Central Symptoms Predict Post-Treatment Outcomes and Clinical Impairment in Anorexia Nervosa: A Network Analysis in a Randomized-Controlled Trial

Background: Anorexia nervosa (AN) is a serious mental disorder associated with high levels of psychological comorbidity, physical complications, and mortality. Treatment outcomes for AN with best available psychological therapies remain poor. Examination of AN comorbidities and prognostic variables have been highlighted as research areas with high potential. Methods: We examined data from the MOSAIC trial, a multi-site RCT for AN including 142 individuals. We conducted network analyses for (1) AN symptomatology and (2) AN comorbidity with depression and anxiety at all available time points. We conducted linear regressions to determine whether centrality of symptoms in the baseline network was predictive of patient recovery status and clinical impairment at 12-months follow-up. Results: Feeling fat, fear of weight gain, discomfort seeing one's own body, dissatisfaction with weight, and desire to lose weight were among the most central symptoms of AN. Feelings of worthlessness had the highest bridge centrality connecting depression to AN. Centrality of symptoms in the network at baseline was highly related to prognostic value in predicting post-treatment recovery status (r2 = 0.52, p < 0.001) and clinical impairment (r2 = 0.55, p < 0.001). Conclusions: We identified central symptoms of AN and bridge symptoms between AN, depression, and anxiety. These analyses generate useful hypotheses about the etiology and maintenance of AN and related comorbidities, and also generate hypotheses concerning ideal treatment. The validity of network theory in identifying important symptoms is supported by the finding that symptom centrality was closely related to prognostic utility.