Sleep terrors are characterized by a sudden awakening from Stage 3/4 sleep that is accompanied by a piercing scream, intense anxiety, and excessive autonomic arousal. These signs are significantly more intense than in nightmares. The individual may sit up in bed but is usually unresponsive to the environment. Disorientation and confusion are predominant. There is no recall of any specific imagery or dreams, and there may be a total amnesia for the sleep terror episode on arising in the morning. Sleep terrors occur mostly in the first third of the night, when deep NREM sleep predominates. Their incidence is more common between ages 4 and 12 years, and they tend to resolve spontaneously during adolescence. Their persistence in adulthood has been associated with a greater likelihood of concomitant psychopathology (J. D. Kales et al. 1980). Although less frequent, sleep terrors with an onset during adulthood can be associated with other conditions such as PTSD and sleep apnea.

Sleepwalking disorder, or somnambulism, is a disorder in which the patient sits up in bed or gets out of bed during sleep and walks about in a poorly coordinated and automatic manner.

Sleep talking also may occur. Sleepwalking occurs early in the night, when deep NREM sleep predominates. It is associated with impaired arousal, and the patient does not respond to the efforts of others to communicate with or to awaken him or her. Full awakening is achieved only through persistent efforts, and confusion, disorientation, and amnesia of sleepwalking usually occur on awakening at night or the next morning. There is a genetic predisposition to sleepwalking, as indicated by increased rates among first-degree relatives. Sleepwalking not only may cause social embarrassment but also carries a significant potential for injuries (A. Kales et al. 1980). Other behaviors, such as eating or violent punching or kicking, can occur in conjunction with sleepwalking episodes. Sleep-related eating appears to be more common among women, and sleep-related violence is more common among men (Guilleminault et al. 1995b; Moldofsky et al. 1995; Schenck et al. 1993b).

Sleep terror and sleepwalking have several common features, and their co-occurrence is also quite frequent. Because they may well be part of the same phenomenon, treatment recommendations are similar and are discussed together.

Behavioral/psychotherapeutic treatment Sleepwalking can be life threatening; therefore, the first-line treatment should consist of safety measures, including sleeping on the ground floor; locking bedroom windows and doors; removing all potentially dangerous objects, including weapons; and rearranging furniture to minimize risk of injury. The second-line intervention modality involves preventive measures. Because both sleepwalking and sleep terror originate from deep NREM sleep, and because this type of sleep increases as a function of wakefulness, it is important to avoid sleep deprivation. When sleepwalking is drug induced, the schedule of medication intake should be altered, if possible (Berlin and Qayyum 1986).

Parents of children with parasomnias should be reassured of the normal developmental nature of these episodes. They should let these events run their natural course and intervene only to prevent injury and minimize sleep deprivation. In both children and adults, it is best not to force arousal on the sleepwalker or individual with sleep terror, but instead to provide reassurance, verbal instructions, and physical guidance in returning the patient to bed.

Hypnosis is one of the few nonpharmacological interventions that has received some empirical support in the treatment of sleepwalking (Reid et al. 1981) and sleep terror (Hurwitz et al. 1991). In the study by Reid et al. (1981), patients were trained to arouse as soon as their feet came in contact with the floor. The results indicated that this procedure was effective in eliminating sleepwalking in about 75% of patients after a brief 3-week intervention. No follow-up data were available.

Clinical hypnosis was also found to be useful in 20 of 26 patients with sleep terrors in a study by Hurwitz et al. (1991). The authors postulated that the benefit derived from hypnosis may result from improved self-control rather than elimination of the actual sleep terror episodes. These results are promising, but, because they are based on retrospective and uncontrolled studies, replication with long-term follow-up evaluations is needed.

Another behavioral technique for treating sleepwalking involves awakening the individual just before the typical time of sleepwalking episodes. This method has been successful in small published case series of children with sleepwalking (Frank et al. 1997; Tobin 1993). When life stressors or psychiatric symptoms play a contributory role in parasomnias, stress management and problem-solving therapy may be useful. Consistent with the notion that persistence of these parasomnias into adulthood may reflect an underlying psychiatric disorder, some authors advocate the use of insight-oriented psychotherapy aimed at helping patients gain insight into the maladaptive nature of their coping mechanisms (J. D. Kales et al. 1982). The empirical support for such recommendations, however, is derived from only two case studies in which almost 2 years of ongoing therapy was required to eliminate sleep terror episodes.

Pharmacological treatment Pharmacological treatment is often considered in individuals with sleep terror and sleepwalking disorder, particularly those with violent behavior during sleep terror episodes. Unfortunately, no recent controlled clinical studies have assessed the efficacy of pharmacological treatment approaches with appropriate placebo controls. In most cases, a benzodiazepine hypnotic medication, administered in moderate doses at bedtime, is the treatment of choice. Empirical research to support this choice is limited. An early study by Fisher et al. (1973) showed no association between suppression of Stage 3/4 sleep and suppression of sleep terror episodes. This leads to the suggestion that the action of benzodiazepines results more from the suppression of arousal than from the suppression of slow-wave sleep. A case series of individuals with combined symptoms of parasomnias, including sleepwalking, found that the benzodiazepines clonazepam and alprazolam were effective for most of those treated (Schenck et al. 1997). Case reports have also suggested that TCAs, such as imipramine, either alone or in combination with benzodiazepines, may be useful for treating night terrors. However, another case series describing the efficacy of TCAs in sleep terrors (Burstein et al. 1983) included patients with PTSD; whether these patients had NREM sleep terrors, REM nightmares, or both is unclear (Cooper 1987).

The clinical dilemma arises from deciding when to treat patients. Because sleepwalking and sleep terror episodes usually do not occur nightly, one must decide to treat the patient every night or only when episodes are more likely to occur, such as at times of psychosocial stress or sleep deprivation.