Sleep. 6(2): 147-155
© 1983 Raven Press. New York
What Do Insomniacs Do, Think, and Feel
During the Day? A Preliminary Study
*Evelyn 1. Marchini, tThomas 1. Coates, Hohn G. Magistad, and
*Shirley 1. Waldum
*Pacific Graduate School of Psychology, Palo Alto, tDivision of General Internal Medicine,
University of California School of Medicine, San Francisco, and MGM Associates, Inc.,
Palo Alto, California, U.S.A.
Summary: Ten sleep-maintenance insomniacs and 11 good sleepers were issued electronic pagers. They activated their devices during their waking hours
and completed the Marchini Monitoring Inventory (MMI) when paged at random
intervals throughout the day. The insomniacs spent significantly more time
shopping, watching television, and relaxing, whereas good sleepers spent significantly more time talking to people, working, and studying. Good sleepers
thought more about their day's routine, everyday problems, work, and their
interpersonal relationships with family members; insomniacs thought more
about their immediate physical environment and numerous forms of passive
relaxation. Good sleepers appeared busier, more active, and more involved in
their work and with other people. Insomniacs seemed more preoccupied with
self. Insomniacs might benefit from treatment strategies that entail increased
activity and involvement with other persons and external events. Key Words:
Insomnia-Daytime activity-Data collection-Pager (electronic).
There is an intimate link between personality, waking actions, and nighttime sleep
(1). Personality traits and daytime actions probably influence, and are influenced by,
sleep at night (2,3). Although the global personality traits of insomniacs have been
described extensively (2,4,5), there is virtually no literature on the daytime activities
of insomniacs and their emotions through the day. Documenting the behavior and
emotions of insomniacs across the day seems essential, because insomnia is defined
in terms of both nighttime sleep and daytime behavior and feelings (6,7).
Our research on nonpharmacologic treatments for primary chronic insomnia has
made us aware of the need to develop treatments more powerful than relaxation-based
therapies (1,8). The present study was designed to develop and test procedures for
describing the daytime actions, moods, and thoughts of insomniacs and good sleepers.
The data from this present study and others to follow from it might provide a basis for
determining whether insomniacs behave and feel differently from good sleepers during
Accepted for publication February 1983.
Address correspondence and reprint requests to Evelyn Marchini, 1151 Compass Lane #205, Foster City,
CA 94404, U.S.A.
147
148
E. J. MARCHINI ET AL.
the day; or at specific times during the day. The data from these descriptive studies
might be used subsequently to develop programs to modify behaviors and mood states
of insomniacs to determine their impact on sleep.
METHODS
Subjects
Subjects were recruited through advertisements placed in area newspapers requesting good sleepers and insomniacs to volunteer for research. Twelve sleep maintenance insomniacs and 12 good sleepers were selected according to the following
criteria. (a) Two male and two female insomniacs were selected within each of the
following age ranges: 20-35 years, 36-50 years, and 51-65 years. Good sleepers were
selected to meet the same age and sex distributions. (b) The insomniacs reported
insomnia (latency to sleep greater than 60 min, or wake after sleep greater than 30
min, or total sleep time less than 6.5 h) for at least the previous 2 years on 4 nights
per week. (c) Any person with a medical condition that could disrupt sleep, such as
diabetes, heart disease, thyroid dysfunction, pregnancy, menopause, or kidney or liver
dysfunction, was excluded. Self-reports obtained during an intensive interview were
corroborated by reports from subjects' family physicians. (d) Insomniacs reported no
symptoms of sleep apnea, nocturnal myoclonus, or narcolepsy, and no evidence of
these disorders was obtained in baseline sleep recordings. (e) The good sleepers had
no complaints of poor sleep, used no sedative medication, had no complicating medical
conditions, and did not abuse alcohol or drugs. (f) During post-adaptation sleep, recorded in three consecutive all-night recordings at home, insomniacs averaged less
than 88% sleep efficiency (time asleep divided by time in bed), whereas good sleepers
showed greater than 88% sleep efficiency (9). (g) The insomniacs were diagnosed as
having persistent psychophysiologic disorders of maintaining sleep (DIMS) (6). Subjects discontinued all daytime napping and sedative medications 4 weeks prior to
undergoing 5 nights of EEG monitoring in their homes.
