Clinical Biomechanics 28 (2013) 757761

Contents lists available at ScienceDirect

Clinical Biomechanics
j o u r n a l h o m e p a g e : w w w . e l s e v i e r .c o m / l o c a t e / c l i n b i o m e c h

Comparison of gluteus medius muscle activity during functional

tasks in individuals with and without osteoarthritis of the hip joint
Maureen K. Dwyer a,, Kelly Stafford a, Carl G. Mattacola a, Timothy L. Uhl a, Mauro Giordani b
a
b

Department of Rehabilitation Sciences, University of Kentucky, Lexington, KY, USA

Department of Orthopedic Surgery, University of Kentucky, Lexington, KY, USA

a r t i c l e

i n f o

Article history:
Received 22 February 2013
Accepted 15 July 2013
Keywords:
Electromyography
Osteoarthritis
Hip
Muscle
Gait
Step test

a b s t r a c t
Background: Neuromuscular alterations have been reported for patients with osteoarthritis of the hip
joint; however, the underlying cause associated with altered gluteus medius muscle function has not been
examined. This study assessed electromyographic amplitudes of the gluteus medius muscles during
function in patients with unilateral end-stage osteoarthritis of the hip joint compared to controls.
Methods: Patients with unilateral end-stage hip joint osteoarthritis (n = 13) and asymptomatic control
partici- pants (n = 17) participated. Average root-mean squared muscle amplitudes represented as a
percent of maxi- mum voluntary isometric contraction for both the involved and uninvolved limb gluteus
medius muscles were analyzed during step up, step down, and gait. The association between muscle
activation and impact forces dur- ing stepping tasks was assessed.
Findings: Patients with hip osteoarthritis exhibited increased gluteus medius muscle electromyographic
ampli- tudes bilaterally during stair ascent, stair descent, and gait compared to controls, regardless of which
limb they led. Involved limb muscle activity was inversely related to impact force during step down onto
the ipsilateral limb.
Interpretation: Patients with hip osteoarthritis demonstrated increased gluteus medius muscle activation
levels during stepping tasks and gait when compared to controls. The increased activation is most likely a
compensa- tory response to muscle weakness. Therefore, application of strengthening exercises which
target the gluteal muscles should assist in neuromuscular control and result in improved strength for
patients with hip joint
osteoarthritis.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
Osteoarthritis (OA) of the hip joint is a frequent cause of functional
disability in the young adult between 3050 years of age. Joint pain,
in- stability and muscle weakness have been identied as potential
risk fac- tors for functional decline in this population (Dekker et al.,
2009; Felson and Zhang, 1998; Shindle et al., 2008). Continued
function in the pres- ence of neuromuscular alterations may hasten
the progression of joint disease and results in functional alterations
(Herzog et al., 2003; Shrier,
2004). For patients who present with end-stage hip osteoarthritis,
bio- mechanical alterations exist during gait (Pustoc'h and Cheze,
2009; Watanabe et al., 1998) and stair climbing (Pustoc'h and Cheze,
2009), and these changes have been shown to directly correlate with
hip abduc- tor muscle weakness (Vaz et al., 1993). Alterations in the
physiological properties of the gluteal muscles may account for the
functional weak- ness, as fewer numbers and smaller cross-sectional
areas of type II mus- cle bers as well as reductions in the radiological
density of the gluteus
Corresponding author at: Kaplan Joint Center, Newton Wellesley Hospital,
2014
Washington Street, Green Building, Newton, MA 02462,
USA.
E-mail address: mkdwyer@partners.org (M.K.
Dwyer).
0268-0033/$ see front matter 2013 Elsevier Ltd. All rights
reserved. http://dx.doi.org/10.1016/j.clinbiomech.2013.07.007

