Brain-scan aging index links cognition and gut biology in three-cohort study

UCLA researchers have linked a brain-scan measure of age differences to cognitive performance, depressive symptoms and gut biological signatures in young and midlife adults. Announced by UCLA Health on September 10, the findings add evidence that brain-age research may capture meaningful differences before old age, while leaving its usefulness for individual prediction unresolved.

The study appeared online in eBioMedicine on September 9. Its relevance to healthspan lies in a central prevention question: can researchers identify useful signals of vulnerability early enough to improve later health? This work identifies associations worth testing, but does not establish that the signals predict dementia or that changing gut bacteria protects cognition.

What the researchers measured

The cross-sectional analysis included 1,462 participants across three cohorts: a discovery group of 674, a replication group of 444 and an independent group of 344. Researchers used resting-state functional MRI, which examines coordinated activity between brain regions while participants are not performing a specific task.

A statistical model estimated chronological age from those connectivity patterns. The researchers then calculated an age-bias-corrected difference between estimated and actual age, termed the Brain Aging Index. A higher value indicated an older estimated brain age relative to chronological age.

The outcomes were associations with cognitive and emotional measures, alongside gut microbiome and metabolite profiles in the independent cohort. These were research measurements rather than clinical endpoints such as new dementia diagnoses, loss of independence or survival.

That distinction matters when interpreting the word “aging.” An age estimate from a scan describes how a person's measured brain pattern compares with patterns associated with age. A single assessment cannot, by itself, show how quickly that person's brain is changing over time.

Consistent brain associations, narrower gut evidence

Across the three groups, higher index values were associated with poorer cognitive performance, particularly working memory and executive function, and greater depressive symptoms. UCLA reported that recurring connectivity patterns involved regions associated with memory and self-referential thought.

The gut analysis identified microbial and stool-metabolite signatures associated with the index. Reported compounds included ceramides and a cholesterol-related molecule, while estetrol showed an inverse association. Exploratory pathway analysis pointed toward immune, vascular, synaptic and mitochondrial biology.

These findings provide possible biological directions for follow-up. They do not demonstrate that a particular bacterium or metabolite caused the brain findings. Nor does the presence of a compound in stool establish how much reaches the brain or what it does there.

The three-cohort structure strengthens the observation that the imaging measure relates to cognition and mood. The gut component has a narrower evidence base because those biological measurements were integrated in one cohort, rather than independently reproduced across all three.

What would make the measure clinically useful?

The next questions concern prediction and action. Following participants over time could test whether the index anticipates meaningful cognitive decline beyond information already available from age, health history and conventional assessments. Repeated scans could help distinguish a persistent individual difference from a changing trajectory.

Intervention studies would address a separate question: whether modifying a proposed biological pathway improves outcomes. Even if an intervention lowered the imaging index, researchers would still need to establish that the change corresponds to better memory, function or long-term health.

For now, the practical contribution is a research framework connecting brain-network measurements with cognitive and gut biology in earlier adulthood. The evidence supports investigating those relationships, while providing no basis for selecting a microbiome treatment or interpreting a brain-age number as a personal forecast of disease.