32 Sleep disorders

Despite major advances in our understanding of its neurobiology, sleep remains an enigma. Its true function and even the amount needed for optimum brain performance remain uncertain (Frank 2006). However, the need to sleep is imperative, reflecting the fact that sleepiness, like hunger and thirst, is a true drive state. Sleepiness can only be satiated by sleep itself. Moreover, severely disordered sleep can profoundly affect cognition, mental health, and physical well-being.

Although sleep medicine has a traditionally low profile in neurology teaching and practice, sleep-related phenomena are frequently associated with numerous neurological disorders. Conversely, sleep problems can adversely affect familiar conditions such as headache and epilepsy. Furthermore, in large surveys, sleep-related symptoms are undoubtedly common with 25 per cent of the population reporting problems that significantly and regularly impact on daily activities.

The traditional view that sleep is largely a passive or necessarily restful process has largely been superseded. By contrast, rather than simply reflecting the absence of wakefulness, sleep is actively orchestrated with a highly reproducible and complex internal architecture (Harris 2005). Sleep is divided into non-rapid eye movement, NREM, stages I to IV, and rapid eye movement, REM, states on the basis of physiological parameters (Fig. 32.1). Light non-rapid eye movement sleep of stages I and II comprises about 55 per cent of total sleep time whereas deep non-rapid eye movement, stages III and IV, sleep fills 25 per cent. The remaining 20 per cent is occupied by the curious state of REM sleep in which vivid dreams predominate.

A typical distribution of healthy adult sleep is shown in Fig. 32.2. Four or five rapid eye movement cycles occur through the night. It should be recognized that occasional arousals from nocturnal sleep may be regarded as normal and that seemingly random body movements or shuffles in position occur regularly throughout the night often at shifts of sleep stage. In rapid eye movement sleep, however, despite high levels of cerebral metabolic activity that loosely correspond to dream mentation, general motor activity is profoundly suppressed and any observable movements are confined to occasional minor jerks.

The precise mechanisms that control shifts between wake and the various sleep stages are poorly characterized although a current theme proposes the existences of multiple sleep switches (Saper et al. 2001). In essence, discrete brainstem centres for wake, non-rapid eye movement, and rapid eye movement sleep interact by mutually inhibitory pathways, creating so-called ‘flip-flop’ switches. With these, transitions between wake and sleep states are made both efficiently and quickly. Switches between non-rapid eye movement and rapid eye movement sleep states may also occur by a similar process (Lu et al. 2006).

Sleep generally deteriorates with age and determining the limits of normal sleep can be difficult. Most classifications of disordered sleep are symptom-based and the recently revised International Classification of Sleep Disorders (ICSD-2) recognizes eight categories (American Academy of Sleep Medicine 2005):

Chronic insomnia is loosely defined as the perception of inadequate sleep for a period of more than 4 weeks. Such inability to fall asleep or maintain continuous sleep is a common symptom and has a number of causes both intrinsic and extrinsic (Table 32.1). It is rare for organic cerebral pathology to underlie primary insomnia and persistently maladaptive attitudes or behaviours are usually responsible. An index event or illness can often be identified. The common forms of primary insomnia are probably best treated by behavioural modification including a combination of cognitive behaviour therapy and relaxation techniques (Morin et al. 1999). The intermittent use of short-acting hypnotics may be helpful although long-term drug treatment is rarely beneficial.

If symptoms of inadequate sleep date back to childhood, the term ‘idiopathic insomnia’ is sometimes used. Although its neurobiology remains obscure, at some level, this disorder probably reflects a constitutionally impaired sleep drive such that the normal homeostatic pressure to sleep is inadequate.

The interplay between psychological distress and chronic insomnia is complex with each element potentially fuelling the other. Psychiatric input to treat any significant mood disorder can therefore be helpful in attempting to resolve sleep-related symptoms.

Significant excessive daytime sleepiness is reported by 5 per cent of the population and is most often due to poor quality or diminished overnight sleep (Table 32.2). It is important to distinguish true sleepiness or drowsiness from fatigue and lethargy which often have different causes. Within the abnormally sleepy population, approximately 2 per cent have a primary sleep disorder in which the most striking complaint is an inability to stay awake appropriately despite the desire to do so.