The sleep of these subjects was documented and described in home all-night sleep
recordings (9). Six months following the recordings, the 24 subjects were contacted
again and asked to participate in the study reported here; 21 (10 insomniacs and 11
good sleepers) consented. As 6 months had elapsed since their participation in the
study of Coates et al. (9), we tested whether there had been significant changes in the
sleep habits of either group, by using self-reports from sleep diaries from five consecutive nights. The insomniacs reported no significant changes in sleep. The good sleepers
reported falling asleep in significantly less time during the current study than in the
previous study (mean = 21.08, t = 2.83, p < .01).
Two of the original good sleepers dropped out of the current study because of personal problems. We recruited two good sleepers, using the identical methods and criteria established by the aforementioned study (9).
Five of the insomniacs were married and lived with their spouses; four were single
and lived alone; one was single and lived with his parents. Five good sleepers were
married and lived with their spouses; three were single and lived with a roommate;
one was single and lived alone; one was divorced and lived with her two children; and
one insomniac was single and lived with his parents. Jobs held by insomniacs included
Sleep. Vol. 6, No.2, 1983
INSOMNIACS DURING THE DAY
149
student, secretary, coffeehouse employee, mechanical engineer, psychologist/teacher,
gem broker, security investor, telephone operator, and pilot (retired). Occupations held
by good sleepers included student, postdoctoral fellow in biochemistry, field engineer,
technical translator, electronics assembler, engineer, secretary, and teachers (two subjects). There was one housewife in each sleeper group. All subjects were employed in
9-to-5 jobs except the student and the retired pilot. The two groups were matched by
socioeconomic status.
Instruments
The Marchini Monitoring Inventory (MMI) was developed to measure subjects' moods
and perceptions of their environment, their physical and mental states, and their thoughts
and activities. A Likert scale was used to monitor 21 variables related to mood, thought,
and perceived environment. Two antonyms were placed on either end of the scale
(e.g., interested and indifferent; well and ill; worried and unconcerned); six boxes
separated the two adjectives. The subject placed an x in the choice which came closest
to his or her perception at the time of assessment. The inventory was printed on a 3.5
x 5.5 in. sheet with questions on both sides, making it compact and easily portable.
Reliability
Two alternative forms of the questionnaire were administered 20 min apart to the
study subjects during the time that the study was being explained to them by the
experimenter. Reliability determined using Cronbach's alpha (10) coefficients was
reasonably high on most variables for good sleepers and insomniacs. Alphas for good
sleepers and insomniacs were, respectively: noisy/quiet, .842, .847; attractive/gloomy,
.737, .889; enjoying/not enjoying self, .592, .789; upset/calm, .946, .910; depressed/
elated, .948, .848; hostile/accepting, .857, .871; tenselrelaxed, .928, .985; well/ill, .922,
.882; energetic/sleepy, .872, .900; mentally active/inactive, .682, .846; worried/unconcerned, .885, .836; self-critical/self-accepting, .603, .502. Alpha coefficients for the
other variables were physically uncomfortable/comfortable, - .989, .871; interested/
indifferent, .684, - .621. Negative scores indicated that group means remained stable
while individual scores changed. Low alphas indicated that those variables would have
to be sampled more extensively to obtain stable estimates of those variables; 12 times
per day was deemed adequate to reduce sampling variability.
Procedure
The experimenter traveled to each subject's home to initiate the study and to obtain
written informed consent. Each subject was issued six sealed envelopes, an electronic
pager, a spare battery, and all needed information concerning the pager's operation.
Each envelope contained 15 copies of the MMI and one daily sleep questionnaire. The
subject was instructed to open one envelope on rising in the morning, to complete one
daily sleep questionnaire about sleep on the previous night, and to complete one MMI.
