- Email: [email protected]
Relationship between muscle tone changes, sawtooth waves and rapid eye movements during sleep
ELSEVIER
Electroencephalography and clinical Neurophysiology 103 (1997) 627-632
Relationship between muscle tone changes, sawtooth waves and rapid eye movements during sleep S u s u m u S a t o a'*, C h a r l o t t e M c C u t c h e n a, B o n n i e G r a h a m a, A m a n d a
F r e e m a n a,
I r e n e v o n A l b e r t i n i - C a r l e t t i a, D a v i d W . A l l i n g b aEEG Section, National Institute of Neurological Disorders and Stroke, Bethesda, MD 20892-1404, USA bOffice of Director, Clinical Center, National Institutes of Health, Bethesda, MD, USA
Accepted for publication: 13 June 1997
Abstract 'Sawtoothed' waves occur during rapid eye movement (REM) sleep, but their source as well as their function is not known. We studied the onset of sawtooth waves (STW) in relation to the onset of muscle tone reduction (MTR) and to REM during 20 polysomnographic recordings in 11 normal volunteers. Seventy-seven (85%) of a total of 91 REM sleep cycles were used for further analysis, because within a 1000 s interval they showed a relatively stereotyped sequence of events: a generalized body movement, followed by MTR, then the appearance of STW, and ending with the first REM. The first STW often occurred during the electrographic stage II period prior to the beginning of the REM, by which time muscle tone was either already at or close to the tone level of the REM sleep period in all subjects. The overall mean onset time of MTR was 267 s (range 89-660 s), STW was 378 s (range 169-779), and REM was 448 s (range 265-849); their differences were highly significant (F = 27.1, df 2,20, P < 0.001). Our data suggest that a predictable sequence of muscle tone reduction, then STW, and then REM, precedes the generally accepted onset of the REM sleep period, and may have implications for the redesignation of physiologic REM sleep onset. © 1997 Elsevier Science Ireland Ltd. Keywords: Sawtooth wave; REM; Muscle tone reduction
1. Introduction During rapid eye movement (REM) sleep, 'sawtoothed' waves are known to occur in close relationship to the onset of REM (Rechtschaffen and Kales, 1968); however, the source of these waves is not known. Sawtoothed wavelike activity has been recorded beginning at 5 weeks of age in normal babies (Curzi-Dascalova, 1997), and at the cortical level coinciding with the ponto-geniculo-occipital (PGO) activity in P a p i o s a n u b i s (Balzamo, 1980). Some investigators have speculated that the PGO spikes that have been routinely recorded in experimental animals might be related to human sawtooth waves (Ishiguro et al., 1979). Topographic analysis of sawtooth waves in humans, however, has shown that they appear to have max* Corresponding author. Tel.: +1 301 4965121; fax: +1 301 4028796; e-mail: [email protected]
00134694/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0013-4694(97)00072-2
imum presentation at the vertex region (Yasoshima et al., 1984; Broughton and Hasan, 1995), which is not typical of PGO spikes. To our knowledge, there have been no reports in which sawtooth waves have been examined in relation to changes in muscle tone. We have studied the onset of sawtooth waves in relation to the onset of muscle tone changes and to REM during polysomnographic recordings in normal volunteers.
2. Subjects Eleven normal volunteers (5 males, 2 2 - 3 7 years, mean 29.4 years; 6 females, 2 1 - 3 2 years, mean 26.5 years) entered the study after signing the NIH Institutional Review Board approved consent. None of them had a history of significant illness or drug use. All subjects were neurologically normal.
EEG 97016
628
S. Sato et al. / Electroencephalograph), and clinical Neurophysiology 103 (1997) 627-632
3. Recording methods
5. Statistical methods
A Nihon Kohden EEG instrument was combined with the Semi-Automated Sleep Scoring System (SASSSY system; Telefactor, Conshohocken, PA) and an infrared video camera to monitor the subject's behavior. EEG was recorded with 7 channels (Fpl-C3, C3-O1, Fp2C4, C4-O2,T3-Cz, CzT4, C3-A2), and the electrodes were placed according to the International 10-20 system. Also recorded were: EOG, 3 channels, EMG, 3 channels covering cheek, mental and submental musculature; leg movements, two channels (right and left anterior tibialis); EKG, one channel. Each subject had two consecutive nights of polysomnography, but only the first night of polysomnography of two subjects was available for analysis due to technical problems.
Onset times (muscle tone reduction, MTR; sawtooth wave, STW; rapid eye movement, REM) were subjected to a three-way analysis of variance (11 subjects, 3 onset types, two nights). Residuals were found to be normally distributed. Relationships between pairs of onset times were studied using linear regression and the corresponding differences consisted of two intervals (STW minus MTR and REM minus STW) which were analyzed in a 3-way layout (11 subjects, two intervals, two nights). Association was estimated using the Spearman rank correlation coefficient. Analyses of variance were carried out using PROC GLM in the SAS library (SAS Institute Inc., 1989); parameters describing a linear regression were estimated using least squares (PROC REG).
