Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The footprint of pain: investigating persistence of altered trunk control in recurrent low back pain
(USC Thesis Other)
The footprint of pain: investigating persistence of altered trunk control in recurrent low back pain
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE FOOTPRINT OF PAIN:
INVESTIGATING PERSISTENCE OF ALTERED TRUNK
CONTROL IN RECURRENT LOW BACK PAIN
By
Hai-Jung Steffi Shih
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOKINESIOLOGY)
AUGUST 2020
Copyright 2020 Hai-Jung Steffi Shih
ii
DEDICATION
To mom, dad, my grandparents, aunt, family, and friends,
who tolerated my absence and believed in me.
iii
ACKNOWLEDGEMENTS
I would like to express my profound gratitude to the many individuals who have
contributed to the completion of this dissertation. It has been an honor to work intimately with so
many outstanding and kind human beings throughout my five years in the program.
I would like to thank my extraordinary committee members who are not only content
experts in their respective fields, but also true mentors that are deeply invested in my
professional growth. Dr. Kornelia Kulig, chair of the committee and my academic advisor, who
always allowed me to explore unconventional ideas, created countless professional opportunities
and connections, and guided me on the road of becoming an independent scientist. Dr. James
Gordon, whose visionary insight and warmth was my greatest inspiration. Dr. Carolee Winstein,
whose wisdom and enthusiasm always motivated me to explore new perspectives on data. Dr.
Jason Kutch, whose genius and creativity facilitated breakthroughs when I was stuck. Last but
not least, Dr. Linda Van Dillen, whose clinical-oriented perspective and profound understanding
of low back pain research helped tremendously with my execution and interpretation of this
dissertation, despite the handicap of long-distance meetings.
My peers in the Division of Biokinesiology and Physical Therapy, and past and present
members of the Jacquelin Perry Musculoskeletal Biomechanics Laboratory have been the
greatest resource and support, serving simultaneously as my Encyclopedias and family away
from home. I have truly learned so much from these wonderful individuals, and would like to
especially thank Michael Rowley, Abbi Fietzer, Jo Armour Smith, Eugene Chang, and Danielle
Jarvis, for being role models and for generously sharing their experience with me as they’ve
walked the path I have been walking. I am also extremely lucky to share learning experiences
with undergrad, masters of science, and doctor of physical therapy students when serving as their
iv
research mentor and teaching assistant. I would not have made it this far if teaching had not been
such an integral part of the program’s training.
I would like to thank my parents, my family, and my friends scattered around the world
for being my biggest cheerleaders. This dissertation would not have been possible without them.
Finally, I would like to acknowledge my funding source, the International Society of
Biomechanics, for their generosity that made this project possible, and the division for matching
the amount. I would also like to thank all the participants for their contribution to this project.
v
TABLE OF CONTENTS
DEDICATION............................................................................................................................... ii
ACKNOWLEDGEMENTS ........................................................................................................ iii
LIST OF TABLES ...................................................................................................................... vii
LIST OF FIGURES ..................................................................................................................... ix
ABSTRACT ................................................................................................................................ xiv
CHAPTER I: Overview ................................................................................................................ 1
CHAPTER II: Background and Significance ............................................................................ 5
The role of the trunk in lateral stability during gait ................................................................... 5
Low back pain and its recurring symptoms as a worldwide health issue ................................... 8
Could alterations in trunk control lead to future symptoms? ................................................... 10
Could alterations in attention become detrimental to daily tasks? ........................................... 13
Significance of this multidisciplinary study in the prevention of low back pain recurrence .... 15
CHAPTER III: Trunk Control at Various Step Widths during Walking ............................ 18
Abstract ..................................................................................................................................... 18
Introduction ............................................................................................................................... 19
Methods ..................................................................................................................................... 21
Results ....................................................................................................................................... 27
Discussion ................................................................................................................................. 36
CHAPTER IV: Trunk Control In and Out of an Episode of Recurrent Low Back Pain .... 41
Abstract ..................................................................................................................................... 41
Introduction ............................................................................................................................... 43
Methods ..................................................................................................................................... 46
Results ....................................................................................................................................... 54
Discussion ................................................................................................................................. 59
CHAPTER V: Attentional Prioritization and Trunk Control Under Dual-Task................. 64
Abstract ..................................................................................................................................... 64
Introduction ............................................................................................................................... 66
Methods ..................................................................................................................................... 68
vi
Results ....................................................................................................................................... 74
Discussion ................................................................................................................................. 79
CHAPTER VI: Summary and Conclusions ............................................................................. 84
REFERENCES ............................................................................................................................ 92
APPENDIX A: Chapter III Detailed Statistical Results ....................................................... 112
APPENDIX B: Chapter IV Step Width Effect Statistical Results ....................................... 129
APPENDIX C: Chapter IV Stability of Performance in Control Participants .................. 130
APPENDIX D: Chapter V Practice Effect Analyses ............................................................. 131
APPENDIX E: Pressure Pain Threshold Testing .................................................................. 133
vii
LIST OF TABLES
Table III.1 Test-retest reliability and SEM for primary variables (ICC: intra-
class correlation coefficient, SD: standard deviation, SEM: standard
error of the measurement).
36
Table IV.1 Participant demographics (mean ± standard deviation). 54
Table IV.2 Mean ± standard deviation (range) for low back pain characteristics.
ODI: Oswestry Disability Index; FABQ-PA: Fear Avoidance Beliefs
Questionnaire – Physical Activity Subscale; FABQ-PA: Fear
Avoidance Beliefs Questionnaire –Work Subscale; VAS: Visual
Analog Scale. Significant p-values are in bold.
54
Table V.1 Mean ± standard deviation (range) for low back pain characteristics.
ODI: Oswestry Disability Index; FABQ-PA: Fear Avoidance Beliefs
Questionnaire – Physical Activity Subscale; FABQ-W: Fear
Avoidance Beliefs Questionnaire –Work Subscale; VAS: Visual
Analog Scale. Significant p-values are in bold.
75
Table A.1 Summary of p-values for pairwise comparisons between step widths
for step width error, trunk excursion, trunk coordination, and peak
longissimus activation and bilateral longissimus co-activation.
111
Table A.2 Pairwise comparisons between step widths for constant step width
error (ANOVA F=12.96, p<0.001).
118
Table A.3 Pairwise comparisons between step widths for variable step width
error (ANOVA F=4.65, p=0.0021).
118
Table A.4 Pairwise comparisons between step widths for transverse plane
thorax excursion (ANOVA F=28.50, p<0.0001).
119
Table A.5 Pairwise comparisons between step widths for transverse plane
pelvis excursion (ANOVA F=4.238, p=0.004).
119
Table A.6 Pairwise comparison between step widths for transverse plane trunk
excursion (ANOVA F=11.50, p=<0.0001).
120
Table A.7 Pairwise comparisons between step widths for sagittal plane pelvis
excursion (ANOVA F=12.977, p<0.001).
120
Table A.8 Pairwise comparisons between step widths for sagittal plane trunk
excursion (ANOVA F=5.534, p<0.001).
121
viii
Table A.9 Pairwise comparisons between step widths for transverse plane anti-
phase coordination (ANOVA F=3.014, p=0.023).
121
Table A.10 Pairwise comparisons between step widths for transverse plane
thorax-only coordination (ANOVA F=3.830, p=0.007).
122
Table A.11 Pairwise comparisons between step widths for transverse plane
pelvis-only coordination (ANOVA F=4.955, p=0.001).
122
Table A.12 Pairwise comparison between step widths for frontal plane in-phase
coordination (ANOVA F=5.007, p=0.001).
123
Table A.13 Pairwise comparisons between step widths for frontal plane thorax-
only coordination (ANOVA F=3.171, p=0.018).
123
Table A.14 Pairwise comparisons between step widths for sagittal plane in-
phase coordination (ANOVA F=27.432, p<0.001).
124
Table A.15 Pairwise comparisons between step widths for sagittal plane anti-
phase coordination (ANOVA F=22.432, p<0.001).
124
Table A.16 Pairwise comparisons between step widths for sagittal plane thorax-
only coordination (ANOVA F=9.349, p<0.001).
125
Table A.17 Pairwise comparisons between step widths for sagittal plane thorax-
only coordination (ANOVA F=2.591, p=0.043).
125
Table A.18 Pairwise comparisons between step widths for fight peak
longissimus activation (ANOVA F=9.049, p<0.001).
126
Table A.19 Pairwise comparisons between step widths for left peak longissimus
activation (ANOVA F=9.318, p<0.001).
126
Table A.20 Pairwise comparison between step widths for bilateral longissimus
co-activation (ANOVA F=8.491, p<0.001).
127
Table B.1 Step width effects statistical results from general mixed-effects
models. CTRL: Back-healthy controls; rLBP-A: recurrent low back
pain group in active pain; rLBP-R: recurrent low back pain group in
remission; rLBP-pooled: pooled data across two testings for
recurrent low back pain group (only when there was no pain effect
between rLBP-A and rLBP-R).
128
Table E.1 Mean ± standard deviation values of pressure pain threshold in the
back-healthy control group (CTRL), the recurrent low back pain
group during a painful episode (rLBP-A), and recurrent low back
pain group during symptom remission (rLBP-R).
132
ix
LIST OF FIGURES
Figure I.1 Framework of the specific aims for this dissertation. 3
Figure II.1 Conceptual model of the extrapolated center of mass and margin of
stability for determining mediolateral foot placement. v: velocity of
the center of mass; xCoM: extrapolated center of mass (prediction of
center of mass position at subsequent heel strike based on current
center of mass position and velocity); MoS: Margin of stability, the
distance between the xCoM and the lateral boarder of the base of
support.
7
Figure II.2 A conceptual framework of how persistent altered motor behavior
identified by testing in and out of painful episodes may contribute to
future low back pain recurrence.
17
Figure III.1 Experimental setup of treadmill walking with prescribed step
widths. A visual feedback was projected on a wall in front of the
treadmill, with a red horizontal bar representing participant’s actual
step width and black vertical lines indicating the target width.
22
Figure III.2 (A) Conceptual graphs illustrating our definition of thorax, pelvis,
and trunk angles. Thorax angle and pelvis angles were defined as the
thorax or pelvis segment angles relative to the lab’s coordinate
system, while the trunk angle is the thorax relative to the pelvis
segment angle. Showing frontal plane for ease of illustration, but
these definitions also apply to the transverse and sagittal planes.
(B1) Demonstration of vector coding analysis on an angle-angle
diagram of a gait cycle in one representative participant. Coupling
angle was defined as the vector angle of two consecutive points in
time relative to the right horizontal. (B2) Cutoffs for binning of the
coupling angles into four coordination patterns. (B3) Illustrations of
the physical implication of the four coordination patterns. In-phase
indicate that both segments are rotating towards the same direction
at similar velocity; anti-phase indicate that the segments are rotating
to the opposite direction at similar velocity; thorax-only and pelvis-
only indicate that the thorax or pelvis segment is rotating
significantly faster than the other segment, while the other segment
may be hardly rotating.
25
Figure III.3 Step width task performance. (A) Square plot presenting the mean
and standard deviation of actual step width performance relative to
the prescribed step widths. The diagonal reference line indicates a
one-to-one fit of the performance with the targets. (B) Mean and
standard deviation of constant and variable step width error. Step
width affected constant error more than variable error.
28
x
Figure III.4 Group mean center of mass (CoM) in 5 prescribed step widths over
gait cycle in the (A) mediolateral direction (L: left, R: right) (B)
anteroposterior direction (Ant: anterior, Post: posterior), and (C)
vertical direction (Sup: superior, Inf: inferior).
29
Figure III.5 Group mean thorax, pelvis, and trunk angles in 5 prescribed step
widths over gait cycle in the transverse, frontal and sagittal planes.
Boxes highlight regions of significant one-way repeated measures
ANOVA revealed by SPM analyses that was more than 10% gait
cycle and consistently occurred during both left and right steps.
30
Figure III.6 Mean and standard deviation of angular excursion of the thorax,
pelvis, and trunk in the transverse, frontal, and sagittal planes for 5
prescribed step widths. Shaded background denotes step widths
wider than preferred.
32
Figure III.7 Mean and standard deviation of thorax-pelvis kinematic
coordination in the transverse, frontal, and sagittal planes for 5
prescribed step widths. Shaded background denotes step widths
wider than preferred.
34
Figure III.8 (A) Mean and standard deviation of the right and left longissimus
peak activation during the contralateral stance phase. (B) Mean and
standard deviation of bilateral co-contraction ratio of the
longissimus throughout the whole gait cycle.
35
Figure IV.1 Participant consort diagram. 46
Figure IV.2 Experimental setup of treadmill walking with prescribed step
widths. A visual feedback was projected on a wall in front of the
treadmill, with a red horizontal bar representing participant’s actual
step width and black vertical lines indicating the target width. The
thorax and the pelvis segments are illustrated.
50
Figure IV.3 (A) A pelvis-thorax angle-angle diagram of a gait cycle in one
representative participant. The “+” denotes right heel strike and the
arrow indicates progression of movement. Coupling angle was
defined as the vector angle of two consecutive points in time relative
to the right horizontal. (B) Cutoffs for binning of the coupling
angles into four coordination patterns. (C) Illustrations of the
physical implication of the four coordination patterns. In-phase
indicates that both segments are rotating to the same direction at
similar rate; anti-phase indicates that the segments are rotating to the
opposite direction at similar rate; thorax-only and pelvis-only
indicates that the thorax or pelvis segment is rotating significantly
faster than the other segment, or the other segment is hardly rotating.
51
xi
Figure IV.4 Self-reported body pain diagram composite indicating rLBP
participants’ primary pain location during the active pain testing
session.
55
Figure IV.5 Step width task performance. Both plots showing data for the back-
healthy controls (CTRL), individuals with recurrent low back pain
while in active pain (rLBP-A), and the same individuals during
symptom remission (rLBP-R). (A) Mean and standard error of mean
of preferred step width (B) Square plot presenting the mean and
standard deviations of actual step width performance relative to the
prescribed step widths. Note that the 1 x preferred step width was
also prescribed with visual feedback. The diagonal reference line
indicates a one-to-one fit of the performance with the targets.
56
Figure IV.6 The estimate and 95% confidence interval for the effect of pain (left
figure, rLBP-R relative to rLBP-A) and effect of group (right figure,
rLBP-A and rLBP-R relative to the CTRL) on frontal plane trunk
excursion.
57
Figure IV.7 The estimate and 95% confidence interval for the effect of pain (left
figure of each plot, rLBP-R relative to rLBP-A) and effect of group
(right figure of each plot, rLBP-A and rLBP-R relative to CTRL or
rLBP-pooled relative to CTRL) on frontal plane coordination
patterns: in-phase, anti-phase, pelvis-only, and thorax-only.
57
Figure IV.8 The estimate and 95% confidence interval for the effect of pain (left
figure of each plot, rLBP-R relative to rLBP-A) and effect of group
(right figure of each plot, rLBP-A and rLBP-R relative to CTRL or
rLBP-pooled relative to CTRL) on (A) peak longissimus activation,
and (B) bilateral longissimus co-activation.
58
Figure V.1 (A) Experimental setup of treadmill walking with prescribed step
widths. A visual feedback was projected on a wall in front of the
treadmill, with a red horizontal bar representing participant’s actual
step width and black vertical lines indicating the target width. (B)
Pelvis-thorax angle-angle diagram of a gait cycle in one
representative participant. The “+” denotes right heel strike and the
arrow indicates progression of movement. Coupling angle was
defined as the vector angle of two consecutive points in time relative
to the right horizontal. (C) Cutoffs for binning of the coupling
angles into four coordination patterns. (D) Illustrations of the four
thorax-pelvis coordination patterns. In-phase indicates that both
segments are rotating to the same direction at similar rate; anti-phase
indicates that the segments are rotating to the opposite direction at
similar rate; thorax-only and pelvis-only indicate that the thorax or
71
xii
pelvis segment is rotating significantly faster than the other segment,
or the other segment is hardly rotating.
Figure V.2 Single task arithmetic and step width performance for the back-
healthy controls (CTRL), individuals with recurrent low back pain
while in active pain (LBP-A), and the same individuals during
symptom remission (LBP-R). Error bars indicate standard error of
mean. n.s. = Non-significant.
75
Figure V.3 Dual-task performance and primacy switch effects on the control
group (CTRL) and the low back pain group in pain (LBP-A) and out
of pain (LBP-R). Each dot and each pair of connected dots indicate a
single participant. (A) Arithmetic dual-task effect during no
instruction condition. (B) Step width dual-task effect during no
instruction condition. (C) The change of arithmetic performance
from step width-prioritization to arithmetic prioritization conditions.
(D) The change of step width performance from step width-
prioritization to arithmetic prioritization conditions. n.s. = Non-
significant.
76
Figure V.4 (A) Arithmetic and (B) step width task performance variability
during the three dual task conditions shown as coefficient of
variation (CV). §: indicates a pain main effect between individuals
with recurrent low back pain when they were in (LBP-A) and out of
pain (LBP-R). *: indicates a group main effect between back-
healthy participants (CTRL) and LBP-R. · : indicates a trend
towards significance between CTRL and LBP-A.
77
Figure V.5 Thorax-pelvis coordination patterns during single task, dual task (no
instruction), step width prioritization (SW-pri) and arithmetic
prioritization (Ari-pri) conditions in the control group (CTRL) and
the low back pain group in pain (LBP-A) and out of pain (LBP-R).
Error bars indicate standard error of mean.
78
Figure VI.1 Conceptual figure summarizing key findings regarding altered trunk
control in recurrent low back pain.
91
Figure C.1 No differences in the test-retest of 6 control participants for trunk
excursion, and trunk in-phase and anti-phase coordination,
indicating no evidence of an effect of previous exposure to the
experimental task.
130
Figure D.1 No differences in the comparisons between test-retest of the control
group (CTRL) for arithmetic and step width (SW) single task
performance and dual-task performance variability, indicating no
evidence of practice effect.
131
xiii
Figure D.2 Correlation between days to retest and the changes in arithmetic and
step width (SW) single task performance and dual-task performance
variability between the in and out of pain testing sessions in
individuals with recurrent LBP. There were no relationships for any
of those variables, indicating no evidence of improved performance
associated with close testing dates.
132
xiv
ABSTRACT
Maintaining lateral stability during walking requires active control, which involves
controlling foot placement as well as center of mass movement. This needs to be achieved, in part,
through complex and precise coordination of multiple within-trunk segments, which may be
aberrant in individuals with low back pain (LBP). The impact of LBP around the globe on both
the societal and personal level is overwhelming due to its long-lasting and recurring nature.
Modifiable factors related to recurrence have yet to be established, and movement impairments
associated with a history of pain are currently a suspected candidate, since the majority of LBP is
non-specific and movement provoked. Altered trunk control strategies that persist beyond painful
episodes may be detrimental to spinal health and impaired adaptability to external task demands
could prevent the return to back-healthy control strategy in persons with recurrent LBP. However,
the persistence of altered trunk control beyond symptom duration warrants further investigation as
previous studies have not tested the same patients both in and out of an episode of pain.
Additionally, pain affects multiple systems and was known to interfere with attention, which may
in turn compromise goal-oriented motor behavior, as well as daily and working activity.
Furthermore, past findings of unchanged motor behavior during concurrent cognitive tasks may
indicate preferential prioritization of motor tasks in individuals with LBP, although these studies
were not designed to test attentional prioritization. Therefore, the current dissertation first
characterized trunk control at different step widths in back-healthy participants, then determined
if altered trunk control persists beyond symptoms by testing in and out of pain in individuals with
recurrent LBP and finally, determined if altered attention and preferential prioritization persist
beyond symptoms in individuals with recurrent LBP.
xv
The purpose of chapter III was to characterize changes in trunk control during different
prescribed step widths, and to compare the effect of wide and narrow step widths on trunk
kinematics and muscle activation. Twenty healthy young adults walked on a treadmill at 1.25 m/s
while matching step widths with real-time visual feedback to five prescribed widths: 0.33, 1.67, 1,
1.33, 1.67 × preferred step width. Motion capture was used to record kinematic data for calculating
step width, center of mass, and thorax and pelvis segmental motions. Thorax-pelvis coordination
was calculated using vector coding. Surface electromyography was used to capture longissimus
muscle activation, then peak longissimus activation and bilateral longissimus co-activation were
calculated after signal processing. Results showed that while center of mass only scaled with
different step widths in the mediolateral direction, trunk kinematics in all three planes varied with
step widths. Furthermore, only wider step widths affected transverse plane trunk kinematics,
leading to increased transverse plane trunk angular excursion, increased thorax-only coordination,
and decreased pelvis-only coordination. On the other hand, narrower step widths were associated
with increased longissimus activation, increased bilateral co-activation, and increased in-phase
coordination as well as decreased pelvis-only pattern in the frontal plane. These findings confirmed
our hypothesis that wide and narrow step widths introduce different demands on the system. The
changes in frontal plane trunk coordination and longissimus activation and co-activation reflected
increased active control of the trunk in response to decreased mediolateral base of support, making
these variables a sensitive measure of trunk control during lateral stabilization and suitable for the
subsequent investigation of recurrent LBP.
The purpose of chapter IV was to investigate the persistence of altered trunk control
associated with recurrent LBP. Twenty young adults with recurrent low back pain were tested once
during a painful episode and once in symptom remission, and twenty matched back-healthy
xvi
participants served as controls. Participants walked on a treadmill at 1.25 m/s with the same 5 step
widths as chapter III. Frontal plane trunk excursion, thorax-pelvis coordination, longissimus
muscle activation, and bilateral longissimus co-activation were first compared between pain status
within the recurrent LBP group using general linear models with pain status and step width as
fixed effects and subject as random effect. Then, if there were no effect of pain status or interaction
effects, data for the two testing sessions were pooled and compared to the control group, otherwise,
each pain status was compared to the control group separately. Results showed that there was no
interaction between pain status or group and step width, suggesting similar adaptability of trunk
control to various step widths. Individuals with recurrent LBP had less trunk excursion and anti-
phase coordination and increased in-phase coordination during a painful episode. This “tighter”
trunk control strategy when they have active pain was, however, not “tighter” when compared to
back-healthy persons. Furthermore, regardless of pain status, the recurrent low back pain group
had greater pelvis-only and less thorax-only coordination, and decreased bilateral longissimus co-
activation compared to the control group, suggesting a “loose” trunk control strategy. These
findings indicate an overall “loose” trunk control strategy in persons with recurrent LBP that was
slightly “tightened” by active pain, but not becoming “tighter” than a typical back-healthy strategy.
The purpose of chapter V was to investigate attention, task prioritization, and trunk control
during gait using a dual-task paradigm, in and out of a painful episode in a cohort of individuals
with recurrent LBP. Participants with recurrent LBP and matched controls walked on a treadmill
while matching a narrow step width and simultaneously performing serial subtractions of 7s, with
and without instructions to prioritize either task. Results suggest that active pain altered attention,
which negatively impacted single task performance and dual-task performance variability in
individuals with recurrent LBP. The ability to switch task prioritization between step width and
xvii
arithmetic tasks was not different between pain status or groups. Compared to the control group,
individuals with recurrent low back pain exhibited greater pelvis-only and less in-phase and
thorax-only patterns regardless of pain status, and across single and dual-task conditions.
The findings from these studies establish that there were increased active lateral
stabilization during walking with narrower widths. This was demonstrated by increased
longissimus activation and co-activation, as well as increased in-phase and decreased pelvis-only
trunk coordination in the frontal plane. Furthermore, persons with recurrent LBP exhibited an
overall “loose” trunk control strategy during walking regardless if they were in or out of pain, a
strategy that was consistent despite mechanical manipulation of step width or attentional
manipulation by dual-task. The presence of active pain somewhat “tightened” the trunk control
strategy in persons with recurrent LBP, but it also interfered with attentional processes and caused
decreased motor and cognitive task performance. This was the first time that altered trunk control
was demonstrated both in and out of pain in the same cohort of individuals with recurrent LBP.
Further work is warranted to test whether these alterations contribute to future symptoms and how
they can be effectively intervened in clinical implementation.
1
CHAPTER I
OVERVIEW
The trunk plays an important role in maintaining stability during human gait. Particularly,
stability during walking in the mediolateral direction requires more active control and relies on
modifications of foot placement and center of mass movement. Among the multiple body
segments, the trunk makes up a significant portion of body mass, thus contributes greatly to the
control of the center of mass. However, the multiple segments and degrees of freedom within the
trunk pose a challenge to the central nervous system such as a need for intricate coordination while
maintaining stability. In populations that have altered trunk control, such as individuals with low
back pain, the modulation of within-trunk coordination in response to changing step widths could
be further compromised.
Low back pain (LBP) is highly prevalent and leads to disability worldwide, largely due to
its long-lasting, recurring nature. Up to 80% of the population experience at least one episode of
LBP in a lifetime, and around 56% of the people that experience LBP exhibit recurrent symptoms
within a year. Despite the attempts to incorporate prevention of recurrence in the management of
LBP, the efficiency of these programs are inconclusive, and effective elements of the intervention
have not yet been identified. Moreover, the lack of established factors associated with future
recurrence is a critical barrier that hinders the development of effective and well-targeted
prevention programs. Since non-specific LBP is often movement-related, aberrant movement
associated with a history of pain is a highly suspected candidate as one of these factors.
Current literature proposes that the motor system adapts to back pain, utilizing suboptimal
control of movement that may be beneficial in the short-term, but may lead to long-term adverse
consequences. For example, one strategy individuals with LBP adopt is a “tight” trunk control
2
strategy, with increased trunk co-activation and thorax-pelvis kinematic coupling during walking,
which could lead to muscle fatigue, increased spinal loading, and induce intervertebral disc
inflammation. On the other hand, there were reports of “loose” trunk control strategies that persons
with LBP adopt, featuring early or excessive motion, specific to the lumbar spinal segments,
potentially creating excessive shear forces that can lead to pathology. These altered motor control
strategies can persist beyond symptom duration in persons with recurrent LBP. Furthermore, a
rigid motor strategy indicated by the lack of movement variability may prevent the return to pre-
LBP motor behavior. However, the notion that individuals with LBP are unable to change motor
strategy has yet to be systematically tested using external perturbations that attempt to induce
changes in motor behavior.
