Skip to content
HabitSeek
Back to results

Get enough sleep regularly

Making regular time for an age-appropriate amount of sleep, with seven to nine hours as adult guidance rather than a mortality-derived optimum.

At a glance

  • Evidence: Habitual sleep near seven hours is associated with lower mortality in cohorts.Habitual sleep duration and all-cause mortality · Limited certainty · Observational association
  • What’s uncertain: Cohorts cannot establish that changing sleep duration changes mortality risk.
  • First step: Protect a regular sleep opportunity and adjust for your own sleep need and daytime functioning.
  • Before you try it: Adapt activity after poor sleep and avoid driving or hazardous tasks when severely sleepy. All precautions
Read the full takeaway

Observed sleep-duration patterns

Seven to nine hours is an expert-panel range for healthy adults, not a range established by mortality trials. Large prospective syntheses consistently show the lowest mortality association near seven hours, with higher observed rates at shorter and longer durations, and that association is substantial evidence about the relationship between habitual sleep and death - the two largest syntheses describe the same data differently rather than disagreeing, because Liu modelled 24-hour sleep duration with 7 hours as the reference while Yin modelled the change in risk per one-hour increment or reduction either side of 7 hours.

Limits of mortality evidence

What that evidence cannot do is show that deliberately changing how long you sleep changes your risk: no trial has tested a sleep duration against death, long sleep may reflect underlying illness, and the association is therefore a reason to take sleep seriously rather than a mortality prescription or proof that seven hours is optimal.

Sleep-extension trial

The randomised evidence points at a different and more practical benefit: a trial in adults with overweight who slept under 6.5 hours found that extending sleep reduced energy intake without any prescribed diet. That trial did not test mortality and does not show that everyone benefits from trying to sleep nine hours.

A woman asleep on cream linen beddingDiscussed byRhonda PatrickView statements ↓
SleepLong-term healthEnergyVaries · Moderate · Free
Reviewed Oct 3, 2026
Save

Scientific evidence

What the research shows

Outcome assessments based on research, including its uncertainties and limitations.

Habitual sleep duration and all-cause mortality

Finding: Benefit supportedCertainty: Limitedstudies found a benefit from doing this

Evidence basis: observational studies (association, not proof of cause)

Habitual sleep near seven hours is consistently associated with lower mortality than very short or very long sleep; this is an observed association, not a tested intervention.

Population: Adults in large prospective cohort studies: 40 cohorts with 2,200,425 participants and 271,507 deaths in the largest synthesis, and generally healthy populations in the dose-response synthesis of prospective cohorts.

Research details

Summary

Across large syntheses of prospective cohorts, habitual sleep duration shows a consistent non-linear relationship with all-cause mortality. In 40 cohorts covering 2,200,425 participants and 271,507 deaths, a J-shaped curve was present with seven hours as the reference: four hours carried a relative risk of 1.05 (95% CI 1.02-1.07), eight hours 1.03 (1.02-1.05), nine hours 1.13 (1.10-1.16), ten hours 1.25 (1.22-1.28) and eleven hours 1.38 (1.33-1.44). A separate dose-response synthesis found U-shaped associations for all four outcomes it examined, with the lowest risk at approximately seven hours per day: for all-cause mortality the pooled relative risk was 1.06 (95% CI 1.04-1.07) per one-hour reduction below seven hours, and 1.13 (95% CI 1.11-1.15) per one-hour increment above it. In Liu's separate analysis restricted to night sleep, the estimate at 9 hours was smaller (relative risk 1.08, 95% CI 1.05-1.11) and 8 hours was null (1.01, 95% CI 0.99-1.02). Short-sleep risk in the 40-cohort analysis was significant in women but not in men.

Why this certainty

Rated Limited. Certainty is not lower, because two large independent syntheses of prospective cohorts agree on a U-shaped association with the lowest risk near seven hours and the sensitivity analyses in one of them survive exclusion of short-follow-up studies and participants with pre-existing disease. It is not Moderate or Strong, because the evidence is entirely observational, because the two syntheses model the same data differently rather than independently confirming each other, because long sleep is plausibly a marker of illness rather than a cause, and because no retrieved study randomises sleep duration against any mortality endpoint. The rating describes confidence in the association; it is not a claim that sleeping seven hours causes longer life.

