- Email: [email protected]
Daytime sleepiness after long-term continuous positive airway pressure (CPAP) treatment in obstructive sleep apnea syndrome
Journal of the Neurological Sciences, 110 (1992) 21-26 © 1992 Elsevier Science Publishers B.V. All rights reserved 0022-510X/92/$05.00
21
JNS 03766
Daytime sleepiness after long-term continuous positive airway pressure (CPAP) treatment in obstructive sleep apnea syndrome E. Sforza and J. Krieger Sleep Disorders Unit, Service d'Explorations Fonctionnelles du Syst~me Nerveux, CHU Strasbourg, Strasbourg, France (Received 30 July, 1991) (Accepted 20 December, 1991)
Key words: Sleep apnea syndrome; Sleepiness; Continuous positive airway pressure treatment; Multiple sleep latency test Summary
A modified maintenance of wakefulness test was performed in 58 patients with obstructive sleep apnea (OSA) syndrome before treatment and after long-term (554 +_28 days) home therapy with nasal continuous positive airway pressure (CPAP). Before treatment the patients had a shorter mean sleep latency than controls (16 + 1 vs. 27 + 1 min, mean + SEM, P < 0.001). After treatment, the mean sleep latency increased to 20 + 1 min (P < 0.002 as compared to baseline), but was still shorter than in controls (P < 0.001). The incomplete normalization of the mean latency contrasted with the patients' claim that they no longer felt sleepy. The improvement in daytime alertness was significantly correlated with the red~etion in sleep fragmentation after CPAP treatment and with the baseline mean sleep latency. These results support the hypothesis that sleep disruption related to respiratory events plays a role in the pathogenesis of daytime sleepiness.
Introduction The most typical symptom in patients with obstructive sleep apnea (OSA) syndrome is excessive daytime sleepiness (Dement et al. 1978). The multiple sleep latency test (Richardson et al. 1978; Carskadon 1982; Mitler et al. 1982a; Reynolds et al. 1982) and the multiple maintenance of wakefulness test (Mitler et al. 1982b; Browman et al. 1983; Browman 1986) have been proposed as objective measures of daytime somnolence~ Indeed, shortened latencies during these tests, as well as a shortened sleep onset latency at night, are correlated with subjective sleepiness. Recently, nasal continuous positive airway pressure (CPAP) has been used as an effective therapy in OSA syndrome (Sullivan et al. 1981). Over the short-term this treatment induces a rapid improvement in daytime alertness, which parallels the normalization of the sleep structure and the resolution of nocturnal hypoxemia (Issa and Sullivan 1986; Rothenberg and Rapaport 1985). However, although the multiple sleep latency increases in most cases, no treatment of OSAS, whether
Correspondence to: Jean Krieger, MD, Service d'Explorations Fonctionnelles du Syst~me Nerveux, CHU, F-67091 Strasbourg Cedex, France. Tel.: (33) 88-16-13-12; Fax: (33) 88-60-75-50.
medical (Roth et al. 1980), surgical (Zorick et al. 1983) or with short-term CPAP (Rajagopal et al. 1986; Di Philippo et al. 1988), results in the normalization of the multiple sleep latency test. No data have been published concerning long-term effects of CPAP on subjective and objective sleepiness. The purpose of our study was to analyse the long-term effects of home-treatment with CPAP on daytime sleepiness in OSA patients.
Methods Protocol Fifty-eight patients, 53 men and 5 women, with a clinical and polysomnographic diagnosis of obstructive sleep apnea were evaluated (see details below) at the time of diagnosis and after long-term home CPAP therapy. The reevaluation was performed after an interval of 554 + 28 days (mean + SEM). The mean CPAP pressure applied was 9 _+0 cm H 2 0 with a range from 4 to 14. The mean rate of use of CPAP, computed from the readings of a time-counter built into the CPAP device, was 5.3 + 0.2 h / d a y (range: 1.1-8.3). For each patient, daytime sleepiness was assessed by a modified maintenance of wakefulness test administered before CPAP (baseline) at the time of the
22 diagnosis, and after CPAP at the time of the follow-up reevaluation. Nocturnal variables were monitored while off CPAP at the baseline evaluation and while on CPAP at the follow-up evaluation. They included the EEG, EOG, EMG of chin muscles and ECG. Respiration was analyzed with a Fleisch No. 2 pneumotachograph, thoracic and abdominal strain gauges or an esophageal balloon, and an ear oximeter (Ohmeda Biox III). Sleep was scored according to the criteria of Rechtschaffen and Kales (1968). The following sleep parameters were measured: total sleep time (TST); wage time after sleep onset (WASO); total sleep period (TSP = TST + WASO); the percentages of stages 1, 2, 3-4 and rapid eye movement (REM) sleep; sleep efficiency (SE), defined as the ratio of TST to TSP; number of awakenings (defined by an alpha activity in the EEG lasting more than 20 sec) and of arousals (defined as a transient resumption of alpha activity a n d / o r an increase in chin EMG amplitude for less than 20 see). An index of sleep stability was computed by dividing the total sleep time by the number of sleep stage shifts (stability index). During the CPAP-treated night, the recording was started with the lowest possible CPAP pressure (3 c m H 2 0 ) . The pressure was progressively increased until apneas and snoring were eliminated; this was achieved within 30-60 min. The period before the efficient CPAP pressure was reached was discarded. Therefore, the measured sleep latency and the total sleep time were biased; they were not taken into account in the data analysis. For the same reason, WASO and the numbers of awakenings and arousals were expressed as a function of the total sleep period, in order to permit between night comparisons. Subjective daytime sleepiness was defined according to its time and conditions of occurrence as mild (class 3: extension of normal diurnal sleepiness, persists despite long night sleep), moderate (class 2: falls asleep at rest (reading, watching TV), can keep awake with physical activity), or severe (class 1: falls asleep during motor activity (talking, eating), incapacitated by sleepiness) (Sullivan 1985). Objective daytime sleepiness was assessed with a modified maintenance of wakefulness test. For six sessions at 10.00, 12.00, 14.00, 16.00, 18.00, and 20.00 h, the patient lay in bed in a quiet room, being left undisturbed for 30 min without instructions either to ~tay awake or ~*o try to fall asleep. The patients were told that the aim of the test was tr examine their behavior when left undisturbed and that they were free to read, work, or just idle, provided that they stayed in bed; no external intervention was allowed during the tests, and the telephone was disconnected. This procedure is close to the one proposed in the modified assessment of sleepiness test (Erman et al. 1987) in that it simulates the sedentary situations during which
the patients most readily fall asleep in their daily life. All tests were terminated 20 min after the first epoch of sleep or after 30 min without sleep. Between the tests the patients were asked to stay awake and out of bed. The sleep latency was the time from the start of the recording to the first epoch of any sleep stage (including stage 1); it was arbitrarily said to be equal to 30 min when no sleep was recorded. The mean of the six sleep latencies in the six sessions was the mean sleep latency. Twenty-nine normal male subjects (age 45.8 + 2.1 years, BMI 26.8 + 2.4 kg/m2), without a history of sleep-wake complaints, excessive daytime sleepiness or snoring or irregular sleep (poor sleepers and shiftworkers were excluded) constituted the control group.
