Secondary Logo

Journal Logo


Neuroplasticity Promoted by Task Complexity

Carey, James R.; Bhatt, Ela; Nagpal, Ashima

Author Information
Exercise and Sport Sciences Reviews: January 2005 - Volume 33 - Issue 1 - p 24-31
  • Free


Skillful motor performance is recognized in routine daily activities, such as handwriting, keyboarding, or steering a vehicle. With tasks involving greater complexity of movement, “professional” skill can be recognized in pianists, baseball batters, ballet dancers, etc. However, skill, or rather the absence of it, can also be recognized in seemingly simple tasks such as standing balance, reaching, or grasping and releasing objects, as individuals with stroke or other neurological disorders train to regain these abilities. Thus motor skill, expressed as precision performance in time and space to meet the movement demands of a specific task, exists in essentially all we do. The training factors that optimize the development of these skills are important for all individuals, from the stroke survivor to the professional athlete. One such factor is the level of task complexity during training, and the neural resources needed to manage such complexity. The purpose of this review is to compare the effects of repetitive execution of simple motor tasks versus complex motor tasks involving in-depth cognitive processing to promote neuroplastic changes subserving skillful motor performance.


This concern for task complexity has been explored extensively in rats. The rat model allows for the examination of a neurological substrate that might account for behavioral observations, thereby yielding stronger validation of a scientific principle. One such neurological substrate generally thought to correlate to motor learning is synaptogenesis. Based on the premise that the cerebellum is integrally involved with motor learning, Black et al. (2) examined whether the synaptic density of Purkinje cells would show a difference in rats trained on a repetitive complex motor task versus those trained on a repetitive simple motor task. Randomized groups of rats were trained under four different conditions. The acrobatic condition (AC) was complex, requiring rats to ambulate across a sequence of balance beams, see-saws, rope bridges, and other obstacles. Rats training in the forced exercise (FX) condition simply walked on a treadmill for progressively longer periods each day. Rats trained in the voluntary exercise (VX) condition had free access to a running wheel. In the fourth condition, the inactive condition (IC), the animals remained in their cages without opportunity for learning or exercise. The AC rats had to repeatedly problem-solve progressively more difficult obstacle paths for 30 d, whereas the FX and VX rats repeatedly executed simple ambulatory movements. The AC rats showed dramatic improvements in the time required to traverse the obstacle course over the 30 d. After 30 d of training the rats were sacrificed and the paramedian lobule of the cerebellar cortex was examined for synaptic density in Purkinje cells. The AC rats had approximately 25% more synapses per Purkinje cell than the other groups (P < 0.003). This finding was particularly remarkable because the total distances traveled by the FX group and the VX group were roughly 10 times and 19 times, respectively, greater than that of the AC group. The authors interpreted these results to indicate that motor learning and, presumably, the associated in-depth information processing rather than mere motor activity, is key to inducing morphological changes lending to the acquisition of skill.

Kleim et al. (7) examined the relationship between synapse number, regional brain function, and learning. Rats were assigned to either a skilled reaching condition (SRC) or unskilled reaching condition (URC). The SRC involved trials requiring the rats to reach for and grasp a food pellet repeatedly over the first 5 d. For the next 5 d, the trials required the same rats to reach and grasp pellets from a table that rotated slowly, creating a moving target. The URC required rats to simply press a lever, dispensing a food pellet that could be retrieved by tongue and mouth movement; thus, grasping with the forepaw was minimized. Four hundred trials occurred per day for each group. Electrophysiological mapping occurred near the end of the experiment. With the animal anesthetized, a stimulation microelectrode was penetrated into multiple sites of the motor cortex contralateral to the paw used most frequently. Intracortical microstimulation (ICMS) was then applied at these sites at a specified level and the resultant response of digit, wrist, or elbow/shoulder movement (or else no movement) was recorded. With repeated penetrations and stimulations of the motor cortex in 250-μm intervals, a motor map indicating the location and extent of the movement representations in the motor cortex was created. Map sizes were compared between the SRC and URC groups. Finally, the investigators determined the number of synapses per neuron within layer V of specified motor areas.

