412 TREATMENT OF SPECIFIC ANXIETY DISORDERS
empirical research relevant to Hypothesis 4 is still preliminary, these early findings are
encouraging for further exploration of the role of metacognitive processing in GAD.
Clinician Guideline 10.12
Individuals with GAD will engage in extensive catastrophizing and consider their worri-
some thoughts as dangerous and uncontrollable. The cognitive therapist must address this
faulty appraisal process in order to achieve the desired therapeutic gains over worry.Hypothesis 5
Individuals with GAD (1) will expend greater effort toward disengagement from or
suppression of worry, (2) are more likely to use faulty control strategies, and (3) will
experience less success controlling their worry than nonanxious worriers.
Given that individuals with GAD tend to appraise their worrisome thoughts as
disturbing and associated with a greater likelihood of negative outcomes, Hypothesis
5 is a natural extension of the previous hypothesis. According to the cognitive model
illustrated in Figure 10.1, we predict unsuccessful and futile efforts to control or sup-
press worry will paradoxically contribute to its persistence, in accord with Wegner’s
ironic process theory of suppression (Wegner, 1994; Wenzlaff & Wegner, 2000). As
predicted by Hypothesis 5, researchers have consistently found that GAD is character-
ized by a heightened subjective experience of worry as an uncontrollable process and
any efforts at control prove futile and unproductive (Craske et al., 1989; Hoyer et al.,
2001; Wells & Morrison, 1994). Despite their acknowledged inability to control worry,
it is interesting that individuals with GAD are highly invested in continuing with their
efforts toward gaining control over worry (Hoyer et al., 2001).
There is now some evidence that deliberate attempts to suppress worrisome thoughts
can have adverse effects on the worry process. For example, we found that university
students instructed to suppress thoughts of failing an exam experienced a rebound in
worry once suppression efforts ceased (Wang & Clark, 2008). Becker, Rinck, Roth,
and Margraf (1998) also found evidence of impaired mental control with GAD patients
having less success suppressing their main worry than speech phobic and nonanxious
controls. However, other studies have failed to find adverse effects with attempts to sup-
press worry (e.g., Mathews & Milroy, 1994; McLean & Broomfield, 2007).
It is possible that thought suppression might not directly influence the frequency
of worry but instead have other untoward effects on the worry experience. Harvey
(2003) reported that individuals with insomnia try to suppress and control their intru-
sive thoughts and worry during the presleep period more than good sleepers. In addi-
tion, individuals instructed to suppress their presleep worrisome thoughts experienced
longer sleep-onset latency and poorer quality of sleep but did not report more unwanted
worrisome thought intrusions. In a recent study, high trait- anxious students were ran-
domly assigned to suppress previously presented threat and neutral words, concentrate
on the words, or just allow the thoughts to wander (Kircanski, Craske, & Bjork, 2008).
Analyses revealed that suppression of threat words generated an enhanced explicit mem-