SPINE Volume 29, Number 3, pp 311–317
©2004, Lippincott Williams & Wilkins, Inc.
Force Production Parameters in Patients With Low
Back Pain and Healthy Control Study Participants
Martin Descarreaux, MSc,*† Jean-Sébastien Blouin, MSc,* and Normand Teasdale, PhD*
Study Design. A control group study with repeated
measures.
Objective. To compare isometric force production pa-
rameters in low back pain and healthy study participants.
Summary and Background Data. Recent evidence sug-
gests that chronic patients with low back pain exhibit
deficits in trunk proprioception and motor control. The
control of force and its between-trial variability are often
taken as critical determinants of performance. We com-
pared various force time characteristics in patients with
low back pain and healthy study participants.
Methods. Fifteen control study participants and 16 pa-
tients with low back pain participated in this study. Study
participants were required to exert 50% and 75% of the
maximal trunk flexion and extension. In a learning phase,
visual and verbal feedback was provided. Following these
learning trials, study participants were asked to perform
10 trials without any feedback. Time to peak force, time to
peak force variability, peak force variability, and absolute
error in peak force were calculated. Time to peak and peak
dF/dt were computed to determine if the first peak of dF/dt
could predict the peak force achieved.
Results. Two subgroups of patients with low back pain
were identified. Controls and patients with low back pain
with more pain showed faster time to peak force than
patients with low back pain with less pain (331 ms and
341 ms vs. 574 ms, respectively). Linear regressions
showed that, for control study participants and low back
pain study participants with more pain, peak dF/dt ex-
plained 94.0% and 97.0% of the variance observed in peak
force while 84.4% was explained for low back pain study
participants with less pain. Peak force variability and ab-
solute error in peak force were similar for all groups.
Conclusions. Patients with low back pain were able to
produce isometric forces with an accuracy similar to con-
trol study participants. The longer time to peak force and
the smaller percentage of variance observed for the linear
regressions suggest that some patients with low back
pain adopted a control mode that was less “open-loop.” It
is possible that this mode of producing forces results
from an adaptation to chronic pain or tissue degenera-
tion. Spine 2004;29:311–317
From *Faculté de Médecine, Division de Kinésiologie, U. Laval, and
†Département de Chiropratique, Université du Québec à Trois-
Rivières, Canada.
Supported by FCQ and FCAR (M.D.) and by C1HR-CFCQ (J.-S.B.).
Acknowledgment date: February 20, 2003. First revision date: May 13,
2003. Acceptance date: May 27, 2003.
The manuscript submitted does not contain information about medical
device(s)/drug(s).
Federal and foundation funds were received in support of this work.
No benefits in any form have been or will be received from a commer-
cial party related directly or indirectly to the subject of this manuscript.
Address correspondence to Martin Descarreaux, MSc, Cité Universi-
taire, Faculté de Médecine, Division de Kinésiolgie, PEPS local 00233,
Québec, Ste-Foy G1K 7P4; E-mail: martin.descarreaux@
kin.msp.ulaval.ca
Epidemiological studies have shown that 50% to 80% of
the population is affected by low back pain (LBP) at least
once in a lifetime.1,2 In the United States, annual costs for
low back disorders have been estimated to $100 billion.3
One factor explaining these enormous costs is the high
rate of recurrence and chronic disability related to LBP
disorders. As reported by Croft et al,4 the majority of
patients with LBP are still symptomatic after 1 year, with
only 21% of patients being pain free and 25% of the
patients completely recovering from disabilities associ-
ated with their low back problems. Work and lifestyle
have changed over the last decades and automation, ro-
botics, and prevention campaign have led to minimiza-
