Download The Current Prevalence of Sleep Disordered Breathing in

Scientific investigations
The Current Prevalence of Sleep Disordered Breathing in Congestive Heart Failure
Patients Treated with Beta-Blockers
Mary Macdonald, RPSGT; James Fang, M.D.; Steven D. Pittman, MSBME; David P. White, M.D.; Atul Malhotra, M.D.
Divisions of Cardiovascular Diseases, Sleep Medicine, and Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital and Harvard
Medical School, Boston, MA
Study Objectives: Although sleep disordered breathing is thought to
be common in patients with systolic heart failure, prior studies are difficult to interpret due to a variety of factors including small sample sizes,
referral bias to sleep laboratories among participants, lack of modern
medical therapy for congestive heart failure, and the failure to use
modern techniques to assess breathing such as nasal pressure. Our
objective was to determine the current prevalence of sleep disordered
breathing in a state-of-the-art congestive heart failure clinic.
Methods: We conducted a prospective study of consecutive patients
who visited our heart failure clinic to assess the prevalence of sleep
apnea in all eligible patients on maximal medical therapy. We used
4-channel recording equipment and modified Chicago criteria for scoring respiratory events (using heart rate response as a surrogate for
arousal from sleep).
Results: We observed that among the 108 participants, 61% had
some form of sleep disordered breathing (31% central apnea with
Cheyne Stokes respiration and 30% obstructive sleep apnea). Sleep
disordered breathing was significantly associated with atrial fibrillation
(OR = 11.56, p = 0.02) and worse functional heart failure class (OR =
2.77, p = 0.02), after adjusting for male sex, age over 60 years, body
mass index, and left ventricular ejection fraction.
Conclusions: We conclude that both obstructive and central sleep
apnea remain common in congestive heart failure patients despite
advances in medical therapy, and that the previously reported high
prevalence values are unlikely to be explained by referral bias or participation bias in prior studies. These data have important clinical implications for practitioners providing CHF therapy.
Keywords: Apnea, sleep, congestive heart failure, obstructive, central, outcome, lung
Citation: Macdonald M; Fang J; Pittman SD; White DP; Malhotra A.
The current prevalence of sleep disordered breathing in congestive
heart failure patients treated with beta-blockers. J Clin Sleep Med
2008;4(1):38-42.
S
leep disordered breathing (SDB) appears to be common
among patients with congestive heart failure (CHF).1,2 Prior
studies have suggested that 40%-50% of patients with CHF and
left ventricular systolic dysfunction will have some form of
SDB, either obstructive or central sleep apnea (CSA).1-3 However, important limitations exist in these previous reports. Many
were based in sleep laboratories and therefore subject to referral and participatory bias.2 To our knowledge, only one study,
using nonconsecutively enrolled subjects, has prospectively
evaluated the prevalence of SDB in outpatients in a heart failure
clinic.4 Other studies have been limited by technical issues such
as the use of respiratory monitoring systems that do not employ
state-of-the-art sensors for ventilation i.e., nasal pressure to assess hypopneas.5 Previous studies have also been difficult to
reconcile due to variable definitions of CSA, and a failure to
distinguish between CSA in general and Cheyne Stokes respiration (CSR) in particular. In many cases, it can be difficult to distinguish obstructive from central apnea type in CHF patients,
leading some authors to suggest that different metrics for sleep
disordered breathing severity may be desirable.6
Perhaps most importantly, many previous studies have not
included study populations with currently optimal medical management including adequately dosed β-blockers, angiotensin
converting enzyme inhibitors (ACEI), angiotensin II receptor antagonists (ARBs) and aldosterone antagonists.7 This lack of optimized medical therapy may have particular relevance since such
therapies may have an impact on the prevalence of SDB such as
CSR.8,9 For example, β-blockers are known to influence hypoxic chemosensitivity which would be predicted to influence the
development of CSR.10,11 Similarly, ACEI and diuretic therapy
can both contribute to lower intracardiac filling pressures, which
could importantly influence the occurrence of CSR.8,12
The goals of this study were to (1) prospectively evaluate the
prevalence of SDB in consecutive medically optimized outpa-
Disclosure Statement
This study was supported in part by Respironics. Dr. White is employed
by Respironics and has received research and consulting fees from Itamar Medical and consulting fees from Aspire Medical. Dr. Malhotra has
received research support from Respironics, Inspiration Medical, NMT
Medical, Restore Medical, Pfizer, and Cephalon. Dr. Fang and Ms. Macdonald have indicated no financial conflicts of interest. Mr. Pittman is an
employee of Respironics.
