Daytime Sleepiness in Patients With Congestive Heart Failure and Cheyne-Stokes Respiration*

Transcription

1 Daytime Sleepiness in Patients With Congestive Heart Failure and Cheyne-Stokes Respiration* Patrick Hanly, MBBCh, FCCP; and Naheed Zuberi-Khokhar, MD, BSc Study objective: To determine whether patients with congestive heart failure who develop Cheyne-Stokes respiration (CSR) during sleep experience excessive daytime sleepiness. This was addressed by comparing sleep quality and daytime sleepiness in three groups: patients with CHF and CSR during sleep (CSR group), patients with CHF without CSR (CHF group), and healthy control subjects (control group). Design: Single-blind, cross-sectional study. Setting: Patients referred by cardiologists and control subjects recruited from the general community. Patients: Twenty-three men: 7 in the CSR group, 7 in the CHF group, and 9 in the control group. Measurements: Each subject had an overnight sleep study and an assessment of sleepiness the following day. Results: The three groups were a similar age: CSR, 68 ± 5 years; CHF, 62±4 years; and control, 65±4 years; and left ventricular ejection fraction was the same in patients with CSR ( ± 1.5% ) and CHF (23 ± 5 %). Sleep latency was significantly shorter in patients with CSR (4 ± 1.1 min) than patients with CHF(11.3± 4.8 min) and healthy controls (12.4 ± 1.9 min) and was within the diagnostic range of severe sleepiness. Patients with CSR had significantly more stage 1 and 2 non-rapid eye movement (NREM) sleep (CSR, 83 ± 7; CHF, 64 ± 9; control, 63 ± 9 % total sleep time), less REM sleep (CSR, 10±3; CHF, 22±8; control, 22±7% total sleep time), and a higher frequency of arousals from sleep (CSR, 30 ± 16; CHF, 18 ± 15; control, 10 ± 2/ h of sleep); 66% of arousals were associated with CSR. Regression analysis revealed that sleep latency was inversely related to the amount of stage 1 and 2 NREM sleep (r= -0.67), arousal frequency (r= -0.46), and the apnea-hypopnea index (r= -0.63) and was positively correlated with the amount of slow-wave sleep (r=0.45) and REM sleep (r=0.56) and the mean oxygen saturation during sleep (r=0.50). Conclusions: Patients with CHF who develop CSR experience excessive daytime sleepiness due to sleep disruption. This should be considered the clinical evaluation of these patients' daytime complaints. (Chest 1995; 107:952-58) AHI=apnea-hypopnea index; CHF=congestive heart failure; CSR=Cheyne-Stokes respiration; EMG=electromyogram; EOG=electro-oculogram; MSLT=multiple sleep latency test; :'IIREM=non-rapid eye movement sleep; REM= rapid eye movement sleep; Sa02=oxygen saturation; SWS =slow-wave sleep; tco2=transcutaneous Pco2; TST=total sleep time Key words: Cheyne-Stokes respiration; daytime sleepiness; heart failure ~ t h oexcessive u g h daytime sleepiness has been observed previously in the clinical presentation of Cheyne-Stokes respiration (CSR),l- 3 these studies were not designed to address this specific issue and consequently were not controlled for potential confounding variables. Daytime sleepiness is a disabling symptom and its consequences range from reduced daytime performance to an increased risk of road traffic accidents. 4-6 Recent studies have shown that CSR during sleep can be reduced both by supplemental oxygen 7 therapy and nasal continuous positive airway pressure. 2 Consequently, demonstration of daytime hypersomnolence associated with CSR in an appropriately controlled study could have important clinical implications for the management of patients with congestive heart failure (CHF). Daytime sleepiness is often found in patients with *From the Sleep Laboratory, The Wellesley Hospital, University of Toronto, Ontario, Canada. Supported by Physicians Services Incorporated Foundation. Manuscript received March 17, 1994; revision accepted August 5. obstructive sleep apnea 8 where its pathogenesis has been related both to sleep fragmentation 9 and nocturnal hypoxemia 10 although predominantly the former. Cheyne-Stokes respiration is a form of periodic breathing with alternating central apnea and hyperpnea that has a characteristic crescendo/ decrescendo pattern. 11 It has been estimated to occur during sleep in 40 to 50% of patients with significantly impaired left ventricular function.iu 2 Apneas are usually associated with oxygen desaturation and hyperpneas are associated with arousal from sleep. 