Persons with lower-limb amputation have impaired trunk postural control while
maintaining seated balance
Brad D. Hendershota, Maury A. Nussbauma,b,*
aVirginia Tech – Wake Forest School of Biomedical Engineering and Sciences, Virginia Tech, Blacksburg, VA 24061, USA
bIndustrial and Systems Engineering, Virginia Tech, Blacksburg, VA 24061, USA
Low back pain (LBP) represents a substantial secondary
impairment among persons with lower-limb amputation (LLA)
. LBP prevalence is substantially higher among persons with LLA
(52–71%) compared to the general population (6–33%) [2,3]. More
than half (52%) of persons with LLA report experiencing at least one
back pain episode in the prior month, 25% describe their LBP as
constant, and 31% describe their pain as severe . LBP may even
be more bothersome than residual limb and phantom limb pain,
two pain sources considered major contributors to post-amputa-
tion morbidity [2,3]. While the problem of LBP among persons with
LLA has been well documented, the underlying causes in this
population remain largely unknown. Given the high prevalence
and debilitating nature of LBP among persons with LLA, it is
important to understand the mechanisms of LBP onset and
recurrence in this population.
Notable alterations and asymmetries in gait following LLA have
been described, including larger or more prolonged forces
generated by and transmitted through the contralateral limb
during the stance phase of gait . Such altered and asymmetric
movement patterns have been associated with increased three-
dimensional trunk kinematics compared to able-bodied controls
[e.g., 5], suggesting that abnormal mechanics of movement
resulting from LLA may increase stability demands on the spinal
column and/or alter existing postural control mechanisms and
neuromuscular responses. Further, changes in spinal posture (e.g.,
lordosis, scoliosis) and/or muscle architecture, sometimes ob-
served in persons with LLA [e.g., 6,7], may also influence such
responses. Despite these alterations/adaptations in gait and
posture, the effects of LLA and subsequent prosthetic use on
trunk postural control and spinal stability are not well understood.
Maintaining spinal stability requires efficient and synergistic
responses from passive structures and active neuromuscular
control . The aforementioned alterations in gait and movement
with LLA may result in new spinal loading patterns, rates, and
magnitudes, which could chronically alter motor control strategies
and functional properties of the passive spine. Numerous studies
have demonstrated disturbances in passive and active trunk
behaviors of healthy persons following acute exposure to
Gait & Posture 38 (2013) 438–442
A R T I C L E
I N F O
Received 5 June 2012
Received in revised form 8 January 2013
Accepted 10 January 2013
Low back pain
A B S T R A C T
Abnormal mechanics of movement resulting from lower-limb amputation (LLA) may increase stability
demands on the spinal column and/or alter existing postural control mechanisms and neuromuscular
responses. A seated balance task was used to investigate the effects of LLA on trunk postural control and
stability, among eight males with unilateral LLA (4 transtibial, 4 transfemoral), and eight healthy, non-
amputation controls (matched by age, stature, and body mass). Traditional measures derived from
center of pressure (COP) time series, and measures obtained from non-linear stabilogram diffusion
analyses, were used to characterize trunk postural control. All traditional measures of postural control
(95% ellipse area, RMS distance, and mean velocity) were significantly larger among participants with
LLA. Non-linear stabilogram diffusion analyses also revealed significant differences in postural control
among persons with LLA, but only in the antero-posterior direction. Normalized trunk muscle activity
was also larger among participants with LLA. Larger COP-based sway measures among participants with
LLA during seated balance suggest an association between LLA and reduced trunk postural control.
Reductions in postural control and spinal stability may be a result of adaptations in functional tissue
properties and/or neuromuscular responses, and may potentially be caused by repetitive exposure to
abnormal gait and movement. Such alterations could then lead to an increased risk for spinal instability,
intervertebral motions beyond physiological limits, and pain.
? 2013 Elsevier B.V. All rights reserved.
