Social profiles among youth with attention-deficit/hyperactivity disorder (ADHD): Evidence from the ABCD study

Abstract

Social functioning difficulties among youth with attention-deficit/hyperactivity disorder (ADHD) have been examined behaviorally; however, limited research has investigated brain networks associated with social difficulties among youth with ADHD. A growing body of literature supports the utility of the NIMH's Research Domain Criteria (RDoC) framework, which emphasizes broad neurobiological based dimensions, allowing for the integration of models of both neural circuitry and behavior when examining externalizing behaviors in youth. We hypothesized that an ADHD classification system based on social functioning would better predict real-world psychosocial and academic outcomes compared to traditional Diagnostic and Statistical Manual of Mental Disorders (DSM-5) nosology of ADHD presentations. First, using data from the Adolescent Brain Cognitive Development (ABCD) Study, we identified four distinct profiles of youth with ADHD ranging from low social functioning to high social functioning. These social-data-derived profiles were linked to differential social challenges associated with caregiver income and mental health disorders. Next, our neuroimaging findings initially revealed differential patterns of functional connectivity across profiles involving attention-control, cingulo-opercular, sensorimotor networks. However, these connectivity differences were not consistently replicated, indicating that social functioning alone may not define neurobiologically distinct subgroups. Finally, in comparing our social functioning profiles to existing DSM-5 nosology with respect to real-world psychosocial outcomes, our social profiles demonstrated greater explanatory power for outcomes related to peer relationships, family conflict, and mental health. Overall, these findings emphasize the heterogeneity in social functioning among ADHD youth and suggest that while behavioral profiles are clinically meaningful, future work should integrate additional dimensions, such as executive functioning, to more precisely capture the neurobiological underpinnings of ADHD.

Keywords: Attention-deficit/hyperactivity disorder; Brain networks; Research domain criteria; Resting state functional, connectivity; Social functioning.

Fig. 1

Analytic Workflow. Participants were selected from the ongoing ABCD Study via items from a behavioral measure (i.e., KSADS-5) to identify ADHD and TD groups (Step 1). Select items related to social functioning were extracted from behavioral questionnaires including the KSADS-5, CBCL, Parent Short-SRS, Youth- and Parent- prosocial behavior survey, and FES - family conflict subscale (Step 2). Social items were then annotated as pertaining to either RDoC social processes domain constructs (i.e., Affiliation & Attachment, Social Communication, Perception and Understanding of Self, Perception and Understanding of Others) or real-world psychosocial and academic outcomes (Step 3). A latent profile analysis was conducted to identify data-driven phenotypic profiles of ADHD youth with social functioning impairment. Then, profile differences in rsFC data were examined utilizing the Gordon parcellation network in relation to four social profiles of youth (Step 4). Finally, eight separate linear regression models were conducted to compare the clinical utility and concurrent validity of data-driven profiles to DSM-5 nosology on outcomes related to peer relationships, family conflict, mental health, and academic outcomes (Step 5).

Fig. 2

Social Functioning Impairment Across Profiles. Latent profile analysis revealed four profiles of social functioning among ADHD youth in the ABCD Study. These four profiles show distinct levels of social functioning among the domains of Affiliation and Attachment, Social Communication, and Perception and Understanding of Others. Higher scores indicate higher levels of impairment among social domains. Profile 1 (Blue) corresponded to Low Social Functioning; Profile 2 (Green) corresponded to Moderate Social Functioning; Profile 3 (White) corresponded to Highly Perceptive, reflecting high overall social functioning with particular strengths in Perception and Understanding of Others; and Profile 4 (Red) corresponded to Highly Connected, reflecting high overall social functioning with particular strengths in Affiliation and Attachment and Social Communication. The number of dots for each group is due to jittering applied to avoid overlap in the visualization; this does not reflect an unequal distribution of participants across groups.

Fig. 3

Standardized DAN and SMH Within-Network Connectivity. Estimates of standardized within-network connectivity (95 % confidence intervals) for the Dorsal Attention Network (DAN) and Sensorimotor-Hand Network (SMH) across social functioning profiles of ADHD youth compared to typically developing controls. TD = Typically Developing controls. Profile 1 = Low Social Functioning (blue); Profile 2 = Moderate Social Functioning (green); Profile 3 = Highly Perceptive (white/gray); Profile 4 = Highly Connected (red). Significant estimates are denoted with asterisks: **.

Fig. 4

Standardized Between-Network Connectivity. Estimates of standardized between-network connectivity (95 % confidence intervals) across social functioning profiles of ADHD youth compared to typically developing controls. Panel A = Auditory and Salience; Panel B = Dorsal Attention and Salience; Panel C = Default Mode and Sensorimotor-Hand; Panel D = Auditory and None; Panel E = Cingulo-Opercular and None. TD = Typically Developing controls. Profile 1 = Low Social Functioning (blue); Profile 2 = Moderate Social Functioning (green); Profile 3 = Highly Perceptive (white/gray); Profile 4 = Highly Connected (red). Significant estimates are denoted with asterisks: **.

Fig. 5

Comparison of proportion of variance explained between data-driven social profiles and traditional DSM-5 nosology to psychosocial and academic outcomes (Peer Relationships, Family Conflict, Mental Health, Academic Performance).