medius muscle have been observed for this population compared to

con- trol subjects (Arokoski et al., 2002; Rasch et al., 2009). However,
limited research describing alterations in
the neuromuscular
function of this muscle associated with OA of the hip joint exists.
Previous research using surface electromyography (SEMG)
reported alterations in timing and duration of gluteus medius
activation during the gait of elderly subjects with advanced hip
joint disease (Long et al., 1993) as well as increased muscle activity
for the gluteus medius muscles of both limbs during stepping for
subjects with early phase hip joint OA compared to healthy
individuals (Sims et al., 2002). However, none of the previous studies
reported the specic magnitude of differ- ence between populations.
Quantication of the amplitude of muscle activation during activity
will assist in identifying the underlying cause for the observed
muscle alterations. An increase in muscle activa- tion levels may
indicate a weak muscle, resulting from a compensatory increase in
neural drive to achieve the required muscle force to com- plete the
given task (Ling et al., 2007; Sims et al., 2002). Higher muscle
activation during submaximal activities may also indicate an inability
to fully activate due to pain. Contrary, a decrease in activation levels
may indicate a reexive inhibition of neural drive to the muscle, the
body's protective response to joint injury (Hart et al., 2010) or lower
demands

M.K.
M.K.
Dwyer
Dwyer
et al.
et /al.Clinical
/ Clinical
Biomechanics
Biomechanics
28 (2013)
28 (2013)
757
757

75

761761
to the muscle from lower body mass or improved
neuromuscular
con- trol. As there is currently no cure for osteoarthritis, therapeutic
options for young patients are few and often involve surgical
interventions along with rehabilitation. A better understanding of the
neuromuscular alterations and
the subsequent biomechanical
consequences during function associated with end-stage joint
disease may facilitate the crea- tion of appropriate rehabilitation
programs. The purpose of our study was to compare average SEMG
muscle amplitude of the gluteus medius muscles during function
between adults with unilateral advanced phase hip joint OA and agecomparable healthy controls. Given that bilateral alterations in
gluteus medius muscle activation have been reported for patients
with early phase joint disease (Sims et al., 2002), we hypothe- sized
that average gluteus medius muscle SEMG amplitudes will be increased for both the involved and uninvolved limb during gait and
stepping tasks for individuals with hip OA compared to controls. In
ad- dition, we hypothesized that SEMG values of the gluteus medius
mus- cles during stepping will be correlated with impact forces
measured during these tasks.

2. Methods
2.1.
Participants
Adults with unilateral, end-stage OA of the hip joint scheduled
for total hip arthroplasty were recruited from
the patient
population of one of the authors (MG). Subjects were included if
they presented with unilateral hip pain, were between 18 and 65
years of age, had no history of vestibular disorders, and presented
with no major co- morbidities. Subjects were excluded if they had
symptomatic bilater- al hip pain, any history of major lower
extremity surgery, or current lower extremity injury. Healthy
control subjects were recruited from the local area via internet
advertisements and clinic posters. Subjects were considered for the
control group if they reported no pain in ei- ther hip joint and had
no history of major lower extremity injury or surgery. The analysis
sample consisted of 13 patients with OA of the hip joint and 17
controls (Table 1). This research was approved by the university's
Institutional Review Board.
2.2. Pain
Subjects were asked to rate their pain using the Visual Analog
Scale (VAS). Hip symptoms were assessed by asking each subject
During the past 7 days, what is the highest level of pain
experienced during your daily activities? The VAS ranged from 0 to
10 cm, with 10 cm indicat- ing worse pain.
2.3.
Electromyography
A 16-lead SEMG system (Run Technologies, Mission Viejo, CA,
USA) was used to record muscle activity. A Myopac transmitter belt
unit (Run Technologies, Mission Viejo, CA, USA) was worn by each
subject during data collection and used to transmit raw SEMG data
via a ber optic cable to its receiver unit. Unit specications include
an amplier gain of 2000 Hz, an input impedance of 1 M, and a
CMRR of 90 dB. Muscle

2.4. Force platform

Impact force data was collected using the 6 foot long force plate
of the NeuroCom Smart Balance Master (NeuroCom Inc, Clackamas,
OR, USA). The subjects performed step up, step down, and level
gait. A member of the research team instructed each subject on the
perfor- mance of each test on the long force plate. Following MVIC
data collec- tion, each subject performed each task, with a 30second rest in between each trial and a 2 minute rest between
exercises to prevent fa- tigue. Order of testing was randomized
between subjects to prevent order bias, and subjects were allowed
to practice the functional tasks until they
felt condent in
performing them.
2.4.1. Step tasks
Step up and step down were performed three times for each limb for

Table 1
Subject
demographics.