Adaptations to pain are not only presented as altered motor control, but also as altered
cognitive processing, an often-overlooked component of motor behavior. Individuals with LBP
may demonstrate greater prioritization to trunk-relevant motor tasks over concurrent cognitive
tasks, indicated by less change, or even improvements of motor performance when dual-tasking.
Using a well-designed dual task paradigm and explicit instructions, I will confirm whether there
is preferential prioritization of attentional resources, and furthermore, if it persists beyond
symptom duration. Persistence of altered trunk control or attention have a potential to not only
contribute to future symptoms, but also other adverse consequences such as falls or injury to other
body regions.
This dissertation, therefore, aims to investigate the persistence of altered motor behavior
associated with recurrent LBP by examining the modulation of trunk control and attentional
prioritization in response to different external task demands, in the ecological context of walking.
This was achieved through repeated testing on 20 young adults with a history of recurrent LBP
3
patients in and out of painful episodes, as well as 20 matched back-healthy controls with the
following aims.
Figure I.1. Framework of the specific aims for this dissertation.
Specific aim 1. To characterize trunk control during walking with various step widths in healthy
young adults (Chapter III).
Hypothesis 1. Wide and narrow step widths will have different effects on trunk control. Trunk
excursion and coordination will be more affected by wider widths due to larger center of mass
displacement, while longissimus muscle activation will be more affected by narrower widths due
to increased demand on active control.
Specific aim 2. To determine if trunk control in response to various step widths differs in and out
of pain in individuals with recurrent LBP, and when compared to back-healthy persons (Chapter
IV).
4
Hypothesis 2a. Individuals with recurrent LBP will have a “tighter” trunk control (decreased trunk
excursions, increased in-phase thorax-pelvis coordination, and increased longissimus activation
and co-activation) than back-healthy individuals, regardless of pain status.
Hypothesis 2b. Individuals with recurrent LBP will show less change in trunk control in response
to different step widths than back-healthy individuals, regardless of pain status.
Specific aim 3. To determine if attentional processing, task prioritization, and trunk control during
a dual-task differs in and out of pain in individuals with recurrent LBP, and when compared to
back-healthy persons (Chapter V).
Hypothesis 3a. Individuals with recurrent LBP will exhibit altered attentional processing
(decreased single task and dual-task performance, altered performance variability) compared to
back-healthy individuals, regardless of pain status.
Hypothesis 3b. Individuals with recurrent LBP will exhibit decreased ability to switch task
prioritization from motor to cognitive task compared to back-healthy individuals, regardless of
pain status.
Hypothesis 3c. Individuals with recurrent LBP will exhibit more in-phase trunk coordination
during dual task conditions compared to back-healthy individuals, regardless of pain status.
5
CHAPTER II
BACKGROUND AND SIGNIFICANCE
The role of the trunk in lateral stability during gait
Maintaining stability is one of the essential goals of human gait. Mediolateral stabilization
requires significantly more active control compared to the anterior-posterior direction, where
passive mechanisms can largely contribute to maintaining stability (Kuo & Donelan, 2010). This
was demonstrated by experiments showing that foot placement variability in the mediolateral
direction is more than the anteroposterior direction, and is more perturbed by sensory deprivation,
indicating higher dependence on sensory-based active feedback control (Bauby & Kuo, 2000;
Collins & Kuo, 2013). The cost of actively stabilizing in the mediolateral direction during gait is
demonstrated by studies using external stabilization to the pelvis (Dean et al., 2007; Donelan et
al., 2004). When walking with a narrow step width, applying external stabilization decreased step
width variability and metabolic cost (Dean et al., 2007; Donelan et al., 2004). The relevance of
mediolateral active control during gait is also supported by the fact that populations with known
balance-related deficits often exhibit wider preferred step widths (Arvin, Mazaheri, et al., 2016;
Hak et al., 2013; Hicks et al., 2017). Additionally, a narrower step width in older adults was
associated with higher lateral fall risk (Ko et al., 2007).
Lateral stability during walking is primarily controlled by two mechanisms – foot
placement and center of mass motion (Bruijn & van Dieën, 2018; MacKinnon & Winter, 1993).
In simple terms, the center of mass should not exceed the base of support determined by the foot
placement to avoid the risk of falling; in reality, however, the center of mass is dynamic and foot
placement needs to be determined before the center of mass reaches its maximum lateral position.
Based on the inverted pendulum model and the use of margin of stability, the foot is usually placed
6
lateral to the extrapolated center of mass, maintaining a minimum margin of stability and allowing
the ground reaction force to decelerate the center of mass (Hof et al., 2005) (Fig II.1). The
orientation of the foot such as its external rotation angle can also play a role in maintaining stability
by changing the base of support, responding to perturbations, and correcting for previous errors
(Rebula et al., 2017). Although walking with a wider step width offers more “reserve” from a
stability standpoint, the energy cost of step-to-step transition increases with the square of step
width (Donelan et al., 2001). Therefore, human preferred step width lies around 0.13 × leg length,
a tradeoff between energy costs from step-to-step transitions in wider widths and active control for
stability and cost of limb swinging laterally during narrow widths (Donelan et al., 2001, 2004;
Perry & Srinivasan, 2017). The center of mass movement can also be controlled to prevent it from
exceeding the base of support, especially when foot placement is already determined or
constrained. The control of the body segments that contribute to the center of mass primarily occur
through regulating joint moments at the subtalar, hip, and spinal joints in the frontal plane during
walking (MacKinnon & Winter, 1993). The hip abductors and adductors and the trunk musculature
are some of the primary muscles contributing to this lateral control of balance (Hof & Duysens,
2013; Kubinski et al., 2015; Neptune & McGowan, 2016; Prince et al., 1994).
7
Figure II.1. Conceptual model of the extrapolated center of mass and margin of stability for determining
mediolateral foot placement. v: velocity of the center of mass; xCoM: extrapolated center of mass (prediction of
center of mass position at subsequent heel strike based on current center of mass position and velocity); MoS:
Margin of stability, the distance between the xCoM and the lateral boarder of the base of support.
The trunk plays a critical role in the control of center of mass as it carries the head and the
arms, accounting for a total of 67% body mass (Winter, 2009). The trunk is multisegmented and
has numerous degrees of freedom, therefore complex coordination of trunk intersegmental
movement likely occurs while maintaining lateral stability in walking. While there were a few
studies that treated the trunk as a single segment or a representation of the whole body (McAndrew
Young & Dingwell, 2012; Perry & Srinivasan, 2017), there were, however, none that investigated
within trunk coordination in the context of lateral stability during walking. A past study reported
that when trunk movement is constrained with an orthosis, both center of mass excursion and step
width decreased (Arvin, van Dieën, et al., 2016). Therefore, the first aim of the current dissertation
is to examine how trunk inter-segmental control alters response to systematically changing
demands of different step widths. This is relevant because the findings provide a foundation for
investigating populations where trunk control is affected, such as individuals with low back pain.
8
Low back pain and its recurring symptoms as a worldwide health issue
Low back pain (LBP), often described as pain located between the lower ribs and inferior
gluteal fold, is highly prevalent worldwide (Deyo et al., 2014; Hartvigsen et al., 2018). It is the
leading cause of disability worldwide and accounts for 60.1 million years lived with disability
(GBD 2015 Disease and Injury Incidence and Prevalence Collaborators, 2016). LBP has a slightly
higher prevalence in women than men, and although the prevalence remains high from 10-89 years
old, it is the highest around 30-79 years old (Hoy et al., 2012). Moreover, the prevalence in North
America continues to rise (Friedly et al., 2010). About 50%-80% of the population will experience
one or more episodes of back pain in a lifetime, with up to 56% of the people that experience LBP
exhibiting recurrent symptoms within a year (da Silva et al., 2017; Pengel et al., 2003; White &
Gordon, 1982). Majority of the patients with LBP have fluctuating symptoms in their clinical
course that do not include complete recovery nor constant pain (Kongsted et al., 2015, 2016). The
high recurrence rate adds to the impact of LBP (da Silva et al., 2020). For individuals persistent
LBP symptoms impact quality of life, increase the risk of depression and risk of developing opioids
dependence (Bener et al., 2015; Deyo et al., 2015; Hirano et al., 2014). The effect of LBP on the
society is also well documented, with an estimated 87.6 billion US dollars of healthcare spending
in 2013 together with neck pain, ranked among the top three costly medical conditions (Dieleman
et al., 2016). LBP is also the second most common pain condition causing lost productive time
(5.28 hours per worker per week) due to its high prevalence within the working-age population
(Pengel et al., 2003; Stewart et al., 2003). With its detrimental effects on the societal level and the
personal level, the long-lasting symptoms of LBP are undoubtedly a critical issue around the globe
that need to be addressed effectively.
9
Development of prevention programs for LBP is difficult due to a lack of established
biomarkers associated with recurrence. LBP is a symptom, not a disease, since it can result from
various diseases or pathology that could be undetected (Hartvigsen et al., 2018). Only a small
proportion of low back pain can be attributed to a specific source of pathology, and the rest, making
up 85%-90%, is termed non-specific LBP (Deyo et al., 1992; Maher et al., 2017; Van Tulder et
al., 2002; White & Gordon, 1982). Current clinical standards are unable to accurately identify the
structural source of these non-specific low back pain (Deyo et al., 2014; Hartvigsen et al., 2018).
Current clinical guidelines suggest a multidisciplinary approach to non-specific LBP that includes
non-pharmacological treatment such as exercise and cognitive behavioral therapy, and avoids
pharmacological and surgical interventions (Foster et al., 2018). However, the cost-effectiveness
of physical therapy is still controversial (Friedly et al., 2010) and specific, effective intervention
programs that target the prevention of future back pain episodes are warranted especially when a
person has begun exhibiting episodes of pain. Our lack of understanding of contributing factors to
recurrence is a major obstacle in designing these intervention programs. Currently, no known
factors other than a history of previous episodes were linked to an increased risk of future
recurrence (da Silva et al., 2017). Although attempts have been made to incorporate prevention of
recurrence in clinical management of LBP, the effectiveness of these programs were overall
promising but inconsistent between studies, and the specific type of exercise intervention with the
highest efficacy was not identified (Choi et al., 2010). Recent protocol of a randomized control
trial also promised to examine the effects of McKenzie method self-treatment on secondary
prevention of LBP, but the results were not yet reported (de Campos et al., 2017).
10
Could alterations in trunk control lead to future symptoms?
Pain affects how we move. Existing reviews suggest that pain or threat of pain causes
changes at multiple levels of the nervous system, causing redistribution of muscle activities and
altered movement patterns, all with a common goal to reduce pain, protect the region, and
potentially to feel more in control (Hodges et al., 2013; Hodges & Tucker, 2011; Van Dieën et al.,
2017).
The majority of the literature points to a “tight” trunk control in individuals with LBP. This
was manifested in both changes in muscle activity and kinematics compared to back-healthy
individual. Studies reporting muscle activation showed that individuals with clinical or
experimental LBP demonstrated increased co-activation and preferential activation of the
superficial muscles during walking (Colloca & Keller, 2001; Crosbie et al., 2013; Gombatto et al.,
2015; Hodges et al., 2009, 2013; Seay et al., 2011a; Van Dieën et al., 2017). Increased muscle
activity was also found in individuals with LBP during anticipation of a perturbation (Van Dieën
et al., 2003). Kinematic findings during walking demonstrated less range of motion and a more
“en bloc” behavior of the thorax and pelvis was observed in persons with LBP (Crosbie et al.,
2013; Gombatto et al., 2015; Seay et al., 2011a).
On the other hand, recent reviews and limited experimental data suggest that a “loose”
trunk control (delayed muscle activation, decreased trunk coupling, increased segmental
movement) may also be presented in individuals with LBP with certain subgroups or context of
tasks (Van Dieën, Reeves, Kawchuk, van Dillen, et al., 2019; Van Dieën, Reeves, Kawchuk, Van
Dillen, et al., 2019). Deep trunk muscles such as transverse abdominis and multifidus often
exhibited delayed onset and offset during postural perturbation (Prins et al., 2018). Studies
utilizing self-initiated tasks such as forward bending and picking up an object reported increased
11
lumbar excursion during the early phase of movement in patients with chronic LBP (Marich et al.,
2017, 2020). Divergence of the findings in LBP motor behavior may depend on diverse context of
the task and subgroups of patients, as well as pain status at the time of testing.
Pain-related movement adaptations, such as altered trunk control, may persist beyond
symptom duration. Participants with a history of LBP who were asymptomatic at the time of
testing still exhibit increased trunk stiffness (Hodges et al., 2009), delayed anticipatory postural
response (MacDonald et al., 2009), and more in-phase trunk coordination during running (Seay et
al., 2011a, 2011b) and walking (Seay et al., 2011a), although results on trunk coordination during
walking are inconsistent between studies (Seay et al., 2011a, 2011b; Smith & Kulig, 2016a). The
persistence of these movement alterations may be due to reinforcement learning in an attempt to
avoid further pain (Van Dieën et al., 2017).
Decreased movement variability and adaptability to external task demands may be
preventing the return to back-healthy control strategies. Pain is a potent stimulus and the body sees
it as a great cost, therefore we may be learning to adapt to pain in less time, fewer repetitions, but
fail to further refine our strategies (Van Dieën et al., 2017). Individuals with LBP exhibit low
movement variability once they have chosen a specific strategy to move with pain (Hedayati et al.,
2014; Van Dieën et al., 2017). An experiment done on healthy participants found that lower
movement variability during the adaptation to experimental pain was predictive of failure to return
to previous strategy when the pain subsides (Moseley & Hodges, 2006). Therefore, individuals
with LBP may adopt a rigid motor strategy which prevents the return to pre-LBP motor control. It
has also been shown that persons during symptom remission of recurrent LBP did not increase
multifidus activation in response to the demand of walking faster, unlike their back-healthy
counterparts, illustrating a decreased adaptability to changing external demands (Smith & Kulig,
12
2016b). Collectively, decreased movement variability and adaptability indicate a sign of rigid
motor strategy that individuals with LBP may be using. If they are “stuck” in their selected
strategy, this could explain the difficulty of existing interventions to improve long-term outcomes
in LBP.
Persistent alterations in trunk control could lead to long-term adverse consequences. In the
case of increased trunk stiffness, the adaptations to pain may be beneficial in the short-term as
lowering the displacement of anatomical structures to a given force can reduce shear stress to the
injured structures. On the other hand, a “loose” control strategy could reduce compression forces
on the spine, but at the expense of increased shear stress. If these altered trunk control strategies
persist, it could impact spinal health and LBP symptoms negatively in several ways. First, co-
contraction and increased activities in the trunk muscles could lead to increased compression and
shear loading on the spine (Marras, Davis, et al., 2001; Marras, Ferguson, et al., 2001), which,
when sustained, could lead to intervertebral disc inflammation and long-lasting nerve injuries
(Miyagi et al., 2012). Second, sustained trunk muscle activation at low variability and level as low
as 2-5% maximum voluntary contraction could result in muscle pain and fatigue, thus lead to
further impairments in postural control due to compromised proprioception, and reduced reflex
adaptations (Abboud et al., 2016; Boucher et al., 2012; Parreira et al., 2013). Third, low variability
of movement could compromise load sharing between passive structures and lead to repetitive
stress on the same tissue (Srinivasan & Mathiassen, 2012). Forth, reduced variability may cause
impoverished feedback and contribute to the changes in the trunk representation area in the cortex
and in turn lead to impaired sensorimotor integration, starting a vicious cycle (H. Tsao et al., 2008;
Henry Tsao et al., 2011). Last, although the intrinsic stiffness of the trunk could be increased,
overall stability could still be compromised because the benefit of flexibly controlling each spinal
13
segment is lost (Mok et al., 2007; Reeves et al., 2011). This could lead to impairments in balance
and thus increase risk falls, injury of the spine, as well as other body regions (Marshall et al., 2016;
Zazulak et al., 2007a, 2007b).
Despite abundant evidence that supports potential mechanisms of how altered motor
control can increase the risk of back pain recurrence, there is currently no longitudinal studies that
prove this. The present dissertation studies can identify promising variables of interest that will
serve as a basis for future studies to determine the association of selected movement adaptations
and recurrence rate.
Could alterations in attention become detrimental to daily tasks?
Pain is such a potent stimulus that it influences not only the traditional sensorimotor
functions, but also attention. Attention can be defined as the behavioral and cognitive process of
selectively focusing on a piece of information in the environment (Fougnie, 2008). Attention and
movement are not distinctly independent. Goal-directed movement requires attention to selectively
process relevant information in the preparatory and execution phases of the action. In fact, motor
memory is stored in the motor cortex as well as the attentional network (dorsolateral prefrontal
cortex and the inferior parietal lobule), where goal-relevant priorities are maintained and task-
relevant inputs are selected (Robertson, 2009). Therefore, when attentional focus is altered or when
attentional resource is occupied, movement planning and execution can be affected. Moreover, a
bigger problem is presented when altered attention impacts executive function that is a
fundamental cognitive process underlying almost every daily and work activity.
Attention can be captured and occupied by LBP, leading to decreased cognitive and dual-
task performance. Attention selection involves both bottom-up and top-down mechanisms
14
(Legrain et al., 2009; Van Damme et al., 2010). Through the bottom-up route, attention can be
captured involuntarily by LBP because of its salient nature. This capture of attention by pain is a
critical survival feature because of its alarming function. However, top-down selection can
regulate how strongly attention is directed to pain. For example, when performing a demanding
and prioritized task, a “goal-shielding” mechanism prevents irrelevant information like pain to
capture attention (Van Damme et al., 2010). On the other hand, an individual with LBP with pain-
catastrophizing beliefs may constantly scan for painful stimuli (Legrain et al., 2009; Seminowicz
& Davis, 2006). Evidence of compromised attention in individuals with LBP is present from brain
to behavioral studies. Patients with chronic LBP showed aberrant structure and functional
connectivity in the attentional networks (Čeko et al., 2015; Mao et al., 2014; Seminowicz et al.,
2011). Persons with LBP also performed worse than their back-healthy counterparts in cognitive
task performance, such as in a Stroop test (Etemadi et al., 2016; Lamoth et al., 2008). During dual-
task conditions, individuals with LBP often exhibit different motor responses compared to the
control group, indicating alterations in attentional capacity, allocation, or prioritization.
When pain is associated with movement, such as the case with LBP, individuals may
prioritize their focus to a motor task over concurrent cognitive tasks, either consciously or
unconsciously. Individuals with LBP demonstrated greater gait variability but less trunk
coordination variability during dual-task walking (Hamacher et al., 2014, 2016; Lamoth et al.,
2008), and faster balance recovery velocity when given a concurrent cognitive task when
compared to controls (Etemadi et al., 2016). This preferential prioritization, like other movement
adaptations to pain, seems to persist beyond symptom duration. Smith et al. found that while back-
healthy adults decreased step length consistency during dual-task walking turns, asymptomatic
participants with a history of recurrent LBP did not show this change (Smith et al., 2017). These
15
research findings may be due to greater prioritization of the motor tasks in persons with LBP,
although these experiments were not specifically designed to test for attentional prioritization.
A more robust way to investigate prioritization is to include different instructions on
prioritization to probe participants’ preference and ability to flexibly switch prioritization on
demand. Some of the dual-task studies in LBP failed to report cognitive performance, or did not
take both speed and accuracy into account when calculating performance, making it difficult to
account for trade-offs in task performances (Hamacher et al., 2016; Mazaheri et al., 2010; Smith
et al., 2017). Additionally, most of the previous research did not take baseline differences of single
task performance into account, while persons with LBP often perform worse on the baseline
cognitive and motor tasks (Etemadi et al., 2016; Lamoth et al., 2008). If individuals with LBP
indeed have persistent attentional prioritization towards motor tasks, it would greatly impact daily
and working activities. Therefore, the current dissertation aims to specifically test whether
individuals with recurrent LBP prioritize a trunk-relevant motor task over a concurrent cognitive
task, and whether this preferential prioritization persists, by implementing primacy switch trials
with instructions on prioritization, and accounting for performance trade-offs and baseline
differences.
Significance of this multidisciplinary study in the prevention of low back pain recurrence
This is the first study to test the same participants in and out of an episode of back pain
with a longitudinal approach, which allows for a robust evaluation of the persistence of alterations
in motor behavior. Previous studies have either compared motor control in participants with
recurrent LBP during their asymptomatic periods to back-healthy controls, or used distinct groups
for individuals with active LBP versus individuals with a history of LBP but recovered (Hooper et
16
al., 2016; Rowley et al., 2019; Seay et al., 2011a, 2011b; Smith & Kulig, 2016b). LBP is
undoubtedly a multi-faceted problem that involves countless factors ranging from biomechanical
to physiological to psychological and beyond (Cholewicki et al., 2019), therefore a
multidisciplinary approach is highly warranted, yet still not popular in the current literature. In this
dissertation the concurrent testing of the motor and cognitive systems was done using mechanical
and attentional manipulations to provide a more complete picture of altered motor behavior in
recurrent LBP. Mediolateral stability demands during walking were systematically manipulated
with different step widths using an innovative step width real-time feedback. Previous studies have
used methods that often requires participants to look down at the target line on the treadmill (Arvin,
Mazaheri, et al., 2016; Donelan et al., 2001), however this would have created movement that
affects the trunk. The novel customized program used in this study projects a real-time feedback
that was calculated from 3D marker-based data in front of the participant at eye level so that they
could match their step width without changing trunk posture. A refined dual-task paradigm with
task prioritization instructions was also used in this dissertation, which will advance our
understanding of attentional prioritization in persons with recurrent LBP.
This work is significant because it provides a basis for future longitudinal studies that can
determine the relationship between persistent alterations and the risk of future recurrence. If these
persistent alterations to recurrent LBP prove to be linked to future episodes, then they would be
considered maladaptive. The long-term goal of this line of research is to develop interventions
aiming to prevent LBP recurrence by targeting these maladaptations. Innovative and precisive
rehabilitative and preventative strategies are warranted to reduce the personal and social impacts
caused by LBP.
17
Figure II.2. A conceptual framework of how persistent altered motor behavior identified by testing in and out of
painful episodes may contribute to future low back pain recurrence.
18
CHAPTER III
TRUNK CONTROL AT VARIOUS STEP WIDTHS DURING WALKING
Abstract
Healthy young adults were tested to characterize trunk control in response to different step
widths and to compare the effects of wide and narrow step widths. Participants walked on a
treadmill at 1.25 m/s while matching step widths with real-time visual feedback to five prescribed
widths: 0.33, 1.67, 1, 1.33, 1.67 times preferred step width. Motion capture was used to analyze
step width, center of mass, and within-trunk segment motions, while surface electromyography
was used to record longissimus activation. Results showed that while center of mass only varied
across step width in the mediolateral direction, trunk kinematics in all three planes were affected.
Angular excursions of the thorax, pelvis, and trunk were affected more by wider widths and this
was most apparent in the transverse plane, demonstrating higher excursions with wider widths.
Thorax-pelvis kinematic coordination was affected more by wider widths in transverse plane and
by narrower widths in the frontal plane. Peak longissimus activation and bilateral co-contraction
increased as step widths became narrower. Peak longissimus activation was more affected by
narrower step widths. Wide and narrow step widths present with different demands on the motor
system and the trunk within-segment behavior was controlled in complex ways to match these
demands. This study provides foundation for future investigations on the trunk during gait in
different populations.
19
Introduction
Maintaining stability is one of the primary goals in human locomotion (MacKinnon &
Winter, 1993). In contrast with sagittal plane stability that could be largely achieved through
passive mechanisms, frontal plane stability during gait requires active control (Bauby & Kuo,
2000; Kuo & Donelan, 2010). Frontal plane stability during walking could be achieved through
manipulating mediolateral foot placement (Bruijn & van Dieën, 2018) and external rotation of the
foot (Rebula et al., 2017), or adjusting the body’s center of mass movements (Arvin, Mazaheri, et
al., 2016). Most likely, a combination of these strategies is used. In real-life situations, however,
we encounter scenarios where foot placement is restricted, for instance, narrow hiking trails or
obstacles on the ground. In these scenarios the body’s center of mass must be controlled so we do
not fall.
The head, arm, and trunk adds up to 67% of the body weight and is the largest contribution
to the center of mass (MacKinnon & Winter, 1993; Prince et al., 1994). Therefore, how the trunk
is controlled and coordinated within its multiple segments will directly influence upper body
movement and center of mass movement. Although the relationship of center of mass and lateral
foot placement has been studied widely (Arvin, Mazaheri, et al., 2016; Bruijn & van Dieën, 2018;
Hurt et al., 2010; McAndrew Young & Dingwell, 2012; Perry & Srinivasan, 2017; Stimpson et
al., 2018; Wang & Srinivasan, 2014), studies investigating the relationship of trunk control and
step width are rare. When trunk movement is constrained with an orthosis, both center of mass and
step width decreased (Arvin, van Dieën, et al., 2016). There are no studies, however, describing
how trunk control and segmental coordination is modified with prescribed foot placement.
Furthermore, most studies that manipulate step width use strings or tape on the ground or the
treadmill, requiring participants to look down, which will affect their trunk motion (Arvin,
20
Mazaheri, et al., 2016; Perry & Srinivasan, 2017). A study that is designed to investigate trunk
control across different step widths is therefore desired.
During gait, the mass of the trunk is controlled in the frontal plane mainly by the spinal
musculature and the hip abductors (MacKinnon & Winter, 1993). The contribution of hip
abductors to modifying foot placement (Rankin et al., 2014) and stability during stance phase
(Kubinski et al., 2015) were documented. However, the relationship of paraspinal muscle
activation with different step widths was not well studied. Literature on typical gait points to
muscles at the lumbar region as the highest activated among the C7 to L4 paraspinal muscles
(Prince et al., 1994), yet how the paraspinal muscles activate differently in different step width is
still unknown.
