Increased muscle coactivation is linked with fast feedback control when reaching in unpredictable visual environments

Abstract

Humans encounter unpredictable disturbances in daily activities and sports. When encountering unpredictable physical disturbances, healthy participants increase the peak velocity of their reaching movements, muscle coactivation, and responses to sensory feedback. Emerging evidence suggests that muscle coactivation may facilitate responses to sensory feedback and may not solely increase stiffness to resist displacements. We tested this idea by examining how healthy participants alter the control of reaching movements and responses to sensory feedback when encountering variable visuomotor rotations. The rotations changed amplitude and direction between movements, creating unpredictable errors that required fast online corrections. Participants increased the peak velocity of their movements, muscle coactivation, and responses to visual and proprioceptive feedback with the variability of the visuomotor rotations. The findings highlight an increase in neural responsiveness to sensory feedback and suggest that muscle coactivation may prime the nervous system for fast responses to sensory feedback that accommodate properties of unpredictable visual environments.

Keywords: cognitive neuroscience; neuroscience; sensory neuroscience.

Graphical abstract

Figure 1

General task description (A) Participants performed goal-directed reaching movements with their dominant arm supported in a robotic exoskeleton. The activity of the monoarticular elbow muscles and biarticular muscles was recorded with surface electromyography. (B) Participants performed reaching movements in the absence or presence of variable visuomotor rotations (VMRs). The VMRs rotated the feedback cursor clockwise or counterclockwise relative to the hand in random order. Lateral cursor jumps (±4 cm) were used to probe the responsiveness to visual feedback in random trials throughout the experiment. The cursor jumped laterally when the participant left the start position (i.e., movement onset). The rotations were not applied in cursor jump trials. (C) Experiment 1 included baseline, exposure, and washout phases. Participants encountered unpredictable visuomotor rotations with different levels of variability in separate subphases of the exposure phase (±20° and ±30°). In addition to cursor jumps, unperturbed trials (absence of visuomotor rotations) were also randomly interleaved during the exposure phase. (D) Exemplary cursor and hand trajectories performed by a representative participant in the absence (baseline and washout) or presence of unpredictable VMRs in Experiment 1. Gray lines indicate cursor jump trials. The unpredictable VMRs elicit lateral cursor displacements and require corrections of the hand trajectory. Thick lines represent the average cursor and hand trajectories. Thin lines are individual trials. Counterclockwise rotations (dashed lines) required clockwise corrections of the hand trajectory and vice versa. Similarly, rightward cursor jumps required leftward corrections of the hand trajectory and vice versa.

Figure 2

Properties of voluntary behavior and associated muscle activity of unperturbed trials in Experiment 1 (A) Group mean ± SE forward velocities in each phase of the experiment. The data are aligned with movement onset (t = 0 ms). Colored horizontal lines indicate the 95% confidence interval of the movement time across phases of the experiment. The side panel displays each individual’s mean peak forward velocity (gray lines) and the corresponding group means (colored diamonds and black line) across experimental phases. (B) Group mean ± SE activity of the elbow extensor muscles (solid lines) plotted in the same format as (A). The data were smoothed with a 10 sample (10 ms), zero-delay moving average for display purposes. The shaded, gray area (−100 to 100 ms) indicates the muscle activity associated with the planning and initiation of voluntary behavior. The side panel displays the average muscle activity surrounding movement onset. (C) Group mean ± SE activity of the elbow flexor muscles (dashed lines) plotted in the same format as (B). Arrows indicate statistically significant contrasts. BL, baseline; WO, washout; nu, normalized unit.

Figure 3

Lateral velocities and muscle responses during visual probes in Experiment 1 (A) Group mean ± SE change in lateral velocities in each phase of the experiment during rightward cursor jumps. The data are aligned with the onset of the visual probes (t = 0 ms). The dashed, vertical lines separate the velocity_early (180–230 ms) and velocity_late (230–280 ms) time windows. The side panel displays the mean lateral velocity in the velocity_early and velocity_late time windows for each participant (gray lines) across experimental phases. The colored diamonds and black line represent the group means. (B) Group mean ± SE change in lateral velocities in each phase of the experiment during leftward cursor jumps plotted in the same format as (A). (C) Group mean ± SE change in activity of the elbow flexor (solid lines, upper panel) and extensor muscles (dashed lines, lower panel) during rightward cursor jumps. The data are aligned with the onset of the visual probe (t = 0 ms). The dashed, vertical lines separate the SLR_visual (90–120 ms) and LLR_visual (120–180 ms). The data were smoothed with a 10 sample (10 ms), zero-delay moving average for display purposes. The side panels display the mean muscle activity in the SLR_visual and LLR_visual time windows for the elbow flexors (solid gray lines, upper panel) and extensors (dashed gray lines, lower panel) for each participant across experimental phases. The colored diamonds and black line represent the group means. (D) Group mean ± SE change in activity of the elbow extensor (dashed lines, upper panel) and flexor muscles (solid lines, lower panel) during leftward cursor jumps plotted in the same format as (C). Arrows indicate statistically significant contrasts. BL, baseline; WO, washout; nu, normalized unit.

