Flexibility of the tissues about a joint can influence the amplitude and economy of a movement. In sports, muscle stretching is common because of the inherent constraints of the muscle-tendon units and joint range of motion in executing technical movements that require extreme motion at some joints. In rehabilitation, stretching is used as a method to maintain or regain the range of motion of a joint that is necessary for movements performed during daily activity. Current opinion acknowledges that acute muscle stretching and long-term stretch training generate both mechanical adaptations of the muscle-tendon unit and neural adjustments in the spinal circuitry. The objective of this review is to examine the neural mechanisms that accompany acute muscle stretching and long-term stretch training.
ACUTE EFFECT OF MUSCLE STRETCHING
In isolated muscle from animals, it has been shown that slow or static stretching of the muscle-tendon unit induces acute changes in its passive length-tension relation (14). This effect is nonlinear during dynamic loading because of the contribution of the viscoelastic properties of the tissues to the progressive increase in the passive torque of the muscle-tendon unit. Structures such as muscle fascicles (structural proteins), tendons, aponeurosis, joint capsules, and ligaments contribute to limit the range of motion that can be achieved with passive muscle lengthening. In vivo, however, the amount of stretch that can be produced is attributable to muscle resistance caused by tonic reflex activity (1,11).
Effects of Static Stretching on Reflex Activity
Passive stretching refers to the technique of lengthening a muscle group by slowly moving a joint to its maximal range of motion and maintaining the position for a period (usually >10 s). Contrary to fast muscle lengthening, static stretching does not increase reflex activity of the stretched muscle, but instead reduces spinal reflex excitability. This inhibition can be tested in the soleus muscle by recording the Hoffmann reflex (H reflex) and the tendon reflex (T reflex) by means of electromyography. The H reflex is induced by the stimulation of the Ia fibers of the tibial nerve in the popliteal fossa (13), whereas percussion of the Achilles tendon with a reflex hammer is used to produce the T reflex. The H-reflex amplitude can be modulated at two levels: the motor neuron and the Ia synaptic transmission (presynaptic inhibition). The presynaptic Ia inhibitory pathways project presynaptically through one or more interneurons onto the Ia terminals. Hence, the amplitude of the H reflex can be modulated by presynaptic inhibition without any changes in motor neuron excitability (4). In contrast, the amplitude of the T reflex is influenced by both the adjustments that occur in the neural circuitry (motor neuron excitability and presynaptic inhibition) and by changes in the sensitivity of the muscle spindle. Therefore, in addition to slight differences in the monosynaptic and oligosynaptic contributions to the H and T reflexes, the comparison of the relative changes in the amplitudes of both reflex responses allows one to identify the contribution of the spindle to the observed changes during a stretch.
This comparison was performed in the soleus muscle during static stretching by progressively moving the ankle joint in increments of 5° (7). In such conditions, the amplitude of the H reflex decreased more than the amplitude of the T reflex (−31% of control value for the H reflex compared with −8% of control value for the T reflex). The amplitudes of the T and H reflexes remain depressed as long as the stretching maneuver was maintained (Fig. 1). Overall, the decline in both reflex responses indicates that the magnitude of the change in muscle length influences the magnitude of the reduction in spinal reflex excitability during muscle stretching (7). After the stretching maneuver, the amplitude of the H reflex returned immediately to its control value, whereas the amplitude of the T reflex remained below its control value. These results suggest that the reduction in the amplitude of the T reflex did not result from changes located in the neural circuitry but was likely caused by either a reduction in muscle spindle sensitivity or increased compliance of the muscle-tendon unit. The decrease in both T and H reflexes was not associated with significant changes in the maximal compound action potential of the muscle (Mmax) that was evoked by supramaximal electrical stimulation of the motor nerve. This stability of the Mmax response ensured that the stimulation conditions remained constant throughout the experiments.
Inhibitory Mechanisms During Static Stretching
The structure of the human muscle-tendon unit is characterized by thixotropic properties, which indicates that the stiffness and viscosity of the system depend on the preceding activity of the muscle (14). Some receptors, such as muscle spindles, also display thixotropic behavior and thus discharge differently according to the immediate history of length change and the type of conditioning contraction. For example, the stretch sensitivity of primary spindle endings is increased after the muscle has been held at long lengths compared with short lengths (6). Muscle stretching induces adjustments that can be of peripheral origin by exerting an action on the receptor (muscle spindle, Golgi tendon organ) or of central origin by influencing the neural circuitry. A decrease in the amplitude of the H reflex during passive stretching suggests that neural changes are induced by the stretch and contribute to muscle lengthening by reducing the afferent input to the motor neuron pool and therefore reducing tonic reflex activity. The decline in the H-reflex amplitude, however, may involve multiple neural mechanisms that are located at presynaptic and postsynaptic sites.
