|Home | About | Journals | Submit | Contact Us | Français|
This study characterized sleep in heart failure (HF) and determined associations with quality of life. Forty stable HF patients and 34 healthy volunteers were studied in a clinical research unit. HF patients had more central apneas/hour [17.6 vs. 5.4 (p≤0.01)] and obstructive apneas/hour [21.7 vs. 8.5 (p≤0.05)], spent more time in stage 1 sleep [54 min. vs. 35 min. (p≤.05)] and had more respiratory awakenings following apneic events [27.2 vs. 4.2 (p≤.01)]. More HF patients were depressed (55% vs. 27.2%, p≤0.01) and had worse fatigue (p≤0.05). In multiple regression analysis, physical functioning quality of life was predicted by reduced LVEF (p≤0.05), shorter distance on a six-minute walk test (p≤.05), greater fatigue (p≤0.01), and more apneas (p’s≤0.05) (model R2 =.672, p≤0.001). Emotional functioning quality of life was predicted by greater fatigue (p≤0.01) (model adjusted R2 =.732, p<0.001). Findings provide evidence that in addition to functional status and ongoing fatigue, poorer quality of life in HF is independently related to the severity of sleep-disordered breathing.
Quality of life is often poor in HF, being attributable to worse functional status and symptom burden 1–3, but also to depression and fatigue, which are common in HF 4, 5. In addition, female gender and ethnicity are associated with poorer quality of life in HF 1–3.
Sleep disordered breathing (SDB) is common in individuals with HF 6. Reported incidence rates of obstructive sleep apnea (OSA) and central sleep apnea (CSA) vary depending on NYHA class, with rates typically increasing with greater severity. While OSA is considered a risk factor for HF 7, HF is the most common cause of CSA 8, 9. An apnea hypopnea index (AHI) cutpoint of ≥ 10 events/hour is a commonly accepted and clinically relevant definition of sleep apnea 10–12. In mild to moderate HF, AHI scores of ≥ 10 events/hour yield OSA prevalence rates of 15% – 53%; and CSA prevalence rates of 15% – 57% 13, 14, depending upon sample age. SDB in HF is associated with increased risk for death and hospitalization 15, 16.
In non-HF populations, SDB is associated with poor quality of life. For example, individuals with obstructive sleep apnea show overall poorer quality of life than age and gender matched populations 17. The poor quality of life is independently associated with fatigue and with severity of SDB - specifically with the degree of respiratory disturbance and number of arousals from sleep 18, 19.
Studies have begun to assess the contribution of poor sleep to quality of life in HF. Brostrom et al. showed that self-reported difficulties with sleep, including difficulties initiating and/or falling asleep, are associated with worse quality of life in HF 20. Other studies suggest that HF patients with SDB have poorer quality of life as compared to HF patients without SDB 21, 22, although such studies have not examined possible confounding factors such as fatigue and depression, both of which also influence quality of life and are tightly linked to sleep quality 1–4. It could be, for example, that the negative influence of SDB on quality of life in HF is an indirect consequence of the fatigue associated with SDB and not with SDB per se 23.
The purpose of this study therefore was to characterize sleep and SDB in patients with HF and determine potential independent contributions of fatigue and depression, as well as functional capacity, to quality of life.
Stable HF patients with left ventricular ejection fraction (LVEF) ≤ 45% were recruited from the UCSD Heart Failure Program and the VA Medical Center Coronary Care Program. Of 74 HF patients eligible to participate in the study, 40 men and women (33 NYHA class II and 7 NYHA class III) between the ages of 31 and 80 years agreed to participate and completed the study. Healthy non-HF volunteers were recruited from the local community via advertisement and referrals. Of 61 non-HF volunteers eligible to participate in the study, 34 men and women between the ages of 35 and 81 years agreed to participate and completed the study. Individuals who did not agree to participate cited lack of interest or time as reasons.