Submitted for publication June, 2007
Accepted for publication October, 2007
Address correspondence to: Atul Malhotra, Sleep Disorders Research
Program @ BIDMC, Brigham and Women’s Hospital, 75 Francis Street,
Boston, MA, 02115; Tel: (617) 732-6488; Fax: (617) 732-7337; E-mail:
[email protected]
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
38
Sleep Apnea in Heart Failure
Table 1—Baseline Clinical Characteristics of Study Participants
and Non-Participants*
Table 2—Clinical Characteristics of Subjects with and without
Sleep Disordered Breathing*
Participants Non-Participants p value
(n=108)
(n=45)
Female n,%
16 (15)
14 (31)
0.04
Age (yr)
57 ± 11
58 ± 11
0.53
BMI (kg/m2)
26.8 ± 5.8
28.9 ± 6.6
0.13
LVEF, %
20 (15-30)
20 (15-25)
0.89
NYHA n, %
Class II
67 (62)
27 (60)
Class III-IV
41 (38)
18 (40)
0.96
Etiology
Ischemic
47 (44)
26 (58)
Idiopathic
61 (66)
19 (42)
0.15
β-blockers n,%
39 (82)
39 (86)
0.68
Other Medications, %
Loop diuretics
88
**
ACE inhibitor
83
**
Digitalis
63
**
Anticoagulant
48
**
Spironolactone
36
**
Statin
35
**
Antiplatelet
34
**
Antiarrhythmic
30
**
Nitrate
25
**
ARB
13
**
AHI (/h)
24.7 ±17.2
**
CSR time, %
24 ±30
**
Atrial fibrillation n, % 15 (14)
**
no SDB
(n=42)
Prevalence
39%
Female n,%
7 (17)
Age (yr)
56 ±9
BMI (kg/m2)
26.8 ±5.8
AHI (/h)
9.5 (6.4-11.5)
CSR time, %
0 (0-7)
LVEF, %
20 (15-25)
NYHA n, %
Class II
31 (74)
Class III-IV
11 (26)
Etiology
Ischemic
15 (36)
Idiopathic
27 (64)
β-blockers n, %
39 (93)
Atrial fibrillation n, % 1 (2)
36 (55)
30 (45)
0.07
32 (49)
34 (51)
50 (77)
14 (21)
0.27
0.06
0.01
*Unless indicated, values expressed as mean ±SD or median (95%
CI range)
since a repeat echocardiogram is performed if any important
change in clinical status occurs. All echocardiograms were
read by board certified cardiologists with advanced training in
echocardiography. The Teichholz formula is routinely used to
assess LVEF, which must be agreed upon by the overreading
echocardiographer.
Subjects underwent an in-home, unattended overnight study
with a standard 4-channel recording device (StarDust, Respironics, Inc, Murrysville, PA). This device records nasal pressure, thoracic excursion (as measured by a piezoelectric crystal), body position, pulse oximetry, and heart rate derived from
pulse oximetry. The data were downloaded and scored by an
experienced scorer.
