Sex Differences in Subcortical Morphometry of Children with ASD

Friday, May 12, 2017: 12:00 PM-1:40 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
K. Hirabayashi1, D. Crocetti2, B. C. Lee3, X. Tang4,5, D. J. Tward3 and S. H. Mostofsky2, (1)Center for Neurodevelopmental and Imaging Research, Kennedy Krieger Institute, Baltimore, MD, (2)Kennedy Krieger Institute, Baltimore, MD, (3)Center for Imaging Science, Johns Hopkins University, Baltimore, MD, (4)Carnegie Mellon University, Pittsburgh, PA, (5)Sun Yat-Sen University, Guangzhou, China
Background:  ASD is associated with impairments in social, goal-directed, and motor function. Subcortical structures contribute to the control of motor and social functions, including networks of communication, emotion, and reward. Due to the underrepresentation of girls in anatomical studies of autism, there is a paucity of literature exploring neuroanatomical correlates of sex-based differences in children with ASD.

Objectives:  We compared the volumes and morphology of subcortical structures in ASD boys and girls to those of their typically-developing peers. We were particularly interested in examining how sex differentially affects the volumes and morphology of subcortical structures of ASD compared to TD populations. We examined differences in the left and right caudate, putamen, globus pallidus, and thalamus.

Methods:  T1-weighted MPRAGE images were acquired from 107 children with ASD (20 girls) and 107 typically-developing children (19 girls). Subcortical parcellations were performed using MRICloud, a high-throughput neuroinformatics platform for automated brain MRI segmentation. Univariate ANCOVA tests were performed on the volumes of each structure, with the between-subject factor of diagnostic group and covariates of age, total cerebral volume (TCV), and head coil type. P-values were FDR-corrected. LDDMM was used to examine morphological features of the subcortical structures. Diagnostic effects were tested using a linear regression model including age, TCV and head coil as covariates. Non-parametric permutation testing was used to obtain corrected p-values for each vertex on a shape’s surface. Partial correlations were conducted to assess associations between subcortical volumes and behavioral measures of motor and social function and autism symptom severity. Age, TCV, and head coil were used as covariates

Results:  ANCOVA revealed that girls with ASD showed larger right caudate and right thalamus volumes compared to TD girls (p=0.012, d = 0.872; p=0.012, d = 0.873). No volume differences were observed in boys. No sex-by-diagnosis interactions on subcortical volumes were found. Correlations between subcortical volumes and ADOS scores were found predominantly in ASD girls. In ASD girls, higher ADOS total scores correlated with larger left caudate volumes (p =0.017, r=0.586). Higher ADOS Communication+Social Interaction scores correlated with larger left caudate, left putamen, and right putamen volumes in ASD girls (p=0.0008, r=0.638; p=0.014; r=0.601; p=0.017, r=0.585). No volume correlations with ADOS were found in ASD boys.

Linear regression revealed the effect of diagnosis on subcortical shape, including localized expansion on the right putamen and right thalamus for children with ASD compared to TD children in the whole group (19/553 vertices; 7/605 vertices). Within ASD girls, we observed expansions in the right putamen and right globus pallidus compared to TD girls (2/512 vertices; 5/234 vertices). Within ASD boys, we observed expansions in the right putamen and left globus pallidus compared to TD boys (3/525 vertices; 3/250 vertices).

Conclusions: Our volumetric findings indicate that girls with ASD display more robust differences within same-sex peers than boys with ASD. An increased sample size of girls may facilitate our understanding of the relationship between sex and diagnosis in subcortical structures. Additional analysis will include examining shape correlations with measures of ASD symptomology, and motor and social function.