Model uses blood pressure swings to better distinguish risk
An observational cohort analysis linked blood pressure fluctuations to heart and blood vessel outcomes; the model better distinguished people with different levels of risk.
The Annals of Applied Statistics
In people in the Multi-Ethnic Study of Atherosclerosis, an observational cohort, researchers analysed blood pressure and heart and blood vessel outcomes using a proposed statistical model. The model combined each person’s average blood pressure with their fluctuations over time. It allowed the amount of fluctuation to differ between people.
The analysis linked these fluctuations to cardiovascular outcomes in this ethnically diverse and generally healthier cohort. Explicitly accounting for differences in variability substantially improved how well the model distinguished risk. The researchers also tested their method in simulations. Conventional models that assumed equal variability across people could produce biased estimates and poorer predictions when variability differed.
Why it matters
For research on age-related cardiovascular disease, the model offers a way to consider fluctuations within a person—not just their average blood pressure—when estimating risk.
Caveats
The cohort analysis was observational, so the associations do not establish cause and effect. Better risk discrimination does not show that using the model improves health outcomes.
The paper
University of California, Los Angeles
The Annals of Applied Statistics · 1 Sep 2026
