Investigation of Changes in Causality Throughout Life—A Magnetoencephalogram Study Using Granger Causality and Transfer Entropy

The use of magnetoencephalogram (MEG) signals in cognitive neuroscience research to investigate the functioning of the brain has increased over recent years. In this sensor space study, Granger Causality (GC) and Transfer entropy (TE) were applied to resting state MEGs from 220 healthy volunteers (aged 7–84) to characterise the possible changes in causality due to age and gender. Additionally, graph theory principles were used to evaluate different network components such as integration (global efficiency), segregation (clustering coefficient and modularity), centrality (betweenness), and resilience (strength and assortativity). Results showed that males had higher GC than females until mid-adulthood (~60 years). However, this gender difference was not observed using TE. Moreover, complex network analysis results of low global efficiency, high clustering coefficient, and low node strength, suggest that at rest, the brain topology resembled a network made up of loosely connected modules that had segregated and disassortative nodes with low resistance to change. Statistical analyses of results from both techniques, using pairwise t-test and two-way ANOVA, showed that age had a significant effect (p < 0.05) in all brain regions for both genders with significant gender differences being observed over the anterior, posterior, left lateral and right lateral regions of the brain. The results from this study could be used to develop a fingerprint of healthy ageing, which can potentially be used to assist with the identification of alterations to background brain activity due to pathology.