How your sleep patterns may signal heart health risks years before symptoms appear.
What the Research Found
Scientists have found that early-stage sleep data—specifically heart electrical activity—can predict future cardiovascular risks up to 10 years in advance.
The study, published in the journal SLEEP, used electrocardiogram (ECG) data collected during sleep. A single ECG lead on the chest during sleep was enough to train an AI model.
Researchers tested the model on three datasets: nearly 16,000 patients from a sleep study at Massachusetts General Hospital, and additional groups from Emory University Hospital and Beth Israel Deaconess Medical Center.
Key Predictive Outcomes
- Only one group showed a significantly increased risk of heart attack.
Adding ECG data from sleep improved current cardiovascular risk models by up to 6 percent for atrial fibrillation, up to 9 percent for heart failure, and up to 7 percent for mortality prediction.
These increases may seem small. But they still help separate people who will and won’t develop heart failure beyond standard methods.
With further development, the AI system could help doctors screen patients with sleep disorders for extra heart failure risk.
How It Works
During sleep, the nervous system changes rapidly. These shifts affect heart rate and electrical conduction.
Changes are especially clear in the lightest sleep stages—stages one and two. These subtle shifts may show early signs of heart dysfunction not seen during waking hours.
The autonomic nervous system regulates heart rate and blood pressure. During sleep, it shifts dynamically. These shifts can reveal abnormalities in heart-rate variability and conduction that may appear before symptoms.
Traditional models use ECG data from awake or walking patients. This study shows sleep-phase ECG data holds valuable signals—like baseline rhythms and subtle conduction delays—missing in standard tests.
The AI model was trained on public health data and tested on independent patient groups. It adjusted for known risks like age, sex, diabetes, high blood pressure, and smoking. This ensures the model’s power isn’t just from these factors.
Why It Matters
Early detection can improve outcomes and reduce long-term healthcare costs.
Polysomnography (PSG) tests often include ECG monitoring. But clinicians rarely act on subtle findings. This research suggests even small heart activity changes during sleep could be early warnings.
David Goff, acting director of the National Heart, Lung, and Blood Institute (NHLBI) at the NIH, said the method could identify at-risk people years before symptoms appear. He added that testing whether this improves traditional risk models is a key next step.
Gari Clifford, a professor of biomedical engineering at Georgia Institute of Technology, said adding the AI score to sleep studies could improve early detection of cardiovascular disease. He said: ‘Adding these scores into sleep studies has great potential for improving early detection of cardiovascular disease.’
Limitations and Open Questions
The study shows strong predictive value. But it doesn’t prove clinical usefulness yet. It’s unclear how the model would perform in diverse groups—like those with different ethnicities, body types, or health conditions—compared to the mostly white, middle-aged cohort in the original data.
The model only showed a significant risk of heart attack in one group. This may reflect data limits or the complex nature of heart attacks, which involve more than just heart rhythm.
More research is needed to see how this tool fits into real-world clinical settings with varied patient conditions and equipment.
What to Watch Next
Future work will test the model in broader, real-world populations. Researchers plan to validate it across different ages, genders, and regions to ensure fair performance.
Experts say the next step is to test whether this AI tool improves patient care by identifying high-risk individuals for heart failure or stroke. If proven, such a tool could become standard in routine sleep evaluations—especially for people with sleep disorders.
For now, this research offers a new way to interpret sleep data—and a reminder that even quiet moments of rest may hold vital health insights.
Original source: ScienceAlert
Sources & further reading
Featured image: The Way You Sleep Can Reveal Your Risk of Heart Failure Up to 10 Years in Advance. Via original publisher. Image source