The percentage of reaches yielding successful pellet retrieval in the SRC showed significant improvement (P < 0.001). Analysis of motor map size showed that the SRC group had significantly larger area compared to the URC group for digit (P < 0.03) and wrist (P < 0.006) movement representation in the caudal forelimb area, whereas the URC group showed significantly larger area for elbow/shoulder representation (P < 0.0006) (Fig. 1). The authors suggested that the enlarged digit and wrist representation in the SRC animals may have occurred at the expense of their elbow/shoulder representation, and that this shift may likely be connected with the development of grasping skill. The number of synapses per neuron within the caudal forelimb area were significantly higher in the SRC animals compared to the URC animals (P < 0.05). These results are important because they show that morphological differences between complex versus simple training conditions exist not only in the cerebellar cortex, as described above by Black et al. (2), but also in the cerebral cortex. Moreover, the difference in representation maps between the two groups in the same location where the synapse count was also different gives further credibility to the notion that morphological differences (synapses) are connected to functional differences (motor neuron excitability). Thus, the case for enhanced motor learning through repetitive complex tasks versus repetitive simple tasks grows stronger.

Figure 1.:
Comparison of areas of motor maps responding to intracortical microstimulation following training in rats in a skilled reaching condition (SRC), involving grasping, and in an unskilled reaching condition (URC), involving no grasping. SRC rats showed significantly (P < 0.05) larger representations for digit and wrist movements and significantly less representations for elbow and shoulder movements in the caudal forelimb area, suggesting that training enlarged the distal representations at the expense of the proximal representations. (Reprinted from Kleim, J.A., S. Barbay, N.R. Cooper, T.M. Hogg, C.N. Reidel, M.S. Remple, and R.J. Nudo. Motor learning-dependent synaptogenesis is localized to functionally reorganized motor cortex. Neurobiol. Learn. Mem. 77:63–77, 2002. Copyright © 2002 Elsevier. Used with permission.)


Similar results in a different animal model strengthen this case. Two separate but related papers addressed the question of simple versus complex repetitive activity and its effect on motor skill enhancement and neuroplasticity. The first study was conducted by Nudo et al. (10). They assigned three monkeys to a digit-training task group and three to a control group. First, a movement representation map was created in all monkeys with ICMS by methodically penetrating the primary motor cortex across a grid with 250-μm intervals and stimulating each site with a specified intensity of current to determine the category (digit, wrist, or forearm) of evoked forelimb movement, if any. Next, digit-training task animals underwent training that required them to retrieve food pellets from a small well. The diameter of the well was such that the animals could not reach in with their full hand to retrieve the pellet. Instead, they had to figure out how to insert a single digit into the well and, through flexion-extension movements of that digit, pull the pellet from the well. This training continued until the number of daily retrievals exceeded 600 pellets for two consecutive days, which required a minimum of 11 d. The control animals received no training. The representation map was then reconstructed, with the area representing digit flexion-extension expressed as a percentage of the entire distal forelimb area. The digit-training task animals showed significant (P < 0.05) improvement in performance, indicated on videotape by the decline in the number of digit flexions required for a successful pellet retrieval over the training period. Also, their relative finger flexion-extension representation area increased significantly (P < 0.05) from roughly 33% of the distal forelimb area at pretraining to 46% at posttraining, compared to no change in control animals (34% pretraining and posttraining).

