inﬂuence of the dynamometer and knee joint axis of ...€¦ · isokinetics and exercise science...

Isokinetics and Exercise Science 13 (2005) 3–4 3IOS Press

Abstract

Influence of the dynamometer and knee jointaxis of rotation alignment on the isokineticmeasurements

J.L. Croisier, P. Ciavatta and B. ForthommeDepartment of Physical Medicine and Rehabilitation, University of Liege, BelgiumE-mail: [email protected]

Objectives: Factors that could influence accuracy ofisokinetic assessment deserve careful attention [1,2].In order to develop accurate and standardized test pro-tocols, many investigations have focused for instanceon gravity correction, visual feedback, range of motion,resistance pad position. The aim of our study was todetermine the consequences of an inappropriate align-ment between the dynamometer and knee joint axis ofrotation on isokinetic measurements.Methods: Twelve healthy male subjects (24± 2 yearsold) participated in one session of testing (Cybex Normdynamometer) on their dominant leg. Four conditionsof proper or intentionally altered (in the horizontalplane) alignment were randomly proposed. Based onproper alignment with the knee flexed at 90, the al-tered alignments were designed as follows: joint axisof rotation placed 6 cm (−6) and 12 cm (−12) behindor 6 cm (+ 6) in front of the dynamometer axis of rota-tion in the horizontal plane (Fig. 1). Knee flexors andextensors performed 3 maximal repetitions for testingat 60/s in the concentric mode throughout a constant100 range of motion. Subjects benefited from 3 pre-liminary submaximal trials for familiarization beforethe successive testing positions. Descriptive statisticsand ANOVA tests were used to analyze the data.Results: For both knee flexors (Fl) and quadriceps (Q),measurements in the properlyaligned and+ 6 positionsshowed no significant difference in peak torque (PT)(Fig. 2). By contrast, the PT corresponding to the joint

- 12 - 6 Aligned + 6

Fig. 1. Representation of proper and altered alignments between theknee joint and dynamometer axis of rotation.

axis of rotation placed behind the device axis of rotation(−6 and−12) were significantly (p < 0.05) decreasedfor both muscle groups, most notably on the quadricepsin the−12 position (−14% for that position comparedto the properly aligned).

The same observations characterized the work pa-rameter, even though the Fl/Q ratio calculated fromPT was not significantly influenced by the alignment

ISSN 0959-3020/05/$17.00 2005 – IOS Press and the authors. All rights reserved

4 J.L. Croisier et al. / Influence of the dynamometer and knee joint axis of rotation alignment on the isokinetic measurements

0

50

100

150

200

250

300

Aligned -6 -12 +6 Aligned -6 -12 +6

PT

(N

.m)

a b b a a a b b

Extensors Flexors

Fig. 2. Concentric peak torques (PT in N.m) developed by knee extensors and flexors at 60/s: comparative study between proper or alteredalignments. Different letters represent statistically significant differences (p < 0.05).

modalities.Conclusion: We highlighted the influence of an im-proper alignment between the knee joint and dy-namometer axis of rotation, particularly when the jointis positioned behind the dynamometer motor in the hor-izontal plane. A lack of precision on that point wouldintroduce a factor confounding accurate isokinetic as-sessment.

References

[1] Z. Dvir, Isokinetics. Muscle testing, interpretation and clinicalapplications, (second edition), Churchill Livingston, 2004.

[2] D.H. Perrin,Isokinetic exercise and assessment, Human Kinet-ics Publisher, 1993.


Activation EMG pattern during isokineticknee flexion-extension assessment:Comparison between healthy subjects andchronic pain patients

D. Maquet, C. Demoulin, B. Forthomme, J.M. Crielaard and J.L. CroisierDepartment of Physical Medicine and Rehabilitation, University of Liege, BelgiumISEPK, B21, Allee des Sports 4, 4000 Liege, Belgium

1. Introduction

Isokinetic testing has been increasingly used to as-sess muscle performances in patients suffering fromchronic pain [7,11,13,14]. Surface electromyography(EMG) is a common scientific tool allowing an objec-tive assessment of muscular fatigability [1,3,15,17,18,26]. Impaired maximal isometric and isokinetic mus-cle performances have been reported in several stud-ies dealing with fibromyalgia and chronic low backpain patients (FM and CLBP patients) [7,10,11,13,20].The respective influence of physiological, psycholog-ical and deconditioning factors still remains contro-versial [16]. EMG pattern activity during isokineticefforts appears little documented [12,24]. Accordingto Robertson et al., isokinetic torque values appearedcorrelated to the agonist muscle activation. Tesch etal. [24] described an increase of vastus lateralis (VL)and rectus femoris (RF) integrated EMG activity duringan endurance knee extension test performed in concen-tric mode at 180/s. EMG has been extensively used toexplore muscle function in patients with chronic pain.Nevertheless, to our knowledge, no study has yet com-pared FM and CLBP patients with regard to muscleperformances and EMG patterns.

Hence the aims of this study were as follows:

1. evaluate the isokinetic concentric strength and fa-tigue of knee flexors and extensors in sedentary

healthy women and in women suffering from fi-bromyalgia or CLBP,

2. measure in these populations the EMG activitypatterns during the isokinetic strength and fatigueassessments.

2. Material and methods

2.1. Populations

All subjects received a complete explanation of theinvestigation purposes and gave their consent. Thisstudy was approved by the Ethical Committee of theUniversity of Liege Medical Center.

2.1.1. Control groupThe control subjects were fifteen women (45± 10

years old, 66± 11 kg) without joint or muscle injury.They performed on average 0.7 hour of recreationalsport activities per week.

2.1.2. FM patients groupTen women (50± 5 years old, 71± 11 kg) meeting

American College of Rheumatology Criteria for FMsyndrome were included. The duration of symptomsranged from 1 to 6 years and 3 patients were in paidjobs. The mean time spent in physical activities onlyreached 0.5 hour per week on average.


6 D. Maquet et al. / Activation EMG pattern during isokinetic knee flexion-extension assessment

Table 1NPT (N.m/kg), CW (J) and NCW (J/kg) developed by knee flexors and extensors (means andSD): comparison between control subjects, FM patients and CLBP patients

Variables Test modalities Control group FM group CLBP group

Strength measurementsFLENPT (N.m/kg) C60/s 0.97 (0.27)a 0.64 (0.21)b 0.88 (0.25)a

C180/s 0.71 (0.18)a 0.49 (0.15)b 0.7(0.14)a

EXTNPT (N.m/kg) C60/s 1.77 (0.37)a 1.34 (0.23)b 1.68 (0.37)a

C180/s 1.07 (0.2)a 0.86 (0.2)b 1.14 (0.28)a

Fatigue measurementsFLE + EXTCW (J) C180/s 3058 (839)a 2278 (394)b 2755 (564)a

NCW (J/kg) C180/s 47 (12)a 33 (7)b 42 (9)a

Different letters represent a significant difference (p < 0.05); C, concentric.

2.1.3. Chronic low back pain patients groupTen women (45± 10 years old, 66± 6 kg) suffering

from chronic low back pain (without pain irradiationin the leg or diffuse pain history) participated in thisstudy. The symptom duration ranged from 1 to 5 years.Only 4 patients were in paid jobs and the mean timespent in physical activities was on average 0.6 hour perweek.

2.2. Methods

After a warming-up consisting in 5 minutes of un-load stationary cycling at 50 rotations per minute (rpm)followed by static stretching of quadriceps and ham-strings, subjects performed a concentric isokinetic teston their dominant leg by means of a Cybex Normdynamometer (Henley Healthcare, Sugerland, Texas,USA). The range of motion during testing consisted in100 degrees of flexion from the full active (0) exten-sion. Each subject was seated with the trunk and thedominant leg fixed. The dynamometer axis of rotationwas aligned with the knee center of rotation (in 90

flexed position).Submaximal efforts of knee flexors and extensors (at

120/s angular velocity) was allowed in order to famil-iarize with the isokinetic exercises. Three preliminarysubmaximal repetitions preceded each test speed.

The testing protocol started with the strength assess-ment by means of 3 maximal repetitions at 60/s and 5maximal repetitions at 180/s. The analysis was basedon peak torques (PT in N.m) and bodyweight normal-ized peak-torques (NPT in N.m/kg). Thereafter, sub-jects performed the fatigue assessment consisting in 30maximal-intensity knee flexions and extensions at theangular velocity of 180/s. Fatigue was expressed bythe cumulative work (CW in J) and the bodyweight nor-

malized cumulative work (NCW in J/kg) determinedby summing up the work developed by both musclegroups.

EMG activity was recorded from the vastus medialis(VM), rectus femoris (RF), internal hamstring (IH) andexternal hamstring (EH). The ground reference elec-trode was secured over the tibia. Prior to electrodeplacement, the skin was vigorously cleaned with alco-hol wipes. In order to locate electrodes (silver/silverchloride surface) accurately, we identified muscle areasthanks to a maximal voluntary isometric effort from theseated (VM and RF) and prone (IH and EH) positions.The distance between the centers of the recording elec-trodes was 20 mm. EMG data and torque were sam-pled at 1000 Hz and recorded using a Noraxon Myosys-tem Software. All EMG signals were rectified andsmoothed (RMS 50). The average root mean square(RMS) was used as a measure of muscular activity. Ag-onist and antagonist EMG activities (e.g. flexor mus-cles recruitment respectively during knee flexion andextension) were expressed in absolute (µV) and peaktorque normalized values (µV/N.m).

Immediately after the isokinetic assessment, partici-pants completed a 10 cm visual analogue scale (VAS)for pain intensity estimate: the distance was scoredfrom 0 to 10 arbitrary units (a.u.), 0 a.u. correspondingto the absence of pain and 10 a.u. to the maximumpain.

2.2.1. Data analysisAn analysis of variance (ANOVA) was used for each

variable in order to examine the difference betweenthe 3 populations. Student’s paired t-test allowed tocompare results recorded at both speeds. An ANOVAtest was used to examine fatigue effects on agonistand antagonist EMG activities of each muscle groupthroughout the thirty repetitions.

D. Maquet et al. / Activation EMG pattern during isokinetic knee flexion-extension assessment 7

Table 2Agonist electrical activities of VM, RF, IH and EH recorded during the isokineticstrength assessment (means and SD): comparison between control subjects, FM patientsand CLBP patients

Muscles EMG activities Control group FM group CLBP group

VMC 60/s Absolute (µV) 177 (58)a 172 (86)a 169 (111)a

Normalized (µV/N.m) 1.61 (0.69)a 1.83 (0.79)a 1.5 (0.75)a

C 180/s Absolute (µV) 167 (56)a 180 (72)a 171 (104)a


RFC 60/s Absolute (µV) 158 (48)a 135 (57)a 132 (75)a




IHC 60/s Absolute (µV) 228 (108)a 155 (63)a 167 (82)a




EHC 60/s Absolute (µV) 177 (83)a 202 (94)a 177 (104)a



Normalized (µV/N.m) 3.51 (1.43)a 6.23 (2.88)b 3.86 (2.14)a

Different letters represent a significant difference (p < 0.05); C, concentric.

3. Results

3.1. Isokinetic strength and fatigue

The PT, NPT, CW and NCW parameters were sig-nificantly lower in FM patients comparatively to thecontrol group but also to the patients suffering fromchronic low back pain. Impairment affected more kneeflexors than knee extensors (Table 1).

By contrast, the isokinetic muscle profile did notdiffer significantly between the control group and thelow back pain patients

3.2. EMG pattern during isokinetic strengthassessment

The agonist electrical activity (expressed in absoluteand normalized values) of VM, RF, IH and EH wasrecorded during the isokinetic strength assessment at60/s and 180/s angular velocities. No significantdifference was observed between the control group andboth groups of chronic pain patients (Table 2).

We determined a specific EMG pattern for knee ex-tensors and flexors, based on EMG activities recordedrespectively from VM and RF and from IH and EH.The agonist recruitment of both muscle groups was sig-

Table 3Agonist electrical activities of knee extensors and flexors recordedduring the isokinetic strength assessment (60/s and 180/s)

EMG activities C 60 C 180 p

ExtensorsControl agonist (µV) 168 (48) 152 (43) 0.003FM agonist (µV) 153 (68) 159 (61) 0.37CLBP agonist (µV) 150 (91) 151 (94) 0.79

FlexorsControl agonist (µV) 203 (92) 176 (65) 0.01FM agonist (µV) 179 (75) 182 (73) 0.72CLBP agonist (µV) 172 (91) 162 (74) 0.22

C, concentric.

Table 4Comparison of normalized EMG activity Ext / Fl ratio recordedduring the isokinetic strength assessments (60/s and 180/s)

Ratio Ext/Fl C 60 C 180 p

Control 0.55 (0.27) 0.68 (0.34) 0.003FM 0.8 (0.1) 1 (0.16) 0.03CLBP 0.92 (0.15) 1.12 (0.15) 0.004

nificantly influenced by velocity in the control group incontrast to the chronic pain populations (Table 3).

The normalized EMG activity Ext/Fl ratio calculatedin the 3 groups of subjects appeared significantly lowerat low velocity (60/s) than at high velocity (180/s).


Evolution profile of RF agonist EMG activity

100110120130140150160170

1 to 3 9 to 11 18 to 20 28 to 30 contractions

EM

G a

ctiv

ity (µ

V)

Controls

FM

CLBP

Fig. 1. Evolution profile of RF agonist EMG activity recorded during the isokinetic fatigue assessment (control subjects, FM patients and CLBPpatients).

Evolution profile of EH agonist EMG activity

100

120

140

160

180

200

220


EM

G a

ctiv

ity

(µV

)

Controls

FM

CLBP

Fig. 2. Evolution profile of EH agonist EMG activity recorded during the isokinetic fatigue assessment (control subjects, FM patients and CLBPpatients).

3.3. EMG pattern during isokinetic fatigueassessment

We determined the agonist and the antagonist EMGactivities of VM, RF, IH and EH during the isokineticfatigue assessment. Four sequences consisting in 3 con-secutive flexion-extension movements were compared:1st to 3rd repetitions, 9th to 11th repetitions, 18th to20th repetitions and 28th to 30th repetitions.

Whatever the sequence, the agonist and antagonistEMG activities did not significantly differ between thethree populations. By contrast with the steady agonistEMG activity of the knee flexors (IH and EH), weobserved a moderate agonist EMG activity increase ofVM and RM in the control group and in the chronicpain patients (Figs 1 and 2). The antagonist EMGactivities increase at the end of the fatigue assessmentreached+2% to+ 29% depending on the muscle andthe population studied (no significant difference).

The curve profiles of agonist (Figs 1 and 2) andantagonist (Figs 3 and 4) EMG activities were similarin control group and chronic pain patients.

4. Discussion

In this study, we compared muscle performances andEMG patterns in a control group and in two groupssuffering from chronic pain: fibromyalgia and chroniclow back pain patients. Variables reflecting isokineticstrength and muscle fatigue resistance were signifi-cantly lower in the FM group compared to the controlgroup and the CLBP patients.

The FM patients muscular performances impairmentis well documented [7,11,13]. It could be related tothe anticipation of pain, pain itself, central alterationsor mechanisms induced by a deconditioning syndrome.The predominant reduction of isokinetic strength onknee flexors may be explained by our testing procedurewhich requires successive repetitions of knee flexionsand extensions without period of rest. In contrast toother studies, knee muscles performances did not dif-fer between healthy women and CLBP patients. Leeet al. [10] found a significant reduction in knee flex-ors and extensors strength in CLBP patients. Schip-plein et al. [21] and Trafimow et al. [25] suggested thatquadriceps muscle performance limits the CLBP pa-


Evolution profile of RF antagonist EMG activity

25

30

35

40

45

50


EM

G a

ctiv

ity (µ

V)

Controls

FM

CLBP

Fig. 3. Evolution profile of RF antagonist EMG activity recorded during the isokinetic fatigue assessment (control subjects, FM patients andCLBP patients).

Evolution profile of EH antagonist EMG activity

50

55

60

65

70

75


EM

G a

ctiv

ity

(µV

)

Controls

FM

CLBP

Fig. 4. Evolution profile of EH antagonist EMG activity recorded during the isokinetic fatigue assessment (control subjects, FM patients andCLBP patients).

tients’ ability to lift with their knee flexed. In our study,the similar isokinetic performances in the control andCLBP groups may be explained partly by our exclusioncriteria (absence of leg irradiation). Furthermore, thepopulations were drastically paired for age, weight anddaily physical activities.

To our knowledge, no comparative study of muscleperformances in FM and CLBP patients appears in theliterature. The comparison of both populations suffer-ing from chronic pain showed higher isokinetic strengthand muscle fatigue resistance in CLBP patients. Thediffuse pain in FM patients may explain these obser-vations. At the end of the isokinetic assessment, VASpain score reached respectively 0.68 a.u. and 2.7 a.u. incontrol and CLBP groups (non significant difference).By contrast, the VAS pain score in FM averaged 5.3a.u. (p < 0.05).

Agonist electrical activities of VM, RF, IH and EHrecorded during the isokinetic strength assessment werenot significantly different among our populations. Si-mons and Mense [23] reported no rest EMG activityincrease in patients suffering from chronic pain. Sihvo-nen et al. [22] described a trunk flexors and extensors

activity decrease, in CLBP patients, when they had anagonist function in dynamic movements.

In the control group, we demonstrated a signifi-cant influence of velocity on agonist EMG activity ofknee flexors and extensors. According to Robertson etal. [19], torque production is strongly related to EMGactivity in the biceps femoris and vastus lateralis. Theseauthors observed that increased torque was associatedwith increased EMG activity. Croce et al. [2] found thatknee flexors PT was higher when the ankle was fixedin dorsiflexion compared to plantarflexion. This in-creased PT was not combined with concomitant EMGactivity increase. However, according to these authors,the results were not in contradiction with Robertson’sstudy. Croce et al. [2] suggested that the higher kneeflexors PT was due to a greater gastrocnemius contri-bution as knee flexors, which explained the absence ofrepercussion in hamstring EMG activity. In our study,no significant relation between PT (which were signifi-cantly reduced at 180/s compared to 60/s) and EMGactivity was observed in the two populations suffer-ing from chronic pain. These results suggested that


chronic pain syndrome may influence the motor unitsrecruitment.

