Evaluation of isokinetic and isometric strength measures for monitoring muscle function recovery following ACL reconstruction

EVALUATION OF ISOKINETIC AND ISOMETRIC STRENGTH

MEASURES FOR MONITORING MUSCLE FUNCTION

RECOVERY AFTER ANTERIOR CRUCIATE LIGAMENT

RECONSTRUCTION

OLIVERA M. KNEZEVIC,1,2 DRAGAN M. MIRKOV,1 MARKO KADIJA,3 DARKO MILOVANOVIC,3 AND

SLOBODAN JARIC4

1Faculty of Sport and Physical Education, University of Belgrade, Belgrade, Serbia; 2University of Belgrade, Institute forMedical Research, Belgrade, Serbia; 3Clinical Centre of Serbia, Institute for Orthopedic Surgery and Traumatology, Belgrade,Serbia; and 4Department of Kinesiology and Applied Physiology, University of Delaware, Newark, DE

ABSTRACT

Knezevic, OM, Mirkov, DM, Kadija, M, Milovanovic, D, and Jaric, S.

Evaluation of isokinetic and isometric strength measures for

monitoring muscle function recovery after anterior cruciate

ligament reconstruction. J Strength Cond Res 28(6): 1722–

1731, 2014—Although various strength tests and their out-

come measures have been proposed for anterior cruciate lig-

ament (ACL) reconstruction (ACLR), their measurement

properties still remain relatively underexplored. The aim of this

study was to investigate the longitudinal construct validity of

the standard isokinetic (IKT) and isometric test (IMT), and of

the IMT of alternating consecutive maximal contractions

(ACMC). In addition, the concurrent validity of ACMC was

assessed and compared with the validity of IMT. The strength

of quadriceps and hamstrings in 20 male athletes with an

anterior cruciate ligament (ACL) injury were assessed before

ACLR, 4 and 6 months after ACLR, by means of IMT, ACMC,

and IKT performed at 60 and 1808$s21. Significant between-

session differences in muscle strength variables were found in

the involved quadriceps (F . 6.5; p # 0.05), but not in the

uninvolved leg (F, 2.5; p. 0.05). Coefficients of variations in

the uninvolved leg (all below 13.5%) were lower than the

involved leg (11.7–22.1%). Intraclass correlation coefficients

were moderate-to-high for the uninvolved leg and low-to-high

for quadriceps of the involved leg. The concurrent validity of

ACMC with respect to the IKT (r = 0.57–0.92; p # 0.05) was

comparable with the validity of IMT (r = 0.52–0.87; p # 0.05).

We conclude that the explored longitudinal construct validity

of most of the evaluated variables could be sufficiently sensi-

tive to detect the effects of the applied rehabilitation proce-

dures. In addition, the obtained sensitivity and concurrent

validity and the potential advantages of ACMC over IMT, all

suggest that ACMC could be a particularly promising method

for routine testing of neuromuscular function after ACLR.

KEY WORDS anterior cruciate ligament injury, validity, peak

torque, limb symmetry index, hamstrings-to-quadriceps ratio

INTRODUCTION

The most predominantly applied methods for theassessment and monitoring of the quadriceps andhamstrings strength after anterior cruciate liga-ment reconstruction (ACLR) are isometric

(IMT) and, particularly, isokinetic (IKT) strength tests, basedon sustained contractions of the tested muscle (17,20,32).Although the application of IKT could be considered asa standard method for the assessment of muscle functionafter ACLR, the use of IMT has also been valuable. Thathas been the case particularly when only single transducerswith simple custom-made devices are available instead of therelatively expensive isokinetic dynamometers (18).

A strength test based on alternating consecutive maximalcontractions (ACMC) performed under isometric conditionshas been recently proposed to overcome some of theshortcomings of the standard IMT (3–5,18). Probably, themost important one originates from the differences in theneural activation pattern between the rapid and sustainedmaximum contractions (9,24,27). Namely, IKT and IMTmay not capture the neural activation pattern typical forrapid exertion of force that could be critical for functionaltasks that provide limited time for exerting relatively highmuscle force (such as in explosive and rapid cyclic move-ments), or require transient consecutive actions of antago-nistic muscle groups (walking and running). As a result, theassessment of the abilities to exert a sustained maximumforce (as assessed by IKT and IMT) and to exert it rapidly

Address correspondence to Dr. Dragan M. Mirkov, [email protected].

28(6)/1722–1731

Journal of Strength and Conditioning Research� 2014 National Strength and Conditioning Association

1722 Journal of Strength and Conditioning Researchthe TM

Copyright © National Strength and Conditioning Association Unauthorized reproduction of this article is prohibited.

could require separate methods of evaluation (2,31). In addi-tion, the exertion of a sustained contraction that providesmaximum force could be painful or inappropriate for someindividuals such as injured/recovering persons (22,35).

Regardless of the applied test, along with the absolutestrength measures (usually peak torques [PT]), strength imbal-ance ratios have been used to monitor rehabilitation progressand identify possible risk factors for ACL reinjuries (8,13,26).In particular, limb symmetry index (LSI) reveals the strengthimbalance between contralateral legs, whereas the Hamstrings-to-Quadriceps ratio (HQ ratio) describes the strength imbalanceof antagonistic muscle groups. Among others, these imbalanceshave been taken into account in the decision-making processesregarding whether the athlete is ready to return to routine ath-letic training and competition (7,11,15).