1 Patients with CSR experience nocturnal hypoxemia and sleep disruption that may be as severe as that associated with significant obstructive sleep apnea. 1 We hypothesized that patients with heart failure who have CSR during sleep develop excessive daytime sleepiness whereas patients with heart failure without CSR do not. To test this hypothesis, we assessed sleep quality and daytime alertness in three groups of patients: Patients with CHF with and without CSR and healthy control subjects. 952 Daytime Sleepiness in Patients With CHF and Cheyne-Stokes Respiration (Hanly, Zuberi-Khokhar)

2 Patient Population METHODS We studied 23 subjects: 14 patients with CHF and 9 healthy control subjects. Each subject had an overnight sleep study and the following day completed a subjective and objective assessment of daytime sleepinesss. The patients with heart failure were recruited from the cardiology clinic at our hospital by asking those who had severe, stable heart failure (New York Heart Association class 3 or 4) with a resting left ventricular ejection fraction less than 35% to participate. The control subjects were recruited from the general population by asking healthy individuals who were age and sex matched with the heart failure group to participate in a research study. A history of abnormal breathing during sleep or excessive daytime sleepiness was not required for entry to the study. Patients were excluded if they had the following: (1) significant pulmonary, renal, or neurologic disease; (2) a known cause of excessive daytime sleepiness; or (3) sedative medication prescribed. Patients with CHF were subsequently divided into two groups based on whether they had CSR on the overnight sleep study: seven patients had CSR with an apnea-hypopnea index (AHI) greater than /h (CSR group) and seven patients had no evidence of CSR (CHF group). Recruitment of patients with heart failure into the CSR group was determined solely by the presence of CSR on the overnight sleep study and was not influenced by daytime symptoms or multiple sleep latency test (MSL T) results. The protocol was approved by the Ethics Committee of our institution and informed consent was obtained from each patient. Sleep Study The following variables were monitored during overnight polysomnography. We recorded two-channel electroencephalogram (EEG) (Cs-A2, C4-A1), electro-oculogram (EOG), and submental electromyogram (EMG) using surface electrodes. Airflow was detected by monitoring expired C02 at the nose and mouth through nasal cannulas adapted for this purpose and attached to a C02 analyzer (PB 223, Puritan-Bennett Corporation). Respiratory effort was monitored by respiratory plethysmography with transducers placed around the chest and abdomen (Respitrace, Ambulatory Monitoring, Ardsley, NY). Arterial oxygen saturation (Sa02) was recorded with a pulse oximeter (Biox 3740, Ohmeda, Boulder, Colo) set at its fastest response. Transcutaneous Pco2 (tcoz) was recorded by a Pco2 sensor placed on the anterior chest wall and attached to a C02 monitor (Kontron 7640, Medilog Ltd). The ECG and heart rate were recorded from standard limb leads. Leg movements were monitored by recording anterior tibialis EMG from both legs using surface electrodes. All variables were continuously recorded on a polygraph (model 78E, Grass Instruments, Quincy, Mass) at a paper speed of 10 mm/s. Transcutaneous Pcoz was displayed on a slow recorder (paper speed, em/h) that was synchronized to the polygraph (Grass). All polysomnograms were scored manually and sleep stage and arousals were determined by established criteria using the EEG, EOG, and EMG. 13 An arousal was defined as an awakening from sleep for more than 5 s, as evidenced by simultaneous alpha activity on the EEG, EMG activation, and eye movements. Arousals associated with the hyperpneic phase of CSR were identified. Central apnea was defined as absence of airflow for more than 10 s due to loss of respiratory effort. Hypopnea was defined as a reduction in the amplitude of respiratory effort of at least 50% from the sleeping baseline level for more than lo s with or without an associated fall in oxygen saturation. The AHI was defined as the number of apneas and hypopneas per hour of