the effect of backpack loading configuration and design ......table 2 mean (standard deviation),...
TRANSCRIPT
i
The effect of backpack loading configuration and design
features on postural stability, energy cost, comfort and
shoulder interface pressure
By
Samira Golriz
BSc, MSc in Physiotherapy
This thesis is presented for the degree of Doctor of
Philosophy, Murdoch University
2013
ii
I declare that this thesis is my own account of my research and contains as
its main content work which has not previously been submitted for a degree
at any tertiary education institution.
Samira Golriz
iii
Abstract
While backpacks are commonly used to transfer extra load, they are
sometimes associated with negative effects on the body. The broad aim of
this thesis was to study the effect of backpack loading configurations on
postural stability, physiological variables, perceived exertion, and interface
pressure and compare them to an unloaded condition. The specific aims of
this thesis were to: 1) systematically review the literature and identify relevant
deficits in knowledge; 2) assess the test-retest reliability and criterion validity
of a force plate which has unknown psychometric properties; 3) examine the
effect of carrying a 20% of body weight loaded backpack and load placement
in a backpack (high vs. low load placement) on postural stability,
physiological variables and backpack-shoulder interface pressure; 4) assess
the effect of hip belt use on postural stability; and 5) investigate the effect of
shoulder strap width on backpack-shoulder interface pressure. These aims
were investigated using force plates, a metabolic cart, pressure sensors, and
questionnaires. Our results indicated that carrying loaded backpacks
decreased postural stability, increased physiological variables and perceived
exertion as compared to an unloaded condition. While a hip belt did not
enhance postural stability, participants reported a perception of increased
stability and less exertion. Load placement did not influence postural stability,
physiological variables, perceived stability or exertion; however, participants
reported lower levels of local perceived exertion at the shoulders and the
upper back regions when the load was placed low in the backpack.
Conversely, low load placement resulted in higher shoulder interface
pressure as compared to high load placement. Shoulder strap width also
affected shoulder interface pressure with wide shoulder straps associated
with lower shoulder interface pressure. While we identified several aspects of
backpack configuration and loading characteristics and their effects on
postural stability, energy cost, interface pressure and perceived exertion,
additional study of backpack configuration and the resulting impact on
biomechanical and physiologic characteristics is required.
iv
Acknowledgements
I would like to express my deepest gratitude to my supervisors, Dr Bruce
Walker and Dr Jeffrey Hebert, for their invaluable understanding, insight,
guidance and support which contributed enormously to my success through
this long process. Without a doubt, they were there for me, even during the
hardest times of my study to be a light, not a judge.
I would also like to thank my external supervisor, Dr Bo Foreman, for his
advice and assistance throughout my work.
I would like to extend my gratitude to Dr Jeremiah Peiffer for his help with a
section of this project.
Thanks must also go to the staff of School of Chiropractic and Sports
Science at Murdoch University.
I would like to offer my sincere appreciation to my parents, Simin and
Hashem, and my sister, Naghmeh, for their constant support and
encouragement in my life.
Last but not least, I would like to thank my best friend and husband, Navid,
whose love, support, encouragement and patience contributed immensely to
the completion of my PhD study, and my son, Kian, who has brought deeper
meanings and purposes to my life with his recent arrival.
v
TABLE OF CONTENTS
Abstract ……………………………………………….………. iii
Acknowledgements…………………………..…….………... iν
List of Tables ………………………………………………... ix
List of Figures ……………………………………………..... xii
Statement of Candidate Contribution………………………. xiv
CHAPTER ONE- Introduction…………………………………………
1
Background …………………………………………………………...…. 2
The broad aim of the thesis…………………………………………… 3
Specific aims and hypotheses of this thesis……………………….... 3
Overview of the thesis…………………………………………………... 5
References………………………………………………………...……... 6
CHAPTER TWO- Backpacks. Several factors likely to influence
design and usage: A systematic literature review………………..
8
Abstract …………………………………………………………...……… 11
Introduction ………………………………………………………………. 12
Methods …………………………………………...……………...……… 13
Results ………………………………….………………………………... 16
Discussion …………………………………...…………………………... 28
References………………………………………………………...……... 32
CHAPTER THREE- Can load carriage system weight, design
and placement affect pain and discomfort? A systematic
review……………………………………………………………………..
38
vi
Abstract ………………………………………………………………...… 41
Introduction ………………………………………………………………. 42
Methods …………………………………………...…………………...… 42
Results ………………………………….………………………………... 45
Discussion …………………………………...…………………………... 57
Conclusion ……………………………………………………………….. 59
References…………………………………………………………...…... 60
CHAPTER FOUR- The reliability of a portable clinical force
plate used for the assessment of static postural control:
repeated measures reliability study…………………………………
66
Abstract …………………………………………………………...……… 69
Introduction ………………………………………………………………. 70
Methods …………………………………………...………………...…… 71
Results ………………………………….………………………………... 75
Discussion …………………………………...…………………………... 77
Conclusion ………………………………………………………..……... 79
References…………………………………………………………...…... 79
CHAPTER FIVE- The validity of a portable clinical force plate
in assessment of static postural control: concurrent validity
study………………………………………………………………………
82
Abstract …………………………………………………………...……… 85
Introduction ………………………………………………………………. 86
Material and methods…………………...………………………………. 87
Results ………………………………….………………………………... 91
Discussion …………………………………...…………………………... 94
References…………………………………………………………...…... 97
vii
CHAPTER SIX- The effect of hip belt use and load placement
in a backpack on postural stability and perceived exertion: a
within subjects trial…………………………………………………….
99
Abstract …………………………………………………………………... 102
Introduction ………………………………………………………………. 103
Methods ……………………………...………………………………...… 105
Results ………………………………….………………………………... 109
Discussion …………………………………...…………………………... 114
References……………………………………………………………...... 117
CHAPTER SEVEN- The effect of backpack load placement on
physiological and self-reported measures of exertion…………..
121
Abstract …………………………………………………………...……… 124
Introduction ………………………………………………………………. 126
Methods …………………………………………...………………...…… 127
Results ………………………………….………………………….......... 131
Discussion …………………………………...…………………….......... 133
References…………………………………………………………......... 136
CHAPTER EIGHT- The effect of shoulder strap width and load
placement on shoulder–backpack interface pressure…………..
140
Abstract …………………………………………………………………... 143
Introduction ………………………………………………………………. 144
Methods …………………………………………...……………………... 145
Results ………………………………….………………………….......... 148
Discussion …………………………………...…………………….......... 151
References……………………………………………………………….. 154
viii
CHAPTER NINE- Conclusion………………………………………... 157
APPENDICES………………………………………………………….... 161
ix
LIST OF TABLES
Chapter 2 Page
Table1 NHMRC level of evidence ………………………….………… 15
Table 2 Descriptive characteristics of studies that evaluated the best
placement of a backpack on the spine………………………..
17
Table 3 Quality appraisal of studies that evaluated the best
placement of a backpack on the spine………………….…….
20
Table 4 Descriptive characteristics of studies which investigated
front pack or double pack vs. backpack………………………
22
Table 5 Critical appraisal of studies which investigated front pack or
double pack vs. backpack…………..………………………….
24
Table 6 Descriptive characteristics of studies that investigated
shoulder straps……………………………………………………
25
Table 7 Quality appraisal of studies that investigated shoulder straps 27
Table 8 Recommendations for best backpack.………………………… 29
Chapter 3
Table 1 NHMRC level of evidence ……………………………………… 44
Table 2 Descriptive characteristics of the studies that assessed the
correlation between backpack use and pain, perceived
exertion and discomfort……………………………..…………
47
Table 3 Results of the studies that assessed the correlation between
backpack use and pain …………………………...…………..
49
Table 4 Results of the studies that assessed the correlation between
backpack use and perceived exertion………………………..
50
Table 5 Quality appraisal of studies that assessed the correlation
between backpack use and pain………………………………
51
Table 6 Descriptive characteristics of studies that examined the
x
effect of load placement on pain……………………………… 53
Table 7 Quality appraisal of studies that examined the effect of load
placement on pain………………………………………………
54
Table 8 Descriptive characteristics of studies that examined if
different designs can reduce pain, perceived exertion and
discomfort………………………………………………………..
54
Table 9 Quality appraisal of studies that examined if different
designs can reduce the pain……………………………….......
55
Chapter 4
Table 1 Reliability results of single measures and means of 2, 3, 4
and 5 measures for each variable………………………………
75
Chapter 5
Table 1 Criterion validity analysis……………………………………….. 92
Chapter 6
Table 1 Demographic and Baseline Characteristics of Participants… 110
Table 2 Mean (standard deviation) and mean difference (95% CI)
values during unloaded, hip belt and no hip belt condition….
110
Table 3 Mean (standard deviation) and mean difference (95% CI)
values during unloaded, high load placement and low load
placement………………………………………………………
111
Chapter 7
Table 1 Demographic and Baseline Characteristics of Participants… 131
Table 2 Mean (standard deviation), mean difference (95% CI) values
during unloaded, high and low load placement conditions…
132
Chapter 8
Table 1 Mean (SD) of peak and average backpack–shoulder
interface pressure under different width shoulder straps
(kPa) and load placements………………………………..….
148
Table 2 Mean (SD) of peak and average backpack–axilla interface
pressure under different width shoulder straps (kPa) and
xi
load placements………………………………………………… 149
xii
LIST OF FIGURES
Chapter 3 Page
Figure 1 Inclusion and exclusion of articles…………………………….… 46
Chapter 4
Figure 1 The Midot Posture Scale Analyser………………………………. 72
Chapter 5
Figure 1 The Midot Posture Scale Analyzer and The Accugait AMTI 88
Figure 2 Differences between filtered and unfiltered average velocity
vs. cut off frequency to decide the best filter frequency………
89
Figure 3 Bland-Altman plot representing comparison of average
velocity between the MPSA and the Accugait…………………..
93
Figure 4 Bland-Altman plot representing comparison of sway area
between the MPSA and the Accugait………………..………….
93
Chapter 6
Figure 1 The Promopak backpack…………………………………………. 108
Figure 2 The loading conditions in the backpack…………………………. 108
Figure 3 Perceived exertion of neck, shoulders, upper back, lower
back, lower extremity in unloaded, hip belt and no hip belt
conditions…………………………………………………………..
113
Figure 4 Perceived exertion of neck, shoulders, upper back, lower back
and lower extremity in unloaded, high and low load
placements................................................................................
130
Chapter 7
Figure 1 The loading conditions in the backpack………………………… 152
Chapter 8
Figure 1 The manikin and position of the pressure sensors…………...... 145
Figure 2 The SAS harness backpack……………………………………… 146
Figure 3 Average backpack-shoulder interface pressure under various 150
xiii
shoulder strap widths in high and low load placement…………
Figure 4 Average backpack-axillary area interface pressure under
various shoulder strap widths in high and low load placement..
151
xiv
Statement of Candidate Contribution
The work involved in designing and conducting the studies described in this
thesis has been conducted principally by Samira Golriz (the candidate). The
thesis, outline and experimental design of the studies was developed and
planned by the candidate, in consultation with her supervisors Associate
Professor Bruce Walker, Dr Jeffrey Hebert, Dr Bo Foreman. All participant
recruitment was carried out by the candidate, along with the implementation
and undertaking of data collections.
The candidate was responsible for all data analysis and original drafting of all
chapters contained within this thesis. The supervisors have provided
feedback on all drafts of chapters contained within this thesis.
1
Chapter One
INTRODUCTION
2
Background
The carrying of heavy loads is often a necessity and several means are
available for transferring the extra load on the body from one place to
another. Personal load carriage systems, typically referred to as backpacks,
are used for moving extra loads. Contemporary backpacks, as we know them
today, date back to the 1920s. They were invented by Lloyd Nelson, who
originated the idea during a trip to Alaska when he borrowed a pack from a
local Indian (Widrig 2004). Before the 1920s, the basic design of a backpack
was a canvas bag, with shoulder straps attached, which needed careful
packing to prevent the contents from digging into the wearer’s back (Widrig
2004). Nowadays, backpacks are frequently used in different environments
by different groups such as the military (Knapik, Harman, and Reynolds
1996; Knapik, Reynolds, and Harman 2004), students (Korovessis, Koureas,
and Papazisis 2004; Pascoe et al. 1997; Sheir-Neiss et al. 2003) and hikers
(Lloyd and Cooke 2000 a, 2000 b).
The drawback of carrying backpacks is that they may have detrimental
effects on the body (Knapik, Reynolds, and Harman 2004; Knapik, Harman,
and Reynolds 1996; Boulware 2003).
The effects that backpacks can potentially have on the body are numerous
(Talbott 2005). Ranging from no noticeable change in posture, postural
stability, energy cost and pain to postural instability, pain, perceived exertion,
increased oxygen consumption and energy cost. There is growing concern
regarding both immediate and more chronic health effects due to backpack
use (Wiersema, Wall, and Foad 2003). While studies into the chronic health
effects would appear to require longitudinal tracking of backpack wear,
assessment of the short term changes is more feasible and can represent the
presence of acute alterations that should be considered when establishing
guidelines for backpack use.
Answers to questions regarding the effects of backpack use on postural
stability, physiological measures and perceived exertion will assist in
identifying a potential link between backpack use and the occurrence of pain
and discomfort, physiological measures, postural stability and potential for
3
injury from falls. In addition, discovering if modifications of a backpack
reduces the incidence or alleviates the negative effects on the body will
assist in improving backpack design. Backpacks need to be designed with an
understanding of the general principles and issues related to human anatomy
and biomechanics. Recognition of the detrimental effects that backpacks may
have on the body and then finding a way to moderate those effects is the
ideal pathway for the development of safe load carriage systems.
The broad aim of the thesis
The broad aim of this thesis was to investigate the optimum load placement
in a backpack regarding postural stability, physiological measures, perceived
exertion and shoulder interface pressure; and to examine the effect of
shoulder strap width on shoulder interface pressure in order to modify and
improve backpack design. These factors might be affected by backpack use
and load placement within a backpack.
The backpacks used in this study were everyday use backpacks and sample
population was recruited among adults who were occasional backpack users.
To achieve this aim, the thesis was completed in three steps. In the first step,
two systematic reviews of the literature were conducted to systematically
review the literature that investigated backpacks and their effects on the body
and to find gaps in the literature and the areas that needed further
investigation. In the second step, two methodology studies were conducted
to assess the test re-test reliability and criterion validity of an instrument that
was going to be used in the following experiment. In the third step, based on
the results of the systematic reviews, areas that needed further investigation
were chosen and three studies were conducted. The three studies examined
the effect of different load placements in a backpack on postural stability,
physiological measures, perceived exertion and interface pressure and
compared it with an unloaded control condition in order to identify the
optimum load placement in a backpack and backpack shoulder strap width.
Specific aims and hypotheses of this thesis
The six main aims and hypotheses of this thesis are as follows:
4
To provide a systematic review of the literature to compare different
backpack designs and their effects on the body (chapter 2).
To provide a systematic review of the literature to investigate if
backpack load, design and load placement affects pain, discomfort
and perceived exertion in the backpack carrier (chapter 3).
To assess the reliability and validity of a force plate that we planned to
use for measurement of postural stability in chapter 6 (chapter 4 and
5).
To assess the effect of load placement in a backpack and backpack
hip belt use on perceived exertion and subjective and objective
measures of postural stability and compare them with an unloaded
condition (chapter 6).
We hypothesized that lower levels of postural stability and higher levels of
perceived exertion would occur for loaded conditions result than for an
unloaded condition. We also hypothesized that hip belt use would provide
physical constraint between the pelvis and thorax. Furthermore, a hip belt
would facilitate the transfer of a substantial amount of load from the
shoulders to the pelvis, possibly allowing the trunk muscles to handle the
imposed load better and focus more on postural stability. This could enhance
postural stability and reduce perceived exertion.
The body is considered as an inverted pendulum (Winter et al. 1993) and we
hypothesized that high load placement would elevate the centre of gravity
(COG) of the body and tend to destabilize the body when compared to low
load placement. The result would be lower postural stability and higher levels
of perceived exertion.
To look for differences in physiological (oxygen uptake, minute
ventilation and heart rate) and self reported measures of exertion,
movement economy and efficiency when carrying a loaded backpack
during both a high and low load placement, compared to an unloaded
control condition (chapter 7).
We hypothesized that both loaded conditions (regardless of the type of load
placement) would lead to higher heart rate, oxygen uptake, minute
5
ventilation, perceived exertion and lower movement economy and efficiency
when compared to the unloaded condition. We also hypothesized that
elevation of the COG in high load placement would cause greater postural
instability compared to low load placement as a result of higher rotational
inertia of high load placement. Therefore, the trunk would tend to lean
forward to a higher degree to bring the body’s COG back to its position and
closer to the base of support. Higher degrees of trunk forward lean and
displacement of the COG would result in higher physiological variables and
lower economy and efficiency of movement.
To evaluate the effect of the shoulder strap width and load placement
in a backpack on the shoulder–backpack and axillary area–backpack
interface pressures (chapter 8).
We hypothesized that wider shoulder straps would distribute the load over a
wider area and thereby reduce interface pressures. We also hypothesized
that higher interface pressure would be generated by high load compared to
low load placement.
Overview of the thesis
This thesis is composed of nine chapters. Following this introduction are two
chapters that provide systematic reviews of the literature. These reviews
identified areas that needed further investigation. The fourth and fifth
chapters describe the investigation of the reliability and validity of a clinical
force plate with unknown psychometric properties which was needed in
chapter six to evaluate postural stability of the body. The sixth chapter
evaluates the effect of hip belt use and load placement in a backpack on
perceived exertion and subjective and objective measures of postural
stability. If hip belt use enhances postural stability or increases the perception
of postural stability, it shows the necessity that backpacks should be
accompanied by a hip belt. If load placement in a backpack affects postural
stability, it indicates that load placement is a factor that needs to be
considered in packing a backpack, i.e. it is the best to pack heavy items at
top or bottom of a backpack. It might also be helpful in designing a backpack,
i.e. to modify the internal compartment of a backpack by embedding a
6
chamber for heavy items at top or bottom of a backpack. The seventh
chapter investigates the effect of load placement on physiological measures,
perceived exertion and the economy and efficiency of movement. If load
placement improves economy and efficiency of movement and physiological
measures and decreases perceived exertion, it indicates the importance of
load placement consideration in packing and/or designing a backpack. The
eighth chapter assesses the effect of shoulder strap width and load
placement on shoulder–backpack interface pressure. The optimum width of
shoulder straps and load placement would alleviate the amount of pressure
exerted on the shoulders and make backpack carrying more comfortable.
Shoulder strap width as well as load placement might be the factors to be
considered in improving backpack design. The ninth chapter provides an
overarching conclusion of the thesis findings.
References
Boulware, D. R. 2003. Backpacking-induced paresthesias. Wilderness and
Environmental Medicine 14 (3):161-166.
Knapik, J., E. Harman, and K. Reynolds. 1996. Load carriage using packs: A
review of physiological, biomechanical and medical aspects. Applied
Ergonomics 27 (3):207-216.
Knapik, J. J., K. L. Reynolds, and E. Harman. 2004. Soldier Load Carriage:
Historical, Physiological, Biomechanical, and Medical Aspects. Military
Medicine 169 (1):45-56.
Korovessis, P., G. Koureas, and Z. Papazisis. 2004. Correlation between
Backpack Weight and Way of Carrying, Sagittal and Frontal Spinal
Curvatures, Athletic Activity, and Dorsal and Low Back Pain in
Schoolchildren and Adolescents. Journal of Spinal Disorders and Techniques
17 (1):33-40.
Lloyd, R., and C. B. Cooke. 2000 a. Kinetic changes associated with load
carriage using two rucksack designs. Ergonomics 43 (9):1331-1341.
Lloyd, R., and C. B. Cooke. 2000 b. The oxygen consumption associated
with unloaded walking and load carriage using two different backpack
designs. European Journal of Applied Physiology 81 (6):486-492.
Pascoe, D. D., D. E. Pascoe, Y. T. Wang, D. M. Shim, and C. K. Kim. 1997.
Influence of carrying book bags on gait cycle and posture of youths.
Ergonomics 40 (6):631-641.
7
Sheir-Neiss, G. I., R. W. Kruse, T. Rahman, L. P. Jacobson, and J. A. Pelli.
2003. The association of backpack use and back pain in adolescents. Spine
28 (9):922-930.
Talbott, N. R. 2005. the effect of weight, location and type of backpack on
posture and postural stability of children. PhD thesis, University of Cincinnati.
Widrig, C. D. A brief of how backpacking got it's name 2004. Available from
http://www.url.biz/Articles/Article-1217.html.
Wiersema, B. M., E. J. Wall, and S. L. Foad. 2003. Acute backpack injuries in
children. Pediatrics 111 (1):163-166.
Winter, D.A., C. D. Mackinnon, G. K. Ruder, and C. Wieman. 1993. An
integrated EMG/biomechanical model of upper body balance and posture
during human gait. Progress in brain research 97:359-367.
8
CHAPTER TWO
BACKPACKS. SEVERAL FACTORS LIKELY TO INFLUENCE
DESIGN AND USAGE: A SYSTEMATIC LITERATURE REVIEW.
Published as
Golriz S, Walker B. (2012). Work, 42(4):519-31.
9
In this chapter, the results of a systematic review “Backpacks. Several factors
likely to influence design and usage” are presented. This systematic review
was conducted to answer three questions: 1) what is the optimum backpack
positioning on the spine/optimum load placement in a backpack; 2) what are
the human effects of front packs and double packs compared to backpacks;
and 3) what is best shoulder strap design.
The work resulted in the following publication:
Golriz S, Walker B. Backpacks. Several factors likely to influence design and
usage: A systematic literature review. Work. 2012;42(4):519-31.
10
Samira Golriz1§MSPT, Bruce Walker2 DC, MPH, DrPH
1PhD candidate, School of Chiropractic and Sports Science, Murdoch
University, Murdoch 6150, Western Australia, Australia
2Senior Lecturer, School of Chiropractic and Sports Science, Murdoch
University, Murdoch 6150, Western Australia, Australia
§Corresponding author
Email addresses:
SG: [email protected] . Tel: + 61 8 9360 1450, Fax: + 61 8 9360
1299
11
Abstract.
Objective: The purpose of this study was to systematically review the
literature to answer three questions: 1) what is the best backpack positioning
on the spine/ what is the best load placement in a backpack; 2) what are the
human effects of front packs and double packs compared to backpacks; and
3) what is best shoulder strap design. Methods: A systematic review of the
literature using eight databases was carried out. Studies relevant to
backpack design were retrieved. Two independent reviewers assessed the
papers; a third party was used for consensus decisions. Descriptive
characteristics, type of research design and level of evidence of papers were
evaluated with a view to pooling data. The trials were also quality appraised
using a modified Crombie tool. Results: Thirty papers met the inclusion
criteria. There were similarities in methods of measurement between some
papers but subject’s age group, tasks performed and backpack usages were
so different between studies that it prevented data pooling and made it
difficult to draw firm generic conclusions. Subsequent qualitative analysis
shows that there are conflicting results on best backpack placement and
shoulder strap design but front packs and double packs provide better
posture than backpacks. Conclusions: Some recommendations for best
practice design are made for children and adults based on elements of
design and correct spinal placement.
Keywords: load carriage system, placement, straps
12
Introduction
Using a backpack is a popular mode of load carriage and it is considered the
most appropriate way of carrying additional weight [4; 48]. Backpacks are
widely used for different purposes, especially by students of different age
groups [15; 32; 35; 37; 42], military personnel [16; 24; 29], leisure and
recreational backpackers [28], NASA employees [48], rescue workers [40]
and even agricultural workers [18]. However, backpacks may have adverse
effects on the body including low back pain.
Adult low back pain is a significant source of long term dysfunction and
absence from work and this put a huge economic, social and emotional
burden on individuals and society [55; 56]. Additionally, back pain is a rising
issue among young people with low back pain prevalence in adolescents
measured between 20 to 50% [25; 38; 42; 53; 54]. While backpacks are an
effective way to carry weight, it has been proposed they can also be a
significant contributing risk factor for discomfort, fatigue, muscle soreness
and musculoskeletal pain especially low back and thoracic spine pain [6; 25;
45]. It has been speculated that backpacks may cause problems not only for
the developing skeletal system but also for a mature spine as a developed
spine is also sensitive to load [41]. Moreover, experiencing back pain in
childhood is a concern as it may lead to more common and severe issues
later in life [5; 33]. Complications from carrying a backpack is not limited to
spinal pain; other health issues may occur such as respiratory problems [8;
27], brachial plexus lesion [3], winged scapula [26], foot blisters [24].
In the literature backpack usage complications have been correlated with
load amount [59], duration of load carrying [59], distance walked [24],
inadequate distribution of weight [37], poor item placement in the backpack
[51] and poor positioning of the backpack on the spine [15] which may
provoke postural changes, producing pain and discomfort. Some studies
have tried to moderate these factors by applying changes to the backpack
design [37; 44] resulting in changes to primary designs and the evolution of
specialised designs for different users.
13
The objective of this study was to systematically review the literature to
answer three questions: 1) what is the best backpack positioning on the
spine/ what is the best load placement in a backpack; 2) what are the human
effects of front packs and double packs compared to backpacks; and 3) what
is best shoulder strap design. Our aim was to use the answers to these
questions to advise on reducing any complications of backpack usage and
finally to add to the literature on best practice in backpack design.
There has been no other systematic review looking at these questions.
Methods
Literature search
A comprehensive search strategy was conducted to identify all relevant
publications on load carriage systems and their design. The search strategy
is seen below.
Allied health, health-research, health-science and medical databases
including Medline, Cochrane library, Science Direct, PubMed, Scopus,
CINAHL, MANTIS and EMBASE were used. The search was performed
using the following key indexing terms independently: ‘backpack’, ‘back
pack’, ‘rucksack’, ‘schoolbag’, ‘school bag’, ‘load carriage system’, ‘front
pack’ and ‘double pack’. Google searches and a hand search of the
reference lists of existing articles were also conducted to find papers that did
not appear in the main database searches. The search covered literature
from 1966 to January 2010.
Selection criteria
Studies that investigated new features and designs of backpack or papers
that assessed factors that can potentially alter the design of backpacks were
included. The inclusion criteria were as follows:
Best backpack placement on the spine/best load placement in a backpack,
Design of shoulder straps,
Front pack and double pack usage.
14
Studies were limited to peer-reviewed journals and conference proceedings.
In the case of research on participants, only studies that investigated healthy
populations were included. All study abstracts meeting these broad criteria
were initially included. In the case that decision could not be made based on
the title and abstract of the paper, the full text of the paper was included for
further decision. Subsequent inclusion was based on inclusion criteria
assessed by two trained reviewers (SG and BW) who reviewed the papers. If
a difference of opinion occurred, consensus was reached on inclusion or
exclusion by discussion and reflection.
Data extraction and management
SG acted as the principal reviewer. A research assistant, SB who was trained
by SG and BW acted as a second reviewer to extract data from the included
papers. Training sessions for SB included clarification of all data items and
the required elements of critical appraisal were provided. Standardisation of
the procedure was required to provide consistency in methods used by the
reviewers. Before starting to extract data, a trial was conducted on two
similar but unrelated papers and the results discussed. Co-investigator (BW)
was consulted when there was disagreement between SG and SB. BW’s
opinion stimulated further discussion and allowed consensus to occur. This
data extraction method (double data extraction) has been shown to have a
lower rate of error than single data extraction [7]. Pooling of data would be
undertaken where adequate homogeneity of results exists.
Level of evidence
Level of evidence of each paper was assessed based on the National Health
and Medical Research Council of Australia guidelines (Table 1). This is
based on the proposition that some designs provide more valid and reliable
findings than others. The lower the ranking in hierarchy of evidence the
greater the risk of bias or error in a study [1].
15
Table 1: NHMRC level of evidence for intervention studies
Level of evidence
Study design
I A Systematic review of randomised controlled trials
II A randomised controlled trial
III-1 A pseudo randomised controlled trial
III-2 A comparative study with concurrent controls: Non-randomised experimental trial Cohort study Case-control study Interrupted time series with a control group
III-3 A comparative study without concurrent controls: Historical control study Two or more single arm study Interrupted time series without a parallel control group
IV Case series with either post-test or pre-test/post-test outcomes
Quality appraisal
The quality of papers was assessed according to a modified version of the
quality appraisal tool by Crombie [11]. In this study, we modified the Crombie
tool by adding three extra appraisal items, ‘attention to calibration of
equipment/ instrument before use, ‘Was the person who carried out the
measurement trained?’ and ‘Discussion of weaknesses or limitations in the
paper’. Moreover, validity and reliability of measurements which were fitted
into one item by Crombie, were split up into two questions as these two
concepts demonstrate two different aspects in research. Modified Crombie
quality appraisal items can be seen in table 3.
Answers to the quality appraisal items were defined as Yes, No, Not
Applicable or Unclear. These answers were used instead of a scoring
system. Often, reviewers combine the scores of individual items from the
critical appraisal tool to present a total score [50]. However, using this
method may be arbitrary as is weighting each item. Moreover, it has been
recommended to investigate each item separately, rather than use a
combined quality score [20; 58]. However, given the dichotomy of views in
the literature we chose to simply classify studies with the notation of how
many critical appraisal items they satisfied. In this way we believe an
16
estimation of the quality of the study can be gained, with the results of
studies that meet less appraisal criteria being treated with more caution.
Results
Literature search
Five hundred and ten articles were identified from the databases using the
search strategy. Titles and abstracts of these articles were manually
screened for relevance and 334 articles were excluded. The remaining
articles (n=176) were studied in detail to see if they satisfied the inclusion
criteria. A further 150 papers were excluded. Six articles were added after
reference checking. Two articles were eliminated as they were presented at a
conference and their full texts were not available. Thirty articles were
included for the final review.
Study results
Level of evidence, descriptive characteristics of participants, equipment and
instruments, variables measured, task, what element of the backpack was
studied and results are reported in table 2.
The included studies used a mixed mode of data collection including
questionnaires, scales and examinations. All of the studies were conducted
in developed countries. In 28/30 studies, experiments were conducted on
subjects and 2/30 studies carried out their experiment on a mannequin [36;
43]. In 13 studies, the age ranges selected were from adult populations and
in 15 studies adolescents and children were examined. In just 12 studies,
subjects were screened for entry into the experiment based on inclusion and
exclusion criteria.
Best positioning of a backpack on the spine/best load placement in a
backpack:
Twelve studies’ aimed to find out where the best positioning of a backpack on
the spine (upper back, middle back or lower back) should be. Placement of a
backpack is a factor that can also potentially alter the design of a backpack.
17
Descriptive characteristics and quality appraisal of these studies can be seen
in table 2 and 3.
Table 2. Descriptive characteristics of studies that evaluated the best
backpack placement on the spine/ best load placement in a backpack
Auth
or(
s),
year
Sam
ple
siz
e
Age
Inclu
sio
n a
nd e
xclu
sio
n c
rite
ria
Equ
ipm
ent
& instr
um
ent
of
measure
men
t
Vari
able
s m
easure
d
Task
Cle
arl
y s
tate
d a
ims
Result
Stuempf
le et al
2004
[52]
1
0
18-
22
N BS,
VO2/EC
G
exercise
system,
HR
monitor,
stadiom
eter
VO2,
respiratory
exchange ratio,
respiratory rate,
minute
ventilation, HR,
RPE
Walking on
treadmill with
positioning of
25% BW in
high, central or
low on the back
Y
Less energy &
oxygen
consumption
in high load
placement
Bobet
and
Norman
1984 [2]
1
1
19-
22
N EMG Electrical
activities of
muscles
Walking under
3 conditions:
WL, wearing
BP below mid-
back or above
shoulders
Y
High load
placement
caused higher
electrical
activity level
Macias
et al
2008
[32]
1
0
13.
2±
0.8
Y BS,
pressure
sensor
Contact
pressure
beneath
shoulder
straps, RPE
Standing and
walking WL or
with 10%, 20%
and 30% BW
on lower or
upper back with
one shoulder
strap and 2
shoulder strap
Y
Low back
position
induced more
perceived
exertion &
higher contact
pressure
18
Sing
and Koh
2009b
[47]
1
7
8-
12
N Camera
motion
analysis
corporati
on,
optical
motion
analysis
system,
Force
plate
Walking
velocity, pelvic
/trunk
transverse
plane rotation,
knee flexion
angle,
normalized first
peak force,
peak
acceleration of
ankle/ knee/
pelvis/ trunk
and head,
shock
transmission
ratio between
ankle-knee/
knee-pelvis/
pelvis-trunk/
trunk-head
/ankle-head
Walking on
treadmill either
WL or carrying
10, 15 and 20%
BW in a BP on
lower or upper
back
Y
Load
placement had
no effect on
gait & posture
Liu
2007
[29]
5 21-
28
N pulmona
ry
function
test
system,
Q
Respiratory
frequency,
minute
ventilation,
respiratory
exchange ratio,
VO2, HR
Walking on
treadmill
carrying 15%
BW at different
velocities &
grade levels at
high & low load
placement
Y More oxygen
consumption
in high load
position
Brackle
y et al
2009 [4]
1
5
N
M
(gr
ad
e
5)
Y Q, 2
dimensi
onal
video,
body
map
diagram
Trunk flexion
angle, cranio-
vertebral angle,
RPE
Standing
unloaded,
standing &
walking while
carrying 15%
BW in a BP
with the centre
of mass of the
pack located on
high, mid or low
back
Y Low load
placement
decreased
postural
adaptations &
spinal
curvatures
Sing
and Koh
2009a
[46]
1
7
9.6
5±
1.5
N Camera,
Force
plate,
motion
analysis
corporati
on
Velocity,
cadence and
stride length,
double support
time, trunk
forward angle
Walking on
treadmill WL or
carrying 10, 15
and 20% BW in
BP on high or
low back
Y Load
placement had
no difference
on gait &
posture
Chow et
al 2009
[10]
2
2
12
±0.
