Abnormally increased intrinsic neural timescales in sensory and default mode networks in cocaine use disorder

Cocaine use disorder (CUD) is associated with abnormal structural and functional brain changes. However, the neurodynamics and molecular underpinnings remain unclear. In this study, we mapped whole-brain intrinsic neural timescales (INTs), reflecting temporal neural processing, using resting-state functional magnetic resonance imaging data from 44 CUD patients and 44 healthy controls (HC). CUD showed increased INTs in visual, somatomotor, and default mode networks compared with HC. Mediation analysis linked local INTs abnormalities to altered dorsal attention network neurodynamics, associated with inhibitory control deficits. Notably, these changes were primarily correlated with alterations in gamma-aminobutyric acid type A receptors and the noradrenaline transporter. Machine learning classifiers based on INTs achieved a maximum accuracy of 75.5% in distinguishing CUD from HC, with a generalization accuracy of 65.0% on an independent dataset. This study elucidates aberrant neural mechanisms underlying CUD and highlights INTs as promising diagnostic biomarkers for clinical detection and intervention.