Sleep Medicine Pearls 2nd

Answer: A trial of modafinil. Discussion: Methylphenidate and amphetamine medications are effective in decreasing daytime sleepiness to satisfactory levels in about 60-80% of patients with narcolepsy. However, some patients experience significant side effects, including nervousness, weight loss, or even sleep disturbance. Modafinil (Provigil) is a new alerting agent that is not a stimulant. Its mechanism of action is unknown. Double-blind placebo-controlled trials have shown that modafinil is effective at increasing objective and subjective measures of sleepiness in patients with narcolepsy. However, it has not been shown to be more effective than the stimulant medications such as methylphenidate. Modafinil has a number of advantages: It can be taken once daily (half life 9-14 hrs); it is not a schedule II medication (refills allowed); it does not disturb sleep if taken in the morning; and it is not associated with tolerance. Moreover, side effects are relatively few. The most common side effects are headache and nausea. Headache can be minimized by starting at a very low dose (100 mg) daily for a few days if necessary. Other side effects include nervousness and palpitations. The drug is metabolized in the liver, and there are a number of potential drug interactions. The most significant one is that modafinil may reduce the effectiveness ofbirth control medications. The usual starting dose of modafinil is 200 mg once daily. If significant daytime sleepiness persists, a higher dose of 400 mg daily may be effective. While not studied systematically, some clinicians have also added small doses of methylphenidate to modafinil, if needed. One recent preliminary study suggested that if patients have breakthrough sleepiness in the afternoon, they can try splitting the dose of modafinil (modafinil 200 mg bid). Another alternative that can be tried in patients not tolerating stimulants is the irreversible MAO type B inhibitor selegiline. At doses of 10-20 mg/day, this drug has been shown to improve narcoleptic symptoms. Unfortunately, at doses over 20 mg/day it loses its MAO inhibitorselectivity, and even at lower doses a low tyramine diet is indicated to avoid the risk of hypertensive reactions. The drug also has anti-cataplectic activity in addition to its alerting ability. Patients who experience intolerable side effects with other agents may benefit from this medication as long as they are willing to adhere to a low tyramine diet. In the present case, the patient was started on modafinil 100 mg daily for I week. He did not feel as awake as he had on the methylphenidate, and the dose of modafinil was increased to 400 mg daily. On this dose he felt better, but still had problems near lunch time. Methylphenidate 10 mg was added at II AM. On this regimen he returned to his normal weight and noted less irritability, Clinical Pearls I. Modafinil offers a useful treatment alternative and is the only FDA-approved medication for narcolepsy. 2. Modafinil has not been proven to be more effective than traditional stimulant medications for treating the daytime sleepiness of narcolepsy. 3. Modafinil has never been directly compared to methylphenidate with respect to side effects. However, modafinil did not impair sleep quality in placebo-controlled studies. Some patients may find modafinil more tolerable than stimulants. 4. Headache is the most common side effect of modafinil. Starting with 100 mg for a few days may minimize this side effect. 5. Modafinil has a number of potential drug interactions. However, the most significant is with hormonal birth control medications. 6. Selegiline, an MAO inhibitor, has both alerting and anti-cataplexic properties. It may be tried in patients intolerant of modafinil or stimulants. A low tyramine diet is required. REFERENCES I. Mayer G, Ewert-Meier K, Hephata K: Selegeline hydrochloride treament in narcolepsy: A double-blind, placebo-controlled study. Clin Neuropharmcol 1995; 18:306-319. 2. U.S. Modafinil in Narcolepsy Study Group: Randomized trial of modafinil for the treatment of pathological somnolence in narcolepsy. Ann Neurol 1998; 43:88-97. 3. Moldofsky H, Broughton RJ, Hill JD: A randomized trial of the long-term, continued efficacy and safety of rnodafinil in narcolepsy. Sleep Med 2000; 1:109-116. 4. U.S. Modafinil in Narcolepsy Study Group: Randomized trial of rnodafinil as a treatment for the excessive daytime somnolence of narcolepsy. Neurology 2000; 53: 1166-1175. 289

PATIENT 90 A 35-year-old man with sleep apnea and a short REM latency A 35-year-old man complained of severe sleepiness of 5-year duration (Epworth Sleepiness Scale score 20/24). His wife reported that he snored heavily and had severe attacks of sleepiness in social situations and even at meals. However, she was unable to comment on the existence of periods of apnea because she slept in a separate bedroom. The patient had recently been fired for falling asleep on the job. He denied sleep paralysis, but reported feeling "funny" when angry. The sensation was more like being dizzy than weak. A multiple sleep latency test (MSLT) performed without nocturnal polysomnography at another hospital showed a mean sleep latency of 4 minutes (severe sleepiness) and two of five naps with REM sleep. The patient was diagnosed with narcolepsy, but treatment with stimulants was ineffective. Sleep Study Time in bed Total sleep time Sleep period time Sleep latency REM latency Arousal index 480 min (414-455) 350 min (400-443) 425 min (405-451) 10 min (2-20 min) 5 min (70-100) 65/hr Sleep Stages Stage Wake Stage 1 Stage 2 Stages 3 and 4 Stage REM AHI PLM index O/OSPT 18(0-3) 13 (2-9) 49 (50-64) 5 (7-18) 10 (20-27) 80/hr « 5) O/hr ( )= normal values for age, PLM = periodic leg movement Questions: What is your diagnosis? Should another MSLT be performed immediately? 290

Diagnosis: Obstructive sleep apnea. An MSLT would not be useful until after the sleep apnea is adequately treated. Discussion: Obstructive sleep apnea (OSA) can be associated with both a short sleep latency and two or more REM periods on an MSLT. In fact, one study of 1145 consecutive patients evaluated for suspected OSA found two or more sleep-onset REM periods in 4.7%. Thus, these MSLT findings cannot be used to support a diagnosis of narcolepsy in patients with significant sleep apnea. If narcolepsy is suspected, the approach is to first treat the sleep apnea, and then repeat the MSLT if clinically indicated. For example, treatment is begun with nasal CPAP. If daytime sleepiness completely resolves, then narcolepsy probably is not present. If some degree of sleepiness persists or there is a history of cataplexy, then further evaluation is indicated. After several weeks of treatment, another sleep study (on nasal CPAP) confirms adequate treatment and REM sleep, and a subsequent MSLT (also on CPAP) determines the severity of residual sleepiness as well as the presence of REM periods. A diagnosis of narcolepsy is supported when there is evidence of persistent, severe sleepiness (mean sleep latency < 5 minutes) and two or more of five naps with REM sleep-assuming that the nocturnal sleep study showed reasonable sleep quality (and no evidence of REM or slow wave sleep rebound). No further evaluation is needed when the MSLT is normal. However, if the MSLT shows significant sleepiness (sleep latency < 10 minutes) and no REM periods, then several possibilities must be considered, including: poor compliance with nasal CPAP, inadequate CPAP pressure, idiopathic hypersomnia, periodic leg movements, insufficient sleep, and narcolepsy. Remember, failure on an MSLT to show two or more REM periods in five naps does not rule out narcolepsy. When unequivocal cataplexy is present, a diagnosis of narcolepsy can be made on clinical grounds. Two studies have suggested that modafinil can improve sleepiness in patients with OSA still sleepy on positive airway pressure. Thus, a case can be made for adding this wake-promoting agent to positive airway pressure treatment even if a diagnosis of narcolepsy is not clearly documented. However, the addition should occur only after positive-pressure treatment is optimized and other sleep disorders eliminated, if possible. Note that while adequate positive pressure improves daytime sleepiness in OSA patients, it does not always return the MSLT mean sleep latency to a normal value (> 10 minutes). In the present patient, the polysomnogram revealed severe OSA and sleep fragmentation (high arousal index), with reduced amounts of slow wave and REM sleep. The REM latency was very short, suggesting narcolepsy - but this finding also can be seen in OSA. There was no history of unequivocal cataplexy, but daytime sleepiness can be present for several years before the onset of cataplexy. The patient underwent a nasal CPAP titration, and on 12 ern H20 the AHI was reduced to 5/hr. A large rebound in slow wave and REM sleep was noted. Treatment with nasal CPAP resulted in a complete resolution of symptoms: This fact alone made narcolepsy unlikely. However, because of the past diagnosis and the equivocal history of cataplexy, a repeat sleep study and MSLT (both on CPAP) were performed several weeks later. The nocturnal study showed a normal REM latency and adequate treatment of OSA. The MSLT showed a sleep latency of 12 minutes and an absence of REM sleep. These findings demonstrated that coexistent narcolepsy was unlikely. Clinical Pearls 1. When significant sleep apnea is present on a nocturnal sleep study, testing for narcolepsy (MSLT) should be delayed until after the sleep apnea is adequately treated. 2. REM deprivation associated with OSA can result in both a short nocturnal REM latency and two REM periods during MSLT testing. 3. A nocturnal study documenting effective treatment of OSA (and a normal amount of REM sleep) coupled with a subsequent MSLT (on treatment) satisfying the diagnostic criteria for narcolepsy supports this additional diagnosis in a patient with OSA. 4. An MSLT without preceding nocturnal polysomnography can be misleading and is rarely indicated. 5. The addition of modafinil to positive-pressure treatment of patients with OSA can be considered if the patients are still sleepy on optimized positive-pressure treatment. 291

REFERENCES I. Walsh JK. Smitson SS. Kramer M: Sleep-onset REM sleep: Comparison of narcoleptic and sleep apnea patients. Clin Electroencephalogr 1982: 13:57-60. 2. American Sleep Disorders Association: The clinical use of the multiple sleep latency test. Sleep 1992: 15:268-276. 3. Chervin RD, Aldrich MS: Sleep onset REM periods during multiple sleep latency tests in patients evaluated for sleep apnea. Am J Resp Crit Care Med 2000: 161:426--431. 4. Kingshott RN, Vennelle M, Coleman EL. et al: Randomized, double-blind, placebo-controlled crossover trial of modafinil in the treatment of residual excessive daytime sleepiness in the sleep apnea/hypopnea syndrome. Am J Respir Crit Care Med 200 I: 163:918-923. 5. Pack AI, Black JE. Schwartz JR. Matheson JK: Modafinil as adjunct therapy for daytime sleepiness in obstructive sleep apnea. Am J Respir Crit Care Med 2001: 164:1675-1681. 292

PATIENT 91 A 40-year-old man with sleep apnea and persistent daytime sleepiness A 40-year-old African-American man with excessive daytime sleepiness since age 20 was diagnosed as having obstructive sleep apnea (AHI 80/hr) at another hospital. Nasal CPAP at 12 cm H20 reduced the AHI to 3/hr. The patient was started on treatment, and he noted some improvement in his symptoms. However, significant daytime sleepiness persisted despite using CPAP for at least 6 hours a night. There was no history of cataplexy or sleep paralysis. Narcolepsy was considered, but the patient was HLA-DR15 (a subtype of HLA-DR2) negative. He was referred for another opinion. Physical Examination: HEENT: dependent palate; 17-inch neck circumference. Chest: clear. Cardiac: normal. Extremities: no edema. Sleep Study (on nasal CPAP 12 em HP) Time in bed 450 min (390-468) Total sleep time 395.5 min (343-436) Sleep period time (SPT) 430 min (378--452) WASO 34.5 min Sleep efficiency (0/0) 88 (85-97) Sleep latency 10 min (2-18) REM latency 15 min (55-78) Sleep Stages Stage Awake Stage 1 Stage 2 Stages 3 and 4 Stage REM AHI « 5/hr) PLM index O/OSPT 8 (1-12) 10 (5-11) 52 (44-66) 10(2-15) 20 (19-27) 3/hr IO/hr ( ) = normal values for age. AHI = apnea + hypopnea index. PLM = periodic limb movement Multiple Sleep Latency Test: On nasal CPAP 12 cm H20 : mean sleep latency 4 minutes; three of five naps with REM sleep. Question: What is causing the persistent sleepiness? 293

