Dreaming: a psychiatric view and insights from the …...gested that REM sleep, and associated dreaming, may perform vital functions. Studies on the child-hood development of dreaming

Summary

Eiser AS,Schenck CH.Dreaming:a psychiatric viewand insights from the study of parasomnias.SchweizArch Neurol Psychiatr 2005;156:440–70.

This paper provides a psychiatric view of the psy-chological, experimental, and medical literature on dreams, along with insights gained from thestudy of parasomnias.The psychoanalytic theory ofdreams developed by Freud considers dreams to behighly meaningful mental products, with uncon-scious processes playing a major role. However, ithas been quite difficult to devise operational crite-ria to rigorously test this theory.With the discoveryof REM sleep and the NREM/REM sleep cycle,sleep laboratory techniques facilitated the system-atic sampling of dreams. Studies carried out toexamine dreams focused on their individual con-text were hampered by small sample sizes. Studiesof dreams from larger samples of subjects involvedin a particular and often disturbing life circum-stance posed various challenges in interpretation.Experimental manipulations of REM sleep andassociated dreaming have been carried out, withvarious theoretical and methodological short-comings. REM sleep deprivation studies have sug-gested that REM sleep, and associated dreaming,may perform vital functions. Studies on the child-hood development of dreaming are also pertinent.The neurobiology of REM sleep as the primarysource of data from which to make inferences on

dreaming was pioneered by Hobson and McCarleyin their “activation-synthesis” model (recently up-dated as the activation/input/modulation-AIM-model by Hobson), in which brainstem activationduring REM sleep randomly stimulates the fore-brain, which then attempts to “synthesize” the in-formation into coherent dreams. In this model,dreams are not intrinsically meaningful, thoughsome meaning may accrue through the forebrain’sefforts to “make sense” of its physiologically de-termined stimulation. Various problems with theactivation-synthesis model have been identified.A recent critique by B. E. Jones discusses howbrainstem activation of the forebrain can originatefrom highly ordered brainstem circuits that gener-ate not only specific motor patterns but also com-plex organized behaviors. The possible relation to dreaming of the “hippocampal theta rhythm”in REM sleep is discussed. The “neuropsychologyof dreaming” proposed by Solms is presented,based on the large study of neurological patientswith widely distributed brain lesions. Neuroimag-ing techniques and their application to dreamingacross the sleep-wake cycle are another new sourceof data. The second part of this article discussesabnormal dreaming with the parasomnias, whichare the behavioral, autonomic nervous system and experiential disturbances that can accompanysleep. REM sleep behavior disorder is the proto-typic dream-enacting parasomnia in which thecharacteristic generalized muscle paralysis of REMsleep, i.e. “REM atonia”, becomes compromised,allowing for behavioral release associated with ste-reotypically altered dreams involving the dreamerbeing threatened or attacked by unfamiliar peopleor animals.Sleepwalking and sleep terrors in adults,obstructive sleep apnea (OSA “pseudo-RBD”),nocturnal seizures and nocturnal dissociative dis-orders can also present with dream-enacting be-haviors. The recently identified “epic dream dis-

S C H W E I Z E R A R C H I V F Ü R N E U R O L O G I E U N D P S Y C H I A T R I E w w w. s a n p . c h 1 5 6 n 8 / 2 0 0 5440

Dreaming: a psychiatric view and insights from the study of parasomnias1

n A. S. Eisera, C. H. Schenckb

a University of Michigan Sleep Disorders Center, Department of Neurology and Psychiatry,University of Michigan Medical School, Ann Arbor (USA)

b Minnesota Regional Sleep Disorders Center, Department of Psychiatry, Hennepin County Medical Center and University of Minnesota Medical School, Minneapolis (USA)

Correspondence:

Alan S. Eiser, PhDUniversity of Michigan Sleep Disorders CenterUH-8D87021500 East Medical Center DriveUSA-Ann Arbor, MI 48109-0117e-mail: [email protected]

Carlos H. Schenck, MDMinnesota Regional Sleep Disorders CenterHennepin County Medical Center701 Park Ave. SouthUSA-Minneapolis, MN 55415e-mail: [email protected]

Original article

1 Acknowledgments: This work was supported in part by a grant from Hennepin Faculty Associates (CHS). Mark W. Mahowald, MD, is acknowledged for his ongoingcollaboration (with CHS). Each author (ASE, CHS) cri-tically reviewed each other’s section of this manuscript.

order” features relentless, neutral-content “epic”dreaming without emotionality that is experiencedto occur throughout sleep.

Keywords: dreaming; psychiatry; REM sleep;non-REM sleep; parasomnias; polysomnography;REM sleep behavior disorder; sleepwalking; sleepterrors; obstructive sleep apnea; dissociative dis-orders; psychoanalysis; unconscious processes;internal conflict; instinctual behaviors; neurobiol-ogy; neuroimaging; adaptation

Introduction

The contemporary student of dreaming is facedwith a wide, disparate, and potentially bewilderingarray of sources of data as well as theoretical views.Workers in this area have drawn material from suchareas as the intensive clinical study of individualdreams, systematic samples of recalled dreams,experimental manipulations of REM sleep anddreaming, neurobiological studies of REM sleep,lesion and imaging studies, and a variety of others.Indeed, in this paper, we will consider a new sourceof data, changes in dreaming seen in certain sleepdisorders.How best to consider material from thesevaried sources, and how to assess the adequacy andusefulness of the different theoretical models avail-able to account for this material, are complex anddifficult questions. In what follows,we shall presentsome of the major areas of data, important findings,inferences drawn,and implications for the differentextant theoretical models. In Part II, various in-sights on dreaming from the study of parasomniaswill be considered, including the presentation ofaltered dreaming associated with dream-enactingbehaviors.

Part I:

Dreaming as seen from a range of approaches

A. S. Eiser

Intensive clinical study of dreams

There has perhaps been no better opportunity forthe intensive, in-depth study of individual dreamsin their subjective context than that afforded inclinical work. Here dreams can be explored withmaximal knowledge of the dreamer’s present-daycircumstances, past history, salient inner themesand conflicts, idiosyncratic subjective meanings andmodes of thought, and, importantly, relationship tothe person to whom the dream is reported. Themost rich, comprehensive, and influential theory

to be devised from clinical work is the psychoana-lytic, initially developed by Freud [1] and elabo-rated and modified by many contributors since.This model will be described in some detail, bothbecause of its importance in thinking about dreamsand the fact that it deals with so many of the keyissues that need to be considered in a theory ofdreaming.

The psychoanalytic model understands dreamsto be highly meaningful mental products. Theirnature is determined both by general principles ofmental functioning and particular circumstancesobtaining during sleep (predominantly REM sleep,according to present understanding). Unconsciousprocesses play a major role in the construction of the dream; such processes follow, as in wakingmental life, different laws, called primary-processmechanisms, than do conscious thought processes(which utilize secondary-process mechanisms).Thedream is an attempt at the fulfilment of a wish; thewishes that serve as the motive force for dreams areconflicted instinctual, i.e. sexual and aggressive,wishes that can ultimately be linked to childhoodsources. During sleep, the controlling/inhibitoryforces that deal with these wishes are weakened butnot absent; a motor blockade helps protect fromenactment of the wishes.The meaning of the dreamis not apparent in the “manifest content” of thedream, or the dream as recalled by the dreamer,but is to be found in the “latent content”, whichincludes the unconscious wishes. Processes of asso-ciation to the manifest elements of the dream arerequired to uncover the latent dream. The latentcontent is transformed into the manifest content by the unconscious, primary-process mechanismswhich include condensation, the combining orfusion of two or more elements into a single one,and displacement, a shifting of emphasis from oneelement to a different one. Symbolisation is alsoinvolved in the transformation of the latent into the manifest content of the dream.Since the wishesare conflicted, they must be disguised so as not toarouse excessive anxiety and disturb sleep; theprimary-process mechanisms subserve this needfor censorship, which is diminished but still pre-sent during sleep. If disguise fails and the dreamarouses excessive anxiety, a nightmare with awak-ening may result. (Another mechanism for night-mares may be the need to actively replay a trau-matic situation that was originally experiencedpassively, i.e.without preparedness or control, in aneffort to obtain mastery, though it is not certainwhether this is truly distinct from the more usualprocesses that enter into dreaming.) The manifestdream also contains “day residues”, or thoughtprocesses typically from the previous day, which

S C H W E I Z E R A R C H I V F Ü R N E U R O L O G I E U N D P S Y C H I A T R I E w w w. a s n p . c h 1 5 6 n 8 / 2 0 0 5441

are the point of contact between the unconsciouswishes and the dreamer’s current concerns. Themental activity of the dream is blocked from motoroutput due to the state of sleep (particularly REMsleep),and undergoes a regressive process in whichthe dream is represented perceptually, as a halluci-nation.

The psychoanalytic model can be seen to en-compass hypotheses about the dream’s meaning-fulness, the kind of mental material (conflictedwishes; diminished inhibitory structures) withwhich it deals, the nature of the thought processeswhich enter into its construction,and its connectionto ongoing life concerns and experiences. A tech-nique for understanding the dream’s meaning, andplacing it in the context of the individual’s present-day life and his or her significant conflicts and pasthistory, is also part of the model. Indeed, with theexpansion over time of general psychoanalytic the-ory, the conceptual richness that can be brought to bear in understanding dreams has also grown.For example, developments in conflict theory,the theory of object relations (internal relatednessto people), and the understanding of narcissism(development of the self and self-esteem) all addscope and depth to the possibilities for compre-hending dreams.

The richness, complexity, and specificity of thepsychoanalytic model are commensurate with andfaithful to the nature and quality of the individualdream experiences which they are intended toelucidate. At the same time, these very strengthshave made it difficult to develop truly adequate andsatisfactory experimental approaches for testingthe model. This has led some to question its scien-tific status, e.g. [2], but the difficulty may be more areflection of the very complexity that gives the the-ory its special value, the fact that it deals withunconscious, covert elements that cannot be ap-prehended in a simple or direct way, and that someof its postulates deal with fundamental elements ofmental functioning in general that are not easilytestable in the narrow realm of dreaming. In anycase, the challenge of devising studies that engageessential aspects of the model and satisfactorilyoperationalize key concepts has proven to be quitedifficult.

In-depth laboratory studies of dreams

When, with the discovery of REM sleep and elab-oration of the NREM/REM sleep cycle [3, 4],sleep laboratory techniques became available forsampling dreams systematically and more imme-diately after their occurrence, a number of studies

were carried out that made serious efforts to ex-amine dreams in an intensive way, with detailedattention to their individual context and potentialsubjective meanings. Associations to dream ele-ments were collected, and in-depth informationabout the dreamers’ mental life was obtained, oroften was available due to subjects being psycho-therapy patients. Interesting preliminary findingswere made. For example, evidence was found thatsuggested the dreams of a particular night may dealwith the same emotional conflict, or a limited num-ber of conflicts [5].Exploration of which of a night’sdreams a subject will or will not report to either theexperimenter who awakens them or a psychiatristinterviewing them the following day suggested thatdreams, or aspects of dreams, that contain areas ofinterpersonal conflict (often not conscious), espe-cially where these conflicts involve the person towhom the dream is to be reported, tend not to bereported or to be defensively altered [6]. Unfortu-nately, conceptually rich studies such as these tendto be highly labour-intensive, such that sample sizesare very small, and difficult to generalize from,given the uniqueness of individual subjects andoften the use of subjective judgments based onlarge amounts of material as measures. Over time,such studies became increasingly rare, to be re-placed by approaches utilizing easily operational-ized measures which unfortunately tended to dealonly with more concrete and narrow aspects ofdreams (e.g. numerical counts of the characters inthe dream, the incidence of recent vs. older memo-ries, etc.) without reference to the dream’s overallmeaning and significance or the guidance of anoverarching theoretical perspective.

Large-sample studies of dreams

Another approach has been to collect dreams fromlarger samples of subjects who are involved in aparticular life circumstance, often a disturbing one,and study how the circumstance,and efforts to copewith it, may be reflected in dreaming. Perhaps bestknown is the work of Cartwright on dreaming in people undergoing divorce [7]. She obtaineddreams in the laboratory from 70 subjects who wereseparated and anticipating divorce, 40 of whomwere judged depressed about their pending divorceand 30 who were seen as not depressed. She report-ed that the depressed subjects had more dreamswith negative feeling than those who were not de-pressed, indicating that daytime mood is reflectedin dreaming.The figure of the spouse was found toappear in the dreams of those who were depressedabout the divorce much more often than in the


dreams of subjects judged to be coping well. Thissuggested that dreams reflect emotionally dis-tressing issues from waking life. The final reportedfinding was that among the depressed subjects,those who incorporated the spouse in their dreamswere more likely to be doing well at one-yearfollow-up than those who did not incorporate thespouse. Cartwright interpreted this finding to indi-cate an adaptive, emotional problem-solving func-tion for dreaming: those subjects who dreamt aboutthe spouse were working on the problem of thedivorce in their dreams, and therefore more likelyto have a favorable outcome. This last interpreta-tion would be more compelling if data were pre-sented that elucidated why, among non-depressedsubjects, absence of the spouse in the dream was areflection of successful coping with the divorce,whereas among depressed subjects, non-appear-ance of the spouse indicated a lack of adaptive,problem-solving efforts. This difficulty reflects thebroader limitation that much potential richness and specificity is lost when single elements areabstracted from dreams and understood only in a rather literal fashion. It is likely that strugglesregarding the divorce were reflected in a variety of other ways in subjects’ dreams than simply bydirect incorporation of the spouse, and that thesestruggles differed qualitatively among subjects inways that might be linked more convincingly to theprognosis for a good outcome. Also, as Cartwrightnotes, it is difficult to know whether adaptive ef-forts reflected in dreaming indicate an adaptivefunction for dreaming or merely that dreams reflectwhat is happening, including adaptive efforts, else-where in a person’s life.

It becomes possible to appreciate some of thedata in a more multidimensional way when a num-ber of dreams from each of several individualsubjects, along with their related life circumstancesand typical modes of functioning, are discussed [8].Most compelling is Cartwright’s demonstrationthat the dreams reflect a subject’s overall copingcapacities and strategies, however effective orlimited. Some of the interpretations of the mean-ings of individual dreams also seem quite cogent,but in other cases, without associations to elucidateindividual elements of the dreams or richer, morein-depth data about the subjects’ inner lives, theinterpretations can seem somewhat arbitrary,deriving from rather generic application of someknown significant themes in the subject’s life,assessment of where the content falls on some“dream dimensions”, etc. In addition, relying a fair amount on subjects’ own interpretations andinferences about their dreams makes appreciationof deeper unconscious levels less likely.

Hartmann [9] has drawn on a similar source ofdata in formulating his view of dreaming. He hasstudied series of dreams reported by subjects overa period of time after experiencing a psychologicaltrauma, and regards severe trauma as a paradig-matic situation for understanding dreaming. Hereports that initially, subjects have dreams thatreflect raw feelings of terror, being overwhelmed,etc., as in a dream of being threatened by a tidalwave. Working with a neural nets model of themind, he sees dreams as acting to make connec-tions, in a much broader way than waking mentalactivity, and in an “autoassociative” mode. Guidedby the dreamer’s emotion, the dream connects thetrauma and related emotional material, in order to“contextualize” the dominant emotion in the formof an explanatory metaphor. Over time, the dreamconnects the experience of the trauma more andmore broadly with emotionally similar materialfrom the dreamer’s life, experience, and mind, andthe trauma becomes more and more integrated.The process functions both to calm the immediateemotional storm and to promote more ready inte-gration of similar traumas in the future.

The findings and ideas are certainly of interest,though a number of questions may be raised, in-cluding some similar to those about Cartwright’swork. The singling out of particular elements fromdreams, and application of a single interpretiveparadigm, at times in the absence of the dreamer’sassociations, may result in a loss of richness, com-prehensiveness, and specificity in understandingthe dream.That dreams employ broad connectionsseems clear; it is less compelling that the functionof the dream is to make such connections, or thatemotions, which are undoubtedly important indreams, act to guide the dream in this process ofmaking connections. The central place that themodel assigns to trauma in dreaming does noteasily fit with the fact that the cyclic occurrence of dreaming during the night and its ontogeneticdevelopment are largely preprogrammed ratherthan reactive to trauma; in addition, the role ofinternal psychological conflict appears somewhatmarginalized. In extending his model to “ordinary”dreams, i.e. dreams that do not follow a trauma,Hartmann suggests that such dreams too functionto deal with the dreamer’s dominant emotionalconcern, but that the concern may not be apparentor there may be a number of such concerns activesimultaneously, making the working of the dreammore obscure. However, it may also be that themodel cannot as readily accommodate ordinarydreaming. When moving from trauma to moretypical dominant emotional concerns, Hartmannmay be in a realm with some conceptual overlap


with persistent internal conflict, but it is not clearthat the level of explanatory power is preserved. Inemphasizing an adaptive function for dreaming,Hartmann is in agreement with a number of con-temporary theorists from a variety of disciplines.

