Background:
Ringo et al (1991) hypothesized that, due to the larger metabolic costs and conduction delays associated with long-distance connections, larger brains would show decreased long-distance connectivity. That hypothesis is supported by computational modeling (Ringo, 1991), MRI across species with a wide range of brain sizes (Rilling and Insel, 1999), and MRI from adults and children (Jancke et al, 1997; 1999). Recent research has, moreover, shown that this scaling relationship develops during childhood (Lewis et al, 2007). Objectives: This motivates the hypothesis that the abnormal early brain overgrowth seen in autism (Courchesne et al, 2001; Hazlett et al, 2005) will lead to a lesser degree of connectivity — consistent with findings of underconnectivity in autism (Lewis et al, 2003, 2004; Just et al, 2004; Herbert et al, 2005). The prediction tested here is that individuals with autism should show a scaling relation between connection length and degree of connectivity, as do controls, but should show a lesser degree of connectivity, overall.
Methods: Using diffusion tensor imaging (DTI) and tractography to detail the patterns of connectivity of the corpus callosum, and to estimate the length of interhemispheric connections in each of five sub-regions, we investigated the relation between the length of the connections and the degree of connectivity in 9 individuals with autism and 22 controls. Regression analyses were used to assess the scaling relation in each of the five sub-regions in both groups, and between group differences were assessed with analysis of variance tests.
Results: Regressions between connection length and degree of connectivity were significant in anterior and posterior regions in both individuals with autism and in controls. Anterior regions showed a significantly reduced degree of connectivity in the autistic group.
Conclusions: The results support the hypothesis that the early brain overgrowth in autism leads to reduced connectivity.