Physiology of sleep

Michael Schupp MD FRCA

Christopher D Hanning MD FRCA

Sleep is a state of reversible unconsciousness in period of wakefulness and are repeated 3–6
which the brain is less responsive to external times each night and are typically displayed as Key points
stimuli. We are functionally blind during sleep an hypnogram (Fig. 1). The majority of deep A genetically determined
with no response to visual stimuli and a (stage 4) NREM sleep occurs in the first and activity–inactivity cycle is
found in all living organ-
decreased threshold of response to auditory second cycles. As the night progresses, the pro- isms.
stimuli. Babies have been exposed to sound of portion of REM sleep in a cycle increases and
In mammals, sleep is
up to 100 dB, which is above the legal limit for the NREM element is of lighter stage 2 sleep. divided into two distinct
ear protection for employees, without waking Age has a major effect on the duration of sleep states (NREM and REM)
up. In adults, the process is selective demon- and the ratio of NREM/REM sleep. Neonates with different mecha-
nisms and electrophysio-
strating continuing cortical function. For exam- sleep 16–18 h. It is widely distributed through- logical characteristics.
ple, a sleeping mother is woken by her crying out the day with REM sleep accounting for 50%
Sleep has profound
baby but not by other louder noises. of total sleep time (TST). This may be even effects on the control of
greater in premature babies. By the age of 2 yr, respiration.
Definitions and sleep children should sleep 10 h per day, mainly at Hypoxaemia is almost
Sleep is distinguished from unconsciousness night with one or two naps during the daytime always worse in REM
and REM sleep has declined to 20–25% of TST. sleep.
and anaesthesia by a characteristic cycle of sleep
Adults normally sleep 6–8 h per day with Anaesthesia and surgery
phases with specific EEG patterns and physio- suppress REM and slow-
logical changes. Natural sleep is divided into 15–20% REM sleep. With increasing age, TST wave sleep (SWS).
two distinctive states: non-rapid eye movement changes little although sleep is more fragmented
(NREM) and rapid eye movement (REM) sleep. with more frequent and longer awakenings
NREM sleep is then further divided into 4 stages (decreased sleep efficiency) with less REM sleep
where stage 1 is the lightest and stage 4 the and more light NREM sleep. Night-time sleep
deepest level of sleep. REM sleep is divided into may be decreased if naps are taken during the day.
phasic and tonic phases. The two distinctive
states follow a regular pattern called a sleep Functions of sleep
cycle which, in an adult, lasts about 90 min and The functions of sleep are still poorly under-
comprises a period of NREM sleep followed by stood. However, the observation that sleep (or, at
REM sleep. The cycles may be separated by a least, an activity–inactivity cycle) is present in

Michael Schupp MD FRCA

Specialist Registrar, Department of
Anaesthesia, Leicester General
Hospital, Leicester LE5 4PW
Christopher D Hanning MD FRCA
Consultant Anaesthetist, Sleep
Disorders Service, Leicester General
Hospital, Leicester LE5 4PW
Tel: 0116 258 4602
Fax: 0116 258 4611
E-mail: [email protected]
Fig. 1 Typical hypnogram of a young adult. (for correspondence)

DOI 10.1093/bjacepd/mkg069 British Journal of Anaesthesia | CEPD Reviews | Volume 3 Number 3 2003
© The Board of Management and Trustees of the British Journal of Anaesthesia 2003 69
Physiology of sleep

all species and has been preserved throughout evolution and that It has been suggested that sleep might conserve energy by
sleep deprivation leads to a drastic deterioration in cognitive reducing core temperature slightly and lowering metabolic rate
function and eventually to mental and physical morbidity proves by 10% compared with quiet wakefulness. Sleep would prevent
its importance. perpetual activity as a response to environmental stimuli leading

Fig. 2 EEG waveforms in different sleep stages.

