The Effect of Fatigue on Multijoint Kinematics,
Coordination, and Postural Stability During a
Repetitive Lifting Test
Patrick ). Sparto, MS' -
Mohamad Parnianpour, PhD
Tom E. Reinsel, M D
Sheldon Simon, M D 4
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
he functional evaluation
and rehabilitation of pa-
tients suffering from oc-
cupationally related low
back disorders are major
challc.nges to the interdisciplinary
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
treating team, including physical
therapists. Due to the multitude of
etiologic factors and the lack of ab-
normal anatomical findings in a ma-
jority of cases, the diagnoses are of-
ten unclear (7,36). This situation is
compounded by a lack of standard-
ized testing procedures to identify
Journal of Orthopaedic & Sports Physical Therapy®
and quantify the relevant ftmctional
capabilities (46). For instance, Mayer
et a1 (38) have shown that while re-
habilitated patient5 may still show
residual impairment in strength dur-
ing isolated trunk strength tests com-
pared with healthy subjects, they are
able to d o as well as or even better
than the healthy subjects during
multilink coordinated lifting tasks.
Although the literature on strength
and endurance of isolated trunk m u s
cles has grown tremendously in the
last two decades with the advent of
computerized dynamometers (45).
the concepts of tnmk muscular con-
trol and coordination have remained
poorly understood (27,29,49). More-
over, the significant effect5 of fatigue
on the control and coordination dur-
ing mdtijoint repetitive lifting has
remained unexplored.
JOSPT Volume 25 Number 1 January 1997
Because of the inability of strength tests to accurately discriminate between low back pain
patients and healthy subjects, a multifactorial evaluation of low back pain patients is warranted. It
is postulated that measurements of endurance, kinematics, postural stability, and coordination, in
addition to strength, are necessary to fully document the patients' functional capabilities. This
research study was conducted in order to understand the effects of fatigue on the above factors.
Twelve healthy male subjects performed a repetitive lifting test in which a submaximal load was
lifted at a maximal rate. Knee, hip, and trunk motion was measured using videography and
electrogoniometry, postural stability was measured using a forceplate, and coordination parameters
were determined using phase-plane analysis. Fatigue was documented by a 31 % reduction in
lit'tins power. At the end of the endurance test, there was less knee and hip range of motion and
greater spine peak flexion, while the coordination measures demonstrated that there was greater
hip and lumbar spine extension earlier in the lifting phase. The postural stability declined as the
test endured. Utilization of these measures may guide physical therapists in their rehabilitation of
low back pain patients.
Key Words: coordination, postural stability, low back pain
' Graduate Research Assistant, Biomedical Engineering Center, The Ohio State University, Columbus, OH.
Address for correspondence: The Ohio State University, Room 210, Baker Systems and Engineering, 1971
Neil Avenue, Columbus, OH 432 10.
' Assistant Professor, Department of Industrial, Welding, and Systems Engineering, The Ohio State University,
Columbus, OH
'Assistant Professor, Department of Orthopaedic Surgery, The Ohio State University, Columbus, OH
' Professor, Department of Orthopaedic Surgery, The Ohio State University, Columbus, OH
Often, the primary goal of reha- nance of that level of trunk activity
bilitation or functional restoration is over a long duration or many repeti-
reducing the fear of injury, resuming tions. There is an abundance of evi-
physical activities, and enhancing re- dence to suggest that repetitive lifting
sidual strength (46). However, the is a risk factor for low back pain (4,
restoration of endurance, especially 13,17,26,35,60).Furthermore, the
after the muscle wasting that may isometric endurance of the trunk
occur after periods of bed rest ( 16), musculature has been highly associ-
should also be a critical factor for ated with first-time occurrence and
determining when a worker should previous history of low back pain,
return to the job, because the job much more so than tnmk strength
may not only require a specific (5,6,23,33,4159). A balanced exercise
strength but may require mainte- program should include a diverse set
 of strength and endurance training
exercises (56).
It is suggested that adaptations in
the motor control and coordination
that occur due to neuromuscular fa-
tigue may increase the risk of injury
for low back pain. For example, alter-
ations in trunk movement patterns
have been shown to occur in re-
sponse to fatigue. Parnianpour et al
(49) documented decreases in pri-
mary plane trunk motion at the end
of a test in which subjects were asked
to perform sagittally symmetric repet-
itive flexion and extension within a
trunk dynamometer as quickly and
accurately as possible. At the same
time, they observed that accessory
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
plane motions increased. They con-
cluded that increases in frontal and
The job may not only
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
require a specitic
strength but may
require maintenance
of that level of trunk
activity over a
long duration or
Journal of Orthopaedic & Sports Physical Therapy®
many repetitions.
transverse trunk range of motion
were due to adaptations in the neuro-
muscular control system as a result of
fatigue. This finding is significant
because it demonstrates a mechanism
whereby fatigue can lead to increased
risk of injury to the spine, which has
been demonstrated in industrial sur-
veillance, epidemiologic, and tissue
tolerance studies for these types of
loading (26,35,57). While these re-
sults were corroborated by Gomez et
al (19) in a similar study, Potvin (51)
did not find any decrease in trunk
motion symmetry during uncon-
strained repetitive lifting tests. How-
ever, Potvin did discover other a d a p
tations, such as increased peak spinal
flexion, which may predispose the
passive tissues of the spine to certain
types of in-jury.
Until recently, the investigation
of how the central nervous system
coordinates it.. many redundant de-
grees of freedom to perform a rela-
tively simple manual lifting task has
been largely unexplored. However,
the application of tools from other
disciplines has allowed biomechanists
to study control and stability issues
that are central to the study of hu-
man movement. For example, the
adaptations of the central nervous
system during repetitive lifting have
been quantified using measures of
interjoint coordination. Boston et al
(g,lO), using set correlation analysis
of temporal parameters of the range
of movement of the hip and knee
joints, documented a difference be-
tween healthy subjects and low back
pain patient.. . Healthy subject. were
shown to complete the movement of
the hip and knee at the same time,
while patients showed different com-
pletion times. Another measure used
to analyze lifting coordination is the
relative phase angle, which is based
on phase-plane analysis of joint mo-
tion ( 11,52,53). This measure quanti-
fies the relative movement between
two joints at each point in time dur-
ing the lift. This technique has been
shown previously to be sensitive to
changes in lifting coordination that
occur when the mass of the load is
changed (52,54). It was hypothesized
that this method would also be sensi-
tive to coordination changes that are
elicited during fatiguing repetitive
lifting.
Although the evaluation of bal-
ance and postural stability is essential
to the rehabilitation of stroke and
neurological disorder patients, their
evaluation has been much less em-
phasized in the rehabilitation of low
back pain patients. However, low
back injuries are often precipitated
by sudden postural perturbations,
such as falls or slips (8). Oddsson et
al (43) have shown silence in erector
spinae activity followed by a large
increase in activity after a forced
backward perturbation, which indi-
cated substantial force gradients in
the muscle that may be injurious to
the passive tissues. Hence, decreased
postural stability may increase the
risk for a more injurious loading to
the spine. It is suggested that the
adaptive response to perturbations to
the load or platform, or slippery con-
ditions, may be less effective due to
fatigue, as indicated by a reduction in
the postural stability.
It is suggested that adaptations in
the motor control of manual lifting
are quantifiable and that changes in
these measures during repetitive lift-
ing may lead to identification of
mechanisms that explain the correla-
tion between risk of low back injury
and endurance. Consequently, the
purpose of this study is to determine
the effect of fatigue on kinematic
and coordination responses during
repetitive lifting of a submaximal load
at a selfselected maximal lifting rate.
METHODS
Twelve healthy male subjects vol-
unteered for the study. Mean (range)
age, height, and mass were 27 (20-
36) years, 180 (164-195) cm, and 80
(61-96) kg, respectively. Before par-
ticipating, they were cleared to partic-
ipate after a physical screening by an
orthopaedic surgeon. The subjects
then read and signed the informed
consent, approved by the Biomedical
Sciences Committee, Research Involv-
ing Human Subject. Committee, Ohio
State University, Columbus, OH.
Figure 1 shows a schematic of the
set-up by which ground reaction
forces and moments, joint trajecto-
ries, and lifting forces were simulta-
neously measured. The lift- were per-
formed with the subject standing on
a forceplate (Bertec Corp., Worthing-
ton, OH) so that ground reaction
forces and moments could be mea-
sured. The lumbar spine motion rela-
tive to the pelvis was measured with a
triaxial electrogoniometer (Lumbar
Volume 2.5 Number 1 *January 1997 JOSPT
 RESEARCH STUDY