Walking with wide and narrow step widths each place unique demands on the motor
system. The preferred step width is selected largely to minimize energy cost without compromising
stability. Moving away from the preferred step width may affect the inverted pendulum-like
motion of the center of mass and influence energy demands (Kuo et al., 2005). A study by Donelan
and Kuo demonstrated an increased cost of step-to-step transition as step width becomes wider,
and on the other hand, an increased cost of lateral limb swing as step width becomes narrower than
preferred (Donelan et al., 2001). In other words, walking with wider step widths requires greater
mechanical work to redirect the center of mass, while walking with narrower step widths requires
the swing leg to swing laterally to avoid the stance leg. Walking with narrower widths presents
greater challenges to stability as the base of support decreases, increasing the demand on active
postural control (MacKinnon & Winter, 1993; Perry & Srinivasan, 2017). When the pelvis was
stabilized externally during narrow step width, metabolic cost was decreased (Donelan et al.,
21
2004). How these unique demands during walking with wide and narrow step width affect trunk
control differently warrants investigation.
The purpose of this study was to characterize changes in trunk control during different
prescribed step widths, and to compare the effect of wide and narrow step widths (relative to the
preferred) on trunk kinematics and muscle activation. We hypothesized that muscle activation will
be more affected by narrower widths due to increased active control, while trunk kinematics will
be greater affected by wider widths due to larger center of mass displacement.
Methods
Participants
Twenty healthy young adults participated in this study (14 females, 6 males; 26.25 ± 3.31
years; 165.54 ± 9.93 cm; 61.39 ± 12.71 kg; BMI = 22.21 ± 2.84 kg/m2). Participants were included
if they are between 18 to 45 years old, and if they have no currently known pathology or history
of lower extremity or spine surgery. Participants were recruited through flyers, public
announcements, and word of mouth in the University of Southern California and its surrounding
community. Participants gave written informed consent that was approved by the institutional
review board of the University of Southern California.
Instrumentation
Participants were instrumented with a lower extremity marker set including markers on the
1st and 5th metatarsal heads, distal foot (2nd toe), medial and lateral malleoli, medial and lateral
femoral epicondyles, greater trochanters, anterior superior iliac spines, iliac crests, L5-S1, and
additional markers on bilateral acromion, sternal notch, and T1. Clusters of 3-4 markers on a rigid
22
plate were placed on the thigh, shank, and heel counter of the shoes to serve as tracking markers.
Kinematic data were recorded by a 11-camera Qualisys motion capture system (Qualisys Inc.,
Gothenburg, Sweden) at 125 Hz. Surface electrodes were placed on bilateral longissimus (2 finger
width para spinous process of L3). Electromyography (EMG) data were collected using Noraxon
wireless EMG system (Noraxon U.S.A, Inc., Scottsdale, AZ, USA) at 1500 Hz. A portable
treadmill (PRO-FORM 505 CST, ICON Health & Fitness, Inc., Logan, UT, USA) was used for
the walking trials.
Experimental procedures
Participants completed a medical history form and the International Physical Activity
Questionnaire (Craig et al., 2003). Leg length from the greater trochanter to the ground of the right
leg was measured. A standing calibration trial for the kinematic data was collected. Participants
were given up to 3 minutes to familiarize with the treadmill, after which a 30 second treadmill
walking trial at 1.25 m/s was collected to determine preferred step width (PSW).
Figure III.1. Experimental setup of treadmill walking with prescribed step widths. A visual feedback was projected
on a wall in front of the treadmill, with a red horizontal bar representing participant’s actual step width and black
vertical lines indicating the target width.
23
Participants then walked with 5 different step widths with the help of real-time visual
feedback projected on the wall in front of the treadmill (Fig III.1). Step width was calculated using
marker data that was streamed real-time into MATLAB (MathWorks, Inc., Natick, MA, USA).
The average heel and 2nd toe marker position at foot flat was calculated to indicate foot position,
then the medial-lateral distance between subsequent foot falls were calculated as step width. This
accounted for external rotation angle in addition to foot placement, which has an effect on base of
support and can be helpful in maintaining stability during walking (Rebula et al., 2017).
Participants were instructed to match the width of the red bar indicating their current step width to
the black dotted line indicating the goal. Five different step widths were prescribed as a ratio of
their PSW, including 0.33, 0.67, 1, 1.33, and 1.67 × PSW. One 30-second familiarization trial at
1PSW was given. Four 30-second practice trials, one trial for each prescribed widths other than
1PSW, were then carried out. More practice trials were provided if the participant was not
comfortable or was not able to perform at an adequate level. Participants then completed one 30-
second trial for each step width in a randomized order. Six participants were tested again at least
one week apart to determine test-retest reliability.
Data Analyses
Kinematic data were low-pass filtered at 10 Hz with a dual-pass 4th order Butterworth
filter. Gait events were identified based on the toe and heel marker position relative to the pelvis
coordinate. Each gait cycle was time-normalized to 101 data points from right heel strike to the
subsequent right heel strike. Each 30-second trial consists of approximately 25-35 strides, but only
the last 20 strides were included in the analyses to allow participants to reach steady state walking
with the prescribed step widths.
24
Foot position of each step was determined as the mid position between the heel and 2nd
toe marker at foot flat. Step width was then calculated as the mediolateral distance between foot
positions of two subsequent foot falls. Constant and variable step width errors were determined
independently. Constant error was calculated as the difference between the mean performed step
width and the prescribed step width. Variable error was calculated as the standard deviation of the
performed step width.
The thorax and pelvis angles were defined as the thorax/pelvis segment relative to the lab
coordinate system, whereas the trunk angle was defined as the thorax relative to the pelvis (Fig
III.2A). Angular excursions were calculated as the peak to peak angles of the thorax, pelvis, and
trunk during each gait cycle. Thorax-pelvis kinematic coordination was calculated using vector
coding analysis described by Needham et al (Needham et al., 2014). An angle-angle diagram was
first constructed with the pelvis on the horizontal axis and the thorax on the vertical axis, then the
coupling angle was determined as the angle of the vector between two adjacent data points in time
relative to the right horizontal (Fig III.2B). Coordination patterns were categorized as in-phase,
anti-phase, thorax-only, and pelvic-only defined by coupling angles falling within each ranges of
angles indicated in Fig III.2C (Fig III.2C&D).
25
Figure III.2. (A) Conceptual graphs illustrating our definition of thorax, pelvis, and trunk angles. Thorax angle and
pelvis angles were defined as the thorax or pelvis segment angles relative to the lab’s coordinate system, while the
trunk angle is the thorax relative to the pelvis segment angle. Showing frontal plane for ease of illustration, but these
definitions also apply to the transverse and sagittal planes. (B1) Demonstration of vector coding analysis on an
angle-angle diagram of a gait cycle in one representative participant. Coupling angle was defined as the vector angle
of two consecutive points in time relative to the right horizontal. (B2) Cutoffs for binning of the coupling angles into
four coordination patterns. (B3) Illustrations of the physical implication of the four coordination patterns. In-phase
indicate that both segments are rotating towards the same direction at similar velocity; anti-phase indicate that the
segments are rotating to the opposite direction at similar velocity; thorax-only and pelvis-only indicate that the
thorax or pelvis segment is rotating significantly faster than the other segment, while the other segment may be
hardly rotating.
26
To avoid heart rate contamination, EMG data were bandpass filtered between 30 Hz and
500 Hz with a dual-pass 4th order Butterworth filter (Drake & Callaghan, 2006). The signal was
then rectified and smoothed using a 100 ms moving window and normalized to the averaged peaks
of the initial treadmill walking trial without prescribed step width feedback (peak gait activation).
Peak lumbar longissimus activation was determined within the contralateral stance phase, where
higher activation was found comparing the bilateral stance phase. Bilateral co-contraction was
determined as the average absolute ratio between right and left longissimus activation.
Statistical Analyses
Data were plotted for visualization and examined for normality, homoscedasticity, and
outliers. Descriptive statistics was performed on all variables. Statistic parametric mapping (SPM)
was used to compare the time series data of trunk kinematics and detect phases in gait that are
significantly different across step widths. In this study, a vector-field equivalent of one-way
repeated measures analysis of variance (ANOVA), the SPM F-statistics was used. For more
detailed information on SPM, see Pataky (Pataky, 2010, 2012). Constant and variable step width
errors, thorax, pelvis, and trunk angular excursions, thorax-pelvis kinematic coordination patterns,
lumbar longissimus EMG activity and bilateral co-contraction were compared between step widths
using one-way repeated measures ANOVA. In the case of a significant ANOVA test, Tukey’s
post-hoc tests were performed for multiple pairwise comparisons. Test-retest intraclass correlation
coefficient for one-way random effects, absolute agreement, and multiple measurements (ICC
(1,k)) were calculated for all variables (McGraw & Wong, 1996), and interpreted based on the
cutoff criteria from Koo and Li (Koo & Li, 2016). Additionally, SEM was calculated as
𝑆𝐷 √1 − 𝐼𝐶𝐶 . The α level was set at 0.05. Statistical analyses were done in R (R Core Team, 2018).
27
Results
Task Performance
The average PSW was 14.77 ± 2.92 cm, equivalent to 0.17 x leg length. Participants were
able to vary their step widths based on the visual feedback (Fig III.3A). Both constant and variable
step width error increased in the narrowest step width (F(4,76) = 18.93, p<0.001; F(4,76) = 4.65,
p=0.002, respectively), which was significantly different from the rest of the step widths (pairwise
p<0.014), with the exception of the widest step width for variable error (p=0.073) (Fig III.3B).
Constant error at 0.67PSW was also higher than at 1.67PSW (p<0.001).
Figure III.3. Step width task performance. (A) Square plot presenting the mean and standard deviation of actual
step width performance relative to the prescribed step widths. The diagonal reference line indicates a one-to-one fit
28
of the performance with the targets. (B) Mean and standard deviation of constant and variable step width error. Step
width affected constant error more than variable error.
Center of Mass (CoM) and Thorax, Pelvis, and Trunk Kinematics
The amplitude of mediolateral CoM scaled with step width, while the anteroposterior and
vertical CoM were hardly affected (Fig III.4). On the other hand, walking with different step widths
resulted in different thorax, pelvis, and trunk angle curves in distinct phases of gait in all three
planes (Fig III.5). The curves that reached a significant p-value with one-way repeated measures
ANOVA using SPM were highlighted with boxes. In the transverse plane, differences between
step widths happened primarily around the peak angles; In the frontal plane, differences in the
pelvis angles included the peaks while differences in the thorax and trunk angles occurred near
mid-range angles; In the sagittal plane, the results were more scattered and differences occurred
across multiple timepoints in the gait cycle.
29
Figure III.4. Group mean center of mass (CoM) in 5 prescribed step widths over gait cycle in the (A) mediolateral
direction (L: left, R: right) (B) anteroposterior direction (Ant: anterior, Post: posterior), and (C) vertical direction
(Sup: superior, Inf: inferior).
30
Figure III.5. Group mean thorax, pelvis, and trunk angles in 5 prescribed step widths over gait cycle in the
transverse, frontal and sagittal planes. Boxes highlight regions of significant one-way repeated measures ANOVA
revealed by SPM analyses that was more than 10% gait cycle and consistently occurred during both left and right
steps.
Thorax, Pelvis, and Trunk Angular Excursions
Peak to peak angular excursions of the thorax, pelvis, and trunk across different step widths
are presented in Figure III.6. In the transverse plane, thorax excursions increased significantly from
1PSW to 1.33PSW (p<0.001), and from 1.33PSW to 1.67 PSW (p=0.001), while no difference
was found in the two narrowest widths. Transverse pelvis excursion at 1.67PSW was significantly
greater than 0.33PSW - 1PSW (p=0.007-0.018). Transverse trunk excursion at 1.33PSW and
1.67PSW were both significantly greater than the rest of the step widths (all p<0.017), although
31
these two wider step widths did not differ from each other. In the frontal plane, there were no main
effects of step width for the thorax (F(4,76)=2.45, p=0.053), pelvis (F(4,76)=2.95, p=0.095), and
trunk (F(4,76)=2.26, p=0.070) excursions. In the sagittal plane, there were no significant
differences between step widths for the thorax excursions (F(4,76)=2.31, p=0.065). Sagittal pelvis
excursion at 1.33PSW and 1.67PSW were both significantly greater than the rest of the step widths
(p<0.035). Sagittal trunk excursion at 1.67 PSW was significantly greater than 0.33PSW and
0.67PSW (p=0.001-0.003), and 1.33PSW was also significantly greater than 0.33PSW (p=0.036).
See Appendix A for more detailed statistical results.
32
Figure III.6. Mean and standard deviation of angular excursion of the thorax, pelvis, and trunk in the transverse,
frontal, and sagittal planes for 5 prescribed step widths. Shaded background denotes step widths wider than
preferred.
Thorax-pelvis kinematic coordination
Thorax-pelvis kinematic coordination across different step widths is presented in Figure
III.7. The influence of step widths on coordination patterns was more pronounced at wider step
widths in the transverse plane. The percentage of transverse plane in-phase pattern was not
33
different across step widths. The percentage of transverse anti-phase was greater at 1.33PSW than
at 0.33PSW (p=0.041). Transverse thorax-only pattern had greater percentage at 1.67PSW than
the three narrowest step widths (p=0.007-0.023), whereas pelvis-only pattern had smaller
percentage at 1.67PSW than the three narrowest step widths (p=0.002-0.005).
The influence of step widths on coordination patterns was more apparent at narrower step
widths. Frontal plane in-phase pattern had a significantly greater percentage at 0.33PSW than
0.67PSW, 1PSW, and 1.33PSW (p=0.002-0.009). There was no difference across step widths for
frontal anti-phase pattern. Frontal plane thorax-only pattern had a greater percentage at 0.33PSW
than 0.67PSW and 1PSW (p=0.015-0.036), while pelvis-only pattern had a smaller percentage at
0.33PSW than the rest of the step widths (p<0.044).
The relationship of sagittal plane coordination with step widths were more linear. The
percentage of sagittal plane in-phase increased as step width became wider, indicated by
significant pairwise comparisons between most step widths (p<0.023) except for some of the
adjacent ones. Sagittal plane anti-phase showed similar results as in-phase but opposite in direction
(p<0.007). The sagittal plane thorax-only pattern had lower percentage at the two widest step
widths than the two narrowest step widths (p<0.014). The sagittal plane pelvis-only pattern had a
significant one-way repeated measures ANOVA test (F(4,76)=2.59, p=0.043), however, all
pairwise comparisons failed to reach significance. See Appendix A for full statistical results.
34
Figure III.7. Mean and standard deviation of thorax-pelvis kinematic coordination in the transverse, frontal, and
sagittal planes for 5 prescribed step widths. Shaded background denotes step widths wider than preferred.
35
Longissimus Activation and Co-contraction
The longissimus activation and co-contraction results are presented in Figure III.8. The
right longissimus peak activation during 0.33PSW was significantly greater than the rest of the
step widths (p<0.025). The left longissimus showed similar results, but activation during 0.33PSW
was only greater than 1PSW, 1.33PSW, and 1.67PSW (p<0.001), and activation at 0.67PSW was
greater than 1.67PSW (p=0.003). Bilateral longissimus co-contraction at 1.67PSW was less than
the three narrowest widths (p<0.007), and co-contraction at 1.33PSW was also less than 0.33PSW
(p=0.005). See Appendix A for full statistical results.
Figure III.8. (A) Mean and standard deviation of the right and left longissimus peak activation during the
contralateral stance phase. (B) Mean and standard deviation of bilateral co-contraction ratio of the longissimus
throughout the whole gait cycle.
36
Test-retest reliability
The test-retest reliability for the measures are presented in Table III.1.
Table III.1. Test-retest reliability and SEM for primary variables (ICC: intra-class correlation coefficient, SD:
standard deviation, SEM: standard error of the measurement).
ICC SD SEM
Preferred Step Width (cm) 0.85 2.56 1.00
Constant Step Width Error (cm) 0.81 1.19 0.52
Variable Step Width Error (cm) 0.85 0.54 0.21
Thorax Angular Excursion (degree) 0.97 0.85 0.15
Pelvis Angular Excursion (degree) 0.93 2.21 0.57
Trunk Angular Excursion (degree) 0.94 2.52 0.61
In-phase Coordination (%) 0.93 8.05 2.13
Anti-phase Coordination (%) 0.78 3.73 1.75
Thorax-only Coordination (%) 0.85 3.99 1.57
Pelvis-only Coordination (%) 0.93 12.22 3.32
Longissimus Activation (%) 0.73 5.38 2.77
Longissimus Co-contraction 0.76 0.03 0.02
Discussion
We tested healthy young adults to investigate the effect of step width on trunk control and
compare the effect of wider step widths and narrower step widths. Participants successfully varied
their step widths based on real-time visual feedback. We found that while center of mass only
varied across step width in the mediolateral direction, trunk kinematics in all three planes were
affected. Consistent with our hypothesis, angular excursions of the thorax, pelvis, and trunk were
affected more by wider widths. This was most apparent in the transverse plane, demonstrating
higher excursions with wider widths. Thorax-pelvis kinematic coordination was affected more by
wider widths in transverse plane and by narrower widths in the frontal plane, while sagittal plane
coordination appears linear to step width. Also, in support of our hypothesis, peak longissimus
activation was greater affected by narrower step widths, resulting in an increased activation.
Bilateral co-contraction of the longissimus decreased as step widths became wider.
37
Participants’ preferred step width was wider than the previously reported value of 0.13 x
leg length (Donelan et al., 2001) due to the different methods of step width calculation (center of
pressure in Donelan et al). Task performance for adhering to the prescribed step widths was good
(Fig III.3). The deviation from the prescribed step width at narrower widths was small and likely
due to increased postural challenges and the range effect (a phenomenon in which the target at the
extreme ranges is under-reached (Poulton, 1975)). Constant error increased as step width becomes
narrower, but overall remained small (less than 3 cm). Variable error was small and consistent
across step width. The consistent and satisfactory performance of step width allowed us to inspect
trunk control variables with confidence.
The clear changes in the center of mass displacement solely in the mediolateral direction
reflected the constraints of the task (modifying step width but maintaining gait speed) (Fig III.4),
while the motor system achieved this by multi-planar adjustments of the within-trunk segments.
In previous studies where different foot placement is prescribed, center of mass behavior changed
correspondingly (Arvin, Mazaheri, et al., 2016; McAndrew Young & Dingwell, 2012; Perry &
Srinivasan, 2017). The mechanical state of the center of mass during the swing phase can also
predict subsequent foot placement (Bruijn & van Dieën, 2018; Hurt et al., 2010; Stimpson et al.,
2018; Wang & Srinivasan, 2014). These studies did not report center of mass behavior in directions
other than mediolateral, nor did they describe detailed trunk kinematics. Our findings showed that
thorax, pelvis, and trunk angles across all three planes were influenced by step widths at different
phases of gait (Fig III.5). It is possible that adjustments are made across multiple planes due to
coupling motion of the spinal structure (Barnes et al., 2009; Legaspi & Edmond, 2007). The motor
system is redundant in that immeasurable combinations of body segment configurations can
38
produce the same center of mass, therefore the multi-planar adjustments of trunk kinematics could
be an exploitation of motor redundancy (Bernstein, 1967).
Changes in angular excursion of the thorax, pelvis, and trunk (thorax-on-pelvis) at different
step widths were most reflected in the transverse plane (Fig III.6). This is interesting given that
step width was mainly manipulated in the frontal plane. The time-series data (Fig III.5) showed
that although differences in kinematics occurred in the frontal plane throughout multiple phases of
gait, they occurred predominantly during mid-ranges of movement, while the differences in
transverse plane kinematics occurred during the peaks and valleys. It is also worth noting that
transverse plane rotation is the dominant motion during walking, and is associated with step length
and arm swings (Pontzer et al., 2009; Saunders et al., 1953). In wider step widths participants
walked with greater step length diagonally since the treadmill speed was fixed, and larger arm
swings were also observed, corresponding to the increased thorax, pelvis, and trunk excursions.
At narrower step widths, step length does not simply become proportionally smaller and the swing
leg would need to move laterally to avoid the stance limb. Therefore, those factors may have
prevented the transverse excursions to further decrease linearly at narrow widths. The trunk
excursions were not simply the difference between thorax and pelvis excursions since these
segments are not always anti-phase. Therefore, an analysis of the thorax-pelvis coordination was
further informative in understanding trunk control at different step widths.
The thorax-pelvis coordination patterns changed with step width, more so during wide
widths in the transverse plane and more so during the narrow widths in the frontal plane (Fig III.7).
Consistent with previous literature (Seay et al., 2011b), in-phase and pelvis-only patterns
dominated the transverse plane. The pelvis-only pattern dominated the frontal plane during
walking, which is logical since the thorax needs to maintain level to stabilize the head for visual
39
input, however, this does not match with the results from Seay and colleagues potentially due to
different walking speed and segment definitions (Seay et al., 2011b). No comparison could be
made to previous literature, however, regarding coordination patterns at different step widths. In
the transverse plane, the time in pelvis-only pattern decreased and thorax-only pattern increased
with wider widths, even though both pelvis and thorax excursions increased. This demonstrated
the fundamental difference in the measurement of angular excursion and coordination, the former
concerning rotation range and the latter concerning the relationship between segments. In the
frontal plane, time in pelvis-only pattern decreased and in-phase pattern increased with narrow
widths. This was supported by increased bilateral longissimus co-contraction which may be a
result of increased active control with smaller base of support (MacKinnon & Winter, 1993).
Longissimus muscle activation and bilateral co-contraction both increased as step width
decreased (Fig III.8). The paraspinal muscles are controlled through a feedforward mechanism and
pre-activate in a top-down manner from cervical to lumbar before heel strike during walking
(Prince et al., 1994). The paraspinal muscles at the lumbar region are the highest activated during
gait (Prince et al., 1994), likely due to its function controlling the large mass and moment arm of
the upper body. The increased peak longissimus activation and co-contraction at narrower step
widths support previous literature pointing to “tightened” postural control (Perry & Srinivasan,
2017) or increased active muscle control contributing to metabolic cost (Donelan et al., 2004). The
peak longissimus activation was further influenced by narrow widths compared to the wide widths.
We speculate that the activation level at wider widths was mostly maintained due to the demands
of counteracting flexion moment of the trunk in the sagittal plane, whereas the co-contraction
continued to decrease because it was a frontal plane-specific measurement.
40
The current study presents how trunk control changes in response to narrow and wide step
widths that has not been previously studied, but not without limitations. Participants were a
convenient sample of college and graduate-level students who are active, young adults, and this
may limit generalizability to other populations. Also, some changes across step widths in trunk
angular excursions or thorax-pelvis kinematic coordination were arguably small (for example, less
than 2 degrees in pelvis transverse excursion). However, they are relatively large when considering
the total excursion or % coordination pattern used in walking and test-retest reliability was good.
This research provides a foundation for future studies on trunk control in various step widths on
different populations, such as individuals with low back pain.
In summary, step width affected trunk control in various ways. Trunk kinematics and
thorax-pelvis coordination in all three planes were affected by step width, with transverse plane
variables greater impacted by wider widths and frontal plane variables greater impacted by
narrower widths. Longissimus activation and bilateral co-contraction both increased with narrow
step widths, and peak activation showed greater changes with narrow step widths. Findings of this
study support the hypothesis that wider and narrow step widths present different demands on trunk
control and offer insight into how the segments within trunk are controlled to maintain stability at
various step widths.
41
CHAPTER IV
TRUNK CONTROL IN AND OUT OF AN EPISODE OF RECURRENT
LOW BACK PAIN
Abstract
Objective: To determine if altered trunk control associated with recurrent LBP persists beyond
symptom duration.
Design: Prospective, laboratory-controlled study with a semi-longitudinal design.
Methods: Twenty young adults with recurrent low back pain were tested once during a painful
episode and once in symptom remission, and twenty back-healthy matched controls were tested.
Motion capture and surface electromyography were used to capture frontal plane trunk kinematics
and muscle activation while participants walked on a treadmill with five prescribed step widths.
Thorax-pelvis coordination was analyzed using vector coding, and longissimus activation and
bilateral co-activation were calculated.
Results: During symptom remission, individuals with recurrent low back pain had greater trunk
excursion and anti-phase coordination and reduced in-phase coordination than when they were in
pain (main pain status effect p<0.002). Compared to back-healthy controls, individuals in
remission also had greater trunk excursion and reduced in-phase coordination (main group effect
p<0.012), while there were no differences when they were in pain. Regardless of pain status, the
recurrent low back pain group had greater pelvis-only and less thorax-only coordination compared
to the control group (main group effect p<0.009). There were no significant interactions between
pain status or group and step width for all dependent variables.
42
Conclusion: Young adults with recurrent LBP exhibited a “loose” trunk control strategy compared
to back-healthy controls in the frontal plane during gait across multiple step widths, regardless of
pain status. This “loose” trunk control strategy was further exaggerated when they were in
symptom remission.
43
Introduction
Low back pain (LBP) is highly prevalent and leads to global burden, and often follow a
recurrent course (GBD 2015 Disease and Injury Incidence and Prevalence Collaborators, 2016).
About 80% of the population is affected by LBP, and up to 56% of the people that experience LBP
exhibit recurrent symptoms within a year (da Silva et al., 2017; Pengel et al., 2003). Recurring
LBP symptoms impact individual’s quality of life, increase the risk of depression and risk of
developing opioid dependence (Bener et al., 2015; Deyo et al., 2015; Hirano et al., 2014).
Furthermore, LBP is responsible for the highest amount of time lived with disability worldwide
and ranks the 3rd together with neck pain on healthcare spending in the United States (Dieleman et
al., 2016; Vos et al., 2016). The majority of LBP is non-specific, with no identifiable cause of
symptoms which makes developing primary prevention programs challenging (Hartvigsen et al.,
2018). Secondary prevention after the initial development of symptoms could greatly reduce
personal and societal cost associated with LBP disability. Despite the attempts to incorporate
prevention of recurrence in clinical management of LBP, the effectiveness of these programs is
still under investigation (Choi et al., 2010; de Campos et al., 2017). The lack of established factors
associated with future recurrence is a critical barrier to the development of effective intervention
programs (da Silva et al., 2017).
