Enteral infusion of glucose at rates approximating EGP enhances glucose disposal but does not cause hypoglycemia

Portal infusion of glucose at rates approximating endogenous glucose production (EGP) causes paradoxical hypoglycemia in wild-type but not GLUT2 null mice, implying activation of a specific portal glucose sensor. To determine whether this occurs in humans, glucose containing [3-³H]glucose was infused intraduodenally at rates of 3.1 mg·kg^-1·min^-1 (n = 5), 1.55 mg·kg^-1·min^-1 (n = 9), or 0/0.1 mg·kg^-1·min^-1 (n = 9) for 7 h in healthy nondiabetic subjects. [6,6-²H₂] glucose was infused intravenously to enable simultaneous measurement of EGP, glucose disappearance, and the rate of appearance of the intraduodenally infused glucose. Plasma glucose concentrations fell (P < 0.01) from 90 ± 1 to 84 ± 2 mg/dl during the 0/0.1 mg·kg^-1·min^-1 id infusions but increased (P < 0.001) to 104 ± 5 and 107 ± 3 mg/dl, respectively, during the 1.55 and 3.1 mg·kg^-1·min^-1 id infusions. In contrast, insulin increased (P < 0.05) during the 1.55 and 3.0 mg·kg^-1·min^-1 infusions, reaching a peak of 10 ± 2 and 18 ± 5 μU/ml, respectively, by 2 h. Insulin concentrations then fell back to concentrations that no longer differed by study end (7 ± 1 vs. 8 ± 1 μU/ml). This resulted in comparable suppression of EGP by study end (0.84 ± 0.2 and 0.63 ± 0.1 mg·kg^-1·. min^-1). Glucose disappearance was higher (P < 0.01) during the final hour of the 3.1 than 1.55 mg·kg^-1·min^-1 id infusion (4.47 ± 0.2 vs. 2.6 ± 0.1 mg·kg^-1·min^-1), likely because of the slightly, but not significantly, higher glucose and insulin concentrations. We conclude that, in contrast to mice, selective portal glucose delivery at rates approximating EGP does not cause hypoglycemia in humans.