4. Analysis 6. Results For the purposes of this study, sawtooth waves, as described by Berger in 1962, were, after a full literature review, defined as follows: fronto-central, bilaterally synchronous and symmetric (Yasoshima et al., 1984), surface positive (Yasoshima et al., 1984), 2-5 Hz waves with a mean duration of 252 ms (Yasoshima et al., 1984), usually in a train of 3 or more consecutive waves with an angle larger than 80 ° (10 mm/s, 7 mm/50 #V) to the baseline, with an amplitude of 20-100/~V, and with a slow incline to a negative peak with a consecutive steep linear decline ending in a positive peak (Geisler et al., 1987; Tafti et al., 1991). They are most prominent at the central vertex (Yasoshima et al., 1984; Broughton and Hasan, 1995). Also for the purposes of this study, the REM were required to be unequivocal with regard to amplitude and steep angulation. After identifying the first unambiguous REM of a REM sleep episode, each investigator then analyzed the preceding epochs to identify the first appearance of sawtooth waves and the onset of muscle tone changes, stopping the retrograde analysis at the point where the last major generalized body movement was initiated in the stage II sleep prior to the REM sleep episode being analyzed. For this study the onset of the muscle tone reduction was defined as the beginning of a sustained muscle tone level of REM sleep. We then calculated the intervals in seconds from the onset of the generalized body movement in question to the onsets of the muscle tone change, sawtooth wave, and REM. Although the morphology of the sawtooth waves could be better appreciated with a 30 mm/s sweep speed, standard polysomnographic speed (10 mm/s) was used for our analysis of the onsets of changes in muscle tone, sawtooth wave, and REM. The recordings were visually analyzed on a computer screen according to Rechtshaffen and Kales' criteria with liberal but consistent modifications. Two investigators (S.S., C.M.) independently scored the polysomnographic recordings.
A total of 20 nocturnal polysomnographic recordings in 11 normal subjects yielded 91 REM periods. Each subject had an average of 4 REM cycles per night. Seventy-seven (85%) out of 91 of REM cycles showed a sequence of events (from onset of the body movement to the onset of REM) with an interval of less than 1000 s, and were used for further analyses. The remaining 14 REM cycles were excluded from further calculation: of these, 9 occurred at an interval longer than 1300 s, and were classified as outliers because one or two such cycles during a single night would skew the mean values significantly, and 5 were not preceded by detectable sawtooth waves (two of the latter were also outliers). The first sawtooth waves sleep occurred often during the electrographic stage II period prior to REM, and at a frequency of 2 - 4 Hz, positive frontocentrally, in trains of 1.59 s (mean 3.6 s). Although usually the muscle tone changes preceded the first sawtooth wave, sometimes muscle tone reduction and sawtooth waves occurred more or less simultaneously, while, even less frequently, the sawtooth waves preceded the onset of muscle tone reduction. The trains of sawtooth waves occurred 1-7 times (mean 2.4 times) prior to each REM cycle. By the time the sawtooth waves occurred, muscle tone was either already at or close to the level of the REM sleep period in all subjects. The muscle tone waxed and waned at least a few times with various degrees of reduction within a relatively short period, or took a course of gradual reduction before reaching a sustained REM level muscle tone. Fig. 1 shows a typical example of progression from body movement to the onset of REM. In Table 1 are shown the mean lengths of time (in seconds) elapsing between the last body movement during stage II sleep and respective onsets of sustained muscle tone reduction, sawtooth wave and REM. There is close agreement between the overall mean onset times measured by the two readers (t = 0.5, P -- 0.62), whereas there is a
629
S. Sato et aL / Electroencephalography and clinical Neurophysiology 103 (1997) 627-632 21 yemr-old ferule
, !
5o~v I _
irE,e-A2 ~ , ~ - . . , ~ k . ~
90seconds . . . . . . . . .
Submental EMG ~
~--~
: -- ~ : -~
It Tibia y movement I rt Tibia - - ~.eod
3 0 ~econds
u¢,
'-
-
lit ~ muacle tone
'
,~
Stwloothwaves
2ad~mu~lctofle
EKG t--t-+-Y-t-t--~t~
Sawloochwaves
R~pideyem o v e m e n t i,
Fig. 1. A
1
typical sequence of events (body movement, changes in muscle tone, sawtooth waves) prior to the onset of REM in
s i g n i f i c a n t d i f f e r e n c e b e t w e e n the overall m e a n s for the t w o n i g h t s o f s t u d y (t = 2.06, P = 0.04), the m e a n for t h e s e c o n d n i g h t b e i n g s o m e 9 0 s longer. Table 1 Mean length of time (s) between last onset of body movement during stage II sleep and sustained muscle tone reduction (MTR), sawtooth waves (STW) and rapid eye movement (REM) MTR
STW
REM
Overall mean
Reade: 1 2
248 + 37 262 + 38
348 _+40 372 + 40
434 _+42 428 _+43
343 + 24 354 _+25
Nightb 1 2
228 _+36 282 _+38
307 _+ 37 412 + 39
374 _+42 488 + 38
303 _+23 394 + 24
Each tabular entry consists of the average of 20 measurements and the corresponding standard error. Comparison of overall means: aReader l versus Reader 2: t = 0.50, P = 0.62, 90 dr. bNight 1 versus Night 2: t = 2.06, P = 0.04, 90 df.
a 21
year old female subject.
