Neural fingerprints of gambling disorder using diffusion tensor imaging

Abstract

Gambling disorder (GD) is a behavioral addiction associated with personal, social and occupational consequences. Thus, examining GD's clinical relationship with its neural substrates is critical. We compared neural fingerprints using diffusion tensor imaging (DTI) in GD subjects undergoing treatment relative to healthy volunteers (HV). Fifty-three (25 GD, 28 age-matched HV) males were scanned with structural magnetic resonance imaging (MRI) and DTI. We applied probabilistic tractography based on DTI scanning data, preprocessed and analyzed using permutation testing of individual connectivity weights between regions for group comparison. Permutation-based comparisons between group-averaged connectomes highlighted significant structural differences. The GD group demonstrated increased connectivity, and striatal network reorganisation, contrasted by reduced connectivity within and to frontal lobe nodes. Modularity analysis revealed that the GD group had fewer hubs integrating information across the brain. We highlight GD neural changes involved in controlling risk-seeking behaviors. The observed striatal restructuring converges with previous research, and the increased connectivity affects subnetworks highly active in gambling situations, although these findings are not significant when correcting for multiple comparisons. Modularity analysis underlines that, despite connectivity increases, the GD connectome loses hubs, impeding its neuronal network coherence. Together, these results demonstrate the feasibility of using whole-brain computational modeling in assessing GD.