Statistical analysis Comparisons between controls and OSA patients were made with Student's t-test. Within patients, differences between baseline and follow-up values were tested with Student's t-test for paired values. Correlations were analysed using least square regression analysis and Pearson's correlation coefficient. The accepted level of statistical significance was P < 0.05 (two-tailed). Results are given as means + SEM.
Results
The distribution of the anthropometric and nocturnal respiratory parameters in the 58 patients at the time of the diagnosis are shown in Fig. 1. Their mean age was 53.0 + 1.3 years with a body mass index (BMI) of 32.4 + 2.3 kg/m 2. Thirty-three patients had a BMI greater than 30.0 k g / m 2. The TABLE I NOCTURNAL SLEEP PARAMETERS BEFORE AND AFTER LONG-TERM TREATMENT WITH CPAP Mean :l: SEM.
Total sleep time
Baseline
Long-term CPAP
274
269
+13
P
+13
(min) Sleep efficiency WASO (min/h TSP) No. of arousals/ h TSP No. of awakenings/ h TSP Stability index (see) Stage 1 (%) Stage 2 (%) Stage 3-4 (%) REM (%)
0.60+ 0.02 24.2 + i.4
0.84+ 0.02 9.7 + 1.0
< 0.001 < 0.001
39
+ 4
9
+ 1
< 0.001
15
+ 1
7
+ 1
< 0.001
14 32.5 56.9 2.9 7.8
+ + + +_ +
+ + + + +
< 0.001 < 0.001 <0.001 < 0.001 <0.001
! 1.9 1.5 0.7 0.9
64 9.9 49.1 18.2 23.1
4 1.0 1.3 1.4 1.2
23 severity of their sleep apnea syndrome covered a wide range (Fig. 1), from mild to severe, with a mean a p , ea index of 71 + 5 apneas per hour of sleep and a me a~! lowest S a O 2 of 86-I-1%. Not surprisingly, they had small amounts of stage 3-4 and REM sleep, with great amounts of stages 1 and 2 non-REM sleep during the baseline nocturnal polygraphic study. Their sleep was disrupted, as evidenced by a large number of arousals and awakenings and a low sleep efficiency and stability index (Table 1). All but 8 patients reported subjective daytime sleepiness ranging from mild (n = 11) to moderate (n = 15) or severe (n = 24). The excessive daytime sleepiness was confirmed by a reduction in the mean sleep latency during the multiple latency test. The patients had a mean latency of 16.5 _+ 1.2 min, ranging from 0.3 to 30. This value was significantly shorter than in the control group (mean sleep latency: 26.9 + 0.8 min, range 13.2-30, P < 0.001). Sleep onset REM periods (SOREMPs) were recorded in six patients, who had a mean number of 1.3 _+0.2 SOREMPs.
Following home therapy, all sleepy patients reported the elimination of daytime sleepiness. No significant change in body mass index was noted. The multiple latency test showed a significant increase in mean sleep latency to 20.2 + 1.0 min ( P < 0.002), but the latency was still shorter than that of the control group (P < 0.001). SOREMPs were recorded in three patients (only one of them also had SOREMPs before treatment), with a mean number of 1.7 + 0.4 episodes. The sleep continuity parameters improved (Table 1). The percentage of slow wave sleep (stage 3-4) and of REM sleep increased, whereas the percentage of light sleep (stages 1 and 2) decreased (Table 1). The best predictor of the improvement in sleep latency with CPAP treatment was the improvement in sleep fragmentation indices. Significant correlations were demonstrated between the change in mean sleep latency and the change in wake after sleep onset (WASO, r = -0.35, P < 0.01), in sleep efficiency (SE, r = 0.35, P < 0.01) and in the sleep stability index (r = 0.36, P < 0.01) (Fig. 2). The increase in mean sleep latency was also correlated with the baseline apnea
liI15............. ~j~ 1 f15 1~.......... ~ " 1 ~86
--~
[IIIIL, ] ;,
253545556575 208040506070 0 10 20 BMI
AGE
6
MSL
30
e 6
4
0~~____~~,~ 0 C 0 % ~20180 OA10Mean 20 SODur4050 0 10 12 8
6OA+H1201leo
1(]
4
7O 8O 90
C
IO0
Mean Sa02
35 55 75 95 6O 70 8O9OIO0 Mean Lowest Min Sa02
Sa02
Fig. 1. Distribution of anthropome;ric and respiratory nocturnal parameters on baseline in the 58 OSA patients (age, in years, BMI -- body mass index, in k g / m 2, AI = apnea index, in n / h , A Mean Dur = apnea mean duration, in s, A + HI = apnea + hypopnea index, in n/h, Mean SaO 2 -- mean SaO 2 over the whole night, in %, Min SaO 2 -- absolute minimal SaO 2 for the whole night, in %, Mean Lowest SaO 2 = mean of the minimal SaO 2 observed after each apnea, in %.
