Sudden Gains and Outcome: A Broader Temporal Analysis of Cognitive Therapy for Depression Andrew M. Busch, Jonathan W. Kanter, Sara J. Landes, a University of Wisconsin-Milwaukee Robert J. Kohlenberg, b University of Washington Individual Beck Depression Inventory (BDI) time courses of 38 clients receiving cognitive therapy (CT) and a modified form of CT were studied in order to investigate temporal changes during CT for depression. The primary aim was to determine if alternative methods of defining and computing gains occurring early in CT would alter the conclusions drawn in the current literature. Three types of gains were studied: sudden gains (previously studied sudden, substan- tial, and stable improvements in depression during 1 between-session interval after Session 2), first-session gains (occurring after first sessions), and pretreatment gains (occurring after pretreatment assessments). Positive out- comes were predicted by first-session gains and by sudden gains occurring in the first half of treatment, highlighting the importance of early change in CT for depression. COGNITIVE THERAPY (CT; Beck, Rush, Shaw, & Emery, 1979) is the most widely used and well researched of the empirically supported psycholog- ical treatments for depression, and its efficacy is well-established (DeRubeis & Crits-Christoph, 1998). However, the exact mechanisms by which CT accomplishes change are still unclear. In efforts to isolate active techniques and mediators of change in CT, several investigators have examined the temporal course of change during treatment. One set of investigators has focused on rapid early response to treatment. For example, Rush, Kovacs, Beck, Weissenburger, and Hollon (1981) found in their sample that over 60% of total symptom reduction occurred by Week 4, and Fennell and Teasdale (1987) found in their sample that more than 66% of overall improvement occurred by Week 3. Fennell and Teasdale also found that clients who experienced at least a 50% reduction in Beck Depression Inventory (BDI; Beck, Ward, Mendelsohn, Mock, & Erbaugh, 1961) scores in the first 2 weeks of treatment had a 100% recovery rate, while those who did not had an 11% recovery rate. Similarly, Blackburn and Bishop (1983) reported that outcomes at the end of treatment could be predicted from scores obtained in Weeks 3 and 4. Ilardi and Craighead (1994), in a review of this literature, concluded that 60% to 80% of symptom reduction occurs in the first 4 weeks of CT. They reasoned that nonspecific factors must be respon- sible for these changes because the application of specific cognitive techniques does not occur until later in treatment. The therapy rationale, therapeu- tic alliance, and assignment of homework were among the nonspecific factors Ilardi and Craighead argued could play a role in this disproportionate early symptom reduction. Tang and DeRubeis (1999a, 1999b) challenged Ilardi and Craighead’s (1994) conclusion. First, Tang and DeRubeis (1999a) pointed out that eight sessions often occur in the first 4 weeks of CT, making the rapid early response in CT seem less remarkable. Second, they cited several studies (DeRubeis & Feeley, 1990; Feeley, DeRubeis, & Gelfand, 1999) indicating that cognitive techniques are used substantially as early as Session 2. Thus, symptom reduction early in treatment should not be Behavior Therapy 37 (2006) 61–68 www.elsevier.com/locate/bt We thank Kris Berlin and Jordan Bonow for helpful reviews of this manuscript. Address correspondence to Jonathan W. Kanter, Ph.D., Univer- sity of Wisconsin-Milwaukee, Department of Psychology, PO Box 413, Milwaukee, WI 53201, USA; e-mail: jkanter@uwm.edu. 0005-7894/06/061–068$1.00/0 n 2006 Association for Behavioral and Cognitive Therapies. Published by Elsevier Ltd. All rights reserved.