Semaglutide Impacts Proteomic Signatures Related to Dementia Risk in the SELECT Cardiovascular Outcomes Trial
Paul Mystkowski1, Marti Jimenez Mausbach2, Jan Christian Refsgaard2, Gabriel Martino2, Teresa Leon2, Betty Tijms3, Clare Paterson4
1Novo Nordisk Inc., 2Novo Nordisk A/S, 3Alzheimer Center Amsterdam, Neurology, Vrije Universiteit Amsterdam, Amsterdam UMC location VUmc, 4Clinical Research and Development, Standard BioTools
Objective:

To assess the effect of once-weekly semaglutide up to 2.4mg on proteomic signatures of dementia risk in a population aged ≥65 years with obesity or overweight and established cardiovascular disease (CVD) without diabetes, using serum samples from the SELECT cardiovascular outcomes trial (NCT03574597).

Background:

Plasma proteomic signatures preceding onset of dementia can yield insights into disease pathobiology and reveal novel biomarkers and avenues for intervention. The Dementia SomaSignal™ Test (dSST) is a blood-based, machine learning-derived 25-protein-only score that predicts 5- and 20-year risk of all-cause dementia. Evidence from real-world studies and clinical trials suggests that glucagon-like peptide-1 receptor agonists (e.g. semaglutide) may have therapeutic potential in Alzheimer’s disease (AD).

Design/Methods:

SELECT trial participants received semaglutide (≤2.4mg) or placebo for up to 5 years. Non-fasted serum samples of 2970 participants aged ≥65 years were collected at baseline, week 20 and week 104, and used to assess the 5- and 20-year risk of dementia diagnosis with dSST.

Results:

After 2 years’ treatment, the increase in dSST-predicted 5-year dementia risk was 2.5 times smaller with semaglutide versus placebo, corresponding to a 26% lower expected 5-year dementia event rate (odds ratio [OR]=0.74; 95% confidence interval [CI]:0.64;0.85). The predicted 20-year dementia risk was 1.67 times smaller with semaglutide versus placebo, corresponding to an 8.8% (OR=0.91; 95% CI:0.88;0.94) lower expected 20-year dementia event rate. An ordinal regression analysis revealed that semaglutide treatment was associated with a significant decrease in the odds of being in the higher dementia risk groups at end of trial compared with placebo (β=−0.44; standard error=0.078; p<0.001). Treated participants had 36% lower odds of being in a higher dementia risk category versus placebo (OR=0.64; 95% CI:0.55;0.75).

Conclusions:

Semaglutide appears to modify dementia-related proteomic signatures in participants with obesity or overweight and established CVD without diabetes.

 Previously presented at CTAD25 and published in JPAD (Jiménez-Mausbach et al. 2025;DOI:TBC )

10.1212/WNL.0000000000215557
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.