Learning from the Physical Consequences of Our Actions Improves Motor Memory

Abstract

Actions have consequences. Motor learning involves correcting actions that lead to movement errors and remembering these actions for future behavior. In most laboratory situations, movement errors have no physical consequences and simply indicate the progress of learning. Here, we asked how experiencing a physical consequence when making a movement error affects motor learning. Two groups of participants adapted to a new, prism-induced mapping between visual input and motor output while performing a precision walking task. Importantly, one group experienced an unexpected slip perturbation when making foot-placement errors during adaptation. Because of our innate drive for safety, and the fact that balance is fundamental to movement, we hypothesized that this experience would enhance motor memory. Learning generalized to different walking tasks to a greater extent in the group who experienced the adverse physical consequence. This group also showed faster relearning one week later despite exposure to a competing mapping during initial learning, evidence of greater memory consolidation. The group differences in generalization and consolidation occurred although they both experienced similar magnitude foot-placement errors and adapted at similar rates. Our results suggest the brain considers the potential physical consequences of movement error when learning and that balance-threatening consequences serve to enhance this process.

Keywords: balance; consolidation; generalization; locomotion; sensorimotor adaptation.

Figure 1.

Experimental tasks and protocol. A, A simulated view of the stepping target through the goggles coupled with zero-diopter (nonvisual-field-shifting) lenses and 20-diopter prism lenses that shift the perceived location of the target 11.4° to the right. B, An illustration of the precision walking task without and with an adverse physical consequence, a slippery surface, present next to the stepping target. Inset, Left, A diagram showing positive (+) and negative (–) ML foot-placement error, defined as the distance between the position marker on the mid-foot and the center of the target. AP, anterior-posterior dimension in laboratory space. Inset, Right, Bird’s eye view of the location of the slippery surface relative to the target. C, An illustration of the generalization tasks. Left, Interlimb transfer test. Note that the left foot is used to step to the target in this task. Inset, Identical to that shown in C. Right, Obstacle avoidance task. Inset, A diagram showing positive (+) and negative (–) ML deviation from the obstacle, defined as the distance between the center of the obstacle and the position marker on the mid-foot of both the trailing foot (i.e., step N–1: right foot) and leading foot (i.e., step N: left foot). D, An illustration of the experimental protocol across both testing sessions (top), as well as predicted foot-placement error (and distance from obstacle) profiles related to initial adaptation, generalization, and consolidation (bottom). Note that for the obstacle avoidance task, only predictions for step N are shown for simplicity. Generalization is evident from a negative (leftward) shift (i.e., opposite to the prism shift participants adapted to earlier) in error or distance. During the first testing session, all participants performed baseline, adaptation, generalization, readaptation, and washout phases. The baseline phase included the precision walking task present in the adaptation phase and the two generalization walking tasks. Depending on the phase, participants wore goggles paired with either zero-diopter or 20-diopter lenses. To assess consolidation, participants repeated baseline and adaptation phases one week later.

Figure 2.

Slip measures. A, Left, Group mean ± SE slip distance across all trials for baseline and adaptation phases for the control (blue) and consequence (orange) groups. Right, Group mean ± SE slip distance for the baseline phase (mean of the last 10 trials) and the first adaptation trial for the control (blue) and consequence (orange) groups. B, Left, Group mean ± SE peak slip velocity across all trials for baseline and adaptation phases for the control (blue) and consequence (orange) groups. Right, Group mean ± SE peak slip velocity for the baseline phase (mean of the last 10 trials) and the first adaptation trial for the control (blue) and consequence (orange) groups. Individual participant values are superimposed. * Indicates that values are significantly different from each other based on post hoc tests (p < 0.05). See Extended Data Figure 2-1 for more detailed post hoc test results. Insets, Individual data points showing significant evidence of a slip perturbation, defined as a slip distance or slip velocity of greater than the mean plus 2 SDs of the last 10 baseline trials. The dark gray shaded box represents the first adaptation trial, and the light gray shaded box represents early adaptation trials. The numbered black boxes represent how many participants slipped during each trial. Every participant in the consequence group slipped during the first two adaptation trials. No participants slipped after the sixth adaptation trial.

Figure 3.

Visuomotor adaptation during session 1. A, Group mean ± SE ML foot-placement error across all trials for baseline and adaptation phases and the first washout trial during the first testing session for the control (blue) and consequence (orange) groups. B, Group mean ± SE foot-placement error for the baseline phase (mean of the last 10 trials), first adaptation trial, late adaptation (mean of the last 10 trials), and first washout trial during the first testing session for the control (blue) and consequence (orange) groups. Individual participant values are superimposed (n = 12 per group). A positive value represents errors in the direction of the prism shift (i.e., to the right of the target) and negative values represent errors in the opposite direction to the prism shift. C, Group mean ± SE gait speed during the precision walking task. * Indicates that values are significantly different from each other based on post hoc tests (p < 0.05). See Extended Data Figure 3-1 for more detailed post hoc test results.

Figure 4.

Generalization. A, Group mean ± SE foot-placement error for the baseline (mean of the last 10 trials) and generalization phases during the precision walking task for the control (blue) and consequence (orange) groups. Foot-placement errors in the direction opposite to the prism shift (i.e., a negative value) indicate generalization. B, Group mean ± SE foot-placement error for the baseline (mean of the last 10 trials) and generalization phases during the obstacle avoidance task for both the trailing foot (i.e., step N–1: right foot) and leading foot (i.e., step N: left foot) for the control (blue) and consequence (orange) groups. A smaller value indicates generalization for step N–1 (right foot), whereas a greater negative value reflects generalization for step N (left foot). Individual participant values are superimposed (n = 12 per group). * Indicates that values are significantly different from each other based on post hoc tests (p < 0.05). See Extended Data Figure 4-1 for more detailed post hoc test results.

Figure 5.

Motor memory consolidation. A, Group mean ± SE foot-placement error for all trials in the baseline and adaptation phases across testing sessions for the control (blue) and consequence (orange) groups (n = 12 per group for both sessions). B, Group mean ± SE for the first adaptation trial error (dark gray shaded box in panel a), early adaptation error (light gray shaded box in panel A), and rate of adaptation across testing sessions for the control (blue) and consequence (orange) groups. One week separated testing sessions. Individual participant values are superimposed (n = 12 per group). * Indicates that values are significantly different from each other based on post hoc tests (p < 0.05). See also Extended Data Figure 5-1 for more detailed post hoc test results.

Figure 6.

Relationship between slip severity and measures of generalization and consolidation. A, Scatter plots of the relationship between slip velocity and measures of generalization (left column) and consolidation (right column). B, Scatter plots of the relationship between slip distance and measures of generalization (left column) and consolidation (right column). Solid lines indicate the linear fits obtained from the regression analyses. See also Extended Data Figure 6-1.

Figure 7.

Summary of the possible mechanisms for the improvement in motor learning observed for the consequence group. The slip perturbation experienced when making a foot-placement error because of the novel visuomotor mapping may have served as a surprise (for the first exposure during adaptation) and increased the threat of losing balance (in subsequent walking trials). These factors may have increased emotional arousal, which led to a strengthening of synaptic connections in relevant sensorimotor areas where memory of the learned mapping was marked for consolidation. The slip perturbation associated with making foot-placement errors (possibly because of the surprising nature of it) may have also increased the error signal itself, leading to greater generalization and consolidation.