Effect of Rate of Distraction on Lossof Range of Joint Movement, Muscle

Stiffness, and IntramuscularConnective Tissue Content During

Surgical Limb-Lengthening:A Study in the Rabbit

PAMELA WILLIAMS,1* HAMISH SIMPSON,2 PETER KYBERD,2JOHN KENWRIGHT,2 AND GEOFFREY GOLDSPINK3

1Department of Biological Sciences, University of Hull, Hull HU6 7RX, England2Nuffield Orthopaedic Centre, Oxford OX3 7LD, England

3Department of Anatomy and Developmental Biology, Royal Free Hospital MedicalSchool, London NW3 2PF, England

ABSTRACTSurgical lengthening of limbs often results in loss of range of joint

movement and this has been shown to be associated with an increase inpassive tension and an increase in collagen content of the muscles. In thisstudy, we have investigated the length/tension properties and the connectivetissue component of muscle distracted at three different rates in order todetermine whether low rates of distraction would enable the connectivetissue component, as well as the contractile component (number of serialsarcomeres), to adapt more completely to the increased functional length ofthe muscle and thus lead to improved range of joint movement. It was foundthat loss of range of movement varied with rate of distraction. At the lowrate, there was no change in the passive tension or collagen contentcompared to muscles from sham-operated animals, and range of movementwas significantly greater than at the other rates. At the medium rate,although the muscles showed good adaptation in terms of serial sarcomerenumber, passive tension and collagen content was increased and range ofmovement reduced, indicating that changes in the connective tissue compo-nent are important factors in loss of joint movement. In the case of muscledistracted at a high rate, failure of the muscle fibres to add on sufficientsarcomeres, combined with changes in the connective tissue, resulted inalmost total loss of joint movement. Anat Rec 255:78–83, 1999.r 1999 Wiley-Liss, Inc.

Key words: passive tension; connective tissue; collagen; fibrosis; musclecontracture

Surgical lengthening of limbs using the Ilizarov method(Ilizarov, 1989) has become an important surgical means ofcorrecting discrepancies in limb length. However, theprocedure often results in soft tissue complications such asmuscle contracture (Paley, 1990; Fitch et al., 1993), thus,there is a need to optimise this method of distraction.Previous studies have shown that when the tibialis ante-rior muscle of the rabbit is progressively stretched duringsurgical limb distraction, the muscle fibres lengthen by

addition of new serial sarcomeres, provided that stretch iscarried out at an appropriate rate. This results in a shift in

Grant sponsor: Action Research.*Correspondence to: Dr. Pamela E. Williams, Department of Biologi-

cal Sciences, University of Hull, Cottingham Road, Hull HU6 7RX,UK. Fax 144 1482 465458; E-mail: p.e.williams@biosci.hull.uk

Received 20 November 1998; Accepted 4 January 1999

THE ANATOMICAL RECORD 255:78–83 (1999)

r 1999 WILEY-LISS, INC.

the active length/tension curve which should enable themuscle to function over the new length (Williams et al.,1993; Simpson et al., 1995). Above a certain rate ofdistraction, however, insufficient sarcomeres are added toallow adaptation to the imposed increase in length. Thisresults in over-stretching of the sarcomeres, muscle weak-ness, and loss of range of joint movement (Williams et al.,1995; Simpson et al., 1995).

In a recent study (Williams et al., 1998), muscles dis-tracted at a rate that allowed good adaptation in terms ofincrease in serial sarcomeres still showed loss of jointmovement. This was shown to be associated with in-creased passive tension in the muscles and increasedintramuscular connective tissue. An increase in intramus-cular connective tissue has been noted in other models ofdistraction (Lee et al., 1992; Gil-Albarova et al., 1997).Thus, it would appear that the parallel elastic componentof muscle does not adapt to prolonged stretch as well as thecontractile component. It seems likely, therefore, that thereduced range of joint movement noted both in animalmodels and in patients who have undergone limb distrac-tion is the result of changes in the connective tissuecomponent of the muscle.

