Time to Glucose Peak during an Oral Glucose Tolerance Test Identifies High Prediabetes Risk: Results from a Multiethnic Study

Presentation Number: OR14-1
Date of Presentation: April 3rd, 2017

Stephanie T Chung*1, Mirella Galvan De La Cruz1, Kannan Kasturi2, Brianna A Bingham1, Anthony Onuzuruike1, Rafeal L Baker1, Jean N Utumatwishima3, Lilian Mabundo1, Madia Ricks4, Joon Ha3, Arthur S Sherman3 and Anne E Sumner4
1NIDDK, NIH, Bethesda, MD, 2National Institutes of Health, Bethesda, MD, 3NIDDK NIH, Bethesda, MD, 4NIDDK/NIH, Bethesda, MD

Abstract

Beyond the diagnostic utility of the oral glucose tolerance test (OGTT), specific morphological characteristics that can be easily derived for population analyses may reflect different metabolic phenotypes of β-cell function relative to insulin sensitivity. Time to glucose peak is a reproducible parameter that could provide additional prediabetes risk stratification. Therefore, we determined the predictive ability of the time to glucose peak for prediabetes and the relationship of glucose peak with β-cell function relative to insulin sensitivity. A standard OGTT with glucose, insulin and C-peptide determined at 0, 30, 60, 90 and 120 min was performed in a multi-ethnic study of 128 adults who self-identified as healthy (47 African-American, 57 African immigrant, 24 white, 79% female, age 43±9y (mean±SD), range 24-62y, BMI 29.2±5.3 kg/m2, range 19.9-45.2 kg/m2). Participants were divided into 2 groups: glucose peak at 30 min vs. glucose peak >30min. The glucose area under the curve (AUC) during the OGTT was calculated using the trapezoid rule and prediabetes defined using ADA 2016 criteria. One to two weeks later, an insulin-modified frequently sampled intravenous glucose tolerance test (IM-FSIGT) was performed. Insulin sensitivity (SI) was calculated by the minimal model and β-cell function measured by both the acute insulin response to glucose (AIRg) and the disposition index (DI). Glucose peak >30 min occurred in 60% (76/128) of participants. Prediabetes was identified in 34% (43/128) and the odds of having prediabetes was 5-times higher if glucose peak was >30min (OR 5.26, 95%CI 2.1-13.1, P<0.01). Glucose peak >30 min was associated with older age (44±9 vs. 41±9y, P=0.03), higher BMI 30.1±5 vs. 27.9±5kg/m2, P=0.02), higher AUC (17011±2615 vs. 14128±2000 mg/dl•min, P=<0.01), lower SI (2.5±1 vs. 3.4±2 x10-4 min-1 (µU/ml)-1, P=0.03). Adults with glucose peak >30 min also had lower β-cell function; lower AIRg (672±530 vs. 982±685 µU/mL, P<0.01) and lower DI (1378±858 vs. 2584±1311, P<0.01). In multiple regression models, time to glucose peak and AUC were independent predictors of DI (adjR2=0.38, P<0.01) and this relationship persisted when individuals with prediabetes and normal glucose tolerance were analyzed separately (prediabetes: adjR2=0.25, P=0.03; normal: adjR2=0.39, P=0.03). Time to glucose peak during an OGTT is a strong indicator of beta cell function relative to insulin sensitivity and is independent of overall glycemia during the test. In short, time to glucose peak during the OGTT may be an important parameter for prediabetes risk stratification, especially in studies in which complex metabolic phenotyping is not feasible.

 

Nothing to Disclose: STC, MG, KK, BAB, AO, RLB, JNU, LM, MR, JH, ASS, AES