Thyroid hormones in relation to polybrominated diphenyl ether and metals exposure among rural adult residents along the Yangtze River, China.

Although several studies indicate that exposure to polybrominated diphenyl ethers (PBDEs) and metals may influence thyroid function, the evidence is limited and inconsistent in general population. The current study was conducted to determine the levels of plasma PBDEs and urinary metals and evaluate the associations of co-exposure to both with thyroid hormones (THs) among rural adult residents along the Yangtze River, China. A total of 329 subjects were included in current analyses, and 8 PBDEs congeners and 14 urinary metals were measured to reflect the levels of environmental exposure. Multiple linear regression models were used to evaluate the association between PBDEs, metals and THs levels. Bayesian Kernel Machine Regression (BKMR) was used to examine PBDEs and metals mixtures in relation to THs. The geometric mean (GM) and 95% confidence interval (CI) of total measured PBDEs was 65.10 (59.96, 70.68) ng/g lipid weights (lw). BDE-209 was the most abundant congener, with a GM (95% CI) of 47.91 (42.95, 53.26) ng/g lw, accounting for 73.6% of the total PBDEs. Free thyroxine (FT4) was significantly negatively associated with BDE-28, 47, 99, 100, 154, and 183, and urinary strontium [β (95% CI): -0.04 (-0.07, -0.02)], but positively associated with selenium [β (95% CI): 0.04 (0.02, 0.06)]. Free triiodothyronine (FT3) was negatively associated with BDE-28 [β (95% CI): -0.03 (-0.05, -0.01)] and urinary arsenic [β (95% CI): -0.01 (-0.02, -0.001)]. The current study did not observe a statistically significant association of thyroid-stimulating hormone (TSH) with PBDEs and urinary metals. BKMR analyses showed similar trends when these chemicals were taken into consideration simultaneously. We found no significant interaction in the association between individual chemical at the 25th versus 75th percentiles and THs estimates, comparing the results when other chemicals were set at their 10th, 50th, and 90th percentile levels. Further study is required to confirm these findings and determine potential mechanisms.