Creating a Professional Resumé

Psychopathology_History_Diagnosis_and_Empirical_Fo..._----_Pages_109_to_144.pdf

Home >Psychology homework help >Creating a Professional Resumé

ASSESSMENT 93

are summed to produce a total score ranging from 0 to 40. In most instances, scores of 0 to 7 represent subclinical OCD symptoms, those from 8 to 15 represent mild symptoms, scores of 16 to 23 relate to moderate symptoms, scores from 24 to 31 suggest severe symptoms, and scores of 32 to 40 imply extreme symptoms.

The Y-BOCS is sensitive to multiple aspects of OCD severity independent of the number or types of different obsessions and compulsions. A limitation, however, is that the symptom checklist contains some items that are not genuine obsessions or compulsions (e.g., hair pulling). Research indicates that the Y-BOCS possesses adequate reliability, validity, and sensitivity to treatment (Taylor, Thordarson, & Sochting, 2002). Conelea, Freeman, and Garcia (2012) have described ways to use the Y-BOCS that are consistent with recent conceptualizations of the links between obsessions and compulsions.

The Brown Assessment of Beliefs Scale (BABS; Eisen et al., 1998) is a seven-item interview assessing the degree of insight into the senselessness of OCD symptoms. The individual’s main obsessional fear (e.g., “If I do not perform a prayer ritual, my mother will die”) is rated along the following parameters of insight: (1) conviction (that the belief is accurate), (2) perception of others’ views (about the fear’s accuracy), (3) explanation for any difference between the patient’s and others’ views of the belief, (4) whether the person could be convinced that the belief is wrong (i.e., fixity), (5) attempts to disprove beliefs, (6) whether the person recognizes that the belief has a psychological cause, and (7) an optional item assessing delusions of reference. Each item is rated from 0 to 4, with higher scores indicating poorer insight; the first six items are summed to create a total score (range, 0 to 24). The seventh item is not included in the total score because referential thinking is characteristic of some disorders but not others. The BABS has strong internal consistency (α = .87), strong interrater and test-retest reliability, and good convergent and discriminant validity.

SELF-REPORT MEASURES

An array of self-report and interview measures has been developed to assess OC symptoms. These include the Obsessive Compulsive Inventory and its revision (OCI-R; Foa et al., 2002), the Padua Inventory (PI; Sanavio, 1988) and its revision (PI-R; Burns, Keortge, Formea, & Sternberger, 1996), and the Vancouver Obsessive Compulsive Inventory (VOCI; Thordarson et al., 2004). These instruments measure the wide range of possible obsessions and compulsions and generally contain items assessing specific and quintessential types of obsessions and compulsions (e.g., “I feel that there are good and bad numbers”; Foa et al., 2002). The items are rated on a Likert-type scale of agreement, personal relevance, or associated distress.

Although widely used, many self-report measures of OCD have drawbacks that detract from their ability to provide a time-efficient, empirically consistent, and conceptually clear assessment of OCD symptom severity. One limitation is that relative to respondents with fewer (or more circumscribed) types of obsessions and compulsions, those with multiple types of symptoms will endorse a greater number of scale items, and therefore (all else being equal) obtain more severe scores. Thus, many OCD symptom measures confound severity with the range of symptoms present. Actual OCD severity, however, is independent of the number of different types of symptoms reported (e.g., McKay et al., 2004). For example, one patient might be singly obsessed with contamination yet be severely impaired by fear, avoidance, and compulsive washing rituals, whereas another

Psychopathology : History, Diagnosis, and Empirical Foundations, edited by Linda W. Craighead, et al., John Wiley & Sons, Incorporated, 2013. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/ashford-ebooks/detail.action?docID=1380177. Created from ashford-ebooks on 2021-10-31 19:43:32.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

94 OBSESSIVE-COMPULSIVE DISORDER

patient might have sexual, religious, and violent obsessions, as well as mental rituals, yet experience only mild fear, avoidance, and impairment.

The heterogeneous and idiosyncratic nature of obsessions and compulsions also challenges authors of self-report scales to pick and choose which symptoms to include in scale items (a scale including items to match all possible OCD symptoms would be extremely long, and many items would not apply to any given respondent). As a result, respondents whose obsessions and compulsions happen to match those chosen by authors to be represented by items on the measure will appear more severe than those whose symptoms do not match. For example, the PI-R contains an abundance of items related to obsessions about harm; thus individuals with more harming obsessions (as opposed to symmetry obsessions, which are not assessed on the PI-R) will have higher scores than those with fewer harming obsessions. Often, obsessions that are uncommon (e.g., obsessional fear of developing schizophrenia) and rituals that are covert (e.g., mental rituals) are underrepresented on self-report measures. Thus, such scales also confound severity with the level of obscurity of the respondent’s obsessions and compulsions.

Another limitation is that existing self-report measures of OCD contain a one- dimensional assessment of severity. For example, on the VOCI, respondents indicate how much each symptom-based item is “true of you” (e.g., “not at all” to “very much”). Similarly, on the OCI-R, respondents rate their level of “distress” associated with 18 different types of obsessions and compulsions. OCD symptom severity, however, is multidimensional; it consists of parameters such as distress, functional interference, and the frequency or duration of obsessions and compulsions. The use of a single rating of item relevance, severity, or distress might account for the fact that some existing self-report measures of OCD symptoms show strong correlations with measures of general anxiety and depression.

A fourth weakness of existing self-report OCD measures is that these instruments assess obsessions separately from compulsions, thus treating these symptoms as dis- connected phenomena. As indicated earlier, however, structural analyses indicate that OCD psychopathology does not distill neatly into obsessions and compulsions, but rather into dimensions characterized by both obsessions and compulsions. Relatedly, avoidance behavior is not adequately captured on most OCD self-report measures. The general fail- ure to account for avoidance, however, leads to underestimating OCD symptom severity because avoidance is often used in place of compulsive rituals to reduce obsessional anxiety (e.g., avoiding public bathrooms in place of lengthy compulsive washing rituals). Thus, individuals with OCD who do not endorse many compulsions might have severe avoidance strategies that, while an important (and functionally debilitating) element of their OCD symptom picture, are not captured on existing rating scales. Finally, with the exception of the PI and PI-R, self-report OCD measures include items assessing hoarding, which appears to be distinct from OCD. As a result, these measures are inconsistent with the most up-to-date empirically derived structural framework of OCD symptoms. More- over, such measures are likely to overestimate OC symptom severity among individuals with hoarding behaviors.

To address many of the aforementioned limitations, Abramowitz et al. (2010) developed the Dimensional Obsessive-Compulsive Scale (DOCS), a 20-item self-report measure that assesses the severity of the four most consistently replicated OCD symptom dimensions (which correspond to four DOCS subscales): (1) contamination, (2) responsibility for harm

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

ASSESSMENT 95

and mistakes, (3) symmetry/ordering, and (4) unacceptable thoughts. To accommodate the heterogeneity of OCD symptoms and the presence of obsessions and rituals within each symptom dimension, each subscale begins with a description of the symptom dimension along with examples of representative obsessions and rituals. The examples clarify the form and function of each dimension’s fundamental obsessional fears, compulsive rituals, and avoidance behaviors. Within each symptom dimension, five items (rated 0 to 4) assess the following parameters of severity (over the past month): (1) time occupied by obsessions and rituals, (2) avoidance behavior, (3) associated distress, (4) functional interference, and (5) difficulty disregarding the obsessions and refraining from the compulsions. The DOCS subscales have excellent reliability in clinical samples (α = .94 to .96) and the measure converges well with other measures of OCD symptoms (Abramowitz et al., 2010).

FUNCTIONAL ASSESSMENT

Functional assessment is the compiling of detailed patient-specific information about the antecedents and consequences of target behaviors and emotions, usually for the purposes of developing a treatment plan. Behavioral or cognitive-behavioral theory dictates what information is collected and how it is organized to form a conceptualization of the problem that will drive therapeutic intervention (usually cognitive-behavioral treatment). We provide a framework for the functional assessment of OCD in the following paragraphs.

Assessment of Obsessional Stimuli

This includes compiling a comprehensive list of the external and internal stimuli that evoke obsessional fear. External triggers include objects, situations, places, and the like that give rise to obsessional thinking and urges to ritualize (e.g., public bathrooms, knives, churches, the number 13, leaving the house, and driving in certain places). Examples of questions to elicit this information include:

• What kinds of situations make you feel anxious?

• What kinds of things do you avoid?

• What triggers you to want to do rituals?

Internal obsessional stimuli include recurring ideas, images, doubts, and impulses that the individual finds unwanted, upsetting, immoral, repulsive, or otherwise unacceptable (i.e., obsessional thoughts). Examples include thoughts of germs, ideas of injuries occurring to loved ones, unwanted sexual or blasphemous images, doubts about making unlikely mistakes, and impulses to harm elderly people. Examples of questions to elicit this information include:

• What intrusive thoughts do you have that trigger anxiety?

• What thoughts do you try to avoid, resist, or dismiss?

It is similarly important to obtain information about the cognitive basis of the individual’s obsessional anxiety (i.e., the feared consequences). For example, what does

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

96 OBSESSIVE-COMPULSIVE DISORDER

the person fear if exposed to obsessional stimuli? Examples of questions to elicit feared consequences include:

• What is the worst thing you imagine happening if you are exposed to ______ [obsessional trigger]?

• What do you think might happen if you don’t do your ______ rituals?

The cognitive-behavioral model proposes that misinterpretations of unwanted thoughts, impulses, and images give rise to OCD symptoms. Therefore, assessment should include identification of mistaken beliefs about the presence and meaning of such stimuli. For example, “Thinking about stabbing my wife could lead me to actually stab her,” “God will punish me for thinking immoral thoughts,” and “I’m a pervert if I have unwanted thoughts about sex.” Examples of questions to elicit this information include:

• What do you think it means that you have this thought?

• What will happen if you think this thought too much?

• Why do you try to avoid or dismiss these thoughts?

Some individuals with OCD fear that if obsessional anxiety is evoked, anxiety and related bodily sensations will persist indefinitely or spiral out of control. For example, “If I don’t arrange the books on the shelf perfectly, I will always feel anxious that things aren’t just right.” Questions to help elicit these types of cognitions include:

• Do you worry that you will become anxious and that the anxiety will never go away?

• What might happen to you if you remained anxious for long periods of time?

Two self-report questionnaires, the Obsessive Beliefs Questionnaire (OBQ) and the Interpretation of Intrusions Inventory (III), have been developed to systematically measure a range of pertinent OCD-related beliefs and misinterpretations (Frost & Steketee, 2002). These psychometrically validated instruments are useful to include in the functional assessment to augment interview data.

Assessment of Avoidance and Compulsive Rituals

The cognitive-behavioral model proposes that avoidance and compulsive behavior main- tain obsessional fear. It is, therefore, necessary to include such behaviors in a functional assessment. Most individuals with OCD avoid obsessional stimuli in order to reduce anxiety over feared disasters. Examples include avoidance of certain people (e.g., AIDS patients), places (e.g., public restrooms), situations (e.g., being the last one to leave the house), and words (e.g., “stab”). Examples of questions to elicit this information include:

• What situations do you avoid because of obsessional fear?

• Can you ever confront this situation?

• How does avoiding ______ make you feel more comfortable?

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

ASSESSMENT 97

If avoidance is impossible, compulsive behaviors are performed to escape from distress or reduce the probability of feared consequences. In addition to gathering detailed information about all rituals (e.g., cleaning, checking, repeating actions, arranging objects, and asking for reassurance), subtle mini rituals such as wiping, using special soaps, and brief checks must be assessed. The cognitive motivation of all rituals should also be clarified as well (e.g., checking to prevent fires, using a certain soap to target certain kinds of chemicals). Examples of questions to elicit this information include:

• What do you do when you can’t avoid ______ [insert situation]?

• Tell me about the strategies or rituals you use to reduce obsessional fear of ______ [insert obsessional fear]?

• How does doing this ritual reduce your discomfort?

• What might happen if you didn’t engage in this ritual?

Mental rituals are often overlooked in the functional assessment of OCD because they are not directly observable. Thus, it is important to ascertain any cognitive strategies the patient uses in response to obsessional stimuli. Examples include thinking “safe” thoughts, repeating prayers in a set manner, excessively mentally reviewing one’s own actions to gain assurance, and habitual thought suppression and mental distraction. As with overt rituals, it is necessary to ascertain the cognitive links between mental rituals and the obsessional thoughts. For example, repeating the phrase “God is good” to avoid punishment for having sacrilegious thoughts, and suppression of violent thoughts to prevent acting violently. Examples of questions to elicit this information include:

• What kinds of mental strategies do you use to dismiss unwanted thoughts?

• What might happen if you didn’t use the strategy?

Self-Monitoring

Self-monitoring of rituals and avoidance behavior is an excellent tool for collecting real-time data on OCD symptoms. A log sheet can be given to the patient on which he or she records (a) date, (b) time, (c) obsessional thought or stimulus that triggers anxiety, (d) level of anxiety on a 0 to 10 scale, and (e) the ritual or avoidance behavior employed. Self-monitoring helps both the clinician and the patient gain an accurate picture of OCD symptom severity and the functional relationship between obsessions and compulsions. It also helps the patient identify obsessions and rituals that he or she might not be aware of.

Case Conceptualization

The main value of a functional assessment is that it yields information from which to synthesize a case conceptualization. The case conceptualization is an individualized blueprint of OCD symptoms and is derived by listing (a) the situations and thoughts that trigger obsessional fear, (b) the associated cognitive variables (i.e., dysfunctional beliefs and appraisals), and (c) avoidance and compulsive rituals. Next, using the cognitive- behavioral model of OCD (as discussed previously) as a framework, links between these phenomena are sketched as shown in Figure 3.1. The individual in this case had obsessional fears of contamination along with decontamination rituals.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

98 OBSESSIVE-COMPULSIVE DISORDER

External Triggers

(e.g., driving, gasoline stations, bleaches, detergents, paints, smells)

Intrusive Obsessional Thoughts

Ideas : I’m contaminated with chemicals. Doubts : I might make my daughter sick from exposure to chemicals. Images : My daughter dying of cancer in 10 years.