Limitations

Every estimate is observational, so the association cannot show that changing how long someone sleeps changes their risk of dying. Reverse causation is a serious concern for the long-sleep arm in particular: illness, depression and undiagnosed sleep apnoea all increase time in bed, and one of the two syntheses found that excluding participants with pre-existing disease and studies with short follow-up left the pooled estimate essentially unchanged (RR 1.13, 95% CI 1.11-1.16, I2 = 14.6%), which argues against reverse causation being the whole explanation without excluding it. Sleep duration is self-reported in most contributing cohorts, which misclassifies exposure, and the reference category (seven hours) is a modelling choice rather than a randomised control. No retrieved trial has tested a sleep duration against death.

Effect estimate

Compared with 7 hours, all-cause mortality relative risk 1.05 (95% CI 1.02-1.07) at 4 hours and 1.38 (95% CI 1.33-1.44) at 11 hours, across 40 cohorts and 2,200,425 participants.

Sources for this outcome:

Energy intake and body weight

Finding: Benefit supportedCertainty: Limitedstudies found a benefit from doing this

Evidence basis: randomised trials

A randomised trial found that extending sleep reduced energy intake and produced weight reduction in adults who habitually underslept.

Population: 80 adults aged 21-40 with overweight (body mass index 25.0-29.9) who habitually slept less than 6.5 hours per night, in a single-centre trial conducted in real-life home settings without a prescribed diet.

Research details

Summary

In a single-centre randomised clinical trial, 80 adults aged 21 to 40 with a body mass index of 25.0 to 29.9 and habitual sleep duration under 6.5 hours per night were assigned either to counselling intended to extend bedtime to 8.5 hours or to continue their habitual sleep, with no prescribed diet or exercise. Sleep duration increased by approximately 1.2 hours per night (95% CI 1.0 to 1.4 hours), the sleep-extension group had a significant decrease in energy intake compared with controls (-270 kcal/d; 95% CI -393 to -147 kcal/d), and the change in sleep duration was inversely correlated with the change in energy intake (r = -0.41; 95% CI -0.59 to -0.20). There was no significant treatment effect on total energy expenditure, and the trial reported weight reduction in the sleep-extension group relative to controls.

Why this certainty

Rated Limited. Randomisation and objective measurement of energy expenditure by doubly labelled water are genuine strengths, and the finding addresses a causal question the cohort data cannot. Certainty is held at Limited because there is a single relatively small trial, the population was restricted to habitually short-sleeping adults with overweight, blinding was impossible, and weight was not the tested primary endpoint.

Limitations

This is one trial in 80 participants, and it enrolled only people with overweight who were already habitually short sleepers, so it does not show that extending sleep benefits normal-weight or already-adequate sleepers. The behavioural counselling could not be blinded, and participants knew their allocation. The primary outcome was change in energy intake derived from total energy expenditure plus change in body energy stores rather than a direct measure of what participants ate, and the trial was not designed or powered to test long-term weight change, so the weight finding is secondary.

Effect estimate

Sleep extension of approximately 1.2 hours per night (95% CI 1.0 to 1.4) reduced energy intake by 270 kcal/d (95% CI -393 to -147) in a randomised trial of 80 adults with overweight.

Sources for this outcome:

Exercise performance with normal sleep versus acute sleep loss

Finding: Benefit supportedCertainty: Limitedstudies found a benefit from doing this

Evidence basis: indirect evidence, not a direct test of this exact habit

Maintaining normal sleep supports better exercise performance than experimental acute sleep loss.

Population: 959 participants, predominantly men, in experimental normal-sleep and sleep-loss comparisons.

Research details

Summary

A synthesis of 69 publications found lower exercise performance after acute sleep loss, with a pooled change of -7.56% (95% CI -11.9% to -3.13%). Effects varied by task, loss pattern and time of day. This is not a trial proving that exercise after a poor night reverses lost sleep or that every person should train in the morning.

