Sleep, 20(12):1127-1134
© 1997 American Sleep Disorders Association and Sleep Research Society
Do Our Methods Lead to Insomniacs' Madness?:
Daytime Testing After Laboratory and Home-Based
Polysomnographic Studies
*tJack D. Edinger, :j:Ana 1. Fins, *tRobert J. Sullivan, Jr.,
tGail R. Marsh, tDorothy S. Dailey, *T. Victor Hope
*Margaret Young, tEdmund Shaw, *Donna Carlson and tDiane Vasilas
*Veterans Affairs Medical Center and tDuke University Medical Center, Durham, North Carolina; and
t.university of Miami School of Medicine, Coral Gables, Florida, U.S.A.
Summary: Complaints of daytime dysfunction are common among chronic insomniacs, but laboratory comparisons of insomniacs and age-matched and gender-matched normal controls have generally failed to document these
complaints. However, a few studies, which allowed subjects to sleep in their homes on the nights before daytime
testing, have shown some relative diurnal deficits among insomniacs. The current study compared the effects of
nocturnal laboratory and home polysomnogram (PSG) studies on subsequent daytime test results among older
insomniacs and normal sleepers. Insomniacs (n = 32) and normal sleepers (n = 32) were randomly assigned to
first undergo three nights of nocturnal PSG monitoring either in the sleep laboratory (16 insomniacs, 16 normal
sleepers) or in their homes (16 insomniacs, 16 normal sleepers). Following the third night of PSG monitoring,
subjects spent I day in the sleep laboratory, where they completed a four-trial multiple sleep latency test along with
four trials of a computer-administered performance test battery. Results showed that insomniacs, as a group, were
slightly, albeit consistently, sleepier than were normal sleepers following nights of home sleep monitoring, but a
reverse of this trend was found among subjects who underwent nocturnal laboratory PSG before daytime testing.
Furthermore, normal sleepers showed faster reaction times on a signal detection task than did insomniacs within
the subgroup who underwent home PSGs prior to such testing. However, within the subgroup that underwent
nocturnal laboratory PSGs, insomniacs' signal detection reaction times were significantly faster than those shown
by normal sleepers. Results provide some support for the speculation that the nocturnal PSG monitoring site, used
as a precursor to daytime testing, may systematically affect daytime comparisons between insomniacs and matched
controls. Moreover, these results suggest that the use of home-based nocturnal PSG monitoring prior to daytime
testing may provide an enhanced understanding of insomniacs' diurnal complaints. Key Words: Normal sleepersInsomniacs-Multiple sleep latency test-Reaction time-Older adults.
Insomniacs often complain bitterly about daytime
fatigue, reduced cognitive performance, and impaired
occupational and social functioning resulting from
their chronic sleep disturbances. However, laboratory
studies designed to document insomniacs' diurnal
complaints have generally failed to demonstrate significant daytime deficits among such individuals. For
example, numerous studies (1-8) have shown that insomniacs do not evidence greater daytime sleepiness
on the multiple sleep latency test (MSLT) than do agematched and gender-matched noncomplaining normal
sleepers. In fact, some studies (2,6) have clearly demonstrated less daytime sleepiness among insomniacs
Accepted for publication August 1997.
Address correspondence and reprint requests to Jack D. Edinger,
Ph.D., Psychology Service (l16B), Veterans Affairs Medical Center,
508 Fulton Street, Durham, NC 27705, U.S.A.
than among age-matched normal controls. Furthermore, studies (8-10) of insomniacs' mental functioning have failed to confirm their reported daytime cognitive difficulties. As a consequence, the origins of insomniacs' diurnal concerns have remained perplexing
and elusive.
As a result of such findings, many sleep experts
have come to view insomnia as a relatively mild form
of sleep disturbance that results in no significant daytime deficits for most affected individuals. This contention, in tum, has led to a variety of speculations to
explain insomniacs' nocturnal and diurnal complaints.
Stepanski et al. (1,2,6), for example, have suggested
that insomniacs suffer from a chronic state of hyperarousal that precludes normal sleep at night and excessive sleepiness during the daytime. In contrast,
Chambers and Kim (7) argue that insomniacs' relative
1127
1128
1. D. EDINGER ET AL.
tendency to complain of somatic distress accounts for
their diurnal complaints, whereas Morin et al. (11)
have suggested that sleep attitudes, beliefs, and expectations may contribute significantly to insomniacs'
self-appraisals. Alternately, various psychometric studies (12-15) showing mild psychopathology among insomniac samples has led others (13,15) to argue that
underlying psychiatric disturbances may exacerbate insomniacs' diurnal complaints. Finally, various studies
(16-18) showing exaggerated subjective impressions
of sleep disturbance among insomniacs have suggested
that perceptual deficits that preclude accurate discrimination of sleep and wakefulness may magnify insomniacs' sleep/wake symptoms.
