Compounds whose parent structure is uric acid. These compounds depress central nervous system activity. Long-acting barbiturates such as pentobarbital have been used to treat epilepsy. Barbital was used during the early 20th century to facilitate sleep in individuals with insomnia.

Benzodiazepines

Compounds whose parent structure is a fusion of a diazepine ring with a benzene ring. Benzodiazepines enhance activity of the inhibitory neurotransmitter γ-aminobutyric acid (GABA). Benzodiazepines, which reduce anxiety and promote muscle relaxation, also promote sleep. The earliest benzodiazepines were chlordiazepoxide (Librium) and diazepam (Valium). Benzodiazepines for insomnia are now being replaced by nonbenzodiazepines such as zolpidem (Ambien), eszopiclone (Lunesta), and others.

Circadian rhythms “Circa” is Latin for “about,” and “diem

” is Latin for “day.” Circadian rhythms refer to the approximately 24-hour cycle of biochemical, physiologic, and behavioral processes.

Electroencephalography (EEG)

Measurement and recording of the gross electrical activity of the brain. During EEG recordings, electrodes are typically placed across multiple scalp regions. The electrodes are connected to amplifiers and filters that detect, magnify, and record the electrical activity of the brain.

Hypersomnia

Presence of excessive sleepiness. Daytime sleepiness is so great that it leads to inappropriate daytime napping or sleep. Excessive sleepiness is not alleviated by prolonged sleep times or by napping.

Hypnotic

Class of drugs used to induce sleep.

Parasomnia

Group of sleep disorders manifested by undesirable motor, sensory, or behavioral phenomena that occur during sleep. The International Classification of Sleep Disorders, Revised (ICSD-R) lists 24 parasomnias. More commonly encountered parasomnias include confusional arousals, sleep terrors, and sleepwalking.

Polysomnography

Measurement and recording of EEG activity, typically coupled with measurement and recording of cardiorespiratory activity and eye movements, during sleep.

The origins of sleep and the meaning of dreams have fascinated people for centuries—from philosophers to poets, ideas of the significance of sleep abound. Edgar Allen Poe described sleep as “little slices of death,” whereas William Shakespeare regarded it to be the “chief nourisher in life’s feast.” Until more recently, sleep was considered to be a passive, dormant counterpart to waking life. We now know that sleep is an active process that may look similar to, but is very different from, either anesthesia or coma.

History of treatment of sleep disorders

Humans spend about a third of their life sleeping, and sleep disorders, which affect a large proportion of the general population and occur in all age groups, represent a major public health and global economic burden.¹ It is estimated that 50 million to 70 million adults in the United States have a chronic sleep disorder that interferes with their daily functioning and adversely affects their health and quality of life. Until the mid-1960s, the field of sleep medicine focused primarily on describing and treating insomnia, parasomnias (e.g., sleep walking, night terrors) and hypersomnias, such as narcolepsy. Patients experiencing these symptoms typically sought consultations from neurologists, psychiatrists, psychologists, or their family physician. Until more recently, treatments were generally based on empiric pharmaceutical intervention and behavioral modification protocols or psychotherapy or both.

The role of the respiratory therapist in sleep medicine is still emerging, and areas of necessary proficiency and expertise are yet to be fully defined. An authoritative knowledge about sleep, sleep disorders, pharmacology, and the therapeutic actions of drugs on sleep architecture is required, together with knowledge of associated side effects or adverse drug reactions and toxicities. In the future, respiratory therapists may be involved with the attending physician in determining appropriate pharmacotherapy for sleep-related disorders, such as narcolepsy or periodic limb movement disorder (PLMD). At a minimum, it is highly probable that during the course of performing sleep diagnostic procedures, the respiratory therapist will assess the patient’s current therapeutic regimen and, after consultation with the physician, determine whether medications should be temporarily suspended before diagnostic procedures. Chapter 23 provides basic knowledge about sleep medicine for both the respiratory therapist who is still developing clinical skills and the experienced therapist who is actively participating in patient care with physicians.

! KEY POINT

The International Classification of Sleep Disorders, Revised (ICSD-R), published in 2005 by the American Academy of Sleep Medicine (AASM), classifies sleep disorders and diagnostic criteria. Diagnostic codes for each disorder are also provided. Currently, more than 80 sleep disorders are described in the ICSD-R.

It is beyond the scope of this chapter to address the entire spectrum of sleep disorders classified in the International Classification of Sleep Disorders, Revised (ICSD-R) or to provide an exhaustive summary of all the pharmacologic, neutraceutical, or cognitive behavioral therapies employed in this field. This chapter describes the classes of drugs that are likely to be encountered when treating patients with sleep disorders. First, a broad but brief overview of the history and evolution of sleep research and sleep pharmacology is presented followed by a brief overview of the brain mechanisms underlying the processes of wakefulness and sleep, including circadian aspects. Some key sleep-related disorders are described, and typical compounds that may be used to treat those disorders are reviewed. Adult sleep apnea syndrome, which is primarily treated with mechanical devices such as oral appliances or application of continuous positive airway pressure (CPAP), are not discussed. Apnea and bradycardia of prematurity, which are treated primarily with respiratory stimulants, are discussed in Chapter 8.

The onset and duration of sleep are orchestrated through multiple brain structures and neurotransmitter substrates, and pathology within those structures or neurotransmitter systems become manifest as a sleep disorder. For example, loss of orexin/hypocretin-producing neurons in the lateral hypothalamus is associated with the inability to maintain prolonged periods of wakefulness or sleep and the intrusion of rapid eye movement (REM) sleep, or the signs of REM sleep, into wakefulness. Sleep pharmacotherapeutics evolved with the understanding that normalization of activity within perturbed brain regions or neurotransmitter systems, or both, leads to a reduction of the signs and symptoms of specific sleep disorders.

! KEY POINT

Compounds to induce and sustain sleep have been in use for almost 5000 years. The oldest may perhaps be opium (which contains morphine), whereas the most recent are nonbenzodiazepines.

As a result of increased public awareness about sleep disorders and scientific advances, sales of sleep-related drugs have increased markedly. Sales of medications to treat insomnia in the United States increased from $1.3 billion in 2001 to approximately $4.6 billion in 2006. However, the interest in pharmaceutical sleep aids is not new; sleep-inducing compounds were discovered and used thousands of years ago. Perhaps the first compound to be used as a sleep-inducing aid was the juice from the opium poppy. Some of the earliest written descriptions of the opium poppy have been found on Sumerian clay tablets dating to approximately 3000 bc. At that time, the juice of the poppy was harvested and consumed because of its ability to induce a euphoric state; this led to the plant being considered a Gil Hul, or “joy plant.” Descriptions of the opium poppy have also appeared in writings of the Assyrians and Persians. The Greeks eventually were introduced to the opium poppy, which may represent the first time it was used expressly for sleep induction. Greek mythology depicts many sleep-related deities, including Hypnos (sleep), Morpheus (dreams), Nyx (night), and Thanatos (death, the twin brother of Hypnos) in association with opium extracted from the poppy. Homer described the properties of opium in both The Iliad and The Odyssey as an intoxicating, pain-relieving, and sleep-inducing substance.

More recent literature also expounds the sleep-inducing power of opium in the Wizard of Oz with Dorothy, her dog Toto, and the Cowardly Lion falling into a deep sleep as they passed through a field of poppies on their approach to the Emerald City. The sleep-inducing effects of the opium poppy can be attributed to numerous alkaloids contained within the plant, including morphine and codeine. Both are central nervous system (CNS) depressants and opioid pain relievers.

