INSTRUMENTATION AND METHODOLOGY
The Maintenance of Wakefulness Test in Normal Healthy Subjects
Siobhan Banks, BA (Hons)1,2; Maree Barnes, MBBS3; Natalie Tarquinio, MA3; Robert J. Pierce, MD3; Leon C. Lack PhD1,4; R. Doug McEvoy, MD1,2
1Adelaide Institute for Sleep Health, Repatriation General Hospital, Daw Park, SA; 2School of Medicine, Flinders University of South Australia,
Bedford Park, SA; 3Institute for Breathing and Sleep, Austin Health, Heidelberg, Vic; 4School of Psychology, Flinders University of South Australia,
Bedford Park, SA
Study Objectives: The Maintenance of Wakefulness Test (MWT) examines an individual’s ability to stay awake in an environment of decreased
sensory stimulation. Only 1 previous study has systematically examined
the MWT in normal healthy subjects.
Setting: Sleep disorders unit laboratory
Participants and Design: 31 subjects (mean age 48.5 years, SD 9.6; 8
women) were randomly selected via the telephone directory within a 30km radius of the test centers. They answered a general screen for health
complaints (respiratory, cardiovascular, and psychiatric disorders) and
sleep difficulties (snoring).
Interventions: N/A.
Measurements and Results: Overnight polysomnography and a 40minute MWT the following day were performed on all subjects. Mean
sleep latency to the first epoch of unequivocal sleep during the 40-minute
trial MWT was 36.9 ± 5.4 (SD) minutes. The lower normal limit, defined as
2 SD below the mean, was therefore 26.1 minutes. Mean sleep latency for
the first 20 minutes of the trial (with sleep latency defined as time to the
first appearance of 1 epoch of stage 1 sleep or a 10-second microsleep)
was 18.6 ± 2.3 minutes, with a lower normal limit of 14.0 minutes.
Conclusions: The mean results are consistent with previously published
normative data. However, the SDs found in this study are smaller, and,
thus, the lower normal limit suggested here is 4 to 6 minutes longer. The
subjects in this study were randomly selected from the general population
and may, therefore, be a truer representation of the normal population
than in the previous study in which subjects were recruited via hospital
advertisements and word of mouth.
Key Words: Maintenance of Wakefulness Test, normative data, daytime
sleepiness
Citation: Banks S; Barnes M; Tarquinio N et al. The maintenance of
wakefulness test in normal healthy subjects. SLEEP 2004;27(4):799-802.
a 40-minute trial and less than 10.9 minutes in a 20-minute trial.6
A drawback of this study was the way in which the subjects were
recruited. They were primarily advertisement respondents and
hospital employees. While the subjects were carefully screened
for good health, a potential sampling bias exists. They may have
volunteered to participate due to a concern about daytime sleepiness or a sleep disorder. The aim of this study was to add to the
existing normative data on the MWT sleep latency by randomly
selecting healthy subjects from the general community, naïve to
the objective of the study.
INTRODUCTION
THE MAINTENANCE OF WAKEFULNESS TEST (MWT)
WAS FIRST INTRODUCED AS A VARIANT OF THE
MULTIPLE SLEEP LATENCY TEST, AND ITS USE AS A
CLINICAL AND RESEARCH TOOL HAS GROWN CONSIDERABLY IN RECENT YEARS. The MWT protocol is similar to
that of the Multiple Sleep Latency Test; however, subjects are
required to stay awake without using any excessive mental or
physical methods while sitting upright in a darkened room.
Originally, the MWT used 4 evenly spaced 20-minute trials.1 In
subsequent investigation of patients with obstructive sleep apnea
(OSA), the MWT trials were lengthened to 40 minutes to try to
eliminate the observed ceiling effect.2 Both trial lengths are currently used, the particular length depending on the subjects’
degree of clinically reported sleepiness.