Basic physiology of REM sleep

The basic parameters and physiology of REMsleep, as the stage of sleep in which most dreamsoccur, have been utilized as another source of datafor approaching an understanding of dreaming.Theregular, cyclic,and essentially universal occurrenceof REM sleep indicates that the occurrence ofdreaming is largely preprogrammed and not, assuggested in earlier psychoanalytic thought,a morehappenstance occurrence dependent on the chancearousal of an instinctual wish. The muscle atoniawhich is distinct to REM sleep is consistent with the idea that a motor block plays an important, inpart protective, role in processes of dreaming, andsupports the idea that it is the state of REM sleepin particular rather than sleep as a whole that ismost facilitative of dreaming.The sheer quantity ofdreams is such as to suggest that defensive pro-cesses, i.e. repression, cannot account for themajority of dream forgetting; indeed, research has indicated that an awakening during or shortlyafter an REM period may be necessary for a dreamto be committed to memory [10], after which for-getting may indeed be determined by psychological defense as well as other factors. Finally,the cyclic episodes of penile erection in men andvaginal blood engorgement in women occurring inclose temporal association with periods of REMsleep suggest processes of physiological drive acti-vation during REM consistent with an importantrole for instinctual drives, and related wishes, indreaming.

Experimental manipulations of REM sleep/dreaming

A variety of experimental manipulations of REMsleep and associated dreaming have been carriedout in an effort to better understand the nature,process, and function of dreaming. Subjects havebeen exposed to a wide range of stimuli both during REM sleep and before going to sleep in an effort to see whether and how dreaming isaffected. Stimuli presented during REM sleep may be incorporated in the ongoing dream bothdirectly and in a transformed manner; the fre-quency of incorporation varies with the kind of

stimulus. In an early study of this kind, by Dementand Wolpert [11], the stimulus most frequently in-corporated in the dream was a spray of water onthe skin, at a rate of 42%; the least frequent was a 1000 hz tone, at a rate of 9%. An example of adirect incorporation of the spray of water was thedreamer “being squirted by someone”; an indirectincorporation was “a leaking roof”. Most often,rates of incorporation of stimuli applied duringREM sleep have been found to be relatively lowand such incorporations as do occur typically haverelatively little impact on the dream as a whole[12]. Rechtschaffen was impressed by this relativeimperviousness of the dream to external stimula-tion, and the dream’s tendency to follow its own,endogenously determined course, in formulatinghis view of “the single-mindedness and isolation”of dreams. The effects of a variety of pre-sleepexperiences on subsequent dreams have also beenstudied, including exposure to films with aggres-sive, sexual, or anxiety-provoking content and de-privation of water and food. Here too, it has oftenbeen reported that direct and straightforwardeffects on the content of dreams are relatively infrequently present. However, many of thesestudies may have been limited by significantmethodological and theoretical shortcomings:without consideration in some depth of a sub-ject’s particular psychology, it is difficult to knowwhether a generic type of pre-sleep experience willimpinge on the subject’s areas of conflict and if so in what way, creating ambiguities both for theexpectation and the assessment of incorporation indreams. A study in which dreams were collectedafter experiences that were stressful in known, per-sonally distinct ways, such as participation in grouptherapy sessions where the subjects’ problems werethe focus for the group,and in which dream contentwas studied in light of knowledge of the dreamer’ssalient issues, modes of adaptation, etc., reportedhigh rates of incorporation in the dreams of theconflicts stirred up by the pre-sleep experiences[13].

A study that is of particular interest both for its theoretical clarity and relevance and the appro-priateness of the measure employed is that ofShevrin and Fisher [14]. Here the pre-sleep stimu-lus to which subjects were exposed was a specialsubliminal stimulus that had previously beendemonstrated to evoke both primary- and sec-ondary-process levels of responses which could bescored in a precise, objective fashion. In free asso-ciations obtained after awakenings from subse-quent REM and NREM sleep, it was found thatprimary-process associates to the pre-sleep stimu-lus predominated after REM-sleep awakenings


whereas secondary-process associates predomi-nated after NREM awakenings. These findingssupport the psychoanalytic view that primary-process modes of cognitive functioning predo-minate in association with dreaming, at least thatwhich occurs during REM sleep. The use of sub-liminal stimulation provided a means for trackingunconscious influences on thought processes inassociation with REM/dreaming sleep and NREMsleep.

A very extensively utilized experimental mani-pulation has been the deprivation of REM sleep,usually by procedures that awaken subjects when-ever they begin to enter REM. Pioneered byDement, this method has consistently demon-strated that subjects deprived of REM sleep makeincreasingly frequent attempts to enter that stage(“increased REM pressure”), and, after REMdeprivation is terminated, show a compensatoryincrease in REM time (“REM rebound”). Thesefindings suggest that REM sleep, and associateddreaming, perform a vital function.An early hypo-thesis, based in part on psychoanalytic thought,was that subjects deprived of REM sleep (andassociated dreaming) would become psychotic,with the dream-like mentation spilling over intowakefulness, but this was not borne out [15].

A number of studies in cats, rats, and mice haveprovided evidence that prolonged REM sleepdeprivation leads to hypersexual and/or hyper-aggressive behavior during wakefulness.A strikingearly description of hypersexuality provided byDement and colleagues was of male cats who,during prolonged REM deprivation, would makepersistent efforts to mount other male cats, awakeor anesthetized, behavior not seen in any othercircumstances [16, 17]. Although some of the earlywork was not presented in fully quantified form,and must be seen as exploratory and suggestive, itis lent additional credence by later, more formalexperimental reports that similarly found hyper-aggressive and hypersexual behavior with REMdeprivation (e.g., in rats, [18–20]). It has also beenfound that REM deprivation leads to an increasein intracranial self-stimulation behavior (ICSS) in rats [21]. These findings are consistent with theview that REM sleep and associated dreaming areimportantly involved with basic drives and plea-sure-seeking,manifestations of which increase dur-ing wakefulness when REM sleep is decreased.And conversely, ICSS behavior in REM-deprivedrats has been reported to result in a reduction ofREM rebound [21], while electrical stimulation of hypothalamic defense (rage) reactions in catsresults in both a decrease in subsequent REM sleepand, in REM-deprived animals, a decrease in sub-

sequent REM rebound [22], findings suggestive ofreciprocal elements in the relationship betweenREM sleep and drive.

REM deprivation has also been used exten-sively to investigate the hypothesis that REM sleep(and perhaps associated dreaming) is important inlearning and memory consolidation.Subjects,oftenanimals, have been exposed to a new learning taskand then deprived of REM sleep, with the expec-tation that impairments in learning would result.Early studies yielded mixed results; more recentefforts, and conceptual refinements, may be yield-ing more clear-cut evidence of such a relationship[23, 24], but this remains an area of controversy[25].

Childhood development of dreaming

An important approach to understanding dream-ing is the study of its childhood development.REMsleep (or active sleep, its developmental precursor)occupies about 50% of sleep time at birth; thisdecreases to around the adult level of 20–25% by ages 3 to 5 years. It is very difficult, however,to ascertain what this means for dreaming; theyounger a child is, the more limitations in con-ceptual and linguistic abilities, and in memory,complicate determination of what the child expe-riences during REM sleep. In laboratory studies,dreams have been obtained from children as youngas 2 years of age [26], though dream recall is re-ported to be much less frequent the younger thechild. Foulkes, who has done the most extensivework in this area, reported that a median figure ofonly 15% of REM awakenings in children ages 3 to 5 yielded dream reports. He characterized thedreams of very young children as brief and static,with barnyard animals the most frequent charac-ters and body states often represented. Only grad-ually with development to the 7 to 9 year age rangedo dreams come to include such features as socialinteraction, an active self character, locomotion or other movement, and a narrative quality [27].Foulkes links these developmental changes indreaming to the child’s overall cognitive develop-ment, according particular importance to the de-velopment of visuo-spatial skills in being able tohave dreams that are complex narratives. In addi-tion to uncertainties stemming from the difficultiesin accessing young children’s dreams mentionedabove, some questions have been raised about howaccurately laboratory-collected dreams reflect ac-tual dreaming experience in very young children,given differences from impressions obtained fromparents, who are much closer to the children than


the experimenter, and from clinical populations[28, 29]. However this may be, it is an observationof central importance that dreaming undergoes aprocess of development that must be seen in thecontext of the child’s developing cognitive capa-cities.

Foulkes seems on much weaker footing in hisview that the contents of children’s (and alsoadult’s) dreams are not very different from ordi-nary cognitive activity, representing the operationof typical waking consciousness except for theabsence of external sensory input and of self-con-trol of ideation. He sees these conditions as lead-ing to relatively mundane, but novel, combinationsof knowledge and memory.Unconscious processes,internal conflicts, intense feelings, disguise, etc.have no place in this model. But absent a means of evaluating, in their individual psychological con-text, what the elements in the dreams he collectsmay signify to the (child) dreamer, there is no wayof determining whether the dreams have meaningsother than what taking their manifest character atface value would suggest. Indeed, one of Foulkes’own findings, that barnyard animals are the mostfrequently occurring characters in preschool chil-dren’s dreams, suggests a view quite different fromhis own. If dreams simply draw upon and combineknowledge and experience in novel yet mundaneand realistic ways, one would expect the child’sparents and siblings, the most frequently present(and important) characters in its life, to be the mostcommon characters in dreams, rather than animalswith which the child may have little direct expe-rience and may know mainly from stories, fairytales, movies, etc. However, such animals appear in imaginative products that appeal to childrenprecisely because of their value in representing,and at one remove, the child’s troubling impulses(e.g. biting, devouring, messing, sexual urges),frightening images of the parents, and other ele-ments of the internal conflicts that occupy such animportant place in the child’s inner life [30]. Theirfrequent appearance in dreams suggests that thedreams too deal with such impulses and internalconflicts in ways that may not be immediately ap-parent.

Neurobiology of REM sleep

Another approach to understanding dreaming hasbeen to utilize the increasing knowledge of theneurobiology of REM sleep as the primary sourceof data from which to make inferences. Here, thedistance between the data and the intended objectof study, dreams and dreaming, grows wider, and

the possibilities for error or bias to lead the in-ference process astray become proportionatelygreater. The most widely known and influentialsuch attempt is the activation-synthesis model ofHobson and colleagues [31],a group who have beenmajor contributors to the understanding of the sitesand neurochemical interactions in the brainsteminvolved in the generation of REM sleep. Themodel was derived by direct interpretation of theneurophysiology, based on what the authors term“mind-body isomorphism”, an assumption of an“identity of form” between physiological andpsychological events. It leans heavily on the fact ofthe generation of REM sleep in the pontine brainstem. In the activation portion of the model, thebrainstem is seen as providing largely random,direct stimulation of the forebrain, for example ofthe oculomotor, vestibular, and motor systems,accounting for the predominance of visual andmovement elements in dreams.The forebrain thenattempts to “synthesize” the information that hasbeen generated in the pontine brain stem: “Best fits to the relative inchoate and incomplete dataprovided by the primary stimuli are called up frommemory, the access to which is facilitated duringdreaming sleep.” In this, “the forebrain may bemaking the best of a bad job in producing even par-tially coherent dream imagery from the relativelynoisy signals sent up to it from the brain stem”.Themodel explains bizarre formal qualities of dreams,such as spatiotemporal distortions, condensations,and discontinuities, as the direct result of the ran-dom pattern of brainstem stimulation. Dreams inthis view are not essentially meaningful, thoughsome meaning may accrue secondarily through the forebrain’s efforts to make sense of its physio-logically determined stimulation.

Fundamental problems with the activation-syn-thesis model were identified from the time of itsinception. In an early critique, Vogel [32] pointedto three such issues. He noted that the assumptionthat a particular temporal discharge pattern ofbrainstem neurons would cause particular quali-ties in mentation, including “bizarre” formal qual-ities, was entirely arbitrary. In laboratory dreamresearch, the bizarre formal qualities in dreamshave been found to be much less common than themodel would suggest [33], and efforts to correlatethese qualities in actual dreams with indicators ofphasic activation from the brainstem have gener-ally shown weak relationships at best [34]. Second,the model overstates the role of the brainstem;evidence was available even at that time that in the intact animal, forebrain structures interact im-portantly with the brainstem sites in the productionof REM sleep. Finally, a considerable amount of


dreaming that is indistinguishable from REM-sleepdreaming occurs in NREM sleep, when the phy-siological processes upon which the activation-syn-thesis model bases its explanation of dreaming arenot found.

Laboratory dream researchers such as Recht-schaffen and Foulkes have particularly emphasizedin their criticisms of the activation-synthesis modelthe lack of fit between the relative coherence andorganisation of most dreams and what the activa-tion-synthesis model would suggest. In Rechtschaf-fen and Siegel’s words [35], “some scholars haveproposed that dreams arise from random brainactivity, but dreams are not kaleidoscopic jumblesof visual fragments; they are organized thematical-ly and perceptually”.

In a recent, important critique, Jones, herself acentral contributor to the understanding of theneurobiology of REM sleep for more than threedecades, argued that Hobson and colleagues hadinterpreted the physiology in an arbitrary fashion[36]. She emphasized two points: first, the absenceof evidence to support the idea, fundamental to the model, that the brainstem activation of the fore-brain is chaotic. She noted that on the contrary,brainstem circuits are highly ordered, generatingnot only specific motor patterns but also “complexorganized behaviors … including sexual and ragebehaviors that persist in decerebrate animals”.The latter behaviors are “instinctual and highlymotivational, perhaps stimulating the wishes thatemerge in dreams”. She also cited evidence that the cortex, far from having no control over thebrainstem in REM sleep and dreaming, has influ-ence “over the very neurons shown to be critical for the initiation and maintenance of REM sleepin the pontine reticular formation”. Elimination of these cortical inputs results “in a complete im-poverishment of both rapid eye movements andPGO spikes, reducing them to the very stereotypic,highly ordered and hence low-information–contentpattern of purely brainstem driven activity”. Shesuggested she “might posit that the physiology ofREM sleep provides considerable support forFreud’s basic assumptions, according to which in-stinctual and highly motivational impulses arisefrom the brainstem and are in turned worked uponby the cortex where condensation, displacement,and symbol formation may control the continuedactivity of the brainstem and provide the complexand seemingly bizarre, though meaningful, contentof dreams” [36].

Other difficulties with the activation-synthesismodel are indicated in the sections below on lesionand imaging studies. Although the model has thevirtue of drawing attention to developments in

neurobiology and the potential for utilizing themin understanding dreaming, the specific effort is soflawed by the arbitrary, even tendentious, qualityof its inferences as to have limited value in ad-vancing our understanding.

Over time, some expansion in the activation-synthesis model began to allow for a greater andmore functional role for the forebrain, particularlyin learning and memory consolidation. In its most recent form, the “AIM” model, it attempts to explain in neurobiological terms the entire spec-trum of “brain-mind” states based on where theyfall on three dimensions: level of brain activation(“A”), thought to account for the quantity, com-plexity, and intensity of mental activity, source ofinput (“I”), either from external sensory or inter-nal sources, and neuromodulatory balance (“M”),the ratio of aminergic to cholinergic influence,thought to account for mode of cognitive func-tioning, including the degree of directed thought,self-reflective awareness, emotion, insight, andmemory [37]. REM sleep and its associated dream-ing are then understood as comprised of a highlevel of activation (hence the intense mental ac-tivity), internal rather than external input, andaminergic demodulation, accounting in part for the diminution of directed thought, self-reflectiveawareness, “insight” into illogical elements, andmemory.

Unfortunately, many of the fundamental short-comings of the activation-synthesis model areretained. The “chaotic nature of the pontine auto-activation process” continues to be seen as an im-portant factor in the bizarre qualities of dreaming.More broadly, the view is maintained that the brainin dreaming functions in a kind of fragmentary,un-integrated fashion; the brainstem stimulatesmany disparate elements of cognitive activitywhich the forebrain then has to “synthesize”. How-ever, more coherence and organisation do comeinto play in the AIM model in the idea that emo-tion has a role in shaping dream plots. The hypo-thesis that aminergic demodulation is connected tosome of the cognitive properties of dreaming willneed to be tested, but it is a confusion of levels ofdiscourse (psychological, neurobiological) to thinkthis would “explain” those cognitive qualities asopposed to elucidating their neural correlates.Explanations of properties of mental activity mustreside in the first instance in grasping their psy-chological nature and significance. The approach is greatly colored by the view that dreams aredeficient, and in many ways pathological, forms of mental activity, seen in the pervasive use of terms like “hallucinosis”, “delusion”, and “lack ofinsight”; the task in understanding dreams becomes


in effect one of searching out their pathophysio-logy.

Other perspectives on dreaming have beendeveloped from the neurobiology that follow moreclosely from the actual data.Winson [38,39], relyingprimarily on animal studies, has drawn creativelyon neurobiology in conjunction with evolutionaryfindings on REM sleep and experiments in learn-ing. His work centres on hippocampal thetarhythm, which has been shown to occur duringwakefulness in a variety of subprimate mammals in situations in which species-specific behaviors im-portant for survival are engaged, and also through-out REM sleep in subprimate mammals. Thetarhythm plays an important role (through “LongTerm Potentiation”) in the storage of memory.And, in unit cell recordings from rats, hippocampalcells that have been active in encoding spatialinformation during prior wakefulness fire morefrequently during subsequent sleep. Winson hypo-thesizes that during REM sleep, experience thatpertains to species-specific behavior important forsurvival is integrated with past experience to formnew behavioral strategies. He links this to findingsregarding the evolutionary origins of REM sleep,and the adaptive necessity for a means of process-ing new information “off-line”, i.e. during sleep. Inhumans (who also may have theta rhythm [40, 41]),he extends the hypothesis to suggest that in REMsleep, and dreams, it is not information pertainingto specific behaviors that must be integrated butrather all waking experience that pertains to “psy-chological survival”. Here he links his thinking toauthors such as Cartwright who propose an adap-tive function for dreaming. He further notes thatthe intricate associations regarding relevant or similar events that emerge during the integrativeprocess in dreaming are “strongly biased towardearly childhood experience”. Winson believes thatit is at about the age of two, when the hippocampusbecomes functional, that REM sleep and dreamingtake on their integrative memory function in humans.

