Semaglutide Linked to Lower Predicted Dementia Risk in SELECT Analysis

A post hoc analysis of the SELECT phase 3 trial reports that semaglutide was associated with a smaller increase in predicted dementia risk among adults aged 65 and older with overweight or obesity and cardiovascular disease who did not have diabetes. The paper appeared on August 8, 2026 in Alzheimer's & Dementia: Diagnosis, Assessment & Disease Monitoring.

How the analysis was run

The analysis drew on data from SELECT (Semaglutide Effects on Heart Disease and Stroke in Patients With Overweight or Obesity). It covered 2,970 participants randomized to semaglutide 2.4 mg or placebo. Non-fasted serum samples collected at baseline and at week 104 were analyzed with the Dementia SomaSignal Test (dSST), a validated 25-protein score that predicts 5- and 20-year all-cause dementia risk.

What the results showed

Compared with placebo, semaglutide reduced the rise in five-year risk 2.5-fold: the predicted event rate was 26.0% lower (odds ratio 0.74; 95% confidence interval 0.65–0.85). The increase in 20-year risk was 1.67-fold smaller (8.8% lower; OR 0.91; 95% CI 0.88–0.94). The odds of falling into a higher dementia risk category fell by 36% (β −0.44; P < 0.001).

What limits the conclusions

This is a post hoc analysis of a trial whose primary aim was cardiovascular outcomes, and the measured endpoint is a proteomics-based risk signature rather than diagnosed dementia. The authors note that plasma proteomics can detect multi-pathway biological changes that precede dementia onset, and that preclinical and clinical data suggest GLP-1 receptor agonists may have neuroprotective effects. Several authors are employees of Novo Nordisk and Illumina; the paper's discussion is limited to the conclusion that semaglutide slowed progression of the validated risk signature.