Misappraisals of Obsessional Thoughts

I can’t take the chance that this will happen.

I must do everything I can to prevent this.

The more I think it, the more likely it is.

I can and should control my thoughts or I am fully responsible for the outcome.

Obsessional Anxiety/Fear

Anxiety-Reduction Strategies

Avoidance (e.g., chemicals, daughter, spouse)

Rituals (e.g., washing, cleaning, showering, changing clothes)

Short -Term Anxiety/Fear Reduction

Rituals and avoidance prevent the correction of dysfunctional beliefs.

Negative reinforcement:

Rituals and avoidance reduce distress.

Dysfunctional Beliefs Overestimates of the probability and severity of danger An inflated sense of responsibility for harm I can and should control my thoughts Thought-action fusion

Hypervigilance leads to noticing more obsessional cues.

FIGURE 3.1 Case Conceptualization of an Individual With OCD

Psychological Treatment

Cognitive behavior therapy (CBT), a set of techniques derived from the cognitive- behavioral theoretical model described earlier, is considered the most effective approach to the psychological treatment of OCD. Two specific CBT methods have been examined in clinical studies: behavior therapy by exposure and response prevention (ERP) and cognitive therapy (CT). This section provides a concise description of these procedures

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

PSYCHOLOGICAL TREATMENT 99

and reviews research substantiating their effectiveness. Detailed guidelines for planning and implementing CBT techniques are provided in various treatment manuals (e.g., Abramowitz, 2006; Clark, 2004).

EXPOSURE AND RESPONSE PREVENTION

Exposure and response prevention (ERP) entail confrontation with stimuli that provoke obsessional fear, but that objectively pose a low risk of harm. Exposure can occur in the form of repeated actual encounters with the feared situations (situational or in vivo exposure), or in the form of imagined confrontation with the feared disastrous consequences of confronting these situations (imaginal exposure). For example, an individual with obsessional fears of hitting pedestrians while driving would be asked to practice driving through crowded streets for situational exposure. She would also practice imaginal exposure to thoughts of possibly having hit someone and being held responsible. A patient with fears of contaminating his daughter by accident might be asked to touch objects of increasing dirtiness—a doorknob, the floor, a toilet seat—for situational exposure. He would then confront thoughts of his daughter coming down with a serious illness as a result of his carelessness with these contaminants.

As might be anticipated, initiating exposure tasks evokes the patient’s anxiety. Patients are encouraged to engage in such tasks completely, and to allow themselves to experience this obsessional distress without resisting the feelings of anxiety. Over time, the anxiety (and associated physiological responding) naturally subsides—a process called habitu- ation. With each repetition of the exposure task, habituation occurs more rapidly. The response prevention component of ERP entails refraining from compulsive rituals and other subtle avoidance behaviors that serve as an escape from obsessive fear. Response prevention helps to prolong exposure and to facilitate the eventual extinction of obses- sional anxiety. In the previous examples, the first patient might practice refraining from any strategies she typically uses to reassure herself that no accidents have occurred, such as going in reverse to check out the scene, gazing in the rearview mirror, asking a pas- senger for assurance, or checking with the police department about hit-and-run accidents in the vicinity. The second patient would be instructed to refrain from decontamination rituals such as washing or cleaning.

The Delivery of ERP

The way ERP is delivered can vary widely, although greater effectiveness is achieved when therapist-guided exposure sessions are held multiple times per week, as opposed to once weekly (Abramowitz, 1996). One format that has been found to produce particularly potent effects includes a few hours of assessment and treatment planning followed by 15 (daily or twice-weekly) treatment sessions, lasting about 90 minutes each (Abramowitz, Foa, & Franklin, 2003). When pragmatic concerns render intensive regimens impractical, conducting the treatment sessions on a weekly basis works very well for individuals with less severe OCD (Warren & Thomas, 2001). Self-supervised exposure homework practice should be assigned for completion between sessions. It is important that home- based self-supervised exposure exercises (as with therapist-supervised exposure) last long enough for the patient to learn that either anxiety dissipates (i.e., habituation occurs) or that he or she can tolerate the anxiety.

A course of ERP ordinarily begins with the assessment of obsessions, compulsive ritu- als, avoidance strategies, and anticipated consequences of confronting feared situations.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

100 OBSESSIVE-COMPULSIVE DISORDER

Information gathered during the assessment sessions is then used to plan the specific exposure exercises that will be pursued. Importantly, the term response prevention does not imply that the therapist physically restrains the patient from performing rituals. Instead, the therapist must convince patients to resist their own urges to carry out these behaviors. The therapist must also provide a cogent rationale for how exposure and response prevention will be helpful in reducing OCD. This rationale must motivate the patient to tolerate the distress that typically accompanies therapy. The treatment rationale also lets the patient know that this distress is temporary and that it usually subsides with repeated practice; although learning how to tolerate anxiety is an important part of long-term success with ERP.

The exposure exercises in ERP typically begin with moderately distressing situations, stimuli, and images, and progress to the most distressing situations. Beginning with less anxiety-evoking exposure tasks increases the likelihood that the patient will learn to manage distress. This also increases confidence in the treatment. At the end of each treatment session, the therapist instructs the patient to continue exposure for several hours and in different environmental contexts without the therapist. Exposure to the most anxiety-evoking situations—which must be faced—is completed during the middle third of the treatment program. This allows the patient ample opportunity to repeat such exposures in different contexts to allow generalization of treatment effects. During later sessions, the therapist emphasizes the importance of generalization and of the individual continuing to apply the ERP procedures learned during treatment.

Mechanisms of Change

Foa and Kozak (1986) hypothesized that ERP produces its effects by correcting patients’ overestimates of danger that underlie obsessional anxiety. These authors point to three requirements for successful outcome with ERP. First, physiological arousal and subjective fear must be evoked during exposure. Second, the fear responses must gradually diminish during the exposure session in the absence of compulsive rituals (within-session habituation). Third, the initial fear response at the beginning of each exposure session should decline across sessions (between-sessions habituation).

More recently, Craske and colleagues (Craske et al., 2008) have emphasized the role of inhibitory learning in exposure-based therapy. According to this approach, original fear-based associations (e.g., bathrooms = danger) are not corrected or replaced during exposure, but rather new safety-based associations (e.g., bathrooms = safety) are formed and they compete with the fear-based associations. The aim of CBT for OCD, then, is to help the patient develop this new learning and then enhance the recall of the new associations relative to the older, threat-based associations (Craske et al., 2008).

Several dismantling studies have examined the individual effects of exposure and response prevention techniques (e.g., Foa, Steketee, Grayson, Turner, & Lattimer, 1984), finding that whereas exposure produces the greatest effects on obsessional anxiety, response prevention produces the greatest reduction of compulsive rituals. Nevertheless, there is an additive effect of combining the two techniques: Using both exposure and response prevention is more effective than using either of its individual components (Foa et al., 1984).

The Efficacy of ERP

A meta-analysis of 24 ERP studies revealed very large posttreatment (ES = 1.16 to 1.41) and follow-up (ES = 1.10 to 1.57) effect sizes (Abramowitz, 1996). Foa and Kozak (1996)

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

PSYCHOLOGICAL TREATMENT 101

found that across 13 ERP studies, 83% of patients were responders (defined as at least 30% symptom reduction) at posttreatment, and across 16 studies, 76% were responders at follow-up (mean follow-up was 29 months). In concert, these findings suggest that the majority of OCD patients who undergo treatment with ERP evidence substantial short- and long-term benefit. Data from individual studies reveal that patients who complete a trial of ERP consistently achieve 50% to 70% symptom reduction.

Randomized controlled studies have found that ERP is superior to waiting list, progres- sive muscle relaxation, anxiety management training, pill placebo, and pharmacotherapy by serotonergic medication (e.g., Foa et al., 2005). These studies indicate that the effects of CBT are due to the specific ERP techniques over and above the effects of nonspecific factors common to all interventions, such as the therapeutic relationship. Moreover, the effects of ERP are not limited to highly selected research samples or to treatment as delivered in specialty clinics. Effectiveness studies conducted with nonresearch patients (e.g., Franklin, Abramowitz, Foa, Kozak, & Levitt, 2000) show that over 80% of patients who complete CBT achieve clinically significant improvement.

COGNITIVE THERAPY

The basis of cognitive therapy (CT) for OCD is the rational and evidence-based challenging and correction of faulty and dysfunctional thoughts and beliefs that underlie emotional distress. As has been discussed, individuals with OCD hold characteristic faulty beliefs that lead to obsessive fear. It is these beliefs that are targeted in CT, including overestimates of the probability and severity of danger and misinterpretations of intrusive thoughts as having implications for responsibility for harm.

Delivery of CT

Cognitive therapy typically begins with the therapist presenting a rationale for treatment that incorporates the notion that intrusive obsessional thoughts are normal experiences and not harmful or significant. The patient is then socialized to the cognitive-behavioral framework for understanding OCD as discussed earlier in this chapter. A central aim of CT is therefore to reduce obsessional fear and the need for compulsive rituals by helping the patient correct dysfunctional thinking and behavioral responses to obsessional stimuli (situations, thoughts, images) so that such situations no longer require avoidance and intrusive thoughts are no longer perceived as needing to be controlled or neutralized via rituals.

Various CT techniques are used to help patients identify and correct their erroneous appraisals, such as didactic presentation of educational material, Socratic dialogue, and cognitive restructuring aimed at helping patients recognize and remedy dysfunctional thinking patterns. Behavioral experiments, in which patients enter situations that exem- plify their fears, are often used to facilitate the acquisition of corrective information about the realistic risks associated with obsessional fears. For a detailed manual describing the use of CT techniques for OCD, see Wilhelm and Steketee (2006).

Van Oppen and Arntz (1994) outlined a 16-session CT intervention for OCD. This protocol included (a) learning to conceptualize obsessive intrusions as normal stimuli, (b) identifying and challenging anxiety-provoking thoughts associated with obsessions with Socratic questioning, (c) changing the dysfunctional assumptions to nondistressing beliefs, and (d) behavioral experiments to test out the new beliefs. This program was found to be effective in reducing OCD symptoms (Van Oppen et al., 1995).

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

102 OBSESSIVE-COMPULSIVE DISORDER

CT Versus ERP

A handful of studies have addressed the relative efficacy of CT and ERP by directly comparing variants of the two interventions. Although most of these studies have found that these treatments produced equivalent effects, methodological problems prevent definitive conclusions. In earlier studies, for example, both ERP and CT yielded minimal improvements in OCD symptoms. The efficacy of ERP was likely attenuated by the lack of therapist-supervised exposure, and CT programs were likely enhanced by the use of supervised behavioral experiments, which mimic the effects of exposure. Vogel, Stiles, and Götestam (2004), however, found that the inclusion of CT was useful in reducing dropout from ERP. Thus, there are likely benefits to incorporating CT techniques along with ERP; perhaps CT techniques improve the acceptability of ERP.

Biological Treatments

SEROTONIN REUPTAKE INHIBITORS

Pharmacotherapy using selective serotonin reuptake inhibitor (SSRI) medication is the most widely used treatment for OCD. The specific agents in this class of drugs include fluoxetine, paroxetine, sertraline, citalopram, and fluvoxamine. Clomipramine, a tricyclic medication that also possesses serotonergic properties, is also used in the treatment of OCD. Problems with the serotonin hypothesis (as discussed earlier in this chapter) withstanding, it is this model that provides the rationale for the use of serotonergic medications to treat OCD. On average, serotonin medications produce a 20% to 40% reduction in obsessions and compulsions (Rauch & Jenike, 1998). The major strengths of a pharmacological approach to treating OCD include the convenience and the requirement of little effort on the patient’s part. Limitations include the relatively modest improvement and likelihood of residual symptoms, high rate of nonresponse (40% to 60% of patients do not show any favorable response), and the prospect of unpleasant side effects (which can often be stabilized, however, by adjusting the dose). Moreover, once SSRIs are terminated, OCD symptoms typically return rapidly (Pato, Zohar-Kadouch, Zohar, & Murphy, 1988).

NEUROSURGICAL TREATMENT

Although they have received little in the way of controlled empirical evaluation, four neurosurgical procedures have been reported with OCD patients: (1) subcaudate trac- totomy, (2) limbic leucotomy, (3) cingulotomy, and (4) capsulotomy. These operations involve severing interconnections between areas of the brain’s frontal lobes and the limbic system. Surgical procedures are recommended only in cases where severe and unmanageable OCD and depressive symptoms persist despite adequate trials of all other available treatments; the risks of neurosurgery include permanent alterations in cognitive functioning and personality. Although clinical improvement has been observed in some cases, it remains unknown why these procedures are successful for only a subset of OCD patients (Jenike, 2000). There is also an increased risk of suicide following failure with this approach.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

SUMMARY AND NEW DIRECTIONS 103

Summary and New Directions

Few syndromes in psychopathology have generated as much curiosity and clinical exploration as has OCD. Since the 1970s, research on OCD has increased exponentially, leading to a clearer understanding of the heterogeneity of the disorder, its boundaries with other syndromes, and the development of increasingly sophisticated theoretical models of etiology and maintenance. Perhaps most importantly, research has led to advances in treatment; and whereas the first-line therapies (CBT and serotonergic medication) are not entirely effective for every patient, they have transformed OCD from an unmanageable lifetime affliction into a treatable problem that need not reduce quality of life.

Despite the aforementioned advances, a number of unresolved disagreements concern- ing OCD have emerged. Differences, for example, have surfaced over phenomenological issues, etiological models, and approaches to treatment. These disagreements occur predominantly along disciplinary lines between biologically oriented and cognitive- behaviorally oriented schools of thought. Biologically inclined theorists view OCD as a medical condition involving abnormal neurological processes, whereas psychosocial formulations emphasize the role of learning and dysfunctional cognitions. Yet, because of the relative insularity within each camp, theoretical conjecture and empirical findings from within different approaches are typically addressed toward distinct and narrow audiences. Clinicians, researchers, and students with broad interests are hindered from gaining a clear understanding of the diverse (and sometimes polarized) perspectives.

