Neural Correlates of Aging-Related Differences in Pro-active Control in a Dual Task

Abstract

Background: Multi-tasking is usually impaired in older people. In multi-tasking, a fixed order of sub-tasks can improve performance by promoting a time-structured preparation of sub-tasks. How proactive control prioritizes the pre-activation or inhibition of complex tasks in older people has received no sufficient clarification so far. Objective: To explore the effects of aging on neural proactive control mechanisms in a dual task. Methodology: To address this question, the psychological refractory period (PRP) paradigm was used. Two 2-alternative-forced-choice reaction tasks with a predefined order (T1 and T2) signaled by a cue had to be executed simultaneously or consecutively by young (mean age 25.1 years, n = 36) and old subjects (mean age 70.4 years, n = 118). Performance indices of dual-task preparation were used to assess the focused preparation of T1 and T2. To compare preparatory mechanisms at the neurophysiologic level, multi-channel electroencephalogram (EEG) was recorded and negative slow cortical potentials (SCPs) were analyzed as objective markers of the amount and localization of cortical pre-activation before sub-task presentation. Results: Dual-task performance was significantly slower in old adults. T1 performance was facilitated in both age groups, but T2 processing in old adults was not optimized by the temporal structure as efficiently as in young adults. Also, only young adults manifested a stable pattern of focused of negative slow-wave activity increase at medial frontal and right-hemisphere posterior regions, which was associated with a coordinated preparatory T1 pre-activation and T2 deferment, while old adults manifested a broad topographic distribution of negative SCPs associated with a pre-activation of sensory and motor processes. Conclusions: These observations demonstrate that the proactive preparation for dual tasking is altered with aging. It is suggested that in young adults, attention-based pre-activation of working memory and inhibitory networks in the right hemisphere synchronizes the simultaneous preparation of the two sub-tasks, whereas in old adults, sensory and motor networks appear to be non-specifically pre-activated for subsequent deferred mode of processing.

Keywords: EEG; ERP; aging; dual task; proactive control; psychological refractory period (PRP) paradigm; slow cortical potentials.

FIGURE 1

Psychological refractory period (PRP) task: (A) Procedure of a single experimental trial. In the example, a blue frame and the letter B are presented as target stimuli. CTI, Cue-target interval; SOA, stimulus onset asynchrony; T1/R1, first target and reaction time in response to the first target; T2/R2, second target and reaction time in response to the second target; RCI, response-cue interval. (B) Trials with SOA = 0 ms and SOA = 750 ms. Response times are provisory.

FIGURE 2

(A) Reaction time (RT) for two groups of participants, young and old. (B) Normalized distribution of Capacity parameter for both groups of young and old adults. Capacity: speed of processing in the dual-task blocks as reflected by the mean RT to S1 and S2 in two SOA conditions (SOA750 and SOA0), in two blocks (B1-LC and B2-CL). (C) Normalized distribution of dual-task preparation parameters: preparation for T1 (PT1), preparation for T2 (PT2), and input interference for T1 (IIT1), presented for young and old subjects.

FIGURE 3

(A) Grand average of slow cortical potentials (SCPs) for two groups, young and old, for two blocks, block 1 (B1-LC), and block 2 (B2-CL) at electrodes Cz and POz (band pass 0.03-30 Hz). Statistical significance of difference in SCPs between groups of young and old subjects is presented in the map on the right side. p, statistical significance; red (negative SCPs young > negative SCPs old), blue (negative SCPs old > negative SCPs young). (B) Grand average maps of the distribution of SCPs for the same two blocks and for the B1–B2 difference in young and old adults.

FIGURE 4

Schematic illustration of electrode positions at which SCPs were extracted by regression models as positive (in red) or negative (in blue) predictors (p < 0.05) of T1 preparation (PT1), T2 preparation (PT2), and input interference (IIT1) for the groups of young and old adults. Statistical details are presented in Supplementary Table 3.