Effect of Clonidine in Obstructive Sleep Apnea 1 •2
FAIQ G. ISSA
Introduction
Obstructive sleepapnea (OSA) is a common disorder affecting 1 to 4% of the male population (1). The disorder is characterized by repetitive closure of the pharyngeal airway during sleep because of an imbalance in the forces acting on the upper airway (2). The dilating forces generated by upper airway muscles (e.g., genioglossus) during sleep fail to counterbalance the constricting luminal forces generated by inspiratory thoracic muscles. Sleep per se induces a variety of changes in the respiratory system that enhance closure of the upper airway in susceptible subjects. For example, compared with the diaphragm, the activity of upper airway muscles preferentially decreases during sleep (3). Furthermore, sleep reduces the overall ventilatory response to hypoxia, hypercapnia, and other respiratory reflexes, thus reducing neural output to both sets of muscles (4-6). At low chemical drive and reduced neural output to the upper airway and thoracic pump muscles, closure of the upper airway is enhanced when chemical drive returns to normal levels (7). These specific effects of sleep on the respiratory system become more pronounced during rapid eye movement (REM) sleep. There is a profound hypotonia of upper airway muscles and a greater reduction in the ventilatory responses to hypoxia, hypercapnia, and mechanical loading (4-6). Furthermore, once upper airway closure occurs, the arousal threshold to asphyxia is greatly reduced during this sleep stage. Thus, during REM sleep higher levels of hypoxemia, hypercapnia, and acidemia are tolerated (8). Currently, the treatment of choice for OSA is nasally applied continuous positive airway pressure (CPAP), introduced in 1981 by Sullivan and coworkers (9). Nasal CPAP pneumatically splints the upper airway during sleep. However, this treatment is not tolerated by every patient, and compliance in most sleep centers is estimated to be approximately 700/0 (10). Clinical experience indicates that nonrespondents tend to be either pa-
SUMMARY The current treatment of choice for obstructive sleep apnea is continuous positive airway pressure. However, not all patients tolerate this form of therapy. We eveluated the effect of clonldlne hydrochloride, an a,-adrenerglc agonist with REM-suppressant activity, In eight male patients with obstructive sleep apnea. In each patient, sleep-stsge distribution and breathing pattern In two ali-night sleep studies performed during a 10-day course of clonldlne _re compared with those of two control and two placebo nights. A dose of 0.2 mg of clonldlne administered orally at bedtime totally suppreased REM sleep In two patients. In the other six patients, the same dose decreased percent time spent In REM sleep from a control of 13.4 ± 1.0 to 8.6 ± 1.4% (mean ± SEM, p < 0.05). The latency to REM sleep Increased In the latter group from a control of 129 ± 9 to 308 ± 24 min (p < 0.001). Clonldlne had no effect on the frequency and duration of non·REM breathing abnormalities. Und9r clonldlne, the level of nocturnal hypoxemia Improved In six patients. This was due to a totsl suppression of REMand the consequent lack of REMapneas In two patients. In four patients, upper airway obstruction disappeared during periods of unsuppressed REM sleep, and Sao, remained above 90% throughout this sleep stsge. Clonldlne transformed the pattern of sleep-disordered breathing during unsuppressed REMIn the other two patients from that of repetitive obstructive hypopneas associated with persistent hypoxemia to occlusive apnea and cyclical hypoxemia. These results _re observed consistently In all patients during both clonldlne-sleep studies. The data suggest that clonldlne does not only alter mechanisms Involved In the Initiation and maintenance of REM sleep but It also Influences breathing pattern during REM sleep. AM REV RESPIR DIS 1992; 145:435-439
tients with mild OSA or asymptomatic OSA. Patients wereassignedat random to one patients with otherwise severe OSA. In of two treatment schedules: placebo followed both circumstances, treatment is essen- by clonidine, or clonidine followed by placetial since retrospective studies indicate bo. In the clonidine-treatment phase, 0.20mg that survival rate decreases in patients . was taken orally at bedtime for 10 days. A similar schedule was followed during the who are not treated (11). placebo-treatment phase. Between the two Clonidine is an uradrenoreceptor ago- treatment phases, a period of 1 wk without nist that was introduced approximately treatments was instituted. A total of six all25 yr ago for the treatment of hyperten- night sleep studies were performed in each sion. The antihypertensive effects of patient. A control study was carried out beclonidine are centrally mediated, acting fore the commencement of each treatment mainly in the lower brainstem region (12). phase, and two studies were performed durClonidine is rapidly and. completely ab- ing each treatment phase: on Days 1, 2, or sorbed after an oral dose and has a Tv, 3 and on Days 7, 8, 9, or 10 of clonidine or placebo treatment. Clonidine blood levels of 5 to 15 h, and its peak plasma levels were not monitored. The protocol was apafter oral dosing occur in 60 to 90 min proved by the Ethics Review Committee of (13). Clonidine produces dryness of the Foothills Hospital and the University of nose and mouth and sedation when tak- Calgary. en during the daytime. Clonidine has a REM-suppressant activity, and has been previously used in narcolepsy (14-16). In this study, we carried out a trial with (Received in original form April 3. 1991 and in clonidine hydrochloride and placebo in revised form August 7, 1991) eight male patients with OSA. WeexamFrom the Faculty of Medicine, University of ined the effect of clonidine on respira- Calgary and Foothills Hospital, Calgary, Alberta, tion and distribution of sleep stages. Canada. 1
Methods Protocol A placebo-drug trial was used to evaluate the effect of clonidine in eight male patients with
2 Correspondence and requests for reprints should be addressed to Faiq G. Issa, M.D., Ph.D., Assistant Professor, Faculty of Medicine, University of Calgary and Foothills Hospital, 3330 Hospital Drive, NW., Calgary, Alberta, Canada T2N
4Nl.
