A neural substrate for sensory over-responsivity defined by exogenous and endogenous brain systems

Hannah L. Choi, Maia C. Lazerwitz, R.a. Powers, Mikaela A. Rowe, Jamie Wren-Jarvis, Amir Sadikov, Lanya T. Cai, Robyn Chu, LaShelle Rullan, Kaitlyn J. Trimarchi, Rafael D. Garcia, Elysa J. Marco, Pratik MukherjeeJournal of Neurodevelopmental Disorders 2025 · DOI

TL;DR

This study identifies distinct patterns of brain connectivity in neurodiverse children with sensory over-responsivity (SOR). They found that SOR is associated with reduced connectivity in "exogenous" (sensory input) brain systems and increased connectivity in "endogenous" (higher-order processing) systems, a pattern that is crucial for behavioral resilience.

What Problem Does This Paper Address?

Understanding the neural underpinnings of neurodevelopmental disorders (NDDs) is challenging due to the heterogeneity within these populations. Sensory over-responsivity (SOR), a common symptom in conditions like autism and ADHD, presents a unique opportunity to study how sensory processing interacts with broader brain network function. While previous research has hinted at altered brain connectivity in SOR, a clear picture of how different brain systems (those processing external sensory information vs. internal cognitive processes) are affected, and how this relates to behavioral outcomes, has been missing. This paper aims to bridge this gap by investigating the functional connectivity of these two distinct brain systems in neurodiverse children with and without SOR.

Methods

The researchers employed a multimodal neuroimaging approach using resting-state functional MRI (rs-fMRI) on 83 neurodiverse children aged 8-12. They used Independent Component Analysis (ICA) to identify 14 functional connectivity networks (FCNs) and then categorized these into "exogenous" (e.g., visual, somatomotor) and "endogenous" (e.g., central executive, default mode) systems. To measure connectivity, they used:

  • Dual Regression (DR): To assess long-range functional connectivity (FC) between brain regions.
  • Fractional Amplitude of Low-Frequency Fluctuations (fALFF): To measure local brain activity and connectivity.
  • Regional Homogeneity (ReHo): To assess the synchronization of brain activity within local regions.

Additionally, they used the Behavioral Assessment Scale for Children (BASC-3) and latent profile analysis to categorize children into "resilient" or "dysregulated" behavioral groups. This allowed them to examine how SOR and behavioral regulation interact with brain connectivity.

Datasets

The study utilized data from 83 neurodiverse children recruited from a community neurodevelopment center. These children underwent structural and functional MRI scans. The data was collected as part of a larger study protocol. No new benchmarks were introduced; rather, existing datasets and assessment tools were used.

Key Insights and Ideas

  • The Brain as Two Interacting Systems: The paper proposes a useful framework of "exogenous" and "endogenous" brain systems, providing a clear way to categorize and understand the complex network interactions in the brain.
  • SOR as a "Double Dissociation" Driver: The most striking finding is the "double dissociation" in local connectivity: SOR children show reduced exogenous FC and elevated endogenous FC, while non-SOR children show the opposite. This suggests SOR fundamentally alters the balance between sensory input processing and higher-level cognitive control.
  • Resilience is Tied to a Specific Connectivity Contrast: This double dissociation is particularly pronounced in children who are behaviorally resilient. This implies that a specific balance between these two brain systems is not just a consequence of SOR, but a mechanism that allows for adaptive functioning despite sensory challenges.
  • Beyond Simple Hypoconnectivity: While reduced long-range connectivity in SOR was observed (as expected from prior work), the local connectivity findings reveal a more nuanced picture of increased endogenous connectivity, suggesting compensatory mechanisms at play.

