Effects Of Hypnagogic Imagery On The Event-Related

Effects of Hypnagogic Imagery on the Event-Related Potential to External Tone
Stimuli
Nanae Michida, MA; Mitsuo Hayashi; PhD Tadao Hori, PhD
Department of Behavioral Sciences, Faculty of Integrated Arts and Sciences, Hiroshima University, Higashi-Hiroshima, Japan
reaction times than 5000 milliseconds were analyzed. Electroencephalograms were recorded from Fz, Cz, Pz, Oz, T5 and T6. There were no
differences in reaction times and electroencephalogram spectra between
the conditions of with and without hypnagogic imagery. These results indicated that the arousal levels were not different between the 2 conditions.
On the other hand, the N550 amplitude of the event-related potentials in
the imagery condition was lower than in the no-imagery condition.
Conclusions: The decrease in the N550 amplitude in the imagery condition showed that experiences of hypnagogic imagery exert some influence on the information processes of external tone stimuli. It is possible
that the processing of hypnagogic imagery interferes with the processing
of external stimuli, lowering the sensitivity to external stimuli.
Keywords: Event-related potential, Hypnagogic imagery, N550
Citation: Michida N; Hayashi M; Hori T. Effects of hypnagogic imagery
on the event-related potential to external tone stimuli. SLEEP 2005;28(7):
813-818.
Study Objectives: The purpose of this study was to examine the influence of hypnagogic imagery on the information processes of external
tone stimuli during the sleep onset period with the use of event-related
potentials.
Design: Event-related potentials to tone stimuli were compared between
conditions with and without the experience of hypnagogic imagery. To
control the arousal level when the tone was presented, a certain criterion named the electroencephalogram stage was used. Stimuli were presented at electroencephalogram stage 4, which was characterized by the
appearance of a vertex sharp wave.
Setting: Data were collected in the sleep laboratory at Hiroshima University.
Participants: Eleven healthy university and graduate school students
participated in the study.
Interventions: N/A.
Measurements and Results: Experiments were performed at night. Reaction times to tone stimuli were measured, and only trials with shorter
is perceived. Components of event-related potentials (ERPs) are
useful to examine the information processing of external tone
stimuli because these reflect intracerebral information processes.
Concerning previous studies about ERPs during the sleep-onset
period, changes of ERPs influenced by declines in the arousal
level, variation of stimulus frequency, or differences of the task
load have been reported.6-14 However, there are no reports on the
relationship between hypnagogic imagery and ERPs. Comparison of ERPs to external tone stimuli between the conditions with
and without hypnagogic imagery makes it possible to estimate the
influence of hypnagogic imagery on the information processing
of tone stimuli.
It must be noted that the amplitude of the ERP components
changes with the fluctuation of the arousal level. According to
a review of ERPs during the sleep-onset period,14 amplitudes of
the negative wave peaking at 100 milliseconds (N100) and the
positive component peaking around 300 milliseconds (P300)
gradually decrease as the arousal level becomes lower. Apparent diminution of these 2 components is observed when the subject no longer signals detection of the external stimulus. On the
other hand, P200 may increase during the sleep-onset period. The
decrease in N100 and the increase in P200 have been explained
by the removal of a long-lasting processing negativity associated with attention and consciousness in the waking state. During late stage 1 sleep of the standard sleep criteria (provided by
Rechtschaffen & Kales),15 N300 begins to emerge. And finally, in
standard stage 2 sleep, which is regarded as definitive sleep, the
large N550-P900 components of the K-complex become clearly visible. Therefore, it is necessary to control the arousal level
when tone stimuli are presented to investigate differences in the
processing of tone stimuli between the presence and absence of
imagery. To objectively judge the arousal level during the sleep-
INTRODUCTION
IT IS KNOWN THAT SIMILAR PHENOMENA TO DREAMS
ARISE DURING THE PERIOD OF TRANSITION FROM
WAKEFULNESS TO NON-RAPID EYE MOVEMENT (NREM)
SLEEP. This dream-like phenomenon during the sleep-onset period is called hypnagogic imagery. The usual way experimental
studies regarding hypnagogic imagery are performed is to awaken
a dozing subject and discuss his or her mental experiences that
occurred before awakening. However, this method relies on the
introspection of the subjects. Therefore, a psychophysiologic
method is needed to more objectively study hypnagogic imagery.
