The term insomnia signifies a chronic inability to sleep despite adequate opportunity to do so; it is used popularly to indicate any impairment in the duration, depth, or restorative properties of sleep. There may be difficulty in falling asleep or remaining asleep, awakening may come too early, or there may be a combination of these complaints. Precision as to what constitutes pathologic insomnia is impossible at the present time because of our uncertainty as to the exact amounts of sleep required, and the role of sleep in the economy of the human body. All that can be said is that some form of sleeplessness is a frequent complaint (20 to 40 percent of the population) and is more prominent in the elderly and in women. Only a small proportion of persons who perceive their sleep to be inadequate seek professional help or use sleeping pills, according to Mellinger and colleagues.
Two general classes of insomnia can be recognized—one in which there appears to be a primary abnormality of the normal sleep mechanism, and another in which the sleep disturbance is secondary to, or perhaps more accurately comorbid with, a medical or psychologic disorder. Polysomnographic studies have defined yet another subgroup who actually sleep enough, but who perceive their sleep time to be shortened or disrupted ("paradoxical insomnia").
This term is reserved for the condition in which nocturnal sleep is disturbed for prolonged periods and none of the symptoms of anxiety, depression, pain, or other psychiatric or medical diseases can be invoked to explain the sleep disturbance. In some patients, like those described by Hauri and Olmstead, the disorder is lifelong. Unlike the rare individuals who seem to be satisfied with 4 h or even less of sleep a night, insomniacs suffer the effects of partial sleep deprivation and resort to medications, alcohol, and their lives come to revolve around sleep to such an extent that they have been called "sleep pedants" or "sleep hypochondriacs." Although statements on the quantity and quality of sleep given by insomniacs are often not to be taken as entirely valid, Rechtschaffen and Monroe have confirmed that most of them do, indeed, sleep for shorter periods, move and awaken more often, spend less time in N3 sleep than normal persons, and show a heightened physiologic arousal. Personality inventories have disclosed a high incidence of psychologic disturbances in this group, but whether these are cause or effect is not clear.
Furthermore, a category of "conditioned', or "psychophysiological" insomnia has been denominated, in which a situational trigger for insomnia has ceased, but the sleep disorder persists. Although insomniacs, regardless of the cause, tend to exaggerate the amount of sleep lost, primary insomnia should be recognized as an entity and not passed off as a neurotic quirk.
This common type of insomnia, which is often transitory, can be ascribed to pain or some other recognizable bodily disorder, such as drug or alcohol abuse or, most commonly, to anxiety, worry, or depression. Of the medical disorders conducive to abnormal wakefulness, certain ones stand out—pain in the joints or in the spine, abdominal discomfort from peptic ulcer and carcinoma, pulmonary and cardiovascular insufficiency, and the nocturia engendered by prostatism. The "restless legs" syndrome and periodic leg movements of sleep are not considered in this category and have their own physiology, phenomenology, and treatment.
Restless Legs Syndrome, Periodic Leg Movements of Sleep, and Related Disorders
The disorder known as the restless legs syndrome may regularly delay the onset of sleep and usually occurs in its early stages. Ekbom called it asthenia crurum paresthetica and also, anxietas tibiarum. This disorder is surprisingly prevalent, affecting more than 2 percent of the population. The patient may complain of unpleasant aching and drawing sensations in the calves and thighs, often associated with creeping or crawling feelings; other descriptions have included "worms," "internal itch," and "coldness," and the legs may feel tired, heavy, and weak. The symptoms are provoked by rest, and rapidly, but temporarily, relieved by moving the legs. An urge to move the legs can be suppressed voluntarily for a brief period but is ultimately irresistible. Moving the limbs alleviates the sensation briefly. It is interesting that a small proportion of patients have similar symptoms in the arms after many years of symptoms. There may be variants of nocturnal restlessness in other parts of the body such as the abdomen, as suggested by Pérez-Díaz and colleagues. Their patients described an unpleasant abdominal musculature restlessness that required movement for relief and was eliminated with dopamine agonists.
Fatigue worsens restless legs syndrome, and there is a tendency for it to be worse in warm weather. In a few patients, mainly older ones with a severe form of the nighttime disorder, these movements and an associated myoclonus spill over into wakefulness and are accompanied by restlessness, foot spasms, foot stamping, body rocking, and marching that are only partly under voluntary control. The daytime phenomena may require several medications used simultaneously for control.
The syndrome is idiopathic and persists for years. Iron-deficiency anemia and low ferritin levels are associated with the syndrome in many instances, as is thyroid disease, pregnancy, and certain drugs, such as antidepressants and antihistamines. Occasionally, it is a prelude to a peripheral neuropathy, particularly in relation to uremia. Also, reduced levels of iron in the cerebrospinal fluid (CSF) have been found with restless legs syndrome and with periodic limb movements of sleep (see below). The basis for this relationship is not well defined, but it makes it advisable to check for reduced iron stores and anemia in most patients. One hypothesis is that a disturbance of iron storage in the basal ganglia causes a decrement in dopamine binding by dopamine receptors and transporters, as has been described in studies using PET and single-photon emission tomography (SPECT) scans. Another potential relationship, unproved, is that iron is a cofactor for the enzyme, tyrosine hydroxylase, which is required to produce dopamine.
A closely related disturbance is periodic leg movements of sleep. Like the restless legs syndrome, it may result in sleep deprivation and daytime somnolence or, more often, in disturbance of a bed partner. However, the diagnosis of periodic leg movements depends on finding them during polysomnographic recordings, whereas restless leg syndrome is identified on clinical grounds. Originally described as "nocturnal myoclonus," periodic leg movements are slower than myoclonic jerks. They consist of a series of repetitive movements of the feet and legs occurring every 20 to 90 s for several minutes to an hour; mainly the anterior tibialis is involved, with dorsiflexion of the feet and big toes, sometimes followed by flexion of the hip and knee. The movements are similar to the triple-flexion (Babinski) response, which can be elicited in normal sleeping persons. These movements produce frequent microarousals or, if severe and periodic, full arousals. The patient, usually unaware of these sleep-related movements at the time they occur, is told of them by a bed mate or suspects their occurrence from the disarray of the bedclothes. Periodic leg movement is closely associated with the restless legs syndrome and many sleep specialists consider it an integral part of the syndrome, but it also occurs independently with narcolepsy, sleep apnea, following the use of tricyclic and serotonin reuptake inhibiting antidepressants, L-dopa, and withdrawal from anticonvulsants and sedative-hypnotic drugs. Approximately eighty percent of individuals with restless leg syndrome will display periodic leg movements, but the opposite is not the case, as only twenty to thirty percent of patients with periodic leg movements have restless leg syndrome.
A seminal genetic finding by Stefansson and colleagues derived from several populations, including the homogenous Icelandic, is that a nucleotide variant in a short segment of chromosome 6p is associated with periodic leg movements of sleep. This was found to hold in those with and without restless legs syndrome. If nothing else, as pointed out by the authors, this establishes that periodic limb movements are a distinct entity as defined in the era of genomics. The biologic significance and frequency in other populations of this variant is not yet known. Nonetheless, we continue to be impressed at the frequent cooccurrence of the two conditions and several shared underlying conditions such as iron deficiency, and treatments that are effective in both.
A search for iron deficiency, and its correction if present, is indicated in almost all cases. A large number of symptomatic medications have proved helpful in the treatment of both the restless legs syndrome and periodic leg movements. As a first choice, many practitioners favor treatment with dopamine agonists such as pramipexole (0.25 to 0.75 mg) or ropinirole (0.5 to 1.5 mg), either one taken 1.5 to 2 h before bedtime. Long-acting combinations of L-dopa/carbidopa (12.5/50 or 25/100 mg dose) taken at bedtime have also been successful, but L-dopa, and sometimes the dopamine agonists, causes some patients to develop the movements earlier, i.e., in the daytime, which become more intense and spread to other body parts. A longer acting dopamine agonist, rotigotine patch is available to treat patients who have this augmentation phenomenon. The latter is also helpful for periodic leg movements.
A major problem, recently recognized, is one of "augmentation," or enhancement of the restless leg syndrome with the long-term use of this class of drugs. This is less prominent with some of the other numerous drugs that have been effective including gabapentin, pregabalin, clonazepam (0.5 to 2.0 mg), temazepam (30 mg), taken 30 min before retiring and extended release gabapentin which are available to minimize the augmentation phenomenon. Other drugs—e.g., baclofen, opioids, carbamazepine seem to be helpful in certain patients, but they are infrequently required. The lengthy list of medications that have been effective is given in Earley's review. It is sometimes useful to give a medication in 2 divided doses, the first early in the evening, and the second just before sleep or, in severe cases, during the night by setting an alarm clock before the anticipated time of symptoms.
Other Causes of Secondary Insomnia
Among the secondary insomnias, those caused by some type of psychologic disturbance are particularly common. Domestic or business worries may keep the patient's mind in turmoil (situational insomnia). A strange bed or unfamiliar surroundings may prevent drowsiness and sleep. Under these circumstances, the main difficulty is in falling asleep, with a tendency to sleep late in the morning. These facts emphasize that conditioning and environmental factors (social and learned) are normally involved in readying the mind and body for sleep.