Most research has focused on MWT scores in patients with
OSA3 and narcolepsy.4 The effects of treatment on MWT mean
sleep latency is also an area of increased research.5 While the
MWT is used increasingly in clinical settings, there is a relative
lack of normative data. Only 1 published study has comprehensively examined MWT mean sleep latency in 64 normal subjects.6 The authors suggest that an abnormal MWT result (< 2 SD
from the mean) is a mean sleep latency less than 19.4 minutes in
METHODS
This study was approved by the relevant research and ethics
committees at the Repatriation General Hospital - Daw Park,
Adelaide, and Austin Health, Melbourne. All subjects gave written informed consent.
Participants
Forty-one subjects were recruited (17 at Austin Health,
Melbourne, and 24 at the Repatriation General Hospital - Daw
Park, Adelaide) to serve as a control group for a trial investigating the effects of mild to moderate OSA on neurocognitive function and the relative treatment effects of continuous positive airway pressure and a mandibular-advancement splint in the group
with OSA.
Subjects were recruited using a computer program to randomly select telephone numbers from the telephone book, within a
30-km radius of each hospital. An assistant then called these
numbers and, using a set script, invited subjects to participate in
the trial. This script was specifically designed for the study and
consisted of a brief introduction and then a general health screen.
In the introduction, the assistant began by explaining that the
study was examining the relationship between cardiovascular
Disclosure Statement
This study was supported by a National Health and Medical Research Council
Grant.
Submitted for publication August 2003
Accepted for publication January 2004
Address correspondence to: Siobhan Banks, Adelaide Institute for Sleep Health,
Repatriation General Hospital - Daw Park, Daws Rd, South Australia, Australia,
5041; Tel: 61 8 8275 1149; Fax: 61 8 277 6890;
E-mail: Siobhan.Banks@rgh.sa.gov.au
SLEEP, Vol. 27, No. 4, 2004
799
The MWT in Normal Healthy Subjects—Banks et al
Ambient temperature in the room was approximately 22oC.
Bedroom doors were closed, and all monitoring was performed
external to the bedroom to keep noise to a minimum. During each
MWT trial, subjects sat semi-upright (10o to 30o back from vertical) in a comfortable lounge chair, which had a high back to support the head and neck. Prior to each trial, subjects were instructed to “keep your eyes open and try not to fall asleep.” Subjects
were asked not to use any extraordinary mental or physical measures (eg, face slapping) to avoid sleep. The recordings were then
started, and the lights dimmed to an illumination of 1 lux. Each
trial was terminated at the first occurrence of sustained sleep (3
consecutive 30-second epochs of stage 1 sleep or 1 epoch of any
other stage) or after 40 minutes if there was no sleep. The 4 sleep
latencies were then averaged over the day to obtain the mean
MWT sleep latency.
measures, memory, IQ, and problem solving (these were all measures taken in examination of the effectiveness of treatment for
the patients with mild to moderate OSA). No mention was made
that the study was investigating sleep, sleep apnea, or daytime
sleepiness. It was explained that participants would need to come
to the hospital for 1.5 days and 1 overnight stay. It was also
explained that the benefit for participating in this study was a
comprehensive health check but no remuneration. Potential subjects then answered a general screen for health complaints (respiratory, cardiovascular, and psychiatric disorders) and sleep difficulties (snoring). Volunteers were excluded from the study if
they were not between the ages of 30 and 70 years; if English was
not their first language; if they were heard to snore more than 1
night per week; if they were on any medications for respiratory,
cardiovascular, or psychiatric disorders; or if they had a history
of drug or alcohol abuse.
Data Analysis
Procedure
In order to compare our data with previously published normative data, we analyzed the data using the same 4 sleep-onset criteria. Sleep latency was calculated separately for the first 20 minutes of recording and for the whole 40-minute recording period.
All MWT trials for each subject were analyzed according to these
4 protocols by 1 trained technician at the Adelaide laboratory.
For the 20-minute period—Sleep onset was defined as (1) the
first appearance of a 10-second burst of theta or the first epoch of
sleep (MICROMWT20) and (2) the first appearance of 3 epochs
of stage 1 sleep or 1 epoch of any other sleep stage (MWT20).
