GLP-1 Drugs and Alzheimer's: What the EVOKE and ELAD Trials Reveal
Adrian Carter·Former metabolic disease researcher turned health writer. Breaks down how hormones like GLP-1 shape your weight, appetite, and energy — no jargon required.··11 min read
GLP-1 Drugs and Alzheimer's: What the EVOKE and ELAD Trials Reveal
Two of the most closely watched clinical trials in neuroscience just reported results — and the story they tell is genuinely complicated. GLP-1 Alzheimer's disease research produced one clear negative, one partial positive, and a mountain of real-world data pointing in a direction that researchers are not ready to dismiss.
Here is what the science actually shows, what it does not, and what it means for anyone watching this space.
What GLP-1 Receptor Agonists Are and How They Reach the Brain
GLP-1 receptor agonists (GLP-1RAs) are a class of drugs originally designed to lower blood sugar in type 2 diabetes. They mimic glucagon-like peptide-1 (GLP-1), a hormone your gut releases after eating that signals the pancreas to release insulin, slows stomach emptying, and tells your brain you are full. Semaglutide (sold as Ozempic and Wegovy) and liraglutide (sold as Victoza and Saxenda) are the two most recognized members of this family.
What caught neuroscientists' attention is where GLP-1 receptors show up inside the brain. They are not just in the hypothalamus regulating appetite. GLP-1 receptors are expressed on neurons, oligodendroglia, astroglia, microglia, and endothelial cells throughout the brain — the full cellular cast of the central nervous system [10]. When a GLP-1RA binds these receptors, it triggers cascading signals (cAMP/PKA, PI3K/Akt, and MAPK pathways) that promote neuronal survival, suppress neuroinflammation, and reduce oxidative stress.
There is another mechanism worth understanding. Alzheimer's disease (AD) involves what researchers sometimes call "type 3 diabetes" — a state of insulin resistance specific to brain tissue. In this state, the normal IRS-1/PI3K/Akt signaling that neurons depend on for energy and survival breaks down. GLP-1RAs appear to restore that signaling, potentially addressing one of the upstream drivers of neurodegeneration rather than just mopping up amyloid plaques after the fact [10]. This is why the idea made biological sense before any trial data came in — and why a negative trial result does not necessarily close the book.
Who Might Be Most Affected by Alzheimer's and Why This Drug Class Was a Logical Candidate
Alzheimer's disease is not a single uniform condition. It is a spectrum, and where someone sits on that spectrum matters enormously for whether any intervention can help. The ideal window for intervention is early — before extensive neuronal death has occurred. The EVOKE and EVOKE+ trials enrolled adults aged 55 to 85 with mild cognitive impairment (MCI) or mild dementia who tested positive for amyloid on biomarker screens, representing the earliest symptomatic stages .
This content is for informational purposes only and is not intended as medical advice, diagnosis, or treatment. Always consult a qualified healthcare provider before starting any supplement or making changes to your health regimen.
AC
Adrian Carter
Former metabolic disease researcher turned health writer. Breaks down how hormones like GLP-1 shape your weight, appetite, and energy — no jargon required.
Former metabolic disease researcher turned health writer. Breaks down how hormones like GLP-1 shape your weight, appetite, and energy — no jargon required.
GLP-1Alzheimer's diseasesemaglutideliraglutide
Related Articles
neurodegeneration
GLP-1 & Metabolics
What Is Allulose and Does It Really Boost GLP-1?
Allulose is a rare sugar linked to lower postprandial glucose and possible GLP-1 stimulation. Here is what the allulose GLP-1 science shows.
People with type 2 diabetes (T2D) have consistently shown elevated Alzheimer's risk in epidemiological data — roughly 50 to 65 percent higher than the general population. This overlap between metabolic dysfunction and neurodegeneration is not coincidental. Insulin resistance, chronic inflammation, and vascular injury are common threads running through both conditions. If GLP-1RAs address these shared mechanisms, T2D patients on these drugs would be a natural group to watch — and real-world databases have now followed over a million of them [6].