At present, we stand at a time when we can look to the past and to the future with hope. In looking back, we can be glad that the days are gone when individuals with OCD had little hope of receiving the help that they needed, or worse, underwent years of ineffective treatment with false hopes of improvement. Today, we can see that there are treatments that are useful and oftentimes highly effective. We can also see that there is energetic disagreement among experts in the OCD research communities, and it is the sort of disagreement that can lead to productive outcomes, fruitful debate, and more refined care of individuals with OCD. Although cognitive-behavioral treatments can be highly effective, one challenge for the future involves helping patients without good access to this therapy. Other challenges include helping individuals with motivational problems, and those with families and other social circles that reinforce OCD symptoms, to enter and succeed with treatment. In looking forward, we can also hope that investigators with differing backgrounds and research agendas will engage with one another as they pursue their own work so that at the close of this century we will have an understanding of OCD that integrates the best methods of behavioral science with the best methods of neurochemistry and neurophysiology.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

References

Abramowitz, J. S. (1996). Variants of exposure and

response prevention in the treatment of obsessive-

compulsive disorder: A meta-analysis. Behavior Therapy, 27, 583–600. doi:http://dx.doi.org/10.1016 /S0005-7894(96)80045-1

Abramowitz, J. S. (2006). Understanding and treat- ing obsessive-compulsive disorder: A cognitive- behavioral approach. Mahwah, NJ: Erlbaum.

Abramowitz, J. S., Deacon, B., Olatunji, B.,

Wheaton, M. G., Berman, N., Losardo, D., . . .

Hale, L. (2010). Assessment of obsessive-compulsive

symptom dimensions: Development and evalua-

tion of the Dimensional Obsessive-Compulsive

Scale. Psychological Assessment, 22, 180–198. doi:10.1037/a0018260

Abramowitz, J. S., Foa, E. B., & Franklin, M. E.

(2003). Exposure and ritual prevention for obsessive-

compulsive disorder: Effects of intensive versus

twice-weekly sessions. Journal of Consulting and Clinical Psychology, 71, 394–398. doi:10.1037/ 0022-006X.71.2.394

Abramowitz, J. S., Khandher, M., Nelson, C., Dea-

con, B. J., & Rygwall, R. (2006). The role of cognitive

factors in the pathogenesis of obsessive-compulsive

symptoms: A prospective study. Behaviour Research and Therapy, 44, 1361–1374. doi:10.1016/j.brat .2005.09.011

Abramowitz, J., Wheaton, M., & Storch, A. (2008).

The status of hoarding as a symptom of obsessive-

compulsive disorder. Behaviour Research and Ther- apy, 46, 1026–1033. doi:10.1016/j.brat.2008.05.006

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author.

Baxter, L. R., Jr., Schwartz, J. M., Mazziotta, J. C.,

Phelps, M. E., Pahl, J. J., Guze, B. H., & Fairbanks, L.

(1988). Cerebral glucose metabolic rates in nonde-

pressed patients with obsessive-compulsive disorder.

American Journal of Psychiatry, 145, 1560–1563. Beck, A. T. (1976). Cognitive therapy of the emotional

disorders. New York, NY: International Universities Press.

Burns, G. L., Keortge, S. G., Formea, G. M., &

Sternberger, L. G. (1996). Revision of the Padua

Inventory for obsessive compulsive disorder symp-

toms: Distinctions between worry, obsessions, and

compulsions. Behaviour Research and Therapy, 34, 163–173. doi:10.1016/0005-7967(95)00035-6

Chambless, D. L., & Steketee, G. (2000). Expressed

emotion and behavior therapy outcome: A prospec-

tive study with obsessive-compulsive and agorapho-

bic outpatients. Journal of Consulting and Clini- cal Psychology, 67, 658–665. doi:10.1037//0022- 006X.67.5.658

Clark, D. A. (2004). Cognitive-behavioral therapy for OCD. New York, NY: Guilford Press.

Conelea, C. A., Freeman, J. B., & Garcia, A. M. (2012).

Integrating behavioral theory with OCD assessment

using the Y-BOCS/CY-BOCS symptom checklist.

Journal of Obsessive-Compulsive and Related Dis- orders, 1, 112–118.

Craske, M. G., Kircanski, K., Zelikowsky, M., Mys-

tknwski, J., Chowdhury, N., & Baker, A. (2008).

Optimizing inhibitory learning during exposure ther-

apy. Behaviour Research and Therapy, 46, 5–27. Di Nardo, P., Brown, T., & Barlow, D. H. (1994).

Anxiety disorders interview schedule for DSM-IV: Lifetime version (ADIS-IV-LV). San Antonio, TX: Psychological Corporation.

Dowson, J. H., & Grounds, A. T. (1995). Personal- ity disorders: Recognition and clinical management. Cambridge, England: Cambridge University Press.

Eisen, J. L., Mancebo, M. C., Chiappone, K. L.,

Pinto, A., & Rasmussen, S. A. (2008).

Obsessive-compulsive personality disorder. In

J. A. Abramowitz, D. McKay, & S. Taylor (Eds.),

Clinical handbook of obsessive-compulsive disor- der and related problems. Baltimore, MD: Johns Hopkins University Press.

Eisen, J. L., Phillips, K. A., Baer, L., Beer, D. A.,

Atala, K. D., & Rasmussen, S. A. (1998). The Brown

Assessment of Beliefs Scale: Reliability and validity.

American Journal of Psychiatry, 155, 102–108. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J.

(2002). Structured clinical interview for the DSM- IV Axis I disorders. New York, NY: Biometrics Research Department, New York State Psychiatric

Institute.

Fitzgerald, K. D., Moore, G. J., Paulson, L. A., Stew-

art, C. M., & Rosenberg, D. R. (2000). Proton

spectroscopic imaging of the thalamus in treatment-

naive pediatric obsessive-compulsive disorder. Bio- logical Psychiatry, 47, 174–182. doi:10.1016/S0006- 3223(99)00286-3

Foa, E. B., Huppert, J. D., Leiberg, S., Langner, R.,

Kichic, R., Hajcak, G., & Salkovskis, P. M. (2002).

104

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

http://dx.doi.org/10.1016/S0005-7894(96)80045-1

REFERENCES 105

The Obsessive-Compulsive Inventory: Development

and validation of a short version. Psychologi- cal Assessment, 14, 485–496. doi:10.1037/1040- 3590.14.4.485

Foa, E. B., & Kozak, M. (1986). Emotional processing

of fear: Exposure to corrective information. Psy- chological Bulletin, 99, 20–35. doi:10.1037//0033- 2909.99.1.20

Foa, E. B., & Kozak, M. J. (1995). DSM-IV field trial: Obsessive-compulsive disorder. American Journal of Psychiatry, 152, 90–96.

Foa, E. B., & Kozak, M. J. (1996). Psychological treat-

ment for obsessive-compulsive disorder. In M. R.

Mavissakalian, R. F. Prien, M. R. Mavissakalian, &

R. F. Prien (Eds.), Long-term treatments of anxiety disorders (pp. 285–309). Washington, DC: American Psychiatric Association.

Foa, E. B., Liebowitz, M. R., Kozak, M. J., Davies, S.,

Campeas, R., Franklin, M. E., . . . Tu, X. (2005). Ran-

domized, placebo-controlled trial of exposure and

ritual prevention, clomipramine, and their combina-

tion in the treatment of obsessive-compulsive disor-

der. American Journal of Psychiatry, 162, 151–161. doi:10.1176/appi.ajp.162.1.151

Foa, E. B., Steketee, G, Grayson, J., Turner, R., &

Lattimer, P. (1984). Deliberate exposure and block-

ing of obsessive-compulsive rituals: Immediate and

long-term effects. Behavior Therapy, 15, 450–472. doi:10.1016/S0005-7894(84)80049-0

Franklin, M., Abramowitz, J., Foa, E., Kozak, M., &

Levitt, J. (2000). Effectiveness of exposure and ritual

prevention for obsessive-compulsive disorder: Ran-

domized compared with nonrandomized samples.

Journal of Consulting and Clinical Psychology, 68, 594–602. doi:10.1037//0022-006X.68.4.594

Frost, R. O., & Steketee, S. (2002). Cognitive approaches to obsessions and compulsions: The- ory, assessment, and treatment. Oxford, England: Elsevier.

Goodman, W. K., Price, L. H., Rasmussen, S. A.,

Mazure, C., Delgado, P., Heninger, G. R., & Char-

ney, D. (1989a). The Yale-Brown Obsessive Com-

pulsive Scale: Validity. Archives of General Psy- chiatry, 46, 1012–1016. doi:10.1001/archpsyc.1989 .01810110054008

Goodman, W. K., Price, L. H., Rasmussen, S. A.,

Mazure, C., Fleischmann, R. L., Hill, C. L.,

Delgado, P., & Charney, D. (1989b). The Yale-

Brown Obsessive Compulsive Scale: Development,

use, and reliability. Archives of General Psychi- atry, 46, 1006–1011. doi:10.1001/archpsyc.1989 .01810110048007

Hodgson, R., & Rachman, S. (1972). The effects of

contamination and washing in obsessional patients.

Behaviour Research and Therapy, 10, 111–117. doi:10.1016/S0005-7967(72)80003-2

Hollander, E., DeCaria, C. M., Nitescu, A., Gully, R.,

Suckow, R. F., Cooper, T. B., . . . Liebowitz, M. R.

(1992). Serotonergic function in obsessive-

compulsive disorder: Behavioral and neuroendocrine

responses to oral m-chlorophenylpiperazine and fen-

fluramine in patients and healthy volunteers. Archives of General Psychiatry, 49, 21–28. doi:10.1001/ archpsyc.1992.01820010021003

Insel, T. R., Mueller, E. A., Alterman, I., Linnoila, M.,

& Murphy, D. L. (1985). Obsessive-compulsive

disorder and serotonin: Is there a connection? Biolog- ical Psychiatry, 20, 1174–1188. doi:10.1016/0006- 3223(85)90176-3

Jenike, M. (2000). Neurosurgical treatment of

obsessive-compulsive disorder. In W. Goodman,

J. Maser, & M. V. Rudorfer (Eds.), Obsessive- compulsive disorder (pp. 457–482). Mahwah, NJ: Erlbaum.

Kessler, R., Berglund, P., Demler, O., Jin, R., Merikan-

gas, K., & Walters, E. (2005). Lifetime prevalence

and age-of-onset distributions of DSM-IV disor- ders in the National Comorbidity Survey Replica-

tion. Archives of General Psychiatry, 62, 593–602. doi:10.1001/archpsyc.62.6.593

McKay, D., Abramowitz, J. S., Calamari, J. E.,

Kyrios, M., Radomsky, A. S., Sookman, D., . . .

Wilhelm, S. (2004). A critical evaluation of

obsessive-compulsive disorder subtypes: Symptoms

versus mechanisms. Clinical Psychology Review, 24, 283–313. doi:10.1016/j.cpr.2004.04.003

Mowrer, O. (1960). Learning theory and behavior. New York, NY: Wiley.

Pato, M. T., Zohar-Kadouch, R., Zohar, J., & Mur-

phy, D. L. (1988). Return of symptoms after

discontinuation of clomipramine in patients with

obsessive-compulsive disorder. American Journal of Psychiatry, 145, 1521–1525.

Pertusa, A., Frost, R. O., Fullana, M. A., Samuels, J. F.,

Steketee, G., Tolin, D. F., . . . Mataix-Cols, D. (2010).

Refining the diagnostic boundaries of compul-

sive hoarding: A critical review. Clinical Psychol- ogy Review, 30, 371–386. doi:10.1016/j.cpr.2010 .01.007

Pfohl, B., & Blum, N. (1991). Obsessive-compulsive

personality disorder: A review of available data

and recommendations for DSM-IV. Journal of Per- sonality Disorders, 5, 363–375. doi:10.1521/pedi .1991.5.4.363

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

106 OBSESSIVE-COMPULSIVE DISORDER

Rachman, S., & de Silva, P. (1978). Abnormal and nor-

mal obsessions. Behaviour Research and Therapy, 16, 233–248. doi:10.1016/0005-7967(78)90022-0

Rachman, S., Elliot, C. M., Shafran, R., & Radom-

sky, A. S. (2009). Separating hoarding from OCD.

Behaviour Research and Therapy, 47, 520–522. doi:10.1016/j.brat.2009.02.014

Radomsky, A. S., Rachman, S., & Hammond, D. (2001).

Memory bias, confidence and responsibility in com-

pulsive checking. Behaviour Research and Therapy, 39, 813–822. doi:10.1016/S0005-7967(00)00079-6

Rasmussen, S. A., & Eisen, J. L. (1992a). The epidemi-

ology and clinical features of obsessive-compulsive

disorder. Psychiatric Clinics of North America, 15, 743–758.

Rasmussen, S. A., & Eisen, J. L. (1992b). The epi-

demiology and differential diagnosis of obsessive-

compulsive disorder. Journal of Clinical Psychiatry, 53, 4–10. doi:10.1007/978-3-642-77608-3_1

Rassin, E., Merckelbach, H., Muris, P., & Spaan, V.

(1999). Thought-action fusion as a causal factor in

the development of intrusions. Behaviour Research and Therapy, 37, 231–237. doi:10.1016/S0005- 7967(98)00140-5

Rauch, S., & Jenike, M. (1998). Pharmacological treat-

ment of obsessive compulsive disorder. In P. E.

Nathan & J. M. Gorman (Eds.), A guide to treatments that work. London, England: Oxford University Press.

Salkovskis, P. (1996). Cognitive-behavioral approaches

to the understanding of obsessional problems. In R.

Rapee (Ed.), Current controversies in the anxiety disorders (pp. 103–133). New York, NY: Guilford Press.