Inclusion criteria specific to HF included NYHA classes II through IV, symptoms of HF for at least 3 months which have been optimally treated with diuretics, angiotensin converting enzyme (ACE) inhibitors, and β-blockers where appropriate (approximately 95% of the patients were being treated with β-blockers), and an ejection fraction ≤ 45%. Other study inclusion criteria included ages between 30 – 85 years, blood pressure < 180/110 mm Hg, and men and women of all ethnicities and races. Exclusion criteria included recent myocardial infarction (1 month), recent stroke or significant cerebral neurological impairment, severe COPD, and psychiatric illness other than depression. Inclusion criteria specific to the non-HF controls included being in the same age range as the HF patients and to be otherwise physically healthy. The control participants were recruited specifically to participate in this study on sleep in HF, and did not have echocardiography measurement or the 6-minute walk test (as described below for the HF patients).
The investigation conformed to the principles outlined in the Declaration of Helsinki. The study was approved by the University of California, San Diego Institutional Review Board. All subjects gave informed written consent.
All HF patients completed the 21 item Minnesota Living with Heart Failure Questionnaire (MLHFQ). The MLHFQ was designed to assess perceived quality of life due to the effects of HF and its treatment. The psychometrically sound instrument questions signs and symptoms of HF, work and emotions, social relationships, and physical and sexual activity 24. All subjects completed the 30-item Multidimensional Fatigue Symptom Inventory - Short Form (MFSI-sf) 25 and the 21-item Beck Depression Inventory (BDI) 26. The MFSI-sf assesses the principal manifestations of fatigue, yielding a total fatigue score. The 21-item BDI assesses symptoms related to sadness, feelings of guilt, perceptions of self-worth, suicidal ideation, and changes in appetite and body weight, among other characteristics 26. Scores ≥ 10 indicate possible depression.
Echocardiography and 6-minute walk studies were performed as previously described 27. The six-minute walk test has been shown to be independently related to self-reported ratings of physical capacity, including MLWHFQ physical functioning 28, 29.
Participants were scheduled for one night of sleep monitoring at the Gillin Laboratory of Sleep and Chronobiology at the UCSD General Clinic Research Center. Participants arrived at 6:00 pm. Sleep set-up began at 8:00 pm and took approximately one hour. Lights were turned off at 10:00 pm.
Sleep data was recorded using standard polysomnography 30. Electroencephalography (EEG), electrooculography (EOG), chin electromyography (EMG), thoracic and abdominal respiration were recorded on a Grass model PSG36-2 (Grass Technologies; West Warick, RI) or Embla model A10 polysomnograph (Broomfield, CO). Oxyhemoglobin saturation (SpO2) was monitored using a pulse oximeter (Biox 3740, Ohmeda, Louisville, CO) and analyzed by software from Profox (Escondido, CA). Anterior tibialis EMG’s were used to rule out periodic limb movements during sleep. Records were scored with the Rechtshaffen and Kales 31 criteria by technicians with inter-rater reliabilities above 90%.
Based on the existing literature 9, 32, we choose a panel of nine sleep parameters to examine: time spent in sleep stage 1, stage 2, slow wave sleep (SWS) (stage 3 and stage 4 combined), and REM sleep as descriptors of sleep architecture. Percent sleep efficiency and sleep onset latency were chosen as objective indices of sleep efficiency. (Percent sleep efficiency was computed as the ratio of total sleep time to time spent in bed multiplied by 100.) Total number of arousals, number of respiratory awakenings, and wake after sleep onset were chosen to represent sleep fragmentation. An arousal was defined as a shift in EEG frequency to alpha or theta for ≥3 seconds but <15 seconds duration as scored from central, occipital or both EEG leads.
The total number of apneas and hypopneas per hour of sleep were determined, as well as the total number of central and obstructive apneas per hour. An apnea was defined as a decrement in airflow ≥90% from baseline for ≥10 seconds. A hypopnea was defined as a decrement in airflow ≥50% but <90% from baseline for ≥10 seconds. AHI was defined as the total number of apneas plus hypopneas per hour of sleep. Central AHI (C-AHI) was defined as the total number of central apneas plus hypopneas per hour of sleep. Obstructive AHI (O-AHI) was defined as the total number of obstructive apneas plus hypopneas per hour of sleep.
