The root cause of Alzheimer’s disease remains contested, but I can tell you where the evidence points and where it frays.

The dominant framework for three decades has been the amyloid cascade hypothesis—that accumulation of amyloid-beta peptides triggers a cascade leading to tau pathology, synaptic loss, and neurodegeneration. This hypothesis generated enormous research investment and several targeted therapies. Yet the repeated failure of anti-amyloid interventions to meaningfully halt cognitive decline in clinical trials suggests either we’ve been targeting the wrong amyloid species, the wrong stage of disease, or that amyloid is better understood as a consequence of earlier upstream dysfunction rather than its origin.

My own bias, shaped by watching neurons fail in other contexts, is that the most parsimonious account points to a convergence of insults: synaptic vulnerability sits at the center. Specific neuronal populations—layer 2/3 excitatory neurons and inhibitory interneurons—are exceptionally metabolically demanding and highly interconnected. When these circuits are stressed by:

• Age-related decline in mitochondrial function

• Chronic neuroinflammation (microglial dysregulation)

• Vascular compromise reducing metabolic delivery

• Possibly persistent pathogens or viral reactivation

...the system reaches a tipping point where amyloid deposition and tau spread become self-reinforcing rather than initiating.

The dissatisfying honest answer is that “root cause” may be multiply determined—Alzheimer’s likely represents several pathophysiological routes converging on a final common phenotype of network collapse. The field is slowly moving toward recognizing this heterogeneity, which explains why monotherapy approaches have disappointed.

What we desperately need are biomarkers that capture the earliest cellular disruptions—before plaques are visible—to actually identify the trigger, not just the downstream debris.

My Perspective: Neuronal Vulnerability as a Circuit Problem

The question presupposes a single root cause, but from where I sit—building taxonomies of cell types and mapping brain-wide connectivity—the evidence increasingly points to Alzheimer’s as a selective vulnerability of specific neuronal populations propagating through circuits.

Here’s why this framing matters:

1. Not all neurons die equally. Cell type atlases reveal that certain excitatory pyramidal neurons in layer II/III of the entorhinal cortex—the first to degenerate—are molecularly distinct. They may share features (connectivity patterns, metabolic profiles, ion channel expression) that make them susceptible before amyloid or tau pathology becomes widespread.

2. The spreading follows connectivity. Tau pathology advances along anatomically connected pathways, not randomly. This suggests the disease exploits trans-synaptic mechanisms. My work on the mesoscale connectome showed that specific circuit architectures are more vulnerable to disruption.

3. Amyloid may be upstream but not sufficient. The amyloid hypothesis dominated for decades, but anti-amyloid therapies have shown modest effects. I suspect amyloid deposition triggers a cascade in predisposed cell types, and the root cause is actually that molecular vulnerability—what makes certain neurons unable to handle proteostatic stress, mitochondrial dysfunction, or inflammatory challenge.

4. Glia set the context. Microglia and astrocytes aren’t passive. My colleagues’ work on single-cell atlases shows microglial states shift dramatically in Alzheimer’s. The “root cause” may ultimately be a failure of glial-neuronal crosstalk that every brain possesses, but some cell type combinations handle poorly.

In short: the root cause isn’t amyloid per se—it’s why particular cell types in particular circuits lose resilience first. That’s a question that only systematic, brain-wide cell type mapping can answer.

The honest answer is: we don’t know with sufficient certainty, and I think the field has spent too long treating a leading hypothesis (amyloid) as settled when it clearly isn’t.

From a genomics perspective, what strikes me is the divergence between where therapeutic effort has gone versus where genetic evidence points. The rare familial mutations in APP and the presenilins gave the amyloid cascade hypothesis genuine plausibility. But GWAS in sporadic Alzheimer’s implicates pathways that aren’t obviously amyloid-centric—immune microglial function, lipid metabolism, endosomal trafficking, membrane dynamics. That’s not noise; that’s the genome telling us something.

My bias as someone who thinks mechanistically: the field has conflated correlation with causation for too long. Amyloid plaques are nearly ubiquitous in aged brains and don’t track tightly with cognitive impairment. Tau pathology tracks better but is downstream. We’ve been treating a correlation like a root cause.

The most useful framing I can offer: Alzheimer’s is probably not one disease. The familial early-onset forms have a clearer amyloid angle. The late-onset sporadic cases likely involve convergence of multiple vulnerability pathways—genetic risk, metabolic history, vascular health, immune competence—that may manifest as similar clinical endpoints but have different upstream drivers.

What we lack isn’t more descriptive data about plaques and tangles. We lack perturbation experiments that establish causation. That’s where genome editing, pooled screens in relevant cell types, and systematic functional validation could actually move the needle—testing whether modulating candidate genes actually modifies disease phenotypes, not just whether they’re associated.

The root cause question may be the wrong frame if the biology is genuinely multifactorial and stochastic.

Disclaimer: My lab studies synaptic plasticity, circuit dynamics, and neural activity during learning using two-photon microscopy—I am not an Alzheimer’s disease researcher. I can offer a basic science perspective, but this is not my domain of expertise.

The short honest answer is: we don’t know the root cause. What we have are several well-supported pathological signatures—amyloid-beta plaques, tau neurofibrillary tangles, neuroinflammation, synaptic loss—and a web of mechanistic hypotheses linking them. My own bias as someone who watches synapses form and disappear in real time is to pay close attention to synaptic vulnerability as a convergent endpoint. Many lines of evidence suggest that before clinical symptoms emerge, spines are lost, neurotransmission dampens, and distributed circuits become dysregulated. Whether that’s downstream of amyloid, tau, inflammation, or some combination is unresolved.

I’m skeptical of any single “root cause” framing. In complex neurodegenerative diseases, the causal architecture is likely to be network-level—multiple initial insults converging on a vulnerability that then propagates across circuits over years. The field has spent decades privileging amyloid; the limited success of anti-amyloid therapies suggests we may be treating a downstream manifestation rather than the trigger.

What I’d want, if I were working in this space: longitudinal, cell-type-specific imaging of the same circuits across disease progression—analogous to what we do in learning studies—to distinguish early from late events. Most human data is endpoint or cross-sectional. The mechanistic insights will come from watching the process unfold.

The Root Cause of Alzheimer’s: An Immunologist’s Perspective

From where I stand—studying immune dysregulation in human patients—the Alzheimer’s field has been asking the wrong question. Framing it as the root cause implies a single upstream event, but decades of failed monotherapies targeting amyloid or tau suggest we’re dealing with convergence points, not origins.

What’s becoming clearer from human data: neuroinflammation and microglial dysfunction are not downstream epiphenomena—they’re integral to disease initiation and progression. The brain’s immune cells, primed by aging, genetic risk (TREM2 variants, ApoE4), and likely peripheral inflammatory signals, adopt a self-perpetuating destructive state. This isn’t my primary disease focus, but the mechanistic overlap with chronic inflammatory conditions I study is striking.

I approach Alzheimer’s the way I approach autoimmunity: multiple upstream inputs (genetic susceptibility, environmental triggers, aging immune system) converge on final common pathways that destroy tissue. The amyloid hypothesis dominated for decades partly because it was testable in mice—yet those models lack the human immune complexity that matters.

My concern: interventions that ignore immune-mediated damage will continue to disappoint. We need longitudinal human cohorts with deep immune phenotyping, not more reductionist models. The root cause may ultimately be the convergence of dysregulated inflammation, protein aggregation, and neuronal vulnerability—interacting over years before symptoms emerge.