Across the literature, no studies have examined asymmetry of cross-education of strength in the lower body. Of interest is that no lower limb training studies have specified the limb dominance of participants (e.g., distribution of right- vs left-legged individuals), no studies have targeted exclusively right- or left-footed individuals, and the vast majority of studies have trained the left leg or nondominant leg (usually determined by kicking preference). Cross-education is present under a variety of training conditions in the lower limbs, and the magnitude of effect varies greatly, ranging from no significant effect up to 77% (13), with the largest magnitudes reported after left-leg training (13). In contrast to the upper limb training studies, it is difficult to discern a behavioral pattern of cross-education of strength.
A noticeable trend in the literature is that the studies reporting the highest magnitudes of cross-education also used more novel strength tasks. There are four cross-education studies where the relative strength increase in the untrained limb met or exceeded the relative strength increase observed in the trained limb (7,9,13,20). All four of these used novel strength tasks for training, and three of them used the dominant limb for training (7,9,20). However, if strength asymmetry is evident before training, a similar absolute change in strength in both limbs will result in a larger relative change in strength for the weaker limb (9). This is likely the case when the strength task is novel and the dominant limb is used for training. Another consideration is that unilateral strength training with a more novel task could result in more associated movements of the nontraining limb. Some cross-education studies report a small amount of activation in the nontraining limb during training (9,13). Although the small amount of activation is unlikely to account for large strength increases in the untrained limb, it is one potential mechanism that could increase the magnitude of effect in some studies.
CORTICAL ADAPTATIONS WITH CROSS-EDUCATION
There is general consensus in the literature that skill transfer between limbs is a motor learning response controlled by cortical adaptations in the brain. Motor learning has been associated with plasticity in the brain, including regions of primary motor cortex (M1), premotor cortex, and supplementary motor area (19,22,25,28). Of these, plasticity in M1 is probably the leading candidate because it plays a role in early learning of motor skills (22). There is evidence from functional magnetic resonance imaging (fMRI) and transcranial magnetic stimulation (TMS) studies that activity occurs in M1 contralateral and M1 ipsilateral to an active limb (27). Specifically regarding skill transfer between limbs, Perez et al. (29) reported that modulation of interhemispheric inhibition between ipsilateral and contralateral M1 after skill acquisition with one hand might lead to improved performance of the opposite hand. We have theorized that similar adaptations could occur in the context of unilateral strength training, especially if the task is more complex or novel (9). The theory that cross-education of strength is controlled by cortical adaptations is not new, but few studies have been able to measure changes at this level with strength training. Hortobágyi et al. (15) recently found that repetitive TMS of M1 delivered during strength training diminished the magnitude of strength increase observed after 4 wk of training. They suggested that M1 may play an important role in neural adaptations associated with strength training. Consistent with these findings, Griffin and Cafarelli (12) found evidence that cortical excitability increases within 4 wk of strength training. However, the role of cortical adaptations in response to strength training is controversial. Studies in both animals and humans have compared skill training with strength training and found evidence of cortical adaptations only after skill training (18).
Using fMRI, our laboratory has documented the first attempt to examine changes in brain activation after unilateral strength training (7). An active region of the brain as detected by fMRI is indicative of an increase in the amount of oxygenated blood used by a cortical region. Therefore, a brain region that is shown to exhibit more activation is thought to be more physiologically active. We used the same isometric ulnar deviation training protocol as our previous study (9) and again found a large cross-education effect (Fig. 2). Unilateral strength training of the right hand was associated with changes in brain activation in both cortical hemispheres. For the untrained left arm, there was an increase in activation of the contralateral sensorimotor cortex and ipsilateral temporal lobe regions (Fig. 3). For the trained right arm, there was also evidence of increased activation of the contralateral sensorimotor cortex. We compared these changes with those of a nontraining control group who performed motor imagery through the intervention period and did not show changes in strength (Fig. 2). The nontraining group did not exhibit similar changes in sensorimotor cortex or temporal lobe for the left arm, and the data were much more variable (Fig. 3). Increased fMRI activation (as an indicator of more physiological activity) in M1 after strength training is somewhat similar to previous studies of motor learning that report M1 plasticity or enlarged M1 activation (19,25) or reduced interhemispheric inhibition after task acquisition by one hand (29). Furthermore, increased M1 activation is consistent with strength training studies where M1 might play a role in neural adaptations (15).
