- Email: [email protected]
Sleep-Related Myocardial Ischemia and Sleep Structure in Patients With Obstructive Sleep Apnea and Coronary Heart Disease
Sleep-Related Myocardial Ischemia and Sleep Structure in Patients With Obstructive Sleep Apnea and Coronary Heart Disease*
Harold
Jorg
Schdfer, MD; Ulrich Koehler, MD; Thomas Ploch; and
Hermann Peter, MD
Study objectives: Patients with coronary heart disease (CHD) and obstructive sleep apnea may have an increased cardiac risk due to nocturnal myocardial ischemia triggered by apneaassociated oxygen desaturation. Sleep structure in patients with obstructive sleep apnea is system (CNS) (arousal) due to obstructive apneas. fragmented by activation of the central nervous Nocturnal myocardial ischemia may lead to activation of the CNS as well. Patients: Fourteen patients with obstructive sleep apnea and CHD disease and seven patients suffering from obstructive sleep apnea without CHD were studied. Overnight sleep studies and simultaneous six-lead ECG recordings were performed. In addition, sleep studies and ECG recordings were performed with administration of a sustained-release nitrate in these patients in a double-blinded crossover design. Results: Analysis of three nights' recordings revealed 144 episodes of nocturnal myocardial ischemia in six subjects. Five patients had underlying CHD and one patient exhibited diffuse wall defects of the coronary arteries; also, 85.4% of ischemic episodes were concomitant with apneas and oxygen desaturation >3%, and 77.8% of ischemicepisodes occurred during rapid eye movement (REM) sleep, although total amount of REM sleep was only 18% of total sleep time. Mean oxygen saturation was significantly lower (p<0.05) during apnea-associated ischemic episodes than during nonapnea-associated ischemia (77.3% vs 93.1%). Nitrate administration did not reduce ischemic episodes. Sleep architecture (macrostructure) exhibited a reduction in sleep stages non-REM 3 and 4 and REM sleep. Comparing the microstructure of sleep (arousals) within episodes with and without ischemia but similar criteria like sleep stage, apnea activity, and oxygen saturation, we found significantly more (p<0.01) and severe (p<0.001) arousals during periods with myocardial ischemia than during control episodes. In addition, microstructure of sleep was disturbed by myocardial ischemia itself in absence of apneas. Conclusion: It is concluded that patients with CHD and obstructive sleep apnea are endangered by apnea-associated ischemia and that these ischemic episodes lead to activation of the CNS and additional fragmentation of sleep. Patients with nocturnal ischemia should be screened for (CHEST 1997; 111:387-93) underlying sleep apnea even if nitrate therapy fails. Key words: arousal; coronary heart disease; nocturnal myocardial ischemia; sleep apnea; sleep structure Abbreviations:
NREM=nonrapid eye movement; REM=rapid eye movement
¥7* xperience with ambulatory ECG monitoring in the last 20 years indicated that most ischemic episodes in patients with coronary heart disease were asymptomatic.1 Recent clinical studies support the -*-J
*From the Department of Internal Medicine, University of Bonn, Bonn, Germany (Dr. Schafer), and the Department of Internal Medicine, Sleep Disorders Center, University of Marburg, (Drs. Koehler, Ploch, and Peter). Marburg, Germany received Mav 3, 1996; revision accepted Septem¬ Manuscript ber 30. Reprint requests: Dr. Schafer, Medizinische Universitatsklinik und Poliklinik Bonn, Innere Medizin/Kardiologie und Pneumologie, Sigmund-Freud-Str. 25, D-53105 Bonn, Germany
relationship between silent myocardial ischemia and sudden death.2 Some believe that the autonomic nervous system triggers such events in these pa¬ tients.3 Obstructive sleep apnea, a condition with increased activation of the autonomic nervous sys¬ tem4 that is frequent in middle-aged men,5 shares many risk factors with coronary heart disease. Un¬ treated patients have a higher, predominantly car¬ diovascular mortality.6 Patients with the combination of coronary heart disease and obstructive sleep apnea may have an increased cardiac risk due to worsening of the relationship between myocardial oxygen deCHEST / 111 / 2 / FEBRUARY, 1997
387
of apnea-associated supply as a result autonomic nervous of and activation the hypoxemia In system. addition, sleep quality may be influenced by apnea-associated activation of theTheCNS (arousals)of and ischemia-associated arousals. objective this study was to determine the association of noc¬ turnal myocardial ischemia and sleep apnea in pa¬ tients with coronary heart disease and obstructive apnea. In addition, macrostructure and microsleep structure of sleep were analyzed with regard to ischemic episodes and the effect of nitrates on ischemic episodes and sleep structure was tested.
mand and
Materials We studied 21 male
and
Methods
patients with obstructive sleep apnea
consecutive nights in the sleep labora¬ (apnea index ^10) in sixhad a combination of obstructive sleep Fourteen patients apnea and coronaiy heart disease diagnosed by coronary angiog¬ raphy (group 1). Seven patients exhibited obstructive sleep apnea without underlying coronary heart disease as demonstrated by coronary angiography (group 2). Indications for cardiac catheter¬ ization were history of chest pain or pathologic findings in exercise ECG. In addition to the usual sleep and cardiorespira¬ tory measurements, all patients had six-lead ECG monitoring were weaned off treatment with all during sleep. Patients calcium-channel blockers, and P-blockers medication, antianginal at least 48 h prior to the first polysomnography. In a double-blind crossover design, patients were administered placebo-controlled a single dose of a sustained-release nitrate or placebo at the and 5,6 to test the effect beginning of the sleep study in nightsof 3,4 of nitrate therapy for suppression nocturnal ischemia. There¬ fore, we chose a dose at the upper limit of a single-dose sustained-release nitrate (100 mg isosorbide dinitrate). Evalua¬ tion of polysomnographies was based on the data obtained in the second night (without medication), fourth night (placebo or active medication), and sixth night (placebo or active medication).
tory.
Informed consent was obtained from all patients.
Exercise Test
Before sleep studies, all patients underwent a maximal graded in supine position. A minimum load of 50 W bicycle stress testwhich was increased by 25 W every 2 min.7 was performed, Twelve-lead ECGs were printed out for each minute during
exercise and for the first 5 min of recovery or until the baseline was achieved. An ischemic ECG response was defined as > 1 mm horizontal or downsloping ST-segment depression below
state
the PR segment for >0.08 s after the J point on at least three consecutive beats. The exercise test was stopped at the onset of ischemic ECG changes with or without symptoms or when the target heart rate was reached.
Coronary Arteriography
Both left and right coronary arteriograms were performed in all patients using the standard Judkins approach and stenotic lesions
considered significant if there was >50% diameter stenosis. Diffuse wall defects of the coronary vessels were considered not significant if the remaining, patent diameter was at least 50%.
were
Sleep
Studies two-channel EEG (C3-A2, C3-02), electroand submental electromyogram using surface elec¬
We obtained
oculogram, 388
a
trodes. Airflow was detected using a nasal thermistor. Arterial oxygen saturation was recorded with an ear oximeter. Respiratory effort was monitored by inductive plethysmography of the chest wall and abdomen. The ECG and heart rate were recorded from standard limb leads. Body movements were detected by means of an actigraph on the forearm. All parameters were registered
continuously by a sleep system (SIDAS 2010 sleep-system; Stimotron; Wendelstein, Germany) and were converted from analog to digital signals at a sampling rate of 100 Hz. The method has been described elsewhere in detail.8 The digital signals were sent to an integrated personal computer, which both presented the data on-line on an ink jet printer and screen display and stored the data at a sampling rate of 1 to 100 Hz for each signal on a hard disk in order to perform an off-line analysis. Scoring Rules
All sleep studies were scored manually and sleep was staged in intervals of 20 s duration by the standard guidelines of Rechtschaffen and Kales.9 We used a modified classification of arousal scoring on the basis of Stepanski et al10 and from the Sleep Disorders Atlas Task Force of the American Sleep Disor¬ ders Association.11 Arousals were defined as abrupt shifts in EEG frequency following a minimum of 10 continuous seconds of intervening sleep. They were assigned to three categories: micro, gross, and complete arousals, whereas awakenings represented a separate category. Arousals were classified as microarousals when the criteria of the general definition of an arousal were found but no alpha activity was present. In a gross arousal, alpha-bursts 3 s at least. In a complete arousal, appeared for <15 s but were defined by alpha-bursts persisted for >15 s.^60Awakenings s. Alpha intrusion in rapid persistence of alpha activity for eye movement (REM) sleep was not scored unless it was activ¬ accompanied by an increase in submental electromyogram ity. Obstructive apnea was defined as cessation of airflow for at least 10 s in the presence of continued respiratory efforts. The apnea index was calculated from the number of apneas per hour of sleep.
