Trunk Response Analysis under Sudden Forward Perturbations Using a Kinematics-
B. Bazrgari1; A. Shirazi-Adl1 and C. Larivière2
1. Department of Mechanical Engineering, École Polytechnique, Montréal, Canada
2. Occupational Health and Safety Research Institute Robert-Sauvé (IRSST), and
Center for Interdisciplinary Research in Rehabilitation of Greater Montreal,
Rehabilitation Institute, Montreal, Canada
Address correspondence to:
Professor, Department of Mechanical Engineering, Ecole Polytechnique,
P.O. Box 6079, station “centre-ville”,
Montreal, Quebec, Canada H3C 3A7
Tel: 514-3404711 Ext 4129;
Correspondence and reprint requests to
Christian Larivière, Ph.D.
Institut de recherche Robert-Sauvé en santé et en sécurité du travail (IRSST)
505, boul. De Maisonneuve Ouest
Montréal (Québec), H3A 3C2
Phone: (514) 288-1551 #217
Fax: (514) 288-6097
Accurate quantification of the trunk transient response to sudden loading is crucial in
prevention, evaluation, rehabilitation and training programs. An iterative dynamic kinematics-
driven approach was used to evaluate the temporal variation of trunk muscle forces, internal
loads and stability under sudden application of an anterior horizontal load. The input
kinematics is hypothesized to embed basic dynamic characteristics of the system that can be
decoded by our kinematics-driven approach. The model employs temporal variation of applied
load, trunk forward displacement and surface EMG of select muscles measured on two healthy
and one chronic low-back pain subjects to a sudden load. A finite element model accounting
for measured kinematics, nonlinear passive properties of spine, detailed trunk musculature with
wrapping of global extensor muscles, gravity load and trunk biodynamic characteristics is used
to estimate the response under measured sudden load. Results demonstrate a delay of ~ 200 ms
in extensor muscle activation in response to sudden loading. Net moment and spinal loads
substantially increase as muscles are recruited to control the trunk under sudden load. As a
result and due also to the trunk flexion, system stability significantly improves. The reliability
of the kinematics-driven approach in estimating the trunk response while decoding measured
kinematics is demonstrated. Estimated large spinal loads highlight the risk of injury that likely
further increases under larger perturbations, muscle fatigue and longer delays in activation.
Key Words: Sudden load, Trunk rotation, Muscle force, Spinal loads, Stability, Finite
Sudden loading is recognized as a risk factor in back pain (Lavender et al., 1989;
Marras et al., 1987) and as such has been the subject of investigation on the stabilizing roles of
neural and muscular mechanisms (Brown et al., 2003; Cholewicki et al., 2000; Cresswell et al.,
1994; Gardner-Morse and Stokes, 2001; Moorhouse and Granata, 2007). Under unexpected
loads, the central nervous system has a tendency to overshoot while attempting to reflexively
compensate for deviations between desired and actual kinematics (Horak et al., 1989; Zeinali-
Davarani et al., 2008) thereby increasing the likelihood of excessive spinal loads. The
voluntary and reflexive muscle activations as well as corresponding intrinsic and reflexive
stiffness values under sudden loading conditions differ between back patients and healthy
individuals (Radebold et al., 2000; Reeves et al., 2008; Stokes et al., 2006) and alter as a result
of rehabilitation (Wilder et al., 1996) and training (Pedersen et al., 2004; Pedersen et al., 2007).
A delayed muscle reflexive response, for example, could be a significant predictor (risk factor)
of future back injuries rather than only an indication of a prior injury (Cholewicki et al., 2005).
Design of effective prevention as well as treatment and rehabilitation procedures in
management of back disorders draw substantial benefits from an improved understanding of
trunk response under sudden perturbations.
Adequate dynamic stability of the spine is essential to safeguard the entire system
against injuries especially in unexpected loading environments. Trunk stability is provided by
passive, intrinsic and reflexive subsystems (Panjabi, 1992). Contribution of the passive
subsystems (i.e. disc, ligaments and facets) has been demonstrated to increase at greater trunk
flexions (Arjmand and Shirazi-Adl, 2006; Bazrgari et al., 2008d; Granata and Wilson, 2001)
and compression loads (Shirazi-Adl, 2006; Stokes and Gardner-Morse, 2003) whereas it
diminishes at near neutral standing postures (El-Rich et al., 2004; Granata and Wilson, 2001).