Diagnosis: Narcolepsy. Discussion: A combination of narcolepsy and obstructive sleep apnea (OSA) is not uncommon. Adequate treatment of both disorders is required for control of daytime sleepiness. If cataplexy is unequivocal, a diagnosis of narcolepsy can be made in patients with OSA on clinical grounds. However, cataplexy makes a delayed appearance in many patients with narcolepsy-sometimes years after sleep attacks begin-and is present in only about 70% of patients with narcolepsy. The multiple sleep latency test (MSLT) is used to support the diagnosis of narcolepsy in these patients with OSA and suspected narcolepsy without cataplexy. However, patients with untreated OSA can exhibit MSLT findings consistent with narcolepsy. Therefore, the first step in all of these cases is successful treatment ofthe GSA. Then a repeat sleep study (on treatment) is followed by an MSLT (also on treatment). The sleep study documents adequate treatment (and adequate sleep), and the MSLT provides objective evidence of continued daytime sleepiness (sleep latency < 5 min) and the presence of two or more REM periods. The differential of a patient with OSA still sleepy on CPAP includes: poor compliance, inadequate CPAP pressure, narcolepsy, periodic limb movements in sleep, idiopathic hypersomnia, insufficient sleep, and depression. "LA haplotyping has been used in evaluation of suspected narcolepsy since early studies showed that most patients with narcolepsy (with cataplexy) were HLA-DR2 positive. Obviously, many patients without narcolepsy are HLA-DR2 positive, so the main utility of haplotyping is excluding the diagnosis in HLA-DR2 negative patients. The first reports were that most Japanese patients with narcolepsy were HLA-DR2 positive. Subsequent studies have found that DRI5 (a subtype of HLA-DR2) and DQ6 (particularly DQB I*0602) are present in 95-100% of Caucasian and Japanese patients with narcolepsy, and DQB I*0602 also is present in 95% of narcoleptic African-American patients. However, 40% of the latter are DR 15 negative. Therefore, DQBl*0602 appears to be the best marker across all races, but the utility of genetic testing is limited by the fact that 1-5% of all patients with narcolepsy are negative for DQB I*0602. Patients with narcolepsy without cataplexy are more likely to be DQB I*0602 negative. Of note, there appear to be factors other than genetic that determine the appearance of the syndrome. Cases of monozygotic twins have been reported where only one twin developed narcolepsy. In the present patient, a negative HLA-DR2 test did not rule out narcolepsy-the patient was, in fact, DQB I*0602 positive. The early age of symptom onset was consistent with narcolepsy. The sleep study on CPAP showed excellent treatment of his sleep apnea and sleep of fairly good quality (no evidence of REM rebound). The MSLT documented both severe daytime sleepiness and REM periods in three of five naps. The absence of other reasons to explain these MSLT findings makes narcolepsy highly likely. The patient was treated with methylphenidate 10 mg tid, with improvement in his symptoms of daytime sleepiness. Clinical Pearls I. While most narcolepsy patients are DQB I*602 positive (all races), a negative result does not rule out narcolepsy. 2. When daytime sleepiness persists on nasal CPAP, consider the possibility of other sleep disorders as well as poor compliance or inadequate pressure. 3. An MSLT can provide objective evidence of persistent sleepiness on nasal CPAP and help support a diagnosis of narcolepsy. 4. If narcolepsy is suspected in a patient with significant OSA, first treat the OSA. Then, a sleep study and MSLT on treatment can be ordered to support a diagnosis of narcolepsy as well as OSA. REFERENCES I. American Sleep Disorders Association: The clinical use of the multiple sleep latency test. Sleep 1992; 15:268-276. 2. Mignot E. Lin X. Arrigoni J. et al: DQB I*0602 and DQA I*0 10I are better markers than DR2 for narcolepsy in Caucasians and African-Americans. Sleep 1994; 17:60-67. 3. Bassetti C. Aldrich MS: Narcolepsy. Neurol Clin 1996; 14:545-569. 294

PATIENT 92 A 35-year-old man requesting stimulant medication A 35-year-old man was evaluated for complaints of excessive daytime sleepiness of 6-year duration. He had been diagnosed with idiopathic hypersomnia by another physician and had been receiving stimulant medications for more than 2 years. After having recently moved, he sought medical attention for medication refills. For the previous 2 months he had not taken stimulant medications, and at work he drank large amounts of coffee to combat sleepiness. His usual bedtime was 11:00 PM, and he awoke at 4:30 AM by alarm. The early awake time was necessary because of a lengthy commute to work. On the weekends, he slept to 9:00 AM and felt somewhat less sleepy. The patient denied a history of cataplexy, hypnagogic hallucinations, sleep paralysis, head trauma, or depression. There was no history of snoring. His previous evaluations included a normal nocturnal polysomnogram, and an MSLT showed a short sleep latency (8 minutes) with no episodes of REM sleep. The patient was asked to keep a sleep log and to obtain at least 7 hours of sleep nightly before a repeat polysomnogram and MSLT. Physical Examination: Normal. Laboratory Findings: Normal thyroid function. Sleep Study Time in bed Total sleep time Sleep period time (SPT) Sleep efficiency (0/0) Sleep latency REM latency Arousal index AHI 440 min (414-455) 417 min (400-443) 430 min (405-451) 95 (95-99) 10 min (2-10 min) 110 min (70-100) 10/hr 2/hr « 5) Sleep Stages Stage Wake Stage I Stage 2 Stages 3 and 4 Stage REM PLM index PLM-arousal index O/OSPT 3 (0-3) 13 (2-9) 49 (50-64) 15 (7-18) 20 (20-27) O/hr O/hr AHI = apnea + hypopnea index, PLM = periodic leg movement. ( ) = normal values for age MSLT: Mean sleep latency 13 minutes, no REM periods in five naps. Question: What is your diagnosis? 295

Diagnosis: Insufficient sleep syndrome. Discussion: In the insufficient sleep syndrome, an inadequate amount of time is allotted for sleep by the patient due to personal or societal (work) schedules. The amount of sleep required for normal function varies considerably between individuals, with a population mean around 7.5 hours. This sleep need is genetically determined. When less sleep is obtained, a sleep debt accumulates. Commonly, such patients sleep considerably more on weekends. A study of patients with the insufficient sleep syndrome found that this disparity between the amount of sleep obtained on weekday nights and on weekends was an important clinical clue. These patients had normal-to-high sleep efficiencies during nocturnal sleep testing, with greater total sleep times than reported for a typical night, and they showed moderate reductions in sleep latency without REM periods on MSLT. One study in normal subjects found that a reduction of the time in bed from 8 to 6 hours reduced the mean sleep latency from approximately 12.5 to 8.5 minutes. Thus, a mild reduction in nocturnal sleep can increase daytime sleepiness, although usually not to a severe degree (i.e., sleep latency < 5 minutes). Remember, though, that any reduction in nocturnal sleep magnifies the sleepiness associated with other sleep disorders, such as narcolepsy or sleep apnea. The present patient's normal duration of sleep was, at most, 5.5 hours. Thus, the possibility of insufficient sleep was considered. This short sleep time would make interpretation of an MSLT difficult, which is why the patient was asked to sleep for at least 7 hours and keep a sleep log prior to testing. The nocturnal polysomnogram documented fairly normal sleep. The MSLT revealed a sleep latency in the "grey" zone: traditionally, a sleep latency> 15 minutes is considered normal, < 10 minutes abnormal, and 10-15 minutes could be either (mild sleepiness). Certainly a sleep latency of 13 minutes is inconsistent with the severe symptoms reported by this patient. When confronted with the results of his testing, the patient admitted that he thinks "sleep is a waste of time" and that he always tries to function on as little sleep as possible. Although he could not remember his sleep habits before the previous sleep testing, he believed he had allotted the usual short amount of time. While not entirely happy with the decision not to prescribe stimulants, the patient did understand that the test proved he would be less sleepy during the day if he had more nocturnal sleep. Clinical Pearls 1. Proper interpretation of the MSLT depends on the patient having an adequate amount of sleep (ideally 7-7.5 hours a night) for at least 1 week before testing. 2. An accurate sleep log is an essential part of the evaluation of daytime sleepiness. It also is helpful in interpreting the results of both the nocturnal sleep study and the MSLT. 3. A modest shortening of nocturnal sleep (to about 6 hours) can shorten the mean sleep latency on the MSLT to < 10 minutes. 4. The insufficient sleep syndrome should be considered in the differential of excessive daytime sleepiness. It may also worsen the impact of other disorders such as narcolepsy and OSA. REFERENCES l. Roehrs T. Zorick F. Sicklesteel J. et al: Excessive daytime sleepiness associated with insufficient sleep. Sleep 1983; 6:319-325. 2. American Sleep Disorders Association: The clinical use of the multiple sleep latency test. Sleep 1992; 15:268-276. 3. Rosenthal L. Roehrs TA. Rosen A. et al: Level of sleepiness and total sleep time following various time in bed conditions. Sleep 1993; 16:226-232. 4. Aldrich MS: The clinical spectrum of narcolepsy and idiopathic hypersomnia. Neurology 1996; 46:383-40 I. 296

PATIENT 93 A 64-year-old man with daytime sleepiness since age 21 A 64-year-old man was evaluated for the complaint of excessive daytime sleepiness present since age 21. He sometimes fell asleep while driving and in social situations. He denied a history of cataplexy, sleep paralysis, or hypnagogic hallucinations. There was a history of mild snoring. The patient retired nightly around 9:30 PM and reported falling asleep in less than 30 minutes. He usually arose at 6:30 AM (awakened by alarm clock). The patient commonly slept for up to 9 hours on the weekend, with no reduction in his symptoms of sleepiness. There was no history of head trauma nor symptoms to suggest depression. Physical Examination: Vital signs: normal. General: thin, in no distress. HEENT: normal. Chest: clear to auscultation and percussion. Cardiac: normal. Abdomen: normal. Extremities: normal. Neurologic: normal. Sleep Study Time in bed Total sleep time Sleep period time (SPT) Sleep efficiency (%) Sleep latency REM latency Arousal index 490 min (414-489) 450 min (363-452) 470 min (404-479) 92 (83-97) 4.5 min (1-15 min) 100 min (65-103) lO/hr Sleep Stages Stage Wake Stage I Stage 2 Stages 3 and 4 Stage REM AHI PLM index %SPT 10 (2-14) 10 (6-14) 55 (48-66) 7 (0-8) 18 (20-27) O/hr « 5) O/hr ( ) = normal values for age, AHI = apnea + hypopnea index, PLM = periodic leg movement MSLT: Mean sleep latency = 6 minutes; no REM sleep in five naps. Question: What is the most likely cause of the patient's daytime sleepiness? 297

Diagnosis: Idiopathic hypersomnia. Discussion: The diagnosis of idiopathic hypersomnia (IHS) is made by documenting excessive daytime sleepiness (EDS) despite adequate sleep and excluding other disorders that cause daytime sleepiness, such as narcolepsy. Patients with IHS may complain of persistent daytime drowsiness or discrete sleep attacks. IHS, like narcolepsy, usually begins in adolescence or the early twenties. The syndrome accounts for 5-10% of patients seen in sleep clinics with complaints of EDS. The sleep period may be long (> 8 hours). Some patients are able to wake normally; others report difficulty waking and/or disorientation at awakening ("sleep drunkenness"). Unlike narcolepsy, naps are not refreshing. Some patients have the onset of symptoms after a viral illness such as hepatitis and mononucleosis. Rarely, familial cases occur and are associated with an increased incidence of HLA-Cw2. Some of this group have associated symptoms suggesting problems with the autonomic nervous system, including headache, syncope, orthostatic hypotension, and peripheral vascular complaints (Raynaud-type symptoms). Other patients with IHS have no history of viral illness or autonomic nervous system problems. Thus, patients with IHS are a heterogeneous group. The polysomnography of IHS patients shows a normal or increased quantity of sleep. Sleep quality usually is normal, and the amount of slow wave sleep may be increased. The REM latency is not decreased. The sleep latency usually is < 10 minutes. In contrast, the sleep of narcoleptics typically is shortened or fragmented, and the REM latency may be very short « 20 minutes). The multiple sleep latency test (MSLT) in patients with IHS documents EDS (sleep latency < 10 minutes) while showing no REM sleep episodes (rarely, one) in five naps. Some have suggested that 24-hour sleep recording may be useful in patients with IHS. Common results are long nocturnal sleep and long daytime naps, with 10-12 hrs of total sleep. Some patients with IHS report hypnagogic hallucinations and sleep paralysis. Thus, eliciting such symptoms does not necessarily differentiate IHS from narcolepsy. Before the diagnosis of IHS can be made with confidence, medical and psychiatric causes of EDS should be excluded. Chronic low-grade depression may be difficult to eliminate as a possibility. A moderately reduced nocturnal REM latency « 60 minutes) might be a clue that depression is present. Sleepiness also can be a symptom of progressive hydrocephalus. Recent onset, worsening of symptoms, or impairment of cognitive functioning suggests a need for neurologic evaluation and computed axial tomography or other studies to rule out this possibility. Posttraumatic hypersomnia is a syndrome in which symptoms and findings of hypersomnia develop 6-18 months after head trauma. Sedative/hypnotic abuse is another possible cause of daytime sleepiness. Urine or blood tests can screen for use of these agents. The insufficient sleep syndrome and the upper airway resistance syndrome (UARS) also should be excluded. The treatment of patients with IHS involves the same stimulant medications as used in narcolepsy, but patients with IHS do not always respond. Patients also are instructed to avoid reductions in sleep time or irregular sleep/wake schedules. The present patient snored lightly but aroused rarely, making UARS unlikely. The nocturnal sleep study did not show a decreased REM latency. There was no history of depression or head trauma. The MSLT documented moderate sleepiness (sleep latency = 6 min) despite fairly normal sleep, but no REM sleep was noted. The MSLT was not consistent with narcolepsy, and there was no history of cataplexy. However, neither of these facts absolutely excludes narcolepsy. Therefore, the diagnosis of IHS always is made with some uncertainty. Comparison ofNarcolepsy and Idiopathic Hypersomnia NARCOLEPSY IHS Symptoms Nocturnal sleep MSLT 298 Cataplexy in 70% Hypnagogic hallucinations Sleep paralysis Fragmented sleep Short sleep latency Short REM latency Refreshing naps Mean sleep latency < 5 min 2: 2 REM onsets in 5 naps No cataplexy Hypnagogic hallucinations Sleep paralysis Normal to long total sleep time Normal sleep architecture REM latency not decreased Non-refreshing naps Mean sleep latency < 10 min < 2 REM onsets

Clinical Pearls I. The diagnosis of idiopathic hypersomnia depends on exclusion of other disorders causing excessive daytime sleepiness, including sleep apnea (and the upper airway resistance syndrome), periodic limb movements in sleep, narcolepsy, affective disorders, stimulant withdrawal, and insufficient sleep. 2. History of recent head trauma suggests the posttraumatic hypersomnia syndrome. 3. When polysomnography fails to explain the recent onset of hypersomnia in an older patient, exclude neurologic disease (e.g., brain tumors, hydrocephalus), medical illness, medication side effects, and depression. 4. Patients with idiopathic hypersomnia have a mean sleep latency < 10 minutes, but less than two REM onsets in five naps on an MSLT. REFERENCES I. Guilleminault C. van den Hoed J. Miles L: Posttraumatic excessive daytime sleepiness. Neurology 1983; 33: 1584-1589. 2. Baker TL. Guilleminault C. Nino-Murcia G. Dement We: Comparative polysomnographic study of narcolepsy and idiopathic central nervous system hypersomnia. Sleep 1986; 9:232-242. 3. Aldrich MS: The clinical spectrum of narcolepsy and idiopathic hypersomnia. Neurology 1996; 46:383-40 I. 4. Billiard M: Idiopathic hypersomnia. Neurol Clin 1996; 14:573-582. 299

