Sleep advice is almost entirely about duration. Seven to nine hours, the number everyone knows, the number every app scores you against.
Then researchers measured both duration and timing in the same very large group of people, and found that the question nobody asks predicted mortality better than the one everyone does.
Key takeaways
Across 60,977 people and more than 10 million hours of accelerometer data, higher sleep regularity was associated with 20% to 48% lower all-cause mortality than the least regular group (Windred et al. 2024, SLEEP).
Sleep regularity was a stronger predictor of mortality than sleep duration in the same dataset.
The finding held after adjusting for age, sex, ethnicity, and sociodemographic, lifestyle and health factors.
This is an association from observational data, not proof that fixing your bedtime extends your life.
What does sleep regularity actually mean?
The day-to-day consistency of when you sleep and wake, rather than how much you get.
Researchers quantify it with a Sleep Regularity Index, which essentially asks: picking any minute at random, how likely is it that you were in the same state (asleep or awake) at that same minute yesterday? Someone who sleeps 11pm to 7am every night scores very high. Someone who sleeps eight hours every night but starting anywhere between 10pm and 3am scores low, despite identical duration.
That distinction is the whole point. Two people can average exactly seven and a half hours and have completely different regularity.
How much does regularity matter?
More than duration did, in the largest study to look at both with objective data.
The analysis calculated regularity scores from over 10 million hours of accelerometer data across 60,977 UK Biobank participants, then followed mortality outcomes (Windred et al. 2024):
Outcome | Association with higher regularity |
|---|---|
All-cause mortality | 20% to 48% lower risk |
Cancer mortality | 16% to 39% lower risk |
Those comparisons are against the least regular fifth of participants, and the results were adjusted for age, sex, ethnicity, and a range of sociodemographic, lifestyle and health factors.
The headline conclusion was explicit: sleep regularity predicted all-cause mortality more strongly than sleep duration did, confirmed by comparing equivalent models with and without each variable.
Three caveats worth carrying. This is observational, so it shows association rather than causation. Irregular sleep travels with shift work, illness and social circumstances that carry their own risks. And the range given (20% to 48%) is wide because it reflects several different model specifications.
Even with all that, it is a large, objectively measured finding, and it inverts the usual advice.
Why would timing matter so much?
Because nearly every system you own runs on an internal clock that works best when it can predict what is coming.
Body temperature, hormone release, digestion and alertness all follow a daily schedule. When your sleep and wake times are steady, the clock can prepare in advance: melatonin rises on cue, temperature dips on schedule, and falling asleep feels like pushing on an open door.
When the schedule swings, you are effectively giving yourself a small dose of jet lag without going anywhere. Same total hours, rougher ride. Researchers call the weekday-to-weekend version social jet lag.
This is the same mechanism that makes evening light so disruptive: ordinary room light before bed shifted melatonin onset in 99% of people tested. Both are ways of confusing a clock that would rather know what time it is.
Is 7 to 9 hours wrong then?
No, but notice how wide that band is, and what the width implies.
It is a range because people genuinely differ. Where you land inside it is personal, and the honest test is how an ordinary day feels rather than whether you hit a number. A steady seven-hour sleeper often feels better than a chaotic nine-hour one, which is exactly what the regularity data would predict.
Treat 7 to 9 as friendly guardrails rather than a nightly exam. And if you are going to work on one thing, the evidence now points at the timing rather than the total.
What does a good pattern look like?
Tight scatter around whatever time suits you.
The typical time itself is yours to choose. Night owls and early birds both do fine; chronotype is largely not a decision. What matters is the wobble around your own usual pattern.
A tight spread, with most nights landing within about half an hour of each other, means your body can predict sleep and prepare for it.
A wide spread means every night is a small negotiation, and the same hours refresh you less.
Weekends count. The weekday-to-weekend swing is where most people's regularity is lost, and it is the version most studied.
Wake time is the easier lever. It is more under your control than the moment you actually fall asleep, and it anchors the whole cycle.
If you have been chasing catch-up sleep at weekends, the evidence on sleep debt is worth reading alongside this, because the two ideas pull in opposite directions.
Frequently asked questions
Is sleep consistency more important than sleep duration?
For mortality risk, the largest objective study to compare them found regularity was the stronger predictor. Higher regularity was associated with 20% to 48% lower all-cause mortality across 60,977 people, and outperformed duration in equivalent models (Windred et al. 2024). Both matter; the point is that duration has been getting nearly all the attention.
How consistent does my bedtime need to be?
There is no threshold in the research, but the practical target most people use is keeping most nights within about half an hour of each other. Regularity was measured as a continuous score rather than a pass mark, and the association improved across the range.
Does going to bed late matter if I still get eight hours?
Late is fine if it is consistently late. The evidence points at variability rather than lateness itself: eight hours starting at a different time each night scores poorly on regularity even though the duration is ideal.
What is social jet lag?
The mismatch created by keeping one sleep schedule on weekdays and another at weekends. It produces the same kind of circadian misalignment as travelling across time zones, without the travel, and it is where most people lose their regularity.
Can I fix my sleep regularity quickly?
Faster than you can fix most health metrics, because it is a scheduling change rather than a physiological adaptation. Anchoring your wake time is usually the most effective single move, since it is more controllable than sleep onset and it sets the rest of the clock.
Related reading
Two practical companions: building a wind-down routine for the levers that get you to bed at the same time, and waking up in the night for what happens once you are there.
Knit shows patterns, not diagnoses. Talk to a clinician about readings that concern you.
