http://poi.sagepub.

com/
International
Prosthetics and Orthotics
http://poi.sagepub.com/content/26/2/101
The online version of this article can be found at:

DOI: 10.1080/03093640208726632
2002 26: 101 Prosthet Orthot Int
C. Sjdahl, G-B. Jarnlo, B. Sderberg and B.M. Persson
amputees before and after special gait re-education
Kinematic and kinetic gait analysis in the sagittal plane of trans-femoral

Published by:
http://www.sagepublications.com
On behalf of:

International Society for Prosthetics and Orthotics

can be found at: Prosthetics and Orthotics International Additional services and information for

http://poi.sagepub.com/cgi/alerts Email Alerts:

http://poi.sagepub.com/subscriptions Subscriptions:
http://www.sagepub.com/journalsReprints.nav Reprints:

http://www.sagepub.com/journalsPermissions.nav Permissions:

What is This?

- Aug 1, 2002 Version of Record >>

at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
Prosthetics and Orthotics International, 2002, 26, 101-112
Kinematic and kinetic gait analysis in the sagittal plane of
trans-femoral amputees before and after special
gait re-education
C. SJDAHL*, G-B. JARNLO*, B. SDERBERG** and B. M. PERSSON***
*Department of Physical Therapy, Lund University, Sweden
**Scandinavian Orthopaedic Laboratory, Lund University Hospital, Sweden
***Department of Orthopedics, Lund University Hospital, Sweden
Abstract
A special gait-training programme, combining
a psychological therapeutic approach with
methods in physiotherapy and body awareness,
was used to re-educate nine unilateral trans-
femoral amputees. All were rehabilitated trauma
or tumour amputees with an age of 16-60 years.
They had worn prostheses for more than 18
months. The re-education aimed at integrating
the prosthesis in normal movements and
increasing body awareness. Gait was measured
before and after treatment and at 6 months
follow-up with a three-dimensional motion
analysis system. Results showed almost
normalised gait speed and increased symmetry
in the hip joints with increased muscle work on
the amputated side both immediately and at
follow-up. At follow-up there were significant
differences in almost all parameters between the
two legs of the subjects and when compared to a
reference group of 18 healthy volunteers of
similar age. Thus, the intact leg compensates for
loss of function in the amputated leg and thereby
works differently compared to the reference
group. For example, during shock absorption the
extension moment in the intact knee increased
from 0.6Nm/kg before to 1.0Nm/kg after
treatment and at follow-up compared to 0.4
Nm/kg in the reference group. The eccentric
power of quadriceps increased from 0.6w/kg
before to 1.8w/kg after treatment and 1.7w/kg at
All correspondence to be addressed to C. Sjdahl,
Scandinavian Orthopaedic Laboratory, Lund
University Hospital, Entrgatan 5, S-221 85 Lund,
Sweden. E-mail: catharina.sjodahl@swipnet.se
follow-up compared to 0.4w/kg in the reference
group. The limp of amputees is usually observed
in the frontal plane, but the authors' special
focus on the sagittal plane here illustrates gait
propulsion influences. The positive training
results remained after six months.
Introduction
The asymmetrical gait pattern of trans-femoral
amputees has been investigated from several
perspectives. Stance phase is shorter and swing
phase is longer on the prosthetic side (Boonstra
et al, 199'4) and stance phase is increased with
decreasing stump length (Jaegers et al, 1995).
Gait speed is slower compared to non-amputees
and the variability in speed was low (Boonstra et
al, 1993; Jaegers et al, 1995). The gait pattern
is often characterised by lateral trunk bending
over the amputated leg because of weak hip
abductors and lack of stabilisation from the
socket (Berger, 1992; Jaegers et al, 1995).
Vaulting by heel-rise on the healthy leg is often
used to secure foot clearance in swing of the
prosthetic side (Berger, 1992; Czerniecki, 1996).
Jaegers et al (1995) also showed atrophy of both
cut and uncut hip muscles.
Trans-femoral amputee gait differs from
normal gait because the muscles directly
controlling the knee-joint are lacking and
because of functional restrictions of the
prosthetic knee joint. Most patients were unable
to flex and extend the knee while putting weight
on the prostheses. The absence of knee flexion
during shock-absorption and at the end of
terminal stance contributed to make the amputee
101 at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
102 C. Sjodahl, G-B. Jarnlo, B. Soderberg and B. M. Persson
gait more energy consuming than normal
(Kamwendo and Oberg, 1986).
