To make movements and perform locomotion, humans and other animals exert torques about joints by transmitting forces from muscles onto bones. If the muscle fibers only had to transmit forces via their tendons to the bones of the human skeleton, as is often assumed implicitly, understanding (myotendinous) force transmission would be relatively simple. However, muscle fibers must also exert forces at joints, for example to ensure joint stability, which makes the analysis of myotendinous force transmission much more complex. The purpose of this review is to show that several additional paths are involved in force transmission and that as a consequence, muscles do not act as independent actuators. Therefore, to understand the function of muscles, muscle groups, synergists, antagonists, and limbs, an integrative approach is required.
MYOTENDINOUS FORCE TRANSMISSION
At least at one end, most muscle fibers have a specialized myotendinous junction (13) that connects them to tendinous collagen fibers (Fig. 1). At apical ends of muscle fibers, collagen fibers of the aponeurosis enter and attach to invaginations of the sarcolemma-basal lamina complex. This complex is attached to the thin filaments of the final sarcomeres of the muscle fiber (Fig. 1, inset) via the transsarcolemmal molecules and the cytoskeleton. Note that the endomysium is not involved in this interaction at this location. In the active muscle fiber, these connections will experience shear forces and stiffen, which allows force transmission to the collagen fibrils of the tendon. In most conditions, a significant, but unknown, fraction of the force exerted by a fiber is likely to be transmitted via the tendon to bone through this path.
MYOFASCIAL FORCE TRANSMISSION
In addition to myotendinous transmission, however, force is also transmitted between sarcomeres and the endomysium. The endomysium is part of the extracellular matrix that surrounds each muscle fiber and is reinforced with collagen fibers (14). Each muscle fiber operates within such a tube of endomysium. When a force that counteracts shortening of the active or passive sarcomere does not originate from sarcomeres in series, but (at least in part) from the extracellular matrix, force is transmitted along this myofascial path.
The elegant work of Street (1962, 1983) yielded the first evidence for such myofascial force transmission. This work has been reviewed recently by several authors (2) and will not be treated any further here. Although the feasibility of force transmission between muscle fibers and their endomysia was suggested previously (8), it was based on the erroneous hypothesis that the longitudinal collagen fibrils of the tendon “change into” the intramuscular connective tissue of muscle. The continuity of the tendinous structures and intramuscular connective tissue, however, is based on the association between the endomysium and perimysium in muscle to similar sheets in the tendon, that is, endotenon (Figs. 1 and 2) and peritenon, respectively.
Intramuscular Myofascial Force Transmission
As soon as force is exerted onto an endomysium, there are three potential paths: (a) tensile transmission in longitudinal direction—because the endomysium is continuous with a similar network surrounding groups of tendinous collagen fibers (Figs. 1 and 2), force can be transmitted onto the intramuscular aponeurosis or tendon (a type of fasciotendinous transmission); (b) transmission in cross-fiber directions onto adjacent endomysial tunnels by shearing of the connective tissue of the fascicle, that is, force can be transmitted between the endomysia of neighboring muscle fibers of a fascicle; (c) transmission in cross-fiber directions onto active or passive sarcomeres within adjacent muscle fibers by shearing between the endomysium and basal lamina of neighboring muscle fibers.
As a result of these pathways, force can be transmitted either onto the tendon (fasciotendinous transmission) or onto neighboring fascicles (2,6) and then to the epimysium surrounding the muscle. If a muscle is dissected fully from other structures, intramuscular myofascial force transmission onto the intramuscular aponeurosis is the only path, in addition to direct myotendinous force transmission. In vivo, the full stroma of intramuscular connective tissues of a muscle (including the epimysium) are involved in the transmission of active and passive forces exerted by all the fibers in the muscle. This property allows groups of muscle fibers that are no longer connected to the tendon at one end to contribute to muscle force (6).
