Elucidating Distinct and Common fMRI-Complexity Patterns in Preadolescent Children With Attention-Deficit/Hyperactivity Disorder, Oppositional Defiant Disorder, and Obsessive-Compulsive Disorder

<h2>ABSTRACT</h2><h3>Objective</h3> The pathophysiology of ADHD is complicated by high rates of psychiatric comorbidities, thus delineating unique versus shared functional brain perturbations is critical in elucidating illness pathophysiology. We investigated resting-state fMRI (rsfMRI)-complexity alterations among children with ADHD, oppositional defiant disorder (ODD), and obsessive-compulsive disorder (OCD), respectively, and comorbid ADHD, ODD, and OCD, within the cool and hot executive function (EF) networks. <h3>Method</h3> We leveraged baseline data from 9-10-year-old children in the Adolescent Brain and Cognitive Development (ABCD) Study. Children who singularly met all DSM-5 behavioral criteria for ADHD (<i>N</i> = 61), ODD (<i>N</i> = 38), and OCD (<i>N</i> = 48), respectively, were extracted, alongside children with comorbid ADHD, ODD, OCD, and/or other psychiatric diagnoses (<i>N</i> = 833). A control sample of age-, sex-, and developmentally-matched children was also extracted (<i>N</i> = 269). Voxel-wise sample entropy (SampEn) was computed using the LOFT Complexity Toolbox. Mean SampEn within all regions of interest (ROIs) of the EF networks was calculated for each participant. Hierarchical models with Generalized Estimating Equations compared SampEn of comorbid-free and comorbid ADHD, ODD, and OCD within the EF networks. <h3>Results</h3> SampEn was reduced in comorbid-free ADHD and ODD in overlapping regions of both EF networks than the healthy controls, including the bilateral superior frontal gyrus, anterior/posterior cingulate gyrus, and bilateral caudate (Wald statistic=5.682 to 10.798, <i>p</i> < 0.05, and Benjamini-Hochberg [BH] corrected), with ADHD additionally affected in the right inferior/middle frontal gyrus and bilateral frontal orbital cortex (Wald statistic=7.231 to 9.420, <i>p</i> < 0.05, and BH corrected). Among comorbid presentations, the presence of ADHD symptomatology was associated with significantly lower SampEn in every ROI (<i>z</i> = -3.973 to -2.235, <i>p</i> < 0.05, and BH corrected). <h3>Conclusion</h3> ADHD and ODD shared common impairments underlying the EF networks in the comorbid-free presentations, with ADHD showing more widespread complexity reduction. When ADHD co-occurred with other psychiatric disorders, the reduction in SampEn extended beyond the regions affected in comorbid-free ADHD, indicating that comorbidities amplify neural complexity deficits. In contrast, no significant SampEn alterations were observed in OCD, whether presented alone or in combination with ADHD.