Sleep Signal Flags Heart Failure Risk Up To 10 Years Early

Buzzline

Researchers at Emory University developed an AI model that reads a heart-rhythm signal recorded during a sleep study to predict disease risk up to 10 years ahead. Trained on nearly 16,000 patients and tested on about 23,000 more, higher scores were linked to 69 to 133 percent higher heart failure risk and 103 to 172 percent higher atrial fibrillation risk. The signal was strongest during the lightest sleep stage. The experimental tool needs more validation before doctors can use it, though it could turn routine sleep tests into early heart screening.  Disclosure: This video contains stock footage and content created or enhanced using AI-assisted tools.

Transcript

00:00Researchers at Emory University have built an AI model that reads one heart signal recorded

00:05during sleep. It predicts the 10-year risk of heart failure, atrial fibrillation, stroke,

00:10and death. The model was trained on nearly 16,000 patients from a Massachusetts General

00:16Hospital sleep study. It was then tested on nearly 10,000 patients from Emory and 13,000

00:22from a Boston hospital, per standard deviation increase in the risk score. Heart failure

00:27risk was 69 to 133% higher. Atrial fibrillation risk was 103 to 172% higher. Results held

00:35after adjusting for age, sex, diabetes, hypertension, and smoking. The signal was most predictive

00:42during the lightest stage of sleep. The authors described the gain over existing models as modest

00:47but meaningful. The tool needs more development and validation before clinical use. The study

00:53appeared in the journal Sleep Disclosure. This video contains stock footage and content created

00:58or enhanced using AI-assisted tools.

Category

Transcript

Comments

Recommended