EVIDENCE FOR A NEUROANATOMICAL DIFFERENCE WITHIN THE OLIVO-CEREBELLAR PATHWAY OF ADULTS WITH DYSLEXIA. Andrew J Finch, Roderick I Nicolson and Angela J Fawcett (Department of Psychology, University of Sheffield, UK) ABSTRACT Recent behavioural evidence has indicated that cerebellar impairment may be strongly associated with dyslexia. Previous neuroanatomical research has shown the presence of anomalies within the cerebral cortex of brains of dyslexic people. This paper reports equivalent analyses on the cerebella of the same brain specimens. Cross sectional areas and cell packing densities of Purkinje cells in the cerebellar cortex, and cells in the inferior olivary and dentate nuclei of four dyslexic and four control brains were measured using the dissector method. A significant difference in mean cell area in medial posterior cerebellar cortex was identified, with the dyslexic cells having larger mean area. Furthermore, analysis of cell size distributions not only confirmed the significant differences in the posterior lobe, with an increased proportion of large neurons and fewer small neurons for the dyslexics, but also revealed significant differences in the anterior lobe, again with a pattern of more large and fewer small cells. Similar distributional differences were seen in the inferior olive. No differences were found in the flocculonodular lobe or the dentate nucleus. While caution is necessary in generalising from the results given the small number of specimens, together with the age difference, the neuroanatomical data established here provides further converging evidence of cerebellar abnormality in dyslexia. Key words: cerebellum, inferior olive, dyslexia, learning disability, cell size INTRODUCTION Developmental dyslexia is traditionally defined as failure “to attain the language skills of reading, writing and spelling commensurate with [the child’s] intellectual abilities” (World Federation of Neurology, 1968). Theoretical research has revealed that dyslexia is also characterised by deficits in phonological processing (e.g., Bradley and Bryant, 1983; Stanovich, 1988), motor skill (e.g. Rudel, 1985; Wolff et al., 1990), ‘automatic’ balance and information processing speed (e.g., Nicolson and Fawcett, 1990; Wolf and Bowers, 1999). The cerebellum has traditionally been considered as a motor area (Holmes, 1917; Eccles et al., 1967) and is also involved in the automatisation of motor skill and in adaptive control (Ito, 1990). Despite early work (Frank and Levinson, 1973), for many years the cerebellum was discounted as a causal factor in dyslexia owing to its supposed lack of involvement in language. However, neuroanatomical and neurophysiological work has indicated that the human cerebellum has links to premotor and prefrontal areas, including Broca’s language area (Leiner et al., 1993). Furthermore, cerebellar involvement in language-related activities has also been observed in functional imaging studies (e.g., Paulesu et Cortex, (2002) 38, 529-539