AI sleep study predicts health risks

A multidisciplinary research team, brought together through the Discovery Accelerator partnership between Cleveland Clinic and IBM, developed an artificial intelligence model that analyzes routine overnight sleep study (polysomnography) data to identify patients' long-term health risks. Published in Tata Communications on August 3, 2026, the study found that the AI model could uncover hidden physiological patterns associated with heart disease, cognitive decline, and mortality. Patients classified in the highest-risk group had twice the five-year mortality risk compared to the lowest-risk group, a distinction not captured by the standard apnea-hypopnea index. The model grouped patients into five risk categories and performed well for both men and women, addressing historical gender biases in sleep apnea assessment. Findings were independently confirmed in a nationwide patient cohort. The researchers emphasize the need for further validation in diverse populations before clinical use. The study suggests that routine medical tests may contain more physiological information than currently extracted, potentially transforming sleep studies into a richer source of prognostic information.