We determined the knee flexors and extensors EMGpattern. The normalized Ext/Fl ratio of EMG activityincreased significantly with speed in the 3 groups. Thisobservation may be related to the muscular typology;Garret et al. [5] demonstrated a higher proportion of fasttwitch motor units in hamstrings than in the quadriceps.Faulkner et al. [4] suggested that, at high contractionvelocities, slow twitch motor units were less able tocontribute to power than at slow contraction velocities.Therefore, FT motor units contribution may be moreimportant at 180/s. This fact may also explain thatknee flexors EMG activity increases less than quadri-ceps muscle EMG activity (composed of a higher num-ber of ST motor units) when speed gets higher. Re-sults of the present study indicated a normalized EMGactivity Ext/Fl ratio inferior to 1 in the control group.This observation may also be related to the differentfiber types proportions in knee flexors and extensors. Infact, Gerdle et al. [6] reported a positive correlation be-tween RMS and the percentage of type II muscle fibersduring 100 maximal dynamic knee extensions at 90/s.The higher ratio found in our chronic pain populationsmay be explained by a modified typology consecutiveto a deconditioning syndrome or a reduction of dailyphysical activities.

In the isokinetic fatigue assessment, we examinedthe agonist and antagonist EMG activities through-out 4 serials of contractions. The signal amplitude(RMS) reflects the recruitment of motor units. Theliterature describes a RMS increase during prolongedstatic submaximal contractions [8]. EMG pattern dur-ing repeated dynamic movements is less documented.Larsson’s study showed good reproducibility for maxi-mal isokinetic knee extension PT and RMS during threesets of 10 contractions [9]. Lindstrom and Gerdle [12]investigated the interrelationships between RMS andpeak torque throughout 100 successive maximal isoki-netic contractions of both knee flexors and extensors.They reported a RMS increase during the initial 20contractions, followed by relatively stable RMS levelsthroughout the subsequent contractions [12]. At theend of the test, a tendency towards decreased levelsappeared [12]. According to these authors, no signifi-cant difference in EMG behavior between extensor andflexor muscles existed. Contrarily, our study revealedthat unlike the knee flexors activity (IH and EH activ-ities), the agonist EMG activity of VM and RM mod-erately increased. Our results may differ from thoseof Lindstrom and Gerdle, due to our specific protocol.

Several mechanisms may be responsible for changesin the RMS and it remains difficult to interpret theEMG activity measured during maximal dynamic ef-fort. Unexpectedly, agonist and antagonist activities ofthe chronic pain groups did not differ from those of thecontrol group for each repetitions serial investigated inthe present study.

Curve profiles of agonist or antagonist EMG activi-ties were similar in the control group and the chronicpain patients. Consequently, the chronic pain syndromeappears to be not related to the EMG pattern duringisokinetic fatigue assessment.

5. Conclusion

FM patients isokinetic muscular performances weresignificantly decreased in comparison with the controland the CLBP groups. Results of these last two popula-tions did not differ significantly. Agonist electrical ac-tivities recorded during the isokinetic strength assess-ment were similar in the 3 studied groups. In the con-trol subjects, we demonstrated a significant influenceof velocity on agonist EMG activity of knee flexorsand extensors. The normalized EMG activity Ext/Flratio increased significantly with speed in the 3 popu-lations. During the isokinetic fatigue assessment, thecurve profiles of agonist and antagonist EMG activitieswere identical in the control group and the chronic paingroups. In contrast with the knee flexors agonist EMGactivity, we observed a moderate agonist EMG activityincrease of VM and RF.

References

[1] B. Bigland-Ritchie, EMG/force relations and fatigue of hu-man voluntary contractions, in:Exercise and sport sciencesreviews, D. Miller, ed., Philadelphia: Franklin Institute Press,1981, pp. 75–117.

[2] R.V. Croce, J.P. Miller and P. St. Pierre, Effect of ankle positionfixation on peak torque and electromyographic activity of theknee flexors and extensors,Electromyogr. Clin. Neurophysiol.40 (2000), 365–373.

[3] C. De-Luca, Myoelectric manifestations of localised muscularfatigue in humans,Crit. Rev. Biomed. Eng.11 (1984), 251–279.

[4] J.A. Faulkner, D.R. Clafin and K.K. McCully, Power output offast and slow fibers from human muscles, in:Human musclepower Human Kinetics, N.L. Jones, N. McCartney and A.J.McComas, eds, Champaign IL., 1986, pp. 81–94.

[5] W. Garrett, J. Califf and F. Basett, Histochemical correlates ofhamstring injuries,Am. J. Sports Med.12 (1984), 98–103.


[6] B. Gerdle, S. Karlsson, A.G. Crenshaw et al., The influences ofmuscle fibre proportions and areas upon EMG during maximaldynamic knee extensions,Eur. J Applied Physiol.81 (2000),2–10.

[7] S. Jacobsen, G. Wildschiodtz and B. Danneskold-Samsoe,Isokinetic and isometric muscle strength combined with tran-scutaneous electrical muscle stimulation in primary fibromyal-gia syndrome,J. Rhematol.18 (1991), 1390–1393.

[8] K.C. Jurell, Surface EMG and fatigue,Electromyography9(1998), 933–946.

[9] B. Larsson, B. Mansson, C. Karlberg et al., Reproducibilityof surface EMG variables and peak torque during three sets often dynamic contractions,J. Electromyogr. Kinesiol.9 (1999),351–357.

[10] J.H. Lee, Y. Ooi and K. Nakamura, Measurement of musclestrength of the trunk and the lower extremities in subjects withhistory of low back pain,Spine18 (1995), 1994–1996.

[11] M.H. Lindh, L.A. Johansson, M. Hedberg et al., Studies onmaximal voluntary muscle contraction in patients with fi-bromyalgia,Arch. Phys. Med. Rehabil.75 (1994), 1217–1222.

[12] B. Lindstrom and B. Gerdle, The interrelationships betweenEMG, peak torque and perceived fatigue during repeated max-imum isokinetic knee flexion with and without active kneeextension,Physiother. Theory Prac.10 (1994), 17–25.

[13] D. Maquet, J.L. Croisier, C. Renard et al., Muscle performancein patients with fibromyalgia,J. Bone Spine69 (2002), 293–299.

[14] A.M. Mengshoel, O. Forre and H.B. Komnaes, Musclestrength and aerobic capacity in primary fibromyalgia,Clin.Exp. Rheumatol.8 (1990), 475–479.

[15] T. Oberg, Muscle fatigue and calibration of EMG measure-ments,J. Electromyogr. Kinesiol.5 (1995), 239–243.

[16] N.J. Olsen and J.H. Park, Skeletal muscle abnormalities in

patients with fibromyalgia,Am. J. Med. Sci.315 (1998), 351–358.

[17] J. Potvin and L. Bent, A validation of techniques using surfaceEMG signals from dynamic contractions to quantify musclefatigue during repetitive tasks,J. Electromyogr. Kinesiol.7(1997), 131–139.

[18] J.A. Psek and E. Cafarelli, Behavior of coactive muscles duringfatigue,J Appl Physiol74 (1993), 170–175.

[19] R.N. Robertson, L.R. Osternig, J. Hamill et al., EMG-torquerelationships during isokinetic dynamometer exercise,SportsTraining Med. Rehab.2 (1990), 1–10.

[20] S.H. Roy, C.J. De Luca and D.A. Casavant, Lumbar musclefatigue and chronic lower back pain,Spine14 (1989), 992–1001.

[21] O.D. Schipplein, J.H. Trafimow, G.B. Andersson et al., Rela-tionship between moments at L5/S1 level, hip and knee jointwhen lifting, J. Biomech.23 (1990), 907–912.

[22] T. Sihvonen, J. Partanen, O. Hanninen et al., Electrical be-haviour of low back muscles during lumbar pelvic rhythm inlow back pain patients and healthy controls,Acta Phys. Med.Rehab.72 (1991), 1080–1087.

[23] D.G. Simons and S. Mense, Understanding and measurementof muscle tone as related to clinical muscle pain,Pain 75(1998), 1–17.

[24] P.A. Tesch, G.A. Dudley, M.R. Duvoisin et al., Force andEMG signal patterns during repeated bouts of concentric oreccentric actions,Acta Physiol. Scand.138 (1990), 263–271.

[25] J.H. Trafimow, O.D. Schipplein, G.J. Novak et al., The effectsof quadriceps fatigue on the technique of lifting,Spine18(1993), 364–367.

[26] J. Viitasalo and P. Komi, Signal characteristics of MEG duringfatigue,Eur. J. Appl. Phys.37 (1977), 111–121.


Abstract

Electromyographic activity of the knee flexorsand extensors in sedentary women duringisokinetic strength and fatigue assessments

D. Maquet, B. Forthomme, C. Demoulin, J.M. Crielaard and J.L. CroisierDepartment of Physical Medicine and Rehabilitation, University of Li ege, Belgium

Introduction: Isokinetic dynamometers are reputed toprovide accurate measurement of muscle strength andfatigue. The study of EMG patterns during suchisokinetic testing represents a complementary approachwhich seems particularly relevant [1]. The objectiveof our study was to assess concentric muscle strengthand fatigue during knee flexion and extension using anisokinetic dynamometer combined with analysis of legmuscles EMG patterns.Methods: Fifteen healthy females (45± 10 years old;66± 11 kg) without history of lower limb injury wereincluded in the study. After a familiarization periodon the dynamometer (Cybex Norm), all subjects per-formed a concentric isokinetic test on their dominantknee flexors (Fl) and extensors (Ext). The assessmentwas conducted along a constant range of motion from0 to 100 of flexion. Isokinetic muscle strength (peaktorque, PT in N.m) was first evaluated by means of 3repetitions at 60/s and 5 repetitions at 180/s. Thefatigue protocol consisted in 30 maximal-intensity con-traction cycles at 180/s. Fatigue was analyzed usingcumulative work (CW, in J) and bodyweight normalizedcumulative work (NCW, in J/kg) determined by sum-ming up the work developed by both muscle groups.Electrical activities of the vastus medialis (VM), rec-tus femoris (RF), internal hamstrings (IH) and exter-nal hamstrings (EH) were recorded from surface elec-trode using a standard surface EMG procedure (No-raxon Myosystem). All electrical activities were recti-

fied and smoothed (RMS 50). They were expressed inabsolute (µV) and PT normalized values (µV/N.m).Results: Intra-session reproducibility of EMG activi-ties was calculated by comparing the 3 first repetitionsof isokinetic strength (180/s) and fatigue assessments.Coefficients of variation did not exceed 7%. PT andEMG activities were influenced by velocity, the wholedata decreasing significantly at the angular velocity of180/s compared to 60/s. We determined a separateEMG pattern of knee flexors and extensors by meansof EMG activities respectively recorded from IH andEH and from VM and RF. Normalized EMG activityExt/Fl ratio appeared significantly influenced by veloc-ity (Ext/Fl ratio equal to 0.55 at 60/s versus 0.68 at180/s). The fatigue protocol showed a significant re-duction of flexors and extensors PT at the end of 30contraction cycles. In contrast, agonist and antago-nist EMG activities (e.g. flexor muscles recruitment re-spectively during flexion and extension) recorded dur-ing isokinetic fatigue assessment did not significantlychange throughout thirty repetitions neither for flexors,nor for extensors (Figs 1 and 2).Discussion and conclusion: In repeated EMG analy-sis procedures, reliability of EMG activities appearedsatisfactory with CV inferior or equal to 7%. Thisstudy demonstrated a significant influence of velocityon torque and EMG activity. Comparison between kneeflexor and extensor muscles showed higher flexor EMGactivity (Ext/Fl ratio< 1). This ratio significantly in-creased with velocity (0.55 at 60/s to 0.68 at 180/s).


D. Maquet et al. / Electromyographic activity of the knee flexors and extensors 13

Agonist EMG activities

100

120

140

160

180

200

1 to 3 9 to 11 18 to 20 28 to 30 repetitions

VM

RF

IH

EH

µV

Fig. 1.

Antagonist EMG activities

30

40

50

60

70

80

1 to 3 9 to 11 18 to 20 28 to 30 repetitions

VM

RF

IH

EH

µV

Fig. 2.

Agonist and antagonist EMG activities were not signif-icantly modified throughout isokinetic fatigue session.However, in contrast with the agonist EMG activity ofIH and EH, we observed a moderate increase of agonistEMG activity of VM and RF.

Reference

[1] R.N. Robertson, L.R. Osternig, J. Hamill and P. Devita, EMG-torque relationships during isokinetic dynamometer exercise,Sports Training Med. Rehab. 2 (1990), 1–10.


Abstract

Isokinetic assessment of complete proximalhamstring tendon rupture: Case reports

J.L. Croisier, V. Bartsch, M. Burnel and B. ZeevaertDepartment of Physical Medicine and Rehabilitation, University of Li ege, BelgiumE-mail: [email protected]

Complete hamstring tendon rupture is a rare clini-cal entity, reported particularly among water skiers [1].The functional outcome seems uncertain and most au-thors recommend surgical repair of the injury [2,3].In our work, we investigated two cases consisting ofwater skiing-related hamstring injury with a completetendon rupture. Patient A (male, 47 years old) was nonoperatively treated and patient B (male, 33 years old)benefited from surgical repair one year after the injury.In both cases, the clinical diagnosis was confirmed bymagnetic resonance imaging. No bony lesion from theischium was observed after the injury and electromyo-graphic examination excluded any motor recruitmentabnormality.

The standardized isokinetic assessment includedconcentric measurements at 60/s and 240/s for bothflexors and quadriceps. Afterward, flexor muscleswere subjected to eccentric angular speeds of 30/sand 120/s. The resulting analysis consisted of bilat-eral comparison of peak torques permitting determi-nation of asymmetries expressed in percentages andflexor/quadriceps (Fl/Q) ratios.

Patient A, managed with conservative treatment(classical physical therapy), still reported marked sub-jective complaints and disability on return to sports5 months after the injury. The first isokinetic assess-ment took place at that time and revealed a major deficitof maximal strength (Fig. 1). He started an isoki-netic strengthening program that was individually es-tablished on the basis of his own strength profile. After20 training sessions, a second isokinetic testing did not

show any improvement in performance, whatever thecontraction mode.

After surgery,patient B was instructed by the sur-geon to progressively perform peddling and swimmingactivities, but he was not allowed to receive any rehabil-itation treatment! Due to a poor functional outcome, hewas submitted 3 years postoperatively to a first isoki-netic assessment that revealed a decrease in strengthperformance higher than 50% (Fig. 1). After a first pe-riod of isokinetic training (10 sessions) the peak torques(measured through a second isokinetic test) improved,mainly in concentric. By contrast, a second period of15 training sessions did not allow any further recoveryin strength. The residual deficit finally reached 29% inconcentric (at 60/s) and 48% in eccentric (at 30/s).

Both patients were characterized by very low Fl/Qratios meaning a persistent agonist/antagonist imbal-ance. These observations concur with the concept thatnon-operative treatment leads to profound impairmentin strength and function among patients with completeproximal hamstring tendon rupture. Our follow-up alsosuggests that surgical repair does not systematicallylead to a total recovery in muscle performance, par-ticularly when the postoperative rehabilitation phase isneglected.

References

[1] Sallay et al.,Am J Sports Med 24 (1996), 130–136.[2] Cross et al.,Am J Sports Med 26 (1998), 785–788.[3] Klingele and Sallay,Am J Sports Med 30 (2002), 742–747.


J.L. Croisier et al. / Isokinetic assessment of complete proximal hamstring tendon rupture: Case reports 15

0

10

20

30

40

50

60

70

PT

def

icit

(%

)

Test 1Test 2Test 3

Patient A Patient B

C60 C60 E30 E30

= Isokinetic strengthening intervention

Fig. 1. Evolution of peak torque (PT) deficits (%) on injured muscles through successive isokinetic assessments. C60 = concentric at 60/s; E30= eccentric at 30/s.


Abstract

Muscle activation after ACL reconstruction:Influence of the resistance pad position

J.L. Croisiera, B. Forthommea, H. Baudoina, J.P. Huskinb, J.M. Crielaarda and D. MaquetaaDepartment of Physical Medicine and Rehabilitation, University of Li ege, BelgiumE-mail: [email protected] of Orthopedic Surgery, University of Liege, Belgium

Objectives: The question of anterior tibiofemoral jointshear force during active extension of the knee must betackled with great attention when managing the rehabil-itation of surgically reconstructed anterior cruciate lig-ament (ACL). The contraction of the quadriceps whenthe resistance pad of the isokinetic dynamometer leverarm is placed at the distal tibia tends to force the proxi-mal end of the tibia anteriorly with respect to the distalfemur [1]. This mechanism could damage the graftafter a surgical repair. The magnitude of anterior shearforce is significantly reduced when the resistance padis moved closer to the proximal extremity of the tibia.Hence, practitioners frequently use either a proximalplacement or an anti-shear device aimed at reducingthe anterior shear, thus protecting the graft in the earlyphase of exercise. Nevertheless, as we demonstrated ina previous publication [2], the position of the resistancepad on the tibia influences the estimation of the strengthdeficit through a bilateral comparison. The aim of thisstudy was to investigate the influence of the resistancepad positioning on the electromyographic (EMG) ac-tivity pattern of the flexor and quadriceps muscles afterACL reconstruction.Methods: Twelve male subjects (28± 7 years old) withunilateral knee surgery were studied. All of them un-derwent an ACL reconstruction using a patellar tendongraft. The mean time between the surgery and the studyexamination was 24 weeks (range 22–26 weeks). Eachpatient benefited from a pain free bilateral isokineticassessment for knee flexors and quadriceps in the con-

centric mode at 60/s and 240/s. Two different posi-tions of the resistance pad either distal (dist) or proxi-mal (prox) on the tibia were successively used for test-ing (order randomly assigned). At the same time, EMGactivity was measured on the vastus medialis (VM),the rectus femoris (RF), the respective mass of internalhamstrings (IH) and external hamstrings (EH) usingsurface electrodes. EMG data were sampled at 1000Hz and recorded using Noraxon Myosystem Software.Signals were rectified and smoothed (RMS 50). Theaverage root mean square (RMS) represented muscleactivity.Results: The isokinetic strength deficits of the quadri-ceps (through a bilateral comparison) appeared signifi-cantly increased when the test was performed in a dis-tal position of the resistance pad compared to a prox-imal position (at low velocity only). The quadricepsdeficit at 60/s averaged respectively 36± 14% and19%± 13% in the distal and proximal positions (p <0.05). Strength deficit magnitude on flexors was notinfluenced by resistance pad position (respectively 10± 7% and 7± 8% in the distal and proximal positionsat 60/s). The VM and RF activation pattern duringisokinetic knee extension at maximal intensity was alsomodified by the resistance pad position on the tibia. Atlow velocity (60/s), the prox/dist EMG activity ratioof both quadriceps muscle heads (Fig. 1) was signifi-cantly increased on the operated leg compared to thecontralateral healthy side. By contrast, the prox/dist


J.L. Croisier et al. / Muscle activation after ACL reconstruction: Influence of the resistance pad position 17

0

0.2

0.4

0.6

0.8

1

1.2

1.4

60˚/s 240˚/s

Pro

x / D

ist

rati

o (

EM

G a

ctiv

ity)

Healthy sideACL surgery

p < 0.05 *

*

Fig. 1. Prox/Dist EMG activity ratio of the VM: comparative study between ACL reconstructed knee and the contralateral healthy joint.

