Sleep regularity and longevity: why when you sleep beats how long

Here’s a number that reorganised how I think about sleep: in a study of roughly 60,000 people, when you sleep predicted your risk of dying better than how long you slept. Same hours, different consistency — and the consistency was the bit that mattered more.

I’d spent years quietly optimising for duration. Chasing the magic 7-to-8. Feeling smug on the nights I hit it, mildly tragic on the nights I didn’t. And it turns out I was tracking the wrong axis — or at least, not the most important one. Apparently the body cares quite a lot about rhythm, and I’d been treating my sleep schedule like a suggestion.

This is the bit nobody puts on the sleep-hygiene posters. Let me explain.

Duration versus regularity — two different things

Sleep duration is obvious: hours asleep per night. Sleep regularity is how consistent your sleep and wake times are from one day to the next. They sound similar. They’re mostly independent.

You can sleep a textbook 7.5 hours every single night and still have terrible regularity — 11pm to 6:30am on Monday, 1am to 8:30am on Tuesday, 2:30am to 10am on Saturday. Same duration, wildly different timing. Your average looks fine. Your body clock is in a state of low-grade chaos.

Researchers needed a way to put a number on this, so in 2017 a group at Brigham and Women’s Hospital and Monash introduced the Sleep Regularity Index (SRI) in Scientific Reports (Phillips et al.). It works by asking, for every minute across two consecutive days, what’s the probability you were in the same state — asleep or awake — at the same clock time? Perfect consistency scores 100. Random scores around 0. Most adults land somewhere in the 70s and 80s — in the UK Biobank accelerometer data the median was about 81.

The elegant thing about the SRI is that it’s blind to how much you sleep. It only cares whether today looks like yesterday. Which means it isolates a variable that duration-based research had been quietly averaging away for decades.

The study that changed the conversation

The headline result came in 2024. Windred and colleagues published an analysis in Sleep using wrist accelerometer data from about 60,000 participants in the UK Biobank — properly measured movement data, not self-reported guesses about bedtime. They calculated each person’s SRI and followed mortality over roughly seven to eight years.

The finding: sleep regularity was a stronger predictor of all-cause mortality than sleep duration. It also outperformed duration for cardiovascular and cancer mortality specifically. The most regular sleepers had something in the order of a 23–48% lower risk of death than the most irregular ones — and crucially, that held after adjusting for how long people slept. Regularity wasn’t just a proxy for getting enough hours. It was doing its own work.

I want to be careful here, because this is the kind of stat that gets flattened into “duration doesn’t matter” on the internet, and that’s not what it says. Duration still predicts mortality. Sleeping five hours a night on an immaculate schedule is not a longevity hack. But if you’ve already sorted your hours and you’re looking for the next lever, regularity is a bigger one than most people realise — and far less discussed than the usual how-many-hours-do-I-need question.

Why the body cares about timing

Bit nerdy, but stay with me, because the mechanism is genuinely the interesting part.

Almost every cell in your body runs a roughly 24-hour clock. These are coordinated by a master pacemaker in the hypothalamus — the suprachiasmatic nucleus — which takes its main cue from light hitting your eyes and then synchronises a sprawling network of peripheral clocks in your liver, gut, pancreas, immune cells, blood vessels. When this orchestra is in time, metabolism, hormone release, body temperature, and cellular repair all happen in a sensible sequence.

Irregular sleep is essentially conducting that orchestra erratically. Eat and sleep at inconsistent times and your peripheral clocks start receiving conflicting signals — light says one thing, food timing says another, your activity pattern says a third. This internal desynchrony, sometimes called circadian misalignment, is the leading candidate mechanism for why irregular sleep tracks with worse health even when total sleep looks adequate.

The downstream effects show up in the systems you’d expect. Glucose regulation is tightly time-of-day dependent; your body handles the same meal worse at midnight than at noon. Blood pressure follows a circadian dip that blunts when sleep timing is erratic. And the autonomic nervous system — the thing that governs your heart rate variability — relies on a predictable day/night rhythm to balance its stress and recovery branches. Scramble the schedule and you nudge all of these in the wrong direction at once.

There’s also a stress angle. A chaotic sleep schedule keeps you in a mild, chronic state of “where am I, what time is it, why am I awake,” which is not far from the low-grade activation we talk about in the chronic stress literature. The body likes to know what’s coming. Regularity is, in a sense, a way of telling it.

Social jetlag — and a Munich connection

My favourite term in all of this is social jetlag, and I’m slightly biased because it was coined practically up the road from me. Till Roenneberg’s chronobiology group at LMU Munich introduced it in Chronobiology International in 2006 (Wittmann et al.), and it describes the gap between your biological clock and your social calendar — most visibly, the difference between your weekday and weekend sleep timing.

Here’s the maths that ruined weekend lie-ins for me. If you wake at 6:30am on workdays and 10am at weekends, the midpoint of your sleep has shifted by nearly two hours. Biologically, that’s the equivalent of flying from Munich to somewhere mid-Atlantic every Friday night and flying back every Monday morning. You’d never voluntarily jet-lag yourself twice a week. Most of us do it accidentally, every single week, and call it a nice weekend.

