Sleep Strategy

As adventure racers, the two most common questions we ask regarding sleep are “What is the least amount of sleep we need to gain the maximum benefits?” and “Is it better to ‘power-nap’ for 15-minutes, 30-minutes or an hour?”

[This article was written back in 2001. I was recently looking for it and today I found it on an old external hard drive. There is a lot of technical sleep info first an then an ‘In English..’ summary. There are some other articles on sleep also on Read them too.]


Most sleep research – and it’s a huge field – centres on sleep disorders; hormones and sleep; whether exercising helps you to sleep and if in fact you need to sleep more; the effect of the contraceptive pill (and many other drugs) on sleep patterns; snoring etc. With a constant debate going on the function of sleep, as in “Why do we sleep?”, and research aimed at helping people achieve a full night of quality rest, there is nothing on snatching the least number of hours in order to roam valleys and mountains day after day.

So, from the available literature, I’ve compiled an introductory study on sleep, outlining the state and associated stages. I’ve drawn my own conclusions regarding sleep strategy in adventure racing and accordingly take no responsibility for anything you may do after reading this article when wandering around in a sleep deprived state during races.

Mammals, birds, reptiles, amphibians and fish and even some insects have been caught napping, a phenomena present throughout the animal kingdom. Even so, there is no definitive answer as to why we sleep.

Sleep has, for decades, been stereotyped as a state in which the brain is turned off and the body rests, allowing for recovery from the vigour’s of wakefulness. However, in the last decade extensive sleep research has shown that in some phases of the sleep cycle the brain reaches peak activity. A highly active and organised state, sleep is divided into cycles, which comprise distinct stages.

Recording Sleep
Traditionally, electroencephalograph (EEG) rhythms generated during sleep are used to define the sleep stages. Recordings of brain activity (EEG), eye movements (electro-oculogram, EOG) and muscle activity (electromyogram, EMG) can be made simultaneously to make up a polysomnogram.1

Based on these recordings, researchers have classified two distinct sleep states – REM sleep, so named for the characteristic rapid eye movements observed and non-REM sleep, which comprises Stages 1, 2, 3 and 4.1 Used in combination, these measures are fairly accurate in distinguishing sleep from wakefulness and the different sleep stages. Although other measures, such as heart rate and respiration may vary during sleep, they do not sufficiently distinguish stages. There are also instances where stages cannot be clearly defined. Also, stage changes occur gradually, not abruptly, making it difficult to draw sharp dividing lines between stages. In addition, particularly in the case of sleeping disorders, the EEG may indicate, for example, a state of wakefulness while the person shows no awareness of his environment.2

Sleep Stages


A person is judged to be awake when their EEG recording alternates between two major patterns. The first activity pattern is low voltage (~10-30uV) and fast (16-25Hz – cycles per second) and is termed an activation or desynchronised pattern. The second has higher voltage (20-40uV) at 8-12Hz. This is termed alpha activity.

The activation pattern is prominent when subjects are awake and they are scanning the visual environment. Alpha activity is abundant when the subject is relaxed and their eyes are closed.2

 An EEG recording of brain activity in an awake state.2

Stage 1

Stage 1 sleep is a light transitional phase from waking to sleep.3 It is characterised by drowsiness and the ability to make responses if necessary i.e. to a question.1 Alpha activity decreases and the activation pattern is rarely noted. The EEG consists mostly of low voltage, mixed frequency activity at around 3-7Hz. Rapid eye movements (REM’s) are absent though slow rolling eye movements can be observed. The EMG shows low to moderate activity.2

An EEG recording of brain activity in a drowsy state.2

An EEG recording of brain activity in stage 1 NREM sleep.2
Stage 2

The deeper stage 2 is normally considered as the true onset of sleep, or sleep onset latency (SOL). Approximately 50% of the normal night is spent in this stage.1, 3 On the EEG, bursts of distinctive 12-14Hz waves, called ‘sleep spindles’ appear above a background of continuing low voltage, mixed frequency activity. Eye movements are rare and the EMG readings remain moderate to low.2

