somnambulism associated with hallucinations
TRANSCRIPT
Case Reports
Somnambulism Associated With Hallucinations
NEaB. KAVEY, M.D.JAMIE WHYTE, M.D.
Sleepwalking can be symptomatic ofa varietyof underlying conditions, including classic
non-rapid eye movement (REM) somnambulisml
•2 with or without night terrors, REM behav
ior disorder,3 and complex-partial seizures.4•s In
these various disorders the behaviors performed,associated mentation, responsiveness to the environment, intranight timing of events, and morning recall vary according to etiology, althoughoverlap does exist. However, because the heterogeneity of sleep-related ambulation with partialarousal has only recently been elucidated, a tendency to consider sleepwalking a homogeneousphenomenon persists.
Until the latter half of this century, the linkbetween somnambulism and mental imagery wasbelieved to be the dream state (i.e., the sleepwalker was thought to be acting upon the mentalcontent present in an ongoing dream). Researchin the early 19605, however, repudiated this ideawhen it was found that virtually all episodes ofsomnambulism and night terrors occurred in deepnon-REM sleep (Stages III and IV). Sleepwalking was considered at that time to be a "disorderofarousal't6 in which the physiological defect layin the neural regulation of generalized corticalactivation. Although the relationship betweensomnambulism and night terrors has never been
Received August 6, 1990; revised June 17, 1991; accepted June 25, 1991. From The Sleep Disorders Center,Columbia-Presbyterian Medical Center, New York, NY; Department of Psychiatry, Columbia University College ofPhysicians and Surgeons, New York, NY. Address reprintrequests to Dr. Kavey, The Sleep Disorders Center, Columbia-Presbyterian Medical Center, 161 Fon Washington Ave.,New York, NY 10032.
Copyright © 1993 The Academy of PsychosomaticMedicine.
86
satisfactorily defmed, they are widely concededto represent points along a continuum of severitywith an identical underlying cause. More recently, a new entity, termed 'the REM behaviordisorder, has been recognized in which grossmotor activity occurs in REM sleep, with overtdream enactment a clear component in at leastsome cases. This relatively uncommon disorderis caused by a failure of neural machinery thatactively suppresses postural muscle tone in normal REM sleep and bears little epidemiological,pathophysiological, or polysomnographic similarity to non-REM sleepwalking.
In differentiating among somnambulism,night terrors, REM behavior disorder, and nocturnal seizures, it was thought that mentationassociated with the behavioral abnormalities isclinically useful. Thus, fragmentary or absentrecall is most consistent with classic somnambulism or seizures, a single vivid image is stronglysuggestive ofnight terror, and a detailed narrativeladen with visual imagery is indicative of REMbehavior disorder. Several recent studies, however, have cast doubt upon this simple view,reporting an increasingly wide range of mentation in association with parasomnias as thephenomena are more intensively investigated.The mentation associated with episodic paraphenomena in sleep is important not only clinically but may hold important clues as to etiology.In particular, non-REM sleepwalking is postulated to have multifactorial causation, with organic and psychogenic factors contributing tovarying degrees in different patients.7
We recently evaluated two patients whocomplained of sleepwalking with associatedmental activity that was subtly but clearly different from mentation in known sleepwalking dis-
PSYCHOSOMATICS
Case Reports
orders. These patients appeared to sleepwalk inconjunction with hallucinatory episodes, a phenomenon that to our knowledge has not beenpreviously described. Known types of visual distortions related to sleep include illusions, a common occurrence in normal subjects with noknown psychopathology; imagery ofREM sleep,also a normal mental event; fleeting REM-likeimagery at the transition between sleep andwakefulness; and hypnagogic and hypnopompichallucinations, which are actually the visual manifestations ofREM intruding into a state ofpartialwakefulness. These latter events may be sporadicand benign or a component of narcolepsy but, ineither case, tend to be associated with behavioralquiescence or frank postural atonia.
Thus, these cases are of interest on two accounts: first, they extend the range of mentationnoted to occur in association with gross motoractivity in sleep and second, they may provideinsight into the etiology and pathophysiology ofthese incompletely understood disorders.
Case Reports
Case 1. Patient A was a 34-year-old female account executive for an advertising agency. She wasreferred by her psychotherapist, who was treatingher for anxiety and mild depression and for the complaint of sleep-related hallucinations often accompanied by gross motor activity or ambulation. Theepisodes began at about age 15 and had occurred intenninendy since that time. She had experienced asmany as several episodes each week for a number ofweeks in succession but at other times had gone amonth or more without any episodes.