The connective tissue framework of muscle consists offibrillar collagen which provides both myofibre to myofibreconnections (Borg and Caulfield, 1980) and a three-dimensional support structure with important elastic prop-erties (Trotter and Purslow, 1992). Thus, even smallchanges in the concentration of this element would beexpected to have a substantial effect on muscle function. Itwould appear that during prolonged stretch the connectivetissue elements remodel less readily than the contractilecomponent. This may lead to damage to the perimysial andendomysial network with subsequent fibrosis and reduc-tion in contractile function (Janiki, 1992). Alternatively, ithas been shown that distraction may be accompanied bymuscle inflammation and necrosis (Williams et al., 1994;Lee et al., 1993), and the fibrotic changes may be subse-quent to muscle fibre damage (Foidart et al., 1981).

It is possible that very low rates of distraction, at a ratejust above that at which the bone prematurely fuses,would enable the connective tissue, as well as the contrac-tile material, to adapt more completely to the increasedfunctional length of the muscle. In this study, therefore, wedistracted the anterior tibialis muscle of the rabbit atthree different rates. The aim was to determine to whatextent loss of range of joint movement was correlated withchanges in passive tension and connective tissue content,and whether a low distraction rate would result in normalmuscle compliance and range of joint movement.

Collagen is the major connective tissue protein in muscle.The tensile properties of passive muscles are primarilydependent on the amount and structure of intramuscularcollagen (Kovanen et al., 1984). For this reason, theendomysial and perimysial collagen content of the dis-tracted muscles was investigated using image analysis ofsections stained for collagen with Picrosirius Red. Colla-gen volume fraction determined in this way has beenshown to be closely related to the concentration of thecollagen-specific amino acid, hydroxyproline (Pickeringand Boughner, 1990). The main intramuscular collagen iscollagen type III and image analysis was therefore alsocarried out on sections which had been immunostainedusing monoclonal antibody to this collagen type.

In order to determine whether connective tissue changesare associated with muscle damage, muscles were exam-ined at the light and electron microscopic level. From theseexperiments, we hoped to gain information concerningwhich parameters are influenced by the progressive stretchthat is exerted by the lengthening of the bones not onlyduring limb-lengthening procedures but also during nor-mal growth.

METHODSAnimal Model

Adult New Zealand white rabbits were anaesthetised byintravenous injection of Hypnorm (0.4 mg kg21) andMedazolam (0.2 mg kg21). The muscles in the lower legs ofthree animals in three groups were subjected to distrac-tion by applying an external fixation device (OrthofixM100) to the medial surface of the tibia. A mid-diaphysealosteotomy was created, the periosteum was repaired, andthe skin was closed (Simpson et al., 1995). The animalswere able to bear weight freely upon the operated legwithin two hr of operation. Food and water were availablead libitum and a cycle of 12 hr light/12 hr dark was used.Lengthening was carried out at 0.8, 1.6, or 2.7 mmd21,daily increases being divided into two increments. Length-ening was continued until a 20% increase of the initiallength of the tibia had been achieved. Contralateral musclesacted as controls; however, in order to distinguish betweenchanges produced by the process of distraction and thoseresulting from the presence of the distractor itself, a shamoperation was carried out in a separate group of threerabbits. In these animals, the external fixation system wasapplied and the osteotomy was created, but no lengtheningwas affected.

Animals were anaesthetised by intravenous injection ofHypnorm (0.4 mg kg21) and Medazolam (0.2 mg kg21), andmeasurements of function were made on the tibialisanterior. Results were compared using Students t-test orMann Whitney U.

Measurement of Range of Joint Motionand Muscle Length

The range of ankle joint movement (ROM) was mea-sured using a goniometer. The anterior tibialis muscleswere exposed and ink marks made on the distal end of thepatellar tendon attachment and on the furthest point ofinsertion of muscle fibres into the distal tendon. Usingthese fixed anatomical landmarks, muscle length wasmeasured with the ankle at the extremes of the movementrange as well as with the ankle at 90°.

