Individual-specific subspace of altered functional connectivity predicts clinical symptoms in Internet gaming disorder.

BACKGROUND

Despite extensive research efforts aimed at unraveling the neural mechanisms underlying Internet gaming disorder (IGD), reproducibility remains a challenge, largely due to overlooking the clinical and biological diversity within individuals affected by IGD. Therefore, investigating the altered brain features associated with IGD within both individual-shared and individual-specific subspaces is crucial for understanding this complex and heterogeneous disorder.

METHODS

This study included 555 participants, comprising 326 individuals with IGD and 229 recreational game users (RGUs). Firstly, we computed altered functional connectivity (AFC) matrices for individuals with IGD and compared them with those of RGUs. Subsequently, we applied the common orthogonal basis extraction algorithm to partition the AFC of individuals with IGD into individual-shared and individual-specific subspaces. Finally, we examined brain regions exhibiting generally abnormal patterns in the individual-shared subspace and employed multiple linear regression analysis to assess the predictive influence of AFC within the individual-specific subspace on clinical symptoms.

RESULTS

Our findings revealed individual-shared altered patterns in the visual network, medial frontal network (MFN), and frontoparietal network (FPN) among individuals with IGD, which are associated with executive control and visual processing. Within the individual-specific subspace, we observed that AFC within the default mode network could predict scores related to fun-seeking behavior in the behavioral activation system (BAS), while AFC within the MFN correlated with reward responsiveness and drive scores in the BAS. Additionally, AFC within the FPN was predictive of scores in the behavioral inhibition system.

CONCLUSIONS

This study successfully decomposed the AFC of IGD into individual-shared and individual-specific subspaces. The AFC within individual-specific subspaces holds promise as potential biomarkers for elucidating clinical symptoms in IGD, thereby offering an analytical framework for investigating heterogeneity in other addictive behaviors.