EVOKE Trials: Why Semaglutide Failed to Treat Alzheimer's Disease

Semaglutide and the Brain: From Hope to Reality

The relationship between GLP-1 receptor agonists and brain health has been one of the most intriguing areas of metabolic research. Epidemiological studies suggested that patients taking semaglutide had significantly lower rates of Alzheimer's disease (AD) diagnosis compared to other diabetes medications. This observational signal, combined with preclinical evidence of GLP-1's neuroprotective effects, led Novo Nordisk to invest in two large Phase 3 trials — EVOKE and EVOKE+ — testing oral semaglutide for early-stage Alzheimer's disease [1].

The Preclinical and Epidemiological Rationale

Before examining the trial results, it is important to understand why there was scientific rationale for testing semaglutide in AD:

Preclinical Evidence:

• GLP-1 receptors are expressed throughout the brain, particularly in the hippocampus (memory center)
• In animal models, GLP-1 receptor agonists reduced amyloid-beta plaque formation
• GLP-1 activation improved synaptic plasticity and neuronal survival in preclinical studies
• Anti-inflammatory effects of GLP-1 could theoretically reduce neuroinflammation, a key driver of AD

Epidemiological Data:

• A large retrospective study published in Alzheimer's & Dementia (2024) found that semaglutide was associated with a significantly reduced risk of first-time AD diagnosis compared to other diabetes medications
• The association was strongest compared to insulin users (HR approximately 0.40)
• Similar signals were observed in multiple independent database analyses [2]

EVOKE and EVOKE+ Trial Design

The two companion trials were designed to rigorously test the AD hypothesis:

EVOKE:

• Phase 3, randomized, double-blind, placebo-controlled
• Oral semaglutide 14 mg vs. placebo
• Patients aged 55-85 with early symptomatic AD (mild cognitive impairment or mild dementia)
• Primary endpoint: Change in CDR-SB (Clinical Dementia Rating-Sum of Boxes) — a standard measure of cognitive and functional decline
• Duration: 104 weeks (2 years)

EVOKE+:

• Same design as EVOKE but with additional biomarker assessments
• Included CSF biomarkers, MRI volumetrics, and amyloid PET imaging
• Provided mechanistic insights into semaglutide's effects on AD pathology

Both trials were conducted across 566 sites globally, representing one of the largest investments in GLP-1 neurological research [1].

Primary Results: Both Trials Failed

The results, presented at the AD/PD 2026 conference and published in The Lancet by Cummings et al., were unambiguously negative for the primary endpoints:

EVOKE Primary Endpoint (CDR-SB change):

• Semaglutide did not significantly slow cognitive decline compared to placebo
• The treatment difference was small and not statistically significant
• Both groups showed progressive cognitive decline over the 2-year period

EVOKE+ Primary Endpoint:

• Similarly failed to demonstrate significant cognitive benefit
• Biomarker analyses showed minimal effects on core AD pathology

Key Takeaway: Oral semaglutide 14 mg does not treat established Alzheimer's disease [1].

What the Biomarker Data Showed

Despite the negative primary endpoints, the EVOKE+ biomarker data provided some interesting signals:

CSF Biomarkers:

• Semaglutide mildly shifted some CSF AD biomarkers in a favorable direction
• However, the magnitude of change was insufficient to translate into clinical benefit
• Amyloid and tau markers showed minimal change

Brain Imaging:

• MRI volumetric analyses did not show significant differences in brain atrophy rates
• Amyloid PET imaging showed no significant reduction in amyloid plaque burden

Metabolic Markers:

• Semaglutide did improve metabolic parameters (weight, HbA1c, insulin resistance) in the AD population
• However, these metabolic improvements did not translate into cognitive benefits

Why Treatment Failed but Prevention Might Work

The disconnect between epidemiological prevention signals and negative treatment results can be explained by several factors:

1. Prevention vs. treatment: The epidemiological data suggested semaglutide might prevent AD development, not reverse established disease. By the time patients have symptomatic AD, irreversible neuronal loss has already occurred.

2. Dose and formulation: The oral 14 mg dose may not achieve sufficient brain concentrations. Higher doses or injectable formulations might have different results.

3. Disease stage: The EVOKE trials enrolled patients with established cognitive symptoms. Earlier intervention — in presymptomatic or preclinical AD — might be more effective.

4. Confounding in observational data: The epidemiological associations may have been influenced by confounders (healthier patients choosing semaglutide, diabetes itself as an AD risk factor) [3].

Ongoing Research Directions

Despite the EVOKE failures, research into GLP-1 and brain health continues:

• Prevention trials: There is interest in testing semaglutide for AD prevention in high-risk individuals without symptoms
• Higher doses: Injectable semaglutide at higher doses might achieve better brain penetration
• Combination approaches: GLP-1 agonists combined with anti-amyloid therapies could provide complementary mechanisms
• Other neurological conditions: GLP-1 agonists are being studied for Parkinson's disease, where preliminary data is more encouraging

Clinical Implications

For clinicians and patients, the EVOKE results provide clear guidance:

1. Do not prescribe semaglutide for Alzheimer's treatment — there is no evidence of cognitive benefit in established AD
2. Metabolic health still matters for brain health — the epidemiological associations suggest that managing metabolic risk factors (obesity, diabetes, insulin resistance) early in life may reduce AD risk
3. Continue semaglutide for approved indications — the EVOKE failure does not diminish semaglutide's proven benefits for obesity, diabetes, cardiovascular disease, and kidney disease

References

1. Cummings JL, et al. "Efficacy and safety of oral semaglutide 14 mg (flexible dose) in early-stage symptomatic Alzheimer's disease (evoke and evoke+): two phase 3, randomised, placebo-controlled trials." The Lancet. 2026. DOI: 10.1016/S0140-6736(26)00459-9

2. Norgaard CH, et al. "Associations of semaglutide with first-time diagnosis of Alzheimer's disease." Alzheimer's & Dementia. 2024. DOI: 10.1002/alz.14313

3. Alzforum. "Semaglutide Does Not Treat Alzheimer's. Could It Prevent Dementia?" December 2025. Link