CHAPTER 3

Sleep Stages and

Scoring Technique
Raman K. Malhotra
Alon Y. Avidan

Introduction to Sleep Stage Scoring right occipital; electrode location: ground (FPZ), ref-
The original Rechtschaffen and Kales sleep scoring manual erence (CZ), A1 or M1 and A2 or M2; C3 and O1, left
of 1968, commonly known as the R and K rules, was used central and occipital, respectively, EEG electrodes; C4
until 2007, at which point the American Academy of Sleep and O2, right central and occipital, respectively, EEG
Medicine (AASM) updated the scoring manual in what electrodes. The exploring reference electrode (F3, F4,
is commonly known as the AASM scoring manual. The C3, C4, O1, and O2) is chosen on the opposite side
Rechtschaffen and Kales method divides sleep into five dis- of the head from the mastoid electrode (M1, M2) or
tinct stages: non–rapid eye movement (non-REM [NREM]) average (AVG).
stages 1, 2, 3, and 4 and stage REM sleep. The AASM scor- Electromyogram electrodes (EMG): LtTib1-LtTib2 and
ing manual recognizes four sleep stages: Stage N1 (formerly RtTib1-RtTib2, left and right tibialis anterior EMG
stage 1 sleep), stage N2 (formerly stage 2 sleep), stage N3 electrodes; Chin1-Chin2, submental EMG signal; chin
(formerly stages 3 and 4 sleep), and stage R sleep (formerly EMG, Chin1-Chin2; limb EMG (left leg, right leg),
stage REM sleep). The reader is reminded that sleep stages LtTib1-LtTib2, RtTib1-RtTib2.
should not be viewed as distinct entities, but rather as a grad- Electrocardiogram electrodes (ECG): ECG1-ECG2,
ual transition of a waveform. The scoring rules were devised ECG2-ECG3.
to allow uniformity between sleep laboratories and to offer a Respiratory electrodes: SNORE, snore sensor sound;
conceptual simplicity rather than a rigid framework. OroNs, oronasal airflow or oronasal thermistor;
Particular signals of interest for sleep scoring include PFLOW, nasal pressure transducer; THOR/CHEST
those generated from the cerebral cortex (electroen- and ABD, chest and abdominal walls motion effort;
cephalogram [EEG]), eye movements (electro-oculogram THOR1-THOR2, thoracic effort channel; ABD1-
[EOG]), and the muscles of the face (picked up by chin ABD2, abdominal effort channel; CFLOW, con-
electromyographic activity). Electrode placement for EEG tinuous positive airway pressure (CPAP) airflow
scoring rules follows the international 10-20 system, which channel; Pco2, mm Hg of carbon dioxide; Spo2,
assigns a number to each EEG electrode to specify the loca- oxygen saturation by pulse oximetry by finger probe;
tion in the left or right hemisphere (Table 3.1). Pleth, plethysmography; Pt Position, patient position
Figures 3.1 and 3.2 depict the updated AASM- (supine, left, right, prone).
recommended derivations for recording the EEG. Figures are provided with a 1-second ruler.
This chapter will include a discussion of the specific
parameters required for staging sleep and a summary of the
various EEG activity needed to score sleep. This discussion
Parameters for Staging Human Sleep
will be followed by a discussion of the stages of sleep using Common to all PSG monitoring is the measure of the
specific polysomnographic (PSG) records. following three physiological parameters:
The following abbreviations are used in the PSG samples 1. EOG leads: left eye and right eye
provided in this chapter: 2. EEG leads: one occipital, one central, and one frontal
Electro-oculogram electrodes (EOG): Left outer canthus: lead
LOC-A2; right outer canthus: ROC-A1; LOC-A2, 3. EMG lead: one submental EMG channel
ROC-A1, left and right electro-oculogram referred
to right and left mastoid leads; M2, right mastoid
Eye Movements (Electro-oculogram Activity)
electrode location; M1, left mastoid electrode loca-
tion; L/ROC-AVG, left and right electro-oculogram The EOG signals measure changes in the electrical potential
referred to an average reference electrode; E1, left of the positive anterior aspect of the eye, the cornea, rela-
outer canthus eye electrode; E2, right outer canthus tive to the negative posterior aspect, the retina. Horizontal
eye electrode. axis electrodes are placed near the outer canthi and vertical
Electroencephalogram electrodes (EEG): C3-A2, C4-A1, axis electrodes 1 cm below (LOC) and 1 cm above (ROC)
O1-A2, O2-A1, left central, right central, left occipital, the eye to measure transient changes in potential during the
77
78 Atlas of Sleep Medicine

Table 3.1  10-20 Electrode Placement

The 10-20 system assigns a number to further specify the location in the left or right hemisphere. Location z
is used to indicate that the location of the electrode is in the midline or “zero,” meaning that it is neither left
hemisphere nor right hemisphere. The electrode placed at Cz is said to be the vertex, meaning that it is the
midcentral or at the top of the head. Fpz and Oz are used in achieving the other measurements and can be used
as additional electrode placements for localization of activity. Fpz may be used as the location of the common or
ground electrode placement.

Nasion

Fpz
Fp1 Fp2

F7 F3 Fz F4 F8

T3 C3 Cz C4 T4
A1 A2

P3 Pz P4
T5 T6

O1 O2
Odd Oz Even

Inion
Brain area Left Midline Right
Frontal pole Fp1 Fpz Fp2
Frontal F3 Fz F4
Inferior frontal F7 F8
Anterior temporal T1 or F9 T2 or F10
Midtemporal T3 T4
Posterior temporal T5 T6
Central C3 Cz C4
Parietal P3 Pz P4
Occipital O1 (Oz) O2

