Abstract
BACKGROUND: The real-world applicability of long-term blood pressure (BP) variability measurements remains underexplored. We evaluated the association between visit-to-visit BP variability and the risk of cardiovascular events and all-cause mortality using electronic health records. METHODS: In this retrospective cohort study at a large academic medical center in Taiwan, we calculated the variability independent of the mean (VIM) and average real variability of BP using electronic health records of 16 945 adults with at least one outpatient BP measurement in any 3 consecutive years from 2012 to 2017. We used Cox proportional hazards models to assess associations between BP variability and cardiovascular events, including cardiovascular deaths, and all-cause mortality through 2020. RESULTS: Over a median follow-up of 4 years, 317 patients experienced cardiovascular events, and 582 died. Adjusted hazard ratios (HRs) for cardiovascular events increased gradually across both VIM and average real variability quartiles of BP. The adjusted HRs (95% CIs) per interquartile range increase in systolic BP variability was 1.24 (1.09-1.41) for VIM and 1.11 (1.01-1.23) for average real variability. For diastolic BP, the HRs (95% CIs) were 1.22 (1.09-1.36) and 1.13 (1.02-1.24), respectively. Similar results were observed for all-cause mortality except a weaker association with average real variability of diastolic BP (HR, 1.08 [95% CI, 0.99-1.17]). The association between VIM of BP and risk of cardiovascular events was consistent across patient subgroups. CONCLUSIONS: In the electronic health records analysis, visit-to-visit BP variability was independently associated with the risk of cardiovascular events and all-cause mortality. Our findings indicate the applicability of BP variability indices in real-world health care settings.