Sanavio, E. (1988). Obsessions and compulsions: The

Padua Inventory. Behaviour Research and Therapy, 26, 169–177. doi:10.1016/0005-7967(88)90116-7

Sheehan, D., Lecrubier, Y., Harnett-Sheehan, K., Amor-

iam, P., Janavs, J., Weiller, E., . . . Dunbar, G. (1998).

The Mini International Neuropsychiatric Interview

(M.I.N.I.): The development and validation of a struc-

tured diagnostic interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59(Suppl. 20), 22–33.

Stanley, M., Swann, A., Bowers, T., & Davis, M.

(1992). A comparison of clinical features in tri-

chotillomania and obsessive-compulsive disorder.

Behaviour Research and Therapy, 30, 39–44. doi:10.1016/0005-7967(92)90094-W

Steketee, G., & Van Noppen, B. (2003). Family

approaches to treatment for obsessive-compulsive

disorder. Brazilian Journal of Psychiatry, 25(1), 43–50.

Storch, E. A., Abramowitz, J. S., & Goodman, W. K.

(2008). Does obsessive-compulsive disorder belong

among the anxiety disorders in DSM-V? Depres- sion and Anxiety, 25, 326–347. doi:10.1002/da .20488

Taylor, S., Thordarson, D., & Sochting, I. (2002).

Obsessive-compulsive disorder. In M. Antony &

D. H. Barlow (Eds.), Handbook of assessment and treatment planning for psychological disorders (pp. 182–214). New York, NY: Guilford Press.

Thordarson, D. S., Radomsky, A. S., Rachman, S.,

Shafran, R., Sawchuk, C. N., & Hakstian, A. (2004).

The Vancouver Obsessional Compulsive Inventory

(VOCI). Behaviour Research and Therapy, 42, 1289–1314. doi:10.1016/j.brat.2003.08.007

Tolin, D. F., Abramowitz, J. S., Kozak, M. J., &

Foa, E. B. (2001). Fixity of belief, perceptual aberra-

tion, and magical ideation in obsessive-compulsive

disorder. Journal of Anxiety Disorders, 15, 501–510. doi:10.1016/S0887-6185(01)00078-0

Van Oppen, P., & Arntz, A. (1994). Cognitive ther-

apy for obsessive-compulsive disorder. Behaviour Research and Therapy, 32, 79–87. doi:10.1016/ 0005-7967(94)90086-8

Van Oppen, P., De Haan, E., Van Balkom, A. J.,

Spinhoven, P., Hoogduin, K., & Van Dyck, R.

(1995). Cognitive therapy and exposure in vivo in the treatment of obsessive compulsive disorder.

Behaviour Research and Therapy, 33, 379–390. doi:10.1016/0005-7967(94)00052-L

Vogel, P. A., Stiles, T. C., & Götestam, K. G.

(2004). Adding cognitive therapy elements to expo-

sure therapy for obsessive compulsive disorder: A

controlled study. Behavioural and Cognitive Psy- chotherapy, 32, 275–290. doi:10.1017/S1352465804 001353

Warren, R., & Thomas, J. C. (2001). Cognitive-behavior

therapy of obsessive-compulsive disorder in pri-

vate practice: An effectiveness study. Journal of Anxiety Disorders, 15, 277–285. doi:10.1016/S0887- 6185(01)00063-9

Wheaton, M. G., Abramowitz, J. S., Berman, N. C.,

Riemann, B. C., & Hale, L. R. (2010). The

relationship between obsessive beliefs and symp-

tom dimensions in obsessive-compulsive disorder.

Behavior Research and Therapy, 48, 949–954. doi:10.1016/j.brat.2010.05.027

Whiteside, S. P., Port, J. D., & Abramowitz, J. S.

(2004). A meta-analysis of functional neuroimag-

ing in obsessive-compulsive disorder. Psychiatry Research: Neuroimaging, 132, 69–79. doi:10.1016/j .pscychresns.2004.07.001

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

REFERENCES 107

Wilhelm, S., & Steketee, G. (2006). Cognitive ther- apy for obsessive-compulsive disorder: A guide for professionals. Oakland, CA: New Harbinger.

Woods, C. M., Vevea, J. L., Chambless, D. L.,

& Bayen, U. J. (2002). Are compulsive checkers

impaired in memory? A meta-analytic review. Clin- ical Psychology: Science and Practice, 9, 353–366. doi:10.1093/clipsy/9.4.353

Zohar, J., & Insel, T. R. (1987). Obsessive-compulsive

disorder: Psychobiological approaches to diagnosis,

treatment, and pathophysiology. Biological Psy- chiatry, 22, 667–687. doi:10.1016/0006-3223(87) 90199-5

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

Chapter 4

Generalized Anxiety Disorder

KAREN ROWA, HEATHER K. HOOD, AND MARTIN M. ANTONY

Introduction

Generalized anxiety disorder (GAD) is characterized by excessive worry about a variety of topics. Until the publication of the revision of the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R; American Psychiatric Association [APA], 1987), the diagnostic features of this disorder were not well established, and GAD was essentially a residual diagnostic category for individuals with persistent anxiety whose symptoms did not meet criteria for another anxiety disorder. Since the publication of DSM-III-R and in each subsequent edition of the DSM, the key feature of GAD has been chronic and excessive worry, and GAD is no longer considered a residual category. However, problems with unreliability in the diagnostic criteria, as well as considerable revisions from one version of DSM to the next, have taken their toll on the empirical status of knowledge regarding this disorder. Changing criteria have made it difficult for researchers to identify the essential biological and psychological underpinnings of GAD. Until recently, treatment efforts were restricted by the inherent instability of the core features of this disorder. Recently, however, GAD has finally begun to receive the research emphasis it deserves, and studies have built a more stable foundation of theory and knowledge regarding the nature and phenomenology of this disorder.

Nature of Generalized Anxiety Disorder

DIAGNOSTIC CONSIDERATIONS

The central feature of GAD according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5; APA, 2013) is excessive worry occurring on more days than not, about a number of different topics. Worry must persist for at least 6 months, and can involve a wide variety of topics such as finances, health, safety, and minor matters. Interestingly, the worries typically seen in GAD are indistinguishable in content from those reported by nonclinical samples (Becker, Goodwin, Hölting, Hoyer, & Margraf, 2003), though they are distinguishable from cognitions seen in panic disorder (Breitholtz, Johansson, & Öst, 1999) and social anxiety disorder (Hoyer, Becker, & Roth, 2001), and appear to be more future oriented than worry found in other anxiety disorders (Dugas, Gagnon, Ladouceur, & Freeston, 1998). However, the features

108

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

NATURE OF GENERALIZED ANXIETY DISORDER 109

that seem to best distinguish the worries found in GAD from normal worries include increased frequency and intensity, and the individual’s perceived inability to control the worry (Craske, Rapee, Jackel, & Barlow, 1989). As a result of such findings, difficulty controlling the worry was added as a diagnostic criterion in DSM-IV (APA, 1994).

In addition to the presence of uncontrollable and persistent worry, the diagnosis of GAD requires the presence of at least three out of six symptoms reflecting physiological or psychological arousal that accompany the worry. These include feeling keyed up, restless, or on edge, difficulty concentrating or having one’s mind go blank due to worry, disrupted sleep due to worry, muscle tension, irritability, and fatigue.

Worry is a common feature of mood disorders. In the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR; APA, 2000), one of the criteria specified that GAD could not be diagnosed if the worry occurred exclusively in the context of a mood disorder. Similarly, DSM-IV-TR specified that GAD should not be diagnosed if symptoms occurred exclusively during the course of posttraumatic stress disorder, a psychotic disorder, or a pervasive developmental disorder. These specific criteria have been removed in DSM-5. Instead, DSM-5 emphasizes that a diagnosis of GAD should not be made if the worry is better explained by another mental disorder and suggests, in the text rather than in the official criteria, that GAD should not be diagnosed if excessive worry occurs only during the course of a mood disorder. In other words, to establish a comorbid diagnosis of GAD with unipolar or bipolar depression, the presence of excessive worry should predate the onset of a mood disorder or continue to be met even during periods of remission from a mood disorder.

As with other diagnoses, the worry and associated symptoms of GAD must lead to significant distress or impairment in a person’s life. Further, the symptoms must not be better accounted for by another disorder (e.g., social anxiety disorder or an eating disorder), by a general medical condition, or by the use of a substance.

HISTORICAL PERSPECTIVES ON GENERALIZED ANXIETY DISORDER

As mentioned earlier, the criteria required to make a diagnosis of GAD according to the DSM-5 (APA, 2013) bear little resemblance to the criteria as first outlined in DSM-III (APA, 1980). To meet criteria for GAD according to the DSM-III, three of the following four symptoms needed to be met for at least 1 month: (1) symptoms reflecting startle, tension, and restlessness; (2) symptoms reflecting autonomic hyperactivity; (3) symptoms of anxious apprehension (e.g., worry, rumination); and (4) symptoms of hypervigilence. Thus, according to this diagnostic scheme, worry could be part of the diagnostic picture, but did not need to be present to meet criteria for a diagnosis of GAD. Further, according to DSM-III, GAD could be diagnosed only if all other anxiety disorders were ruled out, and none of the anxiety disorders could be diagnosed if another more pervasive problem had been diagnosed (e.g., a depressive or psychotic disorder). Thus, GAD was truly a residual category. The 1-month symptom duration also meant that a broader range of symptom presentations were able to meet criteria because of the inclusion of symptoms of a short duration. Accordingly, the reliability of this diagnostic category was poor (Di Nardo, O’Brien, Barlow, Waddell, & Blanchard, 1983). Changes in the DSM-III-R attempted to rectify some of the difficulties with the original diagnostic criteria for GAD, though these changes led to only a small increase in the reliability of the diagnosis (Di Nardo, Moras, Barlow, Rapee, & Brown, 1993). For the first time, worry became the central feature of GAD in DSM-III-R (APA, 1987), with excessive worry about at least

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

110 GENERALIZED ANXIETY DISORDER

two life circumstances having to be present. Further, symptoms had to be present for at least 6 months, raising the stringency for meeting criteria for GAD. Although establishing a minimum 6-month duration of symptoms appeared to be a reasonable strategy for increasing the meaningfulness of the diagnosis, recent research questions whether this duration is necessary (Lee et al., 2009).

DSM-III-R also addressed some of the hierarchical rule-outs that contributed to GAD’s residual status, including allowing the diagnosis of GAD even in the presence of another disorder (except for a mood disorder), as long as the GAD symptoms were not better accounted for by the other disorder. Despite these changes, the associated symptoms of GAD remained somewhat problematic, with the list of possible associated symptoms being too broad (6 out of 18 associated symptoms were required for a GAD diagnosis in DSM-III-R). Subsequent research provided useful suggestions for further revision to identify the most relevant associated symptoms. For example, Marten and colleagues (1993) found that symptoms of autonomic arousal (e.g., palpitations) were the least reliable in distinguishing individuals with GAD from nonanxious controls, and therefore these were removed in DSM-IV (in DSM-IV, three out of six associated symptoms were required for a GAD diagnosis). More recently, studies have provided additional clarity regarding which symptoms are most distinctively associated with GAD (Pluess, Conrad, & Wilhelm, 2009).

EPIDEMIOLOGY AND DESCRIPTIVE PSYCHOPATHOLOGY

It has been difficult to establish prevalence estimates for GAD due to the shifts in diagnostic criteria. The first prevalence estimates based on DSM-III-R criteria came from the National Comorbidity Survey. This study found that GAD was relatively rare in terms of current prevalence (1.6%), but was more common (approximately 5%) when lifetime prevalence was examined (Wittchen, Zhao, Kessler, & Eaton, 1994). The replication of the National Comorbidity Survey suggested a lifetime prevalence of GAD of 5.7% (Kessler et al., 2005). Rates of GAD in primary care appear to be even higher. A review of prevalence rates in primary care indicate a median point prevalence of this disorder of 5.8%, suggesting that people with GAD may be more likely to seek medical attention than individuals with other disorders and are therefore more highly represented in a primary care setting than in the community (Roy-Byrne & Wagner, 2004).

In the replication of the National Comorbidity Survey, the median age of onset of GAD was identified as 31 (Kessler et al., 2005). In a study of individuals who presented at an anxiety disorders clinic, mean age of onset was substantially lower at 21 (Brown, Campbell, Lehman, Grisham, & Mancill, 2001). Studies suggest that earlier onset of this disorder is associated with higher levels of symptom severity, comorbidity, and vulnerability to other disorders (Campbell, Brown, & Grisham, 2003).

Studies of the long-term course of GAD suggest that it is a chronic and relapsing disorder with some fluctuation in course. A 16-month follow-up of clients diagnosed using DSM-III-R criteria found that half of clients continued to be diagnosed with GAD at the time of follow-up (Mancuso, Townsend, & Mercante, 1993). An even less encouraging outcome was found in a 5-year follow-up of clients with GAD, where only 18% of clients achieved full remission at the follow-up assessment compared to 45% of clients with panic disorder (Woodman, Noyes, Black, Schlosser, & Yagla, 1999). A recent 5-year prospective study found that the probability of full remission at some point over the

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

NATURE OF GENERALIZED ANXIETY DISORDER 111

5 years was 38% and the probability of partial remission was 47%, with accompanying high levels of relapse (Yonkers, Dyck, Warshaw, & Keller, 2000). Predictors of a negative clinical course appear to be comorbid Axis I disorders (Bruce et al., 2005), personality disorders (especially cluster C personality features), decreased life satisfaction, and difficult family relationships (Yonkers et al., 2000). Results from a longitudinal study of individuals with GAD in primary care revealed a chronic course of illness, in which many individuals experienced periods of symptom improvement or recovery over a 2-year follow-up, but the majority of these individuals went on to experience a recurrence of their symptoms (Rodriguez et al., 2006). In this study, predictors of chronicity included being female, having greater comorbidity (especially depression and the anxiety disorders), and demonstrating more severe psychosocial impairment.