The biological case extends beyond diabetes. Obesity itself is an independent risk factor for dementia, and GLP-1RAs dramatically reduce body weight and systemic inflammation. A large propensity-matched cohort of 102,935 obese patients found that those on GLP-1RAs had a relative risk of 0.627 for Alzheimer's disease — roughly 37 percent lower than matched controls — along with reductions in Lewy body dementia risk (RR 0.590) and vascular dementia risk (RR 0.438) [7]. The signal was consistent enough across metabolic risk groups that clinical researchers decided to run prospective trials directly in AD patients, regardless of diabetes status.
What the EVOKE and ELAD Trials Actually Found
This is where the data gets interesting, and honest.
The EVOKE and EVOKE+ trials were the largest GLP-1 Alzheimer's disease trials ever conducted. They enrolled 3,808 amyloid-positive participants across multiple countries and randomized them to oral semaglutide 14 mg daily or placebo for 104 weeks. The primary endpoint was the Clinical Dementia Rating Sum of Boxes (CDR-SB), a validated measure of how well someone functions cognitively day-to-day. In November 2025, topline results landed: semaglutide did not achieve superiority over placebo [1]. The CDR-SB curves were described by investigators as "exactly on top of each other."
That is a clear negative result for the clinical outcome. But EVOKE was not a biomarker-free trial. Secondary analyses showed nominally significant improvements: p-tau (a protein associated with the neurofibrillary tangles of Alzheimer's) was approximately 10 percent lower in the semaglutide group, and high-sensitivity C-reactive protein (hs-CRP), a marker of systemic inflammation, dropped by roughly 30 percent [1]. These biomarker shifts tell you something biological is happening. They just did not translate into measurable cognitive protection over two years in this population.
The ELAD trial told a different kind of story. ELAD enrolled 204 non-diabetic participants with mild to moderate Alzheimer's and treated them with daily subcutaneous liraglutide 1.2 to 1.8 mg for 52 weeks [3]. The primary endpoint — preservation of cerebral glucose metabolism as measured by FDG-PET scan — was missed (P=0.14). If you stopped reading there, you would miss the more striking secondary findings. Exploratory MRI showed approximately 50 percent less gray matter volume loss in frontal, temporal, and parietal regions in the liraglutide group compared to placebo. Executive function, measured by the ADAS-Exec subscale, improved significantly (P=0.01). Overall cognitive decline was approximately 18 percent slower over the 12-month period [3]. These are not trivial signals.
The honest interpretation is this: neither trial achieved its primary endpoint, which means neither drug is ready for clinical approval in Alzheimer's disease. But ELAD found structural brain protection and functional cognitive signals that are hard to dismiss. EVOKE found biomarker changes without clinical translation. Together, they do not prove GLP-1RAs work in Alzheimer's — they raise a sharper, more specific set of questions about dose, timing, duration, and which patients might benefit.
The Real-World Evidence That Preceded and Surrounds the Trials
Before EVOKE's results landed, observational data had already been building a case that GLP-1RAs were doing something protective in large populations. A meta-analysis of 26 randomized controlled trials covering 164,531 participants found that GLP-1RAs were the only cardioprotective glucose-lowering drug class with a statistically significant reduction in all-cause dementia risk (odds ratio 0.55; 95% CI 0.35 to 0.86) [5]. SGLT2 inhibitors, despite their own cardiovascular benefits, showed no significant effect (OR 1.20). That is not a small signal in a well-powered meta-analysis.
A target trial emulation study using a nationwide US database of 1,094,761 T2D patients found that semaglutide was associated with a 67 percent reduction in first-time AD diagnosis versus insulin users (HR 0.33) and a 41 percent reduction versus users of other GLP-1RAs (HR 0.59) [6]. That second comparison is particularly important — it suggests semaglutide may have advantages over older GLP-1RAs, possibly due to its greater CNS penetration or potency at GLP-1 receptors.
A multinational real-world cohort of 214,442 matched individuals followed for a mean of four years found that GLP-1RAs were associated with 19 percent lower overall neurodegeneration risk versus DPP4 inhibitors, with an AD-specific hazard ratio of 0.77 and a dementia hazard ratio of 0.76 [8]. A separate analysis of GLP-1RA users versus SGLT2 inhibitor users found GLP-1RA hazard ratios of 0.67 for Alzheimer's and related dementias [9]. These studies all carry the usual limitations of observational research — confounding, indication bias, the healthy-user effect. But the consistency across different databases, countries, and comparator classes is notable [5][6][7][8].
