Subtyping Internet Gaming Disorder Based on Interindividual Deviation of Functional Connectivity: A Network Contingency Analysis.

Internet gaming disorder (IGD) is a behavioural addiction characterized by substantial clinical heterogeneity. The present study aimed to investigate this heterogeneity by analysing the deviation patterns of interindividual brain functional connectivity. Resting-state functional magnetic resonance imaging (fMRI) data from 87 patients with IGD and 56 healthy controls (HCs) were included. First, functional connectivity matrices were constructed, and the interindividual deviation of functional connectivity (IDFC) of each patient with IGD relative to the HC group was quantified at the network level. K-means clustering was then applied to identify IGD subtypes. Edge-wise functional connectivity analysis with network contingency analysis (NCA) was used to characterize subtype-specific patterns of altered connections. Finally, support vector regression (SVR) based on network-level IDFC features was used to predict Internet Addiction Test (IAT) scores in the overall IGD cohort. Two distinct IGD subtypes were identified based on IDFC patterns. Compared with HCs, Subgroup 1 showed a widespread hypoconnectivity pattern, whereas Subgroup 2 showed a relative hyperconnectivity pattern. Network-level IDFC features also predicted IAT scores in the overall IGD cohort using SVR (r = 0.391, one-tailed permutation p = 0.038). Patients with IGD exhibit substantial interindividual variability in brain functional connectivity. Subtyping IGD based on these connectivity deviation patterns may improve the understanding of its neurobiological heterogeneity and facilitate the development of more precise individualized interventions.