A Promising Therapy for Eating Disorders Fails to Outperform Existing Treatments in New Evidence Review
A psychological therapy designed to help people understand their own thoughts, feelings and relationships may improve eating-disorder symptoms, but a new systematic review finds that there is not yet enough evidence to show it works better than established alternatives. Mentalization-based treatment, or MBT, has attracted growing interest because eating disorders often involve far more than food and body image. They can be accompanied by intense emotional distress, unstable relationships, self-harm, anxiety, depression and difficulty interpreting what is happening inside oneself or between oneself and other people. The review, published in the Journal of Eating Disorders, examined seven quantitative studies involving 326 reported participant enrolments and concluded that the treatment’s benefits remain preliminary and uncertain.
Mentalization refers to the capacity to understand behaviour—one’s own and other people’s—as influenced by underlying mental states such as beliefs, feelings, intentions and expectations. Someone with stronger reflective functioning may be better able to recognise, for example, that an overwhelming urge to restrict food is linked to fear, shame or a need for control, rather than treating the urge as an unquestionable command. They may also be more able to consider that another person’s comment or reaction could have several possible explanations. In MBT, therapists work to strengthen this flexible, curious way of thinking. The approach does not simply ask patients to challenge distorted beliefs; it focuses on how people make sense of emotional experiences and interpersonal events, especially when they are distressed.
The underlying theory is particularly relevant to eating disorders because mentalizing can become less reliable under emotional pressure. A person may become highly certain about what someone else thinks, lose access to their own feelings, or interpret ambiguous interactions as rejection or criticism. These disruptions can intensify strategies such as food restriction, binge eating, purging or avoidance. MBT aims to slow down that process and help patients examine the links between emotions, thoughts, relationships and behaviour. The treatment may be delivered in different formats and over different periods, and some interventions in the review were fully MBT while others were merely informed by its principles. That variation makes it difficult to determine whether the therapy itself, rather than general therapeutic support, is responsible for reported improvements.
Katie Gabrielle Salucci of East London NHS Foundation Trust, Sela Scott of the University of Calgary and the University of Alberta, and Amelia Austin of the University of Calgary conducted the review. The researchers prospectively registered their protocol with PROSPERO under registration number CRD42024421136 and followed PRISMA guidelines, an established framework for making systematic reviews transparent and reproducible. They searched bibliographic databases, clinical-trial registries, grey-literature sources and Google Scholar from the start of each database through 1 July 2026. They also searched forward and backward through citations, a technique intended to identify relevant studies that may not appear in conventional database searches. Two reviewers independently handled study selection, data extraction and risk-of-bias assessment.
The seven reports included three randomised controlled trials, two quasi-experimental studies and two prospective observational studies. Participants included children, adolescents and adults with diagnosed eating disorders, but the samples were small and diagnostically mixed. The studies covered different conditions and clinical presentations, and they did not all use the same measures or treatment schedules. Two reports may have drawn on overlapping clinical samples, meaning that the total of 326 reported enrolments may represent fewer unique individuals. This matters because a treatment can appear more strongly supported when the same participants are effectively counted in more than one report. The review therefore treated the evidence cautiously and did not assume that every reported enrolment represented an independent contribution.
Several studies recorded improvements during treatment. Participants sometimes showed reductions in eating-disorder symptoms, as well as gains in emotional wellbeing, relationship functioning or quality of life. Broader psychological outcomes also improved in some reports. Such within-group changes can be clinically meaningful: if people feel less distressed or experience fewer symptoms after receiving therapy, the intervention may be helpful in practice. But this type of result cannot establish that MBT caused the improvement. Symptoms may change naturally, participants may improve because of regular clinical contact, or other elements of treatment may account for the outcome. Without a sufficiently rigorous comparison group, it is also impossible to determine whether MBT offers an advantage over another effective therapy or over the care patients would otherwise receive.
The comparative evidence was notably less encouraging. Studies that directly compared MBT or MBT-informed treatment with another intervention did not consistently find that MBT was superior. The review’s authors emphasised that this does not prove MBT is ineffective; instead, it means the available studies cannot reliably establish comparative effectiveness. All of the randomised trials raised at least some concerns about risk of bias, which can arise from weaknesses in randomisation, missing outcome data, outcome measurement or selective reporting. The non-randomised studies were judged to have a critical overall risk of bias. In those designs, differences between people who receive one treatment and those who receive another may reflect pre-existing characteristics rather than the therapy itself.
