The goal of the present study was to compare the efficacy of cognitive-behavioral treatment for panic control alone versus this treatment containing an additional in vivo exposure component. The sample was comprised of 68 individuals who met diagnosis for panic disorder with agoraphobia. Participants were randomly assigned to one of two 16-week treatment conditions, panic control only and panic control with in vivo exposure. Assessments were repeated at baseline, mid-treatment, posttreatment, and 6-month follow-up using diagnostic and behavioral measures. Results indicated that the two treatment conditions were equally efficacious for both panic disorder and agoraphobia. The intervention explicitly targeting agoraphobia appeared superfluous given the efficacy of panic control alone. On the other hand, reduction in panic frequency predicted reduction in agoraphobic avoidance overall. The practical and theoretical implications are discussed, as are limitations and directions for future research.
Previous research on panic disorder and agoraphobia has consistently demonstrated a strong association between occurrence of panic attacks and agoraphobia avoidance. Specifically, it has been shown that, in clinic samples, agoraphobic avoidance mostly develops following panic attacks or panic-like sensations (e.g., Goisman, Warshaw, Peterson, Rogers, et al., 1994; Pollard, Bronson, & Kenney, 1989), that level of avoidance of a specific agoraphobic situation is predicted by anticipatory anxiety surrounding panic attacks (e.g., Cox, Endler, & Swinson, 1995; Craske & Barlow, 1988; Craske, Rapee, & Barlow, 1988), and that more severely agoraphobic individuals perceive more negative consequences of panic attacks in agoraphobic situations (Telch, Brouilard, Telch, Agras, et al., 1989). Given these findings, it is reasonable to hypothesize that control of panic attacks may mediate reductions in agoraphobia, and, therefore, treatment addressing panic attacks alone may be sufficient for ameliorating agoraphobia.
In contrast, research on the context specificity of extinction (Bouton, 1993) and its application to human fear reduction (e.g., Mineka, Mystkowski, Hladek, & Rodriguez, 1999) would suggest that reduction of fear of panic-related bodily cues occurring outside of agoraphobic contexts may not generalize to these situations. Moreover, second-order conditioning experiments have demonstrated the extinction of fear to a first-order conditioned stimulus does not alter expression of fear to a second-order conditioned stimulus. It has been suggested that a second-order stimulus continues to generate fear following extinction to the first-order stimulus due to a difference in the nature of the underlying associations (Rescorla, 1973). Specifically with regard to panic disorder and agoraphobia, interoceptive cues can be conceptualized as first-order stimuli, whereas agoraphobic contexts can be conceptualized as second-order stimuli, which generate fear by virtue of their association with the first-order cues. Thus, panic-related fears of internal sensations may initially generate agoraphobic avoidance to certain agoraphobic stimuli. However, subsequent alteration of panic-related fears alone might not diminish agoraphobic behavior. In other words, etiological pathways may not dictate the most efficient treatment strategy.
Extant data pertaining to the effect of panic control on agoraphobia are very limited. Van den Hout, Arntz, and Hoekstra (1994) compared four sessions of cognitive therapy to four sessions of nonspecific “associative” therapy prior to eight sessions of in vivo exposure for individuals with moderate to severe agoraphobia. The initial cognitive therapy significantly reduced panic frequency, but self-reported agoraphobia and behavioral avoidance were unaffected. The latter variables decreased only with the subsequent eight sessions of in vivo exposure. This set of findings suggests that panic control is not sufficient for alleviation of agoraphobia. However, four sessions may have been too brief an intervention for a true test of the effects of panic control treatment upon agoraphobia.
Ito, Noshirvani, Basoglu, and Marks (1996) randomized patients with severe agoraphobia and panic disorder to 10 weeks of interoceptive and in vivo exposure or in vivo exposure only. In addition, both groups were trained in slow diaphragmatic breathing and asked to carry out daily self-exposure to agoraphobic situations. The groups improved equally well, suggesting that in vivo exposure is sufficient without addition of panic control treatment (i.e., interoceptive exposure). However, this study did not specifically address whether or not panic control is sufficient for the reduction of agoraphobia and to what degree it mediates this reduction. Of more relevance is the report by Hoffart (1995) that an intensive 3-week panic-focused cognitive therapy was effective for agoraphobia symptoms as well as for panic. However, the panic-focused therapy included daily planned exposure tasks.
As previous research has not yet yielded conclusive findings, the goal of the current study was to compare cognitive-behavioral treatment designed specifically for panic control with and without addition of in vivo exposure treatment for agoraphobia. We evaluated whether the combination of panic control and in vivo exposure to agoraphobia situations was more effective for agoraphobia than panic control alone. In addition, we examined the degree to which agoraphobia reduction was mediated by reduction in panic-related fears.