[Computerized oculomotor training in dyslexia: A randomized, crossover clinical trial in pediatric population]. / Rééducation oculomotrice informatisée dans la dyslexie : essai clinique randomisé en crossover en population pédiatrique.

OBJECTIVE: Several studies have reported abnormal oculomotor capacities leading to reading/writing difficulties among dyslexic children. However, no randomized clinical trial has been conducted to determine whether oculomotor training improves reading/writing skills of these children. The present study aims to evaluate the efficacy of computer-based oculomotor training among dyslexic children. METHOD: Crossover randomized trial with enrollment from January 12, 2015 to July 24, 2015, and follow-up to February 4, 2016. Eleven children (aged 7 to 12 years old) with dyslexia were included in a French psychiatric unit. The computer-based oculomotor training consisted of exercises focused on control of saccadic movements (reflexes and voluntary saccades), vergences and visual attention and memory. At baseline, 3 and 6 months, participants were assessed on reading and writing skills as well as phonological skills, visuo-attentional skills and verbal memory using the French batterie analytique du langage écrit (BALE). Saccadic and antisaccadic ocular movements (latencies and gains) were recorded using a specific device. Several Anova models were performed to test whether oculomotor training improves reading, writing and phonological, verbal memory and visuo-attentional skills. Our analyses were considered exploratory (alpha at 5%). RESULTS: No effect of oculomotor training was found on reading skills. However, oculomotor training was associated with a short-term effect (after 3 months of training) on several tests measuring phonological skills (syllabic suppression; P-value=0.022), visuo-attentional skills (search of anarchic verbal cues; P-value=0.035) and verbal memory (digit span backward; P-value=0.022) and with a long-term effect (3 months after the end of the 3 months of training) on a measure of writing skills (regular words; P-value=0.019). Finally, training was associated with an increase of saccadic latencies indicating an increase of visuo-attentional skills (P-value=0.026). CONCLUSIONS: Our results suggested that computer-based oculomotor training might be effective on writing skills and several cognitive skills among dyslexic children, but future clinical trials are needed to confirm our results.