In T a b l e 2 are p r e s e n t e d the o n s e t t i m e s o f m u s c l e t o n e r e d u c t i o n , s a w t o o t h w a v e s , a n d R E M for e a c h i n d i v i d u a l subject. A l s o s h o w n are the m e a n n u m b e r o f R E M cycles ( a v e r a g e o f t w o n i g h t s ) ; at least o n e R E M c y c l e in e a c h set w a s p r e c e d e d b y a s a w t o o t h wave. E a c h o n s e t t i m e r e p r e sents the average over measurements made by two readers o n r e c o r d s c o l l e c t e d o n t w o s u c c e s s i v e nights. T h e r e are h i g h l y s i g n i f i c a n t d i f f e r e n c e s a m o n g the o v e r a l l m e a n o n s e t times: M T R 2 6 7 s ( r a n g e 8 9 - 6 6 0 ) , S T W 378 s ( r a n g e 1 6 9 - 7 7 9 ) , R E M 4 4 8 s ( r a n g e 2 6 5 - 8 4 9 ) ( F = 27.1, d f 2,20, P < 0.001). T h e 3 o n s e t t i m e s t e n d to vary f r o m s u b j e c t to s u b j e c t in a c o n c o r d a n t pattern; thus it is o f interest to c o m p u t e a n a v e r a g e o f t h e s e t i m e s to assess s u b j e c t to s u b j e c t v a r i a t i o n s . It t u r n s o u t that t h e s e a v e r a g e s s h o w h i g h l y s i g n i f i c a n t d i f f e r e n c e s a m o n g the i n d i v i d u a l s u b j e c t s ( F = 24.6, d f 10,20, P < 0.001). Regression analysis showed a straight-line relationship b e t w e e n the o n s e t o f m u s c l e t o n e c h a n g e s a n d the o n s e t o f s a w t o o t h w a v e s ; f u r t h e r m o r e , the slope o f the line w a s v e r y
S. Sato et al. /Electroencephalography and clinical Neurophysiology 103 (1997) 627-632
630 Table 2
Mean number of REM cycles and mean lengths of time (s) between last onset of body movement during stage II sleep and sustained muscle tone reduction (MTR), sawtooth waves (STW) and REM by subject Subject
No. REM cycles
MTR
STW
REM
Subject mean
1 2 3 4 5 6 7 8 9 l0 11
4.0 3.5 3.5 5.0 4.0 5.0 3.5 2.0 3.5 5.0' 4.5
268 288 292 114 222 284 354 660 213 154 89
306 626 387 342 300 325 429 779 254 169 240
342 683 474 378 371 363 587 849 297 265 317
305 532 384 278 297 324 457 763 254 196 215
Mean SE
3.95 0 27
267 46
378 54
448 56
Each tabular entry is an average of 4 measurements (one for each night and each reader).
eye movements associated with visual experiential phenomena during dreaming. W e have not been able to find any data in the literature describing clear relationships between muscle tone changes, sawtooth waves, and REM. Our data show that, on the average, in 11 normal subjects, a stereotypical sequence of these physiological events, muscle tone reduction and then sawtooth waves, occurs before the first REM, heretofore considered the onset of REM sleep. Furthermore, our data suggest that the muscle tone changes, the sawtooth waves and the REM occur in a predictable order. The muscle tone changes and sawtooth waves typically occurred a significant length of time preceding the eye movement. This is counter to the accepted 800 700 600 soo o
400 300 200
close to 45 ° (see Fig. 2a; slope = 0.98 + 0.21, SE). The same relationship held between the onset of sawtooth waves and the onset o f REM (see Fig. 2b; slope = 1.01 + 0.07). Each relationship implies that the time at which the later event occurs depends only on the timing of the earlier event and not on the length of time elapsing from the last onset of body movement• In other words, the length of the interval between the pair of events varies around a fixed constant• The question arises whether the length of the second interval ( R E M - S T W = 12 say) depends on the length of the first interval (STW MTR = I1; see Fig. 3). Although there is a very gradual slope downward (-0.06 + 0.12, SE)) this is not significant (P = 0.62), indicating that the two intervals are independently distributed. In fact, the correlation between the two intervals was found to close to zero (Spearman P = -0.04, P > 0•50)• Thus it would be expected that the length of the two intervals combined would vary around a fixed constant which would be approximately equal to the sum of the two fixed constants associated with the separate intervals. Regression analysis of the onset of REM on the onset of muscle tone reduction shows (see Fig. 2c; slope = 1.02 + 0.21) that this is indeed the case (slope = 45.5 °, constant = 176 s, compared with 183 = 66 + 117).
100
•
.
,
i
. . . .
. . . .
i
200
. . . .
i
300
. . . .
i
400
. . . .
i
500
. . . .
i
600
700
Onset of Muscle Tone Reduction (sec)
9OO 8O0 700
0
600 500 400
~)
300 200
....
, ....
100
, ........
200
300
, ....
400
, ........
500
600
700
800
OnsetofSawtoo~Waves(sec)
~'900 800 7OO ~600
O
500
~400 300
0 200
o .
.
.
.
i
.
1 O0
Berger et al. (1962) first described 'sawtoothed waves' which preceded the onset of REM and the occurrence o f sawtooth waves without REM. They speculated that the bursts of sawtooth waves and subsequent REM cluster were indicative of a sudden change of neurophysiologic condition and questioned the hypothesis that R E M represent
i
100
~ 7. Discussion
"
Y = 117 + 0.98"X
0
.
.
.
[
200
Y = 176 + 1.02*x .
.
.
.
i
300
.
.
.
.
i
400
.
.
.
.
i
.
500
.
.
.
i
600
.
.
.
.
i
700
Onset of Muscle Tone Reduction (see) Fig. 2. Regression plots. (a) Onset of sawtooth wave on the onset of muscle tone reduction; intercept = 117 + 64 (SE), slope = 0.98 + 0.21. (b) Onset of rapid eye movement on the onset of sawtooth waves; intercept = 66 + 28, slope = 1.01 + 0.07. (c) Onset of REM on the onset of muscle tone reduction; intercept = 176 + 64, slope = 1.02 + 0.21.
S. Sato et al. /Electroencephalograph), and clinical Neurophysiology 103 (1997) 627-632 o Q) ~
160
140
120
t0
100
O O
m~
80
O
E
~ N
0
~o
.o... O
4o 21)
. . . .
o i . . . .
50
= . . . .
100
|
150
. . . .
= . . . .
200
i . . . .
250
i . . . .
