Gut microbiome changes after lifestyle intervention in older adults at dementia risk

A 24-week multidomain lifestyle intervention was associated with changes in the gut microbiome of older adults at risk of dementia in an exploratory sub-study of the SUPERBRAIN trial. The analysis included 130 people aged 60–79 who had at least one modifiable dementia risk factor and no overt dementia; 91 participated in the intervention and 39 were controls. Stool samples collected at baseline and week 24 were examined using 16S rRNA gene sequencing. The intervention group also showed greater improvement in cognitive performance than the control group, as measured with the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) Total Scale Index.

Within the intervention group, observed amplicon sequence variant richness and inverse Simpson diversity both decreased between baseline and week 24, with P values of 0.027 and 0.039, respectively. Neither measure changed significantly over time among controls, for which the corresponding P values were 0.464 and 0.812. Weighted UniFrac beta diversity, which assessed differences in bacterial community composition, also shifted significantly within the intervention group (P < 0.001), while the change among controls was not significant (P = 0.193). A dysbiosis index decreased within the intervention group (P < 0.001), consistent with a shift toward the study’s healthy reference microbiome configuration; the corresponding change among controls was not significant (P = 0.372).

Integrated analyses identified 19 genera showing directionally consistent changes in the intervention group. These included enrichment of genera described as producers of short-chain fatty acids, including Faecalibacterium and Butyrivibrio, alongside reductions in members of the Enterobacteriaceae family. The investigators evaluated microbiome composition alongside predicted microbial functional profiles, meaning that the functional findings were estimates based on taxonomic information rather than direct measurements of metabolites or metabolic activity. The study therefore reported a microbial composition consistent with changes in functional potential, but it did not directly establish changes in short-chain fatty acid production or other biological pathways.

Longitudinal changes in several microbial genera were associated with changes in clinical outcomes. These associations showed that particular patterns of microbiome change coincided with changes in cognitive or physical-function measures, but the analysis did not demonstrate that the microbial changes caused the clinical changes. Similarly, the microbiome findings cannot be said to account for the greater RBANS improvement observed in the intervention group, because the study did not establish that microbiome remodeling mediated the treatment effect. Its design supports characterization of associations within this trial rather than a definitive mechanistic explanation of how the lifestyle program influenced cognition.

The researchers separately assessed whether baseline microbiome and clinical information could identify participants who would respond to the intervention, defined as an RBANS improvement greater than zero. An exploratory machine-learning model combined microbiome composition with clinical features and achieved an area under the receiver operating characteristic curve of 0.83. Baseline Short Physical Performance Battery score and Eubacterium abundance were the strongest contributors. This performance was reported from the study sample rather than an independent validation cohort, making the result preliminary and limiting its generalizability. The study was also exploratory in scope, and the authors called for larger investigations incorporating multi-omics approaches and independent cohorts to confirm the findings and clarify the mechanisms underlying the associations between microbial change, physical function and cognitive response.

Subject of Research: Associations between a multidomain lifestyle intervention, gut microbiome changes and cognitive outcomes in older adults at risk of dementia.

Article Title: Longitudinal changes in the gut microbiome following a multidomain lifestyle intervention in older adults at risk of dementia: an exploratory sub-study of the SUPERBRAIN trial

Article References: Yoon, C. W., Kang, W., Lee, J.-Y., Sung, H., Oh, H.-S., Hyun, D.-W., Jeong, J. H., Na, H. R., Moon, S. Y., Hong, C. H., Park, Y. K., & Choi, S. H. (2026). Longitudinal changes in the gut microbiome following a multidomain lifestyle intervention in older adults at risk of dementia: an exploratory sub-study of the SUPERBRAIN trial. BMC Geriatrics. https://doi.org/10.1186/s12877-026-08421-7

Image Credits: AI Generated

DOI: 10.1186/s12877-026-08421-7

Keywords: gut microbiome, multidomain intervention, older adults, dementia risk, cognitive function, 16S rRNA sequencing, microbiome diversity, SUPERBRAIN trial, machine learning, exploratory study, Longitudinal, changes