GLP-1 Drugs Linked to 37% Lower Suicide Risk in Diabetes — but Cannabis Use Eliminates This Protection | Research

The takeaway

Among 6.4 million T2DM patients, GLP-1 drug users had 37% lower suicide attempt risk, but this protection was completely absent in those with cannabis use disorder, who had 5.5-fold elevated baseline risk.

37% lower suicide attempt risk (aHR 0.63)
Among 6.4 million T2DM patients, GLP-1RA use was associated with reduced suicide risk after controlling for depression, BMI, and other factors — but cannabis use disorder patients showed no protection

What the researchers found

Among 6,424,228 T2DM adults, GLP-1RA use was associated with significantly lower suicide attempt risk (aHR: 0.63; 95% CI: 0.47-0.85). Cannabis use disorder (CUD) dramatically elevated risk (aHR: 5.50; 95% CI: 4.39-6.89). Among CUD patients using GLP-1RAs, the elevated risk persisted (aHR: 5.75; 95% CI: 3.42-9.69) with no protective GLP-1RA effect (aHR: 1.00; 95% CI: 0.37-2.69 for GLP-1RA effect within CUD group).

Why it matters

The European Medicines Agency investigated whether GLP-1 drugs increase suicide risk in 2023 after safety signals were reported. This study — the largest to date examining this question — not only finds no increased risk but actually shows a protective association. This is enormously reassuring for the tens of millions of people taking these drugs. The cannabis use finding adds important clinical nuance: mental health screening and substance use assessment remain critical in this population.

How the study worked

Retrospective cohort study using the TriNetX Research Network (>20 countries, 2003-2023). Adults 30-85 with T2DM were categorized by GLP-1RA exposure and cannabis use disorder. Patients with prior suicidal ideation were excluded. Cox proportional hazards models estimated hazard ratios adjusted for age, sex, depression, BMI, and HbA1c. Kaplan-Meier survival analysis visualized cumulative risk over one year.

What this study cannot tell us

Retrospective observational design cannot establish causation. The TriNetX database may have coding inconsistencies across 20+ countries. GLP-1RA users may differ from non-users in unmeasured ways (healthier patient bias). Cannabis use disorder may be underdiagnosed in claims data. The one-year follow-up may miss longer-term effects. Suicidal ideation exclusion removes patients at already-elevated risk.

How to read the evidence

This is a very large retrospective cohort study (6.4M patients) from a global research network with appropriate confounding adjustment. The massive sample size provides strong statistical power. However, the observational design and potential for unmeasured confounding (healthy user bias) limit causal conclusions.

When this study was published

Published in 2025, this study directly addresses the ongoing regulatory and clinical debate about GLP-1 drugs and psychiatric safety, providing the largest dataset to date on this question.

The bigger picture

This study adds powerful evidence to the debate about GLP-1 drugs and mental health. Rather than causing psychiatric harm, these peptide drugs appear to be associated with reduced suicide risk — potentially through reward pathway modulation, improved metabolic health, weight loss-related quality of life improvements, or direct neuroprotective effects. The cannabis interaction is biologically interesting: cannabis and GLP-1 both affect the brain's reward system, and cannabis use disorder may dysregulate these pathways in ways that GLP-1 drugs cannot overcome.

Questions still open

Is the suicide risk reduction from GLP-1RAs mediated by direct brain effects, improved metabolic health, or weight loss-related psychological benefits?
Why does cannabis use disorder completely negate GLP-1RA's protective association — is there a biological interaction at the cannabinoid-GLP-1 receptor level?
Should clinicians screen for cannabis use disorder when assessing mental health in GLP-1RA patients?

Common questions

Do GLP-1 drugs like Ozempic cause suicidal thoughts?

This study of 6.4 million people with diabetes found the opposite — GLP-1 drug users had 37% fewer suicide attempts than non-users, even after controlling for depression and other factors. While regulatory agencies investigated suicide concerns in 2023, the largest studies consistently show no increased risk. If you experience mood changes on a GLP-1 drug, discuss them with your doctor, but the overall evidence is reassuring.

Why does cannabis use change the GLP-1 drug effect on suicide risk?

Cannabis use disorder was associated with a dramatic 5.5-fold increase in suicide risk in diabetic patients, and GLP-1 drugs provided no protection in this group. Both the cannabinoid and GLP-1 systems affect brain reward circuits, and chronic cannabis use may dysregulate these pathways in ways that prevent GLP-1 drugs from exerting protective effects. This highlights the importance of addressing substance use as part of diabetes care.

Read the original research

Glucagon-like peptide-1 receptor agonists and suicide risk in individuals with diabetes and Cannabis use disorder.

Preventive medicine reports, 58, 103244

View on PubMed ↗Publisher’s paper ↗

Citation

Dhruva, Yesh; Messias, Erick; Lin, Ping-I. (2025). Glucagon-like peptide-1 receptor agonists and suicide risk in individuals with diabetes and Cannabis use disorder.. Preventive medicine reports, 58, 103244. https://doi.org/10.1016/j.pmedr.2025.103244