To study changes in excitability of the soleus motor neuron pool during muscle lengthening, Delwaide (4) recorded a spinal reflex response in a pathway devoid of presynaptic inhibitory inputs. A train of stimuli delivered to the posterior tibial nerve at the ankle induced a reflex response that originated mainly from cutaneous nerve fibers that project polysynaptically onto the same motor neuron pool (exteroceptive or E reflex; Fig. 2A). By comparing the H reflex and the E reflex, Delwaide (4) demonstrated that for small-amplitude passive stretches, induced by a thrust against the Achilles tendon, the size of the H reflex decreased without any change in the E reflex. A systematic investigation of the H and E reflexes during small-amplitude stretches of the ankle confirmed that presynaptic mechanisms contribute to the decrease in the amplitude of the H reflex (8). This conclusion was supported by the observation that neither the E reflex nor the motor evoked potentials (MEPs) induced by transcranial magnetic stimulation changed when the H reflexes were decreased (8) (Fig. 2). For large-amplitude stretches (>10° dorsiflexion), both reflexes (H and E) and MEP decreased similarly, suggesting that postsynaptic inhibitory mechanisms contribute to the observed changes. Again, this inhibition was transient, as both reflex and MEP responses returned to control values within a few seconds after the end of the stretching maneuver.
Thus, the decrease in reflex loop activity during small-amplitude stretching (<10° dorsiflexion) is likely attributable to presynaptic inhibition of the Ia afferents without changes in motor neuron excitability. It is also possible that homosynaptic depression (a decrease in synaptic activity because of a reduced capacity for synaptic transmission from the Ia afferents to the motor neurons) may contribute to the presynaptic inhibition (2). However, because homosynaptic depression lasts approximately 12 s, and the recording of the two reflexes began at least 20 to 30 s after the onset of the passive stretching, this mechanism likely did not play a major role in the reduction in the amplitude of the H reflex. Furthermore, when subjects exerted a maximal voluntary contraction of the triceps surae muscle during a stretching protocol, the size of the H reflex recovered completely, suggesting that the active inhibitory mechanism was under the control of the nervous system.
The decrease in MEP size during large-amplitude stretching (>10° ankle dorsiflexion) could be due to reduced excitability of both the cortical neurons and α motor neurons. The observation of a concurrent and similar decrease in MEP and E reflex responses during large-amplitude stretching suggests a reduced excitability of the motor neuron pool through spinal postsynaptic inhibitory mechanisms. Different spinal inhibitory mechanisms, such as afferent input from the Golgi tendon organs and Renshaw cells, can reduce motor neuron excitability during muscle stretching. The afferents from joint and cutaneous receptors do not seem to play a major role in the inhibition of the spinal reflex excitability.
The Golgi tendon organs primarily respond to the force of a contraction and are less sensitive to the mechanical tension of passive stretching. They appear to be activated only during large-amplitude stretching. In humans, a pure Ib inhibition of soleus motor neurons can be elicited by the electrical stimulation of group I fibers in nerve to the medial gastrocnemius muscle, as it is not contaminated by any Ia excitation. The resulting heteronymous Ib inhibition of the soleus motor neuron pool lasts less than 10 ms. The conditioning stimulation of the nerve to medial gastrocnemius reduces the amplitude of the soleus H reflex because of the heteronymous inhibition by the excitation of the Ib fibers (12). The conditioned H reflex is decreased by 32.3% when the ankle is in a neutral position (0° of dorsiflexion). At 10° and 20° of dorsiflexion, however, the conditioning stimulus was less effective, and the decline in the amplitude of the H reflex corresponded to 9 and 3.4%, respectively (Fig. 3). These results suggest that by increasing the level of dorsiflexion, postsynaptic inhibition onto the motor neurons pool is progressively augmented, and thereby, the contribution of heteronymous inhibitory afferents to the motor neuron pool of the soleus muscle is progressively weakened. In addition to the contribution of the Golgi tendon organs to postsynaptic inhibition of the motor neuron pool during large-amplitude stretches, recurrent inhibition via the Renshaw loop may also contribute to the reduction in the amplitude of the H reflex.
In conclusion, the reduction in the excitation of motor neurons during muscle stretching is caused by mechanisms located at both presynaptic and postsynaptic sites. Presynaptic mechanisms dominate during small-amplitude stretches, whereas postsynaptic mechanisms contribute more to reflex inhibition during large-amplitude stretches.