EMG activity ratio of the IH and EH during flexion didnot show significant difference between the two legs.Conclusion: We concluded that, after an ACL recon-struction, the position of the resistance pad on thetibia during maximal isokinetic knee extension at alow angular velocity significantly influences quadricepstorque production and activation. Based on these ob-servations, isokinetic users must reflect on the follow-ing points when they have to define the resistance padposition: (1) the safety of the graft referring to the ante-

rior shear forces; (2) the true estimation of the quadri-ceps deficit; (3) the optimal modalities in strengthen-ing, taking into account inhibition phenomena.

References

[1] R. Nisell et al.,Am J Sports Med 17 (1989), 49–54.[2] J.L. Croisier et al.,Actualites en Reeducation Fonctionnelle et

Readaptation, Masson, Paris, 1995, pp. 239–245.


Abstract

Randomized comparison of isokineticdynamometry and electromyographic activity(sEmg) after anterior cruciate ligamentreconstruction between semitendinosus andgracilis or patellar tendon autograft

D. Mourauxa, P.-M. Dugaillya, E. Brassinnea and M. VancabekebaDepartment of Physical Therapy and Rehabilitation, Universit e Libre de Bruxelles, Erasmus Hospital, BelgiumbDepartment of Orthopaedic Surgery, Brugmann Hospital, BelgiumE-mail: [email protected]

Objectives: There is still a lack of agreement regardingthe choice between hamstring or patellar tendon auto-graft techniques for the reconstruction of the anteriorcruciate ligament [4]. The goal of our study was to com-pare isokinetic parameters, 3 months post-operatively,after ACL reconstruction with hamstring tendon graftor with patellar tendon versus a healthy control group.Materials and methods: Twenty healthy male studentsparticipated in the study. They were compared to thirtypatients who had surgery for ACL reconstruction withhamstring tendon (n = 20) or patellar tendon (n = 10)autograft precisely 3 months before the assessments.The mean age and morphometric parameters were sim-ilar in each three samples. The tests were performedusing a Cybex Norm and consisted in 3 maximal dy-namic contractions at 180/s and 60/s in concentricmode (Con) for both flexor muscles and quadriceps andat 30/s in eccentric mode (Ecc) for flexors exclusively.During these tests, the sEmg activity was recorded fromthe vastus lateralis (VL), vastus medialis (VM) rectusfemoris (RF) and hamstrings using a Noraxon Myosys-tem 2000. Parameters were sequenced at every 10 ona range of motion from 0 to 105. The influence of

independent variables (group and limb) was assessedusing analysis of variance (ANOVA).Results: In the present study, both surgical proceduresinduced a significant loss of strength of 30% for ham-string muscles in concentric and eccentric modes (Ta-ble 1). For the quadriceps, we observed a significantdeficit for both samples but in the patellar tendon popu-lation, the decrease of strength was more important andreached 54%, 60% at 180/s and 60/s respectively, inthe concentric mode. At 60/s, sEmg activity showedsignificant differences for the quadriceps (RF, VL, VM)from 15 to 85 of knee flexion between patellar tendongraft and the two other groups.Discussion: As expected at 3 months post-op, the in-volved knee presented a significant weakness of thehamstrings and quadriceps compared to the uninvolvedknee. This loss of strength is more important afterpatellar tendon graft. Nevertheless at this stage of re-covery it is probably to early to predict better functionaloutcome between these two surgical procedures. Thus,it is necessary to reevaluate the presented parameters atone year and to correlate them with a functional score.


D. Mouraux et al. / Randomized comparison of isokinetic dynamometry and electromyographic activity (sEmg) 19

Table 1Mean± standard deviation of quadriceps and hamstrings Peak Torque± PT)

Control Hamstring tendon graft Patellar tendon graftUninvolved Involved Uninvolved Involved

Quad PT Con 180/s 156± 21 150± 37 113± 38∗ 109± 36∗ 71± 35∗Quad PT Con 60/s 215± 32 209± 48 151± 55∗ 167± 56∗ 87± 40∗Ham PT Con 180/s 99± 12 89± 22 72± 23∗ 68± 25∗ 59± 27∗Ham PT Con 60/s 136± 15 122± 26 100± 32∗ 97± 31∗ 82± 36∗Ham PT Ecc 30/s 182± 27 185± 27 137± 44∗ 146± 58 121± 52∗

∗ = Comparison between control legs and uninvolved-involved legs for both technical surgeryp <0.05 level.

References

[1] A.K. Aune, I. Holm, M.A. Risberg, H.K. Jensen and H. Steen,Four-strand hamstring tendon autograft compared with patellartendon bone autograft for ACL reconstruction. A randomizedstudy with two years follow-up,Am J Sports Med 29(6) (2001),722–728.

[2] J.A. Feller and K.E. Webster, A randomized comparison ofpatellar tendon and hamstring tendon ACL reconstruction,AmJ Sports Med 31(4) (2003), 564–573.

[3] L.A. Hiemstra, S. Webber and P.B. MacDonald, Knee strengthdeficit after hamstring tendon and patellar tendon ACL recon-struction,Med Sci Sport Exerc 32(8) (2000), 1472–1479.

[4] D.B. O’Neil, Arthroscopically assisted reconstruction of theACL: a prospective randomized analysis of three techniques,JBone Joint Surg Am 78(6) (1996), 803–813.

[5] E. Witvrouw, J. Bellemans, R. Verdonk, D. Cambier, P. Coore-vits and F. Almqvist, Patellar tendon vs semitendinosus andgracilis tendon for anterior cruciate ligament reconstruction,IntOrthop 25(5) (2001), 308–311.


Isokinetic evaluation of anterior cruciateligament reconstruction using a free fascialata graft strengthened by gracilis tendon

I. Rebeyrotte-Bouleguea,b, J.-C. Davieta, A. Oksmanc, J.-Y. Sallea, P. Dudognona and C. MabitcaDepartment of Physical Medecine and Rehabilitation, CHU, 87042 Limoges Cedex, FrancebService of Physical Medecine and Rehabilitation, CH, 40024 Mont-de-Marsan Cedex, FrancecDepartment of Orthopaedics and Trauma Medicine, CHU, 87042 Limoges Cedex, France

Abstract. Anterior cruciate ligament (ACL) reconstruction is used to restore knee stability.Objective: To evaluate the functional outcomes following anterior cruciate ligament reconstruction using a free fascia lata graftstrengthened by gracilis tendon.Methods: 60 patients met the inclusion criteria (age over 16, minimum 1 year follow-up). The isokinetic test on Cybex II consistof a concentric evaluation at 60/s, 180/s and 240/s and an eccentric one at 30/s. A Satisfaction index, Lysholm-Tegner scoreand IKDC score completed the evaluation.Results: 55 patients were examined (91.6%), mean age 31 years, 69% of men and a mean follow up of 30 months; only 50 hadisokinetic test. 81.8% were really/very satisfied. Mean Lysholm knee score is 96/100 with 93% of excellent and good results;mean Tegner activity level is 7 (8 before ACL injury) with 71% return to their previous sport level. IKDC global score noted 98%of good (22% B grade) and very good (76% A grade) results. In concentric test, 56% of patients have a hamstring muscle strengthloss, with a mean deficit of 9% at 60/s and 8.8% at 180/s; 76% had a quadriceps strength loss, with a mean deficit of 15% at60/s and 11% at 180/s. The hamstring and quadriceps deficit is around 9% in eccentric test. The ratio hamstring/quadricepsare symmetrical between the treated knee and the healthy one. No significant difference is related.Discussion-Conclusion: These results are in accordance with the principal series in the literature. The agonist-antagonist muscularobtained balance shows a fearly good analytical muscular recovery. The functional outcomes agree with other studies. Thus, thistechnique of anterior cruciate ligament reconstruction is reliable, respect extensor system and does not disturb hamstring musclerecovery.

1. Introduction

Anterior cruciate ligament (ACL) rupture, a frequentaccident affecting mostly young people practising con-tact and pivot sports (around 30 000 persons per year inFrance) is responsible for a chronic knee instability andexposes to meniscus and osteo-cartilaginous lesions,making the bed for osteoarthritis in the long term.

Anterior cruciate ligament reconstruction is the pro-posed surgical treatment to restore knee functional sta-bility and to limit degenerative lesions. Numerous tech-niques of reconstruction exist, but the most frequentlyused are the bone-patellar tendon-bone graft harvestedfrom the middle third of the patellar tendon, derived

from Kenneth-Jones operation and the 4-strand semi-tendinous and gracilis tendon (hamstring) graft withmetal interference screw fixation; each of them has ad-vantages and drawbacks. Functional outcomes are allin all similar, but morbidity is less with the hamstringgraft and laxity is better corrected by the patellar tendongraft [1,6,7].

A strip of fascia lata has also been used (Lemaire,McIntosh, Jaeger). Since 1999, Mabit has proposedan original alternative to the two reference techniqueswith the use of a fascia lata free graft, with a bonestick harvested from the tibial intercondylar tubercule(bone-tendon graft) strengthened by 2-strand gracilistendon.


I. Rebeyrotte-Boulegue et al. / Isokinetic evaluation of anterior cruciate ligament reconstruction 21

Whatever the technique chosen, the muscularstrength and functional recovery remain the first pre-occupation/priority, a challenge all the more essen-tial these young sportsmen set very high standards forthemselves. The evaluation of the maximal muscularstrength with an isokinetic dynamometer is used bymany surgical or rehabilitative teams to follow theirpatients outcomes.

The purpose of our study is to evaluate the func-tional outcomes and the consequences on hamstringand quadriceps muscular strenght following anteriorcruciate ligament reconstruction using a free fascia latagraft strengthened by gracilis tendon, with a mean 2.5years postoperative follow-up.

2. Methods

This study involves 60 patients with a chronic kneeanterior laxity, and who underwent arthroscopically as-sisted anterior cruciate ligament reconstruction using afree fascia lata graft strenghten by gracilis tendon bythe same surgeon, Mabit, between January 1999 andMay 2002 at Limoges University Hospital.

The patients are all over 16 years old and the post-operative follow-up had to be above one year long.

55 patients were physically re-examined, whichmeans a review rate of 91.6%; for geographical rea-sons, 4 were not examined and only answered phonequestionnaires. Only one person was lost to follow-up.

The 55 group is composed of 38 men (69%) for 17women with a mean age of 31 years (20–59). The meanpostoperative follow-up is 30 months (14–49) and themean delay between initial injury and surgery is 41months (3 months to 20 years). The right knee was in-volved in 30 cases (54.5%) and the left knee in 25 cases.The different circumstances for the ACL rupture werefootball for 27%, handball (17%), basketball (13%),ski (13%), other sports for 23% and traffic accident for9%.

2.1. Isokinetic evaluation

Knee flexors and extensors muscular performance ismeasured by a Cybex II dynamometer. After a 20 min-utes warming-up on cycloergometer, the patient sitson the dynamometer with straps around the trunk; thetested leg is strapped around the thigh, with a counter-support above the malleolus. The controlateral leg isblocked behind a support to avoid it to move freely.The leg weight is taken into account to correct grav-

ity effect. The knee joint motion is adjusted to have arange of motion of 90, with the same reference framefor both knee. The system axis is adjusted to the jointrotation axis.The test is made without visual feed-back.Before starting the test,a training is done at the differentspeeds that will be used for the concentric evaluation.The healthy knee is first tested. The same physiatristhas made all the isokinetic evaluation.

The evaluation on concentric mode is composed of4 repetitions at 60/s, 4 movements at 180/s followedby an endurance test with 20 repetitions at 240/s, with2 minutes’ rest between each series. The parametersanalysed are peak knee flexion and extension torqueor maximum knee flexion and extension moments andthe flexors/extensors ratio. A percentile torque deficitin relation to the corresponding non-involved knee iscalculated.

For the eccentric mode evaluation, after a training, 3repetitions at 30/s are performed, with the analysis ofthe same parameters quoted above.

For professional unavailiability of 5 of them, only50 patients over the 55 were tested on the isokineticdynamometer.

2.2. Clinical and functional evaluation

An orthopaedic surgeon, distinct from the surgeonwho made the ACL reconstruction, records the subjec-tive symptoms and performs the clinical examinationof the patients and then the radiographs to allow forcompletion of the IKDC 1999 score, with each of the 7parameters quoted in 4 grades (A “normal”, B “almostnormal”, C “abnormal”, D “severely abnormal”) and aglobal score also quoted in 4 grades. A measure withthe KT-1000 arthrometer indicates the residual antero-posterior laxity (mean of three successives measures).A Lysholm knee score, with total score over 100 points,determines three levels of performance (good/excellentbetween 84 and 100, medium between 65 and 83 andbad under 65); the Tegner activity level that gives thephysical and sports activity level (from 1 to 10) is eval-uated a posteriori for the pre-injury period and then atthe study time.

At last, a satisfaction index is established by thepatient with three levels (very satisfied, satisfied, nonsatisfied).

2.3. tatistical analysis

The Student t-test, Chi 2 test or Fisher exact test wereused to compare qualitative data. The link betweenordinal quantitative data is studied by the correlationtest (linear correlation).P < 0.05 was regarded assignificant for all the tests.

22 I. Rebeyrotte-Boulegue et al. / Isokinetic evaluation of anterior cruciate ligament reconstruction

0

10

20

30

40

50

60

70

80

90

Verysatisfied

Satisfied Nonsatisfied

Fig. 1. Satisfaction index.

0

10

20

30

40

50

60

70

80

90

100

go o d/ excellent bad

Fig. 2. Lysholm knee score.

3. Results

3.1. Clinical and functional results

The patients consider the results of their operations asvery satisfying; the satisfaction index (Fig. 1) shows 45persons very satisfied (81.8%), for 9 satisfied (16.4%),and one who is not satisfied (1.8%), suffering of insta-bility and subjective and objective pains.

The mean Lysholm knee score (Fig. 2) at follow-up is 96 (75–100), with 51 good and excellent results(93%) and only 4 medium ones.

On Tegner activity level, the mean level of physicaland sports activity preinjury was 8; after the surgery,when review is performed, the mean level is 7 (3–9). Thirty-nine patients (71%) have recovered theirprevious sport activity level and practise sport then at

Equal level71%

Lower level for functional deficiency

7%

Lower level for fear15%

Lower level for professional

reason7%

Fig. 3. Tegner activity level.

0 10 20 30 40 50 60 70 80

A B C D

Fig. 4. Global IKDC score.

the level they had before the injury. On the other hand,16 persons (29%) have taken up sport again but at alower level. An actual functional deficiency of the kneeis the cause for only four of them; 8 patients (7.2%)have slightly reduced their sporting activity for fear ofgetting injured again. The last four have limited theiractivity only for professional reasons (Fig. 3).

According to the clinical criteria of IKDC, the kneefunction is considered as normal (grade A) for 42 pa-tients, almost normal (grade B) for 12 of them, andabnormal (grade C) for only one (Fig. 4). We haveobtained 98% of good and excellent results. None ofthe 7 parameters is quoted D. Only one patient shows aresidual effusion (grade C). The passive motion of theknee is normal for all the patients. A moderate residuallaxity exists for 12 patients (grade B) (Fig. 5); it is con-firmed by the measure with the KT-1000 arthrometer.

3.2. Isokinetics results

At this postoperative 2.5 years follow-up, for isoki-netic concentric muscular evaluation, the extensors

I. Rebeyrotte-Boulegue et al. / Isokinetic evaluation of anterior cruciate ligament reconstruction 23

Table 1Isokinetic concentric muscular evaluation

Extensors Extensors Flexors Flexors Ratio flex/ext Ratio flex/exthealthy operated healthy operated healthy operated

side side side side side side

Peak torque 188+/− 58 170+/− 52 111+/− 36 109+/− 37 61%+/− 13 65%+/− 1260/s (93–313) (58–273) (11–191) (42–209) (37–126) (36–111)Peak torque 130+/− 37 121+/− 35 87+/− 24 83+/− 25 68%+/− 12.8 69.6%+/− 12180/s (67–228) (54–208) (45–145) (40–147) (41–107) (45–110)

Maximal peak torque in Nm.

0

10

20

30

40

50

60

70

80

A B C D

Fig. 5. Laxity IKDC score.

peak torque deficit is moderate, with a mean deficit of15.3% at 60/s and 11.3% at 180/s. But 48 patientsover the 50 tested (76%) present this extension peaktorque deficit. The flexors peak torque deficit is nonsignificant, with a mean deficit of 9.1% at 60/s and8.8% at 180/s; and only 28 persons over fifty show thisflexion peak torque deficit. The flexors/extensors ratioare similar between the operated knee and the healthyone (Table 1).

For isokinetic eccentric muscular evaluation, thepeak torque deficit on the treated side is about 9%, nonsignificant. The flexors/extensors ratio are equivalentcomparing the treated knee and the non-injured one(Table 2).

3.3. Inter-parameter correlation

No significant difference has been pointed up forisokinetic results, IKDC scores, Lysholm ones and Teg-ner activity levels regarding age and sex.

But, even without any significant difference, it seemsthat Lysholm scores and IKDC ones are better whenthe preoperative instability period is shorter.

4. Discussion

The sample of the current study is comparable withother series regarding age, sex, sports involved in theinjury, period of preoperative instability and mean post-operative follow up of 2.5 years [1,5,6,8,10,15].

The findings of the study are generally consistentwith those of similar trials that have been published re-cently. There is no significant difference between thisoriginal surgical technique using fascia lata and gra-cilis and the reference ones using bone-patellar tendon-bone or combined semitendinous and gracilis ham-strings tendon grafts in terms of functional outcomesat 2 or 3 years [4,9,12,15]. Even more our functionalresults are better [3,6,8].

