Вы находитесь на странице: 1из 14

Journal of Psychosomatic Research, Vol. 47, No. 6, pp.

583596, 1999
Copyright 1999 Elsevier Science Inc.
All rights reserved.
0022-3999/99 $see front matter

S0022-3999(99)00062-8

PSYCHOLOGICAL CORRELATES OF SLEEP APNEA


WAYNE A. BARDWELL,* CHARLES C. BERRY,
SONIA ANCOLI-ISRAEL* and JOEL E. DIMSDALE*

AbstractRelationships were examined between psychological and sleep variables in individuals with
sleep apnea (n572, 24 with hypertension; Respiratory Disturbance Index515) and without sleep apnea
(n540, 16 with hypertension; respiratory disturbance index,15). Subjects were 3264 years old, 100
150% of ideal weight, with no other major illness. For subjects with sleep apnea, depression, anger, and
total mood disturbance correlated positively with deep sleep, rapid eye movement sleep, and/or hypoxemia. For subjects without sleep apnea, vigor correlated positively with sleep quantity and negatively
with hypoxemia. When age, body mass, and hypertension were controlled, results changed little for subjects without sleep apnea; for subjects with sleep apnea, depression and total mood disturbance no longer
correlated with sleep measures. Although various psychological measures correlate with sleep variables
in sleep apnea subjects, many are explained by controlling age, body mass, and hypertension. Anger and
vigor, however, remain associated with sleep variables. 1999 Elsevier Science Inc.
Keywords:

Anger; Blood pressure; Mood; Obstructive sleep apnea; Psychological factors.

INTRODUCTION

Obstructive sleep apnea is a devastating illness that leaves patients exhausted from
sleep deprivation and, at times, cognitively impaired [1]. The literature is mixed on
the role psychological factors play in this illness. Some researchers have observed
that sleep apnea is associated with clinical depression [25] or increased levels of
depressive symptoms [612], either as a direct consequence of the sleep deprivation
or indirectly as a consequence of social effects of this illness. Others have reported
that patients with sleep apnea do not show clinically significant levels of depression,
or have levels of depressive symptoms no higher than control groups [1315] or patients with other chronic illnesses [16]. In addition, some investigators have asserted
that sleep apnea patients have certain other associated psychological characteristics,
namely irritability [6, 9, 12, 16], tendency toward conflict with others [9], anxiety [2,
3, 7, 9, 10, 17], and fatigue or diminished energy [1, 7, 11, 14, 18]. Others have reported that sleep apnea patients show a somaticneurotic personality pattern [12,
17]. However, one researcher reported finding no correlation between presence/absence of sleep apnea and any personality factors [19].
Departments of * Psychiatry and Family and Preventive Medicine, University of California, San
Diego, California, USA.
Veterans Affairs San Diego Healthcare System, San Diego, California, USA.
Address correspondence to: Dr. Joel Dimsdale, Department of Psychiatry, University of California,
San Diego, La Jolla, CA 92093-0804. Phone: (619)-543-5592; Fax: 619-543-5462; E-mail: jdimsdale@
ucsd.edu.