435
436
Subjects Patients with OSA presenting with a high resting blood pressure (untreated or treated with antihypertensive drugs) were excluded from the study. Lung function and arterial blood gases weremeasured during wakefulness. The presence of obstructive apneas was confirmed polysomnographically in a diagnostic study conducted prior to the present protocol. Polysomnography All-night sleep studies were performed in the Sleep Laboratory at the Foothills Hospital. Surface electrodes recorded an electrocardiogram (EKG), two electroencephalograms (EEG), two electrooculograms (EOG), and a chin muscle electromyogram (EMG). Sao, was measured continuously with ear oximeter (Ohmeda 3700; Ohmeda, Louisville, CO). Oronasal airflow was monitored with either a thermistor taped close to the nostrils and lips or by measuring end-tidal CO 2 with an infrared capnometer (Datex; Puritan-Bennett Corp, Los Angeles, CA). Chest and abdominal wall movement was detected using an inductance plethysmograph (Respitrace'[; Ambulatory Monitoring, Ardsley, NY). All signals were recorded on a 16-channel recorder (Gould ESloo0; Gould Instruments, Cleveland, OH, or Model 6; Grass Instruments, Quincy, MA). Body position was monitored using an infrared camera positioned over the bed. Body position was marked on the polygraph record by the technician every 15to 20 min, and changes in posture were marked as they occurred. The patient was allowed to sleep undisturbed throughout the night. Two criteria were followed in all studies. First, the patient was allowed to maintain a posture most preferable to him during these studies. Second, all sleep studies commenced at the usual bedtime for each patient and were terminated between 7:00 and 8:00 A.M. This time was chosen to allow recording of long epochs of REM sleep,which are characteristically observed during early morning hours of sleep. The sleep record was scored manually by a qualified polysomnographer who was blind to the treatment schedule. Sleep was classified into non-REM (NREM) and REM sleep using standard criteria (17). Another technician, also blind to the treatment schedule, calculated the apnea/hypopnea index (AHI: obstructive apneas and hypopneas per hour of sleep), mean apnea duration and mean lowest desaturation. Breathing abnormalities were grouped separately into those occurring in NREM and those occurring in REM sleep. An obstructive apnea was defined as a period of ~ 10s with absent inspiratory flow but associated respiratory effort. Obstructive hypopneas represented periods of ~ 10s with a ~ 4070 decrease in Sao, and decreased respiratory effort and/or flow. The Sao, signal was fed into a compu ter (IBM-AT)and stored for subsequent analysis to generate the cumulative percent sleeping time spent at different levels of Sao2.
FAlQ G. ISSA
Statistical Analysis Both parametric (Student's t test and analysis of variance) and nonparametric tests (Wilcoxon's test) were employed to examine the effect of clonidine on apnea indices and sleep architecture. Statistical significance was determined at p ~ 0.05.
sleep time spent in the supine position was not different in control and clonidine studies (78 ± 6070 versus 84 ± 7% for the control and clonidine studies, respectively).
