AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Post-Myocardial Infarction GLP-1 Receptor Agonist Initiation and Incident Dementia in Type 2 Diabetes: A Propensity-Matched Cohort Study
By Bajwa W, Sajawal M, Ben Yezza A et al.
Background: GLP-1 receptor agonists (GLP-1 RA) reduce cardiovascular events in type 2 diabetes (T2DM), and emerging evidence suggests potential neuroprotective effects. Myocardial infarction (MI) may represent a period of heightened vulnerability for cognitive decline; however, whether GLP-1 RA initiation after MI is associated with lower dementia risk, but the mechanism remains unclear. Hypothesis: Among patients with T2DM initiating glucose-lowering therapy after MI, GLP-1 RA initiation would be associated with a lower incidence of dementia than the dipeptidyl peptidase-4 inhibitor (DPP-4i) initiation. Methods: We conducted a retrospective cohort study using the TriNetX US Collaborative Network. Adults aged≥40 years with T2DM and acute MI who initiated a GLP-1 RA or DPP-4i within 180 days after MI were identified. Patients with type 1 diabetes, pre-index dementia, or mild cognitive impairment (MCI) were excluded. Cohorts were balanced using 1:1 propensity score matching for demographics, comorbidities, cardiovascular risk factors, and medication use. The primary outcome was incident dementia over 3 years; secondary outcomes included MCI, vascular dementia, Alzheimer's disease, and all-cause mortality. Longer follow-up windows were assessed in sensitivity analyses. Risk ratios (RRs) and hazard ratios (HRs) with 95% CIs were estimated; p<0.05 was considered significant. Results: After 1:1 propensity score matching, 7,419 patients were included in each cohort with adequate baseline balance (all SMDs <0.10). At 3 years, GLP-1 RA initiation was associated with lower cumulative incidence of dementia compared with DPP-4i initiation (1.2% vs 1.6%; RR, 0.75;95% CI, 0.57-0.98; P=0.04). Time-to-event analysis was directionally concordant but nonsignificant (HR, 0.87; 95% CI, 0.66-1.14). Risk-based dementia findings remained significant at 5 years (RR, 0.62; 95% CI, 0.49-0.79) and at the longest available follow-up (RR, 0.55; 95% CI, 0.44-0.68). Secondary neurocognitive outcomes were directionally lower but not significant at 3 years. All-cause mortality was lower among GLP-1 RA initiators at 3 years (RR,0.82; 95% CI, 0.76-0.88; HR, 0.92; 95% CI, 0.85-0.99). Conclusions: Among patients with T2DM initiating therapy after MI, GLP-1 RA initiation was associated with a lower 3-year cumulative incidence of dementia than DPP-4i initiation, with consistent risk-based findings across longer follow-up windows. Attenuated time-to-event estimates for dementias support cautious interpretation. These hypothesis-generating findings warrant prospective evaluation of GLP-1 RA initiation for long-term neurocognitive outcomes after MI.