Insomnia and Anxiety (Series in Anxiety and Related Disorders) (10 page)

vative time-in-bed restriction) and PTSD (i.e., addressing nightmares). We review

the evidence that treating the comorbid anxiety disorder only (and thus ignoring the

insomnia) may limit the degree of anxiety disorder treatment response and/or result

in residual insomnia, and the evidence for a combined approach (i.e., treating the

insomnia and anxiety problem concurrently).

Now that we have discussed the ways in which insomnia and anxiety disorders are

assessed, we can discuss the different manifestations (both subjective and objec-

tive) of sleep disturbance within anxiety disorders. We discuss the associated fea-

tures of the anxiety disorder and focus especially on those anxiety conditions

presumed to have higher comorbidity with insomnia. We also consider whether

there are any disorder-specific special insomnia treatment considerations within

each diagnostic category.

Panic Disorder

Panic Disorder (PD) is a disorder characterized by discrete episodes of intense fear

or terror (American Psychiatric Association, 1997). Symptoms in a panic attack

include autonomic symptoms of dizziness, choking, palpitations, trembling, chest

discomfort, parathesias, chills, hot flashes, stomach upset, and sweating. Panic

attacks often include cognitive–emotional symptoms such as fear of dying, losing

control, or going “crazy,” as well as derealization (e.g., a sense that things are

C.E. Carney and J.D. Edinger,
Insomnia and Anxiety
, Series in Anxiety and Related Disorders,

33

DOI 10.1007/978-1-4419-1434-7_3, © Springer Science+Business Media, LLC 2010

34

3 Anxiety Disorders and Accompanying Insomnia

unreal) or depersonalization (e.g., a sense of detachment from oneself). The first set

of DSM-IV-TR diagnostic criteria for Panic Disorder requires the occurrence of (1)

recurrent unexpected panic attacks and one of the following: (a) persistent fear of

having more attacks, (b) persistent fear of the implications of the attacks (e.g., “I

may go crazy” or “I may die”), and/or (c) a significant change in behavior related

to the attacks (e.g., avoidance of situations wherein panic attacks may occur)

(American Psychiatric Association, 1997). Lastly, the panic attacks should not be

better accounted for by the physiological effects of a medical condition or sub-

stance, or another mental disorder (e.g., such as the panic attacks that can occur in

Social Phobia). Panic Disorder can occur with or without agoraphobia.

In addition to insomnia, there are many comorbidities associated with PD

including depression, agoraphobia, and alcohol, sedative or hypnotic abuse

(American Psychiatric Association, 1997). The overlap with other disorders has

clouded polysomnographic investigations of sleep in those with PD, as some of

the reported sleep problems may relate to comorbid depression rather than PD

(Stein, Enns, & Kryger, 1993). Generally, PSG has verified the subjective report

of poor sleep in PD sufferers. For example, PSG has revealed a decreased sleep

efficiency (i.e., the percentage of time spent asleep while in bed) and the total

minutes spent asleep (Mellman & Uhde, 1989), and increased movement time in

those with PD (Brown & Uhde, 2003). Interestingly, movement time is decreased

on nights in which there is a nocturnal panic (NP) attack (Brown & Uhde, 2003).

This has led to the speculation that NP might occur in reaction to greater relax-

ation (as evidenced by the perception of decreased movement) in those prone to

relaxation-induced anxiety. Those with PD are often characterized by relaxation-

induced anxiety (Craske, Lang, Tsao, Mystkowski, and Rowe, 2001), thus the

perception of less movement and increased relaxation may be a possible trigger

for some with NP.