(32,47). Briefly, the protocol con-

JOINTS sisted of a series of three maximal
0,. Ankle elbow flexor, knee extensor, and
0,. Knee trunk extensor isometric exertions
0,. Hip (12) and a maximal isokinetic lifting
0,. Lumbar Spine test (47), which provided the esti-
0,. Shoulder mated maximal isoinertial lifting
0,. Elbow strength. The strength measures were
shown to be repeatable from testing
SEGMENTS session to testing session (3). while
A. Shank the simulation of isoinertial lifting
B. Thigh
conditions by the LIDOLift at the
C. Pelvis
tested loads was shown to be within
D. Trunk
95% confidence of the intended re-
E. Upper arm
sponse (58). The mass is subse-
F. Forearm
quently increased from this initial
estimate during isoinertial lifts until
FIGURE 2. Definition of the joint angles andsegmenh the load's maximum acceleration
for the kinematic and coordination measures.
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.

does not exceed 0.3 g (gravitational

acceleration = 9.8 m/sec2).
The video camera was placed
FIGURE 1. Initial lining position o i a subject perform- Before the commencement of
ing the repetitive endurance test. perpendicular to the sagittal plane of the repetitive lifting test, the subject
the subject at a distance of approxi- performed stretching exercises for
mately 4 meters. It is recognized that approximately 5 minutes and the
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

Motion Monitor, Ohio State Univer- some out of plane movements may resting heart rate of the subject was
sity, Columbus, OH). The Lumbar have occurred and introduced error observed with a heart rate monitor
Motion Monitor, has been found to into the measurement of the joint (Polar Favor, Polar Electro, Inc., Port
be as accurate as video-based systems angles. However, it was assumed that Washington, NY). With a small box
(34), and Cronbach correlation coef- nonsagittal motion was kept to a min- (25 cm length, 30 cm width, 23 cm
ficients greater than 0.9 have demon- imum based on the amount of lateral depth, with handles centered along
strated its repeatability for measuring and transverse trunk range of motion the length, 7 cm from the top) at-
sagittal movements on a day-to-day (see Results section) and the lateral tached to the robotic arm, the lift
basis (36). The Lumbar Motion Mon- excursion of the box (approximately simulator was set to provide a con-
Journal of Orthopaedic & Sports Physical Therapy®