A modified version of the 1999 American Academy of Sleep
Medicine (AASM) criteria for scoring respiratory events was
used.13,14 An obstructive apnea was defined as ≥10-sec cessation of airflow as measured by nasal pressure associated with
the continuation of thoracic effort. A central apnea was defined
as ≥10-sec cessation of airflow without thoracic effort. A hypopnea was defined as a 50% reduction in airflow for 10 sec or a
discernable change in airflow with either a ≥3% oxyhemoglobin desaturation or an arousal, as defined by a 10% increase
in heart rate.15 Arousal was defined based on a 10% change in
heart rate because EEG was not available using the StarDust
system. We believe that a relative rather than absolute change
in heart rate is more reliable for comparing patients with and
without β-blockade. The apnea-hypopnea index (AHI) and central apnea index (CAI) were calculated based on total number
of events per hour of total recording time (TRT).
SDB was defined by AHI of ≥15 events per hour. Another
metric of CSR was also used for exploratory purposes. In this
case, CSR was defined as a symmetrical crescendo-decrescendo
respiratory pattern with a >50% difference between peak and
nadir nasal pressure or respiratory effort amplitude, occurring
within a 30 to 90-second period. Primary CSR was defined as a
study with a CSR time (as a percentage of total recording time)
≥33%. Primary OSA was predefined as an AHI ≥15 events per
hour and a CSR time of <33%.
*Unless indicated, values expressed as mean ±SD or median (95%
CI range)
** not compared
tients in a heart failure clinic; (2) examine risk factors for SDB
and specifically CSR in this population; and (3) explore alternative metrics for CSR.
METHODS
One hundred and eight subjects out of 153 consecutively eligible patients from the Heart Failure Clinic of the Brigham &
Women’s Hospital agreed to participate. Prior to enrollment,
subjects were not asked about symptoms or risk factors for
SDB. Subjects were included if they had clinically stable heart
failure; were between the ages of 18 and 85 years; had New
York Heart Association (NYHA) functional class II, III, or IV;
and had a left ventricular ejection fraction (LVEF) of <40%.
Subjects were excluded if they had primary valvular heart disease, primary diastolic failure (i.e., congestive heart failure with
normal systolic function), a history of major lung disease (i.e.,
obstructive pulmonary disease), a history of pneumothorax in
the prior 6 months, current or previous use of positive airway
pressure noninvasive ventilation, current use of supplemental
oxygen, or the presence of an artificial airway. Subjects who
had been hospitalized or had medication changes within the
past 30 days were also excluded. All subjects provided written informed consent. This study was approved by the Partners’
Institutional Review Board.
Subjects underwent echocardiographic assessment of
LVEF 12 months prior to performing a sleep study, during which time LVEF was unlikely to significantly change,
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
SDB
p value
(n=66)
61%
9 (14)
0.88
58 ±12
0.36
28.9 ±6.6
0.13
32.9 (24.3-41.6) <0.001
34 (9-72)
<0.001
20 (15-30)
0.29
39
M Macdonald, J Fang, S Pittman et al
Table 3—Clinical Characteristics of Subjects with and without
Cheyne Stokes Respiration*
no CSR
(n=74)
Prevalence
69%
Female n, %
13 (18)
Age (yr)
56 ± 10
BMI (kg/m2)
27.0 ± 6.8
AHI (/h)
13.6 (8.1-24.7)
CSR time, %
3.1 (0.0-10.3)
LVEF, %
24 (15-30)
NYHA n, %
Class II
52 (70)
Class III-IV
22 (30)
Etiology
Ischemic
31 (66)
Idiopathic
43 (34)
β-blockers n,%
66 (89)
Atrial fibrillation n, % 6 (8)
CSR
p value
(n=34)
31%
3 (9)
0.37
59 ± 13
0.32
28.5 ± 5.7
0.68
37.8 (28.7-46.0) <0.001
68.6 (39.3-79.2) <0.001
15 (13-25)
0.02
15 (44)
19 (56)
0.02
16 (47)
18 (53)
23 (68)
9 (26)
0.77
0.49
0.02
Figure 1—The prevalence of Cheyne Stokes Respirations (CSR)
varies with the definition used. With a rising threshold for percentage of CSR time, there is a falling prevalence of this breathing
abnormality. At a cutoff of 33% CSR time, almost one third of
participants had this disease.