Alterations in movement strategies attributed to a history of pain could lead to long-term
adverse consequences if they persist beyond symptom duration. Studies have reported that a
history of previous back pain episodes predicts future recurrences and persistence of symptoms,
while the underlying mechanisms remain elusive (da Silva et al., 2017; Klyne et al., 2019). A
plausible mechanism may be that the motor system adapts to back pain and utilizes suboptimal
control of movement that may be beneficial in the short-term, but lead to long-term adverse
44
consequences (Etemadi et al., 2016; Marshall et al., 2016; Steffens et al., 2015; Van Dieën et al.,
2017; Zazulak et al., 2007c, 2007a). For example, strategies individuals with LBP have been
reported to adopt include increased trunk muscle co-activation and thorax-pelvis kinematic
coupling during walking, which could lead to muscle fatigue, increased spinal loading, and induce
intervertebral disc inflammation (Crosbie et al., 2013; Granata et al., 2005; Marras, Davis, et al.,
2001; Miyagi et al., 2012). However, a review of the literature suggests that persons with LBP
may show a wide distribution of motor control strategies and be at either extreme or overlap with
back-healthy controls (Van Dieën, Reeves, Kawchuk, Van Dillen, et al., 2019). The concept that
there may be either a “tight” (increased muscle activation, increased trunk stiffness and coupling,
decreased segmental movement, leading to more compression loads), or “loose” (decreased muscle
activation, decreased trunk stiffness and coupling, increased segmental movement, leading to more
shear force) control over the trunk in LBP has also been recently presented (Van Dieën, Reeves,
Kawchuk, van Dillen, et al., 2019; Van Dieën, Reeves, Kawchuk, Van Dillen, et al., 2019).
There is evidence that altered motor control persists beyond symptom durations in persons
with recurrent LBP (Hodges et al., 2009; Hooper et al., 2016; Seay et al., 2011a; Smith & Kulig,
2016a, 2016b). Asymptomatic participants with a history of recurrent LBP exhibit impaired
dynamic balance (Hooper et al., 2016), increased trunk stiffness (Hodges et al., 2009), and a more
in-phase thorax-pelvis coordination during running (Seay et al., 2011a, 2011b) and walking (Seay
et al., 2011a). However, results on trunk coordination during walking are inconsistent (Seay et al.,
2011a, 2011b; Smith & Kulig, 2016a). Moreover, these prior studies tested individuals during
symptom remission but did not provide information on how these individuals would perform when
in pain. Therefore, a study of the same cohort designed to capture trunk behavior in and out of pain
is highly warranted.
45
Pain is a potent stimulus and the body sees it as a great cost, therefore we may learn to
adapt to pain in less time, fewer repetitions, but fail to further refine our strategies (Van Dieën et
al., 2017). Individuals with LBP exhibit low movement variability once they have chosen a specific
strategy to move with pain (Hedayati et al., 2014; Van Dieën et al., 2017). An experiment done on
healthy participants found that a lower movement variability during the adaptation to experimental
pain was predictive of failure to return to previous strategy when pain subsides (Moseley &
Hodges, 2006). These studies relied on observations of natural movement variability. It is
unknown whether they are also less able to adapt to varying tasks demands. Since walking is highly
repetitive and ecological, and requires active balance control in the frontal plane (Kuo & Donelan,
2010), we are interested in using systematically changing step widths to examine if individuals
with LBP have an adaptable movement strategy.
The purpose of this study was to compare trunk control in and out of a painful episode in
individuals with recurrent LBP, and to determine if they were altered when compared to back-
healthy individuals. We achieved this by examining frontal plane trunk kinematics and longissimus
muscle activation in response to different step widths during gait. We hypothesized greater trunk
excursion, in-phase coordination, muscle activation, and bilateral co-activation in individuals with
recurrent LBP regardless of pain status. We also hypothesized a decreased response in trunk
control to different step widths in individuals with recurrent LBP compared to controls.
46
Figure IV.1. Participant consort diagram.
Methods
Participants
Sample size calculation based on pilot data revealed that 14 participants in each group
would be sufficient to reach 80% statistical power to detect differences in most trunk coordination
patterns and longissimus activation between pain status within the LBP group and between the
LBP and control groups. Twenty participants with recurrent LBP and twenty matched back-healthy
individuals completed the study (Fig.IV.1). Participants with recurrent LBP were included if they
47
were between 18 and 45 years old, had pain that was localized to the area between the lower
posterior margin of the rib cage and the horizontal gluteal fold for more than 6 months, but had
less than half of the days in pain. They also had to have back pain episodes that were severe enough
to limit function (at least one item on the Oswestry Disability Index > 0, other than the pain item).
They were excluded if they had a history of leg pain below the knee accompanying their LBP, had
chronic or recurrent pain that lasted more than 6 months in other body regions, had a history of
spine or lower extremity surgery, a radiological or clinical diagnosis of spinal stenosis or scoliosis,
spinal malignancy, spinal infection, or lumbar or cervical radiculopathy (by clinical diagnosis or
had altered sensory, motor, or reflex function with screening of the C5-S1 myotomes, dermatomes,
and deep tendon reflexes), a history of diabetes mellitus that affected peripheral sensation,
rheumatic joint disease, polyneuropathy, spinal fracture or dislocation, ankylosing spondylitis,
active cancer, or current pregnancy. They were also excluded if they consumed alcohol for more
than 10 drinks per week, caffeinated drinks for more than 4 cups per day, tobacco for more than
15 cigarettes per day, or had any condition that would prevent completion of the experimental
tasks or is known to severely affect balance or locomotion.
Back-healthy controls were age (± 5 yrs), sex, body mass index (BMI) (in the same
category), and activity (± 15% metabolic equivalents per week based on the International Physical
Activity Questionnaire) matched to the recurrent LBP group. They were included if they had no
LBP in the previous year. They were excluded based on the same criteria as the recurrent LBP
group.
Participants with recurrent LBP were tested twice, first when their pain had persisted for
more than 24 hours at the level of ≥2/10 on a written numeric pain rating scale (0-10) (Ostelo et
al., 2008), and then when their pain was <1/10 on the numeric pain rating scale for more than 24
48
hours, determined by electronic communication. Given the high test-retest reliability from our pilot
study, the back-healthy group were only tested once. Participants were recruited through flyers,
public announcements, and word of mouth at the University of Southern California campus and
its surrounding community. Participants gave written informed consent that was approved by the
institutional review board of the University of Southern California.
Instrumentation
Participants were instrumented with reflective markers placed on the distal feet, heels,
medial and lateral malleoli, medial and lateral femoral epicondyles, greater trochanters, anterior
superior iliac spines, iliac crest, and S1, bilateral acromion, sternal notch, and T1. Tracking
markers were placed using clusters of 3-4 markers on the foot, shank, and thigh. Kinematic data
were recorded by an 11-camera Qualisys motion capture system (Qualisys Inc., Gothenburg,
Sweden) at 125 Hz. Surface electromyography (EMG) was collected using bipolar silver chloride
electrodes with an interelectrode distance of 22 mm. Electrodes were placed on bilateral
longissimus (2 fingers width from the L3 spinous process bony landmark) per guidelines from
SENIAM (Hermens et al., 2006). Data were collected using a Noraxon wireless EMG system
(Noraxon U.S.A, Inc., Scottsdale, AZ, USA) at 1500 Hz. A portable treadmill (PRO-FORM 505
CST, ICON Health & Fitness, Inc., Logan, UT, USA) was used for the walking trials.
Experimental procedures
Participants underwent a clinical screening of scoliosis using the Adam’s forward flexion
test, and a clinical screening of lower extremity sensory, motor, and reflex function for
radiculopathy. They then completed the International Physical Activity Questionnaire (Craig et
49
al., 2003), and medical history form that incorporated the minimal dataset recommended by the
National Institutes for Health task force on research standards for chronic LBP (Deyo et al., 2014).
Individuals with recurrent LBP also completed the Oswestry Disability Index (Fairbank &
Pynsent, 2000), the Fear Avoidance Beliefs Questionnaire (FABQ) (Waddell et al., 1993), and a
body pain diagram to indicate pain location (Southerst et al., 2013). They were also required to
complete several visual analog scales for pain under different conditions, which was separate from
the written numeric rating scale that we used to determine eligibility. Leg length from the greater
trochanter to the ground of the right leg was measured for normalizing step width.
Participants were given 3 minutes to familiarize themselves to the treadmill, after which a
30 second treadmill walking trial at 1.25 m/s was collected to determine preferred step width
(PSW). Participants were then introduced to step widths with real-time visual feedback projected
on the wall in front of the treadmill (Fig IV.2). Step width was calculated using marker data that
was streamed real-time into a custom program in MATLAB (MathWorks, Inc., Natick, MA, USA)
by finding the medial-lateral distance between the averaged heel and 2nd toe marker position with
every foot flat. Participants were instructed to match the width of a red bar indicating their current
step width to the black dotted lines indicating the target width (Fig IV.2). Five different target step
widths were prescribed as a ratio of their PSW, including 0.33, 0.67, 1, 1.33, and 1.67 × PSW.
After one familiarization trial at 1PSW, one 30-second practice trial for each prescribed width was
then carried out. More practice trials were provided for one participant because the participant did
not fully understand the task and was not comfortable enough to perform at an adequate level.
Participants then completed a 30-second trial for each step width presented in a randomized order.
50
Figure IV.2. Experimental setup of treadmill walking with prescribed step widths. A visual feedback was projected
on a wall in front of the treadmill, with a red horizontal bar representing participant’s actual step width and black
vertical lines indicating the target width. The thorax and the pelvis segments are illustrated.
Data Analyses
Kinematic data were low-pass filtered at 10 Hz with a dual-pass 4th order Butterworth filter.
Gait events were identified based on the toe and heel marker position relative to the pelvis
coordinate. Each gait cycle was time-normalized to 101 data points from right heel strike to the
subsequent right heel strike. Each 30-second trial consisted of 25-35 strides, but only the last 20
strides were included in the analyses as participants reached steady state walking with the
prescribed step widths.
The frontal plane trunk excursion and thorax-pelvis kinematic coordination were analyzed
as this was the predominant plane that was sensitive to the changes in active lateral stabilization
demands accompanying changing step width (see Chapter III). Trunk excursion was defined as the
maximum range of thorax rotation relative to the pelvis during gait. Thorax-pelvis kinematic
coordination was calculated using vector coding analysis described by Needham et al (Needham
et al., 2014). Specifically, an angle-angle diagram was first constructed with the pelvic angle on
the horizontal axis and the thorax angle on the vertical axis, then the coupling angle was calculated
as the angle of the vector between two adjacent data points in time relative to the right horizontal
51
(Fig IV.3A). Coordination patterns were categorized as in-phase, anti-phase, thorax-only, and
pelvic-only defined by coupling angles falling within each range indicated in Fig IV.3B (Fig IV.3B
& C).
Figure IV.3. (A) A pelvis-thorax angle-angle diagram of a gait cycle in one representative participant. The “+”
denotes right heel strike and the arrow indicates progression of movement. Coupling angle was defined as the vector
angle of two consecutive points in time relative to the right horizontal. (B) Cutoffs for binning of the coupling
angles into four coordination patterns. (C) Illustrations of the physical implication of the four coordination patterns.
In-phase indicates that both segments are rotating to the same direction at similar rate; anti-phase indicates that the
segments are rotating to the opposite direction at similar rate; thorax-only and pelvis-only indicates that the thorax or
pelvis segment is rotating significantly faster than the other segment, or the other segment is hardly rotating.
52
EMG data were bandpass filtered between 30 Hz and 500 Hz to avoid heart rate
contamination with a dual-pass 4th order Butterworth filter. The signal was then full-wave rectified
and smoothed using a 100 ms moving window and normalized to the averaged peaks of a treadmill
walking trial without prescribed step width or visual feedback (peak gait activation). The peak
longissimus EMG during the contralateral stance phase (where the greater activation occurred)
was calculated. Bilateral co-activation for longissimus was determined as the average ratio
between right and left muscle activation, where the less activated side was always the numerator.
Statistical Analysis
Data were plotted for visualization and examined for normality, homoscedasticity, and
outliers. Descriptive statistics were performed, and paired t-tests were used to compare participant
characteristics, except for sex as the two groups were identical. The analyses were done in steps.
First, all dependent variables including trunk excursion, 4 patterns of trunk coordination, peak
longissimus activation, and bilateral longissimus co-activation was compared within the recurrent
LBP group when they were in active pain (rLBP-A) versus in remission (rLBP-R). Data were
analyzed using general mixed effect models, with step width (here we used the actual step widths
participants achieved, a continuous variable) and pain status (in and out of pain) as fixed effects,
and subjects as a random effect. Then, interaction between step width and pain status was
examined, and if the term was not significant we removed it from the model to allow for
interpretation of main effects. Finally, if there were no significant effects of pain status or
interaction, then data for rLBP-A and rLBP-P were pooled (rLBP) and compared to the back-
healthy control group (CTRL) using general mixed effect models with the same structure but
53
replacing the fixed effect pain status as groups. If an effect of pain status or interaction existed,
then rLBP-A and rLBP-R were compared with CTRL separately. Again, interaction terms were
dropped if not significant. The α level was set at 0.05. Statistical analyses were performed in R (R
Core Team, 2018). Here we only present the main effect of pain status or group which is of primary
interest, please see Appendix B for step width main effects.
Test-retest reliability and stability of performance
We tested 6 back-healthy participants in and out of pain to determine test-retest reliability
and to assess stability of performance. Intraclass correlation coefficients for one-way random
effects, absolute agreement, and multiple measurements (ICC (1,k))(McGraw & Wong, 1996)
were calculated for test-retest reliability on preferred step width, trunk excursion, trunk
coordination, peak longissimus activation, and bilateral longissimus co-activation. We also
planned a post-hoc analyses using paired t-tests to compare test-retest performance of theses 6
controls on any variable that was significantly different between rLBP-A and rLBP-R to examine
the effect of previous exposure to the task.
54
Results
Table IV.1. Participant demographics (mean ± standard deviation).
rLBP (n=20) Control (n=20) P-value
Sex 6 M, 14 F 6 M, 14 F --
Age (years) 25.3 ± 5.2 26.3 ± 3.3 0.192
Height (cm) 167.0 ± 7.8 165.5 ± 9.9 0.534
Weight (kg) 64.2 ± 12.4 61.4 ± 12.7 0.244
BMI (kg/m2) 23.0 ± 4.0 22.2 ± 2.8 0.132
Activity (MET) 3216 ± 2991 2590 ± 1450 0.393
Table IV.2. Mean ± standard deviation (range) for low back pain characteristics. ODI: Oswestry Disability Index;
FABQ-PA: Fear Avoidance Beliefs Questionnaire – Physical Activity Subscale; FABQ-PA: Fear Avoidance Beliefs
Questionnaire –Work Subscale; VAS: Visual Analog Scale. Significant p-values are in bold.
Active Pain In remission P-value
LBP duration (years) 5.8 ± 6.2 (0.58-18) -- --
ODI (0-100) 17.7 ± 7.9 (8-32) 4.7 ± 7.9 (0-16) <0.001
FABQ-PA (0-24) 13.4 ± 5.7 (3-21) 11.8 ± 6.1 (0-21) 0.034
FABQ-W (0-42) 7.7 ± 5.9 (0-17) 6.6 ± 6.2 (0-18) 0.430
Pain at rest VAS (mm) (0-100) 40.3 ± 17.5 (21-72) 1.4 ± 2.4 (0-7) <0.001
Pain during gait VAS (mm)
(0-100)
36.3 ± 20.6 (0-64) 1.7 ± 2.2 (0-7) <0.001
Average pain this episode
VAS (mm) (0-100)
43.6 ± 20.2 (5-83) -- --
Worst pain this episode VAS
(mm) (0-100)
56.9 ± 20.4 (27-87) -- --
Participant characteristics
Participants’ age, sex, weight, height, BMI, and activity level are summarized in Table
IV.1. By study design, no significant differences were found for these characteristics between
individuals with recurrent LBP and their matched counterparts. There were significant differences
55
between rLBP-A and rLBP-R for ODI, FABQ-physical activity subscale, pain at rest, and pain
during gait (Table IV.2). Participants with rLBP returned for testing in remission after 47.9 ± 44.2
days, and the last time they recalled having pain was 10 ± 6.9 days ago. Participants with rLBP
had pain localized between the 12th rib and gluteal fold (Fig IV.4).
Figure IV.4. Self-reported body pain diagram composite indicating rLBP participants’ primary pain location during
the active pain testing session.
Task Performance
There were no differences in preferred step width between CTRL, rLBP-A, and rLBP-R
(Fig IV.5A). Participants were able to vary their step widths based on the visual feedback and there
were no differences in how well they performed the prescribed widths between groups or testing
time points (Fig IV.5B).
56
Figure IV.5. Step width task performance. Both plots showing data for the back-healthy controls (CTRL),
individuals with recurrent low back pain while in active pain (rLBP-A), and the same individuals during symptom
remission (rLBP-R). (A) Mean and standard error of mean of preferred step width (B) Square plot presenting the
mean and standard deviations of actual step width performance relative to the prescribed step widths. Note that the 1
x preferred step width was also prescribed with visual feedback. The diagonal reference line indicates a one-to-one
fit of the performance with the targets.
Trunk excursion and trunk coordination
There were no significant interactions between step width and pain status or step width and
group for any kinematic variables. Regardless of step widths, trunk excursion was significantly
higher in rLBP-R compared to both rLBP-A and CTRL (Fig IV.6). The rLBP-R had reduced in-
phase coordination compared to both rLBP-A and CTRL and had greater anti-phase coordination
compared to when they were in pain (Fig IV.7). The pooled rLBP had greater pelvis-only
coordination and reduced thorax-only coordination than the CTRL group (Fig IV.7).
57
Figure IV.6. The estimate and 95% confidence interval for the effect of pain (left figure, rLBP-R relative to rLBP-
A) and effect of group (right figure, rLBP-A and rLBP-R relative to the CTRL) on frontal plane trunk excursion.
Figure IV.7. The estimate and 95% confidence interval for the effect of pain (left figure of each plot, rLBP-R
relative to rLBP-A) and effect of group (right figure of each plot, rLBP-A and rLBP-R relative to CTRL or rLBP-
pooled relative to CTRL) on frontal plane coordination patterns: in-phase, anti-phase, pelvis-only, and thorax-only.
58
EMG
There were no significant interactions between step width and pain status or step width and
group for any EMG variables. There was no difference between rLBP-A and rLBP-R, or rLBP and
CTRL for peak longissimus activation (Fig IV.8A). The pooled rLBP had reduced bilateral
longissimus co-activation compared to the CTRL (Fig IV.8B).
Figure IV.8. The estimate and 95% confidence interval for the effect of pain (left figure of each plot, rLBP-R
relative to rLBP-A) and effect of group (right figure of each plot, rLBP-A and rLBP-R relative to CTRL or rLBP-
pooled relative to CTRL) on (A) peak longissimus activation, and (B) bilateral longissimus co-activation.
59
Test-retest reliability and stability of performance
Test-retest reliability for the preferred step width was good (ICC=0.85). Test-retest
reliability for trunk excursion was excellent (ICC=0.94), the four patterns of trunk coordination
was good to excellent (ICC=0.78-0.93), and longissimus activation and co-activation was good
(ICC=0.73, 0.76). Trunk excursion, in-phase, and anti-phase coordination was different between
rLBP-A and rLBP-R, therefore we examined the test-retest performance of these variables in the
6 control participants. There was no difference between the two tastings in any of these variables
(p=0.099, 0.603, 0.536) (see Appendix C).
Discussion
We tested individuals with recurrent LBP, in and out of a painful episode and back-healthy
controls, to determine if alterations in trunk control persist beyond symptom duration. Inconsistent
with our hypothesis, we found that in general participants with recurrent LBP utilized a “looser”
trunk control than back-heathy individuals, especially when they were in remission. Regardless of
pain status, individuals with recurrent LBP had greater pelvis-only and reduced thorax-only
coordination and decreased bilateral longissimus co-activation compared to controls. Individuals
during symptom remission of recurrent LBP had greater trunk excursion and reduced in-phase
coordination compared to when in active pain, and also when compared to controls. Inconsistent
with our second hypothesis, participants with recurrent LBP varied their trunk control in response
to different step width similarly to the back-healthy individuals.
Pain status influenced trunk kinematics but not longissimus muscle activation in
participants with recurrent LBP. Consistent with previous literature reporting increased trunk
stiffness in both clinical and experimental LBP, in an attempt to protect the region from unexpected
60
segmental motion, our participants had decreased trunk excursion and increased in-phase
coordination during a painful episode (Colloca & Keller, 2001; Crosbie et al., 2013; Gombatto et
al., 2015; Hodges et al., 2009, 2013; Seay et al., 2011a). However, peak longissimus activation
and bilateral longissimus co-activation did not differ in and out of pain within our participants with
LBP. A previous study with experimentally induced back pain demonstrated increased global
trunk muscle activity, despite highly variable individual muscle activation changes (Hodges et al.,
2013). Our results may suggest that trunk muscles other than the longissimus may contribute to
the “tightened” control of the trunk during a painful episode in our cohort. For example, changes
to the deep (multifidus) and superficial (longissimus and rectus abdominus) trunk muscle
coordination may also play a role in the modulation of trunk kinematics and stability, as predicted
by simple modeling (Van Dieën et al., 2017). In our data, the lack of change in longissimus activity
and pelvis and thorax-only coordination patterns in and out of pain, is a sign of motor adaptation
to LBP that persists beyond symptoms, considering that these variables were altered compared to
back-healthy individuals. Pain is a potent stimulus and the body sees it as a great cost. Therefore
it is possible that persistent movement strategies are developed through reinforcement learning
when adapting to LBP (Van Dieën et al., 2017).
Interestingly, our participants with recurrent LBP generally exhibited a “loose” trunk
control strategy compared to the back-healthy controls, regardless of pain status. Even when they
“tightened” their trunk during a painful episode, they were not “tighter” than the control group. A
series of recently published reviews suggested that there is divergence in motor strategies, “tight”
or “loose” trunk control, that were adopted by potentially different subgroups of patients (Van
Dieën, Reeves, Kawchuk, van Dillen, et al., 2019; Van Dieën, Reeves, Kawchuk, Van Dillen, et
al., 2019). However, there was no sufficient evidence to link certain patient characteristics to
61
“tight” or “loose” trunk control. The participants of the present study are young, active, and have
a history of mildly disabling recurrent LBP for a considerable length of time (about 5 years). They
have fear-avoidance scores within the typical range reported in LBP patients, which may prevent
them from exhibiting a fear- or pain catastrophizing-induced stiffening strategy (Karayannis et al.,
2013; Van Dieën, Reeves, Kawchuk, Van Dillen, et al., 2019).
Another possibility is that altered control of the trunk may present differently given
different demands and nature of the task. Many experiments utilizing fast, discrete perturbations,
such as posterior-anterior forces to the trunk or a sudden release mechanism causing
anteroposterior perturbation, had reported increased trunk stiffness and neuromuscular reflex in
persons with LBP, corresponding to the “tight” control pattern (Colloca & Keller, 2001; Hodges
et al., 2009; Miller et al., 2013). On the other hand, studies utilizing self-initiated tasks such as
forward bending and picking up an object reported increased lumbar excursion during early phase
of the movement in patients with chronic LBP, consistent with the “loose” control pattern (Marich
et al., 2017, 2020). This increased early lumbopelvic movement is consistent with our findings of
persistently increased pelvic-only coordination regardless of pain status. Walking is a task that is
self-initiated and continuous, while simultaneously faced with external mechanical perturbation of
ground impact and reaction forces. Previous studies on walking gait had inconsistent results on
trunk kinematics and coordination, showing either tight” control (Crosbie et al., 2013; Gombatto
et al., 2015; Lamoth et al., 2006; Müller et al., 2015; Seay et al., 2011a, 2011b; van den Hoorn et
al., 2012) or “loose” control (Lamoth et al., 2006; Smith & Kulig, 2016a) of the trunk, or no
difference (Crosbie et al., 2013; Lamoth et al., 2006, 2008) in participants with a history of LBP,
with some symptomatic and some asymptomatic at the time of testing. We have to be mindful that
these studies have greatly varied methods of computing trunk coordination, reported results in
62
different planes of motion, and adopted diverse definitions of the LBP cohort. A study using a self-
initiated, continuous task reported decreased trunk coupling indicating a loose trunk control
strategy, in the frontal plane, consistent with our results (Rowley et al., 2019). Further research is
warranted to elucidate the trunk control alterations seen in persons with LBP as it relates to
multiple factors of different methodology, subgroups, and tasks.
Participants with recurrent LBP did not demonstrate altered ability to adjust their trunk
control to different step width, regardless of pain status, compared to back-healthy controls.
Previous studies have shown results that point to an impaired adaptability in persons with LBP,
such as less increase in multifidus activation to faster gait speed (Smith & Kulig, 2016b), and
reduced trunk coordination variability in the transverse plane during walking (Lamoth et al., 2006)
and running (Seay et al., 2011a). However, there were also results that suggested otherwise, that
is greater coordination variability in the frontal plane during walking (Lamoth et al., 2006), and an
increased change in trunk coordination in response to different gait speed (Seay et al., 2011b) in
persons with LBP compared to controls.
