Author + information
- Received August 12, 2015
- Revision received September 15, 2015
- Accepted September 21, 2015
- Published online February 1, 2016.
- Michael Arzt, MDa,∗ (, )
- Holger Woehrle, MDb,c,
- Olaf Oldenburg, MDd,
- Andrea Graml, Dipl Statc,
- Anna Suling, PhDe,
- Erland Erdmann, MDf,
- Helmut Teschler, MDg,
- Karl Wegscheider, PhDe,
- SchlaHF Investigators
- aDepartment of Internal Medicine II, University Hospital Regensburg, Regensburg, Germany
- bSleep and Ventilation Center Blaubeuren, Respiratory Center Ulm, Ulm, Germany
- cResMed Science Center, Martinsried, Germany
- dDepartment of Cardiology, Heart and Diabetes Center North Rhine-Westphalia, Bad Oeynhausen, Germany
- eDepartment of Medical Biometry and Epidemiology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
- fClinic III for Internal Medicine, Heart Center University Hospital Cologne, Cologne, Germany
- gDepartment of Pneumology, Ruhrlandklinik, West German Lung Center, University Hospital Essen, University Duisburg-Essen, Essen, Germany
- ↵∗Reprint requests and correspondence:
Dr. Michael Arzt, Schlafmedizinisches Zentrum, Klinik und Poliklinik für Innere Medizin II, Universitätsklinikum Regensburg, Franz-Josef-Strauss-Allee 11, 93053 Regensburg, Germany.
Objectives This prospective study investigated the prevalence of sleep-disordered breathing (SDB) and its predictors in patients with stable chronic heart failure (HF).
Background SDB is increasingly recognized as being important in patients with HF.
Methods The multicenter SchlaHF (Sleep-Disordered Breathing in Heart Failure) registry provides demographic and clinical data on chronic, stable, symptomatic patients with HF (New York Heart Association functional class ≥II; left ventricular rejection fraction ≤45%). Moderate-to-severe SDB (apnea-hypopnea index ≥15/h) was determined by a 2-channel screening device (ApneaLink, ResMed, Sydney, Australia).
Results Data from 6,876 patients were analyzed. The prevalence of moderate-to-severe SDB was 46%, with a significant sex difference: 36% in women (n = 1,448) versus 49% in men (n = 5,428). Prevalence of SDB rose with increasing age (31%, 39%, 45%, 52%, and 59% in those age ≤50, >50 to 60, >60 to 70, >70 to 80, and >80 years, respectively). Risk factors for SDB were body mass index (per 5 units; odds ratio [OR]: 1.29; 95% confidence interval [CI]: 1.22 to 1.36), left ventricular rejection fraction (per 5% decrement from 45%; OR: 1.10; 95% CI: 1.06 to 1.14), age (per 10-year difference to 60 years; OR: 1.41; 95% CI: 1.34 to 1.49), atrial fibrillation (OR: 1.19; 95% CI: 1.06 to 1.34), and male sex (OR: 1.90; 95% CI: 1.67 to 2.17).
Conclusions SchlaHF registry data demonstrate a high prevalence of SDB in a representative population of stable patients with chronic HF receiving contemporary medical management. Male sex, age, body mass index, and the severity of both symptoms and left ventricular dysfunction were clinical predictors for prevalent SDB. (Prevalence, Clinical Characteristics and Type of Sleep-disordered Breathing in Patients With Chronic, Symptomatic, Systolic Heart Failure; NCT01500759)
Chronic heart failure (HF) is a symptomatic, debilitating disease. Patients require frequent hospitalization, and the mortality rate is high (1,2). Guideline-driven pharmacological treatment and cardiac resynchronization therapy (3,4) have improved outcomes for many patients. However, a significant number of those with chronic HF still experience persistent symptoms, and most of them will die from cardiovascular causes, usually progressive HF (1,2). A complementary strategy for the management of HF might be to identify alternative and additional features that contribute to disease progression and impair prognosis, and then develop strategies to target, improve, and overcome these factors (5).
