Escamilla Et Al. Cruciate Ligament Force During the Wall Squat and the One Leg Squat

Copyright @ 200 by the American College of Sports Medicine. Unauthorized reproduction of this article is prohibited.9

Cruciate Ligament Force during the WallSquat and the One-Leg Squat

RAFAEL F. ESCAMILLA1, NAIQUAN ZHENG2, RODNEY IMAMURA3, TORAN D. MACLEOD1,W. BRENT EDWARDS4, ALAN HRELJAC2, GLENN S. FLEISIG5, KEVIN E. WILK6,CLAUDE T. MOORMAN III7, and JAMES R. ANDREWS8

1Department of Physical Therapy, California State University, Sacramento, CA; 2The Center for Biomedical Engineering,Department of Mechanical Engineering and Engineering Science, University of North Carolina, Charlotte, NC; 3Kinesiologyand Health Science Department, California State University, Sacramento, CA; 4Department of Kinesiology, Iowa StateUniversity, Ames, IA; 5American Sports Medicine Institute, Birmingham, AL; 6Champion Sports Medicine, Birmingham, AL;7Duke University Medical Center, Durham, NC; and 8Andrews Research and Education Institute, Gulf Breeze, FL

ABSTRACT

ESCAMILLA, R. F., N. ZHENG, R. IMAMURA, T. D. MACLEOD, W. B. EDWARDS, A. HRELJAC, G. S. FLEISIG, K. E. WILK,

C. T. MOORMAN, and J. R. ANDREWS. Cruciate Ligament Force during the Wall Squat and the One-Leg Squat. Med. Sci. Sports

Exerc., Vol. 41, No. 2, pp. 408–417, 2009. Purpose: To compare cruciate ligament forces during wall squat and one-leg squat

exercises. Methods: Eighteen subjects performed the wall squat with feet closer to the wall (wall squat short), the wall squat with feet

farther from the wall (wall squat long), and the one-leg squat. EMG, force, and kinematic variables were input into a biomechanical

model using optimization. A three-factor repeated-measure ANOVA (P G 0.05) with planned comparisons was used. Results: Mean

posterior cruciate ligament (PCL) forces were significantly greater in 1) wall squat long compared with wall squat short (0-–80- knee

angles) and one-leg squat (0-–90- knee angles); 2) wall squat short compared with one-leg squat between 0-–20- and 90- knee angles;

3) wall squat long compared with wall squat short (70-–0- knee angles) and one-leg squat (90-–60- and 20-–0- knee angles); and 4)

wall squat short compared with one-leg squat between 90-–70- and 0- knee angles. Peak PCL force magnitudes occurred between 80-

and 90- knee angles and were 723 T 127 N for wall squat long, 786 T 197 N for wall squat short, and 414 T 133 N for one-leg squat.

Anterior cruciate ligament (ACL) forces during one-leg squat occurred between 0- and 40- knee angles, with a peak magnitude of 59 T

52 N at 30- knee angle. Quadriceps force ranged approximately between 30 and 720 N, whereas hamstring force ranged approximately

between 15 and 190 N. Conclusions: Throughout the 0-–90- knee angles, the wall squat long generally exhibited significantly greater

PCL forces compared with the wall squat short and one-leg squat. PCL forces were similar between the wall squat short and the one-leg

squat. ACL forces were generated only in the one-leg squat. All exercises appear to load the ACL and the PCL within a safe range in

healthy individuals. Key Words: BIOMECHANICS, KINETICS, CLOSED CHAIN EXERCISES, KNEE

Weight-bearing exercises, such as the squat, arecommonly used by athletes to train the hip andthe thigh musculature. Physical therapists and

trainers also have their patients or clients use squatting-typeexercises during anterior cruciate ligament (ACL) and pos-

terior cruciate ligament (PCL) rehabilitation to allow themto recover faster and return to function earlier (6,29,37).

Several studies involving barbell and body weight squatexercises reported PCL forces between 300 and 2700 Nand no ACL forces throughout the knee range of motion(8,11,12,31,35). In contrast, other squat studies reportedrelatively low magnitude peak ACL forces between 30 and500 N approximately between 0- and 60- knee anglesand PCL forces approximately between 60- and 120- kneeangles (3,14,25,32). These data are supported by otherweight-bearing knee flexion studies, with ACL strain oc-curring at lower knee angles and PCL strain occurring athigher knee angles (9,17). What is consistent in the squatliterature is that PCL loading occurs at higher knee anglestypically greater than approximately 60-. What is incon-sistent in the squat literature is whether or not ACLstrain always occurs at smaller knee angles. Part of the

Address for correspondence: Rafael F. Escamilla, Ph.D., P.T., C.S.C.S.,Department of Physical Therapy, California State University, Sacramento,6000 J Street, Sacramento, CA 95819-6020; E-mail: [email protected] for publication September 2007.Accepted for publication July 2008.

0195-9131/09/4102-0408/0MEDICINE & SCIENCE IN SPORTS & EXERCISE�

Copyright � 2009 by the American College of Sports Medicine

DOI: 10.1249/MSS.0b013e3181882c6d

408

APP

LIED

SCIENCES


inconsistencies in ACL strain during the squat is thatsome studies estimated ACL strain in vivo using strainsensors inserted within the ACL (3,15), whereas otherstudies used biomechanical musculoskeletal models toestimated ACL strain (8,11,12,31,35). However, it is clearthat when ACL strain does occur, it occurs at smallerknee angles and its strain or force magnitudes are relativelylow. Using dynamic optimization techniques, peak ACLforces have been reported to be less than 20 N during bodyweight squatting (30).

Although the effects of exercise technique variations oncruciate ligament strain while performing the barbell squathave been examined (11,12), there are no studies that haveexamined the effects of technique variations on cruciateligament loading while performing the one-leg squat andwall squat. One-leg squat and wall squat exercises are bothperformed in training and rehabilitation settings. Wallsquats can be performed with varying techniques, such aspositioning the heels farther or closer to the wall. Position-ing the heels farther from the wall typically results in theknees being maintained over the feet at the lowest positionof the squat, whereas positioning the heels closer to the walltypically results in the knees moving anterior beyond thetoes at the lowest position of the squat. Performing a one-leg squat also causes the knees to move forward beyond thetoes at maximum knee flexion. Clinicians and trainers oftenbelieve that anterior movement of the knees beyond the toesduring the wall squat or one-leg squat may increase cruciateligament loading, although there are very limited data thatsupport this belief (1). Moreover, it is unclear if the ACL orthe PCL is loaded when anterior knee movement occurs.