sleep. Cheyne Stokes respiration was defined by the following criteria: (1) regularly alternating periods of central apnea (or hypopnea) and hyperpnea; (2) hyperpneic phase exhibited a clear crescendo/ decrescendo pattern; and (3) three or more consecutive cycles of periodic breathing. Although apneas were always associated with oscillation in oxygen saturation, significant hypoxemia was not required. The amount of CSR during sleep was quantified by expressing the duration of CSR as a percentage of total sleep time (TST). Mean oxygen desaturation during sleep (mean Sa02) was calculated by averaging the high and low SaOz for each 30-s epoch. In addition, mean Sa02 during 5 to 15 min of wakefulness at the beginning of the study (SaOz [W]) was similarly calculated. Transcutaneous Pco 2 was measured during two different conditions: (1) mean tcoz during 5 to 15 min of wakefulness at the beginning of the study (tcoz [W]) and (2) mean tco2 during sleep (tco 2 [TST]). All mean values of tcoz were calculated from the average tcoz over 36-s intervals. Periodic leg movements were defined as four or more consecutive, involuntary leg movements during sleep, lasting 0.5 to 5.0 s with an intermovement interval of 5 to 90s. Cardiac Function Left ventricular ejection fraction was estimated at rest by radionuclide angiography. Circulatory delay from the lung to the carotid body was estimated either from the overnight sleep study or during a voluntary breath hold the next day. Circulation time was calculated from the end of an apnea or voluntary breath hold to the nadir of oxygen desaturation estimated with an ear oximeter set at its fastest response. Apneas with an abrupt termination and onset of respiration were chosen so that a clear increase in Sa02 could be detected. The average circulation time from at least ten apneas or breath holds was obtained in each patient. Daytime Sleepiness A detailed sleep history was taken from each subject. No patient was sleep restricted and mean reported sleep time was not significantly different between groups (CSR, 7.2 ± 1.0 hs; CHF, 7.7±0.9 h; control, 7.1±0.5 h). Daytime sleepiness was assessed by a subjective rating (Epworth Sleepiness Scale) in each patient the morning after the overnight sleep study. This is a self-administered questionnaire that asks the patient to rate on a numeric scale the likelihood of falling asleep in different situations such as watching TV or sitting in a car. The scale ranges from 0 ("would never doze") to 3 ("high chance of dozing") and the potential score ranges from 0 to 24 with a higher score indicating increased sleepiness. The Epworth Sleepiness Scale has been correlated with the MSL T and a score of 5.9 ± 2.2 has been reported in normal controls and 16 ± 4.4 in patients with severe obstructive sleep apnea. 14 Daytime sleepiness was also measured objectively by an MSL T that was done in a standardized fashion 15 the day after the overnight sleep study. Scorers were blinded to the presence of CSR since these recordings did not include any respiratory parameters. Sleep stage was determined from surface electrodes that recorded three-channel EEG (Cs-A2, C4-A 1, OrA2), two-channel EOG (F 7 -A1. F 8 -A2), and submental EMG. During the MSLT, the patient lay in bed in a quiet, darkened room and was invited to fall asleep. The patient took four naps throughout the day at 9 AM, ll AM, 1 PM, and 3 PM. Sleep onset was defined as the time from "lights out" to the first of three consecutive epochs of stage 1 NREM sleep or one epoch of any other sleep stage. Once sleep onset was identified, the patient was awakened to prevent consolidated sleep improving his performance on subsequent naps. In addition, the patient was observed between naps to ensure that he did not sleep. Statistical Analysis Significant differences between groups were analyzed by either unpaired t test or one-way analysis of variance (ANOV A) with the Bonferroni test to detect significant differences between means. CHEST /107/4/ APRIL,