Illnesses in which anxiety and fear are prominent symptoms also result in difficulty in falling asleep and in light, fitful, or intermittent sleep. Disturbing dreams are frequent in these situations and may awaken the patient. Exceptionally, a patient may even try to stay awake in order to avoid them. In contrast, depressive illness produces early morning waking and inability to return to sleep; the quantity of sleep is reduced, and nocturnal motility is increased. REM sleep in depression, although not always reduced, comes earlier in the night. If anxiety is combined with depression, there is a tendency for both the above patterns to be observed. Yet another common pattern of disturbed sleep can be discerned in individuals who are under great tension and worry or are overworked and tired out. These people sink into bed and sleep through sheer exhaustion, but they awaken early with their worries and are unable to get back to sleep.
Chronic and even short-term use of alcohol, barbiturates, and certain nonbarbiturate sedative-hypnotic drugs markedly reduces REM sleep as well as stages 3 and 4 of NREM sleep (N3). Following withdrawal of these drugs, there is a rapid and marked increase of REM sleep, sometimes with vivid dreams and nightmares. "Rebound insomnia," a worsening of sleep compared with pretreatment levels, has also been reported upon discontinuation of short-half-life benzodiazepine hypnotics, notably triazolam (Gillin et al) and including the newer sleep agents mentioned below. Furthermore, a form of drug-withdrawal or rebound insomnia may actually occur during the same night in which the drug is administered. The drug produces its hypnotic effect in the first half of the night and a worsening of sleep during the latter half of the night, as the effects of the drug wear off; the patient and the physician may be misled into thinking that these latter symptoms require more of the hypnotic drug or a different one. Alcohol taken in the evening acts in the same way. Rebound insomnia must be distinguished from the early morning awakening that accompanies anxiety and depressive states.
A wide variety of other pharmacologic agents may give rise to sporadic or persistent disturbances of sleep. Caffeine-containing beverages, corticosteroids, bronchodilators, central adrenergic-blocking agents, amphetamines, certain "activating" antidepressants such as fluoxetine, and cigarettes are the most common offenders. Others are listed in the extensive review of Kupfer and Reynolds.
Acroparesthesias, a predominantly nocturnal tingling and numbness of the fingers and palms caused by tight carpal ligaments (carpal tunnel syndrome), may awaken the patient at night (see further on, under "Sleep Palsies and Acroparesthesias"). Cluster headaches characteristically awaken the patient within 1 to 2 h after falling asleep (see Chap. 10 for a fuller discussion). In a few patients, cluster headaches occur only during or immediately after the REM period.
The sleep rhythm is totally deranged in acute confusional states and especially in delirium, and the patient may doze for only short periods, both day and night, the total amount and depth of sleep in a 24-h period being reduced. Frightening hallucinations may prevent sleep. The senile patient tends to catnap during the day and to remain alert for progressively longer periods during the night, until sleep is obtained in a series of short naps throughout the 24 h; the total amount of sleep may be increased or decreased.
In general, a sedative-hypnotic drug for the management of insomnia should be prescribed only as a short-term aid during an illness or some unusual circumstance, i.e., for acute insomnia. For patients who have difficulty falling asleep, a quick-acting, fairly rapidly metabolized hypnotic is useful. The most commonly used medications are the benzodiazepine receptor agonists, which act on the gamma-aminobutyric acid (GABA)-A receptor complex. In the past, benzodiazepines were popular but these have been replaced by newer nonbenzodiazepine receptor agonists with shorter half-lives and fewer side effects (e.g., zolpidem, zaleplon, and eszopiclone). Patients who do not respond to these medications may be given an intermediate-duration benzodiazepine such as temazepam. Hypnotic use is inadvisable during pregnancy and should be used cautiously in patients with alcoholism or advanced renal, hepatic, or pulmonary disease, and should be avoided in patients with sleep apnea syndrome.
Melatonin (3 to 12 mg) has reportedly been as effective as the sedative-hypnotics and may cause fewer short-term side effects, but both of these statements are difficult to confirm. Melatonin has a short half-life, and it has only weak hypnotic effect. Therefore, for sleep rhythm disturbances, it is ideally taken three to four hours before sleep time. Amitriptyline (25 to 50 mg at bedtime) appears to be a sleep-enhancing drug even in those who are not anxious or depressed. Tolerance develops to the drug, and there are morning side effects so the drug may find its best use in patients who are taking it for alternative reasons such as headache or depression. Some practitioners indicate that it may also worsen restless leg or periodic leg movement disorders. When pain is a factor in insomnia, the sedative may be combined with a suitable analgesic. Nonprescription drugs containing diphenhydramine (Benadryl), valerian, or doxylamine, which are minimally or not at all effective in inducing sleep, may impair the quality of sleep and lead to drowsiness the following morning.
The chronic insomniac who has no other symptoms should be discouraged from using sedative drugs. The solution of this problem is rarely to be found in medication. One should search out and correct, if possible, any underlying situational or psychologic difficulty, using medication only as a temporary measure. Patients should be encouraged to regularize their daily schedules, including their bedtimes, and to be physically active during the day but to avoid strenuous physical and mental activity before bedtime. It has been suggested that illumination from broad-spectrum light (television) in the late evening is detrimental. Dietary excesses must be corrected, and all nonessential medications interdicted. Coffee and alcohol should be avoided at night, if not throughout the day. A number of simple behavioral modifications may be useful, such as using the bedroom only for sleeping, arising at the same time each morning regardless of the duration of sleep, avoiding daytime naps, and limiting the time spent in bed strictly to the duration of sleep. A helpful approach is to lessen the patient's concern about sleeplessness by pointing out that he will always get as much sleep as needed and that there is pleasure to be derived from staying awake and reading, or viewing a movie.
Disorders of Sleep Related to Neurologic Disease
Many neurologic conditions seriously derange the total amount and patterns of sleep (see Culebras). Lesions in the upper pons, near the locus ceruleus, are particularly prone to do so. Markand and Dyken have described the most substantial of these, pontine infarction with involvement of the tegmental raphe nuclei. The clinical abnormality took the form of diminished NREM sleep and near abolition of REM sleep lasting for weeks or months. Bilateral lacunar infarctions in the pontine tegmentum, demonstrable by MRI, also appear to be the basis of some instances of the so-called REM sleep behavior disorder (Culebras and Moore) described further on with the other parasomnias. Bilateral paramedian thalamic infarctions are a potent cause of hypersomnia, the result of disruption of both arousal mechanisms and NREM sleep (Bassetti et al).
Medullary lesions may affect sleep by altering automatic ventilation; the most extreme examples occur with bilateral tegmental lesions that may completely abolish breathing during sleep ("Ondine's curse," as described in Chap. 26). Lesser degrees of tegmental damage—as might occur with Chiari malformations, unilateral medullary infarction, syringobulbia, or poliomyelitis—may cause sleep apnea, and daytime drowsiness. Patients with large hemispheric strokes may also be left with daytime lethargy on the basis of inversion of sleep–wake rhythm. Certain instances of mesencephalic infarction that are characterized by vivid visual hallucinations (peduncular hallucinosis) may be associated with disruption of sleep.
von Economo encephalitis, now an extinct illness, was usually associated with a hypersomnolent state but caused persistent insomnia in some instances. The latter was related to a predominance of lesions in the anterior hypothalamus and basal frontal lobes, in distinction to hypersomnia, which was related to lesions mainly in dorsal hypothalamus and subthalamus. This subject and other forms of hypersomnia are elaborated further on under "Pathologic Excessive Sleep (Hypersomnia)."
A remarkable illness, termed fatal familial insomnia, was initially described by Lugaresi and colleagues. This disorder, with onset in middle age and a clinical course of 7 to 36 months, is characterized by a virtual incapacity to sleep and to generate EEG sleep patterns. The cerebral changes consist mainly of profound neuronal loss in the anterior or anteroventral, and mediodorsal thalamic nuclei. These cases apparently represent a usually familial form of prion disease similar to diseases that cause subacute spongiform encephalopathy and Gerstmann-Sträussler-Scheinker disease (see Chap. 33). Interestingly, the alcoholic form of the Korsakoff amnesic state, associated with less severe lesions in the same thalamic nuclei, is also characterized by a sleep disturbance, taking the form of an increased frequency of intermittent periods of wakefulness (Martin et al). When carefully sought, similar sleep–wake disturbances have been found in sporadic Creutzfeldt-Jakob-prion disease (Landolt et al).