CON
OA

75

activation of the involved and uninvolved limb gluteus medius

muscles was collected for each subject using bi-polar AgAgCl surface
electrodes (Ambu Inc., Glen Burnie, MD, USA) measuring 5 mm in
diameter with a center-to-center distance of 2.0 cm. The involved
limb of the subjects with hip OA was dened as the limb which was
undergoing surgery and was matched to the dominant limb of the
control subjects, dened as the leg with which they would kick a ball.
The skin was prepared, and electrodes were placed in parallel
arrangement over the muscle belly for each muscle, as described by
Cram et al. (1998). Electrode placement for both muscles was the
proximal one-third of the distance between the highest point on the
iliac crest and the greater trochanter of the femur. Electrodes were
secured to the skin using Cover-Roll (Beiersdorf-Jobst, Charlotte, NC,
USA). SEMG data were sampled at 1000 Hz and analyzed using
Datapac software (Run Technologies, Mission Viejo, CA, USA). A foot
switch was placed in the shoe of both lower extremities to determine foot contact and foot lift-off during the performance of
the functional activities. Footswitch data was collected at 1000 Hz and
syn- chronized with the SEMG data using the Datapac software.
Following electrode placement,
subjects were asked to
perform three maximum voluntary isometric contractions (MVICs)
contractions. Each trial lasted 3 s with a 30-second break in between
trials. The sub- jects with hip OA were unable to perform MVIC
testing for the gluteus medius in the traditional side lying hip
abduction position because of pain; therefore, MVIC data were
collected for both groups during iso- metric weight-bearing hip
abduction. Given that the primary function of this muscle is pelvic
stabilization in the frontal plane (Gottschalk et al., 1989; Inman,
1947; Neumann, 2010; Ward et al., 2010), we be- lieve the testing
methods we employed were appropriate for acquiring maximal
muscle activation. The subjects were allowed to practice the MVIC
data collection methods until they felt condent in performing
them. Subjects stood in front of a stationary pole with their feet
shoul- der width apart. Resistance to movement was provided by an
immobi- lization strap placed around both ankles. Subjects were
instructed to push out against the immobile strap, attempting to
abduct their leg, as hard as they could for the entire 3-second trial.
MVIC data was collected for the non-pushing limb during each trial.
They were instructed to keep their toes
pointed forward to
primarily challenge the gluteal medius musculature and to avoid
leaning to the side during each trial. Failure to do so resulted in the
trial being discarded and repeated. They were allowed to maintain
balance, but were instructed not to lean into the pole. This
procedure was repeated for the other leg.

Age (yr)

Height (cm)

Mass (kg)

17
13

50.8 (1.4)
51.1 (2.3)

173.1 (2.5)
178.2 (4.3)

77.3 (3.8)
84.2 (6.8)

Data are presented as mean

(SEM). CON = control group.
OA = hip osteoarthritis
group. yr = year.

attempt to control velocity of movement (Brophy et al., 2010) (Fig.

1). each trial, the subject stood on the force plate behind an 8For
inch box. All subjects performed the step tasks using the
high
uninvolved
cm = centimeters. kg = kilograms.

