Adult Hippocampal Neurogenesis: What Is Actually Settled Science

By Anas, Editor at HealthForge · Published 23 August 2026 · 15 min read

Two landmark studies, published in the same journal within weeks of each other in 2018, reached opposite conclusions. One reported that new neurons all but vanish from the human hippocampus after childhood. The other found thousands of them in people well into their seventies. Neither team was careless, and both worked with human brain tissue. That contradiction, more than any single finding, is what makes adult hippocampal neurogenesis one of neuroscience’s genuinely unsettled questions.

The confusion is worth untangling, because the phrase “settled science” gets used loosely here. Parts of this field are firmly established. Others are probable but contested. And a few of the questions that matter most to non-specialists — how much happens in an aging human brain, and whether it shapes memory or mood — remain open.

This piece maps that terrain: what we know, what we reasonably think, and what the evidence still cannot tell us.

In this article

1. The short version
2. What adult hippocampal neurogenesis actually is
3. What is genuinely settled
4. What remains debated in humans
5. Why the studies disagree
6. The evidence hierarchy: what each method can and cannot prove
7. What the field now broadly agrees on
8. What this means for memory, mood, and aging
9. What a thoughtful reader can take from this
10. What the evidence does not establish
11. Common questions
12. Where this leaves us
13. Related reading

What adult hippocampal neurogenesis actually is

Adult hippocampal neurogenesis is the generation of new excitatory granule cells in the dentate gyrus, a small region of the hippocampus. These cells arise from radial glia-like neural stem cells and pass through intermediate progenitors and neuroblasts before maturing into functional granule neurons¹.

A common misconception holds that these new neurons replace dead ones. They don’t, in any meaningful sense. Their contribution is better understood as added plasticity — a form of neuroplasticity in which new units reshape how the dentate gyrus encodes and separates information — rather than a swapping out of lost tissue².

This lineage sequence, from stem cell to mature granule neuron, is the settled backbone of the field. The disagreements begin when researchers try to count these cells in adult human brains.

What is genuinely settled

Three things are not seriously disputed.

First, the process exists and is robust in rodents. Decades of lineage-tracing and labeling work have mapped the dentate gyrus neurogenic niche in mice and rats in fine detail.

Second, adult humans generate at least some new hippocampal neurons. The first direct evidence came from a 1998 postmortem study that found dividing cells and new neurons in the adult human dentate gyrus using BrdU, a label incorporated into newly made DNA³. Fifteen years later, an independent approach reinforced it: carbon-14 birth dating, which exploits the spike in atmospheric radiocarbon from Cold War nuclear testing to timestamp when a cell’s DNA was synthesized. That 2013 Cell Stem Cell study concluded that new neurons are generated in the adult human hippocampus throughout life, with only a modest age-related decline⁴.

Third, immature-neuron markers are detectable across the human lifespan. A 2010 study in the Journal of Comparative Neurology found DCX-expressing cells in the human dentate gyrus from birth to age 100 — but with an exponential decline as people aged⁵.

Notice that even the “settled” evidence points toward decline. The fight is over how steep that decline is, and whether it reaches effectively zero.

What remains debated in humans

Here is the crux. In 2018, a Nature study reported that proliferating progenitors and young neurons drop sharply after birth and were essentially undetectable in the adult dentate gyrus⁷. Weeks later, a Cell Stem Cell study analyzing whole human hippocampi found intermediate progenitors and thousands of immature neurons across ages 14 to 79, with neurogenesis largely preserved in healthy aging — though with a smaller pool of quiescent progenitor cells in the anterior-mid dentate gyrus⁸.

A perspective in Science Translational Medicine captured the moment with a pointed title, asking whether the dogma had been overturned “again and again”⁶. By 2023, reviewers were describing the field as two competing schools of thought rather than one converging consensus²⁰.

The honest status, then, is this: that some neurogenesis persists into adulthood is well supported. Whether it continues at biologically meaningful rates into old age is not resolved.

Why the studies disagree

The disagreement is less about interpretation than about laboratory method — which is oddly reassuring, because methodological problems are tractable.