Loss of muscle function after an amputation
may result in an increased activity in the
remaining muscles. For example, Winter and
Sienko (1988) found that a considerably
increased concentric muscle activity in the hip
extensors at the early heel-strike and at mid-
stance probably was a compensation for the lack
of plantar flexors, normally needed in late stance
for propulsion.
During normal gait the knee starts flexing
before the foot leaves the ground. This flexion,
which is counteracted by quadriceps, is difficult
to achieve with artificial knee-joints. When the
foot is off the ground, quadriceps continues to
work to limit knee flexion to 65 degrees at most.
Then quadriceps starts to work concentrically to
swing the shank forward. During this swing-
phase, activity suddenly stops in quadriceps and
the shank swings forward as a pendulum.
Almost immediately the hamstrings start to work
to reduce the speed of extension and at the same
time make sure that the foot regains contact with
the ground smoothly (Murray et ah, 1966; Perry,
1992). In trans-femoral amputees the hydraulic
prosthetic knee is constructed to imitate
quadriceps' eccentric work in the beginning of
swing-phase and hamstrings' eccentric work in
the end of swing-phase (Czerniecki, 1996). The
amputated person controls the prosthetic knee
mainly using the decelerating power of the hip
extensors (Hale, 1990).
The aim of this study was to describe the
effect of a training programme of unilateral
trans-femoral amputees (Sjodahl etal, 2001) on
temporal parameters and on movements,
moment and power in the sagittal plane
simultaneously in pelvis, hip, knee and ankle
joints. The aim was also to compare the values
from the intact leg to those from the amputated
leg and from a healthy reference group.
Subjects
The inclusion criteria were as follows: 1/ age
16-60, 2/ residents within two health-care
districts of southern Sweden, 3/ amputation
performed more than two years, 4/ amputation
caused by trauma or tumour, 5/ the subjects well
fitted with their prosthesis at least 18 months
previously and finished with their rehabilitation
and 6/ all had to understand written and spoken
instructions in Swedish.
Sixteen subjects met the inclusion criteria.
Seven were excluded, age 19-51, one for
medical reasons, one for failure to participate in
treatment on a regular basis because of his work
and three died before start of treatment. Two
interrupted treatment, one because of an altered
family situation and one because his office was
moved. Five men and four women completed
the treatment and follow-up (Table 1). They had
been amputees since for a mean of 9.7 years
(range 3-27). At inclusion mean and median age
was 33 years (range 16-51). The amputation was
caused by trauma in four men and by tumour in
Table 1. Description of trans-femoral amputees (N=9) and the reference group (N=18), M= male, F= female.
Trans-femoral amputees
(N=9)
Gender
M/F
F
F
F
F
M
M
M
M
M
Mean
Median
Range
Age
(y)
16
34
49
51
40
28
24
33
23
33
34
16-51
Height
(cm)
158
160
160
167
191
186
185
190
179
175
179
158-191
Body
weight
(kg)
51
66
62
82
67
68
78
106
64
71
67
51-106
Years
of amp
(y)
5
16
4.5
27
13.5
7.5
6.5
3
4.5
9.7
6.5
3-27
Reference group
Female (N=9
Age
(y)
33
34
25
37
38
52
50
38
43
39
38
25-52
Height
(cm)
167
166
170
171
160
155
160
172
176
166
167
155-176
Body
weight
(kg)
79
54
65
82
64
54
61
59
68
65
64
54-58
Male (N=9)
Age
(y)
21
37
28
47
36
31
26
35
34
33
34
21-47
Height
(cm)
181
173
192
173
172
184
186
187
183
181
183
172-192
Body
weight
(kg)
68
73
89
68
71
78
85
93
96
80
78
68-96
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
Gait analysis in trans-femoral amputees
Table 2. Stump length, socket and prosthetic components in nine trans-femoral amputees.
103
Case
1
2
3
4
5
6
7
8
9
Stump length
ISO standard*
average
short
short
average
long
average
long
average
average
Socket
ischial containment
quadrilateral
quadrilateral
quadrilateral
quadrilateral
ischial containment
quadrilateral
quadrilateral
ischial containment
Knee
Total knee mechanical
Aqua pend pneumatic
Total knee mechanical
Total knee mechanical
Vaxjo knee hydraulic
Mauch knee hydraulic
T-Ling pneumatic
Aqua pend pneumatic
Aqua pend pneumatic
Foot +
Seattle foot
Flex foot
Flex foot
Multiflex ankle
Seattle foot
Multiaxis Vaxjo foot
Flex foot
Flex foot
Flex foot
Flex foot
short means length less than width at the base, long means length more than twice the width at the base and average
means stump length between 1-2 times the width at the base of stump (18).
