The Scoring of Movements in Sleep
Arthur S. Walters, M.D. ; Gilles Lavigne, D.M.D., Ph.D.2; Wayne Hening, M.D., Ph.D.3; Daniel L. Picchietti, M.D.4; Richard P. Allen, Ph.D.5; Sudhansu Chokroverty,
M.D.1; Clete A. Kushida, M.D., Ph.D.6; Donald L. Bliwise, Ph.D.7; Mark W. Mahowald, M.D.8; Carlos H. Schenck, M.D.8; Sonia Ancoli-Israel, Ph.D.9
Seton Hall University School of Graduate Medical Education, New Jersey Neuroscience Institute at JFK Medical Center, Edison, NJ; 2University of
Montreal Sleep Disorder Laboratory, Sacre Coeur Hospital, Montreal, QC, Canada; 3UMDNJ-Robert Wood Johnson Medical School, New Brunswick,
NJ; 4Carle Clinic, University of Illinois, Urbana, IL; 5Johns Hopkins University, Baltimore, MD; 6Stanford University, Stanford, CA; 7Emory University
Medical School, Atlanta, GA; 8Minnesota Regional Sleep Disorders Center, Hennepin County Medical Center, University of Minnesota Medical School,
Minneapolis, MN; 9University of California, San Diego, San Diego, CA
Abstract: The International Classiﬁcation of Sleep Disorders (ICSD-2) in sleep, REM sleep behavior disorder, and sleep bruxism. Published
has separated sleep-related movement disorders into simple, repetitive scoring criteria for rhythmic movement disorder, excessive fragmentary
movement disorders (such as periodic limb movements in sleep [PLMS], myoclonus, and hypnagogic foot tremor/alternating leg muscle activa-
sleep bruxism, and rhythmic movement disorder) and parasomnias tion were empirical and based on descriptive studies. The literature re-
(such as REM sleep behavior disorder and disorders of partial arousal, view disclosed no published evidence deﬁning clinical consequences of
e.g., sleep walking, confusional arousals, night terrors). Many of the pa- excessive fragmentary myoclonus or hypnagogic foot tremor/alternating
rasomnias are characterized by complex behaviors in sleep that appear leg muscle activation. Because of limited or absent evidence for reli-
purposeful, goal directed and voluntary but are outside the conscious ability and/or validity, a standardized RAND/UCLA consensus process
awareness of the individual and therefore inappropriate. All of the sleep- was employed for recommendation of speciﬁc rules for the scoring of
related movement disorders described here have speciﬁc polysomno- sleep-associated movements.
graphic ﬁndings. For the purposes of developing and/or revising speciﬁ- Keywords: Periodic Limb Movements in Sleep, REM Sleep Behavior
cations and polysomnographic scoring rules, the AASM Scoring Manual Disorder, Sleep Bruxism, Sleep related Rhythmic Movement Disorder,
Task Force on Movements in Sleep reviewed background literature and Excessive Fragmentary Myoclonus, Hypnagogic Foot Tremor, Alternat-
executed evidence grading of 81 relevant articles obtained by a litera- ing Leg Muscle Activation.
ture search of published articles between 1966 and 2004. Subsequent Citation: Walters AS; Lavigne G; Hening W et al. Movements in sleep:
evidence grading identiﬁed limited evidence for reliability and/or valid- A review by the movement task force of the american academy of sleep
ity for polysomnographic scoring criteria for periodic limb movements medicine. J Clin Sleep Med 2007:3(2);155-167
Disclosure Statement Dr. Ancoli-Israel has received research support from Takada; is a consul-
This is was not an industry supported study. Dr. Walters has received re- tant, on the scientiﬁc advisory board, and/or speakers bureau for Ferring
search support and consulting fees from GlaxoSmithKline, Boehringer- Pharmaceuticals, King Pharmaceuticals, Merck, Neurocrine Biosciences,
Ingelheim, Kyowa, Schwarz Pharmaceuticals and Xenoport. Dr. Lavigne Neurogen, Sanoﬁ-Aventis, Sepracor, Somaxon, and Takeda. Drs. Picchietti
has received research or travel support from Respironics. Dr. Hening is a and Schenck have indicated no ﬁnancial conﬂicts of interest.
member of the advisory board for Boehringer Ingelheim, Schwarz-Pharma,
and Xenoport; is on the speakers bureau for Boehringer Ingelheim; and has Submitted for publication February 1, 2007
received research support from Boehringer Ingelheim. Dr. Allen has ﬁnan- Accepted for publication March 15, 2007
cial interest in IM Systems, Inc. and has participated in speaking engage- Address correspondence to: Arthur S. Walters, M.D., New Jersey Neurosci-
ments and/or is on the speakers bureau for GlaxoSmithKline, Boehringer ence Institute, JFK Medical Center, 65 James Street, Edison, NJ 08818, Tel:
Ingelheim, Schwarz Pharma, and Xenoport. Dr. Chokroverty is a member (732) 321-7000 x 68177; Fax: (732) 632-1584; E-mail: email@example.com
of the advisory board of Sanoﬁ-Aventis and has participated in speaking
engagements for Boehringer-Ingelheim. Dr. Kushida is an investigator on 1.0 HISTORICAL PERSPECTIVE
research managed by Stanford University including contract research from
GlaxoSmithKline, Boehringer-Ingelheim, XenoPort, Inc., Schwarz Phar-
maceuticals, and Kyowa Pharmaceuticals and has received travel grants W hen the Rechtschaffen and Kales1 sleep scoring manual
was ﬁrst published in 1968, the rules were limited to the
scoring of normal sleep. Since then, many abnormalities of sleep
from XenoPort, Inc. for oral presentations. Dr. Bliwise has received research
support from GlaxoSmithKline, Sector Medical, and Takeda; is a consul- have been identiﬁed, including involuntary movements during
tant/advisory board member for GlaxoSmithKline, Gerson Lehman Group, sleep. The International Restless Legs Syndrome Study Group
Neurocrine, Pﬁzer, Sanoﬁ Synthelabo, Takeda, Sepracor, Sleep Medicine (IRLSSG),2-4 ﬁrst characterized involuntary movement abnor-
Education Institute, and Cephalon; and has participated in speaking engage- malities such as restless legs syndrome (RLS) and periodic limb
ments for King, School of Sleep Medicine, Sepracor, and Takeda. Dr. Ma- movements in sleep (PLMS). For the ﬁrst time, rules for scoring
howald has received research support from Advanced Medical Electronics, PLMS evolved from a mathematical analysis of the leg move-
Kyowa Pharmaceuticals, Merck, Schwarz Pharmaceuticals, and Xenoport. ments themselves.5 Additional movement disorders have also
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 155
AS Walters, G Lavigne, W Hening et al
been identiﬁed, including REM sleep behavior disorder (RBD),6
excessive fragmentary myoclonus,7 hypnagogic foot tremor,8,9 Table 1—Summary of evidence grading
and alternating leg muscle activation (ALMA).10,11
The new International Classiﬁcation of Sleep Disorders (ICSD- Evidence StudyDesign
2)12 has separated sleep-related movement disorders into simple Levels
repetitive movement disorders (such as PLMS, leg cramps,13 1 Randomized well-designed trials with low-
bruxism,14 and rhythmic movement disorder15) and parasomnias alpha & low-beta errors
(such as REM sleep behavior disorder and disorders of partial 2 Randomized trials with high-beta errors
arousal (sleepwalking, confusional arousals, and sleep terrors16). 3 Nonrandomized controlled or concurrent cohort
Many of the parasomnias are characterized by complex behaviors studies
in sleep that appear purposeful, goal directed, and voluntary but 4 Nonrandomized historical cohort studies
are outside the conscious awareness of the individual and there- 5 Case series
Adapted from Sackett18
Most of the sleep-related movement disorders have speciﬁc
ﬁndings on polysomnography that warrant their inclusion in a
manual of sleep scoring. These are discussed in the review, but bruxism OR tooth grinding OR bruxomania, AND sleep, and
other disorders such as leg cramps, RLS, and disorders of partial bruxism OR tooth grinding AND sleep AND polysomnography
arousal where ﬁndings on polysomnography are less speciﬁc are OR scoring OR diagnosis; for rhythmic movement disorder, the
therefore not addressed. key words were rhythmic movement disorder OR body rocking
OR head banging OR jactatio, AND sleep; For fragmentary my-
2.0 METHODS oclonus, the keywords were fragmentary myoclonus OR partial
myoclonus OR twitching, AND sleep; for hypnagogic foot tremor/
The Movement Task Force (see page 165) was appointed by ALMA the keywords were rhythmic foot movements OR hypna-
the American Academy of Sleep Medicine (AASM) in 2004 for gogic foot tremor OR alternating muscle activity OR ALMA, AND
the purpose of developing standard scoring rules for movement sleep. Over 203 articles were identiﬁed, of which 81 were relevant
disorders occurring during sleep. The task force consisted initially to this review; 44 are included in evidence tables. Because of the
of 9 members with expertise in sleep movements who participat- rigorous but lengthy procedure employed for evidence grading,
ed in evidence review, and one liaison from the Scoring Manual it was not possible to do an updated literature search of all sub-
Steering Committee. An additional 2 members were recruited sequent articles and subject them to the RAND voting process.
to complete consensus balloting, with a total of eleven partici- However, select key articles from 2005 and 2006 were added to
pants in the task force and consensus activities. When evidence this review and played a role in the ﬁnal consensus statement.
review provided only limited evidence for rule development, a
RAND/UCLA Appropriateness Method consensus process17 was 2.3 Evidence selection and grading
employed. For the purposes of this review, the Movements Task
Force: 1) performed a literature search to identify studies for For general discussion and historical information, each task
evidence grading and 2) developed a referenced paper including force member was assigned a clinical topic and was free to re-
background and evidence review that would serve as an annotated view any of the papers for inclusion in their area of review as
supporting text for the purposes of guiding consensus and rule long as the paper included empirical data relevant to the section.
development. When possible, the paper focused on evidence from Exclusion criteria included abstracts, reviews, theoretical papers,
the literature for reliability and validity of speciﬁcations and mea- editorials, and case studies.
sures. Following consensus, recommendations were forwarded to Evidence review required formalized data extraction and ev-
the AASM Scoring Manual Steering Committee for ﬁnal drafting idence grading. A data extraction sheet was developed prior to
of scoring rules which are published in a separate volume.18 review all of the articles chosen for evidence review. Each move-
ment paper was reviewed and summarized on the extraction sheet
2.1 Timeline by 2 task force members for the following information: study
design, number and sex of subjects, types of movement mea-
The task force met via conference call on 15 occasions between sures, study outcomes and outcome measures, signiﬁcant results,
January 2005 and February 2006 and met face-to-face once at the evidence grade, and miscellaneous notes. Evidence grading was
Annual Meeting of the Associated Professional Sleep Societies in modiﬁed from Sackett19 and is presented in Table 1. An evidence
June 2005. Discussions totaled approximately 18 hours. table for each movement disorder (which can be accessed on the
web at www.aasmnet.org), summarizing all the extraction sheets,
2.2 Search terms and articles can be found at the end of each section of the paper.
A computer-based PubMed literature search was performed for 2.4 RAND/UCLA Appropriateness Method Consensus Balloting
all human studies published in English between 1966 and 2004
using the following key words: For periodic limb movements in The task force developed questions following a modiﬁed
sleep the key words were periodic limb movement AND legs, noc- RAND/UCLA appropriateness method,17 resulting in 6 questions
turnal myoclonus, anterior tibialis AND leg, and anterior tibialis covering 6 major topics: 1) periodic limb movements in sleep,
AND sleep; for REM behavior disorder, the key words were REM 2) REM sleep behavior disorder, 3) sleep bruxism, 4) rhythmic
sleep behavior disorder AND polysomnography AND diagnosis movement disorder, 5) excessive fragmentary myoclonus, and 6)
AND scoring AND reliability; for bruxism, the keywords were hypnagogic foot tremor/alternating leg muscle activation. As 9
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 156
Movements in Sleep
Table 3a—ICSD-212 Criteria for the Diagnosis of PLMD Table 3b—Exceptions to ICSD-2 PLMS Rules – Adapted from the
International RLS Study group4,5
A. Polysomnography demonstrates repetitive highly stereotyped
limb movements that are The following rules define a significant leg movement (LM) event:
i. 0.5 to five seconds in duration. (1) The maximum duration of a LM event is 10 seconds.
ii. Of amplitude greater than or equal to 25% of toe dorsiflexion (2) The minimum size of a LM event is an 8 µV increase in EMG
during calibration. voltage above resting EMG.
iii. In a sequence of four or more movements. (3) The onset of a LM event is defined as an 8 µV increase in EMG
iv. Separated by an interval of more than 5 seconds (from limb- voltage above resting EMG.
