Comparison of visual information processing in school-age dyslexics and normal readers via motion-onset visual evoked potentials.

Standard pattern-reversal visual evoked potentials (VEPs) and motion-onset VEPs (M-VEPs) were tested in 19 dyslexics and 19 normal readers aged 7-13 years in order to evaluate the feasibility of M-VEPs for the objective diagnostics of a visual subtype of dyslexia, in which a dysfunction of the magnocellular subsystem/dorsal stream of the visual pathway is suspected. The set of VEPs consisted of the pattern-reversal VEPs with check sizes of 20', two types of translational motion (with low and high contrast) and two types of radial motion (in the full field or the periphery). While the P100 peak parameters in pattern-reversal VEPs did not differ between the group of dyslexics and controls, the group of dyslexics displayed significantly longer N2 latencies in all types of M-VEPs. Abnormal N2 latencies were found in 35-56% of dyslexics in different types of M-VEPs, with translational motion with high contrast being the most sensitive stimulation. A receiver operating characteristic analysis showed that the latencies of M-VEPs displayed higher discrimination potential than M-VEPs amplitudes. The study confirms a "magnocellular pathway/dorsal stream deficit" in approximately half of dyslexics.