Clinical Biomechanics 24 (2009) 833–841

Contents lists available at ScienceDirect

Clinical Biomechanics
journal homepage: www.elsevier.com/locate/clinbiomech

Knee joint kinematics, kinetics and muscle co-contraction in knee

osteoarthritis patient gait
Tamika L. Heiden *, David G. Lloyd **, Timothy R. Ackland
School of Sport Science, Exercise and Health, The University of Western Australia, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Background: Compared to matched controls, knee osteoarthritis patients walk with altered, kinematics,
Received 10 December 2008 kinetics and muscle activity. Studies of osteoarthritis patient gait have focused on individual measures,
Accepted 9 August 2009 and findings from these studies differ due to differences in patient levels of disability and age. Therefore,
aims of this study were to examine kinematic, kinetic and muscle co-contraction gait variables within a
single osteoarthritis patient group, and to determine if alterations in these variables are related to pain,
Keywords: symptom and function measures.
Co-contraction
Methods: Thirty asymptomatic controls and 54 patients with radiographic evidence of knee osteoarthritis
Gait mechanics
Knee joint moments
participated. Self-perceived measures of pain and symptoms, and gait (knee joint angles, moments and
Knee extension strength muscle co-contraction) were analysed and compared.
Findings: Osteoarthritis patients had greater self-perceived pain and symptoms on the questionnaires.
Gait differences in the knee osteoarthritis patients were greater knee flexion at heel strike and during
early stance along with reductions in the peak external knee extension moment in late stance. Co-con-
traction ratios highlighted greater lateral muscle activation in osteoarthritis patients, which were corre-
lated with the magnitude of their adduction moments. Larger adduction moments were related to lower
self-perceived pain and symptoms.
Interpretation: Osteoarthritis patients use predominantly lateral muscle activation during stance which
may aid in stabilising the external knee adduction moment. Kinematic alterations in knee osteoarthritis
patient gait occur without alterations in knee joint moments. Our results also suggest that adduction
moments are lowered to reduce the patients’ pain and symptoms.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction
Osteoarthritis (OA) is the most common rheumatic disease and
the knee is the most often affected weight bearing joint (Martin,
1994). Knee OA causes low quality of life and functioning in activ-
ities of daily living, with increased pain, decreased muscle mass,
proprioception deficits, and altered gait mechanics. In regard to
gait, compared to matched controls, knee OA patients have reduc-
tions in walking speed (Astephen and Deluzio, 2005; Brinkmann
and Perry, 1985) and cadence (Chen et al., 2003; Stauffer et al.,
1977), longer double support time (Chen et al., 2003; Smith
et al., 2004), a smaller stride length (Baliunas et al., 2002), in-
creased knee flexion at heel strike (Childs et al., 2004; Munder-
mann et al., 2005), and reduced knee flexion during the stance
* Corresponding author. Address: School of Sports Science, Exercise and Health,
The University of Western Australia, 35 Stirling Highway, Crawley, Western
Australia 6009, Australia.
**Corresponding author.
E-mail addresses: tamikaheiden@hotmail.com (T.L. Heiden), David.Lloyd@uwa.
edu.au (D.G. Lloyd).
phase of gait (Astephen and Deluzio, 2005; Baliunas et al., 2002).
They also walk with reduced heel strike forces on the affected limb
(Stauffer et al., 1977), lower external knee flexion moments during
early stance (Kaufman et al., 2001), lower external knee extension
moments in terminal stance (Smith et al., 2004), and increased
external knee adduction moments in stance (Astephen and Delu-
zio, 2005). This is accompanied by increased hamstring activation
(Hortobagyi et al., 2005), prolonged muscle activation in stance
(Childs et al., 2004), and increased co-contraction (Schmitt and Ru-
dolph, 2007).
Wide variability exists in the magnitude of impairment in walk-
ing speed, knee kinematics, and knee kinetics between OA and con-
trol groups in the aforementioned studies. This variability may be
attributed to differences between the studies in age, knee align-
ment, disease severity, or walking speed. With increasing age
comes altered neuromuscular function and reductions in walking
speed (Rudolph et al., 2007). Reductions in gait speed are a func-
tion of both age and the disease. Healthy subjects walk at faster
preferred gait speeds than those with knee OA (Hanlon and Ander-
son, 2006) and many gait variables change with walking speed
(Andriacchi et al., 1977). Specifically, ground reaction forces

