Cerebellar Grey Matter and Lobular Measures Correlate with Core Autism Symptoms

Friday, May 16, 2014
Atrium Ballroom (Marriott Marquis Atlanta)
A. M. D'Mello1, D. Crocetti2, S. H. Mostofsky2 and C. J. Stoodley3, (1)Department of Psychology, American University, Washington, DC, (2)Laboratory for Neurocognitive and Imaging Research, Kennedy Krieger Institute, Baltimore, MD, (3)Psychology, American University, Washington, DC
Background: Anatomical differences in the cerebellum are among the most consistent post-mortem findings in autism spectrum disorder (ASD). Previous structural neuroimaging findings show grey matter (GM) decreases in lobules VII (Crus I), VIII and IX, VIII of the cerebellum. Similarly, individuals with autism show reduced activation in right anterior cerebellum, Crus I, and Crus II during a variety of motor, executive functioning and language tasks. In clinical studies, malformations and acquired lesions in the cerebellum can result in ASD symptomology. In particular, lesions affecting the posterior cerebellum can result in language difficulties, affective dysfunction, and problems in executive functioning – symptoms relevant to ASD. Recent evidence also suggests that there are functional subregions within the human cerebellum for sensorimotor and cognitive processing; the anterior lobe connects to primary sensorimotor areas of the cerebral cortex, whereas the posterior lobe connects to prefrontal and parietal regions. This provides an important framework in interpreting cerebellar findings in ASD.

Objectives: We investigated cerebellar regional grey matter (GM) and volumetric measurements of the cerebellar lobules in ASD children compared to typically-developing (TD) children, and examined the relationship between cerebellar structure and core ASD symptoms. 

Methods: Voxel-based morphometry (VBM) was used to compare whole-brain GM in 35 ASD and 35 TD children (mean age 10.4 ±1.6 years; range 8-13 yrs). The cerebellar SUIT atlas was used to compute volumetric measurements of individual cerebellar lobules. Correlations were calculated between scores on the Autism Diagnostic Observation Schedule (ADOS) and Autism Diagnostic Interview (ADI) and the VBM and volumetric data.

 Results: VBM revealed reduced GM in ASD children bilaterally in cerebellar lobule VII (Crus I/II).  More impaired scores on the ADOS Social Interaction subscale correlated with reduced GM in seven cerebellar clusters, including right Crus I/II.  More impaired scores on the ADOS Social-Communication subscale correlated with reduced GM clusters in right Crus I/II and VIIIA/VIIIB. Poorer ADOS Stereotyped Behaviors & Restricted Interests scores correlated with reduced GM in regions of right Crus I/II. In addition, more impaired scores on the ADI Social Interaction and Restrictive, Repetitive & Stereotyped Behaviors subscales correlated with reduced GM in a cluster in the anterior cerebellum. Consistent with the VBM findings, the SUIT volumetric analysis revealed smaller right Crus I in the ASD group. Volumetric analyses also revealed larger vermis lobules VIIIA and VIIIB in children with ASD. Smaller vermis VI volume correlated with more impaired ADOS Social Interaction and ADI Communication scores. Poorer ADOS Stereotyped Behaviors & Restricted Interests scores were associated with smaller volumes in lobule VIIIB bilaterally. Larger vermis VIIB and VIIIA volumes were associated with worse ADI Restrictive, Repetitive & Stereotyped Behaviors.

Conclusions: Using two analytic approaches, we showed reduced cerebellar Crus I/II GM in ASD, a region which connects to prefrontal and parietal association areas. Importantly, cerebellar GM volume and lobule volumes significantly correlated with ASD severity, providing further evidence of a role for the cerebellum in ASD etiology.