Glucose infusion rates from various complex carbohydrates estimated from two models.

Due to the importance of diet in the management of noninsulin dependent diabetes, we are interested in constructing a model to study the glycemic response to different foods in normal and diabetic subjects. In this study, serum glucose and insulin values from individual normal subjects (Diab. 26, 1178) following the ingestion of different foods (corn, rice, bread, potato) high in complex carbohydrate were analyzed to determine glucose infusion rates (J) into the blood from the gut as a function of time following the ingestion of the food item. A Modular Modeling Program was designed to allow selection of data, curve fit, model, and tuning of parameters. The two models used were from Ackerman et al (Phys. Med. Biol. 9, 203) and the minimal model (VI) from Bergman and Cobelli (Fed. Proc. 39,110). The models were modified so that glucose and insulin values were used as input and J was the output. There was a good correspondence between the shape of the J curves predicted by the two models. While the J curves resembled serum glucose curves, they generally reached a peak earlier and declined more rapidly. Two peaks occurred in the J curves for bread, corn and rice. The second peaks may correspond to a delay in the release of glucose during digestion. The resulting infusion rate characterization of these foodstuffs maybe useful for modeling responses in diabetic patients with normal gut function but with impaired peripheral glucose utilization and in detecting abnormal gut function.