In 110 metagenomics and 121 serum metabolomics samples from older Beijing Hospital adults, multivariable-adjusted analyses identified 57 bacterial species and 56 metabolites associated with intrinsic capacity (IC) decline — the WHO framework encompassing cognitive, locomotor, sensory, vitality, and psychological domains. Healthy IC was enriched with Lactobacillus zeae and Paenibacillus glucanolyticus, while decline correlated with disruptions in amino acid pathways (L-serine, cysteine), carnitine metabolism (N6,N6,N6-trimethyl-L-lysine, C5-OH carnitine), and senescence-linked compounds including N1,N8-diacetylspermidine and secondary bile acid 3-epideoxycholic acid.
What makes this work methodologically notable is the multivariable adjustment for age, sex, comorbidity burden, and dietary habits — a step that most microbiome-aging studies omit, leaving findings confounded by lifestyle variables. By doing so, these investigators isolate IC-specific signals rather than generic aging noise. The carnitine pathway findings are particularly compelling: carnitine deficiency is already implicated in sarcopenia and mitochondrial dysfunction, and its appearance here suggests a mechanistic thread worth pursuing therapeutically. Similarly, spermidine's role in autophagy induction means N1,N8-diacetylspermidine as a senescence marker adds a plausible biological story. Critical limitations remain: the cross-sectional design prevents causal inference, the sample is modest in size, and the Beijing cohort limits generalizability. This is best categorized as hypothesis-generating — solid, well-designed foundational work that demands longitudinal replication before any clinical translation.