Note that such myofascial force transmission occurs at lower mean sarcomere lengths, compared with those fibers that connect directly to the tendon, because fiber shortening occurs until sufficient shear strain has accumulated in the basal lamina and endomysium to stretch and stiffen these structures enough to be able to transmit the additional force (2). This enhances the distribution of sarcomere lengths within the muscle, which can result in substantial functional consequences: (a) optimum length of the muscle is increased; (b) force at optimum length is decreased, but at lower lengths force is increased; (c) the length at which muscle exerts an active force (active slack length) approaches zero is decreased; d) and therefore, the range of active force exertion is enhanced, which increases the active range of joint motion.
Such acute changes also describe the mechanism that is manipulated as a consequence of some clinical interventions to relieve various muscular abnormalities. For example, cutting the aponeurosis of a muscle perpendicular to its longitudinal axis, a procedure known as aponeurotomy, is intended to increase the range of motion about a joint in spastic patients. This procedure is used to treat the condition of pes equines (i.e., extreme plantar flexion of the ankle), which occurs in spastic patients.
MYOFASCIAL FORCE IS NOT RESTRICTED TO THE BELLY AND TENDON OF ONE MUSCLE
Our experimental findings (6) on intramuscular myofascial force transmission within a dissected muscle that was left in situ led us to speculate (2) about additional paths for force transmission from muscle, in addition to the two intramuscular pathways to the tendon. Two conceivable major pathways for force transmission can be distinguished: (a) extramuscular myofascial force transmission, that is, transmission to extramuscular connective tissues of the compartment; and (b) intermuscular myofascial force transmission, that is, direct transmission between the intramuscular connective tissue stromata of two muscles. If no force is transmitted to or from the muscle at other locations than the tendons, the forces exerted at the proximal and distal tendons of a muscle are equal. Therefore, any difference in these two forces is attributable to the transmission of force from or onto the muscle–tendon complex anywhere between the two locations. Experiments aimed at measuring such proximodistal force differences can be performed readily in muscles that cross more than one joint without damaging the connective tissues, particularly around the bellies of the target muscle and its neighbors. Rat extensor digitorum longus muscle (EDL; the dorsal flexor of the ankle and extensor of several joints within the foot) was used for such experiments (4). In contrast to humans, the rat EDL also crosses the knee joint, where it exerts an extensor moment. This anatomy allows access to its proximal tendon without interfering with the tissues in the anterior crural compartment and access to the distal tendons (EDL has four parallel heads with one distal tendon each) to be dissected mostly within the foot. Only release of the tendon from the retinaculae at the ankle required some interference with the anterior crural compartment.
EXTRAMUSCULAR MYOFASCIAL FORCE TRANSMISSION
The transmission of force between the intramuscular connective tissue of a muscle and extramuscular connective tissues is referred to as extramuscular myofascial force transmission. This force corresponds to a reaction force that prevents the active or passive sarcomeres within fibers of the target muscle from shortening. Examples of extramuscular connective tissues are the fascia that constitutes the boundaries within a compartment and the connective tissue tracts that contain bundles of nerves and blood vessels (Fig. 3A).
In most physiological experiments, the intermuscular and extramuscular connective tissues are dissected to gain access to the muscle. Although few investigators recognize that such dissection influences the observed muscle properties (1), in vivo such connections are probably important functionally. As a consequence, results of experiments on isolated but in situ muscles, which are extremely important for studying basic properties of muscle, should not be extrapolated directly to the in vivo condition.
“Isolation” of a muscle by performing full lateral compartmental fasciotomy as well as dissection of extramuscular connective tissue has a major effect on the length-force characteristics of rat muscle (12). This particular study was performed to mimic general effects of an operation performed on the arm of human patients with spastic flexor carpi ulnaris muscle (FCU). Such operations are aimed at transposing the FCU tendon to an extensor location to relieve limitations in wrist joint range of motion imposed by that muscle. First, the compartment is opened and tenotomy is performed, followed by subsequent progressive dissection of the FCU belly for about 50% of its length. This procedure alters FCU length-force characteristics. Such altered length-force characteristics suggest that the integrity of the compartment plays an important role in modifying muscular properties. They also suggest the hypothesis that intermuscular and extramuscular tissues actually may cause the limitations in range-of-joint movement in patients with spastic muscles. To demonstrate unequivocally that such effects are related to myofascial force transmission requires simultaneous force measurements at proximal and distal tendons, which is almost impossible to obtain in humans for obvious ethical reasons.