The rates of return to preinjury levels of physicalactivity (71%) is high compared with other studies [6,14,15]. Feller [6] reports that only 55% of hamstringtendon patients return to their preinjury sport comparedwith 69% of bone-patellar tendon-bone graft patients.We used preinjury rather than preoperative levels ofactivity as a baseline to postoperative comparison; wethink that preinjury level is a more relevant parameterbecause the patients’sports level is often lower after theinjury.

Isokinetic performances do not differ from the onesafter bone-patellar tendon-bone or hamstring tendongrafts. In recent studies [1,5,6,10,15] often comparingthe two reference surgical procedures, the quadricepsdeficit is around 10 to 15% at low speed, a little lessat high speed. Feller [6] and Aune [1] found that thisdeficit is higher in the patellar tendon group than inthe hamstring tendon one. Some authors [2,14] notea more important quadriceps deficit, from 25 to 35%.Concerning hamstring muscular strength, the deficit of9% is in accordance with the results of other series [1,9,13,15]. This deficit seems higher in hamstring tendonstechnique [1,6]. In other reports, a reduced hamstringmuscle strength has not been really noted even whenthe hamstring tendons are used as graft material [5,14],which means a good hamstring strength recovery.

24 I. Rebeyrotte-Boulegue et al. / Isokinetic evaluation of anterior cruciate ligament reconstruction

Table 2Isokinetic eccentric muscular evaluation

Extensors Extensors Flexors Flexors Ratio flex/ext Ratio flex/exthealthy operated healthy operated healthy operated

side side side side side side

Peak torque −203+/− 59 −193+/− 61 −133+/− 44 −128+/− 44 66%+/− 19 65%+/− 14.530/s (−369–98) (−351–22) (−238–56) (−241–49) (34–165) (40–100)

Maximal peak torque in Nm.

In our study, only 3 patients (5.5%) suffer fromfemoro-patellar pain. All the authors recognize thatanterior knee pain is more frequent after harvest of apatellar tendon graft [1,6].

A larger sample size would perhaps allow to identifysignificant differences. To maximize the strength of ourstudy we used strict inclusion criteria; a single surgeonperformed all the procedures to reduce variability insurgical technique and the evaluation was always madeby the same independent observers to limit detectionbias.

This report is also limited by the fact that there is nocontrol group. The uninvolved knee is used as controland is so supposed to be healthy.

After the ACL reconstruction using a free fascia latagraft strengthened by gracilis tendon, the rehabilitationprotocol does not differ from usual protocol after ACLreconstruction.

5. Conclusion

The results of this study demonstrate that this surgi-cal procedure of ACL reconstruction using a free fascialata graft strengthened by gracilis tendon is a reliabletechnique, with low morbidity, and low functional andmuscular deficit. The extensors are respected. Thehamstring strength recovery is not disturbed. The kneelaxity is well controlled.

Thus this technique of ACL reconstruction is a goodalternative to the reference surgical procedures.

References

[1] A.K. Aune, I. Holm, M.A. Risberg, H.K. Jensen and H. Steen,Four-strand hamstring tendon autograft compared with patel-lar tendon-bone autograft for anterior cruciate ligament recon-struction. A randomized study with two-year follow-up,Am JSports Med 29(6) (2001), 722–728.

[2] T.R. Carter and S. Edinger, Isokinetic evaluation of anteriorcruciate ligament reconstruction: hamstring versus patellartendon,Arthroscopy 15(2) (1999), 169–172.

[3] W.P. Charlton, D.A. Randolph, S. Lemos and C.L. Shields,Clinical outcome of anterior cruciate ligament reconstructionwith quadrupled hamstring tendon graft and bioresorbable in-terference screw fixation,Am J Sports Med 223;31(4) 518–521.

[4] P. Colombet, M. Allard, V. Bousquet, C. De Lavigne, P.H.Flurin and C. Lachaud, Anterior cruciate ligament recon-struction using four-strand semitendinosus and gracilis ten-don grafts and metal interference screw fixation,Arthroscopy18(3) (2002), 232–237.

[5] L. Ejerhed, J. Kartus, N. Sernert, K. Kohler and J. Karlsson,Patellar tendon or semitendinosus tendon autografts for ante-rior cruciate ligament reconstruction?Am J Sports Med 31(1)(2003), 19-25.

[6] J.A. Feller and K.E. Webster, A randomized comparison ofpatellar tendon and hamstring tendon anterior cruciate liga-ment reconstruction,Am J Sports Med 31(4) (2003), 564–573

[7] K.B. Freedman, M.J. D’Amato, D.D. Nedeff, A. Kaz and B.R.Bach, Arthroscopic anterior cruciate ligament reconstruction:a metaanalysis comparing patellar tendon and hamstring au-tografts,Am J Sports Med 31(1) (2003), 2–11.

[8] K.A. Jansson, E. Linko, J. Sandelin and A. Harilainen, Aprospective randomized study of patellar versus hamstring ten-don autografts for anterior cruciate ligament reconstruction,Am J Sports Med 31(1) (2003), 12–18.

[9] M. Marcacci, S. Zaffagnini, F. Iacono et al., Intra- and extra-articular anterior cruciate ligament reconstruction utilizing au-togeneous semitendinosus and gracilis tendons: 5-year clini-cal results,Knee Surg Sports Traumatol Arthrosc 11 (2003),2–8.

[10] N. Nakamura, S. Horibe, S. Sasaki et al., Evaluation of activeknee flexion and hamstring strength after anterior cruciateligament reconstruction using hamstring tendons,Arthroscopy18(6) (2002), 598–602.

[11] D.D. Nedeff and B.R. Bach, Arthroscopic anterior cruciateligament reconstruction using patellar tendon autografts,Or-thopedics 25(3) (2002), 343–357.

[12] L.A. Pinczewski, D.J. Deehan, L. Salmon et al., A five-tearcomparison of patellar tendon versus four-strand hamstringtendon autograft for arthroscopic reconstruction of the anteriorcruciate ligament,Am J Sports Med 30(4) (2002), 523–536.

[13] J.-M. Parisaux, P. Boileau and C. Desnuelle, Isokinetic eval-uation of the knee flexor muscles after anterior cruciate liga-ment reconstruction using gracilis and semitendinosus tendongrafts,Rev Chir Orthop 90 (2004), 33–39.

[14] S. Robineau, P. Gallien, J. Jan and P. Roshcongar, Isokineticmuscular testing after anterior cruciate ligament reconstruc-tion: results, interest,Ann Readaptation Med Phys 43 (2000),437–449.

[15] T. Tashiro, H. Kurosawa, A. Kawakami, A. Hikita and N.Fukui, Influence of medial hamstring tendon harvest on kneeflexor strength after anterior cruciate ligament reconstruction.A detailed evaluation with comparison of single- and double-tendon harvest,Am J Sports Med 31(4) (2003), 522–529.


Abstract

Isokinetic and functional muscleperformances among football players: Atransversal study

J. Bineta, C. Lehancec, G.Vandenbroekd, T. Buryc and J.L CroisierbaDepartment of physical Medecine and Rehabilitation, CHR, Tournai, BelgiumbDepartment of physical Medecine and Rehabilitation, University of Li ege, BelgiumcDepartment of Physiology of Sports, University of Liege, BelgiumdExcelsior Mouscron Football Team

Athletic physical conditioning demands an ever in-creasing scientific approach. Isokinetics allows usto objectively quantify muscular strength. Howeversuch an evaluation remains analytical and only pro-vides us with partial information about the athlete’sfunctional skills. The opto-jump system assesses thestrength-speed factor in different ways (dash runningand jumps), therefore allowing a functional analysis ofthe muscle performance. The goal of this study deal-ing with football players was the transverse analysis ofthe functional and isokinetical variables in different agegroups and the correlation analysis between analyticaland functional variables.

57 football players, fit for intense sport practice, wereincluded into 3 subsets: pro-group (PG: 22-years-oldor older;n =20), minor leaguers (ML: 18 to 21-years-old; n =20), juniors group (JG: 15 to 17-years-old;n =17). The isokinetical strength of the knee’s flexor-extensor muscles was measured in concentric mode (60and 240/s) as well as eccentric (30 and 120/s). Fieldtests were also achieved: 10-meter dash running andsquat-jump series. These tests were performed by useof an optical measuring system that measures flightand ground contact times (opto-jump). A one criterionvariant analysis, combined with the Bonferroni test,enabled us to compare the performancesand correlativestudies were carried out.

Focusing on the absolute performance of the quadri-ceps strength at low speed, our results showed sig-nificant differences (p < 0.05) between ML and JG.This difference also exists between PG and JG (non-significant). At high speed, the absolute value of thequadriceps strength also showed significant differencesbetween PG and JG as well as differences between MLand JG. The same statistical differences between PG-JG and ML-JG were reported for the squat-jump data.

In the pro-group, the squat-jump appeared signifi-cantly correlated with extensor peak torques at 60/s(p < 0.001 andr = 0.64) and at 240/s (p < 0.0001and r = 0.63). The 10-meter dash running was notlinked to any isokinetic variables of the 3 player groups.

The maturation of the muscle system and intenseweight lifting sessions have most likely contributed tothe significant differences between the 3 tested groups.Intensive body-building surely affects the squat-jumpresults in all groups. The absence of correlation be-tween dash running and isokinetical variables couldbe explained by the not strength-related factors suchas reaction time, ground contact duration, gesture fre-quency.

Conversly to the 10-meter dash running test, thesquat-jump and the isokinetic assessment of the foot-ball player appear to be discriminating tests towards theage feature.


26 J. Binet et al. / Isokinetic and functional muscle performances among football players

0

50

100

150

200

250

300

PG ML JG

Squat JumpQuadriceps strength comparison at 60 and 240˚/s in all groups

240˚/sec

PG ML JG

60˚/sec

PG ML JG

Ave

rage

in

cm

Ave

rage

in

Nm

Despite the close relationship between some of theirparameters, isokinetic and opto-jump can be comple-mentary measurement tools: the opto-jump system al-lows a thorough functional evaluation of the lowerlimbs muscle performances and the isokinetic evalua-tion permits to assess analytical strengh of each mus-

cle group. For instance, if the field test results donot match with normal values, isokinetical evaluationcould identify a muscle group deficiency. By contrast,if the isokinetical values stand normal, a neuromuscularcoordination deficiency would be evoked.


Lateral specificity in knee muscle function incompetitive high jumpers compared withuntrained individuals

Sana Mansoura, Olivier Maisettia,b, Christophe Cornua,c and Pierre Porteroa,b,∗aInstitut de Myologie, Groupe Hospitalier Pitie-Salpetriere - F-75651 Paris Cedex 13, FrancebUFR STAPS, Paris XII-Val de Marne - 80, Avenue du General de Gaulle, 94009 Creteil, FrancecJE 2438 “Motricite, Interactions, Performance”, UFR STAPS, Universite de Nantes, France

Abstract. The preferential training of dominant over non-dominant limbs is usually associated with an increase in musclefunction. The aim of this study was to examine the effect of chronic high jump training on lateral dominance in knee musclefunction. Nine national high jumpers (5 females, 4 males) and fourteen controls performed maximal knee flexions and extensions(0 to 180.s−1) with dominant and non-dominant legs. Male high jumpers were stronger (p < 0.001) but more asymmetricthan controls during maximal isometric knee extension (+17% for the dominant leg,p < 0.01) in contrast with maximalconcentric knee extensions (−5% to+2%,p > 0.05). Additionally, levels in agonist and antagonist EMG activity during kneeextensions were similar across groups. These results suggest that neither neural activation or typology were responsible for thelaterality observed in male high jumpers. According to current knowledge on the effects of plyometric training and functionalcharacteristics of power-athletes, we hypothesized that this muscle- and velocity-specific lateral dominance in competitive malehigh jumpers was related to differences in terms of muscle volume and/or muscle stiffness. Further studies are needed to examinethese assumptions with suitable techniques over a larger sample of competitive high jumpers.

Keywords: Training, high jump, take-off leg, isokinetic torque, countermovement jump, electromyography, gender

1. Introduction

High jumping (HJ) is a complex motor task includ-ing a run-up and a one-legged vertical jump toward abar to clear. HJ involves the lower limbs intensely andparticularly the take-off leg whereas the contralateralleg i.e. the leg raised at take-off has been ascribed tolead the movement with upper limbs [1]. Thus, ac-cording to the principles of training specificity [16], thepreferential use of take-off leg due to chronic unilat-eral training is probably associated with higher func-tional changes and particularly for muscles crossingthe knee joint. Surprisingly, numerous studies have

∗Address for correspondence: Pr Pierre Portero, Institut de My-ologie, Groupe Hospitalier Pitie-Salpetriere, 83, Bd de l’Hopital,75651 Paris Cedex 13, France. Tel.: +33 1 42 16 58 75; Fax: +331.42 16 58 81; E-mail: [email protected].

investigated mechanical and muscular factors influenc-ing vertical jump performance but only during two-legged jumps, assuming laterality balance [30,32,38,39,47]. Indeed, to the best of our knowledge, only twostudies have been carried on one-legged vertical jumpperformance [5,60]. Van Soest et al. [5] reported a one-legged countermovement (CMJ) jump height that wasonly 58.8% of the two-legged CMJ jump. They sug-gested that the performance deficit during two-leggedjump was mainly due to reduced neural drive to kneeextensor and plantar flexor muscles. However, theseauthors did not test the non-preferred limb alone andthe tasks considerably differed in kinematics variables(e.g. push-off duration). Young et al. [60] did not re-strict the raised-leg swinging during one-legged CMJ,which makes difficult comparison between unilateraland bilateral stretch-shortening cycle performances.


28 S. Mansour et al. / Lateral specificity in knee muscle function in competitive high jumpers compared with untrained individuals

Interestingly, unilateral adaptations have been re-ported through single-joint isokinetic testing in athletessuch as soccer, tennis or water-polo players [25,56,61]. More recently, Siqueira et al. [15] did not showfunctional asymmetry in competitive long and triplejumpers during maximal concentric isokinetic actionsof knee flexor and extensor muscles (60–240.s−1).Although these jump events may have technical char-acteristics similar to high jumping, the question ofwhether chronic high jump training induces differenceacross lower limbs has not been addressed yet. OnlyAmiridis et al. [28] compared knee extensor musclesbetween high jumper athletes group and sedentary sub-jects during different maximal isokinetic contractions.However, they did not study the non-dominant leg.

Therefore, the present study was designed to exam-ine whether high jump training induced marked dif-ference in terms of laterality with reference to controlgroup. Maximal isometric and isokinetic torques ofknee extensors and flexors were performed by seden-tary subjects and highly-trained high jumpers engagedin periodic unilateral plyometric training and highjumping. We hypothesized that lateral differenceswould probably be more marked in elite high jumperswithout gender interference.

2. Methods

2.1. Subjects

Two groups of subjects with no prior history of kneepathology participated in the study. A control group(C) consisted of fourteen non-trained but physically ac-tive subjects [7 men (27± 2 years old) and 7 women(25± 1 years old)]. Nine competitive high jumpers [4men (25± 3 years old) and 5 women (23± 2 yearsold)] competing at the French national level in the In-ternational Association of Athletics Federation (IAAF)took also part in this study. All jumpers had trained andcompeted regularly in high jump for 2–6 years aboutfive times a week. The physical characteristics of theparticipants are shown in Table 1. Best high jumpingperformances (seasonal and personal best) and associ-ated points according to the IAAF table are shown inTable 2 for high jumpers. All subjects agreed to par-ticipate in the study on a voluntary basis and signed aninformed consent form.

Table 1Physical characteristics of the male and female high jumpers (HJ)and controls (C)

N Age (years) Body mass (Kg) Height (cm)

HJ women 5 23± 2 65± 9∗∗∗ 176± 10∗∗∗HJ men 4 25± 3 74± 9∗∗∗ 184± 8∗∗∗C women 7 25± 1 58± 6 165± 4C men 7 27± 2 66± 5 177± 4

HJ were taller and weighted more than controls (p < 0.001) withregard to gender (men> women,p < 0.001).

2.2. Functional performance

A force platform (LG-6-4-1, AMTI, Newton, MA)was used to perform vertical jump tests. To estimate theability to jump, subjects were asked to perform maxi-mal two-legged squat jump (SJ) and counter-movementjump (CMJ). The initial position for the SJ was the up-right position with the knee flexed at 90. During CMJ,subjects executed the counter movement until 90 ofknee flexion. All jumps were performed barefooted andthe subjects were instructed to keep their hands claspedbehind their back and to jump as high as possible. Asjump verticality was especially important, the subjectswere instructed to pay particular attention to keep thehead and trunk vertically during the jump and to landat about the same place on the force platform. Trialsin which these instructions were not respected were re-jected. The highest jump of three trials was retained forfurther analysis jump height, defined as the differencebetween the height of the mass center of the body atthe apex of the jump and the height of this mass centerwhen the subject was standing upright with heels on theground, was calculated from the following formula [10,45]:

SJ and CMJ height= 1/8*T2f*g, where g: gravity

force;Tf : flight time.Pre-stretch augmentation was calculated as [100*

(CMJ – SJ)/SJ] to reflect elastic recoil efficiency [4].

2.3. Isokinetic testing

A Biodex System II isokinetic dynamometer(Biodex medical, Shirley, N.Y., USA) was used to per-form knee isometric and concentric isokinetic tests.This apparatus was calibrated according to the pro-cedure recommended by the manufacturer using thestandard calibration (sensitivity level and gravitationalcorrection feature for knee joint analysis). A secondPC unit was used to collect surface electromyographic(SEMG) signals. A specific software was designed inour laboratory to store synchronously Biodexme-

S. Mansour et al. / Lateral specificity in knee muscle function in competitive high jumpers compared with untrained individuals29

Table 2High Jumpers personal best (PB) and seasonal best (SB) and related pointsaccording to the IAAF table score

PB (m) Points SB (m) Points

High M 2.27 1128 2.18 1040jumper N 2.12 982 2.10 962men S 2.07 933 2.05 914

A 2.10 962 1.95 818Mean± SD 2.1± 0.1 1001± 87 2.1± 0.1 934± 93High C 1.78 967 1.75 934jumper M 1.77 956 1.60 773women G 1.88 1076 1.88 1076

A 1.70 880 1.60 773S 1.68 858 1.60 773

Mean± SD 1.76± 0.08 947± 86 1.69± 0.13 866± 137

chanical signals (torque,angle and velocity) and SEMGsignals. Bipolar SEMG signals were recorded usingminiature Ag/AgCl surface electrodes (4 mm diameter,11 mm inter-electrode distance). Each SEMG channelwas connected to a preamplifier (band pass filter: 6–500 Hz) and was amplified (bandwidth: 10–1000 Hz)through an isolated amplifier (Gould 6600).