* Corresponding author at: Department of Industrial and Systems Engineering,
Virginia Tech, 250 Durham Hall (0118), Blacksburg, VA 24061, USA.
Tel.: +1 540 231 6053; fax: +1 540 231 3322.
E-mail address: email@example.com (M.A. Nussbaum).
Contents lists available at SciVerse ScienceDirect
Gait & Posture
jo u rn al h om ep age: ww w.els evier.c o m/lo c ate/g aitp os t
0966-6362/$ – see front matter ? 2013 Elsevier B.V. All rights reserved.
mechanical loading and atypical postures, suggesting an associa-
tion between altered trunk behaviors, reduced postural control,
and the occurrence of LBP. For example, prolonged static or
dynamic trunk flexion reduces passive support of the spine , and
whole-body vibration exposure can compromise trunk proprio-
ception . Mechanical or sensory deficits in passive tissues can
also lead to decreased muscle force output, reduced propriocep-
tion, and inhibition of stretch reflexes . Thus, repeated
exposures to abnormal mechanics of motion could lead to similar,
but chronic, alterations in trunk behaviors, and subsequent
reductions in trunk postural control and spinal stability among
persons with LLA.
Measures of the center of pressure (COP) during seated balance
have been used to assess trunk postural control in both healthy
individuals and LBP patients, and impaired trunk postural control
has been associated with spinal instability and LBP [e.g., 12]. While
LLA is associated with deficits in whole-body postural control
during quiet standing , these are likely a result of lost
musculature in the lower limb (e.g., at the ankle joint). During
quiet upright stance, postural adjustments can be made using a
variety of responses through the ankle, knee, hip, and lumbosacral
joints. It was therefore anticipated that performing a task not
requiring the lower limbs, where inherent differences and
asymmetries are present among persons with LLA, a better
understanding of the effects of LLA on trunk postural control
and spinal stability could be obtained. Therefore, the goal of the
present work was to investigate trunk postural control among
persons with LLA during a seated stability task. It was hypothe-
sized that persons with LLA would have impaired trunk postural
control compared to non-amputation controls, evidenced by
increases in COP-based seated sway measures, and suggesting a
decrement in spinal stability and the potential for increased risk of
low back injury.
Eight males with unilateral LLA (4 transtibial, 4 transfemoral)
and eight male, non-amputation controls participated (Table 1).
The most frequent reason for amputation was trauma (5), followed
by congenital deformity or abnormality (2), and cancer (1). The
mean (SD) duration of prosthetic use among the LLA group was
12.1 (10.1) years. Members of the control group were recruited to
match participants with LLA, at the individual level, in terms of age,
stature, and body mass (within <8 year, <5 cm, and <5 kg,
respectively). Inclusion criteria for participants with LLA, consis-
tent with previous biomechanical studies [14,15], were: (1) adults
with a unilateral above- or below-knee amputation; (2) regular/
daily use of prosthesis (?1 year post-amputation/rehabilitation);
and, (3) independent locomotion without the use of walking aids.
All participants completed the short, self-administered version of
the International Physical Activity Questionnaire (IPAQ) , and
were categorically identified as moderately active. Potential
participants (in both groups) were excluded if they had any
recent history (6 months) of falls, neurologic deficits, or any
underlying musculoskeletal disorders (not including amputation)
that could confound the results. In particular, none of the
participants in the study had low back pain at the time of testing.
Each participant completed initial informed consent procedures
approved by the Virginia Tech Institutional Review Board.
Participants with LLA wore their prosthetic device during all
2.2. Experimental design and procedures
Seated balance was tested using an unstable chair (Fig. 1A) that
pivots on a low-friction ball-and-socket joint. Adjustments to the
seat allow for the participant’s center of mass to be centered over
the ball-and-socket. Four springs are placed circumferentially, in
each cardinal direction, to provide supplemental support .
