Highlights
- Individual GLP-1 receptor agonists and SGLT2 inhibitors demonstrated broadly comparable cardiovascular effectiveness across a cohort of over 1.2 million patients with type 2 diabetes mellitus (T2DM).
- Semaglutide and empagliflozin showed no significant difference in 3-point MACE risk (HR: 1.05; 95% CI: 0.79–1.39) or 4-point MACE risk (HR: 0.95; 95% CI: 0.81–1.12).
- Findings held consistent in a subgroup of 316,242 patients with established cardiovascular disease.
- Clinical decision-making should prioritize factors beyond cardiovascular efficacy, including safety, tolerability, adherence, cost, and patient preferences.
Background
Type 2 diabetes mellitus affects more than 500 million people globally and represents one of the most significant drivers of cardiovascular morbidity and mortality in the modern era. Patients with T2DM face a two- to fourfold increased risk of adverse cardiovascular events compared to those without diabetes, including myocardial infarction, stroke, and heart failure. This elevated risk has driven decades of research into glucose-lowering strategies that can modify cardiovascular outcomes beyond glycemic control.
Two drug classes have emerged as transformational in cardiometabolic medicine: glucagon-like peptide 1 receptor agonists (GLP-1RAs) and sodium-glucose cotransporter 2 inhibitors (SGLT2Is). Both classes have demonstrated robust cardiovascular benefits in large randomized controlled trials, leading to regulatory label expansions and prominent positions in clinical guidelines from the American Diabetes Association, European Association for the Study of Diabetes, and major cardiology societies.
Previous meta-analyses and network analyses have suggested class-level equivalence between GLP-1RAs and SGLT2Is for major adverse cardiovascular events (MACEs). However, these analyses were derived primarily from cardiovascular outcome trials that enrolled highly selected populations, used fixed dosing regimens, and employed rigorous protocols not reflective of routine clinical practice. The question of whether meaningful differences exist between individual agents within and across these drug classes has remained largely unanswered due to a lack of head-to-head comparative effectiveness research.
Study Design
To address this critical evidence gap, Bu and colleagues conducted a multinational, retrospective, new-user active-comparator cohort study using data from 10 U.S. and non-U.S. administrative claims and electronic health record databases. The study enrolled 1,245,211 adults with T2DM who were receiving metformin and initiated second-line therapy with either a GLP-1RA or an SGLT2I between 2013 and 2023.
The GLP-1RA cohort included six agents: albiglutide, dulaglutide, exenatide, liraglutide, lixisenatide, and semaglutide. The SGLT2I cohort included four agents: canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin. Notably, empagliflozin (393,499 patients; 31.6%) and semaglutide (235,585 patients; 18.9%) were the most commonly prescribed agents, followed by dapagliflozin (208,666; 16.8%) and dulaglutide (207,348; 16.8%).
A prespecified subgroup analysis focused on 316,242 patients with established cardiovascular disease (CVD) to assess comparative effectiveness in the highest-risk population. Primary outcomes were 3-point MACE (acute myocardial infarction, stroke, or sudden cardiac death) and 4-point MACE (3-point MACE plus hospitalization or emergency department visit for heart failure). Secondary outcomes included individual MACE components.
The investigators employed propensity score adjustment, negative control calibration, and comprehensive study diagnostics to mitigate confounding—a critical consideration in observational comparative effectiveness research. Random-effects meta-analysis was used to pool hazard ratios across data sources that passed quality diagnostics. On-treatment (per-protocol) and total follow-up hazard ratios were estimated using Cox proportional hazards models.
Key Findings
The central finding of this large-scale real-world study was that individual GLP-1RAs and SGLT2Is demonstrated broadly similar cardiovascular effectiveness, both within and across drug classes. The study had adequate statistical power to detect clinically meaningful differences in hazard ratios between 0.8 and 1.2 for commonly used agents.
Head-to-Head Comparison of Semaglutide versus Empagliflozin
The most clinically relevant comparison—between semaglutide (the most used GLP-1RA) and empagliflozin (the most used SGLT2I)—revealed no significant difference in cardiovascular outcomes:
- 3-point MACE: meta-analytic HR 1.05 (95% CI: 0.79–1.39)
- 4-point MACE: meta-analytic HR 0.95 (95% CI: 0.81–1.12)
The confidence intervals crossed the null, indicating equivalence rather than superiority of either agent for cardiovascular protection. Importantly, these findings were consistent in the subgroup of patients with established CVD, providing reassurance that the findings apply to the highest-risk populations typically targeted in cardiovascular outcome trials.
Within-Class Comparisons
Within-class comparisons similarly demonstrated no significant heterogeneity. GLP-1RAs showed comparable MACE rates across individual agents, as did SGLT2Is. This finding challenges the notion that any specific agent within either class offers meaningfully superior cardiovascular protection relative to its class counterparts—a conclusion that aligns with class-level meta-analyses from randomized trial data.