Effect of Clonidine on Sleep Architecture The anthropometric and lung function In general, clonidine changed sleep patdata of the eight male patients participat- terns in all patients. This was observed ing in the study are summarized in table in both sleep studies performed during 1. All patients received a dose of cloni- the course of clonidine therapy. In all dine (expressed in ug/kg), which is eight patients, clonidine suppressed REM periods in the first half of the night (first known to influence sleep architecture. All patients slept continuously for 6 3.5 to 4.0 h of the study). In six patients to 8 h during each sleepstudy, and a mini- clonidine increased the latency to REM mum of one period of REM sleep (range, sleep from a control of 129 ± 9 min to one to four periods/night) was recorded 308 ± 24 min (P < 0.001). In two patients in each patient during control and dur- (Patients 4 and 8), cIonidine totally suping placebo nights. All patients tolerat- pressed REM sleep in both studies. In ed clonidine and slept undisturbed the six patients in whom REM sleep throughout the night. Three patients emerged in the early morning hours, an awoke momentarily once or twice dur- average of one to three REM episodes ing each clonidine night complaining of were recorded during each sleep study dry mouth. The mean total study time performed during the clonidine therapy. was similar during control, placebo, and However,the percent time spent in REM clonidine nights (control, 417 ± 14; sleep was significantly shorter than that placebo, 433 ± 23; clonidine, 426 ± 16 recorded during control or placebo min). Clonidine was administered first nights. When more than one REM epiin six patients. Although sleep-state dis- sode was recorded during clonidine tribution was not identical in the two con- studies, a tendency for a progressive introl and the two placebo studies, the crease in the duration of consecutive mean percent time spent in NREM and REM episodes was observed. Polygraphiin REM sleep for the whole group was cally,the EEG waves,distribution of rapnot significantly different in these id eye movements, and postural muscle studies, and the results of these four hypotonia during these REM episodes studies (two control and two placebo wereidentical to those 0 f the control and sleep studies) were, therefore, pooled. placebo nights. Clonidine-induced reduction in REM Similarly, the results of the two sleep studies performed during clonidine ad- sleep (expressed as percent of total study time) was not associated with increased ministration were averaged. Analysis of body position during sleep percent time spent in wakefulness (figdemonstrated that the time spent in the ure 1). Instead, there was a compensatosupine position was not statistically ry increase in NREM sleep, which was different in the control and clonidine mainly due to an increase in percent time studies. In particular, the percent of REM spent in Stage 1/11 NREM sleep(figure 1). Results
TABLE 1 ANTHROPOMETRIC AND LUNG FUNCTION DATA IN EIGHT PATIENTS Patient No. 1 2 3 4 5 6 7 8
Age
(yr) 53
45 59 31 45 36 33 43
Weight (kg)
Height (em)
BMI (kglm2j
BP (mm Hg)
FEV, (L)
FVC (L)
Pao, • (mmHg)
Paco, • (mm Hg)
Clonidine (jlglkg)
72 91 108 93 110 92 80 88
167 165 176 176 183 177 184 170
25.8 33.4 34.7 30.0 32.8 29.4 23.6 30.4
124/76 120/75 110n5 128/90 120/80 130/70 118nO 118/80
2.87 3.65 3.30 4.97 5.36 2.83 5.78 3.40
3.75 4.45 4.51 6.30 6.16 3.71 6.40 4.08
81 73 69 78 61 75 80 71
37 38 38 35 36 38 37 42
2.78 2.20 1.86 2.15 1.82 2.17 2.50 2.27
• Normal values for Calgary (1.050 m above sea level):Pso, = 70 to 80 mm Hg; Paco, = 30 to 40 mm Hg.
437
EFFECT OF CLONIDINE IN OBSTRUCTIVE SLEEP APNEA
.
80
*
E
c:
e-
60
~
40
~
~
1;
...c:.. ~
20
~
0 It'
IIII
III/IV
REM
Fig. 1. Effect of clonidine on sleep architecture. Open bars = the mean time spent in each stage (expressed as percent of total study time) in the control and placebo sleep studies; Closed bars = mean time spent in each stage (expressed as percent of total study time) during clonidine sleep studies. NS = not significant; asterisks = p < 0.05.
Effect of Clonidine on Apnea Indices When analyzed for the entire sleep record, clonidine caused a small improvement in apnea indices, with mean apnea duration decreasing from a control of 20.6 ± 0.3 to 19.3 ± 0.9 and mean lowest Sa0 2 increasing from a control of 84 ± 1.0 to 86 ± 1.5070. Both changes were small in magnitude, but they werestatistically significant (p < 0.05). However, the increase in mean AHI (from a control of 26.9 ± 4.2 to 28.3 ± 5.1 events/h during clonidine) was not statistically significant. A more careful examination of
apnea was possible within specific stages of sleep. Preliminary analysis of clonidine sleep studies indicated that, in comparison with the control nights, the pattern of breathing remained unchanged during NREM sleep, and that clonidine particularly altered REM sleep breathing abnormalities. NREMsleep. In general, clonidine did not have an affect on the pattern of breathing abnormalities during NREM sleep. Despite interindividual variability, clonidine had no effect on the group mean AHI, mean apnea duration, or mean lowest Sa0 2 (table 2). REMsleep. In two patients, REM sleep was totally suppressed in both sleep studies performed during the course of clonidine therapy (Patients 4 and 8 in table 2). In the control and placebo studies, these patients developed repetitiveapneas during both NREM and REM sleep. The improvement in nocturnal Sa0 2 was quantified as an increase in the cumulative percent time spent at or below a certain level of Sa0 2 from the control to clonidine nights. In Patient 4, the percent of cumulative percent time spent with Sa0 2 ~ 80% increased from 86 to 93% during control and clonidine nights, respectively. Patient 8 had a milder form of OSA (table 2), and the cumulative percent time spent with Sa0 2 ~ 88% increased from 88 to 94%.
