Clinical manifestations

Presentation and course

Anxiety and trauma-based disorders share common features of “fear,” the response to an actual or perceived imminent threat and “anxiety,” and the anticipation of future threat. Both fear and anxiety are unavoidable parts of the human experience and serve adaptive functions serving to motivate behaviors to avoid negative consequences. Although obviously similar, fear and anxiety differ in their expression, with more autonomic arousal and “fight or flight” response with fear as compared to restlessness and tension in anxiety.

Fear and anxiety become pathological when they no longer serve an adaptive function and instead cause significant distress or impairment of social or occupational functioning due to severity or pervasiveness. Anxiety disorders and trauma-based disorders vary in terms of types of situations or objects that provoke a pathological response and the associated behaviors and cognitions that result.

Anxiety and fear-based disorders are frequently associated with complaints of difficulty initiating and maintaining sleep, restless nonrestorative sleep, and short sleep duration (67). For example, more than 90% of combat veterans diagnosed with posttraumatic stress disorder complain of chronic difficulties with insomnia or nightmares (51). Chronic sleep difficulties are also a risk factor for the development of an anxiety disorder. Some studies have highlighted the independent and bidirectional relationships between anxiety, fear, and sleep disturbances. For example, one study of youth aged 8 to 16 years with a primary anxiety disorder evaluated the relationship between daytime assessments of affect using ecological momentary assessment (EMA) and subsequent nighttime sleep. Higher daytime positive affect and positive to negative affect ratio objectively were associated with less time in bed the subsequent night for youth with a primary anxiety disorder. In addition, more time asleep during the night was associated with more positive next-day affect (16). In midlife, women with histories of comorbid anxiety and depression, or anxiety only, sleep disturbances independently contributed to low health-related quality of life, even after controlling for psychiatric illness, menopausal status, and other potential confounders (30).

Different anxiety disorders are associated with sleep disturbance of varying presentations and severity, thus, each anxiety disorder and its associated predominant sleep profiles are presented below:

Generalized anxiety disorder. The hallmark of generalized anxiety disorder is chronic pervasive anxiety affecting multiple areas of social or occupational functioning. Patients are typically described by friends and colleagues as being "nervous" and prone to excessive worrying. They often complain that they are unable to stop worrying (ruminative thinking) when trying to sleep, resulting in difficulty initiating and maintaining sleep. Insomnia is reported in more than half of patients with generalized anxiety disorder (04). Polysomnography studies in patients with generalized anxiety disorder similarly demonstrate impairment in sleep initiation and maintenance.

Panic disorder. Panic disorder is characterized by recurrent, unexpected panic attacks leading to progressive restriction of activities in which escape may be difficult, including agoraphobia. Panic attacks are acute, self-limited episodes of severe autonomic arousal (eg, shortness of breath, palpitations, tachycardia, chest pain, sweating, lightheadedness, trembling, and paresthesias) accompanied by catastrophic thinking (eg, fear of dying or losing control). Symptoms begin abruptly, reach a peak within 10 minutes, and gradually abate, with typical episodes lasting 20 to 30 minutes. Panic attacks can occur as a part of various mental disorders, or they can be isolated events.

Sleep problems are common in panic disorder. Recurrent insomnia is reported in 67% of patients (46), and most polysomnography studies document difficulties initiating and maintaining sleep. Sleep onset problems may be particularly pronounced among panic disorder patients with high levels of anxiety sensitivity, defined as excessive fear of anxiety-related sensations (27).

Panic attacks can occur when an individual is falling or waking from sleep. Sleep complaints tend to be most severe in patients with nocturnal panic attacks, which are recurrent in 33% of patients. Nocturnal panic attacks begin with an abrupt arousal from sleep without dream mentation and are accompanied by panic attack symptoms that are similar in severity and duration to panic attacks experienced while awake, although in some patients, panic attacks may be experienced predominantly or exclusively during sleep. These nocturnal episodes are distinguishable from sleep terrors, sleep apnea, nightmares, or dream-induced arousals. In some patients, nocturnal panic attacks may lead to a phobic fear of the sleep environment or sleep itself, leading to avoidance of sleep and, consequently.

