VōxSōmaHear a free preview

By Ramūnas Deniušis, founder · September 13, 2026

Sleep Regularity vs Sleep Duration: Which Actually Matters More?

Almost every piece of sleep advice you have ever read is about a number of hours. Seven. Eight. "Nine if you can get it." The entire consumer sleep industry — trackers, supplements, mattresses, apps — is built on the premise that the job is to accumulate more sleep.

Over the last decade a different measurement has quietly moved to the front of the research: not how much you sleep, but how consistently you sleep at the same times. And in the largest study that has directly compared the two, consistency was the stronger predictor.

This article is about what that research actually found, what it does not say, and the one practical implication that follows from it.

New here? This piece sits inside our broader guide to sound and sleep, and follows on from how much deep sleep you actually need.

Is sleep regularity more important than sleep duration?

In the largest prospective comparison to date, yes — for predicting mortality risk. Analysing accelerometer data from 60,977 UK Biobank participants, Windred and colleagues found that sleep regularity was a stronger predictor of all-cause mortality than sleep duration. That is a statement about prediction in a population, not a promise about any individual. Duration still matters; regularity appears to matter at least as much.

The detail is worth reading slowly. The researchers calculated a Sleep Regularity Index score from more than 10 million hours of wrist-accelerometer data, then followed participants for up to 7.8 years, during which 1,859 deaths were recorded. Compared with the least regular fifth of participants, the four more regular groups had a 23–48% lower risk of all-cause mortality, a 17–39% lower risk of cancer mortality, and a 27–57% lower risk of cardiometabolic mortality — adjusted for age, sex and ethnicity, and robust to further adjustment for sociodemographic and lifestyle factors (Windred et al., Sleep, 2024).

Crucially, the authors tested this head-to-head. Regularity beat duration both in equivalent models and when duration was added to and removed from the regularity model. Their own framing of the practical takeaway: asking people to keep similar sleep times between days "may ultimately be a more feasible strategy" than asking them to give a larger share of the day to sleep.

What is the Sleep Regularity Index?

The Sleep Regularity Index (SRI) is a single 0–100 score for how repeatable your sleep pattern is. It calculates the percentage probability that you are in the same state — asleep or awake — at any two time-points exactly 24 hours apart, averaged across the whole recording. Sleep and wake at identical times every day and you score 100. Sleep and wake at random and you score 0.

The metric was introduced by Phillips and colleagues in 2017 (Scientific Reports, 7:3216). What makes it useful is what it doesn't require: it never has to decide which sleep episode is your "main" one. A nap, a 3am wake-up, a Saturday lie-in and an all-nighter all show up as exactly what they are — mismatches between one 24-hour cycle and the next.

For scale: in the UK Biobank cohort, the median SRI was 81.0 (interquartile range 73.8–86.3). So "regular" in the real world is not 100. It is around 80, and the difference between the top and bottom of that distribution is mostly weekends.

What does irregular sleep do to the body?

The most-cited evidence is cardiovascular. In the Multi-Ethnic Study of Atherosclerosis, 1,992 adults aged 45–84 with no cardiovascular disease wore an actigraph for seven days and were then followed for roughly five years, over which 111 cardiovascular events occurred. Participants with the most irregular sleep schedules were close to twice as likely to develop cardiovascular disease as those with more regular ones — an association that held after adjusting for factors including sleep apnoea (Huang, Mariani & Redline, J Am Coll Cardiol, 2020;75(9):991–999).

These are observational associations, not demonstrated cause and effect. Irregular sleep travels with shift work, stress, illness and a dozen other things. But the pattern now replicates across cohorts, outcomes and measurement methods, which is more than can be said for most sleep advice.

Why would timing matter if you still get eight hours?

Because your body is not a battery, it is a clock. Nearly every system that matters overnight — core temperature, cortisol, melatonin release, blood pressure, digestion — runs on a roughly 24-hour rhythm that is anchored by regular cues, chiefly light. Move your sleep window by two hours and you have not just shifted sleep; you have shifted the light you see, and therefore the clock that schedules everything else.

Phillips and colleagues found this directly. In their 30-day study of 61 undergraduates, the most irregular sleepers showed melatonin onset (DLMO) more than two and a half hours later than the most regular sleepers, a later peak in daily sleep propensity, and light-exposure rhythms of noticeably lower amplitude. Modelling suggested the circadian differences were primarily driven by those different light patterns rather than the other way round. Regularity also tracked with grades (r = 0.37, p < 0.004).

The everyday version of this is what chronobiologists named social jetlag: the gap between the schedule your body would choose and the one your calendar imposes, most visible as the Friday-to-Sunday shift (Wittmann et al., Chronobiology International, 2006;23(1–2):497–509). Sleeping in on Saturday is, to your circadian system, indistinguishable from flying two time zones west and back every weekend.

Which is easier to change — regularity or duration?

