Effect of flurazepam on sleep spindles and K-complexes

Electroencephalography and Clinical Neurophysiology, 4 0 ( 1 9 7 6 ) 6 7 - - 7 7 © Elsevier Scientific Publishing C o m p a n y , A m s t e r d a m - - P r i n t e d in The N e t h e r l a n d s

67

EFFECT OF FLURAZEPAM ON SLEEP SPINDLES AND K-COMPLEXES* L.C. J O H N S O N , K. H A N S O N a n d R.G. B I C K F O R D

Naval Health Research Center, San Diego, Calif., and University o f California, San Diego, Calif. 92152 (U.S.A.) ( A c c e p t e d for p u b l i c a t i o n : J u l y 29, 1 9 7 5 )

Studies of the effect of drugs on sleep, especially of hypnotics, have focused primarily on REM (rapid eye movement) sleep and, secondarily, on changes in SWS (slow-wave sleep, stages 3 and 4) (Oswald et al. 1968; Hartmann 1969; Oswald 1969; Kales et al. 1970; Kay et al. 1972). Only passing mention has been made of the drug-related changes in sleep spindles and, even more infrequently, on K-complexes. The advent of the benzodiazepines and their wide acceptance as a sedative, anticonvulsive, central muscle relaxant, and as a hypnotic has raised questions as to the effect of each new benzodiazepine c o m p o u n d on sleep. Recent studies have focused primarily on the influence of flunitrazepam (Ro 5-4200) on sleep (Gaillard et al. 1973; Kales and Scharf 1973; Monti and Altier 1973; Oliveros et al. 1973; Oswald et al. 1973; Monti et al. 1974). Flunitrazepam, like flurazepam, produced a decrease in SWS but, in contrast to flurazepam which has little influence on REM sleep, most studies reported a decrease in REM sleep though the effect varied from study to study. In addition to changes in sleep stages, all of the flunitrazepam

* This s t u d y was s u p p o r t e d in p a r t b y H o f f m a n n - - L a R o c h e Inc., Nutley, N.J., a n d b y t h e B u r e a u o f Medicine a n d Surgery, D e p a r t m e n t of t h e Navy, u n d e r Work U n i t M F 5 1 . 5 2 4 . 0 0 4 - 2 0 0 7 D A 5 G . The o p i n i o n s a n d assertions c o n t a i n e d h e r e i n are t h e private ones o f t h e a u t h o r s a n d are n o t t o be c o n s t r u e d as official or as reflecting t h e views of t h e Navy D e p a r t m e n t .

studies, except Kales and Scharf (1973), also commented on the influence of flunitrazepam on sleep spindles. With the exception of Oliveros et al. (1973), an increase in spindle activity during nonREM sleep was reported. Oliveros et al. reported no change. Of the studies cited, only Monti and Altier (1973), however, reported quantitative data. They reported that in 6 healthy adults, there was a significant increase in the number of spindles from a baseline average of 4.8 + 0.4 spindles per min to 8.2 + 0.2 per min on drug nights. Gaillard et al. (1973) also noted a decrease in "K potentials" along with the decrease in SWS. The purpose of the present study was to measure the effect of flurazepam on sleep spir/dles and K-complexes. The study utilized an automatic spindle detector device, as well as visual and computer analyses of EEG sleep records before, during, and after drug administration. We know of no report of the effect of flurazepam on spindles or K-complexes, aside from the c o m m e n t by Cravens et al. (1974) that fiurazepam significantly increased spindles in an unspecified number of subjects. We also investigated the relation of sleep spindles to b o d y movements before and during drug administration, since Sterman et al. (1970) have inferred a functional similarity between their sensory m o t o r rhythm (SMR) and sleep spindles. In awake cats, the SMR has been associated with m o t o r inhibition (Roth et al. 1967; Wyrwicka and Sterman 1968).

68 Method

Subjects All-night sleep recordings were obtained from 5 subjects (4 females and 1 male), age range 23--42, during 4 placebo baseline nights, during 7 drug nights, and during 3 placebo withdrawal nights, in a single blind study. Four subjects were also recorded 4--6 weeks after the last placebo withdrawal record. With the exception of the single follow-up recording, the subjects were told that on each night they would be given either a sleeping pill or a placebo. All subjects complained of a sleep problem. Four of the subjects complained of a sleep latency of greater than 45 min. One complained of awakening around 2 a.m. and being unable to go back to sleep for periods of up to 2 h, resulting in a total sleep time of less than 6 h. Each subject was interviewed and the sleep complaints were corroborated by sleep logs for approximately 2 weeks before being accepted and before the first all-night record was run. During drug nights, all subjects were scheduled to receive a 30 mg tablet of flurazepam HC1. Because of complaints of daytime drowsiness, the dose of subject AS was reduced to 15 mg after 3 nights. For similax complaints, the dose was reduced to 15 mg for subject JM on the 5th drug night, but 30 mg was given on the 6th and 7th nights.

