REM Sleep and Disease Risk: 95,000-Adult Study Findings
Most of us grade a night's sleep on one number. Six hours, rough. Eight hours, human again. A large study published this month suggests a second number deserves some attention: how much of the night you spend in REM sleep, the stage when most dreaming happens.
Researchers followed 95,559 adults in the UK for roughly nine years and found that people with more REM sleep had a lower risk of 83 different diseases, among them dementia, heart failure and Parkinson's disease. The paper appeared September 17 in PLOS Medicine. It is a well-built study. It also cannot show that REM sleep protects anyone, and that limit matters as much as the big number in the headline. Below is what the team did, what they found, and what is sensible to do with it.
A quick refresher on sleep stages
Over a night, you cycle through light sleep, deep sleep and REM several times. Deep sleep is the heavy, hard-to-wake stage. REM is different. Brain activity looks closer to waking, dreams tend to be most vivid, and the muscles are largely switched off, which is what keeps you from acting out those dreams. In a typical adult night, REM makes up roughly a fifth to a quarter of total sleep, though it varies with age. So a gap of 48 minutes between two sleepers is sizeable.
How researchers measured sleep in 95,000 people
The data came from UK Biobank, a large health database. Participants wore a wrist accelerometer, a small motion sensor, for seven consecutive days and nights. A deep-learning algorithm then turned that movement data into estimates of REM, deep and light sleep, plus total sleep time, time awake after first dozing off, and how much sleep varied from night to night, according to the PLOS summary published by ScienceDaily.
The average participant was 56, and just over half were women, Medscape reported. Over a median follow-up of 8.9 years, the team compared sleep patterns against health records for more than 1,000 diseases and conditions.
The method matters because most sleep research leans on people reporting how they slept, and the study's press summary notes that self-reports line up poorly with objective measures. The alternative, an overnight lab recording, is accurate but usually limited to small groups sleeping in an unfamiliar bed. A week at home with a wrist sensor captures more ordinary nights for far more people. The trade-off is that these sleep stages are estimates from movement, not brain-wave readings.
What they found about REM sleep
Each increase of about 48 minutes in REM sleep (47.6, to be exact) was linked to a lower risk of 83 diseases across 12 major disease categories, per Medscape. The examples the authors highlighted show the range: heart failure (hazard ratio 0.74), dementia (0.54) and Parkinson's disease (0.20). A hazard ratio below 1 means a lower rate of the disease among people with more REM, after the researchers accounted for factors such as age, sex, lifestyle and socioeconomic status. Alzheimer's disease and multiple sclerosis were also on the list, U.S. News reported.
Deep sleep showed a narrower pattern. More of it was tied to lower risk of seven conditions, including type 2 diabetes and major depression. Light sleep was a mixed bag. More of it went with higher risk for 28 diseases, major depression among them (hazard ratio 1.31), yet with lower risk of heart failure and cellulitis. Irregular sleep and more time awake after falling asleep were each linked to higher risk of several conditions, including anxiety and substance use disorders.
Take the Parkinson's figure with some salt. A hazard ratio of 0.20 is dramatic, and dramatic numbers in observational work often have a less exciting explanation, which brings us to the caveats.
Hours still count
REM wasn't the only story. Total sleep time didn't have a straight-line relationship with disease risk. For most of the 86 conditions where duration mattered, risk bottomed out inside a window of six to eight hours. People sleeping under five hours faced a higher risk of 37 conditions.
That window reaches a little lower than official guidance. The CDC recommends adults get at least seven hours, and a joint consensus statement from the American Academy of Sleep Medicine and the Sleep Research Society reached the same conclusion for adults 18 to 60. The two aren't in conflict. The study looked at middle-aged and older adults and reported where risk was lowest, not what any one person needs. Someone averaging six hours sits inside the study's window but below the CDC's recommendation, so treating seven as the target is the safer reading.
Why the results don't prove REM sleep protects you
The authors are clear that this is an observational study, so it can't establish that sleep patterns cause disease risk. There are four reasons to hold the findings loosely.
First, the arrow may point the other way. Medscape noted the study can't tell whether sleep changes contribute to disease or result from developing health problems. That's a real concern for brain diseases, where sleep can shift years before a diagnosis. ScienceDaily has covered research on REM sleep behavior disorder and dementia prediction along those lines.
Second, testing more than 1,000 outcomes invites false alarms. The team applied a statistical correction for that, and 156 associations between sleep patterns and disease survived it. That's reassuring, but it isn't proof.
Third, the sleep stages are algorithm estimates, and the volunteers were mostly middle-aged and older, so the findings may not carry over to younger adults.
Fourth, one study is one study. There is supporting evidence. An earlier analysis of the Sleep Heart Health Study, which used overnight lab recordings from 4,490 people, also tied more REM sleep to a lower risk of heart failure. Agreement across different methods is what you want to see. It still isn't a trial.
For transparency, the work was supported by Chinese national and Beijing research grants, and the paper states that the funders had no role in the design, analysis or decision to publish.
What you can actually do with this
None of this calls for a new gadget or supplement. A few steps fit both the new findings and older, sturdier advice.
Protect the time. Aim for at least seven hours. Sleep specialists generally describe REM periods as growing longer toward morning, so cutting the night short tends to trim REM first.
Keep the schedule steady. Irregular sleep was flagged in this study, and a consistent bedtime and wake time costs nothing.
Don't obsess over your watch's stage chart. Consumer wearables estimate stages too, and this study can't tell you whether nudging your nightly REM number up by a few minutes would change anything about your health.
Get persistent problems checked. Loud snoring, gasping, or waking unrefreshed after a full night can point to a condition such as sleep apnea, which the Sleep Foundation notes can make sleep unrefreshing no matter how many hours you log. A doctor can sort that out. A tracker can't.
The bottom line
The strongest reading of this study is modest and still useful. In a large group of adults followed for years, sleep with more REM, enough deep sleep and a steady six-to-eight-hour pattern tracked with better long-term health. Whether improving REM would change anyone's risk is a question for trials, and those haven't been done. Until then, the boring advice holds: get enough sleep, keep it regular, and see a doctor if it never feels restful.
This article is for general information and isn't medical advice. Talk with a qualified clinician about your own sleep and health.
Sources and further reading