Metformin effect on TSH in subclinical hypothyroidism: randomized, double-blind, placebo-controlled clinical trial

Non-randomized trials suggest that metformin may reduce TSH levels through unknown mechanisms.

To evaluate whether metformin can reduce TSH levels in subjects with subclinical hypothyroidism.

This is a randomized, double-blind, placebo controlled clinical trial with 3 months duration that enrolled 48 individuals, between 18 and 65 years, with subclinical hypothyroidism. The patients were randomized to the use of metformin 850 mg or placebo twice a day for 3 months. The primary outcome was the absolute decrease in TSH levels. Secondary outcomes were changes in the clinical and laboratory assessment, as well as in blood pressure assessed by ambulatory blood pressure monitoring.

After 3 months, 93.75% of participants completed the follow-up. The post treatment value of TSH in the metformin and placebo groups were 6.48 ± 3.11 and 7.02 ± 3.28 mIU/L, respectively (p = 0.57). Patients who achieved status of euthyroidism in the metformin and placebo groups were 21.7 and 18.2%, respectively (p = 0.76). There was no significant reduction of TSH within the groups [delta for TSH of 0.63 ± 0.56 (p = 0.28) and 0.54 ± 0.60 mIU/L (p = 0.38), in metformin and placebo groups, respectively]. There was a small increase in HDL cholesterol (1.62 ± 0.45 vs. 1.34 ± 0.39 mmol/L, p = 0.03) favoring the metformin group.

Since the sample size was small, the study was inconclusive and the results should be considered preliminary data of a study that needs to enroll 1626 patients to show a 0.5 mIU/L difference in TSH between the groups, with 90% power.