Project Summary Polycystic ovary syndrome (PCOS) is the commonest endocrine disorder in reproductive age women. In addition to gynecologic and reproductive problems, it is well established that women with PCOS have significant long term cardiometabolic risk. Evidence of hyperandrogenism is key to appropriate counselling and management due to its association with increased metabolic risk. There is significant heterogeneity amongst patients, as hyperandrogenism is diagnosed by elevated levels of classic androgens of ovarian source namely, total and free testosterone and/or high Ferriman-Gallwey (FG) hirsutism score. This leads to patient dissatisfaction with both the diagnostic experience and counselling. Recently, novel 11-oxygenated (oxy) androgens and their precursors have been shown to be elevated in sera from women with PCOS raising the possibility that both ovarian and adrenal androgens might be significant contributors to this syndrome. Moreover, the enzyme AKR1C3 responsible for the conversion of androgen precursors to potent 11keto- androgens, is increased in the adipose tissues of women with PCOS. As yet, there is no information on the impact of first-line treatments such as oral contraceptive pills (OCPs), metformin and lifestyle modifications (LSM) on serum 11-oxyandrogens levels. This data will be essential to guide the optimal choice of therapies based on the androgen metabolome. We propose to use residual de-identified serum samples from three NIH funded clinical trials for two aims. Aim 1. Determine differences in serum levels of 11-oxyandrogens and classic androgens in three PCOS phenotypes compared to controls using a validated panel of 12 ∆4 steroids. Aim 2. To determine the change in 11-oxyandrogens after interventions with OCPs, metformin or LSM and determine the correlation between change in 11-oxyandrogens and change in cardiometabolic risk factors. We hypothesize that women with elevated FG scores but normal classic androgens levels will exhibit an increase in serum 11-oxyandrogens compared to controls, thereby explaining the observed clinical hyperandrogenism. Also, women with irregular menses, polycystic ovary morphology but normal classic androgens will exhibit an increase in less potent 11-oxyandrogens. This deep androgen phenotyping, if validated in larger and diverse cohorts, will improve the precision of current diagnostic criteria for PCOS. Further, we hypothesize that metformin and LSM will significantly decrease 11-oxyandrogens by improving insulin resistance and altering AKR1C3 activity in the adipose tissue. On the other hand, OCPs will likely have a modest effect on serum levels of 11-oxyandrogens due to inadequate suppression of adrenal precursors. The proposed examination of a broad panel of adrenal androgens and their precursors and assessment of their correlations with several cardiometabolic markers will allow optimization of precision therapy in different PCOS phenotypes.