- Email: [email protected]
Characterization of the mu rhythm during rapid eye movement sleep
Clinical Neurophysiology 112 (2001) 528±531
www.elsevier.com/locate/clinph
Brief communication
Characterization of the mu rhythm during rapid eye movement sleep Stephen P. Duntley a, Albert H. Kim a, Daniel L. Silbergeld b, John W. Miller b,c,* a
Department of Neurology and Neurological Surgery, Washington University School of Medicine, St. Louis, MO, USA b Department of Neurosurgery, University of Washington, Seattle, WA, USA c Department of Neurology, University of Washington, Seattle, WA, USA Accepted 15 December 2000
Abstract Objective: The rolandic mu rhythm, a resting activity of somatosensory cortex, is a striking feature of the waking human electroencephalogram. This study will demonstrate that activity with identical features occurs during rapid eye movement (REM) sleep. Methods: Eye and chin leads were added during prolonged closed circuit television (video) electroencephalographic (EEG) recording with scalp (12 patients) or subdural electrodes including 64 contract grids over the frontoparietal cortices (5 patients). Sleep staging was performed by reformatting into standard polysomnography montages (using two EEG channels, and eye and chin channels) and applying standard scoring criteria. The recordings were then reviewed using all EEG channels to assess rhythmic EEG activity by a reader blinded to the sleep staging. Results: During scalp recordings, 7±10 Hz central rhythms were seen during wakefulness in 7 patients, with 6 of these also having similar rhythms during REM sleep. Similar activity was seen over somatosensory cortex during wakefulness and REM in all invasively recorded patients. This activity was blocked by contralateral body movement or contralateral somatosensory stimuli, even during REM sleep. It was absent in other sleep stages. Conclusions: This REM sleep activity recapitulates all the characteristics of the waking rolandic mu rhythm. This demonstrates functional similarity between the states of wakefulness and REM sleep. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Mu rhythm; Rapid eye movement sleep; Electrocorticography
1. Introduction The rolandic mu rhythm is a well-known and fundamental feature of the waking human electroencephalogram (Gastaut, 1952). It has been characterized as a `wicket shaped', 7±11 Hz activity commonly recorded from the central areas of scalp during the relaxed, alert state (Gastaut, 1952). This rhythm can be attenuated by a number of stimuli, including tactile stimulation of the contralateral body, active or passive movement of the contralateral body or face, active or passive movement of the ipsilateral arm, and planning for active movement of the contralateral body (Arroyo et al., 1993). Enhancement of mu activity has also been documented, as shown in the cortical hand area during active foot or tongue movement and reading of words (Pfurtscheller and Neuper, 1994). In addition to the waking state, the mu rhythm has been described to occur during non-rapid eye movement (non* Corresponding author. University of Washington Regional Epilepsy Center, Box 359745, Harborview Medical Center, 325 9th Avenue, Seattle, WA 98104-2499, USA. Tel.: 11-206-731-6769; fax: 11-206-731-4409. E-mail address: [email protected] (J.W. Miller).
