Mechanical Properties of Aponeurosis and Tendon of the Cat Soleus Muscle During Whole-Muscle Isometric Contractions

JOURNAL OF MORPHOLOGY 224:73-86 (1995)

STEPHEN H. SCOTT AND GERALD E. LOEB MRC Group in Sensory-Motor Physiology, Department of Physiology, Queen's University, Kingston, Ontario, Canada K7L 3N6 ABSTRACT Recent studies have suggested that the mechanical properties of aponeurosis are not similar to the properties of external tendon. In the present study, the lengths of aponeurosis, tendon, and muscle fascicles were recorded individually, using piezoelectric crystals attached to the surface of each structure during isometric contractions in the cat soleus muscle. We used a surgical microscope to observe the surface of the aponeurosis, which revealed a confound- ing effect on measures of aponeurosis length due to sliding of a thin layer of epimysium over the proximal aponeurosis. After correcting for this artifact, the stiffness computed €or aponeurosis was similar to tendon, with both increasing from around 8 Fo/L, (Fo is maximum isometric force and Lc is tissue length) at 0.1 Fo to 30 Fo/L, at forces greater than 0.4 Fo. At low force levels only (0.1 Fo), aponeurotic stiffness increased somewhat as fascicle length increased. There was a gradient in the thickness of the aponeurosis along its length: its thickness was minimal at the proximal end and maximal at the distal end, where it converged to form the external tendon. This gradient in thickness appeared to match the gradient in tension transmitted along this structure. We conclude that the specific mechanical properties of aponeurosis are similar to those of tendon. o 1995 Wiley-Liss, Inc.

