Empirical Status of the Cognitive Model 81
high negative affect, which is a mood– personality dimension common to all emotional
disorders.
There is evidence that specificity may only apply to a subset of anxious cognitions.
Jolly and Dykman (1994) reported that some threat cognitions were more related to a
general negativity factor, whereas other cognitions related to physical or health- related
threat were more specific to anxiety. In other research anxious overconcern (i.e., worry)
emerged as a common feature of all anxiety disorders, whereas negative evaluation of
others or social threat may evidence more subtype- specificity (Becker, Namour, Zayfert,
& Hegel, 2001; Mizes, Landolf- Fritsche, & Grossman-McKee, 1987). Finally, Riskind
(1997) has argued that looming vulnerability, the perception of threat movement, may
offer better precision in distinguishing anxiety from depression because it incorporates
time and rate of change in its conceptualization of threat appraisal. Although still ten-
tative, it appears that only certain types of threat- related cognitions such as concerns
about physical symptoms, health, social evaluation, and impending danger are specific
to anxiety, whereas anxious apprehension or worry may be more evident in both anxi-
ety and depression.
Cognitive Specificity in Anxiety Disorder Subtypes
Less research has investigated whether a specific cognitive content is associated with the
anxiety disorder subtypes. In two studies R. Beck and colleagues found that worry was
common to anxiety and depression and a strong predictor of negative affect, whereas
hopelessness was predictive of low positive affect and panic- related cognitions were
clearly specific to anxiety states (R. Beck, Benedict, & Winkler, 2003; R. Beck et al.,
2001). In a confirmatory factor analysis of self- reported anxious and depressive self-
statements, self- statements reflecting depression/hopelessness and self- statements reflect-
ing anxiety/uncertainty about the future had large and significant loadings on a general
negativity factor (Safren et al., 2000).
One of the most direct tests of cognitive content- specificity among anxiety disorder
subtypes was reported by Woody, Taylor, McLean, and Koch (1998). They found that
patients with panic disorder scored significantly higher on a measure of threat- related
cognitions that were unique to panic (i.e., the UBC Cognitions Inventory—Panic sub-
scale) compared to patients with major depression. However, the two groups did not
differ on the Cognitions Checklist— Anxiety subscale, which the authors claim assesses
more general conceptions of anxious cognitions.
Summary
Over the years numerous studies have shown that automatic thoughts and images of
threat, danger, and harm occur with greater frequency and intensity in the anxiety dis-
orders when fear is activated. Consequently there is ample evidence supporting the basic
assertion of Hypothesis 6. Whether thoughts of threat and danger are a specific marker
of anxiety has been more equivocal, and whether each anxiety disorder has its own
unique cognitive content that distinguishes it from other emotional states has not been
subjected to adequate empirical investigation. However, a number of tentative conclu-
sions can be drawn about cognitive content- specificity in anxiety. It is likely that only
some forms of anxious thought will show the level of specificity predicted by Hypothesis