Antagonistic pre-activation of trunk muscles increases the trunk intrinsic stiffness thereby
decreasing the perturbation magnitude under sudden loads (Andersen et al., 2004; Brown and
McGill, 2008; Cholewicki et al., 2000; Krajcarski et al., 1999; Vera-Garcia et al., 2006). It
hence yields higher effective stiffness and a more stable system but at an increased metabolic
cost (Franklin and Granata, 2007) and loads on the spine passive tissues (Arjmand et al., 2008;
Bazrgari et al., 2008d; Vera-Garcia et al., 2006). Dynamic system-identification approaches
with the entire trunk idealized as a single-degree-of-freedom system have been used to estimate
trunk stiffness (Cholewicki et al., 2000; Gardner-Morse and Stokes, 2001; Lawrence et al.,
2006) as well as reflex feedback and intrinsic contributions (Moorhouse and Granata, 2007) in
post disturbance periods by matching measured kinematics in subjects under sudden loads.
Nevertheless, the relative contributions of various components in trunk stiffness require further
Infeasibility of direct measurement of muscle forces and spinal loads along with
limitations and difficulties associated with their indirect measurements promote the use of
biomechanical models for estimation of spinal loads and stability (Bazrgari, 2008). Previous
studies on the trunk response in sudden loading have, however, focused on the measurement of
muscle electromyographic (EMG) activities (Radebold et al 2000, Wilder et al 1996) with
nearly no reference to the generated internal spinal loads. We have developed a kinematics
driven approach for the estimation of trunk transient response (i.e., muscle forces, spinal loads
and system stability) in which measured kinematics are prescribed into a nonlinear finite
element model of the passive-active trunk that accounts for the external/gravity forces,
inertia/damping loads and muscle forces. The method has successfully been applied for the
study of trunk response under static (Arjmand, 2007; El-Rich, 2005) and transient (Bazrgari,
2008; Bazrgari et al., 2008b; Bazrgari et al., 2008d) loading conditions. The trunk biodynamics
under the sudden release of a posteriorly applied load has also been investigated (Bazrgari et
al., 2008c). In continuation, the current study aims to quantify muscle forces, spinal loads and
trunk stability under the sudden application of an anterior horizontal force in neutral trunk
posture. In this study, using a sudden loading paradigm, measured kinematics and suddenly
applied forces of three subjects (two asymptomatic and one low-back pain patient) with height
and mass values matching those in the model are prescribed as input data while their recorded
EMG data are used for qualitative validation of model predictions. It is hypothesized that, by
prescribing into the model the measured time-dependent kinematics of the trunk under sudden
loading, the kinematics driven approach can accurately capture the hidden transient
characteristics of the system while estimating the trunk response. In other words, the trunk
transient kinematics embeds the essential signatures governing trunk mechanical performance
and as such could be employed to detect component performances within the overall response.
The data of three male subjects, two healthy and one with chronic low-back pain
(CLBP) (see Table 1), from an earlier (yet unpublished) experiment on 31 healthy and 33
CLBP subjects were chosen for the present modeling study. These subjects were selected due
only to the close match of their trunk height and total body mass with the corresponding values
(47 cm and 74 kg) in our model as the objective of this work was not to evaluate the influence
of CLBP on the response. The CLBP was defined as an almost daily non-specific lumbosacral
pain lasting for at least three months. In measurements, individuals were unexpectedly
subjected to an anteriorly applied force at the T8 level while measuring trunk kinematics and
EMG activity of a number of superficial trunk muscles (see Fig. 1). Subjects, while seated in a
sudden loading apparatus with their pelvis constrained, were instructed to keep a reference
upright “zero” position and to minimize the abdominal co-activity before loading. A sudden
load, obstructed by a metal screen, was applied randomly within 2 to 6 seconds after subjects
held the reference position. The magnitudes of pre-load and sudden load were set for each
subject based on their trunk and head inertial properties. Each subject performed 6 preparatory
perturbation trials to familiarize with the protocol and the feedback system, followed after a 10
min rest period by 20 perturbations separated by a one min rest period.
