Background: Hypertension and smoking independently contribute to the risk of stroke. Our objective was to investigate home blood pressure (HBP) levels, day-by-day BP variability, and smoking in the prediction of stroke in Japanese men.
Methods: In this study, 902 men (mean age, 58.6 years) without a past history of stroke were evaluated. HBP was measured once every morning for 4 weeks. Day-by-day BP variability was defined as within-subject standard deviations (SD) of HBP. Smoking history was obtained from a standardized questionnaire. Hazard ratios (HRs) for stroke were examined by Cox regression model, with adjustment for possible confounders.
Results: During 13.1 years (median) of follow-up, 89 cerebral infarctions, 28 intracranial hemorrhages, and six other strokes occurred. Systolic HBP levels (HR = 1.59 per 14.6 mm Hg increase, P < 0.0001) and variability (HR = 1.26 per 3.1 mm Hg increase, P = 0.03) of +1 between-subject SD were significantly associated with cerebral infarction. The relationship between HBP and cerebral infarction differed with smoking status (interaction P = 0.021 and 0.017 for systolic level and variability, respectively). In analyses stratified according to smoking, systolic level (HR = 1.78, P < 0.0001) and variability (HR = 1.38, P = 0.006) were significantly associated with cerebral infarction in ever smokers (N = 511), but not in never smokers (N = 391; P ≥ 0.6 for both). No significant association was found between smoking and the risk of intracranial hemorrhage.
Conclusions: In ever smokers, both HBP levels and variability are significantly associated with the risk of cerebral infarction. Our findings further validate the benefit of smoking cessation in preventing cardiovascular disease (CVD), especially cerebral infarction.