The results are in: What 2 major new studies show about semaglutide and Alzheimer’s

For a brief moment, it looked like one of medicine’s biggest success stories might cross into an entirely new frontier.

Semaglutide—the GLP-1 agonist powering Ozempic, Wegovy, and Rybelsus—had shown just enough suggestive biology to raise a tantalizing question: Could the blockbuster weight-loss drug also slow early Alzheimer’s?

Findings from two major trials delivered some disheartening news.

The trials that shut the door

The evoke and evoke+ trials followed nearly 3,800 adults ages 55–85 with mild cognitive impairment or early Alzheimer’s over two years. Participants were randomized to daily oral semaglutide (Rybelsus) or placebo. ^[]

The results were pretty straightforward, and not what many of us were hoping for:

• No improvement on the Clinical Dementia Rating Sum of Boxes, the gold-standard measure of cognitive and functional decline.

• No meaningful difference in memory, behavior, or daily functioning.

• No slowing of overall cognitive trajectories—despite small shifts in Alzheimer’s-related biomarkers.

Why scientists were hopeful

GLP-1 agonists have a surprisingly rich neurobiology story behind them. In preclinical and observational data, semaglutide and related drugs appeared to ^[]:

• Reduce neuroinflammation

• Improve insulin signaling in the brain

• Support mitochondrial function

• Reduce amyloid and tau accumulation

• Slow decline among people with diabetes in real-world datasets

In mice, this looked promising. In cell models, even better. In humans with Alzheimer’s? Not so much. It’s another case of the translational cliff between bench science and the brain’s messy reality.

Why this might have failed

The biology still makes sense—but the timing may not.

1. Treatment may have started too late. Neuroprotective strategies typically work best before cognitive symptoms arise. Once amyloid and tau are entrenched, targeting inflammation or metabolism alone is unlikely to move the needle. ^[]

2. Alzheimer’s is multifactorial. Even “multi-mechanism” drugs like semaglutide may not be enough to slow a disease that spirals through multiple pathways simultaneously. ^[]

3. Biomarkers aren’t outcomes. Shifts in CSF proteins didn’t translate into clinically meaningful change—at least not over two years. ^[]

4. Trial duration may have been too short. Researchers will debate whether longer follow-up could have revealed a signal. For now, the absence of early effect makes that unlikely. ^[]

Your takeaway

Even if you’re not managing Alzheimer’s directly, this story will affect conversations in primary care, endocrinology, geriatrics, and obesity medicine.

1. Patients and families are already asking. The idea of a weight-loss drug that might also “protect the brain” has made its way into exam rooms. These results give clinicians a clear, evidence-based answer.

2. It tempers expectations around GLP-1s as cure-alls. With GLP-1 agonists rapidly expanding into cardiology, nephrology, addiction research, and more, this is an important reminder: Great mechanisms don’t guarantee clinical outcomes.

3. It reshapes off-label enthusiasm. Some clinicians had wondered whether semaglutide could be justified for patients with MCI or strong family histories. These data argue firmly against it.

4. It will influence future drug pipelines. Pharma’s pivot into neurodegeneration using metabolic drugs may slow or change course, affecting what comes to market in the next decade.