*Values expressed as mean ±SD or median (95% CI range)
Differences between groups were assessed using the unpaired
Student’s t-test for parametric results and the Mann-Whitney
rank sum test for nonparametric data. χ2 was used to compare
proportions between groups. Logistic regressions were used to
assess the odds ratios of SDB, and CSR, conferred by various
independent variables. A value of p < 0.05 was considered statistically significant, with results being reported as the mean ±
SD or median and 95% confidence intervals.
Candidate independent variables were first established using
a univariate model. The final model included only those variables with a significant effect on the dependent variable as measured by the likelihood ratio test statistic (p < 0.05). Multiple
logistic regression was then used to establish the multivariate
variables that remained independently predictive of SDB and
CSR. The same regression model was also used to determine if
the self-reported use of β-blockers independently predicted the
presence or absence of SDB and CSR Odds ratios and 95% confidence intervals were calculated. Calculations were performed
in SigmaStat 3.0 software.
After adjusting for male sex, age over 60 years, BMI, and
LVEF, subjects with SDB had a nearly 12-fold increased odds
for atrial fibrillation (OR = 11.56, 95% CI 1.43 – 93.02, p =
0.02), and a significantly greater odds for a worse functional
class of heart failure (OR = 2.77, 95% CI 1.14 – 6.73, p =
0.02).
Cheyne Stokes Respiration
The group with CSR time ≥33% had a mean CSR time of
68.6% (39.3-79.2); the group without CSR had mean CSR time
3.1% (0.00-10.3) (Table 3, Figure 1).The CSR group had significantly more impaired functional capacity than the group
without CSR (NYHA class III-IV 44% vs 30% class II, p =
0.02), a lower LVEF (15% [13.0-25.0] vs. 25% [15.0-30.0], p
= 0.02), and had more subjects with atrial fibrillation (26% vs.
8%, p = 0.02). There were no significant differences in baseline
demographics, CHF etiology, or β-blocker use. After adjusting
for male sex, age over 60 years, BMI, and LVEF, subjects with
CSR had a nearly 6-fold increased odds for atrial fibrillation
(OR = 5.65 95% CI 1.70 – 18.73, p = 0.01), and a significantly
greater odds for a worse functional class of heart failure (OR =
3.38 95% CI 1.34 – 8.48, p = 0.01).
Using a CAI cut-off of 5/h and 15/h respectively (consistent
with definitions of primary CSA used in previous studies),10,16
the prevalence of CSA in our group is 28% and 17%. NYHA
functional class and the presence of atrial fibrillation still
emerge as significant differences between those with CAI ≥15
and those with a CAI <15/h. (NHYA: 61% Class III-IV vs. 33%,
p = 0.05; a. fib: 39% vs. 9% p = 0.003). β-Blocker use was not
significantly different between groups.
RESULTS
108 subjects (92 males and 16 females) enrolled in the study
(participation rate of 71%). The primary reason for refusal to participate was the perceived inconvenience of the in-home sleep
study. Other than a higher proportion of females among the nonenrolled, there were no significant baseline clinical differences
between those who enrolled and those who did not (see Table 1).
Sleep Disordered Breathing
The prevalence of SDB in this cohort was 61% (57 males and
9 females). 31% of subjects had primary CSR (31 males and 3
females); 30% had primary OSA (26 males and 6 females).
Between the groups with and without SDB (Table 2), the
only significant difference was the presence of atrial fibrillation in the SDB group (21% vs. 2%, p = 0.01). There were no
significant differences in baseline demographics, NYHA class,
LVEF, CHF etiology, or β-blocker use.
