S E M I N A R S I N NEUROLO(;Y-VOL,UME
1 1 , NO. 2 JUNE: 1991
Sleep Disorders and Epilepsy: Differential Diagnosis
Certain sleep disorders share significant common features with epilepsy. Cardinal symptoms of both conditions are recurrent, stereotypical, paroxysmal, transient disturbances of behavior, with return to baseline function between attacks. In most cases, causative underlying systemic or progressive neurologic diseases are absent, and the patient is normal between attacks. Often, in both conditions, neurologic examination and standard clinical neuroimaging studies are unremarkable. Rather than by permanent, fixed neurologic dysfunctional states, or by permanent structural brain abnormalities, both of these disease categories are characterized by episodic central nervous system (CNS) functional disturbances. Therefore, the critical role of the neurologic history of the attacks, from the patients and observers, is common to the evaluation of sleep disorders and epilepsy. In particular, historical aspects similarly emphasized in sleep disorders medicine and epileptology include behavioral content during and after an attack, temporal aspects of the recurrence pattern, attack duration, age at onset, and inheritance pattern. Since episodic C;NS dysfunction occurs in both, clinical neurophysiologic studies of brain function (interictal electroencephalography [EEG],long-term videoEEG monitoring for ictal events, and polysomnography [see later]) are other significant shared evaluative procedures. Powerful reciprocal interactions exist even between normal physiologic sleep, rather than sleep disorders per se, and epilepsy. Sleep is used routinely to activate spike-wave discharges on the EEG in patients undergoing evaluation for epilepsy. Sleep deprivation probably is additive to sleep in the activation of epileptiform discharges on the EEG and can cause transient increases in seizure frequency in patients with chronic epilepsy. A typ-
ical example of this phenomenon is the college student who experiences an exacerbation o f epilepsy while sleeping little while preparing for examinations. Approximately 25% of epilepsy patients experience seizures only in sleep; another 35% experience seizures while asleep and while awake; so that around 60% of epilepsy patients have at least some nocturnal seizures.' Patients with benign rolandic epilepsy of childhood experience simple partial seizures with tongue paresthesias, salivation, anarthria, and facial twitching while awake, but generalized tonic-clonic seizures while asleep. A feature of nonepileptic seizures of psychologic origin is that they do not occur in sleep. Antiepileptic drugs disrupt sleep cycle architecture and may produce symptomatic sleep disturbances.' This article will review the clinical phenomenology and pathophysiology of certain paroxysmal sleep disorders that resemble epilepsy. Differential diagnosis between the two conditions will be emphasized. T h e correct diagnosis is reached easily in a patient with witnessed nocturnal generalized tonic-clonic seizures, therefore emphasis will be placed on differentiating sleep disorders from nocturnal complex partial seizures. To enhance understanding of the sleep disorders, a brief introductory review of polysomnography and normal physiologic sleep stages and sleep architecture will be presented. Keviews of the dynamic interactions between sleep and epilepsy are available."
POLYSOMNOGRAPHY, SLEEP STAGING, AND SLEEP CYCLES Our current understanding of the physiologic content and architecture of sleep has been arrived at largely because of our ability to perform and
Ilirector, Comprehensive Epilepsy Center, New York Hospital-Cornell New York
Medical Center, Ne\\ E'ork,
Reprint requests: Dr. Labar, Comprehensive Epilepsy Center, New York Hospital-(:o~-nelI Medical Center, 525 East 68th Street, New York, NY 10021 Copyright 0 1991 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NE' 100 16. All rights reserved.
Downloaded by: University of British Columbia. Copyrighted material.
Douglas K . Labar, M.D., Ph.D.
S1,EEP 1)ISOKDEKS A N D EPILEPSY-LABAR
is slower than in the awake state. 'I'he EOG reveals clusters of rapid, ballistic eye movements. Penile erections occur. T h e chin EMG shows a baseline reduction in tonic activity, to a sustained atonic state, but with bursts of phasic muscle activity. T h e bursts of phasic muscle activity o n the chin EMG are highly analogous to the clusters of ballistic eye movements. When a subject is awakened from the various stages of sleep, dreaming is reported most frequently from KEM sleep segments, but drearning also occurs in other stages of sleep. These stages of sleep occur repeatedly in a predictable sequence in a normal subject's sleep architecture. Initially stages I to IV are seen in a progressive sequence over the first 90 minutes of sleep. There then follows a period of KEM sleep lasting 5 to 15 minutes. For the rest of the night, repeated cycles of alternating KEM sleep and nonREM sleep occur over 90-minute periods. With each later cycle, the duration (and therefbre the relative sleep fraction) of KEM increases, at the expense of decreasing sleep fractions occupied by stages 111 and IV. Most of the night is spent oscillating between stage I1 sleep and KEM sleep. Stages I11 and IV sleep occupy a greater sleep fraction in children, and decrease substantially with increasing age.
