GLP-1 Drug Safety Review Finds Suicidality Reports but No Proven Causal Link — Lower Rates Than Metformin | Research

The takeaway

FAERS analysis found suicide/self-injury reports associated with semaglutide, liraglutide, and tirzepatide, but rates were lower than metformin, and no causality link could be established with current data.

GLP-1 RAs < metformin for psychiatric reports
While semaglutide showed a reporting odds ratio of 2.03 for suicide/self-injury, all GLP-1 drugs had lower psychiatric adverse event associations than metformin — providing important context for the safety signal.

What the researchers found

Among 209,354 adverse drug reports in FAERS (2005-2023), 5,378 involved psychiatric disorders and 383 were classified as serious. After pharmacovigilance unmasking, 271 cases implicated individual GLP-1 RAs:

- Liraglutide: n=90, ROR=1.64

- Semaglutide: n=61, ROR=2.03

- Exenatide: n=67, ROR=0.80

- Dulaglutide: n=45, ROR=0.84

- Tirzepatide: n=5, ROR=1.76

42 deaths were recorded including 13 completed suicides. Suicidal ideation was reported for 6 of 7 GLP-1 RAs (all except lixisenatide). Critically, metformin had a greater association with these adverse events than GLP-1 drugs, while orlistat did not. No causal link could be established.

Why it matters

With GLP-1 drugs being prescribed to tens of millions of people for weight loss and diabetes, even rare psychiatric side effects could affect many patients. This comprehensive FAERS analysis provides the most detailed drug-by-drug breakdown of suicidality reports across GLP-1 agonists published to date. The finding that metformin actually has higher psychiatric reporting rates provides important context and suggests these signals may reflect underlying disease risk rather than drug effects.

How the study worked

This pharmacovigilance study analyzed adverse drug reports from the FDA Adverse Events Reporting System (FAERS) from 2005 to 2023. Descriptive and disproportionality analyses were performed using reported odds ratios. An unmasking analysis was conducted to identify individual drug contributions. GLP-1 RAs were compared against metformin and orlistat as reference drugs. Selected preferred terms related to suicidal ideation, self-injury, and suicide were analyzed.

What this study cannot tell us

FAERS is a spontaneous reporting system with inherent biases including underreporting, reporting bias (more attention to GLP-1 psychiatric effects increases reporting), and the inability to establish causation. Patients taking these drugs often have comorbid conditions (depression, diabetes) that independently increase suicide risk. The comparison with metformin is informative but both populations differ in many ways. Denominator data (total prescriptions) is not available in FAERS, preventing incidence rate calculations.

How to read the evidence

This is a pharmacovigilance disproportionality analysis from FAERS, which can identify signals but cannot establish causation. The use of comparator drugs (metformin, orlistat) and unmasking analysis strengthens the methodology, but inherent FAERS limitations remain.

When this study was published

Published in 2024 with data through 2023, this is very current and addresses one of the most actively debated safety questions in medicine as GLP-1 drug prescriptions surge globally.

The bigger picture

The psychiatric safety of GLP-1 drugs is one of the most closely watched pharmacovigilance topics in medicine. This study adds nuance to the debate: while suicidality signals exist for some GLP-1 drugs, they're lower than for metformin — a drug that's been used safely for decades without major psychiatric safety concerns. The authors hypothesize that rapid weight loss itself may trigger emotional and psychological responses, suggesting the issue may not be the drugs per se but the physiological changes they induce.

Questions still open

Does rapid weight loss from any cause (surgery, medication, voluntary) increase psychiatric risk, independent of the specific drug used?
Should psychiatric screening become standard before prescribing GLP-1 agonists for weight loss?
Why does semaglutide show a higher reporting odds ratio (2.03) than other GLP-1 drugs — is this a true signal or a reporting artifact due to its popularity?

Common questions

Do GLP-1 weight-loss drugs cause suicidal thoughts?

This large FDA adverse event analysis found some reports of suicidal ideation with GLP-1 drugs, but at lower rates than metformin (a common diabetes drug not known for psychiatric effects). No causal link could be established. The authors suggest rapid weight loss itself, rather than the drug, may trigger emotional responses in some people.

Should I worry about mental health side effects from semaglutide?

While semaglutide had the highest reporting odds ratio (2.03) among GLP-1 drugs for psychiatric events, this was still lower than metformin's rate. The reports represent a tiny fraction of all users. However, if you have a history of psychiatric conditions, discuss this with your doctor before starting treatment, and report any mood changes.

Read the original research

Exploring the association between suicidal thoughts, self-injury, and GLP-1 receptor agonists in weight loss treatments: Insights from pharmacovigilance measures and unmasking analysis.

European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology, 82, 82-91

View on PubMed ↗Publisher’s paper ↗

Citation

Guirguis, A; Chiappini, S; Papanti P, G D; Vickers-Smith, R; Harris, D; Corkery, J M; Arillotta, D; Floresta, G; Martinotti, G; Schifano, F. (2024). Exploring the association between suicidal thoughts, self-injury, and GLP-1 receptor agonists in weight loss treatments: Insights from pharmacovigilance measures and unmasking analysis.. European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology, 82, 82-91. https://doi.org/10.1016/j.euroneuro.2024.02.003