Subjects with no sleep onset within 20 minutes were assigned a
value of 20 minutes.
For the 40-minute period—Sleep onset was defined as (1) the
first appearance of a 10-second burst of theta or the first epoch of
sleep (MICROMWT40) and (2) the first appearance of 3 epochs
of stage 1 sleep or 1 epoch of any other sleep stage (MWT40).
Subjects with no sleep onset were assigned a value of 40 minutes.
Overnight Polysomnography
Subjects were required to keep a sleep diary for a week prior to
the testing and to maintain their typical sleep-wake routine for
that week. They arrived at the laboratory between 8 PM and 9 PM
on the night of testing. Height and weight were measured to calculate body mass index. Both centers used the same equipment
and settings to acquire the data. Polysomnography (PSG)
(Sleepwatch Compumedics, Melbourne, Australia) included
C3/A2 and C4/A1 electroencephalograms, electrooculogram, submental electromyogram, electrocardiogram, airflow, respiratory
effort, and SaO2. The PSGs and MWTs were manually analyzed
in 30-second epochs according to standard criteria,7 and the electroencephalogram arousals were scored using American Sleep
Disorders Association guidelines.8 Sleep-disordered breathing
was defined according the following criteria: an apnea was
defined as a > 80% reduction in airflow lasting at least 10 seconds and was classified as obstructive, central, or mixed depending on whether or not there were ongoing respiratory efforts. A
hypopnea was defined as an event of at least 10 seconds duration
with a > 50% reduction from baseline in at least 2 of the following 3 signals: airflow, thoracic movement, and abdominal movement. The apnea-hypopnea index (AHI) was calculated by dividing the number of apneas plus hyponeas by the hours of sleep.
RESULTS
Ten subjects were found to have an AHI of more than 10 per
hour, and data from these subjects were excluded from further
analysis because of the concern that subclinical OSA may affect
sleep latency. This resulted in a final sample of 31 subjects, 8
women and 23 men (age 48.7 years ± 9.7 SD). Their demographic, subjective sleepiness, and PSG variables are presented
in Table 1. MWT mean sleep-latency results for all 4 scoring criteria are presented in Table 2.
The previous study of normative MWT data6 used a statistical
criterion of 2 SDs below the mean as the cut-off point for normalcy, as suggested by the American Electroencephalographic
Society.9 Using this criterion, the cut-off points from the current
data for normal mean sleep latency for all scoring criteria were
14.0 minutes (MICROMWT20), 15.2 minutes (MWT20), 16.1
minutes (MICROMWT40), and 25.5 minutes (MWT40). The
previous authors’ reported mean sleep latencies, in minutes, were
18.1 ± 3.6, 18.7 ± 2.6, 32.6 ± 9.9 and 35.2 ± 7.9, respectively, and
their suggested cut-offs for normalcy, in minutes, were 10.9,
13.5, 12.9 and 19.4, respectively.6
MWT Protocol
On the day after the PSG, subjects underwent four 40-minute
MWT trials at 2-hourly intervals, with the first beginning
between 8 AM and 10 AM. Smoking was prohibited during the
30 minutes before a trial. Subjects were required to abstain from
consuming caffeinated beverages during the test day. Subjects
also gave a urine sample on the day of testing, which was
screened for habitual drugs of abuse. All subjects had negative
results. Breakfast (cereal, milk, and toast) was served at least 1
hour before the first MWT trial, and lunch (sandwich or roll,
piece of fruit, and noncaffeinated beverage) immediately after the
second trial (approximately 12 noon), 1 hour before the next trial.
All trials were performed in a similar setting using a simplified
recording montage (C3/A2, O1/A2, electromyogram, and electrooculogram). Bedrooms were sound attenuated, insulated from
external light, and equipped with overhead dimmer lights.
SLEEP, Vol. 27, No. 4, 2004
800
The MWT in Normal Healthy Subjects—Banks et al
used volunteers responding to advertisements or word of mouth
between hospital employees. It is possible that some subjects
who volunteered may have selected themselves because they had
some daytime sleepiness or were concerned about a possible
sleep disorder. Because the current study recruited the subjects
without specific mention of daytime sleepiness or sleep disorders, potential sample bias was avoided.
