Learning and interlimb transfer of new gait patterns are facilitated by distributed practice across days

Abstract

Background: Previous studies have shown that the extent to which learning with one limb transfers to the opposite, untrained limb (i.e., interlimb transfer) is proportional to the amount of prior learning (or skill acquisition) that has occurred in the training limb. Thus, it is likely that distributed practice-a training strategy that is known to facilitate learning-will result in greater interlimb transfer than massed practice.

Research question: To evaluate the effects of massed and distributed practice on acquisition and interlimb transfer of leg motor skills during walking.

Methods: Forty-five subjects learned a new gait pattern that required greater hip and knee flexion during the swing phase of gait. The new gait pattern was displayed as a foot trajectory in the sagittal plane and participants attempted to match their foot trajectory to this template. Subjects in the massed practice group (n = 20) learned the task on a single day, whereas subjects in the distributed practice group (n = 25) learned the task that was spaced over two consecutive days (training phase). Following completion of training, subjects in both groups practiced the task with their untrained, opposite leg to evaluate interlimb transfer (transfer phase).

Results: Results indicated that the amount of skill acquisition (i.e., reductions in tracking error) on the training leg was significantly higher (P < 0.05) in the distributed practice group when compared with the massed practice group. Similarly, the amount of interlimb transfer was also significantly higher (P < 0.05) in the distributed practice group both at the beginning and end of the transfer phase.

Significance: The findings indicate that acquisition and interlimb transfer of leg motor skills are significantly greater when the task was learned using distributed practice, which may have implications for gait rehabilitation in individuals with unilateral deficits, such as stroke.

Keywords: Consolidation; Cross education; Intermanual; Skill learning; Spaced practice; Spacing.

Figure 1:

Schematic of the experimental protocol. Two groups of participants volunteered for this study. Participants in the massed practice group performed the training (18 trials) in a single day and the participants in the distributed practice group performed the same task over two consecutive days that were separated by about 24 hours (i.e., 9 trials on Day 1 and 9 trials on Day 2). Thus, the total practice time was constant for both groups. During normal walking (NW) trials participants walked normally on a treadmill at 0.89 m/s (2 mph). The hip and knee kinematic data recorded during the normal walking trials were used to construct the target-templates. During target-matching (TM) trials participants performed a foot trajectory tracking task that necessitated greater hip and knee flexion during the swing phase of gait. The prefixes “Tr” and “Tf” refer to the training and transfer legs, respectively. The suffixes Pre and Post refer to the initial and final training blocks, respectively. Following each block, subjects were given approximately 1 min of rest.

Figure 2:

(A) Schematic of the experimental set-up of the leg motor skill learning task during treadmill walking. (B) Schematic showing creation of target-template. (C) Schematic showing the computation of tracking error (shaded region).

Figure 3:

Ensemble averages of the hip and knee flexion angles during normal walking (NW) and target-matching (TM) trials in the massed and distributed practice group. Panels A and B represent the training leg’s ensemble averages of hip and knee flexion angles, respectively. Panels C and D represent the transfer leg’s ensemble averages of hip and knee flexion angles, respectively. Note that the strategies used for performing the motor learning task were similar between groups for both the training and the transfer legs.

Figure 4:

Mean tracking error in the massed and distributed practice groups for each target-matching (TM) block with the training leg (Tr) and the transfer leg (Tf). The shaded regions indicate the training and testing blocks performed on Day 2 for the distributed practice group. Error bars represent the standard error of the mean. The blocks within the rectangle indicate the trials that were used in the statistical analysis. Tracking error was significantly lower in the distributed practice group when compared with the massed practice group both for the training (Tr-TM Post) and transfer legs (Tf-TM initial and Tf-TM final), indicating that skill acquisition and interlimb transfer were higher after distributed practice.

Figure 5:

Reductions in tracking error in the training (i.e., skill acquisition) and transfer legs (i.e., interlimb transfer) of the massed and distributed practice groups. Bars represent the marginal means of the change in tracking error and error bars represent standard error of the mean. Asterisks indicate a significant difference at α = 0.05.