The neural foundations of the Dark Triad (DT) personality traits—narcissism, Machiavellianism, and psychopathy—remain largely unexplored despite their profound impact on social behavior. Previous research has indicated potential structural and functional alterations in the default mode network (DMN) and reward-related regions in individuals exhibiting high levels of DT traits. However, these studies have often been limited by univariate analytical methods and small, unbalanced sample sizes. To address these gaps, we analyzed brain scans from a robust sample of 200 participants (mean age: 32.43 years; 105 females) using group Independent Component Analysis (gICA), an unsupervised machine learning approach. This method allowed us to extract temporal variability associated with resting-state macro-networks without relying on predefined regions of interest. Our analysis identified 20 macro-networks, of which two were significantly predictive of DT traits: the Central Executive Network (CEN) and the posterior hub of the Default Mode Network (DMN). Notably, higher DT scores correlated with increased blood-oxygen-level-dependent (BOLD) temporal variability within the CEN and decreased variability within the DMN. The heightened activity in the CEN may reflect enhanced manipulative abilities and strategic planning commonly observed in individuals with high DT traits, which is supported by the association between the CEN and the Machiavellianism subscale. Conversely, reduced functionality in the DMN may correspond to diminished self-reflective and emotional capacities in these individuals. These findings contribute to a deeper understanding of the neurobiological basis of DT traits and hold implications for future research and interventions.

Anxiety is a diffuse condition ranging from milder manifestations to severe disorders, impacting individuals’ lives significantly. Specific sensitive periods such as adolescence and young adulthood are particularly vulnerable to anxious states, often associated with psychological traits like impulsivity, aggression, and varying coping strategies. The goal of the present study is to address the need for a comprehensive analysis of trait anxiety by employing Parallel ICA, a data fusion machine learning technique, in a sample of young individuals divided into a lower anxiety group (n=252) and a higher anxiety group (n=302), aiming to identify the joint gray-white matter networks characterizing higher versus lower trait anxiety. Additionally, we aim to characterize higher anxiety individuals for their usage of maladaptive coping strategies, and other affective dimensions. In higher anxious individuals, we identified a fronto-parieto-cerebellar network with decreased gray matter concentration, linked to bodily awareness and threat modulation, and a parieto-temporal network with increased white matter concentration, emphasizing insula and precuneus role. At the psychological level, we found higher stress, cognitive and motor impulsivity, and avoidance/emotional coping in higher anxious individuals. These findings may enhance the understanding of anxiety’s neural underpinnings in young individuals, supporting early interventions.