actual eye movement (Fig. 3.3). During any eye movement • T

 he cornea (front) has a positive polarity. The retina
the cornea moves toward one electrode, while the retina (back) has a negative polarity.
moves away. When the eye is not moving, the change in rela- • E
 OG placement (LOC and ROC) is on the outer can-
tive position is zero, and the eye leads do not record a signal. thus of the eye, offset 1 cm below (LOC) and 1 cm
Slow rolling eye movements (SREMs) occur during drows- above (ROC) the horizon.
iness and light sleep and are recorded as long gentle waves, • E
 lectro-oculography picks up the inherent voltage of
whereas rapid jerking movements are represented by sharply the eye. During eyes-open wakefulness, sharp deflec-
contoured fast waves. Blinking of the eyes produces rapid tions in the EOG tracing may indicate the presence of
vertical movements. During REM sleep, eye movements eye blinks.
again become active and jerky. The intensity of the bursts of
activity is used to describe the density of REM sleep.
Electroencephalographic Recording
Electro-oculographic Recording Wakefulness and sleep are determined by the characteris-
• E
 OG voltages are higher than EEG signal. Because the tic patterns of the scalp EEG signals and are of fundamen-
eye is outside of the skull structure, there is no bone tal importance in interpreting PSG studies. EEG records
to attenuate signal. electrical potentials generated by the cortex but can reflect
 CHAPTER 3  Sleep Stages and Scoring Technique 79

Updated recommended derivations R L

Nasion E2
1 cm
1 cm
1 cm 1 cm

Fpz E1
Fp1 Fp2
Gaze to the right Gaze to the left

F7 F3 Fz F4 F8 + − − +
−
T3 C3 Cz C4 T4
E2−M2
+
E1−M2 −
Mastoid Mastoid +
M1 T5 P3 Pz P4 T6 M2 E1: Left outer canthus eye electrode (previously LOC)
E2: Right outer canthus eye electrode (previously ROC)
M2: Right mastoid electrode location
O1 O2 FIGURE 3.3  The eye can be envisioned as a battery with the positive
Oz pole at the cornea and the negative pole at the retina. The electro-
oculogram (EOG) consists of a bipolar linkage from the right electro-
oculogram (ROC) electrode 1 cm lateral and 1 cm superior to one outer
Inion canthus to the left electro-oculogram (LOC) electrode 1 cm lateral and
FIGURE 3.1  Updated American Academy of Sleep Medicine (AASM) 1 cm inferior to the other outer canthus. The electrode toward which the
recommended derivations for recording the electroencephalogram. eyes move becomes relatively positive, the other relatively negative. As
Electrode locations as recommended in 2007 by the AASM. (Modi- the eyes move during sleep, they produce corresponding changes in the
fied from Kryger MH. Atlas of Clinical Sleep Medicine. Philadelphia: electrical field, producing a correlating potential change in the electro-
Saunders/Elsevier; 2010.) encephalogram electrodes. This can be verified by noting correspond-
ing movements in the EOG channels. (From Avidan AY, Barkoukis T,
eds. Review of Sleep Medicine. 3rd ed. Philadelphia: Elsevier; 2011.)
Updated alternative derivations

Nasion the influence of deeper brain structures, such as the thala-

mus. Measurement of the EEG signal is possible because of
the relative difference in potential between two recording
electrodes in grid 1 and grid 2 of the channel. A negative
Fpz discharge in grid 1 relative to grid 2, by convention, is rep-
Fp1 Fp2 resented by an upwardly deflecting wave in grid 1 of the
channel. The PSG references the left or right electrodes to
electrodes on the opposite right and left ears (A2, A1) or
F7 F3 Fz F4 F8
mastoids (M2, M1). The general rule is to read only from
the right cortical channel. However, when this channel
develops artifact or the validity of the signal is suspected,
T3 C3 Cz C4 T4 comparison is made with the left channel.
Electroencephalographic Recording Criteria:
Mastoid Mastoid • M
 inimum paper speed of 10 mm/sec. One page
M1 T5 P3 Pz P4 T6 M2 equals 30 seconds and is defined as one epoch.
• T
 ime constant of 0.3 seconds or low-frequency filter
of 0.3 Hz.
O1 O2 • P
 en deflections of 7.5 to 10 mm for 50μV are
Oz recommended.
 lectrode impedances should not exceed 5000 Ω.
• E
Inion
FIGURE 3.2 Updated American Academy of Sleep Medicine alter- Electromyographic Recording
native derivations for recording the electroencephalogram. Alterna-
The EMG signals are muscle twitch potentials, which are
tive electrode locations as recommended in 2007. Backup electrode
substitutions: O1 for Oz, Fpz for Fz, and C3 for Cz. Fp, Frontopolar or used in PSG to distinguish between sleep stages based on
prefrontal; F, frontal; C, central; T, temporal; P, parietal; O, occipital; the fact that EMG activity diminishes during sleep. Spe-
A, ear or mastoid; F3, left midfrontal; P3, left parietal; T4, right tempo- cifically, during REM sleep, muscle activity is minimal.
ral; A1, right ear; Cz, vertex. (Modified from Kryger MH. Atlas of Clinical Compounding the problem of interpreting EMG channels is
Sleep Medicine. Philadelphia: Saunders/Elsevier; 2010.) occasional intrusion of EMG artifact into the record. Some
80 Atlas of Sleep Medicine

Table 3.2  Sleep Electroencephalographic Waveforms

Sample Definition Label
Alpha activity 8- to 13-Hz rhythm, usually most prominent in occipital leads. Thought to
be generated by cortex, possibly via dipole located in layers 4 and 5. Used
as a marker for relaxed wakefulness and CNS arousals.

Theta activity 4- to 8-Hz waves, typically prominent in central and temporal leads.
Sawtooth activity (shown in figure) is a unique variant of theta activity
(containing waveforms with a notched or sawtooth-shaped appearance)
frequently seen during REM sleep.

Vertex sharp waves Sharply contoured, negative-going bursts that stand out from the back-
ground activity and appear most often in central leads placed near the
midline.

Steep spindle A phasic burst of 11- to 16-Hz activity, prominent in central scalp leads;
typically last for 0.5-1.5 seconds. Spindles are a scalp representation of
thalamocortical discharges; the name derives from their shape (which is
spindlelike).

K complex Recently redefined in the AASM manual as an EEG event consisting of a well-
delineated negative sharp wave immediately followed by a positive component
standing out from the background EEG with total duration ≥ 0.5 seconds,
usually maximal in amplitude over the frontal regions.