An important strength of this interesting viewis that it anchors its understanding of dreaming inanimal findings and evolutionary considerations,through the association of dreaming with REMsleep. Winson feels that his hypothesis implies avery different view of the unconscious and asso-ciated brain functioning than Freud’s, but theremay be a considerably greater degree of corre-spondence than he recognizes. For in humans,central issues involved in “psychological survival”originate in “early childhood experience” and haveto do with the management of wishes and passionswhich become highly conflicted in the context of

the development of crucially important dependentand loving relationships as well as an integratedand cohesive self. The model emphasizes an adap-tive function, in agreement with a number of otherrecent views. It is not clear whether its concept ofspecies-specific behaviors necessary for survivalcan fully encompass the findings regarding the re-lationship between REM sleep and drive that werediscussed above.

Jones has evolved views of dreaming that drawpredominantly upon a broad and rich range ofelements of the neurobiology of REM sleep, as well as some of the perspectives of Winson. Sheemphasizes that the brainstem may provide “fun-damental, species-specific information of complexsensorimotor behavioral programs”; “… this fun-damental information … is also modulated andelaborated in the normal intact animal by forebrainand cortical processing and descending feedback to the brainstem … Thus, dreams may be made offundamental sensorimotor and behavioral pro-grams important in development, but these mustalso be modulated and elaborated by complexlearned material throughout adult life” [42]. Asindicated above, she notes that the behavioralpatterns are “instinctual and highly motivational”[36], and perhaps relates them more broadly todrive than does Winson. She does emphasize theimportant role of theta rhythm in consolidating and potentiating new learning in regard to species-specific instinctive behaviors, and has contributedto the understanding of the role of cholinergicneurons in the basal forebrain in stimulating thetarhythm in the cortex during REM sleep [43]. Else-where, citing evidence from EEG spectral analysis,she has also noted that “the coherence in gammaactivity between distant cortical areas is as high inREM sleep as during the most aroused wakingstates”, and that there are similar findings for thecoherence of theta activity. This suggests “thepossibility for integrated activity across widelydistributed cortical regions” during REM sleep[44], consistent with a view of dreaming as highlyorganized mental activity.

Lesion studies

A very significant new contribution to the study ofdreaming and its relation to the brain has beenmade through the approach of examining changesin dreaming that occur with lesions in differentareas of the brain. Although numerous scatteredreports of small numbers of subjects or particulartypes of lesions have appeared in the literature for many years, it was Solms’ contribution to sys-


tematically study a large sample of 332 neurologicalpatients with widely distributed lesions and com-bine his findings with the earlier reports collatedfrom the literature [45]. In his clinicoanatomicalapproach, Solms correlated reported changes indreaming with the site of the lesion and the asso-ciated clinical syndrome and deficits, and utilizedthe findings to develop a neuropsychology ofdreaming. An important strength of the approachis that it links data concerning the brain directlywith reported qualities of dreams, rather than re-lying on indirect inferences based on previouslyunderstood relationships between areas of thebrain and aspects of mental functioning and on theassociation between REM sleep and dreaming.A major finding was that there are two areas of thebrain in which lesions led to complete cessation of dreaming. One area is at or near the parietal-temporo-occipital (PTO) junction; both unilateraland bilateral PTO lesions had this effect. Solmsunderstood the loss of dreaming as based on thisarea supporting “various cognitive processes thatare vital for mental imagery”, the right PTO areabeing essential for concrete spatial cognition andthe left PTO area for quasi-spatial (symbolic)thought. The second type of lesion which led tocessation of dreaming was bilateral deep frontallesions. Solms understood this effect as due todamage to fibers that connect the dopaminergicventral tegmental area to the limbic system andfrontal cortex, circuits that have been thought to beessential for instigating appetitive interactions withthe world (the “seeking” or “wanting” commandsystem), as well as for the regulatory elaboration of appetitive drives into adaptive form. Solmsinferred that these motivational functions are alsoessential for dreaming, and that the “wanting” sys-tem may be the instigator of dreams. Both types of lesion associated with cessation of dreaming arelocated in the forebrain, and do not have any effecton REM sleep, as seen in laboratory studies.

A number of additional findings offer con-siderable further clarification of the cognitivefunctions and brain structures involved in dream-ing. Loss of visual imagery in dreams occurs withlesions in visual association, but not primary visu-al, cortex; these patients experience concomitantloss of the capacity to conjure up visual imagerywhile awake. Other imagery modalities in dreamsfor which there were data available, including somatosensory, somatomotor, audioverbal, andmotor speech imagery, appear to be unaffected by lesions in the corresponding areas of unimodalcortex which profoundly affect these functionsduring wakefulness. These findings provide evi-dence that “the perceptual and motor images we

experience in our dreams are not simple, isomor-phic products of nonspecific activation of per-ceptual and motor cortex during sleep; highly com-plex heteromodal representational mechanismsappear to be involved in the active construction ofthe manifest dream” [45]. Lesions in various ante-rior limbic structures lead to an inability to dis-tinguish dreams from reality, often in conjunctionwith increased frequency and vivacity of dreaming,and, during wakefulness, loss of reality testing,visual hallucinations, and delusions. This suggeststhat these structures are involved in exerting aninhibitory influence on the occurrence and inten-sity of dreams and dream-like mentation. Lesionsin the dorsolateral prefrontal cortex, a regionessential for executive functioning, self-monitor-ing,and volitional control,have no effect on dream-ing, suggesting that those functions are not signi-ficantly involved in the dreaming process.

Solms’ findings provide neuropsychological evi-dence that the forebrain has a crucial, specific, andhighly differentiated role in the dreaming process,with a variety of structures acting together in a con-certed fashion. They support the view that dreamsare “actively constructed through complex cogni-tive processes”. Many specific qualities of dreamscan be accounted for on the basis of the particularconfiguration of forebrain structures involved.

Solms [46] argues for an essential detachmentof dreaming from REM sleep and the brainstem,viewing REM sleep as only one of many sourcesthat can activate the dreaming mechanism, whichhe locates entirely in the forebrain. However, thisview cannot encompass the kinds of neurophysio-logical, neurobehavioral, and adaptive considera-tions related to the role of the brainstem in dream-ing that were discussed above in the work of Jonesand Winson, and more generally removes the studyof dreaming from possible links with findings inanimal research and evolutionary considerations.Although the subjects with brainstem lesions inSolms’ study and some other published reports didnot report loss of dreaming, it has not yet beendemonstrated that lesions in the brainstem thateliminate REM sleep do not lead to cessation ofdreaming. Imaging studies (discussed below) haveshown that the complex pattern of activation anddeactivation of forebrain structures that Solms has connected with dreaming is consistently seenduring REM sleep, but has not been found duringNREM sleep (though the possibility that there arebrief periods during NREM sleep when dreamingoccurs and this activation pattern is seen has not been fully investigated). Similarly, the muscleatonia which apparently evolved to help protectagainst acting out of dreams is only found during


REM sleep. Finally, new findings on alterations indreaming in patients with REM behavior disorder(also discussed below), a condition understood toresult from neurodegenerative processes in thebrainstem, provide evidence that the brainstemdoes indeed play a role in dreaming beyond that ofmere activation. These findings strongly point to amore integral connection between REM anddreaming than in Solms’ model.

Imaging studies

The development of neuroimaging techniques andtheir application across the sleep-wake cycle isproviding another valuable new source of data forthe understanding of dreaming and its relation tothe brain.The approach used in most studies to datehas been to investigate the pattern of brain activa-tion/deactivation during REM sleep, and compareit with NREM sleep and/or wakefulness. Priorknowledge of the functions of different regions andareas of the brain is then utilized to form a pictureof mental processes during REM sleep, and byimplication in dreaming. This approach permits arelatively broad-brush view to be developed ofwhat brain regions, functions,and organisation maybe involved in dreams.Although different imagingtechniques and experimental procedures havebeen used by different groups, consistent findingshave emerged [47–49]. A very specific, selectivepattern of activation of forebrain structures duringREM sleep suggests that the brain is organized tocarry out particular functions in a concerted man-ner. Overall level of activity is high; structures inthe brainstem, thalamus, and basal forebrain thatmediate arousal are activated.

Particularly striking is the finding that parts ofthe hypothalamus, and the limbic and paralimbicsystems, are more highly activated than duringwaking,suggesting a major involvement of emotionand drive (in sharp contrast to the activation-synthesis view that the dream process has little orno primary volitional or emotional content). Theactivation of the amygdala may be linked to therole of anxiety in dreaming. On the other hand,widespread deactivation of the dorsolateral pre-frontal cortex found by most, but not all, of thegroups correlates with diminished executive func-tioning in dreams. Overall, the imaging findingsregarding forebrain structures involved in REMsleep are quite consistent with Solms’ lesion find-ings on forebrain structures in dreaming, eachthereby lending additional credence to the other(though the imaging findings have not providedsupport for Solms’ detachment of dreaming from

REM sleep). Maquet et al. [47], emphasizing therole of the amygdala, suggest that REM sleep isinvolved in processing “emotionally significantmemories”. Nofzinger and colleagues [49] inter-pret the pattern of activation as supporting “theview that one function of REM sleep is the inte-gration of neocortical activity with hypothalamic-basal forebrain regulatory and motivational rewardmechanisms”. They see their findings as consis-tent with a role for REM sleep in memory consol-idation, and as “quite in accordance with older,more general views that REM sleep, and specifi-cally dream content, is associated with internallygenerated, or instinctual behaviors that subserveadaptive mechanisms”. Important advances in thedirection of greater specificity will be made pos-sible by the development and employment ofimaging techniques that permit finer time, and also spatial, resolution, and by efforts to correlateimaging findings with qualities of individualdreams recalled upon awakening from REM, andalso NREM, sleep.

Significant parallels can be drawn between theview of dreaming derived from the lesion and imag-ing studies and the psychoanalytic model derivedfrom intensive clinical work. The highly specificand selective pattern of forebrain activation, con-sistent with structures acting together in a con-certed fashion to carry out particular functions,is broadly supportive of the view of dreams asmeaningful, complexly determined,and organized.The findings of activation of the appetitive, “want-ing”system (Solms) and of structures involved withmotivation and emotion, or “internally generated,or instinctual behaviors” (Nofzinger), support theimportance of instinctual wishes in dreaming. Therelative deactivation of the dorslolateral prefrontalcortex, on the other hand, is consistent with thediminishment of controlling/inhibitory (ego) func-tions, and perhaps contributes to the dominance ofa different (primary-process) mode of functioning.The view that dreams involve conflicted wishes,wishes which are associated with danger and arouseanxiety, finds support in the prominent activationof the amygdala, which is involved with anxiety and the response to threat.The lesion and imagingfindings can similarly be seen as broadly consis-tent with Jones’ view of dreaming based primarily on the neurophysiology, and elements of Winson’sadaptive-evolutionary view.

Dreaming in REM sleep behavior disorder

Part II of this paper is comprised of an extensive,in-depth treatment of dreaming in the parasom-


nias, including REM Sleep Behavior Disorder(RBD); here, some comments will be made re-garding elements of dreaming in RBD which haveparticular relevance for issues raised and perspec-tives advanced in this first part. Although RBD ismost often thought of in terms of the failure ofmuscle paralysis and acting out of dreams occurringduring REM sleep, evidence has accumulated froma number of sources that the dreams themselves arestrikingly altered in this disorder. Patients typical-ly report that their dreams are more vivid, active,and especially aggressive or violent than before[50, 51], and the elevated aggression has been con-firmed in comparison with the dreams of normalcontrols by objective raters [52]. Since RBD is veryclosely associated with certain neurodegenerativedisorders and related brainstem pathology, thesefindings provide evidence that brainstem functionhas a greater role in dreaming than merely servingas a source of activation, as proposed by Solms.Theregular, almost stereotyped alteration of dreamingin the direction of more aggression also contradictsthe activation-synthesis view that the brainstemcontributes random,chaotic stimulation;here,evena damaged brainstem is seen to make a specific and patterned contribution to dreaming. The find-ings seem most closely in keeping with the viewsadvanced by Jones regarding the role of instinctualbehavior patterns (in this instance aggressive pat-terns) arising from the brainstem in REM sleep anddreaming. Also of interest, clinical impressionshave been reported to the effect that men withRBD (the disorder has a high male predominance)tend to have placid, mild-mannered personalities[53]. Fantini et al. [52] report findings that add tothese impressions. Using the Aggression Question-naire, which is thought to measure daytime aggres-siveness as a stable personality trait, they foundthat patients with RBD had a lower mean score onthe physical aggressiveness subscale (though not inoverall aggressiveness) than normal controls.Theyalso found a significant negative correlation inRBD patients between the Aggression Question-naire subscale score for hostility and the percent ofdreams with at least one aggression, as well as sig-nificant negative correlations between the subscalescore for anger and the percent of dreams with atleast one misfortune, and between the subscalescore for physical aggression and the percent ofnegative emotions in dreams.These intriguing find-ings of inverse relationships between daytime anddream aggressiveness in RBD suggest that theneurodegenerative processes in RBD may interactwith personality factors concerned with manage-ment of aggression during wakefulness in ways thatrequire further elucidation.

References

1 Freud S. The interpretation of dreams. In: Strachey J,trans. and ed. The Standard Edition of the Complete Psychological Works of Sigmund Freud. Vols. IV and V.London: The Hogarth Press; 1953 (orig. published 1900).

2 Vogel GW. Critique of current dream theories. Behav Brain Sci 2000;23:1014–6.

3 Aserinsky E, Kleitman N. Regularly occurring periods ofeye motility, and concomitant phenomena, during sleep.Science 1953;118:273–4.

4 Dement W, Kleitman N. Cyclic variations in EEG duringsleep and their relation to eye movements, body motility,and dreaming. Electroencephalogr Clin Neurophysiol 1957;9:673–90.

5 Offenkrantz W, Rechtschaffen A. Clinical studies of sequential dreams. I. A patient in psychotherapy. Arch Gen Psychiatry 1963;8:497–508.

6 Whitman RM, Kramer M, Baldridge B. Which dream doesthe patient tell? Arch Gen Psychiatry 1963;8:277–82.

7 Cartwright RD. Dreams and adaptation to divorce. In: Barrett D, editor. Trauma and Dreams. Cambridge, Massachusetts: Harvard University Press;1996. p. 179–85.

8 Cartwright RD, Lamberg L. Crisis Dreaming: Using YourDreams to Solve Your Problems. New York: HarperCollins; 1992.

9 Hartmann E. Outline for a theory on the nature and functions of dreaming. Dreaming 1996;6:147–70.

10 Goodenough DR. Dream recall: history and current status of the field. In: Arkin AM, Antrobus JS, Ellman SJ,editors. The Mind in Sleep: Psychology and Psycho-physiology. Hillsdale, NJ: Lawrence Erlbaum Associates;1978. p. 113–40.

11 Dement W, Wolpert EA. The relation of eye movements,body motility, and external stimuli to dream content. J Exp Psychol 1958;55:543–53.

12 Rechtschaffen A. The single-mindedness and isolation of dreams. Sleep 1978;1:97–109.

13 Breger L, Hunter I, Lane RW. The Effects of Stress on Dreams. New York: International Universities Press; 1971.

14 Shevrin H, Fisher C. Changes in the effects of a wakingsubliminal stimulus as a function of dreaming and nondreaming sleep. J Abnorm Psychol 1967;72:362–8.

15 Vogel GW. A review of REM sleep deprivation. Arch Gen Psychiatry 1975;32:749–61.

16 Dement W, Henry P, Cohen H, Ferguson J. Studies on the effect of REM deprivation in humans and in animals.In: Kety SS, Evarts EV, Williams HL, editors. Sleep and Altered States of Consciousness. Baltimore: Williams and Wilkins; 1967. p. 456–8.

17 Dement W. The biological role of REM sleep (circa 1968).In: Kales A, editor. Sleep: Physiology and Pathology.Philadelphia: JB Lippincott; 1969. p. 245–65.

18 Ferguson J, Dement W. The behavioral effects of amphetamine on REM deprived rats. J Psychiatr Res 1969;7:111–8.

19 Peder M, Elomaa E, Johansson G. Increased aggressionafter rapid-eye-movement sleep deprivation in Wistar rats is not influenced by reduction of dimensions of enclosure. Behav Neural Biol 1986;45:287–91.


20 Morden B, Conner R, Mitchell G, Dement W, Levine S. Effects of rapid-eye-movement (REM) sleep deprivation onshock-induced fighting. Physiol Behav 1968;3:425–32.