However, proximal and distal force can be measured readily in experimental animals: differences in the proximal and distal forces in maximally active and passive rat EDL were reported in several experimental studies (3–5,9). Although differences in these forces indicate an additional path for the transmission of force along the muscle, the results cannot distinguish between extramuscular and intermuscular sources. Additional interventions are required to make this distinction: (a) after the compartment was opened with a full lateral fasciotomy, the proximodistal force difference in the active as well as the passive muscle persisted with a little decreased amplitude (7); (b) similar effects were also found when, after the fasciotomy, all muscles were removed from the anterior crural compartment except for EDL, to leave only the extramuscular connective tissues available for myofascial force transmission. Such conditions caused a decrease in the force, but the proximodistal difference in EDL force persisted, albeit after yet another decrease in amplitude (4); (c) further dissection of the extramuscular connective tissues, except for its most crucial elements containing major blood vessels and nerves to EDL, abolished the proximodistal force difference for active but not for passive EDL (4). Therefore, it appears that extramuscular connective tissues, such as the neurovascular tract, are capable of transmitting force. However, questions regarding the intramuscular effects of such extramuscular force transmission are not presently accessible to experimental research. For example, it is not possible to measure or even to estimate local sarcomere length within the muscle experimentally. It is clear that noninvasive methods, such as nuclear magnetic resonance imaging, are needed for the study of myofascial force transmission.
Finite Element Model Results
Until noninvasive methods are developed and applied, such questions can be addressed only by finite element modeling of extramuscular and other elements of a muscular compartment (Fig. 4A). Any model used for this purpose, however, should have an adequate representation not only of the intracellular domain, but also of the extracellular matrix, including the connections between these domains (15). Obtained with such a model, Figure 4B shows distributions of local strains within the muscle fibers, calculated for locations throughout the muscle (maximal strain range, −24 to +30%). These local strains, which provide estimates of sarcomere lengths at different locations, show two major effects: (a) even in a muscle that is active at high lengths, there are a considerable number of sarcomeres that have shortened substantially because of the local decrease in the extracellular matrix forces; (b) the model is consistent with the hypothesis that extramuscular myofascial force transmission enhances the distribution of sarcomere lengths along the muscle fibers. This means that the sarcomeres at proximal or distal ends of muscle fibers may exert quite different active forces on the aponeuroses, and the extracellular matrix prevents sarcomeres of adjacent muscle fibers from acting completely independently. As a consequence, zones of similar sarcomere lengths are arranged more or less in bands across the fibers (Fig. 4B) so that there is more uniformity of sarcomere length in segments of the fibers arranged in parallel than there is for sarcomeres arranged in series within the same fiber. These types of nonuniformities do influence muscle properties substantially, but usually are not consider in either biomechanical modeling of in vivo movement or in most interpretations of experimentation data.
INTERMUSCULAR MYOFASCIAL FORCE TRANSMISSION BETWEEN SYNERGISTS
Figure 3B shows an example of the collagenous connections between two adjacent muscles within a muscle group. If these connections are stiff enough, force can be transmitted between the two intramuscular stromata of connective tissue, an effect that is known as intermuscular myofascial force transmission. By this mechanism, it is possible for the force originating in a neighboring muscle to supplement that exerted by the target muscle. At which of the two muscles such integration will occur depends on the relative stiffness of the force transmission paths of the two synergists. Because the points of application of the external forces are intermediate between the ends of the muscle fibers, intermuscular myofascial force transmission also should enhance the distribution of sarcomere lengths arranged in series within the muscle fibers.
