![]()
PART I
Introducing CBT for GAD
![]()
1
Introducing Generalised Anxiety Disorder
What is Generalised Anxiety Disorder?
Generalised Anxiety Disorder (GAD) is one of the most common mental health problems, and arguably also one of the most misunderstood and misdiagnosed. As a diagnostic category, GAD is relatively recent compared with other anxiety disorders. Historically, GAD was virtually a diagnosis of exclusion: when the client was anxious, but other anxiety disorders were not applicable, they tended to be given the label of GAD. More recently, the diagnostic criteria have been revised and sharpened and GAD can now be understood as a discrete diagnosis in its own right.
Figure 1 below outlines the formal diagnostic criteria for Generalised Anxiety Disorder as defined in the DSM-IV-TR (APA, 2000). Worry and the perceived uncontrollability of worry are considered to be essential features of GAD. Furthermore, there is the observation that the intensity, frequency and duration of the anxiety and worry are disproportionate to the actual likelihood or impact of the subject of the worry.
Prevalence and co-morbidity
GAD affects a significant number of people, both in its own right and in combination with other psychological disorders. In a UK prevalence study, Jenkins et al. (1997) found a current prevalence rate of 3.1 per cent. A US survey of mental health prevalence in the community found a one-year GAD prevalence rate of 3.1 per cent (Wittchen et al., 1994). An Australian survey found a one-year GAD prevalence of 3.6 per cent (Hunt et al., 2002). In keeping with observations of other anxiety disorders, it appears that GAD is more commonly seen in women than it is in men. For example, in two large-scale US studies of mental health prevalence rates in the community, it was found that twice as many women were identified with GAD compared to men (Blazer et al., 1991; Wittchen et al., 1994).
Figure 1 Diagnostic criteria for Generalised Anxiety Disorder.
Source: APA (2000). Reprinted with permission from the Diagnostic and Statistical Manual of Mental Disorders, 4th edn, text revision (copyright 2000). American Psychiatric Association
Further evidence of the prevalence of GAD comes from Katon et al. (1990) who, in a study of high users of medical care, found that of those people reporting significant emotional distress, 21.8 per cent met criteria for a current diagnosis of GAD and 40.3 per cent of them met the diagnostic criteria for GAD at some point in their lives. It is also worth bearing in mind that prevalence studies typically measure rates of a particular disorder against formal diagnostic criteria, such as those in the DSM-IV-TR (APA, 2000) and the ICD-10 (World Health Organization, 1990). However, a client would not need to meet the full diagnostic criteria of either system for the GAD symptoms they have to make a significant impact on their well-being and quality of life.
Not only is GAD a common psychological problem in its own right, it is also often encountered in combination with one or more other psychological disorders. Dugas and Robichaud (2007) note that over 90 per cent of people with GAD will have at least one other disorder, with over half having depression as well as GAD, and almost a third having social phobia as well as GAD. Davey and Wells (2006) also cite clinical studies of GAD showing people with GAD having other disorders, most commonly depression, followed by social phobia, specific phobia and panic disorder. It isnāt only Axis I disorders that often co-occur with GAD; Dyck et al. (2001) report that 37.7 per cent of their patients with GAD also met the diagnostic criteria for one or more personality or Axis II disorders, with avoidant personality disorder being the most frequent.
Worry: the cardinal symptom of GAD
Diagnostically (see Figure 1 above), and in our clinical experience, the cardinal symptom of GAD is worry, and as diagnostic criteria for GAD have evolved, worry has taken an increasingly prominent role. In GAD, the worry is excessive, perceived (by the worrier) to be uncontrollable and difficult to interrupt, and has a significant emotional and physiological impact on the client and their quality of life. In this section, we will look at what worry is, at what distinguishes ānormalā worry from the worry seen in GAD, and at some of the issues that clients tend to worry about.