Recordings A continuous recording of a six-lead chest wall ECG (MingoMunich, Germany) was performed and was graph; Siemens; out at a paper speed synchronized to the sleep study andofprinted of 25 mm/s. Standard placements the six-lead chest wall ECG were used (V1-V6). ST-segment depressions of >0.1 mV below the isoelectric line for >0.08 s after the J point with horizontal and descending courses and changes of theT wave, such as T inversion, were considered for myocardial ischemia. An ischemic documented only if its minimum duration was 10 s. episode was Isolated T-wave changes and ST-segment shifts occurring after change in body position were excluded. Episodes of myocardial ischemia were examined specifically with respect to the interac¬ tion between respiration (apnea, hyperventilation, oxygen satu¬ ration) and changes in the EEG pattern described previously. Ischemic episodes were compared with "control episodes" in the same patients with regard to arousal pattern. These control as periods without myocardial ischemia episodes were defined but with similar criteria like apnea activity, oxygen saturation, and sleep stage with close proximity to the ischemic events. ECG
Statistical Analysis
Descriptive statistics for continuous variables were expressed the mean ± SD. Differences in mean values were assessed using Student's two-tailed t test and analysis of variance. The x2 as
Clinical
Investigations
test with Yates' correction was used for categorical variables. In all cases, p values of <0.05 were considered to be significant.
Results
The mean age of the 21 patients was 56.3 years (47 67 years) and the mean weight was 89.4 kg (72 to 129 kg). None of the subjects had a clinical history of myocardial infarction. Mean Po2 during wakefulness was 81.9 mm Hg (64.6 to 100.4 mm Hg). The results of coronary arteriograms are listed in Table 1.
to
Respiration and General Sleep Structure Concerning the respiratory parameters, no signif¬
differences among the three recording nights found with regard to apnea index, oxygen saturation, and heart rate (Table 2). Furthermore, no significant differences in sleep stages between the recording nights and between the groups of patients with and without coronary heart disease were found. On the average, of all 21 patients, the amount of sleep stages 1 and 2 was 67.1% of the total sleep time and the amount of sleep stages 3 and 4 was 14.9%. Eighteen percent of total sleep time was spent in REM sleep. Sleep efficiency was 90.5% of total sleep time. Patients with coronary heart disease (group 1) spent an average of 66.8% in sleep stages 1/2, 15.2% in 3/4, and 18% in REM sleep. Patients without icant
were
Table Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
No.1
coronary heart disease (group 2) spent 67.0% in stages 1/2, 14.5% in 3/4, and 18.5% in REM sleep.
Myocardial Ischemia Episodes with nocturnal myocardial ischemia were found in 6 of these 21 patients. In four of the six patients with nocturnal myocardial ischemia, one single coronary vessel was affected in each patient and one
in patient two coronary vessels were af¬ fected. One patient exhibited diffuse wall defects of the coronary vessels. A total of 144 ischemic episodes with a maximum duration of 96 s (10 to 96 s) was recorded within three nights (Table 3). Nitrate ad¬ ministration was without effect with regard to num¬ ber and mean duration of ischemic episodes (Table 3). In the following, all the ischemic episodes of all three nights were pooled. In two patients, only few ischemic episodes occurred (patient 9, four episodes; patient 11, one episode), whereas in the others, more episodes occurred (patient 1, 34 episodes; patient 3,
episodes;These patient 14, 31 episodes; patient 15, 59 episodes). episodes appeared predominantly during periodsin of high apnea activity and oxygen desaturation, the minority within periods of un¬ disturbed ventilation. Of the 144 ischemic episodes, 123 were concomitant with and 15
(85.4%)
oxygen desaturation (>3%). In the
apneas
remaining
21
(14.6%), change breathing pattern or episodessaturation was evident. no
oxygen
in
1.Findings of Coronary Arteriography, Exercise ECG, and Nocturnal Sa02*
Nocturnal Ischemia
Minimal
Sa02
During Sleep, °i <60 64 <60
0
+
75 79 <60 75 75 75 <60 70 <60 80 60
0 0 0 0 0
77
70 78 72 80 80 84
No. of Affected Coronary Arteries (>50%)
Maximal Work Load (Watts) of Exercise ECG Findings1
1 3 1 1 3 3 1 2 1 2 1 2 2 2
No CHD, diffuse wall defects No CHD No CHD No CHD, diffuse wall defects No CHD No CHD No CHD, diffuse wall defects
No No No No
125 100 125 75 50 150 50 75 75 100 75 125 50 100 125
findings findings findings findings
*Sa02=arterial oxygen saturation; CHD=coronary heart disease. fThe first 14 patients represent group 1 (with coronary heart disease). The following seven patients represent group 2 (without coronary heart disease). "The maximal work load (watts)
at
which
pathologic changes of the ST segment occurred is determined.
CHEST / 111 / 2 / FEBRUARY, 1997
389
Table
2.Respiratory Parameters of the Patients With Nocturnal Myocardial Ischemia*
Respiratory Parameters (n=21 Patients)
Apnea index
Mean SD Min Max Duration of apneas, s Mean SD Min Max Mean arterial oxygen saturation during sleep, % Mean SD Min Max Minimal arterial oxygen saturation during sleep, % Mean SD Min Max Maximal drop of arterial oxygen saturation during sleep, Mean SD Min Max Mean heart rate during sleep, bpm Mean SD Min Max
Minimal heart rate Mean SD Min Max
during sleep, bpm
Maximal heart rate during sleep, Mean SD Min Max
bpm
Night 2
Night 4
Night 6
(Baseline)
(Placebo)
(Nitrate)
32.3 11.8 13.0 50.0
30.4 15.0 11.0 59.0
36.9 19.0 12.0 80.0
63.6 26.6 30.0 120.0
58.3 22.2 30.0 100.0
61.4 24.7 20.0 120.0
90.0 2.0 85.0 93.0
90.5 2.8 83.0 94.0
89.8 3.0 83.0 94.0
73.8
10.4 46.0 88.0
74.6 10.7 50.0 88.0
71.4 13.0 45.0 89.0
19.3 9.2 6.0 42.0
17.7 8.3 9.0 38.0
20.9 10.1 8.0 40.0
67.7 8.6 50.0 82.0
65.5 7.3 46.0 78.0
68.8
46.0 94.0
49.5 5.0 38.0 57.0
48.5 6.0 38.0 61.0
50.3 7.0 37.0 62.0
100.3 11.7 84.0 132.0
97.8 10.3 75.0 121.0
103.5 14.9 82.0 142.0
10.7
2 (without medication), night 4 (with placebo), and night 6 (with sustained *bpm=beats per minute. Evaluation was based on the data of night nitrate). Data are expressed in mean value with SD, minimum value (min), and maximum value (max).