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
β-Blocker Usage
Eighty-two percent of all subjects were using β-blockers (Table 4). The group not using β-blockers was significantly more
functionally impaired than the group that was (63% Class III-IV
vs. 33% Class III-IV, p = 0.03). There were no other significant
differences in baseline demographics, severity of SDB, CHF eti40
Sleep Apnea in Heart Failure
Table 4—Clinical Characteristics of Subjects on and off Blockers*
Β-Blockers
(n=89)
Prevalence
82%
Female n, %
11 (12)
Age (yr)
57 ± 10.7
BMI (kg/m2)
28.1 ± 6.4
AHI (/h)
19.4 (10.1-33.2)
CSR time, %
9.0 (0.0-36.5)
LVEF, %
20 (15-30)
NYHA n, %
Class II
60 (67)
Class III-IV
29 (33)
Etiology
Ischemic
38 (43)
Idiopathic
51 (57)
Atrial fibrillation n, % 10 (11)
Table 5—Clinical Characteristics of Subjects with Cheyne Stokes
Respiration vs. Obstructive Sleep Apnea*
no β-Blockers p value
(n=19)
18%
5 (26)
0.23
58 ± 12.9
0.83
28.0 ± 6.4
0.95
31.8 (18.3-37.5) 0.06
17.1 (1.2-43.7)
0.36
25 (15-30)
0.56
7 (37)
12 (63)
0.03
9 (47)
10 (53)
5 (26)
0.91
0.17
CSR
(n=34)
Prevalence
31%
Female n, %
3 (9)
Age (yr)
59 ± 13
BMI (kg/m2)
28.5 ± 5.7
AHI (/h)
37.8 (28.7-46.0)
CSR time, %
68.6 (39.3-79.2)
LVEF, %
15 (13-25)
NYHA n, %
Class II
15 (44)
Class III-IV
19 (56)
Etiology
Ischemic
16 (47)
Idiopathic
18 (53)
β-blockers n,%
23 (68)
Atrial fibrillation n,%
9 (26)
* Values expressed as mean ±SD or median (95% CI range)
ology, or the presence of atrial fibrillation; 77% of subjects with
SDB and 68% with CSR were using β-blockers. After adjusting
for male sex, age over 60, BMI, and LVEF, β-blocker use did not
independently predict the presence or absence of SDB or CSR.
21 (58)
11 (34)
0.13
16 (50)
16 (50)
24 (75)
5 (16)
0.99
0.89
0.44
*Values expressed as mean ±SD or median (95% CI range)
and CSR, an association between NYHA functional class and
CSR has not been previously reported to our knowledge. Although this finding would support the notion that CSR is a
consequence of progressive heart failure, and in fact patients in
our cohort with CSR had significantly lower LVEFs than those
without CSR, the lack of a β-blocker interaction in this study
may suggest otherwise. Another interpretation of these findings
is that aggressive medical management of CHF may improve
LVEF in some patients, whereas a persistently low LVEF may
be a marker of a particularly impaired cardiovascular system.
This impairment could manifest as an elevated filling pressure
which could lead to CSR.9 In addition, as seen in Table 5, there
are some possible differences between those patients with CSR
compared with those with OSA. OSA patients had lower AHI
as well as higher LVEF as compared to those with CSR. However, β-blocker use did not differ between these groups.
One limitation of our study is the use of a four-channel recording device that does not measure sleep. Thus, our measurements of AHI, CAI, and CSR time are based on total recording
time rather than total sleep time (TST). This issue may be relevant in heart failure since sleep efficiency is often low. However, TRT may also be useful (in addition to TST), since many
subjects experience CSR while awake. Furthermore, although
it is recognized that sleep efficiency is low in heart failure, patients were studied at home (using a minimally invasive device)
rather than in the sleep laboratory and therefore sleep efficiency
may have been better than previous reports would indicate. Because our goal was to assess a large group of patients in a CHF
clinic, we believe that our use of home monitoring allowed us
access to the largest and most representative group possible. We
have previously tried to refer all CHF patients for overnight inlaboratory polysomnography, but only the most motivated patients were willing to undergo this test. Thus, although our use
of home monitoring is a potential limitation, it may have helped
us access a more generalizable sample of heart failure patients.
