95,559 wrists tracked for a week: the sleep stage linked to 83 diseases
What happened
Most of us think about sleep as one thing. You go to bed, you get up, and you either slept well or you did not. A new study says the picture is more detailed than that. Researchers looked at 95,559 adults who took part in UK Biobank, a huge British health study. Each person wore a wrist tracker for seven days and nights in a row. A computer program worked out how much time they spent in each stage of sleep: light sleep, deep sleep and REM sleep. REM is the stage where most dreaming happens. The team then followed those people through their NHS records for a median of 8.9 years. They checked more than 1,000 different diseases and health conditions to see who developed what. The headline result is striking. People who got more REM sleep were less likely to go on to develop 83 different diseases. More deep sleep was linked to a lower risk of seven conditions. Altogether the team reported 156 significant links between sleep patterns and later illness. The work was led by Shengzhi Sun at Capital Medical University and was published in the journal PLOS Medicine on 17 September 2026.
Why this matters
Sleep advice is everywhere, and a lot of it is guesswork. What makes this study unusual is its size and its method. Most sleep research asks people to fill in a questionnaire. People are not very good at guessing how long they slept, and almost nobody can guess how much REM sleep they had. This study used wrist sensors instead, recording real movement data night after night. It also did not start with a hunch. Rather than asking 'does sleep affect dementia?', the researchers tested sleep against over a thousand conditions at once and let the data point them somewhere. That is a more honest way of looking. The conditions on the list are not minor. They include heart failure, dementia and Parkinson's disease. For a country where roughly one in three adults says they sleep badly, that is worth paying attention to.
What the evidence actually says
Here are the numbers, plainly. For every extra 47.6 minutes of REM sleep, the risk of 83 diseases was lower. For heart failure the hazard ratio was 0.74. For dementia it was 0.54. For Parkinson's disease it was 0.20, the largest drop the team found. A hazard ratio below 1 means lower risk. But these are averages across a very large group. They are not a promise about any one person. Total sleep time mattered too. Six to eight hours a night was linked to the lowest overall risk. People sleeping under five hours had a higher risk of 37 different conditions. Now the important part. This is an observational study. The authors say so themselves. It can show that two things travel together. It cannot show that one causes the other. That matters here more than usual, because illness can change sleep long before it is diagnosed. Parkinson's disease is a good example. Doctors have known for years that people with Parkinson's often have disturbed REM sleep, sometimes a decade before other symptoms appear. So the very large effect for Parkinson's may partly be the disease affecting sleep, rather than sleep affecting the disease. This is called reverse causation, and it is a real limitation. The study also measured sleep over just one week, near the start of the follow-up period. Sleep changes over a lifetime. None of this makes the findings worthless. It makes them a signal to follow up, not an instruction.
Practical advice
You cannot directly control how much REM sleep you get. There is no switch. What you can do is protect the conditions that let normal sleep happen. Keep your wake-up time steady, including at weekends. A regular body clock does more for sleep quality than almost anything else. Aim for a realistic total. Six to eight hours came out best in this study. Chasing nine hours is not the goal. Watch alcohol in the evening. It makes people fall asleep faster, then suppresses REM sleep in the second half of the night. That is one of the better-established effects in sleep science. Check your medicines with a pharmacist, not with the internet. Some antidepressants, some blood pressure tablets and some painkillers affect sleep. Sometimes the timing of a dose can be adjusted. Never stop or change a prescribed medicine on your own. If you snore heavily, stop breathing in your sleep, or feel exhausted all day despite enough hours in bed, speak to your GP. Sleep apnoea is common, treatable, and often missed. And be careful with your own sleep tracker. Consumer devices estimate sleep stages, they do not measure them the way a sleep clinic does. Watching the numbers can make anxiety about sleep worse, which makes sleep worse. If your tracker is making you tense, put it in a drawer.
What to know
A study of 95,559 UK adults, followed for almost nine years, found that more REM sleep was linked to a lower risk of 83 diseases, and that six to eight hours a night was linked to the lowest overall risk. It is a large, careful, well-designed study. It is also observational, so it cannot prove that better sleep prevents disease, and some of the link may run the other way. The sensible response is not panic, and not a new gadget. It is a steadier bedtime, a word with your pharmacist about any medicine that may be disturbing your nights, and a GP appointment if daytime tiredness is a real problem. Sources: PLOS Medicine, 'Accelerometer-derived real-world sleep stages and risk of incident diseases in the UK Biobank', 17 September 2026, https://journals.plos.org/plosmedicine/ ; ScienceDaily, 'More REM sleep linked to lower risk of 83 diseases', 23 September 2026, https://www.sciencedaily.com/releases/2026/09/260923035930.htm ; Medscape, 'What 95,000 Wrist Trackers Revealed About Sleep and Disease', 2026, https://www.medscape.com/viewarticle/what-95-000-wrist-trackers-revealed-about-sleep-and-disease-2026a1000z7z This article is for general information and does not replace advice from a doctor, pharmacist or other qualified healthcare professional.
Related articles
Six hours, not never: the ICU filter notice the MHRA issued today
The UK's medicines and devices regulator has issued a new safety notice to hospitals, and it is about timing. On 13 October 2026 the Medicines and Healthcare p…
Eight pages to a minister: what the data guardian says trust is worth
On 12 October 2026, the government published a letter about who should be trusted with NHS patient data. It was written by Dr Nicola Byrne, the National Data G…
Update: the tests came back, and it was not plague
This is a follow-up to a story we covered on 5 October, when nearly 200 people were being monitored after the death of a laboratory worker in Siberia. There has…