Development of sleep-inducing compounds was revolutionized in the early 19th century by the synthesis of opium. Shortly after this, chloral hydrate and the bromides were developed. Chloral hydrate, a CNS depressant, rapidly induces deep sleep. Bromides, invented in the mid-19th century, are also CNS depressants and induce sleep relatively quickly. Their popularity as sleep aids increased through the late 19th century and into the early 20th century. Also during the 19th century, nitrous oxide was rediscovered, and ether and nitrous oxide were inhaled as “party favorites” of upper-class Europeans and Americans. The initial discoveries of ether by the Spanish alchemist Lullius in 1275 and nitrous oxide by the English chemist Priestly in 1772 were lost to medical science until their reintroduction in 1842. At that time, Long, a surgeon in Georgia, employed the recreational drug ether in surgical procedures because of its incredibly rapid induction of “sleep, amnesia and pain relief.” In doing so, he unknowingly ushered in the modern era of anesthesia.

Barbiturates, first discovered in the mid-19th century, soon replaced bromides as the “sleeping pills” of choice in the early 20th century. This class of drugs comprises more than 25,000 compounds that were synthesized by combining various compounds with barbituric acid. Although multiple barbituric acid compounds were developed, only a select few (including a diethyl derivative) resulted in sleep-promoting properties. Barbiturates, such as phenobarbital, are very effective at inducing sleep. However, they also have multiple side effects, not the least of which is the potential risk for barbiturate addiction, and, if taken with alcohol, they can result in respiratory suppression and death. These hypnotic agents have now been replaced by newer, more effective, and safer compounds.

Benzodiazepines such as diazepam (Valium), temazepam (Restoril), and clonazepam (Klonopin) were first marketed in the 1970s. Early formulations of these CNS depressants shared similar side-effect profiles to the barbiturates, although their margin of safety was much greater. However, benzodiazepines possess the potential for addiction, and because of the long half-life and the duration needed to eliminate some benzodiazepines from the body (more than 12 to 24 hours), next-day “hangover” sleepiness effects and memory impairments are common. Since the introduction in the 1990s of nonbenzodiazepine and analogues of these compounds (e.g., zopiclone, zolpidem, zaleplon, and eszopiclone), the use of benzodiazepines for insomnia has declined.

The “ideal hypnotic” should possess the following principal characteristics. It should induce sleep rapidly (in about 10 to 15 minutes); maintain sleep over prolonged periods (about 7 to 8 hours); and be devoid of daytime residual side effects on memory, cognition, or alertness. It should also possess additional characteristics such as rapid absorption; optimal half-life; receptor-specific binding; no active metabolite; no potential for abuse, tolerance, or dependence; no respiratory depressive effect; and no interaction with alcohol or other CNS depressants.

In addition to “sleeping pills,” many nondepressant medications are used in sleep medicine. In disorders such as restless legs syndrome (RLS) and PLMD, sleep onset is often delayed and fragmented. First-line therapy for these disorders includes dopamine agonists and, in some cases, opiates. Respiratory stimulants are another class of medications sometimes employed in the treatment of sleep disorders. Medroxyprogesterone and acetazolamide have been used in an effort to enhance ventilation in patients with high altitude–induced central sleep apnea or obesity-hypoventilation syndrome. More recently, the antidepressant mirtazapine has been shown to reduce apnea severity in animal models emulating sleep apnea as well as in some patients.² An effective pharmaceutical treatment for obstructive sleep apnea does not yet exist.

! KEY POINT

Manifestations of sleep disorders can emerge from within different sleep states. REM behavior disorder occurs only during REM sleep. Confusional arousals tend to occur when awakening from slow wave sleep (stage N3). Somnambulism (sleep walking) is also known to occur during the first third of the night, the sleep period predominated by slow wave sleep (stage N3).

Progression of sleep

Sleep was originally thought to be a passive rather than an active process. Canton, a physiologist at the Royal Infirmary in Liverpool, is credited with the discovery of electroencephalography (EEG), an important milestone in the understanding of sleep as an active process. His experimental observations on cortical currents in rabbits and monkeys were published in the British Journal of Medicine.³ Using a mirror galvanometer, Canton observed an increase in the amplitude of waves measured from the cortex during states of sleep as opposed to a decrement in cortical amplitudes during wakefulness. Canton was the first to perform sleep EEG in mammals.⁴ Subsequently, several observations including those of von Economo and the experiments by Morruzi and colleagues clearly showed that sleep is not a passive phenomenon but involves several brain regions, especially the diencephalon and the brainstem, which actively control sleep and states of arousal.^5–8 In 1957, Dement and Kleitman reported the cyclical alternating pattern of non–rapid eye movement (NREM) and REM sleep.⁹

When determining the most appropriate pharmaceutical intervention for a sleep disorder, it is important first to consider how sleep is defined and measured and the normative values of time spent in sleep and in each sleep stage (i.e., sleep architecture). Mammalian sleep can be defined as a cyclical, reversible behavioral state of perceptual disengagement from, and unresponsiveness to, the external environment. Within normal human monophasic sleep, sleep is polygraphically characterized into two distinct stages based on a constellation of behavioral and electrophysiologic parameters. These two stages are NREM and REM sleep. NREM sleep is categorized further into stages N1-N3 (formerly known as stages 1 to 4); N1 is the lightest and stage N3 is the deepest sleep stage. Figures 23-1 to 23-4 show examples of EEG-defined wakefulness followed by examples of stages N1 to N3.

Figure 23-1 Wakefulness. (Recreated from Embla, Bloomfield, Colo.)

Figure 23-2 Sleep state N1, formerly known as stage 1. (Recreated from Embla, Bloomfield, Colo.)

Figure 23-3 Sleep stage N2, formerly known as stage 2. (Recreated from Embla, Bloomfield, Colo.)

Figure 23-4 Sleep stage N3, formerly known as stages 3 and 4. (Recreated from Embla, Bloomfield, Colo.)

The term for REM sleep is derived from the periodic bursts of REMs during sleep. REM sleep has both tonic (persistent) and phasic (episodic) components. During tonic REM sleep, the EEG tracing shows a similar pattern to that of N1, but it may also exhibit increased activity in the theta frequency range (3 to 7 Hz) and sawtooth-type waves. REM sleep is also accompanied by a generalized muscle atonia, with the exception of the extraocular muscles and the diaphragm. Figure 23-5 shows an example of EEG-defined REM sleep.

Figure 23-5 REM sleep. (Recreated from Embla, Bloomfield, Colo.)

Aserinsky and Kleitman were the first to observe the electrophysiologic characteristics of REM sleep and, in particular, the rapid, jerky, and binocularly symmetric eye movements in this sleep stage.¹⁰ EEG patterns similar to wakefulness were noted, showing the characteristic fast desynchronized rhythms in the cortical EEG, and the term paradoxical sleep was introduced by Jouvet and Michel in 1960; the term active sleep was used by other researchers. These terms are used interchangeably in the literature, although subtle differences exist. Additionally, autonomic activation occurs during this state as respiratory and heart rates are increased. Dream recall is also common when subjects are awakened during this stage, whereas dream recall during NREM sleep is relatively rare (Table 23-1).