EEG recording and analysis All-night sleep recordings were done in the sleep laboratory of Reginald Bickford at the University of California, San Diego. Total bed time was held constant at 7.5 h. The sleeprecording procedures recommended by Rechtschaffen and Kales (1968) were used. A left and right electro-oculogram (EOG), C3--A2, C3--O 1, EEG, and chin electromyogram (EMG) were recorded. All EEG analyses were from the recommended C3--A2 electrode placement. In addition to a paper write-out, the data were also recorded on a Vetter frequencymodulated tape recorder. An on-line com-

L.C. JOHNSON ET AL. pressed spectral somnogram (Hanson et al. 1974) was also obtained. All records from placebo baseline nights 3 and 4, drug nights 1, 2, 3 and 7, placebo follow-up night 3, and the 4--6 week postdrug follow-up night were visually scored for sleep stages using the criteria listed in Rechtschaffen and Kales (1968). The FM tapes from these same nights were analyzed by an automatic spindle detection system (Gondeck and Smith 1974; Smith et al. 1975). This detector system counted the waves occurring between 12.25-15.5 c/sec whose amplitude was at least 5 pV. A pattern recognizer in the system determined if a specified number of waves, 87.5%, met the amplitude and duration criteria within a given time period. The minimum time period for the system to detect a spindle burst was one cycle plus 0.37 (0.875 × 0.42) sec. The duration of the spindle burst continued until the activity dropped to 55% of the last 0.42 sec. An electronic counter recorded the number of spindle bursts detected and a total spindle count was obtained for each night analyzed. The duration of the spindle bursts was obtained from the spindle detector paper write-out. An example of the spindle detector write-out is presented in Fig. 1. Silverstein (1974), using this spindle detector, reported 92% agreement between human and automatic detection of sleep spindles from 8 young adults. Agreement between visual scoring and the detector was also high in this study; 93% on a baseline and 97% on a drug night from the same subject, and 92% agreement from a baseline night of a second subject. In most instances, disagreements were due to the automatic system's failure to detect spindle bursts that were identified visually. The reliability of the detector system with itself was examined by reanalysis of 5 nights. The total spindle count, or reliability, varied between 96% and 88% with a mean of 93%. While the automatic spindle detector provided a quantitative spindle count, the compressed spectral somnogram provided spectral data for comparison with the spindle detector and visual analysis. The compressed spectral

FLURAZEPAM AND SLEEP SPINDLES

69 Baseline

Night

4

Spindle Detector

Detector Filter

~

.

:

, A ......~ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ r . . . . 7r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

--

'

.~... ~, . . . . . . . . . . . . . . --, .............

~ '~

"

........... ......

:

~, ........... -,~ . . . . . .

,-

~ ,~

.... .....

sopvL 1 sec Drug

•

"wW

"

~ -

~

"T*

Night

~1 r"

. . . . .

7

~

. . . . . .

~lr-

r

~ "

so/uvl 1 sec

B Fig. 1. E x a m p l e of a u t o m a t i c spindle d e t e c t i o n f r o m same s u b j e c t d u r i n g baseline n i g h t 4 a n d drug n i g h t 7. T h e A tracing u n d e r drug n i g h t 7 was a s e c t i o n r e c o r d e d at 10 m m / s e c . B tracing was f r o m a n o t h e r s e c t i o n r e c o r d e d at 25 m m / s e c . In B, the spindle f r e q u e n c i e s are clearly e v i d e n t in b o t h t h e tracing o f the o u t p u t o f the d e t e c t o r filter a n d the raw E E G trace. T h e d o w n w a r d d e f l e c t i o n o f t h e spindle d e t e c t o r trace i n d i c a t e d t h e p r e s e n c e o f a spindle b u r s t a n d t h e d e f l e c t i o n persisted as long as t h e waves m e t t h e a n a l y z e r criteria.

somnogram is a graphic display of a night of sleep produced by the automated analysis of electrophysiological data and by on-line plotting of the results. An example of the somnogram for a baseline and drug night is presented in Fig. 2. (Not shown in Fig. 2 are the integrated EOG, EMG and EEG outputs which are also part of the somnogram.) In Fig. 2, each spectrum is based on 8 sec of primary EEG data transformed to intensities with 1/4 c/sec resolution. Each point is an EEG spectral

intensity (related to the square of the voltage) at the given frequency. Successive 8-sec spectra are plotted behind the first, using the hiddenline suppression technique (Bickford 1972). The K-complexes were scored visually using the criteria published by Johnson and Karpan (1968). Percent agreement between two raters, of the K-complexes from one subject, was 94% in this study. The primary rater's agreement with a second rater in another study, using the same K-complex criteria, was also