0268-0033/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.clinbiomech.2009.08.005
834 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841
(Andriacchi et al., 1977), peak knee flexion and sagittal plane knee
joint moments (Lelas et al., 2003) have been shown to rise with
increasing gait speed, and gait speed is a reflection of a patient’s
clinical state (Andriacchi et al., 1977; Brinkmann and Perry,
1985). With increasing OA disease severity there are alterations
in the external knee adduction moment (Mundermann et al.,
2004), knee flexion angle, and gastrocnemius muscle activity
(Astephen et al., 2007) during gait. Given these differences, the
selection of similarly aged OA patients and controls, and the con-
trol of confounding factors should be considered when examining
gait alterations in knee OA patients.
Inclusion of patient self-perceived pain, symptoms and function
are important due to their direct links to functioning in activities of
daily living (Maly et al., 2006). The relationship between structural
changes and symptoms in knee OA patients is small (Dieppe,
2004). However, the development of increased pain is associated
with increased external adduction moments in the gait of knee
OA patients (Miyazaki et al., 2002) and these adduction moments
in turn are associated with increased radiographic severity (Shar-
ma et al., 1998). Pain and symptoms may also be related to muscle
activation patterns although this is yet to be tested.
There are only a few studies of knee OA patients compared with
asymptomatic, similarly aged controls, that have comprehensively
examined knee joint kinematics, kinetics and muscle activation,
whilst controlling walking speed and age of comparison group
(Lewek et al., 2004, 2006; Messier et al., 2005; Schmitt and Ru-
dolph, 2007). Importantly, the findings of these studies varied,
most likely due to differences in the OA patient populations exam-
ined. For example, Lewek et al. (2004, 2006) studied ten OA pa-
tients, 49 years old, with medial compartment OA and genu
varum, while Messier et al. (2005) examined ten OA patients,
74 years old, with varying OA severities in all three knee compart-
ments. Schmitt and Rudolph (2007) studied a larger cohort (28 pa-
tients), 60 years old, and OA confined to the medial compartment.
So one must control for confounding factors such as walking speed
and assess the level of pain and function if we are to examine OA
gait and the relationships between gait kinematics, kinetics and
muscle activation patterns.
Co-contraction of the knee muscles provides a means to alter the
stability and articular loading of the joint (Hubley-Kozey et al., 2008).
Additionally and as previously mentioned, knee OA patients may
walk with increased co-contraction of the knee muscles. In general,
there are two forms of knee muscle co-contraction: generalised co-
contraction and directed co-contraction (Lloyd and Buchanan,
2001). In generalised co-contraction all agonists and antagonists of
the knee co-activate equally, whereas in directed co-contraction
medial agonists and antagonists are activated to support abduction
moments and lateral muscles for adduction moments. Directed co-
contraction is believed to directly support the external moment to
prevent condylar lift-off and reduce the concentration of articular
loading in the medial knee compartment (Schipplein and Andriacchi,
1991). Generalised co-contraction can also have this effect but be-
cause of the non-directionality it is less effective in preventing condy-
lar lift-off, and may unduly increase all articular loading (Andriacchi
et al., 1984; Lloyd and Buchanan, 2001; Zhang et al., 2001). General-
ised co-contraction has been identified when people were supporting
the isometric adduction/abductions knee moments (Lloyd and Bu-
chanan, 2001; Zhang et al., 2001), and during sidestepping and cross-
over cutting (Besier et al., 2003a,b). Directed co-contraction has been
demonstrated in ligamento-muscular reflexes to resist adduction/
abduction perturbations at the knee (Buchanan et al., 1996), and vol-
untary directed co-contraction has been shown to the support static
knee abduction–adduction moments (Andriacchi et al., 1984; Zhang
et al., 2001), and abduction moments at the knee during side stepping
(Besier et al., 2003a,b). Studies examining muscle co-contraction in
knee OA patients have found increased levels of medial muscle co-
contraction (Lewek et al., 2004, 2006) and lateral muscle co-contrac-
tion (Schmitt and Rudolph, 2007) in the OA patients compared with
control subjects. However, these studies did not compare the medial
versus lateral muscle co-activations, i.e. directed co-contraction. Re-
cently, greater lateral relative to medial hamstring activation during
gait has been shown in both severe (Hubley-Kozey et al., 2008) and
moderate OA patients (Hubley-Kozey et al., 2006), possibly in an at-
tempt to control medial joint articular loading. However, these inves-
tigations did not control for walking speed or age of the comparison
group (Hubley-Kozey et al., 2006), or did not use a comparison group
(Hubley-Kozey et al., 2008). Therefore it remains unclear if this pat-
tern is distinctive to knee OA patient gait. It is also possible that knee
OA patients may increase their directed co-contraction of the lateral
muscles to support the increased external knee adduction moment
commonly reported in knee OA patients. Finally, directed co-contrac-
tion may be related to the self-perceived pain and symptoms experi-
enced by knee OA patients.
Consequently, the aims of this study were; (1) to determine if
the knee joint kinematic and kinetic variables, previously deter-
mined to be altered in knee OA patient gait, all occur within the
same OA patients, (2) determine levels of net muscle activation
in knee OA patient gait and examine two directed co-contraction
ratios; (a) medial and lateral quadriceps, hamstrings and gastroc-
nemius muscles and (b) medial and lateral hamstring muscles,
and (3) examine the relationships between self-perceived pain,
symptoms and physical function with gait kinematics, kinetics,
and muscle directed co-contraction.
2. Methods
Thirty asymptomatic control subjects (19 females) and 54 phy-
sician-diagnosed knee OA patients (30 females) were recruited.
Public advertisements were used for the recruitment of controls
whilst OA patients were recruited through both local orthopedic
outpatient clinics and advertisements. All OA patients had radio-
graphic signs of knee OA, BMI <35, experienced morning knee stiff-
ness <30 min, knee varus/valgus alignment 65°, could walk
unassisted, and had not received steroid injections in the past
6 months, or regular physiotherapy in the last 12 months. Inclu-
sion in the program for both OA patients and controls was re-
stricted to those aged between 50 and 80. Exclusion criteria were
neurological and/or cardiovascular disorders, surgery or injury to
the back or lower limbs in the past 2 years, and any form of inflam-
matory arthritis (e.g., rheumatoid or psoriatic). All procedures were
approved by the University of Western Australia Human Research
Ethics Committee and all participants gave their informed, written
consent prior to being enlisted in the study.
The Knee Osteoarthritis Outcome Survey (KOOS) (Roos et al.,
1998) and the Medical Outcomes Study 36-item short form health
survey (SF-36) (Ware and Sherbourne, 1992) questionnaires were
administered to all participants prior to testing. The KOOS is a 42
item disease-specific instrument with subscales of pain, other
symptoms, function in daily living (ADL), sport and recreation
(Sport/Rec) and knee related quality of life (QOL) (Roos and Loh-
mander, 2003). Use of the KOOS in a knee OA population has been
found to be a valid and reliable tool (Roos and Toksvig-Larsen,
2003). Due to the age and current activity status of the OA patients,
however, the subscale Sport/Rec was given the additional selection
option of ‘‘not applicable” (N/A) (Roos and Lohmander, 2003; Roos
and Toksvig-Larsen, 2003). KOOS subscales are scored from 0 to
100 with 0 indicating extreme problems and 100 indicating no
problems.
The SF-36 assesses general health and comprises a physical
component score (PCS) and a mental component score (MCS).
PCS and MCS scores were developed using norm based methods
 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841
and then standardised using linear T-score transformations (Ware
et al., 2002) with values ranging between 23.9–61.6 for PCS, and
30.1–64.2 for MCS.
Motion, ground reaction force and electromyographic (EMG)
data from gait were collected synchronously using Workstation
Version 4.6 (Oxford Metrics, Oxford, UK). Motion data were cap-
tured using a seven-camera VICON MX motion analysis system (Ox-
ford Metrics, Oxford, UK) operating at 100 Hz. Ground reaction
force data were collected at 2000 Hz from two AMTI force platforms
(AMTI, Watertown, MA, USA), indistinguishable from the floor, at
the midpoint of the 10 m walkway. OA patients walked at their nat-
urally selected speed, whilst the controls carried out walking trials
at three speeds (a naturally selected speed, slow walking and very
slow walking) for the purpose of matching walking speed. A mini-
mum of six trials with good force plate contact were collected for
each walking speed. To reflect what the participants wore on a daily
basis and subsequently reflect the gait patterns that they experi-
ence daily, participants were required to wear their preferred com-
fortable everyday walking shoes. Retro-reflective markers, 10 mm
in diameter, were placed according to previously established meth-
ods (Besier et al., 2003a,b). This involved clusters of three markers
on both thigh and shank segments, and individual markers on the
landmarks of left and right anterior superior iliac spines, and left
and right posterior superior iliac spine, as well as over the calca-
neous, and the first and fifth metatarsals of each foot.
EMG data were recorded at 2000 Hz using a 16 channel EMG
system (Delsys, Boston, MA). After abrading and cleaning the skin,
ClearTraceTM Ag/AgCI disposable surface EMG electrodes (ConMed
Corporation, Utica, NY, USA), 30 mm diameter, were applied in line
with the muscle fibres over the mid muscle bellies of the quadri-
ceps (rectus femoris [RF], vastus lateralis [VL], vastus medialis
[VM]); hamstrings (biceps femoris [BF] and semimembranosus
[SM]); and gastrocnemius (medial gastrocnemius [MG] and lateral
gastrocnemius [LG]) muscles of the involved leg (most painful or
functionally limited limb in OA patients and matched limb of con-
trols) using double differential electrodes with an inter-electrode
distance of 30 mm. A reference electrode was placed over the head
of the fibula. The raw EMG data were checked for artefacts after
placement of electrodes and prior to data collection.
2.1. Data analysis
The gait parameters of the involved leg of each subject were
used in the data analysis. Marker trajectories and force data were
filtered using a fourth order zero lag Butterworth filter (Vaquita
Butterworth Filter Plugin Version 1.5) with a cut off frequency of
10 Hz. Data were processed using custom kinematic and inverse
dynamic models (Besier et al., 2003a,b) in Bodybuilder (VICON
Peak, Oxford, UK) to calculate 3D knee joint kinetics and kinematics
from the gait trials. The model uses functional hip and knee move-
ments to identify the hip joint centres and knee axes (Besier et al.,
2003a,b) using a custom MATLAB (Mathworks Inc., Natick, USA)
program, which has been shown to produce more repeatable data
(Besier et al., 2003a,b). The knee centre was identified as the mid-
point of the femoral epicondyles along the mean helical knee axis,
defined from the swinging shank’s marker trajectories referenced
relative to the thigh anatomical coordinate system (Besier et al.,
2003a,b). The hip joint centre was determined as the centre of
spheres transcribed by the thigh marker trajectories relative to
the pelvis anatomical coordinate system (Besier et al., 2003a,b). A
foot alignment rig was used to define the subjects foot abduction/
adduction and rear foot inversion/eversion angles (Besier et al.,
2003a,b). Knee kinematics were expressed as per the ISB standard
convention (Wu and Cavanagh, 1995) and external moments were
calculated with inverse dynamics (Besier et al., 2003a,b; Kadaba
et al., 1990) using the body segment parameters of de Leva (1996).
835
Further processing of the raw EMG data and gait kinematic and
kinetic data were carried out using custom software (MatlabTM ver-
sion R2007a). Raw EMG data were full wave rectified and then fil-
tered with a zero lag, fourth order, 6 Hz low pass Butterworth filter
to create linear envelopes. These data were then amplitude nor-
malized to maximum values calculated from maximal isometric
strength trials carried out using a Biodex dynamometer.
Gait strides were normalized to 51 points using a cubic spline
function, and then further broken down into four sub-phases
throughout stance (Fig. 1). These sub-phases were: loading first 15%
of stance, early stance 15–40% of stance, mid-stance 40–60% of stance
and late stance the last 40% of stance. The kinematic variables studied
included knee flexion angle at heel strike, peak knee flexion angle in
early stance, and peak extension angle during mid-stance. Kinetic
variables consisted of the peak adduction moment during early
stance, peak flexion moment in early stance, peak flexion moment
in late stance, peak extension moment in loading and late stance.
There are three characteristics of muscle co-contraction: (1) the
ratio of the agonist to antagonist activation, (2) the total activation
of the agonist and antagonist muscles, and (3) determination of
which agonists or antagonists are the dominant co-contractors.
Two variables were constructed to examine these features of co-
contraction: (1) directed co-contraction ratios (DCCR) of agonists
and antagonists, and (2) net muscle activation. DCCRs were calcu-
lated for the medial (SM, VM, MG)/lateral (BF, VL, LG) muscles
(MLDCCR), medial (SM)/lateral (BF) hamstrings (HAMDCCR) and
the knee flexors (SM, BF, MG, LG)/extensors (VL, VM, RF) (FEDCCR).
The DCCRs were calculated as follows:
If agonist mean EMG > antagonist mean EMG;
DCCR = 1 antagonist mean EMG  agonist mean EMG
Else
DCCR = agonist mean EMG  antagonist mean EMG 1
In these equations if agonists (e.g. extensors or medial muscles)
were more active than the antagonists (e.g. extensors or lateral
muscles) the DCCR would be above zero, and vice versa. Maximum
co-contraction would be represented with a DCCR equal to zero,
while a minimum co-contraction is indicated with a DCCR of 1 or
1. In addition to the DCCR, net muscle activation values (i.e. the
sum of all agonist and antagonist activity) were calculated from
the normalized EMG data during each phase of stance in gait.
Instantaneous values of all co-contraction measures were deter-
mined across stance, and the mean values of these determined in
each of four stance sub-phases.
2.2. Statistical analyses
Between group differences for age, height, body mass, and
walking speed were examined using t-tests. The Mann–Whitney
U-test was used to analyse non-parametric data, whilst analysis
of variance (ANOVA) was used to compare between group differ-
ences on the remaining dependant variables (co-contraction ratios,
and gait variables). Pearson’s correlations (r) were used to examine
relationships between muscle co-contraction and gait parameters
of knee OA patients and controls, and between the self-perceived
measures (KOOS and SF-36) and the gait parameters of the OA pa-
tients only. All statistical analyses were done using SPSS for Win-
dows (Version 12.0; SPSS, Chicago, IL, USA). Due to the number
of statistical tests a significance level of P 6 0.01 was used for all
examinations.
3. Results
There were no differences between controls and OA patients in
age and height (Table 1), but OA patients were 13% heavier than
controls (P = 0.002). Controls slow walking trials were compared
836 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841

