Insulin sensitivity in postmenopausal women - Independent and combined associations with hormone replacement, cardiovascular fitness and body composition

OBJECTIVE - The effects of combined physical activity and hormone replacement therapy (HRT) on insulin sensitivity in postmenopausal (PM) women are unclear. The purpose of the study was to test the following hypotheses: 1) PM women who have undergone vigorous exercise training have greater insulin sensitivity than PM women who are physically active and PM. women who are sedentary and 2) PM women using HRT have greater insulin sensitivity than PM women not using HRT. We also sought to determine whether body composition or cardiovascular fitness was the stronger predictor of insulin sensitivity in these women. RESEARCH DESIGN AND METHODS - Three groups of PM women classified as sedentary (n = 18), physically active (n = 19), and athletic (n = 23) underwent an insulin-modified frequently sampled intravenous glucose tolerance test to determine the insulin sensitivity index (S-l) and dual-energy X-ray absorptiometry to determine body composition. RESULTS - There was a significant association between both physical activity (P = 0.036) and HRT (P = 0.007) and Easting plasma insulin levels. The athletic PM women had the lowest plasma insulin levels and the highest S-l. Across all physical activity levels, PM women using HRT (n = 29) had significantly lower fasting plasma insulin levels and a lower S-l than PM women not using HRT (n = 31), HRT was significantly (P = 0.025) associated with intravenous glucose tolerance (K-G); the women not using HRT had a higher K-G than the PM women using HRT (0.83 +/- 0.08 vs. 0.60 +/- 0.05% per minute). Percent body fat (r = -0.37, P = 0.004) and V-O2max (r = 0.35, P = 0.007) were similar predictors of S-l. CONCLUSIONS - We conclude that, although overall HRT was associated with an attenuated S-l, vigorous exercise training was independently associated with the greatest S-l. In addition, PM women using HRT may benefit from having lower plasma insulin levels, but they may also have a lower S-l.