Large-Scale Cortical Functional Connectivity During Social Scene Processing in ASD

Background:  From appropriate perception and understanding of another individual’s non verbal body behaviour, and its association to various element of the environmental context, humans are able to gain crucial knowledge about the environment and use it to subsequently adapt their own behaviour successfully. This results in a continuous need to rapidly and accurately monitor social contexts and meanings. Two cognitive systems might have a complementary role in understanding social scenes. Motor and emotional perception and simulation mechanisms may provoke a pre-reflective or automatic representation of the visual scene, whereas inferential mechanisms may then elaborate on this initial hypothesis by using previously acquired conceptual knowledge. An altered capacity to interpret socio-emotional context and use it to adapt their behavior accordingly is a hallmark characteristic of individuals with ASD’s behavior. 

Objectives:  We used functional magnetic resonance imaging (fMRI) to assess the functional connectivity of brain networks involved in the explicit processing of emotional information presented in realistic movie clips depicting everyday social interactions. Because of its role in explicit emotional attribution, we hypothesized that the Dorsomedial Prefrontal cortex (DMPFc) would be abnormally activated and functionally connected in participants with ASD compared to controls.

Methods:  Twelve right-handed male adult volunteers (mean age = 21 ± 2 years) and twelve right-handed adults with high functioning autism (HFA) or Asperger Syndrome (AS) (autistic spectrum disorders (ASD); mean age = 26 ± 10) participated in the study. Structural and functional images were acquired on a 3 Tesla Bruker scanner. Stimuli consisted of a set of sixty four 9 sec. video clips of everyday socio-emotional events. We contrasted experimental conditions where subjects were required to process emotional information explicitly or implicitly. To explore large scale connectivity within brain networks, we performed functional connectivity modelling using an atlas of 152 regions of interest derived from a morphometric parcellation of the brain of each individual subject.

Results:  Performance levels were identical for both groups, suggesting that adults with ASD are perfectly able to recognize other’s emotional behavior when asked to do so explicitly. However, this intact performance in explicit attribution of emotion was associated with an absence of activation of the DMPFc in the ASD group. Furthermore, large-scale functional connectivity analysis revealed abnormal long-range connectivity in the antero-posterior axis, more specifically between the DMPFc, lateral prefrontal cortex and precuneus.

Conclusions:  Our results using whole social scenes confirm those obtained in previous studies using emotional faces or body movements. A network of brain regions including the DMPFc seem to be consistently abnormally activated and functionally connected in ASD, suggesting a absence of top-down influence that lead to a difficulty to attribute emotional valence to the otherwise normally perceived stimuli. We discuss these results by underlying the striking similarity between this brain network and both the ‘resting state’ and the ‘contextual’ network. Abnormal emotional processing in ASD may be related to abnormal introspection and an altered capacity to associate the stimulus to its affective meaning by reactivating previously acquired conceptual emotional knowledge.