Measurement of Passive Length/Tension CurvesThe tendon of the anterior tibialis was cut distally and

the muscle belly dissected free up to its origin, which wasleft undisturbed. The tendon was then attached with asuture to a lever arm of a recording device that measuredthe passive force generated by the muscle (CambridgeTensiometer). The muscle was kept moist with a salinedrip and maintained at 35°C using a thermistor-controlledlamp. Using a vernier calliper and screw carriage mecha-nism, the belly of tibialis anterior was set to variouslengths between the extremes of its range (as measuredbefore cutting the tendon). The passive force developed at

79MUSCLE STIFFNESS RESULTING FROM DISTRACTION

each muscle length was measured and the length/tensioncurve plotted.

Serial Sarcomere NumberFollowing the physiological analysis, the animals were

killed with an overdose of anaesthetic and the anteriortibialis muscles were removed. A strip of muscle fibresinserting furthest onto the distal tendon was dissectedfree, pinned to cork, then fixed overnight in 4% formalinmade up in 0.9% NaCl before being digested in 30% nitricacid for 48 hr and stored in 50% phosphate bufferedglycerol (pH 7.0). Small bundles of three to five fibres weredissected out under a dissecting microscope.

The distances between the proximal and distal myoten-dinous junctions were measured. Single fibres were thenmounted in glycerol jelly and viewed under a microscope.At approximately 45 points along each fibre, the lengthcovered by 10 sarcomeres was measured using an imageanalyser (Seescan Solitaire). The total number of sarco-meres in series along the length of each fibre was thencalculated from the mean sarcomere length and musclebundle length. Previous work (Williams and Goldspink,1985), using a control group of three rabbits, has shownthat when sarcomere number measurements are made asdescribed, the between-animal variation is low (coefficientof variation 5 4.9%).

Connective Tissue Analysis

Collagen content. Muscle samples, taken from themid-belly region of each muscle, were frozen in isopentanepre-cooled in liquid nitrogen and sectioned at 10µ using acryostat. Total amount and distribution of collagen weredetermined by image analysis on sections stained withPicrosirius Red (Sweat et al., 1964). Recent work hasshown that modification of the Sirius Red stain by theaddition of Fast Blue prevents any uptake of the SiriusRed by the cytoplasm, thus facilitating image analysis byproviding good contrast between the connective tissue andintracellular elements (unpublished observations). Slideswere fixed in a solution of 10% formalin in 0.9% NaCl for15 to 20 minutes before being rinsed with distilled water.They were then placed in Fast Blue RR solution (0.1% FastBlue RR, 7 mM magnesium sulphate, and 60 mM magne-sium borate) at room temperature for 30 min. Slides werethen washed thoroughly in distilled water and placed in a

solution of Sirius Red (0.1% in saturated aqueous picricacid) for 10 min before being dehydrated, cleared, andmounted. Sections were viewed using light microscopy.The microscope image was detected by a video camera andtransduced to a Seescan Solitaire image analysis systemby which the image was digitised. An image outside thetissue was recorded as a reference for shade correction ofeach section. The threshold was adjusted so as to give thebest correspondence between the video and the thresh-olded image. The sections were scanned using a samplingarea of 0.2 mm2 (excluding dense bands of perimysium)and the area occupied by connective tissue, i.e., thenumber of highlighted pixels, determined. The percentagearea within each section that was occupied by connectivetissue was then calculated. Results were compared usingthe Mann Whitney U test.