24 index ( r = 0.33, P < 0.0125) and the baseline mean sleep latency during the multiple latency test (r = - 0 . 5 8 , P < 0 . 0 0 0 0 1 , Fig. 3). A stepwise multiple regression analysis showed that only 2 variables were independently correlated with the change in mean sleep latency; these were the baseline mean sleep latency and the change in sleep efficiency, which accounted for 34% and 8% of the variance of the change in multiple sleep latency, respectively. No correlation was found between the change in mean sleep latency and the change in BMI nor between the increase in mean sleep
30
•~ MSL ,,
AMSL - 0.07 AI - 0.97
20 10 0
-10 •
-20
'
0
S0
& MSL A MSL , , . 0.26 AWASO + 0.001
30
'
50
& MSL
I
p < '
100
0.0125
I
'
150
200 AI
._
A MSL - .
•
20
r - 0.83 I
0.53 MSL + 12.45
20 ~
:
I
i
i
I
I
i
ig
Ii
10 •
-
•
-10
I
-
I
r -
- 0.35
-20
I
-S0 3O
inl
p •
'
I
-40
'
-30
-10
0.01"
I
'
-20
I
I
'
-10
I
0
'
10
20
& WASO
A MSL
r -- - 0.58
-20
'
0
AMSL = 15.5 A S Elf + 0.001
-
20
-
1o
,, l
•
m
l
I
20
'
" I
30
"
40
Fig. 3. Regression analysis of the change in multiple sleep latency (AMSL, in •in) after long-term CPAP therapy vs. the baseline multiple sleep latency and the baseline apnea index (Al, in n/h).
.'.:"
,
10
p < 0.00001 '
USL
,•
am
I
~
•
l I
•
I
.
• 10
•
•20
,
,
-0,4
,
-0,2
,
,
0,0
• r = 0.35 , , , 0,2
,
0,4
p < 0.01 , 0,6
0,8
A S Eft 30
latency and the various parameters of nocturnal hypoxemia.
Discussion
A MSL A MSL -" 0.1 &Stab Ind - 1.0
20
.
10
.
: • hi= •" "
-.
• o
::-
-10
•
-20
i
-100
0
•
D-
r,
•
0.36
p
•
0.01
,
100
200
& Stab Ind Fig. 2. Regression analysis of the change in multiple sleep latency
(AMSL, in min) after long-term CPAP therapy vs. the changes in sleep continuity parameters (AWASO = change in wake time after sleep onset, in min/h of total sleep period, ,~SE = change in sleep efficiency, AStab lnd = change in stability index, in sec).
Our study demonstrates a significant improvement in multiple sleep latency following long-term CPAP therapy, which is concomitant with the elimination of subjective daytime sleepiness. The improvement in alertness was paralleled by an improvement in sleep structure, similar to what had been reported after short-term CPAP (Issa and Sullivan 1986) or surgical treatment (Zorick et al. 1983). An improvement in alertness has been observed immediately after the initiation of CPAP treatment (Rajagopal et al. 1986; Di Philippe et al. 1988), or with short-term treatment (Wittig et al. 1986; Lamphere et al. 1989), but the results were not compared with those of controls. In our study, the multiple sleep latency was not normalized, as compared to unsleepy controls. A similar incomplete improvement has also been re-
25 ported with medical (Roth et al. 1980) or surgical (Zorick et al. 1983) treatment of OSA, as well as with short-term (26 days to 6 months) CPAP treatment (Gaddy and Doghrajmi 1991). This incomplete improvement contrasts with the patients' subjective feeling, since they all claimed that they were no longer sleepy; it once again underlines the fact that the subjective self-evaluation and objective evaluation of sleepiness are not parallel (Dement and Carskadon 1982). The reasons for the persistence of some degree of shortening of the multiple sleep latency are not clear, since sleep structure improved with CPAP treatment. One possible explanation would be that sleep duration remains short despite the improved sleep quality. Indeed, several patients reported that they woke up early in the morning and did not use CPAP during the early morning hours. Since we discarded the sleep period with inefficient CPAP pressure at the beginning of the night, our study does not permit an evaluation of total sleep time. However, the time during which CPAP was used at home, as evaluated from the readings of the time-counter of the CPAP device, measures the amount of normal sleep the patients actually got. Although it varied widely, this measure was not correlated with the degree of improvement of the mean sleep latency, suggesting that shortened sleep duration was not the main factor of the persistence of a decreased mean sleep latency. Another possible explanation would be that obstructive sleep apnea results in irreversible brain dysfunction. This view would be supported by the recent report of abnormal auditory event-related potentials (decreased P300 amplitude and increased P300 latency), which only partially improved with CPAP treatment (Rumbach et al. 1991). The persistence of some degree of central nervous system dysfunction after the elimination of sleep apneas may also support the early hypothesis that an underlying brain dysfunction plays a causative role in the development of obstructive sleep apnea (Lugaresi et al. 1967). The determinants of daytime somnolence are still debated. Previous studies have shown that the best predictors of daytime sleepiness in OSA patients were the sleep discontinuity and fragmentation due to respiratory events (Guilleminault et al. 1988; Roth et al. 1980; Roehrs 1989). Experimental sleep fragmentation in normal subjects induces sleepiness similar to that observed in OSA (Bonnet 1985, 1986; Stepanski et al. 1987; Magee et al. 1987). The possible role of other factors such as the reduction of slow and REM sleep and the severity of nocturnal hypoxemia has also been suggested (Orr et al. 1979; Issa and Sullivan, 1986; Chesson et al. 1988).