REM) and REM sleep, but de®nitive, functional identi®cation of this activity has not been not performed (Yamada and Kooi, 1975). In this study, we will demonstrate, through analysis of scalp and invasive recordings, that mu-like activity is a prominent feature of REM sleep and that this activity exhibits features identical to the waking mu rhythm. Moreover, in contrast to previous ®ndings, we will show that the mu rhythm is not a characteristic of non-REM sleep. 2. Methods and materials Eye and chin leads were added to recording montages of adult (age range 24±53) patients undergoing prolonged closed circuit television (video) electroencephalographic recording (CCTV/EEG) with scalp (12 patients) or subdural electrodes (5 patients). All patients undergoing dural recording had localization-related epilepsy, with foci distant from somatosensory cortex; similarly 9 of the 12 patients undergoing scalp recordings had electrographic evidence of epilepsy on the study, without evidence of seizure origination in the centroparietal region. An additional 10 patients undergoing routine 1 h daytime scalp EEG studies were
1388-2457/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88-2457(00)0055 9-9
CLINPH 2000140
S.P. Duntley et al. / Clinical Neurophysiology 112 (2001) 528±531
similarly recorded. All recordings were performed with digital recording systems, allowing post-acquisition remontaging and ®ltering. Sleep staging was performed by reformatting into standard polysomnography montages (using two EEG channels, along with eye and chin channels) and applying standard criteria (Rechtschaffen and Kales, 1968). The recordings were then reviewed again by displaying all EEG channels to assess the presence of rhythmic EEG activity. Invasive recordings in patients being evaluated for neurosurgical treatment of epilepsy were performed with 64 contact grids placed over the convexity of the frontal, parietal and lateral temporal cortices, with additional 4±6 contact subdural strips over the subtemporal, medial frontoparietal, and orbitofrontal cortices (Fig. 3). Several (3±4) 15 min segments of consolidated non-REM, REM, and waking time were selected for careful analysis using standard bipolar and referential EEG montages by a board certi®ed EEG reader blinded to the staging. Any epochs showing evidence of arousal from sleep were excluded. 3. Results The review of waking periods on scalp recordings revealed that 7 out of 12 patients displayed identi®able
529
mu-like activity. During REM periods, six of the seven patients had intermittent 7±10 Hz central rhythms (Fig. 1) with the same morphology as the waking mu rhythm. This activity was seen in patients with and without documented epilepsy. The mean frequency of the waking mu was 8.9 Hz with a mean amplitude of 39 mV, and the mean frequency of the rhythm during REM sleep was 8.2 Hz with a mean amplitude of 29 mV. The distribution of these two rhythms was identical, being maximal in the C3 and C4 electrodes, with a much lower amplitude in the P3 and P4 electrodes. A similar rhythm was not observed during non-REM sleep in this group of patients. Additionally, analysis of a separate set of ten patients undergoing routine daytime scalp EEG monitoring also demonstrated no occurrences of mu activity in non-REM sleep. During invasive recordings, all patients had a 7±8 Hz activity recorded over regions of the lateral and medial parietal cortex during waking. The electrodes where this activity appeared corresponded to somatosensory cortex as documented by median nerve somatosensory evoked response testing and also by the effects of electrical stimulation of the cortex (Fig. 3). This activity was blocked by contralateral body movement or contralateral somatosensory stimuli (Fig. 2A). This activity therefore corresponds to the rolandic mu rhythm seen on scalp recordings,
Fig. 1. Scalp EEG recording of a 9 hz central rhythm recorded during REM sleep. This rhythm had the same morphology and distribution as this subject's waking mu rhythm. This recording was done prior to placement of subdural grid (see Fig. 2).
530
S.P. Duntley et al. / Clinical Neurophysiology 112 (2001) 528±531
con®rming past reports (Arroyo et al., 1993) that the mu rhythm originates primarily from somatosensory cortex. During REM sleep, rhythmic activity of identical frequency
and distribution was seen in all of the patients. In one patient, the reactivity of this rhythm to sensory stimuli during REM sleep was tested. Squeezing the contralateral
Fig. 2. (A) Waking mu activity in the same patient as Fig. 1 is seen most prominently in leads G6 and G7 of a subdural recording grid. At the arrow, the contralateral hand was squeezed, attenuating the rhythm. The record is referenced to a scalp electrode (RF). (B) Recording from subdural grid of central activity similar in morphology and distribution to the waking mu during REM sleep in the same patient. At the arrows, the contralateral hand is squeezed, resulting in attenuation of the rhythmic activity.