Fig. 1. A representation of the external adduction moment (top), flexion/extension moment (middle) and sagittal knee joint angle (bottom) throughout the stance phases of
gait.

to the OA patients natural walking speed and there was no signif-
icant difference in these walking speeds between groups
(P = 0.240).
The sport and recreation subscale on the KOOS was not com-
pleted by many participants due to their current level of disability
so these results were not analysed. OA patients scored lower than
controls on all other KOOS subscales with the greatest reduction
seen for QOL scores (63%) (Table 1). The SF-36 revealed a signifi-
cantly lower PCS for the OA patients (P < 0.001) compared with
controls, but no difference between the groups for MCS (P = 0.883).
OA patients exhibited greater knee flexion at heel strike
(P < 0.001), greater peak knee flexion during early stance
(P = 0.005) and less knee extension during late stance (P < 0.001)
(Table 2). The knee flexion excursion from heel strike to
mid-stance was not significantly different (P = 0.100) between OA
patients and controls (effect size = 0.18).
Comparison of peak external knee joint moments in OA patients
and controls (Table 2) revealed peak adduction levels during early
stance that were 7.7% greater than those experienced by the con-
trol group, although this was not significant (P = 0.301). The OA pa-
tients exhibited trends towards higher peak flexion moments in
late stance (P = 0.040) and greater knee extension moments during
loading (P = 0.039), but these variables had small effect sizes. The
OA patients, however, did experience a significantly lower peak
external knee extension moment in late stance (P < 0.001) with a
moderate effect size (0.40).
OA patients exhibited significantly greater net muscle activa-
tion in both loading (P < 0.01) and early stance (P = 0.002) com-
pared with controls (Fig. 2). The increase in net muscle activation
during loading was negatively correlated (r = 0.461) with the
FEDCCR in the same phase showing an increase in the flexor mus-
cle activation. During early stance the increase net muscle
 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841 837

Table 1
Subject demographics and questionnaire data (mean (SD)) comparing knee OA patients and controls.