These springs can be adjusted inward/outward (7–22 cm) from the
center, thereby facilitating control of task difficulty by altering the
resistance of the chair to rotation (Fig. 1B). Following calibration
procedures [17,18], the spring positions were converted to a
percentage of the gravitation gradient (5G) for each participant
seated on the chair. The value of 5G determines the mass (or
weight) distribution of the participant on the chair, with 100% 5G
specifying spring positions that will fully equilibrate the gravita-
tional gradient (i.e., facilitate seated stability with no need for
participant compensation). Here, the task difficulty (i.e., spring
positions) was standardized to 60% 5G for all four springs. A
similar setting has been used previously, and is a difficulty level
sufficient for discerning differences in seated sway measures
between groups or exposure conditions [17,19]. It is slightly
conservative, however, since it was not known a priori to what
degree postural control would be degraded among persons with
LLA. Pelvic motions were minimized using a belt placed across the
hips, and an adjustable footrest limited motion of the lower-limbs
and kept the knees and hips at ?908 angles.
Initially, participants performed seated maximum voluntary
contractions (MVC) in trunk flexion, extension, and left/right
lateral bending. These were done in a separate fixture, with a
custom chest harness connected to a fixed anchor via a rigid rod,
allowing participants to make maximal efforts in the desired
direction. During MVCs, electromyographic (EMG) activities of the
bilateral lumbar (L3) erector spinae, rectus abdominis, and
external oblique muscles were recorded using bipolar Ag/AgCl
surface electrodes, and following existing electrode placement
protocols . Initially, the skin was prepared using abrasion and
cleaned with alcohol, and inter-electrode impedance was main-
tained below 10 KV. Raw EMGs were preamplified (100?) near the
collection site, bandpass filtered (10–500 Hz), amplified, and
converted to RMS in hardware (Measurement Systems Inc., Ann
Arbor, MI, USA), then sampled at 1000 Hz. Peak EMG-RMS values
were identified, and used subsequently for normalization (see
Participants were given five initial practice trials to reduce
learning effects and acclimate to the task , which involved
maintaining seated balance on the chair using (primarily) lumbar
spine motion. Participants then completed three seated balance
trials, and were instructed to keep the chair surface as level as
possible while sitting with an upright posture (no slouching), eyes
open and looking straight ahead, and arms folded across their
Mean (SD) participant characteristics in the lower-limb amputation (LLA) and control groups. Reported p values represent group comparisons (pooled
transtibial + transfemoral vs. control) from unpaired t tests. TTA: transtibial, TFA: transfemoral.
TTA (n = 4)
TFA (n = 4)
LLA (n = 8)
Control (n = 8)
Body mass (kg)
B.D. Hendershot, M.A. Nussbaum / Gait & Posture 38 (2013) 438–442
chest. Each trial lasted 65 s, and at least two minutes of rest was
provided between each to minimize fatigue. During trials, EMGs
were collected (as during MVCs) and triaxial reaction forces and
moments were recorded (1000 Hz) using a force platform (AMTI,
OR6-7-1000, Watertown, MA, USA) located beneath the chair/
spring assembly. The latter were low-pass filtered (second-order,
zero-phase-lag, Butterworth, 10 Hz cut-off frequency) and pro-
cessed to obtain COP time series in the antero-posterior (A-P) and
medio-lateral (M-L) directions.
2.3. Dependent measures and statistical analysis
The initial 10 s and final 5 s of each seated balance trial were
removed prior to analysis to avoid initial adjustments and
anticipatory effects, respectively, and all COP data were de-
meaned prior to analyses. Postural control was then determined
using several traditional measures derived from COP time series:
95% ellipse area (cm2), RMS distance (cm), and mean velocity (cm/
s) . These measures were chosen based on evidence of
relatively good within- and between-day reliability [e.g., 18 and
22], and higher values are typically interpreted as indicating
inferior (or deteriorated) postural control [21,23]. To supplement
these, local dynamic stability was assessed using non-linear
stabilogram diffusion analyses, which characterizes a control
system as two parts, involving both short-term, open-loop control
and long-term, closed-loop control [23,24]. A scaling exponent, H,
evaluates correlations between past and future displacements (of
COP), and is computed from the slope of mean square COP
displacements versus time intervals (log–log). A critical point, CP,
represents the transition from short-term to long-term control.
