OBJECTIVE To investigate the relationship between insulin and adrenal androgens in patients with polycystic ovarian disease (PCOD). METHODS Patients with PCOD and a group of volunteers who attended the department during a period of 6 months were studied. METHODS Department of Gynaecology and Obstetrics, Università Cattolica del Sacro Cuore, Roma, Italy. METHODS Healthy women with ovulatory cycles (hospital personnel, n = 8) and women affected by PCOD (n = 32) were studied on day 5 to 6 of their follicular phase. METHODS All women had an oral glucose tolerance test (OGTT) (75 g) on day 5 to 6 of the cycle. Then plasma samples were collected at 7.00 A.M.; at 11.00 P.M., 2 mg of dexamethasone (DEX) were orally administered with blood samples collected the day after at 7.00 A.M. (effect of DEX). Then adenocorticotropic hormone (ACTH, Synacten; Ciba-Geigy, Varese, Italy) 250 micrograms was injected intravenously (IV) and samples collected 60 minutes later (effect of ACTH). METHODS Plasma glucose and insulin concentration were assayed on OGTT samples collected at time 0, 30, 60, 90, 120, 180, and 240 minutes after glucose ingestion. Data are expressed as area under the curve. Cortisol, 17 alpha-hydroxyprogesterone (17-OHP), testosterone (T), androstenedione (A), and dehydroepiandrosterone sulphate (DHEAS) plasma levels were evaluated on the samples collected before and after DEX or ACTH administration. Data are expressed as absolute concentrations and percent increase in respect to values before the treatment. RESULTS According to the OGTT response, 21 patients were classified as hyperinsulinemic and 11 as normoinsulinemic. The ideal body weight was greater in hyperinsulinemic patients. No differences in baseline hormone levels were found between the two groups. Only sex hormone binding globulin levels were significantly greater in normoinsulinemic patients (P less than 0.05). Also, the plasma concentration of all steroids after DEX were similar in both groups. Intravenous injection of ACTH significantly increased plasma androgens levels. Cortisol, DHEAS, and T enhancement did not differ in normoinsulinemic and hyperinsulinemic patients, whereas significantly greater A (P less than 0.01) and 17-OHP (P less than 0.05) plasma concentrations were observed after ACTH injection in hyperinsulinemic when compared with normoinsulinemic PCOD subjects. Control group after IV ACTH showed an increase of A and 17-OHP similar to those found in normoinsulinemic PCOD group. CONCLUSIONS These data suggest that insulin could be involved in the androgen production by adrenal gland and it could influence the responsiveness of adrenal to its trophic hormones.