The contractile components of muscle ex- from muscles representing a wide range of ert forces on the relatively rigid skeleton via musculotendinous architectures. in-series connective tissue. The mechanical An alternative to the study of whole arrangements and properties of these connec- muscles is to develop models of particular tive tissues may have important effects on muscles based on general mechanical func- the dynamics and energetics of the relation- tions provided by the various individual tis- ship between muscle fascicles and skeleton. sues of which they are composed: contractile For example, spring-like compliance in the components, parallel-elastic components, se- connective tissue may absorb and release elas- ries-elastic components and the mechanical tic strain energy (Alexander and Vernon, '75; coupling among them. The mechanical func- Cavagna, '77) and may distort the relation- tions that reside wholly within the external ship between the overall length of a muscle tendon or individual muscle fibers are ame- and the proprioceptive feedback from length nable to study in isolated preparations of those tissues (e.g., Ker, '81 and Edman, '88, sensors such as muscle spindles (Rack and respectively). However, the muscle fibers of Westbury, '84; Hoffer et al., '89; Griffiths, many muscles are attached not to tendon or '91, but see Elek et al., '90). The magnitude bone, but rather to sheets of internal and/or of these effects usually depends on the spe- external connective tissue called aponeuro- cific morphometry of the muscle, and extrapo- ses. The mechanical properties of an exten- lation to other muscles requires an explicit or sive aponeurosis may dominate the series- implicit mathematical description. However, the selection of the qualitative form and particular quantitative values for the compo- Address reprint requests to Dr. G.E. Loeb, Bio-MedicalEngineer- ingunit, Queen's University, Kingston, Ontario,K7L 3N6 Canada. nents of the model of whole muscle is ham- Dr. S.H. Scott is now at the Departement de Physiologie, pered by the paucity of experimental data Universite de Montreal, Montrdal, Qu&ec, H3C 357 Canada. o 1995 WILEY-LISS, INC. 74 S.H. SCOTT AND G.E. LOEB elastic and mechanical coupling functions, grammed sequences of electrical stimuli to but these properties are not well understood. the nerve entering soleus (50 Hz for 400 ms) The intimate and distributed connections while controlling the position and velocity of between an aponeurosis and terminating a step-motor attached in-series to the muscle. muscle fibers make it difficult to estimate its Whole muscle isometric contractions were mechanical properties directly. Techniques recorded at 16 lengths that evenly spanned such as photogrammetry have been em- the anatomical range of motion of soleus. ployed only recently to record length changes The program recorded the force generated by of connective tissue during changes in muscle the muscle and the position of the motor, force (Huijing and Ettema, '88/89; Ettema which was calibrated in terms of whole- and Huijing, '89; Lieber et al., '91; Zuurbier muscle length. The program also recorded et al. '94). Based on such data, Lieber et al. the length of individual portions of the muscle ('91) suggest that aponeurosis and tendon (fascicle, aponeurosis and tendon) using pi- have very different mechanical properties: ezoelectric crystaIs (Sonomicrometer 120; aponeurosis was found to be four times more Triton Technology) attached to the surface of compliant than tendon. Recently Ettema and the muscle. The crystals transmitted ultra- Huijing ('89) claimed to have shown that the sonic pulses through a pool of parrafin oil; length of an aponeurosis is dependent not the pool was created by tethering the skin only on muscle force, but also on fascicle flaps to a metal ring. Each ultrasonic pulse length. These studies suggest that the organi- was transmitted by one crystal and received zation of an aponeurosis is complex and that by another; the transit time was converted its mechanical properties are not similar to into a measure of length. The transmission external tendon. of ultrasonic pulses through the paraffin oil, In the present study multiple piezoelectric rather than directly through the muscle tis- crystals were attached to the surface of the sue (i.e., Griffiths, '91; Hoffer et al., '89) cat soleus muscle. These provided direct mea- negates the conflicting effect of muscle force surement of the length of the aponeurosis on the velocity of the ultrasonic signals (see and tendon during isometric contractions per- Hatta et al., '88; Caputi et al., '92). Some formed at different whole-muscle lengths. preliminary experiments were performed in These recordings were used to estimate the which the crystals were attached to a Silastic stiffness properties of the tendon and aponeu- sheet (Reinforced Silastic Sheeting, Dow rosis. In addition, anatomical techniques were Corning) that was sutured to the muscle used to relate the structure of the aponeurosis to its mechanical properties. surface. In subsequent experiments, the crystals were attached to the surface of the muscle MATERIALS AND METHODS using cyanoacrylate adhesive (Accu-Flo Crazy Data collection Glue, Lepage). Crystals attached by the adhe- Experiments were carried out on adult cats sive were found to remain secure for long (2.75-4.7 kg; either sex) anesthetized ini- periods of time, although constant monitor- tially with 35 mg/kg of sodium pentobarbital ing was required in order to ensure the stabil- IP with supplemental doses IV as needed to ity and alignment of the crystals and thus the suppress withdrawal reflex. An incision was quality of the output signals. Although the made along the posterior surface of the right glue may tend to augment the stiffness prop- leg from the calcaneum to the popliteal fossa. erties of the aponeurosis, the magnitude The popliteal fat pad was removed and the would be minimal since the size of the drop of plantaris and gastrocnemius muscles were glue was small relative to the distance tra- resected. A tri- or bipolar nerve cuff was versed between crystals (approx. 1:15). placed around the tibial nerve. Nerve stimu- Crystals were positioned initially at the lation was elicited by a biphasic constant proximal and distal ends of the aponeurosis current pulse with a duration of 0.2 ms and to record the entire length of this structure. an amplitude of four times the current to The distal end of the aponeurosis was defined elicit a just-visible contraction. The soleus as the most distal point of muscle-fiber attach- tendon and a small piece of the posterior ment. Most of the aponeurosis appears as calcaneus were removed from the foot and longitudinal white strands of connective tis- attached to a force transducer (linear up to sue that converge to form the external ten- 55 N), which was attached to a stepping don. However, the proximal end of the apo- motor. A computer-controlled data-acquisi- neurosis is less evident; its border can best be tion system was used to deliver prepro- visualized at long muscle lengths when the MECHANICAL PROPERTIES OF AF’ONEUROSIS AND TENDON 75 perimeter of the aponeu:. -sis can be seen as a 0.1 Fo above and below