Gray and White Matter Networks Predict Mindfulness and Mind Wandering Traits: A Data Fusion Machine Learning Approach.

Abstract

Background: Mindfulness and mind wandering are cognitive traits central to attentional control and psychological well-being, yet their neural underpinnings are yet to be elucidated. This study aimed to identify structural brain networks comprising gray matter (GM) and white matter (WM) that predict individual differences in mindfulness and distinct mind wandering tendencies (deliberate and spontaneous).

Methods: Using structural MRI data and self-report measures from 76 participants, we applied an unsupervised data-fusion machine learning technique (parallel independent component analysis) to identify GM and WM networks associated with mindfulness and mind wandering traits.

Results: Our analysis revealed several distinct brain networks linked to these cognitive constructs. Specifically, one GM network involving subcortical regions, including the caudate and thalamus, positively predicted mindfulness and deliberate mind wandering, while negatively influencing spontaneous mind wandering through the mediating role of the mindfulness facet "acting with awareness." In addition, two separate WM networks, predominantly involving frontoparietal and temporal regions, were directly associated with reduced spontaneous mind wandering.

Conclusions: These findings advance our current knowledge by demonstrating that specific GM and WM structures are involved in mindfulness and different forms of mind wandering. Our results also show that the "acting with awareness" facet has a mediating effect on spontaneous mind wandering, which provides supporting evidence for attentional and executive control models. These new insights into the neuroanatomical correlates of mindfulness and mind wandering have implications for ongoing research in the growing topic of mindfulness and mind wandering, mindfulness-based interventions, and other clinical applications.