70639727b6
Pipeline auto-fixer: removed [[ ]] brackets from links that don't resolve to existing claims in the knowledge base.
6.5 KiB
| type | title | author | url | date | domain | secondary_domains | format | status | priority | tags | intake_tier | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| source | VigiBase Pharmacovigilance: GLP-1 Shows Disproportionate Reporting for Depressed Mood (aROR 1.70) and Suicidality (1.45) Despite Protective Cohort Evidence | ScienceDirect / Clinical Nutrition (VigiBase study team) | https://www.sciencedirect.com/science/article/pii/S0261561425001657 | 2025-12-01 | health | research-article | unprocessed | medium |
|
research-task |
Content
Study: "Psychiatric and psychological adverse effects associated with dulaglutide, semaglutide, and liraglutide: A VigiBase study." Published in ScienceDirect (2025, data through December 1, 2024).
Method: Disproportionality analysis of WHO VigiBase (global adverse event reports database). Analyzed spontaneous adverse event reports for semaglutide, dulaglutide, and liraglutide. Calculated adjusted reporting odds ratios (aROR) for psychiatric outcomes.
Key findings:
- Depressed mood disorders: aROR 1.70 (semaglutide) — significant
- Suicidality: aROR 1.45 (semaglutide) — significant
- Anxiety: aROR 1.26 (semaglutide) — significant
- Eating disorders: aROR 4.17-6.80 across all three GLP-1 RAs — HIGHLY significant
- Only semaglutide showed signals in BOTH FAERS and VigiBase for depression (FAERS: ROR 1.26; VigiBase: ROR 1.38)
- Liraglutide and tirzepatide showed NO significant depressive disorder signals
Signal context:
- Pharmacovigilance reporting signals are NOT incidence estimates — they measure disproportionate reporting relative to other drugs, not absolute risk
- Notoriety bias: once media covers GLP-1 and depression/suicidality, reporting rates increase independent of causation
- The Swedish national cohort (n=95,490) uses causal inference methods and shows PROTECTIVE effects for semaglutide on depression/anxiety
- These two bodies of evidence appear contradictory but are measuring different things
Critical finding — eating disorders:
- aROR 4.17-6.80 for ALL THREE GLP-1 RAs studied
- This signal appears across drugs, not just semaglutide — suggests it's a class effect
- GLP-1-induced nausea, appetite suppression, and food aversion may trigger eating disorder pathology in susceptible individuals
- This is the MOST ACTIONABLE safety signal — higher magnitude than depression/suicidality
Concurrent prescribing interaction:
- Concurrent antidepressants: OR 4.45 for suicidal ideation reports
- Concurrent benzodiazepines: OR 4.07 for suicidal ideation reports
- These interaction signals suggest the highest-risk patients are those with pre-existing psychiatric pharmacotherapy
Regulatory context:
- FDA recommends close monitoring for mood changes, emerging/worsening depression, or suicidal behavior during liraglutide and semaglutide treatment
- FDA review (2023-2024) found NO causal evidence linking GLP-1s to suicidality in clinical trial data
- EMA monitoring continues
Methodological limitation for interpretation:
- All patients with "psychiatric instability" are typically excluded from GLP-1 trials — the highest-risk patients are not represented in clinical trial safety data
- Real-world pharmacovigilance thus captures a broader population than trials, but is susceptible to reporting bias
- Indication bias is irreducible in voluntary adverse event reporting
Agent Notes
Why this matters: This is the "other side" of the GLP-1 psychiatric safety debate. While the Lancet Psychiatry Swedish cohort shows 44% reduced depression risk, the VigiBase signals show semaglutide-specific disproportionate reporting for psychiatric adverse events. The eating disorder signal (aROR 4-7) is most actionable — it's consistent across drug class and magnitude suggests a real class effect rather than reporting artifact.
What surprised me: The eating disorder signal magnitude (aROR 4.17-6.80) is dramatically higher than the depression/suicidality signals. GLP-1's appetite-suppressing mechanism may precipitate eating disorders in vulnerable individuals — this is the opposite of the anorexia treatment narrative, but makes mechanistic sense. The drug industry doesn't have strong incentives to study this.
What I expected but didn't find: A clear regulatory response to the eating disorder signal. The FDA focused on suicidality (major media/political attention) but the eating disorder signal appears more consistent and higher-magnitude.
KB connections:
- GLP-1 receptor agonists are the largest therapeutic category launch in pharmaceutical history but their chronic use model makes the net cost impact inflationary through 2035 — safety signal monitoring as market risk
- Big Food companies engineer addictive products by hacking evolutionary reward pathways creating a noncommunicable disease epidemic — GLP-1 addresses food reward pathways; eating disorder risk in vulnerable individuals is the mechanistic flip side
Extraction hints:
- No standalone claim yet — the evidence is too contradictory to write with confidence
- This belongs as a "challenges considered" section in any GLP-1 psychiatric protective claim
- The eating disorder signal (aROR 4-7) may eventually warrant its own claim: "GLP-1 receptor agonists carry a class-effect eating disorder risk in vulnerable individuals that is not captured in metabolic disease clinical trials"
- Extractor should pair this with the Lancet Psychiatry Swedish study archive to capture the full evidence landscape
Context: Novo Nordisk semaglutide MDD program interim data expected late 2026 — this will be the decisive evidence on the depression question. The pharmacovigilance signals will be re-evaluated against MDD RCT results.
Curator Notes (structured handoff for extractor)
PRIMARY CONNECTION: GLP-1 receptor agonists are the largest therapeutic category launch in pharmaceutical history but their chronic use model makes the net cost impact inflationary through 2035 WHY ARCHIVED: Critical counter-evidence to the Swedish cohort protective finding. The eating disorder signal is the most actionable, highest-magnitude psychiatric safety concern from GLP-1s — not adequately covered in current KB. EXTRACTION HINT: Pair with the Lancet Psychiatry Swedish cohort archive. The combined picture is: GLP-1 appears protective for worsening depression/anxiety in metabolic patients, but carries class-effect eating disorder risk. These are not contradictory claims — they cover different populations and outcomes.