Vulnerability to Anxiety 105
nitive perspective with greater specificity to anxiety and its disorders. Anxiety sensitiv-
ity (AS) is the fear of anxiety- related bodily sensations based on enduring beliefs that
negative physical, social, or psychological consequences might result from these anxious
symptoms (Reiss, 1991; Reiss & McNally, 1985; Taylor, 1995a; Taylor & Cox, 1998).
For example, a person with high AS might interpret chest pain as sign of an immi-
nent heart attack and so feel highly anxious when experiencing this bodily sensation,
whereas a person with low AS might interpret the chest pain as muscle tension due to
physical exertion and so experience no anxiety with the bodily sensation.
A propensity to feel anxious about certain bodily symptoms is present in high AS
because individuals believe anxiety and its physical symptoms can lead to serious con-
sequences like heart attacks, mental illness, or intolerable anxiety (Reiss, 1991). Thus
AS is a personality variable that amplifies fear when anxiety sensations and behaviors
are experienced (Reiss, 1997). In this way it is thought to play both an etiological and a
maintaining role in all the anxiety disorders, but particularly panic disorder and agora-
phobia (Reiss, 1991; Taylor & Cox, 1998).
Psychometric Validation
The 16-item Anxiety Sensitivity Index (ASI) is the primary measure for assessing indi-
vidual differences in AS (Reiss, Peterson, Gursky, & McNally, 1986; Reiss & McNally,
1985). Despite considerable debate over its factorial structure, it now appears that the
ASI is a hierarchical multidimensional construct with two or three correlated lower
order factors (i.e., Fears of Mental Catastrophe vs. Fears of Cardiopulmonary Sensa-
tions or Physical Concerns, Mental Incapacitation, and Social Concerns about Being
Anxious) linked to a higher order general factor of AS (Mohlman & Zinbarg, 2000;
Schmidt & Joiner, 2002; Zinbarg, Barlow, & Brown, 1997). There is also controversy
over which dimensions best describe AS. Based on a 36-item ASI-R, only two correlated
factors were replicated across data sets drawn from six countries: Fear of Somatic Symp-
toms and Social- Cognitive Concerns (Zolensky et al., 2003).
The most recent revision of the ASI, the 18-item ASI-3, may provide the best assess-
ment of the three AS dimensions; physical, cognitive, and social concerns (Taylor, Zvo-
lensky, et al., 2007). The ASI-3 subscales had improved internal consistency and good
criterion- related validity, although the three subscales were very highly correlated (r’s >
.83). Nevertheless, findings across the various versions of the ASI indicate that subscales
rather than a total score should be utilized to indicate level of AS.
The ASI measures have good internal consistency, test– retest reliability, and strong
convergent validity with other measures of anxiety (Mohlman & Zinbarg, 2000; Reiss
et al., 1986; Taylor & Cox, 1998; Zvolensky et al., 2003). Moreover, the AS lower order
dimensions are generally consistent across various countries (Bernstein et al., 2006; Zvo-
lensky et al., 2003), although there is some evidence that high AS scores may decrease
over time even in the absence of a specific intervention (Gardenswartz & Craske, 2001;
Maltby, 2001; Maltby, Mayers, Allen, & Tolin, 2005). There has been considerable
debate on whether AS is distinct from trait anxiety (for discussion, see Lilienfeld, 1996;
Lilienfeld, Jacob, & Turner, 1989; McNally, 1994). The current view is that AS is a dis-
tinct lower order construct hierarchically linked to the broader personality disposition
of trait anxiety (Reiss, 1997; Taylor, 1995a).