Generalized Anxiety Disorder 393
the nature of WorryWorry is ubiquitous to the human condition. Who among us has not been preoccupied
about some important task that we face or worried about an anticipated negative or
threatening future situation? Borkovec (1985) noted that because human beings have
the cognitive capacity to create mental representations of past events as well as antici-
pated future events in order to plan and problem solve, we are able to generate internal
representations of future aversive events that cause anxiety in the absence of existing
threat. It is this capacity to symbolically represent threat that is the basis of worry. The
worry experience stems from the production of thoughts and images of exaggerated
anticipation of possible negative outcomes. It is an important component of trait anxi-
ety or neuroticism and can be considered the cognitive component of anxiety (Eysenck,
1992), although worry is distinguishable from anxiety (i.e., Brown et al., 1998; Ruscio,
2002; Mathews, 1990; Zinbarg & Barlow, 1996). Nevertheless, since the publication
of DSM-III-R (APA, 1987), excessive worry is now considered the cardinal feature of
GAD.
Defining Worry
Borkovec and colleagues offered one the earliest definitions of worry that has become
widely accepted in research on generalized anxiety: “Worry is a chain of thoughts and
images, negatively affect-laden and relatively uncontrollable. The worry process rep-
resents 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 fear processes” (Borkovec, Robinson, Pruzinsky, & DePree, 1983,
p. 10). However, in the intervening years a more complicated picture has emerged on the
nature of worry. Worry is predominantly a verbal– linguistic cognitive phenomenon that
may serve an avoidant coping function by suppressing somatic and negative emotional
responses to internally represented threat cues (Borkovec, 1994; Sibrava & Borkovec,
2006). Mathews (1990) defined worry as a “persistent awareness of possible future dan-
ger, which is repeatedly rehearsed without being resolved” (p. 456) and is maintained by
an automatic information- processing bias for threat that underlies high trait anxiety and
vulnerability to GAD. Wells (1999) argued that worry is an intrusive ideational process
that is predominantly ego- syntonic and is perceived as serving an adaptive function.
One of the most important debates about worry as it relates to GAD is whether
worry can be constructive and adaptive whereas the excessive worry in GAD is clearly
maladaptive and pathological. Some researchers have argued that worry can lead to
effective problem solving of stressful life events because it involves problem- focused
active coping, information seeking, and task orientation with at most a minimal level
of associated anxiety (e.g., Davey, 1994; Wells, 1999; see Watkins, 2008). On the other
hand, pathological worry is (1) more pervasive, (2) time- consuming, (3) uncontrolla-
ble, (4) focused on more minor matters and more remote but personal future- oriented
situations, (5) selectively biased for threat, and (6) associated with greater restricted
autonomic variability (Craske et al., 1989; Dugas, Gagnon, et al., 1998; Dupuy et al.,
2001; see also Ruscio et al., 2001). However, attempts to delineate adaptive worry from
pathological worry will be difficult because of the strong association between worry
and heightened anxiety (Roemer et al., 2002) and the dimensional nature of normal and