tion of heavy lifting and repetitive tasks in the working
and living environments. During this period of time,
however, the incidence and costs of LBP kept rising.3
Diminution of mechanical loads and trunk forces devel-
oped during working activities does not seem to prevent
LBP episodes. One could easily argue that muscular
weakness is not the predominant factor in the develop-
ment and recurrence of LBP episodes. Indeed, most stud-
ies have indicated that isometric strength, by itself, is not
related to risk of low back disorders.3
There is much recent evidence suggesting that patients
with chronic LBP exhibit deficits in proprioception and
trunk motor control. Changes in postural control,5 de-
layed muscle responses to sudden trunk loading,6 in-
creased trunk movement detection threshold,7 and in-
creased repositioning errors in patients with LBP were
recently reported.8 Hodges demonstrated that the orga-
nization of the trunk muscle responses (anticipatory pos-
tural responses) is altered in study participants with
chronic LBP.9 When producing a fast unilateral flexion
or abduction of the arm, patients with LBP showed de-
layed transversus abdominis activation compared with
healthy study participants. In healthy study participants,
the transversus abdominis muscle activation preceded
the postural perturbation. Hodges proposed that these
changes could be explained by an alteration or adapta-
tion in the planning of motor responses.9 Alternatively,
Indahl et al have proposed that sensorimotor changes
could result from local alterations of the lumbar spine
complex reflex system.10 Numerous mechanoreceptors
and nociceptors located in the intervertebral discs and
zygapophysial joints can modulate local paraspinal mus-
cle activity to control segmental stabilization.11,12 For
example, degenerative conditions of the lumbar tissues
can lower the threshold at which mechanical and pain
receptors are activated.13 Impaired control of segmental
motion and stabilization resulting from degenerative
311
312 Spine • Volume 29 • Number 3 • 2004
condition, pain, or previous low back injuries are possi-
ble mechanisms that could explain the high recurrence
rate of LBP injuries and the development of chronic pain
and disabilities.10
Since repetitive activities requiring trunk forces pro-
duction are often linked to the development of LBP, it is
possible that alteration of force production parameters
could partially point out the chronic changes and adap-
tations related to low back conditions.1,14 To our knowl-
edge, trunk isometric force production parameters and
their variability have not been studied (in a motor con-
trol context) in LBP populations. The control of force
and its variability are at the heart of several motor con-
trol models aiming at the understanding of skillful be-
havior. When a study participant attempts to produce a
given target force repeatedly, the between-trial variabil-
ity of the force-time curves is often taken as a critical
determinant of performance. The underlying idea is that
the initial impulse for producing the force can be related
to the muscular activity necessary to produce the
action.15,16
Furthermore, force production spatiotemporal char-
acteristics measured in an isometric reproduction task
have been proposed to reflect various limitations of the
neuromuscular system.16 Using the first and second de-
rivative of the force signal, Ghez and Gordon proposed a
model to examine the control strategy adopted by study
participants in modulating the amplitude of isometric
forces aimed at a target.15 To determine whether isomet-
ric trunk forces were similarly programmed in LBP and
healthy study participants, we used the model proposed
by Ghez and Gordon.15 If proprioception and motor
planning are affected by LBP, we should observe differ-
ences in force production parameters between healthy
and LBP study participants. Hence, the aim of this study
was to evaluate if patients with LBP can produce various
isometric extension and flexion forces with an accuracy,
variability, and mode of control similar to those of
healthy control study participants.
