Apple Watch blood pressure alerts: who gets missed?

A modelling study in JAMA estimated that the Apple Watch hypertension notification would miss 59% of US adults who have undiagnosed high blood pressure. It would also alert 7.7% of adults who do not have it. The estimate combines Apple’s reported test performance with national survey data, so it describes a population, not any single wearer.

How well does the Apple Watch hypertension alert work?

In a JAMA research letter published in February 2026, Cohen and colleagues applied the feature’s reported sensitivity (41%) and specificity (92%) to NHANES 2017 to 2020, a nationally representative US health survey. They included adults aged 22 and over who had not been diagnosed with hypertension.

Those two inputs come from Apple’s own validation study, which the authors list as a limitation. This is cross-sectional modelling, not a trial of real wearers.

The result: an estimated 41.2% of people with undiagnosed hypertension would receive an alert, and 7.7% of people without hypertension would too. An alert was informative. A missing alert was much less so.

Why does an alert mean different things at different ages?

Because the same test reads differently depending on how common the condition is. Among adults not yet diagnosed, the model put the probability of hypertension at 14% under age 30 and 45% at 60 and over.

An alert raised those figures to 47% and 81%. No alert lowered them to 10% and 34%. The pattern also varied by race and ethnicity: an alert moved the estimate from 36% to 75% in non-Hispanic Black adults and from 24% to 63% in Hispanic adults, reflecting different starting prevalence in each group.

The authors flag the risk of false reassurance: in none of the groups reported did a quiet watch bring the estimate near zero. They write that alerts should supplement, not replace, validated cuff-based screening.

The same question, what a wearable signal does and does not tell you, runs through the smartwatch atrial fibrillation screening trial. The letter’s closing line calls for rigorous validation and thoughtful integration of cuffless technologies into screening.

The bottom line

• A JAMA modelling study estimated the Apple Watch alert would flag 41.2% of undiagnosed hypertension and falsely alert 7.7% of people without it.
• An alert raised the estimated probability of hypertension from 14% to 47% under age 30 and from 45% to 81% at 60 and over.
• The inputs came from Apple’s own validation study, and the analysis is a model, not a trial of wearers.