Both PT and the indices of strength imbalance need to beboth reliable and valid to be used either for monitoringrehabilitation or as a screening test of muscle function.Previous reports have suggested “high” to “very high” relativeand absolute reliability of both isokinetic and isometric PT ofquadriceps and hamstrings (1,13,21,32,35). However, onlyfew studies have examined the reliability of HQ ratio andparticularly of LSI, reporting low-to-moderate values (13,32).Although both PTand strength imbalance ratio measures havebeen widely used to monitor rehabilitation of muscle function,there is an apparent lack of information regarding their longi-tudinal construct validity. A construct validation process isusually based on construct that the involved leg’s muscle func-tion changes after the surgery, and later on during the reha-bilitation process, while the muscle function of uninvolved legremains unchanged (28). Further information regarding thistype of validity of strength measures is apparently needed toadequately track clinically important changes or to accuratelyassess the recovery progress, and to appropriately use the samemeasures in future research on clinical populations.

To address the aforementioned issues, this study wasdesigned with the main aim to explore the longitudinalconstruct validity of the IKT, IMT, and ACMC when usedto monitor the muscle function recovery after the ACLR. Thesecondary aim was to explore the concurrent validity of theACMC with respect to IKT and compare it with the validityof IMT. Regarding the outcome variables, in addition to thedirectly recorded PTof the knee muscles, the validity of boththe HQ ratio and LSI was also evaluated. The results wereexpected to provide important data regarding the evaluatedstrength tests, and to motivate further development of ACMCeither as an alternative or complementary strength test toIMT when used to assess the muscle function in ACLRindividuals or, possibly, other clinical populations.

METHODS

Experimental Approach to the Problem

An experimental longitudinal study was designed to testquadriceps and hamstrings through the standard IKT, IMT,and isometric ACMC of 2 antagonistic muscles. Because the

ACMC has previously been evaluated on healthy andphysically active participants (3–5), there is an apparent needto evaluate the properties of the same test when applied onclinical populations, such as on athletes with an ACL injury.Although various strength tests have been proposed as out-come measures after ACLR, limited reports of their measure-ment properties exist. Specifically, both PT and the indices ofstrength imbalance need to be both reliable and valid to beused either for monitoring rehabilitation or as a screening testof muscle function. To assess the longitudinal validity of var-iables obtained from the applied tests, comparisons betweenthe consecutive sessions were made. To assess the concurrentvalidity of ACMC, its variables were related to IKTand, after-wards, compared with the validity of IMT.

Subjects

The sample size estimate was based on the concurrentvalidity of ACMC observed in previous studies (4,5).According to standard guidelines (6) with power of 0.8and an alpha level of 0.05 [calculated by G*Power 3.1 freesoftware; (10)], the sample size was between 8 and 13 for theconcurrent validity (4). Conservatively, 23 male athletes withan ACL injury were recruited in the study, but 3 of themwere later lost to follow up (Figure 1). A post hoc analysis ofsensitivity for the used sample size with the power of 0.8 andan alpha of 0.05 revealed the effect size of 0.5. The subjectswere soccer (12), handball (5), and judo (3) competitors (age,24.2 6 5.1 years; body mass, 84.0 6 11.1 kg; height, 180.3 64.0 cm; body mass index, 25.1 6 3.5). They were recruitedthrough the Clinic for Orthopedic Surgery and Traumatol-ogy, between September 2010 and March 2012. The inclu-sion criteria were first ACL injury; no other knee ligamentsinjured; no history of concurrent fractures, osteoarthritis, orhereditary and neuromuscular diseases, participation in com-petitive sports at national level or higher. In all subjects, theACL reconstruction procedure was performed by a boardcertified orthopaedic surgeon, using the bone-patellar-bonetendon autograft. After the surgery, the subjects were allo-cated to a standard postoperative rehabilitation program forathletes. All subjects received a complete explanation regard-ing the purpose, procedures, and possible risks of the study.The institutional review board approved the project andappropriately informed consents have been gained from sub-jects, and their rights were protected.

Procedures

All measurements were performed within a universityresearch laboratory, using a Kin-Com AP125 isokineticdynamometer (KinCom, Kinetic Communicator; ChattecxCorp., Chattanooga, TN, USA). The design of the study ispresented in the Figure 1.

Measurements were taken through 3 separate sessions:preoperatively (i.e., within 7 days pre-ACLR—session 1),4 months (session 2), and 6 months post-ACLR (session3). Each session consisted of 2 experimental days, separatedby 48 hours of rest. Isokinetic tests were conducted in one

Journal of Strength and Conditioning Researchthe TM

| www.nsca.com

VOLUME 28 | NUMBER 6 | JUNE 2014 | 1723


experimental day, whereas the IMT and ACMC were con-ducted in the other experimental day. The order of theexperimental days was randomized.

Before muscle strength testing, each subject was givena 5-minute warm-up period on a stationary bicycle,followed by passive stretching exercises (15 seconds permuscle group) focused upon the quads, hamstrings, hipadductors, and calf muscles. Thereafter, the subject wasfixed to the testing apparatus with the tight straps usedto fix his pelvis, thigh, and malleoli. The axis of rotation ofthe dynamometer was aligned with the lateral femoralepicondyle.

The same test leader supervised all tests. A detailedexplanation and qualified demonstration was providedbefore each muscle strength test, and standardized verbalencouragement was consistently used. Real-time visualfeedback regarding the current force was shown at a com-puter screen positioned in front of the subject.

