ca
c
, A
Department of Biomedical Engineering (ND-20), The Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195, USA
reconstruction, at a cost of almost one billion dollars ate incidence of non-contact ACL injuries based on gen-
der, with females reported to suffer these injuries 5–7
times more frequently than males (Arendt and Dick,
1995). Despite the vast amount of ongoing research into
ACL injuries, the precise mechanisms of non-contact
.
*
Corresponding author.
E-mail address: mcleans@bme.ri.ccf.org (S.G. McLean).
Clinical Biomechanics 19 (20268-0033/$ - see front matter
2004 Elsevier Ltd. All rights reservedReceived 6 February 2004; accepted 9 June 2004
Abstract
Background. Knee joint sagittal plane forces are a proposed mechanism of anterior cruciate ligament injury during sport move-
ments such as sidestep cutting. Ligament force magnitudes for these movements however, remain unknown. The need to examine
injury-causing events suggests elucidation via model-based investigations is possible. Using this approach, the current study deter-
mined whether sagittal plane knee loading during sidestep cutting could in isolation injure the anterior cruciate ligament.
Methods. Experiments were performed on subject-specific forward dynamic musculoskeletal models, generated from data
obtained from 10 male and 10 female athletes. Models were optimized to simulate subject-specific cutting movements. Random pert-
urbations (n=5000) were applied to initial contact conditions and quadriceps/hamstrings activation levels to simulate their effect on
peak 3D knee loads. Injury via the sagittal plane mechanism was based on the criterion of an anterior drawer force greater than
2000 N.
Findings. Realistic neuromuscular perturbations produced significant increases in external knee anterior force and valgus and
internal rotation moments. Peak anterior drawer force never exceeded 2000 N in any model, and thus failed to cause anterior cru-
ciate ligament injuries. Valgus loads reached values that were high enough to rupture the ligament, occurring more frequently in
females than in males.
Interpretation. Sagittal plane knee joint forces cannot rupture the anterior cruciate ligament during sidestep cutting. The inter-
action between muscle and joint mechanics and external ground reaction forces in this plane, places a ceiling on ligament loads.
Valgus loading is a more likely injury mechanism, especially in females. Modifying sagittal plane biomechanics will thus unlikely
contribute to the prevention of anterior cruciate ligament injuries.

2004 Elsevier Ltd. All rights reserved.
Keywords: Anterior cruciate ligament; Forward dynamic simulation; Gender; Sidestep cutting; Knee joint loading; Knee valgus; Neuromuscular
control; Monte Carlo simulations
1. Introduction
Anterior cruciate ligament (ACL) injury is a common
and potentially disabling sports related injury. Approxi-
mately 80,000 ACL injuries occur annually within the
United States, with roughly 50,000 requiring surgical
(Daniel and Fritschy, 1994). Approximately 70% of
ACL injuries occur as a result of a non-contact episode,
typically during the execution of movements characteri-
zed by a sudden deceleration or direction change, such
as sidestep cutting (Arendt and Dick, 1995; Griffin
et al., 2000). Of particular concern, is the disproportion-Sagittal plane biomechanics
sidestep
Scott G. McLean
*
, Xuemei Huangdoi:10.1016/j.clinbiomech.2004.06.006nnot injure the ACL during
utting
nne Su, Antonie J. van den Bogert
www.elsevier.com/locate/clinbiomech
004) 828–838