An EEG recording of brain activity in stage 2 NREM sleep.2
Stage 3
From stage 2, sleep gradually becomes the deeper stage 3 sleep.3 Slow (0.5-2Hz), high amplitude (>75uV) waves, called delta waves appear in the EEG. EOG and EMG recordings continue as before.2

Stage 4

Stages 3 and 4 are collectively refered to as slow wave sleep (SWS).1, 4, 5 There is a quantitative increase in delta waves such that they dominate the EEG tracing.2



An EEG recording of brain activity in stage 4 NREM sleep.2

Made famous by a pop group, REM refers to the rapid eye movements observed in this sleep stage. At the end of the SWS period, the sleeper then moves back through the lighter stages and into REM sleep. This stage is characterised by a low voltage, mixed frequency pattern similar to that observed in stage 1.2, 3 Prominent rapid eye movements appear while the background EMG pattern is virtually absent. Many small muscle twitches may occur.2

An EEG recording of brain activity in REM sleep.2
Sleep Cycles

A sleep cycle comprise an episode of non-REM sleep followed by a period of REM sleep, which alternate cyclically throughout the night. In adults, a night of sleep begins with about 80 minutes of NREM sleep, followed by 10 minutes of REM sleep – a single cycle lasting approximately 90 minutes. In the first 2-3 sleep cycles, SWS (stages 3 & 4) predominate with longer REM periods occuring later in the sleep period. Thus, a person will pass through 4-7 cycles per night, depending on the cycle length and total sleep time.2, 3

Over the course of a night, the sleeper will move through a number of sleep cycles. Stages 3 and 4 dominate the NREM periods in the first part of the night but are completely absent during the later cycles. Towards the end of the night, brief periods of wakefulness may interrupt sleep.2

Generally, a sleeper will go through all the stages in a cycle i.e. from stage 2 to stage 3, to stage 4 and then back up the lighter stages and into REM sleep. But, it is possible that a stage can be shortened or excluded completely. For example: should a sleeper roll-over towards the end of stage 4, sleep may be disturbed such that they move straight into stage 2 and from there into REM.3

Researchers define the depth of sleep as the ‘resistance to being awakened by an external stimulus’. Although it takes a louder sound to awaken someone from stages 3 and 4 than from 1, 2 or REM, sleepers awakened in any of the NREM stages report that they have been in ‘lighter sleep’ than when awakened from REM.2

Thermoregulation and Sleep

Since the beginning of the 90’s there has been quite alot of discussion relating body temperature (thermoregulation), exercise and sleep, creating a “thermogenic hypothesis”.6 Body temperature, a circadian rhythmn,7 declines towards evening, “triggering” sleep onset. In addition, during NREM sleep body temperature is regulated at a lower level than during wakefulness and together with sweating, is inhibited during REM sleep.6 This brain and body cooling mechanism provides the following benefits; lower energy utilisation, brain protection from sustained temperatures of wakefulness, the facilitation of immune defences and the regulation of timing of behavioural activity to the circadian light-dark cycle.3

Exercise elevates body temperature. And, particularly when exercise has been intense or of a long duration, sleep onset latency (SOL) may be decreased, SWS (stages 3 & 4) of sleep is increased and REM sleep is reduced.4, 5, 7

The ‘restorative theory of sleep’ states that sleep in general and SWS in particular may be “related to recovery from fatigue, to reparitive processes and to the build up of energy”.5

Sleep Deprivation
Sleep deprivation is of maximal interest to us, yet there is very little in the literature. I believe research was done a few years back on crew in the “Around the World” yacht race and on aeroplane pilots, but I was unable to get my hands on these papers. However, I did find the following:

