A Functional MRI Study of Imitation and the Mirror Neuron System in ASD

Background: The mirror neuron (MN) theory of ASD posits that impairments in  language, empathy, Theory of Mind and imitation are related to a deficit in the observation-execution matching mechanism afforded by the MN system (e.g. Oberman et al., 2005). Nevertheless, only a limited number of studies have investigated the neurobiological underpinnings of this theory. In the human brain regions with ‘mirror’ properties  include the inferior frontal gyrus (IFG/ventral premotor cortex), the inferior parietal lobe (IPL), and to a lesser extent temporal and sensorimotor regions. Both electrophysiological (e.g. Oberman et al., 2005) and functional MRI (e.g. Williams et al., 2005) studies have indicated reduced mirror -like activity in ASD. Moreover, it has been suggested that altered functional connectivity may contribute to the dysfunction of the ‘mirror’ system (e.g. Perkins et al., 2010; Williams et al., 2005). 

Objectives: To extend previous findings of reduced mirror-like activity in ASD using a carefully controlled imitation paradigm and functional connectivity analyses. 

Methods: Behavioural and fMRI data were acquired from 21 TD individuals and 20 individuals with ASD. Imitation was tested using a paradigm based on the task developed by Iacoboni et al., (1999). The original paradigm included three conditions, 1) imitation, 2) imitation with symbolic cue and 3) symbolically-cued action (Symbol-Static). New stimuli were created for each condition and an additional condition was developed to better control for observed biological motion during symbolically-cued action (Symbol-Bio). Furthermore, participants were required to count target events during observation conditions. 

Results: Analysis of behavioural data revealed no significant group differences in either imitative or symbolically-cued action or in the control task included in the observation conditions. Imitation evoked significantly increased activity relative to symbolically-cued action in the anterior IPL in both groups, although there were subtle group differences in the pattern of activity. Nevertheless, the observation of action alone did not evoke significant activity in the anterior IPL in either group, suggesting that it could not be considered as a prototypical ‘mirror’ area. There was no evidence of mirror-like activity in the IFG. The concurrent observation and execution of incongruent actions (Symbol-Bio) evoked patterns of activity indicative of inhibitory modulation affecting lateral occipitotemporal and sensorimotor regions. This effect was less pronounced in the ASD group. Psychophysiological interaction analyses indicated increased interaction between right lateral occipitotemporal regions and both posterior and anterior ‘mirror’ regions in TD individuals during imitation compared with Symbol-Static. In individuals with ASD, however, increased interaction was observed between the left lateral occipitotemporal region and the left IPL in the uncorrected data. 

Conclusions: The most likely candidate ‘mirror’ region identified in the current study was the anterior IPL. Nevertheless, this region was not a prototypical ‘mirror’ area, nor was there activity in the anterior ‘mirror’ area. Subtle group differences were particularly apparent in functional connectivity analyses, consistent with the theory that altered connectivity may contribute to ‘mirror’ system dysfunction in ASD. The incidental findings indicative of laterality effects and inhibitory modulation will be discussed.