Associations between persistent organic pollutants and thyroid-related hormones and homeostasis parameters in middle-aged to older men and postmenopausal women: The HCHS/SOL.

Background: Previous research has yielded inconsistent results on associations of persistent organic pollutants (POPs) with thyroid hormones, which has been implicated in diabetes development. Studies specifically investigating associations with hypothalamic-pituitary-thyroid (HPT) homeostasis parameters or associations in postmenopausal women are limited.

Methods: Using baseline examination data between 2008 and 2011, we studied 1073 men and 716 postmenopausal women participating in an ancillary study of the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). Survey-weighted multivariable linear regression models examined associations of serum concentrations of eight organochlorine pesticides (OCPs), twenty-four polychlorinated biphenyls (PCBs), ten polybrominated diphenyl ethers and polybrominated biphenyl (PBB) 153 with thyroid stimulating hormone (TSH), triiodothyronine (T3), free thyroxine (FT4), and thyroid homeostasis parameters. Homeostasis parameters included T3-to-FT4 and TSH-to-T3 ratios, thyroid feedback quantile-based index, TSH index, and parametric thyroid feedback quantile-based index. POPs were analyzed in quartiles, except for mirex and 1,1,1-trichloro-2,2-diphenylethane (p,p'-DDT) which were categorized as non-detected (reference), ≤ median, and > median of detected values.

Results: p,p'-DDT was positively associated with T3 and PBB-153 negatively associated with FT4 in postmenopausal women and men. In postmenopausal women only, PBB-153 was positively associated with the T3-to-FT4 ratio (β = 3.34; 95 %CI:0.53,6.14) suggesting increased deiodination or peripheral thyroid hormone sensitivity and negatively associated with PTFQI and TFQI indicating greater central sensitivity to FT4. In men only, PCBs and select OCPs were operative in the disruption of thyroid hormones and homeostasis parameters. For instance, a one quartile increase in ∑24-PCBs was positively associated with TSHI (β = 0.06; 95 %CI:0.001,0.12) and relative increases in the geometric mean of TSH (β = 7.25mIU/L, 95 %CI:1.01,12.8), and the TSH-to-T3 ratio (β = 8.33, 95 %CI:2.02,13.9). Associations between POPs and thyroid homeostasis parameters were negative in men with adiposity measures at or above the median, but positive when adiposity measures were below the median. In postmenopausal women, associations by adiposity measures were opposite to those observed in men.

Conclusion: We report associations of select POPs with disruptions in thyroid hormones as well as measures of central and peripheral thyroid hormone sensitivity. Given the persistence of these chemicals in humans and the environment, understanding their effects on thyroid homeostasis pathways may be critical for diabetes prevention.