the desired force level change in light reflectance due to the angular (Fig. 1).Stiffness estimates were normalized change between the surfxe of the fascicles to Fo for each preparation and to L,, the and aponeurosis. length of the recorded tissue at slack length The length of the entire aponeurosis was and zero muscle force, in order to define a recorded by one set of crystals in the first few dimensionless term called the “specific stiff- experiments. In later experiments, a crystal ness” of the tissue. To determine the stiff- was positioned in the middle of the aponeuro- ness of a particular muscle’s connective tis- sis, and the lengths of the distal and proximal sue, the dimensionlessspecific stiffness would portions of the aponeurosis were recorded be multiplied by that muscle’s Fo (typically simultaneous, In the last three experi- estimated from its physiological cross-sec- ments, crystals ./ re also attached on the tional area in cm2 and the specific tension of medial and lateral edqes of the aponeurosis muscle in N/cm2) and divided by its L, (in so that the width of this structure could be cm) resulting in N/cm. recorded simultaneously with its length. The length of the entire tendon could not Anatomical observations be recorded because the crystals must trans- The superficial surface of the soleus muscle mit their ultrasonic pulses through some me- was examined, using a surgical microscope dium, in this case, a paraffin oil pool. The (Ome Olympus), in four cats. The gastrocne- distal end of the tendon, which was attached mii were resected, and the surface of the to the clamp, was above the paraffin pool for aponeurosis and the epimysium of the soleus contractions at all but the shortest of whole- were observed under static and dynamic con- muscle lengths. Therefore, the crystals were ditions. positioned at the most proximal end of the The right soleus muscles of two cats were external tendon and at a position 1.0-1.5 cm dissected and fixed in 10% formalin for 24 hr. more distally. Measures of this segment of The mass of each muscle was recorded along tendon were possible at most, but not all with the length of the aponeurosis. The whole-muscle lengths. muscles were then cut along the midline of the muscle and orthogonal to the surface of Data analysis the aponeurosis. Each half was trimmed just The force-length relationships of the apo- above the proximal end and just below the neurosis and tendon were computed during distal end of the aponeurosis. The remaining individual isometric contractions at 16 differ- portions were divided into three sections and ent whole-muscle lengths. each section was embedded in a mixture of Stiffness is a measure of the relationship paraffin and plastic polymers (Paraplast, between the change in force transmitted by a Sherwood Medical). A single 5-pm section structure and its corresponding change in was cut from each block orthogonal to the length. For the contractions recorded in the present study, the simplest method to estimate stiffness would be to compare the change in muscle force and tissue length Specific stiffness at 0.5 Fo = dt between adjacent data-collectionsample periods. However, the nonlinear rise and fall in muscle force during tetanic contractions result in large and rapid changes in length at low force levels and very slow changes at $ 0.5 higher force levels when peak muscle force was approached. Therefore, the variability in the estimates of stiffness would increase at high force levels because the change in force I I , and length were small between adjacent :::u0.0 sample periods. In the present study, stiff- 1.00 1.01 1.02 1.03 1.04 1.05 ness was estimated at specific force levels Tendon length (L,) equal to increments of 0.1 Fo (where Fo is the maximal isometric force generated at the op- Fig. 1. Felis cutus. The specific stiffness of connective tissue was estimated at increments of 0.1 Fo from the timal fascicle length, Lo). Each estimate of whole-muscle isometric contractions. Each estimate was stiffness was based on interpolation from tis- based on the tissue lengths recorded at force levels that sue lengths recorded at force levels that were were 0.1 Foabove and below the desired level. 76 S.H.SCOTT AND G.E. LOEB surface of the aponeurosis and stained with tachment of the crystals to the muscle sur- Gomori's trichrome to highlight the connec- face, the large hysteresis between the rising tive tissue of the aponeurosis (Drury and and falling phase of isometric contractions Wallington, '80).The thickness of the aponeu- disappeared (Fig. 2B). Similar inconsisten- rosis was calculated at 1-mm intervals along cies between the rising and falling phases of its surface from each section, using a micro- the contractions were found during prelimi- scope graticule at 250 power (Dialux 20, Leitz nary experiments in which the crystals were Wetzler). Although shrinkage of the connec- attached to a Silastic Sheet (Reinforced Silas- tive tissue was not accounted for in this analy- tic Sheeting, Dow Corning), which was su- sis, we were interested primarily in the tured to the surface of the muscle. In the relative thickness along the extent of this experiments described here, all crystals were tissue. monitored constantly to ensure that they remained attached firmly to the underlying RESULTS tissue by cyanoacrylate adhesive. The force-length properties of the connec- Comparisons of length and computations tive tissue in the soleus were analyzed fully of stiffness presented here for tendon and in five cats. The lengths of the aponeurosis aponeurosis are all based on the rising phase and tendon were 4.4 2 0.6 cm and 2.7 2 0.3 of the contractions, which showed only small cm, respectively, at slack length under pas- and inconsistent differences in stress-strain sive conditions. Therefore, the nominal fas- relationships from the falling phase. We have cicle length of the soleus (3.8 k 0.6 cm) is therefore assumed throughout that these approximately one-half the length of its in- structures act purely elastically and have neg- series connective tissue. ligible viscosity under physiological condi- The measurement of the relationship be- tions. tween the force generated by the muscle and the length of any portion of the muscle de- Aponeurosis and tendon lengths during pended substantially on the quality of the isometric contractions at different attachment of the crystals to the muscle sur- muscle lengths face. Figure 2A displays the recorded force- Figure 3 illustrates the length changes in length relation of the aponeurosis when the the fascicle, aponeurosis and tendon during crystal was seen to be attached poorly to the the rising phase of isometric contractions at underlying surface of the muscle. The record- five different whole-muscle lengths. In gen- ing shows a large difference in the force- eral, the force-length relationship of tendon length relation between the rising and falling was independent of starting muscle length. phase of the contraction, suggesting that However, the starting length of the aponeuro- there was a large hysteresis between loading sis shifted in a way that was not dependent and unloading of the aponeurosis. Upon reat- on muscle force but rather on muscle length, 3' I