Methods
Force production parameters were measured in 16 study par-
ticipants with chronic nonspecific LBP and in 15 control study
participants. Every study participant gave their informed writ-
ten consent, and the study was approved by the Université
Laval (Canada) Ethics Committee. All study participants were
recruited through local advertising. The experimental group
(LBP) included 16 study participants (11 men, 5 women, mean
age 41.1 years) who had a history of chronic recurrent LBP that
lasted for at least 6 months. Exclusion criteria for both groups
were as follows: 1) spondylolisthesis or spondylolysis, 2) anky-
losing spondylitis, 3) spinal osteoarthritis or inflammatory ar-
thritis, 4) nerve root compression, 5) trunk neuromuscular dis-
ease, 6) scoliosis (ⱖ15°), 7) previous spinal surgery, 8)
malignant tumor, 9) hypertension, 10) pregnancy, and 11)
breastfeeding. Lateral and anteroposterior radiographic films
of the lumbar spine (including pelvis) were taken to rule out the
possibility of congenital, degenerative, or inflammatory dis-
eases of the lumbar spine. Pain levels at the beginning and the
Table 1. Basic Data on Low Back Pain and Healthy
Study Participants
Controls [mean (SD)] LBP [mean (SD)]
Subjects n ⫽ 15 n ⫽ 16
Age (yr) 38.2 (10.7) 41.1 (9.1)
Height (cm) 172.8 (6.4) 171.1 (7.0)
Weight (kg) 74.1 (15.2) 72.4 (13.9)
Duration of LBP (mo) — 50.93 (16.4)
Oswestry index (%) — 26.8 (9.9)
VAS (mm)
Before testing — 26 (17)
After testing — 31 (15)
end of the experiment were assessed using a standard 100-mm
visual analog pain scale. Each LBP study participant filled out
the modified Oswestry questionnaire before the experiment.
The control group consisted of 15 healthy study participants (9
men, 6 women, mean age 38.2 years). Table 1 shows charac-
teristics of both the experimental and the control groups.
Study participants were asked to exert 50% and 75% of the
maximal flexion and extension forces of trunk muscles against
a harness connected to an isometric testing apparatus (Loredan
Biomedical, West Sacramento, CA). During the maximal vol-
untary contractions, study participants were encouraged ver-
bally to give their maximum effort in both flexion and exten-
sion. The higher force value obtained in three consecutive
4-second trials was used as the reference for maximal voluntary
contraction. Testing was done in a neutral standing posture.
With their eyes closed, study participants had to learn to pro-
duce each of the four experimental conditions (50% and 75%
of the maximal isometric force in extension and flexion) within
an error margin of ⫾10%. For each trial, study participants
were instructed to produce a rapid isometric force as soon as
they heard the start signal given by the experimenter. They
were encouraged to produce a single impulse (“shoot and re-
lease”) and to make no attempt at correcting the force once the
contraction was initiated. After each learning trial, study par-
ticipants opened their eyes and looked at their result on an
oscilloscope (visual accuracy feedback) located in front of
them. They could then evaluate their performance and correct
it for the next trial, if necessary. The learning sequence stopped
when 5 consecutive trials were successfully made (within the
10% margin). Following these learning trials, study partici-
pants were asked to perform 10 consecutive trials without any
visual feedback. All conditions were performed by block. The
number of learning trials needed for each of the four experi-
mental conditions was also calculated and submitted to a group
by direction by force analysis of variance with repeated mea-
sures on the last two factors. The analysis revealed no signifi-
cant effect (mean number of trials 15.8; P ⬎ 0.05).
For every trial (3 seconds), torque data were recorded at a
sampling rate of 500 Hz. Torque data were digitally filtered
with a seventh-order Butterworth filter (7 Hz low pass cutoff
frequency). Onset of force and peak force were then deter-
mined in every trial for each study participant. Using this in-
formation, time to peak force, time to peak force variability,
peak force variability, as well as absolute error in peak force
were calculated for each condition. Absolute error in peak
force represents the positive difference between the reached
peak force and the goal peak force, whereas time to peak rep-
resents the period of time between the beginning of rising force
and the maximal force obtained in the trial. Peak dF/dt was also
 Force Production Parameters • Descarreaux et al 313

Table 2. Basic Data on Low Back Pain Subgroups indication of a preprogrammed or open-loop mode of
control.17
LBPshort LBPlong
[mean(SD)] [mean(SD)]
Results
Subjects n⫽7 n⫽9
Age (yr) 39.7 (12.7) 42.1 (10.1) Both groups demonstrated similar maximal flexion and
Height (cm) 173.6 (7.4) 169.4 (7.9)
Weight (kg) 68.1 (15.2) 75.4 (13.9) extension forces. On average, the peak flexion and peak
Duration of LBP (mo) 52.4 (12.8) 45.5(17.4) extension forces were 179.5 Nm and 231.5 Nm, respec-
Oswestry index (%) 26.0 (9.3) 27.4 (10.8) tively (P ⬎ 0.05 for the main effects of group). As re-
VAS (mm) before testing* 43 (9) 14 (8)
ported in Table 1, pain levels of the LBP study partici-
* Significant difference (P ⬍ 0.001).
