A voxel based morphometry investigation into brain abnormalities in different subsets of developmental dyslexia

Several previous studies have reported that brain structure in developmental dyslexics differs from that of non-impaired readers. The results, however, have been inconsistent across studies and are not highly significant. The hypothesis driving the present study is that the inconsistent and weak effects are as a consequence of individual variability in how dyslexia manifests. On the basis of a previous unpublished study at the Wellcome Department of Imaging Neuroscience, it has been proposed that adolescents with reading impairments can be subdivided into those who have worse reading than spelling (henceforth referred to as dyslexics) and those that have worse spelling than reading (henceforth referred to as dysgraphics). The aim of the present study was to investigate whether this categorisation was also appropriate for adult university students with a history of reading impairments. The project involved detailed behavioural testing and structural MRI acquisition on 29 adults, 16 of which had previously been diagnosed with reading difficulties. On the basis of their behavioural results, the literacy impaired group was divided into two groups: dyslexics and dysgraphics. As predicted by their diagnosis, both groups had worse spelling and slower reading than the non-impaired readers. In addition, both groups had less grey matter than controls in two left hemisphere regions that have previously been associated with the "phonological loop". Critically, however, the structural MRI data also confirmed a double dissociation between grey matter reductions in the dyslexics and dysgraphics. The dyslexics had less grey matter in the right hemisphere phonological areas and the dysgraphics had less grey matter in the cerebellum. These findings support the hypothesis that abnormal brain structure depends on the type of reading impairment