Rapid eye movement sleep behaviour disorder in children and adolescents


  • Gregory Stores MA MD DPM FRCPsych FRCP

    Corresponding author
      * Correspondence to author at North Gate House, 55 High Street, Dorchester on Thames, Oxfordshire OX10 7HN, UK.
      E-mail: gregory.stores@psych.ox.ac.uk
    Search for more papers by this author

  • See end of paper for list of abbreviations.

* Correspondence to author at North Gate House, 55 High Street, Dorchester on Thames, Oxfordshire OX10 7HN, UK.
E-mail: gregory.stores@psych.ox.ac.uk


Rapid eye movement (REM) sleep behaviour disorder (RBD) is a recently described parasomnia with important implications for diagnosis, treatment, and prognosis. Once thought to occur only in elderly males, it has now been reported in other groups including children and adolescents. The cardinal feature of RBD is the loss of REM atonia, which allows dreams to be acted out. Depending on the dream content, the accompanying behaviour can be dramatic, sometimes causing self-injury or harm to others. In adults, increasing numbers of associated conditions have been reported, notably neurodegenerative disorders of which RBD can be a prodrome. Other associations include narcolepsy, epilepsy, and antidepressant medications. Awareness that RBD can occur in young patients (admittedly rarely, it seems) is important. It is essential to distinguish between RBD and other dramatic parasomnias because each has different implications and treatment requirements. Children and adolescents displaying both clinical and polysomnographic features of RBD, and others with only polysomnographic evidence of loss of REM atonia, have been reported in association with various neurological and neurodevelopmental disorders. Reports of these cases are reviewed, together with suggestions for further reporting and research.

Review of the parasomnias reveals the many forms of disturbed behaviour and strange experiences related to sleep that are now described in children.1 The 2005 revision of the International Classification of Sleep Disorders2 lists more than 30 parasomnias, including some that were once thought to occur only in adults but have now been described at an earlier age. Rapid eye movement (REM) sleep behaviour disorder (RBD) is a case in point.

Initially, RBD was thought to occur exclusively in elderly men. However, more recent reports have demonstrated its existence (or something akin to it) at all ages and also in women, although elderly men still predominate in reported series.

Although seemingly rare at an early age, this condition is of special clinical importance for paediatricians and child psychiatrists. It is essential that it be distinguished from other dramatic parasomnias because of its possible underlying or associated disorders and its relatively specific treatment requirements.

Before discussing RBD in young people, this review outlines the general features of the condition as described in recent publications. Although these accounts have been predominantly concerned with adults, the main points arising convey the nature of RBD and are generally relevant to its consideration in children and adolescents, including the cases described later. The emphasis in this review is on clinical aspects. The complex and often speculative pathophysiological issues involved have been discussed in detail elsewhere.3

Overview of RBD

Having been foremost in bringing RBD in humans to the attention of clinicians, in 2002 Schenck and Mahowald4 provided a comprehensive general review of the condition. Since then, further accounts have been published5 that nevertheless leave some fundamental questions about its nature at least partly unanswered. The following is a summary of the main points made in these publications.

Normally, skeletal musculature is effectively paralysed in REM sleep (REM atonia). The essential feature of RBD is dream enactment made possible by the loss of REM atonia. Because most dreaming occurs in REM sleep, it is not normally possible for dreams to be enacted but, with pathological preservation of muscle tone, this can happen. If the dream content is violent, the patient punches, kicks, leaps or runs about, or displays apparently purposeful violent behaviour in keeping with the content of the dream. This often causes self-injury or injury to anyone nearby, especially the bed partner. Typically, such behaviour is out of character.

Prevalence has been estimated as 0.38% to 0.5%2 but it is not really known because of uncertainties regarding its recognition.

Aetiological factors

An increasing number of conditions associated with RBD are being reported (symptomatic RBD).6 The chronic form is strongly associated with neurodegenerative disorders of the synucleinopathy type (in particular, Lewy body dementia, multiple system atrophy and Parkinson’s disease) but also with narcolepsy. Additional associations include a variety of other neurological conditions including epilepsy.7

An acute form of RBD has been described after withdrawal of alcohol or various sedative-hypnotic drugs, and also with intoxication with certain medications, notably some antidepressants, as well as with caffeine intake.

