Many, but not all, girls with high levels of the male hormone testosterone go on to develop polycystic ovary syndrome (PCOS) as adults. Women with PCOS often have irregular menstrual periods, excess facial and body hair, and weight gain. PCOS is also a leading cause of difficulty becoming pregnant. The investigators do not understand why some girls with high hormones develop PCOS and others do not. In a previous study by our group, some girls with high levels of male hormones had abnormalities in the secretion of another hormone, called luteinizing hormone (LH), that are often seen in women with PCOS. However, another group had normal LH secretion. The girls with the abnormal LH secretion had higher levels of another hormone, called insulin, than the girls with normal LH secretion. The investigators will test whether metformin, an insulin-sensitizing agent, changes the effects of high male hormone levels in adolescent girls, specifically by looking at their LH secretion response following metformin treatment.

A better understanding of the factors that make adolescent girls more or less susceptible to the adverse neuroendocrine effects of elevated androgens will hopefully lead to improved prevention and treatment strategies for PCOS. In this study, we propose to explore the role of hyperinsulinemia on neuroendocrine function in hyperandrogenic adolescent girls by assessing the effect of the insulin sensitizer Metformin on hypothalamic progesterone sensitivity. Other differences between the progesterone sensitive and progesterone insensitive subgroups, including racial and ethnic differences between the two populations and a trend towards older gynecologic age in the progesterone insensitive population, are being pursued through other ongoing studies (IRB-HSR# 8588 and 12160).

Inclusion Criteria:

• Girls ages 10 to 17
• Hyperandrogenemic (free testosterone greater than 2.5 standard deviations above the mean for normal control subjects of the same Tanner Stage)
• Creatinine clearance > 90 ml/min as calculated by the Cockcroft-Gault equation
• Hemoglobin > 12 mg/dL or Hematocrit > 36%
• Normal screening labs (with exception of the expected hormonal abnormalities inherent in hyperandrogenemia)
• Sexually active subjects must agree to abstain or use double barrier contraception during the study
• Subjects must agree not to take any other medications during the course of the study without approval by the study investigators.

Exclusion Criteria:

• Abnormal screening labs (with the exception of the expected hormonal abnormalities inherent in hyperandrogenemia)
• Creatinine clearance less than 90 ml/min as calculated by Cockcroft-Gault equation
• Hemoglobin <12 mg/dL or hematocrit < 36%
• Abnormal liver function tests, including Aspartate Aminotransferase (AST), Alanine Aminotransferase (ALT), Bilirubin, Albumin, and Alkaline Phosphatase
• Weight < 34 kg
• History of renal dysfunction, liver dysfunction, congestive heart failure, deep venous thrombosis, breast cancer, endometrial cancer, or cervical cancer
• Pregnant or breast feeding
• On medications known to affect the reproductive axis within 3 months of the study (including oral contraceptive pills, metformin, and spironolactone)
• Are currently participating in another study or have been in one in the last 30 days.
• Subjects using restricted medication (see restrictions below) are excluded unless the subject's primary care provider approves stopping the medication.

25 subjects enrolled and 12 completed the study. Only 10 subjects were included in the formal analysis.

From start of study procedures until 30 days post study drug.