Bulimia in Japan: A Digital CBT Trial That Shifts the Access Equation
- RCT across 7 Japanese university hospitals (n=61 women, ages 13-65, mean 27.8 years): guided internet-based CBT reduced weekly binge-purge episodes by 9.84 (95% CI 2.49–17.18, p=0.01, Cohen's d=0.73) compared to usual care alone — a medium-to-large effect.
- The ICBT program was delivered over 12 weeks, culturally adapted for Japan, with therapist guidance added to a self-guided digital protocol.
- Mean illness duration was 9.3 years before trial enrollment — consistent with the chronic, treatment-resistant course typical of bulimia nervosa in clinical settings.
- East Asia represents a large and underserved market for eating disorder treatment; this is among the first methodologically rigorous trials of digital BN treatment in the region.
Most RCT evidence for bulimia nervosa treatment comes from North America and Western Europe. The Japanese healthcare context differs substantially: face-to-face CBT for eating disorders is scarce even at university hospitals, waitlists are long, and stigma around psychiatric treatment remains high. This JAMA Network Open trial tests whether guided digital CBT can fill the gap — and gets a meaningful answer.
What the Data Shows
Sixty-one women meeting DSM-5 criteria for bulimia nervosa were recruited across seven university hospitals in Japan over two years (August 2022 to October 2024). Both groups received usual care; the intervention group additionally completed a 12-week guided ICBT program culturally tailored for Japanese patients. The primary outcome was weekly combined frequency of binge eating and compensatory behaviors — the core behavioral metric for BN severity.
The intervention group reduced episodes by a mean of 9.84 per week — an effect size of d=0.73, which falls in the medium-to-large range and is clinically meaningful. The mean illness duration at enrollment was 9.3 years, meaning these were chronically symptomatic patients, not mild presentations. Sensitivity analyses supported the main finding.
The phrase "culturally adapted" is worth unpacking. Japanese CBT adaptations for eating disorders typically adjust framing around perfectionism (which carries different social weight in collectivist contexts), communication of treatment rationale (more directive, less collaborative than Western CBT), and the role of therapist authority. The authors don't detail every adaptation, but the fact that a Western protocol had to be restructured for a Japanese clinical setting is itself a signal for practitioners working with Asian-background patients in any country.
For Your Practice
The 9.3-year mean illness duration in this sample is a clinical landmark. It signals that these patients had been symptomatic through multiple previous treatment attempts. A digital protocol that still produced d=0.73 in this chronically ill group is notable. Digital tools aren't a shortcut — they're a genuine delivery mechanism for evidence-based intervention when in-person care isn't accessible or isn't working.
For clinicians working with patients from East Asian backgrounds, two implications follow. First, the standard CBT-BN frame may need adaptation: consider how shame functions in the patient's cultural context, how secrecy around eating behavior is maintained, and whether the self-monitoring tasks of CBT feel confrontational. Second, patients with long illness duration and multiple treatment failures are not necessarily refractory to CBT — they may simply have never received a culturally calibrated version of it.
Nine years of illness, and a 12-week digital protocol still produced clinically meaningful change — the barrier was access, not treatability.
Small sample (n=61), all-female, Japanese clinical context limits international generalizability. The usual care control means we can't isolate ICBT effects from therapist contact effects. Long-term maintenance data were not reported.