Associations between organophosphate esters and sex hormones among 6-19-year old children and adolescents in NHANES 2013-2014
Kai Luo,Jihong Liu,Yuqing Wang,Ruxianguli Aimuzi,Fei Luo,Junjie Ao,Jun Zhang
DOI: https://doi.org/10.1016/j.envint.2020.105461
Abstract:Background: Organophosphate esters (OPEs) are a class of alternative replacements for polybrominated diphenyl ethers. In vitro and in vivo studies suggested that OPEs may disrupt the homeostasis of sex steroid hormones. However, human evidence in children and adolescents is limited. Objectives: We conducted a cross-sectional analysis of the associations between OPE biomarkers and sex steroid hormones among children (6-11 years) and adolescents (12-19 years) in the U.S. National Health and Nutrition Examination Survey, 2013-2014. Methods: Participants aged 6-19 years who had available data on urinary OPE metabolites, serum sex hormones [total testosterone (TT), estradiol (E2)] and sex hormone binding globulin (SHBG) were included (n = 544). Free androgen index (FAI) calculated as TT divided by SHBG and a ratio of TT to E2 (TT/E2) were generated. Five urinary OPE metabolites were examined. A constructed puberty status was defined as either high steroid hormone levels (TT ≥ 50 ng/dL in males and E2 ≥ 20 pg/ml in females) or onset of menarche. Multiple linear regression and weighted quantile sum (WQS) regression analyses stratified by sex-age and sex-puberty-status groups were conducted to examine the associations of OPE metabolites and its mixture with sex hormone levels. Results: After adjusting for covariates, dibutyl phosphate (DBUP) and dibutyl phosphate (DPHP) were significantly inversely associated with TT (or FAI) and E2; DBUP was negatively associated with SHBG; and DPHP was positively associated with SHBG and TT/E2 in female adolescents. In male adolescents, we observed monotonic negative associations of bis(1,3-dichloro-2-propyl) phosphate (BDCPP), DBUP or DPHP with TT (or FAI) and E2, and positive associations of BDCPP and DPHP with SHBG. Among adolescents, the OPEs index was negatively associated with TT [WQS beta = -0.29 (95% confidence interval: -0.51, -0.07) in males and -0.15 (-0.28, -0.01) in females ], FAI [-0.46 (-0.71, -0.2) in males and -0.23 (-0.41, -0.05) in females] and E2 [-0.25 (-0.41, -0.1) in males and -0.33 (-0.59, -0.08) in females], with stronger associations with TT and FAI in males and a slightly stronger association with E2 in females. In addition, the OPEs index presented a comparable positive association with SHBG in both sexes of adolescents. In contrast, significant associations of individual OPE metabolites or OPEs index with sex hormones were sparse in children. Results by sex-puberty status in single pollutant and WQS regression analyses presented a similar pattern, where most of the significant associations were limited to the pubertal individuals. Of note, stronger inverse associations of the OPEs index with TT and FAI remained in pubertal boys. But the association between the OPEs index and E2 was non-significant in pubertal girls, and only in pubertal boys did the OPEs index show a significant and stronger inverse association with E2. Conclusions: Exposure to OPEs, either individually or as a mixture, was associated with decreased levels of certain sex steroid hormones (TT, FAI, and E2) and increased levels of SHBG in adolescents or pubertal individuals, with the associations presenting somewhat sex-dependent pattern. However, there is little evidence of the significant associations in children or prepubescent ones. Given the cross-sectional nature of the analysis, our findings need further confirmation.