Major head injury is an important cause of sleep disturbance. The abnormalities, which may persist for months or years, consist mainly of a decrease in stages 1 and 2 sleep, and less than the expected amounts of REM sleep and dreaming. Some patients in the persistent vegetative state show a cycle of changes in the EEG, progressing from a picture of abortive spindles and K complexes with cyclic alterations in respiration and pupil size to the acquisition of a more normally structured sleep activity. This sequence usually presages the change from a state of coma to one of minimally conscious state (Chap. 17). Organized sleep activity is absent in virtually all types of coma that are the result of anatomic damage to the brain. An exception, albeit a semantic one, occurs in the unusual condition known as "spindle coma," in which persistent coma, and the electrographic features of sleep coexist. This particular combination of events has been described after head trauma, and rarely, in association with profound metabolic encephalopathies. Despite what appears to be a genuine comatose state (not simply hypersomnolence) from a lesion of the reticular-activating system, the EEG displays frequent spindle activity and vertex waves, attesting to the integrity of thalamocortical pathways for sleep activity (see Nogueira de Melo et al).
Migraine, cluster headaches, and paroxysmal hemicrania all have been linked to certain sleep stages and are discussed in Chap. 10 in relation to other forms of "hypnic headaches." A variety of sleep disturbances may accompany brain tumors or follow surgical resection of an intracranial tumor. These include excessive daytime sleepiness, sleep apnea, and, rarely, nocturnal epilepsy. The location of the lesion, rather than the tumor type, is predictive of such a disturbance; thus, tumors affecting the hypothalamus, and pituitary are associated with excessive daytime drowsiness, whereas medullary lesions cause respiratory disturbances that may affect sleep (Rosen et al). A symptomatic form of narcolepsy is associated with tumors located adjacent to the third ventricle, and midbrain (see below). Schwartz and associates reported transient cataplexy (see further on) following surgery for a craniopharyngioma, but a delirious state has been more common in a few cases for which we have been consulted.
However, it is the disturbed sleep patterns in patients with Alzheimer disease, Huntington chorea, olivopontocerebellar degeneration, and progressive supranuclear palsy that have attracted the most attention by neurologists (Parkes). Dreaming is also absent in some of these conditions. The peculiarities of sleep in Parkinson disease have been extensively studied. Many patients in the early stages of the disease complain of fragmented and unrestful sleep, particularly in the early morning hours; some advanced cases have pathologic insomnia, and this is influenced also by medications used to treat the disease and by deep brain stimulation (see Chap. 39 for a discussion of the nonmotor effects of Parkinson disease). The loss of natural body movements and the alerting effects of L-dopa contribute to the insomnia. The directly acting dopaminergic agonist drugs used for the treatment of Parkinson disease may have the side effect of a pronounced and often rapid daytime sleepiness; however, a similar problem arises in some patients with advancing disease alone.
In striatonigral degeneration (multiple system atrophy), Lewy-body disease, and other parkinsonian syndromes, there is often a characteristic REM sleep disorder, in which the patient moves and speaks violently and aggressively during dreaming. This "REM sleep behavior disorder" does not reflect the personality of the patient or his behavior in the daytime. The majority of patients are able to recall their dream and report that it involved escaping or protecting another person from harm. In these degenerative conditions that affect the basal ganglia the sleep disturbances may precede other symptoms or be an essential part of the illness (see later in this chapter and Chap. 39).
Disorders of Sleep Associated with Changes in Circadian Rhythm
Sleep is also disturbed and diminished when the normal circadian rhythm of the sleep–wake cycle is exogenously altered. This is observed most often in shift workers, who periodically change their work schedule from day to night, and as a result of transmeridianal air travel—i.e., jet lag (Baker and Zee). Eastbound travelers fall asleep late and face an early sunrise. The consequent fatigue is a product of both sleep deprivation and a phase change required by changing time zones. A review of the subject can be found by Sack. One antidote is to reset one's watch on the plane and conform to the routine of the destination—i.e., to stay awake all day until the usual evening hour for sleep—and to take a short-acting sedative at bedtime. Melatonin is also used for this purpose, and a meta-analysis by Herxheimer and Petrie of four trials suggests it is slightly effective. These measures facilitate the resetting of the circadian rhythm. Westbound travelers face a late sunset and a long night's sleep and adjust more readily to resetting of the circadian rhythm than do those traveling east. Exposure to light during the extended day is helpful in entraining the sleep cycle; this adjustment is also accomplished more easily when traveling west than east. Shifting of the circadian rhythm in animals suggests that brief exposure to light at crucial times effectively resets the sleep–wake cycle; apparently, the period just before 4 A.M. is a nodal time for susceptibility to this phase change. The sleep problems caused by shift work are more complicated (see Monk).
The delayed-sleep-phase syndrome is a chronic inability to fall asleep and to arise at conventional clock times. Sleep onset is delayed until 3 to 6 A.M.; the subject then sleeps normally until 11 A.M. to 2 P.M. An imposed sleep period from 11 P.M. to 7 A.M. leads to a prolonged sleep latency and daytime sleepiness. By contrast, the advanced-sleep-phase syndrome is characterized by an early evening sleep onset (8 to 9 P.M.) and early morning awakening (3 to 5 A.M.). Simply delaying the onset of sleep usually fails to prevent early morning awakening. This pattern is not uncommon among healthy elderly persons (and also among college students), in whom it should probably not be defined as an insomnia syndrome. Still other persons show a completely irregular sleep–wake pattern; sleep consists of persistent but variable short or long naps throughout the night and day, with a nearly normal 24-h accumulation of sleep.
Parasomnic Disturbances and Isolated Sleep Symptoms
Although new classifications have rearranged the nosology of these phenomena, included under this title are several diverse disorders: somnolescent starts, sensory paroxysms, nocturnal paroxysmal dystonia, sleep paralysis, night terrors and nightmares, somnambulism, and REM sleep behavior disorder.
Somnolescent (Sleep, Hypnic, Myoclonic) Starts
As sleep comes on, certain motor centers may be excited to a burst of insubordinate activity. The result is a sudden "start" or myoclonic bodily jerk of large amplitude, which rouses the incipient sleeper. It may involve one or both legs or the trunk, less often, the arms. If the start occurs repeatedly during the process of falling asleep and is a nightly event, it may become a matter of great concern to the patient. The starts are more apt to occur in individuals in whom the sleep process develops slowly; they are especially frequent under conditions of tension and anxiety. Polysomnographic recordings have shown that these bodily jerks occur at the moment of falling asleep or during the early stages of sleep. Sometimes they appear as part of an arousal response to a faint external stimulus and are then associated with a frontal K complex in the EEG. These bodily jerks are not variants of epilepsy.
A small proportion of otherwise healthy infants exhibit rhythmic jerking of the hands, arms, and legs or abdomen, both at the onset and in the later stages of sleep (benign neonatal myoclonus). The movements begin in the first days of life and disappear within months. There may be a familial tendency toward these movements. Coulter and Allen differentiate this state from myoclonic epilepsy and neonatal seizures by the absence of EEG changes, and its occurrence only during sleep.
Sensory centers may be disturbed in a similar way to the earlier-described sleep starts, either as an isolated phenomenon or in association with motor phenomena. The patient, dropping off to sleep, may be roused by a sensation that darts through the body, a sudden flash of light, or a sudden crashing sound or thunderclap of head pain—cephalgia fugax, or "the exploding head syndrome" (Pearce). Sometimes there is a sensation of being turned or lifted, and dashed to the ground; conceivably, these are sensory paroxysms involving the labyrinthine-vestibular mechanism. Though obvious causes for concern by patients, these sensory paroxysms are benign.
Nocturnal Frontal Lobe Epilepsy
Numerous forms of epilepsy become more prominent during sleep as noted in a later section and in Chap. 16 on epilepsy. However, one isolated seizure disorder is characterized by paroxysmal bursts of generalized choreoathetotic, ballistic, and dystonic movements occurring during NREM sleep (Lugaresi et al, 1986). Sometimes the patient appears awake and has a fearful or astonished expression, or there are repetitive utterances and an appearance of distress, similar to what is seen in night terrors, the main differential diagnosis discussed further on. The attacks may begin at any age, affect both sexes, and are usually nonfamilial. Two forms of this disorder have been recognized: in one, the attacks last 60 s or less; they may be diurnal as well as nocturnal; some patients in addition have epileptic seizures of the more usual type; and all respond to treatment with carbamazepine. The studies of Tinuper and coworkers, using prolonged video-EEG monitoring, indicate that these brief attacks of nocturnal paroxysmal dystonia may actually be epileptic seizures of frontal lobe origin. In a second and more rare type, the attacks are longer lasting (2 to 40 min). Ictal and interictal EEGs during wakefulness and sleep are normal, and these attacks do not respond to anticonvulsants of any type. Except for the lack of familial incidence and occurrence only during sleep, the disorder is very much the same as the "familial paroxysmal dystonic choreoathetosis" described by Lance (see "Paroxysmal Choreoathetosis and Dystonia" in Chap. 4).
Curious paralytic phenomena, referred to as pre- and post-dormital paralyses, may occur in the transition from the sleeping to the waking state. Sometimes in the morning and less frequently when falling asleep, otherwise healthy persons—though awake, conscious, and fully oriented—are seemingly unable to activate their muscles. Respiratory and diaphragmatic function and eye movements are usually unaffected, although a few patients have reported a sensation of being unable to breathe. They lie as though still asleep, with eyes closed, and may become quite frightened while engaged in a struggle for movement. They have the impression that if they could move one muscle, the paralysis would be dispelled instantly, and they would regain full power. It has been stated that the slightest stimulus, such as the touch of a hand or calling the patient's name, will abolish the paralysis. Sleep deprivation is a common precipitant to the syndrome.