Variable Controls (n = 30) OA patients (n = 54) Test statistic P-value

Subject characteristics t-value
Age (years) 64 (6) 65 (8) 1.041 0.301
Height (m) 1.70 (0.09) 1.70 (0.09) 0.174 0.862
Body mass (kg) 71.2 (13.2) 81.4 (14.2) 3.307 0.002
Gait velocity (m/s) 1.10 (0.11) 1.12 (0.153 1.184 0.240
KOOS questionnaire U-value
Pain 96.5 (6.0) 57.6 (18.2) 21.0 <0.001
Symptoms 94.3 (7.9) 53.6 (17.6) 24.0 <0.001
ADL 96.8 (5.5) 60.0 (20.0) 27.0 <0.001
QOL 89.6 (14.9) 32.4 (16.8) 12.5 <0.001
SF-36 questionnaire
PCS 54 (5) 39 (9) 103.0 <0.001
MCS 53 (9) 51 (12) 750.0 0.883

Note: ADL = activities of daily living, QOL = quality of life, PCS = physical component score, MCS = mental component score. KOOS scores are 0–100 with 0 indicating extreme
problems.

Table 2
Comparison of kinetic and kinematic data for controls and OA patients. Data are shown as mean (SD).

Control OA patients F-value P-value r

Joint moments (%BW  HT)
Peak adduction ES 2.70 (1.03) 2.91 (1.33) 0.492 0.485 0.07
Peak flexion ES 2.34 (1.34) 2.73 (1.45) 1.446 0.233 0.13
Peak flexion LS 1.46 (0.61) 1.75 (0.61) 4.366 0.040 0.23
Peak Extension LD 0.99 (0.32) 1.14 (0.32) 4.403 0.039 0.22
Peak Extension LS 1.13 (0.86) 0.17 (1.18) 15.409 <0.001 0.40
Knee joint angles (degrees)
Flexion HS 1.7 (4.3) 7.8 (5.3) 29.099 <0.001 0.51
Flexion peak ES 14.3 (7.1) 18.6 (6.3) 8.360 0.005 0.31
Extension peak LS 2.6 (4.7) 8.9 (6.9) 19.839 <0.001 0.44
Flexion excursion HS ES 12.6 (5.0) 10.8 (4.4) 2.773 0.100 0.18

Note: LD = loading, ES = early stance, LS = late stance, MS = mid-stance, HS = heel strike, r = effect size (0.1 small, 0.3 moderate, 0.5 large).

activation seen in the OA patient group displayed a significant rela-
tionship with both the HAMDCCR (r = 0.356) and the FEDCCR
(r = 0.344) highlighting greater flexor and lateral muscle activa-
tion patterns.
Examination of the muscle co-contraction ratios in gait showed
the FEDCCR differing only during the mid-stance phase where con-
trols exhibited greater flexor muscle activation (P < 0.001) (Fig. 3C).
The HAMDCCR differed between the controls and OA patients
Fig. 2. Net muscle activation (sum of all normalized muscle activity) for OA
patients and controls throughout the stance phases of gait. Significant differences
in Loading and Early Stance at P < 0.01.
(Fig. 3B), with increased lateral hamstring activation during load-
ing, early stance and mid-stance for the OA patients compared
with controls. Control subjects utilized a medial hamstring strat-
egy during loading and an almost equal HAMDCCR during early
stance. In mid-stance, both controls and OA patients used greater
lateral muscle activation but, the OA patients had significantly
higher lateral muscle activity compared with controls.
The MLDCCR showed a distinctive difference between loading
to mid-stance with OA patients having greater lateral muscle activ-
ity in loading and early stance whilst the controls had greater med-
ial activity. Then in mid-stance the OA patients tended to utilize
the medial and lateral muscles equally whilst the controls experi-
enced high levels of medial muscle activity (Fig. 3A). There were no
between group differences in net muscle activation, nor any of the
co-contraction ratios in late stance.
Correlation coefficients (r) between patient self-perceived mea-
sures and patient knee joint kinematic and kinetic gait variables
tended toward significance P 6 0.05 (Table 3), whilst there were
significant correlations (P < 0.001) between the PCS scores and
net muscle activation during loading, early stance and mid-stance
(Table 3). Control and OA patient net muscle activation and muscle
co-contraction data revealed significant relationships with the
kinematic and kinetic gait data (Table 4). In loading, significant
relationships existed for knee flexion angle at heel strike and
MLDCCR (P < 0.001) and HAMDCCR, (P < 0.01). In early stance the
peak flexion angle was significantly related to HAMDCCR
(P < 0.01) MLDCCR and FEDCCR (P < 0.001), the external knee flex-
ion moment peak was related to the FEDCCR, HAMDCCR and
MLDCCR (P < 0.01), and the peak knee adduction moment was sig-
nificantly correlated with MLDCCR and HAMDCCR (P < 0.01). In
838 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841

Table 3
Pearson’s correlation coefficients for self-perceived disability measures, and knee
joint kinetics, kinematics, and muscle co-contraction variables for OA patient data
only.

Pain Symptom PCS

Kinematics
Flexion HS 0.047 0.061 0.347#
Flexion peak ES 0.036 0.024 0.100
Extension peak LS 0.048 0.031 0.328#
Kinetics
Peak flexion ES 0.011 0.027 0.025
Peak flexion LS 0.058 0.193 0.090
Peak extension LD 0.186 0.159 0.091
Peak extension LS 0.027 0.123 0.254
Peak adduction ES 0.290* 0.288* 0.005
Muscle activity
FEDCCR LD 0.188 0.150 0.211
FEDCCR ES 0.138 0.040 0.182
FEDCCR MS 0.022 0.030 0.118
MLDCCR LD 0.162 0.078 0.215
MLDCCR ES 0.102 0.070 0.144
MLDCCR MS 0.020 0.126 0.101
HAMDCCR LD 0.258 0.218 0.389#
HAMDCCR ES 0.251 0.125 0.319*
HAMDCCR MS 0.138 0.098 0.354#
Net activation LD 0.367# 0.012 0.482**
Net activation ES 0.405# 0.060 0.557**
Net activation MS 0.412# 0.071 0.607**

Note: PCS = Physical component score of the SF-36, HS = heel strike, ES = early
stance, LS = late stance and LD = loading.
*
P < 0.05.
**
P < 0.001.
#
P < 0.02.

Table 4
Correlations between muscle co-contraction and kinematic and kinetic gait variables
for each phase of stance.

