GLP-1 receptor agonists and SGLT-2 inhibitors combined with metformin were associated with reduced risk of cognitive decline, while insulin monotherapy was associated with higher incidence of all three cognitive outcomes in adults with type 2 diabetes.
Key Findings
Results
GLP-1 receptor agonists plus metformin were associated with lower incidence of vascular dementia, mild cognitive disorder, and Alzheimer disease compared to metformin alone.
Study population included 1,528,885 adults with type 2 diabetes; propensity score matching was applied to balance covariates across treatment groups
Participants were aged 40–69 years with a mean age of 58 years and 49.2% women
Results
SGLT-2 inhibitors plus metformin were associated with reduced risk of vascular dementia but not mild cognitive disorder or Alzheimer disease.
Vascular dementia: HR 0.68 (95% CI, 0.54–0.87)
The risk reduction was specific to vascular dementia; associations with mild cognitive disorder and Alzheimer disease were not statistically significant
Outcomes were identified using ICD-10 codes from electronic health records spanning 2010 to 2024 in the TriNetX US Collaborative Network
Results
Insulin monotherapy was associated with higher incidence of all three cognitive outcomes compared to metformin alone.
Authors note this association 'is likely influenced by confounding by indication, unmeasured markers of diabetes severity (including diabetes duration, cardiovascular and renal disease severity, and microvascular and macrovascular complications), and shorter follow-up among insulin users, and should not be interpreted as a direct drug effect'
Mortality and censoring were substantially higher in the insulin group, and because standard Cox models do not account for death as a competing event, reported hazard ratios reflect cause-specific hazards and may not directly correspond to cumulative incidence
Methods
The study used a retrospective cohort design with propensity score matching and Cox proportional hazards models applied to a large real-world electronic health records database.
Data source: TriNetX US Collaborative Network of electronic health records from 2010 to 2024
Inclusion criteria: adults aged 40–69 years with type 2 diabetes and at least 1 year of follow-up
Five treatment groups were defined: metformin only (reference), metformin plus DPP-4 inhibitors, metformin plus GLP-1 receptor agonists, metformin plus SGLT-2 inhibitors, and insulin monotherapy
Exposure was defined by first recorded prescription and continued use during follow-up
Landmark analyses were conducted for follow-up shorter and longer than 5 years
Results
DPP-4 inhibitors plus metformin were included as a treatment group but no significant associations with cognitive outcomes were highlighted in the abstract findings.
Five treatment groups were evaluated in total, with metformin monotherapy as the reference group
DPP-4 inhibitors plus metformin constituted one of the comparison arms; notable findings for this group were not reported in the abstract
The study followed adults for variable durations, with landmark analyses separating follow-up periods under and over 5 years
Conclusions
The authors concluded that antidiabetic medication choice may influence long-term cognitive outcomes and should be considered in diabetes management.
GLP-1 receptor agonists showed the broadest protective associations across all three cognitive outcomes
SGLT-2 inhibitors showed a narrower protective association limited to vascular dementia
The authors cautioned that the insulin monotherapy findings should not be interpreted as a direct drug effect due to confounding by indication and disease severity
The study population had a mean age of 58 years, 49.2% women, and totaled 1,528,885 adults
What This Means
This research suggests that the type of diabetes medication a person takes may be linked to their risk of developing memory and thinking problems later in life. Using health records from over 1.5 million adults with type 2 diabetes in the United States, researchers compared five different diabetes treatment approaches and tracked whether patients developed mild cognitive impairment, Alzheimer's disease, or vascular dementia over up to 14 years. People who took GLP-1 receptor agonists (such as semaglutide or liraglutide) together with metformin had notably lower rates of all three cognitive conditions compared to those on metformin alone — roughly half the risk of vascular dementia and Alzheimer's disease. People on SGLT-2 inhibitors (such as empagliflozin or dapagliflozin) combined with metformin also showed a lower risk of vascular dementia specifically.
People on insulin alone had much higher rates of cognitive disorders, but the researchers were careful to explain that this likely reflects the fact that insulin is typically prescribed to people whose diabetes is more advanced and harder to control — not that insulin itself causes cognitive decline. Factors like longer diabetes duration, kidney disease, and cardiovascular complications — which are more common in insulin users — were not fully accounted for, and the insulin group also had higher death rates, which can affect how results are interpreted.
This research suggests that newer diabetes medications, particularly GLP-1 receptor agonists, may offer benefits for brain health beyond blood sugar control, and that the choice of diabetes treatment could be a meaningful factor in long-term cognitive health. Because this was an observational study using medical records rather than a controlled clinical trial, the findings show associations rather than definitive cause-and-effect relationships, and future clinical trials would be needed to confirm these results.
Nasir A, Kilani Y, Aldiabat M, Abdelghany O, Hafez Y, Desai N, et al.. (2026). Antidiabetic medications and risk of cognitive disorders in type 2 diabetes: A retrospective cohort study.. PloS one. https://doi.org/10.1371/journal.pone.0356138