Neural Bases of Self Representation in High-Functioning Autism

Background: Previous research has indicated that impairments in self understanding in persons with Autism Spectrum Disorders (ASD) may contribute to difficulties in joint attention, social learning, and social cognition. Self-referential processing entails the ability to process information by incorporating self awareness and self memory. In neurotypical individuals, self-referential processing has been linked to several brain areas, including the dorsal medial prefrontal gyrus, left inferior frontal gyrus, anterior and posterior cingulate cortices, and precuneus. Individuals with ASD have been found to show atypical activation pattern in some of these areas (Craik et al., 1999). In addition, activation in these regions has been correlated with difficulties in social communication and social cognition in persons with ASD (Ohnishi, 2000).

Objectives: The primary objective of this study was to measure the neural substrates underlying self representation in persons with ASD. This study also examined whether self-referential processing leads to enhanced memory in ASD, as it does in neurotypical individuals.

Methods: Eight high-functioning adolescents and adults with ASD and twelve age and IQ matched neurotypical individuals participated in this fMRI study (data collection is ongoing). Participants completed three judgment tasks with a button response of "yes" or "no": 1) the ‘self’ block in which participants judged whether the adjective described the participant, 2) the ‘other’ block in which participants judged whether the adjective described their favorite teacher, and 3) the ‘letter’ block in which participants judged whether the adjective contained the letter “e”. Stimuli were positive and negative adjectives. The three judgment tasks were completed in a blocked design format with three blocks for each type of judgment. The fMRI data were acquired from a Siemens 3T Allegra scanner and analyzed using SPM2 software. After the scanning session, participants completed a recognition memory test consisting of 90 adjectives seen during the scan and 90 new adjectives. Additional measures were collected, including fluid reasoning, working memory, and symptomotology surveys and rating scales.

Results: No differences in participant responses to stimuli were found. Compared to controls, participants with ASD showed reliably less activation in several areas while processing self and other adjectives compared to performing letter identification. These include less activation in the medial prefrontal cortex, an area previously associated with social-cognition and self-referential thinking, and the left inferior frontal and left posterior superior temporal regions, areas associated with language processing. In ASD, there was no difference in the number of words recalled in “self” and “other” conditions during the memory test (in contrast to previous research in this area), while controls showed an advantage in memory for words in the self conditions.

Conclusions: These findings suggest less socially and semantically oriented processing of information related to self and others in persons with ASD. They also underscore the important role the medial prefrontal cortex plays in social cognition and self-referential thinking. Moreover, the behavioral (memory) and neuroimaging (hypoactivation of the medial prefrontal cortex) results suggest that self-referential processing may not have a privileged neural status in persons with autism.