Multimodal Evidence for Large-scale Network Imbalance in Etomidate Use Disorder.

Abstract

Background: Etomidate use disorder (EUD) is an escalating public health issue in East and Southeast Asia, yet its neuroimaging features remain poorly understood. This study used simultaneous electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) to examine multimodal functional connectivity alterations in individuals with EUD.

Methods: A total of 54 individuals with EUD and 46 healthy controls (HCs) were recruited and underwent simultaneous resting-state EEG and fNIRS recordings. Psychological dimensions including impulsivity, aggression, and craving were evaluated using standardized and validated scales. Source-level EEG functional connectivity and fNIRS hemodynamic connectivity were analyzed to identify group differences. Associations between altered connectivity features and clinical measures were examined within the EUD group. Machine learning analyses were conducted as exploratory assessments of the discriminative information contained in multimodal connectivity features.

Results: In unadjusted exploratory analyses of BIS-11 and BPAQ subscales, the EUD group showed higher cognitive impulsivity, non-planning impulsivity, physical aggression, anger and self-directed aggression (t=2.32 to 4.54, p<0.05), but lower motor impulsivity, compared with HCs. Source-level EEG analysis revealed significant group differences only in the delta band, with reduced large-scale connectivity in EUD involving default mode, dorsal attention, visual and sensorimotor systems. fNIRS analysis showed reduced hemodynamic connectivity across multiple frontal subregions and motor-related areas. Weaker connectivity between the primary motor cortex and the triangular part of the inferior frontal gyrus was associated with stronger negative reinforcement craving (rho=-0.471, p_FDR<0.05). Exploratory machine learning analysis showed preliminary discriminative performance, with the random forest model producing the numerically highest test-set AUC (0.80) in this particular data split.

Conclusions: Convergent EEG-fNIRS findings suggest delta-band network decoupling and reduced frontal hemodynamic connectivity as multimodal neural features of EUD. These findings provide preliminary evidence for the neural mechanisms of EUD and may inform future objective assessment and treatment-monitoring studies.