Demographic and Clinical Predictors of Multiple Admissions in Inpatient Eating Disorder Treatment

Despite advances in treatment, up to 20% of cases of bulimia nervosa and anorexia nervosa (AN) persist chronically, with about 25% of individuals with AN being re-hospitalized. This study examined demographic and clinical characteristics associated with one versus multiple admissions to an inpatient eating disorder unit.

The sample included 1609 individuals, treated on an inpatient medical stabilization unit, diagnosed with an eating disorder. Of those, 25.4% (n = 326) required multiple admissions. Cox proportional hazard models evaluated adjusted associations of age, gender, eating disorder diagnosis, race, ethnicity, length of stay, presence of non-suicidal self-injury (NSSI), body mass index (BMI), and insurance type at initial admission with the likelihood of multiple admissions.

No significant group differences were found in age, gender, BMI, initial length of stay, or ethnicity. Adjusting for other variables, multiple admissions were more likely for those with NSSI (adjusted HR = 1.46) hazard ratio (HR) and with lower BMI at initial admission (adjusted HR = 0.96) as well as Black patients (adjusted HR = 2.07). Relative to individuals with anorexia nervosa, those with Avoidant Restrictive Food Intake Disorder demonstrated a lower likelihood of multiple admissions (adjusted HR = 0.54).

NSSI, lower BMI at initial admission, and Black race were associated with greater likelihood of multiple admissions, suggesting that emotion dysregulation, greater illness severity, and demographic factors contribute to rehospitalization. Future research may consider integrating interventions such as Dialectical Behavior Therapy into inpatient care to reduce risk of readmission. Early identification of high-risk behaviors (e.g., NSSI) is critical to tailoring treatment and improving long-term outcomes.