In patients with type 2 diabetes (T2D) and moderate risk of cardiovascular disease (CVD), glucagon-like peptide 1 receptor agonists (GLP-1 RAs) provide comparable kidney protection, but semaglutide outperforms other agents in terms of preventing or delaying chronic kidney disease (CKD) and lowering the risk of death, according to a target trial emulation using real-world claims data.
The primary kidney composite outcome, which included incident CKD stages 3-4 and kidney failure (ie, CKD stage 5 and kidney replacement therapy [KRT]), did not significantly differ among dulaglutide, exenatide, liraglutide, and semaglutide. The risk estimates ranged from 0.90 to 1.05 across pairwise comparisons (p=0.200). [Am J Kidney Dis 2026;88:366-377]
For the secondary composite outcome that included the components of the primary composite outcome plus death from any cause, semaglutide was associated with a 12-percent risk reduction compared with exenatide (hazard ratio [HR], 0.88, 95 percent confidence interval [CI], 0.79–0.99) and an 8-percent risk reduction compared with dulaglutide (HR, 0.92, 95 percent CI, 0.87–0.97; p=0.017 overall).
Additionally, semaglutide lowered the risk of death by 19 percent compared with dulaglutide (HR, 0.81, 95 percent CI, 0.70–0.93; p=0.010).
“These findings suggest that semaglutide may be a preferred agent within the GLP-1 RA class when the priority of therapy is prevention of incident CKD and overall survival,” said principal investigator Dr Joshua Neumiller from the Washington State University, Spokane, Washington, US, and colleagues.
“The robust efficacy of semaglutide as a glucose-lowering and weight-loss agent may explain, at least in part, the observed association of semaglutide with benefit for our secondary composite kidney outcome relative to other agents in the class,” Neumiller and colleagues noted.
The investigators also cited a previous work, wherein semaglutide and liraglutide demonstrated greater CVD risk reductions compared with other GLP-1 RAs. [Diabetes Res Clin Pract 2025;229:112910]
“Thus, our findings for the secondary composite outcome could be in part attributable to reductions in CVD-related mortality. Understanding the implications of within-class agent selection is of particular importance in the context of GLP-1 RA drug shortages and with the recent clinical availability of lower cost generic versions of liraglutide and exenatide,” they said.
Some considerations
In an accompanying editorial, Drs Allison Reaves and Mark Sarnak, both from Tufts Medical Center, Boston, Massachusetts, US, commended Neumiller and colleagues for completing an analysis that compared exenatide, dulaglutide, liraglutide, and semaglutide (oral and injectable forms) using a target trial emulation framework and incorporated several sensitivity analyses. [Am J Kidney Dis 2026;88:338-340]
Reaves and Sarnak, however, emphasized that the findings should be considered within the context of several limitations of the study.
“First, albuminuria was not ascertained; therefore, individuals in CKD stages 1 and 2 may have been included in the analysis, and differential levels of albuminuria by GLP-1 RA treatment may theoretically have influenced the results. Second, many of the effect sizes between the comparisons are similar, with some results being statistically significant and others not,” they wrote.
“Third, there are no data on weight loss or glucose control, which, if available, may have provided mechanistic support to the results. Fourth, claims data are inherently limited because they lack sensitivity in identifying CKD, particularly incident CKD stage 3 (the most common outcome in the study), potentially biasing the results,” they continued.
Reaves and Sarnak also highlighted that the analysis did not include dual glucose-dependent insulinotropic polypeptide/GLP-1 RA tirzepatide, which is currently one of the more frequently prescribed GLP RA drugs. The findings, they said, should be reproduced in randomized trials with more detailed ascertainment of albuminuria and laboratory-based glomerular filtration rate rather than relying on claims-based data.
Meanwhile, “for patients with T2D and without CKD stage 3 or higher, the choice of GLP-1 RA does not appear to overtly impact kidney outcomes, so the choice of agent should likely be based on other factors, including reduction of CVD risk, targeted weight loss, and avoiding financial hardship for patients,” they concluded.
For the present analysis, Neumiller and colleagues used Medicare fee-for-service claims data from OptumLabs Data Warehouse. They included adults ≥21 years of age, at moderate cardiovascular risk, who filled a new prescription for a GLP-1 RA. Those with CKD stages 3-4 and kidney failure at baseline were excluded.
Random treatment assignment was emulated using inverse probability of treatment weighting with propensity scores. A total of 74,131 adults were included, of whom 32,727 were in the dulaglutide group, 4,436 in the exenatide group, 6,670 in the liraglutide group, and 30,298 in the semaglutide group. The average age was 64 years, and more than 50 percent were female.