583

584

W. A. BARDWELL et al.

Problems with these studies include sample characteristics, small sample size, lack
of control groups, and inconsistent or undisclosed criteria for diagnosing sleep apnea. Subjects have also suffered from other comorbid disorders and the studies have
rarely controlled for hypertension (HTN)frequently found in sleep apnea patientswhich has certain psychological characteristics in its own right [20, 21]. Previous studies have also not examined the influence of weight and age, which are associated with both sleep apnea and psychological factors.
Table I summarizes findings from several studies of sleep apnea and psychological variables. Eleven studies found elevated levels of depression in patients with
sleep apnea [2, 3, 5, 6, 8, 10] and/or improvement in depressive symptoms after
treatment (continuous positive airway pressure [CPAP], uvulopalatopharyngoplasty [UPPP], or tracheostomy) [4, 7, 9, 11, 16], whereas three studies did not [13
15]. Six studies found elevated levels of anxiety or tension or else improvement in
these symptoms after treatment in patients with sleep apnea [2, 3, 7, 9, 10, 17]. Three
investigations indicated that sleep apnea patients reported or were observed to be
irritable, frustrated, or prone to conflict [6, 9, 16].
Various measures of sleep quality and quantity, such as sleep fragmentation [2,
5, 9, 11] or hypoxemia [2, 9, 11], have been found to be associated with psychological
and cognitive impairment in this population. One study reported that depression,
anxiety, and cognitive deficits were related to time spent in rapid eye movement
(REM) and deep sleep (stages 3 and 4), as well as to hypoxemia [9]. Another study
found that cognitive deficits were correlated with frequency of apneas/hypopneas,
arousals, and hypoxemia [2]. One group reported that depressive symptoms and fatigue decreased after CPAP. They found that the decrease in depressive symptoms
correlated with an increase in deep sleep, whereas the decrease in fatigue correlated
with improvement in oxyhemoglobin saturation [11]. Finally, another team found
depression correlated with REM time, REM latency, and the presence/absence of
HTN medications [5].
Although psychiatrists and behavioral medicine researchers are very interested
in links between various mood states and physiology, it has been striking that observations have almost entirely ignored sleepa behavioral and physiological state
comprising one third of the dayand sleep disorders. Certainly, the most common
of these sleep disorders is obstructive sleep apnea.
For these reasons, we thought it important to examine psychological and sleep
variables in a group of individuals with and without sleep apnea, some normotensive and some hypertensive. All participants fell into a common age range and were
free of other significant illnesses. Compared with most studies, we had a larger sample size and we collected data on hypertension so we could tease out the effect of
this illness from that associated with sleep apnea alone.
METHOD
Study methodology
Subjects (n5112) included 61 men and 11 women with sleep apnea (defined as a respiratory disturbance index [RDI]>15), 24 of whom had HTN; and 27 men and 13 women with no sleep apnea (RDI,
15) or other sleep disorder, 16 of whom had HTN. Subjects ranged from 32 to 54 years of age (see Table
II). Subjects with a history suggestive of sleep apnea were recruited by advertising and word of mouth.
To qualify, subjects had to be in the range of 100150% of ideal body weight as determined by Metropolitan Life Insurance tables [22]. Although sleep apnea is more common among the obese, subjects .150%

1993/ 20 males
1994

1993

Borak

Gall

20 males
w/ mild
apnea

38 males

1994

180 apneic

Edinger

1994

Flemons

100 apneic

32 w/
apnea

1997

Flemons

Subjects

Engelman 1994

Year

Authors
Interview/questionnaire

Instruments

Polysomnography:
AHI . 10/hour

SCL-90-R, POMS, SIP


Katz Adjustment
Scale (KAS), MOSSF-36, Psychosocial
Adjustment to Illness
Scale (PAIS-R)

BDI, Taylor Manifest


Anxiety Scale, Tylko
Assessment of
Psychological
Effectiveness of
Rehabilitation Scale

Polysomnography
GHQ-12
AHI/hour > 10
Polysomnography:
HADS, GHQ-28,
AHI/hour > 5,
Nottingham Health
AHI median 5 28,
Profile (QOL),
AHI range 5
UWIST Mood
7129
Adjective Checklist
Polysomnography
MMPI

Apnea verification

7 males
Polysomnography:
referred
AHI , 20
due to
snoring,
but
negative
polysomnography

Controls

CPAP

CPAP vs.
oral
placebo
(crossover
design)

Treatment

Table I.Summary of findings from the literature

(continued)

CPAP improved: depression and anxiety


on HADS, energetic arousal on
UWIST, social and sex life and ability
to do chores; CPAP did not improve
tension or hedonic tone on UWIST
Apneics have higher MMPI scores: mean
depression 5 75.1 mean
hypochondriasis 5 83.1
Apneics have higher depression and
anxiety
CPAP improved depression and anxiety
Depression and anxiety correlated with
duration of illness, deep sleep, REM,
minimum O2 saturations; apneics
observed to be irritable, prone to
conflicts, dejected apneics reported
poor self-esteem and social functioning
No difference in SCL-90 depression,
current mood state, and total mood
disturbance for apneics vs. controls;
apneics have lower SF-36 energy and
sense of well-being for mental health;
apneics have lower POMS vigor;
apneics have impaired work/social
functioning due to health, more limits
in functioning due to physical/
emotional probs.; apneics have
poorer PAIS-R adjustment due to
illness in domestic, work, and social
environments and worse adjustment in
family; spouses report apneics have
fewer social activities

Apneics reported depression,


irritability, frustration
Apnea not related to depression

Findings

Psychological correlates of sleep apnea

585

1992 25 males, 4
females

Cheshire

30 w/ AHI
,10

Controls

Polysomnography:
AHI > 15/hour

Apnea verification

HADS

Freiberger Personality
Inventory

Instruments

1992 23 males

1987 10 males
preop.
for UPP
1985 43 males, 7
female

Platon

Klonoff

Kales

10 males
preop.
for
CABG
Apnea warranting
tracheostomy

Polysomnography:
AHI > 15/hour

MMPI, SCL-90-R

MMPI, BDI

16 males, 1 AHI > 20/hour


MMPI (Spanish),
female
O2 desat. > 10%,
Adjustment
incapacitating day
Inventory (Spanish)
somnolence, and/or
associated CVD
Lee
1990 60 Chinese
Polysomnography
DSM-III-R clinical
Millman
1989 55 apneic
Polysomnography:
Zung Self-rating
apnea
AHI > 5/hour
Depression Scale
(SDS)
Derderian 1988 7 males
7 males
Polysomnography:
POMS
AHI mean 5 40.7,
AHI range 5 3461