S.P. Duntley et al. / Clinical Neurophysiology 112 (2001) 528±531
531
mu rhythm in REM sleep through both EEG and functional criteria. Furthermore, we could ®nd no evidence of the rhythm in non-REM sleep in multiple patients. The ®nding of the mu rhythm during REM sleep is significant because it demonstrates the functional similarity between the states of wakefulness and REM sleep. REM sleep has been described as `not really sleep at all, but a state in which the subject is awake, but paralyzed and hallucinating' (Dement, 1976). The EEG during REM sleep has been poorly characterized and is commonly described as low amplitude and featureless except for occasional 2-6 Hz, vertex sawtooth waves. While there are clear differences between the waking and REM sleep EEG, such as increased generalized theta activity in the REM EEG, there are also similarities, such as the presence of alpha frequency activity over the posterior head regions in both states (Carskadon, 1982). Animals with experimental pontine lesions eliminating the paralysis associated with REM sleep (Henley and Morrison, 1974), and patients with the REM sleep behavior disorder both `act out' their dreams and are able to move about and react to environmental stimuli, demonstrating that sensorimotor processing may be active during REM sleep dreams. It is therefore not surprising that the rolandic mu rhythm, which is a phenomenon tied to sensorimotor processing, would be present during REM sleep. Fig. 3. Diagram demonstrating electrode placement from subdural electrodes in the same patient as Fig. 1. The distribution of waking and stage REM mu activity is indicated by the shaded circles. The crosses indicate the location of motor and somatosensory responses with extraoperative electrical stimulation of the electrodes.
hand consistently attenuated the rhythm without evidence of arousal while squeezing the ipsilateral hand had no effect (Fig. 2B). This REM sleep rhythm therefore recapitulates all of the characteristics of the waking rolandic mu rhythm. As with the scalp recordings, no evidence of the mu rhythm was seen in other sleep stages. 4. Discussion Although the mu rhythm has been well-studied in the waking state, little is known about the characteristics of this activity in the REM state. Three early EEG studies mentioned the presence of fast theta activity during human REM sleep but did not recognize its signi®cance (Dement and Kleitman, 1957; Johnson et al., 1969; Putkonen and Elomaa, 1972). For example, Putkonen and Elomaa (1972) described a 7 Hz fronto-centro-parietal activity, which they named theta spindles. These reports did not link this activity to the mu rhythm nor determine if it was affected by somatosensory stimulation. Another study observed an activity consistent with the mu rhythm in multiple sleep stages (stages I, II, and REM) but also did not perform functional, mu-de®ning tests (Yamada and Kooi, 1975). In this study, we have identi®ed the presence of the
References Arroyo S, et al. Functional signi®cance of the mu rhythm of human cortex: an electrophysiologic study with subdural electrodes. Electroenceph clin Neurophysiol 1993;87:76±87. Carskadon M. Basics for polygraphic monitoring of sleep. In: Guilleminault C, editor. Sleeping and waking disorders: Indications and techniques, Menlo Park, California: Addison-Wesley Publishing Company, 1982. p. 15. Dement W. Some must watch while some must sleep, San Francisco: San Francisco Book Company, Inc, 1976. p. 26. Dement W, Kleitman N. Cyclic variations in EEG during sleep and their relation to eye movements, body motility, and dreaming. Electroenceph clin Neurophysiol 1957;9:673±690. Gastaut H. Etude electrocorticographique de la reactivite des rythmes rolandiques. Rev Neurol 1952;87:176±182. Henley K, Morrison AR. A reevaluation of the effects of the lesions of the pontine tegmentum and locus coeruleus on the phenomena of paradoxical sleep in the cat. Acta Neurobiol Exp 1974;34:215±232. Johnson L, Lubin A, Naitoh P, Nute C, Austin M. Spectral analysis of the EEG of dominant and non-dominant alpha subjects during waking and sleeping. Electroenceph clin Neurophysiol 1969;26:361±370. Pfurtscheller G, Neuper C. Event-related synchronization of mu rhythm in the EEG over the cortical hand area in man. Neurosci Lett 1994;174:93±96. Putkonen PTS, Elomaa E. Theta spindles in human REM sleep. Revue EEG Neurophysiol Clin 1972;5:223±226. Rechtschaffen A, Kales A. A manual of standardized terminology: techniques and scoring system for sleep stages of human subjects, Los Angeles: UCLA Brain Informaiton Service/Brain Research Institute, 1968. Yamada T, Kooi KA. Level of Consciousness and the Mu Rhythm. Clin Electroenceph 1975;6:80±88.
