An Increase in Postural Load Facilitates an Anterior Shift of Processing Resources to Frontal Executive Function in a Postural-Suprapostural Task

Abstract

Increase in postural-demand resources does not necessarily degrade a concurrent motor task, according to the adaptive resource-sharing hypothesis of postural-suprapostural dual-tasking. This study investigated how brain networks are organized to optimize a suprapostural motor task when the postural load increases and shifts postural control into a less automatic process. Fourteen volunteers executed a designated force-matching task from a level surface (a relative automatic process in posture) and from a stabilometer board while maintaining balance at a target angle (a relatively controlled process in posture). Task performance of the postural and suprapostural tasks, synchronization likelihood (SL) of scalp EEG, and graph-theoretical metrics were assessed. Behavioral results showed that the accuracy and reaction time of force-matching from a stabilometer board were not affected, despite a significant increase in postural sway. However, force-matching in the stabilometer condition showed greater local and global efficiencies of the brain networks than force-matching in the level-surface condition. Force-matching from a stabilometer board was also associated with greater frontal cluster coefficients, greater mean SL of the frontal and sensorimotor areas, and smaller mean SL of the parietal-occipital cortex than force-matching from a level surface. The contrast of supra-threshold links in the upper alpha and beta bands between the two stance conditions validated load-induced facilitation of inter-regional connections between the frontal and sensorimotor areas, but that contrast also indicated connection suppression between the right frontal-temporal and the parietal-occipital areas for the stabilometer stance condition. In conclusion, an increase in stance difficulty alters the neurocognitive processes in executing a postural-suprapostural task. Suprapostural performance is not degraded by increase in postural load, due to (1) increased effectiveness of information transfer, (2) an anterior shift of processing resources toward frontal executive function, and (3) cortical dissociation of control hubs in the parietal-occipital cortex for neural economy.

Keywords: dual-task; event-related potential; functional connectivity; graph analysis; network-based statistics.

Figure 1

Schematic illustration of experimental setup (A) and physiological data (B). Real-time display of precision grip force, ankle displacement, and target signals for concurrent force-matching and postural tasks. By separate scale-tuning of the manual force target and postural target, the target signals of both postural and force-matching tasks could be displayed in an identical position on the monitor. Suprapostural performance was assessed with the reaction time (RT) and normalized force error (NFE) of a force-matching act. The event-related potential (ERP) of the force-matching act was recorded with scalp electroencephalography. ERP between the executive tone and onset of the force-impulse profile was denoted as preparatory ERP, composed of N1 and P2 components. TF, target force; PGF, peak grip force.

Figure 2

Workflow of cortical network construction with ERP in the preparatory period. Functional connectivity was characterized with the synchronization likelihood (SL) of the paired preparatory ERP of different recording electrodes. The adjacent matrix represents the strengths of inter-electrode synchronization likelihood following appropriate threshold selection. With the adjacent matrix, connectomes in the preparatory period in the level-surface, and stabilometer conditions were established.

Figure 3

The contrast of network parameters between the concurrent force-matching and postural tasks in the level-surface and stabilometer conditions at different threshold values. E_glob, global efficiency; E_loc, local efficiency; Sigma, small-world index; ^*, level-surface > stabilometer, p < 0.05; ^†, stabilometer > level-surface, p < 0.05; ^††, stabilometer > level-surface, p < 0.01; ^†††, stabilometer > level-surface, p < 0.001.

Figure 4

The pooled adjacent matrix of synchronization likelihood (SL) of preparatory ERP for the concurrent force-matching and postural tasks in the level-surface and stabilometer conditions (threshold value = 0.3). (A) Population means of SL adjacent matrix in the surface (left) and stabilometer (right) conditions. (B) The adjacent matrix of t-values for contrasting SL between the level-surface and stabilometer conditions (t > 1.771, stabilometer SL > level-surface SL, p < 0.05; t < −1.771, level-surface SL > stabilometer SL, p < 0.05).

Figure 5

(A) The pooled topological mapping of clustering coefficients (C_w) between concurrent force-matching and postural tasks in the level-surface and stabilometer conditions (threshold value = 0.3). (B) The stance-dependent difference in the distribution of clustering coefficients, in terms of the topography of t-values plotted on the scalp (t > 1.771, stabilometer C_w > level-surface C_w, p < 0.05; t < −1.771, level-surface C_w > stabilometer C_w, p < 0.05).

Figure 6

Spectral connectivity analysis for contrasting wiring diagrams between the concurrent force-matching and postural tasks in the level-surface and stabilometer conditions (threshold value = 0.3). (A) Synchronization likelihood in the theta band (4–8 Hz) that connects to the mid-frontal area (F_z and FC_z). (B) Synchronization likelihood in the upper alpha (10–13 Hz) and beta band (13–35 Hz) of recording electrodes that connect to the sensorimotor (C₃, C_z, C₄, CP₃, CP_z, and CP₄) and parietal-occipital (P₃, P_z, P₄, P₃, O₁, O_z, and O₂) areas (thin red line, stabilometer connectivity > level-surface connectivity, p < 0.05; bold red line, stabilometer connectivity of supra-threshold > level-surface connectivity of supra-threshold, p < 0.005; thin blue line, level-surface connectivity > stabilometer connectivity, p < 0.05; bold blue line, level-surface connectivity of supra-threshold > stabilometer connectivity of supra-threshold, p < 0.005).