0 2.20 2.25 2.30 2.35 1.65 1.70 1.75 1.80 Length (crn) Length (cm)

Fig. 2. Felis catus. The recordings of the aponeurosis attached poorly to the underlying surface of the muscle. were affected by the quality of the attachment of the B: The hysteresis during contractions largely disap- piezoelectric crystals. A: A large difference in the force- peared when the crystals were re-attached firmly to the length relationship for the rising and falling phase of muscle surface. contractions was often recorded when the crystals were MECHANICAL PROPERTIES OF APONEUROSIS AND TENDON 77 producing a family of force-length curves of the aponeurosis also increased when the shown in Figure 3B. Similar shifts in the fascicles shortened. length of the aponeurosis associated with the length of the fascicles have been reported in Anatomical observations of the surface of the rat medial gastrocnemius and extensor digi- aponeurosis torum longus muscles (Huijing and Ettema, Observations of the surface of the aponeu- '88/89; Ettema and Huijing, '89). rosis using light microscopy revealed that the In order to study the properties of the surface was composed of a thin layer of aponeurosis further, the length of the proxi- crimped collagen fibers, similar to the colla- mal and distal portions of the aponeurosis gen in the external tendon (Diamont et al., were recorded separately, as was the width '72; Rowe, '85). The collagen was organized across the proximal end of the aponeurosis into longitudinal bands approximately 200 (Fig. 4). The proximal and distal portions of pm wide. The width of the aponeurosis was the aponeurosis both increased in length as relatively uniform along its entire extent: force rose during the contraction. The start- this dimension was maximal near its proxi- ing length of the proximal portion of the mal end and decreased slowly distally. At its aponeurosis continued to shift with changes distal end, the longitudinal bands converged in whole-muscle length. In fact, the relative to form the external tendon. The relatively change in length increased. In contrast, the uniform width of the aponeurosis suggests length of the distal portion of the aponeuro- that termination of muscle fibers and thus sis did not shift with fascicle length. The transmission of muscle fiber force onto the dependence of aponeurosis length on fascicle aponeurosis was relatively consistent and uni- length seemed to be isolated to only the most form along its longitudinal axis. proximal end of the aponeurosis; this corre- The muscle was covered with epimysium sponds with the region of the aponeurosis composed of a thin membrane containing where the strands of connective tissue dimin- collagen fibers oriented in a crossed-ply ar- ish in size and terminate. The width of the rangement similar to the observed mesh- aponeurosis increased during the rising phase work of collagen in the perimysium and endo- of the contractions (Fig. 4D).The initial width mysium (Rowe, '81; Purslow, '89), except

25 ea20 0 15 LL 10 5 0 1.5 2.0 2.5 3.0 3.5 4.0 31 3.3 3:4 3.5

Length (cm) 35 1 tendon C 30

5 0 1.05 1.10 1.15 1.20 Length (cm)

Fig. 3. Felis catus. The lengths ofthe fascicle, aponeu- whole-muscle lengths. B: The length of the aponeurosis rosis and tendon are shown during the rising phase of increased with force during each contraction, but its whole-muscle isometric contractions recorded at differ- starting length shifted for contractions at longer whole- ent muscle lengths. Measurements from the same contrac- muscle lengths. C: The change in length of the tendon tion are denoted with similar line types. A: The fascicle during contractions was consistent at all whole muscle shortened as force increased during each contraction and lengths. its length increased for contractions recorded at longer 78 S.H. SCOTT AND G.E. LOEB

2.5 3.0 3.5 4.0 4.5 2.5 2.6 2.7 2.8 Length (cm) Length (cm) B

0 10 LL 5

0 1.20 1.25 1.30 1.35 0.60 0.65 0.70 0.75 Length (cm) Width (cm)

Fig. 4. Felis catus. The length of different portions of the aponeurosis, as well as the muscle fascicle were recorded simultaneously during the contractions (A-D). Measurements from the same contraction are denoted with similar line types. Note that the length of the proximal, but not the distal portion of the aponeurosis shifted for contractions at different whole-muscle lengths.