A novel finding of our functional imaging study (7) was that left temporal lobe was part of a new activation associated with the untrained left limb after right-limb training (Fig. 3). Because regions of middle left temporal lobe are involved in the retrieval of motion knowledge and semantic memory, we hypothesized that left temporal lobe activation during left-arm contractions might indicate memory retrieval of the strength task previously acquired by the opposite trained right limb. The type of information stored in memory could be related to the specific motor control requirements of the strength task, which would aid in motor planning and execution.
The findings from our successive studies and others support a behavioral connection between cross-education of strength and skills. Our functional imaging study showed that changes in activation of contralateral motor areas are similar for each arm after unilateral strength training (7), suggesting that there may be changes in neural networks in the brain that are shared between cortical hemispheres. This is consistent with previous evidence that modulating interhemispheric inhibition between M1 regions is involved in skill transfer (29). In the near future, TMS studies will likely emerge with the purpose of evaluating interhemispheric inhibition between M1 regions in the context of strength transfer. It is important to recognize that there has yet to be an experiment that has directly compared the cortical adaptations associated with cross-education using strength versus skill training. Jensen et al. (18) compared skill versus strength training and found cortical adaptations only after skill training. Because the contralateral limb was not compared, the role of cortical adaptations in the cross-education of strength versus skill remains unclear. The amount of cortical adaptation that is involved in strength transfer might depend on how skilled or complex the strength task is. The contribution of spinal mechanisms also is important to consider, in conjunction with or independent from cortical mechanisms. The precise role of spinal mechanisms in cross-education of strength is unclear. Electrical stimulation training has been shown to result in a large magnitude of cross-education (104%), which suggests that spinal mechanisms are involved (14). However, Lagerquist et al. (20) showed that unilateral strength training was associated with increased spinal reflex excitability in the trained limb but not the untrained limb, despite showing significant strength increases in both limbs. The authors suggested that supraspinal mechanisms may be more involved in cross-education of strength.
A THEORETICAL MODEL OF CROSS-EDUCATION OF STRENGTH
A theoretical model of cross-education of strength in the upper limbs is proposed in Figure 4. The model represents an adaptation of previous models of skill transfer and applies it in the context of unilateral strength training while incorporating recent evidence from functional imaging and TMS studies that would suggest adaptations resulting from strength task acquisition (i.e., training) in one limb contribute to subsequent performance of the task by the untrained limb. The model is based on the central hypothesis that cross-education of strength is similar to cross-education of skills and is primarily controlled by cortical mechanisms. The model incorporates ideas from the proficiency model and the cross-activation model (24). The cross-activation model states that motor programs for a skill or task are stored in both hemispheres with unilateral acquisition. The proficiency model suggests that task acquisition with the more proficient system (i.e., the limb or side of brain that is dominant for the particular motor task) provides a better stored motor program for the opposite limb. The cross-education of strength model described here supports the hypothesis that training with the more proficient limb (e.g., dominant) would generate the greatest transfer to the untrained limb. The model is presented as a conceptual framework and does not describe the detailed processes of neural control. It describes three distinct levels where adaptation could contribute to the force output of the limb: 1) the motor planning level - which represents sources of input to the motor cortex to prepare for movement including premotor cortex, supplementary motor area, prefrontal cortex, temporal lobe, sensory and visual cortex, and input from periphery such as joint position; 2) the motor command level - which represents the output signals sent from the motor cortex to the periphery to coordinate the movement and includes pathways from motor cortex through the brainstem, spinal cord, and motor units; 3) the peripheral muscle level - which represents the physiological state of the muscles (i.e., primarily referring to muscle size) that are recruited to execute the strength task. The model incorporates the idea that if there is greater proficiency by the dominant limb system at the motor planning level, the motor command level, or at the peripheral muscle level, the amount of force generated during the strength movement would be greater (Fig. 4). In particular, because the dominant limb system is probably more proficient before training and becomes even more proficient with subsequent training, the task is more likely to be mastered by the dominant limb. Therefore, optimal information regarding the goals of the movement and precise patterns of muscle activation can be transferred to the nondominant limb to enhance motor planning and motor commands - resulting in cross-education of strength. In other words, the information acquired at these cortical levels is shared between limbs, and after the task has become mastered (more likely the case if the proficient limb acquires the task), motor control information that is optimal for forced production is stored. After training the dominant limb, the two limbs are virtually equivalent at the motor planning level. Although both limbs improve substantially at the motor command level, there is still an advantage for the execution of the task by the dominant limb because the motor pathways have been augmented through repetitive training over several weeks. In the context of this model, the amount of transfer asymmetry observed would be somewhat contingent upon how proficient the dominant limb is before training. A ceiling effect may be apparent that limits the magnitude of strength increase in the training limb (i.e., dominant).