Sleep starts (hypnic jerks) Nightmares (REM anxiety attacks) FUNDAMENTALS OF SLEEP MEDICINE 19 PARASOMNIAS Determining a cause for abnormal movements or behavior during sleep is often a challenging problem for sleep physicians. A parasomnia is a motor, verbal, or experential phenomenon that occurs during sleep and is often undesirable. Evaluation of these nocturnal "spells" begins with a detailed history of the nature, age of onset, and time of night of the episodes. Also explore factors (sleep deprivation) and medications that may have affected the behaviors. Simultaneous video and sleep monitoring can be of great value in evaluating parasomnias. Today most digital polysomnography equipment manufacturers offer digital video recording or a method of time stamping or synchronizing analog video recording. Additional electrodes are needed to evaluate the possibility of seizure activity (see Fundamentals 20). For the REM behavior disorder, monitoring of hand muscle EMG (flexor digitorum) is often performed in addition to right and left tibialis anterior (leg) EMGs. One problem in monitoring parasomnias is that they typically do not occur every night. Multiple nights of study may be needed. Differential Diagnosis of Unusual Behavior Associated With Sleep DIAGNOSIS USUAL SLEEP STAGE Normal Sleep Phenomena Sleep onset REM» NREM Parasomnias Sleep walking (somnabulism) Sleep terrors Confusional arousal Sleep talking (somniloquy) REM behavior disorder Parasomnia overlap disorder Bruxism Enureis Panic attacks Posttraumatic stress syndrome NREM NREM NREM NREM and REM REM NREM and REM NREM (stage 2) NREM and REM (random) Psychiatric Disorders NREM (transition stage 2 to stage 3) REM and NREM Seizure Disorders Nocturnal seizures NREM > Wake >REM Possible Seizure Disorders Nocturnal paroxysmal dystonia Episodic nocturnal wandering Hypnicjerks (sleep starts) are brief, total-body jerks that occur at sleep onset. These are entirely normal. Nightmares (REM anxiety attacks) are unpleasant dreams that can awaken the individual. Both of these phenomena are present in most normal individuals. Sleep walking, sleep terrors, and confusional arousals occur out of NREM sleep (see Patients 94 and 95), are more common in childhood, frequently have a family history, and often decrease in frequency with increasing age. The REM behavior disorder 300

(see Patient 96), in contrast, is more more common in men in their 50s. Loss of muscle hypotonia during REM sleep enables the patient to act out his or her dreams. In the parasomnia overlap disorder, patients present with elements of both night terrors/sleep walking out of NREM sleep and the REM behavior disorder out of REM sleep. Bruxism, grinding and gnashing of teeth during sleep (see Patient 97), is most common in stage 2 NREM sleep. Enuresis (bed wetting) can occur in any stage of sleep. Sleep talking (somniloquy) is the utterance of speech or sounds during sleep without simultaneous, subjective, and detailed awareness of the event. Sleep talking can occur out of any stage of sleep. Patients with sleep terrors often talk out of slow wave sleep. The REM behavior disorder can be associated with talking as well as violent utterances. Sleep talking can also occur in the obstructive sleep apnea syndrome during arousals from sleep. Nocturnal paroxysmal dsytonia is a syndrome characterized by coarse movements associated with tonic spasms that often occur multiple times per night. The episodes can be violent or associated with vocalization. There is now evidence that this syndrome may in fact be a seizure disorder. Carbamazepine (400-600 mg) at bedtime is an effective treatment for the disorder. Episodic nocturnal wanderings may present with symptoms similar to sleep walking and sleep terrors. Patients may wander, vocalize, and show violent behavior during sleep. These patients respond to antiseizure medications, and this syndrome may represent automatisms secondary to an epileptic disorder. Among psychiatric disorders, panic attacks can present as nocturnal spells. While nocturnal panic attacks usually occur in patients with known daytime attacks, a few patients may have panic attacks only at night. Post-traumatic stress disorder patients also may complain ofterrifying dreams and may awaken with episodes similar to night terrors. They usually are not confused on awakening and may have vivid dream recall. Common Features Associated with Parasomnias SLEEP CONFUSIONAL SLEEP NIGHTTERROR AROUSAL WALKING MARES RBD SEIZURE Time of night Early Early Early Late Late Any Sleep stage at SWS SWS SWS REM REM NREM start > REM Screams Yes No No Rare Rare (talking Rare or yelling) Autonomic Extreme Minimal Minimal Mild Mild Mild activation Walking No No Yes No Rare Common Confusion after Usual Usual Usual Rare Rare Usual episode Age Child Child Child Child!Adult Adult Adult Episodes also in No No No No No Usual Wake CNS lesion No No No No Can occur Common RBD = REM behavior disorder. SWS = slow wave sleep (stages 3 and 4). > = more likely than Adapted from Broughton RJ: NREM arousal parasomnias. InKryger MH. Roth T, Dement WH (eds): Principles and Practice of Sleep Medicine. 3rd ed. Philadelphia, WB Saunders. 2000. pp693-706. Not every patient with a parasomnia requires a sleep study; there are important factors to consider in the decision. Moreover, the type of monitoring will depend on the clinician's impression ofthe most likely diagnosis as well as the local resources. Indications for Sleep Monitoring in Possible Parasomnias Very frequent events Unusual age of onset Injury has occurred to patient or bedpartner Associated with daytime sleepiness Medical/legal issues 301

REFERENCES I. Broughton RJ: NREM arousal parasomnias. ln Kryger MH. Roth T. Dement WH (eds): Principles and Practice of Sleep Medicine. 3rd ed. Philadelphia, WB Saunders. 2000. pp 693-706. 2. Mahowald MW. Schenck CH: REM sleep parasornnias. In Kryger MH. Roth T, Dement WH (eds): Principles and Practice of Sleep Medicine. Philadelphia. WB Saunders, 2000. pp 724-741. 3. Foldvary N: Video-encephalography/polysomnography for monitoring nocturnal events. In Lee-Chiong TL. Sateia MJ, Carsakadon MA (eds): Sleep Medicine. Philadelphia, Hanley and Belfus, 2002, pp 681-688. 302

PATIENT 94 A 20-year-old man with severe "nightmares" A 20-year-old man was evaluated for complaints of awakening with screaming and severe sweating once or twice a week, usually before 3 AM in the morning. According to his roommate, the patient was diaphoretic and difficult to communicate with during these episodes. Total amnesia for the events was reported. The patient admitted that he had severe nightmares as a child, but that they were infrequent until recently. The events typically occurred after he had missed his normal amount of sleep the night before because of social events or studying for tests. Physical Examination: Normal. Question: Are these episodes really nightmares? What is the correct diagnosis? 303

Diagnosis: Sleep terrors (pavor nocturnus). Discussion: Sleep terrors, also called night terrors or pavor nocturnus, consist of sudden arousal, usually from stage 3 or 4 sleep, accompanied by a scream or cry and manifestations ofsevere fear (behavioral and autonomic). The affected individual typically is confused, diaphoretic, and tachycardic, and he or she frequently sits up in bed. It is difficult or impossible to communicate with a person having a night terror, and total amnesia for the event is usual. Night terrors typically occur in prepubertal children (up to 3%) and subside by adolescence; they are uncommon in adults. Some studies have suggested that the presence of night terrors in adulthood indicates psychopathology. However, other authorities disagree with this conclusion. Patients may sleepwalk during episodes of night terrors. Thus, many consider sleepwalking and night terrors to be one syndrome with a spectrum of manifestations. Both are considered disorders of arousal. In adults, night terrors/sleep walking can occur out of stage 2 NREM sleep and during the second part of the night. Stress, febrile illness, sleep deprivation, and heavy caffeine intake have been identified as inciting agents for night terrors. Slow wave sleep rebound, such as occurs with nasal CPAP treatment of GSA, also has been associated with episodes of night terrors. Confusional arousals are also somewhat similar to night terrors. They also tend to occur out of stages 3 and 4 sleep. Individuals are very confused following spontaneous or forced arousals from sleep. In contrast to night terrors, there is no autonomic hyperactivity, signs of fear, or blood-curdling screams. The differential diagnosis of night terrors includes nightmares, nocturnal seizure activity, the REM behavior disorder (RBD), and the posttraumatic stress syndrome. Nightmares (dream anxiety attacks) and RBD occur within REM sleep and are more common in the second part of the night. RBD usually does not begin until after age 40. Differentiation from partial complex seizures is difficult without complete EEG monitoring. Seizures tend to be more stereotypic and may occur during the day. Patients with nightmares, the posttraumatic stress syndrome, and RBD typically can relate complex dream mentation that promoted the event. Polysomnography usually is not required to evaluate night terrors unless the episodes are frequent, violent, or have the potential to result in self-injury. When polysomnography is performed, inclusion of video monitoring (synchronized if possible) is ideal. If seizures are suspected, then a complete clinical EEG montage is needed. When a night terror is captured, it appears as a sudden arousal from slow wave sleep. The EMG amplitude is greatly increased, and alpha waves are present; however, persistent slow wave activity also is noted. If the episodes of night terrors are infrequent, treatment beyond simple environmental precautions is unnecessary. Several medications, including benzodiazepines, tricyclic antidepressants, and selective serotonin reuptake inhibitors, have been used with some success. Avoidance of inciting agents is recommended. The present patient seemed emotionally welladjusted. He was told that irregular sleep patterns were probably responsible for the reappearance of the episodes. As he wanted to avoid medication at all costs, the patient diligently maintained good sleep habits and reported only one minor episode every 2-3 months. Clinical Pearls I. Night terrors usually occur from slow wave sleep and are more common in the first part of the night in children. 2. In adults, night terrors can occur from stage 2 sleep and in the second half of the night. 3. The persistence of night terrors into adulthood or onset in adulthood is not necessarily evidence that psychopathology is present. 4. Unlike nightmares and RBD, patients with night terrors cannot relate dream mentation associated with the event. 5. Night terrors have been described in adults during nasal CPAP treatment of GSA. 304

REFERENCES I. Pressman MR, Meyer TJ .. Kendrick-Mohamed J, et al: Night terrors in an adult precipitated by sleep apnea. Sleep I'!'!J: 18:773-775. 2. Guilleminault C, Moscovithc A, et al: Forensic sleep medicine: Nocturnal wanderings and violence. Sleep 19'15: 18:740-748. 3. Crisp AH: The sleepwalking/nightterrors syndrome in adults. Postgrad Med J 1996; 72:599-604. 4. Mahowald MW, Schenck CH: NREM sleep parasomnias. Neurol Clin 1996: 14:675-696. 5. Broughton RJ: NREM arousal parasornnias. In Kryger MH, Roth T, Dement WH (eds): Principles and Practice of Sleep Medicine, 3rd ed. Philadelphia, WB Saunders, 2000, pp 693-706. 305

PATIENT 95 A 25-year-old woman walking in her sleep A 25-year-old woman was referred for evaluation of sleepwalking. She had a history of sleepwalking beginning at age 10, at which time she had about five episodes a month. These gradually decreased until they were uncommon (one or two a year) from age 13 on. However, recently the episodes had been occurring weekly. During this time she had been sleeping poorly because of stress related to college. She sometimes got as little as 3 hours of sleep because of studying for examinations. The patient sought evaluation because she had read that persistence ofsleepwalking into adulthood implied psychiatric problems. She denied symptoms of depression and anxiety and did not abuse alcohol or stimulant medications. Physical Examination: Normal. Figure: The tracings below occurred when the patient was noted to sit up in bed and pick at the sheets. When the technician entered the room and tried to talk to the patient, she did not respond. Question: Should the patient be referred for psychiatric evaluation? ROC-A 1 LaC - A 2 chin EMG I 50 uv 306