This study becomes more important when considered alongside the next study. Plautz et al. (12) performed a similar experiment using ICMS, except that the size of the wells holding the food pellets was considerably larger, allowing the animals to use whole-hand (all digits at once) grasping, as compared to the individual-digit grasping used by Nudo et al. (10). The training period was equivalent. Their results showed no improvement in performance, as retrieval was successful with only one repetition even at the beginning of the experiment. Also, no changes were found in the representation area from pretraining to posttraining. Figure 2 summarizes the results of the combination of these two studies. In considering the combined results, Plautz et al. (12) proposed a learning-dependent hypothesis for cortical plasticity, emphasizing that skill acquisition, as opposed to simple repetitive movement, is required for cortical reorganization to occur. However, an alternative view is that the depth of cognitive effort involved in processing the complex task to be learned might induce the neuroplastic change, which in turn might lead to the acquisition of skill.

Figure 2.:
Comparison of effects of small-well versus large-well training in monkeys. (A) Shows improved performance, indicated by the declining number of finger flexions required to retrieve food pellets in a monkey training in the small-well condition and no such improvement in a monkey training in the large-well condition. (B) Shows expanded motor map responding to intracortical motor stimulation in one monkey pre- and posttraining with small finger well. (C) Shows the sizeable increase in percentage of total area represented by digits in the group of monkeys that trained with the small finger well. (Fa, forearm; Error bars, 1 standard deviation) (Reprinted from Plautz, E.J., G.W. Milliken, and R.J. Nudo. Effects of repetitive motor training on movement representations in adult squirrel monkeys: role of use versus learning. Neurobiol. Learn. Mem. 74:27–55, 2000. Copyright © 2000 Elsevier. Used with permission.)


Cortical motor neuron excitability changes, assessed in animals with ICMS, can also be assessed in humans using transcranial magnetic stimulation (TMS). Pascual-Leone et al. (11) examined the effects of complex versus simple finger movement tasks on the piano in people with no experience playing piano. The test group performed a five-finger piano exercise in which each digit of one hand pressed a designated piano key with a specified order and time interval that required cognitive effort to accomplish with precision. This exercise was done repeatedly for 5 d. On each day, a special stimulation electrode was applied to the scalp methodically at different positions with 1-cm intervals. At each position, the intensity of the TMS was adjusted over repeated stimulations to determine which sites would produce a motor-evoked potential (MEP), as detected by EMG electrodes at either a finger flexor muscle or finger extensor muscle contralaterally. This testing was done for both the trained and untrained hands. Thus, a motor map was created that revealed the size of the area that was successful in eliciting an MEP for each hand. Two control groups were used. Control 1 had TMS mapping only, without any training. Control 2 had TMS mapping and training, but the training involved executing finger movements in whatever random sequence the subjects wanted, minimizing their cognitive effort. On the last day subjects in the two control groups were also trained in the test sequence to determine their frequency of errors. Results showed improved performance at the task in the test group over the training period, as measured by reduced errors in the sequence. Also, their error rate was considerably lower than the two control groups after they were tested on the last day. The motor maps indicated that the areas of successful stimulation increased sizably for the trained hands of the test group. It also increased for the trained hand in the control 2 group, but the enlargement was significantly greater in the test group compared to the control 2 group (P < 0.001) (Fig. 3). These results are important because they complement animal studies, suggesting now in humans that complex tasks (designated finger sequence and timing) are more effective than simple tasks (random finger sequence and timing) in producing physiological changes (increased cortical excitability) consistent with skillful performance.

Figure 3.:
Representative examples of cortical motor maps showing the area over which transcranial magnetic stimulation succeeded in evoking responses in contralateral finger flexor and extensor muscles on each of five training days. Maps for the Trained Hand show progressively larger areas of cortical excitability compared to the Untrained Hand of the same subject who trained with a specified finger exercise to promote skill. In contrast, maps for the Control Subject, who performed a nonspecific task involving no skill, show some increase in area but not as much compared to subject who trained for skill. (Reprinted from Pascual-Leone, A., K.T. Nguyen, A.D. Cohen, J. Brasil-Neto, A. Cammarota, and M. Hallett. Modulation of muscle responses evoked by transcranial magnetic stimulation during the acquisition of new fine motor skills. J. Neurophysiol. 74:1037–1045, 1995. Copyright © 1995 American Physiological Society. Used with permission.)

