It has been suggested that an early deficit in the human Mirror Neuron System (hMNS) is one of the key problems in autism spectrum disorder (ASD). Early dysfunctions related to the intuitive understanding and mental representation of actions and emotions may result in cascading deficits in social interaction. Alternatively, observed differences in brain activation within the hMNS may be a secondary effect, influenced by social motivation, attentional effects, and stimulus properties (such as emotional and social valence).
The present study aimed at investigating the hypothesis of a deficit in the hMNS in adolescents and young adults with ASD using functional magnetic resonance imaging (fMRI). Furthermore we aimed to explore whether activation in the hMNS and other regions associated with facial imitation and observation is modulated by the social valence of facial stimuli.
We presented 15 patients with ASD and 18 typically developing controls (TDC), aged 12-20, with videos of facial expressions (smile or frown), which they should either imitate or merely observe. Video recordings of the participant's faces during the fMRI scan were used to ensure proper performance. Stimulus faces belonged to two groups differing with respect to social valence, representing either "friends" or "foes". Social valence was implicitly manipulated before the fMRI scan by means of a simple multi-round dictator game using the stimulus faces as co-players. They either generously gave points to the participants ("friends") or retained most of the points for themselves ("foes"). All participants were able to make the distinction between "friends" and "foes" as reflected in their own performance during the game.
For both TDC and ASD subjects we observed significant activation in areas previously associated with the hMNS (right inferior frontal cortex, including pars opercularis, and right premotor cortex (BA 6)), as well as in the right amygdala and bilateral occipital cortex, extending into superior and inferior temporal cortex (conjunction analysis of observation and imitation versus baseline, p<.05 voxel-level, FWE-corrected for the whole brain). In the direct group comparison, no significant decrease in activation could be observed for ASD in areas associated with the hMNS, even at a very low threshold (p<.05 uncorrected, voxel level). In TDC we observed a significantly stronger modulation of brain activation by social valence in the right amygdala as compared to participants with ASD (p<.05 voxel level, FWE-corrected for ROI). In TDC, activation in this area was stronger for facial expressions which were congruent to the social valence of the stimuli (i.e. smiling "friends" and frowning "foes", p<.05 voxel level, FWE-corrected for ROI).
Our results argue against a fundamental deficit of the hMNS in ASD. Irrespective of social valence, areas previously associated with the hMNS were activated during observation and imitation of videos with facial expressions. The modulation of amygdala activation by the social valence of stimulus faces suggests a role for the amygdala in detecting congruency of social valence and displayed emotion. This effect was absent in participants with ASD suggesting a lack of using social valence information in face-to-face situations.
See more of: Brain Imaging
See more of: Brain Structure & Function