In our study, the improvement in daytime alertness was correlated with the improvement in sleep stability and the decrease in sleep fragmentation, as well as with the severity of sleep apnea before treatment. Thus, our data tend to support the role of sleep fragmentation due to respiratory events. The absence of a correlation between the improvement in daytime alertness and the severity of nocturnal hypoxemia does not support a role for nocturnal hypoxemia, in accordance with the results observed with medical (Roth et al. 1980) or surgical (Zorick et al. 1983) treatments of OSA. The persistence of SOREMPs was unexpected. SOREMPs have been considered a direct consequence of sleep disruption (Browman et al. 1983) and could therefore have been expected to be eliminated with the CPAP-induced normalization of sleep structure. References Bonnet, M. (1985) Effect of sleep disruption in sleep, performance and mood. Sleep, 8: 11-19. Bonnet, M. (1986) Performance and sleepiness as a function of frequency and placement of sleep disruption. Psychophysiology, 23: 263-271. Browman, C.P. (1986) Evaluation of daytime somnolence: objective measures of sleep-wake tendency. J. Electrophysiol. Techn., 13: 233-239. Browman, C.P., K.S. Krishnareddy, M.G. Sampson and M.M. Mitler (1983) REM sleep episodes during the maintenance of wakefulness test in patients with sleep apnea syndrome and patients with narcolepsy. Sleep, 6: 23-28. Carskadon, M.A. (Ed.) (1982) Current perspectives on daytime sleepiness. Sleep, 5, suppl. 2: $55-$202. Chesson, A.L., W.M. Anderson, K.C. Richards, L. Bairnsfather, S.C. Kirby and R.B. George (1988) Apnea and apnea plus hypopnea indices correlate differently with symptoms in obstructive sleep apnea than does quantitation of nocturnal hypoxemia. Sleep Res., 17: 61. Dement, W.C. and M.A. Carskadon (1982) Current perspectives on daytime sleepiness: the issues. Sleep, Suppl. 2: $56-$66. Dement, W.C., M.A. Carskadon and G. Richardson (1978) Excessive daytime sleepiness in the sleep apnea syndrome. In: Guilleminault, C. and Dement, W.C. (Eds.), Sleep apnea syndromes, Alan R. Liss, New York, pp. 23-46. DiPhilippo, M.A., J.M. Fry, M.R. Pressman (1988) Objective measurement of daytime sleepiness following treatment of obstructive sleep apnea with nasal CPAP. Sleep Res., 17: 167. Erman, M.K., B. Beckhma, D.A. Gardner and H.P. Roffwarg (1987) The modified assessment of sleepiness test (MAST). Sleep Res., 16: 550. Gady, J.R. and K. Doghramji (1991) Daytime sleepiness after nCPAP treatment. Sleep Res., 20: 245. Guilleminault, C., M. Partinen, M.A. Quera-Salva, B. Hayes, W.C. Dement and G. Nino-Murcia (1988) Determinants of daytime sleepiness in obstructive sleep apnea. Chest, 94: 32-37. Issa, F.G. and C.E. Sullivan (1986) The immediate effects of nasal continuous positive airway pressure treatment, on sleep pattern in patients with obstructive sleep apnea syndrome. Electroenceph. Clin. Neurophysiol., 63: 10-17. Lamphere, J., T. Roehrs, R. Wittig, F. Zorick, W.A. Conway and T. Roth (1989) Recovery of Alertness After CPAP in Apnea. Chest, 96: 1364-1367.
26 Magee, J., J. Harsh and P. Badia (1987) Effects of experimentally induced sleep fragmentation on sleep and sleepiness. Psychophysiology, 234: 528-534. Mitler M.M., G.K., M.G. Sampson, C.P. Browman (1982) Multiple day-time nap approaches to evaluating the sleepy patient. Sleep, 5: S119-S127. Mitler, M.M., S. Gujavarty and C.P. Browmann (1982) Maintenance of wakefulness test: a polysomnographic technique for evaluating treatment efficacy in patients with excessive somnolence. Electroenceph. Clin. Neurophysiol., 53: 658-661. Orr, W.C., R.J. Martin, N.K. lmes, R.M. Rogers and MIL. Stahl (1979) Hypersomnolent and nonhypersomnolent patients with upper airway obstruction during sleep. Chest, 75: 418-422. Rajagopal, K.R., L.L. Dillard, T.A. Bennett, C.J. Tenholder and M.F. Tellis (1986) Overnight nasal CPAP improves hypersomnolence in sleep apnea. Chest, 90: 172-176. Rechtschaffen, A. and A. Kales (1968) A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Brain Information Service. Brain Research Institute, Los Angeles, CA. Reynolds, C.F., I!I, P.A. Coble, D.J. Kupfer and B.C. Holzer (1982) Application of the multiple sleep latency test in disorders of excessive sleepiness. Electroenceph. Clin. Neurophysiol., 53: 443452. Richardson, G.S., M.A. Carskadon, W. Flagg, J. VandenHoed, W.C. Dement and M.M. Mitler (1978) Excessive daytime sleepiness in man: multiple sleep latency measurement in narcoleptics and in control subjects. Electroencepb. Clin. Neurophysiol., 45: 621-627. Roehrs, T., F. Zorick, R. Winig, W. Conway and T. Roth (1989) Predictors of objective level of daytime sleepiness in patients with sleep-related breathing disorders. Chest, 95: 1202-1206.
Roth, T., K.M. Hartse, F. Zorick and W. Conway (1980) Multiple naps and the evaluation of daytime sleepiness in patients with upper airway sleep apnea. Sleep, 3: 425-439. Rothenberg, S.A. and D.M. Rapoport (1985) Recovery from sleep deprivation in the obstructive sleep apnea syndrome (OSAS) treated with nasal continuous positive airway pressure. Sleep Res., 14: 204. Rumbach, L., J. Krieger and D. Kurtz (1991) Auditory event-related potentials in obstructive sleep apnea: effects of treatment with nasal continuous positive airway pressure. Electroenceph. Clin. Neurophysiol., 80: 454-457. Stepanski, E., J. Lamphere, P. Badia, F. Zorick and T. Roth (1984) Sleep fragmentation and daytime sleepiness. Sleep, 7: 18-26. Stepanski, E., J. Lamphere, T. Roehrs, F. Zorick and T. Roth (1987) Experimental sleep fragmentation in normal subjects. Int. J. Neurosci., 33: 207-214. Sullivan, C.E. and F.G. Issa (1985) Obstructive sleep apnea. Ciin. Chest. Med., 6: 633-650. Sullivan, C.E., F.G. Issa, M. Berthon-Jones and L. Eves (1981) Reversal of obstructive sleep apnoea by continuous positive airway pressure applied through the nares. Lancet, i: 862-865. Lugaresi, E., G. Coccagna and G. Berti Ceroni (1967) Syndrome de Pickwick et syndrome d'hypoventilation alveolaire primaire. Rev. Neurol., 116: 678-679. Wittig, R., F. Zorick, W. Conway, J. Ward and T. Roth (1986) Normalization of the MSLT after six weeks of CPAP for sleep apnea syndrome. Sleep Res., 15: 185. Zorick, F., T. Roehrs, W. Conway, S. Fujita, R. Wittig and T. Roth (1983) Effects of uvulopalatopharyngoplasty on daytime sleepiness associated with sleep apnea syndrome. Bull. Eur. Physiopathol. Respir., 19: 600-603.