The subjects in the current study were required to sit upright in
a lounge chair rather than upright in bed, as occurred in the previous study. Some subjects may have been more inclined to fall
asleep while sitting in a bed because of the contextual cue to
sleep in this environment. Also, the illumination level during the
MWT trials in the previous study was much lower (0.10-0.13 lux
compared to 1 lux in this study). We believe that if the MWT is
to be generalizable to real-life situations, then the conditions of
the test need to be comparable. These differences in the study
protocol and recruitment methods may have been enough to
increase the variance of sleep-latency scores in the Doghramji et
al study6 and, therefore, reduce the lower normal limit suggested
by the authors.
Additionally, the different results in our study compared to
Doghramji et al’s6 are unlikely to be explained by different exclusion criteria for OSA. The previous authors used a slightly more
stringent AHI cut off (ie, AHI > 5 for 30- to 39-year-olds and AHI
> 10 for 60- to 69-year-olds) than in this study (AHI > 10 for all
subjects). This exclusion criterion may be expected to result in
the selection of fewer subjects with sleepiness and, therefore, a
DISCUSSION
The range of normal scores found in this study is very similar
to that previously found by Doghramji et al.6 In particular, the
mean sleep latencies for both trial protocols are virtually identical. However the SD of scores in this study is smaller and resulted in higher cut-off points for abnormality. The subjects in this
study were randomly selected from the telephone book in a 30km radius around the hospitals in Adelaide and Melbourne and
were invited to participate via a phone call. The previous study
Table 1—Means and Standard deviations of Demographic and
Polysomnographic Variables
Age (years)
BMI
ESS
SlpD
TST
AHI
ArI
MEAN
SD
48.7
27.0
5.4
419.9
320.2
4.6
16.6
9.7
3.8
3.1
61.3
57.4
3.1
7.4
Note: N=30; BMI = Body Mass Index (kg/m2), ESS = the Epworth
Sleepiness Scale, SlpD = Subjective sleep time from a sleep diary
(minutes) for the week prior to MWT, TST = Total Sleep Time from
overnight PSG (minutes), AHI = Apnoea/Hypopnoea Index (events/hr
sleep), ArI = Arousal Index (events/hour sleep).
Table 2—Mean sleep onset latencies for each MWT trial in both 20 and 40 min protocols.
% achieved
sleep onset
Mean
SD
Min
Max
5th %ile
10th %ile
25th %ile
50th %ile
75th %ile
90th %ile
% achieved
sleep onset
Mean
SD
Min
Max
5th %ile
10th %ile
25th %ile
50th %ile
75th %ile
90th %ile
0900
1100
12.9
19.0
3.2
4.5
20.0
8.4
17.2
20.0
20.0
20.0
20.0
22.6
18.6
3.4
6.0
20.0
8.4
12.6
20.0
20.0
20.0
20.0
0900
1100
19.4
35.2
10.3
4.5
40.0
8.4
17.2
40.0
40.0
40.0
40.0
35.5
32.5
11.5
6.0
40.0
8.4
12.6
21.0
40.0
40.0
40.0
MICROMWT20
1300
1500
19.4
18.2
4.3
2.0
20.0
6.2
10.2
20.0
20.0
20.0
20.0
19.4
18.5
3.9
5.0
20.0
6.2
11.2
20.0
20.0
20.0
20.0
MICROMWT40
1300
1500
45.2
31.1
12.0
2.0
40.0
6.2
10.2
21.0
40.0
40.0
40.0
29.0
32.9
11.8
5.0
40.0
6.2
11.2
22.0
40.0
40.0
40.0
Mean
0900
1100
MWT20
1300
1500
Mean
18.6
18.6
2.3
4.5
20.0
10.2
14.6
17.9
20.0
20.0
20.0
6.5
19.6
2.2
8.0
20.0
14.6
20.0
20.0
20.0
20.0
20.0
3.2
19.9
0.5
17.0
