Atypical Corpus Callosum Development Associated with Autism in Infants and Toddlers

Background:  Connecting left and right cerebral hemispheres, the corpus callosum is the largest white matter pathway in the human brain. This connective structure begins to develop prenatally and undergoes rapid change over the first postnatal year of life. Numerous brain imaging studies have suggested that the corpus callosum is relatively smaller in older children and adults with autism spectrum disorder (ASD). However, there are no published studies examining the morphological development of this connective pathway in infants and toddlers.

Objectives:  To characterize corpus callosum development among infant siblings at low- and high-familial risk for ASD from 6 to 24 months age using fully automated, reproducible approaches to brain image morphometry.

Methods: Structural magnetic resonance imaging data were collected longitudinally from 254 high-risk infant siblings (by virtue of having an older sibling with ASD) and 101 low-risk controls at 6, 12, and 24 months age. Fifty-four children met criteria for ASD based on clinical best estimate diagnosis at age 2 years, yielding 3 groups (low risk, high-risk non-ASD, and high-risk ASD). An automated, inter-hemispheric connectivity-model-based probabilistic subdivision of the midsagittal T1 image was performed to measure total and regional corpus callosum areas. Following this analysis, corpus callosum thickness was measured across 100 equidistant points using a contour-based model. Data were analyzed using a longitudinal mixed-model approach controlling for site, sex, IQ, and total brain volume.

Results: We found increased corpus callosum area in children with ASD starting at 6 months age (p = .01). Probabilistic subdivision data suggested this difference was attributable to the anterior- and posterior-frontal regions of the corpus callosum. Thickness data revealed increased total thickness in children with ASD starting at 6 months age, p = .0001. FDR corrected comparisons indicated a general pattern of high-risk ASD > high risk non-ASD > low risk controls. As with the area data, this difference was strongest in anterior regions of the corpus callosum, and was most pronounced in the high-risk ASD vs. low-risk control comparison. Although there were no significant effects for group X time, growth curves suggested that differences may diminish over the age interval studied.

Conclusions:  In contrast to previous work among older children and adults, our findings suggest that the corpus callosum may be larger in infants who go on to develop ASD. This result was apparent whether or not we controlled for factors including total brain size and sex. The contrast between the current findings with the existing literature may reflect developmental phenomenon. While corpus callosum size in older children is driven in large part by use-dependent myelination and growth of commissural axons, much of early corpus callosum development is characterized by robust axon pruning. In this regard, the present findings are consistent with theory and related findings specific to connectivity in very young children with ASD.