428. ADHD: neurocognitive pathways
Abstract Background ADHD is characterized by its clinical heterogeneity in symptoms and their trajectories. However, both cognitive impairment and motivational dysfunction in the well-known ‘dual-pathway model for ADHD’ can be absent in about 30% of ADHD cases and cannot predict the persistence or remitting of the symptoms. Aims & Objectives To explore neurocognitive pathways to ADHD that contribute to the heterogeneity in its clinical symptom and the underlying neuroimaging alterations. Method We analyzed the data from more than 10,000 participants in two community-based, longitudinal cohorts of adolescents (i.e., ABCD and IMAGEN) and three independent clinical cohorts of ADHD (i.e., ADHD-200, ADHD-1000, and ADHD-Shanghai). Multidimensional data were collected in these cohorts, including the genetic, neuroimaging, behavioural, symptomy, etc. We characterized different ADHD symptom trajectories during adolescence and assessed their associations with the functional traits, neurocognitive performances, emotion regulation traits, etc. We tested whether the different symptom trajectories were related to distinct neuroimaging signatures. Using the longitudinal measurements, we established the neurocognitive pathways to ADHD by identifying mediation effects of neurocognitive functions on the associations between neuroimaging signatures and ADHD symptom trajectories. Furthermore, we built prediction models for ADHD symptom using neuroimaging measures, cognitive performances, and genetic variants. Finally, we tested the generalizability of these models to clinical cohorts. Results We found that the cognitive (working memory and attention) and motivational (Kirby rate) measures were associated with ADHD symptoms in adolescents. The ADHD symptom and these two measures were associated with a common neuroimaging feature, i.e., the gray matter volume of a posterior occipital cluster. In clinical samples, this volumetric difference was most evident in ADHD patients without treatment but this difference compared with typical developing controls reduced after treatment. We found that emotion dysregulation was associated with the ADHD symptom (partial eta2 = 0.21) after controlling for both cognitive and motivational deficits. Emotion dysregulation mediated 35.54% of the association between the right pars orbitalis at the baseline and greater ADHD symptoms at 1-year follow-up, and this mediation pathway was separable from those for the cognitive and motivational deficits. Compared with the model using both the cognitive and the motivational pathways, the new model including the emotional dysregulation pathway significantly improved the prediction accuracy of future symptom persistence or remission (AUC=0.80, 95% CI = [0.71, 0.85]). Longitudinally, we found that the symptom remission was associated with faster hippocampal volumetric expansion. This association enhanced the prediction of future symptom in the ABCD cohort (R2=0.114), was both replicated in an independent adolescent cohort (i.e., the IMAGEN cohort) and observable in the two clinical cohorts (i.e., ADHD-200 and ADHD-1000). Discussion & Conclusions These findings highlight emotion dysregulation and its associated right pars orbitalis forming a neurocognitive pathway to ADHD, and reveal that the expansion of the hippocampus as a potential target for facilitating symptom remission in ADHD.