Cerebellar function in children with and without dyslexia during single word processing

Abstract

The cerebellar deficit hypothesis of dyslexia posits that dysfunction of the cerebellum is the underlying cause for reading difficulties observed in this common learning disability. The present study used functional magnetic resonance imaging (fMRI) and a single word processing task to test for differences in activity and connectivity in children with (n = 23) and without (n = 23) dyslexia. We found cerebellar activity in the control group when word processing was compared to fixation, but not when it was compared to the active baseline task designed to reveal activity specific to reading. In the group with dyslexia there was no cerebellar activity for either contrasts and there were no differences when they were compared to children without dyslexia. Turning to functional connectivity (FC) in the controls, background FC (i.e., not specific to reading) was predominately found between the cerebellum and the occipitaltemporal cortex. In the group with dyslexia, there was background FC between the cerebellum and several cortical regions. When comparing the two groups, they differed in background FC in connections between the seed region right crus I and three left-hemisphere perisylvian target regions. However, there was no task-specific FC for word processing in either group and no between-group differences. Together the results do not support the theory that the cerebellum is affected functionally during reading in children with dyslexia.

Keywords: cerebellum; child; dyslexia; functional magnetic resonance imaging; reading; reading disability; word processing.

Figure 1

Cerebellar and cortical regions of interest used for the activation and connectivity analyses. (a) Cerebellar regions chosen based on the literature and defined with the SUIT atlas (Diedrichsen et al., 2009). These were used in the activation analysis (cerebellar sub‐regions) and again in the connectivity analyses (cerebellar seed regions). (b) Cortical target regions for the functional connectivity analyses are shown as spheres but all were anatomical regions derived from the FSL Harvard‐Oxford Atlas (Desikan et al., 2006)

Figure 2

Cerebellar functional activation maps. (a) Real Word > Fixation and (b) Real Word > False Font contrasts. Significant activation in vermis VI (not shown), left crus I, and right lobule VI in Control group, height threshold p < .001, p < .05 FWE‐corrected. No activation for Real Word > False Font for either group and no between‐group differences for either contrast (A or B)

Figure 3

Functional activation maps constrained to eight cerebellar sub‐regions. (a) Location of the cerebellar sub‐regions: bilateral lobule VI, crus I, crus II, and lobule VIIb. (b) Real Word > Fixation and (c) Real Word > False Font contrasts. Significant activation in left lobule VI, left crus I, and right lobule VI for Real Word > Fixation in controls. Height threshold p < .001, p‐FWE < .05 and Bonferroni‐corrected so that significance was p < .00625. No significant activation for Real Word > False Font in Controls, Dyslexics, nor between‐group differences. Corresponding coordinates in Table 4

Figure 4

Background functional connectivity maps for left and right cerebellar seed regions: lobule VI, crus I, crus II, and lobule VIIb. Cortical target regions, depicted as spheres, are from the Harvard‐Oxford Atlas. Cerebellar seed regions were selected from the SUIT Atlas. In Controls, FC was largely limited to within the cerebellum, and between the cerebellum and left occipital‐temporal cortex. Dyslexics had FC from left and right cerebellar seed regions to several left hemisphere cortical regions, including posterior superior temporal gyrus and superior parietal lobule. Controls > dyslexics is shown in purple, while dyslexics > controls are shown in green. All results corrected for multiple comparisons, p‐FDR < .05, and two‐sided statistic. FC, functional connectivity

Figure 5

Background functional connectivity between‐group differences for right crus I examined in more detail. Taken from Figure 4, (a) Between‐group differences, where purple indicates positive FC Controls > Dyslexics and green indicates positive FC dyslexics > controls. Violin plots of z‐scores (y‐axis) for right crus I with (b) left crus I and (c) the three left cortical regions, IFG tri, AG, and pSTG. Black bars represent the mean of z‐score values and circles are z‐scores of individual participants. FC, functional connectivity; IFG, inferior frontal gyrus; pSTG, posterior superior temporal gyrus