1993 76 w/ AHI
>10

Cassel

Subjects

Year

Authors

Table I.(Continued)

UPP

CPAP

CPAP

CPAP

Treatment

(continued)

CPAP improved POMS depression,


total mood disturbance, and fatigue
compared with controls; change in
depression correlated with change in
deep sleep; change in fatigue
correlated with change in O2
saturation
UPP improved MMPI and BDI
depression but was not different from
CABG controls; apneics reported
reduced irritability after UPP
MMPI showed somaticneurotic
personality (elevated hypochondriasis,
depression, hysteria); 25 of 50 were
depressed (MMPI); cognitive
impairment increased potential for
irritability

No correlation between apnea &


personality for all apneics or severe
apneics
On HADS, 7 of 29 apneics showed
depression, 10 of 29 apneics showed
anxiety; cognitive deficits correlated
with frequency of apneas/hypopneas,
arousals, hypoxemia
Apneics show general neurotic profile
on MMPI, including higher
depression, anxiety, hypochondriasis;
apneics show poorer emotional,
social, professional adjustment
No clinical depression observed
25 of 55 had clinical depression, CPAP
reduced depression

Findings

586
W. A. BARDWELL et al.

1983 15 male, 5
female

1981 20 males

Sachs

Beutler

Polysomnography
Association of
Sleep Disorder
apnea criteria

Apnea verification

10 males

Polygraphic
monitoring

Polysomnography:
50 apneas/night >
10 seconds

63 males, 2 Polysomnography
females

Controls

MMPI

Eyesenck Personality
Questionnaire,
Karolinska Scales of
Personality, CA
Psychological
Inventory,
MarloweCrowne
MMPI, POMS

SADS-L
KDS-1,2,3A

Instruments

Tracheostomy
(8 subjects)

Treatment

Apneics with overall elevated


MMPI, including higher MMPI
hypochondriasis and hysteria;
apneics lower on POMS vigor;
apneics more dependent on
external events, more oriented
to physical well-being and
physical symptoms than
controls
24% had seen psychiatrist for
anxiety or depression; 28%
showed elevated MMPI
depression; tracheostomy
improved personality and
MMPI depression

Of 25 apneics, 3 had major


depression history, 2 had
chronic depression, 1 had
cyclothymia, 4 had alcohol
abuse, 10 met criteria for
psychiatric disorder;
depression correlated with
REM, REM latency, and use of
HTN medications
Apneics higher on neuroticism
with a trend toward higher on
psychoticism; apneics higher on
social desirability; apneics
trend toward higher somatic and
muscular tension

Findings

AHI, Apnea/Hypopnea Index; BDI, Beck Depression Inventory; GHQ, General Health Questionnaire; HADS, Hospital Anxiety & Depression Scale;
KDS, KupferDetre Self-Rating Scale; MMPI, Minnesota Multiphasic Personality Inventory; MOS, Medical Outcome Study; POMS, Profile of Mood
States; QOL, Quality of Life; SADS-L, Schedule for Affective Disorders and Schizophrenia; SCL-90-R, Symptom Checklist90, Revised; SIP, Sickness
Impact Profile; UWIST, University of Wales Institute of Science and Technology.

Guilleminault 1977 25 males

1984 25 male

Reynolds

Subjects

Year

Authors

Table I.(Continued)

Psychological correlates of sleep apnea

587

3359
21.831.1
25
112.5
99133
6489

12062.0
7561.2

Range

44.061.3
26.960.6
3.560.2
5.160.7

Mean

14661.8
9561.1

46.861.4
28.461.0
2.960.3
6.861.0

Mean

131159
84100

4059
22.136.1
15
1.314.5

Range

HTN (n516)

12161.5
7861.0

49.061.2
29.260.6
3.060.1
46.463.7

Mean

90137
6289

3464
2038.9
15
15110.7

Range

No HTN (n548)

14961.8
9361.2

49.961.6
31.160.9
2.760.2
61.066.8

Mean

90170
62104

3264
2040.1
15
17.6142.5

Range

HTN (n524)

Patients with apnea (n572)

HTN, hypertension (mean systolic blood pressure>140 mmHg or mean diastolic blood pressure>90 mmHg).
a
Age: main effect for apnea (p50.008).
b
Body mass index: main effects for apnea (p,0.003) and HTN (p50.039).
c
Respiratory disturbance index: main effects for apnea (p,0.001) and HTN (p50.042).
d
Systolic blood pressure: main effect for HTN (p,0.001).
e
Diastolic blood pressure: main effect for HTN (p,0.001).

