Back

Healthy Aging As Information Divergence In The Multiplex Brain

Ghosh, D.; Ray, D.; Das, M.

2026-06-02 neuroscience
10.64898/2026.05.29.728670 bioRxiv
Show abstract

AO_SCPLOWBSTRACTC_SCPLOWUnderstanding the co-evolution of the human brains structural scaffold and functional traffic across the adult lifespan remains a fundamental challenge in neuroscience. While age-related degradation in grey matter and functional activation is well-documented, the joint trajectory of the structural (SC) and functional (FC) connectomes is often overlooked due to the lack of an integrative framework. Here, we model the brain as a multiplex network to quantify the information-theoretic interdependencies between these two layers in a cross-sectional cohort of 589 healthy individuals (ages 18-88) from the Cam-CAN dataset. Using Jensen-Shannon Divergence and relative entropy metrics, we identify a fundamental organizing principle of healthy aging: a progressive information divergence where functional dynamics increasingly "untether" from their underlying structural constraints. Our results reveal that this decoupling follows a robust linear trajectory, yet is highly spatially hetero-geneous. Meso-scale community analysis using the Multiplex Map Equation identifies subcortical hubs--specifically the putamen, pallidum, caudate, and thalamus--as the primary epicenters of age-related divergence. This topological shift toward functional independence in subcortical "switch-boards" provides a mechanistic connectomic signature for the well-documented decline in fluid intelligence and motor adaptation. In striking contrast, the limbic core (hippocampus and entorhinal cortex) exhibits remarkable stability, suggesting a biological imperative to preserve high-fidelity memory circuits amidst global communicative rewiring. By framing healthy aging as a system-atic subcortical untethering alongside rigid limbic resilience, our work provides a powerful new multiplex baseline to distinguish normative cognitive decline from the early topological signals of neurodegenerative disease.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.