Sleep-disordered breathing (SDB) is 1 condition that is under investigation as a contributor to symptom burden (6), morbidity, and mortality in patients with chronic HF (5,7–9). The prevalence of comorbid SDB in patients with HF is high, at 47% to 76% (10–14). SDB is characterized by 2 different abnormal breathing patterns: obstructive sleep apnea (OSA) and central sleep apnea (CSA) with Cheyne-Stokes respiration (CSR). The prevalence of OSA was reported to be higher in patients with chronic HF (12% to 43%) than in the general population (15–20). The prevalence of CSA-CSR in patients with chronic HF ranges between 21% and 40% in different studies (10–14). Furthermore, both OSA and CSA-CSR have been shown to be markers of disease severity and predictors of increased mortality in chronic HF (8,21–25).
In general, the reported prevalence of SDB in chronic HF is based on either single-center or small-scale studies. To date only 1 multicenter study has been reported (12), and just 5 single-center studies enrolling ≥100 participants have been published (10–14). Furthermore, in all previous studies, women were a minority of participants, which does not allow robust sex-specific estimates of SDB prevalence (10–14). All samples included highly selected patients with chronic HF who were either hospitalized patients or outpatients from specialist HF clinics (10–14,26). This resulted in populations that were predominantly male and younger than the average patient with chronic HF (2,26–28).
The clinical diagnosis of SDB in patients with chronic HF is difficult because key symptoms and characteristics of patients with SDB and normal cardiac function, such as daytime sleepiness and obesity, are often not present (29,30). This means that the rate of sleep testing in patients with chronic HF is very low (31). Clinical predictors for SDB have not been studied in a large sample of stable patients with chronic HF who were not pre-selected according to potential SDB-related symptoms.
This study analyzed a large representative sample of patients with chronic HF from cardiology practices and hospital departments. The aim was to prospectively determine the prevalence of SDB and clinical predictors for the disease that are routinely assessed in the clinical work-up of patients with chronic HF.
In the German SchlaHF (Sleep-Disordered Breathing in Heart Failure) registry, assessment of SDB was performed in 256 centers using a 2-channel respiratory monitor (ApneaLink, ResMed, Sydney, Australia). The rationale and design of the SchlaHF registry have been described in detail elsewhere (32). All German sleep laboratories were invited to participate in the registry. However, most did not have the infrastructure necessary to run such a project and therefore built networks with cardiologists in their referral area. Participating cardiologists, who had been contacted by and were working with sleep laboratories, were asked to consecutively enroll all patients with HF who fulfilled the registry inclusion criteria described later.
Age, body mass index (BMI), sex, left ventricular ejection fraction (LVEF), New York Heart Association (NYHA) functional class, heart rhythm, nocturnal dyspnea symptoms, nocturia, apnea-hypopnea index (AHI), HF etiology, and medication were documented for 6,876 stable patients with chronic HF between February 2008 and January 2011 at 138 centers in Germany (91 cardiology practices and 47 hospital departments). Oxygen desaturation index, average oxygen saturation (mean Spo2), and minimum oxygen saturation (min Spo2) were documented in 97% of patients, respectively. The SchlaHF registry received central ethics approval from the Freiburger Ethikkomission for Germany. All aspects of the registry were conducted within the principles of Good Clinical Practice and in accordance with the Declaration of Helsinki. All participants provided written informed consent.