Age
Body mass indexb
Social class
Respiratory Disturbance Indexc
Screen blood pressure
Systolicd
Diastolice

No HTN (n524)

Patients without apnea (n540)

Table II.Demographic characteristics

588
W. A. BARDWELL et al.

Psychological correlates of sleep apnea

589

of ideal body weight were excluded due to the possibility of confounding by other conditions associated
with obesity. Subjects were also excluded if they had any other major illness such as diabetes or depression.
Subjects had their sleep monitored for two nights in the Clinical Research Center. Polysomnography
included central and occipital electroencephalogram (EEG); bilateral electrooculogram (EOG); submental electromyogram (EMG); airflow, thoracic, and abdominal excursions with Respitrace; and tibialis EMG. Sleep records were scored according to the criteria of Rechtshaffen and Kales [23], and the
numbers of apneas and hypopneas were recorded. The majority of subjects had obstructive apneas only,
but some also showed evidence of central apneas. Subjects were classified as having sleep apnea if their
two-night average RDI was>15.
Blood pressure (BP) was measured with a Dinamap 845 XT monitor. BPs obtained from the Dinamap
correlate well (r.0.95) with mercury sphygmomanometry readings [24]. Subjects were seated for at least
5 minutes and acclimated to the equipment prior to BP determinations. Three consecutive readings were
taken on each of two occasions approximately 1 week apart. The BP used in the current investigation
was the mean of the six readings. Hypertension was defined as mean systolic BP (SBP)>140 mmHg or
mean diastolic BP (DBP)>90 mmHg. Participants who were hypertensive were tapered from their medication prior to participating in the study.

Variable selection
Psychological variables. In our behavioral variables, we employed a set of commonly used measures
of dysphoric mood: the Spielberger Trait Anxiety Scale; Center for Epidemiological StudiesDepression (CESD) Scale; CookMedley (CM) Stress Subscale; Profile of Mood States (POMS) Tension,
Depression, Fatigue, and Vigor Subscales; and POMS Total Mood Disturbance Factor. We also looked
at a set of behavioral medicine variables that focus on aspects of anger: BussDurkee (BD) Irritability
Subscale and BD Total Anger, Experience of Anger, and Expression of Anger Factors; CM Hostility
and Cynicism Subscales; and POMS Anger.
Several scales related to anger were intentionally selected. There are subtle differences in these measures: whereas one may measure a persons experience of anger (e.g., BD Experience of Anger Factor,
POMS Anger), another may measure how one expresses anger (e.g., BD Expression of Anger Factor),
and yet another may include several dimensions of anger (e.g., BD Total Anger Factor).

Sleep variables
The following sleep variables were chosen because they have been commonly used in other studies
of sleep and psychological/cognitive factors: deep sleep and REM sleep (each calculated as a percentage
of total sleep time [TST]); TST; wake time after sleep onset (WASO); RDI; and hypoxemia (defined as
the percent of time that oxygen saturation was ,90%).

Data analysis
Data analysis involved four phases. First, an omnibus test was conducted of the relationship between
sleep variables and psychological variables. This provides protection against type I error due to multiple
comparisons of these variables. Specifically, the maximal canonical correlation between sleep variables
and psychological variables was computed. Its statistical significance was determined from its permutation distribution under the null hypothesis (see chapter 15 of Efron and Tibshirani [25] for a review of
permutation testing).
Second, patients with and without sleep apnea were contrasted in terms of the psychological variables
using t-tests and then using analysis of covariance (ANCOVA), while controlling for age, body mass index (BMI), and hypertension (HTN). Third, correlations between the psychological and sleep variables
were examined as follows:
1.
2.
3.
4.

In
In
In
In

patients
patients
patients
patients

with sleep apnea.


without sleep apnea.
with sleep apnea, while controlling for age, BMI, and HTN.
without sleep apnea, while controlling for age, BMI, and HTN.

Finally, multiple regression analysis was used to determine if the presence/absence of sleep apnea affected the relationships between individual psychological variables and individual sleep variables. Unlike
the correlations, the multiple regression analyses used a single data set consisting of patients with and
without sleep apnea, using a dummy variable to differentiate the groups. Each model included a psychological variable, the dummy variable (indicating presence/absence of sleep apnea), and an interaction
term (dummy variable multiplied by the psychological variable). If the interaction terms proved to be
significant, this would indicate that the presence/absence of sleep apnea is important in understanding
the relationship between sleep and psychological variables.

