An Integrated Analysis of Neural Network Correlates of Categorical and Dimensional Models of Attention-Deficit/Hyperactivity Disorder.

BACKGROUND: Attention-deficit/hyperactivity disorder (ADHD) is a heterogeneous neurodevelopmental disorder, putatively induced by dissociable dysfunctional biobehavioral pathways. Here, we present a proof-of-concept study to parse ADHD-related heterogeneity in its underlying neurobiology by investigating functional connectivity across multiple brain networks to 1) disentangle categorical diagnosis-related effects from dimensional behavior-related effects and 2) functionally map these neural correlates to neurocognitive measures. METHODS: We identified functional connectivity abnormalities related to ADHD across 14 networks within a large resting-state functional magnetic resonance imaging dataset (n = 409; age = 17.5 ± 3.3 years). We tested these abnormalities for their association with the categorical ADHD diagnosis and with dimensional inattention and hyperactivity/impulsivity scores using a novel modeling framework, creating orthogonalized models. Next, we evaluated the relationship of these findings with neurocognitive measures (working memory, response inhibition, reaction time variability, reward sensitivity). RESULTS: Within the default mode network, we mainly observed categorical ADHD-related functional connectivity abnormalities, unrelated to neurocognitive measures. Clusters within the visual networks primarily related to dimensional scores of inattention and reaction time variability, while findings within the sensorimotor networks were mainly linked to hyperactivity/impulsivity and both reward sensitivity and working memory. Findings within the cerebellum network and salience network related to both categorical and dimensional ADHD measures and were linked to response inhibition and reaction time variability. CONCLUSIONS: This proof-of-concept study identified ADHD-related neural correlates across multiple functional networks, showing distinct categorical and dimensional mechanisms and their links to neurocognitive functioning.

The interaction between 5-HTTLPR and stress exposure influences connectivity of the executive control and default mode brain networks.

We recently reported that the serotonin transporter polymorphism 5-HTTLPR moderates the relation between stress exposure and attention-deficit/hyperactivity disorder (ADHD) severity. This gene-environment interaction (GxE) has been previously tied to the processing of emotional stimuli, which is increasingly recognized to be a key factor in ADHD-related impairment. The executive control and default mode brain networks play an important role in the regulation of emotion processing, and altered connectivity of these networks has also been associated with ADHD. We therefore investigated whether resting-state connectivity of either of these networks mediates the relation of 5-HTTLPR and stress exposure with ADHD severity. Resting-state functional magnetic resonance imaging, genetic, and stress exposure questionnaire data was available for 425 adolescents and young adults (average age 17.2 years). We found that 5-HTTLPR S-allele carriers showed a more negative relation between stress exposure and connectivity of the executive control network than L-allele homozygotes, specifically in the pre/postcentral gyrus, striatum, and frontal pole. In the default mode network, we found a positive association between the GxE and supramarginal gyrus connectivity. Connectivity of either network did not significantly mediate the effect of this GxE on ADHD. Opposite effects of stress exposure on connectivity in the executive and default mode networks may contribute to findings that stress exposure is associated with lowered cognitive control and heightened levels of rumination and worrying, for S-allele carriers but not L-allele homozygotes. When combined, these effects on connectivity of both networks may relate to the emotional problems seen in individuals with ADHD.

Attention-Deficit/Hyperactivity Disorder symptoms coincide with altered striatal connectivity.

BACKGROUND: Cortico-striatal network dysfunction in attention-deficit/hyperactivity disorder (ADHD) is generally investigated by comparing functional connectivity of the main striatal sub-regions (i.e., putamen, caudate, and nucleus accumbens) between an ADHD and a control group. However, dimensional analyses based on continuous symptom measures might help to parse the high phenotypic heterogeneity in ADHD. Here, we focus on functional segregation of regions in the striatum and investigate cortico-striatal networks using both categorical and dimensional measures of ADHD. METHODS: We computed whole-brain functional connectivity for six striatal sub-regions that resulted from a novel functional parcellation technique. We compared functional connectivity maps between adolescents with ADHD (N=169) and healthy controls (N=122), and investigated dimensional ADHD-related measures by relating striatal connectivity to ADHD symptom scores (N=444). Finally, we examined whether altered connectivity of striatal sub-regions related to motor and cognitive performance. RESULTS: We observed no case-control differences in functional connectivity patterns of the six striatal networks. In contrast, inattention and hyperactivity/impulsivity symptom scores were associated with increases in functional connectivity in the networks of posterior putamen and ventral caudate. Increased connectivity of posterior putamen with motor cortex and cerebellum was associated with decreased motor performance. CONCLUSIONS: Our findings support hypotheses of cortico-striatal network dysfunction in ADHD by demonstrating that dimensional symptom measures are associated with changes in functional connectivity. These changes were not detected by categorical ADHD versus control group analyses, highlighting the important contribution of dimensional analyses to investigating the neurobiology of ADHD.

