SLEEP IN THE ELDERLY:
EXPERIENCES, EXPECTATIONS, AND ADAPTATIONS
RUTH ANN MOONEY
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Appreciation is expressed to the members of the
dissertation committee: Dr. Marjorie White, Chair, Dr.
Molly Dougherty, and Dr. Melody Marshall from the
College of Nursing, Dr. Wilse Webb, from Psychology, and
Dr. Gordon Streib, form Sociology for their tireless
effort and expert guidance. I am grateful to Dr.
Margaret Wilson for her assistance in data analysis.
Appreciation is further expressed to Dr. William Hale,
M.D. and the staff of the Dunedin Clinic for their
invaluable assistance in data collection. Appreciation
is extended to Dr. Frank May for his assistance and
advice in retrieving background data. Thanks are
especially given to the subjects for their time, effort,
and interest in completing the questionnaire.
Special thanks are extended to my parents and son
for their support, cooperation, assistance and
TABLE OF CONTENTS
ACKNOWLEDGMENTS . . . . . . .
ABSTRACT . . . . . . . . ..
* . ii
* . V
I INTRODUCTION . . . . . ..
Problem Statement . . . . .
Objective Behavioral Theory of Sleep.
Adaptive Theory of Sleep ......
Restorative Theory of Sleep . .
The Sick Role . . . . . .
Operational Definitions . . . .
Summary . . . . . . . .
II REVIEW OF LITERATURE . . . ..
. . 10
. . 16
. . 19
. . 21
* . 22
Sleep Behavior Changes With Age . . .
Sleep Patterns. ... . . . . .....
Subjective Response to Sleep. .. ......
Developmental Changes in Sleep Patterns
Organic States that Affect Sleep . .
Individual Differences . . . ..
of Sleep. .
Self-Rating of Health .
III MATERIALS AND METHODS .
. . . 29
. . . 32
. . . . . 36
Subjects . .
Setting . .
Instruments . .
Procedure . .
Pilot . . .
Power analysis. .
Main study . ..
Data Analysis . .
IV RESULTS AND DISCUSSION.
Sleep Behavior . ....
Sleep latency. . .
Sleep length . .
Sleep continuity .
Sleep evaluation .
Sleep disturbances .
Help-seeking . ..
. . . . . 47
. . 83
i i i
Behavioral facilitators and inhibitors
of sleep . . . . . . . 87
Affective response to sleep disturbances 89
Summary . . . . . . . . . 92
V SUMMARY AND RECOMMENDATIONS . . . .. 95
Implications for Nursing . . . ... 99
Recommendations for Research . . .... 102
REFERENCES . . . . .... . . . ... 103
A SLEEP QUESTIONNAIRE . . . . . .. 108
B FREQUENCY OF USE OF SLEEP FACILITATORS AND
INHIBITORS AND THEIR EFFECTIVENESS. . 117
BIOGRAPHICAL SKETCH . . . . . . .. 124
Abstract of Dissertation Presented to the Graduate
School of the University of Florida in Partial
Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
SLEEP IN THE ELDERLY:
EXPERIENCES, EXPECTATIONS, AND ADAPTATIONS
Ruth Ann Mooney
Chairman: Marjorie White
Major Department: College of Nursing
The purpose of this research is to determine the
relationship between self-report of sleep behavior among
non-institutionalized elderly and their expectations
regarding sleep. Sleep facilitators utilized to manage
sleep disturbances and affective responses to sleep
behavior changes are also explored. It is important
that nurses understand sleep behavior changes
experienced by non-institutionalized elderly so that
effective nursing care can be planned and implemented
for clients in the community, hospitals, and
institutions. Findings from this research help to
validate the objective behavioral theory of sleep
proposed by Webb.
The sample consisted of 282 subjects over the age
of 65 who reported to the Dunedin Program clinic from
September 28, 1987, to November 25, 1987. Subjects took
the Sleep Questionnaires home to complete after
consenting to participate.
Subjects consistently rated their sleep as having
been better when they were younger. Subjects reported
having slept better, gotten enough sleep, felt more
rested upon awakening, enjoyed sleep more, and
experienced deeper sleep when they were younger.
Emotional stress was the most frequently selected
behavioral inhibitor of sleep. A disproportion was
found between the number of subjects who identified
emotional stress as a factor in sleep disturbance (n=67)
and the number of subjects who found relieving
emotional stress to be an effective sleep facilitator
Behavioral facilitators employed when sleep
disturbances occurred included relaxing, reading,
and praying. Few subjects (6%) sought help from health
care professionals or family and friends for sleep
disturbances. When asked how they feel when they
experience sleep disturbances, the most frequent
response indicated they accept the disturbance.
Implications for nursing include utilizing this
information to assess sleep patterns of the elderly, to
plan and implement nursing care, and to educate clients
experiencing sleep disturbances. Gerontological nurses,
as experts on normal age related changes, educate
other health care providers regarding age related
changes in sleep patterns and sleep facilitators the
elderly find useful in dealing with sleep disturbances.
Longitudinal research is needed to substantiate age
related sleep behavior changes. Further research is
also needed to determine the generalizability of sleep
facilitators to larger numbers of the elderly.
In 1980 25.5 million (11.7%) Americans were over
the age of 65 (Atchley, 1985), compared to approximately
50,000, or 2.5% of the population, in 1790. It has been
projected that, based on low birth rates and stable
death rates, in the year 2030, 51.6 million Americans
(20.9% of the population) will be over age 65.
The over-65 population is itself growing older. In
1960 almost 66% of the over 65 population was under 75
years old. Six percent of the over 65 year age group
was over 85 in 1960. By the year 2000, the young-old
(65 to 74 years) will represent 55.7% of the over 65
population, while the proportion of the over 85 year age
group will have risen to 10.4% of the over 65
population. Because of the differences in mortality
rates between males and females, this growing elderly
population will be predominantly female.
An increase in sleep disturbances has been found in
elderly subjects over that reported in other age groups
(McGhie & Russell, 1962). The elderly experience more
difficulty getting to sleep, a decreased ability to
sustain sleep and more early morning wakening than
younger persons (Webb, 1983). Women report sleep
difficulties more frequently and report more frequent
use of sleeping medications than men. Of those elderly
individuals habitually taking sedatives, a large
proportion started taking sedatives while hospitalized
(Lamberg, 1984). Since these medications are usually
ordered as needed, the nurse has a very important role
in assisting elderly clients in decision making
regarding the management of sleep behavior patterns.
Nationally, the elderly comprise 12% of
the population yet account for 31.2% of the total health
care expenditure. (Wolinsky, Mosely, and Coe, 1986).
The cost of care for persons with sleep disturbances
contributes to health care expenditure both directly and
indirectly. Direct costs are incurred for sedatives.
Indirect costs result from the side effects of
sedatives. Side effects of sedatives, including
confusion and incontinence, leading factors in nursing
home admission, are more prevalent and more serious in
the elderly than in younger persons. Decreasing the use
of sedatives in the elderly could substantially decrease
health care expenditures by decreasing the direct and
indirect costs related to the treatment of sleep
Little is known about what the elderly expect of
sleep patterns as they age. Sleep facilitators utilized
by the elderly when they experience sleep disturbances
and the attitudes of elderly subjects to sleep pattern
changes they experience have not been explored. It is
important to discover the normal sleep patterns of the
elderly who reside in the community so that the nurse
will understand when deviations from normal occur in
clients she or he is caring for. Nurses are responsible
for individualizing care for elderly when they are
hospitalized or institutionalized. It is important to
base that plan of care on a sound scientific base. This
includes knowledge of normal age related changes as well
as responses of the elderly to changes they experience.
Nurses play a very active role in educating elderly and
often instruct clients regarding normal age related
changes. In the area of sleep disturbances the nurse is
in a key role to acquaint elderly clients with normal
age related changes and strategies employed and found
effective by other elderly persons. Nurses employed in
the expanded role of nurse practitioner are able to
prescribe medications for the elderly. It is important
to understand sleep related sleep behavior changes in
order to develop protocols that are effective in helping
elderly clients deal with sleep disturbances.
Gerontological nursing is a new specialty area, and
the gerontological clinical specialist often assumes the
function of educating other members of the health care
team regarding normal age related changes and strategies
for dealing with these changes. It is important,
therefore, to determine what these changes are, their
prevalence, and effective strategies being utilized by
elderly living in the community.
This research is undertaken to determine the
relationship between self-report of sleep behavior among
well elderly living in the community as it may be
affected by expectations regarding sleep, sleep
facilitators utilized by elderly who experience sleep
disturbances, and sleep inhibitors common to the
elderly. These data are vital for nurses to develop
nursing interventions to assist the elderly in managing
sleep pattern changes. Physicians who are responsible
for a medical plan of care that includes large numbers
of sleeping medications for elderly clients can also use
these data. Elderly also need to have this information
for better management of sleep pattern changes they
Research to date has focused on sleep behavior
patterns experienced by the elderly. Several surveys
have been conducted (Domino, Blair, & Bridges, 1984;
Hayter, 1983; McGhie, & Russell, 1962; and Tune, 1969).
Electroencephalogram sleep recordings have also been
utilized in the study of sleep of the elderly (Agnew, &
Webb, 1971; Bixler, Kales, Jocoby, Soldatas, & Vela-
Bueno, 1984; Blois, Feinberg, Gaillard, Kupfer, & Webb,
1983; Carskadon, Brown, & Dement, 1982; Karacan,
Thornby, Anch, Holzer, Warheit, Schwab, & Williams,
1979; Roehrs, Zorick, Sicklesteel, Wittig, & Roth,
(1983); Reynolds, Coble, Black, Holzer, Carroll, &
Kupfer, 1980; Webb, & Schneider-Helmert, 1984; Webb,
1982). Webb (1983) conducted a study of sleep patterns
of elderly subjects which included sleep diaries, post-
sleep inventories, EEG recordings, and sleep
questionnaires. Observational studies of sleep in the
elderly have also been conducted (Webb & Swineburne,
1971; and Gress, Bahr, & Hassanein, 1981).
Expectations have been shown to be a factor in
defining a condition as a problem. Further,
expectations are developed based on previous personal
experience and in comparison with what others are
experiencing (Blau, 1964). Although elderly subjects
have been found to experience sleep disturbances, they
often do not report sleep complaints. It has been
suggested (Carskadon, VanDenHoed, & Dement, 1980) that
the elderly may accept the stereotyped view of sleep in
the elderly and believe that their disturbed sleep is
normal for people their age.