Narcolepsy is not a rare disorder with an estimated prevalence of 1 per 2000 in Caucasian populations. However, differences in case ascertainment and the availability of sleep services have led to considerable variance in reported rates worldwide (Silber et al. 2002). Moreover, there is undoubtedly a spectrum of severity and many mildly affected individuals are undiagnosed, misdiagnosed, or only correctly diagnosed after many years of symptoms (Kryger et al. 2002). It most often starts in adolescence with a second minor peak in early middle age and symptoms are life-long. Most sufferers develop coping strategies to minimize the impact of the syndrome although this may lead to social isolation.

Narcolepsy is important not least because it is usually disabling, influencing every aspect of daily living (Dodel et al. 2004). Many narcoleptics feel a sense of underachievement partly because treatment is frequently either delayed or only partially effective. A perceived lack of medical interest in the disease together with the adverse effects on schooling, careers, and relationships understandably produces frustration. Secondary mood disorders are seen in many patients.

Rather than reflecting true hypersomnolence over a 24-h period, narcolepsy is best viewed as a primary disorder of sleep–wake regulation (Fig. 32.3) with an inability to stay awake for more than a few hours as the cardinal symptom (Dauvilliers et al. 2007). Indeed, excessive daytime sleepiness not explained by another disorder remains an essential diagnostic criterion. Many subjects describe sudden and irresistible urges to sleep, often in public or inappropriate situations, invariably worse if they are unoccupied or bored. Short naps lasting minutes can often be restorative, however, in contrast to most other sleep disorders causing hypersomnolence. A minority of sufferers are relatively unaffected by excessive daytime sleepiness and other features of the syndrome predominate.

Cataplexy is a curious phenomenon highly specific to narcolepsy and present to varying degrees in two-thirds of patients (Overeem et al. 2001). It is particularly important to identify typical cataplexy since its presence in a subject with excessive daytime sleepiness is considered diagnostic of narcolepsy. Full-blown episodes reflect an intrusion of profound muscle paralysis that descends over a few seconds from head to the lower limbs, often causing collapse to the floor. Identifiable triggers usually have an emotional content. Laughter and other positive emotions such as pleasant surprise are the most common precipitants although frustration and anger can also reliably provoke episodes. In some subjects, the mere thought or anticipation of an emotional event can cause unpredictable collapse. Reassuringly, attacks are rare in dangerous situations and most sufferers only report cataplexy when relatively relaxed in familiar environments with friends. It is therefore very uncommon for physicians to witness episodes, making a reliable history crucial for confident diagnosis. Importantly, partial or focal attacks are common and can be subtle, perhaps confined to the jaw or neck. Occasionally, an inability to tell the punchline of a joke due to a stuttering dysarthria may be the only manifestation. Facial twitching or head bobbing is very common as an episode commences and can lead to diagnostic confusion. Crucially, awareness is preserved in cataplexy although in rare instances when attacks last more than a minute or so, dream-like intrusions and altered consciousness may intercede. Severely affected individuals may have over 20 attacks a day often reporting that the amusement or frustration induced by the cataplectic episodes themselves can prolong the period of weakness.

It is widely thought that cataplexy occurs because rapid eye movement sleep paralysis intrudes into the wakeful state. Indeed, as in rapid eye movement sleep, a subject is rendered temporarily areflexic during an episode as a result of descending inhibitory neural impulses from lower brainstem centres directly onto motor neurons. Some evidence suggests that this phenomenon may occur to a minor degree during emotion in control subjects, adding credibility to the adage ‘going weak with laughter’ (Overeem et al. 1999).

Sleep paralysis and hallucinations around sleep–wake transition are the other two components of the narcoleptic ‘tetrad’ first described over fifty years ago (Yoss and Daly 1957). Only 25 per cent of subjects have all four elements, however, and the presence of these other symptoms, particularly sleep paralysis, is not specific to narcolepsy. Sleep paralysis is usually frightening primarily because of an inability to take deep breaths voluntarily. Most episodes occur at the point of waking although narcoleptics typically also report episodes at sleep onset. Accompanying sensations of being crushed with or without vivid visual hallucinations may add to the distress of the episodes. Like cataplexy, this phenomenon reflects the intrusion of rapid eye movement sleep elements into the wakeful or drowsy state.

Hallucinations occurring at sleep onset, hypnagogic, or as the subject wakes, hypnopompic, are usually visual and can be both vivid and disturbing, especially in children. They represent fragments of dream mentation intruding into the conscious state, reinforcing the notion that narcoleptics cannot adequately maintain consistent boundaries between states of wakefulness and sleep.