Sensitivity in our study was also optimized by using nasal pressure rather than thermistry as in some of the prior reports.
In conclusion, SDB and CSR remain highly prevalent in subjects with CHF despite optimization of medical therapy. This
Discussion
In this heart failure clinic-based study of SDB in a cohort
of unselected consecutive patients with CHF who were receiving optimally dosed medical management with both β-blockade
and renin-angiotensin system antagonists, a high prevalence of
SDB persisted (61% overall; 31% CSR and 30% OSA) supporting previous work suggesting that SDB is common in CHF. Our
data add to the existing literature by showing that the previously
reported prevalence rates are not likely a product of selection
bias or a result of inadequate or antiquated medical therapy.
Although β-blocker therapy in heart failure is associated
with improvements in LVEF, functional class, and survival,17-19
it did not appear to affect the prevalence of SDB in our cohort,
which was almost uniformly treated with these agents. In fact,
the prevalence of SDB in this study is similar to the rates previously reported in other heart failure cohorts, which range from
41% to 75%. In these previous studies, β-blocker use was either
not reported or limited to a minority of participants. In one of
the only recent true prevalence studies where β-blocker therapy
was reported, only 30% of heart failure patients were on these
agents.4 Furthermore, this prior study was limited to 53 subjects
referred to a sleep laboratory with less advanced heart failure
(NYHA I/II 75%, mean LVEF 34%). Thus, it would appear that
newer therapies for CHF do not have a clinically important impact on overall breathing instability. However, because different
criteria have been used to define the apnea hypopnea index in
each of these studies, further work is clearly needed to determine if variations in prevalence reported in the literature are
methodological or biological.
Patients with both SDB (61%) and CSR (31%) were both
more likely to have atrial fibrillation and poorer NYHA functional class than those without SDB or CSR. These differences
were not explained by differences in demographic variables,
medical therapy, or CHF etiology. Although previous studies
have identified the association of atrial fibrillation with SDB
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
OSA
p value
(n=32)
30%
6 (6)
0.3
57 ± 2
0.59
29.3 ± 1.5
0.62
26.0 (20.1-32.9) <0.001
8.9 (3.1-13.2) <0.001
30 (20-32)
0.001
41
M Macdonald, J Fang, S Pittman et al
observation suggests that the relationship between SDB and
heart failure is still poorly understood since effective therapies
for heart failure appear to have no major effect on SDB. Because available evidence suggests that CSR may contribute to
the progression of heart failure,20 CSR may still represent a potential target of therapy for the persistently symptomatic patient
on optimal medical therapy.21-25
19.
REFERENCES
21.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
20.
Javaheri S, Parker TJ, Wexler L, et al. Occult sleep-disordered
breathing in stable congestive heart failure [published erratum appears in Ann Intern Med 1995 Jul 1;123(1):77]. Ann Intern Med
1995;122:487-92.
Sin D, Fitzgerald F, Parker J. et al. Risk factors for central and obstructive sleep apnea in 450 men and women with congestive heart
failure. Am. J. Respir. Crit. Care Med. 1999;160:1101-1106.
Christ M, Sharkova Y, Fenske H, et al. Brain natriuretic peptide
for prediction of Cheyne-Stokes respiration in heart failure patients. Int J Cardiol 2007;116:62-9.
Ferrier K, Campbell A, Yee B, et al. Sleep-disordered breathing
occurs frequently in stable outpatients with congestive heart failure. Chest 2005;128:2116-22.
Hosselet JJ, Norman RG, Ayappa I, Rapoport DM. Detection of
flow limitation with a nasal cannula/pressure transducer. Am J
Respir Crit Care Med 1998;157:1461-7.