6
Y Cardiop
ulmonar
y
function
FVC, FEV1,
FEV1 /FVC
ratio, peak
expiratory flow,
Exhale during
upright stance
WL or with 15%
BW in a BP
Y
Load
placement had
no difference
on children’s
19
BP, Backpack ; BS, Borg Scale ; BW, Body Weight ; COG, Centre Of Gravity; EMG, Electromyography; FEV1, Forced Expiratory Volume in 1sec ; FVC, Forced Vital Capacity ;GRF, Ground Reaction Force; HR, Heart Rate; N, No; Q, Questionnaire; RPE, Rating of Perceived Exertion; U, Unclear; VO2, Oxygen Consumption; WL, Without Load; Y, Yes
machine forced
expiratory flow
positioned at
T7, T12 and L3
pulmonary
function
Grimme
r et al
2002
[15]
2
5
0
12-
18
Y Camera Changes in
horizontal
positions of
tragus of ear,
lateral iliac
crest of hip, C7,
greater
trochanter, mid
acromion of
shoulder and
fibula relative to
ankle joint
Standing WL or
wearing 3%,
5% and 10%
BW in a BP
centred at T7,
T12, L3
Y
Low
placement
produced less
postural
deviation
Devroey
et al
2007
[13]
2
0
23.
9±
2.5
N Force
plate,
BS,
camera,
EMG
Mean angle of
head/
neck/thorax/spi
ne/pelvis,
Single support/
double support/
stance/swing
duration, stride
time, walking
velocity,
electrical
activity of 7
muscles, HR
Standing and
walking WL or
carrying 5%,
10% and 15%
BW in BP on
thoracic or
lumbar
N
-walking:
lumbar
placement
caused more
postural
adaptations
-standing:
thoracic
placement
caused more
postural
changes
Kellis
and
Arampat
zi 2009
[21]
1
8
8 U Force
platform,
camera
Temporal and
GRF
parameters
Walking WL or
carrying 17%
BW in a BP
high or low on
the back
Y
High position
caused
increase in
vertical GRFs,
no difference
on gait
between 2
placements
Chow et
al 2010
[9]
1
9
11.
4±
0.5
U Accelero
meter
Cervical/ upper
& lower
thoracic/ upper
& lower lumbar
spinal
curvature
Standing under
7 conditions:
WL, wearing
BP anteriorly or
posteriorly
while COG of a
BP is located at
T7, T12 or L3
Y
Carrying BP
anteriorly with
placing COG
at T12 had
least effect on
spinal
deformation
20
Table 3. Quality appraisal of studies that evaluated the best backpack
placement on the spine/ best load placement in a backpack
Auth
or(
s),
year
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Tota
l score
Stu
dy d
esig
n &
level of
evid
ence
Stuempfle
et al 2004
N Y U
N
N N Y N
N N Y N Y Y Y
6/
15
III-
3A
Bobet and
Norman
1984
N Y U N N N N
U
Y Y Y N N Y N
5/
15
III-
2
Macias et
al 2008
N Y U
N N N N Y
N Y Y N N Y Y
6/
15
III-
2
Sing and
Koh
2009b
N N
U N
N N Y N
Y Y Y
N N Y
N
5/
15
III-
2A
Liu 2007 N Y U N N N Y Y N Y Y N Y Y Y
8/
15
III-
3A
Brackley
et al 2009
N Y
U N
Y Y Y Y
Y Y Y N N N Y
9//
15
III-
3A
Sing and
Koh
2009a
N Y U N N N Y N
N Y Y N N Y Y
6/
15
III-
2A
Chow et
al 2009
N Y U Y N N Y N N Y Y
N Y Y N
7/
15
III-
2A
Grimmer
et al 2002
Y
Y U N Y Y Y N Y Y Y N
Y Y
N
10
/1
5
III-
2A
Devroey
et al 2007
N
N
U
N
N
N
Y
Y
Y Y Y
N
Y
Y
Y 8/
15
III-
2A
Kellis &
Arampatzi
2009
N
Y U
N
N
N
Y
Y
Y
Y Y N
Y
Y Y
9/
15
III-
2A
Chow et
al 2010
N Y U Y N Y Y
Y N Y Y N Y Y Y 10
/1
5
III-
2A
1. Justification of sample size; 2. Consistency in the number of subjects reported throughout the paper; 3. The person who carried out the measurement was trained; 4. Was the equipment/instrument calibrated before use; 5. Adequate description of the validity of the
21
instrument/equipment; 6. Adequate description of the reliability of the instrument/equipment; 7. Appropriateness of design to meet the aims; 8. Weakness or limitations mentioned; 9. Interpretation of null findings; 10. Interpretations of important effects; 11. Comparison of results with previous reports; 12. Implication in real life/generalisability; 13. Adequate description of statistical methods; 14. Adequate description of the data; 15. Assessment of statistical significance; N, No; N/A, Not Applicable; U, Unclear; Y, Yes; *, These studies assessed factors that can potentially alter the design of backpack; §, This study has been conducted on a mannequin and not human subjects; III-2, a comparative study with concurrent control; III-2A, a comparative study with concurrent control (phases randomisation); III-3, a comparative study without concurrent control; III-3A, a comparative study without concurrent control (phases randomisation)
Tables 3,5 and 7 (quality appraisal) were completed with Yes, No, Unclear
and N/A. Score 1 was given to each yes answer and 0 to No, Unclear and
N/A. The overall score was reported as a tally of all yes answers out of 15, 14
or 13 based on the applicable answers for each study. In a study where two
or more pieces of equipment or instruments were used, even if details of
calibration, validity and reliability of one of them were provided it was
considered that this paper criterion was adequate.
Front pack and double pack
Nine studies looked at ways to reduce aberrant postural spinal changes, pain
and discomfort caused by carrying a backpack. These studies substituted the
backpack for a front pack or a combination of backpack and front pack to try
to keep the spine in a natural position or distribute the load along the spine.
Two of these studies compared a front pack vs. a backpack, six experiments
evaluated a backpack vs. a double pack and the objective of one study was
comparison between backpacks, front packs and double packs. Descriptive
characteristics and quality appraisal of these studies can be seen in tables 4
and 5.
22
Table 4. Descriptive characteristics of studies which investigated front
pack or double pack vs. backpack
Auth
or(
s),
year
Sam
ple
siz
e
Age o
f p
art
icip
an
ts
Inclu
sio
n &
exclu
sio
n
crite
ria
E
qu
ipm
ent
& I
nstr
um
ents
of
measure
men
t
Vari
able
s m
easure
d
Task
Wha
t e
lem
ent
of b
ackpack
was s
tudie
d/d
esig
ne
d
Cle
arl
y s
tate
d a
ims
Result
LIoy
d
and
Cook
e
2000
a
[30]
9 24.
7±
4.3
N Came
ra,
Force
plate
Support time,
GRFs
Walking on
force plate
under 3
conditions:
WL,
wearing BP
or DP
DP
vs.
BP
Y Less
propulsive
force while
carrying
DPs
LIoy
d
and
Cook
e
2000
b
[31]
9 24.
7±
4.3
U
Gas
analys
is
syste
m
VO2max, HR Walking on
treadmill
uphill,
downhill
WL or while
carrying
load
DP
vs.
BP
Y Less
energy
consumptio
n while
carrying
DPs
Wan
g et
al
2007
[57]
2
7
11
-
15
Y Electr
o
gonio
metric
syste
m
Tilting angle of
spine, upper/
lower thoracic
kyphosis, upper/
lower lumbar
lordosis,
cervical
lordosis,
thoracic
kyphosis,
lumbar lordosis
Standing
upright with
or without
wearing a
BP loaded
at 15% BW
either
anteriorly
or
posteriorly
FP
vs.
BP
Y FP had
less
influence
on sagittal
spinal
curvature
than BP
Knap
ik et
al
1997
[24]
1
5
29.
7±
4.3
Y HR
monit
or,
PRSD
Q
Road march
times, HR,
distance of
grenade
throwing, rating
of pain,
soreness and
discomfort, foot
blister
20 km
march
course
while
wearing
two kinds
of BP and
carrying 34,
48 and 61
kg load
Stan
dard
BP
vs.
DP
Y DPs
caused
less
discomfort
in low
back, lower
incidence
of blisters,
more
optimal
posture but
it resulted
in pain in
23
neck & hips
and it took
longer to
complete
the march
Mot
man
s et
al
2006
[39]
1
9
20.
12
±
2.0
3
N EMG Electrical
activity of rectus
abdominis and
erector spinae
Standing
with
extended
knees &
feet in
frontal
plane
wearing
15% BW in
shoulder
bag, BP,
FP or DP
com
pari
ng
BP,
FP,
shou
lder
bag
&
DP
Y DP had
less
influence
on muscle
activities
Kino
shita
1985
[23]
1
0
N
M
N Came
ra,
Force
plate,
photo
electri
c
timing
syste
m
Double/single
support time,
step length/
width, angle of
gait, GRF,
angular
orientation of
trunk/ thigh/leg
and foot
Walking
under 5
conditions:
WL,
carrying
20% & 40%
BW in BP,
carrying
20% &
40% BW in
DP
BP
vs.
DP
Y Less
changes in
gait pattern
& trunk
forward
inclination
with using
DP
Fiolk
owsk
i et al
2006
[14]
1
3
24.
8±
3.1
N
Video
analys
is
syste
m,
VAS
Knee/ankle/5th
metatarsophala
ngeal joint
angles,
displacement of
body landmarks,
Pressure on the
shoulders/ low
back, ease of
walking, overall
comfort
Walking
WL or
carrying
10% and
15% BW
either in BP
or FP
FP
vs.
BP
N FP had
less effect
on posture
than BP
Kim
et al
2008
[22]
1
5
9-
11
Y EMG,
3D
motio
n
analys
is
syste
m
Electrical
activity of neck
muscles,
forward head
angle and
distance
Standing &
walking
under 4
conditions:
WL or
carrying
15% BW in
a BP, DP
or modified
DP
BP
vs.
DP
Y Less
postural
deviation
was seen
while
carrying
the
modified
DP
24
John
son
et al
1995
[19]
1
5
21
-
36
N Q, 6-
point
scale
March time,
intensity of
fatigue,
discomfort, back
pain
Walking
while
carrying 34,
48 or 61 kg
in an army
BP or DP
BP
vs.
DP
Y No
difference
between 2
pack types
on comfort
BP, Backpack; BW, Body Weight; DP, Double Pack; EMG, Electromyography; FP, Front Pack; GRF, Ground Reaction Force; HR, Heart Rate; N, No; Q, Questionnaire; VAS, Visual Analogue Scale; VO2, Oxygen Consumption; WL, Without Load; Y, Yes.
Table 5. Critical appraisal of studies which investigated front pack or
double pack vs. backpack
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Tota
l score
Stu
dy d
esig
n &
le
vel o
f
evid
en
ce
N Y N U N N Y N Y Y Y N Y Y Y
8/15 III-2A
N Y N Y N N Y N Y Y Y N Y Y Y
9/15 III-2A
N Y N Y N Y Y U Y Y Y N Y Y Y
10/15
III-2A
N N N U N N N N Y Y Y N Y Y Y
6/15 III-3
N Y N U N N Y N Y Y Y N Y Y N
7/15 III-3A
N Y N U N N N N Y Y Y N Y Y N
6/15 III-2
N Y N U N N N N Y Y Y N N Y
Y 6/15 III-2
N N N N N N Y U N Y Y N N Y
Y
5/15 III-2A
N Y N N/A
N N Y N Y Y Y N Y Y Y 8/14 III-2A
Refer to table 3.
Shoulder straps
Seven studies assessed backpack straps. Three of these compared single
strap vs. double strap backpacks while one study investigated the optimal
25
strap angle and best location for attachment of straps. Three studies
investigated the tightness of the straps when comparing loose fit pack vs.
tight fit pack. Descriptive characteristics and quality appraisal of these studies
can be seen in tables 6 and 7.
Table 6. Descriptive characteristics of studies that investigated
shoulder straps
Auth
or(
s)
, year
Sam
ple
siz
e
Age o
f p
art
icip
an
ts
Inclu
sio
n &
exclu
sio
n c
rite
ria
Equ
ipm
ent
&
Instr
um
ents
of
measure
men
t
Vari
able
s m
easure
d
Task
Wha
t e
lem
ent
of b
ackpack
was s
tudie
d/d
esig
ne
d
Cle
arl
y s
tate
d a
ims
Result
Legg
and
Cruz
2004
[27]
1
3
27.
3±
9.3
Y Spirom
eter,
Peak
flow
meter
FVC, FEV1,
FVC1, Peak
Expiratory Flow
Maximally
forced
expiration
during
upright
stance
under 3
situations:
WL,
wearing
SSB or
DSB
SSB
vs.
DSB
Y SSB can
cause
respiratory
dysfunctio
n
Bygr
ave
et al
2004
[8]
1
2
25
±5
Y Spirom
eter
FVC, FEV1,
FVC1, peak
expiratory flow
maximally
forced
expiration
in an erect
relaxed
standing
under 3
conditions:
WL, loose
fit pack,
tight fit
pack
Loos
fit
pack
vs.
tight
fit
pack
(tight
ness
of
strap
s &
belts)
Y Tight fit
pack can
lead to
pulmonary
impairment
.
Reid
et al
2000
§
[43]
N
/
A
N/
A
N
/
A
Load
Distrib
ution
Manne
quin,
Pressu
re
Lumbar shear
force,
compressive
force over all
attachment
locations, load
distribution
N/A strap
angle
&
optim
al
locati
on for
Y The strap
angle
between
24⁰ & 30⁰
with
respect to
the vertical
26
sensor
, force
plate
between upper
& lower torso,
peak/average
anterior
shoulder/
armpit contact
pressure, waist
belt lift
lower
attac
hmen
t
point
of
strap
angle
axis of the
body
caused
least
contact
pressure
beneath
straps &
more
comfort
Mack
ie et
al
2005
§
[36]
N
/
A
N/
A
N
/
A
Load
carriag
e
simulat
or ,
pressu
re
sensor
Shoulder strap
force &
pressure
N/A Loos
eness
of
strap
s +
hip
belt
Y Loose
straps
produced
less
shoulder
strap force
& pressure
Hong
and
Li
2005
[17]
1
3
12.
2±
0.9
Y
In-
shoe
pressu
re
measu
rement
system
Vertical GRF,
force-to-time
ratio of the first
peak force, gait
cycle/stance/sin
gle and double
support
duration, time
to peak force
Climbing
up and
down stairs
while
carrying
different
loads in a
DSB or
SSB
SSB
vs.
DSB
Y The critical
threshold
load was
less while
carrying
SSB to
maintain a
stable gait
pattern
compared
with DSB
Stanf
ord
et al
2002
[49]
1
0
13
-
15
N Motion
analysi
s
system
, global
coordi
nate
system
Trunk & pelvis
motion, Rt & Lt
shoulder
elevation &
swing, mean
angle & range
of motion for
each kinematic
parameters
Walking
under 3
conditions:
WL,
carrying
20% BW in
a DSB or a
SSB
SSB
vs.
DSB
Y SSB
caused
more
postural
changes
Mack
ie
and
Legg
2008
[34]
1
8
13
-
14
N
Camer
a, Q,
Horizontal
displacement of
7 body
landmarks
relative to ankle
joint, Comfort,
RPE
Standing
WL or
wearing 5,
10, 12.5 or
15% BW in
a BP. Also,
a BP
loaded to
10% BW
was carried
with tighter
shoulder
straps
Tight
ness
of
shoul
der
strap
s
Y No
difference
on
posture,
comfort
and
balance
between
loose &
tight
shoulder
straps
27
BP, Backpack; BW, Body Weight ; FEV1, Forced Expiratory Volume in 1sec ; FVC, Forced Vital Capacity ;GRF, Ground Reaction Force; N/A, Not Applicable; N, No; Q, Questionnaire; RPE, Rating of Perceived Exertion; WL, Without Load; Y, Yes; §, This study has been conducted on a mannequin and not human subjects
Table 7. Quality appraisal of studies that investigated shoulder straps
Au
tho
r(s),
ye
ar
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Tota
l score
Stu
dy d
esig
n &
le
vel o
f
evid
en
ce
Legg and Cruz 2004
N Y U U N N Y N N Y Y N N N N
4/15
III-2A
Bygrave et al 2004
N Y U U N N N Y N Y Y N N Y Y 6/15
III-2
Reid et al 2000
N/A
N/A
U Y N N Y Y N N Y N N Y N
5/13
III-3
Mackie et al 2005
N/A
N/A
U Y Y Y Y Y Y Y Y N Y Y Y
11/13
III-3
Hong and Li 2005
N Y U Y N Y Y Y N Y Y N N Y Y 9/15
III-2A
Stanford et al 2002
N Y U Y N N Y N N Y Y N Y Y Y 8/15
III-2A
Mackie and Legg 2008
N N U Y N Y N N Y Y Y N Y Y Y 8/15
III-2
Refer to table 3
28
Discussion
The results of this systematic review show significant variability in the design
of the studies. This variability prevented any meaningful pooling of data. With
83% of studies published since 2000, it shows that interest and demand in
studying the most popular mode of load carriage is rising. A substantial
amount of literature in this review focused on providing the best backpack for
a specific group of people, mostly students or soldiers, while the rest of the
studies targeted others who use backpacks. Nevertheless, all of the studies’
reviewed aimed to discover the elements of a backpack that have the best
effect on posture, perceived pain, lung function and other comfort variables.
A backpack may maintain posture yet lead to respiratory impairment or
higher energy expenditure. So, to design a suitable backpack or to design a
feature of a backpack, the effect of the backpack on multiple important
aspects of bodily function should be considered.
Best backpack placement on the spine/ best load placement in a
backpack
Based on the results of 4 studies, low backpack placement leads to less
postural deviation and postural sway [2; 4; 9; 15]. One of these 4 studies was
conducted on adults and the other 3 on adolescents. Three and two studies
showed that load placement doesn’t have any influence on gait and posture,
respectively. It has been recommended that for carrying more than 15% of
body weight, low load placement should be avoided [46]. Other studies
investigated the effect of load positioning on energy and oxygen
consumption, pain, contact pressure and pulmonary function; however, due
to limited evidence no strong conclusions can be made.
Front pack and double pack
Results of 5 studies showed that front packs and double packs had less
effect on posture than backpacks [14; 22-24; 57]. Double packs help place
the centre of gravity of the load closer to the body and provide more stability
and efficiency for carrying heavy loads by distributing the load between the
back and front of the body; on the other hand, double packs can restrict
respiratory ventilation, prevent movement of upper limbs and reduce the
29
visual field in front of the body [23; 31; 39; 52]. There is limited evidence on
influence of these two types of packs on other variables such as pain,
discomfort, energy consumption, electrical activity of muscles and etc.
Shoulder straps
There was little evidence found about backpack straps. What we did find was
that single strap backpacks caused respiratory dysfunction and more postural
changes than double strap backpacks. Three studies compared looseness
vs. tightness of shoulder straps but the findings of these studies are
conflicting.
General recommendations
Despite the difficulties with heterogeneity among the included studies we can
conclude that there is no one backpack that minimises effects on the body
from all of the aspects. In addition we can also propose some guidelines for
best practice design and use of backpacks (Table 8). We make these
recommendations taking into account the number of appraisal criteria
satisfied in tables 3, 5 and 7 which reflect whether the study should be
regarded as having a high risk of bias. Further discussion of these issues
follow.
Table 8: Recommendations for best backpack
User Population Children Adult
Backpack Fit Loose or tight fit pack made no difference on posture, comfort and balance [34]
Loose fit pack results in postural instability and higher energy consumption while tight fit pack leads to respiratory problems [8]
Loose fit pack resulted in less shoulder strap tension force and less shoulder interface pressure [36]
Best Type Of Shoulder Strap Regarding Lung Function and
No Research Double Strap Backpack is recommended over single strap backpack
Aspects
of design
30
physiological variables
[27]
Best Type Of Shoulder Strap Regarding Gait & Posture
Double Strap Backpack is recommended over single strap backpack [17; 49]
No Research
Optimal Angle of Shoulder Strap
Between 24-30 degrees with respect to the vertical axis of the body [43]
‡ Lung Function and Physiological Variables
No Difference [10] Conflicting [29; 52]
‡ Gait & Posture Conflicting [4, 8, 9, 15, 21, 34, 46, 47]
Conflicting [13]
‡ Shoulder Contact Pressure
Upper Placement is recommended over lower placement [32]
No Research
Lower placement is recommended over upper placement [36]
Backpack Vs. Front Pack Or Double Pack Regarding Posture
Front pack [57] and double pack [22] are recommended over backpack
Front pack [14] and double pack [23; 24] are recommended over backpack
‡, Best Backpack Placement on the Spine/ Best Load Placement in the
Backpack Regarding
Limitations of the studies
Differences in study populations prevented complete combination and
comparison of study findings. Such differences inhibit identifying all of the
best features of a suitable backpack for a general market. These differences
include the variables of age, gender, stature, physical fitness and intended
usage of the backpack. Furthermore, there are other factors which make
combining or comparing the findings of the studies difficult. Study differences
included duration of carrying a backpack, dynamic or static condition of
testing, evenness or unevenness of terrain, and the part or system of the
body that was the focus of the studies.
This review reveals widespread deficiencies in the validity, reliability and
calibration of equipment and instruments used to study backpacks. These
factors are fundamental issues in producing meaningful evidence; therefore,
we recommend more studies need to be conducted to find reliable and valid
instruments for measuring the influence of backpack design on the body.
31
Moreover, none of the studies mentioned training of the examiners or
providing training sessions for subjects in self rated outcome measures. This
may have also influenced the validity and reliability of the studies.
Sample size was just justified in one paper. A study with a small sample size
may not detect significant results. Also, there is a chance of random error
and publication bias in small studies, because interesting and favourable
results from small studies might be reported whereas less interesting findings
from small studies remain unreported [11; 12].
In this review, an appraisal scoring system was not used and studies were
not labelled by low, moderate or high quality; instead, trials were classified
with the notation of how many appraisal criteria they satisfied and were
assessed for every single item separately. It is worth noting that some
studies might be strong in some parts but poor in other aspects.
Inadequate reporting and lack of information of research methods reduced
the quality of studies in this review. More accurate reports of studies might be
achieved if more detailed information was reported. In this regard the
variables listed in the descriptive tables and the critical appraisal tables could
act as a template for improved study design and reporting.
As discussed it does not seem possible to design a backpack suitable for
everyone and usable in every situation. A best suited backpack for short
distance walking may not be the best for marching or hiking. Additionally, the
most comfortable backpack for girls may not also be the most comfortable for
boys. More comfortable and efficient backpack design should provide for
varying backpack sizes to suit different age groups, different average means
of height, weight and torso stature for that specific age group. However, just
a few studies considered backpack size or mentioned that what backpack
sizes were used in their investigations. Generally no explanation was
provided about the reasons why the size of the backpack studied was
chosen. Additionally, no information was provided about the optimal
dimensions especially the width of backpack features such as shoulder
straps, hip belt and sternum belt.
32
Backpack fabric is another factor which can be important for choosing a
backpack. Backpack fabric can make a backpack heavier or lighter and put
more or less pressure on the body. Also, some fabrics provide better air
circulation between the body and pack and prevent less friction and heat
production between them. This has implications in hot weather or for long
distance use.
Limitations of this study
There are a number of limitations to the current study. This review was not a
totally blind review; authors and publication details were disclosed to the
reviewers and this can potentially lead to reviewer bias. However, reviewers
were not aware of the background and previous works of the authors. A
further limitation is that although the search strategy was comprehensive it is
possible that some studies were not found. Also, the validity and reliability of
the critical appraisal tool used in this study has not been established but was
developed from first principles using previously developed tools from related
areas. The suggestion of potential bias in studies using the number of critical
appraisal variables achieved is controversial and readers are invited to use
this as a guide only.
Acknowledgment
The authors thank Stacy Brunn for her independent review of the studies.
References
[1] T.N.H.a.M.R.C. (NHMRC), NHMRC additional levels of evidence and
grades for recommendations for developers of guidelines, in, 2008, p. Stage
2 consultation.
[2] J. Bobet and R.W. Norman, Effects of load placement on back muscle
activity in load carriage, European Journal of Applied Physiology and
Occupational Physiology 53 (1984), 71-75.
[3] D.R. Boulware, Backpacking-induced paresthesias, Wilderness and
Environmental Medicine 14 (2003), 161-166.
[4] H.M. Brackley, J.M. Stevenson, and J.C. Selinger, Effect of backpack
load placement on posture and spinal curvature in prepubescent children,
Work 32 (2009), 351-360.
33
[5] G. Brattberg, Do pain problems in young school children persist into
early adulthood? A 13-year follow-up, European Journal of Pain 8 (2004),
187-199.
[6] A.M. Briggs, A.J. Smith, L.M. Straker, and P. Bragge, Thoracic spine
pain in the general population: prevalence, incidence and associated factors
in children, adolescents and adults. A systematic review, BMC
Musculoskeletal Disorders 10 (2009), 77.
[7] N. Buscemi, L. Hartling, B. Vandermeer, L. Tjosvold, and P. Klassen,
Single data extraction generated more errors than double data extraction in
systematic reviews, Journal of Clinical Epidemiology 59 (2006), 697-703.
[8] S. Bygrave, S.J. Legg, S. Myers, and M. Llewellyn, Effect of backpack
fit on lung function, Ergonomics 47 (2004), 324-329.
[9] D.H.K. Chow, Z.Y. Ou, X.G. Wang, and A. Lai, Short-term effects of
backpack load placement on spine deformation and repositioning error in
schoolchildren, Ergonomics 53 (2010), 56-64.
[10] D.H.K. Chow, J.M.L. Ting, M.H. Pope, and A. Lai, Effects of backpack
load placement on pulmonary capacities of normal schoolchildren during
upright stance, International Journal of Industrial Ergonomics 39 (2009), 703-
707.
[11] I.K. Crombie, The pocket guide to critical appraisal BMJ publishing
group, London, 1996.
[12] H.C.W. De Vet, T. Van Der Weijden, J.W.M. Muris, J. Heyrman, F.
Buntinx, and J.A. Knottnerus, Systematic reviews of diagnostic research.
Considerations about assessment and incorporation of methodological
quality, European Journal of Epidemiology 17 (2001), 301-306.
[13] C. Devroey, I. Jonkers, A. de Becker, G. Lenaerts, and A. Spaepen,
Evaluation of the effect of backpack load and position during standing and
walking using biomechanical, physiological and subjective measures,
Ergonomics 50 (2007), 728-742.
[14] P. Fiolkowski, M. Horodyski, M. Bishop, M. Williams, and L. Stylianou,
Changes in gait kinematics and posture with the use of a front pack,
Ergonomics 49 (2006), 885-894.
[15] K. Grimmer, B. Dansie, S. Milanese, U. Pirunsan, and P. Trott,
Adolescent standing postural response to backpack loads: A
randomisedcontrolled experimental study, BMC Musculoskeletal Disorders 3
(2002), 1-10.
34
[16] M. Holewijn and W.A. Lotens, The influence of backpack design on
physical performance, Ergonomics 35 (1992), 149-157.
[17] Y. Hong and J.X. Li, Influence of load and carrying methods on gait
phase and ground reactions in children's stair walking, Gait and Posture 22
(2005), 63-68.
[18] T. Horton, Backpack spraying tips: Improving vegetation management
on utility ROWs, Public Works 136 (2005), 129.
[19] R.F. Johnson, J.J. Knapik, and D.J. Merullo, Symptoms during load
carrying: effects of mass and load distribution during a 20-km road march,
Perceptual and Motor Skills 81 (1995), 331-338.
[20] P. Juni, A. Witschi, R. Bloch, and M. Egger, The hazard of scoring the
quality of clinical trials for meta-analysis, JAMA 282 (1999), 1054-1060.
[21] E. Kellis and F. Arampatzi, Effects of sex and mode of carrying
schoolbags on ground reaction forces and temporal characteristics of gait,
Journal of Pediatric Orthopaedics Part B 18 (2009), 275-282.
[22] M.H. Kim, C.H. Yi, O.Y. Kwon, S.H. Cho, and W.G. Yoo, Changes in
neck muscle electromyography and forward head posture of children when
carrying schoolbags, Ergonomics 51 (2008), 890-901.
[23] H. Kinoshita, Effects of different loads and carrying systems on
selected biomechanical parameters describing walking gait, Ergonomics 28
(1985), 1347-1362.
[24] J.J. Knapik, P. Ang, H. Meiselman, W. Johnson, J. Kirk, C. Bensel,
and W. Hanlon, Soldier performance and strenuous road marching: Influence
of load mass and load distribution, Military Medicine 162 (1997), 62-67.
[25] P. Korovessis, G. Koureas, S. Zacharatos, and Z. Papazisis,
Backpacks, back pain, sagittal spinal curves and trunk alignment in
adolescents: A logistic and multinomial logistic analysis, Spine 30 (2005),
247-255.
[26] S.G. Lee, J.H. Kim, S.Y. Lee, I.S. Choi, and E.S. Moon, Winged
scapula caused by rhomboideus and trapezius muscles rupture associated
with repetitive minor trauma: A case report, Journal of Korean Medical
Science 21 (2006), 581-584.
[27] S.J. Legg and C.O. Cruz, Effect of single and double strap backpacks
on lung function, Ergonomics 47 (2004), 318-323.
[28] S.J. Legg, L. Perko, and P. Campbell, Subjective perceptual methods
for comparing backpacks, Ergonomics 40 (1997), 809-817.
35
[29] B.S. Liu, Backpack load positioning and walking surface slope effects
on physiological responses in infantry soldiers, International Journal of
Industrial Ergonomics 37 (2007), 754-760.
[30] R. Lloyd and C.B. Cooke, Kinetic changes associated with load
carriage using two rucksack designs, Ergonomics 43 (2000 a), 1331-1341.
[31] R. Lloyd and C.B. Cooke, The oxygen consumption associated with
unloaded walking and load carriage using two different backpack designs,
European Journal of Applied Physiology 81 (2000 b), 486-492.
[32] B.R. Macias, G. Murthy, H. Chambers, and A.R. Hargens, Asymmetric
loads and pain associated with backpack carrying by children, Journal of
pediatric orthopedics 28 (2008), 512-517.
[33] W.G. Mackenzie, J.S. Sampath, R.W. Kruse, and G.J. Sheir-Neiss,
Backpacks in children, Clinical Orthopaedics and Related Research (2003),
78-84.
[34] H.W. Mackie and S.J. Legg, Postural and subjective responses to
realistic schoolbag carriage, Ergonomics 51 (2008), 217-231.
[35] H.W. Mackie, S.J. Legg, J. Beadle, and D. Hedderley, Comparison of
four different backpacks intended for school use, Applied Ergonomics 34
(2003), 257-264.
[36] H.W. Mackie, J.M. Stevenson, S.A. Reid, and S.J. Legg, The effect of
simulated school load carriage configurations on shoulder strap tension
forces and shoulder interface pressure, Applied Ergonomics 36 (2005), 199-
206.
[37] A.B. Marsh, L. DiPonio, K. Yamakawa, S. Khurana, and A.J. Haig,
Changes in posture and perceived exertion in adolescents wearing
backpacks with and without abdominal supports, American Journal of
Physical Medicine and Rehabilitation 85 (2006), 509-515.
[38] S. Masiero, E. Carraro, A. Celia, D. Sarto, and M. Ermani, Prevalence
of nonspecific low back pain in schoolchildren aged between 13 and 15
years, Acta Paediatrica, International Journal of Paediatrics 97 (2008), 212-
216.
[39] R.R.E.E. Motmans, S. Tomlow, and D. Vissers, Trunk muscle activity
in different modes of carrying schoolbags, Ergonomics 49 (2006), 127-138.
[40] D. Nayduch and S.L. Sullivan, The problems of intrahospital transfer of
patients with trauma and one solution: the "Trauma Transfer Backpack",
Journal of Emergency Nursing 18 (1992), 383-389.
36
[41] S. Negrini and R. Carabalona, Backpacks on! Schoolchildren's
perceptions of load, associations with back pain and factors determining the
load, Spine 27 (2002), 187-195.
[42] D.D. Pascoe, D.E. Pascoe, Y.T. Wang, D.M. Shim, and C.K. Kim,
Influence of carrying book bags on gait cycle and posture of youths,
Ergonomics 40 (1997), 631-641.
[43] S.A. Reid, J.T. Bryant, J.M. Stevenson, and J.B. Doan, Biomechanical
assessment of rucksack shoulder strap attachment location: Effect on load
distribution to the torso, in: Proceedings of the RTO-NATO specialist
meeting, 2000, pp. 1-9.
[44] S.A. Reid, J.M. Stevenson, and R.A. Whiteside, Biomechanical
assessment of lateral stiffness elements in the suspension system of a
backpack, Ergonomics 47 (2004), 1272-1281.
[45] G.I. Sheir-Neiss, R.W. Kruse, T. Rahman, L.P. Jacobson, and J.A.
Pelli, The association of backpack use and back pain in adolescents, Spine
28 (2003), 922-930.
[46] T. Singh and M. Koh, Effects of backpack load position on
spatiotemporal parameters and trunk forward lean, Gait and Posture 29
(2009 a), 49-53.
[47] T. Singh and M. Koh, Lower limb dynamics change for children while
walking with backpack loads to modulate shock transmission to the head,
Journal of Biomechanics 42 (2009 b), 736-742.
[48] S.A. Southard and G.A. Mirka, An evaluation of backpack harness
systems in non-neutral torso postures, Applied Ergonomics 38 (2007), 541-
547.
[49] C. Stanford, P. Francis, and H. Chambers, Changes in upper body
and arm kinematics while carrying symmetrical and asymmetrical backpack
loads, in: International society of biomechanics in sports, Extremadura-Spain
2002, pp. 610-613.
[50] E. Steele, A. Bialocerkowski, and K. Grimmer, The postural effects of
load carriage on young people - a systematic review, BMC Musculoskeletal
Disorders 4 (2003), 12.
[51] J.M. Stevenson, J.T. Bryant, S.A. Reid, R.P. Pelot, E.L. Morin, and
L.L. Bossi, Development and assessment of the Canadian personal load
carriage system using objective biomechanical measures, Ergonomics 47
(2004), 1255-1271.
37
[52] K.J. Stuempfle, D.G. Drury, and A.L. Wilson, Effect of load position on
physiological and perceptual responses during load carriage with an internal
frame backpack, Ergonomics 47 (2004), 784-789.