The inclusion criteria were as follows: chronic HF diagnosed and treated according to the European Society of Cardiology guidelines (4) ≥12 weeks before enrollment; moderate-to-severe left ventricular systolic dysfunction (LVEF ≤45% by an imaging method, such as echocardiography, radionuclide angiography, left ventriculography, or cardiac magnetic resonance imaging) documented <12 weeks before enrollment; NYHA functional class III or IV at the time of inclusion, or NYHA functional class II with ≥1 hospitalization for HF in the last 12 months; and patient able to fully understand the information about the SchlaHF registry and give written informed consent. Patients were excluded if they had any of the following: current use of positive airway pressure therapy; life expectancy <1 year for diseases unrelated to chronic HF-with reduced ejection fraction (REF); cardiac surgery, percutaneous coronary intervention, myocardial infarction, or unstable angina within 6 months before randomization; cardiac resynchronization therapy implantation scheduled or performed within 6 months before randomization; transient ischemic attack or stroke within 3 months before enrollment; primary hemodynamically significant uncorrected valvular heart disease (obstructive or regurgitant) or any valvular disease expected to require surgery; and acute myocarditis/pericarditis within 6 months before enrollment.
Assessment of SDB
Nasal flow and pulse oximetry were measured using the ApneaLink device (ResMed) that has been validated in several studies for screening of SDB (33–38). Comparing ApneaLink (using automatic scoring) with the gold standard polysomnography (PSG) in patients without known heart disease, studies have reported a sensitivity of 73% to 94% and a specificity of 85% to 95% using an AHI cutoff value of 15/h.
In a subset of patients with chronic HF from the SchlaHF-Registry (n = 65), ApneaLink (using automatic scoring) was performed simultaneously with PSG as described in detail in the Online Appendix. For detecting SDB (AHI ≥15/h) assessed by PSG (manual scoring), the ApneaLink (automatic scoring) had a sensitivity of 73%, a specificity of 87%, a positive predictive value of 86%, and a negative predictive value of 74%.
The default settings of the screening device were used for the definitions of apnea, hypopnea, and desaturation: apnea was defined as a ≥80% decrease in airflow for ≥10 s; hypopnea was defined as a decrease in airflow by ≥50% to 80% versus baseline for ≥10 s; desaturation was defined as a ≥4% decrease in oxygen saturation; and SDB was defined as AHI ≥15/h. Of the 138 cardiology centers, 8 used respiratory polygraphy and 7 used polygraphy or ApneaLink for screening of SDB. When the ApneaLink was used, automatic scoring of respiratory events was accepted.
Descriptive statistics were absolute and relative frequency or mean ± SD, whatever appropriate. Because prevalences varied considerably between centers (cardiology practices or departments) presumably reflecting differences in patient populations, all inferential statistics were calculated with centers as a random term. Random effects logistic regression was used to model the association between suspected causes of SDB and the risk of developing that disorder. Linear mixed models were applied for the severity analysis of SDB in patients with an AHI ≥15/h. The models included the same set of potential clinical predictors of SDB: age, BMI, sex, LVEF, NYHA functional class, atrial fibrillation, and ischemic origin. For the study of sex differences, interaction terms with sex were added to the model if selected by backward elimination based on likelihood ratio tests. Results are visualized by forest plots showing adjusted odds ratios (ORs) and by marginal means (here, adjusted SDB prevalence). Although age, BMI, LVEF, and EF were modelled as continuous linear predictors for the forest plots, they were classified to allow a graphical presentation of the prevalence.
A p value <0.05 was considered to be statistically significant. Statistical analysis was performed with SPSS version 22.0 (SPSS Inc., Chicago, Illinois) and STATA version 13.0 (STATA Corporation, College Station, Texas).
Data from 6,876 stable patients with chronic HF were included in this analysis. Patients with SDB were older, had a higher BMI, and were more likely to be male (Table 1). Patients with chronic HF with SDB had more severe HF (higher NYHA functional class and higher prevalence of atrial fibrillation). In addition, such symptoms as nocturnal dyspnea and nocturia (≥3 times/night) occurred significantly more often in individuals with SDB. The causes of chronic HF, such as ischemia, were similar in both groups, as was use of most HF medication. However, patients with SDB were significantly more likely to be using diuretic agents and significantly less likely to be taking an aldosterone antagonist compared with those without SDB.
