Peptide News Digest
Evidence Brief 9 min read

GLP-1 Drugs and Eating Disorders: Who May Benefit, Who Is at Risk, and the New Screening Advice

A 60-week Penn trial found semaglutide eased loss-of-control eating and body-image worries in adults with obesity, while a 45-member expert panel says people with active anorexia or restrictive bulimia should generally avoid these drugs. Semaglutide is not approved to treat binge-eating disorder.

Key points

  • In a 60-week Penn trial of 120 adults with obesity, semaglutide 2.4 mg reduced loss-of-control eating, eating-disorder symptoms, drive for thinness, and body dissatisfaction more than placebo at several timepoints
  • Overall binge frequency did not fall significantly more on semaglutide; people with baseline binge episodes (31) or full or subthreshold binge-eating disorder (18) were more likely to stop bingeing on the drug
  • A 2026 meta-analysis of 25 randomized trials found GLP-1 drugs reduced binge eating but increased dietary restraint, and every trial had high or some risk of bias
  • In an online survey of 436 people with eating disorders, 32.1% had ever used a GLP-1 drug and 10.1% reported misuse; lifetime use was 42% among people with atypical anorexia
  • A 45-member Delphi panel in World Psychiatry recommends screening for current and past eating disorders before starting a GLP-1 and generally avoiding the drugs in active anorexia, atypical anorexia, or restrictive bulimia
  • Lisdexamfetamine (Vyvanse), approved in 2015, is the only FDA-approved drug for binge-eating disorder; no GLP-1 label lists the condition

Penn's 60-week semaglutide trial looked for harm and mostly found improvement

Eating-disorder clinicians have asked for years whether a drug that shuts down appetite helps people who binge or feeds the restriction behind anorexia. A trial from the University of Pennsylvania, published online November 15, 2026, in the International Journal of Eating Disorders and announced by Penn Medicine on November 17, gives a 60-week, placebo-controlled answer for people with obesity.

Researchers at Penn's Center for Weight and Eating Disorders randomly assigned 120 adults seeking obesity treatment to weekly semaglutide 2.4 mg (the Wegovy dose) or a placebo shot, with three people getting the drug for every two on placebo. Both groups also got a lifestyle program with diet counseling and activity guidance. At weeks 20, 40, and 60, clinicians interviewed each participant with the Eating Disorder Examination, a standard structured interview, and participants filled out questionnaires on loss of control over eating, drive for thinness, body dissatisfaction, and mood. Loss-of-control eating means feeling unable to stop or limit eating, whether or not the amount is large.

Compared with placebo, the semaglutide group had larger drops in loss-of-control eating, overall eating-disorder symptoms, concern about weight and shape, drive for thinness, and body dissatisfaction at several of those checkpoints. Penn says weight loss averaged about 15% on semaglutide and about 3% on placebo at 60 weeks.

Two results keep this from being a clean win. Across the whole trial, binge frequency did not fall significantly more on semaglutide than on placebo. The binge benefit appeared in subgroups: among the 31 people who reported binge episodes at the start, and the 18 who had full or subthreshold binge-eating disorder, those on semaglutide were more likely to stop bingeing completely. (Subthreshold means symptoms that fall just short of the formal diagnosis.) Subgroups of 18 and 31 people can throw up chance results. The abstract also reports that new binge eating, compensatory behaviors such as vomiting, and reliable worsening of eating symptoms were rare and did not seem more common on semaglutide. They did happen, though, and the authors call for "periodic monitoring in at-risk individuals."

The trial also kept out some of the people most at risk. Its ClinicalTrials.gov record (NCT05548647) excluded anyone diagnosed with bulimia nervosa in the past five years, along with people with diabetes, severe depression, or bipolar disorder. Novo Nordisk, which makes semaglutide, is listed as a collaborator. Penn says the study was funded in part by a Novo Nordisk investigator-sponsored grant and that the company had no role in design, analysis, or reporting.

Binge eating has years of small trials and one approved drug

The National Institute of Mental Health (NIMH) describes binge-eating disorder as regularly losing control of eating and eating unusually large amounts of food. GLP-1 drugs look most promising here, though the trial record is small.

The same Penn group ran an earlier test aimed squarely at the diagnosis. In a 17-week double-blind pilot published in Obesity Science & Practice, 27 adults with binge-eating disorder took liraglutide 3.0 mg (the Saxenda dose) or placebo. "Double-blind" means neither participants nor staff knew who got the drug. Binge episodes fell in both groups, by 4.0 a week on liraglutide and 2.5 on placebo, and the gap was not statistically significant. Remission rates of 44% and 36% did not differ either. Weight loss did: 5.2% versus 0.9%. The authors named a pharmacy dispensing error as a significant limitation.