21
JNS 03766
Daytime sleepiness after long-term continuous positive airway pressure (CPAP) treatment in obstructive sleep apnea syndrome E. Sforza and J. Krieger Sleep Disorders Unit, Service d'Explorations Fonctionnelles du Syst~me Nerveux, CHU Strasbourg, Strasbourg, France (Received 30 July, 1991) (Accepted 20 December, 1991)
Key words: Sleep apnea syndrome; Sleepiness; Continuous positive airway pressure treatment; Multiple sleep latency test Summary
A modified maintenance of wakefulness test was performed in 58 patients with obstructive sleep apnea (OSA) syndrome before treatment and after long-term (554 +_28 days) home therapy with nasal continuous positive airway pressure (CPAP). Before treatment the patients had a shorter mean sleep latency than controls (16 + 1 vs. 27 + 1 min, mean + SEM, P < 0.001). After treatment, the mean sleep latency increased to 20 + 1 min (P < 0.002 as compared to baseline), but was still shorter than in controls (P < 0.001). The incomplete normalization of the mean latency contrasted with the patients' claim that they no longer felt sleepy. The improvement in daytime alertness was significantly correlated with the red~etion in sleep fragmentation after CPAP treatment and with the baseline mean sleep latency. These results support the hypothesis that sleep disruption related to respiratory events plays a role in the pathogenesis of daytime sleepiness.
Introduction The most typical symptom in patients with obstructive sleep apnea (OSA) syndrome is excessive daytime sleepiness (Dement et al. 1978). The multiple sleep latency test (Richardson et al. 1978; Carskadon 1982; Mitler et al. 1982a; Reynolds et al. 1982) and the multiple maintenance of wakefulness test (Mitler et al. 1982b; Browman et al. 1983; Browman 1986) have been proposed as objective measures of daytime somnolence~ Indeed, shortened latencies during these tests, as well as a shortened sleep onset latency at night, are correlated with subjective sleepiness. Recently, nasal continuous positive airway pressure (CPAP) has been used as an effective therapy in OSA syndrome (Sullivan et al. 1981). Over the short-term this treatment induces a rapid improvement in daytime alertness, which parallels the normalization of the sleep structure and the resolution of nocturnal hypoxemia (Issa and Sullivan 1986; Rothenberg and Rapaport 1985). However, although the multiple sleep latency increases in most cases, no treatment of OSAS, whether
Correspondence to: Jean Krieger, MD, Service d'Explorations Fonctionnelles du Syst~me Nerveux, CHU, F-67091 Strasbourg Cedex, France. Tel.: (33) 88-16-13-12; Fax: (33) 88-60-75-50.
medical (Roth et al. 1980), surgical (Zorick et al. 1983) or with short-term CPAP (Rajagopal et al. 1986; Di Philippo et al. 1988), results in the normalization of the multiple sleep latency test. No data have been published concerning long-term effects of CPAP on subjective and objective sleepiness. The purpose of our study was to analyse the long-term effects of home-treatment with CPAP on daytime sleepiness in OSA patients.
Methods Protocol Fifty-eight patients, 53 men and 5 women, with a clinical and polysomnographic diagnosis of obstructive sleep apnea were evaluated (see details below) at the time of diagnosis and after long-term home CPAP therapy. The reevaluation was performed after an interval of 554 + 28 days (mean + SEM). The mean CPAP pressure applied was 9 _+0 cm H 2 0 with a range from 4 to 14. The mean rate of use of CPAP, computed from the readings of a time-counter built into the CPAP device, was 5.3 + 0.2 h / d a y (range: 1.1-8.3). For each patient, daytime sleepiness was assessed by a modified maintenance of wakefulness test administered before CPAP (baseline) at the time of the
22 diagnosis, and after CPAP at the time of the follow-up reevaluation. Nocturnal variables were monitored while off CPAP at the baseline evaluation and while on CPAP at the follow-up evaluation. They included the EEG, EOG, EMG of chin muscles and ECG. Respiration was analyzed with a Fleisch No. 2 pneumotachograph, thoracic and abdominal strain gauges or an esophageal balloon, and an ear oximeter (Ohmeda Biox III). Sleep was scored according to the criteria of Rechtschaffen and Kales (1968). The following sleep parameters were measured: total sleep time (TST); wage time after sleep onset (WASO); total sleep period (TSP = TST + WASO); the percentages of stages 1, 2, 3-4 and rapid eye movement (REM) sleep; sleep efficiency (SE), defined as the ratio of TST to TSP; number of awakenings (defined by an alpha activity in the EEG lasting more than 20 sec) and of arousals (defined as a transient resumption of alpha activity a n d / o r an increase in chin EMG amplitude for less than 20 see). An index of sleep stability was computed by dividing the total sleep time by the number of sleep stage shifts (stability index). During the CPAP-treated night, the recording was started with the lowest possible CPAP pressure (3 c m H 2 0 ) . The pressure was progressively increased until apneas and snoring were eliminated; this was achieved within 30-60 min. The period before the efficient CPAP pressure was reached was discarded. Therefore, the measured sleep latency and the total sleep time were biased; they were not taken into account in the data analysis. For the same reason, WASO and the numbers of awakenings and arousals were expressed as a function of the total sleep period, in order to permit between night comparisons. Subjective daytime sleepiness was defined according to its time and conditions of occurrence as mild (class 3: extension of normal diurnal sleepiness, persists despite long night sleep), moderate (class 2: falls asleep at rest (reading, watching TV), can keep awake with physical activity), or severe (class 1: falls asleep during motor activity (talking, eating), incapacitated by sleepiness) (Sullivan 1985). Objective daytime sleepiness was assessed with a modified maintenance of wakefulness test. For six sessions at 10.00, 12.00, 14.00, 16.00, 18.00, and 20.00 h, the patient lay in bed in a quiet room, being left undisturbed for 30 min without instructions either to ~tay awake or ~*o try to fall asleep. The patients were told that the aim of the test was tr examine their behavior when left undisturbed and that they were free to read, work, or just idle, provided that they stayed in bed; no external intervention was allowed during the tests, and the telephone was disconnected. This procedure is close to the one proposed in the modified assessment of sleepiness test (Erman et al. 1987) in that it simulates the sedentary situations during which
the patients most readily fall asleep in their daily life. All tests were terminated 20 min after the first epoch of sleep or after 30 min without sleep. Between the tests the patients were asked to stay awake and out of bed. The sleep latency was the time from the start of the recording to the first epoch of any sleep stage (including stage 1); it was arbitrarily said to be equal to 30 min when no sleep was recorded. The mean of the six sleep latencies in the six sessions was the mean sleep latency. Twenty-nine normal male subjects (age 45.8 + 2.1 years, BMI 26.8 + 2.4 kg/m2), without a history of sleep-wake complaints, excessive daytime sleepiness or snoring or irregular sleep (poor sleepers and shiftworkers were excluded) constituted the control group.