Answer: No. Sleepwalking in adults is often not associated with psychopathology. Referral is indicated only if the history suggests an emotional problem. Discussion: Sleepwalking (somnambulism) is defined as a series of complex behaviors that are initiated during slow wave sleep and result in ambulation during sleep. Activity can vary from simply sitting up in bed to walking. Patients usually are difficult to awaken during these episodes, and if awakened, are confused. Talking during sleep (somniloquy) can occur simultaneously. In children, sleepwalking usually occurs during the first third of the night, when slow wave sleep is present. However, recent studies in adults have recorded episodes beginning in stage 2 NREM sleep and frequently in the second half of the night. Episodes in children are rarely violent, and movements often are slow, but episodes in adults can be frenzied and violent. Sleepwalking may be terminated by the patient returning to bed or by the patient simply lying down and continuing sleep out of bed. Typically, there is total amnesia for the episodes. Sleepwalking can occur as soon as children can walk, but peaks between the ages of 4 and 8. The onset of sleepwalking can occur in adulthood; however, most adult sleepwalkers had episodes during childhood. Sleepwalking usually disappears in adolescence. Fever, sleep deprivation, and certain medications (e.g., phenothiazines, tricyclic antidepressants, lithium) can precipitate the events. Sleepwalking during slow wave sleep rebound has been reported in a patient with obstructive sleep apnea (OSA) treated with nasal CPAP. While it was once thought that persistence of sleepwalking into adulthood was a manifestation of underlying psychopathology, several studies have found that at least 50% of adult sleepwalkers have no psychopathology. Sleepwalking is considered a disorder of arousal. Because there is some overlap with night terrors, some refer to the syndrome as sleepwalking/night terrors. Although polysomnography rarely is performed to evaluate cases of sleepwalking, the classic finding is a sudden arousal occurring in slow wave sleep. During the prolonged arousal, there usually is tachycardia and persistence of slow wave EEG activity-despite the presence of high-frequency EEG activity and an increase in EMG amplitude. Evaluation of parasomnias in the sleep laboratory is best performed with simultaneous video recording to document body movements. Sleep monitoring is indicated when sleepwalking has resulted in bodily injury or has failed to respond to simple measures. The differential diagnosis of sleepwalking includes the REM behavior disorder, seizure disorders (such as temporal lobe seizures), and dissociative states. The walking associated with the REM behavior disorder occurs during REM sleep usually in the later part of the night. When awakened, subjects generally are not confused and may relate a dream in which they were moving. Patients with nocturnal seizures also may have seizures during wakefulness. However, if seizure activity only occurs during sleep, this diagnosis is more difficult. Diagnosis of temporal lobe seizures may not be possible with conventional scalp electrodes. In one study of 100 adults referred for evaluation of sleep-related injury, 54 had night terrors/sleepwalking, 36 had the REM behavior disorder, and two had nocturnal seizures. Interestingly, 33% of the group with sleepwalking had an age of onset after age 16, and 70% had episodes arising from both stages I and 2 as well as slow wave sleep. The sleepwalking behaviors were variable in duration and intensity. Psychological evaluation identified 50% with psychiatric disorders (e.g., depression, substance abuse, dysthymia). However, 50% of the group had no identifiable psychopathology. The main complications of sleepwalking are social embarrassment and danger of self-injury. Violent behavior (homicide) has been reported. The treatment of sleepwalking includes environmental precautions (e.g., closed doors and windows, sleeping on the first level, avoidance of precipitating causes such as sleep deprivation) and reassurance. If the episodes seem to require medication, then benzodiazepines or tricyclic antidepressants may be tried. Clonazepam 0.5-2 mg qhs or temazepam 30 mg qhs is commonly prescribed. Medications should be given early enough before bedtime so that sleepwalking in the first slow wave cycle is prevented. Selective serotonin reuptake inhibitors also have been reported to work. In the present case, the irregular sleep schedule and sleep deprivation were the most likely causes of the return of sleepwalking. The sample sleep tracing shows evidence of arousal from slow wave sleep. Note that some slow wave activity still is present, despite the large amount of high-frequency EEG activity. As the history did not suggest psychopathology, referral for psychological evaluation was not deemed necessary. The patient was instructed to keep a regular sleep schedule and to take environmental precautions. She was reassured that the return of her sleepwalking did not necessarily imply that she had emotional problems. After she began following instructions, the sleepwalking episodes decreased to less than one every 2-3 months. 307

Clinical Pearls I. Not all adults with sleepwalking had episodes as children. 2. Sleepwalking classically occurs during slow wave sleep in the first part of the night (especially in children). However, in some adults onset can occur in stage 2 sleep and in the second half of the night. 3. The persistence of sleepwalking into adulthood does not necessarily imply underlying psychopathology. 4. Prior sleep deprivation with resulting slow wave sleep rebound (as with nasal CPAP treatment for OSA) can trigger episodes of sleepwalking. REFERENCES I. Schenck CH, Milner DM, Hurwitz TD, et al: A polysomnographic and clinical report of sleep related injury in 100 adult patients. Am J Psych 1989; 146: 1166-1 172. 2. Kavey NB. Whyte J. Resor SA, et al: Somnabulism in adults. Neurology 1990; 40:749-752. 3. Millman RF. Kipp GJ. Carskadon MA: Sleepwalking precipitated by treatment of sleep apnea with nasal CPAP. Chest 1991; 99:750-751. 4. Mahowald MW, Schenck CH: NREM sleep parasomnias. Neural Clin 1996; 14:675-696. 308

PATIENT 96 A 55-year-old man with violent dreams A 55-year-old man complained of violent movements during sleep. The problem had begun 14 months ago. His movements tended to occur during the last half of the night and varied from simply moving his arms to hitting his wife. On some occasions, the patient got up from the bed. When awakened he was not confused, but only rarely remembered dream content. During some of the episodes, the patient also screamed or talked about harming someone. The episodes seemed to be worse after periods of interrupted sleep or a change in sleep schedule. There was no history of head trauma or change in intellectual functioning, motor strength, sensation, or coordination. There was no history of sleepwalking (somnambulism) during childhood. Physical Examination: Normal. Sleep Study: The following tracing was noted. Question: What is your diagnosis? patient yelling C4-Al C3-A2 02-Al ROC-Al LOC-A2 EKG chinEMG Leg EMG FIGURE I 309

Diagnosis: REM sleep behavior disorder. Discussion: The REM behavior disorder (RBD) is characterized by a loss of the normal muscle hypotonia associated with REM sleep or an overactivation of phasic REM phenomenon; thus, dreams can be "acted out." Limb and body movements often are violent (e.g., hitting a wall, kicking) and may be associated with emotionally charged utterances. The movements can be related to dream content ("kicking an attacker"), but the patient may not remember associated dream material when awakened during an episode. Serious injury to the patient or the bed partner can result from these episodes, which typically occur one to four times a week. The median age of onset is about 50 years, and a milder prodrome of sleeptalking, simple limbjerking, or vividly violent dreams may precede the full blown syndrome. Because the episodes occur during REM sleep, they are most common during the early morning hours (the second half of the night). The differential diagnosis of abnormal movement and behavior arising from sleep includes sleeprelated seizure activity, periodic limb movements in sleep, sleepwalking, night terrors, nocturnal panic attacks, nightmares, and the posttraumatic stress disorder. In contrast to RBD, sleepwalking (and variants) classically occurs during slow wave sleep (stages 3 and 4) and, hence, is most common in the early portion of the night. Unlike RBD, most adults with sleepwalking had episodes during childhood. When patients are awakened during sleepwalking or night terror episodes, they are quite confused and tend to have no memory of dream content. If content is remembered, usually it is not as complex as a typical dream. However, note that recent studies of sleepwalking and night terrors in adults have shown that episodes can begin in stage 2 sleep and during the second part of the night. In addition, the separation between sleepwalking/night terrors and RBD is not absolute-some patients have violent behavioral episodes occurring in both NREM and REM sleep (mixed disorder, or parasomnia overlap disorder). Although both nightmares and the posttraumatic stress syndrome can be associated with violent or terrifying dream content and arousal from sleep, complex body movements are uncommon. Nocturnal seizure activity usually occurs in NREM sleep, and behaviors typically are more stereotyped and less complex than in RBD. A few patients with abnormal EEG activity and complex and violent behavior have been described. These patients responded to anti-seizure medication. In animal experiments, lesions in the pons can result in body movements during REM sleep. Thus, 310 degeneration of the brainstem is believed to be one possible cause of RBD in humans. However, even with extensive evaluation, about 60% of cases are idiopathic. Others are associated with multiple sclerosis, subarachnoid hemorrhage, dementia, ischemic cerebrovascular disease, and brain stem neoplasm. In one study, almost 40% of patients with idiopathic RBD later developed Parkinson's syndrome. An acute form of RBD can occur after withdrawal from REM suppressants, such as ethanol. Drug-induced cases also have been reported, with the use of tricyclic antidepressants or selective serotonin reuptake inhibitors (SSRIs; e.g., fluoxetine). Polysomnography mayor may not reveal an episode, as most patients do not have nightly attacks. Some sleep centers routinely perform at least three serial sleep studies. Simultaneous video and sleep recording (including both leg and arm EMG) isrecommended. An episode is evidenced by bursts of limb movement or persistent augmented chin and/or leg EMG activity during REM sleep. At first glance, the episode may appear as stage Wake (eye movements and elevated chin EMG; Fig. I). Clues to the fact that abnormal REM sleep is present include phasic EMG bursts in the limbs and alterations in airflow associated with bursts of eye movements. The heart rate also may remain constant despite the sudden appearance of increased EMG tone (as opposed to an awakening). In other episodes of RBD the chin tone may remain fairly normal, but large increases in the phasic activity of the limbs are seen. In Figure 2, the chin EMG is reduced, allowing easy recognition of REM sleep. However, the leg EMG activity is greatly increased. It differs from PLM activity because it is prolonged. Also note the many fine spikes in the leg EMG, characteristic of phasic activity in REM sleep. You would not make the diagnosis ofRBD solely from a tracing such as this one, which can occur during REM rebound with nasal CPAP treatment. An associated clinical picture must be present. A detailed neurologic evaluation of patients suspected of having RBD is indicated and should include MRI of the brain, a full clinical EEG, and a thorough neurologic examination. Successful treatment of RBD has been achieved with clonazepam 0.5-2.0 mg in approximately 90% of patients. Clonazepam dramatically reduces episode frequency. However, occasional breakthrough attacks can occur, and environmental precautions (e.g., bed mate sleeping in a separate bed, closed windows and doors) are essential. Successful treatment of RBD has also been reported with carbamazepine.

In the present case, the patient was noted to awaken yelling from REM sleep (Fig. I). Shortly before this episode, the chin EMG and leg EMG showed increased phasic activity during REM sleep. Note the fine spikes in the EMG of both legs and chin. Also note that tachycardia is not present following awakening (typical of RBD). The present patient responded well to clonazepam, and his episodes of violent movement during sleep became infrequent, occurring only once every 2 months and at much milder intensity. The patient's wife began sleeping in a separate bed as a precaution. C4-Al 02-Al ROC-Al LOC-A2 EKG chin EMG airflow chest abdomen R. L Leg EMG 5002 , I i \ \ , , , I I '" I 1 II 'I I I FIGURE 2 Clinical Pearls I. Episodes of violent limb or body movements during sleep starting in adulthood suggest the REM behavior disorder (RBD). 2. Although most RBD is idiopathic, episodes can occur with tricyclic antidepressants or SSRIs. 3. A detailed neurologic examination is essential to rule out an associated neurologic problem. Some patients with idiopathic RBD later develop Parkinson's disease. 4. Polysomnography with video recording can help confirm the diagnosis. If seizures are suspected, a full clinical EEG montage should be monitored. 5. Treatment with clonazepam usually is successful, although breakthrough episodes can occur. Environmental precautions are essential. 6. An overlap parasomnia disorder that involves both NREM and REM sleep has been described. Manifestations of both sleep terrors/sleep walking and RBD are present. REFERENCES I. Schenck CH, Bundlie SR, Patterson AL, et al: Rapid eye movement sleep behavior disorder: A treatable parasomnia affecting older males. JAMA 1987; 257:1786-1789. 2. Schenck CH, Mahowald MW: A polysomnographic, neurologic, psychiatric and clinical outcome report on 70 consecutive cases with REM sleep behavior disorder: Sustained clonazepam efficacy in 89.5% of 57 treated patients. Clev Clin J Med 1990; 57(Suppl); 10-24. 3. Bamford CR: Carbamazepine in REM sleep behavior disorder. Sleep 1993; 16:33-34. 4. Schenck CH, Bundlie SR, Mahowald MW: Delayed emergence of a parkinsonian disorder in 38% of 29 older men initially diagnosed with idiopathic rapid eye movement sleep disorder. Neurology 1996; 46:388-393. 5. Schenck CH, Boyd JL, Mahowald MW: A parasomnia overlap disorder involving sleep walking, sleep terrors, and REM sleep behavior disorder in 33 polysomnographically confirmed cases. Sleep 1997; 20:972-981. 311

PATIENT 97 A 50-year-old man with an interesting chin EMG A 50-year-old man underwent sleep monitoring because of a history of heavy snoring. The patient had originally seen a dentist for complaints of soreness of his teeth and temporomandibular joint in the morning. His wife reported that he snored, snorted, and stopped breathing at night. The patient denied recent increased stress in his life or any change in medication. He had gained about 20 pounds over the last 2 years and also noted some increase in daytime sleepiness (Epworth Sleepiness Scale score l4124-mild sleepiness). There was no history of sleep walking or episodes of acting out dreams. Sleep Study: The study showed heavy snoring. The tracing below shows 30 seconds of recording. In the prior epoch, some snoring was noted in stage 2 sleep. Although not shown, the heart rate did not change during this episode. Question: What is causing the rhythmic pattern noted on the tracing? FIGURE I 312