2.4. Experimental procedure

Prior to the actual test, subjects were trained to fa-miliarize with the experimental procedure. The mainexperiment was designed as follows. A full test ses-sion, including rest periods, lasted about 30 minutes andcomprised: (i) an explanation of the tests, (ii) prepara-tion of the subject, (iii) habituation to the tests, (iv) theactual tests.

First, surface electrodes were located on the rectusfemoris (RF) and biceps femoris (BF) muscles of bothlower limbs considering the recommendations of SE-NIAM project (1999). Inter-electrode impedance wasreduced to less than 10 kΩ with standard skin prepa-ration procedures. Then, after a standardized warm-up on the cycle ergometer, the subject was placed onthe seat in order to test knee flexors and extensors ofone lower limb. The trunk, pelvis, shoulders, and theupper portion of the thigh of the leg were stabilizedwith special restraint systems to minimize superfluousbody movements. The axis of the shaft rotation ofthe dynamometer was visually aligned with the lateralfemoral condyle of the knee. The leg was fixed to thedynamometer lever arm with a padded cuff just next tolateral malleolus. The general posture of the subjectduring the experiment was with the hip angle joint set at115 and the knee angle joint set at 120 for isometrictests (180 corresponding to knee maximal extensionposition). For the isokinetic test, a range of motion of

90 (between 90 and 180) was defined for full legextension-flexion movements.

Three maximal isometric knee extension trials fol-lowed by three maximal isometric knee flexion tri-als were performed with each limb randomly chosen.Each isometric contraction lasted maximally 3 secondsand 1 minute of rest was respected between each trial.Finally, the subject was asked to achieve three max-imal isokinetic knee extension-flexion movements at180.s−1. During all tests, the subjects had oral andvisual feedback and were instructed to work as hard aspossible.

2.5. Data processing

Whatever the subject, the take-off leg was definedas the preferred leg used to take an impulse during aone-legged jump. For isometric tests, the best torquevalue of the three trials was retained as the actual maxi-mal voluntary contraction (MVC). Moreover, for isoki-netic contraction, the trial with the highest peak torquewas identified in each subject and selected for furtheranalysis. During isometric and isokinetic contractions,SEMG were used to calculate the root mean square(RMS) amplitude reflecting the level of agonist andantagonist muscle activity. During isometric contrac-tions, RMS amplitude was averaged over a 1-s periodduring which the maximal torque was approximatelyconstant. During concentric contractions, SEMG sig-nals were analyzed over a range of motion which ex-cluded 30 at the beginning and at the end of the move-ment, so that only isokinetic activity was included andcalculated every 0.10 s in consecutive windows. Ag-onist activity was normalized to the maximum RMSEMG obtained during its respective maximal voluntaryisometric torque. Coactivation level was expressed asa percentage of the muscle’s maximal activity recordedduring its respective MVC.


2.6. Data reduction and statistical analysis

When not specified, data were reported as means± SD. Squat jump (SJ) and countermovement (CMJ)height differences were analyzed using a three-wayANOVA (gender by group by type of jump) (Statistica5.1., Statsoft, France). Bravais Pearson linear correla-tions coefficients were calculated between variablesof interest. Four-way ANOVA (gender by Group byside by velocity) was performed for knee extension andflexion separately (torque, agonist and antagonist EMGactivity). Significance was accepted atp < 0.05 for allstatistical tests.

3. Results

3.1. Jump performances

Jump performances were higher in high jumpers withreference to controls (+36%, p < 0.001) and sex-dependant (+23% for men,p < 0.001, Fig. 1). SJheight was significantly lower than that of CMJ (p <0.001). However, pre-stretch augmentation was similaracross groups and genders (+23%,P > 0.05) suggest-ing that the tendon stiffness of the knee extensor mus-cles was comparable among groups (Fig. 1). Pooleddata of CMJ and SJ performances were significantlycorrelated with extension torque of either take-off legor raised leg ranging between 0.36< r2 < 0.69 (p <0.05). For pooled high jumpers, CMJ and SJ heightwere correlated with personal or seasonal best perfor-mances, accounting for 58 to 78% of their explainedvariances.

3.2. Maximal voluntary torque-velocity curves andassociated EMG activity

3.2.1. Knee extensionKnee extensor muscles torques decreased signifi-

cantly with velocity (p < 0.001) and were higher forHJ than controls with respect to gender (p < 0.001).ANOVA revealed a four-factor significant interactionindicating that gender influenced the knee extensortorque-velocity curves when considering laterality (p <0.01). Thus, separate analysis for each gender wasperformed. The take-off leg of HJmen was strongerthan that of controls especially at the highest veloc-ity (15% at 0.s−1, p < 0.05; 29% at 180.s−1, p <0.01). In contrast, maximal isometric torque gener-ated by the raised leg (non-dominant) was not different

across groups (5%,p > 0.05) but HJmen raised legwas stronger than that of their counterparts only duringmaximal concentric contractions (19% to 32%,p <0.05). Then, the take-off leg of HJmen was strongerthan the raised leg during isometric condition (17%,p < 0.01). HJ women were stronger than their coun-terparts (51 to 68%,p < 0.01) and a group by veloc-ity interaction (p < 0.01) indicated a flattened torque-velocity curve i.e. no difference between isometric andlow-velocity torques for untrained women in contrastwith athletes (Fig. 2).

During knee extensions, absolute RMS EMG waslower for women compared with men (mean difference:58%,p < 0.01). The normalized RMS EMG of rec-tus femoris increased significantly with velocity about30–47% from isometric value (p < 0.001) but was notinfluenced by side, gender or group (p > 0.05). Neitherabsolute nor normalized coactivation of biceps femoriswas different upon gender, group, side or velocity ac-counting on average for 29% of EMG activity obtainedduring maximal isometric flexions (p > 0.05).

3.2.2. Knee flexionKnee flexion torques decreased significantly with ve-

locity (p < 0.001) and were 46% higher for HJ groupcompared to controls (p < 0.001) (Fig. 3). Men werestronger than women (+45%,p < 0.001) but there wasno significant side effect (p > 0.05). During knee flex-ions, normalized BF EMG activity increased with ve-locity about 15–23% maximal isometric agonist RMSEMG (p < 0.05) but was unaltered by either group,gender or side (p > 0.05). Absolute RMS EMG ofBF was lower for women compared with men (meandifference: 87%,p < 0.01). RF coactivation reachedon average 10% of maximal RMS EMG obtained dur-ing maximal isometric knee extensions irrespective ofgender, group or side (p > 0.05).

4. Discussion

The aim of this study was to examine whether asym-metry of the strength of the muscles crossing the kneejoints was more prominent in high jumper athletesthan in controls. We expected greater functional dif-ferences between take-off and raised legs in competi-tive high jumpers without gender interaction. Our re-sults revealed that male high jumper athletes were moreasymmetric during maximal isometric knee extensionsthan controls but not females. The following discus-sion will focus on the interpretation of these findings


0.00

0.10

0.20

0.30

0.40

0.50

HJmen HJwomenCmen Cwomen

Hei

ght(

m)

0.00

0.10

0.20

0.30

0.40

0.50

HJmenHJmen HJwomenHJwomenCmenCmen CwomenCwomen

Hei

ght(

m)

Fig. 1. SJ (white), CMJ (black) height for high jumpers and controls. CMJ performances were higher than SJ with respect to group and gender(p < 0.001). Men were stronger than women (p < 0.001) and high jumpers were stronger than controls (p < 0.001).

in relation to current knowledge on plyometric train-ing induced-adaptations and functional characteristicsof power-trained athletes.

A limited number of reports in the literature areavailable on functional asymmetry in well-trained ath-letes and their results are conflicting. Lateral isokinetictorque balance in well-trained athletes have been ob-served for trunk muscles in elite sailors [8] and limbmuscles engaged in sports such as skiing, rowing, run-ning, gymnastic or even in long and triple jumpers [15,20,36,52]. Similarly to non-trained subjects or physi-cal students [15,50], these aforementioned studies didnot show significant differences between dominant andnon-dominant legs. From the available data, lateralasymmetry has not been displayed in sedentary subjectsor athletes practising rather bilateral sports given thatboth legs are equally used either moderately in dailylife or maximally during such sports. In contrast, somestudies indicated higher maximal torque generated bythe dominant upper limb in tennis, water-polo playersand canoeists [7,56,61]. Lateral dominance of lowerlimbs has mainly been examined in soccer players butthe results are equivocal, partly related to gender dif-ferences or level of practise. Indeed, some studies havereported higher hamstring and/or quadriceps muscletorques in favour of the kicking leg for elite male soccerplayers depending on their field position [13,25,29,33,44]. In contrast, this was not the case for younger nineto 21-years-old players [34,35] or for females [2]. Ourfindings are in agreement with those of previous studiesindicating a rather low but significant (<10%) lateraldominance in favour of the dominant leg in athletes [7,13,25,33,56,61]. To the best of our knowledge, this is

the first cross-sectional study examining the effects ofchronic high jumping on functional asymmetry of mus-cles crossing the knee joint. Our findings showed thatthe knee extensor muscles of the dominant leg i.e. thetake-off leg generated greater muscle force than its con-tralateral i.e. the raised leg during isometric conditiononly (+17%,p < 0.01). In contrast, no such lateral-ity was found for concentric knee extension torques aswell as for knee flexion torques. These findings suggestthat there was a muscle- and velocity-specific lateraldominance in competitive high jumper athletes. Fur-thermore, this functional imbalance was not observedfor high jumper females. Indeed, all males but onlythree female among five high jumpers exhibited thisasymmetry in quadriceps MVC. A gender-differencein terms of functional laterality due to training adapta-tions has also been reported for isokinetic elbow flexortorques in junior tennis players and in female soccerplayers [2,56]. In the present study, the higher level ofexpertise of high jumper men according to the currentinternational table (Table 2) may have mainly explainedthis sex-dependence.

Although it is quite possible that this lateral domi-nance was genetically conferred, alternatively it may bethe result of specific adaptations related to the preferen-tial use of the dominant limb. The rather higher torqueimbalance for male high jumpers compared with pre-vious results based on other athletics populations givessupport to this lateral training-induced specificity as re-ported by cross-sectional studies on unilateral training(e.g. Taniguchi [64]). During actual high jump com-petitive event, it has been recently shown that the freelimbs (arms and raised leg) in very high skilled high


Speed (o.s-1)

Speed (o.s-1)

Tor

que

(N.m

)

High jumper me

#

80

130

180

230

280

330

0 60 105 180

80

130

180

230

280

330

0 60 105 180

High jumper wome

50

100

150

200

250

300

0 60 105 180

50

100

150

200

250

300

0 60 105 180

Take-off leg

Raised leg

Tor

que

(N.m

)

*

†

‡

‡

††

‡ ‡

‡ ‡

†

* *

**

†

†

† †

†

†

Tor

que

(N.m

)

High jumper men

#

80

130

180

230

280

330

0 60 105 180

80

130

180

230

280

330

0 60 105 180

High jumper women

50

100

150

200

250

300

0 60 105 180

50

100

150

200

250

300

0 60 105 180

Take-off leg

Raised leg

Tor

que

(N.m

)T

orqu

e (N

.m)

High jumper me Control men

#

80

130

180

230

280

330

80

130

180

230

280

330

0 60 105 180

80

130

180

230

280

330

0 60 105 180

80

130

180

230

280

330

0 60 105 180

High jumper wome Control women

50

100

150

200

250

300

0 60 105 180

50

100

150

200

250

300

0 60 105 180

50

100

150

200

250

300

0 60 105 180

50

100

150

200

250

300

0 60 105 180

50

100

150

200

250

300

0 60 105 180

50

100

150

200

250

300

0 60 105 180

Take-off leg

Raised leg

Tor

que

(N.m

)

*

†

‡

‡

††

‡ ‡

‡ ‡

†

* *

**

†

†

† †

†

†

Fig. 2. Knee extension torque – velocity curves for take-off leg and raised leg in high jumper men and women; in male and female sedentarysubjects. Men were stronger than women (p < 0.001) and high jumpers were stronger than controls (p < 0.01). #p < 0.01 significant betweenleg difference. ∗, †, ‡: p < 0.05, p < 0.01, p < 0.001: significant difference from maximal isometric torque (0.s−1).

jumper men contributedonly 8% to the vertical velocityat take-off, pointing out the importance of the take-offleg in generating muscle power during this particularunilateral jump [1]. Thus, the training regimen for highjumpers consists of bilateral moderate to heavy-loadplyometric exercises that are the essential factors im-proving vertical jump performance combined with re-sistance and flexibility training [26,47,54]. In addition,unilateral stretch-shortening cycle exercises performedwith the take-off leg account for a non-negligible partof the training program.

Although the underlying mechanisms responsible for

this lateral specificity cannot be directly assessed in thepresent study, increases in maximal voluntary torqueare usually associated with adaptations occurring in thecentral nervous system and/or at the muscle level [57].A number of investigations have utilized SEMG to ex-amine the neural adaptations associated with increasein torque indicating either increases [12,51,57] or nochanges with strength training [31,58]. Our results in-dicated that neural activation of rectus femoris (nor-malized RMS SEMG) during maximal knee exten-sions was not different across high jumpers and con-trols whereas torque differences ranged between 15 to


Tor

que

(N.m

)

High jumper men

Take-off leg

Raised leg

High jumper wome

Tor

que

(N.m

)

†

20

40

60

80

100

120

140

160

0 60 105 180

20

40

60

80

100

120

140

160

0 60 105 180

40

60

80

100

120

140

160

180

0 60 105 180

40

60

80

100

120

140

160

180

0 60 105 180

Tor

que

(N.m

)

High jumper me Control men

Take-off leg

Raised leg

Take-off leg

Raised leg

High jumper women Control women

Tor

que

(N.m

)

†

20

40

60

80

100

120

140

160

0 60 105 180

20

40

60

80

100

120

140

160

0 60 105 180

20

40

60

80

100

120

140

160

0 60 105 180

40

60

80

100

120

140

160

180

40

60

80

100

120

140

160

180

0 60 105 180

40

60

80

100

120

140

160

180

0 60 105 180

40

60

80

100

120

140

160

180

0 60 105 180

Speed (o.s-1)

Speed (o.s-1)

Fig. 3. Knee flexion torque – velocity curves for take-off leg and raised leg in high jumper men and women; in male and female sedentarysubjects. Torque decreased significantly with speed (p < 0.001). Men were stronger than women (p < 0.001) and high jumpers were strongerthan controls (p < 0.001). No between-leg difference was found.

30% for athletes males. Similarly, Amiridis et al. [28]showed that EMG amplitude of the dominant mono-articular knee extensor muscles (vastus medialis andlateralis, VM and VL respectively) was not differentacross high jumpers and sedentary subjects during max-imal isometric and concentric extensions even though a28 to 44% higher torque generation was found for highjumpers at 14 angular velocities (−120 to 300.s−1).In contrast, these authors showed that EMG amplitudeof VL and VM was significantly lower during eccentricactions compared with isometric activity only in seden-tary subjects, suggesting the presence of a tension-

regulating mechanism in untrained controls as reportedpreviously using neurostimulation techniques [19,21,28,55]. Recent evidence indicates that the level of an-tagonistic cocontraction is modifiable with training thathas also been proposed as a mechanism responsiblefor the greater torque-producing capabilities of mus-cle after isometric training [6] in contrast with eccen-tric training effect [49]. On competitive high jumpers,Amiridis et al. [28] reported a three-fold lower levelof activation of the internal knee flexor muscle (semi-tendineous, ST) vs. untrained (13% vs. 38%) indepen-dent of velocity or action type. On the contrary, our


findings indicated similar antagonistic coactivation ofthe biceps femoris across groups or between dominantand non-dominant legs (≈29% of agonist maximal ac-tivity) probably due to functional differences in muscleinvestigated (BF vs. ST). However, the observed lateraldifference in high jumpers MVC can not be ascribed to alikely difference in the coactivation level of ST betweenthe take-off leg and raised leg because a similar differ-ence in torque would have also been expected duringconcentric actions (60 to 180.s−1). These above ob-servations suggest that agonistic and antagonistic neu-ral activation may not have been the mechanisms un-derlying the lateral dominance observed in high jumpermales. However, a recent study showed that voluntaryEMG measured during maximal contraction as an indexof the level of neural activation is quite questionablewhen comparing endurance-, power-trained and seden-tary subjects [19]. Despite the lack of difference acrossgroups in the EMG amplitude in VL and VM mus-cles measured during maximal isometric contraction,athletes exhibited a significant higher activation level(≈15%) using superimposed twitch technique [19].

Another interesting finding was the lack of between-leg difference in peak torque at the faster concen-tric speed (180.s−1) probably related to similar fibretype proportion between the take-off and raised leg inhigh jumpers. The higher torque differences with in-creased velocity between athletes and sedentary sub-jects (Fig. 1) indicate a higher percentage of fast fibresin power athletes compared with untrained men [22,25] that has also been significantly related to squatand countermovement performances [9]. Indeed, re-cent studies have demonstrated that maximal isokineticknee extension torque normalized to muscle volumeobtained at velocity of movement higher than 120.s−1

was positively related to the proportion of fast twitch fi-bres [3,25,42]. These observations prompted us to sup-pose that this lateral isometric specificity in male highjumpers is related to other factors than neural activationand fiber type transition.

The maximal force-generating capacity during iso-metric and low-concentric actions (<30.s−1) of agiven muscle is generally related to its physiologi-cal cross-sectional area (PCSA) and have been exten-sively discussed [24,41,46,53]. Hypertrophy is also as-sociated with increased muscle fiber pennation angleand/or muscle thickness [14,43,62]. Recently, Kearnset al. [14] reported significant higher fascicle lengthand muscle thickness of gastrocnemius medialis of thekicking leg in soccer players. Theoretically, an increasein pennation angle and/or muscle thickness allows for

attachment of more contractile tissue to the tendon,which may lead to increased strength. Therefore, it canbe speculated that difference in MVC could be due to agreater PCSA of the dominant knee extensor muscles.

