Multiple comparison corrections were applied for the analyses. For power analyses involving whole-brain voxel-wise comparisons, we used Monte Carlo simulations (with actual Gaussian filter widths extracted from the data) to derive the corrected threshold (p < 0.05): voxel level p < 0.005 (one-tailed) over 11 contiguous voxels. As for connectivity analyses, the 3 × 3 connectivity matrix resulted in six comparisons for each hypothesis testing, for which we applied the false discovery rate (FDR) criterion (FDR p < 0.05). Lastly, for clinical association analyses, we conducted confirmatory correlation analyses constrained to regions demonstrating main effects, for which correction was not applied. We also applied a whole-brain regression of α power on hypervigilance scores, followed by Monte Carlo multiple comparison correction. While the sensory hypothesis implicates a direct association between sensory disinhibition and hypervigilance symptoms, it is possible that this sensory pathology also contributes to other PTSD symptom clusters. However, correlations between reduced source-level α power and total PCL scores (S-RS/M-RS rs > −0.10/−0.14, ps > 0.41/0.25) or other subscale scores (rs > −0.16; ps > 0.19) indicate rather weak effects. We thus focused on the hypervigilance symptoms below. Trend-level and non-corrected effects will be reported but not further discussed.