Prevalence of SDB in patients with chronic HF
The prevalence of SDB was 36% in women, 49% in men, and 46% overall. The prevalence of SDB increased with age and was numerically higher in males in any age group (Table 2). In patients with SDB, mean AHI was higher in men than in women, both overall and in the different age groups (Table 3).
Clinical predictors for SDB
Because predictors of SDB act simultaneously, it is required to analyze them in multivariate models. Random effects logistic regression showed that in chronic HF, SDB was independently associated with male sex (OR: 1.90; 95% confidence interval [CI]: 1.67 to 2.17; p < 0.001), age (per 10 years; OR: 1.41; 95% CI: 1.34 to 1.49; p < 0.001), BMI (per 5-U increment; OR: 1.29; 95% CI: 1.22 to 1.36; p < 0.001), LVEF (per 5% decrease; OR: 1.10; 95% CI: 1.06 to 1.14; p < 0.001), and atrial fibrillation (OR: 1.19; 95% CI: 1.06 to 1.34; p = 0.004). Conversely, ischemic etiology (OR: 0.92; 95% CI: 0.83 to 1.03; p = 0.139) and NYHA functional class ≥III (OR: 1.12; 95% CI: 0.99 to 1.27; p = 0.069) were not significant (Figure 1A). Adjusted SDB prevalence rates by age and sex are shown in Figure 2. Figure 3 shows adjusted SDB prevalence based on BMI and LVEF.
Clinical predictors for AHI in patients with SDB
Multivariate analysis showed that higher AHI in patients with SDB was independently associated with male sex (regression coefficient, 4.90; 95% CI: 3.16 to 6.65; p < 0.001), age (per 10 years; OR: 0.78; 95% CI: 0.30 to 1.27; p = 0.002), BMI (per 5-unit increment; OR: 1.65; 95% CI: 1.18 to 2.13; p < 0.001), NYHA functional class ≥III versus class II (OR: 1.85; 95% CI: 0.72 to 2.99; p = 0.001), and atrial fibrillation (OR: –1.73; 95% CI: –2.77 to –0.68; p = 0.001). LVEF and ischemic etiology in men and women did not show a significant independent association with AHI (p = 0.368, p = 0.194, and p = 0.110, respectively) (Figure 1B).
This analysis of the prospective SchlaHF registry reports several robust findings. First, moderate-to-severe SDB is very common in a large representative population of patients with stable chronic HF receiving optimized medical treatment: 46% in a sample of 6,876 patients. Prevalence was dependent on both age and sex. The sex difference was constant over the total age range after accounting for other risk factors for SDB in HF. Second, male sex, older age, higher BMI (kg/m2), more severe impairment of systolic cardiac function (i.e., lower LVEF), higher NYHA functional class, and atrial fibrillation were clinical predictors for at least moderate SDB. Third, the extent of AHI in patients with SDB is greater in males, patients who are older, with higher BMI, and with higher NYHA functional class.
The observed high prevalence of SDB in the SchlaHF registry is in accordance with most of the 5 previous studies that included more than 100 patients with chronic HF. Discrepancies can be explained in most cases by differences in patient populations and methodology applied to measure SDB. For example, Sin et al. (13) reported an SDB prevalence of 61% in 450 consecutive patients with stable chronic HF (85% male, 15% female). The higher prevalence of SDB in this study can be explained by the fact that this study was not designed to determine the prevalence of SDB in patients with HF. Instead, the HF sample was pre-selected by referral to the sleep laboratory because of a suspected sleep disorder, and scoring rules of PSG in this study allowed classification of respiratory events as hypopneas in the absence of an arousal or oxygen desaturation (13). In another study, the prevalence of SDB of any severity (AHI ≥5/h) in a population of 218 patients with chronic HF (23% of who were female) was 47% (14). This value is very similar to that reported in the present study because although the population was younger (age 56 years vs. 66 years), this was counterbalanced by worse left ventricular function (LVEF 25% vs. 34%). A high prevalence of SDB in HF was reported in other studies as well. The largest study to date prospectively enrolled 700 patients (20% female) (11). SDB was documented in 76% of patients (40% with CSA and 36% with OSA) (11). Looking at moderate-to-severe SDB (AHI ≥15/h), the prevalence was 52% (11). Patients with CSA or OSA were older than those without SDB and were more likely to be male. In addition, patients with CSA had more advanced HF symptoms and were more likely to have atrial fibrillation, and those with OSA had a higher BMI and tended to have a higher prevalence of diabetes (11). The overall prevalence of SDB in another study of 100 male patients with chronic HF was also approximately 50% (10).
