For decades, oligodendrocytes had a simple job description. They wrap nerve fibers in myelin, the fatty sheath that lets electrical signals travel quickly, and they were considered unambiguously helpful. A study in Nature Medicine argues they can also turn against the brain they support.
Researchers at the University of Edinburgh and the UK Dementia Research Institute found that people whose thinking declined fastest in old age had a shift toward smaller nerve fibers, abnormally thick myelin on the largest ones, more oligodendrocytes, and reduced levels of a protective protein inside those cells. When the team removed that protein from oligodendrocytes in mice, the animals developed the same white matter changes and learned less well over time.
"This study has fundamentally shifted how we think about these brain cells in aging," said Georgina Craig, the study's first author and a postdoctoral fellow at St Michael's Hospital in Toronto and the UK Dementia Research Institute. "We have always considered oligodendrocytes as purely beneficial, yet here we surprisingly find that they can become dysfunctional and contribute to cognitive impairment in aging."
A Cohort Tested From Childhood Into Its Ninth Decade
The finding rests on an unusually deep dataset. The Lothian Birth Cohort 1936 is a group of Scots whose cognitive abilities have been assessed from childhood into older age, giving researchers individual trajectories rather than single snapshots.
Cognitive performance was measured from age 70 to 82 using tests of memory, processing speed and spatial skills. Of 1,091 people in the cohort, 866 had follow-up testing after age 70, according to the University of Edinburgh announcement. Almost everyone declined to some degree, which allowed the team to compare the brains of people whose decline was faster or slower than average.
That last point matters more than it sounds. When researchers examined donated post-mortem brain tissue from a subset of participants, the white matter changes tracked the rate of decline rather than how well someone scored at any single moment. Two people with identical test scores of 80 could have very different tissue, depending on how far each had fallen to reach that score.
The specific pattern was counterintuitive. Worse trajectories went with smaller myelinated axons, thicker myelin on the large-diameter fibers, and more oligodendrocytes. More insulation and more insulating cells, in other words, went with worse outcomes. Degenerating fibers, myelin sheaths with no axon left inside them, also became more common as decline steepened.
Why that might matter is a question of physics. The authors note that roughly 85 percent of the variance in human nerve conduction speed is attributed to two factors: myelin thickness and axon size. Those are precisely the two measurements that shifted.
One Protein Connected the Human Tissue to the Mouse Experiments
The molecular thread was NRF2, a protein that regulates hundreds of genes protecting cells from damage. People with more severe cognitive decline had a smaller proportion of oligodendrocytes expressing it.
To test whether that was cause or consequence, the team engineered mice to lose NRF2 specifically in oligodendrocytes. In a water maze, control animals improved their performance by an average of about 58 percent over four days of testing. The knockouts improved by roughly 28 percent. They were not incapable, in other words. They simply gained less from experience. In late aging, their myelinated axons showed the same signature observed in the human brain: smaller fibers and thicker sheaths.
That combination, a human association plus an animal model that reproduces it, is what lifts this above correlation. It remains one pathway rather than the full complexity of human cognitive aging, and female knockout mice were excluded from the cognitive analysis after showing genotype-related differences in anxiety behavior that could have confounded the results.
The broader shift is that researchers have spent years focused on neurons and, more recently, on the brain's immune cells. Work in Nature Neuroscience documented a drop in oligodendrocyte density in the white matter of aging mice, and a plasma proteomics study found that about half of the blood proteins tracking brain age belonged to the oligodendrocyte lineage. Earlier work mapping organ aging signatures in blood pointed the same way. This study puts those cells at the center rather than the periphery.
Why an Existing Multiple Sclerosis Drug Enters the Conversation
The NRF2 pathway is not obscure. It is already targeted by approved drugs, including dimethyl fumarate, used in multiple sclerosis, and the authors note that activating NRF2 with that drug has improved cognitive function in rodents and in people with MS. That raises a repurposing question they are openly interested in.
Nobody is prescribing anything for age-related cognitive decline on the strength of this paper. There are no approved treatments for the condition, the authors frame repurposing as a possibility rather than a plan, and the leap from mouse white matter to a pill that preserves human memory is long. Anyone taking or considering a medication should discuss it with their own clinician rather than acting on preclinical findings.
Two limits deserve mention. The cohort is Scottish and of majority white European ancestry, so how far the pattern generalizes is untested. And post-mortem tissue captures an endpoint, not a process, which is why the mouse experiments were needed.
Key Questions Answered
What are oligodendrocytes?
Brain cells that produce myelin, the insulating sheath around nerve fibers that lets signals travel efficiently. They have long been considered purely supportive of brain function.
What did the researchers actually find?
People with faster cognitive decline had smaller myelinated nerve fibers, abnormally thick myelin on the largest ones, more oligodendrocytes, and lower levels of the protective protein NRF2 inside those cells.
Why is more myelin a bad sign here?
The thickening was observed in large-diameter axons in people who were declining faster. The authors interpret it as dysfunctional rather than protective insulation, not as evidence of a healthier brain.
Does this prove oligodendrocytes cause cognitive decline?
Not in humans. The human data show an association. Mice lacking NRF2 in oligodendrocytes reproduced the tissue changes and learned less over time, supporting a causal role, though this does not establish it in people.
Could an existing drug help?
The NRF2 pathway is already targeted by approved drugs, including one for multiple sclerosis. That is a research lead, not a treatment recommendation, and no approved therapy exists for age-related cognitive decline.
What should people do with this information?
Nothing yet, clinically. Anyone concerned about memory changes should raise the issue with a clinician rather than self-medicating based on preclinical research.