21 Steiner SS, Ellman SJ. Relation between REM sleep andintracranial self-stimulation. Science 1972;177:1122–4.

22 Putkonen PTS, Putkonen A-R. Suppression of paradoxicalsleep (PS) following hypothalamic defence reactions in cats during normal conditions and recovery from PSdeprivation. Brain Res 1971;26:333–47.

23 Smith CT. Sleep states and learning: a review of the animal literature. Neurosci Biobehav Rev 1985;9:157–68.

24 Smith C. Sleep states and memory processes. Behav Brain Res 1995;69:137–45.

25 Vertes RP, Eastman KE. The case against memory consolidation in REM sleep. Behav Brain Sci 2000;23:867–76.

26 Kohler WC, Coddington RD, Agnew HW. Sleep patterns in2-year-old children. J Pediatr 1968;72:228–33.

27 Foulkes D. Children’s Dreaming and the Development of Consciousness. Cambridge, MA: Harvard UniversityPress; 1999.

28 Resnick J, Stickgold R, Rittenhouse CD, Hobson JA. Self-representation and bizarreness in children’s dreamreports collected in the home setting. Conscious Cogn 1994;3:30–45.

29 Salzarulo P, Chevalier A. Sleep problems in children andtheir relationship with early disturbances of the waking-sleeping rhythms. Sleep 1983;6:47–51.

30 Bettelheim B. The Uses of Enchantment: the Meaningand Importance of Fairy Tales. New York: Random House, Inc.; 1975.

31 Hobson JA, McCarley RW. The brain as a dream stategenerator: an activation-synthesis hypothesis of thedream process. Am J Psychiatry 1977;134:1335–48.

32 Vogel GW. An alternative view of the neurobiology ofdreaming. Am J Psychiatry 1978;135:1531–5.

33 Snyder F. The phenomenology of dreaming. In: Madow L,Snow LH, editors. The Psychodynamic Implications of thePhysiological Studies on Dreams. Springfield, IL: CharlesC. Thomas; 1970. p. 124–51.

34 Pivik RT. Tonic states and phasic events in relation tosleep mentation. In: Ellman SJ, Antrobus JS, editors. The Mind in Sleep: Psychology and Psychophysiology.2nd ed. New York: John Wiley & Sons; 1991. p. 214–47.

35 Rechtschaffen A, Siegel JM. Sleep and dreaming. In: Kandel ER, Schwartz JH, Jessel TM. Principles of Neuroscience. 4th ed. New York: McGraw-Hill; 2000. p. 936–47.

36 Jones BE. The interpretation of physiology. Behav Brain Sci 2000;23:955–6.

37 Hobson JA, Pace-Schott EF, Stickgold R. Dreaming andthe brain: toward a cognitive neuroscience of consciousstates. Behav Brain Sci 2000;23:793–842.

38 Winson J. The meaning of dreams. Sci Am 1990;262:86–96.

39 Winson J. The biology and function of rapid-eye-movementsleep. Curr Opin Neurobiol 1993;3:243–8.

40 Raghavachari S, Kahana MJ, Rizzuto DS, Caplan JB,Kirschen MP, Bourgeois B, et al. Gating of human thetaoscillations by a working memory task. J Neurosci 2001;21:3175–83.

41 Cantero JL, Atienza M, Stickgold R, Kahana MJ,Madsen JR, Kocsis B. Sleep-dependent theta oscillationsin the human hippocampus and neocortex. J Neurosci 2003;23:10897–903.

42 Jones BE. Paradoxical sleep and its chemical/structuralsubstrates in the brain. Neuroscience 1991;40:637–56.

43 Lee MG, Hassani OK, Alonso A, Jones BE. Cholinergicbasal forebrain neurons burst with theta during wakingand paradoxical sleep. J Neurosci 2005;25:4365–9.

44 Jones BE. The neural basis of consciousness across the sleep-waking cycle. In: Jasper HH, Descarries L,Castellucci VF, Rossignol S, editors. Consciousness: At the Frontiers of Neuroscience. Philadelphia: Lippincott-Raven Publishers; 1998. p. 75–94.

45 Solms M. The Neuropsychology of Dreams. Mahwah, NJ:Lawrence Erlbaum Associates; 1997.

46 Solms M. Dreaming and REM sleep are controlled by different brain mechanisms. Behav Brain Sci 2000;23:843–50.

47 Maquet P, Peters J-M, Aerts J, Delfiore G, Degueldre C,Luxen A, et al. Functional neuroanatomy of human rapid-eye-movement sleep and dreaming. Nature 1996;383:163–6.

48 Braun AR, Balkin TJ, Wesensten NJ, Carson RE, Varga M,Baldwin P, et al. Regional cerebral blood flow throughoutthe sleep-wake cycle: an H2

15O PET study. Brain 1997;120:1173–97.

49 Nofzinger EA, Mintun MA, Wiseman MB, Kupfer DJ, MooreRY. Forebrain activation in REM sleep: an FDG PET study. Brain Res 1997;770:192–201.

50 Schenck CH, Hurwitz TD, Mahowald MW. REM sleep behavior disorder: an update on a series of 96 patientsand a review of the world literature. J Sleep Res 1993;2:224–31.

51 Olson EJ, Boeve BF, Silber MH. Rapid-eye-movementsleep behavior disorder: demographic, clinical and laboratory findings in 93 cases. Brain 2000;123:331–9.

52 Fantini ML, Corona A, Clerici S, Ferini-Strambi L. Aggressive dream content without daytime aggressiveness in REM sleep behavior disorder. Neurology 2005;65:1010–5.

53 Schenck CH, Mahowald MW. REM sleep behavior disorder: clinical, developmental, and neuroscience perspectives 16 years after its formal identification in SLEEP. Sleep 2002;25:120–38.


Part II:

Insights on dreaming from studying the parasomnias

C. H. Schenck

Parasomnias, which are the behavioral, autonomicnervous system, and experiential disturbances thatcan accompany sleep, are now known to shed con-siderable light on various aspects of dreaming.Thetopics pertaining to dreaming and parasomnias tobe covered in this section include the following:I Parasomnias and

the “mental status exam of sleep”II Dream-enacting disorders:

differential diagnosisIII REM sleep behavior disorderIV Sleepwalking, sleep terrors and

“dreamwalking” in adultsV Epic dream disorderThe author has been practicing psychiatry since1981, and has also been practicing sleep medicineand conducting research on the parasomnias at theMinnesota Regional Sleep Disorders Center since1982.

I Parasomnias and the “mental status exam of sleep”

In the spring of 2005 the American Academy ofSleep Medicine published its first revision of theInternational Classification of Sleep Disorders(ICSD-2) [1]. The author and his colleague MarkW. Mahowald served as Chairmen of the Parasom-nias Committee for ICSD-2, in which parasomniasare defined as “undesirable physical events orexperiences that occur during entry into sleep,within sleep, or during arousals from sleep. Theseevents are manifestations of central nervous systemactivation transmitted into skeletal muscle andautonomic nervous system channels, often withexperiential concomitants. Parasomnias encom-pass abnormal sleep related movements,behaviors,emotions, perceptions, dreaming, and autonomicnervous system functioning”[1].Furthermore,“ba-sic drive states” can emerge in pathologic formswith the parasomnias, as seen with sleep-relatedaggression and locomotion, sleep-related eatingdisorder, and abnormal sleep-related sexual be-haviors. Parasomnias can affect people of any age,and are now known to be surprisingly common.For example, in adults violent behaviors duringsleep have a 2% prevalence, sleepwalking has a 4% prevalence, sleep terrors have a >2% preva-lence, sleep-related eating disorder has a >1% pre-valence, REM sleep behavior disorder (RBD) hasa �0.5% prevalence, and so on [1].

From the author’s cumulative experience withthe parasomnias and the practice of sleep medicine,it has become apparent that a clinical cornerstonein the fields of Neurology and Psychiatry shouldalso be recognized as a clinical cornerstone in the field of Sleep Medicine, which is the “mentalstatus exam”. The assessment of mental status iscomprised of various distinct functions, the suminteractions of which help form the ongoing,moment-to-moment phenomena of our conscious-ness and interactions with the environment. Para-somnias (along with narcolepsy) afford a specialopportunity for considering how the componentfunctions of the mental status can be dissociatedand recombined dynamically across the variousstages of sleep and across the boundaries of sleepand wakefulness, as shown in table 1. The MentalStatus Exam of Sleep can help clinicians identifythe “target symptoms” for treating a parasomnia,or any other category of sleep disorder. In otherwords, the sleep disorder diagnosis is alone notsufficient in either fully conceptualizing the dis-order in a given patient or in formulating a treat-ment plan. A consideration of the mental status ofsleep (and wakefulness) for that patient must also


Table 1 Mental status of sleep.

contexts

within REM sleep

within NREM sleep

during entry into REM or NREM sleep

during arousals from REM or NREM sleep

during mixed states of sleep and wakefulness

categories

1 motor-behavioral

2 dream-cognitive

3 mood-emotional

4 perceptual-hallucinatory

5 autonomic nervous system

6 eating

7 sex

8 convulsive-epileptic

9 vestibular (equilibrium)

10 memory

A full range (and time course) of activations, and dynamic associations, dissociations and recombinations of thesemental status components exists during sleep, during entryinto sleep, or during arousals from sleep, for all age groupsand for both sexes. Transitional sleep-wake (i.e. hypno-pompic) and wake-sleep states (i.e. hypnagogic), and alsowithin-sleep states are represented in this table [41–45].

take place. The clinical challenge will then be to assess the impact of the therapeutic intervention(s)on the severity and frequency of the affected men-tal status components for the given sleep disor-der(s).

II Dream-enacting disorders: differential diagnosis

At least five disorders have been identified toinvolve recurrent dream-enacting behaviors. Be-sides RBD and sleepwalking, sleep terrors (Dis-orders of Arousal), three other conditions, viz. ob-structive sleep apnea (OSA), nocturnal seizuresand nocturnal dissociative disorders can manifestwith dream-enacting behaviors [2]. This is a majorreason why it is necessary for a patient who com-plains of a dream-enacting disorder to have a care-ful clinical interview, together with the bed part-ner, conducted by a sleep clinician that is followedby extensive overnight polysomnographic (PSG)and audio-video monitoring, with a sleep technol-ogist in continuous attendance. The dream abnor-malities with dream-enacting behaviors found withRBD and with sleepwalking, sleep terrors will becovered in the following sub-sections. I will nowdiscuss the other three disorders that can manifestwith dream-enactment.

1) OSA-induced “pseudo-RBD” was first de-scribed in 1996 in a report on 5 cases of OSA-induced arousals from REM sleep, with dream-related complex and violent behaviors [3]. All 5cases were males, 51–69 years old. A 69-year-oldman presented with a history of excessive daytimesleepiness (EDS) and dream-related behaviors,such as turning the pages of a newspaper or screw-ing in a light bulb. A 68-year-old man with a histo-ry of loud snoring and EDS had several episodes of hitting his wife in bed during attempted dreamenactment while dreaming that he was in a fight.A 67-year-old man had a history of EDS and ab-normal nocturnal behaviors in which he wouldshout, grab and injure his wife while dreaming ofprior military experiences and of sporting events.A 52-year-old man had a progressive disorder ofEDS and nocturnal dream enactment, when hewould throw out punches and kick repeatedly whiledreaming that dogs were biting his leg. A 51-year-old man would fall out of bed while asleep, and put his fist through a window.The PSGs of all 5 pa-tients were diagnostic of moderately severe tosevere OSA, but not for RBD (i.e. there was com-plete integrity of the customary REM sleep muscleatonia that prohibited any behavioral releaseduring REM sleep). The “apnea-hypopnea index”

ranged from 51 to 196 per hour, oxygen saturationnadirs ranged from 68 to 82%, and sleep efficiencyranged from 30 to 79%.The 3 patients treated withnasal continuous positive airway pressure therapy(nCPAP) all had resolution of their pseudo-RBDwith control of OSA. Thus, in these cases of OSA,obstructive apnea-induced arousals from REMsleep with vivid dreaming presumably resulted inimmediate, post-arousal, dream-enacting behav-iors with locomotion, agitation and violence thatresolved with nCPAP treatment. This state couldalso be called an OSA-induced “hypnopompicREM sleep parasomnia” that closely resemblesactual RBD behaviors.

In a recent publication entitled, “Severe ob-structive sleep apnea/hypopnea mimicking REMsleep behavior disorder” [4], 16 patients (11 men,5 women, mean age, 59.6 � 7.7 years) presentedwith dream-enacting behaviors and unpleasantdreams (e.g. being attacked or chased by an un-known person or animal, falling down abruptly,arguing with someone) that were highly suggestiveof RBD – but all these patients also had snoring andEDS.Abnormal sleep behaviors emerged through-out the night with a frequency that ranged fromonce every two weeks to several times a night. Thesleep behaviors witnessed by the bed partnersincluded gesturing, punching, kicking, and otheraggressive behaviors. Two patients had assaultedtheir spouses, five had fallen out of bed, and twohad lacerated their face and arms during dream enactment. Whereas video-PSG in the sleep labexcluded RBD and documented the presence ofnormal REM sleep atonia and without REM sleepphasic motor overactivity, it was diagnostic ofsevere obstructive sleep apnea-hypopnea, with amean hourly apnea-hypopnea index of 67.5 � 18.7(range, 41–105), and all abnormal sleep behaviorsoccurred during apnea-hypopnea induced arousalsfrom REM and NREM sleep. Nasal CPAP therapyeliminated the abnormal behaviors, unpleasantdreams, snoring and EDS in all 13 patients who had agreed to undergo this therapy. Repeat video-PSG of these 13 patients while they were on nCPAP therapy documented the complete reso-lution of OSA while again documenting the ab-sence of any “REM sleep without atonia” or frankRBD. Therefore, OSA-induced “pseudo-RBD”was definitively proven in this study.

2) Nocturnal seizures can manifest as recurrentabnormal dreams, nightmares, and dream-enactingbehaviors – including vigorous and violent behav-iors. One dramatic example involved a 65-year-oldman who was reported to have the sudden onset offrequent nightmares and dream-related behaviorsprecipitated by a right temporal lobe infarction [5].


In a typical episode, “he would suddenly boltupright, pace around with a terrified expression onhis face, and shout in a dysarthric voice … At timeshe related frightful visions (men coming into hisroom), and occasionally would give vivid details”.PSG showed that no episode emerged from REMsleep, but typically from stages 2 and 3 NREMsleep, “with approximately 20 seconds of EEGevidence of awakening while the patient remainedbehaviorally asleep. The episode of abrupt move-ments and shouting then obscured the EEGrecord”. Anti-convulsant therapy with diphenyl-hydantoin of these stroke-induced,complex partialseizures produced complete remission of all symp-toms, including nightmares and dream-relatedbehaviors.

Another report described a 58-year-old manwith a 6-year history of vivid dreaming asso-ciated with ambulation and screaming; one time he awakened while striking the molding around his bedroom window [6]. PSG revealed “frequentrepetitive polyspike discharges … in REM sleepaccompanied by generalized clonic activity lastingup to 20 seconds. Frequent brief myoclonic jerkswere also noted during REM sleep … After thelongest episode he mumbled and made purposefulmovements … Otherwise REM atonia was main-tained”.

A third report described a 16-year-old male who experienced recurrent purposeful movements,cursing, and “auto-aggressive behavior” duringsleep, with some subsequent recall of these episodes; PSG revealed slow spike-and-waveactivity arising from REM sleep [7]. Anti-con-vulsant therapy with carbamazepine completelyresolved his sleep and presumably dream-relatedproblems.

3) Nocturnal (sleep-related) dissociative disor-ders comprise a unique, female-predominant cate-gory of “psychiatric parasomnia”, insofar as thisbehavioral-experiential nocturnal disorder is adirect manifestation of a psychiatric disorder (i.e.dissociative disorder) that emerges during sus-tained EEG wakefulness after an awakening from sleep or just prior to the onset of sleep [1, 8].Prolonged (or shorter-lasting) complex behaviorsin bed, or elaborate nocturnal wanderings canoccur, which at times include driving a car, or evenboarding an airplane (a prime example of a “fuguestate”). The events occur with negligible self-awareness and no subsequent recall. Behaviors at times are “sexualized” (e.g. repetitive pelvicthrusting) and can be paired with defensive be-havior and vocalisations. Perceived “dreaming”and dream-enacting behaviors at times occur, withthe “dreams” typically representing memories of

past sexual and physical abuse. These memoriesthat are experienced as “dreams” occur duringaltered states of EEG wakeful consciousness,during drowsiness or after an awakening fromsleep. The abuse-related memory retrieval that isexperienced as a dream resides at the core of thedissociative process, since the emotional distressassociated with reliving psychological trauma isunconsciously lessened by embedding the memoryin a dream-like or a frank dream state while oneremains awake in an altered state of consciousness.On the other hand, perceived dreaming duringnocturnal dissociative episodes may not have anyobvious link with past abuse, as found in the casedescribed below that was contained in our sleepcenter’s initial report on this newly described para-somnia [8].

