Patients who have chronic kidney disease (CKD) together with type 2 diabetes mellitus (T2DM) are at increased risk of cognitive decline, dementia, and Alzheimer disease. Contributors to this risk include vascular dysfunction, insulin resistance and chronic inflammation. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have been reported to possess neuroprotective properties in preclinical and some clinical contexts, but their association with dementia risk among patients with CKD and T2DM has not been well defined. This study compared dementia outcomes among CKD stage 3 or later patients with T2DM who initiated GLP-1RAs versus those who initiated dipeptidyl peptidase-4 inhibitors (DPP4is).
The analysis used the TriNetX global research network, drawing from electronic medical records contributed by 67 healthcare organizations in the US Collaborative Network. The study period for identifying new prescriptions was 1 January 2015 through 31 December 2020. Subjects of interest were patients with CKD stage 3 or later who also had a diagnosis of T2DM and who were newly prescribed either a GLP-1RA or a DPP4i during the study window.
Patients with prior exposure to either GLP-1RAs or DPP4is before the index prescription were excluded. Additional exclusions included a recorded diagnosis of dementia within 12 months before the index date and recent hospitalization prior to the index date. The resulting cohorts comprised patients newly initiated on one of the two drug classes and followed from 90 days up to 5 years after the index prescription.
The primary outcome was incidence of dementia, with specific diagnostic categories analyzed separately: Alzheimer disease, vascular dementia, frontotemporal dementia, Parkinson disease, extrapyramidal and movement disorders, and dementia with Lewy bodies. Follow-up ranged from a minimum of 90 days to a maximum of 5 years after treatment initiation, with events captured through electronic medical record diagnoses recorded in the TriNetX network.
The investigators used Kaplan–Meier survival curves to depict cumulative incidence and compared groups using Cox proportional hazards models to estimate hazard ratios (HRs) with 95% confidence intervals (CIs). These methods provided relative risk estimates for dementia and the specified neurological outcomes comparing GLP-1RA initiators with DPP4i initiators.
GLP-1RA use was associated with a statistically significant reduction in the risk of overall dementia compared with DPP4i use (hazard ratio 0.80, 95% confidence interval 0.71–0.91, P = .001). For Alzheimer disease specifically, GLP-1RA therapy was also associated with a lower risk (hazard ratio 0.76, 95% confidence interval 0.59–0.98, P = .033). These associations indicate that, in this observational US cohort of patients with CKD stage 3 or later and T2DM, initiation of GLP-1RAs correlated with reduced incidence of dementia and Alzheimer disease over the follow-up period.
No significant differences were observed between GLP-1RAs and DPP4is for several other diagnostic categories: vascular dementia, frontotemporal dementia, Parkinson disease, extrapyramidal and movement disorders, or dementia with Lewy bodies. Those null findings indicate that the observed association appeared specific to overall dementia and Alzheimer disease within the limits of the recorded diagnoses and analyses.
The authors conclude that GLP-1RA therapy may lower the risk of dementia and Alzheimer disease in patients with CKD stage 3 or later who have T2DM, suggesting potential neuroprotective effects beyond glucose lowering. They highlight the potential clinical relevance for a vulnerable population in which dementia risk is elevated and note that GLP-1RAs could offer benefits that extend beyond metabolic control.
As a retrospective cohort study based on electronic medical records, the analysis is subject to limitations inherent to observational designs, including potential residual confounding and reliance on diagnostic coding for outcome ascertainment. The article states that additional research is required to confirm these findings and to guide optimal treatment strategies for patients with CKD and T2DM. Specific details on adjustment variables, subgroup analyses, absolute event rates, and medication dosing were not reported in the abstract and would require consultation of the full text for further granularity.