75

limb rst. For step up,

subjects were instructed
to step up onto the box
using the uninvolved limb

M.K.
M.K.
Dwyer
Dwyer
et al.
et /al.Clinical
/ Clinical
Biomechanics
Biomechanics
28 (2013)
28 (2013)
757
757
761761
and then bring their

75

involved limb up onto the box. For step down,

subjects were instructed to step forward off the box onto the force
plate, landing as softly as possible, leading with the involved limb,
followed by the uninvolved limb, and come to

Fig. 1. (A) Step up and (B) step down.

a quiet static stance on the force plate. For involved limb trials, the
process was repeated with limbs in reverse order. If subjects were
unable to complete either task without assistance, the data
was discarded and the trial was repeated.
2.4.2. Gait
Subjects stood at the end of the force plate in a two-legged
stance position. They were instructed to walk across the force plate
at a self- selected natural walking speed while keeping each leg on
either side of the center line. If the subject did not walk the entire
distance of the force plate or crossed the center line, the data was
discarded and the trial repeated.
2.5. Data
analysis

processing

and

For SEMG data collected during MVIC testing, raw signals

obtained during the 3-second trials were band passed ltered
from 20 to
500 ms and full wave rectied using Datapac software. Each trial
was analyzed by dividing the data into 500 ms windows, with each
window overlapping by 100 ms. The 500 ms window with the
highest mean amplitude represented 100% activation. The peak
mean amplitude of the three trials for each muscle was used for
normalization.
For SEMG data obtained during the three tests, raw SEMG
signals were band passed ltered at 20 to 500 Hz, stored on a
personal comput- er, and analyzed using the Datapac software. In
order to determine the appropriate data smoothing parameters, the
delity of the muscle am- plitude after signal smoothing was
evaluated using time constants from 5 ms to 50 ms at 5 ms time
increments. Based on the results of this analysis, ltered SEMG
signals were processed using root-mean squared smoothing with a
30 ms time constant. Data was normalized to the 100% peak mean
amplitude measured during MVIC testing to allow for comparison
between subjects.
For analysis purposes, step up and step down were dened using
the footswitch. Step up was dened from the time of initial
stepping movement of the lead limb to the time when the trail limb
made con- tact with the box. Step down was dened from the time
the trail limb removed contact with the box to the time when the
lead limb came into contact with force plate. For level gait, the
stance and swing phases were dened for both the involved and
uninvolved limb using the foot contact and toe off data from the

footswitch. The stance phase was dened as the time from heel
strike to foot off,

while the swing phase was dened as the time from toe off to heel
strike. Mean root mean squared muscle amplitudes for both the involved and uninvolved limb gluteus medius muscles were averaged
across the three trials each for step up and step down tasks initiated
by both limbs, as well as the stance and swing phases of gait.
Impact force data during step up and step down were averaged
across the three trials for both the involved and uninvolved limb
trials. The dependent variable for step up was lift-up index,
quantied as the average maximal concentric force exerted by the
initial ascending limb (percentage of body weight). The dependent
variable for step down was impact index, quantied as the
maximal vertical impact force of the
initial descending limb
(percentage of body weight).
IBM SPSS v.19.0 (IBM Corporation, New York, USA) was used to
an- alyze all data. All dependent variables were examined for
deviation from normality using the KolmogorovSmirnov and Levene
tests. To compare data between the control group and OA group,
separate inde- pendent t-tests were conducted for normally
distributed dependant variables, and MannWhitney U tests were
conducted for those vari- ables
which were not normally
distributed. Mean difference and 95% condence intervals were
calculated for all comparisons and presented along with the P-value in

the results section (mean difference [95% con- dence intervals]; P

value). To assess the association between average SEMG amplitude
of the gluteus medius muscles and impact forces dur- ing stepping
tasks, separate Pearson Product Moment Correlations were conducted
for both the control and OA group. Level of statistical signif- icance
was set a priori at P 0.05.
3. Results
A total of 30 subjects participated in this study. Subject demographics for each group are presented in Table 1. Data are
presented as mean and standard error of measurement. There were
no signicant differences between the two groups for subject age (P
= 0.92), height (P = 0.35), or mass (P = 0.29). Participants with
hip OA reported higher pain scores (VAS = 6.18 (0.6) cm) than
the control group (VAS = 0.09 (0.02) cm; P b 0.001).
3.1. Step tasks
Tables 2 and 3 summarize mean root mean squared muscle
ampli- tude (%MVIC) during all trials for step up and step down for the
involved