pants, measured with the visual analog pain scale, did
not vary significantly during the course of the experiment
computed to examine if the first peak of dF/dt could predict the (P ⬎ 0.05). A first analysis of the force-time curves, how-
peak force achieved. Linear regressions were calculated for ever, revealed that LBP study participants were not a
each study participant, and a high r2, indicating that the first homogeneous group. They could clearly be separated
peak of dF/dt could predict the peak force, was taken as an according to the time they took to achieve the various

Figure 1. A: Time to peak force

(95% confidence interval) for LBP
subgroups and control group. B:
Time to peak force variability
(95% confidence interval) for LBP
subgroups and control group.
314 Spine • Volume 29 • Number 3 • 2004

Figure 2. A: Typical force and

dF/dt curves illustrating the mean
(SD is represented by the dashed
line) of 10 consecutive flexion tri-
als (without feedback) for a typ-
ical control subject, LBPshort and
LBPlong subject. B: This figure
clearly illustrates that LBPlong
subjects used a longer time to
peak force and showed more
than one peak in the first time
derivative of the force signa.

force levels. Seven LBP study participants showed time to
peak force similar to those observed for healthy study
participants. For these LBP study participants, all times
to peak force were within 1 standard deviation of the
mean time to peak force observed for healthy study par-
ticipants. For the nine other LBP study participants, the
time to peak force was beyond 3 standard deviations of
the time to peak force observed for healthy study partic-
ipants. For this reason, we decided to create two sub-
groups from the original LBP group. Characteristics of
the two subgroups, therein called LBPshort and LBPlong
are presented in Table 2. It is interesting to note that the
LBPshort subgroup reported pain scores (visual analog
pain scale) that were significantly higher than the scores
reported by the LBPlong subgroup (43 mm vs. 14 mm; P
⬍ 0.001). A reanalysis of the maximal peak flexion and
peak extension forces showed that the three groups were
not different (P ⬎ 0.05 for the main effects of group). All
subsequent analyses included three groups (healthy
study participants, LBPshort and LBPlong), and force-
dependent variables were submitted to group by direc-
tion by force level (3 ⫻ 2 x 2) analysis of variance with
repeated measures on the last two factors.
Not surprisingly, the average time to peak force was
significantly longer for the LBPlong than for the LBPshort
and healthy study participants (574 ms vs. 341 ms and
331 ms, respectively; F2,28 ⫽ 11.47, P ⫽ 0.0002 for the
main effect of group). This observation is illustrated in

Figure 1A. For all groups, the time to peak force was
longer when producing extension than flexion forces
(446 ms vs. 385 ms, respectively; F1,27 ⫽ 9.02, P ⫽
0.006). The main effect of force level and all interactions
were not significant (P ⬎ 0.05). Figure 1B shows similar
observations for the time to peak force variability (82 vs.
41 and 40 ms for the LBPlong, LBPshort, and healthy study
participants, respectively; F2,28 ⫽ 7.34, P ⫽ 0.003). The
main effects of force level and direction and all interac-
tions were not significant (P ⬎ 0.05). Figure 2 illustrate
force-time curves and their first derivative for a typical
control study participant and for one typical study par-
ticipant of each LBP subgroup. Each curve is an average
of 10 consecutive flexion trials, and all curves were syn-
chronized on force onset (vertical dashed line).
The first derivative of force shows not only a longer
duration of force production but also a different control
strategy for the LBPlong than for the LBPshort and healthy
study participants. The dF/dt curve for the LBPlong shows
that the impulse was characterized by two peaks, a small
initial peak followed by a more important peak dF/dt.
