Exposure of pregnant women to persistent organic pollutants and cord sex hormone levels

Abstract STUDY QUESTION Is prenatal exposure to persistent organic pollutants (POPs) associated with variations of sex hormone levels in cord blood? SUMMARY ANSWER Prenatal exposure to a number of POPs is associated with a disruption of hormone levels in cord blood, with sex specificities. WHAT IS KNOWN ALREADY Epidemiological studies have reported disorders of reproductive health, in relation with POPs exposure during early life and the endocrine disruption properties of these chemicals have been suggested as possible mechanisms. STUDY DESIGN, SIZE, DURATION A subset of 282 mother-child pairs was selected from the prospective population-based PELAGIE birth cohort (n = 3421, 2002–2006, Brittany, France). Pregnant women were recruited before 19 weeks of gestation and followed until delivery. PARTICIPANTS/MATERIALS, SETTING, METHODS Sex hormone levels including sex hormone-binding globulin (SHBG), estradiol (E2), total testosterone (T), free testosterone (fT = T/SHBG) and the aromatase index (AI = T/E2) were measured in 282 cord blood samples. Anti-Müllerian hormone (AMH) was measured in male newborns only. Pesticide concentrations of α-endosulfan, β-hexachlorocyclohexane (β-HCH), γ-HCH, dieldrin, pp′-dichlorodiphenyldichloroethylene (p,p′-DDE), hexachlorobenzene (HCB), heptachlor epoxide (HCE), as well as PCBs (congeners 153, 187 and the sum of anti-estrogenic PCBs 118, 138, and 170) and decabrominated diphenyl ether (BDE209) were also measured in cord blood. Associations between sex hormones and POPs exposure were explored using multiple linear regressions adjusted for potential confounders. MAIN RESULTS AND THE ROLE OF CHANCE High PCB levels were associated with an increase of SHBG (P-trend < 0.01) and AMH (P-trend < 0.05) and a decrease of fT (P-trend < 0.05) and AI (P-trend < 0.01). High pesticide levels, particularly α-endosulfan and HCE, were associated with an increase of SHBG (P < 0.05) and E2 (P < 0.01) and a decrease of fT (P < 0.05) and AI (P < 0.01). Several of these associations were stronger, or specific, among male or female newborns. The associations were not altered in the sensitivity analyses. LIMITATIONS, REASONS FOR CAUTION The study population was of relatively small sample size, and some compounds rarely detected in cord blood. The high level of correlation between POPs makes it difficult to identify the most contributing POPs. Hormone measurements were performed at birth (in cord blood) and may not adequately represent the infant endo