Pack AI, Goldberg LR. Routine polysomnography is not indicated
in congestive heart failure. Con. J Clin Sleep Med 2005;1:19-22.
Sin D, Logan A, Fitzgerald F, et al. Effects of continuous positive airway pressure on cardiovascular outcomes in heart failure
patients with and without Cheyne-Stokes respiration. Circulation
2000;102:61-66.
Solin P, Bergin P, Richardson M, Kaye DM, Walters EH, Naughton MT. Influence of pulmonary capillary wedge pressure on central apnea in heart failure. Circulation 1999;99:1574-9.
Lloyd TC, Jr. Effect of increased left atrial pressure on breathing
frequency in anesthetized dog. J Appl Physiol 1990;69:1973-80.
Witte KK, Thackray S, Nikitin NP, Cleland JG, Clark AL. The
effects of long-term betablockade on the ventilatory responses to
exercise in chronic heart failure. Eur J Heart Fail 2005;7:612-7.
Warner MM, Mitchell GS. Role of catecholamines and beta-receptors in ventilatory response during hypoxic exercise. Respir
Physiol 1991;85:41-53.
Solin P, Roebuck T, Johns DP, Walters EH, Naughton MT. Peripheral and central ventilatory responses in central sleep apnea with
and without congestive heart failure. Am. J. Respir. Crit. Care
Med. 2000;162:2194-200.
Rechtschaffen A, Kales A. A manual of standardized terminology,
techniques and scoring system for sleep stages of human subjects.
Los Angeles: Brain Information Service/Brain Research Institute,
UCLA, 1968.
AASM. Sleep-related breathing disorders in adults: Recommendations for syndrome definition and measurement techniques in
adults. Sleep 1999;22:667-689.
Pitson DJ, Stradling JR. Autonomic markers of arousal during
sleep in patients undergoing investigation for obstructive sleep
apnoea, their relationship to EEG arousals, respiratory events and
subjective sleepiness. J Sleep Res 1998;7:53-9.
Malhotra A, Berry R, White D. Central sleep apnea. In: Carney
PR, Berry RB, Geyer JD, Clinical Sleep Disorders. Philadelphia:
Lippincott Williams & Wilkins, 2005:331-346.
Krum H, Roecker EB, Mohacsi P, et al. Effects of initiating carvedilol in patients with severe chronic heart failure: results from the
COPERNICUS Study. JAMA 2003;289:712-8.
Krum H, Mohacsi P, Katus HA, et al. Are beta-blockers need-
Journal of Clinical Sleep Medicine, Vol. 4, No. 1, 2008
22.
23.
24.
25.
42
ed in patients receiving spironolactone for severe chronic heart
failure? An analysis of the COPERNICUS study. Am Heart J
2006;151:55-61.
Packer M, Coats AJ, Fowler MB, et al. Effect of carvedilol
on survival in severe chronic heart failure. N Engl J Med
2001;344:1651-8.
Lanfranchi PA, Braghiroli A, Bosimini E, et al. Prognostic value
of nocturnal Cheyne-Stokes respiration in chronic heart failure.
Circulation 1999;99:1435-40.
Bradley TD, Logan AG, Kimoff RJ, et al. Continuous positive
airway pressure for central sleep apnea and heart failure. N Engl
J Med 2005;353:2025-33.
Eckert DJ, Jordan AS, Merchia P, Malhotra A. Central sleep apnea: Pathophysiology and treatment. Chest 2007; 131:595-607.
Javaheri S. Acetazolamide improves central sleep apnea in heart
failure: a double-blind, prospective study. Am J Respir Crit Care
Med 2006;173:234-7.
Javaheri S. Effects of continuous positive airway pressure on
sleep apnea and ventricular irritability in patients with heart failure. Circulation 2000;101:392-397.
Malhotra A, Muse VV, Mark EJ. Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case
12-2003. An 82-year-old man with dyspnea and pulmonary abnormalities. New England Journal of Medicine 2003;348:1574-85.