AI predicts hypertensive crises up to three days early in preliminary study

The American Heart Association detailed preliminary research in an announcement showing that AI analysis of wearable health data may identify the risk of a hypertensive crisis up to three days in advance. The research will be presented at Scientific Sessions 2026 in Chicago from 6 to 9 November.

The study followed 4,190 adults with stage 2 hypertension at one hospital in Sindh, Pakistan, for 12 months. Participants used a portable cuff based blood pressure monitor and a wrist worn fitness tracker, while an AI model analyzed 47 measures including blood pressure during sleep, changes after waking, heart rate and heart rate variability.

During the study, 312 participants experienced 418 hypertensive crises. About 32% of participants who received a positive alert later experienced a crisis, while 97.4% of those without an alert remained free of a crisis. Blood pressure that did not decrease during sleep was associated with nearly five times the odds of an upcoming crisis, while an increase of more than 35 mm Hg after waking was associated with nearly four times the odds.

Among patients whose physicians received an alert and adjusted medication within 24 hours, hypertensive crises were about 61% lower and emergency department visits were about 45% lower than among patients without a timely intervention. The findings come from a research abstract and have not been reviewed by peers or published as a full manuscript.