Poster #3 - Training-induced Plasticity of Frontoparietal Activation and Connectivity During Task Switching in Children

• In Event: Poster Session 11

Abstract

With age, children become better at flexibly switching between tasks and adapting their behavior to changing environments. Children’s difficulties with task switching diminish with training, but the neural correlates of these improvements are largely unknown. We examined whether task-switching training in children led to more efficient rule processing in frontoparietal regions reflected in reduced brain activation (e.g., Dux et al., 2009) and to changes in connectivity between these regions, as previously shown in adults (e.g., Jolles et al., 2013). Alternatively, children may adapt to increased task-switching demands during training via different neural mechanisms, including upregulation of activation in frontoparietal brain regions already involved in task switching prior to training or through the recruitment of additional brain regions.
Children aged 8 to 11 years underwent nine weeks of intensive single-task (SI, N=40/30 with/without MRI) or task-switching (SW, N=40/26) training on tablets at home, or were part of a passive control group (MC, N=39, all with MRI). Children in the training groups performed task switching in the MRI scanner or MRI simulator across four timepoints: before, after 3 weeks, 6 weeks and at the end of training.
Hierarchical drift diffusion models (Wiecki et al., 2013) revealed that children in the SW group showed greater training-related increases in drift rates during task-switching (difference between posterior distributions of timepoints > .99), while drift rates did not change in the SI and MC groups (differences between posterior distributions of timepoints < .95). Thus, intensive practice led to faster evidence accumulation for the correct response in children who practiced the actual switching between tasks, but not in children who practiced the tasks separately.
The left superior parietal lobe (SPL) and dorsolateral prefrontal cortex (dlPFC), brain regions previously associated with task-set maintenance and management (Richter & Yeung 2014), showed training-related reductions in activation in the SW group (session x group interaction; SPL: b = –0.87 (95% CI: –1.73 to –0.02); dlPFC: b = –1.45 (95% CI: –2.46 to –0.44)), suggesting more efficient task-set management to support improved task-switching performance. Compared to the SI and MC group, the SW group showed reduced activation on switch relative to repeat trials in the anterior insula after training (session x group x condition interaction; b = –0.89 (95% CI: –1.71 to –0.05)), suggesting less reliance on a salience-driven response mode on switch trials (Han et al., 2019). While task-based connectivity between frontoparietal regions was increased for repeat relative to single trials prior to training (p < .001, FDR-corrected), this connectivity pattern did not change with training, providing no support for the hypothesis of changes in communication between brain regions with training.
Overall, these results suggest more efficient processing of task rules with training, as evident in faster accumulation of evidence for the correct response along with reduced activation in brain regions supporting the management of multiple relevant task rules. More efficient processing of task rules might enable a shift from salience-driven reactive control to proactive control, as suggested by reduced activation in the anterior insula.

Authors

• Sina Schwarze, Max Planck Institute for Human Development
  Presenting Author

• Yana Fandakova, University of Trier