Obviously, the use of different AHI thresholds to define SDB is associated with variation in prevalence rates. For instance, Schulz et al. (12) performed a multicenter study using an AHI ≥10/h as the SDB definition and found a prevalence of 71% in 203 patients with chronic HF (25% female). Age, BMI, and AHI in this study were similar to those in the current analysis, and therefore the different definition of SDB (AHI ≥10/h vs. ≥15/h) was probably a big contributor to the different prevalence rates reported, although a lower LVEF in the Javaheri (10) study population could also have played a role.
In summary, the prevalence rate observed in this large population analysis was lower than that in previous studies that included younger patients with a greater degree of left ventricular dysfunction (10–14). The older age and better cardiac function in this trial can probably be explained by the fact that patients were identified through cardiology practices, whereas those in previous studies were mostly recruited or referred from specialized HF clinics in hospitals.
In the Sleep Heart Health Study, Young et al. (20) reported the average age-dependent SDB prevalence in community-dwelling adults age 39 to 99 years. The results showed SDB prevalence, defined as AHI ≥15/h, increased between 40 and 60 years of age from 6% to 20%, reaching a plateau between age 60 and 70 years. In contrast, we could not find such a plateau effect in our population of patients with chronic HF. Instead there was a continuous increase in SDB prevalence, from 31% in the subgroup age ≤50 years to 57% in the oldest patients (age >80 years).
Evaluation of which clinical characteristics assessed as part of the routine work-up of patients with chronic HF might help clinicians to predict those likely to have SDB would be helpful in the management of the condition, enabling risk stratification and reducing symptom burden and morbidity in selected patients. The role of treatment for OSA and/or CSA in patients with HF remains to be determined (39). In this context, the recently published results from the SERVE-HF (Treatment of Predominant Central Sleep Apnoea by Adaptive Servo Ventilation in Patients With Heart Failure) trial (NCT00733343) provide new and important information (40).
SERVE-HF was a multinational, multicenter, randomized controlled trial designed to assess whether treatment of moderate-to-severe predominant CSA with adaptive servoventilation, in addition to guideline-based medical therapy, reduced morbidity and mortality in patients with chronic stable HF and reduced ejection fraction (41). The primary endpoint was a composite of all-cause mortality and hospitalization for worsening HF, and there was no statistically significant difference between patients in the adaptive servoventilation group compared with those receiving guideline-based medical therapy alone (control subjects) (hazard ratio: 1.13; 95% CI: 0.97 to 1.31; p = 0.10) (40). However, additional analysis showed that patients in the adaptive servoventilation group had an increased risk of both all-cause mortality (hazard ratio: 1.28; 95% CI: 1.06 to 1.55; p = 0.01) and cardiovascular mortality (hazard ratio: 1.34; 95% CI: 1.09 to 1.65; p = 0.006) versus control subjects (40). Potential mechanisms that might explain these unexpected results remain unclear (42).
Male sex, age, obesity, severe impairment of systolic cardiac function, NYHA functional classes III and IV, and atrial fibrillation were independently associated with SDB or increased AHI in this study. Perhaps 1 of the most interesting results was that there were no sex differences in the significant predictors of SDB and AHI, despite analyses designed to detect such differences. This suggests that although the prevalence of SDB differs in men and women, the underlying disease mechanisms are the same.