Pooling the trials helps a little. A 2026 meta-analysis in eClinicalMedicine from University College London combined 25 randomized trials with 8,069 participants. GLP-1 drugs reduced binge eating, loss-of-control eating, and emotional eating compared with control groups, and the effects were larger in trials that enrolled people with binge-eating disorder. Every included trial was rated at high risk of bias or "some concerns," and results varied a lot from trial to trial.

The same analysis found the drugs increased dietary restraint, meaning deliberate limits on what and how much a person eats. The authors wrote that whether this restraint is pathological "is unclear." That finding sits at the center of the whole debate. More restraint may help someone whose problem is bingeing, and the same shift is what clinicians fear in someone prone to anorexia.

No GLP-1 drug is approved to treat binge-eating disorder, and none of their labels list it. In a search of FDA drug labels, the only products with binge-eating disorder as an approved use are lisdexamfetamine (Vyvanse) and its generics. FDA approved Vyvanse for moderate to severe binge-eating disorder in adults on January 30, 2015, calling it the first approved medicine for the condition. It is a Schedule II stimulant, and its patient guide states that it "is not for weight loss." NIMH lists psychotherapy, medical monitoring, nutrition counseling, and medication as parts of eating-disorder treatment. Constant thoughts about food can overlap with binge eating; our food noise explainer covers that research.

Where the risk sits: restriction, misuse, and weak screening

For anorexia nervosa, atypical anorexia, and bulimia, the worry runs the other way. The Academy of Nutrition and Dietetics wrote in July 2026 that GLP-1 appetite changes and stomach side effects "may reinforce strict food rules," and it flagged concerns about binge eating, purging, and misuse among adolescent males. NIMH notes that no FDA-approved medicine exists for anorexia nervosa symptoms, and that anorexia has an extremely high death rate compared with other mental disorders.

The best numbers on misuse come from a University of Louisville research letter in JAMA Psychiatry. It reported an interim analysis of an online survey of 436 people with eating disorders, collected December 2025 through January 2026. As reported by Healio, 32.1% had ever used a GLP-1 drug and 22.0% were using one at the time. One in ten (10.1%) reported misuse, which the researchers defined to include taking more than prescribed, raising the dose early, staying on longer than prescribed, tampering with the pen, or sharing doses. Another 9.9% had used compounded versions. Use was highest among people with binge-eating disorder (54.6%), but it reached 42% among people with atypical anorexia and 11.2% among those with anorexia nervosa. Atypical anorexia involves dangerous weight loss and many anorexia symptoms in someone whose weight is average or above for their height, according to Cleveland Clinic.

The sample was self-selected, 88.5% white, and mostly female, and everything was self-reported, so these percentages may not hold for all people with eating disorders. Lead author Nicholas Peiper told Healio that some people with eating disorders may misreport on the self-reported intake forms that online pharmacies use, and that more research is needed on how they obtain compounded products. In Australia and New Zealand, the Eating Disorders Alliance said in February 2026 that "a number of stories" showed people being prescribed weight-loss drugs "without any screening for an eating disorder, and in some cases, despite having disclosed an eating disorder." The alliance did not publish case counts.

Health-records studies have not shown a jump in eating-disorder diagnoses. A TriNetX study in Diabetes, Obesity and Metabolism matched 9,222 teenagers with overweight or obesity who started semaglutide, liraglutide, or tirzepatide against 9,222 who got lifestyle treatment and found no significant difference in eating-disorder diagnoses from 30 days to three years after starting. That design has blind spots. Diagnosis codes miss people who never come to clinical attention, and people without obesity who buy the drugs outside normal channels are not in a study of patients with obesity. Our guide to teens on GLP-1s covers adolescent prescribing more broadly, and counterfeit and gray-market GLP-1s covers the supply side.

The current Wegovy and Zepbound prescribing information on DailyMed does not mention eating disorders.

The 45-member panel says screen first, then keep checking

In September 2026, World Psychiatry published consensus recommendations from 45 panelists led by Aaron Keshen of Dalhousie University's psychiatry department in Halifax. The group used a modified Delphi process, in which panelists vote in repeated rounds until they converge. It took three rounds, required at least 70% agreement for each recommendation, and averaged 91.7%. The panel included diabetes, obesity, and eating-disorder specialists plus people with lived experience of an eating disorder or GLP-1 use. Kelly Allison, first author of the Penn trial, and Nicholas Peiper, lead author of the survey, were both on it.

