Altered integrated and segregated states in cocaine use disorder.

Introduction: Cocaine use disorder (CUD) is a chronic brain condition that severely impairs cognitive function and behavioral control. The neural mechanisms underlying CUD, particularly its impact on brain integration-segregation dynamics, remain unclear. Methods: In this study, we integrate dynamic functional connectivity and graph theory to compare the brain state properties of healthy controls and CUD patients. Results: We find that CUD influences both integrated and segregated states, leading to distinct alterations in connectivity patterns and network properties. CUD disrupts connectivity involving the default mode network, frontoparietal network, and subcortical structures. In addition, integrated states show distinct sensorimotor connectivity alterations, while segregated states exhibit significant alterations in frontoparietal-subcortical connectivity. Regional connectivity alterations among both states are significantly associated with MOR and H3 receptor distributions, with integrated states showing more receptor-connectivity couplings. Furthermore, CUD alters the positive-negative correlation balance, increases functional complexity at threshold 0, and reduces mean betweenness centrality and modularity in the critical subnetworks. Segregated states in CUD exhibit lower normalized clustering coefficients and functional complexity at a threshold of 0.3. We also identify network properties in integrated states that are reliably correlated with cocaine consumption patterns. Discussion: Our findings reveal temporal effects of CUD on brain integration and segregation, providing novel insights into the dynamic neural mechanisms underlying cocaine addiction.