Kivorent News

Wednesday, November 18, 2026

Health techNews

Google study: smartwatch sensors caught two in three cases of high blood pressure

Google Research reported in Nature that an AI model reading a smartwatch's pulse and motion sensors found 65.8% of hypertension cases in a 196-person study, ahead of a single office cuff check.

Pulse, covering wearables and health sensors
, 3 min read

65.8%of hypertension cases the watch model flagged

What to know

  • Google Research published the study in Nature on November 5, 2026, testing 196 adults at five US sites against 24-hour ambulatory blood pressure monitoring.
  • The watch model had 65.8% sensitivity and 90.0% specificity; a standard office reading caught 55.3% of cases.
  • Google calls it early feasibility research, not a product or a diagnosis, though it says the work informed the Health Guardian blood pressure trends feature.
Share of hypertension cases detected (sensitivity), initial screening
Smartwatch model65.8%
Office cuff reading55.3%

Source: Google Research blog, Nov 5, 2026. Reference: 24-hour ambulatory blood pressure, 196 participants.

Google Research says an AI model that reads only a smartwatch's built-in sensors picked out about two in three people with high blood pressure (hypertension) in a clinical study, a better hit rate than a single reading taken with an arm cuff in a doctor's office. The results were published in Nature on November 5, 2026, and described the same day on the Google Research blog. AndroidPure also reported the figures below.

How the study worked

The system, which Google calls AI-PPG-ACC-HTN, does not measure blood pressure. It looks at two signals the watch already records: the optical pulse sensor (photoplethysmography, or PPG) and the accelerometer, which picks up the tiny body movements caused by each heartbeat. The person's height is also an input. The model averages what it sees over several days of wear and gives a probability that the wearer has hypertension.

Google first trained the model on 492,347 hours of unlabeled sensor data from 2,109 consenting participants. It then tested it in a prospective study at five US sites, with Beth Israel Deaconess Medical Center and Harvard Medical School, on 196 adults who had no known heart disease and were not taking blood pressure medication. Each person's true status was set by 24-hour ambulatory blood pressure monitoring, a cuff worn around the clock that doctors treat as the reference test. Participants wore the watch for seven days, according to AndroidPure.

Google's blog does not name the watch model used. AndroidPure says the paper describes it only as a widely available consumer smartwatch.

The results

In the first screening test, Google reported:

  • Smartwatch model: sensitivity 65.8% (the share of hypertension cases it caught), specificity 90.0% (the share of people without hypertension it correctly cleared), and positive predictive value 80.6%.
  • Office cuff reading: sensitivity 55.3% and positive predictive value 77.8%. AndroidPure reports its specificity as 90.0%.

Google also ran a confirmation test on 48 participants whose office reading was already high, the step where doctors decide whether a raised number is real. There the watch reached 78.4% sensitivity and 90.9% specificity. A return office visit scored 67.6% and 63.6%, and several days of home cuff readings scored 89.2% and 81.8%, the best sensitivity of the three.

Limits Google states

Google calls the work early feasibility research. Its blog says the methods are exploratory and do not indicate any current or upcoming product feature. It says screening by smartwatch is "not intended to replace standard clinical arm cuffs." The study left out people already on blood pressure drugs or with known heart disease, so it says nothing about tracking treated patients.

The blog does say the findings informed the blood pressure trends part of Health Guardian, the monthly report in the Google Health app for recent Pixel Watch and Fitbit devices. AndroidPure notes that Google's help page describes that feature as for general wellness only. None of the sources says Google has sought FDA clearance for a hypertension alert.

That sets it apart from Apple, whose hypertension notifications were cleared by the FDA in 2025, and from cuffless devices such as the Aktiia Hilo band. Our explainer on what FDA clearance means covers the difference between a cleared feature and a wellness one.

Why it matters

Blood pressure is usually checked with a cuff during an occasional office visit, and in this study that single reading missed nearly half of cases. A peer-reviewed study showing that an ordinary watch can flag hidden cases more often than a one-off cuff check makes passive screening on the wrist more credible, but it is one 196-person study run largely by Google, and the company itself says a cuff is still needed to confirm.

This article is general information about how technology works. It is not medical advice. Talk to a clinician about your own health.

Sources

Heart health techWearablesGoogleDigital health and AI in care

Written by an AI correspondent under standards set by editor Jacob Knight. Spotted an error? Tell us and we will fix it.