Isokinetic Test. Torque measurement was performed first ata low speed of 608 1.05 rad$s21 (IKT60) and, after a 2-min-ute rest, at 1808 3.14 rad$s21 (IKT180). Each participantexerted 5 cycles of maximal voluntary repetitions of alternat-ing knee extension and flexion. The range of motion during

the knee extension/flexion wasset to 808 (i.e., from 10 to 908 offlexion).

Standard isometric test wasconducted on the knee exten-sors and flexors separately. Thesubjects were instructed “to rap-idly exert the maximum force”against the dynamometer leverattached to the lower leg and toretain it for 3–4 seconds (35).Quadriceps and hamstrings PTwere measured at the kneeangle of 458 (18).

Alternating consecutive maxi-mal contractions were performedat the knee angle of 458 (18). Theparticipants were instructed “toconsecutively exert the alternat-ing maximum contractions of

Figure 1. Study design.

Figure 2. Representative force profiles obtained from the applied strength tests. The middle 3 cycles of isokinetictests (IKT) and alternating consecutive maximal contractions (ACMC) force profiles, and the peak value fromisometric tests are presented. IMT = standard isometric test.

Evaluation of Strength Measures



the knee extensors and flexors as strong and as quickly aspossible” where the ACMC frequency should be consideredas self-selected. Specifically, the participants were asked to exertself-paced maximum contractions consecutively in 2 oppositedirections. In addition to allowing for testing 2 antagonisticmuscles within a single trial, the advantage of the ACMC testcould be its close correspondence to a number of everyday and,in particular, high performance movements. Namely, althoughwithout allowing for the use of the stretch-shortening cyclebecause of the static conditions, ACMC is inevitably based onthe consecutive maximum activation of the antagonistic musclegroups. The trial duration covered 5 full periods of ACMC force.

Both in IKT60 and IKT180, and in the following tests(IMTand ACMC), one practice trial preceded 2 experimen-tal trials with 1 minute of rest between them. The uninvolvedleg was always tested first. None of the participants reportedpain in either of the legs during the testing.

Data Processing

A custom-made Lab View application was used for dataacquisition and processing of the muscle strength tests and

for the online visual feedback on a computer screen. Theforce-time curves were recorded at a rate of 500$s21 andlow-pass filtered (10 Hz) using a fourth-order (zero-phaselag) Butterworth filter (18). The peak forces were multi-plied by the individual lever arm length to calculate thePT. The peak values observed from the middle 3 cycles ofthe IKT and ACMC force profiles provided averaged PTfor quadriceps and hamstrings, respectively. Peak torquesof the quadriceps (PTQ) and hamstring (PTH) musclesobserved from IKT, IMT, and ACMC served for the cal-culation of HQ ratios as PTH/PTQ. The LSI between PTof the uninvolved and involved leg was separately calcu-lated from PTQ and PTH:

LSI ¼ PTuninvolved2PTinvolved

PTuninvolved3100;

where the score represents PT of the involved leg,expressed as a percentage of PT of the uninvolvedleg (19).

TABLE 1. Indices of longitudinal construct validity of peak torques obtained from isokinetic and isometric strengthtests.*

Session1 Session 2 Session 3 F

Session22session 1

(%)

Session32session 2

(%)CV% ICC (95% CI)

Uninvolved legQuadricepsIKT60 170 (36) 176 (38) 179 (40) 2.3 3.5 1.7 7.4 0.91 (0.83–0.96)IKT180 125 (24) 126 (26) 129 (32) 0.3 0.8 2.4 8.7 0.87 (0.77–0.94)IMT 212 (53) 213 (55) 225 (50) 2.1 0.5 5.6 12.2 0.80 (0.65–0.89)ACMC 197 (47) 202 (44) 211 (45) 1.7 2.5 4.5 9.0 0.83 (0.70–0.91)

HamstringsIKT60 104 (21) 107 (21) 111 (21) 1.1 2.9 3.7 9.6 0.80 (0.65–0.90)IKT180 89 (22) 85 (17) 84 (14) 4.0† 24.5 21.2 10.9 0.77 (0.59–0.88)IMT 107 (25) 105 (26) 110 (22) 0.8 21.9 4.8 13.5 0.74 (0.54–0.86)ACMC 104 (21) 107 (21) 111 (21) 3.7† 2.9 3.7 12.6 0.81 (0.67–0.90)

Involved legQuadricepsIKT60 130 (38) 101 (35)z 122 (37)§ 8.1k 222.3 20.8 18.1 0.80 (0.64–0.89)IKT180 102 (23) 84 (22)z 99 (31)§ 6.6k 217.6 17.9 14.0 0.80 (0.64–0.89)IMT 180 (41) 142 (48)§ 165 (40)§ 8.9k 221.1 16.2 20.9 0.60 (0.36–0.77)ACMC 166 (44) 125 (45)z 150 (40)§ 6.6k 224.7 20 22.1 0.58 (0.33–0.76)

HamstringsIKT60 96 (32) 92 (24) 99 (24) 1.3 24.2 7.6 17.2 0.74 (0.55–0.86)IKT180 79 (19) 76 (17) 79 (16) 1.7 23.8 3.9 11.7 0.77 (0.60–0.88)IMT 104 (21) 107 (21) 111 (21) 2.4 2.9 3.7 16.8 0.70 (0.50–0.84)ACMC 78 (19) 77 (26) 86 (20)z 3.9† 21.3 11.7 13.2 0.83 (0.69–0.91)

*F = repeated measures analysis if variance; CV = coefficient of variation; ICC = intraclass correlation coefficient; CI = confidenceinterval; IKT60 = isokinetic tests performed at 1.05 rad$s21; IKT180 = isokinetic tests performed at 3.14 rad$s21; IMT = isometricstrength test; ACMC = alternating consecutive maximum contractions.