590

W. A. BARDWELL et al.

Balancing type I and type II error


In any inquiry into the relationship between two sets of variables, such as sleep and mood state variables, one must balance off type I and type II error rates. On the one hand, one needs to make some
adjustments for multiple comparisons in some situations. When searching for relationships among variables not previously studied, control of family-wise type I error rates via multiple comparison adjustments is sensible. On the other hand, such adjustments should not be too stringent, thereby masking a
likely relationship. However, when attempting to verify previously reported negative or positive results,
adjustment for multiple comparisons is less advisable and may lead to spurious results such as the refutation of correct, positive results that were previously reported. In this study, we only investigated relationships that were previously reported, so comparison-wise p-values are reported.

RESULTS

Analysis of demographic variables and data inspection


Table II indicates that there were some demographic differences between the
groups. Patients with sleep apnea were, on average, about 4 years older (p50.008)
and somewhat heavier (p50.003). There was no difference in socioeconomic status
(SES) as measured by the Hollingshead Social Class Scale [26] and in BP. The hypertensive group was slightly heavier (p50.039) and, of course, had higher SBP
(p,0.001) and DBP (p,0.001) than the nonhypertensive group. There was no significant difference in age or social class. Upon inspection of data, we noticed that
one individuals psychological test scores were extraordinarily high (up to 2 sds
higher than the rest of the sample). This patient was undergoing enormous family
turmoil at the time of the study. Because of her unusual circumstances, and their
evident effects on her scores, we felt it prudent to drop her data from the analyses
because of the likely contamination effect of her family crisis.
Analysis of sleep variables and psychological variables
Omnibus test. The omnibus test was significant (p50.014), indicating there was a
relationship between the group of sleep variables and the group of psychological
variables. This provides protection against inflated type I error due to multiple comparisons. Therefore, we proceeded with the additional planned analyses.
Use of t-tests and ANCOVA. Simple t-tests were conducted to see if patients with
and without sleep apnea differed in terms of individual sleep and psychological variables. None of the tests for the psychological variables were significant. For the
sleep variables, patients without sleep apnea showed higher levels of deep sleep
(9.6% vs. 5.7%, p50.017) and less TST (265 minutes vs. 294 minutes, p50.024) than
patients with sleep apnea. As expected, patients with sleep apnea showed higher
RDI (51 vs. 6, p,0.001) and hypoxemia (18.0% vs. 1.3%, p,0.001) than patients
without sleep apnea.
When ANCOVA was used to control for age, BMI, and HTN, there were still no
significant differences for the psychological variables. For the sleep variables, patients without sleep apnea continued to show higher levels of deep sleep (p50.038),
and patients with sleep apnea continued to show greater RDI (p,0.001) and hypoxemia (p,0.001). However, the difference in TST was eliminated.
Correlation analysis
The previous analyses treated sleep apnea dichotomously, as is commonly done
in the clinical literature. However, we wanted to see if we would obtain a different
perspective on the links between sleep apnea and mood if we took advantage of the

Psychological correlates of sleep apnea

591

continuous nature of the sleep variables. Therefore, correlational analyses were


conducted. The following psychological variables were not correlated with any
sleep variables and, therefore, were not studied further in subsequent analyses: BD
Irritability Subscale; CM Hostility, Cynicism, and Stress Subscales; Spielberger
Trait Anxiety Scale; and POMS Tension, Fatigue, and Depression Subscales. In addition, WASO was not correlated with any psychological variables and was not
studied further.
Results for patients with sleep apnea. Overall, for patients with sleep apnea, higher
levels of dysphoric mood were significantly related to increased deep sleep, REM,
and hypoxemia (see Table III). TST and RDI did not correlate with any of the psychological variables.
Results for patients without sleep apnea. In general, for patients without sleep apnea, sleep stages, RDI, and hypoxemia were not important correlates of the psychological variables (see Table III). Also, in contrast to patients with sleep apnea, for
patients without sleep apnea, only one measure of dysphoric mood (anger expression) was correlated with a sleep variable.
Results for patients with sleep apneacontrolling for age, BMI, and HTN. In patients with sleep apnea, after controlling for age, BMI, and HTN, only one measure
of dysphoric mood remained significant: POMS Anger was still positively correlated
with REM sleep (r50.35, p50.005) (see Table IV). In addition, POMS Vigor was
found to be positively correlated with TST (r50.27, p50.035).
Results for patients without sleep apneacontrolling for age, BMI, and HTN.
Controlling for age, BMI, and HTN produced only two changes in the correlation
results for patients without sleep apnea (see Table IV). Anger (BD Experience of
Anger Factor) became significantly correlated with RDI (r50.33, p50.050); however, POMS Vigor was no longer significantly correlated with hypoxemia.