When questioned, the majority of narcoleptics have fragmented nocturnal sleep. Although this may be due to the intrusion of a parasomnia or obstructed breathing, both of which are commoner in narcoleptics, the primary problem is one of sleep regulation and maintenance. The notion that narcoleptics have problems sleeping at night is counterintuitive to some but is an important addition to the original descriptions of the syndrome.

Other symptoms are commonly reported in narcolepsy. In addition to obvious naps, most narcoleptics will experience numerous ‘micro-sleeps’ through the day in which awareness during routine activities is compromised for a few seconds. The resulting lapses lead to automatic and inappropriate behaviours with worrying consequences for complex and potentially dangerous tasks such as driving. Although difficult to characterize, many narcoleptics also report significant problems with memory and concentration as a result of their sleep–wake difficulties. Furthermore, increasing evidence suggests abnormalities of appetite, particularly at night, with cravings for sweet foods. Moderate obesity is commonly seen in childhood narcolepsy and may have a metabolic explanation since there is no clear link with excessive food intake (Kotagal et al. 2004). Indeed, some evidence suggests that overweight narcoleptics eat less than normal controls.

Since the discovery in 1984 that Japanese narcoleptics were extremely likely to carry the human leucocyte antigen, HLA, haplotype DR2 (Juji et al. 1984), an autoimmune basis for the syndrome has been thought likely. The predisposing antigen has since been established as DQB1*0602. This is present in over 90 per cent of narcoleptics with cataplexy and around 50 per cent of those without cataplexy, compared to a frequency of 20 per cent in control populations. Of interest, homozygosity for DQB1*0602 appears to confer an even greater risk for the syndrome. However, there remains no direct evidence for autoimmunity either in the form of serum markers or CSF abnormalities.

A major breakthrough in understanding the neurobiology of narcolepsy occurred in 1999 when two groups independently demonstrated abnormalities of a recently described neuropeptide, hypocretin, also called orexin, in separate animal models. The well-established autosomal recessive Doberman model was shown to have dysfunctional hypocretin receptors (Lin et al. 1999) whilst a mouse hypocretin knockout model developed convincing clinical features of narcolepsy with cataplexy (Chemilli et al. 1999). Subsequently, it has been demonstrated that CSF hypocretin is virtually absent both in sporadic canine models and also human narcoleptics with cataplexy (Ripley et al. 2001). Indeed, a CSF hypocretin level of less than 110 pg/ml is now considered diagnostic. Post-mortem evidence has confirmed that pathology in narcoleptic brains is confined specifically to hypocretin neurons (Thannickal et al. 2000). Confusingly, however, in rare familial narcolepsy and in sporadic cases without typical cataplexy, hypocretin levels can be preserved, implying there is more than one pathogenetic mechanism for certain forms of the syndrome.

Following the unexpected involvement of the hypocretin system in human narcolepsy, it has been intensely studied in intact animals. Around 30 000 neurons containing the peptide are confined to the lateral hypothalamus but innervate all the arousal systems in the brain. Levels of hypocretin rise towards the end of the waking day, especially in the presence of peptide hunger signals or if the subject is expecting food (Saper 2006). Activity of hypocretin neurons, therefore, appears to stabilize a state of wakefulness when the organism needs to be alert. In narcolepsy, their absence leads to inappropriate switches between sleep and wakefulness. Moreover, transitions between behavioural states may be incomplete, explaining the intrusion of rapid eye movement sleep phenomena such as paralysis into wakefulness. The precise mechanism by which emotional stimuli, in particular, trigger cataplexy, however, remains elusive.

If typical cataplexy is absent and CSF hypocretin levels cannot easily be measured, a positive diagnosis of narcolepsy can be made following a multiple sleep latency test. This test measures the propensity for a subject to fall sleep by recording the average length of time to reach light sleep in a conducive environment over four or five nap opportunities between 9 am and 3 pm. If the mean sleep latency is less than 8 min and rapid eye movement sleep is achieved within 15 min on at least two occasions, the criteria for narcolepsy are fulfilled. Reliable results depend on ensuring a reasonable night’s sleep preceding the investigation. The multiple sleep latency test also requires a strict protocol to avoid false negative results.

Narcoleptic symptoms including cataplexy have been reported in a number of neurological conditions (Nishino and Kanbayashi 2005). Given the recent advances in the understanding the neurobiology of the primary syndrome, it is not surprising that pathology in the region of the hypothalamus such as tumours around the third ventricle can lead to secondary narcolepsy, presumably by depletion of hypocretin-containing neurons. However, the mechanism of severe sleepiness or sleep–wake dysregulation after head injury or as components of other conditions such as multiple sclerosis and Parkinson’s disease can be difficult to explain.