Baseline Night 4

Drug Night 7

3

3'~

2

2*

Iq

J,

I,,M

0

0

4

8

12

16

0

4

8

12 16

c/sec EEG Spectra Fig. 2. Somnogram, compressed spectral arrays, for the first 3 h of sleep from one subject on baseline night 4 and drug night 7. The increase in spectral intensity in the 12--14 c/sec range is clearly evident. Note the absence of spindle activity during REM periods. Spectra with clear activity present in all frequencies reflect body movements.

FLURAZEPAM AND SLEEP SPINDLES

71

94% (Johnson et al. 1975). K-complexes were scored only during stage 2 and only for the last baseline and last drug night.

Results

When compared to placebo baseline nights, drug nights showed a decrease in sleep latency (to onset of stage 2), decreased stage 1, increased total sleep time, and fewer awakenings. This confirms previous results (Kales et al. 1970). Three of the 5subjects improved on all sleep measures. Also, as noted in previous studies, t h e r e was a decrease in both absolute time and percent time of stage 4 sleep in 4 of the subjects. The 5th subject had only 2 min of stage 4 on her baseline record. Stage 2 total time and percent time increased in all 5 subjects, but there were no consistent changes in REM sleep time or REM latency. These drug-induced changes in sleep stages were still present on the 3rd placebo withdrawal night but n o t on the post-study follow-up night.

Sleep spindles As in previous studies (e.g., Johnson 1973), few spindles were scored in REM sleep by the automatic spindle detector and 12--15 c/sec peaks were seldom seen in the somnogram. Detected activity in the 12--15 c/sec band during REM were, in most instances, associated with movements. Spindle activity, however, was detected by all methods during stages 2, 3 and 4. To obtain a spindle rate per minute, the total all-night automatic spindle detector

• count was divided by the total time spent in stages 2, 3 and 4 sleep. In addition, a separate analysis was made for rate per minute in stage 2 for baseline night 4 and drug night 7. The mean spindle rate per minute and variability for the nights scored are listed in Table I. As there were no significant differences in spindle rate over the 2 baseline and the 4--6 week follow-up nights, these 3 nights were c o m b i n e d into a single baseline rate for comparison with drug nights. The mean baseline rate of 5.1 for baseline nights is similar to the 4.80 mean rate per min reported by Monti and Altier (1973) and the 4.6 rate found by Feinberg et al. (1967) for their young normal group (mean age 26) for stages 3 and 4. Lester et al. {1968) reported a rate of 6 per min for stage 2 sleep. All 5 subjects showed an increase in spindle activity during drug administration. Linear contrast analysis indicated this increase in rate over drug nights was significant (0.05 level or better). Comparison of individual drug nights with baseline indicated the increase in spindle rate was significant by the 2nd drug night. The spindle rate of the 3rd placebo withdrawal night was significantly higher than the baseline rate, indicating a carry-over of the drug effect. There was no significant difference in spindle rate between the 3rd and 7th drug nights. Since there was an increase in stage 2 during drug nights, the possibility existed that the total increase in spindle bursts might be due to the increased time in stage 2. To examine this question, the spindle rate in only stage 2 was computed for the last baseline and last

TABLE I Mean a n d variability o f spindle rate (N = 5). Baseline

X S.D. *N=4.

Nights f l u r a z e p a m

Withdrawal

Follow-up *

3

4

1

2

3

7

3

4--6 weeks

5.01 1.51

4.90 1.92

7.68 2.18

8.14 1.46

10.52 1.82

11.27 2.14

9.76 3.59

5.63 1.90

72

L.C. JOHNSON ET AL.

drug night. The mean rate, 10.7 + 2.78 per min, during stage 2 on drug night 7 was significantly higher than the 4th baseline night stage 2 rate of 4.6 ± 2.54 (t = 5.20). The spindle burst duration on the drugnight was also significantly longer, 0.90 + 0.11 sec versus 0.76 -+ 0.11 sec (t = 3.21). Silverstein (1974) r epor t e d a significant inter-night reliability (Intra-class r = 0.83) for spindle activity in his 8 subjects over 3 nights. For our 5 subjects, the rank order correlation for baseline nights 3 and 4 was 0.90 reflecting similar night-to-night stability. During drug administration, this correlation was reduced to 0.40 between baseline and last drug night and 0.45 between drug nights 3 and 7.

