Neural Basis of Action and Intention Understanding in Autism and Typical Development

Background:
The mirror neuron system (MNS) is comprised of interconnected brain regions whose neurons respond in a similar way whether an individual is observing or executing an action. Some define the MNS strictly as an action observation-execution matching system, while others assert that it supports intention understanding and empathy. It follows that such a system would be impaired in autism, and several studies have found preliminary evidence for this hypothesis (e.g. Oberman et al., 2005). Recent experimental work suggests that children with autism are not impaired in all the functions thought to be supported by the MNS (Carpenter et al., 2001), making it hard to reconcile the hypothesis of a global MNS impairment in ASD with the ability to understand intentional acts on objects.This study is a counterpart to a recent EEG study from our group that exploring the hypothesis with mu wave suppression as the indicator of MNS activity. In the current fMRI study, we examine the same experimental participant group, using the same experimental stimuli for one paradigm while expanding it with a second, complementary, paradigm.

Objectives:
The objective of this research is to reconcile the aforementioned inconsistency by comparing indices of MNS activation in children with ASD to those with typical development (TD). The study explores MNS activation in response to observation of actions varied by transitivity, visibility and the conventionality of underlying intention. We hypothesize that, compared the matched TD control group, children with ASD will demonstrate abnormal patterns of response during observation of other people’s intransitive actions and actions for unconventional intentions.

Methods:
Neuroimaging and psychological measures were obtained from the two participant groups. During functional neuroimaging, participants viewed videos of transitive and intransitive grasping actions, with the hand-to-object grasp occluded by a screen for half of the stimuli, and videos of actions demonstrating conventional and unconventional intentions. The behavioral measures assessed general cognitive skill, imitative ability and intention understanding.

Results:
Preliminary neuroimaging analyses show that TD participants and participants with ASD display neural activation in the inferior frontal gyrus and inferior parietal lobe, key components of the MNS, while viewing actions on objects and actions that indicate conventional intentions. Both groups show negligible MNS activation when viewing intransitive actions. These early findings also replicate Umilta’s study, showing a main effect for transitivity (Umilta et al., 2001) with negligible effect of the visibility of the final grasping (or mimed grasping) action. Consistent with the behavioral imitation literature, the MNS of participants with ASD was more robust for transitive compared to intransitive actions. Similarly, we are finding different MNS activation patterns for actions demonstrating conventional contrasted with unconventional intentions. The MNS activation of TD participants is showing a main effect of intention conventionality, while that of the ASD is not.

Conclusions:
These preliminary analyses support our hypothesis that there is not a gross absence of MNS activation in children with ASD. Indeed, there seems to be typical MNS activation for paradigms depicting behaviors with which children with ASD have preserved competency, namely imitating transitive actions.