Altered Resting-State Functional Connectivity in Individuals with ASDs and Impaired Cognitive Functioning

Oral Presentation
Thursday, May 10, 2018: 1:57 PM
Willem Burger Zaal (de Doelen ICC Rotterdam)
M. A. Reiter1,2, L. E. Mash1,2, J. Liu2, C. H. Fong3, A. C. Linke3, I. Fishman3 and R. A. Mueller3, (1)Joint Doctoral Program in Clinical Psychology, SDSU / UC San Diego, San Diego, CA, (2)Brain Development Imaging Laboratories, Department of Psychology, San Diego State University, San Diego, CA, (3)Brain Development Imaging Laboratories, San Diego State University, San Diego, CA
Background: General functional level, as measured by IQ, is an important correlate of adaptive functioning (Liss et. al, 2001), a set of cognitive, social, and practical abilities enabling mastery of age-expected daily-living skills. Although >50% of individuals with Autism Spectrum Disorders (ASDs) exhibit IQ scores below 85 (CDC, 2014), little is known about functional brain organization in this population, as high-quality fMRI data can usually only be collected from high-functioning individuals. Improved understanding of brain mechanisms that predict long-term outcome in broader segments of the ASD spectrum is an urgent research priority.

Objectives: To examine intrinsic functional connectivity (iFC) in children and adolescents with ASDs and IQ≤85 (L-ASD) in comparison with typically developing (TD) and ASD peers with higher IQ (H-ASD).

Methods: T1-weighted anatomical and eyes-open fMRI resting-state scans from 88 individuals (44 ASD, 44 TD) were taken from in-house data (SDSU: n=36) and two sites (NYU: n=28; OHSU: n=24) from the Autism Brain Imaging Data Exchange (ABIDE). The full sample was split by diagnosis and IQ scores into 4 groups of 22 individuals (L-ASD [IQ≤85] mean IQ: 77±6; H-ASD: 123±8; Average-TD [A-TD]: 100±7; High-TD [H-TD]: 124±8). All groups were matched within and across scanning-sites on head-motion, age, gender, and handedness.

Following standard preprocessing, mean timeseries were extracted from seeds in the medial prefrontal cortex (mPFC), posterior cingulate cortex (PCC), posterior superior temporal sulcus (pSTS), insula, and amygdala, and entered into a subject-level GLM. Whole-brain group differences in iFC for 4 contrasts of interest (L-ASD vs. H-ASD, L-ASD vs. A-TD, H-ASD vs. H-TD, and A-TD vs. H-TD; Figure 1) were examined using AFNI 3dttest++. All results were significant at a voxelwise p < .005, corrected to control for false-discovery rate (α < .05) using permutation testing.

Results: Compared to the H-ASD group, the L-ASD group showed significantly weaker positive iFC between the mPFC and PCC, and between pSTS bilaterally and occipital cortices (Figure 2). Compared to the A-TD group, the L-ASD group showed increased connectivity between mPFC and cuneal and intracalcarine cortex. Compared to the H-TD group, the H-ASD group had higher connectivity (mostly reduced anti-correlations) between the PCC and right superior frontal gyrus, right anterior insula, and left Crus-1 of the cerebellum. Group differences were largely consistent across scanning sites. There were no significant iFC differences between the two TD groups.

Conclusions: Our findings suggest that within the autism spectrum, general functional level may be associated with specific patterns of functional network abnormalities, affecting social-processing regions and core hubs of the default-mode and salience networks. Remarkable in L-ASD participants was atypical iFC with visual cortices, reduced for social cognition (pSTS), but increased with a default mode region (mPFC). Some earlier findings have suggested that visual cortex may play compensatory roles in ASDs (Simard et al., 2015), and our findings suggest that any such mechanisms may be less engaged in individuals with L-ASDs. Our preliminary study illustrates the importance of including individuals with L-ASDs in research seeking to relate brain functioning in autism to long-term outcome variables.