Skip to main content

Glucagon-like peptide-1 receptor agonists and risk of suicidality among patients with type 2 diabetes: active comparator, new user cohort study.

BMJ (Clinical research ed.)2025-02-27PubMed
Total: 78.5Innovation: 7Impact: 8Rigor: 8Citation: 9

Summary

Using two active-comparator new-user cohorts from linked UK primary care and hospital/death records, GLP-1 receptor agonist use was not associated with higher risk of suicidality (suicidal ideation, self-harm, or suicide) versus DPP-4 inhibitors or SGLT-2 inhibitors. Median follow-up was 1.3–1.7 years in the first cohort with >270,000 participants.

Key Findings

  • Two active-comparator new-user cohorts compared GLP-1 RAs to DPP-4 inhibitors and SGLT-2 inhibitors using propensity score fine stratification weighted Cox models.
  • In the DPP-4 comparator cohort, 36,082 GLP-1 RA users and 234,028 DPP-4 users were included (median follow-up 1.3 vs 1.7 years).
  • GLP-1 RA use was not associated with increased risk of suicidality versus either comparator class.

Clinical Implications

Clinicians can reassure patients that GLP-1 receptor agonists are not associated with increased suicidality relative to DPP-4 or SGLT-2 inhibitors, while still screening for mental health risks as part of routine care.

Why It Matters

Addresses widespread regulatory and public concerns about suicidality with GLP-1 therapies using rigorous comparative designs, informing benefit–risk assessments and prescribing confidence.

Limitations

  • Observational design cannot fully exclude residual confounding and outcome misclassification.
  • Median follow-up was relatively short for rare outcomes like suicide; under-ascertainment is possible.

Future Directions

Extend analyses to longer follow-up, psychiatric subgroups, weight-loss indications, and international datasets; integrate patient-reported outcomes to improve suicidality ascertainment.

Study Information

Study Type
Cohort
Research Domain
Safety/Prognosis
Evidence Level
II - Well-designed observational cohort with active comparator and robust confounding adjustment
Study Design
OTHER