Subjects were then asked to activate their pagers, keep them activated throughout their
waking hours, and to complete an MMI immediately when paged. If subjects were
unable to do so, they were asked to complete one as soon as possible and to write the
actual time completed on the MMI. To prevent mechanical failures, all subjects were
asked to telephone an experimenter if they were not paged by 10:00 a.m. each workday
morning and by 11 :00 a.m. on a leisure day. Subjects were also asked to remain within
Sleep. Vol. 6, Nu. 2, 1983
E. J. MARCHINI ET AL.
150
a 50-mile radius of Stanford University to ensure the functioning of their pagers. Subjects were instructed to complete a final MMI at bedtime, and to deactivate their pager
before retiring.
All subjects were monitored during 3 work days and 2 leisure days. Half the subjects
were monitored from Friday to Tuesday; the other half were monitored from Wednesday
to Sunday. The first day for each subject was considered an acclimation day. All data
were picked up by an experimenter each morning to reduce the probability that the
subjects would notice patterns in their feelings and moods and to allow the experimenters to reinstruct the subjects about correct completion of the forms, it that proved
necessary.
RESULTS
We gathered 1065 out of a possible 1407 (76%) valid MMI responses over the 5-day
experimental period. Totals of 467 repsonses were made on weekdays, 423 on leisure
days, and 175 on acclimation days. Causes for missing data were pager malfunctions
and subject failure in completing part or all of the MMI. Some subjects left items
unanswered when they felt the item did not apply to their feelings, moods, or thoughts
when paged. Some subjects failed to complete MMls because they were asleep when
paged, ignored the page, or had deactivated their pagers. This occurred when subjects
were in situations where beeper noise would be embarrassing (e.g., attending church,
viewing a movie, or interviewing for a new job). The 24% of MMI not completed was
equally represented in both good and poor sleepers.
Daytime activities and thoughts
Good sleepers and insomniacs reported significant differences in waking activities
(see Table 1). The insomniacs spent significantly more time shopping, watching tele-
TABLE 1. Percentage of time good and poor sleepers spent in
various activities
Stud ying/working
Watching television
Relaxing
Talking
Shopping
Other
Household chores
Drivinglriding
Eating/cooking
Going to bed
Dressing/grooming
Waking up
Physical activity
Good
sleepers
Insomniacs
x'
17.6
4.2
11.1
6.9
1.2
8.9
8.7
4.5
12.4
8.2
6.4
4.0
5.7
8.2
9.8
18.6
3.4
2.8
5.7
6.4
5.7
13.9
8.8
5.9
4.4
5.9
13.296"
8.965"
7.391b
4.920'
2.562 d
2.284
1.291
.576
.370
.086
.082
.078
.022
~41.923"
The sample consisted of 11 good sleepers and 10 insomniacs.
" p < .001; b P < .01; , p < .05; d P < .10.
Sleep, Vol. 6. No.2, 1983
151
INSOMNIACS DURING THE DAY
TABLE 2. Locations, social interaction, convenience when paged, and
past, present, and future thoughts
Good sleepers
Insomniacs
X'
Time spent in various
locations
Work
Other
Home
In transit
15.8
17.8
61.4
5.0
7.5
23.3
64.8
4.4
11.575'
2.882
.361
.131
lI4.949'
Number of people
with subject
None
1-2
3-4
5+
52.1
35.5
4.0
8.4
35.6
49.1
5.7
9.6
12.258'
8.658'
1.217
.298
l22.431'
Past, present,
and future thoughts
Don't know
Present
Futre
Past
9.8
62.2
23.1
4.9
7.8
65.3
21.8
4.9
.691
.305
.152
.0000
ll.148
Data are expressed as percentage of time.
a p < .001.
VISIOn, and relaxing; good sleepers spent significantly more time talking to people,
working, and studying. We found good sleepers at work and alone a greater percentage
of time (see Table 2).
Insomniacs and good sleepers differed in reported content of waking thoughts. Good
sleepers spent significantly more time thinking aobut their interpersonal relationships
and their family, whereas insomniacs thought more about their physical environment,
various forms of passive relaxation (watching television and reading), numerous news
items (the activites of the President, the state of the U.S. economy, etc.), and about
getting ready to go to bed (see Table 3).