Sleep facilitators and inhibitors common to the
elderly, with the exception of sedative use, have not
been explored. Decreasing the amount of time spent in
bed at night to consolidate sleep for individuals who
perceive their sleep as disturbed and taking a nap
during the day if day tiredness is a problem have been
suggested (Carskadon, Brown, & Dement, 1982). The
relationship between expectations about sleep, and the
facilitators and inhibitors the elderly employ in
dealing with sleep pattern disturbances has also not
Several study questions were posed. What is the
sleep behavior reported by persons 65 and older residing
in the community? How do the elderly perceive their
sleep to have changed? What is the relationship between
sleep behaviors reported by the elderly and their views
of sleep in others their age? What is the relationship
between current sleep behavior reported by the elderly
and self-report of their previous sleep behavior? What
types of sleep facilitators do the elderly use to deal
with sleep pattern disturbances they experience? What
types of sleep inhibitors do the elderly exhibit? How
effective are sleep facilitators employed by the elderly
in dealing with sleep pattern disturbances? What are
the subjective reactions of the elderly to sleep pattern
disturbances? What is the affective response reported
by the elderly to sleep disturbances they experience?
Objective Behavioral Theory of Sleep
The objective behavioral theory of sleep developed
by Webb (in press) includes behavioral, adaptive, and
restorative components. Two theories of sleep, the
restorative theory and the adaptive theory, were in
existence prior to the development of the objective
behavioral theory of sleep. The objective behavioral
model of sleep is an attempt to reconcile the
differences between the two preexisting theories.
The behavioral component of the objective
behavioral theory of sleep (Webb, in press) consists of
modulating, intervening, and dependent variables.
Sleep behavior is the dependent variable and consists of
the presence or absence of sleep (sleep latency and
sleep length), sleep structure (sleep stages and
continuity of sleep), and subjective responses to sleep
(sleep evaluations and thresholds and dreams) (see
Adaptive Theory of Sleep
The adaptive theory of sleep proposed by Webb
(1974) is a precursor to the objective behavioral theory
of sleep. In the adaptive theory of sleep (Webb, 1974,
p. 1023) sleep is considered to be an adaptive
Presence or Absence of
Sleep (Sleep Patterns)
Subjective Responses to
Thresholds and Dreams
Adapted from Webb (in press)
Figure 1-1. Diagram of the Behavioral Component Within
the Objective Behavioral Theory of Sleep
non-responding and the particular characteristics of the
sleep of each species evolved from the adaptive role of
this process in the ecological niche of each species.
Webb addresses the semantic problem related to
describing sleep. Webb asks, "Is sleep behaving or not
behaving? If an animal is required to avoid continuing
a behavior, is this avoidance behavior or non-behavior?"
Webb resolves this issue by using the term non-
responding defined as "qualitatively minimal behavioral
engagement with the surround" (Webb, 1974, p. 1023) in
describing sleep behavior.
The primary postulate of the adaptive theory of
sleep is that survival requires periods of non-
responding. Secondly, sleep is an active process
controlling the behavior of the animal, resulting in
nonresponding. According to the adaptive theory of
sleep (Webb, 1974), three critical variables associated
with sleep of all species within each twenty-four hour
period are diurnal placement, intermittentcy, and total
amount. Diurnal placement refers to the time of day
sleep usually takes place. Some animals are nocturnal,
sleeping during the day and searching for food at night
while others sleep during the night and are awake during
daylight hours. Intermittentcy refers to the
continuity of sleep. Some species, for example cats,
sleep for short intervals several times during the 24
hour cycle. Total amount of sleep is the total number
of hours the animal sleeps. The primary correlates of
these variables are the search for food, and the
position on the predatory hierarchy of the particular
Species differences are further addressed in the
adaptive theory of sleep in that precocial animals,
those animals that are highly independent at birth, vary
little in the total amount of sleep time from birth to
adulthood. Altricial animals, those animals that are
highly dependent at birth and for a long time afterward,
including humans, vary greatly developmentally, with
infants sleeping much longer than adults.
Restorative Theory of Sleep
The restorative theory of sleep holds that
wakefulness causes a depletion or build up of some
substance that causes sleep to occur (Dann, Beersma, &
Borbely, 1984). Sleep serves to dissipate or buildup
whatever it is that either becomes depleted or increases
to an excess. The theory developed by Dann, et al.
(1984) is based on the presence of a sleep-regulating
variable (S) which increases during wakefulness and
decreases during sleep. Sleep onset is initiated when S
reaches an upper threshold (H) and awakening occurs when
the lower threshold is reached (L). The restorative
theory of sleep further holds that the thresholds
display a circadian rhythm that is controlled by a
single circadian pacemaker. The major shortcoming of
this theory is that S has not been found and cannot be
measured (Webb, in press).
The common elements of adaptive models of sleep and
restorative models of sleep are an interaction of sleep
demand and circadian tendencies and facilitators and
inhibitors of sleep (Webb, in press). Sleep demand
(Webb, in press) is the amount of time the individual is
awake preceding sleep. Prior wakefulness represents
sleep demand in the restorative theory of sleep (Dann et
al., 1984). Sleep demand is solely a time variable and
does not include activity or behavior within that time.
Time asleep decreases sleep demand (Dann, et al., 1984;
Webb, in press). Sleep latency, the time it takes to
get to sleep, has been found to be affected by prior
wakefulness (Webb & Agnew, 1971). The circadian
tendency of sleep behavior is the time of occurrence of
sleep within a twenty-four cycle. Circadian tendencies
have been described in both the adaptive theory of sleep
(Webb, in press), and in the restorative theory of sleep
(Dann et al., 1984).
External stimuli enhance or interfere with the
circadian tendency of the sleep-wake cycle (Webb, 1982).
Behavioral facilitators and inhibitors are behaviors of
the individual which make sleep more or less likely to
occur (Webb, in press). Anything that makes sleep more
likely to occur such as lying down is a sleep
facilitator. A sleep inhibitor is anything that keeps
the individual from falling asleep, for example,
standing upright. Behavioral facilitators and
inhibitors may be voluntary or involuntary. Behavioral
facilitators and inhibitors are complex and multiple and
include such things as body temperature, light, noise,
and stress level. External conditions described by
Dann, et al., 1984, are analogous to behavioral
inhibitors and facilitators (Webb, in press) and can
affect the threshold of S. The upper threshold has been
demonstrated to be altered in sleep deprivation studies.
Such environmental factors as darkness, warmth, and lack
of social stimulation can lower the threshold so that
sleep onset occurs. Naps can cause a depression of the
upper threshold. The ringing of an alarm clock may
precipitate a sudden rise in the lower threshold,
causing awakening (Dann et al., 1984).
The three intervening variables (Webb, in press),
sleep demand, circadian tendencies, and behavioral
facilitators and inhibitors, are modulated by four
additional variables (Webb, in press). These four
modulating variables are species differences,
developmental stages, organismic states, and individual
differences. These modulating variables affect the
intervening and dependent variables. Species
differences have been discussed under the adaptive
theory of sleep. Sleep behavior changes with
developmental stages in humans. Infants sleep a greater
proportion of the 24 hour cycle than adults. Naps are a
part of the infant's sleep cycle and are not a part of
the adult sleep cycle. Nap patterns seem to emerge
again in the elderly. Sleep demand, circadian
tendencies, and behavioral facilitators and inhibitors
change from infancy to old age (Webb, in press).
Organismic states are physiological characteristics
including such things as pain, medications, sedation,
stimulants, sleep apnea, and nocturnal myoclonus (Webb,
in press). Painful conditions such as arthritis are
more prevalent in the elderly than in younger adults.
The use of sedatives is higher in the elderly. Sleep
apneas and nocturnal myoclonus increase in incidence
with age. Individual differences occur within any
species, developmental stage or organismic state (Webb
in press). Some individuals require as little as 4
hours of sleep to function well, while others require
more than the average 8 hours to function well.
The behavioral component of the objective
behavioral theory of sleep (Webb, in press) is selected
for further study in the current project (see Figure
1-2). The intervening variables described in the
objective behavioral theory of sleep addressed in this
study are sleep facilitators and inhibitors and
circadian tendencies. Dependent variables described in
the objective behavioral theory of sleep addressed in
this study include (1) the presence or absence of sleep
as measured by sleep latency and length of sleep
episodes; (2) subjective responses to sleep consisting
of sleep evaluations; and (3) sleep structure limited to
continuity of sleep. Sleep stages, generally measured
using EEG, are not measured in this study. Sleep
thresholds and dreams are not addressed in this study.
The modulating variable species differences is not
dealt with in this research. Only humans are studied.
Individual differences are not dealt with in this study.
The design used provides some information regarding
developmental stages, and organismic states, the two
remaining modulating variables described by Webb (in
press). If sleep behavior changes are developmental,
one would expect to detect changes in all subjects with
Intervening Variables Dependent Variable
Behavioral Facilitators-4 --- Sleep Behavior
or Inhibitors Sleep patterns
Circadian Tendencies Sleep length
Developmental Stages Organismic States
Figure 1-2. Objective Behavioral Theory of Sleep:
Schema of Selected Components
aging. Asking subjects questions about their sleep when
they were younger should shed some light on the
developmental nature of sleep patterns. Data obtained
from the retrospective portion of the questionnaire must
be interpreted with caution. Individuals are being
asked to recall sleep patterns experienced when they
were younger and their responses may be influenced by
memory as well as events that have occurred in the time
interval between then and now.
Organismic states are addressed by reviewing data
from elderly individuals living in a community to
determine the relationship between recent major illness
and sleep behavior. Self-ratings of health are
available for a segment of the population and these will
be utilized in relating organismic differences to sleep
The Sick Role
Although sleep disturbances are not illnesses,
elderly individuals often think of them as illness, and
seek help in the same manner individuals experiencing
other illnesses seek help. The sick role as described
by Parsons (1951) allows a framework for explaining this
perspective. Parsons defines illness as:
a state of disturbance in the "normal" functioning
of the total human individual, including both the
state of the organism as a biological system and of
his personal and social adjustments. It is partly
biologically and partly socially defined. (p. 431).
Parsons describes four aspects of the
institutionalized expectation system related to the sick
role. The first aspect is that the sick individual is
excused from normal social role responsibilities. The
excuse is related to the nature and severity of the
illness. Second, the individual is helpless to "will"
himself well. Third, the individual has an obligation
to get well. Fourth, the individual is expected to seek
competent help and to cooperate in trying to get well.
The urgency of the need for help varies with the
severity of the disability, suffering, and risk of death
or serious, lengthy, or permanent disability. The
situation of illness places the patient and those close
to the patient in a situation of strain.
Sleep disturbances can be discussed within the
framework of Parson's sick role. The person
experiencing sleep disturbances may be excused from
social roles because of being perceived as too tired to
perform these roles adequately. This judgement is
directly related to the severity and length of the sleep
disturbance. Granting of sick role status is related to
the degree of responsibility for the sleep disturbance
assessed to the individual. If the individual does such
things as consume caffeineated drinks immediately prior
to retiring, that individual will less likely be granted
the privileges of the sick role. The individual is
expected to try to help him/herself by reducing caffeine
intake, decreasing stress, and using remedies such as
drinking milk before retiring. If the sleep disturbance
persists the individual is expected to seek competent
help, usually from a physician. The individual is then
expected to cooperate with the physician and carry out
the prescribed therapy. Death or permanent disability
are not expected to be caused by sleep disturbances.
