Eight hours a night is what counts
Short answerNot the way you think
On this page
Not the way you think. How long you sleep counts at the extremes, and very short and very long sleep both go with higher death rates. But in 60,977 UK adults who wore activity trackers, how regular their sleep was predicted death better than how long it was. The most regular sleepers had a 20 to 48% lower death rate than the least regular fifth.
What we’d do
Fix your wake time first. Get up at the same time every day, weekends included, within about an hour. Then go to bed when you’re sleepy enough to get a full night, and let your bedtime settle around that. Don’t chase a perfect eight on your tracker, and don’t worry over one short night.
If you remember three things
- Regular beat long. In 60,977 people, the regularity of sleep predicted death better than its length.
- The extremes still count. In 16 studies of 1.38 million people, short sleepers had a 12% higher death rate, and long sleep went with higher risk too.
- Start with the time you get up. It’s the part of your sleep you can set with an alarm.
Why everyone believes it
Eight hours is the number everyone grows up with, and most sleep trackers lead with how long you slept. Weekends feel like the time to catch up, so it’s common to sleep short on weeknights and long on Saturday, and call it eight on average. Regularity rarely gets mentioned at all.
A picture from daily life
Think of a bus that’s supposed to come at 7:05. If it comes at 7:05 every day, you can plan your whole morning around it. If it comes anywhere between 6:30 and 9, everything else in your day gets thrown off, even if the bus covers the same distance. Sleep that comes at the same time each day looks like it works the same way.
What did the study find?
In 2024 Windred and colleagues worked through more than 10 million hours of activity tracker data from 60,977 people in the UK Biobank, with an average age of 63. For each person they scored how regular their sleep was, meaning how similar their sleep and wake times were from one day to the next. Then they followed them for about 6 years, in which 1,859 died.
| Compared with the least regular fifth | Lower risk in the more regular groups |
|---|---|
| Death from any cause | 20 to 48% |
| Death from cancer | 16 to 39% |
| Death from heart disease and metabolic disease | 22 to 57% |
The results allowed for age, sex, ethnicity, money, lifestyle and health. When the researchers compared models side by side, regularity predicted death better than sleep length did, and adding sleep length on top of regularity didn’t improve the prediction.
What about how long you sleep?
It counts too, at the ends. A 2010 review of 16 studies, with 1,382,999 people followed for 4 to 25 years, found that short sleepers had a 12% higher death rate than people sleeping a normal amount. Long sleepers had a higher death rate as well. Neither end is where you want to be, and the middle covers a range of hours.
12%higher death rate for short sleepers, in 1.38 million people
20 to 48%lower death rate for the more regular sleepers, against the least regular fifth
Don’t regular sleepers just have regular lives?
The sceptic says
People with steady sleep have steady jobs, steady health and fewer kids waking them up. Irregular sleep could be a sign of a harder life rather than its cause.
The reply
Part of it may be. The researchers allowed for income, lifestyle and health, and the link held, but this is still a study that followed people rather than changing their sleep. Nobody has run a trial of a steady wake time with death as the result. A steady wake time is free, and in this study regularity had the clearest link to a longer life, so we’d do it anyway.
2 sources
- Windred DP and others, Sleep regularity is a stronger predictor of mortality risk than sleep duration: A prospective cohort study, Sleep, 2024pubmed.ncbi.nlm.nih.gov
- Cappuccio FP and others, Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies, Sleep, 2010pmc.ncbi.nlm.nih.gov