Some LBPlong study participants showed three peaks for
a number of trials.
No group difference was observed for peak force vari-
ability (6.6, 7.7, and 8.8 Nm for the LBPlong, LBPshort,
and healthy study participants, respectively; P ⬎ 0.05)
and absolute errors (11.1, 11.9, and 14.1 Nm for the
LBPlong, LBPshort, and healthy study participants, respec-
tively; P ⬎ 0.05). Peak force variability was more impor-
tant for the extension than for the flexion forces (8.5 Nm
vs. 6.7 Nm; F1,28 ⫽ 6.84, P ⫽ 0.014). Similarly, absolute
errors were more important for the extension than
for the flexion forces (14.4 Nm vs. 10.7 Nm; F1,28 ⫽
8.40, P ⫽ 0.007). The main effects of force level and
all interactions were not significant for both variables
(P ⬎ 0.05).
As mentioned above, linear regressions between peak
dF/dt and peak force were also computed to examine if
the first peak of dF/dt could predict the peak force
achieved. High r2 indicates that the first peak of dF/dt
could predict the peak force; this can be taken as an
indication of a preprogrammed or open-loop mode of
control. A lower r2 indicates less stereotypical force re-
sponses. On average, peak dF/dt explained 94.2% and
97.0% of the variance observed in peak force for the
control study participants and the LBPshort subgroup,
respectively. The r2 for the LBPlong study participants
was 84.4. A one-way analysis of variance showed that
this difference was statistically significant (F2,28 ⫽ 6.31,
P ⫽ 0.005). Hence, LBPlong study participants were less
consistent in regulating their force rise time.
Discussion
In the current study, isometric force production param-
eters among patients with chronic nonspecific LBP and
healthy control study participants were evaluated to de-
termine if there were any differences between these pop-
ulations. Differences in the temporal characteristics of
Force Production Parameters • Descarreaux et al 315
trunk isometric force production were found within the
initial LBP group. Consequently, two different LBP sub-
groups were defined: one subgroup exhibited time to
peak force values much higher than the control group,
and a second subgroup showed time to peak values sim-
ilar to the control group.
The control group and the LBPshort subgroup demon-
strated high r2 values between the first peak of dF/dt and
the peak force achieved. In other words, the control
study participants and LBP short study participants
showed force parameters that reflected a preplanned
force response characterized by a pulse height control.17
The lower r2 values for the LBPlong study participants
suggest that these study participants used a rise time reg-
ulation strategy. For control study participants, such a
strategy has been suggested to help in reducing response
variability.17,18 As noted above, the LBPlong subgroup
scored lower pain level at the time of testing compared
with the LBPshort subgroup (43 mm vs. 14 mm). At the
present, it is difficult to draw any direct association be-
tween pain levels at the time of testing and the force rise
times observed; it is nevertheless possible that the longer
force rise times for these study participants emerged
from a direct strategy to control and limit pain. Pain at
the time of testing, however, does not always account for
the pattern of sensory pain perception experienced by the
study participants over a defined period of time.4 For
instance, Al-Obaidi et al19 reported results supporting
the hypothesis that spinal physical capacity in chronic
LBP is not explained solely by the sensory perception of
pain at the time of testing. In their experiment, the antic-
ipation of pain and the fear-avoidance belief about phys-
ical activities better predicted the variation in spinal ex-
tension isometric strength than the pain intensity at the
time of testing. An alternative cognitive strategy (fear of
pain) aimed at limiting the rate of force development by
trunk muscles could have been used by LBPlong study
participants to avoid potential pain during the experi-
mentation. Since all study participants with LBP pro-
duced maximal voluntary contractions similar to the
control group and since none showed increased pain dur-
ing the experiment, we suggest that increased pain or fear
of pain related to the experiment is probably not an im-
portant factor explaining the group differences. Oddsson
et al20,21 have proposed that the observed changes in the
trunk muscular activity (median frequency and EMG
root mean square amplitude) could result from subtle
postural adjustments that were developed during the
acute phase to avoid pain. These changes would eventu-
ally become “normal” behavior for the individuals with
chronic LBP.20,21 Similar sensorimotor adaptations in
the force production parameters could explain the differ-
ences observed between the two LBP subgroups. Further
studies are needed to explore the possible acquisition of
different control strategies for the production of trunk
isometric force.
