Blood pressure variability supersedes heart rate variability as a real-world measure of dementia risk

Figures

• 

  Figure 1. Association of blood pressure and heart rate variability with incident Alzheimer’s Disease and Related Dementias in the EKG cohort. Incident dementia risk associated with systolic BPV, diastolic BPV, and HRV (A) in the pooled cohort, and in stratified analyses by (B) sex, (C) age, (D) number of comorbidities (0 vs. ≥ 1), and (E) race/ethnicity (Non-Hispanic White vs. Others). Cox models adjusted for age, sex, race/ethnicity, number of visits, use of antihypertensive medications, smoking status, diabetes mellitus, chronic kidney disease, atrial fibrillation/atrial flutter, coronary artery disease, mean systolic and diastolic blood pressure, presence of any metastatic malignancy, myocardial infarction, heart failure, stroke, number of EKGs, and heart rate. Patients censored at last follow up visit or death prior to the end of the study period, whichever occurred later. Analyses exclude patients of unknown race due to model convergence. Abbreviations: BPV blood pressure variability, CI confidence interval, HRV heart rate variability.

• 

  Figure 2. Association of multiple blood pressure variability metrics with incident Alzheimer’s Disease and Related Dementias. Incident dementia risk associated with systolic measures of BPV (A) in the pooled cohort, (B) stratified by sex, and (C) stratified by age, as well as diastolic measures of BPV with the same stratification (D–F). Cox models adjusted for age, sex, race/ethnicity, number of visits, use of antihypertensive medications, smoking status, diabetes mellitus, chronic kidney disease, atrial fibrillation/atrial flutter, coronary artery disease, mean systolic and diastolic blood pressure, presence of any metastatic malignancy, myocardial infarction, heart failure, stroke, number of EKGs, and heart rate. Patients censored at last follow up visit or death prior to the end of the study period, whichever occurred later. Analyses exclude patients of unknown race due to model convergence. Abbreviations: ARV absolute real variability, CI confidence interval, MMD maximum–minimum difference, SD standard deviation.