Although there are several reports on the functional activity of the
brain immediately before waking with imagery during the sleeponset period using spectral analysis of electroencephalogram
(EEG) records,1-5 these studies do not show a common view corresponding to the presence of imagery. Therefore, it seems worthwhile for intracerebral and intercerebral information processing
of hypnagogic imagery to be investigated from another point of
view. It is thought that the processing of hypnagogic imagery
has an influence on other information processes when imagery
Disclosure Statement
This was not an industry supported study. Drs. Michida, Hayashi, and Hori
have indicated no financial conflicts of interest.
Submitted for publication December 2003
Accepted for publication February 2005
Address correspondence to: Tadao Hori PhD, Department of Behavioral Sciences, Faculty of Integrated Arts and Sciences, Hiroshima University, 1-7-1
Kagamiyama, Higashi-Hiroshima, 739-8521, Japan. Tel: +81-82-424-6580;
Fax: +81-82-424-0759. E-mail: [email protected]
SLEEP, Vol. 28, No. 7, 2005
813
Effects of Hypnagogic Imagery on the ERP—Michida et al
onset period, a certain criterion, named the EEG stage, has been
proposed.3,16,17,18 This criterion divides the sleep-onset period into
9 stages based on the typical appearance of EEG as declining with
the arousal level. Hori et al3 used a 5-stage simplified version of
this criterion (Figure 1) to analyze the occurrence rate of hypnagogic imagery for each EEG stage. According to their study, the
average recall rates of hypnagogic imagery for each EEG stage
showed an inverse U-shaped distribution. The recall rate at EEG
stage 3 was the highest, followed by stage 2 after stage 4. In this
study, EEG stage 4 of the Hori et al criteria3 was used because
there seemed to be a good possibility of obtaining both “with-imagery” and “without-imagery” samples.
The purpose of this study was to examine the influence of hypnagogic imagery on the information processing of external tone
stimuli with the use of ERPs. It has been hypothesized that ERPs
to external tone stimulus show some differences between the 2
cases of the presence and absence of hypnagogic imagery. It must
be noted that the fluctuation of the arousal level affects the ERP
amplitudes. Thus, EEG signals just before each tone stimulus
were subjected to spectral analysis to make sure that the arousal
levels during the stimulus-presentation period were not different
between the conditions of with and without hypnagogic imagery.
pended on the progress of sleep onset of the subjects, though the
minimum time between the stimuli was more than 30 seconds.
Procedure
Subjects were tested in a bed in a dark, electrically shielded,
2950×2800×2200 room. They were instructed to sleep if they became drowsy but to press the microswitch button taped to their
right hand as soon as possible whenever they heard tones. They
were also instructed to explain in detail what they experienced
immediately before pressing the button upon having a cue tone
signal in every trial. They were requested to use one of the following words to explain their experiences: “saw,” “heard,” “felt,”
“thought,” “forgot,” and “nothing.” These words were used to
categorize their mentations.
To control the arousal level when the pip-tone stimuli were presented, the 5 EEG-stage scoring system for the sleep-onset period3
was used. The EEG from Cz for 5 seconds just before the tone
stimuli was used for the determination of the EEG stage. Tone
stimuli were presented at EEG stage 4, which was characterized
by the appearance of a well-defined vertex sharp wave. Subjects
were awakened at each trial because they were required to report
their mentations. Thus, the process of falling asleep was repeated
again and again during the experiment. After the experiments, the
EEG stages of tone-presented trials were scored again, and the
trials that were confirmed as EEG stage 4 were used for further
analysis. If a K-complex was evoked by a tone stimulus, the trial
was not regarded as EEG stage 4 because the stage is defined as
a part of sleep stage 1 of the standard criteria (provided by Rechtschaffen & Kales).15 According to the standard sleep criteria,
the K-complex is one of the characteristics of sleep stage 2.
Experiments performed were excluded when the number of
tones sounded reached 200 or when the tone-presenting period
lasted longer than 2 hours. When the number of adoptable trials
for each condition did not reach 20 times, which was the minimum number of trials to determine ERPs, additional trials were
made on another night.