Data were analyzed using one-way ANOVA, multivariate analysis of variance (MANOVA), Pearson correlations, and multiple linear regression (SPSS 15.0 for Windows (SPSS Inc., Chicago, IL). Data are presented as means ± SD. The level of statistical significance was set at p ≤ .05 (two-tailed). Borderline significance (p ≤ .10) is also presented to show trends that may help to better illustrate effects and relationships.
Demographic characteristics are presented in Table 1. HF and non-HF subjects were similar in age, BMI, race, and incidence of hypertension. There were fewer women in the HF group than the non-HF group (Χ2 = 8.75, p=0.003).
MFSI-sf scores (F= 4.23, p=0.043) and BDI scores (F= 6.62, p=0.012) were higher in HF than non-HF. Fifty-five percent of HF patients were depressed (i.e., BDI scores ≥ 10) as compared to 27.2% of controls (Χ2 = 5.69, p=0.015). HF patients’ MLHFQ physical functioning scores were 20.35 (SD=11.2, range 0 to 40) and emotional functioning scores were 8.85 (SD=7.1, range 0 to 25). These values are comparable to those found in the literature for our NYHA class patients 20 (Table 1).
Sleep characteristics are presented in Table 2. A MANOVA controlling for gender was significant for HF group (F = 2.42; Wilk's Lambda p value = .015). Regarding sleep architecture, HF patients spent more time in stage 1 sleep (F=4.80; p = 0.032) and marginally less time in REM sleep (F=3.71; p=0.058). Regarding sleep fragmentation, HF patients had a greater number of respiratory awakenings (F=6.22; p=0.015). HF patients tended towards a greater number of total arousals (F=3.01; p=0.084) and showed no difference in how much time was spent awake after sleep onset. There were no group differences in sleep efficiency. HF patients had a lower average nocturnal O2 saturation (F=7.19; p=0.009) and spent more time below 90% saturation (F=5.97; p=0.017).
HF patients had a greater total number of apneas/hour as compared to non-HF subjects (AHI = 78.0 (SD=62) versus 65.4 (SD=48) (F= 6.51; p=0.021). HF patients had a greater number of central (F=8.95; p=0.004) and obstructive apneas/hour (F=5.42; p=0.022) (Table 2; Figure 1). The presence of sleep apnea was defined as AHI, C-AHI and O-AHI ≥ 10 14. According to this definition, compared to non-HF, more HF patients were classified as apneic: AHI: 23.4% versus 72.5% (Χ2 = 17.6, p<0.001); C-AHI: 11.7% versus 57.5% (Χ2 = 16.6, p<0.001); O-AHI: 20.5% versus 50.0% (Χ2 = 6.8, p<0.01).
Separate multiple linear regression models were constructed for the dependent variables MLHFQ physical and emotional functioning subscales. Independent variables BMI, gender, and race were entered into the first block of variables. LVEF and the 6-minute walk test were entered in the second block. BDI scores were entered into the third block. MFSI-sf scores were entered into the fourth block, and the sleep variables from the MANOVA were entered into the fifth block of variables.
MLHFQ physical functioning was predicted (model adjusted R2 =.610, F=8.74, p<0.001) by a combination of variables including reduced LVEF (p=0.005), shorter distance on the six-minute walk test (p=0.055), higher MFSI-sf scores (p=0.031), number of obstructive apneas (p=0.027) and number of central apneas (p=0.038) (Table 3). MLHFQ emotional functioning was predicted (model adjusted R2 =.732, F=13.7, p<0.001) by higher MFSI-sf scores (p<0.001) and marginally by the number of central apneas (p=0.070) (Table 4).
Understanding factors that determine quality of life in HF is vital because in addition to the importance of day to day satisfaction with life, quality of life predicts morbidity and mortality in HF 33. Much is already known about determinants of quality of life in HF. HF severity, as determined by functional status and symptom burden 1–3 influences quality of life. Depression and fatigue too are predictors of poorer quality of life in HF 4, 5, as are female gender and ethnicity (poorer in African-Americans and Caucasians as compared to Hispanics) 1–3. Increased incidence of both depression and fatigue are linked to severity of HF, and negatively affect quality of life 5, 34.