[53] S. Taimela, U.M. Kujala, J.J. Salminen, and T. Viljanen, The
prevalence of low back pain among children and adolescents: A nationwide,
cohort-based questionnaire survey in Finland, Spine 22 (1997), 1132-1136.
[54] P. Viry, C. Creveuil, and C. Marcelli, Nonspecific back pain in children:
A search for associated factors in 14-year-old schoolchildren, Revue du
Rhumatisme. English Edition 66 (1999), 381-388.
[55] B.F. Walker, The prevalence of low back pain: A systematic review of
the literature from 1966 to 1998, Journal of Spinal Disorders 13 (2000), 205-
217.
[56] E.J. Wall, S.L. Foad, and J. Spears, Backpacks and back pain:
Where's the epidemic?, Journal of Pediatric Orthopaedics 23 (2003), 437-
439.
[57] C.X.G. Wang, D.H.K. Chow, and M.H. Pope, Biomechanical effect of
LOAuD carriage on spine curvature and repositioning ability in adolescents,
in: Proceedings of the Fifth IASTED International Conference on
Biomechanics, BioMech 2007, 2007, pp. 161-166.
[58] P. Whiting, R. Harbord, and J. Kleijnen, No role for quality scores in
systematic reviews of diagnostic accuracy studies, BMC Medical Research
Methodology 5 (2005), 19.
[59] J.K. Whittfield, S.J. Legg, and D.I. Hedderley, The weight and use of
schoolbags in New Zealand secondary schools, Ergonomics 44 (2001), 819-
824.
38
CHAPTER THREE
CAN LOAD CARRIAGE SYSTEM WEIGHT, DESIGN AND
PLACEMENT AFFECT PAIN AND DISCOMFORT? A
SYSTEMATIC REVIEW.
Published as
Golriz S, Walker B. (2011)
Journal of Back and Musculoskeletal Rehabilitation, 24(1):1-16.
39
In this chapter we report our findings of a systematic review which we
conducted in order to answer three questions: 1) Does usage or the weight of
load carriage systems cause pain, perceived exertion or discomfort? 2) Can
load carriage system placement on the spine influence pain, perceived
exertion or discomfort? 3) Can load carriage system design influence the
amount of pain, perceived exertion or discomfort caused by their use?
The work resulted in the following publication:
Golriz S, Walker B. Can load carriage system weight, design and placement
affect pain and discomfort? A systematic review. Journal of Back and
Musculoskeletal Rehabilitation. 2011;24(1):1-16.
40
Do backpacks cause back pain? A systematic review
Samira Golriz1 MSPT, Bruce Walker2 DC, MPH, DrPH
1PhD Student, School of Chiropractic and Sports Science, Murdoch
University, Murdoch 6150, Western Australia, Australia
2Senior Lecturer, School of Chiropractic and Sports Science, Murdoch
University, Murdoch 6150, Western Australia, Australia
Contact information:
[email protected] ; Tel: + 61 8 9360 1450, Fax: + 61 8 9360 1299
School of Chiropractic and Sports Science, Murdoch University, 90 South
Street, Murdoch, WA 6150, Australia
41
Abstract.
Purpose: A systematic review of the literature was conducted to answer the
questions 1. Does backpack use or backpack weight cause pain or
discomfort? 2. Can backpack placement on the spine influence any pain or
discomfort? 3. Can backpack design influence the amount of any pain or
discomfort caused by their use?
Method: Eight databases were searched. Each included study was analysed
and quality appraised by two independent reviewers.
Results: Forty-five articles that addressed the research questions were
included in the study. Significant variability in the study design and
populations of the studies prevented data pooling and the evidence is
conflicting. However, qualitative synthesis of the studies shows that carrying
loads may provoke low back pain; and it may also trigger neck, thoracic and
shoulder pain. Backpack weight can influence perceived pain, however other
factors are involved.
Discussion: There is conflicting but positive evidence on the correlation
between backpack load carrying and experiencing pain during different
stages of life. The research to date is lacking with the most commonly
identified methodological deficiencies being poor overall design, the lack of
justification of sample size, providing training sessions for examiners, and not
utilising calibrated, valid and reliable instruments for measurement.
Keywords: backpack, pain, systematic review, front pack, double pack
42
Introduction
Adult back pain is a significant source of long term dysfunction and absence
from work which puts a huge economic, social and emotional burden on
individuals and society [56,57]. Additionally, back pain is a current issue
among young people with low back pain prevalence in adolescents
measured between 20% to 72% [29,38]. Young people commonly use
backpacks as they are an effective and most economical way of carrying
weight, however, it has been proposed they can also be a significant
contributing risk factor for discomfort, fatigue, muscle soreness and
musculoskeletal pain especially low back pain [9,29,47-49,57]. It has been
speculated that backpacks may cause problems not only for the developing
skeletal system but also for a mature spine as a developed spine is also
sensitive to load [41,45]. Moreover, experiencing back pain in childhood is a
concern as it may lead to more common and severe issues later in life [41].
Various suggested cut off backpack weights have been recommended by
researchers in order to reduce the risk associated with backpack use.
However, do backpacks really cause pain and discomfort?
The aim of this systematic review is to answer the following questions about
a broad range of population groups:
Does usage or weight of load carriage system cause pain, perceived exertion
or discomfort?
Can load carriage system placement on the spine influence pain, perceived
exertion or discomfort?
Can load carriage system design influence the amount of pain, perceived
exertion or discomfort caused by their use?
Methods
Literature search
A comprehensive search strategy was conducted to identify all relevant
publications on load carriage systems and their design. The search strategy
is seen below.
43
Allied health, health-research, health-science and medical databases
including Medline, Cochrane library, Science Direct, PubMed, Scopus,
CINAHL, MANTIS and EMBASE were used. The search was performed
using the combination of the following key indexing terms: (‘backpack’, ‘back
pack’, ‘rucksack’, ‘schoolbag’, ‘school bag’, ‘load carriage system’) and
(‘pain’, ‘discomfort’, ‘perceived exertion’, ‘comfort’) and (‘design’ or
‘performance’). Google searches were also carried out to find any related
articles, meeting proceedings or links. Furthermore, a hand search of the
reference lists of existing articles was conducted to find papers that did not
appear in the main database searches. The search covered literature from
1966 to February 2010.
Selection criteria
Studies with the main focus on the human effects of load carriage systems
(backpacks, front packs or double packs) on comfort, discomfort, pain or
perceived exertion were included. Only studies that were conducted on
humans and not manikins were included. Also, studies that focussed on
unhealthy subjects (e.g. scoliosis) were excluded from the review. Studies
were limited to peer-reviewed journals and conference proceedings. Case
reports and clinical opinions were excluded. This led to broad inclusion
criteria for study design in order to prevent limitation of potentially relevant
articles. All study abstracts meeting these broad criteria were initially
included. Subsequent inclusion based on the inclusion criteria was then
assessed by two trained reviewers (SG and BW) who reviewed the papers
independently. If the eligibility of studies was not clear from the abstracts,
then full texts of the articles were obtained and assessed independently by
the two authors. If a difference of opinion occurred, consensus was reached
on inclusion or exclusion by discussion and reflection. A third party could be
used in the event of disagreement.
Data extraction and management
SG acted as the principal reviewer. Three other reviewers (MW, NM, and
WN) were trained by SG and BW (an experienced investigator) and acted as
the second reviewers to extract data from the included papers. Training
44
sessions included clarification of all data items and required elements of the
quality appraisal tool were provided. Standardisation of the procedure was
required to provide consistency in methods used by the reviewers; therefore,
before starting to extract data, a trial was conducted on two similar but
unrelated papers and the results discussed. Co-investigator (BW) was
consulted when there was disagreement between SG and MW, NM or WN.
Data extraction form
This form consisted of descriptive characteristics and a quality appraisal tool.
Data were extracted based on the elements of this form which were related
to the research questions and aims of this review and seen in tables 2–9.
Level of evidence
The level of evidence of each paper was assessed based on the National
Health and Medical Research Council of Australia guidelines (table 1). This
was based on the proposition that some designs provide more valid and
reliable findings than others. The lower the ranking in hierarchy of evidence
the greater the risk of bias or error in a study [1].
Table 1: NHMRC level of evidence
Level of evidence
Study design (intervention) Study design (aetiology)
I A Systematic review of randomised controlled trials
A systematic review of level II studies
II A randomised controlled trial A prospective cohort study III-1 A pseudo randomised controlled trial All or none III-2 A comparative study with concurrent
controls: Non-randomised experimental trial Cohort study Case-control study Interrupted time series with a control group
A retrospective cohort study
III-3 A comparative study without concurrent controls: Historical control study Two or more single arm study Interrupted time series without a parallel control group
A case-control study
IV Case series with either post-test or pre-test/post-test outcomes
A cross-sectional study or case series
45
Quality appraisal
The quality of papers was assessed according to a modified version of the
quality appraisal tool by Crombie [11]. In this study, we modified the Crombie
tool by adding three extra appraisal items, ‘attention to calibration of
equipment/ instrument before use’, ‘Was the person who carried out the
measurement trained?’ and ‘Discussion of weaknesses or limitations of the
study in the paper’. Moreover, validity and reliability of measurements which
were fitted into one item by Crombie, were split up into two questions as
these two concepts demonstrate two different aspects in research. The
modified Crombie quality appraisal items can be seen in table 5.
Answers to the quality appraisal items were defined as Yes, No, Not
Applicable or Unclear. In the case that two or more pieces of equipment or
instruments were used, details of calibration, validity or reliability of one of
instruments was considered an adequate description of the validity or
reliability. A score of one was given to each yes answer and zero to no,
unclear and N/A answers. The overall score was reported as a tally of all yes
answers out of 15, 14 or 13 based on the applicable answers for each study.
Often, reviewers add the scores of individual items from the critical appraisal
tool to present a total score [51]. However, using this method may be
arbitrary as is weighting each item. Instead, it has been recommended that
each item be investigated separately, rather than use a combined quality
score [23,58]. However, given the dichotomy of views in the literature we
chose to simply classify studies with the notation of how many critical
appraisal items they satisfied. In this way we believe an estimation of the
quality of the study can be gained, with studies that meet less appraisal
criteria being treated with more caution.
Results
Two hundred and eighty four articles were identified from the databases
using the search strategy. Titles and abstracts of these articles were
manually screened for relevance and 178 articles were excluded. The
remaining articles (n=106) were studied in detail to see if they satisfied the
inclusion criteria. A further 59 papers were excluded as did not meet the
46
selection criteria. Two articles were added after reference checking. Two
articles were excluded as the full texts were unobtainable by the Murdoch
University library staff despite genuine efforts [25,55] and one paper was
excluded as the full text was not in readable form [42]. Forty six articles were
included for the final review.
Figure 1: Inclusion and exclusion of articles
Study results
It is worth noting that noxious human effects in many studies were labelled
with different words such as pain, discomfort and perceived exertion. Also,
these variables were often assessed using a variety of scales such as
regional body diagrams, categorical 5-point or 7-point scales,
musculoskeletal discomfort diagrams, soreness and discomfort figures,
Visual Analogue Scales (VAS) and Borg Scales (BS). The latter two being
the most commonly used. We found 46 suitable trials to include in this
review. Of these, 26 trials examined the correlation between backpack use
and pain, perceived exertion or discomfort, seven studies assessed the
correlation of pain, discomfort and perceived exertion with increasing load,
three studies investigated the effect of load placement on pain, perceived
exertion or discomfort and 10 studies compared the effect of different designs
and features of backpacks on pain, discomfort and perceived exertion.
Articles discovered from online databases (n= 284)
Excluded after screening of
titles and abstracts (n=178)
9
Full texts were studied for
more detail (n=106) Excluded as did not meet
the inclusion criteria (n=59)
Articles added after
hand search (n=2) Excluded as the full texts
were unobtainable (n=2)
9
Excluded as the full text was
not in readable form (n=1)
9
Articles accepted for review
(n=46)
47
What is the relationship between backpack use & pain, perceived
exertion or discomfort?
In this part of the review, 33 papers were included. Twenty out of 33 studies
used cross-sectional design to collect data through use of various
questionnaires and 13 studies utilised experimental design. Twenty-six
studies assessed the correlation between backpack use and pain, perceived
exertion or discomfort and seven studies assessed the association between
pain, perceived exertion or discomfort with increasing load. Twenty-three
studies assessed adolescents while in 10 trials adults were investigated.
Eight studies studied women or men exclusively. In nine studies, it was not
clear if they examined any gender exclusively as they didn’t provide the
number of male or female subjects. In just 10 studies, subjects were
screened for entry into the experiment based on the inclusion and exclusion
criteria. Descriptive characteristics of these studies can be seen in table 2
and results of these studies can be seen in table 3 and 4.
Table2: Descriptive characteristics of studies that assessed the
correlation between backpack use and pain, perceived exertion and
discomfort
Au
tho
rs, P
ub
licat
ion
Ye
ar
Sam
ple
siz
e
Age
(ran
ge o
r M
ean
± S
D)
Ge
nd
er
Incl
usi
on
an
d e
xclu
sio
n
crit
eri
a
Inst
rum
en
ts o
f
me
asu
rem
en
ts
Task
Co
mp
licat
ion
s o
f w
ear
ing
bac
kpac
k h
ave
be
en
re
po
rte
d o
n
Cle
arly
sta
ted
aim
s
Du
rati
on
of
carr
yin
g
bac
kpac
k
Marsh et al 2006 [37]
20 13-16 F:9 M:11
Y VAS, BS Walking while wearing BP with and without an abdominal support (10 & 20% BW)
RPE Y 5 min for each trial
Siambanes et al 2004 [48]
3498 11-15 U N Digital electronic scale, Q
Q back pain Y -
Navuluri & Navuluri 2006 [40]
61 F: 12.9 M: 13.3
F:32 M:29
N Q, BS, scale
Q Back and neck pain
Y -
Chiang et al 2006 [10]
55 13.25±0.4
U N Q, digital scale
Q Low back pain
Y -
Wall et al 2003 [57]
346 6-18 F:237 M:109
Y Q Q Low back pain
Y -
Grimmer & Williams 2000 [15]
1193 8-12 U N digital electronic scale, Q
Q Low back pain
Y -
Moore et al 2007 [39]
531 8-18 F: 287 M: 244
N Electronic scale, interview
Questions at interview
Upper and mid back pain
Y -
48
Negrini & Carabalona 2002 [41]
202 11.06±0.34
U N Q, balance
Q back pain Y -
Madras et al 1998 [36]
11 25±3.38 U Y BS Walking at 0, 5% & 10% grade wearing no pack, waist pack or BP
RPE Y 5 min for each trial
Iyer 2001 [18]
351 9-20.6 U N Q, BS, meter scale with a height rod
Q Shoulder and back pain
N -
Birrell & Haslam 2009 [6]
127 20.61±2.55†
F:29 M:98
N Comfort Q
Marching while carrying load
RPE Y 1 h
Birrell & Hooper 2007 [7]
18 21.2±1.4†
U N Q, interview
Q Upper limb Y 2 h
Korovessis et al 2004 [28]
3441 12±1.5 F:1816 M:1625
Y Q, digital electronic Mettler weightier
Walking for 2hr while students carried their personal BP
Dorsal pain and Low back pain
Y -
Haselgrove et al 2008 [16]
1202 14.1±0.2
F:587 M:615
N Q Q Back and neck pain
Y -
Korovessis et al 2005 [29]
1252 12-18 F:664 M:588
Y VAS, scale
Standing with extended knees & hips, keeping the arms close to the body with and without wearing a BP
Dorsal pain and Low back pain
Y -
Lockhart et al 2004 [34]
127 12-13 F:78 M:49
N Survey Q pain Y -
Bauer & Freivalds 2009 [4]
20 12.9±1 F:10 M:10
N BS Standing & walking while carrying 0, 10, 15 or 20% BW in a BP
RPE Y 3 min for each trial
Sheir-Neiss et al 2003 [47]
1122 12-18 U N Q, scale Q back pain Y -
Kirk & Schneider 1992 [26]
11 18-33 F N BS Walking while carrying 33% BW in an internal or external frame BP
RPE N 1 h
Talbott et al 2009 [53]
871 9-18 F:511 M:359
N Q Q Pain Y -
Al-Hazzaa 2006 [2]
702 6-14 M N Q , digital scale
Questions at interview
Shoulder pain
Y -
Whittfield et al 2005 [59]
140 17.1 F:70 M:70
N Electronic scale, Nordic Musculos-keletal Q
Q Upper body
Y -
Goodgold et al 2002 [13]
345 11-14 F:176 M:169
Y Q, scale Q Back pain Y -
Van Gent 2003 [54]
745 12-14 F:367 M:378
N Q, scale Q Neck, shoulder and back pain
Y -
Young et al 2006 [60]
184 11.2±1.9
M:76 F:108
N Q, scale Q Back pain Y -
Puckree et al 2004 [43]
176 12.2±0.8
U N Scale, Q Q Neck, shoulder and back comfort
Y -
Ling et al
2004§ [32]
7 24.5±3.4
F Y VAS, modified
Marching without load or carrying 20,
Discomfort U 1 h
49
BS 30, 40 & 50 pounds
Beekley et
al 2007§
[5]
10 32.4±1.3†
M Y BS Marching while carrying 30, 50 and 70% of lean body mass
RPE Y 30 min for each trial
Quesada et
al 2000§
[44]
12 22.4±2.3†
M Y BS Marching WL or while carrying 15 and 30% of BW
RPE Y 40 min for each trial
Goslin & Rorke
1986§ [14]
10 24.3±2.8
M N BS Walking while carrying 0, 20, 40% BW in a BP at 2 different speeds
RPE Y 10 min for each trial
Kennedy et
al 1999§
[24]
42 9.1±0.69
F:20 M:22
Y BS, VAS Walking under 5 conditions: WL or carrying 5, 10, 15, 20% of BW in a BP
RPE Y 5 min for each trial
LIoyd et al
2009§ [33] 32 22.3 F N VAS, BS,
Likert scale
Walking while carrying a variety of loads until pain led to voluntary cessation of the session or a load of 70% body mass was carried
RPE, pain and regional discomfort
Y 4 min for each trial
Johnson et
al 1995§
[22]
15 21-36† M N Environmental symptoms Q, 6-point scale
Walking while carrying 34, 48 or 61 kg in a standard army BP or prototype double pack
Tiredness, discomfort, back pain
Y Time needed to finish 20 km
BP, Backpack; BS, Borg Scale; BW, Body weight; H, Hour; Min, Minute; N, No; Q, Questionnaire; RPE, Rating of
Perceived Exertion; VAS, Visual Analogue Scale; WL, Without Load; Y, Yes; §, studies that assessed the
correlation of pain with increasing load; †, studies that were conducted on soldiers
Table 3: Results of the studies that assessed the correlation between
backpack use and pain
Author Correlation between backpack weight and pain
Comment
Siambanes et al 2004 Y Older students and those who walk to and from school experienced more pain
Navuluri & Navuluri 2006
Y Backpack pain was seen just among girls
Moore et al 2007 Y Younger student and girls are more at risk of experiencing pain
Haselgrove et al 2008 Y Pain increased in both genders but it was more prevalent between girls
Sheir-Neiss et al 2003 Y Girls and students with larger body mass index experienced more pain. Correlations between backpack weight and extent of using backpacks with pain
Talbott et al 2009 Y Besides weight, the amount of time carrying was also associated with pain
Van Gent et al 2003 Y Girls and younger students reported pain more often
Grimmer & Williams 2000
Y Pain was associated with weight, time spent carrying backpack, time spent sitting. Girls and younger students were more vulnerable. No correlation was
50
seen between backpack weight and body mass index
Puckree et al 2004 Y
Korovessis et al 2004 Y Girls experienced more pain than boys
Chiang et al 2006 N Association was reported between backpack carrying time and pain
Wall et al 2003 N Intensity of pain increased by backpack carrying
Negrini & Carabalona 2002
N Duration of backpack carrying but not backpack weight was associated with pain
Korovessis et al 2005 N Girls reported pain more often and with higher intensity
Goodgold et al 2002 N
Al-Hazza 2006 N Older students reported pain more often
Whittfield et al 2005 N No association between backpack weight and incidence of pain
Lyer 2001 N No association between backpack weight or age with pain
Young et al 2006 N No association between backpack weight and back pain. Association between age and back pain (pain was more common in older students)
N, No; Y, Yes
Table 4: Results of the studies that assessed the correlation between
backpack use and perceived exertion
Author Correlation between backpack weight and perceived exertion
Comment
Marsh et al 2006 Y Abdominal support decreased RPE
Madras et al 1998 Y
Kirk & Schneider 1992 Y Besides weight, the amount of time carrying was also associated with RPE
Bauer and Freivalds 2009
N Carry 10% bodyweight didn’t have influence on RPE
N, No; Y, Yes
The quality appraisal of the studies can be seen in table 5. Sample size
varied from seven to 3498 but just two of the studies justified their sample
sizes. Also only three articles declared that the person who carried out the
measurements was trained. Eight out of seventeen studies used calibrated
equipment and instruments and nine studies didn’t provide any detail on the
calibration of the equipment they used. Reasonable information and
description of the validity and reliability of equipment and instruments used
were reported in just seven and four studies, respectively. Also, one paper
provided information on the inter-tester and intra-tester reliability of the
51
method they used. The results of none of the papers could be generalised as
they just assessed a specific age range or sex.
Table 5: Quality appraisal of studies that assessed the correlation
between backpack use and pain
Authors, Publication Year
1 2 3 4 5 6 7 8 9 10 11 12
13 14
15 16
Total score
Marsh et al 2006
N Y U Y Y Y Y U Y Y Y N Y Y Y III-3A
11/15
Siambanes et al 2004
N Y U Y N N Y Y N Y Y N N Y Y IV 8/15
Navuluri & Navuluri 2006
N Y U N Y Y Y Y U Y Y N Y Y U IV 9/15
Chiang et al 2006
N Y U N N N Y Y Y Y Y N Y Y Y IV 8/15
Wall et al 2003
N Y U N/A
N N Y Y N N Y N N Y N IV 4/14
Grimmer & Williams 2000
Y Y Y Y N N Y N Y Y Y N Y Y Y IV 11/15
Moore et al 2007
N Y U N N N Y Y N Y Y N N Y Y IV 8/15
Negrini & Carabalona 2002
N Y U N Y N Y Y Y Y N N Y Y U IV 8/15
Madras et al 1998
N N U N/A
N N Y Y N N Y N Y Y Y III-2A
6/14
Iyer 2001 N Y Y Y N N Y N N Y N N N Y Y IV 7/15
Birrell & Haslam 2009
N Y U N/A
Y N Y Y Y Y Y N Y Y Y IV 10/14
Birrell & Hooper 2007
N Y U N/A
N N Y N N Y Y N N Y N IV 5/14
Korovessis et al 2004
N Y U N N N Y Y N N Y N Y Y Y IV 7/15
Haselgrove et al 2008
N N
U N/A
Y Y Y Y N Y Y N Y Y Y IV 9/14
Korovessid et al 2005
N Y U N N N Y Y Y Y Y N Y N Y IV 8/15
Lockhart et al 2004
N Y U N/A
N Y Y Y Y Y Y N Y Y Y IV 10/14
Bauer & Freivalds 2009
Y Y U N/A
N N Y N Y Y Y N Y Y Y III-2A
9/14
Sheir-Neiss et al 2003
N Y U Y Y N Y N N Y Y N Y Y Y IV 9/15
Kirk & Schneider 1992
N Y U N/A
N N Y N Y Y Y N Y Y Y III-3 8/14
Talbott et al 2009
N Y U N/A
N N Y U N Y Y N Y Y Y IV 7/14
Al-Hazzaa 2006
N Y U N N N Y N N N Y N Y Y Y IV 6/15
Whittfield et al 2005
N Y U Y N N Y U N Y Y N Y Y Y IV 8/15
Goodgold et al 2002
N N U N U N Y U Y Y Y N Y Y Y IV 7/15
Van Gent 2003
N N U N N N Y Y Y Y Y N Y Y Y IV 8/15
52
Young et al 2006
N Y U Y N N Y Y Y N/A
Y N Y Y N IV 8/14
Puckree et al 2004
N Y U Y Y U Y N Y Y Y N Y Y N IV 9/15
Ling et al 2004§
N N
U N/A
N N N Y N Y Y N Y Y Y III-2 6/14
Beekley et al 2007§
N Y U N/A
N N Y N Y N Y N N Y Y III-3A
6/14
Quesada et al 2000§
N Y U N/A
N N Y N Y N Y N Y Y Y III-2A
7/14
Goslin & Rorke 1986§
N Y U N/A
N N Y N Y Y Y N N Y N III-2A
6/14
Kennedy et al 1999§
N Y U N/A
N N Y N Y Y Y N N Y Y III-3A
8/15
LIoyd et al
2009§ N Y Y N/
A N N Y N Y Y Y N Y Y Y III-
2A 9/14
Johnson et
al 1995§ N Y U N/
A N N Y Y N Y Y N Y Y Y III-
2A 8/14
1. Justification of sample size; 2. Consistency in the number of subjects reported throughout the paper; 3. The person who carried out the measurement was trained; 4. Was the equipment/instrument calibrated before use; 5. Adequate description of the validity of the instrument/equipment; 6. Adequate description of the reliability of the instrument/equipment; 7.was the design appropriate for stated aims; 8. Weakness or limitations mentioned; 9. Interpretation of null findings; 10. Interpretations of important effects; 11. Comparison of results with previous reports; 12. Implication in real life/generalisability; 13. Adequate description of statistical method; 14. Adequate description of the data; 15. Assessment of statistical significance; 16. Type of experimental design and level of evidence ;§, studies that assessed the correlation of pain with increasing load; *, adequate description of reliability of the method was provided; III-2A, a comparative study with concurrent control (an internal control group) _phases randomisation; III-2, a comparative study with concurrent control (an internal control group); III-3A, a comparative study without concurrent control _phases randomisation; III-3, a comparative study without concurrent control; IV, a cross-sectional study
Seven studies assessed the correlation between pain, discomfort and
perceived exertion with increasing load. Ling et al 2004 reported that level of
discomfort increased as the load increased in adults [32]. Also, Beekley et al
2007 showed that perceived exertion was significantly higher while carrying
70% lean body mass (LBM) than 30% and 50% LBM in adults; however, no
differences in perceived exertion responses were seen between 50% and
30% LBM [5]. LIoyd et al 2009 observed that pain, perceived exertion and
regional discomfort increased with increasing load (from 10% to 70% of body
mass) in most of the body parts while some other parts such as chest, hips,
buttocks and feet only showed significant changes between 15% and 20%
body mass load [33]. Quesada et al 2000 stated that 0 and 15% body weight
load produced similar results of RPE but subjects perceived the work to be
harder during carrying 30% bodyweight [44]. Goslin and Rorke 1986 also
reported that there is a linear relationship between perceived exertion and
increase in the amount of load [14]. Johnson et al 1995 assessed discomfort
when soldiers carried 34, 48 or 61 kg loads in a backpack and double pack. It
53
was reported that as load increased, discomfort soared [22]. In the only study
that investigated adolescents, Kennedy 1999 declared a rise in perceived
difficulty with increasing weight [24]. Descriptive characteristics and quality
appraisal of these studies can be seen in table 2 and 5.
Based on these studies the weight of the evidence suggests that there is a
correlation between backpack weight and pain, perceived exertion or
discomfort. However, other factors such as gender, age and duration of
carrying load can also influence these variables. Further, as the load
increases the level of pain, perceived exertion or discomfort raises but the
beginning point of the pain can be different in various conditions and between
different subjects.
Determining the effect of load placement on pain
Three studies assessed the effect of load placement on pain, perceived
exertion and discomfort. Stuempfle et al 2004 compared the effect of load
placement on perceived exertion in female adults and it was shown that high
back load placement could lead to less perceived exertion compared to mid
or low back load placement [52]; on the other hand, Brackley et al 2009 and
Devroey et al 2007 reported that load placement did not have influence on
perceived exertion in adolescents and adults, respectively [8,12]. Table 6 and
7 shows the descriptive characteristics and quality appraisal of these papers,
respectively.
Table 6: Descriptive characteristics of studies that examined the effect
of load placement on pain
Au
tho
rs, P
ub
licat
ion
Ye
ar
Sam
ple
siz
e
Age
(ran
ge o
r M
ean
± S
D)
Ge
nd
er
Incl
usi
on
an
d e
xclu
sio
n
crit
eri
a
Inst
rum
en
ts o
f
me
asu
rem
en
t
Task
Co
mp
licat
ion
s o
f w
ear
ing
bac
kpac
k h
ave
be
en
rep
ort
ed
on
Cle
arly
sta
ted
aim
s
Du
rati
on
of
carr
yin
g
bac
kpac
k
Stuempfle et al 2004 [52]
10 18-22 F N BS Walking while carrying 25% BW in a high, central or low back position
RPE Y 10 min for each trial
Brackley et al 2009 [8]
15 U (grade 5 student
F:10 M:5
U Q, body map diagram, Wong-Baker Faces pain
Carrying 15% of BW with the centre of BP located on high, mid or low back
RPE N Time needed to finish 1000 m walk
54
s) scale
Devroey et al 2007 [12]
20 23.9±2.59
F:8 M:12
N BS Standing and walking without pack and with 5, 10 and 15% BW
RPE and discomfort
N Static: 1 min Dynamic : 5 min
Refer to table 2
Table7: Quality appraisal of studies that examined the effect of load
placement on pain
Authors, Publication Year
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Total score
Stuempfle et al 2004
N Y U N/A N N Y N N N Y N Y Y Y III-3A
6/14
Brackley et al 2009
N Y U N/A N N Y Y N N Y N N N Y III-3A
5/14
Devroey et al 2007
N N U N/A N N Y Y Y N Y N Y Y Y III-2A
7/14
1-16, Refer to the legend of table 5
The results of this section are inconclusive so it is not possible to conclude
what is the best placement of backpacks on the spine.
Can different designs of backpacks reduce the discomfort?
Ten studies compared the effect of different designs and features of
backpacks on pain, discomfort and perceived exertion. Descriptive
characteristics and quality appraisal of these studies can be seen in table 8
and 9, respectively. Eight out of 10 studies investigated adults and just two
studies were conducted on children. Five studies examined males
exclusively. Only Bauer & Freivalds 2008 justified the sample size they
examined.
Table 8: Descriptive characteristics of studies that examined if different
designs can reduce pain, perceived exertion and discomfort
Au
tho
rs, P
ub
licat
ion
Y
ear
Sam
ple
siz
e
Age
(ra
nge
or
Mea
n ±
SD)
Ge
nd
er
Incl
usi
on
an
d e
xclu
sio
n
crit
eri
a
Inst
rum
ents
of
me
asu
rem
en
t
Task
Co
mp
licat
ion
s o
f w
ear
ing
bac
kpac
ks
hav
e b
ee
n r
ep
ort
ed
on
cle
arly
sta
ted
aim
s
Du
rati
on
of
carr
yin
g
bac
kpac
k
Jacobson & Jones 2000 [20]
20 24.3±3.6 20 N BS, 7 point scale
Walking with 16 obstacles while wearing BPs
comfort Y Time needed to finish 30 m
55
Jacobson et al 2003 [19]
21 20.4±1.41 F: 16 M:5
Y VAS Carrying different backpacks for the whole day
comfort Y 10 days
Jacobson et al 2004 [21]
19 22±1.36 M:10 F:9
N VAS, anatomical illustration rating scale
Carrying 2 BPs for the whole day
Neck, shoulder, back and overall comfort
Y 10 days
Holewijn & Lotens 1992 [17]
10 NM†(adult) M N Q Marching with 10 different BPs
RPE N 1 h
Bauer & Freivalds 2008 [3]
20 11-14 F:10 M:10
Y BS Standing & walking WL, wearing a standard BP & a BP with additional comfort features
RPE Y 3 min for each trial
Southward & Mirka 2006 [50]
15 21-55 M:12 F:3
N Likert scale Bending forward while wearing different BPs
comfort Y Time needed to gradually bending forward until Reaching the designated angle
Knapik et al 1997 [27]
15 29.7±4.3† M Y soreness and discomfort Q, scale
Marching while carrying BP & double pack
discomfort
Y Time needed to finish 20 km
Mackie et al 2003 [35]
12 12.6±1.1 F:6 M:6
N VAS, Q, BS, scale rating method, musculoskeletal discomfort diagram
Walking while wearing each of 4 BPs
RPE and discomfort
Y 20 min for each BP
Legg et al 1997 [31]
10 22.5±6.3 M N regional body diagram, category ratio scale, Q, VAS
Walking while carrying 20 kg in 2 different BPs
Perceived discomfort
Y 30 min for each BP
Legg et al 2003 [30]
10 30.8±11.3 M N regional body diagram, category ratio scale, Q, VAS
Walking while carrying 15 kg in 2 different BPs
Perceived discomfort
Y 15 min for each BP
Refer to table 2
Table 9: Quality appraisal of studies that examined if different designs
can reduce the pain
Authors, Publication Year
1 2 3 4 5 6 7 8 9 10 11
12
13
14
15 16 Total score
Jacobson & Jones 2000
N Y U N/A
N N Y Y N N Y N N N N III-3A
4/14
Jacobson et al 2003
N N U N/A
N N Y Y N/A
Y Y N N Y N III-3A
5/13
Jacobson et al 2004
N Y U N/A
N N Y Y N Y Y N Y Y Y III-3A
8/14
Holewijn & Lotens 1992
N Y U N/A
N N Y N N N Y N Y Y N III-3 5/14
Bauer & Freivalds 2008
Y Y U N/A
N N Y N Y N N N Y N N III-2A
5/14
Southward & Mirka 2006
N Y U N/A
N N Y Y N/A
N N N Y Y Y III-3 6/13
Knapik et al 1997
N N U Y N N N N N N Y N Y N Y III-3 4/15
Mackie et al N Y U N/ N N N N N Y Y N Y Y Y III-3 6/14
56
2003 A Legg et al 1997
N Y U N/A
N N Y N Y N/A
Y N N Y Y III-3 6/13
Legg et al 2003
N Y U N/A
N N Y N Y Y Y N Y Y Y III-3A
8/14
1-16, Refer to the legend of table 5
Jacobson and Jones 2000 reported that there was no significant difference in
level of comfort between internal frame and external frame backpacks [20].
