The effect of dehydroepiandrosterone on insulin resistance in patients with impaired glucose tolerance.

OBJECTIVE Dehydroepiandrosterone (DHEA) and Dehydroepiandrosterone-sulfate (DHEA-S) are the most abundant steroid hormones in the body. Recently, DHEA-S has gained interest as an antidepressant substance, with positive effects on autoimmune disease such as lupus and ulcerative colitis, as well as obesity, cancer, cardiovascular disease and diabetes. Its effect on insulin resistance is also assumed to be positive, but has not as yet been confirmed. The present cross-over clinical trial was conducted to evaluate the efficacy of DHEA and placebo on insulin resistance. DESIGN Participants were selected among relatives of diabetic patients who were referred to the Isfahan Endocrine Research Center because of Impaired Glucose Tolerance (IGT) test. Thirty IGT patients were treated randomly with DHEA (50 mg/day) or placebo by cross-over clinical trial for six months and insulin resistance between the beginning and the end of each three months treatment period was assessed. RESULTS At the end of the first three months, the mean changes from baseline of the various parameters in the drug group were: DHEA-S, 2.5 µmol/l (p=0.008); Homeostatic Model Assessment of Insulin Resistance (HOMA-IR), 0.6 (p=0.6); insulin, 7.1 pmol/l (p=0.3) and FPG, 0.5 mmol/l (p=0.1). The changes in the placebo group were: DHEA-S, 0.08 µmol/l (p=0.6); HOMA-IR, 0.9 (p=0.03); FPG, 0.8 mmol/l (p=0.1); insulin, 25.1 pmol/l (p=0.05). In the second three months, the mean changes in the drug group were: DHEA-S, 4.5 µmol/l (p=0.003); Fasting Plasma Glucose (FPG), 0.1 mmol/l (p = 0.4); insulin, 4.3 pmol/l (p=0.2); HOMA-IR, 0.3 (p=0.1) and the changes in placebo group were: DHEA-S, 0.7 µmol/l (p=0.5); FPG, 0.3 mmol/l (p=0.3); insulin, 10.7 pmol/l (p=0.1); HOMA-IR, 0.6 (p=0.03). CONCLUSION DHEA did not reduce insulin resistance, although there was a tendency to improvement. The data indicate a possible but not clearly favorable effect of DHEA on insulin resistance.