Effect of a Reduced Sampling Protocol on Changes in SI with Exercise Training

Abstract

2267 Purpose: To determine if a reduced blood sampling protocol (13 samples) during an intravenous glucose tolerance test (IVGTT, minimal model analyses) altered the change in the insulin sensitivity index (SI) reported with exercise training. Methods: Thirty-one middle-aged men and women were examined before and after a 6-month exercise training program (50–70% of VO2peak, 3-5 days/wk, 40–60 min/day). An IVGTT was performed when the subjects were sedentary and 24 hours after the final exercise bout. Blood glucose and insulin were determined in the 33 samples obtained during a typical IVGTT and an SI calculated. Plasma insulin and glucose from the 13 samples recommended for a reduced sampling protocol were also utilized to calculate an SI and results compared. Results: SI increased significantly with training (mean ± SE) (P<0.001) when SI was calculated using either the 33- (3.1 ± 0.4 vs. 5.0 ± 0.6 (μU/L)−1 &#8729; min−1) or 13-sample (3.5 ± 0.6 vs. 4.9 ± 0.6 (μU/L)−1 &#8729; min−1) protocol. When comparing the relative change (post/pre) in SI with training, there was no difference between the 33- (1.6 ± 0.1-fold increase) and the 13-sample (1.6 ± 0.2- fold increase) methods. The two methods were related to each other (R=0.72, P<001). However, the minimal model program could not calculate an SI in 11 of the subjects with the 13 sample protocol. Conclusion: The 13-sample IVGTT method with use of the current minimal model program can provide accurate estimates of changes with training in some subjects.