Cumulative blood pressure improves risk prediction

Ten-year cumulative systolic blood pressure modestly improved cardiovascular risk prediction.
*Retrospective cohort prediction study; Level 2b (OCEBM).

Citation

Avrahami D, Zahger D, Arbel R, et al. Effect of cumulative blood pressure exposure on long-term cardiovascular outcomes in the community, a nationwide cohort study. The American Journal of Medicine. 2026;139:1330-1336. doi:10.1016/j.amjmed.2026.05.015

Background

Standard cardiovascular risk calculators usually use one recent blood pressure reading. Because blood pressure varies and acts over years, long-term exposure may better identify risk.

Patients

614,084 Israeli adults aged 45 years or older, without known cardiovascular disease, and with at least 3 blood pressure measurements over the prior 10 years. Patients with prior cardiovascular disease were excluded.

Intervention

Adding 10-year cumulative systolic blood pressure above 140 mmHg to a standard pooled cohort cardiovascular risk score.

Control

The same risk score using the most recent systolic blood pressure measurement only.

Outcome

Hard cardiovascular disease, stroke, all-cause death, and prediction performance using the C statistic (0.5 to 1.0; higher is better).

Follow-up Period

5 years after 10 years of prior blood pressure exposure.

Results

Outcome Odds ratio per cumulative exposure unit Risk reclassification improvement
Hard cardiovascular disease OR 1.027 (95% CI 1.024 to 1.029) 0.492 (95% CI 0.458 to 0.520)
Stroke OR 1.031 (95% CI 1.028 to 1.033) 0.656 (95% CI 0.620 to 0.701)
All-cause death OR 1.014 (95% CI 1.011 to 1.016) 0.539 (95% CI 0.520 to 0.559)

OR: odds ratio. One cumulative exposure unit means an average of 1 mmHg above 140 mmHg for 5 years.

Stroke prediction also improved by C statistic: 0.716 to 0.730, difference 1.3% (95% CI 1.0% to 1.7%). The comparator was the current standard risk-prediction approach.

Limitations

Observational electronic-record study; findings may not generalize outside Israel. Patients needed repeated blood pressure checks, creating selection bias. Prediction gains were statistically clear but modest for C statistic, and no treatment strategy was tested.

Funding

Israeli public and institutional grants; no declared conflicts.

Clinical Application

Use cumulative blood pressure history, when available, to refine risk discussions; do not change treatment solely from this prediction study.