MLDCCR HAMDCCR FEDCCR Net

activation
Loading phase
Flexion angle at heel 0.288# 0.354* 0.178 0.253
strike
Peak extension moment 0.001 0.016 0.056 0.089
Early stance
Flexion angle peak 0.364* 0.331# 0.342* 0.195
Peak adduction moment 0.329# 0.308# 0.100 0.243
Peak flexion moment 0.374* 0.339# 0.308# 0.152
Late stance
Fig. 3. Mean (standard error) for co-contraction ratio data at each stage of stance.
Extension angle peak 0.462* 0.456* 0.488* 0.346*
MLDCCR (A), HAMDCCR (B), and FEDCCR (C). Significant differences at P < 0.001.
Peak flexion moment 0.265 0.284# 0.436* 0.148
Peak extension moment 0.338# 0.380* 0.481* 0.295#

late stance significant relationships existed between peak exten- Note: MLDCCR = medial/lateral directed co-contraction ratio, HAMDCCR = medial
hamstrings/lateral hamstrings directed co-contraction ratio, FEDCCR = flexion/
sion angle and all muscle co-contraction variables, peak extension
extension directed co-contraction ratio, and net activation = sum of all normalized
moment and all muscle co-contraction variables, whist the peak muscle activity.
flexion moment was related to both HAMDCCR and FEDCCR (Table *
P < 0.001.
#
4). P < 0.01.