Jacobson et al 2003 compared the comfort level of an ordinary backpack with
an experimental backpack which had a slanting shelving system and
distributed the weight vertically in adults; more local and overall comfort was
reported by using this system [19]. Moreover, Jacobson et al 2004 compared
the regional and overall comfort of subjects’ personal backpacks and an
experimental backpack. No significant differences in the comfort of
backpacks was seen on a Visual Analogue Scale; however, the experimental
backpack was more comfortable for the back on an Anatomical Illustration
Rating Scale [21]. Southward and Mirka 2007 compared the effect of a basic
and an advanced backpack harness system (a backpack which had lateral
stiffness rods) on comfort in adults. It was shown that the advanced design
which could distribute the weight between shoulders and hips can provide
more local and overall comfort [50]. Knapik et al 1997 assessed the effect of
backpack and double pack. They reported that double pack caused less
discomfort in low back, lower incidence of blisters, but it resulted in pain in
neck & hips and it took longer to complete the march with wearing the double
pack [27]. Also, Mackie et al 2003 evaluated the influence of four backpacks
on perceived exertion and discomfort. In this study a backpack which had two
major compartments, back padding and side compression straps became the
students’ most favoured one [35]. Bauer and Freivalds 2008 evaluated the
impact of two backpacks with different comfort features on perceived exertion
and it was shown that additional comfort features could not provide less
perceived pain [3]. Holewijn and Lotens 1992 compared the effect of different
carrying modes on perceived exertion; they concluded that carrying the same
amount of load in a backpack can cause more perceived exertion than waist
carrying mode [17].
57
The evidence is conflicting but on balance we can conclude that backpack
designs that distribute load between the shoulders and hips or between the
front and back of the body provides more local and overall comfort.
Discussion
This study is the first comprehensive systematic review looking at load
carriage systems and pain, discomfort and perceived exertion. The results of
this systematic review show significant variability in the design and study
populations of studies. This variability prevented any meaningful statistical
pooling of data. However a qualitative synthesis was feasible.
While studies were of various designs, none were randomised controlled
trials and there were other widespread deficiencies in the validity, reliability
and calibration of equipment and instruments of measurement. These
instruments of measurement factors are fundamental to producing
meaningful scientific evidence; therefore, we recommend more rigour and
explanation in trial design and selection and in the use of reliable and valid
instruments for measuring the influence of backpack design on the body.
Moreover, providing training sessions for subjects and examiners may have
also influence the validity and reliability of the study.
Sample size was just justified in only two papers. A study with a small sample
size may not detect significant results. Also there is a chance of random error
and publication bias in small studies, because interesting and favourable
results from small studies might be reported whereas less interesting findings
from small studies remain unreported [11]. If the sample size is too large
there are ethical implications in wasting participant’s time and in some cases
putting them at risk.
In this review, a scoring system was not used and studies were not labelled
by low, moderate or high quality; instead, trials were classified with the
notation of how many appraisal criteria they satisfied and were assessed for
every single item separately. It is worth noting that some studies might be
strong in some parts but poor in other aspects. The number of criteria
58
satisfied clearly reflects whether the study should be regarded as having a
high risk of bias.
Of the 26 studies that examined the correlation between backpack carrying
or backpack weight and pain, discomfort or perceived exertion, 13 trials
declared that there is a significant positive correlation between these two
factors while 10 studies were of the opinion that there is no association
between these two variables. Of particular note was the heterogeneity among
studies with respect to study populations, participants’ age range and gender,
type of the study design, task of the participants during the experiment,
habitual differences and outcome measurements. Due to this diversity it was
not possible to perform statistical pooling of the data. The strength of
evidence of each paper was assessed by the quality appraisal tool by
Crombie. Most of the papers that were in favour of the correlation between
backpack use and pain, perceived exertion and discomfort had qualities in
the range of 6/14 to 11/15 (see table 5) and the trials that didn’t support this
correlation showed qualities in the range of 4/14 to 10/14.
From this review it became apparent that factors other than backpack weight
can generate pain, perceived exertion or discomfort; the reported factors
were gender [15,16,39,40,43,54], age and grade in school [2,15,39,48,54],
subject’s body mass index [5,47], the amount of time using the backpack and
walking to and from school [10,41,47,48,53]. Girls experience more pain that
boys, this could be because, boys have a stronger musculoskeletal system,
also they might have higher threshold of pain based on differences between
physiological and psychological factors between genders [46].
Mostly it was thought that load carrying provokes low back pain but this
review reveals that just less than half of the studies in this review reported
feeling low back pain. It should be noted that other complications such as
neck pain, thoracic pain, shoulder pain, upper limb discomfort, overall
discomfort and perceived exertion are also frequent.
Of the seven studies that examined the effect of increasing load on perceived
exertion, it was shown that increasing weight provokes higher intensity of
pain and exertion; however, different load thresholds as the start point of
59
feeling discomfort and fatigue were reported in these studies. Subjects start
to notice differences in sensation of effort at different load thresholds and
also the level of pain and discomfort threshold varies among individuals. It
seems that age, gender, the circumstances of the load carrying experience
and profession are factors that have an effect on the load threshold [5].
Three studies assessed the effect of load placement on pain, perceived
exertion and discomfort but it is not possible to find out where the load should
be placed in order to reduce the pain. It is hard to draw a conclusion as these
three trials examined different age groups. These studies scored 6/15, 5/14
and 7/14.
Conclusion
The results of this review show that there is conflicting evidence on the
correlation between load carrying and experiencing pain, exertion and
discomfort during different stages of life. However, based on this conflicting
evidence we can say that carrying loads does not always provoke low back
pain; and that it may trigger neck, thoracic or shoulder pain. In addition the
physiological and psychological status of individuals can intensify or reduce
the level and threshold of perceived pain.
Also, there is limited evidence on the effect of load positioning and various
designs on level of perceived pain and exertion. It seems that so far none of
these changes could be helpful in reducing the complications of wearing
backpacks. Moreover, the methodological and quality assessments showed
that most of the included studies in this review were not strong enough and
could not be relied upon. The most commonly identified methodological
deficiencies were the lack of justification of sample size, providing training
sessions for examiners, utilising calibrated, valid and reliable instruments for
measurement.
There are a number of limitations to the current study. This review was not a
totally blind review; authors and publication details were disclosed to the
reviewers and this can potentially lead to reviewer bias. However, reviewers
were not aware of the background and previous works of the authors. A
60
further limitation is that although the search strategy was comprehensive it is
possible that some studies were not found. Indeed two studies that were
identified could not be located and one study was not obtainable in readable
form. Also, the validity and reliability of the critical appraisal tool used in this
study has not been established but was developed from first principles using
previously developed tools from related areas. Although the modified
Crombie instrument has face validity, further research is needed to assess its
validity and reliability. The suggestion of potential bias in studies using the
number of quality appraisal variables achieved is controversial and readers
are invited to use this as a guide only.
Acknowledgment
The authors thank Dr Navid Moheimani, Marly Walker and Wen Qi Ng for
their independent review of the studies.
References
[1] T.N.H.a.M.R.C. (NHMRC), NHMRC additional levels of evidence and
grades for recommendations for developers of guidelines, in, 2008, p. Stage
2 consultation.
[2] H.M. Al-Hazzaa, School backpack: How much load do Saudi schools
boys carry on their shoulders?, Saudi Medical Journal 27 (2006), 1567-1571.
[3] D.H. Bauer and A. Freivalds, A comparison of two middle school
backpacks with varying comfort features under an increasing load, in: Human
factors and ergonomics society annual meeting proceedings, Human factors
and ergonomics society, 2008, pp. 1775-1779.
[4] D.H. Bauer and A. Freivalds, Backpack load limit recommendation for
middle school students based on physiological and psychophysical
measurements, Work 32 (2009), 339-350.
[5] M.D. Beekley, J. Alt, C.M. Buckley, M. Duffey, and T.A. Crowder,
Effects of heavy load carriage during constant-speed, simulated, road
marching, Military Medicine 172 (2007), 592-595.
[6] S.A. Birrell and R.A. Haslam, Subjective skeletal discomfort measured
using a comfort questionnaire following a load carriage exercise, Military
Medicine 174 (2009), 177-182.
[7] S.A. Birrell, R.H. Hooper, and R.A. Haslam, The effect of military load
carriage on ground reaction forces, Gait and Posture 26 (2007), 611-614.
61
[8] H.M. Brackley, J.M. Stevenson, and J.C. Selinger, Effect of backpack
load placement on posture and spinal curvature in prepubescent children,
Work 32 (2009), 351-360.
[9] A.M. Briggs, A.J. Smith, L.M. Straker, and P. Bragge, Thoracic spine
pain in the general population: prevalence, incidence and associated factors
in children, adolescents and adults. A systematic review, BMC
Musculoskeletal disorders 10 (2009), 77.
[10] H.Y. Chiang, K. Jacobs, and G. Orsmond, Gender-age environmental
associates of middle school students' low back pain, Work 26 (2006), 19-28.
[11] I.K. Crombie, The pocket guide to critical appraisal BMJ publishing
group, London, 1996.
[12] C. Devroey, I. Jonkers, A. de Becker, G. Lenaerts, and A. Spaepen,
Evaluation of the effect of backpack load and position during standing and
walking using biomechanical, physiological and subjective measures,
Ergonomics 50 (2007), 728-742.
[13] S. Goodgold, M. Corcoran, D. Gamache, J. Gillis, J. Guerin, and J.Q.
Coyle, Backpack use in children, Pediatric Physical Therapy 14 (2002), 122-
131.
[14] B.R. Goslin and S.C. Rorke, The perception of exertion during load
carriage, Ergonomics 29 (1986), 677-686.
[15] K. Grimmer and M. Williams, Gender-age environmental associates of
adolescent low back pain, Applied Ergonomics 31 (2000), 343-360.
[16] C. Haselgrove, L. Straker, A. Smith, P. O'Sullivan, M. Perry, and N.
Sloan, Perceived school bag load, duration of carriage, and method of
transport to school are associated with spinal pain in adolescents: An
observational study, Australian Journal of Physiotherapy 54 (2008), 193-200.
[17] M. Holewijn and W.A. Lotens, The influence of backpack design on
physical performance, Ergonomics 35 (1992), 149-157.
[18] S.R. Iyer, An Ergonomic Study of Chronic Musculoskeletal Pain in
Schoolchildren, Indian Journal of Pediatrics 68 (2001), 937-941.
[19] B.H. Jacobson, D.A. Cook, T.S. Altena, H.A. Gemmell, and B.M.
Hayes, Comparison of perceived comfort differences between standard and
experimental load carriage systems, Ergonomics 46 (2003), 1035-1041.
[20] B.H. Jacobson and K. Jones, Comparison of selected perceptual
variables for backpacks with internal and external frames, Perceptual and
Motor Skills 90 (2000), 605-608.
62
[21] B.H. Jacobson, B. Oney, B. Redus, B. Edgley, F. Kulling, and H.A.
Gemmell, Comparison of perceived backpack comfort using two rating
scales, Perceptual and Motor Skills 99 (2004), 171-178.
[22] R.F. Johnson, J.J. Knapik, and D.J. Merullo, Symptoms during load
carrying: effects of mass and load distribution during a 20-km road march,
Perceptual and Motor Skills 81 (1995), 331-338.
[23] P. Juni, A. Witschi, R. Bloch, and M. Egger, The hazard of scoring the
quality of clinical trials for meta-analysis, JAMA 282 (1999), 1054-1060.
[24] L. Kennedy, F. Lawlor, A. O'Connor, S. Dockrell, and J. Gormley, An
investigation of the effects of schoolbag carriage on trunk angle,
physiotherapy Ireland 20 (1999), 2-8.
[25] A.Q. Khan, M.S. Nazir, M.Z. Iqbal, and M.M. Khan, Body ache and
pain in school children associated with school bag carriage in Multan -
Pakistan, Medical Forum Monthly 18 (2007), 4-9.
[26] J. Kirk and D.A. Schneider, Physiological and perceptual responses to
load-carrying in female subjects using internal and external frame backpacks,
Ergonomics 35 (1992), 445-455.
[27] J.J. Knapik, P. Ang, H. Meiselman, W. Johnson, J. Kirk, C. Bensel,
and W. Hanlon, Soldier performance and strenuous road marching: Influence
of load mass and load distribution, Military Medicine 162 (1997), 62-67.
[28] P. Korovessis, G. Koureas, and Z. Papazisis, Correlation between
Backpack Weight and Way of Carrying, Sagittal and Frontal Spinal
Curvatures, Athletic Activity, and Dorsal and Low Back Pain in
Schoolchildren and Adolescents, Journal of Spinal Disorders and Techniques
17 (2004), 33-40.
[29] P. Korovessis, G. Koureas, S. Zacharatos, and Z. Papazisis,
Backpacks, back pain, sagittal spinal curves and trunk alignment in
adolescents: A logistic and multinomial logistic analysis, Spine 30 (2005),
247-255.
[30] S.J. Legg, A. Barr, and D.I. Hedderley, Subjective perceptual methods
for comparing backpacks in the field, Ergonomics 46 (2003), 935-955.
[31] S.J. Legg, L. Perko, and P. Campbell, Subjective perceptual methods
for comparing backpacks, Ergonomics 40 (1997), 809-817.
[32] W. Ling, V. Houston, Y.S. Tsai, K. Chui, and J. Kirk, Women's load
carriage performance using modular lightweight load-carrying equipment,
Military Medicine 169 (2004), 914-919.
63
[33] R. Lloyd, B. Parr, S. Davies, and C. Cooke, Subjective perceptions of
load carriage on the head and back in Xhosa women, Applied Ergonomics 41
(2009), 522-529.
[34] R.A. Lockhart, K. Jacobs, and G. Orsmond, Middle school children's
participation in activities and the effects of pain from backpack use on
participation, Work 22 (2004), 155-168.
[35] H.W. Mackie, S.J. Legg, J. Beadle, and D. Hedderley, Comparison of
four different backpacks intended for school use, Applied Ergonomics 34
(2003), 257-264.
[36] D.E. Madras, M.W. Cornwall, and J.R. Coast, Energy cost, perceived
exertion and postural adjustments when treadmill walking with two types of
backpack, Journal of Human Movement Studies 35 (1998), 233-249.
[37] A.B. Marsh, L. DiPonio, K. Yamakawa, S. Khurana, and A.J. Haig,
Changes in posture and perceived exertion in adolescents wearing
backpacks with and without abdominal supports, American Journal of
Physical Medicine and Rehabilitation 85 (2006), 509-515.
[38] S. Masiero, E. Carraro, A. Celia, D. Sarto, and M. Ermani, Prevalence
of nonspecific low back pain in schoolchildren aged between 13 and 15
years, Acta Paediatrica, International Journal of Paediatrics 97 (2008), 212-
216.
[39] M.J. Moore, G.L. White, and D.L. Moore, Association of relative
backpack weight with reported pain, pain sites, medical utilization, and lost
school time in children and adolescents, Journal of School Health 77 (2007),
232-239.
[40] N. Navuluri and R.B. Navuluri, Study on the relationship between
backpack use and back and neck pain among adolescents, Nursing and
Health Sciences 8 (2006), 208-215.
[41] S. Negrini and R. Carabalona, Backpacks on! Schoolchildren's
perceptions of load, associations with back pain and factors determining the
load, Spine 27 (2002), 187-195.
[42] S. Negrini, E. Politano, R. Carabalona, I. Tartarotti, and M.L.
Marchetti, The backpack load in schoolchildren: Clinical and social
importance, and efficacy of a community-based educational intervention: A
prospective controlled cohort study, Europa Medicophysica 40 (2004), 185-
190.
[43] T. Puckree, S.P. Silal, and J. Lin, School bag carriage and pain in
school children, Disability and Rehabilitation 26 (2004), 54-59.
64
[44] P.M. Quesada, L.J. Mengelkoch, R.C. Hale, and S.R. Simon,
Biomechanical and metabolic effects of varying backpack loading on
simulated marching, Ergonomics 43 (2000), 293-309.
[45] M.F. Reneman, B.J.J. Poels, J.H.B. Geertzen, and P.U. Dijkstra, Back
pain and backpacks in children: Biomedical or biopsychosocial model?,
Disability and Rehabilitation 28 (2006), 1293-1297.
[46] J.J. Salminen, J. Pentti, and P. Terho, Low back pain and disability in
14-year-old schoolchildren, Acta Paediatrica, International Journal of
Paediatrics 81 (1992), 1035-1039.
[47] G.I. Sheir-Neiss, R.W. Kruse, T. Rahman, L.P. Jacobson, and J.A.
Pelli, The association of backpack use and back pain in adolescents, Spine
28 (2003), 922-930.
[48] D. Siambanes, J.W. Martinez, E.W. Butler, and T. Haider, Influence of
School Backpacks on Adolescent Back Pain, Journal of Pediatric
Orthopaedics 24 (2004), 211-217.
[49] D.L. Skaggs, S.D. Early, P. D'Ambra, V.T. Tolo, and R.M. Kay, Back
pain and backpacks in school children, Journal of Pediatric Orthopaedics 26
(2006), 358-363.
[50] S.A. Southard and G.A. Mirka, An evaluation of backpack harness
systems in non-neutral torso postures, Applied Ergonomics 38 (2007), 541-
547.
[51] E. Steele, A. Bialocerkowski, and K. Grimmer, The postural effects of
load carriage on young people - a systematic review, BMC Musculoskeletal
disorders 4 (2003), 12.
[52] K.J. Stuempfle, D.G. Drury, and A.L. Wilson, Effect of load position on
physiological and perceptual responses during load carriage with an internal
frame backpack, Ergonomics 47 (2004), 784-789.
[53] N.R. Talbott, A. Bhattacharya, K.G. Davis, R. Shukla, and L. Levin,
School backpacks: It's more than just a weight problem, Work 34 (2009),
481-494.
[54] C. Van Gent, J.J.C.M. Dols, C.M. De Rover, R.A. Hira Sing, and
H.C.W. De Vet, The weight of schoolbags and the occurrence of neck,
shoulder, and back pain in young adolescents, Spine 28 (2003), 916-921.
[55] P. Viry, C. Creveuil, and C. Marcelli, Nonspecific back pain in children:
A search for associated factors in 14-year-old schoolchildren, Revue du
Rhumatisme (English Edition) 66 (1999), 381-388.
65
[56] B.F. Walker, The prevalence of low back pain: A systematic review of
the literature from 1966 to 1998, Journal of Spinal Disorders 13 (2000), 205-
217.
[57] E.J. Wall, S.L. Foad, and J. Spears, Backpacks and back pain:
Where's the epidemic?, Journal of Pediatric Orthopaedics 23 (2003), 437-
439.
[58] P. Whiting, R. Harbord, and J. Kleijnen, No role for quality scores in
systematic reviews of diagnostic accuracy studies, BMC Medical Research
Methodology 5 (2005), 19.
[59] J. Whittfield, S.J. Legg, and D.I. Hedderley, Schoolbag weight and
musculoskeletal symptoms in New Zealand secondary schools, Applied
Ergonomics 36 (2005), 193-198.
[60] I.A. Young, A.J. Haig, and K.S. Yamakawa, The association between
backpack weight and low back pain in children, Journal of Back and
Musculoskeletal Rehabilitation 19 (2006), 25-33.
66
CHAPTER FOUR
THE RELIABILITY OF A PORTABLE CLINICAL FORCE PLATE
USED FOR THE ASSESSMENT OF STATIC POSTURAL
CONTROL: REPEATED MEASURES RELIABILITY STUDY
Published as
Golriz S, Hebert JJ, Foreman KB, Walker BF. (2012)
Chiropractic and Manual Therapies, May 23;20(1):14.
67
To assess the effect of fastening a hip belt of a backpack and to study load
placement within a backpack on postural stability we needed to use a reliable
and valid instrument of measurement. Measurement of postural stability is
usually obtained by assessment of centre of pressure using a force plate. We
chose to use a Midot posture scale analyser (MPSA) which was the only
available force plate to us in our research laboratory at the School of
Chiropractic and Sports Sciences. However, the reliability and validity of this
instrument had not been studied. Therefore, we decided to assess its test-
retest reliability in this chapter and concurrent validity in the next chapter.
The work resulted in the following publication:
Golriz S, Hebert JJ, Foreman KB, Walker BF. The reliability of a portable
clinical force plate used for the assessment of static postural control:
repeated measures reliability study. Chiropractic and Manual Therapies.
2012. May 23;20(1):14.
68
The reliability of a portable clinical force plate used for the assessment
of static postural control: repeated measures reliability study
Samira Golriz1* * Corresponding author
Email: [email protected]
Jeffrey J Hebert2
Email: [email protected]
K Bo Foreman3
Email: [email protected]
Bruce F Walker2
Email: [email protected]
1 School of Chiropractic and Sports Science, Murdoch University, 90 South
Street, Murdoch, WA 6150, Australia
2 School of Chiropractic and Sports Science, Murdoch University, Perth, WA,
Australia
3 Department of Physical Therapy, University of Utah, Salt Lake City, UT,
USA
Human research ethics committee of Murdoch University approved the
research protocol of this study.
69
Abstract
Background: Force plates are frequently used for postural control
assessments but they are expensive and not widely available in most clinical
settings. Increasingly, clinicians are using this technology to assess patients,
however, the psychometric properties of these less sophisticated force plates
is frequently unknown. The purposes of the study were to examine the test-
retest reliability of a force plate commonly used by clinicians and to explore
the effect of using the mean value from multiple repetitions on reliability.
Methods: Thirty healthy volunteer adults were recruited. Postural control
measures were obtained using the Midot Posture Scale Analyzer (MPSA).
Data were collected in 2 sessions. Five successive repetitions each of 60
seconds duration were obtained from each participant in each session.
Results: The reliability coefficients obtained using single measures were low
(ICC3,1 = 0.06 to 0.53). The average of two measures allowed for reliable
measurements of COP mean velocity and average location of COP. The
average of three and five measures was required to obtain acceptable
reliability (ICC ≥ 0.70) of relative weight bearing on legs and sway area,
respectively. Higher measurement precision values were seen by averaging
four or five repetitions for all variables.
Conclusion: Single measures did not provide reliable estimates of postural
sway, and the averaging of multiple repetitions was necessary to achieve
acceptable levels of measurement error. The number of repetitions required
to achieve reliable data ranged from 2 to 5. Clinicians should be wary of
using single measures derived from similar equipment when making
decisions about patients.
Keywords: Reproducibility, Posture, Stability, Balance, Force plate
70
Introduction
Postural control organises the orientation and equilibrium of the body during
upright stance and is essential to the successful performance of daily
movements and activities as well as fall prevention [1]. Postural control
depends on visual, vestibular and proprioceptive input and can be disrupted
by various perturbations experienced in everyday life [2,3]. Moreover,
pathology, medications, alcohol consumption, and the aging process can
adversely affect postural control [4,5] .
Postural control can be measured subjectively or objectively. Subjective
measures of postural control are obtained through the use of questionnaires.
Such questionnaires provide valuable information, however they often have
limitations with some special populations such as the elderly or individuals
with specific physical or cognitive impairment [6,7]. In addition, subjective
methods of measurement may suffer from floor or ceiling effects or lack
optimal reliability, validity and the precision to detect small differences [8,9]. It
has been suggested that these questionnaires be used in combination with
other measures [9].
Objective assessments are the most common method of measuring postural
control. Postural control is usually evaluated by interpretation of centre of
pressure (COP), postural sway and weight bearing distribution
measurements [10,11]. COP estimates the position of the ground reaction
force vector on the base of support of the body and calculation of the total
COP kinematics is frequently used to assess postural sway [12]. Postural
sway is an indicator of the displacement and correction of the centre of
gravity in relation to the base of support [11]. Relative weight bearing is a
gross estimate of postural control. Postural control is influenced by weight
bearing distribution and weight bearing distribution asymmetry results in less
postural stability [10].
Force plates are frequently used to measure COP and postural sway [11].
This approach requires the individual to stand or walk, while transducers
measure ground reaction forces generated by the body. As a clinical tool,
force plates are utilised therapeutically [13], and for longitudinal assessment
71
[14]. Force plates can be used to enhance balance training by providing
visual feedback to the patient [13]. However, the types of force plates used
by clinicians typically lack the sophistication of force plates used in research
environments. Moreover, well known force plates used in research have
known psychometric properties such as reliability and validity, while the
psychometric properties of force plates used in the clinical setting are
frequently unknown or poorly defined.
The Midot posture scale analyser (MPSA) is an example of less sophisticated
force plate used by clinicians such as physical therapists, chiropractors, and
neurologists [15] (This issue has been addressed in the instrument section
where we outlined the specifications of the MPS). The MPSA is relatively
inexpensive and includes simple and easy to use software that allows
individuals with limited experience to obtain and interpret measures of
postural control and weight bearing distribution.
However, when making decisions about patients, it is logical to rely on
assessments obtained by instruments shown to be reliable and valid.
Reliability is a prerequisite of validity and reflects measurement consistency
and the degree to which an instrument is free from errors of measurement
[16]. When clinicians obtain measurements to inform clinical decision making,
it is necessary that the instruments demonstrate adequate reliability.
Therefore, the purpose of this study was to examine the test-retest reliability
of a force plate commonly used by clinicians. Additionally, we explored the
effect of using the mean value from multiple repetitions on reliability.
Methods
Participants
Participant recruitment was by way of posted advertisements on University
bulletin boards. Potential participants were adults between 18 to 60 years of
age. A lower age limit of 18 years was chosen to help ensure maturity of the
skeletal system [17]. Postural control decreases in the elderly [18], therefore
60 years was considered the upper age limit.
72
Participants were excluded from the study if they had a history of
musculoskeletal injury in the previous three months, a balance deficit
stemming from a rheumatologic or neurologic disorder, pregnancy, an ear
infection or fever within 72 hours of testing or were currently taking
medications that could alter sensory perception (see appendices of chapter
4, page 163). The Human Research Ethics Committee of Murdoch University
approved the study protocol (2010/139), and all subjects gave written
consent before enrolling in the study. The rights of all subjects were
protected.
Instrument
The Midot posture scale analyser (MPSA)- (QPS-200, Midot Medical
Technology, Shekel Electronic Scale, Israel) is a portable force plate
consisting of 4 electronic weighing plates set in a rectangular position (Figure
1). Analogue signals were sampled and transferred to a laptop computer via
a USB-to-serial (RS-232) analogue to digital converter. The MPSA
specifications report an acquisition sampling frequency of 200 Hz and cut-off
frequency of 0.5 Hz [15].
The following specifications causes the MPSA to be less sophisticated: a) it
only measures two dimensional applied forces (Fx, Fy) and moments (Mx,
My) applied in balance; b) it can only be used for assessment of balance and
not gait; c) it uses weighing plates to measures balance instead of sensors
which are used in other force plates.
Figure 1 The Midot Posture Scale Analyser
The MPSA is designed to measure, during quiet standing: (1) the average
location of COP with reference to the cross point of the weighing platform
73
(mm), (2) COP mean velocity, which is the average velocity of COP
movement (mm/sec), (3) sway area, which is the area of an ellipse enclosing
95% of COP movement (mm2) and was measured by calculation of the
ellipse area, and (4) the relative weight distribution between the subject’s
right and left sides (%). As the purpose of this study was to assess the test
retest reliability of the MPSA, all those variables were assessed.
Sway area was calculated using the formula: 𝐴𝑃 𝑆𝐷∗𝑀𝐿 𝑆𝐷∗𝜋
4; Where AP SD
was the long axis of the ellipse and ML SD was the short axis of the ellipse.
Procedure
Calibration of the force plate was conducted each day prior to data collection
based on the manufacturer’s instructions. We tested all subjects under the
same conditions. Participants were asked to wear loose fit comfortable
clothing. Participants stood with their feet shoulder width apart on a sheet of
paper placed on the top of the platform. The paper remained in place during
testing. Both feet were outlined to ensure consistent placement across trials.
Participants were asked to remove their shoes, stand upright on the force
plate and remain as still as possible in a relaxed posture. We asked
participants to put their arms to their sides in a comfortable position and to
distribute their body weight evenly on both feet while breathing normally.
Finally, the participants were instructed to look straight ahead at an “X” on
the opposite wall located 2 meters away at eye level. If the patient usually
wore glasses, they continued to do so during this procedure.
Subjects were scheduled for 2 sessions of 5 trials, 5 minutes apart. The
second session replicated the first. The duration of each trial was 60
seconds. A 60-second assessment was chosen to mimic constituent periods
of standing during typical activities of daily living (e.g., waiting for a bus or
elevator). To avoid inconsistencies in the data at transitions, we informed the
participants of the data collection start time 5 seconds before the actual start
time. Mandatory breaks of 1 minute were allocated between each individual
trial during which subjects were allowed to sit. Breaks assured that
participants were refreshed for each trial. Five successive trials were
74
recorded during each session. On average, the overall duration of the
experiment was approximately 25 minutes.
Statistical analysis
Using the approach of Donner and Eliasziw (1987) [19], considering a
minimally acceptable intraclass correlation coefficient (ICC) value of 0.70 [20]
and 5 repetitions in each of two sessions, recruitment of 30 participants at an
alpha level of 0.05 was estimated to provide 80% power to detect a
relationship this strong or stronger.
Data management and statistical analyses were performed using SPSS
version 17. Data were entered and inspected to ensure that there were no
errors of entry. Descriptive statistics were calculated for all variables.
Values for a single repetition and averages of 2, 3, 4 and 5 repetitions were
calculated for all dependant variables. To visually examine for the presence
of systematic error (e. g., fatigue or learning effects) average values of COP
mean velocity and sway area were plotted against the number of repetitions.
Relative reliability of the measures was assessed using ICC3,k [21] and 95%
confidence intervals. The standard error of measurement was calculated to
assess measurement precision using the formula:
Standard error of measurement = pooled standard deviation × 1-ICC .
Additionally, we calculated the minimal detectable difference (MDD) with 95%
confidence intervals. The MDD was calculated using the formula:
MDD =1.96×standard error of measurement× 2 . and estimates the smallest
difference exceeding measurement error [16]. To explore for systematic
differences between the first and second sessions, bias statistics with 95%
confidence intervals were calculated by computing the mean difference of
measures obtained during the two sessions. Levels of agreement (LOA) were
calculated as:
LOA = bias ±1.96 standard deviation of differences between the 2 sessions . It is
expected that 95% of the difference between the first and second sessions
would be between these limits [22]. Finally, to compare the effect of the
number of repetitions on reliability, all of the above calculations were
75
generated using values from a single repetition as well as the mean of the
first 2, 3, 4 and 5 repetitions.
Results
Thirty volunteers aged 20 to 57 years participated in the study. All subjects
were able to complete the protocol and all data were included for statistical
analysis. The final sample was composed of 16 men and 14 women, with a
mean ± SD age of 30.5 ± 7.2 years and BMI of 25.6 ± 5.5.
The profile plots displaying the averages of COP mean velocity and sway
area across repetitions were visually inspected and no learning or fatigue
effects were identified. Means, standard deviations, reliability coefficients,
standard error of measurement, MDD, bias and LOA statistics are presented
in table 1 for each dependant variable.
Table 1 Reliability results of single measures and means of 2, 3, 4 and 5 measures for each variable
Variable Mean ± SD * ICC (95% CI) SEM MDD Bias (95% CI) ± 95% LOA
Relative weight bearing on right leg (%)
1 repetition 50.10 ± 2.7 0.44 (−0.18,0.74) 2.0 5.6 0.3(−0.3,0.9) ±6.5
2 repetitions 49.84 ± 2.4 0.65 (0.25,0.83) 1.4 3.9 0.3(−0.2,0.7) ±4.9
3 repetitions 49.69 ± 2.3 0.75 (0.47,0.88) 1.2 3.2 0.1(−0.3,0.5) ±4.1
4 repetitions 49.69 ± 2.3 0.82 (0.63,0.92) 1.0 2.7 0.2(−0.2,0.5) ±3.6
5 repetitions 49.64 ± 2.3 0.83 (0.64,0.92) 0.9 2.6 0.3(−0.1,0.6) ±3.5
Relative weight bearing on left leg (%)
1 repetition 49.90 ± 2.7 0.44(−0.18,0.75) 2.0 5.6 0.3(−0.3,0.9) ±6.5
2 repetitions 50.16 ± 2.4 0.65 (0.27,0.85) 1.4 3.9 0.3 (−0.2,0.7) ±4.9
3 repetitions 50.31 ± 2.3 0.75 (0.49,0.89) 1.2 3.2 0.1(−0.3,0.5) ±4.1
4 repetitions 50.31 ± 2.3 0.81 (0.59-0.91) 1.0 2.7 0.1(−0.2,0.5) ± 3.7
5 repetitions 50.36 ± 2.3 0.82 (0.63-0.92) 1.0 2.7 0.2(−0.1,0.5) ±3.5
COP mean velocity (mm/sec)
1 repetition 6.0 ± 4.8 0.19 (−0.75,0.62) 4.4 12.1 −0.6(−1.9,0.7) ±13.50
76
2 repetitions 5.6 ± 3.4 0.83 (0.65,0.92) 1.3 3.5 0.2(0.1,0.4)† ±1.81
3 repetitions 5.5 ±3.3 0.95 (0.90,0.98) 0.7 2 0.2(0.0,0.5) ±2.68
4 repetitions 5.5 ± 3.3 0.97 (0.94,0.99) 0.6 1.6 0.09(−0.1,0.3) ±2.09
5 repetitions 5.4 ± 3.2 0.92 (0.84,0.96) 0.9 2.5 0.03(−0.3,0.3) ±3.43
Average location of COP (mm)
1 repetition 35.0 ± 21 0.53 (−0.01,0.78) 14.4 39.7 −1.8(−6.2,2.6) ±47.2
2 repetitions 38.4 ± 31.1 0.92 (0.84,0.96) 8.8 24.4 −1.6(−4.7,1.5) ±33.3
3 repetitions 37.3 ± 26.5 0.91 (0.81,0.96) 8 22.1 −0.4(−3.2,2.4) ±30.2
4 repetitions 38.8 ± 28.9 0.93 (0.85,0.97) 7.7 21.3 −0.7(−3.4,2.1) ±29.8
5 repetitions 39.0 ± 29.6 0.94 (0.88,0.97) 7.3 22 −0.6(−3.2,2.0) ±27.9
Sway area (mm2)
1 repetition 1549.6 ± 1605.5 0.06(−1.02,0.56) 1556.9 4302.7 −265.6(−686.4,155.1) ± 4517.0
2 repetitions 1442.6 ± 1055.5 0.47 (−0.13,0.75) 763.7 2116.8 −73.2 (−302.3,156.3) ±1399.0
3 repetitions 1456.6 ± 1001.9 0.63 (0.28,0.82) 612.7 1698.3 −159.2 (−350.4,32.1) ±2054.0
4 repetitions 1440.3 ± 889.8 0.68 (0.33,0.85) 504.8 1399.2 −159.2 (−278.9,43.8) ± 1732.7
5 repetitions 1466.9 ± 918.2 0.83 (0.64,0.92) 380.4 1054.3 −48.6 (−178.7,81.7) ± 1399.0
CI confidence interval, ICC intra class correlation coefficient, LOA limit of agreement, MDD minimal detectable difference, SD standard deviation, SEM standard error of the measurement; *Pooled from all repetitions;† Statistically significant bias (different from zero); Bold ICCs indicate acceptable reliability values
Each of the five variables measured exhibited unacceptable levels of
measurement error when calculated from single measures. For the relative
weight distribution on the legs, it was necessary to average 3 or more
repetitions to achieve a minimum acceptable reliability value. COP mean
velocity required at least 2 trials to obtain an acceptable ICC value. The ICC
values of the average COP location are acceptable with 2 measurements;
while measures of sway area required 5 repetitions to achieve acceptable
reliability.
