Estimates ofIn VivoInsulin Action in Man: Comparison of Insulin Tolerance Tests with Euglycemic and Hyperglycemic Glucose Clamp Studies*

We compared estimates of in vivo insulin action derived from insulin tolerance tests (ITT) and euglycemic and hyperglycemic glucose clamp studies in 17 normal subjects and 19 patients with various diseases characterized by insulin resistance. Fifteen subjects underwent an ITT and a euglycemic clamp study, 17 subjects underwent an ITT and a hyperglycemic clamp study, and 4 subjects underwent all 3 tests. The ITT consisted of a bolus iv injection of regular insulin (0.1 U/kg BW). The plasma glucose disappearance rate during the 3- to 15-min period following the insulin injection was taken as a measure of insulin action. In both euglycemic and hyperglycemic clamp studies, which were carried out with standard techniques, the ratio between the amount of glucose infused to maintain glycemia at the desired level and the mean plasma insulin concentration from 60-120 min (M) (euglycemic clamp studies) or 20-120 min (I) (hyperglycemic clamp studies) was used as a measure of insulin action. A close correlation was found between plasma glucose disappearance rate and the M/I ratio during either the euglycemic (r = 0.811; P less than 0.001) or the hyperglycemic (r = 0.826; P less than 0.001) clamp studies. These results suggest that the 15-min ITT is suitable as a simple and rapid estimation of in vivo insulin action when glucose clamp studies are not feasible, as in large series of subjects or serial studies.