Despite these prevailing views, some investigations
have suggested that insomniacs may suffer from actual
diurnal deficits. Epidemiological studies, for example,
have shown that insomniacs spend only about half as
much time in productive activities (e.g. working,
studying) and have twice as many fatigue-related traffic accidents as do individuals without sleep complaints (19,20). In addition, two studies, (21,22), which
employed pupillometric measures of daytime sleepiness, found that insomniacs demonstrated greater daytime sleepiness than normal sleepers, whereas another
study, designed to investigate the 24-hour activity patterns, showed insomniacs evidenced significantly less
activity during the daytime than did normal sleepers
(23). Because such findings differ from results of the
afore-cited laboratory studies, it seems reasonable to
speculate that methodological factors, rather than inherent characteristics of insomniacs themselves, may
account for the failure of many previous studies to
document insomniacs' daytime deficits.
In this regard, it is noteworthy that the studies cited
above that failed to support insomniacs' daytime complaints all used laboratory polysomnography (LPSG)
as a precursor to daytime testing. In contrast, those
studies that documented diurnal deficits among insomniacs allowed subjects to sleep in their own homes on
the nights before daytime measures were obtained.
This methodological difference seems significant inasmuch as some reports (24-30) have noted that LPSG
may lead to relative sleep disruption among normal
sleepers and transitory sleep stabilization among insomniacs. As a result, when LPSG is used as precursor
to daytime comparisons of normal sleepers and insomniacs, such sleep monitoring may reduce or eliminate
differences between such groups during subsequent diurnal testing. The current investigation was conducted
to test this hypothesis. Specifically, this study examines the differences between insomniacs and agematched and gender-matched normal controls shown
on MSLT and performance testing following nocturnal
home-based or laboratory-based sleep monitoring.
Sleep. Va!. 20. No. 12. 1997
METHOD
Subjects
Data for this study were obtained from the subjects
who participated in the study described in our preceding report (24). Because the characteristics of these
subjects and the methods (31-34) used in their recruitment and screening were described in the preceding
article, this information will not be repeated here.
Nonetheless, it is noteworthy that the two subject
groups 1) were each composed of 16 women and 16
men; 2) consisted predominantly of Caucasians (the
normal sleepers consisted of 16 Caucasian women, 15
Caucasian men, and 1 African-American man, whereas
the insomniacs included 16 Caucasian women, 14
Caucasian men, and 2 Asian-American men); and 3)
did not differ significantly from each other in regard
to their mean ages and educational levels. Hence, these
groups were well matched on variables that otherwise
might confound their comparisons on the dependent
measures (daytime sleepiness and reaction time) used
herein.
Multiple sleep latency test (MSL T)
All subjects underwent a four-trial MSLT following
three nights of nocturnal sleep monitoring conducted
either in their homes or in the sleep laboratory. In most
respects, the standard MSLT protocol (35) was followed with the first nap commencing 2-3 hours after
the morning rising time following the third consecutive night of lab or home PSG monitoring, with each
subsequent nap occurring at 2-hour intervals. For each
nap, subjects were placed in a darkened room in the
sleep laboratory and were instructed to attempt to fall
asleep. Because this study was conducted between October 1, 1992, and September 30, 1994, traditional
MSLT criteria were used to define the sleep latency
for each nap. Specifically, sleep latency was defined
as the time between the beginning of the nap trial and
either the first three consecutive 30-second epochs of
stage 1 sleep or the first 30-second epoch of any other
sleep stage. If no sleep occurred, the trial was terminated at 20 minutes, and a sleep latency of 20 minutes
was assigned. To minimize carryover effects from one
nap to the next, each nap trial was discontinued when
the subject showed 3 consecutive minutes of sleep of
any stage.
Performance testing
Immediately prior to each nap trial, each subject underwent a 16-minute series of three computer-administered performance tests selected from the neurobe-
DAYTIME TESTING AFTER POLYSOMNOGRAPHY
havioral evaluation system (NES) (36). During this
testing, each subject was placed individually in a testing room. The subject sat in front of a PC computer
that contained the NES software used for test administration. At the beginning of each test, the computer
software provided written instructions for the subject
on the computer screen. The subject read the instructions and then proceeded with the test administered in
a standardized manner as guided by the computer software.
Inasmuch as reaction time testing has frequently
been used in previous comparisons of insomniacs with
normal sleepers, we chose three NES reaction time
tests so that our results might be compared to such
previous studies. However, we attempted to choose
three tests that varied in complexity and that required
different degrees of decision-making and concentration/attention by our subjects. Nonetheless, we recognized these three tests measured an extremely restricted range of human performance; yet a more diverse
and extensive test battery seemed obviated given the
considerable burden placed on subjects to complete all
procedures (i.e. extensive screening procedures, multiple PSGs, daytime testing, etc.) in this project. The
following discussion provides a detailed description of
each of the three tests used in this study.
