Characterize symptoms reported by Reddit users describing their own retatrutide use and identify potential safety signals for formal pharmacovigilance attention.
Approach:
Design and data: Cross-sectional analysis of 148,640 Reddit posts and comments from 38,936 users across 27 communities, collected from May 2021 through December 2025.
Symptom analysis: A large language model identified self-reported use and symptoms, mapped symptoms to MedDRA Preferred Terms, and excluded prespecified weight-loss and appetite-suppression terms. Manual review of 100 posts assessed classification and extraction performance.
Comparison: Semaglutide and tirzepatide findings from prior social-media research were presented descriptively; samples were not matched and statistical comparisons were not performed.
Key Findings:
The analysis identified 13,589 unique users reporting current retatrutide use; after specified exclusions, 7,823 users (57.6%) had at least one mapped symptom term.
Leading reported terms included increased appetite, fatigue, increased energy, nausea, food craving, insomnia, and increased heart rate. Increased appetite was unexpected given weight loss reported in early retatrutide trials.
Retatrutide discussions had greater representation in appetite, general-system, muscle-related, sleep, and heart-rate or pulse categories than prior semaglutide and tirzepatide discussions. Those discussions had greater representation in nausea, vomiting, and gastrointestinal motility categories; comparisons were descriptive, with unmatched samples and no statistical testing.
Gastrointestinal symptoms were reported but were not predominant in the retatrutide social-media data.
Interpretation:
The authors describe the findings as hypothesis-generating. The observed symptom pattern differs from phase 2 trial reports and prior social-media analyses of approved GLP-1 receptor agonists, but the analysis cannot establish whether symptoms were caused by retatrutide. A proposed link to glucagon receptor activity is speculative.
Limitations:
Social-media reports may be selectively reported and may not represent all retatrutide users.
Product identity, purity, dose, route, treatment duration, concomitant medications, and medical history could not be verified.
The analysis cannot determine incidence, causality, dose-response, or comparative safety.
Natural language processing and MedDRA mapping may miss or misclassify symptoms.
Differences from semaglutide and tirzepatide discussions are descriptive; samples were not matched and no statistical comparisons were performed.
Conclusion:
The authors state that reports from users of potentially unregulated retatrutide may identify safety concerns for pharmacovigilance, while emphasizing that the findings do not establish causal drug effects or generalize to all users.
