More REM Sleep, Lower Risk of 83 Diseases — in One Very Large, Very Observational Study
A 95,559-person UK Biobank analysis tied more REM sleep to lower risk of dozens of conditions, from heart failure to dementia. It's a strong signal, but it shows association, not cause.
Singularity Wellness Editorial · 2026-10-17 · 3 min read
More REM Sleep, Lower Risk of 83 Diseases — in One Very Large, Very Observational Study
A study published in PLOS Medicine on September 17, 2026 analyzed 95,559 UK Biobank participants who wore wrist accelerometers for a week and were followed for a median of 8.9 years against more than 1,000 disease outcomes. Each interquartile-range increase in REM sleep (about 47.6 more minutes) was associated with lower risk of 83 diseases, including heart failure, dementia and Parkinson's; more deep sleep tracked with lower risk of 7 conditions. Total sleep of about 6 to 8 hours was tied to the lowest risk. The central limitation is that this is observational — it cannot prove that more REM sleep prevents disease, sleep stages were estimated by a wrist device rather than lab polysomnography, and early undiagnosed illness could be changing sleep rather than the reverse.
A study published in PLOS Medicine on September 17, 2026, by Shengzhi Sun and colleagues at Capital Medical University, is one of the largest attempts yet to connect the architecture of a night's sleep — not just its length — to long-term health. The scale is the story, and so is the caveat that comes with it.
The researchers analyzed 95,559 UK Biobank participants who wore a wrist accelerometer for seven consecutive days and nights, then followed them for a median of 8.9 years, testing their sleep against more than 1,000 disease outcomes. That breadth is what lets a single study speak to dozens of conditions at once.
What They Found
Each interquartile-range increase in REM sleep — roughly 47.6 more minutes a night — was associated with lower risk of 83 separate diseases. The associations were sizable for some serious conditions: a hazard ratio of 0.74 for heart failure, 0.54 for dementia, and 0.20 for Parkinson's disease, where a lower hazard ratio means lower risk. Greater deep (slow-wave) sleep was separately associated with lower risk of 7 conditions, including type 2 diabetes and major depressive disorder.
- Heart failure — hazard ratio 0.74
- Dementia — hazard ratio 0.54
- Parkinson's disease — hazard ratio 0.20
Total sleep duration still mattered. About 6 to 8 hours a night was associated with the lowest overall disease risk, while sleeping less than 5 hours was associated with higher risk of 37 conditions — a reminder that the amount of sleep and its composition are related but distinct questions.
Why 'Associated With' Is Doing a Lot of Work
This is an observational study, which means it can show that patterns travel together but cannot prove one causes the other. Two limitations matter especially here. First, the sleep stages were estimated by a wrist accelerometer, not measured by in-laboratory polysomnography, the gold standard. Second, reverse causation is a live possibility: early, undiagnosed disease — a neurodegenerative process, for instance — can alter sleep well before symptoms appear, which would make disrupted sleep a marker of illness rather than a cause of it. The data also came from a single week of wear.
This study shows a statistical association between sleep patterns and disease risk. It does not prove that increasing REM sleep prevents disease, and it does not validate any wearable or app as a tool that can do so.
The honest takeaway is about protecting total sleep opportunity and consistency — not chasing a REM number on a tracker.
What to Actually Do With This
No wearable or app has been shown to raise REM sleep on command, or to change these disease outcomes by doing so. REM and deep sleep tend to follow from the basics the evidence has long supported: enough time in bed, a consistent schedule, and a sleep environment that doesn't fight you — which is where paying attention to the room, not just the number, earns its place in the longevity loop.