To investigate the association between visit-to-visit variability in blood pressure and the incidence of dementia and its subtypes in a general population, we conducted a population-based retrospective cohort study using the Korean National Health Insurance System database. We identified 7 844 814 subjects without a history of any dementia who underwent ≥3 health examinations from 2005 to 2012 in the Korean National Health Insurance System cohort. Blood pressure variability (BPV) was measured using the variability independent of the mean, coefficient of variation, and SD. During the median follow-up of 6.2 years, there were 200 574 cases of all-cause dementia (2.8%), 165 112 cases of Alzheimer's disease (2.1%), and 27 443 cases of vascular dementia (0.3%). There was a linear association between higher BPV and outcome measures. In the multivariable adjusted model, the hazard ratios and 95% CIs of all-cause dementia were 1.06 (1.04-1.07) for the highest quartile of variability independent of the mean of diastolic blood pressure only, 1.09 (1.08-1.11) for that of systolic blood pressure only, and 1.18 (1.16-1.19) for that of both systolic and diastolic blood pressure compared with subjects having no highest quartile for BPV. Consistent results were noted for Alzheimer's disease and vascular dementia using other indices of variability and in various sensitivity and subgroup analyses. BPV is an independent predictor for developing dementia and its subtypes. A dose-response relationship was noted between higher BPV and dementia incidence. Reducing BPV may be a target for preventing dementia in the general population.
Keywords: Alzheimer’s disease; Blood pressure; Dementia; Variability; Vascular dementia.