movement onset to limb-movement onset) and less than 90 (4) The offset criteria of a LM event is defined as the start of a
seconds (typically there is an interval of 20 to 40 seconds). period of at least 0.5 seconds during which the EMG does not
B. The PLMS index exceeds 5 per hour in children and 15 per hour exceed 2 µV above resting EMG.
in most adult cases. (5) An arousal by ASDA criteria and a LM are assumed to be
C. There is clinical sleep disturbance or a complaint of daytime associated with each other if they overlap or if the end of one
fatigue. event and the beginning of the other event are within 0.5 seconds
D. The PLMS are not better explained by another current sleep or less of each other, regardless of which event is first.
disorder, medical or neurological disorder, mental disorder, (6) An apnea/hypopnea and a LM are assumed to be associated with
medication use, or substance use disorder (e.g., PLMS at the each other if they overlap or if the end of one event and the
termination of cyclically occurring apneas should not be counted beginning of the other event are within 0.5 seconds or less of
as true PLMS or PLMD). each other, regardless of which event is first.
voting members were needed, 2 additional experts in movement Previous versions of the ICSD had recommended a PLMS index
sleep disorders were added to the task force to ensure a full vote at (number of leg kicks per hour of sleep) >5 to determine clinical
each meeting. The task force voted by secret ballot on each ques- signiﬁcance.22 The new ICSD recommended a cut-off of PLMS
tion and then met via telephone conference calls to discuss those index >15. The ICSD pointed out that even a high PLMS index
areas where there was not sufﬁcient agreement, which resulted in with associated arousals had little relationship to symptomatic
modiﬁed questions. A second round vote was then taken to deter- sleep disturbances (e.g., insomnia or hypersomnia) and therefore
mine the ﬁnal rule. recommended that a relationship must be established between the
PLMS and the insomnia and hypersomnia, with no other disorder
3.0 PERIODIC LIMB MOVEMENTS IN SLEEP accounting for the PLMS, for the term periodic limb movement
3.1 Background disorder (PLMD) to be employed. Otherwise, the presence of
PLMS should simply be noted. Evidence from the literature sug-
In 1993, the Atlas Task Force of the American Sleep Disorders gests that true PLMD in adults is rare.12 For children, the recom-
Association (ASDA) developed scoring rules for periodic limb mendation was for a PLMS index of >5 plus symptomatic sleep
movements in sleep (PLMS).20 To be counted as PLMS, there disruption. ICSD-2 criteria for PLMD can be found in Table 3a.
had to be 4 consecutive movements, each lasting 0.5-5 seconds, More recent modiﬁcations to these rules developed by the Pe-
with movement onsets each 5-90 seconds apart. Although previ- riodic Limb Movement task force of the International Restless
ous deﬁnitions suggested by Coleman,21 among others, had sug- Legs Syndrome Study Group (IRLSSG) are listed in table 3b.
gested an upper limit of 120 seconds rather than 90 seconds for Unless otherwise stated in table 3b, rules for PLMS suggested by
the inter-movement interval, the ASDA provided no rationale for the IRLSSG are the same as those in table 3a.
this change in criteria. The ASDA committee also added an am- Periodic arm movements have also been documented in wake-
plitude criterion of at least 25% of the EMG bio-calibration signal fulness and sleep in patients with restless legs syndrome (RLS).23
required for the minimum EMG of the movement event but did In some cases, periodic arm movements had an independent
not provide criteria for starting or ending a movement event. No frequency and were not synchronized with periodic leg move-
suggestions were made for a minimum interval for the measure- ments,23,24 but they were observed only in patients with signiﬁcant
ment from the offset of one movement to the onset of another. periodic leg movements. In such cases, the periodic leg move-
The ASDA committee recommended that leg movements on 2 ments represent the primary disorder.
separate legs separated by less than 5 seconds be counted as a
single leg movement. Periodic leg movements in association with 3.2 Evidence for reliability of rules, specifications, or measures.
episodes of sleep disordered breathing would not be considered
true PLMS. Arousals following a PLMS movement would only The instructions for measuring PLMS are qualitative, in that
be considered as movement related if they followed the leg move- patients exert a varying amount of force for dorsiflexion of the
ment by ≤3 seconds. The Atlas Task Force of the ASDA did not great toe during biocalibrations prior to polysomnography. Some
quote substantive literature to formulate their rules for scoring but investigators have therefore questioned the importance of using
relied primarily on precedents established by Coleman,21 and on amplitude criteria for the measurement of PLMS during wakeful-
expert opinion. Evidence levels for this and all other PLMS ar- ness and during sleep.25 In a study of 24 patients, Gschliesser et
ticles can be found in Table 2 (which can be accessed on the web al found significantly more PLMS of all types when amplitude
at www.aasmnet.org). criteria were eliminated.25 To overcome this problem, the PLM
In 2005, the new International Classiﬁcation of Sleep Disor- Task Force of the International Restless Legs Syndrome Study
ders (ICSD-2) adopted the ASDA deﬁnition with 2 changes.12 Group (IRLSSG) recommended that the baseline EMG values be
obtained from the resting tibialis anterior EMG and not from the
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 157
AS Walters, G Lavigne, W Hening et al
EMG signal obtained from dorsiflexion of the great toe during Five Level 3 studies27-31 and one Level 2 study32 examined night-
biocalibration.4 This group also recommended that PLM onset be to-night variability in PLMS, all ﬁnding signiﬁcant night-to-night
recorded as soon as the EMG signal increased to 8 μV or more variation in PLMS indices within individual subjects, but not in
above baseline and that movement offset be recorded when the group data, and all recommended multiple nights of recording to
EMG signal fell to <2 μV above the baseline for at least 0.5 sec- decrease variability for within-subject evaluations. Speciﬁcally,
onds. The PLM Task Force of the IRLSSG also recommended Bliwise et al,27 based on a study of 45 patients, recommended that
that 1) the definition of PLMs in sleep be extended to 10 seconds, in clinical trials of therapeutic medications for PLMS, at least
making them comparable to PLMs in wakefulness; this change 2 sleep nights per therapeutic condition be recorded to decrease
was supported by data showing a small but significant number variability. Although the Bliwise study compared more than one
of PLMs in sleep lasting 5 to 10 seconds; 2) an arousal and a night of polysomnography, there was no blinded evaluation, and
PLM be considered associated with each other when there is <0.5 the study was therefore rated as Level 3 evidence. Mosko et al28
seconds between the end of one event and the onset of the other showed similar results in a study of 28 patients with PLMS, each
event, regardless of which is first; 3) a leg event not be included in recorded for 3 nights. When a single individual was examined,
the PLM count if it precedes or follows an apnea or hypopnea by the night-to-night variation was considerable, but as a group there
less than 0.5 seconds; 4) at least ≥5 seconds must elapse between was little variability. The average PLMS index for the group was
LMs to score a subsequent LM.4 While the IRLSSG criteria were 32.0 for night 1, 26.6 for night 2, and 31.6 for night 3. The cor-
primarily based upon expert opinion rather than on evidence, the relation coefﬁcient was 0.59 for night 1 vs. night 2 and 0.78 for
task force of the IRLSSG developing these criteria also completed night 2 vs. night 3. The authors therefore cautioned against using
a separate analysis of published PLM literature.5 The IRLSSG de- rigid cut-offs for the PLMS index for making decisions about the
veloped scoring standards using non-standardized consensus dur- clinical signiﬁcance of PLMS in an individual. The design of the
ing face-to-face discussions at a symposium which reviewed and Mosko study was similar to that of Bliwise et al and was there-
analyzed published data.4 The subsequent adoption of IRLSSG fore also rated as Level 3 evidence. Edinger et al32 compared 3
scoring rules for The AASM Manual for the Scoring of Sleep and consecutive nights of polysomnography in 15 individuals. Inter-
Associated Events: Rules, Terminology, and Technical Specifica- night comparisons of PLMS were 0.81 (night 1 vs. night 2), 0.86
tions employed the RAND Appropriateness Method as described (night 1 vs. night 3), and 0.74 (night 2 vs. night 3), all signiﬁcant
in section 2.4 of Methods. at p<0.001. Raters in this study were blinded to subjects’ identi-
The IRLSSG data review employed several methods of math- ties and night numbers. However, subjects were not chosen in a
ematical analyses. One database of about 65 RLS patients and 22 consecutive manner, so the study was rated as having Level 2
controls was available and included in the considerations along evidence. Both Hornyak et al and Sforza et al,30,31 using unblinded
with other data presented by the participants. The data reviewed scoring procedures reported considerable variability in the PLMS
included histogram analyses of event durations, inter-movement index between consecutive nights of recording in single patients,
intervals and various cut-off criteria for deﬁning events. Some but Sforza et al pointed out, as in previous studies, that there was
cases were presented documenting the occurrence of periodic little change across nights when the average for a group of pa-
events with durations >5 seconds, and some as long as 10 seconds tients was considered.
during sleep. A Markovian analysis indicated signiﬁcant periodic- One additional study by Culpepper et al29 compared night-to-
ity for inter-movement intervals of 5–90 seconds, supporting con- night variability, but raters were not blinded. This study examined
tinuing this criteria.5 The IRLSSG scoring rules also deﬁned for 46 patients with PLMS only, 10 with narcolepsy and PLMS, and
the ﬁrst time the onset and end of a movement event, permitting 16 with sleep apnea and PLMS. When night 1 was compared to
better consistency in scoring. All 18 IRLSSG participants were night 2, the pattern was more stable in older patients with idio-
authors on the ﬁnal manuscript that was subsequently published.4 pathic PLMS and insomnia (r = 0.98, p<0.01), as opposed to the
A method known as the suggested immobilization test (SIT) pattern in younger patients with narcolepsy and hypersomnia (r =
has been developed by Montplaisir et al to examine periodic limb 0.68, p= 0.06).
movements in wakefulness (PLMW) in restless legs syndrome In this same study, Culpepper et al29 examined time-of-night
(RLS) under conditions designed to provoke RLS symptoms.26 patterns, ﬁnding 2 types of PLMS patterns across the night. In
During the SIT, patients are asked to sit motionless on a bed at 45 older patients with idiopathic PLMS and insomnia, the number of
degrees, with legs outstretched and eyes open during wakefulness PLMS decreased across the night, while in younger patients with
for one hour while EMG is recorded from anterior tibialis muscle narcolepsy and hypersomnia, the number of PLMS remained rel-
of both legs, and a standard sleep EEG is also recorded to ensure atively constant across the night with a slight peak in the middle
wakefulness. The PLMs during wakefuness are scored if they are of the night.
0.5-10 seconds in duration, compared to the 0.5-5 second dura- Two studies, both rated as Level 2 evidence, have examined
tion criterion for PLMS. The increase in the maximum movement inter-rater reliability in scoring PLMS.32,33 In the study by Edinger
duration criterion for PLMS stemmed from appreciation that a et al,32 2 raters, blinded to conditions, scored some of the same
large number of patients demonstrate PLMW >5 seconds. Six- records for PLMS associated with arousals with resulting inter-
teen patients and 16 controls in one study underwent the SIT test rater reliability of 83%. In a second study of inter-rater reliability,
as well as 2 nights of polysomnography. A SIT index >40 and a Bliwise et al33 had 5 highly experienced polysomnographic tech-
PLMS index >11 had similar power (81% sensitivity and speci- nologists from different parts of the USA blindly score 24 noctur-
ﬁcity in both cases) and were found to discriminate patients with nal polysomnograms. The reliability for scoring the PLMS index
RLS from control subjects.26 This study used a control group, but was 0.95 while the reliability for scoring the PLMS arousal index
scoring was not done in a blinded fashion and thus was rated as was 0.65. In both studies the scorers were blinded to subjects’
having Level 3 evidence. identities and results of the other scorers. However, subjects were
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 158
Movements in Sleep
lain-Barre Syndrome.35 The other 5 cases were considered to be
Table 4—ICSD-2 Criteria for the Diagnosis of RBD idiopathic. In addition to the previously described speciﬁc poly-
somnographic changes in RBD, the authors also noted increased
(A) The presence of REM sleep without atonia: the EMG finding stage 3 and 4 sleep. In 1993, the same authors reported on a series
of excessive amounts of sustained or intermittent elevation of
of 96 cases and reviewed the world literature on RBD.36 These
submental EMG tone or excessive phasic submental or (upper
and lower) limb EMG twitching. studies revealed a male predominance (>63% of RBD cases), an
(B) At least one of the following is present: association with central nervous system disorders in over 48% of
1. Sleep-related injurious, potentially injurious, or disruptive cases, and a mean age of onset in the early to mid-ﬁfties. Their
behaviors by history. series of 96 cases also indicated that 80% of their patients had
2. Abnormal REM sleep behaviors documented during an elevated percentage of stage 3 and 4 sleep and frequent oc-
polysomnographic monitoring. currence (in about 2/3 of cases) of PLMS during NREM sleep.
