43 Predictors of outcome for cognitive behaviour therapy in binge eating disorder 2 This study has several limitations. Although we were as precise as possible in our expertinformed assessment procedure, we did not use a standardised diagnostic instrument to confirm BED and comorbid diagnoses. In addition we did not use BED diagnosis or binge frequency as outcome measures. Moreover, our CBT intervention was broader than the original CBT protocol as developed by Fairburn et al. (1993). These issues possibly limit the generalisability of our conclusions. Several other limitations are consistent with limitations of other prediction studies. Our follow-up was limited to six months after treatment and we did not take into account all potentially relevant predictors that showed up in other studies, such as binge eating frequency, the overvaluation of shape and weight, emotional eating, self-esteem, interpersonal problems and personality disorders. The fact that the sample consisted mainly of female patients might be considered a limitation as it does not reflect the estimated even distribution of BED among men and women (Grucza et al., 2007). This however is commonly seen in treatment-seeking BED samples (e.g. Hilbert et al., 2007; Ricca et al., 2010). Definite strengths of this study are the large sample size, the naturalistic setting and therefore the possible clinical relevance. In addition, contrary to most prediction studies in eating disorder research, this study used end-of-treatment variables to predict 6-month follow-up. An advantage of this approach is that it not only gives an idea of who can be identified at pretreatment as needing extra attention, but also who needs extra attention during and/or after treatment in order to maintain or further improve treatment outcome in the long term. In conclusion, we found that those patients that start treatment with higher levels of binge eating pathology do improve, but end with relatively higher levels of binge eating pathology. Furthermore, our data suggest that higher levels of drive for thinness, higher levels of interoceptive awareness and, in women, lower levels of body dissatisfaction lead to better treatment outcome in the short and longer term. Future research should look into the potential causal effects of these predictors on treatment outcome. Acknowledgements We thank therapists and support staff at Amarum for their assistance in collecting the data and Liesbeth Verschoor for her assistance in entering the data. Declaration of interest None.
RkJQdWJsaXNoZXIy MjY0ODMw