GSPM_27211.

fm Page 1 Thursday, July 29, 2004 9:01 PM

Research in Sports Medicine 12: 1–9, 2004

Copyright © Taylor & Francis Inc.
ISSN:1543-8627 print / 1543-8635 online
DOI:10.1080/15438620490497503

Muscle Activation Pattern of Hip

Research
12:3Taylor
Taylor
GSPM
1
2004
10.1080/15438620490497503
27211
L.
Muscle
5 Vogt&Activation
etin
Francis
al.
&
Sports
FrancisTaylor
Inc.
Pattern
Medicineofand
HipFrancis
Arthroplasty
325 Chestnut
PatientsStreetPhiladelphiaPA191061543-86271543-8635
in Walking

Arthroplasty Patients in Walking

L. VOGT PhD, AND W. BANZER MD, PhD

Johann Wolfgang Goethe-University Frankfurt/Main,
Dept. of Sports Medicine, Germany

K. PFEIFER PhD
Otto-von-Guericke University Magdeburg, Training and Health,
Germany

R. GALM MD
Wirbelsäulenklinik Bad Homburg v.d.H., Germany

This study examined the hip abductor activation pattern of 14 hip replacement
patients and 10 age-matched healthy controls by measuring surface elec-
tromyography (EMG) onset and cessation times. Stride characteristics, surface
EMG from bilateral gluteus medius, and 3D pelvis kinematics were evalu-
ated during treadmill ambulation. EMG onset times were normalized with
regard to the individual stride time for each gait cycle. An ANOVA revealed
significantly delayed EMG onset times (p < .001) in comparing hip abduc-
tors of the operated side with the unimpaired side and the healthy controls.
Between subject effects also demonstrated significant differences (p < .01)
for stance duration and sagittal pelvis range of motion. No significant differ-
ences were found for EMG cessation times and angular pelvis peak-to-peak
ranges in the frontal and transverse planes. The results indicated deficien-
cies in the hip abductor recruitment pattern of hip arthroplasty patients.
Further analysis should explore whether specific exercises, or rehabilitation
programs can facilitate adequate muscle activation.

The authors gratefully acknowledge the work and assistance of K. Brettmann and
C. Allerlei during this research project.
Address correspondence to Lutz Vogt, Ph.D., Johann Wolfgang Goethe-University,
Institute for Sport Sciences, Dept. of Sports Medicine, Ginnheimer Landstrasse 39, 60487
Frankfurt/Main, Germany. E-mail: L.Vogt@sport.uni-frankfurt.de.

1
GSPM_27211.fm Page 2 Thursday, July 29, 2004 9:01 PM

2 L. Vogt et al.

Keywords hip arthroplasty, clinical gait analysis, hip abductor, recruitment

pattern, electromyography

Introduction
Following total hip replacement, the emphasis in rehabilitation lies on the optim-
ization of independence in activities of daily living. Therefore hip abductor
strength training is a standard recommendation after total hip arthroplasty
(Horstmann, Martini, Mayer, et al. 1995; Vaz, Kramer, Rorabeck, et al. 1993).
However, studies have demonstrated continued weakness on the side of the oper-
ated hip up to several months after surgery (Shih, Du, and Lin 1994). Despite
concern about the importance of hip abductor strength training to improve
patients’s gait performance, ambulatory independence and pelvis control only
few studies have analyzed the neuromuscular activation of hip surrounding mus-
cles (Long, Dorr, Healy, et al. 1993; Perron, Malouin, Moffet, et al. 2000). No
study has clearly concentrated on the contractile pattern of the hip abductor mus-
cles in walking and their control function for pelvis balance in the frontal plane.
For this reason the present study intends to provide a more detailed look at
the hip abductor recruitment pattern of patients with a single hip arthroplasty by
measuring surface EMG onset and cessation times.

Methods
Fourteen total hip replacement (THR) patients (9 m., 5 f.) with a mean age of 63
(56–72) years and 10 age-matched healthy volunteers (Ctrl; 7 m., 3 f.; average
age 61 [47–71] years) participated in the investigation after the study was
approved by the university’s ethics committee. Each subject of the control group
was asymptomatic and denied any previous neurological, muscular, or skeletal
disability. Hip arthroplasty subjects were 31 to 46 days postsurgery (average
41 days) and had completed their prescribed 3 week inpatient rehabilitation regi-
men before testing. The rehabilitation program included active stretching and
strengthening exercises for the involved muscles, functional activities, ambula-
tion, and stair climbing. The therapeutic exercises were initiated at the first post-
operative day and carried out 5 days per week until rehabilitation discharge.
Patients with bilateral hip arthroplasty, vascular insufficiency, or systemic
problems (e.g. cancerous, cardiovascular, or endocrinologic diseases) as well as
patients with concomitant ipsilateral knee, ankle, or foot problems were excluded
from the study. Left and right leg length—measured as the distance between the
superior anterior spine of the iliac bone and the medial malleolus—was not
allowed to show more than 1cm inequality. All patients were rigorously screened
and inducted into the study by one clinician skilled in the evaluation of hip dis-
eases. The indication for surgery was degenerative arthritis of the hip. The hip
arthroplasty surgeries were performed at three different hospitals. All hips were
GSPM_27211.fm Page 3 Thursday, July 29, 2004 9:01 PM

