A transdiagnostic data-driven study of children's behaviour and the functional connectome

Abstract

Behavioural difficulties are seen as hallmarks of many neurodevelopmental conditions. Differences in functional brain organisation have been observed in these conditions, but little is known about how they are related to a child's profile of behavioural difficulties. We investigated whether behavioural difficulties are associated with how the brain is functionally organised in an intentionally heterogeneous and transdiagnostic sample of 957 children aged 5-15. We used consensus community detection to derive data-driven profiles of behavioural difficulties and constructed functional connectomes from a subset of 238 children with resting-state functional Magnetic Resonance Imaging (fMRI) data. We identified three distinct profiles of behaviour that were characterised by principal difficulties with hot executive function, cool executive function, and learning. Global organisation of the functional connectome did not differ between the groups, but multivariate patterns of connectivity at the level of Intrinsic Connectivity Networks (ICNs), nodes, and hubs significantly predicted group membership in held-out data. Fronto-parietal connector hubs were under-connected in all groups relative to a comparison sample and children with hot vs cool executive function difficulties were distinguished by connectivity in ICNs associated with cognitive control, emotion processing, and social cognition. This demonstrates both general and specific neurodevelopmental risk factors in the functional connectome.

Keywords: Behavioral difficulties; Executive function; Functional connectivity; Neurodevelopment.

Fig. 1

Results of the consensus community detection. The behavioural network (left) shows the child-by-child correlation matrix across the six scales of the Conners 3 after sorting by the community assignments: C1 (n = 339), C2 (n = 260), and C3 (n = 178). The behavioural profiles for the three communities and Controls in the whole sample (top right) and MRI sample (bottom right). Points indicate mean scores and error bars denote 95% confidence intervals from 1000 bootstrapped samples. The dashed horizontal line indicates the threshold for elevated scores and the solid line indicates very elevated scores. Note: Executive Function (EF), Hyperactivity/Impulsivity (Hyperactivity), Learning Problems (Learning).

Fig. 2

PLS components of ICN connectivity that predicted group membership. The heatmaps (top) show the 25% largest loadings of ICN connections onto each component relative to their standard error over 1000 bootstrapped samples. The boxplots (bottom) show the bootstrapped distribution of component scores for each group and significant group differences assessed by a permutation test. ICN abbreviations: Somatomotor (SomMot), dorsal attention network (DAN), ventral attention network (VAN), fronto-parietal (Control), and default mode network (DMN). ***p < 0.001, *p < 0.05.

Fig. 3

PLS components of nodal strength that predicted group membership. The brain plots (left) show the 5% largest loadings of nodes onto each component (PLS1 top, PLS2 bottom) relative to their standard error over 1000 bootstrapped samples. The size of the node is proportional to its absolute loading and the colour corresponds to its ICN: default-mode (red), limbic (cream), visual (purple), fronto-parietal (orange), and subcortical (black). The boxplots (right) show the bootstrapped distribution of component scores for each group and significant group differences assessed by a permutation test. ***p < 0.001, **p < 0.01, ⁺p < 0.06. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

PLS components of connector hub strength that predicted group membership. The brain plots (left) show the 30% largest loadings of connector hubs onto each component (PLS1 top, PLS2 bottom) relative to their standard error over 1000 bootstrapped samples. The size of the node is proportional to its absolute loading and the colour corresponds to its ICN: fronto-parietal (orange), subcortical (black), dorsal attention (green), and limbic (cream). The boxplots (right) show the bootstrapped distribution of component scores for each group and significant group differences assessed by a permutation test. * *p < 0.01, *p < 0.05. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

PLS components of provincial hub strength that predicted group membership. The brain plots (left) show the 10% largest loadings of provincial hubs onto each component (PLS1 top, PLS2 bottom) relative to their standard error over 1000 bootstrapped samples. The size of the node is proportional to its absolute loading and the colour corresponds to its ICN: default-mode (red), visual (purple), limbic (cream), fronto-parietal (orange), and somatomotor (blue). The boxplots (right) show the bootstrapped distribution of component scores for each group and significant group differences assessed by a permutation test. **p < 0.01, *p < 0.05, ⁺p < 0.06. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)