Relation

of spindles to body movements

To see if there was a drug-related change in b o d y movements and to see if the sleep spindle could be viewed as an inhibitor of movements, movements on the last placebo baseline and last drug nights were scored. Movements were scored visually, utilizing the occurrence of muscle potential artifacts in the EEG (Schieber et al. 1971; Naitoh et al. 1973). The movements initially were grouped into those with a duration o f less than 10 sec and those lasting longer than 10 sec. The duration of the movement, however, provided no significant differences and a single-movement score was used. The m o v e m e n t rate per minute was c o m p u t e d separately for nonREM, for REM, and for total sleep. As in previous studies (Sassin and Johnson 1968; Naitoh et al. 1973), the rate of movements was significantly higher in stage REM than in nonREM on bot h baseline and drug nights. There was, however, no difference between baseline and drug night in the rate of b o d y movements per minute for either nonREM, for REM, or for total sleep. The possible inhibitory function of spindles on movements was investigated in two ways. First, the time interval between the last spindle prior to each b o d y m o v e m e n t in nonREM was measured. Second, the n u m b e r of spindles in the 30 sec preceding each m o v e m e n t in nonREM was counted. If several brief movements

occurred, separated by less than 30 sec, the spindle rate was c o u n t e d in the 30-sec period before the first of these movements. For a comparison, 2 min before each real m o v e m e n t or each burst o f movements, a line was drawn, and, for comparison with the real movements, these lines were labeled pseudomovements (see Johnson and Lubin 1972). The time between the last spindle prior to this p s e u d o m o v e m e n t was compared with the time of the last spindle prior to the real m o v e m e n t . The n u m b e r of spindles in the 30-sec period before the p s e u d o m o v e m e n t was com pared to the n u m b e r before the real movement. There were no significant differences between real movements and pseudom ovem ent s in time to last spindle for either baseline or drug nights. There was also no difference in time between last spindle prior to either real or pseudom ovem ent s when the baseline and drug nights were compared. There was also no significant difference between pseudo- and real movements in the n u m b e r of spindles that occurred in the 30 sec prior to each. No significant differences were found when comparisons were made between baseline and drug nights. In these data, there was thus no relation between the p r o x i m i t y of a spindle burst and the occurrence of a m o v e m e n t as defined in this paper.

K-complex K-complexes were visually scored during stage 2 for the last baseline and last drug nights and a rate per minute computed. The average baseline rate of 1.48 + 0.64 per min was close to the rate of 1.21 reported by Johnson and Karpan (1968). This 1.48 per rain baseline rate was significantly higher than the 0.71 -+ 0.58 rate per min found on the last drug night (t = 4.00).

Discussion Flurazepam, like o t h e r benzodiazepines, increases the rate o f spindle bursts, decreases the n u m b e r of K-complexes, and decreases

FLURAZEPAM AND SLEEP SPINDLES the time spent in stage 4 sleep. Of specific interest in this paper was the change in spindle activity. Spindle bursts increased during the 1st drug night, and the rate of occurrence was significantly higher than baseline by the 2nd drug night. The spindle rate by the 7th drug night was, in one subject, three times that seen during nondrug sleep. The augmented spindle rate was still present on the 3rd withdrawal night, but no drug effects were found 4--6 weeks postdrug. As noted in the introduction, most studies on the influence of drugs on sleep have not closely examined changes in sleep spindles and K-complexes. The available data, however, suggest that drugs other than the benzodiazepines may also influence spindle activity. Lewis and Oswald (1969) reported accentuated spindle activity in the sleep records of 3 patients following overdose of tricyclic antidepressants. Lewis and Oswald also cite the report of T o y o d a (1964} as another instance of drug accentuated sleep spindling. T o y o d a studied the influence of chlorpromazine and imipramine. Barbiturate studies have led to conflicting results. Kay et al. (1972) reported a significant increase in spindle bursts after 4.0 ml doses of pentobarbital were given intramuscularly. They also cited Haider (1969) who reported an increase in spindles with sodium amylobarbitone. Lester et al. (1968), however, reported a significant decrease in spindle rate following 200 mg doses of secobarbital. Finally, Gross and Goodenough {1968) have noted an absence of sleep spindles during high blood alcohol levels. Johnson et al. (1970) reported an average spindle rate of 3 per min and diminished K-complexes during alcohol withdrawal in chronic alcoholics. All the various drugs used in the above studies influenced one or more of the sleep measures, but there was no consistent pattern of change. The almost universal increase in sleep spindles seen with benzodiazepines is obviously not restricted to this particular drug. Until it is clearly shown that the action of the benzodiazepines is unique, it would be premature to attempt an explanation based upon the specific