Statistical analysis Comparisons between controls and OSA patients were made with Student's t-test. Within patients, differences between baseline and follow-up values were tested with Student's t-test for paired values. Correlations were analysed using least square regression analysis and Pearson's correlation coefficient. The accepted level of statistical significance was P < 0.05 (two-tailed). Results are given as means + SEM.
Results
The distribution of the anthropometric and nocturnal respiratory parameters in the 58 patients at the time of the diagnosis are shown in Fig. 1. Their mean age was 53.0 + 1.3 years with a body mass index (BMI) of 32.4 + 2.3 kg/m 2. Thirty-three patients had a BMI greater than 30.0 k g / m 2. The TABLE I NOCTURNAL SLEEP PARAMETERS BEFORE AND AFTER LONG-TERM TREATMENT WITH CPAP Mean :l: SEM.
Total sleep time
Baseline
Long-term CPAP
274
269
+13
P
+13
(min) Sleep efficiency WASO (min/h TSP) No. of arousals/ h TSP No. of awakenings/ h TSP Stability index (see) Stage 1 (%) Stage 2 (%) Stage 3-4 (%) REM (%)
0.60+ 0.02 24.2 + i.4
0.84+ 0.02 9.7 + 1.0
< 0.001 < 0.001
39
+ 4
9
+ 1
< 0.001
15
+ 1
7
+ 1
< 0.001
14 32.5 56.9 2.9 7.8
+ + + +_ +
+ + + + +
< 0.001 < 0.001 <0.001 < 0.001 <0.001
! 1.9 1.5 0.7 0.9
64 9.9 49.1 18.2 23.1
4 1.0 1.3 1.4 1.2
23 severity of their sleep apnea syndrome covered a wide range (Fig. 1), from mild to severe, with a mean a p , ea index of 71 + 5 apneas per hour of sleep and a me a~! lowest S a O 2 of 86-I-1%. Not surprisingly, they had small amounts of stage 3-4 and REM sleep, with great amounts of stages 1 and 2 non-REM sleep during the baseline nocturnal polygraphic study. Their sleep was disrupted, as evidenced by a large number of arousals and awakenings and a low sleep efficiency and stability index (Table 1). All but 8 patients reported subjective daytime sleepiness ranging from mild (n = 11) to moderate (n = 15) or severe (n = 24). The excessive daytime sleepiness was confirmed by a reduction in the mean sleep latency during the multiple latency test. The patients had a mean latency of 16.5 _+ 1.2 min, ranging from 0.3 to 30. This value was significantly shorter than in the control group (mean sleep latency: 26.9 + 0.8 min, range 13.2-30, P < 0.001). Sleep onset REM periods (SOREMPs) were recorded in six patients, who had a mean number of 1.3 _+0.2 SOREMPs.
Following home therapy, all sleepy patients reported the elimination of daytime sleepiness. No significant change in body mass index was noted. The multiple latency test showed a significant increase in mean sleep latency to 20.2 + 1.0 min ( P < 0.002), but the latency was still shorter than that of the control group (P < 0.001). SOREMPs were recorded in three patients (only one of them also had SOREMPs before treatment), with a mean number of 1.7 + 0.4 episodes. The sleep continuity parameters improved (Table 1). The percentage of slow wave sleep (stage 3-4) and of REM sleep increased, whereas the percentage of light sleep (stages 1 and 2) decreased (Table 1). The best predictor of the improvement in sleep latency with CPAP treatment was the improvement in sleep fragmentation indices. Significant correlations were demonstrated between the change in mean sleep latency and the change in wake after sleep onset (WASO, r = -0.35, P < 0.01), in sleep efficiency (SE, r = 0.35, P < 0.01) and in the sleep stability index (r = 0.36, P < 0.01) (Fig. 2). The increase in mean sleep latency was also correlated with the baseline apnea
liI15............. ~j~ 1 f15 1~.......... ~ " 1 ~86
--~
[IIIIL, ] ;,
253545556575 208040506070 0 10 20 BMI
AGE
6
MSL
30
e 6
4
0~~____~~,~ 0 C 0 % ~20180 OA10Mean 20 SODur4050 0 10 12 8
6OA+H1201leo
1(]
4
7O 8O 90
C
IO0
Mean Sa02
35 55 75 95 6O 70 8O9OIO0 Mean Lowest Min Sa02
Sa02
Fig. 1. Distribution of anthropome;ric and respiratory nocturnal parameters on baseline in the 58 OSA patients (age, in years, BMI -- body mass index, in k g / m 2, AI = apnea index, in n / h , A Mean Dur = apnea mean duration, in s, A + HI = apnea + hypopnea index, in n/h, Mean SaO 2 -- mean SaO 2 over the whole night, in %, Min SaO 2 -- absolute minimal SaO 2 for the whole night, in %, Mean Lowest SaO 2 = mean of the minimal SaO 2 observed after each apnea, in %.
24 index ( r = 0.33, P < 0.0125) and the baseline mean sleep latency during the multiple latency test (r = - 0 . 5 8 , P < 0 . 0 0 0 0 1 , Fig. 3). A stepwise multiple regression analysis showed that only 2 variables were independently correlated with the change in mean sleep latency; these were the baseline mean sleep latency and the change in sleep efficiency, which accounted for 34% and 8% of the variance of the change in multiple sleep latency, respectively. No correlation was found between the change in mean sleep latency and the change in BMI nor between the increase in mean sleep
30
•~ MSL ,,
AMSL - 0.07 AI - 0.97
20 10 0
-10 •
-20
'
0
S0
& MSL A MSL , , . 0.26 AWASO + 0.001
30
'
50
& MSL
I
p < '
100
0.0125
I
'
150
200 AI
._
A MSL - .