Mean oxygen saturation was 91.0% at the onset of the 123 apnea-associated myocardial ischemic epi¬ sodes vs 77.3% during ischemia. Heart rate was monitored between 70 beats/min before and 77 beats/min during the ischemic episode. In the 21 ischemic episodes, oxygen sat¬ nonapnea-associated uration was 93.1% before and 92.9% during isch¬ emia. Heart rate was 75 beats/min before and 94 beats/min during ischemia. The differences between and nonapnea-associated ischemic apnea-associated in mean values and changes of oxygen the episodes saturation (mean change apnea associated, 13.6±10.3%; nonapnea associated, 0.3±0.7%; 390
and heart rate (mean change apnea associ¬ p<0.05) 7.4±16.5 beats/min; nonapnea associated: ated, -19.8±14.5 beats/min; p<0.001) were statistically significant. One hundred twelve ischemic episodes (77.8%) occurred during REM sleep, 10 (6.9%) in non-REM (NREM) stage 1, 20 (13.9%) in NREM stage 2, and two (1.4%) in NREM stages 3/4 (Fig 1). No statisti¬ cal differences were found between apnea-associ¬ ated and nonapnea-associated ischemic episodes with regard to sleep stage. Roth groups of ischemic episodes were frequent during REM sleep. Apneaassociated ischemic episodes occurred in 78.1% in Clinical
Investigations
Table 3.Number and Duration of Ischemic Episodes* Nocturnal
Myocardial Ischemia
(n=6 Patients)
No. of ischemic episodes Duration of ischemic episodes,
*
Night 2
Night 4 Night 6
(Baseline) (Placebo) (Nitrate)
Association*
45 40 59 s
29.3 Mean SD17.5 Min12.0 96.0 Max
28.2 13.4 12.0
73.0
27.7 16.3 10.0 91.0
Evaluation was based on the data of night 2 (without medication), night 4 (with placebo), and night 6 (with sustained nitrate). Data are expressed in mean value with SD, minimum value (min), and maximum value
Table 4.Comparison of Ischemic Episodes and Control Episodes With Regard to Arousal Association, Arousal Specification, Sleep Stage, and Apnea
(max).
Association +
0 Sum
Arousal type Microarousal Gross arousal Complete arousal
Awakening
sleep and nonapnea-associated ischemic epi¬ sodes occurred in 76.2% in REM sleep. Ischemic periods were asymptomatic in all patients. Microstructure of Sleep A total of 175 arousals were recorded during the 144 myocardial ischemic episodes. Seven ischemic periods (4.9%) revealed no arousal, and 137 (95.1%) resulted in at least one. Of these 137 ischemic episodes, 99 events (72.3%) were concomitant with one arousal and 38 events were REM
(27.7%)
concomitant
with a second arousal. During the 144 control epi¬ sodes, only 127 arousals were recorded (p<0.01). In 29 control episodes (20.1%), no arousal was de¬ tected. Of the 127 episodes, 115 events (90.5%) were concomitant with one arousal and 12 events (9.5%) were concomitant with a second one. Comparing the intensity of arousals, significantly more severe ones (complete arousals, p<0.001) were detected during ischemic episodes than during control episodes (Ta¬ ble 4). A statistically significant correlation (p<0.01; r=0.5) was found between oxygen saturation before
Sum
*
Control
Ischemic
Arousal
Episodes,
(%) 137(95.1) 7 (4.9) 144 (100)
p Value
37(21.1)
NS NS <0.001 NS <0.01
No.
57 (32.6) 75 (42.9)
6 (3.4) 175(100)
<0.01 <0.01 NS
Significant differences between ischemic and control indicated by p values.
Episodes,
(%) 115 (79.9) 29(20.1) 144 (100) No.
47 (37.0)
62 (48.8) 18 (14.2)
0(0)
127 (100)
episodes are
and during the ischemic episodes and the severity of arousals. The differences in oxygen saturation before and during ischemic episodes were 1.4% when no arousal occurred, 10.2% in microarousals, 11.4% in gross arousals, and 14.5% in complete arousals. Only 3.7% was the difference in awakenings. No statistically significant correlation was found between heart rate before and during the ischemic and arousals. episodes Microstructure of sleep during the 21 ischemic not apnea associated was signifi¬ episodes that weredifferent from control episodes. cantly (p<0.01) During ischemic episodes, only in 19% no arousal was detected, whereas in 71.4% of control episodes, no arousal was detected. Classification of arousals revealed in all of the control episodes microarousals, whereas in ischemic episodes, 40% were gross arous¬
als, 35% were complete arousals, 15% croarousals, and 10% were awakenings.
were
mi¬
Discussion
asymptomatic nocturnal ST-segment depression patients with obstructive sleep apnea in those patients with additional coronary heart only disease and in one patient with diffuse coronary vessel defects. The vast majority of these episodes We found
in
Sleep stages
BI Sleep stages H REM SLEEP
1/2 3/4
Total Sleep Time Ischemic Episodes
Figure 1. Distribution of ischemic episodes and total sleep time
in
relation to
sleep stages.
were
associated with apnea-related oxygen desatura-
predominantly in REM sleep. were not correlated to severity and extent of They the angiographic findings of the coronary vessels. Administration of a sustained-release nitrate had no significant influence on the occurrence of these episodes. Microstructure of sleep was disturbed to a greater extent in ischemic episodes than in control episodes tions and occurred
CHEST / 111 / 2 / FEBRUARY, 1997
391
Ischemic episodes lead to by activation of the CNS. than control episodes with more and severe arousals a correlation to the extent of oxygen desaturation. However, nonapnea-associated ischemic episodes as well lead to more and intense arousals than control of the CNS by silent episodes, showing an activation of ischemia independent apnea. The frequency of ST-segment depressions in pa¬ tients with obstructive sleep apnea is unclear.1213 In the study of Hanly et al,12 asymptomatic ST-segment in 7 of 23 during sleep were observedwithout depressions evi¬ obstructive with sleep apnea patients dence of coronary heart disease by three-channel Holter ECG. The authors concluded that these episodes are very common in patients with obstruc¬ tive sleep apnea and may reflect true myocardial ischemia or nonspecific changes associated with re¬ current obstructive apneas. We believe that STsegment depressions in our patients reflect true ischemia. In order for best visual analysis myocardial of the ST segments, we performed a standard sixlead chest wall ECG monitoring with a paper speed of 25 mm/s and an exact electrode placement with¬ out change during the study. ST-segment shifts after changes in body position were excluded; 85.4% of these ischemic episodes were concomitant with ob¬ structive apneas and oxygen desaturations. Myocardial oxygen supply is diminished by hypox¬ emia and may therefore be a cause of myocardial ischemia. The so-called anoxemia test with hypoxic respiration and ECG registration was first described Levy et al14 as a diagnostic method for detection by of coronary insufficiency. Obstructive apneas, how¬ ever, worsen myocardial oxygen demand by mechan¬ ical changes due to reduced intrathoracic pressure.15 These pressure and volume changes increase left ventricular preload and afterload.16 In addition, interventricular septum shift leads to impediment of and stroke
the diastolic function of the left ventricle volume.17 We defined coronary heart disease by an angiographic stenosis of at least 50%, bearing in mind that diffuse wall defects exclude not necessarily clinically significant coronary heart disease. In fact, completely normal angiograms and normal results of exercise tests were seen in only three patients. In these patients, results of the night studies also were nor¬ mal, whereas the patient with diffuse wall defects and nocturnal ischemia had an abnormal result of an exercise test. No significant difference was found between patients with and without nocturnal isch¬ emia with regard to risk factors (smoking history,
diabetes). hypertension, rate as
Heart an important determinant of myo¬ cardial oxygen consumption was significantly lower in apnea-associated ischemic episodes than in non¬
392
apnea-associated episodes and therefore did not contribute seriously to these events. In contrast, Quyyumi and coworkers18 found episodes of STsegment depressions preceded by an increase in heart rate as a result of arousal, lightening of sleep, body movements, and REM sleep in patients with coronary heart disease and nocturnal angina without evidence of obstructive sleep apnea. The strong association of ischemic episodes with apneas and oxygen desaturations suggests that these factors con¬ tribute to myocardial stress, which is not diminished by the administration of a sustained-release nitrate. In this circumstance, nasal continuous positive air¬ way pressure therapy for reversal of apneas and
oxygen desaturation may be effective to treat myo¬ cardial ischemia. Hanly and coworkers12 found a significantly reduced duration of ST-segment de¬ pression by nasal continuous airway pressure in their
study.
