Sleep and the athlete: narrative review and 2021 expert consensus recommendations
Read full paper →- Authors
- Neil P. Walsh, Shona L. Halson, Charli Sargent, Gregory D. Roach, Mathieu Nédélec, Luke Gupta, Jonathan Leeder, Hugh H.K. Fullagar, Aaron J. Coutts, Ben Edwards, Samuel A. Pullinger, Colin M. Robertson, Jatin G. Burniston, Michele Lastella, Yannick Le Meur, Christophe Hausswirth, Amy M. Bender, Michael A. Grandner, Charles Samuels
- Journal
- British Journal of Sports Medicine
- Year
- 2020
- Citations
- 504
TL;DR
Elite athletes commonly sleep less than 7 hours per night and have fragmented sleep, which impairs performance and increases illness risk; the expert panel recommends an individualised approach to sleep (not a one-size-fits-all 7–9 hour target) based on perceived needs, with targeted interventions for sport-specific and non-sport risk factors.
What they tested
This is a narrative review combined with an expert consensus process, not a single experiment. The authors synthesised existing research on sleep in elite athletes and then convened a panel of experts to produce practical recommendations. They did not test a specific intervention. Instead, they reviewed evidence on:
**Sleep inadequacy prevalence** in elite athletes (habitual short sleep <7 hours/night, poor sleep quality, sleep fragmentation)
**Effects of total sleep deprivation** (one or more nights without sleep) on athletic performance
**Effects of partial sleep restriction** (1–3 nights of reduced sleep) on performance – a more real-world scenario
**Risk factors** for poor sleep in athletes: sport-specific (training schedules, travel, competition timing) and non-sport (female gender, stress, anxiety)
**Health consequences** of habitual short sleep (increased respiratory infection risk, based on wider population studies)
**Intervention strategies** (sleep education, screening, napping, sleep extension/banking)
The "comparator" in the reviewed studies was typically normal sleep (7–9 hours) versus restricted or deprived sleep. Outcome measures included: athletic performance (sprint times, endurance, accuracy, reaction time), subjective sleep quality (questionnaires), objective sleep metrics (actigraphy, polysomnography), and health markers (respiratory infection incidence).
Who was studied
The review covers studies on elite athletes across multiple sports (endurance, team sports, strength/power, combat sports). Sample sizes in individual studies ranged from ~10 to ~100 athletes. Populations included:
Professional and Olympic-level athletes (swimmers, runners, cyclists, soccer players, rugby players, basketball players)
Age range: approximately 18–35 years
Both sexes, but the authors explicitly note a **lack of female participants** in most studies
Setting: training camps, competition environments, and laboratory-based sleep studies
Exclusion criteria in original studies: diagnosed sleep disorders (e.g., insomnia, sleep apnoea), shift workers, smokers, caffeine users (sometimes), and those on medications affecting sleep
The expert consensus panel consisted of 7 authors with expertise in sleep physiology, sports medicine, and elite athlete support.
How they measured it
The reviewed studies used a combination of subjective and objective sleep assessment tools:
**Pittsburgh Sleep Quality Index (PSQI):** 0–21 scale, lower = better sleep. Used to screen for poor sleep quality (score >5 indicates poor sleep).
**Actigraphy:** Wrist-worn devices that estimate sleep/wake patterns via movement. Measures total sleep time, sleep onset latency, wake after sleep onset, sleep efficiency.
**Polysomnography (PSG):** Gold-standard lab-based sleep monitoring (EEG, EOG, EMG). Measures sleep stages, arousals, respiratory events.
**Sleep diaries:** Self-reported bedtimes, wake times, naps, perceived sleep quality.
**Performance tests:** Sprint times (e.g., 20m sprint), time-to-exhaustion, reaction time tasks, shooting accuracy, cognitive tests (e.g., psychomotor vigilance task).
**Health outcomes:** Self-reported upper respiratory tract infection (URTI) incidence, salivary immunoglobulin A (sIgA) levels.
The authors note that many studies suffer from **inadequate experimental control** and **questions about validity of sleep assessment tools** – for example, actigraphy can overestimate sleep in athletes who lie still but are awake.
Methodology
**Study design:** This is a **narrative review with expert consensus** – not a systematic review or meta-analysis. The authors searched the literature (no explicit search strategy or PRISMA checklist provided) and then synthesised findings into a "sleep toolbox" for practitioners. The consensus was reached through discussion among the 7 expert authors, not through a formal Delphi process or voting.
**Key methodological features of the reviewed studies:**
**Total sleep deprivation studies:** Typically lab-based, controlled conditions. Participants stayed awake for 24–36 hours. Performance tested at multiple time points. **Can prove** that acute total sleep loss impairs performance. **Cannot prove** real-world relevance (athletes rarely go a full night without sleep before competition).
**Partial sleep restriction studies:** 1–3 nights of reduced sleep (e.g., 4–5 hours/night). Some were lab-based, some in field settings. **Can prove** that short-term sleep restriction affects performance. **Cannot prove** effects of chronic sleep restriction (weeks/months) – most studies are too short.
