Some people reach their eighties with episodic memory that would be impressive in someone decades younger. Researchers call them SuperAgers, but the name does not mean they escape every ordinary effect of ageing. It refers to performance on a specific kind of memory test.

A 2026 study in Nature found an intriguing difference inside six donated SuperAger brains. In the hippocampus, a region central to forming and retrieving memories, these donors had roughly twice the abundance of immature neurons found in cognitively healthy older adults. Their cells also carried patterns of gene regulation that looked unusually resilient.

The result supports the idea that exceptional cognitive ageing can involve more than retaining neurons formed early in life. The brain may also preserve a pipeline that produces and matures new ones.

That interpretation needs two important limits. The researchers did not watch neurons being born in living people. They inferred developmental stages from molecular signatures in postmortem tissue. And although the twofold difference is a striking descriptive result, the comparison between six SuperAgers and eight healthy older adults was not statistically significant because the counts varied widely. The comparison with Alzheimer’s disease was significant.

“SuperAger” has a narrow scientific meaning

The participants were not selected because family members thought they seemed unusually sharp. They came from the Northwestern SuperAging Program, which applies a defined testing threshold.

SuperAgers had to be at least 80 years old and score at or above the average for people in their fifties on delayed recall from a standard word-list test. They also had to remain within one standard deviation of the expected range for their own age and education on other cognitive measures. Their daily functioning was preserved, with no clinical history of neurological or psychiatric disease.

The formal benchmark in the peer-reviewed Nature paper was episodic memory equal to or better than that of adults aged 50 to 59. Some descriptions refer more broadly to people in their fifties and sixties, reflecting the wider norms used by the programme. Neither version means an 80-year-old SuperAger performs like a younger adult on every aspect of attention, reasoning or processing speed.

It is an unusual phenotype defined around memory, not a diagnosis or a promise of protection from future decline.

The evidence came from 38 donated hippocampi

The team analysed tissue from five groups: eight cognitively intact young adults aged 20 to 40, eight older adults with healthy cognition for their age, six people with preclinical intermediate pathology, ten with Alzheimer’s disease and six SuperAgers. In total, researchers profiled 355,997 nuclei from the hippocampus.

They focused on the dentate gyrus, a part of the hippocampal memory system and the main candidate site for adult neurogenesis. Each tissue block was checked for the presence of this region before it was isolated for analysis.

One method, single-nucleus RNA sequencing, measured which genes were active in individual nuclei. A paired technique called ATAC sequencing measured chromatin accessibility, identifying stretches of DNA open to regulatory machinery. Those two views allowed researchers to separate neural stem cells, neuroblasts, immature granule neurons and mature neurons, then examine the regulatory networks associated with each stage.

The accompanying Northwestern University account describes the SuperAger hippocampus as a cellular environment that supports both the birth and survival of new neurons. That is a plausible synthesis of the data, but it is still an interpretation of tissue collected at one point after death.

Young neurons were identified by a molecular trajectory

“Young neuron” is convenient language, but the researchers were not reading a date stamped on each cell. They classified cells by molecular features associated with successive developmental stages.

The analysis traced a trajectory from neural stem cells to neuroblasts, then to immature granule neurons and finally mature granule neurons. RNA velocity, a computational method that compares newly made and processed RNA, helped establish the likely direction of this progression. Chromatin patterns and reference datasets provided additional checks.

This is strong evidence consistent with ongoing adult neurogenesis. It is not equivalent to labelling a dividing cell in a living person and following its descendants over time, an experiment routinely done in animals but generally impossible in healthy human brains.

That methodological gap explains why the field has argued for decades. A 1998 Nature Medicine study used a DNA tracer given to cancer patients and reported newly generated hippocampal neurons after death. A 2018 Nature study found a sharp childhood decline and concluded that neurogenesis was absent or extremely rare in adults. Differences in tissue preservation, markers and cell annotation can change what remains detectable.

In 2019, another team reported thousands of immature neurons in neurologically healthy people into their ninth decade, with a decline in Alzheimer’s disease. The 2026 study adds single-cell gene activity and chromatin accessibility to that evidence rather than relying on one marker alone.

The twofold result did not clear every statistical test

At first, the SuperAger group showed a significant increase in immature neurons compared with the other cohorts combined. The researchers then noticed that one donor had an especially high count and repeated the analysis without that sample.

The pattern remained. Excluding the outlier, the SuperAger group still showed a 2.5-fold increase in immature neurons. Compared separately with healthy older adults, young adults and people with preclinical pathology, the difference was about twofold.

But those separate comparisons were not statistically significant. The authors explicitly attribute the limitation to the small number of brains and high person-to-person variability in cell abundance. The SuperAger comparison with the Alzheimer’s group was statistically significant whether or not the outlier was included. SuperAgers also had significantly more neuroblasts than the Alzheimer’s group.

This makes “roughly twice as many” a fair description of the observed sample, not a precise estimate for all exceptional agers. With six SuperAgers, one unusual brain can strongly affect an average. Larger postmortem collections will be needed to determine the typical effect size and how often the pattern occurs.

A resilience signature extended beyond cell counts

The more persuasive part of the study may be the convergence of several kinds of evidence. SuperAger immature neurons contained 7,058 regions of more accessible chromatin, while their neuroblasts contained 674. A few key genes, including BDNF and CALB1, were more active. BDNF encodes a protein involved in neuronal survival and plasticity, though this study does not show that raising BDNF would create a SuperAger phenotype.

The team’s regulatory analysis also found that SuperAgers and young adults shared several of the same active networks in stem cells and immature neurons. In Alzheimer’s samples, many programmes involved in neuronal development, synaptic plasticity and communication were reduced. Preclinical samples showed earlier versions of some of those changes.

An NIH summary of the work describes this as a distinct molecular signature associated with cognitive resilience. The study also implicated astrocytes, support cells that help regulate the neuronal environment, and CA1 neurons, another hippocampal population central to memory. Preserved excitatory signalling between these cells distinguished healthy ageing from pathological decline.

In other words, the finding is not simply that SuperAgers had a larger pile of immature cells. Their hippocampi appeared to preserve a wider system supporting development, communication and survival.

The study reveals an association, not a longevity recipe

The study cannot establish which direction the relationship runs. Continued neurogenesis might help preserve episodic memory. A broader protective biology might independently preserve both memory and the neurogenic environment. Genetics, lifelong behaviour, medical history and unmeasured exposures could contribute to both.

Because tissue was examined after death, the researchers could not measure each person’s neurogenesis rate over time. Cell abundance reflects a balance among production, maturation, integration and death. Having more immature neurons could mean more were being generated, more were surviving in an immature state, or some combination of the two.

Nor does the paper show that exercise, puzzles, supplements or a drug can double neuron production in an older human hippocampus. Animal experiments can test such interventions directly; this human study cannot. It identifies molecular networks that might become therapeutic targets, but target discovery is many steps removed from a safe treatment.

A previous ScienceBlog report introduced the study’s central comparison. The closer reading is both more cautious and more interesting: six exceptional brains preserved a developmental signature that ordinary ageing often diminishes, while Alzheimer’s disease disrupted it much more sharply.

This is one postmortem study, not settled consensus. It strengthens the evidence that adult human hippocampi can continue making neurons and connects that capacity to exceptional memory, but it does not prove that neurogenesis is the cause of SuperAging. This is not treatment advice.