However, individual suffering from severe and prolonged
sleep disturbances may fear that their mental health
will be affected. Side effects of medications used to
treat sleep disturbances may be severe.
Sleep disturbances experienced by one member of the
family may place other members of the family under a
strain. Family members may be asked to alter their
routines to accommodate the sleep disturbed individual.
The sleep disturbed individual may be excused from
family roles which need to be adopted by other family
members. It seems relevant to address help-seeking
behavior of the elderly with regard to sleep
disturbances from the perspective of Parson's sick role.
Several operational definitions are utilized in
this research. They are as follows:
Sleep behavior. Sleep behavior is comprised of
sleep patterns, sleep structures, and subjective
responses to sleep.
Sleep Pattern. Sleep pattern is the presence or
absence of sleep as identified by sleep latency and
length of sleep episode.
Sleep latency. Sleep latency is the length of time
subjects report it takes them to get to sleep.
Length of sleep episode. The length of the sleep
episode is the duration of sleep as measured by the
average number of hours subjects report sleeping at
Sleep structure. Sleep structure is limited to
continuity of sleep and omits sleep stages in this
Continuity of sleep. Continuity of sleep consists
of the number of awakenings, the length of time awake
after falling asleep, and the frequency and length of
Subjective response to sleep. Subjective response
to sleep is limited to sleep evaluations and eliminates
thresholds and dreams in this study.
Sleep evaluations. Sleep evaluations include the
subject's estimate of how well they sleep; self-rating
of the amount of sleep they get; how rested they feel
when they wake up; how much they enjoy sleep; and how
light or deep they report their sleep.
Circadian tendency. Circadian tnedency is the
placement of sleep within a 24 hour cycle as measured by
the time subjects report as their bedtime and time of
getting out of bed on weekdays and on weekends.
Circadian tendency also includes the reported weekday
bedtime of subjects compared to their average weekend
bedtime; the average weekday time of awakening compared
to their average weekend time of awakening; the use of
an alarm clock; the number of days per week subjects
report going to bed more than one hour earlier or later
than their average bedtime.
Sleep disturbance. Sleep disturbances include the
reported frequency of; (1) difficulty getting to sleep;
(2) awakening during the night without being able to
return to sleep: and (3) awakening earlier than expected
without being able to return to sleep.
Behavioral facilitators. Behavioral facilitators
are strategies elderly subjects identify as those they
use to help them when they experience sleep disturbances
and which are considered effective in dealing with sleep
Behavioral inhibitors. Behavioral inhibitors are
those variables identified by subjects as possible
causes of overall sleep problems; and strategies elderly
subjects employ which have been shown to have a
detrimental effect on sleep (i.e. alcohol or
Affective responses to sleep disturbances.
Affective responses to sleep disturbances are the
reported reactions to sleep pattern disturbances.
Elderly. Elderly are individuals age 65 or older.
The objective behavioral theory of sleep as
described by Webb (in press) is used as the foundation
for this research with the focus on exploring the
behavioral components of the theory. This research is
undertaken to determine the relationship between sleep
behavior of the elderly and expectations regarding
sleep. Sleep facilitators employed by the elderly as
they encounter sleep disturbances are also investigated.
Affective reactions to sleep disturbances experienced by
elderly subjects are also explored. Help-seeking
behavior for sleep disturbances will be examined within
the framework of Parson's sick role.
REVIEW OF LITERATURE
Sleep Behavior Changes With Age
Research findings related to sleep pattern changes
in the elderly suggest that there is a great variability
in sleep behavior in elderly subjects (Hayter, 1983).
As clients age, changes in sleep behavior are reported
in subjects over age 75 with still greater changes seen
in subjects over 85 year of age (Hayter, 1983; McGhie &
Russell, 1962). The primary sleep related changes that
occur with aging are difficulty getting to sleep,
awakening within sleep, sleep stage liability, longer
time to return to sleep following arousal and early
awakenings. (Webb, 1983).
Sleep latency, the time it takes to fall asleep, is
longer in older subjects (Agnew and Webb, 1971; Hayter,
1983; and Webb and Schneider-Helmert, 1984). Agnew and
Webb (1971) compared the sleep latencies of seven age
groups and found that subjects aged 16 to 69 had sleep
latencies of one to five minutes. The oldest age group,
60 to 69 years, exhibited the longest sleep latencies
with a mean of 15 minutes to fall asleep. More frequent
night awakenings occur in those over 85 (Hayter, 1983;
Tune, 1969; Webb, 1982a). Sleep patterns were evaluated
in the laboratory and total wake time was found to be
correlated positively with age (Bixler, et al., 1984).
This difference was due primarily to an increase in the
number and duration of night awakenings in subjects in
the 50 to 80 year group. Webb and Swineburne (1971)
observed night awakenings in subjects ranging from two
to five per night. Awakening during the night increased
with age (Tune, 1969) with subjects over 50 having the
most disturbed sleep.
Differences in sleep behavior changes between men
and women have not been found to be significant
(Carskadon, et al., 1982; Webb & Schneider-Helmert,
1984; Webb, 1982c). Women report more pre-sleep
difficulties (McGhie & Russell, 1962) including longer
sleep latency (Hayter, 1983; Webb & Schneider-Helmert,
1984). A greater number of awakenings for men has been
reported (Bixler, et al., 1984; Reynolds, Kupfer, Taska,
Hoch, Sewitch, & Spiker, 1985; and Webb, & Schneider-
An increase in the number of naps has been found in
those over 75 years of age (Hayter, 1983) In subjects
over 85, the total amount of nap time increases with no
increase in the number of naps (Hayter, 1983). Tune
(1969) reported an increase in the number and duration
of naps in older subjects. Webb and Swineburne (1971)
observed the sleep of 19 elderly subjects for two 36
hour periods and found that almost all of the subjects
took one or more naps. Great variability was found with
one subject not taking any naps while four subjects took
three naps a day. The length of naps was not related to
night sleep and was judged to be neither compensatory
nor limiting relative to night sleep.
Webb and Agnew (1974) studied 14 subjects for 14
days in a time-free environment to determine if sleep
patterns were dependent on exogenous or endogenous
factors. All subjects displayed a rhythm greater than
24 hours. Large individual difference in the amount of
variance from 24 hours were seen. The overall
distribution of sleep stages did not differ from
baseline recordings. Webb and Agnew (1974) concluded
that overall sleep patterns remain stable in a time-free
environment. Prior wakefulness, length of sleep, and
sleep onset time continue to affect sleep. This is
taken as-evidence for the stability of the structure of
the sleep process.
Subjective Response to Sleep
Karakan, et al., (1979) studied 1645 subjects
ranging in age from 18 to over 70. A steady increase in
the percentage of subjects reporting trouble sleeping
occurred with age. Seventy-five percent of the 18-19
year age group reported never or seldom having trouble
sleeping while 50% of those over 70 reported never or
seldom having trouble sleeping. Twenty-five percent of
subjects over 70 reported having trouble sleeping often
or all the time. In the 18-19 year age group, 6% of the
subjects reported trouble sleeping often or all of the
time. Females in the older age ranges reported more
difficulty sleeping than males. When asked what kind of
sleep problems they experienced, a higher percentage of
respondents over age 40 indicated trouble staying
asleep or waking too early.
McGhie and Russell (1962) found that despite sleep
changes, the aged tended to complain less of morning
tiredness until age 75 when there was a significant
increase in the incidence of reports of day tiredness.
Carskadon, et al., (1982) reported that only transient
arousal and respiratory events had a statistically
significant relationship to daytime sleep tendency.
Continuity of sleep was of greater significance to
daytime well-being than total sleep time.
Webb and Levy (1982) found that older subjects were
affected more by acute sleep deprivation than younger
subjects. Although many sleep pattern changes occur
with aging, few are related to daytime tiredness (McGhie
& Russell, 1962). Day tiredness does not significantly
increase until age 75. Women and those over 75 are more
likely to take sedatives. This is of great importance
because the over 75 age group is the most rapidly
growing cohort, and women constitute a high percentage
of that cohort.
Webb and Schneider-Helmert (1984) use the terms
"disturbance" rather than disease and "category" rather
than symptom when describing sleep changes. This
reflects their belief that sleep changes are associated
with the aging process.
Developmental Changes in Sleep Patterns
Webb (1982b) was able to study a second time the
sleep of five subjects who had undergone sleep
recordings of four nights 15 years earlier. The
recordings at an older age showed a sharp increase in
awakenings. Rank order correlations for time one and
time two resulted in .55 correlation on number of
Organic States That Affect Sleep
Organic states that affect the sleep of the elderly
include sleep apnea, nocturnal myoclonus, arthritis,
heart disease, respiratory disease, Alzheimer's disease,
blindness, and nocturia. (Milne, 1982; Quan, et al.,
1984; and Schirmer, 1983). Roehrs, Zorick, Sicklesteel,
Wittig, and Roth (1983) conducted a study of 562
patients referred to a sleep disorder clinic because of
sleep disorders. Of those patients seen, 97 were in the
61 to 81 year age group, 264 were in the 41 to 60 year
age group and 202 were in the 20 to 40 year age group.
Nocturnal myoclonus was the most frequent problem of the
elderly. Sleep disturbances related to drug and alcohol
use were higher in the oldest age group than in the
other two groups. Other disorders that contributed to
sleep disorders included sleep apnea, psychophysiologic
and psychiatric disorders, medical disorders and
circadian rhythm disturbance.
Sleep apnea is a condition in which the individual
experiences episodes of apnea throughout the night
(Quan, et al., 1984). Sleep is interrupted by apneic
episodes that are often followed by loud snoring. Sleep
apnea is often most apparent to the bed partner, and may
go unrecognized by the individual experiencing sleep
apnea. Because of the sleep fragmentation, the
individual often experiences daytime hypersomnolence.
Nocturnal myoclonus is a condition in which the
individual experiences periodic leg movements throughout
the night that result in sleep fragmentation (Quan, et
al., 1984). Excessive daytime hypersomnia may result.
Arthritis and other painful conditions may prevent
the elderly individual from getting to sleep easily or
awaken the individual during the night. (Quan, et al,
1984). Arthritis is more frequent in the elderly than
in younger age groups. Elderly individuals with heart
disease may be awakened from sleep because of pain or
because of difficulty breathing (Quan, et al., 1984).
Respiratory disorders, Alzheimer's disease, and
blindness may also lead to sleep fragmentation and
increased daytime somnolence (Quan, et al., 1984).
Nocturia, getting out of bed at night to void,
is more common in the elderly than in younger
individuals (Schirmer, 1983). This may be due to
prostatism in males or to an increased use of diuretics
in either males or females.
Great variations in sleep behaviors are seen
between individuals. The average number of hours adults
sleep is 7.5 with a standard deviation of one hour
(Webb, 1983). This means that an individual may sleep
6.5 to 8.5 hours and be within one standard deviation of
the norm for sleep. Ninety-nine percent of individuals
will sleep between 5.5 and 9.5 hours. Webb and
Swineburne (1971) observed the sleep of nine men and 10
women. They found great variability between subjects.