+ all prosthetic feet are energy-storing.
the other. All were independent community
walkers. They were considered stable in their
adjustment and did not any longer require any
special attention from the prosthetist or
physiotherapist. Two amputees were dependent
on one stick and one on two crutches while
walking before treatment. To avoid confounding
factors all kept their same prosthesis throughout
the training and follow-up. The prostheses were
serviced as needed and only worn out parts were
replaced. No socket or alignment changes were
made during the study period. The prosthetic
components and stump lengths according to ISO
(8548-2, 1993) are shown in Table 2.
Pain in the stump was slight to moderate in all
patients and all reported low-back pain almost
every day limiting daily activities.
A convenient sample of healthy volunteers,
nine men and nine women, was chosen among
hospital employees to create a reference group
(Table 1). Both mean and median age was 36
years (range 21-52).
Method
Gait-analysis was performed by use of
VICON 370 (version 1.2, Oxford Metrics Ltd,
Oxford, UK), a three-dimensional motion
analysis system recently introduced in Sweden.
The system consisted of five 50Hz cameras with
infrared strobes, one Kistler force-plate (a
piezoelectric transducer, type 9284) and one
data-station (Pentium II, 350MHz processor)
where the information was gathered and
processed in VICON Clinical Manager (VCM,
1995).
Twenty-two special lightweight surface
markers were attached directly to the skin or the
prosthesis and placed over standardized
landmarks or corresponding spots on the
prosthesis (anterior superior iliac spine, lower
lateral third of the thigh, lateral epicondyle of
femur, lower lateral third of the calf, lateral
malleolus of fibula and between second and
third metatarsal heads and one marker between
the posterior superior iliac spines) according to
the biomechanical model of Davis et al. (1991).
On the prosthesis the knee marker was placed
over the joint centre and in all the prosthetic feet,
the ankle marker was attached to the spot
corresponding to the lateral malleolus on the
intact side. In static positions markers were also
used bilaterally over the greater trochanter and
on the posterior calcaneus at the same distance
from ground level as the forefoot markers.
Subjects wore their own normal, comfortable
walking shoes when measured before, after
treatment and at follow-up. The same shoes
were used at all three occasions.
The gait-path was 12m long with a force-plate
built in at floor level and covered with a thin
rubber carpet. Subjects were asked to walk at
self-selected comfortable gait-speed. Recordings
of five strikes on the force-plate of healthy and
amputated leg respectively were made. The most
representative graph out of the five (i.e. the
median) was chosen and the mean value from
the nine subjects was used in the calculations
and figures. Due to a technical problem the data
from the before treatment session is missing in
one participant.
All documented values are maximum values
of different phases in the gait cycles and values
for moment and power were normalised by body
weight. Internal moments were calculated and
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
104 C. Sjodahl, G-B. Jarnlo, B. Soderberg and B. M. Persson
Table 3. Median values and ranges in parameters in trans-femoral amputees before and after treatment and at follow-up.
Corresponding values for the reference group are also shown.
Parameter
Gait speed (m/s)
Median
Range
Cadence
(step/min)
Median
Range
Step length (m)
Median
Range
Single support
(% of gait cycle)
Median
Range
Double support
(% of gait cycle)
Median
Range
Amputated side
Before
1.10
0.63-1.65
95
79-105
0.73
0.55-0.95
35
30-40
22
12-34
After
1.54*
1.13-1.81
107*
95-113
0.93*
0.76-1.05
37
32-39
20
12-25
Follow-up
1.45+
1.06-1.64
103+
88-109
0.92
0.70-1.01
35
33-40
18
17-25
Intact side
Before
1.06
0.66-1.67
95
79-107
0.61
0.44-0.93
39
30-43
25
18-45
After
1.52*
1.10-1.73
103*
91-111
0.85*
0.62-1.01
40*
38-45
22*
16-28
Follow-up
1.49+
1.12-1.60
103
91-111
0.84+
0.63-0.97
40++
37-43
22+
15-27
Reference
group
1.67
0.99-1.97
137
104-156
0.71
0.56-0.86
39
35-43
20
13-26
p-values < 0.05 * before and after treatment, + before treatment and follow-up, ++ after treatment and follow-up
(Wilcoxon signed ranks test).
interpreted as muscles and ligaments
counteracting the external moments produced by
ground reaction force (Gage, 1991). The product
of joint net moment and joint angular velocity,
the net muscle power, was calculated, where
positive values represent generating power i.e.
concentric muscle power and negative values
represent absorbing power i.e. eccentric muscle
power (Gage, 1991).