3 Table l-patient Data in Those With CSR, Those Without CSR (CHF), and Healthy Control Subjects* No Sex, M:F 7:0 7:0 9:0 Age, yr 68±5 62±4 65±4 BMI, kg/ m 2 28±4 26.2± ±3 CSR, %TST 58± EF,% ± ±5 CT, s 16.8±3.9t 15.2±7lt 9± 1.5 Cardiac medications, % of group D Di A N v AA *BMI=body mass index; TST=total sleeptime; EF=left ventricular ejection fraction; CT=circulation time; D=digoxin; Di=diuretics (furosemide); A=angiotensin-converting enzyme inhibitor (captopril); N =nitrates; V =vasodilators (hydralazine); AA =antiarrhythmics (am iodarone, mexiletine). fp<0.05 vs control. Differences between sample proportions were analyzed by x 2 test. Both simple regression and stepwise linear regression analysis were used to examine the relationship between mean sleep latency (MSLT) and sleep study variables. A p value <0.05 was considered significant. RESCLTS The 14 patients we studied had severe, stable congestive heart failure (NYHA class 3 or 4). Left ventricular ejection fraction was less than 35% in all patients. All patients with heart failure and control subjects were male with a similar age and weight (Table 1). The CSR group had a large amount of periodic breathing during sleep that occupied 58% of the TST. By definition, the CHF group had no CSR during sleep, nor was periodic breathing seen in our control subjects. Both the CSR and CHF groups had similar impairment of left ventricular function as evidenced by their ejection fraction and circulation time that was almost twice as long as that in our control subjects. Ejection fraction was presumed to be normal in the control subjects since they had no history or physical evidence of impaired ventricular function. Although the patients with CSR and CHF were taking a variety of cardiac medications, the proportion of each group taking individual drugs was quite similar. None of our patients with heart failure were taking beta-blockers or medication known to cause excessive sleepiness. None of our control subjects were taking medication. The sleep data in the three groups of patients are shown in Table 2. Although TST (lights out to lights on) was 6 to 7 h in our patients, mean TST was 5 h or less in the CSR and CHF groups due to a reduced sleep efficiency. Our patients with CSR had signifi- Table 2-Sleep Data in Patients With CSR, Patients Without CSR (CHF), and Healthy Contol Subjects* TST, h 4.4± ± ±0.9 Sleep efficiency, % 71±26 77±8 88±10 Stage 1, 2, %TST 83±7t 64±9 63±9 SWS, %TST 7±7 14±8 15±7 REM, %TST 10±3t 22±8 22±7 AR (total)/ h 30± ±15 10±2 AR (CS R)/ h ±14 PLM >25/ h (% of group) ll PLM/ h 48±501 40± ±12 AR (PLM)/ h 10±181 9± 121 1±1 *Stage 1,2=stage 1 and 2 REM sleep; SWS=slow-wave sleep; AR (total)=total movement arousals from sleep; AR (CSR)=arousals associated with the hyperpneic phase of CSR; PLM=periodic leg movements; AR (PLM )=arousals associated with PLM. fp<0.05 vs control and CHF. tp<0.05 vs control. cantly more stage 1 and 2 NREM sleep and less REM sleep than the CHF and control groups. In addition, the CSR group tended to have less slow-wave sleep (SWS), although this did not reach statistical significance. The reduction in sleep quality in the CSR group was most likely due to the higher frequency of arousals from sleep that was three times higher than in the control group although not statistically different from the CHF group. Sixty-six percent of these arousals were associated with the hyperpneic phase of Cheyne-Stokes breathing, and the remainder were associated with periodic leg movements or occurred spontaneously. These differences in sleep quality were not caused by periodic leg movements since the proportion of patients in the CSR and CHF groups who had more than 25 leg movements per hour of sleep were identical. In addition, both the frequency of periodic leg movements and associated arousals were not different between CSR and CHF groups. These results suggest that the impairment of sleep quality that some patients with CHF suffer is due to the development of CSR during sleep rather than heart failure itself. Respiratory data during wakefulness and sleep in all patients are summarized in Table 3. Mean oxygen saturation during wakefulness was higher in control subjects than in the CSR and CHF groups. Although oxygen saturation was not different between CSR and CHF groups during wakefulness, it fell significantly during sleep in patients with CSR. However, the extent of nocturnal hypoxemia was not severe and was predominantly due to one patient whose mean Sa02 during sleep was 75.7%. Transcutaneous Pcoz tended to be lower both during wakefulness and sleep in patients with CSR. This has been reported previously as an important determinant of which patients with 954 Daytime Sleepiness in Patients With CHF and Cheyne-Stokes Respiration (Hanly, Zuberi-Khokhar)

