This cut has not been rendered. Farm manifest: sleep-study-heart.json.
A Sleep-Study Heart Tracing and Future Risk
The paper came out Friday in Sleep. Here is what the authors and NIH actually reported.
Primary sources
National Institutes of Health · agency · published 2026-10-09
Sleep (Oxford University Press) · journal · published 2026-10-09
No video file yet. Placeholder poster only. Manifest: sleep-study-heart.json.
Script
A paper out Friday in the journal Sleep says a heart tracing from a routine sleep study can help sort long-term heart risk. The National Institutes of Health, which backed the work, says the team used one electrocardiogram lead plus the night's sleep stages. They tuned the model on 15,809 patients at Massachusetts General Hospital, then checked it on 9,810 patients at Emory and 12,576 at Beth Israel Deaconess. In the Emory group, a one-standard-deviation rise in the score lined up with about twice the hazard of atrial fibrillation, a hazard ratio of 2.03. The score still added information after age, sex, body mass, diabetes, blood pressure, and sleep apnea were accounted for. It also tracked heart failure and death from any cause. Heart attack and stroke, the Institutes say, still need more work. One lead, not twelve, means the check can ride along with a test people already take.
Claims
Koscova and colleagues published in Sleep on October 9, 2026 (DOI 10.1093/sleep/zsag229). NIH posted its release the same day.
The model uses a single-lead ECG from overnight polysomnography plus expert-annotated sleep stages. It was fine-tuned on 15,809 Massachusetts General Hospital patients and tested on 9,810 Emory patients and 12,576 Beth Israel Deaconess patients.
In the Emory cohort, a one-standard-deviation increase in the model score had an adjusted hazard ratio of 2.03 (95% CI 1.82–2.27) for atrial fibrillation.
The score kept predictive value after adjustment for age, sex, BMI, diabetes, hypertension, and sleep measures including apnea severity.
NIH says the model could predict atrial fibrillation, heart failure, and death from any cause, and that additional work is needed for myocardial infarction and stroke.