Treatment interventions for longstanding and severe eating disorders: an updated systematic review and narrative synthesis

To update the evidence on treatment outcomes for adults with longstanding and severe eating disorders (ED) by reviewing quantitative studies published since 2019 and identifying priorities for future research. Systematic searches were conducted in June 2024 and updated in April 2025 and May 2026 (PROSPERO CRD42024579804). Eligible studies included quantitative intervention studies involving adults with longstanding and severe ED, with no restrictions on diagnosis, intervention, or treatment settings. The findings were narratively synthesized by intervention type, study design, methodological quality, and outcome. Seventeen main empirical studies, reported across 20 publications, and eight case studies were included. Studies predominantly included female participants with anorexia nervosa. Interventions included specialist inpatient treatment, high-calorie re-alimentation, integrated enhanced cognitive behavioral therapy (CBT-E) pathways, outpatient care, digital or virtual transition support, pharmacological interventions, neuromodulation, and neurosurgery. Two level II randomized controlled trials were identified: one found no added benefit from low-intensity digital self-management after intensive treatment, while the other reported preliminary benefits of adjunctive infra-low frequency neurofeedback in a small inpatient sample with ED and comorbid PTSD. Level III cohort studies of specialist inpatient treatment and intensive integrated pathways reported improvements in BMI, eating disorder psychopathology, psychological symptoms, functioning, and, where assessed, quality of life. The clearest signal of benefit was observed in integrated pathways combining collaborative psychological treatment, weight restoration to the normal range, and continuity across levels of care. Illness duration was not a consistent predictor of outcome, whereas discharge BMI within the healthy range was associated with good follow-up outcomes. Standalone outpatient care and low-intensity digital transition support showed limited benefits. Pharmacological, neuromodulation, and neurosurgical studies were small, largely uncontrolled, and showed inconsistent effects on core ED outcomes. Adults with longstanding and severe ED can achieve meaningful clinical improvement, particularly within intensive, integrated, multidisciplinary pathways. These findings challenge the assumption of futility based on illness duration. However, this heterogeneous group has high comorbidity, and the evidence is limited by inconsistent definitions, observational designs, and short follow-up. Future research should prioritize standardized reporting, implementation, longer follow-up, lived experience, comorbidity, legal status, and continuity of care. This review examined research published since 2019 on treatments for adults with longstanding and severe eating disorders, most commonly anorexia nervosa. The aim was to understand which treatments may help, what predicts better outcomes, and where important gaps in evidence remain. The main synthesis included 17 empirical studies reported across 20 publications, with eight additional case reports or case studies. Most reported real-world clinical data from services, and only two were randomized trials. Most participants were women, and many studies provided limited information on ethnicity, socioeconomic background, comorbidity, or legal treatment status. Overall, the findings show that people with longstanding and severe eating disorders can improve. The most consistent benefits were observed in intensive, integrated care pathways that combined psychological treatment, nutritional rehabilitation, weight restoration to the normal range, and continued support after discharge. These approaches were associated with improvements in weight, eating disorder symptoms, psychological symptoms, functioning, and quality of life. In contrast, standalone outpatient treatment and low-intensity digital support showed limited benefits. The length of illness alone did not reliably predict outcome. Achieving and maintaining a healthy weight, together with continued psychological support, appeared more important for sustaining recovery and reducing relapse. This review challenges the assumption that longstanding illnesses render treatment futile. It highlights the need for better-quality research, longer follow-up, more inclusive samples, and stronger reporting of outcomes that matter to patients and families, including quality of life, functioning, relapse, readmission, and treatment acceptability.