4 Table 3-Respiratory Data in Patients With CSR, Patients Without CSR (CHF), and Healthy Control Subjects* SaOz (W) 91.4± ± ± 1.0 tcoz (W) 34.3± ± ±5.3 SaOz (TST) 87.6±5.4t 91.6± ± 1.2 tco 2 (TST) 38.4± ± ±3.6 CSR (%TST) 58± AHI, h 33±91 9±7 9±6 *SaOz (W), tcoz (W)=mean oxygen saturation and transcutaneous Pcoz during 5 to 15 min of wakefulness at the beginning of the sleep study; SaOz (TST), tcoz (TST)=mean oxygen saturation and transcutaneous Pcoz during sleep; CSR (%TST)=Cheyne-Stokes respiration expressed as a proportion of the total sleep time; fp<0.05 vs control.!p<0.05 vs control and CHF. p<0.05 vs CSR and CHF. heart failure develop CSR during sleep.l 6 By definition, only the CSR group had periodic breathing during sleep and consequently the AHI was significantly higher in this group. The mean AHI was not normal in the CHF and control groups due to a low frequency of posthyperventilation central apnea or obstructive apnea during sleep. However, this sleepdisordered breathing was mild and was not believed to be clinically significant. Results of our assessment of daytime sleepiness in all patients are shown in Table 4. Subjective rating of daytime sleepiness by the Epworth Sleepiness Scale tended to be higher in the CSR group, although this did not reach statistical significance. However, mean sleep latency on the MSL T was significantly lower in the CSR group and within the diagnostic range of severe sleepiness.l 5 A more detailed presentation of the MSL T data is provided in Figure 1 that indicates that mean sleep latency was persistently lower in the CSR group throughout the day. These data indicate that patients with heart failure without sleep-disordered breathing have similar daytime alertness to healthy individuals of the same age and that patients with heart failure with CSR during sleep are consistently sleepier. The excessive sleepiness that patients with heart failure with CSR experience is most likely Table 4-Epworth Sleepiness Scale (ESS) Scores and Mean Sleep Latency on MSLT in Patients with CSR, Patients Without CSR (CHF), and Healthy Control Subjects ESS MSLT, min 8.9±3.5 4±1.1* *p<0.05 vs control and CHF. MSL T=multiple sleep latency test. 8.1 ± ± ± ± 1.9 due to the nocturnal hypoxemia and sleep fragmentation caused by recurrent apnea and hyperpnea, respectively (Tables 2 and 3). To examine the relationship between sleep latency on the MSL T and sleep fragmentation and nocturnal hypoxemia, we performed both simple and stepwise linear regression analysis between mean sleep latency (dependent variable) and the polysomnographic data (independent variables) in all patients. Simple linear regression analysis indicated a significant inverse relationship between sleep latency on the MSL T and the duration of stage 1 and 2 NREM sleep (r= -0.67), the frequency of arousals from sleep (r= -0.46), and the AHI (r= -0.63). There was a significant positive correlation between sleep latency and the duration of SWS (r=0.45) and REM sleep (r=0.56) and mean nocturnal Sa0 2 during sleep (r=0.50) (Fig 2). Stepwise linear regression analysis for all patients where group classification was not considered indicated that only the duration of stage 1 and 2 NREM sleep was significantly correlated with sleep latency. When group classification was included in the analysis, the only variable that was significantly correlated with sleep latency was the presence of CSR. DISCUSSION Patients with heart failure with CSR during sleep are hypersomnolent in contrast to patients with heart failure without CSR who have a similar tendency to fall asleep as healthy, age-matched control subjects. Although excessive daytime sleepiness has been reported previously in the clinical presentation of CSR, 1-3 these studies have not been controlled for the _,_ CONTROL FIGt:RE l. Mean sleep latency on multiple sleep latency test (MSL T) in patients with Cheyne-Stokes respiration (CSR group), patients without Cheyne-Stokes respiration (CHF group), and healthy subjects (control group). Asterisk= p<0.05 vs CHF and control. SLEEP LATENCY (min., MSLT) :00 14:00 16:00 T l (hours) ~ 1 l t * - - CI-F...t..- CSR CHEST /107/4/ APRIL,

