Methods (mean 60) years, and chosen because they. part. They had taken no CNS drugs in the preceding

Transcription

1 Br. J. clin. Pharmac. (1982), 14, A COMPARISON OF THE EFFECTS OF CHLORMEZANONE AND NITRAZEPAM ON SLEEP KIRSTINE ADAM & I. OSWALD University Department of Psychiatry, Royal Edinburgh Hospital, Edinburgh EH1 5HF, Stland 1 Twelve volunteers, of mean age 6 years, took part in a double-blind, balanced cross-over study, to mpare effects of chlormezanone 4 mg and nitrazepam 5 mg on electrophysiologically-rerded and subjectively-rated sleep. 2 In the first week of administration nitrazepam caused a significant shortening of the time to fall asleep, but following withdrawal subjects took longer to fall asleep than during the baseline period. 3 Both chlormezanone and nitrazepam initially caused increase of sleep duration and less interruption of sleep by wakefulness. By the third week, for chlormezanone this effect was no longer significant, and for nitrazepam there was a significant decline in the effect. There was no statistically significant difference between the two drugs for these measures. 4 The drugs differed little in their effects on the amount of the various sleep stages, except that nitrazepam significantly reduced the duration of slow wave sleep, whereas chlormezanone had no significant effect on slow wave sleep. Both drugs reduced the amount of REM sleep in the first 6 h of sleep but only nitrazepam reduced the percentage of the time spent in REM sleep of the whole night. 5 Subjects' own ratings of sleep quality showed that both of the drugs improved sleep, but following withdrawal it was only after nitrazepam that there was impairment of the quality of sleep. Neither drug affected subjective alertness in the morning. Introduction Chlormezanone (Tranpal, Winthrop) has been used for some 25 years in the treatment of anxiety, tension and muscle spasm. Drowsiness by day has been reported as a side-effect (Champlin et al., 1968; Rickels et al., 1974). The present study aimed to answer two questions. Does chlormezanone, taken at bedtime, have any measurable effects on the overnight pattern of sleep and wakefulness and if so how do the effects mpare with those of nitrazepam? An hypnotic should increase the time spent asleep, reduce the amount of wakefulness interrupting sleep, patients should feel that they have slept better, and they should not be impaired throughout the next day (Oswald et al., 1979). Electrophysiological rerding of sleep makes possible the precise measurement of the duration of sleep and the distribution of sleep stages, including REM (rapid eye movement) or paradoxical sleep, and NREM (non-rem) or orthodox sleep. The latter is sub-divided into stages 1,2,3 and 4. Stages are known as slow wave sleep (SWS) and are thought by many to be 'worth more; in terms of restorative processes (Dement & Greenberg, 1966; Oswald, 198) /82/757-9 $1. Methods Selection ofsubjects Twelve volunteers, three men and nine women aged (mean 6) years, and chosen because they nsidered themselves not to be good sleepers, took part. They had taken no CNS drugs in the preceding months, and were asked not to take any drugs other than those given by us and to nsume no alhol during the study. The study was approved by the Royal Edinburgh Hospital Ethics Committee. Experimental design Chlormezanone 4 mg and nitrazepam 5 mg nightly were mpared in a cross-over design, single-blind for the placebo periods and double-blind for the drugs. Each drug was taken for a 3 week period that had been preceded by 2 weeks of matched placebos and was followed by 1 week of placebos. The order of receiving the drugs was balanced among the 12 subjects. During the first week of placebos there were two nights at the sleep laboratory for adaptation. In the next week, two nights were rerded to give baseline The Macmillan Press Ltd 1982

