Experts call for eating-disorder screening before GLP-1 treatment

As GLP-1 drugs reach more patients, experts are confronting difficult questions about appetite suppression, eating-disorder risk, and how clinicians should monitor people who may be vulnerable.

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. Image Credit: Jack_the_sparow / Shutterstock

In a recent ‘Article in Press’ in the journal World Psychiatry, a group of researchers developed expert-informed clinical and research recommendations for the use of glucagon-like peptide-1 receptor agonists (GLP-1RAs) in people with eating disorders.

Background

What happens when a medication designed to suppress appetite is used by someone vulnerable to disordered eating? GLP-1RAs, first developed for type 2 diabetes and later expanded to obesity treatment, reduce appetite, lower blood glucose and blood pressure, slow gastric emptying, and promote weight loss.

Early studies have reported reductions in binge eating and eating-disorder psychopathology, mainly among people with binge eating disorder. However, findings have been mixed, and the most rigorously controlled trial did not find a greater reduction in binge frequency than placebo despite producing greater weight loss.

Evidence remains limited, with most studies being case reports, pilot trials, or retrospective studies with small sample sizes or limited long-term follow-up. Reports of relapse or emergence of eating disorders have also raised safety concerns. 

About the study

The investigation used an adapted three-round Delphi process to develop consensus-based clinical and research recommendations concerning GLP-1RA use in eating-disorder contexts. Consensus was defined as at least 70% agreement, consistent with commonly recommended Delphi thresholds. The Nova Scotia Health research ethics board reviewed the project and granted an exemption from full ethics review.

A purposively selected core research team of 12 members was expanded to a 45-member panel through snowball sampling. The panel comprised 16 clinicians or researchers specializing in diabetes, obesity, or endocrinology; 22 clinicians or researchers specializing in eating disorders; and seven individuals with lived experience of an eating disorder and/or GLP-1RA use. Lived-experience representatives were recruited through social media.

The core team held two virtual meetings in January and February 2025 to identify themes using recent literature. Seven areas were incorporated into the initial structure: eating-disorder screening, educational materials, symptom monitoring, concurrent nonpharmacological treatment, epidemiological research, regulation and patient safety, and clinical trials for binge eating disorder.

Panel members attended one or more of three virtual meetings in June 2025 to provide verbal feedback, followed by anonymous written feedback. Recommendations were then circulated in an anonymous online survey in August 2025.

Participants rated statements using a four-point Likert scale and provided comments. Agreement percentages and qualitative feedback were used to refine the recommendations before final panel review.

Study results

The multidisciplinary panel of 45 participants achieved at least 70% consensus on the recommendations, with an average consensus of 91.7%. For clinical care, 94.7% agreed that clinicians should assess every individual for a current or previous eating disorder before initiating a GLP-1RA. An 89.5% consensus supported a brief three-to-five-question screening instrument addressing eating disorders, previous disorders, and risk factors such as weight or shape concerns and restrictive behaviors.

Screening was supported in prescribing settings, including primary care, by 94.7% of participants. Positive screens were considered to warrant specialist referral, increased monitoring, education, or treatment modification, depending on symptom severity, with 92.1% agreement.

The panel reached 100% consensus that GLP-1RAs should generally be avoided in people with active eating disorders, particularly anorexia nervosa, atypical anorexia nervosa, and bulimia nervosa accompanied by marked dietary restraint and/or significant weight suppression. The panel also recommended caution among people with a previous history of these disorders. The recommendation was not considered an absolute contraindication for every eating-disorder presentation.

For monitoring, 84.2% supported monitoring physical and psychological signs of eating disorders during treatment and after discontinuation. Monitoring intensity should be individualized according to baseline screening and emerging concerns, such as extreme appetite loss or unexpectedly rapid weight loss; this recommendation received 97.4% consensus.

Monitoring domains included eating-disorder psychopathology, weight-loss rate, nutrition, hunger and satiety, physical health measures, mood, and quality of life, with 86.8% agreement. If symptoms emerged, 92.1% supported psychoeducation, specialist referral, or consideration of GLP-1RA discontinuation, depending on symptom type and severity. Any decision to discontinue treatment should be individualized and collaborative, balancing emerging eating-disorder symptoms against the potential medical benefits of continued GLP-1RA use.

The panel also supported standardized educational materials for prescribers and patients, with 92.1% consensus. Recommended approaches included evidence-based resources, provider quick guides and training, accessible patient materials, social-media communication, dietary psychoeducation, and harm-reduction strategies for remote prescribing. Concurrent evidence-based psychotherapy for individuals with diagnosed eating disorders receiving GLP-1RAs received 81.6% agreement. Examples included cognitive behavioral therapy and dialectical behavior therapy for full-threshold disorders, while guided self-help, group therapy, telehealth, or dietitian counseling were suggested within stepped care for people at risk or with subthreshold symptoms.

The recommendation to compare eating-disorder incidence, risk factors, and outcomes among GLP-1RA users and nonusers received 100% consensus. Longitudinal cohorts, case-control studies, and electronic medical record (EMR) reviews also received 100% consensus. Validation of brief screening tools received 97.4% agreement, while neuroimaging and genetic studies received 89.5%.

Regulatory studies and policies addressing patient safety received 91.9% consensus. Proposed areas included a national outcome register, which received 73.7% consensus, an evaluation of telehealth and compounded prescribing, an analysis of Internet and social media trends, and referral guidelines.

Clinical trials examining GLP-1RA safety, efficacy, and psychological outcomes in binge eating disorder received 100% consensus. Trial designs were recommended to assess binge frequency, “food noise,” restrictive eating, physiological and laboratory measures, psychological mechanisms, neurobiological effects, comparisons with other medications, and combination with evidence-based psychotherapy.

Conclusion

The modified Delphi process generated expert-informed recommendations addressing GLP-1RA use in eating-disorder contexts. The recommendations focused on screening before treatment, individualized monitoring during and after treatment, standardized education, and concurrent evidence-based psychotherapy for people with diagnosed eating disorders receiving these medications.

Research priorities included epidemiological studies, validation of screening tools, regulatory and patient safety investigations, and clinical trials for binge eating disorder. The authors acknowledged limitations, including incomplete representation of all experts, varying panel participation, the absence of a standardized Delphi consensus threshold, and limited applicability across all clinical scenarios. The recommendations are expected to evolve as additional evidence becomes available. Further research is needed to clarify benefits, risks, incidence, and long-term outcomes.