Stanford Researchers Build AI That Predicts Disease Risk From One Night of Sleep

Stanford Researchers Build AI That Predicts Disease Risk From One Night of Sleep

A new artificial intelligence model developed by Stanford Medicine researchers can use data from a single night of sleep to predict a person's risk of developing more than 100 different health conditions—including Parkinson's disease, dementia, heart attack, and several types of cancer.

The model, called SleepFM, represents a significant advance in using AI for preventive healthcare. Rather than waiting for symptoms to appear, it can identify disease risk years before traditional diagnostic methods would detect anything wrong.

How SleepFM Works

SleepFM was trained on nearly 585,000 hours of sleep data collected from 65,000 participants. The data comes from polysomnography—the comprehensive sleep assessment used in sleep clinics that records brain activity, heart activity, respiratory signals, leg movements, eye movements, and more.

"We record an amazing number of signals when we study sleep," said Dr. Emmanuel Mignot, Craig Reynolds Professor in Sleep Medicine and co-senior author of the study, in a Stanford Medicine news release. "It's a kind of general physiology that we study for eight hours in a subject who's completely captive. It's very data rich."

The researchers built SleepFM as a foundation model—similar to how large language models like ChatGPT are trained on vast amounts of text. In this case, SleepFM learned what they describe as "the language of sleep," training on five-second increments of sleep data.

Predictive Accuracy

After training, the researchers fine-tuned SleepFM to predict future disease onset by pairing sleep data with decades of health records from patients at the Stanford Sleep Medicine Center, which has been operating since 1970.

The results were notable. SleepFM achieved a C-index (a measure of predictive accuracy) higher than 0.8 for several conditions:

• •Parkinson's disease: 0.89
• •Prostate cancer: 0.89
• •Breast cancer: 0.87
• •Dementia: 0.85
• •Hypertensive heart disease: 0.84
• •Death: 0.84
• •Heart attack: 0.81

A C-index of 0.8 means the model's prediction is correct 80% of the time when comparing any two individuals to determine which one will experience an event first.

What the Model Is Detecting

The AI identified that certain patterns in sleep physiology correlate with future disease risk. According to Dr. James Zou, associate professor of biomedical data science and co-senior author, the most predictive signals came from contrasting different data modalities.

"Body constituents that were out of sync—a brain that looks asleep but a heart that looks awake, for example—seemed to spell trouble," Mignot noted.

This suggests that the model is picking up on subtle physiological mismatches that wouldn't be apparent to human clinicians reviewing standard sleep reports.

Implications for Healthcare

The researchers acknowledge that SleepFM is not yet ready for clinical use. The study, published in Nature Medicine in January 2026, demonstrates proof of concept rather than a diagnostic tool.

However, the implications could be significant. If validated in broader populations, sleep-based disease prediction could offer a non-invasive, relatively accessible way to identify health risks early. Unlike blood tests or imaging, polysomnography is already widely available at sleep clinics, and the researchers are exploring whether consumer wearables could eventually capture enough data for similar predictions.

"We were pleasantly surprised that for a pretty diverse set of conditions, the model is able to make informative predictions," Zou said. "From an AI perspective, sleep is relatively understudied. There's a lot of other AI work that's looking at pathology or cardiology, but relatively little looking at sleep, despite sleep being such an important part of life."

Sources

• •Stanford Medicine News: New AI model predicts disease risk while you sleep (January 6, 2026)
• •ScienceDaily: Stanford's AI spots hidden disease warnings that show up while you sleep