Figure 4

Properties of voluntary behavior and associated muscle activity of unperturbed trials in Experiment 2 (A) Exemplary cursor and hand trajectories performed by a representative participant in the absence (baseline and washout) or presence of unpredictable VMRs. Increasing the amplitude of the VMRs imposes larger cursor deviations and requires larger corrective responses to guide the feedback cursor into the goal target within the time constraints of the task. (B) Group mean ± SE forward velocities in each phase of the experiment. The data are aligned with movement onset (t = 0 ms). Colored horizontal lines indicate the 95% confidence interval of the movement time across phases of the experiment. The side panel displays each individual’s mean peak forward velocity (gray lines) and the corresponding group means (colored diamonds and black line) across experimental phases. (C) Group mean ± SE activity of the elbow extensor muscles (solid lines) plotted in the same format as (B). The data were smoothed with a 10 sample (10 ms), zero-delay moving average for display purposes. The shaded, gray area (−100 to 100 ms) indicates the muscle activity associated with the planning and initiation of voluntary behavior. The side panel displays the average muscle activity surrounding movement onset. (D) Group mean ± SE activity of the elbow flexor muscles (dashed lines) plotted in the same format as (C). Arrows indicate statistically significant contrasts. BL, baseline; WO, washout; nu, normalized unit. See also Figure S1.

Figure 5

Lateral velocities and muscle responses during visual probes in Experiment 2 (A) Group mean ± SE change in lateral velocities in each phase of the experiment during rightward cursor jumps. The data are aligned with the onset of the visual probes (t = 0 ms). The dashed, vertical lines separate the velocity_early (180–230 ms) and velocity_late (230–280 ms) time windows. The side panel displays the mean lateral velocity in the velocity_early and velocity_late for each participant (gray lines) across experimental phases. The colored diamonds and black line represent the group means. (B) Group mean ± SE change in lateral velocities in each phase of the experiment during leftward cursor jumps plotted in the same layout as (A). (C) Group mean ± SE change in activity of the elbow flexor (solid lines, upper panel) and extensor muscles (dashed lines, lower panel) during rightward cursor jumps. The data are aligned with the onset of the visual probe (t = 0 ms). The dashed, vertical lines separate the SLR_visual (90–120 ms) and LLR_visual (120–180 ms). The data were smoothed with a 10 sample (10 ms), zero-delay moving average for display purposes. The side panels display the mean muscle activity in the SLR_visual and LLR_visual time windows for the elbow flexors (solid gray lines, upper panel) and extensors (dashed gray lines, lower panel) for each participant across experimental phases. The colored diamonds and black line represent the group means. (D) Group mean ± SE change in activity of the elbow extensor (dashed lines, upper panel) and flexor muscles (solid lines, lower panel) during leftward cursor jumps plotted in the same format as (C). Arrows indicate statistically significant contrasts. BL, baseline; WO, washout; nu, normalized unit.

Figure 6

Lateral displacements and muscle responses during mechanical probes in Experiment 3 (A) Group mean ± SE change in lateral hand displacements in each phase of the experiment during extension probe trials. The data are aligned with the onset of the mechanical probes (t = 0 ms). The side panel displays the mean peak lateral hand displacement of each participant (gray lines) across experimental phases. The colored diamonds and black line represent the group means. (B) Group mean ± SE change in lateral hand displacement in each phase of the experiment during flexion probe trials. The data are plotted in the same layout as (A). (C) Group mean ± SE change in activity of the elbow flexor (solid lines, upper panel) and extensor muscles (dashed lines, lower panel) during extension probes. The data are aligned with the onset of the mechanical probes (t = 0 ms). The dashed, vertical lines separate the SLR_mechanical (25–50 ms) and LLR_mechanical (50–100 ms) time windows. The data were smoothed with a 10 sample (10 ms), zero-delay moving average for display purposes. The side panels display the mean muscle activity in the SLR_mechanical and LLR_mechanical time windows for the elbow flexors (solid gray lines, upper panel) and extensors (dashed gray lines, lower panel) for each participant across experimental phases. The colored diamonds and black lines represent the group means. (D) Group mean ± SE change in activity of the elbow extensor (dashed lines, upper panel) and flexor muscles (solid lines, lower panel) during flexion probes plotted in the same format as (C). Arrows indicate statistically significant contrasts. BL, baseline; WO, washout; nu, normalized unit. See also Figures S1–S3.

Figure 7

Relationship between muscle activity and responses to sensory feedback based on linear mixed effects models (A) Relationship between the background muscle activity (BKG) of the elbow flexor and extensor muscles with the slope of the lateral velocity during rightward (RW) cursor jumps. Colored background indicates the planar fit of the model with brighter colors indicating a steeper slope of the lateral velocity. The side panel displays the 95% confidence intervals of the fixed effects of the model and each participant’s individual intercept (I) and slopes (β₁: flexors; β₂: extensors). (B) Relationship between BKG activity of the elbow flexor and extensor muscles and the slope of the lateral velocity during leftward (LW) cursor jumps. The data are plotted in the same layout as (A). (C and D) Relationship between BKG of the elbow flexor and extensor muscles with the peak hand displacement during mechanical probes that caused extension (Ext) or flexion (Flx) of the elbow. The data are plotted in the same layout as (A) with brighter colors of the background indicating reduced peak hand displacements. (E and F) Relationship between the LLR_visual of the elbow flexor and extensor muscles with the slope of the lateral velocity, plotted in the same layout as (A). (G and H) Relationship between the LLR_mechanical of the elbow flexor and extensor muscles with the peak hand displacement, plotted in the same layout as (C). Agonist and antagonist muscles were defined relative to their action during the visual or mechanical probes. See also Table S1.