Urinary cadmium and endometriosis prevalence in a US nationally representative sample: results from NHANES 1999–2006

Abstract STUDY QUESTION Is exposure to toxic metal cadmium associated with increased endometriosis prevalence among a nationally representative sample of the US population? SUMMARY ANSWER Concentrations of urinary cadmium, a long-term biomarker (10–30 years) of cadmium exposure, were associated with an increased prevalence of endometriosis. WHAT IS KNOWN ALREADY Cadmium exhibits estrogenic properties and may increase the risk of endometriosis, a gynecologic condition associated with substantial morbidity, for which estrogen has a central pathogenic role. Previous epidemiological studies of cadmium and endometriosis have yielded mixed results, with null, positive, and inverse associations being reported. STUDY DESIGN, SIZE, DURATION We conducted a cross-sectional study using data from four cycles of the National Health and Nutrition Examination Survey (NHANES) 1999–2006. PARTICIPANTS/MATERIALS, SETTING, METHODS The study population comprised participants aged 20–54 years who had an endometriosis diagnosis, available data on urinary cadmium, and a glomerular filtration rate ≥60 ml/min/1.73 m2 (unweighted n = 1647). Urinary cadmium was measured by inductively coupled plasma–mass spectrometry, and we used urinary creatinine concentrations and covariate-adjusted standardization to account for urinary dilution. Self-reported diagnosis of endometriosis was ascertained by interview. We examined the association between quartiles of urinary cadmium and endometriosis using log-binomial regression to estimate adjusted prevalence ratios (aPRs) and 95% CIs. MAIN RESULTS AND THE ROLE OF CHANCE We observed twice the prevalence of endometriosis for participants with cadmium concentrations in the second quartile (versus the first quartile) (aPR 2.0, 95% CI: 1.1, 3.9) and the third quartile (versus the first quartile) (aPR 2.0, 95% CI: 1.1, 3.7). Our data also suggested a 60% increased prevalence of endometriosis with urinary cadmium concentrations in the fourth quartile (versus the first quartile) (aPR 1.6, 95% CI: 0.8, 3.2). In a sensitivity analysis, restricting the study population to premenopausal participants with an intact uterus and at least one ovary (unweighted n = 1298), stronger associations accompanied by wider CIs were observed. LIMITATIONS, REASONS FOR CAUTION We were limited by the ascertainment of urinary cadmium and endometriosis diagnosis at a single time point, given the cross-sectional study design, and we relied on self-report of endometriosis diagnosis