itor was placed on each subject using 2 cm). Nigg and Cole (42) have stant load equal to 25% of the sub-
a set of harnesses scaled to the size of shown that the error in measuring ject's maximal isoinertial lifting capa-
the individual. Likewise, the hip mo- the length of an object is less than bility as determined by the maximal
tion relative to the pelvis was mea- 0.5% for the amount of trunk lateral isoinertial lifting capacity test. After
sured with a biaxial electrogoniom- bending measured in this study and the test instructions, the subject per-
eter (hip monitor), which has been less than 4.0% for motion that devi- formed a few warm-up lifts to be-
documented to be accurate through- ates as much as 15" from the plane come accustomed to the new load.
out its range of motion (27,28). The perpendicular to the camera axis. In Then, at the command of the experi-
axis of sagittal flexion and extension addition, the robustness of the planar menter, the subject began lifting and
was aligned with the greater trochan- model is demonstrated by the strong lowering the box repetitively at a
ter of the femur. The planar sagittal correlation coefficient ( r = .95) and maximal rate as selected by the sub-
motion of the ankle, knee, s h o ~ ~ l d e r , the small relative error (root mean ject. The starting and destination
and elbow was measured using a two- square = 7.3%) between the mea- heights of the lift were at midshank
dimensional video analysis system sured and predicted vertical ground and waist level, respectively. The lift-
(LIDOKAS, Loredan Biomedical, reaction forces using methodology ing style was self-selected by the s u b
Inc., Davis, CA). Reflective markers reported in a previous paper (58). ject, and the subject was instructed
were placed over the approximate Each subject was tested using the not to move his feet. The subject was
joint centers of rotation in the sagit- LIDOLift lifting simulator (Loredan instructed to lift at his maximal lift-
tal plane of the ankle, knee, hip, Biomedical, Inc., Davis, CA) to deter- ing rate until he could no longer
shoulder, elbow, and wrist (58). Fig- mine his maximal isoinertial (con- continue. With encouragement to
ure 2 provides the joint angle defini- stant mass) lifting capacity, employ- maintain the same pace, the subject
tions for clarity. ing a previously described algorithm continued to lift and lower until

JOSPT Volume 2.5 Number 1 January 1997

 Phase plane analysis of joint motion the forceplate (31,55). If the ante-
Nomalized Peak 1 rior-posterior excursion of the center
Exiens~onVeloctly

..,:.
of pressure exceeds the base of s u p
Liing ,, port, the subject will fall due to loss
,
,I of balance (15,31). Since the center
3
.- 0.5 ,,';m' m X
of pressure is a measure of whole
0 body dynamics (15) and is also
bounded by the size of the base of
support of the subject, it provides a
convenient parameter for examining
the postural stability. Previous pos-
tural research has compared the ex-
cursion of the center of pressure to a
maximum range based on an in-
verted pendulum model. However,
-1 -0.5 0 0.5 1 because the multijoint movement
Nomallzed Peak Normahzed Peak
Flewon Posfl~on Normalized Position ~xtenslonP O ~ I ~ I C ~
during lifting cannot be represented
by an inverted pendulum, using such
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.

FIGURE 3. Description of how the interjoint coordination measures are determined from the phase-plane a maximum range is not justified,
construction of the joint motion profiles. a#) = Hip phase angle. u,(t) = Knee phase angle (see text for details).
and the method of determining such
a limit during lifting is unclear.
1) the subject felt that he could no comparison on equal time bases (58). Phase-plane analysis has been
longer continue, or 2) the heart rate Furthermore, data from the three used to describe the interjoint coor-
of the subject, as monitored by the repetitions were averaged together to dination during lifting tasks (1 1,52,
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

tester, attained 180 beats/minute.
Mayer et al (37) had also used an
aerobic limit for their progressive
isoinertial lifting evaluations. The
heart rate was monitored continu-
ously by the experimenter and re-
corded at 30-second intervals.
The analog voltage signals from
the LIDOLift, Lumbar Motion Moni-
Journal of Orthopaedic & Sports Physical Therapy®
create ensemble curves that guarded 53). The trajectories of angular posi-
against any influential extreme values tion and velocity constitute the phase
in the individual cycles. These aver- portrait in the phase plane. Due to
aged data sets from the start and end reproducibility of the lifting tasks, the
of the endurance test formed the portraits will be almost closed curves
bases of comparison for the statistical for each cycle representing periodic-
analysis. The 128 point curves were ity of the lift cycle. A complete de-
analyzed to determine the mean, scription of the measurement of co-
maximum, minimum, and range ordination by phase-plane analysis is

tor, hip monitor, and forceplate were (where applicable) of the lifting and fully detailed in Burgess-Limerick et
digitized at 60 Hz (to match the data lowering phases. For the sake of a1 (11) and Scholz (52,53). Figure 3
capture rate of the video system) by a brevity, only the lifting results will briefly demonstrates the phase-plane
12-bit A/D hoard (Scientific Solutions, be presented. analysis procedure.
Solon, OH) and stored on a Compaq The effect of fatigue on the de- Phase portraits for the knee and
Portable 386 computer (Compaq, scriptive statistic measures of the tri- hip for a typical lifting and lowering
Houston, TX). Data from the video axial trunk motion, biaxial hip mo- cycle are plotted. The angular posi-
and A/D collection were synchro- tion, and planar knee motion was tion is normalized so that the peak
nized and input into a program that studied. Based on previous investiga- flexion angle equals - 1, the peak
divided the data into individual cy- tions (19,49,51), it was hypothesized extension angle equals 1, and the
cles. A complete cycle consisted of a that adaptations to fatigue would re- midpoint of the range equals zero.
lifting and lowering phase, beginning sult in reduced primary plane knee, The normalized position for any an-
from the lowest position of the box. hip, and trunk motion and increased gle is determined by a linear interpo-
The first three cycles and the last cy- secondary plane trunk motion. lation of the raw and normalized po-
cle were discarded to dismiss any arti- Similarly, it waq postulated that sitions. Similarly, the peak flexion
fact due to boundary conditions (36); decreased motor control as a result velocity is normalized to - 1, the
this ensured that the subjects had of fatigue would be evident by a re- peak extension velocity is normalized
settled into their lifting rhythms. duction in the postural stability. The to 1, and the zero velocity is set to
Data from each cycle belonging to a postural stability has been measured zero. For each phase of movement
subset of the first three and last three in balance studies by the anterior- (flexion or extension), the instanta-
of the remaining cycles were interpo- posterior excursion (range) of the neous angular velocity is linearly in-
lated to 128 point curves to allow for center of pressure as measured using terpolated between 0 and - 1 for

