Punishment sensitivity and altered brain network centrality in cross-platform problematic online behaviors

Problematic online behaviors often span multiple digital platforms, yet previous research has focused on single forms of problematic use. It remains unclear whether distinct cross-platform behavioral profiles are associated with differential reward and punishment sensitivity and alterations in brain network topology. This study combined latent profile analysis with resting-state functional magnetic resonance imaging to identify latent risk profiles of cross-platform problematic online behaviors among Chinese university students and examine their motivational and neural correlates. In Study 1, 947 students completed measures of problematic usage of smartphones, video gaming, and social networking, and reward/punishment sensitivity. Latent profile analysis identified three graded profiles: Low-Risk, Medium-Risk, and High-Risk. In Study 2, a stratified subsample of 150 participants underwent resting-state functional magnetic resonance imaging; whole-brain eigenvector centrality (EC) indexed hub-like network properties. The Medium-Risk and High-Risk groups showed higher reward sensitivity than the Low-Risk group, with no difference between at-risk groups. Punishment sensitivity increased stepwise across profiles. Neuroimaging analyses revealed reduced EC in the left superior frontal gyrus and left precuneus in both at-risk groups relative to the Low-Risk group. Lower EC in these regions was associated with greater punishment sensitivity, whereas reward sensitivity showed no significant association with EC. These findings suggest that punishment sensitivity is not only a psychological marker of POB severity but is also associated with reduced hub centrality in the left superior frontal gyrus and left precuneus, indicating a potential neural correlate of heightened sensitivity to negative outcomes.