that it was more sparse. The epimysial colla- the length of the aponeurosis related to fas- gen fibers over the fascicles were oriented cicle length was artifactual and created by symmetricallyat equal angles to the long axis movement of the epimysium over the surface of the fascicles. When the fascicles were of the aponeurosis. lengthened the crossed-ply was reoriented to The crossed-ply collagen of the epimysium become more aligned with the long axis of the along the side of the muscle was also oriented fascicles in order to accommodatethe fascicle symmetrically about the longitudinal axis of length change. A similar re-orientation of the the fascicles. Therefore, epimysial sliding crossed-ply arrangement of collagen fibers should alter the apparent width of the apo- has been observed for the perimysium (Purs- neurosis: fascicle shortening would lead to a low, '89). decrease in the apparent width of the aponeu- The most important feature of the epimy- rosis. However, the apparent width of the sium is its termination onto the aponeurosis. aponeurosis was observed to increase when The collagen fibers of the epimysium were the fascicle shortened based on the piezoelec- observed to cover the perimeter of the aponeu- tric crystals. This suggests that the width of rosis for a couple of millimeters prior to their the aponeurosis does increase when the fas- attachment to the collagen fibers of the apo- cicle shortens and may in fact increase more neurosis. Therefore, when the fascicles than was recorded. lengthened, piezoelectric crystals that were The midline-stained sections of the aponeu- attached to the epimysium would move away rosis displayed a marked variation in the from the aponeurosis due to the lengthening thickness of the aponeurosis along its length of the epimysium (Fig. 5).Apparent changes (Fig. 6). The thickness was minimal (less in aponeurosis length would be observed only than 10 pm) at its proximal end, but in- at the most proximal end because the colla- creased gradually at more distal positions. gen fibers of the epimysium terminate only Near the tendon, the thickness of the aponeu- on the perimeter of the aponeurosis. Indeed, rosis increased more rapidly reaching 200 this was what was observed in the records of pm at its most distal end. This rapid increase aponeurosis length; length changes related in the aponeurosis thickness over the distal to fascicle length were restricted to the most quarter of the aponeurosis coincides with a proximal end"of the muscle (Fig. 4). This rapid narrowing of the aponeurosis surface strongly suggests that the apparent shift in as the loneitudinal- bands converge- to form MECHANICAL PROPERTIES OF APONEUROSIS AND TENDON 79

Fig. 5. Felis cutus. Schematic lateral cross-sectional hatch) that is free to slide over the anatomical junction and superior views of muscle showing relationships be- (dashed lines). A Recorded length of aponeurosis (J2l tween piezoelectric crystals used for length measure- and fascicle (La are correct when central crystal is above ments (circles) and their mounting sites on the distal the junction. B: When the muscle lengthens, the length edge of the aponeurosis (right), bony origin of the muscle recorded by the crystals spanning the fascicle underesti- fascicles (left) and junction between fascicles and aponeu- mates the true fascicle length (Lf < Ld and the crystals rosis (center). Note that the central crystals are actually over the aponeurosis erroneously record an increase in mounted on a slip of epimysial connective tissue (cross- length for the aponeurosis (L: > La). the external tendon. In general, the gradient Stiffness measures in the thickness of the aponeurosis appeared The specific stiffness of the soleus tendon to balance the gradient in the force transmit- is shown in Figure 7. Each value equals the ted across this structure, supporting the no- tion that strain will be distributed uniformly stiffness averaged from all recorded contrac- in this structure when active tension is gener- tions. Specific stiffness of the tendon was ated homogeneously in this muscle. approximately 8 Fo/L, at 0.1 Fo but rose to

200 1 0.

00 0

vI 0. v) OO0 .. x l5OI100 . 0. ,,..-a0 1 5OLOl 0d.O .o: do I , *.go%e*s 0 0.0 0.2 0.4 0.6 0.8 1 .o proximal distal Normalized Length

Fig. 6. Felis catus. The thickness of the soleus aponeurosis varies along its length. The two symbols denote the recorded thickness for two different cats (soleus muscle mass = 2.67 and 2.70 g). The abrupt fluctuations, particularly at the thick end, reflect collagen bundles crossing at a slightly oblique angle to the plane of section. 80 S.H. SCOTT AND G.E. LOEB

0 58 0 s9

V 0 s10

‘3 s11 i s12 B v 4 4 I 1 0.0 0.2 0.4 0.6 0.8 1 .o Force (Fo)

Fig. 7. Felis catus. The specific stiffness of the tendon relative to muscle force. The values for each muscle are identified by unique symbols. around 30 Fo/L, at forces greater than 0.3 Fo. displayed, but with corrections for the epimy- Estimates of stiffness were consistent be- sial movement (see Appendix). The stiffness tween trials as displayed by the small size of of the aponeurosis consistently increased with the SD bars in Figure 7 (vertical lines). Even fascicle length at low force levels. At 0.1 Fo, though specific stiffness is normalized to FO specific stiffness increased from 5 to 15 FdL, and L,, there was considerable variability in as fascicle length increased from 0.8 to 1.1 the mechanical properties between speci- LO.At 0.2 Fo, the stiffness of the aponeurosis mens; maximal values ranged from 25 to 40 became less dependent on fascicle length: Fo/L,. specific stiffness averaged around 15 Fo/L,, The specific stiffness of the distal portion but ranged from 7 to 30 Fo/L,. At 0.6 Fo, the of the aponeurosis was similar to that of the specific stiffness of the aponeurosis averaged tendon (Fig. 8). Specific stiffness increased 25 to 30 Fo/L,. As in the estimates of the with muscle force up to 0.4 Fo, whereas at specific stiffness of tendon, there was consid- higher force, its magnitude remained con- erable variability in the elastic properties of stant. The elastic properties were unaffected the aponeurosis between specimens even by the length of the fascicles except at 0.1 Fo, though stiffness was normalized to Fo and L,. where stiffness tended to increase with fas- However, note that there are no systematic cicle length. variations in the specific stiffness estimated Figure 9 displays the specific stiffness of for different portions of the aponeurosis. the aponeurosis plotted against fascicle length Figure 10A displays the relationship be- at three different levels of force, 0.1, 0.2 and tween tendon and aponeurosis stiffness in 0.6 Fo, estimated from all muscles studied. units of N/L, at 0.6 Fo of muscle force and at Values based on recordings that include the a fascicle length of 1.0 LO.The stiffness of the proximal portion of the aponeurosis are also entire aponeurosis, S,, was strongly corre- lated with the stiffness of the tendon, St (r = 0.96;P < 0.01). The correlation was not different from unity (P > 0.20) so that the stiffness of the aponeurosis and tendon ap- pear to be similar at least under moderate-to- high physiological loads. Figure 10B demon-

