Risk of fatal stroke according to blood pressure level: an 18-year follow-up of the Oslo Study.

OBJECTIVE: To determine how blood pressure level predicts the incidence of fatal stroke. DESIGN: The Oslo Study is a prospective cohort study of preventive and epidemiological aspects of cardiovascular disorders in middle-aged men. Of 25,915 men invited, 16,209 aged 40-49 years attended the screening. A 7% random sample of men aged 20-39 years were also invited to attend. METHODS: The screening started in May 1972 and the analysis presented is an 18-year follow-up for fatal strokes. Men with a history of stroke were excluded from the analyses. RESULTS: Of 16,173 men with no history of stroke 85 died from stroke. Results from Cox proportional hazards regression analysis confirm diastolic (DBP) and systolic blood pressure (SBP) as strong independent risk factors of fatal stroke, with DBP being the stronger predictor. Analyses of risk of fatal stroke by quintile values show SBP to give significantly increased risk from the third quintile (136 mmHg), and DBP from the fifth quintile (95 mmHg) relative to the first quintile. No levelling off at highest levels can be seen when analysing decile values. No J-shape of the curve was evident. Men on drug treatment for hypertension with no stroke history (n = 440) had 4.7-fold (crude) and 2.8-fold (adjusted for age and DBP) the rate of stroke mortality of men not on drug treatment for hypertension. CONCLUSION: DBP was a stronger predictor than SBP, with increasing risk from the fifth quintile of DBP and the third quintile of SBP. Men on drug treatment for hypertension at screening were at increased risk during the follow-up period, indicating that their treatment did not sufficiently reduce their risk of stroke.