Alternatively, a difference in terms of torque-angleshape may have partly explained this lateral domi-nance in isometric condition only. Inter-subject dif-ferences (61± 4) have been reported on the samemuscle group for untrained individuals [63]. Similarly,cross-sectional studies have shown different torque-angle curve across different athlete’s populations suchas runners and cycling athletes or skaters [27,59] oreven for non-trained subjects [40] of their dominantlimb. These observations give support to the length-at-use hypothesis that specific training can modify theshape of the MVC-angle relationship. This might beexplained by between-leg difference in tendon stiffnessbecause during even fixed-end isometric conditions,muscle fibers shorten until the slack of series elasticcomponents such as tendinous tissues is taken up [37].Thus, a lower tendon stiffness of the non-dominantquadriceps femoris would result in a leftward shift ofthe torque-angle curve of the knee extensors. Thiswould place the contractile component at a non-optimalpoint on the force-length curve [18] and thus alteringtheir potential for joint moment generation. Lower ten-don stiffness would also delay the angle at peak torqueduring dynamic contractions [63]. In the present ex-periment, high jumpers exhibited a systematic angleat peak torque during isokinetic contractions at moreextended positions for the raised leg than the take-offleg (up to 11, data not reported) that gives support tothe aforementioned assumption. In addition, previousstudies have shown an increase in musculo-tendinousstiffness of upper and lower limbs measured during ac-tive conditions by quick-release or oscillations tech-niques after plyometric training probably related to anincrease in tendon stiffness [11,23,48]. A higher mus-culotendinous stiffness of the take-off leg may be suit-able for transmitting the force more effectively [17] andfurther enhance the release of potential energy duringthe push-off by shortening the coupling time i.e. thetime between touchdown and take-off. No data aboutadaptations of tendon properties before and after plyo-metric training in humans are available to support sucha hypothesis and requires further study.

5. Conclusion

Male high jumpers were more asymmetric than non-trained subjects during maximal isometric knee exten-


sions but not during concentric action type or kneeflexions. These results suggest that chronic high jumptraining may have induced greater imbalance in kneeextensor torque probably due to the preferential use ofthe leg at take-off over the raised leg. Our results seemto indicate that this lateral dominance is probably ex-plained by peripheral differences (muscle volume, pen-nation angle, tendon stiffness) rather than either neu-ral factors or fibre composition. Future studies exam-ining these assumptions on a larger sample size withsuitablein vivo technique such as ultrasonography arewarranted.

References

[1] A. Lees, J. Rojas, M. Ceperos, V. Soto and M. Gutierrez, Howthe free limbs are used by elite high jumpers in generatingvertical velocity,Ergonomics43 (2000), 1622–1636.

[2] A. Ostenberg, E. Roos, C. Ekdahl and H. Roos, Isokineticknee extensor strength and functional performance in healthyfemale soccer players,Scandinavian Journal of Medicine &Science in Sports8 (1998), 257–264.

[3] A. Thorstensson, G. Grimby and J. Karlsson, Force-velocityrelations and fiber composition in human knee extensor mus-cles,Journal of Applied Physiology40 (1976), 12–16.

[4] A.D. Walshe, G.J. Wilson and A.J. Murphy, The validity andreliability of a test of lower body musculotendinous stiff-ness,European Journal of Applied Physiology & Occupa-tional Physiology73 (1996), 332–339.

[5] A.J. Van Soest, M.E. Roebroeck, M.F. Bobbert, P.A. Huijingand G.J. van Ingen Schenau, A comparison of one-legged andtwo-legged countermovement jumps,Medicine and Sciencein Sports and Exercise17 (1985), 635–639.

[6] B. Carolan and E. Cafarelli, Adaptations in coactivation afterisometric resistance training,Journal of Applied Physiology73 (1992), 911–917.

[7] B. Humphries, G.A. Abt, R. Stanton and N. Sly, Kinanthro-pometric and physiological characteristics of outrigger canoepaddlers,Journal of Sports Sciences18 (2000), 395–399.

[8] B. Larsson, N. Beyer, P. Bay, L. Biond, P. Aagaard and M.Kjaer, Exercise performance in elite male and female sailors,International Journal of Sports Medicine17 (1996), 504–508.

[9] C. Bosco and P.V. Komi, Potentiation of the mechanical be-havior of the human skeletal muscle through prestretching,Acta Physiologica Scandinavica106 (1979), 467–472.

[10] C. Bosco, J. Tihanyi, P.V. Komi, G. Fekete and P. Apor, Storeand recoil of elastic energy in slow and fast types of humanskeletal muscles,Acta Physiologica Scandinavica116 (1982),343–349.

[11] C. Cornu, M.I. Almeida Silveira and F. Goubel, Influenceof plyometric training on the mechanical impedance of thehuman ankle joint,European Journal of Applied Physiology& Occupational Physiology76 (1997), 282–288.

[12] C. Thepaut-Mathieu, J. Van Hoecke and B. Maton, Myoelectri-cal and mechanical changes linked to length specificity duringisometric training,Journal of Applied Physiology64 (1988),1500–1505.

[13] C. Tourny-Chollet and D. Leroy, Isokinetic knee musclestrength of soccer players according to their position,Isoki-netics and Exercise Science8 (2000), 187–193.

[14] C.F. Kearns, M. Isokawa and T. Abe, Architectural character-istics of dominant leg muscles in junior soccer players,Euro-pean Journal of Applied Physiology85 (2001), 240–243.

[15] C.M. Siqueira, F.R. Pelegrini, M.F. Fontana and J.M. Greve,Isokinetic dynamometry of knee flexors and extensors: com-parative study among non-athletes, jumper athletes and runnerathletes,Revista do Hospital das Clinicas57 (2002), 19–24.

[16] D.G. Sale, Neural adaptation to resistance training,Medicineand Science in Sports and Exercise20 (1988), 135–145.

[17] D.L. Morgan, U. Proske and D. Warren, Measurements ofmuscle stiffness and the mechanism of elastic storage of energyin hopping kangaroos,The Journal of Physiology282 (1978),253–261.

[18] F.E. Zajac, Muscle and tendon: properties, models, scaling,and application to biomechanics and motor control,CriticalReviews Biomedical Engineering17 (1989), 359–411.

[19] G. Lattier, G.Y. Millet, N.A. Maffiuletti, N. Babault andR. Lepers, Neuromuscular differences between endurance-trained, power-trained, and sedentary subjects,Journal ofStrength and Conditioning Research Res17 (2003), 514–521.

[20] G. Neumayr, H. Hoertnagl, R. Pfister, A. Koller, G. Eibl andE. Raas, Physical and physiological factors associated withsuccess in professional alpine skiing,International Journal ofSports Medicine24 (2003), 571–575.

[21] G.A. Dudley, R.T. Harris, M.R. Duvoisin, B.M. Hather andP. Buchanan, Effect of voluntary vs. artificial activation onthe relationship of muscle torque to speed,Journal of AppliedPhysiology69 (1990), 2215–2221.

[22] G.G. Sleivert, R.D. Backus and H.A. Wenger, Neuromusculardifferences between volleyball players, middle distance run-ners and untrained controls,International Journal of SportsMedicine16 (1995), 390–398.

[23] G.J. Wilson, A.J. Murphy and J.F. Pryor, Musculotendinousstiffness: its relationship to eccentric, isometric, and concen-tric performance,Journal of Applied Physiology76 (1994),2714–2719.

[24] H. Akima, S. Kuno, H. Takahashi, T. Fukunaga and S. Kat-suta, The use of magnetic resonance images to investigate theinfluence of recruitment on the relationship between torqueand cross-sectional area in human muscle,European Journalof Applied Physiology83 (2000), 475–480.

[25] H. Gur, L. Gransberg, D. vanDyke, E. Knutsson and L. Lars-son, Relationship between in vivo muscle force at differentspeeds of isokinetic movements and myosin isoform expres-sion in men and women,European Journal of Applied Physi-ology88 (2003), 487–496.

[26] H. Toumi, T.M. Best, A. Martin, S. F’Guyer and G. Poumarat,Effects of eccentric phase velocity of plyometric training onthe vertical jump,International Journal of Sports Medicine25(2004), 391–398.

[27] H.H. Savelberg and K. Meijer, Contribution of mono- andbiarticular muscles to extending knee joint moments in runnersand cyclists,Journal of Applied Physiology94 (2003), 2241–2248.

[28] I.G. Amiridis, A. Martin, B. Morlon, L. Martin, G. Cometti,M. Pousson and J. van Hoecke, Co-activation and tension-regulating phenomena during isokinetic knee extension insedentary and highly skilled humans,European Journal ofApplied Physiology & Occupational Physiology73 (1996),149–156.

[29] J. Magalhaes, J. Oliveira, A. Ascensao and J. Soares, Concen-tric quadriceps and hamstrings isokinetic strength in volley-ball and soccer players,The Journal of Sports Medicine andPhysical Fitness44 (2004), 119–125.


[30] J.P. Van Zandwijk, M.F. Bobbert, M. Munneke and P. Pas,Control of maximal and submaximal vertical jumps,Medicineand Science in Sports and Exercise32 (2000), 477–485.

[31] J.P. Weir, T.J. Housh and L.L. Weir, Electromyographic evalu-ation of joint angle specificity and cross-training after isomet-ric training, Journal of Applied Physiology77 (1995), 197–201.

[32] J.T. Viitasalo and C. Bosco, Electromechanical behaviour ofhuman muscles in vertical jumps,European Journal of AppliedPhysiology & Occupational Physiology48 (1982), 253–261.

[33] K. Masuda, N. Kikuhara, H. Takahashi and K. Yamanaka, Therelationship between muscle cross-sectional area and strengthin various isokinetic movements among soccer players,Jour-nal of Sports Sciences21 (2003), 851–858.

[34] L. Capranica, G. Cama, F. Fanton, A. Tessitore and F. Figura,Force and power of preferred and non-preferred leg in youngsoccer players,The Journal of Sports Medicine and PhysicalFitness32 (1992), 358–363.

[35] L. Hansen, J. Bangsbo, J. Twisk and K. Klausen, Developmentof muscle strength in relation to training level and testosteronein young male soccer players,Journal of Applied Physiology87 (1999), 1141–1147.

[36] M. Amato, F. Lemoine, J. Gonzales, C. Schmidt, P. Afriat andP.L. Bernard, Influence of age and physical activity on isoki-netic characteristics of hamstring and quadriceps muscles ofyoung gymnasts and soccer players,Annales de Readaptationet de Medecine Physique44 (2001), 581–590.

[37] M. Ito, Y. Kawakami, Y. Ichinose, S. Fukashiro and T. Fuku-naga, Nonisometric behavior of fascicles during isometric con-tractions of a human muscle,Journal of Applied Physiology85 (1998), 1230–1235.

[38] M. Voigt, E.B. Simonsen, P. Dyhre-Poulsen and K. Klausen,Mechanical and muscular factors influencing the performancein maximal vertical jumping after different prestretch loads,Journal of Biomechanics28 (1995), 293–307.

[39] M.F. Bobbert and G.J. van Ingen Schenau, Coordination invertical jumping,Journal of Biomechanics21 (1988), 249–262.

[40] M.J. Pavol andM.D. Grabiner, Knee strength variability be-tween individuals across ranges of motion and hip angles,Medicine and Science in Sports and Exercise32 (2000), 985–992.

[41] M.V. Narici, G.S. Roi, L. Landoni, A.E. Minetti and P. Cer-retelli, Changes in force, cross-sectional area and neural ac-tivation during strength training and detraining of the humanquadriceps,European Journal of Applied Physiology & Oc-cupational Physiology59 (1989), 310–319.

[42] P. Aagaard and J.L. Andersen, Correlation between contrac-tile strength and myosin heavy chain isoform composition inhuman skeletal muscle,Medicine and Science in Sports andExercise30 (1998), 1217–1222.

[43] P. Aagaard, J.L. Andersen, P. Dyhre-Poulsen, A.M. Leffers,A. Wagner, S.P. Magnusson, J. Halkjaer-Kristensen and E.B.Simonsen, A mechanism for increased contractile strengthof human pennate muscle in response to strength training:changes in muscle architecture,The Journal of Physiology15(2001), 613–623.

[44] P. Mognoni, M.V. Narici, M.D. Sirtori and F. Lorenzelli, Isoki-netic torques and kicking maximal ball velocity in young soc-cer players,The Journal of Sports Medicine and Physical Fit-ness34 (1994), 357–361.

[45] P.V. Komi and C. Bosco, Utilization of stored elastic energyin leg extensor muscles by men and women,Medicine andScience in Sports and Exercise10 (1978), 261–265.

[46] R.J. Maughan, J.S. Watson and J. Weir, Strength and cross-sectional area of human skeletal muscle,The Journal of Phys-iology 338 (1983), 37–49.

[47] R.U. Newton, W.J. Kraemer and K. Hakkinen, Effects of bal-listic training on preseason preparation of elite volleyball play-ers,Medicine and Science in Sports and Exercise31 (1999),323–330.

[48] R.W. Spurrs, A.J. Murphy and M.L. Watsford, The effect ofplyometric training on distance running performance,Euro-pean Journal of Applied Physiology89 (2003), 1–7.

[49] S. Colson, M. Pousson, A. Martin and J. Van Hoecke, Isoki-netic elbow flexion and coactivation following eccentric train-ing, Journal of electromyography and kinesiology9 (1999),13–20.

[50] S. Demura, S. Yamaji, F. Goshi and Y. Nagasawa, Lateraldominance of legs in maximal muscle power, muscular en-durance, and grading ability,Perceptual and Motor Skills93(2001), 11–23.

[51] S. Garfinkel and E. Cafarelli, Relative changes in maximalforce, EMG, and muscle cross-sectional area after isomet-ric training,Medicine and Science in Sports and Exercise24(1992), 1220–1227.

[52] S. Parkin, A.V. Nowicky, O.M. Rutherford and A.H. McGre-gor, Do oarsmen have asymmetries in the strength of theirback and leg muscles?,Journal of Sports Sciences19 (2001),521–526.

[53] S.E. Alway, J. Stray-Gundersen, W.H. Grumbt and W.J.Gonyea, Muscle cross-sectional area and torque in resistance-trained subjects,European Journal of Applied Physiology &Occupational Physiology60 (1990), 86–90.

[54] S.E. Blattner and L. Noble, Relative effects of isokineticand plyometric training on vertical jumping performance,Re-search Quarterly50 (1979), 583–588.

[55] S.H. Westing, A.G. Cresswell and A. Thorstensson, Muscleactivation during maximal voluntary eccentric and concentricknee extension,European Journal of Applied Physiology &Occupational Physiology62 (1991), 104–108.

[56] T. Ellenbecker and E.P. Roetert, Age specific isokinetic gleno-humeral internal and external rotation strength in elite juniortennis players,Journal of Sports Science & Medicine6 (2003),63–70.

[57] T. Moritani and H.A. deVries, Neural factors versus hypertro-phy in the time course of muscle strength gain,The AmericanJournal of Physical Medicine58 (1979), 115–130.

[58] T.K. Evetovich, T.J. Housh, G.O. Johnson, D.B. Smith,K.T. Ebersole and S.R. Perry, Gender comparisons of themechanomyographic responses to maximal concentric and ec-centric isokinetic muscle actions,Medicine and Science inSports and Exercise30 (1998), 1697–1702.

[59] W. Herzog, A.C. Guimaraes, M.G. Anton and K.A. Carter-Erdman, Moment-length relations of rectus femoris musclesof speed skaters/cyclists and runners,Medicine and Science inSports and Exercise23 (1991), 1289–1296.

[60] W.B. Young, C. MacDonald and M.A. Flowers, Validity ofdouble- and single-leg vertical jumps as tests of leg exten-sor muscle function,Journal of Strength and ConditioningResearch15 (2001), 6–11.

[61] W.C. McMaster, S.C. Long and V.J. Caiozzo, Isokinetic torqueimbalances in the rotator cuff of the elite water polo player,The American Journal of Sports Medicine19 (1991), 72–75.

[62] Y. Kawakami, T. Abe and T. Fukunaga, Muscle-fiber penna-tion angles are greater in hypertrophied than in normal mus-cles,Journal of Applied Physiology74 (1993), 2740–2744.


[63] Y. Kawakami, T. Muraoka, S. Ito, H. Kanehisa and T. Fuku-naga, In vivo muscle fibre behaviour during counter-movementexercise in humans reveals a significant role for tendon elas-ticity, The Journal of Physiology540 (2002), 635–646.

[64] Y. Taniguchi, Lateral specificity in resistance training: theeffect of bilateral and unilateral training,European Journalof Applied Physiology & Occupational Physiology75 (1997),144–150.

Isokinetics and Exercise Science 13 (2005) 38 38IOS Press

Abstract

Relation between muscle strength and age ina sample population of 80 soccer playingchildren aged from 10 to 15 years

P. Calmels, R. Richard, R. Ouillon, D. Teillol and G. CarrotService MPR et Service Physiologie et Medecine du Sport, Hopital Bellevue, CHU de Saint Etienne, F. 42055,France

Introduction: The evolution of muscle strength withage is very well documented for adult, but not for chil-dren. Gobelet reports a decrease of the knee flex-ors/extensors ratio between 10 to 15 years, then an in-crease under the influence of training [1]. Kellis etal. [2] report for young soccer players a correlation be-tween dominant knee flexors and extensors and age,height, weight and puberty.Objectives: To analyse the evolution of knee flexorsand extensors of young soccer players and to researcha relation with age, morphology and puberty.Population: 80 young subjects, without any disease,voluntary, who take part in a special school for practiceand education to sports, and aged from 10 to 15 years(age mean 12,78 years +/− 2,88).Method: For each subject are performed a clinical eval-uation (age, weight, height, BMI, Tanner scale of pu-berty) and a bilateral evaluation of knee flexors and ex-tensors muscle strength with an isokinetic dynamome-ter.Results: There is a strong significant correlation (p <0.0001) between age, height, weight, BMI and Tannerlevel of puberty and all peak torques at different an-gular velocities. Between 11 and 15 years, the peaktorque values increase from 50% and more between12 to 14 years. The knee flexors/extensors ratios areequivalent but there decrease significantly between 14

an 15 years. The ANOVA reveals that puberty levelis an important factor on the different peak torques forall angular velocities (p <0.0001), but without effecton the ratio values. Muscle strength increases largelybetween Tanner level 1 and level 5,particularly betweenlevel 2 and level 4.Discussion: The evaluation of muscle strength duringgrowth is difficult and requires longitudinal studies.There is a large individual variation. The increase ofmuscle strength between 11 to 15 years is in accordwith others results in the literature. But, for you, nostudy has very well analysed the evolution of pubertyfor this sample of young sportsmen. But, this periodis clinically difficult to analyse, and it will be probablynecessary to associated to our analyse, an hormonalmeasure of testosterone and a radiological evaluationof the puberty level.

