Influence of basal androgen levels in euandrogenic women on glucose homeostasis

Objective: To evaluate possible relationships between insulin action and the normal variations of serum androgens in euandrogenic women. Design: Prospective evaluation of insulin action in normal nonobese women using hyperglycemic and euglycemic hyperinsulinemic clamp techniques, correlating insulin action to serum testosterone (T), free T, androstenedione (A), and dehydroepiandrosterone sulfate (DHEAS). Statistical analysis used Spearman's rank correlation. Setting: Yale University Clinical Research Center. Participants: Nonobese females with normal oral glucose tolerance tests, on no medications known to affect glucose metabolism, having the following range of serum androgen levels: T, 0.69 to 3.12 nmol/L; free T, 0.17 to 1.25 nmol/L; A, 2.48 to 11.31 nmol/L; DHEAS, 0.68 to 10.61 μmol/L. Total number of patients studied: hyperglycemic clamps, n = 58; euglycemic hyperinsulinemic clamps, n = 43. Interventions: None. Main Outcome Measures: Pancreatic insulin secretion in response to hyperglycemia and insulin action as assessed by insulin-mediated glucose utilization using the euglycemic, hyperinsulinemic clamp technique. Results: We identified no significant correlation between serum androgens and either glucose uptake or insulin- mediated glucose utilization. Glucose-stimulated insulin release was negatively correlated with serum T and free T throughout the normal range of these hormones. Conclusion: We conclude that, within the normal range, variations of serum androgens are not correlated with changes in the response to insulin. It seems unlikely, therefore, that modest increases of serum androgens within the normal range are responsible for inducing insulin resistance.