The literature often suggests that insomniacs are unable to sleep because they ruminate over past or future problems. To test this, we scored both groups' thoughts
according to whether the subjects were thinking and/or worrying about a past, present,
or future problem. Table 2 shows that we found no significant differences during the
day between groups on this dimension.
Based upon the reported content of their waking activities and thoughts, good sleepers
appeared busier, more active, more involved in their work, and with other people.
Insomniacs seemed to be preoccupied with self.
Discriminant analysis
Each subject's mean score on each variable at each time was derived by averaging
responses across all 4 days for that particular hour of day. A total mean score on each
Sleep, Vol. 6, No.2, 1983
E. 1. MARCHINI ET AL.
152
TABLE 3. Percentage of responses of good and poor sleepers to
"What are you thinking?"
Response
Passive Relaxation
Family/personal
relationships
The news
Getting ready for bed
Your environment
Personal grooming
Working/studying
Household chores
General problems
Active relaxation
The day's routine
This research study
Physical/psychological
illness
Others and dreams
In transit
Nothing in particular
Hobbies
Eating/cooking
Good
sleepers
Insomniacs
8.4
16.6
10.852"
17.7
2.0
2.5
2.2
2.0
16.7
6.1
7.1
6.4
7.9
3.2
9.3
3.9
5.2
4.4
0.8
13.7
4.4
5.4
7.5
6.7
2.6
10.213"
5.16Jb
3.89P
2.846'
2.095
l.l51
1.102
.872
.353
.334
.251
6.1
2.5
1.2
1.7
2.0
4.4
5.4
2.4
1.6
2.1
2.3
4.4
.170
.164
.l31
.4
.3
X'
-..JlillL
~39.717a
a
p < .001;
b
P < .05;' p < .10.
variable for each individual was derived by averaging scores across time slots (hours
of day).
We wanted first to determine which variables discriminated between good sleepers
and insomniacs. Individual responses on each variable were compared, first by using
the chi-square test. The results are presented in Table 4. Significant differences between
the two groups of sleepers were found on all variables except attractive/gloomy. Insomniacs reported their environments to be quieter, more gloomy, and more physically
uncomfortable. In terms of mood states, insomniacs reported significantly more calmness, indifference, and depression, and less enjoyment and hostility. In reporting their
physical states, insomniacs indicated that they were less active, more relaxed, and
more likely to be ill or sleepy. Finally, in terms of mental state, insomniacs reported
being less active and concerned, and more self-accepting.
A multiple discriminant function analysis with free entry of variables was completed
next to determine if responses on individual variables could be used to classify good
sleepers and insomniacs into correct categories. The results are presented in Table 4.
The rank was determined by the magnitude of the Wilks' lambda; a smaller lambda
indicates greater discrimination between the two groups. A combined ranking was
computed as the absolute sum of the ranking of the lambda (smallest to largest) and
the chi-square (largest to smallest). As indicated in Table 4, insomniacs were passive
and accepting: they were more unconcerned, relaxed, quiet, accepting, self-accepting,
calm, mentally inactive, physically uncomfortable and inactive, sleepy, indifferent, not
enjoying self, and depressed.
Sleep, Vol. 6, No.2, 1983
153
INSOMNIACS DURING THE DAY
TABLE 4. Ranking of MMI variables
Variables
Worried/unconcerned
Tense/relaxed
NoisY/quiet
WelVill
Hostile/accepting
Self-critical/
self-accepting
Upset/calm
Mentally active/
inactive
Attractive/gloomy
Physically comfortable/uncomfortable
Physically active/
inactive
Energetic/sleepy
Interested/
indifferent
Enjoying self/not
enjoying self
Depressed/elated
Wilks
lambda
Ranking
0.149
0.565
0.445
0.354
0.635
0.713
I
I
6
4
2
9
12
2
3
4
5
6
0.656
0.575
10
7
7
8
15
5
0.356
0.807
3
14
9
10
14.84
14
0.543
5
11
23.02
25.07
13
11
0.593
0.668
8
11
12
13
33.92
8
0.826
15
14
25.05
12
0.764
13
15
Chi-square
Ranking
82.62
68.30
39.68
32.43
51.27
76.02
3
6
9
4
2
39.42
27.28
10
8.01
41.l5
I
7
Combined
Ranking
Variables in italics show responses more characteristic of insomniacs.
