Acute, Longer-Term, and Transdiagnostic Outcomes After Digital Interventions for Eating Disorders

Importance: Digital interventions are increasingly promoted as scalable options for reducing the treatment gap in eating disorders, with the evidence base expanding in recent years to include new populations, delivery formats, and therapeutic approaches. A comprehensive, up-to-date synthesis is needed to clarify the current evidence for digital treatment delivery formats in eating disorders. Objective: To evaluate the association of digital interventions for eating disorders with core and transdiagnostic symptom outcomes in the acute and longer-term phases. Data Sources: MEDLINE, PsycINFO, Web of Science, and Scopus were searched (October 2025) using terms related to eating disorder, digital health, and randomized clinical trials. Study Selection: Randomized clinical trials evaluating a digital intervention for threshold or subthreshold eating disorders were eligible. Interventions had to be delivered via digital technologies (eg, websites, applications, chatbots), with or without support, and compared against a control. Data Extraction and Synthesis: Two reviewers extracted data. Risk of bias was assessed using 4 Cochrane risk of bias criteria. Meta-analyses were conducted using random-effects models, calculating Hedges g for continuous outcomes and odds ratios for symptom abstinence. Main Outcomes and Measures: Primary outcomes included core eating disorder symptoms (global eating disorder psychopathology, binge eating frequency, compensatory behaviors, abstinence, and symptom-specific subscales). Secondary outcomes included comorbid mental health symptoms (depression, anxiety, general distress) and general well-being (quality of life, clinical impairment, self-esteem). Results: A total of 36 trials were included. At posttreatment assessment, digital interventions compared with controls produced significant improvements in primary eating disorder psychopathology (Hedges g = 0.49; 95% CI, 0.38-0.60) and objective binge eating (Hedges g = 0.37; 95% CI, 0.24-0.51) outcomes, as well as other symptom-specific and comorbid mental health outcomes. Effect sizes largely remained significant when adjusting for various sources of biases. Significant benefits were mostly observed across specific clinical populations (eg, bulimia nervosa, binge-eating disorder). Effect sizes were largest for trials that used a waiting list relative to other controls. At follow-up, digital interventions produced weaker but statistically significant sustained improvements for 7 of 9 outcomes. Conclusions and Relevance: In this study, digital interventions were associated with consistent and durable benefits across numerous symptom-specific and transdiagnostic outcomes. These results highlight their potential to expand access to evidence-based support and to inform future clinical implementation efforts.