The purpose of this study was to compare cruciate liga-ment tensile forces among squat types (wall squat with thefeet farther away from the wall—wall squat long; wall squatwith the feet closer to the wall—wall squat short; and theone-leg squat) and squat phases (squat descent and squatascent) at specific knee angles (0-, 10-, 20-, 30-, 40-, 50-,60-, 70-, 80-, and 90-). It was hypothesized that 1) ACLtensile force would occur at knee angles 30- or less in theone-leg squat and wall squat short; 2) PCL tensile forcewould occur throughout the knee angle range of motionin the wall squat long; 3) PCL forces would generally begreater in the wall squat long compared with the wall squatshort and one-leg squat; 4) PCL forces would generally notbe significantly different between the wall squat short andthe one-leg squat; and 5) for all three squat types, ACL andPCL forces would generally be greater at specific kneeangles during the squat ascent compared with the corre-sponding knee angles during the squat descent. Quadricepsand hamstrings muscle force magnitudes will also be pre-sented to help better understand ACL and PCL force mag-nitudes. Understanding how cruciate ligament tensile forcesvary among squatting techniques will allow physical thera-pists, physicians, and trainers to prescribe safer and moreeffective knee rehabilitation to patients during ACL or PCLrehabilitation.

METHODS

Subjects

Eighteen healthy individuals (nine males and ninefemales) without a history of cruciate ligament pathologyparticipated with an average age, mass, and height of 29 T7 yr, 77 T 9 kg, and 177 T 6 cm, respectively, for males,and 25 T 2 yr, 60 T 4 kg, and 164 T 6 cm, respectively, forfemales. All subjects were required to perform wall squatand one-leg squat exercises pain-free and with proper formand technique for 12 consecutive repetitions using their 12repetition maximum (12 RM) weight.

To control the EMG signal quality, the current study waslimited to males and females that had average or belowaverage body fat, which was assessed by Baseline skinfoldcalipers (Model 68900; Country Technology, Inc., GaysMill, WI) and body fat standards set by the AmericanCollege of Sports Medicine. Average body fat was 12% T4% for males and 18% T 1% for females. All subjectsprovided written informed consent in accordance with theInstitutional Review Board at California State University,Sacramento, which approved the research conducted andinformed consent form.

Exercise Description

Wall squat (Figs. 1 and 2). The wall squat began withthe right foot on a force platform and their left foot on theground, both knees fully extended (0- knee angle), the backflat against the wall, and a dumbbell weight held in bothhands with the arms straight and at the subject’s side. Fromthis position, the subject slowly flexed both knees andsquatted down until the thighs were approximately parallelto the ground with the knees flexed approximately 90-–110-, and in a continuous motion the subject returned backto the starting position. A metronome was used to help en-sure that the knees flexed and extended at approximately45-Isj1. The surface of the wall was smooth, and a towelwas positioned between the wall and the subject to mini-mize friction as the subject slid down and up the wall. Thestance width (distance between inside heels) was 32 T 6 cmfor males and 28 T 7 cm for females, and the foot anglewas approximately 0- (feet pointing approximately straightahead), and both stance and foot angle were according tosubject preference.

The wall squat was performed with two technique varia-tions, wall squat long (Fig. 1) and wall squat short (Fig. 2).The foot position relative to the wall for the wall squat longwas determined using a heel-to-wall distance that resultedin the legs being approximately vertical and the knees po-sitioned above the ankles when the thighs were parallel withthe ground (Fig. 1), which is commonly recommended byclinicians and trainers. The average heel-to-wall distance forthe wall squat long was 45 T 3 cm for males and 41 T 3 cmfor females. The heel-to-wall distance for the wall squatshort was one half the distance of the heel-to-wall distance

ONE-LEG SQUAT AND WALL SQUAT Medicine & Science in Sports & Exercised 409

APPLIED

SCIEN

CES


for the wall squat long. The shorter heel-to-wall distance forthe wall squat short resulted in the anterior surface of theknee moving beyond the distal end of the toes 9 T 2 cm atthe lowest position of the wall squat short (Fig. 2).

One-leg squat. The one-leg squat started with the sub-ject standing on one leg with the right foot on a forceplatform, the right knee fully extended, the left knee bentapproximately 90-, and a single dumbbell weight held withboth hands in front of the chest. From this position, thesubject slowly flexed the right knee and squatted down untilthe right knee was flexed approximately 90–100- with thetrunk tilted forward approximately 30–40- (Fig. 3), and ina continuous motion the subject returned back to the start-ing position. A metronome was used to help ensure that theright knee flexed and extended at approximately 45-Isj1. Atthe lowest position of the one-leg squat, the anterior surfaceof the knee moved 10 T 2 cm beyond the distal end of thetoes (Fig. 3).

Data Collection

Each subject came in for a pretest 1 wk before the test-ing session. The experimental protocol was reviewed, thesubject was given the opportunity to practice the one-legsquat and wall squat exercises, and each subject’s heel-to-wall distances for the wall squat short and wall squat longwere determined. In addition, to normalize intensity be-tween the wall squat and the one-leg squat exercises, eachsubject’s 12 RM was determined. To determine the weight

used for the wall squat short and long, each subject usedtheir 12 RM weight while performing the wall squat using aheel-to-wall distance that was halfway between the heel-to-wall distances for the wall squat short and wall squat long,and this weight was used for both the wall squat short andthe wall squat long during the testing session. The meantotal dumbbell mass used was 56 T 9 kg for males and 36 T8 kg for females for the wall squat short and wall squat longand 15 T 3 kg for males and 10 T 3 kg for females for theone-leg squat.

Blue Sensor (Ambu Inc., Linthicum, MD) disposable sur-face electrodes (type M-00-S) were used to collect EMGdata. These oval-shaped electrodes (22 mm wide and 30 mmlong) were placed in a bipolar electrode configuration alongthe longitudinal axis of each muscle, with a center-to-centerdistance of approximately 3 cm. Before positioning the elec-trodes over each muscle, the skin was prepared by shaving,abrading, and cleaning with isopropyl alcohol wipes to re-duce skin impedance. As previously described (2), electrodepairs were then placed on the subject’s right side for thefollowing muscles: a) rectus femoris, b) vastus lateralis, c)vastus medialis, d) medial hamstrings (semimembranosus andsemitendinous), e) lateral hamstrings (biceps femoris), and f)gastrocnemius.