Volume 25 Number 1 January 1W7 JOSF'T


flexion and 0 and 1 for extension.
For each data point sampled, radial
lines are drawn from the origin to
the phase portrait's coordinate. The
angle between the negative abscissa
and the radial line is the phase angle.
The phase angles at 0, 90, 180, and
270" are depicted in Figure 3. The
knee phase angle for a point in the
lifting cycle is designated ak(t),while
the hip phase angle for the same
point is a h ( t ) .The difference be-
tween the phase angles for two joints
is the relative phase angle. For this
case, the knee phase angle is shown
to lead the hip phase angle.
Because the phase angle and rel-
ative phase angle are computed from
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
the phase portrait, the repeatability
of these measures will be dependent
on the repeatability of the joint angu-
lar position and velocity measures.
Consequently, it is assumed that Io-
cally at the start and end of the lift,
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
the measures of the joint angular po-
sition and velocity were relatively sta-
ble. The variability in the joint posi-
tion trajectories over the three cycles
at the start and end of the endurance
test was quite low; the variance ratio,
a measure of the repeatability of a
signal (22), was 8.5% for the knee
angular position, 2.5% for the hip,
Journal of Orthopaedic & Sports Physical Therapy®
and 3.5% for the lumbar spine.
It is assumed that goaldirected
movement (in this case, lifting a box
from midshank to waist at a maximal
lifting rate) is completed by an ideal
sequence of coordination (14). Any
departure from this pattern indicates
a change in coordination. The phase
angles for the knee, hip, and lumbar
spine, as well as relative phase angles
between the knee and hip, hip and
lumbar spine, and knee and lumbar
spine, were calculated to investigate
any alterations in the lifting coordi-
nation as a result of fatigue.
The dependent variables (knee,
hip, and spine kinematics, phase an-
gles and relative phase angles, and
trunk center of mass and center of
pressure) were tested using an analy-
sis of variance with a repeated mea-
sures design to determine whether
JOSPT Volume 25 Number 1 January 1997
RESEARCH STUDY
Start End
power, respectively (Table 1). The
P effect of fatigue on the kinematic
x SD x SD adaptations of the knee, hip, and
Lift duration (sec) .71 .09 .77 .09 NS lumbar spine are provided in Table 2
(lifting phase (due to equipment malfunction, lum-
only)
Average lifting .82 .12 .73 .09 <.05 bar spine motion data could be ana-
velocity (mlsec) lyzed for only eight of the 12 sub-
Averagelifting 272 103 202 34 <.05 jects). The knee and hip had
force (N) significantly less range of motion at
Average lifting 157 72 108 20 <.05 the end of the test. While there was
power (W)
virtually no change in the range of
NS = Not signiricant. motion of the lumbar spine with re-
TABLE 1. Results from the repeated measures spect to the pelvis, the amount of
analysis of variance for the effect of fatigue on lihing peak flexion was greater at the end
performance measures. of the test. Because of their smaller
range of motion, knee and hip veloc-
fatigue significantly affected their val- ity were also reduced at the end of
ues between the start and end of the the test.
endurance task. There were no discernible differ-
ences in lateral or transverse lumbar
RESULTS spine motion from the start of the
test to the end of the test as mea-
The mean (SD) endurance time sured with the Lumbar Motion Moni-
of the repetitive lifting task was 130 tor (Table 3). Alternatively, the ex-
(37) seconds, and the mean lift rate cursion of the center of pressure in
was 39 (4) lifts/minute. Fatigue was the anterior-posterior direction in-
documented by a 26% and a 31% creased significantly at the end of the
reduction in average lifting force and test. This measurement was con-
Start End
P
X SD X SD
Position (degrees)
Knee range of motion
Knee peak extension
Knee peak flexion
Hip range of motion
Hip peak extension
Hip peak flexion
Lumbar spine range of motion
Lumbar spine peak extension
Lumbar spine peak flexion
Velocity (Olsec)
Knee peak extension velocity
Knee average extension velocity
Hip peak extension velocity
Hip average extension velocity
Lumbar spine peak extension velocity
Lumbar spine average extension velocity
Acceleration V/sec2)
Knee peak extension acceleration
Knee peak extension deceleration
Hip peak extension acceleration
Hip peak extension deceleration
Lumbar spine peak extension acceleration
Lumbar spine peak extension deceleration
NS = Not signiiicant.
TABLE 2. Results from the repeated measures analysis of variance for the effect of fatigue on joint kinematics.
 RESEARCH STUDY

Start End firmed by increased anterior-poste-

- P rior excursion of the trunk center of
mass.
Motor control The mean phase angle for each
Lumbar spine lateral range of motion (degrees) 3.3 1. joint wa.. tested for the effect of fa-
Lumbar spine transverse range of motion (degrees) 1.2 0.
Postural stability tigue to investigate if there were any
A.*.A.- +-. :A- A"
lu~
-am-
n l l l r l ~ u ~ - p x n carurxull
-.:
ul
.dm--&.-
LCIIICI
A
UI
..-en.-
p~cmurc\LIIII
I--) C L 2 differences in the relative timing of
Anterior-posterior excursion of trunk center of mass (cm) each joint's motion. For example, the
Coordination phase angles for each joint for one
Knee average phase angle (degrees) subject at both the start and end of
Hip average phase angle (degrees)
Lumbar spine average phase angle (degrees) the test are plotted in Figure 4. This
Knee-hip average relative phase angle (degrees) p p h illustrates that for each joint,
Hiplumbar spine average relative phase angle (degrees) the phase angle at the end of the
Knee-lumbar spine average relative phase angle (degrees) endurance test was greater, indicating
NS = Not significant. that the relative amount of extension
occurred earlier in the duration of
TABLE 3. Results from the repeatedmeasures analysis of variance for the effect of fatigue on motor control,
postural stability, and coordination measures. the lifting cycle. In fact, the amount
of extension occurred significantly
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.

earlier for the hip and lumbar spine

as demonstrated by the mean phase
angles (Table 3).
The phase-plane analysis demon-
strated that for most subjects, motion
was initiated from the distal to proxi-
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.

mal joints. That is, knee extension

preceded hip extension, which pre-
ceded lumbar spine extension. An
example of this sequential activation
for one subject is shown with the aid
of phase angles and relative phase
angles in Figure 5. The phase angles
show that for any point in time, the
knee is farther along in its phase por-
Journal of Orthopaedic & Sports Physical Therapy®

