Shared brain and genetic architectures between mental health and physical activity

Abstract

Physical activity is correlated with, and effectively treats various forms of psychopathology. However, whether biological correlates of physical activity and psychopathology are shared remains unclear. Here, we examined the extent to which the neural and genetic architecture of physical activity and mental health are shared. Using data from the UK Biobank (N=6,389), canonical correlation analysis was applied to estimate associations between the amplitude and connectivity strength of sub-networks of three major neurocognitive networks (default mode, DMN; salience, SN; central executive networks, CEN) with accelerometer-derived measures of physical activity and self-reported mental health. We estimated the genetic correlation between mental health and physical activity measures, as well as putative causal relationships by applying linkage disequilibrium score regression, genomic structural equational modeling, and latent causal variable analysis to genome-wide association summary statistics (GWAS N=91,105-500,199). Physical activity and mental health were associated with connectivity strength and amplitude of the DMN, SN, and CEN (all r>0.13, all p<0.048). These neural correlates exhibited highly similar loading patterns across mental health and physical activity models even when accounting for their shared variance. This suggests a largely shared brain network architecture between mental health and physical activity. Mental health and physical activity were also genetically correlated (|rg|=0.085-0.121), but we found no evidence for causal relationships between them. Collectively, our findings provide empirical evidence that mental health and physical activity have shared brain and genetic architectures and suggest potential candidate sub-networks for future studies on brain mechanisms underlying beneficial effects of physical activity on mental health.

The Effect of Neuroepo on Cognition in Parkinson’s Disease Patients Is Mediated by Electroencephalogram Source Activity

“…Nevertheless, the studies cited evaluate cognitive or qEEG variables as completely separate modalities. Recent work in the multi-omics and multidomain search for brain biomarkers has stressed, to great advantage, extracting underlying latent variables to focus on shared sources of variability ( Smith and Nichols, 2018 ; Liu et al, 2021 ; Zhang et al, 2022 ). Many of these proposals focus on modern variants of Canonical Correlation Analysis (CCA), a significant application to PD described in Cai et al (2022) .…”