Sleep Regularity Predicts Mortality Better Than Sleep Hours
In this article
For decades the sleep question has been about hours. Seven? Eight? Is six enough? Our own sleep and longevity guide covers that research, and it holds up. But a 2024 study in the journal SLEEP asked a different question, and its answer is worth knowing if you care about the long run. It found that when you sleep, night after night, predicted mortality risk more strongly than how long you sleep.
This post walks through what that study measured, what it found, what a second analysis added, and where the evidence stops. The short version is encouraging. Keeping a steadier schedule is one of the more doable sleep changes there is.
Lower risk of death from any cause for the most regular sleepers compared with the least regular fifth, after full adjustment for lifestyle and social factors
Windred et al., SLEEP, 2024, 60,977 UK Biobank participants measured by wrist accelerometer
The short answer
- In 60,977 UK adults whose sleep was measured by a wrist device for a week, the most regular sleepers had a lower risk of death over the following years than the least regular (hazard ratio 0.70 after full adjustment, 0.52 with only age, sex and ethnicity adjusted).
- Regularity was the stronger predictor. Its best-to-worst gap was larger than the gap for sleep duration, and adding duration to a regularity model did not meaningfully improve the prediction.
- A second team analyzing 88,975 people from the same UK Biobank cohort found the extra risk sat mostly with the most irregular sleepers. Above a middling level of regularity, the benefit leveled off.
- Both studies are observational. They show an association, not proof that fixing your schedule will add time to your life.
- The practical target from the study itself is modest. The most regular fifth fell asleep and woke within roughly 1-hour windows on most days. The least regular fifth were closer to 3-hour windows.
What "sleep regularity" actually means
Sleep regularity is the day-to-day consistency of when you are asleep and when you are awake. The study used a score called the Sleep Regularity Index (SRI). It compares every moment of one day with the same moment 24 hours later and asks how often your state (asleep or awake) matched. A score of 100 means a perfectly repeating pattern. A score of zero means a random one.
That design matters. The SRI is not just "same bedtime." It also drops when sleep is broken up, when you nap at different times, or when your wake time drifts. Someone who sleeps a solid 7.5 hours but goes to bed at 10pm on Tuesday and 1am on Saturday will score lower than someone whose nights look alike.
What the study did
The researchers (Windred and colleagues, published in SLEEP in 2024) used the UK Biobank, a long-running study of about half a million British adults. Between 2013 and 2016, a subset wore an accelerometer on their dominant wrist for 7 days under normal living conditions. From more than 10 million hours of that data, the team calculated regularity scores for 60,977 participants, average age 62.8, 55% women.
They then followed deaths recorded in national health registers. Over a mean follow-up of about 6.3 years (up to 7.8 years), 1,859 participants died. The team split everyone into fifths by regularity and, separately, fifths by sleep duration, and compared death rates across the groups.
Two levels of adjustment ran side by side. The "minimal" model accounted for age, sex and ethnicity. The "full" model also accounted for physical activity, employment, income, neighborhood deprivation, social life, smoking, urban or rural home, shift work, and medication for blood pressure or cholesterol.
What it found
More regular sleep tracked with lower risk, step by step. The relationship was monotonic. Each more regular fifth had lower all-cause mortality risk than the least regular fifth, and the top fifth had the lowest.
| Most regular fifth vs least regular fifth | Hazard ratio |
|---|---|
| Minimal adjustment (age, sex, ethnicity) | 0.52 |
| Full adjustment (plus lifestyle and social factors) | 0.70 |
The authors are candid about which number to trust. Some of the factors in the full model, like smoking, could be partly caused by irregular sleep patterns rather than just confounding them. So they write that the true strength of the relationship "likely lies between" the two models. That is why we quote the more conservative 30% in the headline above and not the larger figure.
Regularity beat duration as a predictor. For sleep duration, the best-to-worst comparison was smaller (hazard ratio 0.69 with minimal adjustment, 0.76 with full). Formal model comparisons favored regularity. And when both measures were put in the same model, duration did not explain significant additional mortality risk beyond what regularity already captured. The authors put their conclusion in the title. Regularity is a stronger predictor of mortality risk than duration.
It showed up across causes of death. Of the 1,859 deaths, 1,092 were from cancer and 377 from cardiometabolic causes (heart disease, stroke, diabetes and related conditions). Regular sleep was associated with lower risk of cancer death, cardiometabolic death, and death from other causes. Two details keep this honest.
- Sleep duration was not a significant predictor of cancer death in this study. Regularity was.
- For cardiometabolic death, both mattered. In the fully adjusted model that included both measures, regularity was no longer a significant predictor of cardiometabolic death.
Duration still counts. This is not a finding that hours don't matter. The authors write that their results "confirm an important role for sleep duration in predicting mortality," and the shortest-sleeping fifth had the highest risk among the duration groups. The longest-sleeping fifth in this sample averaged a little over 8 hours, so the study could not test very long sleep.
A second look at the same cohort
A separate team (Cribb and colleagues, eLife, 2023) analyzed regularity and mortality in 88,975 UK Biobank participants, with 3,010 deaths over a median of 7.1 years. They used their own version of the SRI and modeled it as a smooth curve rather than fifths.
They also found irregular sleep was associated with higher mortality. The shape is the useful part. Risk was highest among the most irregular sleepers and fell as regularity approached the middle of the pack, "after which the decrease began to plateau." Compared with a person at the median, someone at the most irregular end (5th percentile) had a hazard ratio of 1.53. Someone at the most regular end (95th percentile) had 0.90, a confidence interval that reached 1.00.