The hallucinations usually occurred at sleeponset but had occurred also with arousals in the middle of the night. The hallucinatory content was variable as were the associated emotional and motorresponses. She often was amused by the hallucination. In the week preceding the consultation, she hadseen a Iinle flying saucer whipping about the bedroom. She sat up in bed and anempted to catch it asit flew by, laughing all the while at its elusiveness.Fish were recurring hallucinatory elements. Shortlyafter purchasing a fish tank she saw fish crawling upthe bedroom curtains. She walked over to the curtains and began grabbing at the fish, laughing and
VOLUME 34· NUMBER I • JANUARY - FEBRUARY 1993
telling them that they must go back to their tank. Despite repeated efforts, she could never quite get holdof them.
On other occasions, however, the hallucinationswere frightening. as when the patient reported seeingweird machinery in the room. The machinery beganmoving toward the bed and was perceived as threatening. In response she jumped out of bed and pushedher husband out of bed as wel1. Other hallucinationshad been of a cow floating in the room. of 6-inch spiders dropping from the ceiling, and of a man comingthrough the bedroom transom.
In addition to the hallucinatory episodes, the patient reported other sleep-related symptoms. As anadult she had experienced at least two night terrors(i.e., events in which she awoke from sleep with apiercing scream) without any wel1-developed visualimagery. Over the preceding 2 years, several episodes of sleep paralysis had occurred, always in association with a hallucination. She also reported achronic and moderate level of excessive daytimesleepiness; falling asleep at her desk at work was aproblem but driving. she felt, did not represent ahazard in this regard.
The medical history was unremarkable exceptfor obesity, mild hypertension. and a slightly hypoactive thyroid treated with levothyroxine sodium. AnEEG and an MRI were within nonnallimits.
The family history was relevant for a sister withoccasional nocturnal wanderings of unclear etiologyin a state of partial arousal. The patient's 4-year-oldson had frequent nightmares from which he awakened. He had reported seeing witches and skeletonsin his bedroom, spiders dropping down on his head,and, on one occasion. a banana on his pillow.
Two nights of polysomnographic study and amultiple sleep latency test (MSLT) were perfonned.Sleep was staged according to standard methodsusing EEG channels C4-A I. 01-A2. submental electromyogram (EMG), and electrooculogram. ECG,chest expansion. and anterior tibialis EMG were alsoincluded in the polysomnographic montage. The patient did not report any hallucinations on eithernight, consistent with her claim that episodes awayfrom home were rare. Irregularities in REM sleepwere observed, however. including a short REM latency on Night 2 and an even distribution of REMperiods across both study nights (see Table I). Themean time to sleep onset in the MSLT was 3.5 minutes. wel1 within the pathological range. REM anddelta sleep were absent during the daytime naps.
The patient was treated with 0.5-2.0 mg ofc10nazepam I hour before bedtime, without response.
87
Case 2. Patient B was a 34-year-old female cosmetician with a complaint during the preceding year ofnightmares, night terrors, and terrifying sleep-associated hallucinations that were often accompaniedby gross motor activity or ambulation. The episodesbegan approximately 2 months after a serious automobile accident in which the patient suffered multiple injuries, including a fractured spine. Persistentsequelae from the accident included a neurogenicbladder that required self-eatheterization, basilarmigraine, and foot drop. She was referred by herpsychotherapist who was treating her for anxietyrelated to the accident.
The episodes had occurred with some consistency 2-3 times each week, invariably within 1-2hours of sleep onset. They were always terrifyingbut varied in manifest content, motor response, andextent of recall. Several images repeated themselvesin the hallucinations, such as a redheaded man in thebedroom who might terrorize her with a knife. In response, she would rise from bed and dash about ashe pursued her. On at least one occasion she attempted to climb over the balcony railing of her 23rdstory apartment. A second common image was of alarge string that two people would knit into a hugeweb that threatened to close in on her. Recall of theepisodes were precisely detailed as above, vague, orabsent.
Since the accident, Patient B had also sufferedfrom intermittent sleep paralysis, usually at sleeponset but not infrequently at morning awakening.Chronic daytime sleepiness was also present sincethat time as well as a persistent onset insomnia, possibly related to a fear of falling asleep. Medicationstaken during the past year included cyclobenzaprinefor back spasms, butalbital for migraines, ranitidinefor stomach ailments, alprazolam for anxiety, andtriazolam or amitriptyline for the insomnia. The hallucinatory episodes were somewhat reduced by the
Case Reports
triazolam. The family history was negative for sleepdisorders.
