Sleep recordings can help identify future health risks: Study

A study published in Nature Medicine, titled 'A multimodal sleep foundation model for disease prediction,' developed SleepFM, an AI model trained on more than 585,000 hours of sleep recordings from approximately 65,000 participants that identified patterns associated with future risk of 130 conditions, including dementia, heart failure and chronic kidney disease. Doctors commenting on the findings, including pulmonologist Dr Majeed Pasha, cautioned that identifying statistical associations differs from predicting an individual's future health and that the model must be validated across different populations and clinical settings, such as Indian sleep-laboratory data. The research suggests sleep data may offer health insights beyond conventional sleep-disorder detection, though clinicians said sleep changes alone cannot establish disease risk and decisions on further testing should remain guided by medical history and clinical evaluation.

Sleep recordings could help identify patterns associated with a person’s future risk of diseases ranging from heart attacks to dementia, according to a study published in the medical journal, Nature Medicine..The findings suggest that sleep data may offer insights into health beyond the detection of conventional sleep disorders, though identifying risk patterns is not the same as predicting whether an individual will develop a disease..The study, entitled ‘A multimodal sleep foundation model for disease prediction’, developed SleepFM, an Artificial Intelligence AI model trained on more than 5,85,000 hours of sleep recordings from approximately 65,000 participants. The model identified patterns associated with the future risk of 130 conditions, including dementia, heart failure and chronic kidney disease..Dr Chirag D, Consultant – Interventional Cardiology, said changes during sleep, such as pauses in breathing, fluctuations in oxygen levels and sudden spikes in heart rate, could point to underlying problems such as obstructive sleep apnoea..“AI combined with multiple data streams could identify patterns indicating predisposition to disease, even when individual measurements appear normal,” he said..Sleep apnoea can trigger the body’s stress response and raise blood pressure, while changes in heart rate may reflect alterations in autonomic nervous system activity. However, these changes do not necessarily mean that a particular disease is developing, the doctors note..RIP Bappi Lahiri: Sleep apnea, which caused Bappi Da's death, is common but serious disorder, say experts.Dr Majeed Pasha, Consultant Pulmonologist, said identifying statistical associations was different from accurately predicting an individual’s future health. Sleep patterns vary with work schedules, socio-economic conditions, lifestyle and cultural practices, making independent validation using Indian sleep-laboratory data important, he noted..The model must be tested across different populations and clinical settings, with doctors continuing to guide decisions on further testing and treatment, he added..Dr B K Madhusudhan, Lead Consultant – Neurology, said disturbances such as frequent waking, altered sleep cycles and changes in brain-wave activity were being studied for possible links with neurodegenerative conditions such as Alzheimer’s disease but sleep changes alone cannot establish dementia risk..Research has examined the role of slow-wave sleep, rapid eye movement REM sleep and sleep-related brain activity in memory and brain function..Dr G R K Sarma, Senior Consultant Neurologist, said: “Sleep recordings capture multiple physiological signals, including brain activity, heart rate, breathing and oxygen levels. AI could help identify patterns that conventional interpretation may miss.”.Dr Sarma said the findings must be assessed alongside a person’s medical history, symptoms and other risk factors, with further tests guided by clinical evaluation..The technology could eventually support preventive healthcare by helping clinicians identify people who may need further evaluation.