(C) Absence of EEG epileptiform activity during REM sleep unless Olson et al indicated that 90 of 93 patients with clinical RBD had
RBD can be clearly distinguished from any concurrent REM either increased phasic or tonic EMG activity in REM sleep with
sleep related seizure disorder.
abnormal motor activity recorded on videotape in 42 (45%) of
(D) The sleep disturbance is not better explained by another sleep
disorder, medical or neurological disorder, mental disorder, patients.37
medication use, or substance use disorder. In 2002, Schenck and Mahowald elaborated on the associa-
tion of RBD with narcolepsy and with Parkinson disease, mul-
not chosen in a consecutive manner, so the studies were rated as tiple system atrophy, progressive supranuclear palsy, Shy-Drager
having Level 2 evidence. syndrome, and olivopontocerebellar atrophy.6 RBD is thought
The scoring of PLMS has some validity based on limited evi- to be a harbinger of neurodegenerative disease in some cases. A
dence. PLMS in the absence of restless legs syndrome (RLS) rare- substantial minority of patients with RBD go on to develop typi-
ly results in clinically signiﬁcant insomnia or hypersomnia for the cal Parkinson disease or atypical parkinsonisms such as dementia
individual with PLMS. However in cases when, in the absence with Lewy body disease and multiple system atrophy, which are
of RLS, the insomnia or hypersomnia can deﬁnitely be attributed characterized neuropathologically by alpha synuclein inclusion
to the PLMS, the diagnosis of PLMD is made. PLMS commonly bodies.6
results in insomnia of the bed partner.3,12 Periodic limb movements While previous versions of the ICSD included only clinical
in wakefulness in RLS patients may result in insomnia and dif- criteria, the ICSD-2 also contains polysomnographic criteria that
ﬁculty remaining seated during the day.3,12 are required for the diagnosis of RBD.12 The full ICSD-2 criteria
Computerized scoring of PLMS has been developed, but these for the diagnosis of RBD are shown in Table 4.
scoring techniques were not considered as part of the review pro-
cess which focuses on reliability, variability, and validity of visual 4.2. Evidence for reliability and validity of rules, specifications, or
scoring of PLM. measures
Three Italian groups looked at inter-observer reliability for
3.3 Summary making a clinical diagnosis of RBD using older ICSD criteria for
The scoring of periodic limb movements has evolved to a more RBD. Six physicians looked at videotaped interviews of patients
sophisticated procedure including amplitude, duration, and fre- with sleep motor-behavior disorders. Inter-rater reliability for the
quency criteria that are primarily based on the previous ASDA diagnosis of RBD was substantial at 83% with a kappa of 0.65.
and IRLSSG consensus recommendations.4,20 Reliability of scor- Nevertheless, some criteria showed a lower inter-rater reliability
ing PLMS is acceptable when standard criteria are employed by such as “limb or body movement associated with dream menta-
trained individuals. tion” (79% agreement with a kappa of 0.58) and “sleep behaviors
that disrupt sleep continuity” (74% agreement with a kappa of
4.0 RAPID EYE MOVEMENT SLEEP BEHAVIOR DISORDER 0.44). The authors recommended further clariﬁcation of termi-
nology of the criteria to improve the reliability.38 This study was
4.1 Background blinded but did not employ consecutive patients and was therefore
rated as having Level 2 evidence. Evidence levels for all articles
In 1986 Schenck et al34 reported on 4 men (aged 67-72 years) on RBD are presented in Table 5 (which can be accessed on the
and a woman (aged 60 years), with a 4-month to 6-year history web at www.aasmnet.org).
of aggressive behaviors during sleep, often occurring during at- In 1992, Lapierre and Montplaisir were the ﬁrst investigators
tempted dream enactment, which injured them or their spouses. to report on a quantitative polysomnographic scoring method
Polysomnography documented REM sleep pathology with vari- that could distinguish RBD patients from control subjects. The
able loss of chin atonia, extraordinarily increased limb-twitch ac- proportion of 20-second REM sleep epochs containing at least
tivity, and increased REM ocular activity and density. A broad 10 seconds of increased muscle tone on the chin EMG, and the
range of REM sleep behaviors were recorded on videotape, in- proportion of 2-second REM mini-epochs that contained phasic
cluding stereotypical hand motions, reaching and searching bursts on the chin EMG were compared in records of 5 male pa-
gestures, punches, kicks, and veriﬁed dream-related behaviors. tients (mean age 58.4 years) and 5 normal controls. Phasic bursts
Schenck et al named this disorder REM sleep behavior disorder were deﬁned as chin EMG activity lasting 0.1-5 seconds with
(RBD).35 In a second publication in 1987 on 9 men and 1 woman an amplitude exceeding 4 times the background EMG activity.
with RBD, Schenck et al reported that in 5 of the 10 cases, there Patients with RBD exhibited increased tonic chin EMG activity
was a closely associated neurologic disorder, such as dementia, during more than 98% of 20-second REM epochs. In contrast,
olivopontocerebellar atrophy, subarachnoid hemorrhage, or Guil-
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 159
AS Walters, G Lavigne, W Hening et al
controls showed increased tonic chin EMG activity in only 6% of disorders, including 19 with Parkinson disease and 273 without
20-second REM epochs. Phasic chin EMG activity was also sig- Parkinson disease. Patients with dementia, multiple system at-
niﬁcantly increased in RBD patients compared to controls (23% rophy, or any other neurodegenerative disease were excluded.
of 2-second REM mini-epochs compared to only 4%). Interest- They used clinical criteria from the previous ICSD, employing
ingly, there was no difference between RBD patients and controls interviews with the patients and their bed partners to diagnose
in the percentage of 2-second mini-epochs of REM sleep con- RBD and using the following polysomnographic criteria to
taining at least one rapid eye movement.39 This study employed identify RBD: REM sleep without muscle atonia seen on the
a control group but was not blinded and was therefore rated as polysomnography in over 50% of epochs associated with motor
Level 3 evidence. behavior visible on videotape. Nine of 19 PD patients (47%)
Consens et al40 validated the same polysomnographic scor- had RBD. RBD occurred in only 4 of 273 patients without PD
ing method for RBD but speciﬁed 30-second epochs instead of (1.8%). The sensitivity of the clinical interview for identifying
20-second epochs. Twenty-three individuals (11 M, 12 F) with RBD when compared to the polysomnographic diagnosis was
an average age of 63 years were studied. Seventeen of these pa- 100% for non-PD patients with a speciﬁcity of 99.6%. Howev-
tients were thought to be at risk for RBD secondary to neuro- er, the sensitivity of the clinical interview for identifying RBD
degenerative disease, and 6 without neurodegenerative disease when compared to the polysomnographic diagnosis was only
were considered to be unlikely to have RBD. Based upon clini- 33% for PD patients, with a speciﬁcity of 90%. Eisensehr et al
cal interview, 9 of the individuals were thought to have probable concluded that the clinical interview may be adequate for the
or possible RBD, and the other 14 were judged to be unlikely to diagnosis of RBD in patients without PD but that polysomnog-
have RBD. Probable RBD was used for subjects with frequent raphy is needed to make the diagnosis of RBD in patients with
and clearly-identiﬁed dream enactment. Possible RBD was used PD. This study quantitatively compared clinical to polysom-
for subjects who had behavioral episodes that were not clearly nographic data and employed consecutive patients but was not
repeated dream enactment. Bed partners completed a question- blinded; it was therefore rated as Level 2 evidence.
naire that quantiﬁed RBD symptom severity. Two different Eisensehr and coworkers also performed a study on muscle
polysomnographic RBD scores were generated from 2 consecu- activity during REM sleep in patients with RBD.43 They found a
tive nocturnal studies in each patient. The ﬁrst was the percent- correlation between increased muscle activity during REM sleep
age of 30-second REM epochs with at least 15 seconds of toni- and decreased striatal presynaptic dopamine transporters (i.e., de-
cally maintained chin EMG, and the second was the percentage creased IPT uptake) as detected by SPECT brain imaging. This
of 3-second REM mini-epochs that contained phasic bursts on study of RBD presented a new method for scoring muscle activ-
the chin EMG. Both of these measures combined were higher ity in REM sleep; for the ﬁrst time extremity EMG activity was
in patients with clinical determination of probable or possible scored (in addition to chin EMG activity), and muscle activity
RBD than in subjects judged unlikely to have RBD. If patients lasting 5-10 seconds was scored. Each 10-second epoch of REM
with a polysomnographically determined RBD score of <10% sleep was scored for increased chin or (upper/lower) extremity
(of total REM sleep time) were eliminated from the analysis (as EMG activity, deﬁned as increased muscle activity of >1 second
suggested by the receiver operator curve) then the RBD score in during each 10-second REM sleep epoch (i.e., lasting at least 10%
comparison to the physician clinical impression had a sensitivity of each 10-second REM sleep epoch). However, “each persistent
of 89% and a speciﬁcity of 57%. The overall polysomnographic increase of muscle activity lasted at least 0.5 seconds.” In other
measures correlated with symptom severity as determined by words, at least 2 separate 0.5-second bursts of muscle activity or
the bed partners of the patients (r = 0.42). The clinical impres- at least a single 1 second burst of muscle activity would score a
sion by the clinician and the RBD symptom score as determined 10-second epoch of REM sleep with increased muscle activity.
by the patient’s bed partner were highly correlated. The authors Eisensehr et al developed their scoring methods because of their
concluded that this quantitative method for scoring RBD from interest in identifying “long-lasting muscle activity” during REM
the polysomnogram showed good correlation with the clinical sleep in RBD, and also because they identiﬁed some limitations
impression of RBD and with a rating scale for the severity that with previous scoring methods.
was ﬁlled out by the bed partner.40 This study compared clinical Focusing on the opposite end of the spectrum, a recent study
to polysomnographic data in a quantitative manner, employed analyzed short-lasting (approximately 0.1 second) phasic EMG
consecutive patients, and was blinded; it is therefore rated as activity during REM and NREM sleep in normal subjects and in
Level 2 evidence. patients with Parkinson disease.44 Five different muscle groups had
Sforza et al41 in their study of RBD, pointed out that although EMG recordings during sleep; submental, bilateral anterior tibialis,
all 13 patients with primary (i.e. idiopathic) RBD had complete and bilateral brachioradialis muscles. A “phasic electromyograph-
absence of atonia for at least 90% of REM sleep epochs, only half ic” (REM) metric was developed, with the authors suggesting its
of 39 patients with secondary (i.e. symptomatic) RBD showed the future usefulness in identifying patients with neurodegenerative
pattern of complete absence of atonia for at least 90% of REM disorders who develop motor disinhibition during sleep as a marker
sleep epochs. However, the other half of the secondary cases did of disinhibition of midbrain dopaminergic pathways.
have loss of REM atonia for at least the majority of REM sleep
epochs. This study did not employ a control group; it was not 4.3 Summary
blinded nor did it employ consecutive patients. It is therefore rat-
ed as Level 4 or observational/case series evidence. The clinical diagnosis of the REM behavior disorder (RBD)
Eisensehr et al42 compared the sensitivity and speciﬁcity of has relatively good reliability, but the reliability of polysomno-
the clinical diagnosis of RBD to polysomnographic diagnosis. graphic criteria for REM without atonia is largely unexplored.
They reviewed the polysomnograms of 292 patients with sleep Polysomnographic scoring of REM without atonia does correlate
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 160
Movements in Sleep
5.2 Evidence for reliability and validity of rules, specifications, or
Table 6—ICSD-212 definition of Sleep bruxism measures (see Table 7—which can be accessed on the web at www.
A. The patient reports or is aware of tooth-grinding sounds or tooth
clenching during sleep.
Sleep bruxism is identiﬁed when repetitive (phasic/rhythmic)
B. One or more of the following is present:
1. Abnormal wear of the teeth.
or prolonged (tonic/sustained) electromyographic (EMG) poten-
2. Jaw muscle discomfort, fatigue or pain and jaw lock upon tials of the masseter muscle (alternatively, the temporalis muscle)
awakening. are seen on PSG recording, if the amplitude and duration exceed
3. Masseter muscle hypertrophy upon voluntary forceful various criteria; e.g., a rise above a certain µV level, a certain
clenching. amount of time background activity or above a threshold set at a
C. The jaw muscle activity is not better explained by another current fraction of voluntary maximum voluntary clenching activity per-
sleep disorder, medical or neurological disorder, medication use, formed while awake. Episodes are classiﬁed as phasic/repetitive,
or substance use disorder. tonic/sustained, or mixed episodes according to empirical criteria
with the clinical diagnosis of RBD based on bed partner ratings. as described below.
Standardized scoring demonstrating validity has employed the Reding and colleagues, in 1968, were the ﬁrst to suggest a
percentage time with elevated chin EMG during REM sleep and “visual” method for SB scoring.61 In an evidence Level 3 case
the presence of phasic EMG bursts in the chin and/or limb EMG control study, they suggested a criteria of 40 μV for amplitude of
during REM sleep. Guidelines for PSG scoring in RBD will re- masseter muscle EMG potential and a minimum episode length
quire further validation studies. 1.5 seconds. Episodes with ≥5 EMG potentials were labeled as
rhythmic (now named phasic) if no EMG potential exceeded 1.5
5.0 SLEEP BRUXISM seconds or, otherwise as nonrhythmic. The scoring discrimination
between rhythmic and nonrhythmic episodes was possible with
5.1 Background an inter-scorer agreement of 92%. Interestingly, the mean episode
duration was estimated at 9 seconds (range 2.7 to 66.5 seconds).
