Reduced Sampling Schedules for Calculation of an Insulin Sensitivity Index from the Liquid Meal Tolerance Test

Abstract

Previous results from our laboratory indicate the Matsuda insulin sensitivity index (MISI) calculated from a liquid meal tolerance test (LMTT) correlates well with the minimal model insulin sensitivity index (SI). The present analysis evaluated the influence of reduced LMTT sampling schedules on the relationship between MISI and SI. Subjects included 38 men and 31 women, mean±SEM age 35.1±1.0 yr and BMI 32.5±0.5 kg/m2. For the LMTT, subjects consumed a liquid meal (Carnation Instant Breakfast® and whole milk) providing 7 kcal/kg (24% fat, 59% carbohydrate, and 17% protein). SI was calculated from the 30‐ sample, insulin‐modified intravenous glucose tolerance test (MINMOD, v3). The Pearson correlation coefficient between SI and the MISI using values at 0, 30, 60, 90, 120 min was 0.519, p<0.0001. Correlation coefficients for selected reduced sampling schedules were: 0, 60, 120 min (r=0.493); 0, 120 min (r=0.534); 0, 30, 120 min (r=0.476); 0, 30, 60 min (r=0.450), 0, 60 min (r=0.461; all comparisons p<0.001). Correlation coefficients for the full and reduced schedule MISI values were r=0.953 for the 0, 60, 120 min and r=0.909 for the 0, 120 min schedules, respectively (both p≤0.0001). These results suggest that the MISI calculated from reduced sampling schedules (0, 60, 120 min and 0, 120 min) may be useful for application in population studies.