Downloaded by: University of British Columbia. Copyrighted material.
analyze nocturnal polysomnography (PSG). This physiologic assessment technique allows for recording some o r all electrical signals of EEG, eye rnovernents (electrooculography: EOG), electromyography (EMG) of the chin and tibialis anterior muscles, respiratory oral and nasal airflow, respiratory chest wall and abdominal movements, electrocardiogram, penile tumescence, galvanic skin response, and blood oxygen saturation by pulse oximetry, all continuously in time synchrony with videotaped images of the patient's behaviors and vocalizations. 'The PSG is carried out as an all-night sleep study in the clinical physiology laboratory, and attempts are made to simulate the patient's nocturnal sleeping environment at home. Based on the EEG, EOG, and chin EMG findings on the PSG, sleep is divided into two general stages, rapid eye movement (KEM) sleep and slowwave (non-KEM) sleep. Slow-wave sleep is further divided into stages I, 11, 111, and IV. In stage I sleep, the well-organized, 9 to 10 cyclesisec (Hz) dominant occipital EEG rhythm (alpha rhythm) disappears, and background activity over all head regions is replaced by medium voltage semirhythmic slower (4 to 7 Hz) theta activity. Kolling, slow lateral eye movements occur con.jugately, with an EOC; frequency of around 0.25 to 0.50 Hz. If' awakened during this stage of sleep, the patient may report being drowsy, o r may report being asleep. Stage I1 sleep is marked by the appearance of sleep spindles, which are bilateral, vertex maximal runs of monorhythmic 14 Hz activity seen on the EEG for periods lasting 1 to 3 seconds. Sleep spindles may persist into stages 111 and IV sleep, but become less frequent, of lower amplitude, and less well organized in the deeper stages. 'The emergence of stages 111 arid IV sleep is marked by increasing amounts of very slow EEG activity (delta, less than 4 Hz); in stage 111, there is 25 to 50% of EEG activity in the delta range, and in stage IV sleep, there is more than 50% of EEG activity in the delta frequency range. Kelatively few eye movements are detected by the EOG in stages I1 to IV. A low level of tonic muscle activity is recorded by the chin EMC; in stage I to IV sleep. Vertex waves on the EEG are high-voltage monophasic o r polyphasic, vertex-maximal, negative-polarity, deltafrequency waves best seen in stages I and I1 sleep, which fade into the background delta activity in stages 111 and IV sleep. In REM sleep, the EEG acquires a relatively fast-frequency content, with low-voltage activity composed of' mixed theta and alpha waves predominating. T h e EEG activity is frequently described as "desynchronized" and exhibiting "sawtooth waves." Although faster than in the other stages of sleep, the background frequency content
SOMNAMBULISM T h e behaviors typifying the sleepwalking attacks in somnambulism are moderately complex, semipurposeful, at times repetitive motor acts, such as leaving the bed and entering an adjacent bathroom; going to the kitchen, opening the cabinets o r the refrigerator, and eating; and even leaving the house and walking around the neighborhood. This behavior typically continues for u p to 30 minutes. T h e patient then may voluntarily return to the bed, be led back to bed, o r return to peaceful sleep elsewhere in the house. Upon awakening, the patient is amnesic for the entire episode. Although it is not widely recognized, injuries, usually of mild to moderate severity, occur frequently during episodes of somnambulism.' This disorder is largely restricted to children in the first decade of life, occurring in u p to 15% of normal children.When present in adults, it can be associated with one of several abnormal personality traits in the general category of emotional lability, particularly poor impulse control, and excessive anger." In most cases, a family history of this o r some other sleep disorder is present.' T h e sleepwalking attacks occur within 1 to 2 hours of sleep onset. With PSG monitoring, they can be seen to be an arousal from stage 111 o r IV
129
S E M I N A R S I N NEUKOI,O
1 1 , NO. 2 JUNE: 1991
Sleep Disorders and Epilepsy: Differential Diagnosis
