Neuroanatomical and neurofunctional markers of social cognition in autism spectrum disorder

Abstract

Social impairments in autism spectrum disorder (ASD), a hallmark feature of its diagnosis, may underlie specific neural signatures that can aid in differentiating between those with and without ASD. To assess common and consistent patterns of differences in brain responses underlying social cognition in ASD, this study applied an activation likelihood estimation (ALE) meta-analysis to results from 50 neuroimaging studies of social cognition in children and adults with ASD. In addition, the group ALE clusters of activation obtained from this was used as a social brain mask to perform surface-based cortical morphometry (SBM) in an empirical structural MRI dataset collected from 55 ASD and 60 typically developing (TD) control participants. Overall, the ALE meta-analysis revealed consistent differences in activation in the posterior superior temporal sulcus at the temporoparietal junction, middle frontal gyrus, fusiform face area (FFA), inferior frontal gyrus (IFG), amygdala, insula, and cingulate cortex between ASD and TD individuals. SBM analysis showed alterations in the thickness, volume, and surface area in individuals with ASD in STS, insula, and FFA. Increased cortical thickness was found in individuals with ASD, the IFG. The results of this study provide functional and anatomical bases of social cognition abnormalities in ASD by identifying common signatures from a large pool of neuroimaging studies. These findings provide new insights into the quest for a neuroimaging-based marker for ASD. Hum Brain Mapp 37:3957-3978, 2016. © 2016 Wiley Periodicals, Inc.

Keywords: activation likelihood estimation; autism; brain; meta-analysis; neuroimaging; social brain; social cognition.

Figure 1

Examples of social cognition tasks used in studies included in the meta‐analysis. [Color figure can be viewed at http://wileyonlinelibrary.com.]

Figure 2

ALE estimation of social brain activity across ASD, TD, and ASD and TD participants combined (p < 0.05, FDR cluster‐forming threshold). Activity is seen in regions, such as the MPFC, bilateral STG, posterior cingulate/precuneus, fusiform gyrus, and bilateral inferior frontal gyrus. [Color figure can be viewed at http://wileyonlinelibrary.com.]

Figure 3

ALE analysis for TD > ASD (orange) and ASD > TD (green) group differences across studies: (p < 0.05, FDR cluster‐forming threshold). [Color figure can be viewed at http://wileyonlinelibrary.com.]

Figure 4

Between‐group differences in social task requiring face processing (A), and social tasks that do not require face processing (B). TD > ASD (orange), ASD > TD (green). All areas p < 0.05, FDR cluster‐forming threshold. [Color figure can be viewed at http://wileyonlinelibrary.com.]

Figure 5

Group differences in surface area (top) and cortical thickness (bottom) between a sample of T1 images of ASD and TD participants within social brain ROIs computed from the ALE mask. Red denotes decrease ASD and blue denotes increases in ASD. [Color figure can be viewed at http://wileyonlinelibrary.com.]

Figure 6

The social brain mask produced across theory of mind task type across both TD and ASD participants is displayed as a yellow overlay. Results of the surface based analysis on the mask found regions of decreased surface area (dark blue), volume (green) and increased thickness (pink). This is displayed in conjunction with ALE computed VBM meta‐data displaying decreased (dark blue) and increased (red) volume in ASD. [Color figure can be viewed at http://wileyonlinelibrary.com.]