Sleep Regularity Outperforms Sleep Duration: Evidence from 60,977 People

What the study found: 60,977 people, 10 million hours of data

Windred, Burns, Lane et al. (Sleep, 2023, DOI 10.1093/sleep/zsad253) conducted a prospective cohort study in a sample of 60,977 UK Biobank participants (mean age 62.8 ± 7.8 years, 55% women). Each participant wore a wrist accelerometer (Axivity AX3) for 7 consecutive days; more than 10 million hours of data were accumulated in total. The median follow-up period was 6.30 ± 0.83 years.

During this time, 1,859 all-cause deaths were recorded: 377 cardiometabolic, 1,092 cancer-related, and 390 other.

Key result: participants in the top quintile by Sleep Regularity Index (SRI) had a mortality hazard ratio of 0.70 (95% CI 0.59–0.83, p<0.001) compared to the bottom quintile. This translates to a 30% lower all-cause mortality risk. For cardiometabolic mortality the effect was even stronger: HR=0.62 (a 22–57% risk reduction).

The authors' principal finding: in nested model testing, SRI proved statistically a stronger predictor of mortality than sleep duration.

What is the Sleep Regularity Index and what score is considered good?

The Sleep Regularity Index (SRI) is a metric from 0 to 100. It is calculated as the average consistency of state (sleep or wake) at the same time points exactly 24 hours apart. SRI=100 means an absolutely identical pattern every day. SRI=0 means completely random. The algorithm accounts for naps and fragmented sleep, and works with polyphasic patterns.

The median value in the UK Biobank cohort was 81.0 (interquartile range 73.8–86.3). Dividing into quintiles showed a monotonic decrease in mortality risk as SRI increased: even moving from the bottom quintile to the second reduced risk (HR=0.80, CI 0.69–0.93). The effect is linear — every step toward greater regularity is associated with lower mortality.

Importantly: SRI does not assess whether a person goes to sleep early or late — only the consistency of that time from day to day.

Why irregular sleep is harmful: circadian mechanisms

The central circadian clock in the suprachiasmatic nucleus of the hypothalamus controls the expression of approximately 10–15% of all genes. It synchronizes peripheral tissue clocks — in the pancreas, liver, immune cells, and blood vessels — via neuroendocrine signals. An irregular sleep schedule disrupts this synchronization.

The specific consequences are well documented. The systematic review by Kalkanis et al. (Sleep Medicine Reviews, 2025, DOI 10.1016/j.smrv.2025.102203), covering 59 studies, recorded the following associations of sleep irregularity with moderate-quality evidence:

• Metabolic disturbances: elevated BMI, insulin resistance, hypertension.
• Cardiovascular events: sleep irregularity is linked to higher 10-year cardiovascular disease risk.
• Mental health: symptoms of depression and anxiety correlate significantly with schedule instability.
• Inflammation: elevated white blood cell counts and C-reactive protein.
• Mortality: in five large cohorts, irregular sleepers had 20–88% higher all-cause mortality independently of sleep duration and quality.

One mechanism is reduced heart rate variability (HRV): sleep schedule irregularity is associated with poorer autonomic regulation, which is itself a marker of cardiovascular risk.

Irregular sleep and dementia: what 2024–2025 data show

A study of 82,391 UK Biobank participants (BMC Public Health, 2025, PMID 39920677, mean follow-up 7.9 years) showed: participants with SRI ≥70 had a 26% lower relative dementia risk (HR=0.74, 95% CI 0.63–0.87) compared to those with SRI below 70. The association was linear, and the protective effect of regularity was especially pronounced in people with insufficient (less than 7 hours) or excessive (8 or more hours) sleep duration — meaning regularity partially compensated for suboptimal duration.

Yiallourou et al. (Neurology, 2024, DOI 10.1212/WNL.0000000000208029) in 88,094 participants confirmed: sleep irregularity is associated with smaller hippocampal volume — a structure critical for memory formation and most vulnerable in Alzheimer's disease.

Regularity vs. duration: what to prioritize?

The data do not say that duration is unimportant — recommendations of 7–9 hours for adults remain in effect. But they suggest the following: a person who sleeps a consistent 6.5–7 hours every night at the same time may have a better health prognosis than someone who sleeps 5–6 hours on weekdays and tries to "catch up" with 9–10 hours on weekends. "Social jet lag" — a shift in schedule of 1.5–2 hours on weekends compared to weekdays — is well studied as a factor of metabolic risk.

This study does not prove causality: people with chronic illness may also have more irregular sleep as a consequence of their condition (reverse causality). The authors Windred et al. conducted a sensitivity analysis excluding early deaths and obtained similar results, which reduces but does not eliminate this limitation.