As shown by Hodges,9 changes in the organization of
feedforward postural responses of trunk muscles in study
316 Spine • Volume 29 • Number 3 • 2004
participants with LBP could be explained by alterations
in motor planning rather than by alterations of the de-
scending drive. Such alterations could explain the differ-
ences in force rise times observed in LBPlong subgroup.
While our data on isometric force production parameters
cannot point out the neurophysiological origin of the
observed differences between LBP subgroups, adaptation
to chronic pain, presence of pain at the time of testing,
and modification in motor planning are plausible mech-
anisms explaining the observed behavior.
Maximal voluntary trunk flexion and extension were
similar for the three groups. These results contrast with
those of Alston et al22 and Thorstensson and Arvidson23
who observed reduced trunk muscle maximal voluntary
contractions in patients with LBP. For both groups, how-
ever, maximal trunk extension force was higher than
maximal trunk flexion force. This seems to be a general
feature of healthy and LBP populations.24 Peak force
variability and absolute errors were greater for extension
forces than flexion forces. When expressed in percentage
of peak force, however, peak force variability was simi-
lar for both positions. The differences observed are a
normal feature of isometric force contractions where
peak force variability increases with higher generated
forces.16 Furthermore, there are no significant position
differences in absolute errors when they are expressed as
a percentage of the peak force. With the exception of
maximal voluntary contraction, we can conclude that
isometric spatiotemporal characteristics were similar for
both trunk extension and flexion.
It is important to note, however, that our study in-
cluded only study participants with chronic nonspecific
low back. It represents only a portion of the total LBP
population. Our results cannot be extended to more
complicated LBP such as LBP with radicular symptoms,
severe osteoarthritis, or inflammatory disorders of the
lumbar spine. Further investigations would be needed to
explore whether such patients show similar adaptations
in the production of trunk isometric forces.
Conclusion
Our results demonstrate that all patients with LBP were
able to produce isometric forces with levels of accuracy
similar to control study participants. The patients with
LBP with less pain at the time of testing, however, took
more time to reach the peak forces. The longer time-to-
peak force and the smaller percentage of variance ob-
served for the linear regressions suggest that some LBP
study participants modified their feedforward force pro-
duction parameters and adopted a control mode that
was less “open-loop.” Future studies will be needed to
document the force production parameters throughout
the evolution of LBP as well as the general effect of pro-
ducing less impulsive trunk forces on lower back symp-
toms and functional status. Appropriate clinical evalua-
tion of the motor control changes and adaptations
related to low back injuries could be useful in identifying
the stage of the condition as well as in the prescription of
appropriate rehabilitation protocols.
Key Points
● Force production parameters were measured in
LBP and healthy subjects.
● Some LBP subjects used a different motor control
strategy to produce isometric trunk forces.
● Adaptation to chronic pain, presence of pain at
the time of testing and modification in motor plan-
ning are plausible mechanisms explaining the ob-
served behavior.
● Future studies will be needed to document the
force production parameters throughout the evolu-
tion of LBP.
References
1. Frymoyer JW, Ducker TB. The Adult Spine: Principles and Practice, 2nd ed.
Philadelphia: Lippincott-Raven, 1997.
2. Waddell G. Volvo award in clinical sciences: a new clinical model for the
treatment of low-back pain. Spine. 1987;12:632– 644.
3. Marras WS. Occupational low back disorder causation and control. Ergo-
nomics. 2000;43:880 –902.