300
i
350
Onset STW minus onset MTR (see)
Fig. 3. Regression plot of length of time between onsets of STW and REM on length of time between onsets of MTR and STW; intercept = 77 _+ 18 (SE), slope = -0.06 _+0.12. notion that sawtooth waves occur simultaneously with the eye movements, although the importance of muscle tone changes has been mentioned (Rechtschaffen and Kales, 1968). A totally unexpected finding was that all three types of onsets tended to occur later on the second night than on the first night. This has no readily apparent explanation, but may represent an exquisite objective measure of the 'first night effect'. There is little published literature regarding sawtooth waves, and the source of these waves in humans is unknown. It has been speculated that they have a close relationship with PGO spikes regularly recorded in the lateral geniculate nuclei and generated in the upper brain-stem of cats (Ishiguro et al., 1979). In a study by the same group, sawtooth wave-like activity in the central region associated with eye movements during REM sleep in 3 narcoleptic patients was thought to be equivalent to the PGO spikes in cats, even though each had quite different distributions. Further, these authors thought that the sawtooth waves consisted of a negative sharp wave, a positive deflection and an irregular negative component, and resembled triphasic waves (Ishiguro et al., 1979). In a study of infants during active sleep (Curzi-Dascalova, 1997), sawtooth-like waves (rhythmic vertex-occipital sharp waves) were first recorded beginning at 5 weeks of age and became well defined after 11 weeks of age. However, there was no description concerning the relationship between the parameters that are the subjects of this study. The PGO waves can be recorded from the lateral geniculate body, pons, and occipital cortex, and serve as a neural marker for REM sleep in cats. The generation of PGO waves in the lateral geniculate body has been traced to presumed cholinergic neurons in the pedunculo-pontine tegmental and laterodorsal tegmental nuclei. Serotonin (5HT) appears to inhibit PGO wave generation (Sanford et al., 1994). The occurrence of numerous PGO waves in non-REM sleep usually indicates the beginning of transition to REM, but PGO wave frequency does not change very
631
much across various sleep stages, including wakefulness and REM sleep period (Sanford et al., 1994). This may have some relevance for our observation that not all REM bursts are associated with sawtooth waves even during REM sleep. In the present study, only 5 of 91 REM cycles were not preceded by sawtooth waves, suggesting that these two phenomena are closely associated with each although not completely interdependent. In a study of P a p i o anubis, at the cortical level, phasic sleep 'ponto-geniculo-cortical' activities were said to resemble the sawtooth waves described in man, but no further details were given (Balzamo, 1980). Vertex sharp wave transients of sleep differ from sawtooth waves in that their main component is surface negative. While they sometimes occur in a train of a few seconds duration with a lower amplitude and interhemispheric phase reversal, they are rarely in a train of well-formed sharp sawtoothed appearance, or in a train of several seconds in the age group we studied. Sawtooth waves classically occur in association with REM sleep, whereas vertex sharp waves classically occur during non-REM sleep. At times, the muscle tone reduction and sawtooth waves occurred more or less simultaneously, and, occasionally, sawtooth waves preceded the onset of muscle tone reduction. The majority of the time, however, the sequence of events began with sustained changes in muscle tone, followed by sawtooth wave appearance, finishing eventually with the onset of REM. Our observations suggest that this is a fairly stereotyped cascade of events heralding the onset of REM sleep, and some or all of its parts may actually constitute the real physiological beginning of REM sleep, rather than its beginning with the eye movements as previously thought. 8. S u m m a r y Two consecutive nights of polysomnographic recording in 11 normal subjects showed a fairly predictable cascade of events leading to the onset of REM sleep, consisting of muscle tone change, followed by sawtooth waves and eventually REM. This sequence of events may represent physiological processing for REM sleep substantially in advance of the eye movements themselves, and may have implications for the redesignation of REM sleep onset.
References
Balzamo, E. States of wakefulness and ponto-geniculo-corticalactivities (PGC) in Papio anubis (author's translation). Electroenceph. clin. Neurophysiol., 1980, 48: 694-705. Berger, R.J., Olley, P. and Oswald, I. The EEG, eye-movements and dreams of the blind. Q. J. Exp. Psychol., 1962, 14: 183-186. Broughton, R. and Hasan, J. Quantitative topographic electroencephalographic mapping during drowsiness and sleep onset. J. Clin. Neurophysiol., 1995, 12: 372-386.
632
S. Sato et aL / Electroencephalography and clinical Neurophysiology 103 (1997) 627-632
Curzi-Dascalova, L. Waking and sleeping EEG in normal babies before 6 months of age (author's translation). Rev. Electroencephalogr. Neurophysiol., 1997, 7(3): 316-326. Geisler, P., Meier-Ewert, K. and Matsubayashi, K. Rapid eye movements, muscle twitches and sawtooth waves in the sleep of narcoleptic patients and controls. Electroenceph. clin. Neurophysiol., 1987, 67: 499-507. Ishiguro, T., Hanamura, S. and Otaka, T. The saw-tooth wave associated with small nystagmus. A study on a narcoleptic patient and her family. Sleep Res., 1979, 8: 195. Rechtschaffen, A. and Kales, A. A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. NIH Publication No. 204, 1968. Sanford, L.D., Ross, R.J., Seggos, A.E., Morrison, A.R., Ball, W.A. and
Mann, G.L. Central administration of two 5-HT receptor agonists: effect on REM sleep initiation and PGO waves. Pharmacol, Biochem. Behav., 1994, 49: 93-100. SAS Institute Inc. SAS/STATs User's Guide, Version 6, 4th edn. SAS Institute Inc., Cary, NC, 1989, pp. 1135-1194, 1351-1456. Tafti, M., Olivet, H. and Billiard, M. Phasic events in narcolepsy and sleep apnea syndrome. In: M.G. Terzano, P.L. Halasz and A.C. Declerck (Eds.), Phasic Events and Dynamic Organization of Sleep. Raven Press, New York, 1991, pp. 151-166. Yasoshima, A., Hayashi, H., Iijima, S., Sugita, Y., Teshima, Y., Shimizu, T. and Hishikawa, Y. Potential distribution of vertex sharp wave and sawtoothed wave on the scalp. Electroenceph. clin. Neurophysiol., 1984, 58: 73-76.