TABLE 23-1

Electroencephalographic Correlates of Sleep Stages

		CHARACTERISTICS
SLEEP STAGES	TST (%)	EEG	EOG	EMG	OTHER VARIABLES
Stage awake (relaxed wakefulness)		Alpha activity (8-12 Hz) or low-amplitude beta (13-35 Hz), mixed-frequency waves	REM (in sync or out of sync deflections), eye blinks	Relatively high tonic EMG activity	Alpha activity in occipital leads compared with central leads, eye opening suppresses alpha activity, movement artifacts
N1, formerly known as stage 1	2-5	Low-voltage, mixed-frequency waves (2-7 Hz range), mainly irregular theta activity, triangular vertex waves	SEMs, waxing and waning of alpha rhythm	Tonic EMG levels typically below range of relaxed wakefulness	Alpha ≤50%, vertex sharp waves in central leads, absence of spindles and K complexes
N2, formerly known as stage 2	45-55	Relatively low-voltage, mixed-frequency waves, some low-amplitude theta and delta activity	No eye movement	Low chin muscle activity	Sleep spindles (7 to 14 Hz) and K-complexes occur intermittently
N3, formerly known as stages 3 and 4	5-20	≥20%-50% of epoch consists of delta (0.5-2 Hz) activity	No eye movement	Chin muscle activity is lower than N1 and N2	Sleep spindles may be present
Stage REM	20-25	EEG is relatively low voltage with mixed frequency resembling N1 sleep	Episodic rapid, jerky, and usually lateral eye movements in clusters	EMG tracing almost always reaches its lowest levels owing to muscle atonia	Phasic and tonic components, presence of sawtooth waves, alpha waves are 1-2 Hz slower than waves occurring during wakefulness and non-REM sleep

EEG, Electroencephalography; EMG, electromyography; EOG, electrooculography; REM, rapid eye movement; SEMs, slow eye movements; TST, total sleep time.

Sleep stages occur in cycles that repeat approximately every 90 to 120 minutes. A normal sleep cycle begins with N1 and proceeds through N3. Sleep rapidly passes through the same stages in reverse order before REM sleep is initiated, usually first occurring about 90 minutes after sleep onset. Although significant interindividual variation in sleep need is noted, adult humans typically sleep about 7 to 9 hours per night and spend almost one-third of their life sleeping. Figure 23-6 provides a graphic representation of the cyclical distribution of sleep states across a single night in a normal healthy adult.

Figure 23-6 Sleep hypnogram displaying distribution of sleep stages across the night.

Neurophysiologic mechanisms

Arousal and wakefulness

! KEY POINT

The neural arousal or activating mechanisms that produce the state of wakefulness reside within multiple brain regions, and each region uses a different neurotransmitter. Activities within brain regions using histamine, norepinephrine, and acetylcholine are correlated in the production and maintenance of wakefulness.

In a series of postmortem examinations of the brains of patients who had died as a result of the outbreak of encephalitis lethargica after World War I, von Economo ⁸ observed that lesions in the rostral midbrain and posterior hypothalamus had a profound effect on sleep and wakefulness. He derived two significant correlates from his observations. The first was that lesions in the preoptic and basal forebrain (BF) areas caused severe insomnia. The second was that lesions in the posterior and lateral hypothalamus (LH) caused severe hypersomnia. On the basis of these data, von Economo hypothesized the existence of a group of sleep-promoting neurons around the hypothalamic optic chiasm and, conversely, a group of wake-promoting neurons in the area of the posterior hypothalamus. Both observations have been proved to be essentially correct, but it was only toward the end of the 20th century that the hypothalamic influences on sleep and wakefulness were integrated into the mechanisms of vigilance state control. Before that, the emphasis had been on brainstem mechanisms and the ascending reticular activating system (ARAS).

Ascending reticular activating system

Physiologic analysis of the mechanisms of EEG arousal and wakefulness began with the classic studies of Bremer, who, in 1935, showed that if the brainstem of a cat was completely transected at the level of the midbrain (i.e., to produce the cerveau isolé), the result was that the cat maintained a persistent state of sleep. Different interpretations of this result were possible until Moruzzi and Magoun ⁶ demonstrated the existence of an active arousal center below the level of the transaction, in the pons (i.e., the brainstem). Subsequent lesion and electrical stimulation studies identified the brainstem core, or pontine reticular system, as a critical component of arousal and wakefulness. This system was termed the ascending reticular activating system and was morphologically defined by cell bodies that projected from the brainstem to innervate the midbrain and cortex.

A more recent advance has been the identification of the importance of a cholinergic activating system in EEG arousal. This is one of the major components of the ARAS, and its identification depended on methods that were developed for labeling neurons that contain specific neurotransmitters. Steriade and colleagues ¹¹ identified cells located near the pons-midbrain junction that increased their discharge rate about 60 seconds before the first change to an aroused state was noted on the EEG. These neurons were found to project to the thalamus, and the change in their discharge rate was the first indication of arousal. Subsequent work identified these neurons as containing the neurotransmitter acetylcholine and being localized to the laterodorsal pontine tegmentum/pedunculopontine tegmentum (LDT/PPT) region.

Cholinergic systems are not the exclusive substrate of EEG arousal, however, and evidence that multiple systems are involved in arousal and wakefulness came from the inability of lesions of any single one of these systems to disrupt EEG arousal on a permanent basis.¹² Other brainstem reticular neuronal projections to the thalamus using glutamate neurotransmission and noradrenergic and serotoninergic projections from the locus caeruleus and raphe nuclei also play important roles in maintaining wakefulness. In addition to brainstem nuclei, a cholinergic input to the cortex that ascends from the BF nuclei, especially the nucleus basalis of Meynert, plays an important role. Histaminergic neurons localized in the tuberomammillary nucleus (TMN) of the posterior hypothalamus also promote wakefulness. Discovery of the orexin/hypocretin system in 1999 (see section on Narcolepsy later) led to another CNS arousal system being identified. It is probably the latter hypothalamic systems that were affected in the brains examined by von Economo.⁸

The current conception of the mechanisms of arousal and wakefulness can be summarized by noting that wakefulness and the concomitant EEG arousal is a state of brain activation resulting from the influence of several excitatory neurotransmitters. The term ARAS has been replaced by ascending activating system (AAS). NREM sleep is the absence of such excitatory drive, and from the onset of sleep through to the deep stages of slow wave sleep (SWS) (N3), NREM sleep is marked by the gradual reduction of this arousing influence. REM sleep then occurs as a different aroused state, but one that is still modulated by some of the same excitatory pathways that are active during wakefulness. Wakefulness is supported by several, apparently redundant parallel neurotransmitter pathways, which include glutamate, acetylcholine (projecting from both the LDT/PPT nuclei in the brainstem and the BF), and the monoamines (i.e., norepinephrine, serotonin, and histamine). With the exception of the hypothalamic TMN and LH projections, which also innervate the cortex, and the cholinergic BF projection, which exclusively innervates the cortex, these ascending projections of the AAS innervate the thalamus.

Thalamic mechanisms of arousal

Most AAS projections that mediate EEG arousal and wakefulness synapse in the thalamus, which is an essential center for the organization of EEG arousal and for maintaining activation at a cortical level. Thalamic mechanisms at a cellular level influence the differences between wakefulness and NREM sleep. A detailed consideration of these thalamic mechanisms is beyond the scope of this chapter; they depend primarily on the cells in the thalamus that project to the cortex (i.e., thalamocortical neurons).¹³ Thalamocortical neurons differ in their rate and pattern of discharge depending on the vigilance state. When the arousal-related glutamatergic, cholinergic, noradrenergic, and serotoninergic projections are active, they drive the thalamocortical neurons to discharge in single spike mode. This discharge keeps the EEG in an active state, which contributes to arousal and wakefulness.

In contrast, in the absence of activating or arousing inputs, thalamocortical cells modify their discharge rate to a burst mode. This bursting drives oscillations in thalamic and cortical loop circuits, and these oscillations are the substrate of the slowing and increasing amplitude of the EEG that characterizes NREM sleep. The gradual and continuing reduction in the arousing input results in the gradual deepening of NREM sleep until delta waves dominate the EEG during SWS.

Sleep onset and processes that maintain sleep

! KEY POINT

The neural mechanisms that produce the state of sleep also reside within several brain regions, and each region uses a different neurotransmitter. Onset of activity within the ventrolateral preoptic nuclei orchestrates the onset and maintenance of NREM sleep.