Evaluation of ICA-AROMA and alternative strategies for motion artifact removal in resting state fMRI.

We proposed ICA-AROMA as a strategy for the removal of motion-related artifacts from fMRI data (Pruim et al., 2015). ICA-AROMA automatically identifies and subsequently removes data-driven derived components that represent motion-related artifacts. Here we present an extensive evaluation of ICA-AROMA by comparing our strategy to a range of alternative strategies for motion-related artifact removal: (i) no secondary motion correction, (ii) extensive nuisance regression utilizing 6 or (iii) 24 realignment parameters, (iv) spike regression (Satterthwaite et al., 2013a), (v) motion scrubbing (Power et al., 2012), (vi) aCompCor (Behzadi et al., 2007; Muschelli et al., 2014), (vii) SOCK (Bhaganagarapu et al., 2013), and (viii) ICA-FIX (Griffanti et al., 2014; Salimi-Khorshidi et al., 2014), without re-training the classifier. Using three different functional connectivity analysis approaches and four different multi-subject resting-state fMRI datasets, we assessed all strategies regarding their potential to remove motion artifacts, ability to preserve signal of interest, and induced loss in temporal degrees of freedom (tDoF). Results demonstrated that ICA-AROMA, spike regression, scrubbing, and ICA-FIX similarly minimized the impact of motion on functional connectivity metrics. However, both ICA-AROMA and ICA-FIX resulted in significantly improved resting-state network reproducibility and decreased loss in tDoF compared to spike regression and scrubbing. In comparison to ICA-FIX, ICA-AROMA yielded improved preservation of signal of interest across all datasets. These results demonstrate that ICA-AROMA is an effective strategy for removing motion-related artifacts from rfMRI data. Our robust and generalizable strategy avoids the need for censoring fMRI data and reduces motion-induced signal variations in fMRI data, while preserving signal of interest and increasing the reproducibility of functional connectivity metrics. In addition, ICA-AROMA preserves the temporal non-artifactual time-series characteristics and limits the loss in tDoF, thereby increasing statistical power at both the subject- and the between-subject analysis level.

ICA-AROMA: A robust ICA-based strategy for removing motion artifacts from fMRI data.

Head motion during functional MRI (fMRI) scanning can induce spurious findings and/or harm detection of true effects. Solutions have been proposed, including deleting ('scrubbing') or regressing out ('spike regression') motion volumes from fMRI time-series. These strategies remove motion-induced signal variations at the cost of destroying the autocorrelation structure of the fMRI time-series and reducing temporal degrees of freedom. ICA-based fMRI denoising strategies overcome these drawbacks but typically require re-training of a classifier, needing manual labeling of derived components (e.g. ICA-FIX; Salimi-Khorshidi et al. (2014)). Here, we propose an ICA-based strategy for Automatic Removal of Motion Artifacts (ICA-AROMA) that uses a small (n=4), but robust set of theoretically motivated temporal and spatial features. Our strategy does not require classifier re-training, retains the data's autocorrelation structure and largely preserves temporal degrees of freedom. We describe ICA-AROMA, its implementation, and initial validation. ICA-AROMA identified motion components with high accuracy and robustness as illustrated by leave-N-out cross-validation. We additionally validated ICA-AROMA in resting-state (100 participants) and task-based fMRI data (118 participants). Our approach removed (motion-related) spurious noise from both rfMRI and task-based fMRI data to larger extent than regression using 24 motion parameters or spike regression. Furthermore, ICA-AROMA increased sensitivity to group-level activation. Our results show that ICA-AROMA effectively reduces motion-induced signal variations in fMRI data, is applicable across datasets without requiring classifier re-training, and preserves the temporal characteristics of the fMRI data.