Exercise-Induced Improvement in Insulin Sensitivity Is Not Mediated by Change in Cardiorespiratory Fitness

Although exercise-induced improvement in cardiorespiratory fitness (CRF) and reduction in abdominal obesity are putative mechanisms by which chronic exercise attenuates insulin resistance (1), the independent contribution of each is unknown. We therefore investigated whether improvement in abdominal obesity and/or CRF mediates the effect of exercise on insulin sensitivity in abdominally obese adults.

Participants included sedentary, abdominally obese men and women who participated in two previously published exercise interventions (2,3). Participants were randomized to control (n = 18) or supervised aerobic exercise (n = 59) for 3 (women) or 4 (men) months. CRF (VO_2peak) was measured using a maximal treadmill test, abdominal obesity by magnetic resonance imaging and waist circumference (WC), and insulin sensitivity by the hyperinsulinemic-euglycemic clamp.

Simple mediation analysis (Table 1) revealed that with the exception of abdominal subcutaneous adipose tissue (AT) and CRF, all variables were significant mediators of the association between exercise and insulin sensitivity (P < 0.05). In multiple mediation, after adjusting for total AT, visceral AT, subcutaneous AT, and BMI, WC remained a significant mediator of the association (P < 0.05), whereas all other variables were no longer mediators after adjusting for WC (Table 1).

Our primary finding is that exercise-induced change in CRF did not mediate change in insulin sensitivity, whereas the change in WC was a strong mediator independent of change in all other AT depots and/or CRF. Our observations suggest that CRF is a characteristic that improves consequent to increases in physical activity, but does not represent the pathway by which exercise improves insulin sensitivity.

Since improvement in CRF is a central adaptation driven in large measure by improvement in cardiac output, it follows that change in insulin sensitivity, a peripheral adaptation in skeletal muscle glucose uptake, is not associated with changes in CRF as measured by VO_2peak. However, it is possible that alternative measures of CRF, including submaximal adaptations and/or exercise tolerance, may better reflect peripheral adaptations in skeletal muscle and hence be stronger mediators of exercise-induced change in insulin sensitivity than VO_2peak. Consistent with this notion, Sénéchal et al. (4) did not observe a significant association between exercise-induced change in VO_2peak and improvement in glycemic control measured by HbA_1c; however, they did observe a significant association between increase in exercise capacity as estimated by METs and change in HbA_1c.

In a clinical context, our finding here does not negate the importance of measuring CRF in practice as it is well established that CRF measured by VO_2peak is an independent predictor of morbidity and mortality (5). Our finding only suggests that change in CRF measured by change in VO_2peak that occurs in response to exercise does not inform clinicians regarding the individual’s ability to manage blood glucose. Alternative measures of cardiorespiratory adaptations to exercise, including submaximal and exercise tolerance, may be more suitable to interpret short-term adaptations in insulin-mediated glucose uptake in response to exercise.