Target Trial Emulation of Metformin for Dementia Prevention in All of Us: Robustness to Cohort and Confounder Definitions

In the NIH All of Us cohort, metformin initiation is linked to lower death risk (HR 0.61, 95% CI 0.49–0.75) and a reduced dementia incidence (HR 0.90, 95% CI 0.82–1.02).

• Why it matters: Prior observational work suggests metformin may reduce incident dementia, but results can depend on how cohorts and confounders are defined.
• What they did: Using target trial emulation with causal competing risk analysis, the study compared metformin versus sulfonylurea initiation and applied stabilized inverse propensity score weighting, plus Cox proportional hazards and nonparametric models.
• The result: The dementia and death hazard estimates remained aligned across robustness checks, including excluding baseline type 2 diabetes and varying propensity score covariates, supporting metformin as a candidate for dementia prevention.

AI summary of the abstract · may contain errors · not medical advice · read the paper.