73 chemical structure or unique pharmacological action of the benzodiazepines. In this paper, we would like to offer a more general theory as to the possible function of sleep spindles and speculate as to the behavioral correlate of the increased rate following drug administration. When hypotheses as to the function of sleep spindles have been made, the hypotheses offered have implied an inhibitory function (Andersen and Sears 1964; Andersen et al. 1967; Rhodes 1969; Sterman et al. 1970; Yamadori 1971). Andersen and his colleagues (1964, 1967) based their conclusions on neurophysiological data in which the IPSP played an important role. Their data, in most instances, were recorded from individual cells in various thalamic nuclei in anesthetized cats. Rhodes (1969) based his conclusions on the increased spindle activity in retardates, and Sterman et al. (1970) supported their inhibitory hypothesis by noting the correlation of the 12--14 c/sec SMR with the suppression of movement in the awake cat (Roth et al. 1967; Wyrwicka and Sterman 1968). Based upon the finding that cats given SMR training when awake had more spindling when asleep, Sterman et al. (1970) stated "the neural mechanism c o m m o n to the SMR and the sleep spindle burst is concerned with the suppression of movement, as indicated by the absence of phasic m o t o r behavior when these rhythms are present in wakefulness and in sleep" (p. 1148). We found no association between the rate of sleep spindle activity and number of movements during sleep; further, we found that the presence of a spindle did not delay or inhibit the occurrence of a movement during both nondrug and drug-induced sleep. Many movements followed within I sec of a spindle burst. While Gaillard et al. (1973) reported a decrease in b o d y movements simultaneous with an increase in sleep spindles, Lester et al. (1968) reported fewer b o d y movements during sleep with decreased sleep spindle rate. Though our N is small and we used only movements clearly delineated by EEG artifact, instead of an inhibition of m o t o r activity, we

74

would like to offer the hypothesis that sleep spindles serve to raise the arousal threshold which, in turn, helps to maintain sleep by reducing the probability of being awakened. We believe that the onset of the unique sleep state is closely associated with the appearance of the first spindle (Johnson 1973). Our hypothesis as to the function of spindles is similar to that of Yamadori (1971), who has stated that "the underlying spindle producing mechanism has an active role in the sleep inducing process". In support of this hypothesis, Yamadori reported that auditory stimulation given synchronously with spindles failed to produce any effect on the transition to stages 3 and 4 and did not evoke K-complexes. In contrast, stimuli presented in the absence of spindles, i.e., during interspindle periods, elicited clear K-complexes and delayed the subject's entering stages 3 and 4. The subject remained in stage 2 or briefly went to stage 1. Spindles thus appear to block or inhibit the stimulus arousal effect, as reflected by the absence of the K-complex and by the subject's continuing his normal progression into stages 3 and 4. Both spontaneous and evoked Kcomplexes are associated with an increase in heart rate and finger vasoconstriction, b u t spontaneous spindle bursts show no such autonomic responses. Also, stimuli do n o t evoke spindle bursts (Johnson and Karpan 1968). Our speculation that the spindle mechanism is associated with an increase in arousal threshold is consistent with the subjective reports that hypnotic drug-induced sleep is "deeper", "sounder", "more restful", and with the fewer awakenings seen during sleep recordings. The absence of spindles in patients with high blood alcohol levels and the lower rate in chronic alcoholics during withdrawal would be consistent with the fragmented sleep often seen in these patients. Also consistent is the finding by Feinberg et al. (1967) that their aged normal and chronic brain syndrome subjects had more awakenings and fewer spinles than their young normal sample. The question may be raised, particularly

L.C. J O H N S O N ET AL.