first collected with all joints in the neutral position, fol-
lowing which, the forehead, left and right anterior supe-
BiomACL injury, and the extent to which they may be gender
specific, remain unclear. Theories continue to evolve
alongside this research however, as to the most likely
contributors to ACL injury risk.
Sagittal plane mechanisms for non-contact ACL in-
jury have been proposed previously for sports move-
ments (Chappell et al., 2002; DeMorat et al., 2004;
Griffin et al., 2000). Such postulates are based on the
fact that the landing phase of these movements typically
incorporates large quadriceps force at relatively small
flexion angles, a combination known to induce anterior
force on the tibia (Durselen et al., 1995; Pandy and Shel-
bourne, 1997). Women are often observed to perform
these movements with less knee flexion than males
(Chappell et al., 2002; Malinzak et al., 2001), which is
thus viewed as a likely contributor to their increased risk
of ACL injury (Colby et al., 2000; Griffin et al., 2000;
Lephart et al., 2002). The neuromuscular control and
strength ratio of the hamstrings and quadriceps are also
viewed as important components of a sagittal plane in-
jury mechanism (Colby et al., 2000; Griffin et al.,
2000). Both of these variables have similarly been found
to differ across gender (Wojtys et al., 2003).
Another important component of the sagittal plane
loading mechanism during execution of sports move-
ments is the presence of a large ground reaction force
(GRF), which is directed posteriorly with respect to
the tibial axis (McLean et al., 2004a). This force would
help protect the ACL during the landing phase of these
movements, but has not been taken into account in cur-
rent theories on sagittal plane contributions to ACL in-
jury. Thus, the potential for sagittal plane biomechanics
to induce ACL injury may be overestimated.
If the sagittal plane biomechanics associated with
sporting postures can produce an ACL injury, then pre-
vention strategies could focus on teaching women to
perform movements with more knee flexion, and more
hamstrings activation. However, the true potential for
ACL injury via this mechanism remains unclear, as lig-
ament forces have not been measured or estimated dur-
ing an injury-causing event. Furthermore, the need to
examine the knee joint loading response to controlled
systematic movement variations, or to evaluate injury
scenarios, makes elucidation via human experimenta-
tion unfeasible. The recent development and validation
of subject-specific forward dynamic simulations of
sporting postures such as sidestep cutting, has made it
possible to predict the effect of perturbations in neuro-
muscular control on resultant knee movement and load-
ing (McLean et al., 2003). Models of this type provide a
fast and relatively inexpensive means to study acute
knee joint injuries while controlling all aspects of neuro-
muscular control (NMC). Using such an approach, the
current study determined the effects of random varia-
tions in NMC during the stance phase of sidestep cut-
S.G. McLean et al. / Clinicalting on 3D knee loading. From these data, therior iliac spine (ASIS), medial femoral condyle and
medial and lateral malleoli markers were removed prior
to the motion trials. Synchronized 3D GRF data were
collected during each trial at 1000 Hz via an AMTI force
plate (Model OR6-5, Serial # 4068, Watertown, MA,potential for the sagittal plane loading mechanism, com-
prising quadriceps and hamstring forces, flexion angle,
and external anterior–posterior joint loads, to produce
ACL injuries during sidestep cutting was evaluated
and compared across gender.
2. Methods
Twenty subject-specific forward dynamics models of
the stance phase (0–200 ms) of a sidestep cut were gen-
erated for the current study. Subject data implemented
within each model were obtained from 10 male and 10
female NCAA Division 1 basketball players, whom were
matched for experience level (Table 1). Prior to experi-
mentation, approval for the research was gained
through the Institutional Review Board of the Cleveland
Clinic Foundation and written informed consent for all
subjects was obtained. Subject inclusion in the study was
based on no history of operable lower limb joint injury.
A summary of subject characteristics is presented in
Table 1.
2.1. Data collection
Three-dimensional (3D) kinematic and GRF data
were recorded for each subject across 10 sidestep cut-
ting trials. Approach speeds were monitored and re-
quired to fall between 4.5 and 5.5 ms

1
, reflecting
speeds at which these movements are typically executed
in the game situation (McLean et al., 1999). Sidestep
cutting angles were required to be 35–55 from the
original direction of motion, again reflecting values typ-
ically demonstrated in the game situation, and adopted
previously (McLean et al., 2004a,b). Angles were meas-
ured from the center of the force plate and the corre-
sponding line was marked (using tape) so that it
could be clearly seen by the subjects (Fig. 1). Specifi-
cally, subjects were required to land and sidestep cut
off the right leg, such that that the cutting action moved
the subject forward and to the left of the force plate at
the appropriate angle (McLean et al., 1999, 2004a,b)
(see Fig. 1). Kinematic data were obtained from the
3D coordinates of skin-mounted markers secured to
various anatomical locations (Fig. 2), recorded via six
electronically shuttered high-speed video cameras at
240 fps and Eva 6.0 tracking software (Motion Analysis
Corp., Santa Rosa, CA, USA). A standing trial was
echanics 19 (2004) 828–838 829USA).