The various subtypes of narcolepsy are shown in Table 32.3.

Advice on lifestyle helps a proportion of narcoleptics. Planned naps, especially after meals, may improve wakefulness. Furthermore, the avoidance of large meals rich in refined carbohydrates is reportedly beneficial to some.

The majority of narcoleptics, however, benefit from medication to improve daytime wakefulness (Table 32.4) although few are normalized (Mignot and Nishino 2005). Modafinil is the most widely used wake-promoting agent that has partly replaced traditional psychostimulants. Its mechanism of action remains obscure although a direct effect on arousal centres in the hypothalamus is postulated. It has no definite positive effect on cataplexy. Side-effects are rare and include headache or gastrointestinal upset. Interactions with the oral contraceptive pill and uncertainty over safety in pregnancy may limit its use in young women.

In severe sleepiness or if modafinil is unsuccessful, central stimulants with a predominantly dopaminergic action such as dexamfetamine are helpful, especially if used flexibly. Despite prescriber concerns, it is rare for psychological addiction to occur in narcolepsy although tolerance may require increasing doses with time. At a practical level, cardiovascular side-effects such as hypertension are relatively rare but necessitate caution in the elderly. Given the different mechanisms of action, a combination of modafinil and a psychostimulant is appropriate. Additional use of caffeine and setting aside time for planned naps may reduce the need for medication.

About a half of narcoleptic subjects also require specific treatment for cataplexy. Although the evidence base is small, most anti-depressants will suppress cataplexy by increasing cerebral monoaminergic activity and inhibiting the tendency to enter rapid eye movement sleep. The side-effect profile of most anti-depressant drugs, particularly the tricyclics, may limit their usefulness in cataplexy. A new approach for troublesome cataplexy is to use sodium oxybate although emerging trial evidence suggests this drug helps daytime sleepiness as well. It is a liquid preparation taken before bedtime and, due to its short half-life, once during the night, if the subject is awake. After several weeks of therapy, the effects on cataplexy are striking with almost 90 per cent of attacks abolished (Xyrem International Study Group 2005). Inadvertent daytime naps, objective and subjective measures of daytime sleepiness also improve. The agent appears to work, in part, by inducing deep restorative sleep early in the night such that the sleep drive is effectively dissipated by the following morning. The drug should be used with extreme caution in any patient living alone or with young children in case confusional episodes from deep sleep are provoked. However, if disturbed nocturnal sleep is a major symptom, it appears a logical treatment given that standard benzodiazepine hypnotic agents rarely induce refreshing sleep in narcolepsy.

Following the recent findings that most narcoleptics are deficient of the neuropeptide, hypocretin, a future goal will be to develop replacement therapy. Indeed, if hypocretin levels are increased in animal models by intracerebral infusions, there appear to be clinical effects. In humans, the development of an oral analogue that will penetrate the blood–brain barrier is a current pharmacological goal.

Idiopathic hypersomnia is a diagnosis of exclusion most often made when excessively sleepy patients do not fulfil the criteria for narcolepsy. Depending on precise definitions, it is probably 10 times less common than narcolepsy. Classical cases report difficulty waking in the morning followed by prolonged unrefreshing daytime naps despite long and deep nocturnal sleep. Low mood and frequent automatic behaviours are commonly reported (Bassetti and Aldrich 1997). However, no specific narcoleptic features such as cataplexy are present and CSF hypocretin levels are generally normal. Sleep investigations should confirm a shortened daytime sleep latency of less than 8 min preceded by normal, yet prolonged, nocturnal sleep. A new category of idiopathic hypersomnia without prolonged overnight sleep has been proposed although this is controversial and distinction from atypical or monosymptomatic narcolepsy can be difficult.

As in narcolepsy, although usually with less satisfactory results, the treatment of idiopathic hypersomnia consists of modafinil alone or in combination with traditional psychostimulants.

