SAT-245 Estimated Insulin Sensitivity Score Predicts Post-OSTT Insulin Secretion and GI Hormone Differences in Adolescents with Obesity and PCOS

Abstract

Women with polycystic ovarian syndrome (PCOS) have an increased risk for type 2 diabetes and obesity increases this risk. Insulin resistance (IR), along with beta-cell failure, is a precursor for diabetes, yet is hard to measure clinically. The estimate of insulin sensitivity (eIS) score (e^[4.647 -0.0203 (waist, cm) - 0.0977 (HbA1c, %) -0.00235 (TG, mg/dl)]) that we previously developed in adolescents utilizes clinical measures. An eIS of <6.3 in obese girls with PCOS represents the ≤50th percentile of insulin sensitivity assessed by hyperinsulinemic euglycemic clamp. Past attempts to prevent progression of dysglycemia in obese youth have suffered from a lack of understanding of postprandial responses across the spectrum of IR. Our goal was to assess differences in post-oral glucose load responses in a cohort of obese girls with PCOS stratified by eIS score. We recruited 55 untreated obese girls with PCOS (age 16 ± 0.2 years, BMI Z-score 2.1 ± 0.04, free testosterone 9.3 ± 0.7 mg/dL). Fasting lipids and hormones were measured followed by a frequently-sampled 6-hr oral sugar tolerance test (OSTT, 75g glucose +25g fructose). Participants were categorized as eIS < or ≥ 6.3. Baseline and post-OSTT responses were compared, including impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) categories. eIS group differences were compared by t-test, and metabolite curves and timepoint differences by repeat measures ANOVA. In this cohort, 67% (N=37) had an eIS score <6.3. The low eIS group was more obese [BMI Z-score 2.23 (2.04, 2.41) vs. 1.77 (1.62, 2.02); p<0.001] and more Hispanic (57% Hispanic vs 16% Hispanic, p=0.01), but was similar in terms of PCOS clinical severity (acne, hirsutism, menses/year and free testosterone). The groups had similar proportions of IFG (13% vs 12%, p=0.80; <6.3 vs ≥6.3) and IGT (49% vs. 29%, p=0.30), but the glucose curve trended to be greater in eIS<6.3 (p=0.05) in particular at times 60 and 90 min. The insulin and c-peptide curves were greater in eIS<6.3 (p=0.01, p=0.002), and the insulin concentrations from timepoints 45-150 min were greater in eIS<6.3. The insulin peak was delayed to 120 min in eIS <6.3 vs. 30 min in eIS≥6.3. The glucagon curve was also greater eIS<6.3 (p=0.01), but the GLP-1 curves were not different (p=0.61).Obese girls with PCOS who had an eIS<6.3 had expected hypersecretion of insulin and c-peptide. The delayed insulin peak and trend for increased glucose at 60-90 min suggest a relatively inadequate first-phase insulin release relative to the degree of IR. Those with a low eIS also had evidence of alpha-cell dysfunction. Insulin hypersecretion was not mediated by increased GLP-1. Understanding these early hormone differences may be key to the development of strategies to prevent progression of dysglycemia in this high-risk population. Funding: BIRCWH K12, NIDDK K23, Boettcher Webb Waring, Doris Duke FCRS, Colorado CTSI, Children’s Hospital Colorado