Slow waves High-amplitude (≥75 μV) and low-frequency (≤2 Hz) variants of delta (1-4
Hz) activity. Slow waves are the defining characteristics of stage N3 sleep.

REM Rapid eye movements are conjugate saccades occurring during REM sleep
correlated with the dreamer's attempt to look at the dream sensorium.
They are sharply peaked with an initial deflection usually <0.5 second in
duration.

SEM Slow eye movements are conjugate, usually rhythmical, rolling eye move-
ments with an initial deflection usually ≥0.5 second in duration.

Modified from Kryger MH. Atlas of Clinical Sleep Medicine. Philadelphia: Saunders/Elsevier; 2010.
AASM, American Academy of Sleep Medicine; CNS, central nervous system; EEG, electroencephalogram; REM, rapid eye movement; SEM, slow eye
movement.

of these intrusions are in the form of yawns, swallows, and Electroencephalographic Activity
teeth grinding (bruxism).
During Wakefulness and Sleep
Submental Electromyographic Recording Six EEG wave patterns are used to differentiate wake and
Criteria sleep states and classify sleep stages: (1) alpha activity,
• M
 ental (mentalis muscle) and submental (mylohyoid (2) theta activity, (3) vertex sharp waves, (4) sleep spindles,
and anterior belly of the digastrics muscle) place- (5) K complexes, and (6) slow wave activity. These are sum-
ments are acceptable. marized in Tables 3.2 and 3.3.
• T
 hese are used to detect muscle tone changes for Cortical activity can be characterized by specific fre-
scoring REM versus NREM sleep. quencies. Frequency is defined as the number of times a
• M
 uscle tone is high during wakefulness and NREM sleep. repetitive wave recurs in a specific time period (typically
It is lower in NREM sleep than in wakefulness. It is gen- 1 second). Frequency is noted as cycles per second (i.e.,
erally lower in slow wave sleep then in stage N1 or N2. Hertz [Hz]). EEG activity has been divided into four bands
• M
 uscle tone is lowest during stage REM. based on the frequency and amplitude of the waveform, and
 CHAPTER 3  Sleep Stages and Scoring Technique 81

Table 3.3  Definitions and Examples of Sleep Figures Encountered on an Electroencephalogram

Please note that although all slow waves are in the delta frequency range, not all delta waves are slow waves.
EEG Rhythm Characteristics Best Seen Examples
Posterior 8.5-13 Hz Occipital LOC-A1
dominant
rhythm ROC-A1

F4-A1

C4-A1

O2-A1

Slow waves 0.5-2 Hz; amplitude ≥ 75 μV Frontal LOC-A1

ROC-A1

F4-A1

C4-A1

O2-A1

Spindle 11-16 Hz; duration ≥ 0.5 sec Central LOC-A1

ROC-A1

F4-A1

C4-A1

O2-A1

K complex Diphasic; large amplitude, Frontal LOC-A1

duration ≥0.5 sec

ROC-A1

F4-A1

C4-A1

O2-A1

Modified from Vaughn BV, Giallanza P. Technical review of polysomnography. Chest. 2008;134:1310-1319.
EEG, Electroencephalogram.

Alpha Activity (see Table 3.3)

the bands are assigned Greek letters (alpha, beta, theta, and
delta). The EEG frequencies are defined slightly differently • I t is also known as posterior dominant rhythm.
according to the reference used. The following conven- • A lpha EEG is 8 to 13 Hz.
tion is used to define EEG frequencies: beta is greater than • I t originates in the parietooccipital regions bilaterally.
13 Hz; alpha is between 8 and 13 Hz; theta is between 4 and • A normal alpha rhythm is synchronous and symmetri-
less than 8 Hz, and delta is the slowest activity at less than cal over the cerebral hemispheres.
4 Hz. Another EEG activity is gamma, which ranges from • I t is seen during quiet alertness with eyes closed.
30 to 45 Hz. • E ye opening causes the alpha waves to “react” or
decrease in amplitude.
• I t has a sinusoidal appearance.
Beta Activity
• D ecrease in frequency occurs with aging.
• B eta EEG is defined as a waveform between 14 and
30 Hz but is usually between 18 and 25 Hz.
Theta Activity
• I t originates in the frontal and central regions but can
also occur more diffusely. • T heta activity has a frequency of 4 to less than 8 Hz.
• I t is present during wakefulness and drowsiness. • I t originates in the central vertex region.
• I t may be more persistent during drowsiness, dimin- • T here is no amplitude criteria for theta.
ish during deeper sleep, and reemerge during REM • I t is the most common EEG sleep frequency.
sleep.
• T he amplitude over the two hemispheres should not
Sleep Spindles
vary by more than 50%.
• E nhanced or persistent activity suggests use of sedative- • S
 leep spindles are 11 to 16 Hz.
hypnotic medications. • T
 hey originate in the central vertex region.
82 Atlas of Sleep Medicine

Table 3.4  Major Differences Between Rechtschaffen and Kales Manual and the AASM Scoring Manual from 2007
Differences R and K Manual AASM Scoring Manual
EEG electrodes Score sleep stages using central (C3, C4) leads Score using frontal, central, and occipital leads
Major body movements Movement time can be scored if more than half the No movement time staging exists
epoch is obscured
Slow wave sleep Consists of both stage 3 and stage 4 sleep with delta Only recognizes stage N3 sleep with delta wave
wave amplitude measured using central leads amplitude measured using frontal leads
Terminology of stages Stage 1, stage 2, stage 3, stage 4, and stage REM sleep Stage N1, stage N2, stage N3, and stage R
sleep
Reference electrode Left and right ear or mastoid, termed A1 or A2 Left and right mastoid, termed M1 or M2
Scoring stage 2 (or N2) 3-minute rule that states if greater than 3 minutes No 3-minute rule exists
sleep pass between spindles or K complexes, then score
stage 1 sleep

AASM, American Academy of Sleep Medicine; EEG, electroencephalogram; R and K, Rechtschaffen and Kales; REM, rapid eye movement.

