The results of the current cross-sectional study in 1275 euthyroid subjects with obesity showed that both FT3 and FT4 levels were positively correlated with BMI and PBF, whereas they were differentially associated with IR and atherogenic lipid profiles, as FT3 and FT4 were positively and negatively correlated with HOMA-IR and atherogenic lipid profiles, respectively. In addition, this study showed that TSH and thyroid antibody positivity were not associated with any metabolic parameters linked to obesity. To our knowledge, this is the largest study with a high number of adult obese (BMI ≥30) participants to evaluate the effect of thyroid hormones in the euthyroid range on metabolic parameters. Our study has shown that thyroid function in the euthyroid range has an effect on determining the metabolic phenotype of obesity.
Various studies on subjects with obesity have reported an increase in FT3 levels in obesity; however, they reported low, normal and high levels of FT4.[4,6,16–19] In our study, subjects with morbid obesity had significantly higher serum FT3 and FT4 compared with grade I obese subjects, but after adjustment for confounders, the significance of the difference was lost. Several theories have been proposed to explain the increased levels of T3 and low, normal and high levels of T4 in obese subjects. Firstly, an increased deiodinase activity with a high conversion rate of T4 to T3 in obese individuals might be responsible for this pattern. Leptin, a hormone produced by adipocytes, alters the activity of deiodinases, thus promoting the conversion of T4 to T3.[26,27] This mechanism may be interpreted as a defence mechanism that restrains the accumulation of fat by increasing the basal metabolic rate and total energy expenditure. Furthermore, it is known that T4 turnover rate is proportional to body size.[27,28] An increased rate of thyroid hormone disposal due to a large body size would promote the activation of the hypothalamus–pituitary–thyroid axis to maintain serum thyroid hormone levels within the euthyroid range. Eventually, this sequence of events would result in a low-normal serum FT4. A second reason might be a reduction of TSH and thyroid hormone receptor expression in adipocytes of obese subjects. Decreased tissue responsiveness to circulating thyroid hormones would cause a consequent increase in TSH and FT3 levels to surpass peripheral resistance.[27,30] Thirdly, the effects of leptin on the hypothalamus might play a role. Obesity is accompanied by leptin resistance. Increased leptin levels have been shown to stimulate thyrotrophin-releasing hormone (TRH) in the hypothalamus. Therefore, the levels of TSH and thyroid hormones would be increased in obesity. Moreover, there is an impaired feedback mechanism in obesity due to a lower number of T3 receptors in the hypothalamus. All of these disturbances are a result of weight excess and studies have shown that they improve after weight loss.[11–13]
In our study, the median TSH level in the antibody-negative group was 1·8 μIU/ml. This value was higher than that found in another study from Turkey, with 408 normal-weight participants aged ≥18 who were negative for thyroid antibodies, which reported a median TSH level of 1·4 μIU/ml.
The association between thyroid function and metabolic parameters is complex. Thyroid hormones act on the liver, white adipose tissue, skeletal muscle and pancreas to influence plasma glucose levels, insulin sensitivity and carbohydrate metabolism. A recent study that evaluated the association between thyroid function and BMI in genetically confirmed obese children showed that FT3 and FT4 had opposing strong correlations with body composition in childhood, with FT3 and FT4 being positively and negatively associated, respectively, with fat mass and BMI. In our study, FT3 and FT4 were both positively associated with BMI and PBF, although their metabolic effects differed. Insulin resistance is distinguished into peripheral and hepatic subtypes. The effect of T3 on hepatocytes is antagonistic to insulin and stimulates glucose production in the liver (gluconeogenesis and glycogenolysis). Low levels of thyroid hormone cause impaired sensitivity of all tissues to insulin. High leptin levels due to a high mass of adipose tissue may cause much more conversion of T4 to T3 by increasing deiodinase activity. Hence, low levels of FT4 and high levels of T3, which is related with an unfavourable metabolic profile, might be a result of the higher levels of adipose tissue and leptin levels.
A previous study demonstrated that in subjects with morbid obesity, serum cholesterol levels were lower than in lean controls with similar degrees of serum TSH elevation, which suggests that a higher serum TSH level in subjects with morbid obesity is not associated with peripheral hypothyroidism. In our study, TC, HDL-C and LDL-C were similar in the upper and lower TSH quantiles. TG was slightly and significantly higher in the upper TSH quantile after adjustment for confounders, although correlation analysis showed no association between TSH and TG. Moreover, logistic regression analysis revealed no association between TSH within the normal range and atherogenic dyslipidaemia. Furthermore, FT3 within the normal range was positively correlated with atherogenic dyslipidaemia and FT4 was negatively associated with atherogenic dyslipidaemia, although the association lost its significance after adjusting for confounders.
In a review investigating the relationship between measures of adiposity and serum TSH within the normal range, 18 of 29 studies showed a positive correlation. In our study, TSH in the normal range was not associated with BMI or PBF. Although adjusted HOMA-IR was higher in the upper quantile of TSH, there was no association between TSH within the normal range and HOMA-IR in logistic regression analysis. Higher HOMA-IR in the upper quantile TSH group may be caused by other factors associated with obesity, which we did not study.
Leptin is an inflammatory cytokine that influences immune reactivity and has been implicated in the increased rates of autoimmune diseases and thyroid autoimmunity in obesity.[6,35] Previous studies have demonstrated that the prevalence of autoimmune thyroid disease in obese individuals was 10% to 16%. In our study, anti-TPO positivity was detected in 14% and anti-TG positivity in 15% of euthyroid subjects with obesity. Anti-TPO and anti-TG positivity were similar in the morbidly obese group compared with the grade I obesity group. Logistic regression analysis showed no association between thyroid antibody positivity, IR and atherogenic dyslipidaemia.
The strengths of this study include the large number of subjects with obesity who did not use medications known to influence metabolism and did not have overt thyroid disease or diabetes mellitus. Furthermore, most results were adjusted for potential confounders such as age, gender, current smoking status and thyroid autoantibody positivity.
The limitations of our study are as follows. First, the cross-sectional study design is not able to establish a definite cause-and-effect relationship. Second, although HOMA-IR correlates well with the hyperinsulinaemic euglycaemic clamp, the gold standard for evaluating IR is still the hyperinsulinaemic euglycaemic clamp.
Clin Endocrinol. 2016;85(4):616-623. © 2016 Blackwell Publishing