To our knowledge, the current study is the first to examine trunk control during gait in the
same cohort of people in and out of an episode of recurrent LBP. However, the results need to be
interpreted with caution due to the following limitations. To maximize feasibility, we tested
participants first in pain then out of pain, which may introduce a potential issue where participants
performed differently out of pain simply because of previous exposure to the task. We performed
planned post-hoc analyses to check for stability and performance and did not find evidence
indicating previous exposure having an effect. Despite this, it will be ideal to have the pain
condition testing sequence randomized or counterbalanced in future studies. Additionally, stability
of performance in individuals with recurrent LBP may not be the same as the controls, therefore
63
test-retest reliability in this group should be examined in future studies. An important direction to
advance our understanding of recurrent LBP is to explore if the trunk control alterations seen in
pain or during remission can predict future symptoms through a longitudinal design. These studies
may uncover opportunities for designing effective interventions.
In conclusion, young individuals with mildly disabling recurrent LBP exhibited a “loose”
trunk control strategy compared to back-healthy controls including decreased bilateral longissimus
co-activation and greater pelvis-only coordination in the frontal plane during gait, regardless of
pain status. This “loose” trunk control strategy was further exaggerated when they were in
symptom remission, demonstrated by an increased trunk excursion, decreased in-phase and
increased anti-phase coordination.
64
CHAPTER V
ATTENTIONAL PRIORITIZATION AND TRUNK CONTROL UNDER
DUAL-TASK
Abstract
Background: Dual-tasking while walking requires appropriate utilization and allocation of
attentional resources, an ability that may be compromised in clinical conditions such as low back
pain. Pain influences attention and can alter normally largely automatic behaviors such as walking
to become more cognitively demanding, therefore impacting performance. Whether the influence
of pain on attention and performance persists into symptom remission in recurrent low back pain
(LBP) is unclear.
Research question: Does low back pain influence attention, task prioritization, and trunk
coordination during narrow-based walking and if so, does the influence persist after symptoms
subside?
Methods: Twenty young adults with recurrent LBP were tested once in and once out of a painful
episode, and twenty matched back-healthy individuals served as controls. Participants walked on
a treadmill while matching a narrow step width and simultaneously performed serial subtractions
of 7s, with and without instructions to prioritize either task. A motion capture system was used to
record kinematic data, and frontal plane trunk coordination was analyzed using vector coding on
the thorax and pelvis angles. Dependent variables were first compared between pain status within
the LBP group, and then compared between the groups. Paired t-tests were performed on single
and dual-task performance, performance variability, and task prioritization switch, while repeated
65
measures two-way ANOVAs were performed on dual-task performance variability and trunk
coordination.
Results: Active pain negatively impacted single task performance and dual-task performance
variability in both arithmetic and step width tasks for individuals with recurrent LBP. Dual-task
performance and task prioritization switch was not different between pain status or groups.
Compared to the control group, individuals with recurrent low back pain exhibited altered trunk
coordination, including greater pelvis-only and less in-phase and thorax-only patterns regardless
of pain status, and across single and dual-task conditions.
Significance: Task performance that requires attention was impacted only when active pain was
present in young individuals with recurrent low back pain, however, trunk coordination during
walking was altered regardless of pain status, suggesting persistent changes in motor strategies.
These findings inform future clinical research and will contribute to the design of novel
intervention.
66
Introduction
The ability to perform dual-tasks while walking is crucial in daily life, but this ability may
be impacted by pain. Pain is a potent stimulus that captures our attention and occupies executive
resources, which may impact dual-task performance (Legrain et al., 2009; Van Damme et al.,
2010). Furthermore, when pain is associated with movement, such as the case in many
musculoskeletal conditions, this may elicit a greater degree of executive control in normally less
attention-demanding movements (e.g. walking) to avoid inducing additional discomfort (Clark,
2015).
Task prioritization during dual-tasking may also be influenced by movement-related pain.
When a postural task is involved as one of the dual-tasks, appropriate allocation of attentional
resources between tasks is crucial for safe and successful execution of both tasks. Previous
literature has demonstrated that as postural threat increases (e.g., narrowing base of support during
walking), participants shift their priority towards walking over a cognitive task (Kelly et al., 2013).
This effect may be even more profound in individuals with decreased postural reserve (Yogev-
Seligmann et al., 2012), or with increased threat associated with the motor task such as movement-
related pain. Low back pain (LBP) is usually movement-provoked; as such LBP serves as a
reasonable model to investigate the influence of pain on attention and task-prioritization during
dual-task walking.
Studies using a dual-task paradigm on persons with LBP suggest alterations to different
aspects of dual-task performance. Different variables such as spatiotemporal gait variability
(Hamacher et al., 2016; Smith et al., 2017), trunk movement or coordination variability (Hamacher
et al., 2014; Lamoth et al., 2008), and postural control (Etemadi et al., 2016; Mazaheri et al., 2010)
were affected differently in individuals with LBP compared with controls when under dual-task
67
conditions. Due to the location of LBP, we are also interested in how the trunk is controlled during
walking and if it is impacted by dual-task manipulation. Trunk coordination was previously shown
to be more in-phase between the thorax and pelvis during walking, perhaps to reduce movement
between spinal segments for protection (Crosbie et al., 2013; Seay et al., 2011a, 2011b). There
were inconsistent findings, however, reported by studies that introduced a dual-task condition,
with two studies reporting no difference between groups (Lamoth et al., 2008; Smith et al., 2017)
and one study reporting increased in-phase coordination in the LBP group (Rowley et al., 2020).
Dual-task study designs used in previous research often lack a comprehensive evaluation
of both the motor and cognitive task performance, a critical limitation that overlooks the trade-off
between tasks (Plummer & Eskes, 2015). Furthermore, without an explicit manipulation of task
priority, dual-task effects on performance are uninterpretable—is the performance decrement due
to pain (group comparison), interference (dual-task vs single task), and/or task priority (implicit
vs explicit)?
Neuroplasticity may drive persistent changes in dual-task performance when pain becomes
long-term. Alterations in functional connectivity and white matter anatomy in the dorsal-lateral
prefrontal cortex, part of the cognitive/attentional network, has been shown in individuals with
chronic LBP (Mao et al., 2014; Seminowicz et al., 2011). Some dual-task studies tested
participants with LBP during a painful episode (Etemadi et al., 2016; Hamacher et al., 2014, 2016;
Lamoth et al., 2008), while others tested them during symptom remission (Mazaheri et al., 2010;
Smith et al., 2017), but their results were too scattered to draw conclusions in the comparisons.
Whether changes to dual-task performance are solely due to active pain, or if there is a persistent
impact on attentional resources beyond symptom duration is an important question that we aim to
answer with the present study. Due to previous reports on only partial recovery of brain
68
morphology after pain subsides (Čeko et al., 2015) and motor behavior alterations in asymptomatic
individuals with a history of pain (Rowley et al., 2020; Smith et al., 2017; Smith & Kulig, 2016b),
our working hypothesis is that the impact of an active pain episode will persist beyond symptom
remission in those with recurrent LBP.
This study was designed to investigate attention, task prioritization, and trunk control
during gait in and out of a painful episode in a cohort of individuals with recurrent LBP and to
compare them to back-healthy individuals. We hypothesized that attention (measured by dual-task
performance and performance variability) and the ability to switch task priorities during a
challenging walking task (i.e., narrow-based walking) will be diminished in individuals with
recurrent LBP, regardless of pain status, compared to their back-healthy counterparts. Further, we
tested the hypothesis that individuals with recurrent LBP, regardless of pain status, will exhibit
more in-phase trunk coordination than controls, across single and dual-task conditions.
Methods
Participants
According to sample size calculation that was based on pilot data, 16 participants in each
group would be sufficient to reach 80% statistical power to detect differences in step width dual-
task performance, change in step width performance between prioritization instructions, and most
trunk coordination patterns between pain status within the LBP group and between the two groups.
Twenty young adults with recurrent LBP (14 females, 6 males; 25.3 ± 5.2 years; 167.0 ± 7.8 cm;
64.2 ± 12.4 kg; Body Mass Index = 23.0 ± 4.0 kg/m2, Activity= 3216 ± 2991 MET) and twenty
age, sex, Body Mass Index, and activity- matched back-healthy individuals (14 females, 6 males;
26.25 ± 3.31 years; 165.54 ± 9.93 cm; 61.39 ± 12.71 kg; Body Mass Index = 22.21 ± 2.84 kg/m2,
69
Activity= 2590 ± 1450 MET) participated. Participants with recurrent LBP were included if they
are between 18 to 45 years old, have had activity-limiting back pain for more than 6 months, but
had less than half of the days in pain. Back-healthy participants were included if they have had no
back pain in the previous year. Participants were excluded if they have a history of leg pain below
the knee accompanying back pain, have chronic or recurrent pain in other body regions, a history
of low back, lower extremity, or cervical surgery, any known spinal fracture or pathology, a history
of diabetes mellitus, active cancer, current pregnancy, or any condition that is known to affect
balance or locomotion. To eliminate potential effects of stimulants, they were also excluded if they
consume alcohol for more than 10 drinks per week, caffeinated drink for more than 4 cups per day,
or tobacco for more than 15 cigarettes per day.
Participants with recurrent LBP were tested twice, first when their pain persisted for more
than 24 hours at the level of ≥2/10 on the numeric rating scale (Ostelo et al., 2008), and then again
when their pain is <1/10 on the numeric rating scale for more than 24 hours. Given the high test-
retest reliability in the pilot control group, back-healthy individuals were only tested once.
Participants provided written informed consent that was approved by the University of Southern
California institutional review board.
Instrumentation
Participants were instrumented with a typical lower extremity marker-set, with additional
markers placed on the pelvis and thorax including anterior superior iliac spines, iliac crest, and S1,
bilateral acromion, sternal notch, and T1. Kinematic data were recorded by a 11-camera Qualisys
motion capture system (Qualisys Inc., Gothenburg, Sweden) at 125 Hz. A portable treadmill (PRO-
FORM 505 CST, ICON Health & Fitness, Inc., Logan, UT, USA) was used for the walking trials.
70
Experimental procedures
Participants were screened for scoliosis and radiculopathy, and completed a medical
history form, the International Physical Activity Questionnaire (IPAQ) (Craig et al., 2003), and
the Montreal Cognitive Assessment (moCA) (Pike et al., 2016). Individuals with recurrent LBP
also completed the Oswestry Disability Index (ODI) (Fairbank & Pynsent, 2000) and the Fear
Avoidance Beliefs Questionnaire (FABQ) (Waddell et al., 1993).
For all treadmill walking trials, the speed was set at 1.25 m/s. Participants were given 3
minutes to familiarize themselves with the treadmill, after which their preferred step width was
determined by a 30 second treadmill walking trial. Participants were then introduced to the step
width real-time visual feedback projected on the wall in front of the treadmill (Fig V.1A). Step
width was calculated using marker data that streamed real-time into MATLAB (MathWorks, Inc.,
Natick, MA, USA) by finding the medial-lateral distance between the averaged heel and 2nd toe
marker positions every foot flat. Participants were instructed to match their current step width to
the target width indicated on the visual feedback as accurately and consistently as possible. The
target step width was prescribed as 0.33 times their preferred step width. They were also introduced
to the arithmetic task, which required subtracting 7s continuously from a random three-digit
number as fast and as accurately as possible. Participants completed five 30-second step-width
matching practice trials, three 30-second serial subtraction practice trials while seated, and one
dual-task familiarization trial in which they perform the two tasks simultaneously. They then
performed under 5 different conditions, 3 trials each: arithmetic single task (seated), step width
single task, dual-task (without prioritization instructions), step width prioritization (SW-pri), and
arithmetic prioritization (Ari-pri). During the prioritization trials, participants were instructed to
71
do their best on the either the step width or arithmetic task while dual-tasking. The single and dual-
task conditions were randomized and performed first, before the two prioritization conditions were
randomized and administered to avoid contamination of the prioritization condition to the no-
priority-instructed dual-task condition.
Figure V.1. (A) Experimental setup of treadmill walking with prescribed step widths. A visual feedback was
projected on a wall in front of the treadmill, with a red horizontal bar representing participant’s actual step width and
black vertical lines indicating the target width. (B) Pelvis-thorax angle-angle diagram of a gait cycle in one
representative participant. The “+” denotes right heel strike and the arrow indicates progression of movement.
Coupling angle was defined as the vector angle of two consecutive points in time relative to the right horizontal.
(C) Cutoffs for binning of the coupling angles into four coordination patterns. (D) Illustrations of the four thorax-
pelvis coordination patterns. In-phase indicates that both segments are rotating to the same direction at similar rate;
anti-phase indicates that the segments are rotating to the opposite direction at similar rate; thorax-only and pelvis-
only indicate that the thorax or pelvis segment is rotating significantly faster than the other segment, or the other
segment is hardly rotating.
72
Data Analyses
Task performance of the step width task was assessed using root mean square error of
participant’s actual step width relative to the target width. Performance of the arithmetic task was
calculated as the rate of correct response (the number of correct response where they correctly
subtracted 7 from their last answer divided by 30 seconds). Since there was a positive relationship
between single task performance and the magnitude of change from single to dual-task
performance, we calculated the dual-task effect (DTE) =
𝑠𝑖𝑛𝑔𝑙𝑒 𝑡𝑎𝑠𝑘 𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 − 𝑑𝑢𝑎𝑙 𝑡𝑎𝑠𝑘 𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑠𝑖𝑛𝑔𝑙𝑒 𝑡𝑎𝑠𝑘 𝑝𝑒𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 for both tasks to control for single task differences
while assessing the impact of dual-task on participant’s performance. There was no relationship
between single task performance and the change of task performance between prioritization
instructions, therefore we assessed the effect of task prioritization instructions using the raw
difference in task performance between step width- and arithmetic- prioritization conditions. Dual-
task performance variability was calculated as coefficient of variation (CV) of task performance
under the different dual-task conditions from each individual.
Kinematic data were low-pass filtered at 10 Hz with a dual-pass 4th order Butterworth filter.
The frontal plane trunk kinematic coordination was chosen because of its sensitivity in response
to narrow step width (see Chapter III), and was calculated using vector coding analysis described
by Needham et al (Needham et al., 2014). An angle-angle diagram was first constructed, then
coupling angle was determined as the angle of the vector between two adjacent data points in time
relative to the right horizontal (Fig V.1B). Coordination patterns were categorized as in-phase,
anti-phase, thorax-only, and pelvic-only, defined by coupling angles falling within each range
indicated in Fig V.1C (Fig V.1C&D).
73
Statistical Analysis
Descriptive statistics were performed and paired t-tests were used to compare participant
demographics. First, to examine if pain status affects behavior, all dependent variables were
compared within the recurrent LBP group when they were in active pain (LBP-A) and in remission
(LBP-R). Paired t-tests were performed on single task performance, dual-task effect, and changes
in task performance under different prioritization instructions. Two-way repeated-measures
analysis of variance (ANOVA) were used for dual-task performance variability and trunk
coordination, with pain status by dual-task conditions design. A result of no difference between
pain status would support our hypotheses on the persistence of motor behavior in recurrent LBP.
Then, to compare behavior between groups, if there were no significant effects of pain status, data
for LBP-A and LBP-R were pooled and compared to the back-healthy controls (CTRL). If a pain
effect existed, then LBP-A and LBP-R were separately compared with CTRL. If there was a
significant main or interaction effect in the ANOVA results, the Tukey’s test for post-hoc pairwise
comparison was performed. A main effect of group would allow us to confirm our hypotheses
regarding alterations in motor behavior in those with recurrent LBP. A main effect of condition
would indicate the effect of dual-task manipulation on behavior across groups, whereas a
significant interaction effect would indicate that there is a different effect of dual-task manipulation
on those behaviors in the LBP-A, LBP-R, or pooled LBP vs. CTRL. The α level was set at 0.05.
Statistical analyses were done in R (R Core Team, 2018).
Reliability and checking for practice effects
Six back-healthy participants were re-tested at least one week apart to determine test-retest
reliability and to assess potential practice effects of the tasks. For test-retest reliability, intraclass
74
correlation coefficients for one-way random effects, absolute agreement, and multiple
measurements (ICC (1,k)) were calculated for step width and arithmetic DTE and for trunk
coordination (McGraw & Wong, 1996). To assess whether differences between pain status in
individuals with recurrent LBP were merely due to practice effects, we planned a post-hoc analyses
on any variable that was different in and out of pain by performing (1) paired t-tests between the
test-retest sessions in the controls and (2) Pearson’s correlations between the time to re-test and
changes in performance between sessions in those with recurrent LBP. If there was a practice
effect, we would expect to see (1) an improvement of performance, and (2) the closer the two
sessions were, the more improvement in participant’s performance.
Results
Participant characteristics
Per design, no significant differences were found for participants’ age, sex, weight, height,
BMI, and activity level between individuals with recurrent LBP and their matched counterparts
(p>0.05). There was also no significant difference between groups on MoCA scores (LBP: 28.80
1.11; CTRL: 28.65 1.67, p=0.651). Back pain characteristics in the recurrent LBP group are
summarized in Table V.1. There were significant differences in and out of pain for ODI, FABQ-
physical activity subscale, pain at rest, and pain during gait. Participants with recurrent LBP
returned for testing in remission after 47.9 ± 44.2 (8-172) days, and the last time they recalled
having pain was 10 ± 6.9 (range = 1-23) days earlier. There was no difference in preferred step
width for individuals with recurrent LBP in and out of pain, or when compared to the back-healthy
group (p>0.05), therefore the narrow step width should provide a similar degree of challenge to all
participants.
75
Table V.1. Mean ± standard deviation (range) for low back pain characteristics. ODI: Oswestry Disability Index;
FABQ-PA: Fear Avoidance Beliefs Questionnaire – Physical Activity Subscale; FABQ-W: Fear Avoidance Beliefs
Questionnaire –Work Subscale; VAS: Visual Analog Scale. Significant p-values are in bold.
Active Pain In remission P-value
LBP duration (years) 5.8 ± 6.2 (0.58-18) -- --
ODI (0-100) 17.7 ± 7.9 (8-32) 4.7 ± 7.9 (0-16) <0.001
FABQ-PA (0-24) 13.4 ± 5.7 (3-21) 11.8 ± 6.1 (0-21) 0.034
FABQ-W (0-42) 7.7 ± 5.9 (0-17) 6.6 ± 6.2 (0-18) 0.430
Pain at rest VAS (mm) (0-100) 40.3 ± 17.5 (21-72) 1.4 ± 2.4 (0-7) <0.001
Pain during gait VAS (mm) (0-100) 36.3 ± 20.6 (0-64) 1.7 ± 2.2 (0-7) <0.001
Attention – assessed by task performance
When in pain, participants with recurrent LBP performed worse on the arithmetic single
task then when they were in remission (p=0.044) and their back-healthy counterparts (p=0.040)
(Fig V.2). When in pain, individuals with recurrent LBP also performed worse on the step width
single task than when they were out of pain (p=0.006), however, their performance both in and out
of pain did not differ from the control group (Fig V.2).
Figure V.2. Single task arithmetic and step width performance for the back-healthy controls (CTRL), individuals
with recurrent low back pain while in active pain (LBP-A), and the same individuals during symptom remission
(LBP-R). Error bars indicate standard error of mean. n.s. = Non-significant.
76
There were no statistical differences between pain status or groups on arithmetic and step
width dual-task effect (Fig V.3 A&B).
Figure V.3. Dual-task performance and primacy switch effects on the control group (CTRL) and the low back pain
group in pain (LBP-A) and out of pain (LBP-R). Each dot and each pair of connected dots indicate a single
participant. (A) Arithmetic dual-task effect during no instruction condition. (B) Step width dual-task effect during no
instruction condition. (C) The change of arithmetic performance from step width-prioritization to arithmetic
prioritization conditions. (D) The change of step width performance from step width-prioritization to arithmetic
prioritization conditions. n.s. = Non-significant.
Results of dual-task performance variability are presented in Figure V.4. Individuals with
recurrent LBP exhibited greater variability of arithmetic performance during dual-task conditions
when they were in pain compared to in remission (main effect of pain status p=0.018). The
recurrent LBP group in remission had less variability of arithmetic performance compared to the
control group (main effect of group p=0.009). There was no difference between the recurrent LBP
77
group in pain and the control group (p=0.544). Individuals with recurrent LBP exhibited greater
variability of step width performance during dual-task conditions when they were in pain
compared to in remission (main effect of pain status p=0.031). There was a trend that the recurrent
LBP group in pain had more variability of step width performance compared to the control group
(p=0.069). There was no difference between the recurrent LBP group in remission and the control
group (p=0.731). There was no main effect of condition or pain/group x condition interaction for
all comparisons (p>0.05).
Figure V.4. (A) Arithmetic and (B) step width task performance variability during the three dual task conditions
shown as coefficient of variation (CV). §: indicates a pain main effect between individuals with recurrent low back
pain when they were in (LBP-A) and out of pain (LBP-R). *: indicates a group main effect between back-healthy
participants (CTRL) and LBP-R. · : indicates a trend towards significance between CTRL and LBP-A.
Task prioritization
There were no statistical differences between pain status or groups on the effect of task
prioritization switch on arithmetic and step width task performance (Fig V.3 C&D). Both groups
were effectively able to switch task priority to the same degree.
78
Trunk coordination
Results of trunk coordination are presented in Figure V.5. There was no effect of pain status
on any of the four coordination patterns, therefore the results for the LBP group were pooled and
compared with the control group. Regardless of pain status, individuals with recurrent LBP
exhibited greater pelvis-only and less in-phase and thorax-only patterns in gait compared to
controls (p ≤0.001). Single task coordination resulted in less anti-phase pattern compared to the
other three conditions (p<0.05). Single task resulted in more pelvis-only pattern compared to the
dual-task condition (p<0.05). There was no pain/group x condition interaction for any of the
comparisons (p>0.05).
Figure V.5. Thorax-pelvis coordination patterns during single task, dual task (no instruction), step width
prioritization (SW-pri) and arithmetic prioritization (Ari-pri) conditions in the control group (CTRL) and the low
back pain group in pain (LBP-A) and out of pain (LBP-R). Error bars indicate standard error of mean.
79
Reliability and potential practice effects
Test-retest reliability for step width dual-task effect was moderate (ICC=0.668), similar to
the ICCs reported for gait speed dual-task effect (Plummer et al., 2015). Reliability for arithmetic
DTE was poor (ICC=0.232), consistent with a previous report of cognitive task ICCs, and may be
due to the novelty and the challenge of cognitive tasks (Plummer et al., 2015). The ICC for the
four patterns in trunk coordination was 0.78-0.93, representing good to excellent reliability.
Arithmetic and step width single task performance and dual-task performance variability
were the only variables that were different in and out of pain in those with recurrent LBP. There
were no differences between test-retest of the controls in any of those variables (p=0.172-0.953),
and there were no relationships between time to re-test and change in performance in the recurrent
LBP group (p=0.280-0.708), indicating no evidence of a practice effect (Appendix D).
Discussion
We examined dual-task performance, task prioritization, and trunk coordination during
narrow-based walking, for the first time, in and out of a painful episode in a cohort of individuals
with recurrent LBP and compared those variables to those of back-healthy individuals. We found
that participants with recurrent LBP had altered attention when they were in pain indicated by
decreased single task performance and increased dual-task performance variability compared to
when they were in symptom remission. Dual-task performance and the ability to switch task
prioritization were not different in and out of pain in individuals with recurrent LBP, nor when
compared to their back-healthy counterparts, a finding that was inconsistent with our hypothesis.
Additionally, individuals with recurrent LBP, regardless of pain status, exhibited more pelvis-only
coordination and less in-phase and thorax-only patterns than the control group, across single and
80
dual-task conditions. Although this finding is in part consistent with our hypothesis regarding
persistent changes in trunk coordination beyond symptom remission, the direction of the changes,
being less in-phase, is inconsistent with what we predicted.
Active pain negatively affected attention. Our results showed a decline in both the
arithmetic and step width single task performance in individuals with recurrent LBP when they
were in a painful episode compared to in symptom remission, and their arithmetic performance
was also worse than back-healthy participants. Several previous studies found that persons with
active LBP performed worse on cognitive tasks than controls (Etemadi et al., 2016; Lamoth et al.,
2008), while a study on asymptomatic participants with a history of LBP did not find a difference
in single task cognitive performance compared to the control group (Smith et al., 2017). An fMRI
study had also shown that patients with chronic LBP have decreased activation of the cingulo-
frontal-parietal brain network during an attention-demanding task (Mao et al., 2014). Additionally,
our findings indicate that participants with recurrent LBP were less consistent in their dual-task
performance when they were in a painful episode, and their step width performance variability
also trending to be greater than controls. One viable explanation for these variable results is that
active pain may have introduced noise into the sensorimotor system thereby altering the signal to
noise ratio and attenuating relevant sensori-motor signals needed for consistent task performance.
The decreased arithmetic performance variability observed for those with recurrent LBP in
symptom remission could be explained by a practice effect. However, we did not see the same
decrease in back-healthy individuals through repeated testing, thereby ruling out a practice effect
explanation (Appendix D). We found that after normalizing for single task performance, the effect
of dual-task on performance did not differ between pain status in those with recurrent LBP or
between groups. This finding could indicate that even though attention was impacted, the functions
81
involving dual-task performance such as procedural processing and goal-selection were preserved
(Salvucci & Taatgen, 2008). Our results, consistent with previous findings, showed that active pain
affected task performance even in this young, otherwise healthy cohort. The impact of LBP on
daily life and work performance in the long-term may be an issue that warrants further attention.
Inconsistent with our hypothesis, having active LBP or a history of LBP did not impact
participants’ ability to switch task prioritization. Previous literature has shown that individuals
with LBP had a potential tendency to prioritize the motor task (Etemadi et al., 2016; Mazaheri et
al., 2010; Smith et al., 2017). Armour-Smith and colleagues reported no change in step length
consistency during divided attention (Smith et al., 2017), while Etemadi and colleagues reported
increased balance recovery velocity during dual-task in persons with LBP (Etemadi et al., 2016).
If there was prioritization to the motor task, we would expect to see a trade-off between tasks
demonstrated by decreased cognitive performance during dual-task conditions in those with LBP
compared to controls, but this was not observed in these studies. Moreover, the goal of the motor
task was not explicitly defined in previous studies, therefore no conclusions about task
prioritization could be made. Our data showed that participants with recurrent LBP, regardless of
pain status, were able to perform the dual-task and switch task priorities as instructed and to a
similar level as their back-healthy counterparts, based on performance of explicitly defined task
goals.