Bias estimates were small and not significantly different from zero indicating
that there were no statistically significant differences between the two
sessions. The only exception was the COP mean velocity calculated of two
repetitions, which reached the threshold of statistical significance, however
the magnitude of this difference was small.
77
Discussion
The objectives of this study were to examine the test-retest reliability of
postural control measurements of a portable force plate commonly used by
clinicians and to explore the effect of using the mean value from multiple
repetitions on reliability. Our results demonstrated that for measures obtained
by the MPSA, single trials do not provide reliable estimates of postural
control and that averaging multiple measures is necessary to achieve
acceptable levels of measurement error. The number of repetitions
necessary to achieve reliable results varied depending of the outcome
variable and ranged from two to five. Clinicians should take this into account
when measuring postural control on their patients.
ICC values for measures of relative weight bearing appeared lower than ICC
values of COP mean velocity and average location of COP. However,
inspection of the descriptive statistics (mean and SD) in table 1 for each of
these variables indicates less inter-subject variability in relative weight
bearing as compared to the other dependent variables. Low levels of inter-
subject variability are known to artificially lower ICC estimates, as this
increases the relative magnitude of the error term in the ICC equation [16].
Thus, it can be difficult to interpret ICC values derived from homogenous
measures such as relative weight distribution.
We also assessed COP mean velocity, average location of COP, and sway
area. Consistent with our results, COP mean velocity has been reported by
others to be the most reliable estimate of COP [23-25]. In contrast, others
have examined samples of healthy participants and reported low reliability for
measures of COP mean velocity [26]. However, it should be noted that in that
study, the ICC statistics were calculated by averaging data from 3 10-second
repetitions. The longer duration of recording used in our protocol may explain
our higher reliability estimates. While longer duration trials of up to 120
seconds are recommended to reduce measurement error [24], sampling
duration should be matched to the abilities of participants. For instance,
children with cerebral palsy or the elderly may not tolerate standing for an
ideal duration of time.
78
For measures of sway area, it was necessary to average values from five
repetitions to achieve an acceptable level of measurement error. Two studies
reported similarly low ICC values for sway area [27,28]. Alternatively, others
have reported acceptable levels of ICC for this variable; however, these latter
studies were conducted under eyes closed.
A potential issue with relying on measures obtained from a suboptimal
number of repetitions can arise in clinical practice. For instance, when
examining for differences in postural stability over using single measures, a
practitioner needs to observe an improvement of at least 12.1 mm/sec in
COP mean velocity to be 95% confident that a true change has occurred.
However, when using a mean of 3 repetitions, a practitioner can be just as
confident that true change has occurred with a change of 2 mm/sec.
The results of this study are limited by several factors. This study was
conducted on healthy individuals and the results may or may not generalize
to clinical populations. Moreover, the MPSA as an instrument of
measurement has several limitations. The MPSA is limited to a fixed duration
of data acquisition of between 5 and 60 seconds. It is not possible to set up
the recording time to durations longer than 60 seconds, which may be
desirable is some circumstances. This technology also has a fixed sampling
rate and cut-off frequency and altering these frequencies is not possible.
Finally, the MPSA software does not report some variables such as sway
area, which we calculated from the raw data.
Future research should examine the validity of the force plates commonly
used by clinicians by comparing their measures to those obtained using force
plates with known validity. Additionally, it would be useful to assess the
reliability of similar force plates in a clinical population, such as those
individuals with neurological impairments. Future studies should also
examine the sensitivity of the MPSA to ensure that it can adequately
measure change over different time periods.
79
Conclusion
For measures of postural sway obtained by the MPSA, single trials do not
provide reliable estimates, and the averaging of multiple repetitions was
necessary to achieve acceptable levels of measurement error. Depending on
the variable, the number of repetitions required to achieve reliable data
ranged from 2 to 5. Clinicians should be wary of using single measures
derived from similar equipment when making decisions about patients.
Abbreviations
COP, Centre of pressure; ICC, Intraclass correlation coefficient; LOA, Level
of agreement; MDD, Minimal detectable difference; MPSA, Midot posture
scale analyser
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
SG participated in the design of the study, collected data, performed the
statistical analysis and drafted the manuscript. JH participated in the design
of the study, helped to perform the statistical analysis and drafted the
manuscript. BF helped to draft the manuscript. BW participated in the design
of the study and helped to draft the manuscript. All authors read and
approved the final manuscript.
References
1. Lafond D, Corriveau H, Prince F: Postural Control Mechanisms during
Quiet Standing in Patients with Diabetic Sensory Neuropathy. Diabetes
Care 2004, 27:173–178.
2. Brocklehurst JC, Robertson D, James-Groom P: Clinical correlates of
sway in old age, Sensory modalities. Age Ageing 1982, 11:1–10.
3. Horak FB, Henry SM, Shumway-Cook A: Postural perturbations: New
insights for treatment of balance disorders. Phys Ther 1997, 77:517–533.
4. Haas BM, Burden AM: Validity of weight distribution and sway
measurements of the Balance Performance Monitor. Physiother res
international 2000, 5:19–32.
80
5. Li X, Aruin AS: The effect of short-term changes in body mass
distribution on feed-forward postural control. J Electromyogr Kinesiol
2009, 19:931–941.
6. Berg K, Wood-Dauphinee S, Williams JI, Gayton D: Measuring balance
in the elderly: Preliminary development of an instrument. Physiother Can
1989, 41:304–311.
7. Kairy D, Paquet N, Fung J: A postural adaptation test for stroke
patients. Disabil Rehabil 2003, 25:127–135.
8. Mercer VS, Hankins CC, Spinks AJ, Tedder DD: Reliability and validity
of a clinical test of reaction time in older adults. J Geriatr Phys Ther
2009, 32:103–110.
9. Blum L, Korner-Bitensky N: Usefulness of the Berg Balance Scale in
stroke rehabilitation: A systematic review. Phys Ther 2008, 88:559–566.
10. Anker LC, Weerdesteyn V, van Nes IJW, Nienhuis B, Straatman H,
Geurts ACH: The relation between postural stability and weight
distribution in healthy subjects. Gait Posture 2008, 27:471–477.
11. Prieto TE, Myklebust JB, Myklebust B: Characterization and modeling
of postural steadiness in the elderly: a review. IEEE Trans Rehabil Eng
1993, 1:26–34.
12. Schiffman JM, Bensel CK, Hasselquist L, Gregorczyk KN, Piscitelle L:
Effects of carried weight on random motion and traditional measures of
postural sway. Appl Ergon 2006, 37:607–614.
13. Geiger RA, Allen JB, O’Keefe J, Hicks RR: Balance and mobility
following stroke: Effects of physical therapy interventions with and
without biofeedback/forceplate training. Phys Ther 2001, 81:995–1005.
14. Lajoie Y: Effect of computerized feedback postural training on
posture and attentional demands in older adults. Aging Clin Exp Res
2004, 16:363–368.
15. Clark RA, Bryant AL, Pua Y, McCrory P, Bennell K, Hunt M: Validity and
reliability of the Nintendo Wii Balance Board for assessment of
standing balance. Gait Posture 2010, 31:307–310.
16. Portney LG, Watkins MP: Foundations of clinical research: Application to
practice. 3 edition. Upper Saddle River, NJ: Prentice Hall Health; 2008.
17. Malina RM, Pena Reyes ME, Eisenmann JC, Horta L, Rodrigues J, Miller
R: Height, mass and skeletal maturity of elite Portuguese soccer
players aged 11–16 years. J Sports Sci 2000, 18:685–693.
81
18. Era P, Sainio P, Koskinen S, Haavisto P, Vaara M, Aromaa A: Postural
balance in a random sample of 7,979 subjects aged 30 years and over.
Gerontology 2006, 52:204–213.
19. Donner A, Eliasziw M: Sample size requirements for reliability
studies. Stat Med 1987, 6:441–448.
20. Lohr KN, Aaronson NK, Alonso J, Burnam MA, Patrick DL, Perrin EB,
Roberts JS: Evaluating quality-of-life and health status instruments:
Development of scientific review criteria. Clin Ther 1996, 18:979–984.
21. Shrout PE, Fleiss JL: Intraclass correlations: Uses in assessing rater
reliability. Psychol Bull 1979, 86:420–428.
22. Bland JM, Altman DG: Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet 1986, 1:307–310.
23. Lin D, Seol H, Nussbaum MA, Madigan ML: Reliability of COP-based
postural sway measures and age-related differences. Gait Posture 2008,
28:337–342.
24. Lafond D, Corriveau H, HeÌbert R, Prince F: Intrasession reliability of
center of pressure measures of postural steadiness in healthy elderly
people. Arch Phys Med Rehabil 2004, 85:896–901.
25. Hadian MR, Negahban H, Talebian S, Salavati M, Jafari AH, Sanjari MA,
Mazaheri M, Parnianpour M: Reliability of center of pressure measures of
postural stability in patients with unilateral anterior cruciate ligament
injury. J Appl Sci 2008, 8:3019–3025.
26. Doyle TL, Newton RU, Burnett AF: Reliability of traditional and fractal
dimension measures of quiet stance center of pressure in young,
healthy people. Arch Phys Med Rehabil 2005, 86:2034–2040.
27. Salavati M, Hadian MR, Mazaheri M, Negahban H, Ebrahimi I, Talebian
S, Jafari AH, Sanjari MA, Sohani SM, Parnianpour M: Test-retest reliabty of
center of pressure measures of postural stability during quiet standing
in a group with musculoskeletal disorders consisting of low back pain,
anterior cruciate ligament injury and functional ankle instability. Gait
Posture 2009, 29:460–464.
28. Takala EP, Korhonen I, Viikari-Juntura E: Postural sway and stepping
response among working population: Reproducibility, long-term
stability, and associations with symptoms of the low back. Clin Biomech
1997, 12:429–437.
82
CHAPTER FIVE
THE VALIDITY OF A PORTABLE CLINICAL FORCE PLATE IN
ASSESSMENT OF STATIC POSTURAL CONTROL:
CONCURRENT VALIDITY STUDY
Published as
Golriz S, Hebert JJ, Foreman KB, Walker BF. (2012)
Chiropractic and Manual Therapies, May 23;20(1):15.
83
In chapter 4, we realized that the MPSA had an acceptable level of test-retest
reliability for assessment of centre of pressure variables. However,
concurrent validity is another factor that needed to be tested before using the
MPSA in the main experiments. Fortunately, the School research laboratory
was enhanced with the purchase of an AMTI force plate which has been
widely used in research and could be used as a gold standard by which we
could compare the MPSA unit. Therefore, we investigated the concurrent
validity of the MPSA in this chapter.
The work resulted in the following publication:
Golriz S, Hebert JJ, Foreman KB, Walker BF. The validity of a portable
clinical force plate in assessment of static postural control: concurrent validity
study. Chiropractic and Manual Therapies. 2012 May 23;20(1):15
84
The validity of a portable clinical force plate in assessment of static
postural control: concurrent validity study
Samira Golriz1* * Corresponding author
Email: [email protected]
Jeffrey J Hebert2
Email: [email protected]
K Bo Foreman3
Email: [email protected]
Bruce F Walker2
Email: [email protected]
1 School of Chiropractic and Sports Science, Murdoch University, 90 South
Street, Murdoch, WA 6150 Perth, Australia
2 School of Chiropractic and Sports Science, Murdoch University, Perth, WA,
Australia
3 Department of Physical Therapy, University of Utah, Salt Lake City, UT,
USA
85
Abstract
Background: The broad use of force plates in clinical settings for postural
control assessment suggests the need for instruments that are easy to use,
affordable and readily available. In addition, these instruments of
measurement should be reliable and valid as adequate reliability and validity
are prerequisites to making correct inferences. The aim of this study was to
examine the concurrent validity of postural control measures obtained with a
clinical force plate.
Methods: Thirty-one healthy adults were recruited. Participants completed 1
set of 5 trials on each force plate. Postural control measures (centre of
pressure [COP] average velocity and sway area) were collected and
compared using the Midot Posture Scale Analyzer (clinical force plate) and
the Accugait force plate (criterion measure). Intra class correlation coefficient
(ICC), standard error of measurement, and paired t-tests were calculated and
Bland-Altman plots were constructed to compare the force plates and assess
consistency of measurement and agreement between them.
Results: The ICC values (ICC = 0.14-0.60) between the two force plates were
lower than the acceptable value for both COP average velocity and sway
area. There was significant difference (p > 0.05) in COP average velocity and
sway area between the force plates. Examination of the plots revealed that
there is less difference between the force plates in lower magnitudes of COP
for average velocity and sway area however, the greater the average velocity
and sway area, the greater the difference between the measures obtained
from the two force plates.
Conclusion: Findings of this study showed poor concurrent validity of the
clinical force plate. This clinical force plate cannot be a replacement for
known reliable and valid force plates and consequently measures obtained
from this force plate should be treated with caution especially in a clinical
population.
Keywords: Concurrent validity, Force plate, Postural control
86
Introduction
Postural control is a crucial factor in maintaining balance during standing,
walking, task performance and when responding to the unexpected
perturbations experienced in everyday life [1]. Postural control assessment
can provide useful information when identifying individuals who are
susceptible to postural control deficits [2,3]. Furthermore, postural control
assessment has been used in sports medicine for selection of talented
athletes, identification of athletes at high risk for injury, and for the prevention
of sports related injuries [4]. Postural control is usually assessed by
interpretation of parameters derived from the centre of pressure (COP) such
as velocity and area of COP displacement [5]. COP is defined as the point of
application of ground reaction forces under the feet.
In a clinical setting, force plates are regularly used to objectively assess
postural control [6]. However, the reliability and validity of these force plates
is often unknown. Reliability represents the consistency of measures or the
level to which an instrument is free from errors of measurement [7]. Validity
represents the extent to which an instrument measures what it is supposed to
measure [7]. Concurrent validity examines the validity of an instrument
against a criterion measure to test if the new instrument can be used instead
of the criterion [7]. Understanding the reliability and validity of a measurement
tool is crucial for making correct inferences from the data collected.
The Midot posture scale analyser (MPSA) is a lower cost portable force plate
commonly used in the clinical setting. The MPSA has previously
demonstrated acceptable reliability when averaging at least 5 values
(ICC > 0.70) [8]. However, the validity of the MPSA has not been previously
reported. Therefore, the purpose of this study was to examine the concurrent
validity of postural control measures obtained with a clinical force plate when
compared to a “gold standard” valid instrument. It is to be noted that the “gold
standard” was by decision by consensus among the research team rather
than a standard reported in previous research.
87
Materials and methods
Participants
Participants were recruited through advertisements on University bulletin
boards. The inclusion criteria were 18 to 60 years of age. A lower age limit of
18 years was set because the development of the skeletal system reaches its
maturity at this age [9]. An upper age limit of 60 years was set because the
ability to maintain postural control decreases in the elderly [10]. Potential
participants were excluded from the study if they reported having balance
deficits stemming from rheumatological or neurological disorders, a recent
musculoskeletal injury (within three months), an ear infection or fever within
72 hours of the testing session, current pregnancy or the use of medications
that could alter sensory perception (see appendices of chapter 5, page 166).
The Human Research Ethics Committee of Murdoch University approved the
experimental protocol (2010/220), and all participants gave written consent
before enrolment in the study.
Instruments
The MPSA (QPS-200, Midot Medical Technology, Shekel Electronic Scale,
Israel) is a portable force plate consisting of four electronic weighing plates
set in a rectangular position (Figure 1). The MPSA records data using an
internal sampling frequency of 200 Hz and filter frequency of 0.5 Hz.
Calibration was conducted in accordance with the manufacturer’s
recommendations using a 20 kg certified weight.
The Accugait (Advanced Mechanical Technology Inc., Watertown, MA,USA)
was used as the criterion measure or reference standard. This force plate is
a portable square force plate. There are flat rubber pads in each corner of the
force plate to make the force plate less sensitive to vibration from most floor
surfaces (Figure 1). The Accugait measures the three dimensional applied
forces (Fx, Fy, Fz) and moments (Mx, My, Mz) involved in balance and uses
established algorithms to compute the location of the COP and its associated
variables from the forces and moments applied to the force plate. Data were
acquired, recorded and analyzed using Balance Clinic software (balance
88
software for AMTI’s Accusway plus balance platform, version 2.02.01) loaded
on a Dell laptop.
Figure 1 The Midot Posture Scale Analyzer and The Accugait AMTI
The Accugait force plate was initially factory validated prior to its release
(personal communication with AMTI). However, to confirm the factory results
we validated the Accugait based on the validation manual and validation
report provided by AMTI [11]. The validation test represented an absolute
COP error (the cumulative effect of noise and drift) of 3.6 mm for the X
average, 3.4 mm for the Y average, 33.6 mm/s for average velocity and 2.9
mm2 for sway area over 40 seconds of data acquisition at a 50 Hz sampling
frequency. All of these values were sufficiently close to the values reported in
the AMTI validation report and as such the instrument was considered valid.
We used data filtration to remove noise [12] and chose a 5Hz cut off
frequency as the best level to filter data by performing a residual analysis.
Residual analysis computes the differences between filtered and unfiltered
signals over a wide range of cut off frequencies. To accomplish this we
calculated differences between filtered and unfiltered COP average velocity
signals over a wide range of cut off frequencies (1, 2, 3, 5, 10, 20 Hz) and
plotted them (Figure 2). The plot was composed of linear and non linear
parts. The linear part mostly represented random noise and the non linear
part represented true signal. The cut off frequency where the plot turned from
linear to non linear was chosen as the best cut off frequency for this
experiment [12].
89
Figure 2 Differences between filtered and unfiltered average velocity vs. cut off frequency to decide the best filter frequency
The acquisition sampling frequency of the Accugait was set at 100 Hz with
the cut-off frequency of 5 Hz. Data were filtered using a fourth order
Butterworth filter [13]. As we considered the Accugait as the criterion
measure, the acquisition sampling frequency and cut off frequency of the
Accugait was set at the optimal setting regarding the task and environment of
our study. The force plate was calibrated with a calibration CD provided by
the manufacturer and it was zeroed before each recording.
The COP average velocity (mm/sec) and the sway area, which is the area of
an ellipse enclosing 95% of COP movements (mm2), are the two variables
that were assessed and compared.
Procedure
The two force plates were placed on the laboratory floor next to each other.
The manufacturer recommends use of the force plate on any flat surface and
adds there is no need to install the force plate permanently in the ground
[13]. All participants were tested under the same conditions. Paper was
placed on top of force plates and the force plates were zeroed. Participants
90
were asked to wear loose fit comfortable clothing, remove their shoes and
stand upright on the force plate and remain as still as possible with a relaxed
posture. The participants were asked to put their arms to their sides in a
comfortable position and distribute their body weight evenly on both feet.
Also, they were asked to breathe normally and look straight ahead at an “X”
on the opposite wall that was located two meters away at their eye level.
Before commencement of data acquisition and after inspection for position
symmetry, both feet were outlined on the piece of paper located on top of the
force plate to ensure consistent foot placement across trials and between the
two force plates. The same procedure was used for testing on the second
force plate using the outline of the feet to ensure consistent foot placement.
Participants completed 1 set of 5 trials on each force plate as outlined above.
To reduce the potential for bias due to ordering effects, we randomly
allocated and counterbalanced the order in which the measurements were
obtained on each force plate.
Five successive identical trials of 60 seconds were acquired on each force
plate. Mandatory breaks of one minute and 5 minutes were allocated
between trials and between force plates, respectively. Data were averaged
across five trials, this number of repetitions was informed by a study where
we assessed the reliability of the MPSA and found acceptable reliability when
averaging at least 5 values [8].
Statistical analysis
We conducted an a priori sample size estimation. Using the approach of
Donner and Eliasziw (1987), assuming an alpha level of 0.05 and a minimally
acceptable ICC value of 0.70 [7], recruitment of 30 participants provided 80%
power [14].
Data management and statistical analyses were conducted using SPSS
(version 17, Chicago, IL, USA). Data from 31 participants across 2 sets, 5
minutes apart were included for statistical analysis. Averages of 5 trials of
COP average velocity and sway area were used for the analyses.
91
We calculated intraclass correlation coefficients (ICC3,5) for comparison
between the two force plates. Paired t-tests were also performed to examine
for differences in average velocity and sway area between the force plates.
We examined the consistency of measurements between the 2 force plates
by calculating the standard error of measurement using the formula:
Standard error of measurement = pooled standard deviation × √1 − ICC.
Additionally, bias statistics with 95% confidence intervals and limits of
agreement (LOA) were calculated and Bland-Altman plots [15] were
constructed for COP average velocity and sway area. Plots were constructed
by plotting the mean difference between the two force plates against the
mean of the two force plates. Plots were examined for the magnitude of the
difference between the force plates and the distribution around the mean line
(bias). Bias represents the mean difference between the two force plates,
while the LOA examines agreement between the two force plates. LOAs are
defined as bias ± 1.96 SD, where SD is the standard deviation of the
difference between measures of the two force plates.
Results
Thirty-one healthy participants consisting of 18 males and 13 females, aged
23 to 58 were recruited from the staff and students of Murdoch University. All
participants were able to complete all test repetitions and all data were
included for statistical analysis. The mean (SD) age of the participants was
32.7 (8.5) years and BMI was 24.7 (5.4). Descriptive statistics, ICC, standard
error of measurement, bias and limits of agreement are presented in table 1
for average velocity and sway area. The ICC values between the two force
plates are lower than the acceptable value with wide confidence intervals for
both average velocity and sway area. Paired t-tests showed significant
differences in average velocity and sway area between the 2 force plates
(p < 0.05).
92
Table 1 Criterion validity analysis
Mean (SD) t-test (p value)
SEM ICC (3,5)(CI’s) Bias (95% CI) ± 95% LOA
COP average velocity (mm/s)
Accugait 8.3 (2.3)
0.004* 2 0.60(0.16-0.80) 1.7† (1.1,2.2) ± 5.9
MPSA 6.6 (3.5)
Sway area (mm2)
Accugait 274 (155)
0.000* 870 0.14 (−0.24-0.48)
-1660† (-1370,-1950) ± 3134
MPSA 1936 (1720)
CI, confidence Interval; *p < 0.05 indicates a significant difference in values between the 2 force plates; †statistically significant bias (different from zero)
The Bland–Altman plots for the average velocity and sway area are provided
in figures 3 and 4 and demonstrate more variation between the measures
obtained from the two force plates when the amount of COP average velocity
and sway area were higher.
93
Figure 3 Bland-Altman plot representing comparison of average velocity between the MPSA and the Accugait
Figure 4 Bland-Altman plot representing comparison of sway area
between the MPSA and the Accugait
Mean
Mean + 1.96SD
Mean – 1.96SD
Mean
Mean + 1.96SD
Mean – 1.96SD
94
Discussion
The purpose of this study was to investigate the concurrent validity of the
MPSA by measuring COP variables and comparing these with a validated
reference standard. The ICCs of the postural control measures were equal to
0.60 for the average velocity and 0.14 for the sway area between the force
plates. Results of the reliability study of the MPSA also reported higher
agreement for the average velocity and a lower agreement for sway area [8].
The ICC is a ratio estimate and there is no widely agreed upon thresholds for
identifying an acceptable level of agreement with ICC reporting. As a result,
there are various interpretations of ICC values available in the literature.
Landis and Koch [16] suggested the following qualitative approach to ICC
interpretation: 0.00-0.20 slight, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80
substantial and 0.81-1.00 almost perfect. Alternatively, Portney and Watkins
[7] suggest that values over 0.75 indicate good agreement, and values below
0.75 are indicative of poor to moderate agreement. The highest level of
agreement identified in the current study was ICC = 0.60. Therefore, these
results represent evidence that the MPSA does not possess sufficient
concurrent validity.
The high standard error of measurement values of both variables especially
the sway area, when compared to the mean values obtained by the two force
plates, indicates a lack of consistency of measurement between the force
plates. Furthermore, our results demonstrated significant differences
between the COP measurements obtained from the force plates for the COP
average velocity and sway area.
Comparison of standard deviations to means of sway area captured by the
two force plates shows that the large standard deviation of MPSA sway area
indicates a high degree of random error in the MPSA data. Additionally, the
large mean difference between MPSA and Accugait sway area might be a
sign of a systematic error in the MPSA when assessing postural control.
Visual inspection of the Bland-Altman plot (Figure 3) of average velocity
shows a funnel effect, meaning there is more variability when the magnitude
of the average velocity is greater. Our results represent excessive variation
95
between the force plates in higher magnitudes of average velocity and less
variation in lower magnitudes of average velocity. Differences between the
measures of COP velocity between the force plates depend on the
magnitude of the measurement. The excessive variations in higher
magnitudes of average velocity indicate that data captured by the MPSA
contain errors for measuring higher COP velocities.
The bias estimates for COP velocity were small but statistically different from
zero. Statistically significant bias was found for the average velocity variable
where the MPSA values underestimate the criterion in most of the cases with
an estimate of 1.7 and a 95% confidence interval of 1.1 - 2.2. One possible
explanation for this is that data captured by the Accugait was filtered with a 5
Hz cut off frequency whereas data obtained by the MPSA were filtered with a
fixed 0.5 Hz cut off frequency. This may have resulted in the MPSA
eliminating true signal. The lower cut off frequency causes removing higher
proportion of true signals.
The plot of the sway area (Figure 4) shows a systematic trend. With lower
magnitudes of sway area the differences between the two force plates are
relatively low as compared to higher magnitudes, but the greater the sway
area the bigger the difference between the two force plate measures.
Moreover, the statistically significant bias of sway area showed that the
MPSA overestimated sway area. Overall, larger measures of average
velocity and sway area result in increased systematic and random error in
MPSA.
While the MPSA previously demonstrated acceptable reliability [8], its validity
is not satisfactory; therefore, it cannot be considered a replacement of a
known valid force plate for the assessment of postural control. In addition,
results obtained by the MPSA on clinical populations should be treated with
caution. Clinical populations are potentially less stable and show higher
magnitudes of average velocity and sway area [17,18] and the MPSA was
incapable of measuring accurate data with higher magnitudes of average
velocities and sway areas. Additionally, to reduce measurement error, we
used a mean of 5 trials, which is unlikely to be the case with clinical use
96
given the time constraints of clinical practice. Clinicians should consider all
the above mentioned issues when assessing postural control using the
MPSA.
Although we cannot rely on the MPSA for assessment of postural control, it
may possibly be used in obtaining qualitative estimates of postural control,
for instance as a biofeedback training tool and to enhance motivation level of
patients with balance defects. Having said that, we suggest clinicians
intending to use a force plate purchase a reliable and valid instrument.
Limitations exist within this study. The study sample consisted of healthy
individuals and the findings may or may not generalize to clinical populations.
This study did not attempt to assess the validity of the MPSA in different
testing positions such as eyes closed, single limb standing and narrow
stance (feet together) although we hypothesis that this would not enhance
the MPSA performance.
Future research could be done to assess the validity of the MPSA in different
balance testing positions and clinical populations. Estimates of reliability and
validity should be known prior to using any type of force plates.
In conclusion, postural control parameters cannot be validly measured in
clinical or research settings using the MPSA. The MPSA is a lower cost force
plate with a low-technology design and easy to use software but it did not
fulfil the criteria to be regarded as a valid force plate for clinical use.
Abbreviations
COP, Centre of pressure; ICC, Intraclass correlation coefficient; LOA, Level
of agreement; MPSA, Midot posture scale analyser
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
SG participated in the design of the study, collected data, performed the
statistical analysis and drafted the manuscript. JH participated in the design
of the study, helped to perform the statistical analysis and drafted the
97
manuscript. BF helped to draft the manuscript. BW participated in the design
of the study and helped to draft the manuscript. All authors read and
approved the final manuscript.
References
1. Westcott SL, Lowes LP, Richardson PK: Evaluation of postural stability
in children: Current theories and assessment tools. Phys Ther 1997,
77(6):629–645.
2. McGuine TA, et al: Balance as a predictor of ankle injuries in high
school basketball players. Clin J Sport Med 2000, 10(4):239–244.
3. Lafond D, et al: Postural control during prolonged standing in
persons with chronic low back pain. Gait Posture 2009, 29(3):421–427.
4. Rogind H, et al: Comparison of Kistler 9861A force platform and
Chattecx Balance System® for measurement of postural sway:
Correlation and test-retest reliability. Scand J Med Sci Sports 2003,
13(2):106–114.
5. Prieto TE, et al: Measures of postural steadiness: Differences
between healthy young and elderly adults. IEEE Trans Biomed Eng 1996,
43(9):956–966.
6. Guskiewicz KM, Perrin DH: Research and clinical applications of
assessing balance. J Sport Rehabil 1996, 5(1):45–63.
7. Portney LG, Watkins MP: Foundations of clinical research: Application to
practice. 3rd edition. Upper Saddle River: Prentice Hall Health; 2008.
8. Golriz S, et al.: The reliability of a portable clinical force plate used for
the assessment of static postural control: repeated measures reliability
study. Chiropractic and Manual Therapies. 2012 May, 20:14
(doi:10.1186/2045-709X-20-14)"
9. Malina RM, et al: Height, mass and skeletal maturity of elite
Portuguese soccer players aged 11–16 years. J Sports Sci 2000,
18(9):685–693.
10. Era P, et al: Postural balance in a random sample of 7,979 subjects
aged 30 years and over. Gerontology 2006, 52(4):204–213.
11. AMTI: Calibration validation manual, I.A. Advanced Mechanical
Technology, Editor: Watertown, MA, USA.
12. Winter DA: Biomechanics and motor control of human movement. 4th
edition. Hoboken, New Jersey: John Wiley & Sons INC; 2009.
98
13. AMTI: Balance software for AMTI's AccuSway plus balance platfrom.
Watertown: I. A. Advanced Medical Technology; 2006.
14. Donner A, Eliasziw M: Sample size requirements for reliability
studies. Stat Med 1987, 6(4):441–448.
15. Bland JM, Altman DG: Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet 1986,
1(8476):307–310.
16. Landis JR, Koch GG: The measurement of observer agreement for
categorical data. Biometrics 1977, 33(1):159–174.
17. Blaszczyk JW, et al: Assessment of postural instability in patients
with Parkinson's disease. Exp Brain Res 2007, 183(1):107–114.
18. Ruhe A, Fejer R, Walker B: Center of pressure excursion as a
measure of balance performance in patients with non-specific low back
pain compared to healthy controls: a systematic review of the literature.
Eur Spine J 2011, 20(3):358–368
99
CHAPTER SIX
THE EFFECT OF HIP BELT USE AND LOAD PLACEMENT IN A
BACKPACK ON POSTURAL STABILITY AND PERCEIVED
EXERTION: A WITHIN SUBJECTS TRIAL
Submitted to
Gait and Posture
100
In chapter 4 we found out that although the MPSA has acceptable test retest
reliability, but does not have an acceptable level of concurrent validity and
therefore is not suitable for assessing postural stability. Thus, we decided to
use the AMTI force plate to record postural stability measures. We validated
the AMTI force plate in chapter 5 and evaluated its test–retest reliability
which will be mentioned in this chapter. In this chapter we examine the effect
of load placement and hip belt use on perceived exertion and objective and
subjective measures of postural stability.
This chapter has resulted in the following submission for publication;
Golriz S, Jeffrey JJ, Foreman KB, Walker BF. The effect of hip belt use and
load placement in a backpack on postural stability and perceived exertion: a
within subjects trial. Submitted to Gait and Posture
101
The effect of hip belt use and load placement in a backpack on postural
stability and perceived exertion: a within subjects trial
Samira Golriz a, Jeffrey J. Hebert a, K. Bo Foreman b, Bruce F. Walker a
a, School of Chiropractic and Sports Science, Murdoch University, Perth,
Australia
b, Department of Physical Therapy, University of Utah, Salt Lake City, UT,
USA
Samira Golriz* * Corresponding author
Email: [email protected]
90 South Street, Murdoch 6150, WA, Australia
Tel: +61 8 93601450
Fax: +61 8 93601299
Jeffrey J. Hebert
Email: [email protected]
K. Bo Foreman
Email: [email protected]
Bruce F. Walker
Email: [email protected]
Acknowledgement
This research was funded by Promopak Pty Ltd. Perth, WA; however, the
manufacturer played no part in the design, conduct or reporting of this study.
102
Abstract
Carrying a backpack is a popular mode of load carriage and is considered
the most appropriate way of carrying additional weight. However, carrying a
backpack can cause discomfort and affect postural stability. Hip belt use and
load placement in a backpack may affect comfort and stability, as it helps to
distribute the load over the body. The purpose of this study was to assess the
effects of hip belt use and load placement in a backpack on subjective and
objective measures of postural stability, as well as perceived exertion. Thirty
participants were instructed to stand on a force plate for one minute and then
walk along a designated route under five conditions: unloaded, high load
placement in a backpack, low load placement in a backpack, hip belt on and
hip belt off. Backpacks were loaded with 20% of participants’ body weight.
