WHAT THE STUDY ACTUALLY SAYS

The Apple Watch hypertension alert catches four in ten. The silence means little.

FDA-cleared in September 2025, the feature notifies fewer than half of people with hypertension in Apple's own validation. Hypertension specialists warn the bigger risk is false reassurance.

Share of people notified, by reference blood-pressure categoryStage 1 hypertension: 29.6%; Stage 2 hypertension: 53.7%; All with hypertension: 41.2%0%30%60%Stage 1 hypertension29.6%Stage 2 hypertension53.7%All with hypertension41.2%
Share of people notified, by reference blood-pressure category
GroupValue (%)
Stage 1 hypertension29.6 (24.5 to 35.1)
Stage 2 hypertension53.7 (47.7 to 59.7)
All with hypertension41.2 (37.2 to 45.3)
Share of people notified, by reference blood-pressure category Apple's pivotal validation study, against home cuff readings over 30 days; whiskers are 95% CIs. Source: Apple Inc.

The Apple Watch's hypertension notification, cleared by the FDA in September 2025, does not measure blood pressure and does not tell a wearer theirs is high. It watches the optical pulse sensor over 30-day windows and, if the pattern suggests chronic high blood pressure, sends a single notification [s1]. In Apple's own validation study it caught a minority of people who had hypertension — and that is the number to understand before reading anything into a quiet watch.

What the feature is

The feature runs on the Apple Watch Series 9 and Ultra 2 and later, passively during waking hours, analysing photoplethysmography — the same optical signal behind heart-rate tracking — over non-overlapping 30-day periods [s1]. Apple is explicit that it "does not measure blood pressure directly, and does not surface a blood pressure reading" [s1]. It is aimed at adults 22 and older who have not been diagnosed with hypertension, and not at use during pregnancy [s1]. The FDA cleared it as a Class II "Cardiovascular Machine Learning-Based Notification Software" device under the 510(k) pathway [s2].

Apple says the algorithm was developed using data from more than 100,000 study participants and then tested in a pivotal, decentralised validation study of more than 2,000 people, with performance judged against home blood-pressure cuff readings taken twice daily over 30 days [s1].

What the validation showed

The headline result is a trade-off that Apple chose deliberately, tuning the algorithm for high specificity — few false alarms — at the cost of sensitivity. Overall sensitivity was 41.2% (95% CI, 37.2 to 45.3), and overall specificity was 92.3% (95% CI, 90.6 to 93.7) [s1]. In plain terms, among people who met the study's definition of hypertension, the watch notified about four in ten; among people with normal or merely elevated pressure, it correctly stayed silent about nine times in ten.

Sensitivity rose with severity. For Stage 2 hypertension — the more dangerous category — the feature notified 53.7% (95% CI, 47.7 to 59.7), against 29.6% (95% CI, 24.5 to 35.1) for Stage 1 [s1]. Specificity among people with genuinely normal pressure was 95.3% [s1]. Apple reports consistent performance across age, sex, race and skin tone, with somewhat higher sensitivity and lower specificity in older participants and those with higher body-mass index [s1].

The silence is the problem

Because it was built to avoid false positives, the feature's weakness is false negatives. Writing in the American Heart Association journal Hypertension, a group of hypertension specialists made the central point directly: the greater danger is false reassurance, where someone with undiagnosed high blood pressure takes the absence of an alert as a clean bill of health and skips conventional screening [s3]. Roughly 8% of people without hypertension still received an alert in the validation data, and the authors stress that notifications cannot replace traditional blood-pressure measurement [s3].

This is the opposite of how wearables are usually marketed. The notification is most useful when it fires — it is a reasonable prompt to start monitoring at home or see a clinician — and close to useless as reassurance when it does not. A cuff remains the only way to know a number.

Where it sits among the alternatives

The hypertension notification is a different kind of tool from the cuffless blood-pressure devices that claim to read a number off the wrist — a category cardiologists have declined to endorse for clinical decisions. Apple sidestepped that fight by not reporting a reading at all, which is also why the FDA could clear a screening notification rather than a measurement device. It belongs with the watch's other machine-learning notifications that flag a risk pattern without claiming to quantify it.

For anyone tracking their blood pressure, the definition of high blood pressure itself is what the alert is gesturing at — and the thresholds, confirmation readings and repeat measurement that a diagnosis requires are exactly what a watch notification cannot supply. The feature can open the conversation. It cannot close it, and it was never cleared to.

Sources

  1. Hypertension Notification Feature on Apple WatchApple Inc. , September 1, 2025
  2. 510(k) Substantial Equivalence Determination for Hypertension Notification Feature (K250507)U.S. Food and Drug Administration , September 11, 2025
  3. Apple Watch for Hypertension ScreeningHypertension (American Heart Association) , January 21, 2026

More on

Related coverage