Multiple regression analysis


Multiple regression analysis was used to determine if the presence/absence of
sleep apnea affected the relationships between individual psychological variables
and individual sleep variables. A single data set, which included patients with and
without sleep apnea, was utilized, using a dummy variable to differentiate the patient groups. Each model tested included a psychological variable, the dummy variable (indicating presence/absence of sleep apnea), and an interaction term (the
dummy variable multiplied by the psychological variable).
With TST as the dependent variable, significant interactions were found between
presence/absence of sleep apnea and BD Total Anger (R250.09, p50.049) and between presence/absence of sleep apnea and BD Anger Expression Factor (R250.12,
p50.010). In both cases, it was the patients without sleep apnea whose level of anger
varied with sleep time; that is, as TST decreased, patients anger increased. Patients
with sleep apnea showed no relationship between total sleep time and anger.
DISCUSSION

Several design aspects characterize this study: ample sample size; control of confounders likely related to sleep apnea; and demonstration of the implications of different methods of analysis. Our sample consisted of 72 subjects with sleep apnea
and 40 subjects without sleep apnea. Most previous studies have used fewer subjects, often with no control group.

r50.26, p50.036

p,0.01 (all other p-values,0.05).

Patients without apnea


Depression
CESD
Anger
BD experience
BD expression
BD total anger
POMS anger
Mood disturbance
POMS total mood
Energy
POMS vigor

Patients with apnea


Depression
CESD
Anger
BD experience
BD expression
BD total anger
POMS anger
Mood disturbance
POMS total mood
Energy
POMS vigor

Deep sleep
(% of TST)

r50.26, p50.039

r50.37, p50.003a

r50.28, p50.022

r50.27, p50.026

REM
(% of TST)

r50.32, p50.043

r520.39, p50.017

Total sleep
time
(TST)

Respiratory
disturbance
index

Table III.Correlations between sleep variables and psychosocial variables

r520.36, p50.042

r50.26, p50.041
r50.26, p50.046

Percentage
of time
with oxygen
saturation of
,90%

592
W. A. BARDWELL et al.

p,0.01 (all other p-values,0.05).

Patients without apnea


Depression
CESD
Anger
BD experience
BD expression
BD total anger
POMS anger
Mood disturbance
POMS total mood
Energy
POMS vigor

Patients with apnea


Depression
CESD
Anger
BD experience
BD expression
BD total Anger
POMS anger
Mood disturbance
POMS total mood
Energy
POMS vigor

Deep
Sleep
(% of TST)

r50.35, p50.005a

REM
(% of TST)

r50.39, p50.016

r520.35, p50.040

r50.27, p50.035

Total sleep
time
(TST)

r50.33, p50.050

Respiratory
disturbance
index

Percentage of time
with oxygen
saturation,90%

Table IV.Correlations between sleep variables and psychosocial variablescontrolling for age BMI, and HTN

Psychological correlates of sleep apnea

593

594

W. A. BARDWELL et al.

Table V.Overview of correlations between psychological and sleep variables


Psychological
variable
Depression
CESD
CESD
Anger
BD experience
BD experience
BD expression
BD expression
BD total anger
POMS anger
Mood disturbance
POMS total
Vigor
POMS vigor
POMS vigor

Without covariates

Sleep
variable

Apnea

Deep sleep
REM

1
1

REM
RDI
Hypoxemia
TST
Hypoxemia
REM

REM

Nonapnea

1
1
1

TST
Hypoxemia

With covariates of Age, BMI, HTN


Apnea

Nonapnea

1
2

2
1

1
2

1, Positive correlation; 2, negative correlation.

