This study examined the relationship between therapeutic alliance and primary symptom change (weight gain) during CBT for anorexia nervosa. The aims were threefold: (1) to establish the strength of the therapeutic alliance across the treatment, (2) to determine whether early therapeutic alliance is associated with the completion of CBT for this client group, and (3) to determine the direction of the relationship between therapeutic alliance and weight gain. Adult outpatients (N = 65) with a diagnosis of anorexia nervosa (or atypical anorexia nervosa) completed a measure of alliance at session six and at the end of treatment. Weight was recorded at the start of treatment, session six and at the end of treatment. The strength of the alliance was consistently high in the sample. However, early therapeutic alliance was not associated with either the likelihood of completing treatment or subsequent weight gain. In contrast, both early and later weight gain were associated with the strength of subsequent alliance. These findings indicate that it might be advisable to focus on techniques to drive weight gain rather than rely on the therapeutic alliance to bring about therapeutic change.
It has been suggested that a number of factors hinder the development of an effective therapeutic alliance in the treatment of anorexia nervosa, including the patient's fear of change and development, investment in anorexia nervosa as a primary identity, denial of the problem, and impaired cognitive functioning due to malnutrition (e.g., Bruch, 1973; Couturier & Lock, 2006; Katzman, Christensen, Young, & Zipursky, 2001). Despite those proposed factors, the therapeutic alliance in the psychological treatment of anorexia nervosa is actually relatively strong (e.g., Pereira, Lock, & Oggins, 2006; Waller, Evans, & Stringer, 2012). What is not clear is the causality of the relationship between the therapeutic alliance and symptom change during treatment for anorexia nervosa. The evidence to date is that the two are correlated (Isserlin & Couturier, 2012; Pereira et al., 2006), but the widespread assumption that the therapeutic relationship drives symptom change (e.g., Miller & Mizes, 2000) is unsupported to date in the eating disorders as a whole (e.g., Loeb et al., 2005; Waller et al., 2012; Wilson et al., 1999; Zaitstoff, Doyle, Hoste, & Le Grange, 2008).
In order to understand this relationship, it is important to consider the relevant research in other areas of psychological treatment. Two meta-analyses have found a robust association between therapeutic alliance and symptom change (Horvath & Symonds, 1991; Martin, Garske, & Davis, 2000). However, the size of the alliance–outcome correlation is modest at best (r = .22–.26), explaining approximately 5% of therapeutic gains. Furthermore, the order of change is not clear from most such studies – does the therapeutic relationship drive symptom change, or does symptom change drive the therapeutic alliance (as suggested by Safer & Hugo, 2006)? In the depression literature, there is evidence that changes in symptoms precede improvements in therapeutic alliance ( DeRubeis & Feeley, 1990; Feeley, DeRubeis, & Gelfand, 1999; Strunk, Brotman, & DeRubeis, 2010; Tang & DeRubeis, 1999). There is also evidence of a more complex pattern, where symptom change predicts therapeutic alliance, but that alliance then goes on to predict subsequent symptom change ( Barber, Connolly, Crits-Chrisoph, Gladis, & Siqueland, 2000; Webb et al., 2011). However, it is not known whether this pattern would be found in other disorders, including the eating disorders.
This study examined patients' perceptions of the strength of the therapeutic alliance during CBT for anorexia nervosa. It also considered whether early therapeutic alliance is associated with the completion of CBT for this client group. However, the primary aim of the study was to determine the direction of the relationship between therapeutic alliance and key symptom change during CBT for anorexia nervosa – does therapeutic alliance drive weight change or vice versa?
