Short Range Over-Connectivity and Long Range Under-Connectivity in the Resting State Network in Autism Spectrum Disorders

Background:  There is converging evidence of atypical functional brain connectivity within task specific and idle neural networks in individuals with Autism Spectrum Disorders (ASD).  It has been reported that individuals with ASD have greater short to medium range connections and fewer long range connections. However, previous studies have been small in size and/ or have not been carried out in children and young adolescents with ASD. 

Objectives:  We examined (1) functional connectivity among regions of the resting state network [i.e. precuneus (central hub) with anterior cingulate cortex, and middle temporal gyrus] and (2) functional connectivity between the precuneus and neighbouring regions [i.e. posterior cingulate cortex, calcarine, cuneal, lingual gyri and the primary visual cortex].

Methods:  We report here on data from 65 participants between the ages of 9-15 years (ASD: n=31, and typically developing controls: n=34). A significant between-group difference was found in IQ (ASD: 93.14 ± 18.2 and TD: 116.61 ± 9.07, p<0.05). Functional Magnetic Resonance Imaging data were pre-processed using a combination of AFNI, FSL and locally developed software tools. Each participant’s functional scan was coregistered to their anatomical scan, and ROIs were drawn using the Harvard-Oxford cortical atlas. ROIs were iteratively eroded until the mean intravoxel cross-correlation within an ROI reached a threshold of 0.8 (based on the methods of Golestani and Goodyear, 2011). For each participant, we calculated the coefficient of determination (R^2, or the square of the correlation coefficient) among all regions, bilaterally. We tested for significant mean differences between control and ASD subjects using a two-tailed t-test.

Results:  Significantly higher correlations were found between the precuneus and neighbouring regions in ASD compared to controls (all p< 0.05) and higher connectivity in the resting state network was observed in controls compared to individuals with ASD (all p<0.05).

Conclusions:  These preliminary results suggest short range over-connectivity in children and adolescents with ASD, in the context of a under-connected resting state network. These findings extent our current knowledge of atypical functional connectivity in ASD into a younger population.