Social and specific phobia. In social phobia, there is a marked and persistent fear of one or more social situations in which the person is exposed to possible scrutiny by others. In specific phobia, there is a marked and unreasonable fear precipitated by the presence or anticipation of a specific object or situation. Sleep complaints are relatively uncommon in social phobia and specific phobias, unless the content of the specific phobia is linked to the sleeping environment. Although little systematic investigation has been conducted, one study found that as many as two thirds of a sample of treatment-seeking patients with social phobia reported insomnia symptoms, with one third falling in the moderate to severe range (56). Moreover, general and social anxiety severity contributed to insomnia severity, after controlling for confounders including depression.

Posttraumatic stress disorder. Exposure to trauma that is beyond the scope of ordinary life events and threatens the physical integrity of the self or others can lead to variable clinical phenotypes that when persistent and significantly interfering may constitute posttraumatic stress disorder. Some patients demonstrate primarily fear-based, re-experiencing, and others may predominantly experience dysphoria and emotional numbing. Individuals try to avoid internal and external reminders of the trauma and often demonstrate alterations in arousal and reactivity.

Sleep complaints are nearly universal in posttraumatic stress disorder and may be uniquely related to the severity of symptoms (07). Patients often complain of difficulties initiating and maintaining sleep, although the findings from objective overnight polysomnography studies have been conflicting. One of the more consistent findings has been disruptions of REM sleep continuity (eg, increased arousals and motor activity) and increased REM activation (eye movement density) (43). Anxiety arousals are also more common in these patients. Most of these events are REM sleep-related nightmares, although nightmares in these patients may also occur out of NREM sleep (68). Repetitive nightmares are reported in many patients and may be the presenting complaint (45). These nightmares often represent a "reliving" of the original trauma and associated emotions. This may further result in a conditioned fear toward sleep, and accompanying avoidance behaviors may develop, especially if the original trauma occurred in the sleep setting. A study using the National Comorbidity Survey Replication (NCS-R) found that respondents meeting diagnostic criteria for posttraumatic stress disorder reported more weeks per year of insomnia than individuals with adult separation anxiety, alcohol dependence, and major depression (37).

A systematic review of nightmares in psychiatric conditions suggested that the presence of nightmares increased the risk of suicidality and treatment of nightmares could possibly mitigate this risk (61).

In addition to nightmares and insomnia, posttraumatic stress disorder may be associated with sleep-disordered breathing, with one study suggesting that the rates of sleep-related breathing disorder (defined as an apnea-hypopnea index greater than 10) could be as high as 70% (71). Another study found an association between nocturnal blood pressure nondipping, an established risk factor of hypertension, which commonly occurs during nocturnal breathing events, and lifetime full and subthreshold posttraumatic stress disorder (44).

Several studies have also examined the role of sleep disturbances in the development or exacerbation of posttraumatic stress disorder symptoms (36). In a cross-sectional study of a combat team surveyed 90 to 180 days post-deployment, Luxton and colleagues found that nearly 75% of respondents slept 6 hours or less per night and, after controlling for combat exposure, short sleep was associated with symptoms of depression, posttraumatic stress disorder, panic attacks, and high-risk health behaviors (41). In another study, insomnia at 4 months post-deployment was found to be a significant predictor of change in depression and posttraumatic stress disorder at 12 months post-deployment (70). The relationship between sleep disturbances and posttraumatic stress disorder severity has also been found in youth 8 to 15 years of age who experienced a natural disaster (13). It is possible that sleep disruption and perhaps short sleep duration may interfere with healthy emotional or cognitive adaptation and regulation, contributing to the development or exacerbation of posttraumatic stress disorder and other psychiatric symptoms.

Polysomnography data. Polysomnographic findings in patients with anxiety and trauma-based disorders generally parallel subjective complaints. In generalized anxiety disorder, patients exhibit prolonged sleep latencies, decreased sleep efficiency, increased wakefulness after sleep onset, and increased stage 1 sleep (59). Patients with panic disorder demonstrate increased sleep latency and waking time as well as an increase in movement time. Nocturnal panic attacks arise during NREM sleep, especially during the transition between stage 2 and deep slow-wave sleep, most commonly during the first few hours of sleep (26). Polysomnographic studies in social and specific phobias are limited, but a small pilot study of patients with social phobia reported normal nighttime sleep (12).

