A multicohort study examining healthy adults aged 18 to 65 has found that a brain aging index—an imaging-based estimate of how old the brain appears relative to a person’s chronological age—maps onto thinking skills, mood and gut biology. Conducted by Kanhao Zhao and colleagues with support from National Institutes of Health funding, the research spans data drawn from groups in the United States and Japan and hinges on cross-cohort analysis.
The brain aging index is produced by training a computer model to predict age from brain scans; a positive score suggests an older-appearing brain, while a negative score indicates a younger-looking brain. The concept has been proposed as a potential marker of brain health, with earlier work linking older-looking brains to poorer cognitive and physical aging and a higher risk of death.
Lead author Zhao, who conducted the study during postdoctoral work at the University of California, Los Angeles (the Church Lab) and is now a postdoctoral researcher at Stanford University School of Medicine, said the project sought to test whether meaningful brain-age differences could be detected earlier in life and whether these differences related to gut biology as well as cognition and mood.
To probe these questions, the team analysed resting-state functional magnetic resonance imaging data and a battery of cognitive and emotional assessments from three samples: a discovery group of 674 participants, a replication group of 444, and an independent group of 344.
In the independent sample, stool samples were collected and analysed using metagenomics and metabolomics to profile gut bacteria and metabolites, respectively.
Across the three cohorts, the brain-age estimates tracked participants’ actual ages reasonably well, though with imperfect precision: predictions were typically off by about six to nine years. Those with older-appearing brains showed a consistent pattern of brain connectivity changes, including stronger connections within the default mode network and the frontoparietal network, with the strongest links in the precuneus and posterior cingulate cortex—regions associated with internal thought and memory retrieval.
Conversely, individuals with older-appearing brains exhibited weaker connections in the ventral attention network, a pattern associated with poorer performance on cognitive tests. They tended to perform more poorly on measures of performance IQ, working memory and executive function, and reported more depressive symptoms. The researchers note the associations are modest: correlations between brain-age scores and individual cognitive or mood measures ranged roughly from 0.12 to 0.38, meaning brain-age scores explained at most about 14 percent of the variance in any single outcome.
“The effects should be viewed as meaningful population-level associations rather than as a diagnostic test for an individual person,” Zhao noted. “An important strength of the study is that the brain-age pattern was examined across multiple independent cohorts, but the associations with cognition, mood, and gut biology were generally modest. At this stage, the brain aging index is better understood as a research marker of individual variability in brain health than as something that could tell a person how quickly their brain is aging in a clinical setting.”
The gut data yielded similarly modest signals. In 182 participants with complete data, a combined profile of gut bacteria and stool chemicals accounted for roughly 2 percent of the differences in brain-age scores, a finding that exceeded chance. Bacteria such as Clostridium and Collinsella and stool metabolites including ceramides and 24-hydroxycholesterol were among the strongest contributors to older-looking brains, while younger-looking brains tended to relate to metabolites linked to neurotransmitter function and amino acid processing. Notably, estetrol—a form of estrogen—was among the molecules most strongly tied to a younger brain, the authors said, hinting at a possible protective role for reproductive hormone pathways.
“What I found particularly interesting was that the signal was not confined to the brain,” Zhao said. “We observed coordinated associations involving functional brain networks, cognition and mood, and gut-derived biological features, including microbial and metabolic signatures related to lipid metabolism, immune and vascular processes, synaptic function, and mitochondrial pathways. This convergence across very different biological levels suggests that understanding brain aging may ultimately require looking beyond the brain alone.”
Crucially, the study found that cognitive and mood measures aligned with brain-network patterns rather than with any single gut bacterium or metabolite. Zhao emphasised that the findings reflect broad, systemic processes rather than straightforward gut-brain causality.
“One important message is that people of the same chronological age can show meaningful differences in how their brain networks resemble patterns associated with aging,” Zhao explained. “We also found that these brain-age differences were linked to patterns in gut microbes and metabolites, suggesting that brain aging may be related to broader biological processes across the brain-gut system rather than occurring in isolation.”
As with all observational work, the authors stress that the study is cross-sectional and cannot establish causality. They caution against inferring that particular gut microbes or metabolites cause the brain to age faster, or that changing the microbiome would necessarily make the brain younger. Diet and other factors could influence both gut biology and brain connectivity at the same time, they note, and the brain-age measure used is a statistical index derived from functional connectivity rather than a direct measurement of biological brain age.
The gut analyses were conducted in 182 participants and were drawn from datasets in the United States and Japan, with the majority of participants in the group being women. The researchers acknowledge that larger and more balanced studies are needed to confirm the signatures observed and that sequencing identifies broad bacterial species rather than specific strains that might have different health effects.
Looking ahead, Zhao said the team aims to move from cross-sectional associations to longitudinal and prospective work to determine whether brain-body signatures can predict future cognitive trajectories and whether they change in response to modifiable factors or interventions. They also envisage expanding beyond the brain-gut axis toward a more integrated view of brain-body ageing, noting that brain function is closely related to cardiovascular physiology and other organ systems.
“The next step is to move from cross-sectional associations toward longitudinal and prospective studies,” Zhao said. “We would like to determine whether these brain-body signatures can predict future cognitive trajectories and whether they change in response to modifiable factors or interventions. More broadly, we are interested in expanding beyond the brain-gut axis toward a more integrated view of brain-body aging. Emerging research suggests that brain function is also closely related to cardiovascular physiology and other organ systems.”
“In future studies, we hope to combine brain imaging with cardiac imaging, physiological signals such as ECG and PPG, and molecular measurements to better understand how aging processes across different organ systems interact,” he added. “Ultimately, our goal is to identify multimodal biomarkers that explain why individuals age differently and may eventually help guide personalised strategies for maintaining brain health.”
An ECG records the heart’s electrical activity, while PPG uses light sensors, like those in many smartwatches, to track pulse and blood flow.
The study, “Brain-gut crosstalk associated with brain ageing in young and mid-life adults: a multicohort cross-sectional study,” was authored by Kanhao Zhao, Gabriel A. Vignolle, Jennifer S. Labus, Emeran A. Mayer, Allison Vaughan, Marika Dy, Priten Vora, Ming W. Hung, Keith Vossel, Chris Gill, Daniele Del Rio, Catherine Stanton, R. Paul Ross, John F. Cryan, Rima Kaddurah-Daouk, Yu Zhang, and Arpana Church.