Hayter (1983) studied sleep of 212 noninstitutionalized
subjects aged 65 to 93 and found great variability even
among subjects in the same age group. Hayter found
variability in time spent in bed, sleep latency, number
and amount of time awake after sleep onset, number and
length of daytime naps, time going to bed, and time
Behavioral Facilitators and Inhibitors of Sleep
Organic disease such as arthritis, sleep apnea,
nocturnal myoclonus, cardiac disease, respiratory
conditions have been indicated in sleep disturbances.
Treating the underlying cause of the sleep disturbance
may be all that is necessary to restore adequate sleep
(Quan, et al., 1984). Exercise programs and relaxation
training have also been suggested (Milne, 1982; Quan, et
al., 1984, Schirmer, 1983). Praying may be a soothing
activity for some individuals (Schirmer, 1983).
Education regarding normal sleep pattern changes that
occur with age has also been recommended (Milne, 1982;
Quan et al., 1984, Schirmer, 1983). Eating a light
snack or drinking a glass of milk before retiring have
been suggested (Milne, 1982; Schirmer, 1983). Drinks
which contain caffeine should not be ingested within two
hours of bedtime (Milne, 1982). A regular sleep
schedule to regulate the biological clock has been
recommended (Schirmer, 1983). For difficulty getting to
sleep, getting out of bed and listening to music or
reading may be helpful (Schirmer, 1983).
Older subjects were more likely to report sedative
use than younger subjects (McGhie & Russell, 1962;
Hayter, 1983; Karacan, et al., 1979). In 1981, twenty-
one million prescriptions were written for sleeping
pills (Lamberg, 1984). An increase in the use of
sedatives has been found in the 85-year-old age group
(Hayter, 1983). Many people who report habitually
taking sleeping pills state that they began taking
sleeping pills while they were in the hospital (Lamberg,
1984). Sedatives are usually ordered as needed,
therefore the nurse has great latitude in administering
these medications. Sedative use can result in side
effects such as ataxia, paradoxical excitement,
confusion, incontinence, and even death in individuals
with sleep apnea (Quan, Bamford, & Beutler, 1984).
Confusion and incontinence are leading factors in
nursing home admissions of the elderly. Nurses play a
pivotal role in sedative use among the elderly.
Lamberg (1984, p.139) advises against the use of
sleeping pills if one is "middle-aged" or older because
of the increased incidence of ataxia, the decreased
ability of the liver to detoxify the drug and the
incidence of interaction with other drugs. Other
conditions frequent to the elderly in which the authors
warn the use of sleeping pills include the presence of
respiratory disease, liver or kidney disorders, alcohol
consumption, and a history of having taken sleeping
pills before without improvement in sleep.
Problems associated with sedative use include
confusion, paradoxical excitement, morning "hangover"
perhaps accompanied by ataxia, interactions with other
drugs, dangers in specific illness such as sleep apnea,
and distortion of natural sleep patterns (Hartman,
1980). Because of decreased ability to adapt, and
increased incidence of chronic illness, the elderly are
more likely to experience adverse effects from sedatives
(Quan, et al., 1984).
Most sleep inducing agents lose their effectiveness
within a week or two and influence sleep stage pattern
during their administration (Borkovec and Fowles, 1973).
Many cases of insomnia do not involve physiological
causes such as pain but are caused by psychological or
environmental factors. (Borkovec and Fowles, 1973).
Psychological factors include such things as stress,
anxiety, and fear, while environmental factors include
such things as noise, light, and temperature. The most
important biological cause of insomnia has been
identified as drugs (Borkovec et al., 1973). Sedatives
are of questionable value in the elderly and may
contribute to a reduced quality of sleep rather than
improved sleep (Colling, 1983). In 1974, one-third of
all prescriptions for secobarbital and diazepam were for
patients over 60 years of age (Quan, et al., 1984).
Excessive use of alcohol is associated with the
same kind of sleep problems associated with sedatives
(Milne, 1982; Quan, et al., 1984). Individuals may fall
asleep more readily after consuming alcohol, but are
likely to awaken and be unable to return to sleep (Quan
et al., 1984).
In this research, sleep facilitators employed by
the elderly when they experience disruptions of sleep
patterns are explored. Sleep inhibitors exhibited by
the elderly are also explored.
Self-Rating of Health
An older person with a given set of health status
characteristics is more likely than a younger person
with similar characteristics to rate his/her health
better than that of peers the same age (Stoller, 1984).
Older persons are more likely to rate their health
positively than younger persons (Cockerham, Sharp, &
Wilcox, 1983). The old-old, those over age 75, assess
their health at a higher level than do the young-old
exhibiting similar health characteristics (Linn, & Linn,
Symptoms themselves do not define a person as sick,
but rather serve as "triggers that raise questions for
further exploration" (Zola, 1966). Self-assessments of
health among older people seem to be influenced by
comparisons with their peers (Cockerham, et al., 1983).
In a study conducted by Stoller (1984) most older
respondents rated their health on the upper (positive)
rungs of the Cantril ladder (Cantril, 1965). In this
study subjects were presented with a picture of a ladder
with the rungs numbered from zero on the bottom rung to
nine on the top rung. Subjects were asked where they
would place themselves, where they would place others
their age, and where they thought they would be in five
years. The majority (63.8%) of the elderly subjects in
this study (Stoller, 1984) rated their own health better
than they rated the health of other people their own
age. Stoller concludes that self-assessments of health
by the elderly as better than others of the same age may
reflect misconceptions about health by the elderly. The
elderly may believe that a marked decrease in level of
health and functional decline are inevitable with aging.
Because of these misconceptions, when the elderly do not
experience the declines they expect with aging, they
consider themselves the exception and rate their health
better than they rate the health of their peers.
It can be concluded, therefore, that expectations
of health influence the way the elderly describe their
health. In this research, it is anticipated that
expectations regarding sleep disturbances in aging will
correlate with self-rating of sleep by the elderly. It
is anticipated that elderly individuals experiencing
sleep disturbances will rate their sleep more poorly
than they rate the sleep of others their age.
Some elderly persons may accept sleep pattern
changes as part of aging and not define sleep changes
they are experiencing as problems (Quan, et al., 1984).
Other elderly individuals interpret sleep pattern
changes as being abnormal and define changes as problems
(Quan, et al., 1984). Lack of knowledge on the part of
the elderly concerning age related sleep changes
(Hayter, 1983; and Quan, et al., 1984) and the great
variability of sleep patterns exhibited by the elderly
(Hayter, 1983) contribute to the confusion. This
research is undertaken to further elucidate the
behavioral aspects of Webb's Objective behavioral Theory
of Sleep (in press) by describing sleep behaviors
experienced by the elderly, their expectations regarding
sleep behavior, help-seeking for sleep disturbances,
sleep facilitators and inhibitors common to the elderly
and affective responses to sleep behavior changes.
MATERIALS AND METHODS
Using a survey design, 336 subjects from the
Dunedin Clinic, Dunedin, Florida were given
questionnaires to complete. All eligible clients of the
Dunedin Clinic who had their annual health assessment
during the period from September 28, 1987 to November
25, 1987 were invited to participate in the main study.
Subjects ranged in age from 65 to 94 with a mean age of
78.3. A total of 282 subjects returned the sleep
questionnaire in a useable form. One hundred sixty-five
(57.9%) respondents were female and 87 (35.4%) were
male. Demographic data was not available for 33
subjects. All subjects were living in the community.
The Dunedin Program is a longitudinal study of 4200
elderly residents of Dunedin, Florida, funded in part by
the James Hilton and Emma Austin Manning Foundation
since 1975. The Dunedin Program is conducted at the
Dunedin Clinic. The sole function of the Dunenin Clinic
is to conduct the Dunedin Program. The purpose of the
Dunedin Program is to study healthy elderly individuals
living in the community for unidentified health
disorders. The study is also conducted to generate
information about the prevalence of disease, signs and
symptoms, and to determine biochemical norms for this
age group. (Hale, Marks, & Stewart, 1980).
Each year subjects complete a detailed
questionnaire about their health including a detailed
medication assessment. Subjects then report to the
Dunedin Clinic where an electrocardiogram is performed,
and biochemical analysis of a blood sample is conducted.
Each subject's blood pressure is assessed. Subjects are
referred to their private physician for treatment if
health problems are discovered.
The Dunedin Clinic provides access to a readily
available subject pool from which subjects can be drawn.
All subjects are over the age of 65, and are living
independently in the community. Longitudinal data are
available on subjects and it may be possible to continue
to study subjects in this project in the future.
Dunedin, Florida is located on the western coast of
Florida. Dunedin is located in Pinellas county, the
most densely populated county in Florida with 4.4
persons per acre. The economy of Dunedin is dependent
on tourism and retirement living. According to the
Chamber of Commerce (1986) the permanent population
estimate of Dunedin is 35,292. Approximately 43 percent
of Dunedin's population is sixty years of age or older
with females making up approximately 54.7% of the total
population and 58.1% of the over sixty population.
White persons constitute 98.7% of the population.
Educational information indicates that 72.4 percent of
the Dunedin residents 25 years of age and older are high
school graduates. Eighteen percent have had some
college and 16 percent have had at least 4 years of
The Sleep Questionnaire developed by White (1975)
(See Apprendix A) was utilized to assess subjects'
current sleep patterns, their previous sleep patterns,
and their thoughts about the sleep patterns of others
their age. Items were added to the Sleep Questionnaire
to determine mechanisms employed by the elderly when
they encounter sleep disturbances of sleep onset
latency, awakening during the night, and early morning
awakening. The frequency of the use of these
mechanisms, and the subjects' perception of the
effectiveness of these mechanisms were assessed.
Subjects' affective responses to sleep disturbances were
also assessed. The final questionnaire consisted of 94
items. Webb and Stone (1963) administered a
questionnaire related to sleep to 104 undergraduate
students over a two year period. Test-retest
reliability was .82 for the first class and .83 for the
second class for length of sleep. Sample two was asked
a question regarding length of time to function at full
capacity upon awakening. Test-retest reliability was
Reliability of the Sleep Questionnaire used in this
study has been established in an investigation which
compared the measurement of sleep behavior using this
questionnaire versus using a daily diary (White, 1975).
Subjects were asked to complete the Sleep Questionnaire
as a pre-log and as a post-log. Subjects were asked to
maintain a sleep diary for two or four weeks. Sleep
length had a correlation of .70 between the pre-log
questionnaire and the diary. Average weekday bedtime
correlated .78 between the pre-log questionnaire
estimate and the diary. Average weekend bedtime
correlated .66 between pre-log questionnaire and diary.
Sleep diary mean weekday wake-up time and questionnaire
correlated .74. Average weekend wake-up time and
questionnaire estimate correlated .58. Post-log
questionnaire and diary correlations were also reported.
These correlations tended to be higher, presumably
because subjects had been paying attention to their
sleep behaviors for two or four weeks. The more
conservative pre-log questionnaire correlations were
selected for this study because they more accurately
correspond to the method used in this study.
