VA study finds liraglutide comparable to semaglutide on kidney and heart outcomes

A comparative effectiveness study published in JAMA Network Open found that veterans with type 2 diabetes who started liraglutide, semaglutide or dulaglutide had similar risks of kidney failure and major heart events over several years of follow-up ^[1]. The authors concluded that head-to-head randomized trials are still needed to confirm the findings ^[1].

The study used what researchers call a "target trial emulation" — an approach that tries to mimic the design of a randomized trial using real-world records. It drew on national data linked across the Department of Veterans Affairs, Medicare and the US Renal Data System ^[1]. Participants were veterans with type 2 diabetes who had never used a GLP-1 receptor agonist, did not have end-stage kidney disease, were being treated with metformin, and started one of the three drugs between January 1, 2018, and December 31, 2021 ^[1]. Outcomes were tracked through March 31, 2023, and the data were analyzed from September 2024 to June 2025 ^[1].

What the numbers showed

Of 21,790 veterans included, the mean age was 63.5 years and 19,823 (91.0%) were male ^[1]. A total of 5,425 (24.9%) started liraglutide, 10,838 (49.7%) started semaglutide and 5,527 (24.9%) started dulaglutide ^[1].

Compared with starting semaglutide, starting liraglutide carried similar hazards for kidney failure (hazard ratio 0.93; 95% CI, 0.60-1.44), for the combined cardiovascular-kidney-metabolic outcome (HR 0.96; 95% CI, 0.84-1.10) and for major adverse cardiovascular events, or MACE (HR 0.95; 95% CI, 0.83-1.09) ^[1]. Results were similar for the liraglutide-versus-dulaglutide and dulaglutide-versus-semaglutide comparisons ^[1]. Kidney failure was defined as a sustained estimated glomerular filtration rate below 15 mL/min/1.73 m2 or the start of kidney replacement therapy; MACE included heart attack, heart failure, or stroke/transient ischemic attack ^[1].

Death results were more mixed. Liraglutide was linked to a lower hazard of all-cause death than semaglutide in the intent-to-treat analysis (HR 0.83; 95% CI, 0.69-0.99), but that difference lost statistical significance in per-protocol models, which account for whether people stayed on the drug ^[1]. Compared with dulaglutide, liraglutide was associated with lower all-cause mortality in both intent-to-treat (HR 0.69; 95% CI, 0.58-0.83) and per-protocol analyses (HR 0.50; 95% CI, 0.31-0.82) ^[1]. Dulaglutide showed a higher hazard of death than semaglutide only in the per-protocol model (HR 1.72; 95% CI, 1.20-2.47) ^[1].

For gastrointestinal side effects — gastroparesis, intestinal obstruction, gallstones, acute cholecystitis and acute pancreatitis — the only difference the researchers found was a lower risk of gallstones (HR 0.72; 95% CI, 0.54-0.95) and acute cholecystitis (HR 0.62; 95% CI, 0.39-0.99) with dulaglutide compared with semaglutide ^[1].

Why it matters for patients

GLP-1 drugs are often discussed as if newer means better, but the authors note that head-to-head comparative effectiveness and safety data for individual GLP-1 receptor agonists are not well understood ^[1]. This analysis suggests that, at least for kidney failure and major heart events in a mostly male, older veteran population with type 2 diabetes on metformin, the three drugs performed similarly ^[1].

That can matter when supply, formulary coverage or cost pushes a switch from one GLP-1 to another. It does not settle questions about weight loss, blood sugar control or head-to-head benefit in people without diabetes, because those outcomes were not the focus of this study ^[1].

The death findings are the least consistent part of the results and should be read with care. A signal that appears in intent-to-treat analysis but disappears when researchers account for staying on the drug can reflect how the analysis was done rather than a true drug effect, and the study cannot prove cause and effect ^[1].

Several limits are worth noting. The population was 91.0% male with a mean age of 63.5 years, so the results may not describe younger patients or women as well ^[1]. Everyone had type 2 diabetes and was taking metformin ^[1]. Tirzepatide (Mounjaro, Zepbound) was not part of this comparison ^[1].

What happens next

The authors call for randomized head-to-head trials to confirm the findings ^[1]. No such trial dates are given in the published report ^[1]. JAMA Network Open also published an accompanying commentary on the study ^[1].