The centre of pressure (COP) average velocity and sway area from the force
plate were measured. Participants rated their perceived stability and exertion
using a perceived sense of postural sway and instability scale, as well as the
Borg scale. The variables were analysed using separate, one–way repeated
measures analysis of variance (ANOVA). Compared to the unloaded
condition, all loaded conditions significantly increased COP average velocity,
sway area, perceived stability and exertion. Hip belt use did not affect COP
average velocity and sway area; however, participants reported higher levels
of stability and lower levels of exertion with hip belt use. Load placement did
not affect COP average velocity, sway area, perceived stability or exertion.
This study showed that wearing a loaded backpack reduced postural stability,
while manipulation of load placement in a backpack did not affect subjective
and objective measures of postural stability. Also, this study demonstrated
hip belt use did not improve objective measures of postural stability;
however, it did help the participants to feel more stable and they reported
lower perceived exertion.
Keywords: postural stability, backpack, hip belt, load placement, exertion
103
Introduction
The human bipedal stance is inherently unstable due to the relatively high
position of the body’s centre of mass (COM) above a small base of support
[1]. To remain stable and avoid injury during daily living activities, the body
needs to keep the COM within the base of support [2]. The central nervous
system controls the maintenance of postural stability by obtaining information
from vestibular, visual, and somatosensory inputs [3]. Disease, medications,
the aging process and carrying loads can affect postural stability [4, 5].
Carrying loads can also decrease postural stability and falls have been
reported by backpack users [6].
Early backpack designs have evolved into specific models for children,
recreational backpackers, military and emergency personnel. As part of this
development, features such as a hip belt have been added in an attempt to
make backpacks more secure, comfortable, improve posture and decrease
energy costs. Without a hip belt, the majority of the vertical force is imparted
on the shoulders. Whit a hip belt, approximately 30% of the force is
transferred to the pelvis [7]. The transfer of the vertical force from the
shoulders to the pelvis is speculated to reduce activity in the shoulders and
trunk muscles (e.g. trapezius and erector spinae muscles) [8] as well as
reduce the incidence of “rucksack palsy” and shoulder–backpack interface
pressure [9-11]. Also, it results in the increased stability of the pelvis–thorax
coordination pattern [12], thus providing more comfort and improving
performance [8, 13]. A further advantage of hip belt use is that the pelvis is
less sensitive to contact pressure than the shoulders. As a result, load
bearing by the pelvis is generally more comfortable than load bearing at the
shoulders [14].
Backpack users traditionally carry heavy loads while simultaneously dealing
with the perturbations experienced in daily living. It has been demonstrated
that carrying a backpack has an impact on static and dynamic postural
stability and may lead to loss of balance and fall in adults and children. In
everyday activities, backpack users stand still, encounter stairs, walk on
multiple surfaces and may have to lift the loaded bag and often bend over to
104
pick up fallen objects with the pack on. Performing these activities affects
postural stability and may result in slips or falls. If hip belt use improves
postural stability, it may have important implications regarding decreasing the
probability of fatigue, fall and injury. While it may appear to be a logical
assumption that using a hip belt improves postural stability, a review of the
literature shows that the effect of hip belt use on postural stability is unknown
[15].
We hypothesized that lower levels of postural stability and higher levels of
perceived exertion would occur for loaded conditions result than for an
unloaded condition. We also hypothesized that hip belt use would provide
physical constraint between the pelvis and thorax. Furthermore, a hip belt
would facilitate the transfer of a substantial amount of load from the
shoulders to the pelvis, possibly allowing the trunk muscles to handle the
imposed load better and focus more on postural stability. This could enhance
postural stability and reduce perceived exertion.
In addition to the consideration of a hip belt, optimising load placement within
a backpack may also be important in backpack design improvements to
enhance static and dynamic postural stability. Without knowing the best load
placement, the alteration of other backpack features is premature [16]. In the
literature, the effect of load placement on energy consumption [17, 18],
muscle electrical activity [19], and posture [16] has been studied with the
ultimate goal of making a backpack more comfortable and efficient to carry.
However, there is limited evidence regarding the effects of load placement on
objective and subjective measures of postural stability.
We hypothesized that high load placement would elevate the centre of
gravity (COG) of the body and tend to destabilize the body when compared
to low load placement. The result would be lower postural stability and higher
levels of perceived exertion. [19]. Accordingly, the purpose of this study was
to assess the effect of hip belt use and load placement in a backpack on
rating of perceived exertion (RPE) and subjective and objective measures of
postural stability.
105
Methods
Study design
This study used a one–way repeated measures design in which participants
served as their own controls. The Murdoch University Ethics Committee
approved the study protocol (2010/221), and all participants provided written
consent before study enrolment.
Participants
Potential participants were occasional backpack users who were recruited
from a university campus and screened based on predefined selection
criteria. We recruited participants if they were aged between 18 and 60 years
and excluded participants who: 1) had balance deficits stemming from
rheumatological or neurological disorders, 2) had a recent musculoskeletal
injury, or spinal or lower extremity pain (within three months), 3) had an ear
infection or fever within 72 hours of testing, 4) were currently pregnant or 5)
were taking medications that could alter sensory perception (see appendices
of chapter 6, page 168).
Procedures
Participants completed two study sessions of approximately one hour
duration. The study sessions consisted of a hip belt session and a load
placement session and were scheduled seven days apart to minimise
carryover effects. The hip belt session comparison consisted of three
conditions: unloaded, hip belt and no hip belt. The load placement session
comparison consisted of three conditions: unloaded, high load placement
within the backpack and low load placement within the backpack. During
each session, participants completed dynamic and static tests.
The sequencing of each comparison (hip belt or load placement) and
experimental condition (unloaded, hip belt, no hip belt, high and low load
placement) were randomised and counterbalanced using an online random
number generator [20].
106
Static testing condition
The Accugait– AMTI (Advanced Mechanical Technology Inc., Watertown,
USA) force plate and Balance Clinic software (Advanced Mechanical
Technology Inc., Watertown, USA) were used to acquire, record and analyse
the data. We extracted two variables from the software: COP average
velocity defined as the path length of the COP divided by the measurement
time (cm/s); and sway area, defined as the area of an ellipse enclosing 95%
of COP movements (cm2).
We validated and calibrated the force plate before acquiring the data. The
process of validating and determining the cut–off frequency has been
described elsewhere [21]. The acquisition sampling frequency was 100 Hz
with a cut–off frequency of 5 Hz and data were filtered using a fourth–order
low–pass Butterworth filter.
In addition to assessment of the validity of the force plate, we examined the
test–retest reliability of the force plate on a separate cohort of 30 participants,
using average values from three repetitions. The intraclass correlation
coefficient ICC(3,3), standard error of measurement and minimal detectable
difference (MDD) of the average velocity (cm/s) were 0.95 (95% confidence
interval = 0.88,0.98), 0.01, 0.03, respectively. The ICC, standard error of
measurement and MDD of sway area (cm2) were 0.80 (95% confidence
interval = 0.58, 0.90), 0.36, 0.99, respectively.
To account for factors that could influence our data, we asked participants
not to consume alcohol for 12 hours before, not to drink coffee three hours
before and not to smoke 30 minutes before testing. Participants stood
barefoot and upright on the force plate while relaxed with their arms at their
sides. They were asked to distribute their body weight evenly on both feet
and breathe normally while looking straight ahead at a target that was
located at eye level two metres away on the opposite wall. We placed a piece
of paper on the top of the plate and outlined both feet to ensure consistent
foot placement across trials. To reduce measurement error, three 60–second
measurements were obtained and mean values were calculated. Mandatory
breaks of three minutes took place between repetitions to minimise fatigue
107
effects. Participants were allowed to sit down and take the backpack off
during the breaks.
Dynamic testing condition
Upon completion of the static testing conditions, participants walked 250
metres along a designated route at a self–selected pace. The route included
variable surfaces such as brick paving, fine gravel, garden bark and mulch,
as well as changing slopes and steps. The average time for completion was
around four minutes. All participants wore athletic footwear such as running
shoes and loose fit comfortable clothing.
Upon completion of the dynamic testing condition, the perceived sense of
postural sway and instability scale (PSPSI) was completed to assess
perceived stability. The PSPSI is a subjective measure of postural stability
and has been used to estimate psychophysical safety during task
performance. The PSPSI consists of four questions, each with possible
scores ranging from 0 to 2 and a maximum summated score of 8, with higher
scores representing lower levels of postural stability [22] (see appendices of
chapter 6, page 171). This scale has demonstrated adequate test–retest
reliability and criterion validity in healthy subjects between the age of 20 to 60
during standing upright, moving, bending and lifting [22].
The Borg scale was also completed upon completion of the dynamic testing
condition to measure overall and local RPE. Local RPE was measured for
the neck, shoulder, upper back, lower back, and lower extremity regions (see
appendices of chapter 6, page 172). To fill out the Borg Scale, participants
were clearly asked to rate overall and local perceived exertion and not
discomfort or pain. This scale consists of 15 points with potential scores
ranging from 6 (very, very light) to 20 (maximum effort). This scale has
acceptable test–retest reliability and criterion validity [23]. Visual analogue
scale was also completed to measure discomfort.
Backpack configuration
We used the Promopak backpack (Promopak Pty Ltd, Australia) with
dimensions of 47 x 21 x 15 cm and a hip belt with 4 cm width (see figure 1).
108
The backpack was loaded to 20% of the participant’s body weight, up to a
maximum of 20 kg. We divided the backpack into upper, middle and lower
compartments using high–density foam. In the hip belt session, the weights
were placed in the middle compartment of the backpack. During load
placement session, we fixed the load in the upper or lower backpack
compartments (see figure 2). Loads were placed as close to the spine as
possible and distributed evenly on the right and left sides of the backpack.
The unloaded condition was defined as wearing no backpack.
Figure 1. The Promopak backpack
Figure 2. The loading conditions in the backpack
Left, Middle load placement; Centre, High load placement and Right, Low load placement
To place the backpacks, a single investigator used the same location relative
to the shoulders with the top of the backpack 10 cm below shoulder level on
all participants. The strap lengths were tailored to each participant’s body
type and maintained throughout the measurement sessions.
109
Statistical analysis
Using sway area data obtained from our examination on the reliability of the
force plate, we performed an a priori power analysis using G-Power 3.1.3
software [24]. We found that recruiting 26 participants, at an alpha level of
0.05, would provide a 90% power to detect an MDD of 0.99 cm2. However,
we recruited 30 participants to account for the possibility of participant loss to
follow up.
All statistical analyses were conducted using SPSS version 17.0 (Chicago,
IL, USA). Data from three repetitions, in each condition, were averaged and
used in the postural stability analyses. Descriptive statistics were reported for
all dependent variables. To examine for differences in postural stability
measures, perceived stability and RPE we conducted separate one–way
repeated measures analyses of variance (ANOVA) with least significant
difference pairwise post hoc comparisons. The first independent variable was
the hip belt loading condition with three levels of unloaded, hip belt and no
hip belt. The second independent variable was the load placement loading
condition with three levels of unloaded, high load placement and low load
placement. The dependent variables were COP average velocity, sway area,
perceived stability and RPE. The Kolmogorov–Smirnov test was used to test
data normality and Mauchly’s test of sphericity was used to assess data
sphericity. We treated perceived stability scores as a continuous variable
since they were normally distributed [25]. The alpha level was set at 0.05 for
all comparisons.
Results
Thirty healthy adults (14 females), aged between 18 and 55 years,
volunteered to participate in the study. None of the participants dropped out
of the study. All participants completed both testing sessions. Participant
demographic and baseline data are provided in table 1. All data were
normally distributed. The means, standard deviations and difference scores
with 95% confidence intervals for each dependent variable are shown in
tables 2 and 3.
110
Table 1. Demographic and baseline characteristics of participants (N=30)
Variable Value
Age (years) 32.8 (8.3)
Sex (%) Female 47%
Weight (kg) 69.3 (19.4)
Height (cm) 170.1 (10.6)
BMI (kg/m2) 23.9 (5.1)
Loaded backpack weight (kg) 13.6 (3.2)
Values are mean (standard deviation) unless otherwise indicated.
Table 2. Mean (standard deviation) and mean difference (95% CI) values during unloaded, hip belt and no hip belt condition
Mean
(SD)
Mean difference
(95% CI)
Variable Unloaded Hip
belt
No hip
belt
Unloaded-
hip belt
Unloaded-no
hip belt
Hip belt-
no hip
belt
Average
velocity
0.73
(0.15)
0.88
(0.23)
0.89
(0.20)
-0.15
(-0.22,-0.90)
-0.16
(-0.22, -0.10)
-0.01
(-0.04,0.03)
Sway
area
2.50
(1.26)
4.12
(2.40)
4.31
(2.68)
-1.62
(-2.38,-0.86)
-1.81
(-2.68, -0.94)
-0.19
(-0.87,0.48)
PSPSI 0.25
(0.54)
1.48
(1.44)
2.28
(1.96)
-1.23
(-1.69,-0.77)
-2.03
(-2.66, -1.40)
-0.8
(-1.17,-
0.42)
Overall
RPE
7.67
(0.39)
11.33
(0.52)
12.50
(0.51)
-3.67
(-4.72,-2.62)
-4.83
(-5.67, -3.99)
-1.17
(-1.99,-
0.35)
Neck RPE 7.33
(2.35)
10.43
(2.76)
11.43
(3.00)
-3.10
(-4.26,-1.94)
-4.10
(-5.22, -2.98)
-1.00
(-1.75,-
0.25)
111
shoulder
RPE
6.83
(1.18)
12.67
(2.50)
13.67
(2.37)
-5.83
(-6.81,-4.86)
-6.83
(-7.77, -5.89)
-1.00
(-1.64,-
0.36)
Upper
back RPE
7.33
(1.95)
11.43
(3.14)
12.60
(2.98)
-4.10
(-5.20, -3.00)
-5.27
(-6.33, -4.20)
-1.17
(-1.94,-
0.39)
Lower
back RPE
7.67
(2.17)
10.63
(2.62)
11.07
(2.70)
-2.97
(-3.85, -2.09)
-3.40
(-4.24, -2.56)
-0.43
(-1.12,
0.25)
Lower
extremity
RPE
8.47
(2.61)
10.03
(2.63)
10.80
(3.08)
-1.57
(-2.53, -0.61)
-2.33
(-3.12, -1.55)
-0.77
(-1.38,-
0.15)
PSPSI, perceived sense of postural sway and instability; RPE, rating of perceived exertion;
Significant differences are in bold (p<0.05).
Table 3. Mean (standard deviation) and mean difference (95% CI) values during unloaded, high load placement and low load placement
Mean
(SD)
Mean difference
(95% CI)
Variable Unloaded
HLP
LLP
Unloaded-
HLP
Unloaded-
LLP
HLP- LLP
Average
velocity
0.73
(0.15)
0.88
(0.19)
0.88
(0.22)
-0.15
(-0.19, -
0.10)
-0.16
(-0.21, -0.1)
-0.01
(-0.06, 0.04)
Sway
area
2.43
(1.21)
4.27
(2.28)
3.90
(1.80)
-1.84
(-2.63, -
1.05)
-1.48
(-2.06, -
0.89)
0.36
(-0.34, 1.07)
PSPSI 0.21
(0.42)
2.29
(1.82)
1.90
(1.49)
-2.08
(-2.72, -
1.44)
-1.69
(-2.20, -
1.18)
0.39
(-0.24, 1.01)
Overall
RPE
7.43
(1.48)
12.17
(2.20)
11.80
(2.22)
-4.73
(-0.54, -
3.93)
-4.37
(-5.11, -
3.62)
0.37
(-0.27, 1.01)
Neck 7.03 11.70 11.03 -4.67 -4.00 0.67
112
RPE (1.27) (3.09) (3.00) (-5.88, -
3.45)
(-5.16, -
2.84)
(-0.90, 1.42)
shoulder
RPE
6.67
(1.74)
13.50
(2.56)
13.10
(2.45)
-6.83
(-7.98, -
5.68)
-6.43
(-7.55, -
5.32)
0.40
(-0.30, 1.10)
Upper
back
RPE
7.00
(2.12)
12.23
(2.78)
11.17
(2.34)
-5.23
(-6.47, -
3.99)
-4.17
(-5.33, -
3.01)
1.07
(0.29, 1.84)
Lower
back
RPE
7.33
(2.37)
11.33
(2.29)
11.73
(2.59)
-4.00
(-5.13, -
2.87)
-4.40
(-5.65, -
3.15)
-0.40
(-1.16, 0.36)
Lower
extremity
RPE
8.17
(1.90)
10.63
(2.54)
10.63
(2.40)
-2.47
(-3.29, -
1.65)
-2.47
(-3.24, -
1.70)
0.00
(-0.90, 0.90)
HLP, High load placement; LLP, Low load placement, PSPSI, perceived sense of postural
sway and instability; RPE, rating of perceived exertion; Significant differences are in bold
(p<0.05).
Static testing conditions
The omnibus tests identified significant differences in average velocity and
sway area in both hip belt and load placement comparisons. Subsequent
post hoc testing identified significant increases in average velocity and sway
area between the unloaded and loaded conditions. There were no significant
differences between the hip belt and load placement conditions.
Dynamic testing conditions
The omnibus test indicated significant differences in perceived stability,
discomfort and overall and local RPE. Post hoc analyses identified lower
perceived stability and comfort and higher overall and local RPE in the
loaded conditions than in the unloaded condition. Additionally, lower
perceived stability and higher RPE in most bodily regions, except the lower
back, was reported in the no hip belt condition compared to the hip belt
condition (see figure 3).
113
Figure 3. Perceived exertion of neck, shoulders, upper back, lower back, lower extremity in unloaded, hip belt and no hip belt conditions
While upper back RPE was higher during the high load placement condition
compared to the low load placement condition, no differences were found in
perceived stability as well as overall, neck, shoulder, lower back and lower
extremity RPE between the two placements (see figure 4).
114
Figure 4. Perceived exertion of neck, shoulders, upper back, lower back and lower extremity in unloaded, high and low load placements
Discussion
The purpose of this study was to assess the effect of load placement in a
backpack and hip belt use on subjective and objective measures of postural
stability. Our findings demonstrated that postural stability decreased when
wearing a backpack. However, load placement and hip belt use did not affect
measures of postural stability.
Decreased postural stability can result from mechanical and physiological
mechanisms. From a physiological perspective, it has been shown that
respiration affects postural stability [1]. Carrying external loads results in
increased demand on the respiratory system and as a result of this,
cardiovascular systems are also influenced. This influence is reflected by
increased heart rate, blood pressure and oxygen consumption [26]. As these
changes occur, the motion of blood mass and cardiopulmonary organs
115
increases, resulting in lower postural stability as a result of the displacement
of the COM [27].
From a mechanical point of view, the human frame is intrinsically unstable,
[1] and carrying a backpack results in decreased postural stability [5]. When
a backpack is carried, a vertical projection of the COM of the body shifts
anteriorly, [28] and the displacement of the COM reduces postural stability.
Decreased postural stability in loaded conditions compared to the unloaded
condition in the current study is comparable with other studies [6, 29] but
conflicts with one study [30]. Despite the similarities between the
methodology of the current study and that study [30], it is not clear why they
failed to observe any difference in COP variables between the unloaded and
loaded conditions.
The participants in our study perceived lower stability in the loaded conditions
compared to the unloaded condition. This finding agrees with a study that
reported decreased perceived stability when carrying 10% and 20% of body
weight load compared to an unloaded condition [6].
We did not identify differences in average velocity and sway area between
the hip belt and no hip belt conditions. However, participants reported higher
perceived stability when using a hip belt. Maintaining balance depends in part
on muscular activity to counterbalance torque resulting from internal and
external forces [31]. While we did not measure muscle activity in this study,
previous research has identified lower levels of muscle activation when a hip
belt was used [8, 32]. Changes in postural balance in the no hip belt
condition did not deviate significantly from what we observed in the hip belt
condition, which is likely to be due to the compensatory increased muscle
activation in the no hip belt condition.
While we found no differences in objective measures of postural stability
between the hip belt and no hip belt conditions, subjective measures of
postural stability showed that in the hip belt condition, participants perceived
greater stability compared to the no hip belt condition. No information is
available in the literature on the effect of hip belt use on objective and
subjective measures of postural stability. However, a study that compared
116
the level of comfort between different backpacks stated that subjective
measures might be more sensitive as well as more useful and appropriate
when comparing different backpacks, especially when the difference between
the backpacks is small [33]. However, subjective measures may also include
information from other domains that differ from objective measures of
postural stability.
Not surprisingly, RPE was greater during the loaded conditions compared to
the unloaded condition, and is consistent with previous research [34, 35]. In
our study, overall and local RPE measured at the neck, shoulder, upper back
and lower extremity decreased with hip belt use. Conversely, lower back
region RPE was not different between the hip belt and no hip belt conditions.
Decreased local RPE at the shoulder and upper back regions may result
from decreased interface pressure under the shoulder straps. It has been
noted that carrying a 10 kg backpack resulted in a shoulder–backpack
interface pressure of 203 mmHg, which decreased to 15 mmHg when a hip
belt was applied [32]. Additionally, hip belt use results in less shoulder
muscular activation [8, 32].
No differences in postural stability between the high and low load placement
conditions were found. These findings are consistent with another study
comparing COP variables between the high and low load placement
conditions when a 10% and 20% body weight load was applied to a
backpack [6]. Similarly, participants reported no difference in perceived
stability during the high and low load placement conditions. Although
previous research has not compared subjective measures of stability during
walking between different load placements, our findings accord with studies
reporting objective measures [36, 37]. One explanation for why participants
did not feel less stable in the high load placement in the current study might
be that the body maintains its stability in high load placement by decreasing
walking velocity and increasing double limb support [37]. However,
spatiotemporal parameters have not been investigated in this study and we
cannot confirm this assumption.
117
The results of this study are limited, as we only examined a load of 20% body
weight for a short duration. Therefore, our results may not generalise to other
loading conditions. Additionally, it was not possible to blind participants to the
measurement condition, and this may be a source of bias in the results.
A potential limitation was that there was not full standardization regarding
footwear and clothing of the participants and this may have led to a decrease
in the internal validity of the study. However we chose the middle ground
standardisation to increase the external validity of the study and to generalise
the results to the broader population.
Future research should incorporate electromyography measures of postural
and lower extremity musculature in different load placements, hip belt and no
hip belt conditions.
In conclusion carrying a loaded backpack reduces postural stability, while low
or high load placement in a backpack does not affect subjective and objective
measures of postural stability. While the use of a hip belt resulted in
improvements in perceived stability and exertion, there were no
corresponding improvements in postural stability. Therefore, based on the
findings of this study, hip belt use can be recommended, while high or low
load placement cannot.
Conflict of interest statement
This research was funded by Promopak Pty Ltd. Perth, WA; however, the
manufacturer played no part in the design, conduct or reporting of this study.
References
1. Schmid, M., et al., Respiration and postural sway: Detection of phase
synchronizations and interactions. Human Movement Science, 2004. 23(2):
p. 105-119.
2. Winter, D.A., Human balance and posture control during standing and
walking. Gait and Posture, 1995. 3(4): p. 193-214.
3. Horak, F.B., S.M. Henry, and A. Shumway-Cook, Postural
perturbations: New insights for treatment of balance disorders. Physical
Therapy, 1997. 77(5): p. 517-533.
118
4. Haas, B.M. and A.M. Burden, Validity of weight distribution and sway
measurements of the Balance Performance Monitor. Physiotherapy research
international : the journal for researchers and clinicians in physical therapy,
2000. 5(1): p. 19-32.
5. Li, X. and A.S. Aruin, The effect of short-term changes in body mass
distribution on feed-forward postural control. Journal of Electromyography
and Kinesiology, 2009. 19: p. 931-941.
6. Talbott, N.R., the effect of weight, location and type of backpack on
posture and postural stability of children. 2005, University of Cincinnati.
7. LaFiandra, M. and E. Harman, The Distribution of Forces between the
Upper and Lower Back during Load Carriage. Medicine and Science in
Sports and Exercise, 2004. 36(3): p. 460-467.
8. Southard, S.A. and G.A. Mirka, An evaluation of backpack harness
systems in non-neutral torso postures. Applied Ergonomics, 2007. 38(5): p.
541-547.
9. Mackie, H.W., et al., The effect of simulated school load carriage
configurations on shoulder strap tension forces and shoulder interface
pressure. Applied Ergonomics, 2005. 36(2): p. 199-206.
10. Vacheron, J.J., et al., The effect of loads carried on the shoulders.
Military Medicine, 1999. 164(8): p. 597-599.
11. Bessen, R.J., V.W. Belcher, and R.J. Franklin, Rucksack paralysis
with and without rucksack frames. Military Medicine, 1987. 152(7): p. 372-
375.
12. Sharpe, S.R., et al., Effects of a hip belt on transverse plane trunk
coordination and stability during load carriage. Journal of Biomechanics,
2008. 41(5): p. 968-976.
13. Knapik, J.J., K.L. Reynolds, and E. Harman, Soldier Load Carriage:
Historical, Physiological, Biomechanical, and Medical Aspects. Military
Medicine, 2004. 169(1): p. 45-56.
14. Holewijn, M. and W.A. Lotens, The influence of backpack design on
physical performance. Ergonomics, 1992. 35(2): p. 149-157.
15. Golriz, S. and B.F. Walker, Bacakpacks. Several factors likely to
influence design and usage: A systematic literature review. Work, 2012.
42(4): p. 1-12.
119
16. Brackley, H.M., J.M. Stevenson, and J.C. Selinger, Effect of backpack
load placement on posture and spinal curvature in prepubescent children.
Work, 2009. 32(3): p. 351-360.
17. Stuempfle, K.J., D.G. Drury, and A.L. Wilson, Effect of load position on
physiological and perceptual responses during load carriage with an internal
frame backpack. Ergonomics, 2004. 47(7): p. 784-789.
18. Liu, B.S., Backpack load positioning and walking surface slope effects
on physiological responses in infantry soldiers. International Journal of
Industrial Ergonomics, 2007. 37(9-10): p. 754-760.
19. Bobet, J. and R.W. Norman, Effects of load placement on back muscle
activity in load carriage. European Journal of Applied Physiology and
Occupational Physiology, 1984. 53(1): p. 71-75.
20. Urbaniak, G.C. and S. Plous. Research Randomizer (version 3.0).
2011; Available from: http://www.randomizer.org/.
21. Golriz, S., et al., The validity of a portable clinical force plate in
assessment of static postural control: concurrent validity study. Chiropractic
and Manual therapies, 2012. 20(1): p. 15.
22. Chiou, S.Y., A. Bhattacharya, and P.A. Succop, Evaluation of workers'
perceived sense of slip and effect of prior knowledge of slipperiness during
task performance on slippery surfaces. American Industrial Hygiene
Association Journal, 2000. 61(4): p. 492-500.
23. Skinner, J.S., et al., The validity and reliability of a rating scale of
perceived exertion. Medicine and Science in Sports and Exercise, 1973. 5(2):
p. 94-96.
24. Faul, F., et al., Tests for correlation and regression analyses. Behavior
Research Methods, 2009. 41: p. 1149-1160.
25. Streiner, D.L. and G.R. Norman, Health measurement scales. A
practical guide to their development and use. Third ed. 2003: Oxford Medical
Publictions.
26. Hong, Y., et al., Effects of load carriage on heart rate, blood pressure
and energy expenditure in children. Ergonomics, 2000. 43(6): p. 717-727.
27. Conforto, S., et al., Hemodynamics as a possible internal mechanical
disturbance to balance. Gait and Posture, 2001. 14(1): p. 28-35.
28. Goh, J.H., A. Thambyah, and K. Bose, Effects of varying backpack
loads on peak forces in the lumbosacral spine during walking. Clinical
Biomechanics, 1998. 13(SUPPL. 1): p. S26-S31.
120
29. Heller, M.F., J.H. Challis, and N.A. Sharkey, Changes in postural sway
as a consequence of wearing a military backpack. Gait and Posture, 2009.
30(1): p. 115-117.
30. Zultowski, I. and A. Aruin, Carrying loads and postural sway in
standing: The effect of load placement and magnitude. Work, 2008. 30(4): p.
359-368.
31. Balasubramaniam, R. and A.M. Wing, The dynamics of standing
balance. Trends in Cognitive Sciences, 2002. 6(12): p. 531-536.
32. Holewijn, M., Physiological strain due to load carrying. European
Journal of Applied Physiology and Occupational Physiology, 1990. 61(3-4): p.
237-245.
33. Legg, S.J., L. Perko, and P. Campbell, Subjective perceptual methods
for comparing backpacks. Ergonomics, 1997. 40(8): p. 809-817.
34. Bauer, D.H. and A. Freivalds, Backpack load limit recommendation for
middle school students based on physiological and psychophysical
measurements. Work, 2009. 32(3): p. 339-350.
35. Madras, D.E., M.W. Cornwall, and J.R. Coast, Energy cost, perceived
exertion and postural adjustments when treadmill walking with two types of
backpack. Journal of Human Movement Studies, 1998. 35(5): p. 233-249.
36. Simpson, K.M., B.J. Munro, and J.R. Steele, Does load position affect
gait and subjective responses of females during load carriage? Applied
Ergonomics, 2012. 43(3): p. 479-485.
37. Devroey, C., et al., Evaluation of the effect of backpack load and
position during standing and walking using biomechanical, physiological and
subjective measures. Ergonomics, 2007. 50(5): p. 728-742.
121
Chapter Seven
THE EFFECT OF BACKPACK LOAD PLACEMENT ON
PHYSIOLOGICAL AND SELF–REPORTED MEASURES OF
EXERTION
Submitted to
Human Movement Sciences
122
In chapter 6, we examined the effect of backpack load placement on
subjective and objective measures of postural stability and discovered that
there is no significant difference between different load placements. In this
chapter we will examine the effect of load placement on physiological and
self reported measures of exertion and economy of movement to further
explore the difference between load placements.
This chapter has resulted in the following submission for publication:
Golriz S, Peiffer JJ, Walker BF, Foreman KB, Jeffrey JJ. The effect of
backpack load placement on physiological and self–reported measures of
exertion. Submitted to Human Movement Sciences
123
The effect of backpack load placement on physiological and self–
reported measures of exertion
Samira Golriz 1, Jeremiah J Peiffer 1, Bruce F Walker 1, K Bo Foreman 2,
Jeffrey J Hebert 1
1, School of Chiropractic and Sports Science, Murdoch University, Perth,
Australia
2, Department of Physical Therapy, University of Utah, Salt Lake City, UT,
USA
Samira Golriz* * Corresponding author
Email: [email protected]
90 South Street, Murdoch 6150, WA, Australia
Tel: +61 8 93601450; Fax: +61 8 93601299
Jeremiah Peiffer
Email: [email protected]
Bruce F Walker
Email: [email protected]
K Bo Foreman
Email: [email protected]
Jeffrey J Hebert
Email: [email protected]
Conflicts of Interest and Source of Funding
This research was funded by Promopak Pty Ltd. Perth, WA; however, the
manufacturer played no part in the design, conduct or reporting of this study.
Results of the present study do not constitute endorsement by the American
College of Sports Medicine.
124
The effect of backpack load placement on physiological and self–
reported measures of exertion
Abstract
Introduction: Backpacks are the most popular way of carrying additional
weight; however, it puts the body under physical stress and may cause
discomfort. It may also increase oxygen demand and energy cost.
Manipulation of load placement may alleviate the effects carrying a backpack
has on the body.
Purpose: The purpose of this study was to examine for differences in the
physiological and self-reported measures of exertion, movement economy
and efficiency of carrying a loaded backpack in both high and low placement
compared to an unloaded control condition.
Methods: Fifteen healthy adults were screened with the American Heart
Association/ American College of Sports Medicine health-fitness facility
participation screening questionnaire and instructed to walk on a treadmill for
ten minutes during three sessions under three load conditions: no load,
carrying a 20% of their body weight in high load placement and low load
placement. Dependent variables were measured using a metabolic
measurement system and participants rated their perceived exertion on a
Borg scale. The order of conditions was randomised. The variables were
analysed using separate, one–way repeated measures ANOVA.
Results: Carrying a 20% body weight load produced a significant increase in
VO2, minute ventilation, heart rate, movement economy and overall
perceived exertion in both the high and low load placements compared with
the unloaded condition. However no difference was observed between the
high and low load placement conditions. Nevertheless, perceived exertion
reported at the shoulder region was higher for high load placement as
compared to low load placement condition.
Conclusion: While altering load placement did not influence physiological
variables or overall RPE (rating of perceived exertion), participants reported
125
lower perceived exertion on the shoulders in low load placement and low
load placement might be preferable in this respect.
Keywords: load carriage, loading configuration, economy, VO2
126
Introduction
Carrying heavy loads is a necessity for many individuals and backpacks are
often considered the most convenient and appropriate way of achieving this
[1-3]. Nevertheless, carrying heavy loads in a backpack can induce increased
physical stress and cause discomfort and pain [4]. Additionally, increased
oxygen consumption and energy cost have been observed when carrying
heavy loads in backpacks compared with unloaded conditions [5, 6].
Quesada et al. [5] reported increased energy cost during walking while
carrying 15% and 30% body weight in comparison to an unloaded condition.
Patton et al. [6] also compared the energy cost of carrying a 31.5 and 49.4 kg
load with an unloaded condition and reported higher energy costs when
carrying extra loads.
In addition to the weight of the load, factors such as load placement may
influence physical stress and energy cost [7, 8]. It has been suggested loads
should be located centrally on the trunk and not carried asymmetrically, as in
a briefcase, handbag, shoulder bag or backpack carried unilaterally [8, 9].
Furthermore, in relation to the sagittal plane, the load needs to be carried
close to the centre of mass of the body [10], thus minimising the moment
created by the load [11].
Despite the recommendations in relation to frontal and sagittal load carrying,
controversy exists in relation to the optimal load placement in the axial plane
and the importance of load placement within a backpack [12]. Stuempfle et
al. [7] reported large differences between high and low load placement in a
backpack in terms of oxygen uptake (VO2) and minute ventilation; however,
they only examined female participants. Load placement is even more
important during walking as dynamic movements can increase load moment
up to 40% due to rotational inertia [13]. On the other hand, Liu [14] found no
differences in physiological measures between high and low load placement
while carrying 15% body weight at different speeds and grades.