RBD as a prodrome

RBD can be an early sign of a neurodegenerative condition which in time declares itself, sometimes after many years. In addition, there can be a prodromal period before the development of the classical manifestations of RBD, characterized by persistent sleep talking, loud vocalizations, limb twitching, or gross limb and body jerking.

Idiopathic or cryptogenic RBD?

Because of its strong association with organic factors, RBD is increasingly viewed as a secondary parasomnia rather than a primary, idiopathic sleep disorder. The term ‘cryptogenic’ is increasingly preferred to ‘idiopathic’ on the grounds that, in the fullness of time, the condition underlying the sleep disorder might well come to light.

Associated sleep disorders

Narcolepsy has already been mentioned as an associated disorder of RBD. Some patients with RBD have been reported to have coexisting sleepwalking or sleep terror episodes (‘parasomnia overlap disorder’), which in some cases seemed to have started in childhood.8 RBD is one of the sleep disorders that may be accompanied by periodic limb movements in sleep.

The possible pathophysiology of RBD has recently been discussed by Boeve and colleagues on the basis of findings from animal experimentation and human studies.3 They freely admit that much remains to be clarified about the structures, networks, and neurochemical systems involved. Brainstem nuclei that are particularly implicated include the locus coeruleus, the pedunculopontine nucleus, and the laterodorsal tegmental nucleus, but other centres at higher and lower levels are also represented in their provisional model.

The mechanisms producing RBD in the wide range of pathological conditions mentioned above (from direct structural interference with the nuclei involved in the generation of REM sleep to complex functional disturbance in other associated conditions) must be diverse. It is hoped that clarification of the different mechanisms underlying RBD will also be shed light on the nature of some of the associated conditions including neurodegenerative diseases.9


The criteria for the diagnosis of RBD that are given in the 2005 revision of the International Classification of Sleep Disorders2 are essentially a combination of the clinical characteristics described earlier and polysomnographic (PSG) evidence of REM sleep without atonia. Recall of dream content corresponding to the abnormal REM sleep-related behaviour is useful additional evidence of RBD. There should be no epileptiform activity during REM sleep unless RBD can be clearly distinguished from any concurrent REM sleep-related seizure disorder. In addition, the sleep disturbance should not be better explained by another sleep disorder, medical or psychiatric condition, medication, or substance misuse.

Dream enactment alone is not evidence of RBD, because similar behaviour has been reported in obstructive sleep apnoea10 and also in sleepwalking and sleep terrors, post-traumatic stress disorder, and as an effect of alcohol or other substance use or withdrawal.5

‘Subclinical RBD’ is described, by which is meant the electrophysiological finding of loss or atonia during REM sleep but without evidence of dream enactment.3 It remains an open question how often such a finding proves to be the precursor to the eventual development of clinical RBD.

Differential diagnosis

Other (and, in some instances, much more common) conditions that might be confused with RBD because, although they have their own distinctive features, they also can involve dramatic behaviour1, include the agitated form of sleepwalking, sleep terrors, confusional arousals, nightmares, the dramatic awakenings that may occur in obstructive sleep apnoea, nocturnal frontal lobe epilepsy and certain other sleep-related seizures, nocturnal panic attacks, sleep-related psychogenic states, and also ‘pseudoparasomnias’ in which physical events are simulated.

RBD and these other sleep disorders may be misdiagnosed as quite different conditions.11 Of course, different parasomnias can coexist.


Even in the presence of a neurodegenerative disorder, RBD is usually very treatable. Clonazepam is the usual initial treatment and is effective in most cases. Alternatives include melatonin and pramipexole. All concerned need to be encouraged to view the disturbed behaviour in RBD as part of an illness rather than a personality problem.

RBD in children and adolescents

Published accounts of RBD have dwelt almost exclusively on adults with this disorder. However, the admittedly limited literature about younger patients serves to emphasize the importance of knowing that the condition can occur in children and adolescents.

It also raises further challenges about its recognition, underlying pathology and pathophysiology, and also appropriate treatment strategies in young patients, although much of what has already been outlined previously mainly about adults with RBD applies (or might well apply) at an earlier age.