Participants with recurrent LBP demonstrated persistent alteration of trunk coordination
during narrow-based walking across single and dual-task conditions compared to the back-healthy
group that was not influenced by pain status. Contrary to our hypothesis, they increased time
during gait spent in the pelvis-only pattern and decreased time spent in in-phase and thorax-only
patterns in the frontal plane. There are very few studies on trunk coordination in LBP during dual-
82
task walking, and none of them could be directly compared with our results due to discrepancies
in methodology or study population. Lamoth and colleagues found that patients with chronic LBP
had smaller pelvis-thorax relative phase (more in-phase and coupled) during treadmill walking
across single and dual-task conditions, but the difference was not statistically significant (Lamoth
et al., 2008). Armour Smith and colleagues reported no difference in trunk inter-segmental
coordination between asymptomatic persons with a history of recurrent LBP and controls during
walking turns with and without divided attention (Smith et al., 2017). Both studies mentioned
above reported axial plane coordination, while we chose to focus on the frontal plane given the
sensitivity it demonstrated to the demands of narrowing step width (see Chapter III). A study
reporting frontal plane movement during a continuous balance-dexterity task showed decreased
trunk coupling (corresponding to less in-phase) in persons with recurrent LBP in symptom
remission (Rowley et al., 2019, 2020). More studies are warranted to disambiguate how trunk
coordination is affected by attentional focus and task constraints, in different subgroups of LBP.
Our results do demonstrate, however, that changes in trunk coordination patterns in the recurrent
LBP group were independent of pain status and attentional manipulations. Individuals with
recurrent LBP may have adopted a trunk control strategy through reinforcement learning, which
persisted beyond symptom duration (Van Dieën et al., 2017).
Our repeated measures study design offers a unique opportunity to answer two important
questions. 1) How does active pain perception influence narrow-based walking performance and
trunk coordination under dual-task conditions? and 2) does the adapted behavior persist beyond
symptom duration in recurrent low back pain? Due to feasibility issues we always tested
participants in pain first, which created the risk of a practice effect for the second test administered
during remission. While the post-hoc analyses did not show evidence of a practice effect, our
83
results should be interpreted with caution. Replication studies should aim to randomize the testing
sequence in and out of pain. The participants in our study were young, active adults with only
mildly disabling recurrent LBP, therefore the findings may not be generalizable to those who are
older or have more severe or chronic forms of LBP. In fact, when the effect of aging presents itself
in those young individuals with recurrent LBP, the alterations in attention will likely become
greater. The present study provides insight on the relationship between pain status and attention
in a clinical population of recurrent LBP, expanding on the already abundant literature using
experimentally induced pain.
In conclusion, being in pain negatively impacted cognitive and motor single task
performance and increased the variability of dual-task performance in individuals with recurrent
LBP. On the other hand, compared to age-matched back-healthy participants, individuals with
recurrent LBP consistently demonstrated more pelvis-only and less in-phase and thorax-only
coordination patterns in the frontal plane during narrow-based walking, regardless of pain status
and across single and dual-task conditions. Testing during an active pain episode significantly
influenced task performance in the dual-task conditions, whereas changes in trunk coordination
strategies were found to persist beyond symptom duration.
84
CHAPTER VI
SUMMARY AND CONCLUSIONS
This dissertation examined, for the first time, how trunk control and attention may be
altered in the same cohort of young individuals with recurrent LBP in and out of pain. A solid
investigation of how the trunk is controlled and coordinated during human walking with different
mediolateral stability demands enabled the use of a novel step width matching task on the LBP
population. In chapter III, I characterized trunk control during walking with various step widths in
healthy young adults. In chapter IV, I then determined if trunk control in response to various step
widths differs in and out of pain in individuals with recurrent LBP, and when compared to back-
healthy persons. In chapter V, I determined if attentional prioritization differs in and out of pain in
individuals with recurrent LBP, and when compared to back-healthy persons.
Chapter III presented a detailed characterization of trunk control at five different step
widths (two narrower, one preferred, and two wider) in young adults. Task performance of the
novel step width matching task was good, as participants successfully varied their step widths
based on the target. There was greater error causing wider-than-target step widths during the
narrower prescribed step widths, driven by constant error rather than variable error. This could be
explained by a range effect common in target reaching tasks, and the increased balance demands
at the narrower widths. At the whole-body level, center of mass mediolateral displacement scaled
with step width, while displacements in the vertical and anteroposterior directions did not vary,
potentially to optimize energy retention during walking and minimize metabolic cost. However,
when examining the thorax and pelvis segments independently and in relation to each other, the
effect of different step widths was evident across all three planes of motion. This may be due to
coupling motion of the functional spinal units that occurs because of the anatomy of the vertebrae,
85
and/or the exploitation of multiple degrees of freedom within the segments to increased load
sharing and reduce the risk of injury. Statistical parametric mapping analyses were able to detect
differences in the time series excursion curves of the thorax, pelvis, and trunk (thorax relative to
pelvis) between step widths in all three planes. The deviations of the curves mainly occurred
around the peaks in the transverse and sagittal planes, but around mid-range in the frontal plane.
As a result, the peak-to-peak excursions of the thorax, pelvis, and trunk only differed between step
widths in the transverse and sagittal planes, with the transverse plane showing the largest
difference due to its typically greater excursion during walking.
In chapter III, we also compared the effects of wide and narrow step widths on trunk
kinematics, coordination, and muscle activation, as these two conditions place different demands
on the system. Transverse plane thorax, pelvis, and trunk excursions significantly increased only
in step widths wider than preferred, while narrow step widths did not have an effect on transverse
plane excursion. Transverse plane thorax-pelvis coordination was also only affected by wider
widths, with decreased pelvis-only pattern and increased thorax-only pattern. On the other hand,
narrow step widths were associated with increased peak longissimus activation and bilateral
longissimus co-activation, accompanied by decreased pelvis-only and increased in-phase and
thorax-only patterns in the frontal plane. These findings were mostly in line with our hypothesis.
Furthermore, these results demonstrated that frontal plane trunk coordination and longissimus
activation best reflected the change in active lateral stabilization as step width changes, suitable as
variables of interest for the following chapters.
Chapter IV provided evidence of altered trunk control in individuals with recurrent LBP
that was even more apparent during symptom remission than in pain. Regardless of pain status,
young adults with recurrent LBP had decreased bilateral longissimus co-activation, and greater
86
pelvis-only and less thorax-only coordination during walking compared to their back-healthy
counterparts, across the five different step widths during testing. Although inconsistent with our
hypothesis of a stiffening strategy, this was consistent with a “loose” trunk control strategy that
was proposed to appear in subgroups of patients with LBP during self-initiated, low demanding
tasks. Greater pelvis-only coordination highlighted movement in the lower lumbar spine.
Decreased thorax-only pattern could indicate a less mobile thorax, or a need to further stabilize the
thorax due to increased reliance on visual and vestibular sensory inputs. Furthermore, individuals
during symptom remission of recurrent LBP had greater trunk excursion and reduced in-phase
coordination than when they were in pain, as well as when compared to back-healthy participants.
This indicated that without the stimulus of active pain, persons with recurrent LBP exhibited an
even “looser” trunk control strategy. Based on previous literature that showed decreased
movement variability or dampened response to changing external task demands, we also tested
whether individuals with recurrent LBP had a more rigid strategy that does not respond to changing
step widths. The results showed that there was no interaction between step width and pain status
or group, which was inconsistent with our hypothesis. This study was the first to demonstrate that
individuals with recurrent LBP exhibit persistent alterations in trunk control beyond symptom
duration by testing the same cohort in and out of an episode. Furthermore, it showed that trunk
control may be further altered during symptom remission, when active pain is not present.
Chapter V demonstrated how attention and trunk control during walking was altered in and
out of pain in persons with recurrent LBP by using a dual task paradigm. Participants were required
to perform narrow-based walking by matching their step width to a target, and also perform an
arithmetic task consisting of serial subtractions, both separately as single tasks and simultaneously
as dual task. Active pain negatively impacted single task performance and dual-task performance
87
variability in both arithmetic and step width tasks for individuals with recurrent LBP. This did not
support our hypothesis that changes in attention would persist beyond symptom duration. This
was, however, in line with previous literature indicating that active pain (e.g. experimentally
induced or chronic LBP) captures attention and therefore disrupts task performance and its
consistency. Also inconsistent with our hypothesis, when instructed to switch prioritization
between the two tasks, there was no difference in the ability to switch attentional focus between
pain status within the recurrent LBP group, or between the recurrent LBP and control groups.
Previous findings from other studies of unchanging motor performance in response to dual task
may not be a result of attentional prioritization, but rather a fixed motor strategy that these
individuals with LBP adopted. This speculation was supported by an examination of thorax-pelvis
coordination, which revealed that compared to back-healthy participants, persons with recurrent
LBP, regardless of their pain status, had greater pelvis-only and less in-phase and thorax-only
patterns consistent across single and dual task conditions. The nature of these alterations, however,
was not supporting our original hypothesis of increased in-phase coordination during dual task in
individuals with recurrent LBP, regardless of pain status. Alternatively, these results supported the
presence of a “loose” trunk control strategy that was previously observed in different cohorts of
patients with LBP or during certain tasks that were self-initiated and minimally demanding.
There were consistent findings of a “loose” trunk control strategy used by our cohort of
individuals with recurrent LBP in both chapter IV and V. In chapter IV, less bilateral longissimus
co-activation and greater pelvis-only coordination pattern was found in persons with recurrent LBP
than the control group across different step widths (mechanical manipulation). In chapter V,
greater pelvis-only and less in-phase coordination patterns were found in persons with recurrent
LBP than the control group across single and dual task conditions (attentional manipulation).
88
These repeated findings indicated an altered, yet unchanging trunk control strategy that was a
signature of individuals with recurrent LBP, unperturbed by mechanical or attentional
manipulations. Moreover, this “loose’ trunk control strategy was present both in and out of a
painful episode, indicating persistent changes in motor control in this population.
Revisiting the initial hypotheses. Supporting our hypothesis 1, we found that wide and
narrow step widths have different effects on trunk control in healthy young adults. Trunk excursion
and coordination was more affected by wider widths, while longissimus muscle activation was
more affected by narrower widths. An exception being that frontal plane trunk coordination was
more affected by narrower widths, accompanying the increased bilateral longissimus activation.
In contrast to hypothesis 2a, individuals with recurrent LBP did not have a “tighter” trunk control
(decreased trunk excursions, increased in-phase thorax-pelvis coordination, and increased
longissimus activation and co-activation) than back-healthy individuals, regardless of pain status.
Instead, we found that individuals with recurrent LBP had a “looser” trunk control (increased
pelvis-only coordination, and decreased longissimus co-activation) compared to back-healthy
individuals, and that this strategy was even more exaggerated during symptom remission
(increased trunk excursion and decreased in-phase coordination pattern). Hypothesis 2b, stating
that individuals with recurrent LBP will show less change in trunk control in response to different
step widths than back-healthy individuals, regardless of pain status, was not supported by the
results as there was no difference between pain status or groups. Partially supporting hypothesis
3a that predicted individuals with recurrent LBP to exhibit altered attentional processing
(decreased single task and dual-task performance, altered performance variability) compared to
back-healthy individuals, regardless of pain status, the results showed decreased single task
performance and increased dual-task performance variability in individuals with LBP, but only
89
during a painful episode. The results do not support hypothesis 3b, that Individuals with recurrent
LBP will exhibit decreased ability to switch task prioritization from motor to cognitive task
compared to back-healthy individuals, regardless of pain status. In hypothesis 3c, we hypothesized
that individuals with recurrent LBP will exhibit more in-phase trunk coordination during dual task
conditions compared to back-healthy individuals, regardless of pain status. This was not supported
by the results, which showed a “looser” trunk coordination pattern indicated by increased pelvis-
only and decreased in-phase patterns across single and dual tasks, regardless of pain status in
individuals with recurrent LBP.
Many hypotheses, informed by previous literature, were not met in our study, speaking to
the issue of reproducibility in science. This was potentially due to the inconsistent use of definition
of the study population, diverse use of tasks and methods, and the inherent heterogeneity of low
back pain conditions. Our findings of a “loose” trunk control strategy used by this cohort of persons
with recurrent LBP was consistent with several previous studies, but it is uncertain whether this
strategy was associated with certain subgroups of patients or to certain types of tasks.
There is a significant amount of work that needs to be considered in the future to expand
our current knowledge in LBP and to contribute to clinical implementation. Systematic
examination of different subgroups of patients as well as different types of tasks (e.g. perturbation-
based or self-initiated, level of perceived or actual mechanical challenge) could be done to extract
factors that are associated with a “tight” or “loose” trunk control. Additionally, longitudinal studies
are warranted to determine whether the altered trunk control that persists beyond symptoms, either
a “tight” strategy shown in some previous studies, or a “loose” strategy demonstrated by the
present dissertation, could lead to a higher risk of future episodes. If this relationship could be
established, then these alterations in trunk control are maladaptive and therefore should be targeted
90
by interventions. It would also be interesting to explore if these alterations are present before the
first onset of an episode, making it a risk factor of LBP. Additionally, as the participants in the
current study were young and only mildly disabled, whether the alterations become exaggerated
or present completely different with the effect of aging or prolonged history of LBP warrants
further investigation. Finally, high quality intervention studies that consider the biopsychosocial
model and target maladaptive factors should be conducted to evaluate potential treatment and
prevention programs for LBP. The development of efficient and targeted intervention programs is
desired to provide a cost-effective solution for the huge societal and personal impact caused by
LBP.
This was the first study, to the author’s knowledge, that examined trunk control and
attention in and out of pain in the same cohort of individuals with recurrent LBP and compared
those with a back-healthy control group. A sample size of 20 per group exceeded the 16 per group
indicated by power-analyses based on pilot data, and a 100% retention rate ensured the inclusion
of all participants in the final analyses. The testing always started first when participants were in
pain then in remission to maximize recruitment and retention. This created, however, a potential
risk of practice effect where participants in remission behaved differently due to previous exposure
to the tasks. An analysis of test-retest reliability of 6 control participants indicated generally good
to excellent reliability (Chapter III), and the second testing did not result in statistically better
performance of the tasks (Appendix C&D). Additionally, in the recurrent LBP group there was no
relationship between time to retest and the change of their behavior between the two testing, which
indicated no “fading” of a practice effect should it be present in the first place (Appendix D).
Taking these results together, there was no evidence of a practice effect and therefore any
91
difference between the two testing sessions that could be confidently attributed to an effect of pain
status.
In conclusion, this was the first study that demonstrated the presence of altered trunk
control both in and out of an episode of pain in the same cohort of individuals with recurrent LBP.
Individuals with recurrent LBP, regardless of pain status, exhibited a “loose” trunk control strategy
characterized by decreased muscle activation, increased pelvis-only, and decreased in-phase
coordination pattern during walking. This strategy was consistently present despite mechanical
(step width) and attentional (dual task) manipulations. The presence of active pain in individuals
with recurrent LBP was associated with a trunk control strategy that was slightly more similar to
their back-healthy counterparts’, however, it was also associated with altered attention that
impacted the performance of motor and arithmetic tasks. These findings serve as a steppingstone
for future work aiming to develop intervention programs for LBP.
Figure VI.1. Conceptual figure summarizing key findings regarding altered trunk control in recurrent low
back pain.
92
REFERENCES
Abboud, J., Nougarou, F., Lardon, A., Dugas, C., & Descarreaux, M. (2016). Influence of
lumbar muscle fatigue on trunk adaptations during sudden external perturbations. Frontiers
in Human Neuroscience, 10(November), 576. https://doi.org/10.3389/fnhum.2016.00576
Arvin, M., Mazaheri, M., Hoozemans, M. J. M., Pijnappels, M., Burger, B. J., Verschueren, S.
M. P., & van Dieën, J. H. (2016). Effects of narrow base gait on mediolateral balance
control in young and older adults. Journal of Biomechanics, 49(7), 1264–1267.
https://doi.org/10.1016/j.jbiomech.2016.03.011
Arvin, M., van Dieën, J. H., & Bruijn, S. M. (2016). Effects of constrained trunk movement on
frontal plane gait kinematics. Journal of Biomechanics, 49(13), 3085–3089.
https://doi.org/10.1016/j.jbiomech.2016.07.015
Barnes, D., Stemper, B. D., Yogananan, N., Baisden, J. L., & Pintar, F. A. (2009). Normal
coupling behavior between axial rotation and lateral bending in the lumbar spine - biomed
2009. Biomedical Sciences Instrumentation, 45, 131–136.
http://www.ncbi.nlm.nih.gov/pubmed/19369752
Bauby, C. E., & Kuo, A. D. (2000). Active control of lateral balance in human walking. Journal
of Biomechanics, 33(11), 1433–1440. https://doi.org/10.1016/S0021-9290(00)00101-9
Bener, A., Dafeeah, E. E., & Salem, M. O. (2015). Determinants of depression and somatisation
symptoms in low back pain patients and its treatment: global burden of diseases. Journal of
the Pakistan Medical Association, 65(5), 473–479.
Bernstein, N. (1967). The Co-ordination and Regulation of Movements. Pergamon Press.
Boucher, J.-A., Abboud, J., & Descarreaux, M. (2012). The influence of acute back muscle
fatigue and fatigue recovery on trunk sensorimotor control. Journal of Manipulative and
93
Physiological Therapeutics, 35(9), 662–668. https://doi.org/10.1016/j.jmpt.2012.10.003
Bruijn, S. M., & van Dieën, J. H. (2018). Control of human gait stability through foot placement.
Journal of the Royal Society, Interface, 15(143). https://doi.org/10.1098/rsif.2017.0816
Čeko, M., Shir, Y., Ouellet, J. A., Ware, M. A., Stone, L. S., & Seminowicz, D. A. (2015).
Partial recovery of abnormal insula and dorsolateral prefrontal connectivity to cognitive
networks in chronic low back pain after treatment. Human Brain Mapping, 36(6), 2075–
2092. https://doi.org/10.1002/hbm.22757
Choi, B. K., Verbeek, J. H., Tam, W. W.-S., & Jiang, J. Y. (2010). Exercises for prevention of
recurrences of low-back pain. The Cochrane Database of Systematic Reviews, 1,
CD006555. https://doi.org/10.1002/14651858.CD006555.pub2
Cholewicki, J., Breen, A., Popovich, J. M., Reeves, N. P., Sahrmann, S. A., Van Dillen, L. R.,
Vleeming, A., & Hodges, P. W. (2019). Can biomechanics research lead to more effective
treatment of low back pain? A point-counterpoint debate. Journal of Orthopaedic and
Sports Physical Therapy, 49(6), 425–436. https://doi.org/10.2519/jospt.2019.8825
Clark, D. J. (2015). Automaticity of walking: functional significance, mechanisms, measurement
and rehabilitation strategies. Frontiers in Human Neuroscience, 9(May), 246.
https://doi.org/10.3389/fnhum.2015.00246
Collins, S. H., & Kuo, A. D. (2013). Two independent contributions to step variability during
over-ground human walking. PloS One, 8(8), e73597.
https://doi.org/10.1371/journal.pone.0073597
Colloca, C. J., & Keller, T. S. (2001). Stiffness and neuromuscular reflex response of the human
spine to posteroanterior manipulative thrusts in patients with low back pain. Journal of
Manipulative and Physiological Therapeutics, 24(8), 489–500.
94
https://doi.org/10.1067/mmt.2001.118209
Craig, C. L., Marshall, A. L., Sjöström, M., Bauman, A. E., Booth, M. L., Ainsworth, B. E.,
Pratt, M., Ekelund, U., Yngve, A., Sallis, J. F., & Oja, P. (2003). International physical
activity questionnaire: 12-Country reliability and validity. Medicine and Science in Sports
and Exercise, 35(8), 1381–1395. https://doi.org/10.1249/01.MSS.0000078924.61453.FB
Crosbie, J., de Faria Negrão Filho, R., Nascimento, D. P., & Ferreira, P. (2013). Coordination of
spinal motion in the transverse and frontal planes during walking in people with and without
recurrent low back pain. Spine, 38(5), E286-92.
https://doi.org/10.1097/BRS.0b013e318281de28
da Silva, T., Mills, K., Brown, B. T., Herbert, R. D., Maher, C. G., & Hancock, M. J. (2017).
Risk of recurrence of low back pain: a systematic review. The Journal of Orthopaedic and
Sports Physical Therapy, 47(5), 305–313. https://doi.org/10.2519/jospt.2017.7415
da Silva, T., Mills, K., Kongsted, A., Maher, C., & Hancock, M. (2020). What is the personal
impact of recurrences of low back pain? Subanalysis of an inception cohort study. The
Journal of Orthopaedic and Sports Physical Therapy, 53(9), 1–18.
https://doi.org/10.2519/jospt.2020.9345
de Campos, T. F., Maher, C. G., Clare, H. A., da Silva, T. M., & Hancock, M. J. (2017).
Effectiveness of McKenzie Method-based self-management approach for the secondary
prevention of a recurrence of low back pain (SAFE Trial): protocol for a pragmatic
randomized controlled trial. Physical Therapy, 97(8), 799–806.
https://doi.org/10.1093/ptj/pzx046
Dean, J. C., Alexander, N. B., & Kuo, A. D. (2007). The effect of lateral stabilization on walking
in young and old adults. IEEE Transactions on Biomedical Engineering, 54(11), 1919–
95
1926. https://doi.org/10.1109/TBME.2007.901031
Deyo, R. A., Dworkin, S. F., Amtmann, D., Andersson, G., Borenstein, D., Carragee, E.,
Carrino, J., Chou, R., Cook, K., DeLitto, A., Goertz, C., Khalsa, P., Loeser, J., Mackey, S.,
Panagis, J., Rainville, J., Tosteson, T., Turk, D., Von Korff, M., & Weiner, D. K. (2014).
Report of the NIH task force on research standards for chronic low back pain. Spine (Phila
Pa 1976), 39(14), 1128–1143. https://doi.org/10.1097/BRS.0000000000000434
Deyo, R. A., Rainville, J., & Kent, D. L. (1992). What can the history and physical examination
tell us about low back pain? JAMA, 268(6), 760–765.
http://www.ncbi.nlm.nih.gov/pubmed/1386391
Deyo, R. A., Von Korff, M., & Duhrkoop, D. (2015). Opioids for low back pain. Bmj, 350(jan05
10), g6380–g6380. https://doi.org/10.1136/bmj.g6380
Dieleman, J. L., Baral, R., Birger, M., Bui, A. L., Bulchis, A., Chapin, A., Hamavid, H., Horst,
C., Johnson, E. K., Joseph, J., Lavado, R., Lomsadze, L., Reynolds, A., Squires, E.,
Campbell, M., DeCenso, B., Dicker, D., Flaxman, A. D., Gabert, R., … Murray, C. J. L.
(2016). US Spending on personal health care and public health, 1996-2013. JAMA, 316(24),
2627. https://doi.org/10.1001/jama.2016.16885
Donelan, J. M., Kram, R., & Kuo, A. D. (2001). Mechanical and metabolic determinants of the
preferred step width in human walking. Proceedings. Biological Sciences, 268(1480),
1985–1992. https://doi.org/10.1098/rspb.2001.1761
Donelan, J. M., Shipman, D. W., Kram, R., & Kuo, A. D. (2004). Mechanical and metabolic
requirements for active lateral stabilization in human walking. Journal of Biomechanics,
37(6), 827–835. https://doi.org/10.1016/j.jbiomech.2003.06.002
Drake, J. D. M., & Callaghan, J. P. (2006). Elimination of electrocardiogram contamination from
96
electromyogram signals: An evaluation of currently used removal techniques. Journal of
Electromyography and Kinesiology, 16(2), 175–187.
https://doi.org/10.1016/j.jelekin.2005.07.003
Etemadi, Y., Salavati, M., Arab, A. M., & Ghanavati, T. (2016). Balance recovery reactions in
individuals with recurrent nonspecific low back pain: effect of attention. Gait & Posture,
44, 123–127. https://doi.org/10.1016/j.gaitpost.2015.11.017
Fairbank, J. C. T., & Pynsent, P. B. (2000). The Oswestry Disability Index. Spine (Phila Pa
1976), 25(22), 2940–2953. https://doi.org/10.1097/00007632-200011150-00017
Foster, N. E., Anema, J. R., Cherkin, D., Chou, R., Cohen, S. P., Gross, D. P., Ferreira, P. H.,
Fritz, J. M., Koes, B. W., Peul, W., Turner, J. A., Maher, C. G., Buchbinder, R., Hartvigsen,
J., Cherkin, D., Foster, N. E., Maher, C. G., Underwood, M., van Tulder, M., … Woolf, A.
(2018). Prevention and treatment of low back pain: evidence, challenges, and promising
directions. The Lancet, 6736(18), 1–16. https://doi.org/10.1016/S0140-6736(18)30489-6
Fougnie, D. Y. (2008). The relationship between attention and working memory. In N. B.
Johansen (Ed.), New research on short-term memory (pp. 1–45). Nova Sicence Publishers.
Friedly, J., Standaert, C., & Chan, L. (2010). Epidemiology of spine care: the back pain
dilemma. Physical Medicine and Rehabilitation Clinics of North America, 21(4), 659–677.
https://doi.org/10.1016/j.pmr.2010.08.002
GBD 2015 Disease and Injury Incidence and Prevalence Collaborators. (2016). Global, regional,
and national incidence, prevalence, and years lived with disability for 310 diseases and
injuries, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015.
