Neuroanatomical Correlates of Cognitive Flexibility in Adolescents with Autism and the Broader Phenotype

Background:  

Autism is associated with impairment in some aspects of executive function. Conflicting reports concern whether subtle or related deficits are present in their unaffected first degree relatives. Structural and functional brain imaging correlates of executive function have been identified and found to differ in individuals with autism.

Objectives:  

To explore differences in cognitive flexibility in adolescents with autism and their siblings compared to controls. Structural brain imaging correlates of executive function were investigated and compared between the groups.

Methods:  

40 adolescents with autism, 40 of their unaffected siblings, and 40 typically developing controls completed an intra-dimensional/ extra-dimensional shift task (ID/ED), a test of set-shifting ability. Structural MRI data was acquired which was analysed using the multivariate analysis partial least squares (PLS). This was further followed by a permutation test of familial variation of the MRI markers.

Results:  

Significant differences in performance were found between the autism and control groups on the ID/ED (p =.003), at the highest level of difficulty. Significant differences in performance on these measures were not apparent between the siblings and controls. PLS identified a grey matter system associated with set-shifting performance. Further analysis revealed significant correlation between grey matter volume in this system and performance on the ID/ED task; however this was only true for the control group. Variation in grey matter volume within this system was significantly more alike between related sibling pairs compared to randomly permutated sibling pairs, as determined by the familial permutation test (p=0.008).

Conclusions:  

Impairments in set-shifting were found in individuals with autism. In contrast, set-shifting ability was intact in unaffected siblings. Set-shifting per se therefore does not qualify as a cognitive endophenotype for autism. We found an association between brain structure and cognitive flexibility in the control group, an association that was not apparent in the autism group. This suggests that the structure to function relationship evident in controls is disrupted in autism. Grey matter volume within this system was found to be heritable between autism and sibling groups which could be further explored for its value as a candidate endophenotype for autism.