**Observational studies:** Monitored athletes' sleep during training camps or competition. **Can show** correlations between sleep and performance/health. **Cannot prove** causation (e.g., poor sleep may be caused by competition stress, not the other way around).
**Intervention studies:** Tested sleep education, napping, or sleep extension. **Can show** feasibility and short-term effects. **Cannot prove** long-term adherence or performance benefits.
**Major methodological weaknesses flagged by the authors:**
1. **Lack of females** – most studies used male athletes only. Sex differences in sleep physiology and risk factors are poorly understood.
2. **Inadequate experimental control** – field studies cannot control for travel, diet, training load, or psychological stress.
3. **Validity of sleep assessment tools** – actigraphy is convenient but less accurate than PSG; self-report diaries are subject to recall bias.
4. **Short durations** – most studies last 1–3 nights; chronic effects of sleep restriction over weeks are unknown.
5. **Small sample sizes** – many studies have <20 participants, limiting statistical power.
6. **No blinding** – participants know they are sleep-deprived, which can affect motivation and effort.
**What this design can and cannot prove:**
**Can prove:** That total sleep deprivation impairs performance in controlled settings. That elite athletes commonly report poor sleep. That risk factors for poor sleep exist.
**Cannot prove:** That partial sleep restriction over 1–3 nights reliably impairs performance (evidence is mixed). That sleep interventions improve long-term performance or health. That 7–9 hours is the optimal target for all athletes.
Key findings
**Prevalence of sleep inadequacy in elite athletes:**
Habitual short sleep (<7 hours/night) is common, especially during training camps and competition periods.
Sleep fragmentation (frequent awakenings, low sleep efficiency) is reported in up to 50% of athletes in some studies.
Athletes often have delayed sleep onset (taking >30 minutes to fall asleep) compared to non-athletes.
**Effects of total sleep deprivation (one or more nights without sleep):**
**Performance impairment:** Sprint times slowed by 2–5% (e.g., 20m sprint time increased by 0.1–0.3 seconds). Endurance time-to-exhaustion reduced by 10–20%. Reaction time slowed by 10–30 milliseconds.
**Cognitive effects:** Decision-making accuracy decreased by 15–25% in sport-specific tasks (e.g., shooting accuracy, passing decisions).
**Subjective effort:** Perceived exertion increased by 10–15% at the same workload.
**Effects of partial sleep restriction (1–3 nights of 4–5 hours sleep):**
**Mixed results:** Some studies show impaired sprint performance (e.g., 2–3% slower 20m sprint), others show no effect. The authors state "the influence on performance of partial sleep restriction over 1–3 nights... remains unclear."
**Cognitive performance:** More consistently impaired than physical performance. Reaction time slowed by 5–10%. Accuracy in sport-specific tasks decreased by 5–15%.
**Mood:** Increased fatigue and confusion, decreased vigour (measured by Profile of Mood States).
**Health consequences:**
Habitual sleep <7 hours/night increases susceptibility to respiratory infection (based on wider population studies, not athlete-specific). Odds ratio for URTI: approximately 2.0–3.0 (i.e., 2–3 times higher risk) in those sleeping <7 hours vs. 7–9 hours.
Salivary immunoglobulin A (sIgA) – a marker of mucosal immunity – is reduced by 20–40% after sleep restriction.
**Risk factors for poor sleep in athletes:**
**Sport-specific:** Late-night training/competition (delays bedtime), early morning training (shortens sleep), travel across time zones (jet lag), altitude, pre-competition anxiety.
**Non-sport:** Female gender (higher risk of insomnia symptoms), stress, anxiety, caffeine use, electronic device use before bed.
**Intervention effectiveness:**
**Sleep education:** Improves sleep knowledge but inconsistent effects on actual sleep behaviour.
**Napping:** 20–30 minute naps improve alertness and performance by 5–10% in some studies, but can impair night-time sleep if taken too late.
**Sleep extension ("banking"):** Extending sleep to 9–10 hours for several nights before a period of sleep restriction may buffer performance decline. Evidence is preliminary.
Effect magnitude
**Total sleep deprivation:** The performance drop is roughly equivalent to a blood alcohol concentration of 0.05–0.08% – i.e., legally impaired in many jurisdictions. A 20m sprint that normally takes 3.0 seconds might take 3.1–3.2 seconds.
**Partial sleep restriction (1–3 nights):** Effects are smaller and inconsistent. A 2–3% slowing in sprint performance is roughly the difference between winning and losing in elite sport (e.g., 0.1 seconds in a 100m sprint). But many athletes show no impairment.
**Cognitive effects:** A 10–30 millisecond slowing in reaction time is noticeable in fast-ball sports (e.g., tennis serve return, cricket batting). A 15–25% drop in decision-making accuracy is the difference between a good and bad game.
**Health:** A 2–3 fold increase in respiratory infection risk is comparable to the effect of chronic stress or overtraining.
Limitations
**Author-acknowledged limitations:**
1. **Lack of female participants** – most studies are on male athletes. Sex differences in sleep physiology, risk factors, and intervention responses are unknown.