The reference group was measured with the
same procedure for left and right leg. As no
significant difference was found between left
and right leg, mean values for both legs were
used in the calculations.
All subjects were interviewed, prior to the gait
test, regarding stump and low-back pain.
The study group completed a special gait-
training programme (Sjodahl et al, 2001). The
intention was to increase the function of the
amputated leg with decreased asymmetry and
compensatory mechanisms in the healthy leg.
Another aim was to make better use of the
technically advanced prosthetic components.
Ability to carry and control the body weight
during single support and hip-extension at push-
off were emphasised and therefore dynamic
changes in the sagittal plane were especially
expected.
Wilcoxon signed ranks test was used to
calculate the training effect and Mann-Whitney
U-test to analyse differences comparing the
amputated to the intact side and these two to the
reference group, respectively. A p-value <0.05
was regarded as significant.
Results
All the nine amputees had stable walking-
patterns and there was almost no sign of intra-
variations at the three separate occasions.
There was a significant increase in gait speed,
cadence and step length on both sides after
treatment, most remaining at follow-up and an
overall tendency to reduced ranges (Table 3).
There were no significant differences between
the amputated and the intact leg in these
parameters before or after treatment or at follow-
up. However, gait speed and cadence remained
significantly higher for the reference group than
for the subjects. A significantly longer step
length was found after treatment and at follow-
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
Gait analysis in trans-femoral amputees 105
Table 4. Median peak values and ranges of movement, moment and power in hip, knee and ankle related to the gait
cycle in trans-femoral amputees before and after treatment and at follow-up. Corresponding values for the reference group
are also shown.
Parameter
Hip flexion (degrees)
range
% of gait cycle
Concentric power in
hip extensors (w/kg)
range
% of gait cycle
Eccentric power in hip
flexors (w/kg)
range
% of gait cycle
Hip flexion moment
(Nm/kg)
range
% of gait cycle
Knee flexion (degrees)
range
% of gait cycle
Knee extension moment
(Nm/kg)
range
% of gait cycle
Eccentric power q-ceps
(w/kg)
range
% of gait cycle
Concentric- power in
q-ceps (w/kg)
range
% of gait cycle
Concentric power in
ankle plantar flexors
(w/kg)
range
% of gait cycle
Amputated side
Before
31
21-47
2
0.8
0.0-1.5
8
0.6
0.1-1.2
44
0.9
0.3-1.1
50
0
14
-0.3
-0.5-0.4
22
0.0
0.0-0.0
10
0.0
-0.1-0.2
20
0.2
0.0-0.7
52
After
38*
25-55
0
1.5*
0.1-4.6
6
0.9*
0.4-2.5
44
1.0*
0.7-1.8
48
0
16
-0.2
-0.6-0.3
18
0.0
-0.1-0.0
10
0.0
0.0-0.2
20
0.5*
0.3-0.8
50
Follow-up
37++
25-52
0
1.3++
0.4-3.4
8
0.9
0.6- 1.8
46
1.0
0.5-1.6
50
0
14
-0.2
-0.6-0.0
16
0.0
0.0-0.1
10
0.0
-0.1-0.1
20
0.5
0.2-0.8
52
Intact side
Before
49
42-56
4
1.2
0.1-2.2
10
0.5
0.1-1.01
50
0.9
0.3-2.0
50
33
20-35
16
0.6
0.5-0.9
14
-0.6
0.3-0.9
10
0.5
0.3-1.2
20
1.7
0.6-2.1
56
After
57*
44-68
4
1.6
0.2-3.4
12
0.7
0.2-2.1
44
1.2*
0.7-1.4
56
39
23-41
14
1.0
0.4-1.4
14
-1.8*
0.3-3.0
10
1.2*
0.6-2.3
20
2.9*
1.2-3.5
56
Follow-up
52
39-66
4
1.6
1.2-2.2
12
0.8
0.2-1.4
50
1.1
0.1-1.6
54
36
16-47
14
1.0
-0.1-0.3
14
-1.7+
0.3-3.2
10
1.0++
0.3-1.5
20
2.9+
1.2-3.4
56
Ref group
34
26-54
2
1.0
0.1-3.3
14
0.8
0.2-1.7
42
1.0
0.4-1.4
52
22
10-31
14
0.4
0.0-0.7
14
0.4
-1.0-0.0
10
0.4
-0.1-1.0
20
2.1
0.5-3.8
54
P-values < 0.05 * before and after treatment, + before treatment and follow-up, ++ after treatment and follow-up
(Wilcoxon signed ranks test).
up on both the amputated and the intact side
compared to the reference group (Tables 3
and 5).
