Consistent sleep times

What is behind it

The internal clock is set by light, and regular sleep times create regular light patterns. In 75 adolescents who wore actigraphy over two school weeks and two vacation weeks, a higher Sleep Regularity Index was significantly associated with a more regular Light Regularity Index in both periods.

Anyone who gets up at the same time every day receives morning light at the same internal time every day. The zeitgeber becomes stable instead of alternately setting the clock forward and back. How strongly light acts here is shown by the dose-response curve of the melatonin phase: even dim room light achieves half the maximum phase shift of bright light.

What the Sleep Regularity Index measures

All the large cohorts work with a single metric. On a scale from 0 to 100, the Sleep Regularity Index indicates how likely a person is to be in the same state, that is, either asleep or awake, at two time points 24 hours apart. It is averaged over seven days of actigraphy.

This is more than bedtime: the index captures sleep onset time, wake-up time and lying awake at night simultaneously, and the weekend inevitably counts too. A target value cannot be derived from it — in the same UK Biobank, the median is 81.0 or 60, depending on the calculation.

Where the benefit lies

The association is nonlinear. Relative to the median index, the 5th percentile carried a hazard ratio of 1.53 (95 % CI 1.41–1.66), whereas the 95th percentile carried only 0.90 (95 % CI 0.81–1.00). So the big step is from very irregular to average, not from good to perfect.

More sleep does not replace regularity. In the analysis of major cardiovascular events, meeting the sleep duration recommendation offset the risk in the moderately irregular (HR 1.07; 95 % CI 0.96–1.18), but not in the markedly irregular (HR 1.19; 95 % CI 1.06–1.35). The reverse applies just the same: short sleep remains a risk marker in its own right.

What the studies show

Regularity predicts mortality better than duration

Prospective cohort study, 60,977 UK Biobank participants, 62.8 ± 7.8 years, 55.0 % female, median index 81.0. Over a mean of 6.30 years, 1,859 deaths occurred. Result: 20–48 % lower risk of all-cause mortality in the upper four quintiles, 16–39 % for cancer mortality, 22–57 % for cardiometabolic mortality. In nested models with and without sleep duration, regularity proved to be the stronger predictor — as predictive performance, not as a causal effect.

The same association, nonlinear

A second analysis of the same cohort with 88,975 participants, mean age 62 years, median index 60. Over a mean follow-up of 7.1 years, 3,010 deaths occurred. After adjustment, there was a nonlinear association with all-cause mortality: hazard ratio 1.53 (95 % CI 1.41–1.66) for the 5th percentile and 0.90 (95 % CI 0.81–1.00) for the 95th percentile relative to the median.

Major cardiovascular events

Prospective cohort with 72,269 adults aged 40 to 79, seven days of wrist accelerometry, 8 years of follow-up, without prior events and without an event in the first year of follow-up. Irregular sleepers with an index below 71.6 had a hazard ratio of 1.26 (95 % CI 1.16–1.37) compared with regular sleepers above 87.3, and the moderately irregular 1.08. The authors argue for including sleep regularity as an independent risk factor in guidelines.

Dementia

Population-based prospective cohort with 82,391 adults aged 43 to 79 from the accelerometry subsample, mean follow-up 7.9 years, 694 new cases of dementia. The group with an index of 70 or above had a hazard ratio of 0.74 (95 % CI 0.63–0.87). Important: the favorable association appeared only in short and long sleepers, not with medium sleep duration.

What the cohorts do not show

There is no randomized study showing that improving sleep regularity improves outcomes. The entire evidence consists of prospective observational cohorts; the literature search found no completed intervention study with hard endpoints, only a study protocol. Reverse causality is therefore not ruled out: incipient diseases make sleep irregular before they are diagnosed.

In addition, there are four practical limitations. The cohorts are older and predominantly of European descent, with age ranges of 40–79 and 43–79 years, and the UK Biobank has a known participation bias toward healthier people. The measurement covers only one week in each case, from which conclusions are drawn about years. The benefit is unevenly distributed: for dementia, the favorable association appeared only in short and long sleepers. And for the common tolerance of plus/minus 30 minutes there is no source — the index cannot be translated into a tolerance in minutes.

How to do it

In all cohorts, what was measured was the Sleep Regularity Index over seven consecutive days of actigraphy, including the weekend. It captures sleep onset time, wake-up time and lying awake at night simultaneously — so sleeping in on Saturday goes directly into the value. The thresholds from the largest analysis of cardiovascular events are: regular above 87.3, moderately irregular between 71.6 and 87.3, irregular below 71.6. These are cohort quantiles, not individual target values. The plausible lever in everyday life is light — a consistent wake-up time stabilizes the zeitgeber, not just the calendar.

Safety

Consistent sleep times are the lowest-risk measure in this topic area: no substance, no dose, no exposure. There are nevertheless two limitations. First, regularity must not turn into sleep restriction. Anyone who keeps a fixed wake-up time but goes to bed later in the evening shortens their sleep — and in the same cohort, short sleep was associated with an increased risk of dementia, with a hazard ratio of 1.19 for the median of short sleepers of 6.5 h compared with the reference point of 7.9 h. Second, a rigid schedule is counterproductive in insomnia if it leads to long periods of lying awake in bed; treatment belongs in professional hands. Anyone who cannot fall asleep for weeks or snores loudly with breathing pauses should have this checked by a physician. Shift workers cannot implement the tip because of their work.

Sources

• Windred et al., Sleep 2024
• Cribb et al., eLife 2023
• Chaput et al., J Epidemiol Community Health 2025
• Bian et al., BMC Public Health 2025
• Potts et al., Nutrients 2025
• Hand et al., Sleep 2023
• Zeitzer et al., J Physiol 2000