Generalized Anxiety Disorder 391
related to the co- occurring anxiety disorder. Subsequent research based on the DSM-
III-R criteria supported the central role of worry but revealed that autonomic hyperac-
tivity is the least reliable and frequently endorsed of the GAD symptoms (Roemer et al.,
2002). Thus DSM-IV (APA, 1994) introduced a further revision in which the number
of physical symptoms of anxiety needed to meet diagnostic criteria was reduced from
six out of 18 to three out of six symptoms. Although this led to an improvement in the
reliable diagnosis of GAD, many of these physical symptoms overlap with depression,
making differential diagnosis with major depression more difficult (see Roemer et al.,
2002, for discussion). For example, we found that two- thirds of our sample with GAD
was misclassified as having major depression or panic disorder based on a discriminant
function analysis of common symptom and cognition measures of anxiety and depres-
sion (D. A. Clark, Beck, & Beck, 1994). Unfortunately, GAD lacks symptom specificity
which can make it difficult to differentiate from other disorders.
Generalized Anxiety and Depression
In recent years there has been considerable debate among researchers on whether GAD
is an anxiety disorder or whether it fits more closely with the affective disorders, espe-
cially major depression. Although it has been argued that GAD may be the basic “anxi-
ety disorder” because worry, its central feature, is common across the anxiety disorders
(Roemer et al., 2002), many others have questioned the diagnostic distinctiveness of
GAD because none of its features are exclusive or specific to the disorder (Rachman,
2004). Moreover, GAD appears to have a particularly close relationship with depres-
sion. High comorbidity rates for GAD and major depression have been reported in the
NCS-R (Kessler, Chiu, et al., 2005), as well as in large-scale surveys of primary care
practice (Olfson et al., 2000). In a large sample of treatment- seekers, 40% of individu-
als with GAD had a secondary mood disorder and the rate jumped to 74% for lifetime
co- occurrence (Brown, Campbell, et al., 2001; see also Mofitt et al., 2007). Moreover,
there was no temporal priority of one disorder over the other.
Research on symptom structure indicates there is a great deal of overlap between
GAD and major depression, with GAD having the highest associations of the anxiety
disorders with the higher order nonspecific negative affect (NA) dimension and minimal
or reverse associations with autonomic arousal (Brown, Chorpita, & Barlow, 1998;
McGlinchey & Zimmerman, 2007; see also Krueger, 1999). Mineka et al. (1998) pro-
posed that GAD and major depression are both distress-based disorders that contain a
large nonspecific NA component. More recently Watson (2005) concluded that GAD is
misplaced within the anxiety disorders because GAD and major depression are indistin-
guishable phenotypically and genetically. He recommended reconceptualizing DSM-IV
anxiety and depression into a quantitative hierarchical organization with major depres-
sion, dysthymia, GAD, and PTSD categorized together as distress disorders. In support
of this view there is evidence that panic disorder can be distinguished from GAD and
major depression by its close association with physiological hyperarousal (e.g., Joiner et
al., 1999).
On the other hand, there is a large body of cognitive research showing that GAD is
associated with an automatic attentional bias for threat (see Chapter 3 and discussion
below) and that worry is distinct but intricately related to anxious apprehension and fear
(Barlow, 2002). Thus we contend that strong arguments exist for retaining GAD within
the anxiety disorders but we also must recognize that it has a much closer relationship