Electroencephalography and clinical Neurophysiology 103 (1997) 627-632
Relationship between muscle tone changes, sawtooth waves and rapid eye movements during sleep S u s u m u S a t o a'*, C h a r l o t t e M c C u t c h e n a, B o n n i e G r a h a m a, A m a n d a
F r e e m a n a,
I r e n e v o n A l b e r t i n i - C a r l e t t i a, D a v i d W . A l l i n g b aEEG Section, National Institute of Neurological Disorders and Stroke, Bethesda, MD 20892-1404, USA bOffice of Director, Clinical Center, National Institutes of Health, Bethesda, MD, USA
Accepted for publication: 13 June 1997
Abstract 'Sawtoothed' waves occur during rapid eye movement (REM) sleep, but their source as well as their function is not known. We studied the onset of sawtooth waves (STW) in relation to the onset of muscle tone reduction (MTR) and to REM during 20 polysomnographic recordings in 11 normal volunteers. Seventy-seven (85%) of a total of 91 REM sleep cycles were used for further analysis, because within a 1000 s interval they showed a relatively stereotyped sequence of events: a generalized body movement, followed by MTR, then the appearance of STW, and ending with the first REM. The first STW often occurred during the electrographic stage II period prior to the beginning of the REM, by which time muscle tone was either already at or close to the tone level of the REM sleep period in all subjects. The overall mean onset time of MTR was 267 s (range 89-660 s), STW was 378 s (range 169-779), and REM was 448 s (range 265-849); their differences were highly significant (F = 27.1, df 2,20, P < 0.001). Our data suggest that a predictable sequence of muscle tone reduction, then STW, and then REM, precedes the generally accepted onset of the REM sleep period, and may have implications for the redesignation of physiologic REM sleep onset. © 1997 Elsevier Science Ireland Ltd. Keywords: Sawtooth wave; REM; Muscle tone reduction
1. Introduction During rapid eye movement (REM) sleep, 'sawtoothed' waves are known to occur in close relationship to the onset of REM (Rechtschaffen and Kales, 1968); however, the source of these waves is not known. Sawtoothed wavelike activity has been recorded beginning at 5 weeks of age in normal babies (Curzi-Dascalova, 1997), and at the cortical level coinciding with the ponto-geniculo-occipital (PGO) activity in P a p i o s a n u b i s (Balzamo, 1980). Some investigators have speculated that the PGO spikes that have been routinely recorded in experimental animals might be related to human sawtooth waves (Ishiguro et al., 1979). Topographic analysis of sawtooth waves in humans, however, has shown that they appear to have max* Corresponding author. Tel.: +1 301 4965121; fax: +1 301 4028796; e-mail: [email protected]
00134694/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved PII S0013-4694(97)00072-2
imum presentation at the vertex region (Yasoshima et al., 1984; Broughton and Hasan, 1995), which is not typical of PGO spikes. To our knowledge, there have been no reports in which sawtooth waves have been examined in relation to changes in muscle tone. We have studied the onset of sawtooth waves in relation to the onset of muscle tone changes and to REM during polysomnographic recordings in normal volunteers.
2. Subjects Eleven normal volunteers (5 males, 2 2 - 3 7 years, mean 29.4 years; 6 females, 2 1 - 3 2 years, mean 26.5 years) entered the study after signing the NIH Institutional Review Board approved consent. None of them had a history of significant illness or drug use. All subjects were neurologically normal.
EEG 97016
628
S. Sato et al. / Electroencephalograph), and clinical Neurophysiology 103 (1997) 627-632
3. Recording methods
5. Statistical methods
A Nihon Kohden EEG instrument was combined with the Semi-Automated Sleep Scoring System (SASSSY system; Telefactor, Conshohocken, PA) and an infrared video camera to monitor the subject's behavior. EEG was recorded with 7 channels (Fpl-C3, C3-O1, Fp2C4, C4-O2,T3-Cz, CzT4, C3-A2), and the electrodes were placed according to the International 10-20 system. Also recorded were: EOG, 3 channels, EMG, 3 channels covering cheek, mental and submental musculature; leg movements, two channels (right and left anterior tibialis); EKG, one channel. Each subject had two consecutive nights of polysomnography, but only the first night of polysomnography of two subjects was available for analysis due to technical problems.
Onset times (muscle tone reduction, MTR; sawtooth wave, STW; rapid eye movement, REM) were subjected to a three-way analysis of variance (11 subjects, 3 onset types, two nights). Residuals were found to be normally distributed. Relationships between pairs of onset times were studied using linear regression and the corresponding differences consisted of two intervals (STW minus MTR and REM minus STW) which were analyzed in a 3-way layout (11 subjects, two intervals, two nights). Association was estimated using the Spearman rank correlation coefficient. Analyses of variance were carried out using PROC GLM in the SAS library (SAS Institute Inc., 1989); parameters describing a linear regression were estimated using least squares (PROC REG).
