Structural brain measures among children with and without ADHD in the Adolescent Brain and Cognitive Development Study cohort: a cross-sectional US population-based study
Joel Bernanke,Alex Luna,Le Chang,Elizabeth Bruno,Jordan Dworkin,Jonathan Posner
DOI: https://doi.org/10.1016/S2215-0366(21)00505-8
Abstract:Background: Structural neuroimaging research has identified a variety of abnormalities in cortical and subcortical structures in children with ADHD. However, studies to date have not employed large, non-referred samples, complete with data on potential confounding variables. Here, we tested for differences in structural MRI measures among children with and without ADHD using data from the Adolescent Brain and Cognitive Development (ABCD) Study, the largest paediatric brain imaging study in the USA. Methods: In this cross-sectional study, we used baseline demographic, clinical, and neuroimaging data from the ABCD Study, which recruited children aged 9-10 years between Sept 1, 2016, and Aug 31, 2018, representative of the sociodemographic features of the US population. ADHD was diagnosed by parent report of symptoms. Neuroimaging data underwent centralised quality control and processing by the ABCD team. Linear mixed effects models were used to estimate Cohen's d values associated with ADHD for 79 brain measures of cortical thickness, cortical area, and subcortical volume. We used a novel simulation strategy to assess the ability to detect significant effects despite potential diagnostic misclassification. Findings: Our sample included 10 736 participants (5592 boys, 5139 girls; 5692 White, 2165 Hispanic, 1543 Black, 221 Asian, and 1100 of other race or ethnicity), of whom, 949 met the criteria for ADHD and 9787 did not. In the full model, which included potential confounding variables selected a priori, we found only 11 significant differences across the 79 brain measures after false discovery rate correction, all indicating reductions in brain measures among participants with ADHD. Cohen's d values were small, ranging from -0·11 to -0·06, and were not meaningfully changed by using a more restrictive comparison group or alternative diagnostic methods. Simulations indicated adequate statistical power to detect differences even if there was substantial diagnostic misclassification. Interpretation: In a sample representative of the general population, children aged 9-10 years with ADHD differed only modestly on structural brain measures from their unaffected peers. Future studies might need to incorporate other MRI modalities, novel statistical approaches, or alternative diagnostic classifications, particularly for research aimed at developing ADHD diagnostic biomarkers. Funding: Edwin S Webster Foundation and Duke University, NC, USA.