Scientists have developed a new artificial intelligence model capable of assessing the risk of developing more than one hundred diseases by analyzing physiological data collected during a single night of sleep.

The study, published in Nature Medicine, shows that recordings of physiological signals during sleep—such as brain electrical activity, heart rate, and breathing—may contain early signs of future health problems, even before symptoms appear.

The AI model was trained on a large dataset and identifies patterns linking sleep characteristics to more than one hundred categories of disease, including cardiovascular conditions, diabetes, certain types of cancer, and mental health disorders. The authors believe that analyzing nighttime physiological signals could become a convenient and non-invasive method for early disease risk detection, as such data are readily obtained using polysomnography, the standard clinical method for recording sleep parameters.

The researchers emphasize that poor sleep, beyond its immediate effects such as fatigue and reduced concentration, may also reflect deeper disruptions in the body that can eventually develop into serious diseases. By analyzing a combination of these signals, the new AI model can detect complex patterns that are difficult to identify using conventional approaches.

This method opens the door to personalized disease prevention: if the system predicts a high risk for a specific condition, doctors and patients could adjust lifestyle factors or begin targeted monitoring in advance to help prevent the disease from developing.

The authors note that the technology still requires further validation and clinical trials, but it already shows potential for integration into healthcare systems and use in preventive medicine. Such an approach could significantly change strategies for early disease detection and help millions of people receive risk warnings years before clear symptoms emerge.