It should be noted, however, that simple repetition can influence cortical excitability. Classen et al. (4) studied the direction of thumb movements evoked by TMS applied to healthy human subjects before and after approximately 30 min of training involving simple voluntary repetitive thumb movements. The direction of the voluntary thumb movements was opposite the direction of the TMS-evoked movements at pretraining. They found that the evoked movements at posttraining were in the direction that was trained—not in the direction observed at pretraining. However, this reorganization of cortical excitability was short-lived, as continued testing with posttraining TMS showed that the evoked movements reverted to the original pretraining direction within 30–35 min following training (Fig. 4). Thus, although repetition of simple movements does show evidence of inducing neuroplastic changes, these changes appear to be transient, and therefore not likely to lend to the acquisition of skillful motor performance.

Figure 4.:
Shows the direction of thumb movements responding to transcranial magnetic stimulation (TMS) at pretraining, and the change to the opposite direction at posttraining, resulting from 30 min of training with voluntary movements in the direction opposite to the pretraining direction. The posttraining records show that the TMS-evoked movements revert to the original (pretraining) direction in 30–35 min. (Reprinted from Classen, J., J. Liepert, S.P. Wise, M. Hallett, and L.G. Cohen. Rapid plasticity of human cortical movement representation induced by practice. J. Neurophysiol. 79:1117–1123, 1998. Copyright © 1998 Elsevier. Used with permission.)

A mechanism has not been established explaining why complex tasks, with their inherently greater depth of cognitive processing, may promote greater neuroplastic change than simple tasks. However, one possible explanation is the expression of trophic factors in the brain. Gomez-Pinilla et al. (5) used the Morris water maze to examine for differences in the expression of basic fibroblast growth factor messenger RNA (FGF-2 mRNA), which is thought to contribute to healthy brain function. In the Morris water maze, rats swim through a maze while processing external spatial cues as they search for a submerged platform to find safety. The investigators explored whether training involving the combination of physical activity and learning activity could induce a greater expression of FGF-2 mRNA than physical activity alone. One group of rats (learning) was placed inside the maze at different locations repeatedly. They swam and processed spatial cues until they found the platform. Another group of rats (yoked) swam for the same amount of time as their matched mates in the learning group but no processing of cues occurred, as there was no platform to locate. After training, the animals were sacrificed and the results showed that both of these groups had higher levels of FGF-2 mRNA expression than a sedentary control group for select regions of the brain and time points. Furthermore, the learning group showed higher levels of FGF-2 mRNA expression than the yoked group in the hippocampus and cerebellum under certain conditions. The results suggest that the more complex task involving both physical activity and spatial processing potentiated the effects of physical activity alone on the expression of trophic factors in certain brain regions. Thus it seems plausible that elevation of such trophic factors to some threshold level might be an important ingredient to promote long-lasting neuroplastic change and skill acquisition and, furthermore, that the in-depth cognitive processing associated with complex tasks may achieve this threshold more readily than simple repetitive tasks.

The importance of this information is that it might inform training strategies to optimize the stimulus for morphological and physiological changes that presumably serve as the groundwork for skillful performance. Our laboratory has employed this strategy in a stroke recovery training program (3). Subjects with stroke who had impaired finger function were assigned to either a training group or a control group. The training group received approximately 20 1-h sessions involving finger movement tracking training. A computer screen displayed several cycles of a target pathway (squarewave, triangle wave, or sawtooth wave). With an electrogoniometer attached to the index finger and connected to the computer, subjects attempted to track the target as accurately as possible with careful finger extension and flexion movements. A host set of 60 different tracking protocols was created in advance with varying levels of difficulty depending on the waveform, movement speed, movement amplitude, and hand position. For each session the subject tracked 20 different protocols assigned randomly from the host set. For each protocol the subject performed a block of three tracking trials before moving to the next randomly assigned protocol and its block of three trials. Thus a total number of 60 tracking trials occurred at each session. The frequent switching of protocols after only three trials created what we considered to be a reasonable level of difficulty. Making the task too difficult by switching protocols after each trial rather than every three trials could be counterproductive in training skillful performance (15).

