hide
Free keywords:
-
Abstract:
Children acquire their native language at a rapid pace during a time when the brain is also undergoing major changes. We set out to explore the relationship between learning during this period of language acquisition and structural brain changes. The current study investigated white matter changes after eight training sessions over three weeks of language training in 4-year-old children, using diffusion weighted MRI images collected before and after training. Children were divided into three groups, word learning, active control, and passive control. During the training period, children in the word learning group learned 60 novel fictional characters and their labels. We used a word-picture matching task with answers recorded by button press of the participant. Employing the same set- up, the active control group performed a sentence-picture matching task without novel items. The children excelled at the word learning task, the group reaching above chance performance in the second training session. The active control group did not show improvement on the trained task. To examine white matter changes, we conducted a whole brain tract-based spatial statistics (TBSS) analysis to investigate the differences between groups in fractional anisotropy (FA) change from pre- to post-scan. Additionally, we correlated performance over all training sessions with FA change in the word learning group. In the group comparison between word learning and passive control groups, the word learning group showed greater FA changes in bilateral pre- and postcentral white matter, stretching posterior to parietal white matter in the left hemisphere. To ensure that this effect was not driven by button pressing in the word learning group, we included the active control group in the analysis. When comparing all three groups, only the left hemisphere FA increase remained. Interestingly, the correlational analysis in the word learning group, correlating FA increase from scan 1 to scan 2 with accuracy over all training sessions, reveals a negative correlation between FA increase and overall accuracy in left pre- and postcentral white matter, which overlaps with the group analysis result. The correlation shows that children for whom learning was more effortful showed a larger FA increase in this region. That is, children who needed more exposure to learn the association between character and label showed larger FA increases in this region. While children at the age of four have not yet mastered all aspects of language, for example the development of understanding complex syntax stretches through the early school years, they are adept word learners. This is reflected in the behavioral data, where ceiling performance was reached already by the sixth training session. Previous studies investigating structural brain changes following word learning in adults have found changes in the white matter following training. Our results provide evidence that it is possible to measure plasticity of the white matter in children using diffusion weighted MRI images following a short word learning training program. Furthermore, the changes in white matter reflect individual differences in the ability to learn new word-object pairings.
