Default Mode Connectivity in Children with and without Autism, and Their Siblings

Background: Recent fMRI research has used temporal correlations between BOLD signals in the resting-state as an assay of functional connectivity in the human brain.  In particular, this research has identified two large-scale, anticorrelated functional networks: the task-positive network, active during the performance of cognitively or attentionally demanding tasks, and the default mode network (DMN), engaged during reflective or ruminative cognition (such as self-reflective thought, theory of mind reasoning, and autobiographical episodic recall).  Given that the default mode network has been linked to cognitive functions that are putatively impaired in autism, this network may function abnormally in those with autism; prior research has shown reductions in DMN connectivity in adults with autism, but research has not investigated this network in children with autism.  Such research will be relevant to the question of whether connectivity abnormalities play a role in the etiology of autism.

Objectives: To assess functional connectivity in the default mode and task-positive networks in children with autism, and their siblings.

Methods: A collection of children (age 5-15) with and without autism, and unaffected siblings of kids with autism, were scanned in a 3T magnet for 6:40 minutes in the absence of stimulus presentation.  Seed voxel-based connectivity analyses were performed in each subject, using time series from precuneus and left anterior insula as predictors, to assess connectivity in default mode and task-positive networks, respectively.  Motion artifacts are common in fMRI data acquired in children, and have a particularly strong effect on connectivity analyses; thus any dataset with more than 1mm of motion within a period of 3 TRs (6s) was discarded from the analysis.

Results: Preliminary results suggest that default mode and task positive network connectivity are both largely intact in children with autism, and their siblings, with minimal differences in correlation strengths between kids with and without autism.

Conclusions: Together with prior studies demonstrating DMN abnormalities in adults with autism, these results suggest either that these abnormalities develop after childhood, or that the previously observed differences resulted from the presence of artifacts (e.g., from head motion) in data from subjects with autism.  The findings suggest against the claim that abnormalities in long-range connectivity play a central role in the development of autism.