An important consideration is the mechanism that begins this process of reducing the drive from the activating (wakefulness) systems.¹⁴ In other words, how does sleep begin? Electrophysiologic recordings of cells in the BF and anterior hypothalamic regions showed that some of these neurons discharge only during sleep, and this was hypothesized to be an active sleep-promoting mechanism. Confirmation came from studies by Sherin and colleagues,¹⁵ who used anatomic techniques to detect neurons in the ventrolateral preoptic (VLPO) area that were selectively active during NREM sleep. Subsequent immunohistochemical studies identified the neurotransmitters contained in these cells as inhibitory γ-aminobutyric acid (GABA) and glycine. Anatomic work showed neurons containing GABA and glycine projected not only to wakefulness-promoting histaminergic neurons in the TMN and other hypothalamic centers including the BF, but also to all the brainstem nuclei important in EEG arousal.¹⁶ This group of cells coordinates the inhibition of activity in all components of the AAS to facilitate sleep onset. Current studies continue to investigate the interaction of these neurons with other systems that are important in sleep and in particular how other cells within a region around the VLPO area, named the extended VLPO area, are involved with initiating the onset of REM sleep.

Two-process model of sleep regulation

Homeostatic sleep drive.

Sleep propensity is determined by homeostatic and circadian sleep drives. The homeostatic component of sleep need is the sleepiness that follows prolonged wakefulness, and there is now considerable evidence to support the role of adenosine as a mediator of this component. This role of adenosine seems to depend primarily on its inhibitory action on the BF wakefulness-promoting neurons. Commonsense evidence for a sleep-enhancing effect of adenosine comes from the ubiquitous use of coffee and tea to increase alertness because these beverages contain caffeine, an adenosine receptor antagonist.¹⁷ McCarley and colleagues¹⁸^,¹⁹ hypothesized that during prolonged wakefulness adenosine accumulates selectively in the BF and promotes the transition from wakefulness to sleep by inhibiting the wakefulness-promoting BF neurons through its action at the adenosine A1 receptor. Regulation of the levels of extracellular adenosine depends primarily on metabolic rate: Increased metabolism leads to reduced high-energy phosphate stores and increased adenosine, which, via an equilibrative transporter, leads to increased extracellular adenosine. The BF wakefulness-promoting neurons inhibit the VLPO area, and the adenosine-mediated inhibition of BF cells is one mechanism by which the VLPO cells begin to discharge as sleepiness increases and the sleep episode begins.

! KEY POINT

The suprachiasmatic nuclei (SCN) are the neuroanatomic sites of the primary mammalian biologic clock. Subpopulations of SCN neurons exhibit spontaneous patterns of discharge activity and are described as self-sustaining neural oscillators, or pacemakers. When maintained in culture and devoid of afferent input, these neural pacemakers follow a “free-running” activity pattern that is slightly less than 24 hours.

Circadian process.

In addition to the homeostatic need for sleep, sleepiness depends on a second major influence, circadian phase. Humans, similar to many other species, continue to show regular, circadian sleep-wake cycles and other physiologic and hormonal rhythms in the absence of a 24-hour light/dark (LD) cycle. These rhythms must depend on an internal “clock” or pacemaker that is self-sustaining in the absence of external time cues and can be reset by changes in the environment. The mechanisms of this internal clock have been subject to research over many years, and significant progress has been made using genetic data obtained from a wide variety of species, including the bread mold Neurospora, the fruit fly Drosophila, and the mouse.²⁰ The diversity of the species from which these results have been obtained emphasizes a remarkable conservation of function in time-keeping in biologic systems during evolution.

This circadian influence on sleep also acts through the VLPO area to work in concert with the homeostatic drive to maintain sleep by consolidating the sleep period. The circadian pacemaker achieves this consolidation by a mechanism that, at first sight, seems to be in a paradoxical phase relationship to the normal timing of the sleep period. This assumption follows from the fact that the circadian drive for wakefulness is strongest in the evening hours, just before the normal time of sleep onset. Conversely, the circadian drive for sleepiness is strongest in the morning hours, just before the usual waking time. This process helps to consolidate the sleep phase despite the homeostatic drive for sleepiness in the evening and homeostatic drive for wakefulness in the morning.

Combining the characteristics of the endogenous sleep-wake rhythm with those of a circadian oscillator has led to testable mathematical models of sleep propensity. One of the most significant of these models was developed by Borbely,²¹ who based his model on a two-process single oscillator model. In this model, sleep is seen as the net result of two processes. One, process S, or sleep propensity, builds during wakefulness and declines exponentially during sleep and is indexed by the delta power of the EEG. As noted earlier, adenosine is a likely candidate for the endogenous mediator of process S. The second process, process C, is an endogenous circadian oscillator that closely parallels core body temperature. The output from the endogenous clock is probably the mediator of process C.

Circadian sleep-wake and physiologic rhythms are treated as sinusoidal variables in this model, assumptions that are not supported by actual data. The sleep-wake state is essentially a binary process, and the actual shape of the variation in physiologic variables across the nycthemeron is asymmetric and is modulated, under normal conditions, by changes related to activity and sleep onset. The recording of core body temperature under several different conditions, including normal expression of the sleep-wake cycle, sleep deprivation with constant activity over 24 hours, and continuous bed rest with minimal activity but normal sleep-wake behavior, might provide more accurate modeling data after appropriate subtractive manipulation.²² This point is addressed in more detail subsequently.

Circadian processes and chronobiology

The circadian processes of sleep and wakefulness are also important for considerations related to pharmacotherapy. The basic underlying biologic oscillators affect the response to a drug in addition to driving the biologic rhythms such as sleep propensity and core body temperature.

Circadian timing system

Biologic rhythms vary systematically across the 24 hours of the nycthemeron. In particular, circadian rhythms are driven by endogenous pacemakers that have periods (τ, tau) approximating 24 hours.²³ As noted previously, these are self-sustained, internally generated biologic signals that, in the natural environment, are normally synchronized or entrained to the 24-hour LD cycle. Both sleep and temporal organization are evolutionarily conserved behaviors, although they can change during the life span of an organism.²⁴ Such evolutionarily conserved, intrinsic temporal order is crucial for human health and well-being, and disturbances in these rhythms result in behavioral, physiologic, psychological, biochemical, and endocrinologic abnormalities.

Studies over many years have attempted to derive an accurate estimation of the period of the endogenous pacemaker. To do so required the subjects not only to be placed under “free-running” conditions in which the environment was completely devoid of any time cues, but also under conditions in which the period of the endogenous pacemaker could not be entrained to the rest-activity cycle. These considerations led to the adoption of the forced desynchrony protocol, originally developed by Kleitman in 1938.⁹ In this type of study, subjects were kept on a rest-activity cycle that was sufficiently long (e.g., 28 hours) to prevent the entrainment of the endogenous pacemaker to this rhythm. Results were variable, however, until Czeisler and colleagues²⁵ in 1999 changed the protocol so that their subjects were exposed to very dim light (about 10 to 15 lux) throughout. In this way, Czeisler and colleagues were able to determine the period of the human circadian pacemaker at 24.18 hours; they also reported that healthy older subjects had the same periodicity, with the same stability and precision, as younger subjects.

Under normal circumstances, circadian rhythms become synchronized or entrained to the environmental LD cycle, which acts as a pervasive and prominent synchronizer, or zeitgeber.²⁶ Light signals are perceived in the retina and are transmitted via a monosynaptic pathway, the retinohypothalamic tract (RHT), to the suprachiasmatic nucleus (SCN). Non–image forming effects of retinal light exposure range from effects on various physiologic measures, such as shifts in the circadian rhythms of melatonin and body temperature, to effects on psychological measures—for example, high environmental light intensity increases arousal and alertness.