TABLE 2 APNEA INDICES RECORDED DURING NREM AND REM SLEEP DURING CONTROL (AVERAGE OF CONTROL AND PLACEBO STUDIES) AND CLONIDINE NIGHTS' NREM Sleep Patient No.
2 3 4 5 6 7 8 Mean ± SEM
Study
AHI
Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine
13.7 14.8 34.6 37.4 7.8 10.6 59.0 46.2 39.6* 45.2 21.2§ 21.2 10.7 10.3 23.2 19.3 26.2 ± 6.1 25.7 ± 5.3
Duration
17.9 ± 21.2 ± 21.9 ± 22.2 ± 14.0 ± 14.8 ± 20.8 ± 18.5 ± 18.9 ± 18.6 ± 16.1 ± 21.9 ± 18.7± 25.2 ± 16.5± 13.4 ± 18.1 ± 19.5 ±
1.7 2.1 1.9 2.3 1.4 1.7 1.2 1.9 1.7 1.7 1.7 2.8t 1.9 1.8t 1.9 1.4t 1.9 2.4
REM Sleep Sao,
86 90 83 81 90 91 81 84 82 85 85 84 85 87 88 90 85 87
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
AHI
2.3 18.6 2.2 0.0 1.6 30.3 2.0 0.0 2.3 19.3 2.6 0.0 2.1 65.2 2.7t NO 1.5 33.3* 2.6t 52.8 24.6§ 2.5 1.8 61.3 1.7 13.2 2.6 6.1 2.6 25.7 3.2 NO 1.1 28.8 ± 5.7 1.3 40.1 ± 17.2
Duration
Sao,
24.3 ± 1.9
83 ± 3.7
26.3 ± 3.1
81 ± 3.1
18.4 ± 3.9
88 ± 0.6
22.2 ± 2.0
80 ± 2.1 NO
NO
32.7 20.1 32.1 19.3 25.3 13.6 23.3
± ± ± ± ± ± ±
NO
2.9 2.8t 4.1 1.1t 3.7 1.6 2.2
83 71 79 75 85 88 87
± ± ± ± ± ± ±
NO
25.6 ± 3.7 82 ± 2.9 17.7 ± 2.0t 83 ± 3.9
Definition of abbreviations: AHI = apnealhypopnea index; NO = not detected during REM sleep in clonidine studies. • Values are mean ± SEM. t p < 0.05. compared with the control values. AHI for the entire night increased from a control of 38.5 to 46.2 during clonidine therapy. § AHI for the entire night increased from a control of 21.5 to 24.9 during clonidine therapy.
t
3.0 1.2t 2.7 2.2 2.2 2.9 2.0
REM sleep was not eliminated in the other six patients. In four of these patients, the number and severity of apneas occurring during REM sleep was improved. 1Wopatients (Patients 3 and 7) had a mild form of OSA, with repetitive obstructive apneas limited to REM sleep (table 2). These apneas disappeared completely in Patient 3 during clonidine nights. This patient breathed without snoring and was lying supine during 33 min of REM sleep recorded in two clonidine sleep studies (figure 2). In Patient 7, REM-apneas became shorter in duration, and the frequency of apneas during REM sleep decreased to a value below that used for the definition of OSA (18) (table 2). A similar phenomenon was also observed in two other patients (Patients 1and 2) who developed apneas during both NREM and REM sleep during the control nights (table 2). Regular breathing during clonidine therapy was not a brief phenomenon but was observed consistently during a total period of 134min of REM sleep recorded in both patients. The patency of the upper airway was compromised during NREM sleep, but regular breathing ensued once these patients entered into REM sleep. The predominant breathing abnormality during REM sleep in Patients 5 and 6 was that of relatively long periods of obstructive hypopneas associated with persistant hypoxemia. This abnormality was transformed during clonidine intake into complete occlusive apneas that were of shorter duration and resulted in a more profound fall in the Sao, (table 2 and figure 3). Despite the decrease in duration of apneas and the associated significant increase in REM -AHI, clonidine did not have a major impact on all-night apnea indices in these two patients (table 2). All patients tolerated this dose of clonidine without reporting any unpleasant daytime side effects. There was no improvement in daytime symptoms, e.g., hypersomnolence, during clonidine therapy. Blood pressure remained unchanged throughout the protocol in these normotensive patients. Discussion
This is the first report on the use of clonidine in patients with OSA. The results indicate that clonidine administered at night alters sleep stage distribution and specifically influences breathing pattern during REM sleep in patients with OSA. Clonidine at a dose of 2.15 and 2.27 ug/kg body weight (for Patients 4 and
FAIQ G. ISSA
438
[[G
••
FAIQ G. ISSA
Introduction
Obstructive sleepapnea (OSA) is a common disorder affecting 1 to 4% of the male population (1). The disorder is characterized by repetitive closure of the pharyngeal airway during sleep because of an imbalance in the forces acting on the upper airway (2). The dilating forces generated by upper airway muscles (e.g., genioglossus) during sleep fail to counterbalance the constricting luminal forces generated by inspiratory thoracic muscles. Sleep per se induces a variety of changes in the respiratory system that enhance closure of the upper airway in susceptible subjects. For example, compared with the diaphragm, the activity of upper airway muscles preferentially decreases during sleep (3). Furthermore, sleep reduces the overall ventilatory response to hypoxia, hypercapnia, and other respiratory reflexes, thus reducing neural output to both sets of muscles (4-6). At low chemical drive and reduced neural output to the upper airway and thoracic pump muscles, closure of the upper airway is enhanced when chemical drive returns to normal levels (7). These specific effects of sleep on the respiratory system become more pronounced during rapid eye movement (REM) sleep. There is a profound hypotonia of upper airway muscles and a greater reduction in the ventilatory responses to hypoxia, hypercapnia, and mechanical loading (4-6). Furthermore, once upper airway closure occurs, the arousal threshold to asphyxia is greatly reduced during this sleep stage. Thus, during REM sleep higher levels of hypoxemia, hypercapnia, and acidemia are tolerated (8). Currently, the treatment of choice for OSA is nasally applied continuous positive airway pressure (CPAP), introduced in 1981 by Sullivan and coworkers (9). Nasal CPAP pneumatically splints the upper airway during sleep. However, this treatment is not tolerated by every patient, and compliance in most sleep centers is estimated to be approximately 700/0 (10). Clinical experience indicates that nonrespondents tend to be either pa-
SUMMARY The current treatment of choice for obstructive sleep apnea is continuous positive airway pressure. However, not all patients tolerate this form of therapy. We eveluated the effect of clonldlne hydrochloride, an a,-adrenerglc agonist with REM-suppressant activity, In eight male patients with obstructive sleep apnea. In each patient, sleep-stsge distribution and breathing pattern In two ali-night sleep studies performed during a 10-day course of clonldlne _re compared with those of two control and two placebo nights. A dose of 0.2 mg of clonldlne administered orally at bedtime totally suppreased REM sleep In two patients. In the other six patients, the same dose decreased percent time spent In REM sleep from a control of 13.4 ± 1.0 to 8.6 ± 1.4% (mean ± SEM, p < 0.05). The latency to REM sleep Increased In the latter group from a control of 129 ± 9 to 308 ± 24 min (p < 0.001). Clonldlne had no effect on the frequency and duration of non·REM breathing abnormalities. Und9r clonldlne, the level of nocturnal hypoxemia Improved In six patients. This was due to a totsl suppression of REMand the consequent lack of REMapneas In two patients. In four patients, upper airway obstruction disappeared during periods of unsuppressed REM sleep, and Sao, remained above 90% throughout this sleep stsge. Clonldlne transformed the pattern of sleep-disordered breathing during unsuppressed REMIn the other two patients from that of repetitive obstructive hypopneas associated with persistent hypoxemia to occlusive apnea and cyclical hypoxemia. These results _re observed consistently In all patients during both clonldlne-sleep studies. The data suggest that clonldlne does not only alter mechanisms Involved In the Initiation and maintenance of REM sleep but It also Influences breathing pattern during REM sleep. AM REV RESPIR DIS 1992; 145:435-439
tients with mild OSA or asymptomatic OSA. Patients wereassignedat random to one patients with otherwise severe OSA. In of two treatment schedules: placebo followed both circumstances, treatment is essen- by clonidine, or clonidine followed by placetial since retrospective studies indicate bo. In the clonidine-treatment phase, 0.20mg that survival rate decreases in patients . was taken orally at bedtime for 10 days. A similar schedule was followed during the who are not treated (11). placebo-treatment phase. Between the two Clonidine is an uradrenoreceptor ago- treatment phases, a period of 1 wk without nist that was introduced approximately treatments was instituted. A total of six all25 yr ago for the treatment of hyperten- night sleep studies were performed in each sion. The antihypertensive effects of patient. A control study was carried out beclonidine are centrally mediated, acting fore the commencement of each treatment mainly in the lower brainstem region (12). phase, and two studies were performed durClonidine is rapidly and. completely ab- ing each treatment phase: on Days 1, 2, or sorbed after an oral dose and has a Tv, 3 and on Days 7, 8, 9, or 10 of clonidine or placebo treatment. Clonidine blood levels of 5 to 15 h, and its peak plasma levels were not monitored. The protocol was apafter oral dosing occur in 60 to 90 min proved by the Ethics Review Committee of (13). Clonidine produces dryness of the Foothills Hospital and the University of nose and mouth and sedation when tak- Calgary. en during the daytime. Clonidine has a REM-suppressant activity, and has been previously used in narcolepsy (14-16). In this study, we carried out a trial with (Received in original form April 3. 1991 and in clonidine hydrochloride and placebo in revised form August 7, 1991) eight male patients with OSA. WeexamFrom the Faculty of Medicine, University of ined the effect of clonidine on respira- Calgary and Foothills Hospital, Calgary, Alberta, tion and distribution of sleep stages. Canada. 1
Methods Protocol A placebo-drug trial was used to evaluate the effect of clonidine in eight male patients with
2 Correspondence and requests for reprints should be addressed to Faiq G. Issa, M.D., Ph.D., Assistant Professor, Faculty of Medicine, University of Calgary and Foothills Hospital, 3330 Hospital Drive, NW., Calgary, Alberta, Canada T2N
4Nl.