Analysis of collagen type III.Using frozen sections,endogenous peroxidase was quenched with 1.2% hydrogenperoxide at room temperature for 30 min. Followingblocking with normal rabbit serum, sections were incu-bated with primary antibody. The sections were thenincubated with biotinylated rabbit anti-mouse immuno-globulins antibody (1:300) (DAKO Ltd) at room tempera-ture for 30 min, followed by streptavidin complexed withbiotinylated peroxidases (DAKO Ltd) at room temperaturefor 30 min. The peroxidase complexes were visualisedusing Diaminobenzidine (DAB); washing and dilutionbuffers were either Tris-buffered saline (TBS) or Tris. Thismethod of visualisation has been shown to give strongerpositive staining without increased background when com-pared to the indirect peroxidase method, thus providinggreater contrast for image analysis (Lowry et al., 1997).Sections were scanned using the Seescan image analyserand the relative area occupied by collagen III determined.

Muscle Damage

Sections were stained with haemotoxylin and eosin forgeneral histology. The percentage of damaged fibres wasdetermined by counting a minimum of 300 fibres from eachsection in several different microscope fields across thewhole area of the section. A fibre was recorded as damagedwhen it showed one or more of the following features: apronounced shift from a polygonal to a rounded shape;hypereosinophilia; invasion of mononuclear cells; hyalini-

TABLE 1. Changes in range of joint movement, sarcomere number,and passive tension in distracted and sham-operated muscles

Rate of distractionSham

mean 6 se0.8 6 0.1 mm d21

mean 6 se1.6 mm d21

mean 6 se2.7 6 0.2 mm d21

mean 6 se

Loss of joint movement (%) 6 6 1 34 6 8 64 6 11* 89 6 7*Increase in muscle belly length at

90° ankle position (mm) 21 6 1 13 6 1 13 6 2 12 6 1Change in serial sarcomere

number (%) 1.0 6 3.9 [19.0 6 1.7 17.6 6 3.2] 13.23 6 1.49*Passive tension at the 90° ankle

position (Newtons) 0.0 1.6 6 0.9 [8.4 6 3.8 7.2 6 1.4]*Percentage of damaged fibres ,1 [2.3 6 1.2 4.3 6 2.9] 10.8 6 3.1*Area stained by Sirius Red (%) 12.1 6 2.5 14.8 6 1.6 [19.3 6 1.5 18.9 6 2.6]*Area stained by antibody to col-

lagen III (%) 10.0 6 2.0 9.4 6 1.6 [16.9 6 1.7 17.8 6 1.3]*

*Significantly different from the low rate P , 0.05. Results in brackets not significantly differentfrom each other.

80 WILLIAMS ET AL.

sation; vacuolation; disruption of the cytoplasm; internallyplaced myonuclei (Cumming et al., 1994).

RESULTSRange of Joint Movement (ROM)and Muscle Length

In the sham-operated animals the mean loss of ROMwas only 6% (Table 1). All the distracted limbs showed asignificant loss of ankle joint movement compared withthose of these sham-operated animals. Limbs distracted atthe highest rate showed the greatest loss of ROM. Measure-ment of muscle belly length at the 90° ankle positionshowed that all distracted muscles were longer than thecontralateral controls (mean increase 13 mm) with nosignificant difference among the rates.

Increase in Serial Sarcomeres and Changesin Length/Tension Parameters

All distracted muscles showed an increase in serialsarcomeres when compared with control muscles. How-ever, the amount of new contractile material added dif-fered among muscles distracted at the different rates. Atthe two lower rates, a similar number of sarcomeres wereadded (19.0 6 1.47% and 17.6 6 3.2%) whereas at thehighest rate, significantly fewer sarcomeres were added(13.25 6 1.49). In this latter case, therefore, it wouldappear that increase in length of existing sarcomeres, i.e.,decreased overlap of the actin and myosin filaments, wouldaccount for some of the observed increase in muscle length.

Passive Length/Tension PropertiesAt the 90° limb position, contralateral muscles and

muscles distracted at the lowest rate exerted very littlepassive tension (Table 1 and Fig. 1) (0.0 N and 1.6 6 0.9 Nrespectively). In comparison, the passive tension exertedby muscles distracted at the two higher rates was found tobe greatly increased (8.4 6 3.8 N and 7.2 6 1.4 N).