- 210
trait than the hip and, likewise, the
hip farther along than the lumbar
spine. These results are confirmed by
the mean relative phase angles,
which show that the difference be-
: 60 tween the knee and hip and hip and
2 30
P
o lumbar spine is positive (Table 3).
0% 25% 50% 75% 100% During the repetitive test, the h i p
lumbar spine relative phase angle
significantly increased in magnitude,
Lumbar Spine indicating that the hip led the spine
- 210
by a greater magnitude as the test

! 1i
-
g
2
60
30
other
ences
progressed.
combinations
in relative
There phase
were
of joint
angle
no differ-
angles.
for the

a 0
0% 25% so% 7596 100% DISCUSSION
Percent of LM Cycle The hypothesis of decreased
FIGURE 4. Change in the phase angles from start to end of the repetitive lifting test for one subject. The phase range of motion for the knee and
angles are shown for the knee, hip, and lumbar spine. hip was confirmed. The decreased

8 Volume 25 Number 1 January 1997 JOSPT


Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
0% 25%
Percent of lifting cycle
FIGURE 5. The relationshipbetween the normalized joint position, phase angle, and relativephase angle for the tially greater fatigue of the knee ex-
knee, hip, and lumbar spine. All data shown were taken from the same cycle at the start of one subject's repetitive tensor musculature. Further exami-
lifting test.
Journal of Orthopaedic & Sports Physical Therapy®
knee and hip motion combined with
increased peak lumbar flexion indi-
cate a change from a squat lift to a
stoop lift. A change toward a stoop
lift (ie., increased peak lumbar flex-
ion) was also found in several s u b
jects by Potvin (51). The stoop lift
has been shown to be physiologically
less demanding (18,20,50). Although
there is no direct method of mea-
surement of energy expenditure in
this study, the vertical excursion of
the body center of mass, which was
estimated using the video joint kine-
matics and body segment parameters, subjects, there was substantial bending
was less at the end. Hence, it is in-
ferred that less energy was spent in
raising the body against gravity. Al-
though the range of motion of the
lumbar spine was unchanged, the
JOSFT Volume 25 Number 1 January 1997
I x Knee
50% 75% 100%
trunk experienced 4" more flexion at range of motion may also have been
the end of the endurance task.
The greater amount of lumbar
spine peak flexion at the end of the
repetitive lifting task (33 vs. 29") may motion at the start and end, suggest-
indicate an increased risk for low
back injury. Based on range of m e
tion measurements, it is estimated
that the lumbar spines of the subjects been shown that the erector spinae
were flexed approximately 34%
(SD = 23%) of their elastic limit ( 1 ) . resistant slow twitch fibers compared
The range in percentage of flexion
compared with the elastic limit (2-
73%) indicated that for several of the
moment resisted by the passive tissues
(1). This may predispose the interspi-
nous and capsular ligaments for
greater risk of injury (39). If applied
repetitively, the increase in bending
RESEARCH STUDY
moment due to greater flexion at the
end of the test may also provide condi-
tions for gradual disc prolapse (2) and
larger facet contact forces (57).
Less range of motion in the knee
may also signify significant fatigue in
the knee extensors. If the knee ex-
tensors are the limiting factor in lift-
ing, the reduction of demand on the
knee extensors may increase the de-
mand on the low back (63). There-
fore, the evaluation of knee extensor
endurance may be necessary in addi-
tion to trunk endurance evaluation,
because the motion of the knee and
lower back are highly interdependent
during lifting (61).
The effects of fatigue were also
manifested in the temporal profiles
of the joint kinematics. For example,
while all 12 subjects had ballistic (ie.,
no dips in the angular position pro-
file and only one zero crossing in the
velocity profile) knee motion at the
start of the exercise, only eight of the
12 subjects had ballistic motion at the
end. An example of a subject whose
knee motion became less ballistic is
depicted in Figure 6. It is hypothe-
sized that these subjects had substan-
nation revealed that these subjects
had a lesser amount of knee range of
motion during the test. The smaller
an adaptation to prevent fatigue in
the knee extensors. Conversely, all of
the 12 subjects had ballistic lumbar
ing that the trunk extensors may not
fatigue as much as the knee exten-
sors during this type of lift. It has
have a greater proportion of fatigue-
with the knee extensors (24), which
substantially increases their endur-
ance capacity (25).
One of the significant findings of
the Pamianpour et al (49) study was
the increased motion in the second-
ary planes of movement. These find-
ings were not replicated in our study,
perhaps due to differing meamre-
 RESEARCH STUDY
FIGURE 6. The effect of fatigue on the knee motion of one subject. The reduction in the range of motion, as well
as the loss of the "typical" extension and flexion phases, is seen from start to end.
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
ment techniques. The B200 dyna-
mometer constrains the pelvis and
lower extremities while the Lumbar
Motion Monitor, as used in this
study, measures the trunk motion
without constraining the motion (re-
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
fer to references 49 and 34 and 35,
respectively, for further descriptions
of the two devices). Meanwhile,
Pamianpour et al (49) and Gomez et
al (19) both discovered increased
torque in secondary axes. Although it
was not possible to predict secondary
axis torque in this study, given the
small movement in accessory motion,