435
436
Subjects Patients with OSA presenting with a high resting blood pressure (untreated or treated with antihypertensive drugs) were excluded from the study. Lung function and arterial blood gases weremeasured during wakefulness. The presence of obstructive apneas was confirmed polysomnographically in a diagnostic study conducted prior to the present protocol. Polysomnography All-night sleep studies were performed in the Sleep Laboratory at the Foothills Hospital. Surface electrodes recorded an electrocardiogram (EKG), two electroencephalograms (EEG), two electrooculograms (EOG), and a chin muscle electromyogram (EMG). Sao, was measured continuously with ear oximeter (Ohmeda 3700; Ohmeda, Louisville, CO). Oronasal airflow was monitored with either a thermistor taped close to the nostrils and lips or by measuring end-tidal CO 2 with an infrared capnometer (Datex; Puritan-Bennett Corp, Los Angeles, CA). Chest and abdominal wall movement was detected using an inductance plethysmograph (Respitrace'[; Ambulatory Monitoring, Ardsley, NY). All signals were recorded on a 16-channel recorder (Gould ESloo0; Gould Instruments, Cleveland, OH, or Model 6; Grass Instruments, Quincy, MA). Body position was monitored using an infrared camera positioned over the bed. Body position was marked on the polygraph record by the technician every 15to 20 min, and changes in posture were marked as they occurred. The patient was allowed to sleep undisturbed throughout the night. Two criteria were followed in all studies. First, the patient was allowed to maintain a posture most preferable to him during these studies. Second, all sleep studies commenced at the usual bedtime for each patient and were terminated between 7:00 and 8:00 A.M. This time was chosen to allow recording of long epochs of REM sleep,which are characteristically observed during early morning hours of sleep. The sleep record was scored manually by a qualified polysomnographer who was blind to the treatment schedule. Sleep was classified into non-REM (NREM) and REM sleep using standard criteria (17). Another technician, also blind to the treatment schedule, calculated the apnea/hypopnea index (AHI: obstructive apneas and hypopneas per hour of sleep), mean apnea duration and mean lowest desaturation. Breathing abnormalities were grouped separately into those occurring in NREM and those occurring in REM sleep. An obstructive apnea was defined as a period of ~ 10s with absent inspiratory flow but associated respiratory effort. Obstructive hypopneas represented periods of ~ 10s with a ~ 4070 decrease in Sao, and decreased respiratory effort and/or flow. The Sao, signal was fed into a compu ter (IBM-AT)and stored for subsequent analysis to generate the cumulative percent sleeping time spent at different levels of Sao2.
FAlQ G. ISSA
Statistical Analysis Both parametric (Student's t test and analysis of variance) and nonparametric tests (Wilcoxon's test) were employed to examine the effect of clonidine on apnea indices and sleep architecture. Statistical significance was determined at p ~ 0.05.
sleep time spent in the supine position was not different in control and clonidine studies (78 ± 6070 versus 84 ± 7% for the control and clonidine studies, respectively).