Why this certainty

Limited because experimental comparisons generally favour normal sleep but are highly heterogeneous and indirect for choosing a recovery routine.

Limitations

Heterogeneity was extremely high (I² 98.1%) and women were underrepresented. Experimental sleep loss and athletic tasks do not define the safest individual workout after a bad night.

Effect estimate

Acute sleep loss versus normal sleep: exercise performance -7.56% (95% CI -11.9% to -3.13%).

Sources for this outcome:

What people say

Who discusses this?

Public statements, with positions specific to each source and topic. These views do not determine the scientific evidence assessment.

What the research shows ↑

Practical steps and precautions

How to try it

  1. Allow for roughly seven to nine hours of sleep as adult expert guidance, then adjust for your own sleep need and daytime functioning.
  2. Work back from a fixed wake time to set your bedtime rather than adding sleep opportunistically, and judge your duration over a typical week rather than a single night.
  3. If you habitually sleep under six and a half hours and have overweight, the randomised evidence for benefit came from extending bedtime towards eight and a half hours in bed, not from occasional long lie-ins.
  4. After a poor night, preserve the next sleep opportunity and adapt exercise to your alertness: comfortable movement or a lighter session may be more appropriate than hard intervals, heavy lifts or technical work. Exercise is not a replacement for sleep, and severe sleepiness makes driving or hazardous activity unsafe.

Precautions

  • Persistent long sleep, especially with unrefreshing sleep, snoring or daytime sleepiness, is a reason to seek assessment for a sleep disorder rather than to set an alarm earlier.
  • Needing far more sleep than usual can be a symptom of another condition, including depression, hypothyroidism or anaemia, and is not treated by willpower.
  • People with insomnia should not simply extend time in bed, which can worsen sleep efficiency; cognitive behavioural therapy for insomnia is the first-line treatment.
  • If you take sleep-affecting medication, or have a cardiovascular or respiratory condition, discuss changes to your sleep schedule with a clinician.

References

Sleep duration and risk of all-cause mortality: A flexible, non-linear, meta-regression of 40 prospective cohort studies

Liu TZ, Xu C, Rota M, Cai H, Zhang C, Shi MJ, Yuan RX, Weng H, Meng XY, Kwong JS, Sun X · Sleep medicine reviews · 2017

Non-linear dose-response meta-regression of prospective cohort studiesVerified from abstract
Open original source

Relationship of Sleep Duration With All-Cause Mortality and Cardiovascular Events: A Systematic Review and Dose-Response Meta-Analysis of Prospective Cohort Studies

Yin J, Jin X, Shan Z, Li S, Huang H, Li P, Peng X, Peng Z, Yu K, Bao W, Yang W, Chen X, Liu L · Journal of the American Heart Association · 2017

Systematic review and dose-response meta-analysis of prospective cohort studies. Quoted values marked "full text" come from the open-access article at PMC5634263; the sensitivity analysis excluding studies with short follow-up and participants with pre-existing disease left the pooled estimate essentially unchanged, which the audit report reads as arguing against reverse causation being the whole explanation.Verified from abstract
Open original source

Effect of Sleep Extension on Objectively Assessed Energy Intake Among Adults With Overweight in Real-life Settings: A Randomized Clinical Trial

Tasali E, Wroblewski K, Kahn E, Kilkus J, Schoeller DA · JAMA internal medicine · 2022

Randomized clinical trialVerified from full text
Open original source

National Sleep Foundation's sleep time duration recommendations: methodology and results summary

Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, Hazen N, Herman J, Katz ES, Kheirandish-Gozal L, Neubauer DN, O'Donnell AE, Ohayon M, Peever J, Rawding R, Sachdeva RC, Setters B, Vitiello MV, Ware JC, Adams Hillard PJ · Sleep health · 2015

Expert-panel consensus recommendations using the RAND/UCLA Appropriateness MethodVerified from abstract
Open original source

Effects of Acute Sleep Loss on Physical Performance: A Systematic and Meta-Analytical Review.