Of ischemic episodes, 77.8% occurred during sleep, whereas the total amount of REM sleep was only 18% of total sleep time. Obstructive apneas are longer in REM sleep; therefore, oxygen satura¬ tion falls to lower levels in this sleep state suggesting the occurrence of ischemic episodes as a result of apnea and oxygen desaturation rather than sleep stage. However, experimental data from animal stud¬ ies exhibited an increase in heart rate and coronary blood flow in REM sleep due to an increased sympathetic drive to the heart.19 Furthermore, there was a partial mismatch between the percent increase in coronary blood flow and concomitant cardiac metabolic demand in phasic REM sleep.19 In the presence of coronary stenosis, the heart rate accel¬ eration is associated with a decrease rather than an increase in coronary blood flow20 and in this circum¬ stance, myocardial ischemia may be promoted by REM
sleep. Also, however, nonapnea-associated in REM predominantly episodes occurred was described that an association by Nowlin sleep, and coworkers as well.21 Macrostructure of sleep was similar in both groups with and without coronary heart disease and in both groups, the amount of slow-wave sleep (NREM stage 3 and 4) and REM sleep was reduced. This sleep pattern with reduction in slow-wave sleep and REM sleep and increase in light sleep (NREM stage 1 and 2) is described in patients with coronary heart disease22 and in patients with obstructive sleep ap¬ nea.23 Episodes with ST-segment depression during more often associated with sleep werethansignificantly control episodes. Furthermore, the arousals quality and intensity of arousals were different in ischemic and control episodes. A second arousal more often in the former ones than in the REM
ischemic
occurred
Clinical
Investigations
indication of a disturbing effect of ischemia on microstructure of sleep. myocardial Standard sleep stage scoring systems are intended to identify state and not transient interruptions of Arousals are transient and generally do not sleep. result in behavioral awakening, but the important fact is that they result in fragmented sleep and this leads to increased daytime sleepiness. The arousal classification we used in our study was based on a on scoring rules11 since to our preliminary report no standard criteria for classification knowledge, exist. The rationale for taking the duration of alpha bursts10 into consideration was that this phenome¬ non reflects an activation of the CNS and it can be assumed that a longer alpha burst disturbs microstructure of sleep more than a shorter one does. This should be demonstrated by measurements of day¬ time sleepiness using standard vigilance tests in a next step. Since ischemic episodes were compared with control episodes, which were similar with re¬ gard to sleepin stage, apnea duration, and oxygen desaturation the same patient, it is a reasonable assumption that ischemia itself took part in an arousing stimulus and not apnea alone. In summary, patients with the combination of coronary heart disease and obstructive sleep apnea are considered as a particular group at risk, because apnea-associated oxygen desaturation can trigger nocturnal myocardial ischemia. This "silent" myocar¬ dial ischemia leads to an activation of the CNS, which may result in increased daytime sleepiness. In these patients, timely diagnosis and treatment of ischemic heart disease as well as sleep apnea are necessary. Furthermore, patients with nocturnal ischemia should be screened for underlying sleep apnea even if nitrate therapy fails.
latter ones
as an
Pepine CJ. Silent myocardial ischemia: definition, magnitude and scope of the problem. Cardiol Clin 1986; 4:577-81 2 Whitaker MP, Sheps DS. Prevalence of silent myocardial ischemia in survivors of cardiac arrest. Am J Cardiol 1989;
64:591-93 3 Schwartz PJ, La Rove re
MT, Vanoli E. Autonomic nervous system and sudden cardiac death. Circulation 1992; 85(suppl
I):77-91
Ejnell H, Sellgren J, et al. Is high and fluctuating muscle nerve sympathetic activity in the sleep apnea syn¬ drome of pathogenetic importance for the development of hypertension? J Hypertens 1988; 6:529-31
4 Hedner J,
Palta M,
Dempsey J,
et
H, Grote L, Penzel T, et al. A mobile 10-channel
(Sidas 2010) for the diagnosis of sleep-related breathing disorders. In: Peter JH, Penzel T, Podszus T, et al, eds. Sleep
unit
and health risk. Rerlin: Springer, 1991; 37-49 A, Kales A, eds. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Bethesda, Md: National Institute of Neurological Disease and Blindness, 1968 (NIH publication
9 Rechtschaffen
No.
204)
Stepanski E, Lamphere J, Badia P, et al. Sleep fragmentation and daytime sleepiness. Sleep 1984; 7:18-26 11 American Sleep Disorders Association. EEG arousals: scoring rules and examples. A preliminary report from the Sleep Disorders Atlas Task Force of the American Sleep Disorders Association. Sleep 1992; 15:174-84 12 Hanly P, Sasson Z, Zuberi N, et al. ST-segment depression during sleep in obstructive sleep apnea. Am J Cardiol 1993; 71:1341-45 13 Philip P, Guilleminault C. ST segment abnormality, angina during sleep and obstructive sleep apnea. Sleep 1993; 16: 558-59 10
14
Levy RL, Williams NE, Bruenn HG, et al. The 'anoxemia test' the diagnosis of coronary insufficiency. Am Heart J 1941;
in
21:634-56 15 Scharf SM, Graver LM, Balaban K. Cardiovascular effect of periodic occlusions of the upper airway in dogs. Am Rev Respir Dis 1992; 146:321-29 16 Tilkian AG, Guilleminault C, Schroeder JS, et al. Sleep induced apnea syndrome: hemodynamic studies during wakefulness and sleep. Ann Intern Med 1976; 85:714-19 17 Shiomi T, Guilleminault C, Stoohs R, et al. Leftward shift of
the interventricular septum and pulsus paradoxus in obstruc¬ sleep apnea syndrome. Chest 1991; 100:894-902 18 Quyyumi AA, Efthimiou J, Quyyumi A, et al. Nocturnal angina: precipitating factors in patients with coronary artery disease and those with variant angina. Br Heart J 1986; tive
56:346-52
1
Young T,
Springer, 1982; 9-18
8 Schneider
19 Dickerson LW,
References
5
sleep-disordered breathing among middle-aged adults. N Engl J Med 1993; 328:1230-35 6 Partinen M, Guilleminault C. Daytime sleepiness and vascu¬ lar morbidity at 7-year follow-up in obstructive sleep apnea patients. Chest 1990; 97:27-32 7 Holmgren A, Ivert T. Exercise protocol. In: Loogen F, Seipel L, eds. Detection of ischemic myocardium with exercise. Rerlin:
al. The
occurrence
of
20 21
Huang AH, Thurnher MM, et al.