5 r P r 0.45 P ~ MSLT NREM sws. '.. 5 r r P 0.05 p O.OOS '. MSLT AR AHI 0 7. p ,... r REM Sa02 FICCRE 2. Relationship between mean sleep latency on multiple sleep latency test (MSL T) and polysomnographic data in all patients. 1,2 NREM=stage 1 and 2 non-rapid eye movement sleep; SWS= slow-wave sleep; REM=rapid eye movement sleep; AR =number of arousals from sleep (per hour); AHI=apnea-hypopnea index (per hour); Sa02=mean nocturnal oxygen saturation during sleep. Sleep stage data are expressed as a percentage of total sleep time. effect of other potential influences on daytime alertness such as age, study protocol, coexisting sleep disorders, and heart failure itself. All of our patients were subjected to the same study protocol. We excluded patients with other potential causes of daytime sleepiness such as sleep restriction, narcolepsy, obstructive sleep apnea, other medical and psychiatric illnesses, alcohol abuse, and sedative medicationy Although periodic leg movements, which may be associated with excessive daytime sleepiness, 18 were found in many patients, the proportion of the CSR and CHF groups with a significant periodic leg movement disorder were identical (Table 2) and relatively few leg movements were associated with arousals from sleep. In addition, both the frequency of leg movements and associated arousals were not different between CSR and CHF groups. The severity of heart failure was similar in both the CSR and CHF groups as evidenced by their NYHA functional status, ejection fraction, and circulatory delay (Table I). The only significant difference between the CSR and CHF groups was the presence of periodic breathing during sleep in the former. Consequently, we are confident that the hypersomnolence manifested by our patients was related to CSR and was not caused by the patients' age, study protocol, coexisting sleep disorders, or heart failure itself. Furthermore, the association between hypersomnolence and CSR was not due to a selection bias since the recruitment of patients with heart failure into the CSR group was based solely on the presence of CSR on the overnight sleep study and was not influenced by daytime symptoms or MSLT results. We used quite conservative criteria to score sleep onset during the MSL T. The conventional requirement for scoring sleep onset in the MSL T is one 30-s epoch of NREM sleep. 19 In effect, only 16 s of unin- terrupted sleep are required since each epoch is scored as the sleep stage that occupies more than 50% of its duration. We believed that more conservative criteria were required to identify established sleep. Despite our stricter criteria, we found that patients with heart failure with CSR had a mean sleep latency of less than 5 min that has been acknowledged to indicate a pathologic level of daytime sleepiness both in patients with sleep apnea and narcolepsy 21 and in sleep-deprived normal individuals. 22 This degree of sleepiness has been associated with significantly impaired daytime performance. 23 We found that patients with heart failure who develop CSR during sleep experience significant sleep disruption (Table 2). Although only two parameters of sleep quality were statistically different between CSR and CHF groups, there was -a strong trend for the remaining indices to be worse in the CSR group. This was due to changes associated with CSR since patients with heart failure without CSR slept as well as healthy individuals of the same age. Consequently, it appears that patients with severe heart failure whose conditions have been stabilized with medical therapy do not experience sleep disruption unless they develop periodic breathing. The cause of sleep disruption is related to the recurrent arousals during the hyperpneic phase of CSR 1 that accounted for two thirds of the arousals in the CSR group (Table 2). Mean oxygen saturation during sleep was significantly lower in patients with CSR than patients with CHF and healthy subjects. However, this statistical difference was strongly influenced by one patient with CSR who had more severe hypoxemia than others in that group. This is consistent with previous studies that reported relatively mild hypoxemia associated with CSR 24 and attributed the beneficial effect of supplemental oxygen on sleep more to stabilization of breathing rather than correction of 956 Daytime Sleepiness in Patients With CHF and Cheyne-Stokes Respiration (Hanly, Zuberi-Khokhar)