With regard to the peripheral muscle level described in the model, if the primary agonist muscles of the dominant limb are larger before training or become larger as a result of training, this enhances force production for that limb. The model predicts marginal differences in muscle size between the limbs before training, but only the training limb experiences muscle hypertrophy that contributes to increased force production as a result of unilateral training. The trained limb increases force production through muscle hypertrophy and other possible muscle adaptations (i.e., metabolic, contractile properties) in addition to neural adaptation. The vast majority of studies agree that morphological changes are not observed in the untrained homologous muscles. Therefore, in theory, the untrained muscles experience only nervous system adaptation. This is an obvious difference compared with models of skill transfer where no morphological change is apparent in the trained muscles. This is important because in skill transfer studies when the performance improvement of each limb is the same after acquiring the task with one limb, transfer has occurred without morphological change to the muscles of either limb. In the context of unilateral strength training, when a similar or greater relative increase in strength is observed in the untrained limb compared with the trained limb (7,9,13,20), it could be argued that the relative contribution of the nervous system to performance improvement was greater for the untrained limb because the trained limb has experienced muscle hypertrophy that contributed to its strength increase, whereas the untrained limb did not (7,9). In situations where the dominant limb is considerably more proficient than the nondominant limb before training, subsequent training of the dominant limb could theoretically result in adaptations in the nervous system that lead to a greater relative strength increase in the untrained limb than the trained limb (9). However, a greater improvement in performance does not directly imply that the amount of motor learning was greater. The amount of learning that occurs for each limb is very difficult to quantify, especially when the performance measure is force output and each limb performs differently at baseline and is exposed to different stimuli that lead to improved force output.
The theoretical model can incorporate an important factor that seems to influence the magnitude of cross-education of strength - the characteristics of the strength task. More skilled or complex strength tasks involve coordination and activation of multiple muscle groups and theoretically involve greater motor learning adaptations at the motor planning and motor command level in response to strength training. In theory, strength training using more simple strength tasks would require less cortical adaptation and involve much less motor learning. However, if the strength task is very novel (i.e., uncommon movement in strength training programs and rarely encountered in activities of daily living), substantial adaptation would be predicted to occur at the motor planning and motor command level in the context of strength training, regardless of task complexity. In cases where the strength task is novel and more complex (e.g., isometric ulnar deviation), the dominant limb would be more proficient (i.e., better able to execute the motor plan and is stronger) at baseline and would be better at mastering the task. Subsequent training of the dominant limb would provide a better reference of motor information in cortex to be accessed when executing the task and this improves the pattern of muscle activation in the untrained limb (e.g., coordination of agonists and antagonists, postural control, synergist muscle activity). The highest magnitudes of cross-education of strength are predicted under these conditions.
Although the theoretical model is primarily constructed using evidence from upper limb studies, the general principles could be applied to cross-education in the lower limbs. However, no cross-education of strength studies have directly compared dominant versus nondominant leg training, and there are few lower limb skill transfer studies from which to compare. Strength asymmetry between legs might be less pronounced because there is more equal contribution of each leg to normal everyday activities. This would make it difficult to distinguish the more proficient limb, but as with the upper limbs, this is probably influenced by the characteristics of the task. More research is required to evaluate how well the theoretical model works for the lower limbs.