M.K.
M.K.
Dwyer
Dwyer
et al.
et /al.Clinical
/ Clinical
Biomechanics
Biomechanics
28 (2013)
28 (2013)
757
757
761761

76

Table 2
Average root-mean squared amplitude of the both gluteus medius muscles during
stair ascent represented as a percentage of maximal voluntary isometric contraction.
Group

Step upinvolved (%)

GMI

GMU

Table 4
Mean lift-up index of stair ascent and impact index of stair descent represented as
a percentage of body weight.
Lift-up index (%)

Step upuninvolved (%)

GMI

CON
29.4 (2.4)
21.8 (2.0)
21.8 (1.3)
a
a,b
b
(2.5) OA
43.2 (5.0)
45.2 (5.1)
31.0 (4.0)
a
39.5 (3.8)
Data are presented as mean
(SEM). CON = control group.
OA = hip osteoarthritis group.
GMI = involved limb gluteus medius muscle.
GMU = uninvolved limb gluteus medius
muscle.
4. Discussion
a
Indicates a signicant group difference (P b 0.05).

GMU
28.9

76

CON
OA

Impact index (%)

Involved

Uninvolved

Involved

Uninvolved

38.7 (2.7)
a
49.8 (4.6)

34.9 (1.8)
30.7 (3.1)

43.8 (3.8)
51.8 (4.6)

52.1 (3.9)
a
38.9 (3.2)

Data are presented as mean

(SEM). CON = control group.
OA = hip osteoarthritis group.
a
Indicates a signicant group difference (P b 0.05).

and uninvolved limb gluteus medius muscles. Increased gluteus

medius muscle SEMG amplitudes were observed in participants with
hip OA for the involved limb muscle during step up initiated
with that limb (+ 13.7% [3.1%, 24.4%]; P = 0.025) and for the
uninvolved limb muscle during step up initiated with both the
involved (+ 23.4% [13.1%, 33.8%]; P b 0.001) and uninvolved (+
10.6% [4.4%, 21.5%]; P = 0.027) limb when compared to the
control
group. Similarly,
increased
gluteus medius SEMG
amplitudes were observed in participants with hip OA for the
involved limb muscle during step down initiated with the involved (+ 4.0% [9.4%, 25.7%]; P b 0.001) and uninvolved (+
12.2% [2.8%, 21.6%]; P = 0.018) limb and for the uninvolved limb
muscle dur- ing step down initiated with both the involved (+ 12.9%
[4.4%, 21.5%]; P = 0.004) and uninvolved (+ 11.2% [1.6%, 20.7%]; P
= 0.007) limb when compared to the control group.
Table 4 summarizes mean lift-up index (step up) and impact
index (step down) during all trials. Participants with hip OA
exhibited signif- icantly higher lift-up index during step up initiated
with the involved limb (+ 11.2% [0.9%, 21.5%]; P = 0.035) and lower
impact index during step down initiated with the uninvolved limb
( 13.3% [2.6%, 24.0%]; P = 0.017) when compared to the control
group. For participants with hip OA, impact index was inversely
correlated (r = 0.639; P = 0.019) with SEMG amplitude of the
involved limb gluteus medius muscle during step down initiated
with the involved limb (Fig. 2).
3.2. Gait
Table
5
summarizes mean root mean squared muscle
amplitude (%MVIC) for the dominant and non-dominant gluteus
medius muscles during the stance and swing phases of gait.
Participants with hip OA exhibited signicantly higher SEMG muscle
amplitudes for the domi- nant gluteus medius muscle during both
the stance (+ 15.9% [2.6%,
29.2%]; P = 0.022) and swing (+ 14.6% [4.2%, 24.9%]; P =
0.001) phases of gait when compared to controls. Participants with
hip OA exhibited signicantly higher SEMG muscle amplitudes for
the non- dominant gluteus medius muscle during only the
stance phase (+ 9.9% [2.0%, 17.8%]; P = 0.008) of gait when
compared to controls.