4. Analysis 6. Results For the purposes of this study, sawtooth waves, as described by Berger in 1962, were, after a full literature review, defined as follows: fronto-central, bilaterally synchronous and symmetric (Yasoshima et al., 1984), surface positive (Yasoshima et al., 1984), 2-5 Hz waves with a mean duration of 252 ms (Yasoshima et al., 1984), usually in a train of 3 or more consecutive waves with an angle larger than 80 ° (10 mm/s, 7 mm/50 #V) to the baseline, with an amplitude of 20-100/~V, and with a slow incline to a negative peak with a consecutive steep linear decline ending in a positive peak (Geisler et al., 1987; Tafti et al., 1991). They are most prominent at the central vertex (Yasoshima et al., 1984; Broughton and Hasan, 1995). Also for the purposes of this study, the REM were required to be unequivocal with regard to amplitude and steep angulation. After identifying the first unambiguous REM of a REM sleep episode, each investigator then analyzed the preceding epochs to identify the first appearance of sawtooth waves and the onset of muscle tone changes, stopping the retrograde analysis at the point where the last major generalized body movement was initiated in the stage II sleep prior to the REM sleep episode being analyzed. For this study the onset of the muscle tone reduction was defined as the beginning of a sustained muscle tone level of REM sleep. We then calculated the intervals in seconds from the onset of the generalized body movement in question to the onsets of the muscle tone change, sawtooth wave, and REM. Although the morphology of the sawtooth waves could be better appreciated with a 30 mm/s sweep speed, standard polysomnographic speed (10 mm/s) was used for our analysis of the onsets of changes in muscle tone, sawtooth wave, and REM. The recordings were visually analyzed on a computer screen according to Rechtshaffen and Kales' criteria with liberal but consistent modifications. Two investigators (S.S., C.M.) independently scored the polysomnographic recordings.
A total of 20 nocturnal polysomnographic recordings in 11 normal subjects yielded 91 REM periods. Each subject had an average of 4 REM cycles per night. Seventy-seven (85%) out of 91 of REM cycles showed a sequence of events (from onset of the body movement to the onset of REM) with an interval of less than 1000 s, and were used for further analyses. The remaining 14 REM cycles were excluded from further calculation: of these, 9 occurred at an interval longer than 1300 s, and were classified as outliers because one or two such cycles during a single night would skew the mean values significantly, and 5 were not preceded by detectable sawtooth waves (two of the latter were also outliers). The first sawtooth waves sleep occurred often during the electrographic stage II period prior to REM, and at a frequency of 2 - 4 Hz, positive frontocentrally, in trains of 1.59 s (mean 3.6 s). Although usually the muscle tone changes preceded the first sawtooth wave, sometimes muscle tone reduction and sawtooth waves occurred more or less simultaneously, while, even less frequently, the sawtooth waves preceded the onset of muscle tone reduction. The trains of sawtooth waves occurred 1-7 times (mean 2.4 times) prior to each REM cycle. By the time the sawtooth waves occurred, muscle tone was either already at or close to the level of the REM sleep period in all subjects. The muscle tone waxed and waned at least a few times with various degrees of reduction within a relatively short period, or took a course of gradual reduction before reaching a sustained REM level muscle tone. Fig. 1 shows a typical example of progression from body movement to the onset of REM. In Table 1 are shown the mean lengths of time (in seconds) elapsing between the last body movement during stage II sleep and respective onsets of sustained muscle tone reduction, sawtooth wave and REM. There is close agreement between the overall mean onset times measured by the two readers (t = 0.5, P -- 0.62), whereas there is a
629
S. Sato et aL / Electroencephalography and clinical Neurophysiology 103 (1997) 627-632 21 yemr-old ferule
, !
5o~v I _
irE,e-A2 ~ , ~ - . . , ~ k . ~
90seconds . . . . . . . . .
Submental EMG ~
~--~
: -- ~ : -~
It Tibia y movement I rt Tibia - - ~.eod
3 0 ~econds
u¢,
'-
-
lit ~ muacle tone
'
,~
Stwloothwaves
2ad~mu~lctofle
EKG t--t-+-Y-t-t--~t~
Sawloochwaves
R~pideyem o v e m e n t i,
Fig. 1. A
1
typical sequence of events (body movement, changes in muscle tone, sawtooth waves) prior to the onset of REM in
s i g n i f i c a n t d i f f e r e n c e b e t w e e n the overall m e a n s for the t w o n i g h t s o f s t u d y (t = 2.06, P = 0.04), the m e a n for t h e s e c o n d n i g h t b e i n g s o m e 9 0 s longer. Table 1 Mean length of time (s) between last onset of body movement during stage II sleep and sustained muscle tone reduction (MTR), sawtooth waves (STW) and rapid eye movement (REM) MTR
STW
REM
Overall mean
Reade: 1 2
248 + 37 262 + 38
348 _+40 372 + 40
434 _+42 428 _+43
343 + 24 354 _+25
Nightb 1 2
228 _+36 282 _+38
307 _+ 37 412 + 39
374 _+42 488 + 38
303 _+23 394 + 24
Each tabular entry consists of the average of 20 measurements and the corresponding standard error. Comparison of overall means: aReader l versus Reader 2: t = 0.50, P = 0.62, 90 dr. bNight 1 versus Night 2: t = 2.06, P = 0.04, 90 df.
a 21
year old female subject.
