Relational Mechanisms in Residential Eating Disorder Treatment: A Moderated Mediation Analysis

Background. Eating Disorders (ED), previously described as a “disease of disconnection,” appear to have differing trajectories of treatment and behavior change based, in part, on feelings of connectedness to others during and outside of treatment. Patient experience of Relational Connectedness is not well-explored in ED treatment compared to other diagnoses but may be a valuable mechanism of change. Interpersonal problems are also known to impact outcome trajectories. Patients high in interpersonal and relational difficulties exhibit negative compensatory behaviors such as avoidance and hostility, which can deleteriously affect building a trusting relationship. Individual difference variables likely play a role in relationship-outcome associations; hence, we sought to understand how the rate of change in Relational Connectedness impacts outcomes, and how interpersonal problems further impact this association. This study explored the impact of Relational Connectedness on symptom outcomes in residential ED treatment, and furthermore, assessed how initial interpersonal functioning of patients impacts their ability to form relationships during treatment. Method. Participants were adolescent and adult females diagnosed with a range of severe EDs who attended residential ED treatment between 2017-2018 (N = 1054). Measures included the Eating Disorder Examination Questionnaire (EDE-Q), Progress Monitoring Tool for Eating Disorders (PMED), and Inventory of Interpersonal Problems (IIP). A confirmatory factor analysis supported differentiation among the constructs, and two sets of parallel multilevel structural equation models evaluated the impact of rate of change in Relational Connectedness (mediation) as well as the added influence of baseline IIP (moderated mediation) on outcomes at discharge and follow-up. Results. All models demonstrated a statistically significant effect of baseline eating symptoms on subsequent symptom outcomes (p < 0.001). Rate of change in Relational Connectedness evidenced a partial indirect effect on 6-month follow-up ED severity (indirect effect, ab = -0.253, p = 0.004); however, the relationship was no longer statistically significant when the impact of interpersonal problems was included as a moderator. Additionally, interpersonal problems at baseline were statistically significantly associated with discharge symptom outcomes (b2 = 0.034, p = 0.017). Conclusion. Interpersonal factors play a role in ED treatment outcomes at both discharge and 6-month follow-up. Enhancing efforts for identifying individuals experiencing limited or worsening Relational Connectedness and building relational skills during treatment may be a valuable buffer against poorer long-term treatment outcomes.