More REM Sleep Was Linked to Lower Risk of 83 Diseases in a 95,000-Person Study
Evidence review & editorial synthesis
Most people reduce sleep to one number: hours.
A new study of 95,559 adults suggests that may miss much of the story. Researchers reconstructed REM, deep and light sleep from a week of wrist-accelerometer recordings, then tracked participants' health for a median of 8.9 years. Greater REM sleep was associated with lower incidence of 83 diseases across 12 disease categories. Deep sleep showed its own pattern, including lower risks of type 2 diabetes and major depression.
The implication is not that everyone should chase a particular REM score. It is that the internal architecture of sleep may contain far more information about long-term health than total sleep time alone.
REM sleep produced the broadest disease signal
Participants were an average of 56 years old and wore a wrist accelerometer continuously for seven days. A deep-learning system called SleepNet estimated when they were awake and how much time they spent in REM, light and deep sleep.
Researchers then examined 1,049 subsequent health outcomes.
The strongest stage-specific pattern appeared in REM sleep. An interquartile difference of 47.6 minutes more REM was associated with lower incidence of 83 diseases.
The associations crossed organ systems. Heart failure risk was 26% lower, atrial fibrillation 17% lower and ischemic heart disease 13% lower across that REM difference. Dementia was associated with a 46% lower risk, Alzheimer disease with 31% lower risk and multiple sclerosis with 52% lower risk.
The largest association was with parkinsonism, where the hazard ratio was 0.20.
That last number is striking, but it is also exactly where caution matters. Neurodegenerative diseases can alter sleep years before their classic symptoms appear. Less REM may therefore sometimes be an early marker of disease already developing rather than the cause of what follows.
Still, the REM finding is not entirely new. Earlier studies using full polysomnography have independently associated lower REM percentages with greater dementia risk and higher mortality. The new research dramatically expands the scale and breadth of those observations.
Deep sleep told a different story
REM was not simply a proxy for all good sleep.
Deep sleep produced a much narrower disease signature. A 47.5-minute interquartile difference in deep sleep was associated with lower incidence of seven conditions across five categories.
Among the most recognizable were type 2 diabetes, with an 11% lower risk, major depressive disorder with a 14% lower risk and Parkinson disease with a 30% lower risk.
Light sleep showed a more complicated pattern. Greater light-sleep duration was associated with a 31% higher risk of major depressive disorder, yet with lower incidence of some other conditions including heart failure.
That contrast is important. Sleep does not appear to behave like one uniform block of unconscious time. REM, deep sleep and lighter stages are biologically different states, with different patterns of brain activity, autonomic function and neurochemistry.
REM is strongly involved in memory and emotional processing. Deep non-REM sleep has been linked to restorative physiology and cerebrospinal-fluid dynamics associated with metabolic waste exchange in the brain. These mechanisms are still being worked out, but they offer plausible reasons why changing sleep architecture could accompany very different long-term health trajectories.
Six to eight hours repeatedly landed in the lowest-risk zone
Sleep stages were only part of the study.
Total sleep duration showed nonlinear relationships with 86 disease phenotypes. For 69 of them, the estimated minimum-risk duration clustered between six and eight hours per night.
The opposite extreme was clearer still. People sleeping fewer than five hours showed the broadest pattern of elevated disease risk, accounting for 37 of the 41 significant adverse associations identified when sleep-duration categories were compared.
That does not mean six hours is ideal for everyone. The American Academy of Sleep Medicine recommends that adults regularly obtain at least seven hours, and individual needs vary. The important point is that very short sleep performed badly across an unusually wide range of outcomes.
The study also examined continuity and timing. More wakefulness after sleep onset was associated with higher incidence of six conditions. Greater night-to-night irregularity was associated with 23% higher anxiety risk and 26% higher major-depression risk.
So the emerging picture is broader than simply getting enough hours. Duration, continuity, regularity and architecture may all matter.