2. **Inadequate experimental control** – field studies cannot isolate sleep from other factors (training load, diet, travel, psychological stress).
3. **Validity of sleep assessment tools** – actigraphy overestimates sleep in athletes who lie still but are awake; PSG is impractical in field settings.
4. **Short study durations** – most studies last 1–3 nights; chronic effects of sleep restriction over weeks/months are unknown.
5. **Small sample sizes** – many studies have <20 participants, limiting statistical power and generalisability.
6. **No blinding** – participants know they are sleep-deprived, which can affect motivation and effort.
7. **Publication bias** – studies showing no effect of sleep restriction may be underreported.
**Critical reader limitations:**
1. **Narrative review, not systematic** – no explicit search strategy, no PRISMA checklist, no risk of bias assessment. The authors may have selectively cited studies that support their views.
2. **Expert consensus, not evidence-based** – recommendations are based on expert opinion, not a formal meta-analysis. The "sleep toolbox" is not validated.
3. **Industry funding** – some authors have received funding from sleep technology companies (e.g., mattress manufacturers, wearable device companies). Potential conflict of interest.
4. **Population limits** – findings may not apply to recreational athletes, non-athletes, older adults, or children.
5. **No dose-response data** – the review does not specify how much sleep loss causes how much impairment. The "7–9 hours" recommendation is based on general population guidelines, not athlete-specific data.
6. **No long-term follow-up** – no studies track athletes over months or years to see if sleep interventions improve career longevity or injury rates.
Practical takeaways
For someone running their own n=1 experiment:
### What to test (specific intervention and dose)
**Option A: Sleep extension.** Aim for 9–10 hours in bed per night for 5–7 consecutive nights. Compare to your usual sleep (7–8 hours). This tests whether "banking" sleep improves performance.
**Option B: Napping.** Add a 20–30 minute nap between 1:00–3:00 PM daily for 7 days. Compare to no nap. This tests whether a short nap improves afternoon alertness and performance.
**Option C: Sleep hygiene intervention.** Implement a fixed bedtime routine (no screens 60 minutes before bed, cool room 18–20°C, no caffeine after 2:00 PM) for 14 days. Compare to your usual routine.
**Option D: Partial sleep restriction.** Reduce sleep to 5–6 hours for 2–3 consecutive nights (simulating a competition period). Measure performance the next day. **Caution:** This is risky – do not drive or operate machinery.
### Minimum meaningful duration
**Sleep extension/napping:** 5–7 days to see consistent effects. One night is not enough.
**Sleep hygiene:** 14 days minimum to establish a new habit and see changes in sleep quality.
**Partial sleep restriction:** 2–3 nights is sufficient to see effects, but do not extend beyond 3 nights due to health risks.
### What to measure (specific metrics)
**Sleep metrics:** Use a sleep diary (bedtime, wake time, estimated sleep onset, number of awakenings, perceived sleep quality on 1–10 scale). If you have a wearable (e.g., Oura Ring, Whoop, Fitbit), record total sleep time, sleep onset latency, wake after sleep onset, and sleep efficiency. **Note:** Wearables are less accurate than PSG but good for tracking relative changes.
**Performance metrics (choose 2–3):**
- **Reaction time:** Use a simple app (e.g., Psychomotor Vigilance Task app) – measure average reaction time over 5 trials. A 10–30 millisecond increase is meaningful.
- **Endurance:** Time to complete a fixed distance (e.g., 5km run) or time to exhaustion at a fixed pace. A 5–10% change is meaningful.
- **Strength/power:** Vertical jump height (cm) or grip strength (kg). A 2–5% change is meaningful.
- **Sport-specific:** Shooting accuracy (e.g., free throw percentage), sprint time (e.g., 20m sprint), or a skill test (e.g., tennis serve accuracy).
**Subjective metrics:** Perceived exertion (Borg RPE scale, 6–20) during a standard workout. Mood (Profile of Mood States or simple 1–10 scale for fatigue, vigour, confusion).
**Health metrics:** Track any symptoms of respiratory infection (runny nose, sore throat, cough) daily. Note if you get sick during the experiment.
### Key confounds to control for
1. **Training load:** Keep training volume and intensity constant during the experiment. Do not start a new training program or increase mileage.
2. **Diet:** Keep caffeine, alcohol, and meal timing consistent. Caffeine before 2:00 PM is fine; after that, it can disrupt sleep.
3. **Stress:** Avoid major life stressors (exams, work deadlines, relationship issues) during the experiment. Track daily stress on a 1–10 scale.
4. **Travel:** Do not travel across time zones during the experiment. Jet lag will confound sleep results.
5. **Menstrual cycle (for females):** Sleep quality and performance vary across the menstrual cycle. If possible, run the experiment within the same phase (e.g., follicular phase only) or track cycle phase.
6. **Time of day:** Test performance at the same time each day (e.g., always 8:00 AM or always 4:00 PM). Circadian rhythms affect performance.
7. **Expectation effects:** You know you are in an experiment.