The EMG signals from three pairs of back muscles and two pairs of abdominal muscles
were collected (bandpass filter at 20-450 Hz, pre-amplification gain at 1000, sampling rate at
1024 Hz) with active surface electrodes (Delsys Inc., MA); multifidus (MF) at the L5 level,
iliocostalis lumborum (IC) at the L3 level, longissimus (LG) at the L1 level, rectus abdominis
(RA) and external oblique (EO) as detailed elsewhere (Larivière et al., 2001). Trunk kinematics
was simultaneously recorded using a potentiometer (Fig. 1).
Table-1 Data on the three subjects used in the current model studies.
Subject – 1
Subject – 2
Subject – 3
Model -- 74
24 76 190 47.5
22 74 178 45
48 73 180 48
Figure 1: Experimental setup. The force is applied via a harness at the T8 level and is recorded using a load cell
located between the cable and the harness in front of the subjects while a potentiometer measures trunk
displacement. The sudden load system is released by a noiseless electro-magnet allowing the hooks from the pre-
load and sudden-load, separated by ~ 1 cm, to transmit the load to the trunk. Visual feedbacks of EMG raw data
from right external oblique (EO) and rectus abdominus (RA) are used to assist subjects in minimizing, as
instructed, coactivities in abdominal muscle.
The kinematics driven approach uses a finite element model that accounts for nonlinear
passive properties of the ligamentous spine, dynamic characteristics of the trunk (i.e. mass,
mass moment of inertia and damping), detailed muscle architecture, wrapping of the global
extensor muscles and satisfaction of equilibrium at all spinal levels and directions (Bazrgari,
2008; Bazrgari et al., 2008d). The finite element model includes six nonlinear beam elements
representing the stiffness of T12-S1 lumbar motion segments and seven rigid elements
simulating lumbosacral vertebrae (L1-S1). Nonlinear direction-dependent mechanical
properties of beam elements are chosen to represent the passive properties of motion segments
(including discs, ligaments and facet joints) in different directions. Thorax-neck-head
assembly (C1-T12) is represented by one rigid element. Concentrated mass and mass moments
of inertia for trunk segmental dynamic properties and connector elements are employed to
simulate linear/angular inter-segmental damping as well as buttocks’ nonlinear stiffness and
damping (see Fig. 2).
Figure 2: Trunk model used in the study including global and local musculatures in the sagittal plane (on the
right), frontal plane (in the middle, fascicles on one side are shown) and vertebral column (on the left). ICpl:
iliocostalis lumborum pars lumborum, ICpt: iliocostalis lumborum pars thoracic, IP: iliopsoas, LGpl: longissimus
thoracis pars lumborum, LGpt: longissimus thoracis pars thoracic, MF: multifidus, QL: quadratus lumborum, IO:
internal oblique, EO: external oblique, RA: rectus abdominus. The pelvis and sacrum are restrained in all
Measured kinematics data (i.e. sagittal rotations) along with external and gravity loads
are applied onto the model and the nonlinear transient equations of motion are solved using
implicit integration algorithm (time increment of 0.1 ms) with unconditionally stable Hilber-
Hughes-Taylor integration operator (Hilber et al., 1978). This provides reaction moments (i.e.
net moments minus passive ligamentous moments) at all levels where rotations are prescribed.
To partition a reaction moment among muscles attached to a specific level, an optimization
approach with the cost function of minimum sum of quadratic muscle stresses is employed at
each spinal level to circumvent the redundancy. In a comparative study on various optimisation
cost functions, those of sum of cubed and quadratic muscle stresses were found plausible
yielding results comparable with EMG activities and disc pressure measurements (Arjmand
and Shirazi-Adl, 2005b). Moreover, muscle forces are bound to remain greater than their
passive force components (calculated based on instantaneous length and a tension-length
relationship (Davis et al., 2003)) while remaining smaller than the sum of foregoing passive
components plus maximum active forces (i.e., 0.6 MPa times muscle’s physiological cross-
sectional area, pcsa (Winter, 2005)).
