Visit-to-visit blood pressure variability and dementia risk after considering antihypertensive treatment: real-world data from the Japanese National Health Insurance

This retrospective cohort study evaluated the association between visit-to-visit blood pressure (BP) variability and dementia risk, considering antihypertensive drug classes and medication adherence using Japanese National Health Insurance data (2015-2023) of 301,448 participants (age: 66.6 years, male: 38.6%). Visit-to-visit systolic BP (SBP) variability was assessed using the coefficient of variation (CV) from five annual health check-ups. The outcome was antidementia drug initiation as a proxy for dementia diagnosis and was analyzed using Fine-Gray models, with death as the competing outcome. For participants untreated/treated with antihypertensive medication, 366/298 initiated antidementia drugs during 2.20 ± 1.19/2.11 ± 1.19 years of follow-up, respectively. The highest SBP-CV sextile was associated with dementia risk regardless of treatment status: hazard ratio of the 6^th sextile (SBP-CV ≥ 9.83%) vs. 1^st-5^th sextiles was 1.50 (95% confidence interval [CI]: 1.17-1.92) for untreated participants, while for treated participants, the hazard ratio of the 6^th sextile (SBP-CV ≥ 10.67%) was 1.43 (95%CI: 1.09-1.89) after further adjusting for antihypertensive drug classes and medication adherence assessed by medication possession rate. In the stratification analyses by baseline characteristics, only HbA1c in treated participants demonstrated a significant interaction with SBP-CV for dementia risk; this association was pronounced among treated participants with HbA1c ≥ 6.5% (interaction P = 0.024). No significant interactions were observed among antihypertensive drugs, poor adherence, and SBP-CV in relation to dementia risk. High visit-to-visit BP variability, indicated by a CV of approximately ≥10%, is associated with dementia, regardless of treatment status. In the treated participants, this association persisted even after accounting for antihypertensive drug classes and medication adherence.