EVOKE Trial Results: Semaglutide Fails Alzheimer's Test — But the Prevention Story May Be Different
The largest GLP-1 trials ever conducted in Alzheimer's disease missed their primary endpoints. Here's what the biomarker data and observational studies actually show.
Semaglutide did not slow cognitive decline in people who already had early Alzheimer's disease — the EVOKE trials failed their primary endpoints. However, the drug showed ~10% reductions in key Alzheimer's biomarkers in cerebrospinal fluid, and separate observational studies show 40-70% lower Alzheimer's diagnosis rates in diabetic patients taking semaglutide. The emerging consensus: GLP-1s may prevent rather than treat Alzheimer's.
The EVOKE Trials: What Happened
On November 24, 2024, Novo Nordisk announced that its Phase 3 EVOKE and EVOKE+ trials had failed. Oral semaglutide — the same active ingredient in Ozempic and Rybelsus — did not slow disease progression in patients with early-stage Alzheimer's disease compared to placebo.
Full results were presented at the Clinical Trials on Alzheimer's Disease (CTAD) conference in San Diego on December 3, 2024.
Trial Design
EVOKE (NCT04777396) and EVOKE+ (NCT04777409) were randomized, double-blind, placebo-controlled trials — the gold standard for clinical evidence. Together, they represent the largest GLP-1 trials ever conducted in a neurodegenerative disease.
| Parameter | EVOKE / EVOKE+ |
|---|---|
| Total Participants | 3,808 adults (ages 55-85) |
| Patient Population | Mild cognitive impairment (MCI) or mild dementia due to Alzheimer's with confirmed amyloid positivity |
| Countries | Nearly 40 |
| Treatment Duration | 104 weeks (main phase) + 52-week extension (cancelled) |
| Drug | Oral semaglutide 14mg daily vs placebo |
| Primary Endpoint | Change in CDR-SB (Clinical Dementia Rating - Sum of Boxes) at Week 104 |
| Key Difference | EVOKE+ enrolled patients with evidence of small vessel cerebrovascular disease |
Primary Results: No Cognitive Benefit
The trials failed on every cognitive and functional endpoint.
Semaglutide did not demonstrate superiority over placebo on change in CDR-SB score at Week 104. Curves for the semaglutide and placebo groups were essentially indistinguishable throughout the study.
Secondary endpoints — including ADCS-ADL-MCI (activities of daily living), Montreal Cognitive Assessment, ADAS-Cog13, MMSE, and time to progression from MCI to dementia — all showed no benefit from semaglutide treatment.
Hazard ratio: 0.96 (95% CI, 0.86-1.06)— No statistically significant difference between semaglutide and placebo. Source: NeurologyLive CTAD 2024 Coverage
Based on these negative results, Novo Nordisk cancelled the planned 1-year extension period of both trials.
The Biomarker Data: A Glimmer of Biological Activity
Despite the clinical failure, something interesting showed up in the 199-patient cerebrospinal fluid (CSF) substudy: semaglutide did appear to affect Alzheimer's-related biomarkers.
CSF Biomarker Changes
| Biomarker | Category | Treatment Ratio (Semaglutide vs Placebo) |
|---|---|---|
| pTau181 | AD Pathology | 0.92 (~8% reduction) |
| pTau217 | AD Pathology | 0.91 (~9% reduction) |
| npTau181 | AD Pathology | 0.90 (~10% reduction) |
| npTau205 | AD Pathology | 0.91 (~9% reduction) |
| YKL-40 | Neuroinflammation | 0.93 (~7% reduction) |
| Total Tau | Neurodegeneration | 0.93 (~7% reduction) |
| Neurogranin | Synaptic Integrity | 0.92 (~8% reduction) |
Why didn't biomarker changes translate to clinical benefit?
At the CTAD presentation, lead investigator Peter Johannsen addressed this discrepancy. His explanation, as reported by AlzForum:
The 10 percent shift in CSF AD biomarkers in response to semaglutide was in the same ballpark as biomarker changes observed in trials of other drugs that showed no efficacy, while drugs that were approved—lecanemab and donanemab—had triggered larger biomarker responses.— Peter Johannsen, Novo Nordisk, CTAD 2024
Johannsen suggested future Phase 2 trials should require larger biomarker changes — 20%, 25%, perhaps 30% — before advancing to Phase 3.
Inflammatory Markers
One notable finding: semaglutide significantly reduced plasma high-sensitivity C-reactive protein (hs-CRP), a marker of systemic inflammation.
| Trial | hs-CRP Treatment Ratio |
|---|---|
| EVOKE | 0.76 (24% reduction) |
| EVOKE+ | 0.71 (29% reduction) |
Principal investigator Jeffrey Cummings, MD, noted: "Our underlying biological hypothesis was, if we can suppress peripheral inflammation, we will have a beneficial impact on cognition."
That hypothesis was not supported by the clinical outcomes.
Did Semaglutide Reach the Brain?
Yes. A CSF substudy confirmed that semaglutide crossed the blood-brain barrier, though at low concentrations:
After 12 weeks of treatment, median CSF concentration was 0.14 nmol/L versus 34.8 nmol/L in plasma — a ratio of approximately 0.4% CSF to plasma.
The Prevention Paradox: Observational Data Tells a Different Story
Here's where it gets interesting. While semaglutide failed to treat existing Alzheimer's disease, observational studies consistently show people taking GLP-1 medications are diagnosed with dementia at dramatically lower rates.