There is only little awareness in the muscle literature on the connections that do exist between muscles and how these attachments may influence function. For example, Pond (11) noted:
… in adult insects[,] each muscle was anatomically distinct and only very loosely attached to adjacent muscles. Even its constituent fibers could easily be teased apart, and each one could, and apparently in life often did, contract independently of its neighbors. I was therefore shocked, on first dissecting a horse, to find that muscles of the limbs and trunk were tightly bound to each other with fascia—intermuscular fat and that tough connective tissue which binds the skin to the muscles. Even muscles that clearly had opposing functions were firmly attached to their neighbors …. A slight twitch in one muscle will set up a web of uncharted forces in the whole bundle of muscles. When tissues are required to move relative to each other there seems to be no difficulty in forming low-friction surfaces, as for example, tendons in tendon sheets. But a naive observer, accustomed to invertebrate systems, is impressed by the ubiquitous fascia, everywhere linking muscle to muscle and muscle to skin. Even when the anatomy of the limb is such that nearest-neighbor muscles do not lie closely opposed, chunks of intermuscular fat and connective tissue bind the muscles together. In preparing tissues for physiological study of the muscle receptors or the mechanical response of the muscle, many investigators routinely “free” the muscle to be studied from its associated connective tissue. I have recently done some in vivo experiments on muscles in the guinea pig hind leg which suggest that there are quite large differences in the force recorded through the insertion of “freed” muscles and those in which the associated connective tissue is left intact. …
Unfortunately, at the time, these observations and speculations were not quantified or otherwise substantiated and subsequently have been largely ignored in literature.
Presently, there is both direct and indirect evidence that force can be transmitted between synergists located within one compartment. The direct evidence is based on the observation that when the length of the rat EDL muscle–tendon complex is kept constant, but after the neighboring tibialis anterior and extensor hallucis longus muscles were lengthened, there is a decrease in the EDL isometric force during (9). The indirect evidence is based on the observation that compartmental fasciotomy of the anterior crural compartment causes the force exerted by tibialis anterior–extensor hallucis longus complex when kept at a constant length to increase by approximately 30% at all EDL lengths examined.
For many conditions, there are clear indications that when one muscle of a group is stretched, it will “attract” force from neighboring muscles to be exerted at the tendon of the stretched muscle. This appears to be a general feature of intermuscular interactions. This effect is caused by an increase in the stiffness of the intermuscular connections resulting from a lengthening of the target muscle. Because the pathways of force transmission depend on the relative stiffness of the different elements, less force will be transmitted to the muscle’s own tendon and more via the intermuscular connections to the extracellular matrix of lengthened muscle to be exerted at the distal tendon of that muscle.
Finite Element Model Results
Figure 5B shows distributions of local strains for a muscle that contains only intermuscular connections to a neighboring muscle (15). Note that an enhanced distribution is found for sarcomere lengths arranged in series within fibers. This is apparent from the increased range of local strains (strain range, −20 to +67%) than for the muscle with only extramuscular connections (Fig. 4B). This enhanced distribution is present in all modeled muscle fibers. Thus, force originating from many sources outside of a muscle may contribute to the force exerted at one of its tendons. Because the location of such integration depends on the relative stiffness of the available force transmission paths, the selection of the muscle in a group of synergists that actually exerts the myofascial force at its tendon is influenced by a number of factors that will alter the length of intermuscular and extramuscular connections. These factors include muscle length and differences in length between neighboring muscles.
EFFECTS OF MUSCLE LENGTH AND RELATIVE POSITION
When a muscle is kept within a myofascial context, the effects of muscle length on force exerted may differ at proximal and distal tendons (4,5). To identify the mechanism responsible for such effects, EDL length was changed by symmetrical displacement of either the proximal or distal tendons. This resulted in quite different effects (3,5): at most lengths studied, isometric force at the tendon that had been displaced was higher than at the fixed tendon. Hence, the difference in the proximodistal force also reversed its sign, depending on the location of length change of the muscle. Because the muscle–tendon complex length was not different between these conditions, some other factor must be responsible for these effects. In this experiment, the only factor manipulated was the position of (parts of) the target muscle relative to extramuscular structures and other muscles. Figure 6 schematically illustrates an even more constrained condition: without changing its length, a muscle is moved through its connective tissue context. Such an experiment showed that force exerted at proximal and distal EDL tendons is highly dependent on the relative position of the muscle. Also, the proximodistal force difference is dependent on the relative position of the muscle, and therefore also the quantity and direction of myofascial force transmission.