•
20
r - 0.83 I
0.53 MSL + 12.45
20 ~
:
I
i
i
I
I
i
ig
Ii
10 •
-
•
-10
I
-
I
r -
- 0.35
-20
I
-S0 3O
inl
p •
'
I
-40
'
-30
-10
0.01"
I
'
-20
I
I
'
-10
I
0
'
10
20
& WASO
A MSL
r -- - 0.58
-20
'
0
AMSL = 15.5 A S Elf + 0.001
-
20
-
1o
,, l
•
m
l
I
20
'
" I
30
"
40
Fig. 3. Regression analysis of the change in multiple sleep latency (AMSL, in •in) after long-term CPAP therapy vs. the baseline multiple sleep latency and the baseline apnea index (Al, in n/h).
.'.:"
,
10
p < 0.00001 '
USL
,•
am
I
~
•
l I
•
I
.
• 10
•
•20
,
,
-0,4
,
-0,2
,
,
0,0
• r = 0.35 , , , 0,2
,
0,4
p < 0.01 , 0,6
0,8
A S Eft 30
latency and the various parameters of nocturnal hypoxemia.
Discussion
A MSL A MSL -" 0.1 &Stab Ind - 1.0
20
.
10
.
: • hi= •" "
-.
• o
::-
-10
•
-20
i
-100
0
•
D-
r,
•
0.36
p
•
0.01
,
100
200
& Stab Ind Fig. 2. Regression analysis of the change in multiple sleep latency
(AMSL, in min) after long-term CPAP therapy vs. the changes in sleep continuity parameters (AWASO = change in wake time after sleep onset, in min/h of total sleep period, ,~SE = change in sleep efficiency, AStab lnd = change in stability index, in sec).
Our study demonstrates a significant improvement in multiple sleep latency following long-term CPAP therapy, which is concomitant with the elimination of subjective daytime sleepiness. The improvement in alertness was paralleled by an improvement in sleep structure, similar to what had been reported after short-term CPAP (Issa and Sullivan 1986) or surgical treatment (Zorick et al. 1983). An improvement in alertness has been observed immediately after the initiation of CPAP treatment (Rajagopal et al. 1986; Di Philippe et al. 1988), or with short-term treatment (Wittig et al. 1986; Lamphere et al. 1989), but the results were not compared with those of controls. In our study, the multiple sleep latency was not normalized, as compared to unsleepy controls. A similar incomplete improvement has also been re-
25 ported with medical (Roth et al. 1980) or surgical (Zorick et al. 1983) treatment of OSA, as well as with short-term (26 days to 6 months) CPAP treatment (Gaddy and Doghrajmi 1991). This incomplete improvement contrasts with the patients' subjective feeling, since they all claimed that they were no longer sleepy; it once again underlines the fact that the subjective self-evaluation and objective evaluation of sleepiness are not parallel (Dement and Carskadon 1982). The reasons for the persistence of some degree of shortening of the multiple sleep latency are not clear, since sleep structure improved with CPAP treatment. One possible explanation would be that sleep duration remains short despite the improved sleep quality. Indeed, several patients reported that they woke up early in the morning and did not use CPAP during the early morning hours. Since we discarded the sleep period with inefficient CPAP pressure at the beginning of the night, our study does not permit an evaluation of total sleep time. However, the time during which CPAP was used at home, as evaluated from the readings of the time-counter of the CPAP device, measures the amount of normal sleep the patients actually got. Although it varied widely, this measure was not correlated with the degree of improvement of the mean sleep latency, suggesting that shortened sleep duration was not the main factor of the persistence of a decreased mean sleep latency. Another possible explanation would be that obstructive sleep apnea results in irreversible brain dysfunction. This view would be supported by the recent report of abnormal auditory event-related potentials (decreased P300 amplitude and increased P300 latency), which only partially improved with CPAP treatment (Rumbach et al. 1991). The persistence of some degree of central nervous system dysfunction after the elimination of sleep apneas may also support the early hypothesis that an underlying brain dysfunction plays a causative role in the development of obstructive sleep apnea (Lugaresi et al. 1967). The determinants of daytime somnolence are still debated. Previous studies have shown that the best predictors of daytime sleepiness in OSA patients were the sleep discontinuity and fragmentation due to respiratory events (Guilleminault et al. 1988; Roth et al. 1980; Roehrs 1989). Experimental sleep fragmentation in normal subjects induces sleepiness similar to that observed in OSA (Bonnet 1985, 1986; Stepanski et al. 1987; Magee et al. 1987). The possible role of other factors such as the reduction of slow and REM sleep and the severity of nocturnal hypoxemia has also been suggested (Orr et al. 1979; Issa and Sullivan, 1986; Chesson et al. 1988).
