We previously proposed the insulin sensitivity index, S1as an absolute measure of whole body tissue sensitivity to insulin. S1 is defined, in the physiologic range of insulin action, as the effect of insulin to augment glucose's ability to reduce its own plasma level. This parameter can be determined from the frequently sampled intravenous glucose tolerance test (IVGTT) by using a digital computer to find the coefficients of a minimal mathematical model of glucose kinetics. In the present study we compared, in normal dogs, S, determined from the IVGTT to an analogous parameter “S1(clamp)” calculated from the euglycemic glucose clamp (EGC). Fifteen pairs of experiments (1 IVGTT and 1 EGC) were performed on 12 animals. IVGTTs: After glucose injection (0.3 g/kg), frequent blood samples were taken over the subsequent 3 h. KG ranged from 1.7 to 4.7 %/min, and integrated insulin from 0.4 to 5.9 (μU/ml) min for 0–60 min. S1 varied over a ninefold range from 1.0 to 9.1 × 10−4 (min−1)/μU/ml), with a mean of 4.3 ± 0.7 × 10−4. Fractional glucose disappearance rate independent of insulin (p1) was 4.3 ± 0.5 %/min. EGCs: Low-rate insulin infusion (8 mU/min from 1 to 150 min) elevated plasma insulin (INS) from 15 ± 4 to 43 ± 10 μU/ml. Glucose was infused (GINF) at 119 ± 20 mg/min to maintain euglycemia. Moderate insulin infusion (40 mU/min: 151–300 min) further elevated plasma insulin (to 176 ± 37 μU/ml) and the requisite glucose infusion (to 372 ± 36 mg/min). S1(clamp) was calculated in steady state as the increase in glucose infusion per increment in insulin above basal, normalized to the basal glucose and body weight [GINF/(Gb × BW × δINS)]. S1(clamp) ranged from 3.1 to 40.5 × 10−4 (dl/kg/min)/μU/ml). The values of S1 and S1(clamp) were highly correlated (r = 0.82, P < 0.005).
These studies suggest that the insulin sensitivity index, S1determined from the IVGTT is a useful measure of insulin sensitivity, which is capable of differentiating sensitivities among a normal population.