The question this newspaper reported last summer, whether the adult human hippocampus still makes neurons, now has a disease chapter. In Nature on 25 February 2026, Ahmed Disouky, Mark Sanborn, K. R. Sabitha and colleagues in Orly Lazarov's laboratory at the University of Illinois Chicago, working with Northwestern University, the University of Washington and the Rush Alzheimer's Disease Center, reported profiles of 355,997 cell nuclei taken from post-mortem hippocampi and analysed with paired single-nucleus RNA sequencing and single-nucleus ATAC sequencing, which reads which stretches of DNA are open for use. The team identified neural stem cells, neuroblasts and immature granule neurons in five groups: young adults with intact memory, older adults with no cognitive impairment, so-called SuperAgers over 80 whose episodic memory matches or exceeds that of people decades younger, adults with preclinical intermediate pathology, and adults with Alzheimer's disease. Reports from the researchers and their university put the total at 38 donated brains.

The pattern the authors describe is a pipeline that jams near the end. In the Alzheimer's group there were fewer neuroblasts and fewer immature neurons than in young adults or healthy older adults, and fewer immature neurons than in the preclinical group, while stem cell numbers were not depleted; differences already appeared in the preclinical brains, before dementia. Differences in chromatin accessibility, particularly in the stem cells, separated the groups more sharply than gene expression did. The SuperAgers ran the other way. "Superagers had twice the neurogenesis of the other healthy older adults," Lazarov said in her university's account of the work, and the paper attributes that profile largely to thousands of differentially accessible regions of DNA in immature neurons and neuroblasts, with brain-derived neurotrophic factor expression among the features the authors call a resilience signature.

The caveats sit inside the design. Every brain was studied after death, once, so the work compares groups rather than following anyone over time, and it cannot show whether more new neurons produced the preserved memory, whether preserved brains simply permit more new neurons, or whether some third factor drives both. The groups are small, and the SuperAger tissue came from Northwestern's SuperAging Program while the other samples came from the University of Washington, meaning group differences and source differences travel together. Commentators on Alzforum also noted that these SuperAgers had little plaque and tangle pathology, which makes them a picture of resistance to disease rather than of resilience in its presence, and Hongjun Song of the University of Pennsylvania said he was surprised by the abundance of stem cells reported in Alzheimer's tissue.

The wider dispute is not settled by one paper. Sorrells and colleagues argued in Nature in 2018 that new neuron production drops to undetectable levels in adult humans, and a review in Cell Stem Cell on 5 March 2026 by Doludda and 16 co-authors, among them Jonas Frisen, whose Karolinska team reported dividing progenitors in Science on 3 July 2025, and Lazarov herself, frames the existence question as increasingly answered by single-cell methods while the functional question stays open. That review counts 38 registered clinical trials using the keyword adult neurogenesis, a reminder that products and programmes promising to boost new neurons will reach families well before any trial shows a cognitive benefit.