Polybrominated diphenyl ether (PBDE) exposures and thyroid hormones in children at age 3 years

Abstract

Background: Polybrominated diphenyl ethers (PBDEs) reduce serum thyroid hormone concentrations in animal studies, but few studies have examined the impact of early-life PBDE exposures on thyroid hormone disruption in childhood.

Methods: We used data from 162 mother-child pairs from the Health Outcomes and Measures of the Environment Study (2003-2006, Cincinnati, OH). We measured PBDEs in maternal serum at 16 ± 3 weeks gestation and in child serum at 1-3 years. Thyroid hormones were measured in serum at 3 years. We used multiple informant models to investigate associations between prenatal and early-life PBDE exposures and thyroid hormone levels at age 3 years.

Results: Prenatal PBDEs were associated with decreased thyroid stimulating hormone (TSH) levels at age 3 years. A 10-fold increase in prenatal ∑PBDEs (BDE-28, -47, -99, -100, and -153) was associated with a 27.6% decrease (95% CI -40.8%, -11.3%) in TSH. A ten-fold increase in prenatal ∑PBDEs was associated with a 0.25 pg/mL (0.07, 0.43) increase in free triiodothyronine (FT3). Child sex modified associations between prenatal PBDEs and thyroid hormones, with significant decrements in TSH among females and decreased free T4 (FT4) in males. Prenatal ∑PBDEs were not associated with TT4, FT4, or total T3.

Conclusions: These findings suggest an inverse relationship between prenatal ∑PBDEs and TSH at 3 years. Associations may be sexually dimorphic, with an inverse relationship between prenatal BDE-47 and -99 and TSH in females and null associations among males.

Keywords: Brominated flame retardants; Endocrine disruption; Epidemiology; Polybrominated diphenyl ethers (PBDEs); Thyroid function; Thyroid hormones.

Conflict of interest statement

Competing financial interest declaration

The authors declare they have no competing financial interests.

Copyright © 2018 Elsevier Ltd. All rights reserved.

Figures

Fig. 1

β-coefficients and 95% CIs from multiple informant models for associations of PBDEs (ng/g lipid) and thyroid hormones at age 3 years. Adjusted by maternal age, race/ethnicity, education, smoking status during pregnancy, alcohol consumption during pregnancy, vitamin use, maternal BMI, parity, maternal serum ΣPCBs, and child sex.

Fig. 2

β-coefficients and 95% CIs from linear regression models for associations of prenatal PBDEs (ng/g lipid) and thyroid hormones at 3 years by child sex. Adjusted by maternal age, race/ethnicity, education, smoking status during pregnancy, alcohol consumption during pregnancy, vitamin use, maternal BMI, parity, maternal serum ΣPCBs, and child sex.