70 British Journal of Anaesthesia | CEPD Reviews | Volume 3 Number 3 2003

 Physiology of sleep

to excessive energy consumption. However, sleep is a state of later cycles. It is the most abundant sleep stage in adults
starvation and there is no evidence that sleep is important for tis- accounting for up to 50% of TST.
sue repair. Sleep has been implicated as an important factor in
storage of long-term memory. Facts learned during the day are
Stages 3 and 4
usually better remembered the next morning whereas facts Deep NREM sleep stages 3 and 4, sometimes combined as
learned shortly before going to sleep are often poorly recalled. slow wave sleep (SWS) are characterized by high amplitude
low frequency delta waves (> 75 µV and 0.5–2 Hz) with stage
Electrophysiological characteristics of sleep 3 having between 20–50% and stage 4 more than 50% delta
The stages of sleep are characterised by typical patterns of activity. EMG activity is low and eye movements are rare.
electroencephalogram (EEG; Fig. 2), electro-myogram Arousal through auditory stimuli from this stage of sleep is
(EMG) and electro-oculogram (EOG) activity (Table 1). difficult and, if awakened, the individual is often disorientat-
Wakefulness with open eyes is characterised by an EEG with ed and slow to react. Return to sleep is easy and short arousals
dominant low amplitude, high frequency beta activity of (< 30 sec) are rarely remembered.
16–25 Hz. Muscle tone is usually high with high-to-moderate
REM sleep
EMG activity.
NREM sleep is followed by REM sleep, the proportion
Stage 1 increasing with each cycle. REM sleep is characterised by a fast
Sleep is usually initiated by a transition from wakefulness to mixed frequency low voltage EEG with saw-tooth waves and
a state of drowsiness with closed eyes and a shift from EEG rapid eye movements on the EOG. During the tonic phases of
beta activity to alpha activity of 8–12 Hz passing to Stage 1 REM sleep, there is marked reduction of muscle tone and EMG-
NREM sleep with a mixed frequency EEG-pattern with low activity in skeletal muscles. The tonic phases of REM sleep are
amplitude theta waves of 3–7 Hz accompanied by slow rolling interrupted by short episodes of phasic REM sleep with increased
eye movements. Involuntary muscle clonus occurs frequently, EMG activity and limb twitches. The atonia of REM sleep affects
resulting in jerky movement of the whole body (hypnic jerks) all skeletal muscles, except the diaphragm and the upper airway
and EMG activity is moderate-to-low. This stage lasts typical- muscles, and is associated with hyperpolarisation of the α-motor
ly only 5–10 min, during which time minor auditory stimuli neurones. The purpose of this may be to prevent the acting out of
will cause arousal. dreams. About 10% of the population have experienced sleep
paralysis (i.e. wakening from sleep and finding that the atonia has
Stage 2 persisted into wakefulness). It can be frightening but is entirely
Stage 2 is characterised by short bursts of high frequency harmless. Natural wakening usually occurs from REM sleep.
activity (12–15 Hz – sleep spindles) and K-complexes (large Subjects woken from REM sleep are much more likely to recall
amplitude biphasic waves). Bodily movements continue and dream content than those awakened from NREM sleep. NREM
the EMG activity is low-to-moderate. This stage is generally dreams are generally vague and formless in contrast to REM
short (10–20 min) in the first 1–2 cycles but predominates in dreams.

Table 1 Electrophysiological changes during different stages of sleep

Awake NREM sleep REM sleep

Stage 1 Stage 2 Stage 3 Stage 4
EEG Beta rhythm Mixed frequency with Sleep spindles 12–14 Hz Delta waves Delta waves Desynchronised low
16–25 Hz theta waves 3–12 Hz and high amplitude 0.5–2 Hz > 50% amplitude mixed
K-complexes (20–50%) frequency
EMG High/moderate Moderate/low Low/moderate Low Low Absent except small
muscle twitches in phasic
REM sleep
EOG Move with gaze Slow rolling Rare Rare Rare Bursts of REM in phasic
REM sleep