Sleep bruxism (SB) is a stereotyped movement disorder occur- SB occurred mainly in stage 2 NREM sleep.
ring during sleep and characterized by tooth grinding (TG) and/ The next signiﬁcant paper came from Ware and Rugh,62 who
or clenching.12 SB should be distinguished from daytime-awake suggested criteria based on an evidence Level 5 case control study.
bruxism, which is a jaw muscle clenching habit or tic that is It was proposed that SB could be scored from electroencephalo-
mainly related to stress or anxiety. Awake/daytime tooth grinding graphic (EEG) artifacts rather than from EMG. SB-related activi-
is rare in otherwise healthy subjects. In the presence of medical ty was scored if the EEG potential exceeded 75 mV (probably μV
disorders, medication or drug use, tooth grinding is labeled sec- after ampliﬁcation correction) or twice the background activity.
ondary or iatrogenic.14,45,46 The most important ﬁnding from the Ware and Rugh62 study was
SB usually begins in childhood with the prevalence decreasing the classiﬁcation of SB episodes into phasic, tonic, and tonic-pha-
with age.47,48 No sex differences have been observed. The con- sic (later termed mixed). Phasic episodes included burst durations
sequences of SB are tooth destruction, temporomandibular joint lasting >0.25 seconds with ≥2 discrete bursts; tonic or sustained
pain or lock (SB patients are 2-3 times more at risk of lock jaw, SB required an EMG potential lasting longer than 2.0 seconds.
and 20% of them report moderate pain upon awakening), tem- Their results revealed that SB occurred in both REM and NREM
poral headaches, or cheek-biting (which is worse if xerostomia sleep, but their sample of SB patients had a large age range and
is present).14,49 The loud noise made by the tooth grinding can included cases that would now be considered as “secondary” SB
greatly disturb the rest of sleeping partners. (due to use of medication, or patients presenting with insomnia,
The following major risk factors have been shown to exacer- excessive daytime sleepiness, or depression).
bate SB-tooth grinding: 1) smoking, caffeine, and heavy alcohol In a series of evidence Level 5 clinical observational studies
drinking (odds ratio [OR] = 1.3 to 1.950,51) ; 2) type A personality in geriatric subjects or in patients with respiratory disturbances,
– anxiety14,50,52-54; 3) other sleep disorders such as snoring (OR = Okeson et al63,64 proposed that masseter EMG potential needed to
1.4), sleep apnea (OR=1.8), or periodic limb movements (co-oc- exceed 40% of maximum voluntary clenching and to last for at
curring in 10%)14,48-50,55; 4) approximately 60%–80% of SB epi- least 2 seconds. It was found that SB episodes had a mean dura-
sodes are associated with brief nonperiodic body movements.56,57 tion of 5.95 seconds (range 2 to 375 seconds). Furthermore, an
Clinically, SB is diagnosed following the report of recent/fre- index for SB, i.e., number of episodes/hour of sleep, was pro-
quent tooth grinding by a sleeping partner or parent, which is the posed and found to have a mean value of 3.03. The mean age
most reliable criterion, the presence of tooth wear or jaw muscle of subjects was 70 and 50 years, respectively, in their 2 studies.
hypertrophy, and awareness of jaw clenching while awake.14 The The Montreal sleep laboratory, in an evidence Level 5 case series
sleep laboratory diagnosis of SB requires that it be distinguished based on Reding,61 Ware and Rugh,62 and Okeson,63,64 suggested
from other oromandibular activity during sleep. The use of au- that SB episodes can be classiﬁed as phasic (at least 3 repetitive
dio and video together with polygraphic recordings in the sleep EMG bursts; each burst >0.25 and <2 seconds), tonic (bursting
laboratory allows us to control quantitatively for the presence of lasting >2 seconds), or mixed. These authors required an interval
non-SB-related oromandibular activities such as myoclonus (ob- of 3 seconds between episodes, proposed an EMG amplitude cri-
served in 10% of SB patients) or swallowing, coughing, grunting, terion of >40% of the maximal voluntary clenching contraction
tooth tapping. These non-SB events can represent up to 40% of awake,59 and excluded jaw muscle myoclonic contractions de-
oromandibular EMG events.58-60 The ICSD12 deﬁnition of SB is ﬁned as isolated contractions lasting less than 0.25 seconds.7,60 In
shown in Table 6. an evidence Level 4 case control study, Sjoholm and colleagues57
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 161
AS Walters, G Lavigne, W Hening et al
Table 8—ICSD-212 Definition of Sleep Related Rhythmic Movement Table 10—ICSD-212 Definition of fragmentary myoclonus
(A) The patient exhibits small movements of the fingers, toes, or
(A) The patient exhibits repetitive, stereotyped, and rhythmic motor corners of the mouth or small muscle twitches resembling either
behaviors. physiologic hypnic myoclonus or fasciculations. The movements
(B) The movements involve large muscle groups. may be present during “…sleep/wake transitions…..” or sleep.
(C) The movements are predominantly sleep related, occurring (B) Polysomnographic monitoring demonstrates recurrent and
near naps or bedtime, or when the individual appears drowsy or persistent very brief (75 to 150 millisecond) EMG potentials in
asleep. various muscles occurring asynchronously and asymmetrically
(D) The behaviors result in a significant complaint as manifest by at in a sustained manner without clustering.
least one of the following: (C) More than 5 potentials per minute are sustained for at least 20
1. Interference with normal sleep. minutes of NREM stages 2, 3, or 4 sleep.
2. Significant impairment in daytime function (D) The disorder is not better explained by another sleep disorder,
3. Self-inflicted bodily injury that requires medical treatment medical or neurological disorder, medication use, or substance
(or would result in injury if preventable measures were not use disorder.
(E) The rhythmic movements are not better explained by another night.61,64 Lavigne et al71 in an evidence Level 5 case series re-
current sleep disorder, medical or neurological disorder, mental ported that the across night coefﬁcient of variation of the index
disorder, medication use, or substance use disorder. of SB episodes was 25% while that of tooth grinding sounds was
53%. Currently, the presence of tooth wear is no longer consid-
proposed scoring criteria based on masseter muscle activity that ered sufﬁcient to diagnose SB.14,72
had an EMG elevation twice the background activity if it lasted
>2 seconds. Later, in a published evidence Level 3 abstract of 5.3 Summary
a comparative study of normal and SB patients, they suggested
that bruxism episodes only needed an amplitude >20% of awake Sleep bruxism is a stereotyped movement disorder occurring
maximal voluntary jaw clenching (about 30 μV).65 They classi- during sleep and characterized by tooth grinding. Polysomno-
ﬁed the episodes as phasic/rhythmic if ≥3 bursts were present. In graphic scoring criteria using a minimum of bursts as well as in-
1996, a non-controlled study using ambulatory recording of mas- terval and amplitude criteria has high reliability and reasonably
seter muscle proposed that EMG potentials over 3% or 10% the distinguishes the clinical diagnosis of bruxism in a limited num-
maximal voluntary contraction be adequate to discriminate SB.66 ber of controlled studies.
An evidence Level 4 comparative study done with a large age
spectrum of SB patients (65% with reports of morning headaches) 6.0 SLEEP RELATED RHYTHMIC MOVEMENT DISORDER
used empirical amplitude criteria of twice the level of the EMG 6.1 Background
signal just before the episodes, or more than 25% of maximal vol-
untary jaw clenching.67 The episodes were scored when a mini- Sleep related rhythmic movement disorder (RMD) consists of
mum of 3 bursts of less than 1 second apart occurred in series; an stereotyped and repetitive movements such as body rocking and
episode ended when there was an interval of 3 seconds without head banging usually in the wake/sleep transition, stages 1 and
SB activity. As in previous studies, a clear tachycardia was associ- 2, and less frequently in REM sleep and slow wave sleep.15,73,74
ated with the SB episodes.61,68,69 Patients are amnestic for episodes of RMD when they occur in the
In an evidence Level 2 study design to assess the reliability of deeper stages of sleep.15 Sleep related rhythmic movements are
SB scoring criteria, Lavigne and colleagues 58 found that empiri- common in childhood but only rarely persist into adulthood.74-78
cal criteria derived from Reding,61 Ware and Rugh,62 Okeson,64 as Although RMD that persists beyond childhood has been reported
described by Velly–Miguel,59 correctly classiﬁed 83.3% of severe to be more frequently associated with mental retardation or au-
bruxers (grinding > 5 nights/week) and 81.3% of controls.58 The tism, it has been reported in adults with normal mentation.74-78
requirement for 4 episodes of SB/hour of sleep or 25 EMG bursts/ The average age of onset is 9 months, and by 10 years of age
hour reliably differentiated frequent tooth grinders from controls. the majority of subjects no longer complain of headbanging or
A minimum of 2 audible tooth-grinding episodes conﬁrmed with body rocking.74 Some patients may experience injury as a result
audio and video recordings during PSG needed to be present for of RMD.15 In one series of 10 patients, age range 7-24 years, 6
the diagnosis. This study was based on data collected in 18 young of the patients had symptoms of Attention Deﬁcit Hyperactivity
controls and SB patients (primary SB only). The differentiation Disorder (ADHD).78 The current ICSD12 deﬁnition of RMD can
between phasic, tonic, and mixed episodes was possible. An in- be found in Table 8.
terscorer kappa=0.6 and re-scoring twice of SB episodes/hour of
sleep by the same observer was high with an ICC of 0.95.58 Later, 6.2 Evidence for reliability of rules, specifications, or measures.
in a larger evidence 3 sample study, it was found that the 40%
amplitude criteria used in visual scoring analysis (derived from Few studies have been conducted on RMD. The majority have
Okeson study64) was set too high; the post hoc analysis of the been observational case reports or small case series and are all
scoring criteria used in practice with a computer system revealed rated as Level 4 evidence (see Table 9—which can be accessed
that the minimum EMG amplitude selected was >10% but closer on the web at www.aasmnet.org). Dyken et al studied seven pa-
to 20%.14 Interestingly, this was similar to the empirical 20% used tients, aged 1-12 years, who experienced 37 distinct episodes
by Sjöholm et al.70 of RMD with a frequency of 0.5- 2 Hz.15 For all seven patients,
It is well accepted that SB tooth grinding varies across the RMD was present from 1 to 59 minutes during a typical night’s
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 162
Movements in Sleep
7.0 EXCESSIVE FRAGMENTARY MYOCLONUS
Table 12—ICSD-212 Definitions of Hypnagogic Foot Tremor and
Alternating Leg Muscle Activation 7.1 Background
Hypnagogic Foot Tremor Excessive fragmentary myoclonus (EFM) was ﬁrst described
A. The patient reports foot movements (directly experienced or by Broughton and Tolentino in an observational case report with
observed by others) that occur at the transition between wake an evidence Level 4.81 The duration of the EFM bursts was <150
and sleep or during light sleep. msec and was multi-focal involving arms, legs, and face in an
B. Polysomnographic or activity monitoring demonstrates:
asynchronous and random fashion. Such short duration EMG
1. Recurrent EMG potentials or foot movement typically at 1
to 2 Hz (range 0.5-3 Hz) in one or both feet.
bursts are usually associated with small amplitude movements
2. Burst potentials longer than the myoclonic range (greater or no visible movement and are not associated with large move-
than 250 milliseconds) and usually less than one second. ments across large joint spaces. The muscle bursts occurred as
3. Trains lasting 10 or more seconds. often as 2/second. EFM, phasic REM twitches, and physiologi-
C. The disorder is not better explained by another sleep disorder, cal hypnic myoclonus in stage 1 were all characterized by either
medical or neurological disorder, mental disorder, medication no visible movement or by small movements of the ﬁngers, toes,
use, or substance use disorder. or corners of the mouth. The brief duration of the bursts most
Alternating Leg Muscle Activation resembled phasic REM twitches, but EFM occurred in NREM
(A) Polysomnography demonstrates a pattern of brief, repeated
sleep and did not occur in clusters of phasic bursts the way that
activation of the anterior tibialis in one leg alternating with
similar activation in the other leg.
the phasic bursts of REM sleep do. Rather, the phasic bursts of
(B) At least 4 discrete and alternating muscle activations occur with EFM occurred in a more dispersed pattern. The burst duration of
less than two seconds between activations. EFM also resembled that of physiological fragmentary myoclo-
(C) Individual activations last between 0.1 and 0.5 seconds and nus (also called physiological hypnic myoclonus) seen in stage 1
occur at a frequency of 0.5 to 3 Hz (usually 1 to 2 Hz) sleep,82 but EFM was present in all NREM stages and not just in
(D) Sequences of alternating activations last between one and 30 stage 1.
seconds and may recur periodically (e.g. one to four times per Broughton and Tolentino reported on another 38 cases and
minute). made the point that almost all the patients were older males, and
the EMG pattern was infrequent or absent in pre-sleep wakeful-
recording. Individual uninterrupted episodes of RMD lasted from
ness.83 The characteristics of the EMG pattern seen in this series
4 seconds to 21 minutes. In a case series study of 5 female and 5
was similar to those previously described with some additional
male patients aged 7-24 years by Stepanova et al,78 the frequency
features. The amplitude of the movements was usually >50 μV
of the movements was again noted to be 0.5-2 Hz with 1 Hz as
and occasionally 150 μV. Also, to be labeled as having EFM, pa-
the prevailing frequency. Absence of REM atonia was sometimes
tients had to have 5 such burst potentials/minute for ≥20 consecu-
seen for 30-60 seconds before and after a series of movements, a
tive minutes of stage 2, 3, or 4 sleep. Although Broughton and
ﬁnding not seen in the Dyken series.15 Using quantitative indices,
Tolentino found EFM in association with various sleep disorders,
REM atonia was thought to be normal in another 2 patients with
they also found EFM in 4 of 10 controls, although to a much
RMD despite having RMD episodes in REM79 In a study of a
more limited degree. In addition, the 38 patients with EFM were
single 9-year-old patient, RMD was associated with the A-phase
referred to the sleep laboratory for reasons other than suspected
of the cyclic alternating pattern.80
EFM. Thus, the causal role of EFM in sleep disruption is very
Kohyama et al79 summarized the literature on RMD and found
open to question. In addition, patients are often unaware of the
that in 33 children and adults, exclusive of RMD in wakeful-
movements, and EFM is often an incidental ﬁnding on EMG.
ness, 15 had episodes in NREM only, 10 had episodes during
Concurrence with other sleep disorders such as PLMS, OSA,
both NREM and REM, and 8 had episodes exclusively in REM.
