Long-term motor skill training with individually adjusted progressive difficulty enhances learning and promotes corticospinal plasticity

Abstract

Motor skill acquisition depends on central nervous plasticity. However, behavioural determinants leading to long lasting corticospinal plasticity and motor expertise remain unexplored. Here we investigate behavioural and electrophysiological effects of individually tailored progressive practice during long-term motor skill training. Two groups of participants practiced a visuomotor task requiring precise control of the right digiti minimi for 6 weeks. One group trained with constant task difficulty, while the other group trained with progressively increasing task difficulty, i.e. continuously adjusted to their individual skill level. Compared to constant practice, progressive practice resulted in a two-fold greater performance at an advanced task level and associated increases in corticospinal excitability. Differences were maintained 8 days later, whereas both groups demonstrated equal retention 14 months later. We demonstrate that progressive practice enhances motor skill learning and promotes corticospinal plasticity. These findings underline the importance of continuously challenging patients and athletes to promote neural plasticity, skilled performance, and recovery.

Figure 1

Motor performance and EMG activity during training. (A) Box- and density plots of performance from the first (left), middle (middle) and last (right) 4 min of the first training session. Each data point (green dot) represents the accumulated score from 4 min of training from one participant. Asterisk denotes a significant effect of time (p < 0.05, n = 24). (B) Box- and density plots of long-term comparisons of performance on the ‘Day 1’ task level for both groups. Asterisks denote significant difference from baseline performance and from Retention 1 to 2 across the two groups (p < 0.05, n = 24). (C) Box- and density plots of motor performance at ‘End’ task Level (n = 12). Asterisks denote a significant difference between the two groups (p < 0.05) and from Retention 1 to 2 within the PT group. (A–C) Coloured dots represent individual data. Whiskers represent highest and lowest value within 1.5 inter quartile range. Bold horizontal lines signify median values, the lower and upper hinges correspond to the first and third quartiles, and the means are represented by black dots. Please note that the scaling of Y-axes differs between the three violin plots. (D) ADM EMG activity, normalized to M_max, during the first and last training session for one participant from each group. No statistical comparisons were made for EMG activity obtained during training.

Figure 2

Electrophysiological results for the progressive (red) and non-progressive (blue) training group. (A) Pooled recruitment curves for both the PT Group (n = 12) and the NPT Group (n = 12 before and after the first day of training, as well as after 2, 4 and 6 weeks of motor practice. Dots represent motor evoked potential amplitudes normalized to M_max and then to baseline MEP_max. Green dots and curve fit (left) illustrate measurements obtained immediately after the first training session. Groups are merged for this fit because the protocol for the first training session was identical for all participants. Plots represent global fits to the complete dataset, and dotted lines represent confidence bands. No statistical tests were performed based on the presented global fits, which primarily serve to illustrate the data set. (B) Global recruitment curves obtained at Retention test 1 following 8 days of detraining. (C) Group mean area under recruitment curve (AURC) normalized to individual baseline values (represented by the dashed line). Asterisks denote a significant difference from ‘Baseline’ and asterisks next to vertical black lines denote a difference between the two groups. (D) Group mean MEP_max values normalized to individual baseline value (represented by the dashed line). Asterisks denote a significant difference from ‘Baseline’ and asterisks next to vertical black lines denote a difference between the two groups. (E) Group mean resting Motor Threshold (rMT) normalized to individual baseline rMT (represented by the dashed line). The asterisk denotes a significant difference from ‘Baseline’ across the two groups for all time points. For (C–E) error bars represent s.e.m.

Figure 3

General design of the study. Flow chart illustrating the design of the study. Please note that the study was conducted in two ‘batches’ of 12 participants each. The velocity and paddle size of the ‘End’ task level was established based the attained performance progression of the six participants engage in progressive skill training from the first ‘batch’, whereas the group-comparison of performance on the ‘End’ task level was based on six participants from each group of the last ‘batch’(N = 12).

Figure 4

Behavioural task and experimental Setup. The visuomotor training task consisted of a game called “BreakOut”, a spin-off from a classical arcade game. (A) A screen shot from the game shows a random in-game situation. (B) The paddle at the bottom of the screen was moved by rolling the trackball to the right or left by abducting or adducting the fifth digit. Hand position during motor practice. (C) Both of the subject’s arms and hands were strapped during electrophysiological measurements in order maintain stable hand and arm position. (D) Representative Motor Evoked Potentials (MEPs) normalized to M_max at different transcranial magnetic stimulation (TMS) intensities (normalized to resting motor threshold (rMT) for one subject.