Simple reaction time test
The first and least difficult test presented in each
trial was the simple reaction time test (SRT). During
this test, the subject was required to press a specially
marked key on a computer keyboard whenever shelhe
saw a figure (i.e. a small square) appear on the computer screen. Once the test was begun, the figure appeared at intervals varying between 1,000 milliseconds
and 2,500 milliseconds. Upon each presentation, the
figure remained on the screen either until the subject
responded or until 1,000 milliseconds had elapsed. The
total trial lasted approximately 5 minutes and consisted
of 90 presentations of the target stimulus. For each
presentation of the figure, the computer software automatically recorded the time (in milliseconds) between the appearance of the figure on the screen and
the subject's computer key-press response. For each 5minute SRT trial, the computer software computed a
mean response latency and a within-subject standard
deviation of the subject's response latencies. These two
indices served as measures of SRT performance on
each test trial.
Continuous performance test
This somewhat more challenging test, which immediately followed the SRT, consisted of a signal de-
1129
tection task during which a target (i.e. the letter "S")
and background signals (i.e. the letters "1>(', "C",
"E", and "T") were presented in rapid fashion on the
computer screen. Once the continuous performance
test (CPT) was begun, target and background letters
appeared in a random sequence during which a 1:4
target-to-background ratio was maintained. Target and
background letters were presented at the rate of one
per second, and each letter remained on the computer
screen for 50 milliseconds. The total test lasted approximately 5 minutes and included 60 presentations
of the target and 240 presentations of the background
stimuli. In responding to this test, the subject was required to press a specially marked key on the computer
keyboard when and only when the preidentified target
letter appeared on the screen. For each trial of this test,
the subject's mean response latency and a within-subject standard deviation for the response latencies were
derived using the NES computer software for this specific test.
Switching attention test
The most difficult task included in the performance
battery was the switching attention test (SWAT),
which lasted approximately 6 minutes and included
three subtests. Each subtest required the subject to
press speciallY marked keys on the right and left sides
of the keyboard in response to stimuli presented on the
computer screen. During part I (side condition) of the
SWAT, a square appeared on either the right or left
side of the computer screen, and the subject was required to press a marked key on the corresponding side
of the computer keyboard. Upon each presentation, the
stimulus remained on the screen until either the subject
responded or 2,500 milliseconds had elapsed. This
portion of the SWAT included six practice and 16 test
presentations of the stimulus.
During part II (direction condition), an arrow, pointing right or left, appeared in the center of the screen,
and the subject was required to make a right-side or
left-side key response in response to the direction in
which the arrow was pointing. As in the previous section of the SWAT, the stimulus remained on the screen
either until the subject responded or until 2,500 milliseconds had elapsed. This portion of the test included
four practice and 16 test presentations of the stimulus.
Finally, during part III (switching condition), an arrow (pointing right or left) appeared on either the right
or left side of the screen. Preceding each presentation
of this arrow by 1,000 milliseconds, one of two command words, "SIDE" or "DIRECTION", appeared on
the screen. This command word served to instruct the
subject to respond by pressing a key on the side of the
keyboard corresponding either to the side of the screen
Sleep, Vol. 20, No. 12, 1997
1130
1. D. EDINGER ET AL.
TABLE 1.
Spearman correlations among mean MSLT latencies and the performance measures
SRT-L
MSLT
SRT-L
SRT-SO
CPT-L
CPT-SO
SWAT I-L
SWAT I-SO
SWAT II-L
SWAT II-SO
-0.19
SRT-SO
-0.17
0.81 **
CPT-L
CPT-SO
-0.29*
0.71 **
0.59**
-0.31*
0.42**
0.46**
0.79**
SWAT I-L
SWAT I-SO
-0.27*
0.76**
0.65**
0.74**
0.58**
-0.12
0.57**
0.61 **
0.53**
0.55**
0.76**
SWAT II-L SWAT II-SO
-0.14
0.63**
0.59**
0.58**
0.50**
0.73**
0.55**
0.01
0.35**
0.35**
0.36**
0.51**
0.43**
0.50**
0.70**
MSLT, multiple sleep latency test; SRT, simple reaction time test; CPT, continuous performance test; SWAT I, switching attention test,
part I; SWAT II, switching attention test, part II; L, mean response latency; SO, within-subject standard deviation of response latencies.
* p < 0.05, ** P < 0.005.
on which the arrow appeared or the direction in which
the arrow was pointing. On 50% of the presentations,
the side of the screen on which the arrow appeared
and the direction in which it was pointing agreed. On
the remaining presentations, these two stimulus characteristics were in conflict. Throughout the test, these
non conflict and conflict presentations occurred in a
random sequence. Overall, the switching condition included eight practice and 48 test presentations of the
command-stimulus combination. For each test trial, the
NES computer software computed a mean response
latency and a within-subject response latency standard
deviation for each section (i.e. side, direction, and
switching conditions) of the SWAT.