Case: Exclusively animalistic sleep-related (nocturnal) dissociative disorderA 19-year-old male, accompanied by his adoptiveparents, presented with a 4-year history of elabo-rate episodes emerging from the nocturnal sleepperiod approximately 1–2 times weekly, when hewould act like a large jungle cat. These episodestypically began 1–2 hours after falling asleep, whenhe would leave his bed while growling, hissing,crawling, leaping about, and biting objects, for aslong as one hour. He then would collapse abruptlyon the floor, perspiring profusely, and be com-pletely unresponsive.Although always subsequent-ly amnestic for his actions during the night, themorning after an episode he would invariably re-call a particular recurrent “dream” of being a lionor tiger let out of his cage by a woman zookeeperwhom he then followed down a path. She held a“piece of raw meat” in her hand, but he would never be able to leap up to snatch it “because of an invisible force field” holding him back, whichalways made him feel “disappointed and frustrat-ed”. These “dreams” always ended with “someoneshooting a tranquilizer gun at me” and then hewould fall down and become unconscious – whichis when his “dream” merged with reality, as he layon the floor in a room in his home, completely unresponsive.

He considered the “dream” action, which closely mirrored his actual behaviors (for which he was ostensibly amnestic), to be “very vivid andreal”, and expressed interest in recreating thedream during his waking life by making a movie ofit.Although the “dream” plot was always identical,the face of the woman zookeeper varied fromepisode to episode and was never a familiar face of someone in his life. He emphasized that duringthis “dream” he felt completely like an animal and


had no awareness of actually being human, nor was he ever aware of interacting with members of his household while moving around during his “dreams”. (In fact, his parents and often also his sister attempted to intervene in each of hisepisodes.)

His family had never heard him speak nor inter-act verbally with them during these spells, and henever acted in any sexual manner. His postures andbehavioral repertoire (including loud, eerie growl-ing and an impressive transformation of his handsinto facsimiles of paws) were consistently those ofa large cat, as documented by home videotaping.He successfully navigated the environment, aswhen he would open the refrigerator with hismouth, put uncooked bacon between his teeth (that presumably corresponded with the piece of“raw meat” in his “dream”), and then proceed toprowl around the house.

His family commented on his recurrent feats of“super-human strength”, such as leaping far fromhis bed, lifting a mattress with his jaws and drag-ging it across a room, or even lifting a marble tablewith his jaws. He frequently left imprints of histeeth on the wood furniture during these nocturnalepisodes, but never left the house, nor did he sit orstand up: his movements were constantly quadri-pedal. He remained unresponsive to people duringand immediately after these spells, despite havinghis family talk,shout,shake him or splash cold wateron his face. Sometimes he demonstrated opposi-tional behavior, such as when he actively avoidedhis mother or father or resisted his mother’s at-tempts to wipe his perspiring brow,or when he triedto jerk a towel away that she was holding in hismouth to prevent him from cracking his teeth whenhe bit the furniture. Whenever his father held himupright, by standing behind him during an episode,he would engage in prolonged, continuous, quadri-pedal locomotion, as was demonstrated on a homevideotape.(This is “classic”for activation of “Loco-motor Pattern Generators” in the brainstem ofmammals [9],which can be experimentally inducedthrough electrical or specific chemical stimulationof that lower brain region.)

He never lost control of his bowel or bladderduring these episodes, and he never demonstratedrepetitive “tonic-clonic” movements or stereo-typies suggestive of an epileptic attack.

He had injured his lips and gums on numerousoccasions from biting sharp objects, and had alsobruised and lacerated himself all over his bodyduring these nocturnal episodes.

The onset of his nocturnal sleep-related dis-order had no recognized precipitant, nor did anyparticular episode have an identifiable trigger.

Numerous neurological examinations and wakingEEGs had been normal. He was never observed tohave a daytime dissociative spell.

This patient was generally regarded to be per-sonable and well-behaved,and he had several closefriends. He never demonstrated strange or objec-tionable daytime behaviors. There was no historyof physical, sexual,or verbal-emotional abuse sincethe time he was adopted at the age of 10 months.During his interviews with me, he was polite and friendly, did not appear psychologically “dis-turbed”, and did not exhibit any peculiar manner-isms or ways of acting.

He was studied overnight in our sleep lab fortwo consecutive nights. On the first night, two of his characteristic episodes were recorded,and eacharose from clear-cut EEG wakefulness. The firstepisode occurred 53 minutes after he had fallenasleep, during which time he had cycled through all4 stages of NREM sleep. Beginning with a sponta-neous EEG awakening from Stage 1 NREM sleep(that had been present for several minutes), hegrowled intermittently for 2 minutes while he re-mained motionless in bed with eyes closed andbehaviorally appeared to be asleep. This beganseveral minutes after an epoch of Stage 1 NREMsleep had terminated in sustained EEG wake-fulness – and with his remaining in bed and stillappearing asleep.He then abruptly left the bed andcrawled around the room, hissing, growling, loudlygrinding his teeth and pulling the mattress with his jaws. He also chewed and swallowed portionsof the airflow monitoring device.After 6.5 minutes,he abruptly collapsed on the floor while his handscontinued to be contorted as if they were paws.After another 2 minutes he was up again, crawlingand growling for 4.5 minutes before a final collapseand 30 seconds of clinical unresponsiveness. Nor-mal wakeful EEG activity continued throughoutthis entire event. He then consciously emergedfrom (or “snapped out” of) this spell and describedhis typical “dream” in full detail, but was amnesticfor all his actions.

The second episode emerged 1 hour 15 minutesafter the first episode (the patient having againcycled through NREM and REM sleep). DuringEEG wakefulness, he growled for 15 seconds before crawling out of bed, knocking a lamp over,biting that lamp, chewing and swallowing por-tions of another airflow monitoring device andrepeatedly banging his head against a wall. Theentire episode lasted over 9 minutes and stoppedabruptly with a collapse; 20 minutes later he be-came responsive and reported “the same dreamagain, nothing different”. A wakeful EEG wasmaintained throughout this second episode,and no


seizure-like EEG activity was evident during eitherof the two episodes.

Therefore, both episodes in the sleep lab thatemerged during well-established wakefulness werevirtually exact replicas of what was recorded athome on videotape.

This 19-year-old male was the first patient everdocumented, both clinically and by PSG monitor-ing, to have an exclusively (nocturnal) sleep-re-lated dissociative disorder, and also an exclusivelyanimalistic dissociative disorder. In fact, he wasdiagnosed to have a nocturnal Multiple Personal-ity Disorder, since he fulfilled official psychiatricdiagnostic criteria that included: (i) the existencewithin a person of two or more distinct persona-lities or personality states (each with its own rela-tively enduring pattern of perceiving, relating to,and thinking about the environment and self); (ii)at least two of these personalities or personalitystates recurrently take full control of the person’sbehavior.

The videotaped behaviors of this young man,both during the two episodes in the sleep lab and during the episode recorded at home, clearlyindicate that during these lengthy episodes thebehaviors were exclusively animalistic, whichwould seem to indicate that the mode of perceivingand interacting with the environment was also ex-clusively animalistic.The mechanism(s) underlyingan exclusively animalistic and sleep-related disso-ciative disorder are unknown,especially since therewas no known history of sexual or physical abuse.

He was offered treatment with hypnotherapy,with the goal being for him to learn and utilize aself-hypnosis technique at bedtime, as a means of gaining more control over himself during thesleep period while he was in the process of fallingasleep. However, the patient and his family werenot interested in another form of “psychologicaltherapy”. (Obviously, the content of his “dream”involving a female zookeeper holding a piece of“raw meat” that he could not snatch from herbecause of an “invisible force field”, which culmi-nated with “someone shooting a tranquilizer gun atme”, could easily be subjected to psychodynamicinterpretation.)

The ICSD-2 has officially added sleep-relateddissociative disorders as a recognized parasomnia[1], with the essential features being that they canemerge throughout the sleep period during well-established EEG wakefulness, either at the transi-tion from wakefulness to sleep or within severalminutes after an awakening from stages 1 or 2NREM sleep or from REM sleep. Sleep-relateddissociative disorders comprise a sleep-relatedvariant of dissociative disorders, which are defined

in the DSM-IV as “… a disruption in the usuallyintegrated functions of consciousness, memory,identity, or perception of the environment” [10].The general similarity of the behaviors found with sleep-related dissociative disorders to thebehaviors found with various parasomnias justifiestheir inclusion within the parasomnias section ofICSD-2 and indicates how they comprise a distinctsleep-related variant of dissociative disorders.

III REM sleep behavior disorder

REM sleep behavior disorder (RBD) is charac-terized by abnormal behaviors emerging during REM sleep that cause injury or sleep disruption.RBD is also associated with electromyographic(EMG) abnormalities during REM sleep that serveas the neurophysiologic substrate for allowingRBD behavioral release. The EMG demonstratesan excess of muscle tone or phasic EMG twitch activity during REM sleep [1, 2]. A complaint of sleep-related injury is common with RBD, whichusually manifests as an attempted enactment ofdistinctly altered, unpleasant, action-filled, and violent dreams in which the individual is beingconfronted, attacked, or chased by unfamiliarpeople or animals. Typically, at the end of anepisode, the individual awakens quickly and be-comes rapidly alert (as is common for all awaken-ings from REM sleep), and reports a dream with acoherent story, with the dream action correspond-ing to any sleep behaviors that are observed by a bed partner, roommate, or sleep laboratory tech-nician.

An important aspect of RBD worth empha-sizing is that it is a dream disorder almost as muchas it is a sleep behavior disorder, since approxima-tely 90% of reported patients complain of abnor-mal dreaming and dream-enacting behaviors thatbegan with the onset of RBD [2]. Furthermore,successful treatment of RBD virtually always involves the tandem control of the abnormaldreaming and the abnormal (dream-enacting)sleep behaviors, and relapse of these core featuresof RBD also occurs in tandem whenever the pa-tient fails to take clonazepam at bedtime (i.e. thestandard therapy of RBD). Thus, the dream andsleep behavioral disturbances of RBD appear toshare a common pathophysiology, the implicationsof which will be discussed later in this section. It is also of interest to note, vis-à-vis the Freudiantheory of dreams, that release from the shackles of REM atonia with the consequent ability to act-out dreams has unveiled a realm of dream-enact-ment that appears to be devoid of sexuality [2],


although definitive research on this topic has notyet been conducted, and also the dreams in RBDpatients (apart from dream-enactment) have alsonot yet been systematically examined.

Sleep and dream-related behaviors in RBD re-ported by history and documented during PSGmonitoring include talking, laughing, shouting,swearing, gesturing, reaching, grabbing, arm flail-ing, slapping, punching, kicking, sitting up, leapingfrom bed,crawling,and running.Walking,however,is quite uncommon with RBD, and leaving theroom is especially rare and probably occurs bychance, with the individual running through anopen doorway. The eyes usually remain closedduring an RBD episode, with the person attendingto the dream action and not to the actual environ-ment; this is a major reason for the high rate ofinjury in RBD. Also, chewing, feeding, drinking,sexual behaviors, urination, and defecation havenot been documented to occur in REM sleep – norin the REM sleep associated dreaming duringdream-enactment. The findings in human RBD inregards to the categories of behaviors observed in REM sleep, and categories of behaviors not observed in REM sleep, mirror the findings froman experimental animal model of RBD producedby pontine tegmental lesions [2].

Because RBD occurs during REM sleep, it usu-ally appears at least 90 minutes after sleep onsetunless there is coexisting narcolepsy, in which caseRBD can emerge shortly after sleep onset duringa sleep-onset REM period. RBD is usually a long-standing, progressive condition.

The chronic form of RBD is male-predominantand typically emerges after the age of 50 years,although any age group can be affected. Our center has reported that 65% of men �50 years of age who were initially diagnosed with “idio-pathic” RBD eventually developed a parkinsoniandisorder, with a mean interval of 13 years from the onset of the RBD to the onset of the parkinso-nian disorder, and with a range extending from 2 to29 years [11, 12]. Furthermore, in some of thesepatients, the characteristic dream disorder of RBDemerged months before the behavioral disorder of RBD, leading to the observation that a changein (dream) consciousness can be the first symptomof parkinsonism, the classic signs of which may notemerge for more than a decade.

An increasingly recognized precipitating factorof RBD in males and females of any age is psycho-tropic medication use, particularly venlafaxine,selective serotonin reuptake inhibitors, mirta-zapine, and other antidepressant agents – with theexception of bupropion (a dopaminergic anti-depressant) [13].

Diagnostic criteria for RBD are listed below,with time-synchronized audio-video-PSG moni-toring being essential for establishing the diagno-sis of RBD [1]:A. Presence of REM sleep without atonia: the

EMG finding of excessive amounts of sustainedor intermittent elevation of submental EMGtone or excessive phasic submental or (upper orlower) limb EMG twitching.

B. At least one of the following is present:i Sleep-related injurious, potentially injuri-

ous, or disruptive behaviors by historyii Abnormal REM sleep behaviors document-

ed during polysomnographic monitoringC. Absence of EEG epileptiform activity during

REM sleep unless RBD can be clearly distin-guished from any concurrent REM sleep-relat-ed seizure disorder.

D. The sleep disturbance is not better explained byanother sleep disorder, medical or neurologicaldisorder, mental disorder, medication use, orsubstance use disorder.

Patients with RBD are typically described ashaving calm and pleasant personalities, and do not display irritability or anger while being awake.Also, these people did not suddenly become thisway with the onset of their RBD, but always have had this personality style, as verified by theirspouses of 25–50 years and by first-degree familymembers. They have been successfully employedand have enjoyed the company of loving familiesand good friends. There is no evidence that theyhave been building up repressed anger or aggres-sion over the course of their lives, and then in late-life ultimately release these negative feelingsduring their REM sleep. The strong association of RBD with Parkinson’s disease and other braindisorders also downplays any proposed “repressedaggression” theory of RBD.There thus seems to bean intriguing set of brain-mind-personality-para-somnia interactions with RBD.An important studythat has recently been published explores this topicin a systematic manner [14].

In this study, dream characteristics were sys-tematically assessed in RBD patients and con-trols, along daytime aggressiveness tendencies,by having them recall their most recent dreams and by having them complete the Aggression Ques-tionnaire (AQ).Verbatim dream descriptions werescored and analysed according to the Hall and VanDe Castle method. The results showed that RBDpatients had a higher percentage of dreams with atleast one aggressive episode than controls (66 vs.15%, p <0.00001), a higher aggression/friendlinessinteraction ratio (86 vs. 44%, p <0.0001), and agreater frequency of animal characters (19 vs. 4%,


p = 0.0001). On the other hand, compared to con-trols, none of the RBD patients had dreams withsexual elements (0 vs. 9%, p <0.0001). In regards toaggressive tendencies during wakefulness, the twogroups did not differ in total AQ scores, except for a lower score on the physical aggressivenesssubscale in RBD patients compared to controls(16.5 � 6.4 vs. 20.4 � 8.3, p = 0.034). There was no correlation between dream aggressiveness andeither age, duration or frequency of RBD symp-toms. Therefore, despite normal levels of daytimeaggressiveness, patients with RBD had dreamscharacterized by an increased proportion of ag-gressive elements. In their discussion, the authorspointed out previous research indicating that chil-dren have higher rates of aggression and higherrates of animals in their dreams compared to adults,presumably related to the development of threat-avoidance skills. Given the authors’ current find-ings on dreams in RBD, “it may be hypothesizedthat chronic RBD, as a part of a widespread neuro-degenerative process, would lead to a release ofontogenetically early dream patterns” [14]. Never-theless, the validity of this conclusion could bequestioned on several fronts, such as the quality ofaggression in the dreams of children vis-à-vis RBDpatients could differ substantially. In a related vein,as elucidated by Corner in the Afterword to arecent book by this author [15], the phasic EMGabnormalities in human RBD appear to representthe pathological release in later life of the mostprimitive form of motor activity found in the onto-genetic development of organisms throughout the animal kingdom. This motor activity in earlydevelopment is subsequently suppressed to a largeextent during REM sleep with the adaptive emer-gence of motor inhibitory systems [16, 17].

In 1977 Hobson and McCarley published aseminal and provocative paper entitled,“The brainas a dream state generator: an activation-synthesishypothesis of the dream process” [18]. A core fea-ture of that hypothesis is that during REM sleepthere is ongoing activation of the cerebral cortexby ascending activating signals from the brainstemthat allow the cortex to formulate coherent dreams.What is relevant about this dream hypothesis inregards to RBD is that in both the experimentalanimal model and in the chronic human disorderthere is brainstem pathology and the activation ofspecific categories of dream-enacting (or “sleephallucinatory”) behaviors along with the absenceof other categories of dream-enacting behaviors.(In the experimental animal model, the site andextent of the brainstem lesion correlated with thecategory of the expressed REM sleep behavioralrepertoire that emerged as presumed “hallucina-

tory”, i.e. dream-related, behaviors [2].) Further-more, with human RBD, there is activation ofparticular categories of dreams along with suppres-sion of other categories of dreams (e.g. physicallyactive and aggressive dreams predominate, where-as sexual dreams are suppressed). Hobson hasrecently refined and upgraded this “Activation-Synthesis” dream model with the “AIM” model(activation, input, modulation), whereby “brain-mind space” across sleep and wakefulness is de-fined along three axes: level of central nervoussystem activation, sensory input (endogenous vs.exogenous), and neurotransmitter modulation(aminergic vs. cholinergic). Mahowald has pro-vided an excellent overview of Hobson’s AIMmodel in the context of understanding conscious-ness across sleep and wakefulness [19], with an in-depth consideration of Edelman and Tononi’s“dynamic core hypothesis” of the neural substratesof consciousness [20,21].(The latter hypothesis ana-lyses RBD as a behavioral and experiential statethat does not originate in the external world [21].)