Effect of Clonidine on Sleep Architecture The anthropometric and lung function In general, clonidine changed sleep patdata of the eight male patients participat- terns in all patients. This was observed ing in the study are summarized in table in both sleep studies performed during 1. All patients received a dose of cloni- the course of clonidine therapy. In all dine (expressed in ug/kg), which is eight patients, clonidine suppressed REM periods in the first half of the night (first known to influence sleep architecture. All patients slept continuously for 6 3.5 to 4.0 h of the study). In six patients to 8 h during each sleepstudy, and a mini- clonidine increased the latency to REM mum of one period of REM sleep (range, sleep from a control of 129 ± 9 min to one to four periods/night) was recorded 308 ± 24 min (P < 0.001). In two patients in each patient during control and dur- (Patients 4 and 8), cIonidine totally suping placebo nights. All patients tolerat- pressed REM sleep in both studies. In ed clonidine and slept undisturbed the six patients in whom REM sleep throughout the night. Three patients emerged in the early morning hours, an awoke momentarily once or twice dur- average of one to three REM episodes ing each clonidine night complaining of were recorded during each sleep study dry mouth. The mean total study time performed during the clonidine therapy. was similar during control, placebo, and However,the percent time spent in REM clonidine nights (control, 417 ± 14; sleep was significantly shorter than that placebo, 433 ± 23; clonidine, 426 ± 16 recorded during control or placebo min). Clonidine was administered first nights. When more than one REM epiin six patients. Although sleep-state dis- sode was recorded during clonidine tribution was not identical in the two con- studies, a tendency for a progressive introl and the two placebo studies, the crease in the duration of consecutive mean percent time spent in NREM and REM episodes was observed. Polygraphiin REM sleep for the whole group was cally,the EEG waves,distribution of rapnot significantly different in these id eye movements, and postural muscle studies, and the results of these four hypotonia during these REM episodes studies (two control and two placebo wereidentical to those 0 f the control and sleep studies) were, therefore, pooled. placebo nights. Clonidine-induced reduction in REM Similarly, the results of the two sleep studies performed during clonidine ad- sleep (expressed as percent of total study time) was not associated with increased ministration were averaged. Analysis of body position during sleep percent time spent in wakefulness (figdemonstrated that the time spent in the ure 1). Instead, there was a compensatosupine position was not statistically ry increase in NREM sleep, which was different in the control and clonidine mainly due to an increase in percent time studies. In particular, the percent of REM spent in Stage 1/11 NREM sleep(figure 1). Results
TABLE 1 ANTHROPOMETRIC AND LUNG FUNCTION DATA IN EIGHT PATIENTS Patient No. 1 2 3 4 5 6 7 8
Age
(yr) 53
45 59 31 45 36 33 43
Weight (kg)
Height (em)
BMI (kglm2j
BP (mm Hg)
FEV, (L)
FVC (L)
Pao, • (mmHg)
Paco, • (mm Hg)
Clonidine (jlglkg)
72 91 108 93 110 92 80 88
167 165 176 176 183 177 184 170
25.8 33.4 34.7 30.0 32.8 29.4 23.6 30.4
124/76 120/75 110n5 128/90 120/80 130/70 118nO 118/80
2.87 3.65 3.30 4.97 5.36 2.83 5.78 3.40
3.75 4.45 4.51 6.30 6.16 3.71 6.40 4.08
81 73 69 78 61 75 80 71
37 38 38 35 36 38 37 42
2.78 2.20 1.86 2.15 1.82 2.17 2.50 2.27
• Normal values for Calgary (1.050 m above sea level):Pso, = 70 to 80 mm Hg; Paco, = 30 to 40 mm Hg.
437
EFFECT OF CLONIDINE IN OBSTRUCTIVE SLEEP APNEA
.
80
*
E
c:
e-
60
~
40
~
~
1;
...c:.. ~
20
~
0 It'
IIII
III/IV
REM
Fig. 1. Effect of clonidine on sleep architecture. Open bars = the mean time spent in each stage (expressed as percent of total study time) in the control and placebo sleep studies; Closed bars = mean time spent in each stage (expressed as percent of total study time) during clonidine sleep studies. NS = not significant; asterisks = p < 0.05.
Effect of Clonidine on Apnea Indices When analyzed for the entire sleep record, clonidine caused a small improvement in apnea indices, with mean apnea duration decreasing from a control of 20.6 ± 0.3 to 19.3 ± 0.9 and mean lowest Sa0 2 increasing from a control of 84 ± 1.0 to 86 ± 1.5070. Both changes were small in magnitude, but they werestatistically significant (p < 0.05). However, the increase in mean AHI (from a control of 26.9 ± 4.2 to 28.3 ± 5.1 events/h during clonidine) was not statistically significant. A more careful examination of
apnea was possible within specific stages of sleep. Preliminary analysis of clonidine sleep studies indicated that, in comparison with the control nights, the pattern of breathing remained unchanged during NREM sleep, and that clonidine particularly altered REM sleep breathing abnormalities. NREMsleep. In general, clonidine did not have an affect on the pattern of breathing abnormalities during NREM sleep. Despite interindividual variability, clonidine had no effect on the group mean AHI, mean apnea duration, or mean lowest Sa0 2 (table 2). REMsleep. In two patients, REM sleep was totally suppressed in both sleep studies performed during the course of clonidine therapy (Patients 4 and 8 in table 2). In the control and placebo studies, these patients developed repetitiveapneas during both NREM and REM sleep. The improvement in nocturnal Sa0 2 was quantified as an increase in the cumulative percent time spent at or below a certain level of Sa0 2 from the control to clonidine nights. In Patient 4, the percent of cumulative percent time spent with Sa0 2 ~ 80% increased from 86 to 93% during control and clonidine nights, respectively. Patient 8 had a milder form of OSA (table 2), and the cumulative percent time spent with Sa0 2 ~ 88% increased from 88 to 94%.
