Complexity, variability, and the behavioral layer of gerophysics: a commentary on Unfried et al. (2026)

This is the authors' abstract. We don't add key points for this paper.

The report of the Global Conference on Gerophysics [1] aims to "bridge changes across biological scales: from subcellular to organism level", with dynamical systems, entropy, network theory, and stochastic processes as its theoretical pillars.This ambition can be extended to account for the organism-environment interactions as expressed through behavioral dynamics, thereby adding a continuous layer to the field.Gerophysics already spans scales from the molecular to organism-level (frailty, resilience, physiological decline, and population-level health trajectories); yet, capturing how the aging organism copes in real time with everyday demands requires specific measures, formalized within the same framework.Well before the term "Gerophysics" was introduced, Lipsitz and Goldberger [2] proposed that aging is accompanied by a measurable reduction in the dimensionality and adaptive richness of physiological and behavioral output.They argued that healthy physiological systems generate fractal-like fluctuations across multiple time scales and that aging degrades this complexity, yielding outputs that are either more random or more deterministic, reflecting a loss of adaptive functional solutions.This age-related loss of complexity may offer a testable behavioral correlate of the reductions in stability, resilience and robustness addressed by Fedichev and Gruber and by Rutenberg and Pridham within the Gerophysics framework.Whether this loss of complexity is formally equivalent to a reduction in attractor dimensionality in these models remains an open hypothesis rather than an established correspondence.Entropy-based and fractal indices of time-series complexity have been applied across sub-levels of the integrated brain-mind-body system, including the way in which aging reshapes cognitive-motor interactions [3].Findings ranging from brain [4] to muscular [5] signal complexity, and from movement variability [6] to the prediction of future falls [7], collectively constitute a coherent organism-level framework rather than isolated observations.A further argument for its irreducibility comes from Kerepesi's preliminary and as-yet unpublished findings suggesting that network entropy decreases with age in adults, in contrast to molecular entropy, which increases.If confirmed, this divergence would suggest that entropy and complexity are level-specific constructs and that a decrease in network entropy need not be a simple mirror image of molecular entropy.On this view, the loss of behavioral complexity would be neither equivalent to nor inferable from molecular disorder or network entropy.