Reaction times to tone stimuli were measured with a microswitch button. Only those trials with a shorter reaction time than
5000 milliseconds were analyzed. Two seconds after the press of
the button, a tone stimulus (1,500 Hz, 60 dB SPL, 1 second) was
given to subjects as a cue to report their hypnagogic mentations.
Based on the report, each trial was divided into 2 categories for
analysis: imagery and no imagery. All trials of every experiment
were categorized by 1 experimenter. There were some trials that
were excluded from the analysis. It has been reported that visual
imagery is included in the majority of experiences in hypnagogic
imagery.3,19-21 Therefore, only trials with visual imagery were categorized as the imagery condition to control the homogeneity of
imagery experiences. Trials with auditory or kinesthetic imagery
without visual imagery were excluded from the analysis. When
subjects did not recall details of their mental experiences, the
trial was also excluded from analysis. Meanwhile, trials with no
mental experience and logical or illogical thought without clear
imagery were categorized as the no-imagery condition.
METHODS
Subjects
Eleven university and graduate school students (3 men and 8
women) between the ages of 18 and 25 years (mean=21.3 years)
participated in this study. They were screened for health and sleep
problems. All of them were right handed and had some experience
with sleep experiments. Informed consent was obtained.
Stimuli
Pip-tone stimuli (1,000 Hz, 60 dB SPL, 50 milliseconds) were
presented from a speaker set up 1.4 meters above a pillow. The
stimuli were manually presented by the experimenter under the
condition as follows. Each tone stimulus was produced within 5
seconds of the appearance of a well-defined vertex sharp wave in
EEG from Cz that appeared more than 30 seconds after the preceding stimulus. Therefore, the length of the stimulus interval de-
EEG Recording
Figure 1—Representative electroencephalogram (EEG) patterns for 5
electroencephalographic stages and correspondence to standard sleep
stages.
SLEEP, Vol. 28, No. 7, 2005
EEGs were recorded from Fz, Cz, Pz, Oz, T5, and T6 with AgAgCl surface electrodes referred to linked earlobes with the forehead as the ground. Electrooculogram recordings were obtained
814
Effects of Hypnagogic Imagery on the ERP—Michida et al
from electrodes affixed at the corner of both eyes referred to the
ipsilateral earlobes for horizontal eye movements and attached to
the supraorbital and infraorbital positions of the left eye, referred
to each other for vertical eye movement. A time constant of 3.2
seconds was used for both EEG and electrooculogram recordings.
An electromyogram was recorded from surface electrodes placed
on the chin (mentalis), referred to each other. The time constant
for EMG was 0.03 seconds. In all recordings, 120-Hz high cut-off
filtering was used and interelectrode impedances were kept below
5 KΩ. All electrophysiologic parameters were simultaneously recorded on an 18-channel polygraph (Model 1A97, NEC San-Ei,
Tokyo, Japan) and a 14-channel FM tape recorder (SR-50, TEAC,
Tokyo, Japan).
Statistical Analysis
A 2-tailed paired t test was used for comparing the reaction
times between the 2 conditions: imagery and no imagery. Mean
amplitudes of EEG frequency bands and peak amplitudes of ERP
components were analyzed using a 2-way repeated measures
analysis of variance. The independent variables were condition
and electrode site. Significant levels were determined following
the adjusted Greenhouse-Geisser approach for repeated observations.22 As for the posthoc test, all comparisons between individual cell means were performed using the Newman-Keuls procedure. All statistical significances of comparisons were based on a
.05 level of confidence.
RESULTS
EEG Spectrum Analysis
Rate of Imagery and Reaction Time
A signal processor (Model 7T18A, NEC San-Ei) was used for
the subsequent off-line analysis of EEG data reproduced by the
FM recorder. EEGs were recorded during a 5000-milliseconds
epoch prior to each stimulus tone. Autopower spectra with a 0.2Hz resolution were computed using a fast Fourier transform and
smoothed with a Hanning window. Spectra that averaged more
than 20 epochs, which corresponded to trials used for ERP calculation, were respectively computed for each condition. The average amplitudes (root power) for EEG frequency bands, delta (2.43.4 Hz) theta (3.6-7.4 Hz), alpha (7.6-13.4 Hz), beta (13.6-20.0
Hz) and sigma (13.0-14.8 Hz) were computed from each EEG
channel.
