Getting enough REM and deep sleep may be associated with a lower risk of dozens of diseases, while sleeping fewer than five hours a night is linked to substantially higher disease risk, according to a new study published September 17 in the open-access journal PLOS Medicine. The research was led by Shengzhi Sun of Capital Medical University in China and colleagues.
Sleep quantity and quality have become important public health concerns in China. The 2026 China Sleep Health White Paper reports that surveyed Chinese adults sleep an average of 6.97 hours a night, while 23% sleep fewer than six hours. Only 26% reported maintaining a consistent sleep schedule, highlighting growing concerns about disrupted sleep and its potential effects on health.
Despite increasing attention to sleep, researchers still have a limited understanding of how specific sleep stages and everyday sleep patterns are related to long-term health. One challenge is that self-reported measures of sleep do not always correspond closely with objective measurements, making it difficult to assess how people actually sleep accurately.
In the new study, researchers analysed data from 95,559 UK Biobank participants who wore wrist accelerometers continuously for seven days and nights. Using a deep-learning algorithm, the researchers estimated REM, deep and light sleep, as well as total sleep duration, wakefulness after falling asleep and night-to-night sleep irregularity.
Participants were followed for a median of 8.9 years, allowing the researchers to examine associations between sleep patterns and more than 1,000 health outcomes. Overall, the analysis identified 156 significant associations between sleep patterns and newly diagnosed diseases after applying a stringent statistical correction.
More REM sleep was associated with a lower risk of 83 diseases, including heart failure, dementia and Parkinson’s disease. Greater amounts of deep sleep were associated with a lower risk of seven conditions, including type 2 diabetes and major depressive disorder. In contrast, greater sleep irregularity and more time awake after initially falling asleep were associated with increased risks of several conditions, including anxiety and substance use disorders.
Total sleep duration also showed an important relationship with health. For many conditions, the lowest disease risk was concentrated within a six-to-eight-hour sleep window. Researchers found significant non-linear associations between sleep duration and 86 disease outcomes, with the minimum-risk duration for 69 of them falling within that range. People who slept fewer than five hours had an increased risk of 37 conditions.
The researchers said the findings provide additional evidence that six to eight hours of sleep may be beneficial for middle-aged and older adults, potentially because this duration supports more favourable distributions of different sleep stages. However, the study was observational, meaning it cannot establish that particular sleep patterns directly prevent or cause disease. The findings nevertheless suggest that both how long people sleep and the quality and structure of that sleep may be important considerations for long-term health.
More information: Jingsong Luo et al, Accelerometer-derived real-world sleep stages and risk of incident diseases: A UK Biobank cohort study and phenome-wide association analysis, PLOS Medicine. DOI: 10.1371/journal.pmed.1005213
Journal information: PLOS Medicine Provided by PLOS