Two nights of polysomnographic study wereperformed. On the fIrSt night there occurred out ofStage m sleep an episode in which the patientopened her eyes and cried out suddenly. She laterreported that at that time she had seen a Martian bythe infrared light source in the room. The head of theMartian moved close to her, and, as she was about tostrike it, the technician entered the room and the hallucination resolved. In addition to this florid episode,the patient was observed on multiple occasions overboth nights to open her eyes in all stages of nonREM sleep and lie quietly, still asleep, in bed. Thisis consistent with home events in which she had hallucinations but remained motionless so that the hallucinatory characters would not notice her and,therefore, would leave her alone. Other than the hallucination described above, however, the patient reported no other episodes while in the sleeplaboratory.
Sleep onset was delayed (49 and 64 minutes)on both study nights, consistent with the history ofinsomnia. Sleep efficiency was low (78%) and REMlatency was prolonged on the first study night. Theseparameters returned to normal on the second night,suggesting that lab effect was responsible. Sleeparchitecture was otherwise unremarkable.
The patient was treated with clonazepam0.5-2.0 mg I hour before bedtime and respondedwith near total resolution of the episodes.
Discussion
In diagnosing the complaint of sleepwalking, attention to the episode-associated mentation andimagery is an important clue to the etiology.Thus, classic somnambulism nonnally produces
TABLE 1. Sleep data on two patients with somnambulism associated with baUucinatiom
SOL TST SE SI% SU% SWS% REM% REML
Patient ANight I 19 450 93 5 57 17 17 90Nighl2 13 464 96 8 56 14 21 54
Patient BNight I 49 356 70 6 43 21 16 156Night 2 64 357 99 7 57 16 19 80
Note: SOL =sleep onsel latency; TST =total sleep time; SE =sleep efficiency; SI% =percentage Stage 1of sleepperiod time; SU% =percentage Stage U sleep; SWS% =percentage slow wave sleep; REM% =percentage REM sleep;REML =l81ency 10 REM sleep.
88 PSYCHOSOMATICS
Case Reports
no morning recall, activity in REM behavior disorder often reflects an ongoing REM dream witha relatively well-developed narrative,3 and theclassic night terror is often accompanied by asingle, fear-eliciting image.6 Terrifying, sleeprelated mental imagery in posttraumatic stressdisorder (PTSD) consists of a replaying of thetraumatic incident.8 Conversely, complex-partialseizures do not produce significant recall.
The imagery reported by these patients doesnot resemble that reported in association with theaforementioned causes of somnambulism but,rather, is hallucinatory. The sleepwalker's awareness of his or her environment and the placementof the hallucinated entity in the environment ismore clearly demonstrated than in classic somnambulism, REM behavior disorder, or PTSD.Because the episodes emerge directly from sleep,they should be classified as sleepwalking inassociation with hypnopompic (following awakening) or hypnagogic (at sleep onset) hallucinations, depending upon their time ofoccurrence inrelation to the sleep period.
To our knowledge, the concurrence ofsleepwalking with hypnagogic or hypnopompichallucinations has not been reported. These hallucinations occur often in narcoleptics. eitheralone or in association with sleep paralysis.9
•lo As
such, they are considered to be dissociated elements of REM sleep. In normal subjects, hallucinations at sleep onset can occur in Stage I sleep. II
Although we could find no previous reports ofsomnambulism in association with hallucinations, there is one case report in which non-ambulatory gross motor activity accompaniediatrogenic hypnopompic hallucinations.12
The prolonged, hallucinatory images ofteninvolving conversation with others are clearlydifferent from imagery normally seen in classicnon-REM parasomnias or in dreams that accompany episodes ofREM behavior disorder. Nevertheless, it is not clear that they represent anentirely new phenomenon. One possibility is thatthey are part of an atypical presentation of narcolepsy (i.e., a phenomenon of dissociated REMsleep). With Patient A, narcolepsy is suggested
VOLUME 34· NUMBER I • JANUARY - FEBRUARY 1993
by several factors: the history of sleep paralysis,the excessive daytime sleepiness confirmed byMSLT, and the somewhat short REM latency onthe second sleep study night. However, there wasno history of cataplexy, REM sleep did not occurduring the MSLT, and true REM onset periodswere absent from nocturnal sleep. The suggestionof narcolepsy in Patient B is weaker, althoughposttraumatic narcolepsy could have a less clearpresentation.
It is also possible that the episodes representcomplex-partial seizures. Nevertheless, in patients with such frequent episodes, the complexity of behavior concurrent with reportableimagery and the absence of any behavior moretypical of seizure make that diagnosis highlyunlikely. Similarly, in patients with high eventfrequency, the presence of a normal interictalEEG as in Patient A and the normal intraepisodicEEG recorded in Patient B argue against a diagnosis of seizure. Although onset of the episodesreported by Patient B occurred following trauma,her mental imagery did not contain replays of thetraumatic incident that would be suggestive ofPTSD.