RLS, narcolepsy, or insomnia is not uncommon. This study did
They suggest that the underlying mechanism of RMD that ap-
employ a control group but was not blinded and is therefore rated
pears exclusively in NREM is fundamentally different than RMD
as Level 3 evidence.81 Table 10 shows the current ICSD deﬁnition
that appears exclusively in REM. It should be pointed out that the
of excessive fragmentary myoclonus.12
stage distribution from this review should not be considered to
be reﬂective of the stage distribution of RMD in the population
7.2 Evidence for reliability of rules, specifications, or measures
as a whole, since the review summarizes a compendium of case
(see Table 11—which can be accessed on the web at www.aasmnet.
6.3 Summary Lins et al84 reported on the use of a fragmentary myoclonus
index (FMI) which they used to quantitate EFM in 11 male pa-
Rhythmic movements such as body rocking and head banging
tients. Each 30-second epoch of sleep was divided into ten 3-sec-
in and around the time of sleep are common in normal infants
ond mini-epochs. The number of mini-epochs with one or more
and children but may also occur in adults. Case series describe
EFM potentials exceeding 50 μV was then counted for each 30-
a characteristic frequency range for these movements. Rhythmic
second epoch, the number therefore ranging from 0 to 10. The
movements uncommonly result in significant sleep disturbance,
FMI was then the calculated mean of individual epoch counts.
impaired daytime function, or sleep related injury, but, if they do,
Using this methodology, EFM was less in the ﬁrst hour of the
the term sleep related rhythmic movement disorder is employed.
night than the remaining hours. They found EFM in all stages
of sleep with progressively decreasing amounts from REM to 1
to 2 to slow wave sleep (SWS). With removal of SWS from the
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 163
AS Walters, G Lavigne, W Hening et al
analysis, it was found that the decrease of EFM in the ﬁrst hour of another and was recorded from both anterior tibialis muscles. In
sleep was due to a suppressant effect of SWS. It is not clear with both patients the tremor occurred during pre-sleep wakefulness,
this methodology that Lins et al had adequately discriminated during arousals, usually persisted in stage 1 at reduced amplitude,
phasic REM twitches from EFM in REM sleep. In this study there occasionally occurred in stage 2, and was always absent in stages
was no increase in EFM with increasing age. Patients with EFM 3, 4, and REM. Both patients had a history of sleep disturbance.
did not display more sleep fragmentation or lighter sleep than an This was an observational study with an evidence Level of 4.
age- and sex-matched patient control group without EFM, again In 2001, Wichniak et al8 looked at the prevalence of this same
suggesting that EFM may be a coincidental ﬁnding on polysom- disorder in 375 consecutive sleep-disordered patients and found
nography. Further support of this contention is the fact that none a prevalence of 7.5%. The HFT usually occurred in short series
of the patients in any other studies had symptoms that speciﬁcally of 10 to 15 seconds but sometimes persisted for up to 2 minutes.
pointed to EFM as a possible diagnosis.81,82,84 The study by Lins et They had a high night-to-night variability and were detected as
al employed a control group but was not blinded and therefore is rhythmic, oscillating movements of the whole foot or toes. The
rated as Level 3 evidence. frequency of the tremor in the 28 patients studied was usually 1-2
Vetrugno et al85 reported on an additional 2 patients with EFM Hz (range 0.3 to 3.3 Hz). The individual muscle bursts usually
and noted the similarity between those EFM movements not as- varied from 300 to 700 msec in duration (range 75 to 2200 msec).
sociated with movement across a joint space and fasciculations. The amplitude of the bursts was variable and ranged between
They found that EFM prevailed in NREM sleep stages and the 60 and 650 μV. The HFT occurred at highest intensity during
second part of the night and was completely absent in wakeful- sleep/wake transitions, during arousals and usually persisted in
ness. Daytime EMG in wakefulness was negative for any muscle stages 1 and 2. They did not occur during stages 3 and 4. All sub-
pathology including fasciculations. EEG-EMG back-averaging jects who were aware of the movements reported that they were
did not show any cortical potentials related to the twitches, sug- able to be repressed voluntarily. Although there were a variety
gesting that EFM does not have a cortical origin. The characteris- of other sleep disorders associated with HFT, the patients were
tics of the movements were as in previous studies,81,83,84 however chosen from those who came to the sleep laboratory with other
visible movements were associated with longer duration (>150 complaints. A normal control group of 20 subjects had a similar
msec) or repetitive potentials or with higher amplitude (>200 μV) prevalence of HFT (5%). This study employed a control group,
potentials.85 This study was strictly an observational case series utilized consecutive patients, but was not blinded. It is therefore
and is rated as Level 4 evidence. rated as Level 2 evidence.
Chervin et al10 reported similar movements but noted that they
7.3 Summary alternated from one leg to another and coined the term alternating
leg muscle activation (ALMA). The records of 16 patients with
As described in case series, excessive fragmentary myoclonus ALMA were identiﬁed from 1500 polysomnographic records per-
(EFM) has minimum frequency and EMG amplitude criteria on formed for other reasons. ALMA was deﬁned as more than 4 dis-
polysomnography. There are no studies deﬁning reliability and crete and alternating muscle activations with <2 seconds between
validity of scoring of EFM using these criteria. If EFM is ob- activations. The frequency of the contractions was approximately
served on the EMG of the polysomnogram, very small move- 1-2 Hz (range 0.5-3.0 Hz) with a burst duration of 100-500 msec.
ments are observed or, in some cases, no visible movement is Sequences of movements usually lasted between several and 20
observed. It should be made clear that there has been no deﬁnite seconds (range 1.4 to 22.2 seconds). The number of ALMA se-
link to any clinical consequence for this phenomenon and it may quences during sleep per record ranged from 1 to 40. ALMA oc-
simply represent a benign movement occurring during sleep with curred primarily in stages 1, 2, and REM sleep, but particularly
a distinctive pattern on electromyography. during arousals. When patients had repeat sleep studies, ALMA
was almost always present on both studies. Ten of the 16 patients
8.0 HYPNAGOGIC FOOT TREMOR AND ALTERNATING LEG had >5 PLMS/hr of sleep, and 12 of the 16 patients were taking
MUSCLE ACTIVATION antidepressant medications. The authors speculated that ALMA
8.1 Background could represent transient facilitation of a spinal central pattern
generator for locomotion perhaps due to serotonergic effects of
Hypnagogic foot tremor (HFT) and alternating leg muscle ac- antidepressants. Because the frequency of the movements and the
tivation (ALMA) are similar both clinically and in EMG pattern. duration of the movements are similar to HFT, it is possible that
They may represent variations of the same basic underlying dis- ALMA is a form of HFT. As mentioned above, some patients with
order, but further work needs to be done to determine this. Most HFT have it in both legs.9 In addition, some of the patients in the
studies done in both HFT and ALMA are observational, although Chervin study had activation only in one leg.10 A ﬁnal interesting
some case series exist. ICSD-212 deﬁnitions are shown in Table observation by Chervin et al was that in some cases, 5- to 10-
12 and evidence ratings are shown in Table 13 (which can be ac- second sequences of ALMA recurred at intervals that resembled
cessed on the web at www.aasmnet.org). those between PLMS (e.g. 20-30 seconds). The study of Chervin
et al is an observational case series and therefore rated as Level
8.1 Evidence for reliability of rules, specifications, or measures 4 evidence. Consentino et al performed an electrophysiological
analysis of ALMA. The alternating leg movements were preced-
HFT was ﬁrst reported in 1988 in 2 patients by Broughton.9 In ed by cyclic alternating pattern A phases and increased heart rate
the ﬁrst patient the frequency of the tremor was 0.5-1.0 Hz, how- in most instances.11
ever the frequency was not reported for the second patient. In both
patients the tremor was bilateral, occurred in one leg more than
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 164
Movements in Sleep
8.3 Summary 4:101-119.
4. Zucconi M, Ferri R, Allen R, et al. The ofﬁcial World Association
As described in case series, alternating leg muscle activa- of Sleep Medicine (WASM) standards for recording and scoring pe-
tion (ALMA) and hypnagogic foot tremor (HFT) have a range riodic leg movements in sleep (PLMS) and wakefulness (PLMW)
developed in collaboration with a task force from the International
of frequencies and EMG burst durations on polysomnography.
Restless Legs Syndrome Study Group (IRLSSG). Sleep Med 2006;
There are no studies deﬁning reliability and validity of scoring 7:175-83.
of ALMA or HFT using speciﬁc criteria. It should be made clear 5. Ferri R, Zucconi M, Manconi M, Plazzi G, Bruni O, Ferini-Strambi
that there have not been any reported clinical consequences of L. New approaches to the study of periodic leg movements during
these 2 newly described phenomena to date, and they may simply sleep in restless legs syndrome. Sleep 2006; 29:759-69.
be benign movements during sleep associated with characteristic 6. Schenck CH, Mahowald MW. REM sleep behavior disorder: clini-
electromyographic patterns. Further work is needed to determine cal, developmental, and neuroscience perspectives 16 years after its
if these 2 disorders are variants of the same underlying disorder. formal identiﬁcation in SLEEP. Sleep 2002;25:120-38.
7. Broughton RJ, Tolentino MA, Krelina M. Excessive fragmentary
myoclonus in NREM sleep: a report of 38 cases. Electroencepha-
9.0 UNRESOLVED ISSUES AND FUTURE RESEARCH
logr Clin Neurophysiol 1985;61:123-33.
The extent of the clinical consequences of many of the sleep 8. Wichniak A, Tracik F, Geisler P, Ebersbach G, Morrissey SP, Zul-
related movement disorders remains to be determined. In RBD, ley J. Rhythmic feet movements while falling asleep. Mov Disord
and to some extent in bruxism, RMD, and PLMD, there is de-
9. Broughton RJ. Pathological fragmentary myoclonus, intensiﬁed
veloping evidence of physiologic or clinical consequences that hypnic jerks and hypnagogic foot tremor: three unusual sleep-re-
supports the evolution of scoring of these sleep-associated move- lated movement disorders. In: Koella WP, Obal F, Schulz H, Visser
ment disorders. Although past research has mostly focused on P, eds. Stittgart: Gustav Fischer Verlag, 1988: 240-3.
the relationship of these disorders to insomnia, hypersomnia, or 10. Chervin RD, Consens FB, Kutluay E. Alternating leg muscle ac-
sleep related injury, current research is focusing on the relation- tivation during sleep and arousals: a new sleep-related motor phe-
ship of these disorders to other medical, neurological, and psychi- nomenon? Mov Disord 2003;18:551-9.
atric problems. For example, REM sleep behavior disorder may 11. Consentino FI, Iero I, Lanuzza B, Tripodi M, Ferri R. The neuro-
be a predictor of Parkinson disease,42 PLMS may be a predictor physiology of the alternating leg muscle activation (ALMA) during
sleep: study of one patient before and after treatment with prame-
of mortality in end-stage renal disease,86 and rhythmic movement
pexole. Sleep Med 2006; 7:63-71.
disorder is frequently associated for unknown reasons with atten- 12. American Academy of Sleep Medicine. The international classiﬁca-
tion deﬁcit hyperactivity disorder (ADHD).15,78 Current research tion of sleep disorders. 2nd ed. Westchester: American Academy of
is focused on the relationship of PLMS to anxiety and depres- Sleep Medicine, 2005.
sion,87 to ADHD,88,89 as well as to hypertension,90 heart disease,91 13. Butler JV, Mulkerrin EC, O’Keefe ST. Nocturnal leg cramps in
and stroke. Medical comorbidities of all movement disorders need older people. Postgrad Med J 2002; 78:596-8.
to be more fully examined. 14. Lavigne GJ, Manzini C, Kato T. Sleep bruxism. In: Kryger MH,
The metrics for describing movements during sleep have been Roth T, Dement WC, editors. Principles and practice of sleep medi-
based predominantly on observational studies and consensus cine. Philadelphia: Elsevier Saunders, 2005: 946-59.
