Challenge to promote change: the neural correlates of the contextual interference effect

When learning motor tasks across the lifespan, it is crucial to know which types of training lead to the best possible outcome. Learning and retention can be facilitated by presenting an optimal level of desirable difficulty during practice by means of manipulating practice structure, an effect called contextual interference (CI). CI is generally induced by using a randomized rather than blocked schedule for practicing different task variants, thereby creating a richer practice context. Although random practice is more challenging during the acquisition phase, thereby leading to inferior performance levels relative to blocked practice, it will eventually lead to better long-term retention. The CI effect appears to be a robust phenomenon in simple (unimanual) task learning, however, as mastering bimanual coordination skills is particularly important to remain functionally independent across the lifespan, the generalizability of the CI effect in learning more complex bimanual tasks is of great interest. Therefore, in the first part of the present thesis, we aimed to establish the behavioral effects of CI in complex bimanual task conditions and different age groups. The behavioral part served as a foundation for the second part of the thesis where we sought to identify the underlying neural mechanisms of the CI effect. Throughout the thesis, the CI effect was examined using a bimanual task with variants of different levels of complexity. Our behavioral work (part 1 of the doctoral thesis) revealed that when task demands were relatively low, typical CI effects were confirmed in both age groups. More specifically, random practice resulted in inferior performance during the acquisition phase, but better retention performance as compared to blocked practice. Nevertheless, absolute retention benefits after randomized as compared to blocked practice were not evident when task demands increased. Yet, random practice led to the formation of a more stable memory trace, i.e. stable performance from the end of acquisition to retention one week later, irrespective of task complexity and age group. Therefore, we believe the lack of absolute retention benefits could not be explained by a cognitive overload resulting from increased task demands in combination with a more challenging practice schedule. Rather, a more demanding task variant tended to be a desirable difficulty during blocked practice, as performance depletions after blocked practice were less pronounced in the more demanding task variant. Interestingly, although older adults had more difficulty with performing the task in general, resulting in lower levels of performance as compared with young adults, the benefits of CI were independent of age. In the second part of this doctoral thesis, fundamental knowledge about the neural basis of training-induced neuroplasticity underlying the behavioral effects of CI was gained. Functional imaging revealed that across age groups, blocked practice evoked higher motor-related brain activity during the acquisition phase as compared to random practice. On the contrary, we observed a higher dependency on visual processing regions during both the acquisition and retention phase in randomized relative to blocked practice in both age groups. Also from a neurochemical perspective, γ-aminobutyric acid (GABA) level in the occipital cortex was differentially modulated through different types of practice in both young and older adults. More specifically, occipital GABA level was decreased as a function of random practice, which is indicative of cortical LTP-like plasticity. Since the bimanual visuo-motor tracking task, used in the present thesis, places a strong emphasis on the visual modality for performing the task, our findings give rise to a task-dependent perspective on the beneficial CI effects. In the context of aging, we demonstrated that different types of practice structure modulated activity in default mode network (DMN) and task-related brain regions. More specifically, by providing a richer practice context, i.e. random practice, activity in the DMN regions decreased whereas task-related (i.e. motor) brain regions became more active from early to late phases of learning, i.e. increased segregation of brain activity. This finding may reflect an indirect signature of increased task engagement during random practice, ultimately leading to positive behavioral effects. The fact that the DMN appears to show increased vulnerability with age, but that it can nevertheless be modulated as a result of practice with behavioural consequences, is highly exciting for practical implications. In conclusion, the current doctoral thesis provided evidence that across age groups, high levels of CI (i.e. random practice) led to the formation of a stronger memory trace as compared to blocked practice during bimanual skill learning. This finding indicates that potential learning benefits through the implementation of an optimized training protocol (i.e. random practice) are preserved at older age. At the neural level, our findings underscore the task-dependent nature of the beneficial CI effect. Furthermore, providing a more challenging practice structure triggered neuroplastic changes in older adults resulting in increased segregation of the task-related and DMN activations, ultimately leading to better retention performance.status: publishe