The average hit rate to tone stimuli was 96.9% (SEM=1.60%).
Only trials with a behavioral response to tone were analyzed
after this. The average frequencies and rates of hypnagogic experiences across subjects were as follows: Imagery: 65.5 trials (SEM=15.90), 54.1%; no imagery: 38.6 trials (SEM=6.84),
38.2%; others: 8.0 trials (SEM=1.65), 7.7%. Mean reaction times
were not significantly different between the 2 conditions (imagery: 891.6 milliseconds, SEM=118.62 milliseconds; no imagery:
905.1 milliseconds, SEM=116.88 milliseconds).
FFT Analysis
Figure 2 shows the EEG spectra in the last 5000 milliseconds
before the tone. The spectra of both conditions almost overlap
each other at every site. Average amplitudes of each EEG frequency band were statistically analyzed. For all frequency bands,
only the main effects of the electrode sites were significant, but
significant differences between the conditions were not seen at
every site.
ERP Analysis
Averaging over 20 times for ERP analysis was performed on
each condition for EEG signals from 200 milliseconds before to
1,500 milliseconds after each stimulus tone at a 5-millisecond
sampling rate. Peak amplitudes of the ERP components were
measured from the baseline, which was determined as the mean
voltage during the last 200 milliseconds prior to the tone. The
peak latency windows of ERP were defined as follows: N100: 75150 milliseconds, P200: 150-250 milliseconds, N300: 200-400
milliseconds, P400: 300-500 milliseconds, N550: 400-750 milliseconds, and P900: 750-1200 milliseconds.
ERP Analysis
Figure 3 presents the grand average ERPs across all subjects
at each site. Clear peaks of the P900 component did not appear
at every site, although a slow positive potential, which followed
N550 and did not return to the baseline, was observed at Fz and
Table 1—Peak Amplitudes of Event-Related Potential Components
Fz
N100
P200
N300
P400
N550
Imagery
-4.38
(1.18)
11.08
(2.37)
-20.22
(3.50)
-15.08
(5.23)
-21.58
(5.67)
Noimagery
-3.68
(1.14)
11.96
(1.71)
-20.70
(3.93)
-20.21
5.11)
-27.88*
(5.48)
Cz
Imagery
-6.69
(1.66)
13.70
(2.09)
-22.13
(3.89)
-2.28
(5.92)
-7.62
(5.51)
Noimagery
-5.54
(1.33)
15.64
(1.93)
-23.12
(4.50)
-5.91
(5.08)
-13.17*
(4.99)
Pz
Imagery
-4.90
(0.96)
12.65
(1.48)
-10.41
(3.55)
15.25
(4.60)
-3.69
(3.61)
Oz
Noimagery
-4.02
(1.28)
13.96
(1.60)
-11.61
(3.86)
10.78
(3.49)
-7.35*
(2.73)
Imagery
-2.31
(0.59)
6.71
(1.80)
-0.91
(1.87)
18.57
(2.87)
-10.15
(2.80)
Noimagery
-1.26
(0.88)
7.52
(1.51)
-1.68
(1.92)
16.44
(2.05)
-10.57
(1.96)
T5
NoImagery imagery
-1.13
-0.90
(0.40)
(0.44)
3.35
3.71
(0.99)
(0.68)
-2.36
-3.28
(1.29)
(1.13)
10.40
8.00
(2.39)
(1.97)
-6.94
-8.62
(2.05)
(1.43)
T6
NoImagery imagery
-0.98
-0.69
(0.58)
(0.87)
2.79
3.91
(1.07)
(1.15)
-3.64
-3.61
(1.66)
(1.38)
9.95
8.26
(2.17)
(1.93)
-9.41
-10.19
(2.01)
(1.81)
Data are presented as mean (SEM) in 11 subjects.
*Statistically significant at P<.05.
SLEEP, Vol. 28, No. 7, 2005
815
Effects of Hypnagogic Imagery on the ERP—Michida et al
Figure 3—Total averages of event-related potentials across all subjects. Black lines are the event-related potentials for the imagery condition. Gray lines are the event-related potentials for the no-imagery
condition.
onset period vary their amplitudes with the decline in arousal
level. To put it concretely, N100 and P300 decrease in amplitude
while P200 and N300 increase throughout the transitional period
from arousal to definitive sleep. If experiences of hypnagogic
imagery are associated with arousal level, it would be expected
that each ERP component in the imagery condition would show a
distinctive feature that reflects the difference of the arousal level.