TABLE 2 APNEA INDICES RECORDED DURING NREM AND REM SLEEP DURING CONTROL (AVERAGE OF CONTROL AND PLACEBO STUDIES) AND CLONIDINE NIGHTS' NREM Sleep Patient No.
2 3 4 5 6 7 8 Mean ± SEM
Study
AHI
Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine Control Clonidine
13.7 14.8 34.6 37.4 7.8 10.6 59.0 46.2 39.6* 45.2 21.2§ 21.2 10.7 10.3 23.2 19.3 26.2 ± 6.1 25.7 ± 5.3
Duration
17.9 ± 21.2 ± 21.9 ± 22.2 ± 14.0 ± 14.8 ± 20.8 ± 18.5 ± 18.9 ± 18.6 ± 16.1 ± 21.9 ± 18.7± 25.2 ± 16.5± 13.4 ± 18.1 ± 19.5 ±
1.7 2.1 1.9 2.3 1.4 1.7 1.2 1.9 1.7 1.7 1.7 2.8t 1.9 1.8t 1.9 1.4t 1.9 2.4
REM Sleep Sao,
86 90 83 81 90 91 81 84 82 85 85 84 85 87 88 90 85 87
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
AHI
2.3 18.6 2.2 0.0 1.6 30.3 2.0 0.0 2.3 19.3 2.6 0.0 2.1 65.2 2.7t NO 1.5 33.3* 2.6t 52.8 24.6§ 2.5 1.8 61.3 1.7 13.2 2.6 6.1 2.6 25.7 3.2 NO 1.1 28.8 ± 5.7 1.3 40.1 ± 17.2
Duration
Sao,
24.3 ± 1.9
83 ± 3.7
26.3 ± 3.1
81 ± 3.1
18.4 ± 3.9
88 ± 0.6
22.2 ± 2.0
80 ± 2.1 NO
NO
32.7 20.1 32.1 19.3 25.3 13.6 23.3
± ± ± ± ± ± ±
NO
2.9 2.8t 4.1 1.1t 3.7 1.6 2.2
83 71 79 75 85 88 87
± ± ± ± ± ± ±
NO
25.6 ± 3.7 82 ± 2.9 17.7 ± 2.0t 83 ± 3.9
Definition of abbreviations: AHI = apnealhypopnea index; NO = not detected during REM sleep in clonidine studies. • Values are mean ± SEM. t p < 0.05. compared with the control values. AHI for the entire night increased from a control of 38.5 to 46.2 during clonidine therapy. § AHI for the entire night increased from a control of 21.5 to 24.9 during clonidine therapy.
t
3.0 1.2t 2.7 2.2 2.2 2.9 2.0
REM sleep was not eliminated in the other six patients. In four of these patients, the number and severity of apneas occurring during REM sleep was improved. 1Wopatients (Patients 3 and 7) had a mild form of OSA, with repetitive obstructive apneas limited to REM sleep (table 2). These apneas disappeared completely in Patient 3 during clonidine nights. This patient breathed without snoring and was lying supine during 33 min of REM sleep recorded in two clonidine sleep studies (figure 2). In Patient 7, REM-apneas became shorter in duration, and the frequency of apneas during REM sleep decreased to a value below that used for the definition of OSA (18) (table 2). A similar phenomenon was also observed in two other patients (Patients 1and 2) who developed apneas during both NREM and REM sleep during the control nights (table 2). Regular breathing during clonidine therapy was not a brief phenomenon but was observed consistently during a total period of 134min of REM sleep recorded in both patients. The patency of the upper airway was compromised during NREM sleep, but regular breathing ensued once these patients entered into REM sleep. The predominant breathing abnormality during REM sleep in Patients 5 and 6 was that of relatively long periods of obstructive hypopneas associated with persistant hypoxemia. This abnormality was transformed during clonidine intake into complete occlusive apneas that were of shorter duration and resulted in a more profound fall in the Sao, (table 2 and figure 3). Despite the decrease in duration of apneas and the associated significant increase in REM -AHI, clonidine did not have a major impact on all-night apnea indices in these two patients (table 2). All patients tolerated this dose of clonidine without reporting any unpleasant daytime side effects. There was no improvement in daytime symptoms, e.g., hypersomnolence, during clonidine therapy. Blood pressure remained unchanged throughout the protocol in these normotensive patients. Discussion
This is the first report on the use of clonidine in patients with OSA. The results indicate that clonidine administered at night alters sleep stage distribution and specifically influences breathing pattern during REM sleep in patients with OSA. Clonidine at a dose of 2.15 and 2.27 ug/kg body weight (for Patients 4 and
FAIQ G. ISSA
438
[[G
••