Trunk rotation at the T12, verified to yield the measured trunk translation at the T8
level, is partitioned among lumbar vertebrae based on the reported ratios in the literature (i.e.
11.5%, 15%, 14%, 18%, 21.5%, 20 % from the T12-L1 to L5-S1 level, respectively) (Bazrgari,
2008). The pelvis is constrained in accordance with measurements. At each time step,
calculated muscle forces at different levels along with the external loads are re-applied to the
model and the procedure is repeated till the convergence is reached. To investigate the system
stability, trunk muscles are then replaced by uniaxial elements assuming a linear stiffness-force
relation (i.e. k=q F/L) in which the muscle stiffness is proportional to the instantaneous muscle
force, F, and inversely proportional to its current length, L, with q as a dimensionless muscle
stiffness coefficient taken to be the same for all muscles (Bergmark, 1989). Under different q
values, trunk stability is examined at all loaded deformed configurations by both natural
frequency and linear perturbation analyses. In this manner, the minimum (critical) muscle
stiffness coefficient, q, below which the structure becomes unstable, is determined at all time
steps. Finite element analyses are performed using ABAQUS (Simulia Inc., Providence, RI,
Version 6.5) while the optimization procedure is analytically solved using an in-house program
based on the Lagrange Multiplier method.
Measured/applied loads and resulting trunk motions at the T8 level (Fig. 3),
predicted/measured muscle forces/activities (Fig. 4) along with estimated peak loads at
different lumbar spine levels (Table 2) are given for all three subjects considered in this work.
Due however to the similarity in trends among all three subjects, the focus hereafter is placed
on the response of one subject (i.e., subject 1). In measurements of the subject 1, the horizontal
constant preload of 61 N is suddenly increased to reach, after some fluctuations, to ~ 155 N as
seen in Fig. 3 (left) where corresponding measured horizontal translation is also depicted. The
predicted anterior translation, velocity, and acceleration profiles at the T8 vertebra, calculated
by a forward dynamic analysis (i.e. unconstrained model subject to the forgoing measured load
at the T8 with muscle forces remaining unchanged from their initial values calculated under
preload and gravity) deviate from those of measurements at ~ 220, 180 and 160 ms,
respectively (with corresponding periods of ~ 200, 180 and 110 ms in subject 2 and ~220, 180
and 140 ms in subject 3). These instances point to the initiation of muscle reflexive
intervention to control trunk forward movement under sudden load (Fig. 3, top). Measurements
indicate latency periods for different muscles ranging between ~ 50 to 220 ms.
Figure 3: Measured temporal variation of the sudden load and trunk horizontal translation along with estimated
trunk translation in forward dynamic (FD) analyses for different subjects. Note that the results for the isometric
pre-perturbation periods that last between 2 to 6 seconds would be the same as those at t = 0 s. Moreover, the FD
analyses were performed for a short post-perturbation period only to delineate the reflexive response.
Figure 4: Temporal variation of measured muscle activities (left) and predicted muscle forces (right) in extensor
muscles for different subjects. Abdominal muscles (not shown) remain nearly silent with negligible activities.
Predicted temporal variations of forces in global extensor muscles are in good
qualitative agreement with measured activities of extensor muscles with cross-correlation
values of 0.41 (0.11) between predicted forces and measured activities in the IC (LG) as well
as cross-correlation of 0.62 between predicted force in LG and measured muscle activity in MF
(R=1 indicates identical curves) (Fig. 4, top). Cross-correlation values between predicted IC
(LG) forces and measured IC (LG) activities increase to 0.7 (0.9) when a delay of 0.2 s is
considered in measurement data. Cross-correlations are even higher in other two subjects.
Table- 2 Peak required moment (balanced by muscle forces), wrapping contact forces and
internal local spinal loads at disc mid-height at different levels.
