Association between use of sodium-glucose co-transporter-2 inhibitor and the risk of incident dementia: a population-based cohort study.

Objectives: To assess the association between sodium-glucose co-transporter-2 inhibitor (SGLT-2i) use and the risk of incident dementia compared with dipeptidyl peptidase-4 inhibitors (DPP-4i) use among individuals with type 2 diabetes.

Design: A population-based retrospective cohort study.

Setting: The Clinical Practice Research Datalink (CPRD) Aurum database from the UK.

Participants: Individuals with type 2 diabetes, aged 40 years or older, newly prescribed SGLT-2i or DPP-4i on or after 2013-2021, registered in the CPRD Aurum database.

Main outcome measure: The primary outcome was incident dementia, and the secondary outcome was incident mild cognitive impairment (MCI). Cox proportional hazard models were used to estimate the HR and corresponding 95% CI for the primary and secondary outcomes. Propensity score fine stratification weights were used to adjust for confounding.

Results: Among a cohort of 118 006 individuals, the incident rate (IR) of dementia was 0.56/1000 person-years over a median follow-up period of 1.54 years among SGLT-2i users compared with 2.67/1000 person-years in DPP-4i users, over a median follow-up period of 1.79 years. The adjusted HR for SGLT-2i use compared with DPP-4i use for dementia was 0.78 (95% CI 0.55 to 1.12), while for MCI was 0.86 (95% CI 0.80 to 0.92). The age-specific stratified analysis demonstrated the adjusted HR for SGLT-2i use compared with DPP-4i use for the risk of incident dementia among elderly, aged ≥65 years, was 0.50 (95% CI 0.31 to 0.80).

Conclusion: Primary findings did not yield conclusive evidence to infer an association between SGLT-2i use and the risk of incident dementia.