Generalized Anxiety Disorder 409
nerability beliefs that take into account the current concerns and life tasks of the indi-
vidual. Negative beliefs about ambiguity and uncertainty feature strongly in GAD but
are unlikely to be specific to the disorder. And it is clear that negative beliefs about
worry are more pathognomonic to GAD than the positive beliefs. However, research
into these constructs is still preliminary, and more experimental research is needed to
determine how these beliefs might interact with other cognitive processes that contrib-
ute to the persistence of worry.
Clinician Guideline 10.10
Cognitive therapy of GAD must target personal beliefs about the perceived threat and nega-
tive consequences associated with ambiguous and uncertain future negative outcomes as
well as negative beliefs about the uncontrollable and dangerous nature of worry.Hypothesis 3
Individuals with GAD will exhibit an automatic attention and interpretation threat bias
when processing information relevant to valued goals and personal life concerns.
The proposition that an automatic preferential encoding, interpretation, and
retrieval bias for threat is a causal contributor to the development and maintenance of
GAD is a central aspect of most cognitive theories of GAD (MacLeod & Rutherford,
2004). Numerous information- processing experiments have been conducted that are
supportive of this hypothesis; much of this material is reviewed in Chapter 3. In this sec-
tion we take a brief look at encoding and interpretation of ambiguity studies that have
utilized GAD samples.
There is strong empirical support for an encoding bias for threat in GAD patients
and high trait- anxious individuals that occurs both at the automatic and the elabora-
tive processing levels. Various studies using the emotional Stroop task have found that
color- naming latencies for threat stimuli were significantly longer for GAD or high trait-
anxious individuals than nonanxious groups (e.g., Bradley, Mogg, et al., 1995; Edwards,
Burt, & Lipp, 2006; Martin et al., 1991; Mogg, Bradley, et al., 1995; Mogg et al., 1993;
Richards et al., 1992; Rutherford, MacLeod, & Campbell, 2004). Moreover, threat bias
is apparent at both subliminal and supraliminal levels but exposure to a current stressor
may enhance automatic but not elaborative threat bias for high trait- anxious individuals
(Edwards et al., 2006). In addition there is evidence that the encoding bias in GAD may
not be specific to threat but to negative information more generally (Martin et al., 1991;
Mogg, Bradley, et al., 1995; Mogg et al., 1993; Rutherford et al., 2004).
A number of visual and semantic dot probe experiments have found an automatic
attentional vigilance for threat in GAD patients (e.g., MacLeod et al., 1986; Mogg,
Bradley, & Williams, 1995; Mogg et al., 1992) as well as high trait- anxious individuals
(e.g., Koster et al., 2006; Mogg et al., 2000; Wilson & MacLeod, 2003). However, at
slower presentation rates, high trait- anxious individuals may show attentional avoid-
ance of threat (Koster et al., 2006) and findings by Wilson and MacLeod (2003) suggest
that high trait- anxious individuals may show disproportionate vigilance for threat only
at moderate levels of threat intensity (see Chapter 3, Figure 3.3, for further discussion).
In addition it has been suggested that the dot probe effects may be partly explained by