British Journal of Anaesthesia | CEPD Reviews | Volume 3 Number 3 2003 71

Physiology of sleep

Physiological changes during sleep greater during phasic REM sleep. Cerebral metabolic rate, oxy-
gen consumption and neuronal discharge rate are reduced during
Respiratory system NREM sleep but increased above resting values during REM
During NREM sleep, there is a decrease in respiratory drive and a sleep. The autonomic nervous system shows a general decrease
reduction in the muscle tone of the upper airway leading to a 25% in sympathetic tone and an increase in parasympathetic tone,
decrease in minute volume and alveolar ventilation and a doubling except in phasic REM sleep.
of airway resistance accompanied by a small (0.5 kPa) increase in
PaCO2 and decrease in PaO2. Hypercarbic and hypoxic ventilatory Renal system
drives are reduced compared with wakefulness. The breathing pat- The glomerular filtration rate and filtration fraction are reduced
tern is regular except at the transition from wakefulness into sleep and ADH secretion is increased resulting in a low volume con-
when brief central apnoeas are common. centrated urine.
During REM sleep there is a further decrease in hypercarbic and,
particularly, hypoxic ventilatory drives. The breathing pattern is
Endocrine system
irregular especially during phasic REM sleep. The loss of skeletal The secretion of several hormones is directly linked to the
muscle tone in REM sleep affects the intercostal and other muscles sleep/wake cycle. Melatonin is released from the pineal gland
which stabilise the chest wall during inspiration. In infants, this under the control of the supra-chiasmatic nuclei (SCN) in a 4–5
may be seen as paradoxical movement of the rib cage and h pulse, usually beginning at the onset of darkness (~9 pm). The
abdomen. In adults, there may be maldistribution of ventilation and pulse is inhibited or delayed by exposure to bright light in the
impaired ventilation–perfusion matching with consequent arterial evening. It is best regarded as being permissive of sleep (‘open-
hypoxaemia. In normal subjects, this is unimportant but it may be ing the gate to sleep’) rather than as an hypnotic, as it is possible
very important in patients with chronic lung disease or abnormali- to maintain wakefulness during this period. Growth hormone is
ties of the thoracic (e.g. kyphoscoliosis). The great majority of mostly secreted during the first episode of SWS, particularly
patients with impaired respiratory function will be at their worst during puberty. Prolactin concentrations also increase shortly
during REM sleep. after sleep onset and decrease with wakefulness. Sleep phase
delay delays secretion of both of these hormones. The secretion
Cardiovascular system of cortisol decreases with the onset of sleep and reaches a trough
Blood pressure decreases during NREM and tonic REM sleep in the early hours of the morning and a peak just after waking.
but may increase above waking values during phasic REM
sleep. Cardiac output is generally decreased during all sleep
Temperature control
phases. Systemic vascular resistance (SVR) and the heart rate In contrast to anaesthesia, thermoregulation is maintained during
are both reduced during NREM and tonic REM sleep and sleep. However, the shivering threshold is decreased and body
increased during phasic REM sleep. core temperature decreases by about 0.5°C in humans and 2°C
in hibernating mammals. Body temperature is linked to the cir-
Central nervous system cadian rhythm and reaches its nadir at about 3 am.
Cerebral blood flow (CBF) increases by 50–100% above the Thermoregulation is quite good in human infants compared with
level of resting wakefulness during tonic REM sleep and is even other species.
Table 2 Physiological changes in sleep

NREM sleep Tonic REM sleep Phasic REM sleep

vs wakefulness vs NREM sleep vs tonic REM sleep
EEG activity ↓ ↑ ↑↑
Cerebral blood flow Unchanged ↑ ↑↑
Heart rate ↓ Unchanged ↑ and variable
Blood pressure ↓ Unchanged ↑ and variable
Respiratory rate ↓ ↑ and variable ↑ and variable with central apnoeas
Airway resistance ↑ ↑ and variable ↑ and variable
Responsiveness to ↑CO2 ↓ ↓ No difference
Upper airway muscle tone ↓ ↓
Penile and clitoral tumescence ++ +++