Differences in the moment arm to the joint will also lead to in vivo changes of the relative position of two synergists. For a given angular displacement about a joint, the muscle furthest from the joint will experience the greater amount of shortening. Furthermore, the position of multiarticular muscles relative to monoarticular synergists will depend on the relative displacement that occurs about the involved joints. Because of these factors, in vivo changes of the relative position of muscles must be frequent events.
MYOFASCIAL FORCE TRANSMISSION BETWEEN ANTAGONISTS
Because relative position is a major determinant of muscle force and myofascial force transmission, an analysis should be performed to determine when the biggest changes in this parameter occur in vivo. Any such analysis will yield the same conclusion: the most sizable changes in the relative position of two muscles will occur between antagonist muscles. For example, when the agonist shortens actively, the antagonist will have to lengthen either passively or in an eccentric contraction. Whether a mechanical interaction occurs will depend on the nature and stiffness of connections between such muscles, if any exist. There cannot be intermuscular myofascial force transmission, as defined above, because direct interaction between the intramuscular stromata of antagonists is not possible because of an intermuscular septum that forms part of the compartment boundaries. However, if both intramuscular stromata attach to the septum via extramuscular connective tissue, such as the neurovascular tracts, intermuscular interaction by means of extramuscular myofascial force transmission is possible. Morphologic analysis confirms the existence of such extramuscular connective tissues (Fig. 7). Nonetheless, only preliminary experimental data has been published on the mechanical interaction between rat antagonist muscles, namely, for peroneal muscles and muscles within the anterior crural compartment (3). This issue awaits further inquiry.
SOME FUNCTIONAL CONSEQUENCES OF MYOFASCIAL FORCE TRANSMISSION
It has already been recognized (10) that integration of multiple levels of organization may yield unique mechanical properties of motor units. Thus, the properties of a single motor unit differ when the unit is examined in isolation, when assessed within the extracellular matrix with other motor units, and when the relative positions of muscles are altered. These effects will also depend on the relative location of the various motor units within the muscle as force is increased by volition from zero to maximum.
Perfect coordination of muscular activity to perform a certain task precisely at minimal metabolic cost is possible only if relative stiffness of different intramuscular, extramuscular, and intermuscular pathways, which determine the partition of force transmitted via the major pathways, is optimal for exertion of moments and forces at the joints. However, joint stability is a condition sine qua non for movement. Myofascial force transmission may help to exert forces at the joint. Previously, it has been pointed out that joint capsule and capsular ligaments cannot be described as separate entities, but should be considered in unity with muscles that are arranged mechanically in series with them: some muscle force is transmitted onto the periost around the bones, as well as onto intermuscular septa. This occurs because many muscle fibers have an origin or insertion not on an intramuscular aponeurosis, but on such structures (myotendinous pathways). Even if this is not the case, extramuscular myofascial pathways will transmit force to these structures, and such structures are continuous with the capsule and ligaments of the joint.
It seems likely that a major role of motor control is the tuning of the stiffness of many connective tissue elements of a limb. This may seem an incredibly complex task for the central nervous system because the number and degree of interactions seem almost limitless. Three factors simplify this problem: (a) intramuscular and muscle-related receptors of different kinds are located at specialized positions within muscles to monitor muscle–connective tissue units, in a way that does not coincide with the topography of muscles as morphological entities; (b) many receptors and nerve endings are located within the walls of the compartment (general fascia, intermuscular septa, and interosseal membrane); (c) some specific interactions are likely to occur more frequently and may be learned in the course of many years of development or training.