6 hypoxemia per se? We used a relatively small number of patients in our study. Although a larger sample size would have strengthened the statistical analysis, we were still able to find significant differences in the polysomnographic data and MSL T results between groups. Comparison of data from a single sleep study may also be regarded as a methodologic limitation, particularly if abnormalities such as CSR were not a consistent finding. However, five of the seven patients in the CSR group were restudied within 1 month as part of another research protocol and their polysomnographic findings were remarkably consistent (author's unpublished observations, P.H., 1992). Consequently, we believe that within the limitations of our study design, our findings are significant. The causes of excessive daytime sleepiness in obstructive sleep apnea have been extensively investigated. Although these patients frequently experience both significant hypoxemia and sleep fragmentation, the cause of their daytime sleepiness is now considered to be predominantly related to the latter Most patients in our study who were hypersomnolent experienced significant sleep disruption and relatively mild hypoxemia. Stepwise regression analysis, without classification of patients into our three groups, indicated that only the duration of light sleep was significantly correlated with sleep latency. An increased proportion of light sleep is a correlate of disrupted sleep. Unfortunately, we were unable to explore this further by stepwise regression analysis within the CSR group due to an inadequate sample size. Nevertheless, these results support the suggestion that hypersomnolence in patients with sleep apnea is predominantly due to sleep fragmentation rather than hypoxemia. There was a discrepancy between the subjective and objective assessment of daytime sleepiness in our patients (Table 4). Despite shorter sleep_ latency on the MSL T in patients with CSR (Fig l ), their Epworth Sleepiness Scale scores were not significantly different from patients with heart failure and control subjects. These differences may reflect a deficiency in the assessment of daytime sleepiness by the Epworth Sleepiness Scale questionnaire, although it has been correlated with the MSL Tin patients with narcolepsy and obstructive sleep apnea. 14 Alternatively, these results may reflect a true difference between the physiologic assessment and behavioral assessment of sleepiness that has been previously reported in those with obstructive sleep apnea 26 and normal individuals. 27 What are the clinical implications of our findings for the management of patients with heart failure and CSR? Daytime sleepiness can be a disabling symptom that is \vell recognized to reduce the quality of life when left untreated in patients with narcolepsy and obstructive sleep apnea. 4 5 Patients with severe CHF have many somatic complaints such as generalized fatigue, insomnia, and impairment of neuropsychologic function. 28 Although many of these symptoms can be attributed to their cardiac dysfunction, they may also be influenced by sleep disruption and excessive sleepiness due to CSR. The coexistence of heart failure and daytime sleepiness may make the latter more difficult to diagnose without formal, objective testing such as MSL T. If excessive daytime sleepiness does complicate CSR in patients with heart failure, it can be treated through correction of periodic breathing by supplemental oxygen 7 or nasal continuous positive airway pressure. 2 Failure to recognize it deprives these patients of a chance to improve their daytime sleepiness and quality of life. In summary, we found that patients with heart failure who develop CSR during sleep are excessively sleepy. We believe that daytime sleepiness is predominantly caused by sleep disruption associated with periodic breathing and that it may be an unrecognized cause of impaired daytime performance and morbidity in these patients. We suggest that sleep quality and daytime alertness be carefully assessed in patients with heart failure suspected of having CSR during sleep. ACKNOWLEDGMENT: The authors wish to thank Elvie Garcia for typing the manuscript. R