Sensory input is essential for the optimal performance of motor tasks. Feedback from proprioceptors in muscle, tendon, ligaments, and joint capsules provides the nervous system with information that is used by spinal and supraspinal centers to refine motor performance. In addition to the ascending component of afferent input that gives rise to conscious perception of movement, many afferent inputs also make local synapses at the level of the spinal cord. Spinal circuitry is comprised of a complex system of neural networks and reflexes that regulate motor output via their influence on spinal motoneurons. The spinal circuits are, in turn, regulated by supraspinal centers. When the mechanisms of reflex regulation are disrupted, as they are in the case of upper motor neuron pathology (e.g., stroke, spinal cord injury, traumatic brain injury), the spinal reflexes become disorganized, and motor disorders such as spasticity, clonus, and cocontraction emerge (see (11) for review). While reflexes have traditionally been thought to be immutable and “hard-wired,” evidence gathered over the last two decades suggests this is far from the case. It is now clear that reflexes are naturally modulated on a short-term basis as movements are performed, and demonstrate plasticity over the long term based on the manner in which they are used. This article will review how reflex modulation and plasticity occur as results of the sensory and cortical input accompanying movement and practice. A review of the literature will explore how electrical stimulation might be used to simulate this input, thereby providing a tool to enhance motor performance. Examples from my work will be used to illustrate how electrical stimulation affects spinal reflexes in able-bodied individuals, and how it may be used as part of rehabilitation for individuals with neurological deficits.
MOTOR ACTIVITY MODIFIES SPINAL REFLEXES
Short-Term Modulation of Reflex Gain
The traditional textbook conceptualization of spinal reflexes is that they are formed by hard-wired neural circuitry. But far from being immutable, reflexes are highly modulated during movement. Short-term, “phase-dependent” reflex modulation is the change in strength (i.e., the degree of amplification, or “gain”) of a reflex over the course of a behavior. A case in point is the stretch reflex. This reflex, mediated largely by the Ia afferents from muscle spindles, is a monosynaptic reflex that results in excitation of the homonymous muscle when the muscle is subjected to a rapid stretch. In able-bodied individuals, the gain of the soleus H-reflex (the experimental equivalent of the stretch reflex) is high during the stance phase of the gait cycle, but low during the swing phase (9). This phase-dependent modulation is thought to be functionally important to allow the body to move over the foot during walking, a movement that would be hindered if the biomechanically-imposed stretch on the soleus in terminal stance resulted in soleus contraction and ankle plantarflexion. While phase-dependent modulation merely alters the gain of the reflex in this example, in other cases there may actually be phase-dependent reversal of the reflex. That is, the sign of the reflex (e.g., excitatory) may be reversed to the opposite sign (e.g., inhibitory), depending on the phase of the movement in which the stimulus occurs.
Forssberg et al. (8) were among the first to investigate the reversal in sign of a spinal reflex. Cats with complete transection of the thoracic spinal cord (i.e., spinal cats) are able to perform hindlimb walking on a treadmill. When a pulse of electrical stimulation was applied to the dorsum of the foot during the swing phase of the step cycle, relative to a nonstimulated step cycle (Fig. 1, A1; nonstimulated cycle), the ipsilateral limb was lifted higher (Fig. 1, A2; stimulation onset indicated by “s”) as a result of increased knee and ankle flexion. However, the same stimulus pulse applied during the stance phase of the ipsilateral limb (Fig. 1, B1; nonstimulated cycle) results in increased knee and ankle extension (Fig. 1, B2; stimulation onset indicated by “s”) with accompanying changes in associated muscle activity.
The phase-dependent modulation of reflex activity is critical to the optimal performance of motor behaviors. This modulation may be impaired or lost following upper motor neuron damage. In individuals with spastic paresis due to spinal cord injury, for example, the gain of the soleus H-reflex may be high throughout the gait cycle (9). This loss of regulation is associated with excessive ankle plantarflexion in terminal stance and early swing, contributing to the characteristic spastic gait pattern often observed in individuals with stroke. Thus, short-term, phase-dependent modulation of reflex gain contributes to the optimal performance of movement.