CROSS-EDUCATION AND UNILATERAL INJURY
There are obvious applications of the cross-education of effect in a rehabilitation setting. One such application is during recovery from unilateral injury (e.g., fractured arm) where prolonged periods of disuse are required during therapeutic intervention. Very few studies have examined contralateral strength exercise in a rehabilitation setting. Stromberg (32) examined the effect of contralateral strength training on postoperative recovery of function in the affected hand. The participants who completed a detailed home regimen of contralateral exercises showed much greater strength in the postoperative hand during recovery. However, the data could be misinterpreted because preoperative strength data are not shown, and postoperative hand strength at 1 wk postsurgery was higher in the exercise group and seemed to recover in a similar fashion for both groups throughout the 12-wk postsurgery follow-up. Because the baseline data are not presented, it is difficult to discern whether or not the contralateral exercise altered the time course of recovery of function or if the exercise group had better function than controls before surgery. Further to this, specific details about the home exercise program were not provided so the extent to which patients completed strength training is unclear. Nonetheless, the contralateral exercise group had better strength at all time points for the entire 12-wk follow-up.
We have recently applied the cross-education effect using a model of unilateral injury in healthy volunteers (10). We had two groups of right-handed participants wear a nondominant forearm/wrist cast for a 3-wk period. One group completed unilateral strength training of the free limb (right) during the immobilization period, whereas a second group did no training. A third group served as a control (no casting or training). The main finding was that strength was maintained at baseline levels (preimmobilization) in the immobilized limb of the unilateral training group, whereas the nontraining group showed a significant decline in strength (−14.7%) of the immobilized limb (Fig. 5). Unilateral strength training of the free limb seemed to prevent strength loss in the opposite immobilized limb through the cross-education effect. Although this study does not involve the context of a real injury, it does provide a convincing indication that cross-education could provide a benefit for the immobilized limb in this situation. In the case of a real injury, cross-education might be useful to reduce, but not eliminate, the strength loss and functional decrements encountered in the injured limb.
An important consideration of this work is that we used isometric ulnar deviation training of the dominant right hand. As previously discussed, this type of dominant hand training in right-handers elicits a high magnitude of cross-education. Whether training of the nondominant left hand would provide a benefit for a casted right hand remains unclear. The asymmetry of strength transfer we observed earlier (9) would suggest that the same benefit of contralateral training would not be observed for dominant arm immobilization. This is an important question for future research but is appreciably difficult because it would require volunteers to immobilize their dominant limb. The therapeutic use of the cross-education effect might be limited to scenarios where the nondominant limb is injured. However, as the theoretical model suggests (Fig. 4), the higher magnitude of transfer observed with dominant hand training is likely related to enhanced proficiency of the dominant limb. In both our dominant right-hand training studies, the dominant hand was stronger at baseline (7,9). In the context of unilateral injury, such as a broken limb, it could be argued that the more proficient limb becomes the uninjured limb regardless of handedness. In this case, left-arm training might still provide a benefit for an immobilized right limb, even though we have observed a preferential dominant-to-nondominant cross-education effect. The benefit for the immobilized limb in this situation might be reduced but not absent. Another important problem is the considerable lack of evidence in the literature for left-handed individuals. Behavioral studies of motor skills suggest that right and left-handers exhibit different patterns of transfer (24). Although left-handers might be expected to show similar benefits of contralateral training during a period of nondominant arm immobilization, we still have a general lack of understanding of the behavioral effects of cross-education in this population.
Another potential clinical application of the cross-education phenomenon is during stroke recovery. Strength training is well documented as an effective strategy to improve functional ability in stroke rehabilitation (11). However, strength training is often applied bilaterally or in the context of constraint-induced movement therapy (CIT). CIT is a strategy that involves the deliberate constraint of the less affected limb to force the use of the more affected limb. This strategy is known to be effective in restoring functional abilities after stroke, and it seems to involve the opposite strategy compared with cross-education. The potential use of cross-education of strength after stroke is not clear, but a recent animal study showed that training the less affected limb was maladaptive for the more affected limb in the early stages after a stroke (2). These results would suggest avoiding strength training of the less affected limb during the acute stages of recovery. Similar to how limb dominance influences the cross-education effect in healthy individuals, hand dominance and the side of the stroke have been shown to impact outcomes during stroke rehabilitation (21). The use of unilateral training poststroke might be dependent on whether the dominant or nondominant limb is more affected. McCombe Waller and Whitall (21) found that right-handed stroke patients with left, but not right, hemispheric lesions showed improvements in the more affected limb for several arm strength measures after a 6-wk bilateral arm strength training program. If there is a benefit of cross-education of strength in stroke recovery, it would likely be used in conjunction with existing effective strategies (e.g., bilateral strength training, CIT), introduced several months postincident and perhaps only in the context of a left hemisphere lesion.