The critical question observational data cannot answer is whether this represents genuine neuroprotection or simply better metabolic control reducing downstream vascular risk. That is precisely what EVOKE and ELAD were designed to resolve — and why their mixed results matter.
What to Watch Out For: Side Effects in Older Adults
Understanding safety in the Alzheimer's population matters especially because this is a vulnerable group. A pooled analysis of semaglutide safety data specifically in adults aged 65 and older — the demographic most likely to develop AD — provides a useful picture [11].
Gastrointestinal (GI) disorders were the most frequent adverse events, affecting 44.6 to 73.8 percent of participants aged 65 and older depending on the formulation and dose [11]. This includes nausea, vomiting, and diarrhea — familiar side effects to anyone following GLP-1RA research. Overall, 73.6 to 92.4 percent of older adults experienced at least one adverse event. Discontinuation due to adverse events ran 9.3 to 12.4 percent in older adults, compared to 5.7 to 8.7 percent in the overall adult population. Older adults are more sensitive, and the GI burden is real.
One side effect worth flagging specifically for the AD context is weight loss. Semaglutide produced an estimated 3.8 percent body weight reduction at 52 weeks in adults 65 and older [11]. In younger, obese adults trying to lose weight, that is a feature. In an Alzheimer's patient who may already be at risk for malnutrition, unintentional weight loss is a concern that warrants monitoring.
EVOKE also detected small increases in two neurological biomarkers — GFAP (glial fibrillary acidic protein, a marker of astrocyte activation) and NfL (neurofilament light chain, a marker of neuronal injury), both approximately 5 percent higher in the semaglutide group versus placebo [1]. The clinical significance of these changes is considered low, but they add a note of caution about the brain's response to the drug over long timeframes. Importantly, EVOKE data showed no clinically significant interactions between semaglutide and cholinesterase inhibitors, the standard-of-care drugs used by 41.7 percent of North American participants and 61.6 percent of participants in Asia [2].
What This Means for You Right Now
Let's be direct about where things stand. As of March 2026, no GLP-1 receptor agonist has an FDA or EMA approval for Alzheimer's disease. The largest trial to date — EVOKE — did not show cognitive benefit. That is the regulatory and clinical reality, and it matters.
What the full body of evidence suggests, taken carefully, is that GLP-1RAs may offer protection at a population level — particularly for people with T2D or obesity who have elevated AD risk. If you are already taking semaglutide or liraglutide for weight management or diabetes, the real-world data indicating reduced dementia risk is genuinely encouraging news. You are not getting a drug with proven anti-Alzheimer's efficacy, but you may be getting a metabolic effect that happens to reduce downstream neurological risk. That is a meaningful distinction [5][6].
If you have a family history of Alzheimer's or carry known risk factors, the smartest conversation to have right now is with a neurologist or geriatrician who follows this literature. The ELAD structural findings and the consistent real-world signals are enough to make this a legitimate topic for clinical discussion. What they are not is sufficient evidence to take or adjust any medication without that guidance. The EVOKE results remind us that biological plausibility and biomarker improvement do not always translate into the clinical outcomes patients care about most. Science is still working this one out.
Frequently Asked Questions
Did the EVOKE trial prove that GLP-1 drugs do not work in Alzheimer's?
Not exactly. EVOKE showed that oral semaglutide 14 mg did not outperform placebo on the CDR-SB cognitive scale over 104 weeks in early-stage Alzheimer's patients. That is a negative result for that drug, at that dose, in that population, over that timeframe. It does not rule out benefit in other populations, at different doses, over longer periods, or with different GLP-1RAs. Biomarker changes in EVOKE and structural brain changes in ELAD suggest the biology is not inert.
Why did the ELAD trial show brain structure improvements if the primary endpoint was negative?