Long-Term Plasticity of Reflex Gain
Long-term reflex plasticity, an enduring change in the strength of a reflex circuit, results from the manner in which the spinal circuits are used over a prolonged period of time. In an elegant series of experiments, Wolpaw and colleagues (15) demonstrated that the spinal stretch reflex (SSR) and its electrical analog the H-reflex can be operantly conditioned such that the level of excitation it provides to the agonist alpha motoneuron is either increased or decreased. The change in spinal reflex amplitude (either increase or decrease) in monkeys over the course of 80 d of training is illustrated in Figure 2. The upper half of the graph represents the spinal stretch reflex amplitude over the course of training in those animals that were rewarded only for increased reflex amplitude (uptrained), whereas the lower half represents animals that were rewarded only when the reflex amplitude decreased (downtrained). While the spinal circuits must have access to supraspinal centers for operant conditioning to occur, the neural changes that result in the excitability change are located at the level of the spinal cord (15) as it persists even after the removal of descending input (i.e., following spinal transection).
There is an abundance of research indicating that the type of activity one is involved in has a lasting effect on the spinal reflex circuitry. This effect is such that excitability of spinal reflex circuits is modified by the type and intensity of activity the individual regularly engages in. Individuals trained in endurance activities (e.g., triathletes, long-distance swimmers, or cross country skiers) have higher H-reflex activity compared to nontrained individuals. Conversely, power-trained athletes (high- or long-jumpers, sprinters, basketball players) have lower H-reflex activity compared to nontrained individuals (12). Although some authors (12) have suggested the type of training alters the proportion of motor neurons activated by the Ia afferent stimulation, these long-term changes in reflex activity are probably due to training-related alterations in the strength (or gain) of the reflex circuits. Training is therefore associated with an alteration in the level of excitability of the reflex (via pre- or postsynaptic effects at the level of the spinal cord). This means that training-related sensory input, supraspinal input, and perhaps the interaction between the sensory and supraspinal inputs, somehow influence the spinal circuitry, either enhancing or reducing spinal reflex activity. Presumably this alteration in reflex function serves an adaptive purpose that benefits or supports the customary activity the individual is involved in.
The converse may also be true, wherein the alteration to the sensory or supraspinal input to the reflex may be maladaptive, and lead to impaired motor function. This is the case in individuals with upper motor neuron pathology. While the most obvious deficit resulting from the disruption or loss of supraspinal input in these individuals is motor paresis or paralysis, the loss of the modulatory influence of the supraspinal centers on spinal reflex circuitry is also disrupted, resulting in relatively permanent changes in spinal motor output. Motor dysfunctions such as hyperflexia, clonus, and flexor or extensor spasms arise from the loss of descending inhibition and disordered regulation of the spinal reflex activity. This is consistent with the idea that the spinal cord circuitry is modified in a use-dependent manner, with the alteration in input to the spinal cord resulting in a maladaptive reorganization of the spinal reflex circuitry. Maladaptive changes to spinal reflex circuitry are detrimental to motor function and result in disordered motor control.
ELECTRICAL STIMULATION OF SENSORY FIBERS MODIFIES NEURAL CIRCUITRY
The Use of Sensory Input to Modify Spinal Reflex Circuitry
Given that sensory input modulates spinal reflex output during movement, it follows that it may be possible to modulate spinal (and supraspinal) circuitry by artificially activating the same sensory afferents that are activated during movement. This is accomplished by artificial stimulation (via electrical stimulation or mechanical stimulation (e.g., vibration)) of the same sensory fibers that convey sensory information to the spinal cord during movement (Fig. 3). While the use of electrical stimulation to modulate stretch reflex activity in able-bodied individuals has shown mixed results (see (10) for review), there is evidence that it may be effective in reducing stretch reflex excitability in individuals with upper motor neuron injury. Because the loss of regulation of the stretch reflex is one of the primary contributors to spasticity, it is not surprising that the application of electrical stimulation is associated with a reduction in clinical measures of spasticity (10). Furthermore, it appears that some forms of sensory input may be effective in restoring the normal phase-dependent modulation of the stretch reflex. Fung and Barbeau (9) investigated, in individuals with spastic paresis due to incomplete spinal cord injury, the effect of repeatedly activating the ipsilateral flexion withdrawal reflex on phase-dependent modulation of the soleus H-reflex modulation. They demonstrated that activation of the flexion withdrawal reflex (timed to assist with stepping) resulted in more normal modulation of the H-reflex over the gait cycle.