That said, the published accounts of RBD in children and adolescents require close scrutiny. The patients described consist of a miscellany of ages and associated conditions. In addition, they are best separated into those reported to have both clinical and PSG features of RBD (the ‘clinical’ group) and others described as having ‘subclinical’ RBD in the sense that, at the time of assessment, they displayed only the PSG features of loss of REM atonia.

Inevitably, assessing the clinical aspects (especially subjective reports of dream content) can pose problems in children with impaired ability to communicate because they are young or limited in other ways. In such circumstances reliance has to be placed on inferences from the child’s behaviour at the time of the dreaming episode.

Treatment is either not mentioned in some reports or given at unspecified dosages; neither is the patient’s sex always stated. Considering that some associated disorders need treatment in their own right, further reports would be improved by a discussion of wider aspects of care.

As implied earlier, it cannot be assumed that the subclinical group will necessarily develop clinical RBD eventually. One of the drawbacks of the present literature is the very little long-term follow-up information it contains. This also makes it difficult to know how many of the young ‘idiopathic’ cases eventually declared some underlying definitive disorder, including the types described in many adults with RBD.

Clinical RBD

The first report of this type concerned an 8-year-old female with an infiltrating tumour of the pons.13 In addition to cerebellar and other neurological signs, she exhibited agitated periods at night and mouth movements, as well as sleep talking and laughter accompanied by bursts of REM sleep. PSG findings demonstrated loss of REM atonia.

A few years later, a 10-year-old female was described who developed severe sleep disturbance after the removal of a midline cerebellar astrocytoma.14 Her complex sleep disturbance included excessive restlessness (including smooth and jerky movements, complicated stereotyped behaviours such as hand waving, reaching out or fumbling movements and body turning), falling out of bed (often during a dream), and an increase in vivid dreaming. The movements tended to begin about 2 hours after onset of sleep, becoming more pronounced afterwards. PSG findings showed intermittent loss of REM atonia accompanying some of these behaviours and frequent aperiodic limb movements in both REM and non-REM sleep. Interestingly, the female’s healthy 8-year-old brother showed similar aperiodic movements during both types of sleep but no comparable clinical sleep disturbance.

Shortly afterwards, a brief report was published about an 18-month-old male with hereditary quivering chin syndrome or hereditary essential chin myoclonus (a dominantly inherited condition, thought to involve dysfunction of the pontine reticular formation), frequent tongue-biting during sleep, prolonged screaming, crying, and stiffness at night without alteration of consciousness, and periods of restlessness and ‘disorganized behaviour’, followed by awakenings.15 Prolonged electroencephalographic (EEG) monitoring revealed no evidence of seizure activity, but PSG findings demonstrated a loss of REM atonia associated with vigorous body movements. Clonazepam is said to have improved the child’s sleep disturbance but not the tongue-biting.

The same group then described yet another case of this rare condition of hereditary quivering chin syndrome.16 This male, almost 2 years old, had severely lacerated his tongue repeatedly but was also described by his parents as having ‘violent jerking, agitated crawling, and loud vocalizations’ during sleep. PSG recordings showed a loss of REM atonia. Complex behaviour, including crawling and vocalizations, occurred during REM sleep. Again, the recordings provided no evidence of epilepsy. In this case, 0.5mg of clonazepam suppressed both the abnormal behaviour and the tongue-biting.

The brainstem location of the lesions (or the putative location in the last two cases) is in keeping with the site of the nuclei particularly implicated in the pathophysiology of RBD.3 Different mechanisms (as yet ill-defined) must be involved in the remaining cases.

A 16-year-old male was diagnosed as having narcolepsy (primarily a disorder of REM sleep) in view of daytime hypersomnolence and multiple sleep latency test results characteristic of this condition.17 Nocturnal PSG findings included excessive electromyographic (EMG) twitching during REM sleep. During one REM period during the multiple sleep latency test, the patient threw his arms up in front of his face and, on waking up, said that he had been having an ‘action packed dream’. The limited EEG coverage during the multiple sleep latency tests showed no evidence of epileptic activity. The combined clinical features and PSG findings were considered to be compatible with RBD.