However, since only the short-term exponent (Hs) has been shown
to differentiate between participant groups [12,18] and long-term
stability parameters are not representative of a bounded system
, long-term parameters were not used in the present study. The
short-term scaling exponent (Hs) and the critical point coordinate
[CP; time (s) and amplitude (cm2)] were calculated, separately in
the A-P and M-L directions.
All EMG responses were normalized to peak levels from MVCs,
and mean EMG activity within each seated balance trial was
obtained for each muscle. Values for bilateral pairs of muscles were
combined, since initial analyses indicated no bilateral differences.
To address the study hypothesis, one-way analyses of variance
(ANOVAs) were used to compare the COP-based seated sway
measures (separately for each direction) between participants
with LLA and controls. Muscle activities from one participant with
LLA were excluded from these analyses due to measurement
errors, and no violations of parametric models assumptions were
evident. Initial analyses, using mixed-factor ANOVAs, revealed no
significant differences in any of the dependent measures between
participants with transtibial and transfemoral amputations, and
thus all participants with LLA were analyzed as a single group.
There were no significant or substantial differences (i.e., that might
indicate learning effects) across the three replications, with one
exception: mean rectus abdominis values were ?5% higher in the
second of the three replications. All statistical analyses were
performed using JMP (Version 9, SAS Institute Inc., Cary, NC, USA)
with statistical significance set at p < 0.05.
All traditional measures of postural control (95% ellipse area,
RMS distance, and mean velocity) were significantly larger among
participants with LLA (Table 2). RMS distances were larger in the A-
P direction for both groups. Hs was significantly larger among
controls in the A-P direction, but similar in the M-L direction.
CPtime and amplitude were respectively longer and larger among
participants with LLA in the A-P direction, but both were similar
between groups in the M-L direction (Table 3). Mean normalized
RMS muscle activities were larger among participants with LLA
(Fig. 2) in the L3 erector spinae (p < 0.0001), rectus abdominis
(p < 0.0001), and external oblique (p = 0.0045).
The present study investigated trunk postural control in
persons with LLA during a seated stability task, and hypothesized
that it would be impaired among participants with LLA compared
to non-amputation controls. All traditional and non-linear
Fig. 1. (A) Unstable chair with pelvic strap and adjustable foot rest, rigidly attached above a force platform. (B) Base of the unstable chair showing the central low-friction ball-
and-socket joint and the adjustable A-P and M-L springs.
Mean (SD) center of pressure sway measures of seated balance among participants
with LLA and matched controls. p values represent group comparisons (pooled
transtibial + transfemoral vs. control), and bolded values indicate significant
95% Ellipse area (cm2)
RMS distance – AP (cm)
RMS distance – ML (cm)
Mean velocity – AP (cm/s)
Mean velocity – ML (cm/s)
B.D. Hendershot, M.A. Nussbaum / Gait & Posture 38 (2013) 438–442
COP-based sway measures were larger among participants with
LLA during seated balance (with the exception of Hs; see below),
supporting our hypothesis and suggesting an association between
LLA and reduced trunk neuromuscular control and spinal stability.
Increases in traditional COP-based sway measures among
participants with LLA suggest a less tightly controlled postural
control system in both the A-P and M-L directions (Table 2). Non-
linear stabilogram diffusion analyses revealed similar differences
in postural control in the A-P direction between groups, but similar
control in the M-L direction. Comparable findings have been
observed in seated balance performance between patients with
LBP and healthy controls, suggesting that the larger range of spinal
motion in the sagittal plane better discriminates deficits in
postural control between groups . Although both groups here
displayed persistent short-term movements (Hs> 0.5) in both
directions, participants with LLA had significantly less persistence
in the A-P direction. Less persistence implies that deviations in a
time series are less likely to be followed by subsequent deviations
in the same direction . Given the increases in traditional COP-
based sway measures, such reductions in persistence suggest a less
healthy system .
