Abstract
BACKGROUND: Self-guided interventions may broaden the dissemination of evidence-based prevention and treatment protocols for eating disorders. We conducted a meta-analysis comparing self-guided prevention and treatment approaches for eating disorders to (1) control groups and (2) professionally guided self-help programs. METHODS: Forty-six trials were included. Interventions ranged from web, to app, to CD-ROM to book-based programs. Random effects meta-analyses were conducted on numerous symptom and risk outcomes. RESULTS: Only four trials recruited an unselected sample, with self-guided programs reducing shape/weight concerns over control groups (g = 0.18). Among high risk/symptomatic samples (k = 25), self-guided interventions reduced several behavioural and cognitive symptoms (gs = 0.31-0.50) over control groups, with effects being robust when adjusting for higher risk of bias and small sample trials. Among clinical samples (k = 17), evidence for the effectiveness of self-guided interventions over control groups on symptom measures was only found for binge-eating disorder, as too few studies sampled other diagnostic subtypes. Among 10 trials that compared guided to unguided self-help, we observed a significant effect in symptom reduction in favour of guided self-help (g = -0.26). Dropout did not differ between guided and unguided self-help (OR = 0.97). CONCLUSION: Self-guided interventions may be an effective, low intensity intervention format for high risk individuals or for binge-eating disorder presentations.