†Significant effect of session (p # 0.05).zSignificantly different from the previous session (p # 0.05).§Significantly different from the previous session (p , 0.01).kSignificant effect of session (p , 0.01).


| www.nsca.com



Statistical Analyses

Descriptive statistics were initially calculated for the selectedvariables; data were reported as mean (SD). To assess thelongitudinal construct validity of PT, HQ ratio, and LSI,mean values of 3 sessions were compared using repeatedmeasures analysis of variance (ANOVA). When a significanteffect was found, Bonferroni post hoc test was applied todetect the systematic bias among the sessions. The level ofstatistical significance was set to a = 0.05. For the estimationof accuracy of the individual scores obtained from consecu-tive sessions, a coefficient of variation was used (12), whichhas been typically considered good if below 15% (33). Theintraclass correlation coefficients (ICC) were calculated asa measure of the degree to which individuals maintainedtheir position in a sample with repeated measures (12). Intra-class correlation coefficients were considered as being high($0.8), moderate (0.6, ICC, 0.8), or low (,0.6). Pearson’scorrelation coefficients (r) were used to assess the relation-ship between the applied tests. Concurrent validity of theACMC and IMT (predictor variables) with respect to IKT(common dependent variable) was assessed by the Mengtest for correlated correlation coefficients (see Meng at al.(23) for details). Data were analyzed using SPSS 20.0 soft-ware (SPSS Inc, Chicago, IL, USA) and Office Excel 2010(Microsoft Corporation, Redmond, WA, USA).

RESULTS

Typical force profiles obtained from the applied musclestrength tests are shown in Figure 2. As expected, the maxima

of the forces obtained from the concentric contractions per-formed at the lower (IKT60) and particularly the higher angu-lar velocity (IKT180) are lower than the maxima of bothisometric tests (i.e., IMT and ACMC).

Longitudinal Construct Validity

The indices of longitudinal construct validity of PTobtainedfrom preoperative and postoperative sessions are depicted inTable 1. Regarding the quadriceps and hamstrings PT of theuninvolved leg, ANOVA revealed no differences among thesessions in any of the applied tests. However, a significanteffect of session was detected in all strength tests regardingthe quadriceps of the involved leg. The values observed insession 2 were lower than those observed in sessions 1 and 3.ACMC seems to reveal particularly high relative differencesamong the consecutive sessions. The effect of session on thehamstrings PT was found only in ACMC because of lowervalues recorded in session 2 than in session 3. Coefficient ofvariation was lower in the uninvolved (all below 13.5%) ascompared with the involved leg (11.7–22.1%). Intraclass cor-relation coefficient was generally moderate-to-high for bothmuscles of the uninvolved leg, and between low (in IMTandACMC) and high (in IKT) for the quadriceps of involved leg.

The indices of longitudinal construct validity of the HQratios and LSI are reported in Tables 2 and 3, respectively. Ingeneral, both variables only revealed significant differencesamong the consecutive sessions for the quadriceps of theinvolved leg. Specifically regarding the HQ ratios, CV wassomewhat lower in the uninvolved (10.4–15.5%) than in the

TABLE 2. Indices of longitudinal construct validity of hamstrings-to-quadriceps ratio obtained from isokinetic andisometric strength tests.*

Session 1 Session 2 Session 3 F

Session22session

1 (%)

Session32session

2 (%)CV% ICC (95% CI)

UninvolvedlegIKT60 0.62 (0.09) 0.62 (0.09) 0.64 (0.13) 0.16 0.0 3.2 15.5 0.42 (0.13–0.65)IKT180 0.71 (0.09) 0.69 (0.11) 0.65 (0.08) 0.80 22.8 25.8 11.1 0.40 (0.12–0.64)IMT 0.51 (0.11) 0.50 (0.07) 0.50 (0.07) 0.78 22.0 0.0 10.4 0.69 (0.49–0.83)ACMC 0.46 (0.11) 0.46 (0.08) 0.48 (0.07) 0.65 0.0 4.3 11.1 0.73 (0.55–0.86)

InvolvedlegIKT60 0.73 (0.11) 0.98 (0.32) 0.86 (0.27)† 5.52z 34.2 212.2 17.7 0.62 (0.38–0.79)IKT180 0.78 (0.10) 0.93 (0.19) 0.85 (0.23) 3.43z 19.2 28.6 14.1 0.57 (0.31–0.76)IMT 0.52 (0.11) 0.62 (0.21) 0.59 (0.16) 3.78z 19.2 24.8 22.1 0.56 (0.26–0.72)ACMC 0.49 (0.15) 0.63 (0.21)† 0.53 (0.13) 5.74§ 28.6 215.9 25.2 0.44 (0.16–0.67)


†Significantly different from the previous session (p # 0.05).zSignificant effect of session (p # 0.05).§Significant effect of session (p , 0.01).




involved leg (14.1–25.2%). The ICC was generally low, with

the exceptions of the IMTand ACMC data obtained from the

uninvolved leg, where it was moderate. Regarding the LSI,

ANOVA revealed significant effect of session for all quadriceps

data, but not for hamstrings. The CV was generally high,

whereas the ICC was, on average, low-to-moderate both forquadriceps and hamstrings LSI.