In a further report about narcolepsy, of 10 patients meeting strict criteria for this condition and also for RBD, three were aged 12, 17, and 19 years.18 All three were males. Chance association between the two disorders was considered statistically unlikely. The report does not make a clear distinction between the adults and other patients in the series. General reference is made to the RBD behaviours including running, wrestling or clawing movements in sleep (sometimes causing injury), kicking, elbowing, arm waving, or talking and arguing during sleep. These and other behaviours were thought to represent dream enactments. It was felt that treatment of the cataplexy component of the narcolepsy syndrome with clomipramine could have induced or promoted RBD in three patients. Treatment of the RBD was not reported.

A recent report19 described two females aged 7 and 9 years in whom narcolepsy of recent onset was diagnosed on the basis of detailed assessment, apparently accompanied by RBD. In other respects, both children were neurologically normal. The two conditions seem to have developed at about the same time. Long-term follow-up information was not available at the time of reporting.

Nightmares and restless sleep with frequent nocturnal awakenings were described in the 7-year-old female, together with PSG features of loss of REM atonia, sometimes accompanied by sleep talking and ‘complex movements’. In addition, arousals from deep non-REM sleep with disorientation were observed. A diagnosis of parasomnia overlap disorder was made and treatment with modafinil (150mg) and clonazepam (0.5mg in the evening) was introduced, the effect of which was not stated.

The 9-year-old female was reported to have had restless nocturnal sleep with ‘harmful’ behaviour, aggressively attacking her sister and kicking and striking the wall, together with sleep talking. Again, PSG findings compatible with RBD were found, associated with ‘gross extremity movements’. At the time of reporting, her narcolepsy symptoms were partly controlled with a combination of modafinil (200mg) and sodium oxybate (2g), and the RBD was being treated with an evening dose of clonazepam (0.5mg).

RBD has been linked with juvenile Parkinson’s disease.20 A 16-year-old female began to show signs of idiopathic Parkinson’s disease and also excessive daytime sleepiness, which on investigation was considered to be somehow related to her parkinsonism rather than other causes such as narcolepsy. Regularly, her behaviour at night was disturbed. She would talk in her sleep and flail her limbs as if enacting a dream, in the process frequently striking nearby objects and injuring herself. On waking she described vivid movement-laden dreams. PSG findings demonstrated EMG findings compatible with RBD. To a smaller extent, the same type of EMG findings were seen in the patient’s healthy monozygotic twin.

A particularly intriguing report described RBD in children with autism21– nowadays clearly established as a neurodevelopmental disorder. Eleven children with autism diagnosed by the DSM-IV22 were investigated for disrupted sleep with frequent awakenings. With a certain amount of difficulty because of the children’s mental state, limited PSG recordings were obtained. Five of the children (overall ages 3–9y, and including three males) were reported to show both clinical and PSG signs of RBD.

Their nighttime disturbances included repeated waking up in very distressed, ‘dream-like’ or agitated and violent states. One child was said to move, twitch, jerk her limbs, look round and sit up and cry for several minutes without appearing fully awake before falling asleep again. Another wandered about the house still asleep. There is no mention of seizure activity in any of the recordings. The other six children out of the 11 showed none of the PSG features of RBD.

Four of the five children in whom RBD was diagnosed (or thought probable) began to sleep well after treatment with 0.25 to 0.5mg clonazepam. This treatment made the remaining child agitated and restless, but his sleep problems resolved with carbamazepine.

There seemed to be no relationship between the presence or absence of RBD in this series and whether or not the children’s development had regressed. However, as various medical conditions can underlie autism, long-term follow-up might reveal the presence of such a condition, especially in those with PSG signs of RBD, on the grounds that such signs indicate some type of pathophysiology.

An account of five cases of childhood RBD23 included one child with ‘documented narcolepsy’ (details not provided). This series of children (ages and sexes not specified) were all investigated because of unusual sleep-related violent or injurious behaviours or vigorous body movements (including running in circles as if trying to escape, screaming, or thrashing about in bed) associated with frightening dream reports. PSG recordings revealed excessive muscle tone, frequent gross body movements, and increased phasic EMG activity during REM sleep. Clonazepam (0.25mg at bedtime) led to a complete resolution of the nighttime disturbances.