Although the mechanism by which light exerts these alerting effects is unknown, a more recently discovered network of blue light–sensitive retinal ganglion cells (RGCs)²⁷^,²⁸ is likely part of the input system for the physiologic effects. In animals²⁹^,³⁰ and in humans,³¹ the suppression of melatonin and shifts in circadian rhythms are particularly sensitive to a short-wavelength, blue component of light. The blue light–sensitive RGCs express the photopigment melanopsin and a neuropeptide, pituitary adenylate cyclase activating polypeptide (PACAP).²⁷^,²⁸^,³² Significantly, the RGCs project to brain regions implicated in sleep mechanisms, including the SCN and the VLPO area.²⁸^,³³ This finding suggests that they might mediate the effect of light on sleepiness.³⁴

Suprachiasmatic nucleus: the central oscillator

The SCN, which is localized to the anterior hypothalamus, acts as the “biologic clock” to coordinate the circadian rhythm.35–40 The SCN receives photic information via the glutamatergic RHT and the geniculohypothalamic tract (GHT), which contains NPY and nonphotic information via serotoninergic neurons originating in the dorsal raphe nucleus (DRN). SCN neurons, which project to the dorsomedial and posterior hypothalamic areas and the VLPO area, actively promote and maintain wakefulness during the day and sleep at night (Figure 23-7). The SCN is involved in the regulation of the timing of sleep-wake states and in the expression of the sleep-wake cycle and may play a role in the coordination of specific sleep stages.³⁷^,^40–42 The role of the SCN in the control of sleep has been studied extensively in several species. In squirrel monkeys, a diurnal species similar to humans, the circadian signal produced by the SCN promotes wakefulness during the subjective day and consolidation of sleep at night. Lesions of the SCN have been found to disrupt the consolidation of both sleep and wakefulness as a result of a disrupted circadian rhythm.⁴³ Neurons in the SCN express two melatonin receptors (MT1 and MT2) that have different functional roles.

Figure 23-7 Primary afferent and efferent pathways of suprachiasmatic nuclei *(SCN).* In addition to the well-described retinohypothalamic tract, which provides SCN with extrinsic stimuli (light), the lesser described serotoninergic afferent pathway arising from the raphe nuclei and cholinergic afferent pathways originating in the pedunculopontine, parabigeminal, and laterodorsal tegmentum nuclei *(LDT)* are illustrated. This figure graphically illustrates a conceptual model through which intrinsic behavioral state–related stimuli could affect neuronal activity in SCN. (From Decker MJ, Lee SY, Rye DB, et al: Paradoxical sleep suppresses immediate early gene expression in the rodent suprachiasmatic nuclei, *Front Neurol* 1:122, 2010.)

Chronopharmacology

Chronobiology is concerned with the mechanisms of periodic biologic influences on health and disease ⁴⁴; pharmacology refers to the medical discipline concerned with the biochemical and physiologic aspects of drug effects, including absorption, distribution, metabolism, elimination, toxicity, and specific mechanisms of drug action. The effectiveness of drugs, also a critical aspect of pharmacology, depends on pharmacodynamics (i.e., what the drug does to the body) and pharmacokinetics (i.e., what the body does to the drug). These considerations involve the quantitative aspects of drug absorption, distribution, and excretion that are crucial for the design of rational dosage regimens.

Chronopharmacology, or the study of time-dependent variations in pharmacology,⁴⁵ was developed from the inclusion of chronobiologic principles in the study of pharmacology. Traditionally, drug delivery has assumed that a chemical is absorbed predictably from the site of administration. A second-generation drug delivery goal has been the achievement of a continuous constant rate (i.e., zero-order) delivery of drugs. However, living organisms are not “zero-order” in their response to drugs. As mentioned earlier, living organisms are predictable resonating dynamic systems governed by intrinsic oscillators, so they require different amounts of drug at different times within the circadian cycle to maximize the desired and undesired (i.e., chronotoxicity) effects of the drug. Two concepts are important when considering changes in drug efficacy over the 24-hour period. The first is circadian changes in drug bioavailability (i.e., chronokinetics), and the second is circadian changes in the susceptibility to the drug (chronesthesy). In brief, clinical chronopharmacology, or chronotherapeutics, is the purposeful alteration of drug levels to match biologic rhythms and to optimize therapeutic outcomes and minimize side effects.

Melatonin as a chronobiotic and chronohypnotic agent

Drugs that directly influence circadian mechanisms are often referred to as chronobiotics.⁴⁶ The prototype for this type of drug is melatonin (N-acetyl-5-methoxytryptamine), a pineal hormone that has been identified as an important endogenous regulatory factor, with levels that vary with circadian time. In humans, melatonin is an important signal for maintaining endogenous rhythms in synchrony with the environmental LD cycle. Melatonin is exclusively secreted during the subjective night, and its plasma level increases during the evening and declines in the early morning.⁴⁷ The finding that melatonin is secreted primarily during the night and the close relationship between the nocturnal increase in endogenous melatonin and the timing of human sleep have suggested that melatonin might be important in sleep regulation. The onset of melatonin secretion occurs approximately 2 hours before bedtime and has been shown to correlate with the onset of evening sleepiness. In other words, the transition phase from wakefulness and arousal to high sleep propensity coincides with the nocturnal increase in endogenous melatonin. As noted earlier, SCN neurons express high concentrations of both melatonin receptors, MT1 and MT2, although both receptors are also found widely expressed throughout the CNS. Signaling via MT1 leads primarily to inhibition of activity in SCN neurons.⁴⁸^,⁴⁹ In contrast, the principal MT2-mediated actions in the SCN are related to circadian phase shifts and constitute activation of protein kinase C and increased cell activity.⁵⁰^,⁵¹

It is possible that melatonin contributes to sleep initiation by inhibiting the circadian wakefulness-generating mechanisms, an effect that could be mediated by MT1 receptors in the SCN. Melatonin release is pulsatile during light sleep, and the hormone could function to induce deeper sleep and prevent awakening by continuing to inhibit arousal at the level of the SCN. Overall, the hypnotic and chronobiotic effects of melatonin might be mediated in the SCN and possibly through the MT1 receptor.52–54 Induction and maintenance of sleep at the appropriate circadian phase, which could be MT1-mediated, is different from shifting the phase of sleep, which could be MT2-mediated, as a result of exogenous change in the zeitgeber.

It is relevant that the hypnotic effect of melatonin depends on the circadian phase of administration. Stone and coworkers,⁵⁵ using a double-blind placebo-controlled study, found that melatonin administered at night (23:30 hours) had no significant effect on sleep in healthy individuals, whereas melatonin administered in the evening (18:00 hours) exerted a hypnotic activity. Despite such evidence for the hypnotic action of melatonin, its efficacy in promoting sleep is still controversial, especially because most results show only borderline significance or are otherwise difficult to evaluate because of methodologic inconsistencies.⁵⁶ However, the relatively poor outcomes in these studies in terms of sleep efficiency or total sleep time may be due to the short half-life of melatonin in plasma (less than 30 minutes). It is possible that a melatonin receptor agonist with a longer half-life and occupying receptors in the SCN for a longer duration could be more effective than melatonin in promoting sleep in insomniac patients.⁵⁷

Ramelteon is a melatonin receptor agonist that has been shown to be selective for MT1 and MT2 receptors but without affinity for the melatonin-binding site, quinone reductase 2, previously denoted MT3.⁵⁸^,⁵⁹ Ramelteon has no affinity for other major CNS receptors, including binding sites for neurotransmitters, neuropeptides, regulatory enzymes, or ion channels.⁵⁸^,⁵⁹ However, various additional non–membrane binding sites of melatonin remain to be tested.⁶⁰

Sleep disorders: causes and treatments

As noted at the beginning of this chapter, there are more than 80 identified sleep disorders. Epidemiologic data reveal that the incidence and prevalence of these disorders vary in the general population. Although some of these disorders can be effectively managed by nonpharmacologic therapies or medical management (e.g., CPAP for obstructive sleep apnea), others are treated effectively with pharmacologic agents. Available pharmacologic treatment options for some of these disorders are outlined in this section.

