Diverse pathogenesis pathways to type 2 diabetes make optimal care and prevention challenging. Identifying heterogeneous metabolic parameters of patients would be beneficial for developing personalized therapies and prediction of future glycemia. As such, we developed a mathematical model based on Ha et al. Endo 2016,) to extract metabolic parameters of patients and predict their future glycemic trajectories. We fitted longitudinal Oral Glucose Tolerance Test data from a cohort of Pima Indians to predict future glycemic trajectories. First, we estimated three major metabolic parameters from a single OGTT: peripheral insulin sensitivity, hepatic insulin sensitivity, and beta-cell function. Insulin sensitivity estimated by the fitting algorithm correlated well with insulin sensitivity measured by insulin clamp and MINMOD, R2 =0.5 in both cases. Second, we used the fits to two OGTTs separated in time by several years to estimate capacity of beta-cell function to compensate for insulin resistance. Third, we found a strong correlation between an individual’s BMI and insulin sensitivity, R2 =0.75. Using a range of projected BMIs, the model predicts future glycemia. When measured BMIs from later time points were used, the future glycemic trajectory was accurately predicted. The mathematical model has great potential for clinical application in guiding therapy.
J. Ha: None. P.H. Chen: None. A. Sherman: Other Relationship; Self; Eli Lilly and Company.