An Effective Connectivity Approach to Autism Spectrum Disorders

Thursday, May 12, 2016: 11:30 AM-1:30 PM
Hall A (Baltimore Convention Center)
S. Vettori1,2,3, D. Marinazzo2, B. Boets1,3 and R. Siugzdaite2,4, (1)Child and Adolescent Psychiatry, KU Leuven, Leuven, Belgium, (2)Data analysis Department, Ghent University, Gent, Belgium, (3)Leuven Autism Research consortium (LAuRes), KU Leuven, Leuven, Belgium, (4)Experimental psychology department, University of Ghent, Ghent, Belgium

Many neuroimaging studies indicate that autism spectrum disorders (ASD) are characterized by aberrant functional connectivity. However, the nature and directionality of these aberrant connections remain largely unknown and under-investigated.


With a novel combination of advanced neuroimaging data analysis methods, we aim to understand the directionality of these aberrant functional connections.


We used resting state fMRI data of the Autism Brain Imaging Data Exchange database (ABIDE) from 21 participants with ASD and 21 IQ-matched typically developing (TD) children (age range: 7-12 years).  First, we used functional connectivity density mapping (FCD), a whole-brain data-driven method, to identify local and global differences in functional connectivity density between resting-state fMRI data of the ASD and the TD group. Then we used partially conditioned Granger causality analysis to investigate the directionality of those connections that showed group differences in functional connectivity density. With this method, a reliable estimate of group differences in directed connectivity was obtained. 


Local and global functional connectivity density mappings reveal clusters with more dense short-range local and global connections in four subcortical and five frontal regions in the ASD group, as compared to the TD group. Granger causality indicates that both the incoming and the outgoing information flow is increased in ASD in 7 of these 9 regions. However, in the right caudate and the left inferior frontal gyrus (two regions known to be involved in ASD), the incoming information flow is higher than the outgoing flow in ASD, while this pattern is reversed in the TD group.

Conclusions:  With a novel combination of advanced neuroimaging data analysis methods, we investigated differences in dynamical functional connectivity and in the directionality of information flow between children with ASD and TD controls. For the right caudate and the left inferior frontal gyrus (relevant for respectively sensorimotor control and language processing), we found that children with ASD show more incoming and less outgoing functional connections than TD controls. Accordingly, this may indicate that information from these two regions is less transferred to control other regions in a top-down manner, which may relate to deficits in sensorimotor control and language processing in ASD. We conclude that this pattern of inverted effective connectivity in right caudate and left inferior frontal gyrus in ASD may explain some of the inconsistencies in the literature, and that this novel finding provides important insights in the neurobiological origin of ASD.