Insomnia

Insomnia is characterized by difficulty in falling asleep (i.e., a sleep latency of greater than 30 minutes), insufficient sleep (i.e., total sleep time of less than 5.5 to 6 hours), multiple nocturnal awakenings, early morning awakening with inability to resume sleep, or nonrestorative sleep. Common daytime complaints include somnolence, fatigue, irritability, and difficulty concentrating and performing everyday tasks. In addition, subjects with a diagnosis of insomnia are at higher risk for illness and for injury caused by drowsiness while driving. Because insomnia is associated with difficulty in concentration, it is a major risk factor for accidents.⁶¹ Insomnia is a common disorder that affects 30% to 35% of the U.S. adult population and is chronic in about 10%.⁶² The risk of insomnia is greatest in elderly adults. The adverse physiologic and psychological sequelae of insomnia have a major negative impact on the quality of life in affected individuals.⁶³ Almost 10% of people with chronic insomnia have daytime consequences of fatigue, irritability, and impaired concentration that affect health, mood, and normal functioning. With reduced productivity and an increased risk of accidents, the overall economic burden of insomnia is estimated to be 1% of gross domestic product.⁶⁴^,⁶⁵ Insomnia is also experienced as a stressor by patients who have major depressive disorders, and disturbed sleep has been identified as a hallmark of depression; this is not always recognized in clinical practice.

Pharmacologic treatment of insomnia in the last few decades has been based on several classes of medication (Table 23-2). Benzodiazepines were introduced in the 1970s and rapidly increased in popularity because of their efficacy and relative safety compared with barbiturates, carbamates, chloral derivatives, and methaqualone. In recent years, however, prescriptions for benzodiazepines have declined because of their associated side-effect profile, including the tendency of benzodiazepines to promote dependence, the occurrence of rebound insomnia after withdrawal of short-acting and intermediate-acting derivatives, and the loss of efficacy after a few weeks of treatment. The reduction in benzodiazepine use has also coincided with the introduction of a structurally dissimilar group of nonbenzodiazepine derivatives, including the cyclopyrrolone agents zopiclone and eszopiclone, the imidazopyridine derivative zolpidem, and the pyrazolopyrimidine compound zaleplon.

TABLE 23-2

Medications Approved by U.S. Food and Drug Administration for Treatment of Insomnia

MEDICATION	TRADE NAME	DOSE (mg)	HALF-LIFE (hr)	DEA SCHEDULE
BZD Receptor Agonists
*Immediate-Release BZDs*
Estazolam	ProSom	1, 2	8-24	IV
Flurazepam	Dalmane	15, 30	48-120	IV
Quazepam	Doral	7.5, 15	48-120	IV
Temazepam	Restoril	7.5, 15, 22.5, 30	8-20	IV
Triazolam	Halcion	0.125, 0.25	2-4	IV
*Immediate-Release Non-BZDs*
Eszopiclone	Lunesta	1, 2, 3	5-7	IV
Zaleplon	Sonata	5, 10	1	IV
Zolpidem	Ambien	5, 10	1.5-2.4	IV
*Modified-Release Non-BZDs*
Zolpidem CR	Ambien CR	6.25-12.5	2.8-2.9	IV
Selective Melatonin Receptor Agonist
Ramelteon	Rozerem	8	1-2.6	None

BZD, Benzodiazepine; DEA, Drug Enforcement Administration.

Restless legs syndrome and periodic limb movement disorder

The first clinical description of RLS was made in the 17th century by Willis, an English physician, who stated, “Wherefore to some, when being abed they betake themselves to sleep, presently in the Arms and Legs, leaping and Contractions of the tendons, and so great a Restlessness and Tossing of their members ensue, that the diseased are no more able to sleep, than if they were in a Place of greatest torture.” More than 200 years later, the physician Ekbom coined the phrase “restless legs” and stated that “ . . . the syndrome is so common and causes such suffering that it should be known to every physician.”

RLS is now recognized as a chronic and progressive neurologic disorder characterized by unpleasant sensations in the legs and a compelling urge to move them while the patient is awake. These symptoms occur most frequently during the evening or at night as well as during periods of rest. Approximately 5% to 15% of European and U.S. populations are affected by RLS. Adverse outcomes associated with RLS include hypertension, alcohol abuse, neurocognitive deficits, and decrements in mental and physical health. Patients with RLS report an urge to move their limbs during the daytime if they become confined in a delineated space for extended periods, for example, having to sit at a desk. Unpleasant sensations in the limbs typically combined with urges to move the legs occurring in the evening often lead to difficulties with sleep onset or sleep maintenance. When these symptoms are experienced by children, they have often been incorrectly identified as “growing pains.”

PLMD is frequently associated with RLS. PLMD is defined as periodic episodes of spontaneous, repetitive, and highly stereotyped involuntary limb movements that occur during sleep. Four or more repetitive muscle contractions lasting 0.5 to 5 seconds and separated by 4 to 90 seconds are conventionally regarded as indicative of PLMD. The night-to-night variability in their occurrence and variability in the intensity of daytime and evening sensory symptoms often lead to PLMD being underappreciated or missed during a clinical evaluation. Chronic sleep restriction and sleep fragmentation is one adverse outcome attributed to both RLS and PLMD; this is believed to contribute to the deleterious physical, mental, and social effects associated with the disorders. Consequently, the negative impact of RLS on quality of life is similar to that observed with other chronic disorders such as depression, heart failure, and diabetes.

More recent epidemiologic and genetic linkage studies distinguish two forms of RLS: early onset, or primary, RLS, and late onset, or secondary, RLS. The early onset form typically begins in childhood or young adulthood with gradual progression in symptom severity. Early onset RLS is inherited in an autosomal dominant fashion, shows genetic anticipation, is twice as prevalent in females, and is associated with at least two separate genetic loci.⁶⁶ In addition, PLMD is more typically associated with this form of RLS.

Late onset RLS is characterized by a later age of symptom onset, usually older than 45 years; an equal female to male ratio; a rapidly progressive course; and a relationship with anemia and other associated identifiable causes such as diabetes, kidney disease, neuropathy and even nervous system trauma.⁶⁶ In addition to these intrinsic pathophysiologic mechanisms, certain extrinsic factors are associated with increased frequency and severity of symptoms. H₂ histamine antagonists such as ranitidine or cimetidine have been reported to worsen symptoms of RLS.⁶⁷ Caffeine, alcohol, and some antidepressants including fluoxetine also are reported to worsen the frequency and severity of RLS or PLMD symptoms.⁶⁸^,⁶⁹ Epidemiologic studies suggest that the onset of RLS increases with age, with a prevalence rate of 2% in children, 3% in 30-year-olds, and up to 20% in 80-year-olds. Genetic studies and linkage analyses also show that early onset RLS is a heritable trait, but the pathophysiologic mechanisms of RLS remain unclear.⁶⁶

Because of the essential motor component of the disorder, dopamine deficiencies may contribute to the etiology of RLS and PLMD. Research has focused on determining if reductions in extracellular dopamine levels within the CNS or deficiencies in postsynaptic responsivity to dopamine might contribute to the symptoms of RLS and PLMD. Reduced levels of extracellular dopamine are central to several hypotheses regarding neurochemical substrates contributing to RLS. However, elucidation of any actual dopaminergic dysfunction has remained enigmatic. It is possible that reduced synthesis or increased sequestration of dopamine within cell bodies and terminals leads to diminished extracellular dopamine. Alternatively, dopamine production and release may be normal, but the number or type, or both, of postsynaptic dopamine receptors may be altered and so result in the symptoms.