Finally, this author (CHS) has written a bookthat contains 26 first-hand accounts of RBDdream-enactment provided by patients and theirspouses, in order to allow a more broad appre-ciation of the phenomenology of the recurrent,altered, nocturnal sleep and dream experiencesfound with RBD [15]. The following are someexamples of RBD dream-enactment contained inthis book:

One dream when I kicked my wife was a scene frommy childhood when I rode a bicycle and dogs chasedme, biting at my legs, and I kicked one dog, but I wasreally kicking my wife. In another dream I was back-packing, and somebody had stolen some of my gearand I ran up to them and kicked them as hard as Icould. They were above me and I was kicking up.What I actually did was kick the wall and I brokemy big toe. It was black-and-blue and swollen andhurt badly.

The dreams are always basically the same sort oftype, and could be illustrated by the dream I had theother night: I was in a car, I was parking in a ramp,and the car started to move backwards, as if it wereabout to go down the ramp. I jumped out of the carto try to stop its movement, and as I became awakeand oriented, I realized that I had jumped out of bedand I was pushing against the bed. That is a verytypical kind of dream for me. Sometimes if it’s not a car, then it’s a train. Usually, I have a sense ofaccelerated motion on an inclined plane and I willdo something to prevent what seems to me to bedangerous.


He has done this when it wasn’t a violent dream.Onenight I awakened and the bed covers were just goingup and down, up and down, so I woke him up and I said, ‘What in the world are you doing?’ And hestarted to laugh, and he laughed and laughed and I said, ‘What is so funny?’ He said, ‘I’m riding abicycle.’ So what he was actually doing was non-violent – he was just riding a bicycle, just pedaling.Another time he was sitting up in bed, pushing outwith his arms, and I awakened him and he said hewas playing volleyball. And then he was stompinghis feet up and down on the bed, stomping his feet.He said he was killing a cockroach; it was about sixinches long.

Once when I hit her, I had a dream that somebodywas on the other side of a screen door from me andwanted me to hit him. Every time I’d go to hit him,he’d move his face first, so I’d keep trying to hit theface on the other side of the screen door.A week agowe were playing hockey in my dream, and I wasgoing to body-check two guys at once,and I thought,‘Well, I’m going to have to throw a cross-bodyblock,’ so I did that and ended up on the floor. I don’tknow how, but I hit my arm on the nightstand and Ihit my head on something else. I bruised myself upquite badly.

I thought I was with a childhood buddy,and we werestanding on the wing of an airplane. Both he and Iliked airplanes a lot when we were young. Anotherairplane came right beside us and started movingahead and we were standing on the wing, but lowerthan the wing on the other airplane, and the wingswere going to overlap and knock us off. I rememberhollering, ‘Get the hell out of here,’ and we just dovehead-first off the wing, but what I actually did wasdive head-first over the end of the bed with the blan-kets, pillows – everything going with me. There wasa table sitting there and I just scraped the edge of thattable with the side of my head and ear. It woke meup immediately and I thought, ‘Oh my God, if Iwould have hit that table with my forehead, I wouldhave either caved it in or I’d have broken my neck.’

The dreams that I really remember are when thereare spiders all over me.They are golden spiders, andI’m knocking them off. When I’ve gone camping, Iwould be in a sleeping bag that was zipped up whenI fell asleep, and then I would wake up after I bit my way through the sleeping bag, while dreamingthat animals were after me – but no animal was ac-tually around. I have jumped far from the bed intoa dresser table with a mirror on top during dreamsof tigers coming after me.

IV Sleepwalking, sleep terrors and “dreamwalking” in adults

Classic descriptions of sleepwalking and sleep ter-rors mention that episodes occur early in the night,there is virtually no self-awareness during theepisodes, there is often no associated dreaming(and when present,consists of brief fragments with-out plot), and there is either no subsequent recallor else just limited recall of events from the pre-ceding night. While these observations may holdtrue for most children and adolescents with thedisorders of arousal from NREM sleep, the situa-tion can be – and often is – quite different in adultswith sleepwalking and sleep terrors. In fact, asubstantial number of the 25 adult patients withsleepwalking, sleep terrors whose first-hand ac-counts are shared in my previously mentionedbook on RBD and other parasomnias [15],describevivid, and at times lengthy and plot-driven, dream-enacting behaviors during their episodes, and haveconsiderable recall of these events the next day.Also, some felt completely awake during theirepisodes – until they were awakened from sleep. Inaddition, these episodes not infrequently occurredduring the middle stages – and at times even dur-ing the later stages – of sleep, and not just withinthe first hour or two (when slow-wave sleep pre-dominates), as emphasized in the “classic” descrip-tions.

“Dreamwalking”Dreaming while walking about (“dreamwalking”)is rarely found with RBD – and particularly whengoing from one room into another room, which hasnever been reported in RBD (in fact, with RBD,going into another room occurs by chance, whenthe affected person happens to run through anopen doorway rather than into a wall or furniture).Instead, when “dreamwalking” is present, it isusually found with sleepwalking in adults (alsodescribed as “somnambulism associated with hal-lucinations” by Kavey and Whyte [22]). The mis-conception of “dreamwalking” being ipso facto amanifestation of RBD has recently surfaced in a publication entitled, “An early description ofREM sleep behavior disorder” [23]. The authorsreported finding an early description of RBD bythe celebrated French writer named AnthelmeBrillat Savarin who presented the case of a “som-nambulist” monk in his book, Physiologie du Gout[Physiology of Taste] that was published in 1825.

Quoting from the original text by Savarin, theauthors of the journal article presented the follow-ing “very singular fact told me [i.e.Savarin] by DomDuhaget, once prior of the Chartreuse convent”:


There was a monk … who was looked upon as asomnambulist. He used often to leave his cell, andwhen he went astray, people were forced to guidehim back again. Many attempts had been made tocure him, but in vain. One evening I had not gone tobed at the usual hour, but was in my office … whenI saw this monk enter in a perfect state of somnam-bulism. His eyes were open but fixed and … he hada huge knife in his hand. He came at once to my bed,the position of which he was familiar with, and afterhaving felt my hand, struck three blows which pen-etrated the mattress on which I lay … I saw anexpression of extreme gratification pervaded hisface. The light of two lamps on my desk made noimpression, and he returned as he had come, open-ing the doors which led to his cell, and I soon be-came satisfied that he had quietly gone to bed … Onthe next day I sent for the somnambulist and askedhim what he had dreamed of during the precedingnight … ‘Father’, said he,‘I had scarcely gone to sleepwhen I dreamed that you had killed my mother,and when her bloody shadow appeared to demandvengeance, I hurried into your cell, and as I thoughtstabbed you.Not long after I arose … I thanked Godthat I had not committed the crime I had meditated.’I then told him what had passed, and pointed out tohim the blows he had aimed at me …

A number of arguments can be presented in sup-port of sleepwalking rather than RBD as the basisfor the parasomnia episode just described.

1) The monk was a known somnambulist, whoon the night in question opened his cell door twiceduring the same parasomnia episode, once to leavehis cell at the beginning of the episode and then toreenter his cell at the end of the episode. Somnam-bulists open doors, but never RBD patients, whorarely have their eyes open, as they are attendingto their dream world with their eyes shut. In fact,the prior observed the monk to have his eyes openand “fixed” without any reactivity to the light fromtwo lamps on the prior’s desk: this is a classic ob-servation for sleepwalking. The rare RBD patientwho has his eyes open during an episode wouldprobably awaken when exposed to the light of alamp.

2) The monk knew exactly where he was going,based on the horrifying dream that he had justexperienced immediately prior to leaving his cell.In fact, his purpose in sleepwalking was to go to theprior’s cell (and he also knew where the bed waslocated in that cell) to avenge the murder of hismother perpetrated by the prior. Sleepwalkers canengage in such purposeful behavior when there isassociated dreaming, but not a person with RBD,who would throw punches in bed or charge out of

bed and run into a wall or furniture while goingafter a murderer in his or her dream. The problemwith sleepwalking is that while a person can nego-tiate the actual environment, and open doors andgo from one room to another (or leave the houseand even drive an automobile), there is a majorsuspension of critical thinking and judgment, withautomatic behavior that may or may not be asso-ciated with dreaming. If dreaming is present, thecontents of the dream are fully accepted as fact,andthe affected person proceeds accordingly, which inthe monk’s case meant that he had to avenge themurder of his mother, and he knew exactly whereto go, and he went there, i.e. the prior’s room. Suchbehavior is very inconsistent with RBD behavior.

3) The episode occurred very shortly after themonk fell asleep, which is far more indicative of aNREM sleep parasomnia (which can emerge assoon as 15 minutes after sleep onset) than of RBD(which typically emerges at least 90 minutes aftersleep onset, when the first REM sleep period ap-pears, unless narcolepsy – with sleep-onset REMperiods – is also present, but there was no evidencesuggestive of narcolepsy in the description of themonk).

(A full critique of this case has recently beenpublished [24].)

I will now share some case vignettes of vividdream-enacting behaviors, including plot-drivenand dreamwalking scenarios, in patients with sleep-walking, sleep terrors described in my aforemen-tioned book [15].The first case involved a “positivestress” (associated with Christmas Eve) promotingan episode in a woman with lifelong sleepwalking.The second case involved enormous negative stresspromoting an episode in a man with lifelong sleep-walking, sleep terrors. The third case also involvedenormous negative stress promoting the de novoonset of sleep terrors. What is so intriguing aboutthe third case is that instead of developing sleeponset and/or sleep maintenance insomnia in theface of severe stress, as would be expected, this 34-year-old man developed a de novo parasomniamanifesting as sleep terrors, sleepwalking.

A 25-year-old married woman, with lifelongsleepwalking, while asleep on Christmas Eve hadan elaborate dream about Christmas, and in thatdream she placed a doll under the Christmas treeas a gift for her one-year-old daughter – but onChristmas morning she awakened to discover thatduring the night she had actually carried her infantdaughter from her crib and placed her unscathedunder the tree.

A 38-year-old man with lifelong sleepwalking,sleep terrors describes the following episode thatoccurred 16 days after the sudden death of his


brother and 14 days after his wife separated fromhim:

I start having a very vivid, terrifying, horrible dreamof a huge stockinged-face man, about 350 pounds,and six-and-a-half or seven feet tall, holding a hugemachete in his hand which he is bringing up, thenswinging it down on me. All of a sudden – out ofnowhere – I was outside somewhere in a field, it was very dark, I could barely see the outlines of thetrees, but the moon was so bright, I don’t know if itwas a full moon, but it was so bright. So when he islifting and about to swing the machete full force atme, I am up looking around and see what is aboutto happen (while actually standing up in the doublebed – but in my mind, in my dream, I am out in afield). I run away full-bore in the dark – and hit thewall so hard that an extension cord gets ripped offthe plug, and only the metal prongs stay in the wallsocket. Just before I hit the wall, or simultaneouslywith it, I felt the pain of the machete cutting me inthe back – in the exact same place where the rippedelectrical extension cord cut me. It was a one-inchdeep gouge off my skin that ran 8 inches down frommy shoulder blade to my lower back. I came within6 inches of hitting the corner of a dresser, whichwould have then smashed my head in the templearea and I probably would have died. I almostpushed the wall all the way to the outside. There was a hole in the wall from my shoulder blade areadown to my lower back, where I had smashed into it.

A 34-year-old man developed acute-onset,persistent sleepwalking, sleep terrors triggered by major stress at work:

I had injured my back at work in January, her-niating a disk, and the assistant administrator at the nursing home where I worked began doingeverything she could do to get me fired.What she didwas very obvious to the other department heads andsome of the staff. The assistant administrator really“turned up the heat” in the spring and the nightterrors started in May.Some of the situations at workwere reenacted in the dreams, so it was easy to seehow they related. In the first dreams, it was often thepeople with blank eyes, or some other trait, whichtipped me off to the fact they were the ones whowould harm me when the opportunity arose. Some-times I would be walking down the street and rea-lize I was being shadowed by one or more of theseindividuals. I would try to escape, they would attack,and I would be screaming when I woke up awayfrom my bed.

In another dream I was walking down an openplank staircase, already aware someone wanted to

attack me. When I looked back over my shoulder,one of the individuals with blank eyes was hidingdirectly behind the stairs and burst out through thestairway at me. I awoke shrieking and trying to runaway.Many nights I would end up smashing into thewall directly at the foot of the bed, indicating that Ihad hurled myself straight out over the foot of thebed. Other times I would fly out of the side of thebed and seem to crash horizontally against the wall.

The worst dream I ever had occurred when wewere camping. I dreamed I was walking through anolder Western-style town and there were numeroushead of cattle in the street. I had been forewarned tobe on the lookout as one of the cows would try toattack me. I looked behind and a large cow wasapproaching. She opened her mouth and instead ofa tongue, a large knife came out of her mouth withwhich she tried to stab me. I could not run away,I was actually in a sleeping bag, so I tried to fight.I was screaming,‘What are you,who are you?’ WhenI woke up, I could not stop screaming and cryinghysterically, even though I was fully awake. My wifekept trying to calm me but to no avail. My sleepingbag was soaked with sweat though it was only about 40 degrees that night. It must have taken well overan hour to calm me. It was not difficult to interpretthat dream. Shortly before I was told to resign frommy position at the nursing home, effective threeweeks prior to our wedding, a good friend hadwarned me that the director of nursing, who weigh-ed about 250 pounds, would stab me in the back if she had the chance.

Since that time,I’ve had a really bad dream aboutonce a month. Sometimes I’m being attacked by aperson or some other being which I can fight, whileother times the dream is more like an unfoldingimpersonal plot which I have to foil.One night whenwe had company, my wife and I were sleeping on the floor in my den when a portion of the baseboardflipped down and two slender missiles were aimedat us through the opening. The looked like the tail-lights on a ’59 Cadillac. I knew they were about to befired at us, so I whispered to my wife, ‘Get up, we’vegot to go.’ I was trying to drag her out of the room asshe was stumbling to her feet. When I turned on thelight, I was aware it had just been another dream.

Many times the dreams involve the juxtapositionof lights. In these dreams I am aware of a situationwhen at some point the moving lights would line upor form a certain pattern and a bomb had been set to go off at that time. Sometimes I would dreamI was walking and the lights would appear on myclothing or on the side of a building and I would bescreaming as I tried to run away.

One night I had a dream in which there were twosmall, flashing red lights, which I knew would cause


a bomb to go off when they moved together. I sawthem start to move closer together, so I tried to getout of the bed and run. I awoke immediately, how-ever,and realized the lights were actually the flashingred column on my digital alarm. I lay down again togo back to sleep, but as a safeguard I put my extrapillow over the clock just in case it really was a bomb.Two or three nights ago, I dreamed I was in a factory,working at a conveyor belt which carried suitcases.The situation was that when this suitcase decoratedwith flowers on it came around, it was going to be a bomb. A few minutes later here it comes down theline. I grabbed it and tried to get away from the areabefore it blew up.When I got down the hall and intothe living room, the light from the lamp awakenedme, and my wife was asking me what I was doing. Itwas then I realized the ‘bomb’ was just a pillow.

V Epic dream disorder

Beginning in 1986, as our sleep center was becom-ing more familiar with the characteristic dream dis-turbances of RBD, and as we were also becomingincreasingly aware that adults with sleepwalkingand sleep terrors could have precipitous and some-times elaborate dreaming during their recurrentNREM sleep episodes, we also started seeingpatients who presented with a distinct set of dream-related complaints that we eventually named “epicdream disorder”. In 1995 we published findingsfrom our initial series of 20 patients in an abstractentitled,“A disorder of epic dreaming with daytimefatigue, usually without polysomnographic abnor-malities that predominantly affects women” [25].I will now summarize the salient findings from thisseries, present two clinical vignettes, and describesubsequent reports from three other centers in two additional countries on this intriguing, newlyrecognized dream disorder.

In the introduction to our 1995 report,we statedthat “nightmares” were officially regarded at thetime as frightening REM sleep dreams, with sleepterrors and RBD serving as the two main differen-tial diagnoses for disturbing dreams. We then pro-posed another category of disordered dreaming –identified during routine clinical practise at oursleep disorders center – characterized by relentless,“epic” dreaming that was experienced to occurthroughout sleep, leaving the affected individualfeeling exhausted upon arising from sleep in themorning, followed by ongoing fatigue during thedaytime.

Twenty patients with epic dream disorder weregathered over a 9-year period during routineclinical practice at our multi-disciplinary sleep dis-

orders center. All patients but one were evaluatedby this author (CHS), to whom all new patients withdream and/or parasomnia complaints at that timewere preferentially referred. Patients completed astandard, comprehensive sleep center question-naire, and were clinically interviewed. PSG studiesusing standard recording and scoring methods with expanded EEG and EMG monitoring werescheduled, since we wanted to investigate whetherthe complaint of epic dreaming was associated with a motor-behavioral parasomnia, nocturnal seizures, or another type of prominent sleep-dis-ruptive disorder that could account for the com-plaint of non-restorative sleep in the morning. Amultiple sleep latency test (MSLT) was conductedthe day after the overnight, hospital-based PSGstudies.Patients also completed psychometric tests.