Answer: Bruxism. Discussion: Bruxism, defined as clinching or grinding of the teeth, is a common parasomnia. The phenomenon is caused by contractions of the masseter and temporalis muscles. Bruxism is common in young children before the adult teeth erupt, but also is quite common in adults. Prevalence in children ranges from 20% to 88% and declines over time to 3% in adults over age 60. Smokers are said to be more likely to brux than non-smokers. In the average patient there are about eight episodes of bruxing per night. Typically, the bedpartner reports the sound of teeth grinding. Patients often complain ofsoreness in the teeth, jaw, or temporomandibularjoint as well as headache or pain in the neck muscles in the morning. Rarely are they aware that they brux. Patients may be identified by their dentist, who notes abnormal tooth wear or peridontal disease. The cause of bruxism is not known, but the disorder has been associated with stress, malocclusion, and certain medications (serotonin reuptake inhibitors. levodopa). Polysomnography shows a rhythmic increase (about 1 per second) in EMG tone associated with unusual EEG and EOG activity resulting from muscle artifact, or shows transmission of very high EMG activity to the EEG and eye leads. If EMG electrodes are placed over the masseter muscles instead of the lower chin, the EMG activity may be even more prominent. Bruxism may occur in any sleep stage, but is most common in stage 2 NREM sleep or in tonic REM sleep. One study reported that bruxism was common in patients with sleepdisordered breathing, but that episodes of bruxism were not necessarily occuring at apnea termination. Figure 2 (from another patient) shows an episode of bruxism without the characteristic EMG pattern at the end of a period ofsnoring. The technician heard loud teeth grinding at this time and made a comment. Rhythmic changes in the EEG and eye leads are seen. No completely satisfactory treatment for bruxism exists. First, assess the degree of damage, perhaps with the help of a dentist. A mouth-guard protector or bite-splints can be used to prevent tooth damage. Correcting malocclusion may help in some cases. Medications that have been used include benzodiazepines, muscle-relaxers, levodopa, and botulinum toxin. Psychotherapy (stress reduction), relaxation therapy, biofeedback, and hypnosis have all been tried. No treatment has been demonstrated to be effective in a controlled trial. In the present patient, the tracing shows a rhythmic increase in EMG activity with rhythmic muscle artifact in the EEG and eye movements. The tracing has a "checkerboard pattern" that is said to be typical of bruxism. The episode did not occur out of REM sleep and was not associated with violent behavior. Therefore, the REM behavior disorder was not present. The episodes was not associated with screaming or body movements and autonomic hyperactivity. Thus, the illustrated episode is not a night terror. No spike and wave activity was found in the EEG during the night. The sleep technologist clearly heard the sound of grinding teeth during the episode. The patient was found to have snoring and positional sleep apnea with an overall AHI of 25/hr. He underwent treatment with nasal CPAP with resolution of snoring and daytime sleepiness. His dentist constructed a bite splint for tooth protection. The patient's wife reports that he still has some episodes, but the patient reported less mouth discomfort in the morning. C4-Al C3-A2 02-Al ROC-A1 LOC-A2 EKG chin snore nasal pressure FIGURE 2 313

Clinical Pearls I. Bruxism is a common parasomnia in both adults and children. 2. Polysomnographic features include a rhythmic increase (about I per second) in EMG activity, which can be transmitted in the EEG and EGG leads. 3. No treatment of bruxism has been documented to be effective in controlled trials. 4. Sleep technologist observations, such as hearing "teeth grinding," are very helpful in identifying abnormal EEG-EMG activity as bruxism. REFERENCES I. SjoholmTT, Lowe AA, MiyamotoK, et al: Sleep bruxism in patients with sleep-disordered breathing. Arch Oral BioI 2000; 45:889-896. 2. Eng-King T, Jankovic J: Treating severe bruxism with botulinum toxin. JADA 2000; 131:211-216. 3. Lavigne GJ, Manzini C: Bruxism. In Kryger M, Roth T, Dement W (eds): Principles and Practice of Sleep Medicine, 3'" ed. Philadelphia, WB Saunders Co, 2000. 314

FUNDAMENTALS OF SLEEP MEDICINE 20 Monitoring for Nocturnal Seizures The international 10-20 system for electrode placement is illustrated in Figure I. Each electrode is represented by a letter that indicates the underlying region of the brain (Fp = frontopolar, F = frontal, P = parietal, 0 = occipital, T = temporal) and by numerical subscripts indicating position. The odd subscripts are on the left and the even on the right. Note that the new terminology, in which T7, T8, P7, and P8 replace T3, T4, T5, and T6, is illustrated. In the new terminology, all electrodes in a given sagittal plane have the same subscript (F7, T7, P7), and most electrodes in the same coronal plane have the same letter (P7, P3, Pz, P4, P8). new 10-20 .~ ( /F7 F3 Fz F4 F8\\ ( T7 C3 Cz C4 T8 )) P7 P3 pz P4 P8 / '~// AI' bipolar (double banana) /-: ..--L-~ 0Pl Fp~'-. " I? V, '::1, / F7 F3 ~z F4 t \1 /(J J + , (T{C3CzC4 18 ) " ~~,L.~~J/ PZ Pi Pz P4 P8 , \a -L-.( . ,.~/ old 10-20 ~-..:~ /Fp1 Fp2' ( / F7 F3 Fz F4 F8\ ( T3 C3 Cz C4 T4 \)) T5 P3 Pz P4 T6 ) .,01 V "------ transverse bipolar Jf/FPl .r.:FP2~ (I1~T::::::::;Jj:10 \ \P7~P3~PZ~P4~P8 \ ", .~ 01~02 / FIGURE I A derivation is the voltage difference between electrodes. For example Fpl - F3 is the voltage difference between electrodes Fpl and F3. By convention, if Fpl is more negative than F3, the deflection is up. A set of derivations is called a montage. Montages are designed for a particular purpose in mind. Standard montages to detect respiratory events and stage sleep have already been illustrated (see Fundamentals 2). Bipolar montages compare two standard electrodes sequentially, covering the head in an AP or transverse direction (see table). Different labs display the electrodes in different sequences. In the referential scheme, each electrode is referenced to the ipsilateral mastoid electrode. In modern digital EEG recording, usually 315

all electrodes are recorded against a common reference. Then any two electrodes may be compared by subtracting the signals (F7-ref)-(P7-ref) = F7-P7. Digital recording also allows one to visualize multiple time scales. The usual polysomnography paper speed is 10 mm/sec or a 30-second page. In contrast, EEG for seizure evaluation is usually 30 mm/sec or a lO-second page. The latter allows detection of brief, sharply contoured waveforms that may signify seizure activity. Standard EEG Montages AP BIPOLAR ("DOUBLE BANANA") TRANSVERSE BIPOLAR REFERENTIAL Fpl-F7 F7-F3 FPI-AI F7-T7 F3-Fz F7-AI T7-P7 Fz-F4 T7-AI P7-01 F4-F8 P7-AI Fp2-F8 Ff9-T7 FP2-A2 F8-T8 T7-C3 F8-A2 T8-P8 C3-Cz T8-A2 P8-02 Cz-C4 P8-A2 C4-T8 Fpl-F3 T8-Ff10 Fpl-AI F3-C3 F3-AI C3-P3 P7-P3 C3-AI P3-01 P3-Pz P3-AI pz-P4 Fp2-F4 P4-P8 Fp2-A2 F4-C4 F4-A2 C4-P4 FPI-Ff9 C4-A2 P4-02 FP2-FfIO P4-A2 Ol-Ff9 Fz-Cz 02-Ff10 Fz-A2 Cz-Pz Cz-A2 If the capacity to add a few electrodes to traditional sleep monitoring exists, the ability to detect interictal epileptiform activity is increased. For example, four electrodes (F3, F4, T7, T8) could be added. The derivations F3-T7, T7-0 I, F4-T8, and T8-02 would add coverage over much ofthe frontal and temporal areas. These areas are the predominant foci of seizures occurring mainly during sleep. The terminology and identification of epileptic seizures and interictal activity is challenging for physicians without extensive training in EEG. A spike is defined as a transient (any isolated wave or complex that stands out compared to background activity), with a pointed peak and a duration of 20-70 milliseconds (Fig. 2). At polysomnography paper speed, or on a 30-second page, spikes look like a single vertical line. A sharp wave is a transient with a pointed peak and a deflection of 70-200 millseconds. Interictal activity (or interictal discharge) is defined as abnormal EEG activity that occurs between seizures. Because seizures do not always appear during recording, the physician reading an EEG searches for the "interictal footprint" of epilepsy - the epileptiform spike. Spikes represent abnormal brain activity that is seen as an area of negativity at the scalp. Spikes can be localized (negativity at the scalp over one area of the brain) or appear diffusely. Focal seizures usually, though not invariably, begin at the same location as the interictal spikes. The usual spike is followed by a slow wave. However, spikes should not be thought of as pre-seizure activity, because they more commonly follow than precede seizures. Localized spikes will show phase-reversal if the bipolar chains cross the area of the seizure focus. This may help differentiate them from artifact. For example, in Figure 3 negative spike activity is seen under electrode T7 (s = spike, w = wave). This results in downgoing deflections in F7-T7 as T7 is more negative than F7. This pattern reverses for T7-P7. If the spike focus is between two monitoring electrodes (F8 and T8; Fig. 4), the derivation connecting them may show little or no activity (F8-T8), and the derivations on either side will show phase reversal ([Fp2-F8] to [T8-P8]). If you notice interictal activity in the usual polysomnography display (C4-Al, etc.), remember that 316

10 sec page 30 sec page (30 rnrrvsec) (10 rnrrvsec) EEG speed PSG speed sharp waves Jl')vi v spikes 4~~ spike and wave ~~'~ ~~ 1 second polyspikes polyspike and wave ~~ 1 second FIGURE 2 the mastoid electrodes are not neutral. For example, a left temporal seizure focus may be picked up in AI. Thus C4-A I may actually show more activity from a left temporal focus than C3-A2. Also, at the usual polysomnographic time base (paper speed), you might miss spikes. If you are using digital equipment, a switch from a 30-second to a lO-second page view is indicated to scrutinize any suspicious paroxysmal activity. Seizure activity may be manifested by rhythmic activity of many types. While the spike and wave activity is the most familiar, the pattern of repetitive sharp waves of various frequencies is also common. On traditional sleep monitoring montages, ictal activity can even be mistaken for alpha or faster activity or artifact. In Figure 5, a portion of a IO-second page shows a spike and wave complex (SW) followed by rhythmic activity (RA) of 8-9 Hz. At point A, oral autornatisms were noted. The rhythmic activity is differentiated from normal alpha rhythm by being more prominent in the eye leads than occipital leads. This is in fact a portion of a frontal seizure manifested by oral automatisms and loss of responsiveness. You nasion Fp1 - F7 V F7 - T7 T7 - P7 s P7 - 01 \ '.j\right 1/ IA2 I I T8 P8 F8 Fp2 Fz F4 Cz C4 pz P4 01 02 . ---._--._---.> Fp1 F3 Seizure focus ~~7 left(( -(I} C3 \ I - \ A1 \ P7 P3 inion FIGURE 3 317

nasion Fp1 Fp2 Fp2 - F8 F7 F3 Fz F4 F8 -c)- F8 -T8 left T7 C3 Cz C4 T8 \1) right , ' rA1 P7 P3 pz P8 /A2 T8 - P8 P4 01 02 P8 - 02 inion FIGURE 4 might suspect the frontal nature, as the activity is higher amplitude in the eye leads (near the frontal lobes). A complete EEG montage documented a right frontal location (in Figure 5, note the slightly higher amplitude in ROC-A1 than LOC-A2). SW RA A LOC-A2 ROC-Al C3-A2 C4-Al 01-A2 02-Al ( ) 1 second FIGURE 5 REFERENCES I. Fisch BJ: Spehlrnans EEG Primer. New York, Elsevier, 1991. 2. American Electrophysiological Society: Guideline Thirteen: Guidelines for standard electrode position nomenclature, J Clin Neuorphysiol1994: 11:111-113. 3. Chesson AL. DellaBadia J: Seizure disorders and sleep. In Lee-Chiong TL, Sateia MJ, Carsakadon MA (eds): Sleep Medicine. Philadelphia, Hanley and Belfus, 2002, pp 521-531. 4. Foldvary N: Video.encephalography/polysomnography for monitoring nocturnal events. In pp 681-688. 318

PATIENT 98 A 55-year-old man with unusual movements during sleep A 55-year-old man was evaluated for arm movements and confusion during sleep. The episodes, which occurred once or twice weekly, had begun 1 year previously. During an episode the patient did not get out of bed, but was unresponsive. Afterwards, he was groggy. There was no history of daytime sleepiness or insomnia. The patient had no recall of the events in the morning. Physical Examination: Unremarkable. Sleep Study: No overt body movements or periodic limb movements (PLM) were noted. Figure: The following was noted on a tracing when the patient had just fallen asleep. Question: What is causing the body movements during sleep? C3-A2 01 - A2 ROC-A 1 LOC - A 2 EKG ~~~~~~~ ~v-i~ '-~~ 15 319