The purpose of this study was to assess the muscle activation

levels of the gluteus medius muscles during gait and step tasks for
subjects with end-stage hip OA compared to healthy controls. The
results sup- port our hypothesis and demonstrate that subjects with
end-stage OA of the hip joint demonstrate higher mean gluteus
medius muscle SEMG amplitudes for both limbs compared to
healthy controls during gait and step tasks. The increase in mean
muscle amplitudes during function in patients with end-stage OA
may be indicative of the hip ab- ductor muscles recruiting a greater
number of motor units or recruiting the available motor units at a
greater extent than matched controls to accomplish the same
functional task (Ling et al., 2007). A heightened neural drive of the
gluteus medius muscle was found to be the only bio- mechanical
factor which differed between subjects with diminished hip abductor
muscle strength and controls during a landing task (Homan et al.,
2012). Theoretically, in the presence of pathology, the muscle has a
diminished capacity to generate a given force (Edgerton et al.,
1996). As reductions in the cross-sectional area and force-producing
ca- pabilities of the gluteus medius muscle have been previously
reported for subjects with hip osteoarthritis (Arokoski et al., 2002), we
can spec- ulate that the increased amplitudes observed in this study
may be the result of alterations to the anatomical properties of the
muscle as well as muscle weakness.
While isolated strength of the hip abductor muscles was not
mea- sured, lower extremity functional strength during step up and
step down was assessed via the force plate. During step up initiated
by the involved limb, the lift-up index was greater for the OA group
compared to controls, which indicates that the OA group using their
pathological limb applied a greater percentage of their body weight
to the step. These results demonstrate that, for patients with hip OA,
altered move- ment patterns are present during stair ascent;
however, without accom- panying strength or kinematic data,
interpretation of the higher forces applied to the involved limb of
patients with hip OA remains speculative and requires further
exploration. During step down initiated with the uninvolved limb, the
impact index was reduced for the OA group com- pared to the
control group. These results suggest that the subjects with hip OA
controlled the amount of body weight applied to the painful joint
when stepping down to protect the pathological side from incur- ring
high forces. An inverse relationship was observed between impact

Table 3
Average root-mean squared amplitude of both gluteus medius muscles during stair
de- scent represented as a percentage of maximal voluntary isometric contraction.
Group

Step downinvolved (%)

GMI

GMU

Step downuninvolved (%)

GMI

CON
19.9 (1.7)
23.2 (2.1)
21.4 (1.5)
a,b
a,b
b
OA
37.4 (4.0)
36.1 (3.9)
33.7 (4.1)
a,b
(4.5)
Data are presented as mean
(SEM). CON = control group.
OA = hip osteoarthritis group.
GMI = involved limb gluteus medius muscle.
GMU = uninvolved limb gluteus medius
muscle.
a
Indicates a signicant
group difference (P b 0.05).
MannWhitney
U test.

GMU
22.1 (4.5)
33.2

Fig. 2. Scatter plot for impact index (y axis) and average SEMG muscle amplitude (x
axis).

Table 5
Average root-mean squared amplitude of both gluteus medius muscles during the
stance and swing phases of gait represented as a percentage of maximal voluntary
isometric contraction.
Stance phase (%)
GMI

Swing phase (%)

GMU

GMI

CON
18.2 (1.4)
20.4 (2.2)
16.1 (1.5)
a,b
a,b
a,b
(2.2) OA
34.1 (7.3)
30.3 (3.3)
30.6 (5.5)
21.0 (2.3)
Data are presented as mean
(SEM). CON = control group.
OA = hip osteoarthritis group.
GMI = involved limb gluteus medius muscle.
GMU = uninvolved limb gluteus medius
muscle.
b a Indicates a signicant group difference (P b 0.05).
Mann Whitney U test.