.-. - .- .- 1.04 strates the relationship between muscle force ] :/ ____--0.93 and the stiffness of the aponeurosis and ten- ...... 0.82 -0.71 don. Peak muscle force was found to be corre- lated with both the stiffness of the aponeurosis (r = 0.94; P < 0.02) and tendon (r = 0.88; 0.0 0.2 0.4 0.6 0.8 1.0 P < 0.05). Therefore, the stiffness of the con- Force (Fo) nective tissue was matched closely to the force capabilities of the muscle presumably Fig. 8. Felis cutus. The specific stiffness of the distal by modifying the amount of connectivetissue portion of the aponeurosis for contractions at four different muscle lengths. Stiffness increases with force up to rather than its specific material properties. 0.4 Fo and then remains constant at higher force levels. Surprisingly, when the stiffness estimates MECHANICAL PROPERTIES OF APONEUROSIS AND TENDON 81

A B C Stiffness at 0.1 FO Stiffness at 0.2 Fo Stiffness at 0.6 FO

Fascicle Length (LO)

Fig. 9. Felis cutus. The specific stiffness of the aponeu- denote estimates based on measurements of the proximal rosis relative to fascicle length at three levels of muscle and distal portions of the aponeurosis, respectively. Stiff- force, 0.1, 0.2, and 0.6 Fo, estimated from all muscles ness of the aponeurosis varies with fascicle length at studied. Circles denote estimates based on measure- 0.1 Fo. ments of the entire aponeurosis. Squares and triangle were normalized for peak muscle force there DISCUSSION was still a significant positive correlation be- Overlying epimysium alters the recorded tween muscle force and the specific stiffness length of the aponeurosis of the aponeurosis (r = 0.89; p < 0.05). In It has been claimed that the recorded length Other words, the stiffness Of the aponeurosis normalized to muscle force was greater for oftheaponeurotic sheet in rat gastrocnemius stronger muscles and suggests that the elas- and extensor longus muscles de- tic prope-ties ofthe aponeurosis overcornpen- pended n& only On force, but also on sated for the force-generating capacity of muscle length (HuiJing and Ettema, '88/ 89; these stronger muscles. This would also ex- Ettema and Huijing, '89). Our data demon- plain the large variability in the estimates of Strate a similar interaction between fascicle specific stiffness in Figure 9. There was a length and the ~~o~-dedlength ofthe aPoneu- similar trend between the specific stiffness of rosis but show further that this effect was the tendon and maximal muscle force, but isolated to the most proximal end of the this was not significantly different from zero aponeurosis. Concurrently, Zuurbier et al. (P > 0.10). ('94) have also shown similar results for the

Stiffness at 0.6 FOand 1.O LO A B

1200 S, = 0.88 S, + 172 1200 r = 0.96 - p -= 0.01 0 5.2 900 v) a, aponeurosis - - - % 600 S, = 63 Fo - 890 cn p c 0.02

300 300 300 600 900 1200 10 20 30 40 Aponeurosis stiffness (N/L) Maximal muscle force (N)

Fig. 10. Fdis catus. A: Relationship between tendon and aponeurosis stiffness for each muscle. Symbols for each muscle are the same as in Figure 5. B: Relationships between tendon (solid) and aponeurosis (dashed) stiffness and maximal active force, Fo, for each muscle. 82 S.H. SCOTT AND G.E. LOEB aponeurosis of the gastrocnemius medialis overestimated by up to 5%. Alternatively, (GM) muscle of Wistar rats. However, our stiffness measures have been based on the observations of the aponeurosis surface re- length of aponeurosis when muscle was vealed an overlying layer of epimysium along lengthened passively (Lieber et al., '91). Esti- its perimeter that changed shape and slid mates of the stiffness of aponeurosis were over the aponeurosis when fascicle length less than 1N/mm at low force levels. The low changed. We propose that this epimysial stiffness values were a consequence of the movement may be responsible for the previ- relatively flat relationship between passive ously hypothesized interdependence between muscle force and fascicle length. As a result, aponeurosis and fascicle length. This hypoth- the lengths used to estimate stiffness spanned esis appears to conflict with the study by large changes in fascicle length that resulted Zuurbier et al. in which they removed epimy- in large epimysial movement and thus artifac- sial tissue from the aponeurosis. However, tual length changes in the recorded length of they did not verify that the crossed-ply por- the aponeurosis. When the passive force- tion of the epimysium had been removed length curve increases rapidly at longer (C.J. Zuurbier, personal communication). lengths, the stiffness levels rise. The stiffness Therefore, future studies must verify whether of the aponeurosis could also be based on its epimysial collagen is responsible for the ap- length at the peak of isometric contractions parent length changes in the aponeurosis. recorded at different muscle lengths. This Estimates of the stiffness of aponeurosis approach estimated that aponeurosis stiff- could be adversely affected by movement of ness was around 5 N/mm at 15 N of force, the epimysium. Any technique that does not much lower than estimates from length compensate for epimysial movement will in- changes during a single contraction. correctly estimate the stiffness of tissue. To Not only did the epimysial movement cause illustrate this point, the stiffness of the apo- artifactual measurement of movement of the neurosis was estimated from records where aponeurotic sheet, but it also reduced the one piezoelectric crystal was attached to the recorded velocity of the fascicles (Scott and overlying epimysium (Fig. 11). The triangles Loeb, personal observation). In the present in Figure 11B show the stiffness of the apo- study, the crystal at the most proximal end of neurosis based on its length during a single the aponeurosis was used to defined the dis- contraction (dashed line in Fig. 11A) cor- tal end of the fascicle. The observed move- rected for the movement of the epimysium. ment of the epimysium covering the aponeu- The circles denote the stiffness estimated rosis suggests that fascicle velocity would be from the same single contraction, but with- underestimated by the ultrasonic technique. out correcting for epimysial movement. Epi- The percentage error can be computed from mysial movement during a single contraction the slope of the linear regressions relating appears to have caused the stiffness to be fascicle and aponeurosis length. These slopes