In T a b l e 2 are p r e s e n t e d the o n s e t t i m e s o f m u s c l e t o n e r e d u c t i o n , s a w t o o t h w a v e s , a n d R E M for e a c h i n d i v i d u a l subject. A l s o s h o w n are the m e a n n u m b e r o f R E M cycles ( a v e r a g e o f t w o n i g h t s ) ; at least o n e R E M c y c l e in e a c h set w a s p r e c e d e d b y a s a w t o o t h wave. E a c h o n s e t t i m e r e p r e sents the average over measurements made by two readers o n r e c o r d s c o l l e c t e d o n t w o s u c c e s s i v e nights. T h e r e are h i g h l y s i g n i f i c a n t d i f f e r e n c e s a m o n g the o v e r a l l m e a n o n s e t times: M T R 2 6 7 s ( r a n g e 8 9 - 6 6 0 ) , S T W 378 s ( r a n g e 1 6 9 - 7 7 9 ) , R E M 4 4 8 s ( r a n g e 2 6 5 - 8 4 9 ) ( F = 27.1, d f 2,20, P < 0.001). T h e 3 o n s e t t i m e s t e n d to vary f r o m s u b j e c t to s u b j e c t in a c o n c o r d a n t pattern; thus it is o f interest to c o m p u t e a n a v e r a g e o f t h e s e t i m e s to assess s u b j e c t to s u b j e c t v a r i a t i o n s . It t u r n s o u t that t h e s e a v e r a g e s s h o w h i g h l y s i g n i f i c a n t d i f f e r e n c e s a m o n g the i n d i v i d u a l s u b j e c t s ( F = 24.6, d f 10,20, P < 0.001). Regression analysis showed a straight-line relationship b e t w e e n the o n s e t o f m u s c l e t o n e c h a n g e s a n d the o n s e t o f s a w t o o t h w a v e s ; f u r t h e r m o r e , the slope o f the line w a s v e r y
S. Sato et al. /Electroencephalography and clinical Neurophysiology 103 (1997) 627-632
630 Table 2
Mean number of REM cycles and mean lengths of time (s) between last onset of body movement during stage II sleep and sustained muscle tone reduction (MTR), sawtooth waves (STW) and REM by subject Subject
No. REM cycles
MTR
STW
REM
Subject mean
1 2 3 4 5 6 7 8 9 l0 11
4.0 3.5 3.5 5.0 4.0 5.0 3.5 2.0 3.5 5.0' 4.5
268 288 292 114 222 284 354 660 213 154 89
306 626 387 342 300 325 429 779 254 169 240
342 683 474 378 371 363 587 849 297 265 317
305 532 384 278 297 324 457 763 254 196 215
Mean SE
3.95 0 27
267 46
378 54
448 56
Each tabular entry is an average of 4 measurements (one for each night and each reader).
eye movements associated with visual experiential phenomena during dreaming. W e have not been able to find any data in the literature describing clear relationships between muscle tone changes, sawtooth waves, and REM. Our data show that, on the average, in 11 normal subjects, a stereotypical sequence of these physiological events, muscle tone reduction and then sawtooth waves, occurs before the first REM, heretofore considered the onset of REM sleep. Furthermore, our data suggest that the muscle tone changes, the sawtooth waves and the REM occur in a predictable order. The muscle tone changes and sawtooth waves typically occurred a significant length of time preceding the eye movement. This is counter to the accepted 800 700 600 soo o
400 300 200
close to 45 ° (see Fig. 2a; slope = 0.98 + 0.21, SE). The same relationship held between the onset of sawtooth waves and the onset o f REM (see Fig. 2b; slope = 1.01 + 0.07). Each relationship implies that the time at which the later event occurs depends only on the timing of the earlier event and not on the length of time elapsing from the last onset of body movement• In other words, the length of the interval between the pair of events varies around a fixed constant• The question arises whether the length of the second interval ( R E M - S T W = 12 say) depends on the length of the first interval (STW MTR = I1; see Fig. 3). Although there is a very gradual slope downward (-0.06 + 0.12, SE)) this is not significant (P = 0.62), indicating that the two intervals are independently distributed. In fact, the correlation between the two intervals was found to close to zero (Spearman P = -0.04, P > 0•50)• Thus it would be expected that the length of the two intervals combined would vary around a fixed constant which would be approximately equal to the sum of the two fixed constants associated with the separate intervals. Regression analysis of the onset of REM on the onset of muscle tone reduction shows (see Fig. 2c; slope = 1.02 + 0.21) that this is indeed the case (slope = 45.5 °, constant = 176 s, compared with 183 = 66 + 117).
100
•
.
,
i
. . . .
. . . .
i
200
. . . .
i
300
. . . .
i
400
. . . .
i
500
. . . .
i
600
700
Onset of Muscle Tone Reduction (sec)
9OO 8O0 700
0
600 500 400
~)
300 200
....
, ....
100
, ........
200
300
, ....
400
, ........
500
600
700
800
OnsetofSawtoo~Waves(sec)
~'900 800 7OO ~600
O
500
~400 300
0 200
o .
.
.
.
i
.
1 O0
Berger et al. (1962) first described 'sawtoothed waves' which preceded the onset of REM and the occurrence o f sawtooth waves without REM. They speculated that the bursts of sawtooth waves and subsequent REM cluster were indicative of a sudden change of neurophysiologic condition and questioned the hypothesis that R E M represent
i
100
~ 7. Discussion
"
Y = 117 + 0.98"X
0
.
.
.
[
200
Y = 176 + 1.02*x .
.
.
.
i
300
.
.
.
.
i
400
.
.
.
.
i
.
500
.
.
.
i
600
.
.
.
.
i
700
Onset of Muscle Tone Reduction (see) Fig. 2. Regression plots. (a) Onset of sawtooth wave on the onset of muscle tone reduction; intercept = 117 + 64 (SE), slope = 0.98 + 0.21. (b) Onset of rapid eye movement on the onset of sawtooth waves; intercept = 66 + 28, slope = 1.01 + 0.07. (c) Onset of REM on the onset of muscle tone reduction; intercept = 176 + 64, slope = 1.02 + 0.21.
S. Sato et al. /Electroencephalograph), and clinical Neurophysiology 103 (1997) 627-632 o Q) ~
160
140
120
t0
100
O O
m~
80
O
E
~ N
0
~o
.o... O
4o 21)
. . . .
o i . . . .
50
= . . . .
100
|
150
. . . .
= . . . .
200
i . . . .
250
i . . . .
