AI Analysis of Wearable Blood Pressure Data May Flag Hypertensive Crises Up to Three Days Ahead

A single blood pressure reading is a snapshot of one moment. That is the problem a team of researchers set out to address in a study of 4,190 adults with stage 2 hypertension, all treated at one hospital in Sindh, Pakistan, and followed from January 2025 to January 2026. The question was whether artificial intelligence, fed a steady stream of blood pressure and wearable data, could see a hypertensive crisis coming before it arrived.

The findings are preliminary and independent, and they will be presented at the American Heart Association's Scientific Sessions 2026, held Nov. 6-9 in Chicago. The participants were ages 40 to 85, with an average age of 62.4, and 46% were women. Each wore a portable cuff-based blood pressure monitor that took repeated readings through the day and night, including during daily activities and sleep. They also wore a wrist device with light-based sensors that tracked cardiovascular signals such as heart rate and heart rate variability.

Researchers analyzed 47 characteristics in that data, among them how much blood pressure dropped during sleep, how much it rose after waking, how it changed over time, and how heart rate varied. An AI tool then looked for combinations of patterns linked to a hypertensive crisis within the next 72 hours. A crisis here means readings above 180/120 mm Hg together with acute end-organ damage. It can lead to stroke, heart attack, kidney damage, aortic dissection or pulmonary edema, and it is a leading cause of heart-related emergency department visits. When the model flagged a patient, the physician was alerted, contacted the patient, reviewed symptoms and adjusted medication if needed.

Over the 12 months, 312 participants, about 7% of the total, had 418 hypertensive crises. The tool could separate higher-risk from lower-risk patients up to 72 hours ahead. Among those who triggered an alert, about 32% went on to have a crisis, while 97.4% of those with no alert stayed crisis-free.

Two patterns stood out. Participants whose blood pressure did not fall during sleep, as it normally should, had nearly five times the odds of an upcoming crisis. Those whose pressure climbed more than 35 mm Hg after waking, compared with nighttime levels, had nearly four times the odds.

The intervention numbers drew attention too. Among patients whose physicians received an alert and adjusted medication within 24 hours, hypertensive crises were about 61% lower and emergency department visits about 45% lower than in people who did not get timely, alert-triggered intervention. "Everyone's blood pressure changes throughout the day and night, and a single reading only captures one moment in time," said lead author Waleed Razzaq Chaudhry, M.D., an internal medicine resident at the University of Maryland Capital Region Medical Center in Largo. He said earlier detection could give patients and clinicians more time to act.

Maria Delgado-Lelievre, M.D., a hypertension specialist at the University of Miami Miller School of Medicine, noted that many patients have no warning symptoms before a serious complication. She said such tools should complement, not replace, clinical decision-making.

The limits are plain. Every participant was an adult in Pakistan treated at a single hospital, and the researchers say the approach must be tested in more diverse groups and healthcare settings before wide use. More work is also needed on how clinicians should respond to a high-risk alert: confirming pressure, reviewing symptoms and medication adherence, deciding on adjustment, and judging when emergency evaluation is warranted. A 2025 AHA scientific statement said cuffless devices need more research and standardized testing before guiding care. The Association recommends automatic upper-arm cuff monitors, not wrist or finger devices, and lists validated models at validatebp.org.