In addition to the aforementioned stretch reflex hyperexcitability, individuals with upper motor neuron damage often experience a loss of agonist-antagonist reciprocal inhibition. The reciprocal inhibition circuit mediates relaxation of the antagonist muscle when the agonist is activated. This circuit is functionally important for the modulation of reciprocal inhibition and thereby for the expression or suppression of cocontraction about a joint. Carnstam et al. (3) assessed, in individuals with spastic paresis due to multiple sclerosis, the effect of electrical stimulation to the nerve innervating the antagonist muscle on reflex-generated muscle torque. They found that, following 10 min of common peroneal nerve stimulation, the torque produced by the plantarflexors to an Achilles tendon tap was reduced. In some individuals the stimulation was also effective in decreasing the torque produced by quadriceps in response to a patellar tendon tap. Crone et al. (6) investigated the strength of agonist-antagonist reciprocal inhibition in individuals with spastic paresis due to multiple sclerosis. They demonstrated that in these individuals the reciprocal inhibition is typically less effective than in able-bodied individuals. However, four individuals who routinely used common peroneal stimulators as a neural prosthesis to assist with dorsiflexion during walking showed normal levels of reciprocal inhibition. We have used the flexion withdrawal reflex extensively in our body-weight–supported gait training studies in individuals with spinal cord injury. We have shown this form of training to be associated with improvements in functional walking ability, muscle strength, and coordination (7). Given that the flexor withdrawal reflex is thought to be a component of the central pattern generator for walking (2), it is logical to try to incorporate its use into locomotor training programs.
Taken together this research clearly suggests that electrical stimulation is associated with an adaptive reorganization of the spinal neural circuitry. The reorganization results in more normal reflex activity, which may form the foundation for the production of more functional movement. In order to gain a better understanding of the relationship between sensory input and enduring changes to spinal reflex activity, we (13) recently investigated the effects of prolonged electrical stimulation on the reciprocal inhibitory circuit. We examined the effect of the pattern of antagonist stimulation (common peroneal nerve to tibialis anterior muscle) and the relative contribution of supraspinal input versus sensory input in modulating reciprocal inhibition of the agonist (soleus) H-reflex amplitude. The stimulation protocol involved three different conditions:
- 1) Patterned nerve stimulation of the common peroneal nerve using bursts of stimulation. This condition simulates the rate and timing of sensory input that would accompany walking;
- 2) Combined stimulation, which used the patterned stimulation together with stimulation of the leg area of the motor cortex (using transcranial magnetic stimulation). This condition was designed to simulate concurrent sensory and cortical input to spinal centers;
- 3) Uniform stimulation, in which the common peroneal nerve was stimulated using the same number of pulses and the same intensity as in the patterned protocol, but with the pulses uniformly spaced. This condition was designed to provide uniform excitatory input.
The strength of reciprocal inhibition from the tibialis anterior to the soleus was determined before and after application of the stimulation protocol. We found that prolonged patterned stimulation was effective in increasing the ability of the tibialis anterior to inhibit the soleus via the Ia-mediated reciprocal inhibition pathway. The patterned stimulation protocol resulted in increased efficacy of Ia inhibition (relative to baseline) and this effect persisted for 5 min following termination of the stimulation period. This was true both for the patterned stimulation-only condition (Fig. 4 (triangles)) and for the patterned stimulation plus supraspinal stimulation condition (Fig. 4 (squares)); the addition of the supraspinal stimulation to the patterned stimulation condition did not add any additional effect in the pathways investigated in this study. The uniform stimulation condition had no effect on the strength of Ia-mediated reciprocal inhibition (Fig. 4 (circles)). The fact that the effects of the stimulation protocol persist after the cessation of stimulation reflects at least a short-term change in neural elements mediating the efficacy of this circuitry. These studies suggest that activation of sensory afferents via electrical stimulation can result in modulation of spinal reflex circuits, both in able-bodied individuals and in individuals with upper motor neuron pathology.
The Use of Sensory Input to Modify Cortical Circuitry
While the studies highlighted to this point have all assessed changes to spinal reflexes that occur with stimulation, there is also evidence that stimulation can affect the excitability of cortical circuits. Charlton et al. (4) demonstrated that prolonged facilitation of cortically-evoked motor potentials could be induced by various protocols of peripheral nerve stimulation. That is, following peripheral nerve stimulation, the excitability of the motor cortex is increased such that subsequent stimulation of the motor cortex results in a larger amplitude of motor response than that which was elicited before the application of peripheral nerve stimulation. This suggests that sensory stimulation can induce enduring effects on cortical excitability that may have implications for motor function. Conforto et al. (5) showed that prolonged electrical stimulation can provide somatosensory input that influences motor performance in individuals with stroke. They studied the effects of two different stimulation protocols on motor function in individuals with hemiparesis from ischemic stroke. One protocol involved stimulation intensities strong enough to elicit parasthesia in the median nerve distribution; the other was set at an intensity below that required to elicit parasthesia. Following the application of the higher-intensity protocol, subjects demonstrated increased pinch strength and reported improved function in tasks such as writing.
