ABSTRACT
Reports of regional grey matter volume (GMV) anomalies in patients with gambling disorder (PGD) are inconsistent, which can be attributed to methodological disparity and inattention to individual variability. Voxel-based morphometry was used to compare GMV between 25 PGD and 25 healthy controls. Additionally, the study explored associations of interpretative bias (IB, the tendency to reinterpret gambling outcomes) and negative urgency (NU, the tendency to act rashly under negative affect) with GMV in patients. These measures were chosen based on their sound association with gambling disorder in related studies. GMV tests were corrected across the whole brain (using a combination of voxel and cluster-level thresholds for a clusterwise-equivalent p ≤ 0.05). GMV was smaller in PGD than in controls in the dorsomedial prefrontal cortex. In PGD, a stronger cognitive distortion (higher IB) was associated with reduced GMV in the dorsal anterior cingulate; and patients with higher levels of impulsivity (higher NU) presented reduced GMV in the right ventrolateral prefrontal cortex. These findings are consistent with recent studies exploring individual differences in GD. However, the area discriminating between groups showed no overlap with the ones associated with IB and NU. These traits are thus unlikely to be responsible for between-group GMV differences.
Acknowledgements
We thank Warren Tierney for his assistance in revising the English version of this article.
Notes
1. Specifically, we used the Unified Segmentation algorithm as implemented in SPM to isolate tissue probability maps (which has shown higher sensitivity and specificity than other packages; Kazemi & Noorizadeh, Citation2014), and applied processing parameters shown to improve sensitivity and spatial accuracy (Radua, Canales-Rodríguez, Pomarol-Clotet, & Salvador, Citation2014). Generating a study template from participants’ tissue probability maps using DARTEL also improves spatial accuracy, and the use of a 0.2 absolute threshold mask further ensures the restriction of analysis to GM tissue (Ashburner, Citation2007). Finally, we applied a multiple-comparison correction method with a significance threshold of p < 0.001 at the voxel level, to avoid false positive results.
2. In addition, we did find a significant association between NU and the MC gambling subscore. In contrast to SOGS, MC is highly sensitive to present-day symptoms and, specifically, to the current occurrence of cravings. Therefore, the lower correlation of NU with SOGS relative to MC gambling severity scores can be accounted for by differential sensitivity to separate aspects or timings of GD manifestations.