Kleine–Levin syndrome is a rare and poorly characterized sleep disorder most commonly seen in adolescents (Arnulf et al. 2005a). The primary feature is periodic hypersomnia lasting days to weeks, recurring at intervals of weeks to months. During symptomatic periods, the subject is generally drowsy and usually displays abnormal behaviours. These include simple irritability, hallucinations, hypersexuality, and abnormal appetite producing hyperphagia. Investigations are generally unhelpful although an excess of rapid eye movement sleep is occasionally seen during episodes. Intermittent hypothalamic dysfunction is a plausible, if speculative, mechanism to explain the symptom complex. Secondary causes are very rare and reportedly include a wide variety of neurological conditions such as multiple sclerosis and Prader–Willi syndrome.

Treatments are empirical and usually ineffective although the syndrome tends to resolve spontaneously after several years. Amphetamine may help during episodes and lithium may be used as a prophylactic agent.

A nocturnal apnoea is defined as a cessation of breathing for 10 s or more and is most often secondary to an obstruction in the soft palate. Obstructive sleep apnoea is the commonest cause of severe daytime sleepiness and usually occurs on the background of severe snoring. Central adiposity in males is the commonest risk factor although enlarged tonsils and a receding jaw can also predispose to the problem (Stradling and Crosby 1991). In deep non-rapid eye movement and rapid eye movement sleep, either the hypoxia generated by the breathing difficulty or the effort made by respiratory control centres to overcome the obstruction will partially arouse the subject, severely disrupting sleep.

Apart from causing daytime sleepiness, it is thought obstructive sleep apnoea has significant consequences for general health and may be an independent risk factor for vascular disease, diabetes, and hypertension (Shamsuzzaman 2003). Although neurologists will not usually be involved in treating obstructive sleep apnoea, it is important to recognize the syndrome in any patient and to refer to a sleep service for assessment. The best treatment is usually continuous positive airways pressure, or CPAP, delivered by a nasal mask. Mandibular advancement devices, palatal surgery, wake-promoting drugs, and even tracheostomy may be considered in individual cases.

Central sleep apnoea may occur as an isolated phenomenon secondary to brainstem pathology or as part of severe obstructive sleep apnoea, especially in morbidly obese subjects. It is probably under-recognized in conditions with autonomic dysfunction such as multiple system atrophy. It is a complex problem to treat especially since drug therapies rarely help to increase respiratory drive.

Parasomnias can simply be defined as intermittent undesirable events arising from sleep that are not epileptic in nature. The spectrum is large, ranging from visual imagery at sleep onset to complex motor behaviours, occasionally with violent components. Family members and bed partners are usually more concerned than the subjects themselves who often remain oblivious to any nocturnal disturbance. Parasomnias are generally classified according to the sleep stage from which they arise although some parasomnias are not ‘state dependent’.

A simple yet valid scheme to explain most parasomnias is shown in Fig. 32.4. The brain normally exists in three distinct and mutually exclusive states, namely wakefulness, non-rapid eye movement sleep, and rapid eye movement sleep. Normally, switches occur seamlessly and relatively quickly between these states through the sleep–wake cycle. The majority of parasomnias result from abnormal state transitions such that elements of one state intrude into the boundary of another. A subject can be considered ‘caught’ for a variable period of time in a separate abnormal state somewhere in between wake and sleep.

With the exception of certain rapid eye movement sleep disorders, the neuroanatomical basis of parasomnias remains entirely obscure. The high prevalence of familial aggregation suggests genetic factors and predominance in childhood implies a maturational component, particularly in non-rapid eye movement parasomnias.

It is almost a universal experience to have occasional unpleasant sensations of falling through space at the point of sleep onset with resulting brief muscular contractions. In some subjects these hypnic jerks can regularly interfere with sleep onset and recur through the night. In others there are accompanying explosive sensory phenomena, sometimes with severe head pain as a component. Treatments with short-acting hypnotic agents may be justified in severe cases.

More complex and prolonged phenomena comprising a variety of rhythmical movements also tend to occur during extreme drowsiness just before sleep. Head banging is the commonest manifestation in children. The problem tends to resolve with time although can persist into adulthood and disturb bed partners. Various patterns of movement are seen with the head, neck, and trunk most commonly involved. The view that the movements are semi-voluntary, as part of a sleep-inducing habit, does not concord with the observation that the phenomenon arises only from deep or even rapid eye movement sleep in a minority of subjects (Stepanova et al. 2005).

Non-rapid eye movement parasomnias are characterized by sudden but partial arousals from deep sleep, usually stage 4, resulting in behaviours for which the subject usually has no subsequent recollection. Based on clinical features, sleep-walking, confusional arousals, and night terrors are recognized as three separate phenomena all due to abnormal arousal from deep sleep. Within this notional spectrum, however, there may be considerable overlap and the type of episodes may change with age.