15. Dyken ME, Lin-Dyken DC, Yamada T. Diagnosing rhythmic
though reliability of scoring amplitude, duration, and periodicity
movement disorder with video-polysomnography. Pediatr Neurol
of many measures is beginning to be explored. Further testing of 1997;16:37-41.
inter- and intra-scorer reliability of standardized scoring strategies 16. Ohayon MM, Guilleminault C, Priest RG. Night terrors, sleepwalk-
as well as correlation with physiologic events, clinical diagnosis ing, and confusional arousals in the general population: their fre-
and outcomes will be necessary in order to fully understand the quency and relationship to other sleep and mental disorders. J Clin
role of polysomnographic scoring of movements as research and Psychiatry 1999;60:268-76.
clinical tools. 17. Fitch K, Berstein SJ, Aguilar MD, Burnand B, Lacalle JR, Lazaro L.
The RAND/UCLA Appropriateness Method User’s Manual. Santa
MOVEMENTS TASK FORCE MEMBERS Monica, Ca: RAND Corporation, 2001.
18. Iber C, Ancoli-Israel S, Chesson AL, Quan SF. The AASM manual
Movements Task Force members participating in evidence review and/or for the scoring of sleep and associated events: rules, terminology,
consensus balloting included: Arthur S. Walters, chair, Richard P. Allen, and technical speciﬁcations, 1st ed.: Westchester, Illinois: American
Donald L. Bliwise, Sudhansu Chokroverty, Wayne A. Henning, Clete A. Academy of Sleep Medicine, 2007.
Kushida, Gilles Lavigne, Mark W. Mahowald, Daniel Picchietti, Sonia 19. Sackett DL. Rules of evidence and clincal recommendations for the
Ancoli-Israel, and Carlos H. Schenck management of patients. Can J Cardiol 1993;9:487-9.
20. Bonnet MH, Carley D, Carskadon MA, et al. Recording and scoring
REFERENCES leg movements: The Atlas Task Force. Sleep 1993; 16:748-59.
21. Coleman RM, Pollak CP, Weitzman ED. Periodic movements in
1. Rechtschaffen A, Kales A. A manual of standardized terminology, sleep (Nocturnal myoclonus): relation to sleep disorders. Ann Neu-
techniques and scoring system for sleep stages of human subjects. rol 1980; 8:416-21.
US Department of Health, Education, and Welfare Public Health 22. American Academy of Sleep Medicine. The international classiﬁ-
Service - NIH/NIND; 1968. cation of sleep disorders, revised, Rochester, MN: American Sleep
2. Walters AS, Aldrich MS, Allen R, et al. Toward a better deﬁnition of Disorders Association, 1997.
the restless legs syndrome. Mov Disord 1995;10:634-42. 23. Chabli A, Michaud M, Montplaisir J. Periodic arm movements in
3. Allen RP, Picchietti DL, Hening WA, et al. Restless legs syndrome: patients with the restless legs syndrome. Eur Neurol 2000; 44:133-
diagnostic criteria, special considerations, and epidemiology. A re- 8.
port from the Restless Legs Syndrome diagnosis and epidemiology 24. Yokota T, Shiojiri T, Hirashima F. Sleep-related periodic arm move-
workshop at the National Institutes of Health. Sleep Med 2003; ment. Sleep 1995;18:707-8.
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 165
AS Walters, G Lavigne, W Hening et al
25. Gschliesser V, Brandauer E, Ulmer H, Poewe W, Hogl B. Periodic 46. Vollenweider FX, Gamma A, Liechti M, Huber T. Psychological
limb movement counting in polysomnography: effects of amplitude. and cardiovascular effects and short-term sequelae of MDMA (“ec-
Sleep Med 2006;7:249-54. stasy”) in MDMA-naive healthy volunteers. Neuropsychopharma-
26. Montplaisir J, Boucher S, Nicolas A, et al. Immobilization tests and col 1998;19:241-51.
periodic leg movements in sleep for the diagnosis of restless leg 47. Hublin C, Kaprio J, Partinen M, Koskenvuo M. Sleep bruxism based
syndrome. Mov Disord 1998;13:324-9. on self-report in a nationwide twin cohort. J Sleep Res 1998;7:61-7.
27. Bliwise DL, Carskadon MA, Dement WC. Nightly variation of peri- 48. Lavigne GJ, Montplaisir JY. Restless legs syndrome and sleep
odic leg movements in sleep in middle aged and elderly individuals. bruxism: Prevalence and association among Canadians. Sleep 1994
Arch Gerontol Geriatr 1988;7:273-9. 17:739-43.
28. Mosko SS, Dickel MJ, Ashurst J. Night-to-night variability in sleep 49. Bader G, Lavigne G. Sleep bruxism; an overview of an oroman-
apnea and sleep in sleep-related periodic leg movements in the el- dibular sleep movement disorder. Review Article. Sleep Med Rev
derly. Sleep 1988;11:340-8. 2000;4:27-43.
29. Culpepper WJ, Badia P, Shaffer JI. Time-of-night patterns in PLMS 50. Ohayon MM, Li KK, Guilleminault C. Risk factors for sleep brux-
activity. Sleep 1992;15:306-11. ism in the general population. Chest 2001;119:53-61.
30. Hornyak M, Kopasz M, Peigfe B, Riemann D, Voderholzer U. 51. Lavigne GL, Lobbezoo F, Rompre PH, Nielsen TA, Montplaisir J.
Variability of periodic leg movements in variious sleep disorders: Cigarette smoking as a risk factor or an exacerbating factor for rest-
implications for clinical and pathophysiological studies. Sleep less legs syndrome and sleep bruxism. Sleep 1997;20:290-3.
2005;28:331-5, 52. Major M, Rompre PH, Guitard F, et al. A controlled daytime chal-
31. Sforza E, Haba-Rubio J. Night-to-night variability in periodic leg lenge of motor performance and vigilance in sleep bruxers. J Dent
movements in patients with restless legs syndrome. Sleep Med Res 1999;78:1754-62.
2005;6:259-67. 53. Pingitore G, Chrobak V, Petrie J. The social and psychologic factors
32. Edinger JD, McCall WV, Marsh GR, Radtke RA, Erwin CW, of bruxism. J Prosthet Dent 1991;65:443-6.
Lininger A. Periodic limb movement variability in older DIMS 54. Pierce CJ, Chrisman K, Bennett ME, Close JM. Stress, anticipatory
patients across consecutive nights of home monitoring. Sleep stress, and psychologic measures related to sleep bruxism. J Orofac
1992;15:156-61. Pain 1995;9:51-6.
33. Bliwise DL, Keenan S, Burnburg D, et al. Inter-rater reliability for 55. Sjoholm TT, Lowe AA, Miyamoto K, Fleetham JA, Ryan CF. Sleep
scoring periodic leg movements in sleep. Sleep 1991; 14:249-51. bruxism in patients with sleep-disordered breathing. Arch Oral Biol
34. Schenck CH, Bundlie SR, Ettinger M, Mahowald MW. Chronic 2000;45:889-96.
behavioral disorders of human REM sleep: a new category of para- 56. Bader G, Kampe T, Tagdae T. Body movement during sleep in
somnia. Sleep 1986;9:293-308. subjects with long-standing bruxing behavior. Int J Prosthodont
35. Schenck CH, Bundlie SR, Patterson AL, Mahowald MW. Rapid eye 2000;13:327-33.
movement sleep bahavior disorder. A treatable parasomnia affecting 57. Sjoholm TT, Polo OJ, Alihanka JM. Sleep movements in teethgrind-
older adults. JAMA 1987;257:1786-9. ers. J Craniomandib Disord 1992;6:184-91.
36. Schenck CH, Hurwitz TD, Mahowald MW. Symposium: normal 58. Lavigne GJ, Rompre PH, Montplaisir JY. Sleep bruxism: validity
and abnormal REM sleep regulation: REM sleep behaviour disor- of clinical research diagnostic criteria in a controlled polysomno-
der: an update on a series of 96 patients and a review of the world graphic study. J Dent Res 1996;75:546-52.
literature. J Sleep Res 1993;2:224-31. 59. Velly-Miguel AM, Montplaisir J, Rompre P, Lund JP, Lavigne GJ.
37. Olson EJ, Boeve BF, Silber MH. Rapid eye movement sleep behav- Bruxism and other orofacial movements during sleep. J Cranioman-
iour disorder: demographic, clinical and laboratory ﬁndings in 93 dib Dis Fac Oral Pain 2006;6:71-81.
cases. Brain 2000;123:331-9. 60. Kato T, Montplaisir JY, Blanchet PJ, Lund JP, Lavigne GJ. Idio-
38. Bologna, Geneva, Parma, and Pisa Universities group for the study pathic myoclonus in the oromandibular region during sleep: a pos-
of REM Sleep Behaviour Disoders in Parksinson’s Disease. Interob- sible source of confusion in sleep bruxism diagnosis. Mov Disord
server reliability of ICSD-R criteria for REM sleep behaviour disor- 1999;14:865-71.
der. J Sleep Res 2003;12:255-7. 61. Reding GR, Zepelin H, Robinson JE Jr., Zimmerman SO, Smith
39. Lapierre O, Montplaisir J. Polysomnographic features of REM sleep VH. Nocturnal teeth-grinding: all-night psychophysiologic studies.
behavior disorder: development of a scoring method. Neurology J Dent Res 1968;47:786-97.
1992;42:1371-4. 62. Ware JC, Rugh JD. Destructive bruxism: sleep stage relationship.
40. Consens FB, Chervin RD, Koeppe RA, et al. Validation of a Sleep 1988;11:172-81.
polysomnographic score for REM sleep behavior disorder. Sleep 63. Okeson JP, Phillips BA, Berry DT, Cook YR, Cabelka JF. Nocturnal
2005;28:993-7. bruxing events in subjects with sleep-disordered breathing and con-
41. Sforza E, Krieger J, Petiau C. REM sleep behavior: clinical and trol subjects. J Craniomandib Disord 1991;5:258-64.
physiopathological ﬁndings. Sleep Med Rev 1997;1:57-69. 64. Okeson JP, Phillips BA, Berry DT, Cook Y, Paesani D, Galante J.
42. Eisensehr I, Lindeiner H, Jager M, Noachtar S. REM sleep behavior Nocturnal bruxing events in healthy geriatric subjects. J Oral Reha-
disorder in sleep-disordered patients with versus without Parkin- bil 1990;17:411-8.
son’s disease: is there a need for polysomnography? J Neurol Sci 65. Sjoholm TT, Piha SJ, Mantyvaara J, Lehtinen I, Lehtinen I. Rhyth-
2001;186:7-11. mic jaw movements during sleep - an autonomic reﬂex? J Sleep Res
43. Eisensehr I, Linke R, Tatsch K, et al. Increased muscle activity dur- 24, A367. 2006.
ing rapid eye movement sleep correlates with decrease of striatal 66. Ikeda T, Nishigawa K, Kondo K, Takeuchi H, Clark GT. Criteria for
presynaptic dopamine transporters, IPT and IBZM SPECT imaging the detection of sleep-associated bruxism in humans. J Orofac Pain
in subclinical and clinically manifest idiopathic REM sleep behavior 1996;10:270-82.
disorder, Parkinson’s disease and controls. Sleep 2003;26:507-12. 67. Bader GG, Kampe T, Tagdae T, Karlsson S, Blomqvist M. De-
44. Bliwise DL, He L, Ansari FP, Rye DB. Quantiﬁcation of electromyo- scriptive physiological data on a sleep bruxism population. Sleep
graphic activity during sleep; a phasic electromyographic metric. J 1997;20:982-90.
Clin Neurophysiol 2006; 23:59-67. 68. Kato T, Thie NM, Montplaisir JY, Lavigne GJ. Bruxism and orofa-
45. Micheli F, Fernandez PM, Gatto M, Asconape J, Giannaula R, Pare- cial movements during sleep. In: Attanasio A, Bailey DR, eds. Den-
ra IC. Bruxism secondary to chronic antidopaminergic drug expo- tal Clinics of North America. Philadelphia: W.B. Saunders, 2001:
sure. Clin Neuropharmacol 1993;16:315-23. 657-84.
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 166
Movements in Sleep
69. Macaluso GM, Guerra P, Di Giovanni G, Boselli M, Parrino L, Ter- 91. Hanly P, Zuberi-Khokhar NS. Periodic limb movements dur-
zano MG. Sleep bruxism is a disorder related to periodic arousals ing sleep in patients with congestive heart failure. Chest 1996;
during sleep. J Dent Res 1998;77:565-73. 109:1497-1502.
70. Sjoholm TT, Lehtinen I, Piha SJ. The effect of propranolol on sleep
bruxism: hypothetical considerations based on a case study. Clin
Auton Res 1996;6:37-40.
71. Lavigne GJ, Guitard F, Rompre PH, Montplaisir JY. Variability in
sleep bruxism activity over time. J Sleep Res 2001; 10:237-44.
72. Kato T, Thie NM, Huynh N, Miyawaki S, Lavigne GJ. Topical re-
view: sleep bruxism and the role of peripheral sensory inﬂuences. J
Orofac Pain 2003;17:191-213.