Muscle Activation Pattern of Hip Arthroplasty Patients in Walking 3

operated through a posterolateral approach. At the time of testing, all patients

were able to ambulate without gait assistive devices in their everyday life.
Postoperative pain, stiffness, and function during routine activities were
assessed by the Western Ontario and McMaster Universities Osteoarthritis Index
(WOMAC). Reliability and validity of this condition-specific instrument, in
which lower scores are better, has been extensively described (Bellamy, Buchanan,
Goldsmith, et al. 1988; Stucki, Meier, Stucki, et al. 1996). Postoperative gait ana-
lysis evaluated stride characteristics, surface EMG from bilateral gluteus medius,
and 3D pelvis kinematics. Bipolar (Ag/AgCl) surface electrodes (BlueSensor®)
with an interelectrode distance of 20mm were used to sample EMG activity dur-
ing treadmill ambulation (Woodway® Reha ES2, Germany) at self-selected
walking speeds (2.4–4.5 km/h). The reference electrode was attached to the sub-
jects’ posterior superior iliac spine. The skin of the recording site was prepared
by shaving, sanding, and rubbing with gauze saturated with alcohol. All electrode
cables were secured with tape to reduce any possibility of artifact produced by
cable movement. Prior to testing the subjects had time to practice treadmill walk-
ing until they reported they had gotten used to the walking conditions. The tread-
mill habituation time ranged from 4 to 10 min. The EMG activity was recorded
by a multichannel EMG datalogger system (BIOVISION®, Germany) operating
at 1000 Hz. A minimum of 20 successful walking cycles formed the basis for all
subsequent analysis. Muscle onset/cessation was considered to have occurred
when 25 consecutive data points of a sliding window exceeded the current mean
baseline by three-standard deviations. EMG onset and cessation times were nor-
malized with regard to the individual stride time (% stride) for each gait cycle
(Vogt, Pfeifer, and Banzer 2003) and ensembled averaged. The different phases
of the gait cycle were registered by four pressure-sensitive footswitches. The
ZEBRIS CMS70® ultrasound movement analysis system, with an absolute accu-
racy better than 0.6 mm (Himmelreich, Stefanicki, and Bunzer 1998), was used
for movement data acquisition. Sufficient test–retest reliability of the meas-
urement system has been demonstrated for gait analysis and gross motions of the
trunk (Vogt and Banzer 1997; Portscher, Vogt, Pheifer, et al. 2000). Three ultra-
sound microphones, determining a local cartesian coordinate system, were used
to track (50Hz) a small lightweight T-plate with three ultrasonic markers attached
to the posterior midline of the sacrum (S1). Angles were defined as cardanic
angles of a single body segment. The three coordinate axes were defined using
the three markers on the rigid T-plate. The medial–lateral axis was orientated as a
line passing through the two markers on the lateral ends of the T-plates’ horizon-
tal bar with its positive direction to the left. The vertical axis was constructed to
lie in the plane formed by all three markers of the rigid T-plate, orthogonal to the
medial-lateral axis with its positive direction upward. The body embedded ante-
rior–posterior axis was perpendicular to both the medial–lateral and the vertical
axes with its positive direction forward. Prior to testing, subjects were asked to
maintain a relaxed upright posture with surface markers visible to the ultrasound
GSPM_27211.fm Page 4 Thursday, July 29, 2004 9:01 PM

4 L. Vogt et al.

microphones. The anatomical orientation of the pelvis was defined as zero for
computing relative angular displacement during the movement. The obtained net
angular displacements were low-pass filtered (2nd order, critical damped, double
pass) using a cutoff frequency of 6 Hz, normalized with regard to the stride time
(one stride corresponds to 100%) and ensembled averaged at intervals of 1% of
the gait cycle, to give the mean and standard deviation for that subject. Data were
processed using LabView (National Instruments®) and ALEA® Solution soft-
ware (Switzerland). Levene’s test for homogenity of variance and univariate
two-factor (group × body side) ANOVA for repeated measures were selected to
determine significant differences in stance durations and EMG measurements.
Scheffé test was used for post-hoc analysis of significant findings. Mean rotation
amplitudes of the pelvis were compared using Student t-tests. P < 0.05 was
regarded as significant.