• T
 hey have a duration criterion of at least 0.5 seconds drowsy, with the eyes closed, the EEG will show predomi-
for the purpose of sleep stage scoring. nant alpha activity, whereas the EMG activity will become
• T
 hey characterizes stage N2 sleep but can be seen in less prominent. The EOG channels may show SREMs. If
other stages. the patient moves in bed or rolls, the record will reflect
this as a paroxysmal event characterized by high-amplitude
activity with sustained increased artifact. The patient
K Complexes (see Table 3.3)
may enter stage N1 sleep for one or two epochs and then
• T
 hey are sharp, slow waves, with a biphasic morpho- reawaken. Transitions may be difficult to score. From stage
logic structure (negative then positive deflection). W, patients typically proceed to stage N1, but infrequently
• T
 hey are predominantly central vertex in origin. they may enter REM sleep or stage N2 sleep directly if the
• D
 uration must be at least 0.5 seconds. pressure to do so is high (reflecting a state of pathological
• T
 hey do not have an amplitude criteria. sleep deprivation).
• T
 hey are indicative of stage N2 sleep.
Stage N1 NREM Sleep (Fig. 3.6)
Delta Activity (see Table 3.3)
Stage N1 NREM sleep may also be termed transitional
• D elta activity has a frequency of 0.5 to 2 Hz for the sleep or light sleep. Transition into sleep occurs follow-
purpose of sleep scoring. ing stage W sleep. Stage N1 NREM sleep is a transitional
• I t is seen predominantly in the frontal region. state characterized by low-voltage, fast EEG activity. The
• D elta activity has an amplitude criterion of 75μV or EEG patterns may be quite variable and may shift rap-
greater. idly, making it sometimes difficult to interpret. Stage N1
• T here is no duration criterion. sleep is scored when more than 15 seconds (≥50%) of the
epoch is made up of theta activity (4 to 7 Hz), sometimes
intermixed with low-amplitude beta activity replacing the
Stages of Sleep alpha activity of wakefulness. Amplitudes of EEG activity
When scoring a record for stage of sleep, the reader should are less than 50 to 75 μV. Paroxysms of 4 to 7 Hz less than
initially scroll through the entire record quickly to evaluate 75 μV may occur. The alpha activity in the EEG drops to
the quality of the recording and the usefulness of specific less than 50%. Stage N1 is of very short duration, lasting
channels. He or she should observe the specific shape of the for 1 to 7 minutes.
features that represent the stages in that particular individ- Vertex sharp waves may occur toward the end of stage
ual and gain an overall picture of the cycles for that record. N1, but sleep spindles or K complexes are never a part of
Specifically observe for sleep spindles, K complexes, slow stage N1 sleep and neither are rapid eye movements. Vertex
waves, and rapid eye movements. waves have a characteristic high-voltage sharp surface nega-
Table 3.4 illustrates the major differences between the tive followed by surface positive component and are maxi-
R and K manual and the AASM scoring manual from 2007. mal over the Cz electrode. The EMG shows less activity
than in wake stage, but the transition is gradual and of little
assistance in scoring. Arousals are paroxysms of activity last-
Stage Wake (Figs. 3.4 and 3.5)
ing 3 seconds but less than 15 seconds. If an arousal occurs
Typically the first several minutes of the record will consist in stage N1 sleep, and if the burst results in alpha activity
of wake (W) stage. Stage W is recorded when more than for greater than 50% of the record, then the epoch is scored
50% of the epoch has scorable alpha EEG activity. The as stage W.
EEG will show mixed beta and alpha activities as the eyes During drowsiness and stage N1 sleep, the eyes begin
open and close, and predominantly alpha activity when the to slowly roll—SREMs. Sometimes eye movements during
eyes remain closed. Submental EMG is relatively high tone drowsiness and stage N1 NREM sleep may be jerky, irregu-
and will reflect the high-amplitude muscle contractions lar, or gently rolling. Theta activity may start to enter into
and movement artifacts. The EOG channels will show eye the EOG tracing as an artifact. The submental EMG tone
blinking and rapid movements. As the patient becomes is relatively high.
 CHAPTER 3  Sleep Stages and Scoring Technique 83

2
4
1

FIGURE 3.4  Stage wake (30-second epoch).  Wakefulness, eyes open. More than 50% of the epoch has scorable alpha electroencepha-
logram (EEG) activity (1). Electromyogram (EMG) activity is reduced, consistent with relaxed wakefulness (2). Note the posterior dominant
alpha frequency in the O1-A2 and O2-A1 leads (3). An electrocardiogram (ECG) artifact is noted in the electro-oculogram (EOG), EEG, and
EMG leads (4). OroNs, Oronasal airflow; PFLOW, nasal pressure transducer; Spo2, pulse oximetry.

FIGURE 3.5  Stage wake (60-second epoch).  Wakefulness with faster rolling eye movements. The electromyogram activity is elevated
consistent with wakefulness.
84 Atlas of Sleep Medicine

LOC-A2
128 uV
2
ROC-A1
128 uV

Chin1-Chin2
68.3 uV 3
C3-A2
85.3 uV 1
C4-A1
85.3 uV

O1-A2
85.3 uV

O2-A1
85.3 uV

ECG2-ECG3
1.37 mV

LtTib1-LtTib2
170.7 uV

RtTib1-RtTib2 4
170.7 uV

30 sec/page

SNORE
512 uV

OroNs
409.6 uV

THOR
1.09 mV

ABD
1.09 mV

PFLOW
100 mV
100
Spo2
% 0
FIGURE 3.6  Stage N1 sleep (30-second epoch).  There is presence of low-voltage, mixed-frequency theta activity demarcated by the
arrows (1). Slow rolling eye movements are evident (2) and so is a more substantial reduction in chin electromyogram tone (3), which happens
to capture activity from the electrocardiogram (ECG) leads in the form of an ECG artifact (4).