Connective Tissue ContentImage analysis of collagen content using sections stained

with Picrosirius Red showed that there was no differencein perimysial and endomysial connective tissue volumefraction in the muscles distracted at the lowest ratecompared to the sham-operated controls (Table 1 and Fig.2). There was, however, a significant increase in connectivetissue in the muscles distracted at the two higher rates.Collagen type III staining showed a similar pattern.

Muscle DamageSections stained with haemotoxylin and eosin showed

that there was a small but significant increase in damagedfibres in muscles distracted at the lowest rate compared tosham-operated controls. The number of damaged fibreswas found to increase with rate of distraction. At the highrate, many of the muscles showed considerable variabilityin muscle fibre size, internally located nuclei and evidenceof fibre splitting. There was scattered necrosis, the necroticfibres being infiltrated with and surrounded by macro-phages.

DISCUSSIONSkeletal muscle is a tissue which is able to adapt to

changes in function. It has been shown that even adult

muscle is able to adapt to an increase or decrease infunctional length by the addition or removal of serialsarcomeres (Tabary et al., 1972; Williams and Goldspink,1985b). However, although surgical lengthening of thelimbs in patients has been shown often to result in musclecontractures (Paley et al., 1991) and loss of range of anklemovement (Sofield et al., 1958), little is known about theresponse of muscle to progressive stretch. The resultspresented here, concerned with adaptation to continued

Fig. 1. Passive tension exerted by muscles distracted at differentrates. Arrows mark muscle length at the 90° ankle position. ---r---contralateral muscles; ——j—— distracted muscles.

81MUSCLE STIFFNESS RESULTING FROM DISTRACTION

lengthening, are important for understanding both thetype of cellular adaptation that is required to optimiselimb lengthening procedures and also the mechanismsinvolved in normal muscle fibre growth.

All distracted limbs showed some loss of joint move-ment. However, this was found to increase dramaticallywith rate of distraction; muscles distracted at the highestrate showed almost a 90% loss of movement. Muscles fromthe sham-operated animals did not show comparablechanges, demonstrating that loss of movement was associ-ated mainly with limb lengthening rather than the opera-tive procedure or the presence of the distracting device.

All distracted muscles showed a similar increase inmuscle length when measured at the 90° limb position.However, the number of serial sarcomeres added variedaccording to rate of distraction. There was no significantdifference between the number of new sarcomeres addedat the low and medium rates, but at the high rate, thenumber was significantly less. This is in agreement withprevious studies, which indicated that above a certain rateof distraction, insufficient sarcomeres are added to allowadaptation to the imposed increase in length (Simpson etal., 1995).

It has been shown previously that the passive tensiondeveloped by contralateral and sham-operated control

anterior tibialis muscles, when held at the length whichrepresents the 90° limb position, is extremely low (,0.5 N)(Williams et al., 1998). Similarly, in this current study,muscles distracted at the lowest rate developed only verylow passive force (1.6 N). The passive force was, however,greatly increased at both medium and high rates with nosignificant difference between the two (8.4 and 7.2 N,respectively). Because it was found that sarcomere addi-tion was equally great at the medium and low rates ofdistraction, it would appear that the increased passiveforce and greater loss of movement in muscles distractedat this rate compared to the low rate is not due toinsufficient addition of new sarcomeres. Results for theconnective tissue analysis paralleled those for passivetension, i.e., collagen content was increased in musclesdistracted at the medium compared to the low rate. Thus,it would seem likely that the increase in passive force andloss of movement is related to an increase in the intramus-cular connective tissue component.

