Sleep Study Heart Data Could Flag Future Cardiac Risk
Adults who undergo sleep studies may one day have their heart recordings screened for future cardiovascular outcomes.
Updated on Nov. 2, 2026 in Heart Disease
Live Poll
Would you want heart-risk screening added to your next overnight sleep study?
NIH-supported researchers developed a deep learning model that predicts cardiovascular outcomes from single-lead ECG recordings taken during sleep studies. The findings were published in the journal Sleep, and the approach could eventually add a cardiac screening step to tests people already undergo for sleep disorders.
Why it matters
Sleep studies are already widely performed, so extracting cardiac risk information from recordings that are already being collected could expand early detection without a separate test. The research asks whether this added signal improves on the risk factors clinicians already use, which is the key question before any change in practice.
The study was a deep learning analysis trained on about 15,809 patients and validated on roughly 22,386 patients at two other hospitals, using single-lead ECGs from sleep studies with outcomes drawn from electronic health records. The model sorted people into groups with different long-term cardiovascular risk and retained predictive value after adjusting for age, sex, BMI, diabetes, hypertension, and sleep apnea severity.
The players
National Institutes of Health (NIH)
The U.S. federal agency that funds and conducts biomedical and health research, including heart and sleep research.
Massachusetts General Hospital
A Boston hospital affiliated with Harvard Medical School whose clinical data was used to train the model.
Emory University Hospital
An Atlanta hospital whose patient dataset was used to validate the model's performance.
Beth Israel Deaconess Medical Center
A Boston teaching hospital whose patient dataset was used to validate the model's performance.
The details
The system combines a single-lead electrocardiogram recorded during sleep with expert-annotated sleep stage information, allowing the model to analyze heart rhythm patterns across the night. Because it uses one lead rather than the standard 12-lead setup, it can work with the recording equipment already used in sleep studies. The model predicted atrial fibrillation, heart failure, and death from any cause, and adding its score improved prediction beyond established risk factors.
Timeline
NIH announced the study findings on November 2, 2026.
Health Landscape
Sleep studies are typically used to diagnose disorders such as sleep apnea, yet they generate continuous heart rhythm data that has rarely been used for cardiac risk assessment. This work points toward a population-level shift in which routine sleep diagnostics might also serve as a cardiovascular screening tool.
If you have a sleep study scheduled, it is reasonable to ask your sleep physician whether cardiac findings from the recording are reviewed as part of your results. This research does not change current care, so people with heart concerns should continue to discuss their risk with their doctor.
The takeaway
Electrocardiograms recorded during sleep studies may one day help flag future heart risk without an added test. Ask your doctor about your cardiovascular risk factors, and track blood pressure and heart symptoms as you always would.
Further reading
Explore more on Heart Disease and how cardiac risk is being assessed.
Live Poll
Would you want heart-risk screening added to your next overnight sleep study?
