Education and Alzheimer Disease Genetic Risk in Associations of GLP-1 Receptor Agonists With Dementia Among Adults With Type 2 Diabetes

@misc{wang2026education,
  title = {{Education and Alzheimer Disease Genetic Risk in Associations of GLP-1 Receptor Agonists With Dementia Among Adults With Type 2 Diabetes}},
  author = {J. Wang and A. M. Pederson and M. D. Flanders and M. Choi and P. Buto and K. D. Sims and R. Chen and E. Couch and P. Gilsanz and K. N. Hayes and A. R. Zullo and A. Stokes and M. M. Glymour and S. F. Ackley},
  howpublished = {medRxiv},
  year = {2026},
  month = sep,
  doi = {10.64898/2026.09.04.26362304},
  url = {https://www.medrxiv.org/content/10.64898/2026.09.04.26362304v1},
  note = {CC BY},
}

TY  - UNPB
TI  - Education and Alzheimer Disease Genetic Risk in Associations of GLP-1 Receptor Agonists With Dementia Among Adults With Type 2 Diabetes
AU  - J. Wang
AU  - A. M. Pederson
AU  - M. D. Flanders
AU  - M. Choi
AU  - P. Buto
AU  - K. D. Sims
AU  - R. Chen
AU  - E. Couch
AU  - P. Gilsanz
AU  - K. N. Hayes
AU  - A. R. Zullo
AU  - A. Stokes
AU  - M. M. Glymour
AU  - S. F. Ackley
PY  - 2026
DA  - 2026/09/04
PB  - medRxiv
DO  - 10.64898/2026.09.04.26362304
UR  - https://www.medrxiv.org/content/10.64898/2026.09.04.26362304v1
AB  - ObjectiveTo evaluate whether educational attainment and Alzheimer disease genetic risk were associated with GLP-1 receptor agonist initiation and dementia incidence and whether adjustment for these measured factors materially changed the estimated association between GLP-1 receptor agonist initiation and incident dementia among adults with type 2 diabetes.  Research Design and MethodsWe conducted an observational cohort study using linked electronic health record, survey, and genetic data from 14,364 All of Us Research Program participants with type 2 diabetes. We estimated associations of educational attainment and Alzheimer disease genetic risk with treatment initiation and incident dementia and compared GLP-1 receptor agonist initiators with initiators of non-sodium-glucose cotransporter 2 inhibitor second-line therapies, with a separate sodium-glucose cotransporter 2 inhibitor comparison. Models were estimated before and after additional adjustment for educational attainment, APOE {varepsilon}4, and non-APOE genetic risk.  ResultsAmong 14,364 participants (mean age, 60.2 years; 54.2% female), the mean follow-up duration was 4.3 years. The estimated hazard ratio for dementia comparing GLP-1 receptor agonist initiation with non-SGLT2 inhibitor second-line therapy was 0.85 (95% CI 0.64-1.12) before adjustment for education or Alzheimer disease genetic risk and 0.84 (95% CI 0.64-1.12) after adjustment for educational attainment, APOE {varepsilon}4, and non-APOE genetic risk.  ConclusionsAmong adults with type 2 diabetes, adjustment for measured educational attainment and Alzheimer disease genetic susceptibility produced little change in the estimated association between GLP-1 receptor agonist initiation and incident dementia. These findings do not exclude confounding by these factors in other populations or residual confounding from socioeconomic, clinical, behavioral, and health-care-related factors.  HighlightsO_LIObservational studies have reported associations between GLP-1 receptor agonist use and lower dementia risk, but important social and genetic factors are often unavailable in electronic health record data. C_LIO_LIWe examined whether educational attainment and Alzheimer disease genetic susceptibility were associated with treatment initiation and dementia and whether adjustment for these factors changed GLP-1 receptor agonist-dementia estimates. C_LIO_LIAdding measured education and genetic susceptibility produced little change in the estimated conditional hazard ratios. C_LIO_LIThese results do not exclude residual confounding and should not be interpreted as evidence that GLP-1 receptor agonists prevent dementia. C_LI
ER  -