Spheres (3.8 cm in diameter) were attached to adhesivesand positioned over the following bony landmarks: a) third

FIGURE 1—Wall squat with feet farther from wall (wall squat long).

FIGURE 2—Wall squat with feet closer to wall (wall squat short).

http://www.acsm-msse.org410 Official Journal of the American College of Sports Medicine

APP

LIED

SCIENCES


metatarsal head of the right foot, b) medial and lateralmalleoli of the right leg, c) upper edges of the medial andthe lateral tibial plateaus of the right knee, d) postero-superior greater trochanters of the left and the right femurs,and e) lateral acromion of the right shoulder.

Once the electrodes and the spheres were positioned, thesubject warmed up and practiced the exercises as needed,and data collection was commenced. A six-camera peakperformance motion analysis system (Vicon-Peak Perfor-mance Technologies, Inc., Englewood, CO) was used to col-lect 60-Hz video data. Force data were collected at 960 Hzusing a force platform (Model OR6-6-2000; Advanced Me-chanical Technologies, Inc.). EMG data were collected at960 Hz using a Noraxon Myosystem unit (Noraxon USA,Inc., Scottsdale, AZ). The EMG amplifier bandwidth fre-quency was 10–500 Hz. Video, EMG, and force data wereelectronically synchronized and simultaneously collected aseach subject performed in a randomized manner one set ofthree continuous repetitions (trials) during the wall squatshort, wall squat long, and one-leg squat.

After completing all exercise trials, EMG data werecollected during maximum voluntary isometric contractions(MVIC) to normalize the EMG data collected during eachexercise (11). The MVIC for the rectus femoris, vastuslateralis, and vastus medialis were collected in a seatedposition at 90- knee and hip flexion with a maximum ef-fort knee extension. The MVIC for the lateral and themedial hamstrings were collected in a seated position at90- knee and hip flexion with a maximum effort kneeflexion. MVIC for the gastrocnemius was collected duringa maximum effort standing one-leg toe raise with theankle positioned approximately halfway between neutraland full plantarflexion. Two 5-s trials were randomly col-lected for each MVIC.

Data Reduction

Video images for each marker were tracked and digitizedin three-dimensional space with peak performance software.Ankle, knee, and hip joint centers were mathematicallydetermined using the external markers and appropriateequations as previously described (11). Testing of theaccuracy of the calibration system resulted in markers thatcould be located in three-dimensional space with an errorless than 4–7 mm. The raw position data were smoothedwith a double-pass fourth-order Butterworth low-pass filterwith a cutoff frequency of 6 Hz (11). Joint angles, linearand angular velocities, and linear and angular accelerationswere calculated using appropriate kinematic equations (11).

Raw EMG signals were full-waved rectified andsmoothed with a 10-ms moving average window through-out the knee range of motion for each repetition. TheseEMG data were then normalized for each muscle and wereexpressed as a percentage of each subject’s highest cor-responding MVIC trial. The MVIC trials were calculatedusing the highest EMG signal over a 1-s time intervalthroughout the 5-s MVIC. Normalized EMG data were thenaveraged over the three repetition trials performed for eachexercise as a function of knee angle and were used in thebiomechanical model described below.

Biomechanical Model

As previously described (11,41), a biomechanical modelof the knee (Figs. 4 and 5) was used to continuously esti-mate cruciate ligament forces throughout a 90- knee rangeof motion during the knee flexing (squat descent) phase(0-–90-) and the knee extending (squat ascent) phase (90-–0-) of the lunge. Resultant force and torque equilibriumequations were calculated using the inverse dynamics andthe biomechanical knee model (11,41). Anteroposteriorshear forces in the knee were calculated and adjusted toligament orientations to estimate ACL or PCL forces (16).Moment arms of muscle forces and angles for the line ofaction for the muscles and the cruciate ligaments were ex-pressed as polynomial functions of knee angle using datafrom Herzog and Read (16). Knee torques from cruciateand collateral ligament forces and bony contact were as-sumed to be negligible as were forces and torques out ofthe sagittal plane.

Quadriceps, hamstrings, and gastrocnemius muscleforces were estimated as previously described (11,41). Be-cause the accuracy of estimating muscle forces dependson accurate estimations of a muscle’s physiological cross-sectional area (PCSA), maximum voluntary contractionforce per unit PCSA, and the EMG–force relationship, re-sultant force and torque equilibrium equations may not besatisfied. Therefore, each muscle force Fm(i) was modifiedby the following equation at each knee angle:

FmðiÞ ¼ ciklikviAiRmðiÞ½EMGi=MVICi�;

FIGURE 3—One-leg squat.


APPLIED

SCIEN

CES


where Ai is the PCSA of the ith muscle, Rm(i) is the MVICforce per unit PCSA of the ith muscle, EMGi and MVICi

are the EMG window averages of the ith muscle EMGduring exercise and MVIC trials, ci is a weight factor(values given below) adjusted in a computer optimizationprogram to minimize the difference between the resultanttorque from the inverse dynamics (Tres) and the resultanttorque calculation from the biomechanical model (Tmi)(Fig. 4), kli represents each muscle’s force–length rela-tionship as function of hip and knee angles (based onmuscle length, fiber length, sarcomere length, pennationangle, and cross-sectional area) (33), and kvi represents eachmuscle’s force–velocity relationship based on a Hill-typemodel for eccentric and concentric muscle actions usingthe following equations from Zajac (38) and Epstein andHerzog (10):

kv ¼ ðb� ða=F0ÞvÞ=ðbþ vÞ concentrickv ¼ C � ðC � 1Þðbþ ða=F0ÞvÞ=ðb� vÞ eccentric ;

with F0 representing the isometric muscle force, l0 is themuscle fiber length at rest, v is the velocity, and a = 0.32F0,b = 3.2l0Is

j1, and C = 1.8.PCSA data from Wickiewicz et al. (33) were used to

determine the ratios of PCSA between muscle groups (41).According to Narici et al. (24), the total PCSA of the quad-riceps was approximately 160 cm2 for a 75-kg man. TotalPCSA of the quadriceps was scaled up or down by indi-vidual body mass (41). Forces generated by the knee flexorsand extensors at MVIC were assumed to be linearly propor-tional to their PCSA (41). Muscle force per unit PCSA atMVIC was 35 NIcmj2 for the knee flexors and 40 NIcmj2

for the quadriceps (7,23,24,34).