SubjectT12 L1 L2 L3 L4 L5
1 52.9 17.6 15.8 15.7 17.8 19.8
2 59.8 20.1 17.9 17.0 20.8 22.4
3 48.3 17.7 18.1 18.3 18.8 18.9
1 0 68 101 98 122 92
2 0 73 108 92 106 64
3 0 111 129 125 150 134
T12-L1 L1-L2 L2-L3 L3-L4 L4-L5 L5-S1
1 1218 18282267 26493142 3525
1410 2212 270131203704 4132
395 426 249 417 347 1182
30.0 28.1 25.1 21.5 20.2 17.2
41.0 40.6 39.0 35.2 31.4 26.1
Net external moment at the S1 level reaches its peak value of 143.8 Nm at ~ 0.54 s with
relative contributions from the external force (40 %), trunk gravity (44 %) and inertia (16 %)
(Fig. 5). This moment, on the other hand, is internally resisted primarily by active (65%) and
passive (23 %) components of extensor muscle forces with a smaller contribution from the
ligamentous spine (12 %) (Fig. 5). Temporal variations of required moments (i.e. net moment
minus ligamentous moment) at the other lumbar levels, T12 to L5, follow similar trends with
peak values listed for all three subjects in Table 2. Under the load, the trunk reaches the
maximum T12-S1 rotation of 39.5° that results in wrapping of global extensor muscles over the
deformed spine thereby activating reaction forces at different levels in contact (see Table 2 and
Fig. 6). Internal spinal loads along with the net external moment increase caudally reaching
their maximum at the lowermost L5-S1 level (Table 2). Compression and shear forces at the
L5-S1 are generated primarily due to forces in muscles with much smaller contributions from
gravity and inertia (Fig. 7). As far as the stability is concerned, the trunk is less stable at the
beginning and end periods of the task while much more stable at the middle due to associated
greater muscle forces and trunk flexion angle (Fig. 7).
Figure 5: Temporal variation of the net external moment at the S1 level in the model generated by various loading
components (top) and the associated internal moment resisted by different components (bottom) for the subject 1.
Figure 6: Initial (in presence of preload before the release of sudden load) and deformed (under sudden load at the
time of maximum sagittal rotation) configurations of the spine along with corresponding global extensor muscles
accounting for the wrapping at the deformed position for the subject 1.
Figure 7: Predicted temporal variations of spinal loads as well as relative contributions of muscle forces, inertia,
gravity and external load at the L5-S1 disc mid-height in local directions (top and middle) plus the temporal
variation of the minimum (critical) muscle stiffness coefficient q (bottom) for the subject 1. The system is much
more stable during the periods with larger muscle activities and trunk flexion.
Sudden perturbation of the trunk, apart from its role as a back pain risk factor
(Lavender et al., 1989; Marras et al., 1987), has attracted attention for its likely potential in
discriminating healthy individuals from chronic low back pain patients (Radebold et al., 2000;
Reeves et al., 2008; Stokes et al., 2006) and in the evaluation of rehabilitation and training
programs (Pedersen et al., 2004; Pedersen et al., 2007; Wilder et al., 1996). An improved
assessment of the trunk response and parameters affecting it under sudden loads, beyond the
measurement of few EMG activities and kinematics response, is beneficial in contributing to
more effective management of spinal disorders. This study, in line with an earlier one (Bazrgari
et al., 2008c), confirms our hypothesis on the suitability of the kinematics-driven approach in
reliable estimation of the trunk transient response under sudden loads. The current model is
also the first to estimate muscle forces, spinal loads and trunk stability under a transient sudden
loading condition. In accordance with our objectives, measured force and translation data
collected on three subjects with spinal height and weight values closely matching those in our
model were used in the present study. Incorporation of measured data on remaining male-
female and healthy-patient subjects could be considered in future works requiring the
modification of the model to match, as closely as possible, the individual subjects under
consideration. Lack of appropriate data such as the subject-specific values for passive
properties of the ligamentous spine and muscles, amongst others, will as expected hinder such
The kinematics-based approach employs as input data the measured kinematics while
realistically accounting at the same time for the nonlinear passive properties of the spine and
muscles, complex musculature with wrapping of global extensor muscles, gravity/external
loads as well as dynamic characteristics of the trunk. By using an iterative finite element
model, predictions satisfy equations of motion at different levels and directions while assessing
trunk stability margin at each instance of time. The satisfaction of equilibrium simultaneously
at all levels and of trunk stability requirements are issues that are completely overlooked in
commonly-used single-level biomechanical model studies (see Arjmand et al., 2007). The
kinematics-driven approach, in contrast to the EMG-driven methods, does not either require as
input data the collection of EMG activities of trunk muscles that exist in the model. It is hence
much more practical for evaluation of various regular daily, occupational and sportive
activities. Moreover, by employing the measured kinematics, the model ensures the mechanical
consistency between the applied loads, model characteristics and posture.