In muscle distracted at the high rate, ROM was found tobe greatly reduced, to approximately 10% of the normalrange. However, the muscles were found to develop similarpassive tension and have the same connective tissuecontent as those distracted at the medium rate. In thiscase, sarcomere number was significantly reduced com-

Fig. 2. Connective tissue in muscle distracted at 0.8 mm d21 (A) and 2.7 mm d21 (B). Frozen sectionsstained with Picrosirius Red. Scale bar 5 20µ.

82 WILLIAMS ET AL.

pared to those distracted at the medium or low rate. Thus,it would seem that at this high rate, the extreme loss ofROM results from a combination of an increase in endo-mysial and perimysial connective tissue content and fail-ure of the muscle fibres to add on sufficient sarcomeres.

One of the objectives of this study was to determinewhether the increase in connective tissue occurred inresponse to myocyte damage and necrosis or whether thefibrotic changes occurred independently. When assessingthe degree of damage to muscle fibres it is important torecognise the limitations of the method used. The methodemployed in this study was to use the mid-region of themuscle belly and to determine the percentage of fibresshowing morphological evidence of damage in a singlecross-section. Obviously, if only small sections along eachfibre were affected, this would be an underestimate of thetotal number of fibres showing some damage. In addition,it is possible that the stretch could cause more damage tothe ends of the fibres near the attachment to the tendonswhere the formation of new sarcomeres takes place (Wil-liams and Goldspink, 1971). Thus, this method can only beused to make measurements of the relative damage tomuscle fibres in the belly region of muscles distracted atdifferent rates. It was found that while connective tissuewas increased at the medium rate, the number of damagedfibres was low. Indeed, the number was not significantlydifferent from that in muscles distracted at the low ratewhere no connective tissue changes were found. It seemslikely, therefore, that the increase in connective tissuerepresents a direct response to the muscle stretch. How-ever, at the highest rate of distraction, in addition to thegenerally high levels of collagen, pockets of dense connec-tive tissue were noted where it appears that myocyteshave been replaced with fibrotic tissue. Similar observa-tions were made by Yong-Lee et al. (1993) in a study ofchanges in the gastrocnemius muscle of the rabbit follow-ing distraction of the tibia. Although this study has notbeen able to totally dissociate a fibrotic response frommyocyte necrosis, it seems likely that where distraction iscarried out at a rate that allows good adaptation in termsof sarcomere number, connective tissue accumulation oc-curs independently of myocyte loss.

This study demonstrates that the intramuscular connec-tive tissue framework does not readily adapt to passivestretch beyond the normal physiological range, but at avery low rate of distraction, muscle shows good adaptationof both the contractile and the connective tissue elements.However, if the rate is too slow, the bone will heal and theosteotomy procedure will need to be repeated. In thisstudy, the lengthening of the limbs was carried out bymanually adjusting the distractors twice daily. We are nowdeveloping electronic linear motor distractors which willbe capable of producing continuous length changes andwhich may result in better adaptation for a given dailyrate of distraction. Further work is needed to determinethe nature of the link between the physical stretch signaland the regulation of expression of the genes for thedifferent collagen types of the intramuscular connectivetissue matrix.

LITERATURE CITEDBorg TK, Caulfield JB. 1980. Morphology of connective tissue in

skeletal muscle. Tissue & Cell 12:197–207.

Cumming WJK, Fulthorpe J, Hudgson, P Mahon M. 1994. Colour atlasof muscle pathology. London: Mosby-Wolfe. pp 44–60.

Fitch RD, Thompson JG, Rizk WS, Seaber AV, Garrett WE. 1993. Theeffects of the Ilizarov distraction technique on bone and muscle in acanine model: a preliminary report. Iowa Orthop J 16:10–19.

Foidart M, Foidart J, Engel WK. 1981. Collagen localisation in normaland fibrotic human skeletal muscle. Arch Neurol 38:152–157.

Gil-Albarova J, Melgosa M, Gil-Albarova O, Canadell J. 1997. Softtissue behaviour during limb-lengthening: an experimental study inlambs. J Paed Orthop 6:266–273.