The objective function used to determine each ith mus-cle’s coefficient ci was as follows:

min f ðci Þ ¼ ~nm

i¼1ð1� ciÞ2 þ ðTres � ~

nm

i¼1TmiÞ2;

subject to clow e ci e chigh, where clow and chigh are thelower and the upper limits for ci, and L is a constant. Theweight factor c was to adjust the final muscle force cal-culation. The bounds on c were set between 0.5 and 1.5.The torques predicted by the EMG driven model matchedwell (G2%) with the torques generated from the inversedynamics.

Data Analysis

To determine the effects of squat type (wall squat long,wall squat short, and one-leg squat), squat phase (squatdescent and squat ascent), and knee angles (0-–90- in10- intervals) on cruciate ligament forces, a three-factorrepeated-measure ANOVA with planned comparisons wasused. Bonferroni t-tests were used to evaluate the signifi-cance of pairwise comparisons. The level of significanceused was P G 0.05.

RESULTS

Mean cruciate ligament force curves are shown inFigure 6. Main effect differences were identified amongthe three squat types (P G 0.001), between the two squatphases (P G 0.001), and among the 10 knee angles (P G0.001). When examined at each knee angle, a significantsquat type by squat phase interaction was identified at 0-(P = 0.039), 10- (P = 0.002), 20- (P = 0.003), 30- (P =0.011), 40- (P = 0.010), 50- (P G 0.001), 60- (P = 0.048),80- (0.003), and 90- (P G 0.001). Pairwise comparisons of

FIGURE 4—Computer optimization with input from measured kneetorque from inverse dynamics and predicted knee torque from musclemodel, where TK is the resultant knee torque, FK is the resultant kneeforce, I is the moment of inertia about leg center of mass, > is theangular acceleration of leg, m is the mass of leg, a is the linearacceleration of leg, g is the gravitation constant 9.80 mIsj2, Fext is theexternal force acting on foot, Text is the external torque acting on foot,FQ is the quadriceps force, FP is the patellar tendon force, FH is thehamstrings force, and FG is the gastrocnemius force. Note that tosimplify the drawing, the equal and the opposite forces and torquesacting on the distal leg and proximal ankle are not shown.

FIGURE 5—Forces acting on the proximal tibia: FH = force fromhamstrings; FG = force from gastrocnemius (note that this force doesnot act directly on proximal tibia); FPT = force from patellar tendon;FACL = force from ACL; FPCL = force from PCL; and FTF = forcefrom femur.

L


APP

LIED

SCIENCES


mean cruciate ligament forces at specific knee angles (0-–90-) between squat exercises and between squat descentand ascent phases are shown in Table 1. During the squatdescent phase, mean PCL forces were significantly greaterin the wall squat long (259–757 N range) compared withthe wall squat short (100–786 N range) between 0- and 80-knee angles, significantly greater in the wall squat longcompared with the one-leg squat (64–414 N) between 0-and 90- knee angles, and significantly greater in the wallsquat short compared with the one-leg squat between 0-–20- and 90- knee angles. During the squat ascent phase,mean PCL forces were significantly greater in the wallsquat long compared with the wall squat short between 70-and 0- knee angles, significantly greater in the wall squatlong compared with the one-leg squat between 90-–60- and20–0- knee angles, and significantly greater in the wall

squat short compared with the one-leg squat between 90-–70- and 0- knee angles. For all three squat exercises, meanpeak PCL force magnitudes occurred between 80- and 90-knee angles during the squat ascent and were 723 T 127 Nfor the wall squat long, 786 T 197 N for the wall squatshort, and 414 T 133 N for the one-leg squat. ACL forces,which were generated only during the one-leg squat (31–59N range), occurred between 0- and 40- knee angles duringthe squat descent and at 0- knee angle during the squatascent. The mean peak ACL force magnitude during theone-leg squat was 59 T 52 N and occurred at 30- knee angleduring the squat descent.

Significant differences (P G 0.05) in cruciate ligamentforce at specified knee angles between the descent and theascent phases of each squat exercise are shown in Table 1.Mean PCL force was significantly greater in the ascent

TABLE 1. Mean T SD cruciate ligament forces (N) among the three squat types (wall squat long, wall squat short, and one-leg squat) and between the two squat phases (squat decentand squat ascent) as a function of knee angle.

Knee Angles for Descent PhaseWall Squat Long

(WSL)Wall Squat Short

(WSS)One-Leg Squat

(OLS) Significant Differences (P G 0.05) between Squat Types

0- 482 T 209 297 T 152 j31 T 52 WSL 9 WSS (P = 0.002); WSL 9 OLS (P G 0.001); WSS 9 OLS (P G 0.001)10- 423 T 205 243 T 140 j36 T 54 WSL 9 WSS (P = 0.011); WSL 9 OLS (P G 0.001); WSS 9 OLS (P G 0.001)20- 316 T 135 143 T 131 j51 T 77* WSL 9 WSS (P = 0.049); WSL 9 OLS (P G 0.001); WSS 9 OLS (P G 0.004)30- 261 T 124 100 T 100 j59 T 52* WSL 9 WSS (P = 0.023); WSL 9 OLS (P G 0.001)40- 259 T 121 109 T 77 j22 T 66* WSL 9 WSS (P = 0.014); WSL 9 OLS (P = 0.002)50- 295 T 122 157 T 70 64 T 93* WSL 9 WSS (P = 0.005); WSL 9 OLS (P G 0.001)60- 348 T 133* 231 T 81 160 T 97* WSL 9 WSS (P = 0.034); WSL 9 OLS (P G 0.001)70- 445 T 136* 324 T 101* 227 T 81* WSL 9 WSS (P = 0.022); WSL 9 OLS (P G 0.001)80- 573 T 149* 439 T 129* 326 T 118 WSL 9 WSS (P = 0.017); WSL 9 OLS (P G 0.001)90- 659 T 150 578 T 158* 386 T 121 WSL 9 OLS (P G 0.001); WSS 9 OLS (P = 0.003)Knee angles for ascent phase90- 723 T 127 786 T 197* 414 T 133 WSL 9 OLS (P G 0.001); WSS 9 OLS (P G 0.001)80- 757 T 185* 702 T 200* 391 T 169 WSL 9 OLS (P G 0.001); WSS 9 OLS (P G 0.001)70- 714 T 181* 529 T 177* 368 T 157* WSL 9 WSS (P G 0.001); WSL 9 OLS (P G 0.001); WSS 9 OLS (P = 0.035)60- 542 T 144* 366 T 146 374 T 178* WSL 9 WSS (P G 0.001); WSL 9 OLS (P = 0.002)50- 408 T 137 267 T 141 329 T 172* WSL 9 WSS (P G 0.001)40- 355 T 120 223 T 174 266 T 159* WSL 9 WSS (P = 0.020)30- 363 T 141 206 T 158 231 T 132* WSL 9 WSS (P = 0.007)20- 436 T 180 222 T 139 209 T 142* WSL 9 WSS (P G 0.001); WSL 9 OLS (P = 0.005)10- 539 T 223 253 T 155 88 T 130 WSL 9 WSS (P G 0.001); WSL 9 OLS (P G 0.001)0- 529 T 249 274 T 178 j37 T 146 WSL 9 WSS (P = 0.003); WSL 9 OLS (P G 0.001); WSS 9 OLS (P G 0.001)