Craven J, McCartney D, Desbrow B, Sabapathy S, Bellinger P, Roberts L, Irwin C. · Sports medicine (Auckland, N.Z.) · 2022

Systematic reviewSystematic review and meta-analysis of experimental sleep-loss studies from 69 publicationsVerified from abstract
Open original source

Community

Discussion

Have you tried this? What helped or made it difficult?

Reader experiences are separate from the scientific evidence. Explore all discussions

No discussions yet.

Be the first to ask a question or share your experience.

Sign in to vote or join the discussion.

Related habits

Explore other habits in a shared category. Each has its own evidence and precautions.

Editorial history and review method

Revision 11 · Approved for publication · Codex (publication approval explicitly authorized by site owner; category migration)

Prepared and checked with AI. Publication approval does not imply independent expert review. Request a correction.

Revision 4, applying the high-risk audit corrections (_report-focus-a.md). Changes: (1) the all-cause-mortality outcome is relabelled from "Benefit supported" to "Insufficient direct evidence" and its certainty from Moderate to Limited, because no trial has randomised a sleep duration and measured death; the association is still stated in full in the headline, summary and effectEstimate. (2) The limitations sentence that said no cited synthesis demonstrates the reverse-causation point was factually wrong about Yin 2017 and has been replaced with Yin's own position and its sensitivity analysis excluding studies with short follow-up and participants with pre-existing disease (RR 1.13, 95% CI 1.11-1.16, I2 = 14.6%); those values were added to the yin-2017-sleep-mortality record from the open-access full text at PMC5634263. (3) The claim that the two syntheses disagree on J-shape versus U-shape is replaced by what actually differs (Liu modelled 24-hour duration against a 7-hour reference; Yin modelled risk per 1-hour change either side of 7 hours) in both the verdict and the limitations. (4) Liu's night-sleep-only model (9 h RR 1.08, 95% CI 1.05-1.11; 8 h RR 1.01, 95% CI 0.99-1.02) was added so the long arm is not represented by the 24-hour estimate alone. (5) The card summary no longer reads as a mortality benefit. A second reviewer must re-check: (a) whether "Insufficient direct evidence" is the label this catalogue wants for an observational card outcome, or whether the association itself should be rated - this is the one editorial judgement in the change and the audit report itself marks it as an editor's decision; (b) that the Limited certainty now attached to an "Insufficient direct evidence" finding is the intended pairing; (c) that no sentence now implies 8-9 hours of sleep is itself risky, given the added night-sleep null at 8 hours; (d) that Liu 2017's full text still cannot be read, so its covariate set and its 6-hour/8-hour equality tests remain unchecked. Directness pass: the all-cause-mortality outcome no longer reads 'Insufficient direct evidence'. The association between habitual sleep duration and mortality is measured directly, consistently and in millions of participants; what has never been tested is the effect of changing sleep duration, which is a limitation on causal inference rather than an absence of evidence. The finding is now Benefit supported on an Observational association basis at Limited certainty, and the verdict, card summary, limitations and certainty rationale all separate the observed association from the untested intervention and state that seven hours is a cohort nadir rather than a proven optimum. A second reviewer must check: (a) that Benefit supported on cohort evidence does not read as an intervention claim anywhere on the card; (b) that the higher risk observed at long sleep is attributed to possible reverse causation rather than asserted as harm. October 3 video follow-up: Added a conservative exercise-after-poor-sleep step and a separate performance assessment from the experimental sleep-loss synthesis. Do not imply exercise replaces sleep or use subgroup timing as a universal prescription. Publication release 2026-10-04: site owner authorized a bounded AI audit and release of this exact revision. See docs/release-audit-2026-10-04.md for source-check findings and limits. No independent health-professional or peer review is claimed. Category migration 2026-10-04: removed the separate goals field; categories: Sleep, Long-term health → Sleep, Long-term health, Energy. Research assessments, sources, guidance, precautions, and evidence review dates are unchanged. This revision awaits publication approval; no new evidence review is claimed. Publication approval 2026-10-04: site owner explicitly accepted all current category-migration revisions for deployment and requested that remaining limitations be retained in docs/content-review-backlog.md. Approval applies to this exact content revision. No new evidence review, independent expert review, or clinical validation is claimed; existing precautions and disclosures remain in effect.