  • Sleep deprivation enhances SWS.4
  • Sleep deprivation is known to decrease vigilance and mental ability and a number of researchers have shown a deterioration in physical performance.3
  • Following a loss of 1 or 2 nights, physical performance does not appear to be significantly impaired provided that participants are sufficiently motivated.3
  • VO2max decreased with sleep deprivation, suggesting a shorter time to exhaustion and a detrimental effect on endurance ability.3
  • … performance decrements are minor when sleep loss is less than 54 consecutive hours…7
  • Thus, it is likely that desynchronisation of circadian rhythms will decrease mental and physical performance more often than sleep loss per se.7
  • Missing 4h of sleep reduces the immunological activity of cancer- and virus-fighting cells by 28%.7
  • The risk of heatstroke is increases when sleep loss is combined with moderate to severe exercise.7
  • Stressful exercise may distrupt sleep due to stiff, sore muscles, blisters and other physical discomfort incurred during intensive exercise.4

Now in English…
As can be seen from the above information, sleep is not a uniform state. In the course of a night, you will cycle through the NREM and REM sleep stages, completing each cycle every 90 minutes on average – 80 minutes NREM sleep and 10 minutes REM sleep. According to the literature, NREM sleep dominates the first 2-3 cycles, with REM sleep increasing later in the night.

Slow wave sleep (SWS), stages 3 and 4 of the NREM cycle, is believed to be the most beneficial – a time when the body recovers from fatigue, repairs itself and replenishes its energy. The duration of SWS is enhanced by exercise induced body temperature increases, exercise of high intensity and exercise of long duration. Subjects awoken from SWS report feeling as if they were in a lighter phase of sleep than when awoken from REM.

In Conclusion

My theory is that particularly for the first sleep cycle, you’ll drop-off to sleep almost instantly, shooting through stage 2, straight into the beneficial SWS stages, 3 and 4. Since the REM phase is minimised in the first cycle anyway, you may bypass it altogether, beginning your second cycle after approximately 90 minutes. Thus, if you want a quick nap, utilise this cyclic pattern to maximise the benefits by sleeping in 90 minutes blocks i.e. 1.5hrs, 3hrs, 4.5hrs.

I am rather tempted by the notion that you awake more easily from the NREM phase and should thus sleep for 80 minutes (instead of 90min) in the first cycle, though I have a feeling that when you’re dead tired, you probably skip REM in the first cycle, lengthening the SWS stage. This could perhaps come into play if you’re having a long (4.5hr) sleep. So, instead of waking up after 4h30, you should awake at 4h20…

The flaw in my theory regarding sleep strategy is that the sleep cycles are 90 minutes, on AVERAGE, so you may already be into your second cycle – or finishing your first – when you are awoken. But, this research provides a guideline and will pretty much ensure that you get through at least one sleep cycle. Scientifically based, sleeping in 90minute blocks certainly makes more sense than sleeping for a random 20min, 50min or an hour at a time.

But… in a short race (24hr) where I (or one of my team mates) am unable to stay awake and we cannot risk the time for a 90min sleep, I’d take a quick 20min “power nap”. Whether this short sleep is physiologically beneficial or whether it is just a placebo effect remains in question. It may be a case of “I’ve slept for 20 minutes, thus I am rested”. Either way, if a power nap keeps your team going, then it would be worth the time taken.

In a longer race, instead of taking 20minutes here and there, it would make more sense to use your time wisely, co-ordinating sleep stops at the most strategically suitable times to get the most benefit from a long (1.5hr or 3hr) sleep.

Drawing assumptions from Armstrong’s (2000) findings that over a continuous 54hr period, “performance decriments are minor” and that “desynchronisation of circadian rhythms will decrease mental and physical performance more often than sleep loss per se”, you can try to ‘trick your body into thinking that you’re on a regular day-night circadian rhythm by not sleeping during the day and that, when you do sleep at night, to utilise the ‘sleep cycle’ system.