ACL forces are listed as negative values, and PCL forces are listed as positive values. An asterisk (*) implies that there is a significant difference (P G 0.05) in cruciate ligament force atthe specified knee angle between the squat descent and the squat ascent phases of a squat exercise.

FIGURE 6—Mean (SD) cruciate ligament force during the one-leg squat and wall squat.


APPLIED

SCIEN

CES


phase compared with the descent phase between 60- and80- knee angles for the wall squat long, 70-–90- kneeangles for the wall squat short, and 20-–70- knee angles forthe one-leg squat. Descriptive data of mean quadriceps andhamstrings force values during wall squat and one-leg squatexercises are shown in Table 2. Quadriceps force rangedapproximately between 30 and 720 N and generally in-creased with knee flexion, whereas hamstring force rangedapproximately between 15 and 190 N. At each knee angle,quadriceps and hamstrings forces were generally greaterduring the ascent compared with the descent.

DISCUSSION

It is not well understood what PCL or ACL force magni-tudes become injurious to the healthy or reconstructed ACLand PCL. In healthy adults, the ultimate strength of theACL and PCL is approximately 2000 N (36) and 4000 N(27), respectively, although these values depend on age andanatomical factors. Therefore, the ACL and the PCL loadsgenerated during the one-leg squat and the wall squatexercises appear to be well within a safe limit for thehealthy ACL and PCL. The reconstructed ACL and PCLhave similar ultimate strengths compared with the healthyACL or PCL, although these values can change consider-ably depending on graft type and donor characteristics(e.g., autograft vs allograft; patellar tendon vs hamstringsgraft) (4,28). However, the healing graft site may be injuredwith considerably less force compared with the ultimatestrength of the graft, although it is not well understood howmuch force to the graft site is too much and how soon forcecan be applied after reconstruction. Therefore, the meanpeak PCL forces of approximately 400 N during the one-legsquat and approximately 750 N during the wall squat exer-cises may be problematic early after PCL reconstructionwhen the graft site is still healing. Moreover, during PCL

reconstruction, at the same relative intensity, it may be ap-propriate to use the one-leg squat before wall squat exer-cises due to less PCL loading during the one-leg squat,especially compared with the wall squat long. In addition,it may be prudent to use smaller knee angles (e.g., 0-–50-) before progressing to larger knee angles (e.g., 50-–100-) because PCL forces generally increase as knee angleincreases. In contrast, wall squat exercises may be a betterchoice compared with the one-leg squat early after ACLreconstruction due to ACL forces generated during the one-leg squat. However, because peak ACL force duringthe one-leg squat were only approximately 60 N, it is notlikely that the one-leg squat will produce forces thatwould be injurious to the healing ACL graft, and mildstrain to the graft may enhance the healing process (13).Nevertheless, after ACL reconstruction, it may be safer tostart with wall squat exercises and progress to the one-leg squat and use larger knee angles (e.g., 50-–100-)before progressing to smaller knee angles (e.g., 0-–50-)because ACL forces may be generated at smaller kneeangles less than 50-.

As hypothesized, ACL forces were greater in the one-leg squat compared with the wall squat long and occurredat knee angles between 0- and 40- with a peak magnitudeof approximately 60 N at 30- knee angle. During the one-leg sit-to-stand, which is similar to ascent phase of theone-leg squat, Heijne et al. (15) reported a peak 2.8% ACLstrain (calibrated to approximately 100 N) at 30- kneeangle. Moreover, Kvist and Gillquist (19) reported a peakanterior shear ACL force of less than 90 N at 30- kneeangle during the one-leg bodyweight squat, which is similarto the results in the current study. Butler et al. (5)demonstrated that the ACL provides 86% of the totalresistance to anterior drawer (caused by an anterior shearforce) and the PCL provides approximately 95% of the totalrestraining force to posterior drawer (caused by a posterior

TABLE 2. Mean T SD quadriceps and hamstrings force values during wall squat and one-leg squat exercises.

Quadriceps Force (N) Hamstrings (N)

Knee Angles for Descent Phase Wall Squat Long Wall Squat Short One-Leg Squat Wall Squat Long Wall Squat Short One-Leg Squat

0- 63 T 57 31 T 29 50 T 50 35 T 24 19 T 9 47 T 2910- 83 T 68 53 T 51 63 T 59 36 T 23 19 T 10 40 T 2820- 108 T 88 84 T 74 90 T 64 35 T 21 22 T 24 39 T 2930- 150 T 114 116 T 83 145 T 84 34 T 22 25 T 32 51 T 2440- 209 T 145 164 T 98 253 T 139 31 T 20 21 T 17 59 T 3050- 290 T 174 235 T 123 357 T 160 29 T 19 21 T 18 61 T 3360- 384 T 190 318 T 145 542 T 181 23 T 13 19 T 16 61 T 3370- 478 T 203 408 T 155 668 T 172 18 T 8 16 T 12 50 T 3380- 502 T 171 486 T 156 645 T 178 21 T 12 13 T 9 47 T 2590- 419 T 158 559 T 159 593 T 168 15 T 11 15 T 9 56 T 26Knee angles for ascent phase90- 505 T 151 595 T 205 450 T 154 32 T 21 53 T 35 97 T 5780- 475 T 145 705 T 240 469 T 144 49 T 24 60 T 41 81 T 5170- 684 T 196 717 T 286 570 T 160 63 T 38 55 T 40 97 T 5160- 632 T 227 643 T 260 594 T 157 54 T 28 49 T 39 132 T 6050- 499 T 217 525 T 217 548 T 132 49 T 26 45 T 38 152 T 6540- 358 T 161 403 T 175 442 T 121 47 T 27 42 T 39 166 T 7230- 263 T 110 304 T 139 344 T 98 50 T 31 40 T 40 182 T 8220- 197 T 86 221 T 102 253 T 78 55 T 33 40 T 38 192 T 10010- 139 T 62 146 T 73 161 T 75 61 T 36 39 T 37 178 T 1120- 93 T 48 86 T 51 97 T 83 66 T 48 44 T 40 149 T 112