Lancet (London, England), 388(10053), 1545–1602. https://doi.org/10.1016/S0140-
6736(16)31678-6
97
Gombatto, S. P., Brock, T., DeLork, A., Jones, G., Madden, E., & Rinere, C. (2015). Lumbar
spine kinematics during walking in people with and people without low back pain. Gait &
Posture, 42(4), 539–544. https://doi.org/10.1016/j.gaitpost.2015.08.010
Granata, K. P., Lee, P. E., & Franklin, T. C. (2005). Co-contraction recruitment and spinal load
during isometric trunk flexion and extension. Clinical Biomechanics, 20(10), 1029–1037.
https://doi.org/10.1016/j.clinbiomech.2005.07.006
Hak, L., Houdijk, H., van der Wurff, P., Prins, M. R., Mert, A., Beek, P. J., & van Dieën, J. H.
(2013). Stepping strategies used by post-stroke individuals to maintain margins of stability
during walking. Clinical Biomechanics (Bristol, Avon), 28(9–10), 1041–1048.
https://doi.org/10.1016/j.clinbiomech.2013.10.010
Hamacher, D., Hamacher, D., Herold, F., & Schega, L. (2016). Are there differences in the dual-
task walking variability of minimum toe clearance in chronic low back pain patients and
healthy controls? Gait & Posture, 49, 97–101.
https://doi.org/10.1016/j.gaitpost.2016.06.026
Hamacher, D., Hamacher, D., & Schega, L. (2014). A cognitive dual task affects gait variability
in patients suffering from chronic low back pain. Experimental Brain Research, 232(11),
3509–3513. https://doi.org/10.1007/s00221-014-4039-1
Hartvigsen, J., Hancock, M. J., Kongsted, A., Louw, Q., Ferreira, M. L., Genevay, S., Hoy, D.,
Karppinen, J., Pransky, G., Sieper, J., Smeets, R. J., Underwood, M., & Lancet Low Back
Pain Series Working Group. (2018). What low back pain is and why we need to pay
attention. Lancet (London, England), 391(10137), 2356–2367.
https://doi.org/10.1016/S0140-6736(18)30480-X
Hedayati, R., Kahrizi, S., Parnianpour, M., Bahrami, F., Kazemnejad, A., & Mobini, B. (2014).
98
The study of the variability of anticipatory postural adjustments in patients with recurrent
non-specific low back pain. Journal of Back and Musculoskeletal Rehabilitation, 27(1), 33–
40. https://doi.org/10.3233/BMR-130416
Hermens, H., Freriks, B., Merletti, R., Rau, G., Disselhorst-Klug, C., Stegeman, D., & Hagg, G.
(2006). The Surface ElectroMyoGraphy for the Non-Invasive Assessment of Muscles
Project (SENIAM).
Hicks, G. E., Sions, J. M., Coyle, P. C., & Pohlig, R. T. (2017). Altered spatiotemporal
characteristics of gait in older adults with chronic low back pain. Gait & Posture, 55, 172–
176. https://doi.org/10.1016/j.gaitpost.2017.04.027
Hirano, K., Imagama, S., Hasegawa, Y., Ito, Z., Muramoto, A., & Ishiguro, N. (2014). Impact of
low back pain, knee pain, and timed up-and-go test on quality of life in community-living
people. Journal of Orthopaedic Science, 19(1), 164–171. https://doi.org/10.1007/s00776-
013-0476-0
Hodges, P. W., Coppieters, M. W., MacDonald, D., & Cholewicki, J. (2013). New insight into
motor adaptation to pain revealed by a combination of modelling and empirical approaches.
European Journal of Pain (United Kingdom), 17(8), 1138–1146.
https://doi.org/10.1002/j.1532-2149.2013.00286.x
Hodges, P. W., & Tucker, K. (2011). Moving differently in pain: a new theory to explain the
adaptation to pain. Pain, 152(3 Suppl), S90-8. https://doi.org/10.1016/j.pain.2010.10.020
Hodges, P. W., van den Hoorn, W., Dawson, A., & Cholewicki, J. (2009). Changes in the
mechanical properties of the trunk in low back pain may be associated with recurrence.
Journal of Biomechanics, 42(1), 61–66. https://doi.org/10.1016/j.jbiomech.2008.10.001
Hof, A. L., & Duysens, J. (2013). Responses of human hip abductor muscles to lateral balance
99
perturbations during walking. Experimental Brain Research, 230(3), 301–310.
https://doi.org/10.1007/s00221-013-3655-5
Hof, A. L., Gazendam, M. G. J., & Sinke, W. E. (2005). The condition for dynamic stability.
Journal of Biomechanics, 38(1), 1–8. https://doi.org/10.1016/j.jbiomech.2004.03.025
Hooper, T. L., James, C. R., Brismée, J.-M., Rogers, T. J., Gilbert, K. K., Browne, K. L., &
Sizer, P. S. (2016). Dynamic balance as measured by the Y-Balance Test is reduced in
individuals with low back pain: A cross-sectional comparative study. Physical Therapy in
Sport, 22, 29–34. https://doi.org/10.1016/j.ptsp.2016.04.006
Hoy, D., Bain, C., Williams, G., March, L., Brooks, P., Blyth, F., Woolf, A., Vos, T., &
Buchbinder, R. (2012). A systematic review of the global prevalence of low back pain.
Arthritis and Rheumatism, 64(6), 2028–2037. https://doi.org/10.1002/art.34347
Hurt, C. P., Rosenblatt, N., Crenshaw, J. R., & Grabiner, M. D. (2010). Variation in trunk
kinematics influences variation in step width during treadmill walking by older and younger
adults. Gait and Posture, 31(4), 461–464. https://doi.org/10.1016/j.gaitpost.2010.02.001
Karayannis, N. V., Smeets, R. J. E. M., van den Hoorn, W., & Hodges, P. W. (2013). Fear of
movement is related to trunk stiffness in low back pain. PloS One, 8(6), e67779.
https://doi.org/10.1371/journal.pone.0067779
Kelly, V. E., Eusterbrock, A. J., & Shumway-Cook, A. (2013). Factors influencing dynamic
prioritization during dual-task walking in healthy young adults. Gait and Posture, 37(1),
131–134. https://doi.org/10.1016/j.gaitpost.2012.05.031
Klyne, D. M., Moseley, G. L., Sterling, M., Barbe, M. F., & Hodges, P. W. (2019). Are signs of
central sensitization in acute low back pain a precursor to poor outcome? Journal of Pain,
20(8), 994–1009. https://doi.org/10.1016/j.jpain.2019.03.001
100
Ko, S.-U., Gunter, K. B., Costello, M., Aum, H., MacDonald, S., White, K. N., Snow, C. M., &
Hayes, W. C. (2007). Stride width discriminates gait of side-fallers compared to other-
directed fallers during overground walking. Journal of Aging and Health, 19(2), 200–212.
https://doi.org/10.1177/0898264307299308
Kongsted, A., Kent, P., Axen, I., Downie, A. S., & Dunn, K. M. (2016). What have we learned
from ten years of trajectory research in low back pain? BMC Musculoskeletal Disorders,
17(1). https://doi.org/10.1186/s12891-016-1071-2
Kongsted, A., Kent, P., Hestbaek, L., & Vach, W. (2015). Patients with low back pain had
distinct clinical course patterns that were typically neither complete recovery nor constant
pain. A latent class analysis of longitudinal data. Spine Journal, 15(5), 885–894.
https://doi.org/10.1016/j.spinee.2015.02.012
Koo, T. K., & Li, M. Y. (2016). A guideline of selecting and reporting intraclass correlation
coefficients for reliability research. Journal of Chiropractic Medicine, 15(2), 155–163.
https://doi.org/10.1016/j.jcm.2016.02.012
Kubinski, S. N., McQueen, C. A., Sittloh, K. A., & Dean, J. C. (2015). Walking with wider steps
increases stance phase gluteus medius activity. Gait & Posture, 41(1), 130–135.
https://doi.org/10.1016/j.gaitpost.2014.09.013
Kuo, A. D., & Donelan, J. M. (2010). Dynamic principles of gait and their clinical implications.
Physical Therapy, 90(2), 157–174. https://doi.org/10.2522/ptj.20090125
Kuo, A. D., Donelan, J. M., & Ruina, A. (2005). Energetic consequences of walking like an
inverted pendulum: step-to-step transitions. Exercise and Sport Sciences Reviews, 33(2),
88–97. https://doi.org/10.1097/00003677-200504000-00006
Lamoth, C. J. C., Meijer, O. G., Daffertshofer, A., Wuisman, P. I. J. M., & Beek, P. J. (2006).
101
Effects of chronic low back pain on trunk coordination and back muscle activity during
walking: changes in motor control. European Spine Journal, 15(1), 23–40.
https://doi.org/10.1007/s00586-004-0825-y
Lamoth, C. J. C., Stins, J. F., Pont, M., Kerckhoff, F., & Beek, P. J. (2008). Effects of attention
on the control of locomotion in individuals with chronic low back pain. Journal of
Neuroengineering and Rehabilitation, 5, 13. https://doi.org/10.1186/1743-0003-5-13
Legaspi, O., & Edmond, S. L. (2007). Does the evidence support the existence of lumbar spine
coupled motion? A critical review of the literature. Journal of Orthopaedic and Sports
Physical Therapy, 37(4), 169–178. https://doi.org/10.2519/jospt.2007.2300
Legrain, V., Damme, S. Van, Eccleston, C., Davis, K. D., Seminowicz, D. A., & Crombez, G.
(2009). A neurocognitive model of attention to pain: behavioral and neuroimaging
evidence. Pain, 144(3), 230–232. https://doi.org/10.1016/j.pain.2009.03.020
MacDonald, D., Moseley, G. L., & Hodges, P. W. (2009). Why do some patients keep hurting
their back? Evidence of ongoing back muscle dysfunction during remission from recurrent
back pain. Pain, 142(3), 183–188. https://doi.org/10.1016/j.pain.2008.12.002
MacKinnon, C. D., & Winter, D. A. (1993). Control of whole body balance in the frontal plane
during human walking. Journal of Biomechanics, 26(6), 633–644.
https://doi.org/10.1016/0021-9290(93)90027-C
Maher, C., Underwood, M., & Buchbinder, R. (2017). Non-specific low back pain. The Lancet,
389(10070), 736–747. https://doi.org/10.1016/S0140-6736(16)30970-9
Mao, C. P., Zhang, Q. L., Bao, F. X., Liao, X., Yang, X. L., & Zhang, M. (2014). Decreased
activation of cingulo-frontal-parietal cognitive/attention network during an attention-
demanding task in patients with chronic low back pain. Neuroradiology, 56(10), 903–912.
102
https://doi.org/10.1007/s00234-014-1391-6
Marich, A. V., Hwang, C. T., Salsich, G. B., Lang, C. E., & Van Dillen, L. R. (2017).
Consistency of a lumbar movement pattern across functional activities in people with low
back pain. Clinical Biomechanics, 44, 45–51.
https://doi.org/10.1016/j.clinbiomech.2017.03.004
Marich, A. V., Hwang, C. T., Sorensen, C. J., & van Dillen, L. R. (2020). Examination of the
lumbar movement pattern during a clinical test and a functional activity test in people with
and without low back pain. PM and R, 12(2), 140–146. https://doi.org/10.1002/pmrj.12197
Marras, W. S., Davis, K. G., Ferguson, S. A., Lucas, B. R., & Gupta, P. (2001). Spine loading
characteristics of patients with low back pain compared with asymptomatic individuals.
Spine (Phila Pa 1976), 26(23), 2566–2574. https://doi.org/10.1097/00007632-200112010-
00009
Marras, W. S., Ferguson, S. A., Burr, D., Davis, K. G., & Gupta, P. (2001). Spine loading in
patients with low back pain during asymmetric lifting exertions. The Spine Journal, 4(1),
64–75. https://doi.org/10.1016/S1529-9430(03)00424-8
Marshall, L. M., Litwack-Harrison, S., Cawthon, P. M., Kado, D. M., Deyo, R. A., Makris, U.
E., Carlson, H. L., Nevitt, M. C., & Study of Osteoporotic Fractures (SOF) Research Group.
(2016). A prospective study of back pain and risk of falls among older community-dwelling
women. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences,
71(9), 1177–1183. https://doi.org/10.1093/gerona/glv225
Mazaheri, M., Salavati, M., Negahban, H., Sanjari, M. A., & Parnianpour, M. (2010). Postural
sway in low back pain: effects of dual tasks. Gait & Posture, 31(1), 116–121.
https://doi.org/10.1016/j.gaitpost.2009.09.009
103
McAndrew Young, P. M., & Dingwell, J. B. (2012). Voluntarily changing step length or step
width affects dynamic stability of human walking. Gait & Posture, 35(3), 472–477.
https://doi.org/10.1016/j.gaitpost.2011.11.010
McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intraclass correlation
coefficients. Psychological Methods, 1(1), 30–46. https://doi.org/10.1037/1082-989X.1.1.30
Miller, E. M., Bazrgari, B., Nussbaum, M. A., & Madigan, M. L. (2013). Effects of exercise-
induced low back pain on intrinsic trunk stiffness and paraspinal muscle reflexes. Journal of
Biomechanics, 46(4), 801–805. https://doi.org/10.1016/j.jbiomech.2012.11.023
Miyagi, M., Ishikawa, T., Kamoda, H., Suzuki, M., Murakami, K., Shibayama, M., Orita, S.,
Eguchi, Y., Arai, G., Sakuma, Y., Kubota, G., Oikawa, Y., Ozawa, T., Aoki, Y., Toyone,
T., Takahashi, K., Inoue, G., Kawakami, M., & Ohtori, S. (2012). ISSLS prize winner: disc
dynamic compression in rats produces long-lasting increases in inflammatory mediators in
discs and induces long-lasting nerve injury and regeneration of the afferent fibers
innervating discs: a pathomechanism for chronic discogeni. Spine, 37(21), 1810–1818.
https://doi.org/10.1097/BRS.0b013e31824ffac6
Mok, N. W., Brauer, S. G., & Hodges, P. W. (2007). Failure to use movement in postural
strategies leads to increased spinal displacement in low back pain. Spine, 32(19), E537–
E543. https://doi.org/10.1097/BRS.0b013e31814541a2
Moseley, G. L., & Hodges, P. W. (2006). Reduced variability of postural strategy prevents
normalization of motor changes induced by back pain: a risk factor for chronic trouble?
Behavioral Neuroscience, 120(2), 474–476. https://doi.org/10.1037/0735-7044.120.2.474
Müller, R., Ertelt, T., & Blickhan, R. (2015). Low back pain affects trunk as well as lower limb
movements during walking and running. Journal of Biomechanics, 48(6), 1009–1014.
104
https://doi.org/10.1016/j.jbiomech.2015.01.042
Needham, R., Naemi, R., & Chockalingam, N. (2014). Quantifying lumbar-pelvis coordination
during gait using a modified vector coding technique. Journal of Biomechanics, 47(5),
1020–1026. https://doi.org/10.1016/j.jbiomech.2013.12.032
Neptune, R. R., & McGowan, C. P. (2016). Muscle contributions to frontal plane angular
momentum during walking. Journal of Biomechanics, 49(13), 2975–2981.
https://doi.org/10.1016/j.jbiomech.2016.07.016
Ostelo, R. W. J. G., Deyo, R. A., Stratford, P., Waddell, G., Croft, P., Von Korff, M., Bouter, L.
M., & de Vet, H. C. (2008). Interpreting change scores for pain and functional status in low
back pain: towards international consensus regarding minimal important change. Spine,
33(1), 90–94. https://doi.org/10.1097/BRS.0b013e31815e3a10
Parreira, R. B., Amorim, C. F., Gil, A. W., Teixeira, D. C., Bilodeau, M., & Da Silva, R. A.
(2013). Effect of trunk extensor fatigue on the postural balance of elderly and young adults
during unipodal task. European Journal of Applied Physiology, 113(8), 1989–1996.
https://doi.org/10.1007/s00421-013-2627-6
Pataky, T. C. (2010). Generalized n-dimensional biomechanical field analysis using statistical
parametric mapping. Journal of Biomechanics.
https://doi.org/10.1016/j.jbiomech.2010.03.008
Pataky, T. C. (2012). One-dimensional statistical parametric mapping in Python. Computer
Methods in Biomechanics and Biomedical Engineering, 15(3), 295–301.
https://doi.org/10.1080/10255842.2010.527837
Pengel, L. H. M., Herbert, R. D., Maher, C. G., & Refshauge, K. M. (2003). Acute low back
pain: systematic review of its prognosis. BMJ (Clinical Research Ed.), 327(7410), 323.
105
https://doi.org/10.1136/bmj.327.7410.323
Perry, J. A., & Srinivasan, M. (2017). Walking with wider steps changes foot placement control,
increases kinematic variability and does not improve linear stability. Royal Society Open
Science, 4(9), 160627. https://doi.org/10.1098/rsos.160627
Pike, N. A., Poulsen, M. K., & Woo, M. A. (2016). Validity of the Montreal Cognitive
Assessment Screener in adolescents and young adults with and without congenital heart
disease. Nursing Research, 66(3), 222–230.
https://doi.org/10.1097/NNR.0000000000000192
Plummer, P., & Eskes, G. (2015). Measuring treatment effects on dual-task performance: a
framework for research and clinical practice. Frontiers in Human Neuroscience, 9(April),
1–7. https://doi.org/10.3389/fnhum.2015.00225
Plummer, P., Grewal, G., Najafi, B., & Ballard, A. (2015). Instructions and skill level influence
reliability of dual-task performance in young adults. Gait & Posture, 41(4), 964–967.
https://doi.org/10.1016/j.gaitpost.2015.03.348
Pontzer, H., Holloway, J. H., Raichlen, D. A., & Lieberman, D. E. (2009). Control and function
of arm swing in human walking and running. Journal of Experimental Biology, 212(4),
523–534. https://doi.org/10.1242/jeb.024927
Poulton, E. C. (1975). Range effects in experiments on people. The American Journal of
Psychology, 88(1), 3–32. https://doi.org/https://doi.org/10.2307/1421662
Prince, F., Winter, D., Stergiou, P., & Walt, S. (1994). Anticipatory control of upper body
balance during human locomotion. Gait & Posture, 2(1), 19–25.
https://doi.org/10.1016/0966-6362(94)90013-2
Prins, M. R., Griffioen, M., Veeger, T. T. J., Kiers, H., Meijer, O. G., van der Wurff, P., Bruijn,
106
S. M., & van Dieën, J. H. (2018). Evidence of splinting in low back pain? A systematic
review of perturbation studies. European Spine Journal, 27(1), 40–59.
https://doi.org/10.1007/s00586-017-5287-0
R Core Team. (2018). R: A language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.
Rankin, B. L., Buffo, S. K., & Dean, J. C. (2014). A neuromechanical strategy for mediolateral
foot placement in walking humans. Journal of Neurophysiology, 112(2), 374–383.
https://doi.org/10.1152/jn.00138.2014
Rebula, J. R., Ojeda, L. V., Adamczyk, P. G., & Kuo, A. D. (2017). The stabilizing properties of
foot yaw in human walking. Journal of Biomechanics, 53, 1–8.
https://doi.org/10.1016/j.jbiomech.2016.11.059
Reeves, N. P., Narendra, K. S., & Cholewicki, J. (2011). Spine stability: lessons from balancing
a stick. Clinical Biomechanics (Bristol, Avon), 26(4), 325–330.
https://doi.org/10.1016/j.clinbiomech.2010.11.010
Robertson, E. M. (2009). From creation to consolidation: a novel framework for memory
processing. PLoS Biology, 7(1), e19. https://doi.org/10.1371/journal.pbio.1000019
Rowley, K. M., Smith, J. A., & Kulig, K. (2019). Reduced trunk coupling in persons with
recurrent low back pain is associated with greater deep-to-superficial trunk muscle
activation ratios during the balance-dexterity task. Journal of Orthopaedic and Sports
Physical Therapy, 49(12), 887–898. https://doi.org/10.2519/jospt.2019.8756
Rowley, K. M., Winstein, C. J., & Kulig, K. (2020). Persons in remission from recurrent low
back pain alter trunk coupling under dual-task interference during a dynamic balance task.
Experimental Brain Research, Advanced online publication.
107
https://doi.org/10.1007/s00221-020-05772-4
Salvucci, D. D., & Taatgen, N. A. (2008). Threaded cognition: an integrated theory of concurrent
multitasking. Psychological Review, 115(1), 101–130. https://doi.org/10.1037/0033-
295X.115.1.101
Saunders, J., Inman, V., & Eberhart, H. (1953). The major determinants in normal and
pathological gait. The Journal of Bone and Joint Surgery. American Volume, 35-A(3), 543–
558. https://doi.org/10.1007/978-1-4471-5451-8_102
Seay, J. F., Van Emmerik, R. E. A., & Hamill, J. (2011a). Low back pain status affects pelvis-
trunk coordination and variability during walking and running. Clinical Biomechanics,
26(6), 572–578. https://doi.org/10.1016/j.clinbiomech.2010.11.012
Seay, J. F., Van Emmerik, R. E. a, & Hamill, J. (2011b). Influence of low back pain status on
pelvis-trunk coordination during walking and running. Spine (Phila Pa 1976), 36(16),
E1070-9. https://doi.org/10.1097/BRS.0b013e3182015f7c
Seminowicz, D. A., & Davis, K. D. (2006). Cortical responses to pain in healthy individuals
depends on pain catastrophizing. Pain, 120(3), 297–306.
https://doi.org/10.1016/j.pain.2005.11.008
Seminowicz, D. A., Wideman, T. H., Naso, L., Hatami-Khoroushahi, Z., Fallatah, S., Ware, M.
A., Jarzem, P., Bushnell, M. C., Shir, Y., Ouellet, J. A., & Stone, L. S. (2011). Effective
treatment of chronic low back pain in humans reverses abnormal brain anatomy and
function. Journal of Neuroscience, 31(20), 7540–7550.
https://doi.org/10.1523/JNEUROSCI.5280-10.2011
Smith, J. A., Gordon, J., & Kulig, K. (2017). The influence of divided attention on walking turns:
effects on gait control in young adults with and without a history of low back pain. Gait &
108
Posture, 58(June), 498–503. https://doi.org/10.1016/j.gaitpost.2017.09.019
Smith, J. A., & Kulig, K. (2016a). Trunk-pelvis coordination during turning: a cross sectional
study of young adults with and without a history of low back pain. Clinical Biomechanics
(Bristol, Avon), 36, 58–64. https://doi.org/10.1016/j.clinbiomech.2016.05.011
Smith, J. A., & Kulig, K. (2016b). Altered multifidus recruitment during walking in young
asymptomatic individuals with a history of low back pain. The Journal of Orthopaedic and
Sports Physical Therapy, 46(5), 365–374. https://doi.org/10.2519/jospt.2016.6230
Southerst, D., Côté, P., Stupar, M., Stern, P., & Mior, S. (2013). The reliability of body pain
diagrams in the quantitative measurement of pain distribution and location in patients with
musculoskeletal pain: a systematic review. Journal of Manipulative and Physiological
Therapeutics, 36(7), 450–459. https://doi.org/10.1016/j.jmpt.2013.05.021
Srinivasan, D., & Mathiassen, S. E. (2012). Motor variability in occupational health and
performance. Clinical Biomechanics (Bristol, Avon), 27(10), 979–993.
https://doi.org/10.1016/j.clinbiomech.2012.08.007
Steffens, D., Ferreira, M. L., Latimer, J., Ferreira, P. H., Koes, B. W., Blyth, F., Li, Q., & Maher,
C. G. (2015). What triggers an episode of acute low back pain? A case-crossover study.
Arthritis Care & Research, 67(3), 403–410. https://doi.org/10.1002/acr.22533
Stewart, W. F., Ricci, J. a, Chee, E., Morganstein, D., & Lipton, R. (2003). Lost productive time
and cost due to common pain conditions in the US workforce. JAMA, 290(18), 2443–2454.
https://doi.org/10.1001/jama.290.18.2443
Stimpson, K. H., Heitkamp, L. N., Horne, J. S., & Dean, J. C. (2018). Effects of walking speed
on the step-by-step control of step width. Journal of Biomechanics, 68, 78–83.
https://doi.org/10.1016/j.jbiomech.2017.12.026
109
Tsao, H., Galea, M. P., & Hodges, P. W. (2008). Reorganization of the motor cortex is
associated with postural control deficits in recurrent low back pain. Brain, 131(8), 2161–
2171. https://doi.org/10.1093/brain/awn154
Tsao, Henry, Danneels, L. A., & Hodges, P. W. (2011). ISSLS prize winner: Smudging the
motor brain in young adults with recurrent low back pain. Spine, 36(21), 1721–1727.
https://doi.org/10.1097/BRS.0b013e31821c4267
Van Damme, S., Legrain, V., Vogt, J., & Crombez, G. (2010). Keeping pain in mind: A
motivational account of attention to pain. Neuroscience and Biobehavioral Reviews, 34(2),
204–213. https://doi.org/10.1016/j.neubiorev.2009.01.005
van den Hoorn, W., Bruijn, S. M., Meijer, O. G., Hodges, P. W., & van Dieën, J. H. (2012).
Mechanical coupling between transverse plane pelvis and thorax rotations during gait is
higher in people with low back pain. Journal of Biomechanics, 45(2), 342–347.
https://doi.org/10.1016/j.jbiomech.2011.10.024
Van Dieën, J. H., Flor, H., & Hodges, P. W. (2017). Low-Back Pain Patients Learn to Adapt
Motor Behavior With Adverse Secondary Consequences. Exercise and Sport Sciences
Reviews, 45(4), 223–229. https://doi.org/10.1249/JES.0000000000000121
Van Dieën, J. H., Kingma, I., & Van Der Bug, J. C. E. (2003). Evidence for a role of antagonistic
cocontraction in controlling trunk stiffness during lifting. Journal of Biomechanics, 36(12),
1829–1836. https://doi.org/10.1016/S0021-9290(03)00227-6
Van Dieën, J. H., Reeves, N. P., Kawchuk, G., van Dillen, L., & Hodges, P. W. (2019). Analysis
of motor control in low-back pain patients: a key to personalized care? The Journal of
Orthopaedic and Sports Physical Therapy, 49(6), 380–388.
https://doi.org/10.2519/jospt.2019.7916
110
Van Dieën, J. H., Reeves, N. P., Kawchuk, G., Van Dillen, L. R., & Hodges, P. W. (2019).
