Type 2 diabetes (T2D) increases dementia risk, but clear evidence to recommend interventions that can mitigate that risk remains lacking. This population-based retrospective cohort study aimed to determine whether new use of sodium–glucose cotransporter 2 (SGLT2) inhibitors compared with dipeptidyl peptidase 4 (DPP-4) inhibitors was associated with lower dementia risk.


Ontario residents aged ≥66 years who were new users of an SGLT2 inhibitor or a DPP-4 inhibitor from 1 July 2016 to 31 March 2021 entered the cohort. Incident dementia was identified using a validated algorithm for Alzheimer’s disease and related dementias. Propensity score–weighted Cox proportional hazards models were used to obtain adjusted hazard ratios (aHR) and CIs for time to incident dementia. To address reverse causality and disease latency, the observation window started at 1-year lag time from cohort entry. The primary analysis followed intention-to-treat exposure definition, and a secondary as-treated analysis was performed.


Among 106,903 individuals, SGLT2 inhibitors compared with DPP-4 inhibitors were associated with lower risk of dementia (14.2/1,000 person-years; aHR 0.80 [95% CI 0.71–0.89]) over a mean follow-up of 2.80 years from cohort entry. When stratified by different SGLT2 inhibitors, dapagliflozin exhibited the lowest risk (aHR 0.67 [95% CI 0.53–0.84]), followed by empagliflozin (aHR 0.78 [95% CI 0.69–0.89]), whereas canagliflozin showed no association (aHR 0.96 [95% CI 0.80–1.16]). The as-treated analysis observed a larger association (aHR 0.66 [95% CI 0.57–0.76]) than the intention-to-treat analysis.


SGLT2 inhibitors showed an association with lower dementia risk in older people with T2D. Randomized controlled trials are warranted.

This article contains supplementary material online at https://doi.org/10.2337/figshare.21445284.

This article is featured in a podcast available at diabetesjournals.org/journals/pages/diabetes-core-update-podcasts.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.
You do not currently have access to this content.