APP

LIED

SCIENCES


shear force). Therefore, the anterior shear force is resistedprimarily by the ACL, and posterior shear force is resistedprimarily by the PCL. Moreover, ACL forces as a functionof knee angle in the current study are similar to ACL forcesand knee angles in the squat literature (3,15,25,32).However, both the ACL and the PCL forces that aregenerated while performing squatting exercises are depen-dent on which exercise technique is used and whetherexternal resistance is used. For example, in Beynnon et al.(3), it appears that the subjects may have squatted using amore upright trunk position with relatively little forwardtrunk tilt, which suggests that these subjects may use theirquadriceps to a greater extent than their hamstrings (26).This is important because hamstrings force has been shownto unload the ACL and to load the PCL during the weight-bearing squat exercise (11,21,26). Ohkoshi et al. (26)reported no ACL strain at all knee angles tested (15-, 30-,60-, and 90-) while maintaining a squat position with thetrunk tilted forward, with 30- or more forward trunk tiltbeing optimal for eliminating or minimizing ACL strainthroughout the knee range of motion and recruitingrelatively high hamstrings activity.

The exercises that had the greatest amount of anteriorknee movement beyond the knees, the one-leg squat (10 T2 cm) and wall squat short (9 T 2 cm), also generated thegreatest ACL forces and least PCL forces. These exercisesmay be preferable to the wall squat long during PCL re-habilitation. In contrast, as hypothesized, the wall squatlong, in which the knees did not move beyond the toes,generated the highest PCL forces and no ACL forces andmay be problematic during PCL rehabilitation. Anteriorknee movement beyond the toes can influence quadricepsactivity and patellar tendon force, which in turn can influ-ence cruciate ligament loading. Zernicke et al. (40) esti-mated the force in the patellar tendon at approximately 17times bodyweight in a subject that used a considerableexternal load during a squat descent with excessive ante-rior knee movement beyond the toes. Although 17 timesbodyweight may be an over estimate of the actual force inthe patella tendon, large patellar tendon forces tend to loadthe ACL at smaller knee angles less than approximately60- (primarily between 0- and 30-) but load the PCL atlarger knee angles greater than approximately 60- (9,17,18).Although patellar tendon force from quadriceps activitycan load either the ACL or the PCL depending on kneeangle, it is difficult to make definite conclusions regardinghow quadriceps activity and anterior knee movement mayinfluence cruciate ligament loading while performing squatexercises, and additional research in this area is needed.

Although the wall squat short and one-leg squat bothresulted in similar amounts of anterior knee movement atmaximum knee flexion, PCL forces were significantly lowerin the one-leg squat compared with the wall squat shortbetween 90- and 70- knee angles during the squat ascent(Table 1 and Fig. 6). One explanation of the greater PCLforces between 90- and 70- knee angles in the wall squat

short compared with the one-leg squat is greater quadricepsforces that are generated during the wall squat short becausequadriceps forces at knee angles greater than 60- load thePCL (9,17,18). Between 90- and 70- knee angles during theascent, the estimated quadriceps forces in the current studywere approximately 30–50% greater in the wall squat shortcompared with the one-leg squat. Although hamstrings forcesbetween 90- and 70- knee angles also load the PCL, ham-strings forces were only 20–30 N greater in the one-leg squatcompared the wall squat short. In contrast, quadriceps forcemagnitudes were approximately 150 N greater in the wallsquat short compared with the one-leg squat, therefore load-ing the PCL to a great extent compared with the hamstrings.

Although hamstrings forces were greatest in the one-legsquat between 0- and 30- knee angles, the hamstrings arenot effective in either unloading the ACL or loading thePCL due to a small insertion angle into the tibia that resultsin most of the hamstrings force being directed parallelinstead of perpendicular to the tibia. Hamstrings force ismost effective in generating posterior shear force and inloading the PCL when the knee is flexed approximately90- (20). The relatively low hamstrings force (typicallyless than 50 N) generated during the wall squat exercisesthroughout the knee range of motion implies that wallsquat exercises primarily target the quadriceps and not thehamstrings, whereas the one-leg squat is more effective inrecruiting the hamstrings. One reason for greater quadricepsforce and less hamstrings force in the wall squat shortcompared with the one-leg squat is because the trunk iserect in the wall squat short (greater knee extensor torqueand less hip extensor torque needed to overcome the effectsof gravity) but tilted forward 30-–40- in the one-leg squat(less knee extensor torque and greater hip extensor torqueneeded to overcome the effects of gravity).

The friction and the normal forces that the wall applied tothe subject may also help explain why quadriceps forceswere greater in the wall squat short compared with the one-leg squat during the squat ascent. Although friction wasminimized during the wall squat by using a smooth wall,the normal force that the wall exerted on the subject’s backduring the wall squat exercises resulted in an increasedfriction force on the subject as they slid down and up thewall. Because the friction force opposes motion, it actedopposite the force of gravity during the squat descent butin the same direction as the force of gravity during thesquat ascent. Therefore, the friction force made it easier forthe subject to control the rate of sliding down the wall byproducing a knee extensor torque but made it more difficultfor the subject to slide up the wall by producing a kneeflexor torque. Because the one-leg squat did not have afriction force compare to the wall squat, this provides oneplausible explanation why quadriceps force and PCL forcewere greater in the ascent phase of the wall squat exercisescompared with the one-leg squat.

The friction force also differed between the wall squatlong and short. Because during the wall squat long the heels


APPLIED

SCIEN

CES


were twice as far from the wall compared with the wallsquat short, the normal force must be greater in the wallsquat long. Because friction force is directly proportional tothe normal force, the downward-acting friction force on thesubject during the squat ascent was greater in the wall squatlong compared with the wall squat short, which makes thewall squat long more difficult to perform. This may par-tially explain why PCL forces were greater in the wall squatlong compared with the wall squat short.

