GLP-1s for Alzheimer’s. Here’s What the Data Say
GLP-1 drugs may do more than drive weight loss. Emerging research suggests real potential for protecting the brain and slowing Alzheimer’s pathology. The data are early, but optimistic
As many of you know by now, I’m a neuroscientist who is himself at high risk for Alzheimer’s disease.
This is due to the fact that I carry two copies of the Alzheimer’s risk variant ApoE4, which is thought to increase my risk of developing Alzheimer’s disease by roughly 10- to 15-fold.
As a result, I leave no stone unturned when evaluating interventions that might help prevent it.
And one of the most interesting, provocative, and promising possibilities to cross my radar recently has been GLP-1 drugs.
Yes, the same blockbuster weight-loss medications you hear about in the news practically every day.
But the truth is, these aren’t just weight-loss drugs.
Weight is the metric our society fixates on, so it takes center stage.
But these compounds are potent biomodulators that do far more than reduce body weight. In fact, you could argue that weight loss is simply one downstream effect of their broader impact on human metabolism.
We’ve touched on this in prior letters, but today I want to zoom in on one specific domain: the potential role of GLP-1s in preventing Alzheimer’s disease.
To do that, we’ll draw from a new systematic review of the literature focused on this exact question.
Quick Roadmap
Biological background you need to understand Alzheimer’s pathogenesis.
Results from the new systematic review
How GLP-1s interact with those processes, independent of weight loss.
Other high-yield accessible options that I use to protect my brain.
Biological Background: Alzheimer’s Pathologies and Enzymes
Let’s get into some biological background. I’ll do my best to keep this from getting too dry, but stick with me, because this part matters.
I want to focus on two key pathological hallmarks of Alzheimer’s disease—Aβ and neurofibrillary tau tangles—and two key enzymes that drive them: BACE and GSK3β.
Two Pathological Hallmarks: Aβ and neurofibrillary tau tangles
When I say, “pathological hallmarks,” I mean this is what you tend to see when you look at the brains of patients with Alzheimer’s.
Aβ starts as toxic protein fragments and later develops into larger plaques— sticky aggregates outside of neurons.
Tau tangles, by contrast, form inside neurons. Tau itself is normally a functional protein. It helps stabilize microtubules, which you can think of as part of the cell’s internal skeleton. But when tau becomes excessively tagged with phosphate groups, it detaches from that structure and begins to clump together, forming tangles that essentially choke the neuron from the inside out.
A simplified way to think about this is that Aβ fragments, which come from a larger precursor protein, may facilitate or accelerate the formation and spread of these tau tangles.
*Nuance note: the plaques themselves may be less directly toxic than smaller Aβ fragments, and may serve more as a marker that these fragments are present and causing damage. That’s part of the reason why therapies that simply remove plaques haven’t been as effective as we might have hoped.
But let’s be clear, this isn’t a simple, one-directional pathway.
Biology rarely is. It’s more of a web. Aβ may contribute to tau tangles, but tau tangles can also feedback and promote more amyloid production. And both processes are intertwined with broader metabolic dysfunction, especially insulin resistance and inflammation.
I’m sorry if that’s confusing. But I don’t want to present to you a web and pretend it’s a single straight string.
Now, let’s talk about the two enzymes, and then tie the mechanistic picture together.
Two Enzymes: BACE and GSK3β
BACE is involved in processing a protein called the “amyloid precursor protein.”
This protein can be broken down in one of two ways.
One pathway is relatively benign, even protective.
The other generates Aβ, the fragments that go on to form plaques.
BACE initiates this second amyloidogenic (toxic) pathway, ultimately leading to the formation of Aβ.
So, in simple terms: BACE promotes Aβ formation.
Then we have GSK3β. This enzyme is also known as “tau kinase 1,” which is actually a more helpful name. Kinases are enzymes that add phosphate groups to proteins, and tau kinase 1 (GSK3β) does this to tau.