4. Discussion with moderate to severe OA (Thorstensson et al., 2005). The differ-

This study examined the knee joint kinematic and kinetic gait
alterations most commonly associated with knee OA patients,
compared with a similarly aged asymptomatic control group,
whilst controlling for walking speed. Further, we examined the
muscle co-contraction levels during gait, and how they related to
the knee joint kinematic and kinetic features of gait, and the pain
and symptoms experienced by the knee OA patients.
The OA patients in this examination exhibited levels of patient
self-perceived disability and physical function, reported on the
KOOS and SF-36, similar to those previously reported by patients
ence in OA patient self-perceived measures, compared with con-
trols, shows that knee OA patients are subjected to higher levels
of disability than those resulting from age alone.
Knee joint kinematic measures in knee OA patients have been
widely reported. In moderate OA patients kinematic differences
have been found to differ very little to those of controls with the
outward appearance of gait being similar (Astephen et al., 2008).
OA patients in this study exhibited greater knee flexion angles at
heel strike, early stance and during late stance. Heel strike knee
flexion angles, within the knee OA literature, are varied with some
studies finding greater knee flexion (Baliunas et al., 2002; Childs
 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841
et al., 2004; Joss, 2006), some studies reporting greater knee exten-
sion (Mundermann et al., 2005) and others finding no difference
(Rudolph et al., 2007; Smith et al., 2004) in knee OA patients com-
pared to controls. Similarly, peak knee flexion during early stance
also differs between reports with greater knee flexion for controls
(Al-Zahrani and Bakheit, 2002; Lewek et al., 2006) or no difference
in the peak knee flexion values (Baliunas et al., 2002; Joss, 2006).
Reductions in knee flexion excursion have been found in several
examinations of knee OA patients (Childs et al., 2004; Lewek
et al., 2006; Rudolph et al., 2007) a finding that was not replicated
in the current investigation. Zeni and Higginson (2009) found that
patients with moderate OA did not differ from controls in knee
joint excursion when statistically controlling for walking speed,
although severe OA patients had significantly reduced knee joint
excursion. These findings suggest that knee excursion is related
to OA grade and walking speed, both of these factors probably
the cause for limited between group differences in knee kinematics
in the current study.
The knee joint kinematic model used in our study may have
contributed to the observed between group differences in the knee
kinematics. However, other studies have also shown greater knee
flexion at heel strike and similar knee flexion excursion in early
stance, in OA subject’s, using other marker sets and gait analysis
methods (Baliunas et al., 2002; Childs et al., 2004). In addition,
we did not see any gross knee flexion deformities in this group
and inclusion criteria for knee alignment probably excluded exces-
sive deformities in our OA population. Finally, we also analysed
data from trials that were used to determine the knee mean flex-
ion–extension helical axis. In these trials participants cyclically
moved their knee from a comfortable position near full extension
to comfortable flexed knee posture. All patients had a knee flexion
range of at least 90° from full extension.
There was little difference in the knee flexion–extension mo-
ments between the knee OA patients and asymptomatic controls
in the current study. The only difference was for the external
extension moment in late stance; this variable not being often re-
ported in the knee OA literature, although Baliunas et al. (2002)
found a non-significant reduction in this variable in knee OA pa-
tients compared to controls. We found no differences in the peak
flexion moments in early stance, these being typical of an internal
extension moment avoidance gait, often reported in people with
knee OA (Astephen et al., 2008; Smith et al., 2004; Zeni et al.,
2009). However, not all knee OA patients demonstrate extension
moment avoidance (Smith et al., 2004). It is also well known that
walking speed affects flexion–extension moments in the gait of
normal subjects (Kirtley et al., 1985; Lelas et al., 2003) and people
with knee OA (Astephen et al., 2008; Mundermann et al., 2004;
Zeni et al., 2009). Zeni and Higginson (2009) showed that all differ-
ences in flexion–extension moments between controls and pa-
tients with differing knee OA severity could be statistically
accounted for by their different walking speeds. Using matched
walking speeds of OA participants and controls, Baliunas et al.
(2002) also found no differences in these flexion–extension mo-
ments. Therefore, controlling for walking speed in our control
and knee OA groups probably contributed to our lack of between
group differences in the flexion–extension moments.
The often reported increase in peak external knee adduction mo-
ment during early stance was not present in the current OA patient
group. This may have been due to location and grade of knee OA of
our patients. Studies that have found increases in knee adduction
moments, compared to age matched controls walking at the same
speed, have studied OA patients with specific medial compartment
pathology (Baliunas et al., 2002; Lewek et al., 2004; Rudolph et al.,
2007; Schipplein and Andriacchi, 1991; Schmitt and Rudolph,
2007). In contrast, studies that have not found increased knee
adduction moments have used groups with OA located in various
839
compartments of the knee (Kaufman et al., 2001; Messier et al.,
2005). It seems likely that the reported increase in knee adduction
moment is specific to medial compartment knee OA patients. How-
ever, we did not have radiographic information to reliably identify
the location of OA and therefore we cannot confirm this affect.
Although there was no radiographic data, our OA patients’ had
self-perceived disability consistent with a moderate to severe
grouping (Thorstensson et al., 2005), and self-perceived disability
has been associated with the size of the adduction moments (Shar-
ma et al., 1998). We additionally used a strict inclusion criterion of
65° varus/valgus alignment to ensure that our OA patient group
had close-to-normal alignment. Hurwitz et al. (2002) found that
varus alignment of knee OA patients was positively associated with
the peak knee adduction moments. Walking speed also affects the
size of the peak adduction moments in OA patients (Mundermann
et al., 2004; Zeni et al., 2009). Therefore, the moderate level of knee
OA and neutral limb alignment of our OA patient group, and con-
trolling for walking speed in the control group, contributed to the
lack of adduction moment differences between groups.
The OA patients in this current study employed greater levels of
net muscle activation during loading and early stance. Previous re-
search has shown increased muscle activity in the hamstrings
(Astephen et al., 2008; Hortobagyi et al., 2005; Schmitt and Ru-
dolph, 2007), quadriceps (Astephen et al., 2008; Hubley-Kozey
et al., 2006; Schmitt and Rudolph, 2007), and gastrocnemius (Aste-
phen et al., 2008) muscles during the stance phase of OA patient
gait. The greater levels of muscle activation seen the OA patients
in the current investigation were associated with greater levels
of flexor activation in loading and greater flexor and lateral muscle
activation in early stance. Additionally, net muscle activation val-
ues in loading, early stance and mid-stance were shown to increase
as patient self-perceived physical function deteriorated. Our exam-
ination of muscle co-contraction differed from previously used co-
contraction indices (Lewek et al., 2004, 2006; Rudolph et al., 2007;
Schmitt and Rudolph, 2007) and is the first to directly examine
medial/lateral differences in knee OA patients compared with con-
trols of the same age. The use of both medial/lateral and flexion/
extension DCCR comparisons revealed large differences between
controls and OA patients in this investigation. A directed co-con-
traction strategy was observed with knee OA patients exhibiting
greater lateral muscle activation during loading, early stance and
mid-stance that was clearly different to the predominantly medial
muscle activation exhibited by the controls. A similar pattern has
been found in both severe and mid stage knee OA patients (Hub-
ley-Kozey et al., 2008; Hubley-Kozey et al., 2006) and shows a dis-
tinctive difference in the muscle activation strategies utilized by
the OA patient and control groups.
This is also the first study to show that laterally directed co-con-
traction of the knee muscles increases with increasing knee adduc-
tion moments in those with knee OA. The knee flexion angle at heel
strike and peak adduction moment in early stance revealed signif-