with the benzodiazepines, as to why there is often an absence of "deep sleep", i.e., stage 4 sleep. The absence of stage 4 does not negate the sleeper's subjective estimates as to the quality of his sleep or to the recuperative value of the sleep. One of our subjects, age 27, had little SWS on both baseline and drug nights but reported improved sleep on drug nights. In young adult subjects, deprivation of stage 4 was not associated with waking m o o d or performance changes (Johnson et al. 1974; Lubin et al. 1974). Also, some have questioned the concept of stage 4 as deep sleep (Pisano et al. 1966; Keefe et al. 1971). In the study by Keefe et al., the dB level necessary to produce EEG signs of arousal from stage 4 was not significantly higher than that for stages 2 and I~EM. The subject's reaction time after awakening, however, was slower when awakened from stage 4 than when awakened from other stages. What constitutes good sleep is still a matter for debate, and the effect of drugs on sleep is complex. Their effects appear in different ways depending on the substance given, and the mechanism(s) of the drugs are generally unknown. We offer the above hypothesis of the function of sleep spindles as an alternative to the preoccupation with REMs and with time spent in the various sleep stages as the crucial parameters of sleep and look toward future studies to place our hypothesis in proper perspective. Summary In this research, a quantitative study of the EEG from 5 subjects permitted a detailed analysis of the effect of 30 mg of flurazepam administered over 7 nights. Four placebo baseline nights and 3 placebo withdrawal nights were also recorded. For 4 of the subjects, a nondrug and nonplacebo follow-up record was obtained 4 to 6 weeks later. The subjects were 4 females, 1 male, age range 23--42. All complained of either sleep onset greater than 45 min, sleep length of less than 6 h, or two or more sleep awakenings.

FLURAZEPAM AND SLEEP SPINDLES C o m p r e s s e d spectral analysis y i e l d e d a c o m p u t e r - g e n e r a t e d s o m n o g r a m on each o f the 15 nights o f sleep, and an a u t o m a t e d spindle d e t e c t o r was used t o c o u n t and measure the d u r a t i o n o f spindle bursts with frequencies o f 1 2 . 2 5 - - 1 5 . 5 c/sec on baseline nights 3 and 4, drug nights 1, 2, 3 and 7, o n the 3rd withdrawal night, and on the 4 - 6 w e e k followup record. K - c o m p l e x e s were scored visually o n the 4 t h baseline and 7th drug nights. T h e r e were n o significant differences in spindle rate per m i n u t e a m o n g baseline nights and the follow-up record. By the 2nd drug night, spindle rate had significantly increased over the baseline rate. L i n e a r c o n t r a s t analysis indicated t h e r e was a significant increase o f spindle rate over drug nights. All 5 subjects s h o w e d this p a t t e r n o f increase. In c o n t r a s t to the increase in spindle activity, the rate per m i n u t e o f K ~ o m p l e x e s significantly decreased during drug administration.

Rdsumd Effet du Flurazdpam sur les spindles et les complexes-K du sommeil Dans c e t t e r e c h e r c h e , une ~tude quantitative de I'EEG chez 5 sujets a permis une analyse d~taill~e des effets de 30 mg de F l u r a z e p a m administr~ p e n d a n t 7 nuits. 4 nuits de contrSle avec p l a c e b o et 3 nuits de sevrage du p l a c e b o o n t ~galement ~t~ enregistr~es. Chez q u a t r e de ces sujets, un e n r e g i s t r e m e n t sans d r o g u e et sans p l a c e b o a ~t~ r~alis~ 4 ~ 6 semaines plus tard. Les sujets c o m p o r t a i e n t 4 f e m m e s et 1 h o m m e , ~gds de 23 ~ 42 ans. T o u s se plaign a i e n t soit d ' u n t e m p s d ' e n d o r m i s s e m e n t sup& rieur fi 45 min, o u d ' u n t e m p s de sommeil inf~rieur ~ 6 h soit de 2 r~veils o u plus en cours de sommeil. L'analyse spectrale c o m p r i m ~ e a b o u t i t ~ u n h y p n o g r a m m e g~n~r~ par un o r d i n a t e u r p o u r c h a c u n e des 15 nuits de sommeil; un d ~ t e c t e u r automatis~ des spindles est utilis~ p o u r c o m p t e r et p o u r m e s u r e r la dur~e des bouff~es de spindles sup~rieures ~ 0,42 sec e t se carac-

75 t~risant par des fr~quences de 1 2 , 2 5 - - 1 5 , 5 c/sec, au cours des nuits de contrSle 3 et 4, des nuits avec d r o g u e 1, 2, 3 et 7, de la 3~me nuit de sevrage et de l ' e n r e g i s t r e m e n t pratiqu~ 4--6 semalnes u l t ~ r i e u r e m e n t . Les c o m p l e x e s K s o n t quantifi~s visuellement lors de la 4~me nuit de c o n t r S l e et de la 7~me nuit avec drogue. I1 n ' y a pas de d i f f e r e n c e significative dans le t a u x de spindles par m i n u t e au cours des nuits de c o n t r S l e e t de l ' e n r e g i s t r e m e n t ult& rieur. Au cours de la 2~me nuit avec drogue le t a u x de spindles a u g m e n t e de fa~;on significative par r a p p o r t au t a u x de contr61e. Une analyse lin~aire de constraste indique qu'il y a une a u g m e n t a t i o n significative du t a u x de spindles au cours des nuits avec drogue. Ce p a t t e r n d ' a u g m e n t a t i o n s'observe p o u r tous les sujets. E n c o n t r a s t e avec l ' a u g m e n t a t i o n de l'activit~ de spindles, le t a u x par m i n u t e de c o m p l e x e s - K diminue de faqon significative au cours de l ' a d m i n i s t r a t i o n de drogue. We wish to thank Jack R. Smith, University of Florida, Gainesville, for the loan of the spindle detector, and Marion Austin, Naval Health Research Center, for his technical assistance and help in analyzing data. The computer programming assistance of Mike Kalichman, University of California, San Diego, was crucial for the somnogram.

