Effects of Acute Sleep Loss on Physical Performance: A Systematic and Meta-Analytical Review
Participants
69 publications, 227 outcome measures, n=959 participants (89% male). Sleep loss = ≤6 h in any 24 h; control = >6 h.
Key Findings
Overall: sleep loss lowered exercise performance by a mean 7.56% (95% CI −11.9 to −3.13%, p=0.001; I²=98.1%). The effect was significant in every exercise category: anaerobic power, speed/power endurance, HIIE, strength, endurance, strength-endurance, skill.
The pattern of sleep loss matters: consistent harm only with total deprivation and late restriction (waking earlier than normal). Early restriction (going to bed later) was not consistently negative.
Time awake before the task: for deprivation and late restriction, performance fell ~0.4% for every hour awake before exercise.
Time of day: the abstract says tasks in the afternoon/evening were consistently impaired and morning tasks largely unaffected; the full text's pooled analysis shows −8.31% for PM tasks (p = 0.001) and −5.42% for AM tasks (p = 0.013), a negative effect at both times, larger in PM; for endurance tasks AM −6.50% (p = 0.010) and PM −3.56% (p < 0.001), which the authors read as an effect irrespective of time of day (corrected 2026-10-12, full text).
Practical Application
Across 69 studies (n=959, 89% men), ≤6 h of sleep cut performance by ~7.6% on average in every exercise category; the harm came from total deprivation and from waking early, grew ~0.4% per extra hour awake, and hit afternoon/evening sessions harder (pooled PM −8.31% vs AM −5.42%; the abstract calls morning tasks "largely unaffected", Table 3 does not); endurance tasks were impaired at either time (AM −6.50%, PM −3.56%; not formally compared — the authors read it as irrespective of time of day) — after a short night, train early rather than late for non-endurance work, avoid cutting sleep by getting up early, and do not expect a morning endurance session to be spared (corrected 2026-10-12, full text).
If a short night is unavoidable, the authors' own advice: avoid deprivation and late restriction, and prioritise morning exercise.
A key session scheduled for the evening after a short night is the one most likely to suffer in the pooled data; an endurance session was impaired in the morning too (AM −6.50%, PM −3.56%; not formally compared), so moving it to the next day is the option most consistent with these data (corrected 2026-10-12, full text). Inference, not tested: which of the two is better was not studied.
Limitations
Acute sleep loss only (≤6 h, single nights). Says nothing about chronic partial restriction, which is the pattern most athletes actually live with.
Pooled across heterogeneous protocols, populations and tests (I²=98%); each stratified estimate rests on fewer studies than the headline.
Laboratory performance tests, mostly male — transfer to a race after a bad night is inference.
Sleep was manipulated, not observed in the field, and this literature cannot blind an athlete to being sleep-deprived; expectancy is uncontrolled.
No intervention arm: this establishes the cost of losing sleep, not the benefit of any countermeasure such as napping, caffeine or sleep extension.
How this entry was checked and changed
Why this level
regraded 2026-09-27 from A: meta-analysis with extreme heterogeneity (I²=98%)
Content correction 2026-09-27
"Endurance reduced independent of time of day" -> the abstract says afternoon/evening tasks were consistently impaired and morning tasks largely unaffected (Table 3 of the full text: AM −5.42%, p = 0.013, PM −8.31%; endurance tasks were impaired at either time (AM −6.50%, PM −3.56%; not formally compared — the authors read it as irrespective of time of day)) (corrected 2026-10-12, full text).
"Partial and complete deprivation alike" -> harm was consistent only for deprivation and late restriction (early waking), not early restriction; ~0.4% loss per hour awake added.
"RPE +5-8% at identical power", "anaerobic power most resistant" and "8-12% reduction in target workouts" removed -> not in the abstract; the pooled mean is −7.56% and every category, anaerobic power included, was significantly affected.
Plan-week bullets ("Weeks 3-4 and 6-10"), "8+ hours for 3 nights" and "sleep banking" removed -> not from this paper. Sport "cycling" -> general; Level A -> A- (I²=98%).
Cross-references to duplicate stubs repointed to their canonical entries (REF-012->REF-370, REF-132->REF-361); the stubs were emptied on 2026-09-27.
Full text (checked 2026-10-12)
Pooled change by category: anaerobic power −6.26%, speed/power endurance −2.90%, HIIE −6.15%, strength −2.85%, endurance −5.55%, strength-endurance −9.85% and skill −20.9% (full text, Table 2).
By protocol: deprivation −5.25% and late restriction −7.39% (both p < 0.001), early restriction −5.85% (95% CI −13.4 to 1.66, p = 0.125) (full text, Table 2).
For endurance, deprivation gave −6.75% and restriction −3.27% (full text, Table 2).
Correction: the entry says morning tasks were "largely unaffected"; the full text's pooled analysis shows −5.42% for AM tasks (p = 0.013) vs −8.31% for PM tasks (p = 0.001), stating a "consistent negative influence" in both AM and PM with the larger effect in PM (full text, Table 3).
Correction: for endurance specifically, the AM effect was −6.50% (p = 0.010) and the PM effect −3.56% (p < 0.001), so the entry's "morning sessions were largely spared" does not hold for endurance tasks (full text, Table 3).
Late restriction (waking early) cost −9.67% for PM tasks vs −2.48% for AM tasks (full text, Table 3).
Per hour awake, the meta-regression slope was −0.55% for late restriction and −0.36% for restriction (both p < 0.001); for deprivation it was −0.30% (p = 0.051), becoming −0.27% (p = 0.015) after removal of one outlier (full text, Table 4).
Only adults (≥ 18 years) with single-night (acute) sleep loss were included, and mean study quality was a Rosendal score of 67 ± 9% (full text).