In plain terms, most of the signal in this analysis came from getting out of the most chaotic end, not from chasing a perfect score. The association held after adjusting for sleep duration and nighttime wakefulness, and it shrank (but remained) when body weight, blood pressure and physical activity were added.
One caution on reading these two papers together. They draw on the same UK Biobank accelerometer data, so the second is a reanalysis with different methods, not an independent replication in a new population.
What this does not prove
Both research teams say it directly. Windred and colleagues acknowledge "the correlational nature of our findings," and write that sleep regularity "may be both a cause and marker" of mortality risk. Cribb and colleagues write that because the study was observational, "we are, therefore, unable to establish cause and effect," and that it is "not clear" whether interventions that improve regularity extend lifespan.
A few specific limits are worth naming.
- One week of data. Each person's regularity came from a single 7-day recording, a snapshot of a habit that can change over years.
- An older, mostly white British sample. Participants averaged about 63, and 97.2% identified as white. The authors themselves call for replication in other cohorts and cultures.
- Illness can cause irregular sleep. Someone becoming unwell may sleep more erratically before a diagnosis. Both teams tried to account for this with adjustments for existing disease, and the association held, but observational data cannot fully rule it out.
- No years-of-life figure exists. Neither study reports sleep regularity in years of life expectancy, so we will not print one. The verified currency here is relative risk.
That is why our habit library grades the regular wake-time habit as Emerging evidence, not Strong evidence. The signal is large and measured by device rather than memory, which is a real strength. It is also two analyses of one cohort. You can see how we draw that line in how we vet longevity recommendations.
What a steadier schedule looks like in practice
The study gives a concrete benchmark. People in the most regular fifth fell asleep and woke up within roughly 1-hour windows on most days of their recording week. People in the least regular fifth were closer to 3-hour windows. The authors suggest aiming for 1-hour windows "may therefore be a feasible strategy," especially for people with one main sleep period per day.
What that means day to day.
- Anchor your wake time first. A fixed wake time, weekends included, is the simplest lever because it is under your control in a way that falling asleep is not. It is one of the habits on our sleep habits page.
- Move gradually. If your weekday and weekend schedules are far apart, shift in 15 to 30 minute steps rather than one big jump.
- Mind the other things that lower the score. The authors note that night awakenings and fragmented sleep also pull regularity down, as do naps at scattered times.
- Shift workers are a special case. If your job moves your schedule, or you have a diagnosed sleep disorder, work out a plan with your clinician rather than forcing a fixed clock.
The authors make one more point worth repeating. Extending sleep duration is hard for many people in practice. Asking people to keep "more similar sleep times between days" may be a more feasible target than devoting a greater share of the day to sleep. If you have struggled to add an extra hour of sleep, a steadier schedule is a different door into the same room.
Where this fits in your Longlevity numbers
Full transparency. The Longlevity assessment asks how many hours you sleep on a typical night, because duration is the sleep factor with the deepest mortality evidence behind it, from meta-analyses covering millions of people. It does not ask about regularity. We think regularity is promising and we keep watching the research, but one cohort is not yet enough for us to put a number on it in your estimate.
So treat a steadier schedule as a strong companion to your sleep hours, not a replacement for them. To see where sleep ranks against your other habits, the calculator takes about two minutes. For how every habit compares in years, see our ranked table.
Is sleep your biggest lever? Find out
The 2-minute assessment estimates your biological age and shows the one habit worth the most years to you. Free, no signup.
Take the Assessment →Frequently asked questions
Does going to bed at the same time every night really matter?
It appears to. In a 2024 study of 60,977 UK adults whose sleep was tracked by a wrist device, the most regular sleepers had a lower risk of death over about six years than the least regular, and regularity was a stronger predictor than sleep duration. The study is observational, so it shows an association rather than proof, but keeping a consistent schedule is low risk and doable for most people.
Is sleep regularity more important than getting 7 to 8 hours?
In that UK Biobank study, regularity was the stronger predictor of mortality, but the authors are clear that duration still plays an important role. Short sleep remains linked to higher mortality across large meta-analyses. The honest answer is to aim for both, enough hours and a steady schedule, rather than trading one for the other.
How consistent does my sleep schedule need to be?
You do not need perfection. In the 2024 study, the most regular fifth fell asleep and woke within roughly 1-hour windows on most days, while the least regular fifth were closer to 3-hour windows. A separate analysis of the same cohort found most of the extra risk sat with the most irregular sleepers, with the benefit leveling off around average regularity. Moving out of the chaotic end appears to matter most.
Do weekend lie-ins count against sleep regularity?
They can. The Sleep Regularity Index compares each moment with the same moment 24 hours later, so a wake time that jumps by several hours on Saturday lowers the score, as do scattered naps and broken sleep. A fixed wake time across all seven days is the simplest way to keep your schedule steady. If you need to shift, move in 15 to 30 minute steps.
What is a good Sleep Regularity Index score?
The SRI runs from zero (random sleep and wake times) to 100 (a perfectly repeating pattern). In the 60,977-person study the median was 81, and the most regular fifth scored above about 87.
Related reading
Citations
- Windred DP, Burns AC, Lane JM, et al. Sleep regularity is a stronger predictor of mortality risk than sleep duration: A prospective cohort study. SLEEP. 2024;47(1):zsad253. doi:10.1093/sleep/zsad253.
- Cribb L, Sha R, Yiallourou S, et al. Sleep regularity and mortality: a prospective analysis in the UK Biobank. eLife. 2023;12:RP88359. doi:10.7554/eLife.88359.