Résumé
Aims: To evaluate the effectiveness of visual training aimed at reducing crowding in dyslexic children. Study Design: Single-masked crossover pilot study. Place and Duration of Study: University of Turin and the Gradenigo Hospital, Department of Ophthalmology, Turin, between January and November 2013. Methodology: 15 dyslexic children underwent a visual training devised to reduce crowding. Patients were asked to recognize trigrams of letters with different spacing displayed at different eccentricities on both sides of the fixation point. As a placebo half of the sample was administered a contrast sensitivity test. Average reading rate for words and non-words with different interletter spacing was measured before and after the visual training and the placebo. The sample was divided into two subgroup: G1, who was first administered the training, and G2, who underwent first the placebo trial. Results: After the training in G1 reading rate for words increased from 1.54 syl/sec (±0.60) to 1.74 syl/sec (± 0.64) (P= .001). Reading rate for non-words improved from 0.94 (0.68-1.55) syl/sec to 1.03 (0.85-1.63) syl/sec. No significant improvement was found after the administration of the placebo (T2) when testing words and non-words Analysis of variance showed a significant placebo x treatment effect for words (P= .001) and a barely significant effect for non-words (P= .05). In G2 no significant improvement was found after the placebo both at words and non-words (from 1.69 syl/sec [±0.83] to 2.01 syl/sec [±0.94] for words, from 1.07 syl/sec [±0.51] to 1.08 syl/sec [±0.50] for nonwords). In this group the training increased the reading rate for words and non-words (from 2.01 syl/sec to 2.12 syl/sec [±1.13]; non-words: from 1.08 syl/sec to 1.22 syl/sec [±0.59]). However, analysis of variance did not show a significant effect of the treatment (words: P= .70; non-words: P= .85). Conclusion: Factors other than visuoperceptive, in particular the phonological impairment, could help to account for the controversial results obtained in the small group of dyslexic children recruited in this study. In future investigations, performed on a larger sample, a classification aimed at ruling out patients mainly affected by phonological defects should be considered, in order to select the appropriate target suitable for such kind of approach.