2 58 KIRSTINE ADAM & I. OSWALD values. In the subsequent week the first and the fifth drug nights were to give 'early drug' data. At the end of the fourth week there was a further adaptation night and in the fifth week (the third drug week) the seventeenth and twenty-first nights on the drug were rerded to give 'late drug' data. In the sixth week subjects again took placebos and the first, send and fifth withdrawal nights were rerded. After 12 weeks the 6 week sequence was repeated with the alternative treatment. Electrophysiological rerding ofsleep On all nights the electroencephalogram (EEG), eye movements (EOG) and sub-mental muscle tone (EMG) were rerded. Lights-out was at approxiimately 22.3 h and rerdings were of 8 h 45 min duration. Subjects slept in mfortable, airnditioned bedrooms. When all the rerdings had been made, the electrophysiological rerds were ded and categorized 'blind' into the different stages of sleep and wakefulness (Rechtshaffen & Kales, 1968). Thereafter the de was broken, the raw data analysed by mputer, and the mean for the 12 subjects for each variable was calculated for each of the 18 experimental nights i.e. for the two baseline, the two early drug, the two late drug, and the three withdrawal nights for each of the 6-week periods. Subjective ratings On each morning of the two 6-week sequences subjects rated the quality of their sleep by making a mark on a 1 mm line where zero represented the worst imaginable, 5 mm their own average and 1 mm the best imaginable sleep quality. Similarly they rated how alert and fresh they felt ('morning vigilance') by making a mark on another 1 mm line, where zero represented feeling 'marvellously alert and energetic', and 1 mm feeling 'awfully sleepy and lack-lustre' Subsequently we measured the distance in mm from the left hand of the line to the mark. The data for each person was averaged ( for the 14 nsecutive initial nights on placebo, (b) for the 21 nsecutive nights on drug and (c) for the first five nsecutive withdrawal nights. Statistical analysis For the purpose of statistical analysis the mean of the two baseline nights, the mean of the two early drug nights, the mean of the two late drug and, for some of the sleep measures, the mean of the three rerded withdrawal nights were first calculated for each subject for each of the drugs. Each drug was analysed separately by an analysis of variance with repeated measures to test the overall significance of the differences among the four sets of means for each subject, i.e. the analysis took into acunt that it was the same twelve subjects that were mpared under different nditions. In addition the baseline mean and the three individual withdrawal nights following each drug-taking period were mpared by analysis of variance for total sleep time, total wakefulness after sleep onset and sleep onset latency, in order to reveal any disturbance of sleep relative to baseline levels caused by withdrawing the active drug. Sendly, t-tests for paired observations, with 11 degrees of freedom, were used to mpare the different periods. All P-values quoted are based on the 2-tailed level of significance. Most of the data were analysed using these parametric tests. However, neither values for sleep onset latencies nor REM sleep latencies are normally distributed. Parametric tests can only be used to analyse the sleep onset latency data if the data values are first 'normalized' by transformation into natural logarithms. REM sleep latency is bimodal in its distribution and so the nonparametric Friedman's analysis of variance by ranks was used to test the overall significance of the difference among the four periods. Thereafter, the Wilxon matched pairs signed ranks test was used to mpare the different periods. To mpare the drugs an analysis of variance with repeated measures was used to test for any significant differences between the two drugs. Thereafter the data for each drug was transformed into the differences from the baseline mean. These mean differences for the twelve subjects for chlormezanone were mpared with the rresponding mean differences from baseline means for nitrazepam, using a rrelated t-test. In the case of REM latency the Wilxon matched pairs signed ranks test was used. Results On analysis of variance both chlormezanone and nitrazepam were each found to have a highly significant effect on the total time spent asleep (Table 1). Correlated t-tests revealed that, for chlormezanone, the source of this significant difference was mainly due to a significant increase in total sleep from the baseline to the early drug period (t = 3.91, P <.1). The increase in total sleep time from baseline, during the late drug period failed to reach statistical significance on a 2-tailed test (t = 1.97) although there was no significant difference between the early and late drug periods. In the case of nitrazepam both the early drug (t = 4.68, P <.1) and the late drug (t = 2.37, P <.5) total sleep durations were significantly longer than the baseline values. However, the late nitrazepam sleep durations were significantly shorter than those

