Effects of long-term blood pressure variability on cerebral microbleeds

Objective To assess the reproducibility of long-term blood pressure variability in ischemic stroke and the association between the variability and cerebral microbleeds. Methods Patients with ischemic stroke at the previous 1—6 months were consecutively recruited and followed up 12—18 months. Blood pressure measurements were taken at every interview. Blood pressure variability indicated visit-to-visit variability and was quantified by calculating the maximum (Max), standard deviation (SD) and coefficient of variation (CV). For these variability might positively correlated with mean of blood pressure, the additional variability measure, SD independent of mean (SDIM), was also calculated. To determine the reproducibility of mean and variability measurements, the intraclass correlation (ICC) was also calculated. MRI was performed at baseline and the end of the study. Cerebral microbleeds were rated using Microbleed Anatomical Rating Scale (MARS). Multiple Logistic regression was used to assess the association between the visit-to-visit blood pressure variability and cerebral microbleeds. Results A total of 720 patients were recruited, of whom 595 (82.6%) subjects were present for 14 blood pressure readings during follow-up. The visit-to-visit blood variability measurements were moderately reproducible according to the ICC: 0.46—0.72 for systolic blood pressure (SBP) and 0.42—0.69 for diastolic blood pressure (DBP), respectively, P<0.01 for all measures. Patients with cerebral microbleeds were more likely to have higher mean blood pressure and variability for both SBP and DBP regardless the distribution of cerebral microbleeds. Being SBP Max,SBP SD,SBP CV,SBP SDIM（OR=1.036, 95% CI 1.021—1.052, P=0.000; OR=1.060, 95% CI 1.001—1.122, P=0.046; OR=1.084, 95% CI 1.000—1.175, P=0.049；OR=1.065, 95% CI 1.002—1.132, P=0.044）and DBP SD,DBP CV (OR=1.111, 95% CI 1.000—1.233, P=0.049; OR=1.091, 95% CI 1.001—1.190, P=0.047) were the independently risk factors of cerebral microbleeds at deep region; SBP Max（OR=1.049, 95% CI 1.029—1.068, P=0.000）and DBP SD、DBP CV（OR=1.236, 95% CI 1.107—1.379, P=0.000；OR=1.188, 95% CI 1.087—1.298, P=0.000) independently associated with cerebral microbleeds at infratentorial location. There was no significant relation between the long-term variability of blood pressure and cerebral microbleeds at lobar region. Conclusions This study indicates that long-term blood pressure variability is substantial and independently associated with cerebral microbleeds in deep or infratentorial but not with that in lobar region. The different relations between the variability and cerebral microbleeds might indicate the heterogenic mechanisms of cerebral microbleeds. Key words: Blood pressure; Cerebral hemorrhage; Brain infarction