Alzheimer’s Paradox: Why One-Third of Patients Never Lose Their Minds

Something peculiar happens in roughly 30 percent of people who develop Alzheimer’s disease. The plaques accumulate, the tangles form, the characteristic wreckage of neurodegeneration spreads across the hippocampus, and yet their minds stay sharp. They remember faces. They follow conversations. They die, eventually, with full Alzheimer’s pathology in their heads and no dementia on their records.

Evgenia Salta has been thinking about these people for a long time. A neuroscientist at the Netherlands Institute for Neuroscience in Amsterdam, she calls them “resilient,” and she finds them, frankly, baffling. “We really don’t know why,” she says. “That’s a big mystery, and a very important one.” If she can figure out what their brains are doing differently, the implications for Alzheimer’s treatment could be considerable.

Her latest answer involves a cell type that most neuroscientists had written off as essentially irrelevant in adults. Immature neurons (cells that look and behave like young, not-yet-finished neurons) were long thought to vanish from the human brain sometime in childhood, their brief window of utility closing well before middle age. But Salta’s team, working with donated brain tissue from the Netherlands Brain Bank, kept finding them in people well into their 80s. In every group they examined: healthy controls, Alzheimer’s patients, and the resilient individuals who had the disease without the decline.

What the cells were doing, though, turned out to be the real story.

A Garden Under Siege

The team focused on the dentate gyrus, a curved structure tucked inside the hippocampus. It’s one of the few regions in the mammalian brain where new neurons might form in adulthood, a process known as adult neurogenesis, well documented in rodents but fiercely debated in humans for decades. To hunt for immature neurons there, the researchers used single-nucleus RNA sequencing, essentially reading the molecular identity of tens of thousands of individual cells to find the rare ones still carrying the transcriptional fingerprint of youth.

They found them. But the surprise wasn’t the cells themselves; it was what those cells were doing differently in resilient versus Alzheimer-affected brains. The team had expected resilient individuals to simply have more of these immature neurons. More cells, presumably, means more repair capacity. The difference, when it appeared, was subtler than that and, in some ways, more interesting. In people who had Alzheimer’s pathology but no dementia, the immature neurons seemed to activate molecular programs associated with survival, anti-inflammatory signaling, and support for surrounding tissue. In severe Alzheimer’s patients, those same programs were quieter. The cells were still there; they just weren’t behaving the same way.

Fertilizer, Not Replacement Parts

The distinction matters enormously for how scientists have been framing the question. The dominant hypothesis about adult neurogenesis and Alzheimer’s has always been a replacement story: new neurons grow in to compensate for dying ones, and resilient brains are better at this compensatory renewal. Salta thinks the truth might be more complicated. “It might not be (only) about replacing lost neurons,” she says. Instead, she suspects these cells are doing something more ecological, acting, as she puts it, as “a sort of fertilizer in a garden that has started falling apart.” They may be maintaining the homeostasis of the local environment, keeping inflammation at bay, supporting the cells around them, rather than simply substituting for the ones that are gone.

The molecular data support this reading, at least provisionally. In the resilient group’s immature neurons, the team found higher expression of genes linked to anti-inflammatory signaling, neurotrophic support (the chemical scaffolding neurons use to stay healthy), and DNA repair. Two proteins in particular stood out: clusterin, coded by a gene that genome-wide studies have already implicated in Alzheimer’s risk, and prosaposin, a lysosomal protein that, when knocked out in laboratory neurons, triggers a cascade of oxidative stress and cell death that doesn’t occur in other cell types. Both were significantly elevated in resilient individuals compared with severe Alzheimer’s patients. “We also see lower signals related to inflammation and cell death,” Salta says.

What the Cells Cannot Tell Us

Salta is careful, perhaps unusually so for a scientist discussing a finding with obvious clinical appeal. She acknowledges that reading a cell’s function from its gene expression alone is an interpretive act, not a direct observation: the team can see what programs are active, but cannot watch the cells doing anything in a living brain. She also notes that this is cross-sectional work, snapshots of postmortem tissue, not longitudinal tracking of people as their cognitive fate unfolds. “This is one piece of a very large puzzle,” she says flatly. “There will never be just one factor that explains resilience.”

