Correlation Between Cerebellar White Neuroanatomy and a Motor Coordination Task in Autism Spectrum Disorder
Structural neuroimaging studies suggest that Autism Spectrum Disorder (ASD) is accompanied by neuroanatomical differences in the cerebellum. For example, it has been shown that total cerebellar volume is significantly reduced in ASD (Hallahan et al., 2009). Also, a reduction in total cerebellar white matter volume in ASD has been observed (McAlonan et al., 2002).
However, most prior studies were based on measures of cerebellar volume, which is a neuroanatomical highly unspecific feature as it is computed as a product of cortical thickness and surface area. Also, most studies investigated the cerebellum globally (e.g. total cerebellar volume) rather than locally (i.e. different cerebellar sub regions). In addition, it is currently unknown how neuroanatomical cerebellar differences mediate specific autistic symptoms such as impaired motor control and coordination.
Here we aim to establish neuroanatomical differences in the cerebellum between individuals with ASD and controls using a spatially unbiased, surface-based approach based on the cerebellar white matter, and to correlate distinct morphometric features with autistic symptoms.
We included 75 male adults, 38 diagnosed with ASD and 37 healthy controls who did not differ significantly in mean age and full scale IQ (respectively 26 + 7 years and 110 + 14; and 28 + 6 years and 114 + 14). The FreeSurfer image analysis suite was used to derive 3D models of the cerebellar white matter surface in each T1-weighted image. Initially, we examined between-group differences in total cerebellar white matter volume. At each cerebellar vertex (i.e. point on the surface), we also examined between-group differences in curvature (C), cortical folding (i.e. sulcal depth) (SD) and surface area (SA) using an exploratory vertex-level threshold of P < .05 (uncorrected). Furthermore, correlations between distinct neuroanatomical features and performance in the Purdue Pegboard test were examined at each cerebellar vertex using a general linear model (GLM) with a main effect of group, a main effect of task performance, and their interaction. A t-test for independent samples was used to compare task performance between groups.
There were no significant differences in total cerebellar white matter volume between individuals with ASD and controls (p=0.076). However, we observed significant between-group differences in geometric features of the cerebellar white matter including curvature, sulcal depth and surface area, which occurred predominantly in cerebellar hemisphere lobule V and VI (p<0.05, uncorrected). Individuals with ASD performed significantly worse in the Perdue Pegboard test than controls (t=3.578, p=0.001). Last, we found a significant association between morphometric measures and general task performance in cerebellar hemisphere lobules IV, V, VI and VII, where we also found a significant interaction between performance measures and diagnostic category.
In summary, our data suggests that individuals with ASD have significant differences in the geometry of their cerebellar white matter, which may contribute to impairments in motor coordination. Our study therefore provided an important first step into describing cerebellar white matter neuroanatomy using a surface-based approach in individuals with ASD, which may aid future investigations into the specific underlying neural mechanisms of ASD.