Studies of patients with posttraumatic stress disorder have yielded inconsistent results, although increased arousals and greater movement in REM sleep and greater eye movement density during REM sleep have been consistently found (43). A meta-analysis concluded that these patients have more light stage 1 sleep, less deep slow wave sleep, and greater REM density compared to non-posttraumatic stress disorder control samples (35). One study found that, after controlling for relevant variables, an increased number of earlier-life traumatic events among a sample of military veterans was associated with increased REM sleep fragmentation, and this REM fragmentation related to later-life disruptive nocturnal behaviors (28).

Diagnosis of sleep disorders in patients with anxiety disorders. Patients with anxiety disorders with severe enough sleep complaints may qualify for a comorbid sleep disorder. Prior to DSM-5, “insomnia due to mental disorder” could be diagnosed. This has been replaced with “insomnia disorder” (03).

Specific criteria for diagnosis of insomnia disorder per DSM-5 include (03):

	1. A complaint of poor sleep quality or quantity associated with:
		a. difficulty initiating sleep b. difficulty maintaining sleep c. early morning waking
	2. Significant distress or interference in function 3. At least 3 nights per week for greater than 3 months 4. There is adequate opportunity for sleep 5. Insomnia is not better accounted for by another sleep disorder (eg, narcolepsy, parasomnia, etc.) 6. Insomnia is not due to a substance 7. Symptoms are in excess of what is expected from comorbid medical or mental disorders

The specific comorbid psychiatric, medical, or other sleep disorder is specified, as is the chronicity (episodic, persistent, or recurrent).

The criteria for “chronic insomnia disorder” as per the International Classification of Sleep Disorders Third Edition includes (02):

1. Problems with sleep initiation or maintenance 2. Adequate opportunity for sleep 3. Daytime somnolence or other sequelae

Symptoms must occur at least three times per week over a 3-month period. This diagnosis should only be used if there is no other cause of the diagnosis, or if the sleep problems exceed expected sleep problems expected for a related medical or psychiatric condition and are the focus of clinical management.

Prognosis and complications

The prognosis of insomnia disorder with non-sleep disorder mental comorbidity (previous diagnosis of insomnia due to mental disorder) has not been specifically studied. It often waxes and wanes with the severity of the comorbid anxiety or trauma-based disorder, which is often a chronic condition. The course of panic disorder is variable, with 30% to 40% of patients being symptom free, and 10% to 20% continuing to have significant symptoms at long-term follow-up. Good premorbid occupational and social functioning and the presence of a clear precipitating event are good prognostic features. In posttraumatic stress disorder, more than one-third of patients remain symptomatic after many years (34). Extremely young or old patients tend to have a worse prognosis than those with rapid onset, short duration, and good premorbid function, and those with no history of substance abuse. Comorbid depression and substance abuse are common. Generalized anxiety disorder is often a lifelong condition, with only 38% demonstrating full remission after 5 years (72). Comorbidity with depression, other anxiety disorders, and substance abuse is high. When insomnia is specifically treated in the context of an anxiety disorder, it may resolve more quickly than other symptoms. On the other hand, insomnia can frequently persist despite resolution of the primary psychiatric disorder, suggesting that an alternate insomnia diagnosis and independent treatment may be indicated.

