The role of endogenous androgens in enhancing the body's protein anabolic capacity has been controversial. To examine this question we chose to study whole-body protein and glucose kinetics in a group of 21 young, postpubertal females (16.3 +/- 0.6 yr), 8 of whom had clinical and laboratory evidence of ovarian hyperandrogenism (OH) (BMI = 37.8 +/- 1.3 kg/m2). We used L-[1-13C]leucine and [6,6,2H2]glucose tracer infusions before and after suppression of their endogenous androgens with estrogen/progesterone supplementation in the form of Triphasil for 4 weeks. Their baseline data were also compared with those of similar aged girls, 7 obese (OB) (BMI = 36.4 +/- 1.5) and 6 lean (LN) (BMI = 20.9 +/- 0.7) who were normally menstruating and had no evidence of androgen excess. Despite comparable glucose concentrations, both OH and OB groups had significant hyperinsulinemia (OH > OB), both basally and after iv glucose stimulation, as compared to LN controls (basal insulin: OH, 252 +/- 52 pmol/L; OB, 145 +/- 41; LN, 60 +/- 9, P = 0.009 OH vs. LN; peak insulin: OH, 2052 +/- 417; OB, 1109 +/- 127, LN, 480 +/- 120, P = 0.0009 OH vs. LN). The rate of appearance (Ra) of glucose, a measure of glucose production, was greater in the LN controls than in the OH or OB groups (OH, 2.0 +/- 0.1 mg/kg.fat free mass.min; OB, 1.9 +/- 0.1; LN, 3.3 +/- 0.1, P < 0.004 vs. LN). Calculated total rates of whole-body protein breakdown (leucine Ra), oxidation, and protein synthesis (nonoxidative leucine disposal) were substantially higher in the OH and OB groups as compared with LN controls (P < 0.04 vs. LN); however, when data are expressed on a per kilogram of fat free mass basis, the OH group had higher rates of proteolysis than the OB and LN, with indistinguishable rates between the latter two groups. None of the above-mentioned parameters changed after 1 month of administration of Triphasil, despite marked improvement in circulating testosterone and free testosterone concentrations after treatment (testosterone, -50%, P = 0.003; free testosterone, -70%, P = 0.02). We conclude that obesity in young postpubertal females is associated with insulin resistance for both peripheral carbohydrate and protein metabolism, and that patients with the OH syndrome have even greater insulin resistance as compared with simple obesity, regardless of treatment for the androgen excess. Carefully designed studies targeting interventions to improve both the hyperandrogenic and hyperinsulinemic state may prove useful even in the early juvenile stages of this disease.