Such attacks are also observed in patients with narcolepsy (discussed later in this chapter) and with the hypersomnia of the pickwickian syndrome and other forms of sleep apnea. Some cases are familial.
The weakness or paralysis is thought to be a dissociated form of the atonia of REM sleep. Usually, the attacks are brief (minutes or less); if they occur in isolation and only on rare occasions, they are of no special significance. If frequent, as in narcolepsy, they can be prevented by the use of tricyclic antidepressants, particularly clomipramine, which has serotonergic activity.
Night Terrors and Nightmares
The night terror (pavor nocturnus) is mainly a problem of childhood. It usually occurs soon after falling asleep, during stage 3 or 4 sleep. The child awakens abruptly in a state of intense fright, screaming or moaning, with marked tachycardia (150 to 170 beats/min) and deep, rapid respirations. Children with night terrors are often sleepwalkers as well, and both kinds of attack may occur simultaneously. The entire episode lasts only a minute or two, and in the morning the child recalls nothing of it or only a vague unpleasant dream. It has been suggested that night terrors and somnambulism represent impaired or partial arousal from deep sleep, as EEGs taken during such episodes show a waking type of mixed frequency and alpha pattern. Children with night terrors and somnambulism do not show an increased incidence of psychologic abnormalities and tend to outgrow these disorders. The persistence of such problems into adult life, however, has, in a small number of cases, been associated with psychopathology (Kales et al). It has been found that diazepam, which reduces the duration of the deep stages of sleep, will prevent night terrors. Selective serotonin reuptake inhibitors have also been used successfully, especially when night terrors are associated with sleepwalking. Frequent night terrors have reportedly been eliminated by having parents awaken the child for several successive nights, just prior to the usual time of the attack or at the first sign of restlessness and autonomic arousal (Lask).
Frightening dreams or nightmares are far more frequent than night terrors and affect children and adults alike. They occur during periods of normal REM sleep and are particularly prominent during periods of increased REM sleep (REM rebound) following the withdrawal of alcohol or other sedative-hypnotic drugs that had suppressed REM sleep chronically. Autonomic changes are slight or absent, and the content of the dreams can usually be recalled in considerable detail. Some of these dreams (e.g., the ones occurring in the alcohol-withdrawal period) are so vivid that the patient may later have difficulty in separating them from reality; indeed, they may merge with the hallucinations of delirium tremens. Nightmares are of little significance as isolated events. Fevers dispose to them, as do conditions such as indigestion and the reading of bloodcurdling stories or exposure to terrifying movies or television programs before bedtime (truly). Some patients report nightmares and extremely vivid dreams when first taking certain medications such as beta blockers and, particularly in our experience, L-dopa. We have also consulted on a few patients who complained of almost nightly nightmares and concurrent severe headaches, but without apparent depression or other psychiatric illness; the nature of their problem was obscure. Persistent nightmares may be a pressing medical complaint and are often accompanied by other behavioral disturbances or anxieties.
Childhood Somnambulism and Sleep Automatism
This condition occurs far more commonly in children (average age: 4 to 6 years) than in adults, and is often associated with nocturnal enuresis and night terrors, as indicated above. It is estimated that 15 percent of children have at least one episode of sleepwalking, and that 1 in 5 sleepwalkers has a family history of this disorder. Motor performance and responsiveness during the sleepwalking incident vary considerably. The most common behavioral abnormality is for a patient to sit up in bed or on the edge of the bed without actually walking. When walking about the house, he may turn on a light or perform some other familiar act. There may be no outward show of emotion, or the patient may be frightened (night terror), but the frenzied, aggressive behavior of some adult sleepwalkers, described below, is rare in the child. Usually the eyes are open, and such sleepwalkers are guided by vision, thus avoiding familiar objects; the sight of an unfamiliar object may awaken them. Sometimes they make no attempt to avoid obstacles and may injure themselves. If spoken to, they make no response; if told to return to bed, they may do so, but more often they must be led back. Sometimes they repeatedly mutter strange phrases or perform certain repetitive acts, such as pushing against a wall or turning a doorknob back and forth. The episode lasts for only a few minutes, and the following morning, they usually have no memory of it, or only a fragmentary recollection.
A popular belief is that the sleepwalker is acting out a dream. The observations of sleep laboratories are entirely at variance with this view, as somnambulism has been found to occur almost exclusively during deeper stages of NREM sleep (stage N3) and during the first third of the night when dreaming is least likely to occur. In fact, the entire nocturnal sleep pattern of such individuals does not differ from normal. Also, there is no evidence that somnambulism is a form of epilepsy. It is probably allied to talking in one's sleep, although the two conditions seldom occur together. Sleepwalking must be distinguished from fugue states and ambulatory automatisms of complex partial seizures discussed in Chap. 16.
The major consideration in the treatment of childhood somnambulism is to guard patients against injury by locking doors and windows, removing dangerous objects from the patients' usual routes of march, having them sleep on the ground floor, etc. Children usually outgrow this disorder; parents should be reassured on this score and disabused of the notion that somnambulism is a sign of psychiatric or any other disease.
The onset of sleepwalking or night terrors for the first time in adult life is most unusual, and, in an occasional case, may suggest the presence of psychiatric disease or drug intoxication. Almost always, the adult sleepwalker has a history of sleepwalking as a child, although there may have been a period of freedom between the childhood episodes and their reemergence in the third and fourth decades. Adult somnambulism also occurs during N3 of NREM sleep, but unlike the childhood type, is not confined to the earlier part of the night. If one extends the category of somnambulism to all forms of nocturnal wandering, it seems to be remarkably common, with a lifetime prevalence of 29% of U.S. adults according to the survey by Ohayon and colleagues.
Somnambulism in the adult, as in the child, can be a purely passive event unaccompanied by fear or other signs of emotion. More frequently, however, the attack is characterized by frenzied or violent behavior associated with fear and tachycardia, like that of a night terror and sometimes with self-injury. Very rarely, crimes have reportedly been committed during sleepwalking, but the authors are skeptical that organized and planned sequential activity is possible. The finding of normal sleep patterns on polysomnography distinguishes these attacks from complex partial seizures. They can be eliminated or greatly reduced by the administration of clonazepam (0.5 to 1.0 mg) at bedtime. Some patients respond better to a combination of clonazepam and phenytoin or to flurazepam (Kavey et al).
An associated but unclassifiable disorder is "sleep eating" in which the individual seeks out mainly carbohydrates and is only aware of their actions on the following morning when they see the mess they have left. Also, in the provocatively named "sexomnia," the individual, male or female, engages in sexual activity, sometimes forcefully, and has no recollection of the events. The status of these syndromes as authentic parasomnias is unclear.
REM Sleep Behavior Disorder
This is a recognized parasomnic disorder, occurring in adult life, most commonly in older men without a history of childhood sleepwalking. It is characterized by attacks of vigorous, agitated, and often dangerous motor activity accompanied by vivid dreams (Mahowald and Schenck). The characteristic features are angry speech with shouting, violent activity with injury to self and bed mate, a very high arousal threshold, and the variable but sometimes detailed recall of a nightmare of being attacked and fighting back or attempting to flee. The episodes vary in frequency in affected individuals, occurring once every week or two or several times nightly. The episodes, which occur exclusively during REM sleep, usually in the second half of the night, are out of keeping with the patient's waking personality. Polysomnographic recordings during these episodes have disclosed augmented muscle tone but no seizure activity.
The rare appearance of this disorder with pontine infarctions has been mentioned earlier in the chapter. However, in a series of 93 cases of REM sleep behavior disorder reported by Olson and colleagues, more than half were associated with some other neurologic disorder, in particular Parkinson disease, multiple system atrophy, and Lewy-body dementia, but in other series, with a number of degenerative and varied neurologic conditions. A more systemic polysomnographic examination of 457 patients with Parkinson disease by Sixel-Döring and colleagues found REM sleep behavior disorder in 46%. Viewed from another perspective, Postuma and coworkers have reported that one-quarter of individuals with idiopathic REM sleep disorder later developed a neurodegenerative disorder, similar to or slightly lower than other series. These observations have led to the suggestion that this disorder is an early manifestation of a degenerative brain disease characterized by the deposition of alpha-synuclein in certain neuronal systems, as summarized by Boeve and associates.
The episodes can be suppressed by the administration of clonazepam in doses of 0.5 to 1.0 mg at bedtime and by melatonin, 3 to 12 mg. The advantage of the latter is that sleep apnea is not affected as it is with benzodiazepines. Discontinuation of medication, even after years of effective control, has resulted in relapse. Antidepressants are said to exacerbate the disorder with the possible exception of bupropion.
The patient and family should probably be advised about safety including removal of weapons, sharp objects and try to insure spousal safety by this and by means of sleeping in another room.