ERP components preceding N550 indicated that the arousal level
did not differ between the with- and without-imagery conditions.
This was supported by the results of the behavioral reaction time
and fast Fourier transform analysis mentioned above.
On the other hand, N550 amplitudes in the imagery condition
were observed to be smaller than those in the no-imagery condition at Fz, Cz and Pz. The differences in ERPs between the
with- and without-imagery conditions showed that experiences of
hypnagogic imagery during the sleep-onset period exerted some
influence on the information processes of external tone stimuli.
The N550 amplitude was the largest at Fz as a main effect of the
site. That is, the difference of the N550 amplitude between the
conditions appeared where N550 was dominant. Using a loudspeaker as the stimulus-input medium might have produced a
less-constant tone to the subject’s ear than using earphones. However, the difference between these 2 conditions appeared only in
1 component in spite of the emergence of various components in
the ERPs. Therefore, the difference of N550 can be explained by
the influence of imagery rather than by the position of the ear.
N550 is regarded as one of the major components of the Kcomplex, along with an earlier N300 and a later P900.6,13,23-35 The
K-complex occurs in the non-rapid eye movement sleep state at
standard sleep stage 2 or deeper. EEG stage 4, at which the tone
was presented in this study, is included in standard sleep stage 1.
In fact, the detection rate of the stimuli was more than 96%. This
provided evidence that EEG stage 4 trials in this study did not
enter standard sleep stage 2, regarded as definitive sleep. Therefore, the N550 of this study is probably not a component of the
K-complex, or at least the K-complex defined as the negativepositive complex that can only be elicited in definitive sleep. This
EEG stage 4 should not show the K-complex; however, it was
reported that the appearance of N550 was already observed at
EEG stage 2 of the simplified 5-stages criterion and then grew in
proportion as the arousal level declined.7,8 According to Peszka
Figure 2—Electroencephalogram spectra averaged across all subjects. Black lines are the spectra for the imagery condition. Gray lines
are the spectra for the no-imagery condition.
Cz. Table 1 shows the mean peak amplitudes of all subjects. Amplitudes of 5 components except P900 were statistically analyzed.
All components, with the exception of N550, only showed the
significant main effect of the sites but no main effect for the conditions or interactive effect. For the N550 amplitudes, both main
effects of the sites and conditions were found to be significant. A
significant effect of conditions (F1,10=9.74, P<.05) indicated that
the amplitude in the imagery condition (-9.90µV) was smaller
than that in the no-imagery condition (-12.96µV). A significant
interaction effect was also observed (F5,50=4.20, ε=0.284, P<.05).
Newman-Keuls comparisons between conditions showed significant differences at Fz, Cz, and Pz. At these sites, N550 amplitudes
in the imagery condition (Fz: -21.58, Cz: -7.62, Pz: -3.69µV)
were smaller than those in the no-imagery condition (Fz: -27.88,
Cz: -13.17, Pz: -7.35 µV).
DISCUSSION
Reaction Time and FFT Analysis
The behavioral reaction time and the EEG power spectrum are
well known as indicators of fluctuations of the arousal level. In
this study, significant differences in the reaction times between
the with- and without-imagery conditions were not seen. On the
other hand, the mean amplitudes of all frequency bands showed a
significant main effect only of the sites but no main effect of the
conditions or interactive effect. It follows from the results of reaction time and fast Fourier transform analysis that arousal levels of
behavior and EEG did not differ between the 2 conditions (with
and without imagery).
ERP Analysis
All observed components except N550 did not show significant
differences between the conditions of with and without imagery.
According to previous studies, ERP components during the sleepSLEEP, Vol. 28, No. 7, 2005
816
Effects of Hypnagogic Imagery on the ERP—Michida et al
cognitive activity. Sallinen et al41 suggested that the threshold of
perceiving external stimuli should be higher during phasic than
tonic rapid eye movement sleep because of immersion in internal
mental activity (dreaming).