The forward dynamic analysis performed just for a short duration at the beginning
under the sudden load with muscle forces remaining constant at their pre-sudden loading levels
yielded anterior linear displacement, velocity and acceleration values at the T8 level that
matched measurements very well for the first ~160-220 ms (Fig. 3). The dynamic balance
between external (i.e., gravity, inertia and applied perturbation) and internal (i.e., active/passive
muscle forces along with passive spine) loads (Fig. 5) during this initial period results in
computation of nearly constant muscle forces (Fig. 4). Any additional muscle forces during this
period, either reflexively or voluntarily, would have otherwise altered trunk kinematics away
from that predicted in forward dynamics analysis. Reflexive muscle response at ~160 ms
became visually detectable initially in the acceleration profile (not shown) and later on in the
associated velocity and displacement profiles (Fig. 3). Following this delay period during
which initial muscle forces remained nearly unchanged at their preload levels, the forward
dynamics response deviated more and more from the measured response as the muscle
reflexive activation kicked in to control the response. These observations demonstrate once
more the relative strength of the kinematics-driven approach in estimating the trunk response
while decoding input measured kinematics that contain the hidden system signatures. The more
accurate the measured kinematics is, the more reliable will the predictions become. It should be
noted that the predicted delay in additional activation of muscles after the sudden loading
accounts both for the latency period in muscles’ stretch reflex and the electromechanical delay
(EMD) between muscle activity and generated muscle force (Cavanagh and Komi, 1979; van
Dieen et al., 1991). Apart from the electromechanical delay, a direct comparison with recorded
EMG activities shown in Fig. 4 should not overlook the likely effects of filtering process on the
onset of activities.
At the beginning and ending periods of the task, the net external moment was resisted
primarily by activation in local and global extensor muscles (Fig. 5). In between, however, and
under relatively large trunk flexion, this relative contribution of active muscle forces dropped
in favor of that of passive structures. Further increases in the contribution of passive muscle
forces would have relieved the foregoing activity resulting in flexion-relaxation of the extensor
muscles had subjects undergone greater lumbar forward flexion (Bazrgari et al., 2008d). It
should be noted that, in accordance with earlier studies (Jorgensen et al., 2003; Macintosh et
al., 1993), a 10% reduction in initial lever arm of global extensor muscles was allowed before
the occurrence of contact during wrapping between the spine and global extensor muscles.
Spinal loads followed the same trend as that of the net moment with the maximum
occurring at the lowermost L5-S1 level (Fig. 7). While the first peak in the sudden load (Figs. 3
and 5) was counter-balanced mainly by the inertial forces during the forward trunk acceleration
at the onset of sudden loading, its second peak coincided with the trunk deceleration phase due
to the recruitment of muscles. These latter activities in muscles along with the increased net
moment due to the gravity resulted in a substantial increase in compression and shear forces at
different spinal levels; this situation could further deteriorate in presence of longer delay in
activation (Solomonow et al., 2008), larger disturbances and muscle fatigue (Granata et al.,
2004; Herrmann et al., 2006). Despite negligible direct contribution on spinal loads (Fig. 7) of
external force, gravity and inertia as compared with muscle forces, they had considerable
indirect contributions of, respectively at the time of maximum loading, 40, 44 and 16 % in the
net external moment that in turn generated muscle forces. Comparison with the reported tissue
tolerance thresholds of 4-6 kN in compression (Jager and Luttmann, 1991), > 1kN in shear
(Cyron et al., 1976; Miller et al., 1986) and 70 Nm in moment (Miller et al., 1986; Osvalder et
al., 1993) demonstrates the risk of injury.
