A small cross-sectional study explored whether gut microbiome changes and associated blood metabolites could aid early dementia detection. Researchers analyzed blood and stool from 150 adults aged 50 and over, spanning no cognitive impairment, mild cognitive impairment (MCI), and subjective cognitive impairment (SCI).
They identified 33 microbiome- and diet–related molecules and gut microbe–derived metabolites. Using AI-driven analysis, six metabolites formed a model that classified participants into the three groups with 79% overall accuracy and distinguished healthy individuals from those with MCI with over 80% accuracy.
The work supports the premise that alterations in the gut–brain axis occur early in cognitive aging and that blood and stool profiling can be less invasive than brain imaging or spinal fluid testing. The authors caution that the study is early-stage and cross-sectional, so it demonstrates association rather than proven predictive capability for progression to dementia.
Uncertainty remains regarding generalizability and long-term predictive value; further longitudinal validation is needed to determine clinical utility.
Gut–brain signals as a preclinical clue to cognitive aging: study design and key signals
Specific longitudinal or outcome data were not reported in the source.
Blood samples were assayed for metabolites linked to gut microbiome metabolism.
They then applied AI-powered machine learning to evaluate combinations of these metabolites for group classification (healthy, MCI, SCI).
The objective was to determine whether a small metabolite panel could differentiate cognitive status categories.