Most of us treat sleep as a single number. Seven hours on a good night, six if the week got away from us, five when a deadline wins. A large new study says people at the bottom of that range stand out, and their medical records show more than you might guess.
Researchers at Peking University and Capital Medical University in Beijing followed 95,559 adults in the UK Biobank to see who later developed illnesses.
People sleeping less than five hours a night had higher rates of 37 conditions. Some belong in conversations about sleep already. Others might make you look twice.
Published in PLOS Medicine in September, the study also explored sleep stages and how much people’s sleep changed from one night to the next. Those findings included some striking figures. The fine print matters as much as the numbers.
The List Runs From Dementia to Constipation
Researchers compared the shortest sleepers with people who slept six to eight hours, and 37 diagnoses were more common in the short-sleep group.
The findings survived strict statistical correction for testing more than 1,000 conditions at once, as detailed in the study. Grouped by body system, they are:
- Brain and cognition: Parkinson’s disease, parkinsonism, dementias, delirium, symptoms involving cognitive function and awareness
- Mental health: alcohol use disorders, panic disorder
- Heart and circulation: heart failure, atrial fibrillation and flutter, peripheral vascular disease
- Lungs and infections: COPD, acute lower respiratory infection, lobar pneumonia, pleural effusion, sleep apnea, SARS-CoV-2 infection, urinary tract infection, bacterial infections, E. coli, cellulitis and abscess
- Metabolic, kidney and blood: type 2 diabetes, acute kidney failure, acidosis, anemia, vitamin D deficiency
- Mobility and skin: repeated falls, abnormal gait and mobility, wheelchair dependence, osteoarthritis, urinary incontinence, pressure ulcers, other chronic ulcers
- Digestive and other: constipation, gastrointestinal hemorrhage, edema, antibiotic allergy (excluding penicillin), general symptoms and other findings
The sharpest jump was in parkinsonism, with a rate nearly six times higher among short sleepers, though that estimate rests on just 31 cases. Alcohol use disorders followed at about 2.7 times, and dementias at roughly 2.6.
Constipation and antibiotic allergy may look like strange company here. But these diagnoses were recorded during hospital stays, not for every illness participants experienced.
Related: 8 Common Bedroom Mistakes That Could Be Could Be Sabotaging Your Sleep
One Week of Tracking, Nearly Nine Years of Follow-Up
All of this rests on a short sleep snapshot, so it helps to know how it was taken.
Participants averaged 56 years old and wore a wrist accelerometer for seven days and nights. Researchers fed the movement data into SleepNet, a deep-learning program that estimated total sleep, REM, deep, and light sleep; time awake after falling asleep; and how much sleep varied from night to night.
From there, the team followed health records for a median of 8.9 years and placed diagnoses into 1,049 disease categories. They excluded diagnoses from the first six months to lessen the possibility that an illness already underway was influencing sleep.
REM Sleep Had an Even Longer List
Duration got the headline, but people connected sleep type to more conditions than to hours.
Greater estimated REM sleep was linked to lower rates of 83 diseases. For every additional 47.6 minutes, heart failure rates were 26% lower, dementia rates 46% lower, and parkinsonism rates 80% lower.
Deep sleep had a shorter list: lower rates of seven conditions, including type 2 diabetes, major depressive disorder, and sleep apnea. Sleep that varied more from night to night was associated with higher rates of anxiety, depression and abdominal pain.
Save this article
Enter your email address and we'll send it straight to your inbox.
More time awake after falling asleep was linked to higher rates of alcohol abuse, substance dependence, and osteoarthritis.
Related: How Sleep, Stress, and Diet Affect Your Sexual Energy
Why the Biggest Numbers Need Fine Print
The 80% figure is the one most likely to spread, and it carries the widest margin of error.
Its statistical range runs from roughly 60% lower to roughly 90% lower. And 47.6 minutes isn’t a prescribed extra dose of REM: it’s the difference between the group’s 25th and 75th percentiles, used to express an association. No one was assigned extra REM and then followed to see what happened.
The stages themselves were estimated, too. In comparisons with laboratory sleep tests, SleepNet underestimated REM by about 17 minutes on average. The authors caution that its estimates may not fully align with biological sleep stages. They didn’t test consumer wearables.
There is another complication: illness can disrupt sleep well before anyone puts a name to it. Dr. Chelsie Rohrscheib, head sleep expert at the sleep analysis company Wesper, explained: “Conditions such as Parkinson’s disease, dementia and heart failure can begin affecting sleep years before they are diagnosed.”
Even after researchers excluded diagnoses made within two years of sleep tracking, 95 of the original 156 significant associations remained.
Six to Eight Hours Is Not a New Sleep Rule
The study’s lowest-risk range sits a little below the guidance most adults are used to hearing.
For 86 conditions, the relationship between sleep time and risk formed a curve rather than a straight line. In 69, the lowest point was between six and eight hours, slightly shorter than the commonly suggested seven to nine. But there was no single ideal number across all diagnoses: the low point for type 2 diabetes was about 7.3 hours, for dementias about 7.5.
Longer sleep wasn’t automatically better, either. More than eight hours was linked to higher rates of major depressive disorder. The authors note that excessive sleep may itself be an early symptom of depression.
The researchers conclude that sticking to six to eight hours can effectively reduce the risk of multiple diseases.
Yet this study cannot establish cause and effect. The participants were 97% White, and UK Biobank volunteers tend to be healthier and better off than the wider population.
The authors call the analysis exploratory and hypothesis-generating, and it had no pre-registered analysis plan. No trial has yet answered the big practical question: would changing someone’s sleep actually change those disease rates?
Which condition on the list surprised you most, and why?
More Articles You Might Be Interested in:
- 10 Unexpected Things You Can Do When You Can’t Sleep at Night
- Having Trouble Sleeping? This Japanese Cypress Oil May Help You Sleep Better, Study Finds
- Running on No Sleep? A 20-Minute Workout or Nap May Help Protect Your Memory, Study Finds