Ilizarov GA. 1989. The tension-stress effect on the genesis and growthof tissues: Part I—The influence of stability of fixation and soft-tissue preservation. Clin Orthop & Rel Res 238:249–280.

Janiki JS. 1992. Myocardial remodelling and left ventricular diastolicfunction. Braz J Med Biol Res 25:975–982.

Kovanen V, Suominin H, Heikkinen E. 1984. Mechanical properties offast and slow skeletal muscles with special reference to collagen andendurance training. J Biomech 409:438–447.

Lee DY, Choi IH, Chung CY, Chung PH, Chi JG, Suh YL. 1993. Effectof tibial lengthening on the gastrocnemius muscle. A histopathologicand morphometric study in rabbits. Acta Orthop Scand 64:688–682.

Lowry A, Wilcox D, Masson E, Williams PE. 1997. Immunocytochemi-cal methods for semiquantitative analysis of collagen content inhuman peripheral nerve. J Anat 191:367–375.

Paley D. 1990. Problems, obstacles and complications of limb lengthen-ing by the Ilizarov technique. Clin Orthop 250:81–104.

Paley D, Rumley TO, Kovelman H. 1991. The Ilizarov technique: amethod to regenerate bone and soft tissue. Adv in Plastic andReconstructive Surg 7:1–43.

Pickering JG, Boughner DR. 1990. Fibrosis in the transplanted heartand its relation to donor ischemic time: assessment with polarizedlight microscopy and digital image analysis. Circ 81:949–958.

Simpson AHRW, Williams PE, Kyberd P, Goldspink G, Kenwright J.1995. The response of the rabbit anterior tibialis muscle to stretchduring surgical limb lengthening. J Bone & Joint Surg 77-B:630–636.

Solfield HA, Blair SJ, Millar EA. 1958. Leg lengthening: a personalfollow-up of forty patients some twenty years after the operation. JBone & Joint Surg (Am) 40:311–322.

Sweat F, Pulcher H, Rosenthall SI. 1964. Sirius Red F3BA as a stainfor connective tissue. Arch Path 78:69–72.

Tabary JC, Tabary G, Tardieu C, Tardieu G, Goldspink G. 1972.Physiological and structural changes in the cat soleus muscle due toimmobilisation at different lengths by plaster cast. J Physiol 244:231–244.

Trotter JA, Purslow PP. 1992. Functional morphology of the endomy-sium in series-fibred muscles. J Morphol 212:109–122.

Williams PE, Goldspink G. 1971. Longitudinal growth of skeletalmuscle fibres. J Cell Sci 9:751–767.

Williams PE, Goldspink G. 1985a. Effect of stretch combined withelectrical stimulation on the type of sarcomeres produced at theends of the muscle fibres. Exp Neurol 93:500–509.

Williams PE, Goldspink G. 1985b. Changes in sarcomere length andphysiological properties in immobilised muscle. J Anat 127:459–468.

Williams PE, Simpson AHRW, Kyberd P, Goldspink G, Kenwright J.1993. Changes in length/tension properties and fatigue resistance indistracted muscles. J Anat 183:184.

Williams PE, Simpson AHRW, Kyberd P, Kenwright J, Goldspink G.1994. Structural changes in the rabbit Tibialis Anterior musclefollowing distraction of the tibia. Basic & Applied Myology 4:431–437.

Williams PE, Simpson AHRW, Kenwright J, Goldspink G. 1995. Theinfluence of passive stretch on muscle fibre phenotype and contrac-tile characteristics of the rabbit tibialis anterior. J Physiol 479P:6–7.

Williams PE, Simpson AHRW, Kyberd P, Goldspink G, Kenwright J.1998. The morphological basis of increased stiffness of rabbit tibialisanterior muscles during surgical limb-lengthening. J Anat 193:131–138.

83MUSCLE STIFFNESS RESULTING FROM DISTRACTION