[The clinical estimate of adiposity is not a good predictor of insulin tissue sensitivity measured with a minimal model analysis].

Fifteen male volunteers, aged 30 to 40 years old, were classified according to body mass index (BMI) as lean (n = 5, BMI less than 20 kg/m2), normal (n = 5, BMI 20-25) or obese (n = 5, BMI over 30). Glucose intolerance was ruled out by a normal oral glucose tolerance test and insulin sensitivity (SI) and glucose effectiveness (SG) were estimated by a minimal model analysis of a frequently sampled intravenous glucose tolerance test modified by an intravenous insulin injection at minute 20. The MINMOD program was fed with 29 or 12 values (reduced sampling schedule). Despite a significant inverse correlation between BMI and SI (r = -0.533 p < 0.05), the latter parameter overlapped among groups and the correlation was lost when obese individuals were not considered. Waist/hip ratio correlated modestly with SI (r = -0.52 p < 0.05). SG did not correlate with BMI. Using the reduced sampling schedule. SI values had a correlation coefficient of 0.78 with those calculated using the usual sampling schedule, although they were 82% lower. We conclude that only a BMI of over 30 accurately predicts a low SI, and that waist/hip ratio does not have a better predictive power.