The function of hypnagogic imagery is unknown but very interesting. From the difference in N550 amplitude according to
the condition, it can be said that the sleep-protective mechanism
against the arousal associated with the processing of tone stimuli
was weakened when hypnagogic imagery was experienced. However, the behavioral reaction times in these 2 conditions were not
different. That is, the tolerance to the disturbance of sleep while
experiencing hypnagogic imagery did not differ from that without imagery in spite of a lower function of the sleep-protective
mechanism reflected in the N550. This led to the conjecture that
an allocation of cognitive-processing resources to hypnagogic
imagery during the sleep-onset period should support a smooth
transition from wakefulness to sleep. If hypnagogic imagery has
the role of sleep protection from arousal associated with the processing of external stimuli, it is no wonder that the sleep-protective mechanism reflected by the N550 was more evident in the
no-imagery condition to compensate for the role of imagery.
and Harsh,36 N550 was evident in waking EEG state trials, sleep
stage 1 trials and without K-complex trials in sleep stage 2. They
stated that the N550 and K-complex may be closely related but
are separable phenomena. These previous studies indicated that
the appearance of N550 in this study is possible. On the other
hand, Colrain et al.13 did not report the appearance of N550 in
sleep stage 1, which was divided into alpha and theta divisions,
although their stage 1-theta division was similar to this study in
respect to the vertex sharp wave emergence. This contradiction
of the N550 emergence in sleep stage 1 can be explained by the
different ways of arousal level division. Sleep stage 1 is a transition period from wakefulness to sleep and its state rapidly varies.
For example, stage 1-theta of the Colrain group surely included
stage 3 and 4 of the simplified 5-stage EEG criterion. EEG stage
4 is even later sleep stage 1, just prior to stage 2, rather than EEG
stage 3. Another explanation is the differences in the properties
of the stimulus presentations. The tone is probably more salient
in the single tone paradigm, in which the stimulus frequency may
be lower, than in the odd-ball paradigm. It is still unclear whether
the N550 at EEG stage 4 of this study (in standard sleep stage 1)
is related to the K-complex or not, and this requires further investigation.
What brought about the difference in the N550 amplitude between the two conditions should be considered. According to recent studies, delta frequency large waveforms, such as the N550
and K-complex, is a phasic manifestation of slow wave sleep
pressure and reflects the “sleep protective” mechanism.35-38 This
theory, that the underlying mechanism facilitates the initiation
and maintenance of sleep, can be applied to explain why N550
increased accompanied by the progress of sleep in previous studies.7,8,9,39 Concerning the results of the present study, it was interpreted that the arousal associated with the processing of tone
stimuli in the no-imagery condition was blunter than in the imagery condition. In the imagery condition, the sleep protective
mechanism was weaker than in the no-imagery condition.
According to previous studies of auditory ERP in non-rapid
eye movement sleep, the N550 amplitude becomes larger when
the frequency of stimuli presentation is lower,13,23-25,39 when deviance from standard stimuli is larger,25-27 and when the button
press task of reacting to stimuli is set.24,39 These stimuli can be rephrased as physically and/or psychologically salient stimuli. That
is, the sleep-protective mechanism reflected in N550 is strengthened when external stimuli are salient to facilitate the initiation
and maintenance of sleep. In this study, the N550 reflecting this
mechanism was smaller in the imagery than in the no-imagery
condition; therefore, it can be considered that the same tone was
less prominent when imagery was perceived. If so, the reason why
the sensitivity to external tone stimuli was lower in the imagery
condition than in the no-imagery condition should be clarified.
Although the relationship between internal mental activity and
physiologic sleep-protective mechanisms such as N550 has not
yet been sufficiently clarified in this study, the possibility that internal processes related to the experience of hypnagogic imagery
attenuate the sensitive response to external stimuli was observed.
It can be considered that a part of the resources for cognitive processes at EEG stage 4 is allocated to hypnagogic imagery, resulting in the allocation of processing resources to the tone stimuli
being diminished. This idea is neither unusual nor unexpected.
Burton et al40 indicated that the reduced responsiveness to external stimulation during sleep is, at least in part, due to ongoing
SLEEP, Vol. 28, No. 7, 2005
CONCLUSION
In this study, we investigated how hypnagogic imagery influenced the information processes of external tone stimuli in the
sleep-onset period. The behavioral results of button pressing did
not differ between the with- or without-imagery conditions. Also,
there was no difference in the EEG spectral indexes between
whether or not hypnagogic imagery was experienced. Based on
these behavioral and physiologic results, it can be said with fair
certainty that there was no difference in the arousal level during
the periods when stimuli were presented between the 2 conditions—with hypnagogic imagery and without hypnagogic imagery.
