A participant sleeps during the 2026 Power Nap contest at Han River park in Seoul, South Korea, May 2, 2026. REUTERS/Kim Soo-hyeon
Published : 21 Oct 2026, 10:19 PM
Updated : 21 Oct 2026, 10:19 PM
A standard overnight sleep test could double as an early warning system for heart disease, according to new research reported by The Independent.
Scientists have shown that artificial intelligence can extract cardiovascular risk signals from ECG recordings taken during routine sleep studies -- the same tests used to diagnose insomnia and sleep apnoea -- and use them to forecast a patient’s likelihood of heart attack, stroke, irregular heart rhythm, heart failure or death up to a decade later.
Researchers trained the AI model on data from more than 15,000 patients at Massachusetts General Hospital in Boston before validating it against two further datasets totalling over 22,000 patients from Emory University Hospital in Atlanta and Beth Israel Deaconess Medical Centre in Boston.
The model’s predictive power held up even after accounting for established risk factors such as age, sex, BMI, diabetes and hypertension, The Independent reports.
“The score from our deep learning model improved prediction of cardiac outcomes beyond established risk factors,” Gari Clifford, a study author from Emory University, said. “Adding these scores into sleep studies has great potential for improving early detection of cardiovascular disease.”
The findings, published in the journal Sleep, suggest the approach could be delivered through simple, low-cost wearable patches -- with a single night’s data sufficient for a meaningful risk assessment.