Other potential confounding factors (age, BMI, and HTN) were measured,
allowing us to compare results when these variables were controlled and when they
were not controlled in the analyses. Our results showed that these covariates are important for patients with sleep apnea, but relatively unimportant for patients without sleep apnea. Because reproductive hormonal status (i.e., menstrual cycle, menopause, and hormone replacement therapy) can potentially influence both sleep and
mood, it might be of interest for future studies to examine this area explicitly. We
did not examine this area in the current study and feel that the statistical power
is lacking to make these judgments with confidence, given the small number of
women included.
We found several noteworthy patterns of relationships between sleep and psychological measures, and these patterns differed for patients with and without sleep
apnea. For patients with sleep apnea, dysphoric mood (anger, depression, and total
mood disturbance) was primarily associated with amount of REM sleep. Also, depression was related to deep sleep, and anger was related to hypoxemia. All of these
correlations were positive. These results agree with some previously reported findings in studies of patients with sleep apnea, but only partially agree with a recent
metaanalysis of sleep disorders in psychiatric patients. Keep in mind we did not
study psychiatric patients and most mood scores for our subjects were not in the
pathological range. Nonetheless, Benca et al. found that patients with affective disorders showed increased REM but less deep sleep and total sleep time [27].
Table V pictorially summarizes how relationships between sleep and mood variables are affected by the use of covariates. When we controlled for age, BMI, and
HTN in patients with sleep apnea, a new variable, vigor (i.e., ebullience, high energy), became positively correlated with TST. However, many of the relationships
between sleep and mood disappeared. The previously noted links between sleep
and depression, total mood disturbance, and all but one of the anger relationships
vanished when controlling for these factors. This suggests that depression and total
mood disturbance are related more to age, weight, and/or HTN than they are to
sleep apnea. It also suggests that anger is related to the amount of REM sleep in

Psychological correlates of sleep apnea

595

patients with sleep apneaat least when using the POMS Anger Subscale, a measure of experience of anger.
For patients without sleep apnea, total sleep time was positively related to vigor
and negatively related to anger expression. These relationships remained even after
controlling for age, BMI, and HTN. It is plausible that the less one sleeps, the less
vigorous one feels and the more likely one is to express anger. However, the direction of causality may be the opposite, with increased anger and, less probably, a decreased sense of vigor, resulting in diminished total sleep time.
We have increased confidence that, for individuals without sleep apnea, deep
sleep, REM sleep, and hypoxemia are of relatively little importance vis-a`-vis dysphoric mood states. Total sleep time is the more important sleep variable for patients without sleep apnea. On the other hand, for patients with sleep apnea,
amount of REM sleep is related to anger, whereas total sleep time appears to be
important only in terms of the level of vigor they report. This supports findings by
other investigators who also found links between REM and anger [28] and between
TST and vigor [2932].
Our results lend some support to previous findings that patients with sleep apnea
experience more anger than patients without sleep apnea. However, when patients
without sleep apnea lose sleep, anger expression seems to be involvedeither as the
cause or the result of reduced sleep time. Both patients with and without sleep apnea experience increased vigor the more they sleep. Although this seems intuitively
obvious, it indirectly supports previous findings that patients with sleep apnea show
higher levels of fatigue. However, we were not able to find relationships between
sleep variables and our direct measure of fatigue, nor with our measures of stress,
tension, or anxiety.
Our reading of the data is that many of the previously reported links between
mood and sleep in individuals with sleep apnea disappear after controlling for covariates (age, BMI, HTN). Our second observation is that mood and sleep are variously linked, depending on the population being studied. For individuals with sleep
apnea, there are several more links between measures of dysphoric mood and sleep,
but most of these vanish after covarying age, BMI, and HTN. For individuals without sleep apnea, there are fewer relationships between dysphoric mood and sleep.
Psychiatrists and behavioral medicine researchers are likely to pay increasing attention to sleep. Our findings point to the implications of considering variables long
known to behavioral medicine researchers to be important covariates (i.e., age,
BMI, HTN). Table V demonstrates the profound impact they can haveparticularly for patients with sleep apnea. Future studies examining psychological correlates of sleep apnea must control for the effects of these confounders, which may
also be important in resolving disparities in findings from previous research.
AcknowledgmentsThis work was supported by grants HL44915, RR00827, AG08415, and AG02711
from the National Institutes of Health.

REFERENCES
1. Brown LK. Sleep apnea syndromes: overview and diagnostic approach. Mt Sinai J Med 1994;
61:99112.
2. Cheshire K, Engleman H, Deary I, Shapiro C, Douglas NJ. Factors impairing daytime performance
in patients with sleep apnea/hypopnea syndrome. Arch Intern Med 1992;152:538541.

