To test the hypothesis that insulin has a greater effect on glucose metabolism when given as pulsatile than as continuous infusion, a 354-min euglycaemic clamp study was carried out in 8 healthy subjects. At random order soluble insulin was given intravenously either at a constant rate of 0.45 mU/kg \ m=. \min or in identical amounts in pulses of 1\m=1/2\ to 2\m=1/4\ min followed by intervals of 10\m=1/2\to 9\m=3/4\min. Average serum insulin levels were similar during the two infusion protocols, but pulsatile administration induced oscillations ranging between 15 and 62 \g=m\U/ml. Glucose uptake expressed as metabolic clearance rate (MCR) for glucose was significantly increased during pulsatile insulin delivery as compared with continuous administration (270\p=n-\294 min: 8.7 \ m=+-\ 0.7 vs 6.8 \ m=+-\ 0.9 ml/kg \ m=. \min, P < 0.01, and 330\p=n-\354 min: 8.9 \ m=+-\0.5 vs 7.4 \m=+-\0.9 ml/kg \ m=. \ min, P <0.05). The superior efficacy of pulsatile insulin delivery on glucose uptake was not consistently found until after 210 min of insulin administration. In both infusion protocols, endogenous glucose production as estimated by the [3-3H]glucose infusion technique was suppressed to insignificant values. Finally, the effect of insulin on endogenous insulin secretion and lipolysis as assessed by changes in serum C-peptide and serum FFA was uninfluenced by the infusion mode. In conclusion, insulin infusion resulting in physiological serum insulin levels enhances glucose uptake in peripheral tissues in healthy subjects to a higher degree when given in a pulsed pattern mimicking that of the normal endocrine pancreas than when given as a continuous infusion. In man basal plasma insulin oscillates regularly with a mean period of 13 min (Lang et al. 1979). The frequency of oscillations remains stable after intravenous glucose or tolbutamide challenge al¬ though the amplitude increases (Matthews et al. 1983). This contrasts insulin secretion in type II diabetics in whom brief and irregular oscillations are superimposed on fluctuations of long dura¬ tion (Lang et al. 1981). The role of this disorgan¬ ized insulin secretion in the insulin resistance of type II diabetes needs to be determined. Recently, Matthews et al. (1983) demonstrated an aug¬mented hypoglycaemic effect of insulin applied as pulsatile infusion compared with continuous infu¬ sion. It is uncertain whether the improved insulin action was due to effects in the liver or in peri¬ pheral tissues.The present study was performed mainly to determine the impact of pulsatile insulin adminis¬ tration on glucose uptake in healthy subjects. The effect of insulin on lipolysis and endogenous insu¬ lin secretion was also measured. We aimed at insulin levels similar to those observed in the postprandial state and in type I diabetics supplied with insulin through conventional routes.
Subjects and Methods
SubjectsEight healthy males participated in the study. Mean age was 26 years (range 23 to 33 years) and mean weight