Age-related task-switch costs indexed by slow negative potentials

Abstract

1 Universitat de les Illes Balears, Mallorca, Spain Age-related changes in cognitive control correspond with distinct event-related potentials (ERPs), with elderly adults often showing enhanced target P3 amplitudes over frontal regions, but reduced P2 and P3 activity to informative contextual cues. Here we explored the influence of Age and Cognitive control on the neural dynamics of task-set switching and updating. Seventy-two healthy participants were grouped according to their Age and Cognitive control level, estimated from their mean Z scores in several tests of executive function (i.e., Trail Making, Brixton, Stroop, etc). A task-cueing version of the Wisconsin card sorting test was employed to measure switch-specific (local) and switch-unspecific (restart, mixing, global) behavioral costs, as well as several cue- and target-locked ERPs, with a special interest in the P300 component. All measures were submitted to a 2 x 2 x 3 mixed ANOVA design with Age (Middle-aged vs Elderly: 55.6 ±2.8 y.o. vs 67.3 ±5.1 y.o.), Cognitive Control (High vs Low control) and Trial (Switch vs Stay1 vs Stay2). Elderly adults showed larger local and mixing costs than Middle-aged adults, whereas the Lower control scores the larger global and restart costs. Trial influenced cue-locked P3 activity in all groups, with Switch trials eliciting larger P3 amplitudes compared to stay trials. On the contrary, target P3 decreased abruptly in switch trials at all electrode sites. Neither cue-locked P2 and P3, nor target P3 amplitudes showed main effects for Age, Cognitive control, nor interacted with Trial. Interestingly, High control –but not Elderly– adults showed enhanced cue-locked frontal-central slow negativities, and these were more pronounced during switch compared to stay trials. In contrast, target-locked negativities were larger in Middle-aged as compared with Elderly adults. We conclude that cue- and target-locked P3 activity reflect two distinct mechanisms related to task preparation and implementation, respectively. However, task-switch costs in Low control and Elderly adults were best indexed by reduced slow frontal negativities to cues and targets, respectively.