A B -16 .g 12 z 15 z z g0 10 a, LL cc '4 5 tj

0 0 3.2 3.3 3.4 3.5 3.6 3.7 0 5 10 15 20 Aponeurosis length (crn) Force (N) Fig. 11. Felis catus.-l!€&!!L A The relationship between estimates of the aponeurosis are dependent on the type of aponeurosis length and muscle force is complex when force-length relationship used. + ,X, same as in A, circles, length measures include epimysial movement: +, passive stiffness estimates based on dashed line in A, but with no force-length relationship; X, active force-length relation- adjustments for changes in fascicle length; triangles, ship; solid and dashed lines, force-length relationship estimates based on dashed line adjusted for the effects during contractions at four different lengths. B: Stiffness from the overlying epimysium (see text). MECHANICAL PROPERTIES OF APONEUROSIS AND TENDON 83 range from 0.05 to 0.10, suggesting that ap- The specific stiffnessof aponeurosis is proximately 5-10% of fascicle movement similar to tendon would be recorded as movement within the The length-tension relation for tendon can aponeurosis. Therefore, fascicle velocity be divided into two regions: a compliant toe would be 5-10% greater than recorded. region at low forces where tissue stiffness Widthchanges of the aponeurosis depend on increases with force, and a linear region at fascicle length higher force levels where stiffness remains The observed width of the aponeurosis in- constant (Ker, '81; Proske and Morgan, '87). creased when the fascicles shortened. The This relationship would be expected from increase in aponeurosis width was most likely tendon given the crimped organization of its larger than recorded using the piezoelectric constituent collagen fibers (Diamont et al., crystals because the crossed-ply network of '72). The nonlinear phase of the length- collagen within the epimysium would have tension relation is associated with straighten- tended to decrease, not increase the width of ing out the collagen crimps whereas the lin- the aponeurosis when the fascicles short- ear phase is associated with extension of the ened. The long, narrow dimensions of the collagen fibrils (Diamont et al., '72). aponeurotic sheet of cat soleus make it diffi- We obsehed that the organization of the cult to interpret quantitatively estimates of collagen in the aponeurosis was similar to its aponeurosis width. Therefore, the sign, but organization in the tendon, except that it was not the magnitude, of the width changes can spread across the surface of the muscle in be interpreted from the present study. A longitudinal bands. The aponeurosis had a more comprehensiverecording of the aponeu- gradient of thickness from its proximal edge rosis using the piezoelectric crystals would be to its junction with the tendon that appears best suited on a broader, larger muscle such to match the accumulation of active tension as the lateral or medial gastrocnemii of the that could be imposed on it by inserting cat. muscle fibers. This would account for the The increase in the width of the aponeuro- similar strains that we observed in proximal sis when the fascicles shortened was prob- and distal sections of aponeurosis. A more ably related to the increase in cross-sectional quantitative analysis of aponeurotic mor- area of the fibers beneath the aponeurosis. phometry would require correction for the The volume of a muscle fiber remains con- tapered and irregular cross section produced stant during changes in its length. When the by its banded structure and measurement of fiber shortens, the cross-sectional area of the the cross-sectional area of muscle fibers ter- fiber must increase (Elliott et al., '63; Trot- minating at various levels. Moreover, the spe- ter, '91). The observed increases in the width cific stiffness of the aponeurosis was similar of the aponeurosis suggest that the fibers to tendon, at least at moderate to high loads. expand laterally to help compensate for their This suggests that the specific stiffness along increased cross-sectional area. The diameter the entire length of the aponeurosis is con- of the fiber in the sagittal plane probably also stant and similar to that of the tendon. increases, even though no change in the The present estimates of the mechanical length of the aponeurosis was observed. The properties of the in-series connective tissue longitudinal projection of the muscle fiber on of cat soleus are similar to the results from the surface of the aponeurosis is proportional previous studies (Walsmley and Proske, '81; to the diameter of the fiber in the longitudi- Proske and Morgan, '84;Purslow, '89). Mea- nal plane and inversely related to the angle sures of stiffness in-series to the soleus fas- between the long axis of the fiber and the cicles by Rack and Westbury ('84) demon- surface of the aponeurosis. The length of the strated a gradual increase in stiffness with aponeurosis remains constant when the force; assuming Fo equalled 24 N, the specific muscle shortens, but the angle between the stiffness rose to around 30 Fo/L, at 0.3 Fo. At muscle fibers and the aponeurosis pennation higher force levels, Proske and Morgan ('84) angle is known to increase (Muhl, '82). There- recorded the stiffness of aponeurosis and ten- fore, the diameter of the fiber in the longitu- don based on quick releases. Their estimate dinal plane must increase. Presumably the of tendon and aponeurosis compliance was diameter of the fiber increases in all direc- 0.09 mm/N, which would also convert to tions simultaneously, with perhaps a small around 30 Fo/L,. These values are quite simi- change in its cross-sectional shape (Trotter, lar to our results based on direct measures of '91). the length of the aponeurosis and tendon 84 S.H. SCOTT AND G.E. LOEB using ultrasound techniques. The advantage eccentric function during locomotion, which of the present technique is that stiffness esti- may produce forces greater than Fo. mates were possible for all levels of muscle The stiffness of the aponeurosis (but not force. its length) appears to vary with fascicle The mechanical properties of tendon have length. Previous studies have suggested that been extensively studied in many muscles there is no change in the stiffness of the and species (e.g., Abrahams, '67; Ker, '81; entire series elastic component of muscle at Lieber et al., '91). Assuming a cross-sectional different muscle lengths (Morgan, '77; area of 1.8 mm2 (Rack and Westbury, ,841, Ettema and Huijing, '89). However, these the Young's modulus of the soleus tendon in studies concentrated on high force levels, the present study equaled 420 N/mm2 at where our results also showed no effect. higher loads. This is within the range of The variability in the stiffness of the apo- values recorded by other studies. Lieber et al. neurosis at low force levels is probably re- ('91) found values as low as 200 N/mm2 for lated to its interaction with the terminating frog semitendinosus tendon, whereas Ker muscle fibers. It has been suggested that the ('81) found stiffness estimates as high as stiffness of the myotendinous junction itself 1,600 N/mm2 for the plantaris tendon of may change with fascicle length (Proske and sheep. The variability between measure- Morgan, '87). However, the very small size of ments from different muscles and different this structure in comparison to the length of species is probably related to differences in the entire aponeurosis (microns compared to the actual properties of the connective tissue centimetres) suggests that the junction would (Elliott and Crawford, '65). The apparent have negligible affects on the total stiffness of over-design of the aponeurosis and perhaps the aponeurosis. It is more likely that the tendons of the largest muscles (Figs. 7, 10) change in stiffness was related to the change may reflect such differences arising from in the shape of the aponeurosis when the sexual dimorphism; cats S9 and S10 that muscle shortens. The crimped collagen in the dominate this relationship were both large aponeurosis is organized into longitudinal males. bands. The width of the aponeurosis in- Ker ('81) suggested that energy storage in creases when the muscle shortens. A portion tendon was largely due to the latter, linear of this lateral expansion may increase the phase associated with extension of the colla- width of the longitudinal collagen bands and gen fibers. This statement was based upon perhaps increase the crimp of the collagen. recordings of the length-tension relationship An increase in the crimp of the collagen at of plantaris tendon of sheep where the transi- shorter fascicle lengths would increase the tion between the two phases occurred at low range of the exponential spring behaviour, force levels (probably less than 0.1 Fo). In decreasing the apparent stiffness of the apo- contrast, Leiber et al. ('91) found that the neurosis. At longer muscle lengths, the crimp length-tension relationship of tendon was angle would be reduced and thus the stiffness never linear and its stiffness continuously of the aponeurosis would rise more rapidly increased up to Fo. Rack and Westbury ('84) with applied force and would increase over- found that stiffness continued to increase all. with force for loads up to around 0.4 Fo for a It is unlikely that the thin profile and sparse portion of soleus tendon. In the present study, organization of the collagen network in epi- stiffness increased with force for loads less mysium could transmit any significant ten- than 0.4 Fo and remained constant at greater sile load between muscle fascicles and adjoin- force levels. The large variability between the ing aponeurosis. More likely, the major role transition between the two phases of tissue of epimysium is to permit tendons and fascial stretch relative to the force generating capac- planes to slide with respect to each other, ity of each muscle suggests that energy stor- acting as a lubricated layer that prevents age in the tendons of some muscles or species adhesions from forming between such layers. may be attributable to straightening of the Could there be a mechanical role for epimy- crimped collagen as well as to extension of sium? Recently developed mathematical mod- the collagen. Moreover, these differences may els have suggested that certain muscle archi- be related to functional differences between tectures are unstable if superficial muscle muscles. For example, the robust properties fibers attach at an angle to the aponeurotic of the aponeurosis and tendon of the soleus sheet (van Leeuwen and Spoor, '92). How- muscle of the cat may be related to its normal ever, such theoretical models have not consid- MECHANICAL PROPERTIES OF APONEUROSIS AND TENDON 85 ered the potential stabilizing role of epimy- The shift in the aponeurosis length equals sium. It is possible that the mesh-like collagen the change in length of the epimysium, AL,. network in epimysium is involved in stabiliz- Since ALe is proportional to the length change ing the anatomical organization of the under- of the fascicles, ALf, the length of the epimy- lying muscle fascicles and aponeurosis. As a sium can be calculated using consequence, an architectural model of muscle would be mechanically stable only if ALf/L, = AL,/L, the overlying epimysial tissue were explicitly where Lo and L, are the lengths of the fascicle included in the model. and epimysium at optimal length, respectively. Subtraction of L, from the recorded ACKNOWLEDGMENTS length of the aponeurosis provides an esti- We appreciate the helpful comments of Dr. mate of the actual length of aponeurosis re- F.J.R. Richmond and the technical assis- corded by the piezoelectric crystals. These tance of J. Creasy, C. Simbirski and M. Yeoh. two procedures reduced the stiffness esti- This work was supported by the Medical Re- mates by a total of 5-10%, compared to unad- search Council of Canada. justed values.

