Insulin secretion and hepatic extraction in humans by minimal modeling of C-peptide and insulin kinetics.

Methods for measuring insulin secretion and hepatic insulin extraction in vivo, e.g., hepatic vein catheterization, are invasive, and can be applied during steady state only. We introduce a noninvasive method for measuring in vivo insulin secretion and its extraction by the liver during an intravenous glucose tolerance test (IVGTT). This method is based on a minimal model of C-peptide secretion and kinetics that is used for interpreting plasma C-peptide concentration data during an IVGTT in normal humans. The model allows the reconstruction of the time course of insulin secretion and, used in conjunction with a minimal model of insulin delivery and kinetics (described in a previous study), provides a noninvasive measure of the time course of hepatic insulin extraction [H(t)]. The C-peptide model also provides a direct prehepatic measure of beta-cell sensitivity to glucose, expressed by two parameters related to first (phi IC)- and second (phi IIC)-phase insulin secretion. In the 11 healthy volunteers we studied, these parameters were 61 +/- 11 pM.min-1.mg-1.dl and 0.0154 +/- 0.0034 pM.min-2.mg-1.dl, respectively. H(t) showed an initial decrement for approximately 30-50 min (from a fasting value of 63 +/- 8% to a nadir of 53 +/- 9%) after the glucose stimulus, then a steady value of approximately 62% was reestablished and maintained throughout the experiment. The validity of the C-peptide model was further assessed by comparing its estimate of the fractional plasma clearance rate (k01) with that obtained in experiments in which biosynthetic human C-peptide was administered.