The effect of this training was evaluated with a finger tracking test using a sine wave as the target wave form. This test was performed before and after training and involved simultaneous functional magnetic resonance imaging (MRI) of the brain to visualize the extent of cortical activation. Also, a transfer test involving repeated grasp and release of small blocks was performed. The results showed significant improvement in tracking performance (P = 0.022) and in grasp and release of the blocks (P = 0.014) in the training group, with no improvement in the control group. Furthermore, cortical activation was shown to switch from predominantly ipsilateral activation before training to predominantly contralateral activation after training (Fig. 5). The control group, which received no training between the pretest and posttest, showed predominately ipsilateral activation on both of these tests. However, a partial crossover design ensured that, after their posttest, the control group received 20 tracking training sessions. On the postcrossover test, these subjects did demonstrate a switch from ipsilateral to contralateral activation.

Figure 5.:
Examples of pre- and posttest finger tracking responses for one subject with stroke using his paretic right hand before and after 20 treatments of finger movement tracking training. Tracking responses occurred simultaneously with functional MRI, which revealed a shift in cortical activity from predominantly ipsilateral activation at pretest to predominantly contralateral activation at posttest. (Reprinted from Carey, J.R., T.J. Kimberley, S.M. Lewis, E. Auerbach, L. Dorsey, P. Rundquist, and K. Ugurbil. Analysis of fMRI and Finger Tracking Training in Subjects with Chronic Stroke. Brain 125:773–788, 2002. Copyright © 2002 Guarantors of Brain. Used with permission.)

Although this study controlled for training versus no training, it did not control for task difficulty (complexity). Consequently, we could not conclude with certainty that the in-depth cognitive processing associated with the tracking task served as the stimulus for the behavioral improvement and brain reorganization, or whether the physical activity of repeated finger extension/flexion motions caused the change. The repetitive finger motions, irrespective of the cognitive processing required for accurate tracking, might have triggered the changes, although this is doubtful based on the studies cited above. Indeed, in a subsequent study in subjects with stroke, Kimberley et al. (6) showed that repetitive movement of the fingers produced by a combination of electrical stimulation and voluntary activation, without any cognitive effort associated with temporal or spatial processing of the task, did not produce such a shift in cortical activation. Thus we believe that in-depth cognitive processing of the task complexities, rather than mere repetition of simple movements, has merit in strategies for stroke rehabilitation. Further studies are underway to explore this more definitively. Figure 6 depicts the purported advantage of training with problem solving (tracking) versus training with simple movements alone.

Figure 6.:
Speculated advantage of training with repetitive movement and problem-solving (finger tracking) versus repetitive movement alone. The postulate is that the cognitive engagement during tracking invokes greater potentiation of molecular mechanisms (e.g., expression of neurotrophins) than repetitive movement alone. These mechanisms may serve as the stimulus for neuroplastic changes (e.g., synaptogenesis, synaptic efficacy, neurogenesis). Ultimately, activation in neural centers of the brain associated with certain movements enlarge or intensify, which may form the substrate for improved behavioral performance.