Inhibitory Mechanisms During Different Methods of Stretching
Muscle stretching performed at slow velocities appears to be more effective than rapid and repetitive movements (11) because the former avoids reflex responses from the stretched muscles. Some years ago, proprioceptive neuromuscular facilitation (PNF) techniques were designed to inhibit reflex activities and thereby reduce passive muscle resistance during stretching maneuvers. Since that time, a number of studies have shown that PNF techniques increase the range of motion of a joint more effectively than passive stretching (7). The PNF techniques often involve static stretching used in combined procedures: either preceded by a maximal isometric contraction of the muscle (contract-relax, CR) or accompanied by a maximal isometric contraction of the antagonistic muscle (antagonist-contract, AC).
When static stretching is preceded by a maximal voluntary contraction of the muscle to be stretched, the length of the muscle usually increases more than when the stretch is not preceded by a maximal contraction (7). In addition to a possible reduction in the compliance of the elastic components following the isometric contraction, neural changes also contribute to the increased range of motion. The duration of the preceding contraction is not critical as the amplitude of the H reflex decreases by a similar proportion immediately after maximal contractions of the triceps surae ranging from 1 to 30 s (7). Furthermore, the reduction in H-reflex amplitude is not influenced by the magnitude of the contraction that precedes the stretching maneuver. Indeed, Enoka et al. (5) observed a similar reduction in H-reflex amplitude for voluntary contractions performed at 50 and 100% of the subjects' maximal capacity. The decrease in excitability of the spinal reflex recorded after the isometric contraction is brief (<5 s) but appears to be long enough to provide an advantage for a subsequent stretching maneuver. The mechanisms responsible for this transient hypoexcitability are probably located at the presynaptic level (6). In addition, possible reduction in muscle stiffness immediately after the maximal voluntary contraction, the advantage of the CR method over the static stretching method, results from the benefit of postcontraction inhibition of the motor neuron pool.
When static stretching is associated with a contraction of the antagonistic muscle, the mechanisms related to reciprocal inhibition can be added to those considered previously. Although opposite results have been obtained for the hamstrings muscles (11), the AC method increases the inhibition of the motor neurons that belong to the muscle being stretched and allows a greater muscular extensibility (3). For both stretching techniques (CR and AC), the H reflex returns immediately to the initial level of excitability regardless of the duration of the stretch.
LONG-TERM EFFECT OF STRETCH TRAINING
Few studies have investigated the mechanisms of adaptation to chronic muscle stretching. In the past, studies mainly focused on the effect of stretch training on the range of motion or passive torque (15). The reduction in passive torque recorded at a given muscle length after a program of stretch training has been attributed to changes in the compliance of the tendon, the connective tissues, and myofibrils. In addition, the increased joint range of motion after training has been attributed to an increase in the tolerance to stretch (10).
To determine the contribution of neural and mechanical mechanisms to the limitations in the range of motion about a joint, we examined the effects of a 6-week training program of static stretch training (10 min/d) of the triceps surae (9). The training program produced an increase in the ankle dorsiflexion range of motion of 30.8%, with 56% of the gain attained after only 10 sessions. Thirty days after the end of the training program, 74% of the gain in ankle dorsiflexion remained. The training program progressively shifted the torque-angle relation toward a reduced passive stiffness of the muscle-tendon unit (Fig. 4). For all subjects, there was a significant positive relationship between the gain in ankle dorsiflexion and the reduction in passive stiffness (r2 = 0.88; P< 0.001). These results suggest that an increase in the compliance of the muscle-tendon unit in a relaxed state after training may be attributable to either a change in viscoelastic properties or a decrease in reflex stiffness.
As observed during acute stretching, tonic reflex activity appears to decline after a stretch-training program, potentially contributing to the observed decrease in passive torque. Consistent with this hypothesis, the amplitude of the H and T reflexes declined significantly during the course of the training program. However, the T-reflex amplitude decreased to a greater extent (36 vs 14%) and more rapidly than the H reflex, reaching significant values after 20 training sessions. The greatest decline in the H reflex was only observed after 30 training sessions. The reduction in the amplitude of the H reflex suggests that motor neuron excitability was decreased or that synaptic transmission from Ia afferents to the motor neuron pool was reduced after stretch training.