Relation¬
ship between coronary7 hemodynamic changes and the phasic events of rapid eye movement sleep. Sleep 1993; 16:550-57 Verrier LR, Dickerson LW. Autonomic nervous system and coronary blood flow changes related to emotional activation and sleep. Circulation 1991; 83(suppl 2):81-9 Nowlin JB, Troyer WG, Collins WS, et al. The association of nocturnal angina with dreaming. Ann Intern Med 1965; 63:1040-46
22 Cassano
GB, Maggini C, Guazzelli M. Nocturnal angina and
sleep. Prog Neuropsychopharmacol 1981; 5:99-104 23 Issa FG, Sullivan CE. The immediate effects of nasal contin¬ uous positive airway pressure treatment on sleep pattern in patients with obstructive sleep apnea syndrome. Electroenceph Clin Neurophysiol 1986; 63:10-7 '
CHEST / 111 / 2 / FEBRUARY, 1997
393
Harold
Jorg
Schdfer, MD; Ulrich Koehler, MD; Thomas Ploch; and
Hermann Peter, MD
Study objectives: Patients with coronary heart disease (CHD) and obstructive sleep apnea may have an increased cardiac risk due to nocturnal myocardial ischemia triggered by apneaassociated oxygen desaturation. Sleep structure in patients with obstructive sleep apnea is system (CNS) (arousal) due to obstructive apneas. fragmented by activation of the central nervous Nocturnal myocardial ischemia may lead to activation of the CNS as well. Patients: Fourteen patients with obstructive sleep apnea and CHD disease and seven patients suffering from obstructive sleep apnea without CHD were studied. Overnight sleep studies and simultaneous six-lead ECG recordings were performed. In addition, sleep studies and ECG recordings were performed with administration of a sustained-release nitrate in these patients in a double-blinded crossover design. Results: Analysis of three nights' recordings revealed 144 episodes of nocturnal myocardial ischemia in six subjects. Five patients had underlying CHD and one patient exhibited diffuse wall defects of the coronary arteries; also, 85.4% of ischemic episodes were concomitant with apneas and oxygen desaturation >3%, and 77.8% of ischemicepisodes occurred during rapid eye movement (REM) sleep, although total amount of REM sleep was only 18% of total sleep time. Mean oxygen saturation was significantly lower (p<0.05) during apnea-associated ischemic episodes than during nonapnea-associated ischemia (77.3% vs 93.1%). Nitrate administration did not reduce ischemic episodes. Sleep architecture (macrostructure) exhibited a reduction in sleep stages non-REM 3 and 4 and REM sleep. Comparing the microstructure of sleep (arousals) within episodes with and without ischemia but similar criteria like sleep stage, apnea activity, and oxygen saturation, we found significantly more (p<0.01) and severe (p<0.001) arousals during periods with myocardial ischemia than during control episodes. In addition, microstructure of sleep was disturbed by myocardial ischemia itself in absence of apneas. Conclusion: It is concluded that patients with CHD and obstructive sleep apnea are endangered by apnea-associated ischemia and that these ischemic episodes lead to activation of the CNS and additional fragmentation of sleep. Patients with nocturnal ischemia should be screened for (CHEST 1997; 111:387-93) underlying sleep apnea even if nitrate therapy fails. Key words: arousal; coronary heart disease; nocturnal myocardial ischemia; sleep apnea; sleep structure Abbreviations:
NREM=nonrapid eye movement; REM=rapid eye movement
¥7* xperience with ambulatory ECG monitoring in the last 20 years indicated that most ischemic episodes in patients with coronary heart disease were asymptomatic.1 Recent clinical studies support the -*-J
*From the Department of Internal Medicine, University of Bonn, Bonn, Germany (Dr. Schafer), and the Department of Internal Medicine, Sleep Disorders Center, University of Marburg, (Drs. Koehler, Ploch, and Peter). Marburg, Germany received Mav 3, 1996; revision accepted Septem¬ Manuscript ber 30. Reprint requests: Dr. Schafer, Medizinische Universitatsklinik und Poliklinik Bonn, Innere Medizin/Kardiologie und Pneumologie, Sigmund-Freud-Str. 25, D-53105 Bonn, Germany
relationship between silent myocardial ischemia and sudden death.2 Some believe that the autonomic nervous system triggers such events in these pa¬ tients.3 Obstructive sleep apnea, a condition with increased activation of the autonomic nervous sys¬ tem4 that is frequent in middle-aged men,5 shares many risk factors with coronary heart disease. Un¬ treated patients have a higher, predominantly car¬ diovascular mortality.6 Patients with the combination of coronary heart disease and obstructive sleep apnea may have an increased cardiac risk due to worsening of the relationship between myocardial oxygen deCHEST / 111 / 2 / FEBRUARY, 1997
387
of apnea-associated supply as a result autonomic nervous of and activation the hypoxemia In system. addition, sleep quality may be influenced by apnea-associated activation of theTheCNS (arousals)of and ischemia-associated arousals. objective this study was to determine the association of noc¬ turnal myocardial ischemia and sleep apnea in pa¬ tients with coronary heart disease and obstructive apnea. In addition, macrostructure and microsleep structure of sleep were analyzed with regard to ischemic episodes and the effect of nitrates on ischemic episodes and sleep structure was tested.
mand and
Materials We studied 21 male
and
Methods
patients with obstructive sleep apnea
consecutive nights in the sleep labora¬ (apnea index ^10) in sixhad a combination of obstructive sleep Fourteen patients apnea and coronaiy heart disease diagnosed by coronary angiog¬ raphy (group 1). Seven patients exhibited obstructive sleep apnea without underlying coronary heart disease as demonstrated by coronary angiography (group 2). Indications for cardiac catheter¬ ization were history of chest pain or pathologic findings in exercise ECG. In addition to the usual sleep and cardiorespira¬ tory measurements, all patients had six-lead ECG monitoring were weaned off treatment with all during sleep. Patients calcium-channel blockers, and P-blockers medication, antianginal at least 48 h prior to the first polysomnography. In a double-blind crossover design, patients were administered placebo-controlled a single dose of a sustained-release nitrate or placebo at the and 5,6 to test the effect beginning of the sleep study in nightsof 3,4 of nitrate therapy for suppression nocturnal ischemia. There¬ fore, we chose a dose at the upper limit of a single-dose sustained-release nitrate (100 mg isosorbide dinitrate). Evalua¬ tion of polysomnographies was based on the data obtained in the second night (without medication), fourth night (placebo or active medication), and sixth night (placebo or active medication).
tory.
Informed consent was obtained from all patients.
Exercise Test
Before sleep studies, all patients underwent a maximal graded in supine position. A minimum load of 50 W bicycle stress testwhich was increased by 25 W every 2 min.7 was performed, Twelve-lead ECGs were printed out for each minute during
exercise and for the first 5 min of recovery or until the baseline was achieved. An ischemic ECG response was defined as > 1 mm horizontal or downsloping ST-segment depression below
state
the PR segment for >0.08 s after the J point on at least three consecutive beats. The exercise test was stopped at the onset of ischemic ECG changes with or without symptoms or when the target heart rate was reached.
Coronary Arteriography
Both left and right coronary arteriograms were performed in all patients using the standard Judkins approach and stenotic lesions
considered significant if there was >50% diameter stenosis. Diffuse wall defects of the coronary vessels were considered not significant if the remaining, patent diameter was at least 50%.
were
Sleep
Studies two-channel EEG (C3-A2, C3-02), electroand submental electromyogram using surface elec¬
We obtained
oculogram, 388
a
trodes. Airflow was detected using a nasal thermistor. Arterial oxygen saturation was recorded with an ear oximeter. Respiratory effort was monitored by inductive plethysmography of the chest wall and abdomen. The ECG and heart rate were recorded from standard limb leads. Body movements were detected by means of an actigraph on the forearm. All parameters were registered
continuously by a sleep system (SIDAS 2010 sleep-system; Stimotron; Wendelstein, Germany) and were converted from analog to digital signals at a sampling rate of 100 Hz. The method has been described elsewhere in detail.8 The digital signals were sent to an integrated personal computer, which both presented the data on-line on an ink jet printer and screen display and stored the data at a sampling rate of 1 to 100 Hz for each signal on a hard disk in order to perform an off-line analysis. Scoring Rules
All sleep studies were scored manually and sleep was staged in intervals of 20 s duration by the standard guidelines of Rechtschaffen and Kales.9 We used a modified classification of arousal scoring on the basis of Stepanski et al10 and from the Sleep Disorders Atlas Task Force of the American Sleep Disor¬ ders Association.11 Arousals were defined as abrupt shifts in EEG frequency following a minimum of 10 continuous seconds of intervening sleep. They were assigned to three categories: micro, gross, and complete arousals, whereas awakenings represented a separate category. Arousals were classified as microarousals when the criteria of the general definition of an arousal were found but no alpha activity was present. In a gross arousal, alpha-bursts 3 s at least. In a complete arousal, appeared for <15 s but were defined by alpha-bursts persisted for >15 s.^60Awakenings s. Alpha intrusion in rapid persistence of alpha activity for eye movement (REM) sleep was not scored unless it was activ¬ accompanied by an increase in submental electromyogram ity. Obstructive apnea was defined as cessation of airflow for at least 10 s in the presence of continued respiratory efforts. The apnea index was calculated from the number of apneas per hour of sleep.