Diagnosis: Seizure activity. Discussion: Seizure disorders are part of the differential diagnosis of "nocturnal spells" -episodes of abnormal motor activity during sleep. Depending on the type of patients studied, as many as 10-40% of seizures occur exclusively or mainly during sleep. The incidence of nocturnal seizures has two peaks: one about 2 hours after bedtime and another between 4 and 5 am. Daytime seizures are most prevalent in the first hour after awakening. In general, all manifestations of nocturnal seizure disorders are much more common in NREM than REM sleep. Prior sleep deprivation activates seizures (see table below); therefore, patients often undergo clinical EEG monitoring in a sleep-deprived state to increase the likelihood of recording seizure activity. Seizures are classified as partial (focal) onset, arising from a localized area of the brain (with or without subsequent generalization), and generalized onset, arising from both hemispheres simultaneously. Partial seizures can become generalized and result in generalized tonic-clonic seizures. Focal seizure disorders are sometimes called sleeping epilepsies because they frequently are associated with interictal discharges or seizure activity in NREM sleep. Simple partial seizures result in focal motor, sensory, autonomic, or psychophysiologic manifestations without a change of consciousness. Complex partial seizures usually arise from the mesial or lateral part of the temporal lobe or adjacent parts of the frontal lobe. The symptoms consist of changes in the content of consciousness that reduce the patient's ability to interact with his or her surroundings. They can occur with only a change in conciousness or can include automatisms (repetitive movements which may be purposeful, but serve no obvious purpose in the actual situation). For example, lip smacking is a common automatism. Patients with complex partial seizures may have no recollection of the events or only partial memory. Patients may remember an aura, a subjective sensation such as a visual or olfactory disturbance that precedes the start of the event. Primary generalized epilepsies include idiopathic Generalized tonic-clonic (GTC) seizures, absence seizures (petit mal), and juvenile myoclonic seizures. GTC seizures consist of a sudden loss of conciousness, a tonic phase of intense muscle contraction, and then a clonic phasic consisting of bilaterally synchronous jerking of the entire body. After the seizure, there is a post-ictal period of disorientation lasting a variable amount of time. Absence seizures are manifested as a blank stare during which the patient is unresponsive. The characteristic waking EEG pattern is a 3-Hz spike and wave pattern. Absence seizures start in childhood and rarely persist into adulthood. Juvenile myoclonic epilepsy is a genetically determined condition involving myoclonic jerks in the arms shortly after awakening. Primary generalized seizures are sometimes are called awakening epilepsies because they commonly occur when the patient is in a drowsy state upon awakening from sleep. Patients with both absence and juvenile myoclonic disorders also can have GTC seizures. GTC seizures associated with these disorders usually occur on awakening. Seizure activity comprises interictal and ictal phases (see table at right, top). Interictal refers to transient focal or generalized discharges between seizure events. Ictal discharge refers to the event itself, which, depending on the type ofseizure, may be manifested by partial motor activity (limb jerking and twitching), a GTC seizure, myoclonic jerking, an absence seizure, or complex motor behavior. When these symptoms occur during sleep, they may not be recognized. PARTIAL EPILEPSY* Seizure Disorders Affected By Sleep PRIMARY GENERALIZED EPILEPSY • Frontal lobe epilepsy (FLE) Nocturnal frontal lobe epilespy Autosomal dominant FLE • Temporal lobe epilespy 'Can secondarily generalize 320 • Generalized tonic clonic seizures on awakening • Absence epilepsy (petit mal) • Juvenile myoclonic epilepsy

Typical Seizure Occurrence INTERICTAL DISCHARGE ICTAL DISCHARGE SEIZURE TYPE Primary generalized Focal *After awakening from NREM sleep NREM Common Common REM Rare Rare NREM Rare Common REM Rare Rare AFfER NREM* Common Possible Temporal and frontal lobe epilepsies are often mislabeled as othersleep-related conditions (e.g., PLMS, bruxism, sleepwalking). EEG-video monitoring during sleep can be useful in making the correct diagnosis. Of focal seizure disorders, approximately 20% have onset from the frontal lobes. The clinical manifestations offrontal lobe seizures may vary depending on localization of the epileptic focus. Typically, patients with electrographic onset from the midline regions will have supplementary motor cortex involvement, eliciting complex motor manifestations such as dystonic posturing, vocalizations, or speech arrest, with variable loss of consciousness and minimal post-ictal confusion. A classic manifestation is the ''fencing posture" with the head turned toward an outstretched arm. Seizures arising from the supplemental motor area may involve thrashing with maintenance of consciousness, and often are misdiagnosed as psychogenic seizures. Seizures originating from the orbitofrontal areas and the cingulate gyrus often resemble those originating from the temporal lobes, with staring, nonresponsiveness, and automatisms. Thus, the differentiation of frontal lobe and temporal lobe seizures (see table below) is not absolute. In addition, seizures originating from the cingulate gyrus may also have autonomic features such as tachycardia, tachypnea, pallor, and sweating. Seizures originating from the dorsolateral frontal lobes may have minimal symptoms or have motor manifestations depending on the extend of spread of the seizure activity. On the contrary, seizure onset in the posterior frontal lobes from the primary motor cortex may have discrete motor manifestation that have a jacksonian march (begins in the distal muscles of an extremity and moves up the extremity). Unfortunately, patients with frontal lobe seizures frequently do not exhibit interictal EEG activity. Temporal lobe seizures begin focally and impair consciousness. Staring, orofacial or limb automatisms, and head and body movements frequently occur. Temporal lobe seizures are more common in NREM sleep, but also occur at the transition from NREM to REM sleep. Interictal activity can often be seen even using traditional sleep EEG monitoring as the mastoid electrodes are near the temporal lobes. Of note, no abnormal EEG activity may be observed in some patients with temporal lobe epilepsy using scalp electrodes. Partial seizures with complex automatisms have been described in a few patients; unusual sleepwalking episodes, vocalization, and violent behavior were noted. These patients responded to anti-epileptic medications. Comparison ofFrontal and Temporal Lobe Epilepsy* OCCUR MAINTENANCE ONLY DURING OF SLEEP CONCIOUSNESS BODY MOVEMENTS POST-ICTAL CONFUSION Frontal Lobe Temporal Lobe Usually Often Usually Conciousness impaired Tonic posturing ("fencing posture") Automatisms (lip-smacking) Minimal Sometimes *Note that these differences arenotabsolute-see text. 321

Diagnosis of nocturnal seizures requires a full EEG montage and, ideally, simultaneous synchronized video recording. Temporal lobe epilepsy is especially difficult to document and often requires intracranial electrodes. Sometimes a diagnosis is elusive, and an empiric trial of anti-epileptic medications is needed. In routine clinical EEG monitoring, the paper speed is faster (30 mm/sec) and the EEG amplitude is less sensitive than in sleep monitoring. Therefore, on routine sleep monitoring interictal activity appears sharper and often with a higher amplitude. Unlike usual sleep patterns, interictal activity often manifests as repetitive occurrences of nearly identical patterns. If a computerized system is used to record sleep, the tracing can be reviewed at a simulated clinical EEG paper speed (30 mm/sec or lO-second page). The differential of nocturnal seizures includes bruxism, PLMS, night terrors, sleepwalking, and REM behavior disorder. General motor activity arising from seizures is simpler and more stereotypic than motor activity associated with sleepwalking, night terrors, and REM behavior disorder. In the present case, a routine sleep study was initially performed (see Figure). This showed no PLMS, but frequent spike and wave complexes (5) occurred during NREM sleep. Note that the activity does not coincide with the EKG complexes. The complexes were not the usual vertex sharp waves, which are upgoing and sporadic. The spike and wave complexes in the tracing occurred in derivations with A2 (near the right temporal lobe). The patient was referred to a neurologist, and a full clinical EEG confirmed the presence of interictal activity in the right temporal area. The patient was treated with carbamazepine, with resolution of the episodes. An MRI of the brain showed increased size of the temporal horn of the right lateral ventricle, as well as scarring. The damage to the temporal lobe was felt to have developed from repeated seizure activity. Clinical Pearls I. Seizures are part of the differential diagnosis of abnormal motor behavior occurring during sleep. 2. Optimal diagnosis of nocturnal seizures requires a full EEG montage, with simultaneous video recording if possible. 3. Episodes of repetitive, high-amplitude, sharp EEG activity during a routine sleep study could represent interictal seizure activity. 4. Identification of interictal activity on a polysomnogram should prompt a more extensive evaluation for epilepsy even if a frank seizure is not recorded. 5. Frontal lobe and temporal lobe epilepsies are often called sleeping epilepsies because they may occur more frequently or only at night. 6. Complex partial seizures are associated with a change in the content of conciousness as well as localized motor or sensory manifestations. REFERENCES l. Guilleminault C. Moscovithc A. et al: Forensic sleep medicine: Nocturnal wanderings and violence. Sleep 1995; 18:740-748. 2. Malow BA: Sleep and epilepsy. Neural Clin 1996; 14:765-789. 3. Shouse MN, Martins da Silva A, Sarnmaritano M: Circadian rhythm, sleep, and epilepsy. J Clin Neurophysiol 1996; 13:32-50. 322

, , I , I , I I I I , I , I , I , i , I I , , I I I I I I , PATIENT 99 A 60-year-old man with a rhythmic EEG pattern during a CPAP titration A 60-year-old man was evaluated for snoring, daytime sleepiness, and awakening with body jerks during sleep. Obstructive sleep apnea was noted during the initial portion of the sleep study, and a CPAP titration was initiated. During the titration, paroxysmal rhythmic activity lasting 15-30 seconds was noted. Physical Examination: Normal. Sleep Study: The 30-second tracing shown below was noted during stage 1 sleep. The second figure shows a lO-second segment of the same epoch. No unusual body movements occurred during this activity. Question: What is your diagnosis? LOC-A2,.".".- ,..,....--.....-.......----......---...- ----"II'I ROC-A1 -" ....,..'--..-..........-.r_....... ............,.---' ....-'......----...,....~"o'dr'f'"ttr.lliIIP'IttN'.,,\Jl C3-A2 C4-A1 """'...-_--...__........_ ....... V-.A_""-__""-_-_~_"'iN\I',. 01-A2 -----.......- .---.r- .-J__.-..-......",~iIilI'I~ftw,A~rlI 02-At chin EKG Legs airflow chest abdomen 5002 I I I I , , , I I I I I I , , I , , I , I I I I I I I ., I I I I i I I I I , 94% Spikeand wave J ... 94% 323

Answer: Seizures (spike and wave activity). Discussion: Spike and wave activity is a common form of ictal discharge. Spikes are transient waves that are 20-70 milliseconds in duration. At conventional polysomnography paper speed of 10 mm/sec (30-second page), spikes appear as nearly vertical lines or may be difficult to separate from adjoining slow wave activity. They are usually surfacenegative (usually upward deflections) in central or eye leads. Spikes are frequently followed by slow waves. Isolated spike discharges are often seen in patients with seizures (interictal activity), but may occasionally be seen in patients with a family history of seizures who never have clinical seizures. About 10-40% of seizure activity occurs exclusively during sleep. Seizure thresholds are highest during REM sleep, followed by wakefulness, and then NREM sleep. Thus, the likelihood of seizure activity is NREM > Wake> REM. Generalized tonic clonic (GTC) seizures are most common soon after awakening in many patients. Nocturnal seizures present in many ways (see Table), including unusual behaviors or multiple arousals leading to complaints of insomnia or daytime sleepiness. Seizure-associated behaviors are often recurrent, stereotypic, and/or inappropriate. The patient usually has no recall of the events. Seizures can also result in automatisms and may include nocturnal wandering or other behavior that mimics sleepwalking, sleep terrors, or REM behavior disorder. Screaming, vocalizations, and violent autisms with the possibility of injury may also be seen. Seizures may also present as a recurrent nightmare or as isolated symptoms such as choking or laryngospasm. However, nocturnal seizures can also be entirely asymptomatic. Presentations ofNocturnal Seizures Frequent arousals Daytime sleepiness Complaints of insomnia Nocturnal "spells" -parasomnia Asymptomatic The non-neurologist may not be familiar with the elements of a clinical history that are important for differentiating seizures from other parasomnias. Abnormal focal movements, auras (a subjective sensation such as a smell or visual disturbance that precedes attacks), exact description of the attack, non-responsiveness during the episode, and the presence and absence of post-ictal confusion are all important historical elements. Unfortunately, if seizures occur only at night, the patient may not be observed doing anything unless 324 body movements or sounds awaken the bedrnate. Partial complex seizures can result in wanderings. However, patients are poorly responsive during the attacks and usually do not report complex dreams on awakening as in REM behavior disorder. Nocturnal seizures are also more likely to occur out of NREM than during REM sleep. Distinguishing seizure activity from artifact on a polysomnogram can be difficult for the non-neurologist. The ability to view the tracing at a faster paper speed (IO-second page) is very helpful in determining the morphology and frequency of the activity. This is an advantage of digital recording systems. If a spike and wave pattern is seen during the rhythmic activity in question, this is virtually diagnostic for seizure activity. Bursts of delta or theta activity can look somewhat similar, but do not have spikes. They also are rarely as regular as seizure activity. Electrode popping can give high-amplitude sharp waves, but often involves a bad electrode and has a slower periodicity. Of note, seizure activity sometimes is not seen on surface EEG recording (temporal or frontal lobe epilepsy). In such cases, video recording and technologist observation are essential in establishing the presence of subtle abnormal automatisms such as lip smacking or mouth movement. These can be important clues that a seizure is occurring. If no seizure activity can be documented despite extensive monitoring with additional leads, MR imaging can show a focal abnormality in the temporal or frontal lobe areas. Even if seizure activity is seen during traditional polysomnography, identification of a focal onset or localization may be difficult using the typical polysomnography EEG montage. Of note,frontal activity is often seen in the eye leads (ROC. LaC), and temporal activity is often noted in the mastoid electrodes (AI.A2). Some focal seizures generalize extremely rapidly and may appear to be generalized seizures at first glance. While frontal lobe seizures classically present as tonic posturing and temporal lobe seizures as partial complex seizures with automatisms or behavioral arrest, manifestations alone may not allow localization. For example, seizures originating from the orbitofrontal areas and the cingulate gyrus of the frontal lobes often resemble those originating from the temporal lobes-with staring, nonresponsiveness, and automatisms. In the present patient, nocturnal seizures were unrecognized by the patient's wife or the sleep lab technologist because the seizures had no motor manifestations. The patient had many 15- to 20-second bursts of seizure activity during sleep. At the usual sleep study page length (30-second page), the

abrupt onset ofrhythmic activity is noted. However, changing the page duration from 30 seconds to 10 seconds allowed identification of a clear-cut spike and wave activity with a frequency of about 4 Hz. Of note, the activity was generalized (involved left and right sides of the brain). A subsequent full EEG montage study during the day also showed episodes of similar activity. The EEG technologist was able to document that the patient had a brief period of unresponsiveness during the episodes. The seizure activity appeared to have a generalized onset, and a diagnosis of generalized seizure disorder of the absence type (no motor manifestations) was made. Initiation of primary generalized seizures (without focal onset) in adulthood is distinctly rare. On the other hand, secondary generalization of initially focal seizures is very common in adults. While classic petit-mal epilepsy is characterized by spike and wave of 3 Hz, the frequency can be faster in adults. It is always possible that this disorder was present from childhood, but went unrecognized. Although such a scenario is unlikely, neither the patient nor his wife were aware ofthe episodes. An alternate view is that the epilepsy was in fact a focal seizure with very rapid generalization. This possibility is suggested by the slightly earlier appearance of rhythmic activity in LaC (near the left frontal brain area). However, an MRI was within normal limits. The patient was treated with topiramate, an anti-epileptic medication that would be active against both focal and generalized seizures, and nasal CPAP. Clinical Pearls I. Rhythmic activity during sleep should be examined at a faster rate or virtual paper speed to look for spike and wave activity. The traditional EEG paper speed is equivalent to a lO-second page on digital monitoring. 2. Spike and wave activity can be isolated (interictal) or occur in bursts (ictal or seizure activity). Recognition of the characteristic shape can differentiate a burst of seizure activity from artifact or EEG changes that mimic seizure activity. 3. About 10-40% percent of seizures occur primarily or only during sleep. 4. If seizures occur only or mostly during sleep and do not cause abnormal motor activity, they are frequently unrecognized. 5. Use of a complete seizure montage and video recording can assist in the diagnosis of seizures. 6. Nocturnal seizures can present with complaints of excessive daytime sleepiness or insomnia secondary to arousals during seizure activity. REFERENCES I. Aldrich MS. Jahnke B: Diagnostic value of video-EEG polysomnography. Neurology 1991; 41: 1060-1066. 2. Malow B. Fromes A. Gail A. et al: Usefulness of polysomnography in epilepsy patients. Neurology 1997; 48: 1389-1394. 3. Chesson AL. DellaBadia J: Seizure disorders and sleep. In Lee-Chiong TL, Sateia MJ. Carsakadon MA (eds): Sleep Medicine. Philadelphia, Hanley & Belfus, 2002. pp 521-531. 325