Nocturnal panic occurs in up to 70% of those with panic disorder, although only

about half of these cases experience nocturnal panic on a regular basis (Craske &

Barlow, 1989; Mellman & Uhde, 1989). The CBT model of nocturnal panic is

essentially the same as it is for PD. That is, panic attacks are postulated to occur in

response to physiologic changes in an individual fearful of particular bodily symp-

toms. Normal physiologic changes in breathing, heart rate, or muscle activity dur-

ing sleep are perceived as a possible problem producing an arousal and subsequent

panic attack. Those with nocturnal panic appear to engage in catastrophic thinking

about bodily sensations, a process hypothesized to rouse them out of sleep (Craske,

Lang, & Rowe, 2002). While it is intriguing to think that those with NP represent

a more severe form of PD, there is little evidence to support this contention (Craske,

Lang, Mystkowski et al., 2002).

There are limited data regarding the efficacy of nocturnal panic treatments.

Some data support the use of antidepressant medications (Mellman & Uhde, 1990)

or single-agent anxiety medications such as alprazolam (Cameron & Thyer, 1985).

Cognitive Behavioral Therapy (Craske, Lang, Aikins, & Mystkowski, 2005) for

nocturnal panic appears to be an effective treatment, although randomized

controlled clinical trials are needed to validate these preliminary efficacy findings.

Generalized Anxiety Disorder

35

There are no currently published trials of CBT for treating the insomnia in PD suf-

ferers. Specific treatment of the sleep disturbance may be needed, since Cervena

et al. reported that conventional therapy of PD in 20 subjects was not sufficient to

treat the coexisting insomnia (Cervena, Matousek, Prasko, Brunovsky, & Paskova,

2005). Smith and colleagues raise an interesting concern with cognitive behavioral

insomnia therapy in this population (Smith, Huang, & Manber, 2005). They argue

that the sleep restriction component could potentially precipitate daytime PD, and

thus, this component should be used with caution in patients suffering from this

condition. There are a few studies supporting the concern that partial sleep depriva-

tion lowers PD thresholds (Mellman & Uhde, 1989; Roy-Byrne, Uhde, & Post,

1986), so clinicians may want to restrict the time spent in bed to a lesser extent in

those with frequent NP. An abbreviated treatment protocol for NP is more thor-

oughly presented in Chap. 8.

Generalized Anxiety Disorder

Generalized Anxiety Disorder (GAD) is a disorder characterized by excessive anxi-

ety and pervasive worry. These symptoms are reflected in the DSM-IV-TR criterion

A, which specifies excessive anxiety and worry about a number of matters

(American Psychiatric Association, 1997). The worry and anxiety must be frequent

(i.e., on more days than not for at least 6 months). Criterion B indicates that the

experience of the worry is difficult to control. criterion C stipulates the cooccur-

rence of at least three related symptoms, including insomnia, restlessness, fatigue,

impaired concentration, irritability, and muscle tension. GAD and insomnia are

intricately tied in the literature perhaps because of the apparent frequency of worry

present in insomnia (Borkovec, 1982; Harvey & Greenall, 2003) and the frequency

of insomnia in GAD (Culpepper, 2002). In addition, the list of symptoms in crite-

rion C (e.g., restlessness; easily fatigued; difficulty concentrating; irritability;

muscle tension; insomnia) overlap considerably with the day and nighttime symp-

toms of insomnia (Edinger et al., 2004). One key way to distinguish GAD from an

insomnia sufferer who is worried about his/her sleep and daytime functioning is

that criterion D, which state that “the focus of the anxiety and worry, is not confined

to features of an Axis I disorder” (p. 436); thus, if the worry is confined to insom-

nia-specific topics such as sleep and daytime functioning, an insomnia diagnosis

may be more appropriate.

Since sleep disturbance is a symptom in the diagnostic criteria of GAD

(American Psychiatric Association, 1997), it is no surprise that people with GAD

often have subjective sleep complaints. These complaints have also been found on

objective sleep measures. When compared with healthy controls, polysomno-

graphic investigations in those with GAD have generally found evidence of mark-

edly disrupted and fragmented sleep (Monti & Monti, 2000). For example, there is

evidence of increased sleep onset latency, reduced sleep efficiency, increased

amounts of “light” stages of sleep (i.e., 1 and 2 NREM sleep), reduced “deep” sleep

