Objective: Metformin treatment may induce a decrease/suppression in serum TSH levels, mimicking sub-clinical hyperthyroidism (SHT). The aim of the present study was to retrospectively evaluate changes in several electrocardiographic indices in euthyroid subjects with diabetes who, after starting metformin treatment, developed a low serum TSH as compared to patients with SHT resulting from an underlying thyroid disease or TSH suppressive treatment with L-thyroxine.
Design: Heart rate, P wave duration, P wave dispersion, QTmax, QTmin and QT-dispersion were assessed in 23 patients with diabetes treated with metformin before and after 6 months of TSH-suppression and in 31 control patients with SHT.
Results: No significant changes in electrocardiographic parameters were observed from baseline to the TSH-suppression measurement. A significant difference in P wave duration (102.9 ± 7.4 vs. 92.1 ± 5.8 ms, p<0.001), P wave dispersion (13.1 ± 3.4 vs. 7.1 ± 3.5 ms, p<0.001), QTmax (399 ± 18 vs. 388 ± 16 ms, p=0.024), QTmin (341 ± 14 vs. 350 ± 17 ms, p=0.038) and QT dispersion (49.9 ± 9.6 vs. 30.9 ± 9.2 ms, p<0.001) were observed between the control group with SHT and the group of diabetic patients with low serum levels of TSH.
Conclusions: Our results show that the TSH-suppressive effect observed in patients taking metformin is not associated with peripheral markers of thyroid hormone excess, at least at the cardiac level.