73. Kempenaers C, Bouillon E, Mendlewicz J. A rhythmic movement
disorder in REM sleep: a case report. Sleep 1994;17:274-9.
74. Chisholm T, Morehouse RL. Adult headbanging: sleep studies and
treatment. Sleep 1996;19:343-6.
75. Bastuji H. [Rhythms of falling asleep persisting in adults. Two cases
without mental deﬁciency]. Neurophysiol Clin 1994; 24:160-6.
76. Alves RS, Aloe F, Silva AB, Tavares SM. Jactatio capitis noctur-
na with persistence in adulthood. Case report. Arq Neuropsiquiatr
77. Happe S, Ludemann P, Ringelstein EB. Persistence of rhythmic
movement disorder beyond childhood: a videotape demonstration.
Mov Disord 2000;15:1296-8.
78. Stepanova I, Nevsimalova S, Hanusova J. Rhythmic movement
disorder in sleep persisting into childhood and adulthood. Sleep
79. Kohyama J, Matsukura F, Kimura K, Tachibana N. Rhythmic move-
ment disorder: polysomnographic study and summary of reported
cases. Brain Dev 2002;24:33-8.
80. Manni R, Terzaghi M, Sartori I, Veggiotti P, Parrino L. Rhythmic
movement disorder and cyclic alternating pattern during sleep: a
video-polysomnographic study in a 9-year-old boy. Mov Disord
81. Broughton R, Tolentino MA. Fragmentary pathological myoc-
lonus in NREM sleep. Electroencephalogr Clin Neurophysiol
82. Montagna P, Liguori R, Zucconi M, et al. Physiological hypnic my-
oclonus. Electroencephalogr Clin Neurophysiol 1988;70:172-6.
83. Broughton R, Tolentino MA, Krelina M. Excessive fragmentary
myoclonus in NREM sleep: a report of 38 cases. Electroencephalogr
Clin Neurophysiol 1985;61:123-33.
84. Lins O, Castonguay M, Dunham W, Nevsimalova S, Broughton R.
Excessive fragmentary myoclonus: time of night and sleep stage
distributions. Can J Neurol Sci 1993;20:142-6.
85. Vetrugno R, Plazzi G, Provini F, Liguori R, Lugaresi E, Montagna P.
Excessive fragmentary hypnic myoclonus: clinical and neurophysi-
ological ﬁndings. Sleep Med 2002;3:73-6.
86. Benz RL, Pressman MR, Hovick ET, Peterson DD. Potential novel
predictors of mortality in end-stage renal disease patients with sleep
disorders. Am J Kidney Dis 2000;35:1052-60.
87. Saletu B, Anderer P, Saletu M, Hauer C, Lindeck-Pozza L, Saletu-
Zyhlarz G. EEG mapping, psychometric, and polysomnographic
studies in restless legs syndrome (RLS) and periodic limb move-
ment disorder (PLMD) patients as compared with normal controls.
Sleep Med 2002;3(Suppl):S35-S42.
88. Picchietti DL, Underwood DJ, Farris WA, et al. Further studies on
periodic limb movement disorder and restless legs syndrome in
children ith attention-deﬁcit hyperactivity disorder. Mov Disord
89. Picchietti DL, England SJ, Walters AS, Willis K, Verrico T. Periodic
limb movement disorder and restless legs syndrome in children with
attention-deﬁcit hyperactivity disorder. J Child Neurol 1998;13:588-
90. Espinar-Sierra J, Vela-Bueno A, Luque-Otero M. Periodic leg move-
ments in sleep in essential hypertension. Psychiatry Clin Neurosci
Journal of Clinical Sleep Medicine, Vol. 3, No. 2, 2007 167
Table 2—Evidence table for PLMS reliability studies
First Evidence Study Design Pertinent Measures Subjects Signiﬁcant Findings Conclusions
Author/ Level N/age/sex
Bliwise/ 3 Reliability/ PLMS/hr 45/49-87/ 2-night diff in PLMS Pearson Large night to night variation in PLMS (10.3 ±13.2
1988/27 validity 12M33F correlation 0.68, p<0.005 mean ± SD) 4 x larger than SDB events (average
studies diff of 1.2 ± 3.7)
Bliwise/ 3 Reliability/ Interscorer reliability, arousal/hr 24/elderly/ Means range pair wise inter-scorer PLMS interscorer reliability excellent, PLMs
1991/33 validity index, Num of PLM episodes, NS reliability PLMS/hr 0.95 0.91-0.99, arousal interscorer reliability marginal for trained
studies % number movements/episode, % PLMS arousal/hr index 0.65 0.48- scorers
inter-movement interval. 0.83, Num of PLM episodes 0.83
0.71-0.95, % number movements/
episode 0.91 0.83-0.96, % inter-
movement interval 0.66 0.51-0.92.
Bonnet/ NS Reliability/ NS NS NS Deﬁnes PLM scoring criteria
1993/20 validity Requires scoring PLM in ALL stages of sleep and
studies IN WAKING
Chabli/ NS Reliability/ Movement duration, pattern of 18/NS/ Duration of movements - range 0.2 PLM in waking are variable length 0.2 – 15.6
2000/23 validity activation. 14M4F – 15.6 seconds - mean 0.55 sec. seconds. Greater variability than allowed for
studies Anterior tibialis usually identiﬁed No ﬁxed pattern of EMG recruitment PLMS.
as EMG for PLM event No set EMG recruitment pattern.
Culpepper/ 3 Reliability Repeated unblinded measures of 72/NS/ Average and across the night group Group measures stable across night and between
1992/29 PLMs on 2 nights 47M25F measures are reproducible on night-to- nights
Edinger/ 2 Reliability/ Repeated measures of PLMs and 16/61.7/ Interrater scoring agreement 83%. Acceptable interscorer agreement. Group measures
1992/32 Validity movement arousal indices on 3 6M9F Substantial night to night variation stable but substantial night-to-night intrasubject
Study consecutive nights by 2 readers within subjects despite good group variability in movement indices
Ferri/ 4 Reliability/ Leg Movements (LM): duration, 65/50.1/ Inter-movement intervals in RLS The current study only looked at Periodic Limb
2006/5 validity amplitude, area under the curve, 30M35F patients were most commonly 15- Movements in Sleep (PLMS) and not Periodic
studies sleep stage, side, interval, and 30 seconds and extending from 5 Limb Movements (PLMs) in wakefulness. The
bilaterality, inter-LM intervals. -90 seconds. This was not present analysis suggested that the minimum inter-
Periodicity deﬁned as ratio of all in controls. Periodicity measures movement interval for PLMS should be 0.5 sec,
PLM to all movements and was validated by means of a Markovian since it was part of the criteria used to analyze
validated by a Markovian analysis. analysis. the movements, and the movements that were
The differences in inter-LM chosen proved to have validity by the Markovian
intervals, LM duration, and area analysis. Another implication from the data is
under the curve between normal that the current period rule of 5-90 seconds for
controls and patients, and between PLMS is valid. The criteria chosen for amplitude
the 3 patient subgroups identiﬁed were part of the conﬁrmation of periodicity by
on the basis of their periodicity the Markovian analysis. The study also suggests
were statistically analyzed. that PLMS of amplitudes as little as 10 μV minus
Movements were analyzed that a 2 μV baseline = 8 μV should be accepted for
were 0.5-15 sec. In order to be counting on a routine clinical basis.
analyzed movement had to be
at least 10 uV tall by rectiﬁed
EMG signal. A minimum inter-
movement interval was set for
movements on the same leg of 0.5
Gschliesser/ 4 Reliability/ Periodic Limb Movements awake 24/57.5 plus Statistically more Periodic Limb Ignoring amplitude criteria may be useful in
2006/25 validity and asleep that were > 5/hr were or minus 12 Movements were seen without monitoring PLMs in treatment studies and to
studies scored twice in 24 patients—once years/NS amplitude criteria in either sleep as a investigate the periodicity of PLMs.
with and once without amplitude whole, NREM sleep, REM sleep, or
criteria. When amplitude criteria during wakefulness.
were employed PLMs had
to be 25% as tall as the EMG
Hornyak/ 3 Reliability Within subject comparison of 115/NS/NS 27% had variability >10/hour between Variability in PLMS and PLMS arousal indices are
2005/59 PLMS index and PLMS arousal nights; highest variability in RLS. higher in RLS.
index on 2 nights.
Montplaisir/ 3 Reliability/ RLS diagnostic sensitivity and 16/SEM Source criteria sensitivity speciﬁcity SIT provides a better diagnostic test for RLS than
1998/26 validity speciﬁcity used ROC curves to 2.6/8M8F means ± SD PLMS; both are reasonable but not great.
studies choose optimum cut offs. SIT PLMW >40 81 ± 19; 81 ± 19
FIT PLMW>25 69 ± 23; 56 ± 24
PSG PLMS 1 night >8 68±23; 75 ± 21
Max PLMS 2 nights>11
81±19; 81 ±19
Mosko/ 3 Reliability/ One way ANOVA: no signiﬁcant 46/NS/ 43% diagnosis on one and not the other Night-night variability- no signiﬁcant trend but
1988/28 validity effects for PLMS/hr and night 30M16F high enough to alter diagnosis for criteria of
studies of study showed no signiﬁcant PLMS/hr >5.
effect for nights. But 20 of 46
were normal (<5) on one night and
abnormal on the other.
Sforza/ 3 Reliability PLM duration and interval, PLM 28/53.4/NA Group data reliable between nights but Individual inter-night variability is independent of
2004/31 index during wakefulness and intra-individual correlation coefﬁcient age, severity, and duration of the disease.
sleep on 2 nights between nights was 0.60 for PLM index
during wakefulness and 0.54 for PLM
index during sleep.
Yokota/ 4 Observational Periodic arm movements 2/41-81-2M NS Periodic arm movements occur in sleep and
1995/24 outcome wakefulness. Muscle contraction was tonic and
studies lasted 0.2-5 sec, sometimes preceded by muscle
twitch. More often only myoclonic muscle twitches
were recorded. PLMS were also seen;they usually
were independent and at different frequencies.
K complexes were seen in association with the
movements. The authors concluded that since both
patients had high spinal cord lesions that the PLMs
were a result of disinhibition.
Zucconi 5 Observational NS NS NS Expert consensus for criteria for periodic limb
/2006/4 studies movements in sleep and wakefulness. Criteria were
developed based on a separate publication where
PLMS were subjected to Markovian analysis.
Table 5—Evidence table for REM behavior disorder reliability studies
First Evidence Study Design Pertinent Measures Subjects N/ Signiﬁcant Findings Conclusions
Author/ Level age/sex
Bologna / NS NS NS NS NS NS
Consens/ 2 Reliability/ Clinical impression and a clinical 23/63 yrs/ The 2 nights of PSG were very similar in This method of scoring PSG for RBD is
2005/40 validity rating of RBD severity ﬁlled out 11M12F amount of tonicand phasic activity. PSG RBD quantitative and shows good correlation
studies by the bed partner were compared score showed an association with the clinical with the clinical impression of RBD and
to a polysomnographic method of impression (p < 0.023).If patients with PSG with a rating scale for the severity of RBD
scoring RBD. RBD score of < 10% are eliminated from the ﬁlled out by bed partner.
analysis (as suggested by the Receiver Operator
Curve) the PSG score compared to the clinical
impression gold standard has a sensitivity of
89% and a speciﬁcity of 57%.
PSG RBD score showed an association with
the RBD symptom score (Spearman rho = 0.42,
p =.048). Clinical impression and the RBD
symptom score highly correlated (p <0.0001).
Eisensehr/ 2 Reliability/ Percent of patients with PSG 19 PD 273 no Negative predictive value: sensitivity of clinical Clinical interview is excellent in
2001/42 validity evidence of RBD who had clinically PD/SD 9 PD/ interview was 33% in PD patients and 100% in predicting PSG evidence of RBD in non-
studies positive interview for RBD. M13 PD 168 non PD. Speciﬁcity of clinical interview was Parkinson patients but poor in Parkinson
no PD, F6 PD 90% in PD patients and 99.6% in non PD patients. Patients with suspected RBD
105 no PD with PD should undergo PSG to be sure
of diagnosis. Interviews are sufﬁcient for
patients without PD.
Lapiere/ 3 Reliability/ The authors determined the 5/44-65/3M2F NS RBD patients had more tonic REM
1992/39 validity percentage of time spent in tonic than controls. Phasic EMG activity was
studies versus phasic REM for RBD increased in RBD patients compared to
patients versus controls. They controls, but there was no increase in
compared RBD patients and controls phasic eye movement activity in patients
for two types of phasic activity compared to controls. RBD patients
present: (a) The percentage of 2 had Periodic Limb Movements (PLMS)
sec mini-epochs of sleep in REM equally distributed in both REM and
containing EMG bursts 100 msec NREM sleep, compared with controls
– 5sec. The phasic EMG bursts with only PLMS mostly in NREM sleep.
had to be 4 x the amplitude of the
background EMG; and (b) the
percentage of 2-sec mini-epochs of
sleep in REM containing rapid eye
Olson / 4 Observational Study described the clinical features 93/32- 90 or 93 patients had increased tonic or The clinical features of RBD and seen
2000/37 outcome in association with the PSG features 85/81M12F phasic activity in REM sleep. 45% of patients together with the polysomnographic
studies of RBD had abnormal behavior in REM sleep in the features of RBD.
laboratory on the night of the sleep study. 57%
of the patients had associated neurological
Schenck/ 4 Observational Association of dream enacting 5/60-72/4M1F The constellation of features listed in outcome Dreams corresponding in content to
1986/34 outcome behaviors with dreams and increase measures occurred together dream enacting behaviors, loss of chin
studies in chin tone in REM and limb atonia in REM, and increased limb
twitching in REM. twitching occurred together. There was
also an increase in the density of rapid
eye movements in REM. Three men had
PLMS. Neurologic disease was seen in 4
cases. Delta sleep was increased for age
in all subjects.
