• Citer

The impact of very premature birth and the role of statistical learning in reading skills

par Fourdin, Lauréline ;Christiaens, Florence ;Grossman, Dominique ;Tackoen, Marie ;Vuckovic, Aline ;Le Brun, Annick ;De Tiege, Xavier ;Bertels, Julie
Référence Budapest CEU Conference on Cognitive Development (15-17/01/2026: Budapest)
Publication Non publié, 2026-01-06

Poster de conférence

• Détails
• Contenu

Résumé :

Very preterm birth (VPB) abruptly changes brain developmental conditions, raising the risk of complications and potential brain abnormalities1. Although many preterm children achieve developmental outcomes comparable to term-born peers, VPB is frequently associated with long-term cognitive and academic difficulties2, particularly in reading3. Statistical learning (SL) is believed to support reading acquisition by enabling the extraction of probabilistic regularities inherent in language input.4,5 Moreover, the brain regions most affected by VPB6,7 are those underlying SL8. This raises the question of whether SL is altered by VPB, and whether it may influence reading outcomes. We have tested 34 children aged 8–12 years so far: 25 born at term (>36 wGA) and 8 born very preterm (28–32 wGA). Participants were exposed to nine cartoon aliens9 presented in fixed triplets. Learning of the triplets was then assessed through a rapid serial visual presentation (RSVP) detection task, and a three-alternative forced-choice (3AFC) completion. Reading ability was assessed using the standardized French task Alouette-R.Preliminary analyses suggest that both groups were able to learn regularities: reaction times tended to decrease for aliens in the third compared to the first triplet position in the RSVP task, and average accuracy in the 3AFC task was above chance. No robust group differences were observed for SL. Likewise, no difference in reading ability emerged between groups. Regression analyses indicated an association between SL performance and reading efficiency, independently of birth status.These findings must be interpreted with caution given the small sample size, which limits statistical power. Nonetheless, current trends point toward SL as a possible mechanism contributing to individual variability in reading outcomes4, supporting previous findings. They also suggest that VPB has a negligible impact on reading and SL performance. These findings will require confirmation in a larger cohort, which will be available in the coming months.