Cruciate ligament forces tended to be higher in the ascentphase compared with the descent phase, in part becausequadriceps and hamstrings forces were also greater duringthe ascent phase. For the wall squat exercises, significantPCL force differences between squat descent and ascentoccurred only at higher knee angles between 60- and 90-.As previously mentioned, quadriceps force at knee anglesgreater than 60- loads the PCL, and the greater quadricepsforce was greater during the ascent than the descent in partdue to having to overcome gravity and the downward-acting friction force. A different pattern occurred during theone-leg squat, in which between 20- and 70- knee anglesPCL forces were significantly greater during the squat as-cent compared with the squat descent. These findings are inagreement with the squat literature, in which cruciate forceshave been reported to be greater in the squat ascent com-pared with the squat descent (11,12).

There are limitations in this study. Firstly, muscle andcruciate ligament forces were estimated from biomechanicalmodeling techniques and not measured directly because itis currently not possible to measure cruciate ligament forcesin vivo while performing wall squat and one-leg squat ex-ercises in healthy subjects. However, both Beynnon et al.(3) and Heijne (15), who implanted strain sensors in pa-tients within the anteromedial bundle of an ACL duringarthroscopic surgery for partial minisectomies or capsule/patellofemoral joint debridement, after surgery had these pa-tients perform one- and two-leg squat-type exercises. Theseauthors reported a peak ACL strain of approximately 2.8–4% (approximately 100–150 N) at knee angles between 0-and 30-. These ACL force magnitudes and knee anglesfrom Beynnon et al. (3) and Heijne (15) are similar to thecurrent study. Unfortunately, there are no studies that havequantified PCL forces in vivo while performing a squatexercise, so it is not possible to compare the modeled PCLforce results in the current study to in vivo PCL forces.

The current study was limited to sagittal plane motiononly, and only subjects who could perform all exerciseswithout discernable frontal or transverse plane movementswere used in this study. Future three-dimensional biome-chanical analyses of the knee during squatting are needed toinvestigate the effects of transverse plane rotary motionsand frontal plane valgus and varus motions on cruciateligament loading. Slightly different cruciate ligament load-ing patterns during squatting may occur between two- andthree-dimensional analyses, although normal squatting isprimarily sagittal plane movements. A normal range ofmotion of 5-–7- knee valgus and 6-–14- of knee varus hasbeen reported during the one-leg squat (39), although theserelatively small amounts of valgus and varus may not affectcruciate ligament loading. However, excessive knee valgushas been shown to be associated with an increased risk ofACL ruptures (22,39). Transverse and frontal plane hipjoint motions have also been shown to be associated with anincreased risk of ACL ruptures and are relatively commonin individuals with weak hip abductors and externalrotations (22).

In conclusion, throughout the 0-–90- knee angles, thewall squat long generally exhibited significantly greaterPCL forces compared with the wall squat short and the one-leg squat. There was generally no significant difference inPCL force between the wall squat short and the one-legsquat, except at 80- and 90- knee angles, where PCL forceswere greater in the wall squat short. Throughout the 0-–90-knee angles, the wall squat exercises generated PCL forcemagnitudes ranging approximately from 100 to 790 N, withPCL magnitudes generally decreasing between 0- and 30-knee angles and increasing between 40- and 90- kneeangles. Moreover, the one-leg squat generated PCL forcemagnitudes ranging approximately from 60 to 410 N, withPCL magnitudes generally increasing between 50- and 90-knee angles during the descent and 10-–90- knee anglesduring the ascent. ACL forces were only found in the one-leg squat, which generated relatively small magnitudes ofapproximately 20–60 N between 0- and 40- knee angles.The one-leg squat, the wall squat long, and the wall squatshort all appear to load the ACL and the PCL within a saferange in healthy individuals.

The authors would like to thank Lisa Bonacci, Toni Burnham,Juliann Busch, Kristen D’Anna, Pete Eliopoulos, and Ryan Mowbrayfor all their assistance during data collection and analyses.

REFERENCES

1. Ariel BG. Biomechanical analysis of the knee joint during deepknee bends with heavy loads. In: Nelson R, Morehouse C, editors.Biomechanics IV. Baltimore: University Park Press; 1974. p. 44–52.

2. Basmajian JV, Blumenstein R. Electrode Placement in EMGBiofeedback. Baltimore: Williams and Wilkins; 1980. p. 79–86.

3. Beynnon BD, Johnson RJ, Fleming BC, Stankewich CJ, RenstromPA, Nichols CE. The strain behavior of the anterior cruciate liga-ment during squatting and active flexion–extension. A comparisonof an open and a closed kinetic chain exercise. Am J Sports Med.1997;25(6):823–9.

4. Brown CH Jr, Steiner ME, Carson EW. The use of hamstringtendons for anterior cruciate ligament reconstruction. Techniqueand results. Clin Sports Med. 1993;12(4):723–56.

5. Butler DL, Noyes FR, Grood ES. Ligamentous restraints toanterior–posterior drawer in the human knee. A biomechanicalstudy. J Bone Joint Surg Am. 1980;62(2):259–70.

6. Bynum EB, Barrack RL, Alexander AH. Open versus closedchain kinetic exercises after anterior cruciate ligament reconstruc-tion: a prospective randomized study. Am J Sports Med. 1995;23(4):401–6.


APP

LIED

SCIENCES


7. Cholewicki J, McGill SM, Norman RW. Comparison of muscleforces and joint load from an optimization and EMG assistedlumbar spine model: towards development of a hybrid approach.J Biomech. 1995;28(3):321–31.

8. Dahlkvist NJ, Mayo P, Seedhom BB. Forces during squatting andrising from a deep squat. Eng Med. 1982;11(2):69–76.

9. DeFrate LE, Gill TJ, Li G. In vivo function of the posteriorcruciate ligament during weightbearing knee flexion. Am J SportsMed. 2004;32(8):1923–8.

10. Epstein M, Herzog W. Theoretical Models of Skeletal Muscle:Biological and Mathematical Considerations. New York: JohnWiley & Sons; 1998. p. 238.

11. Escamilla RF, Fleisig GS, Zheng N, Barrentine SW, Wilk KE,Andrews JR. Biomechanics of the knee during closed kinetic chainand open kinetic chain exercises.Med Sci Sports Exerc. 1998;30(4):556–69.

12. Escamilla RF, Fleisig GS, Zheng N, et al. Effects of techniquevariations on knee biomechanics during the squat and leg press.Med Sci Sports Exerc. 2001;33(9):1552–66.

