Transdiagnostic Symptom Domains Have Distinct Patterns of Association With Head Motion During Multimodal Imaging in Children

Abstract

Background: It is unclear how transdiagnostic symptoms including attention, disruptive behavior, and internalizing problems are linked to in-scanner motion in children across structural and functional magnetic resonance imaging (fMRI). In the current study, we examined whether transdiagnostic symptoms of attention, disruptive behavior, and internalizing problems were associated with scanner motion in children during multimodal imaging.

Methods: In 9045 children ages 9 to 10 years in the ABCD (Adolescent Brain Cognitive Development) Study, logistic regression and linear mixed-effects models were used to examine associations between motion and behavior. Motion was indexed using ABCD Study quality control (QC) metrics and mean framewise displacement for T1- and T2-weighted structural, resting-state, and diffusion MRI; stop-signal task; monetary incentive delay task; and emotional n-back task. The Child Behavior Checklist was used as a continuous measure of symptom severity.

Results: Greater severity of attention and disruptive behavior problems was associated with a lower likelihood of passing motion QC across imaging modalities, while increased internalizing severity was associated with a higher likelihood of passing. There was also an interaction between sex and attention-related problems in passing QC for T2-weighted and diffusion MRI scans. Increased attention and disruptive behavior problems were associated with increased mean motion, whereas increased internalizing problems were associated with decreased mean motion. Greater severity of attention problems was associated with worse performance across the fMRI tasks.

Conclusions: These findings have implications for advancing the development of computational and behavioral approaches for mitigating motion effects in youths, enhancing accessibility of imaging protocols and representativeness influences across child psychiatric disorders, and identifying brain-based biomarkers.

Keywords: Development; Diffusion MRI; MRI; Motion; Multimodal imaging; Resting-state.

Figure 1

Flowchart illustrating data structure and participants from the ABCD (Adolescent Brain Cognitive Development) Study dataset. Analyses were conducted for each of the imaging modalities: T1- and T2-weighted structural magnetic resonance imaging (MRI), resting-state functional MRI, diffusion MRI (dMRI), stop signal task (SST), monetary incentive delay (MID) task, and the emotional version of the n-back task (EN-back).

Figure 2

Correlations among study variables. Pearson correlations and significance values are shown for age, Child Behavior Checklist scales (Attention, Externalizing, and Internalizing Problems subscale scores), cognition (IQ), motion indexed by mean framewise displacement (mm), and sex assigned at birth. The matrix values indicate the correlation coefficient (top) and the corresponding p values (bottom). The strength and direction of the correlations are color coded, with red representing positive correlations and blue representing negative correlations. Statistically significant correlations are denoted by asterisks: ∗p < .05, ∗∗p < .01, ∗∗∗p < .001.

Figure 3

Severity of transdiagnostic symptom domains in youths are associated with motion during functional and structural imaging, accounting for the effect of scanner manufacturer. Scatterplots depict results of linear mixed-effects models (see Table 3) for Child Behavior Checklist Attention, Externalizing, and Internalizing Problems scores for resting-state functional magnetic resonance imaging (rsfMRI) (A), diffusion MRI (dMRI) (B), the stop signal task (SST) (C), the monetary incentive delay (MID) task (D), and the emotional version of the n-back task (EN-back) (E). The red trendline represents the regression line based on the linear mixed-effects models fit. Statistical significance is denoted by asterisks: ∗p < .05, ∗∗p < .01, ∗∗∗p < .001.

Figure 4

Attention problems were associated with behavioral performance on the stop signal task (SST), monetary incentive delay (MID) task, and the emotional version of the n-back (EN-back) task in youths. Scatterplots depict results of linear mixed-effects models (see Table 4) for Child Behavior Checklist Attention, Externalizing, and Internalizing Problem scores for task-based functional magnetic resonance imaging: SST (A), MID task (B), and the EN-back task for the 0-back (C) and 2-back (D) rounds, respectively. The trendline represents the regression line based on the linear mixed-effects model fit. Statistical significance is denoted by asterisks: ∗p < .05, ∗∗p < .01, ∗∗∗p < .001.