PATIENT 100 A 30e:year-old man who got out of bed during sleep A 30-year-old man was evaluated for abnormal behavior during sleep, which had been ongoing for about 2 years. Sometimes he would awaken and was noted by his wife to be confused. Other times he would awaken in another room in his house and not remember how he got there. The awakenings could occur during any part of the night. The patient never exhibited any violent behavior during sleep or had any recall for the episodes. He had not recently begun any new medications and had no history of head trauma or sleepwalking as a child. His wife did not notice any mouth movements or talking during sleep. Physical Examination: Normal neurological examination Sleep Study: Included video monitoring. Initially, the tracings shown in Figure I were noted. Later in the night the patient awoke. A "strange artifact" was noted just before the awakening (Fig. 2; tracing at polysomnography/30-second page paper speed [left] and EEG/IO-second page paper speed [right]). The patient was observed by the technician to move his mouth in an odd manner. Question: What is causing the unusual behavior? LOC-A2 ROC-A1 C3-A2 C4-Al ( 15 sec of 30 sec page FiGURE 1 LOC-A2~ ROC-Al~,J~~ C3-A2~ C4-Al~~ Ol-A2~ ( ) 15 seconds of 30 sec page ( ) 5 seconds of 10 second page (EEG speed) 326 FiGURE 2

Diagnosis: Temporal lobe epilepsy. Discussion: Temporal lobe epilepsy can present with seizures that occur only at night. The seizures are typically complex partial seizures and may be manifested by lip smacking, episodes of an altered state of consciousness, confused awakenings, and automatic behavior such as wandering through the house. Even with a full seizure montage, the temporal focus may not always be seen with surface electrodes. The appearance of interictal spike activity from the temporal lobe in the typical sleep montage can be misleading. Remember that the mastoid electrodes are quite close to the temporal lobes. Therefore bipolar derivations containing the mastoid on the same side as the temporal lobe may show prominent activity even if the other part of the derivation is on the other side of the brain. For example C4-AI might show prominent activity from a left temporal focus despite the fact that C4 is central in location and on the other side of the brain. In general, approximately 60-70% of patients with partial seizure disorders and 80-85% of patients with primary generalized seizure disorders have seizure remission with anti-epileptic medication (AED). Typically, a single AED is advocated to optimize treatment and minimize adverse effects, but some patients may require two or more AEDs. Avoid sedative hypnotic agents such as phenobarbital, primidone, and clonazepam in patients with suspected concurrent obstructive sleep apnea. In the present patient, the first tracing shows prominent spikes (arrowheads) in derivations containing A2 (near the right temporal lobe). The deflections are downward as A2 is the second electrode in the derivation. Recall that by convention in the derivation (X-V), if X is negative with respect to Y, the deflection is upward. If Y is negative with respect to X, the deflection is downward. Spikes are surface negative, and here A2 is negative compared to the other electrodes. At the typical polysomnograph page size, the spikes appear as a straight line, but at a to-second page size the typical spike morphology is clear (Fig. I). A spike and wave morphology is shown most clearly in LOC-A2. What looks like artifact at a normal 30-second page size (IS-second portion) in Figure 2 is seen on a portion of a lO-second page size to be ictal (seizure) activity consisting of repetitive sharp waves. Note that the rhymic activity is more prominent in derivations containing A2. The patient underwent a full EEG monitoring montage on another night. Figure 3 shows a portion of a "double banana" bipolar montage. You can see that the rhythmic activity is present on the right side. There is phase reversal at the F8-T8 area, localizing the seizure to the right temporal area. An MRI showed a decrease in size of the hippocampal area of the right temporal lobe, which was thought to be secondary to chronic seizure activity. The patient was treated with carbamazepine, an anti-epileptic medication, with a good response. The episodic nocturnal wanderings ended. 1 sec FP1-F7 F7-T7 T7-P7 P7-01 FP2-F8 F8-T8 T8-P8 P8-02 FIGURE 3 327

Clinical Pearls I. Temporal lobe epilepsy manifested as complex partial seizures may present as a parasomnia. 2. If interictal activity such as spikes from the temporal lobe are seen in the usual sleep montage of central and occipital derivations with a mastoid reference, the spikes will often be most prominent in derivations containing the ipsilateral mastoid electrode (A I, A2). 3. Changing the page size from 30 seconds to 10 seconds may help distinguish ictal rhythmic activity from artifact. REFERENCES I. Fisch B1: Spehlmari's EEG Primer. New York, Elsevier, 1991. 2. Malow BA: Sleep and epilepsy. Neurol Clin 1996; 14:765-789. 3. Shouse MN, Mahowald MW: Epilepsy and sleep disorders. In Kryger MH, Roth T, Dement WH (eds): Principles and Practice of Sleep Medicine, 3rd ed. Philadelphia, WB Saunders, 2000, pp 707-723. 328

FUNDAMENTALS OF SLEEP MEDICINE 21 Evaluation of Insomnia Insomnia is a broad term denoting unsatisfactory sleep. Itincludes difficulty initiating sleep (sleeponset insomnia), difficulty maintaining sleep (sleep-maintenance insomnia), early morning awakening (short sleep period), and nonrestorative sleep. Most causes of insomnia are associated with problems both initiating and maintaining sleep. Some, like the delayed sleep-phase syndrome, are associated mainly with sleep-onset insomnia. Patients with depression tend to report early morning awakening. The causes of insomnia are many and diverse, complicating the evaluation of patients with this complaint (see table). Common Causes ofInsomnia PRIMARY INSOMNIA Psychophysiological Acute (adjustment sleep disorder) Chronic Idiopathic Sleep state misperception SECONDARY INSOMNIA* Sleep disorders (sleep apnea, PLMD, RLS) Psychiatric disorder (depression, panic attacks) Inadequate sleep hygiene Environmental sleep disorder Drugs (nicotine, ethanol, caffeine) Medical conditions/medications Fibromyalgia and chronic pain syndromes COPD and other respiratory disorders Medications (beta blockers, theophylline) Circadian disorders Delayed sleep-phase syndrome Advanced sleep-phase syndrome Shift work or jet lag syndrome *"Secondary" means another disorder can be diagnosed. PLM = periodic leg movement. COPD = chronic obstructive pulmonary disease An exhaustive history and collection of a sleep diary are the key elements in making a diagnosis. Obtaining a good history from a patient with insomnia can be difficult because of the many factors to be addressed (see table next page). 329

Insomnia History Nature and duration of problem Sleep habits Time in bed, lights out, sleep onset, waketime Bedroom environment Timing and duration of naps Changes on weekends Effects of a new sleep environment (vacations) Medication/beverage history Symptoms of depression History of leg jerks, restless leg syndrome, snoring, apnea Another challenging problem is identifying a mood disorder (depression). Many patients focus on their sleep disturbance and ignore manifestations of depression such as failure to get pleasure out of life and feeling sad. Others blame fatigue on their sleep disturbance despite the fact that depression often is manifested by this symptom. Some patients have difficulty sleeping because of noise, excessive light, or an uncomfortable temperature (environmental sleep disorder). Others suffer from poor sleep hygiene such as irregular bedtime and wake time, long naps, and/or working in bed. Rather than relying on the patient's memory, a sleep log (diary) is an essential tool in the evaluation of insomnia. Most sleep centers have patients complete such a diary for 2 weeks prior to the initial evaluation. There are many types of sleep diaries. The example below shows sleep behavior Monday night through Tuesday morning. The patient got into bed at 10 PM (1 ), but did not try to fall asleep until 11 PM (X). The first sleep did not occur until around 1 AM and lasted until 3 AM. A prolonged awakening between 3 and 5 AM was noted. Another episode of sleep occurred from 5 to 7 AM. The patient got out of bed at 8 AM(i). PM Midnight AM Noon PM 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 Sun Mon Tues Wed Thu Fri Sat I' to X Lightsout- trying to sleep ( Mon Tues Wed Thurs Fri Sat Sun In bed t Out of bed Polysomnography has a minor role in evaluation of most types of insomnia. Conditions with specific polysomnographic findings are sleep apnea, periodic limb movements in sleep (PLMS), and alphadelta sleep. Although excessive daytime sleepiness usually is the major complaint of patients with obstructive sleep apnea, a few patients complain mainly of insomnia. Complaints of insomnia are more prominent in central sleep apnea. PLMD may result in complaints of both insomnia and excessive sleepiness, but a complaint of insomnia is more common. Alpha-delta sleep is a polysomnographic finding of alpha intrusion into slow wave sleep. It has many causes, such as fibromyalgia (see Patient 107). In most cases of insomnia, polysomnography is not indicated unless sleep apnea or PLMS is suspected. However, if insomnia is severe or does not respond to empiric treatment, then a sleep study may be warranted. This is especially true if a history ofsnoring is elicited. For cases of insomnia in which polysomnography is obtained, the study tends to simply confirm patient complaints of a prolonged sleep latency (> 30 minutes), frequent prolonged awakenings, frequent arousals, reduction in total sleep time, reduced sleep efficiency, and decreased amount of slow wave and REM sleep. A short REM latency or early morning awakening suggests the diagnosis of depression. In 330

sleep misperception, a fairly normal night of sleep is recorded, but the patient believes that good sleep was not obtained. In psychophysiologic insomnia, sometimes both the sleep study and the patient confirm a good night of sleep, suggesting that the home sleep environment is either suboptimal or has become a stimulus for anxiety regarding sleep. REFERENCES I. Buysse OJ, Reynolds CF: Insomnia. In Thorpy MJ (ed): Handbook of Sleep Disorders. New York, Marcel Dekker, 1990, pp 375-433. 2. Kupfer OJ, Reynolds CF: Management of insomnia. N Engl J Med 1997; 336:341-346. 331

PATIENT 101 A 30-year-old woman having difficulty falling asleep A 30-year-old woman was referred for complaints of an inability to sleep (insomnia). This problem had been severe for more than 5 years. The patient usually retired at 10 PM each night, but did not fall asleep until I AM. Three to four awakenings occurred each night, with the final awakening at 6:30 AM (spontaneous). After each, the patient required at least 30 minutes to fall asleep. Self-medication with over-the-counter sleeping pills and alcohol sometimes was effective. The only good night of sleep occurred when the patient went on vacations. The sleep environment was reported to be quiet and dark. The patient did keep a lighted clock at bedside. During the day, fatigue but not definite sleepiness was noted. No naps were taken. There were no symptoms of depression and no history of marital conflicts. The patient's husband reported that his wife did not snore, kick, or jerk during sleep. Physical Examination: General: thin and nervous. Otherwise unremarkable. Sleep Study Time in bed (monitoring time) Total sleep time Sleep period time (SPT) Wake after sleep onset Sleep efficiency (0/0) Sleep latency REM latency 460 min (425-462) 411 min (394-457) 432.5 min (414-453) 21.5 min 89 (90-100) 20 min (0-19) 85 min (69-88) Sleep Stages Stage Wake Stage 1 Stage 2 Stages 3 and 4 Stage REM AHI PLM index O/OSPT 5 (0-6) 11.8 (3-6) 45 (46-62) 18 (7-21) 20.2 (21-31) O/hr O/hr ( ) = normal values for age, AHI = apnea + hypopnea index, PLM = periodic limb movement Question: Why is the sleep study relatively normal? 332