On the amputated side the single and double
support did not change after treatment, but
showed less variability than before treatment.
Single support remained significantly shorter
than on the intact side and in the reference
group. Both values changed significantly on
the intact side and did not differ from
the reference group at follow-up (Tables 3
and 5).
Results in Tables 4 and 5 and in Figs. 1-5 are
presented as either a/ movement in degrees b/
moment in Nm/kg or c/ power in w/kg and
related to the gait cycle.
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
106 C. Sjodahl, G-B. Jarnlo, B. Soderberg and B. M. Persson
Table 5. Confidence levels for differences between the amputated and the intact side and each side to the reference
group at follow-up (Mann-Whitney U test). (Conc=concentric, ecc=eccentric)
Parameters
General
Gait speed
Cadence
Step length
Single support
Double support
Hip
Hip flexion
Cone power
in hip ext
Ecc power
in hip flex
Hip flexion
moment
Knee
Knee flexion
Knee extension
moment
Ecc power
q-ceps
Cone power
q-ceps
Ankle
Cone power
ankle
plantar
flexors
Amputated/Intact side
Pre
0.959
1.000
0.161
0.279
0.328
0.005
0.195
0.328
0.442
<0.001
<0.001
<0.001
<0.001
<0.001
Post
0.730
0.489
0.258
<0.001
0.258
0.004
1.000
0.190
0.546
<0.001
<0.001
<0.001
<0.001
<0.001
Follow
-up
0.605
0.796
0.297
0.001
0.161
0.014
0.258
0.161
1.000
<0.001
<0.001
<0.001
<0.001
<0.001
Amputated side/ Reference
Pre
<0.001
<0.001
0.622
0.002
0.179
0.709
0.360
0.622
0.075
<0.001
0.001
<0.001
<0.001
<0.001
group
Post
0.123
<0.001
<0.001
<0.001
0.459
0.172
0.146
0.116
0.665
<0.001
<0.001
<0.001
<0.001
<0.001
Follow
-up
0,010
<0.001
<0.001
<0.001
0.476
0.363
0.459
0.116
0.706
<0.001
<0.001
<0.001
<0.001
<0.001
Intact side/ Reference
Pre
<0.001
<0.001
0.169
0.665
0.008
<0.001
0.520
0.092
0.011
0.001
<0.001
0.111
0.111
0.011
group
Post
0.035
<0.001
0.005
0.092
0.425
<0.001
0.071
0.686
0.123
<0.001
<0.001
<0.001
<0.001
0.172
Follow
-up
0.016
<0.001
0.013
0.625
0.334
<0.001
0.009
0.686
0.748
0.006
0.006
<0.001
<0.001
0.138
Pelvis
The shape of the movement graph was
unchanged after treatment and at follow-up on
both sides. During stance-phase, the pelvis was
tilted forward on the amputated side and during
pre-swing it was tilted backwards. The tilting of
the pelvis was more than twice as large as in the
reference group, where it was stable at 8 degrees
(Fig. 1).
Hip
On both sides hip flexion increased
significantly after treatment. On the intact side
the hip flexion was significantly larger
compared to the amputated side and to the
reference group at follow-up (Tables 4 and 5).
The hip extensors on the amputated side
started to work earlier compared to the intact
side and to the reference group. Both sides had
larger amplitudes than the reference group at the
beginning of stance phase after treatment,
although timing was different (Figs. 2a and 2b).
Concentric power in the hip extensors almost
doubled on the amputated side after treatment to
about the same level as the intact side. At
follow-up the subjects showed no differences
here between the legs but concentric work in the
intact leg increased compared to the reference
group (Tables 4 and 5).
The eccentric power of the hip flexors of the
amputated side at terminal stance increased after
treatment and remained at follow-up (Fig. 2a).
There were no significant differences between
the sides and compared to the reference group
(Tables 4 and 5).
Concentric power in the hip flexors of the
amputees during pre-swing does not reach the
same peak as the reference group (Fig. 2a and 2b).
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
Gait analysis in trans-femoral amputees 107
-
e
e
s
t
d
e
g
i
d
t
i
flj
o
20 -
15 -
10 .
5
,
> .
Post 6
months
Reference
group
N*
percent of gait cycle
Fig 1. Mean values of pelvic tilt (degrees) in the sagittal plane of the amputated side at self-selected comfortable
speed (N=9).
Post
Post 6
months
Reference
group
-1,5
percent of gait cycle
Fig 2a. Mean values of hip power (w/kg) in the sagittal plane of the amputated side at self-selected comfortable
speed (N=9).