Recordings A continuous recording of a six-lead chest wall ECG (MingoMunich, Germany) was performed and was graph; Siemens; out at a paper speed synchronized to the sleep study andofprinted of 25 mm/s. Standard placements the six-lead chest wall ECG were used (V1-V6). ST-segment depressions of >0.1 mV below the isoelectric line for >0.08 s after the J point with horizontal and descending courses and changes of theT wave, such as T inversion, were considered for myocardial ischemia. An ischemic documented only if its minimum duration was 10 s. episode was Isolated T-wave changes and ST-segment shifts occurring after change in body position were excluded. Episodes of myocardial ischemia were examined specifically with respect to the interac¬ tion between respiration (apnea, hyperventilation, oxygen satu¬ ration) and changes in the EEG pattern described previously. Ischemic episodes were compared with "control episodes" in the same patients with regard to arousal pattern. These control as periods without myocardial ischemia episodes were defined but with similar criteria like apnea activity, oxygen saturation, and sleep stage with close proximity to the ischemic events. ECG
Statistical Analysis
Descriptive statistics for continuous variables were expressed the mean ± SD. Differences in mean values were assessed using Student's two-tailed t test and analysis of variance. The x2 as
Clinical
Investigations
test with Yates' correction was used for categorical variables. In all cases, p values of <0.05 were considered to be significant.
Results
The mean age of the 21 patients was 56.3 years (47 67 years) and the mean weight was 89.4 kg (72 to 129 kg). None of the subjects had a clinical history of myocardial infarction. Mean Po2 during wakefulness was 81.9 mm Hg (64.6 to 100.4 mm Hg). The results of coronary arteriograms are listed in Table 1.
to
Respiration and General Sleep Structure Concerning the respiratory parameters, no signif¬
differences among the three recording nights found with regard to apnea index, oxygen saturation, and heart rate (Table 2). Furthermore, no significant differences in sleep stages between the recording nights and between the groups of patients with and without coronary heart disease were found. On the average, of all 21 patients, the amount of sleep stages 1 and 2 was 67.1% of the total sleep time and the amount of sleep stages 3 and 4 was 14.9%. Eighteen percent of total sleep time was spent in REM sleep. Sleep efficiency was 90.5% of total sleep time. Patients with coronary heart disease (group 1) spent an average of 66.8% in sleep stages 1/2, 15.2% in 3/4, and 18% in REM sleep. Patients without icant
were
Table Patient 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
No.1
coronary heart disease (group 2) spent 67.0% in stages 1/2, 14.5% in 3/4, and 18.5% in REM sleep.
Myocardial Ischemia Episodes with nocturnal myocardial ischemia were found in 6 of these 21 patients. In four of the six patients with nocturnal myocardial ischemia, one single coronary vessel was affected in each patient and one
in patient two coronary vessels were af¬ fected. One patient exhibited diffuse wall defects of the coronary vessels. A total of 144 ischemic episodes with a maximum duration of 96 s (10 to 96 s) was recorded within three nights (Table 3). Nitrate ad¬ ministration was without effect with regard to num¬ ber and mean duration of ischemic episodes (Table 3). In the following, all the ischemic episodes of all three nights were pooled. In two patients, only few ischemic episodes occurred (patient 9, four episodes; patient 11, one episode), whereas in the others, more episodes occurred (patient 1, 34 episodes; patient 3,
episodes;These patient 14, 31 episodes; patient 15, 59 episodes). episodes appeared predominantly during periodsin of high apnea activity and oxygen desaturation, the minority within periods of un¬ disturbed ventilation. Of the 144 ischemic episodes, 123 were concomitant with and 15
(85.4%)
oxygen desaturation (>3%). In the
apneas
remaining
21
(14.6%), change breathing pattern or episodessaturation was evident. no
oxygen
in
1.Findings of Coronary Arteriography, Exercise ECG, and Nocturnal Sa02*
Nocturnal Ischemia
Minimal
Sa02
During Sleep, °i <60 64 <60
0
+
75 79 <60 75 75 75 <60 70 <60 80 60
0 0 0 0 0
77
70 78 72 80 80 84
No. of Affected Coronary Arteries (>50%)
Maximal Work Load (Watts) of Exercise ECG Findings1
1 3 1 1 3 3 1 2 1 2 1 2 2 2
No CHD, diffuse wall defects No CHD No CHD No CHD, diffuse wall defects No CHD No CHD No CHD, diffuse wall defects
No No No No
125 100 125 75 50 150 50 75 75 100 75 125 50 100 125
findings findings findings findings
*Sa02=arterial oxygen saturation; CHD=coronary heart disease. fThe first 14 patients represent group 1 (with coronary heart disease). The following seven patients represent group 2 (without coronary heart disease). "The maximal work load (watts)
at
which
pathologic changes of the ST segment occurred is determined.
CHEST / 111 / 2 / FEBRUARY, 1997
389
Table
2.Respiratory Parameters of the Patients With Nocturnal Myocardial Ischemia*
Respiratory Parameters (n=21 Patients)
Apnea index
Mean SD Min Max Duration of apneas, s Mean SD Min Max Mean arterial oxygen saturation during sleep, % Mean SD Min Max Minimal arterial oxygen saturation during sleep, % Mean SD Min Max Maximal drop of arterial oxygen saturation during sleep, Mean SD Min Max Mean heart rate during sleep, bpm Mean SD Min Max
Minimal heart rate Mean SD Min Max
during sleep, bpm
Maximal heart rate during sleep, Mean SD Min Max
bpm
Night 2
Night 4
Night 6
(Baseline)
(Placebo)
(Nitrate)
32.3 11.8 13.0 50.0
30.4 15.0 11.0 59.0
36.9 19.0 12.0 80.0
63.6 26.6 30.0 120.0
58.3 22.2 30.0 100.0
61.4 24.7 20.0 120.0
90.0 2.0 85.0 93.0
90.5 2.8 83.0 94.0
89.8 3.0 83.0 94.0
73.8
10.4 46.0 88.0
74.6 10.7 50.0 88.0
71.4 13.0 45.0 89.0
19.3 9.2 6.0 42.0
17.7 8.3 9.0 38.0
20.9 10.1 8.0 40.0
67.7 8.6 50.0 82.0
65.5 7.3 46.0 78.0
68.8
46.0 94.0
49.5 5.0 38.0 57.0
48.5 6.0 38.0 61.0
50.3 7.0 37.0 62.0
100.3 11.7 84.0 132.0
97.8 10.3 75.0 121.0
103.5 14.9 82.0 142.0
10.7
2 (without medication), night 4 (with placebo), and night 6 (with sustained *bpm=beats per minute. Evaluation was based on the data of night nitrate). Data are expressed in mean value with SD, minimum value (min), and maximum value (max).
