Connectome-based predictive modeling of Internet addiction symptomatology

Abstract

Internet addiction symptomatology(IAS) is characterized by persistent and involuntary patterns of compulsive Internet use, leading to significant impairments in both physical and mental well-being. Here, a connectome-based predictive modeling (CPM) approach was applied to decode IAS from whole-brain resting-state functional connectivity (rsFC) in healthy population. The findings showed that IAS could be predicted by the functional connectivity between prefrontal cortex with the cerebellum and limbic lobe, connections of the occipital lobe with the limbic lobe and insula lobe. The identified edges associated with IAS exhibit generalizability in predicting IAS within an independent sample. Furthermore, we found that the unique contributing network, which predicted IAS in contrast to the prediction networks of alcohol use disorder symptomatology (AUDS, the range of symptoms and behaviors associated with alcohol use disorder), prominently comprised connections involving the occipital lobe and other lobes. The current data-driven approach provides the first evidence of the predictive brain features of IAS based on the organization of intrinsic brain networks, thus advancing our understanding of the neurobiological basis of IAD susceptibility, and may have implications for the timely intervention of people potentially at risk of IAD.