This was a female-predominant clinical group,with 85% (17/20) being female. Mean age at refer-ral was 34.8 (� SD 11.2) years. Mean duration ofthe epic dream disorder was 3.6 (� 6.7) years in 13 patients, with 7 patients reporting a longstand-ing dream disorder having an indeterminate dura-tion. Almost all the patients were employed andmore than half were married.

Epic dreaming was the primary complaint in70% (14/20) of patients. Epic dreaming withoutnightmares occurred in 30% (6/20) of patients;70% (14/20) complained of both nightmares and of relentless, neutral-content dreaming, such asendlessly walking through snow or mud, or end-lessly working on the job or at home with house-hold chores, without attendant emotions, apartfrom the sensation of exhaustion. The dream dis-turbance shared by all patients was the perceptionof endless dreaming with constant physical activitywithin the dreams. Patients described having a“dream motor running all night long”, or “nothaving the mind shut down during the night”, orexperiencing intense acceleration and spinning intheir dreams. Nightmares were rarely related toany real-life situation.

In 35% (7/20) of patients, epic dream disorderhad an abrupt onset that was either spontaneous orlinked to the onset of benign positional vertigo,multiple sclerosis (MS), rheumatoid arthritis, orsevere, acute stress. In addition, a male patient hadpreviously been reported to suffer from “patho-logic lucid dreaming” [26] (to be described below)that evolved to become an epic dream disorder.Epic dreaming occurred nightly in 90% (18/20) of patients, and 4 nights per week in 10% (2/20) ofpatients. Caffeine and alcohol consumption wasnegligible or moderate in nearly all patients.

PSG studies, completed in 80% (16/20) of pa-tients, were interpreted as being clinically unre-


markable in 75% (12/16); n = 2 had periodic limbmovement disorder (PLMD) and n = 2 had sleep-disordered breathing. REM sleep was not disor-dered in any patient, with REM sleep latencyalways �64 minutes, and with REM sleep per centof total sleep time of being 20–25% in 13/16 pa-tients (range, 14–31%), apart from one patienttaking amitriptyline who had 0% REM sleep.Patients had dream recall after awakenings fromboth NREM and REM sleep, although dream re-call was non-systematically elicited by the atten-dant sleep technologist during spontaneous awak-enings. Neither sleep terrors, RBD, or narcolepsywas detected by the PSG and MSLT studies.MSLTs, completed in 75% (15/20) of patients, werenormal, with a mean sleep latency during eachMSLT of >6 min, without REM sleep.

A history of psychiatric disorder (e.g. dysthy-mia;major depression,anxiety) was present in 35%(7/20) of patients, but there was no direct associa-tion with epic dreaming, apart from one schizo-phrenic patient whose epic dream disorder was adirect extension of her daytime delusional symp-tomatology. Treatment was generally ineffective,despite a broad range of interventions (cognitive-behavioral, hypnosis, relaxation and pharmaco-logic therapies).

To summarize the three core complaints relatedto epic dreaming: first, the perception of relentless,neutral-content dreaming – without emotionality –throughout the night; second, the dream charactersengage in constant physical activity throughout the dreams; third, the feeling of exhaustion uponarising in the morning, with subsequent daytime fatigue. Our conclusion was that epic (i.e. continu-ous, relentless) dreaming with daytime fatigue is achronic, idiopathic, female-predominant, and rela-tively treatment-resistant disorder without charac-teristic PSG correlates.We also considered that ourdata supported findings from various dream inves-tigators that sustained dreaming can occur in bothNREM and REM sleep, a topic that has subse-quently received increased formal attention [27].

At this point, it is worth reviewing the case of pathological lucid dreaming that had evolvedinto epic dreaming in three distinct stages: fromuncomplicated lucid dreaming, to problematiclucid dreaming, to epic dream disorder. The firstdescription of this case encompassed the first twostages in a report entitled, “A case of pathologicallucid dreaming presenting to a sleep disorders cen-ter” [26].

Lucid dreaming – formally discovered byLaBerge – is a unique state of coexisting wakefuland dream consciousness in which a person notonly is aware of dreaming while dreaming, but

he or she can communicate with others throughvolitional movements of the eyes and fingersduring REM sleep, as verified by PSG studies insleep labs [28, 29]. Emotional arousal usuallyaccompanies lucidity:“even the most prosaic luciddreams tend to begin with an unmistakable senseof excitement and delight” ([30], p. 118). Problem-atic lucid dreams have been reported, e.g. “piouslucid dreams were very frequently followed bywhat [van Eeden] called ‘demon-dreams’, in whichhe was typically mocked, harassed, and attacked bywhat he supposed to be ‘intelligent beings of a verylow moral order’.” ([30], p. 175) It is unfortunatethat van Eeden “was never able to free himselffrom his demon-dreams, and his efforts to rid him-self of them met resistance throughout his life”([30], p. 176). In contrast, Saint-Denys was able to overcome lucid dreams of “abominable mon-sters” through wilful resolve in the lucid state ([30],p. 176–9).

To our knowledge, we were unaware of any pre-vious report involving a PSG and clinical sleepevaluation of problematic lucid dreaming [26].

A 28-year-old single, male theatre manager hadlifelong lucid dreaming that became problematic 2 years prior to referral, without identified pre-cipitant. He reported that “I’ve always been ableto dream what I wanted to dream about – I couldplay 18 holes of golf by myself or with anybodyelse”. He would enter lucidity only when flying ina dream,which is a common entry into lucidity [30].He had considered the lucid state to be “a lot of fun … I put a big smile on my face in the dream …I always looked forward to sleep”. He had up to 3 lucid dreams nightly, and often would be chasedby others – but was always able to fly away.

Two years before referral, “things got out ofcontrol in my dreams, and I wasn’t having fun any-more”. He could no longer escape from peoplechasing him, shooting at him or “beating me up – itwas terrible, and I would wake up feeling that I hadbeen hit by a truck, with severe headaches, musclepains, and exhaustion”. However, medical andneurologic examinations and extensive diagnostictesting (including EEGs looking for epileptic ac-tivity) were unremarkable.

Treatment with psychotherapy and with varioushypnotic medications were ineffective. However,the anti-convulsant diphenylhydantoin, which wasprescribed on the suspicion by his physician thatatypical epilepsy was causing his dream disorder,“was a wonder drug for me, I had no more disturb-ing dreams and awakened feeling great”. Of noteis that both normal and pathological lucid dream-ing were suppressed by diphenylhydantoin.When-ever he discontinued this medication, there was full


relapse of abnormal and normal lucid dreamingwithin four days, which remitted promptly with resumption of the medication. However, fourmonths before referral to our center, a sustained,partial relapse of his abnormal lucid dreamingoccurred, with the emergence of dreams in whichhe would fall off buildings, and be “beaten to apulp”.He also experienced uncontrollable sobbing,i.e. intense negative emotionality, during theselucid dreams.

This patient underwent an extensive clinical andPSG evaluation at our center, which included asearch for epileptiform activity in the scalp EEG.However, these studies were unremarkable and his sleep was objectively normal. Epilepsy was notdiagnosed. From a psychiatric perspective, therewas no history of major depression, anxiety dis-order, or any other condition that required treat-ment with psychotherapy or medication, and hehad never been hospitalized. His clinical interviewwas unremarkable. The only abnormality was de-tected by the Minnesota Multiphasic PersonalityInventory (MMPI), which suggested tendencies for paranoia, impulsivity, and “somatizing” (i.e.channeling psychologic distress into physical symp-toms). These findings may relate to his disturbingdreams that involved being chased, shot at, andbeaten up,which are consistent with paranoia.Also,his experience of extensive physical symptomsupon awakening is consistent with somatizing. It isunfortunate that he had not previously completedan MMPI, since it would have been of interest tocompare the current MMPI findings with thosefrom an MMPI completed prior to the onset of hisdream disturbance two years beforehand, in orderto address the issue of whether the current MMPIfindings reflected a longstanding predisposingfactor for his dream disturbance, or instead were (in part) a consequence of his persistent dream dis-turbance.

When his diphenylhydantoin dose was in-creased, and his blood level reached the therapeu-tic range (for treating epilepsy), he then once againachieved control of his abnormal lucid dreaming.This benefit was maintained during twice-yearlyclinic visits for more than 3 years.

We quoted from Hunt that “nightmares andlucidity are linked along the dimension of affectenhancement – the one is capable of producing anecstasy of terror, the other of bliss. Common toboth is a vivid intensification of the dreamingprocess and an enhancement of kinesthetic sen-sation … Intensification of the dreaming processwould thus tend toward either a maximum inte-gration of experience (lucidity …) or maximum dis-integration (nightmares …)” ([31], p. 124). This

statement may possibly have applied to our pa-tient, in whom pathologic lucidity developed forunknown reasons.

The response to an anti-convulsant medicationeither suggested an underlying epileptic contribu-tion to his abnormal lucidity, since cases of complexpartial seizure disorder (viz. “temporal lobe epi-lepsy”) have been reported in which the primary or even sole clinical manifestation of the epilepsywas abnormal dreaming that was controlled withanticonvulsant medication (in such cases, the“epileptic equivalent” was the abnormal dream-ing), or may have been consistent with a non-epileptic basis for the abnormal lucidity, sincevarious non-epileptic clinical disorders are knownto respond to anticonvulsant medication.

Ultimately, the patient developed epic dream-ing, without any associated lucidity, which per- sisted after discontinuation of diphenylhydantointherapy, but he refused to consider any therapy forhis epic dream disturbance.

Epic dreaming – clinical vignettesI) A 60-year-old widowed woman with multiplesclerosis [32] and no psychiatric history reported“working hard in my dreams, all night long. Beforemultiple sclerosis I would dream, but I wouldn’twork in my dreams”.The following comments weredocumented during an audiotaped recording of ourclinical interview, with the patient’s permission:

I’ve had lots of physical activity in my dreams sincehaving MS: walking in hard snow, walking in mud,cleaning walls, scrubbing floors, cleaning up closets– I am always cleaning closets.Sometimes I will havelost something, and I am looking and looking andlooking and looking for it in my dreams. I work inmy dreams, and wake up so tired from them. BeforeMS, I would dream, but I wouldn’t work in mydreams. I would dream, but my dreams were alwaysfun, and I never got tired from them. They werealways little dreams that you forgot about. [Now] Inever get the work done, it is just too hard, workingand working and working, but I never get the jobdone.The closets are so messy that I can never get tothe bottom of them. The walls are so big, I never getdone painting them.There is never an end to the mudI am walking in, I just never get done walking. Fortwo years, there was a faceless man always helpingme. I was in a room that I had to clean, and it wasdeep with junk all over. I did not know where to start,so I was cleaning and cleaning, and here came thisfaceless man, and he was with me for about twoyears in my dreams, helping me clean. Isn’t thatfunny?


II) A 34-year-old man (encountered after publica-tion of our initial series of 20 epic dreaming pa-tients) reported on the dream consequences fromdrinking a great excess of espresso coffee one nightafter dinner: “It was so much work for me. I neverstopped moving in my dreams.” The following com-ments were written down during our discussion:

After downing six small cups of espresso to end a fine dinner, I still managed to fall asleep with accustomed speed, but then experienced continuousdreaming throughout the entire night, with thedream characters – including myself – being inperpetual motion. It was so much work for me, Inever stopped moving in my dreams, and the nextmorning I awoke with my body feeling completelyexhausted.

The man’s wife confirmed that he did not movearound in bed that night while he was continuouslydreaming. Nor did he have nightmares in the usualsense of the word. Rather, he just kept dreamingand dreaming in relentless fashion, with non-stopmovement in his dreams of himself and the otherdream characters, and without associated emo-tionality.

In this man, a night of epic dreaming was in-duced by caffeine excess shortly before he fellasleep. There have been reports of RBD being induced or aggravated by caffeine excess, includ-ing coffee and chocolate over-consumption, as reviewed [13]. However, RBD is characterized byboth a dream disturbance and sleep behavioraldisturbance. Furthermore, the dream disturbanceof epic dreaming is quite different from the abnor-mal dreaming in RBD, in which the affected per-son is being threatened or attacked by unfamiliarpeople or animals in a nightmarish scenario repletewith emotionality. It is at present not clear whycaffeine excess may induce epic dreaming in somepeople and RBD in others – or no apparent dreamor sleep behavioral dysfunction.

The second report on epic dreaming was a case inCanada described in 1996 by the dream researchersZadra and Nielsen [33]. A 35-year-old woman without psychiatric history had presented to theauthors’ dream and nightmare center with a longstanding complaint of epic dreaming. Herexperience of “relentless dreaming” was reportedto occur 4 to 5 times a week. On two occasions (13 years previously and earlier during the year ofpresentation) the woman had consulted her familyphysician with the complaint of “dreaming all nightlong” with subsequent daytime fatigue. The med-ical evaluation revealed no physical problems.

Anti-anxiety medication had failed to improve herepic dreaming.

The subject reported that the thematic contentof her epic dreams typically involved performingjob-related work or running great distances. Incontrast, nightmares occurred approximately onceevery 3 to 4 weeks and were described as beingeither anxiety-producing chase and pursuit dreamsor else extremely sad dreams from which she wouldawaken crying.The subject’s husband reported thatshe would frequently talk in her sleep and that thesleeptalking was often clear and elaborate. He alsoconfirmed that she frequently had vivid dreamswithin 10–30 minutes of falling to sleep and in the morning she would invariably recall long anddetailed dreams.

The patient was monitored for two consecutivenights in the sleep laboratory.PSG findings showedno clinical abnormalities. None of the dreamsreported in the sleep lab involved the themes thattypically characterized her dreams at home. Sleep-talking occurred on two occasions,both out of stage3 NREM sleep, and consisted of a few loud butunintelligible words.Therefore, epic dreaming wasabsent from the dreams collected during two PSGstudies. However, this woman did not report expe-riencing epic dreaming on a nightly basis at home,and only two PSG studies were conducted. Theauthors concluded that home dream diaries andthat 3–5 or more consecutive PSG studies in thesleep laboratory might provide a more fruitfulapproach for better understanding the patterns ofdream recall and dream content associated with theclinical complaint of epic dreaming.

The final two reports on epic dream disorderoriginated in Taiwan, allowing for an Asian per-spective to compare with the first two reports from North America.The first of these was entitled,“Polysomnographic and clinical correlates of epicdream complaints: a retrospective study of 28 Tai-wanese adults” [34].

Over a 1.5-year period,28 healthy adult patientsreferred themselves to the first author’s sleep clinicin Taiwan with the dual chief complaints of constantdreaming throughout sleep on most nights, butwithout increased physical activity of the dreamcharacters; and also daytime fatigue, but withoutexcessive daytime sleepiness. The evaluation of all28 patients was conducted by one author, a board-certified neurologist with specialty training in sleepdisorders medicine, whose sleep clinic routinelyevaluated a full range of adult sleep complaints.A clinical evaluation was followed by an overnightPSG study of all patients in the hospital sleep lab.A sleep lab technologist questioned each patient inthe morning about any dreams during the preced-


ing PSG study, but did not attempt to elicit dreamrecall during spontaneous nocturnal awakenings.PSG arousals were not counted, and psychometrictests were not administered.The data in this reportwere analysed in a retrospective manner.

This group of 28 epic dreamers was 61% male(n = 17) and 39% female (n = 11), with a mean ageof 39.0 � SD 10.3 years (males 41.5 � 10.5; females35.2 � 9.1 years; range, 25–63 years). The meanduration of epic dreaming was nearly 8 years, butthere was a very broad range of 1–30 years for themales and 1–16 years for the females. This series of 28 epic dreaming patients comprised approxi-mately 13% of all new patients presenting to thatsleep clinic during the same time period, of which55% were male.

Two groups of epic dreamers could be distin-guished on the basis of their dream content. The“story” epic dreamers formed a group of 5 patients(4 male) who reported that their constant dream-ing was like a never-ending television “soap opera”,which would immediately resume after an awaken-ing. The dreamer usually experienced emotionalneutrality during these dreams, with the contentinvolving a cohesive, linear story (with emotionsdisplayed by the dream characters, but not felt bythe actual dreamer who was the observer in thedream).The story involved activities on the part offamily, friends, office staff at work, etc.“Story” epicdreaming occurred most nights in all 5 patients.

The second group comprised 23 patients with“non-story” epic dreaming, who also perceiveddreaming throughout sleep nearly every night, butthe dream content did not involve a story with aplot.This group reported dreams with random, dis-connected content, devoid of emotionality, whichnever formed a discernible story. After returning to sleep after a nocturnal awakening, a differentdream would immediately begin, in contrast to the “story” epic dreamers, who would resume analready established, ongoing dream.