Further, CPamplitudes for participants with LLA were nearly
double in the A-P direction compared to controls, suggesting that
the thresholds for closed-loop corrective mechanisms are much
larger among participants with LLA [23,24]. Likewise, longer CP
times among participants with LLA, and between the A-P and M-L
directions, suggest increased delays in corrective responses to
postural perturbations (i.e., transition from open-loop to closed-
loop control). According to Collins and De Luca , CPamplitude
may be determined physiologically by a proprioceptive ‘‘dead
zone’’, whereby slight variations in trunk position and orientation
are left uncorrected. Similar to the interpretation provided by
Radebold et al.  with respect to LBP patients, we speculate that
participants with LLA generated larger COP movements during
short-term intervals to overcome larger sensory thresholds. This
interpretation is consistent with the current CPamplitudes, and
could indicate larger spinal neutral zones among participants with
LLA. This zone is a region of intervertebral motion where passive
tissues provide little resistance to movement . Increases in the
neutral zone, which may occur with losses of passive stiffness,
muscle weakness, and/or inhibition of stabilizing musculature,
have been associated with spinal instability and risk of injury .
Increased trunk muscle recruitment can decrease postural
control during seated balance , and is related to decreased
postural control and spinal stability among persons with LBP .
Here, participants with LLA had higher levels of superficial trunk
muscle activity during seated balance compared to controls (?12%
MVC vs. 5% MVC). Of note, the observed levels among controls are
consistent with data from earlier studies [29,31]. When the spinal
neutral zone is increased (as suggested above) the spinal column
may become unstable, and require increasing compensations from
active trunk muscles. Such increases among participants with LLA
could therefore indicate an attempt to stiffen the trunk, perhaps to
compensate for reductions in passive contributions to spinal
stability. Similar increases in muscle responses have been
associated with aging or neurologic disease and suggested to be
a response to decreased confidence in balance ability [e.g., 32,33].
Regardless of origin, a stiffer trunk caused by increased trunk
muscle activities could consequently result in overcorrections in
posture during seated balance, and thus lead to the observed
increases in traditional COP sway measures. Decreases in
proprioception and/or reductions in reflexive muscle responses
could similarly explain the need for increased muscle activation
among persons with LLA, though future work is needed to
determine if such decreases are indeed present.
4.1. Study limitations
We attributed changes in postural control directly to aspects
related to spinal stability (i.e., passive trunk stiffness and active
neuromuscular responses). However, participants with LLA could
have inherent differences in other sensory systems (e.g., vestibular
or visual) that affect seated balance performance. Despite
constraining the pelvis, balance strategies originating in the hips
could have also influenced seated stability performance. Specifi-
cally, reduced use of the hip flexors may have contributed to the
observed alterations in seated stability among persons with LLA.
Further, proximal strength and stability training are often
emphasized during post-amputation rehabilitation . It is
unclear whether the observed increases in trunk muscle activities
are a result of such training or rather are compensating for
reductions in passive contributions to spinal stability. Lower back
injuries are possible with altered gait and movement biomechanics
occurring with LLA, because of the linkages between the pelvis,
torso, and lower limbs . However, the current study could not
discriminate if such injuries are causes or consequences of reduced
postural control and spinal stability among persons with LLA.
Challenges in recruiting persons with LLA made it difficult to
obtain a large homogeneous group with respect to age, reason for
amputation, amputation level, and the duration of prosthetic use.
The small sample size may limit generalizability of the results, and
group heterogeneity could explain some of the variability in
responses among participants with LLA. Future work should
explore these potential relationships through more systematic
recruitment criteria with better access to such a population.
Nevertheless, non-amputation control participants were carefully
matched, which should have limited important potential sources
Mean (SD) non-linear stabilogram diffusion analysis parameters of seated balance
among participants with LLA and matched controls. p values represent group
comparisons (pooled transtibial + transfemoral vs. control), and bolded values
indicate significant differences.
