Validation of the low dose short insulin tolerance test for evaluation of insulin sensitivity

OBJECTIVES The assessment of insulin sensitivity requires an accurate and reproducible technique. The short insulin tolerance test is a simple and rapid method for screening large numbers of subjects when the fasting glucose level is normal. Conventionally, an insulin dose of 0·1 units/kg is used, but this may result in symptomatic hypoglycaemia in healthy thin subjects who are insulin sensitive. In order to overcome this problem we have employed a lower dose of insulin and have studied the reproducibility of this modified technique comparing it with the euglycaemic hyperinsulinaemic clamp. DESIGN Subjects were studied on two separate occasions, once by a short insulin tolerance test and on a second occasion by either a euglycaemic hyperinsulinaemic clamp (insulin infusion of 40 mU/m²/min) or a repeat short insulin tolerance test. PATIENTS Eleven healthy subjects were studied twice with a short insulin tolerance test. A further 10 healthy subjects received a short insulin tolerance test on one day and a euglycaemic hyperinsulinaemic clamp study on another occasion. MEASUREMENTS Insulin sensitivity was measured in the short insulin tolerance test using the slope of arterialized blood glucose concentration from 3 to 15 minutes after an intravenous bolus of short-acting insulin, 0·05 units/kg body weight. In the clamp study, insulin sensitivity was derived from the average amount of glucose infused at steady state (M) and the mean plasma insulin level (l). RESULTS In the short insulin tolerance test no subject developed symptomatic or biochemical hypoglycaemia, defined as a blood glucose < 2·2 mmol/l. The (mean ± SEM) insulin sensitivities for the 11 subjects studied twice were 174±10 and 179±11 μmol/l/min with a coefficient of variation of 6·9±2·6%. There was a close correlation between insulin sensitivity derived from the short insulin tolerance test and that obtained from the euglycaemic clamp studies (so-called M/l ratio) in the same subjects (r= 0·81; P < 0·005). CONCLUSION The short insulin tolerance test employing 0·05 units/kg insulin is a safe, valid and reproducible method for the assessment of insulin sensitivity.