| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
INVITED EDITORIAL
Department of Integrative Physiology
University of Colorado
Boulder, Colorado Prince of Wales Medical Research Institute and the
University of New South Wales
Sydney, New South Wales, Australia
e-mail: s.gandevia.jap{at}unsw.edu.au
THE PURPOSE OF THE MINI-REVIEWS on this Highlighted Topic is to provide a contemporary synthesis of knowledge on the changes that occur in the physiological properties of the nervous system in response to repeated exposure to physical training such as exercise. The focus of the series is on the motor system, which corresponds to those parts of the nervous system involved in the production of movement. A key goal in this field is to identify the adaptations that are responsible for the changes in performance evoked by a training intervention. As indicated in the mini-reviews, the level of inquiry on most questions is limited to examination of associations between neural adaptations and changes in performance, because it is more difficult to identify causal relations.
Two features of this work deserve emphasis: the approaches that are used to study the neural changes associated with training and the methodological issues that constrain this field. One end of the range of approaches is exemplified by work that describes a training adaptation and subsequent studies that attempt to identify the mechanisms responsible for the change in performance. This approach is used by Tim Carroll and colleagues to characterize the phenomenon known as the contralateral strength training effect, which appears to be mediated by mechanisms that reside within the nervous system. The performance of a strength-training program by muscles in one limb can evoke a small strength gain in the homologous muscles of the untrained limb. Carroll et al. propose that the contralateral strength training effect might be caused by either of two classes of neural mechanisms: one is related to the spread of cortical activity to the pathways for the contralateral limb, and the other involves adaptations in the control system for the trained limb that can be accessed by the contralateral limb. Although the specific mechanisms responsible for the contralateral effect have not been identified, the adaptations are likely to be distributed among cortical, subcortical, and spinal levels. This approach provides a template for the investigation of neural mechanisms that cause improvements in performance in response to training.
The other end of the range can be characterized by studies that examine the effects of training on the function of the nervous system and seek to determine the mechanisms that produce the adaptation. This approach is exemplified by studies on the effects of exercise on cognition and dementia and is reviewed by Art Kramer. Both observational and intervention studies indicate that aerobic exercise can improve performance on various measures of cognition in humans. Although these studies have produced mixed results, meta-analyses suggest that aerobic fitness training has a positive influence on cognition. Interestingly, the training effects are greater when aerobic training is combined with strength and flexibility training. Attempts to identify the responsible mechanisms have involved imaging studies in humans and invasive studies on experimental animals. The animal models have provided evidence that voluntary exercise improves performance on learning and memory tasks, enhances a cellular process in learning (long-term potentiation) that upregulates molecular factors associated with brain plasticity, and promotes growth of new neurons and vasculature in old animals. These results indicate that physical activity and aerobic exercise can have a positive influence on cognition, brain function, and brain structure. Furthermore, levels of social interaction that can accompany aging and exercise are likely to modulate this positive effect (9).
Most studies use approaches that lie along this spectrum and tend to focus either on assessment of the adaptability of various components of the nervous system, such as motoneurons (considered in the mini-review by Phillip Gardiner and colleagues), motor units (Jacques Duchateau and colleagues), somatosensory reflex pathways (Paul Zehr), and the corticospinal system (DeAnna Adkins and colleagues), or on identification of the constraints imposed by the nervous system on the adaptations evoked by training (Richard Carson). These studies often do not attempt to determine the functional consequences of a specific adaptation. For example, Gardiner et al. describe work that focuses on molecular mechanisms that might be responsible for changes in the biophysical properties of motoneurons in response to exercise programs. An increase in voluntary activity can influence dendrite structure, protein synthesis, axonal transport, neuromuscular propagation, and other biophysical properties. These technically difficult studies demonstrate that motoneurons are responsive to the demands imposed by physical activity, but they have not yet indicated a causal relation between a cellular adaptation and an improvement in performance. Similarly, Zehr describes studies that found evidence of plasticity in muscle afferent pathways in response to various training interventions. For example, several weeks of strength training increase strength and the amplitude of the H reflex. These studies, however, do not indicate the relative significance of the change in a reflex pathway to the accompanying strength gains or improvements in coordination.
Fortunately, there are some paradigms that do enable a more direct comparison between changes in the function of the nervous system and motor performance. For example, Van Cutsem et al. (10) found that a training intervention with rapid, submaximal contractions improved both the rate of torque development and the instantaneous discharge rate of motor units. This result was interpreted to indicate that the more rapid increase in torque after training was attributable to an increase in the rate at which motor units could discharge action potentials. Of course, the observation does not indicate whether the adaptation was due either to an increase in the intrinsic capacity of motoneurons to discharge rapidly or to a change in the synaptic input received by the motoneurons.