Erecto r Spin ae
tus Abdominis Extern al O
Normalized Muscle Activity(%MVC)
Fig. 2. Normalized trunk muscle activities during seated balance tasks. Error bars
indicate standard deviations, and the symbol * indicates a significant difference
between groups (LLA vs. control).
B.D. Hendershot, M.A. Nussbaum / Gait & Posture 38 (2013) 438–442
In summary, all traditional and non-linear COP sway measures Download full-text
were larger among participants with LLA during seated balance
(with the exception of Hs), suggesting an association between LLA
and reduced trunk postural control. Reduced postural control and
spinal stability may be a result of adaptations in functional tissue
properties and/or neuromuscular responses, and could be caused
by repetitive exposure to abnormal gait and movement. Such
alterations could then lead to an increased risk for spinal
instability, intervertebral motions beyond physiological limits,
Source of support: None.
Conflict of interest statement
 Gailey R, Allen K, Castles J, Kucharik J, Roeder M. Review of secondary physical
conditions associated with lower-limb amputation and long-term prosthesis
use. Journal of Rehabilitation Research and Development 2008;45:15–29.
 Ehde DM, Smith DG, Czerniecki JM, Campbell KM, Malchow DM, Robinson LR.
Back pain as a secondary disability in persons with lower limb amputations.
Archives of Physical Medicine and Rehabilitation 2001;82:731–4.
 Smith DG, Ehde DM, Legro MW, Reiber GE, del Aguila M, Boone DA. Phantom
limb, residual limb, and back pain after lower extremity amputations. Clinical
Orthopaedics and Related Research 1999;361:29–38.
 Sagawa Jr Y, Turcot K, Armand S, Thevenon A, Vuillerme N, Watelain E.
Biomechanics and physiological parameters during gait in lower-limb ampu-
tees: a systematic review. Gait and Posture 2011;33:511–26.
 Goujon-Pillet H, Sapin E, Fode ´ P, Lavaste F. Three-dimensional motions of trunk
and pelvis during transfemoral amputee gait. Archives of Physical Medicine
and Rehabilitation 2008;89:87–94.
 Burke MJ, Roman V, Wright V. Bone and joint changes in lower limb amputees.
Annals of the Rheumatic Diseases 1978;37:252–4.
 Gaunaurd I, Gailey R, Hafner BJ, Gomez-Marin O, Kirk-Sanchez N. Postural
asymmetries in transfemoral amputees. Prosthetics and Orthotics Interna-
 Panjabi MM. The stabilizing system of the spine. Part I. Function, dysfunction,
adaptation, and enhancement. Journal of Spinal Disorders 1992;5:383–9.
 McGill SM, Brown S. Creep response of the lumbar spine to prolonged full
flexion. Clinical Biomechanics 1992;7:43–6.
 Li L, Lamis F, Wilson SE. Whole-body vibration alters proprioception in the
trunk. International Journal of Industrial Ergonomics 2008;38:792–800.
 Solomonow M. Ligaments: a source of musculoskeletal disorders. Journal of
Bodywork and Movement Therapies 2009;13:136–54.
 Radebold A, Cholewicki J, Polzhofer GK, Greene HS. Impaired postural control
of the lumbar spine is associated with delayed muscle response times in
patients with chronic idiopathic low back pain. Spine 2001;26:724–30.
 Rougier PR, Bergeau J. Biomechanical analysis of postural control of persons
with transtibial or transfemoral amputation. American Journal of Physical
Medicine and Rehabilitation 2009;88:896–903.
 Segal AD, Orendurff MS, Czerniecki JM, Schoen J, Klute GK. Comparison of
transtibial amputee and non-amputee biomechanics during a common turn-
ing task. Gait and Posture 2011;33:41–7.
 Vrieling AH, van Keeken HG, Schoppen T, Otten E, Hof AL, Halbertsma JPK, et al.