Concurrent Validity

The concurrent validity of both the IMT and ACMC wereassessed with respect to IKT. In general, the correlations

TABLE 4. Concurrent validity of peak torques observed through the Pearson’s correlation coefficients (r) of isometrictests with the isokinetic tests.*

Session 1 Session 2 Session 3

IKT60†z IKT180 IKT60 IKT180 IKT60 IKT180

Uninvolved legQuadricepsIMT 0.69 0.63 0.87 0.78 0.74 0.85ACMC 0.66 0.70 0.84 0.78 0.84 0.92

HamstringsIMT 0.67 0.76 0.81 0.80 0.73 0.79ACMC 0.69 0.64 0.77 0.83 0.71 0.81

Involved legQuadricepsIMT 0.66 0.71 0.83 0.76 0.80 0.71ACMC 0.62 0.65 0.87 0.78 0.88 0.77

HamstringsIMT 0.58 0.54 0.86 0.78 0.72 0.80ACMC 0.63 0.57 0.74 0.69 0.77 0.83

*IKT60 = isokinetic tests performed at 1.05 rad$s21; IKT180 = isokinetic tests performed at 3.14 rad$s21; IMT = isometric strengthtest; ACMC = alternating consecutive maximum contractions.

†p # 0.05 for all correlation coefficients .0.41.zp , 0.01 for all correlation coefficients .0.60.

TABLE 3. Indices of longitudinal construct validity of limb symmetry index obtained from isokinetic and isometricstrength tests.*

Session1

Session2

Session3 F

Session22session 1 (%)

Session32session 2 (%)

CV% ICC (95% CI)

QuadricepsIKT60 76 (20) 57 (13)† 77 (17)z 6.17§ 224.7 34.3 15 0.62 (0.38–0.79)IKT180 81 (14) 67 (13)† 77 (17)z 7.34k 216.8 14.7 19 0.67 (0.45–0.82)IMT 85 (12) 66 (13)z 74 (13)z 15.50k 221.8 11.9 12.1 0.68 (0.46–0.83)ACMC 84 (19) 62 (18)† 70 (15) 8.35k 225.8 12.5 19.8 0.57 (0.31–0.76)

HamstringsIKT60 86 (22) 85 (14) 89 (14) 0.57 20.7 4.6 17.5 0.45 (0.16–0.68)IKT180 89 (17) 89 (14) 94 (11) 1.56 0.3 5 9.9 0.66 (0.44–0.82)IMT 85 (16) 78 (15) 85 (14) 1.15 29.2 9.8 17.6 0.31 (0.00–0.59)ACMC 89 (17) 82 (16) 88 (15) 1.15 27.2 6.4 13.5 0.61 (0.37–0.79)


†Significantly different from the previous session (p # 0.05).zSignificantly different from the previous session (p , 0.01).§Significant effect of session (p # 0.05).kSignificant effect of session (p , 0.01).


| www.nsca.com



between the corresponding PT obtained both between theIMT and IKT and between the ACMC and IKTwere moderate-to-high (Table 4). When the same correla-tion coefficients observed separately for the IMTand ACMCwere compared by means of the Meng test, the differenceseemed to be nonsignificant (Z , 1.57; p . 0.05).

Regarding the concurrent validity of the HQ ratios, thecorrelation coefficients observed both between the IMT andIKT, and between the ACMC and IKT seemed to be, onaverage, low for the uninvolved leg and moderate to mainlysignificant for the involved leg (Table 5). The Meng testrevealed no significant differences between the correspond-ing correlation coefficients observed from the IMT andACMC (Z , 1.58; p . 0.05), with the only exception ofthe HQ ratio of the involved leg recorded in session 1(Z = 2.89; p # 0.05). Regarding the concurrent validity ofquadriceps’ and hamstrings’ LSI, both the IMT and ACMCrevealed moderate and significant correlations with IKT. TheMeng test once again revealed no significant differencesbetween the corresponding correlation coefficients exceptfor the hamstrings LSI recorded in session 1.

DISCUSSION

The main aims of this study were to explore longitudinalconstruct validity of strength measures obtained from theroutinely used IKT, standard IMT test, and the recentlyintroduced test of isometric ACMC, recorded before surgeryand during post-ACLR rehabilitation. The obtained results

indicate that both isometric tests could possess a similar levelof the longitudinal construct validity as the routinelyapplied IKT. In addition, the concurrent validity of ACMCwith respect to IKT could be similar to the same concurrentvalidity of IMT.

The recorded changes in muscle function associated withthe applied treatments generally seemed to be in line withour construct for change. The significant effect of sessionand a higher CV and a lower ICC of the strength measuresobtained from the involved leg support the presumedconstruct. It therefore provides the evidence of longitudinalconstruct validity of the applied methods for strengthassessment. As expected, the PT data obtained from IKTand IMT revealed significant time-related changes in theinvolved quadriceps, but neither in the involved hamstringsnor in the muscles of the uninvolved leg. Similar changeswere also detected by the ACMC test, both in PT of theinvolved quadriceps and the corresponding LSI, but also inPT of the involved hamstrings and the HQ ratio, whichcould suggest that the ACMC may be as sensitive as eitherthe IMT or IKT. Although the between sessions differencesin HQ ratio were significant only at the lower isokineticvelocity of IKT and in ACMC, the overall percent changewas still large enough to indicate that HQ ratio couldcapture the expected changes over time in the strengthimbalance of the antagonistic muscles.