Roenneberg’s group went on to show that larger social jetlag is associated with higher BMI, worse metabolic markers, and increased likelihood of smoking and other risk behaviours. It’s not the sole cause of any of those — correlation, the usual caveats — but the pattern is consistent and the mechanism is plausible. Social jetlag is basically self-inflicted circadian misalignment, dressed up as a Sunday treat.

I still have the occasional lie-in. I’m not a monk. But I think about it differently now — an hour is fine, four hours is a self-imposed transatlantic flight.

What irregularity does to your metabolism and heart

The mortality data is the dramatic headline, but the mechanistic studies underneath it are where the story gets convincing, because they show how irregularity gets under the skin before it ever shows up as a hard outcome.

On the metabolic side, Huang and Redline analysed actigraphy data from the Multi-Ethnic Study of Atherosclerosis (Diabetes Care, 2019) and found that greater night-to-night variability in sleep duration and timing was associated with a markedly higher prevalence of metabolic syndrome — elevated blood sugar, blood pressure, waist circumference, the lot. Each hour of increased variability in sleep timing came with meaningfully higher metabolic risk. Same cohort, different paper: Full and colleagues reported in the Journal of the American Heart Association (2023) that sleep irregularity was associated with subclinical markers of cardiovascular disease, including atherosclerosis, independent of sleep duration.

And the SRI itself has been validated specifically in older adults. Lunsford-Avery and colleagues (Scientific Reports, 2018) showed that a lower regularity index in older people tracked with worse cardiometabolic risk profiles — higher blood pressure, higher fasting glucose, greater obesity risk — which matters because this is exactly the demographic for whom small longevity levers compound the most.

Put together: irregular sleep doesn’t wait until you’re old to do damage. It shows up first as a slightly worse metabolic and cardiovascular picture, the kind of thing that quietly nudges your biological age ahead of your birthday count, long before it becomes anything a doctor would name.

How to actually become a more regular sleeper

Right, so — the practical bit, because this is where most sleep advice falls apart into a list of things nobody does. The good news is that regularity is, in my experience, easier to fix than duration, because it depends on a single decision you make once rather than a battle you fight every night.

Anchor your wake time, not your bedtime. This is the whole game. Pick a wake time you can hold seven days a week — yes, including weekends — and keep it within about an hour. Wake time is the more powerful anchor because morning light is your body clock’s strongest input. Bedtime can drift to follow your natural sleepiness; the wake time holds the rhythm in place. I land on 7am most days, give or take. Weekends included, which my past self would find appalling.

Get light early, dim it late. Within an hour of waking, get bright light — ideally outdoors, even on a grey Munich morning, where overcast daylight still blows away anything your ceiling can produce. In the evening, drop the lights and ease off screens. You’re giving the master clock a clear, consistent “day starts now / day ends now” signal so it stops having to guess.

Keep meal timing roughly consistent too. Food is a clock signal for your peripheral organs, especially the liver and gut. Eating your largest meals at randomly scattered times — particularly very late — sends a conflicting cue. You don’t need rigid meal-prep discipline; just avoid a 10pm dinner three nights one week and a 6pm dinner the next.

Treat the weekend lie-in as a small luxury, not a reset. A modest extra hour at the weekend is genuinely fine and won’t show up in any meaningful way. The thing to avoid is the four-hour swing — the Saturday where you “make up for the week” and effectively re-jet-lag yourself.

Don’t chaos-correct with naps and caffeine. Late, long naps and late-afternoon caffeine both push your next sleep later, which feeds the irregularity. A short early-afternoon nap is fine. A 6pm coffee to power through is borrowing against tomorrow’s schedule.

None of this requires getting up at 5am or any of the rise-and-grind nonsense. It’s almost the opposite — it’s about being boring on purpose. Pick a wake time, defend it gently, let everything else follow.

How to measure it

You don’t need a sleep lab. Most wearables that estimate sleep — Apple Watch, Oura, the rest — record the timing of your sleep and wake, which is all the SRI needs. Some apps now surface a consistency or regularity score directly; if yours does, that’s the number to watch.

As with heart rate variability, the single-night value is mostly noise — one late night doesn’t define you. The trend over weeks is the signal. What you’re looking for is the shape of your sleep across a fortnight: are your bedtimes and wake times clustering into a tight band, or are they scattered all over the clock?

We pull sleep timing into Sarvita from Apple Health, so the picture builds passively — you don’t have to log anything. I’m biased, obviously, but the genuinely useful thing isn’t the score on any given morning. It’s noticing, over a month, that your “regular” schedule had a Tuesday-night hole in it you’d never have spotted by feel.

The honest summary

I find this oddly freeing, actually. Duration is a nightly performance — every evening another chance to fall short, with the entire wellness internet ready to tell you that anything under eight hours is slowly killing you. Regularity is a structural decision. You make it roughly once, defend a wake time, and the rhythm largely takes care of itself.

So if you’ve already done the obvious work on your sleep — enough hours, dark room, the screens thing you’re definitely still working on — and you want the next lever, this is a strong one and almost nobody’s pulling it. Same hours, arranged consistently, appears to buy you a meaningfully lower risk of the things none of us want to die of.

Pick a wake time. Keep it at the weekend. Watch what your timing does over a month.

Anyway. The data’s there if you’re curious. No pressure.