On the other hand, in ERP analysis, the N550 amplitudes at Fz,
Cz, and Pz in the imagery condition were lower than in the noimagery condition. The results show that experiences of hypnagogic imagery during the sleep-onset period exert some influence
on information processes to external tone stimuli. It is possible
that the underlying sleep-protective mechanism is weaker when
hypnagogic imagery is experienced than when not. It was suggested that processing of hypnagogic imagery may interfere with
the processing of external stimuli and facilitate the initiation and
maintenance of sleep.
REFERENCES
1.
2.
3.
4.
5.
817
Germain A, Nielsen TA. EEG power associated with early sleep onset images differing in sensory content. Sleep Res Online 2001;4:8390.
Grass P, Lehmann D, Meier B, Meier CA, Pal I. Sleep onset: Factorization and correlations of spectral EEG parameters and mentation
rating parameters. Sleep Res 1987;16:231.
Hori T, Hayashi M, Morikawa T. Topographical EEG changes and
the hypnagogic experience. In: Ogilvie RD, Harsh JR, eds. Sleep
onset: normal and abnormal processes. Washington, DC: American
Psychological Association, 1994:237-53.
Lehmann D, Meier B, Meier CA, Mita T, Skrandies W. Sleep
onset mentation related to short epoch EEG spectra. Sleep Res
1983;12:180.
Lehmann D, Meier B, Meier CA, Mita T, Brandeis D. Sleep onset
Effects of Hypnagogic Imagery on the ERP—Michida et al
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
mentation characteristics related to lateralized EEG spectral power.
Sleep Res 1988;17:105.
Harsh J, Voss U, Hull J, Schrepfer S, Badia P. ERP and behavioral changes during the wake/sleep transition. Psychophysiology
1994;31:244-52.
Hori T, Ushio Y, Hayashi M. Event-related potential changes at
sleep onset. Psychiatr Clin Neurosci 1995;49:S45.
Hori T, Abe A, Hayashi M. Spatial and temporal behavior of ERP in
the sleep onset period. In: Book of Abstracts, Second Congress of
Asian Sleep Research Society, Jerusalem, Israel. 1997:18.
Ogilvie RD, Simons IA, Kuderian RH, MacDonald T, Rustenburg J.
Behavioral, event-related potential, and EEG/FFT changes at sleep
onset. Psychophysiology 1991;28:54-64.
Winter O, Kok A, Kenemans JL, Elton M. Auditory event-related
potentials to deviant stimuli during drowsiness and stage 2 sleep.
Electroencephalogr Clin Neurophysiol 1995;96:398-412.
Ogilvie RD. The process of falling asleep. Sleep Med Rev
2001;5:247-70.
Cote KA, de Lugt DR, Campbell KB. Changes in the scalp topography of event- related potentials and behavioral responses during the
sleep onset period. Psychophysiology 2002;39:29-37.
Colrain IM, Di Parsia P, Gora J. The impact of prestimulus EEG
frequency on auditory evoked potentials during sleep onset. Can J
Exper Psychol 2000;54:243-54.
Campbell KB, Colrain IM. Event-related potential measures of the
inhibition of information processing: the sleep onset period. Int J
Psychophysiol 2002;46:197-214.
Rechtschaffen A, Kales A. A Manual of Standardized Terminology,
Techniques and Scoring System for Sleep Stages of Human Subjects. Washington, DC: Public Health Service, U.S. Government
Printing Office; 1968.
Hori T, Hayashi M, Kato K. Changes of EEG patterns and reaction
time during hypnagogic state. Sleep Research 1991;20:20.
Tanaka H, Hayashi M, Hori T. Statistical features of hypnagogic
EEG measured by a new scoring system. Sleep 1996;19:731-8.
Tanaka H, Hayashi M, Hori T. Topographical characteristics and
principal component structure of the hypnagogic EEG. Sleep
1997;20:523-34.
Foulkes D, Vogel G. Mental activity at sleep onset. J Abnorm Psychol 1965;70:231-43.