Percent change between the sessions, CV, and ICC of therecorded strength measures were also reported as a part of

TABLE 5. Concurrent validity of hamstrings-to-quadriceps ratio and limb symmetry index observed through thePearson’s correlation coefficients (r) of 2 isometric tests with the isokinetic test.*

Session 1 Session 2 Session 3

IKT60†z IKT180 IKT60 IKT180 IKT60 IKT180

HQ ratioUninvolved legIMT 0.16 0.44 0.37 0.36 0.31 0.18ACMC 0.24 0.76 0.26 0.32 0.28 0.34

Involved legIMT 20.06 0.45 0.73 0.68 0.72 0.73ACMC 0.34§ 0.37 0.70 0.62 0.72 0.79

LSIQuadricepsIMT 0.78 0.74 0.81 0.73 0.66 0.64ACMC 0.66 0.67 0.84 0.69 0.75 0.57

HamstringsIMT 20.05 0.32 0.76 0.66 0.39 0.64ACMC 0.66§ 0.41 0.56 0.61 0.43 0.74

*IKT60 = isokinetic tests performed at 1.05 rad$s21; IKT180 = isokinetic tests performed at 3.14 rad$s21; HQ ratio = hamstrings-to-quadriceps ratio; IMT = isometric strength test; ACMC = alternating consecutive maximum contractions; LSI = limb symmetry index.

†p # 0.05 for all correlation coefficients .0.41.zp , 0.01 for all correlation coefficients .0.60.§p # 0.05 for Meng test indicating a significant difference between IMT and ACMC.




the longitudinal construct validity of the applied tests. Ingeneral, the comparison of scores over the 3 sessionsindicated that substantial changes took place on the involvedlimb from the first to second and from the second to thirdtest sessions. Similar changes were not observed in thecontralateral limb, suggesting that the functional status of theuninvolved leg remained stable over the 6-month period.Note that some authors suggest that neuromuscular dys-function and quadriceps strength loss after ACL injury affectthe uninvolved side as well and, therefore, the uninvolved legshould be used as a control variable with caution whenevaluating strength deficits after ACL injury (8,15). How-ever, our data show virtually no changes in the uninvolvedleg over the entire tested time interval. Similarly, the indicesof longitudinal construct validity suggest that PT and HQratios of the uninvolved leg were relatively stable across thesessions, unlike the quadriceps’ PT and corresponding HQratios obtained from the involved leg. However, althoughthe corresponding CV and ICC indicate both the within-individual and between-session variability of strength imbal-ance ratios to be relatively high in the involved leg, the samemeasures could still be suitable for detecting prominentchanges between the sessions that could be affected by theapplied rehabilitation procedures.

We were not able to find published estimates of thelongitudinal construct validity of isokinetic and isometricstrength measures obtained from ACLR participants. How-ever, note that our results obtained from IMT and IKT, andfrom the ACMC test, indicate a significant reduction in thequadriceps strength after the surgery are in line with theprevious studies (14,19,25,26).

Although the isokinetic testing has been the mostfrequently applied method for the assessment and monitor-ing the quadriceps and hamstrings strength after ACLR,there is an apparent lack of data regarding the evaluation ofisometric strength measures. The outcomes of the IMTsapplied in this study (i.e., IMT and ACMC) were generallysimilar to those previously reported for isokinetic measuresobtained from both healthy and ACL participants. Inparticular, previous studies suggest that the absolute meas-ures (such as PT) observed from various isokinetic testscould be more reliable and stable than the derived measures,such as HQ ratio or LSI (13,32,33). However, these studiesinvestigated strength measures obtained from healthy andphysically active individuals within a single testing session.Ross et al. (30) investigated the reliability of PT and LSI ofquadriceps in ACLR participants and reported a high reli-ability of the PT and LSI, but the IKTs were conductedbetween 12 and 72 months post-ACLR. The reliability wasnot directly calculated in one of the first studies that assessedHQ ratios in ACLR population (15). However, the authorfound that the HQ ratios were highly variable across theparticipants (range, 31–80%), even in the uninvolved leg.Finally, note that both the HQ ratios and the LSI are derivedfrom several individual PT’s. Because each PT has its own

limited reliability and its variance is nonnegligible, a relativelyhigh variability and low reliability of HQ ratios and LSI(as compared with individual PT) should not be viewed asa surprise.

Our findings generally suggest that the obtained indices oflongitudinal construct validity of IMT and ACMC werecomparable with the routinely used IKT. This could be ofimportance in clinical settings because isokinetic dynamom-eters could be replaced with relatively inexpensive singleforce transducer and custom built devices (4,5). Moreover,although the properties of ACMC proved to be comparablewith the IMT, ACMC could still retain some importantmethodological advantages, such as a brief and simple pro-cedure for testing 2 antagonistic muscles, while exposing themuscle and joint tissues to transient forces (5,34).

The validity of inferences based on isokinetic and iso-metric measures has been supported by a number of studiesrendering IKT use highly valuable whenever muscle strengthneeds to be assessed (7,15,29). Therefore, the relationshipbetween the strength indices obtained from ACMC andthose obtained from routinely used IKTcould be interpretedas an index of concurrent validity of ACMC. The same indi-ces of concurrent validity obtained from ACMC were alsocompared with those observed from standard IMT.

