Abstract
BACKGROUND: New-generation antidiabetic medications may have therapeutic potential for dementia, beyond their glycemic effects. However, information from observational studies exploring the association between new-generation antidiabetic use and dementia risk is limited. OBJECTIVES: To examine the association between new-generation antidiabetic medication use and dementia risk. DESIGN: Retrospective cohort study using electronic health records of a large non-profit health maintenance organization. PARTICIPANTS: 84,798 dementia-free individuals aged ≥65y with type 2 diabetes. MEASUREMENTS: Antidiabetic medication exposure was based on purchased prescriptions and was used as a time-varying variable. Exposure periods were defined as periods in which either dipeptidyl peptidase-4 inhibitors (DPP-4i), sodium-glucose cotransporter-2 inhibitors (SGLT-2i), or glucagon-like peptide-1 analogs (GLP-1a) or their combinations were used, otherwise unexposed. Dementia classification was based on the International Classification of Diseases, Ninth Revision codes or antidementia medication prescriptions. Cox regression models were fitted to quantify the association between antidiabetic medication use and incident dementia. Models were adjusted for 13 potential sources of confounding using inverse-probability weighting. RESULTS: Among 84,798 individuals with a mean diabetes onset age of 66.4 ± 7.5 years, the median follow-up for dementia risk was 8.7 years (Q1-Q3: 5.4-12.8). Dementia was diagnosed in 11,642 (13.7%) individuals. New-generation medication use was associated with reduced dementia risk (HR = 0.69; 95% CI, 0.66-0.73) and by drug classes (DPP-4i, HR 0.67 [95% CI 0.63-0.71]; SGLT-2i, 0.63 [95% CI 0.56-0.70], GLP-1a, 0.61 [95% CI 0.54-0.69]. CONCLUSIONS: The results of this large-scale study suggest that new-generation antidiabetic medication use may be associated with lower dementia risk in older adults with T2D.