The available data indicate that insomnia improves following cognitive-behavioral treatment for anxiety disorders, even though it is not targeted in treatment. In a meta-analysis, Belleville and colleagues found that the effect size of cognitive behavioral therapy for anxiety disorders on sleep problems was large for generalized anxiety disorder (0.849) and small to moderate for panic disorder with agoraphobia (0.216) and posttraumatic stress disorder (0.404), respectively (05). Despite these improvements, the majority of patients with anxiety disorder will continue to experience residual sleep disturbances following treatment (05). In one study, cognitive-behavioral therapy for posttraumatic stress disorder improved self-reported sleep quality, but as many as 70% of patients who reported baseline sleep difficulties continued to experience sleep problems following treatment. These individuals were more likely to report less improvement in posttraumatic, anxious, and depressive symptoms relative to those who reported no residual sleep problems (08). Similarly, sleep-focused interventions have only a small to moderate impact on associated anxiety symptoms in patients with and without accompanying anxiety disorders (06). These findings suggest that patients with comorbid insomnia and anxiety disorders are likely to benefit most from a combined approach to therapy.

Clinical vignette

A 25-year-old white male presented with a 1-year history of severe sleeping difficulty. He initially experienced episodes of abrupt awakening from sleep associated with a feeling that he was about to die. He would also experience shortness of breath, palpitations, and sweating. Symptoms would worsen over several minutes and then abate over the next 10 minutes. Symptoms initially occurred only during sleep, but after several months he began to experience occasional daytime symptoms that were similar to the nocturnal episodes. As the frequency of nocturnal episodes increased, he began to fear going to sleep. An extensive medical evaluation by his internist revealed no significant abnormalities. On presentation to the sleep center, he was experiencing severe difficulty initiating and maintaining sleep and would only attempt to sleep in the daytime in a sitting position while watching TV. Attempting to lie in bed elicited severe anxiety. In addition to avoiding the bedroom environment, he was beginning to avoid crowded public places. A polysomnogram revealed a marked increase in sleep latency and decreased sleep efficiency. He was diagnosed as having panic disorder with agoraphobia. Initial treatment consisted of escitalopram, which, after a month, led to dramatic reduction in the number of panic attacks. However, the patient still experienced severe anxiety when attempting to lie in bed. The patient underwent six sessions of cognitive-behavioral therapy for insomnia, with marked symptomatic improvement. An attempt was made to taper the escitalopram after 9 months, but with a recurrence of symptoms. He subsequently remained symptom-free on escitalopram.

Biological basis

Etiology and pathogenesis

The etiology of sleep disturbances in anxiety and trauma-based disorders is not well understood. Patients often report excessive rumination and difficulty "turning off" their minds, leading to insomnia. Many patients with anxiety disorders and posttraumatic stress disorder may have elevated levels of physiologic arousal that may directly interfere with sleep mechanisms. The autonomic activation or psychological distress of nocturnal panic attacks or nightmares in posttraumatic stress disorder may directly fragment sleep and lead to conditioned fears.

Genetic factors clearly contribute to vulnerability to anxiety disorders, with nearly half of all patients with panic disorder having an affected relative (31). In generalized anxiety disorder, 19.5% of patients have an affected first-degree relative, compared to 3.5% of control patients (52). Twin studies have confirmed the genetic component with monozygotic twins having a higher concordance rate than dizygotic twins (29). Some predisposing genes have been identified such as the serotonin transporter gene; the STin2.12 allele is found more frequently in patients with generalized anxiety disorder, obsessive-compulsive disorder, and depression than in control subjects (53). Subsequent twin studies in pairs aged 8 to 16 years have identified significant overlap between the genetics of insomnia and the genetic underpinnings associated with anxiety and depression (23).

Environmental factors must also play a role in the precipitation and maintenance of anxiety disorders. Negative life events may be precipitants in many anxiety disorders, a concept supported by the finding that patients with generalized anxiety disorder report a history of trauma involving injury or death at a greater rate than control subjects. Once precipitated, patients demonstrate abnormalities of information processing, which may help maintain the disorder (01). Patients with generalized anxiety disorder display selective attentional biases, preferentially focusing on topics with potential threat, and on potentially threatening interpretation of ambiguous situations. The persistent worry in these patients could possibly serve an adaptive function by preventing more severe distress that would occur without the worry.