GAD appears to be more common in women than men, with some estimates indicating that women are almost twice as likely to meet diagnostic criteria for GAD at some time in their lives (McLean, Asnaani, Litz, & Hofmann, 2011). Importantly, these gender differences are associated with different patterns of comorbidity and disability. In one study, men with GAD had higher rates of comorbid alcohol and substance use, whereas women with GAD had higher rates of comorbid mood and anxiety disorders, and greater degree of disability (Vesga-López et al., 2008). One explanation for this gender difference is that women may use less effective strategies to manage worry. One study found that women were more likely than men to use thought suppression and to have a more negative problem orientation in response to worry (Robichaud, Dugas, & Conway, 2003). Other hypotheses suggest that there is a complex, bidirectional interaction of genetic vulnerabilities, such as higher trait anxiety and negative affectivity, coupled with environmental factors, such as gender role socialization, that increases the likelihood that women will develop anxiety and worry (for a review, see McLean and Anderson, 2009).

Minimal research has examined differences in symptom presentation across different ethnic or cultural groups. Theories of the impact of sociocultural factors on the onset of psychological disorders suggest that some racial or ethnic groups may have a greater vulnerability to develop psychological symptoms because of differential exposure to stressful events, lower socioeconomic position, and culturally specific beliefs about coping (Dohrenwend & Dohrenwend, 1969). In a demographically diverse, urban sample of individuals at risk for psychiatric difficulties, there were no differences among African American, Hispanic, and Caucasian individuals with regard to GAD symptom severity (Ghafoori, Barragan, Tohidian, & Palinkas, 2012). Other studies indicate that the content and severity of worry domains are not uniform across ethnic groups. For example, Scott, Eng, and Heimberg (2002) found that African American individuals reported less worry than Asian or Caucasian individuals on the topics of self-confidence, future goals, work competence, and relationship stability. Asian individuals reported more worry than the other groups on future goals. While Caucasians and Asians reported similar levels of worry across various domains of worry, African Americans reported the most frequent worry about financial issues, despite similar levels of socioeconomic status across the groups. Another study compared somatic symptoms of anxiety in individuals with GAD from the United States and Nepal. This study found that individuals from Nepal endorsed higher scores on somatic symptoms of anxiety, whereas individuals from the United States scored higher on psychological symptoms of anxiety (i.e., being nervous) (Hoge et al., 2006). Thus, there is some preliminary evidence that different cultural groups may demonstrate differences in both the content of worries as well as in the focus of GAD symptoms.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

112 GENERALIZED ANXIETY DISORDER

Although once considered a minor disorder, the significant impairment associated with GAD is now well established. Individuals with GAD have similar functional impairment to individuals with major depression, and the impairment is worse for those with both conditions (Wittchen, 2002). Some individuals with GAD report severe disability across domains, but particularly in the area of romantic relationships (Henning, Turk, Mennin, Fresco, & Heimberg, 2007). In addition, compared to nonanxious controls, individuals with GAD report significantly lower overall quality of life and life satisfaction on a wide range of indices, including self-esteem, work, health, and social relationships (Henning et al., 2007).

COMORBIDITY

Even though GAD was once thought of as a residual or minor disorder that was most likely to occur as a secondary problem, research suggests that a principal diagnosis of GAD is common and that numerous other psychiatric problems often co-occur with it. For example, data from the first National Comorbidity Survey found that 80% of their respondents with a principal diagnosis of GAD also had a comorbid mood disorder (Judd et al., 1998). A recent 32-year longitudinal study found that the onset of GAD was as likely to precede as to follow the onset of a major depressive episode, indicating that GAD is not simply a secondary condition (Moffitt et al., 2007). Research regarding the comorbidity of GAD using DSM-IV criteria suggests that a current principal diagnosis of GAD is highly comorbid with other current Axis I disorders, including panic disorder with or without agoraphobia (41%), social phobia (42%), and major depressive disorder (MDD) (29%) (Brown et al., 2001). Comorbidity rates are even higher when lifetime diagnoses are studied. Some research suggests that high rates of comorbidity in GAD are associated with high levels of trait anxiety and negative affect, variables that are consistently elevated in GAD patients (Chambers, Power, & Durham, 2004). In addition, rates of comorbid personality disorders are also elevated in individuals with GAD, including avoidant (26%), paranoid (10%), and schizotypal (10%) personality disorders (Brawman-Mintzer et al., 1993). The authors also found that anxiety symptoms, per se, did not contribute to personality disorder symptoms, suggesting that it is likely that the presence of other variables in these patients (e.g., interpersonal sensitivity) explains these high comorbidity rates (Mavissakalian, Hamann, Haidar, & de Groot, 1995).

The high rates of comorbidity found in GAD are associated with a more severe and chronic course of the disorder. Prospective studies find that rates of comorbidity increased over a 4-year follow-up period for individuals with GAD, and that this comorbidity had negative implications for the likelihood of remission from GAD symptoms (Bruce, Machan, Dyck, & Keller, 2001). Importantly, comorbidity rates appear to diminish with successful treatment (Borkovec, Abel, & Newman, 1995).

Neurobiological and Psychological Underpinnings

NEUROBIOLOGY

Studies on the neurobiology of GAD are limited, and therefore we do not have an integrated neurobiological theory of the development of this disorder. However, studies do implicate certain brain regions and neurotransmitters in understanding this disorder

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

NEUROBIOLOGICAL AND PSYCHOLOGICAL UNDERPINNINGS 113

(see Sinha, Mohlman, & Gorman, 2004). A number of neurotransmitter systems have been studied in GAD, in part because each of these systems has previously been implicated in understanding fear and anxiety. One such neurotransmitter is gamma-aminobutyric acid (GABA). Hypotheses involving GABA in GAD revolve around GABA’s inhibitory role in the brain, aiding inhibition of subcortical circuits that are stimulated by threat (Thayer & Lane, 2000). It is thought that individuals with GAD and other anxiety disorders may have decreased GABA activity (Friedman, 2007), which leads to less inhibition of these threat-activated structures. Further, individuals with GAD may also have reduced benzodiazepine receptor sensitivity, which contributes to anxiety because the binding of a benzodiazepine receptor actually facilitates GABA binding, which then inhibits excitatory responses in the brain (Sinha et al., 2004). Research to support these ideas includes the findings that (a) GABA receptors appear to be densely congregated in brain areas implicated in fear and anxiety, such as the frontal cortex, hippocampus, and amygdala (Petrovich & Swanson, 1997); (b) binding of a benzodiazepine to its receptor appears to increase the ability of GABA receptors to bind with available GABA (Goddard & Charney, 1997); and (c) benzodiazepines appear to be an effective treatment for GAD (reviewed later in this chapter).

Norepinephrine has also been implicated in GAD; however, data regarding the role of norepinephrine in GAD are mixed. Norepinephrine is the primary neurotransmitter in the sympathetic nervous system, the system responsible for the “fight or flight” response. Studies have demonstrated that norepinephrine levels are elevated in certain other anxiety disorders (e.g., Ballenger, 2001), raising the possibility that it may also be elevated in GAD. However, other studies are equivocal thus far, and appear to include confounds that make definitive conclusions about the role of norepinephrine in GAD difficult (see Sinha et al., 2004). To date, the majority of studies that have examined the role of norepinephrine in GAD have found no differences between clinical and control groups despite the use of various methodologies and biological challenges (Kalk, Nutt, & Lingford-Hughes, 2011).

Serotonin (5-HT) is also broadly implicated across the anxiety disorders, with studies suggesting that low levels of serotonin or serotonin receptor dysfunction, particularly in the midbrain region, are linked with increased anxiety (Nikolaus, Antke, Beu, & Müller, 2010). However, studies of serotonin in GAD have also yielded inconclusive results. Some studies have found that serotonin agonists lead to increased anxiety in GAD (Germine, Goddard, Woods, Charney, & Heninger, 1992) and that paroxetine (a selective serotonin reuptake inhibitor) shows reduced platelet binding in people with GAD compared to controls (Iny et al., 1994). However, other studies do not demonstrate differences in GAD patients versus controls on platelet binding of imipramine, which also affects the serotonergic system (Schneider, Munjack, Severson, & Palmer, 1987).

Cholecystokinin (CCK) has been linked to the panic attacks sometimes seen in GAD. One study found that a CCK agonist induced panic attacks in 71% of participants with GAD as compared to 14% of control participants (Brawman-Mintzer et al., 1997). However, studies have failed to find an anxiolytic effect of CCK receptor antagonists (Adams, Pyke, Costa, & Cutler, 1995). CCK is thought to be linked with both the serotonin and norepinephrine systems, and therefore may exert its effects in GAD through both of these systems, though one study suggested the effects of CCK appeared to be independent of the serotonin system (Goddard et al., 1999).

Cortisol levels, which suggest increased functioning of the limbic-hypothalamic- pituitary-adrenal (LHPA) axis, are also hallmark biochemical markers of stress and anxiety. Some evidence of overproduction of cortisol in GAD exists, though research

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

114 GENERALIZED ANXIETY DISORDER

is not consistent (Tiller, Biddle, Maguire, & Davies, 1988). A recent study provided evidence implicating hyperactivity of the HPA axis in GAD, in which elevated salivary cortisol levels were significantly reduced following SSRI treatment for GAD, and reduced cortisol was associated with lower reported anxiety (Lenze et al., 2011). Although there is some inconsistency in studies of cortisol in GAD, there tends to be a more consistent relationship between GAD symptom severity and elevated cortisol levels in older adults, perhaps due to age-associated alterations in the HPA axis (Mantella et al., 2008).

There have been few neuroimaging studies completed in GAD. Those that have been completed suggest larger amygdala volumes in children with GAD (De Bellis et al., 2000), larger superior temporal gyrus volumes (De Bellis et al., 2002), lower metabolic rates in the basal ganglia (Wu et al., 1991), and hypermetabolism in the prefrontal cortex (Wu et al., 1991). Recent studies have also identified decreased hypothalamus volumes as a possible risk factor for lifetime GAD. For example, Hettema and colleagues (2012) collected neuroimaging data from a small sample of female monozygotic twin pairs and found that lifetime GAD diagnosis and genetic risk for internalizing disorders significantly predicted smaller left hippocampal volumes. Other studies have supported this association, but noted decreased bilateral hippocampal volumes among individuals with GAD (Terlevic et al., 2013). In addition, the amygdala has been implicated in detecting and regulating emotional processing of threat across anxiety disorders. Structural and functional neuroanatomical studies have identified increased amygdala volumes among individuals with GAD compared to healthy controls, and disrupted connectivity of the amygdala with other cortical and subcortical regions involved in anticipatory anxiety and emotional processing (Etkin, Prater, Hoeft, Menon, & Schatzberg, 2010; Etkin, Prater, Schatzberg, Menon, & Greicius, 2009). Although it appears that the amygdala is a core component involved in directing threat and regulating emotional processing, more research is needed to clarify the role of this brain region in GAD, possibly by looking beyond the structural features of the amygdala toward its function and connectivity.

In a study of adolescents with GAD, neuroimaging revealed greater activation in response to angry faces in the right ventrolateral prefrontal cortex, but this activation had a negative relationship with subjective anxiety levels, a finding that is inconsistent with other imaging studies, which have found a positive relationship between anxiety levels and prefrontal cortex activation (Monk et al., 2006). To explain this discrepancy, the authors postulated that activation in the prefrontal cortex may play a compensatory func- tion, helping to regulate activation in subcortical structures. In support of this hypothesis, a follow-up study by Monk et al. (2008) found that adolescents with GAD showed hyperactivation of the right amygdala in response to masked angry faces compared to healthy controls. Further, the functional connectivity between the right amygdala and the right ventrolateral prefrontal cortex was weaker in the GAD group, suggesting that the ventrolateral prefrontal cortex downregulates the amygdala response to threatening emotional stimuli. Studies have also noted a similar inhibitory role of the prefrontal cortex in adults with GAD, particularly in the anterior cingulate cortex, a region of the medial prefrontal cortex. For example, Tromp et al. (2012) found structural abnormalities in the connectivity of the anterior cingulate cortex and the amygdala. Further, two studies have found that pretreatment activity in the anterior cingulate cortex was associated with better response to an 8-week trial of venlafaxine for patients with GAD (Nitschke et al., 2009; Whalen et al., 2008). Taken together, these studies suggest that the structures comprising this circuit may be integral to understanding the emotional processing deficits seen in GAD.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

COGNITIVE AND BEHAVIORAL CORRELATES 115

Chronic worry is also associated with reduced autonomic variability during stressful tasks. This more recent focus on the inhibitory function of the parasympathetic nervous system in addition to the excitatory function of the sympathetic nervous system has helped in understanding central nervous system functioning in GAD. Earlier studies tended to focus more exclusively on the excitation end of the equation (Friedman, 2007). However, as Friedman (2007) points out, physiological variability is actually the most effective way that organisms can maintain stability of their systems. The more variability an organism can show in responding to threat cues (a sympathetic system function) and recovering from threat cues (a parasympathetic system function), the more able to adapt to the environment the organism is likely to be. Thus, autonomic rigidity is thought to lead to less adaptive behavioral and emotional responses to stressful events in comparison to an adaptive system where a more flexible autonomic system can help a person organize his or her physiological responses (Porges, 1995).

The most commonly used marker of autonomic rigidity is lower levels of heart rate variability in response to various stressors. Research suggests that individuals with GAD demonstrate lower heart rate variability during periods of worry and rest, decreased parasympathetic activity during periods of worry and rest (Thayer, Friedman, & Borkovec, 1996), and impaired habituation of heart rate activity to neutral words (Thayer, Friedman, Borkovec, Molina, & Johnsen, 2000). Although there is evidence that all anxiety disorders are characterized by reduced baseline heart rate variability compared to healthy controls, heart rate variability is associated only with symptom severity in GAD (Pittig, Arch, Lam, & Craske, 2013). This finding may suggest that decreased heart rate variability reflects a shared biological dysfunction across anxiety disorders, but the specific manifestation of this feature may vary across disorders. However, autonomic rigidity in worry has not been consistently demonstrated, with one study finding no evidence for autonomic rigidity as measured by heart rate variability in periods of relaxation, nonstressful cognitive tasks, worry, and negative imagery (Davis, Montgomery, & Wilson, 2002). Thus, more research is still necessary to fully understand the rigidity of the physiological system underlying GAD.

