Network Diffusion Modeling Identifies the Postcentral Gyrus as an Epicentre of Brain Atrophy in Tobacco Use Disorder

Background Tobacco use disorder (TUD) is associated with widespread structural brain alterations. Network diffusion models (NDM) propose that neuropathology spreads via brain networks, but this model's applicability to TUD-related gray matter volume (GMV) changes remains unexplored. Methods We recruited structural and resting state fMRI data from 249 males (126 with TUD, 123 healthy controls (HCs)). Regional GMV differences were calculated. The NDM was applied to the healthy functional connectome to simulate pathology spread from potential seed regions. Graph theory analysis identified vulnerable connections (edges), and results were validated against 1000 randomized null networks. Results Compared with HCs, individuals with TUD showed significant GMV reductions in regions including the cingulate gyrus, precuneus, and frontal areas. The NDM revealed that seeding the model from the postcentral gyrus produced the strongest correlation between predicted and observed GMV abnormalities. Furthermore, connections linked to the thalamus were identified as the most vulnerable. The results based on the real functional connectome were significantly different from those obtained using null models. Conclusions This study demonstrates that the NDM can effectively predict the spatial pattern of GMV alterations in TUD based on the healthy brain's functional architecture. The postcentral gyrus is a key epicenter of these changes, and thalamus-related connections are highly vulnerable, offering new insights into TUD's neural mechanisms and potential targets for intervention.