Nocturnal Epilepsy (See Also Chap. 16)
It has long been known that seizures, both convulsive and nonconvulsive, often occur during sleep, especially in children. This is such a frequent occurrence that the practice of inducing sleep has been adopted as an activating EEG procedure to obtain confirmation of epilepsy. Seizures may occur soon after the onset of sleep or at any time during the night, but mainly in stages 1 and 2 of NREM sleep or, rarely, in REM sleep. They are also common during the first hour after awakening. On the other hand, deprivation of sleep may be conducive to a seizure.
Sleeping epileptic patients may attract attention to their seizures by a cry, violent motor activity, unusual but stereotyped actions, such as sitting up and crossing the arms over the chest, the adoption of a "fencing" posture, or labored breathing. After the tonic-clonic phase, patients become quiet and fall into a state resembling deep sleep, but they cannot be aroused from it for some minutes or longer. If the nocturnal seizure is unobserved, the only indication of it may be disheveled bedclothes, a few drops of blood on the pillow from a bitten tongue, wet bed linen from urinary incontinence, or sore muscles. Or the occurrence of a seizure may be disclosed only by confusion, muscle soreness, or headache, the common aftermaths of a major generalized seizure. Rarely, a patient may die in an epileptic seizure during sleep, sometimes from smothering in the bed clothes or aspirating vomitus or for some obscure reason (possibly respiratory or cardiac dysrhythmia). Epilepsy occasionally occurs in conjunction with night terrors and somnambulism; the question then arises whether the latter disorders represent postepileptic automatisms. Usually no such relationship is established. EEG studies during a nocturnal period of sleep are most helpful in such cases. Measurement of serum creatine kinase concentration in the hours following an event may distinguish seizure from night terrors, and the other described sleep-related motor behaviors.
Pathologic Excessive Sleep (Hypersomnia)
Encephalitis lethargica, or von Economo "epidemic encephalitis," the remarkable illness that appeared on the medical horizon as a pandemic following World War I, provided some of the most dramatic instances of pathologic somnolence. Protracted sleep lasting for days to weeks was such a prominent symptom of this disease that it was called sleeping sickness (a term also applied to African trypanosomiasis, as noted below). The patient appeared to be in a state of continuous sleep, or somnosis, and could be kept awake only by constant stimulation. Although the infective agent of von Economo disease was never isolated, the pathologic anatomy was fully disclosed by many excellent studies, all of which demonstrated a destruction of neurons in the midbrain, subthalamus, and hypothalamus. Patients who survived the acute phase of the illness often had difficulty in reestablishing their normal sleep–wake rhythm. As the somnolence disappeared, some patients exhibited a reversal of the normal pattern, tending to sleep by day, and stay awake at night; many of them also developed a parkinsonian syndrome months or years later. The hypersomnia was possibly related to destruction or functional paralysis of dopamine-rich neurons in the substantia nigra, resulting in overactivity of the raphe (serotonergic) neurons.
Hypersomnia is also a manifestation of trypanosomiasis, the common cause of sleeping sickness in Africa, and of other diseases localized to the mesencephalon, and the floor and walls of the third ventricle. Small tumors in this area have been associated with arterial hypotension, diabetes insipidus, hypo- or hyperthermia, and protracted somnolence lasting many weeks. Such patients can be aroused; but if left alone, they immediately fall asleep. Traumatic and vascular lesions and other diseases affecting the mesencephalon may have a similar effect.
Sleep drunkenness is the name given to a special form of hypersomnia, characterized by a failure of the patient to attain full alertness for a protracted period after awakening. Unsteadiness, drowsiness, disorientation, and automatic behavior are the main features. This disorder is usually associated with idiopathic hypersomnia, and sometimes with sleep apnea or other forms of sleep deprivation, but often no such connection can be discerned, in which case a motivational factor should be suspected.
An interesting type of transient unresponsiveness in elderly patients, as described by Haimovic and Beresford, has been in our experience akin to a deep sleep, but the EEG has not shown sleep patterns.
Kleine in 1925 and Levin in 1936 described an episodic disorder characterized by somnolence and overeating. For days or weeks, the patients, mostly adolescent boys, sleep 18 h or more a day, awakening only long enough to eat and attend to toilet needs. They appeared dull, often confused, and restless, and were sometimes troubled by hallucinations. In the series of 18 cases collected by Critchley, the age range was from puberty to 45 years. There may be a brief prodromal period of inertia and drowsiness. The duration of nocturnal sleep may be greatly prolonged, or, as in our patients referred to below, they may sleep for days on end. Food intake during and around the period of hypersomnia may exceed three times the normal (bulimia) and occurs almost compulsively during brief periods of semiwakefulness; to a variable extent, there are other behavioral changes such as social withdrawal, negativism, slowness of thinking, incoherence, inattentiveness, and disturbances of memory. The somnolence has been well studied by modern laboratory methods; except for the total duration of sleep, the individual components of the NREM and REM cycles are normal. Between episodes these patients are behaviorally and cognitively normal.
The basis of this condition has never been clarified. A psychogenic mechanism has been proposed, without foundation in our opinion. The syndrome usually disappears during adulthood, and there is limited pathologic material (see further discussion in the context of hypothalamic syndromes in Chap. 27). A series of 108 cases reviewed by Arnulf and colleagues frame the clinical features; a predominance of males, higher C-reactive protein than controls, and a history of early childhood developmental problems. There was no human leukocyte antigen (HLA) clustering, but children of Jewish heritage were overrepresented. We have cared for a sibling pair who had the illness into young adulthood (Katz and Ropper). The case reported by Carpenter and coworkers, in which an acute and chronic inflammation in the medial thalamus, but not the hypothalamus, was found, must be questioned as representative of the idiopathic adolescent condition. Their patient was a man 39 years of age who had episodes of diurnal drowsiness, hyperphagia (intermittently relieved by methylphenidate), and hypersexuality over a period of months. In some patients with this disorder, schizophrenic and sociopathic symptoms have been recorded between attacks, raising doubt as to whether all the reported cases are of the same type. We have seen variants of this syndrome manifesting themselves in drowsiness and extreme inactivity lasting for a few weeks, then with a complete return to normalcy. In two of our patients, the use of serotonergic antidepressants lengthened the interval between episodes.
No consistent change in the level of hypocretin (orexin) has been found in the spinal fluid, as occurs in narcolepsy (see further on), and the two disorders are distinct. In one typical case, there was pronounced hypoperfusion of the left medial temporal lobe both during and between attacks, but the interpretation of this finding is unclear (Portilla et al).
No treatment has been consistently effective (e.g., antidepressant drugs), but some of the stimulants that are used for the treatment of narcolepsy may be useful (see further on).
Finally, it should be mentioned that sleep laboratories now recognize a form of idiopathic hypersomnia in which there are repeated episodes of drowsiness throughout the day. This condition is discussed further on, in relation to the diagnosis of narcolepsy, with which it is most often confused.
A related disorder has been described of "menstrual related hypersomnia" that has a cyclic catamenial nature.
Sleep Apnea and Excessive Daytime Sleepiness
Excessive daytime sleepiness is a common complaint in general medical practice (Table 19-2). Certainly, the most frequent causes are inadequate sleep and the use of any one of the large variety of medications that are not prescribed primarily for their sedative effect. Abuse of alcohol and illicit drugs should also be included in this category. Most conditions associated with severe fatigue produce daytime sleepiness and a desire to nap. A notable medical cause is infectious mononucleosis but many other viral infections have the same effect. Certain chronic neurologic conditions can produce fatigue and sleepiness, multiple sclerosis being the outstanding example. Among general medical conditions, hypothyroidism and hypercapnia must always be considered when daytime sleepiness is a prominent feature. One must not overlook the possibility that excessive daytime drowsiness is the result of repeated episodes of sleep apnea, discussed below, or the disruption of nocturnal sleep by disorders such as the restless legs syndrome.
Table 19-2 Causes of Daytime Sleepiness ||Download (.pdf)
Table 19-2 Causes of Daytime Sleepiness
1. Medications (including many types of sedatives, tranquilizers, anticonvulsants, antihistaminics, antidepressants, β-adrenergic blockers, and atropinic drugs), L-dopa and dopaminergic agonists, abuse of alcohol and illicit drugs
2. Acute medical illness of the mononucleosis type, including mundane respiratory and gastrointestinal infections
3. Postsurgical, postconcussive, and postanesthetic states
4. Chronic neurologic diseases: multiple sclerosis, dementias
6. Metabolic derangements: hypothyroidism, Addison disease, severe diabetes
7. Encephalitic diseases
A. Following viral encephalitis
C. Encephalitis lethargica (historical)
8. Lesions of the hypothalamus
A. Kleine-Levin syndrome
B. Hypothalamic tumor or granuloma
9. Sleep apnea syndromes; central and obstructive
11. Idiopathic hypersomnia
As mentioned above, REM sleep is characterized by irregular breathing, and this may include several brief periods of apnea up to 10 s in duration. Such apneas and those occurring at the onset of sleep are not in themselves considered to be pathologic. In some individuals, however, sleep-induced apneic periods are particularly frequent and prolonged (> 10 s) and such a condition is responsible for a variety of clinical disturbances in children and adults. This pathologic form of sleep apnea may be the result of a reduction of respiratory drive (so-called central apnea), an obstruction of the upper airway, or a combination of these two mechanisms.
