Altered responsiveness to cortisol, epinephrine, and glucagon in insulin-infused juvenile-onset diabetics. A mechanism for diabetic instability

To evaluate the mechanism of rebound and stress hyperglycemia in diabetes, cortisol, epinephrine, and glucagon were infused into normal and juvenile-onset diabetic subjects for 5-6 hr in doses designed to simulate the increments observed in insulin hypoglycemia or severe stress. Plasma glucose and glucose kinetics were normalized in the diabetics before each experiment by infusion of insulin (0.25-1.0 U/hr) for 4-6 hr, which was then continued during anti-insulin hormone administration. Infusion of cortisol and epinephrine produced comparable elevations in plasma cortisol (35-45 μg/dl) and plasma epinephrine (350-450 pg/ml) in normal and diabetic subjects. However, the increment in plasma glucose in diabetics after cortisol (86 ± 15 mg/dl) and epinephrine (129 ± 12 mg/dl) infusion was 5-fold to 7-fold greater than it was in normals (P < 0.001), despite ongoing insulin infusion. Whereas cortisol failed to alter glucose production in normals, in diabetics there was a progressive 35% increase in glucose production (P < 0.001). Similarly, epinephrine produced only a transient (<90 min) 60% rise in glucose production in normal subjects, but caused a sustained 50%-90% rise in glucose production in diabetics throughout the entire 5-hr study. In contrast, both cortisol and epinephrine infusion was accompanied by a 25%-40% (P < 0.001) decline in glucose clearance, no different than in normal and diabetic subjects. Glucagon infusion produced a transient 35%-40% rise in glucose production, which was of comparable magnitude and duration in normal controls and diabetics. However, the glucagon-induced rise in plasma glucose was 3-fold greater in the diabetics due to failure of endogenous insulin secretion and consequent lack of an increase in glucose clearance as observed in normal controls. When saline was administered to the insulin-infused diabetics, neither plasma glucose nor glucose production were altered from baseline values. It is concluded that (1) the hyperglycemic effect of each of the anti-insulin hormones is increased in diabetics, in spite of prior normalization of plasma glucose concentration, by continuous insulin infusion and (2) hyperglycemia is particularly exaggerated by elevations of cortisol and epinephrine in diabetes as a consequence of an altered response of the liver to these hormones.