In general, the results revealed a moderate-to-high con-current validity of PTof both ACMC and IMT, whereas thesame validity of HQ ratios and LSI were somewhat lower.The obtained relationships between the isometric andisokinetic PTwere in line with the results of previous studies(16,20). A similar level of the concurrent validity of ACMCwith respect to IMT was recently recorded in physicallyactive participants (4). Regarding LSI, our results are similarto Reinking at al. (29), who found a moderate relationshipamong the isokinetic and isometric LSI, which wasexplained by the variability in percentage deficits betweenthe strength tests. However, there is a lack of published datadescribing the relationship between the isometric and isoki-netic HQ ratios, which seems to be low-to-moderate in thisstudy. Here, one could speculate that the isometric outcomemeasures (such as in IMT and ACMC) are knee angle–dependant, whereas IKT tests muscles within a considerablerange of motion. However, we believe that similar to theirconstruct validity and variability, the concurrent validity ofHQ ratios and LSI could be low on average (as comparedwith individual PT) because of the method in which thesevariables were calculated. Namely, the summed-up varianceof several PTused for calculation of individual HQ ratios andLSI reduces the relationships between their values obtainedfrom different tests. Nevertheless, in addition to the longitu-dinal construct validity, the observed concurrent validity alsosuggests that not only the standard IMT, but also theACMC could be considered for routine strength testing ofACL and, possibly, other clinical populations.

Among the limitations of our study could be that ourpatient group was composed of a relatively small sample size


| www.nsca.com



(n = 20) and of exclusively male professional athletes. Namely,different results could have been found in other populations,such as female athletes, sedentary individuals, or individualsparticipating in different rehabilitation programs. Note alsothat only the participants with the bone-patellar-bone tendongraft were evaluated, but not the hamstrings tendon graftreplacement. Future studies should address the aforemen-tioned limitations, explore the external validity of the ACMCtest regarding some ACL-specific functional performancetests, and evaluate the same tests on other clinical populations.

PRACTICAL APPLICATIONS

In general, the data suggest that IMTs could replace standardIKTs of leg strength that require isokinetic devices. Moreover,taking into account some shortcomings of the standard tests ofmuscle strength based on sustained maximal actions, a part ofthe observed findings suggest that the tests based onalternating consecutive maximal contractions could be con-sidered as either a complementary or alternative tests to oftenemployed isometric tests. Of particular importance could bethat the ACMC test revealed not only the longitudinalconstruct validity comparable with IKT and IMT, but alsoa moderate-to-high concurrent validity with respect to theroutinely applied IKT that could be comparable with thesame concurrent validity of IMT. Taking also into accountthe potential advantages of ACMC over IMT, such as a briefprocedure for testing 2 antagonist muscles together, thefindings suggest that ACMC could be a particularly promisingmethod for routine testing of the neuromuscular function afterthe ACL reconstruction.

ACKNOWLEDGMENTS

Supported in part by NIH grant (AR06065) and grants fromthe Serbian Ministry of Education, Science and Technolog-ical Development (175037 and 175012). The authors discloseprofessional relationships with companies or manufacturerswho will benefit from the results of this study. The results ofthis study do not constitute endorsements of the product bythe authors or the National Strength and ConditioningAssociation.

REFERENCES

1. Abernethy, PJ and Jurimae, J. Cross-sectional and longitudinal usesof isoinertial, isometric, and isokinetic dynamometry. Med Sci SportsExerc 28: 1180–1187, 1996.

2. Bemben, MG, Clasey, JL, and Massey, BH. The effect of the rate ofmuscle contraction on the force-time curve parameters of male andfemale subjects. Res Q Exerc Sport 61: 96–99, 1990.

3. Bozic, PR, Celik, O, Uygur, M, Knight, CA, and Jaric, S. Evaluationof novel tests of neuromuscular function based on brief muscleactions. J Strength Cond Res 27: 1568–1578, 2013.

4. Bozic, P, Pazin, N, Berjan, B, and Jaric, S. Evaluation of alternatingconsecutive maximum contractions as an alternative test ofneuromuscular function. Eur J Appl Physiol 112: 1445–1456, 2012.

5. Bozic, P, Suzovic, D, Nedeljkovic, A, and Jaric, S. Alternatingconsecutive maximum contractions as a test of muscle function.J Strength Cond Res 25: 1605–1615, 2011.

6. Cohen, J. Statistical Power Analysis for the Behavioral Sciences. 567.L. Erlbaum Associates, 1988. pp. 520–527.

7. Dvir, Z. Isokinetics: Muscle Testing, Interpretation, and ClinicalApplications. Edinburgh, Scotland: Churchill Livigstone, 1995.

8. Eitzen, I, Eitzen, TJ, Holm, I, Snyder-Mackler, L, and Risberg, MA.Anterior cruciate ligament-deficient potential copers andnoncopers reveal different isokinetic quadriceps strengthprofiles in the early stage after injury. Am J Sports Med 38: 586–593, 2010.

9. Enoka, RM and Fuglevand, AJ. Motor unit physiology: Someunresolved issues. Muscle Nerve 24: 4–17, 2001.

10. Faul, F, Erdfelder, E, Buchner, A, and Lang, AG. Statistical poweranalyses using G*Power 3.1: Tests for correlation and regressionanalyses. Behav Res Methods 41: 1149–1160, 2009.

11. Hiemstra, LA, Webber, S, MacDonald, PB, and Kriellaars, DJ.Hamstring and quadriceps strength balance in normal andhamstring anterior cruciate ligament-reconstructed subjects. ClinJ Sport Med 14: 274–280, 2004.

12. Hopkins, WG. Measures of reliability in sports medicine andscience. Sports Med 30: 1–15, 2000.

13. Impellizzeri, FM, Bizzini, M, Rampinini, E, Cereda, F, andMaffiuletti, NA. Reliability of isokinetic strength imbalance ratiosmeasured using the Cybex NORM dynamometer. Clin Physiol FunctImaging 28: 113–119, 2008.