Webb, Bonnet, and Blume (1976) tested a post-sleep
inventory to determine construct validity. The
researchers found that subjects rated their sleep "good"
or "bad" by comparing their sleep with their "usual"
night, their sleep over a few weeks or several years, or
by comparing their sleep to the way they thought others
slept. The scale developed was found to have high
Content validity was claimed for the instruments.
The Sleep Questionnaire was developed by an expert in
the field of sleep research. The questionnaire reflects
the literature of sleep research. Following a study of
102 college students, White (1975) concluded that
estimates of the expected range of average sleep
behavior for a group utilizing the Sleep Questionnaire
Items related to adaptations to sleep disturbances
were developed in consultation with two sleep research
experts. Literature was reviewed to determine
recommendations for improving sleep. Items which could
be used in the questionnaire were then developed. Sleep
difficulties were divided into difficulties getting to
sleep, difficulty staying asleep and awakening earlier
than desired. Subjects were asked to indicate which
measures they used when they had difficulty at any or
all of the three times. Subjects were also asked to
rate the effectiveness of strategies they employ when
they encounter sleep difficulties.
Demographic data as well as information regarding
current health status, recent illnesses, and recent
hospitalizations were obtained from the Dunedin Program
Questionnaire. This questionnaire is completed annually
by all clients of the Dunedin Clinic.
Pilot. A pilot study was conducted during the week
of September 7, 1987 to test the use of the Sleep
Questionnaire with the participants enrolled in the
Dunedin Program. All subjects (n=24) who visited the
Dunedin Clinic for their annual assessment were asked to
complete the questionnaire and mail it to the
investigator within two weeks. Fourteen of the subjects
who returned the questionnaire to the researcher were
interviewed by telephone. Subjects were asked to
describe any difficulties they had with the
questionnaire, how they rate their sleep, how they
compare their sleep to others their age, what they do if
they have trouble sleeping, how they define "younger",
and what additional questions they might have asked.
Data obtained in the questionnaires and the telephone
interviews were compared to determine consistency.
Methodological problems encountered in conducting
the pilot study were corrected before the main study
began. One question was added to the questionnaire and
several questions were modified as a result of this
The added question required subjects to indicate
whether or not they have a friend or relative who
experiences trouble sleeping. Many of the subjects in
the pilot did not answer questions about sleep of other
elderly persons. This question was added to provide a
focus for subjects related to sleep of other elderly
persons. It was hoped that this would increase the
number of responses to questions about sleep of other
elderly. Questions about sleep of others were changed
from comparative questions to questions requiring a
numeric answer. For example, in the pilot questionnaire
subjects were asked "Do you think others your age sleep
better, worse, or the same as you do?". This question
was changed to "How well do you think others your age
sleep?" The choices for responses were; 1) very well,
2) satisfactorily, 3) some problems, 4) poorly, and 5)
don't know. This was to increase the consistency of the
questions and to increase the response rate of those
items. A "don't know" response was included in all
questions related to sleep of others.
Power analysis. A power analysis as described by
Marks (1982) and Cohen (1977) was conducted to determine
the required number of subjects for this study. A pilot
study was conducted on 24 subjects and from the data
obtained in the pilot study, it was determined that 39%
of the pilot sample chose the response "very well" to
describe their sleep at the present time. The value 39
was used as the estimated P (the percentage of
observations that fall into a particular category) to
determine the number of subjects required for the main
study. The confidence coefficient of .95 was chosen.
The confidence coefficient "reflects our confidence that
the computed interval will contain the parameter of
interest" (Marks, 1982, pg. 122). In this study it is
95% probable that the estimated sleep length obtained,
for example, will fall within the true, unknown, average
sleep length. A bound on error of .06 was selected. To
continue with the previous example, a bound on error of
.06 means that the estimate of sleep length obtained
will be within 6% of the true sleep length. A sample
size of 257 was estimated to be needed for the study
(Marks, 1982, pg. 132) based on the return rate
experienced in the pilot study.
Main study. The 336 subjects attending the clinic
during the time of this study represent 14% of the 2400
subjects currently being seen at the Dunedin Clinic.
Clients with physical or psychological problems such as
blindness or Alzheimer's disease were excluded. All
other clients who presented at the clinic during the
time period of the study were invited to participate in
this study. The time period was defined to end prior to
Thanksgiving and the beginning of the Christmas holidays
because of the increased incidence of depression during
the Christmas holiday. The relationship between
depression and sleep disturbances is well documented.
The Sleep Questionnaire was given to each subject
who consented to be included in the study at the
completion of his/her annual visit. A total of 336
questionnaires were distributed over a six week
interval. The nurse who gave the questionnaire to
subjects described the purpose of the study, and
encouraged the subject to complete and mail the
questionnaire to the researcher within two weeks.
Subjects were asked to sign a consent form permitting
access to data from their most recent Dunedin Program
Questionnaire. Questionnaires for this study were
identified using the same subject number that is used in
the Dunedin Program. The subject number was known only
to the researcher and the research team of the Dunedin
Clinic. Subjects were advised that the same subject
number was being used for this study as is used in the
main Dunedin Program study. A stamped, addressed
envelope was given to subjects for returning the
questionnaire to the investigator at her home.
Telephone calls were made as reminders to subjects who
had not returned their questionnaires within two weeks.
A total of 285 questionnaires were returned for a return
rate of 84.8%.
Much of the data are presented using descriptive
techniques. The number of subjects who responded to
each item is different. The highest number of subjects
responded to questions regarding their sleep patterns at
present, with fewer subjects responding to questions
about their sleep when they were younger, and still
fewer subjects responded to items about the sleep of
others. Of the 336 questionnaires distributed, 282 were
completed and returned in a useable form. Frequencies
and percentages of subjects responding are reported.
Several subscales exist within the Sleep
Questionnaire. These subscales were used to assess the
rhythmic quality of sleep, sleep patterns, behavioral
facilitators or inhibitors of sleep, and affective
responses to sleep disorders. These subscales were
utilized in making comparisons in data analysis.
RESULTS AND DISCUSSION
A pilot study was conducted at the Dunedin Clinic
from September 8, 1987 to September 11, 1987. All of
the 24 subjects who reported to the clinic were
considered for inclusion in the pilot study. Of the 24
subjects who visited the clinic during the time of the
pilot study, all but two were given the questionnaire to
take home and complete. One blind subject and one
subject with Alzheimer's disease were not asked to
participate. Eighteen (81.8%) of the 22 subjects who
were invited to participate returned the completed
questionnaire. The researcher was able to contact 14 of
these subjects by telephone.
Subject's self-rating of sleep at time of interview
was consistent for 13 of the 14 subjects. One subject
rated her sleep as satisfactory on the questionnaire and
good during the interview. When asked what they do when
they have trouble sleeping, 13 of the 14 subjects
interviewed answered the same as they had answered the
questionnaire. One subject indicated on the
questionnaire that she never had trouble sleeping but in
the interview, she answered that when she has trouble
sleeping she stays in bed until she falls asleep.
Subjects were less consistent in their rating of
others' sleep. Seven subjects rated the sleep of others
as compared to their own sleep consistently and five
were not consistent in their replies.
Based on the pilot, it was decided to include a
question to determine if subjects had a friend who had
sleep problems. This was done to provide a focus for
subjects and provide a basis for comparison of subjects
in the main study. Items about the sleep of others were
changed to be consistent with items which immediately
preceded them regarding subject's current and past sleep
A convenience sample of the subjects of the Dunedin
Clinic was obtained. A total of 285 subjects returned
the questionnaire. Of these, three subjects had not
answered any questions and returned blank questionnaires
leaving 282 usable questionnaires. Subjects who
reported to the Dunedin Clinic during the time of the
study ranged in age from 65 to 94 with a mean of 78.1
(SD=5.1). The mean age of subjects who returned the
completed questionnaire was 78.1 (SD=5.4) (see Table 1).
The mean age of subjects who did not return the
questionnaire was 78.1 (SD=5.4) There was a higher
percentage of females in both the subjects who returned
Gender of Respondents and Non Respondents
Respondents Non respondents Total
Age n Mean(SD) n Mean(SD) n Mean(SD)
Female 165 77.4(5.7) 70 78.1(5.9) 235 77.6(5.7)
Male 87 79.4(4.9) 36 78.2(5.6) 123 79.1(5.0)
Total 252 78.1(5.4) 106 78.1(5.4) 358 78.1(5.1)
Gender n % n % n %
Female 165 46.1 70 19.6 235 65.7
Male 87 24.3 36 10.1 123 34.4
Total 252 60.4 106 29.7 358 100.1a
aTotal is greater than 100% due to rounding.
the questionnaire and the subjects who did not return
the questionnaire. Of the subjects reporting to the
Dunedin Clinic at that time for whom demographic data is
known, 235 (65.7%) were female and 123 (34.4%) were
Fifty (14.0%) of the subjects reporting to the
Dunedin Clinic during the time of the study had been
hospitalized at least once during the previous year (see
Table 2). Three subjects had been hospitalized twice,
and one subject had been hospitalized three times within
the last year.
The total number of subjects who indicated that
they had experienced a heart attack within the last year
was 20 (5.6 %). Fifteen (4.2%) of the subjects
indicated they had experienced a cerebrovascular
accident within the past year. No statistically
significant differences were found between those who
responded to the Sleep Questionnaire and those who did
not on any of the demographic variables evaluated.
Subjects who reported to the Dunedin Clinic during
the time of the study included 41 who were entering the
study for the first time. Subjects complete a different
questionnaire each year they are a part of the Dunedin
Program. On questionnsires eight through twelve an
item is included in which subjects are asked to rate
Illness Related Variables of Respondents and Non
Respondents by Gender
Respondents Non respondents Total
% n %
3.6 7 10.0
6.9 1 2.8
4.8 8 7.5
their health. In this subsample of 160 subjects, 23
(6.4%) rated their health excellent; 131 (36.6%) rated
their health good; 66 (18.4%) rated their health fair;
and 6 (1.7%) rated their health poor (see Table 3).
Statistically significant differences existed between
respondents and nonrespondents to the questionnaire (chi
square=14.3 df=2, p=.007). Nonrespondents were more
likely to rate their health as fair or poor than
respondents. Significant differences existed between
male respondents and nonrespondents (chi square=15.7
df=3, p=.004) with a higher percentage of male
respondents rating their health as good, while a greater
proportion of male nonrespondents rated their health as
fair or poor. No differences were found between female
respondents and nonrespondents or between male and
female respondents or nonrespondents.