Physiologically the patient’s breathing becomes shallow, K complexes and sleep spindles occur for the first time
heart rate becomes regular, blood pressure falls, and the and are typically episodic. The EOG leads mirror EEG
patient exhibits little or no body movement. This portion activity. Submental EMG activity is tonically low. Excessive
of sleep is distinguished by drifting thoughts and dreams spindles activity may indicate the presence of medications
that move from the real to the fantastic, along with a kind (such as benzodiazepines). K complexes (see Fig. 3.8) are
of floating feeling. The sleeper is still easily awakened and sharp, monophasic or polyphasic slow waves, with a sharply
might even deny having slept. In general the time spent in negative (upward) deflection followed by a slower positive
Stage N1 increases with age. (downward) deflection. They characteristically stand out
from the rest of the background. K complexes have a dura-
tion criterion and must persist for at least 0.5 seconds. No
Stage N2 NREM (Figs. 3.7 to 3.9)
minimum amplitude criterion exists for K complexes.
Stage N2 NREM sleep may also be termed sigma, spindle, K complexes, even without the presence of sleep spindles, are
or intermediate sleep (see Fig. 3.7). It is an intermediate sufficient for scoring stage N2 sleep. They are predominantly
stage of sleep, but it also accounts for the bulk of a typi- central vertex in origin.
cal PSG recording (up to 50% in adult patients). It follows K complexes may occur with or without stimuli such as a
stage N1 NREM sleep and initially lasts about 20 minutes. sudden sound, and in this respect they may represent a form
It is characterized by predominant theta activity (4- to 7-Hz of cortical evoked potential in a brain still minimally respon-
EEG activity) and occasional quick bursts of faster activity. sive to external stimuli. K complexes may be labeled sponta-
The EEG may show minimal alpha activity. Amplitude may neous if they arise from an unknown reason, indicating that
increase from that seen in stage N1 sleep. Delta activity is their origin is of an endogenous brain activity. They may be
only allowed to occur for less than 20% of the epoch. The labeled evoked if they are clearly triggered by an external
threshold triggering slow wave sleep scoring is reached if stimulus such as sound/noise. K-alpha complexes may be
20% of the epoch consists of delta activity. triggered by other entities such as periodic limb movements
 CHAPTER 3  Sleep Stages and Scoring Technique 85

FIGURE 3.7  Stage N2: the morphological structure and characteristics of sleep spindles (30-second epoch). The illustrated montage is accord-
ing to American Academy of Sleep Medicine scoring criteria.

FIGURE 3.8  Stage N2: the morphological structure and characteristics of K complexes (30-second epoch). The illustrated montage is accord-
ing to American Academy of Sleep Medicine scoring criteria.
86 Atlas of Sleep Medicine

Cursor: 23:50:19, Epoch: 172 . AWAKE 30 sec/page

2
LOC-A2
128 uV

ROC-A1
128 uV

Chin1-Chin2
1
273.1 uV

C3-A2
85.3 uV

C4-A1
85.3 uV

O1-A2
85.3 uV

O2-A1
85.3 uV

ECG2-ECG1
682.7 uV

ECG2-ECG3
682.7 uV

LtTib1-LtTib2
341.3 uV

RtTib1-RtTib2
341.3 uV

30 sec/page
SNORE

OroNs

THOR

ABD

PFLOW

Spo2

FIGURE 3.9  Transition of stage N2 sleep into wakefulness (30-second epoch).  This epoch depicts transition of stage N2 into stage
W. K complexes and sleep spindles are noted earlier in the record (1). An arousal (2) is 3 seconds or longer in duration, and the resulting alpha
electroencephalogram activity persists for more than 50% of the record. The epoch is therefore scored as stage W sleep.

in sleep, an apneic event, or in association with an arousal alpha activity persists for less than 50% of the record, then
(alpha EEG activity) immediately following the complex. the epoch is scored as stage N1 sleep. If the alpha activity
Sleep spindles (see Fig. 3.7), which are also termed persists for greater than 50% of the record, the epoch is
sigma waveforms, may appear here. They are generated scored as stage W. If the first half of the following epoch
in and controlled by activity within the midline thalamic demonstrates stage N2 characteristics (i.e., sleep spindles,
nuclei (reticular thalamic nucleus) and represent an inhibi- K complexes, high-amplitude theta/delta activity), then
tory activity. Sleep spindles are characterized by 12- to14- that epoch is scored as stage N2 sleep.
Hz sinusoidal EEG activity in the central vertex region Stage N2 sleep is associated with a relative diminution
and must persist for at least 0.5 seconds (i.e., six to seven of physiological bodily functions. Blood pressure, brain
small waves in 0.5 seconds), but are rarely longer than 1 metabolism, gastrointestinal secretions, and cardiac activity
second. Sleep spindles possess a high degree of synchrony decrease. The patient descends deeper into sleep, becoming
and symmetry between the two hemispheres in patients more and more detached from the outside world and pro-
older than 1 year. Although classically described as spindle gressively more difficult to arouse.
shaped, they may be polymorphic and may attach as a tail
to a K complex. Normal variant for scoring human sleep is
Stage N3 NREM Sleep (Fig. 3.10)
the appearance in stage N2 sleep of low K complex quan-
tity and high-amplitude spindle activity. Central nervous Stage N3 NREM sleep may also be termed deep sleep, slow
system (CNS) depressant drugs (such as benzodiazepines) wave sleep (SWS), or delta sleep. The new AASM stage
often increase the frequency of the spindle activity, whereas N3 includes R and K stages 3 and 4 together and does not
advancing age often diminishes their frequency. No specific make a distinction between them because such distinction
criteria exist for EOG and EMG in this stage. probably does not have clear clinical significance. SWS is
Arousal from stage N2 may default the scoring into stage marked by high-amplitude slow waves. No specific criteria
N1 or into wakefulness (W) as seen in Figure 3.9. If EEG for EOG and EMG exist for SWS, but in general, muscle
 CHAPTER 3  Sleep Stages and Scoring Technique 87

FIGURE 3.10  Stage N3 (30-second epoch).  Demonstrated in this figure is the highly synchronized low-frequency activity characterizing
stage N3 according to the new American Academy of Sleep Medicine guidelines.