One of the pioneers of work on worry as a psychological disorder was Professor Tom Borkovec. His work produced one of the early attempts to define worry:
Worry is a chain of thoughts and images, negatively affect-laden and relatively uncontrollable; it represents an attempt to engage in mental problem-solving on an issue whose outcome is uncertain but contains the possibility of one or more negative outcomes; consequently, worry relates closely to the fear process. (Borkovec et al., 1983: 10)
More recently, Sibrava and Borkovec (2006) summarise what they perceive to be the central cognition in GAD:
The world is potentially dangerous and I may not be able to cope with whatever comes from the future, so I must anticipate all bad things that might happen so that I can avoid them or prepare for them. (Sibrava and Borkovec, 2006: 239)
These definitions highlight a number of important characteristics about worry. The chain of thoughts is a succession of fearful scenarios, building upon each other, and chaining into new and increasingly catastrophic directions, each trying to foresee and trouble-shoot multiple catastrophes or disasters. What you will notice, then, is that worriers spend much of their time in a theoretically possible but as yet nonexistent future (Borkovec, 2002). And because of what the worrier is generally thinking about and responding to (emotionally and physically), they are by definition spending less time attending to and engaging with the real world that is around them.
So what do people with GAD worry about? People with GAD tend to worry about health, finances, relationships, family, work, school. Relationships seem to be a consistent source of worry in people with GAD. For example, Dugas et al. (1998a) found that over 70 per cent of people with GAD endorsed frequent worry about family and relationships. And if we look at the major worry themes cited above, all of them could potentially have significance for a personās family and social network.
An important point here is that people with GAD worry about the same issues that any one of us can worry about from time to time. If the themes of worry do not distinguish ānormalā worry from clinical levels of worry, what does? Rather than thinking of normal and pathological worry as separate categories, it is perhaps more helpful to understand them as being at opposite ends of a continuum or spectrum (Ruscio et al., 2001). However, there are some characteristics of āpathologicalā worry that distinguish it from ānormalā worry. Pathological worry is usually about more unlikely and/or remote future events (Dugas et al., 1998a). Compared to pathological worriers, normal worriers generally worry in response to a specific trigger, have a subjective sense of more control over their worry and, compared to pathological worriers, devote less time to worrying (Craske et al., 1989). People without GAD may have times when particular circumstances arise or events unfold and they experience worry. Their worry is generally time-limited: there will come a point when they stop worrying, and at these points their worrying may actually serve a purpose. People with GAD tend to worry about most things, most of the time: they get locked into a spiral of worry and find it difficult to disengage from the worry process.
Beyond worry and other diagnostic criteria: other indicators of GAD
In addition to the diagnostic criteria mentioned above, there are other common symptoms or experiences that are seen clinically. These include experiences such as an upset stomach, headaches, frequent urination or diarrhoea. Some symptoms can be problems in their own right, and these will be dealt with in later chapters, for instance alcohol problems (see Chapter 12). Others are more like behavioural manifestations of worry, such as pacing, smoking more cigarettes or nail biting. There are yet other symptoms that seem to relate to the idea of living in worry rather than in the here and now ā for instance, feeling tuned out or separated from the world, having a sense of impending doom or experiencing psychic or psychological isolation. Importantly, the therapist should think beyond the diagnostic criteria when assessing a clientās problems. On the other hand, we also know that some individuals do not meet full diagnostic criteria but still suffer with high levels of worry, and in such cases these clients may still benefit from treatment.
CBT for GAD: a brief history
In the next chapter, we will outline a cognitive-behavioural model of GAD that has been in development since the early 1990s and that guides the treatment methods we cover in this book. We refer to it as the Laval model of GAD because it was originally developed by clinicians and researchers at lāUniversitĆ© Laval in Canada. However, further work on this model and the derived treatment continues in other research centres around the world. Before looking at the Laval model in detail, it may be helpful to offer a brief history of cognitive therapy for GAD and excessive worry.
Early cognitive-behavioural treatments for GAD were strongly influenced by Beckās work on depression (e.g. Beck, 1967, 1976) and his subsequent applications of it to anxiety (e.g. Beck et al., 1985). This was, therefore, not a model that was designed specifically for GAD, but rather it constituted a generic cognitive model of anxiety.
The generic cognitive model of anxiety was built on the premise that anxious individuals exhibit both a preoccupation with danger and an underestimation of their capacity to cope. This preoccupation with danger is evident in a number of cognitive ways. Firstly, it is evident at the level of negative automatic thoughts. These are appraisals or interpretations of events, often characterised by identifiable cognitive biases, e.g. overgeneralisation, emotional reasoning, personalising, etc. However, the preoccupation with danger is also said to be represented at ādeeperā cognitive levels, such as unconditional beliefs (e.g. āI am āthe world is ...ā) and conditional assumptions, i.e. āif ... thenā statements. Early cognitive-behavioural treatments of anxiety worked with these levels of cognition alongside behavioural approaches such as graded exposure and applied relaxation.