Results
All patients walked without walking aids and were able to ambulate on the tread-
mill while not holding onto the handrail for support. The average WOMAC
Osteoarthritis Index was 8.52 (range: 0–17.25) for the operated side and 0.73
(range: 0–7.35) for the nonoperated limb on the day of investigation.
In comparing the hip abductors of the operated side with the nonoperative
side and the healthy controls repeated measures ANOVA revealed a significantly
(p < .001) delayed EMG onset on the operative side (Table 1). Thus, due to the
extremely shortened preactivation phase of the gluteus medius muscle of the
replaced hip, the electrical onset occurred almost simultaneously with the ipsilat-
eral heel contact (Figure 1). Additionally the analysis indicated significant differ-
ences (p < .01) for stance duration (THR: operated 58.1 +/− 6.8%; nonoperated
65.2 +/− 3.9% vs. Ctrl: right 64.8 +/− 2.7%; left 64.6 +/− 2.8%) and a increased
pelvis range of motion in the sagittal plane (p < .01) after hip arthroplasty. There
were no significant differences found for EMG cessation times and angular pel-
vis peak-to-peak ranges in the frontal and transverse planes (Table 1). No signifi-
cant side differences in gait cycle durations were detected. Bivariate correlation
analysis revealed a negative linear association (r = –.761; p < .05) between the
delayed onset in hip abductor activity and the reduced stance time duration.

Discussion
The present findings, obtained at rehabilitation discharge, indicated that patients
who had undergone a total hip replacement operation, following hip osteoarthri-
tis, continue to walk with particular disorders even though they received a guided
3-week inpatient exercise program. The determined changes in stride character-
istics are consistent with those of other authors (James, Nicol, and Hamblen
1994; McCrory, White, and Lifeso 2001), who also found that hip arthroplasty
 Table 1
Average Range of Motion of Angular Sacrum Movements and EMG Onset
and Cessation Times for THR and Control Subjects

EMG, onset/cessation [% gait cycle]

GSPM_27211.fm Page 5 Thursday, July 29, 2004 9:01 PM

Sacrum, range of motion [degree] onset cessation

frontal sagittal transverse operated/left non-op/right operated/left non-op/right

THR 3.3+ /− 1.7 *4.9 +/− 1.4 4.4 +/− 1.6 *1.1(≈101.1) +/− 4.2 96.9 +/− 2.2 23.8 +/− 9.4 25.8 +/− 8.4
Ctrl 4.5 +/− 1.3 3.3 +/− 0.4 3.9 +/− 1.9 95.6 +/− 3.8 95.3 +/− 2.2 28.9 +/− 9.8 21.9 +/− 5.9

Note: (Mean, 1SD) (** p < .01; *p < .001).

5
GSPM_27211.fm Page 6 Thursday, July 29, 2004 9:01 PM

6 L. Vogt et al.

bilateral gluteus medius

heel heel
contact contact
0-100% duration of muscle
gait cycle activity

operated
[s]
side

[µV]

unoperated [s]
side

0-100%
gait cycle
heel heel
contact contact

Figure 1. Characteristic raw EMG’s (operated and unoperated side) of one hip arthro-
plasty patient during iterated walking trials.

patients walk with a considerable asymmetry of stance times. At this early post-
operative stages the majority of patients had shorter stance periods on the oper-
ated limb. However, studies (Murray, Brewer, and Zuege, 1972) revealed
continued postoperative improvement where the stance times for the two limbs
become more equal after the third and sixth postoperative months (Murray, Gore,
Brewer, et al., 1979).
The results showed that the calculated reference means and standard devia-
tions of pelvis movement amplitudes meet normal reference limits, reported in
other studies (Vachalathiti, and Smith Thurston 1982; Crosbie, 1997). The
patients’ increased anterior–posterior pelvic tilting, seen in the current investiga-
tion, might compensate for movement reductions in the hip joint because of post-
operative hip flexion contractures, or due to the failure to correct the antalgic
preoperative gait pattern. These speculations are supported by Jacobs, Shorecki,
and Charnley (1972), who also noted that patients produced larger movements of
the pelvis helping to minimize movement of the hip joint.
The postoperative analysis of EMG peak characteristics and the linear rela-
tionship between stance time and EMG onset provided preliminary evidence that
hip arthroplasty patient’s gait patterns are characterized by some deficiencies in
the hip abductor recruitment. These findings are in agreement with data of Franz
and coworkers (2002), who described later preactivation of the gluteal muscles
during walkway ambulation. Even though the patients in the present research
GSPM_27211.fm Page 7 Thursday, July 29, 2004 9:01 PM