So far, only 3 other studies have reported independent predictors of SDB obtained using multivariate analysis (Table 4) (13,14,26). Male sex, atrial fibrillation, low awake carbon dioxide pressure, age in general, and age >60 years were independently associated with CSA-CSR (13,14). In OSA, both Sin et al. (13) and Yumino et al. (14) demonstrated that BMI was a predictor only in men. However, these 2 studies reported different observations on the role of age; Sin et al. (13) found that age was an independent predictor of OSA in women only, whereas Yumino et al. (14) reported this association for both sexes. Furthermore, Yumino et al. (14) showed that atrial fibrillation was independently associated with OSA. MacDonald et al. (26) identified atrial fibrillation and NYHA functional class III and IV as risk factors for SDB, independent of sex, age ≥60 years, BMI, and LVEF. Other studies, which did not perform multivariate analyses, reported worse functional classification and LVEF <20% as markers for CSA (10,11).
This study has 3 major strengths. First, because of the large sample size, the prospective recruitment of patients by both cardiology practices and hospital cardiology departments, and the use of a portable screening device for SDB instead of PSG, the results are more representative and applicable to contemporary chronic HF population than previous studies (13,14,26). This study cohort is similar to the community chronic HF population described by Redfield et al. (28) in terms of age (63 years vs. 66 years), BMI (28.4 kg/m2 vs. 28.4 kg/m2), and proportion of women (26% vs. 21%). Assessment of SDB with PSG as it was done by Sin et al. (13), Yumino et al. (14), and other investigators is the gold standard for diagnosing SDB. However, performing a time-consuming in-laboratory/in-hospital sleep study confers a major selection bias itself. For example, in the “HypnoLaus study” (43), 875 of the 3,043 patients (29%) who were invited for PSG refused to undergo this diagnostic test. Therefore, the prevalence and risk factor estimates in the SchlaHF study are likely to be the most representative reported to date. Second, because this was the first large-scale study in a representative study cohort, it was possible to present age- and sex-dependent prevalence and risk factor estimates; this is in contrast to previous studies (13,14,26). Third, we could confirm predictors, such as male sex, age, obesity, and atrial fibrillation as reported by previous studies (13,14) in a larger patient population. Furthermore, this is the first study that presents an independent association between SDB and both severe impairment of systolic cardiac function and higher NYHA functional class.
Our study does also have some limitations. We studied SDB prevalence by using the screening device ApneaLink instead of gold standard PSG. PSG was not performed in a representative sample of patients with HF and therefore it was not possible to differentiate between OSA and CSA with respect to prevalence or risk factors. The device was evaluated in several studies versus PSG (33–38). These studies reported an excellent correlation between results obtained by PSG and ApneaLink. Hence, ApneaLink seems to be an accurate screening tool for SDB in a population with high prevalence of the disorder. Although Ragette et al. (37) showed that ApneaLink might slightly overestimate AHI, the results of the validation study in a subset of patients from the SchlaHF Registry (Online Appendix) suggest that ApneaLink might underestimate AHI. This is plausible because the number of apneas and hypopneas are reported in relation to “total recording time” with ApneaLink compared with “total sleep time” with PSG. Thus, the use of ApneaLink is likely to result in conservative prevalence estimates. Because we reported prevalence for SDB overall, and not for the subgroups of OSA and CSA, we cannot exclude the possibility that the risk factors identified relate only to 1 of these conditions, rather than both.
Growing evidence suggests that SDB may play an important role in patients with HF and preserved ejection fraction (5). The SchlaHF registry only includes patients with HF and reduced ejection fraction because at the time the registry was designed, the relationship between SDB and HF with preserved ejection fraction was not well known.
The SchlaHF registry data demonstrate a high prevalence of SDB in a representative population of stable patients with chronic HF with contemporary medical management. Male sex, age, BMI, and the severity of both symptoms and left ventricular dysfunction were clinical predictors for the presence of SDB.