Hip flexion moment at pre-swing was almost
symmetrical but was higher on the intact side
after treatment and comparable to the reference
group at follow-up (Tables 4 and 5). On the
amputated side, the graph showed a much
steeper curve from pre-swing to mid-swing
compared to the reference group, as if the
prosthesis was being rapidly lifted (Fig. 3). This
pattern was not altered after treatment.
Knee
On the amputated side the prosthetic knee was
continuously extended during the whole stance-
phase (Table 4).
During shock absorption maximum knee
flexion in the intact knee increased after
treatment and was higher than in the reference
group. Knee extension moment during shock
absorption in the intact knee increased almost
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
108 C. Sjodahl, G-B. Jamlo, B. Soderberg and B. M. Persson
* Post
Post 6
months
Reference
group
percent of gait cycle
Fig 2b. Mean values of hip power (w/kg) in the sagittal plane of the intact side at self-selected comfortable speed (N=9).
Pre
Post
Post 6
months
Reference
group
percent of gait cycle
Fig 3. Mean values of hip flexion/extension moment (Nm/kg) of the amputated side at self-selected comfortable
speed (N=9).
twofold after treatment and was higher
compared to the amputated leg and to the
reference group (Tables 4 and 5). The eccentric
power of quadriceps increased from 0.6 to 1.8
w/kg after treatment and remained at that level at
follow-up. Compared to the reference group
(Fig. 4, Tables 4 and 5) the value was more than
four times higher.
Just after shock absorption, the concentric
power of the quadriceps increased after
treatment and remained at that level at follow-
up. It was significantly higher than in the
reference group (Fig.4, Tables 4 and 5).
Ankle
At 26% of the gait cycle there was a bump
in the plantar flexion moment curve indicating
that the participants used vaulting before
treatment. After treatment this curve had almost
been smoothed out, but at follow-up there was
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
Gait analysis in trans-femoral amputees
109
w
/
k
t
r
i
e
c
Q)
O
O
r
i
c
/
c
e
n
t
o
o
o>
1.5-
1 -
0.5-
0
-1 -
-1,5-
< ? -
Pre
-XPost
- - . Post 6
months
Reference
"group
-2,5
percent of gait cycle
Fig 4. Mean values of knee power (w/kg) in the sagittal plane of the intact side at self-selected comfortable speed (N=9).
P r e
P o s t
10 20 30 40 50 60 70 80 90 1C0
m
Post 6
months
Reference
group
percent of gait cycle
Fig 5. Mean values of ankle dorsi-/plantarflexion moment (Nm/kg) of the intact side at self-selected comfortable
speed (N=9).
a tendency for this pattern to re-appear
(Fig. 5). Peak plantar flexion moment increased
during treatment and remained at follow-up
at the same level, which was higher compared
to the reference group (Fig. 5).
Maximum plantar flexion power on the intact
side increased after treatment and was not
significantly different compared to the reference
group at follow-up (Tables 4 and 5). There was
a significant increase (p< 0.05) in plantar flexion
power in the prosthetic foot after treatment, but
not compared to the intact side and the reference
group (Tables 4 and 5).
After treatment all patients reported that
almost all low back pain was gone, except for
two subjects who felt some pain occasionally.
Discussion
After training it was observed clinically that
the gait was normal. According to the results the
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
110 C. Sjodahl, G-B. Jarnlo, B. Soderberg and B. M. Persson
prosthesis was used with a greater symmetry, for
example in the work of the hip extensors at the
beginning and the end of stance phase. Although
the gait pattern was observed to be normal, there
were some signs of increased compensatory
responses on the intact side.
There was almost no change in pelvic
movements over time (Fig. 1), as the curves had
the same shape at all three measurements. The
forward tilt during stance-phase on the
amputated side was probably an attempt to
balance the upper body over the base of support
and to maintain knee stability. Another
explanation may be an altered use of m.
iliopsoas as the eccentric power in the hip
flexors increased almost twofold. A third
explanation could be that the femur was slightly
extended relative to the soft tissues of the stump
and the socket position. The increased backward
tilt of the pelvis during swing phase may be due
to the back being rounded as abdominal muscles
were probably used to support hip flexion.
