Prenatal exposure to polybrominated diphenyl ethers: Predictors of exposure and relation to neonatal thyroid hormone measures

Background. Polybrominated diphenyl ethers (PBDEs) are a class of compounds similar in chemical shape and structure to polychlorinated biphenyls (PCBs). Additional similarities include their fire-resistance, chemical stability, environmental persistence, ability to bioaccumulate and capacity for long-range transport. However, while PCBs were banned in the late 1970's, PBDEs are currently in use. Studies examining exposure over time indicate that PBDE levels in human samples are increasing exponentially while PCB concentrations are on the decline. Toxicological studies have indicated that both PCBs and PBDEs have the ability to disrupt thyroid hormone levels in laboratory animals. Epidemiologic studies examining the impact of perinatal exposure to PCBs on thyroid hormone levels have been equivocal while similar studies examining the impact of PBDE exposure have not been conducted.;Objectives. The overall objectives of this dissertation were three-fold: (1) to explore the determinants of prenatal PBDE exposure; (2) to examination of the factors that influence thyroid hormone levels at and shortly after birth; and (3) to explore the impact of prenatal PCB and PBDE exposure on neonatal thyroid hormone levels.;Methods. Umbilical cord blood was collected from a convenience sample of 300 newborns delivered at the Johns Hopkins Hospital (Baltimore, Maryland) between November 2004 and March 2005. The blood was analyzed for 8 PBDE and 35 PCB congeners as well as 3 thyroid hormone indicators: thyroxine (total and free T4) and thyrotropin (TSH). An additional total T4 measurement at 36 hours was assessed from the Maryland Department of Health and Mental Hygiene (MD DHMH) newborn screening program. Demographic and relevant biological measures, including medical history, were abstracted from maternal and infant medical records.;Results. Cord blood levels of PBDEs in this population were similar to levels found in a smaller sample of cord blood from Indiana, but higher than concentrations measured in Swedish cord blood. There were low correlations between PCB and PBDE congeners and differences in exposure patterns. Among the most significant difference, PCBs increased with age while PBDEs were highest among the youngest mothers. Additionally, mothers with lower pre-pregnancy BMI tended to have lower PBDE levels, but higher PCB levels. Neonatal total T4 levels were correlated and increased from birth and 36 hours of life. As expected, various maternal, intrapartum, and infant characteristics were associated with neonatal thyroid hormone levels, including gestational age, mode of delivery, gestational diabetes, pregnancy induced hypertension, and factors indicative of stress during delivery (i.e. low apgar scores and having a pediatrician present at the delivery). There were no univariate associations between cord blood levels of PCBs and PBDE on thyroid hormones. When various determinants of exposure and hormone levels were incorporated as confounders in multivariate models, there was still no statistically significant association between exposure and thyroid function. However, stratifying by delivery mode, there was evidence of associations between increasing PCBs and decreasing TSH, total T4 (in cord blood) and free T4, of which the relationships between lower free T4 with increased PCBs and lower total T4 in cord blood and increased PCBs as well as increased PBDEs are more convincing. There was also evidence of a negative association between PBDEs and TSH and total T4 (in cord blood) but not free T4.;Conclusions. Although structurally similar, PCBs and PBDEs appear to have different exposure determinants and a differing impact on neonatal thyroid health. There is an observed relationship between prenatal PCB and PBDE exposure and neonatal thyroid hormone levels that is consistent with the notion of hormonal perturbation, as indicated by the toxicological models. However, with no other human data available for comparison, this relationship should be explored further and needs to be confirmed in other epidemiologic investigations.