How frequently you should sleep? … there is little available on this subject. My rule of thumb is that if you or your team-mates are unable to function optimally, then it is time for some shut-eye. If you’re lost at night, then, instead of making silly decisions, sleep – even if only for 3hrs. If you reach a dark-zone, sleep. And, knowing the route and disciplines ahead, if you know that you’ll need to be mentally alert, sleep. The decision is yours and one which could ultimately win, or cost, you the race.


  • EEG (electroencephalograph) – An instrument that records ‘brain waves’. These electrical oscillations result from changes in voltage across the membranes of neurons. ‘Brain waves’ are recorded as small changes in voltage between two electrodes placed in contact with the scalp. The recordings are amplified and assessed for variations in duration – expressed as Hz (cycles per second, cps) – and amplitude, expressed in microvolts (uV).1, 2
  • EMG (electromyogram) – A recording of electrical activity from active muscles. Muscular activity can be recorded by placing electrodes on the skin overlying a muscle. An EMG is typically recorded from the muscles under the chin. These muscles show dramatic changes during the various sleep stages. 2
  • EOG (electro-oculogram) – A recording of eye movements. The eye is like a small battery. The retina is negative in relation to the cornea and thus, an electrode placed on the skin near the eye will record a change in voltage as the eye rotates in it’s socket. 2
  • Placebo Effect – In drug trials, there is always a control group – a group of subjects who are given non-active tablet (or other form of drug administration), not knowing that what they are taking is free of the active ingredient. This is done to assess whether the drug being tested is making a difference or not. It is a scientifically documented phenomena whereby subjects taking the placebo respond the same or better than those taking the drug, a purely psychosomatic (‘all in the mind’) reaction.
  • Sleep Onset Latency (SOL) – The time taken, from lights out, for the subject to fall asleep. The subject is determined to be asleep when the polysomnogram readings indicate the appearance stage 2 sleep.1, 3
  • Slow Wave Sleep (SWS) – A collective term for stages 3 and 4 of non-REM sleep. SWS is believed to be the most ‘restorative’ sleep stage, related to recovery from fatigue, repair and to the build up of energy.4, 5


  1. Taylor, S.R. and Driver, H.S. (1995) Is Sleep Affected by Physical Exercise and Fitness? Critical Reviews in Physical and Rehabilitation Medicine. 7 (2): 131-145
  2. Online sleep textbook –
  3. Driver, H.S. and Taylor, S.R. (1996) Sleep disturbances and exercise. Sports Medicine. 21 (1): 1-6
  4. Driver, H.S., Borrow, S.J., Bentley, A.J. and Shapiro, C.M. (1987) The Effect of Endurance Exercise on Sleep Patterns. Second South African Sports Medicine Association CongressApril 1987
  5. Driver, H.S., Rogers, G.G., Mitchell, D., Borrow, S.J., Allen, M., Luus, H.G. and Shapiro, C.M. (1994) Prolonged Endurance Exercise and Sleep Disruption. Medicine and Science in Sports and Exercise. 26 (7): 903-907.
  6. Driver, H.S. and Taylor, S.R. (2000) Exercise and Sleep. Sleep Medicine Reviews. 4 (4): 387-402
  7. Armstrong, L.E. (2000) Biorythmic Disturbances. In: Performing in Extreme Environments. Human Kinetics. USA.

Author: Lisa de Speville | This article was originally written in early 2001.

1 Comment

  1. Great article with some real info to back up the theory.
    In the recent Itera we did both power nap and longer sleeps still only getting 6:30 over the 5 days. Using both planned and emergency strategy, ie recognising when mistakes were happening due to SD.
    One critical issue I noticed in myself was the onset of hallucinations where trees and hedges became huge granite walls, disturbing but not threatening. The other was intense deja vu where I was utterly convinced of having done the whole thing before. I had to wrestle with the logic that of course I hadn’t but could help feeling all the time that I had but didn’t know how it ended.


Comments are closed.