The decision not to combine the results statistically was itself an important finding about the state of the evidence. A meta-analysis requires studies to be sufficiently similar in their participants, interventions, comparisons and outcome measurements for a pooled estimate to be meaningful. Here, clinical and methodological heterogeneity was too great. One study might measure eating-disorder symptoms with a structured interview, another with a questionnaire, and another focus on emotional regulation or interpersonal functioning. Treatment duration and delivery could also differ. Combining such results into a single numerical effect size might create an appearance of precision that the underlying research does not support. Instead, the investigators used a narrative synthesis, comparing patterns across studies while preserving the uncertainty surrounding each result.
That uncertainty is especially important for people with severe or persistent illness, for whom treatment decisions can carry substantial physical and psychological consequences. The review does not recommend replacing established care with MBT, nor does it show that the approach is unsuitable for patients with eating disorders. It suggests that MBT may be a useful option for some people, particularly those whose symptoms are closely intertwined with difficulties in emotion regulation, reflective functioning or relationships, but the current evidence cannot identify those patients with confidence. The review also cannot establish whether improvements in mentalizing are the mechanism through which symptoms change. Demonstrating that mechanism would require studies that measure reflective functioning before, during and after treatment and test whether changes in it predict later clinical improvement.
Future research will need larger samples, clearer diagnostic descriptions and carefully standardised interventions. Adequately powered randomised trials should compare MBT with credible treatments, treatment as usual and, where appropriate, active psychological controls that match the amount of therapist contact. Researchers will also need to track outcomes beyond the end of therapy to discover whether any benefits are sustained. Consistent measurement of eating-disorder symptoms, emotional regulation, interpersonal functioning, self-harm, quality of life and mentalization could make results easier to compare across studies. Patient experiences may provide another crucial piece of the puzzle: interviews and other qualitative methods could reveal how participants understand the therapy, which components they find useful and why some people disengage or fail to benefit. For now, MBT remains a scientifically plausible and clinically intriguing approach, but the new review makes clear that promise has not yet become proof.
Subject of Research: Mentalization-based treatment and MBT-informed interventions for eating disorders
Subject of Research: Medicine
Article Title: Mentalization-based treatment for eating disorders: a systematic review and narrative synthesis
Article References: Salucci, K. G., Scott, S., & Austin, A. (2026). Mentalization-based treatment for eating disorders: a systematic review and narrative synthesis. Journal of Eating Disorders. https://doi.org/10.1186/s40337-026-01755-y
Image Credits: AI Generated
DOI: 10.1186/s40337-026-01755-y
Keywords: eating disorders, mentalization-based treatment, reflective functioning, emotion regulation, psychotherapy, treatment outcomes, systematic review, narrative synthesis
Cite Scienmag News
APA MLA Chicago
Silas E. (August 28, 2026). Review examines mentalization-based treatment for eating disorders. Scienmag. https://scienmag.com/review-examines-mentalization-based-treatment-for-eating-disorders/
Silas E. “Review examines mentalization-based treatment for eating disorders.” Scienmag, 28 August 2026, https://scienmag.com/review-examines-mentalization-based-treatment-for-eating-disorders/. Accessed 28 August 2026.
Silas E. “Review examines mentalization-based treatment for eating disorders.” Scienmag. August 28, 2026. https://scienmag.com/review-examines-mentalization-based-treatment-for-eating-disorders/
Copy citation Download RIS
Tags: comparison of eating disorder therapiescomparison of MBT with established treatments for eating disorderseating disorder treatment outcomeseffectiveness of MBT in eating disorder recoveryemotional regulation in eating disordersevidence-based approaches for complex eating disorder casesevidence-based eating disorder interventionsMBT effectivenessmentalization and eating disorder symptomsmentalization and emotional regulation in eating disorder patientsmentalization and relationship stabilitymentalization capacity and mental healthmentalization capacity and self-harm in eating disordersmentalization-based treatment for eating disordersmentalization-informed interventions for anxiety and depressionpsychological therapies for emotional distress in eating disorderspsychological treatment for emotional distressrole of mentalization in understanding eating disorder behaviorsself-harm and mentalizationsystematic review of MBTsystematic review of mentalization therapytreatment outcomes of mentalization-based therapy
