A new model to fit glucose concentration during the insulin tolerance test improving the predictive capability to estimate insulin sensitivity.

摘要

The insulin tolerance test (ITT) is a simple method to estimate in vivo insulin action, which consists in measuring glucose levels after intravenous insulin administration and calculating the rate of disappearance of glucose from blood (KITT) with the formula 0.693/t1/2 X 100; tVi is the half-life of glucose concentration . There are two proposals to quantify t1/2, the exponential and the mixed models; both of them are fitted using least-squares analysis. The first one considers an exponential decrement of glucose concentration during the 15min after insulin administration ; the mixed model assumes that the glucose level linearly decreases during the test . We have unpublished data, which, based on experimental evidence, have shown that the glucose concentration does not follow the above-mentioned trends, particularly during the first minutes of the test, suggested by other . Therefore, we propose a new phenomenological model to fit glucose concentrations during the ITT with a predictive capability higher than those obtained by means of the traditionally models used allowing an improvement of the insulin sensitivity estimation.