www.elsevier.com/locate/clinph
Brief communication
Characterization of the mu rhythm during rapid eye movement sleep Stephen P. Duntley a, Albert H. Kim a, Daniel L. Silbergeld b, John W. Miller b,c,* a
Department of Neurology and Neurological Surgery, Washington University School of Medicine, St. Louis, MO, USA b Department of Neurosurgery, University of Washington, Seattle, WA, USA c Department of Neurology, University of Washington, Seattle, WA, USA Accepted 15 December 2000
Abstract Objective: The rolandic mu rhythm, a resting activity of somatosensory cortex, is a striking feature of the waking human electroencephalogram. This study will demonstrate that activity with identical features occurs during rapid eye movement (REM) sleep. Methods: Eye and chin leads were added during prolonged closed circuit television (video) electroencephalographic (EEG) recording with scalp (12 patients) or subdural electrodes including 64 contract grids over the frontoparietal cortices (5 patients). Sleep staging was performed by reformatting into standard polysomnography montages (using two EEG channels, and eye and chin channels) and applying standard scoring criteria. The recordings were then reviewed using all EEG channels to assess rhythmic EEG activity by a reader blinded to the sleep staging. Results: During scalp recordings, 7±10 Hz central rhythms were seen during wakefulness in 7 patients, with 6 of these also having similar rhythms during REM sleep. Similar activity was seen over somatosensory cortex during wakefulness and REM in all invasively recorded patients. This activity was blocked by contralateral body movement or contralateral somatosensory stimuli, even during REM sleep. It was absent in other sleep stages. Conclusions: This REM sleep activity recapitulates all the characteristics of the waking rolandic mu rhythm. This demonstrates functional similarity between the states of wakefulness and REM sleep. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Mu rhythm; Rapid eye movement sleep; Electrocorticography
1. Introduction The rolandic mu rhythm is a well-known and fundamental feature of the waking human electroencephalogram (Gastaut, 1952). It has been characterized as a `wicket shaped', 7±11 Hz activity commonly recorded from the central areas of scalp during the relaxed, alert state (Gastaut, 1952). This rhythm can be attenuated by a number of stimuli, including tactile stimulation of the contralateral body, active or passive movement of the contralateral body or face, active or passive movement of the ipsilateral arm, and planning for active movement of the contralateral body (Arroyo et al., 1993). Enhancement of mu activity has also been documented, as shown in the cortical hand area during active foot or tongue movement and reading of words (Pfurtscheller and Neuper, 1994). In addition to the waking state, the mu rhythm has been described to occur during non-rapid eye movement (non* Corresponding author. University of Washington Regional Epilepsy Center, Box 359745, Harborview Medical Center, 325 9th Avenue, Seattle, WA 98104-2499, USA. Tel.: 11-206-731-6769; fax: 11-206-731-4409. E-mail address: [email protected] (J.W. Miller).