Average daily coffee consumption was reported by
subjects. The amount of coffee consumption ranged from
none to eight cups per day. The modal response was 1 to
2 cups of coffee per day. Subjects were also asked how
much alcohol they ingest. Responses available were
none; occasionally; 2-3 cocktails weekly; 1-2 cocktails
nightly; three or more cocktails nightly; wine
occasionally; and wine daily. One hundred thirteen
subjects denied alcohol consumption; 114 indicated they
Self Rating of Health by Gender for Respondents and
Excellent Good Fair Poor
Gender n % n % n % n %
Female 12 7.3 58 35.2 31 18.8 1 .6
Male 6 6.9 44 50.6 12 13.8 -
Total 18 7.1 102 40.5 43 17.1 1 .4
Female 1 1.4 23 32.9 16 22.9 3 4.3
Male 4 11.1 6 16.6 7 19.4 2 5.6
Total 5 4.7 29 27.4 23 21.7 5 4.7
Female 13 5.5 81 34.5 47 20.0 4 1.7
Male 10 8.1 50 40.1 19 15.5 2 1.6
Total 23 6.4 131 36.6 66 18.4 6 1.7
Chi square for respondents vs non respondents =14.3
df=3 p=.007. Chi Square for male respondents vs male
nonrespondents =15.7 df=3 p=.004.
use alcohol occasionally; 27 indicated they drink 2-3
cocktails weekly; 44 indicated they consume 1-2
cocktails nightly; 4 selected three or more cocktails
nightly; 39 drink wine occasionally; and 16 drink wine
Chi square statistics were computed for all
variables obtained from the Dunedin Program
Questionnaire and self-rating of sleep. The only
statistically significant relationship was between use
of medications that effect sleep and self-rating of
sleep. Twenty-two (8.9%) of the subjects for whom
background data were available reported using
medications that would affect sleep, including sedatives
and pain medications used at bedtime (see Table 4). A
higher percentage of subjects who reported taking
medications which affect sleep also rated their sleep as
poor (chi square=10.71, df=3, p=0.013).
Findings are reported within the framework of the
objective behavioral theory of sleep. Sleep behavior
consists of sleep patterns (sleep latency and length of
sleep), sleep structure limited to continuity of sleep,
and subjective response to sleep confined to sleep
evaluation. Continuity of sleep includes awakenings
within sleep and naps.
Quality of Sleep and Use of Medications That Affect
Use of Medications
Yes No Total
Quality of n % n % n %
Very Well 3 13.6 65 28.9 68 27.5
Fairly Well 5 22.7 84 37.3 89 36.0
Poorly 14 63.6 67 29.8 81 32.8
Some Problems -- 9 4.0 9 3.6
Total 22 8.9 225 91.1 247 99.9
Note. Does not equal 100% due to rounding.
Sleep latency. In response to an item regarding
sleep latency, subjects reported an average length of
time to get to sleep as 26.2 (SD=25.7) minutes. Great
variability is seen in this response with a range from
one to 150 minutes to get to sleep (see Table 5). The
mean sleep latency reported by Webb and Schneider-
Helmert (1984) for subjects 60 to 69 years of age was 15
The mean sleep latency subjects reported when they
were younger was much shorter with a mean of 10.6
minutes (SD=8.5). Sleep latency subjects reported for
their younger years was also less variable. The mean
sleep latency of peers was very similar to the subjects'
current sleep latency with a mean of 28.7 minutes
(SD=24.1). It must be noted however that far fewer
subjects estimated sleep latency of peers than reported
their own present or past sleep latency. The finding of
differences between now and younger supports earlier
findings (Agnew & Webb, 1971; Hayter, 1983; and Webb &
Schneider-Helmert, 1984) that sleep latency is longer in
Sleep length. The average sleep length reported by
subjects was 7.1 hours (see Table 5). This varied
little from the report of sleep length when younger
(7.5) and the sleep length of peers (7.0).
Sleep continuity. In addressing the concept of
continuity of sleep, subjects were asked the frequency
and duration of awakenings within sleep and the
frequency and duration of naps. Subjects reported
awakening 0 to 10 times per night with an average of 2.1
(SD=1.3) (see Table 6). Subjects reported fewer
awakenings when they were younger (Mean=.4, SD= 1.0).
Previous studies (Hayter, 1983; Tune, 1969; and Webb
1982a) have reported an increase in the number of
wakenings during the night in elderly subjects. In an
observational study of elderly subjects, Webb &
Swineburne (1971) found that elderly subjects awoke 2 to
5 times per night. Those subjects who reported the
number of times awake for peers estimated a mean of 2.1
awakenings (SD =1.9), which is very similar to the
mean number of times awake per night as subjects
reported for themselves.
Subjects report a mean time awake after sleep onset
of 30.4 minutes (SD=39.4). Again a large variability is
seen with a range from no time awake after sleep onset
to a total of 3 hours spent awake after sleep onset.
Subjects recalled an average of 4.2 minutes awake after
sleep onset when they were younger (SD=19.1). Thirty-
one subjects estimated the amount of time peers spend
awake after sleep onset as 50.2 minutes (SD=40.1).
Sleep Patterns Now, When Younger, and of Peers
Age Group n Mean(SD) Range
Sleep latency (minutes)
Now 269 26.2(25.7) 1-150
Younger 187 10.6( 8.5) .5-60
Peers 64 28.7(24.1) 1-120
Sleep length (hours)
Now 279 7.1(1.2) 4-10
Younger 235 7.5(0.9) 4-12
Peers 71 7.0(2.1) 4-10.5
Great variability is observed in estimates of the length
of time awake after sleep onset when subjects were
younger as well as estimates of wake time for peers.
Time awake after sleep onset was reported as shorter
when younger and longer for peers. Bixler et al.,
(1984) report an increase in both the frequency of night
awakenings and the total time awake after sleep onset.
The frequency and length of naps was explored.
Subjects were also asked to indicate the number of hours
per week they spent taking naps. Two hundred sixty-
seven subjects answered the item related to length of
current naps (see Table 6). The mean number of hours
spent taking naps was 3.4 hours per week (SD=3.8). The
mean number of hours reported by subjects when they were
younger was .7 hours per week (SD=2.9). The finding of
an increased number of naps is consistent with findings
of a study conducted by Hayter (1983). Tune (1969)
reported an increased number of naps in older subjects.
In an observational study of elderly subjects Webb &
Swineburne (1971) found that all subjects took naps.
More subjects responded to items regarding current
naps than responded to items dealing with past nap
patterns or nap patterns of others their age. One
hundred ninety-four subjects (71.6%) reported that
currently take naps (see Table 7). Forty-two (17.1%
Continuity of Sleep Now, When Younger, and of Peers
Age Group n Mean(SD) Range
Times per night awake
Now 277 2.1(1.3) 0-10
Younger 178 .4(1.0) 0-10
Peers 69 2.1(1.9) 0-15
Minutes awake after sleep onset
Now 228 30.4(39.4) 0-180
Younger 140 4.2(19.1) 0-210
Peers 31 50.2(40.1) 0-120
Hours of naps per week
Now 267 3.4(3.8) 0-21
Younger 185 .7(2.9) 0-21
Others 81 6.2(4.9) 0-21
reported having taken naps when they were younger, and
98 (86.0%) indicated that others their age take naps.
Sleep evaluation. To explore the concept of
subjective response to sleep behavior subjects were
asked to evaluate how well they sleep now, the amount of
sleep they get now, how rested they feel when they
awaken, how much they enjoy sleep, and how light or deep
they consider their sleep. Overall, 76 (27.4%) subjects
indicated that they sleep very well, 99 (35.7%)
satisfactorily, 91 (32.9%) some problems, and 11 (4.0%)
indicated that they sleep poorly (see Table 8). When
asked how well they slept when they were younger, 190
asked to indicate how well they felt others their age
sleep, and 141 (50.2%) indicated that others their age
have some problems sleeping. Statistically significant
differences were found between self-ratings of current
and former sleep (chi square=90.7, df=3, p=0.0002) and
between ratings of current sleep and sleep of peers (chi
square=44.6, df=3, p=0.000). The tendency of subjects
to report better sleep when they were younger is
consistent with findings reported by Karakan, et al.,
(1979) from a study which involved subjects ranging in
age from 18 to over 70. In that study, subjects in the
Subjects Taking Naps Now, When younger, and Peers
no naps naps
Age Group n % n %
Now 77 28.4 194 71.6
Younger 204 82.9 42 17.1
Peers 6 5.2 98 86.0
Chi Square now vs younger=209.5 df=l p=0.0000.
Chi Square now vs others=21.8 df=l p=0.0000.
Evaluations of Sleep Now, When Younger, and
now younger others
Response n % n % n %
Quality of Sleep
Note. Does not equal 100% because don't know and don't
remember responses not shown, aChi square now vs
younger=90.7 df=3 p=0.0000. bchi square now vs others=
44.6 df=3 p=0.0000. Cchi square=20.2 df=2 p=0.0000.
dchi square=38.3 df=2 p=0.0000.
now younger others
Response n % n % n %
Feel rested upon awakening
89.0e 115 40.9f
Enjoyment of sleep
Not At All
Note. Does not equal 100% because don't know and don't
remember responses not shown, eChi square now vs
younger=19.5 df=3 p=0.0002. fChi square now vs
others=46.5 df=3 p=0.0000. gChi square=163.9 df=3
p=0.0000. hchi square=46.0 df=3 p=0.0000.
now younger others
Response n % n % n %
Depth of Sleep
Very Deep 12 4.3 46 16.6 1 .4
Deep 145 52.2 152 54.9 9 3.3
Light 113 40.6 45 16.2 55 20.0
Very Light 8 2.9 4 1.4 3 1.1
Total 278 100.0 247 89.1i 68 24.8J
Note. Does not equal 100% because don't know and don't
remember responses not shown.. -Chi square now vs
younger=46.0 df=2 p=0.0000. JChi square now vs
others=59.8 df=2 p=0.0000.
older age groups tended to report more frequent problems
An original question in this research, addressing
the issue of whether subjects who experienced sleep
disturbances when younger continue to exhibit sleep
disturbances when they are older was not answered. Only
12 subjects (4.2%) were found to have experienced sleep
problems when they were younger. The more significant
finding was the increased number of subjects who
reported sleep problems now (n=102, 36.9%).
In response to questions about the amount of sleep
they get, 245 (87.9%) reported getting enough sleep, 25
(8.9%) indicated they do not get enough sleep, and 9
(3.2%) reported getting too much sleep. Subjects were
asked about the amount of sleep they got when they were
younger. Two hundred twenty-six (81.6%) indicated they
got about enough and 27 (9.7%) indicated they had not
gotten enough sleep when they were younger. Fewer
subjects answered the item regarding sleep of others.
Of those who answered, 88 (31.9%) selected about enough.
Statistically significant differences were found between
evaluation of the amount of current sleep and the amount
of sleep when younger (chi square=20.2, df=2, p=0.0000).
Significant differences were found between ratings of
amount of sleep subjects got and the amount of sleep
others obtained (chi square=38.3, df=2, p=0.0000).
