Effective treatments for anorexia nervosa (AN) are available, yet many patients do not respond to treatment or experience relapse. Cognitive bias modification aims to ameliorate cognitive biases that are assumed to contribute to the development and maintenance of AN. This study examines the efficacy of a novel mobile approach-avoidance bias modification training (ABMT) with food cues for AN. In this RCT, 90 inpatients with AN received six sessions of active or sham ABMT with food cues or no training, alongside treatment-as-usual. Primary outcome was self-reported eating disorder psychopathology; secondary outcomes included food-related approach-avoidance bias, fear of food, and BMI. Outcomes were assessed at baseline, end-of-training and at 6-month follow-up. Active ABMT did not result in greater reductions in eating disorder psychopathology. Similarly, treatment conditions did not differ regarding in change in BMI, approach-avoidance bias or fear of food, neither in the short term nor in the long term. Changes in approach-avoidance bias did not mediate training effects on any outcomes. This first RCT on ABMT for AN found no superiority of active ABMT over sham or no ABMT in reducing eating disorder psychopathology. ABMT also appeared insufficient to alter food-related approach bias. Mobile ABMT with food cues in its current form does not seem to be an efficacious adjunct treatment to inpatient TAU for patients with AN. Further research may explore whether modified training protocols and personalized approaches are more promising in this regard. This study's design was registered at the German Clinical Trials Register [https://trialsearch.who.int/Trial2.aspx?TrialID=DRKS00022078].

Clinically derived theoretical models of dissociative identity disorder (DID) suggest that these individuals experience shifts in cognitive functioning between identity states. While prior research has found no objective inter-identity differences in memory, the present study investigated whether such differences are evident in a related cognitive domain: attentional processing. It was hypothesized that individuals with DID would exhibit differences in their attentional bias (AB) toward concern-related stimuli. They were compared to individuals with PTSD who experienced interpersonal childhood trauma, simulators enacting different identity states, and non-clinical adults. Based on cognitive models in anxiety, individuals with PTSD were expected to show a vigilance-avoidance AB pattern for concern-related stimuli. Non-clinical comparisons were expected to show a weaker AB effect than the clinical groups. Participants (n=95) completed a visual spatial dot-probe task with both short (200 ms) and long (1250 ms) exposure conditions, displaying angry (i.e., concern-related stimuli) and happy facial expressions (i.e., control stimuli). In addition, participants completed post-task measures of state anxiety and state dissociation. The results did not reveal significant AB differences between identity states in individuals with DID, although differences were observed between identity states in the simulator group. Moreover, no significant AB differences were found between individuals with DID, PTSD, and non-clinical groups. State anxiety and state dissociation scores were not significantly related to AB scores. Findings provided no support for shifts in attentional processing between identity states in DID. Furthermore, we found no evidence for a hypervigilant-avoidant AB pattern for concern-related stimuli in PTSD.