Cortical Thickness Differences in Autistic Children With and Without Intellectual Disability

Abstract

Of the 1 in 36 individuals in the United States who are diagnosed with autism spectrum disorder, nearly 40% also have intellectual disability (ID). The cortex has been widely implicated in neural processes underlying autistic behaviors as well as intellectual ability. Thus, neuroimaging features such as cortical thickness are of particular interest as a possible biomarkers of the condition. However, neuroimaging studies often fail to include autistic individuals with ID. As a result, there are few studies of cortical thickness in autistic individuals across the entire range of intellectual abilities. This study used MRI to evaluate cortical thickness in young autistic children (n = 88, mean age 5.37 years) with a large range of intellectual ability (IQ 19-133) as well as nonautistic, nondevelopmentally delayed (referred to here as typically developing [TD]) peers (n = 53, mean age 5.29 years). We first investigated associations between full scale IQ and cortical thickness in both autistic and TD children. Autistic children had significant negative associations (i.e., thinner cortex, higher IQ) in bilateral entorhinal cortex, right fusiform gyrus, superior, middle and inferior temporal gyri, and right temporal pole that were not present in TD children. Significantly thicker cortex was also observed in these regions for autistic children with ID (i.e., IQ ≤ 70) compared with those without. Last, given the reported correspondence between the severity of autism symptoms and intellectual ability, we compared cortical thickness associations with both IQ and ADOS Calibrated Severity Scores and found these patterns overlapped to a significant degree across the cortex.

Keywords: IQ; MRI; autism; cortical thickness; intellectual disability.

FIGURE 1

Random field theory corrected (top) and uncorrected (bottom) maps indicating cortical regions (A) with a significant negative association between cortical thickness and full‐scale IQ (FSIQ) in autistic (ASD) participants and (B) regions where autistic individuals without ID had significantly thicker cortex compared with autistic individuals with intellectual disability (ID). Maps are shown in both folded and inflated views. Clusters > 100 vertices are visualized.

FIGURE 2

Uncorrected main effect of full‐scale IQ maps for autistic children with (ID) and without (noID) intellectual disability (ID). Spin tests found no spatial correspondence between these two maps above chance level, suggesting that autistic children with and without ID associations between cortical thickness and IQ that do not overlap to a significant degree. Maps are shown in both folded and inflated views. Clusters > 100 vertices are visualized.

FIGURE 3

Correspondence between cortical thickness associated with full‐scale IQ and autism symptom severity: Uncorrected main effect maps of full‐scale IQ (FSIQ) and ADOS calibrated severity scores (CSS) were found to correspond to a greater degree than would be expected by chance (spin test p < 0.001) indicating that autism symptoms and intellectual ability are largely associated with similar cortical regions. Maps are shown in both folded and inflated views. Clusters > 100 vertices are visualized.