Higher REM sleep duration associated with reduced risk of 83 health conditions
A study of over 95,000 UK Biobank participants used wrist accelerometers and a deep-learning algorithm to measure sleep stages. Over a median follow-up of 8.9 years, greater REM sleep was linked to lower risks of 83 diseases including dementia, heart failure, and Parkinson's, while more deep sleep was associated with reduced risks of type 2 diabetes and major depression. Sleeping less than five hours per night was tied to higher disease risk.
The study used wrist accelerometers and a deep-learning algorithm to objectively measure sleep stages in over 95,000 UK Biobank participants—an improvement over self-reported sleep, which often correlates poorly with objective assessments. Participants were followed for a median of 8.9 years, with health records covering more than 1,000 disease outcomes. The context includes China's National Health Commission prioritizing sleep, with a 2026 white paper showing 23 percent of surveyed citizens sleep under six hours nightly.
Beyond REM and deep sleep findings, the study linked greater sleep irregularity and wakefulness after sleep onset to higher risks of several conditions, including anxiety and substance use disorders. Total sleep duration showed a non-linear relationship with disease risk, with the lowest risk in a 6-to-8-hour window and the highest vulnerability among those sleeping under five hours.
This research could reshape how clinicians and public health officials approach sleep guidance, moving beyond simple duration recommendations to emphasize sleep architecture—the balance of REM and deep stages. For middle-aged and older adults, the findings may encourage more consistent sleep schedules and highlight risks of chronic short sleep. However, as an observational study, causality remains unproven, and individuals should not interpret these results as a guarantee of disease prevention, but rather as additional motivation for healthy sleep habits.