Both groups of epic dreamers commonly com-plained that their brains and minds were beingoverworked during sleep, which they believedmade them feel fatigued in the morning. Never-theless, there was no complaint of increased phys-ical activity of the dream characters within thedreams (although the characters did move around),nor was there any complaint of bizarre or night-marish dream content. Although a few patientsreported a history of abrupt-onset epic dreaming,there was no apparent association with stress, amedical or psychiatric disorder, although psycho-logical testing was not performed nor was a psy-chiatric interview conducted. During nocturnalawakenings at home, patients in both groups near-

ly always reported having just been dreaming. The“story” epic dreaming in 5 patients (4 males) beganon average around the age of 23 years, whereas the“non-story”epic dreaming in 23 patients (13 males)began on average around the age of 33 years.

Most of the patients (78.6% [22/28]) had clini-cally unremarkable PSG studies, with less than aquarter (21.4% [6/28]) having sleep fragmentationdisorders consisting of either moderately severeobstructive sleep apnea (OSA) during REM sleep(n = 5), or severe Periodic Limb Movement Dis-order (PLMD) with a high arousal index (n = 1).There was also a major shift towards light sleep inall patients, with increased percentages of stages 1and 2 NREM sleep, decreased percentages of bothstages 3/4 NREM sleep and of REM sleep, and amodestly increased number of awakenings. (How-ever, this could have represented a “first-night”adaptation effect in the sleep lab.) Nevertheless,total sleep time and sleep efficiency were unre-markable in all sub-groups of patients, and REM-latency was not reduced in any sub-group.

An MSLT was not performed on any patient,given the lack of any history of daytime hyper-somnolence. Most of the patients (22/28) had a bedpartner, but there was never a bed partner com-plaint of problematic sleep-related behaviors orsleep-talking.The morning after their PSG studies,all patients confirmed having had their usual epicdreaming throughout the preceding night.

Therefore, this series of 28 Taiwanese patientswith epic dreaming was male predominant, in con-trast to the previously reported series of 20 epicdreamers from our sleep center in the UnitedStates which was 85% female. The two series hadcomparable mean ages at presentation (35 vs. 39years), with mildly divergent mean ages of epicdreaming onset (27 vs. 35 years). Another dif-ference between the Taiwanese and American epicdream groups concerns the complaint of increasedphysical activity of the dream characters, which was a universal finding in the American group, buta non-existent finding in the Taiwanese group.Whereas patients in the American group attributedtheir daytime fatigue to having increased physicalactivity during their relentless dreaming, patientsin the Taiwanese group attributed their fatigue tohaving their brains and minds overworked duringtheir relentless dreaming. The basis for this dif-ference in the subjective attribution of fatigue is un-known,but may be related to differences in culture,sex, daytime activity level, and other factors.

This series of Taiwanese epic dreaming patientswas further divided into three groups: first, the“story” epic dream group, with onset during ado-lescence and early 20s, and normal PSG findings;


second, the “non-story” epic dream group, with asomewhat older age of onset,and normal PSG find-ings; third, a “non-story” epic dream group, with aconsiderably older age of onset,and abnormal PSGfindings consisting of sleep fragmentation duringREM or NREM sleep on account of OSA orPLMD. By “normal PSG findings” we refer to clin-ically unremarkable findings, since there was con-siderable suppression of stages 3/4 and REM sleep,which may have been a “first-night effect” or per-haps a PSG correlate of epic dreaming.

The striking divergence between the universalcomplaint of relentless dreaming throughout sleepand the reduced per cent of REM sleep for theentire group should be noted. It is now known thatdreaming does not occur exclusively during REMsleep; in fact, �50% of awakenings from NREMsleep are accompanied by recall of cognitive activ-ity or frank dreaming [27]. Therefore, epic dream-ing could be a disorder of exaggerated NREMdreaming which, when paired with REM sleepdreaming, would result in the perception of relent-less dreaming throughout the night.

Two questions raised by these data concernwhether the prominent lightening of sleep (shifttowards stages 1 and 2 NREM sleep) found innearly all 28 patients could have enhanced therecall of dreams, and whether the excessive awak-enings from REM and NREM sleep in the sixpatients with OSA or PLMD also could have en-hanced the recall of dreams.

An active debate exists as to whether there isone dream generator or two dream generators (oneeach for NREM and REM sleep) [27]. Further-more, the question of “phantom REM sleep”, withphysiologic components of REM sleep (i.e. incom-pletely declared REM sleep) intruding into NREMsleep as a dream-promotor, has been raised re-garding the genesis of dreaming in NREM sleep[27]. Therefore, epic dreaming may result fromexcessive activity of dream generators, includingexcessive “phantom REM” intrusions into NREMsleep.The presence of both “story”and “non-story”epic dreamers suggests a differential activation of complex dream production systems that areresponsible for organizing story-like plots withindreams [35].

It is currently unknown whether these patientswere, in fact, dreaming more-or-less continuallythroughout the night and across all sleep stages,since that retrospective study did not test the rela-tionship between dream recall and sleep stageduring nocturnal awakenings in the sleep lab. Abasic question regarding epic dreaming, therefore,is whether dreaming actually occurs during most of sleep (including slow-wave NREM sleep [36]),

or whether dreaming is misperceived to occurthroughout most of sleep. In other words, epicdreaming could be a form of “dream state misper-ception” analogous to “sleep state misperception”[37–39] in which the affected person incorrectlyestimates the amount of PSG-documented sleepobtained (i.e. major discrepancy between subjec-tive and objective sleep). This brings up the issueof the “Mental Status of Sleep” already proposedearlier in this section. In regards to epic dreaming,a perceptual disturbance may be linked withdreaming and recall (i.e. memory) of the dreaming,so that a heightened perception of dreaming maysomehow result in the complaint of epic dreaming.This group of 28 epic dreamers may thus be aheterogenous group consisting of bona fide epicdreamers (with etiologic heterogeneity) and thosewith “dream state misperception”. Nevertheless,irrespective of the presence of bona fide epicdreaming or of dream state misperception, there isa nearly inextricable link between the complaint ofrelentless dreaming and morning/daytime fatigue.Methodologic limitations of this retrospectivestudy included the lack of systematic dreamcollection, the lack of correlating dream recall withREM vs. NREM sleep stages, and the absence of a psychiatric interview and psychologic testing.

The fourth report on epic dream disorder waspublished in 2005 from a different Taiwanese cen-ter [40]. Twenty-two patients (14 females, 8 males;mean age 34 years, range 18–61 years) who com-plained of epic dreaming were recruited from aneurology clinic. A psychiatric interview and theHealth Personality and Habit (HPH) Scale (self-rating) were administered. One night PSG wasconducted, presumably in-hospital, and the re-ported findings were generally unremarkable,with a 13.1% wake time after sleep onset pre-sumably being predominantly a “first-night effect”.The slow-wave NREM sleep per cent was mildlyreduced (9.9%), perhaps being another “first-nighteffect”.Four patients had a respiratory disturbanceindex of �5/hour, and another patient had a PLMindex of 18/hour, but there was no comment on anyassociated arousals. Five patients had alpha EEGintrusions in NREM sleep, a finding without anyknown specific clinical correlate. Three additionalpatients had a spontaneous arousal index through-out sleep of �15/hour, which may also have beento some extent a “first-night effect”.Approximate-ly two-thirds of the patients demonstrated a ten-dency for psychopathology on the HPH scale (an-xiety, depression, personality disorder), and one-third of the patients were clinically diagnosed witheither an anxiety or depressive disorder. Althoughthe authors concluded that epic dreaming appeared


to emerge in the context of increased arousals from“heterogeneous pathology”, it is unclear to whatextent their data supported that conclusion. Fur-thermore, the preponderance of patients with in-creased arousals from sleep across the spectrum of sleep disorders do not report epic dreaming, soother predisposing factors must also be present.

Therefore, epic dreaming has now been re-ported in 71 patients encompassing two racialgroups (Asian and Caucasian) residing in Taiwanand North America, with the North Americangroup being female-predominant (85.7% [18/21]),and the Taiwanese group being gender-neutral(50% [25/50]). The mean age at presentation wasremarkably similar in all four publications (34–39years), and in most cases epic dreaming had beenpresent for years, if not many years, beforehand,indicating that epic dreaming is usually a long-standing, persistent condition, with an underlyingsleep disorder uncommonly being identified, andwhen present not being severe. Medical and psy-chiatric disorders likewise rarely play a contri-buting role. Phenomenological subtypes of epicdreaming have been suggested, such as “story vs.non-story”epic dreaming;or epic dreaming with orwithout increased physical activity of the dreamcharacters; etc. Successful therapy to date has notbeen identified.Thus, epic dream disorder remainsan intriguing and enigmatic condition that has not yet been sufficiently described in the medicalliterature to be included as a diagnostic entity in thecurrent revision of the International Classificationof Sleep Disorders (ICSD-2).

Prospective research on epic dreaming, withsystematic clinical evaluations, PSG monitoringand dream diaries and analyses are thus encour-aged across racial, ethnic, sex, and age groups, andalso across various medical, psychiatric, and sleepdisorder groups. There should be quantification ofdream data, with utilization of physical activityscales from a scaling system adapted for languagedifferences. Also, the basis for continuous and dis-continuous dream stories would need to be statedand quantified, so that the level of continuity or dis-continuity is clearly understood.

References

1 ICSD-2 – International Classification of Sleep Disorders,2nd ed.: Diagnostic and coding manual. Westchester, Illinois: American Academy of Sleep Medicine; 2005.

2 Schenck CH, Mahowald MW. REM sleep behavior disorder:clinical, developmental, and neuroscience perspectives16 years after its formal identification in SLEEP.Sleep 2002;25:120–38.

3 Nalamalapu U, Goldberg R, DiPhillipo M, Fry JM. Behaviors simulating REM behavior disorder in patientswith severe obstructive sleep apnea. Sleep Res 1996;25:311.

4 Iranzo A, Santamaria J. Severe obstructive sleep apnea/hypopnea mimicking REM sleep behavior disorder. Sleep 2005;28:203–6.

5 Boller F, Wright DG, Cavalieri R, Mitsumoto H. Paroxysmal “nightmares”: sequel of a stroke responsiveto diphenylhydantoin. Neurology 1975;25:1026–8.

6 Louden MB, Morehead MA, Schmidt HS. Polyspikes inREM sleep mimicking REM sleep behavior disorder. SleepRes 1994;23:283.

7 Silvestri R, De Domenico P, Musolino R, Mento G,Marabello L, Longo M, et al. Nocturnal complex partialseizures precipitated by REM sleep. Eur Neurol 1989;29:80–5.

8 Schenck C, Milner D, Hurwitz T, Bundlie S, Mahowald M.Dissociative disorders presenting as somnambulism:video and clinical documentation (8 cases). Dissociation 1989;2:194–204.

9 Skinner RD, Garcia-Rill E. Brainstem modulation of rhythmic functions and behaviors. In: Klemm WR,Vertes RP, editors. Brainstem Mechanisms of Behavior.New York: Wiley-Interscience; 1990. p. 465–96.

10 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.

11 Schenck CH, Bundlie SR, Mahowald MW. Delayed emergence of a parkinsonian disorder in 38% of 29 oldermales initially diagnosed with idiopathic REM Sleep behavior disorder. Neurology 1996;46:388–93.

12 Schenck CH, Bundlie SR, Mahowald MW. REM behaviordisorder (RBD): delayed emergence of parkinsonismand/or dementia in 65% of older men initially diagnosedwith idiopathic RBD, and an analysis of the minimum & maximum tonic and/or phasic electromyographic abnormalities found during REM sleep. Sleep 2003;26:A316.

13 Mahowald MW, Schenck CH. REM Sleep Parasomnias. In:Kryger MH, Roth T, Dement WC, editors. Principles andPractice of Sleep Medicine. 4th ed. Philadelphia, Pennsyl-vania: Elsevier Saunders; 2005. p. 897–916.

14 Fantini ML, Corona A, Clerici S, Ferini-Strambi L. Increased aggressive dream content without increaseddaytime aggressiveness in REM sleep behavior disorder.Neurology 2005;65:1010–5.

15 Schenck CH. Paradox Lost: Midnight in the Battlegroundof Sleep and Dreams. Minneapolis, Minnesota: Extreme-Nights, LLC (www.parasomnias-rbd.com); 2005.

16 Corner MA. Ontogeny of brain sleep mechanisms. In:McGinty DJ, Drucker-Colin R, Morrison A, Parmeggiani PL,editors. Brain Mechanisms of Sleep. New York: Raven Press; 1985. p. 175–97.

17 Corner MA, Van Pelt J, Wolters PS, Baker RE, Nuytinck R.Physiological effects of sustained blockade of excitatorysynaptic transmission on spontaneously active developing neuronal networks – an inquiry into the reciprocal linkage between intrinsic biorhythms and neuroplasticity in early ontogeny. Neurosci Biobehav Rev 2002;26:127–85.

18 Hobson JA, McCarley RW. The brain as a dream stategenerator: an activation-synthesis hypothesis of thedream process. Am J Psychiatry 1977;134:1335–48.


19 Mahowald MW. Understanding consciousness: implications and consequences. Perspect Biol Med 2003;46:297–304.

20 Tononi G, Edelman GM. Consciousness and complexity.Science 1998;282:1846–51.

21 Edelman GM, Tononi G. A Universe of Consciousness.How Matter Becomes Imagination. New York: Basic Books; 2000.

22 Kavey NB, Whyte J. Somnambulism associated with hallucinations. Psychosomatics 1993;34:86–90.

23 Schulz P, Curtin F. An early description of REM sleep behavior disorder. Sleep 2004;27:1216–7.

24 Pressman MR, Mahowald MW, Schenck CH. Sleep terrors/sleepwalking – not REM behavior disorder.Sleep 2005;28:278–9.

25 Schenck CH, Mahowald MW. A disorder of epic dreamingwith daytime fatigue, usually without polysomnographicabnormalities that predominantly affects women. Sleep Res 1995;24:137.

26 Schenck CH, Mahowald MW. A case of pathological luciddreaming presenting to a sleep disorders center. Sleep Res 1990;19:155.

27 Nielsen TA. Mentation during sleep: the NREM/REM distinction. In: Lydic R, Baghdoyan HA, editors. Handbookof Behavioral State Control: Cellular and Molecular Mech-anisms. Boca Raton: CRC Press; 1999. p. 101–28.

28 LaBerge SP, Nagel LE, Dement WC, Zarcone VP. Lucid dreaming verified by volitional communication during REM sleep. Percep Mot Skills 1981;52:727–32.

29 Fenwick P, Schatzman M, Worlsey A, Adams J, Stone S,Baker A. Lucid dreaming: correspondence betweendreamed and actual events in one subject during REMsleep. Biol Psychiatry 1984;18:243–52.

30 LaBerge S. Lucid Dreaming. New York: Ballantine Books; 1985.

31 Hunt HT. The Multiplicity of Dreams. New Haven, CT: Yale University Press; 1989.

32 Schenck CH, Slater GE, Sherman R, Anderson DC,Ettinger MG, Mahowald MW. Multiple sclerosis and sleep:survey report and polygraphic detection of REM andNREM motor abnormalities. Sleep Res 1986;15:163.

33 Zadra AL, Nielsen TA. Epic dreaming: a case report.Sleep Res 1996;25:148.

34 Yeh S-B, Schenck CH. Polysomnographic and clinical correlates of epic dream complaints: a retrospectivestudy of 28 Taiwanese adults. Tzu Chi Medical Journal [Taiwan] 2000;12:159–64.

35 Cipolli C, Bolzani R, Tuozzi G. Story-like organization of dream experience in different periods of REM sleep. J Sleep Res 1998;7:13–9.

36 Cavallero C, Cicogna P, Natale V, Occhionero M, Zito A.Slow wave sleep dreaming. Sleep 1992;15:562–6.

37 Salin-Pascual RJ, Roehrs TA, Merlotti LA, Zorick F, Roth T.Long-term study of the sleep of insomnia patients withsleep state misperception and other insomnia patients.Am J Psychiatry 1992;149:904–8.

38 Bonnet MH, Arand DL. Physiological activation in patientswith sleep state misperception. Psychosom Med 1997;59:533–40.

39 Mendelson WB. Effects of time of night and sleep stage on perception of sleep in subjects with sleep statemisperception. Psychobiology 1998;26:73–8.

40 Shih-Wei L, Chien-Ming Y, Hsiao-Hsui L, Ming-Ceh T.Polysomnographic and psychiatric features in patientscomplain of epic dreaming. Sleep 2005;28:A53–4.

41 Mahowald MW, Schenck CH. Dissociated states of wakefulness and sleep. Neurology 1992;42:44–52.

42 Mahowald MW, Schenck CH. Nocturnal dissociation –awake? asleep? both? or neither? Sleep Hypnosis 2001;3:129–30.

43 Mahowald MW, Schenck CH. Evolving concepts of humanstate dissociation. Arch Ital Biol 2001;139:269–300.

44 Mahowald, MW, Woods S, Schenck CH. Sleeping dreams,waking hallucinations, and the central nervous system.Dreaming 1998;8:89–102.

45 Cochen V, Arnulf I, Demeret S, Neulat ML, Gourlet V,Drouot X, et al. Vivid dreams, hallucinations, psychosisand REM sleep in Guillain-Barré syndrome. Brain 2005;128:2535–45.


Documents

Dreaming: a psychiatric view and insights from the …...gested that REM sleep, and associated dreaming, may perform vital functions. Studies on the child-hood development of dreaming