Summary: New research shows that children with autism exhibit differences in neuron density across specific brain regions compared with their peers. Using advanced MRI techniques applied to data from the ABCD study, researchers identified lower neurite density in cortical areas involved in memory and reasoning, alongside higher neurite density in emotion-related regions such as the amygdala.
These differences were specific to children with autism and were not observed in groups with other common psychiatric diagnoses, suggesting a distinctive neuroanatomical profile. The results offer fresh insight into brain development in autism and may help guide future, more targeted interventions.
Key Facts
- Children with autism show altered neuron density in particular brain regions.
- Reduced neurite density appeared in cortical areas linked to memory, learning, and problem-solving.
- Some limbic regions, including the amygdala, showed relatively increased neuron density.
- These anatomical patterns were specific to autism and were not reproduced in groups with other psychiatric conditions.
Source: University of Rochester
Overview
Researchers at the Del Monte Institute for Neuroscience, University of Rochester, analyzed diffusion-weighted MRI data to study neuronal cytoarchitecture in vivo. Rather than relying on postmortem tissue, the team used advanced imaging methods to estimate neurite and neuron density in living children, revealing measurable differences associated with autism.
“For many years, studies focused on large-scale features such as cortical thickness, volume, and curvature,” said Zachary Christensen, MD/PhD candidate at the University of Rochester School of Medicine and Dentistry and first author of the study published in Autism Research. “New MRI approaches now allow us to probe cellular-level architecture noninvasively and reveal previously hidden complexity across development.”
Methods and main findings
The investigators examined brain imaging from more than 11,000 children aged 9–11 collected through the Adolescent Brain Cognitive Development (ABCD) Study. Within that dataset, 142 children had a recorded autism diagnosis and were compared with thousands of controls. Using a restriction spectrum imaging (RSI) framework applied to diffusion-weighted imaging, the team estimated total neurite density (TND) and its components—restricted normalized directional diffusion (RND) and restricted normalized isotropic diffusion (RNI).
Key observations included a significant decrease in overall neurite density in portions of the cerebral cortex tied to memory, reasoning, and complex cognition. By contrast, limbic structures such as the amygdala—central to emotion processing—displayed relatively increased neurite density. These regional patterns were robust when the autism group was compared not only to typically developing controls but also to a subgroup of children with other psychiatric conditions (for example, ADHD and anxiety), indicating specificity to autism.
“Many individuals with autism also experience anxiety, depression, or ADHD. The fact that these neurite measures differentiated autism specifically points to a potentially valuable biomarker for characterizing the condition,” Christensen said. He noted that reliable, noninvasive measures of neuronal structure could help track developmental trajectories and identify children who might benefit from specialized therapeutic approaches.
Implications and technological advances
The ability to assess neuronal microstructure in living children represents a major step forward. Historically, fine-grained cytoarchitectural differences were typically inferred from postmortem tissue, limiting sample sizes and reproducibility. Modern diffusion MRI processing makes it possible to evaluate neurite architecture at scale and incorporate these physiological measures into developmental and clinical research.
The ABCD Study, the largest long-term investigation of brain development and child health in the United States, provided the extensive imaging dataset that made this analysis possible. The University of Rochester is one of 21 national sites contributing data to ABCD, which began in 2015 and continues to yield insights as participants move from childhood into adolescence and early adulthood.
“The extraordinary data collected by the ABCD Study are already reshaping our understanding of brain development and will only grow more informative as the cohort matures,” said John Foxe, PhD, senior author of the study and director of the Del Monte Institute for Neuroscience.
Additional authors include Edward Freedman of the University of Rochester Medical Center.
Funding: This research was supported by the Adolescent Brain Cognitive Development (ABCD) Study and the Translational Neuroimaging and Neurophysiology Core of the University of Rochester Intellectual and Developmental Disabilities Research Center.
About this autism and neuroscience research news
Author: Kelsie Smith Hayduk
Source: University of Rochester
Contact: Kelsie Smith Hayduk – University of Rochester
Image: Image credited to Neuroscience News
Original Research: Open access. “Autism is associated with in vivo changes in gray matter neurite architecture” by Zachary Christensen et al., Autism Research. The study used ABCD diffusion MRI data and RSI modeling to compare neurite measures across children with autism and large control samples.
Abstract
Autism is associated with in vivo changes in gray matter neurite architecture
Previous postmortem studies have reported anomalies in neural cytoarchitecture across limbic, cerebellar, and neocortical networks in autism, including narrower cell minicolumns and variable neuron density. Limited availability of postmortem tissue has hindered reproducible, large-scale characterization.
Advances in processing diffusion-weighted MRI now permit in vivo assessment of neuronal cytoarchitecture. Using extensive DWI data from the ABCD Study, 142 individuals with an autism diagnosis were compared with 8,971 controls via an RSI framework that quantified total neurite density (TND), restricted normalized directional diffusion (RND), and restricted normalized isotropic diffusion (RNI).
The analysis identified a significant decrease in TND in autism within the right cerebellar cortex, along with regionally distributed decreases in RNI and increases in RND across posterior and anterior brain areas, respectively. These regional differences persisted in post-hoc comparisons against a subset of 1,404 individuals with other psychiatric diagnoses. The findings emphasize the value of characterizing neuron cytoarchitecture in autism and incorporating such physiological measures as covariates in future studies.