MON-212 Relevance of Fasting and Oral Glucose Tolerance Testing (OGTT)-derived Metrics of Insulin Resistance on Capturing the Impact of Glucoregulatory Status on Reproductive Disturbance in Women

Rationale. Hyperandrogenism, a cardinal feature of polycystic ovary syndrome (PCOS), is associated with metabolic comorbidities, particularly insulin resistance (IR). Compensatory hyperinsulinemia is posited to augment ovarian androgen production, either via increased luteinizing hormone (LH) or directly on ovarian thecal cells. The degree to which this pathophysiology can be detected in women without extreme manifestations of hyperinsulinemia is unclear, and contributes to uncertainty regarding the impact of IR on ovarian folliculogenesis. Further, the use of numerous surrogate indices of IR have likely contributed to the conflicting evidence on the strength and directionality of associations between IR, androgen production and ovarian follicle populations. Objective. To investigate the degree to which fasting- and OGTT-derived surrogate indices of IR are associated with LH, total testosterone (TT) or ovarian morphology in a sample of women spanning regular cycles, normoandrogenic anovulation (NA-Anov) and hyperandrogenic anovulation (HA-Anov). Methods. Women with HA-Anov (biochemical and/or clinical HA and irregular cycles; N=44), NA-Anov (irregular cycles and absence of clinical and/or biochemical HA; N=31) and controls (regular cycles and absence of clinical or biochemical HA; N=42) underwent a transvaginal ultrasound, OGTT and physical exam to assess hirsutism and anthropometry. Follicle number per ovary (FNPO) were determined from offline assessments using the Grid Method. Blood samples were analyzed for glucose, insulin, LH, and TT. The degree of association between IR metrics and LH, TT and FNPO 2-9mm were evaluated using Pearson correlation coefficients or partial correlations, accounting for BMI where appropriate. Results. AUC(insulin) and WBISI, as OGTT-derived indices of IR, were associated with increased LH (P<0.05) and TT (P<0.05) whereas the Stumvoll indices (original and Tosi) were only associated with LH (P<0.05). None of the OGTT-derived measures were associated with FNPO. In contrast, no fasting measure of IR was associated with LH and only HOMA-IR showed a weak correlation with TT (P<0.05). Further, increasing IR as judged by HOMA-IR, QUICKI and FGIR were associated with fewer follicles (P<0.05). Conclusions. OGTT-derived measures reflect perturbations in LH and TT production, versus fasting measures which more closely associate with changes in ovarian morphology. These data suggest that markers that capture peripheral IR and compensatory hyperinsulinemia best align with endocrinological disturbances whereas changes in ovarian morphology may parallel hepatic IR and β-cell function which reflect more progressed disruptions in glucoregulatory status. Support: Cornell, CIHR, NIH/NCATS, USDA