DISCUSSION
Good sleepers were busier, more active, and more involved with their work and with
other people than were the insomniacs. This description is similar to the one proposed
by Beutler et al. (4), who described sleep-maintenance insomniacs as emotionally seclusive and socially withdrawn.
Insomniacs spent more time thinking about and preparing for bed, various forms of
relaxation, and their immediate physical environment. The good sleepers reported
thinking more about their day's activities, their personal appearance, their everyday
problems, jobs, and interpersonal relationships.
Insomniacs reported being more unconcerned, quiet, ill, mentally and physically
inactive, uncomfortable, sleepy, indifferent, not enjoying themselves, and depressed.
Seligman et al. (11) have described three symptoms of depression similar to the
behaviors and moods reported here by the insomniacs in our sample: (a) passivity and
lowered frequency of response initiation; (b) a negative cognitive set, characterized in
our sample by reports of being ill, inactive, not enjoying themselves, and depressed;
(c) lack of aggression, characterized in our sample by reports of being calm, indifferent,
accepting, and not hostile. Depressed persons have also been described as more selfcritical and indecisive.
Increased activity, especially in the mornings and early evenings, may be helpful in
treating these insomniacs. Beck's (12) cognitive therapy is aimed at engineering success
experiences to increase activity and to change negative cognitive sets. Klein and Seligman (13) used a similar strategy to reverse learned helplessness and depression.
Sleep. Vol. 6. No.2. 1983
154
E. J. MARCHINI ET AL.
Clients may also be helped to plan activities and anticipate the positive experiences
that will follow from these activities (14).
It is not certain, of course, whether these results follow from or lead to poor sleep.
It seems reasonable, however, to intervene in specific moods and activities to determine their effects on sleep. By the same token, it would be interesting to assess experimentally the impact of the amount of sleep on these daytime variables.
Since the insomniacs reported functioning at lower levels of physical and mental
energy throughout the day, future researchers may want to determine if these daytime
mood differences and fluctuations between groups are related to circadian cycle differences. A replication of this study would be enhanced by simultaneous gathering of
physiological data (pulse rate, body temperature, and blood pressure) from each subject
at each MMI paging. Researchers then could plot the subjects' circadian cycles along
with their activities, thoughts, and perceptions of mood and physical and mental states
and then determine whether insomniacs' circadian rhythms were more depressed and
less varying than those of normals.
These data concerning insomniacs appear to differ from the more common clinical
descriptions of insomniacs as persons who (a) internalize and ruminate over problems;
(b) are in a continual state of emotional and physiological arousal; and (c) frequently
feel anxious, tense, self-doubting, insecure, worried, and agitated throughout the day
and especially at night (2). De la Peiia (15) hypothesized that there may be hyperactive
and hypoactive insomniacs. Hyperactive insomniacs may have sleep difficulties because of high physiologic, cognitive, and/or emotional arousal, whereas hypoactive
insomniacs have low physiologic, cognitive, and/or emotional arousal. The insomniacs
in this study did report being less involved with others, less physically and mentally
active and energetic, and more calm, relaxed, and unconcerned than the normals.
It is also possible that insomniacs are hypoactive during the day, but hyperactive at
night. Coates et al. (16) established that these same insomniacs reported increased
mental activity at night in comparison with good sleepers. Future research should test
the relationship between daytime hypoactivity and nighttime hyperactivity.
The discrepancy between these data and the more common clinical findings concerning insomniacs also may result from our innovative data collection methods. The
description of the insomniac as a worried, tense, ruminating, depressed, and anxious
person has been derived from the MMPI, the Taylor Manifest-Anxiety Scale, The Zung
Depression Scale, and the Cornell Medical Index (2,4,5). Future researchers may want
to correlate MMI and scores from the other personality inventories to see if both are
measuring the same constructs. Coursey et al. (5) found that the trait of sensationavoidance was able to predict levels of sleep efficiency. Insomniacs may generally be
unaware of their emotions and environments on a moment-to-moment basis, and thus
may report different moods and perceptions on different instruments.