The central hypothesis presented in this article is that cross-education of strength and cross-education of skills are closely related and involve similar cortical mechanisms. The hypothesis originated from recent evidence of asymmetry of cross-education of strength that is similar to some types of skill transfer, and it is supported by data from functional imaging and TMS studies. The behavioral nature of cross-education of strength is complex and might be more easily understood using a theoretical approach similar to the model presented in this review. The model suggests that unilateral training will lead to cortical adaptations that are shared by both limbs and that optimal information for producing strength is stored when the more proficient limb (e.g., dominant) is used for training. However, the model represents one perspective, and there is evidence in the literature to support the hypothesis that the mechanisms of cross-education reside at multiple sites within the nervous system (4). Importantly, depending on the context in which the cross-education effect is induced (e.g., complex vs simple strength task, upper vs lower body), the mechanisms contributing to the effect could be very different. As a final perspective, we presented evidence that suggests that cross-education provides a benefit for an opposite immobilized limb. Although this is important, it does not represent a real injury or rehabilitation setting, and the study was limited to right-handers who had their left arm immobilized. There is an apparent lack of information regarding the use of cross-education of strength in rehabilitation, and to explore it further, there is a need for more research into the behavioral changes elicited by unilateral strength training in left-handed populations and in the lower limbs.
The author thanks the peer reviewers of the manuscript who provided helpful comments that greatly improved this work. This research was supported by operating and equipment funds from the University of Saskatchewan and operating and equipment grants from the Saskatchewan Health Research Foundation.
1. Adamson M, Macquaide N, Helgerud J, Hoff J, Kemi OJ. Unilateral arm strength training improves contralateral peak force and rate of force development. Eur. J. Appl. Physiol.
2. Allred RP, Jones TA. Maladaptive effects of learning with the less-affected forelimb after focal cortical infarcts in rats. Exp. Neurol.
3. Cannon RJ, Cafarelli E. Neuromuscular adaptations to training. J. Appl. Physiol.
4. Carroll TJ, Herbert RD, Munn J, Lee M, Gandevia SC. Contralateral effects of unilateral strength training: Evidence and possible mechanisms. J. Appl. Physiol.
5. Davies J, Parker DF, Rutherford OM, Jones DA. Changes in strength and cross sectional area of the elbow flexors as a result of isometric strength training. Eur. J. Appl. Physiol. Occup. Physiol.
6. Ebersole KT, Housh TJ, Johnson GO, Perry SR, Bull AJ, Cramer JT. Mechanomyographic and electromyographic responses to unilateral isometric training. J. Strength Cond. Res.
7. Farthing JP, Borowsky R, Chilibeck PD, Binsted G, Sarty GE. Neuro-physiological adaptations associated with cross-education of strength. Brain Topogr.
8. Farthing JP, Chilibeck PD. The effects of eccentric training at different velocities on cross-education. Eur. J. Appl. Physiol.
9. Farthing JP, Chilibeck PD, Binsted G. Cross-education of arm muscular strength is unidirectional in right-handed individuals. Med. Sci. Sports Exerc.
10. Farthing JP, Krentz JR, Magnus CR. Strength training the free limb attenuates strength loss during unilateral immobilization. J. Appl. Physiol.
