Developmental aspects of ADHD

It is increasingly recognized that ADHD is a heterogeneous disorder, both in its clinical presentation (phenotype) and the underlying aetiology. This heterogeneity makes it difficult to identify causal pathways that link the phenotype to brain structure and functioning. In an attempt to go beyond the current categorical definition of the disorder, there is an increasing interest in a dimensional approach to ADHD. Yet an issue that is often overlooked is that a dimensional approach requires an understanding of neurodevelopmental changes in the typical population, in order to understand how deviations in development may lead to ADHD symptoms. The work in this thesis describes the neurodevelopmental context in which ADHD occurs. We investigated both aspects relevant to the interpretation of developmental imaging data, such as the context in which developmental changes are assessed (chapter 2) and the developmental overlap between different cognitive functions (chapter 4), but also aspects relevant to the ontology of ADHD and development: which specific functions underlie the observed cognitive changes in development (chapter 3 and 4), and how specific are these to ADHD (chapter 5)? Results show that the context in which developmental changes are assessed matters: we found that improved task performance over age is reflected in network activity during a task, but not in rest, and, vice versa, that developmental changes in network activity during rest are unrelated to task performance. This may seem obvious to an outsider, yet the majority of studies of developmental changes in neural connectivity rely on resting-state data. The results of these studies should therefore be interpreted with caution. (chapter2). Another important factor is the level of description of the cognitive functions under investigation: as chapter 4 and 5 show, intra-individual response variability (IIV) is an important predictor of performance on a response inhibition task in typically developing subjects, and related to age related increase in the dorsal anterior cingulate (dACG) in this group. This relation was absent in the ADHD group. Children with ADHD additionally show marked differences in the developmental changes in IIV, both compared to typically developing subjects, and compared to children with Autism Spectrum Disorders (ASD), a developmental disorder that shares symptoms of ADHD. These findings suggest that IIV is a good candidate for a mechanism to link neural development to response inhibition. Lastly, the results presented in chapter 3show that neuroimaging data can equally inform theory: this study shows that there are different neural networks related to proactive and reactive inhibition, in addition to a set of common networks. This finding confirms he idea that these concepts capture the underlying neural organization in more detail than the construct of response inhibition as a whole. In sum, we show that the contexts in which behavioural changes are measured, as well as the operationalization of constructs that reflect developmental changes, affect the interpretation of neurodevelopmental imaging studies. Given the evidence that ADHD reflects extreme behaviour on a dimensional scale, future studies should aim to map developmental pathways that lead to ADHD-like behaviour across the population