Journal of Orthopaedic & Sports Physical Therapy®
it is doubtful that there was signifi-
cantly higher loading in the accessory
planes of motion.
A significant increase in the ex-
cursion of the center of pressure in
the anterior-posterior direction was
found at the end of the test, which
was confirmed by greater anterior-
posterior excursion of the trunk cen-
ter of mass. The increase in excur-
sion may be an alarming sign that
the stability of the system is declin-
ing, because a decrease in the lifting
velocity should be accompanied by a
decrease in the center of pressure
excursion (62). The increase in ex-
cursion also reduces the stability mar-
gin, which is the minimum distance
from the location of the center of
pressure to either the heel or toe of
the foot (31). The decrease in stabil-
ity margin may increase the chance
10
of a loss in balance. In addition, the
reduction in postural stability may
amplify the reaction to a slip or per-
turbation. In such a case, there may
be an increase in the risk for a more
injurious loading to the spine while a
lift is performed under fatigued con-
ditions. (43)
Interestingly, the results of this
study indicated that in the fatigued
state, the joint5 reached their exten-
sion at an earlier time of the cycle
(Table 3, Figure 4). Earlier mathe-
matical modeling of trunk extension
and flexion, based on minimization
of energy, had predicted a movement
profile that reached the upright pos-
ture much faster during the exten-
sion phase (48). The results satisfy
biomechanical principles, because by
earlier completion of the extension,
the postural load due to the upper
body weight and the external load is
minimized. This further indicates
that, in part, these adaptive strategies
may subserve the energy minimiza-
tion in the fatigued conditions.
Meanwhile, the relative phase
angles demonstrated that knee exten-
sion led hip extension, which led
lumbar spine extension. For lifting
tasks, this sequential activation a p
pears to be typical coordination (30,
40,44,52-54). Sequential coordina-
tion such as this may maximize the
objective of achieving the maximal
rate of speed during the lifting.
Herring and Chapman (21) have
shown that to optimize the maximum
velocity that a ball can be thrown, the
motion of the arm must be coordi-
nated such that the shoulder exten-
sors are activated before the elbow
extensors and, subsequently, wrist
flexors. The findings of earlier joint
extension and increased lead of hip
extension compared with lumbar
spine extension suggest a change in
the motor control program. If one
assumes coordination determined at
the onset of lifting represents the
"ideal" motor program, then any sig-
nificant change in this coordination
would represent a loss of motor con-
trol or adaptation to the deleterious
effects of fatigue on the capabilities
of the peripheral apparatus (fatigued
muscles). The biomechanical and
epidemiological implications of these
adaptations require further investiga-
tion.
The findings of earlier
joint extension and
increased lead of hip
extension compared
with lumbar spine
extension suggest a
change in the motor
control program.
This study presents methodolo-
gies for quantifying coordination and
postural stability during a multijoint
lifting task. To demonstrate the feasi-
bility of the use of these measures,
healthy subjects were recruited. Fu-
ture work needs to consider if the
same techniques can be used with
the patient population. Boston et al
(9,lO) have demonstrated success in
examining the coordination of low
back pain patients using other similar
measures. However, future biome-
Volume 23 Number 1 January 1997 JOSPT
 chanical and industrial epidemiologi-
cal studies are needed to determine
the association between the decline
in the margin of stability and loss of
coordination to injury mechanisms in
the workplace. These measurements
may be used by the physical therapist
to evaluate the extent of disability of
the low back pain patient and to
guide the rehabilitation strategy us-
ing a multifactorial paradigm.
CONCLUSIONS
A repetitive lifting test in the sag-
ittal plane was performed with a sub-
maximal load at a maximal lifting
rate in order to understand the ef-
fects of fatigue on lower extremity
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
and trunk motion, postural stability,
and coordination parameters. Fatigue
was documented by a 31% reduction
in lifting power. While the joint
ranges of motion indicated that there
was a trend toward a stoop lift, the
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
coordination measures suggested that
the timing of the movement was also
affected by fatigue. For instance, the
hip and lumbar spine demonstrated
significantly earlier extension at the
end of the endurance test. In addi-
tion, knee extension led hip exten-
sion, which led lumbar spine exten-
Journal of Orthopaedic & Sports Physical Therapy®
sion. It is postulated that measure-
ments of endurance, kinematics, p o s
tural stability, and coordination, in
addition to strength, are necessary to
fully document the functional capa-
bilities during lifting tasks. These
novel measures of coordination and
stability need further evaluation to
enhance the methodologies for quan-
titative assessment of low back pain
patients. JOSPT
ACKNOWLEDGMENTS
We would like to thank Professor
MTilliamS. Marras and the rest of the
Biodynamics Lab staff for their in-
valuable assistance.
REFERENCES
1. Adams MA, Dolan P: A technique for
quantifying the bending moment acting
JOSPT Volume 25 Nomher 1 January 1W7
on the lumbar spine in vivo. 1 Biomech
24(2):ll7- 126, 1991
Adams MA, Hutton WC: Gradual disc
prolapse. Spine 1O(6):524-53 1, 1 985
