Abstract P034: Comparison of the Prognostic Value of 1,5-anhydroglucitol (1,5-AG) and the Oral Glucose Tolerance Test (OGTT) in the Atherosclerosis Risk in Communities (ARIC) Study
Introduction: 1,5-AG is a biomarker that reflects hyperglycemic excursions. Unlike the OGTT, 1,5-AG requires only a single blood draw and is a non-fasting test. It is unknown if 1,5-AG could serve as a substitute for OGTT and whether it provides complementary information to fasting glucose (FG) for prediction of diabetes and long-term clinical outcomes. Methods: We included 6,711 ARIC participants without diagnosed diabetes, chronic kidney disease (CKD), and CVD that attended visit 4 (1996-98). Participants were followed for up to 18 years for incident diagnosed diabetes, CKD, CVD, and all-cause mortality. We used Harrell’s C-statistic from Cox models to compare the prognostic value of 1,5-AG to OGTT beyond a base model of demographic factors and body mass index. Restricted cubic splines (4 knots) were used to flexibly model the biomarkers with each of the outcomes. Results: Both OGTT and 1,5-AG provided information beyond the base model for risk discrimination of incident diagnosed diabetes ( p <0.05; Table). However, OGTT provided statistically significantly more information than 1,5-AG (difference in C-statistic: 0.087 (95%CI, 0.075, 0.099)). While 1,5-AG otherwise did not provide more information for future outcomes than the base model, OGTT statistically significantly improved the base model for prediction of CKD, CVD, and all-cause mortality. For incident diagnosed diabetes, inclusion of FG in the models maintained that OGTT provided more information than 1,5-AG for risk discrimination. Inclusion of all three biomarkers (FG, 1,5-AG, and OGTT) was not statistically significantly better than a model with FG and OGTT for future diagnosed diabetes ( p =0.687). Conclusion: 1,5-AG could not sufficiently substitute for the OGTT as a test to identify those at risk of future diabetes. Additionally, glycemic excursions captured by 1,5-AG did not provide additional prognostic value beyond glucose-based tests among those without diagnosed diabetes, suggesting the utility of 1,5-AG is limited to persons with overt diabetes.