Vulnerability to Anxiety 117
clusion reached by Mathews and MacLeod (2002) is that high trait- anxious individuals
have a cognitive vulnerability to anxiety in the form of a lower threshold for switching
from an avoidance to a vigilant information- processing mode.
A second source of supportive evidence for a schematic vulnerability to anxiety
comes from the anxiety sensitivity and diminished control studies reviewed in this chap-
ter. Although it would be inaccurate to describe the ASI as a beliefs measure, it does
assess appraisals that are based on a variety of preexisting beliefs about physical sen-
sations and anxiety. For example, the ASI item “It scares me when I become short of
breath” would be based on a preexisting belief such as “I am putting myself at severe
risk of being unable to breath when I feel short of breath.” If high ASI scores pre-
dict elevated likelihood of subsequent anxiety, we can generalize from these findings to
the beliefs that underlie ASI appraisals as supportive evidence that these beliefs consti-
tute vulnerability for anxiety. The same generalization can be made from the research
on diminished control and negative attributional style in anxiety. Certain preexisting
beliefs about lack of control over anticipated threats will underlie control perceptions,
making these beliefs an important element in the proposal that diminished sense of
personal control is a vulnerability factor in anxiety. To summarize, the notion of pre-
existing dysfunctional beliefs that predispose to anxiety is a common feature of many
cognitive theories of anxiety disorders (e.g., D. A. Clark, 2004; Ehlers & Clark, 2000;
Wells, 2000; Wells & Clark, 1997).
DYSFUNCTIoNAL ANXIETY bELIEFS
In order to investigate the role of dysfunctional beliefs in the etiology of anxiety, specific
belief measures are needed that directly assess threat schema content. Unfortunately,
research in this area is not as well developed as the experimental studies on attentional
bias or the brief prospective diathesis– stress studies found in depression. Nevertheless,
we are beginning to see more research on the role of threat- relevant schemas and beliefs
in clinical anxiety.
In recent years there has been considerable research on the belief structure of OCD.
An international group of researchers called the Obsessive Compulsive Cognitions
Working Group (OCCWG) proposed six belief domains as constituting a cognitive vul-
nerability to OCD: inflated responsibility, overcontrol of thoughts, overimportance of
thoughts, overestimated threat, perfectionism, and intolerance of uncertainty (OCCWG,
1997). Definitions of these belief domains can be found in Table 11.3.
An 87-item self- report questionnaire, the Obsessive Beliefs Questionnaire (OBQ),
was developed to assess the six OCD belief domains. Later factor analysis indicated
it could be reduced to 44 items that assessed three belief dimensions: responsibility/
threat estimation, perfectionism/intolerance of uncertainty, and importance/control
of thoughts (OCCWG, 2005). Two large-scale multisite clinical studies based on the
87-item OBQ revealed that OCD patients scored significantly higher than other non-
obsessional anxious and nonclinical comparison groups on OBQ Control of Thoughts,
Importance of Thoughts, and Responsibility subscales, in particular, and the six OBQ
belief scales correlated better with self- reported OCD measures than with the BAI or
BDI (OCCWG, 2001, 2003; see Steketee, Frost, & Cohen, 1998, for similar results).
However, the six OBQ subscales are highly intercorrelated and they have strong correla-
tions with other non-OCD measures like the Penn State Worry Questionnaire. At pres-