APPENDIX LITERATURE CITED Removal of the effects from the overlying Abrahams, M. (1967) Mechanical behaviour of tendon in epimysium on records of the aponeurosis vitro. Med. Biol. Eng. 5433443. Alexander, R.McN., and A. Vernon (1975)The mechanics The contributions of the overlying epimy- of hopping by kangaroos (Macropodidae). J. Zool. 177: sium to the recorded length of the aponeuro- 265-303. Caputi, A.A., J.A. Hoffer, and I.E. Pose (1992) Technical sis can be estimated and corrected for. This note: velocity of ultrasound in active and passive cat correction is based on the assumption that medial gastrocnemius muscle. J. Biomech. 25r1067- the apparent shift in the aponeurosis length 1074. related to changes in the length of the fas- Cavagna, G.A. (1977) Storage and utilization of elastic energy in skeletal muscle. Exerc. Sport Sci. Rev. 589- cicles is due entirely to epimysial sliding. To 129. begin, the fascicles shortened slightly during Diamont, J., A. Keller, E. Baer, M. Litt, and R.G.C. the isometric contractions; the correspond- Arridge (1972) Collagen; ultrastructure and its rela- ing epimysial movement diminished the tion to mechanical properties as a function of ageing. Proc. R. Soc. Lond. 180:293-315. stretch imposed on the aponeurosis as muscle Drury, R.A.B., and E.A. Wallington (1980) Carleton's force increased. To remove the effect of epimy- Histological Technique. 5th Ed. Oxford; Oxford Univer- sial movement on measures of aponeurosis sity Press. length, a linear regression that related the Edman, K.A.P. (1988) Double-hyperbolic force-velocity relation in frog- muscle fibres. J. Physiol. (Lond.) 404: length of the fascicle and of the aponeurosis 301-321. was estimated at a muscle force equal to 0.5 Elek, J.,A. Prochazka, M. Hulliger, and S. Vincent (1990) Fo,one-half the maximal isometric force gen- In-series comuliance of eastrocnemius muscle in cat erated by the muscle (McCall, '86). The slopes step cycle: Do spindles sighorigin-to-insertion length? J. Physiol. (Lond.)4.29237-258. of these lines were always significantly differ- Elliott, D.H., and G.N.C. Crawford (1965) The thickness ent from zero (P < 0.001). The linear regres- and collagen content of tendon relative to the strength sions were used to adjust the recorded length and cross-sectional area of muscle. Proc. R. Soc. Lond. of the aponeurosis due to changes in fascicle 16.2137-146. Elliott, G.F., J. Lowry, and C.R. Worthingtonm (1963) length during each contraction. An X-ray and light-diffraction study of the filament The regression equation removed the ef- lattice of striated muscle in the living state and in rigor. fect of epimysial slip over the aponeurosis, J. Mol. Biol. 6r295-305. but it did not remove the absolute length of Ettema, G.J.C., and P.A. Huijing (1989) Properties of the tendinous structures and series elastic component of epimysium recorded in the measures of apo- EDL muscle-tendon complex of the rat. J. Biomech. neurosis length. The epimysium is not in 22:1209-1215. series with the muscle fibers and thus would Griffiths, R.I. (1991) Shortening of muscle fibres during not stretch when force was transmitted by stretch of the active cat medial gastrocnemius muscle: the role of tendon compliance. J. Physiol. (Lond.) 436: the aponeurosis. Therefore, aponeurosis 219-236. stretch would be underestimated because a Hatta, I., H. Sugi, andY.Tamura (1988) Stiffnesschanges portion of the recorded length did not contrib- in frog skeletal muscle during contraction recorded ute to the length change. The approximate using ultrasonic waves. J. Physiol. (Lond.) 403: 193- 209. length of epimysium recorded along with the Hoffer, J.A., A.A. Caputi, I.E. Pose, and R.I. Griffiths aponeurosis can be calculated by the ratio (1989) Roles of muscle activity and load on the relation- between epimysial and fascicle movement. ship between muscle spindle length and whole muscle 86 S.H. SCOTT AND G.E. LOEB length in the freely walking cat. Prog. Brain Res. Proske, U., and D.L. Morgan (1987) Tendon stiffness: 80.75-85. methods of measurement and significance for the con- Huijing, P.A., and G.J.C. Ettema (1988/89) Length char- trol of movement. A review. J. Biomech. 20r75-82. acteristics of aponeurosis in passive muscle and during Purslow, P.P. (1989) Strain-induced reorientation of an isometric and dynamic contractions of rat gastrocne- intramuscular connective tissue network implications mius muscle. Ada Morphol. Neer1.-Scand. 26.5142. for passive muscle elasticity. J. Biomech. 22r21-31. Ker, R.F. (1981) Dynamic tensile properties of the plan- Rack, P.M.H., and D.R. Westbury (1984) Elastic proper- taris tendon of sheep (OvisAries). J. Exp. Biol. 93.283- ties of the cat soleus tendon and their functional impor- tance. J. Physiol. (Lond.) 347r479-495. 302. Rowe, R.W.D. (1981) Morphology of perimysial and endo- Lieber, R.L., M.E. Leonard, C.G. Brown, and C.L. Trestik mysial connective tissue in skeletal muscle. Tissue Cell (1991) Frog semitendinosis tendon load-strain and 13r681-690. stress-strain properties during passive loading. Am. J. Rowe, R.W.D. (1985) The structure of the rat tail tendon. Physiol. 261:C86-C92. Connect. Tissue Res. 14.9-20. McCall, R.B. (1986) Fundamental Statistics for Behav- Trotter, J.A. (1991) Dynamic shape of tapered skeletal ioural Sciences. 4th Ed. San Diego; Harcourt Brace muscle fibres. J. Morphol. 207.211-223. Jovanovich. Van Leeuwen, J.L., and C.W. Spoor (1992) Modelling Morgan, D.L. (1977) Separation of active and passive mechanically stable muscle architectures. Phil. Trans. components of short-range stiffness of muscle. Am. J. R. SOC.Lond. B336.275-292. Physiol. 232:C45-C49. Walsmley, B., and U. Proske (1981) Comparison of stiff- Muhl. Z.F. (1982) Active leneth-tension relation and the ness of soleus and medial gastrocnemius muscles in effect of muscle pinnation on fibre lengthening. J. cats. J. Physiol. (Lond.)46r250-259. Morphol. 173285-292. Zuurbier, C.J., A.J. Everard, P. Van der Wees, and P.A. Proske, U., and D.L. Morgan (1984) Stiffness of cat Huijing (1994) Length-force characteristics of the apo- soleus muscle and tendon-during activation of part of neurosis in the passive and active muscle condition and muscle. J. Physiol. (Lond.) 52r459-468. in the isolated condition. J. Biomech. 27:445-453.