Signatures of Critical Brain Dynamics in Adolescents: A Functional Renormalization Analysis on Depression and Anxiety

According to the National Alliance on Mental Illness (NAMI), approximately 50% of mental health disorders emerge by age 14, and 75% by age 24. Therefore, understanding brain patterns during adolescence (a period of high plasticity and reorganization) is crucial to identify early markers of vulnerability. This is particularly relevant for conditions such as anxiety and depression, which are the leading causes of disability in this age group.

An emerging approach to characterize brain organization, both in healthy and clinical populations, is the critical brain hypothesis. This framework proposes that the healthy brain behaves like a physical system undergoing a phase transition, where it self-regulates to remain near a critical point (a phenomenon known as self-organized criticality). This critical state is reflected in power-law scaling in brain activity and has been suggested to support optimal cognitive function by enabling a balance between stability and flexibility in information flow.

In this project, we propose to apply the phenomenological renormalization group (PRG) method to resting-state fMRI data from the Boston Adolescent Neuroimaging of Depression and Anxiety (BANDA) dataset of the Human Connectome Project, consisting of adolescents aged 14 to 17 diagnosed with anxiety, depression, both conditions, or no psychopathology. This functional coarse-graining approach allows us to examine how collective brain dynamics evolve across scales and to estimate three scaling exponents: α, describing how the variance increases as signals are aggregated; β, related to the probability of silence within clusters (interpreted as a form of free energy); and μ, which characterizes the decay of the eigenvalue spectrum of the covariance matrix and reflects the distribution of activity modes.

Our goal is to explore the variation of these exponents across clinical groups, which may provide evidence of altered collective brain dynamics in adolescent mental health. This work integrates concepts from statistical physics with neuroimaging tools to contribute to a mechanistic understanding of the developing brain.