A week of sleep may reveal years of health information
One of the most intriguing aspects of the study is methodological.
Participants were not monitored for years in sleep laboratories. Researchers took approximately one week of real-world movement data and extracted a multidimensional sleep profile from it. That profile was then associated with diagnoses occurring over nearly nine years of follow-up.
After stringent multiple-testing correction, sleep measures were linked to 156 incident disease outcomes. Even after researchers excluded the first two years of follow-up to reduce the possibility that undiagnosed illness was already affecting sleep, 95 of those associations remained significant.
That raises a future possibility: sleep architecture could eventually become a broad physiological biomarker rather than simply a wellness metric.
A future wearable might not merely tell someone they slept seven hours. It could potentially identify persistent changes in sleep structure that warrant closer medical attention.
But today's consumer trackers are not there yet. The study used a research algorithm on raw accelerometer data, while clinical sleep staging still relies on signals such as brain electrical activity, eye movements and muscle tone. A watch reporting less REM on Tuesday morning is not evidence that someone is developing dementia or heart disease.
Do not try to hack REM yet
The headline number makes an obvious question tempting: how do I get more REM?
The study cannot answer that.
Its hazard ratios compare people with different sleep patterns. They do not show that experimentally adding REM sleep reduces disease incidence. Low REM could arise from sleep disorders, medications, alcohol, medical illness, neurodegeneration or other factors that themselves affect health.
For now, the practical signal is much less exotic: protect adequate, regular and uninterrupted sleep rather than trying to manipulate a single stage based on a wearable score. Persistent snoring, breathing pauses, severe insomnia, excessive daytime sleepiness or chronically short sleep deserve more attention than optimizing a nightly REM percentage.
But scientifically, the new study changes the scale of the question.
Sleep may not be valuable merely because the brain and body stop being awake for several hours. What happens inside those hours appears to matter too.
And in this enormous dataset, REM sleep carried a health signature spanning dozens of diseases that reached from the heart to the brain.
Scientific References & Further Reading
- Anchor Study: “Accelerometer-derived real-world sleep stages and risk of incident diseases: A UK Biobank cohort study and phenome-wide association analysis.” (Luo J, Liu R, Yin J, Cao W, Sun S, Chen R. PLoS Med. 2026;23(9):e1005213. doi:10.1371/journal.pmed.1005213. PMID:42752434. The source candidate was discovered through a ScienceDaily report describing more REM sleep as associated with lower risk of 83 diseases. :contentReference[oaicite:0]{index=0})
- REM Sleep and Mortality: “Association of Rapid Eye Movement Sleep With Mortality in Middle-aged and Older Adults.” (Leary EB, Watson KT, Ancoli-Israel S, et al. JAMA Neurol. 2020;77:1241-1251. doi:10.1001/jamaneurol.2020.2108.)
- REM Sleep and Dementia: “Sleep architecture and the risk of incident dementia in the community.” (Pase MP, Himali JJ, Grima NA, et al. Neurology. 2017;89:1244-1250. doi:10.1212/WNL.0000000000004373.)
- Adult Sleep Duration Guidance: “Recommended Amount of Sleep for a Healthy Adult: A Joint Consensus Statement.” (Watson NF, Badr MS, Belenky G, et al. J Clin Sleep Med. 2015;11:591-592.)
- Sleep and Brain Clearance: “Sleep, cerebrospinal fluid, and the glymphatic system: A systematic review.” (Chong PLH, Garic D, Shen MD, Lundgaard I, Schwichtenberg AJ. Sleep Med Rev. 2022;61:101572. doi:10.1016/j.smrv.2021.101572.)
- REM and Memory: “REM refines and rescues memory representations: a new theory.” (Neuroscience review of REM sleep, memory refinement and emotional processing.)
Disclaimer: This article is for informational purposes only and does not constitute medical advice. It is not a substitute for diagnosis, treatment, or individualized medical care.

