In nearly 1.5 million patients with type 2 diabetes and cardiovascular disease, GLP-1 receptor agonists and SGLT2 inhibitors provided equivalent cardiovascular protection, both significantly outperforming DPP-4 inhibitors and sulfonylureas.
GLP-1RA ≈ SGLT2i for heart protectionAcross 1.5 million patients with diabetes and heart disease, both drug classes provided equivalent cardiovascular benefit (HR 1.06, not significant), both significantly outperforming older agents
What the researchers found
Over 5.2 million patient-years of follow-up with 25,982 three-point MACE events: GLP-1RAs showed lower 3-point MACE risk than DPP4is (HR: 0.83, 95% CI: 0.70-0.98) and SUs (HR: 0.72, 95% CI: 0.58-0.88). SGLT2is showed similar patterns vs DPP4is (HR: 0.89, 95% CI: 0.79-1.00) and SUs (HR: 0.76, 95% CI: 0.65-0.89). No significant difference between SGLT2is and GLP-1RAs (HR: 1.06, 95% CI: 0.96-1.17). Same hierarchy for 4-point MACE (adding heart failure hospitalization). DPP4is outperformed SUs (HR: 0.87, 95% CI: 0.79-0.95).
Why it matters
No head-to-head randomized trials compare GLP-1RAs and SGLT2is for cardiovascular outcomes — and none are planned. This massive real-world analysis fills that evidence gap for the two most important modern diabetes drug classes. The finding that they're equally effective for cardiovascular protection is hugely practical: clinicians can choose between them based on other factors (kidney protection, weight loss, cost, patient preference) without sacrificing heart benefits.
How the study worked
Federated analysis across 10 international databases (LEGEND-T2DM network, 1992-2021). 1,492,855 patients with T2DM and CVD on metformin who initiated SGLT2is, GLP-1RAs, DPP4is, or sulfonylureas were identified. Large-scale propensity score models with active-comparator target trial emulation conducted pairwise comparisons. Cox proportional hazards models with random-effects meta-analysis combined site-specific estimates.
What this study cannot tell us
Despite the massive sample size, this is observational data using electronic health records, which cannot fully replicate randomized trial conditions. Propensity score matching reduces but cannot eliminate confounding. The federated design means data quality varies across the 10 databases. Specific drugs within each class (e.g., semaglutide vs. liraglutide) were not compared. Follow-up duration and treatment adherence varied. The analysis covers data through 2021 and may not reflect newer agents or doses.
How to read the evidence
Published in JACC, this is an exceptionally large federated analysis across 10 international databases with rigorous methodology including propensity score matching and target trial emulation. While observational, the scale and methodological rigor approach the strength of randomized trials for this comparison.
When this study was published
Published in 2024 in the Journal of the American College of Cardiology, this LEGEND-T2DM analysis represents the definitive real-world comparison of diabetes drug classes for cardiovascular outcomes using data through 2021.
The bigger picture
This study settles one of the biggest debates in diabetes cardiology: whether GLP-1RAs or SGLT2is should be preferred for cardiovascular protection. The answer — they're equivalent — has major implications for clinical guidelines and drug reimbursement decisions worldwide. It supports the approach of using either class based on individual patient characteristics rather than arbitrary preference. The clear hierarchy (SGLT2i ≈ GLP-1RA > DPP4i > SU) also strengthens the case for moving away from sulfonylureas entirely in patients with cardiovascular disease.
Questions still open
- Should sulfonylureas be formally deprioritized in guidelines for patients with type 2 diabetes and cardiovascular disease?
- Do individual GLP-1RAs (semaglutide, liraglutide, dulaglutide) differ in cardiovascular effectiveness when compared within the class?
- Would combining SGLT2i and GLP-1RA provide additive cardiovascular protection beyond either alone?
Common questions
Which is better for heart health — GLP-1 drugs or SGLT2 inhibitors?
Should I switch from sulfonylureas to a newer diabetes drug?
Read the original research
Comparative Effectiveness of Second-Line Antihyperglycemic Agents for Cardiovascular Outcomes: A Multinational, Federated Analysis of LEGEND-T2DM.
Journal of the American College of Cardiology, 84(10), 904-917
Citation
Khera, Rohan; Aminorroaya, Arya; Dhingra, Lovedeep Singh; Thangaraj, Phyllis M; Pedroso Camargos, Aline; Bu, Fan; Ding, Xiyu; Nishimura, Akihiko; Anand, Tara V; Arshad, Faaizah; Blacketer, Clair; Chai, Yi; Chattopadhyay, Shounak; Cook, Michael; Dorr, David A; Duarte-Salles, Talita; DuVall, Scott L; Falconer, Thomas; French, Tina E; Hanchrow, Elizabeth E; Kaur, Guneet; Lau, Wallis C Y; Li, Jing; Li, Kelly; Liu, Yuntian; Lu, Yuan; Man, Kenneth K C; Matheny, Michael E; Mathioudakis, Nestoras; McLeggon, Jody-Ann; McLemore, Michael F; Minty, Evan; Morales, Daniel R; Nagy, Paul; Ostropolets, Anna; Pistillo, Andrea; Phan, Thanh-Phuc; Pratt, Nicole; Reyes, Carlen; Richter, Lauren; Ross, Joseph S; Ruan, Elise; Seager, Sarah L; Simon, Katherine R; Viernes, Benjamin; Yang, Jianxiao; Yin, Can; You, Seng Chan; Zhou, Jin J; Ryan, Patrick B; Schuemie, Martijn J; Krumholz, Harlan M; Hripcsak, George; Suchard, Marc A. (2024). Comparative Effectiveness of Second-Line Antihyperglycemic Agents for Cardiovascular Outcomes: A Multinational, Federated Analysis of LEGEND-T2DM.. Journal of the American College of Cardiology, 84(10), 904-917. https://doi.org/10.1016/j.jacc.2024.05.069