Glucose kinetics during a meal: one vs two compartment minimal model

During IVGTT the two compartment model performs better than the single compartment one. Here the aim is to test 1 vs 2 compartment model during a more gentle perturbation such an oral test. 1 and 2 compartment minimal model, both unlabeled and labeled, were identified with glucose rate of appearance, i.e. the input of the model, assumed known from a model-independent method. For unlabeled data both models perform similarly while for labeled data the 2 compartment model is better. Among model-based metabolic indices differences were noted in the inhibition of endogenous glucose production due to glucose itself predicted by one and two compartment models.