Brain Activation During Inferential and Theory of Mind Processing In Children with Autism

Background: Individuals with autism have a deficit in Theory of Mind (ToM) processing manifested both behaviorally and in brain activation.  Adults with autism were reported to indiscriminately engage the right hemisphere ToM network during discourse processing that required inferencing about physical, emotional, and mental states, irrespective of the demand for ToM processing (Mason et al., Neuropsychologia, 2008).

Objectives: This functional magnetic resonance imaging (fMRI) study used a paradigm that required inferential processing in physical and social conditions with children with autism to examine whether the inefficient use of the ToM network occurs earlier in the developmental process.  Because children with autism may have more difficulty with reading than typically developing (TD) children, the paradigm used visual scenarios.

Methods: Twelve children with autism (ages 8 to 16 years, mean = 13.08; FSIQ range 87-135, mean 112.08) and 13 TD children (ages 7 to 15 years, mean = 11.23; FSIQ range 97-128, mean = 114.92) participated in this event-related fMRI study.  The two groups were matched on age, gender, VIQ, PIQ, and FSIQ.  Participants simultaneously viewed two pictures labeled as the first and second of a sequence for 6s. Then, they read the question “What comes next?” for 2s. Two more pictures were displayed for 4s while the participants selected which image logically completed the sequence. Ten trials required knowledge about basic physical properties (e.g., that a falling egg would crack). The other ten trials required social knowledge, particularly an understanding of other people’s intentions and mental states (e.g., a boy would cry after a girl knocks over his block tower). Brain activation was analyzed during the viewing of the first two pictures, when inferential processing was thought to occur.

Results: In the Social-Physical comparison within each group, both the autism and the TD group had brain activity in bilateral posterior superior temporal sulcus (pSTS) and right middle temporal gyrus (p<.001). In addition, the autism group had bilateral activation in inferior frontal gyrus and medial frontal gyrus.  In the reverse Physical-Social comparison, both groups had frontal and parietal activation; however, this activation was bilateral in the TD group but left-lateralized in the autism group. In the between-group comparison for Social-Physical, the autism group had greater bilateral inferior frontal gyrus activity than the TD group (p<.005). In the group comparison for Social-Fixation, the autism group had reliably less activation than the TD group in right pSTS (p<.005). Both groups performed the tasks without difficulty (Physical means: autism 93%, TD 95%; Social means: autism 91%, TD 94%).

Conclusions: Both groups differentially used ToM areas during the Social condition. The more extensive brain activity pattern of the autism group during inferencing about Social scenarios, as indicated by the recruitment of frontal language and medial frontal areas, suggests that they required more support from reasoning and language systems to perform the ToM tasks.