Alternatively, the occurrence of episodes inPatient B during Stage II and Stage III sleep,which is not characteristic of hypnopompic orhypnagogic hallucinations, suggests that a closerrelationship with classic non-REM parasomniasdoes exist. It is possible that the hallucinatorymentation only indicates a relative shift to theright along a continuum of arousal during a partial arousal from non-REM sleep. Description ofsomnambulistic episodes with atypical, storylike imagery involving earthquakes, fires, andother catastrophes have been described,13 whichsuggests that the range ofmentation possible withclassic non-REM somnambulism is wider thangenerally believed. Thus, these episodes mayprove to be part of a non-REM parasomnia inwhich the mentation varies among episodes. Further study with detailed accounts of event phenomenology is needed to clarify the relationshipsamong these episodes and known types ofparasomnia.
89
Case Reports
References
I. Gastaut H. Broughton R: Aclinical and polygraphic studyofepisodic phenomena during sleep. Recent Advances inBiological Psychiatry 1965; 7: 197-221
2. Kales A. Jacobson A. Paulson MJ. et a1: Somnambulism:psychophysiological correlates I. All-night EEG studies.Arch Oen Psychiatry 1966; 14:586-594
3. Schenck CH, Bundlie SR. Patterson AL. et a1: Rapid eyemovement sleep behavior disorder: a treatable parasomnia affecting older adults. JAMA 1987; 257:17861789
4. Dahl M. Dam M: Sleep and epilepsy. Annals of ClinicalResearch 1985; 17:235-242
5. Cadilhac JC: Complex partial seizures and REM sleep. inSleep and Epilepsy. edited by Sterman MB, Shouse MN.Passouant P. New York. Academic. 1982. pp 315-324
6. Broughton R: Sleep disorders: disorders of arousal? Science 1968; 159:1070-1078
7. Kavey NB. Whyte J, Resor SR, et a1: Somnambulism inadults. Neurology 1990; 40:749-752
8. Hanmann E: The Nighunare. New York, Basic Books,1984
9. Zarcone V: Narcolepsy: a review ofthe syndrome. NEng(J Med 1973; 288:1156-1165
10. Aldrich MS: Narcolepsy. N Eng( J Med 1990; 323:389394
II. Gastaut H. Broughton R: Aclinical and polygraphic studyofepisodic phenomena during sleep. Recent Advances inBiological Psychiatry 1964; 7: 197-221
12. Schlauch R: Hypnopompic hallucinations and treaunentwith imipramine. Am J Psychiatry 1979; 136:219-220
13. Guilleminault C, Silvestri R: Disorders of arousal andepilepsy during sleep, in Sleep and Epilepsy, edited bySterman MB. Shouse MN. Passouant P. New York. Academic.1982.pp513-531
Three-Dimensional CT Validation of PhysicalComplaints in "Psychogenic Pain" Patients
NELSON HENDLER, M.D., M.S.
JAMES ZINREICH, M.D.JOSEPH G. KOZIKOWSKI, B.E.E.
A large percentage of the patients who seekmedical advice for low back pain are inac
curately diagnosed.' These patients deserve multidisciplinary diagnostic evaluations, which mayuncover previously undiagnosed physical abnormalities. In one report, 50% of the patients withchronic pain were referred for further surgery.2However, for some patients, negative fmdings do
Received October 19, 1990; revised June 20. 1991;accepted July 27. 1991. From the Mensana Oinic. Stevenson.MD; the Johns Hopkins University School of Medicine, Baltimore. MD; and the University of Maryland Dental School.Baltimore. MD. Address reprint requests to Dr. Hendler,Clinical Director, the Mensana Oinic, 1718Greenspring Valley Road. Stevenson. MD 21153.
Copyright e 1993 The Academy of PsychosomaticMedicine.
90
not adequately correlate with the clinical featuresofpain. When this happens, physicians often relyon physical tests rather than clinical judgmentand make a diagnosis such as chronic low backpain, lumbosacral strain, psychogenic pain, orconversion reaction. l .2 The patient, meanwhile,continues to have pain and progresses throughfour stages of psychological responses due to thepain.3 Each stage begins to appear after a setperiod of time, and by 6 months the patient hasentered the third stage of chronic pain in whichhypochondriasis, depression, and hysterical features emerge as the result of pain.3
•4 At this point,
because psychiatric problems are evident even ina previously well-adjusted individual, physiciansbecome tempted to ascribe a large part of thepatient's pain complaints to psychogenic pain,
PSYCHOSOMATICS