GLP-1 eating disorder pharmacovigilance signal (aROR 4.17-6.80) is a class effect that emerged specifically in the obesity treatment population after June 2021, not in the prior metabolic population

Analysis of 2,061,901 adverse event reports through December 2024 found eating disorder signals with adjusted Reporting Odds Ratios between 4.17 and 6.80 across dulaglutide, semaglutide, and liraglutide—the highest magnitude psychiatric signal in the study. Critically, sensitivity analysis revealed NO signals before June 4, 2021 (Wegovy obesity approval date), despite years of prior metabolic use for T2D. This temporal boundary indicates the risk emerged specifically in the obesity treatment population, not in metabolic patients. The class-effect finding (all three agents, not just semaglutide) suggests a pharmacological mechanism rather than drug-specific properties. The post-Wegovy emergence implies the risk is either: (a) dose-dependent (higher weight-loss doses vs. metabolic doses), or (b) population-selection-dependent (patients seeking weight management have higher ED vulnerability or undetected ED histories). Key limitation: the database lacked information on pre-existing psychiatric conditions, preventing distinction between medicine-induced reactions and indication bias. The aROR magnitude (4.17-6.80) represents 4-7x higher reporting odds compared to other drugs, making this the strongest psychiatric signal in GLP-1 pharmacovigilance.

Supporting Evidence

Source: PMC cross-national pharmacovigilance analysis, FAERS/CVAROD/DAEN 2025

Cross-national disproportionality analysis confirms eating disorder signal across US (FAERS), Canadian (CVAROD), and Australian (DAEN) databases. Dulaglutide ROR = 1.47 (95%CI 1.26-1.71) in FAERS and ROR = 17.66 (95%CI 2.45-127.37) in Australian DAEN. Tirzepatide ROR = 1.58 (95%CI 1.14-2.20) in FAERS. Geographic consistency across three national regulatory databases increases biological plausibility of the signal beyond single-database findings.

Extending Evidence

Source: PMC cross-national analysis vs. VigiBase comparison

Methodological discrepancy between studies reveals sensitivity of signal detection to analytical approach: this study's unadjusted ROR (1.47-1.58) is 2.8-4.6x lower than VigiBase's adjusted ROR (4.17-6.80). The VigiBase aROR controls for co-reported adverse events while this analysis may not, suggesting the adjusted analysis provides more reliable effect size estimates. The cross-study discrepancy is itself informative about pharmacovigilance methodology limitations.

Extending Evidence

Source: PMC DAEN analysis, Australia 2025

Australian DAEN database shows exceptionally high ROR (17.66) for dulaglutide eating disorder reports compared to US/Canadian databases, suggesting either: (1) small denominator effects in lower-volume database, (2) population-specific differences in drug response or reporting patterns, or (3) higher clinical awareness/reporting rates in Australian healthcare system. This geographic heterogeneity in signal strength warrants investigation of population-level moderators.

Extending Evidence

Source: NBC News 2024-08-15

FDA adverse event analysis found 'greater risk of abuse among patients taking semaglutide' compared to other weight-loss drugs (not quantified). Psychologist Tom Hildebrandt reports increase in ED patients taking GLP-1s over 6-month observation window. Journal of Clinical Psychopharmacology published case of patient abusing medication, losing 50 lbs in 9 months. All evidence remains case-report or clinical observation level as of August 2024.

Supporting Evidence

Source: ISPOR study via Timmerman Report, November 2025

ISPOR study of 60,000+ GLP-1 users found 1.275% cumulative eating disorder incidence (mainly anorexia nervosa), with prior mental health conditions doubling risk. This confirms the pharmacovigilance signal is detectable in large-scale observational data and identifies behavioral substrate as primary risk stratifier.

Extending Evidence

Source: STAT News, April 27, 2026; ISPOR real-world analysis

ISPOR real-world analysis of 60,000+ GLP-1 users found 1.28% diagnosed with eating disorder within two years. This is total incidence in GLP-1 user population without control group comparison. STAT News characterizes the evidence base as 'scant' and quotes expert assessment that 'physicians, trialists, regulators, policymakers, and drug developers are unprepared for this coming wave.' The 1.28% rate, if applied to 1-in-8 Americans taking GLP-1s, projects to 420,000+ people developing eating disorders.