13. Fitzgerald GK. Open versus closed kinetic chain exercise: issuesin rehabilitation after anterior cruciate ligament reconstructivesurgery. Phys Ther. 1997;77(12):1747–54.

14. Hattin HC, Pierrynowski MR, Ball KA. Effect of load, cadence,and fatigue on tibio-femoral joint force during a half squat. MedSci Sports Exerc. 1989;21(5):613–8.

15. Heijne A, Fleming BC, Renstrom PA, Peura GD, Beynnon BD,Werner S. Strain on the anterior cruciate ligament during closedkinetic chain exercises. Med Sci Sports Exerc. 2004;36(6):935–41.

16. Herzog W, Read LJ. Lines of action and moment arms of themajor force-carrying structures crossing the human knee joint.J Anat. 1993;182(Pt 2):213–30.

17. Jordan SS, DeFrate LE, Nha KW, Papannagari R, Gill TJ, Li G.The in vivo kinematics of the anteromedial and posterolateral bun-dles of the anterior cruciate ligament during weightbearing kneeflexion. Am J Sports Med. 2007;35(4):547–54.

18. Kaufman KR, An KN, Litchy WJ, Morrey BF, Chao EY. Dynamicjoint forces during knee isokinetic exercise. Am J Sports Med.1991;19(3):305–16.

19. Kvist J, Gillquist J. Sagittal plane knee translation and electro-myographic activity during closed and open kinetic chain exer-cises in anterior cruciate ligament-deficient patients and controlsubjects. Am J Sports Med. 2001;29(1):72–82.

20. Markolf KL, O’Neill G, Jackson SR, McAllister DR. Effects ofapplied quadriceps and hamstrings muscle loads on forces in theanterior and posterior cruciate ligaments. Am J Sports Med. 2004;32(5):1144–9.

21. More RC, Karras BT, Neiman R, Fritschy D, Woo SL, DanielDM. Hamstrings—an anterior cruciate ligament protagonist. An invitro study. Am J Sports Med. 1993;21(2):231–7.

22. Myer GD, Chu DA, Brent JL, Hewett TE. Trunk and hip controlneuromuscular training for the prevention of knee joint injury.Clin Sports Med. 2008;27(3):425–48.

23. Narici MV, Landoni L, Minetti AE. Assessment of human kneeextensor muscles stress from in vivo physiological cross-sectionalarea and strength measurements. Eur J Appl Physiol. 1992;65(5):438–44.

24. Narici MV, Roi GS, Landoni L. Force of knee extensor and flexor

muscles and cross-sectional area determined by nuclear magneticresonance imaging. Eur J Appl Physiol. 1988;57(1):39–44.

25. Nisell R, Ekholm J. Joint load during the parallel squat in power-lifting and force analysis of in vivo bilateral quadriceps tendonrupture. Scand J Sports Sci. 1986;8(2):63–70.

26. Ohkoshi Y, Yasuda K, Kaneda K, Wada T, Yamanaka M. Bio-mechanical analysis of rehabilitation in the standing position. AmJ Sports Med. 1991;19(6):605–11.

27. Race A, Amis AA. The mechanical properties of the two bundlesof the human posterior cruciate ligament. J Biomech. 1994;27(1):13–24.

28. Schatzmann L, Brunner P, Staubli HU. Effect of cyclic precondi-tioning on the tensile properties of human quadriceps tendons andpatellar ligaments. Knee Surg Sports Traumatol Arthrosc. 1998;6(suppl 1):S56–61.

29. Shelbourne KD, Nitz P. Accelerated rehabilitation after anteriorcruciate ligament reconstruction. Am J Sports Med. 1990;18(3):292–9.

30. Shelburne KB, Pandy MG. A dynamic model of the knee andlower limb for simulating rising movements. Comput MethodsBiomech Biomed Engin. 2002;5(2):149–59.

31. Stuart MJ, Meglan DA, Lutz GE, Growney ES, An KN. Com-parison of intersegmental tibiofemoral joint forces and muscleactivity during various closed kinetic chain exercises. Am J SportsMed. 1996;24(6):792–9.

32. Toutoungi DE, Lu TW, Leardini A, Catani F, O’Connor JJ.Cruciate ligament forces in the human knee duringrehabilitation exercises. Clin Biomech. 2000;15(3):176–87.

33. Wickiewicz TL, Roy RR, Powell PL, Edgerton VR. Muscle archi-tecture of the human lower limb. Clin Orthop Relat Res. 1983;(179):275–83.

34. Wickiewicz TL, Roy RR, Powell PL, Perrine JJ, Edgerton VR.Muscle architecture and force–velocity relationships in humans.J Appl Physiol. 1984;57(2):435–43.

35. Wilk KE, Escamilla RF, Fleisig GS, Barrentine SW, Andrews JR,Boyd ML. A comparison of tibiofemoral joint forces and electro-myographic activity during open and closed kinetic chain exer-cises. Am J Sports Med. 1996;24(4):518–27.

36. Woo SL, Hollis JM, Adams DJ, Lyon RM, Takai S. Tensileproperties of the human femur-anterior cruciate ligament-tibiacomplex. The effects of specimen age and orientation. Am JSports Med. 1991;19(3):217–25.

37. Yack HJ, Collins CE, Whieldon TJ. Comparison of closed andopen kinetic chain exercise in the anterior cruciate ligament-deficient knee. Am J Sports Med. 1993;21(1):49–54.

38. Zajac FE. Muscle and tendon: properties, models, scaling, andapplication to biomechanics and motor control. Crit Rev BiomedEng. 1989;17(4):359–411.

39. Zeller BL, McCrory JL, Kibler WB, Uhl TL. Differences in kine-matics and electromyographic activity between men and women dur-ing the single-legged squat. Am J Sports Med. 2003;31(3):449–56.

40. Zernicke RF, Garhammer J, Jobe FW. Human patellar-tendon rup-ture: a kinetic analysis. J Bone Joint Surg [Am]. 1977;59A(2):179–83.

41. Zheng N, Fleisig GS, Escamilla RF, Barrentine SW. An analyticalmodel of the knee for estimation of internal forces during exercise.J Biomech. 1998;31(10):963–7.


APPLIED

SCIEN

CES

Escamilla Et Al. Cruciate Ligament Force During the Wall Squat and the One Leg Squat

Documents

Transcript of Escamilla Et Al. Cruciate Ligament Force During the Wall Squat and the One Leg Squat