36

3 Anxiety Disorders and Accompanying Insomnia

(i.e., SWS/delta sleep), and increased frequency and duration of awakenings

(Papadimitriou, Kerkhofs, Kempenaers, & Mendlewicz, 1988; Reynolds III, Shaw,

Newton, Coble, & Kupfer, 1983; Rosa, Bonnet, & Kramer, 1983). In addition, there

is a correlation between ratings of anxiety and polysomnographic indices of sleep

disruption including the number of awakenings, the latency to stage 1 sleep, and the

percentage of stage 2 sleep (Rosa et al., 1983). As noted by Fuller and colleagues,

polysomnographic investigations of those with anxiety disorders sometimes can be

confounded by factors such as comorbidity and the duration of their anxiety prob-

lems (Fuller, Waters, Binks, & Anderson, 1997). Studies designed to tease apart

these potential confounds have compared those with high worry and no Axis I

disorder to those with low worry and no Axis I disorder. Results of this research has

shown that those with high levels of trait worry evidenced longer sleep onset laten-

cies, decreased slow wave sleep, and more frequent transitions into 1 NREM sleep

than did those with low levels of worry. These results suggest that the polysomno-

graphic findings of delayed sleep onset and decreased sleep continuity/depth

reported in those with GAD are not attributable to comorbid problems or adjust-

ments to longstanding psychopathology. These results in concert with other similar

findings (Gross & Borkovec, 1982) suggest that increased worry can disrupt sleep

and sleep disruption can increase worry.

There are no published trials that specifically test CBT for insomnia in those

with GAD. Perhaps, such studies have not been conducted since one study found

that residual insomnia may be less of an issue in those successfully treated with

Cognitive Behavior Therapy for GAD (Belanger, Morin, Langlois, & Ladouceur,

2004). If GAD treatment successfully addresses insomnia, it suggests that insomnia

treatment would be unnecessary. However, one caveat to this study is that the

sample included in this previous trial was relatively mild with respect to insomnia

severity, as the mean Insomnia Severity Index scores were below the suggested

clinical cutoff for a probable insomnia diagnosis. Another caveat is that there was

no follow-up period to assess whether sleep improvements were maintained after

treatment. Interestingly, this study found that the severity of the GAD symptoms

was not significantly related to the severity of insomnia symptoms suggesting at

least some independence of the two syndromes. This picture becomes more com-

plicated by adding results from a large scale combined medication trial for GAD

that suggested treating insomnia concurrently with GAD (escitalopram plus eszopi-

clone) produces greater (and faster) anxiety reduction than treating GAD alone

(escitalopram only) (Pollack et al., 2008). In the dual therapy group, the improve-

ment (i.e., change scores) on sleep onset latency and wakefulness after sleep onset

were approximately double that of the monotherapy (treating GAD alone). The

number of patients with a clinically significant treatment response (i.e., a 50% or

greater decline in anxiety scores at posttreatment) was 10% higher in the combined

therapy compared to mono-therapy that targeted GAD alone. Although, anxiety

improvements were maintained into the follow-up period, sleep improvements

were not. Most participants (75%) had insomnia severity index scores that were

above the clinical range for insomnia. Hence, the study sample for this trial was a

more severely sleep disturbed group than the sample included in the Belanger et al.

Posttraumatic Stress Disorder: PTSD

37

(2004) trial. Unfortunately, these two trials are the only published treatment studies

concerning insomnia GAD. It would appear that dual therapy that targets both sleep

and anxiety would be the treatment of choice in those with clinically significant

insomnia (Monti & Monti, 2000). In mild insomnia cases, a CBT for GAD inter-

vention may be sufficient, although it is unknown if sleep improvements are main-

tained after treatment. CBT for insomnia has more durable treatment effects than

insomnia-focused pharmacologic interventions (Morin et al., 2006), so it would be

interesting to test a treatment for GAD combined with CBT for insomnia against a

GAD mono-therapy.

Posttraumatic Stress Disorder: PTSD