3 THE EFFECTS OF CHLORMEZANONE AND NITRAZEPAM ON SLEEP 59 Table 1 Comparison of chlormezanone (4 mg) and nitrazepam (5 mg) nightly on wholenight sleep measures. Mean results + s.d. (or range) of twelve subjects for two nights each during baseline, early and late drug and three nights during the withdrawal period. (min) Total wakefulness after sleep onset (min) Sleep onset latency* (mean min and range) REM latency (mean min and range) (min) Total wakefulness after sleep onset (min) Sleep onset latency* (mean min and range) REM latency (mean min and range) Chlormezanone Baseline Early drug Late drug ± ± ± ± ± ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Nitrazepam Analysis Baseline Early drug Late drug Withdrawal of variance df = 3, ± ± ±3.5 F = P < ± F = 8.66 P < ( ) 27.2 ( ) 33.4 ( ) 52.6 ( ) F = 9.52 P < Xr2 = 1.28, ( ) ( ) ( ) ( ) df = 3 NS * Raw data transformed into natural logarithms prior to statistical analysis Analysis With- of variance drawal df = 3, F = 7.81 ±55.1 P< F = 6.44 ±47.6 P <.2 F= 1.86 NS XIr2 = 9., df = 3 P<.5 ANOVA used to mpare the drugs (df = 1,11) showed no significant difference between chlormezanone and nitrazepam for any of the above measures. during early nitrazepam intake (t = 2.47, P <.5). Following withdrawal, neither drug led to a significant reduction in the total length of sleep when mparisons were made using the mean withdrawal data (the average of withdrawal nights 1, 2 and 5). However, when the baseline mean and the three individual withdrawal nights (Table 2) for chlormezanone were subjected to analysis of variance a significant difference was found (F = 3.7, d,f, = 3.33, P <.4). Withdrawal night 5 was found to be significantly shorter than the baseline mean (t = 2.35, P <.4). In ntrast, the three nights rerded following the withdrawal of nitrazepam did not differ from the baseline mean (F =.55, d.f. = 3,33, NS). No significant differences were found when chlormezanone and nitrazepam were mpared with one another for their effects on total sleep time (Table 1). Total minutes ofwakefulness afterfirst sleep onset On analysis of variance both drugs were found to have a significant effect on this measure of disturbed sleep (Table 1). There was a significant reduction in wakefulness during the early chlormezanone period, when mpared with the baseline mean (t = 2.88, P <.2), with a reduction still present during late drug intake, however, the latter reduction did not reach statistical significance (t = 1.91). The late drug wakefulness was not significantly different from that rerded during the early drug period. Nitrazepam administration also led to a significant reduction in wakefulness and this was true both during the early drug (t = 4.3, p <.5) and the late drug (t = 2.52, P <.5) periods when mpared with baseline values. Inspection of Table 2 suggests that the withdrawal of chlormezanone led to an

4 6 KIRSTINE ADAM & I. OSWALD Table 2 Comparison of the mean of the two baseline nights prior to and the three individual withdrawal nights rerded following a three week period of taking chlormezanone 4 mg nightly and similarly before and after nitrazepam 5 mg nightly for 3 weeks. Mean min + s.d. (or range) for twelve subjects. Total wakefulness after sleep onset Sleep onset latency (and range) Baseline mean ± Chlormezanone Withdrawal nights ±42.8 ± ± ± ( ) 38.4 ( ) 36.9 ( ) 54.5 ( ) Total wakefulness after sleep onset Sleep onset latency (and range) Nitrazepam Baseline mean I Withdrawal nights ± ± ± ± ( ) 53.4 ( ) 53.7 ( ) 5.8 ( ) ANOVA used to mpare the drugs (df = 1,11) showed no significant difference between chlormezanone and nitrazepam for any of the