20.0
18.8
20.0
20.0
20.0
20.0
20.0
9.7
19.1
3.3
4.5
20.0
8.4
16.0
20.0
20.0
20.0
20.0
9.7
19.6
1.6
11.5
20.0
14.8
19.2
20.0
20.0
20.0
20.0
7.3
19.4
2.1
8.6
20.0
13.4
18.1
20.0
20.0
20.0
20.0
Mean
0900
1100
MWT40
1300
1500
Mean
32.3
32.9
8.4
9.1
40.0
12.7
19.6
28.0
36.3
40.0
40.0
19.4
37.1
7.6
8.0
40.0
14.6
22.2
40.0
40.0
40.0
40.0
16.1
38.2
5.0
17.0
40.0
22.4
31.0
40.0
40.0
40.0
40.0
25.8
35.3
9.7
4.5
40.0
8.4
16.6
37.0
40.0
40.0
40.0
12.9
38.2
7.7
11.5
40.0
14.8
20.2
40.0
40.0
40.0
40.0
18.6
36.9
5.4
16.5
40.0
19.8
31.1
35.8
40.0
40.0
40.0
Note: N=30; mean sleep latencies in minutes.
SLEEP, Vol. 27, No. 4, 2004
801
The MWT in Normal Healthy Subjects—Banks et al
narrower SD for MWT sleep latency. In fact, the findings in the
2 studies were opposite, suggesting that the different exclusion
criteria did not affect the resulting spread of mean sleep-latency
scores.
In summary, this study extends the base of normative data for
the MWT. The results for mean sleep latency, in our Australian
population, are remarkably similar to those obtained from 6
North American and 1 South African center years earlier.
However, the present results, which were obtained in a randomly
selected sample of the normal healthy population, suggest that
the lower limit for normal sleep latency may be higher than previously reported. This study illustrates the need for further investigation of these cut-off points and, perhaps, the introduction of a
grey area similar to that in current usage for the Multiple Sleep
Latency Test.
REFERENCES
1. Mitler MM, Gujavarty KS, Browman CP. Maintenance of wakefulness test: a polysomnographic technique for evaluation treatment
efficacy in patients with excessive somnolence. Electroencephalogr
Clin Neurophysiol 1982;53:658-61.
2. Poceta JS, Timms RM, Jeong DU, Ho SL, Erman MK, Mitler MM.
Maintenance of Wakefulness Test in obstructive sleep apnea syndrome. Chest 1992;101:893-7.
3. Tiihonen M, Partinen M. Polysomnography and Maintenance of
Wakefulness Test as predictors of CPAP effectiveness in obstructive
sleep apnea. Electroencephalogr Clin Neurophysiol 1998;107:3836.
4. Harsh J, Peszka J, Hartwig G, Mitler M. Night-time sleep and daytime sleepiness in narcolepsy. J Sleep Res 2000;9:309-16.
5. Sangal RB, Thomas L, Mitler MM. Disorders of excessive sleepiness. Treatment improves ability to stay awake but does not reduce
sleepiness. Chest 1992;102:699-703.
6. Doghramji K, Mitler MM, Sangal RB, et al. A normative study of
the Maintenance of Wakefulness Test (MWT). Electroencephalogr
Clin Neurophysiol. 1997;103:554-62.
7. Rechtshaffen A, Kales A. A Manual of Standardized Terminology,
Techniques, and Scoring System for Sleep Stages of Human
Subjects. Los Angeles: UCLA Brain Information Services/ Brain
Research Institute; 1968.
8. EEG arousals: scoring rules and examples: a preliminary report
from the Sleep Disorders Atlas Task Force of the American Sleep
Disorders Association. Sleep 1992;15:173-84.
9. American Electroencephalographic Society (1994). Guidelines in
electroencephalography, evoked potentials, and polysomnography. J
Clin Neurophysiol 11:1-147.
SLEEP, Vol. 27, No. 4, 2004
802
The MWT in Normal Healthy Subjects—Banks et al