Contrary to the contention of Cox, Cohen, Direnfeld and Swinson (1996,
Behaviour Research and Therapy. 34, 949-954) that the Beck Anxiety In
ventory (BAI; Beck & Steer, 1993, Manual for the Beck Anxiety Inventor
y) measures primarily symptoms associated with panic attacks rather th
an anxiety in general, we propose that the higher level of anxiety fou
nd in patients with panic disorders not only is not an artifact of the
BAI's symptom content, but patients with panic disorders truly have m
ore anxiety than patients with other types of anxiety disorders. Furth
ermore, the BAI contains symptoms present in other anxiety disorders,
besides panic disorder, and specifically includes 11 symptoms of gener
alized anxiety disorder (GAD). The BAI and revised Hamilton Anxiety Ra
ting Scale (HARS-R; Riskind, Beck, Brown & Steer, 1987, Journal of Ner
vous and Mental Disease, 175, 474-479) scores of 274 (69%) outpatients
with panic disorders and 123 (31%) outpatients with GAD were found to
differentiate these two diagnostic groups equally and significantly.
The panic disorder outpatients had higher scores on both the BAI and t
he HARS-R than did the GAD patients. Thus, Cox et al.'s (1996) specula
tion about the BAI's yielding spuriously high levels of anxiety in pat
ients with panic disorders revives an important issue relevant to the
relation of panic disorder to GAD. Copyright (C) 1996 Elsevier Science
Ltd