The increased hip flexion on the intact side at
the beginning of stance phase may have been a
consequence of the tilted pelvis, but may also be
an attempt to increase the energy input. Hip
flexion moment was symmetrical after treatment
and at follow-up on the intact side compared to
the reference group. The immediate action of the
hip extensors on the amputated side at heel-
strike may be a compensatory response to
counteract an external flexion moment and to
stabilise the knee j'oint. On the amputated side,
the steep curve at the end of terminal stance may
indicate a need for a quick transition from
extension to flexion to swing the prosthetic foot
forward. Another explanation may be a
combination of a forward flexion of the upper
body to support hip flexion. The power
produced at both hip joints during pre-swing was
lower compared to the reference group, which
may also be caused by the forward tilt of the
pelvis. Another explanation may be the assisted
knee flexion/extension during swing-phase by
the artificial knee joint, requiring less work of
the hip flexors.
The increased knee flexion at shock
absorption on the intact side resulted in a longer
lever arm for the compression force. The knee
extension moment was slightly higher than the
reference group before treatment, but increased
to more than twice the reference value after
treatment. The eccentric power of quadriceps
during this phase was almost five times as high
after treatment as the reference group during this
phase. This meant a considerable increase of
power that had to be produced and absorbed by
the muscles of the knee joint in spite of lower
walking speed compared to the reference group
and needs to be investigated further.
Increased gait speed was mainly achieved by
an increased step length on both sides, which
was significantly longer than in the reference
group. Cadence increased significantly, but was
still less than in the reference group. Jaegers et
al. (1995) made the same finding when studying
prosthetic gait in trans-femoral amputees and
interpreted it as a typical sign of pathological
gait as opposed to normal gait, where cadence
increases with increased gait speed.
Despite the fact that the subjects still took
longer steps with both legs than the reference
group, there was no significant difference in step
length between the legs. On the amputated side
single and double support did not change after
treatment, but the values showed less variability
than before treatment in both legs. This
indicated a more stable and symmetrical pattern,
although the single and double support values
may have been slightly high before treatment as
three participants were dependent on walking-
aids before treatment. Also, an error in the
measurements may have been introduced, as
some parts of the processing, for example
marking heel-strike and toe-off, are done
manually, and they occur within one hundredth
of a second. This may account for some of the
calculated asymmetries in double support
values. Another explanation may be that values
for the amputated and intact side represent
median values assessed at different trials and
therefore some differences may have been
caused by slight variations of gait speed.
Gait speed was equivalent of normal speed
according to Perry (1992) and was matched to
the normalised gait speed used to estimate the
interval of pre-set traffic lights at pedestrian
crossings in Sweden (Lundgren-Lindquist et al.,
1983). Eight subjects had increased their gait
speed matching their gender and age group and
one subject walked slightly slower compared to
the results of Bohannon (1997). The speed of
the reference group was still higher, which may
be explained by this group consisting mostly of
hospital employees, who are walking a lot
during working hours and may therefore be
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
Gait analysis in trans-femoral amputees
111
more fit than the average person. Also, this
reference group, chosen to match the group of
trans-femoral amputees, was younger than
Perry's, which ranged from 20-85 years. This
could be one explanation for the higher speed of
the reference group as Perry (1992) noted a
significant decline in walking ability after 70
years of age.
The construction of the energy-storing
prosthetic foot is very different from the "normal
ankle joint" and may therefore be difficult to
measure with the VICON-system, as the
biomechanical model of VICON, described in
VCM users manual (1995), is based on healthy
subjects. One intention of the training
programme was to put more load on the energy-
storing construction of the prosthetic foot by
increasing the use of the hip extensors. There
was a small but a significant increase in plantar
flexion power on the prosthetic ankle joint after
treatment, which was unexpected as there was
limited ankle joint movement of the prosthetic
foot and the system defines power as the product
of net moment and joint angular velocity. The
obtained results may be due to properties of the
energy-storing prosthetic foot and the increased
gait speed. To be sure about the effect of the
prosthetic foot, the biomechanical model should
have been adjusted to reflect its design, which
may be possible in future studies.
Also, there is a lot of movement of the femur
inside the socket that the VICON-system cannot
capture and it is not known how this may have
affected the results. Convery and Murray (2000)
have attempted to investigate stump movements
during gait with ultrasound, but the result was
not conclusive.
On the intact side there was an obvious use of
vaulting before treatment, which had almost
disappeared after. At follow-up there was again
a tendency of vaulting though not as strong as
before. This finding might imply a need of
repeated training to prevent amputees from
returning to vaulting.
All subjects complained of low back pain
before treatment, shown to be significantly more
common in trans-femoral amputees (71%) than
in the general population (Smith et al., 1999).
After treatment none of the participants had
these problems. At follow-up two subjects felt
some pain occasionally, but took it as a sign of
warning that they had been careless with their
gait technique and they managed to correct
themselves to make the pain go away. The
increased symmetry might be one explanation of
why these secondary problems have
disappeared.