Gibson E, Perry F, Redington D, Kamiya, J. Discrimination of sleep
onset stages: Behavioral responses and verbal reports. Percept Mot
Skill 1982;55:1023-37.
Vogel G.W. Sleep-onset mentation. In: Ellman SJ, Antrobus JS,
eds. The Mind in Sleep, 2nd ed. New York: John Wiley & Sons;
1991:125-37.
Winer BJ, Brown DR, Michels KM. Statistical Principles in Experimental Design, 3rd ed. New York: McGraw-Hill; 1991.
Colrain IM, Webster KE, Hirst G. The N550 component of the
evoked K-complex: A modality non-specific response? J Sleep Res
1999;8:273-80.
Niiyama Y, Fushimi M, Sekine A, Hishikawa Y. K-complex evoked
in NREM sleep is accompanied by a slow negative potential related to cognitive process. Electroencephalogr Clin Neurophysiol
1995;95:27-33.
Salisbury DF. Stimulus processing awake and asleep: Similarities
and differences in electrical CNS responses. In: Ogilvie RD, Harsh
JR, eds. Sleep Onset: Normal and Abnormal Process. Washington,
DC: American Psychological Association, 1994:289-308.
Salisbury D, Squires NK. Response properties of long-latency
event-related potentials evoked during NREM sleep. J Sleep Res
1993;2:232-40.
Salisbury D, Squires NK, Ibel S, Maloney T. Auditory event-related potentials during stage 2 NREM sleep in humans. J Sleep Res
1992;1:251-7.
Bastien C, Campbell K. The evoked K-complex: All-or-none phenomenon? Sleep 1992;15:236-45.
SLEEP, Vol. 28, No. 7, 2005
29. Bastien C, Campbell K. Effects of rate of tone-pip stimulation on
the evoked K-complex. J Sleep Res 1994;3:65-72.
30. Campbell K, Bell I, Bastien C. Evoked potential measures of information processing during natural sleep. In: Broughton RJ, Ogilvie
RD, eds. Sleep, Arousal, and Performance. Boston: Birkhäuser;
1992:88-116.
31. Campbell K, Bell I, Deacon-Elliott D. Stimulus-related influences
on the evoked K-complex. In: Koella WP, Rüther E, Schulz H, eds.
Sleep ‘84. Stuttgart: Gustav Fischer Verlag; 1985:235-7.
32. Campbell K, Rouillard L, Bastien C. Component structure of the
evoked K-complex. In: Horne J, ed. Sleep ‘90. Bochum: Pontenagel
Press; 1990:17-9.
33. Ujszászi J, Halász P. Late component variants of single auditory
evoked responses during NREM sleep stage 2 in man. Electroencephalogr Clin Neurophysiol 1986;64:260-8.
34. Nielsen-Bohlman L, Knight RT, Woods DL, Woodward K. Differential auditory processing continues during sleep. Electroencephalogr Clin Neurophysiol 1991;79:281-90.
35. Bastien CH, Crowley KE, Colrain IM. Evoked potential components unique to non-REM sleep: relationship to evoked K-complexes and vertex sharp waves. Int J Psychophysiol 2002;46:257-74.
36. Peszka J, Harsh J. Effect of sleep deprivation on NREM sleep ERPs
and related activity at sleep onset. Int J Psychophysiol 2002;46:27586.
37. Wauquier A, Aloe L, Declerck A. K-complex: are they signs of
arousal or sleep protective? J Sleep Res 1995;4:138-43.
38. Amzica F, Steriade M. The functional significance of K-complexes.
Sleep Med Rev 2002;6:139-49.
39. Kogure T, Hayakawa T, Nishio M, Tomiyama M, Urata J, Enomoto
T. Mechanism of attention in sleep: the changes of wave forms of
event-related potentials in sleep. Psychiatr Neurol 1992;94:553-7.
40. Burton SA, Harsh JR, Badia P. Cognitive activity in sleep and responsiveness to external stimuli. Sleep 1988;11:61-8.
41. Sallinen M, Kaartinen J, Lyytinen H. Processing of auditory stimuli
during tonic and phasic periods of REM sleep as revealed by eventrelated brain potentials. J Sleep Res 1996;5:220-8.
818
Effects of Hypnagogic Imagery on the ERP—Michida et al