Toward Personalized Psychotherapy in Anorexia Nervosa: Evidence from Long-Term Moderator Analyses of the ANTOP study

Background: The Anorexia Nervosa Treatment of OutPatients (ANTOP) randomized controlled trial (RCT) involved 242 adult patients with anorexia nervosa. Patients from 10 German university hospitals were randomly allocated to 10 months of treatment with focal psychodynamic therapy (FPT), enhanced cognitive behaviour therapy (CBT-E), or optimized treatment as usual (TAU-O). At the end of treatment and follow-up no significant BMI differences between groups were reported. The present study is an explorative, secondary analysis of moderators of weight gain for FPT and CBT-E. Methods: N=160 patients – randomized to FPT and CBT-E – were included in the analysis. Moderator analysis was conducted by using a mixed model approach for repeated measures (MMRM). The outcome variable was BMI at the last follow-up T5 (mean time of 5·96 years after the beginning of the study). Results: Low depression severity and high mental quality of life (QoL) at baseline moderated the outcome in favour of CBT-E, whereas high depression severity and low mental QoL was associated with a better outcome in FPT. Sensitivity analyses to predict BMI at earlier time points T1-T4 produced the same results. The direction of the moderator effects was the same for CBT-E and FPT across all time points. Graphical visualisation confirms that there is a wide range where both treatments perform similarly, while CBT or FPT perform differently in the extreme ranges of these two variables. Interpretation: These findings suggest that baseline psychological burden may inform the selection of psychotherapy modality in anorexia nervosa, supporting a move toward personalized, mechanism-informed treatment allocation.