In sleepwalking, the subject will typically leave the bedroom and may well engage in complex behaviours such as cooking and eating. Communication is possible at a basic level although it is usually clear to observers that the subject is not fully alert or responsive. Concerns often arise when there are attempts to leave the house or if there are any violent elements to the episodes.

Confusional arousals refer to brief episodes of disorientation in which the subject may sit up in bed and survey the environment before returning to sleep.

Night terrors are dramatic episodes often lasting several minutes in which the subject suddenly arouses from sleep typically with a loud scream and extreme agitation. Motor and autonomic indications of extreme fear are usually alarming to parents and observers.

All these arousal disorders tend to occur within an hour of sleep when non-rapid eye movement sleep is at its deepest. It is rare for events to recur through the night. If there is any recall, it is usually vague and related to a non-specific fear or urge to leave the bedroom in the case of night terrors. Particularly deep sleep following a period of deprivation or induced by drugs including alcohol may increase the likelihood of events. General stress, changes in schedule, or sleeping in a new environment are further recognized precipitants.

Non-rapid eye movement parasomnias are common in the first decade of life, affecting at least 6 per cent of children on a regular basis (Mason and Pack 2007). Persistence into adulthood occurs in around 15 per cent of these. A confident distinction between nocturnal epilepsy and parasomnias can usually be made from clinical features alone although investigations and video analysis may be required in certain cases. Particularly in adults, overnight investigations may reveal an additional sleep disorder such as sleep apnoea or periodic leg movements that may partially arouse the subject and help trigger a parasomnia.

It is rarely necessary to treat non-rapid eye movement parasomnias with medication, especially in children. However, if disturbances are frequent or likely to cause danger, short courses of benzodiazepines such as low-dose clonazepam before bed are usually helpful. In the absence of any substantial evidence, antidepressants such as paroxetine are also used to good effect, presumably by effects on sleep architecture.

Rapid eye movement parasomnias include nightmares, rapid eye movement sleep paralysis, and rapid eye movement sleep behaviour disorder. Given the propensity for rapid eye movement sleep to occur late in the night, these parasomnias are typically reported between 3 am and 6 am., in contrast to the earlier occurrences of arousal disorders from non-rapid eye movement sleep.

Nightmares represent arousals from unpleasant dreams and are universal experiences. However, up to 4 per cent of adults have frequent or intrusive nightmares often in the context of psychological stress or substance abuse. Nightmares with recurring themes are a hallmark of post-traumatic stress disorder. Certain drugs such as beta-blockers can trigger nightmares as may the sudden withdrawal of anti-depressant agents that normally suppress rapid eye movement sleep.

Symptoms of sleep paralysis seen in around 40 per cent of narcoleptics can also occur as an isolated phenomenon, occasional with a familial pattern. As in narcolepsy, the profound paralysis is usually disturbing. Typically, prolonged episodes can be aborted by a tactile stimulus from a bed partner. If treatment is thought necessary, tricyclic antidepressants are usually helpful

An increasingly recognized rapid eye movement parasomnia occurs when abnormal motor activity intrudes into rapid eye movement sleep, reflecting a fault in the normal mechanisms that render dreaming subjects completely atonic. So-called rapid eye movement sleep behaviour disorder was first formally described in 1986 (Schenck et al. 1986). It is predominantly an affliction of middle-aged or elderly males and has an intimate relation to several neurodegenerative diseases, particularly Parkinsonism. Over 65 per cent of subjects free of any movement disorder during wakefulness at the onset of symptoms will develop Parkinson’s disease within 10 years of follow-up (Schenk et al. 2003). The nocturnal episodes are brief and generally explosive, usually involving the upper limbs. There is often an apparently aggressive intent but injuries to bed partners are incidental and violence is rarely directed. In mild cases, episodes are confined to vocalization or swearing with little observable movement. If awoken during an event, dream recall is the norm although most remain oblivious to their behaviours if their sleep remains continuous. Intriguingly, pleasant dreams or those with a sexual content are very rare whereas reports of being chased by aggressors or attacked by animals are typical themes (Olson et al. 2000).

It has been proposed that when rapid eye movement sleep behaviour disorder is part of an established neurodegenerative syndrome it may indicate underlying synuclein pathology given its predominance in Parkinsonian syndromes. However, if overnight studies are performed, rapid eye movement sleep without atonia and even full-blown rapid eye movement sleep behaviour disorder are also common in other pathologies including tauopathies such as progressive supranuclear palsy (Arnulf et al. 2005b).