11. Gordon NF, Gulanick M, Costa F, et al. Physical activity and exercise recommendations for stroke survivors: An American Heart Association scientific statement from the Council on Clinical Cardiology, Subcommittee on Exercise, Cardiac Rehabilitation, and Prevention; the Council on Cardiovascular Nursing; the Council on Nutrition, Physical Activity, and Metabolism; and the Stroke Council; American Heart Association Council on Clinical Cardiology, Subcommittee on Exercise, Cardiac Rehabilitation, and Prevention; the Council on Cardiovascular Nursing; the Council on Nutrition, Physical Activity, and Metabolism; and the Stroke Council. Stroke.
12. Griffin L, Cafarelli E. Transcranial magnetic stimulation during resistance training of the tibialis anterior muscle. J. Electromyogr. Kinesiol.
13. Hortobágyi T, Lambert NJ, Hill JP. Greater cross education following training with muscle lengthening than shortening. Med. Sci. Sports Exerc.
14. Hortobágyi T, Scott K, Lambert J, Hamilton G, Tracy J. Cross-education of muscle strength is greater with stimulated than voluntary contractions. Motor Control
15. Hortobágyi T, Richardson SP, Lomarev M, et al. Chronic low-frequency rTMS of primary motor cortex diminishes exercise training-induced gains in maximal voluntary force in humans. J. Appl. Physiol.
16. Housh DJ, Housh TJ, Johnson GO, Chu W. Hypertrophic response to unilateral concentric isokinetic resistance training. J. Appl. Physiol.
17. Imamizu H, Shimojo S. The locus of visual-motor learning
at the task or manipulator level: Implications from intermanual transfer. J. Exp. Psychol. Hum. Percept. Perform.
18. Jensen JL, Marstrand PC, Nielsen JB. Motor skill training and strength training are associated with different plastic changes in the central nervous system. J. Appl. Physiol.
19. Karni A, Meyer G, Jezzard P, Adams MM, Turner R, Ungerleider LG. Functional MRI evidence for adult motor cortex plasticity during motor skill learning. Nature
20. Lagerquist O, Zehr EP, Docherty D. Increased spinal reflex excitability is not associated with neural plasticity underlying the cross-education effect. J. Appl. Physiol.
21. McCombe Waller S, Whitall J. Hand dominance and side of stroke affect rehabilitation in chronic stroke. Clin. Rehabil.
22. Muellbacher W, Ziemann U, Wissel J, et al. Early consolidation in human primary motor cortex. Nature
23. Munn J, Herbert RD, Hancock MJ, Gandevia SC. Training with unilateral resistance exercise increases contralateral strength. J. Appl. Physiol.
24. Parlow SE, Kinsbourne M. Asymmetrical transfer of training between hands: Implications for interhemispheric communication in normal brain. Brain Cogn.
25. Pascual-Leone A, Grafman J, Hallet M. Modulation of cortical motor output maps during development of implicit and explicit knowledge. Science
26. Patten C, Kamen G, Rowland DM. Adaptations in maximal motor unit discharge rate to strength training in young and older adults. Muscle Nerve
27. Perez MA, Cohen LG. Mechanisms underlying functional changes in the primary motor cortex ipsilateral to an active hand. J. Neurosci.
28. Perez MA, Tanaka S, Wise SP, et al. Neural substrates of intermanual transfer of a newly acquired motor skill. Curr. Biol.
29. Perez MA, Wise SP, Willingham DT, Cohen LG. Neurophysiological mechanisms involved in transfer of procedural knowledge. J. Neurosci.
30. Steenhuis RE. The relation between hand preference and hand performance: What you get depends on what you measure. Laterality
31. Stoddard J, Vaid J. Asymmetries in intermanual transfer of maze learning in right- and left-handed adults. Neuropsychologia
32. Stromberg BV. Contralateral therapy in upper extremity rehabilitation. Am. J. Phys. Med.
33. Wang J, Sainburg RL. Interlimb transfer of visuomotor rotations depends on handedness
. Exp. Brain Res.
34. Yue G, Cole KJ. Strength increases from the motor program: Comparison of training with maximal voluntary and imagined muscle contractions. J. Neurophysiol.
35. Zhou S. Chronic neural adaptations to unilateral exercise: Mechanisms of cross-education. Exerc. Sport Sci. Rev.
Keywords:©2009 The American College of Sports Medicine
unilateral training; strength transfer; skill transfer; motor learning; handedness; brain activation; fMRI