300
i
350
Onset STW minus onset MTR (see)
Fig. 3. Regression plot of length of time between onsets of STW and REM on length of time between onsets of MTR and STW; intercept = 77 _+ 18 (SE), slope = -0.06 _+0.12. notion that sawtooth waves occur simultaneously with the eye movements, although the importance of muscle tone changes has been mentioned (Rechtschaffen and Kales, 1968). A totally unexpected finding was that all three types of onsets tended to occur later on the second night than on the first night. This has no readily apparent explanation, but may represent an exquisite objective measure of the 'first night effect'. There is little published literature regarding sawtooth waves, and the source of these waves in humans is unknown. It has been speculated that they have a close relationship with PGO spikes regularly recorded in the lateral geniculate nuclei and generated in the upper brain-stem of cats (Ishiguro et al., 1979). In a study by the same group, sawtooth wave-like activity in the central region associated with eye movements during REM sleep in 3 narcoleptic patients was thought to be equivalent to the PGO spikes in cats, even though each had quite different distributions. Further, these authors thought that the sawtooth waves consisted of a negative sharp wave, a positive deflection and an irregular negative component, and resembled triphasic waves (Ishiguro et al., 1979). In a study of infants during active sleep (Curzi-Dascalova, 1997), sawtooth-like waves (rhythmic vertex-occipital sharp waves) were first recorded beginning at 5 weeks of age and became well defined after 11 weeks of age. However, there was no description concerning the relationship between the parameters that are the subjects of this study. The PGO waves can be recorded from the lateral geniculate body, pons, and occipital cortex, and serve as a neural marker for REM sleep in cats. The generation of PGO waves in the lateral geniculate body has been traced to presumed cholinergic neurons in the pedunculo-pontine tegmental and laterodorsal tegmental nuclei. Serotonin (5HT) appears to inhibit PGO wave generation (Sanford et al., 1994). The occurrence of numerous PGO waves in non-REM sleep usually indicates the beginning of transition to REM, but PGO wave frequency does not change very
631
much across various sleep stages, including wakefulness and REM sleep period (Sanford et al., 1994). This may have some relevance for our observation that not all REM bursts are associated with sawtooth waves even during REM sleep. In the present study, only 5 of 91 REM cycles were not preceded by sawtooth waves, suggesting that these two phenomena are closely associated with each although not completely interdependent. In a study of P a p i o anubis, at the cortical level, phasic sleep 'ponto-geniculo-cortical' activities were said to resemble the sawtooth waves described in man, but no further details were given (Balzamo, 1980). Vertex sharp wave transients of sleep differ from sawtooth waves in that their main component is surface negative. While they sometimes occur in a train of a few seconds duration with a lower amplitude and interhemispheric phase reversal, they are rarely in a train of well-formed sharp sawtoothed appearance, or in a train of several seconds in the age group we studied. Sawtooth waves classically occur in association with REM sleep, whereas vertex sharp waves classically occur during non-REM sleep. At times, the muscle tone reduction and sawtooth waves occurred more or less simultaneously, and, occasionally, sawtooth waves preceded the onset of muscle tone reduction. The majority of the time, however, the sequence of events began with sustained changes in muscle tone, followed by sawtooth wave appearance, finishing eventually with the onset of REM. Our observations suggest that this is a fairly stereotyped cascade of events heralding the onset of REM sleep, and some or all of its parts may actually constitute the real physiological beginning of REM sleep, rather than its beginning with the eye movements as previously thought. 8. S u m m a r y Two consecutive nights of polysomnographic recording in 11 normal subjects showed a fairly predictable cascade of events leading to the onset of REM sleep, consisting of muscle tone change, followed by sawtooth waves and eventually REM. This sequence of events may represent physiological processing for REM sleep substantially in advance of the eye movements themselves, and may have implications for the redesignation of REM sleep onset.
References
Balzamo, E. States of wakefulness and ponto-geniculo-corticalactivities (PGC) in Papio anubis (author's translation). Electroenceph. clin. Neurophysiol., 1980, 48: 694-705. Berger, R.J., Olley, P. and Oswald, I. The EEG, eye-movements and dreams of the blind. Q. J. Exp. Psychol., 1962, 14: 183-186. Broughton, R. and Hasan, J. Quantitative topographic electroencephalographic mapping during drowsiness and sleep onset. J. Clin. Neurophysiol., 1995, 12: 372-386.
632
S. Sato et aL / Electroencephalography and clinical Neurophysiology 103 (1997) 627-632
Curzi-Dascalova, L. Waking and sleeping EEG in normal babies before 6 months of age (author's translation). Rev. Electroencephalogr. Neurophysiol., 1997, 7(3): 316-326. Geisler, P., Meier-Ewert, K. and Matsubayashi, K. Rapid eye movements, muscle twitches and sawtooth waves in the sleep of narcoleptic patients and controls. Electroenceph. clin. Neurophysiol., 1987, 67: 499-507. Ishiguro, T., Hanamura, S. and Otaka, T. The saw-tooth wave associated with small nystagmus. A study on a narcoleptic patient and her family. Sleep Res., 1979, 8: 195. Rechtschaffen, A. and Kales, A. A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. NIH Publication No. 204, 1968. Sanford, L.D., Ross, R.J., Seggos, A.E., Morrison, A.R., Ball, W.A. and
Mann, G.L. Central administration of two 5-HT receptor agonists: effect on REM sleep initiation and PGO waves. Pharmacol, Biochem. Behav., 1994, 49: 93-100. SAS Institute Inc. SAS/STATs User's Guide, Version 6, 4th edn. SAS Institute Inc., Cary, NC, 1989, pp. 1135-1194, 1351-1456. Tafti, M., Olivet, H. and Billiard, M. Phasic events in narcolepsy and sleep apnea syndrome. In: M.G. Terzano, P.L. Halasz and A.C. Declerck (Eds.), Phasic Events and Dynamic Organization of Sleep. Raven Press, New York, 1991, pp. 151-166. Yasoshima, A., Hayashi, H., Iijima, S., Sugita, Y., Teshima, Y., Shimizu, T. and Hishikawa, Y. Potential distribution of vertex sharp wave and sawtoothed wave on the scalp. Electroenceph. clin. Neurophysiol., 1984, 58: 73-76.