Overall, the above studies provide evidence of a neural substrate that may account for improved skill acquisition following repetitive training on motor tasks requiring in-depth information processing to manage the task complexities, compared to tasks without such cognitive demand. One theory for why motor tasks involving in-depth cognitive processing are more effective than simpler tasks in training for skill emphasizes contextual interference (CI). CI is defined as the degree of functional interference found in a practice situation when several tasks must be learned and are practiced together (9). Practice at a task involves multiple conditions, such as the kinematic requirements of the task itself, the practice schedule, feedback, etc., and all are viewed as potential sources of interference that can affect the acquisition of skill. For example, a random practice schedule involving high CI has been shown to produce better retention and increased adaptability to novel transfer tasks, as compared to a blocked practice schedule (low CI) (9,15). The advantage of using a random schedule rather than a blocked one is attributed to the increasingly complex and effortful cognitive processing that is used in random practice conditions. These multiple and variable encoding processes accompanying the contextual interference in random practice are proposed by Shea and Zimny (13) as the “Elaboration Benefit” explanation. During practice on any trial, in a high CI effect situation, multiple tasks reside in the working memory simultaneously, increasing the chances for inter- and intraskill comparisons to achieve distinctiveness. As multiple encoding strategies are employed during a high CI condition, there is elaboration of memory representations of the skills being learned. This elaborate range of memory of movements benefits retention and transfer tests, thanks to the availability of multiple retrieval routes (9). Thus subjects engaging in random practice will be able to compare multiple variations of the task and possess an efficient repertoire of strategies. Those engaging in blocked practice show more automated processing, leading to impoverished encoding and strategy retrieval, because the various skills are never challenged together in the working memory.

An alternative explanation for the favorable effect of CI is the “Action Plan Reconstruction” hypothesis proposed by Lee and Magill (8). In a random practice schedule, decay of information related to a recently produced movement occurs because variations of that task are presented on the next trial. Therefore, subjects have to make frequent calls to the long-term memory to effortfully reconstruct the action plan for that trial when it comes up again. This aids novel task transfer because action plan reconstruction is a necessary attribute of novel tasks (9).

The concept of contextual interference has been put to use both in laboratory and sports skills, but these have been applied typically to gross motor skills. However, Ste-Marie et al. (14) applied these concepts to fine motor skills requiring spatial accuracy. They established that a random practice schedule in handwriting skill acquisition in children was effective in promoting higher retention test scores and better writing speeds on the transfer test, as compared to a blocked practice schedule. They argued that random practice provided a more challenging learning environment for children. Random training required that the children execute various motor patterning demands simultaneously as opposed to blocked training, in which subjects repeatedly executed the same motor commands. This discussion demonstrates the value of deep, elaborate cognitive effort exerted by the learner during random practice.

But does increasing the cognitive effort guarantee better learning? This questions was explored by Albaret et al. (1) in a drawing task in which subjects had to reproduce patterns presented on a screen without visual control in a random or a blocked training schedule. Results indicated that the random training group was more accurate than the blocked training group on delayed retention and transfer tests when the tasks practiced were the simplest drawing movements composed of two and three segments. However, there was no difference in the accuracy of the groups when the task was the most complex (involving four segments). The authors defended their results by indicating that in cases of highly complex tasks, the benefits of intertask interference created by a random schedule can be obscured by the intratask interference associated with the difficulty of adding a fourth segment to the task.

Relatedly, Wulf and Shea (15) provided an extensive review of whether the principles of skill acquisition developed for learning simple motor skills apply equally to learning complex skills. Ultimately, they contended that for difficult tasks, involving extensive attention, memory, and/or motor demands, conditions of high contextual interference may overload the information processing capacity of the individual and remove the benefit that such interference normally fosters during skill acquisition. Despite the substantial evidence showing that motor tasks involving more effortful cognitive processing compared to simpler tasks produce greater neuroplastic changes consistent with acquisition of skill, the possibility exists that cognitive overload could thwart these biological changes and hinder skill acquisition. More research is needed to discover the cascade of intervening mechanisms invoked by in-depth cognitive processing that leads to beneficial neuroplastic changes (synaptogenesis, synaptic efficacy, neurogenesis, etc.) during motor skill acquisition, and how best to optimize these mechanisms.