The purpose of this study was to determine possible independent contributions of SDB to quality of life in individuals with HF. As a first step, compared to individuals without HF, we sought to confirm prior studies showing that sleep is poor in HF, particularly showing high rates of central and obstructed sleep apnea 6. Studies indicate that the severity of SBD in HF reflects the severity of HF, as indicated by pulmonary artery pressure and pulmonary capillary wedge pressure 35. Also consistent with the literature, we observed fairly high rates of sleep apnea in our otherwise healthy group of comparison subjects 36. Regarding quality of life in HF, our findings extend the few existing studies in this area by showing that SDB is independently associated with poorer quality of life in HF. We found that independent of demographic factors, functional status, fatigue and depression, the severity of apnea is associated with poorer quality of life in terms of physical functioning, but less so with emotional functioning.
Although there are no currently endorsed strategies for the treatment of SDB in HF 37, 38, there are several different treatments approaches that are promising. Studies examining the effects of continuous positive airway pressure (CPAP) for OSA are generally positive. For example, studies examining 3 months of CPAP for OSA report normalized AHI and improved LVEF in NYHA classes II or greater 14, 39. Studies using CPAP or auto-titrating CPAP for OSA and/or CSA report improvements in apnea severity, as well as several domains of quality of life including physical functioning, emotional well being, and fatigue 39–41. In addition to improving SDB and quality of life, CPAP might also significantly reduce risk of death and hospitalization among patients with HF 15. Since CSA appears to be a consequence of HF, some standard therapies for HF such as β-blockers and diuretics may attenuate CSA severity 38.
In addition to CPAP, physical training can improve SDB and associated indices of quality of life. Six months of exercise training in HF patients with both systolic dysfunction (left ventricular ejection fraction <45%) and SDB (AHI >10) led to a significant reduction in AHI (from 24.9 to 8.8), as well as a reduction in the number of CSA events/night (from 152 to 50) 42. Other studies show that exercise training improves quality of life in HF 43. Twelve weeks of nocturnal oxygen therapy for SDB leads to an improvement in AHI (from 21.0 to 10.0) in NYHA classes II and III HF patients, and an associated improvement in quality of life 44.
We found that the association of quality of life to SDB was independent of fatigue or depression. Fatigue is common in OSA, and can be associated with severity of apnea 45. Depression too is common in HF and has been found to be predictive of quality of life independent of LVEF 5. While the purpose of this study was not to identify predictors of depression in HF, other studies report that severity of depression among patients with HF is associated with fatigue and excessive daytime sleepiness, independent of physical functioning. In our study, depression was initially a predictor of physical functioning related quality of life, but then fell out of the regression model once fatigue was entered. Fatigue remained a significant predictor above and beyond the independent contributions of SDB to physical functioning. In contrast to physical functioning, only fatigue was found to be a significantly associated with emotional functioning. The ongoing fatigue resulting from repeated sleep disruption in individuals with OSA may be a mediator of poor quality of life observed in OSA 46. In pursuit of this possibility in HF, we ran mediator analyses 47 and found that fatigue was not a mediator of the relationship between SDB and poorer physical functioning in our sample of HF patients. If not fatigue and depression, what might be mechanisms linking SDB to poorer quality of life in HF? More severe sleep apnea, likely as a result of its associated daytime hypoxemia, is associated with neurocognitive deficits 11, 48. Perhaps the more severe SDB seen in HF adversely affects quality of life through its effects on cognition, although little attention has been paid to this area in HF 49, which merits further study.
There are several limitations of this study that should be addressed. We had relatively modest sample sizes for each study group. In addition, our comparison group was not matched for gender. Finally, the vast majority of our HF patients had more ‘mild’ HF (i.e., NYHA II), and thus the findings might not generalize to more severe HF.
In summary, independent of HF severity, and independent of associated fatigue and depression, SDB adversely affects health-related quality of life in patients with HF, particularly physical functioning quality of life. In addition to helping delineate associations between sleep and quality of life, the findings provide more support for the sometimes controversial position of more aggressively identifying and treating SDB in HF patients.
This work was supported by grants HL-073355 and HL-57265 from the National Institutes of Health and the UCSD General Clinical Research Center (MO1RR-00827)