Influence of Speech Onset Delay on Cortical Gyrification in Adolescent and Young Adults with Autism Spectrum Disorders

Background: Autism spectrum disorders (ASD) particularities in cortical organization are not restricted to early childhood. For instance, gyrification, defined as the ratio of the inner to the outer cortical contour, locally differs between ASD and typically developing adolescents and young adults (Libero et al., 2014; Schaer et al., 2013; Wallace et al., 2013). Wallace et al. (2013) have reported gyrification increases in association cortices in ASD together with a discrepancy in the association between vocabulary scores and gyrification. Others have also found frontal gyrification increases in autism but not Asperger (Jou et al., 2010). This difference might be underlined by diverging language development, as symptomatic and structural heterogeneity in autism is believed to arise from increased neuro-plasticity that could target either speech or perceptual brain regions (Mottron et al., 2014).

Objectives: To examine whether the categorization of ASD individuals according to the presence or the absence of a speech onset delay can account for gyrification variability in ASD, especially in the perceptual association cortices.

Methods: High-resolution anatomical magnetic resonance imaging T1-weighted scans were obtained from 41 typically-developing young adults (5 females) and 61 individuals with ASD (7 females). ASD Subjects were divided according to the report of a Speech Onset Delay during childhood (SOD group, N=31) or not (NoSOD group, N=30) regardless of their DSM IV subgrouping (autistic or Asperger). The three groups were matched on age (14-38, mean≈22), FSIQ (66-131, mean≈104) and handedness. Following the 3-D methodology by Schaer et al. (2008), accurate measurements of local Gyrification Indexes (lGI) were computed using FreeSurfer. Age influences were modeled separately for each group. Group*age interaction and main effect of group on local gyrification were explored. Clusters of significant differences are reported at p=.01 but were not corrected for multiple comparisons.

Results: Both ASD groups showed areas of higher lGI compared to controls: right anterior cingulate and precuneus (SOD) as well as several clusters in bilateral medial and orbito-frontal cortex, right precuneus, superior frontal and temporal cortices (noSOD). Differences between the ASD groups and controls in relation to age were observed, in orbito-frontal, superior frontal, middle temporal, anterior cingulate and mainly in the right precuneus, where the unfolding rate was lower in controls compared to both the SOD and noSOD groups. Gyrification was higher in the noSOD versus the SOD group in some frontal and temporal clusters, in bilateral occipital cortices and in a larger region encompassing parts of the right inferior parietal and occipital lobes. This perceptual region overlaps with the most extended cluster showing age-related folding differences, i.e. where gyrification decreased faster in the noSOD than in the SOD group.

Conclusions: While some regions of increased gyrification in ASD correspond to recent results, those involving frontal and cingulate cortices, together with differences between ASD subgroups based on language acquisition in age-related changes, represent new findings. Evidencing the relevance of gyrification measures in ASD structural studies, this work, by detecting differences in parieto-occipital perceptual associative regions between ASD subgroups also highlights the importance of considering speech onset categorization in autism research.