72 British Journal of Anaesthesia | CEPD Reviews | Volume 3 Number 3 2003

 Physiology of sleep

Control of sleep brain where stimulation or ablation would lead to sleep; and (ii)
a hormone or transmitter which would reliably induce sleep.
Sleep follows a circadian (~1 day) cycle, the periodicity of Neither have been found because the mechanisms resulting in
which is regulated by an independent genetically determined sleep are complex and diffuse.
‘intrinsic clock’ which is entrained to a 24 h cycle by external During wakefulness, the CNS is dominated by activity of the
cues (Zeitgebers) such as light, darkness, clock time, working ascending reticular activating system (RAS) in the brain stem.
patterns and meal times. When a human being is deprived of all This formation receives sensory input from all peripheral sen-
external time clues and is exposed to constant levels of illumi- sors and projects to the thalamus and the cortex. Its main neuro-
nation (‘free running’), the wake/sleep cycle typically lengthens transmitters are acetylcholine, noradrenaline, dopamine and his-
to about 24.5 h. Subjects who are born blind without any appre- tamine which explains the sedative effect of antagonists to these
ciation of light generally free run while those blinded in later life substances. A decrease in its activity permits sleep to be initiat-
or who retain some perception of light remain entrained. ed by suppressing incoming external stimuli.
All living organisms, including plants and fungi, have been The induction of SWS is associated with the secretion of γ-
found to have clock genes and to show an inactivity/activity aminobutyric acid (GABA) from basal forebrain neurones.
cycle. In mammals, control of the intrinsic clock is located in the Therefore, it is not surprising that benzodiazepines and barbitu-
SCN on either side of the third ventricle, just above the optical rates, which act through stimulation of GABA receptors in the
chiasm. In animal experiments, its destruction leads to a change CNS, induce sleep or anaesthesia. Cholinergic mechanisms ini-
from the normal sleep cycle into several shorter sleep/activity tiate REM sleep through stimulation of pontine neurones in the
periods during the day. As noted above, melatonin secretion is lateral portion of the pontine tegmentum and the nucleus reticu-
prompted by the SCN just before the usual time of sleep onset. A laris pontis oralis. In animal experiments, injection of carbachol
mismatch of this pattern with sleeping time, as occurs in shift (acetylcholine agonist) induces instantaneous REM sleep.
workers and after trans-meridian flights, leads to sleep distur- Recently, orexins (hypocretin) have been isolated in the hypo-
bance (‘jet lag’) as the subject is trying to sleep during their cir- thalamus and appear to be important in the control of REM sleep
cadian day. Light therapy can be helpful in re-setting the circadi- and appetite. CSF concentrations of orexins have been found to
an clock and the interested reader is referred to the bibliography. be very low in patients with narcolepsy.
The propensity to fall asleep varies throughout the day and
depends upon both circadian factors (process C) and time since Influence of surgery and anaesthesia on sleep
the last sleep period (process S). The longer the time since the Anaesthesia and surgery can have a profound effect upon
last sleep period, the greater will be process S. However, its sleep. On the first night after surgery, sleep architecture is
propensity will be modulated by process C. The circadian pres- severely disrupted with little or no SWS and REM sleep. The
sure to sleep is greatest at ~2 am with a secondary peak at ~2 pm. light Stage 2 sleep is fragmented with frequent awakenings.
It is least at ~6 am and ~6 pm. If a subject elects to stay awake The degree of disruption appears to be related to the severity
throughout the night, they will feel most sleepy in the small of the surgical insult. The mechanism is unclear but it is prob-
hours of the morning but will get a ‘second wind’ as morning ably due to a combination of the surgical stress and the effects
approaches and the circadian pressure to sleep declines. If wake- of opioid analgesics.
fulness is maintained, a second period of sleepiness and relative Recovery of lost SWS and REM sleep occurs on postopera-
alertness will follow in early afternoon and early evening, tive nights 2–5, being later after major surgery. This coincides
respectively. Some of the 8-h sleep debt will be recovered that with the nadir of postoperative pulmonary function and several
night but process C will ensure that awakening will occur at or studies have demonstrated marked hypoxaemia associated with
shortly after the normal waking time. the rebound of REM sleep. It was a logical step to attribute post-
Sleep is normally an actively initiated and not a passive operative myocardial ischaemia, myocardial infarction, pul-
process. Unless a subject is sleep deprived, successful initiation monary embolism and cerebral impairment (delirium and cogni-
of sleep depends both upon the phase of the circadian clock and tive impairment) to nocturnal hypoxaemia. However, a number
external factors (recumbent position, darkness, reduction of sen- of studies have failed to confirm these presumed associations,
sory input). Over the years, considerable effort has been focused although this does not exclude the possibility that the hypox-
on a search for: (i) a ‘sleep centre’, a nucleus or region in the aemia may be important in some individuals.

British Journal of Anaesthesia | CEPD Reviews | Volume 3 Number 3 2003 73

Physiology of sleep

Key references Shneerson JM. Handbook of Sleep Medicine. Oxford: Blackwell, 2000
Ambrosini MV, Giuditta B. Learning and sleep: the sequential hypothesis. Williams JM, Hanning CD. Obstructive sleep apnoea, BJA CEPD Rev 2003;
Sleep Med Rev 2001; 5: 477–90 3: 75–78
Dijk DJ, Lockley SW. Functional genomics of sleep and circadian rhythm:
integration of human sleep-wake regulation and circadian rhythmici-
ty. J Appl Physiol 2002; 92: 852–62
Web resources
Basics of Sleep Behaviour <www.sleephomepages.org/sleepsyllabus>
Douglas N. Clinician’s Guide to Sleep Medicine. Edinburgh: Arnold, 2002
International Sleep Medicine Association <www.1sleep.com>
Ebrahim IO et al. The hypocretin/orexin system. J R Soc Med 2002; 95:
227–30 Medsleep Educational Materials <www.aasmnet.org/MEDSleep/med-
sleephome>
Kryger MH, Roth T, Dement WC. (eds) Principles and Practice of Sleep
Medicine, 3rd edn. Philadelphia: 2000. The Sleep Home Pages <www.sleephomepages.org>
Nicolau MC et al.Why we sleep: the evolutionary pathway to the mam- The Sleep Medicine Home page <www.cloud9.net/~thorpy>
malian sleep pattern. Prog Neurobiol 2000; 62: 379–406 The Sleep Net <www.sleepnet.com>
Saper CB, Chou TC, Scammell TE.The sleep switch: hypothalamic control
of sleep and wakefulness. Trends Neurosci 2001; 24: 726–31 See multiple choice questions 50–52.

74 British Journal of Anaesthesia | CEPD Reviews | Volume 3 Number 3 2003