Procedure
As noted in our preceding report (24), 32 randomly
selected subjects (16 normals, 16 insomniacs) underwent home sleep monitoring during the three nights
prior to the daytime testing, whereas the remaining 32
subjects underwent three consecutive lab PSG studies
prior to this testing. All performancelMSLT trials were
conducted in the sleep laboratory under the supervision of trained laboratory technologists. Each performancelMSLT trial was commenced on the instruction
of the assigned technologist, and subjects were supervised between trials to prevent unscheduled sleep episodes. PSG electrodes were worn for the entire day
of laboratory testing and were not removed until after
the final trial was completed. After the fourth performancelMSLT trial, electrodes were removed and the
subject was allowed to leave the laboratory. Once all
subjects had completed the study, factorial statistical
analyses were conducted to determine whether the
nocturnal recording site used as a precursor to daytime
testing had an effect on the MSLT/performance data
obtained.
Sleep, Vol. 20, No. 12, 1997
RESULTS
Preliminary analysis
Prior to conducting the statistical tests relevant to
this study's hypothesis, we first conducted a preliminary correlational analysis to determine the relationship among our dependent measures. This correlational analysis was conducted specifically to determine if
our various performance and MSLT measures reflected
distinctive aspects of daytime functioning. Table 1
shows Spearman rank order correlations among subjects' mean sleep latencies across trials and their respective mean performance test scores (i.e. within-subject means and standard deviations for the response
latency data). Although several of the MSLT/performance score correlations were statistically significant,
their magnitudes consistently indicated less than 10%
shared variance between MSLT latencies and each of
the performance measures. In contrast, the intercorrelations among the performance test scores were all significant. However, the mean correlation between pairs
of the performance measures was 0.58 (range = 0.350.81); this finding suggested that only about 34% of
the variance in one performance measure was predictable from another of these measures. This observation,
in tum, suggested that the MSLT and separate performance tests provided related but not identical measures
of daytime functioning. Hence, separate analyses of
data from the MSLT and each of the performance tests
seemed appropriate.
MSL T results
As noted previously, this study was designed to determine whether the nocturnal recording site (lab vs.
home) wherein PSG studies are conducted prior to
daytime testing might affect the results obtained from
daytime comparisons of normal sleepers and insomniacs. Nonetheless, the daytime testing conducted after
home PSG required our subjects to sleep at home and
1131
DAYTIME TESTING AFTER POLYSOMNOGRAPHY
TABLE 2. MSLT means and standard deviations (minutes) for four subject groups across trials
MSLT after lab PSG
Nonnals
Trial
Trial
Trial
Trial
I
2
3
4
average were sleepier on each trial than were normals
who slept in their homes before such testing. Moreover, the opposite trend was consistently the case
among subjects who slept in the lab prior to the MSLT.
The fact that all eight insomniac vs. normal MSLT
comparisons shown in Table 2 were in the predicted
direction was found to be significant by the nonparametric Fisher's exact test (p = 0.029, two-tailed test).
Hence, the nocturnal recording site used as a precursor
to daytime testing did appear to have some influence
on our MSLT results.
MSLT after home PSG
Insomniacs
Nonnals
Insomniacs
Mean
SO
Mean
SO
Mean
SO
Mean
SO
lL2
10.3
9.9
10.6
5.3
6.2
7.5
7.2
I\.6
10.8
I\.6
12.4
6.6
6.3
6.5
6.9
13.9
10.3
8.7
11.0
6.2
6.3
5.5
6.3
13.1
8.9
7.1
9.3
5.4
5.1
4.7
6.3
MSLT, multiple sleep latency test; PSG, polysomnography.
then arise and travel to the laboratory for their daytime
testing. For such subjects, we recognized that the added travel time to the lab might have delay the beginning of their daytime testing relative to those subjects
undergoing such testing following lab PSG. As a consequence, comparisons of the home and lab PSG
groups vis-a-vis their daytime measures could be confounded by time of day (circadian) factors that would
make such comparisons uninterpretable. However, we
found that the starting times (i.e. beginning times for
trial 1) of MSLT testing for subjects undergoing home
PSG prior to their daytime testing were not significantly later [F (1,63) = 0.99, P > 0.30] than the MSLT
starting times for subjects undergoing daytime testing
after lab PSG. In fact, the average MSLT starting time
(mean time = 8:56 a.m., SD = 39 minutes) for subjects undergoing home PSG prior to daytime testing
was only 9 minutes later than the average starting time
(mean time = 8:47 a.m., SD = 27 minutes) for subjects undergoing lab PSG prior to their daytime testing. Given this observation, it seemed unlikely that
circadian factors influenced the outcome of MSLT testing.