A number of biological abnormalities have been documented in patients with anxiety disorder. Some patients with anxiety have demonstrated increased sympathetic tone, slow adaptation to repetitive stimuli, and excessive responsiveness to stimuli (31). Abnormal GABAergic and serotoninergic mechanisms have been documented in patients, as well as abnormalities in the regulation of norepinephrine, corticotropin-releasing factor, and neurosteroids (29). Functional imaging studies have demonstrated an increase in cortical activity and decreased basal ganglia activity in patients with generalized anxiety disorder that is reversed after treatment (29). Imaging studies provide consensus on the crucial role of the amygdala, anterior cingulate cortex, and the insula in the pathophysiology of anxiety and trauma-based disorders (48). Debate exists over whether these biological changes are in response to psychological conflicts, or whether the biological abnormalities precede and predispose to the psychological abnormalities. It is likely that both mechanisms are operative.

Sleep problems may promote substance use. Cannabis use is frequently used by patients with posttraumatic stress disorder with the goal of improving sleep (10). Alcohol is also commonly used to initiate sleep; however, it can lead to disrupted sleep and addiction.

Epidemiology

Anxiety disorders are one of the most common psychiatric disorders, with nearly one in three individuals suffering from an anxiety disorder in their lifetime. Patients with anxiety disorders also frequently meet criteria for comorbid psychiatric disorders, such as mood and substance use disorders. The lifetime prevalence rate in women is 30.5%, compared to 19.2% in men. The prevalence of anxiety disorders decreases with increasing socioeconomic status. The most current estimates of prevalence rates of anxiety disorders come from the National Comorbidity Survey Replication (NCS-R). Lifetime prevalence of panic disorder is estimated to be 4.7% (32). Women are two to three times more likely to experience panic disorder; onset is usually late adolescence or early adulthood. Generalized anxiety disorder has a lifetime prevalence of 5.7%. Onset is typically from the teens through late 20s and is more common among women, unmarried individuals, and minority groups (33). There have been varying estimates of the prevalence of trauma-based disorders, such as posttraumatic stress disorder, but the NCS-R study found a lifetime prevalence rate of 6.8%. Women have higher rates than men (34). Although posttraumatic stress disorder can occur at any age, it is most common in young adults, reflecting the high incidence among male combat veterans and female rape victims. Specific and social phobias have the highest lifetime prevalence rates of the anxiety disorders, estimated at 12.5% and 12.1% respectively in the NCS-R. Age of onset for social phobia is often in early adolescence, and the disorder can contribute to associated academic difficulties.

An analysis of data from the Primary Care Anxiety Project found that, after controlling for comorbid major depressive disorder and anxiety disorder, sleep disturbance at intake predicted the course of posttraumatic stress disorder but not generalized anxiety disorder, social phobia, or panic disorder (42). More specifically, only one-third of posttraumatic stress disorder subjects with sleep disturbances at intake remitted at 5-year follow-up, compared to more than half of those without sleep problems at baseline.

Little information is available on the epidemiology of insomnia disorder. This type of insomnia is more common among women than men and appears more frequently among middle-aged adults than among adolescents, young adults, and the elderly.

Prevention

There is insufficient knowledge about the predisposing and precipitating factors of most anxiety disorders to allow reliable prevention. Early treatment and education of anxiety disorders may help prevent comorbidities such as depression, substance abuse, poor sleep hygiene, conditioned fear of the sleep environment, and associated sleep disturbances.

Differential diagnosis

The primary differential diagnosis for insomnia disorder with comorbid mental disorder is other causes of insomnia. Primary sleep disorders such as restless legs syndrome, periodic limb movement disorder, sleep related breathing disorders, or circadian rhythm abnormalities may coexist with anxiety disorders and contribute to insomnia. Other medical causes of insomnia may coexist with anxiety disorders, including pain disorders, prostatic hypertrophy, nocturnal asthma, or cardiopulmonary disease. Hyperthyroidism may worsen anxiety and insomnia. The sleep disturbance of anxiety disorders must also be distinguished from those associated with common comorbid conditions such as substance abuse, inadequate sleep hygiene, or depression. The anxiety arousals in posttraumatic stress disorder and panic disorder must be distinguished from other causes of abrupt arousal such as sleep terrors, sleep choking syndrome, gastroesophageal reflux, sleep apnea, cardiac and pulmonary disorders, and seizures. Medications used to treat anxiety disorders, such as SSRIs, can result in insomnia, either directly through their activating effects or indirectly through worsening of restless legs syndrome or periodic limb movement disorder.