Diagnosis: Psychophysiologic insomnia. Discussion: Psychophysiologic insomnia is defined as a disorder of sornatized tension and learned sleep-preventing associations. In most sleep disorder centers, up to 15% of insomniacs receive a diagnosis of psychophysiologic insomnia. These individuals tend to react to stress with increased tension. and there is a marked overconcern and frustration with an inability to sleep. The bedroom and lightsout time become stimuli for increased tension and anxiety. The insomnia usually is fairly fixed, although it may vary in severity. A precipitating event may have caused the problem's onset. but it now has taken on a life of its own. Patients with this disorder frequently have a history of being "light sleepers" for many years. Inadequate sleep hygiene also may be present, but even after correction the problem persists. This diagnosis is not made if the patient can be classified as having an anxiety disorder, obsessive-compulsive neurosis, or major depression. The diagnosis of adjustment sleep disorder (transient psychophysiologic insomnia) is made if the insomnia is transient (usually less than 6 months). clearly follows an acute stress or conflict, and is a change from the patient's norm. Environmental sleep disorder is the diagnosis when insomnia is clearly secondary to problems with the sleep environment, such as noise, bed-partner disturbance, or the necessity of remaining vigilant (e.g., sick children). Polysomnography is of limited utility in evaluating most cases of insomnia; therefore, it is not routinely recommended and often is not reimbursed by health insurance plans. The results usually corroborate the patient's complaints (long sleep latency. low sleep efficiency, frequent arousals. prolonged awakenings) and seldom reveal a specific reason for the sleep disturbance. However, identification of periodic limb movement in sleep (PLMS), a shortened REM latency (possible depression), or, rarely, central or obstructive sleep apnea can provide clues to the cause of the insomnia. Sometimes polysomnography results in an evaluation of insomnia are amazingly normal. In such a case, if the patient believes it was a poor night of sleep, then the diagnosis is sleep state misperceplion. In this disorder, patients do not seem to recognize that they were asleep. Conversely, if the patient recognizes that sleep was fairly normal and, in fact, expected a good night of sleep, then either the home sleep environment is suboptimal or it has become a conditioned stimulus for sleep difficulty. This phenomenon is called the reverse first-night effect, as normal subjects tend to sleep poorly in a novel environment (sleep lab). The present patient complained of both sleeponset and sleep-maintenance insomnia. There was no historical information to suggest sleep apnea, PLMS, or depression. A sleep study was performed at the patient's insistence. The study showed a near normal night of sleep in the sleep laboratory and an absence of evidence for other etiologies, making psychophysiologic insomnia the most likely disorder. The patient was treated with improved sleep hygiene and stimulus control therapy (see Patient 102). Treatment resulted in better sleep latency and sleep continuity. Clinical Pearls I. Diagnosis of the cause of insomnia usually is made on the basis of a careful history and review of a patient sleep diary (log). 2. Polysomnography generally is not indicated in evaluation of insomnia. Two exceptions are when there is a suspicion of PLMS or sleep apnea and when the insomnia is severe and does not respond to empiric therapy. 3. A better-than-normal night of sleep in the sleep laboratory (a reverse first-night effect) suggests that the home sleep environment is suboptimal or has become a conditioned stimulus for sleep difficulty. 4. In psychophysiologic insomnia, sleeping in a novel location may temporarily improve insomnia. 333

REFERENCES I. Reynolds CF. Taska LS, Sewitch DE, et al: Persistent psychophysiological insomnia: Preliminary diagnostic criteria and EEG sleep data. Am J Psych 1984; 141:804-805. 2. American Sleep Disorders Association: The International Classification of Sleep Disorders: Diagnostic and Coding Manual. Lawrence, Kansas, Allen Press, 1990, pp 28-32. 3. Reite M, Buysse D, Reynolds C, Mendelson W: The use of polysomnography in the evaluation of insomnia. Sleep 1995; 18: 58-70. 4. Thorpy M, et al: Standards of Practice Committee of the American Sleep Disorders Association: Practice parameters of the use of polysomnography in evaluation of insomnia. Sleep 1995; 18:55-57. 5. Chesson A Jr, et al: AASM Standards of Practice Committee. Practice parameters for the evaluation of chronic insomnia. An American Academy of Sleep Medicine report. Standards of Practice Committee of the American Academy of Sleep Medicine. Sleep 2000; 23:237-241. 334

PATIENT 102 A 30-year-old woman with insomnia A 30-year-old woman with complaints of insomnia was diagnosed with psychophysiologic insomnia after an evaluation that included a polysomnogram. She admitted that she was a tense person and had problems relaxing. When she had problems falling asleep, she became very anxious: "I look at the clock and am upset that the night is almost over and I haven't fallen asleep." The patient denied drinking coffee, but admitted that she drank wine at bedtime to help her fall asleep. She reported being less tense about falling asleep on the weekends because she could sleep later the next day. Interestingly, the patient reported sleeping better on vacations than in her own bedroom. There was no history of snoring or leg movements during sleep. Sleep Diary PM Midnight AM 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 Noon PM 9 10 11 12 1 2 3 4 5 6 Sun Man Tues Wed Thu Fri Sat ! !,~ .'lE~~ I.... . I' , ! ,1 " fE-~1 )i \/1 I.... , ~)i I.... ... ... If I' W =3 glasses of wine, X =lights out and trying to sleep, ~--+) =asleep, ! =in bed, t =out of bed Man Tues Wed Thurs Fri Sat Sun Question: What treatment options, other than medication, would you recommend? 335

Answer: Good sleep hygiene, stimulus control therapy, relaxation therapy. Discussion: The treatment of insomnia must be individualized. The mainstay of any treatment is to optimize sleep hygiene by educating patients about habits that interfere with good sleep. Good sleep hygiene includes maintaining a favorable sleep environment (e.g., quiet, dark, comfortable), keeping a regular sleep routine (constant bedtime and waketime), avoiding stimulants such as caffeine and other medications that interrupt sleep (e.g., ethanol), and avoiding long naps. Note that caffeine can impair sleep up to /0 hours later, and some patients are quite sensitive to just a tiny amount. Therefore, patients with insomnia should be questioned carefully about their caffeine intake, including colas and tea. Ethanol frequently is used to help promote sleep onset; however, ethanol intake near bedtime can cause awakenings and fragmented sleep later in the night, even at low doses. Behavioral techniques, although widely recommended as treatment for insomnia, are applied less commonly than the pharmacologic approach. One reason is that they are time-intensive for both clinician and patient. Readily available educational materials and instruction by knowledgeable ancillary personnel may reduce clinician involvement and make these techniques more cost-effective. Relaxation therapy is a commonly used behavioral treatment. Many patients with insomnia rep0l1 physiologic (tension) and cognitive/emotional (racing thoughts and worrying) arousal at bedtime. Progressive muscle relaxation (originated by Jacobson) consists of first tensing then relaxing each muscle group in a systematic way. Patients receive instruction in this technique and then practice twice daily, with the last session at bedtime. Biofeedback treatment uses feedback from EMG monitoring of a muscle, such as the frontalis muscle, to teach the patient how to relax. Patients with cognitive arousal at bedtime may benefit from meditation or guidedimagery techniques (refocusing on a pleasant mental target). For some patients, regular exercise can improve sleep. Exercise should not be within 2 hours of bedtime as it raises the body temperature, making sleep onset more difficult. A second behavioral option is stimulus control therapy. This treatment recognizes that insomniacs typically associate the bedtime (temporal cue) or the bedroom (environmental cue) with difficulty falling asleep: they become anxious as they stay in bed and "watch the clock," and over several nights the bedroom itself becomes a stimulus for anxiety and insomnia. This association explains why some patients with insomnia sleep better in a new setting. 336 Stimulus control therapy, developed by Bootzin and Nicassio, seeks to create a conditioned association between the bedroom and sleep. Activities in the bedroom are restricted to sleep and sex. Patient do not get in bed unless sleepy and do not remain in bed unless drowsy or asleep. If they fail to fall asleep in a reasonable time, they are instructed to get out of bed until they feel sleepy. Regular bedtime and waketime as well as avoidance of naps are part of the instructions. A third option is called sleep restriction therapy. The clinician looks at the sleep diary and estimates the time spent in bed and the time asleep. The patient is then asked to restrict the time in bed to match the previous time spent asleep. This induces mild sleep deprivation and increases sleep efficiency. As the efficiency is improved the time allowed in bed is slowly increased. Paradoxical intention is a fourth treatment technique. Patients are encouraged to engage in the most feared activity ("staying awake"). The central idea is that performance anxiety over being able to fall asleep may prevent sleep onset. By concentrating on staying awake, natural sleep processes may be allowed to work. Cognitive-behavioral therapy consists of 8- 10 weekly sessions that provide education about sleep hygiene, stimulus control, sleep restriction, and relaxation techniques. Cognitive therapy focuses on changing unrealistic beliefs and fears regarding the loss of sleep. One common fear is that a large amount of sleep is necessary, or illness will result. The last approach is combined behavioral and pharmacological treatments. Pharmacological treatments of insomnia are discussed in Patient 103. Some patients can temporarily be treated with hypnotics while behavioral techniques are mastered. Others can be treated on a chronic basis with behavioral techniques and pro hypnotics. For some patients, knowing an effective hypnotic rescue is available decreases their anxiety about falling asleep. The standard of practice committe of the American Academy of Sleep Medicine reviewed the evidence for non-pharmacological treatments of insomnia. They gave a standard recommendation for stimulus control therapy. Progressive muscle relaxation therapy, biofeedback, and paradoxical intention received a "guideline" recommendation, which means a moderate degree of clinical certainty exists. Sleep restriction was deemed optional (conflicting or inconclusive evidence). The levels of recommendation are based on exisiting clinical studies.

Behavioral Treatment ofInsomnia Relaxation techniques (progressive muscle relaxation, biofeedback, guided imagery) Stimulus control Sleep restriction Paradoxical intention Cognitive-behavioral treatment Combined behavioral and pharmacological treatment The present patient, when questioned in detail, reported consumption of at least ten caffeine-containing carbonated beverages a day. Her sleep diary shows a lights-out time of 11 PM during the week and a long sleep latency. The patient typically was in bed an hour before lights out, and she sometimes read work-related materials during this time. On the weekends, bedtime was delayed and ethanol was consumed, resulting in a shorter sleep latency. One or two awakenings most nights were recorded. The history of sleeping better on vacation suggested that the patient had an association between her bedroom and problems falling asleep. Treatment included instructions to switch to noncaffeinated drinks and avoid ethanol near bedtime. The patient removed the clock from her bedroom and went to bed just before lights out. If unable to fall asleep within a reasonable time, she got out of bed and read (recreational reading) until she felt sleepy, and then she returned to bed. If she awakened during the night and was unable to return to sleep, she again got out of bed until sleepy (stimulus control). In addition, the patient was given tapes instructing her in relaxation techniques. She practiced relaxation rather than engaging in work-related activities near bedtime. Initially, despite this combined approach, the patient still had some difficulty falling asleep. However, within a few weeks she reported falling asleep within 20 minutes on most nights and having fewer awakenings. Clinical Pearls I. Every patient with insomnia must be questioned in detail about sleep habits and intake of beverages that can disturb sleep. 2. Treatment of insomnia always should begin with establishment of good sleep hygiene. 3. Relaxation therapy may be especially helpful in patients with emotional or physical tension at bedtime. 4. Stimulus control treatment can break the association between the bedroom and poor sleep. 5. Sleep restriction therapy increases sleep efficiency by creating mild sleep deprivation. 6. Cognitive-behavioral treatment employs elements of other approaches (Pearls 3-6) and reverses unrealistic fears and beliefs about sleep. REFERENCES 1. Buysse DJ, Reynolds CF: Insomnia. In Thorpy MJ (ed): Handbook of Sleep Disorders. New York. Marcel Dekker. 1990. pp 375-433. 2. Morin CM, Culbert JP, Schwartz SM: Nonpharrnacologic interventions for insomnia: A meta-analysis of treatment efficacy. Am J Psych 1994; 151:1172-1180. 3. Chesson AL Jr, et al: AASM Standards of Practice Committee. Practice parameters for the nonpharmacologic treatment of chronic insomnia. An American Academy of Sleep Medicine report. Sleep 1999; 22: 1128-33 4. Morin Clvl, Hauri PJ, Espie CA, et al: Nonpharmacologic treatment of chronic insomnia. An American Academy of Sleep Medicine review. Sleep 1999; 22: I 134-1156. 5. Stepanski EJ: Behavior therapy for insomnia. In Kryger M, Roth T, Dement W (eds): Principles and Practice of Sleep Medicine, 3rd ed. Philadelphia, WB Saunders 2000, pp 647-656. 337

PATIENT 103 A 40-year-old man with difficulty falling asleep after the death of his brother A 40-year-old man who previously had no sleep problems developed difficulty falling asleep and staying asleep after the death of his brother 3 months previously. He had tried improvement in sleep hygiene and did not remain in bed unless sleepy. However, he was still unable to sleep for at least I hour after retiring at his normal bedtime. He had seen a psychiatrist who had started him on amitriptyline, but this made the patient very groggy the next day. The patient's primary care physician had given him triazolam, which enabled him to fall asleep quickly, but he sometimes felt very anxious in the mornings. The patient's wife noted that his legs jerked on occasion when he was asleep. Sleep Study Time in bed (TIB) Total sleep time (TST) Sleep period time (SPT) Wake after sleep onset Sleep efficiency (%) Sleep latency REM latency 440 min (390-468) 314 min (343-436) 370 min (378-452) 55.5 min 71 (90-100) 50 min (2-18) 85 min (55-78) Sleep Stages Stage Wake Stage I Stage 2 Stages 3 and 4 Stage REM AHI PLM index % SPT 15 (1-12) 20 (5-11) 45 (44-66) 5 (2-15) 15 (19-27) 3/hr 5/hr ( ) = normal values for age. AHI = apnea + hypopnea index. PLM = periodic limb movement Questions: What is your diagnosis? Which hypnotic would you suggest for this patient? 338