A 2 (subject type) X 2 (nocturnal recording site)
analysis of variance (ANOVA) was used to compare
normalized (via logarithmic transformation) mean
MSLT latencies across trials of our four subgroups of
subjects. Results of this analysis were all nonsignificant. However, as shown by Table 2, insomniacs who
slept in their homes on the nights before the MSLT on
TABLE 3.
Performance test results
Given the design of this study, a total of 256 performance test trials (64 subjects X four trials each)
were conducted with the entire sample. However, some
of the data were lost during the course of the study.
Technician error accounted for the failure to save some
or all of the data from trial 1 for three subjects. The
remainder of the data loss occurred as a result of default parameters included in the NES software. These
default parameters caused the NES software to record
and store CPT reaction times :::;900 milliseconds and
SWAT performances that included no more than four
errors out of each block of eight stimulus presentations
throughout the test. Some subjects exceeded these default parameters on some or all performance test trials.
In these cases, the NES software failed to record and
store subjects' performance data, and information
about how poorly the subjects had performed was lost.
As a result, it was not possible to assign subjects
meaningful performance scores for the these test trials.
Table 3 shows the numbers of subjects and trials in
each experimental condition with missing data.
Because relatively little data was lost from the SRT
« 1%), the CPT «2%), and the initial two portions
(i.e. side and direction conditions) of the SWAT
(3.1 %) test trials, the missing data did not seem sufficient to bias our planned statistical comparisons for
these tests. In contrast, the 27 (10.5%) test trials lost
Distribution of missing performance data
Subjects tested after lab PSG
SWAT
SRTCPT
Subjects with complete data (n)
Subjects missing one trial (n)
Subjects missing two trials (n)
Subjects missing three trials (n)
Subjects missing four trials (n)
Total trials missing
Subjects tested after home PSG
Insomniacs
Normals
SWAT
II
III
12
2
I
16
16
13
3
13
3
0
0
3
3
Normals
I
8
SRT CPT
15
I
14
I
3
15
I
Insomniacs
SWAT
II
III
SRT CPT
15
I
II
2
2
16
16
9
0
0
SWAT
II
III
14
2
14
2
12
3
2
2
6
SRT CPT
15
II
III
14
2
14
2
14
2
12
4
2
2
2
4
PSG, polysomnography; SRT, simple reaction time test; CPT, continuous performance test; SWAT, switching attention test.
Sleep. Vol. 20. No. 12, 1997
1132
1. D. EDINGER ET AL.
for part III (i.e. switching side/direction) of the SWAT
seemed more problematic. Hence, we conducted statistical comparisons with only those data derived from
the SRT, CPT, and initial two portions of the SWAT.
In conducting these analyses, we used the SAS general
linear models ANOVA [SAS user's guide, SAS Institute, Inc., Cary, NC, 1988 (37)], which provides unbiased variance estimates for data sets that are unbalanced due to missing or lost data points.
Two multivariate statistical tests were conducted
with the subjects' performance data. One analysis consisted of a 2 (normals vs. insomniacs) X 2 (home vs.
lab nocturnal PSG) X 4 (test trials) X 4 (SRT vs. CPT
vs. SWAT part I vs. SWAT part II) conducted with
subjects' within-trials mean latencies, whereas the second consisted of a similar analysis conducted with the
within-subject standard deviations for these latencies.
To allow for the statistical comparisons across these
tests, subjects' data for each measure were standardized by converting them to T scores (i.e. mean = 50,
SD = 10) prior to conducting these analyses. The former of the analyses showed significant subject type X
test [F (3,877) = 4.11, P < 0.01] and subject type X
prior nocturnal recording site X test [F (3,877) = 4.41,
P < 0.005] interaction effects; the latter analysis
showed no significant main or interaction effects. The
mean raw response latencies (milliseconds) across tests
are shown in Fig. 1. A posteriori tests (with Bonferroni
correction) showed that insomniacs tended to perform
significantly (p < 0.0125) better (i.e. to have shorter
mean latencies) on part I of the SWAT than did normals, regardless of the nocturnal PSG recording site
employed. Mean response latencies for normals and
insomniacs derived from each of the remaining three
tests were not significantly different.
To test for the effects contributing to the three-way
interaction, we conducted a separate series of a posteriori comparisons with each of the four performance
tests. For each test we compared the performances of
1) normals who underwent home PSG with normals
who underwent lab PSG; 2) insomniacs who underwent home PSG with insomniacs who underwent lab
PSG; 3) normals and insomniacs who underwent home
PSG; and 4) normals and insomniacs who underwent
lab PSG. Because these a posteriori tests included a
total of 16 group comparisons, a Bonferroni correction
was used to reduce the chance of type I error. Results
of these comparisons showed that normals who slept
at home on the night before daytime testing performed
significantly (p < 0.0001) better (i.e. had faster reaction times) on the CPT than did normals who slept in
the lab 01} pretesting nights. In contrast, insomniacs
who slept in the lab on the nights before daytime testing performed significantly better (p < 0.0001) than
did insomniacs who slept at home on these nights.