References Andersen, P., Andersson, S.A. and L~mo, T. Nature of thalamoeortieal relations during spontaneous barbiturate spindle activity. J. Physiol. (Lond.), 1967, 192: 283--307. Andersen, P. and Sears, T.A. The role of inhibition in the phasing of spontaneous thalamo-eortieal diseharge. J. Physiol. (Lond.), 1964, 174 : 459--480. Biekford, R. The compressed spectral array (CSA) -a pictorial EEG. Proceedings of the Biomedical Symposium, San Diego, 1972: 365--370. Cravens, J., Sheehan, J. and Hartmann, E. The sleep spindle: effects of ehlordiazepoxide; effeets of other conditions. In M.H. Chase, W.C. Stern and P.L. Walter (Eds.), Sleep researeh, Vol. 3. Brain Information Service/Brain Research Institute, University of California, Los Angeles, 1974: 48. Feinberg, I., Koresko, R.L. and Heller, N. EEG sleep patterns as a function of normal and pathological aging in man. J. psyehiat. Res., 1967, 5: 107--144.

76 Gaillard, J.-M., Schulz, P. and Tissot, R. Effects of three benzodiazepines (nitrazepam, flunitrazepam and bromazepam) on sleep of normal subjects, studied with an automatic sleep scoring system. Pharmakopsychiat., 1973, 6 : 207--217. Gondeck, A.R. and Smith, J.R. Dynamics of human sleep sigma spindles. Electroenceph. clin. Neurophysiol., 1974, 37 : 293--297. Gross, M.M. and Goodenough, D.R. Sleep disturbances in the acute alcoholic psychoses. Psychiatric Research Report 24, American Psychiatric Association, New York, 1968: 132--147. Haider, I. Effects of a hypnotic (sodium amylobarbitone 200 mg) on human sleep -- an electroencephalographic study. Pak. med. Forum, 1969, 4: 21--30. Hanson, K., Stockard, J.J., Kalichman, M. and Bickford, R.G. Compressed spectral s o m n o g r a m - - a multiparameter spectral sleep display. In J.I. Martin (Ed.), Proceedings of the Biomedical Symposium, San Diego, 1974: 545--548. Hartmann, E. Pharmacological studies of sleep and dreaming: chemical and clinical relationships. Biol. Psychiat., 1969, 1: 243--258. Johnson, L.C. Are stages of sleep related to waking behavior? Amer. Sci., 1973, 61: 326--338. Johnson, L.C., Burdick, J.A. and Smith, J. Sleep during alcohol intake and withdrawal in the chronic alcoholic. Arch. gen. Psychiat., 1970, 22: 406--418. Johnson, L.C. and Karpan, W.E. Autonomic correlates of the spontaneous K-complex. Psychophysiology, 1968, 4: 444--452. Johnson, L.C. and Lubin, A. On planning psychophysiological experiments. In N.S. Greenfield and R.A. Sternbach (Eds.), Handbook of psychophysiology. Holt, Rinehart and Winston, New York, 1972: 125--158. Johnson, L.C., Naitoh, P., Moses, J.M. and Lubin, A. Interaction of REM deprivation and stage 4 deprivation with total sleep loss: Experiment 2. Psychophysiology, 1974, 11: 147--159. Johnson, L.C., Townsend, R.E. and Wilson, M.R. Habituation during sleeping and waking. Psychophysiology, 1975, 12: 574--584. Kales, A., Kales, J.D., Scharf, M.B. and Tan, T.-L. Hypnotics and altered sleep--dream patterns. II. All-night EEG studies of chloral hydrate, flurazepare, and methaqualone. Arch. gen. Psychiat., 1970, 23: 219--225. Kales, A. and Scharf, M.B. Sleep laboratory and clinical studies of the effects of benzodiazepines on sleep: flurazepam, diazepam, chlordiazepoxide, and R0 5-4200. In S. Garattini, E. Mussini and L.O. Randall (Eds.), The benzodiazepines. Raven Press, New York, 1973: 577--598. Kay, D.C., Jasinski, D.R., Eisenstein, R.B. and Kelly, O.A. Quantified human sleep after pentobarbital. Clin. Pharmacol. Ther., 1972, 13: 221--231.