above measures. increase in the amount of total wakefulness after sleep onset. However, when the baseline mean for chlormezanone was mpared with the individual withdrawal nights 1, 2, and 5, using analysis of variance, it was found that there was no significant effect of withdrawing the drug (F = 1.34, d.f. = NS). Similarly, when the baseline mean for nitrazepam was mpared with the three individual withdrawal nights using analysis of variance, differences were again non-significant (F =.12, d.f. = 3,33, NS). There was no significant difference between the two drugs in their effects on this measure of broken sleep (Table 1). Sleep onset latency Chlormezanone had no statistically significant effect on sleep onset latency (Table 1), whereas nitrazepam was asociated with a significant reduction in latency during the early drug period (t = 3.48, P <.1). Withdrawal of nitrazepam led to a significant lengthening of sleep onset latency above baseline values (t = 2.89, P <.2). These significant effects of nitrazepam on sleep latency did not differ significantly from the pattern found with chlormezanone. REM sleep latency Nitrazepam had no significant effect on the number of minutes of sleep from the first onset of sleep to the first appearance of REM sleep (Table 1). Chlormezanone, however, did have a significant effect; mainly due to a reduction in REM latency between the late drug and the withdrawal periods (T = 9, P <.2, Wilxon test). There was no significant difference between the drugs in their effects on REM sleep latency (using Wilxon test). The distribution ofsleep stages The analysis of the time spent in the different stages of sleep was done in three ways: ( the total minutes spent in each stage, (b) the percentage of total sleep spent in each stage, (c) the number of minutes of each stage in the first 6 h of accumulated sleep. Table 3 shows the total number of minutes spent in the different stages of sleep. During administration of both drugs there was a significant increase in stage 2 sleep with a significantly greater increase for nitrazepam than for chlormezanone (P <.5). Nitrazepam, but not chlormezanone administra-

5 THE EFFECTS OF CHLORMEZANONE AND NITRAZEPAM ON SLEEP 61 CA :F c1 4) cn oo 4) +1. 4) -C CU. C). 4), EI I S: 'E IQ -~~~~ O!u ^ No6 4 oo Z lv 21 N toon +1 IV o + N- e el +1 > en+l +l +l Xn CD O W) NO N ~r~1+ l %. "l Ii Z 1 z. 5 _. 4s +1 ' i.f] or " b o.. o oo In "t r- m +1 CN+l W W I O lt %mr 11-1! A I- L4t +1 t1l A, O, cc-o m- +1 " +l +1 +l "Ct 2.S: +. V- en >- 4) 4) 9 *a4o t tion led to a decrease in total min of slow wave sleep (stages 3 + 4), sustained into the withdrawal period mpared with the pre-nitrazepam baseline period (t = 3.59, P <.5). The two drugs were significantly different in their effects on SWS (P <.3). Nitrazepam was associated with a reduction in SWS between the baseline and early nitrazepam nights (mean difference, min) and this reduction was significantly greater (t = 2.29, P <.5) than the change from baseline to early chlormezanone (+ 7.2 min). Similarly, the reduction in SWS from the baseline to the later nitrazepam period ( min) was significantly different (t = 2.42, P <.4) from the baseline-late chlormezanone differences (+ 2.7 min). Table 4 shows that when the sleep stage data are expressed as percentages of total sleep time then the picture alters somewhat. Both drugs led to a reduction in stage 1 (drowsiness) percentage, while nitrazepam led to a higher percentage of stage 2 and a lower percentage of slow wave sleep (SWS) and REM sleep. The percentage of total sleep time in SWS was significantly lower with nitrazepam than with chlormezanone (P <.15) and nversely the percentage stage 2 was significantly higher when nitrazepam was given (P <.3). Table 5 displays the different stages of sleep in the first 6 h of accumulated sleep. Drowsiness (stage 1) was reduced and stage 2 increased by both drugs. The amount of slow-wave sleep (stages 3 + 4) was unaffected by chlormezanone, but significantly reduced by nitrazepam. This reduction was still present relative to baseline in the 5 days following withdrawal of nitrazepam (t = 3.59, P <.1). The two drugs differed significantly in their effects on the amounts