Cardiorespiratory Fitness and Incidence of Type 2 Diabetes in United States Veterans on Statin Therapy

Document Type

Journal Article

Publication Date



American Journal of Medicine








Cardiorespiratory fitness; Statins; Type 2 diabetes incidence


© 2017 Background Impact of cardiorespiratory fitness on statin-related incidence of type 2 diabetes has not been assessed. We assessed the cardiorespiratory fitness and diabetes incidence association in dyslipidemic patients on statins. Methods We identified dyslipidemic patients with a normal exercise test performed during 1986 and 2014 at the Veterans Affairs Medical Centers in Washington, DC or Palo Alto, Calif. The statin-treated patients (n = 4092; age = 58.8 ± 10.9 years) consisted of 2701 Blacks and 1391 Whites. None had evidence of type 2 diabetes prior to statin therapy. We formed 4 fitness categories based on age and peak metabolic equivalents achieved: Least-fit (n = 954), Low-fit (n = 1201), Moderate-fit (n = 1242), and High-fit (n = 695). The non-statin-treated cohort (n = 3001; age = 57.2 ± 11.2 years) with no evidence of type 2 diabetes prior to the exercise test served as controls. Results Diabetes incidence was 24% higher in statin-treated compared with non-statin-treated patients (P <.001). In the statin-treated cohort, 1075 (26.3%) developed diabetes (average annual incidence rate of 30.6 events/1000 person-years). Compared with the Least-fit, adjusted risk decreased progressively with increasing fitness and was 34% lower for High-fit patients (hazard ratio [HR] 0.66; 95% confidence interval [CI], 0.53-0.82; P <.001). Compared with the nonstatin cohort, elevated risk was evident only in the Least-fit (HR 1.50; 95% CI, 1.30-1.73; P <.001) and Low-fit patients (HR 1.22; 95% CI, 1.06-1.41; P =.006). Conclusions Risk of diabetes in statin-treated dyslipidemic patients was inversely and independently associated with cardiorespiratory fitness. The increased risk was evident only in relatively low-fitness patients. Improving fitness may modulate the potential diabetogenic effects of statins.