Motor control changes in low back pain: divergence in presentations and mechanisms.
Journal of Orthopaedic and Sports Physical Therapy, 49(6), 370–379.
https://doi.org/10.2519/jospt.2019.7917
Van Tulder, M., Koes, B., & Bombardier, C. (2002). Low back pain. Best Practice and
Research: Clinical Rheumatology, 16(5), 761–775. https://doi.org/10.1053/berh.2002.0267
Vos, T., Allen, C., Arora, M., Barber, R. M., Brown, A., Carter, A., Casey, D. C., Charlson, F. J.,
Chen, A. Z., Coggeshall, M., Cornaby, L., Dandona, L., Dicker, D. J., Dilegge, T., Erskine,
H. E., Ferrari, A. J., Fitzmaurice, C., Fleming, T., Forouzanfar, M. H., … Zuhlke, L. J.
(2016). Global, regional, and national incidence, prevalence, and years lived with disability
for 310 diseases and injuries, 1990–2015: a systematic analysis for the Global Burden of
Disease Study 2015. The Lancet, 388(10053), 1545–1602. https://doi.org/10.1016/S0140-
6736(16)31678-6
Waddell, G., Newton, M., Henderson, I., Somerville, D., & Main, C. J. (1993). A Fear-
Avoidance Beliefs Questionnaire (FABQ) and the role of fear-avoidance beliefs in chronic
low back pain and disability. Pain, 52(2), 157–168. https://doi.org/10.1016/0304-
3959(93)90127-b
Wang, Y., & Srinivasan, M. (2014). Stepping in the direction of the fall: the next foot placement
can be predicted from current upper body state in steady-state walking. Biology Letters,
10(9), 1–5. https://doi.org/10.1098/rsbl.2014.0405
White, A. A., & Gordon, S. L. (1982). Synopsis: workshop on idiopathic low-back pain. Spine,
7(2), 141–149. https://doi.org/10.1097/00007632-198203000-00009
Winter, D. A. (2009). Biomechanics and Motor Control of Human Movement (4th ed.). John
111
Wiley & Sons.
Yogev-Seligmann, G., Hausdorff, J. M., & Giladi, N. (2012). Do we always prioritize balance
when walking? Towards an integrated model of task prioritization. Movement Disorders,
27(6), 765–770. https://doi.org/10.1002/mds.24963
Zazulak, B. T., Hewett, T. E., Reeves, N. P., Goldberg, B., & Cholewicki, J. (2007a). Deficits in
neuromuscular control of the trunk predict knee injury risk: a prospective biomechanical-
epidemiologic study. American Journal of Sports Medicine, 35(7), 1123–1130.
https://doi.org/10.1177/0363546507301585
Zazulak, B. T., Hewett, T. E., Reeves, N. P., Goldberg, B., & Cholewicki, J. (2007b). The effects
of core proprioception on knee injury: a prospective biomechanical-epidemiological study.
American Journal of Sports Medicine, 35(3), 368–373.
https://doi.org/10.1177/0363546506297909
Zazulak, B. T., Hewett, T. E., Reeves, N. P., Goldberg, B., & Cholewicki, J. (2007c). The effects
of core proprioception on knee injury: A prospective biomechanical-epidemiological study.
American Journal of Sports Medicine, 35(3), 368–373.
https://doi.org/10.1177/0363546506297909
112
APPENDIX A
Chapter III Detailed Statistical Results
Table A.1. Summary of p-values for pairwise comparisons between step widths for step width error, trunk excursion,
trunk coordination, and peak longissimus activation and bilateral longissimus co-activation.
Variable
Step
Width
SW=0.33 SW=0.67 SW=1 SW=1.33 SW=1.67
Constant SW
Error
0.33 -- 0.003 <0.001 <0.001 <0.001
0.67 -- 0.347 0.062 <0.001
1 -- 0.931 0.060
1.33 -- 0.338
1.67 --
Variable SW
Error
0.33 -- 0.014 0.001 0.004 0.073
0.67 -- 0.972 0.996 0.980
1 -- 0.999 0.760
1.33 -- 0.880
1.67 --
Transverse
Thorax
Excursion
0.33 -- 0.998 0.455 <0.001 <0.001
0.67 -- 0.667 <0.001 <0.001
1 -- 0.006 <0.001
1.33 -- 0.001
1.67 --
0.33 -- 0.999 1.000 0.390 0.018
113
Transverse
Pelvis
Excursion
0.67 -- 0.999 0.248 0.007
1 -- 0.401 0.018
1.33 -- 0.683
1.67 --
Transverse
Trunk
Excursion
0.33 -- 0.999 0.850 <0.001 <0.001
0.67 -- 0.943 0.001 <0.001
1 -- 0.017 <0.001
1.33 -- 0.781
1.67 --
Frontal
Thorax
Excursion#
0.33 -- na na na na
0.67 -- na na na
1 -- na na
1.33 -- na
1.67 --
Frontal
Pelvis
Excursion#
0.33 -- na na na na
0.67 -- na na na
1 -- na na
1.33 -- na
1.67 --
Frontal
Trunk
Excursion#
0.33 -- na na na na
0.67 -- na na na
1 -- na na
114
1.33 -- na
1.67 --
Sagittal
Thorax
Excursion#
0.33 -- na na na na
0.67 -- na na na
1 -- na na
1.33 -- na
1.67 --
Sagittal
Pelvis
Excursion
0.33 -- 0.996 0.953 0.003 <0.001
0.67 -- 0.815 <0.001 <0.001
1 -- 0.035 <0.001
1.33 -- 0.297
1.67 --
Sagittal
Trunk
Excursion
0.33 -- 0.999 0.222 0.036 0.001
0.67 -- 0.335 0.067 0.003
1 -- 0.944 0.404
1.33 -- 0.857
1.67 --
Transverse
In-phase#
0.33 -- na na na na
0.67 -- na na na
1 -- na na
1.33 -- na
1.67 --
115
Transverse
Anti-phase
0.33 -- 0.995 0.808 0.041 0.111
0.67 -- 0.957 0.115 0.258
1 -- 0.433 0.680
1.33 -- 0.995
1.67 --
Transverse
Thorax-only
0.33 -- 0.997 1.000 0.810 0.007
0.67 -- 1.000 0.944 0.023
1 -- 0.874 0.012
1.33 -- 0.163
1.67 --
Transverse
Pelvis-only
0.33 -- 1.000 1.000 0.685 0.005
0.67 -- 1.000 0.589 0.003
1 -- 0.565 0.002
1.33 -- 0.191
1.67 --
Frontal In-
phase
0.33 -- 0.002 0.002 0.009 0.197
0.67 -- 1.000 0.990 0.498
1 -- 0.992 0.517
1.33 -- 0.794
1.67 --
Frontal Anti-
phase#
0.33 -- na na na na
0.67 -- na na na
116
1 -- na na
1.33 -- na
1.67 --
Frontal
Thorax-only
0.33 -- 0.655 0.015 0.036 0.217
0.67 -- 0.393 0.570 0.944
1 -- 0.999 0.848
1.33 -- 0.947
1.67 --
Frontal
Pelvis-only
0.33 -- 0.033 <0.001 0.011 0.044
0.67 -- 0.730 0.997 1.000
1 -- 0.899 0.669
1.33 -- 0.992
1.67 --
Sagittal In-
phase
0.33 -- 0.023 <0.001 <0.001 <0.001
0.67 -- 0.304 <0.001 <0.001
1 -- 0.136 <0.001
1.33 -- 0.143
1.67 --
Sagittal Anti-
phase
0.33 -- 0.007 <0.001 <0.001 <0.001
0.67 -- 0.876 0.007 <0.001
1 -- 0.116 <0.001
1.33 -- 0.230
117
1.67 --
Sagittal
Thorax-only
0.33 -- 0.980 0.055 0.002 <0.001
0.67 -- 0.209 0.014 <0.001
1 -- 0.844 0.092
1.33 -- 0.582
1.67 --
Sagittal
Pelvis-only
0.33 -- 1.000 0.763 0.376 0.056
0.67 -- 0.821 0.441 0.075
1 -- 0.973 0.559
1.33 -- 0.901
1.67 --
Right
Longissimus
Peak
Activation
0.33 -- 0.025 <0.001 <0.001 <0.001
0.67 -- 0.408 0.406 0.197
1 -- 1.000 0.994
1.33 -- 0.994
1.67 --
Left
Longissimus
Peak
Activation
0.33 -- 0.370 <0.001 <0.001 <0.001
0.67 -- 0.122 0.190 0.003
1 -- 1.000 0.724
1.33 -- 0.597
1.67 --
118
#One-way repeated measures ANOVA F-test not significant.
Longissimus
Co-
contraction
0.33 -- 0.935 0.413 0.005 <0.001
0.67 -- 0.878 0.063 <0.001
1 -- 0.436 0.007
1.33 -- 0.453
1.67 --
119
Table A.2. Pairwise comparisons between step widths for constant step width error (ANOVA F=12.96, p<0.001).
*Significant at the 0.05 level
Table A.3. Pairwise comparisons between step widths for variable step width error (ANOVA F=4.65, p=0.0021).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -9.970* 2.483 <0.001
1 -14.867* 2.483 <0.001
1.33 -14.059* 2.483 <0.001
1.67 -14.705* 2.483 <0.001
0.67
1 -4.898 2.483 0.279
1.33 -4.090 2.483 0.467
1.67 -4.736 2.483 0.313
1
1.33 0.808 2.483 0.998
1.67 0.162 2.483 1.000
1.33 1.67 -0.646 2.483 0.999
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -3.850* 1.222 0.014
1 -4.609* 1.222 0.002
1.33 -4.313* 1.222 0.003
1.67 -3.161 1.222 0.073
0.67
1 -0.758 1.222 0.972
1.33 -0.463 1.222 0.996
1.67 0.689 1.222 0.980
1
1.33 0.296 1.222 0.999
1.67 1.447 1.222 0.760
1.33 1.67 1.15 1.222 0.880
120
Table A.4. Pairwise comparisons between step widths for transverse plane thorax excursion (ANOVA F=28.50,
p<0.0001).
*Significant at the 0.05 level
Table A.5. Pairwise comparisons between step widths for transverse plane pelvis excursion (ANOVA F=4.238,
p=0.004).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 0.133 0.407 0.998
1 0.678 0.407 0.455
1.33 2.068* 0.407 <0.001
1.67 3.627* 0.407 <0.001
0.67
1 0.545 0.407 0.667
1.33 1.935* 0.407 <0.001
1.67 3.494* 0.407 <0.001
1
1.33 1.390* 0.407 0.006
1.67 2.949* 0.407 <0.001
1.33 1.67 1.559* 0.407 0.001
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -0.121 0.456 0.999
1 0.008 0.456 1.000
1.33 0.808 0.456 0.390
1.67 1.406* 0.456 0.017
0.67
1 0.129 0.456 0.999
1.33 0.929 0.456 0.248
1.67 1.527* 0.456 0.007
1
1.33 0.799 0.456 0.401
1.67 1.398* 0.456 0.018
1.33 1.67 0.599 0.456 0.683
121
Table A.6. Pairwise comparison between step widths for transverse plane trunk excursion (ANOVA F=11.50,
p=<0.0001).
*Significant at the 0.05 level
Table A.7. Pairwise comparisons between step widths for sagittal plane pelvis excursion (ANOVA F=12.977,
p<0.001).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 0.077 0.303 0.999
1 0.307 0.303 0.850
1.33 1.246* 0.303 <0.001
1.67 1.594* 0.303 <0.001
0.67
1 0.230 0.303 0.943
1.33 1.169* 0.303 0.001
1.67 1.517* 0.303 <0.001
1
1.33 0.940* 0.303 0.017
1.67 1.287* 0.303 <0.001
1.33 1.67 0.348* 0.303 0.781
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -0.049 0.133 0.996
1 0.095 0.133 0.953
1.33 0.475* 0.133 0.003
1.67 0.732* 0.133 <0.001
0.67
1 0.144 0.133 0.815
1.33 0.524* 0.133 <0.001
1.67 0.781* 0.133 <0.001
1
1.33 0.380* 0.133 0.035
1.67 0.637* 0.133 <0.001
1.33 1.67 0.258 0.133 0.297
122
Table A.8. Pairwise comparisons between step widths for sagittal plane trunk excursion (ANOVA F=5.534,
p<0.001).
*Significant at the 0.05 level
Table A.9. Pairwise comparisons between step widths for transverse plane anti-phase coordination (ANOVA
F=3.014, p=0.023).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -0.036 0.158 0.999
1 -0.331 0.158 0.222
1.33 -0.451* 0.158 0.036
1.67 -0.609* 0.158 0.001
0.67
1 -0.296 0.158 0.335
1.33 -0.415* 0.158 0.067
1.67 -0.572* 0.158 0.003
1
1.33 -0.119 0.158 0.944
1.67 -0.277 0.158 0.404
1.33 1.67 -0.158 0.158 0.857
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 0.323 0.808 0.995
1 0.888 0.808 0.808
1.33 2.263* 0.808 0.041
1.67 1.953 0.808 0.111
0.67
1 0.565 0.808 0.957
1.33 1.940 0.808 0.115
1.67 1.630 0.808 0.258
1
1.33 1.375 0.808 0.433
1.67 1.065 0.808 0.680
1.33 1.67 -0.310 0.808 0.995
123
Table A.10. Pairwise comparisons between step widths for transverse plane thorax-only coordination (ANOVA
F=3.830, p=0.007).
*Significant at the 0.05 level
Table A.11. Pairwise comparisons between step widths for transverse plane pelvis-only coordination (ANOVA
F=4.955, p=0.001).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 0.553 1.615 0.997
1 0.220 1.615 1.000
1.33 1.765 1.615 0.810
1.67 5.393* 1.615 0.008
0.67
1 -0.333 1.615 1.000
1.33 1.213 1.615 0.944
1.67 4.840* 1.615 0.023
1
1.33 1.545 1.615 0.874
1.67 5.173* 1.615 0.012
1.33 1.67 3.628 1.615 0.163
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 0.278 1.855 1.000
1 0.345 1.855 1.000
1.33 -2.430 1.855 0.685
1.67 -6.455* 1.855 0.005
0.67
1 0.068 1.855 1.000
1.33 -2.708 1.855 0.589
1.67 -6.733* 1.855 0.003
1
1.33 -2.775 1.855 0.565
1.67 -6.800* 1.855 0.002
1.33 1.67 -4.025 1.855 0.191
124
Table A.12. Pairwise comparison between step widths for frontal plane in-phase coordination (ANOVA F=5.007,
p=0.001).
*Significant at the 0.05 level
Table A.13. Pairwise comparisons between step widths for frontal plane thorax-only coordination (ANOVA F=3.171,
p=0.018).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -5.570* 1.484 0.002
1 -5.528* 1.484 0.002
1.33 -4.865* 1.484 0.009
1.67 -3.198 1.484 0.197
0.67
1 0.043 1.484 1.000
1.33 0.705 1.484 0.990
1.67 2.373 1.484 0.498
1
1.33 0.663 1.484 0.992
1.67 2.330 1.484 0.517
1.33 1.67 -1.668 1.484 0.794
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -1.438 1.059 0.655
1 -3.310* 1.059 0.015
1.33 -3.015* 1.059 0.036
1.67 -2.233 1.059 0.217
0.67
1 -1.873 1.059 0.393
1.33 -1.578 1.059 0.570
1.67 -0.795 1.059 0.944
1
1.33 0.295 1.059 0.999
1.67 1.078 1.059 0.848
1.33 1.67 0.783 1.059 0.947
125
Table A.14. Pairwise comparisons between step widths for sagittal plane in-phase coordination (ANOVA F=27.432,
p<0.001).
*Significant at the 0.05 level
Table A.15. Pairwise comparisons between step widths for sagittal plane anti-phase coordination (ANOVA F=22.432,
p<0.001).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 3.902* 1.301 0.023
1 6.407* 1.301 <0.001
1.33 9.435* 1.301 <0.001
1.67 12.435* 1.301 <0.001
0.67
1 2.505 1.301 0.304
1.33 5.533* 1.301 <0.001
1.67 8.532* 1.301 <0.001
1
1.33 3.028 1.301 0.136
1.67 6.027* 1.301 <0.001
1.33 1.67 3.000 1.301 0.143
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -3.330* 0.985 0.006
1 -4.270* 0.985 <0.001
1.33 -6.633* 0.985 <0.001
1.67 -8.680* 0.985 <0.001
0.67
1 -0.940 0.985 0.876
1.33 -3.303* 0.985 0.007
1.67 -5.350* 0.985 <0.001
1
1.33 -2.363 0.985 0.116
1.67 -4.410* 0.985 <0.001
1.33 1.67 -2.048 0.985 0.230
126
Table A.16. Pairwise comparisons between step widths for sagittal plane thorax-only coordination (ANOVA F=9.349,
p<0.001).
*Significant at the 0.05 level
Table A.17. Pairwise comparisons between step widths for sagittal plane thorax-only coordination (ANOVA
F=2.591, p=0.043).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -0.668 1.182 0.980
1 -3.180 1.182 0.055
1.33 -4.128* 1.182 0.002
1.67 -6.123* 1.182 <0.001
0.67
1 -2.513 1.182 0.209
1.33 -3.723* 1.182 0.014
1.67 -5.460* 1.182 <0.001
1
1.33 -1.210 1.182 0.844
1.67 -2.948 1.182 0.092
1.33 1.67 -1.738 1.182 0.582
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 0.095 0.884 1.000
1 1.043 0.884 0.763
1.33 1.588 0.884 0.376
1.67 2.373 0.884 0.056
0.67
1 0.948 0.884 0.821
1.33 1.493 0.884 0.441
1.67 2.278 0.884 0.075
1
1.33 0.545 0.884 0.973
1.67 1.330 0.884 0.559
1.33 1.67 0.785 0.884 0.901
127
Table A.18. Pairwise comparisons between step widths for fight peak longissimus activation (ANOVA F=9.049,
p<0.001).
*Significant at the 0.05 level
Table A.19. Pairwise comparisons between step widths for left peak longissimus activation (ANOVA F=9.318,
p<0.001).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -14.873* 5.012 0.025
1 -23.601* 5.012 <0.001
1.33 -23.624* 5.012 <0.001
1.67 -25.673* 5.012 <0.001
0.67
1 -8.728 5.012 0.408
1.33 -8.752 5.012 0.406
1.67 -10.800 5.012 0.197
1
1.33 -0.024 5.012 1.000
1.67 -2.072 5.012 0.994
1.33 1.67 -2.048 5.012 0.994
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -11.759 6.511 0.370
1 -27.218* 6.511 <0.001
1.33 -25.909* 6.511 <0.001
1.67 -35.330* 6.511 <0.001
0.67
1 -15.458 6.511 0.122
1.33 -14.149 6.511 0.190
1.67 -23.570* 6.511 0.003
1
1.33 1.309 6.511 1.000
1.67 -8.112 6.511 0.724
1.33 1.67 -9.421 6.511 0.597
128
Table A.20. Pairwise comparisons between step widths for bilateral longissimus co-activation pairwise comparison
(ANOVA F=8.491, p<0.001).
*Significant at the 0.05 level
Step Width (I) Step Width (J) Mean difference (J-I) Std. Error P-value
0.33
0.67 -0.011 0.014 0.945
1 -0.025 0.014 0.413
1.33 -0.049* 0.014 0.005
1.67 -0.073* 0.014 <0.001
0.67
1 -0.014 0.014 0.878
1.33 -0.038 0.014 0.063
1.67 -0.062* 0.014 <0.001
1
1.33 -0.024 0.014 0.436
1.67 -0.048* 0.014 0.007
1.33 1.67 -0.024 0.014 0.453
129
APPENDIX B
Chapter IV Step Width Effects Statistical Results
CTRL and rLBP-R
p-
value
0.262
0.026
0.675
--
--
--
--
<0.001
95%
CI
[-0.80,
0.22]
[-4.29,
-0.29]
[-2.08,
1.35]
--
--
--
--
[13.60,
31.09]
Esti-
mate
-0.29
-2.29
-0.37
--
--
--
--
22.34
CTRL and rLBP-A
p-value
0.139
0.037
0.526
--
--
--
--
<0.001
95%
CI
[-0.96,
0.13]
[-4.39,
-0.16]
[-2.20,
1.12]
--
--
--
--
[8.26,
27.77]
Esti-
mate
-0.41
-2.27
-0.54
--
--
--
--
18.02
CTRL and rLBP-pooled
p-
value
--
--
--
0.0289
0.0907
<0.001
<0.001
--
95% CI
--
--
--
[0.48,
8.31]
[-3.53,
0.25]
[-29.21,
-12.50]
[-0.09,
-0.05]
--
Esti-
mate
--
--
--
4.4
-1.64
-
20.86
-0.07
--
rLBP-A and rLBP-R
p-
value
0.239
0.012
0.435
0.012
0.016
0.015
<0.001
<0.001
95% CI
[-0.56,
0.14]
[-4.36,
-0.57]
[-2.75,
1.18]
[1.11,
8.74]
[-2.95,
-0.32]
[-27.61,
-3.18]
[-0.11,
-0.05]
[8.62,
32.55]
Esti-
mate
-0.21
-2.47
-0.78
4.92
-1.64
-
15.39
20.59
-0.08
Trunk
Excursion
In-phase
Anti-phase
Pelvis-only
Thorax-only
Peak
Longissimus
Activation
Bilateral
Longissimus
Co-
activation
Peak
Gluteus
Medius
Activation
Table B.1. Step width effects statistical results from general mixed-effects models. CTRL: Back-healthy controls; rLBP-A: recurrent low
back pain group in active pain; rLBP-R: recurrent low back pain group in remission; rLBP-pooled: pooled data across two testings for
recurrent low back pain group (only when there was no pain effect between rLBP-A and rLBP-R).
130
APPENDIX C
Chapter IV Stability of Performance in Control Participants
Figure C.1. No differences in the test-retest of 6 control participants for trunk excursion, and trunk in-phase and
anti-phase coordination, indicating no evidence of an effect of previous exposure to the experimental task.
131
APPENDIX D
Chapter V Practice Effect Analyses
Figure D.1. No differences in the comparisons between test-retest of the control group (CTRL) for arithmetic and
step width (SW) single task performance and dual-task performance variability, indicating no evidence of practice
effect.
132
Figure D.2. Correlation between days to retest and the changes in arithmetic and step width (SW) single task
performance and dual-task performance variability between the in and out of pain testing sessions in individuals
with recurrent LBP. There were no relationships for any of those variables, indicating no evidence of improved
performance associated with close testing dates.
133
APPENDIX E
Pressure Pain Threshold Testing
Methods
To assess the participants’ potential central sensitization of pain, three trials of pressure
pain threshold measurements were performed on the dominant side of the upper trapezius and
tibialis anterior muscles using the MicroFET2 handheld dynamometer (Hoggan Health, Salt Lake
City, UT).
Results
Pressure pain threshold was not statistically different between individuals with rLBP-A
and rLBP-R, and it was also not different between pooled rLBP data and back-healthy controls
(Table E.1).
Table E.1. Mean ± standard deviation values of pressure pain threshold in the back-healthy control group (CTRL),
the recurrent low back pain group during a painful episode (rLBP-A), and recurrent low back pain group during
symptom remission (rLBP-R).
CTRL rLBP-A rLBP-R
p-value:
rLBP-A vs
rLBP-R
p-value:
rLBP pooled vs.
CTRL
Trapezius (N)
49.80 ±
20.07
49.23 ±
25.25
58.08 ±
41.05
0.195 0.631
Tibialis
Anterior (N)
74.69 ±
28.70
75.24 ±
31.96
76.28 ±
36.24
0.826 0.902
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Trunk control during dynamic balance: effects of cognitive dual-task interference and a history of recurrent low back pain
PDF
Trunk neuromechanics during turning: a window into recurrent low back pain
PDF
The structure of kinematic variability during hopping: a window into motor control factors in overuse injury
PDF
Hemisphere-specific deficits in the control of bimanual movements after stroke
PDF
Multifidus morphology, fatigability and activation in persons with chronic unilateral low back pain
PDF
Development of postural strategies for the control of forward momentum
PDF
Hip and pelvis kinematics and kinetics in persons with femoroacetabular impingement
PDF
The influence of hip muscle performance on postural stability & ankle biomechanics: implications for ankle injury
PDF
The neuromechanical adaptations to Achilles tendinosis
PDF
Central nervous system generation of flexor bias in the control of leg movements during locomotor-related behavior in chick embryos
PDF
Pathomechanics of femoroacetabular impingement syndrome: utilizing subject-specific modeling approaches to investigate the influence of hip joint morphology and neuromuscular control
PDF
Investigating the role of muscle physiology and spinal circuitry in sensorimotor control
Asset Metadata
Creator
Shih, Hai-Jung Steffi
(author)
Core Title
The footprint of pain: investigating persistence of altered trunk control in recurrent low back pain
School
School of Dentistry
Degree
Doctor of Philosophy
Degree Program
Biokinesiology
Publication Date
01/22/2021
Defense Date
05/04/2020
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
adaptation,attention,coordination,low back pain,muscle activity,OAI-PMH Harvest,step width,Walking
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Kulig, Kornelia (
committee chair
), Gordon, James (
committee member
), Kutch, Jason (
committee member
), Van Dillen, Linda (
committee member
), Winstein, Carolee J. (
committee member
)
Creator Email
haijungs@pt.usc.edu,steffijong@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-346735
Unique identifier
UC11665841
Identifier
etd-ShihHaiJun-8734.pdf (filename),usctheses-c89-346735 (legacy record id)
Legacy Identifier
etd-ShihHaiJun-8734.pdf
Dmrecord
346735
Document Type
Dissertation
Rights
Shih, Hai-Jung Steffi
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
adaptation
attention
coordination
low back pain
muscle activity
step width