Sleep, Vol. 20, No. 12, 1997
SRT Data
265
260
255
250
LPSG
HPSG
CPT
Data
395
3 8 0
365
3 5 0
L PSG
315
HPSG
SWAT
Partl
3 0 0
2 8 5
2 7 0
L PSG
H PSG
S W AT - Part II
485
470
455
440
LPSG
HPSG
FIG. 1. Mean response latencies for performance tests. SRT, simple reaction time test; CPT, continuous performance test; SWAT,
switching attention test; LPSG, lab polysomnography; HPSG, home
polysomnography.
Among those subjects who slept in their homes prior
to daytime testing, CPT performances of normals were
superior (p < 0.0001) to insomniacs, whereas the CPT
performances of insomniacs were significantly better
than those of the normals among subjects who slept in
the lab prior to daytime testing. Results of the 12 remaining group comparisons did not reach the Bonferroni level of significance.
DISCUSSION
The current study was conducted to determine if the
recording site in which nocturnal PSG recordings are
conducted has an effect on subsequent daytime comparisons of normal sleepers and insomniacs. Specifically, we predicted that laboratory PSG conducted prior to daytime testing may reduce differences between
normal sleepers and insomniacs on measures of day-
DAYTIME TESTING AFTER POLYSOMNOGRAPHY
1133
time sleepiness and cognitive performance. In contrast, chosen for comparisons of insomniacs and normal
we speculated that home PSG, used as the precursor sleepers, results of these comparisons may make into daytime testing, would lead to greater sleepiness and somniacs' daytime complaints appear trivial. Hence,
performance differences between these two groups. investigators interested in examining insomniacs' dayAlthough results obtained with a portion of our find- time complaints may benefit by employing a variety
ings did not support these contentions, the results ob- of performance measures in their research.
tained with some measures supported our speculations.
Given the minimal home vs. lab sleep time differResults derived from the MSLT, for example, ences noted in our preceding report (24), it is, perhaps,
showed that insomniacs, on average, were slightly, al- surprising that such differences could account for the
beit consistently, more alert than a well-matched group daytime test results obtained herein. However, at least
of normal sleepers when such daytime testing was con- one previous study provides some guidance in explainducted after nocturnal laboratory PSG. These MSLT ing our findings. In their study of middle-aged and
findings, in fact, are very similar to those reported by older normal sleepers, Monk and Moline (38) found
Stepanski et al. (2), who compared a mixed group of that subjects, changing from entrained to free-run
insomnia subtypes with a group of noncomplaining sleeping conditions, showed increased time in bed and
normal sleepers. Such findings imply, as Stepanski et reduced daytime sleepiness without concomitant inal. have suggested, that insomniacs may suffer from creases in nocturnal sleep time. As these authors sugchronic hyperarousal that reduces their sleep propen- gested such benefits to daytime alertness may result
sity during both nighttime and daytime sleep attempts. from the benefits of quiet bed rest or overall changes
However, the MSLT data we gathered from subjects in the circadian system. Whereas both of our subject
who underwent home PSG prior to daytime testing groups spent more time in bed when sleeping in their
provide a contrasting impression. Indeed, these latter homes than they did in the sleep laboratory, it is posfindings suggest that the insomniacs are slightly, albeit sible that normals and insomniacs react very differconsistently, sleepier during the daytime than are nor- ently to wake time in bed. Normal sleepers may reap
mal sleepers. These results appear consistent with the the benefits of increased wake time in bed noted by
previously referenced pupillometric (21,22) and actig- Monk and Moline (38), but insomniacs may fail to do
raphy (23) studies, which compared insomniacs and so or may even suffer from it. If this speculation is
normal sleepers after or during periods of sleep in their correct, the more restrictive sleep laboratory setting
homes. Our findings, considered in conjunction with may favor insomniacs' subsequent daytime perforthese previous studies, suggest that insomniacs' hy- mance, whereas the more liberal home sleep setting
perarousal may be confined primarily to the nighttime may favor normal sleepers' daytime functioning.
sleep period. Moreover, these findings imply that inOverall, our results provide a slightly different view
somniacs' complaints of daytime sleepiness and fa- of insomniacs' daytime complaints than have some
tigue may merit serious attention and should not be previous reports. Our MSLT and CPT testing results
disregarded or summarily dismissed when such pa- imply that insomniacs may evidence consistent, albeit
small, daytime deficits relative to normal sleepers
tients are encountered clinically.