In summary, neurobiological indices suggest that individuals with GAD have a dysregulated central nervous system that may take longer to recover from a stressor than others do. Further, the suppressed and rigid autonomic activity that is associated with worry may have implications for the maintenance of worry, providing individuals with short-term avoidance of physiological responses to stress, but impairing long- term adaptation to these stressors. These ideas will be discussed more thoroughly in the following sections (see, especially, the section on avoidance theories of worry). It is important to note, however, that there still exist many inconsistencies across studies on the neurobiology of GAD, and further research is necessary to address these inconsistencies and aid in the development of a more coherent understanding of the neurobiology of GAD.

Cognitive and Behavioral Correlates

LIFE EVENTS

A small body of research has investigated the role that stressful life events might play in the development of chronic worry and GAD. An early study found that the presence of at least one stressful life event defined as unexpected, negative, and very important

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

116 GENERALIZED ANXIETY DISORDER

was associated with an increased risk of developing GAD symptoms (Blazer, Hughes, & George, 1987). This association was even more dramatic for men who had experienced more than four stressful life events; these men had a risk of GAD that was over 8 times greater than men reporting zero to three life events. Further, elevated levels of generalized anxiety have been found in women who have experienced abortions (Cougle, Reardon, & Coleman, 2005) and in new mothers (Wenzel, Haugen, Jackson, and Robinson, 2003). In addition, stressful life events appear to predict relapse in previously remitted GAD patients, although specific worry themes were not strongly related to the likelihood of relapse (Francis, Moitra, Dyck, & Keller, 2012). It appears that stressful life events may be a general risk factor for the development and relapse of a depressive or anxiety disorder, perhaps because stressful life events alter the stress response and, consequently, activity of the HPA axis, creating a vulnerability to psychopathology (Faravelli et al., 2012).

PROBLEM-SOLVING ABILITY

Researchers have also questioned the association between chronic, elevated worry and problem-solving ability in understanding GAD. Worrying has been conceptualized as an attempt to anticipate or solve real-life problems (Tallis, Davey, & Capuzzo, 1994) and research suggests that worry frequency is related to constructive problem-focused coping (Davey, Hampton, Farrell, & Davidson, 1992). However, at some point, the worry process as constructive problem solving breaks down, and becomes pathological. Davey et al. (1992) hypothesized that pathological worry may be associated with either poor problem-solving ability or poor confidence in problem solving, which would prevent individuals from reaching acceptable solutions to perceived problems, and thus lead to further worry.

Interestingly, research suggests that individuals with chronic worry do not have deficits in problem-solving ability (Dugas, Freeston, & Ladouceur, 1997), but rather have less confidence about their problem-solving abilities (Davey, 1994) and hold a negative problem orientation in response to problems (Belzer, D’Zurilla, & Maydeu-Olivares, 2002). Negative problem orientation refers to how a person thinks and feels about his ability to solve real-life issues. Individuals with GAD tend to have a more negative problem orientation than individuals with other anxiety disorders (Ladouceur et al., 1999) and nonanxious controls (Dugas, Gagnon, et al., 1998). Negative problem orientation, together with intolerance of uncertainty, were the strongest predictors of worry severity, and differentiated individuals with moderate and severe GAD symptoms from those with mild GAD symptoms (Dugas et al., 2007). Further, the manipulation of problem-solving confidence has an impact on levels of anxiety and the tendency to catastrophize a personal worry, with lower confidence yielding higher anxiety and personal worry (Davey, Jubb, & Cameron, 1996). This study lends credence to the notion that meta-aspects of worry such as low problem-solving confidence may play a causal role in determining subsequent levels of worry.

Results for problem-solving ability versus orientation are similar to those found when examining the relationship between time management and worry. Worry has been found to have no relationship with time management behaviors (e.g., setting goals, organizational skills), but does have a negative relationship with time structure and purpose (e.g., possessing the notion that one’s time use has a purpose, being present- focused, demonstrating persistence when completing tasks) (Kelly, 2003). Thus, one maintenance factor of pathological worry as seen in GAD is likely influenced by the

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

COGNITIVE AND BEHAVIORAL CORRELATES 117

thoughts and feelings individuals have toward their problem-solving abilities and the purpose they attribute to the use of time.

Studies have also examined the strategies people with GAD use in response to their chronic worry. Similar to strategies used by individuals with other anxiety disorders, individuals with GAD reported a greater use of worry (e.g., thinking about more minor problems) and punishment (e.g., shouting at oneself) strategies than did nonanxious controls and less use of distraction (e.g., thinking about something else) and social control (e.g., talking to a friend) (Coles & Heimberg, 2005). Chronic worriers also appear to catastrophize about positive aspects of their lives as well as hypothetical situations (Davey & Levy, 1998).

PROBABILITY OVERESTIMATION AND CATASTROPHIZING

Excessive worry, the defining feature of GAD, is essentially a cognitive phenomenon. Research confirms that adults with GAD (Provencher, Freeston, Dugas, & Ladouceur, 2000) and childhood worriers (Suarez & Bell-Dolan, 2001) exhibit cognitive errors of probability overestimation (thinking a feared consequence is more likely to occur than it really is) and catastrophizing (assuming that an outcome will be much less manageable than it actually is). Estimates of the cost of one’s worry are related to worry severity, such that greater cost estimates are linked with greater worry severity (Berenbaum, Thompson, & Pomerantz, 2007).

INFORMATION-PROCESSING BIASES IN WORRY

Information-processing theories of the development and maintenance of GAD suggest that individuals with GAD pay greater attention to threatening stimuli, preferentially encode this information, interpret ambiguous stimuli as threatening, and have biases in memory for threatening events. Through such processes, individuals have greater access to frightening and threatening information, leading to increased states of anxiety. A recent meta-analysis examined threat-related attentional biases in anxiety and found that the anxiety disorders, including GAD, are characterized by heightened attentional and memory biases for threat (Bar-Haim, Lamy, Pergamin, Bakermans-Kranenburg, & van IJzendoorn, 2007). Importantly, these biases are robust across a variety of experimental paradigms, age groups, and whether the stimuli require conscious processing or are presented subliminally. Specific characteristics of these attentional biases include selective attention toward, difficulty disengaging from, and attentional avoidance of threatening stimuli (Cisler & Koster, 2010).

Research using the emotional Stroop paradigm has provided evidence for a cognitive bias in GAD (see Mogg & Bradley, 2005, for a more detailed review of this topic). In the Stroop paradigm, participants are presented with threatening and neutral words presented in different colors. Participants are asked to quickly name the color of the word, while ignoring the meaning of the word. The inability to ignore the meaning of the word (thus suggesting preferential encoding of threat meanings) is deduced from slower color naming. Individuals with GAD appear to have difficulty with the Stroop task, taking longer to name the colors of threatening words (e.g., “disgrace”) than nonthreatening words (e.g., “carpet”) (Mathews & MacLeod, 1985). This result is consistent when words are presented in blocks (i.e., numerous words presented at once)

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

118 GENERALIZED ANXIETY DISORDER

(Mogg, Mathews, & Weinman, 1989); when stimuli involve general threat words (Martin, Williams, & Clark, 1991); when stimuli are relevant to the participant’s idiosyncratic concerns (Mathews & Klug, 1993); and when stimuli are generally emotional words, including positive emotional words (Becker, Rinck, Magraf, & Roth, 2001). Further, the Stroop effect is also seen when word stimuli are blocked; that is, when words are followed by a masking stimulus designed to restrict the participant’s conscious awareness of the word (Bradley, Mogg, Millar, & White, 1995). This attentional bias appears to be modifiable by treatment, such that color-naming interference in GAD participants is ameliorated by cognitive behavior therapy (CBT; Mogg, Bradley, Millar, & White, 1995).

The visual dot probe task is another paradigm to study attentional biases in GAD. Individuals are asked to respond to a probe stimulus (e.g., a small dot) presented on a screen. Just prior to this probe stimulus, participants are presented with stimuli on the screen, one of which is the experimental stimulus (e.g., a threatening word or face). If participants respond quicker to a probe stimulus that appears in the same spot as the threatening stimulus, it is assumed that the participant was attending to that threat cue, and is therefore demonstrating an attentional bias for the threat cue. Research suggests that individuals with GAD demonstrate an attentional bias on this task, using threat words (MacLeod, Mathews, & Tata, 1986); negative words (Mogg, Bradley, & Williams, 1995); emotional faces (Bradley, Mogg, White, Groom, & de Bono, 1999); and aversive pictures (MacNamara & Hajcak, 2010). Further, this bias is also demonstrated when stimuli are masked, suggesting that the processing of threat cues may not be at a conscious level (Mogg et al., 1995). Individuals with GAD were more likely to look toward threat faces rather than neutral faces in a visual probe task compared to nonanxious controls and individuals with depression (Mogg, Millar, & Bradley, 2000).

Individuals with GAD also appear to interpret ambiguous stimuli in a negative or threatening manner. For example, when presented with a number of homophones with both a threatening and nonthreatening meaning (e.g., “die” versus “dye”) and asked to write out these words, individuals with GAD tend to write down a greater proportion of the threatening spellings of these words than do control participants (Mogg, Baldwin, Brodrick, & Bradley, 2004). Similarly, when GAD participants are presented with sentences that have a threatening or nonthreatening interpretation and are then asked to pick out recognized sentences whose meaning has been clarified by context, they have better recognition for the sentences that were given a threatening meaning by context (Eysenck, Mogg, May, Richards, & Mathews, 1991).

Biases in memory for threatening cues have also been studied in GAD. Researchers have hypothesized that individuals with GAD will demonstrate better memory for threatening than nonthreatening stimuli, and also enhanced memory for threatening stimuli than control participants. Results of a memory bias in GAD are not as consistent as those for other types of cognitive biases. Research has generally not found an explicit memory bias in GAD, such that individuals with GAD do not appear to preferentially recall more threatening stimuli (e.g., Becker, Roth, Andrich, & Margraf, 1999), though there are exceptions to these findings. For example, the explicit memory bias in GAD tends to be more pronounced when the stimulus words are ideographically selected and more personally relevant to the participant (Coles, Turk, & Heimberg, 2007).

On the other hand, several studies have found an implicit memory bias in GAD, such that nonconscious memory for threatening stimuli has been demonstrated to affect performance on other tasks, even if these stimuli are not preferentially remembered. For example, MacLeod and McLaughlin (1995) presented threatening and nonthreatening

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

COGNITIVE AND BEHAVIORAL CORRELATES 119

words to individuals with GAD and nonanxious controls, and then asked participants to identify briefly presented words. Individuals with GAD demonstrated an implicit memory bias for threatening words, such that they were more readily able to identify these threatening words. Although the evidence for an implicit memory bias in GAD is more robust than that for an explicit memory bias, results are still equivocal (for a review, see MacLeod & Rutherford, 2004).

Clearly, individuals with GAD demonstrate a number of biases for threat-relevant information, suggesting that attentional biases are an important aspect of the development and maintenance of this disorder. What is still unclear is the relationship between these cognitive biases and the physiology of GAD. One would assume that preferentially encoding threat information would lead to sympathetic arousal, but, as reviewed earlier, individuals with GAD do not demonstrate such arousal. In addition, the mechanisms underlying these attentional biases are not fully understood. Various models propose that attentional biases are mediated by impairment in the attentional control system, inappropriate use of emotion regulation strategies, specific neural mechanisms (e.g., amygdala, prefrontal cortex), and stage of information processing (for a review, see Cisler & Koster, 2010). However, although each of these mediating variables account for some facets of the observed attentional biases, a comprehensive interactive model incorporating these aspects has yet to be tested. Given that one of the hallmark features of cognitive-behavioral interventions for GAD involves trying to modulate biases in cognition, such a model would assist in better understanding the mechanisms of action in CBT and possibly identify specific treatment targets.

AVOIDANCE THEORIES OF WORRY

Researchers have proposed that worry is actually one of the ways that individuals respond to threat (see Borkovec, Alcaine, & Behar, 2004, for a comprehensive review of this theory). According to the cognitive avoidance theory of worry, worry is an attempt at cognitive avoidance, which is similar in function to the behavioral avoidance seen in many types of anxiety disorders. In other words, the verbal activity of worry is thought to distract individuals from the full experience of fear (e.g., feared imagery, sensations of arousal, etc.) and other negative emotional states. Evidence for the notion of worry as cognitive avoidance stems from a number of lines of research. Studies suggest that worry is a primarily verbal and linguistic activity, compared to an image-based process (Behar, Zuellig, & Borkovec, 2005). Further, imagery associated with worry is more concrete (i.e., less vivid, slower, and more difficult to access; Stöber, 1998), making it less likely to evoke a fear reaction. Verbal recounting of feared material is thought to produce less sympathetic nervous system response than feared images. Thus, worry in the face of perceived threat would be negatively reinforced because it allows a person to experience a reduction in distress and arousal in the short term when confronted by a stressful trigger.

Self-reports by college students meeting criteria for GAD suggest that they use worry to distract themselves from more emotional topics, and this motivation for using worry distinguished the GAD group from control participants (Borkovec & Roemer, 1995). People with GAD report higher levels of experiential avoidance than do control individuals (Roemer, Salters, Raffa, & Orsillo, 2005), and worriers avoid anxiety-provoking images even when instructed to attend to them (Laguna, Ham, Hope, & Bell, 2004). Unfortunately, worrying appears to provide only short-term benefit, and is associated with increased arousal in the long term (Wells & Papageorgiou, 1995). Borkovec et al. (2004) suggest

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

120 GENERALIZED ANXIETY DISORDER

that worrying may ultimately interfere with emotional processing of stressful stimuli. Thus, just as situational avoidance provides only short-term but not long-term benefit for anxiety, worry does the same.