With regard to load placement in a backpack, the literature provides
conflicting results on physiological variables and is limited regarding
movement economy, efficiency and subjective measures such as perceived
127
exertion. Most studies have concentrated on the effect of load placement on
physiological variables, while it has also been stated that subjective
measures are more sensitive and appropriate for comparing load carriage
systems.[2] In addition, when it comes down to daily use, it is the subjective
measures that present our perception of exertion. Therefore, subjective
measures should be a powerful source of information for comparing different
load placements in a backpack. Moreover, load placement conditions have
only been compared against each other in the literature and not against an
unloaded control condition [7, 14]. It would be helpful to investigate what type
of load placement is the closest to an unloaded condition.
The purpose of this study was to examine for differences in physiological and
self–reported measures of exertion, movement economy and efficiency when
carrying a loaded backpack in both a high and low load placement compared
to an unloaded control condition. The body can be thought of as an inverted
pendulum [15] and we hypothesised that elevation of the centre of mass in
high load placement will maximise postural displacement and attenuate
imbalance, meaning the body requires more energy to maintain equilibrium
[16]. In addition, it has been reported high load placement led to a greater
trunk inclination angle [17, 18] and it is believed a greater trunk inclination
angle and displacement of the centre of mass results in lower movement
economy [19].
Methods
Participants
We performed a one–way repeated measures study examining the effects of
three loading conditions on physiological and self–reported measures of
exertion, movement economy and efficiency. Fifteen healthy participants
volunteered to participate in this study. All participants were screened with
the American Heart Association/American College of Sports Medicine
health/fitness facility participation screening questionnaire [20] prior to
inclusion. Participants were included if they were 18 to 45 years of age.
Individuals were excluded from participation if they had heart disease,
asthma, lung disease, diabetes and/or musculoskeletal problems that limited
128
their physical activity or if they were pregnant. They were also excluded if
they were deemed moderate or high risk according to the screening
questionnaire (see appendices of chapter 7, page 174).
Prior to data collection, all participants signed a consent form approved by
the Human Research Ethics Committee of Murdoch University (approval
number 2011/165).
Baseline evaluation and maximal graded exercise test
All participants completed a baseline maximal graded exercise test and three
experimental exercise sessions. To prevent carryover effects, we allocated a
duration of at least 48 hours between the maximal graded exercise test and
the first experimental session. All experimental sessions were separated by
no less than one and no more than seven days. Participants wore
comfortable light clothing and abstained from intense physical activity for 24
hours prior to testing and fasted for three hours prior to each testing session.
Participants then completed a maximal graded exercise test on a
Trackmaster treadmill (JAS fitness systems, model TMX 55, KS, USA) during
which the volume of expired oxygen (VO2) and carbon dioxide (VCO2) was
measured using a Pravo TrueOne 2400 metabolic cart (ParvoMedics, Sandy,
UT, USA). The TrueOne metabolic measurement system has been shown to
be a reliable and accurate system to measure gas exchange variables [21].
We calibrated the metabolic cart before use using gasses of known
concentration and through a range of flow rates. Prior to testing, participants
were fitted with a Polar heart rate monitor telemetry chest strap (model
number T31 Coded- Polar Electro, Kempele, Finland) and heart rate was
recorded continuously throughout the duration of the test. The maximal
graded exercise test started at a speed of 3 km.h-1 with a 1% grade for the
first minute and an increase of 1 km∙h-1 each minute thereafter until a
respiratory exchange ratio (RER) equal to 1.0 was achieved and maintained
for one minute. After identifying the RER, the treadmill grade was increased
by 2% each minute until VO2max was achieved. Participants were deemed to
have achieved VO2max if their RER exceeded 1.1, heart rate was greater
129
than 85% of age–predicted max and rating of perceived exertion (RPE)
exceeded 17 [22].
Experimental sessions
During the remaining three sessions, the participants completed three 10-
minute walking trials, in a randomised and counterbalanced order, using
three backpack loading conditions: unloaded, high load placement and low
load placement. Participants walked for ten minutes on a treadmill at a grade
of 1.0% at the individualised speed while continuously measuring heart rate
and expired ventilation. Walking speed was identified by a RER of 1.0 during
the maximal graded exercise test and was constant during all experimental
conditions. We chose a constant grade of 1.0% as it accurately simulates
outdoor running and compensates for the lack of air resistance in treadmill
running [23]. A duration of ten minutes was selected to ensure steady–state
physiological measures were achieved during exercise [3, 5]. Only the final
two minutes of data collection were used in the data analysis to ensure all
variables represented steady–state values.
We used the following formulas to calculate our variables of interest:
- Efficiency (%)= (work rate / energy expended) x 100% [24].
Work rate (kcal.min-1) = weight (kg) [body weight + backpack weight] x speed
of treadmill (m/min) x sin Θ (the angle of treadmill inclination).
Energy Expenditure (kcal.min-1) = 3.9 VO2+ 1.1 VCO2 [25].
-Economy (kj.L-1) = speed (m.min-1) / VO2 [24].
In addition, participants rated their overall and regional (neck, shoulder, back
and lower extremities) perceived exertion using a Borg scale (6-20) [26]
immediately upon completion of each testing session.
Backpack configuration
A backpack (Promopak Pty Ltd, Australia) measuring of 47 x 21 x 15 cm was
used for all experimental sessions. The backpack included two main
compartments, adjustable padded shoulder straps, adjustable non-elastic hip
130
and elastic sternum belts, two side stabiliser straps and plastic quick-release
buckles. Consistent placement of the backpack on the participant across the
loading conditions was achieved by measuring the length of the shoulder
straps and checking their equality. We divided the interior of the backpack
into upper and lower compartments (see figure 1). In the high and low load
placement, we placed loads in the upper or lower compartments of the
backpack respectively and loads were held in place with high–density foam
inserts. The unloaded condition was defined as wearing no backpack. We
utilised a backpack load equal to 20% of the participant’s bodyweight and not
exceeding 20 kg as 20% body weight has been previously used in the
relevant literature [3, 27] and reflects common recommendations that the
backpack weight should be in the range of 15% - 30% body weight [5, 28].
Figure 1. The loading conditions in the backpack
a) high load placement b) low load placement
Statistical analysis
Using VO2 data from a previous study [29], we performed an a priori power
analysis using G-Power 3.1.3 software [30] for a priori sample size
estimation. Assuming an alpha level of 0.05 and within–group standard
deviation of 0.58 ml.kg-1.min-1, recruiting 13 participants provides 80% power
to detect a change in VO2. To account for an expected drop out rate of 15%,
we recruited 15 participants to ensure sufficient power. We entered data into
SPSS version 17.0 and calculated descriptive statistics including mean and
standard deviation for all dependent variables.
131
We assessed the normality of the data with the Shapiro–Wilk test and
sphericity using Mauchly’s test. The data from the high and low load
placement and unloaded condition were compared using separate one–way
repeated measures analysis of variance (ANOVA) to examine for differences
between loading conditions. Least significant difference was adopted for post
hoc pair wise comparisons. The independent variable was the loading
condition with three levels (unloaded, high load placement, low load
placement), and the dependent variables were VO2 (L.min-1), minute
ventilation (L.min-1), heart rate (beats.min-1), movement economy, efficiency
and RPE (overall, neck, shoulders, upper back, lower back, lower extremity).
An alpha level of 0.05 was used to indicate statistical significance.
Results
All data measured in this study were normally distributed. None of the
participants dropped out of the study. Participant demographic and baseline
data are presented in table 1. The means, standard deviations and mean
differences (95% CI) of the dependent variables between loading conditions
are reported in table 2.
Table 1: Demographic and baseline characteristics of participants (N=15)
Variables Value
Age (years) 30.4 (5.5)
Sex (N) 8 female
BMI 23.3 (5.2)
Weight (kg)
Height (cm)
71.1 (22.7)
173.3 (11.3)
VO2 max (L.min-1) 2.5 (1.0)
Heart rate max (beats.min-1) 167 (16)
Backpack mass (kg) 13.1 (3.5)
Values are mean (standard deviation) unless otherwise indicated
132
The omnibus tests revealed significant differences in VO2 (p<0.05), however
the magnitude of the difference was very small and unlikely to represent an
important difference. The omnibus tests also revealed significant differences
in minute ventilation (p<0.001), movement economy (p<0.001), heart rate
(p<0.05) and overall (p<0.001) and local RPEs. Post hoc analysis identified
increased VO2 (p<0.05), minute ventilation (p<0.05) and decreased economy
(p<0.001) in loaded conditions as compared to the unloaded condition, while
no difference (p>0.05) was observed between the high and low load
placement conditions. Increased heart rate was seen in high load placement
as compared to the unloaded condition (p<0.05), while no significant
difference (p>0.05) was observed between unloaded–low load placement
and high–low load placements. Overall and local RPEs increased in loaded
conditions as compared to the unloaded condition (p<0.001). Increased
shoulder RPE was identified in high load placement as compared to low load
placements (p<0.05), while no difference (p>0.05) was observed in overall,
neck, back and lower extremity RPEs between high and low load
placements. There were no between–group differences in efficiency
(p>0.05).
Table 2: Mean (standard deviation), mean difference (95% CI) values
during unloaded, high and low load placement conditions
Mean (SD) Mean difference
(95% CI)
Variables Unloaded
High load
Low load
Unloaded-
high load
Unloaded-
low load
High
load-
low load
VO2
(L.min-1) 1.1 (0.4) 1.2 (0.5) 1.2 (0.5)
-0.1
(-0.2, 0.0)
-0.2
(-0.3, 0.0)
0.0
(-0.1,0.0)
MV
(L.min-1) 27.5 (9.3) 31.7(11.0) 32.7(11.9)
-4.2
(-5.8, -2.5)
-4.7
(-7.4,-1.9)
-0.5
(-2.4,1.3)
HR
(beats.
min-1)
117 (20) 127 (21) 124 (16) -11
(-19, -3)
-8
(-19, 2)
2
(-5, 10)
Economy 92.8(25.0) 82.9(22.4) 81.6(22.2) 9.9 11.2 1.3
133
(kj.L-1) (4.7, 15.1) (5.3, 16.9) (-1.1,3.6)
Efficiency
(%) 18.9 (2.3) 19.6 (1.9) 19.4 (2.3)
-0.7
(-1.7, 0.3)
-0.5
(-1.6, 0.7)
0.3
(-0.3,0.8)
Overall
RPE 8.6 (2.0) 13.1 (2.3) 12.3 (3.3)
-4.5
(-5.6, -3.3)
-3.7
(-5.3, -2.1)
0.80
(-0.5,2.1)
Shoulder
RPE 7.7 (0.4) 14.7 (0.7) 13.0 (0.9)
-7.0
(-8.4, -5.6)
-5.3
(-6.9, -3.8)
1.7
(0.2, 3.2)
HR, heart rate; MV, minute ventilation, RPE, rating of perceived exertion; VO2, oxygen
uptake; Significant differences are in bold (p<0.05).
Discussion
We examined for differences in physiological and self-reported measures of
exertion, movement economy and efficiency of carrying a loaded backpack in
high and low load placements compared to an unloaded condition. VO2,
minute ventilation and overall RPE increased and movement economy
decreased in loaded conditions compared to the unloaded condition;
however, there were no differences between high and low load placements.
Additionally, shoulder region RPE was higher in high load placement as
compared to low load placement and yet load and load placement did not
influence efficiency.
In our study we observed an increase in VO2 and minute ventilation and a
decrease in movement economy from the unloaded to loaded conditions that
were consistent with previous research [1, 3, 5], although the sample
population of the studies were different. Carrying a loaded backpack causes
some modifications in the body, such as increased trunk forward inclination
[28], as well as increased activity of respiratory and postural muscle activity
[18, 31]. It has been reported that increased muscular activity and changes in
posture can lead to higher oxygen consumption, minute ventilation and lower
movement economy [31, 32]. The results of the current study showed
increased VO2, decreased movement economy and unchanged efficiency.
The lack of change in efficiency across the conditions evidenced that, in
contrast to the abovementioned theory, increased VO2 and decreased
economy is mostly as a result of the additional loading and a small
134
component is due to the normal resultant changes in posture or muscle
activity associated with carrying load.
Load placement did not elicit a difference in VO2, minute ventilation or
movement economy. Carrying an additional 20% of body weight led to
greater VO2 and minute ventilation and lower movement economy regardless
of load placement. Our findings are consistent with some studies [14, 33] but
not the study conducted by Stuempfle et al. [7]. Stuempfle et al. studied
females and used a constant weight of 25 kg which was about 35% of
participants’ body weight and was considerably heavier than the load used in
the current study.
Consistent with previous research, we did not identify differences in overall
RPE between the high and low load placement conditions [17, 18, 34] but
contradicts the finding of Stuempfle et al. [7], who reported high load
placement is the most desirable backpack placement on the spine. Unlike the
overall RPE, the shoulder region RPE was greater during the high load
placement condition as compared to low load placement. This finding is
comparable to the results of Frank et al. [35], who reported higher forces on
the shoulders of children while carrying backpacks with high load placement.
Previous research has reported RPE incorporates sensations of exertion
originating from two sources, the active muscles and the cardiopulmonary
system [36]. RPE values of the shoulder area were higher during high load
placement; we observed no corresponding increase in VO2, heart rate and
overall RPE. One explanation for this could be higher levels of activity of the
shoulder girdle muscular system. While we measured physiological variables,
we did not investigate muscular activity in this study. However, Bobet and
Norman [37], reported higher activity in the upper trapezius muscle during
high load placement (the load’s centre of mass was located at the level of the
ear lobe) compared to low load placement (the load’s centre of mass was
located at the level of the xiphoid process). They suggested this difference
may arise from the angular and linear accelerations of the load and trunk
[37]; however, their definitions of high and low load placement differed from
those of the current study. The second explanation could be localised
muscular fatigue due to strap pressure during high load placement. Fatigue
135
of local muscles has been reported as a limiting factor in load carriage [38].
Local muscular fatigue is more likely due to local ischemia or lactate
accumulation and not limitations in aerobic muscular processes [2]. Third,
load carriage leads to mental detection of fatigue even when physiological
variables are not influenced. Higher perception of effort in the shoulder area
during high load placement may override constant physiological variables
and overall RPE values [39].
This study had several strengths and weaknesses. We obtained both
objective physiological measures and self-reported measures which may
measure different aspects of exertion. We used a standardised method;
participants walked at a constant forced pace and were therefore unable to
modify their walking velocity. They might have adapted differently to load and
load placement if they had been able to walk at a self-selected pace. Walking
on a treadmill may not replicate a real-world scenario (speed, gradient,
duration). Therefore, while this approach supports the internal validity of our
study, the generalisability of our results is limited in this sense.
This study revealed that physiological and perceptual variables of exertion
are more vulnerable to backpack load than load placement. Although it has
previously been recommended to pack heavier items at the top and lighter
items at the bottom of the backpack [7], this study showed no difference
exists between the high and low load placements in most of the variables
studied. While altering the load placement may not change physiological
variables or overall RPE, participants reported lower perceived exertion on
the shoulders in low load placement, thus low load placement might be
preferable in this respect.
Conflicts of Interest and Source of Funding
This research was funded by Promopak Pty Ltd., Perth, WA; however, the
manufacturer played no part in the design, conduct or reporting of this study.
Results of the present study do not constitute endorsement by the American
College of Sports Medicine.
136
References
1. Pal, M.S., et al., Optimum load for carriage by soldiers at two walking
speeds on level ground. International Journal of Industrial Ergonomics, 2009.
39(1): p. 68-72.
2. Legg, S.J., L. Perko, and P. Campbell, Subjective perceptual methods
for comparing backpacks. Ergonomics, 1997. 40(8): p. 809-817.
3. Hong, Y., et al., Effects of load carriage on heart rate, blood pressure
and energy expenditure in children. Ergonomics, 2000. 43(6): p. 717-727.
4. Golriz, S. and B.F. Walker, Can load carriage system weight, design
and placement affect pain and discomfort? A systematic review. Journal of
back and musculoskeletal rehabilitation 2011. 24(1): p. 1-16.
5. Quesada, P.M., et al., Biomechanical and metabolic effects of varying
backpack loading on simulated marching. Ergonomics, 2000. 43(3): p. 293-
309.
6. Patton, J.F., et al., Physiological responses to prolonged treadmill
walking with external loads. European Journal of Applied Physiology and
Occupational Physiology, 1991. 63(2): p. 89-93.
7. Stuempfle, K.J., D.G. Drury, and A.L. Wilson, Effect of load position on
physiological and perceptual responses during load carriage with an internal
frame backpack. Ergonomics, 2004. 47(7): p. 784-789.
8. Zultowski, I. and A. Aruin, Carrying loads and postural sway in
standing: The effect of load placement and magnitude. Work, 2008. 30(4): p.
359-368.
9. Macias, B.R., et al., Asymmetric loads and pain associated with
backpack carrying by children. Journal of pediatric orthopedics, 2008. 28(5):
p. 512-517.
10. Knapik, J.J., et al., Soldier performance and strenuous road marching:
Influence of load mass and load distribution. Military Medicine, 1997. 162(1):
p. 62-67.
11. Knapik, J., E. Harman, and K. Reynolds, Load carriage using packs: A
review of physiological, biomechanical and medical aspects. Applied
Ergonomics, 1996. 27(3): p. 207-216.
12. Golriz, S. and B.F. Walker, Bacakpacks. Several factors likely to
influence design and usage: A systematic literature review. Work, 2012.
42(4): p. 1-12.
137
13. Bloom, D. and A.P. Woodhull-McNeal, Postural adjustments while
standing with two types of loaded backpack. Ergonomics, 1987. 30(10): p.
1425-1430.
14. Liu, B.S., Backpack load positioning and walking surface slope effects
on physiological responses in infantry soldiers. International Journal of
Industrial Ergonomics, 2007. 37(9-10): p. 754-760.
15. Winter, D.A., et al., An integrated EMG/biomechanical model of upper
body balance and posture during human gait. Progress in brain research,
1993. 97: p. 359-367.
16. Pedersen, A.V., R. Stokke, and A. Mamen, Effects of extra load
position on energy expenditure in treadmill running. European Journal of
Applied Physiology, 2007. 102(1): p. 27-31.
17. Brackley, H.M., J.M. Stevenson, and J.C. Selinger, Effect of backpack
load placement on posture and spinal curvature in prepubescent children.
Work, 2009. 32(3): p. 351-360.
18. Devroey, C., et al., Evaluation of the effect of backpack load and
position during standing and walking using biomechanical, physiological and
subjective measures. Ergonomics, 2007. 50(5): p. 728-742.
19. Lloyd, R. and C. Cooke, Biomechanical differences associated with
two different load carriage systems and their relationship to economy. Human
Movement, 2011. 12(1): p. 65-74.
20. ACSM, ACSM's guidelines for exercise testing and prescription. 8th
ed. 2009, Philadelphia: Wolters Kluwer Health Ltd.
21. Crouter, S.E., et al., Accuracy and reliability of the ParvoMedics
TrueOne 2400 and MedGraphics VO2000 metabolic systems. European
Journal of Applied Physiology, 2006. 98(2): p. 139-151.
22. Peiffer, J.J., et al., Physiological characteristics of masters-level
cyclists. Journal of Strength and Conditioning Research, 2008. 22(5): p.
1434-1440.
23. Jones, A.M. and J.H. Doust, A 1% treadmill grade most accurately
reflects the energetic cost of outdoor running. Journal of Sports Sciences,
1996. 14(4): p. 321-327.
24. Moseley, L. and A.E. Jeukendrup, The reliability of cycling efficiency.
medicine and science in sports and exercise, 2001. 33(4): p. 621-627.
25. Weir, J.B., New methods for calculating metabolic rate with specia;
reference to protein metabolism. The journal of physiology, 1949. 109(1-9).
138
26. Borg, G.A.V., Perceived exertion: a note on 'history' and methods.
Medicine and Science in Sports and Exercise, 1973. 5(2): p. 90-93.
27. Singh, T. and M. Koh, Effects of backpack load position on
spatiotemporal parameters and trunk forward lean. Gait and Posture, 2009 a.
29(1): p. 49-53.
28. Goh, J.H., A. Thambyah, and K. Bose, Effects of varying backpack
loads on peak forces in the lumbosacral spine during walking. Clinical
Biomechanics, 1998. 13(SUPPL. 1): p. S26-S31.
29. Wallman, K., et al., Test-retest reliability of the aerobic power index
component of the tri-level fitness profile in a sedentary population. Journal of
Science and Medicine in Sport, 2003. 6(4): p. 443-454.
30. Erdfelder, E., F. Faul, and A. Buchner, GPOWER: A general power
analysis program. Behavior Research Methods, Instruments, and Computers,
1996. 28(1): p. 1-11.
31. Li, J.X., Y. Hong, and P.D. Robinson, The effect of load carriage on
movement kinematics and respiratory parameters in children during walking.
European Journal of Applied Physiology, 2003. 90(1-2): p. 35-43.
32. Yoshizaki, H., et al., Effect of posture change on control of ventilation.
Japanese Journal of Physiology, 1998. 48(4): p. 267-273.
33. Johnson, R.C., et al., The effect of load position on biomechanical and
physiological measures during a short duration march. 2000.
34. Simpson, K.M., B.J. Munro, and J.R. Steele, Does load position affect
gait and subjective responses of females during load carriage? Applied
Ergonomics, 2012. 43(3): p. 479-485.
35. Frank, E., J.M. Stevenson, and P. Stothart, The effect of load
placement on static posture and reaction forces in youth. medicine and
science in sports and exercise, 2003. 35(5): p. S21.
36. Ekblom, B. and A.N. Goldbarg, The influence of physical training and
other factors on the subjective rating of perceived exertion. Acta Physiologica
Scandinavica, 1971. 83(3): p. 399-406.
37. Bobet, J. and R.W. Norman, Effects of load placement on back muscle
activity in load carriage. European Journal of Applied Physiology and
Occupational Physiology, 1984. 53(1): p. 71-75.
38. Legg, S.J. and A. Mahanty, Comparison of five modes of carrying a
load close to the trunk. Ergonomics, 1985. 28(12): p. 1653-1660.
139
39. Kirk, J. and D.A. Schneider, Physiological and perceptual responses
to load-carrying in female subjects using internal and external frame
backpacks. Ergonomics, 1992. 35(4): p. 445-455.
140
CHAPTER EIGHT
THE EFFECT OF SHOULDER STRAP WIDTH AND LOAD
PLACEMENT ON SHOULDER–BACKPACK INTERFACE
PRESSURE
Submitted to
Applied Ergonomics
141
In chapter 6 and 7, we found out that load placement within a backpack did
not affect postural stability and energy cost of the body. In this chapter we
examined the effect of load placement on contact pressure under the
shoulder straps. In addition, we investigated if width of shoulder strap affects
contact pressure under the straps. Unlike the previous experiments that were
conducted on participants, this experiment was carried out on a manikin.
This chapter has resulted in the following submission for publication;
Golriz S, Jeffrey JJ, Foreman KB, Walker BF. The effect of shoulder strap
width and load placement on shoulder–backpack interface pressure.
Submitted to Applied Ergonomics
142
The effect of shoulder strap width and load placement on shoulder–
backpack interface pressure
Samira Golriz a, Jeffrey J. Hebert a, K. Bo Foreman b, Bruce F. Walker a
a, School of Chiropractic and Sports Science, Murdoch University, Perth,
Australia
b, Department of Physical Therapy, University of Utah, Salt Lake City, UT,
USA
Samira Golriz* * Corresponding author
Email: [email protected]
90 South Street, Murdoch 6150, WA, Australia
Tel: +61 8 93601450
Fax: +61 8 93601299
Jeffrey J Hebert
Email: [email protected]
K Bo Foreman
Email: [email protected]
Bruce F Walker
Email: [email protected]
143
Abstract
Introduction: Pressure on the shoulder area can be a major limiting factor
related to backpack carriage. The purpose of this study was to evaluate the
effect of the shoulder strap width and load placement in a backpack on the
shoulder–backpack and axillary area–backpack interface pressure.
Methods: A manikin was used in this study and was equipped with a
backpack loaded with a 20 kg mass. One backpack with four different width
straps (5-6-7-8 cm) was used for testing. The load was placed high or low in
the backpack. A pressure sensor was placed over both the shoulder and on
the chest wall aspect of the axilla to measure the shoulder–backpack and
axilla–backpack interface pressure. At these two locations, we recorded
average and peak interface pressures. Data were analysed using 4 x 2
repeated measures ANOVA.
Results: A significant interaction was observed between the shoulder strap
width and load placement. The positive effect of wide straps on shoulder
pressure is greater when high load placement is used and the benefit of wide
straps on axilla pressure is improved when low load placement is used.
Interface pressure increased significantly from the narrow straps to the wide
straps. We observed the lowest and highest peak and average interface
pressure under the 8 cm and 5 cm strap, respectively. High load placement
resulted in lower shoulder–backpack interface pressure and higher axilla–
backpack interface pressure when compared to low load placement. A large
difference was noted between the interface pressure on high and low load
placement with narrow straps; however, as the shoulder strap width
increased, the difference between the two load placements decreased.
Conclusion: The least amount of interface pressure was observed with the 8
cm shoulder strap; however, an 8 cm strap might not always be the optimal
width, especially for people with narrow shoulders. High load placement
reduced the amount of interface pressure; therefore, it needs to be taken into
account when packing backpacks.
Keywords: pressure, shoulder strap, load placement, backpack
144
Introduction
Backpack loading in excess of recommended limits can lead to pain and
injury (Guyer, 2001, Negrini et al., 1999, Knapik et al., 1997, Epstein et al.,
1988). The estimates of prevalence of shoulder pain related to load carriage
among backpack users ranged from 14.7% to 34.5% (Pascoe et al., 1997,
Chiang et al., 2006). In most situations, the shoulder girdle bears, and then
transfers, a substantial proportion of the backpack load to the body. Shoulder
pain (Iyer, 2001, Van Gent et al., 2003), brachial plexus injury (Makela et al.,
2006) and higher levels of activity of shoulder muscles (Piscione and Gamet,
2006) have been reported after using a backpack to carry loads for prolonged
periods of time. When heavy loads are supported on the shoulders, injury
can occur due to forces at the backpack–shoulder interface (Makela et al.,
2006).
Direct measurement of backpack–skin interface pressure is a method for
measuring external forces exerted on the body by the backpack. A
continuous pressure of 14 kPa has been recommended as the threshold to
prevent tissue damage and discomfort during prolonged load carriage (Doan
et al., 1998). However, the shoulder interface pressure reported in the
literature during the carrying of a backpack has been higher than the
recommended threshold (Holewijn, 1990, Macias et al., 2008, Harman et al.,
1999) and skin pressure under the shoulder straps was reported as the factor
that limited backpack carrying (Holewijn, 1990). Therefore, development of
backpack designs that help in distributing and alleviating pressure could be
critical in offsetting discomfort and pain (Jones and Hooper, 2005). Poor
design characteristics of backpacks, such as shoulder straps without
ergonomic consideration, have been proposed as a main contributing factor
for upper body injuries (Fergenbaum, 2007). Widening of shoulder straps has
been recommended to reduce interface pressures and increase comfort
(Harman et al., 1999).
The literature has shown that load placement within a backpack affects
physiological and biomechanical measurements (Stuempfle et al., 2004,
Brackley et al., 2009). High load placement resulted in lower oxygen
145
consumption (Stuempfle et al., 2004), while low load placement led to lower
changes in posture (Brackley et al., 2009). In addition to shoulder strap width,
load placement within a backpack might also influence interface pressure.
Nevertheless, a systematic literature review revealed that the effect of load
placement and widening of shoulder straps on interface pressure has not
been investigated (Golriz and Walker, 2012). Therefore, the aim of the
present study was to evaluate the effect of backpack shoulder strap width
and load placement within a backpack on the shoulder–backpack interface
pressure. We hypothesised that wider shoulder straps would distribute the
load over a greater area and thereby reduce interface pressures and load
placement would affect the amount of interface pressure.
Methods
We used a manikin model to maximise the internal validity of the results. The
manikin consisted of a firm shell that simulated a male torso; it was
supported by two steel rods attached at the thigh stumps, with the lower ends
fixed to a solid wooden base (see figure 1).
Figure 1. The manikin and position of the pressure sensors
We compensated for the shift in the centre of gravity (COG) resulting from an
external load (Goh et al., 1998), by adjusting the anterior lean of the manikin
to replicate a balanced position with a 20 kg load in a backpack. We
estimated the correct angle of anterior lean by placing the manikin (with the
20 kg loaded backpack) on an Accugait-AMTI force plate (Advanced
Mechanical Technology Inc., Watertown, USA), and wedges were placed
under the base of the manikin to lean the manikin forward. Movements of the
COG of the manikin were followed to the point where the COG was aligned
146
with the centre of pressure in the X and Y directions (i.e. the manikin was
“balanced”). At this point, a Dasco Pro angle meter (Dasco Pro, Inc.
Rockford, IL, USA) was placed on the base stand of the manikin and the
angle was recorded. We repeated this 5 times with consistent results of 10
degrees; therefore, the mannequin was fixed to a table with an anterior tilt of
10 degrees to replicate as normal a situation as possible and to resemble the
human body more closely. This degree of forward angle was constant across
all combinations. Our choice of 10 degrees for a 20 kg loaded backpack is
supported by a study that shows that 10 degrees of trunk forward flexion is
recommended as an angle to balance anterior and posterior moments for
backpack loads approximating 25 kg (Reid et al., 2004).
Backpack configuration
Four identical backpacks manufactured by Explore Planet Earth TM, SAS
harness model (New South Wales, Australia), with dimensions of 45 x 32 x
20 cm (H x W x D) and weight of 1.4 kg were used in this study (Figure 2).
The backpacks were identical except for their shoulder straps. Shoulder strap
widths were 5, 6, 7 and 8 cm in width.
Figure 2. The SAS harness backpack
We standardised shoulder strap length and backpack position to ensure
consistency across all comparisons. Loads were positioned into the
innermost compartment of the backpack to maintain the backpack COG as
close to the manikin as possible. We divided the internal space of the
backpack into high (high load placement) and low (low load placement)
147
space and loads were fixed with high–density foams inserts. For the high
load placement, loads were placed in the upper space and for the low load
placement the loads were placed in the lower space of the backpack. The
backpacks were loaded to 20 kg and the hip belt of the backpack was
unfastened.
Interface pressure measurement
We obtained contact pressure measures using Tactile sensor pads
manufactured by Pressure Profile Systems, Inc. (Los Angeles, CA,
USA).This system uses capacitive based pressure sensing technology. The
sensor pads were 50x50 mm2, constructed of a thin (1.1 mm), conformable
and flexible conductive cloth, with relief cuts to allow for additional
conformability around multicurved surfaces. The pads are designed to
accommodate moderate flexing without affecting sensor performance. The
sensors were rated at pressure range of zero to 138 kPa and calibrated for
this range by the manufacturer. The data were acquired and recorded using
Chameleon software (Los Angeles, CA, USA) at a sampling frequency of 30
Hz. Each measurement was recorded for a two-minute duration. The data
were recorded from the entire area that the sensors covered. When using
capacitance pressure sensors, two minutes of measurement is an accurate
reflection of a longer duration of 1 hour (Fergenbaum, 2007).
We standardised sensor placement across all measurements by outlining the
locations of the inner and outer boundaries of the shoulder straps on the
manikin. Two central and common areas between the outlines of the four
shoulder strap configurations were considered for sensor placement. The
sensors were placed on the superior area of the shoulders at the mid
trapezius line and the chest wall aspect of the axilla (see figure 1). The
sensors were secured in these locations to ensure consistent placement
across trials.
We measured interface pressure under the right shoulder strap. The inactive
sensor was placed on the left side to ensure backpack placement symmetry.
We undertook a preliminary assessment of test–retest reliability, which
yielded standard error of measurement values of 0.034 kPa for a mean of
148
five measurements. Therefore, the mean of five measurements resulted in
optimal level of reliability.
Statistical analysis
Statistical analyses were conducted using SPSS version 17.0 (Chicago, IL,
USA). We conducted a 4 x 2 repeated measures analysis of variance
(ANOVA) with least significant difference pair-wise post hoc comparisons.
The independent variables were shoulder strap width with four levels (5, 6, 7
and 8 cm) and load placement with two levels (high and low load placement).
The dependent variables were shoulder and axillary average and peak
pressures (kPa). Data were assessed for violations of sphericity with
Mauchly’s test. Alpha was set at 0.05 for all comparisons (see appendices of
chapter 8, page 177).
Results
The results of average and peak interface pressures in the high and low load
placements across different strap widths are presented in tables 1 and 2. As
average and peak pressure demonstrated similar trends, only the results of
the average pressure are described below and presented in figures 3 and 4.
Table 1. Mean (SD) of peak and average backpack – shoulder interface pressure under different width shoulder straps (kPa) and load placements
Peak-HLP Peak-LLP Average-HLP Average-LLP
5 cm 17.2 (0.1) 30.8 (0.1) 16.8 (0.1) 30.5 (0.1)
6 cm 14.8 (0.1) 27.7 (0.1) 14.4 (0.1) 27.4 (0.2)
7cm 13.1 (0.2) 15.1 (0.2) 12.7 (0.3) 14.7 (0.2)
8 cm 11.5 (0.2) 14.3 (0.1) 11.1 (0.2) 13.8 (0.2)
HLP, high load placement; LLP, low load placement
149
Table 2. Mean (SD) of peak and average backpack – axilla interface pressure under different width shoulder straps (kPa) and load placements
Peak-HLP Peak-LLP Average-HLP Average-LLP
5 cm 6.8 (0.1) 6.1 (0.1) 6.4 (0.1) 5.8 (0.1)
6 cm 5.1 (0.1) 4.6 (0.1) 4.8 (0.1) 4.3 (0.2)
7cm 4.6 (0.1) 4.5 (0.1) 4.3 (0.2) 4.0 (0.2)
8 cm 3.9 (0.1) 4.0 (0.1) 3.6 (0.1) 3.6 (0.1)
HLP, high load placement; LLP, low load placement
Average pressure over the shoulder
A significant interaction (p<0.001) was found between shoulder strap width
and load placement. Significant main effects were observed for shoulder
strap width (p<0.001) and load placement (p<0.001) on the average and
peak pressure with narrow shoulder straps and on low load placement
associated with higher interface pressure. The mean (SD) of the average
shoulder pressure across different shoulder strap widths was 12.5 (1.9) kPa
for 8 cm, 13.7 (1.4) kPa for 7 cm, 20.9 (9.1) kPa for 6 cm and 23.6 (9.6) kPa
for 5 cm. Mean (SD) of average shoulder pressure was 13.8 (2.4) kPa for
high load placement and 21.6 (8.6) kPa for low load placement.