Joint Range of Movement
Some of the expected functional effects are mediated by changes in joint angle-moment curves. The range of motion over which a muscle–tendon complex can exert active forces will increase with a broadening of sarcomere length distributions within the muscle. In shortened and lengthened conditions, even limited enhancement of the active force may have quite important effects: limitations to movement are removed by enhancing the length range of active force generation. Even within a mid range of joint movement, it is likely that extramuscular forces and possibly also intermuscular forces will cause high strains locally in muscles and connective tissue of a compartment, particularly for greater changes in relative muscle location. Such local deformations may play a major role in the cause of afflictions of the human locomotion apparatus, such as repetitive strain injury, tennis elbow, musicians arm, and the adaptation of muscle and connective tissues to such conditions.
The effects of myofascial force transmission have a major impact on our understanding of in vivo muscle function. The most appropriate strategy to examine these interactions is using an integrative approach to the study of muscle function: that is, to combine the knowledge of functional properties of isolated elements of the locomotor system with the knowledge of effects of nearby structures belonging to a higher level of organization and their interactions. Such an approach will be fruitful at all levels of organization. It seems that such an approach is essential, because without it we are not likely to be able to describe functional characteristics adequately and to enhance understanding of many aspects of human movement sufficiently.
1. Denny-Brown, D.E. The histological features op striped muscle in relation to its functional activity. Proc. Royal Soc. B104: 371–411, 1929.
2. Huijing, P.A. Muscle as a collagen fiber reinforced composite material: force transmission in muscle and whole limbs. J. Biomech. 32: 329–345, 1999.
3. Huijing, P.A. Intra-, extra- and intermuscular myofascial force transmission of synergists and antagonists: effects of muscle length as well as relative position. Internat. J. Mech. Med. Biol. 2: 1–15, 2002.
4. Huijing, P.A., and Baan, G.C. Extramuscular myofascial force transmission within the rat anterior tibial compartment: proximo-distal differences in muscle force. Acta Physiol. Scand. 173: 1–15, 2001.
5. Huijing, P.A., and Baan, G.C. Myofascial force transmission: muscle relative position and length determine agonist and synergist muscle
force. J. Appl. Physiol. 94: 1092–1107, 2003.
6. Huijing, P.A., Baan, G.C., and Rebel. G. Non myo-tendinous force transmission in rat extensor digitorum longus muscle. J. Exp. Biol. 201: 682–691, 1998.
7. Huijing, P., H. Maas, and G.C. Baan. Compartmental fasciotomy and isolating a muscle from neighbouring muscles interfere with extramuscular myofascial force transmission within the rat anterior tibial compartment. J. Morphol.
(in press, 2003).
8. Lindhard, J. Der Skeletmuskel und seine Funktion. Ergeb. Physiol. 33: 22–557, 1931.
9. Maas, H., Baan, G.C. and Huijing, P.A. Intermuscular interaction via myofascial force transmission: effects of tibialis anterior and extensor digitorum longus length on force transmission from rat extensor digitorum longus muscle. J. Biomech. 34: 927–940, 2001.
10. Monti, R.J., Roy, R.R., and Edgerton. V.R. Role of motor unit structure in defining function. Muscle Nerve. 24: 848–866, 2001.
11. Pond, C.M. The importance of connective tissue within and in between muscles. Behav. Brain Sci. 5: 562, 1982.
12. Smeulders, M.J.C., Kreulen, M. Baan, G.C. and Huijing. P.A. Progressive surgical dissection for tendon transposition affects length-force characteristics of rat flexor carpi ulnaris muscle. J. Orthoped. Res. 20: 863–868, 2002.
13. Tidball, J.G. Force transmission across muscle cell membranes. J. Biomech. 24: 43–52, 1991.
14. Trotter, J.A., and Purslow. P.P. Functional morphology of the endomysium in series fibered muscles. J. Morphol. 212: 109–122, 1992.
15. Yucesoy, C., Koopman, H.J. Grootenboer, and P.A. Huijing. Inter- and extramuscular myofascial force transmission affect muscular mechanics substantially: a review of recent results. In:Proceedings of the Dutch Annual Conference on Biomedical Engineering
, edited by J. Feijen. Enschede: Institute for Biomedical Technology, University of Twente, 2002.