Minimal model for determination of insulin sensitivity: repeatability in control and obese subjects

A mathematical model was recently developed (minimal model) to estimate insulin sensitivity from the analysis of blood glucose and insulin concentration after a modified intravenous glucose tolerance test. This approach was successfully compared with the euglycemic glucose clamp in humans. In this study, we have analysed the distribution and reproducibility of this method in 10 healthy subjects and seven obese patients. We have observed a lower insulin sensitivity (SI) in obese patients (1.96 +/- 0.38 x 10(-4)/min/microU/ml) than in healthy subjects (8.19 +/- 1.16 x 10(-4)/min/microU/ml) with a small overlap between the two groups. The test was repeated twice and good repeatability was found with the method of Blank and Altman when a log transformation was applied to the data.