596

W. A. BARDWELL et al.

3. Guilleminault C, Eldridge FL, Tilkian A, Simmons FB, Dement WC. Sleep apnea syndrome due to
upper airway obstruction: a review of 25 cases. Arch Intern Med 1977;137:296300.
4. Millman RP, Fogel BS, McNamara ME, Carlisle CC. Depression as a manifestation of obstructive
sleep apnea: reversal with nasal continuous positive airway pressure. J Clin Psychiatry 1989;
50:348351.
5. Reynolds C, Kupfer D, McEachran A, Taska L, Sewitch D, Coble P. Depressive psychopathology
in male sleep apneics. J Clin Psychiatry 1984;45:287289.
6. Flemons WW, Tsai W. Quality of life consequences of sleep-disordered breathing. J Allerg Clin Immunol 1997;99(suppl):S750S756.
7. Engleman HM, Martin SE, Deary IJ, Douglas NJ. Effect of continuous positive airway pressure
treatment on daytime functioning in sleep apnea/hypopnea syndrome. Lancet 1994;343:572575.
8. Edinger J, Carwile S, Miller P, Hope V, Mayti C. Psychological status, syndromatic measures, and
compliance with nasal CPAP therapy for sleep apnea. Percept Motor Skills 1994;78:11161118.
9. Borak J, Cieslicki J, Szelenberger W, Wilczak-Szadkowska H, Koziej M, Zielinski J. Psychopathological characteristics of the consequences of obstructive sleep apnea prior to and three months after
CPAP. Psychiatria Polska 1994;28:3344.
10. Platon MJ, Sierra JE. Changes in psychopathological symptoms in sleep apnea patients after treatment with nasal continuous positive airway pressure. Int J Neurosci 1992;62:173195.
11. Derderian SS, Bridenbaugh RH, Rajagopal KR. Neuropsychologic symptoms in obstructive sleep apnea improve after treatment with nasal continuous positive airway pressure. Chest 1988;94:10231027.
12. Kales A, Caldwell AB, Cadieux RJ, Vela-Bueno A, Ruch LG, Mayes SD. Severe obstructive sleep apneaII: associated psychopathology and psychosocial consequences. J Chronic Dis 1985;38:427434.
13. Flemons WW, Whitelaw WA, Brant R, Remmers JE. Likelihood ratios for a sleep apnea clinical prediction rule. Am J Respir Crit Care Med 1994;150:12791285.
14. Gall R, Isaac L. Quality of life in mild sleep apnea. Sleep 1993;16(suppl):S59S61.
15. Lee S. Depression in sleep apnea: a different view. J Clin Psychiatry 1990;51:309310.
16. Klonoff H, Fleetham J, Taylor DR, Clark C. Treatment outcome of obstructive sleep apnea: physiological and neuropsychological concomitants. J Nerv Ment Dis 1987;175:208212.
17. Sachs C, Levander S. Personality dimension in patients with sleep-apnea: comparison with narcolepsy. Pers Indiv Diff 1983;4:563567.
18. Beutler LE, Ware JC, Karacan I, Thornby JI. Differentiating psychological characteristics of patients with sleep apnea and narcolepsy. Sleep 1981;4:3947.
19. Cassel W. Cognitive effects and daytime sleepiness (a) psychosocial sequelae of sleep disordered
breathing: sleep apnea and personality. Sleep 1993;16(suppl):S56S58.
20. Diamond E. The role of anger and hostility in essential hypertension and coronary heart disease.
Psychol Bull 1982;92:410433.
21. Dimsdale J, Pierce C, Schoenfeld D, Brown A, Zusman R, Graham R. Suppressed anger and blood
pressure: the effects of race, sex, social class, obesity, and age. Psychosom Med 1986;48:430436.
22. Metropolitan Life Foundation. 1983 Metropolitan height and weight tables. Stat Bull (Metropolitan
Life Insurance) 1983;64:1.
23. Rechtshaffen A, Kales A, eds. A manual of standardized terminology: techniques and scoring system
for sleep stages of human subjects. Los Angeles, California: UCLA Brain Information Service/Brain
Research Institute 1968.
24. Silas J, Barker A, Ramsay L. Clinical evaluation of Dinamap 845 automated blood pressure recorder. Br Heart J 1980;43:202205.
25. Efron B, Tibshirani RJ. An introduction to the bootstrap. New York: Chapman & Hall, 1993.
26. Hollingshead A. Social class and mental illness. New York: John Wiley & Sons 1958:398407.
27. Benca RM, Obermeyer WH, Thisted RA, Gillin JC. Sleep and psychiatric disorders: a meta-analysis.
Arch Gen Psychiatry 1992;49:651668.
28. Madigan MF Jr, Dale JA, Cross JD. No respite during sleep: heart-rate hyperreactivity to rapid eye
movement sleep in angry men classified as Type A. Percept Motor Skills 1997;85:14511454.
29. Bonnet MH, Arand DL. The consequences of a week of insomnia. Sleep 1996;19:453461.
30. Bonnet MH, Arand DL. The consequences of a week of insomnia. II: Patients with insomnia.
Sleep 1998;21:359368.
31. Penetar D, McCann U, Thorne D, Kamimori G, Galinski C, Sing H, Thomas M, Belenky G. Caffeine
reversal of sleep deprivation effects on alertness and mood. Psychopharmacology 1993;112:359365.

Вам также может понравиться