Examining the association between urinary triclosan levels and menopausal status: results from the National Health and Nutrition Examination Survey, 2003 to 2016

Abstract

Objective To examine the association between urinary levels of triclosan (TCS), a ubiquitous endocrine disrupter, and menopausal status using the National Health and Nutrition Examination Survey. Methods A retrospective cross-sectional study from 2003 to 2016 was conducted among US female participants who completed the reproductive health questionnaire and provided TCS-level measurements. Exposure was assessed by urinary TCS levels adjusted for urinary creatinine; levels were log-transformed to achieve normal distribution for parametric analyses. Menopausal status was based on participants' responses to: “What is the reason that you have not had a period in the past 12 months?” Multivariable linear regression analyses examined the association between creatinine-adjusted urinary TCS levels and menopausal status after adjusting for age at survey completion, body mass index, race, ethnicity, and smoking exposure. Results Of the final sample of female participants (n = 6,958), 40% identified as postmenopausal, of whom 60% had experienced natural menopause, and of these, 11% had become menopausal at under 40 years of age. Triclosan levels correlated positively with advancing age (r = 0.09, P < 0.001) and inversely with body mass index (r = −0.09, P < 0.001). Smoking exposure was associated with significantly lower TCS levels (P < 0.001). Compared with premenopausal women, postmenopausal women had significantly higher log-transformed, creatinine-adjusted TCS levels (mean, −1.22 ± 1.79 vs −1.51 ± 1.79 ng/mg creatinine; P < 0.001). Triclosan levels were unrelated to the duration of menopause and did not differ between women who underwent natural versus surgical menopause, and premature menopause versus menopause at 40 years or older. In unweighted multivariate linear regression analyses, menopausal status was independently associated with higher urinary TCS levels after adjusting for covariates (β coefficient, 0.17; 95% CI, 0.020-0.323; P = 0.026). Conclusions In a nationally representative sample, postmenopausal status was associated with higher urinary TCS levels, observations that merit further investigation into potential exposures and health consequences.