Illustrative stock photograph by Sahand Babali on Unsplash; it does not depict study participants, brain scans, laboratory samples, or results.
Older adults who produced larger increases in certain walnut-related gut metabolites also tended to preserve more cortical thickness over two years. That is an intriguing microbiome signal—not proof that walnuts prevent dementia.
What changed
This MRI substudy included 104 cognitively healthy older adults from the Walnuts and Healthy Aging trial. Fifty-five were assigned to a walnut-rich diet. Within that group, larger increases in urinary urolithin A, urolithin C, and a related conjugate were associated with better preservation of cortical thickness, mainly in right-hemisphere regions. Urolithins are produced when gut microbes transform compounds found in walnuts and some other plant foods.
How strong is the evidence?
Longitudinal observational analysis inside a randomized trial. The two-year follow-up, urine measurements, and MRI scans are strengths. But the key comparison linked changes in metabolites with changes in brain structure; it was not a direct randomized test showing that walnuts preserved cortex.
What remains uncertain
Only a subset of the larger trial was analyzed, and cortical thickness is a surrogate marker—not a diagnosis, memory score, or dementia outcome. People differ in their ability to produce urolithins, and unmeasured diet or health differences may influence the association. Funding included the California Walnut Commission, so independent confirmation is valuable.
Why it matters
The study offers a plausible food–microbiome–brain connection and may help explain why the same food affects people differently. It does not support buying urolithin supplements or treating walnuts as dementia prevention.
Related: What a 12-week walnut trial found about visceral fat.
Sources
Original study in Food & Function; PubMed record. Published November 2, 2026.