tone is further decreased. SWS constitutes the deepest, women experience clitoral engorgement. The body seems to
most refreshing and restorative sleep type, which tends to have abandoned its effort to regulate its temperature during
diminish with age. the REM phase and resembles a state of poikilothermy, drifting
Physiologically a patient going through SWS has the gradually toward the temperature of the environment.
highest threshold for arousal. SWS may be associated with If patients are awakened from this stage of sleep, they
diffuse dreams (20% of dreams), and many parasomnias may often recall dreaming. Pathologically short REM sleep
(sleep terrors, sleepwalking) manifest themselves here. latency may point to a state of acute or cumulative sleep
Eye movements may cease altogether in this stage of sleep. deprivation, may be caused by abrupt discontinuation of
Physiologically SWS is often linked with a peak in growth REM sleep–suppressing agents (such as antidepressants),
hormone secretion. narcolepsy-cataplexy syndrome, and may also suggest a
The arousal threshold of this stage of sleep is far greater major affective disorder. A variety of sleep disorders are
than in stage N1 or N2 sleep. Both K complexes and sleep strongly associated with REM sleep, including a variety of
spindles may be seen in stage N3 sleep, but no specific cri- parasomnias (REM sleep behavior disorder, REM night-
teria exist for EOG and EMG. If the patient wakes up from mares) and obstructive sleep apnea, which may be more
slow wave sleep, he or she may appear confused or disori- pronounced during this sleep period.
ented. The patient may experience “sleep inertia” or “sleep
drunkenness,” seeming unable to function normally for sev- Scoring Stage R Sleep
eral minutes. The duration of sleep inertia depends on prior Epochs that display low-amplitude, mixed-frequency EEG,
sleep deprivation and CNS-active medications. low chin EMG tone, and rapid eye movements should be
scored as Stage R (see Figs. 3.11 and 3.12). Stage R should
continue to be scored for subsequent epochs even without
Stage REM Sleep (Figs. 3.11 to 3.18)
rapid eye movements, assuming the EEG shows low-ampli-
Stage R or REM sleep may also be termed paradoxical sleep tude, mixed-frequency activity without K complexes or sleep
or active sleep. REM sleep typically occurs about 90 to 120 spindles, and the chin EMG tone remains low (see Fig. 3.13).
minutes after sleep onset in adults. The first REM period is Rapid eye movements are conjugate, irregular, sharply
typically brief with subsequent REM periods becoming pro- peaked eye movements with an initial deflection lasting less
gressively longer and more robust. It typically occupies 20% than 500 milliseconds. This produces rapid conjugate eye
to 25% of the major period of sleep and is characterized movements that appear as out-of-phase EOG channel deflec-
by relatively low-amplitude, mixed-frequency EEG theta tions on the PSG (see Figs. 3.11 and 3.12). During REM sleep
waves, intermixed with some alpha waves, usually 1 to 2 Hz the eyes move rapidly under closed eyelids while dreaming
slower than wake (see Fig. 3.11). Brain waves are small and (see Fig. 3.14, A). Rapid eye movement can also be seen
irregular, with pronounced bursts of eye activity (rapid eye during wakefulness with the eyes open when a person is look-
movements), which are seen in the EOG leads. ing around his or her environment (see Fig. 3.14, B). This can
Unlike the progressive relaxation noted during the NREM usually be distinguished from rapid eye movements of stage
sleep stages N1, N2, and N3, physiological activity during R sleep because of the high EMG tone or video evidence of
REM sleep is significantly higher. Blood pressure and pulse the patient being awake (see Fig. 3.14, B). The EOG activity
rate may increase dramatically or may show intermittent is not needed to mark the start of a REM period.
fluctuations. Breathing becomes irregular, and brain oxygen Stage R sleep is characterized by sawtooth waves (see Fig.
consumption increases. Men exhibit penile erections, whereas 3.15). These are 2- to 6-Hz, sharply contoured triangular
88 Atlas of Sleep Medicine

FIGURE 3.11  Transitioning into stage rapid eye movement (REM) sleep: 30-second epoch demonstrating the sudden decrease of chin elec-
tromyogram tone (arrow, 1) and a change in electroencephalogram frequency from delta slowing of stage N3 sleep to low-amplitude, mixed-
frequency waves of stage R sleep (2). Rapid eye movements are also noted (3). Montage illustrated is according to the American Academy of
Sleep Medicine scoring system.

EEG patterns that are jagged in morphologic structure and T-REM sleep generally consists of low-voltage activated
evenly formed (see Fig. 3.15). These may occur serially for EEG and is characterized by a marked decrease in skeletal
a few seconds and are highest in amplitude over the central muscle EMG activity, without obvious EOG activity (see
leads. REM sleep may be preceded by a series of sawtooth Fig. 3.13). T-REM appears to be mediated by areas near the
waves, though they are not required to score stage R sleep. locus coeruleus.
Low chin EMG tone is noted during REM sleep (see
Figs. 3.11 to 3.13). Though typically lower than in other End Scoring Stage R Sleep
stages of sleep, baseline EMG activity in the chin should be Stage R scoring should continue until there is a clear change
no higher than in any other sleep stage. No specific ampli- to another sleep stage. Stage R ends if an epoch meets criteria
tude criteria currently exist for this determination. Short, for stage W or N3 sleep. In addition, an increase in chin
irregular bursts of EMG activity (less than 0.25 seconds), EMG tone with an absence of spindles or K complexes indi-
called transient muscle activity, may be seen in the chin, cates the end of stage R (see Fig. 3.17) and the beginning of
limb, EEG, or EOG leads, usually in association with rapid stage N1 if the epoch meets criteria for stage N1. An arousal
eye movements (see Fig. 3.15). In REM sleep behavior dis- or major body movement followed by slow eye movements
order, loss of normal REM sleep atonia is noted in patients, also indicates termination of stage R and beginning N1 (see
resulting in injuries and sleep disruption caused by dream Fig. 3.17). Slow eye movements are defined as having their
enactment behavior during sleep (see Fig. 3.16). initial deflection lasting less than 500 milliseconds. If the
REM sleep is sometimes divided into phasic (P) and tonic chin EMG tone returns to a low level and the epoch demon-
(T) components. P-REM sleep is characterized by phasic strates an absence of slow eye movements, this would char-
twitching in the EMG channel occurring concurrently with acterize the continuation of stage R (as seen in Fig. 3.18).
bursts of rapid eye movements (see Fig. 3.15). This activity If a K complex or sleep spindle appears in the first half of
has been suggestively correlated with dream content. The the epoch without any subsequent rapid eye movements,
phasic EMG twitchings in this stage are very short muscle the epoch will be scored as stage N2 even if chin EMG tone
twitches that may occur in the middle ear muscles, genio- remains low. If a sleep spindle or K complex appears in the
glossal muscle, and facial muscles and are associated with second half of the epoch, then the epoch would still be
increased penile and clitoral tumescence. scored R and the next epoch would be scored N2.
Text continued on p. 95
 CHAPTER 3  Sleep Stages and Scoring Technique 89