Mean oxygen saturation was 91.0% at the onset of the 123 apnea-associated myocardial ischemic epi¬ sodes vs 77.3% during ischemia. Heart rate was monitored between 70 beats/min before and 77 beats/min during the ischemic episode. In the 21 ischemic episodes, oxygen sat¬ nonapnea-associated uration was 93.1% before and 92.9% during isch¬ emia. Heart rate was 75 beats/min before and 94 beats/min during ischemia. The differences between and nonapnea-associated ischemic apnea-associated in mean values and changes of oxygen the episodes saturation (mean change apnea associated, 13.6±10.3%; nonapnea associated, 0.3±0.7%; 390
and heart rate (mean change apnea associ¬ p<0.05) 7.4±16.5 beats/min; nonapnea associated: ated, -19.8±14.5 beats/min; p<0.001) were statistically significant. One hundred twelve ischemic episodes (77.8%) occurred during REM sleep, 10 (6.9%) in non-REM (NREM) stage 1, 20 (13.9%) in NREM stage 2, and two (1.4%) in NREM stages 3/4 (Fig 1). No statisti¬ cal differences were found between apnea-associ¬ ated and nonapnea-associated ischemic episodes with regard to sleep stage. Roth groups of ischemic episodes were frequent during REM sleep. Apneaassociated ischemic episodes occurred in 78.1% in Clinical
Investigations
Table 3.Number and Duration of Ischemic Episodes* Nocturnal
Myocardial Ischemia
(n=6 Patients)
No. of ischemic episodes Duration of ischemic episodes,
*
Night 2
Night 4 Night 6
(Baseline) (Placebo) (Nitrate)
Association*
45 40 59 s
29.3 Mean SD17.5 Min12.0 96.0 Max
28.2 13.4 12.0
73.0
27.7 16.3 10.0 91.0
Evaluation was based on the data of night 2 (without medication), night 4 (with placebo), and night 6 (with sustained nitrate). Data are expressed in mean value with SD, minimum value (min), and maximum value
Table 4.Comparison of Ischemic Episodes and Control Episodes With Regard to Arousal Association, Arousal Specification, Sleep Stage, and Apnea
(max).
Association +
0 Sum
Arousal type Microarousal Gross arousal Complete arousal
Awakening
sleep and nonapnea-associated ischemic epi¬ sodes occurred in 76.2% in REM sleep. Ischemic periods were asymptomatic in all patients. Microstructure of Sleep A total of 175 arousals were recorded during the 144 myocardial ischemic episodes. Seven ischemic periods (4.9%) revealed no arousal, and 137 (95.1%) resulted in at least one. Of these 137 ischemic episodes, 99 events (72.3%) were concomitant with one arousal and 38 events were REM
(27.7%)
concomitant
with a second arousal. During the 144 control epi¬ sodes, only 127 arousals were recorded (p<0.01). In 29 control episodes (20.1%), no arousal was de¬ tected. Of the 127 episodes, 115 events (90.5%) were concomitant with one arousal and 12 events (9.5%) were concomitant with a second one. Comparing the intensity of arousals, significantly more severe ones (complete arousals, p<0.001) were detected during ischemic episodes than during control episodes (Ta¬ ble 4). A statistically significant correlation (p<0.01; r=0.5) was found between oxygen saturation before
Sum
*
Control
Ischemic
Arousal
Episodes,
(%) 137(95.1) 7 (4.9) 144 (100)
p Value
37(21.1)
NS NS <0.001 NS <0.01
No.
57 (32.6) 75 (42.9)
6 (3.4) 175(100)
<0.01 <0.01 NS
Significant differences between ischemic and control indicated by p values.
Episodes,
(%) 115 (79.9) 29(20.1) 144 (100) No.
47 (37.0)
62 (48.8) 18 (14.2)
0(0)
127 (100)
episodes are
and during the ischemic episodes and the severity of arousals. The differences in oxygen saturation before and during ischemic episodes were 1.4% when no arousal occurred, 10.2% in microarousals, 11.4% in gross arousals, and 14.5% in complete arousals. Only 3.7% was the difference in awakenings. No statistically significant correlation was found between heart rate before and during the ischemic and arousals. episodes Microstructure of sleep during the 21 ischemic not apnea associated was signifi¬ episodes that weredifferent from control episodes. cantly (p<0.01) During ischemic episodes, only in 19% no arousal was detected, whereas in 71.4% of control episodes, no arousal was detected. Classification of arousals revealed in all of the control episodes microarousals, whereas in ischemic episodes, 40% were gross arous¬
als, 35% were complete arousals, 15% croarousals, and 10% were awakenings.
were
mi¬
Discussion
asymptomatic nocturnal ST-segment depression patients with obstructive sleep apnea in those patients with additional coronary heart only disease and in one patient with diffuse coronary vessel defects. The vast majority of these episodes We found
in
Sleep stages
BI Sleep stages H REM SLEEP
1/2 3/4
Total Sleep Time Ischemic Episodes
Figure 1. Distribution of ischemic episodes and total sleep time
in
relation to
sleep stages.
were
associated with apnea-related oxygen desatura-
predominantly in REM sleep. were not correlated to severity and extent of They the angiographic findings of the coronary vessels. Administration of a sustained-release nitrate had no significant influence on the occurrence of these episodes. Microstructure of sleep was disturbed to a greater extent in ischemic episodes than in control episodes tions and occurred
CHEST / 111 / 2 / FEBRUARY, 1997
391
Ischemic episodes lead to by activation of the CNS. than control episodes with more and severe arousals a correlation to the extent of oxygen desaturation. However, nonapnea-associated ischemic episodes as well lead to more and intense arousals than control of the CNS by silent episodes, showing an activation of ischemia independent apnea. The frequency of ST-segment depressions in pa¬ tients with obstructive sleep apnea is unclear.1213 In the study of Hanly et al,12 asymptomatic ST-segment in 7 of 23 during sleep were observedwithout depressions evi¬ obstructive with sleep apnea patients dence of coronary heart disease by three-channel Holter ECG. The authors concluded that these episodes are very common in patients with obstruc¬ tive sleep apnea and may reflect true myocardial ischemia or nonspecific changes associated with re¬ current obstructive apneas. We believe that STsegment depressions in our patients reflect true ischemia. In order for best visual analysis myocardial of the ST segments, we performed a standard sixlead chest wall ECG monitoring with a paper speed of 25 mm/s and an exact electrode placement with¬ out change during the study. ST-segment shifts after changes in body position were excluded; 85.4% of these ischemic episodes were concomitant with ob¬ structive apneas and oxygen desaturations. Myocardial oxygen supply is diminished by hypox¬ emia and may therefore be a cause of myocardial ischemia. The so-called anoxemia test with hypoxic respiration and ECG registration was first described Levy et al14 as a diagnostic method for detection by of coronary insufficiency. Obstructive apneas, how¬ ever, worsen myocardial oxygen demand by mechan¬ ical changes due to reduced intrathoracic pressure.15 These pressure and volume changes increase left ventricular preload and afterload.16 In addition, interventricular septum shift leads to impediment of and stroke
the diastolic function of the left ventricle volume.17 We defined coronary heart disease by an angiographic stenosis of at least 50%, bearing in mind that diffuse wall defects exclude not necessarily clinically significant coronary heart disease. In fact, completely normal angiograms and normal results of exercise tests were seen in only three patients. In these patients, results of the night studies also were nor¬ mal, whereas the patient with diffuse wall defects and nocturnal ischemia had an abnormal result of an exercise test. No significant difference was found between patients with and without nocturnal isch¬ emia with regard to risk factors (smoking history,
diabetes). hypertension, rate as
Heart an important determinant of myo¬ cardial oxygen consumption was significantly lower in apnea-associated ischemic episodes than in non¬
392
apnea-associated episodes and therefore did not contribute seriously to these events. In contrast, Quyyumi and coworkers18 found episodes of STsegment depressions preceded by an increase in heart rate as a result of arousal, lightening of sleep, body movements, and REM sleep in patients with coronary heart disease and nocturnal angina without evidence of obstructive sleep apnea. The strong association of ischemic episodes with apneas and oxygen desaturations suggests that these factors con¬ tribute to myocardial stress, which is not diminished by the administration of a sustained-release nitrate. In this circumstance, nasal continuous positive air¬ way pressure therapy for reversal of apneas and
oxygen desaturation may be effective to treat myo¬ cardial ischemia. Hanly and coworkers12 found a significantly reduced duration of ST-segment de¬ pression by nasal continuous airway pressure in their
study.
