Comparison of Behavior Therapy and Cognitive Behavior Therapy in the Treatment of Generalized Anxiety Disorder Gillian Butler Department of Psychiatry University of Oxford Melanie Fennell Department of Psychiatry University of Oxford Philip Robson Department of Psychiatry University of Oxford Michael Gelder Department of Psychiatry University of Oxford ABSTRACT In a controlled clinical trial, 57 Ss meeting DSM—III—R criteria for generalized anxiety disorder, and fulfilling an additional severity criterion, were randomly allocated to cognitive behavior therapy (CBT), behavior therapy (BT), or a waiting-list control group. Individual treatment lasted 4—12 sessions; independent assessments were made before treatment, after treatment, and 6 months later, and additional follow-up data were collected after an interval of approximately 18 months. Results show a clear advantage for CBT over BT. A consistent pattern of change favoring CBT was evident in measures of anxiety, depression, and cognition. Ss were lost from the BT group, but there was no attrition from the CBT group. Treatment integrity was double-checked in England and in Holland, and special efforts were made to reduce error variance. Possible explanations for the superiority of CBT are discussed. This research was supported by the Medical Research Council of Great Britain. We would like to thank Ivana Klimes and Anne Cullington for rating three samples of therapy tapes and Sue Simkin for coding and helping to analyze the data. Correspondence may be addressed to Gillian Butler, Department of Psychiatry, University of Oxford, Oxford, England, OX3 7JX. Received: February 28, 1990 Revised: June 17, 1990 Accepted: June 18, 1990 Compared with the substantial progress that has been made in the psychological treatment of phobic and panic disorders (e.g., Barlow, 1988 ; Clark, 1986 ; Foa and Kozak, 1985 ), progress in the development of effective treatments for generalized anxiety disorder (GAD) has been disappointing. Indications that cognitive techniques may be particularly effective for this population (e.g., Borkovec et al., 1987 ; Durham and Turvey, 1987 ; Woodward and Jones, 1980 ) have not been consistent (e.g., Barlow et al., 1984 ; Borkovec and Mathews, 1988 ; Lindsay, Gamsu, McLaughlin, Hood, & Espie, 1987 ). In general, treatment effects are small, and treatments from different theoretical backgrounds appear to be equally, nonspecifically effective (see Öst, 1990 , for a review). Journal of Consulting and Clinical Psychology © 1991 by the American Psychological Association February 1991 Vol. 59, No. 1, 167-175 For personal use only--not for distribution. Page 1 of 1410/30/2000http://spider.apa.org/ftdocs/ccp/1991/february/ccp591167.htmlOne treatment that has been shown to be promising is anxiety management. This led to relatively substantial and clinically valuable improvement in patients with GAD that was significantly greater than the change observed in a randomly allocated waiting-list control group ( Butler, Cullington, Hibbert, Klimes, & Gelder, 1987 ; Butler, Gelder, Hibbert, Cullington and Klimes, 1987 ). However, the study that showed this effect left a number of questions unanswered. First, anxiety management contains both cognitive and behavioral components, and it was not clear how much each component contributed to the results, or whether behavioral treatment alone would be equally effective. The issue is important because behavior therapy is a simpler, and thus more economical, treatment than cognitive therapy, which in practice makes use of both cognitive and behavioral methods. Secondly, the cognitive techniques in anxiety management were relatively uncomplicated and presented in the context of an overall rationale for anxiety management. This proposed that anxiety is maintained by vicious circles relating to physical symptoms, avoidance, and loss of confidence, and that it can be controlled by breaking into these circles,for example by using applied relaxation and graded exposure and by identifying and examining anxiety provoking thoughts. It seemed possible that a more extensive cognitive treatment, based on an explicit cognitive rationale, would produce better results. Specifically, it might reduce the long-standing tendency to worry excessively that characterizes GAD and is now part of the definition of the condition according to the Diagnostic and Statistical Manual of Mental Disorders (3rd ed., rev., DSM—III—R , American Psychiatric Association, 1987 ). Unlike behavioral treatment, cognitive—behavior therapy (CBT) might have the flexibility and range of application to help patients deal better with common consequences of GAD (such as demoralization, loss of confidence, social anxiety, and depression) and with stressful life circumstances. The present investigation was designed to provide information relevant to these issues. Two treatments were compared. The first was a cognitive—behavioral treatment modeled on cognitive therapy as described by Beck (e.g., Beck, 1976 ; Beck, Emery, & Greenberg, 1985 ; Beck, Rush, Shaw, & Emery, 1979 ). The treatment contained cognitive procedures more elaborate than those used in anxiety management. The second treatment was behavioral; it contained the behavioral techniques of anxiety management but none of its cognitive elements. This comparison permitted direct evaluation of the relative efficacy of cognitive and behavioral treatments and hence allowed separate analysis of the main components of anxiety management. The study also considered four subsidiary questions concerned with the processes involved in change: zIs the amount of cognitive change the same after behavior therapy (BT) as it is after CBT? zDoes cognitive change include changes in long-standing patterns of thinking (values, attitudes, assumptions)? zIs CBT more effective than BT for patients who are depressed as well as anxious? zAre relapse rates lower after CBT than after BT, given CBT's potential for changing long-standing cognitive patterns? Method Overview Patients meeting DSM—III—R diagnostic criteria for generalized anxiety disorder were allocated randomly to one of two therapists and to one of three groups: cognitive behavior therapy (CBT); behavior therapy (BT), or waiting list (WL). Those in the WL group were allocated at the same time to Page 2 of 1410/30/2000http://spider.apa.org/ftdocs/ccp/1991/february/ccp591167.htmlsubsequent treatment with
View Full Document
Unlocking...