Certain sleep disorders share significant common features with epilepsy. Cardinal symptoms of both conditions are recurrent, stereotypical, paroxysmal, transient disturbances of behavior, with return to baseline function between attacks. In most cases, causative underlying systemic or progressive neurologic diseases are absent, and the patient is normal between attacks. Often, in both conditions, neurologic examination and standard clinical neuroimaging studies are unremarkable. Rather than by permanent, fixed neurologic dysfunctional states, or by permanent structural brain abnormalities, both of these disease categories are characterized by episodic central nervous system (CNS) functional disturbances. Therefore, the critical role of the neurologic history of the attacks, from the patients and observers, is common to the evaluation of sleep disorders and epilepsy. In particular, historical aspects similarly emphasized in sleep disorders medicine and epileptology include behavioral content during and after an attack, temporal aspects of the recurrence pattern, attack duration, age at onset, and inheritance pattern. Since episodic C;NS dysfunction occurs in both, clinical neurophysiologic studies of brain function (interictal electroencephalography [EEG],long-term videoEEG monitoring for ictal events, and polysomnography [see later]) are other significant shared evaluative procedures. Powerful reciprocal interactions exist even between normal physiologic sleep, rather than sleep disorders per se, and epilepsy. Sleep is used routinely to activate spike-wave discharges on the EEG in patients undergoing evaluation for epilepsy. Sleep deprivation probably is additive to sleep in the activation of epileptiform discharges on the EEG and can cause transient increases in seizure frequency in patients with chronic epilepsy. A typ-
ical example of this phenomenon is the college student who experiences an exacerbation o f epilepsy while sleeping little while preparing for examinations. Approximately 25% of epilepsy patients experience seizures only in sleep; another 35% experience seizures while asleep and while awake; so that around 60% of epilepsy patients have at least some nocturnal seizures.' Patients with benign rolandic epilepsy of childhood experience simple partial seizures with tongue paresthesias, salivation, anarthria, and facial twitching while awake, but generalized tonic-clonic seizures while asleep. A feature of nonepileptic seizures of psychologic origin is that they do not occur in sleep. Antiepileptic drugs disrupt sleep cycle architecture and may produce symptomatic sleep disturbances.' This article will review the clinical phenomenology and pathophysiology of certain paroxysmal sleep disorders that resemble epilepsy. Differential diagnosis between the two conditions will be emphasized. T h e correct diagnosis is reached easily in a patient with witnessed nocturnal generalized tonic-clonic seizures, therefore emphasis will be placed on differentiating sleep disorders from nocturnal complex partial seizures. To enhance understanding of the sleep disorders, a brief introductory review of polysomnography and normal physiologic sleep stages and sleep architecture will be presented. Keviews of the dynamic interactions between sleep and epilepsy are available."
POLYSOMNOGRAPHY, SLEEP STAGING, AND SLEEP CYCLES Our current understanding of the physiologic content and architecture of sleep has been arrived at largely because of our ability to perform and
Ilirector, Comprehensive Epilepsy Center, New York Hospital-Cornell New York
Medical Center, Ne\\ E'ork,
Reprint requests: Dr. Labar, Comprehensive Epilepsy Center, New York Hospital-(:o~-nelI Medical Center, 525 East 68th Street, New York, NY 10021 Copyright 0 1991 by Thieme Medical Publishers, Inc., 381 Park Avenue South, New York, NE' 100 16. All rights reserved.
Downloaded by: University of British Columbia. Copyrighted material.
Douglas K . Labar, M.D., Ph.D.
S1,EEP 1)ISOKDEKS A N D EPILEPSY-LABAR
is slower than in the awake state. 'I'he EOG reveals clusters of rapid, ballistic eye movements. Penile erections occur. T h e chin EMG shows a baseline reduction in tonic activity, to a sustained atonic state, but with bursts of phasic muscle activity. T h e bursts of phasic muscle activity o n the chin EMG are highly analogous to the clusters of ballistic eye movements. When a subject is awakened from the various stages of sleep, dreaming is reported most frequently from KEM sleep segments, but drearning also occurs in other stages of sleep. These stages of sleep occur repeatedly in a predictable sequence in a normal subject's sleep architecture. Initially stages I to IV are seen in a progressive sequence over the first 90 minutes of sleep. There then follows a period of KEM sleep lasting 5 to 15 minutes. For the rest of the night, repeated cycles of alternating KEM sleep and nonREM sleep occur over 90-minute periods. With each later cycle, the duration (and therefbre the relative sleep fraction) of KEM increases, at the expense of decreasing sleep fractions occupied by stages 111 and IV. Most of the night is spent oscillating between stage I1 sleep and KEM sleep. Stages I11 and IV sleep occupy a greater sleep fraction in children, and decrease substantially with increasing age.
