Across three cohorts totaling 1,462 adults, a Bayesian ridge regression model applied to resting-state fMRI data produced a Brain Aging Index (BAI) — the residual between predicted and chronological brain age — that correlated with chronological age at r = 0.50–0.59. Higher BAI mapped consistently to dysregulated connectivity in posterior cingulate/precuneus and medial frontal regions, worse working memory and executive function, and greater depressive symptoms. Critically, multi-omics integration linked elevated BAI to stool ceramides, 24-hydroxycholesterol, and dicarboxylic acids, while estetrol showed an inverse association. KEGG enrichment implicated neuroimmune, vascular, synaptic, and mitochondrial pathways.
The finding that ceramides — sphingolipids already implicated in neuroinflammation and insulin resistance — appear in a gut-brain aging signature before midlife is notable. It suggests peripheral lipid metabolism may modulate default-mode network integrity years before clinical symptoms emerge, a hypothesis consistent with animal data linking gut-derived ceramides to hippocampal neurogenesis impairment. The estetrol inverse association is intriguing given that estrogen metabolites are known neuroprotective agents, though causal direction cannot be established in this cross-sectional design. Limitations are significant: causality is unresolvable, cohort composition and dietary data are unspecified, and fMRI-derived brain age models carry inherent scanner-site variance. Still, replicating BAI associations across three independent samples elevates this beyond typical single-cohort neuroimaging work. For longevity-focused adults, this incrementally strengthens the case for monitoring gut lipid profiles — particularly ceramides — as early brain-health indicators, pending longitudinal validation.