References

[1] C.H. Gobelet, Force isocinetique: de l’enfant a l’adulte, in:Actualite en reeducation fonctionnelle et readaptation, Masson,Ed. Paris, 1985, pp. 49–54.

[2] S. Kellis, E. Kellis, V. Manou and V. Gerodimos, Predictionof knee extensor and flexor isokinetic strength in young malesoccer players, Journal of Orthopaedic Sports and PhysicalTherapy 30(11) (2000), 693–701.



Interest of eccentric isokinetic exercises incases of calcanean tendinosis and thighmuscular injuries: Prospective study results

C. Queiros Da Silva, T. Cotte, L. Vicard, L. Chantelot and J.-M. FerretCentre de Medecine du Sport de Lyon Gerland (CMSLG), France

Abstract. Musculo tendinous injuries are frequent pathologies in sport medicine. Rest and classical treatments are, very often,not enough to succeed a good healing and return to sport activities. The aim of this prospective study is to demonstrate theinterest of eccentric isokinetic exercise in the treatment of sub acute injuries of the calcanean tendon and of the thigh muscles.Complementary with classical kinesitherapy treatment, isokinetic work allowed a complete healing and return to anterior sportactivities, without any recurrent injuries for a majority of the patients. We suggest in conclusion a specific protocol to care thesepathologies using an isokinetic dynamometer in eccentric mode.

Keywords: Calcanean tendon, muscle, reeducation protocol, isokinetic, eccentric, prospective study, sport medicine

1. Introduction

Tendinitis and muscle injuries occur frequently insport medicine: with professional soccer players, mus-cle injuries represent 34% of accidents and tendonitis4% [6,14]; hamstrings injuries correspond to 50%of muscular injuries for runners and 40% for soc-cers. Main etiological factors consist in: static unbal-ance [11], muscular unbalance (bilateral asymmetry,unbalance ratio between agonist and antagonist) [4],impairment of musculotendinous stretchness [11,12,16], poor hydratation, poor practise conditions,duringsport activity (equipment, environmental conditions,training) [11]. William Stanish, used the first, eccen-tric exercises for the treatment of patella tendinitis andcalcanean tendinosis (CT) [15], showing that tractionexercises in eccentric conditions allow to improve thetendinous eccentric resistance to stretch when occur-ring during sport activities. More precisely, eccentricexercises contribute to improve tendinous and muscu-lar injuries healing process. These eccentric exercis-es stimulate the fibroblast and Mauro’s satellite cellsgrowup and help to the alignment of the new fiber in theaxis of traction [8,9]. Alfredson reports in 2003 excel-

lent results using manual and daily eccentric exercisesto care chronic CT [1].

The 3 aims of this prospective study [13] are:

1. To confirm the interest of eccentric exercises atour clinic (Centre de Medecine du Sport de LyonGerland-cmslg), thanks to the data of an isoki-netic dynamometer to evaluate patient’s and PT’sperceptions along the period of treatment.

2. To precise the value of data allowing the return tosport activities.

3. To elaborate appropriate eccentric protocol on theisokinetic dynamometer in conjunction to the re-sults obtained.

2. Materials and methods

2.1. Patient population (Table 1)

Twenty-one patients aged 30+/− 9 years (18 men,2 females, 1 teenager); 11 soccer players, 5 runners, 5others sports.

Nine of them suffer calcanean tendinosis (8 of midportion tendon, 1insertional calcanean tendon pain).


40 C. Queiros Da Silva et al. / Interest of eccentric isokinetic exercises in cases of calcanean tendinosis

Table 1Study patient caracteristics

Calcanean tendon Quadriceps Hamstrings TOTAL(n = 9) (n = 4) (n = 8) (n = 21)

Men Females Men Females Men Females Men Females(n = 8) (n = 1) (n = 3) (n = 1) (n = 8) (n = 0) (n = 19) (n = 2)

Age (years) 34± 5 30 33± 11 24 25± 8 30± 9 27± 3Weight (kg) 82± 9 70 78± 11 60 72± 9 78± 10 65± 5Anteriority of 64 182 72 9 30 51 95.5symptomatology (days) (2–135) (20–165) (7–120) (2–165) (9–182)Sport 3 Runners Runners 1 Runners Soccers 5 Soccers 4 Runners 1 Runner

3 Soccers 2 Soccers 3 Others 10 Soccers 1 Soccer2 Others 5 Others

Anterior treatment 3 0 0 0 1 4 0

Eight of them suffer hamstrings muscular injuries(tears or level 3).

Four of them suffer quadriceps muscular injuries (3level 3, 1 level4-partial rupture).

All these musculotendinous pathologies lasted lessthan 6 months. It was either a first care or consecutiveto anterior treatment failure (sport rest, kinesitherapy,NSAI).Positive diagnosis result from clinical examina-tion and echography or RMI for 50% of the patientgroup.

2.2. Isokinetic equipment

The isokinetic dynamometer is a CYBEXNorm.It is operated in CPM mode. The patient is taught toresist (eccentrically) against the dynamometer move-ment with the muscle involved and to be passive duringthe movement back to the initial position.

Patient positioning depends on muscle involved:

– calf contraction is realised in prone position, kneestraight; max ROM is from 50 plantar flexion to20 dorsiflexion.

– soleus contraction in supine position, hip and knee90 flexed.

– hamstrings contraction in seated position, fromanatomical zero to 135 flexion max.

– quadricep contraction in seated position, fromanatomical zero to 135 flexion max.

– hip flexors contraction in supine position, fromanatomical zero to 120 flexion max

At any time the controlateral limb is fixed andstrapped to avoid join center of rotation displacement.

3. Experimental protocol

– number of treatment session depends of each pa-tient and symptomatology evolution but minimumfrequency is 2 session per week.

– each session of eccentric exercise on dynamome-teris associated with classical kinesitherapy (cryo-therapy, physiotherapy, nsai, deep tranvers mas-sage, progressive passive musculotendinous stre-tching, manual eccentric exercise and propriocep-tion exercises).

– The 1st day of treatment, patients benefit of isoki-netic eccentric assessment on non involved side atthe 3 speeds used on reeducation (5/sec, 10/secand 15/sec).

– Initial force assessment is only on non involvedside to prevent pain and myo tendinous healing.

– No pain rule is strictly respected during all thetreatment period. If such occurs during the ses-sion or after, we came back to none pain anteriorlevel of exercise and slowed the progression of thetreatment.

– 10 minutes of warm up on a bicycle is achievedbefore each session, cool down, recoveryand mus-cular passive stretching follows each of the ses-sion.

– each session consists in an eccentric exercise forhealing goal at 5/sec of speed , ROM is 45%of max, with torque survey around−50 Nm (av-erage). The different parameters (torque, ROM,speed and number of repetitions) are increase pro-gressively staying under the pain level; only oneparameter is increase at the same time. During thetreatment speed is increased from 5/sec to 10/secand then to 15/sec. ROM are patient dependantinside the conditions above described.

– each of the session has different number of repeti-tions depending of the speed: 4 series of 8 reps at5/sec, 5series of 8 reps at 10/sec and 6 series of8 reps at 15/sec.

– Rest time between series is from 30 sec to 60sec.

C. Queiros Da Silva et al. / Interest of eccentric isokinetic exercises in cases of calcanean tendinosis 41

Tabl

e2

Mea

ndu

ratio

nof

mus

culo

-ten

dino

usre

educ

atio

n

Dia

gnos

isS

ympt

omat

olog

yT

reat

men

tdur

atio

nR

etur

nto

run

Ret

urn

totr

aini

ngR

etur

nto

com

petit

ion

dura

tion

Num

ber

ofR

eedu

catio

nN

umbe

rof

Ree

duca

tion

PT

(%no

nN

umbe

rof

Ree

duca

tion

PT

(%no

nN

umbe

rof

Ree

duca

tion

PT

(%no

nse

ssio

nspe

riod

sess

ions

perio

din

volv

edsi

de)

sess

ions

perio

din

volv

edsi

de)

sess

ions

perio

din

volv

edsi

de)

Ent

heso

path

yT.

calc

anee

n13

days

203.

5m

onth

s(n

=1)

Tend

inop

athy

2m

onth

s19

3m

onth

s12

2m

onth

s76

%16

2.5

mon

ths

83%

204

mon

ths

114%

Cor

pore

ale

(±1.

4)(±

5)(±

1)(±

4)(±

0.7)

(±5)

(±0.

8)(±

3)(±

1)2

patie

nts

(n=

8)

Ham

strin

gs1

mon

ths

1245

days

731

days

77%

835

days

91%

1159

days

98%

(n=

8)(±

1)(±

2)(±

7)(±

2)(±

8)(±

1)(±

7)(±

2)(±

16)

Qua

dric

eps

2m

onth

s18

3m

onth

s9

31da

ys79

%14

2.5

mon

ths

79%

163

mon

ths

92%

(n=

4)(±

2)(±

5)(±

1)(±

3)(±

17)

(±5)

(±1)

(±4)

(±1)


4. Evaluation procedure

The prospective study at the CMSLG lasted 6months. Each of the patient benefit a weekly follow-up by the clinician. An enquiry form was filled up,including comfort score, pain score and daily activitylevel items. It was done: at the beginning of the treat-ment, during the treatment session and when return tosport activities. Usual clinical examinations followedthe questionary (inspection, palpation, isometric con-traction against resistance, passive stretching).

The weekly examination was associated to a regularclinical survey operated by the sport clinician specialistreferent.

After 6 months, each patient has been interviewed toevaluate the sport activities return conditions.

After 9 months (3 months later the first interview)a new interview has been done for eventual recurrentproblem.

5. Results

Clinical evolution (Table 2)

5.1. Calcanean tendinosis (CT)

Five among the 9 patients recovered clinical normal-isation (pain free during daily life and clinical exami-nation without anormality), 3 among the 9 shown clini-cal improvement (pain free during daily life, persisten-cy of tendon thickness or sensitivity during palpation)Theses improvements occurred 2 months (average 12sessions) after the beginning of the treatment with apeak torque value around 76% of controlateral value.This period of time corresponds to the return to run.

One among the 9 patients (the subject with insertion-al calcanean tendon pain) didn’t get any clinical im-provement in spite of correct torque isokinetic values.

Patients get back to sport training with a peak torquevalue around 83% of the controlateral value corre-sponding to 16 sessions of isokinetic eccentric exercise.Fifty percent of the patients with CT were runners, theycame back to their specific sport activity one week afterthe run recovery. Only 2 soccer players achieved up to20 sessions of treatment, so they return to competitionwith an isokinetic peak torque at 114% of controlateralvalue.

Among the 9 patients,

– 5 patients recover the same level of sport activitythan before injury.

– 2 of the 9 return to competition with a differentfeeling concerning the tension and thickness ofcalcanean tendon while 2 patients did not succeedto recover their previous sport: one because of aintercurrent pathology, the other one because ofthe absence of improvement in his enthesopathyof calcanean tendon.

To summary during an average of 8 months (3 to 16months) period of follow up only one among the 9 pa-tients presented a recurrent pathology. This patient hada CT and complained about discomfort at his soleusinsertion on involved side during the reeducation time.Echography did not show any lesion. This patient re-ported a recurrent pain on soleus 2 months after returnto sport activities.

5.2. Muscles injuries (MI)

After 10 sessions (5 weeks average) of reeducation,9 patients among the 12, succeed clinical normalisation(i.e. no pain at all during daily life and clinical exami-nation without abnormality). 3 patients among the 12reported clinical improvement (i.e. no pain during dailylife, tension feeling on muscular envelop without signat clinical examination).

Return to run has been effective when involved isoki-netic peak torque reached 78% of controlateral side.This result happened after an average time of 31 dayscorresponding to 8 eccentric isokinetic sessions (aver-age).

Sport training return has been effective when in-volved isokinetic peak torque reached 85% of contro-lateral side corresponding to 11eccentric isokinetic ses-sions.

Competition return has been effective at 95% of con-trolateral peak torque, corresponding to 14 sessions(average).

Nine patients among the 12 recorvered the same lev-el sport activity as before the pathology, even better(better feeling, increase in muscular performances). 2patients return with apprehension feeling but withoutpain. 1 patient faced to intercurrent pathology duringthe reeducation period and interrupted sport activity.

During 8 months period of follow up no one of thepatients reported recurrent lesion.

5.3. Follow up of isokinetic dynamometer results

At the end of reeducation, CT patients reported an in-crease of involved side isokinetic peak torque of 295%compared to initial values.

C. Queiros Da Silva et al. / Interest of eccentric isokinetic exercises in cases of calcanean tendinosis 43

These values correspond to 102% of controlateralside.

MI patients reported an increase of 196% comparedto initial value and 95% of controlateral side.

6. Discussion

First of all we faced a long delay between the begin-ning of the pathology and the beginning of our treat-ment: 3 months for CT, 2 months for MI.

This long delay results from patient’s negligence orfailure of anterior classical treatments.

In spite of this long delay, the functional recovery ofmusculo tendinous structures after eccentric isokineticexercises has been completed, in comparison to the in-volved side either for clinical approach as for the mus-cle force (respectively 102% for CT, 93% for muscleinjury). These results are a good point to prevent therecurrent risk, as Croisier [5] results.

Back to running exercises is possible early (2 monthsfor CT, 1 month for muscles injuries), but return toanterior sport training level and to competition levelis delayed significantly except for runners (probablybecause of no need of technical exercises comparedto other sports with specific needs, as typically teamsports).

Total time for complete recovery is longer for CT (4months) as for muscle injury (2 to 3 months). This dif-ference explains the higher peak torque obtains for CTcompared to MI. But the MI group (majority of soccer)needs, because of this sport requirement, progressiveand specific training before to return to competition.This specificity explains the longer time between returnto training and return to competition.

The number of eccentric session per week is a keyfactor, 3 times a week is a minimum as shown in aprevious study by Cotte [3] CMSLG.

Patients who had only 2 eccentric sessions per weekbecause of professional or personal commitments cameback later to sport activities.

Delay between injury and treatment doesn’t seem tointeract the quality of the healing but increase the totaltime of rehabilitation.

Literature point out that total rest and too early mo-bilisation is negative for correct healing [10]. Contrary,several authors reported the interest of early work ontraction (eccentric), aligned with muscle mechanicalaxis, to care CT and MI [2,8]: thanks to proteinic syn-thesis activation, reinforcement of conjunctive tissuesenvelope and muscle fiber alignement.

Isokinetic dynamometer is an accurate and precisetool for adjustment and management of protocols pa-rameters (speed, ROM, duration, torque- threshold andlimits) allowing individual and precise exercises. It isalso used to quantify objectively the force level requiredfor sport activity return [7].

These are the reasons why we used isokinetic dy-namometer for CT and MI care, with positive results.

This treatment failed on the case with enthesopathyof the calcanean tendon (no pain decrease no return tosport activity in spite of the increase of peak torque aftertreatment). This confirms the results of Alfredson [1]who reports positive results of eccentric work for 89%of CT and only 32% success with insertional pain.

Our results allow the establishment of specific proto-cols parameters for sub-acute muscular and tendinousinjuries of lower limb, using eccentric contraction withisokinetic dynamometer.Protocol must include

– the 1st day of treatment a muscle force assessmenton non involved side, on eccentric conditions atthe same speeds used during for the treatment: 5,10, and 15/sec.

– No pain rule respected.– Warm up before each session and cool down re-

covery on ergometric bicycle and stretching exer-cise after each session. Treatment must includeclassical manual physiotherapy.

– 12 to 20 isokinetic eccentric sessions are recom-mended at a frequency of 3 times per week. Treat-ment starts with isokinetic eccentric exercise forhealing goal: at 5/sec of speed, ROM limited at45% of max ROM, intensity at 35% of peak torqueof non involved side (this corresponds to−50 Nmor−0.66 Nm/kg, if the 2 sides are involved.

– Increase of speed is 5/sec to 10/sec after 5 isoki-netic eccentric sessions. This change of speedmust correspond at the time where peak torquereach 75% of none involved value.

– Increase of speed from 10/sec to 15/sec is whenpeak torque reach 90% of none involved side peaktorque, in the same conditions.Treatment must go on at 15ª/sec up to the time toobtain the same torque value as the non involvedside.Then the eccentric exercise must be continued withhigher speed (30/sec and 60/sec) in order to tar-get more physiological conditions.

Return to run can be suggested, with no pain rules,when peak torque reaches 83% of bilateral performancefor CT and 78% for MI.


Return to sport training and to competition are linkedto patient’s clinical feeling and total recovery of normaleccentric isokinetic values.

7. Conclusions

This 21 patients study on CT and MI pathologiesshows that musculo-tendinous injuries are still to muchneglected: medical examination practised too late (av-erage 2 months) and after inefficiency of a rest sportperiod and/or failure of classical medical care. Restonly is not efficient and classical care is often also notefficient.

Any delay in start of care increases the after-effectsand recurrent risks.

So we do remind the importance of diagnosis andearly care of musculo-tendinous injuries.

We succeed proven results, in case of late treatmentand case of failure of classical care for 4 patients:

– Healing has been completed for 20 patients (painfree, peak torque balance with non involved sideor normal at the end of treatment).

– Only one of the 21 patients did not have improve-ment in spite of normalization of his isokineticpeak torque values and his insertional calcaneanpain tendon needed surgery.

– 18 patients of the group return to their anteriorsport activities: 14 of them in the same conditionsas before injury (same performances, totally painfree) or even better (better feeling of muscular ca-pabilities); 4 patients return to sport with subjec-tive asymmetry feeling or apprehension.

– 1 patient did not return to anterior sport activitiesbecause of inter-current pathology.

– 1 patient faced to recurrent injury 2 months afterreturn to sport.

Return to sport practises is based on clinical exam-ination and values of isokinetic data collected duringtreatment [7]. The aim of this prospective study is tohelp the optimization of musculo-tendinous treatment.

Acknowledgements

We warmly thank the medical team of Centre deMedecine du Sport de Lyon Gerland (Lyon, France)

and the paramedical team of Centre de Kinesitherapiedu Sport de Lyon Gerland (Lyon, France) for help andcontribution to this study.