Alpertl, Matheson L, Beam W, Mooney
V: The reliability and validity of two
new tests of maximum lifting capacity.]
Occup Rehabil 1(1):13-29, 1991
Andersson GB]: Epidemiologic aspects
on low back pain in industry. Spine
6(1):53-60, 198 1
Battit! MC, Bigos S], Fisher LD, Hansson
TH, Jones ME, Wortley MD: Isometric
litling strength as a predictor of industrial
back pain reports. Spine 14(8):851-856,
1989
Biering-Sorensen F: Physical measure-
ments as risk indicators for low back
trouble over a one-year period. Spine
9(2):106-119, 1984
Bigos S], Battit! MC: Risk factors. In:
Weinstein IN, Weisel SW (eds), The
Lumbar Spine, pp 846-859. Philadel-
phia: W.B. Saunders Company, 1990
Bigos S], Spengler DM, Martin NA, Zeh
1, Fisher L, Nachemson A, Wang MH:
Back injuries in industry: A retrospec-
tive study. 11. Injury factors. Spine 1 l(3):
246-25 1, I986
Boston ]R, Rudy TE, Lieber S], Stacey
BR: Measuring treatment effects on re-
petitive lifting for patients with chronic
low back pain: Speed, style, and coor-
dination. ] Spinal Disord 8(5):342-35 1,
1995
Boston ]R, Rudy TE, Mener SR, Kubinski
]A: A measure of body movement coordi-
nation during m i t i v e dynamic lihing.
IEEE Trans Rehabil Eng l(3):137- 144, 1993
Burgess-Limerick R, Abernethy B, Neal
R]: Relative phase quantifies interjoint
coordination. ] Biomech 26(1):91-94,
1993
Chaffin DB, Herrin GD, Keyserling
WM: Preemployment strength testing. ]
OCCUP Med 20(6):403-408, 1978
Chaffin DB, Park KS: A longitudinal
study of low back pain as associated
with occupational weight lifting fac-
tors. Am Ind Hyg Assoc 1 34(12):513-
525, 1973
Chapman AE, Sanderson Dl: Muscular
coordination in sporting skills. In:
Winters ]M, Woo SL-Y (eds), Multiple
Muscle Systems: Biomechanics and
Movement Organization, pp 608-620.
New York: Springer-Verlag, 1990
Collins I], De Luca C]: Open-loop and
closed-loopcontrol of posture: A random
walk analysis of center of pressure trajec-
tories. Exp Brain Res 95:308-3 18, 1993
Cooper RG, Forbes WS, layson MIV:
Radiographic demonstration of para-
spinal muscle wasting in patients with
chronic low back pain. Br 1 Rheumatol
3 l:389-394, 1992
Frymoyer ] W, Pope MH, Clements ]H,
Wilder DG, MacPherson B, Ashikaga
T: Risk factors in low-back pain. ] Bone
Joint Surg 65A(2):213-2 18, 1983
Garg A, Saxena U: Effects of lifting fre-
quency and technique on physical fa-
tigue with special reference to psycho-
physical methodology and metabolic
rate. Am Ind Hyg Assoc ] 40:894-903,
1979
Gomez T, Beach G, Cooke C, Hrudey
W, Goyert P: Normative database for
trunk range of motion, strength, veloc-
ity, and endurance with the lsostation
B-200 lumbar dynamometer. Spine
16(1):15-21, 1991
Hagen KB, Hallen 1, Harms-Ringdahl K:
Physiological and subjective responses
to maximal repetitive lifting employing
stoop and squat technique. Eur J Appl
Physiol 67:29 1-297, 1993
Herring RM, Chapman AE: Effects of
changes in segmental values and tim-
ing of both torque and torque reversal
in simulated throws. ] Biomech 25(10):
1173-1 184, 1992
Hershler C, Milner M: An optimality
criterion for processing electromyo-
graphic (EMG) signals relating to hu-
man locomotion. IEEE Trans Biomed
Eng 25(5):413-420, 1 978
Holmstrom E, Moritz U, Andersson M:
Trunk muscle strength and back muscle
endurance in construction workers
with and without low back disorders.
Scand 1 Rehabil Med 24:3-10, 1992
lohnson MA, Polgar 1, Weightman D,
Appleton D: Data on the distribution of
fibre types in thirty-six human muscles:
An autopsy study. ] Neurol Sci 18:1 1 1-
129, 1973
lorgensen K, Nicolaisen T, Kato M:
Muscle fiber distribution, capillary den-
sity, and enzymatic activities in the
lumbar paravertebral muscles of young
men-Significance for isometric endur-
ance. Spine 18(11):1439-1450, 1993
Kelsey ]L, Githens PB, White AA,
Holford TR, Walter SD, O'Connor T,
Ostfeld AM, Weil U, Southwick WO,
Calogero]A: An epidemiologicstudy of
lifting and twisting on the job and risk
for acute prolapsed lumbar interverte-
bra1 disc. J Orthop Res 2:6 1-66, 1984
Kim ]Y: Patterns of trunk neuromuscu-
lar performance in normal subjects and
low-back pain patients. Unpublished
doctoral dissertation, The Ohio State
University, Columbus, OH, 1995
Kim ]Y, Lee ML, Miller R: The accuracy
and reliability of a biaxial hip electro-
goniometer. Unpublished results, The
Ohio State University, Columbus, OH,
1994
Kim JY, Parnianpour M, Marras WS:
Development of a new experimental
 RESEARCH STUDY
protocol to quantitatively assess the
neuromuscular control capability of the
trunk muscles. In: Proceedings of the
annual American Society of Biome-
chanics Conference, Columbus, OH,
October 13- 15, 1994, pp 173- 174
30. Kippers V, Parker A W : Posture related
to myoelectric silence of erectores spi-
nae during trunk flexion. Spine 9(7):
740 -745, 1984
31. Koozekanani SH, Stockwell CW,
McGhee RB, Firoozmand F: O n the
role of dynamic models in quantitative
posturography. IEEE Trans Biomed Eng
27(l O):605-609, 1980
32. Loredan Biomechanical, Inc. :LlDOLift
Operations Manual, Version 2.4,
Davis, CA: Loredan Biomedical, Inc.,
1989
33. Luoto S, Heliovaara M , Hurri H,
Alaranta H: Static back endurance and
the risk of low back pain. Clin Biomech
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
10(6):325-330, 1995
34. Marras WS, Fathallah F: Accuracy of a
three-dimensional lumbar motion
monitor for recording dynamic trunk
motion characteristic. Int J lndust Ergo-
nomics 9:75-87, 1992
35. Marras WS, Lavender SA, Leurgans SE,
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Rajulu SL, Allread WG, Fathallah FA,
Ferguson SA: The role of dynamic
three-dimensional trunk motion in oc-
cupationally related low back disor-
ders. Spine 18(5):617- 628, 1993
