Tirzepatide versus GLP-1 receptor agonists and stroke in patients with atrial fibrillation and type 2 diabetes

Figures

• 

  Fig. 1. STROBE-compliant patient flow diagram. The diagram illustrates cohort identification, exclusion criteria, propensity-score matching, and the final analytic cohorts of 14,118 patients in each treatment group. Source data are provided as a Source Data file. Abbreviations: AF, atrial fibrillation; GLP-1 RA, glucagon-like peptide-1 receptor agonist; ICD-10, International Classification of Diseases, 10th Revision; PSM, propensity score matching; SD, standard deviation; T2D, type 2 diabetes.

• 

  Fig. 2. Cumulative Incidence of All Stroke among Patients with Atrial Fibrillation and Type 2 Diabetes Receiving Tirzepatide versus Non-Tirzepatide GLP-1 RAs after Propensity-Score Matching. Kaplan–Meier curves for all stroke in the 1:1 propensity-score-matched cohort of 14,118 patients per group. Curves were estimated using cause-specific hazards with death treated as a censoring event; death-restricted sensitivity analyses addressed competing mortality. Source data are provided as a Source Data file.

• 

  Fig. 3. Cumulative Incidence of Hemorrhagic Stroke among Patients with Atrial Fibrillation and Type 2 Diabetes Receiving Tirzepatide versus Non-Tirzepatide GLP-1 RAs. Kaplan–Meier curves for hemorrhagic stroke in the 1:1 propensity-score-matched cohort of 14,118 patients per group. Curves were estimated using cause-specific hazards with death treated as a censoring event; death-restricted sensitivity analyses addressed competing mortality. Source data are provided as a Source Data file.

• 

  Fig. 4. Cumulative Incidence of Ischemic Stroke among Patients with Atrial Fibrillation and Type 2 Diabetes Receiving Tirzepatide versus Non-Tirzepatide GLP-1 RAs. Kaplan–Meier curves for ischemic stroke in the 1:1 propensity-score-matched cohort of 14,118 patients per group. Curves were estimated using cause-specific hazards with death treated as a censoring event; death-restricted sensitivity analyses addressed competing mortality. Source data are provided as a Source Data file.

• 

  Fig. 5. Adjusted Associations of Tirzepatide versus Non–Tirzepatide GLP-1 RAs across Primary, Secondary, Specificity, and Control Outcomes in Patients with Atrial Fibrillation and Type 2 Diabetes. Cox proportional-hazards models estimated adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) in the 1:1 propensity-score-matched cohort of 14,118 patients per group (28,236 distinct patients). Squares denote aHR point estimates and horizontal lines denote 95% CIs; the vertical reference line denotes an aHR of 1. Two-sided confidence intervals are shown. Outcomes were classified as primary (stroke), secondary (all-cause mortality), specificity outcomes (hemorrhagic and ischemic stroke), positive outcome controls (major adverse cardiovascular events and major adverse kidney events), negative outcome controls (burn injury, traumatic brain injury, suicide attempt or ideation), positive exposure control (metformin), and negative exposure control (other antiseptics or disinfectants). Source data are provided as a Source Data file.