Diagnostic workup

A full psychiatric diagnostic interview and physical examination is performed to understand the history of the complaints and to screen for comorbid conditions such as poor sleep hygiene, substance use disorders, or mood disorders. Symptoms and signs of medical or neurologic conditions that may be associated with the underlying anxiety disorder or insomnia are also sought. Diagnostic tests may be indicated to rule out common medical causes of anxiety. The patient should be questioned about symptoms of other specific sleep disorders, such as sleep apnea, restless legs syndrome, or periodic limb movement disorder. Bedmates should be interviewed when possible. Polysomnography is not routinely indicated in uncomplicated cases where the diagnosis is clear by history. In cases where the history suggests the possibility of an additional sleep disorder such as obstructive sleep apnea (OSA), or in cases associated with significant daytime sleepiness, polysomnography may be indicated. Polysomnography, often with specialized monitoring (for instance, full montage EEG if seizures are suspected), may be necessary in some cases where the etiology of anxiety arousals at night is unclear.

Management

The management of insomnia in the context of another anxiety or trauma-based disorder depends in large part on effective treatment of the associated condition. Cognitive behavioral therapy to correct the cognitive distortions and to teach relaxation or biofeedback techniques are helpful in patients with generalized anxiety disorder (20). Newer approaches to generalized anxiety disorder include integrating mindfulness meditation with cognitive behavioral therapy and “metacognitive” approaches. A study of a group exercise intervention improved anxiety and sleep quality as compared to passive controls (75).

Cognitive behavioral therapy in panic disorder has equivalent effectiveness to pharmacotherapy with more durable benefits. The cognitive component is directed at false beliefs about panic attacks, particularly reinforcing the knowledge that they are transient and not life-threatening. Relaxation techniques may help reduce the distress associated with panic episodes. An important behavioral component of therapy involves interoceptive or in vivo exposure to perceived threatening stimuli, with the goal of habituation to panic symptoms in their presence. Exposure therapy is also the preferred treatment for specific phobias.

A variety of cognitive behavioral techniques has proven beneficial in patients with posttraumatic stress disorder. Supportive psychotherapy may be a helpful component of the treatment plan for any of the anxiety disorders. Targeting the psychotherapy toward the specific underlying disorder may be more effective in improving sleep in posttraumatic stress disorder than supportive therapy (69).

Pharmacological therapy is often a component of treatment for anxiety disorders. In generalized anxiety disorder, numerous studies have demonstrated the efficacy of various benzodiazepines (11), but current guidelines indicate that SSRIs, in particular escitalopram, paroxetine, and sertraline, should be first-line pharmacological treatments (60). Venlafaxine and duloxetine are alternatives, and tricyclic antidepressants (eg, nortriptyline) may also be beneficial. Despite FDA approval, some reports conclude that buspirone is not well-established as a monotherapy for generalized anxiety disorder. In panic disorder, SSRIs appear to be particularly effective, although the activating effects of SSRIs can transiently worsen some symptoms of anxiety. Data also support the efficacy of the extended-release form of the SNRI venlafaxine. Tricyclic antidepressants, especially nortriptyline, are also effective but are second-line treatments. If rapid onset of action is needed, alprazolam or lorazepam have been shown to be effective, but these are considered second-line or adjunctive treatments only due to the risk of dependence and effects on cognition. Beta-blockers and SSRIs/SNRIs may be useful for social phobia. Beta-blockers and SSRIs/SNRIs may be useful for social phobia. Medications effective for posttraumatic stress disorder include SSRIs and SNRIs.

Certain medications such as theophylline and beta-agonists may worsen anxiety, and alternate agents should be used when possible; caffeine and nicotine may worsen anxiety and precipitate panic attacks and should be avoided. Sleep hygiene should be taught to all patients. Judicious use of sedative hypnotics may be appropriate in many patients but should be avoided in patients who have any history of substance use.

