Stroke Incidence According to Cardiorespiratory Fitness: A Cohort Study of 483,379 Hypertensive Patients

ObjectivesWe assessed stroke incidence in hypertensive patients according to cardiorespiratory fitness (CRF) and changes in CRF. MethodsA prospective cohort study of 483,379 US Veterans. Participants completed a maximal standardized Exercise Treadmill test (ETT) performed within the Veterans Affairs medical centers across the United States between 1999 and 2020. None exhibited evidence of unstable cardiovascular disease during the ETT. Participants were stratified into 5 age-and-gender specific CRF categories based on the peak metabolic equivalents (METs) achieved. A subgroup of participants with two ETT evaluations (n=110, 576) were also assigned to 4 categories based on MET changes from the initial ETT to the final ETT. Multivariable Cox models, adjusted for age, and co-morbidities were used to estimate HRs and 95% CIs for stroke risk. ResultsThe mean age {+/-} standard deviation (SD) was 59.4{+/-}9.0 years. During the median follow-up time of 10.6 years (5,182,179 person-years), there were 15,925 stroke events with an average annual rate of 3.1 events per 1,000 person-years. In a final adjusted model, relatively poor CRF was the strongest predictor of stroke risk than any other comorbidity (HR: 2.24; 95% CI: 2.10-2.40; P< 0.001). For each 1-MET higher exercise capacity, the risk was 10% lower (0.90, 95% CI 0.90-0.91, p<0.001). Compared to the Least-fit, stroke risk was 23% lower for Low-fit individuals (HR 0.77; 95% CI, 0.73-0.80; p<0.001); and declined progressively to 55% for those in the highest CRF category (HR 0.45; 95% CI 0.42-0.48; p<0.001). We also assessed stroke incidence according to change in CRF. Compared to fit individuals during both evaluations, the risk was 27% higher for those who became unfit (HR 1.27, 95% CI 1.15-1.41, p<0.001), and not significantly different for unfit who became fit (HR 1.10, 95% CI 0.97-1.25, p=0.13). ConclusionsPoor CRF was the strongest predictor of stroke incidence in hypertensive patients, regardless of age race, or gender. The association was independent, inverse, and graded for all stroke types. Changes in CRF over time reflected inverse changes in stroke risk, suggesting that risk of stroke can be modulated by improved CRF.