3 3 3

FIGURE 3.12  Stage rapid eye movement (REM) (60-second epoch) according to the new American Academy of Sleep Medicine guidelines.
The specific characteristics of REM sleep include low-amplitude, mixed-frequency electroencephalogram (1), low chin electromyogram tone
(2), and rapid eye movements (3).
90 Atlas of Sleep Medicine

FIGURE 3.13  This 30-second epoch followed a previous epoch of stage R sleep.  This epoch demonstrates tonic rapid eye movement
(REM) sleep, consisting of low-amplitude, mixed-frequency electroencephalogram (EEG) with no rapid eye movements in this epoch. This
would still be scored as stage R because the previous epoch contained rapid eye movements, the chin electromyogram tone is low, and the
EEG showed mixed frequency without spindles or K complexes. Montage illustrated is according to the American Academy of Sleep Medicine
scoring system.
 CHAPTER 3  Sleep Stages and Scoring Technique 91

LOC-A2
128 uV

ROC-A1 2
128 uV

Chin1-Chin2
68.3 uV

C3-A2
85.3 uV

C4-A1
1
85.3 uV

O1-A2
85.3 uV

O2-A1
85.3 uV

ECG2-ECG1
1.37 mV

LtTib1-LtTib2
170.7 uV 3
RtTib1-RtTib2
170.7 uV

30 sec/page
SNORE
512 uV
OroNs
146.6 uV

THOR
4.68 mV

ABD
1.02 mV

PFLOW
25 mV
Spo2100
A (%) 0

LOC
4
ROC

CHIN 5
F4-M1

C4-M1

O2-M1

EKG

L-LEG

R-LEG 6
SNORE

CPAP

PTAF

AIR-Flow

CHEST

ABD

(%)

B
FIGURE 3.14  Rapid eye movements: stage rapid eye movement (REM) sleep versus wake.  A, Stage REM sleep characterized by
relatively low-amplitude, mixed-frequency electroencephalogram theta waves (1), intermixed with alpha waves. The electro-oculogram leads
depict rapid eye movements that are paroxysmal, relatively sharply contoured, high-amplitude activity occurring in all eye leads simultane-
ously (2). Electromyogram (EMG) tone (3) should show the lowest tone in the record, but no specific amplitude or frequency criteria are in
place. Montage depicted is according to R and K scoring criteria (30-second epoch). B, Stage W also demonstrates rapid eye movements
(4) but has an elevated chin (5) and limb (6) EMG tone associated with it compared to during REM sleep as seen by the two arrows. Montage
illustrated is according to the American Academy of Sleep Medicine scoring system.
FIGURE 3.15  Thirty-second epoch of stage R sleep displaying sawtooth waves (*) and rapid eye movements (1), with transient muscle activity
in the chin electromyogram lead (arrows). Montage depicted is according to American Academy of Sleep Medicine scoring criteria.

FIGURE 3.16  This 30-second epoch has abnormally high chin electromyogram tone for stage R sleep (1). This REM sleep without atonia can
be seen in patients with REM sleep behavior disorder and should be reviewed clinically. Montage depicted is according to American Academy
of Sleep Medicine scoring criteria.
 CHAPTER 3  Sleep Stages and Scoring Technique 93

FIGURE 3.17  Stage R transitioning into stage N1 sleep.  The figure shows stage R with arousal ( ) demonstrating rapid (1) and slow
eye movement (2). Stage N1 is scored on next epoch if the background rhythm is theta. Montage depicted is according to American Academy
of Sleep Medicine scoring criteria.
94 Atlas of Sleep Medicine

FIGURE 3.18  Rapid eye movement (REM) sleep with increased chin electromyogram and arousal with return back to REM sleep. Montage
depicted is according to American Academy of Sleep Medicine scoring criteria.
 CHAPTER 3  Sleep Stages and Scoring Technique 95

FIGURE 3.19  This 30-second epoch contains a major body movement (beginning at  ) obscuring most of the electroencephalogram signal
(purple rectangle). There is a small portion during the beginning and end of the epoch (noted by the arrows) that is consistent with an alpha
rhythm in the occipital leads; hence this epoch would be scored stage W. Montage depicted is according to American Academy of Sleep
Medicine scoring criteria.