ELAD's primary endpoint was FDG-PET cerebral glucose metabolism, which did not show a statistically significant difference (P=0.14). But secondary MRI analyses found approximately 50 percent less gray matter volume loss, and executive function scores improved significantly (P=0.01). Primary endpoints in clinical trials are pre-specified for regulatory approval purposes. Missing the primary endpoint means the drug does not qualify for approval on that basis, but the secondary findings are still scientifically meaningful and warrant follow-up.
Are GLP-1 drugs safe for older adults with Alzheimer's?
Safety data from pooled semaglutide trials in adults aged 65 and older shows the drug is generally tolerable, but GI side effects are frequent (44.6 to 73.8 percent) and discontinuation rates are higher in older adults than in the general population. Unintentional weight loss is also a concern in AD patients. Any decision about using a GLP-1RA in an older adult should be made with a healthcare provider who can weigh individual risk and benefit.
Does taking semaglutide for diabetes or weight loss protect against Alzheimer's?
Real-world data consistently shows reduced AD risk in T2սD patients taking GLP-1RAs. One large study found a 67 percent reduction versus insulin users and a meta-analysis of 26 trials found an odds ratio of 0.55 for all-cause dementia. These are observational findings with limitations, but the signal is strong and consistent. Whether this extends to people without T2D is not yet established by clinical trial data.
Are there other GLP-1 trials in Alzheimer's currently underway?
Yes. The research landscape has not closed. The ELAD structural findings in particular have generated interest in follow-up work with modified protocols. Researchers are examining whether longer treatment durations, earlier intervention, higher doses, or injectable formulations might produce different clinical outcomes than EVOKE achieved with oral semaglutide.
References
[1] Cummings et al., "evoke and evoke+: design of two large-scale, double-blind, placebo-controlled, phase 3 studies evaluating efficacy, safety, and tolerability of semaglutide in early-stage symptomatic Alzheimer's disease," Alzheimer's Research & Therapy, 2025. DOI: 10.1186/s13195-024-01666-7
[2] Scheltens et al., "Baseline characteristics from evoke and evoke+," Alzheimer's & Dementia: TRCI, 2026. DOI: 10.1002/trc2.70200
[3] Edison et al., "Liraglutide in mild to moderate Alzheimer's disease: a phase 2b clinical trial," Nature Medicine, 2026. DOI: 10.1038/s41591-025-04106-7
[4] Edison et al., "Evaluating the effects of the novel GLP-1 analogue liraglutide in Alzheimer's disease: study protocol (ELAD)," Trials, 2019. DOI: 10.1186/s13063-019-3259-x
[5] Seminer et al., "Cardioprotective Glucose-Lowering Agents and Dementia Risk: A Systematic Review and Meta-Analysis," JAMA Neurology, 2025. DOI: 10.1001/jamaneurol.2025.0360
[6] Wang et al., "Associations of semaglutide with first-time Alzheimer's diagnosis in T2D patients: Target trial emulation," Alzheimer's & Dementia, 2024. DOI: 10.1002/alz.14313
[7] Siddeeque et al., "Neuroprotective effects of GLP-1 receptor agonists in neurodegenerative disorders: A large-scale propensity-matched cohort study," International Immunopharmacology, 2024. DOI: 10.1016/j.intimp.2024.113537
[8] Schechter et al., "Neurodegeneration onset with GLP-1 receptor agonists in T2D: a real-world multinational cohort," Cardiovascular Diabetology, 2025. DOI: 10.1186/s12933-025-02962-8
[9] Tang et al., "GLP-1RA and SGLT2i Medications for T2D and Alzheimer Disease and Related Dementias," JAMA Neurology, 2025. DOI: 10.1001/jamaneurol.2025.0353
[10] Liang et al., "Clinical Evidence for GLP-1 Receptor Agonists in Alzheimer's Disease: A Systematic Review," Journal of Alzheimer's Disease Reports, 2024. DOI: 10.3233/ADR-230181
[11] Sabbagh et al., "Safety considerations of semaglutide in the potential treatment of Alzheimer's disease: A pooled analysis of semaglutide in adults aged 65 years," Alzheimer's & Dementia TRCI, 2025. DOI: 10.1002/trc2.70076
This content is for informational purposes only and is not intended as medical advice, diagnosis, or treatment. Always consult a qualified healthcare provider before starting any supplement or making changes to your health regimen.