One hundred sixty-four subjects (58.8%) indicated
they almost always feel rested when they awaken, 71
(25.5%) indicated they often feel rested when they
awaken, 35 (12.5%) chose occasionally, and 9 (3.2%)
indicated that they almost never feel rested when they
awaken. Two hundred five subjects (73.9%) indicated
they almost always felt well rested upon awakening when
they were younger. Analysis of data revealed
statistically significant differences between self-
report of feeling rested upon awakening now and when
younger (chi square =19.5, df=3, p=0.0002). Twenty-five
subjects (8.9%) indicated that others their age feel
rested upon awakening. Significant differences were
found between self-rating of feeling rested upon
awakening and rating of others on this item (chi square=
46.5, df=3, p=0.0000).
Although 36.9% of respondents reported having some
problems with sleep or sleeping poorly, only 15.7% of
respondents indicated they occasionally or almost never
felt rested upon awakening. This supports findings
reported by McGhie and Russell (1962) that the elderly
did not report morning tiredness despite sleep changes.
When asked how much they enjoy sleep, 166 (59.3%)
of subjects selected much, 104 (37.1%) chose moderately
9 chose a little, and only one subject replied not at
all. When asked how much they enjoyed sleep when they
were younger, 190 (67.9%) chose much. When asked how
much others their age enjoy sleep, 41 (14.7%) indicated
much, and 47 (16.8%) selected moderately. Statistically
significant differences were found between reports of
enjoyment of sleep now as compared to younger (chi
square=163.9, df=3, p=0.0000) and between subjects' own
enjoyment sleep and others enjoyment of sleep (chi
square=46.5, df=3, p=0.0000)
The concept of depth of sleep was explored by
asking subjects to rate the depth of their sleep now,
when they were younger, and of others their age. Twelve
subjects (4.3%) rated their sleep as very deep, 145
(52.2%) rated their sleep deep, 113 (40.6%) indicated
that they consider their sleep light, and 8 (2.9%) rated
their sleep as very light. When asked how deep their
sleep was when they were younger, 152 (54.9%) selected
deep. When asked to indicate how deep the sleep of
others their age is, 55 (20.0%) selected light.
Data analysis revealed statistically significant
differences between self-report of depth of sleep now
and depth of sleep when younger (chi square=46.0, df=2,
p=0.0000) and depth of sleep now and depth of sleep of
peers (chi square=59.8, df=2, p=0.0000). Statistically
significant relationships were found between all
variables related to subjective response to sleep using
Spearman's rho (see Table 9).
Subjects consistently rated their sleep as having
been better when they were younger. Subjects reported
having slept better, gotten enough sleep, felt rested
upon awakening, enjoyed sleep more, and experienced
deeper sleep. These findings lend support to the
developmental aspect of the Objective Behavioral Theory
of Sleep (Webb, in press) just as the findings of sleep
pattern changes lend support to this aspect of the
The pattern of self-evaluation of current sleep and
the sleep of peers was less clear when considering
subjective evaluation of sleep. Many subjects chose the
don't know response in relation to the sleep of others
their age. Statistically significant differences were
found, but they must be interpreted with caution. Fifty
percent of subjects who responded to the item regarding
how well others their age sleep at night responded that
their peers have some problems. This is a higher rate
than subjects indicated that they themselves have
Correlations Matrix for Subjective Responses to Sleep
and Circadian Subscale
Sleep Feeling Enjoy Sleep Circadian
Subscale Rating Rested Sleep Depth Tendency
Feeling .429 1.000
Rested .000 .000
Enjoy .147 .156 1.000
Sleep .014 .009 .000
Sleep .433 .344 .161 1.000
Depth .000 .000 .007 .000
Circadian .218 .226 .112 .150 1.000
Tendency .008 .000 .180 .071 .000
problems sleeping. This adds more evidence that the
elderly expect sleep problems as they age.
Circadian tendencies. In order to evaluate
circadian tendencies, the placement of sleep within the
twenty four hour cycle, subjects were asked to respond
to items regarding their usual bedtime and wake up time.
Subjects reported going to bed from 1800 (6:00 P.M.) to
0100 (1:00 A.M.) with an average bedtime of 2246 (10:46
P.M.) (see Table 10). This also varied little from the
estimated bedtime of peers and bedtime when they were
younger. Subjects report an average weekday wake up
time of 0651 (6:51 A.M.) an average weekend wake up time
of 0709 (7:09 A.M.). There is little difference
reported between average weekday and weekend wake up
time now, however there is a difference between weekend
wake up time now and weekend wake up time when younger.
To further assess circadian tendencies, questions
were asked to determine if subjects go to sleep at the
same time, wake up at the same time, and use an alarm
clock. In response to the item dealing with going to
sleep at the same time, 44 (15.7%) indicated they always
go to bed at the same time and 195 (11.8%) stated they
usually go to sleep at the same time (see Table 11).
Twenty-six subjects (9.3%) indicated that they always
went to sleep at the same time when they were younger,
Circadian Tendencies Now, When Younger, and of Peers
Age Group n Mean(SD) Range
Weekday wake up time
Weekend wake up time
Circadian Tendency of Sleep of Elderly Subjects Now,
When Younger, and Others
now younger others
Frequency n % n % n %
Going to sleep at the same time
Total 280 100.0
Awakening at the same time
Total 277 100.0
Note. Columns do not equal 100% because don't know and
don't remember responses have not been included.
aChi Square now vs younger=48.8 df=3 p=0.000. bChi
Square now vs others=28.5 df=3 p=0.000. cChi Square
now vs younger=9.7 df=3 p=0.02. dChi Square now vs
others=37.3 df=3 p=0.000.
now younger others
Frequency n % n % n %
Using an alarm clock or radio to help wake up
Almost Always 24 8.7 139 50.1 8 2.9
Often 15 5.4 42 15.2 21 7.7
Occasionally 84 30.4 26 9.4 47 17.2
Almost Never 153 55.4 59 21.3 23 8.4
Total 276 99.9 266 96.0a 99 36.2b
Note. Columns do not equal 100% due to rounding and
omitting don't know and don't remember responses.
aChi Square now vs younger=142.5 df=3 p=0.000. bChi
Square now vs others=285.5 df=3 p=0.000.
while 108 (38.6%) indicated they usually went to sleep
at the same time when they were younger. Chi square
analysis of data revealed statistically significant
differences between present and previous sleep patterns
with subjects reporting more consistency in going to bed
at the same time now than when they were younger.
Subjects were also asked if they always awaken at
the same time. Sixty-one subjects (22%) indicated they
always awaken at the same time; 169 (61.0%) indicated
they usually awaken at the same time; 43 (15.5%) stated
they sometimes awaken at the same time, and 4 (1.5%)
indicated they never awaken at the same time (see Table
Sixty-six subjects (23.9%) indicated they always
got up at the same time when they were younger, and 122
(44.2%) indicated they usually got up at the same time
when they were younger. Chi Square analysis reveals
statistically significant differences for this item
between now and younger. The greatest difference is
evident between the percentage of subjects who reported
that they usually awaken at the same time now than when
they were younger, with a higher percentage of subjects
reporting they usually awaken at the same time now.
Large differences are seen in the rate of use of an
alarm clock or radio to awaken between what subjects
reported they were doing at the time of the study and
what they reported doing when they were younger. One
hundred fifty-three subjects (55.4%) reported they
almost never used an alarm clock or radio to help them
awaken now, while 139 (50.1%) indicated that they had
almost always used an alarm clock when they were younger
(chi square=142.5, df=3, p=.000). Differences between
what subjects report they are doing versus what they
think others their age are doing are not so dramatic,
however, differences between both of these categories
are statistically significant (chi square=285.5, df=3,
The concept of circadian tendencies was further
explored by utilizing the circadian subscale imbedded
within the Sleep Questionnaire to make comparisons with
other items. Data from the items regarding bedtime,
wake up time, and use of an alarm clock comprised the
circadian subscale. This subscale included items
related to bedtime (Do you go to bed at the same time
now? and How many days per week do you go to bed more
than one hour earlier or later then your usual
bedtime?); wake up time (Do you always awaken at the
same time now? How many days per week do you awaken
more than one hour earlier or later then usual?); and
the regularity of the use of an alarm. All items
comprising the circadian subscale were found to
significantly correlated with the circadian score on the
circadian subscale (see Table 12). The score obtained
on the circadian subscale was also found to be
significantly related to several other items on the
Sleep Questionnaire. It is of interest to note that
this score is related to current and past sleep patterns
as well as to ratings of sleep of others.
The score on the circadian subscale was used to
investigate the relationship between circadian
tendencies and quality of sleep variables. The
circadian quality of sleep was found to be significantly
related to rating of sleep and the reported frequency of
feeling rested upon awakening (see Table 9).
Sleep disturbances. Several items addressed the
issue of sleep disturbances directly. Subjects were
asked how often they have trouble getting to sleep as
quickly as they would like, how many days per week the
usually wake up earlier than they expect and are unable
to return to sleep, and if they have sought help from
either a health care professional or a friend or
relative for sleep problems. In response to the item
which addressed trouble getting to sleep as quickly as
they would like, 71 subjects (25.6%) indicated they
almost never have difficulty getting to sleep, 137
Relationship Between Circadian Tendencies of Sleep
and Quality of Sleep and Other Variables
Going to bed at same time
Waken at the same time
Frequency of use of alarm clock
Going to bed earlier or later
Waking up earlier then expect
Going to bed earlier or later
Others going to bed earlier
Waking up earlier then expect
Others waking up earlier than
Frequency of trouble getting to
Times awaken per night
Minutes awake per night
(49.5%) stated they occasionally have trouble getting to
getting to sleep, and 19 (6.9%) indicated they always
have trouble getting to sleep (see Table 13).
Subjects were also asked how often they had trouble
getting to sleep when they were younger and how often
they think others their age have trouble getting to
sleep. One hundred sixty-six (60.8%) replied they
almost never had trouble getting to sleep when they were
younger. The most frequent responses given in relation
to others their age were occasionally (n=38, 14.1%) and
often (n=43, 16.0%). Statistically significant
differences were found between responses to items about
trouble sleeping now and trouble sleeping when younger
(chi square=203.7, df=3, p=0.0000) as well as trouble
sleeping now and others their age experiencing trouble
(chi square=51.1, df=3, p=0.0000).
Spearman's rho correlations were performed to
determine the relationship between sleep latency and
other qualitative sleep variables. A positive
correlation was found between sleep latency and overall
evaluation of sleep, feeling rested upon awakening and
circadian tendencies (see Table 14). When sleep latency
was reported as short, the evaluation of sleep was more
likely to be positive, the score on the circadian
subscale was smaller, and the subject was more likely to
Frequency of Trouble Getting to Sleep Now, When
Younger, and Others
n % n % n %
now younger others
Almost Never 71 25.6 166 60.8 5 1.9
Occasionally 137 49.5 59 21.6 38 14.1
Often 50 18.1 7 2.6 43 16.0
Almost Always 19 6.9 2 .7 5 1.9
Total 277 100.0 234 85.7a 91 33.91b
Note. Does not equal 100% because don't know and don't
remember responses not shown.
aChi Square now vs younger =203.7 df=3 p=0.000. bchi
Square now vs others = 51.1 df=3 p=0.000.