EFERENCES Hanly PJ, Millar TW, Steljes DG, et al. Respiration and abnormal sleep in patients with congestive heart failure. Chest 1989; 96: Takasaki T, Orr D, Popkin J, et al. Effect of nasal continuous positive airway pressure on sleep apnea in congestive heart failure. Am Rev Respir Dis 1989; 140: Dowdell WT, Javaheri S, McGinnis W. Cheyne-Stokes respiration presenting as sleep apnea syndrome. Am Rev Respir Dis 1990; 141: Guilleminault C, Carskadon M. Relationship between sleep disorders and daytime com plaints. In: Koeller WP, Oevin PW, eds. Sleep Basel: S. Karger, 1977; Broughton R, Ghanem Q, Hishikawa Y. Life effects of narcolepsy in 180 patients from North America, Asia and Europe compared to matched controls. J Can Sci Neurol1981; 8 : Findley L, Unverzagt M, Suratt P. Automobile accidents in patients with obstructive sleep apnea. Am Rev Respir Dis 1988; 138: Hanly P, Miller T, Steljes D, et al. Effect of oxygen on sleep and respiration in patients with severe congestive heart failure. Ann Intern Med 1989; 111: Kryger MH. Management of obstructive sleep apnea. Clin Chest Med 1992; 13: Roehrs T, Zorick F, Wittig R, et al. Predictors of objective level of daytime s leepiness in patients with sleep-related breathing disorders. Chest 1989; 95: Sink J, Bliwise DL, Dement WC. Self-reported excessive daytime somnolence and impaired respiration in sleep. Chest 1986; 90: CHEST 1107 I 4 I APRIL,

7 11 Harrison TR, King CE, Calhoun JA, et al. Congestive heart failure: Cheyne-Stokes respiration as the cause of paroxysmal dyspnea at the onset of sleep. Arch Intern Med 1934; 53: Yamashiro Y, Kryger MH. Sleep and heart failure. Sleep 1993; 16: Kales A, Rechtschaffen A, eds. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. NIH publication No. 4. Bethesda, Md: National Institute of Neurological Disease and Blindness, 1968; Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991; 14: Carskadon M. Measuring daytime sleepiness. In: Kryger MH, Roth T, Dement WC, eds. Principles and practice of sleep medicine. Philadelphia: WB Saunders, 1989; Hanly P, Zuberi N, Gray R. Pathogenesis of Cheyne-Stokes respiration in patients with congestive heart failure: relationship to arterial Pco2. Chest 1993; 104: Moldofsky H. Evaluation of daytime sleepiness. Clin Chest Med 1992; 13: American Sleep Disorders Association. The international classifications of sleep disorders: diagnosis and coding manual. Lawrence, Kan: Allen Press Inc, 1990; Thorpy MJ. The clinical use of the multiple sleep latency test (report from the American Sleep Disorders Association). Sleep 1992; 15: Dement WC, Carskadon MA, Richardson GS. Excessive daytime sleepiness in the sleep apnea syndrome. In: Guilleminault C, Dement WC, eds. Sleep apnea syndromes. New York: Alan R. Liss, 1978: Richardson GS, Carskadon MA, Flagg W, et al. Excessive daytime sleepiness in man: multiple sleep latency measurement in narcoleptic and control subjects. Electroencephalogr Clin Neurophysiol1978; 45: Carskadon MA, Dement WC. Cumulative effects of sleep restriction on daytime sleepiness. Psychophysiology 1981; 18: Roth T, Roehrs T, Carskadon M, et al. Daytime sleepiness and alertness. In: Kryger MH, Roth T, Dement WC, eds. Principles and practice of sleep medicine. Philadelphia: WB Saunders, 1989; Bradley TD. Right and left ventricular functional impairment and sleep apnea. Clin Chest Med 1992; 13: Guilleminault C, Dement WC. Sleep apnea syndromes and related sleep disorders. In: Williams RL, Karacan I, eds. Sleep disorders: diagnosis and treatment. ew York: John Wiley, 1978; Dement WC, Carskadon MA, Richardson GS. Excessive daytime sleepiness in the sleep apnea syndrome. In: Guilleminault C, Dement WC, eds. Sleep apnea syndromes. New York: Alan R. Liss, 1978; Johnson LC, Freeman CR, Spinweber CL. Subjective and objective measures of sleepiness: effect of benzodiazepine and caffeine on their relationship. Psychophysiology 1991; 28:65 28 Braunwald E. Clinical manifestations of heart failure. In: Braunwald E, ed. Heart disease. Philadelphia: WB Saunders, 1988; {1995} Mark Your Calendar! October 29- November 2, 1995 New York. New York The Sixty-First Annual International Scientific Assembly American College of Chest Physicians 958 Daytime Sleepiness in Patients With CHF and Cheyne-Stokes Respiration (Hanly, Zuberi-Khokhar)