GLP-1 receptor agonist treatment reduces the risk of developing dementia.
- Current status
- Plausible unprovenEvidence unclear · Dementia
- Why
- Large observational cohorts in type 2 diabetes and a meta-analysis of RCT secondary outcomes suggest lower dementia incidence, but confounding by indication and healthy-user effects are plausible, and the meta-analysis rests on few events. No randomized trial has tested prevention: evoke/evoke+ (semaglutide) and ELAD (liraglutide) treated people who already had Alzheimer's disease, so their null primary results bear on treatment, not on prevention. Prevention in cognitively normal people has not been tested in a randomized trial.
- What would change it
- A randomized prevention trial with incident dementia or validated cognitive decline as the primary outcome.
Observational replication using active comparators with negative-control outcomes and without diabetes as an entry requirement. - Linked studies
- 6 · 4 supports · 2 doesn't directly test the claim
- Last updated
- Oct 19, 2026
The status is the collection's own judgment on its assessment scale, not a GRADE certainty rating and not medical advice.
The evidence, sorted by what it shows
Supports (4)
- GLP-1RA and SGLT2i Medications for Type 2 Diabetes and Alzheimer Disease and Related DementiasRetrospective cohort · 2025 · Study tier 3 · Different population · Benefit
Target-trial emulation in T2D: ADRD HR 0.67 vs other glucose-lowering drugs; no difference vs SGLT2i.
- Cardioprotective Glucose-Lowering Agents and Dementia Risk: A Systematic Review and Meta-AnalysisMeta-analysis · 2025 · Study tier 3 · Different population · Benefit
Meta-analysis of randomized trials of glucose-lowering drugs: GLP-1 RAs were associated with lower dementia (OR 0.55, 95% CI 0.35-0.86), while SGLT2 inhibitors (OR 1.20, 95% CI 0.67-2.17) and all cardioprotective glucose-lowering drugs pooled (OR 0.83, 95% CI 0.60-1.14) were not.
- Mapping the effectiveness and risks of GLP-1 receptor agonistsRetrospective cohort · 2025 · Study tier 3 · Different population · Mixed
VA atlas: reduced neurocognitive disorders vs usual care; observational.
- Associations of semaglutide with first-time diagnosis of Alzheimer's disease in patients with type 2 diabetes: Target trial emulation using nationwide real-world data in the USRetrospective cohort · 2024 · Study tier 4 · Different population · Benefit
EHR emulation in T2D: semaglutide associated with 40-70% lower first AD diagnosis; short 3-year window; diagnosis-code outcome.
Doesn't directly test the claim (2)
- Efficacy and safety of oral semaglutide 14 mg (flexible dose) in early-stage symptomatic Alzheimer's disease (evoke and evoke+): two phase 3, randomised, placebo-controlled trialsRandomized trial · 2026 · Study tier 1 · Partial match · Null · Industry funded
evoke/evoke+ tested treatment of established early Alzheimer's disease (no significant difference in CDR-SB at 104 weeks in 3808 people); it does not test prevention of dementia in people without it.
- Liraglutide in mild to moderate Alzheimer's disease: a phase 2b clinical trialRandomized trial · 2026 · Study tier 3 · Partial match · Mixed · Partly industry funded
ELAD phase 2b in 204 people with mild to moderate Alzheimer's disease without diabetes: no difference in cerebral glucose metabolism (primary), ADAS-Exec nominally better, and no difference in ADCS-ADL or CDR-SoB. It treated established disease, so it does not test prevention of dementia.
How this assessment has changed
- Plausible unprovenOct 19, 2026 · Evidence unclear
Owner decision 2026-10-18: CONTRADICTORY/weakening -> PLAUSIBLE_UNPROVEN/unclear. evoke/evoke+ re-linked as does_not_directly_test (treatment of established AD, not prevention); no linked record contradicts the claim.
- ContradictorySep 13, 2026 · Evidence weakening
initial curated status