The existing pre-activity in extensor muscles in response to the preload before the
sudden loading yielded an initially more stable spine with critical (minimum) muscle stiffness
coefficient of 18 as compared to 60 required under no preloads. After the sudden loading and
as time progressed, trunk stability substantially increased (i.e., critical q decreases) with
reflexive activity in extensor muscles (Moorhouse and Granata, 2007) and greater trunk flexion
(Arjmand and Shirazi-Adl, 2006; Bazrgari et al., 2008d). Trunk stability deteriorated (i.e.,
critical q increases) once more at the returning phase due to smaller extensor muscle activities
and trunk flexion. The linear relationship between muscle stiffness and total muscle force
(Bergmark, 1989) considered in the current study did not differentiate between the passive,
intrinsic and reflexive components of the muscle force and stiffness.
In summary, the kinematics-driven approach was found reliable in decoding the trunk
characteristics from prescribed measured kinematics in order to estimate the trunk transient
response under sudden loads. Spinal loads increased substantially under sudden loads due to
the compensatory muscle activities. This indicates, hence, the risk of tissue injury associated
with sudden trunk loading conditions especially at larger perturbations, activation delays and
load magnitudes. Results indicate that reflexive muscle actions kick in after a delay period
following the onset of the sudden load. As a result and due also to the large trunk flexion, the
trunk stability substantially improves. Accurate estimation of muscle forces, spinal loads and
trunk stability is crucial in proper management of spinal disorders and in suggestion for
prevention, evaluation, rehabilitation and training programs.
ACKNOWLEDGMENTS: This work is supported by the NSERC-Canada.
Conflict of Interest: None.
Andersen, T.B., Essendrop, M., Schibye, B., 2004. Movement of the upper body and muscle
activity patterns following a rapidly applied load: the influence of pre-load alterations.
Eur J Appl Physiol 91, 488-92.
Arjmand, N., 2007. Computational biomechanics of the human spine in static lifting tasks. PhD
Thesis, Ecole Polytechnique, Montreal (Canada).
Arjmand, N., Shirazi-Adl, A., 2005a. Biomechanics of changes in lumbar posture in static
lifting. Spine 30, 2637-48.
Arjmand, N., Shirazi-Adl, A., 2005b. Sensitivity of kinematics-based model predictions to
optimization criteria in static lifting tasks. Med Eng Phys 28, 504-14.
Arjmand, N., Shirazi-Adl, A., 2006. Model and in vivo studies on human trunk load
partitioning and stability in isometric forward flexions. J Biomech 39, 510-521.
Arjmand, N., Shirazi-Adl, A., Parnianpour, M., 2007. Trunk biomechanical models based on
equilibrium at a single-level violate equilibrium at other levels. Eur Spine J 16, 701-9.
Arjmand, N., Shirazi-Adl, A., Parnianpour, M., 2008. Relative efficiency of abdominal
muscles in spine stability. Comput Methods Biomech Biomed Engin 11, 291-9.
Bazrgari, B., 2008. Biodynamic of the human spine. Mechanical Engineering. PhD Thesis,
Ecole Polytechnique, Montreal (Canada).
Bazrgari, B., Shirazi-Adl, A., Kasra, M., 2008a. Computation of trunk muscle forces, spinal
loads and stability in whole-body vibration. Journal of Sound and Vibration 318, 1334-
Bazrgari, B., Shirazi-Adl, A., Kasra, M., 2008b. Seated whole body vibrations with high-
magnitude accelerations--relative roles of inertia and muscle forces. J Biomech 41,
Bazrgari, B., Shirazi-Adl, A., Parnianpour, M., 2008c. Transient Analysis of Trunk Response
in Sudden Release Loading Using Kinematics-Driven Finite Element Model. Clin
Biomech (in press).
Bazrgari, B., Shirazi-Adl, A., Trottier, M., Mathieu, P., 2008d. Computation of trunk
equilibrium and stability in free flexion-extension movements at different velocities. J
Biomech 41, 412-21.
