Sleep is made up of two physiological states: non-rapid eye movement (NREM) sleep and rapid eye movement (REM) sleep. In NREM sleep, which is composed of stages 1 through 4, most physiological functions are markedly lower than in wakefulness. REM sleep is a qualitatively different kind of sleep, characterized by a high level of brain activity and physiological activity levels similar to those in wakefulness. About 90 minutes after sleep onset, NREM yields to the first REM episode of the night. This REM latency of 90 minutes is a consistent finding in normal adults; shortening of REM latency frequently occurs with such disorders as depressive disorders and narcolepsy.

For clinical and research applications, sleep is typically scored in epochs of 30 seconds, with stages of sleep defined by the visual scoring of three parameters: electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) recorded beneath the chin. The EEG records the rapid conjugate eye movements that are the identifying feature of the sleep state (no or few rapid eye movements occur in NREM sleep); the EEG pattern consists of low-voltage, random, fast activity with sawtooth waves; the EMG shows a marked reduction in muscle tone. The criteria defined by Allan Rechtschaffen and Anthony Kales in 1968 are accepted in clinical practice and for research around the world.

In normal persons, NREM sleep is a peaceful state relative to waking. The pulse rate is typically slowed five to ten beats a minute below the level of restful waking and is very regular. Respiration is similarly affected, and blood pressure also tends to be low, with few minute-to-minute variations. The body musculature resting muscle potential is lower in REM sleep than in awaking state. Episodic, involuntary body movements are present in NREM sleep. There are few, if any, REMs, and seldom do any penile erections occur in men. Blood flow through most tissues, including cerebral blood flow, is slightly reduced.

The deepest portions of NREM sleep—stages 3 and 4—are sometimes associated with unusual arousal characteristics. When persons are aroused 30 minutes to 1 hour after sleep onset—usually in slow-wave sleep—they are disoriented, and their thinking is disorganized. Brief arousals from slow-wave sleep are also associated with amnesia for events that occur during the arousal. The disorganization during arousal from stage 3 or stage 4 may result in specific problems, including enuresis, somnambulism, and stage 4 nightmares or night terrors.

Polygraphic measures during REM sleep show irregular patterns, sometimes close to aroused waking patterns. Otherwise, if researchers were unaware of the behavioral stage and happened to be recording a variety of physiological measures (aside from muscle tone) during REM periods, they undoubtedly would conclude that the person or animal they were studying was in an active waking state. Because of this observation, REM sleep has also been termed paradoxical sleep. Pulse, respiration, and blood pressure in humans are all high during REM sleep—much higher than during NREM sleep and often higher than during waking. Even more striking than the level or rate is the variability from minute to minute. Brain oxygen use increases during REM sleep. The ventilatory response to increased levels of carbon dioxide is depressed during REM sleep, so that no increase in tidal volume occurs as the partial pressure of carbon dioxide increases. Thermoregulation is altered during REM sleep. In contrast to the homeothermic condition of temperature regulation during wakefulness or NREM sleep, a poikilothermic condition (a state in which animal temperature varies with the changes in the temperature of the surrounding medium) prevails during REM sleep. Poikilothermia, which is characteristic of reptiles, results in a failure to respond to changes in ambient temperature with shivering or sweating, whichever is appropriate to maintaining body temperature. Almost every REM period in men is accompanied by a partial or full penile erection. This finding is clinically significant in evaluating the cause of impotence; the nocturnal penile tumescence study is one of the most commonly requested sleep laboratory tests. Another physiological change that occurs during REM sleep is the near-total paralysis of the skeletal (postural) muscles. Because of this motor inhibition, body movement is absent during REM sleep. Probably the most distinctive feature of REM sleep is dreaming. Persons awakened during REM sleep frequently (60 to 90 percent of the time) report that they had been dreaming. Dreams during REM sleep are typically abstract and surreal. Dreaming does occur during NREM sleep, but it is typically lucid and purposeful.

The cyclical nature of sleep is regular and reliable; a REM period occurs about every 90 to 100 minutes during the night. The first REM period tends to be the shortest, usually lasting less than 10 minutes; later REM periods may last 15 to 40 minutes each. Most REM periods occur in the last third of the night, whereas most stage 4 sleep occurs in the first third of the night.