Another example of an association between changes in nervous system function and motor performance was obtained in a study of coordination. Carson and Riek (2) trained subjects to perform abduction-adduction movements with the index finger to match the increasing frequency of a pacing metronome. The improvement in performance was accompanied by a change in the coordination of the intrinsic and extrinsic hand muscles involved in the task. Although the greatest changes occurred in the extrinsic muscles, the specific adaptation varied across individuals. The conclusion was that improved performance required learning of a new pattern of muscle activation. This is a first step, and the challenge is now to identify the central adaptations that drive the change in strategy.
These examples of a neural adaptation and its contribution to a change in performance underscore a basic tenet in science: progress depends critically on the technology available to address a question (5, 7). The observation by Van Cutsem et al. (10), for example, begs the question of what adaptations enabled the motoneurons to discharge action potentials at a greater rate. The classic approach to evaluate the potential role of changes in the biophysical properties of the motoneurons is described in the minireview by Gardiner et al. In this paradigm, the properties of motoneurons in experimental animals are quantified when the preparation is quiescent and usually anesthetized. One obvious concern with such measurements, as emphasized by Gardiner et al., is the relevance of the findings to the in vivo state when the neurons are bathed in neuromodulators that can alter their function (6). It would be preferable to determine the motoneuron properties in the behaving state, but this technology is not yet available. In the absence of this, an alternative strategy is to develop a computational model. They provide an example of this strategy with a model of motoneurons to estimate the changes in ionic conductances that could produce the observed changes in electrophysiological properties. In general, agreement between model output and experimental observations does not mean that the intermediate processes in the model replicate actual events (11).
Another constraint indicated by the study of Van Cutsem et al. (10) is sample size. In most motor unit studies it is only possible to record the behavior of a few motor units, and these differ before and after an intervention. A common assumption in this work is that the function of a motor unit population can be characterized by the behavior of a few motor units. One approach, that can yield qualitative information about motoneuronal output across different groups of subjects is to record the discharge properties of large samples of units during the same task (3). However, when an intervention imposes a differential effect across the motoneuron pool, the minimal requirement is for the study to record a sufficient number of motor units to span the range of properties present in the population. Ideally, the same motor units would be recorded before and after the intervention. Technical capabilities limit both prospects: spanning the range of properties and recording from the same units.
Methodological issues in this field also involve the continual validation of standard techniques as new methods are developed. Two examples illustrate this point: interference electromyogram and the Hoffmann (H) reflex. Despite decades of advising caution in the interpretation of the interference EMG (1), it is only within the last decade that the magnitude of the limitations has become evident (4). Similar concerns apply to the H reflex. Although it is incorrectly regarded as an electrical analog of the stretch reflex (8), the readily evoked H reflex is often used as a measure of the neural changes that can occur during training. Although some investigators use the amplitude of the H reflex as an index of the excitability of a motoneuron pool, this interpretation is overly simplistic. The amplitude of the H reflex is influenced by a number of factors, including presynaptic inhibition, disynaptic inhibition, axonal excitability of both motor and sensory axons, as well as the procedures used to normalize the size of the reflex (8). Most importantly, as emphasized in the mini-review by Zehr, reflex amplitudes vary with the functional state of the individual, and this complicates the interpretation of the response.
Perhaps the most significant methodological limitation in this field involves the inability to access central circuits during behavior. Advances in medical imaging appear capable of providing some new insight at this level of inquiry. As described in the mini-review by Kramer et al., for example, functional magnetic resonance imaging has revealed an increase in neural activity in the frontal and parietal regions in the brain of aerobically trained older adults during a focused attention task. Such procedures will not reveal the cellular adaptations responsible for the changes in coordination described in the mini-review by Carson. Instead, it will be necessary to use the approaches described in the mini-review by Adkins et al. to examine the cellular and molecular events associated with changes in performance. Although it is challenging to demonstrate a direct link between a cellular adaptation and a change in performance, such reductionism is necessary and a hallmark of progress in science (11).
Despite the advances in field, the state of the knowledge is rather rudimentary. Although it is possible to demonstrate neural changes in response to a training intervention, few causal relations have been identified. Furthermore, these relations largely involve details about the motor output from the spinal cord, and there are not yet links to the underlying cellular and molecular events. The challenge is formidable, but the reviews indicate that progress is inevitable.
GRANTS
The authors’ work receives major support from the National Institutes of Health (to R. M. Enoka) and the National Health and Medical Research Council (to S. C. Gandevia).
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
