Long-term predictors of stroke in healthy middle-aged men.

Background There are few data on risk factors for stroke during long-term follow-up of healthy individuals. Aims We aimed to investigate the long-term predictive impact on stroke risk of baseline variables including hemodynamic variables measured at rest and during exercise in middle-aged, healthy men. Methods We performed a prospective cohort study of 2014 healthy Norwegian men aged 40-59 years, recruited during the period 1972-1975 and followed until 2007. Participants underwent a comprehensive clinical assessment at baseline, including a bicycle exercise test. Data on stroke, transient ischemic attack, and death were collected on all participants from follow-up visits, medical records, and the National Cause of Death Registry. We used Cox regression for analysis and estimated hazard ratios with 95% confidence intervals, adjusting for traditional risk factors and hemodynamic variables measured at rest and during exercise. Results During 35 years' follow-up, 316 participants (16%) had stroke, of which 287 (91%) were ischemic and 29 (9%) were hemorrhagic. Age (hazard ratio 2.70 per increase in one standard deviation, 95% confidence interval 2.13-3.43), resting systolic blood pressure (hazard ratio 1.24, 95% confidence interval 1.11-1.39), body mass index (hazard ratio 1.14, 95% confidence interval 1.02-1.29), and atrioventricular conduction time (hazard ratio 1.11, 95% confidence interval 1.03-1.19) were significantly associated with long-term risk of stroke, as were maximal systolic blood pressure and heart rate during exercise (hazard ratio 1.28, 95% confidence interval 1.13-1.46, and hazard ratio 0.86, 95% confidence interval 0.74-0.99, respectively). Conclusions Hemodynamic variables at rest and during exercise testing add to the predictive value of clinical variables in healthy, middle-aged men, and should be included in the assessment of long-term risk of stroke, when available.