36. Marras WS, Parnianpour M, Ferguson
SA, Kim JY, Crowell RR, Bose S, Simon
SR: The classification of anatomic- and
symptom-based low back disorders us-
ing motion measure models. Spine 20:
Journal of Orthopaedic & Sports Physical Therapy®
253 1-2546, 1995
37. Mayer TG, Barnes D, Kishino NO,
Nichols G, Gatchel Rl, Mayer H,
Mooney V: Progressive isoinertial lift-
ing evaluation: I. A standardized proto-
col and normative database. Spine
13(9):993-997, 1988
38. Mayer TG, Gatchel R, Betancur I,
Bovasso E: Trunk muscle endurance as-
sessment-Isometric contrasted to iso-
kinetic testing in normal subjects. Spine
20(8):920-927, 1995
39. McGill SM: Estimation of force and ex-
tensor moment contributions of the
disc and ligaments at L4-L5. Spine
13(12):1395-1402, 1988
40. Nelson JM, Walmsley RP, Stevenson
JM: Relative lumbar and pelvic motion
during loaded spinal flexion/extension. 5 1. Potvin JR: The influences of fatigue on
Spine 20(2):1 99-204, 1 995 hypothesized mechanisms of injury to
41. Nicolaisen T, lorgensen K: Trunk the low back during repetitive lifting.
strength, back muscle endurance and Unpublished doctoral dissertation,
low back trouble. Scand / Rehabil Med University of Waterloo, Ontario, Can-
17:121-127, 1985 ada, 7 992
42. Nigg BM, Cole GK: Optical methods. 52. Scholz JP: The effect of load scaling on
In: Nigg BM, Herzog W (eds), Biome- the coordination of manual squat lift-
chanics of the Musculo-skeletal Sys- ing. Hum Move Sci 12:427-459, 1993
tem, pp 254-286. Chichester, England: 53. Scholz /P: Organizational principles for
john Wiley & Sons, 1994 the coordination of lifting. Hum Move
43. Oddsson L, Cresswell AG, Thorstens- Sci 12537-576, 1993
son A: Conflict between voluntary and 54. Scholz JP, Millford JP, McMillan AG:
postural motor programs during per- Neuromuscular coordination of squat
turbed lifting-A potential injury mech- lifting. I: Effect of load magnitude. Phys
anism. Presented at the International Ther 75(2):7 79-132, 1995
Symposium on Posture and Gait, Mat- 55. Schumann T, Redfern MS, Furman JM,
sumoto, Japan, October 3-7, 1994 El-Jaroudi A, Chaparro LF: Time-fre-
44. Paquet N, Malouin F, Richards C: Hip- quency analysis ofpostural sway. J Bio-
spine movement interaction and mus- mech 28(5):603-607, 1995
cle activation patterns during sagittal 56. Scutter S, Fulton I, Trott P, Parnianpour
trunk movements in low back pain pa- M, Grant R, Brien C: Effects of various
tients. Spine l9(5):596- 603, 1994 isoresistive training programmes on
45. Parnianpour M : Applications of quan- trunk muscle performance. Clin Bio-
titative assessment of human perfor- mech 10(7):379-384, 1995
mance in occupational medicine. In: 57. Shirazi-Ad1 A: Finite-element evalua-
Bronzino JD (ed), The Biomedical En- tion of contact loads on facets of an
gineering Handbook, pp 2306-2320. L2-L3 lumbar segment in complex
Boca Raton, FL: CRC Press, 1995 loads. Spine 16(5):533-54 1, 199 1
46. Parnianpour M, Engin A€: A more 58. Sparto PI, Khalaf KA, Parnianpour M,
quantitative approach to classification Simon SR: The reliability and validity of
of impairments, disabilities, and hand- a lift simulator and its functional equiv-
icaps. J Rheumatol Med Rehabil 5(1): alence with free weight lifting tasks.
52- 64, 1994 IEEE Trans Rehabil Eng 3(2):155- 165,
47. Parnianpour M, Hasselquist L, Fagan L, 1995
Aaron A: Correlation among isometric, 59. Suzuki N, Endo S: A quantii~tivestudy
isokinetic and isoinertial muscle perfor- of trunk muscle strength and fatigability
mance during multi-joint coordinated in the low-back-pain-syndrome. Spine
exertions and isolated single joint trunk 8(1):69-74, 1983
exertion. Eur J Phys Med Rehabil3(3): 60. Svensson H, Andersson GBJ:Low-back
114-122, 1993 pain in 40- to 47-year-old men: Work
48. Parnianpour M, Lafferriere G: Mini- history and work environment factors.
mum jerk and near energy optimal tra- Spine 8(3):272-276, 1983
jectories of sagittal trunk movement. 61. Toussaint HM, van Baar CE, van
Am Soc Mech Engineers Bioengineer- Langen PP, de Looze MP, van Dieen
ing Div 17:159-162, 1990 JH: Coordination of the leg muscles in
49. Parnianpour M, Nordin M, Kahanovitz backlift and leglih. / Biomech 25(11):
N, Frankel V: The triaxial coupling of 7279-1 289, 1992
torque generation of trunk muscles dur- 62. Toussaint HM, Commissaris DACM,
ing isometric exertions and the effect of van Dieen JH, Reijen IS, Praet SF€,
fatiguing isoinertial movements on the Beek PI: Controlling the ground reac-
motor output and movement patterns. tion force during lifting. J Motor Behav,
Spine 13(9):982-992, 1988 1996 (in press)
50. Petrofsky JS, Lind AR: Metabolic, car- 63. Trafimow JH, Schipplein OD, Novak
diovascular and respiratory factors in GJ,Andersson GBJ:The effects of quad-
the development of fatigue in lifting riceps fatigue on the technique of lift-
tasks. J Appl Physiol45(1):64 -68, 1978 ing. Spine 18(3):364-367, 1993
Volume 25 Number 1 January 1W7. JOSPT
 This article has been cited by:

1. Luke J. Santamaria, Kate E. Webster. 2010. The Effect of Fatigue on Lower-Limb Biomechanics During Single-Limb Landings:
A Systematic Review. Journal of Orthopaedic & Sports Physical Therapy 40:8, 464-473. [Abstract] [Full Text] [PDF] [PDF Plus]
2. Eleftherios Kellis, Christina Liassou. 2009. The Effect of Selective Muscle Fatigue on Sagittal Lower Limb Kinematics and Muscle
Activity During Level Running. Journal of Orthopaedic & Sports Physical Therapy 39:3, 210-220. [Abstract] [PDF] [PDF Plus]
Downloaded from www.jospt.org at on November 5, 2017. For personal use only. No other uses without permission.
Copyright © 1997 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Journal of Orthopaedic & Sports Physical Therapy®