In partial support of the cognitive avoidance model of worry, Stapinksi, Abbott, and Rapee (2010) found that individuals instructed to worry in response to an anxiety- provoking trigger maintained threat expectancies and decreased perceptions of control compared to those who used imaginal processing or relaxation/distraction. However, contrary to predictions, worry did not reduce physiological arousal relative to imaginal processing or relaxation, denoted by increased skin conductance. In addition, there was no long-term maintenance of worry among those in the worry condition, suggesting that worry is not necessarily a self-perpetuating process and the effects of worry may be a short-term phenomenon. Further, Newman and Llera (2011) argued that worry prolongs negative emotionality, perhaps because it maintains an anxiety-inducing cognitive representation of a stressor.

To account for these results, Newman and Llera (2011) proposed an extension and modification of the cognitive avoidance theory that focuses on experiential avoidance in GAD. Specifically, their contrast avoidance model draws on evidence that worry does not necessarily enable avoidance of negative emotional experiences (Llera & Newman, 2010), but rather individuals with GAD use worry to avoid further increases, or negative contrast, in negative emotions. They posited that the chronic distress associated with worry helps individuals prepare for negative events and precludes further distress should the negative outcome occur. Further, the positive contrast between anxiety-related distress and relief is heightened when the anticipated negative outcome fails to occur. Although this model requires further empirical support, it is consistent with other cognitive avoidance theories of GAD and appears to provide a synthesis of the extant literature.

INTOLERANCE Of UNCERTAINTY

Dugas and colleagues (e.g., Dugas, Buhr, & Ladouceur, 2004) have proposed that a central difficulty in GAD involves an individual’s tendency to react negatively to uncertain or ambiguous situations, sometimes preferring a negative outcome to an uncertain one. They suggest that intolerance of uncertainty is a cognitive filter through which a person with GAD views his or her world. Because many events and situations are characterized by uncertainty, the outcome of having this filter is an enhanced likelihood of worrying.

Intolerance of uncertainty is increasingly being recognized as a transdiagnostic feature across anxiety and mood disorders. In a recent study comparing intolerance of uncertainty across diagnostic groups, individuals with GAD, social anxiety disorder (SAD), obsessive- compulsive disorder (OCD), and major depressive disorder (MDD) reported significantly greater intolerance of uncertainty compared to undergraduate and community samples, though there were no differences among the clinical groups (Carleton, Mulvoge, et al., 2012). This study suggests that intolerance of uncertainty may represent a vulnerability factor for mood and anxiety symptoms, which is consistent with other data (e.g., McEvoy & Mahoney, 2011). A recent taxometric analysis indicated that intolerance of uncertainty is not a categorical variable that differentiates diagnostic groups, but rather a continuous variable that appears to be present in varying degrees in the population (Carleton, Weeks, et al., 2012). These data are consistent with the interpretation that intolerance of uncertainty may be an underlying vulnerability factor for the development of psychopathology.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

COGNITIVE AND BEHAVIORAL CORRELATES 121

Studies suggest that intolerance of uncertainty is related to worry in nonclinical participants (Freeston, Rhéaume, Letarte, Dugas, & Ladouceur, 1994) and is elevated in individuals with GAD (Dugas, Gagnon, et al., 1998). Although some studies indicate that elevated levels of intolerance of uncertainty are specific to GAD (e.g., Dugas, Marchand, & Ladouceur, 2005), this is likely an artifact of the measures used to assess intolerance of uncertainty. For example, a recent meta-analysis found that intolerance of uncertainty was more strongly related to GAD when using a definition of intolerance of uncertainty developed specifically for GAD, and similarly associated with OCD, MDD, and GAD when using an OCD-specific definition (Gentes & Ruscio, 2011).

Experimental manipulations of intolerance of uncertainty have yielded changes in level of worry, such that increasing participants’ level of intolerance of uncertainty leads to increased catastrophic worrying and low mood (Meeten, Dash, Scarlet, & Davey, 2012). Further, time-series analyses of the temporal sequence of change in CBT for GAD found that, for the majority of participants, changes in intolerance of uncertainty preceded changes in time spent worrying (Dugas, Langlois, Rhéaume, & Ladouceur, 1998). Thus, evidence is accumulating to suggest that intolerance of uncertainty may play a causal role in the development or exacerbation of pathological worry in GAD.

Researchers have also studied the relationship between intolerance of uncertainty and other variables demonstrated to be important in understanding GAD. For example, intolerance of uncertainty and negative problem orientation both contribute to the prediction of worry in nonclinical (Dugas et al., 1997) and clinical (Dugas et al., 2007) samples. Intolerance of uncertainty has also been linked to an information-processing bias, such that participants high in intolerance of uncertainty recalled a greater proportion of words involving uncertainty (e.g., “inconclusive” and “unclear”) and were more concerned about ambiguous situations than participants low in intolerance of uncertainty (Dugas et al., 2005). In a replication and extension of this study, Koerner and Dugas (2008) found that individuals high in intolerance of uncertainty were more likely to make negative appraisals of ambiguous scenarios than those low in intolerance of uncertainty, even after controlling for gender, low mood, and GAD severity. Further, negative appraisals mediated the relationship between intolerance of uncertainty and worry, and, conversely, worry was a partial mediator of the relationship between intolerance of uncertainty and negative appraisals. This indicates that intolerance of uncertainty represents a cognitive vulnerability for the development of psychopathological symptoms only when cognitive biases are present. In addition, worry appears to be a self-perpetuating process by contributing to the cognitive biases thought to be involved in the onset of worry in the first place. Although it is not clear how these cognitive biases develop, Dugas et al. (2004) theorized that intolerance of uncertainty might affect people’s perception of their problem-solving abilities, such that they feel less capable of effectively solving problems as they arise, leading to pathological worry instead of problem solution.

METACOGNITION AND GAD

Wells (e.g., Wells, 1999) has proposed a metacognitive model of GAD that involves positive and negative beliefs about worry, as well as two types of worry content. Metacognition refers to the act of thinking about one’s thought processes. Wells (1999) suggests that it is not typical worry per se that is the most significant problem in GAD, but rather the way one thinks about and reacts to this typical worry. Type 1 worry refers to this so-called typical worry—worry triggered by everyday events (e.g., worries about

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

122 GENERALIZED ANXIETY DISORDER

health, safety, or relationships). Type 2 worry consists of worry about these cognitive processes, or in other words, worry about worry. In this model, Wells proposes that positive beliefs that a person holds about worry (e.g., “Worry will help me cope”) cause a person to actively select worry as a coping strategy when faced with some sort of stressor. As a result, Type 1 worry is activated. In some circumstances, the process of worry is terminated by a felt sense that the person will be able to cope with the particular stressor. However, if this felt sense is not achieved, Type 2 beliefs about worry might be activated. These might include beliefs that worry is uncontrollable and/or beliefs that worrying might have negative mental or physical consequences for an individual. The activation of Type 2 worry is thought to produce excessive levels of anxiety, as well as to lead to unhelpful behavioral coping and thought control strategies, such that Type 2 worries cannot be disconfirmed and the worry cycle is maintained. For example, an increase in anxiety interferes with a person’s ability to achieve a felt sense that he or she can cope, so the person begins to avoid feared situations as a means of coping. Unfortunately, avoidance of the situations that trigger worry does not allow the person to experience disconfirmatory information about the impact of worry on his or her functioning.

Studies confirm that individuals hold both positive and negative beliefs about worrying. For example, Davey, Tallis, and Capuzzo (1996) found that individuals’ beliefs about worrying clustered into three factors representing negative consequences (i.e., worry disrupting performance, worry exaggerating the problem, and worry causing emotional distress), and two factors representing positive consequences (i.e., worry as motivation and worry as helping analytical thinking). Cartwright-Hatton and Wells (1997) developed the Meta-Cognitions Questionnaire and found that three subscales uniquely predicted worry: positive beliefs about worry, negative beliefs about the controllability of thoughts, and cognitive confidence (i.e., lack of confidence in one’s cognitive skills). Positive beliefs about worry have been found to uniquely predict worry above and beyond the prediction of worry by general negative beliefs (Francis & Dugas, 2004). In a case study of CBT for an individual with GAD, modifying positive beliefs about worry led to positive treatment outcomes (Borkovec, Hazlett-Stevens, & Diaz, 1999). Although evidence is clear that positive beliefs about worry are an important aspect of GAD, they are not the most robust predictor of GAD severity (Dugas et al., 2007), nor are they considered unique to GAD (Penney, Mazmanian, & Rudanycz, 2012; Wells, 2005).

In contrast, negative beliefs that worry is uncontrollable or dangerous appear to be central to the maintenance of GAD symptoms. Studies have provided support for the presence of Type 2 worry and its unique association with pathological worry, even when Type 1 worries and trait anxiety are controlled (Wells & Carter, 1999). For example, Ruscio and Borkovec (2004) compared the appraisals of worry following a stressor task by high worriers with and without GAD. They found that individuals with GAD were more likely to endorse beliefs that worry was uncontrollable and dangerous, and reported lower confidence in their cognitive abilities. However, the groups did not differ in their positive beliefs about worry. In a follow-up study with a nonclinical sample, negative beliefs about the uncontrollability and dangerousness of worry, but not positive beliefs about worry, mediated the relationship between trait worry and GAD symptoms (Penney et al., 2012). These results suggest that individuals with high trait worry who perceive their worry as dangerous or uncontrollable are more likely to experience GAD. Other studies have similarly identified meta-worry as an important factor in the severity and maintenance of GAD. Ruggiero and colleagues (2012) found that intolerance of uncertainty and negative beliefs about worry were reciprocally reinforcing and the interaction of these

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

EMOTIONAL CORRELATES 123

variables had a more profound effect on the severity of worry than either variable alone. There is also evidence of the specificity of meta-worry to GAD. Individuals with GAD scored higher on measures of meta-worry than did nonclinical control participants and nonworried anxious controls (Davis & Valentiner, 2000), as well as individuals with social phobia, panic disorder, depression (Wells & Carter, 2001), and OCD (Barahmand, 2009). Clearly, the role of metacognitive factors in GAD has provided a useful way of understanding why individuals with GAD routinely choose worry as a coping strategy (Coles & Heimberg, 2005).

Emotional Correlates

Some research suggests that individuals with GAD may have difficulties in regulating their emotional experience (Mennin, Heimberg, Turk, & Fresco, 2005). More specifically, individuals with GAD may experience more intense emotions, have more difficulty naming and understanding emotions, have more difficulty accepting their emotional experience (especially when this involves negative emotions), and demonstrate difficulty managing and regulating negative emotions when they do occur (Mennin et al., 2005). As a result of these difficulties, individuals often use strategies such as worry in the face of discomforting negative emotions. As worry becomes a more consistently used strategy, individuals have even more difficulty identifying and understanding their underlying emotional experience, motivating further emotional avoidance through worry. Initial research supports the notion that individuals with GAD symptoms, as well as individuals with a clinical diagnosis of GAD, experience more intense emotions than controls, have a more negative reaction to these emotions, and have fewer strategies to manage their emotions (Mennin et al., 2005). Although each of these emotion regulation deficits appears to be present in other disorders, heightened intensity of emotion is a particularly strong predictor of GAD, differentiating GAD from social anxiety and major depressive disorder and suggesting that there may be some disorder specificity in emotion regulation deficits (Mennin, Holaway, Fresco, Moore, & Heimberg, 2007).

Individuals with GAD report being more fearful of depression and having more intense emotions than do those with social phobia and nonanxious controls (Turk, Heimberg, Luterek, Mennin, & Fresco, 2005). In an experimental study in which fear of anxiety was manipulated, nonclinical participants in the increased fear of anxiety condition reported significantly greater worry following an anxiety-provoking task than those in the decreased fear of anxiety group (Buhr & Dugas, 2009). Further, self-reported worry has been associated with general and specific emotion regulation deficits, including the ability to engage in goal-directed activity when distressed, control over impulses, acceptance of emotions, clarity of emotional state, and access to self-regulation strategies, even when controlling for general negative affect (Salters-Pedneault, Roemer, Tull, Rucker, & Mennin, 2006). Finally, Mennin et al. (2005) found that students meeting criteria for GAD had a stronger physiological reaction to a mood induction than did controls, and demonstrated more difficulty managing this reaction (i.e., individuals endorsed that they had more difficulty accepting and influencing their emotional state than did controls). Thus, there is preliminary evidence that emotion regulation deficits may be an important variable for understanding the chronic worry seen in GAD.

Individuals with GAD may also find their worry process affected by preexisting negative emotional states. Davey and colleagues (e.g., Startup & Davey, 2001, 2003)

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

124 GENERALIZED ANXIETY DISORDER

have found evidence for a “mood-as-input” approach to understanding perseverative, catastrophic worry like that found in GAD. They suggest that individuals who have pathological worry use rules in order to decide when to stop worrying. The stop rules they have identified are (a) “feel like continuing,” which is when individuals continue to try to solve a problem until they do not feel like continuing to work on it, and (b) “as many as can,” which is when individuals persist with problem-solving attempts until they feel like they have generated as many possible responses as they possibly can (Davey, 2006). When the “as many as can” stop rule is paired with negative mood, the person is unlikely to feel that he or she has generated as many possible responses as possible after reasonable effort, leading to perseveration on a given task. Specifically applied to pathological worry, Davey (2006) suggests that worriers are likely candidates for perseveration. He notes that worriers are often in a negative mood state, the task of worrying has no obvious end point, and worriers often apply an “as many as can” rule to determine when to stop worrying. Thus, this model predicts that pathological worriers will frequently be subject to entering an iterative worry cycle that is not easily exited.

