Optimizing Home Blood Pressure Monitoring: Findings from the ARIC Study

Introduction: The 2025 AHA/ACC High Blood Pressure (BP) Guidelines recommend home BP monitoring (HBPM) to confirm hypertension diagnosis, guide therapy, and support long-term control. However, the optimal number of HBPM measurements needed for accurate and reliable assessment is not established. Hypothesis: We hypothesized that repeated HBPM, when based on multiple readings per session, would plateau within 3 days. Methods: At ARIC study visit 10 (2021-23) participants completed 8 days of HBPM with two daily sessions (morning and evening) each consisting of 3 readings taken at 1-minute intervals after 5-minutes of rest using an Omron Series 10 BP7450 monitor. We assessed agreement between the 8-day mean (reference) and shorter protocols (fewer days, <3 readings per session, or only morning or evening measurements) using concordance correlation coefficients (CCC) and the proportion of participants with a <10 mm Hg difference from the reference. We then evaluated whether the number of days to reach BP stability (average systolic BP [SBP] and diastolic BP [DBP] <5 mm Hg from reference) differed by age, sex, and anti-hypertensive medication use. Results: Among 812 participants (median age 83 years, 24% Black adults, 40% male), 29% had normal BP during the study visit, and 84% were on anti-hypertensive medications ( Table ). Median SBP and DBP remained within a 2 mm Hg range across each day of the 8-day protocol. Using cumulative averages, the largest improvement in agreement occurred when extending the averaging window from 1 to 2 days, with the CCC increasing for SBP from 0.856 (95% CI: 0.837, 0.873) to 0.929 (0.919, 0.937) and for DBP from 0.846 (0.826, 0.864) to 0.929 (0.919, 0.938) ( Figure 1 ). With 3 days of HBPM, the CCC exceeded 0.95 and the proportion of participants within 10 mm Hg difference achieved ≥97% for both SBP and DBP. Beyond 4 days, agreement metrics plateaued. Averaging ≥2 readings per session, including both morning and evening measurements, yielded better agreement than using a single reading or only one timepoint per day. On average, using 3 readings per session, BP stabilized in 1.28 (1.24, 1.33) days, with results similar within subgroups ( Figure 2 ). Conclusions: In this community-based cohort of very old adults, HBPM provided stable BP estimates within 2-3 days (morning and evening sessions with 2-3 readings each). These findings support shorter HBPM protocols to guide hypertension management while reducing burden in older adults.