02/22/2026
Sleep is not just rest. It's data. And it may be one of the earliest signals your body gives you about future disease.
In a January 2024 study published in Nature Medicine, Stanford Medicine researchers introduced SleepFM, an AI model trained on nearly 600,000 hours of polysomnography sleep data from 65,000 people. After analyzing just one night of sleep, the model was able to predict risk for more than 100 health conditions, often years before symptoms appeared.
The strongest predictive signals were linked to dementia, Parkinson’s disease, heart disease, cancer, and mental health disorders. This is because sleep captures a rare, uninterrupted snapshot of whole-body physiology, including brain activity, heart rhythms, breathing patterns, and more, all interacting at once.
As the researchers explain, sleep is a data-rich window into general physiology. When these systems fall out of alignment, disease risk can surface long before a diagnosis is made.
Sleep is not optional in longevity, it's foundational.
To learn more about how you can support sleep, tap the link in bio.
(For educational purposes only)