Myelin-making cells implicated in age-related cognitive decline

Researchers have linked dysfunction in oligodendrocytes—the cells responsible for producing and maintaining the brain’s myelin insulation—to cognitive decline during aging. The study, published August 25 in *Nature Medicine*, combined decades of cognitive measurements, postmortem human brain tissue and a genetically engineered mouse model.

The human evidence associated faster decline with thicker, structurally abnormal myelin around large nerve fibers, fewer large myelinated axons and reduced activity of the protective protein NRF2 in oligodendrocytes. Selectively removing NRF2 from mouse oligodendrocytes produced similar white-matter abnormalities and blunted improvement during a learning task.

Together, the experiments suggest that myelin-producing cells can become dysfunctional rather than uniformly supportive in later life. They do not demonstrate that activating NRF2 will prevent dementia or cognitive decline in people.

A cohort followed from childhood

The researchers drew on the Lothian Birth Cohort 1936, whose members took a standardized cognitive test at age 11 and underwent repeated testing in later life. Of the 1,091 adults enrolled around age 70, 866 returned for follow-up cognitive assessment. Tests covering memory, processing speed and spatial ability were administered through approximately age 82, allowing investigators to estimate each participant’s trajectory rather than relying on a single late-life score.

Postmortem corpus-callosum tissue was available from much smaller subsets of participants. Electron microscopy connected more severe decline with unusually thick myelin on remaining large-diameter axons and a relative shift toward smaller myelinated fibers.

Single-nucleus RNA sequencing of tissue from 14 donors examined 45,437 nuclei after quality control. Two oligodendrocyte subpopulations showed gene-expression changes associated with more severe decline. Computational pathway analysis identified NRF2 as a shared regulator predicted to be less active in both populations. Immunostaining in samples from 18 donors then found a lower proportion of NRF2-expressing oligodendrocytes among people with more severe decline.

NRF2 regulates cellular defenses involving oxidative stress, metabolism and protein or organelle disposal. The human results nevertheless remain associations: postmortem tissue cannot establish when the cellular changes began or whether they caused, followed or merely accompanied cognitive deterioration.

The mouse experiment tested causality more directly

To investigate mechanism, the team created mice in which NRF2 could be removed selectively from oligodendrocytes. Deletion was induced at six months, after most developmental oligodendrocyte production had occurred.

At 12 months, 38 mice completed a Morris water-maze task. Control and NRF2-deficient animals did not differ by a simple overall genotype comparison on individual testing days, but their learning patterns over time differed: the knockout mice showed less improvement in the time spent searching the target quadrant. The distinction is important because the experiment did not show a uniform loss of all cognitive performance.

Structural analyses found thicker myelin and fewer large-diameter axons in poorer-performing animals. At 18 months, a separate comparison involving 14 mice found that NRF2-deficient animals had smaller average axon diameters and thicker myelin, reproducing central features observed in the human tissue.

These results support a causal role for oligodendrocyte NRF2 loss in the mouse phenotype. A targeted genetic deletion, however, is not equivalent to naturally occurring human aging or treatment with an NRF2-activating drug.

What remains unanswered

The human tissue samples were small and came from deceased members of a geographically and generationally specific cohort. Different analyses used different donor subsets, and survival, participation and brain-donation selection could limit generalizability. Inflammation, vascular disease, Alzheimer’s pathology and other aging processes may affect both NRF2 signaling and cognition despite statistical adjustment for selected factors.

The work also focused on the corpus callosum, leaving open whether the same relationship appears across other brain regions. The pathway connecting NRF2 loss to excessive myelin remains unresolved; impaired cellular recycling is one hypothesis, not an established explanation.

NRF2 is already pharmacologically targetable, prompting interest in possible drug repurposing. But this study included no human intervention, treatment comparison, safety assessment or evidence that changing NRF2 activity improves age-related cognition. Its practical contribution is therefore a testable cellular target and a warning that more myelin is not necessarily healthier myelin—not a clinical strategy.