of SWS (P <.35). The differences from the baseline means were mpared and the fall in SWS from the baseline to the early period on nitrazepam was significantly different from the baseline-early drug difference on chlormezanone (t = 2.4, P <.5). There was a similar finding for the baseline-late drug difference when mpared for the two drugs (t = 2.36, P <.5). Both drugs significantly reduced REM sleep in the first 6 h of sleep. The early chlormezanone (t = 3.35, P <.1) and the late chlormezanone nights (t = 4.74, P <.1) had less REM sleep than the predrug baseline nights. Similarly on early nitrazepam (t = 3.2, P <.2) and late nitrazepam (t = 6.1, P <.1) nights there was less REM sleep in the first 6 h of sleep than during the baseline period. The two drugs did not differ in their effects on this measure. Subjective ratings Both chlormezanone and nitrazepam made subjects feel that they had slept better (Table 6 and Figure 1) and the improvement in the subjective ratings of

6 62 KIRSTINE ADAM & I. OSWALD 4-) U; -d N b. WC 2-Q l: ty ll en 11 "t _; U: 11 o; LO V cl. xo N - IM4 C Z V V.. -1 o C; 11 v V4 'I. 11 LL1V Ci. * N o F- I- - C) C; o "-V -+1 W"+I -+1+l +l C.). +l W -- ;z O e 11 Z : 4-3 Nl t3 tn8 I S 11 V 11 V W P A. CA. 11 v V. PC 1,V NO O aslt tn ON o +l C1 +l +l +1 +l +1 oo W ) o t- ot Irn V) +l +l T-_ e~ CJv _% %O ao Co C %N r 1 ON t Q r WI r%.ọri +1 I +l *C.t.a ON l WI +1 +l I- C..- B rl _ ~ < w +1 a; i +1 t-dbo~ +1 O,1 +1 t en CZ, C4 _.4. W. ON. n. n.. NQ +l -- +l +l +l. N %) tn.4 " C: ts a..^, -.6 'r; 6 cn ts k. m V z 1: ::h,!;., lq" W;4 C14 is, z z en -4 - r- T- ON W t - F o C-+l. %) C+l +1 WI Co. cn * s:. ce 1 II. -~ 1I V V Z L4. NO 11 A4 V cr% C1 C11 en C -. r- c so M en 4 "- F; t 4 C14 ' eq ' en ' +1 +l Co C~u 1 O3 fnt- r4 Nt- _- 6 e- oo o -+ 1 In) O%- en ' O t z. * *. * *. elco W +l In +l -11+l c +l.2 CO - N t) s N _ en N - N en - +l " +l +l +l ;L~ ~ -o l-l ~~ Int W? K z N en o t -N.X en N O W) e o_ +1 ~ I N W C a -+l In +1 -+l +l * 4 O - e)on e. * * *.... a*-c ~+l [I c CIS _- _. _. _a V3 (A U) C1 {.C - ^ aa B Q QQ) n -

7 THE EFFECTS OF CHLORMEZANONE AND NITRAZEPAM ON SLEEP 63 best ever = 1I 71-6 H Ratings of sleep quality (MM) average night 4 F worst ever 3 F e4 Placebo Time (weeks) Figure 1 Chlormezanone (S 4 mg nightly) and nitrazepam (A 5 mg nightly) both led to a significant improvement in subjectively-rated sleep quality. Withdrawal of nitrazepam led to a significant worsening of sleep quality mpared to the baseline level. Mean daily rating (mm) for twelve subjects. Drug 4 I Placebo 6 Table 6 Subjective ratings of sleep quality and morning vigilance before, during and after chlormezanone 4 mg nightly and nitrazepam 5 mg nightly. Each value mean is for mm ± s.d. for twelve subjects Chlormezanone Nitrazepam Baseline Drug Withdrawal Analysis Baseline Drug Withdrawal Analysis mean mean mean of variance mean mean mean of variance 14 nights 21 nights S nights df = 2,22 14 nights 21 nights $ nights df = 2,22 Subjective F = F = 23.6 sleep quality ±5.8 ±9. ±7.3 P < ±8.3 P <.1 Subjective NS NS morning ± vigilance sleep quality, when the baseline period was mpared with the drug period, were statistically significant for both chlormezanone (t = 4.3, P <.1) and nitrazepam (t = 4.27, P <.1). Following withdrawal of chlormezanone, there was no significant impairment of sleep, whereas when nitrazepam was withdrawn there was a significant impairment in sleep quality mpared with baseline (t = 2.777, P <.2). However, there were no significant differences between the drugs. Neither drug led to any significant change in the subjects' assessment of morning vigilance - that is, in how alert and fresh they felt in the morning (Table 6). Discussion The volunteers were selected because they thought they were poor sleepers and to mirror those who frequently take sleeping pills, in that they were middle-aged and included more women than men (McGhie & Russell, 1962). The baseline rerdings show that, as a group, they took about 4 min to fall asleep. Their mean total sleep time was about 7 h, a figure that should cause no surprise as objective measures of poor sleep are usually less dramatic than subjective opinions (Monroe, 1967; Carskadon et al., 1976).