It is often necessary to treat rapid eye movement sleep behaviour disorder on a long-term basis to prevent injury either to the sufferer or the bed partner. Clonazepam in a dose range 0.25–2 mg is usually effective and melatonin has been used as a second-line agent. If there are suspicions of an additional breathing-related sleep disorder, overnight investigations are warranted as clonazepam may worsen obstructive sleep apnoea, for example.

There exist a number of rare parasomnias that can arise from any sleep stage. Nocturnal catathrenia is characterized by high-pitched monotonous groans that occur on expiration after deep inspirations. The events last around 10 s and invariably disturb bed partners, recurring in clusters through the night, even in rapid eye movement sleep (Vetrugno et al. 2001). No consistently effective treatments have been reported.

Bruxism refers to teeth grinding and is common. In sleep, it may affect up to 8 per cent of the population to varying degrees and be an important cause of sleep disruption especially if associated with another sleep disorder (Lavigne and Montplaisir 1994). It occurs at any age and in any sleep stage. In neurological patients, it may trigger facial pain and precipitate migraine attacks or other chronic headache syndromes. If not suspected clinically, it is often picked up as 1-Hz interferences on overnight EEG recordings. Dental occlusal appliances are usually helpful.

Nocturnal enuresis may be considered as a parasomnia in some. It can occur during any sleep stage and is commoner in children with other parasomnias. It can be difficult to treat although behavioural techniques including scheduled awakenings are usually effective. Drugs such as tricyclics or desmopressin may be needed in resistant cases.

Periodic leg movements of sleep are characterized by stereotyped leg movements occurring in clusters every 30 s or so throughout sleep, especially in the light non-rapid eye movement stages. The movements themselves tend to be fairly slow, evolving over 1–5 s and typically involving both legs although one or the other may predominate. An episode tends to start with great toe extension and spreads to include ankle dorsiflexion, followed by knee and hip flexion in severe cases. It is relatively rare for subjects to be aware of the leg movements although bed partners may complain. The phenomenon increases dramatically with age and is strongly associated with restless legs syndrome (Section 40.11.1).

If periodic leg movements of sleep are demonstrated after overnight investigation, it can be difficult to gauge their clinical significance, especially if there are no associated EEG arousals (Silber 2001). Further complications may arise if there are other reasons for fragmented sleep such as obstructive sleep apnoea, in which case periodic leg movements of sleep may be triggered as a secondary epiphenomenon.

Treatments for restless legs syndrome also ameliorate periodic leg movements of sleep. Dopamine agonists are usually effective although it is difficult to predict in advance whether any response will be clinically meaningful.

If both quality and quantity of sleep are normal over 24 h but a subject is unable to sleep or stay awake at the desired or expected time, a circadian rhythm disorder may be diagnosed. Most commonly this problem has a clear extrinsic cause such as shift work or jet travel but in some situations there is almost certainly dysfunction of the internal clock mechanism. Behavioural or motivational factors may contribute to the generation of highly irregular sleep–wake patterns especially in younger subjects.

In mammals, the primary biological clock is sited in an area of the hypothalamus called the suprachiasmatic nucleus. The mechanism of the clock at a sub-cellular level has been extensively researched and appears very similar across all animal species, including humans (Turek 2004). In strict isolation with no external cues, the periodicity of the human clock is around 24.3 h. In real life this rhythm is entrained precisely to 24 h primarily by light cues acting on retinal cells that contain a newly discovered retinal pigment, melanopsin (Hattar et al. 2003). A retinal tract to the hypothalamus allows this information to influence the clock mechanism. Subjects blind from birth frequently report difficulty in adapting to a conventional sleep–wake cycle because their internal clocks run a little ‘slower’ than average without light entrainment. Very rarely, sighted individuals also suffer from a similar non-24 h sleep–wake disorder although the precise mechanism remains obscure.

Subjects diagnosed with delayed sleep phase syndrome can be considered as extreme ‘night owls’ such that they are simply unable to sleep before 2 am or later (Lu and Zee 2006). The main concern is usually the subsequent inability to wake effectively for school or work. It is important to exclude significant mood disorder as a driver for the abnormal cycle. Similarly, delayed sleep phase syndrome would not be diagnosed in those who simply prefer the solitude of night and avoid daytime interactions. Sleep diaries and wrist actigraphy can help confirm the diagnosis which mostly affects adolescents with a prevalence estimated at 1 per cent. Sufferers and their families are very commonly frustrated by this sleep disorder and the relative lack of its recognition.