icant positive correlations with the MLDCCR and HAMDCCR. In
early stance an increased peak adduction moment corresponded
to increased lateral muscle activity across patients. This dominant
lateral pattern of muscle activation continued throughout mid-
stance. Such a response may aid in controlling lateral knee joint
opening (Schipplein and Andriacchi, 1991), and provide greater
knee joint stability against the prevailing adduction moment. This
activity suggests a laterally directed co-contraction strategy, in re-
sponse to the knee adduction moment, similar to that found in side-
stepping in healthy individuals (Besier et al., 2003a,b). However, in
this population this strategy is not simply due to a perturbation, as
found previously (Buchanan et al., 1996; Dhaher et al., 2003) and
may be a developed motor pattern (Dhaher et al., 2003) in response
to the high levels of pain and disability associated with an increased
knee adduction moment as suggested by our results.
840 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841
We found a relationship between pain, symptoms and the
external knee adduction moment whereby increased levels of pain
or symptom coincided with decreases in the adduction moment.
These findings are in agreement with Hurwitz et al. (2000) who
showed that pain was related to the size of the adduction moment
in knee OA patients by washing out NSAIDs and Ebert et al. (2008)
who found a low to moderate correlation for high knee pain, on the
KOOS questionnaire, with lower adduction moments in patients
following matrix-induced autologous chondrocyte implantation.
Therefore, our study shows that in addition to altered knee loading,
increased levels of pain and disability may trigger muscle activa-
tion patterns that redistribute the medial versus lateral loading
of the knee, possibly reducing pain and adduction moments.
Muscle redundancy at a joint, allows a person to generate the
same net joint moment with the same joint posture and movements
but with different muscle activation patterns (Buchanan et al.,
1996; Tax et al., 1990). This is especially the case with co-contrac-
tion. The antagonist muscles counter the agonist muscles, therefore
with higher levels of co-contraction the same net moments are gen-
erated but with higher net muscle activity (Lloyd et al., 2008; Lloyd
and Buchanan, 2001). This is exemplified in the current study that
when compared to the controls, matched for walking speed, age
and knee alignment, knee OA patients had increased levels of net
muscle activity and laterally directed co-contraction, while still
having similar joint moments and posture. Zeni and Collgues
(2009) have recently shown that generalized co-contraction in-
creases even when statistically controlling for walking speed in con-
trol subjects and patients with differing knee OA severities. They
also showed, in a sister paper (Zeni et al., 2009), that essentially
all kinematic and kinetic differences between controls and patients
could be statistically accounted for by their different walking
speeds. Our results are therefore consistent with Zeni and Collgues
(2009) recent research, although we further explored different fac-
ets and types of co-contraction in the knee OA patients.
Muscle activation from electromyography does not reflect the
actual mechanical action of the muscles. For example the MLDCCR
and total activation does not reveal the extent of the lateral muscle
contribution to support the knee adduction moment. This moment
is affected by many factors; e.g. muscle physiological cross section
area, muscle fibre force–length and force–velocity relationships,
and muscle moment arm (Lloyd and Buchanan, 2001; Lloyd
et al., 2005, 2008), in addition to muscle activation. The effective-
ness of muscle stabilization and the affect on articular loading re-
quires studies that employ neuromuscular skeletal modelling
(Lloyd et al., 2005, 2008), which is beyond the scope of this paper,
but needs to be performed in further research. Nevertheless, the
characteristics of the muscle activation and co-contraction pat-
terns still provide information about the motor patterns used by
the knee OA patients.
Patients in this current study wore their own comfortable every-
day walking shoes. This was to ensure their laboratory gait patterns
reflected those they experienced daily. The only stipulation was
participants shoes had to fit securely to their foot and that they
had only a low or no heel as higher heel shoes in women have been
shown to increase peak adduction moments (Kerrigan et al., 1998).
Recent investigations have also found footwear to affect the knee
joint loading in OA patient gait (Shakoor et al., 2008). Using specific,
common shoes in our study may have altered the results and not
have reflected the normally experienced gait patterns.
5. Conclusions
In conclusion, this examination of knee OA patient gait com-
pared with controls of similar age, whilst controlling for confound-
ing factors of walking speed, and limb alignment, has shown that
muscle activation patterns are varied despite only little alteration
in kinematics and kinetics. We have established that the medial/
lateral patterns of muscle activation in a group of knee OA patients
differ from asymptomatic controls. Patients in this cohort exhib-
ited higher levels of lateral muscle activation throughout the
stance phase of gait possibly as a protective mechanism against
pain experienced with external knee adduction moments. In addi-
tion, this laterally directed co-contraction increased with increas-
ing levels of the external adduction moment in people with knee
OA. The presence of this control of directed co-contraction may
have large ramifications for the stabilisation of the knee and artic-
ular loading, and should be the focus of future studies.
Conflict of interest statement
None of the authors have any financial and personal relation-
ships with other people or organisations that could inappropriately
influence (bias) this work.
Acknowledgements
The authors wish to acknowledge The University of Western
Australia for their funding assistance for this work. We would also
like to thank the orthopaedic surgeons involved in recruitment of
patients; Riaz Khan, Sean Williams, and Bo Nivbrant. Thanks must
go to Dr Chris Winby for his Matlab programming skills.
References
Al-Zahrani, K.S., Bakheit, A.M.O., 2002. A study of the gait characteristics of patients
with chronic osteoarthritis of the knee. Disability and Rehabilitation 24, 275–
280.
Andriacchi, T.P., Andersson, G.B.J., Ortengren, R., Mikosz, R.P., 1984. A study of the
factors influencing muscle activity about the knee joint. Journal of Orthopaedic
Research 1, 266–275.
Andriacchi, T.P., Ogle, J.A., Galante, J.O., 1977. Walking speed as a basis for normal
and abnormal gait measurements. Journal of Biomechanics 10, 261–268.
Astephen, J.L., Deluzio, J., Caldwell, G.E., Dunbar, M., Hubley-Kozey, C., 2008. Gait
and neuromuscular pattern changes are associated with differences in knee
osteoarthritis severity levels. Journal of Biomechanics 41, 868–876.
Astephen, J.L., Deluzio, K.J., 2005. Changes in frontal plane dynamics and the loading
response phase of the gait cycle are characteristic of severe knee osteoarthritis
application of a multidimensional analysis technique. Clinical Biomechanics 20,
209–217.
Astephen, J.L., Deluzio, K.J., Caldwell, G.E., Dunbar, M.J., 2007. Biomechanical
changes at the hip, knee, and ankle joints during gait are associated with knee
osteoarthritis severity. Journal of Orthopaedic Research.
Baliunas, A.J., Hurwitz, D.E., Ryals, A.B., Karrar, A., Case, J.P., Block, J.A., Andriacchi,
T.P., 2002. Increased knee joint loads during walking are present in subjects
with knee osteoarthritis. Osteoarthritis and Cartilage 10, 573–579.
Besier, T.F., Lloyd, D.G., Ackland, T.R., 2003a. Muscle activation strategies at the knee
during running and cutting maneuvers. Medicine and Science in Sports and
Exercise 35, 119–127.
Besier, T.F., Sturnieks, D.L., Alderson, J.A., Lloyd, D.G., 2003b. Repeatability of gait
data using a functional hip joint centre and a mean helical knee axis. Journal of
Biomechanics 36, 1159–1168.
Brinkmann, J.R., Perry, J., 1985. Rate and range of knee motion in healthy and
arthritic subjects. Physical Therapy 65, 1055–1060.
Buchanan, T.S., Kim, A.W., Lloyd, D.G., 1996. Selective muscle activation following
rapid varus/valgus perturbations at the knee. Medicine and Science in Sports
and Exercise 28, 870–876.
Chen, C.P.C., Chen, M.J.L., Pei, Y.C., Lew, H.L., Wong, P.Y., Tang, S.F.T., 2003. Sagittal
plane loading response during gait in different age groups and people with knee
osteoarthritis. American Journal of Physical Medicine and Rehabilitation 82,
307–312.
Childs, J.D., Sparto, P.J., Fitzgerald, G.K., Bizzini, M., Irrgang, J.J., 2004. Alterations in
lower extremity movement and muscle activation patterns in individuals with
