© 2004 American Thoracic Society
Obstructive Sleep Apnea and Heart FailureTwo Unhappy BedfellowsAdelaide Institute for Sleep Health Repatriation General Hospital Daw Park, South Australia, Australia Congestive heart failure is a major cause of morbidity and mortality and leads to one fifth of hospital admissions in those over 65 years (1). While the average age at diagnosis increased substantially between the 1950s and 1990s, and the death rate after diagnosis declined by approximately 30%, one-year mortality for all heart failure cases remains high at 25–30% (2). For symptomatic heart failure, one-year mortality is close to 45% (1). ACE-inhibitors, ß-blockers, and spironolactone have been shown to improve cardiac function, quality of life, and longevity without making major inroads into overall mortality (1). Cross-sectional data from the Sleep Heart Health Study showed an adjusted odds ratio of 2.2 for self-reported heart failure amongst subjects with obstructive sleep apnea (the upper quartile of apnea–hypopnea index greater than 11 as compared with lower quartile less than 1.3) (3). Two clinical studies in the pre ß-blocker treatment era found the prevalence of sleep-disordered breathing (apnea–hypopnea index greater than 15 per hour) in severe systolic heart failure to be 50–60% (4, 5); 11 to 37% of all patients had obstructive apnea–hypopnea. Others with sleep disordered breathing had mainly central apneas or Cheyne-Stokes respiration. These studies demonstrated associations, not causality. Sleep is normally a period of cardiac rest. Apart from brief bursts of sympathetic activity in REM sleep, sleep is characterized by decreased sympathetic activity and increased vagal activity, which lowers heart rate and blood pressure. This is not so for patients with obstructive sleep apnea (6). Large negative swings in intrathoracic pressure increase left ventricular afterload. Reflex sympathetic activation secondary to hypoxia and hypercapnia, and arousal from sleep, cause acute surges in blood pressure and heart rate, further increasing left-ventricular afterload and wall stress. Increased venous return, plus acute hypoxic pulmonary vasoconstriction, increase right-ventricular volume and pressure, which may compromise left-ventricular filling. Vascular pressor responses to hypoxia are increased in patients with obstructive sleep apnea (7, 8). Also, patients are more likely to have elevated daytime sympathetic activity and systemic blood pressure (9), which may further perpetuate heart failure. Thus, regardless of whether obstructive sleep apnea is sufficient to cause heart failure, there is ample reason to believe that it could adversely affect left ventricular function in those with an already failing heart. Two recent clinical trials provide strong support for the latter hypothesis. In the first study (10), 24 patients with severe systolic heart failure and co-existing obstructive sleep apnea (apnea–hypopnea index, 42 per hour) were randomly assigned to one month of CPAP treatment or usual medical care. CPAP-treated patients showed a 9% improvement in left-ventricular ejection fraction and reduced daytime systolic blood pressure and heart rate compared with control patients who demonstrated no changes. Symptoms and quality-of-life measures were not reported. In this issue of the Journal (pp. 361–366), Mansfield and colleagues (11) report results of a similar study, which confirm and extend these findings. Fifty-five patients with mild to moderate obstructive sleep apnea (mean apnea–hypopnea index, 28 per hour) and moderately impaired left-ventricular ejection fraction (35.5%) were randomized to 3 months of CPAP versus usual care. Among the 40 patients that completed the study, CPAP treatment was associated with a 5% improvement in left-ventricular ejection fraction, significantly greater than in the control group. CPAP reduced overnight urinary norepinephrine levels, and improved general and disease-specific quality of life but not dyspnea or exercise capacity. Awake blood pressure and heart rate were unchanged. What are the implications of these findings for clinical practice? Given that changes in systolic function of this magnitude have been a good marker of improved morbidity and mortality in most pharmacological trials, should all heart-failure patients now be screened for obstructive sleep apnea, and CPAP treatment aggressively pursued? I believe some degree of circumspection, if not caution, is warranted. Mansfield and coworkers (11) showed symptomatic improvement with CPAP, but improvements were in factors such as mood, sleepiness, and fatigue, and not exercise capacity or dyspnea. This suggests a symptom response to lessened sleep apnea rather than one resulting from improved cardiac function. While improved vigilance and vitality is important, the cardinal symptom of heart failure is dyspnea, and patients and their cardiologists will ultimately be interested to know how CPAP will influence this symptom. Impaired diastolic filling (abnormal ventricular relaxation) and pulmonary venous congestion is a significant cause of dyspnea in this population (12). Neither study provided information on diastolic function. Beta blocker trials in heart failure have shown that systolic function improves ahead of reverse ventricular remodeling and improved exercise tolerance (1). It is possible that CPAP treatment for longer than 3 months will produce similar changes, but this requires further study. A methodological weakness of both studies was the absence of a placebo control group. This is relatively unimportant for "hard" end points, such as left-ventricular ejection fraction, but is significant when considering neurobehavioral changes. Previous studies have shown similar magnitude responses in Epworth sleepiness scores and quality of life using a placebo tablet (13) and sham CPAP (14) to those reported by Mansfield and coworkers following active CPAP. CPAP compliance was relatively high (6.2 and 5.6 hours per night) (10, 11) and remarkable in the presence of minimal symptoms of sleep apnea in one of the studies (10). These short-term studies, however, were conducted among selected, highly motivated patients. The long-term adherence to CPAP may prove more problematic in this severely ill population, particularly if treatment does not translate into improvements in breathlessness. Nonetheless, these are exciting preliminary findings. Defining the place of obstructive sleep apnea and CPAP therapy in modern heart failure management will require the following: prevalence studies in present-day heart failure populations, including those with isolated diastolic (normal systolic function) heart failure; and long-term, randomized placebo-controlled studies using intention-to-treat analyses and inclusion of cardiac endpoints such as dyspnea, exercise capacity, hospital admissions, and, ideally, mortality. Randomized studies may be ethically difficult to justify among patients with severe symptomatic sleep apnea but presumably not among patients with mild to moderate symptoms. Because considerable overlap exists between obstructive and central apnea in heart-failure patients (15), studies would need to account for this and might employ positive-pressure devices that could potentially treat both disorders (16). FOOTNOTES Conflict of Interest Statement: R.D.M. is chief investigator of a multicenter research study, which has pending an untied research grant from ResMed consisting of the loan of equipment valued at Aus $40,000 and direct costs of Aus $50,000. Masimo has loaned equipment for the same study valued at Aus $60,000. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