Relationship Between Sleep Problems and Subjective
Evaluation of Sleep Using Spearman's rho
Subjective evaluation of sleep r p n
Overall evaluation of sleep .58 .0001 151
Feeling rested upon awakening .43 .0001 153
Depth of sleep -.35 .0001 153
Number of times awake
Overall evaluation of sleep .37 .0001 161
Feeling rested upon awakening -.32 .0001 163
Depth of sleep -.24 .0025 162
Length of time awake
Overall evaluation of sleep .58 .0001 123
Feeling rested upon awakening -.38 .0001 124
Depth of sleep -.23 .0025 162
report feeling rested upon awakening. A negative
correlation was found between sleep latency and depth of
sleep. The longer the sleep latency, the more likely
the subject was to report light sleep.
Subjects were also asked how many times they awaken
during the night and how long it takes them to get back
to sleep. Responses to these items have been addressed
earlier. Positive correlations were found between
both the frequency and duration of night awakening and
overall evaluation of sleep, rhythmic quality of sleep,
and feeling rested upon awakening. Negative
correlations were found between depth of sleep and both
frequency and duration of wakefulness after sleep onset.
Early morning awakening was addressed by asking
subjects how often they wake up earlier than they would
like and are unable to return to sleep. One hundred
thirteen (40.0%) subjects indicated they never have this
problem. Most of those who reported they do wake up
earlier than they would like indicated that this occurs
once or twice a week (n=116, 42.7%). Thirty-five
subjects (12.9%) indicated they have this problem 3 to 4
times per week and 12 subjects (4.4%) experience this
problem 5 to 7 times per week.
Help-seeking. In order to assess help-seeking for
sleep disturbances, subjects were asked if they had
sought help from health care professionals or friends
and relatives for trouble sleeping. Thirteen (4.9%)
indicated they had sought help from a health care
professional and 3 (1.1%) had sought help from family or
friends. One hundred seventeen subjects indicated they
do not have trouble sleeping in response to these items
(see Table 15).
Behavioral inhibitors of sleep. To explore the
concept of behavioral inhibitors of sleep, subjects were
asked to select items they felt might be affecting their
overall sleep. They were then asked to rate the
effectiveness of various items dealing with overall
sleep that they might have tried. Emotional stress,
caffeine intake, pain, going to bed too early, lack of
exercise, and naps were cited as problem areas by
greater than 25% of subjects who responded to that item
(see Table 16). While 67 subjects indicated that
emotional stress may have been affecting their overall
sleep behavior, only 24 indicated that reducing
emotional stress worked very well or fairly well in
improving the quality of their sleep (see Table 17).
Three subjects indicated that reducing emotional stress
worked poorly or was no help. Perhaps subjects who
indicated emotional stress was affecting their sleep,
but did not respond to effectiveness of reducing
Frequency of Help Seeking for Sleep Problems From Health
Professionals and Peers
Response n % n %
Yes 13 4.9 3 1.1
No (but trouble sleeping) 136 50.7 148 55.4
No (no trouble sleeping) 119 44.4 116 43.4
Total 268 100.0 267 99.9
Note. Totals do not equal 100% due to rounding.
Behavioral Inhibitors of Overall Sleep Patterns
Inhibitor n % n %
Emotional stress 67 57.3 50 42.7
Caffiene 54 46.2 63 53.8
Pain 52 34.8 58 38.9
Bed too early 44 40.7 64 59.3
Lack of exercise 42 40.4 62 59.6
Naps 38 36.9 65 59.6
Wake up too late 14 11.2 64 51.2
Bed too late 10 13.2 66 86.8
Subjects' Evaluations of Effectiveness of Facilitators
for Overall Sleep Patterns
Works well Works poorly
Facilitator n % n %
Decrease caffiene 38 97.4 1 2.5
Relieve pain 32 86.5 3 8.1
Decrease stress 24 88.9 3 11.1
Decrease naps 19 82.6 4 17.4
Increase exercise 19 90.5 2 10.0
Awaken earlier 4 100.0 -
Go to bed earlier 3 75.0 1 25.0
Go to bed later 1 33.3 2 66.6
emotional stress were unable to reduce their levels of
emotional stress, or have not attempted to do so.
Fifty-four subjects indicated that they felt caffeine
intake might be affecting their overall sleep, and 38
subjects indicated that reducing caffeine intake worked
very well or fairly well. One subject indicated that
reducing caffeine intake worked poorly or was no help.
Fifty-two subjects responded they felt pain was
affecting their overall sleep behavior. Thirty-two
subjects indicated that relieving pain worked well
in improving their sleep behavior. Three subjects
indicated that relieving pain worked poorly.
Lack of exercise was indicated as affecting overall
sleep behavior by 42 subjects. Nineteen subjects
indicated that increasing exercise worked well in
helping overall sleep behavior. Two subjects indicated
increasing exercise worked poorly.
Thirty-eight subjects indicated that naps might be
a factor in their overall sleep behavior. Nineteen
subjects stated that decreasing naps worked well in
improving their overall sleep behavior. Four subjects
indicated that decreasing naps worked poorly.
Behavioral facilitators and inhibitors of sleep.
In addition to overall sleep behavior three areas of
difficulty sleeping were assessed separately. These
three areas were (1) difficulty falling asleep; (2)
difficulty staying asleep; and (3) waking up in the
morning earlier than desired and being unable to return
to sleep. Subjects were first asked to respond to an
item that addressed the frequency with which they
employed various facilitators or inhibitors with each
category of disturbance. A second item addressed how
well those facilitators or inhibitors worked.
Behavioral facililatators included in the items were
relax, read, listen to the radio, take a pain pill,
pray, count sheep, and think of something else.
Behavioral inhibitors included get out ot bed and do
something, watch TV, take a sleeping pill, drink
alcohol, and talk to someone.
The faciliitators subjects always use when they
experience difficulty getting to sleep were: try to
relax, read, pray, think of something else, and listen
to the radio (see Appendix B). Seventy subjects
indicated they never have trouble getting to sleep at
night. The facilitatorss most frequently identified as
working well or very well were read, try to relax,
listen to the radio, pray, and think of something else.
When subjects were asked to indicate what they do
when they wake up at night and can't get back to sleep,
ninety-seven responded that they never wake at night
without being able to get back to sleep. The
facilitators most frequently identified as always used
included read, try to relax, listen to the radio, pray,
and think of something else. Inhibitors frequent used
included getting up and do something, and watching
television. The facilitators most often identified as
working well or very well were read, try to relax,
listen to the radio, and pray. The inhibitor of sleep,
getting up and doing something, was also found to be
effective in dealing with waking up at night and being
unable to get back to sleep. Subjects were also asked
to indicate what they do if they wake up in the morning
earlier than they like. Sixty-seven subjects indicated
that they never wake up in the morning earlier than they
like without being able to get back to sleep. The
facilitators used most frequently were try to relax,
read, listen to the radio, and count sheep. The
inhibitor of getting up and doing something was also
frequently used. The facilitators most frequently
indicated as working well or very well included read,
and try to relax. Getting up and doing something was
also found to be an effective strategy.
Affective response to sleep disturbances. To
explore the concept of affective responses to sleep
disturbances, subjects were asked how they feel when
they don't fall asleep as quickly as they would like,
wake up at night and can't get back to sleep, or wake up
in the morning earlier than they would like. Sixty-five
subjects indicated they always fall asleep as quickly as
they like, 50 indicated that they never wake up at night
and are unable to return to sleep, and 52 subjects
indicated that they never wake up in the morning earlier
than they like. As can be seen in Table 18 the most
frequent response in all three instances indicated that
subjects accepted the sleep disturbance. Of those
subjects who reported having difficulty getting to sleep
at night. 114 (42.7%) indicated they accepted the
disturbance. One hundred twenty-one subjects (46.5%) of
those who reported waking up at night and having
difficulty returning to sleep reported they accepted the
disturbance, and 153 (56%) of subjects who reported
awakening earlier in the morning than they would like
reported they accept the disturbance. The next most
frequent response selected by subjects who did not fall
asleep as quickly as they liked at night (n=47, 17.6%)
and those who woke up at night without being able to get
back to sleep (n=50, 19.2%) was that they feel
frustrated. Thirty-two subjects who reported having
difficulty getting to sleep, 32 subjects who wake up at
night and have difficulty returning to sleep, and 20
Affective Response to Difficulty Falling Asleep,
Awakening During the Night, and Early Morning Awakening
Falling Asleep Wake during Wake too
the night early
Reactions n % n % n %
Not a 65 50 52
Accept 114 42.7 121 46.5 153 56.0
Frustrated 47 17.6 50 19.2 9 3.3
Tired 32 11.9 32 12.3 20 7.3
Don't think 31 11.6 27 10.4 45 16.5
Depressed 19 7.1 10 3.8 5 1.8
Worried 10 3.7 7 2.7 1 .3
Angry 8 3.0 5 1.9 7 2.5
Happy to have 6 2.2 8 3.1 33 12.1
Total 267 99.8 260 99.9 273 99.3
Note. Columns do not equal
100% due to rounding.
subjects who woke up earlier in the morning than they
liked, indicated that they felt physically tired when
this occurred. Some subjects do not think about their
sleep disturbance. Forty-five subjects reported waking
up in the morning earlier than they would like, 31
subjects who reported having difficulty getting to
sleep, and 27 subjects who woke up at night indicated
they don't think about it. Thirty-three subjects who
reported waking up earlier than they like in the morning
reported they were happy to have more time.
Data obtained in this research were analyzed to
determine the relationship between current sleep
behavior, recalled sleep behavior when younger, and
perceived sleep behavior of others. Analysis of
background data as related to sleep revealed a
statistically significant relationship between overall
evaluation of sleep and use of medications that affect
sleep. Analysis of overall sleep behavior revealed that
subjects reported their sleep behavior when they were
younger in a more positive light than their current
sleep behavior. Subjective evaluations of sleep
revealed a similar pattern.
Sleep patterns appear to have great variability.
The mean length of time to get to sleep is longer now
(26.2 minutes) than when subjects were younger (10.6
minutes). The number of reported awakenings during the
night increased from an average of .4 times per night to
2.1 times per night. The amount of time spent awake
after sleep onset also increased from a mean of 4.2
minutes awake after sleep onset when younger, to a mean
of 30.4 minutes spent awake now. More subjects report
taking naps now than when they were younger. Naps are
also longer now than when subjects were younger.
In relation to the circadian quality of sleep,
findings indicated that subjects go to bed and get up at
the same time more frequently now than when they were
younger. Differences between alarm clock usage when
younger and now are dramatic with the majority of
subjects reporting frequent use of an alarm clock when
younger and the majority of subjects reporting not using
an alarm clock now. Statistically significant
relationships exist between circadian tendencies and
evaluations of sleep.
Sleep disturbances were approached from a global
perspective as well as from the perspective of specific
problem areas including difficulty getting to sleep,
difficulty maintaining sleep, and early morning