TT. In adult females, a higher concentration of 3-PBA was associated with higher levels of TT. The BKMR model showed that in female adolescents, the concentrations of pesticide metabolite mixtures at or above the 55th percentile were negatively related to the levels of E2 compared with their mixtures at 50th percentile, and an inverse U-shaped exposure-response function between PNP and E2 was found.Conclusions: Associations between the four non-persistent pesticide metabolites and serum sex hormones were identified in the US females from NHANES 2013-2014 and these associations were age dependent, especially in adolescents. Large-scale cohort studies are needed to confirm these findings and elucidate the potential biological mechanisms.