knee osteoarthritis. Clinical Biomechanics 19, 44–49.
de Leva, P., 1996. Adjustments to Zatsiorsky-Seluyanovs segment inertia
parameters. Journal of Biomechanics 29, 1223–1230.
Dhaher, Y.Y., Tsoumanis, A.D., Rymer, W.Z., 2003. Reflex muscle contractions can be
elicited by valgus positional perturbations of the human knee. Journal of
Biomechanics 36, 199–209.
Dieppe, P.A., 2004. Relationship between symptoms and structural change in
osteoarthritis. What are the important targets for osteoarthritis therapy?
Journal of Rheumatology-Supplement 70, 50–53.
 T.L. Heiden et al. / Clinical Biomechanics 24 (2009) 833–841
Ebert, J.R., Robertson, W.B., Lloyd, D.G., Zheng, M.H., Wood, D.J., Ackland, T.R., 2008.
Traditional vs accelerated approaches to post operative rehabilitation following
matrix-induced autologous chondrocyte implantation (MACI): comparison of
clinical, biomechanical and radiographic outcomes. Osteoarthritis and Cartilage
16, 1131–1140.
Hanlon, M., Anderson, R., 2006. Prediction methods to account for the effect of gait
speed on lower limb angular kinematics. Gait Posture 24, 280–287.
Hortobagyi, T., Westerkamp, L., Beam, S., Moody, J., Garry, J., Holbert, D., DeVita, P.,
2005. Altered hamstring-quadriceps muscle balance in patients with knee
osteoarthritis. Clinical Biomechanics 20, 97–104.
Hubley-Kozey, C., Deluzio, K.J., Dunbar, M., 2008. Muscle co-activation patterns
during walking in those with severe knee osteoarthritis. Clinical Biomechanics
23, 71–80.
Hubley-Kozey, C.L., Deluzio, K.J., Landry, S.C., McNutt, J.S., Stanish, W.D., 2006.
Neuromuscular alterations during walking in persons with moderate knee
osteoarthritis. Journal of Electromyography and Kinesiology 16, 365–378.
Hurwitz, D.E., Ryals, A.B., Case, J.P., Block, J.A., Andriacchi, T.P., 2002. The knee
adduction moment during gait in subjects with knee osteoarthritis is more
closely correlated with static alignment than radiographic disease severity, toe
out angle and pain. Journal of Orthopaedic Research 20, 101–107.
Hurwitz, D.E., Ryals, A.R., Block, J.A., Sharma, L., Schnitzer, T.J., Andriacchi, T.P., 2000.
Knee pain and joint loading in subjects with osteoarthritis of the knee. Journal
of Orthopaedic Research 18, 572–579.
Joss, B.K., 2006. Clincal and biomechanical outcomes following unicondylar knee
arthroplasty with preservation fixed and mobile bearing tibial components,
thesis. University of Western Australia, Perth.
Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., 1990. Measurement of lower
extremity kinematics during level walking. Journal of Orthopaedic Research 8,
383–392.
Kaufman, K.R., Hughes, C., Morrey, B.F., Morrey, M., An, K.-N., 2001. Gait
characteristics of patients with knee osteoarthritis. Journal of Biomechanics
34, 907–915.
Kerrigan, D.C., Todd, M.K., Riley, P.O., 1998. Knee osteoarthritis and high heeled
shoes. Lancet 351, 1399–1401.
Kirtley, C., Whittle, M.W., Jefferson, R.J., 1985. Influence of walking speed on gait
parameters. Journal of Biomedical Engineering 7, 282–288.
Lelas, J.L., Merriman, G.J., Riley, P.O., Kerrigan, D.C., 2003. Predicting peak kinematic
and kinetic parameters from gait speed. Gait and Posture 17, 106–112.
Lewek, M.D., Rudolph, K.S., Snyder-Mackler, L., 2004. Control of frontal plane knee
laxity during gait in patients with medial compartment knee osteoarthritis.
Osteoarthritis and Cartilage 12, 745–751.
Lewek, M.D., Scholz, J., Rudolph, K.S., Snyder-Mackler, L., 2006. Stride-to-stride
variability of knee motion in patients with knee osteoarthritis. Gait Posture 23.
Lloyd, D.G., Besier, T.F., Winby, C.R., Buchanan, T.S., 2008. Neuromusculoskeletal
modelling and simulation of tissue load in the lower extremities. In: Hong, Y.,
Bartlett, R. (Eds.), Routledge Handbook of Biomechanics and Human Movement
Science. Taylor & Francis Books Ltd., Oxford UK, pp. 3–17.
Lloyd, D.G., Buchanan, T.S., 2001. Strategies of the muscular contributions to the
support of static varus and valgus loads at the human knee. Journal of
Biomechanics 34, 1257–1267.
Lloyd, D.G., Buchanan, T.S., Besier, T.F., 2005. Neuromuscular biomechanical
modelling to understand knee ligament loading. Medicine and Science in
Sports and Exercise 37, 1939–1947.
Maly, M.R., Costigan, P.A., Olney, S.J., 2006. Determinants of self-report outcome
measures in people with knee osteoarthritis. Archives of Physical Medicine and
Rehabilitation 87, 96–104.
Martin, D.F., 1994. Pathomechanics of knee osteoarthritis. Medicine and Science in
Sports and Exercise 26, 1429–1434.
Messier, S.P., DeVita, P., Cowan, R.E., Seay, J., Young, H.C., Marsh, A.P., 2005. Do older
adults with knee osteoarthritis place greater loads on the knee during gait? A
preliminary study. Archives of Physical Medicine and Rehabilitation 86, 703–
709.
841
Miyazaki, T., Wada, M., Kawahara, H., Sato, M., Baba, H., Shimada, S., 2002. Dynamic
load at baseline can predict radiographic disease progression in medial
compartment knee osteoarthritis. Annals of the Rheumatic Diseases 61, 617–
622.
Mundermann, A., Dyrby, C.O., Andriacchi, T.P., 2005. Secondary gait changes in
patients with medial compartment knee osteoarthritis: increased load at the
ankle, knee, and hip during walking. Arthritis & Rheumatism 52, 2835–2844.
Mundermann, A., Dyrby, C.O., Hurwitz, D.E., Sharma, L., Andriacchi, T.P., 2004.
Potential strategies to reduce medial compartment loading in patients with
knee osteoarthritis of varying severity: reduced walking speed. Arthritis &
Rheumatism 50, 1172–1178.
Roos, E.M., Lohmander, L.S., 2003. The knee injury and osteoarthritis outcome score
(KOOS): from joint injury to osteoarthritis. Health and Quality of Life Outcomes
1, 64–72.
Roos, E.M., Roos, H.P., Lohmander, L.S., Ekdahl, C., Beynnon, B.D., 1998. Knee injury
and osteoarthritis outcome score (KOOS) – development of a self administered
outcome measure. Journal of Orthopaedic and Sports Physical Therapy 28, 88–96.
Roos, E.M., Toksvig-Larsen, S., 2003. Knee injury and osteoarthritis outcome score
(KOOS) – validation and comparison to the WOMAC in total knee replacement.
Health and Quality of Life Outcomes 1, 17.
Rudolph, K.S., Schmitt, L.C., Lewek, M.D., 2007. Age-related changes in strength,
joint laxity, and walking patterns: are they related to knee osteoarthritis?
Physical Therapy 87, 1422–1432.
Schipplein, O.D., Andriacchi, T.P., 1991. Interaction between active and passive knee
stabilizers during level walking. Journal of Orthopaedic Research 9, 113–119.
Schmitt, L.C., Rudolph, K.S., 2007. Influences on knee movement strategies during
walking in persons with medial knee osteoarthritis. Arthritis & Rheumatism 57,
1018–1026.
Shakoor, N., Lidtke, R.H., Sengupta, M., Fogg, L.F., Block, J.A., 2008. Effects of
specialized footwear on joint loads in osteoarthritis of the knee. Arthritis &
Rheumatism 59, 1214–1220.
Sharma, L., Hurwitz, D.E., Thonar, E.A., Sum, J.A., Lenz, M.E., Dunlop, D.D., Schnitzer,
T.J., Kirwan-Mellis, G., Andriacchi, T.P., 1998. Knee adduction moment, serum
hyaluronan level, and disease severity in medial tibial femoral osteoarthritis.
Arthritis & Rheumatism 41, 1233–1240.
Smith, A.J., Lloyd, D.G., Wood, D.J., 2004. Pre-surgery knee joint loading patterns
during walking predict the presence and severity of anterior knee pain after
total knee arthroplasty. Journal of Orthopaedic Research 22, 260–266.
Stauffer, R.N., Chao, E.Y.S., Gyory, A.N., 1977. Biomechanical gait analysis of the
diseased knee joint. Clinical Orthopaedics 126, 246–255.
Tax, A.A., Denier van der Gon, J.J., Erkelens, C.J., 1990. Differences in coordination of
elbow flexor muscles in force tasks and movement tasks. Experimental Brain
Research 81, 567–572.
Thorstensson, C.A., Roos, E.M., Petersson, I.F., Ekdahl, C., 2005. Six-week high-
intensity exercise program for middle-aged patients with knee osteoarthritis: a
randomized control trial. BMC Musculoskeletal Disorders 6, 27.
Ware, J.E., Kosinski, M.A., Dewey, J.E., 2002. How to Score Version 2 of the SF-36
Health Survey (Standard and Acute Forms). QualityMetric Incorporated, Lincoln,
RI.
Ware, J.E., Sherbourne, C.D., 1992. The MOS 36-item short-form health survey (SF-
36). I. Conceptual framework and item selection. Medical Care 30, 473–483.
Wu, G., Cavanagh, P.R., 1995. ISB recommendations for standardization in the
reporting of kinematic data. Journal of Biomechanics 28, 1257–1261.
Zeni, J.A., Higginson, J.S., 2009. Differences in gait parameters between healthy
subjects and persons with moderate and severe knee osteoarthritis: a result of
altered walking speed? Clinical Biomechanics 24, 372–378.
Zeni, J.A., Rudolph, K., Higginson, J.S., 2009. Alterations in quadriceps and
hamstrings coordination in persons with medial compartment knee
osteoarthritis. Journal of Electromyography and Kinesiology, in press, doi:10.
1016/j.jelekin.2008.12.003.
Zhang, L.Q., Xu, D., Wang, G., Hendrix, R.W., 2001. Muscle strength in knee varus and
valgus. Medicine & Science in Sports & Exercise 33, 1194–1199.