In time, cognitive therapy moved beyond these generic processes and models based on emerging knowledge about specific cognitive and behavioural processes in particular disorders began to emerge, for example panic disorder (Clark, 1986), obsessive-compulsive disorder (Salkovskis, 1985) and social phobia (Clark and Wells, 1995).
Early treatments of GAD, due to limited understanding of the specific psychological processes in the disorder, tended to retain elements of a generic cognitive-behavioural approach. Treatments such as those developed by Gillian Butler and colleagues (e.g. Butler et al., 1987, 1991), Tom Borkovec and colleagues (e.g. Borkovec and Costello, 1993) and David Barlow and colleagues (e.g. Zinbarg et al., 2006) include applied relaxation and incorporate many of Beckās cognitive techniques into their overall treatment package for GAD, with a view to helping the client attain a more realistic perspective on their problems and worries and to modify their perceptions of vulnerability, for example by eliciting probability estimates for the likelihood of the feared situation arising. We could consider these as first-generation treatments.
Second-generation treatments of GAD such as the Laval protocol and the metacognitive model advanced by Adrian Wells (e.g. Wells, 1997) developed these earlier treatments. One could summarise the factors that influence the change in CBT for GAD over the years as an increasing emphasis on cognitive process rather than simply cognitive content. Because the focus of worry shifts in GAD, any techniques that work with a specific concern will have limited use. It may be possible to correct someoneās anxious predictions about Event A, but we know from the nature of GAD that people tend to worry about lots of everyday situations; when one particular situation is re-evaluated, they are very likely to begin worrying about the next. More recent approaches to worry treatment have therefore targeted the cognitive process of worry itself, rather than the content of particular worries.
CBT for GAD: the evidence base
Compared to other anxiety disorders, the evidence base for GAD is relatively small. We have chosen the Laval model and treatment protocol for GAD for the purposes of this book because, although its evidence base is still growing, it is nevertheless arguably stronger than other contemporary models and treatments for GAD at the time of writing. The evidence base is of two types. First, there is a significant evidence base for the different components of the model, both from the initial Laval team and then by various groups led by alumni of that team, particularly Michel Dugas and his research team at Concordia University in Montreal. Second, there are a number of treatment studies, both randomised control trial (RCT) studies and single case designs (with adolescents and older adults), testing the treatment in both individual and group formats against waiting-list and against credible control treatments. As an individual therapy, the Laval protocol was found to be superior to a waiting-list control condition on all outcomes (Ladouceur et al., 2000b). Furthermore, as a group therapy it was also found to decrease the level of worry in the post-treatment phase (Dugas et al., 2003a). In comparison to non-directive therapy, the Laval protocol had a more significant impact on medication discontinuation in long-term benzodiazepine users and led to greater gains in terms of diagnostic remission and symptomatic improvement than the non-directive therapy (Gosselin et al., 2006). Dugas et al. (2009) compared this CBT treatment with applied relaxation and found that, once again, CBT was a successful treatment in comparison to waiting-list controls. The success of the CBT and the applied relaxation were similar; however, unlike the applied relaxation protocol, CBT following the Laval model led to continued improvement during the follow-up period.
There are, of course, other CBT treatment approaches for GAD for which the evidence base is increasing. The first RCT for metacognitive therapy for GAD was published in 2010 (Wells et al., 2010), there is one RCT for an acceptance-based approach to GAD (Roemer et al., 2008), and there are trials of later variants of the basic approach developed by Borkovec and colleagues with various adjunct treatments (e.g. Newman et al., 2008). These studies all support the overall efficacy of CBT approaches for GAD and may provide therapeutic options in some circumstances. However, particularly when developing the knowledge, skill and confidence to treat particular disorders, we believe that it is preferable to learn one treatment approach first. Thus, we have chosen the Laval protocol. It is one of the three treatments for GAD scoped by Roth and Pilling (2...
