Semaglutide fails to slow Alzheimer’s disease in two large trials

The result

Novo Nordisk said its two phase 3 trials, evoke and evoke+, missed their main goal. The trials gave oral semaglutide or placebo to adults with mild cognitive impairment or mild dementia caused by Alzheimer’s disease and followed them for two years. Decline on the Clinical Dementia Rating, a combined score of memory and daily function, was about the same in both groups.

The company said the drug was well tolerated and that some Alzheimer’s-related markers in the blood improved. It stopped a planned one-year extension.

Why people hoped it would work

GLP-1 drugs lower inflammation and blood sugar, and several large health-record studies had linked them to fewer dementia diagnoses in people with type 2 diabetes. Animal work suggested effects on brain inflammation. That was enough for one of the biggest bets yet on a metabolic drug in brain disease.

What it tells us

Health-record links are a starting point, not proof. People who get a new drug may differ from people who do not in ways a database cannot see. A randomised trial removes most of those differences, and here the effect did not appear.

The trials do not settle every question. They used the oral form, studied people who already had symptoms and ran for two years. Prevention in people without symptoms, or other GLP-1 drugs, were not tested. Researchers will look closely at the full data for subgroups, but a subgroup signal after a failed main result is a reason for a new trial, not a conclusion.

Source Novo Nordisk company announcement, 24 Nov 2025; presentation at the Clinical Trials on Alzheimer’s Disease conference

Amino Weekly reports research. It is not medical advice and never gives doses.