Increased gyrification, but comparable surface area in adolescents with autism spectrum disorders

Abstract

Autism spectrum disorders are associated with atypically excessive early brain growth. Recent studies suggest that later cortical development, specifically cortical thickness, during adolescence and young adulthood is also aberrant. Nevertheless, previous studies of other surface-based metrics (e.g. surface area and gyrification) at high-resolution in autism spectrum disorders are limited. Forty-one males with autism spectrum disorders and 39 typically developing males matched on age (mean ≈ 17; range = 12-24 years) and IQ (mean ≈ 113; range = 85-143) provided high-resolution 3 T anatomical magnetic resonance imaging scans. The FreeSurfer image analysis suite quantified vertex-level surface area and gyrification. There were gyrification increases in the autism spectrum disorders group (relative to typically developing subjects) localized to bilateral posterior cortices (cluster corrected P < 0.01). Furthermore, the association between vocabulary knowledge and gyrification in left inferior parietal cortex (typically developing group: positive correlation; autism spectrum disorders group: no association) differed between groups. Finally, there were no group differences in surface area, and there was no interaction between age and group for either surface area or gyrification (both groups showed decreasing gyrification with increasing age). The present study complements and extends previous work by providing the first evidence of increased gyrification (though no differences in surface area) at high resolution among adolescents and young adults with autism spectrum disorders and by showing a dissociation in the relationship between vocabulary and gyrification in autism spectrum disorders versus typically developing subjects. In contrast with previous findings of age-related cortical thinning in this same autism spectrum disorders sample, here we find that increases in gyrification are maintained across adolescence and young adulthood, implicating developmentally dissociable cortical atypicalities in autism spectrum disorders.

Keywords: MRI; autism; cortical folding; gyrification; surface area.

Figure 1

Coronal slice of MRI scans showing greater gyrification in left precuneus for (A) a male with an autism spectrum disorder versus (B) a typically developing male.

Figure 2

Inflated and pial surface maps (dark grey = sulci; light grey = gyri) of both the (A) lateral and (B) medial surfaces of the left hemisphere and the (C) lateral surface of the right hemisphere showing greater cortical gyrification in the autism spectrum disorders group as compared to the group of typically developing control subjects.

Figure 3

Inflated and pial surface maps (dark grey = sulci; light grey = gyri) of the lateral and medial views of the (A) left and (B) right hemispheres showing decreasing cortical gyrification with increasing age in the combined group of individuals with autism spectrum disorders and typically developing control subjects.

Figure 4

(A) Inflated and pial surface maps (dark grey = sulci; light grey = gyri) of the cluster-corrected (P < 0.01) group difference in associations between gyrification and Vocabulary score (peak: x = −44.8, y = −55.7, z = 40.7) and (B) scatterplot of these correlations for each group (autism spectrum disorders group r² = 0.03; typically developing group r² = 0.24). TD = typically developing.

Figure 5

Within the typically developing males only, (A) inflated and pial surface maps (dark grey = sulci; light grey = gyri) of the cluster-corrected (P < 0.01) association between gyrification and Vocabulary score (peak: x = −46.6, y = −40.1, z = 40.6) and (B) scatterplot of this correlation (R² = 0.26).

Figure 6

Within the typically developing males only, (A) inflated and pial surface maps (dark grey = sulci; light grey = gyri) of the cluster-corrected (P < 0.05) association between gyrification and total score from the Social Responsiveness Scale (peak: x = 45.8, y = −39.0, z = 12.4) and (B) scatterplot of this correlation (R² = 0.16), which does not change after removing the three highest gyrification values.