Age-Associated Changes in Functional Networks in ASD: Is There a Shift from Overconnectivity in Childhood to Underconnectivity in Young Adulthood?

Saturday, May 16, 2015: 11:30 AM-1:30 PM
Imperial Ballroom (Grand America Hotel)
N. Ray1, A. Jahedi1, C. P. Chen1, I. Fishman1 and R. A. Müller2, (1)Brain Development Imaging Laboratory, Dept. of Psychology, San Diego State University, San Diego, CA, (2)San Diego State University, San Diego, CA

It has recently been suggested that developmental changes in network connectivity may explain seemingly inconsistent findings of functional over- vs. underconnectivity in ASD. A recent review (Uddin et. al. 2013) of intrinsic functional connectivity (iFC) studies in ASD suggests that there might be an age-related shift from increased connectivity in children with ASD to reduced connectivity in adolescents and young adults with the disorder.


This study aims to empirically test this notion in a large sample of individuals with ASD and typically developing (TD) controls, utilizing the Autism Brain Imaging Data Exchange (ABIDE) dataset.


Utilizing a low-motion subset of the ABIDE resting-state fMRI data (94 individuals with ASD and 94 TD participants; ages 7-34 years; head motion <0.2mm), within-network iFC for three commonly observed networks (Default Mode, Mirror Neuron, and Language) was assessed. Following standard preprocessing procedures (slice-time and motion correction; co-registration and standardization to the MNI space; application of the bandpass filter and spatial smoothing; removal of nuisance regressors including motion, white matter, ventricular and global signals, and their derivatives), canonical regions of interest (ROIs) for each network were identified based on previous reports. Using average time series extracted from each ROI, whole-brain correlation maps were created, cluster corrected (p<1-7) and Fisher-transformed to z’. For each network node (used as seed), the mean z’ was extracted from all other nodes of the respective network to determine the within-network functional connectivity. The relationship between these scores and age was examined within each group with linear and polynomial (quadratic) models, to determine the change in connectivity across age.  


For none of the three networks, a significant effect of age on iFC was found. Although age-related slopes were slightly more positive in the ASD than in the TD group for each network, none of these interaction effects approached significance.


Our findings suggest that effects of individual variability (and other sources of variability that may be hard to control in multisite datasets) may dominate more subtle age-related changes in within-network iFC between ages 7 and 34 years. Specifically, no evidence supporting a crossover from overconnectivity in childhood to underconnectivity in adolescence or adulthood could be detected in ASD.