Muscle Activation Pattern of Hip Arthroplasty Patients in Walking 7

experienced only moderate pain intensities at the time of testing, their hip abduc-
tor activation pattern was marked by a significant delay probably adopted when
these subjects had intense and usually constant pain in the hip joint before sur-
gery. Because it is assumed that muscular dysfunction around the hip joint is sec-
ondary to pathologic hip pain (Andersson, Kammerere, and Greer 1979; Shih,
Du, and Lin 1994), the delayed recruitment strategy of the hip abductors could be
interpreted as a preoperative functional adaptation of the neuromuscular system
to decrease joint pressure and to prevent additional pain.
Although it is likely that a preoperative gait habit is carried on after joint
arthroplasty, there are other possible interrelated causes for the subnormal abduc-
tor function. Among these are surgical trauma because of stretching and retrac-
tion force inherent in the surgical approach; injury to its innervation (Baker and
Bitounis 1989); and decreased mechanical and physiological advantage of the hip
muscles, which may result from bony leverage changes or from shortened dis-
tances between the origin and insertion of the hip muscles.
Another possible explanation for the altered muscle function may be that
patients’ proprioception in the hip joint was affected (Grigg, Finerman, and Riley
1973; Skinner 1993). This complex process of peripheral feedback and afferent
information is likely to play an important role in the control of locomotion. Fol-
lowing suggestions from previous investigations (Skinner 1993), receiving less
or possibly inappropriate spatial, temporal, or kinetic information might impair
the precise control of the timing of events in the gait cycle.
On the other hand, it is worth noting that hip abductor dysfunction might be
more of a problem in the anterolateral or Hardinge approach (Fagerson 1998),
which can involve more disruption to gluteus medius.
Regardless of whether the mentioned changes could be attributed to preoper-
ative adaptations, surgical trauma, neural impairment, inadequate or diminished
joint proprioception, altered muscle activity could change the hip joint forces and
lead to joint instability or a loss of balance (MacKinnon and Winter 1993), prob-
ably increasing the risk of injury and falls. Moreover atypical gait patterns could
have an impact on the physiological loading of adjacent structures and may con-
tribute to arthritic changes at the contralateral hip level.
Long-term investigations should evaluate if the EMG onset differences
between limbs will diminish with increasing stance time symmetry (Murray et al.
1979). Studies also should overcome one limitation of the present research by
additionally measuring muscle strength.

Conclusions
Although patients with THR can make significant improvements in hip abductor
strength following surgery (Kilgus, Amstutz, and Wolgin 1990; Sicard-Rosenbaum,
Light, and Behrman 2002), additional rehabilitative emphasis should be placed on
the optimization of muscle activation patterns. It could be speculated that hip
GSPM_27211.fm Page 8 Thursday, July 29, 2004 9:01 PM

8 L. Vogt et al.

arthroplasty patients may profit from specific exercises or rehabilitation pro-

grams to facilitate adequate muscle activation in different types of motor activi-
ties. Therefore, strengthening exercises for the hip abductors could be combined
with specific therapeutic interventions, as they have been shown by Bullock-
Saxton, Janda, and Bullock (1993), to influence the intensity and timing of mus-
cle activation.
Further analysis should focus on possible ways of therapeutic interventions
and explore whether specific exercises or rehabilitation programs can facilitate
adequate muscle activation.