Tissue preservation

Human brain tissue is not collected under ideal conditions. There is a delay between death and preservation, and tissue is often fixed in formaldehyde for long periods. Both matter enormously. Two independent 2023 studies — one in Frontiers in Neuroanatomy, one in Communications Biology — showed that prolonged fixation and postmortem delay substantially reduce or abolish detection of DCX-positive immature neurons⁹¹⁰. The implication is sharp: a study reporting “no neurogenesis” may be reporting a failure to detect it, not its absence.

Detection also depends on tissue quality, fixation time, antigen retrieval, how autofluorescence is handled, and which markers are chosen¹¹¹². Change the protocol and you can change the answer.

The marker problem

Most human studies lean heavily on DCX as a proxy for young neurons, yet DCX is an imperfect stand-in. A 2024 study found that DCX transcript is broadly expressed beyond classic immature-neuron populations in the human hippocampus and cortex¹³. Work in mice showed that DCX is not even required for newborn neurons to survive or mature — it is a marker, not a driver¹⁴. And a 2023 Hippocampus study found that DCX levels can diverge from actual neurogenesis under some conditions¹⁵. Relying on DCX alone, in other words, can mislead in either direction.

The evidence hierarchy: what each method can and cannot prove

Not all methods carry equal weight, and each answers a slightly different question.

Carbon-14 birth dating is the closest thing to direct proof that new neurons are born in adults, because it timestamps DNA synthesis⁴. Its limitation is that it works at the population level and cannot pinpoint individual newborn cells.

Immunohistochemistry — staining tissue for markers like DCX — can localize candidate cells, but it is exquisitely sensitive to tissue handling and marker choice, which is exactly where studies diverge¹¹¹².

Single-cell and spatial transcriptomics are the newest tools. A 2024 spatial transcriptomics study of the human dentate gyrus detected very few cells expressing neural stem cell and proliferative markers from childhood to middle age, yet substantial DCX expression at all ages — a pattern the authors read as very low neurogenesis rates¹⁶. But these methods have an inherent ceiling: they profile cell states, and cannot by themselves prove that a new neuron was born without complementary validation. Current human datasets are also confounded¹⁷¹⁸.

A very recent 2026 Nature study, using updated cell-annotation methods, reported identifying adult hippocampal neural stem cells and immature neurons — strengthening the case for ongoing neurogenesis¹⁹. It is new and awaits independent replication, so it should be read as a promising data point rather than a verdict.

What the field now broadly agrees on

Strip away the headlines and a workable middle ground emerges.

Adult neurogenesis in the human hippocampus is real but appears far rarer than in rodents, and it declines with age. The rodent-to-human leap is precisely where caution is warranted; species differences make external validity a persistent problem across biology, not just here³². One mechanism likely driving the age-related drop is that neural stem cells become increasingly quiescent — dormant rather than dividing — which limits their activation, though this is established largely in rodents²¹.

Most researchers now accept that negative human studies cannot be taken at face value without ruling out technical failure, and that DCX should not serve as a lone proxy. What remains is a quantitative dispute: how much persists, and whether that amount is enough to matter functionally.

What this means for memory, mood, and aging

Memory and learning

The strongest functional evidence comes from animals. A 2017 meta-analysis found that most ablation studies — in which neurogenesis is experimentally shut down — support a role in behavioral pattern separation, the ability to tell similar experiences apart²². Related work argues that adult-born granule neurons support cognitive flexibility and the discrimination of similar inputs²³. In humans, this link is inferred, not demonstrated directly.

Depression

The neurogenesis theory of depression is more nuanced than popular accounts suggest. A critical review concluded that stress can reduce neurogenesis and antidepressants can increase it under some conditions — but the effects are not uniform across depression models²⁷. An influential earlier proposal reframed the relationship: adult neurogenesis may matter more for antidepressant response than for the cause of depression²⁸. That distinction is important, and frequently lost.