Apnea of the obstructive type in which the posterior pharyngeal muscles collapse and narrow the upper airway is far more common than the central variety. Obstructive apnea is associated with obesity and also accompanies acromegaly, hypothyroidism or myxedema, micrognathia, and myotonic dystrophy. In children, more than in adults, adenotonsillar hypertrophy may be a factor. Instances occur as a result of neuromuscular diseases that weaken the posterior pharyngeal musculature; motor neuron disease is the most common example of this group. Obstructive sleep apnea is characterized by noisy snoring of a cyclic type. After a period of regular albeit noisy breathing, there occurs a waning of breathing efforts; then, despite repeated inspiratory efforts, airflow ceases. Following a prolonged period of apnea (10 to 30 s or even longer), the patient makes a series of progressively greater breathing efforts until breathing resumes, accompanied by very loud snorting sounds and a brief arousal.
Obstructive sleep apnea occurs during both REM and NREM sleep. The upper respiratory muscles (genioglossus, geniohyoid, tensor veli palatini, and medial pterygoid) normally contract just before the diaphragm contracts, resisting the collapse of the oropharynx. If the airway is obstructed or the muscles are weakened and then go slack, the negative intrathoracic pressure causes narrowing of this passage. Sedative medications, alcohol intoxication, excessive tiredness, a recent stroke, head trauma or other acute neurologic disease, and primary pulmonary disease may all exaggerate obstructive sleep apnea, particularly in the obese patient with a tendency to snore.
Hypoxia or perhaps other stimuli induce an arousal response, either a lightening of sleep or a very brief awakening, which is followed by an immediate resumption of breathing. The patient quickly falls asleep again and the sequence is repeated, several hundred times a night in severe cases, greatly disrupting the sleep pattern and reducing the total sleep time. Paradoxically, these patients are very difficult to rouse at all times during the night.
Obstructive sleep apnea is predominantly a disorder of overweight, middle-aged men and usually presents as excessive daytime sleepiness, a complaint that is sometimes mistaken for narcolepsy (see below). Other patients, usually those with the much-less-common central form of apnea, complain mainly of a disturbance of sleep at night, or insomnia, which may be incorrectly attributed to anxiety or depression. The occurrence of an obstructive sleep apnea is accompanied after a period of weeks or months by progressive hemoglobin oxygen desaturation, hypercapnia and hypoxia, a transient increase in systemic and pulmonary arterial pressures, and sinus bradycardia or other arrhythmias. Morning headache, inattentiveness, grogginess, and decline in school or work performance are other symptoms attributable to sleep apnea. Ultimately, systemic and pulmonary arterial hypertension, cor pulmonale, polycythemia, and heart failure may develop. When combined with obesity, these symptoms have been referred to as the "pickwickian syndrome," so named by Burwell and coworkers, who identified this clinical syndrome with that of the extraordinarily sleepy, red-faced, fat boy described by Dickens in The Pickwick Papers.
The full-blown syndrome of obstructive sleep apnea is readily recognized by the features of daytime sleepiness, loud snoring, and the typical habitus of affected individuals. However, in patients who complain only of excessive daytime sleepiness or insomnia, the diagnosis may be elusive and require all-night polysomnographic sleep monitoring.
This disorder has been observed in patients with a variety of severe and life-threatening lower brainstem lesions—bulbar poliomyelitis, lateral medullary infarction, spinal (high cervical) surgery, syringobulbia, brainstem encephalitis, as well as with striatonigral degeneration, Creutzfeldt-Jakob disease, anoxic encephalopathy, and olivopontocerebellar degeneration. When a unilateral lesion (e.g., infarction) of the medulla is the cause, there is almost always involvement of crossing fibers between respiratory nuclei (see discussion in Chap. 26).
In addition to these symptomatic forms of sleep apnea, there is a disorder referred to as primary, or idiopathic, hypoventilation syndrome ("Ondine's curse," as described in Chap. 26). This last term is now applied to many forms of total loss of automatic breathing, especially during sleep. Awakenings during the night are frequent, usually after an apneic period, and insomnia is a common complaint. Snoring is mild and intermittent. In the few autopsied cases of congenital central hypoventilation of childhood, Liu and colleagues found the external arcuate nuclei of the medulla to be absent, and the neuron population in the medullary respiratory areas to be depleted.
Complex sleep apnea, or "treatment emergent central sleep apnea," occurs most often in patients with cardiovascular conditions, particularly congestive heart failure, wherein, after sleep apnea is treated with positive airway pressure, central apnea emerges.
The approach is governed by the severity of symptoms and the predominant type of apnea, central or obstructive. In the treatment of obstructive apnea, continuous positive airway pressure (CPAP) or bilevel positive airway pressure (BIPAP) is the most useful measure. These therapies are delivered by a tight-fitting nasal mask that is worn at night and connected to a pressure-cycled ventilator that is triggered by the patient's breath. The increased airway pressure maintains patency of the naso- and oropharynx, thereby reducing the obstruction. A nasal device that passively resists nasal collapse by providing expiratory resistance is used by some specialists for mild sleep apnea. All of these approaches have a level of discomfort that is not tolerable to some patients.
Patients benefit from losing weight, lateral positioning during sleep, and avoidance of alcohol and other sedative drugs. Surgical correction of an upper airway defect may be helpful, but it is difficult to predict which patients will benefit. There are no clear guidelines for procedures such as uvulopalatopharyngoplasty and related surgeries or uvulectomy and tonsillectomy except in children. These may obliterate snoring more than it ameliorates the sleep apnea. Oral alignment devices that are produced by dentists, aimed at advancing the mandible, have been helpful to some patients, especially for those who cannot tolerate positive pressure.
Those few patients with the most severe hypersomnia and cardiopulmonary failure who cannot tolerate nocturnal positive pressure ventilation require tracheostomy and nocturnal respirator care. (See Parkes for a fuller account of therapeutic measures.) Some patients with non-obstructive apnea may also benefit from nighttime treatment with CPAP, but the results are far less consistent than with the obstructive type.
In central apnea, any underlying abnormality, such as congestive heart failure or nasal obstruction, should, of course, be treated insofar as possible. Where no underlying cause can be found, one of several medications—acetazolamide, medroxyprogesterone, protriptyline, and particularly clomipramine—may be helpful in the short run (Brownell et al). However, drug treatment has proven generally unsatisfactory. Low-flow oxygen may also be useful in reducing central sleep apnea.
This clinical entity has long been known to the medical profession. Gélineau gave it the name narcolepsy in 1880, although several authors had described the recurring attacks of irresistible sleep even before that time. Gélineau had also mentioned that the sleep attacks were sometimes accompanied by falls ("astasias"), but it was Loewenfeld, in 1902, who first recognized the common association between the sleep attacks and the temporary paralysis of the somatic musculature during bouts of laughter, anger, and other emotional states; this was referred to as cataplectic inhibition by Henneberg (1916), and later as cataplexy by Adie (1926). The term sleep paralysis—used to designate the brief, episodic loss of voluntary movement that occurs during the period of falling asleep (hypnagogic, or predormital) or less often when awakening (hypnopompic, or postdormital)—was introduced by S.A. Kinnier Wilson in 1928. Actually, Weir Mitchell had described this latter disorder in 1876, under the title of night palsy.
Sometimes sleep paralysis is accompanied or just preceded by vivid and terrifying hallucinations (hypnagogic hallucinations), which may be visual, auditory, vestibular (a sense of motion), or somatic (a feeling that a limb or finger or other part of the body is enlarged or otherwise transformed). These four conditions—narcolepsy, cataplexy, hypnagogic paralysis, and hallucinations—constitute a clinical tetrad. Wilson has reviewed the historical aspects and early accounts of this subject. The most important observations regarding the pathophysiology of this process have been special relationship to a disordered pattern of REM sleep, and the more recent finding of abnormalities in hypothalamic substances that induce sleep, as discussed below.
This syndrome is encountered regularly by neurologists; Daly and Yoss recorded about 100 new cases a year at the Mayo Clinic. Dement and colleagues have estimated the prevalence at 50 to 70 per 100,000 in the San Francisco and Los Angeles areas. Men and women are affected equally. Several papers have found an epidemiologic relationship to just preceding outbreaks of H1N1 pandemics and the vaccine (see Ham et al and Dauvilliers et al).