GMU
18.9

index and involved limb gluteus medius SEMG muscle amplitude

dur- ing step down initiated with the involved limb. These results
suggest that greater muscle activation was associated with reduced
impact forces. As this relationship was only present in participants
with hip OA, this could suggest the presence of muscle weakness in
the patholog- ical limb.
Our ndings have clinical signicance as hip abductor muscle
force has been identied as a primary contributor to increased hip joint
contact forces (Schache et al., 2003). For individuals with hip joint
disease, in- creases in the forces applied to the joint structures may
have the potential to progress the disease. Given that the alterations in
the force-producing capabilities of the hip abductor muscles occur
bilaterally for patients reporting with unilateral hip disease,
addressing these impairments may be important to preserve the
health of both hip joints. However, the relationship between muscle
weakness, joint loading, and disease progression requires further
study to make denitive conclusions.
4.1.
limitations

Study

The cross sectional design of our study was a limitation of our

study. Future research should examine alterations associated with
joint dis- ease in a prospective design. In addition, it is possible that
MVIC testing performed during our study was inuenced by pain. All
subjects report- ed a moderate to severe level of daily pain as a
result of their OA. Pain may play a factor in the ability of patients
with hip joint OA to produce a maximum contraction; therefore their
values may not give adequate representation of a maximum
contraction. This was also a reason why we chose not to collect
strength data for our subjects, which we recog- nize makes the
interpretation of our results difcult. Pain may have also affected
ability of the subjects with OA to perform the stepping and
walking tasks. Thirdly, the traditional procedure for MVIC testing
of the gluteus medius muscle is a resisted hip abduction performed
in a side-lying position (Bolgla and Uhl, 2007). Because the patients
with hip OA could often not lie on their side due to pain, MVICs
were mea- sured during isometric weight-bearing hip abduction.
Lastly, we chose to match the surgical limb of our patients to the
dominant limb of our control subjects. We did not inquire about limb
dominance of our pa- tient group, thus we could not determine if the
surgical limb was also the dominant limb. This may have resulted in
comparing some non- dominant limbs of our patients to the
dominant limb of our control sub- jects, and we recognize this may
have affected our results.
5. Conclusions
Based on the results of this study, signicantly greater gluteus
medius muscle amplitudes existed bilaterally during gait and step

