Brain Activity and Local Connectivity Underlying Action Understanding in Autism Spectrum Disorders

Background: Detecting the means (how) and intent (why) of others’ actions may involve motor simulation of their action, possibly mediated by the mirror neuron system. Another neural network that may be involved in this process is the theory-of-mind (ToM) network consisting of regions, such as the temporoparietal junction (TPJ), superior temporal sulcus (STS), and medial prefrontal cortex (MPFC). While functional deficits in both of these systems have been reported in people with ASD, they are seldom studied in conjunction. 

Objectives: The main objective of this study was to characterize the neural network involved in interpreting the how and why of actions in high-functioning adults with ASD 

Methods: fMRI data were acquired from 24 high-functioning adults with ASD and 26 typically developing (TD) control participants while they made action and intention judgments about a series of static images of a model using household objects. The participants’ task was to view the model’s action and determine whether the means (how the action was carried out) or intent (the model’s goal) of the action was ordinary or unusual. fMRI data were preprocessed using AFNI and FSL. Local connectivity was calculated using the Regional Homogeneity (ReHo) approach for every voxel and its 26 immediate neighbors. This analysis included low-pass filtering, “scrubbing”, and removal of task-related effects. In addition, parameter estimates extracted from activated brain regions were correlated with assessment scores (Reading the Mind in the Eyes [RME] test, Empathy Quotient [EQ], and the Ritvo Autism Asperger Diagnostic Scale-Revised [RAADS-R]) to establish brain-behavior relationships. 

Results: Overall, interpreting unusual actions generated more activity in both ASD and TD groups, compared to ordinary actions. The TD participants, relative to ASD adults, showed increased activation in right middle frontal gyrus while processing unusual actions. The ASD group, on the other hand, showed increased activation in left middle occipital area for processing intentions. ReHo analysis revealed overconnectivity in participants with ASD in right middle frontal gyrus (p <.05, FWE corrected). Significant positive correlations were found between the RME scores and activation in left fusiform and bilateral superior temporal gyri (STG) in the TD groups; ASD group showed positive correlation between RME and right fusiform activation. 

Conclusions: Our findings of altered activation in occipital and frontal regions in the ASD group may suggest more strategic and visual approach to interpret actions. The overconnectivity in RMFG (BA8, BA9) in participants with autism is consistent with previous findings of excess frontal connectivity (Courchesne & Pierce, 2005), increased neuron number (Courchesne et al., 2011), and increased number of cortical minicolumns (Casanova, 2004) in frontal regions in ASD. While TD participants had a significant positive predictive relationship between RME scores and STG activity, this relationship was not seen in ASD adults, perhaps suggesting their reliance on altered neural routes in reading intention from action. Given that we did not find a group difference in  mirror neuron system activity, this neural system may not differ in high-functioning ASD individuals while interpreting others’ actions.