Mediation by Thyroid Hormone in the Relationships Between Gestational Exposure to Methylmercury and Birth Size

Abstract

Our previous studies have linked gestational methylmercury exposure, originating from seafood, to changes in maternal thyroid hormones and infant birth size in a Swedish birth cohort. Herein we aimed to determine associations between maternal thyroid hormones and infant birth size and elucidate if maternal hormones could mediate the relationship between methylmercury and lower birth size. In 515 women, without known thyroid disease, we assessed metal exposure by erythrocyte mercury concentrations (mainly methylmercury, reflecting exposure over the past months) in early third trimester measured with inductively coupled plasma-mass spectrometry. Plasma concentrations of total and free thyroxine (tT4 and fT4) and triiodothyronine (tT3 and fT3), and thyroid-stimulating hormone (TSH) were measured at an accredited clinical laboratory. In multivariable-adjusted linear regression models, maternal tT3 (per 1 nmol/L increase) was positively associated with birth weight (B: 125 g; 95% CI 36, 214) and length (B: 0.59 cm; 95% CI 0.21, 0.97). Maternal fT4 was inversely associated with birth weight (B: − 33 g; 95% CI − 57, − 9.5), driven by obese women (n = 76). Causal mediation analyses suggested that a doubling of erythrocyte mercury (> 1 µg/kg; n = 374) was associated with a mean tT3-mediated decrease in birth weight of 11 g (95% CI − 25, − 1.6) and in birth length of 0.1 cm (95% CI − 0.12, − 0.01), both equivalent to about 12% of the total effect. To conclude, tT3 was positively associated with infant birth size. Reduced tT3 levels appeared to mediate a minor part of the inverse association between methylmercury exposure and birth size.