Glucagon-like peptide-1 receptor agonists and the risk of ventricular arrhythmia in patients with acute myocardial infarction and type 2 diabetes mellitus: a propensity score-matched cohort study

Ventricular arrhythmia (VA) substantially contributes to sudden cardiac death in patients with acute myocardial infarction (AMI) and type 2 diabetes mellitus (T2DM). Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have demonstrated cardiovascular benefits; however, their influence on post-AMI ventricular arrhythmogenesis remains unclear. This retrospective cohort study included 1,873 patients with concurrent AMI and T2DM admitted between January 2018 and December 2022. Using propensity score matching (1:2, nearest-neighbor), 277 GLP-1RA users were matched with 554 controls. The primary endpoint was incident VA. Cox proportional hazards regression, inverse probability of treatment weighting (IPTW), Firth’s penalized likelihood regression, and Fine-Gray competing risk analysis were performed. Over a median follow-up of 2.5 years, GLP-1RA use was associated with a reduced risk of VA (adjusted HR 0.61, 95% CI 0.37–0.98, P = 0.040), MACE (HR 0.66, 95% CI 0.45–0.96, P = 0.030), cardiovascular death (HR 0.52, 95% CI 0.32–0.86, P = 0.010), and heart failure rehospitalization (HR 0.42, 95% CI 0.26–0.68, P < 0.001). All-cause mortality showed a non-significant trend (HR 0.84, 95% CI 0.59–1.18, P = 0.310). Firth’s penalized regression confirmed the primary finding (HR 0.58, 95% CI 0.36–0.95, P = 0.029). The Fine-Gray model yielded a subdistribution HR of 0.63 (95% CI 0.39–1.02, P = 0.058) after accounting for competing mortality risk. The 90-day landmark analysis (HR 0.65, 95% CI 0.39–1.06, P = 0.085) and the time-varying exposure Cox model (HR 0.66, 95% CI 0.40–1.08, P = 0.097) were directionally concordant but did not reach statistical significance. GLP-1RA therapy was associated with a lower risk of VA and improved cardiovascular outcomes in patients with AMI and T2DM. Because several sensitivity analyses, including the Fine-Gray competing-risk, landmark, and time-varying models, were borderline non-significant, these findings should be interpreted with caution and regarded as hypothesis-generating, warranting confirmation in prospective randomized trials.