Thyroid-stimulating hormone levels within the reference range are associated with serum lipid profiles independent of thyroid hormones.

Furong Wang,Yinyin Tan,Chenggang Wang,Xu Zhang,Yuanfei Zhao,Xinhong Song,Bingchang Zhang,Qingbo Guan,Jin Xu,Ji Zhang,Dongzhi Zhang,Haiyan Lin,Chunxiao Yu,Jiajun Zhao
DOI: https://doi.org/10.1210/jc.2012-1133
2012-01-01
Abstract:Context and Objective: Dyslipidemia in thyroid dysfunction has always been attributed to changes in thyroid hormone (TH) levels. We hypothesized that TSH plays an important role in lipid metabolism independent of TH. Design and Setting: We conducted a cross-sectional study to investigate the relationship between serum TSH levels and lipid profiles after controlling for free T-3, free T-4, total T-3, total T-4 and nonthyroid factors relevant to lipid metabolism in euthyroid Chinese subjects. Main Outcome Measures: General linear analysis was performed to determine whether the impact of TSH on serum lipid levels is independent of the TH levels. Moreover, path analysis, an evolutionary multivariable regression technique, was conducted to test whether there is a direct and/or indirect effect between serum TSH and total cholesterol (TC) levels. Additionally, the odds ratios (95% confidence interval) for hypercholesterolemia in relation to TSH categories were calculated. Results: A total of 3664 euthyroid subjects were finally analyzed. There was a significant linear trend toward higher log TC (P = 0.021) and log triglyceride (P = 0.001) levels with increasing serum TSH levels within the reference range, which remained significant after adjusting for factors such as TH levels, age, and smoking. Most importantly, the total effect of TSH on TC levels (total effect(TC,) (TSH) = 0.05253) includes a direct effect (direct effect(TC,) (TSH) = 0.05979) and an indirect effect via TH. Compared with subjects in the lower part of the reference range (TSH level, 0.27-0.61 mIU/liter), the adjusted odds ratio for hypercholesterolemia was 3.239 (95% confidence interval, 1.392-7.538; P = 0.007) for those in the upper category (TSH level, 4.61-5.5 mIU/liter). Conclusions: The variation in normal TSH levels is partially related to the lipid components and hypercholesterolemia in euthyroid subjects and includes both TH-dependent and TH-independent effects. Our study suggests the importance of controlling TSH in hypothyroid subjects. (J Clin Endocrinol Metab 97: 2724-2731, 2012)
What problem does this paper attempt to address?