The MMI might serve as a sensitive instrument for measuring the impact of daytime
behavior on improvements in sleep produced through drugs or nonpharmacological
therapies. Because insomnia is defined in terms of sleep and daytime fatigue, remedies
should presumably influence both. The next steps anticipated by this research should
involve (a) replications with new samples, (b) comparison of MMI responses with the
MMPI and other personality inventories and, perhaps, structured clinical interview,
Sleep, Vol. 6, No.2, 1983
INSOMNIACS DURING THE DAY
155
(c) studies of the relationship between variation throughout the day and circadian
cycles, (d) change in sleep correlated with changes in MMI responses produced through
therapy, and (e) change in MMI responses correlated with improvement in sleep.
Acknowledgment: This research was supported in part by grant no. MH 27551-03 from the
Clinical Research Branch of the National Institute of Mental Health, to Carl E. Thoresen.
REFERENCES
1. Coates TJ, Thoresen CE. Behavorial treatment for insomnia: some answers, few suggestions, and many
questions. In: Turner SM, Calhoun K, Adams HE, eds, Handbook of clinical behavior therapy. New
York: Wiley, 1980:240-89.
2. Kales A, Kales J. Sleep disorders: recent findings in the diagnosis and treatment of disturbed sleep. N
Engl J Med 1974;290:487-97.
3. Mendelson WB: The use and misuse of sleeping pills. New York: Plenum, 1980.
4. Beutler LE, Thornby n, Karacan I: Psychological variables in the diagnosis of insomnia. In: Williams
RL, Karacan I, eds, Sleep disorders: Diagnosis and treatment. New York: Wiley, 1978:383-94.
5. Coursey R, Buchsbaum M, Frankel B. Personality measures and evoked responses in chronic insomniacs. J Abnorm Psychol 1975;84:239-49.
6. Association of Sleep Disorders Centers. Diagnostic classification of sleep and arousal disorders, first
edition, prepared by the Sleep Disorders Classification Committee, HP Roffwarg, Chairman. Sleep
1979;2: 1-137.
7. Institute of Medicine. Sleeping pills, insomnia, and medical practice. Report of a study by the Institute
of Medicine, Washington, D.C., National Academy of Science, 1979.
8. Killen JD, Coates n. The complaint of insomnia: what is it and how do we treat it? Clin Behav Ther
Rev 1979;1(4):1-15.
9. Coates n, Killen JD, George J, et al. Discriminating good sleepers from insomniacs using all-night sleep
recordings conducted at home. J Nerv Ment Dis 1982;170:224-30.
10. Cronbach LF. Essentials of psychological testing. New York: Harper and Row, 1970.
II. Seligman MEP, Klein DC, Miller WR. Depression. In: Leitenberg H, ed, Handbook of behavior modification and behavior therapy. Englewood Cliffs, NJ: Prentice-Hall, 1975:92-143.
12. Beck AT. Cognitive therapy: nature and relation to behavior therapy. Behav Ther 1970:2:284-300.
13. Klein DC, Seligman MEP. Reversal of performance deficits in learned helplessness and depression. J
Abnorm Psychol 1976:85: 11-26.
14. Anton JD, Dunbar J, Friedman L. Anticipation training in the treatment of depression. In: Krumboltz
JD, Thoresen CE, eds, Counseling methods. New York: Holt, 1976:21-36.
15. De La Peiia A. Toward a psychophysiologic conceptualization of insomnia. In: Williams RL, and Karacan I, eds, Sleep disorders: diagnosis and treatment. New York: Wiley, 1978:101-44.
16. Coates n, Killen JD, Silverman S, et al. Cognitive activity, sleep disturbance, and stage specific differences between recorded and reported sleep. Psychophysiology (in press).
Sleep, Vol. 6, No.2, 1983