13 Download full-text
Bergmark, A., 1989. Stability of the lumbar spine. A study in mechanical engineering. Acta
Orthop Scand Suppl 230, 1-54.
Brown, S.H., Haumann, M.L., Potvin, J.R., 2003. The responses of leg and trunk muscles to
sudden unloading of the hands: implications for balance and spine stability. Clin
Biomech (Bristol, Avon) 18, 812-20.
Brown, S.H., McGill, S.M., 2008. The intrinsic stiffness of the in vivo lumbar spine in
response to quick releases: Implications for reflexive requirements. J Electromyogr
Cavanagh, P.R., Komi, P.V., 1979. Electromechanical delay in human skeletal muscle under
concentric and eccentric contractions. Eur J Appl Physiol Occup Physiol 42, 159-63.
Cholewicki, J., Silfies, S.P., Shah, R.A., Greene, H.S., Reeves, N.P., Alvi, K., Goldberg, B.,
2005. Delayed trunk muscle reflex responses increase the risk of low back injuries.
Spine 30, 2614-20.
Cholewicki, J., Simons, A.P.D., Radebold, A., 2000. Effects of external trunk loads on lumbar
spine stability. J Biomech 33, 1377-1385.
Cresswell, A.G., Oddsson, L., Thorstensson, A., 1994. The influence of sudden perturbations
on trunk muscle activity and intra-abdominal pressure while standing. Exp Brain Res
Cyron, B.M., Hutton, W.C., Troup, J.D., 1976. Spondylolytic fractures. J Bone Joint Surg Br
Davis, J., Kaufman, K.R., Lieber, R.L., 2003. Correlation between active and passive isometric
force and intramuscular pressure in the isolated rabbit tibialis anterior muscle. J
Biomech 36, 505-12.
El-Rich, M., 2005. Modeling of the human thoraco-lumbar spine in standing posture: Analyses
of the spinal load and stability (in French). PhD Thesis, Ecole Polytechnique, Montreal
El-Rich, M., Shirazi-Adl, A., Arjmand, N., 2004. Muscle activity, internal loads, and stability
of the human spine in standing postures: combined model and in vivo studies. Spine 29,
Franklin, T.C., Granata, K.P., 2007. Role of reflex gain and reflex delay in spinal stability--a
dynamic simulation. J Biomech 40, 1762-7.
Gardner-Morse, M.G., Stokes, I.A., 2001. Trunk stiffness increases with steady-state effort. J
Biomech 34, 457-63.
Granata, K.P., Slota, G.P., Wilson, S.E., 2004. Influence of fatigue in neuromuscular control of
spinal stability. Hum Factors 46, 81-91.
Granata, K.P., Wilson, S.E., 2001. Trunk posture and spinal stability. Clin Biomech 16, 650-9.
Herrmann, C.M., Madigan, M.L., Davidson, B.S., Granata, K.P., 2006. Effect of lumbar
extensor fatigue on paraspinal muscle reflexes. J Electromyogr Kinesiol 16, 637-41.
Hilber, H.M., Hughes, T.J.R., Taylor, R.L., 1978. Collocation, dissipation and overshoot for
time integration schemes in structural dynamics. Erthquake Engineering and Structural
Dynamics 6, 99-117.
Horak, F.B., Diener, H.C., Nashner, L.M., 1989. Influence of central set on human postural
responses. J Neurophysiol 62, 841-53.
Jager, M., Luttmann, A., 1991. Compressive strenght of lumbar spine elements related to age,
gender, and other influencing factors. In: Anderson, P.A., Hobart, D.J., Danoff, J.V.
(Eds.), Electromyographical Kinsiology, Elsevier Science, 291-294
Jorgensen, M.J., Marras, W.S., Gupta, P., Waters, T.R., 2003. Effect of torso flexion on the
lumbar torso extensor muscle sagittal plane moment arms. Spine J 3, 363-9.
Krajcarski, S.R., Potvin, J.R., Chiang, J., 1999. The in vivo dynamic response of the spine to
perturbations causing rapid flexion: effects of pre-load and step input magnitude. Clin
Biomech (Bristol, Avon) 14, 54-62.