Dynamic Inter-Modality Source Coupling Reveals Sex Differences in Children based on Brain Structural-Functional Network Connectivity: A Multimodal MRI Study of the ABCD Dataset

Aline Kotoski, Sir-Lord Wiafe, Julia M. Stephen, Yu‐Ping Wang, Tony W. Wilson, Vince D. CalhounbioRxiv 2025 · DOI

TL;DR

This paper introduces a novel method, dynamic inter-modality source coupling (dIMSC), to analyze how brain structure and function interact dynamically over time. They found distinct patterns of this coupling between males and females in children, suggesting sex-specific brain organization during development.

What Problem Does This Paper Address?

While we know that male and female brains differ in structure and function, especially during development, it's been challenging to understand how these differences manifest in the dynamic interplay between brain anatomy and its activity. Previous studies often looked at structure and function in isolation. This paper aims to bridge that gap by investigating the transient coupling between dynamic functional network connectivity (dFNC) and structural MRI (sMRI) gray matter volume, specifically looking for sex differences in this coupling during childhood.

Methods

The core innovation here is the dynamic inter-modality source coupling (dIMSC) method. It builds upon their previous work (IMSC) by incorporating the temporal dimension. Here's how it works:

  • Structural MRI (sMRI): They used Source-Based Morphometry (SBM) to identify spatially independent gray matter sources, essentially creating a structural "fingerprint" for each individual.
  • Resting-State fMRI (rs-fMRI): They computed dynamic functional network connectivity (dFNC) using a sliding window approach. This captures how brain network connections change over time, offering a more nuanced view than static connectivity.
  • dIMSC Calculation: For each subject, they calculated the cross-correlation between the dFNC matrix (representing dynamic functional states) and the SBM vector (representing structural patterns) at each time point. This results in a time-resolved measure of how strongly structure and function are coupled.
  • Categorization: These coupling values were then categorized into positive (structure and function increase together), neutral (weak coupling), and negative (inverse relationship).
  • Sex Differences Analysis: They used two-sample t-tests with corrections for multiple comparisons to identify significant sex differences in these coupling patterns across brain regions.

Datasets

The study utilized data from the Adolescent Brain Cognitive Development (ABCD) study. This is a large, longitudinal study of child development. For their analysis, they selected a random subset of 500 unrelated individuals, matched for age, to manage computational demands.

Key Insights and Ideas

  • Dynamic Coupling Matters: The study highlights that brain structure-function relationships aren't static; they fluctuate over time, and these dynamics can reveal important differences between individuals and groups.
  • Sex-Specific Patterns Emerge: The dIMSC method successfully identified distinct patterns of structure-function coupling between males and females in children, suggesting that biological sex influences how brain anatomy and activity are integrated.
  • Different Regions, Different Dynamics: Sex differences were not uniform across the brain. Specific regions showed stronger coupling in one sex over the other, depending on whether the coupling was positive, neutral, or negative. This points to sex-specific organizational principles in brain development.
  • Beyond Static Views: By moving beyond static analyses, dIMSC provides a more sensitive tool for understanding the complex, time-dependent interplay in the developing brain.

Key Results

  • Overall Coupling: The majority of timepoints (around 47%) fell into the "neutral" coupling category, indicating periods of weak structure-function alignment. Positive and negative coupling each accounted for about 27% of timepoints.
  • Positive Coupling Differences:
    • Females showed stronger positive coupling in the inferior parietal lobule and precuneus. These regions are involved in multimodal integration, attention, and self-referential processing.
  • Males showed stronger positive coupling in the postcentral gyrus and precuneus, areas related to sensorimotor integration.
  • Neutral Coupling Differences:
    • Males showed stronger neutral coupling in the cerebellum and superior temporal gyrus.
  • Females showed stronger neutral coupling in the caudate, precentral gyrus, and middle frontal gyrus.
  • Negative Coupling Differences:
    • Males showed stronger negative coupling in the paracentral lobule, superior parietal lobule, and middle temporal gyrus.
  • Females showed stronger negative coupling in the superior temporal gyrus and inferior parietal lobule.

Limitations and Future Work

  • Behavioral Link: The authors acknowledge that their current analysis does not directly link coupling strength to specific behaviors or cognitive performance. Future work should incorporate ABCD cognitive and behavioral measures to clarify the functional significance of these observed differences.
  • Interpretation of Neutral Coupling: While neutral coupling is characterized by weak correlation, the consistent differences observed between sexes in this state might suggest a form of "structural anchoring" or stabilizing influence. However, this interpretation is speculative.
  • Longitudinal Studies: The authors suggest that future longitudinal studies are needed to determine if these coupling patterns are transient or stable features of development.
  • Environmental/Genetic Factors: Individual variability in coupling strength might be influenced by environmental, genetic, or hormonal factors, which could be explored in future research.
  • Methodological Extensions: The dIMSC approach is presented as a powerful framework that can be extended to investigate individual trajectories in neurodevelopmental conditions, especially those with known sex biases (e.g., ADHD, autism).

Summary updated

Abstract

Together these findings suggest distinct coupling in brain structure-function coupling between sexes, potentially reflecting sex-specific organization of functional networks and their structural substrates. The dIMSC method advances our earlier work by enabling time-resolved analysis of brain structure-function coupling, providing a powerful framework for investigating neurodevelopmental processes.
BibTeX
@misc{kotoski2025dynamic,
  author = {Aline Kotoski and Sir-Lord Wiafe and Julia M. Stephen and Yu‐Ping Wang and Tony W. Wilson and Vince D. Calhoun},
  title = {{Dynamic Inter-Modality Source Coupling Reveals Sex Differences in Children based on Brain Structural-Functional Network Connectivity: A Multimodal MRI Study of the ABCD Dataset}},
  year = {2025},
  doi = {10.1101/2025.07.23.666366},
  howpublished = {bioRxiv},
}
Figure