REM) and REM sleep, but de®nitive, functional identi®cation of this activity has not been not performed (Yamada and Kooi, 1975). In this study, we will demonstrate, through analysis of scalp and invasive recordings, that mu-like activity is a prominent feature of REM sleep and that this activity exhibits features identical to the waking mu rhythm. Moreover, in contrast to previous ®ndings, we will show that the mu rhythm is not a characteristic of non-REM sleep. 2. Methods and materials Eye and chin leads were added to recording montages of adult (age range 24±53) patients undergoing prolonged closed circuit television (video) electroencephalographic recording (CCTV/EEG) with scalp (12 patients) or subdural electrodes (5 patients). All patients undergoing dural recording had localization-related epilepsy, with foci distant from somatosensory cortex; similarly 9 of the 12 patients undergoing scalp recordings had electrographic evidence of epilepsy on the study, without evidence of seizure origination in the centroparietal region. An additional 10 patients undergoing routine 1 h daytime scalp EEG studies were
1388-2457/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S13 88-2457(00)0055 9-9
CLINPH 2000140
S.P. Duntley et al. / Clinical Neurophysiology 112 (2001) 528±531
similarly recorded. All recordings were performed with digital recording systems, allowing post-acquisition remontaging and ®ltering. Sleep staging was performed by reformatting into standard polysomnography montages (using two EEG channels, along with eye and chin channels) and applying standard criteria (Rechtschaffen and Kales, 1968). The recordings were then reviewed again by displaying all EEG channels to assess the presence of rhythmic EEG activity. Invasive recordings in patients being evaluated for neurosurgical treatment of epilepsy were performed with 64 contact grids placed over the convexity of the frontal, parietal and lateral temporal cortices, with additional 4±6 contact subdural strips over the subtemporal, medial frontoparietal, and orbitofrontal cortices (Fig. 3). Several (3±4) 15 min segments of consolidated non-REM, REM, and waking time were selected for careful analysis using standard bipolar and referential EEG montages by a board certi®ed EEG reader blinded to the staging. Any epochs showing evidence of arousal from sleep were excluded. 3. Results The review of waking periods on scalp recordings revealed that 7 out of 12 patients displayed identi®able
529
mu-like activity. During REM periods, six of the seven patients had intermittent 7±10 Hz central rhythms (Fig. 1) with the same morphology as the waking mu rhythm. This activity was seen in patients with and without documented epilepsy. The mean frequency of the waking mu was 8.9 Hz with a mean amplitude of 39 mV, and the mean frequency of the rhythm during REM sleep was 8.2 Hz with a mean amplitude of 29 mV. The distribution of these two rhythms was identical, being maximal in the C3 and C4 electrodes, with a much lower amplitude in the P3 and P4 electrodes. A similar rhythm was not observed during non-REM sleep in this group of patients. Additionally, analysis of a separate set of ten patients undergoing routine daytime scalp EEG monitoring also demonstrated no occurrences of mu activity in non-REM sleep. During invasive recordings, all patients had a 7±8 Hz activity recorded over regions of the lateral and medial parietal cortex during waking. The electrodes where this activity appeared corresponded to somatosensory cortex as documented by median nerve somatosensory evoked response testing and also by the effects of electrical stimulation of the cortex (Fig. 3). This activity was blocked by contralateral body movement or contralateral somatosensory stimuli (Fig. 2A). This activity therefore corresponds to the rolandic mu rhythm seen on scalp recordings,
Fig. 1. Scalp EEG recording of a 9 hz central rhythm recorded during REM sleep. This rhythm had the same morphology and distribution as this subject's waking mu rhythm. This recording was done prior to placement of subdural grid (see Fig. 2).
530
S.P. Duntley et al. / Clinical Neurophysiology 112 (2001) 528±531
con®rming past reports (Arroyo et al., 1993) that the mu rhythm originates primarily from somatosensory cortex. During REM sleep, rhythmic activity of identical frequency
and distribution was seen in all of the patients. In one patient, the reactivity of this rhythm to sensory stimuli during REM sleep was tested. Squeezing the contralateral
Fig. 2. (A) Waking mu activity in the same patient as Fig. 1 is seen most prominently in leads G6 and G7 of a subdural recording grid. At the arrow, the contralateral hand was squeezed, attenuating the rhythm. The record is referenced to a scalp electrode (RF). (B) Recording from subdural grid of central activity similar in morphology and distribution to the waking mu during REM sleep in the same patient. At the arrows, the contralateral hand is squeezed, resulting in attenuation of the rhythmic activity.