The first link made between dopamine and RLS was based on the observation that many patients derived relief from dopamine augmenting drugs. This was acknowledged in the “Practice Parameters for the treatment of RLS and PLM,”⁷⁰ which stated that dopaminergic agents are the best-studied and most successful agents for treatment of RLS and PLMD. Following multiple clinical trials with dopamine-enhancing compounds, levodopa with decarboxylase inhibitors and dopamine agonists such as pergolide were found to be the most effective for treatment of RLS and PLMD.⁷⁰ Despite the promise that dopamine-enhancing compounds can reduce the symptoms of RLS, it should be noted that their use for RLS is currently approved only for adults; data are lacking with regard to their use for RLS in pediatric populations and during pregnancy. When prescribing any type of dopaminergic medication for RLS symptom relief, the potential for side effects, such as nausea, gastrointestinal distress, reduced blood pressure, and sleepiness, should be taken into account and discussed with a sleep medicine physician. In addition, because dopamine modulates mood, cognition, wakefulness, and sleep, any dopamine precursor, agonist, or antagonist can feasibly result in acute thought and behavioral changes. Given that most RLS patients need very low doses of dopaminergic medications for symptomatic relief, the likelihood of a serious or adverse outcome is remote (Table 23-3).

TABLE 23-3

Pharmacologic Management of Restless Legs Syndrome

DRUG	DOSE (mg/day)	TIME TO PEAK PLASMA LEVEL (min)	HALF-LIFE (hr)	MODE OF ELIMINATION
Levodopa	100-400	30	1.5-3	Hepatic
Carbidopa/Levodopa	10/100-25/250	120	6-8	Hepatic
Bromocriptine	2.5-10	45-60	3-4 (up to 40)	Hepatic
Pergolide	0.1-0.75	60	27	Renal
Cabergoline	0.25-3.0	120	63-68	Hepatic
Pramipexole	0.25-1.5	120	8-12	Renal
Ropinirole	0.5-4.0	60-120	Approximately 6	Hepatic

Narcolepsy

The earliest clinical descriptions of narcolepsy were documented by the German physician Westphal in 1877 and by Fisher in 1878. However, the French physician Gelineau, writing in 1880, is generally acknowledged as the first clinician to recognize narcolepsy as a distinct clinical entity. Initial descriptions included characterization of excessive daytime sleepiness and “sleep attacks.” Although all three of these early descriptions documented the appearance of sleep attacks, they did not discern the symptoms of sudden muscle weakness triggered by an emotional event (i.e., cataplexy) as separate from the sleep attack event. This is most evident from the first description of narcolepsy as a clinical entity: “I propose to give the name of narcolepsy (‘narco = somnolence’ and ‘lepsy = seized by’) to a rare neurosis or at least little known until now, characterized by a mandatory need to sleep, sudden and of short duration, that recurs at more or less close intervals.”⁷¹ In 1902, Loewenfeld recognized that the emotion-induced muscle weakness was a separate feature of the disorder and first used the term cataplexy to describe it.

Today, narcolepsy is characterized by a tetrad of clinical symptoms. Two features of the tetrad—persistent excessive daytime sleepiness and cataplexy—were documented in the original descriptions of the disorder and remain the only two clinical features essential for diagnosis. The recognition of two additional features—hypnagogic hallucinations (the onset of dreams while still awake) and sleep paralysis (a temporary loss of muscle tone or an inability to perform voluntary movements either at sleep onset or on awakening)—were added later.⁷² Another common symptom of narcolepsy is fragmented sleep with multiple arousals and awakenings at night.

Narcolepsy may not be as rare a disorder as once thought; according to the National Institute of Neurological Disorders and Stroke (NINDS), narcolepsy is an under-recognized and underdiagnosed condition. Nevertheless, in the United States, narcolepsy is the third most frequently diagnosed primary sleep disorder after sleep apnea and RLS. Current estimates suggest that 1 in about 2000 Americans are narcoleptic. The prevalence of narcolepsy is apparently significantly greater in Japan, affecting 1 in 600 people, but substantially less in Israel, affecting only 1 in 500,000, although differences in diagnosis may account for at least part of this variation. There is no gender difference in the prevalence of the disorder.

The onset of narcoleptic symptoms typically occurs between the ages of 15 and 30, although there are reports of symptom onset occurring in very young children and in adults older than 30. In addition, it is not unusual for 12 years to elapse between the initial onset of symptoms and a definitive diagnosis. Up to 10% of patients diagnosed with narcolepsy report that a close relative has similar symptoms. The familial association of narcolepsy was recognized in the early descriptions, as noted in the mother of the narcoleptic patient first examined by Westphal and the sister of the first narcoleptic patient described by Fisher. Such familial clustering suggests a genetic origin for this disorder. Although immediate family members of narcoleptics are at a statistically greater risk of developing the disorder, this risk is low compared with purely genetic disorders and indicates that other factors must be involved.

Most cases of narcolepsy are sporadic and occur without evidence of genetic inheritance. Until more recently, the etiology of narcolepsy was unknown, but it was associated with the specific human leukocyte antigen (HLA) allele, DQB10602, and often in combination with HLA-DR2 (DRB115). In 1998, two research groups first described, clustered around the lateral and perifornical hypothalamus, a group of cells that contained a previously unknown neurotransmitter.⁷³^,⁷⁴ These cells contained orexin/hypocretin, and based on the neuroanatomy they were hypothesized to be involved in the regulation of sleep and wakefulness. However, the physiology suggested a major role in food intake and energy balance, which was the focus of research during the first 12 months. Within 1 year of the discovery, however, two independent groups discovered in 1999 that the clinical symptoms of narcolepsy were associated with a loss or dysfunction of orexin/hypocretin-containing cells.⁷⁵^,⁷⁶

Following the initial descriptions of narcolepsy, various empiric treatments were tried with little success. Among these were spinal fluid taps, intrathecal air injection, x-ray irradiation of portions of the hypothalamus, and later ephedrine administration. In 1935, Prinzmetal and Bloomberg synthesized a new compound, benzedrine, the original drug in the class later known as amphetamines. Although originally developed to treat nasal congestion, it had no effect on the nasal mucosa, but when given orally, benzedrine led to a reduction in weight, and it was soon routinely used as an appetite suppressant. Subsequently, the CNS-stimulating effect of the amphetamines was recognized, leading to their use to treat hypersomnolence. For many years, a regimen of amphetamines in conjunction with a tricyclic antidepressant such as imipramine, which reduces REM sleep, was the standard pharmaceutical intervention for narcolepsy. Then modafinil, a nonamphetamine wake-promoting agent, was developed.

With a side-effect profile free from addiction, tolerance, and other adverse outcomes associated with amphetamines, modafinil soon became the treatment of choice for alleviating the symptoms of excessive daytime sleepiness associated with narcolepsy. In 2002, sodium oxybate, (Xyrem), a CNS depressant, gained U.S. Food and Drug Administration (FDA) approval for the treatment of excessive daytime sleepiness and cataplexy in narcoleptic patients. The use of a CNS depressant may seem counterintuitive as a strategy for treating excessive daytime sleepiness. However, sodium oxybate, taken immediately before sleep and again 2 to 4 h later, causes an increase in SWS, a reduction in the number of nocturnal awakenings, and enhanced sleep continuity. The result is a reduction in daytime sleepiness and cataplexy and a less fragmented sleep period. The mechanism by which sodium oxybate reduces excessive daytime sleepiness and cataplexy is unknown, but it may involve activation of GABA_B receptors.