Regularity, and this is the quietly radical part. Duration is largely not under your direct control: you cannot decide to fall asleep, and "get 90 more minutes" usually means lying in the dark failing. Timing is under your control. You choose when to set the alarm and when to start winding down. This is why the UK Biobank authors argue regularity may be the more actionable public-health target.

Sleep duration Sleep regularity
What you control When you go to bed (not when you sleep) Both ends of the window
Feedback speed Same night, and easy to misread Weeks — you feel it as easier sleep onset
Weekend cost A lie-in adds hours A lie-in is the single biggest hit
Failure mode Lying awake trying harder None — you simply keep the schedule
What the 2024 UK Biobank comparison found Weaker predictor of all-cause mortality Stronger predictor

Note what is not in that table: any claim that changing your schedule will change your risk of anything. Nobody has run that trial. What the evidence supports is a target worth aiming at, not an outcome you can buy.

How do you actually make sleep more regular?

Anchor the wake time first, seven days a week, and let bedtime follow. Wake time is the one end of the window you can enforce with an alarm, and it is what sets your light exposure for the day. Keep weekend wake-ups within about an hour of weekday ones. Then give the hour before bed a fixed, repeatable shape so that "time to sleep" becomes a cue rather than a decision.

The practical layer, in order of leverage:

  1. Same wake time, including weekends. Within ±60 minutes. This single habit does most of the work on your SRI.
  2. Morning light soon after waking. Outdoor light is orders of magnitude brighter than indoor lighting and is the strongest anchor your clock has.
  3. A fixed wind-down start, not a fixed bedtime. Deciding when to begin slowing down is achievable; deciding when to fall asleep is not. See our evening wind-down routine.
  4. Make the wind-down identical each night. Repetition is what converts a routine into a cue — the same principle behind how repetition builds familiar patterns.
  5. Deal with the real cause of your late bedtime. For most people it isn't insomnia, it's revenge bedtime procrastination — the hour you take back at the end of the day because it's the only one that was yours.
  6. Stop optimising the metrics. Chasing stage percentages is its own kind of stimulation; sleepmaxxing tends to make the bedroom a place of evaluation rather than rest.

Slow, paced breathing is a reasonable on-ramp for step 3 — see slow breathing before sleep — and understanding what happens to your brain waves as you fall asleep helps explain why abrupt bedtimes rarely work.

Where does a nightly audio ritual fit into this?

It fits at step 4, and nowhere else. The value of a fixed audio track before sleep is not that it induces anything — it is that it is identical every night, which is what makes it usable as a timing cue. A 36-minute track has one useful property here that a playlist does not: it starts and ends at the same clock time whenever you press play.

That is the role VōxSōma is designed for. You record seven affirmations in your own voice; they are woven into a five-layer, 36-minute Evening Wind-Down that runs the same way every night, with the affirmation window in the same place each time. It is a repeatable ritual, not a treatment. It does not measure your sleep, it will not raise any score, and nothing about it has been shown to change the health outcomes described above — those studies are about sleep schedules, not about audio. What it can do is give the start of your wind-down a fixed shape, in your own voice rather than a stranger's. You can hear a preview, read why it was built this way, or see the one-time pricing.

FAQ

Is it better to go to bed late and sleep eight hours, or go to bed on time and sleep six? Neither study answers this directly, and it is a false choice over the long run. The consensus recommendation remains 7+ hours for adults (Watson et al., J Clin Sleep Med, 2015;11(6):591–592). The regularity finding does not replace that; it says that of two people both averaging seven hours, the one whose timing varies less has the better risk profile in these cohorts.

Does a weekend lie-in really undo the week? It is the largest single input to a low regularity score, because the index compares each 24-hour cycle to the next. Two hours later on Saturday and Sunday is a measurable circadian shift. Whether it "undoes" anything is not established — but if you want a higher SRI, this is the first place to look.

What counts as a good Sleep Regularity Index score? There is no validated clinical cut-off. For context, the median in a 60,977-person UK Biobank cohort was 81.0, with the middle half falling between 73.8 and 86.3. Treat those as population context, not a target to hit.

Do naps ruin sleep regularity? They lower the index, because the index counts any asleep/awake mismatch between 24-hour cycles. A nap at the same time every day is far less disruptive to the score than an occasional one — consistency again, not abstinence.

Can a sleep tracker measure my regularity? Reasonably well, yes — better than it measures your sleep stages. Regularity only needs sleep/wake timing, which wearables estimate far more reliably than they estimate N3 or REM. If you are going to look at one number on your app, timing consistency is the more defensible one.

Is this the same as having a "good sleep hygiene" routine? Overlapping but narrower. Sleep hygiene is a long list of dos and don'ts. Regularity is one variable, it is measurable, and it currently has the stronger prospective evidence behind it. If you only change one thing, change the wake time.

Sources

VōxSōma is a personal wellness audio tool — not a medical device, not therapy, and not intended to diagnose, treat, cure, or prevent any condition. Individual experiences vary. If you have a sleep, attention, or mental-health condition, please speak with a qualified clinician.