These sleep patterns change over a person’s life span. In the neonatal period, REM sleep represents more than 50 percent
of total sleep time, and the EEG pattern moves from the alert state directly to the REM state without going through stages 1 through 4. Newborns sleep about 16 hours a day, with brief periods of wakefulness. By 4 months of age, the pattern shifts so that the total percentage of REM sleep drops to less than 40 percent, and entry into sleep occurs with an initial period of NREM sleep. By young adulthood, the distribution of sleep stages is as follows:

This distribution remains relatively constant into old age, although a reduction occurs in both slow-wave sleep and REM sleep in older persons.

Most researchers think that there is not one simple sleep control center but a small number of interconnecting systems or centers that are located chiefly in the brainstem and that mutually activate and inhibit one another. Many studies also support the role of serotonin in sleep regulation. Prevention of serotonin synthesis or destruction of the dorsal raphe nucleus of the brainstem, which contains nearly all of the brain’s serotonergic cell bodies, reduces sleep for a considerable time. Synthesis and release of serotonin by serotonergic neurons are influenced by the availability of amino acid precursors of this neurotransmitter, such as L-tryptophan. Ingestion of large amounts of L-tryptophan (1 to 15 g) reduces sleep latency and nocturnal awakenings. Conversely, L-tryptophan deficiency is associated with less time spent in REM sleep. Norepinephrine-containing neurons with cell bodies located in the locus ceruleus play an important role in controlling normal sleep patterns. Drugs and manipulations that increase the firing of these noradrenergic neurons markedly reduce REM sleep (REM-off neurons) and increase wakefulness. In humans with implanted electrodes (for the control of spasticity), electrical stimulation of the locus ceruleus profoundly disrupts all sleep parameters. Brain acetylcholine is also involved in sleep, particularly in the production of REM sleep. In animal studies, the injection of cholinergic-muscarinic agonists into pontine reticular formation neurons (REM-on neurons) results in a shift from wakefulness to REM sleep. Disturbances in central cholinergic activity are associated with the sleep changes observed in major depressive disorder. Compared with healthy persons and nondepressed psychiatric controls, patients who are depressed have marked disruptions of REM sleep patterns. These disruptions include shortened REM latency (60 minutes or less), an increased percentage of REM sleep, and a shift in REM distribution from the last half to the first half of the night. Administration of a muscarinic agonist, such as arecoline, to depressed patients during the first or second NREM period results in a rapid onset of REM sleep. Depression can be associated with an underlying supersensitivity to acetylcholine. Drugs that reduce REM sleep, such as antidepressants, produce beneficial effects in depression. Indeed, about half of the patients with major depressive disorder experience temporary improvement when they are deprived of sleep or when sleep is restricted. Conversely, reserpine (Serpasil), one of the few drugs that increase REM sleep, also produces depression. Patients with dementia of the Alzheimer’s type have sleep disturbances characterized by reduced REM and slow-wave sleep. The loss of cholinergic neurons in the basal forebrain has been implicated as the cause of these changes. Melatonin secretion from the pineal gland is inhibited by bright light, so the lowest serum melatonin concentrations occur during the day. The suprachiasmatic nucleus of the hypothalamus may act as the anatomical site of a circadian pacemaker that regulates melatonin secretion and the entrainment of the brain to a 24-hour sleep-wake cycle. Evidence shows that dopamine has an alerting effect. Drugs that increase dopamine concentrations in the brain tend to produce arousal and wakefulness. In contrast, dopamine blockers, such as pimozide (Orap) and the phenothiazines, tend to increase sleep time. A hypothesized homeostatic drive to sleep, perhaps in the form of an endogenous substance—process S—may accumulate during wakefulness and act to induce sleep. Another compound—process C—may act as a regulator of body temperature and sleep duration.

The functions of sleep have been examined in a variety of ways. Most investigators conclude that sleep serves a restorative, homeostatic function and appears to be crucial for normal thermoregulation and energy conservation. Given that NREM sleep increases after exercise and starvation, this stage may be associated with satisfying metabolic needs.

