Association between glucagon-like peptide-1 receptor agonists and risk of dementia in older adults with type 2 diabetes: A target trial emulation

Abstract

Aim: Type 2 diabetes (T2D) is associated with increased dementia risk, but comparative data across newer glucose-lowering therapies remain limited. We examined whether the initiation of GLP-1 receptor agonists (GLP-1 RAs) was associated with incident dementia compared with DPP4 inhibitors (DPP4is) and SGLT2 inhibitors (SGLT2is) in older adults with T2D. Materials and methods: We conducted a retrospective cohort study emulating a target trial using electronic health records from the University of Pennsylvania Health System (2019–2024), with external validation in the TriNetX U.S. Collaborative Network. Adults aged ≥50 with T2D, no prior dementia, and no GLP-1 RA, DPP4i, or SGLT2i use in the previous year were eligible. New GLP-1 RA users were propensity-score matched (1:1) to DPP4i users (n = 6677 pairs) and SGLT2i users (n = 8434 pairs). The outcome was incident dementia, defined by ICD-10-CM codes. Cox proportional hazards models estimated hazard ratios (HRs), with multiple sensitivity and subgroup analyses performed. Results: Over a median follow-up of 3.0 years (GLP-1 RA vs. DPP4i) and 2.4 years (GLP-1 RA vs. SGLT2i), GLP-1 RA use was associated with a lower dementia risk versus DPP4is (109 vs. 148 events; HR 0.76, 95% CI 0.59–0.97), but a higher risk versus SGLT2is (127 vs. 64 events; HR 1.53, 95% CI 1.13–2.07). Findings were consistent across sensitivity and subgroup analyses. External validation confirmed reduced dementia risk versus DPP4is but not versus SGLT2is. Conclusions: In this large real-world cohort, GLP-1 RAs were associated with lower dementia risk than DPP4is but higher risk than SGLT2is. Prospective, biomarker-informed studies are needed to clarify mechanisms and inform treatment choices in older adults with T2D.