Study Diagnostics and Validity
The investigators employed rigorous methodological safeguards to ensure the validity of these observational findings. Adequate equipoise between comparison groups was confirmed, indicating that prescribers did not systematically select patients with different baseline risk profiles for different agents. Covariate balance was achieved through propensity score weighting, and negative control calibrations suggested minimal residual confounding from unmeasured factors.
Expert Commentary
The findings from this study represent a pivotal contribution to the cardiometabolic literature, providing real-world evidence that bridges the gap between highly controlled randomized trial environments and everyday clinical practice. The scale of the study—over 1.2 million patients across 10 databases spanning multiple countries—provides unprecedented insight into comparative effectiveness under conditions of routine clinical use.
The equivalence demonstrated between semaglutide and empagliflozin is particularly noteworthy given their distinct mechanisms of action and the different cardiovascular pathways they primarily modulate. GLP-1RAs reduce atherosclerotic events through effects on glucose control, body weight, blood pressure, and inflammation, while SGLT2Is primarily reduce heart failure hospitalizations and renal progression through glycosuric effects, neurohormonal modulation, and improvements in cardiac energetics. Despite these mechanistic differences, the clinical outputs—reduction in composite MACE endpoints—appear remarkably similar.
For clinical guideline developers and prescribers, these findings carry important implications. While both drug classes remain strongly recommended for patients with T2DM and established cardiovascular disease per American Diabetes Association guidelines, the absence of clear cardiovascular superiority for any individual agent suggests that selection decisions should incorporate non-efficacy considerations. These include:
- Safety profiles: GLP-1RAs are associated with gastrointestinal side effects and rare reports of pancreatitis, while SGLT2Is carry risks of genital mycotic infections, euglycemic diabetic ketoacidosis, and bone fractures.
- Tolerability and adherence: Oral versus injectable administration may influence patient preference and persistence.
- Weight effects: GLP-1RAs, particularly semaglutide and tirzepatide, offer substantial weight loss benefits that may be prioritized in patients with obesity.
- Heart failure optimization: SGLT2Is demonstrate particular strength in reducing heart failure hospitalizations, making them potentially preferable in patients with reduced ejection fraction.
- Cost and access: Formulary considerations, insurance coverage, and out-of-pocket expenses influence real-world prescribing.
The study’s limitations include its retrospective observational design, despite extensive confounding mitigation. Unmeasured confounders such as diet, exercise, socioeconomic status, and detailed medication adherence data were not uniformly captured. Additionally, the on-treatment analysis may underestimate effects that persist after discontinuation. The heterogeneity across databases, while addressed through random-effects meta-analysis, introduces some uncertainty in summary estimates.
Conclusion
This landmark comparative effectiveness study provides robust real-world evidence that individual GLP-1 receptor agonists and sodium-glucose cotransporter 2 inhibitors offer comparable cardiovascular benefits in patients with type 2 diabetes mellitus. The equivalence demonstrated between semaglutide and empagliflozin—two of the most commonly prescribed agents in their respective classes—challenges assumptions about meaningful inter-agent differences and aligns with class-level estimates from randomized trial networks.
For clinicians managing patients with T2DM and cardiovascular risk, these findings reinforce that both drug classes provide substantial cardiovascular protection, and that choice between agents should be individualized based on patient preferences, comorbidities, safety considerations, and practical factors. The field may now shift from debates about which class or agent is superior toward optimization of combination therapy, sequencing strategies, and implementation science to improve real-world uptake of these transformative medications.
Future research should examine longer-term outcomes beyond five years, comparative effectiveness in special populations such as frail elderly or those with advanced chronic kidney disease, and the potential synergistic benefits of combined GLP-1RA and SGLT2I therapy—a hypothesis currently under investigation in dedicated randomized trials.
Funding and Disclosures
The study was supported by grants from the National Institutes of Health, the FDA’s Sentinel System, and institutional research funds. Several authors reported conflicts of interest relevant to pharmaceutical industry relationships; full disclosures are available in the original publication. This study was registered on ClinicalTrials.gov under NCT identifier NCT[XXXXXXX].
References
- Bu F, Wu R, Ostropolets A, et al. Comparative Cardiovascular Effectiveness of Glucagon-Like Peptide 1 Receptor Agonists and Sodium-Glucose Cotransporter 2 Inhibitors in Diabetes Mellitus. J Am Coll Cardiol. 2026. PMID: 41984016.
- American Diabetes Association. Standards of Care in Diabetes—2024. Diabetes Care. 2024;47(Suppl 1):S1-S456.
- Zinman B, Wanner C, Lachin JM, et al. Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes. N Engl J Med. 2015;373(22):2117-2128.
- Marso SP, Daniels GH, Brown-Frandsen K, et al. Liraglutide and Cardiovascular Outcomes in Type 2 Diabetes. N Engl J Med. 2016;375(4):311-322.
- Kosiborod M, Cavender MA, Fu AZ, et al. Lower Cardiovascular Risk Associated with SGLT2i Treatment in Real-World Clinical Practice. J Am Coll Cardiol. 2017;69(3):273-282.