Insomnia is difficulty initiating or maintaining sleep. It is the most common sleep complaint and may be transient or persistent. Population surveys show a 1-year prevalence rate of 30 to 45 percent in adults.

A brief period of insomnia is most often associated with anxiety, either as a sequela to an anxious experience or in anticipation of an anxiety-provoking experience (e.g., an examination or an impending job interview). In some persons, transient insomnia of this kind may be related to grief, loss, or almost any life change or stress. The condition is not likely to be serious, although a psychotic episode or a severe depression sometimes begins with acute insomnia. Specific treatment for the condition is usually not required. When treatment with hypnotic medication is indicated, both the physician and the patient should be clear that the treatment is of short duration and that some symptoms, including a brief recurrence of the insomnia, may be expected when the medication is discontinued.

Persistent insomnia is composed of a fairly common group of conditions in which the problem is most often difficulty falling asleep rather than remaining asleep. This insomnia involves two sometimes separable but often intertwined problems: somatized tension and anxiety, and a conditioned associative response. Patients often have no clear complaint other than insomnia. They may not experience anxiety per se but discharge the anxiety through physiological channels; they may complain chiefly of apprehensive feeling or ruminative thoughts that appear to keep them from falling asleep. Sometimes (but not always) a patient describes the condition’s exacerbation at times of stress at work or at home and its remission during vacations.

Hypersomnia manifests as excessive amounts of sleep, excessive daytime sleepiness (somnolence), or sometimes both. The term somnolence should be reserved for patients who complain of sleepiness and have a clearly demonstrable tendency to fall asleep suddenly in the waking state, who have sleep attacks, and who cannot remain awake; it should not be used for persons who are simply physically tired or weary. The distinction, however, is not always clear. Complaints of hypersomnia are much less frequent (5 percent of adults) than complaints of insomnia, but they are by no means rare if clinicians are alert to them. More than 100,000 persons with narcolepsy are estimated to live in the United States, and narcolepsy is just one well-known condition that clearly produces hypersomnia. If substance-related conditions are included, hypersomnia is a common symptom.

As with insomnia, hypersomnia is associated with conditions that are hard to classify and idiopathic cases. According to a recent survey, the most common conditions responsible for hypersomnia sufficiently severe to be evaluated by all-night recordings at a sleep disorders center were sleep apnea and narcolepsy.

Transient and situational hypersomnia is a disruption of the normal sleep-wake pattern; it is marked by excessive difficulty in remaining awake and a tendency to remain in bed for unusually long periods or to return to bed to nap frequently during the day. The pattern is experienced suddenly in response to an identifiable recent life change, conflict, or loss and is much less common than insomnia. It is seldom marked by definite sleep attacks or unavoidable sleep but, rather, is characterized by tiredness or by falling asleep sooner than usual and by difficulty arising in the morning.

Parasomnia is an unusual or undesirable phenomenon that appears suddenly during sleep or that occurs at the threshold between waking and sleeping. Parasomnia usually occurs in stages 3 and 4 and, thus, is associated with poor recall of the disturbance.

The DSM-IV-TR classifies sleep disorders on the basis of clinical diagnostic criteria and presumed etiology. The three major categories of sleep disorders in DSM-IV-TR are primary sleep disorders, sleep disorders related to another mental disorder, and other sleep disorders (due to a general medical condition or are substance induced). The disorders described in DSM-IV-TR are only a fraction of the known sleep disorders; they provide a framework for clinical assessment.

The DSM-IV-TR defines primary sleep disorders as those not caused by another mental disorder, a physical condition, or a substance but, rather, are caused by an abnormal sleep-wake mechanism and often by conditioning. The two main primary sleep disorders are dyssomnias and parasomnias. Dyssomnias are a heterogeneous group of sleep disorders that includes primary insomnia, primary hypersomnia, narcolepsy, breathing-related sleep disorder, circadian rhythm sleep disorder (sleep-wake schedule disorder), and dyssomnia not otherwise specified. Parasomnias include nightmare disorder (dream anxiety disorder), sleep terror disorder, sleepwalking disorder, and parasomnia not otherwise specified.