Alzheimer’s disease and injury

A 2019 Nature Medicine study reported that neurogenesis is abundant in neurologically healthy people and drops sharply in Alzheimer’s disease, with fewer and less-mature newborn neurons as the disease advances²⁴. Later reviews echo reduced neurogenesis in Alzheimer’s patients, associated with cognitive impairment — but describe the evidence as largely correlative, not causal²⁵²⁶. Traumatic brain injury also alters hippocampal neurogenesis, including acute loss of newborn neurons and progenitor hyperproliferation, though that evidence is largely preclinical²⁹.

A useful reality check on translation: even directly targeting Alzheimer’s biology has proven hard. Amyloid-targeting monoclonal antibodies produce, at most, small cognitive benefits that don’t exceed minimal clinically important thresholds, while raising the risk of brain imaging abnormalities³⁰³¹. That gap between mechanism and meaningful clinical benefit is a caution worth carrying into any neurogenesis-based therapeutic hope.

What a thoughtful reader can take from this

There is no validated way to measure your own neurogenesis, and no supplement or product has been shown to boost adult hippocampal neurogenesis in humans in a way that meaningfully changes cognition. Claims to the contrary run well ahead of the evidence.

What the science does support is modest and familiar. The behaviors long associated with hippocampal and cognitive health — aerobic exercise, sleep, managing chronic stress, staying mentally and socially engaged — remain reasonable on their own merits, whether or not neurogenesis is the mechanism. Treat “boosts neurogenesis” marketing as a red flag, not a selling point.

The more useful takeaway is interpretive. When you see a headline declaring adult neurogenesis “debunked” or “confirmed,” it is almost always reporting a single study using a single method. Learning how to read health research helps here: the truth lives in the pattern across methods, and that pattern says some, declining, and hard to measure — not none, and not abundant forever.

What the evidence does not establish

The human evidence cannot yet tell us how many new neurons a healthy older adult actually produces, or whether that number is functionally significant. It has not shown that changes in neurogenesis cause depression or Alzheimer’s disease, as opposed to accompanying them. And it cannot confirm that any lifestyle intervention increases human hippocampal neurogenesis, because we lack a way to measure it in living people.

The rich functional findings — pattern separation, cognitive flexibility — come overwhelmingly from rodents, and species differences make direct extrapolation unreliable³². The newest transcriptomic and 2026 findings are exciting but await independent replication, so they should shift confidence, not settle the debate.

Common questions

Does adult hippocampal neurogenesis really happen in humans?

Yes, at least to some degree. Independent methods — carbon-14 birth dating⁴, early BrdU labeling³, and marker studies⁵ — support the generation of new hippocampal neurons in adults. The disagreement is about how much occurs, especially in older age, not whether it happens at all.

Is the science settled or still controversial?

Both, depending on the question. That the process exists is well established; its magnitude in the aging human brain is genuinely contested, with credible studies on each side⁶²⁰.

Why do some studies find it and others don’t?

Largely for technical reasons. Prolonged tissue fixation and postmortem delay can erase the very markers used to detect young neurons, so a negative result may reflect method rather than biology⁹¹⁰. Marker choice, especially reliance on DCX, adds further inconsistency¹³¹⁵.

Does neurogenesis decline with age?

The evidence consistently points toward decline. Marker-based work shows an exponential drop across the lifespan⁵, and carbon-14 dating suggests a more modest decline⁴. One likely driver is increasing stem cell quiescence with age, established mainly in rodents²¹.

What should a non-specialist believe right now?

That adult humans make some new hippocampal neurons, that the amount falls with age and is hard to measure, and that confident claims in either direction — “debunked” or “abundant” — overstate what any single study can show.

Where this leaves us

Adult hippocampal neurogenesis is a case study in how science actually works: a real phenomenon, firmly established in animals, whose human details are being negotiated one method at a time. The settled core is that new neurons arise in the adult dentate gyrus and add plasticity rather than replacing lost cells. The unsettled core is how much of this persists in the aging human brain, and whether it shapes memory and mood in ways that matter clinically.

The most honest position is neither the triumphant “confirmed” nor the dismissive “debunked.” It is that the question is method-sensitive and biologically variable, and that better, standardized techniques — not louder headlines — are what will eventually resolve it.

