Widespread Associations between Behavioral Metrics and Brain Microstructure in ASD
Poster No:
409
Submission Type:
Abstract Submission
Authors:
Benjamin Newman1, Haylee Ressa1, Zachary Jacokes1, Jason Druzgal1, Kevin Pelphrey1, John Van Horn1
Institutions:
1University of Virginia, Charlottesville, VA
First Author:
Co-Author(s):
Introduction:
Autism spectrum disorder (ASD) is a complex, multifaceted condition involving a number of behavioral and cognitive components and diagnosed via behavioral and cognitive tests administered by a trained clinician1,2. A recent paper3 demonstrated that changes in two diffusion MRI cellular microstructural metrics of neuronal capacity, termed aggregate g-ratio and aggregate conduction velocity, are significantly different in autistic individuals compared to non-autistic individuals. If differences in these metrics are representative of genuine neurological differences contributing to ASD, then similar relationships should be observed in validated behavioral tests used to evaluate ASD.
Methods:
Participants: 273 subjects (mean age = 154.3 months ±35.21 S.D., 133 female [49%]) were included in this study. 148 were diagnosed with ASD (mean age = 150.8 months ±34.31 S.D., 70 female [47%]) and 124 non-autistic participants (mean age = 154.3 months ±35.21 S.D., 62 female [50%]).
Behavioral and Cognitive Metrics: All subjects were evaluated using several widely utilized neuropsychiatric metrics: Clinical Evaluation of language Fundamentals (CLEF-4), Behavior Rating Inventory of Executive Function (BRIEF), Repetitive Behavior Scale (RBS-R), Child Behavior Checklist (CBCL), Adolescent/Adult Sensory Profile, Differential Ability Scales (DAS-School Age), Vineland adaptive behavior scales, and individuals with ASD were further evaluated using the Autism Diagnostic Observation Schedule (ADOS-2), age at language acquisition, and the Autism Diagnostic Interview-Revised for a total of 94 different metrics when subscales from each evaluation were included.
Image Acquisition: Diffusion, T1-weighted, and T2-weighted images were acquired from each subject. Diffusion images were acquired with an isotropic voxel size of 2×2×2mm, 64 non-colinear gradient directions at b=1000 s/mm2, and 1 b=0. T1-weighted MPRAGE images with a FOV of 176×256×256mm and an isotropic voxel size of 1×1×1mm, TE=3.3; T2-weighted images were acquired with a FOV of 128×128×34 with a voxel size of 1.5×1.5×4mm, TE=35.
Image Processing: As described in more detail in previous work3 images were preprocessed with MRtrix34, FSL5, and Freesurfer6 to calculate voxel-wise aggregate g-ratio and aggregate conduction velocity. The mean value of these metrics was measured within each of the 164 regions of the Destrieux Cortical Atlas and 48 regions of the JHU WM Atlas. Linear models tested the association between the score on each component of each behavioral test and the mean microstructural value in each ROI while controlling for age, sex, scanning site, total brain volume, and IQ with a Benjamini and Hochberg multiple comparison correction.
Behavioral and Cognitive Metrics: All subjects were evaluated using several widely utilized neuropsychiatric metrics: Clinical Evaluation of language Fundamentals (CLEF-4), Behavior Rating Inventory of Executive Function (BRIEF), Repetitive Behavior Scale (RBS-R), Child Behavior Checklist (CBCL), Adolescent/Adult Sensory Profile, Differential Ability Scales (DAS-School Age), Vineland adaptive behavior scales, and individuals with ASD were further evaluated using the Autism Diagnostic Observation Schedule (ADOS-2), age at language acquisition, and the Autism Diagnostic Interview-Revised for a total of 94 different metrics when subscales from each evaluation were included.
Image Acquisition: Diffusion, T1-weighted, and T2-weighted images were acquired from each subject. Diffusion images were acquired with an isotropic voxel size of 2×2×2mm, 64 non-colinear gradient directions at b=1000 s/mm2, and 1 b=0. T1-weighted MPRAGE images with a FOV of 176×256×256mm and an isotropic voxel size of 1×1×1mm, TE=3.3; T2-weighted images were acquired with a FOV of 128×128×34 with a voxel size of 1.5×1.5×4mm, TE=35.
Image Processing: As described in more detail in previous work3 images were preprocessed with MRtrix34, FSL5, and Freesurfer6 to calculate voxel-wise aggregate g-ratio and aggregate conduction velocity. The mean value of these metrics was measured within each of the 164 regions of the Destrieux Cortical Atlas and 48 regions of the JHU WM Atlas. Linear models tested the association between the score on each component of each behavioral test and the mean microstructural value in each ROI while controlling for age, sex, scanning site, total brain volume, and IQ with a Benjamini and Hochberg multiple comparison correction.
Results:
When examined using data from all subjects, conduction velocity was associated with 47 different subscales in at least 1 ROI (Fig. 1a). The BRIEF, RBS-R, CELF-4, CBCL, SRS-2 and Vineland-II were significantly associated with conduction velocity measured across a wide variety of cortical ROIs but particularly in the superior and frontal cortex, and subcortical gray matter (Fig. 2a). G-ratio was not as widely nor strongly associated across ROIs with the behavioral metrics, with the largest associations found in the deep WM in the BRIEF and DAS. However this pattern was reversed when the associations were considered exclusively within the autistic participants (Fig. 1b). G-ratio was more strongly associated with metrics across a number of ROIs, particularly the CBCL, than conduction velocity. G-ratio relationships were largely located in the motor cortex and WM (Fig. 2b).
·Figure 1
·Figure 2
Conclusions:
Despite differences in evaluations there was a great deal of overlap in brain regions associated with the various metrics, particularly when non-autistic subjects are included. However the switch to more significant g-ratio measurements when evaluating exclusively autistic subjects suggests that behavioral severity in autism may not follow the same neurological mechanism as diagnosis.
Disorders of the Nervous System:
Neurodevelopmental/ Early Life (eg. ADHD, autism) 1
Lifespan Development:
Early life, Adolescence, Aging 2
Keywords:
Autism
Cognition
Development
MRI
Myelin
Neurological
PEDIATRIC
Treatment
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
1|2Indicates the priority used for review
Provide references using author date format
Constantino, J. N., & Charman, T. (2016). Diagnosis of autism spectrum disorder: Reconciling the syndrome, its diverse origins, and variation in expression. The Lancet Neurology, 15(3), 279–291.
Fischl, B. (2012). FreeSurfer. Neuroimage, 62(2), 774–781.
Jenkinson, M., Beckmann, C. F., Behrens, T. E., Woolrich, M. W., & Smith, S. M. (2012). Fsl. Neuroimage, 62(2), 782–790.
Newman, B. T., Jacokes, Z., Venkadesh, S. T., Webb, S. J., Kleinhans, N. M., McPartland, J. C., Druzgal, T. J., Pelphrey, K. A., Van Horn, J. D., & Consortium, G. R. (2023). Conduction Velocity, G-ratio, and Extracellular Water as Microstructural Characteristics of Autism Spectrum Disorder. bioRxiv, 2023.07. 23.550166.
Tournier, J.-D., Smith, R., Raffelt, D., Tabbara, R., Dhollander, T., Pietsch, M., Christiaens, D., Jeurissen, B., Yeh, C.-H., & Connelly, A. (2019). MRtrix3: A fast, flexible and open software framework for medical image processing and visualisation. NeuroImage, 202, 116137.