The authors of that report stopped short of saying that the children’s condition was the same as adult RBD, but the correspondence between the two seems to be close. Apart from the case associated with narcolepsy, their condition could be categorized as idiopathic, subject to revision in the light of possible further developments.

The same can be said of another patient who first began to act out his dreams of riding a motorbike at the age of 14 years (Stores G, unpublished data). In subsequent years his frequent dream enactment took an increasingly violent form, in keeping with the content of his dreams. However, he did not come to medical attention until the age of 30 years when he not only injured himself when dreaming that he was fighting off a bedroom intruder but also hurt his girlfriend while protecting her from a similar threat. An additional concern was that he might harm their new baby with whom they shared their bedroom.

Although wayward earlier in life, he had been well-adjusted, responsible, and otherwise healthy in recent times. Investigations included PSG recordings which showed the characteristic features of RBD together with periodic limb movements. Clonazepam (0.25mg at bedtime) stopped the nighttime episodes promptly without any ill effects. Five years later, and 21 years after the apparent onset of his RBD, he remained physically and psychiatrically well.

‘Subclinical’ RBD

Loss of REM atonia, unaccompanied by convincing clinical signs of RBD, have been reported in young patients in association with various conditions. Uncertainty about the nature of several of these cases arises from insufficiently detailed reporting and lack of information about subsequent developments.

There is an outline account of 19 patients who met PSG, but not clinical, criteria for RBD.24 The pathological categories in which they were placed included non-REM parasomnias, narcolepsy, and olivopontocerebellar degeneration, in each of which some children and adolescents were mentioned but their details were not provided. The other categories that contained only adults were obstructive sleep apnoea and fluoxetine treatment for major depression.

According to another report, an otherwise healthy female began to have daytime sleep attacks at the age of almost 5 years followed a few months later by other problems including frequent nightmares and restless sleep.25 PSG findings are said to have revealed ‘REM-related motor disturbances at night’. Later she developed classical cataplexy, and narcolepsy was diagnosed. Because of the lack of detail in the descriptions of both the clinical features and the PSG findings in this case, at best it can also be placed in the subclinical RBD category.

Similar diagnostic difficulties are encountered in a preliminary report of a series of 16 patients with Tourette syndrome, age range 6 to 65 years, said to have RBD.26 Again, insufficient information is provided in a report of a 4-year-old male and a 5-year-old female with posttraumatic stress disorder, contentiously thought by some to be basically a disorder of REM sleep,27 and REM sleep motor abnormalities.28

More detailed accounts, focusing on early developmental aspects of brainstem neurophysiology rather than claims to have demonstrated RBD features, have described REM abnormalities of an RBD type in children with West syndrome, associated with many forms of neurological disorders,29 and in others with group A xeroderma pigmentosum, a serious skin condition that is, however, also characterized by cortical and subcortical neuronal degeneration.30


From these reports, it seems that (rarely) RBD occurs in children and adolescents with similarities to the condition as described in adults. However, its frequency may be underestimated because of limited awareness of its existence and owing to the fact that many of its clinical features overlap with other dramatic parasomnias, leading to its misdiagnosis. It is also possible that (as in some women with the condition) there are less clinically dramatic forms of childhood RBD that do not come to medical attention.

It is important to distinguish RBD from other sleep disorders (including by means of PSG monitoring) because it has very different significance, including aetio-pathological implications, and because of the relatively specific treatment required that, on present evidence, seems likely to be highly effective as it is in adults. However, response to the main therapeutic agent, clonazepam, does not in itself have diagnostic force because of its non-specific efficacy in various disorders.

In view of the established prodromal implications of clinical RBD in adults, long-term follow-up is necessary in children and adolescents with this condition in the hope of identifying and treating possible underlying disorders at an early stage.

Similarly, cases of subclinical RBD need to be kept under close scrutiny to help the interpretation of this finding early in development, in particular, whether clinical RBD eventually develops with all the implications of its doing so.

RBD in young people is a fertile subject for further research aimed at clarifying its occurrence, pathophysiology, and clinical and predictive significance. Research into the management of this fascinating sleep disorder is also needed.

List of abbreviations



REM sleep behaviour disorder


Rapid eye movement