Major Body Movements Transitions Between Stages of Sleep

Major body movement is movement and muscle artifact To begin scoring stage N1 from stage W, the background EEG
obscuring the EEG for more than half an epoch, making rhythm (usually alpha frequency) seen maximal in the occip-
determination of sleep stage difficult. In the new AASM ital region needs to be replaced by low-amplitude, mixed-
scoring manual from 2007, if an epoch contains a major body frequency activity (usually theta frequency) for more than
movement, it can be scored as stage W if an alpha rhythm is half of the epoch. In patients who do not have an identifiable
present for any part of the epoch (even less than 15 seconds, alpha rhythm, stage N1 can commence with slowing of the
as seen in Fig. 3.19). If no alpha rhythm is noted, but stage wake background frequency by 1 Hz or more, presence of
W precedes or follows the epoch with a major body move- vertex sharp waves, or slow eye movements. Vertex sharp
ment, it can be scored as stage W. Lastly, if these situations waves, or V waves, are sharply contoured waves maximal over
are not noted, the epoch can be scored the same stage as the the central regions with duration less than 500 milliseconds.
epoch that follows it. To begin scoring stage N2 sleep, a K complex or sleep
Arousals can be scored during sleep stages if there is an spindle must be present in the first half of the epoch or the
abrupt change in the EEG frequency noted in the occipital second half of the previous epoch. Scoring of stage N2 can
or central derivations. This can be alpha, theta, or beta be continued even without the presence of K complexes or
frequency (with the exclusion of spindle frequency) that sleep spindles assuming the EEG continues to demonstrate
lasts at least 3 seconds with at least 10 seconds of stable low-amplitude, mixed theta frequency activity.
sleep preceding the change. During stage R, an arousal Stage N2 ends if the epoch meets criteria for stage W,
requires an increase in chin EMG lasting for at least 1 stage N3, or stage R. In addition, following an arousal from
second. Other recording channels such as limb EMG, stage N2 sleep, the epoch is scored stage N1 until a K com-
respiratory channels, and ECG may be used to help make plex or sleep spindle occurs again. Stage N2 ends after a
the determination of an arousal, but EEG criteria must major body movement followed by slow eye movements
be met. and absence of subsequent sleep spindles or K complexes.
96 Atlas of Sleep Medicine

FIGURE 3.20  Stage N2 with arousal ( ) and return to stage N2 caused by a K complex noted at the end of the epoch (♦). Montage depicted
is according to American Academy of Sleep Medicine scoring criteria.

Stage N2 continues to be scored after a body movement if chin EMG (i.e., the patient already has low EMG tone during
there are no slow eye movements following the major body stage N2), the rule provides for scoring of the epoch that
movement. Figure 3.20 demonstrates an arousal from Stage follows the last K complex or sleep spindle as stage R.
N2 sleep; however, in the same epoch a K complex is subse-
quently noted, keeping this epoch scored as N2. Figure 3.21
shows an arousal from stage N2, with a slow eye movement,
Conclusion
which then makes the subsequent epoch stage N1 given that The AASM scoring manual (which slightly modified the R and
there are no K complexes or spindles noted in the interim. K sleep stage scoring that previously was considered the gold
As state earlier, stage N3 is scored when more than 20% of standard for such scoring and expanded the scope of other
the epoch consists of slow wave activity. Slow wave activ- physiological events) provides a methodology to standardize
ity is defined as 0.5 to 2 Hz with an amplitude of 75 μV or sleep recording and should serve as the most critical initial
more measured over the frontal regions. stepping stone for all sleep medicine trainees and practitioners.
When sleep transitions between stage N2 and stage R, The following abbreviations are used in the sleep mon-
epochs between K complexes and sleep spindles or between tage recordings:
rapid eye movements can be difficult to distinguish. The LOC, left electro-oculogram
new scoring manual requires that the scorer “look back” to ROC, right electro-oculogram
determine correct scoring. The first epoch after a decrease Chin, chin electromyogram
in chin EMG should be scored as stage R even if there are ECG, electrocardiogram
no rapid eye movements provided that there are no sleep LtTib, left anterior tibialis surface electromyogram
spindles or K complexes in the epoch. If the record shows RtTib, right anterior tibialis surface electromyogram
sleep spindles or K complexes in the absence of rapid eye SNORE, snore sensor
movements, the epoch is scored as N2 sleep, following a OroNs, oronasal airflow
drop in chin EMG. Furthermore, an epoch with a K com- THOR, thoracic respiratory effort
plex or spindle is scored as stage R if either is followed by ABD, abdominal respiratory effort
a rapid eye movement while chin EMG tone is still low, as PFLOW, nasal pressure transducer
seen in Figures 3.22 and 3.23. Because there is no drop in Spo2, pulse oximetry
 CHAPTER 3  Sleep Stages and Scoring Technique 97

FIGURE 3.21  Stage N2 ending in arousal and slow eye movement (*), with subsequent scoring of N1. Montage depicted is according to
American Academy of Sleep Medicine scoring criteria.
98 Atlas of Sleep Medicine

FIGURE 3.22  This first of two epochs shows the termination of stage N2 with a decrease in chin electromyogram (arrow), rapid eye move-
ment, and low amplitude, mixed theta electroencephalogram frequency (dotted line). This epoch would be scored as stage R. Montage
depicted is according to American Academy of Sleep Medicine scoring criteria.
 CHAPTER 3  Sleep Stages and Scoring Technique 99

FIGURE 3.23  This is the subsequent epoch following Figure 3.22 showing a K complex ( ). This epoch is still scored as stage R sleep, and not
stage N2, because there are rapid eye movements (*) following the K complex and the chin electromyogram tone remains low (dotted line).
Montage depicted is according to American Academy of Sleep Medicine scoring criteria.

Pressman MR. Primer of Polysomnogram Interpretation. Boston: Butter-

Bibliography worth & Heineman; 2002.
Iber C, Ancoli-Israel S, Chesson AL, Quan SF. The AASM Manual for the Rechtschaffen A, Kales A, eds. A Manual of Standardised Terminology, Tech-
Scoring of Sleep and Associated Events. Westchester, Ill: American Academy niques and Scoring System for Sleep Stages of Human Subjects. Los Angeles:
of Sleep Medicine; 2007. Brain Information Service/Brain Research Institute; 1968.
Kryger MH. Atlas of Clinical Sleep Medicine. Philadelphia: Saunders/ Shepard JW, ed. Atlas of Sleep Medicine. Armonk, NY: Futura Publishing
Elsevier; 2010. Co; 1991.
Kryger MH, Roth T, Dement WC. Principles and Practice of Sleep Medicine. Silber MH, Ancoli-Israel S, Bonnet MH, et al. The visual scoring of sleep in
5th ed. Philadelphia: Saunders/Elsevier; 2011. adults. J Clin Sleep Med. 2007;3(2):121-129.