The neural correlates of developmental dyslexia: a new meta-analysis of 48 neuroimaging studies

Over the last two decades, developmental dyslexia has been the focus of much research using functional neuroimaging. A wide range of paradigms, tackling different neurocognitive domains, has been used to assess its neural correlates. The main trust of this work is that typical developmental dyslexics have dysfunction of the phonological and orthography to phonology conversion systems, normally housed in the left occipitotemporal cortex. It remains to be seen whether, besides this well replicated finding, there is a systematic co-occurrence of dysfunctional patterns of different functional systems perhaps converging on the same brain regions associated with the reading deficit. Such evidence would be relevant for theories of dyslexia like, for example, the magnocellular one. To address this problem, we performed a meta-analysis based on an optimized hierarchical clustering algorithm (Cattinelli et al., 2012) which automatically grouped 1982 activation peaks, extracted from 48 neuroimaging studies (fMRI and PET), into clusters in which the activation peaks had minimized spatial variance. The data was based on the literature published up to December 2011 including experiments on reading but also on phonological awareness, motor control, visual motion perception and so forth. The clustering analysis identified 82 clusters. The functional role of the clusters was assessed on the basis of statistical criteria. In particular, the binomial test was used to identify which clusters showed a specific activation effect for dyslexics rather than for controls. The left inferior parietal lobule, the left middle temporal gyrus, the left fusiform gyrus and the left cerebellum showed a specific association with the normal control groups, not being active in the dyslexics; on the other hand early subdivisions of visual cortices, bilaterally, the left insula and the right opercular portion of the inferior frontal gyrus showed a specific association with the dyslexic subjects. A qualitative analysis of each cluster was also performed to evaluate the distribution of the activations peaks in relation to the experimental task (e.g. reading, phonological awareness, motor learning, visual motion discrimination, etc.). We found that the aforementioned difference in left fusiform gyrus was selectively due to a lack of commitment to reading in adult dyslexics while the same region showed occasional activation for more basic visual tasks. On the other hand, the specific activation of early visual cortices in dyslexics was associated with a variety of visual tasks with reading tasks playing a major role. We conclude that the available literature demonstrates a specific lack of activation of the left occipitotemporal cortex in dyslexics that is specific for reading and for visuo-phonological tasks. The larger early occipital activations in dyslexics may represent a compensatory effort for the visual analysis of printed words in the absence of a higher-level visual-word form neural expertise. References Cattinelli I., Borghese A.N., Gallucci M., Paulesu E. (In Press) Reading the reading brain: a new meta-analysis of functional imaging data of reading. J. Neurolinguistic