tasks for patients with end-stage hip joint OA compared to healthy

controls. These results suggest that the gluteus medius muscles are
being recruited to a greater degree to carry out the same level of
function of age and sex matched controls. The increased activation
indicates signicantly greater muscular recruitment for completion
of simple daily tasks.
Acknowledgments
This study was nancially supported through University of
Kentucky Department of Orthopaedics grant. The sponsors had no
involvement in the study design, in the collection, analysis, or
interpretation of data, in the writing of the manuscript, or in the
decision to submit the manu- script for publication.
References
Arokoski, M.H., Arokoski, J.P., Haara, M., Kankaanpaa, M., Vesterinen, M., Niemitukia,
L.H., et al., 2002. Hip muscle strength and muscle cross sectional area in men
with and without hip osteoarthritis. J. Rheumatol. 29, 21852195.
Bolgla, L.A., Uhl, T.L., 2007. Reliability of electromyographic normalization methods
for evaluating the hip musculature. J. Electromyogr. Kinesiol. 17, 102111.
Brophy, R.H., Backus, S., Kraszewski, A.P., Steele, B.C., Ma, Y., Osei, D., et al., 2010.
Differ- ences between sexes in lower extremity alignment and muscle activation
during soc- cer kick. J. Bone Joint Surg. (Am.) 92, 20502058.
Cram, J., Kasman, G., Holtz, J., 1998. Introduction to Surface Electromyography. Jones
and
Bartlett.
Dekker, J., Van Dijk, G., Veenhof, C., 2009. Risk factors for functional decline in
osteoarthri- tis of the hip or knee. Curr. Opin. Rhematology 21, 520524.
Edgerton, V.R., Wolf, S.L., Levendowski, D.J., Roy, R.R., 1996. Theoreti c al basis for patte rning EMG amp l itude s to assess mu sc le dysf unc ti on. Me d. Sci. Spor ts Exer c . 28,
744751.
Felson, D.T., Zhang, Y., 1998. An update on the epidemiology of knee and hip
osteoarthri- tis with a view to prevention. Arthritis & Rhematism 41, 13431355.
Gottschalk, F., Kourosh, S., Leveau, B., 1989. The functional anatomy of tensor fasciae
latae and gluteus medius and minimus. J. Anat. 166, 179189.
Hart, J.M., Pietrosimone, B., Hertel, J., Ingersoll, C.D., 2010. Quadriceps activation
following knee injuries: a systematic review. J. Athl. Train. 45, 8797.
Herzog, W., Longino, D., Clark, A., 2003. The role of muscles in joint adaptation and
degen- eration. Langenbeck's Arch. Surg. 388, 305315.
Homan, K., Norcross, M., Goerger, B., Prentice, W., Blackburn, J., 2012. The inuence of
hip strength on gluteal activity and lower extremity kinematics. J. Electromyogr.
Kinesiol. http://dx.doi.org/10.1016/j.jelekin.2012.11.009.
Inman, V.T., 1947. Functional aspects of the abductor muscles of the hip. J. Bone Joint
Surg.
Br. 29, 607619.
Ling, S.M., Conwit, R.A., Talbot, L., Shermack, M., Wood, J.E., Dredge, E.M., et al., 2007.
Elec- tromyographic patterns suggest changes in
motor unit physiology
associated with early osteoarthritis of the knee. Osteoarthr. Cartil. 15, 11341140.
Long, W., Dorr, L., Healy, B., Perry, J., 1993. Functional recovery of noncemented total
hip arthroplasty. Clin. Orthop. Relat. Res. 288, 7377.
Neumann, D.A., 2010. Kinesiology of the hip: a focus on muscular actions. J. Orthop.
Sports
Phys. Ther. 40, 8294.
Pustoc'h, A., Cheze, L., 2009. Normal and osteoarthritic hip joint mechanical
behaviour: a comparison study. Med. Biol. Eng. Compu. 47, 375383.
Rasch, A., Bystrm, A.H., Daln, N., Martinez-Carranza, N., Berg, H.E., 2009. Persi st ing
mu scle atro phy two yea r s after re pla c emen t of th e hip. J. Bone Joi nt Su rg. (Br.)
91, 583588.
Schache, A.G., Blanch, P., Rath, D., Wrigley, T., Bennell, K., 2003. Differences between
the sexes in the three-dimensional angular rotations of the lumbopelvichip
complex during treadmill running. J. Sports Sci. 21, 105118.
Shindle, M.K., Voos, J.E., Nho, S.J., Heyworth, B.E., Kelly, B.T., 2008. Arthroscopic
manage- ment of labral tears in the hip. J. Bone Joint Surg. Am. 90-A, 219.
Shrier, I., 2004. Muscle dysfunction versus wear and tear as a cause of exercise related
os- teoarthritis: an epidemiological update. Br. J. Sports Med. 38, 526535.
Sims, K., Richardson, C., Brauer, S., 2002. Investigation of hip abductor activation in
sub- jects with clinical unilateral hip osteoarthritis. Ann. Rheum. Dis. 61, 687692.
Vaz, M., Kramer, J., Rorabeck, C., Bourne, R., 1993. Isometric hip abductor strength
follow- ing total hip replacement and its relationship to functional assessments.
J. Orthop. Sports Phys. Ther. 18, 526531.
Ward, S.R., Winters, T.M., Blemker, S.S., 2010. The arch itectu r al d es ign of th e gluteal
mu scle gr oup : imp lic ati ons for mov emen t and rehabilitati on. J. O rth op. Sp orts
Phys. Th er . 40, 95 102.
Watanabe, H., Shimada, Y., Sato, K., Tsutsumi, Y., Sato, M., 1998. Gait analysis before
or after varus osteotomy of the femur for hip osteoarthritis. Biomed. Mater.
Eng. 8,
177186.