In addition, results from our performance testing when both groups are allowed to sleep in their usual
suggest previous methods (3,4,10,12) for substantiat- home sleeping environments. In contrast, many preing insomniacs' reported daytime concentration and at- vious studies, which employed laboratory PSG prior
tention deficits may also contribute to such patients' to daytime testing, have suggested that insomniacs'
presumed madness. In this regard, our data show that daytime complaints may merely represent somatized
home and laboratory PSG recording sites may lead to anxiety (7) or underlying psychopathology (15). Of
different impressions about insomniacs' performance course, our clinical experience suggests that somatideficits, particularly on such signal detection tasks like zation processes, psychological traits, and physiologithe CPT used in the present study. When used as a cal arousal all may be important factors to consider in
precursor to daytime testing, laboratory PSG may understanding various insomnia SUbtypes. Nonetheplace insomniacs at an advantage and/or normal sleep- less, the findings reported herein suggest insomniacs
ers at a disadvantage. However, when subjects are al- may have veridical daytime deficits that require conlowed to sleep in their usual home sleeping environ- sideration as well.
ments, insomniacs may be more prone to show the
As noted at length in the accompanying report (24),
relative attention/concentration deficits of which they this investigation had several limitations that merit
complain. Furthermore, our data suggest that, regard- consideration. Briefly stated, these include 1) the inless of the nocturnal PSG recording site, some atten- clusion of insomniacs, who in most cases were not
tion tests (e.g. SWAT part I) may favor insomniacs clinical patients; 2) the selection of subjects, who by
over normal sleepers. As a result, if such measures are virtue of their advanced ages, may have been relatively
Sleep, Vol. 20, No. 12, 1997
1. D. EDINGER ET AL.
1134
prone to show daytime sleepiness and performance deficits; and 3) the finding noted in the previous report
(24) that both groups of subjects elected to stay in bed
longer during home PSG studies than they did during
laboratory recordings. In addition, the overall research
burden imposed on subjects enrolled in this investigation led us to limit the number and variety of performance measures used. As a result, replications of
this study with clinical insomniacs, younger age
groups, a wider range of performance measures, and
more rigid time-in-bed prescriptions for lab and home
PSG recordings may be useful. Nevertheless, the results reported herein suggest that a research paradigm
that employs home PSG monitoring prior to daytime
testing may provide new insights into the specific daytime deficits that contribute to insomniacs' diurnal
complaints.
Acknowledgements: This research was supported by
the Department of Veterans Affairs Merit Review Program.
Dr. Sullivan's efforts were supported in part by the Claude
D. Pepper Older Americans Independence Center grant #5
P60 AG 11268 and by the NIH National Center for Research
Resources, General Clinical Research Program, grant #M01RR-30.
REFERENCES
1. Stepanski E, Lamphere J, Badia P, Zorick F, Roth T. Sleep fragmentation and daytime sleepiness. Sleep 1984;7: 18-26.
2. Stepanski E, Zorick F, Roehrs T, Young D, Roth T. Daytime
alertness in patients with chronic insomnia compared with
asymptomatic control subjects. Sleep 1988; 11 :54-60.
3. Horne JA, Wilkinson S. Chronic sleep reduction: daytime vigilance performance and EEG measures of sleepiness with particular reference to "practice" effects. Psychophysiology 1985;
22:69-78.
4. Seidel WF, Ball S, Cohen S, Patterson N, Yost D, Dement We.
Daytime alertness in relation to mood, performance and nocturnal sleep in chronic insomniacs and noncomplaining sleepers.
Sleep 1984;7:230-8.
5. Seidel WF, Dement WC. Sleepiness in insomnia: evaluation and
treatment. Sleep 1982; Suppl 2: 180-90.
6. Stepanski E, Zorick F, Sicklesteel J, Young D, Roth T. Daytime
alertness-sleepiness in patients with chronic insomnia. Sleep Res
1986;15:174.
7. Chambers MJ, Kim Jy. The role of state-trait anxiety in insomnia and daytime restedness. Behav Med 1993;19:42-6.
8. Mendelson WB, Garnett D, Gillin JC, Weingartner H. The cxperience of insomnia and daytime and nighttime functioning.
Psychiatry Res 1984:12:235-50.
9. Hauri PJ, Wisbey J. Daytime psychomotor performance in insomnia. Presented at the annual meeting of the Association of
Professional Sleep Societies, Los Angeles, 1993.
10. Sugarman JL, Stern JA, Walsh JK. Daytime alertness in subjective and objective insomnia: some preliminary findings. Bioi
Psychiatry 1985;20:741-50.
II. Morin CM, Stone J, Trinkle D, Mercer J, Remsberg S. Dysfunctional beliefs and attitudes about sleep among older adults
with and without insomnia complaints. Psychol Aging 1993;8:
463-7.