4. Croft PR, Macfarlane GJ, Papageorgiou AC, et al. Outcome of low back pain
in general practice: a prospective study. Br Med J. 1998;316:1356 –1359.
5. Mientjes MI, Frank JS. Balance in chronic low back pain patients compared
to healthy people under various conditions in upright standing. Clin Bio-
mech. 1999;14:710 –716.
6. Hodges P, Richardson C, Jull G. Evaluation of the relationship between
laboratory and clinical tests of transversus abdominis function. Physiother
Res Int. 1996;1:30 – 40.
7. Taimela S, Kankaanpaa M, Luoto S. The effect of lumbar fatigue on the
ability to sense a change in lumbar position: a controlled study. Spine. 1999;
24:1322–1327.
8. Newcomer KL, Laskowski ER, Yu B, et al. Differences in repositioning error
among patients with low back pain compared with control subjects. Spine.
2000;25:2488 –2493.
9. Hodges PW. Changes in motor planning of feedforward postural responses
of the trunk muscles in low back pain. Exp Brain Res. 2001;141:261–266.
10. Indahl A, Kaigle AM, Reikeras O, et al. Interaction between the porcine
lumbar intervertebral disc, zygapophysial joints, and paraspinal muscles.
Spine. 1997;22:2834 –2840.
11. Dreyer SJ, Dreyfuss PH. Low back pain and the zygapophysial (facet) joints.
Arch Phys Med Rehabil. 1996;77:290 –300.
12. Roberts S, Eisenstein SM, Menage J, et al. Mechanoreceptors in interverte-
bral discs: morphology, distribution, and neuropeptides. Spine. 1995;20:
2645–2651.
13. Avramov AI, Cavanaugh JM, Ozaktay CA, et al. The effects of controlled
mechanical loading on group-II, III, and IV afferent units from the lumbar
facet joint and surrounding tissue: an in vitro study. J Bone Joint Surg Am.
1992;74:1464 –1471.
14. Macfarlane GJ, Thomas E, Papageorgiou AC, et al. Employment and phys-
ical work activities as predictors of future low back pain. Spine. 1997;22:
1143–1149.
15. Ghez C, Gordon J. Trajectory control in targeted force impulses: I. Role of
opposing muscles. Exp Brain Res. 1987;67:225–240.
16. Newell KM, Carlton LG. Force variability in isometric responses. J Exp
Psychol Hum Percept Perform. 1988;14:37– 44.
17. Gordon J, Ghez C. Trajectory control in targeted force impulses: II. Pulse
height control. Exp Brain Res. 1987;67:241–252.
18. Newell KM, Carlton LG, Hancock PA. Kinetic analysis of response variabil-
ity. Psychol Bull. 1984;96:133–151.
19. Al-Obaidi SM, Nelson RM, Al-Awadhi S, et al. The role of anticipation and

fear of pain in the persistence of avoidance behavior in patients with chronic
low back pain. Spine. 2000;25:1126 –1131.
20. Oddsson LI, De Luca CJ. Activation imbalances in lumbar spine muscles in
the presence of chronic low back pain. J Appl Physiol. 2003;94:1410 –1420.
21. Oddsson LI, Giphart JE, Buijs RJ, et al. Development of new protocols and
analysis procedures for the assessment of LBP by surface EMG techniques. J
Rehabil Res Dev. 1997;34:415– 426.
22. Alston W, Carlson KE, Feldman DJ, et al. A quantitative study of muscle
Force Production Parameters • Descarreaux et al 317
factors in the chronic low back syndrome. J Am Geriatr Soc. 1966;14:1041–
1047.
23. Thorstensson A, Arvidson A. Trunk muscle strength and low back pain.
Scand J Rehabil Med. 1982;14:69 –75.
24. Takemasa R, Yamamoto H, Tani T. Trunk muscle strength in and effect of
trunk muscle exercises for patients with chronic low back pain: the differ-
ences in patients with and without organic lumbar lesions. Spine. 1995;20:
2522–2530.