This group of researchers has provided empirical support for the assertion that a negative mood induction actually increased the number of catastrophizing steps individuals made when asked to worry about a particular topic. Worriers who were asked to use an “as many as can” stop rule generated more catastrophizing steps than nonworriers (Startup & Davey, 2001). In a sample of worriers, the interaction of negative mood and increased responsibility yielded greater catastrophizing, suggesting that levels of inflated responsibility may be one reason worriers use an “as many as can” stop rule (Startup & Davey, 2003).

Research has investigated differences between worry and related symptoms such as rumination and obsessions. Researchers have long noted the similarities between these cognitive phenomena. Worry and rumination both involve repetitive, negative thinking and are correlated in nonclinical (Fresco, Frankel, Mennin, Turk, & Heimberg, 2002) and clinical samples (Segerstrom, Tsao, Alden, & Craske, 2000). On a measure assessing numerous ways in which these cognitive processes might differ, only seven significant differences emerged between ruminative and worrisome thoughts (Watkins, Moulds, & Mackintosh, 2005). Specifically, worries were reported as having a longer duration, the content of worries was more strongly associated with future events than past events, worries were rated as more disturbing and less realistic, and worries were associated with greater feelings of insecurity. The authors suggest that worry, while similar in process to rumination, involves more future-oriented, threat-relevant content. More recent research suggests that worry is related to both anxious and depressive symptoms, whereas rumination is more uniquely associated with depression (Hong, 2007). In that study, the perceived inability to cope mediated the relationship between worry and anxious and depressed mood. There is also evidence for physiological differences in the manifestation of worry and rumination, with worry, but not rumination, having a strong negative relationship with heart rate variability across emotional contexts (Aldao, Mennin, & McLaughlin, in press). Thus, although worry and rumination share core features, they appear to be functionally different constructs.

Though some overlap exists, research suggests that worry and obsessional thinking are similar but distinct processes. Comer, Kendall, Franklin, Hudson, and Pimentel (2004) reviewed the similarities and differences in process, form, content, and metacognitive beliefs between obsessions and worry. They concluded that worrisome thoughts can be best differentiated from obsessions by the presence of an identifiable trigger, less mental

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

EMOTIONAL CORRELATES 125

imagery, and metacognitive beliefs about the dangerousness and uncontrollability of worry. Langlois, Freeston, and Ladouceur (2000) found that nonclinical participants rated their worrisome thoughts as being more realistic, less ego-dystonic, more persistent, and more verbally oriented than obsessions. Worry appears to be more related to reactive obsessions (i.e., relatively realistic thoughts that are more triggered by environmental cues and are associated with perceived negative consequences) than to autogenous obsessions (i.e., strongly ego-dystonic thoughts that are less related to environmental triggers); see Lee, Lee, Kim, Kwon, and Telch (2005). Thus, although similarities exist between worry and related symptoms such as rumination and obsessions, worry appears to be a distinct cognitive phenomenon.

ENVIRONMENTAL AND FAMILY FACTORS

Environmental and family factors appear to play a number of roles in understanding the development, phenomenology, and maintenance of GAD. For example, among individuals with a propensity to worry, actual situations of enhanced risk appear to play a role in triggering a tendency to overestimate the likelihood of risk associated with a task (Constans, 2001).

Factors related to family environment and parenting styles may also be relevant for understanding GAD. Although most studies on this topic have relied on self-report or retrospective reports of family environments, results converge to suggest that people with GAD may have experienced unpleasant, negative, and rejecting family environments. For example, a study of self-reported parenting styles and GAD symptoms found a relationship between perceived parental alienation and rejection with GAD symptoms in a sample of community adolescents (Hale, Engels, & Meeus, 2006). Further, college students who met self-reported criteria for GAD endorsed less secure attachment to their parents than did control participants (Eng & Heimberg, 2006). A recent longitudinal study found a bidirectional relationship between perceived father-adolescent attachment relationships and GAD symptoms, but a unidirectional relationship for the mother- adolescent attachment relationship, with GAD symptoms leading to a lower-quality relationship (van Eijck, Branje, Hale, & Meeus, 2012). Although this study relies on adolescents’ self-reported GAD symptoms and perceived attachment, it suggests that the effects of attachment are more nuanced than previously believed.

Other researchers have argued that interpersonal problems, in general, may play a causal role in the development of GAD. For example, Crits-Christoph, Gibbons, Narducci, Schamberger, and Gallop (2005) have suggested that there is a great deal of indirect evidence that individuals with GAD have interpersonal issues, including a high frequency of interpersonal themes in worry content and biases to social threat cues. Research suggests that those with GAD show more parent–child boundary problems than do individuals without GAD (Cassidy & Shaver, 1999). Drawing from these ideas, these investigators examined the change in interpersonal problems across the course of brief psychodynamic treatment for GAD and found that changes in interpersonal problems were significantly related to improvements in worry symptoms. Similarly, Borkovec, Newman, Pincus, and Lytle (2002) found that a subset of participants in a CBT trial for GAD demonstrated interpersonal difficulties at posttreatment, and that these difficulties were associated with fewer gains and improvements across follow-up. However, other research suggests that interpersonal problems in GAD may be more focused, with no evidence of disrupted peer relationships in a college sample meeting criteria for GAD

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

126 GENERALIZED ANXIETY DISORDER

(Eng & Heimberg, 2006). The role of interpersonal difficulties in GAD, therefore, is not clearly understood, including whether interpersonal difficulties are a cause or a consequence of excessive worry and anxiety.

Assessment

As noted earlier, the inherent instability in the diagnostic criteria for GAD has made it difficult to reliably diagnose and assess these disorders. As diagnostic criteria have become more stable, reliable means of assessing GAD and chronic worry have also become more available. It has been suggested that a comprehensive assessment of GAD should include the following: worry, beliefs about worry, intolerance of uncertainty, anxiety, associated features (e.g., tension), comorbid symptoms, goals and areas of behavioral inactivation, and emotional avoidance (Roemer & Medaglia, 2001). In this section, we review several useful tools for the assessment of GAD according to DSM-IV diagnostic criteria, including those used to assess these constructs. No changes were made in DSM-5 for GAD so these instruments remain appropriate for that diagnosis, though the instruments will need to be modified in terms of other diagnoses.

Two of the most extensively studied semistructured interviews for diagnosing anxiety- related problems including GAD are the Anxiety Disorders Interview Schedule for DSM-IV (ADIS-IV; Brown, Di Nardo, & Barlow, 2004) and the Structured Clinical Interview for DSM-IV/Axis I Disorders (SCID-IV; First, Spitzer, Gibbon, & Williams, 2007). Both the SCID-IV and ADIS-IV provide systematic questions to establish a reliable current diagnosis of GAD and comorbid conditions (Brown, Di Nardo, Lehman, & Campbell, 2001). Semistructured interviews also provide information about the content of a person’s worry, areas of avoidance or behavioral inactivation, and the presence of comorbid conditions.

ANXIETY DISORDERS INTERVIEW SCHEDULE FOR DSM-IV (ADIS-IV)

The ADIS-IV (Brown et al., 2004) is a clinician-administered semistructured interview that provides both diagnostic and dimensional information about a range of psycholog- ical problems, including anxiety disorders, mood disorders, somatoform disorders, and substance use disorders. Screening questions are provided for psychotic disorders, con- version symptoms, and the presence of a family history of psychiatric illness. Diagnoses are assigned based on DSM-IV criteria. Depending on the version of the ADIS-IV used (standard vs. lifetime version), current and lifetime diagnoses can be assigned. Clinicians require extensive training in the administration of this interview, and the interview duration can be lengthy, particularly for the lifetime version (i.e., several hours). Despite these drawbacks for everyday practice, the ADIS-IV has the benefit of providing clear criteria to help determine the presence or absence of GAD (as well as common comorbid disorders).

The ADIS-IV has demonstrated good reliability. A study by Brown and colleagues examined the interrater reliability of diagnostic decisions made using the ADIS-IV (Brown et al., 2001). Interrater reliability was good both when GAD was diagnosed as the primary problem (kappa = .67) as well as when it was diagnosed as an additional clinical problem (kappa = .65). Sources of unreliability included one interviewer diagnosing GAD

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

ASSESSMENT 127

while the other diagnosed a disorder with overlapping clinical features (e.g., depression or anxiety disorder not otherwise specified—subthreshold GAD), as well as shifts in patient report. More recently, Gordon and Heimberg (2011) examined reliability and validity of particular features of GAD using the ADIS-IV. Interrater reliability was strongest when assessing excessiveness and uncontrollability of worry, interference, and severity of fatigue, but was lowest when evaluating subjective distress. There was generally poor internal consistency with the six associated symptoms, but a high correlation between fatigue and concentration difficulties with the majority of other GAD features, suggesting that these two associated symptoms may be integral to the diagnosis. The authors concluded that GAD can be reliably diagnosed using the ADIS-IV, though there are limitations with reliability of the independent features of the disorder.

STRUCTURED CLINICAL INTERVIEW FOR DSM-IV (SCID-IV)

The SCID-IV (First et al., 2007) is also a clinician-administered semistructured inter- view that provides diagnostic decisions about a wide range of psychiatric disorders. Two versions are available—a clinician version (SCID-CV) and a research version (SCID-I). The clinician version was designed for use in clinical settings and has a less extensive coverage of disorders. The research version has a broader focus and provides for assessment of mood disorders, anxiety disorders, somatoform disorders, substance disorders, eating disorders, adjustment disorders, and psychotic disorders. Current and lifetime diagnoses are obtained. Extensive training is also required to administer the SCID-IV, and administration can be lengthy, especially for the research version (i.e., 2 to 3 hours for a typical outpatient administration).

Studies using DSM-III-R criteria suggest that the SCID demonstrates adequate interrater reliability for GAD (kappa = .56) (Williams et al., 1992). A more recent study found interrater reliability of .63 and test-retest reliability of .44 for GAD using the SCID-IV (Zanarini et al., 2000). Thus, reliability is adequate but not strong when using the SCID to diagnose GAD.

SELF-REPORT MEASURES

Self-report measures of GAD have been developed for a number of purposes, including assessing the presence of diagnostic criteria, the severity of worry, and the range and content of worry topics. Further, there are a number of self-report measures designed to measure constructs thought to be important in the theoretical understanding of this disorder (e.g., measures of intolerance of uncertainty and meta-worry). In this section, we briefly review some of the most often used self-report measures of worry and GAD.

Generalized Anxiety Disorder Questionnaire

The Generalized Anxiety Disorder Questionnaire (GADQ-IV; Newman et al., 2002) is a self-report measure designed to assess the presence of DSM-IV criteria for GAD. It is a brief and simple measure to administer and score, making it a popular tool in research involving analogue samples. Most questions involve “yes/no” responses, two questions involve a severity scale, and one question is open-ended, asking for a list of frequent worries. Thus, the GADQ-IV provides both diagnostic and content-related information about a person’s symptoms. The psychometric properties of this measure are quite strong,

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

128 GENERALIZED ANXIETY DISORDER

as it has demonstrated good test-retest reliability and good convergence with diagnoses made using the ADIS-IV and with scores on self-report measures of worry, while showing weaker relationships with theoretically distinct measures (Newman et al., 2002). The GADQ-IV has also demonstrated strong internal consistency and good convergence with diagnostic decisions made using the ADIS-IV in a clinical sample (Luterek, Turk, Heimberg, Fresco, & Mennin, 2002). The GADQ-IV is an efficient and useful tool to assess the presence of the diagnostic criteria of GAD, especially when it is not practical to use a semistructured interview.

Generalized Anxiety Disorder—7-Item Scale

The Generalized Anxiety Disorder—7-Item Scale (GAD-7; Spitzer, Kroenke, Williams, & Löwe, 2006) is a brief self-report screening measure of DSM-IV GAD diagnosis and symptom severity intended for use in primary care settings. Individuals rate how often over the prior 2 weeks they have been bothered by symptoms on a scale of 0 (not at all) to 3 (nearly every day), with total scores ranging from 0 to 21. Based on the development and validation study (Spitzer et al., 2006), a total score of 10 or greater provides the optimal cut score for identifying cases of GAD. Further, the GAD-7 provides an index of symptom severity, with cases identified as mild (5–9), moderate (10–14), or severe (15–21), which corresponds with increases in functional impairment and disability. The scale has good psychometric properties when used with patients in primary care (Spitzer et al., 2006), community (Löwe et al., 2008), and partial hospital settings (Kertz, Bigda-Peyton, Rosmarin, & Björgvinsson, 2012), although specificity was poor for the highly comorbid partial hospital patients, indicating that it may not be an adequate GAD screening measure with this population.

Self-Report Measures of Worry Severity and Worry Content

There are a few commonly used self-report measures of the general features of worry. For example, the Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990) is a brief self-report measure that was designed to measure a person’s tendency to worry excessively, without assessing the content of the worry. It is a widely used measure that has demonstrated strong psychometric properties (see Roemer, 2001, for a review). In contrast, the Worry Domains Questionnaire (WDQ; Tallis, Eysenck, & Mathews, 1992) provides an assessment of the content of worry. More specifically, it assesses the degree to which a person worries about relationships, lack of confidence, the future, work, and finances. Its psychometric properties are also strong (Roemer, 2001). The content of worry can also be assessed with the Anxious Thoughts Inventory (AnTI; Wells, 1994), which assesses social worry, health worry, and meta-worry. As reviewed earlier, social and health worry are seen as Type 1 worries whereas meta-worry is a Type 2 worry (i.e., having concerns about the consequences of one’s worries on mental health, etc.). Consistent with this idea, scores on the meta-worry subscale have been shown to predict scores on the PSWQ, even after trait anxiety, social worry, and health worry are controlled (Wells & Carter, 1999).

Self-Report Measures of Theoretical Constructs Relevant to GAD

In addition to the meta-worry subscale of the AnTI, several other self-report measures assess theoretical constructs believed to be important in understanding a person’s GAD symptoms. For example, other metacognitions about worry can be assessed by the Meta- Cognitions Questionnaire (MCQ; Cartwright-Hatton & Wells, 1997), and its 30-item short

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .