Age-Related Brain Changes Associated with Social Impairments In Children, Adolescents and Adults with Autism Spectrum Disorders

Background: Brain regions involved in perceiving and recognizing affect show structural abnormalities in adults with autism spectrum disorders (ASD) and are related to impaired social cognition. It remains unclear whether these structural atypicalities change with age or are associated with ASD symptomatology. 

Objectives: We assessed the relation between the severity of social deficits and brain morphometry in cohort of individuals with ASD that spanned a large age range. 

Methods:  28 children, adolescents and adults between the ages of 7 and 39 (mean age=22 ±8 years; 22 males) who carried a clinical diagnosis of ASD, according to the Diagnostic and Statistical Manual of Mental Disorders, 4^th Edition (DSM-IV), were enrolled in this study.  Their diagnoses were then confirmed using the Autism Diagnostic Observational Schedule - Generic (ADOS-G) and the Autism Diagnostic Interview - Revised (ADI-R). High-resolution anatomical Magnetic Resonance Imaging (MRI) scans were obtained for each participant. Cortical thickness measurements were made on the anatomical MRIs using an automated pipeline. The cortical thickness data were analyzed using a general linear model controlling for sex. We performed a directed search in limbic areas (rostral and dorsal anterior cingulate cortices, medial prefrontal cortices, orbitofrontal cortices), and brain areas implicated in social cognition (inferior frontal gyri, superior temporal gyri, fusiform gyri) correcting for multiple comparisons using the False Discovery Rate (q=0.05). Scores for the social domain of the ADI algorithm were correlated with the thickness with the ROIs. 

Results:  Greater impairment on the social domain of the ADI was significantly correlated with increased cortical thickness in the rostral ACC, even when the variance associated with age was removed from the model. A significant interaction was found between the age of the participants and the social scores on the ADI and thickness in the left orbitofrontal cortex, suggesting that the relation between cortical thickness in this region and impaired social cognition varies with age. Younger participants with greater social cognition impairments showed decreased thickness in the orbitofrontal cortex relative to adults. 

Conclusions: Atypical ACC morphology is associated with impaired social cognition in children, adolescents and adults with ASD; however this abnormality was not related to age and may be a result of abnormal development in this region. Interaction effects between age and social cognition were seen in the orbitofrotal cortex, a region involved in social functioning and reward. Previous work has theorized that socialization deficits may be due to disruptions in connections between the orbitofrontal cortex and the amygdala. Given the strong age-related changes in the orbitofrontal cortex seen in the current study, future work should examine fronto-amygdalar connections across the life-span in ASD.