Further controlled studies are needed to show
if this effect can be generalised. A report on gait
analysis in the frontal and transversal plane is in
progress.
Conclusions
For the nine trans-femoral amputees the
training programme led to normalised gait speed
and gait pattern with increased symmetry in the
hip joints in the sagittal plane produced by
increased muscle work on the amputated side.
The intact side differed from the reference group
concerning knee flexion during shock absorption
and knee kinetics during shock absorption and
mid-stance, which needs further investigation.
Acknowledgements
This study was supported by grants from
Malmoehus County Council, Council for
Medical Health Care Research in South Sweden
and the Medical Faculty, Lund University. The
authors also thank Scandinavian Orthopaedic
Gait Laboratory for valuable support.
REFERENCES
BERGER N (1992). Analysis of amputee gait. In: Atlas of
limb prosthetics: surgical, prosthetic rehabilitation
principles./edited by Bowker JH, Michael JW./ 2nd
edition - St Louis: Mosby-Year Book, Inc. p 371-379.
BOHANNON RW (1997). Comfortable and maximum
walking speed of adults aged 20-79 years: reference
values and determinants. Age Ageing 26, 15-19.
BOONSTRA AM, FIDLER V, EISMA WH (1993). Walking
speed of normal subjects and amputees: aspects of
validity of gait analysis. Prosthet Orthot Int 17, 78-82.
BOONSTRA AM, SCHRAMA J, FIDLER V, EISMA WH
(1994). The gait of unilateral trans-femoral amputees.
Scand J Rehabil Med 26, 217-223.
CONVERY P, MURRAY KD (2000). Ultrasound study of
the motion of the residual femur within a trans-femoral
socket during gait. Prosthet Orthot Int 24, 226-232.
CZERNIECKI JM (1996). Rehabilitation in limb deficiency.
1. Gait and motion analysis. Arch Phys Med Rehabil
77, S3-S8.
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from
112 C. Sjodahl, G-B. Jarnlo, B. Soderberg and B. M. Persson
DAVIS RB, UNPUU S, TYBURSKI D, GAGE J (1991). A gait
analysis data collection and reduction technique.
Human Move Sci 10, 575-587.
GAGE JR (1991). Gait analysis in cerebral palsy. London:
MacKeith Press.
HALE SA (1990). Analysis of the swing phase dynamics
and muscular effort of the above-knee amputee for
varying prosthetic shank loads. Prosthet Orthot Int 14,
125-135.
INTERNATIONAL STANDARDS ORGANIZATION, ISO 8548-2
(1993): Prosthetics and orthotics - Limb deficiencies -
Part 2: Method of describing lower limb amputation
stumps.
JAEGERS SMHJ, ARENDZEN H, DE JONGH HJ (1995).
Changes in hip muscles after above-knee amputation.
Clin Orthop 319, 276-284.
JAEGERS SMHJ, ARENDZEN H, DE JONGH HJ (1995).
Prosthetic gait of unilateral trans-femoral amputees: a
kinematic study. Arch Phys Med Rehabil 76, 736-743.
KAMWENDO K AND BERG K (1986). Prosthetic knee-
joints for amputees. - Bromma Handikappinstitutet.
(In Swedish).
L u n d g r e n - L i n d q u i s t B, Aniansson A, Rundgren A
(1983). Functional studies in 79-year-olds. III.
Walking performance and climbing capacity. Scand J
Rehabil Med 15, 125-131.
MURRAY MP, KORY RC, CLARKSON BH, SEPIC SB (1966).
Comparison of free and fast speed walking patterns.
Am J Phys Med 45, 8-23.
PERRY J (1992). Gait analysis: normal and pathological
function. Thorofare: Slack Inc.
SJODAHL C, JARNLO GB, PERSSON BM (2001). Gait
improvement in unilateral trans-femoral amputees by a
combined psychological and physiotherapeutic
treatment. J Rehabil Med 33, 114-118.
SMITH DG, EHDE DM, LEGRO MW, REIBER GE, DEL
AGUILA M AND BOONE DA (1999). Phantom limb,
residual limb and back pain after lower extremity
amputations. Clin Orthop 361, 29-38.
VICON CLINICAL MANAGER (VCM) (1995). - Oxford:
Oxford Metrics Ltd.
WINTER DA AND SIENKO SE (1988). Biomechanics of
below-knee amputee gait. J Biomech 21, 361-367.
at Fund Diag.Est Imstico PARENT on July 1, 2014 poi.sagepub.com Downloaded from