References
Andersson L, Kammerere WS, Greer RB (1979) Risk factors in 101 total hip arthroplasties—
A medical perspective. Clinical Orthopaedics and Related Research 141:50–54.
Baker AS, Bitounis VC (1989) Abductor function after total hip replacement. Journal of
Bone and Joint Surgery, British Volume 71(1):47–50.
Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW (1988) Validation study
of WOMAC: A health status instrument for measuring clinically important patient
relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the
hip or knee. Journal of Rheumatology 15(12):1833–1840.
Bullock-Saxton JE, Janda V, Bullock MI (1993) Reflex activation of gluteal muscles in
walking. Spine 6:704–708.
Crosbie J, Vachalathiti R, Smith R (1997) Age, gender, and speed effects on spinal kine-
matics during walking. Gait and Posture 5:13–20.
Fagerson TL (1998) The Hip Handbook. Oxford: Butterworth-Heinemann.
Franz A, Stapelmann J, Schmidt J, Brüggemann GP (2002) Prospective study on gait pat-
tern changes after total hip arthroplasty. Gait and Posture 16(Suppl. 1):117–118.
Grigg P, Finerman GA, Riley LH (1973) Joint-position sense after total hip replacement.
Journal of Bone and Joint Surgery, American Volume 55(5):1016–1025.
Himmelreich H, Stefanicki E, Banzer W (1998) Die ultraschallgesteuerte anthropometrie
(UGA)—Zur entwicklung eines neuen verfahrens in der asymmetriediagnostik.
Sportverletzung Sportschaden 2:60–65.
Horstmann T, Martini F, Mayer F, Sell S, Knak J, Zacher J (1995) Kraftverhalten der hüf-
tumgreifenden muskulatur und gehfähigkeit bei patienten nach implantation einer
zementfreien hüftendoprothese. Zeitschrift für Orthopädie 133:562–567.
Jacobs NA, Skorecki J, Charnley J (1972) Analysis of the vertical component of force in
normal and pathological gait. Journal of Biomechanics 5:11–34.
James PJ, Nicol AC, Hamblen DL (1994) A comparison of gait symmetry and hip move-
ments in the assessment of patients with monarticular hip arthritis. Clinical Biome-
chanics 9(3):162–166.
Kilgus DJ, Amstutz HC, Wolgin MA, Dorey FD (1990) Joint replacement for ankylosed
hips. Journal of Bone and Joint Surgery, American Volume 57(1):45–54.
Long WT, Dorr LD, Healy B, Perry J (1993) Functional recovery of noncemented total
hip arthroplasty. Clinical Orthopaedics and Related Research 288:73–77.
MacKinnon CD, Winter DA (1993) Control of whole body balance in the frontal plane
during human walking. Journal of Biomechanics 26(6):633–644.
GSPM_27211.fm Page 9 Thursday, July 29, 2004 9:01 PM

Muscle Activation Pattern of Hip Arthroplasty Patients in Walking 9

McCrory JL, White SC, Lifeso RM (2001) Vertical ground reaction forces: objective
measures of gait following hip arthroplasty. Gait and Posture 14:104–109.
Murray MP, Brewer BJ, Zuege RC (1972) Kinesiologic measurements of functional per-
formance before and after McKee-Farrar total hip replacement. Journal of Bone and
Joint Surgery, American Volume 54(2):237–256.
Murray MP, Gore DR, Brewer BJ, Gardner GM, Sepic SB (1979) A comparison of the
functional performance of patients with Charnley and Müller total hip replacement.
Acta Orthopaedica Scandinavica 50:505–509.
Perron M, Malouin F, Moffet H, McFadyen BJ (2000) Three-dimensional gait analysis in
women with a total hip arthroplasty. Clinical Biomechanics 15:504–515.
Portscher M, Vogt L, Pfeifer K, Banzer W (2000) Reproduzierbarkeit kinemetrischer
messungen der lumbalregion in der klinischen ganganalyse. Sportverletzung Sports-
chaden 14:1–5.
Shih C, Du YK, Lin Y (1994) Muscular recovery around the hip joint after total hip arthro-
plasty. Clinical Orthopaedics and Related Research 302:115–120.
Sicard-Rosenbaum L, Light KE, Behrman AL (2002) Gait, lower extremity strength, and
self-assessed mobility after hip arthroplasty. Journal of Gerontology 57A:47–51.
Skinner HB (1993) Pathokinesiology and total joint arthroplasty. Clinical Orthopaedics
and Related Research 288:78–86.
Stucki G, Meier D, Stucki S, Michel BA, Tyndall AG, Dick W, Theiler R (1996) Evaluation
einer deutschen version des WOMAC (Western Ontario und McMaster Universities)
arthroseindex. Zeitschrift Rheumatologie 55:40–49.
Thurston AJ (1982) Repeatability studies of a television/computer system for measuring
spinal and pelvic movements. Journal of Biomedical Engineering 4:129–31.
Vaz MD, Kramer JF, Rorabeck CH, Bourne RB (1993) Isometric hip abductor strength fol-
lowing total hip replacement and its relationship to functional assessments. Journal of
Orthopaedic and Sports Physical Therapy 18(4):526–531.
Vogt L, Banzer W (1997) Reproduzierbarkeit thorakaler und lumbaler wirbelsäulenbewe-
gungen mit der 3-D-ultraschalltopometrie. Physikalische Medizin, Rehabilitations-
medizin, Kurortmedizin 7:21–25.
Vogt L, Pfeifer K, Banzer W (2003) Neuromuscular control of walking with chronic low
back pain. Manual Therapy 8(1):21–28.