Schenck/ 4 Observational Association of dream content with 10/62/9M1F Nine of 10 cases were in males. There was an RBD is a condition primarily of older
1987/35 outcome dream enacting behaviors, older intermittent increase in EMG tone in REM males, often associated with neurological
studies males, increased REM phasic sleep. Sleep related injury was present in 9/10 disease and characterized by potentially
activity, neurological disease, cases. 8 cases had dreams involving motor injurious dream-enacting behavior. The
PLMS, sleep related injury. overactivity and violent confrontations of dream EMG during the PSG in REM sleep
characters. In 5 cases the RBD was related to shows intermittent interruption of the
neurological disease. For all 10 cases there atonia of REM sleep that is associated
was high REM density, increased numbers of with dream enacting behaviors. RBD is
periodic limb movements in sleep (PLMS) exquisitely sensitive to clonazepam.
and aperiodic twitching movements in NREM
sleep, clonazepam showed improvement in
dream enacting behaviors in 7 patients and
desimpramine in one patient.
Schenck / 4 Observational Age at onset, sex distribution, 96/9- RBD is more prominent in older males. There RBD is characterized by complex motor
1993/36 outcome association with neurological 83/84M12F is often a prodrome of several years of sleep activity arising out of REM sleep. Patients
studies disease, dream enacting behavior talking and aperiodic limb twitching prior to the may act out their dreams and injure
in association EMG increase in onset of RBD. RBD is characterized by complex themselves. Clonazepam is the treatment
tone, PLMS, aperiodic twitching in behavior arising out of REM sleep when patients of choice
REM sleep, sleep architecture, sleep act out their dreams. There is loss of REM atonia
related injury, prodrome of sleep or intermittent increases in EMG tone in REM
talking and less complex motor sleep associated with abnormal behaviors. There
activity is often an increase in percentage of stage 3 and
4 sleep for age. Sleep related injury is common.
Patients with RBD more frequently have PLMS
as well as aperiodic twitching of the limbs
in NREM sleep. The condition is exquisitely
responsive to clonazepam. There are often
associated neurological disorders
Schenck/ 4 NS NS NS NS NS
Sforza/ 3 Observational NS 13 no neuro NS More patients had neurological disease.
1997/41 Outcome disease 39 One third of patients with neurological
studies neuro disease/ disease had symptoms of RBD before
Sd 2.1 no onset of neurological disease. Sleep
neuro disease disturbance was the same in both groups
Sd 1.2 neuro Motor activity was more varied in the
disease/NS symptomatic group. More patients in the
idiopathic group had complete absence of
atonia during REM periods. Preservation
of atonia was more frequent in patients
having simple attacks
Table 7—Evidence table for sleep bruxism reliability studies
First Author/Yr/Ref Evidence Level Study Design Pertinent Subjects N/age/ Signiﬁcant Findings Conclusions
Bader/2000/56 3 Case control study NA NA NA Video – PSG including masseter EMG and body
movements in 11 bruxers (4M, 7W; 23-63 yrs) and 8
controls revealed more body movements in bruxers
without any relationship with masseter EMG activity,
suggesting a central etiology common to both.
Bader/1997/67 4 Observational studies NA NA NA Video-PSG including multiple cranial muscle EMGs
in 24 selected bruxers (9M, 15W, ages 23-67) showed
a mean of 24 bruxism episodes, one hr of sleep,
mostly in stage 2 NREM, preceded by alpha activity
in the EEG and tachycardia at onset of episodes.
Broughton/1988/9 5 Case series NS 41/NS/NS NS Describes potentials as occurring being brief (<200
ms) phasic potentials of up to 250 μV occurring at
sleep onset and in NREM & REM; involve different
body regions asymmetrically and asynchronously;
resembles REM myoclonus but is not clustered;
usually no movement or very slight or EEG impact
(i.e. no arousal or K complex for most)
Ikeda/ 5 Clinical series EMG & out rate N=9 EMG at 3 MVC Technical study suggesting minimal criteria for
1996/66 changes/ 5 M, 4 F levels: 3, 10 and 20% ambulatory recordings of EMGs. EMG change >3% =
ambulatory Mean age: 26.8 threshold interval between events >5 seconds.
Kato/2001/68 3 Observational studies 10/20-41/6M4F NA Majority of RMMA in stage 2 NREM sleep were
preceded by autonomic (increased HR – one cycle
before EMG onset ) and cortical (changed alpha-
delta power – 4 seconds) activation. SB patients had
more RMMA during sleep (second night in lab) than
Lavigne/1996/58 2 Reliability/validity NS 18/NA/9M9F A series of cutoffs Bruxers have increased numbers of bruxing episodes
studies for different bruxing per night & per hour of sleep; also greater number
parameters had a of bursts per episode. No difference in sleep stages
sensitivity of at least between groups; most episodes and greatest difference
72% and speciﬁcity of occurred in stages 1 and 2 sleep with most episodes
at least 94% including associated with body movements. Speciﬁc PSG
– cutoffs could be produced that would discriminate
More than 30 bruxers from controls.
bruxing episodes per
More than 4
bruxing episodes per
More than 6
bruxism bursts per
More than 1
bruxing episode with
A combination of
predicted 83.3% of
bruxers and 81.3% of
Lavigne/2001/71 5 Reliability/validity NS 9/NA/5M4F NS The clinical deﬁnition of SB did not change over
the span of repeated studies. There was signiﬁcant
variability in speciﬁc PSG measures, studied by
coefﬁcient of variation: SB episodes per hour
(25.3%), SB bursts per hour (30.4%), and number
of episodes with noise (53.5%). However, 36 of 37
nights had measures higher than proposed thresholds
for episodes per hour and episodes with noise; on 34
nights, the results were higher than 3 criteria (plus
SB bursts per hour), and in 37 nights the results were
higher than at least one criteria.
Lavigne/1997/51 5 Case series NS 13/NS/8M5F NS Those with pain – similar at bedtime and awakening
– had fewer episodes of bruxism per hour, but equal
number of bursts per episode and amplitude.
Okeson/1990/64 5 Clinical series NS 39/60- NS Overall older patients do present 3.03 episodes of
87/19M11F masseter activity hr/sleep; those with dentures had 2.6
times less events; each events lasted 5.95 sec/mean.
Interestingly, 20/30 slept on their back.
Okeson/1991/63 5 Cohort study NS 0/57/NS NS Most SB events occurs in supine position, in light
sleep and in relation to sleep arousals (97 and 100%).
Table 9—Evidence table for rhythmic movement disorder
Author/Yr/Ref Evidence Study Design Pertinent Subjects N/age/sex Signiﬁcant Conclusions
Level Measures Findings
Alves/1998/76 4 Observational NS 1/19/1M NS Single case report in an adult with past history of RMD since
infancy. Overnight PSG shows head banging movements at
4/min rate during wake and NREM stages 1 and 2.
Bastuji/1994/75 4 Observational NS 2/23&24/1M1F NS Two young adults with head banging since the ﬁrst year of life
showed polysomnographic rhythmic head movements at 0.5-1
HZ lasting for 6-16 min. These movements occurred after a
period of REM sleep.
Chisholm/1996/74 4 Observational NS 2/18&24/2M NS Two young adults with history of head banging since childhood
were reported. Overnight PSG study in one patient documented
head banging during all stages of NREM sleep occupying 14%
of the epochs.
Dyken/1997/15 4 Observational NS 7/NA/1-12 NS The authors captured 37 episodes during overnight video-PSG
recording of rhythmic 0.5-2hz head banging, body rocking,
and leg banging movements lasting for 1-59 minutes during
stage 2 NREM sleep in 7 children referred for violent nocturnal
Happe/2000/77 4 Observational NS 3/15,18,&59/2M1F NS Adults with history of head banging and body rolling since
childhood had these episodes during overnight video-PSG
recording in NREM stages 1and 2; in one case these occurred
during wakefulness and PSG sleep drowsiness.
Kempenaers/1994/73 4 Observational NS 1/34/1F NS Video-PSG in a 34-year-old woman with a lifelong history of
body rocking movements during sleep documented rhythmic
head and body rolling movements preferentially during REM
sleep at a frequency of 0.5/ sec.
Movements also occurred during stages 1 and 2
Kohyama/2002/79 4 Observational NS 2/9&13/1M1F NS Video-PSG recording in 2 adolescents documented rhythmic
EMG burst at a frequency of 1-1.6 per second and a duration of
2-93 seconds during NREM stages 1 & 2 and REM sleep, with
occasional bursts in slow wave sleep.
Manni/2004/80 4 Observational NA 1/9/1M NS Video-PSG in a 9 yr old boy with rhythmic movement disorder
documented body rocking and rolling and head banging
movements using all stages of NREM sleep. CAP analysis
showed a close association with A-phase of CAP, mostly A2 an
Table 11—Evidence levels for excessive fragmentary myoclonus
First Author/Yr/Ref Evidence Study Design Pertinent Subjects N/age/ Signiﬁcant Conclusions
Level Measures sex Findings
Lins/1993/84 3 Observational Studies NS 19/34-67/19M NS The investigators found that those with EFM had a higher index
than controls. Highest index was in REM, lowest in SWS, no wake
studied. Controls had index around 1 (i.e. one 3 second bin per
epoch with qualifying potential), EFM group around 4 with increase
over the night due to decrease in SWS.
Montagna/1988/82 5 Observational Studies NS 7/19 - 27/7M NS Describe fragmentary myoclonus as fasciculation type potential (no
quantitation) that appears sporadically in different muscles, present
in wake and all sleep stages at no great difference, different muscles
varied with sleep stages with soleus most in wake and tibialis
anterior/triangularis labii higher in stage 1 and REM. Approximate
frequency 4/hr (highest 1 and REM).
Vetrugno/2002/85 NS Observational Studies NS 2/52 - 71/2M NS Potentials present in all sleep stages and wake, increase over
course of the night Actual movements associated with longer (>150
msec) or repetitive potentials or with higher amplitude (>200 μV)
potentials. Backaveraging showed no cortical potential.
Table 13—Evidence tables for hypnagogic foot tremor/alternating leg muscle activation
First Author/Yr/Ref Evidence Study Design Pertinent Subjects N/ Signiﬁcant Conclusions
Level Measures age/sex Findings
Broughton /1984/81 5 Observational NS 1/42/1M NS Asynchronous potentials occurring asynchronously in 2 legs, arms, face
with EMG potentials of more than 50 μV, up to 200 μV, lasting <150
millisecond, as frequent as every 2 seconds, can be associated with
small limb movements Noted potentials rare in wakefulness, present in
drowsiness, equally present in all stages of NREM, different pattern in
REM (associated in clusters with REMs).
Chervin/2003/10 5 Observational studies NS 16/12 - NS There exists a set of tremor-like leg movements at 1 to 2 Hz with
70/12M4F alternation between legs (sometimes single leg), full range 0.5 to 3 Hz,
predominantly during stages 1 and 2, but also REM, 1 in stage 3 (text says
4, but table shows no subject with 4), no mention of wake (?studied), train
of ALMA lasted 1.4 to 22.2 seconds, often occurred with arousals, usually
following, trains could be periodic, burst durations of 100 to 500 msec.
12 of 16 taking antidepressants, 10 with PLMI >5/hr, some repeat studies
showed most had same phenomena.
Wichniak/2001/8 5 Observational studies NS 28/14 - NS Found 7.5% of those getting PSG for sleep complaint had at least one
72/21M 7F 10-second train of periodic anterior tibialis burst at typical rate of 1 to 2
Hz (0.3 to 3.3), with bursts typically 300 to 700 msec (75 to 2000 msec),
lasting typically 10 to 15 seconds (up to 2 minutes). Trains were most often
seen during wake before sleep, but also during stages 1 and 2 and often
associated with arousals. Bursts were bilateral, usually asynchronous. Video
showed oscillating movements of feet and toes. Authors suggest this is a
form of rhythmic movement disorder.
Key For Evidence Tables: ALMA=alternating leg muscle activation, EFM=excessive fragmentary myoclonus, EMG=electromyogram, HF=hypnagogic foot tremor, HR=heart rate, NS=not stated,
PD=Parkinson disease, PLMS=periodic leg movements of sleep, RBD=REM sleep behavior disorder, RMD=rhythmic movement disorder, RMM= rhythmic masticatory muscle activity, SB=sleep
bruxism, SD=standard deviation, WASO=wake after sleep onset