References

[1] H. Alfredson, M. Fahlstrom et al., Chronic Achilles tendonpain treated with eccentric calf-muscle training,Knee surgery,Sports Traumatologie, Arthroscopy 11 (2003), 327–333.

[2] D.B. Allbrook, Skeletal muscle regeneration,Muscle andNerve 234 (1981).

[3] T. Cotte, Isocinetisme et Reeducation, Journees surl’Isocinetisme en medecine du sport, organisees par la SocieteMEDIMEX, Lyon: 2003, (article non publie).

[4] J.L. Croisier and J.M. Crielaard, Hamstring muscle tear withrecurrent complaints: an isokinetic profile,Isokinetics andExercise Science 8 (2000), 175–180.

[5] J.L. Croisier, Exploration fondamentale et clinique del’exercice isocinetique excentrique, Memoire d’agregation deMedecine Physique – Kinesitherapie et Readaptation. Univer-site de Liege, 2002, p. 150.

[6] J.M. Ferret, C. Bruge and Mathieu R. Prevention, in: Mus-culaires recentes, J.M. Ferret, C. Bruge and Mathieu R. Leslesions, eds, Lipha, 1987, pp. 70–73.

[7] J.M. Ferret, Lesions musculaires, in:Medecine du Sport:Prevention, traitements, homeopathie et nutrition, J.M. Ferretand H. Koleckar, eds, Paris: Boiron, 2000, pp. 79–113.

[8] I. Fyfe and W.D. Stanish, The use of eccentric training andstretching in the treatment and prevention of tendon injuries,Clinics in Sports Medicine 11(3) (1992), 601–624.

[9] T. Hurme et al., Healing of skeletal muscle injury: an ul-trastructural and immunohistochemical study,Medicine andscience in sports and exercise 23(7) (1991), 801–810.

[10] M. Jarvinen and Lehto M. UK, The effects of early mobil-isation and immobilisation on the healing process followingmuscle injuries,Sports Medicine 15 (1993), 78–89.

[11] P. Kannus, Etiology and pathophysiology of chronic tendondisorders in sports,Scandinavian Journal of Medicine andScience in Sports 7 (1997), 78–85.

[12] U.M. Kujala et al., Hamstring injuries. Current trends in treat-ment and prevention,Sports Medicine 23 (1997), 397–404.

[13] C. Queiros da Silva, Apport du travail isocinetique excentriquedans le traitement des lesions musculo-tendineuses subaigues.Synthese bibliographique. Etude prospective portant sur 25 casde tendinopathie du tendon calcaneen et de lesions musculairesdes membres inferieurs. These en medecine. Universite Lyon1: 2004, p. 132.

[14] P. Rochcongar et al., Etudeepidemiologique du risque trau-matique des footballeurs franc¸ais de haut niveau,Science andSports 19 (2004), 63–68.

[15] W.D. Stanish et al., Eccentric exercice in chronic tendonitis,Clinical Orthopaedics and releated research 208 (1986), 65–68.

[16] T.W. Worrell and D.H. Perrin, Hamstring muscle injury: theinfluence of strength, flexibility, warm-up, and fatigue,Jour-nal of Orthopaedics and Sport Physical Therapy 16 (1992),12–18.


Review

Pertinent issues regarding the isokinetics ofthe ankle complex

Zeevi DvirDepartment of Physical Therapy, Sackler Faculty of Medicine, Tel Aviv University, Israel 69978

1. Introduction

Serious research concerning isokinetic testing of theankle goes back to the early 80’s with a series of stud-ies coming from Sweden [6,8]. This series has beenfollowed by extensive research that focused on differ-ent aspects of ankle isokinetics. Surprisingly even to-date, most of the technical problems are not resolved,a tribute to the complexity of this joint system, its opti-mal testing protocol and the incompatibility of variouscommercially available isokinetic dynamometers. Inthis short review I will try to dwell on a number of per-tinent points, specifically on the use of ankle musclesstrength ratios.

2. Protocol aspects

These concern the muscles under consideration, testposition, the range of motion and test velocities. Typi-cally, 4 muscle groups have been investigated: the plan-tarflexors (PF), dorsiflexors (DF) Invertors (IN) andEvertors (EV). Among these, the PF and DF are moreamenable to isokinetic evaluation whereas IN and EVpose a serious methodological problem.

As for the PF and DF, the design of the specific dy-namometer dictates often the test position. However,it seems that most professionals preferred the supineor prone lying position. The advantage of the ’hori-zontal’ testing is twofold: first it allows a measure ofself stabilization by virtue of the weight of the sub-ject. On the other hand care must be taken prevent

forward/backward sliding of the subject on the plinthwhich is often observed in eccentric testing of the DF.In addition this test position enables the placing of theknee in a comfortable degree of flexion while strap-ping the shank in order to provide added distal stabi-lization. Admittedly, the latter may not equally applyupon testing at the prone position. Second, it allowsthe examiner more freedom in aligning the bimalleo-lar axis with that of the lever-arm. This test positionhas been adopted in most of ankle isokinetics studies.An alternative test position for the PF which has beenoriginally proposed by the author was later examinedin a systematic manner [16]. In this position, the kneeis placed at 90 flexion while the foot is supported bya wedge (from below) which is located on an elevatedsurface (stool). With a fulcrum now at the region ofthe heads of metatarsal bones, concentric contractionof the PF lifts the shank while eccentric contractionis associated with lowering of the foot, much like theevents that take place in functional activities of the foot.The scores derived from this protocol have been shownto be highly reproducible. The RoM used for PF/DFtesting is variable, typically between 30 and 45. How-ever the velocities applied ranged 60–180/s which forthe high end and the corresponding RoM would indi-cate a largely non-isokineticconditions. Unfortunately,this issue was never properly addressed with respect tothese muscles.

The situation regarding IN/EV strength is somewhatmore complicated, resulting first and foremost fromthe unavoidable misalignment between the abovemen-tioned axes. Eversion and inversion are compound


46 Z. Dvir / Pertinent issues regarding the isokinetics of the ankle complex

movements involvingsimultaneously numerous instan-taneous axes that cannot be fitted with a singular axis.Moreover, it would be particularly difficult to adjust theposition of the lever-arm to accommodate substantiallythese axes. As a result the IN/EV ‘plane’ of move-ment is defined based on the performance capabilityof the dynamometric system as a whole (motor, seat,and attachments). Typically the subject is in the supineposition while the lever arm is facing the feet ratheralongside them. Maintenance of some flexion in theknee prevents contribution from hip rotators since witha fully extended such co-activation could take place.The action that is being measured may be described interms of rotation in a plane that is generally orthogonalto the axis of the tibia. The RoMs and velocities thatwere used in previous studies were similar to those ap-plied in PF/DF testing with the obvious implications.A different approach using short RoM testing has beenrecently applied to IN/EV testing. In a study of patientswho have undergone ORIF for bimalleolar fractures,the RoM was limited to 20 while the velocities cor-responded to nominal movement times of 2 and 0.5snamely 10 and 40/s, thus rendering alignment lessproblematic [5].

3. Strength ratios

In assessing the integrity of the musculoskeletal ap-paratus the best estimates may be derived when thecompromise in unilateral namely the involved sideis compared to the uninvolved side using identicalstrength testing protocol (Sapega 1990). This axiomis equally valid for ankle muscles. However when thecompromise is bilateral comparison to so-called ‘nor-mative’ values is largely irrelevant.

Another issue that in recent years saw a conspicu-ous development is that of within joint (side) strengthratios. These relate initially to the conventional ago-nist/antagonist strength or power ratios namely concen-tric (agonist) vs. concentric (antagonist) or eccentricvs. eccentric, respectively. However in view of the par-ticular importance attributable to the dynamic controlratio (DCR) defined as the concentric strength of theagonist vs. the eccentric strength of the antagonist withrespect to knee musculature [1,4], similar applicationhas already been attempted with respect to the ankleINs and EVs. Both the conventional ratios and theDCRs were applied for investigating what may proba-bly is the most pressing problem of the ankle complex:chronic instability.

4. Isokinetic correlates of chronic ankle instability

The issue of chronic ankle instability (CAI) whichrefers to repetitive incidences of lateral ankle insta-bility resulting in numerous ankle sprains [10] occu-pies, justifiably, a specific niche in the isokinetic lit-erature. CAI engulfs two well established concepts,that of mechanical and functional ankle insufficiency(MAI and FAI, respectively). Rather than two distinctentities MAI and FAI are probably complementary innature. Hertel [10] suggests that MAI, which relatesto structural changes following an initial ankle sprainthat predispose the ankle to CAI, consists of 4 ele-ments: pathologic laxity, arthro-kinematic restrictions,degenerative changes and synovial changes. On theother hand, FAI, which relates to a compromised neu-romuscular apparatus consists of impaired propriocep-tion, postural and neuromuscular control and strengthdeficits. It is the latter element that has been the fo-cus of a number of studies which have largely appliedisokinetic dynamometry. The reason for attaching spe-cific importance to the ankle muscle complex and itslevel of performance is the critical dependence of an-kle stability on dynamic muscular control. Althoughits ligamentous protection is far more robust comparedto other joints that suffer from inherent instability (e.g.the glenohumeral) the mechanical demands put on thisjoint system are immense. The combination of a highcenter of gravity and hence a relatively large gravita-tional lever arm, uneven surfaces and speed of motionas well as sudden high velocity impacts may result ina biomechanical combination that exceeds the physio-logical tolerance of the lateral aspect. This in turn leadsto various degrees of collapse of the passive defenseprovided by the lateral ligaments of the ankle. Togetherwith the lateral ankle ligaments: the anterior talofibu-lar, calcaneofibular, posterior talofibular, talocacanealand cervical, the peronei muscles act to counteract theexternal inversion-supination moment. As the peroneicontract they also elongate and given the fast nature ofthis contraction it is feasible that the force generatedwithin these muscles is near or at its maximum. Hencethe pivotal role eccentric testing has in the muscular as-sessment of the ankle in general and cases of instability,in particular.

4.1. Strength variations in CAI

There is a controversy regarding weakness of theperonei in CAI. In one of the first studies of its kindTropp [19] compared EV strength in subjects with FAI.

Z. Dvir / Pertinent issues regarding the isokinetics of the ankle complex 47

Significant reduction in peak moment was revealedand attributed to inadequate rehabilitation and atrophy.Using concentric tests, Lentell et al. [13], Ryan [18],Lentell et al. [14], Wilkerson et al. [20], McKnight andArmstrong [15] and Porter et al. [17] reported no selec-tive weakness in the EVs of the involved side in patientspresenting with CAI although tests covered substan-tially the full spectrum of test velocities (30–210/s). Invariance with the findings associated with EV strengthRyan [18] and Wilkerson et al. [20] reported a seem-ingly unexpected weakness in IN strength among CAIpatients. This weakness was explained by compressionof the deep peroneal nerve,an outcome of lateral sprain,and reduced ability to control the lateral displacementof the foot due to compromised lateral ligaments. Asa result, eccentric conditioning of the IN group wassuggested [20]. Eccentric testing in patients with CAIwas conducted by Bernier et al. [2] and Kaminski atal. [10]. In the previous study no differences in eitherthe INs or the EVs was noted. The latter study returnedsimilar findings with respect to EV strength.

On balance, these studies deny a specific weaknessof the EVs but point out to a possible compromise tothe INs. Though the absence of expected EV insuf-ficiency may reflect a genuine situation, it is possiblethat a number of factors conceal a real problem. First,as mentioned above, the possible misalignment of axes(biological vs. motor) as well as the test position forEV-IN and may not impose a pure external counter-moment in the same way it does, for instance, duringtesting of the PF/DF. This misalignment, which dueto their design and construction takes place in all dy-namometers, could put the force vectors of the EVsand INs in an off-plane position and consequently al-low recording of only a fraction of their true potential.In much the same way testing in the seated position isquite in variance with the functional demand and thestatic effect of gravity. In other words, the operatingmuscles may already be under a ‘bias level’ of tension.It is not at all clear whether the test protocols employedhave incorporated a sufficient isometric pre-activationbias. In addition to these lateral ankle sprain normallyoccurs within a very short duration, typically shorterthan what is needed for a reflex arch to activate themuscle. Thus, enhanced strength may not offer enoughprotection in such situations, anyway. On the otherhand, lower velocity threats may be averted by higherperformance muscles and a better tuned proprioceptiveapparatus. These indeed are the major objective of CAIrehabilitation.

4.2. Dynamic control ratios (DCR) applied to anklemuscles

Whereas same mode strength ratios were largely un-remarkable with respect to CAI, there was some hopethat DCR type ratios will allow a clear differentiationbetween involved and uninvolved sides. Specificallytargeted was the EVecc/INcon which relates to the effectthe peronei exert while attempting to control the turn-ing (inversion) moment produced by the invertors [3,11] upon performance of an angular (‘open chain’)isokinetic test. However, both studies failed to estab-lish a specific role for either the EVecc/INcon or theEVcon/INecc ratios.

A different light was shed on the ankle-related DCRsby a recent study which focused on isokinetic perfor-mance of the EVs and INs in patients 6 to 12 monthsfollowing surgery of malleoli fractures [5]. Inver-tors and EVs strength was bilaterally measured con-centrically (con) and eccentrically (ecc). Absolutestrength of the involved limb was significantly lowerthan values of the sound limb, for both muscle groups(p <0.001). However, strength reduction was signifi-cantly greater for the INs (27.6–37.5%) than the EVs(14.4–24.3%) (p =0.0005). The EV/IN conventionalratios of the sound limb (89–102%) were significantlylower than ratios obtained for the involved limb (111–141%,p <0.0001). Of particular note, the clinical sta-tus of the limb was found to have a significant effect onall DCRs: EVecc/INcon, INecc/EVcon, EVcon/INecc andINcon/EVecc: p <0.0001,p =0.0032,p =0.0001 andp =0.0015, respectively. Thus, a significant compro-mise in the DCR was apparent in this group of patientswhich still exhibited a conspicuous functional deficitas indicated by Ankle-Hindfoot Scale. Thus althoughDCR-type ratios may not have a parallel role in ankleneuromotor evaluation as they fulfill in the shoulderand knee, further research is certainly deserving.

References

[1] K. Bak and S.P. Magnusson, Shoulder strength and range ofmotion in symptomatic and pain-free elite swimmers,Am JSports Med 24 (1997), 454–459.

[2] J.N. Bernier, D.H. Perrin and A.M. Rijke, Effect of unilateralfunctional instability of the ankle on postural sway and inver-sion and eversion strength,J Athl Train 32 (1997), 226–232.

[3] B.D. Buckley, T.W. Kaminski, M.E. Powers, C. Ortiz and T.J.Hubbard, Using reciprocal muscle group ratios to examineisokinetic strength in the ankle: a new concept,J Athl Train36 (2001), S93.

[4] Z. Dvir, G. Eger, N. Halperin and A. Shklar, Thigh musclesactivity and ACL insufficiency,Clin Biomech 4 (1989), 87–91.

48 Z. Dvir / Pertinent issues regarding the isokinetics of the ankle complex

[5] Z. Dvir, Y. Steinfeld-Cohen, S. Shabat, C. Peretz and M.Nyska, Ankle rotators strength and functional indices in pa-tients following operative intervention for malleoli fractures.Submitted.

[6] A.R. Fugl-Meyer, Maximum isokinetic ankle plantar and dor-siflexion torque in trained subjects,Eur J Appl Physiol 47(1981), 393–404.

[7] A.R. Fugl-Meyer, L. Gustavsson and Y. Burstedt, Isokineticand static plantarflexion characteristics,Eur J Appl Physiol 45(1980), 221–234.

[8] A.R. Fugl-Meyer, K.H. Mild and J. Hornsten, Output of skele-tal muscle contraction: A study of isokinetic plantarflexion inathletes,Acta Physiol Scand 115 (1982), 193–199.

[9] J. Hertel, Functional anatomy, pathomechanics and patho-physiology of lateral ankle instability,J Athl Train 37 (2002),364–373.

[10] T.W. Kaminski, D.H. Perrin and B.M. Gansneder, Eversionstrength analysis of uninjured and functionally unstable an-kles,J Athl Train 34 (1999), 239–245.

[11] T.W. Kaminski, B.D. Buckley, M.E. Powers, T.J. Hubbard,B.M. Hatzel and C. Ortiz, Eversion and inversion strengthratios in subjects with unilateral ankle instability,Med SciSports Exer 33S (2001), 135.

[12] T.W. Kaminski and H.D. Hartsell, Factors contributing tochronic ankle instability: A strength Perspective,J Athl Train37 (2002), 394–405.

[13] G.L. Lentell, L. Katzman and M. Walters, The relationshipbetween muscle function and ankle stability,J Orthop SportsPhys Ther 11 (1990), 605–611.

[14] G.L. Lentell, B. Bas, D. Lopez, L. McGuire, M. Sarrels andP. Snyder, The contribution of proprioceptive deficits, musclefunction and anatomic laxity to functional instability of theankle,J Orthop Sports Phys Ther 21 (1995), 206–215.

[15] C.M. McKnight and C.W. Armstrong, The role of anklestrength in functional ankle instability,J Sport Rehabil 6(1997), 21–29.

[16] M. Moller, K. Lind, N. Styf and J. Karlsson, The test-retestreliability of concentric and eccentric muscle action duringplantar flexion of the ankle in a closed kinetic chain,IsokinExer Sci 8 (2000), 223–228.

[17] G.K. Porter, T.W. Kaminski, B. Hatzel, M.E. Powers and M.Horodyski, An examination of the stretch-shortening cycleof the dorsiflexors and evertors in uninjured and functionallyunstable ankles,J Athl Train 37 (2002), 494–500.

[18] L. Ryan, Mechanical stability, muscle strength and proprio-ception in the functional unstable ankle,Austral J Physiother40 (1994), 41–47.

[19] H. Tropp, Pronator muscle weakness in functional instabilityof the ankle joint,Int J Sports Med 7 (1986), 291–294.

[20] G.B. Wilkerson, J.J. Pinerola and R.W. Caturano, Invertor vs.evertor peak torque and power deficiencies associated withlateral ankle ligament injury,J Orthop Sports Phys Ther 26(1997), 78–86.

[21] M.S. Yeung, K.M. Chan, C.H. So and W.Y. Yuan, An epidemi-ological survey on ankle sprain,Br J Sports Med 28 (1994),112–116.

inﬂuence of the dynamometer and knee joint axis of ...€¦ · isokinetics and exercise science...

Documents