The clinical recommendations, as stated in the paper's abstract:

  • Screen for current and past eating disorders with a brief validated questionnaire before starting a GLP-1. The abstract does not name a specific questionnaire.
  • Monitor eating-disorder symptoms during treatment, more often for people at higher risk.
  • Generally avoid these drugs in people with active anorexia nervosa, atypical anorexia, or bulimia with marked dietary restraint or significant weight suppression (being well below a previous highest weight). Proceed cautiously in people with a past history of these disorders.
  • Offer evidence-based psychotherapy alongside the drug when someone with an eating disorder is prescribed one.

The panel also asked for standard education materials for patients and prescribers, long-term studies comparing eating-disorder rates in users and non-users, and trials in binge-eating disorder. The authors call the evidence "limited and inconsistent." A consensus statement records expert agreement, which is weaker evidence than a trial, and the Penn results arrived two months after it.

The Australian alliance wants screening written into rules. It called for mandatory comprehensive medical assessments in telehealth consultations and said anyone requesting weight-management products should be screened for an eating disorder "no matter their size or body weight."

Questions to ask a prescriber

These follow the panel's recommendations and the gaps in the research. They can open a conversation; they cannot replace an eating-disorder assessment.

  • Will you screen me for eating disorders before prescribing, and with which questionnaire? If the visit is a telehealth intake form, ask whether a clinician reviews it and follows up. Our telehealth landscape guide explains how those services work.
  • Does my history change the plan? Mention a past diagnosis of anorexia, atypical anorexia, or bulimia even if you are in recovery. The panel advises caution in that situation. The Penn trial excluded anyone diagnosed with bulimia in the past five years, so its results say little about that group.
  • If I binge, what else should I try first or alongside? Ask about psychotherapy, which the panel recommends for people with eating disorders who are prescribed a GLP-1, and about lisdexamfetamine, the only FDA-approved drug for binge-eating disorder.
  • How will we track eating as well as weight? Ask how often you will check in about skipped meals, bingeing, vomiting, and body image, and who to contact if those change.
  • What happens if I stop? The Academy of Nutrition and Dietetics notes researchers' concerns about weight regain and a return of disordered eating after people stop GLP-1s for binge-eating disorder.
  • Could this affect my mood? Our GLP-1 and mood guide covers depression, suicidal-thought, and emotional-flattening data. People on antipsychotics have a separate set of trials, covered in GLP-1 drugs for antipsychotic weight gain.

Warning signs for families

Appetite loss is the point of these drugs, which makes restriction easy to hide. These signs combine NIMH's lists for anorexia, bulimia, and binge-eating disorder with the misuse behaviors counted in the JAMA Psychiatry survey:

  • Eating far less than nausea or low appetite would explain, skipping meals on purpose, or citing side effects as the reason for not eating.
  • Taking more than prescribed, raising the dose early, keeping a second supply from an online seller or compounder, hiding pens, or getting doses from a friend.
  • Wanting to keep losing weight after reaching an average weight, intense fear of gaining weight, or seeing themselves as larger than they are.
  • Vomiting, laxative use, fasting, or exercising to an extreme.
  • Eating in secret, or seeming ashamed or distressed after eating.
  • A teenager or young adult without obesity trying to get a GLP-1.

Body size is a poor guide. People with atypical anorexia start at a higher weight, so dangerous weight loss can be harder to notice, according to Cleveland Clinic. NIMH says families can encourage a person to seek help and that family-based treatment can improve outcomes, particularly for teenagers. A primary care clinician can refer someone to a qualified mental health professional. If someone is in crisis or having thoughts of suicide, call or text 988 in the US, or call 911 in an emergency.

Updates and corrections

No corrections yet.

Limitations

  • The Penn trial excluded people with bulimia nervosa in the past five years, diabetes, severe depression, or bipolar disorder, and its binge-eating subgroups were small (31 and 18 people)
  • This article relies on the Penn trial abstract and Penn's press release; detailed subgroup percentages from the full paper are not reported here
  • The JAMA Psychiatry survey is an interim, self-selected, self-reported online sample; figures come from the published letter as reported by Healio
  • The TriNetX adolescent study is observational and depends on diagnosis codes, which miss undiagnosed eating disorders and people using GLP-1s outside medical care
  • The Delphi recommendations reflect expert agreement, which is weaker evidence than controlled trials
  • The Penn trial was partly funded by Novo Nordisk, which makes semaglutide; Penn says the company had no role in the study

Citations

  1. 1.
  2. 2.
  3. 3.
    A pilot randomized controlled trial of liraglutide 3.0 mg for binge eating disorder
    Randomized Controlled Trial Obesity Science & Practice 2023
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.

Peptides in this article

Full peptide profiles with evidence levels, dosing data, and safety notes live on peptidelist.org.

Related insights