8 64 KIRSTINE ADAM & I. OSWALD In general, hypnotic and anxiety-relieving drugs are one and the same, the supposed differences being a matter of nvention in dose and hour of intake. Chlormezanone has long been used for the treatment of tension states and of musculoskeletal pain, but our results suggest that if in such ways it has been useful, it may have been because of effects that are shared with hypnotic drugs. The drugs we studied were both effective in enhancing sleep, causing increase of total sleep and decrease of wakefulness. In the dosage used, the nitrazepam appeared more potent. Sleep latency, always subject to wide variation, was reduced by both drugs, and by about 1 min, but this reduction only reached significance for nitrazepam. In nformity with this it was only after the withdrawal of nitrazepam that sleep latency was significantly lengthened mpared with baseline durations. Although chlormezanone showed itself in the objective and in the subjective studies to enhance sleep, it would not appear to have any advantage over nventional hypnotics. Both chlormezanone and nitrazepam were associated with some tolerance by the third week, and in the case of chlormezanone there was no longer a significant difference from baseline by that time. Nitrazepam alone suppressed SWS, a time often held to be the most restorative stage of sleep, but it is not known if a reduction in the duration of SWS by electrophysiological criteria means that the restorative value of sleep is impaired (Adam, 1979). Most hypnotic drugs reduce REM sleep (Kay et al., 1976), as did the two hypnotics of the present study in the first 6 h of sleep. It is interesting that, when the whole-night totals were analysed, neither drug had a significant effect on REM sleep, which suggests that both drugs were suppressing REM sleep in the earlier hours, with mpensation later. We have found this within-sleep rebound phenomena before for nitrazepam (Adam et al., 1976). The withdrawal of chlormezanone led to a significant reduction in total sleep time when the baseline mean was mpared with the three withdrawal nights and it was the fifth withdrawal night that was significantly shorter. Nitrazepam withdrawal was not here found to make sleep significantly shorter than during baseline, though in an earlier electrophysiological study with nitrazepam (Adam et al., 1976), we found a significant worsening of sleep following withdrawal of nitrazepam 5 mg that had been taken for 1 weeks. In the latter study we had rerded 6 nights during the withdrawal period and on average it had been the send and third withdrawal nights that had been associated with most increase in wakefulness. In the present study only nights 1, 2 and 5 were rerded and information may possibly have been missed. It is interesting that our subjects felt that their sleep was impaired following withdrawal of nitrazepam but not after withdrawal of chlormezanone, again suggesting that we may have missed the main withdrawal effects of nitrazepam with our electrophysiological rerdings. We are grateful to Shona Adam for her help in the analysis of date and Winthrop for financial support. References ADAM, K. (1979). Do drugs alter the restorative value of sleep? In Pharmalogy of the States of Alertness, eds Passouant, P. & Oswald, I., pp Oxford: Pergamon Press. ADAM, K., ADAMSON, L., BkEZINOVA, V., HUNTER, W.M. & OSWALD, I. (1976). Nitrazepam: lastingly effective but trouble on withdrawal. Br. med. J., 1, CARSKADON, M.A., DEMENT, W.C., MITLER, M.M., GUILEMINAULT, C., ZARCONE, V.P. & SPIEGEL, R. (1976). Self-reports versus sleep laboratory findings in 122 drug-free subjects with mplaints of chronic insomnia. Am. J. Psychiat., 133, CHAMPLIN, F.B., COTTER, C.F., MOSKOWITZ, M.D., ROSSMAN, M., SHEPPARD, C. & MERLIS, S. (1968). A mparison of chlormezanone, meprobamate and placebo. Clin. Pharmac. Ther., 9, DEMENT, W.C. & GREENBERG, S. (1966). Changes in total amount of stage four sleep as a function of partial sleep deprivation. Electroenceph. clin. Neurophysiol., 2, KAY, D.C., BLACKBURN, A.B., BUCKINGHAM, J.A. & KARACAN, I. (1976). Human pharmalogy of sleep. In Pharmalogy of Sleep, eds Williams, R.L. & Karacan, I., pp New York: John Wiley. McGHIE, A. & RUSSELL, S.M. (1962). The subjective assessment of normal sleep patterns. J. ment. Sci., 18, MONROE, L. (1967). Psychological and physiological differences between good and poor sleepers. J. abnorm. Psychol., 72, OSWALD, 1. (198). Sleep as a restorative process: human clues. Progress in Brain Research, eds McConnel, P.S., Boer, G.J., Romijn, H.J., Van de Poll, N.E. & Corner, M.A., 53, OSWALD, I., ADAM, K., BORROW, S. & IDZIKOWSKI, C. (1979). The effects of two hypnotics on sleep, subjective feelings and skilled performance. In Pharmalogy of the States of Alertness, eds Passouant, P. & Oswald, I., pp Oxford: Pergamon Press. RECHTSCHAFFEN, A. & KALES, A. (1968). A Manual of Standardized Terminology, Techniques and Sring System for Sleep Stages ofhuman Subjects. Washington, D.C.: U.S. Government Printing Office, NIH Publica-

9 THE EFFECTS OF CHLORMEZANONE AND NITRAZEPAM ON SLEEP 65 tion, No. 24. RICKELS, K., PEREIRA-OGAN, J.A., CASE, W.G., CSANALOSI, I., MIRMAN, M.J., NATHANSON, J.E. & PARISH, L.C. (1974). Chlormezanone in anxiety: a drug redisvered? Am. J. Psych., 131, (Received December22, 1981, accepted February 18, 1982)