Sleep regularity

Sleep regularity means going to sleep and waking at the same times from one day to the next. In large studies that tracked people with wrist devices, it predicted death better than how long people slept, and it also tracked with heart disease and dementia. Every one of those studies is observational, though, and the only trial of making sleep more regular so far is tiny, so treat it as a strong warning sign rather than a proven treatment.

Keep the same wake time every day, weekends included, and let bedtime follow. Don't count on enough total hours to make up for an erratic schedule.

What it measures

The usual measure is the Sleep Regularity Index: the chance that you are in the same state — asleep or awake — at any two moments 24 hours apart, worked out from a week of data from a wrist-worn activity tracker.^[1] Someone who sleeps and wakes at the same times every night scores high; someone whose sleep moves around scores low, even if their total hours are the same. Other studies measure the night-to-night swing in sleep length or in the time of falling asleep instead.

The headline study, and why its number is too good

In about 61,000 UK adults with a week of tracker data, the most regular fifth of sleepers had a 48% lower death rate than the least regular fifth, 55% lower death from heart and metabolic disease, and 39% lower death from cancer.^[2] Regularity predicted death better than sleep duration did.

Those are the figures usually quoted, and they are adjusted only for age, sex and ethnicity. The paper's fuller model also accounts for physical activity, employment, income, deprivation, social contact, smoking, where people lived, shift work, and blood-pressure or cholesterol medication. That shrinks each figure by about a third: about 30% lower death from any cause, 38% lower from heart and metabolic disease and 24% lower from cancer. The signal is still clear, and the authors place the true effect somewhere between the two models. Two senior authors co-direct a circadian-lighting company and have received funding from lighting firms.

Other studies point the same way

Heart attacks and strokes. In about 72,000 adults followed for around eight years, irregular sleepers had 26% more heart attacks, strokes and cardiovascular deaths — and getting enough total sleep did not offset it.^[3]
A dose-response. Across about 86,000 adults, each extra hour of night-to-night swing in sleep length carried about 19% higher cardiovascular risk, independent of inherited risk.^[4]
Blood pressure. Irregular timing of falling asleep alone has been linked to new high blood pressure.^[5]
Dementia. The most irregular sleepers had about 53% higher risk of developing dementia, and smaller grey-matter and hippocampal volumes at the extremes of regularity.^[6] See Dementia prevention.
Mood. Regular sleepers had 38% less new depression and 33% less new anxiety than irregular sleepers.^[7]

A review of 59 studies found 20% to 88% higher mortality in the least regular sleepers across the five cohorts at lowest risk of bias — a range across studies, not a pooled figure, because the measures were too varied to combine.^[8] An expert review concluded that irregular sleep is independently associated with heart and metabolic disease, and likely a stronger predictor than duration.^[9]

What hasn't been shown: that fixing it helps

All of the above is observational: people with erratic sleep may differ in ways the studies couldn't measure, or poor health may disrupt sleep rather than the other way round. The only randomised test so far had 16 adults keep regular hours for twelve nights. Resting heart rate and heart-rate variability moved in the right direction; blood pressure did not change.^[10] The signal is consistent, but the test of whether changing your schedule changes your risk has barely begun.

Weekend lie-ins

Sleeping in at weekends doesn't repair a week of short nights. In a laboratory study of 14 healthy young adults, short weekday sleep with free weekend recovery still left insulin sensitivity 9% to 27% lower, depending on whether the whole body, liver or muscle was measured, along with a later body clock, more after-dinner eating and weight gain.^[11]

Sleeping in isn't harmful in itself, though: across about 73,500 adults followed for eight years, two or more hours of weekend catch-up sleep was linked to neither higher nor lower mortality or heart disease.^[12] The problem is the short weekday nights, not the lie-in.

What to do

Fix your wake time first, seven days a week. Bedtime tends to follow.
Don't trade regularity for hours. An erratic schedule with enough total sleep still carried the higher risk.
Use morning light and an earlier evening wind-down to make the schedule stick; see Circadian rhythms.
Expect a modest gain, not a halving of risk. The fully adjusted figures, and the lack of trials, argue for keeping your expectations moderate.