L.C. JOHNSON ET AL. Keefe, F.B., Johnson, L.C. and Hunter, E.J. EEG and autonomic response pattern during waking and sleep stages. Psychophysiology, 1971, 8: 198--212. Lester, B.K., Coulter, J.D., Cowden, L.C. and Williams, H.L. Secobarbital and nocturnal physiological patterns. Psychopharmacologia (Berl.), 1968, 13: 275--286. Lewis, S.A. and Oswald, I. Overdose of tricyclic antidepressantsand deductions concerning theircerebral action. Brit. J. Psychiat., 1969, 115: 1403--1410. Lubin, A., Moses, J.M., Johnson, L.C. and Naitoh, P. The recuperative effects of REM sleep and stage 4 sleep on human performance after complete sleep loss: Experiment 1. Psychophysiology, 1974, 11: 133--146. Monti, J.M. and Altier, H. Flunitrazepam (Ro 5-4200) and sleep cycle in normal subjects. Psychopharmacologia (Berl.), 1973, 32: 343--349. Monti, J.M., Trenchi, H.M., Morales, F. and Monti, L. Flunitrazepam (Ro 5-4200} and sleep cycle in insomniac patients. Psychopharmacologia (Beri.), 1974, 35: 371--379. Naitoh, P., Muzet, A., Johnson, L.C. and Moses, J. Body movements during sleep after sleep loss. Psychophysiology, 1973, 10: 363--368. Oliveros, J.C., Peraita, M.R., Reyero, F., Calleja, J., Vela, A. and Oliveros, A. Comparative study of the electroneurophysiological parameters of sleep induced by several hypnotic drugs. Electroenceph. clin. Neurophysiol., 1973, 34: 753. Oswald, I. Human bran protein, drugs and dreams. Nature (Lond.), 1969, 223: 893--897. Oswald, I., Jones, H.S. and Mannerheim, J.E. Effects of two slimming drugs on sleep. Brit. reed. J., 1968, 1: 796--799. Oswald, I., Lewis, S.A., Tagney, J., Firth, H. and Haider, I. Benzodiazepines and human sleep. In S. Garattini, E. Mussini and L.O. Randall (Eds.), The benzodiazepines. Raven Press, New York, 1973: 613--625. Pisano, M., Rosadini, G., Rossi, G.F. and Zattoni, J. Relations between threshold of arousal and electroencephalographic patterns during sleep in man. Physiol. and Behav., 1966, 11: 55--58. Rechtschaffen, A. and Kales, A. (Eds.), A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Public Health Service Publ. 204, U.S. Government Printing Office, Washington, D.C., 1968. Rhodes, J.M. Electrical activity of the brain as a reflection of inhibitory phenomena. Neuropsychologia, 1969, 7: 209--216. Roth, S.R., Sterman, M.B. and Clemente, C.D. Comparison of EEG correlates of reinforcement, internal inhibition, and sleep. Electroenceph. clin. Neurophysiol., 1967, 23 : 509--520. Sassin, J.F. and Johnson, L.C. Body motility during

FLURAZEPAM AND SLEEP SPINDLES sleep and its relation to the K-complex. Exp. Neurol., 1968, 22: 133--144. Schieber, J.P., Muzet, A. and Ferriere, P.J.R. Les phases d'activation transitoire spontandes au cours du sommeil normal chez l ~ o m m e . Arch. Sci. physiol., 1971, 25: 443--465. Silverstein, L.D. Temporal and intra-stage distribution of spindle activity during sleep. MA thesis, University of Florida, 1974. Smith, J.R., Funke, W.F., Yeo, W.C. and Ambuehl, R.A. Detection of human sleep EEG waveforms. Electroenceph. clin. Neurophysiol., 1975, 38: 435--437. Sterman, M.B., Howe, R.C. and Macdonald, L.R.

77 Facilitation of spindleburst sleep by conditioning of electroencephalographic activity while awake. Science, 1970, 167: 1146--1148. Toyoda, J. The effects of chlorpromazine and imipramine on the human nocturnal sleep electroencephalogram. Folia psychiat, neur. jap., 1964, 18: 198--221. Wyrwicka, W. and Sterman, M.B. Instrumental conditioning of sensorimotor cortex EEG spindles in the waking cat. Physiol. Behav., 1968, 3: 703-707. Yamadori, A. Role of the spindles in the onset of sleep. Kobe J. med. Sci., 1971, 17: 97--111.