Key Results

  • Reduced Long-Range Exogenous FC: Neurodiverse children with SOR generally showed reduced long-range functional connectivity in exogenous brain systems.
  • Local FC Double Dissociation: In terms of local connectivity (fALFF and ReHo), a clear double dissociation was observed:
  • ND/SOR children: Reduced exogenous FC and elevated endogenous FC.
  • ND/NO-SOR children: Elevated exogenous FC and reduced endogenous FC.
  • Resilience Specificity: This double dissociation was most evident in behaviorally resilient children. Dysregulated children showed a different, less distinct pattern.
  • Classification Potential: Machine learning models using fALFF features achieved high accuracy (up to 87%) in distinguishing between SOR and non-SOR children, highlighting the potential for neuroimaging biomarkers.

Limitations and Future Work

  • Cross-Sectional Design: The study is cross-sectional, meaning it captures a snapshot in time. Longitudinal studies are needed to understand how these connectivity patterns emerge and evolve over development and how they relate to treatment outcomes.
  • Sample Size for Autism Subgroup: Children meeting autism criteria were excluded due to small sample size, limiting direct comparisons to this specific population.
  • Generalizability: While the findings are significant, further research in larger and more diverse cohorts is needed to confirm generalizability.
  • Mechanisms of Resilience: The study highlights the importance of the exogenous-endogenous balance for resilience but doesn't fully elucidate the precise mechanisms. Future work could explore how this balance is achieved and maintained.
  • Clinical Application: The authors suggest exploring the interaction of SPD and ADHD, investigating the role of autonomic nervous system function, and utilizing advanced MRI techniques for higher-resolution imaging. They also propose interventional trials to assess if remediating SOR can alter these connectivity patterns.

Summary updated

Abstract

Abstract Background Exogenous (outward-directed) and endogenous (inward-directed) neural systems are essential for cognition and behavior. However, how they are altered in neurodiverse (ND) children remains unanswered in part due to heterogeneity. Sensory over-responsivity (SOR), the most prevalent form of sensory processing disorder (SPD), serves as a quintessential paradigm for investigating the interaction between exogenous and endogenous brain networks given that both basic and higher-order sensory processing are substantially implicated in this condition. Methods Neurodiverse children ages 8–12 years old (n = 83; 30 females and 53 males) were directly assessed for SOR using a structured clinical evaluation, the Sensory Processing 3 Dimensions Assessment (SP3D:A), and underwent 3 Tesla MRI. 39 ND children presented with SOR (ND/SOR) and 44 ND children presented without SOR (ND/NO-SOR). Exogenous and endogenous functional connectivity networks (FCNs) were generated through independent component analysis and investigated with two local functional connectivity (FC) measures, fractional amplitude of low-frequency fluctuations (fALFF) and regional homogeneity (ReHo), as well as a long-range FC measure, dual regression (DR). Moreover, we examined FC in the context of behavioral regulation as assessed through the Behavioral Assessment System for Children, 3rd edition (BASC-3), categorizing children as “resilient” or “dysregulated” through latent profile analysis. Results In general, ND/SOR children exhibit reduced long-range exogenous FC. However, in terms of local FC, we find that ND/SOR children have reduced exogenous and elevated endogenous FC which is diametrically opposed to ND/NO-SOR children. Furthermore, this double dissociation is specific to ND children who are behaviorally resilient, while emotionally dysregulated ND children possess a distinct pattern. Conclusion Achieving optimal brain system connectivity—a balanced contrast—is influenced by sensory over-responsivity and essential for resilience.
BibTeX
@article{choi2025neural,
  author = {Hannah L. Choi and Maia C. Lazerwitz and R.A. Powers and Mikaela A. Rowe and Jamie Wren-Jarvis and Amir Sadikov and Lanya T. Cai and Robyn Chu and LaShelle Rullan and Kaitlyn J. Trimarchi and Rafael D. Garcia and Elysa J. Marco and Pratik Mukherjee},
  title = {{A neural substrate for sensory over-responsivity defined by exogenous and endogenous brain systems}},
  journal = {Journal of Neurodevelopmental Disorders},
  year = {2025},
  doi = {10.1186/s11689-025-09656-y},
}
Figure