Wu · Journal of the American Medical Directors Association 2025 · Retrospective active-comparator new-user cohort study · n=165,378

Glucagon-Like Peptide-1 Receptor Agonists and Dementia Risk Reduction in Older Adults With Type 2 Diabetes: A Retrospective Cohort Study.

Cited 3 times in the scientific literature.

Level 3 - non-randomized controlled study

Retrospective active-comparator cohort study with propensity score matching.

PubMed 41075815 · doi:10.1016/j.jamda.2025.105901 · record verified 2026-08-26

What was done

Researchers conducted a retrospective, active-comparator, new-user cohort study using electronic health records from 134 healthcare organizations in the TriNetX Global Collaborative Network (January 2017 to November 2024). Adults aged 65 years and older with type 2 diabetes who initiated either a GLP-1 receptor agonist (GLP-1RA) or a dipeptidyl peptidase-4 inhibitor (DPP-4i) were matched 1:1 using propensity scores. The primary outcome was incident dementia, with secondary outcomes evaluating Alzheimer's disease, vascular dementia, and prescriptions for dementia-related medications using Cox proportional hazards models.

What was found

After matching, 82,689 patients were analyzed in each group (total n = 165,378). Compared with DPP-4i use, GLP-1RA use was associated with a significantly lower risk of incident dementia (HR 0.58, 95% CI 0.55–0.61, P < .0001). GLP-1RA initiation was also associated with decreased risks of dementia-related drug prescriptions (HR 0.76, 95% CI 0.70–0.81), Alzheimer's disease (HR 0.62, 95% CI 0.56–0.70), and vascular dementia (HR 0.62, 95% CI 0.55–0.70).

Why it matters

This study provides large-scale real-world evidence supporting potential neuroprotective effects of GLP-1RAs over DPP-4 inhibitors in older diabetic populations. It reinforces the rationale for ongoing and future randomized controlled trials investigating GLP-1RAs for dementia prevention.

Limits

As a retrospective electronic health record analysis, the study is vulnerable to residual confounding, diagnostic coding inaccuracies, and missing clinical variables. Follow-up duration, medication adherence, and socioeconomic factors were not specified in the abstract, and the observational design cannot establish causality.

Cited by