Morphological, physiological, and behavioral evidence existing across animals and humans suggests that acquisition of motor skill is enhanced by training conditions involving complex tasks and in-depth cognitive processing, compared to less difficult tasks. However, there may be a limit above which further increase in task complexity and cognitive effort is not beneficial. These results call for careful judgment in incorporating the proper level of CI in skill training; such judgment might be one of the distinguishing features of outstanding coaches, teachers, and therapists.


1. Albaret, J.M. and B. Thon. Differential effects of task complexity on contextual interference in a drawing task. Acta Psychologica 100:9–24, 1998.
2. Black, J.E., K.R. Isaacs, B.J. Anderson, A.A. Alcantara, and W.T. Greenough. Learning causes synaptogenesis, whereas motor activity causes angiogenesis, in cerebellar cortex of adult rats. Proceedings of the National Academy of Sciences of the United States of America. 87:5568–5572, 1990.
3. Carey, J.R., T.J. Kimberley, S.M. Lewis, E. Auerbach, L. Dorsey, P. Rundquist, and K. Ugurbil. Analysis of fMRI and Finger Tracking Training in Subjects with Chronic Stroke. Brain 125:773–788, 2002.
4. Classen, J., J. Liepert, S.P. Wise, M. Hallett, and L.G. Cohen. Rapid plasticity of human cortical movement representation induced by practice. Journal of Neurophysiology 79:1117–1123, 1998.
5. Gomez-Pinilla, F., V. So, and J.P. Kesslak. Spatial learning and physical activity contribute to the induction of fibroblast growth factor: neural substrates for increased cognition associated with exercise. Neuroscience 85:53–61, 1998.
6. Kimberley, T.J., S.M. Lewis, E.J. Auerbach, L.L. Dorsey, J.M. Lojovich, and J.R. Carey. Electrical Stimulation driving functional improvments and cortical changes in subjects with stroke. Exp Brain Res. 154:450–460, 2004.
7. Kleim, J.A., S. Barbay, N.R. Cooper, T.M. Hogg, C.N. Reidel, M.S. Remple, and R.J. Nudo. Motor learning-dependent synaptogenesis is localized to functionally reorganized motor cortex. Neurobiol. Learn. Mem. 77:63–77, 2002.
8. Lee, T.D., and R.A. Magill. The locus of contextual interference in motor skill acquisition. J. Exp. Psychol. Learn. Mem. Cogn. 9:730–746, 1983.
9. Magill, R.A., and K.G. Hall. A review of the contextual interference effect in motor skill acquisition. Hum. Mov. Sci. 9:241–289, 1990.
10. Nudo, R., G. Milliken, W. Jenkins, and M. Merzenich. Use-dependent alterations of movement representations in primary motor cortex of adult squirrel monkeys. J. Neurosci. 16:785–807, 1996.
11. Pascual-Leone, A., K.T. Nguyen, A.D. Cohen, J. Brasil-Neto, A. Cammarota, and M. Hallett. Modulation of muscle responses evoked by transcranial magnetic stimulation during the acquisition of new fine motor skills. J. Neurophysiol. 74:1037–1045, 1995.
12. Plautz, E.J., G.W. Milliken, and R.J. Nudo. Effects of repetitive motor training on movement representations in adult squirrel monkeys: role of use versus learning. Neurobiol. Learn. Mem. 74:27–55, 2000.
13. Shea, J.B., and S.T. Zimney. Context effects in memory and learning movement information. In R.A. Magill (Ed.), Memory and the Control of Action. Amsterdam: North-Holland, 345–366, 1983.
14. Ste-Marie, D.M., S.E. Clark, L.C. Findlay, and A.E. Latimer. High levels of contextual interference enhance handwriting skill acquisition. J. Mot. Behav. 36:115–126, 2004.
15. Wulf, G. and C.H. Shea. Principles derived from the study of simple skills do not generalize to complex skill learning. Psychon. Bull. Rev. 9:185–211, 2002.

motor learning; neuroplasticity; psychomotor skill; rehabilitation; contextual interference

©2005 The American College of Sports Medicine