Stroke risk of blood pressure indices determined by home blood pressure measurement: the Ohasama study.

BACKGROUND AND PURPOSE The purpose of this prospective cohort study was to investigate associations between stroke and blood pressure (BP) indices (systolic BP [SBP], diastolic BP [DBP], mean BP, and pulse pressure [PP]) determined by home BP measurement. METHODS Associations between stroke and BP indices were examined in a rural Japanese population. Home BP data of 2369 subjects (40% men) > or =35 years of age (mean, 59 years) without a history of stroke were obtained. Associations between stroke and each index were determined using Cox proportional hazards regression and the likelihood ratio (LR) test. RESULTS During follow-up (mean, 11.7 years), 238 strokes occurred. The LR test showed that SBP and mean BP were significantly more strongly associated with total and ischemic stroke than DBP and PP (LR chi2 > or =9.3, P<0.01 for SBP/mean BP, LR chi2 < or =3.8, P> or =0.05 for DBP/PP). SBP tended to be more strongly associated with total/ischemic stroke than mean BP (LR chi2=3.8, P=0.05 for SBP, LR chi2 < or =0.2, P>0.6 for mean BP). PP tended to be slightly more strongly associated with ischemic stroke than DBP (LR chi2=7.5, P<0.01 for DBP, LR chi(2)=9.3, P<0.01 for PP), whereas DBP was significantly more strongly associated with hemorrhagic stroke than PP (LR chi2=9.2, P<0.01 for DBP, LR chi2=2.5, P=0.01 for PP). CONCLUSIONS PP obtained from home BP measurements was weakly associated with stroke, whereas SBP showed the strongest association. Additionally, DBP and PP may be associated with different stroke types.