Dynamic symptom networks in etomidate use disorder: a cross-lagged panel network analysis of craving, compulsive drug seeking, and cognitive control

Background: Etomidate abuse has emerged as a novel psychoactive-substance threat in China, yet the symptom architecture and temporal dynamics of etomidate use disorder (EUD) remain poorly characterized. This study aims to elucidate the static and dynamic interactions among craving, compulsive drug seeking and cognitive-control deficits in individuals with EUD through cross-sectional and longitudinal network analyses. Methods: From June 2023 to December 2024, a total of 361 male individuals diagnosed with EUD, with a median age of 21 years (interquartile range: 18-26), were assessed upon admission (T0) and one month thereafter (T1) at a rehabilitation center in Hunan Province. Craving was measured with the Drug Desire Questionnaire; compulsive seeking with the Obsessive Compulsive Drug Use Scale; impulsivity with the Barratt Impulsivity Scale; and cognitive control via the Stop Signal Task. Cross-sectional symptom networks were estimated using Gaussian graphical models with GLASSO regularization. Temporal dynamics were examined with cross-lagged panel network modeling. Network centrality and stability were evaluated through bootstrap analyses. Results: Both at T0 and T1, craving, impulsivity, and compulsive drug seeking formed a stable network, while the network structure dynamically reconstructed during abstinence. Drug craving and intention acted as the core node (highest centrality across abstinence stages). Longitudinal network analysis showed that resistance to thoughts and intentions strongly predicted subsequent impulsivity. The results highlight the dynamic interplay between craving and impulsivity in driving EUD symptoms over time. Conclusion: The findings provide insights into the temporal dynamics of EUD symptoms and suggest that craving and impulsivity play pivotal roles in the disorder's progression. These results emphasize the need for targeted interventions that address these central symptoms to improve treatment outcomes.