IMPORTANCE: The head-to-head comparative effectiveness and safety of individual glucagon-like peptide-1 receptor agonists (GLP-1RAs) are not well understood.
OBJECTIVE: To compare risks of kidney, cardiovascular, and death outcomes among patients with type 2 diabetes initiating GLP-1RAs in the Department of Veterans Affairs (VA) health system.
DESIGN, SETTING, AND PARTICIPANTS: This comparative effectiveness study used an active-comparator, new-user target trial-emulation design with national data linked among the VA, Medicare, and US Renal Data System. Participants were GLP-1RA-naive veterans with type 2 diabetes and without end-stage kidney disease treated with metformin who started liraglutide, semaglutide, or dulaglutide between January 1, 2018, and December 31, 2021. Data were analyzed from September 2024 to June 2025.
EXPOSURE: Liraglutide, semaglutide, or dulaglutide.
MAIN OUTCOMES AND MEASURES: Kidney failure (sustained estimated glomerular filtration rate <15 mL/min/1.73 m2 or initiation of kidney replacement therapy), composite cardiovascular and kidney metabolic (CKM) events (kidney failure or major adverse cardiovascular events [MACE]; myocardial infarction, heart failure, or stroke/transient ischemic attack), MACE, all-cause death, and adverse gastrointestinal events (gastroparesis, intestinal obstruction, gallstones, acute cholecystitis, acute pancreatitis) were evaluated separately through March 31, 2023.
RESULTS: Of 21 790 included veterans (mean [SD] age, 63.5 [10.8] years, 19 823 [91.0%] male), 5425 (24.9%), 10 838 (49.7%), and 5527 (24.9%) initiated liraglutide, semaglutide, and dulaglutide, respectively. In weighted Cox regression models, compared with initiation of semaglutide, liraglutide initiation had similar hazards for kidney failure (hazard ratio [HR], 0.93; 95% CI, 0.60-1.44), the CKM composite outcome (HR, 0.96; 95% CI, 0.84-1.10), and MACE (HR, 0.95; 95% CI, 0.83-1.09). Results were similar with liraglutide vs dulaglutide and dulaglutide vs semaglutide comparisons. Liraglutide had significantly lower hazard of all-cause death compared with semaglutide under intent-to-treat analyses (HR, 0.83; 95% CI, 0.69-0.99), which lost significance in per-protocol models. Compared with dulaglutide, liraglutide was associated with a lower risk of all-cause mortality in both the intent-to-treat (HR, 0.69; 95% CI, 0.58-0.83) and per-protocol (HR, 0.50; 95% CI, 0.31-0.82) analyses, but compared with semaglutide, dulaglutide had higher hazard of mortality only in the per-protocol model (HR, 1.72; 95% CI, 1.20-2.47). The only observed difference for the gastrointestinal adverse events was a decreased risk for gallstones and acute cholecystitis with dulaglutide vs semaglutide (gallstones: HR, 0.72; 95% CI, 0.54-0.95; acute cholecystitis: HR, 0.62; 95% CI, 0.39-0.99).
CONCLUSIONS AND RELEVANCE: In this comparative effectiveness study in veterans with diabetes, liraglutide, semaglutide, and dulaglutide initiators had similar risks for kidney and cardiovascular outcomes. Head-to-head randomized trials are needed to confirm these findings.