In individuals with paresis due to incomplete spinal cord injury, some fraction of the corticospinal pathways may be preserved. However, in addition to the disruption of the spinal pathways (both ascending and descending) and the maladapative reorganization of the spinal reflex circuitry, there is also maladaptive reorganization of cortical circuitry much like that seen after stroke (14). This maladaptive reorganization may result in ineffective activation of the remaining descending pathways, thereby contributing to the motor paresis experienced by these individuals. Studies based on this hypothesis are currently underway (1) to investigate the effect of sensory-level electrical stimulation combined with activity-based training on upper-extremity function and cortical excitability in individuals with incomplete cervical spinal cord injury. Preliminary results suggest that electrical stimulation enhances the beneficial effects on grip strength, arm function and cortical excitability that are observed with activity-based training alone. In sum, evidence suggests that in both able-bodied individuals and those with upper motor neuron pathology, electrical stimulation to the peripheral nervous system evokes plastic changes in cortical circuitry that may be beneficial for functional motor performance.
Improvements in motor performance hinge on the plasticity of the motor control system. In both the short and long term, the neural elements underlying motor performance are highly influenced by the sensory input associated with activity. Electrical stimulation provides a valuable tool to influence these neural elements at both the cortical and spinal levels.
1. Beekhuizen, K.S., and E.C. Field-Fote. Effects of massed practice and somatosensory stimulation in individuals with spinal cord injury. Proceedings of the 2004 Combined Sections Meeting of the American Physical Therapy Association,
2. Bussel, B., A. Roby-Brami, O.R. Neris, and A. Yakovleff. Evidence for a spinal stepping generator in man. Electrophysiological study. Acta. Neurobiol. Exp. (Wars).
3. Carnstam, B., L.E. Larsson, and T.S. Prevec. Improvement of gait following functional electrical stimulation. I. Investigations of changes in voluntary strength and proprioceptive reflexes. Scan. J. Rehabil. Med.
4. Charlton, C.S., M.C. Ridding, P.D. Thompson, and T.S. Miles. Prolonged peripheral nerve stimulation induces persistent changes in excitability of human motor cortex. J. Neurol. Sci.
5. Conforto, A.B., A. Kaelin-Lang, and L.G. Cohen. Increase in hand muscle strength of stroke patients after somatosensory stimulation. Ann. Neurol.
6. Crone, C. J., Nielsen, N. Petersen, M. Ballegaard, and H. Hultborn. Disynaptic reciprocal inhibition of ankle extensors in spastic patients. Brain
7. Field-Fote, E.C., and D. Tepavac: Improved intra-limb coordination in individuals with incomplete spinal cord injury following training with body weight support and functional electrical stimulation. Phys. Ther.
8. Forssberg, H., S. Grillner, and S. Rossignol. Phasic gain control of reflexes from the dorsum of the paw during spinal locomotion. Brain Res.
9. Fung, J., and H. Barbeau. Effects of conditioning cutaneomuscular stimulation on the soleus H-reflex
in normal and spastic paretic subjects during walking and standing. J. Neurophysiol.
10. Goulet, C., A.B. Arsenault, D. Bourbonnais, M.T. Laramee, and Y. Lepage. Effects of transcutaneous electrical nerve stimulation on H-reflex
and spinal spasticity. Scand. J. Rehabil. Med.
11. Hultborn, H. Changes in neuronal properties and spinal reflexes during development of spasticity following spinal cord lesions and stroke: studies in animal models and patients. J. Rehabil. Med.
41 (Suppl):46–55, 2003.
12. Maffiuletti, N.A., A. Martin, N. Babault, M. Pensini, B. Lucas, and M. Schieppati. Electrical and mechanical H (max)-to-M (max) ratio in power- and endurance-trained athletes. J. Appl. Physiol.
13. Perez, M.A., E.C. Field-Fote, and M.K. Floeter. Patterned sensory stimulation induces plasticity
in reciprocal Ia inhibition in humans. J. Neurosci
. 23:2014–2018, 2003.
14. Raineteau, O., and M.E. Schwab. Plasticity
of motor systems after incomplete spinal cord injury. Nature Reviews Neuroscience.
15. Wolpaw, J.R., and A.M. Tennissen. Activity-dependent
spinal cord plasticity
in health and disease. Annu. Rev. Neurosci.