An alternative possibility is that there was a reduction in tonic Ia (facilitatory) afferent feedback from muscle spindles as the maximal amplitude of the T reflex (Tmax) also declined during the training program. This reduction in T-reflex amplitude after training could be related to a reduced sensitivity of the muscle spindles or an increased compliance of the passive elastic components of the muscle-tendon unit. In the latter case, the tendon tap would have been less effectively transmitted to the muscle spindles. However, our data show that the decline of the amplitude of the T reflex did not parallel the reduction in passive stiffness during the stretch-training program. Furthermore, the decrease in the Tmax/Hmax ratio (Fig. 5), which measures changes in the sensitivity of the muscle spindle and tendon stiffness, is not correlated (P > 0.05) with changes in passive stiffness after 10, 20, or 30 training sessions (9). These results indicate that the decrease in passive stiffness of the muscle-tendon unit cannot entirely account for the decrease in T-reflex (Tmax) amplitude and that spinal reflex excitability is intrinsically reduced after stretch training. This conclusion is also supported by the observation that Tmax recovered to its pretraining values before passive stiffness. Collectively, these results suggest that the neural input to the motor neuron pool is reduced after 30 sessions of stretching and partly contributes to the gain in flexibility.
SUMMARY AND FUTURE DIRECTION
Both mechanical and neural factors influence the response to stretching. The increased flexibility during a prolonged stretch results mainly from reduced passive stiffness of the muscle-tendon unit. Because passive resistance is attributable to both intrinsic stiffness and neural mechanisms (15), changes at both levels can increase the flexibility around a joint. Proprioceptive neuromuscular facilitation techniques can increase flexibility, but the underlying neural mechanisms and efficacy with long-term use are unknown. The initial changes that are produced by stretch training involve mechanical adaptations that are followed by neural adaptations, which contrasts with the sequence observed during strength training. The timing of these adaptations and their relative contributions to the gains in range of motion about different joints remain largely unexplored.
The authors are particularly grateful to Dr. Carol Mottram for useful comments on the paper and to A. Desseir for assistance in the preparation of the manuscript. This study was supported by the Université Libre de Bruxelles and the Fonds National de la Recherche Scientifique of Belgium.
1. Avela, J., T. Finni, T. Likavainio, E. Niemelä, and P. Komi. Neural and mechanical responses of the triceps surae muscle group after 1 h of repeated fast passive stretches. J. Appl. Physiol.
2. Condon, S.M., and R.S. Hutton. Soleus muscle electromyographic activity and ankle dorsiflexion range of motion during four stretching procedures. Phys. Ther.
3. Crone, C., and J. Nielsen. Methodological implications of the post activation depression of the soleus H-reflex in man. Exp. Brain Res.
4. Delwaide, P.J. Human monosynaptic reflexes and presynaptic inhibition. In: New Developments in Electromyography and Clinical Neurophysiology
, edited by J.E. Desmedt. Basel: Karger, 1973, 3, pp. 508-522.
5. Enoka, R.M., R.S. Hutton, and E. Eldred. Changes in excitability of tendon tap and Hoffmann reflexes following voluntary contractions. Electroencephalogr. Clin. Neurphysiol.
6. Gregory, J.E., R.F. Mark, D.L. Morgan, A. Patak, B. Polus, and U. Proske. Effects of muscle history on the stretch reflex in cat and man. J. Physiol.
7. Guissard, N., J. Duchateau, and K. Hainaut. Muscle stretching and motoneuron excitability. Eur. J. Appl. Physiol.
8. Guissard, N., J. Duchateau, and K. Hainaut. Mechanisms of decreased motoneurone excitation during passive muscle stretching. Exp. Brain Res.
9. Guissard, N., and J. Duchateau. Effect of static stretch training on neural and mechanical properties of the human plantar-flexor muscles. Muscle Nerve
10. Magnusson, S.P., E.B. Simonsen, P. Aagard, H. Sorensen, and M. Kjaer. A mechanism for altered flexibility in human skeletal muscle. J. Physiol.
11. Moore, M.A., and R.S. Hutton. Electromyographic investigation of muscle stretching techniques. Med. Sci. Sports Exerc.
12. Pierrot-Deseilligny, E., R. Katz, and C. Morin. Evidence for Ib inhibition in human subjects. Exp. Brain Res.
13. Schieppati, M. The Hoffmann reflex: a means of assessing spinal reflex excitability and its descending control in man. Prog. Neurobiol.
14. Taylor, D.C., J. Dalton, A.V. Seaber, and W.E. Garrett. The viscoelastic properties of muscle-tendon units. Am. J. Sports Med.
15. Toft, E., T. Geert, M.D. Espersen, M. Soren Kalund, T. Sinkjaer, and B.Horemann. Passive tension of the ankle before and after stretching. Am. J. Sports Med.