Treatment is difficult and starts with a strict schedule and general sleep hygiene measures. Melatonin taken around 2 h before desired sleep onset time may help with sleep onset although long-term use of hypnotics is usually unsuccessful. Phototherapy from a light box on waking may also help ‘reset’ the internal clock.

This is an extremely rare disorder but of interest because a familial form has been identified and the relevant gene analysed (Toh et al. 2001). The point mutation occurs in a period gene, hPer2, such that the circadian sleep–wake period is 23.3 h. This results in subjects sleeping and waking at least 4 h earlier than expected. Other indications of disturbed circadian rhythm include melatonin secretion and core temperature.

Humans also generally show ‘phase advance’ with increasing age. Common experience suggests that many elderly subjects will fall asleep in the evening and wake early in the day, especially in institutions where this may be encouraged as part of a convenient regime.

An increasing number of people are employed in jobs requiring shift work in a variety of patterns. Rotating shifts, in particular, do not allow circadian rhythms to adapt and frequently lead to difficulties either in staying awake for employment or sleeping effectively during daylight hours. Of potential concern are the secondary effects of sleep deprivation on cognitive performance in tasks demanding sustained attention or decision making, especially in occupations involving heavy industry or transportation. Most shift workers find it increasingly difficult to adapt their sleep–wake cycle as they age. Moreover, additional sleep problems such as obstructive sleep apnoea may worsen the situation.

If shift work is causing significant symptoms and cannot be avoided, treatment is a challenging area if simple sleep hygiene advice fails to help. Planned naps may be beneficial and shift patterns that rotate by delaying work time rather than advancing it are generally easier to cope with. Regular medication is controversial with concerns over dependency, especially with regard to hypnotic agents. Regular caffeine may be used and wake-promoting drugs such as modafinil have been licensed in severe shift work sleep disorder. However, the concept of shift work sleep disorder as a problem requiring drug treatment lies uncomfortably with many physicians.

In assessing a patient with a sleep disorder, the importance of a detailed history from the subject and ideally a bed partner or close family member cannot be over-emphasized. Together with a sleep diary, when appropriate, the majority of diagnoses can be made with moderate confidence on history alone. With important exceptions such as sleep apnoea, where quantification of the problem is important, it is relatively rare for investigations to add useful diagnostic information. Clearly, if a reliable history is not available in the case of a subject sleeping alone, for example, tests can be invaluable. The availability of sleep facilities varies dramatically throughout the world, often dependent on how the tests are financed. The following section is based on a United Kingdom perspective where sleep medicine is relatively under-resourced.

Where insomnia is an isolated symptom, overnight tests are rarely useful. However, in subjects complaining of extremely reduced overnight sleep, surrogate monitoring of sleep using wrist actigraphy may be useful in demonstrating paradoxical insomnia in which there is a misperception of the amount of sleep obtained.

An algorithm for assessing chronic insomnia is shown in Fig. 32.5. Chronic insomnia also associated with daytime sleepiness and frequent naps is likely to have a secondary identifiable cause.

If excessive daytime sleepiness is the primary complaint, it is normally possible to identify an underlying cause, even if the answer is simply insufficient overnight sleep. Care should be taken in establishing that sleepiness per se is the symptom of concern and not lethargy or fatigue which are more likely to have psychological or motivational substrates.

An algorithm for assessing a sleepy subject is shown in Fig. 32.6.

Non-rapid eye movement parasomnias are difficult to investigate and reliance on a good history is central to a confident diagnosis. Capturing an event on overnight recording is rare and investigations on asymptomatic nights are usually unremarkable. Particularly in adults, an additional sleep disorder may sometimes be precipitating a parasomnia. If so, it is advisable to perform overnight investigations to detect arousals secondary to apnoeas or leg movements, for example. Differentiating non-rapid eye movement parasomnias from nocturnal epilepsy can be difficult and video analysis, ideally of several episodes, can be crucial for diagnosis. The provision of video recorders to patients’ families in order to capture events at home may be more productive and cost-effective than formal overnight recording in a hospital setting. Table 32.5 outlines some differentiating features between non-rapid eye movement parasomnias and nocturnal epilepsy.