The question of where these immature neurons come from adds another layer of uncertainty the paper doesn’t resolve. Some may be genuine adult-born cells, newly generated throughout life; others might be developmentally arrested neurons, born in fetal life and simply maturing extraordinarily slowly, as the human brain is known to do (a phenomenon called neoteny). The transcriptional evidence hints that at least a subset could arise later in life, but direct proof would require lineage-tracing experiments that postmortem brain tissue simply cannot provide.

One finding did catch the team genuinely off-guard: a previously unrecognized subpopulation of immature neurons, marked by high expression of a gene called OTOF, that appears to be specific to the adult human hippocampus and absent in macaques and mice. It’s not clear yet what these cells do differently, or why this particular molecular signature is ours alone, but it adds a further wrinkle to the already complicated question of whether animal models of adult neurogenesis really tell us much about what’s happening in human brains.

A Decision Point in the Dark

Perhaps what’s most striking about Salta’s framing is where she locates the fundamental mystery. Not in the plaques, not in the tangles, but in a moment of divergence the field hasn’t yet found a way to see. “Somewhere along this trajectory, there’s a kind of decision point,” she says. “Some people remain stable, others develop dementia. We want to understand what drives that difference.”

The cell-communication analysis in the new paper offers a hint about what that decision might involve. In the resilient group, immature neurons appeared to be embedded in a richer web of molecular signaling with surrounding cells (oligodendrocytes, excitatory neurons, microglia), including pathways linked to anti-amyloid processing and tau regulation. In Alzheimer’s patients, that crosstalk was depleted, in some cell types almost entirely absent. The implication, still tentative, is that cognitive resilience might be less about the absolute number of neurons you have, or even their molecular state, and more about whether they remain in conversation with the rest of the tissue around them.

Whether clusterin and prosaposin can be pharmacologically targeted, or whether the signaling environment of resilient dentate gyrus can be chemically induced, are questions nowhere near answering. But after decades in which Alzheimer’s research focused overwhelmingly on clearing pathology rather than on why some brains tolerate it, the pivot toward resilience represents something genuinely different. The aging brain, it keeps turning out, is harder to write off than we thought.

DOI: 10.1016/j.stem.2025.04.001 | Cell Stem Cell, April 2026

Frequently Asked Questions

Why do some people never get dementia even though they have Alzheimer’s in their brain?

Researchers don’t fully know yet, but a new study points to the behavior of a rare class of cells called immature neurons in the hippocampus. In people who carry Alzheimer’s pathology but remain mentally sharp, these cells appear to activate protective molecular programs tied to reducing inflammation and supporting surrounding tissue. Whether that cellular behavior is a cause of resilience or a marker of something deeper is still being worked out.

Do adults actually grow new brain cells?

The debate has run for decades and isn’t settled. What the Netherlands study found is that certain neurons in the adult hippocampus retain a molecular profile more typical of young, still-developing cells, even in people in their 80s. Whether those cells were genuinely born in adulthood or were generated earlier and simply matured very slowly remains an open question, one that postmortem tissue alone can’t answer.

Could boosting these immature neurons actually prevent Alzheimer’s dementia?

It’s a long way from the current evidence to any kind of treatment. The study identified two proteins, clusterin and prosaposin, that were notably elevated in resilient individuals, both of which have roles in neuronal survival and anti-inflammatory signaling. Those proteins are now candidates for further investigation, but whether they can be safely and effectively manipulated in living patients is a different question entirely.

Is there something special about the human brain that mice studies have been missing?

Possibly, yes. The new study identified a previously unknown subpopulation of immature neurons, marked by a gene called OTOF, that appears in the adult human hippocampus but not in macaques or mice. If human immature neurons have unique functions that animal models don’t replicate, that could explain why so many promising Alzheimer’s therapies that work in mice have failed in clinical trials.

What would it actually mean to target cognitive resilience instead of Alzheimer’s pathology?

Most current Alzheimer’s research aims to reduce amyloid plaques or tau tangles, the physical hallmarks of the disease. A resilience-focused approach would instead ask how to keep the brain functional in the presence of those hallmarks, essentially raising the threshold at which pathology translates into decline. Whether that’s achievable pharmacologically, through lifestyle, or some combination, remains an open and increasingly active area of research.