Of ischemic episodes, 77.8% occurred during sleep, whereas the total amount of REM sleep was only 18% of total sleep time. Obstructive apneas are longer in REM sleep; therefore, oxygen satura¬ tion falls to lower levels in this sleep state suggesting the occurrence of ischemic episodes as a result of apnea and oxygen desaturation rather than sleep stage. However, experimental data from animal stud¬ ies exhibited an increase in heart rate and coronary blood flow in REM sleep due to an increased sympathetic drive to the heart.19 Furthermore, there was a partial mismatch between the percent increase in coronary blood flow and concomitant cardiac metabolic demand in phasic REM sleep.19 In the presence of coronary stenosis, the heart rate accel¬ eration is associated with a decrease rather than an increase in coronary blood flow20 and in this circum¬ stance, myocardial ischemia may be promoted by REM
sleep. Also, however, nonapnea-associated in REM predominantly episodes occurred was described that an association by Nowlin sleep, and coworkers as well.21 Macrostructure of sleep was similar in both groups with and without coronary heart disease and in both groups, the amount of slow-wave sleep (NREM stage 3 and 4) and REM sleep was reduced. This sleep pattern with reduction in slow-wave sleep and REM sleep and increase in light sleep (NREM stage 1 and 2) is described in patients with coronary heart disease22 and in patients with obstructive sleep ap¬ nea.23 Episodes with ST-segment depression during more often associated with sleep werethansignificantly control episodes. Furthermore, the arousals quality and intensity of arousals were different in ischemic and control episodes. A second arousal more often in the former ones than in the REM
ischemic
occurred
Clinical
Investigations
indication of a disturbing effect of ischemia on microstructure of sleep. myocardial Standard sleep stage scoring systems are intended to identify state and not transient interruptions of Arousals are transient and generally do not sleep. result in behavioral awakening, but the important fact is that they result in fragmented sleep and this leads to increased daytime sleepiness. The arousal classification we used in our study was based on a on scoring rules11 since to our preliminary report no standard criteria for classification knowledge, exist. The rationale for taking the duration of alpha bursts10 into consideration was that this phenome¬ non reflects an activation of the CNS and it can be assumed that a longer alpha burst disturbs microstructure of sleep more than a shorter one does. This should be demonstrated by measurements of day¬ time sleepiness using standard vigilance tests in a next step. Since ischemic episodes were compared with control episodes, which were similar with re¬ gard to sleepin stage, apnea duration, and oxygen desaturation the same patient, it is a reasonable assumption that ischemia itself took part in an arousing stimulus and not apnea alone. In summary, patients with the combination of coronary heart disease and obstructive sleep apnea are considered as a particular group at risk, because apnea-associated oxygen desaturation can trigger nocturnal myocardial ischemia. This "silent" myocar¬ dial ischemia leads to an activation of the CNS, which may result in increased daytime sleepiness. In these patients, timely diagnosis and treatment of ischemic heart disease as well as sleep apnea are necessary. Furthermore, patients with nocturnal ischemia should be screened for underlying sleep apnea even if nitrate therapy fails.
latter ones
as an
Pepine CJ. Silent myocardial ischemia: definition, magnitude and scope of the problem. Cardiol Clin 1986; 4:577-81 2 Whitaker MP, Sheps DS. Prevalence of silent myocardial ischemia in survivors of cardiac arrest. Am J Cardiol 1989;
64:591-93 3 Schwartz PJ, La Rove re
MT, Vanoli E. Autonomic nervous system and sudden cardiac death. Circulation 1992; 85(suppl
I):77-91
Ejnell H, Sellgren J, et al. Is high and fluctuating muscle nerve sympathetic activity in the sleep apnea syn¬ drome of pathogenetic importance for the development of hypertension? J Hypertens 1988; 6:529-31
4 Hedner J,
Palta M,
Dempsey J,
et
H, Grote L, Penzel T, et al. A mobile 10-channel
(Sidas 2010) for the diagnosis of sleep-related breathing disorders. In: Peter JH, Penzel T, Podszus T, et al, eds. Sleep
unit
and health risk. Rerlin: Springer, 1991; 37-49 A, Kales A, eds. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Bethesda, Md: National Institute of Neurological Disease and Blindness, 1968 (NIH publication
9 Rechtschaffen
No.
204)
Stepanski E, Lamphere J, Badia P, et al. Sleep fragmentation and daytime sleepiness. Sleep 1984; 7:18-26 11 American Sleep Disorders Association. EEG arousals: scoring rules and examples. A preliminary report from the Sleep Disorders Atlas Task Force of the American Sleep Disorders Association. Sleep 1992; 15:174-84 12 Hanly P, Sasson Z, Zuberi N, et al. ST-segment depression during sleep in obstructive sleep apnea. Am J Cardiol 1993; 71:1341-45 13 Philip P, Guilleminault C. ST segment abnormality, angina during sleep and obstructive sleep apnea. Sleep 1993; 16: 558-59 10
14
Levy RL, Williams NE, Bruenn HG, et al. The 'anoxemia test' the diagnosis of coronary insufficiency. Am Heart J 1941;
in
21:634-56 15 Scharf SM, Graver LM, Balaban K. Cardiovascular effect of periodic occlusions of the upper airway in dogs. Am Rev Respir Dis 1992; 146:321-29 16 Tilkian AG, Guilleminault C, Schroeder JS, et al. Sleep induced apnea syndrome: hemodynamic studies during wakefulness and sleep. Ann Intern Med 1976; 85:714-19 17 Shiomi T, Guilleminault C, Stoohs R, et al. Leftward shift of
the interventricular septum and pulsus paradoxus in obstruc¬ sleep apnea syndrome. Chest 1991; 100:894-902 18 Quyyumi AA, Efthimiou J, Quyyumi A, et al. Nocturnal angina: precipitating factors in patients with coronary artery disease and those with variant angina. Br Heart J 1986; tive
56:346-52
1
Young T,
Springer, 1982; 9-18
8 Schneider
19 Dickerson LW,
References
5
sleep-disordered breathing among middle-aged adults. N Engl J Med 1993; 328:1230-35 6 Partinen M, Guilleminault C. Daytime sleepiness and vascu¬ lar morbidity at 7-year follow-up in obstructive sleep apnea patients. Chest 1990; 97:27-32 7 Holmgren A, Ivert T. Exercise protocol. In: Loogen F, Seipel L, eds. Detection of ischemic myocardium with exercise. Rerlin:
al. The
occurrence
of
20 21
Huang AH, Thurnher MM, et al.
Relation¬
ship between coronary7 hemodynamic changes and the phasic events of rapid eye movement sleep. Sleep 1993; 16:550-57 Verrier LR, Dickerson LW. Autonomic nervous system and coronary blood flow changes related to emotional activation and sleep. Circulation 1991; 83(suppl 2):81-9 Nowlin JB, Troyer WG, Collins WS, et al. The association of nocturnal angina with dreaming. Ann Intern Med 1965; 63:1040-46
22 Cassano
GB, Maggini C, Guazzelli M. Nocturnal angina and
sleep. Prog Neuropsychopharmacol 1981; 5:99-104 23 Issa FG, Sullivan CE. The immediate effects of nasal contin¬ uous positive airway pressure treatment on sleep pattern in patients with obstructive sleep apnea syndrome. Electroenceph Clin Neurophysiol 1986; 63:10-7 '
CHEST / 111 / 2 / FEBRUARY, 1997
393