Wang et al., 2024: 67% Lower Alzheimer's Risk
Published in Alzheimer's & Dementia (October 2024), this target trial emulation analyzed electronic health records from 116 million U.S. patients:
| Semaglutide vs. | Hazard Ratio for AD Diagnosis | Risk Reduction |
|---|---|---|
| Insulin | 0.33 (95% CI: 0.21-0.51) | 67% lower |
| Metformin | ~0.50 | ~50% lower |
| Other GLP-1 RAs | 0.59 (95% CI: 0.37-0.95) | 41% lower |
The study included over 1 million patients with type 2 diabetes followed for 3 years. Results were consistent across obesity status, gender, and age groups.
Wang et al., 2025: Similar Findings for All Dementia Types
A follow-up study published in the Journal of Alzheimer's Disease (June 2025) expanded the analysis to 1.7 million patients and examined dementia subtypes:
| Semaglutide vs. | Hazard Ratio (All ADRD) |
|---|---|
| Insulin | 0.54 (46% lower risk) |
| Metformin | 0.67 (33% lower risk) |
| Other GLP-1 RAs | 0.80 (20% lower risk) |
The strongest protective effect was seen for vascular dementia. No significant associations were found for frontotemporal dementia or Lewy body dementia.
JAMA Network Open, 2025: Semaglutide and Tirzepatide
A study of over 60,000 adults with type 2 diabetes and obesity found patients taking semaglutide or tirzepatide had a 37% lower risk of developing dementia (HR 0.63) over 7 years compared to other antidiabetic drugs.
Treatment vs. Prevention: The EVOKE trials tested whether semaglutide could slow cognitive decline in people who already had Alzheimer's pathology (confirmed amyloid). The observational studies track whether people taking semaglutide are later diagnosed with dementia. These answer fundamentally different questions.
Expert Reactions
Even negative trials move the field forward as they still teach us something. If we look to the early anti-amyloid studies, which were also negative, they offered critical lessons that informed later trials and ultimately helped bring drugs like Leqembi and Kisunla to market... As we dig further into the data, there is much we can learn, including exploring the potential of GLP-1 drugs as a preventive therapy, which may still hold promise.— ADDF Statement, November 24, 2024
Although GLP-1 drugs had shown signs they could prevent neurodegenerative diseases, many researchers thought they were unlikely to help once the disease was underway.— Science, November 24, 2024
Based on the significant unmet need in Alzheimer's disease as well as a number of indicative data points, we felt we had a responsibility to explore semaglutide's potential, despite a low likelihood of success.— Novo Nordisk Press Release, November 24, 2024
What This Means Going Forward
For Alzheimer's Treatment: Limited Immediate Impact
Semaglutide will not become an Alzheimer's treatment based on EVOKE data. The trials definitively showed no benefit for patients with existing disease.
For Prevention Research: The Question Remains Open
Several factors support continued investigation of GLP-1s for dementia prevention:
- Consistent observational data showing 40-70% lower dementia diagnosis rates
- Modest but consistent biomarker reductions in CSF
- Anti-inflammatory effects confirmed by hs-CRP reductions
- Treatment of multiple modifiable risk factors (diabetes, obesity, cardiovascular disease)
Experts have called for prevention trials in high-risk populations — particularly APOE4 homozygotes — before symptoms develop.
Combination Therapy Potential
As Fillit noted: "Existing anti-amyloid drugs slow cognitive decline by around 30%, so therapies aimed at other pathways will be crucial as we chip away at the remaining 70%."
With over 70% of the Alzheimer's drug pipeline now focused on non-amyloid targets, GLP-1s may eventually find a role in combination with approved treatments like lecanemab (Leqembi) and donanemab (Kisunla).
Limitations and Caveats
On the EVOKE trials: These were well-designed, rigorous trials that definitively answered the treatment question. The negative result is reliable.
On the observational studies: Target trial emulations cannot establish causation. Potential confounders include:
- Healthy user bias (patients well enough to take semaglutide may be healthier overall)
- Detection bias (weight loss medications may lead to more medical contact)
- Diabetes and obesity management itself may reduce dementia risk independent of semaglutide
- Dementia diagnoses were based on ICD-10 codes, not confirmed with biomarkers
Only randomized prevention trials can determine if GLP-1s actually prevent dementia.
Sources
- ClinicalTrials.gov: NCT04777396 (EVOKE)
- ClinicalTrials.gov: NCT04777409 (EVOKE+)
- Cummings JL, et al. "evoke and evoke+: design of two large-scale, double-blind, placebo-controlled, phase 3 studies." Alzheimer's Research & Therapy, January 2025
- Wang W, et al. "Associations of semaglutide with first-time diagnosis of Alzheimer's disease in patients with type 2 diabetes." Alzheimer's & Dementia, December 2024
- Wang W, et al. "Associations of semaglutide with Alzheimer's disease-related dementias." Journal of Alzheimer's Disease, June 2025
- Alzheimer's Drug Discovery Foundation: Readout of Phase 3 Semaglutide Trials Marks Critical Moment
- Novo Nordisk: Topline Results Press Release
Considering GLP-1 Medications?
Research into cognitive benefits is ongoing. If you're interested in GLP-1 medications for their approved uses (weight loss and diabetes), see our comparison resources.
Affiliate Disclosure: SourceGLP-1 may receive compensation from partner sites linked above. This does not influence our research or editorial content.