Podcast episode 108 (in German)
Semaglutide against Alzheimer’s – the biggest hope has a result
October 16, 2026
One of the largest trials in Alzheimer’s research, 3,808 people over 2 years, has a result for semaglutide that fits in one line: minus 0.08 points. The episode explains why that is a clean no — and why it still does not lower the evidence score.
What it is about
The hope that GLP-1 drugs could slow Alzheimer’s rested on 3 good arguments: animal studies, a plausible mechanism and registry data according to which people with type 2 diabetes on such drugs developed dementia less often. This was tested in 2 trials, Evoke and Evoke plus, in people with mild cognitive impairment or mild dementia and previously confirmed Alzheimer’s deposits. After 2 years, both groups had deteriorated to the same extent, and no subgroup showed a benefit. At the same time, biomarkers in the cerebrospinal fluid fell. The episode explains why a fallen marker is not yet a result and what the trial says about prevention: nothing, because it did not ask that question. The manufacturer has ended the program and published the data in full.
Key points
- Evoke and Evoke plus enrolled 1,855 and 1,953 people aged 55 to 85, spread across 566 centers in 40 countries.
- After 104 weeks, the drug and placebo groups had each worsened by about 2.3 points on the dementia scale, starting from 3.7 points — the curves lie on top of each other.
- The difference was minus 0.08 points in the first trial, range minus 0.35 to plus 0.2, p-value 0.57; in the second, plus 0.1 in favor of placebo, p-value 0.46.
- In 199 participants, tau proteins and inflammatory markers in the cerebrospinal fluid fell by around 10 percent, statistically significant — memory did not follow.
- Adverse events occurred in 91.2 percent on the drug and 84.8 percent on placebo; of 5 deaths attributed to the study medication, 4 were in the placebo group.
- In December the figure was still minus 0.06, as a conference presentation; the peer-reviewed full publication in the Lancet gives minus 0.08 with a narrower range.
- The episode leaves semaglutide at human evidence 10, plus safety 9, use 9, mechanism 9 and hype 7 with a positive direction — that applies to weight and cardiovascular health, not to the brain.
Where the data stop
The trial says something smaller and more precise than “GLP-1 does nothing for the brain”: in people who are already ill and have confirmed deposits, semaglutide does not slow the course of the disease. It says nothing about healthy people, because every participant had a diagnosis. That the substance might work 20 years earlier is plausible, but before this trial everything was plausible — prevention has not been tested, and “not tested” makes it an open question, not a hope. Why the figure shifted between the presentation and the publication, the publication does not say.
The forums do not help either. An analysis of just over 400,000 posts from 9 forums over 6 years counted only complaints; cognitive problems appear in 267 reports, 0.9 percent. People who experience a clearer head do not feature in it, and these are self-reports without a control group in a preprint without peer review. Anyone taking semaglutide for weight or diabetes has reasons this trial does not touch.
Listen (in German)
Listen on Spotify Listen on Apple Podcasts Listen on Amazon Music
Published on October 16, 2026.
The topic in the database
- The blockbuster among the “weight-loss injections”: approved as Ozempic/Wegovy, highly effective for weight and with a proven cardiovascular benefit – a legal, …Semaglutide (Ozempic / Wegovy)
All episodes in the app (German)
Information only, not medical advice and not a usage or dosage recommendation. Prescription-only and unapproved substances belong in the hands of a physician. Last updated: 2026-09-22.