Balance control on a moving platform in unilateral lower limb amputees. Gait
and Posture 2008;28:222–8.
 Craig CL, Marshall AL, Sjo ¨stro ¨m M, Bauman AE, Booth ML, Ainsworth BE, et al.
International physical activity questionnaire: 12-country reliability and va-
lidity. Medicine and Science in Sports and Exercise 2003;35:1381–95.
 Slota GP, Granata KP, Madigan ML. Effects of seated whole-body vibration on
postural control of the trunk during unstable seated balance. Clinical Biome-
 Lee H, Granata KP. Process stationarity and reliability of trunk postural
stability. Clinical Biomechanics 2008;23:735–42.
 Tanaka ML, Nussbaum MA, Ross SD. Evaluation of the threshold of stability for
the human spine. Journal of Biomechanics 2009;42:1017–22.
 Granata KP, Wilson SE. Trunk posture and spinal stability. Clinical Biomechan-
 Prieto TE, Myklebust JB, Hoffmann RG, Lovett EG, Myklebust BM. Measures of
postural steadiness: differences between healthy young and elderly adults.
IEEE Transactions on Biomedical Engineering 1996;43:956–66.
 Lin D, Seol H, Nussbaum MA, Madigan ML. Reliability of COP-based postural
sway measures and age-related differences. Gait and Posture 2008;28:
 Norris JA, Marsh AP, Smith IJ, Kohut RI, Miller ME. Ability of static and
statistical mechanics posturographic measures to distinguish between age
and fall risk. Journal of Biomechanics 2005;38:1263–72.
 Collins JJ, De Luca CJ. Open-loop and closed-loop control of posture: a random-
walk analysis of center-of-pressure trajectories. Experimental Brain Research
 Delignie `res D, Deschamps T, Legros A, Caillou N. A methodological note on
nonlinear time series analysis: is the open-and closed-loop model of Collins
and De Luca (1993) a statistical artifact? Journal of Motor Behavior
 Dingwell JB, Cusumano JP. Re-interpreting detrended fluctuation analyses of
stride-to-stride variability in human walking. Gait and Posture 2010;32:
 Peng C-K, Buldyrev S, Hausdorff J, Havlin S, Mietus J, Simons M, et al. Non-
equilibrium dynamics as an indispensable characteristic of a healthy biologi-
cal system. Integrative Physiological and Behavioral Science 1994;29:283–93.
 Panjabi MM. The stabilizing system of the spine. Part II. Neutral zone and
instability hypothesis. Journal of Spinal Disorders 1992;5:390–7.
 Reeves NP, Everding V, Cholewicki J, Morrisette D. The effects of trunk stiffness
on postural control during unstable seated balance. Experimental Brain Re-
 Hodges P, van den Hoorn W, Dawson A, Cholewicki J. Changes in the mechan-
ical properties of the trunk in low back pain may be associated with recur-
rence. Journal of Biomechanics 2009;42:61–6.
 Preuss RA, Grenier SG, McGill SM. Postural control of the lumbar spine in
unstable sitting. Archives of Physical Medicine and Rehabilitation 2005;
 Carpenter MG, Allum JHJ, Honegger F, Adkin AL, Bloem BR. Postural abnor-
malities to multidirectional stance perturbations in Parkinson’s disease. Jour-
nal of Neurology Neurosurgery and Psychiatry 2004;75:1245–54.
 Melzer I, Benjuya N, Kaplanski J. Age-related changes of postural control: effect
of cognitive tasks. Gerontology 2001;47:189–94.
 Esquenazi A, DiGiacomo R. Rehabilitation after amputation. Journal of the
American Podiatric Medical Association 2001;91:13–22.
 Dananberg H. Gait style as an etiology to chronic postural pain. Part II. Postural
compensatory process. Journal of the American Podiatric Medical Association
B.D. Hendershot, M.A. Nussbaum / Gait & Posture 38 (2013) 438–442