S.P. Duntley et al. / Clinical Neurophysiology 112 (2001) 528±531
531
mu rhythm in REM sleep through both EEG and functional criteria. Furthermore, we could ®nd no evidence of the rhythm in non-REM sleep in multiple patients. The ®nding of the mu rhythm during REM sleep is significant because it demonstrates the functional similarity between the states of wakefulness and REM sleep. REM sleep has been described as `not really sleep at all, but a state in which the subject is awake, but paralyzed and hallucinating' (Dement, 1976). The EEG during REM sleep has been poorly characterized and is commonly described as low amplitude and featureless except for occasional 2-6 Hz, vertex sawtooth waves. While there are clear differences between the waking and REM sleep EEG, such as increased generalized theta activity in the REM EEG, there are also similarities, such as the presence of alpha frequency activity over the posterior head regions in both states (Carskadon, 1982). Animals with experimental pontine lesions eliminating the paralysis associated with REM sleep (Henley and Morrison, 1974), and patients with the REM sleep behavior disorder both `act out' their dreams and are able to move about and react to environmental stimuli, demonstrating that sensorimotor processing may be active during REM sleep dreams. It is therefore not surprising that the rolandic mu rhythm, which is a phenomenon tied to sensorimotor processing, would be present during REM sleep. Fig. 3. Diagram demonstrating electrode placement from subdural electrodes in the same patient as Fig. 1. The distribution of waking and stage REM mu activity is indicated by the shaded circles. The crosses indicate the location of motor and somatosensory responses with extraoperative electrical stimulation of the electrodes.
hand consistently attenuated the rhythm without evidence of arousal while squeezing the ipsilateral hand had no effect (Fig. 2B). This REM sleep rhythm therefore recapitulates all of the characteristics of the waking rolandic mu rhythm. As with the scalp recordings, no evidence of the mu rhythm was seen in other sleep stages. 4. Discussion Although the mu rhythm has been well-studied in the waking state, little is known about the characteristics of this activity in the REM state. Three early EEG studies mentioned the presence of fast theta activity during human REM sleep but did not recognize its signi®cance (Dement and Kleitman, 1957; Johnson et al., 1969; Putkonen and Elomaa, 1972). For example, Putkonen and Elomaa (1972) described a 7 Hz fronto-centro-parietal activity, which they named theta spindles. These reports did not link this activity to the mu rhythm nor determine if it was affected by somatosensory stimulation. Another study observed an activity consistent with the mu rhythm in multiple sleep stages (stages I, II, and REM) but also did not perform functional, mu-de®ning tests (Yamada and Kooi, 1975). In this study, we have identi®ed the presence of the
References Arroyo S, et al. Functional signi®cance of the mu rhythm of human cortex: an electrophysiologic study with subdural electrodes. Electroenceph clin Neurophysiol 1993;87:76±87. Carskadon M. Basics for polygraphic monitoring of sleep. In: Guilleminault C, editor. Sleeping and waking disorders: Indications and techniques, Menlo Park, California: Addison-Wesley Publishing Company, 1982. p. 15. Dement W. Some must watch while some must sleep, San Francisco: San Francisco Book Company, Inc, 1976. p. 26. Dement W, Kleitman N. Cyclic variations in EEG during sleep and their relation to eye movements, body motility, and dreaming. Electroenceph clin Neurophysiol 1957;9:673±690. Gastaut H. Etude electrocorticographique de la reactivite des rythmes rolandiques. Rev Neurol 1952;87:176±182. Henley K, Morrison AR. A reevaluation of the effects of the lesions of the pontine tegmentum and locus coeruleus on the phenomena of paradoxical sleep in the cat. Acta Neurobiol Exp 1974;34:215±232. Johnson L, Lubin A, Naitoh P, Nute C, Austin M. Spectral analysis of the EEG of dominant and non-dominant alpha subjects during waking and sleeping. Electroenceph clin Neurophysiol 1969;26:361±370. Pfurtscheller G, Neuper C. Event-related synchronization of mu rhythm in the EEG over the cortical hand area in man. Neurosci Lett 1994;174:93±96. Putkonen PTS, Elomaa E. Theta spindles in human REM sleep. Revue EEG Neurophysiol Clin 1972;5:223±226. Rechtschaffen A, Kales A. A manual of standardized terminology: techniques and scoring system for sleep stages of human subjects, Los Angeles: UCLA Brain Informaiton Service/Brain Research Institute, 1968. Yamada T, Kooi KA. Level of Consciousness and the Mu Rhythm. Clin Electroenceph 1975;6:80±88.