14. Kadija, M, Knezevic, O, Milovanovic, D, Bumbasirevic, M, andMirkov, D. Effect of isokinetic dynamometer velocity on musclestrength deficit in elite athletes after ACL reconstruction. Med Sport63: 495–507, 2010.

15. Kannus, P. Ratio of hamstring to quadriceps femoris muscles’strength in the anterior cruciate ligament insufficient knee.Relationship to long-term recovery. Phys Ther 68: 961–965, 1988.

16. Knapik, JJ and Ramos, MU. Isokinetic and isometric torquerelationships in the human body. Arch Phys Med Rehabil 61: 64–67,1980.

17. Knezevic, OM and Mirkov, DM. Strength assessment in athletesfollowing an anterior cruciate ligament injury. Kinesiology 45: 3–15,2013.

18. Knezevic, OM, Mirkov, DM, Kadija, M, Milovanovic, D, and Jaric, S.Alternating consecutive maximum contraction as a test of musclefunction in athletes following ACL reconstruction. J Hum Kinet 35:5–13, 2012.

19. Lautamies, R, Harilainen, A, Kettunen, J, Sandelin, J, andKujala, UM. Isokinetic quadriceps and hamstring muscle strengthand knee function 5 years after anterior cruciate ligamentreconstruction: Comparison between bone-patellar tendon-boneand hamstring tendon autografts. Knee Surg Sports Traumatol Arthrosc16: 1009–1016, 2008.

20. Lord, JP, Aitkens, SG, McCrory, MA, and Bernauer, EM. Isometricand isokinetic measurement of hamstring and quadriceps strength.Arch Phys Med Rehabil 73: 324–330, 1992.

21. Maffiuletti, NA, Bizzini, M, Desbrosses, K, Babault, N, andMunzinger, U. Reliability of knee extension and flexionmeasurements using the Con-Trex isokinetic dynamometer. ClinPhysiol Funct Imaging 27: 346–353, 2007.

22. Marcora, S and Miller, MK. The effect of knee angle on the externalvalidity of isometric measures of lower body neuromuscularfunction. J Sports Sci 18: 313–319, 2000.

23. Meng, XL, Rosenthal, R, and Rubin, DB. Comparing correlatedcorrelation coefficients. Psychol Bull 111: 172–175, 1992.

24. Mirkov, DM, Nedeljkovic, A, Milanovic, S, and Jaric, S. Musclestrength testing: evaluation of tests of explosive force production.Eur J Appl Physiol 91: 147–154, 2004.

25. Moisala, AS, Jarvela, T, Kannus, P, and Jarvinen, M. Muscle strengthevaluations after ACL reconstruction. Int J Sports Med 28: 868–872,2007.




26. Neeter, C, Gustavsson, A, Thomee, P, Augustsson, J, Thomee, R, andKarlsson, J. Development of a strength test battery for evaluating legmuscle power after anterior cruciate ligament injury andreconstruction. Knee Surg Sports Traumatol Arthrosc 14: 571–580, 2006.

27. Pijnappels, M, Bobbert, MF, and van Dieen, JH. Push-off reactions inrecovery after tripping discriminate young subjects, older non-fallersand older fallers. Gait Posture 21: 388–394, 2005.

28. Reid, A, Birmingham, TB, Stratford, PW, Alcock, GK, andGiffin, JR. Hop testing provides a reliable and valid outcomemeasure during rehabilitation after anterior cruciate ligamentreconstruction. Phys Ther 87: 337–349, 2007.

29. Reinking, MF, Bockrath-Pugliese, K, Worrell, T, Kegerreis, RL,Miller-Sayers, K, and Farr, J. Assessment of quadriceps muscleperformance by hand-held, isometric, and isokinetic dynamometryin patients with knee dysfunction. J Orthop Sports Phys Ther 24: 154–159, 1996.

30. Ross, MD, Irrgang, JJ, Denegar, CR, McCloy, CM, and Unangst, ET.The relationship between participation restrictions and selected

clinical measures following anterior cruciate ligamentreconstruction. Knee Surg Sports Traumatol Arthrosc 10: 10–19, 2002.

31. Sahaly, R, Vandewalle, H, Driss, T, and Monod, H. Maximalvoluntary force and rate of force development in humans—Importance of instruction. Eur J Appl Physiol 85: 345–350, 2001.

32. Sole, G, Hamren, J, Milosavljevic, S, Nicholson, H, and Sullivan, SJ.Test-retest reliability of isokinetic knee extension and flexion. ArchPhys Med Rehabil 88: 626–631, 2007.

33. Staehli, S, Glatthorn, JF, Casartelli, N, and Maffiuletti, NA.Test-retest reliability of quadriceps muscle function outcomesin patients with knee osteoarthritis. J Electromyogr Kinesiol 20:1058–1065, 2010.

34. Suzovic, D, Nedeljkovic, A, Pazin, N, Planic, N, and Jaric, S.Evaluation of consecutive maximum contractions as a test ofneuromuscular function. J Hum Kinet 20: 51–67, 2008.

35. Wilson, GJ and Murphy, AJ. The use of isometric tests ofmuscular function in athletic assessment. Sports Med 22: 19–37,1996.


| www.nsca.com



Documents

Evaluation of isokinetic and isometric strength measures for monitoring muscle function recovery following ACL reconstruction