Abstract
Long-term retention of a motor skill has received relatively little systematic study, even though lasting neuroplasticity is the holy grail of any clinical intervention. This study examined the acquisition and retention of a novel bimanual polyrhythmic skill, practiced with sparse explicit feedback mimicking real-life scenarios. Self-paced and metronome-paced practice conditions were compared in their effect on long-term retention. Two groups of subjects first underwent extensive practice of 20 practice sessions over 2 mo, then followed up with three retention sessions after 3 mo. Results showed that subjects developed robust spatiotemporal patterns, despite the lack of reward and little quantitative error feedback about their performance (Hypothesis 1). These movement patterns were reproduced after a 3-mo interval, frequently even in the first trial, with no intermediate practice (Hypothesis 2). Self-paced training of movement patterns led to slightly less variability in the retention test (Hypothesis 3). These results document the specificity and stability of kinematic patterns and their underlying neuroplastic changes and underscore the effectiveness of self-guided practice. The findings are discussed in the context of current neuroimaging results and their clinical implications.