In our study, the improvement in daytime alertness was correlated with the improvement in sleep stability and the decrease in sleep fragmentation, as well as with the severity of sleep apnea before treatment. Thus, our data tend to support the role of sleep fragmentation due to respiratory events. The absence of a correlation between the improvement in daytime alertness and the severity of nocturnal hypoxemia does not support a role for nocturnal hypoxemia, in accordance with the results observed with medical (Roth et al. 1980) or surgical (Zorick et al. 1983) treatments of OSA. The persistence of SOREMPs was unexpected. SOREMPs have been considered a direct consequence of sleep disruption (Browman et al. 1983) and could therefore have been expected to be eliminated with the CPAP-induced normalization of sleep structure. References Bonnet, M. (1985) Effect of sleep disruption in sleep, performance and mood. Sleep, 8: 11-19. Bonnet, M. (1986) Performance and sleepiness as a function of frequency and placement of sleep disruption. Psychophysiology, 23: 263-271. Browman, C.P. (1986) Evaluation of daytime somnolence: objective measures of sleep-wake tendency. J. Electrophysiol. Techn., 13: 233-239. Browman, C.P., K.S. Krishnareddy, M.G. Sampson and M.M. Mitler (1983) REM sleep episodes during the maintenance of wakefulness test in patients with sleep apnea syndrome and patients with narcolepsy. Sleep, 6: 23-28. Carskadon, M.A. (Ed.) (1982) Current perspectives on daytime sleepiness. Sleep, 5, suppl. 2: $55-$202. Chesson, A.L., W.M. Anderson, K.C. Richards, L. Bairnsfather, S.C. Kirby and R.B. George (1988) Apnea and apnea plus hypopnea indices correlate differently with symptoms in obstructive sleep apnea than does quantitation of nocturnal hypoxemia. Sleep Res., 17: 61. Dement, W.C. and M.A. Carskadon (1982) Current perspectives on daytime sleepiness: the issues. Sleep, Suppl. 2: $56-$66. Dement, W.C., M.A. Carskadon and G. Richardson (1978) Excessive daytime sleepiness in the sleep apnea syndrome. In: Guilleminault, C. and Dement, W.C. (Eds.), Sleep apnea syndromes, Alan R. Liss, New York, pp. 23-46. DiPhilippo, M.A., J.M. Fry, M.R. Pressman (1988) Objective measurement of daytime sleepiness following treatment of obstructive sleep apnea with nasal CPAP. Sleep Res., 17: 167. Erman, M.K., B. Beckhma, D.A. Gardner and H.P. Roffwarg (1987) The modified assessment of sleepiness test (MAST). Sleep Res., 16: 550. Gady, J.R. and K. Doghramji (1991) Daytime sleepiness after nCPAP treatment. Sleep Res., 20: 245. Guilleminault, C., M. Partinen, M.A. Quera-Salva, B. Hayes, W.C. Dement and G. Nino-Murcia (1988) Determinants of daytime sleepiness in obstructive sleep apnea. Chest, 94: 32-37. Issa, F.G. and C.E. Sullivan (1986) The immediate effects of nasal continuous positive airway pressure treatment, on sleep pattern in patients with obstructive sleep apnea syndrome. Electroenceph. Clin. Neurophysiol., 63: 10-17. Lamphere, J., T. Roehrs, R. Wittig, F. Zorick, W.A. Conway and T. Roth (1989) Recovery of Alertness After CPAP in Apnea. Chest, 96: 1364-1367.
26 Magee, J., J. Harsh and P. Badia (1987) Effects of experimentally induced sleep fragmentation on sleep and sleepiness. Psychophysiology, 234: 528-534. Mitler M.M., G.K., M.G. Sampson, C.P. Browman (1982) Multiple day-time nap approaches to evaluating the sleepy patient. Sleep, 5: S119-S127. Mitler, M.M., S. Gujavarty and C.P. Browmann (1982) Maintenance of wakefulness test: a polysomnographic technique for evaluating treatment efficacy in patients with excessive somnolence. Electroenceph. Clin. Neurophysiol., 53: 658-661. Orr, W.C., R.J. Martin, N.K. lmes, R.M. Rogers and MIL. Stahl (1979) Hypersomnolent and nonhypersomnolent patients with upper airway obstruction during sleep. Chest, 75: 418-422. Rajagopal, K.R., L.L. Dillard, T.A. Bennett, C.J. Tenholder and M.F. Tellis (1986) Overnight nasal CPAP improves hypersomnolence in sleep apnea. Chest, 90: 172-176. Rechtschaffen, A. and A. Kales (1968) A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Brain Information Service. Brain Research Institute, Los Angeles, CA. Reynolds, C.F., I!I, P.A. Coble, D.J. Kupfer and B.C. Holzer (1982) Application of the multiple sleep latency test in disorders of excessive sleepiness. Electroenceph. Clin. Neurophysiol., 53: 443452. Richardson, G.S., M.A. Carskadon, W. Flagg, J. VandenHoed, W.C. Dement and M.M. Mitler (1978) Excessive daytime sleepiness in man: multiple sleep latency measurement in narcoleptics and in control subjects. Electroencepb. Clin. Neurophysiol., 45: 621-627. Roehrs, T., F. Zorick, R. Winig, W. Conway and T. Roth (1989) Predictors of objective level of daytime sleepiness in patients with sleep-related breathing disorders. Chest, 95: 1202-1206.
Roth, T., K.M. Hartse, F. Zorick and W. Conway (1980) Multiple naps and the evaluation of daytime sleepiness in patients with upper airway sleep apnea. Sleep, 3: 425-439. Rothenberg, S.A. and D.M. Rapoport (1985) Recovery from sleep deprivation in the obstructive sleep apnea syndrome (OSAS) treated with nasal continuous positive airway pressure. Sleep Res., 14: 204. Rumbach, L., J. Krieger and D. Kurtz (1991) Auditory event-related potentials in obstructive sleep apnea: effects of treatment with nasal continuous positive airway pressure. Electroenceph. Clin. Neurophysiol., 80: 454-457. Stepanski, E., J. Lamphere, P. Badia, F. Zorick and T. Roth (1984) Sleep fragmentation and daytime sleepiness. Sleep, 7: 18-26. Stepanski, E., J. Lamphere, T. Roehrs, F. Zorick and T. Roth (1987) Experimental sleep fragmentation in normal subjects. Int. J. Neurosci., 33: 207-214. Sullivan, C.E. and F.G. Issa (1985) Obstructive sleep apnea. Ciin. Chest. Med., 6: 633-650. Sullivan, C.E., F.G. Issa, M. Berthon-Jones and L. Eves (1981) Reversal of obstructive sleep apnoea by continuous positive airway pressure applied through the nares. Lancet, i: 862-865. Lugaresi, E., G. Coccagna and G. Berti Ceroni (1967) Syndrome de Pickwick et syndrome d'hypoventilation alveolaire primaire. Rev. Neurol., 116: 678-679. Wittig, R., F. Zorick, W. Conway, J. Ward and T. Roth (1986) Normalization of the MSLT after six weeks of CPAP for sleep apnea syndrome. Sleep Res., 15: 185. Zorick, F., T. Roehrs, W. Conway, S. Fujita, R. Wittig and T. Roth (1983) Effects of uvulopalatopharyngoplasty on daytime sleepiness associated with sleep apnea syndrome. Bull. Eur. Physiopathol. Respir., 19: 600-603.