Heart HealthResearch PaperPaywall

Semaglutide linked to lower mortality in adults with congenital heart disease and diabetes

In a 1,422-pair TriNetX cohort of adults with congenital heart disease and T2DM, semaglutide users had fewer deaths and hospitalizations, but a failed negative control urges caution.

Friday, November 20, 2026 1 view
Published in Am J Cardiol
a cardiologist reviewing an echocardiogram of an adult patient in a clinic while holding a semaglutide injector pen

Summary

Adults with congenital heart disease are living longer and increasingly develop type 2 diabetes and obesity, yet they are left out of the big heart-outcome trials of GLP-1 drugs like semaglutide. This retrospective study used a large U.S. health-records network to compare adults with congenital heart disease and diabetes who started semaglutide with similar adults who started a DPP-4 inhibitor. After matching 1,422 pairs and following them for 3 years, semaglutide users had lower death rates (7% vs 12%), fewer hospitalizations (40% vs 55%) and fewer emergency visits (35% vs 43%). Heart failure and stroke were also lower, but heart attack and arrhythmia were not. However, a control outcome that should not have been affected by the drug was also lower in the semaglutide group, suggesting semaglutide users differ in how they engage with healthcare. The authors call the results hypothesis-generating, not proof of benefit.

Detailed Summary

Adults with congenital heart disease (ACHD) now routinely survive into middle and older age, and many develop type 2 diabetes and obesity. They are typically excluded from cardiovascular outcome trials, so whether semaglutide, a GLP-1 receptor agonist that improves cardiometabolic risk in the general population, helps or harms them has been unknown.

This study used the TriNetX Research Network for a retrospective new-user, active-comparator analysis of adults with ACHD and type 2 diabetes who started semaglutide (n = 2,385) or a DPP-4 inhibitor (n = 2,889) between December 2017 and December 2025. Propensity score matching produced 1,422 pairs followed for 3 years, with a sensitivity analysis matched on 27 lesion categories.

Semaglutide was associated with lower all-cause mortality (7% vs 12%; HR 0.66, 95% CI 0.51 to 0.85), hospitalization (40% vs 55%; HR 0.68) and emergency department visits (35% vs 43%; HR 0.85). Major adverse cardiovascular events (HR 0.71), heart failure (HR 0.74) and stroke (HR 0.65) were also lower. Myocardial infarction and arrhythmia were not. Mortality, hospitalization and ED visits held at 1, 3 and 5 years and in every sensitivity analysis; the cardiovascular associations did not.

This is an association, not a demonstrated drug effect, and no mechanism was tested. A prespecified negative control outcome was elevated in every analysis, pointing to residual differences in healthcare engagement between groups. Congenital aortic valve disease was twice as common among semaglutide initiators before matching (23% vs 12%), and the highest-complexity lesions had 10 or fewer patients per arm, so the results mainly describe simple-to-moderate ACHD.

The finding is worth watching, since a plausible cardiometabolic benefit exists in this neglected group, but it needs prospective or better-controlled confirmation. The authors describe the findings as hypothesis-generating.

Key Findings

  • Semaglutide was associated with lower 3-year all-cause mortality than DPP-4 inhibitors (7% vs 12%; HR 0.66, 95% CI 0.51 to 0.85) in matched ACHD adults.
  • Hospitalization (40% vs 55%) and ED visits (35% vs 43%) were lower, and stayed significant at 1, 3 and 5 years.
  • MACE, heart failure and stroke were lower (HR 0.71, 0.74, 0.65), but myocardial infarction and arrhythmia were not, and these were not robust across sensitivity analyses.
  • A prespecified negative control was elevated in every analysis, indicating residual confounding by healthcare engagement.

Methodology

Retrospective new-user, active-comparator cohort study in the TriNetX network, comparing semaglutide with DPP-4 inhibitors in adults with ACHD and T2DM (Dec 2017 to Dec 2025). 1:1 propensity matching gave 1,422 pairs followed 3 years, with sensitivity analyses matched on 27 lesion categories and a prespecified negative control outcome.

Study Limitations

This summary is based on the abstract only. The study is observational and retrospective with electronic-record data; the failed negative control shows residual confounding, and the aortic valve disease imbalance before matching adds concern. Few patients had highly complex lesions, and the cardiovascular outcome associations were not consistently robust.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: