A Comparison of Tibiofemoral Joint Forces
and Electromyographic Activit During
Open and Closed Kinetic Chain Exercises*
Kevin E. Wilk,† PT, Rafael F. Escamilla, PhD, Glenn S. Fleisig, PhD,
Steve W. Barrentine, MS, James R. Andrews, MD, and Melissa L. Boyd
From the American Sports Medicine Institute,
ABSTRACT
We chose to investigate tibiofemoral joint kinetics
(compressive force, anteroposterior shear force, and
extension torque) and electromyographic activity of the
quadriceps, hamstring, and gastrocnemius muscles
during open kinetic chain knee extension and closed
kinetic chain leg press and squat. Ten uninjured male
subjects performed 4 isotonic repetitions with a 12
repetition maximal weight for each exercise. Tib-
iofemoral forces were calculated using electromyo-
graphic, kinematic, and kinetic data. During the squat,
the maximal compressive force was 6139 ± 1708 N,
occurring at 91° of knee flexion; whereas the maximal
compressive force for the knee extension exercise was
4598 ± 2546 N (at 90° knee flexion). During the closed
kinetic chain exercises, a posterior shear force (poste-
rior cruciate ligament stress) occurred throughout the
range of motion, with the peak occurring from 85° to
105° of knee flexion. An anterior shear force (anterior
cruciate ligament stress) was noted during open kinetic
chain knee extension from 40° to full extension; a peak
force of 248 ± 259 N was noted at 14° of knee flexion.
Electromyographic data indicated greater hamstring
and quadriceps muscle co-contraction during the squat
compared with the other two exercises. During the leg
press, the quadriceps muscle electromyographic activ-
ity was approximately 39% to 52% of maximal velocity
isometric contraction; whereas hamstring muscle ac-
tivity was minimal (12% maximal velocity isometric
contraction). This study demonstrated significant differ-
*
Presented at the interim meeting of the AOSSM, Orlando, Florida, Feb-
ruary 1995.
t Address correspondence and reprint requests to Kevin E. Wilk, PT,
HealthSouth Rehabilitation Center, 1201 llth Avenue South Suite 100, Bir-
mingham, AL 35205.
No author or related institution has received any financial benefit from
research in this study.
518
Biomechanical Laboratory,
Birmingham, Alabama
ences in tibiofemoral forces and muscle activity be-
tween the two closed kinetic chain exercises, and be-
tween the open and closed kinetic chain exercises.
The clinical use of closed kinetic chain exercise has signif-
icantly increased during the past several years. One of the
reasons closed kinetic chain exercises have received in-
creased attention within the rehabilitation community is
that they simulate and replicate many functional move-
ments, such as squatting, stooping, and ascending or de-
scending stairs. Additionally, it has been suggested that
closed kinetic chain exercises are safer than open kinetic
chain exercises because the former minimizes anteropos-
terior tibiofemoral shear forces and thus reduces stress on
both the ACL and PCL. Closed kinetic chain exercises
have been strongly recommended as the best form of ex-
ercise for the ACL-reconstructed knee.8, 10,17, 29, 30
Terminology used for kinetic chain exercises was origi-
nally used to describe linkage analysis in mechanical en-
gineering. Steindler,34 in 1955, suggested that the human
body could be thought of as a chain consisting of rigid
overlapping segments of limbs connected by a series of
joints. He observed that when the foot or hand meets
considerable resistance, muscular recruitment and joint
motion occurred differently from that seen when the foot
or hand was free to move without restriction. Thus, closed
kinetic chain exercise occurs when the terminal or distal
segment of an appendage is fixed, such as during a squat,
leg press, or pull up.26 Conversely, open kinetic chain
exercise occurs when the terminal or distal segment is free
to move, such as during a knee extension or flexion
maneuver.26
With these concepts firmly established, numerous clini-
cians have formulated exercise drills and techniques to
use the closed kinetic chain exercise concept during reha-
bilitation of the knee. 12,17,22,28,36,39 palmitier et al.2s
have documented that during closed kinetic chain exer-

cises, the foot is fixed and motion at the knee joint is
accomplished by motion at the hip and ankle joints in a
predictable manner. Examples of such exercises include
the squat, leg press, step ups, and stair climbing simula-
tion. It has also been suggested that by performing these
specific closed kinetic chain exercises, the patient may be
able tostrengthen the quadriceps and hamstring muscles
simultaneously and also protect the reconstructed ACL.
One of the most common side effects from any type of
ACL reconstructive surgery is atrophy and weakness of
the quadriceps femoris muscle. Hence, one of the major
goals of the rehabilitation program is to strengthen this
muscle while protecting the ACL graft from potentially
deleterious stresses. Grood et a1.13 examined the forces on
the tibiofemoral joint in the ACL-deficient cadaveric knee
during simulated knee extension exercise. They docu-
mented an increase in anterior tibial displacement during
the last 30° of knee extension. Grood et a1.13 and Suntay et
al.35 suggested performing exercises in the upright posi-
tion to allow weightbearing forces to decrease the magni-
tude of tibial displacement during an isolated quadriceps
contraction. Later, Shoemaker and Markolf&dquo; determined
that the amount of anterior tibial displacement in ACL-
deficient cadaveric knees was greatly reduced by loading
the tibiofemoral joint with high compressive forces.
Henning et a1.14 studied ACL strain in vivo during spe-
cific exercises and functional movements. They placed
strain gauges in the ACLs of two subjects and measured
ACL strain during various movements. They reported
that an isometric knee extension at 22° of knee flexion
produced five to seven times more ACL strain than spe-
cific weightbearing exercises such as biking, walking,
jumping rope, or performing a half squat.
Recently, Lutz et al.21 analyzed the tibiofemoral joint
forces and electromyographic (EMG) activity of the quad-
riceps and hamstring muscles during open and closed
kinetic chain exercise. They studied five uninjured sub-
jects performing maximal isometric contractions at 30°,
60°, and 90° of flexion during open kinetic chain extension,
open kinetic chain flexion, and closed kinetic chain modi-
fied leg press. The results indicated the open kinetic chain
knee extension resulted in the greatest amount of anterior
shear force (ACL stress), with 285 ± 120 N of force occur-
ring at 30° of flexion. The closed kinetic chain exercise
produced less posterior and anterior shear forces than the
two open kinetic chain exercises. Additionally, the closed
kinetic chain exercise produced significantly greater com-
pressive forces and increased muscular co-contraction at
the same angles during which the open kinetic chain ex-
ercises produced a maximal amount of shear force and
minimal muscular co-contraction. The authors recom-
mended closed rather than open kinetic chain exercises as
a means of strengthening the quadriceps and hamstring
muscles after injury or reconstruction of the ACL.
These studies have documented the potentially delete-
rious effects of open kinetic chain and resisted knee ex-
tension exercises, and the significant differences between
shear forces and muscular recruitment of open and closed
kinetic chain exercises. However, no study to date has
evaluated the tibiofemoral joint forces and EMG activity
519
during dynamic isotonic resisted exercise movements. In
addition, no study has examined the differences in joint
forces and muscular recruitment between various forms of
closed kinetic chain exercises.
The purpose of this study was to investigate the tib-
iofemoral joint kinetics (compressive force, anteroposte-
rior shear force, and extension torque) and the EMG ac-
tivity of the quadriceps, hamstring, and gastrocnemius
muscles during three commonly performed rehabilitative
exercises: knee extension, leg press, and the squat. In this
study, we conducted three comparisons: 1) comparison of
open versus closed kinetic chain exercise, 2) comparison of
two closed kinetic chain exercises (the squat versus leg
press), and 3) comparison of the ascent and descent phases
of exercises (concentric versus eccentric muscular
contractions).
MATERIALS AND METHODS
Subjects
The left knees of 10 healthy male subjects were analyzed
for this study. The average age of the participants was 30
± 6 years (average ± SD; range, 20 to 37). The average
body mass was 93 ± 14 kg (range, 68 to 113), and the
average height was 177 ± 9 cm (range, 155 to 188). None
of the subjects had a history of knee disorders or chronic
knee pain. All subjects had normal musculoskeletal knee
examinations, with no evidence of ligamentous injury. All
subjects were experienced and highly skilled weight lift-
ers, having regularly trained in the squat, leg press, and
knee extension exercises for at least 2 years. They had an
average of 11 years of weight lifting experience. Amateur
bodybuilders and weight lifters were chosen as the sub-
jects because they provided a homogeneous sample. The
subjects’ extensive years of training enabled them to per-
form the selected exercises with consistency and proper
technique and without apprehension. The resistance loads
used by the subjects were 1) squat, mean 147 ± 39 kg
(range, 102 to 229); 2) leg press, mean 146 ± 30 kg (range,
100 to 181); and 3) knee extension, mean 78 ± 17 kg
(range, 41 to 91). Before testing, all subjects signed a
human subjects consent form in accordance with human
subjects regulations.
Instrumentation
Motion Analysis. An automated motion analysis system
(Motion Analysis, Inc., Santa Rosa, California) was used
to quantify the three-dimensional video data. Reflective
markers were attached to each subject at eight bony land-
marks : slightly posterior to the greater trochanter of each
side of the body and to the third metatarsal head, the
lateral and medial malleoli, the lateral and medial femo-
ral epicondyles, and the acromion tip on the left side.
Video images of the reflective markers were transmitted
from four 60-Hz charged coupled device cameras into a
video processor. Marker images were digitized and their
three-dimensional motion was computed using ExpertVi-
sion 3D software (Motion Analysis, Inc.).
520
External Load Measurement Devices. Two force plat-
forms (model OR6-2000, Advanced Mechanical Technol-
ogies, Inc., Newton, Massachusetts) were used to measure
forces and torques applied by each subject’s left foot. While
performing the squat, the subject placed his left foot on a
force platform mounted even with the testing room floor. A
second force platform was mounted to the foot plate of a
Body Master leg press machine (model MD-117, Body
Masters, Rayne, Louisiana). Each force platform produced
three force and three torque outputs, allowing for calcu-
lation of external loads applied to the foot. A load cell
(model LCCA-500, Omega Engineering, Stanford, Con-
necticut) was installed to directly measure force applied
by the left leg onto the resistance pad of an isotonic knee
extension machine (model 2055, Hoggan Health Indus-
tries, Draper, Utah).
Electromyography. Electromyographic activity of the
knee extensors (quadriceps muscles) and knee flexors
(hamstring and gastrocnemius muscles) were monitored
and quantified using a Noraxon Myosystem 2000 EMG
unit (Noraxon USA, Inc., Scottsdale, Arizona). The EMG
signals were collected, full-wave rectified, and integrated
from surface electrodes attached over three quadriceps
muscles (vastus medialis, vastus lateralis, and rectus fem-
oris), the lateral hamstring muscle (biceps femoris), the
medial hamstring muscle group (semitendinosus and
semimembranosus), and the gastrocnemius muscle. Elec-
tromyographic data were collected at 960 Hz, and data
from the video cameras, force plates, and load cell were
collected at 60 Hz. All data collection devices were elec-
tronically synchronized. Quadriceps and hamstring mus-
cle EMG data were used in quantifying muscle activity
and tibiofemoral kinetics, and gastrocnemius muscle
EMG data were used exclusively in quantifying tibiofemo-
ral kinetics.
Experimental Procedures
Each subject was examined and pretested approximately
1 week before the actual testing session. The subject’s
height, weight, and age were recorded. During the pretest
session, the subject was familiarized with the test proto-
col. A 12-repetition maximal effort was determined for
each exercise. Additionally, during the pretest, each sub-
ject practiced each lift to ensure consistent cadence and
range of motion.
Adequate warm-up time was allowed before beginning
the actual testing session. Then, the subject’s skin was
shaved, abraded, and cleaned to prepare the sites for
surface EMG electrode placement.2,4 Surface electrode
pairs were attached over the quadriceps, hamstring, and
gastrocnemius muscles. An additional electrode was
placed on the patella as a ground for the EMG system.
Impedance values were measured for each electrode pair
to ensure that skin impedance was minimal.
The EMG data were collected while the subject per-
formed maximal voluntary isometric contractions on the
knee extension and leg press machines. The subjects were
positioned with 90° of hip flexion and 90° of knee flexion.
This testing position was based on a pilot study in which
five subjects performed maximal voluntary isometric con-
tractions at 16 different combinations of knee and hip
flexion. The greatest mean EMG amplitude was found at
90° of knee flexion and 90° of hip flexion. After maximal
voluntary isometric contractions testing, reflective mark-
ers were attached to the subject.
Each subject performed 4 repetitions of their 12-repeti-
tion maximal resistance load for the leg press, squat, and
knee extension. The testing order for the three different
exercises was randomized for each subject. Data were
collected for only the last three repetitions of each set;
thus, the first repetition was not analyzed for any subject.
Between each repetition, subjects were instructed to
pause approximately 1 second to allow their body and
weight lifting mass to stabilize and come to a rest.
The starting position for the squat was with the subject
standing erect. In a continuous consistent motion, the
subject descended to a position of approximately 100° of
knee flexion and then ascended to the upright position.
The leg press was performed with the subject supine and
the feet securely positioned on a stationary footplate. Dur-
ing the leg press, the subject started with his knees fully
extended, then flexed his knees to approximately a 100°
position, and finally extended them to the starting (fully
extended) position. The starting position for the knee ex-
tension exercise was with the knees flexed to approxi-
mately 100° and anterior aspects of the distal tibia se-
curely pressed against the resistance pads. The subject
then extended his knees to full extension and descended
slowly to the starting (flexed) position.
Biomechanical Model. Using the three-dimensional mo-
tion analysis data, the distal direction of the leg was
calculated at each instant as a vector from the midpoint of
the femoral epicondyle markers to the midpoint of the
malleoli markers. The distal direction of the thigh was
calculated as a vector from a point 20% of the inward
distance from the left greater trochanter marker to the
right greater trochanter marker, to the midpoint of the
femoral epicondyle markers.3 Knee flexion was then de-
termined as the angle between the leg’s distal direction
and the thigh’s distal direction. Although the placement of
the hip markers slightly posterior to the greater trochan-
ters probably caused a 6° to 8° increase in measured knee
flexion, this location was used to compensate for marker
motion due to skin movement as the hips flexed.
The EMG data during the exercise trials were normal-
ized into a percentage of the maximum voluntary isomet-
ric contraction. Tibiofemoral joint kinetic data and EMG
data were then related to knee flexion angle. Each sub-
ject’s data values for all three repetitions were averaged
together as a function of knee flexion angle during the
knee-flexing and knee-extending portions of each exercise.
Shear force, compressive force, and extension torque
exerted onto the leg were calculated using a three-dimen-
sional inverse dynamics model. Shear force and compres-
sive force represented the cumulative effect on the leg
from the mass5 and inertia’ of the leg, foot, and resistive
weights, as well as the quadriceps, hamstring, and gas-
trocnemius muscles. Extension torque represented the ef-
fect of the muscles on the leg. Muscle forces were esti-

mated using EMG data and physiologic cross-sectional
area ratios.3’ The force and torque applied by each muscle
across the knee were calculated using mathematical equa-
tions that express muscle location and orientation as func-
tions of knee flexion.15,33 Compressive and shear forces
were then determined from estimated muscle forces and
the sagittal components of resultant joint force and
torque.
Statistical Analysis
Maximal compressive force,shear force, and extension
torque weredetermined for each subject during the ascent
and descent portions of each exercise. The EMG data for
each quadriceps or hamstring muscle were divided into six
intervals during the knee-flexing (descending) portion of
each exercise, based on knee flexion angle. Because each
subject exhibited approximately 90° of knee motion, each
of the six intervals represented approximately 15° of mo-
tion. Data were divided into six intervals during the knee-
extending (ascending) portion of the exercise in a similar
fashion. For each subject, data for each variable were
averaged within each interval.
A one-factor analysis of variance test was used to deter-
mine whether there were significant differences (P < 0.05)
in the data between all comparison groups. The compari-
sons conducted included 1) open kinetic chain knee exten-
sions to closed kinetic chain squat, 2) open kinetic chain to
closed kinetic chain leg press, 3) closed kinetic chain squat
to closed kinetic chain leg press, and 4) the ascent (extend-
ing) to the descent (flexing) phase of each exercise.
RESULTS
Knee Flexion Angles
In performing the squat, the subjects’ knees went through
a measured range of motion from 12° ± 7° to 104° ± 11° of
flexion. In the leg press, knee flexion started at 18° ± 7°
and flexed to 101° ± 8°. During open kinetic chain knee
extension, the range of motion of knee flexion was from
12° ± 7° to 101° ± 6°. Each interval therefore correlated to
14° ± 2° for the squat, 14° ± 1° for the leg press, and 15°
± 1° for the knee extension.
Tibiofemoral Forces
Compressive
Large magnitudes of compressive force were produced
throughout the motion of both closed kinetic chain exer-
cises (the squat and leg press) (Fig. 1). The compressive
force during the closed kinetic chain exercises was signif-
icantly greater (P < 0.05) than the compressive force
produced during the open kinetic chain knee extension
exercise. Additionally, there was a significant difference
(P < 0.05) between the compressive forces imparted dur-
ing the squat and the leg press. The compressive force
reached a maximal value of 6139 ± 1708 N during the
squat, 5762 ± 1508 N during the leg press, and 4598 ±
2546 N during the open kinetic chain knee extension ex-
ercise (Table 1). Because maximal values occurred at dif-
521
Figure 1. Graph of the average tibiofemoral compressive
forces observed during the closed kinetic chain squats (filled
triangles), leg presses (small points), and open kinetic knee
extension (open circles).
ferent times for different subjects, the average of the max-
imal values shown in the tables are greater than the
average maximal values shown at each flexion angle in
the figures. The maximal compressive forces occurred dur-
ing the knee flexing (descent) portion of the closed kinetic
chain exercises at a knee flexion angle of approximately
85° to 95° knee flexion (Fig. 1).
Tibiofemoral Shear Forces
Posterior shear force was generated during the entire
motion of each of the closed kinetic chain exercises and
from 101° to 40° of knee flexion during the open kinetic
chain knee extension. Posterior shear force represented
force displacing the tibia posteriorly relative to the femur.
On the basis of anatomic considerations, posterior shear
force is restrained by the PCL and other structures. Sig-
nificantly greater maximal posterior shear force was pro-
duced during the closed kinetic chain exercises than dur-
ing the open kinetic chain knee extension (Fig. 2).
Maximal posterior shear force of 1178 ± 594 N occurred
from 93° to 88° of knee flexion during the open kinetic
chain knee extension (Table 1). Conversely, during the
closed kinetic chain exercises, the maximal posterior
shear force occurred from 88° to 102° of knee flexion.
During the squat, the maximal posterior shear force ob-
served was 1783 ± 634 N; during the leg press, the max-
imal force was 1667 ± 556 N (Table 1). During both closed
kinetic chain exercises, slightly greater force was seen
during the knee-extending (ascending) portions of the ex-
ercises (Table 2). In contrast, significantly greater (P <
0.05) anterior tibiofemoral shear force was noted during
the knee-flexing (descending) portion of the open kinetic
chain knee extension compared with the knee-extending
(ascending) portion (Table 2).
An anterior shear force (an ACL strain) was generated
during the knee extension movement from approximately
38° to 10° of flexion. This anterior shear force can be
viewed as a force applied to the ACL and other capsular
522
TABLE 1
Maximal Tibiofemoral Joint Kinetics
a
All values reported as means ± SD of the maximum.
b a, knee extension values were significantly different (P < 0.05) from the squat and leg press values; b, squat values were significantly
different (P < 0.05) compared with the leg press values.
Figure 2. Graph of the average tibiofemoral shear force (fe-
mur on tibia) versus knee flexion angle during the knee
flexing and extending phases for closed kinetic chain squats
(filled triangles), leg presses (small points), and open kinetic
chain knee extension (open circles).
structures that prevent anterior tibial displacement rela-
tive to the femur. A maximal anterior shear force of 248 ±
259 N occurred from 20° to 11° of flexion during the knee
extending (ascending) portion of the exercise (Table 1).
There was no significant difference between the ascent
and descent portions of the open kinetic chain knee exten-
sion exercise. During the two closed kinetic chain exer-
cises, no evidence of an anterior shear force was noted at
any portion of the exercise range of motion.
Extension
Torque
During the closed kinetic chain exercises, maximal exten-
sion torque was produced with a knee flexion angle of 74°
to 102° (Table 1). The greatest extension torque occurred
during the knee-extending (ascending) portion of the open
kinetic chain knee extensions, when the knee was flexed
approximately 56° to 70° (Fig. 3).
During Exercises’
Electromyographic Activity
The EMG activity of the quadriceps and hamstring mus-
cles expressed as the percentage of the maximal
were
voluntary isometric contraction during each exercise.
Peak values for each muscle and when it occurred are
listed in Table 3, and Figures 4 and 5 graphically illus-
trate the average activity measured for all muscles. The
maximal EMG activity of the quadriceps muscle was 59%
± 10% of maximal voluntary isometric contraction and
occurred during the squat, at 88° to 102° of flexion during
the extending (ascending) phase of the exercise (Fig. 4).
During the leg press, maximal EMG activity of 51% ± 8%
maximal voluntary isometric contraction occurred at 88°
to 102° of flexion during the extending phase. In contrast,
the maximal EMG for the quadriceps muscle occurred
from 35° to 11° (52% ± 11% of maximal voluntary isomet-
ric contraction) during the knee extending phase of the
open kinetic chain knee extension exercise.
The maximal EMG activity for the hamstring muscle
group was 39% ± 12% of maximal voluntary isometric
contraction during the extending (ascending) phase of the
closed kinetic chain squat exercise (Fig. 5). Peak ham-
string muscle EMG activity occurred at 60° to 74° of knee
flexion during the squat. Conversely, the hamstring mus-
cle EMG activity during the leg press and knee extension
was significantly less (P < 0.05) compared with the squat
(Table 3).
DISCUSSION
We have attempted to study the tibiofemoral joint forces
and EMG activity of the quadriceps and hamstring mus-
cles dynamically. Knowledge of these forces is critical to
the practitioner when developing a rehabilitative program
after cruciate ligament injury or reconstruction. Addition-
ally, knowledge of quadriceps and hamstring muscle ac-
tivity during these specific exercises is vital for the reha-
bilitation specialist when devising a rehabilitative
program to restore muscle integrity. We believe there are
several significant findings of the present study worthy of
further discussion.

TABLE 2
Tibiofemoral Joint Kinetics During the Ascent and Descent Phases of Each Exercise
~ a, significant difference (P < 0.05) between ascent phase and descent phase of the squat; b, significant difference (P
ascent and descent phase of the knee extension exercise.
Figure 3. Graph of the average extension torque applied by
the femur on the tibia during the closed kinetic chain squats
(filled triangles), leg presses (small points), and open kinetic
knee extension (open circles).
The tibiofemoral compressive forces were significantly
greater during both closed kinetic chain exercises when
compared with the open kinetic chain knee extension ac-
tivity. Dynamic tibiofemoral compressive forces during
the squat and leg press appear similar, with the squat
producing greater compressive forces overall than the leg
press. The squat was performed with the subject erect,
thus weightbearing forces were applied vertically onto the
tibiofemoral joint. The maximal compressive force ob-
served during the squat (6139 ± 1708 N) was produced
with the knee flexed to 90° (Table 1). However, with the
knee near full extension and flexion, the compressive
forces approached the values previously cited by Lutz et
al.21 Conversely, during the horizontal leg press, the axial
weightbearing force was less compared with the vertical
squat. Consequently, the compressive forces were less
during the leg press than during the squat.
During the open kinetic chain knee extension, we noted
what we thought was a significant compressive force (4598
± 2546 N occurring at 90° ± 5° of knee flexion. The
compressive force noted was attributed to the significant
quadriceps muscle force exerted during resisted knee ex-
tension. Contractions of the quadriceps, hamstring, and
523
< 0.05) between
gastrocnemius muscles in excess of the minimal force
needed to prevent knee joint distraction would, in effect,
generate additional compression between the femur and
tibia. During the kinetic force calculations, if the muscle
forces were not calculated and modeled into the equations,
no compressive force would be noted during the open ki-
netic chain knee extension. Hence, during closed kinetic
chain exercises, such as the squat and leg press, both axial
loading and muscle contraction produce an increase in
joint compressive forces.
The greatest benefit of joint compressive forces in the
rehabilitative process is the facilitation of a co-contraction
among the quadriceps and hamstring muscles at the knee
joint. We believe that knee joint compressive forces in-
curred during closed kinetic chain exercises depend
greatly on the trunk position relative to the knee and
ankle joint. When the trunk was positioned posterior to
the knee joint (as during the horizontal leg press), less
tibiofemoral compressive force was produced compared
with the vertical squat (when the body was positioned
directly over the knees).
The positioning of the trunk relative to the knee and
ankle joints also significantly affects the muscular activity
of the quadriceps and hamstring muscles during closed
kinetic chain exercise. There was a gradual decrease in
compressive forces on the tibiofemoral joint near full ex-
tension and during the initiation of knee extension from
the fully flexed position during the open kinetic chain
knee extension (Fig. 1). Conversely, during the vertical
squat the compressive forces were approximately 2650 N
at 30° of flexion, a common angle used during rehabilita-
tion immediately after ACL reconstruction surgery.
Comparisons of tibiofemoral shear forces indicated sig-
nificantly less shear force during open kinetic chain knee
extensions when compared with both closed kinetic chain
exercises. During both closed kinetic chain exercises a
posterior shear force (PCL stress) was observed through-
out the arc of motion, with the greatest amount of poste-
rior shear force occurring from 83° to 105° of knee flexion.
These findings agree with those of Andrews et al.,l who
reported maximal shear force at the lowest position of the
vertical squat. Greater posterior shear force was gener-
ated during the vertical squat exercise than during either
the leg press or knee extension. This indicates that during
524
TABLE 3
Maximal EMG Activity of the Quadriceps and the Hamstring
Muscles During Closed and Open Knee Chain Exercises&dquo;
’
All values reported as a percentage of the maximal voluntary
isometric contraction (mean ± SD).
b
Squat values were significantly different (P < 0.05) from knee
press and knee extension values.
Figure 4. Graph of the average quadriceps femoris muscle
EMG activity (percentage of maximal voluntary isometric
contraction) compared with knee flexion angle during the
knee flexing and extending phases of closed kinetic chain
squats (filled triangles), leg presses (small points), and open
kinetic knee extension (open circles).
both the leg press and the squat movement loads are being
applied to the PCL.
Our findings agree with those of Lutz et al.,21 who also
reported a posteriorly directed shear force during closed
kinetic chain exercise. These findings suggest that during
normal daily activities, such as squatting, stooping to pick
up an object, or ascending or descending stairs, there is a
posteriorly directed shear force and thus strain on the
PCL. When a person ascends or descends stairs, approxi-
mately 83° to 90° of knee flexion is necessary 20; when a
person stoops to lift an object, at least 117° of knee flexion
is necessary.24 These angles should be considered when
designing a PCL rehabilitation program because certain
closed kinetic chain exercises at varying knee flexion an-
gles may deleteriously increase PCL stress.38
Figure 5. The average of hamstring muscle EMG activity
(percentage of maximal voluntary isometric contraction)
compared with knee flexion angle during the knee flexing and
extending phases of closed kinetic chain squats (filled trian-
gles), leg presses (small points), and open kinetic knee ex-
tension (open circles).
As stated previously, we noted a PCL load throughout
the range of motion during both closed kinetic chain ex-
ercises, and no ACL load was noted. It is also noteworthy
that during the first 45° of the squat (0° to 45°) the pos-
terior shear force was relatively low (ranging from 245 to
565 N). The rapid increase in posterior shear force was
recorded from 45° to 72° of knee flexion. Vertical squats in
a range from 0° to 45° of knee flexion may be safer for the
knee in early rehabilitation after PCL reconstruction.
Recently, Fleming et al.9 reported the strain measure-
ments of the anteromedial bundle of the ACL in vivo
during the squat. They observed strain within the ACL
from approximately 45° to 0° of extension, with the max-
imal strain of 4% occurring at approximately 10° of flex-
ion. These conclusions conflict with our findings and those
of Lutz et al., 21 who noted no anterior shear force during
closed kinetic chain exercises. The differences may be due
to the method of calculations and collection of the data.
Further research is needed to resolve this question.
In this study, we have noted that during open kinetic
chain knee extensions there is a posterior shear force from
60° to 101° of flexion. This is extremely important for the
rehabilitative specialist to realize when implementing a
program for the PCL-injured knee. During resisted knee
extension, we noted an anterior shear force from approx-
imately 38° of flexion to full extension. We noted the
greatest amount of anterior shear force at approximately
14° ± 2°. These findings agree with those of other inves-
tigators who have reported the greatest amount of tibial
displacement occurring within 30° to 10° of full exten-
sion. 13,16,19,27,32,35,40 Numerous investigators have docu-
mented anterior shear force during resisted knee exten-
sion from approximately 60° to 0° of knee
flexion. 13,16,18,19,21,27,32,40,42 Therefore, an ACL stress
may only occur during open kinetic chain knee extension
from 60° to 0° extension, and not at all during closed
kinetic chain exercise. More research is needed to inves-
tigate this further.

The shear force significantly increases in the last 30° of
extension and is proportional to the load applied, location
of resistance, and speed of movement.4° Wilk and An-
drews,4° in a study conducted on ACL-deficient knees,
reported the greatest amount of anterior tibial displace-
ment occurred at 30° to 15° of flexion, with slower isoki-
netics producing greater tibial displacement and proximal
resistance reducing tibial displacement. As stated previ-
ously, we have noted less posterior shear force during
open kinetic chain knee extensions when compared with
both closed kinetic chain exercises. At first, this may ap-
pear surprising to the clinician; however, it agrees with
other biomechanical studies analyzing closed kinetic
chain and open kinetic chain exercises. 21,26 Lutz et al. 21
reported a maximal posterior shear force of 538 ± 476 N
during the closed kinetic chain squat compared with 387
± 67 N during the open kinetic chain knee extension at
90° of flexion. In addition, they noted greater compressive
forces (3765 ± 1313 N) during knee extension compared
with the closed kinetic chain exercise (3453 ± 1313 N).
Our results indicated greater compressive forces during
closed kinetic chain compared with open kinetic chain
exercises. Thus, we believe it is important for the reader to
realize that closed kinetic chain exercises do not automat-
ically reduce shear force; rather, the reduced tibial dis-
placement noted during closed kinetic chain exercise is
due to the combined effects of the joint’s geometry, com-
pressive forces, weightbearing, and muscular forces.
We observed an anterior shear force of 248 ± 259 N at
14° of knee flexion during open kinetic chain knee exten-
sion. Lutz et a1.21reported a maximal anterior shear force
of 285 ± 120 N at 30° of knee flexion during open kinetic
chain knee extension. Our results are consistent with
those of Lutz et al. 21 Other authors have reported the
ultimate strength of the ACL before failure is 1730 N.4° 25
Hence, the question must be asked, is open kinetic chain
knee extension deleterious to a mature ACL graft with
suitable tissue graft strength in the terminal knee exten-
sion range (30° to 0°)? Perhaps this risk occurs only during
the early phases of the rehabilitation, when graft strength
is diminishing and fixation is a concern.
The extension torque values closely replicate the EMG
activity of the quadriceps muscle during all three exer-
cises. The maximal extension torque was noted during the
knee extension exercise and occurred near midrange, be-
tween 56° to 70° of knee flexion. Additionally, greater
extension torque was noted during the ascending (extend-
ing) or concentric phase of the movement when compared
with the descending or eccentric phase.
In the present study, analysis of the integrated EMG
activity revealed that co-contraction of the quadriceps and
hamstring muscles occurred during the first 30° of the
squat. During the other two exercises (the leg press and
knee extension), there was minimal co-contraction. Dur-
ing the extending phase of the squat, peak EMG ham-
string muscle activity occurred from 70° to 40° of knee
flexion. This coincides with the functional role the ham-
string muscle performs through eccentric control of the
hip and knee flexion angle during the ascending (extend-
ing) phase of the squat. During the first 20° to 25° of the
525
vertical squat, the hamstring muscle activity was greater
than that of the quadriceps musculature. These findings
agree with those of Lutz et al., who reported a quadriceps-
to-hamstring ratio of 1:1.3 during the early phase of their
closed kinetic chain exercise. Graham et al.11 reported a
hamstring-to-quadriceps EMG ratio of 62% during a quar-
ter squat. These findings suggest that the clinician may
prefer to delay the use of closed kinetic chain squats from
0° to 45° of knee flexion in patients where hamstring
muscle activity must be minimized, such as during the
early stages of rehabilitation after PCL reconstructions
and meniscal repairs. Additionally, it appears that closed
kinetic chain squats promote a co-contraction whereas a
closed kinetic chain leg press produces a quadriceps mus-
cle dominant contraction.
Conversely, peak EMG activity of the quadriceps mus-
cle occurred during the squat from 88° to 102° of knee
flexion. High levels of quadriceps muscle EMG activity
were recorded during the leg press (from 88° to 102° of
flexion) and knee extension (from 41° to 11° of flexion).41
During both the vertical squat and leg press we noted a
sharp rise in quadriceps muscle EMG activity from 42° to
102° of knee flexion (Fig. 4). However, during the leg press
and knee extension exercises, hamstring muscle EMG
activity was minimal (approximately 10% maximal volun-
tary isometric contractions). During the leg press from 70°
to 105° of knee flexion, the quadriceps-to-hamstring mus-
cle activity ratio was approximately 4:1; during knee ex-
tension from 40° to 0° of flexion, the quadriceps-to-ham-
string ratio was 5:1. Several other investigators have also
noted a similarly high ratio of quadriceps-to-hamstring
EMG activity during knee extensions.ll° 21 Hence, it would
appear that two excellent exercises for isolated quadriceps
muscle activity are the leg press and knee extension,
where the hamstring muscular activity is minimal.
For all three exercises studied, quadriceps muscle activ-
ity was significantly greater during the extending (concen-
tric) phase of the movement. Quadriceps muscle peak
activity during closed kinetic chain exercises appears
nearly the opposite of that demonstrated in the open ki-
netic chain exercise. During the closed kinetic chain exer-
cises, the higher the knee flexion angle, the greater the
quadriceps muscle activity; conversely, during the knee
extension exercise, maximal quadriceps muscle EMG ac-
tivity was greater near full extension (Fig. 4). During the
open kinetic chain knee extension the quadriceps muscle
EMG activity gradually increased as the knee extended,
the sharpest rise in muscle activity was recorded from 40°
to full extension.
A possible reason for the increase seen in the quadri-
ceps-to-hamstrings muscle activity ratio during the leg
press compared with the squat is related to body position
relative to the knee joint. During the horizontal leg press,
the body is positioned behind the knee joint (especially as
knee flexion increases). With this body position, the quad-
riceps musculature must contract to control the increasing
knee flexion angle. Conversely, during the vertical squat,
the body is positioned only slightly posterior to the knee
joint (i.e., smaller perpendicular distance from knee joint
center to time of action of resultant force), resulting in
526
more of a co-contraction between the quadriceps and ham-
string muscles. McLaughlin et al. 23 reported a decrease in
quadriceps muscle activity with an increase in trunk flex-
ion angle. They also noted that skilled lifters perform the
squat with the trunk in a more erect position compared
with less skilled lifters. Therefore, based on these results,
we believe there are significant differences in joint forces
and EMG activity during the closed kinetic chain squat
and leg press.
Clinically, patients who flex their hips excessively and,
in effect, lean forward onto their hands while performing
the vertical squat, increasing trunk flexion, are reducing
the demands placed on the quadriceps muscle by enhanc-
ing the hamstring muscle effect.
There appear to be significant differences in both the
joint forces and muscle activities during the ascending
and descending phases of the closed kinetic chain exer-
cises. Dahlkvist et a1.6 reported significant differences in
joint forces during the ascent and descent phases of spe-
cific exercises. There were greater compressive and ante-
rior shear forces during the extending phase for all three
exercises studied here. Quadriceps muscle activity was
also greater during the extending (ascending, concentric)
phase for all three exercises. We noted low levels of ham-
string muscle EMG activity during all three exercises,
except for the extending (ascending) phase of the vertical
squat.
Unlike previous biomechanical investigations, this
study used a three-dimensional dynamic analysis of knee
biomechanics during exercise. In addition to information
about concentric and eccentric muscle function, this study
presented total joint kinetics for the knee. There are cer-
tain advantages and disadvantages to presenting the data
as total joint kinetics. By using precise motion analysis
and load measurement devices, total joint kinetics could
be calculated with a great degree of confidence. However,
because the data were not separated into different com-
ponents (e.g., bones, muscles, ligaments) it is somewhat
difficult to verify the magnitude of stress applied to the
ACL and PCL. Nevertheless, although models currently
used to separate kinetics into components introduce a
number of assumptions and estimates, such modeling may
be significant in furthering our understanding of the cru-
ciate ligaments.
Another area for future study involves quantifying the
effect that varying exercise technique has on muscle ac-
tivity and joint kinetics. All subjects used in this study
were well trained in the exercises performed. Although
this may have enabled us to ensure exercise movement
consistency and to quantify proper exercise technique, it
may not provide an optimal description for the typical
rehabilitation patient. The average weight used by the
subjects in this study to perform the squat was 146 kg
(range, 102 to 203) and 79 kg (range, 41 to 91) during the
knee extension. Most rehabilitation patients never use
this much weight. The excessive amount of resistance
used by the subjects in this study may contribute to the
significantly high magnitude of compressive and shear
force noted in our study when compared with other
studies.21
The magnitude of the joint forces noted was directly
related to the amount of resistance applied. Thus, the
reader should not interpret the study solely on the abso-
lute joint forces but rather the direction of the shear
forces, changes in forces, and the recruitment of muscles.
Additionally, this information is germane to clinicians
involved with high school, college, and professional
athletes.
The closed kinetic chain leg press used in this study was
the Body Master model. This system uses a 15° angle and
is not completely horizontal. Thus, muscle activity and
joint kinetics may be different with model design
differences.
Before interpreting the kinetic results, however, it is
important to consider limitations in the methods. Muscle
force was assumed to be proportional to EMG activity. It
was also assumed that the force per cross-sectional area
generated was constant for all muscles. Muscle location
and orientation were based on a limited number of small-
sample studies. Finally, ligaments were assumed to con-
tribute shear force only, whereas bones were assumed to
contribute compressive force only. These approximations
are common in biomechanics; nevertheless, it is important
to realize that the forces calculated represent estimations
of actual joint dynamics.
SUMMARY OF RESULTS
Based on the results of this study, the authors conclude
the following:
1. Isotonic closed kinetic chain exercise produced signif-
icantly greater compressive forces than the open kinetic
chain knee extension.
2. During closed kinetic chain exercises, a posterior
shear force (PCL stress) was generated throughout the
entire range, with maximal shear occurring from 85° to
105° of flexion. During knee extension there is an anterior
shear force (ACL stress) from 38° to 0° and a posterior
shear force from 40° to 101°.
3. There was greater posterior shear force produced
during the closed kinetic chain exercises compared with
open kinetic chain exercise.
4. Not all closed kinetic chain exercises produce a co-
contraction of the hamstring and quadriceps muscles, the
magnitude of the co-contraction activity depended on
trunk position relative to the knee joint and the applica-
tion of force and knee flexion angle.
5. There are significant differences in joint forces and
EMG activity during the ascending and descending
phases of both open kinetic chain and closed kinetic chain
exercises.
These results are clinically relevant in that they provide
the scientific rationale for the judicious use of both closed
and open kinetic chain exercises when developing a reha-
bilitation program for the patient with an injured knee.
Additionally, this study documents that not all closed
kinetic chain exercises produce the same muscle recruit-
ment pattern and tibiofemoral joint forces, and with spe-
cific circumstances (e.g., PCL injury or rehabilitation) may
be deleterious.

ACKNOWLEDGMENTS
The authors thank Body Masters of Rayne, Louisiana, and
Hoggan Health Industries of Draper, Utah, for their do-
nation of exercise equipment. Additionally, the authors
thank Ting Ma and Swapna Singireddy for their computer
programming; Phillip Sutton, Andy DeMonia, Larry Strip-
lin, and Tomoyuki Matsuo for their help in data collection
and analysis; and Robin Lonkert-Scott for her assistance
in manuscript preparation.
REFERENCES
1. Andrews JG, Nay JG, Vaughan CL: Knee shear forces during a squat
exercise using a barbell and a weight machine, in Matsui H, Kobayashi K
(eds): Biomechanics VIII-B. Champaign, IL, Human Kinetics Publishers,
1983, pp 923-927
2. Basmajian JV, Blumenstein R: Electrode Placement in EMG Biofeedback.
Baltimore, Williams and Wilkins, 1980
3. Brinckmann P, Hoefert H, Jongen HT: Sex differences in the skeletal
geometry of the human pelvis and hip joint. J Biomech 14: 427-430, 1981
4. Butler DL, Grood ES, Noyes FR, et al: On the interpretation of our anterior
cruciate ligament data. Clin Orthop 196:
26-34, 1985
5. Clauser CE, McConville JT, Young JW: Weight, volume, and center of
mass of segments of the human body. Aerospace Medical Research
Laboratory, AMRL-TR-69-70 Wright-Patterson Air Force Base, Ohio,
1969
6. Dahlkvist NJ, Mayo P, Seedhom BB: Forces during squatting and rising
from a deep squat. Eng Med 11: 69-76, 1982
7. Dempster WT: Space requirements of the seated operator. Wright Air
Development Center, WADC-TR-55-159 Wright-Patterson Air Force
Base, Ohio, 1955
8. DePalma BF, Zelko RR: Knee rehabilitation following anterior cruciate
ligament injury or surgery. Athl Train 21: 200-204, 1986
9. Fleming B, Beynnon B, Peura G, et al: Anterior cruciate ligament strain
during an open and a closed kinetic chain exercise: An in vivo study. Trans
Orthop Res Soc 20: 631, 1995
10. Giove TP, Miller SJ III, Kent BE: Non-operative treatment of the torn
anterior cruciate ligament. J Bone Joint Surg 65A: 184-192, 1983
11. Graham VL, Gehlsen GM, Edwards JA: Electromyographic evaluation of
closed and open kinetic chain knee rehabilitation exercises. J Athl Train
23-30,1993
28:
12. Gray G: Rehabilitation of running injuries: Biomechanical and propriocep-
tive considerations, in Topics in Acute Care and Trauma Rehabilitation.
Rockville, MD, Aspen Publishers, 1986
13. Grood ES, Suntay WJ, Noyes FR, et al: Biomechanics of the knee
extension exercise. Effect of cutting the anterior cruciate ligament. J Bone
Joint Surg 66A: 725-734, 1984
14. Henning CE, Lynch MA, Glick KR: An in vivo strain gage study of elon-
gation of the anterior cruciate ligament. Am J Sports Med 13: 22-26, 1985
15. Herzog W, Read LJ: Lines of action and moment arms of the major
force-carrying structure crossing the human knee joint. J Anat 182 (pt 2):
213-230, 1993
16. Hirokawa S, Solomonow M, Lu Y, et al: Anterior-posterior and rotational
displacement of the tibia elicited by quadriceps contraction. Am J Sports
Med 20: 299-306, 1992
527
17. Irrgang JJ, Rivera J: Closed kinetic chain exercises for the lower extrem-
ity : Theory and application, in Sports Physical Therapy Section Home
Study Course. Volume 1, Issue 4. LaCrosse, WI, Sports Physical Therapy
Section, 1994
18. Jurist KA, Otis JC: Anteroposterior tibiofemoral displacements during
isometric extension efforts. Am J Sports Med 13: 254-258, 1985
19. Kaufman KR, An KN, Litchy WJ, et al: Dynamic joint forces during knee
isokinetic exercise. Am J Sports Med 19: 305-316, 1991
20. Laubenthal KN, Smidt GL, Kettelkamp DB: A quantitative analysis of knee
motion during activities of daily living. Phys Ther 52: 34-42, 1972
21. Lutz GS, Palmitier RA, An KN, et al: Comparison of tibiofemoral joint
forces during open-kinetic-chain and closed-kinetic-chain exercises.
J Bone Joint Surg 75A: 732-739, 1993
22. Mangine RE, Noyes FR: Rehabilitation of the allograft reconstruction.
J Orthop Sports Phys Ther 15: 294-302, 1992
23. McLaughlin TM, Lardner TJ, Dillman CJ: Kinetics of the parallel squat.
Res Q 49:
175-189, 1978
24. Nordin M, Frankel VH: Biomechanics of the knee, in Frankel VH, Nordin M
(eds): Basic Biomechanics of the Skeletal System. Philadelphia, Lea &
Febiger, 1980, pp 117-118
25. Noyes FR, Butler DL, Grood ES, et al: Biomechanical analysis of human
ligament grafts used in knee ligament repairs and reconstructions. J Bone
Joint Surg 66A: 344-352, 1984
26. Palmitier RA, An KN, Scott SG, et al: Kinetic chain exercise in knee
rehabilitation. Sports Med 11: 402-413, 1991
27. Renstr&ouml;m P, Arms SW, Stanwyck TS, et al: Strain within the anterior
cruciate ligament during hamstring and quadriceps activity. Am J Sports
Med 14: 83-87, 1986
28. Rivera JE: Open versus closed kinetic chain rehabilitation of the lower
extremity: A functional and biomechanical analysis. J Sports Rehab 3:
154-167, 1994
29. Seto JL, Brewster CE, Lombardo SJ, et al: Rehabilitation of the knee after
anterior cruciate ligament reconstruction. J Orthop Sports Phys Ther 11:
8-18, 1989
30. Shelbourne KD, Nitz P: Accelerated rehabilitation after anterior cruciate
ligament reconstruction. Am J Sports Med 18: 292-299, 1990
31. Shoemaker SC, Markolf KL: Effects of joint load on the stiffness and laxity
of ligament-deficient knees. J Bone Joint Surg 67A: 136-146, 1985
32. Smidt GL: Biomechanical analysis of knee flexion and extension. J Bio -
mech 6: 79-92, 1973
33. Spoor CW, Van Leeuwen JL: Knee muscle moment arms from MRI and
from tendon travel. J Biomech 25: 201-206, 1992
34. Steindler A: Kinesiology of the Human Body Under Normal and Patholog-
ical Conditions. Springfield, IL, Charles C Thomas, 1955
35. Suntay WJ, Grood ES, Butler DL, et al: Biomechanics of knee extension
exercise. Trans Orthop Res Soc 6: 33-38, 1981
36. Voight ML, Tippett SR: Plyometric exercises of the lower extremity, in
Sports Physical Therapy Section Home Study Course. Volume 1, Issue 3.
LaCrosse, WI, Sports Physical Therapy Section, 1994
37. Wickiewicz TL, Roy RR, Powell PL, et al: Muscle architecture of the
human lower limb. Clin Orthop 179: 275-283, 1983
38. Wilk KE: Rehabilitation of isolated and combined posterior cruciate liga-
ment injuries. Clin Sports Med 13: 649-677, 1994
39. Wilk KE: Current concepts in the treatment of anterior cruciate ligament
disruption. J Orthop Sports Phys Ther 15: 279-293, 1992
40. Wilk KE, Andrews JR: The effects of pad placement and angular velocity
on tibial displacement during isokinetic exercise. J Orthop Sports Phys
Ther 17:24-30, 1993
41. Wilk KE, Simoneau G, McGraw J: The electromyographic activity of the
quadriceps femoris vastus medialis/lateralis ratio during squats, leg press,
and knee extension exercises. Phys Ther 73: 580, 1993
42. Yasuda K, Sasaki T: Exercise after anterior cruciate ligament reconstruc-
tion. The force exerted on the tibia by the separate isometric contractions
of the quadriceps or the hamstrings. Clin Orthop 220: 275-283, 1987