Sex Differences in Interhemispheric Resting State Functional Connectivity in Autism Spectrum Disorder in Early Childhood

Background: Sex differences in the neuropathology of autism spectrum disorder (ASD) are currently poorly understood. While prior research reported sex and ASD related differences in diffusion properties and organization of the corpus callosum in pre-school age children (Nordahl et al., 2015), these interhemispheric connectivity differences have not been examined using resting-state functional magnetic imaging (rfMRI).

Objectives: The current research examines sex-differences in interhemispheric resting state functional connectivity in preschool aged children with and without ASD, and how individual differences ASD symptom severity relates to interhemispheric connectivity.

Methods: The sample included 119 children with ASD (86 male, 33 female) and 68 age-matched typically developing (TD) controls (40 male, 28 female) (mean age 3.7 years). Diagnostic assessments for ASD were carried out by expert clinicians using the ADOS and ADI-R. Structural and resting state EPI BOLD images were acquired during natural nocturnal sleep. Resting state images were preprocessed using tools from AFNI, FSL, and ANTS in the Configurable Pipeline for the Analysis of Connectomes (C-PAC; Craddock et al., 2013). In brief, EPI images were time-shifted, motion corrected, and band-pass filtered (.008 < f < .08 Hz). Volumes with frame-wise displacement greater than 0.25mm were scrubbed. EPIs were then co-registered to the participant’s structural T1-weighted image and then to symmetric MNI space, and smoothed at 6 mm FWHM. Interhemispheric connectivity was assessed using voxel-mirrored homotopic connectivity (VMHC). Cluster based correction for multiple comparisons was carried out using Gaussian Random Field theory using FSL (Z >2.58, p_GRF <.05). For each interaction cluster, mean interhemispheric connectivity was extracted for each participant to test simple effects and slopes; reported p-values for these tests are adjusted for multiple comparisons using Bonferroni.

Results:

Preliminary results reveal atypical interhemispheric functional connectivity in both boys and girls with ASD. Both males and females with ASD exhibit decreased interhemispheric connectivity between left and right posterior cingulate and precuneus cortices relative to TD (p = .01), with a cluster in the precuneus exhibiting a larger decrease in ASD males relative to TD males than do ASD females relative to TD females (Interaction: p = 1.0e-4). Reductions in interhemispheric connectivity relative to TD were also observed in paracingulate cortex and frontal operculum in males (ps ≤ .002). Greater ADOS severity scores predict greater interhemispheric connectivity in the cerebellum in females (b = .12, p = .048), but not males (b = -.07, p = .12), (interaction: p = .006). Greater ADOS severity scores predict decreased interhemispheric connectivity in two clusters in precuneus and superior temporal gyrus in males (bs = ≤ -.07, ps ≤ .001), but positively in the middle temporal gyrus for females (b = .08, p = .03) (interactions: ps ≤ 3.1e-5). These results suggest that the neural phenotype of ASD in young children is differentially presented in males and females compared to their respective typically developing peers.

Conclusions: These preliminary data suggest that females and males with ASD have partially dissimilar patterns of interhemispheric functional connectivity.