Studies have found that pharmacological and nonpharmacological treatments targeting sleep disturbances in patients with anxiety disorders improve both sleep and anxiety symptoms. In a multicenter randomized placebo-controlled trial, coadministration of eszopiclone 3 mg and escitalopram 10 mg improved anxiety symptoms more after 8 weeks than placebo and escitalopram in patients with generalized anxiety disorder (55). Three weeks of eszopiclone 3 mg monotherapy was also found to improve both sleep quality and posttraumatic stress disorder symptoms more than placebo in 24 patients with posttraumatic stress disorder (54). An open-label study in partial responders with generalized anxiety disorder found ramelteon 8 mg at bedtime improved subjective sleep quality and reduced anxiety symptoms after 10 weeks of treatment (25).

Pilot studies of a brief behavioral treatment for sleep disturbances in posttraumatic stress disorder (24) and for mindfulness-based cognitive therapy targeting insomnia in generalized anxiety disorder and panic disorder (73) have been promising but also require more systematic evaluation. A randomized controlled trial showed that a digital administration of cognitive behavioral therapy may be effective in treating patients with generalized anxiety disorder, improving anxiety symptoms, including insomnia (14). Cognitive behavioral therapy specifically for insomnia (CBTi) has been found to be efficacious for improving sleep as well as daytime symptoms of anxiety, mood, and overall functioning (09).

Other studies have evaluated therapies that address both sleep disturbances and accompanying symptoms, such as nightmares. Prazosin 3 mg nightly for 8 weeks improved posttraumatic stress disorder symptoms, subjective sleep quality, and trauma nightmare frequency more than placebo in combat veterans (58). A follow-up study with in-home polysomnography found that 3 weeks of prazosin 3 mg increased objective total sleep time and REM sleep time and reduced REM sleep latency more than placebo (64). A moderately-sized randomized control trial of prazosin for combat posttraumatic stress disorder failed to demonstrate improvement in nightmares or sleep quality (57). Although these studies of prazosin were discouraging, subsequent papers suggest that further data are needed. A meta-analysis of imagery rehearsal therapy and prazosin did not show significant differences in comparative efficacy (74). Another systematic review and meta-analysis of prazosin demonstrated an improvement in nightmares but not in overall posttraumatic stress disorder symptoms, suggesting continued use of prazosin in this condition (76).

Non-medication approaches, such as imagery rehearsal therapy (40; 50; 66) exposure, relaxation, rescripting treatment (19; 63), and imagery rescripting and exposure therapy (39) have been evaluated for posttraumatic stress disorder-related nightmares, but few of these treatments have been rigorously evaluated in veteran populations (49). Response to some of these non-medication nightmare treatments has been associated with physiological changes in heart rate, skin conductance, and corrugator activity (18). In a randomized controlled trial, Cook and colleagues compared a 6-week imagery rehearsal therapy to a credible psychotherapy control in 124 male Vietnam War veterans with chronic, severe posttraumatic stress disorder (15). The results indicated no substantive improvements in sleep quality, posttraumatic stress disorder symptoms, or nightmare frequency after 1, 3, and 6 months. The efficacy of these therapies in controlled trials, therefore, requires further study. Pilot studies of a brief behavioral treatment for sleep disturbances in posttraumatic stress disorder (24) and for mindfulness-based cognitive therapy targeting insomnia in generalized anxiety disorder and panic disorder (73) have been promising but also require more systematic evaluation.

Although the choice for medication or psychotherapy may depend on different factors, a systematic review of interventions for posttraumatic stress disorder suggested that sleep quality is significantly improved by psychological interventions as compared to controls and are more robust than the reductions with medication (17).

Treatment of comorbid sleep issues is also critical. Untreated obstructive sleep apnea is associated with increased posttraumatic stress disorder severity and decreased response to cognitive processing therapy (47).

Special considerations

Pregnancy

Anxiety symptoms are often most severe premenstrually; although pregnancy can precipitate anxiety, symptoms tend to be particularly severe in the postpartum period, with an incidence of 10.3% between 14 and 30 weeks postpartum (62). Anxiety during pregnancy may have an adverse impact on pregnancy outcome (38).