Consensus Guidance Published on GLP-1 Use in Eating Disorders

A multidisciplinary expert panel has published consensus recommendations on the use of glucagon-like peptide-1 receptor agonists (GLP-1RAs) in patients with or at risk for eating disorders, in a new modified Delphi study.¹ The panel recommends that clinicians assess all patients for current or past eating disorders before starting a GLP-1RA and that these agents generally be avoided in active anorexia nervosa, atypical anorexia nervosa, and bulimia nervosa with marked dietary restraint or significant weight suppression.

Agreement on the avoidance recommendation was 100%, and paper authors cautioned against reading it as an absolute contraindication. They said it is meant to support risk-aware prescribing guided by clinical judgment, including an individualized assessment of symptom profile, illness severity, nutritional status, and treatment context. Clinicians should also proceed cautiously in patients with a past history of these disorders.

The authors noted that obesity does not preclude an eating disorder diagnosis. Patients with obesity may have co-occurring atypical anorexia nervosa, binge eating disorder (BED), or bulimia nervosa and, in such cases, may not be appropriate candidates for GLP-1RA treatment. Experts also said individuals with BED are generally at lower theoretical risk of harm and in some cases may be more likely to benefit, but caution remains warranted because some have clinically significant dietary restraint or weight suppression.

Study Background

Investigators noted that no guidelines currently exist to help clinicians navigate GLP-1RA use in an eating disorder context, and the evidence on risk is limited.¹ The authors cited case reports of 2 patients with a history of anorexia nervosa who relapsed after GLP-1RA treatment for obesity, one with tirzepatide and one with semaglutide misuse.² They also cited a retrospective cohort study in which individuals with obesity and a history of mental health diagnoses were twice as likely to develop an eating disorder within 2 years of starting a GLP-1RA as those without such a history, with the highest incidence for anorexia nervosa.

The study used an adapted 3-round Delphi method. A core research team of 12 held 2 virtual sessions in January and February 2025, and the group expanded through snowball sampling to a 45-member panel: 16 diabetes, obesity, or endocrinology clinicians and researchers; 22 eating disorder clinicians and researchers; and 7 representatives with lived experience of an eating disorder or GLP-1RA use. Panelists gave feedback in meetings in June 2025 and rated draft statements in an anonymous online survey in August 2025 on a 4-point Likert scale. Consensus was defined as 70% or greater agreement.

Clinical Recommendations

Average agreement across recommendations was 91.7%. The panel reached the following clinical consensus:

Screening before initiation reached 94.7% agreement. The panel supported a brief tool of 3 to 5 questions (89.5%) covering full-threshold eating disorders, history, and risk factors such as weight and shape concerns and restrictive behaviors, and suggested the Eating Disorder Screen for Primary Care as an option. Positive screens should prompt referral to an eating disorder specialist, increased monitoring, patient education, or treatment modification based on severity (92.1%). In high-risk cases, experts suggested delaying or avoiding initiation.

Monitoring during and after treatment reached 84.2%, with intensity tailored to baseline screening results (97.4%). Rapid weight loss beyond what is expected for a given agent, typically more than 1 to 2 lb per week, may signal emerging symptoms. If symptoms appear, options include psychoeducation, specialist referral, or discontinuing the GLP-1RA (92.1%). The authors said discontinuation should be weighed against the medical benefits of continued use.

Standardized educational materials for prescribers and patients reached 92.1%, and concurrent evidence-based psychotherapy for patients with diagnosed eating disorders reached 81.6%. Panelists who disagreed with the latter said therapy should be good practice but not a barrier to access, given long wait times and costs.

Research Priorities

All research recommendations met the consensus threshold. Epidemiologic studies comparing eating disorder incidence and outcomes in GLP-1RA users and nonusers reached 100%, as did clinical trials of GLP-1RAs in BED. Trial designs should track binge frequency, “food noise,” and maladaptive dietary restriction, and compare GLP-1RAs with agents such as topiramate and lisdexamfetamine. The lowest agreement, 73.7%, was for a national register of GLP-1RA prescriptions and eating disorder outcomes.

The authors noted that most empirical evidence in eating disorders comes from small, preliminary BED studies. One placebo-controlled pilot trial of liraglutide did not significantly reduce binge frequency despite greater weight loss. A phase 2, placebo-controlled trial comparing tirzepatide with placebo and lisdexamfetamine in BED (NCT06847399) is underway.

Limitations and Next Steps

The authors acknowledged that the panel was not representative of all experts and that no standardized consensus threshold exists; a higher threshold might have strengthened agreement. They said the recommendations may not apply to every clinical scenario, apply only to GLP-1-based medications, and will need updating as evidence emerges. Risk-stratification criteria for screening and monitoring still require development and validation before routine clinical use.

References

1. Keshen A, Koning E, Allison KC, et al. Consensus-based clinical and research recommendations for use of glucagon-like peptide-1 receptor agonists in the context of eating disorders: a modified Delphi study. World Psychiatry. 2026;25(3):465-474.

2. Guerdjikova AI, Dlugosz H, Wolterman R, et al. Relapse in anorexia nervosa associated with tirzepatide: a case report. J Clin Psychopharmacol. 2025;45:287-288.