12. Monroe LJ. Psychological and physiological differences between good and poor sleepers. J Consult Clin Psychol 1967;72:
255-64.
Sleep, Vol. 20, No. 12, 1997
13. Kales A, Caldwell AB, Preston TA, Healey S, Kales JD. Personality patterns in insomnia. Arch Gen Psychiatry 1976;33:
1128-34.
14. Edinger JD, Stout AL, Hoelscher TJ. Cluster analysis of insomniacs' MMPI profiles: relation of subtypes to sleep history and
treatment outcome. Psychosom Med 1988;50:77-87.
15. Kales A, Caldwell AB, Soldatos CR, Bixler EO, Kales JD.
Biopsychobehavioral correlates of insomnia, II: MMPI pattern
specificity and consistency. Psychosom Med 1983;45:341-56.
16. Frankel BL, Coursey RD, Buchbinder R, Snyder F. Recorded
and reported sleep in chronic primary insomnia. Arch Gen Psychiatry 1976;33:615-23.
17. Carskadon MA, Dement We, Mitler MM, Guilleminault C, Zarcone VP, Speigel R. Self-reports verses sleep laboratory findings
in 122 drug-free subjects with complaints of chronic insomnia.
Am J Psychiatry 1976;133:1382-8.
18. Edinger JD, Fins AI. The distribution and clinical significance
of sleep time perceptions among insomniacs. Sleep 1995;18:
232-9.
19. Institute of Medicine. Sleeping pilis, insomnia, and medical
practice. Washington, DC: National Academy 'of Sciences,
1979.
20. The Gallup Organization. Sleep in America: a national survey
of us. adults. Princeton, NJ: The Gallup Organization, 1991.
21. Lichstein KL, Johnson RS. Pupillometric discrimination of insomniacs. Behav Res Ther 1993;32:123-9.
22. Lichstein KL, Johnson RS, Sen Gupta S, O'Laughlin DL, Dystra TA. Are insomniacs sleepy during the day?: a pupillometric
assessment. Behav Res Ther 1992;30:283-92.
23. Pollak Cp, Perlick D, Linsner JP. Daily sleep reports and circadian rest-activity cycles of elderly community residents with
insomnia. Bioi Psychiatry 1992;32:1019-27.
24. Edinger JD, Fins AI, Sullivan RJ Jr, et al. Sleep in the laboratory
and sleep at home: comparisons of older insomniacs and normal
sleepers. Sleep 1997;20: 1119-26.
25. Rechtschaffen A, Verdone P. Amount of dreaming: effect of
incentive and adaptation to laboratory and individual differences. Percept Mot Skills 1964;19:947-58.
26. Agnew HW, Webb WB, Williams RL. The first night effect: an
EEG study of sleep. Psychophysiology 1966;2:263-6.
27. Mendel J, Hawkins DR. Sleep laboratory adaptation in normal
subjects and depressed patients ("first night effect"). Electroencephalogr Clin Neurophysiol 1967;22:556-8.
28. Schmidt HS, Kaelbling R. A differential laboratory adaptation
of sleep parameters. Bioi Psychiatry 1971 ;3:33-45.
29. Hauri P, Olmstead E. Reverse first night effect in insomnia.
Sleep 1989;12:97-105.
30. Kales A, Kales JD. Evaluation and treatment of insomnia. New
York: Oxford, 1984.
31. Schramm E. Test-retest reliability and validity of the structured
interview for sleep disorders according to SM-IlI-R (SIS-D).
Presented at the annual meeting of the Association of Professional Sleep Societies, Phoenix, AZ. 1991.
32. Schramm E, Hohagen F, Grasshoff MA, et at Test-retest reliability and validity of the structured interview for sleep disorders according to DSM-III-K Am J Psychiatry 1993;150:86772.
33. American Psychiatric Association. Diagnostic and statistical
manual of mental disorders, 4th edition, revised. Washington,
DC: American Psychiatric Association, 1994.
34. American Psychiatric Press. Structured clinical interview for
DSM-III-R (SCID). Washington, DC: American Psychiatric
Press, 1990.
35. Richardson GS, Carskadon MA, Flagg W, Van Den Hoed J,
Dement WC, Miter MM. Excessive daytime sleepiness in man:
multiple sleep latency measurement in narcoleptic and control
subjects. Electroencephalogr Clin NeurophysioI1978;45:621-7.
36. Letz R, Baker EL. NES2 Neurobehavioral evaluation system.
University of North Carolina, Chapel Hill, 1988.
37. SAS Institute, Inc. SAS user's guide. Cary, NC: Statistical Analysis Systems Institute, Inc., 1988.
38. Monk TH, Moline ML. Removal of temporal constraints in the
middle-aged and elderly: effects on sleep and sleepiness. Sleep
1988; I 1:513-20.