Downloaded by: University of British Columbia. Copyrighted material.
analyze nocturnal polysomnography (PSG). This physiologic assessment technique allows for recording some o r all electrical signals of EEG, eye rnovernents (electrooculography: EOG), electromyography (EMG) of the chin and tibialis anterior muscles, respiratory oral and nasal airflow, respiratory chest wall and abdominal movements, electrocardiogram, penile tumescence, galvanic skin response, and blood oxygen saturation by pulse oximetry, all continuously in time synchrony with videotaped images of the patient's behaviors and vocalizations. 'The PSG is carried out as an all-night sleep study in the clinical physiology laboratory, and attempts are made to simulate the patient's nocturnal sleeping environment at home. Based on the EEG, EOG, and chin EMG findings on the PSG, sleep is divided into two general stages, rapid eye movement (KEM) sleep and slowwave (non-KEM) sleep. Slow-wave sleep is further divided into stages I, 11, 111, and IV. In stage I sleep, the well-organized, 9 to 10 cyclesisec (Hz) dominant occipital EEG rhythm (alpha rhythm) disappears, and background activity over all head regions is replaced by medium voltage semirhythmic slower (4 to 7 Hz) theta activity. Kolling, slow lateral eye movements occur con.jugately, with an EOC; frequency of around 0.25 to 0.50 Hz. If' awakened during this stage of sleep, the patient may report being drowsy, o r may report being asleep. Stage I1 sleep is marked by the appearance of sleep spindles, which are bilateral, vertex maximal runs of monorhythmic 14 Hz activity seen on the EEG for periods lasting 1 to 3 seconds. Sleep spindles may persist into stages 111 and IV sleep, but become less frequent, of lower amplitude, and less well organized in the deeper stages. 'The emergence of stages 111 arid IV sleep is marked by increasing amounts of very slow EEG activity (delta, less than 4 Hz); in stage 111, there is 25 to 50% of EEG activity in the delta range, and in stage IV sleep, there is more than 50% of EEG activity in the delta frequency range. Kelatively few eye movements are detected by the EOG in stages I1 to IV. A low level of tonic muscle activity is recorded by the chin EMC; in stage I to IV sleep. Vertex waves on the EEG are high-voltage monophasic o r polyphasic, vertex-maximal, negative-polarity, deltafrequency waves best seen in stages I and I1 sleep, which fade into the background delta activity in stages 111 and IV sleep. In REM sleep, the EEG acquires a relatively fast-frequency content, with low-voltage activity composed of' mixed theta and alpha waves predominating. T h e EEG activity is frequently described as "desynchronized" and exhibiting "sawtooth waves." Although faster than in the other stages of sleep, the background frequency content
SOMNAMBULISM T h e behaviors typifying the sleepwalking attacks in somnambulism are moderately complex, semipurposeful, at times repetitive motor acts, such as leaving the bed and entering an adjacent bathroom; going to the kitchen, opening the cabinets o r the refrigerator, and eating; and even leaving the house and walking around the neighborhood. This behavior typically continues for u p to 30 minutes. T h e patient then may voluntarily return to the bed, be led back to bed, o r return to peaceful sleep elsewhere in the house. Upon awakening, the patient is amnesic for the entire episode. Although it is not widely recognized, injuries, usually of mild to moderate severity, occur frequently during episodes of somnambulism.' This disorder is largely restricted to children in the first decade of life, occurring in u p to 15% of normal children.When present in adults, it can be associated with one of several abnormal personality traits in the general category of emotional lability, particularly poor impulse control, and excessive anger." In most cases, a family history of this o r some other sleep disorder is present.' T h e sleepwalking attacks occur within 1 to 2 hours of sleep onset. With PSG monitoring, they can be seen to be an arousal from stage 111 o r IV
129
S E M I N A R S I N NEUKOI,O