Related reading

• BDNF — a growth factor closely tied to neurogenesis, memory, and mood.
• Meditation and neurogenesis — how contemplative practice intersects with hippocampal plasticity.

Sources

1. Frontiers in Neuroscience, 2015: Neurogenesis in the Adult Hippocampus
2. Annual Review of Neuroscience, 2014: Functions and dysfunctions of adult hippocampal neurogenesis
3. Nature Medicine, 1998: Neurogenesis in the adult human hippocampus
4. Cell Stem Cell, 2013: Dynamics of hippocampal neurogenesis in adult humans
5. Journal of Comparative Neurology, 2010: Murine features of neurogenesis in the human hippocampus across the lifespan from 0 to 100 years
6. Science Translational Medicine, 2018: Adult neurogenesis in humans: Dogma overturned, again and again?
7. Nature, 2018: Human hippocampal neurogenesis drops sharply in children to undetectable levels in adults
8. Cell Stem Cell, 2018: Human Hippocampal Neurogenesis Persists throughout Aging
9. Frontiers in Neuroanatomy, 2023: Prolonged fixation and post-mortem delay impede the study of adult hippocampal neurogenesis in human tissue
10. Communications Biology, 2023: Prolonged fixation and post-mortem delay impede the study of adult hippocampal neurogenesis in the human hippocampus
11. Frontiers in Neuroscience, 2022: Methods to study adult hippocampal neurogenesis in humans
12. Trends in Neurosciences, 2018: Adult Human Hippocampal Neurogenesis: Controversy and Evidence
13. International Journal of Molecular Sciences, 2024: Reconsidering Neurogenetic Indication in the Human Brain: Broad Expression of Doublecortin Transcript
14. PLoS One, 2016: Doublecortin (DCX) is not Essential for Survival and Differentiation of Newborn Neurons in the Adult Mouse Dentate Gyrus
15. Hippocampus, 2023: A complex relation between levels of adult hippocampal neurogenesis and expression of the immature neuron marker doublecortin
16. Cell Reports, 2024: Spatial transcriptomic analysis of adult hippocampal neurogenesis in the human dentate gyrus across lifespan
17. Frontiers in Neuroscience, 2023: Mapping human adult hippocampal neurogenesis with single-cell transcriptomics
18. Frontiers in Cell and Developmental Biology, 2022: Single-Cell and Single-Nucleus RNAseq Analysis of Adult Hippocampal Neurogenesis
19. Nature, 2026: Human hippocampal neurogenesis in adulthood, ageing
20. Trends in Cognitive Sciences, 2023: Adult human neurogenesis: A view from two schools of thought
21. Cell Stem Cell, 2020: Mechanisms of enhanced quiescence in neural stem cell aging
22. Hippocampus, 2017: Hippocampal neurogenesis and pattern separation: A meta-analysis
23. Nature Reviews Neuroscience, 2017: Adult hippocampal neurogenesis and cognitive flexibility
24. Nature Medicine, 2019: Adult hippocampal neurogenesis is abundant in neurologically healthy subjects and drops sharply in patients with Alzheimer’s disease
25. Brain Research Bulletin, 2024: The relationship between adult hippocampal neurogenesis and cognitive impairment
26. Molecular Neurobiology, 2021: Adult hippocampal neurogenesis in Alzheimer’s disease
27. Neuropsychopharmacology, 2011: Depression, antidepressants, and neurogenesis: a critical reappraisal
28. Nature Neuroscience, 2007: Adult hippocampal neurogenesis in depression
29. Cell and Tissue Research, 2020: Traumatic brain injury and hippocampal neurogenesis
30. Cochrane Database of Systematic Reviews, 2026: Amyloid-beta-targeting monoclonal antibodies for people with mild cognitive impairment or dementia due to Alzheimer’s disease
31. JAMA Internal Medicine, 2024: Clinically Important Benefits and Harms of Monoclonal Antibodies Targeting Amyloid for the Treatment of Alzheimer Disease
32. PLOS Biology, 2018: Is it possible to overcome issues of external validity in preclinical animal research?