Network localization of gray matter alterations in chronic smokers using the normative functional connectome

Chronic smoking has well-documented impacts on brain structure. Voxel-based morphometry (VBM) investigations have revealed diverse regional gray matter (GM) changes in chronic smokers, hindering a unified understanding of smoking-induced neuropathology. To reconcile these findings, this study aimed to identify common intrinsic functional networks underlying these structural alterations using a functional connectivity network mapping (FCNM) approach. We further explored potential exposure-dependent variations to characterize how brain network architecture relates to cumulative smoking dose.

We utilized coordinate-based FCNM to quantitatively integrate heterogeneous findings from previous VBM studies. We systematically reviewed VBM studies reporting GM differences between chronic smokers and non-smokers. We identified peak coordinates from 27 studies, encompassing 36 contrasts with 1,336 smokers and 1803 non-smokers. Resting-state fMRI from 1,093 healthy participants (Human Connectome Project) were utilized to create individual functional connectivity maps based on seed coordinates. Maps were combined to identify a shared alteration network and evaluated for spatial overlap with established canonical brain networks. Sensitivity analysis were conducted with different seed radii. Crucially, subgroup analysis stratified studies into higher-exposure and lower-exposure groups to investigate exposure-dependent mechanisms.

Functional connectivity network mapping identified a widespread network linked to smoking-induced GM changes. Key nodes included the supramarginal gyrus, insula, anterior cingulate cortex, caudate nucleus, putamen, and superior temporal gyrus. Spatial overlap analysis revealed predominant involvement of the posterior Salience Network (51.59%), anterior Salience Network (32.15%), basal ganglia network (31.52%), and auditory network (24.19%). Sensitivity analysis confirmed the robustness of these findings. Subgroup analysis revealed exposure-dependent patterns: while the Salience and basal ganglia networks were consistently affected in both groups, the auditory network and ventral Default Mode Network showed markedly greater involvement in the higher-exposure group, largely spared in the lower-exposure group.

This FCNM approach identified consistent brain networks, predominantly the Salience, basal ganglia, and auditory networks, associated with chronic smoking-related GM alterations. These findings offer network-level insight into the structural effects of smoking, helping to resolve discrepancies and potentially guiding tailored interventions. Furthermore, the findings suggest a progressive neuropathological expansion, characterized by the concurrent recruitment of sensory (auditory) and high-order cognitive systems (ventral Default Mode Network) with cumulative smoking exposure.