As a rule, narcolepsy has a gradual onset between the ages of 15 and 35 years; in fully 90 percent of narcoleptics, the condition is established by the 25th year of life. Narcolepsy is usually the first symptom, less often cataplexy, and rarely sleep paralysis. The essential disorder is one of frequent attacks of irresistible sleepiness. Several times a day, usually after meals or while sitting in class or in other boring or sedentary situations, the affected person is assailed by an uncontrollable desire to sleep. The eyes close, the muscles relax, breathing deepens slightly, and by all appearances, the individual is dozing. A noise, a touch, or even the cessation of the lecturer's voice is enough to awaken the patient. The periods of sleep rarely last longer than 15 min unless the patient is reclining, when he may continue to sleep for an hour or longer. At the conclusion of a nap, the patient feels somewhat refreshed. It should be emphasized that there are many narcoleptics who tend to be pervasively drowsy throughout the day. What distinguishes the typical narcoleptic sleep attacks from commonplace postprandial drowsiness and napping is the frequent occurrence of the former (two to six times every day as a rule), their irresistibility, and their occurrence in unusual situations, as while standing, eating, or carrying on a conversation. Blurring of vision, diplopia, and ptosis may attend the drowsiness and may bring the patient first to an ophthalmologist.
In addition to episodes of outright sleep, narcoleptics, like other very drowsy persons, may experience episodes of automatic behavior and amnesia. Initially the patient feels drowsy and may recall attempts to fight off the drowsiness, but gradually he loses track of events. The patient may continue to perform routine tasks automatically but does not respond appropriately to a new demand or answer complex questions. Often there is a sudden burst of words, without meaning or relevance to what was just said. Such an outburst may terminate the attack, for which there is complete or nearly complete amnesia. In many respects, the attacks resemble episodes of nocturnal sleepwalking. Such attacks of automatic behavior and amnesia are common, occurring in more than half of a large series of patients with narcolepsy-cataplexy (Guilleminault and Dement). Affected patients are frequently involved in driving accidents, even more frequently than epileptics.
Nocturnal sleep is often disrupted and reduced in amount. The number of hours in a 24-h day spent in sleep by the narcoleptic is no greater than that of a normal individual. Narcoleptics have an increased incidence of sleep apnea and periodic leg and body movements, but not of somnambulism.
Approximately 70 percent of narcoleptics first seeking help will report having some form of cataplexy, and about half of the remainder will develop cataplexy later in life. Cataplexy refers to a sudden loss of muscle tone brought on by strong emotion—that is, circumstances in which hearty laughter or, more rarely, excitement, surprise, anger, or intense athletic activity cause the patient's head to fall forward, the jaw to drop, the knees to buckle, even with sinking to the ground—all with perfect preservation of consciousness. Cataplectic attacks occur without provocation in perhaps 5 percent of cases. The attacks last only a few seconds or a minute or two and are of variable frequency and intensity. In most of our patients, they have appeared at intervals of a few days or weeks. Exceptionally, there are many attacks daily and even status cataplecticus, in which the atonia lasts for hours. This is more likely to happen at the beginning of the illness or upon discontinuing tricyclic medication.
Most attacks of cataplexy are partial (e.g., only a dropping of the jaw or "weakening of the knees"). Wilson found that the tendon reflexes were abolished during the attack. Pupillary reflexes are absent in some cases.
Rarely, cataplexy precedes the advent of sleep attacks, but usually it follows them, sometimes by many years. Sleep paralysis and hypnagogic hallucinations together are stated to occur in about half the patients. Of course, hypnagogic paralysis and hallucinations occur in otherwise normal persons, and normal children, especially when tickled, may laugh to the point of cataplexy. About 10 percent of persons with sleep attacks indistinguishable from those of narcolepsy have none of the associated phenomena ("independent narcolepsy"), and in these cases, REM periods are not found consistently at the onset of sleep (see further on).
Once established, narcolepsy and cataplexy usually continue for the remainder of the patient's life. The degree of sleepiness rarely lessens, although cataplexy, sleep paralysis, and hallucinations improve or disappear with age in about one-third of patients who have those features (Billiard and Cadilhac). No other condition is consistently associated with narcolepsy-cataplexy, and none develops later.
A familial component has been recognized for years; the risk of narcolepsy in a first-degree relative of an affected individual is 1 to 2 percent, more than 25 times that in the general population. As reviewed by Chabas and colleagues, important insights into the pathogenesis have come from studies of recessively inherited narcolepsy in three species of dogs, in which mutations have been identified in a gene encoding a receptor for the protein hypocretin (Lin et al). These studies implicate the peptide hypocretin in the control of sleep. The hypocretins were thought in the past to regulate feeding behavior and energy metabolism; indeed, they were also designated "orexins," from the Greek word for appetite. In mice, inactivation of two hypocretin receptors reproduces narcolepsy. In both humans and animals, hypocretin-containing neurons in the hypothalamus send projections widely through the brain and particularly to structures implicated in control of sleep as discussed earlier and shown in Fig. 19-4: the locus ceruleus (noradrenergic), the tuberomammillary nucleus (histaminergic), the raphe nucleus (serotonergic), and the ventral tegmental area (dopaminergic).
A number of compelling observations implicate hypocretin and its receptors in human narcolepsy. First, a narcoleptic patient has been described with a mutation in the gene encoding human hypocretin. Second, hypocretin-secreting neurons are depleted in the brains of human narcoleptics, and CSF hypocretin levels are reduced or absent in affected patients. In some studies, the absence of CSF hypocretin distinguished narcoleptic individuals from patients with other categories of sleep disorders.
Perhaps surprisingly, several lines of evidence suggest an autoimmune causation for narcolepsy. For example, it has long been known that there is an almost universal association with specific alleles of the histocompatibility antigen HLA-DQ (B1-0602) (Neely et al; Kramer et al). Therapeutic approaches to narcolepsy based on a presumed autoantibody have also been developed as noted below. Because the mode of inheritance of narcolepsy is not clearly mendelian (Kessler et al), it has been proposed that the disease reflects a genetic predisposition, possibly with a superimposed autoimmune reaction that impairs the function of hypocretin neuronal systems or damages the neurons that secrete the peptide.
Finally, the approximate similarities to the post infectious sleep states of von Economo encephalitis and, as provocatively, the aforementioned increased incidence of narcolepsy after outbreaks of H1N1 respiratory infection, or after administration of the vaccine, implicates an infectious or post infectious inflammatory cause, ostensibly affecting the hypothalamic nuclei (Dauvillers et al, 2010).
As mentioned earlier, a secondary or symptomatic narcolepsy syndrome on occasion results from cerebral trauma, multiple sclerosis, craniopharyngioma, or other tumors of the third ventricle or upper brainstem, head trauma, or a sarcoid granuloma within the hypothalamus (Servan et al).
Our understanding of narcolepsy was greatly advanced by the demonstration by Dement and his group that this disorder is associated with a reversal in the order of the two states of sleep, with REM rather than NREM sleep occurring at the onset of the sleep attacks. Not all the sleep episodes of the narcoleptic begin with REM sleep, but almost always a number of sleep attacks with such an onset can be identified in narcoleptic-cataplectic patients in the course of a polysomnographic sleep study. The hypnagogic hallucinations, cataplexy, and sleep-onset paralysis (caused by inhibition of anterior horn cells) all coincide with the REM period. These investigators have also shown that the night sleep pattern of patients with narcolepsy-cataplexy may begin with a REM period. This may occur in normal subjects, though infrequently and usually with severe sleep deprivation. Furthermore, the nocturnal sleep pattern is altered in narcoleptics, who have frequent body movements and transient awakenings and a decrease in sleep stage N3, as well as in total sleep duration. Another important finding in narcoleptics is that sleep latency (the interval between the point when an individual tries to sleep and the point of onset of EEG sleep patterns), measured repeatedly in diurnal nap situations, is greatly reduced. Thus, narcolepsy is not simply a matter of excessive diurnal sleepiness (essential daytime drowsiness) or even a disorder of REM sleep but a generalized disorganization of sleep–wake function.
The greatest difficulty in diagnosis relates to the problem of separating narcolepsy from the daytime sleepiness of sedentary, obese adults who, if unoccupied, doze readily after meals, while watching television or in the theater. Many of these patients prove to have obstructive sleep apnea. Excessive daytime somnolence, easily mistaken for narcolepsy, may also attend heart failure, hypothyroidism, excessive use of soporific, other medications including antihistamines, use of alcohol, cerebral trauma, and certain brain tumors (e.g., craniopharyngioma; see Table 19-2). A more serious form of recurrent daytime sleepiness, referred to as independent narcolepsy or essential narcolepsy, is described further on. However, both of these forms of daytime drowsiness are isolated disturbances, lacking the other disturbances of sleep and motor activity that characterize the narcolepsy syndrome. The brief attacks of automatic behavior and amnesia of the narcoleptic must be distinguished from hysterical fugues and complex partial seizures.
Cataplexy must also be distinguished from syncope, drop attacks (Chap. 17), and atonic seizures; in atonic seizures, consciousness is temporarily abolished. The careful documentation of narcolepsy by laboratory techniques is imperative when the diagnosis is in doubt, in part because of the potential for abuse of stimulant drugs used for treatment. Overnight polysomnography followed by a standardized multiple sleep latency test,
in which the patient is afforded opportunities for napping at 2-h intervals, permit the quantification of drowsiness and increase the probability of detecting short-latency REM activity (within 15 min from the onset of each sleep period). According to some investigators, a reduced level (below 110 pg/mL) of hypocretin in the spinal fluid is virtually diagnostic of narcolepsy in the proper clinical circumstances (see Mignot et al). We would comment, however, that it is not necessary to resort to any of these studies in clinically typical cases.
