Music training and language learning improve verbal memory performance but do not change white matter characteristics...

Previous research has suggested associations between memory performance, white matter characteristics, and training in music performance. Associations of memory performance and white matter characteristics have also been found with language learning. Given the proposed links between music training and language learning, we investigate here, whether a year of different types of training (music, language, both, or other), related to white matter characteristics in the corpus callosum and the fornix and verbal and visuospatial memory performance changes.

We obtained verbal and visuospatial memory performance scores (California Verbal Learning Test Second Edition; Wechsler Memory Scale Fourth Edition, Designs 1 and 2) and diffusion tensor imaging data from 65 young adult participants before and after a year during which they enrolled in music performance classes, language learning classes, both types of classes, or other types of classes.

Analyses revealed a significant linear contrast of class type showing improvements in verbal memory scores for participants who had taken either music performance training or language learning classes, and the biggest improvement for participants that had taken both types of classes. No significant effects were observed for visuospatial memory. Regression analyses further indicated that fractional anisotropy in the splenium at time point 1 significantly predicted verbal memory change but that the class type factor did not add explanatory power.

Our findings indicate that both music training and language learning can lead to verbal memory improvements and that both types of training can compound, for example, in the form of singing or opera training to lead to greater improvements. Thus, training in music performance and language learning may have additive effects on verbal memory improvements. While callosal white matter characteristics related to verbal memory changes in our sample, the neural mechanism of the shown training effects is presumably another.