COMPETENCY IN MEDICAL KNOWLEDGE: The pretest probability of a patient with chronic HF and reduced ejection fraction being treated with optimized medical treatment having clinically relevant SDB with ≥15 apneas and hypopneas per hour is approximately 50%. Data from the SchlaHF registry, and the fact that patients with chronic HF with SDB often do not show typical SDB symptoms, suggest that the presence of 1 or more predictors of SDB, such as male sex, older age, obesity, ejection fraction <25%, or NYHA functional class III/IV, should prompt clinicians to perform device-based screening for SDB.
TRANSLATIONAL OUTLOOK: In view of the various phenotypes of SDB and their potential differential clinical implications in chronic HF, the development of easy to use screening devices for SDB with the ability to discriminate between OSA and CSA with and without periodic breathing is warranted. This could be useful for risk stratification in patients with chronic HF.
English language editing assistance was provided by Nicola Ryan, independent medical writer. This assistance was funded by ResMed.
For supplemental information, please see the online version of this article.
The SchlaHF registry was funded by ResMed Ltd. (Sydney, Australia) and ResMed Germany Inc. (Martinsried, Germany). ResMed also provided funding for English language editing assistance by an independent medical writer (Nicola Ryan). Dr. Arzt received grant support from ResMed, Philips Home Healthcare Solutions, and the German Foundation for Cardiac Research (Deutsche Stiftung für Herzforschung); is the holder of an endowed professorship from the Free State of Bavaria at the University of Regensburg that was donated by ResMed and Philips Home Healthcare Solutions; and has previously received lecture fees from AstraZeneca, Philips Home Healthcare Solutions, and ResMed. Drs. Woehrle and Graml are employees of ResMed. Dr. Oldenburg has acted as a consultant for ResMed and Respicardia; has received a research grant from ResMed; and has received lecture honoraria from ResMed, Respicardia, and Weinmann. Dr. Erdmann received consulting fees or honoraria and travel grants from ResMed. Dr. Teschler received grant support from ResMed, the ResMed Foundation, and Linde; and has previously received lecture fees from AstraZeneca, Novartis, Linde, Boehringer Ingelheim, Berlin Chemie, and ResMed. Dr. Wegscheider received consulting fees or honoraria and travel grants from ResMed. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Arzt and Woehrle contributed equally to this work.
- Abbreviations and Acronyms
- apnea-hypopnea index
- body mass index
- confidence interval
- central sleep apnea
- Cheyne-Stokes respiration
- heart failure
- left ventricular rejection fraction
- New York Heart Association
- odds ratio
- obstructive sleep apnea
- sleep-disordered breathing
- Received August 12, 2015.
- Revision received September 15, 2015.
- Accepted September 21, 2015.
- 2016 American College of Cardiology Foundation
- Cubbon R.M.,
- Gale C.P.,
- Kearney L.C.,
- et al.
- Hunt S.A.,
- Abraham W.T.,
- Chin M.H.,
- et al.
- Swedberg K.,
- Cleland J.,
- Dargie H.,
- et al.
- Mentz R.J.,
- Kelly J.P.,
- von Lueder T.G.,
- et al.
- Costanzo M.R.,
- Khayat R.,
- Ponikowski P.,
- et al.
- Wang H.,
- Parker J.D.,
- Newton G.E.,
- et al.
- Oldenburg O.,
- Bitter T.,
- Fox H.,
- Horstkotte D.
- Schulz R.,
- Blau A.,
- Borgel J.,
- et al.
- Javaheri S.,
- Shukla R.,
- Zeigler H.,
- Wexler L.
- Lanfranchi P.A.,
- Braghiroli A.,
- Bosimini E.,
- et al.
- Solin P.,
- Bergin P.,
- Richardson M.,
- Kaye D.M.,
- Walters E.H.,
- Naughton M.T.
- Arzt M.,
- Harth M.,
- Luchner A.,
- et al.
- Hastings P.C.,
- Vazir A.,
- O'Driscoll D.M.,
- Morrell M.J.,
- Simonds A.K.