Why Suicidality Monitoring and Eating Disorder Surveillance Are Non-Negotiable in Pediatric GLP-1 Trials

Looking beyond weight and metabolic outcomes 

Adolescent obesity sits within a complex clinical, social, and emotional context. Depression, body size, sleep, body image, eating behaviors, peer relationships, and participation in daily life can influence one another. A systematic review identified a potentially bidirectional relationship between depressive symptoms and body size, with pathways involving body image, victimization, eating disorders, and sleep.1 

In a cross-sectional analysis of US adolescents who completed the 2019 Youth Risk Behavior Survey, obesity was associated with higher adjusted odds of past-year suicide attempt and suicide planning than no obesity. The adjusted odds ratios were 1.65 for suicide attempt and 1.27 for suicide planning, after adjustment for demographic and measured psychosocial factors, including feeling sad or hopeless, alcohol and illegal drug use, bullying, and sexual abuse.9 

Some adolescents living with obesity may also encounter weight stigma, teasing, bullying, social isolation, body-image dissatisfaction, or reduced quality of life.10 These experiences vary widely, but they are part of the context in which treatment is studied and received. The American Academy of Pediatrics describes Pediatric obesity as a chronic disease shaped by biological, social, and structural factors, and recommends comprehensive evaluation that includes mental and behavioral health screening.2

What the evidence shows about suicidality and GLP-1s 

Suicidality has received understandable regulatory attention in the GLP-1 receptor agonist class. Reports of suicidal thoughts or actions prompted formal review as use of these therapies expanded in diabetes and obesity care. 

In January 2024, FDA reported that its preliminary evaluation had not found evidence that GLP-1 receptor agonists cause suicidal thoughts or actions. FDA also noted that it could not definitively rule out a small risk at that stage because relatively few events had been observed [3]. The European Medicines Agency’s Pharmacovigilance Risk Assessment Committee subsequently concluded that available evidence did not support a causal association between GLP-1 receptor agonists and suicidal or self-injurious thoughts and actions.4 

FDA’s later completed review similarly did not identify an increased risk of suicidal ideation or behavior. In January 2026, FDA requested removal of suicidality warning language from the labels of Saxenda, Wegovy, and Zepbound.5 These findings are reassuring. They also leave a practical question for trial teams: can the study identify meaningful changes in wellbeing and ensure that concerns are reviewed and addressed? 

This is particularly relevant when interpreting data from selected pivotal-trial populations. In STEP TEENS, adolescents were not eligible if they had severe psychiatric disorders, bulimia nervosa, a history of suicide attempt, or major depressive disorder within two years before screening.6 Such criteria protect participants, while limiting direct evidence in adolescents with more complex mental-health histories. 

Reassuring findings, remaining questions 

Recent adolescent observational data provide an encouraging contribution to this discussion. In a retrospective, propensity-score-matched cohort study published in JAMA Pediatrics, adolescents with obesity prescribed a GLP-1 receptor agonist had a lower recorded incidence of suicidal ideation or attempts than matched adolescents receiving lifestyle intervention without a GLP-1 receptor agonist. 

Over 12 months, suicidal ideation or attempts were recorded in 1.45% of adolescents prescribed a GLP-1 receptor agonist and 2.26% of matched controls. This was associated with a hazard ratio of 0.67, representing a 33% lower observed hazard.7 

This is reassuring observational evidence, rather than evidence of a treatment effect. The study used retrospectively identified electronic health record data, and propensity-score matching cannot remove all residual confounding or establish causality. It also cannot provide a complete view of mood, social functioning, body image, eating-related distress, treatment burden, or health-related quality of life. 

Prospective trials can address some of these gaps through developmentally appropriate assessment of how adolescents experience treatment over time. Each study needs a fit-for-purpose plan that sets out what to measure, when to measure it, and how to interpret concerning results. 

Appetite change is only part of the picture 

Reduced appetite is an intended pharmacological effect of GLP-1 receptor agonists. On its own, however, it says little about eating behavior, psychological wellbeing, or the participant’s experience of treatment. In adolescence, changes in appetite and food intake need to be understood alongside body image, flexibility around eating, emotional distress, daily functioning, and any pre-existing or emerging eating-disorder symptoms. 

Eating disorders occur across body sizes and may be under-recognized in higher-weight young people. The American Academy of Pediatrics advises clinicians to recognize early signs and symptoms of eating disorders and to evaluate young people for disordered eating and unhealthy weight-control behaviors.2,8 

For Pediatric GLP-1 trials, this means looking beyond the magnitude of appetite reduction or weight change. Protocol teams should be able to recognize eating-related distress, increasingly rigid restriction, loss-of-control eating, compensatory behaviors, distress around weight regain, or deterioration in body image and social functioning. Clinical context is essential when interpreting these experiences. 

Making monitoring work in practice 

A questionnaire or electronic assessment is only useful when it leads to clinical review, follow-up, and proportionate escalation where needed. Selecting the right instruments for a given population and endpoint is addressed elsewhere in this series. The focus here is what needs to happen once an instrument is administered. 

For Pediatric GLP-1 trials, this starts with an understanding of baseline context, including relevant mental-health symptoms, eating-related experiences, body-image distress, existing support, and social factors that could affect wellbeing during the study. 

Assessment should extend beyond screening and enrolment. Mental health, quality of life, eating behaviors, and treatment experience can change during a long-duration obesity study, including during dose escalation, after substantial weight change, or in response to treatment burden. Planned longitudinal assessment should be clinically meaningful, age appropriate, and proportionate to participant burden. 

Protocols and site materials should define who reviews a concerning response, the timeframe for review, when the investigator or medical monitor is involved, and how urgent-risk assessment, referral, and follow-up are managed. Workflows also need to account for variation in local mental-health resources, site experience, and multinational delivery. 

For sponsors and CROs, the operational task is to connect assessment completion with clinical interpretation and participant support. This requires alignment across protocol design, eCOA workflows, site training, safety oversight, medical monitoring, and participant communications. Digital collection can improve the visibility of completion and response patterns, while clinical judgement and local duty of care remain central. 

The key question for Pediatric GLP-1 trials is whether they can capture the full experience of treatment for adolescents. Available FDA and EMA reviews have not established a causal association between GLP-1 receptor agonists and suicidal thoughts or actions.4,5 Observational adolescent evidence is also reassuring, with GLP-1 receptor agonist prescribing associated with a lower recorded incidence of suicidal ideation or attempts in one matched cohort study.7 

These findings support careful, prospective assessment of mental health, quality of life, body image, and eating-related experience. Weight change may be meaningful, but it does not capture every relevant aspect of benefit, burden, or risk. 

Suicidality monitoring and eating-disorder surveillance are non-negotiable because trials need to spot concerns early, respond appropriately, and make sense of what adolescents report.  

Next Chapter 

About the authors

Marcela Roy, MA, ACC serves as Executive Director of Clinical Science & Medicine, bringing over 20 years of clinical and research expertise to her role. She provides strategic leadership across CNS and non-CNS therapeutic areas, with particular expertise in psychedelic research and central ratings data quality monitoring methodologies.

Joan Busner, PhD, has more than 35 years of experience as an academic psychiatric researcher and psychopharmacology principal investigator and founded and directed two university psychiatric clinical trials units. Dr. Busner served continuously on University Institutional Review Boards for 20 years. At Signant she has scientific and clinical responsibility for studies in mood, anxiety, pediatrics, and rare/orphan diseases. Dr. Busner is currently Affiliate Associate Professor of Psychiatry at Virginia Commonwealth University.