That phosphorylation is what causes tau to detach, misfold, and ultimately form those toxic tau neurofibrillary tangles.
So again, simplified:
BACE → Aβ
GSK3β → tau tangles
The New Systematic Review: Preclinical and Clinical Trials
Now let’s talk about this new systematic review.
The researchers searched three major scientific databases (PubMed, Embase, and the Cochrane Library) and ultimately identified 32 studies of interest. Most were preclinical, meaning conducted in cells, rodents, or non-human primates, although there were two clinical trials.
Across the preclinical studies, the results were rather consistent.
22/30 showed a decrease in Aβ
19/30 showed a decrease in tau tangles
The authors summarize it plainly:
“Collectively, these results point overwhelmingly to the view that administration of GLP-1RAs in preclinical models of Alzheimer’s disease lead to significant reductions in the pathophysiological hallmarks of Alzheimer’s disease .”
That’s a strong claim. And, in my view, generally supported.
That said, we can only go so far with preclinical studies.
What happens in humans?
Of the two clinical studies included, the results were unimpressive.
One was a 26-week randomized, double-blind, placebo-controlled trial in 38 patients with Alzheimer’s disease. The other was an 18-month randomized, double-blind, placebo-controlled trial in individuals with mild cognitive impairment or early dementia.
Overall, there wasn’t much of an effect on Aβ or tau, at least based on the measurement tools available. And to be clear, in living humans, we’re not staining brain tissue directly like we do in animal models. (Typically, clinical trial ethics boards don’t approve trials when the methodology section reads, “decapitation.”) Instead, we rely on advanced imaging techniques like PET scans, which are inherently less precise.
That said, there were small signals.
For example, one study showed decreases in Aβ within extracellular vesicles, which are essentially tiny packages released by neurons that circulate in the bloodstream. But overall, the results were modest.
Why The Gap?
Why do the preclinical studies look so promising, while the clinical trials seem underwhelming?
The answer is simple: timing.
Most human trials enroll patients who already have cognitive decline: at least mild cognitive impairment, if not full-blown Alzheimer’s disease. By that point, the pathology is well underway.
I think of it like a boulder rolling down a hill. Alzheimer’s disease begins decades before symptoms appear. By the time cognitive decline is noticeable, that boulder has already picked up momentum.
The most effective strategy isn’t to stop the boulder at the bottom, it’s to prevent it from rolling in the first place.
So, if you think of GLP-1 receptor agonists as putting up a protective fence, they may be highly effective when the boulder is still sitting quietly at the top of the hill. But a small fence at the base won’t do much once a two-ton boulder is already in motion. It’ll go straight through it.
And this highlights a real challenge in medicine.
We tend to think in terms of a hierarchy of evidence, with randomized, double-blind clinical trials at the top. That’s the “gold standard.” But in some cases, that standard is simply not feasible.
It’s not realistic to enroll thousands of healthy individuals, control all variables, and follow them for 20 or 30 years in a blinded trial.
So, put simply, I do think there’s real potential here for GLP-1s in neuroprotection.
And I don’t think we should expect that potential to show up cleanly in the clinical trials we currently have. Negative results in this context don’t necessarily mean absence of a biological effect, especially if the intervention needs to happen early.
So, if the clinical trials don’t show benefit… but the preclinical data strongly suggest it… what exactly is going on?
And how can we leverage new knowledge and new tools for effective, long-term strategies for Alzheimer’s prevention?
In the rest of this letter for StayCurious Metabolism Premium members, we resolve the intellectual tension and dive deep into:
Promising signals in the human data
How GLP-1s might protect the brain (4 mechanisms)
Minimal effective dosing
High-yield accessible tools for brain health, including:
My favorite form of omega-3
My favorite element for neuroprotection
Boosting brain energy with NAD
The cholesterol drug that people misunderstand
Come find out why StayCurious Metabolism is a Top Overall Bestseller in Science, and why members rave it’s the best investment they’ve made in their health.
