Brain development involves dynamic changes in structural and functional connectivity, crucial for cognitive functions. Disruptions can lead to neurodevelopmental disorders like ADHD. Structural connectivity refers to white matter tracts connecting brain regions, while functional connectivity measures correlated spontaneous activity fluctuations between regions. Structure-function coupling describes the relationship between these two. Prior cross-sectional studies suggested an inverse relationship between coupling strength and regional functional complexity, with higher coupling in sensory regions and lower coupling in higher-order executive function regions. While some cross-sectional and one longitudinal study examined age and structure-function coupling, understanding developmental trajectories, especially in ADHD, remains limited. This study aimed to investigate longitudinal changes in structure-function coupling in typically developing children and compare these changes to those in children with ADHD, hypothesizing increased coupling with age in higher-order cognitive regions in typical development and significant differences between groups.

Existing literature highlights the importance of both structural and functional connectivity for cognitive development. Cross-sectional studies revealed an inverse correlation between structure-function coupling and functional complexity, with stronger coupling in sensory areas than in higher-order cognitive regions. Age-related changes in functional and structural connectivity are similar within brain networks, suggesting structural maturation supports functional communication. While adult studies showed strong structure-function relationships in frontal, parietal, and cerebellar networks, childhood development of structure-function coupling is less understood. Two cross-sectional studies found no age effects in default mode, frontoparietal, or salience networks. One longitudinal study reported increased coupling in frontal regions associated with executive functions. Regarding ADHD, research has consistently demonstrated aberrant structural and functional network organization, mainly in higher-order cognitive and sensory regions. Limited longitudinal studies on structure-function coupling in ADHD highlight the need for further research to understand neural development in this population.

This study utilized data from the Rewiring of the Children’s Attention Project (NICAP), a longitudinal neuroimaging cohort of 175 children (84 typically developing, 91 with ADHD) aged 9-14, with up to three waves of MRI scans at ~18-month intervals. ADHD diagnosis was confirmed via parent and teacher reports and diagnostic interviews. After excluding scans with excessive head motion or poor quality, the final sample included 278 scans (131 ADHD, 147 control). Structural images (3D T1-weighted) and diffusion-weighted imaging (DWI) data were acquired. Resting-state fMRI data underwent standard preprocessing steps (motion correction, spatial smoothing, normalization). The multi-modal parcellation of human cerebral cortex (HCP-MMP) atlas (360 regions) was used to define regions of interest (ROIs). Functional connectivity (FC) matrices were calculated using Pearson correlation coefficients, and structural connectivity (SC) matrices were derived using tractography. Structure-function coupling for each region was calculated using Spearman's rank correlation between non-zero elements of SC and FC matrices. Developmental trajectories were analyzed using generalized additive mixed models (GAMM), incorporating age, coupling strength, sex, and medication as covariates. Model comparisons used likelihood ratio tests (LRT) and Akaike Information Criterion (AIC).

In typically developing children, age-related increases in structure-function coupling were observed across the cortex, including prefrontal, anterior cingulate, posterior cingulate, inferior parietal, medial temporal, and visual cortices. Higher-order cognitive regions showed increasing coupling from 9 to 14 years, while sensory/visual regions showed increases from 9 to 12 years followed by a plateau. Children with ADHD showed weaker coupling compared to controls in the left superior temporal gyrus, right inferior parietal cortex, and right medial prefrontal cortex. They also exhibited different developmental trajectories, showing increased coupling with age in bilateral inferior frontal gyrus, left medial prefrontal cortex, left superior parietal cortex, left precuneus, left inferior temporal cortex, right inferior parietal, right mid-cingulate, right medial temporal cortex, and right visual regions, unlike controls.

The findings demonstrate typical development of structure-function coupling across late childhood to mid-adolescence, particularly in regions supporting cognitive automation (DMN, SAL, FPN). Increased coupling in higher-order cognitive regions might reflect maturation and specialization. Early increases followed by plateaus in sensory/visual networks suggest earlier maturation. While correlation doesn't imply causation, the strong association between structural and functional connectivity suggests that white matter development supports functional communication, or vice-versa. Children with ADHD showed reduced coupling in regions associated with the salience and default mode networks (SAL, DMN), consistent with previous findings of abnormalities in these networks. The increased coupling with age in specific regions in ADHD may reflect compensatory mechanisms, which highlights the dynamic nature of ADHD. However, further research is needed to determine the precise causal relationship between coupling and behavioral deficits.

This longitudinal study provides evidence of the joint maturation of structural and functional brain connections in typical development and reveals deficient structure-function coupling patterns in children with ADHD. The findings suggest atypical development of coordinated white matter and functional connectivity in ADHD, offering insights into the neural underpinnings of the disorder. Future research should investigate causal relationships and explore the relationship between structure-function coupling and specific ADHD symptoms.

The study's relatively small sample size could limit the power to detect subtle effects. The cross-sectional nature of the study's design might limit some inferences about cause and effect. The use of medication in a subset of ADHD participants could potentially confound the results. Finally, a more detailed investigation into the specific subtypes of ADHD and their corresponding neural mechanisms would further refine our understanding of the disorder.