Determining the most appropriate treatment for a narcoleptic patient is influenced by several factors, including age, severity of symptoms, presence or absence of cataplexy, other medical conditions, and concomitant medications. Conservative treatment of narcolepsy involves administration of two or more medications with a stimulant for excessive daytime sleepiness, a tricyclic antidepressant for cataplexy, and a hypnotic for insomnia and fragmented nocturnal sleep. A young narcoleptic patient without cataplexy may achieve some relief of symptoms by maintaining a fixed schedule of sleep time combined with prescheduled daytime naps, if feasible. This approach ensures an adequate opportunity for sleep, and coupled with an alerting compound such as modafinil or armodafinil (the R-enantiomer of modafinil with a longer half-life of 10 to 15 hours) it may help to restore a functional level of daytime alertness.

In contrast, a patient presenting with more severe symptoms, including cataplexy, hypnagogic hallucinations, and sleep fragmentation may require aggressive treatment. In addition to good sleep hygiene, an amphetamine such as methylphenidate might be prescribed to help sustain daytime wakefulness, together with sodium oxybate. This regimen could be combined with a selective serotonin reuptake inhibitor (SSRI) or a tricyclic antidepressant to treat cataplexy and the other symptoms of REM sleep dysregulation, including hypnagogic hallucinations. In a patient with sleep fragmentation, sodium oxybate is often useful because sleep continuity is enhanced, and excessive daytime sleepiness and cataplexy are controlled. Cataplexy and hypnagogic hallucinations require additional agents such as sodium oxybate, an SSRI, or a tricyclic antidepressant. Nonpharmacologic therapies, such as scheduled naps, regular sleep and wake schedules, and proper sleep hygiene, are essential elements for any successful treatment regimen.

Parasomnias

Parasomnias are undesirable motor, sensory, or behavioral phenomena that occur primarily during sleep.⁷⁷ These phenomena range from normal to abnormal and from benign to potentially lethal and can be associated with normal developmental processes or neurodegeneration. The ICSD-R lists 24 parasomnias encompassing NREM sleep or arousal disorders, REM sleep-related disorders, sleep-wake transition disorders, and other parasomnias. The focus here is on more common NREM and REM sleep parasomnias. NREM sleep parasomnias, also termed arousal disorders, include confusional arousals, sleep terrors, and sleepwalking. Arousal disorders can be either primary or secondary if they are associated with an identifiable pathology such as a seizure disorder, obstructive sleep apnea, nocturnal cardiac ischemia, or nocturnal paroxysmal dystonia. The pathophysiologic mechanisms underlying arousal disorders are still unknown, but current hypotheses suggest that they may result from the brain being simultaneously in a state of partial wakefulness and NREM sleep. This state leads to an ability to perform complex motor or verbal actions without conscious awareness of the actions.

Primary arousal disorders share several common factors, including familial clustering, which suggests a genetic predisposition; childhood predominance; and a tendency to occur during NREM sleep. Confusional arousals are characterized by episodes of marked mental confusion during or after an arousal from sleep. They usually occur during the first third of the night, last 30 seconds to 5 minutes, and may be accompanied by mumbling or automatic behaviors or both. During the event, the person does not leave the bed, and there are no signs of fear or terror. Following a confusional arousal, the individual usually falls back to sleep with no recollection of the event (i.e., retrograde amnesia) on awakening. Triggers for confusional arousals include anything that either fragments sleep or enhances SWS. Examples include environmental factors such as noise or temperature, stress, fever, pain, pregnancy, recovery from sleep deprivation, or CNS-active medications. As noted earlier, youth, a family history, and a history of being a deep sleeper are predisposing factors.

Sleep terrors, which are observed primarily in children, are similar to confusional arousals and occur during the first third of the night. They also begin in NREM sleep, typically during SWS at a time when an episode of REM sleep would be expected, and last 30 seconds to 5 minutes. The triggers for sleep terrors are similar to those for confusional arousals; the principal difference is that sleep terrors are accompanied by an abrupt awakening, intense vocalization, and inconsolable fear or terror. Sleep terrors most frequently occur in children 5 to 7 years old and appear with equal prevalence in boys and girls. Most children with sleep terrors outgrow them by 8 years of age, although about 30% may continue to experience them into adolescence; only about 1% experience sleep terrors as adults.

Another parasomnia, somnambulism or sleepwalking, also occurs during NREM sleep but is characterized by the presence of automatic behaviors of varying complexity, including walking, eating, mumbling, and, rarely, violence. The duration of these episodes can be 15 minutes to several hours. The episode is usually self-limiting and terminates with a return to sleep. Clinical evidence shows that attempts to intervene may be met with resistance and outbursts.

As with other NREM sleep parasomnias, the familial clustering of somnambulism suggests a genetic predisposition. Triggers for somnambulism, such as sleep fragmentation and increased depth or duration of SWS, are similar to those of the other arousal disorders. The age of onset for somnambulism is about 5 years with the highest prevalence at about 12 years of age. Somnambulism can occur in 15% to 30% of children and young adolescents, with boys and girls equally affected. Most children who are sleepwalkers typically outgrow the events by age 15, but 1% may continue to experience episodes in adulthood.

Another clinically identifiable category of parasomnias occurs during REM sleep and for this reason typically in the second half of the night. These include REM sleep behavior disorder (RBD), nightmare disorder, and isolated sleep paralysis. In contrast to NREM arousal disorders, REM sleep parasomnias usually affect adults more frequently than children, and they do not exhibit a genetic pattern of inheritance. RBD, in particular, is associated with neurodegenerative disorders such as Parkinson disease and multiple systems atrophy and usually occurs in older men. Symptoms include violent dream enactment behavior owing to a loss of atonia during REM sleep. If left untreated, RBD can cause serious injury to the patient and the sleeping partner.

Treatment for NREM and REM sleep parasomnias frequently includes avoidance of potential triggers and, in the case of somnambulism and RBD, necessitates a safe, well-monitored sleeping environment. The most common pharmaceutical treatment for REM and NREM sleep parasomnias is usually a longer acting benzodiazepine. Through reduction of both SWS and REM sleep time and the number of transitions between sleep states, benzodiazepines essentially reduce the occurrence of the state in which an arousal disorder or RBD episode can occur. Benzodiazepines were initially developed as anxiolytics and subsequently as hypnotics. The success of the first compounds led to further research and development, and many compounds of this class eventually became available. Notable hypnotic benzodiazepine compounds are nitrazepam (Mogadon), temazepam (Restoril), flurazepam (Dalmane), and midazolam (Versed); others, such as clonazepam (Klonopin), are frequently used in treatment of parasomnias and as antiseizure medications.

Considerations for employing benzodiazepine compounds in treatment include their half-life (i.e., the time required for one-half of the active drug to be metabolized or eliminated from the body). Short-acting benzodiazepines have half-lives of 12 hours or less, but long-acting benzodiazepines have half-lives that often exceed 24 hours. A gradual increase in the blood levels of a longer acting drug has the potential to cause residual effects. A benzodiazepine taken in the evening to reduce the likelihood of experiencing an arousal event may induce residual sleepiness the next day. With regard to their use for the treatment of arousal disorders, apart from their role in the reduction of the overall duration of time in states as noted earlier, no definitive conclusions are yet possible concerning their mechanism of action. Clonazepam, as a longer acting benzodiazepine, is a frequent first choice, although careful selection of the appropriate benzodiazepine for a particular patient is essential to reduce the likelihood of residual daytime sleepiness. This is especially important to consider when treating parasomnias because many patients are children or elderly adults.

SELF-ASSESSMENT QUESTIONS