Variations in Serum Free Thyroxine Concentration Within the Reference Range Predicts the Incidence of Metabolic Syndrome in Non-Obese Adults: A Cohort Study

Background: The association of changes in thyroid hormone values over time with the incidence of metabolic syndrome (MetS) has not yet been evaluated. For the first time, this study assessed the effect of thyroid hormone variations in the subclinical and euthyroid range on the incidence of MetS and its components over a 10-year follow-up in an adult population. Methods: Data were analyzed from the prospective population-based Tehran Thyroid Study. Of 5786 randomly selected subjects aged ≥20 years, after excluding subjects with MetS (n = 1403), those with serum thyrotropin (TSH) >10 or <0.1 mIU/L (n = 104), those taking thyroid drugs (n = 85) or corticosteroids (n = 97), those with a body mass index (BMI) <18.5 kg/m2, those with a glomerular filtration rate <30, and those with a history of cancer (12), data for 2393 subjects were analyzed. Body weight, waist circumference, and blood pressure were measured, and serum concentrations of lipids and lipoproteins, fasting blood glucose, insulin, free thyroxine (fT4), and TSH were assayed at baseline and during three follow-up studies at three-year intervals. MetS was determined using definition of the Joint Interim Statement, adjusted for the Iranian population. Results: An increase in fT4 values overtime was associated with lower odds of abdominal obesity (odds ratio [OR] = 0.49 [confidence interval (CI) 0.35–0.69]) and hypertriglyceridemia (OR = 0.57 [CI 0.41–0.78]), and with higher odds of hypertension (OR = 1.35 [CI 1.05–1.74]), adjusted for age, sex, smoking, BMI, and Homeostasis Model Assessment Index for Insulin Resistance. fT4 was associated with lower odds of MetS in the crude model, and after adjustment for age, sex, and smoking (OR = 0.59 [CI 0.39–0.9]). This association lost its significance after further adjusting for BMI. In a subgroup analysis of obese (i.e. BMI ≥30 kg/m2) and non-obese (i.e., BMI <30 kg/m2) subjects, fT4 was a significant predictor of MetS only in non-obese subjects after adjusting for age, sex, and smoking (β = 0.49 [CI 0.29–0.83], p = 0.007) and also after further adjustment for Homeostasis Model Assessment Index for Insulin Resistance (β = 0.57 [CI 0.34–0.96], p = 0.03). Serum TSH variations over time were not associated with any of the MetS components or with odds of MetS. Conclusion: A decrease in serum fT4 values is associated with an increased risk for MetS, especially in non-obese adults.