No single therapy will control all the symptoms. Narcolepsy responds best to (1) strategically placed 15- to 20-min naps (during lunch hour, before or after dinner, etc.); (2) the use of stimulant drugs—modafinil, dextroamphetamine sulfate, or methylphenidate hydrochloride to heighten alertness; and (3) antidepressants (sertraline, venlafaxine, protriptyline, imipramine, or clomipramine) for control of cataplexy. All these drugs are potent suppressants of REM sleep. Monoamine oxidase (MAO) inhibitors also inhibit REM sleep and can be used if they are tolerated. Modafinil (200 mg daily, up to 600 mg in divided doses) may prove to be the safest of the stimulants (Fry), but experience with this agent is still being acquired. Methylphenidate, because of its prompt action and relative lack of side effects, is also widely used. It is usually given in doses of 10 to 20 mg tid on an empty stomach. Alternatively, amphetamine 5 to 10 mg may be given 3 to 5 times a day; this is ordinarily well tolerated and does not cause wakefulness at night. Pemoline, a potent stimulant (50 to 75 mg daily) is no longer available in the United States because of potential hepatic toxicity. The tricyclic antidepressants had been used to reduce cataplexy, but they have been overtaken by selective serotonin reuptake inhibitors such as sertraline and by norepinephrine reuptake inhibitors such as venlafaxine. Sodium oxybate, whose active agent is gamma-hydroxybutyrate, is also beneficial for cataplexy and narcolepsy in many
The combined use of these stimulant and tricyclic antidepressant drugs is often indicated. A problem with the stimulant drugs is the development of tolerance over a 6- to 12-month period, which requires the switching and periodic discontinuation of drugs. Excessive amounts of amphetamines may induce a schizophreniform psychosis. The stimulant drugs and the tricyclic antidepressants increase catecholamine levels; their chronic administration may produce hypertension.
An entirely different approach, based on a presumed autoimmune attack on hypothalamic neurons, has introduced immune globulin infusions in early cases of narcolepsy. This must still be considered preliminary but the results are interesting (see Dauvillers et al, 2004).
Narcoleptics must be warned of the dangers of falling asleep and lapses of consciousness while driving or during engagement in other activities that require constant alertness. The earliest feeling of drowsiness should prompt the patient to pull off the road and take a nap. Long-distance driving should probably be avoided completely.
Idiopathic Hypersomnia (Essential Narcolepsy; NREM Narcolepsy)
As has been indicated, recurrent daytime sleepiness may be the presenting symptom in a number of varied disorders other than narcolepsy. When chronic daytime sleepiness occurs repeatedly and persistently without known cause, it is classified as essential or idiopathic hypersomnolence. Roth distinguishes this state from narcolepsy on the basis of longer and unrefreshing daytime sleep periods, deep and undisturbed night sleep, difficulty in awakening in the morning or after a nap ("sleep drunkenness"), all of these occurring in the absence of REM-onset sleep and cataplexy. Admittedly, this condition proves difficult to distinguish from narcolepsy unless laboratory studies exclude the latter, and even then, there is overlap between the two syndromes (Bassetti and Aldrich). Treatment, however, is the same as that for narcolepsy. Idiopathic hypersomnia, as defined in this manner, proves to be a rare syndrome once narcolepsy, and all other causes of daytime sleepiness have been excluded.
This state, as remarked earlier, has been induced in animals by lesions in the tegmentum (median raphe nuclei) of the pons. Comparable states are known to occur in humans but are rare. Asomnia in hospital practice is a result of delirium of any type, including delirium tremens and drug-withdrawal states. Drug-induced psychoses and mania may induce a similar state. We have seen a number of patients with a delirious hyperalertness lasting a week or more after temporofrontal cerebral contusions or in association with a hypothalamic lymphoma. None of the various treatments we have tried has been successful in suppressing this state. It was transitory in the traumatic cases.
Additional Syndromes Occurring during Sleep
Sleep Palsies and Acroparesthesias
Several paresthetic disturbances, sometimes distressing in nature, may arise during sleep. Everyone is familiar with the phenomenon of an arm or leg "falling asleep." Immobility of the limbs and maintenance of uncomfortable postures, without any awareness of them, permit undue pressure to be applied on peripheral nerves (especially the ulnar, radial, and peroneal). Pressure of the nerve against the underlying bone may interfere with intraneural function in the compressed segment of nerve. Sustained pressure may result in a sensory and motor paralysis—sometimes referred to as sleep or pressure palsy. Usually, this condition lasts only a few hours or days, but if compression is prolonged recovery may be delayed. Deep sleep or a stupor, as in alcohol intoxication or anesthesia, renders patients especially liable to pressure palsies merely because they are not able to heed the discomfort of a sustained unnatural posture.
Acroparesthesias are frequent in adult women and are not unknown in men. The patient, after being asleep for a few hours, is awakened by numbness or a tingling, prickling, "pins-and-needles" feeling in the fingers, and hands. There are also aching, burning pains or tightness, and other unpleasant sensations. With vigorous rubbing or shaking of the hands or extension of the wrists, the paresthesia subsides within a few minutes, only to return later or upon first awakening in the morning. At first, there is a suspicion of having slept on an arm, but the frequent bilaterality of the symptoms and their occurrence regardless of the position of the arms dispels this notion. Usually the paresthesia is in the distribution of the median nerves, and almost invariably proves to be caused by carpal tunnel syndrome.
Nocturnal grinding of the teeth, sometimes diurnal as well, occurs at all ages and may be as distressing to the bystander as it is to the patient. It may also cause serious dental problems unless the teeth are protected in some way. There are many hypothetical explanations, all without proof. Stress is most often blamed, and claimants point to EMG studies that show the masseter and temporalis muscles to be excessively contracted. When present in the daytime, it may also represent a fragment of segmental dystonia or tardive dyskinesia.
Nocturnal Enuresis (See Also Chap. 26)
Nocturnal bedwetting with daytime continence is a frequent disorder during childhood, which may persist into adult life. Approximately 1 of 10 children 4 to 14 years of age is affected, boys more frequently than girls (in a ratio of 4:3); even among adults (military recruits), the incidence is 1 to 3 percent. The incidence is much higher if one or both parents were enuretic. Although the condition was formerly thought to be psychogenic, the studies of Gastaut and Broughton revealed a peculiarity of bladder physiology. The intravesicular pressure periodically rises to much higher levels in the enuretic than in normal persons, and the functional bladder capacity of the enuretic is smaller than normal. This suggests a maturational failure of certain modulating nervous influences.
An enuretic episode is most likely to occur 3 to 4 h after sleep onset, and usually, but not necessarily, in stages 3 and 4 sleep. It is preceded by a burst of rhythmic delta waves associated with a general body movement. If the patient is awakened at this point, he does not report any dreams. Imipramine (10 to 75 mg at bedtime) has proved to be an effective agent in reducing the frequency of enuresis. A series of training exercises designed to increase the functional bladder capacity and sphincter tone may also be helpful. Sometimes all that is required is to proscribe fluid intake for several hours prior to sleep and to awaken the patient and have him empty his bladder about 3 h after going to sleep. One interesting patient, an elderly physician with lifelong enuresis, reported that he had finally obtained relief (after all other measures had failed) by using a nasal spray of an analogue of antidiuretic hormone (desmopressin) at bedtime. This has now been adopted for the treatment of intractable cases. Diseases of the urinary tract, diabetes mellitus or diabetes insipidus, epilepsy, sleep apnea syndrome, sickle cell anemia, and spinal cord or cauda equina disease must be excluded as causes of symptomatic enuresis.
Relation of Sleep to Medical Illnesses
The high incidence of thrombotic stroke that is apparent upon awakening, a phenomenon well known to neurologists, has been studied epidemiologically by Palomaki and colleagues. These authors have summarized the evidence for an association between snoring, sleep apnea, and an increased risk for stroke. As already mentioned, cluster headache and migraine have an intricate relationship to sleep, the former almost always occurring during or soon after the first REM period, and the latter often curtailed by a sound sleep.
Patients with coronary arteriosclerosis may show electrocardiogram (ECG) changes during REM sleep, and nocturnal angina has been recorded at this time. Snoring is strongly associated with chronic hypertension. Asthmatics frequently have their attacks at night, but not concomitantly with any specific stage of sleep; they do have a decreased amount of stage N3 sleep and frequent awakenings, however. Patients with hypothyroidism have shown a decrease of stages N3 sleep, and a return to a normal pattern when they become euthyroid. Demented patients generally exhibit reduced amounts of REM and slow-wave sleep, as do children with Down syndrome, phenylketonuria, and other forms of brain damage. Alcohol, barbiturates, and other sedative-hypnotic drugs that suppress REM sleep produce extraordinary excesses of REM during withdrawal periods. This may, in part, account for the hyperactivity and confusion, and perhaps the hallucinosis, seen in withdrawal states.