Post hoc comparisons demonstrated differences between all of the
combinations of load placements and strap widths. The combination of wide
straps–high load placement resulted in lower pressure and the combination
of narrow straps–low load placement generated higher pressure (see figure 3
and table 1).
150
Figure 3: Average backpack-shoulder interface pressure under various shoulder strap widths in high and low load placement
Shoulder strap width (cm)
4 5 6 7 8 9
Ave
rag
e b
ackp
ack-s
ho
uld
er
inte
rfa
ce
pre
ssure
(K
pa
)
5
10
15
20
25
30
35
High load placement o low load placement
* Significantly different from other strap widths p<0.05 † Significantly different from high load placement p<0.05 Error bar shows 95% CI
Average pressure on the axillary area
A significant interaction (p<0.001) was noted between shoulder strap width
and load placement. Significant main effects were observed for shoulder
strap width (p<0.001) and load placement (p<0.001) on average and peak
axillary interface pressures with narrow straps and on high load placement
associated with higher interface pressure. The mean (SD) of the average
axillary pressure across different shoulder strap widths was 3.6 (0.03) kPa for
8 cm, 4.2 (0.2) kPa for 7 cm, 4.5 (0.4) kPa for 6 cm and 6.1 (0.4) kPa for 5
cm. The mean (SD) of the average axillary pressure was 4.8 (1.2) kPa for
high load placement and 4.4 (0.9) kPa for low load placement.
Post hoc comparison revealed no significant differences between the high
and low load placements of the 8 cm strap. Significant differences between
all of the other combinations of load placements and strap widths were
*†
*† *†
*†
151
observed. The combination of wide straps–low load placement resulted in
lower pressure and the combination of narrow straps–high load placement
generated higher pressure (see figure 4 and table 2).
Figure 4: Average backpack-axillary area interface pressure under various shoulder strap widths in high and low load placement
Shoulder strap width (cm)
4 5 6 7 8 9
Ave
rag
e b
ackp
ack-a
xilla
ry inte
rfa
ce
pre
ssure
(K
pa
)
3
4
5
6
7
8
9
10
High load placement o low load placement
* Significantly different from other strap widths p<0.05 † Significantly different from low load placement p<0.05 Error bar shows 95% CI
Discussion
The purpose of this study was to evaluate the effect of the shoulder strap
width and load placement in a backpack on shoulder and axilla interface
pressure. Our first hypothesis was that wider shoulder straps would distribute
the load over a greater area and thereby reduce interface pressures. Our
second hypothesis was that load placement would affect the interface
pressures.
Our results support our first hypothesis. The 8 cm and 5 cm shoulder straps
resulted in the lowest and highest amount of interface pressure, respectively.
Wide straps help distribute the load exerted on the shoulders, thereby
*†
*†
*†
*
152
preventing concentration of the load and occlusive pressure over a small
area. This increased pressure under narrow straps can lead to discomfort
and pain for users, as significant correlation has been reported between
increased contact pressure and pain (Macias et al., 2005). Our finding was
consistent with another study that showed a lower interface pressure for an 8
cm shoulder strap than for a 5 cm strap (Holewijn, 1990). In that study,
backpack–shoulder interface pressure was compared between two loaded
army backpacks and shoulder strap width, among other backpack
configurations, was one of the characteristics that affected interface
pressure.
Another advantage of wide straps is that doubling the amount of load did not
result in a significant increase in shoulder interface pressure when wide
straps were used. In contrast, it resulted in an increase of 36% when narrow
straps were used (Holewijn, 1990). Furthermore, the narrowness of shoulder
straps has been reported as a negative factor by participants since the
narrow shoulder straps cut into the front of the shoulders and make backpack
carrying uncomfortable (Legg et al., 2003).
Our second hypothesis was also supported by our findings. Load placement
influenced interface pressure, as high load placement generated lower
shoulder pressure and higher axilla pressure, and low load placement
resulted in higher shoulder pressure and lower axilla pressure. A potential
explanation for this involves the differing moment arms between the high and
low load placements to the backpacks axis of rotation. The thoracic kyphotic
curve results in a greater moment arm in the high loading position relative to
the low load placement. This increases the moment caused by the backpack
and will cause the backpack to attempt to rotate posteriorly, thereby
increasing the pressure on the axilla. Conversely, the reduction in moment
with the low load placement causes less pressure on the anterior part of the
shoulder and more over the shoulders.
A significant interaction was observed between the shoulder strap width and
load placement. The positive effect of wide straps on shoulder pressure is
greater when high load placement is used and the benefit of wide straps on
153
axilla pressure is improved when low load placement is used. The
combination of high load placement–8 cm strap generated the lowest
shoulder interface pressure. The 8 cm shoulder strap, regardless of load
placement, generated the lowest axilla interface pressure.
We observed higher pressure over the shoulders and lower pressure on the
axilla in all conditions, which agrees with other studies (Holewijn, 1990, Hadid
et al., 2012). Continuous pressure higher than the recommended threshold of
14 kPa has been suggested to result in tissue damage; therefore, 14 kPa is
considered to be the safe upper limit (Doan et al., 1998). In the current study,
axilla interface pressure was below this recommended threshold in all of the
load placement and strap width conditions. Shoulder interface pressure for
the high load placement–8 cm, low load placement–8 cm and high load
placement–7 cm conditions were also below this threshold and within the
safe limit, while the shoulder interface pressures for the 5 cm and 6 cm
straps in both load placement conditions and for the 7 cm–low load
placement condition exceeded this threshold. In all conditions, the axilla
interface pressure was within the safe limit; therefore, basing the
recommendations for the optimal strap width and load placement on the
findings of shoulder interface pressure would seem logical. Thus, the
combination of high load placement and 8 cm strap is the optimal
configuration with respect to interface pressure between shoulder straps and
the body.
Our findings favour the use of wider shoulder straps, but the stature and
physique of the wearer should also be taken into account when choosing the
optimal width of shoulder straps. The manikin used in this study represented
a male body and the 8 cm strap resulted in the lowest contact pressure.
However, the 8 cm strap might be too wide for smaller framed individuals.
The findings of this study are limited as they were generated using a manikin
and may not generalise to humans. We only measured pressure at two
locations and this might have limited our findings. Future research should
seek to replicate our findings in vivo while incorporating person–centred
information such as perceived pain, discomfort and exertion. In addition,
154
future research should undertake more comprehensive pressure
measurements to look at the effect of load placement and shoulder strap
width on interface pressure.
In conclusion, the current study has shown that shoulder strap width and load
placement within a backpack influenced shoulder interface pressure, such
that wide straps and high load placement generated lower pressure.
Therefore, these factors should be important considerations in backpack
design and when packing a backpack for use. We recommend wide shoulder
straps and high load placement, since these generated the lowest interface
pressure in our study.
Conflict of interest
The backpacks used in this study were donated by Explore Planet EarthTM,
NSW, Australia; however, the manufacturer played no part in the design,
conduct or reporting of the research.
References
Brackley, H. M., Stevenson, J. M. & Selinger, J. C. 2009. Effect of backpack
load placement on posture and spinal curvature in prepubescent children.
Work, 32, 351-360.
Chiang, H. Y., Jacobs, K. & Orsmond, G. 2006. Gender-age environmental
associates of middle school students' low back pain. Work, 26, 19-28.
Doan, J. B., Stevenson, J. M., Bryant, J. T., Pelot, R. P. & Reid, S. A. 1998.
Developing a performance scale for load carriage designs. Proceedings of
the 30th Annual Conference of the Human Factors Association of Canada.
Epstein, Y., Rosenblum, J., Burstein, R. & Sawka, M. N. 1988. External load
can alter the energy cost of prolonged exercise. Eur. J. Appl. Physiol., 57,
243-247.
Fergenbaum, M. A. 2007. Development of safety limits for load carriage in
adults. Ontario, Canada, Queen's university.
Goh, J. H., Thambyah, A. & Bose, K. 1998. Effects of varying backpack loads
on peak forces in the lumbosacral spine during walking. Clin. Biomech., 13,
S26-S31.
Golriz, S. & Walker, B. F. 2012. Backpacks. Several factors likely to influence
design and usage: A systematic literature review. Work, 42, 519-531.
155
Guyer, R. L. 2001. Backpack = back pain. Am. J. Public Health, 91, 16-19.
Hadid, A., Epstein, Y., Shabshin, N. & Gefen, A. 2012. Modeling mechanical
strains and stresses in soft tissues of the shoulder during load carriage based
on load-bearing open MRI. J. Appl. Physiol., 112, 597-606.
Harman, E., Frykman, P., Pandorf, C., Tharion, W., Mello, R. P., Obusek, J.,
et al. 1999. Physiological, biomechanical, and maximal performance
comparisons of female soldiers carrying loads using prototype U.S. marine
corps modular lightweight load-carrying equipment (MOLLE) with interceptor
body armor and U.S.army all-purpose lightweight individual carrying
equipment (ALICE) with PASGT body armor Natick, MA, U.S. army research
institute of environmental medicine.
Holewijn, M. 1990. Physiological strain due to load carrying. Eur. J. Appl.
Physiol., 61, 237-245.
Iyer, S. R. 2001. An Ergonomic Study of Chronic Musculoskeletal Pain in
Schoolchildren. Indian J. Pediatr., 68, 937-941.
Jones, G. R. & Hooper, R. H. 2005. The effect of single- or multiple-layered
garments on interface pressure measured at the backpack-shoulder
interface. Appl. Ergon., 36, 79-83.
Knapik, J. J., Ang, P., Meiselman, H., Johnson, W., Kirk, J., Bensel, C., et al.
1997. Soldier performance and strenuous road marching: Influence of load
mass and load distribution. Mil. Med., 162, 62-67.
Legg, S. J., Barr, A. & Hedderley, D. I. 2003. Subjective perceptual methods
for comparing backpacks in the field. Ergonomics, 46, 935-955.
Macias, B. R., Murthy, G., Chambers, H. & Hargens, A. R. 2005. High
contact pressure beneath backpack straps of children contributes to pain.
Arch. Pediatr. Adolesc. Med., 159, 1186-1187.
Macias, B. R., Murthy, G., Chambers, H. & Hargens, A. R. 2008. Asymmetric
loads and pain associated with backpack carrying by children. J. Pediatr.
Orthop., 28, 512-517.
Makela, J. P., Ramstad, R., Mattila, V. & Pihlajamaki, H. 2006. Brachial
plexus lesions after backpack carriage in young adults. Clin. Orthop., 205-
209.
Negrini, S., Carabalona, R. & Sibilla, P. 1999. Backpack as a daily load for
schoolchildren. Lancet, 354, 1974.
156
Pascoe, D. D., Pascoe, D. E., Wang, Y. T., Shim, D. M. & Kim, C. K. 1997.
Influence of carrying book bags on gait cycle and posture of youths.
Ergonomics, 40, 631-641.
Piscione, J. & Gamet, D. 2006. Effect of mechanical compression due to load
carrying on shoulder muscle fatigue during sustained isometric arm
abduction: An electromyographic study. Eur. J. Appl. Physiol., 97, 573-581.
Reid, S. A., Stevenson, J. M. & Whiteside, R. A. 2004. Biomechanical
assessment of lateral stiffness elements in the suspension system of a
backpack. Ergonomics, 47, 1272-1281.
Stuempfle, K. J., Drury, D. G. & Wilson, A. L. 2004. Effect of load position on
physiological and perceptual responses during load carriage with an internal
frame backpack. Ergonomics, 47, 784-789.
Van Gent, C., Dols, J. J. C. M., De Rover, C. M., Hira Sing, R. A. & De Vet,
H. C. W. 2003. The weight of schoolbags and the occurrence of neck,
shoulder, and back pain in young adolescents. Spine, 28, 916-921.
157
CHAPTER NINE
CONCLUSION
158
The broad purpose of this thesis was to examine the effect of backpack
loading configurations and design features on postural stability, energy cost,
perceived exertion and shoulder interface pressure, when compared to an
unloaded condition. This information was required for further exploration of
the effects of backpacks on the body and to provide recommendations
regarding appropriate backpack load placement, optimum usage of backpack
features (such as a hip belt) and the optimum dimensions of backpack
features (such as shoulder straps).
Our systematic reviews of the literature (chapter 2 and 3) allowed us to
conclude that, in addition to the backpack weight, other factors might also
contribute to the reported pain and perceived exertion of backpack carrying.
The literature also showed that, in addition to the general belief that it leads
to low back pain, carrying a backpack also triggers pain in other parts of the
body, especially the shoulder area. In addition, our systematic reviews
revealed that tightening vs. loosening of the shoulder straps and a single
shoulder strap vs. double shoulder straps have been studied and compared
against each other in the literature, while shoulder straps, their designs and
dimensions have not yet been investigated. Nevertheless, the reviews show
the necessity of shoulder strap design improvement in order to alleviate the
risk of shoulder pain. We also concluded that backpack designs that
distribute the load (i.e. backpacks with a hip belt or double packs) generate
less discomfort and pain, but that no consensus has been reached regarding
the best load placement in a backpack/best backpack placement on the
spine. The systematic reviews also revealed widespread deficiencies in the
methodology of the included studies, such as inadequate statistical power
and a lack of reliable, valid and calibrated instruments of measurement.
We intended to use a clinical force plate (Midot posture scale analyser),
which is commonly used by clinicians but had unknown psychometric
properties, to assess postural stability. Therefore, we investigated the test–
retest reliability and concurrent validity of this clinical force plate in chapters 4
and 5. The Midot posture scale analyser showed an acceptable level of
reliability, but the level of validity was not satisfactory. Consequently, we
chose to use a validated AMTI force plate.
159
Using reliable and valid instruments, we compared postural stability, oxygen
consumption and perceived exertion between loaded conditions and an
unloaded control condition. Carrying a backpack with a load equivalent to
20% of body weight, regardless of the type of load placement and hip belt
usage (with/without), resulted in a lower level of subjective and objective
measures of postural stability (chapter 6) and higher levels of oxygen
consumption, energy cost (chapter 7) and perceived exertion (chapter 6 and
7). These changes can be attributed directly to the extra load imposed on the
body.
In addition to the effect of carrying a backpack on postural stability and
perceived exertion, we investigated whether hip belt use improved postural
stability and perceived exertion (chapter 6). We observed that hip belt use did
not enhance objective measures of postural stability; however, participants
reported a perception of increased stability and decreased exertion.
Therefore, we recommend that individuals should use hip belts when wearing
backpacks.
In addition to hip belt use, we also investigated the effects of load placement
within a backpack. Load placement did not influence postural stability
(chapter 6), heart rate, oxygen uptake, energy cost (chapter 7) or overall
perceived exertion (chapter 6 and 7); however, participants reported a higher
level of local perceived exertion in the shoulders and the upper back during
high load placement. One experiment reported in chapter 6 was conducted in
an outdoor and non-controlled condition, where participants were free to walk
at a self selected pace and for a short time (4 minutes), while the other
experiment was carried out in a controlled laboratory environment, at a
forced pace and for a longer period of time (10 minutes). The only difference
observed between the load placements in either experiment was in the
perceived exertion in the shoulder and upper back. In our questionnaires, we
asked the participants to rate the level of exertion they perceived on different
body regions; however, we did not provide any explanation that would have
clarified the boundaries between the regions. Therefore, participants may
have considered their shoulders and upper back areas to be
interchangeable.
160
We proposed in chapters 6 and 7 that the higher level of perceived exertion
in the shoulders during the high load placement condition was likely due to
increased pressure on the shoulders under that condition. However, in
chapter 8, we determined that high load placement resulted in lower shoulder
interface pressure. This discrepancy might show that perceived exertion is
not indicative of interface pressure, and that the amount of exertion is related
to some other construct that we did not investigate.
In choosing between high and low load placement conditions, a
determination must be made about which factor is more important for the
user: interface pressure or exertion. The findings of chapters 6, 7 and 8
indicate that a low load placement might be preferable if the goal of the
backpack user is to minimize shoulder and upper back exertion, whereas
high load placement might be best if the backpack user prefers to reduce
shoulder–backpack interface pressure. Alternatively, since the difference
between the load placements was small when wide shoulder straps were
used, backpack users may choose the load placement that they feel is the
most comfortable as long as they wear backpacks with wide shoulder straps.
Wide shoulder straps result in decreased interface pressure without
adversely affecting perceived exertion.
Our findings favoured wide straps but did not determine “how wide is wide
enough?”. Further research needs to be conducted to answer this question.
However, in general, physique and anthropometric characteristics need to be
considered to determine the dimensions of a backpack and its features. We
may obtain the optimal results if a backpack and its features are tailored for a
given individual. For elite use (e.g., military, sport), this may be warranted;
however, for general use, individuals should choose a backpack that is
appropriate to their size and wear it properly to reduce the negative effects of
carrying loads.
161
APPENDICES
162
Appendices of chapter 4
163
Health Questionnaire
Please take your time answering this questionnaire. If you have a question
regarding any of the points below, please ask one of the investigators prior to
handing it in.
Gender: Male Female
Age: ______ years
Have you previously suffered from ankle, knee or hip arthritis, weakness,
strain or sprain?
Yes No
Have you ever had surgery on your back, neck or lower extremities?
Yes No
If yes, what was the surgery?
Do you have any history of treatment for spinal and lower extremity pain (legs
or feet) lasting for more than a week?
Yes No
Do you have any history of neurological problems like epilepsy?
Yes No
Do you have any history of joint problems like rheumatoid arthritis?
Yes No
Have you ever been involved in a serious road traffic accident where you
were injured
Yes No
Have you suffered from dizziness, nausea or headache as a consequence?
Yes No
164
Do you have any history of balance problems, dizziness or loss of balance?
Yes No
If yes, please elaborate:
Do you at the moment have cold, ear infection or fever?
Yes No
Do you take any medications for pain suppression?
Yes No
Are you pregnant?
Yes No
Dominant hand: Right Left
Dominant foot: Right Left
Office use only
Height (cms):
Weight (Kgs):
BMI:
165
Appendices of chapter 5
166
Health Questionnaire
Please take your time answering this questionnaire. If you have a question
regarding any of the points below, please ask one of the investigators prior to
handing it in.
Gender: Male Female
Age: ______ years
Have you previously suffered from ankle, knee or hip arthritis, weakness,
strain or sprain?
Yes No
Have you ever had surgery on your back, neck or lower extremities?
Yes No
If yes, what was the surgery?
Do you have any history of treatment for spinal and lower extremity pain (legs
or feet) lasting for more than a week?
Yes No
Do you have any history of neurological problems like epilepsy?
Yes No
Do you have any history of joint problems like rheumatoid arthritis?
Yes No
Have you ever been involved in a serious road traffic accident where you
were injured
Yes No
167
Have you suffered from dizziness, nausea or headache as a consequence?
Yes No
Do you have any history of balance problems, dizziness or loss of balance?
Yes No
If yes, please elaborate:
Do you at the moment have cold, ear infection or fever?
Yes No
Do you take any medications for pain suppression?
Yes No
Are you pregnant?
Yes No
Dominant hand: Right Left
Dominant foot: Right Left
Office use only
Height (cms):
Weight (Kgs):
BMI:
168
Appendices of chapter 6
169
Health Questionnaire
Please take your time answering this questionnaire. If you have a question
regarding any of the points below, please ask one of the investigators prior to
handing it in.
Gender: Male Female
Age: ______ years
Have you previously suffered from ankle, knee or hip arthritis, weakness,
strain or sprain?
Yes No
Have you ever had surgery on your back, neck or lower extremities?
Yes No
If yes, what was the surgery?
Do you have any recent history of treatment for spinal and lower extremity
pain (legs or feet) lasting for more than a week?
Yes No
Do you have any history of neurological problems like epilepsy?
Yes No
Do you have any recent history of joint problems like rheumatoid arthritis?
Yes No
Have you ever been involved in a serious road traffic accident where you
were injured
Yes No
Have you suffered from dizziness, nausea or headache as a consequence?
170
Yes No
Do you have any history of balance problems or loss of balance?
Yes No
If yes, please elaborate:
Do you feel dizzy or have vertigo at time of participation?
Yes No
Do you at the moment have cold, ear infection or fever?
Yes No
Do you take any medications for pain suppression?
Yes No
When was the last time that you consumed alcohol?
Are you pregnant?
Yes No
Dominant hand: Right Left
Dominant foot: Right Left
Office use only
Height (cms):
Weight (Kgs):
BMI:
171
Perceived Sense of Postural Sway and Instability Scale
1. How much did you feel yourself slip (i.e., loss of foot traction)?
A little Some A lot
0 0.5 1 1.5 2
Did you feel at any time that you would slip and fall?
A little Some A lot
0 0.5 1 1.5 2
3. Did you have any difficulty in maintaining balance while performing the task?
A little Some A lot
0 0.5 1 1.5 2
4. What would you say was the overall difficulty of this task?
A little (very
easy)
easy
Some
(moderate)
A lot
(somewhat
hard)
0 0.5 1 1.5 2
172
Borg Scale
Please score the rating of exertion that you perceived for the following zones
based on the below table:
Body zone Rating of perceived exertion
Overall
Neck
Shoulders
Upper back
Low back
Lower extremity
6 Very, very light How you feel when lying in bed or sitting in a chair relaxed
7
8 Very light Little or no effort
9
10
11 Fairly light Target range: how you should feel with exercise or activity
12
13 Somewhat hard
14 Hard
15
16 Very hard How you felt with the hardest work you have ever done
17
18 Very, very hard
19
20 Maximum effort Don’t work this hard!
173
Appendices of chapter 7
174
Health Questionnaire
Please take your time to answer this questionnaire honestly. If you have a
question regarding any of the points below, please ask one of the
investigators prior to handing it in.
Gender: Male Female
Age: ______ years Blood pressure: ______
Weight: ______ kg BMI: ______
Height: ______ cm
History (please tick)
You have had:
…….a heart attack
…….heart surgery
……coronary angioplasty (balloon artery expansion)
…..pacemaker
…..heart valve disease
…..heart failure
…..heart transplantation
…..congenital heart disease
Symptoms
…..You experience chest discomfort with exertion.
….You experience unreasonable breathlessness.
….You experience dizziness, fainting, or blackouts.
….You take heart medications.
Other health issues
175
….You have diabetes.
….You have asthma or other lung disease.
….You have burning or cramping sensation in your lower legs when walking
short distances.
….You have musculoskeletal problems that limit your physical activity.
….You take prescription medication(s). Please list here: ………………
….You are pregnant.
….You are claustrophobic
Section 2
Cardiovascular risk factors
….You smoke, or quit smoking within the previous 6 months.
….Your blood pressure is >140/90 mm Hg*. (We will check this before
starting)
….You take blood pressure medication.
….Your blood cholesterol level is >200 mg/dl.
….You do not know your cholesterol level.
….You have a close blood relative who had a heart attack or heart surgery
before age 55 (father or brother) or age 65 (mother or sister)
….You are physically inactive (i.e.; you get <30 minutes of Moderate-intensity
physical activity on at least 3 days per week).
….You are>10 kgs overweight (BMI ≥ 30).
Section 3
…. None of the above
176
Borg Scale
Please score the rating of exertion that you perceived for the following zones
based on the below table:
Body zone Rating of perceived exertion
Overall
Neck
Shoulders
Upper back
Low back
Lower extremity
6 Very, very light How you feel when lying in bed or sitting in a chair relaxed
7
8 Very light Little or no effort
9
10
11 Fairly light Target range: how you should feel with exercise or activity
12
13 Somewhat hard
14 Hard
15
16 Very hard How you felt with the hardest work you have ever done
17
18 Very, very hard
19
20 Maximum effort
Don’t work this hard!
177
Appendices of chapter 8
Results of a 4 x 2 repeated measures ANOVA with least significant difference
pair wise post hoc comparisons. The independent variables were shoulder
strap width with 4 variations (5, 6, 7 and 8 cm) and load placement with two
levels of high (H) and low (L) load placement.
Average shoulder interface pressure
Descriptive Statistics
Mean Std. Deviation N
aveshoulder5H 16.8291 .12333 5
aveshoulder5L 30.4748 .11321 5
aveshoulder6H 14.4449 .07857 5
aveshoulder6L 27.3876 .15095 5
aveshoulder7H 12.6806 .30805 5
aveshoulder7L 13.7971 .21428 5
aveshoulder8H 11.1231 .16580 5
aveshoulder8L 14.7203 .22138 5
178
Tests of Within-Subjects Effects
Source Type III
Sum of
Squares df
Mean
Square F Sig.
Partial Eta
Squared
strap Sphericity
Assumed
885.029 3 295.010 10998.239 .000 1.000
Greenhouse-
Geisser
885.029 1.694 522.389 10998.239 .000 1.000
Huynh-Feldt 885.029 2.806 315.363 10998.239 .000 1.000
Lower-bound 885.029 1.000 885.029 10998.239 .000 1.000
Error(strap) Sphericity
Assumed
.322 12 .027
Greenhouse-
Geisser
.322 6.777 .047
Huynh-Feldt .322 11.226 .029
Lower-bound .322 4.000 .080
placement Sphericity
Assumed
612.386 1 612.386 61729.738 .000 1.000
Greenhouse-
Geisser
612.386 1.000 612.386 61729.738 .000 1.000
Huynh-Feldt 612.386 1.000 612.386 61729.738 .000 1.000
Lower-bound 612.386 1.000 612.386 61729.738 .000 1.000
Error(placement) Sphericity
Assumed
.040 4 .010
Greenhouse-
Geisser
.040 4.000 .010
Huynh-Feldt .040 4.000 .010
Lower-bound .040 4.000 .010
179
strap * placement Sphericity
Assumed
307.370 3 102.457 7222.170 .000 .999
Greenhouse-
Geisser
307.370 1.267 242.608 7222.170 .000 .999
Huynh-Feldt 307.370 1.586 193.799 7222.170 .000 .999
Lower-bound 307.370 1.000 307.370 7222.170 .000 .999
Error(strap*placement) Sphericity
Assumed
.170 12 .014
Greenhouse-
Geisser
.170 5.068 .034
Huynh-Feldt .170 6.344 .027
Lower-bound .170 4.000 .043
180
Strap
Estimates
Measure:MEASURE_1
Strap
(cm)
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
5 23.652 .052 23.508 23.796
6 20.916 .045 20.792 21.041
7 13.239 .116 12.916 13.562
8 12.922 .063 12.748 13.096
Pairwise Comparisons
Measure:MEASURE_1
(I) strap (J) strap
(cm) Mean Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
5
dimension2
6 2.736* .039 .000 2.629 2.843
7 10.413* .070 .000 10.219 10.607
8 10.730* .047 .000 10.599 10.861
6
dimension2
5 -2.736* .039 .000 -2.843 -2.629
7 7.677* .080 .000 7.457 7.898
8 7.995* .077 .000 7.781 8.208
7
dimension2
5 -10.413* .070 .000 -10.607 -10.219
6 -7.677* .080 .000 -7.898 -7.457
8 .317* .106 .041 .022 .613
8
dimension2
5 -10.730* .047 .000 -10.861 -10.599
6 -7.995* .077 .000 -8.208 -7.781
7 -.317* .106 .041 -.613 -.022
181
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
182
Placement
Estimates
Measure:MEASURE_1
placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 13.769 .074 13.563 13.976
L 21.595 .045 21.470 21.720
Pairwise Comparisons
Measure:MEASURE_1
(I) placement (J) placement
Mean
Difference (I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
H dimension2 L -7.826* .031 .000 -7.913 -7.738
L dimension2 H 7.826* .031 .000 7.738 7.913
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
183
Strap * placement
3. strap * placement
Measure:MEASURE_1
strap placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
5 dimension2
H 16.829 .055 16.676 16.982
L 30.475 .051 30.334 30.615
6 dimension2
H 14.445 .035 14.347 14.542
L 27.388 .068 27.200 27.575
7 dimension2
H 12.681 .138 12.298 13.063
L 13.797 .096 13.531 14.063
8 dimension2
H 11.123 .074 10.917 11.329
L 14.720 .099 14.445 14.995
Estimates
placement Strap
(cm)
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 5 16.829 .055 16.676 16.982
6 14.445 .035 14.347 14.542
7 12.681 .138 12.298 13.063
8 11.123 .074 10.917 11.329
L 5 30.475 .051 30.334 30.615
6 27.388 .068 27.200 27.575
7 13.797 .096 13.531 14.063
8 14.720 .099 14.445 14.995
184
Pairwise Comparisons
Measure:MEASURE_1
Strap
(cm)
(I) placement (J) placement
Mean
Difference (I-
J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
5 dimension2
H dimension3 L -13.646* .022 .000 -13.706 -13.585
L dimension3 H 13.646* .022 .000 13.585 13.706
6 dimension2
H dimension3 L -12.943* .059 .000 -13.107 -12.778
L dimension3 H 12.943* .059 .000 12.778 13.107
7 dimension2
H dimension3 L -1.117* .046 .000 -1.244 -.989
L dimension3 H 1.117* .046 .000 .989 1.244
8 dimension2
H dimension3 L -3.597* .122 .000 -3.936 -3.258
L dimension3 H 3.597* .122 .000 3.258 3.936
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
185
Pairwise Comparisons
placement (I) strap (J)
strap
Mean Difference
(I-J)
Std.
Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
H
dimension2
5
dimension3
6 2.384* .028 .000 2.308 2.461
7 4.149* .091 .000 3.896 4.401
8 5.706* .026 .000 5.635 5.777
6
dimension3
5 -2.384* .028 .000 -2.461 -2.308
7 1.764* .104 .000 1.477 2.052
8 3.322* .040 .000 3.210 3.433
7
dimension3
5 -4.149* .091 .000 -4.401 -3.896
6 -1.764* .104 .000 -2.052 -1.477
8 1.557* .068 .000 1.370 1.745
8
dimension3
5 -5.706* .026 .000 -5.777 -5.635
6 -3.322* .040 .000 -3.433 -3.210
7 -1.557* .068 .000 -1.745 -1.370
L
dimension2
5
dimension3
6 3.087* .067 .000 2.902 3.272
7 16.678* .053 .000 16.530 16.825
8 15.754* .108 .000 15.454 16.055
6
dimension3
5 -3.087* .067 .000 -3.272 -2.902
7 13.590* .067 .000 13.405 13.776
8 12.667* .156 .000 12.233 13.101
7
dimension3
5 -16.678* .053 .000 -16.825 -16.530
6 -13.590* .067 .000 -13.776 -13.405
8 -.923* .155 .004 -1.354 -.493
186
8
dimension3
5 -15.754* .108 .000 -16.055 -15.454
6 -12.667* .156 .000 -13.101 -12.233
7 .923* .155 .004 .493 1.354
Based on estimated marginal means Adjustment for multiple comparisons: Least Significant Difference
(equivalent to no adjustments).
187
Peak shoulder interface pressure
Descriptive Statistics
Mean Std. Deviation N
maxarmpit5H 6.7538 .06865 5
maxarmpit5L 6.1059 .11535 5
maxarmpit6H 5.1278 .06200 5
maxarmpit6L 4.6178 .09756 5
maxarmpit7H 4.6450 .12489 5
maxarmpit7L 4.4520 .15068 5
maxarmpit8H 3.9284 .04901 5
maxarmpit8L 3.9971 .08435 5
188
Tests of Within-Subjects Effects (Measure:MEASURE_1)
Source Type III
Sum of
Squares df
Mean
Square F Sig.
Partial Eta
Squared
strap Sphericity
Assumed
33.318 3 11.106 2066.914 .000 .998
Greenhouse-
Geisser
33.318 1.162 28.668 2066.914 .000 .998
Huynh-Feldt 33.318 1.343 24.810 2066.914 .000 .998
Lower-bound 33.318 1.000 33.318 2066.914 .000 .998
Error(strap) Sphericity
Assumed
.064 12 .005
Greenhouse-
Geisser
.064 4.649 .014
Huynh-Feldt .064 5.372 .012
Lower-bound .064 4.000 .016
placement Sphericity
Assumed
1.027 1 1.027 969.826 .000 .996
Greenhouse-
Geisser
1.027 1.000 1.027 969.826 .000 .996
Huynh-Feldt 1.027 1.000 1.027 969.826 .000 .996
Lower-bound 1.027 1.000 1.027 969.826 .000 .996
Error(placement) Sphericity
Assumed
.004 4 .001
Greenhouse-
Geisser
.004 4.000 .001
Huynh-Feldt .004 4.000 .001
Lower-bound .004 4.000 .001
189
strap * placement Sphericity
Assumed
.777 3 .259 97.353 .000 .961
Greenhouse-
Geisser
.777 1.148 .677 97.353 .000 .961
Huynh-Feldt .777 1.312 .592 97.353 .000 .961
Lower-bound .777 1.000 .777 97.353 .001 .96
Error(strap*placement) Sphericity
Assumed
.032 12 .003
Greenhouse-
Geisser
.032 4.594 .007
Huynh-Feldt .032 5.249 .006
Lower-bound .032 4.000 .008
190
Strap
Estimates
Measure:MEASURE_1
Strap
(cm) Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
5 6.430 .040 6.318 6.541
6 4.873 .036 4.774 4.972
7 4.548 .061 4.380 4.717
8 3.963 .022 3.902 4.023
Pairwise Comparisons
Measure:MEASURE_1
(I) strap (J) strap
(cm) Mean Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
5
dimension2
6 1.557* .013 .000 1.522 1.593
7 1.881* .023 .000 1.817 1.946
8 2.467* .034 .000 2.374 2.561
6
dimension2
5 -1.557* .013 .000 -1.593 -1.522
7 .324* .026 .000 .253 .395
8 .910* .032 .000 .822 .999
7
dimension2
5 -1.881* .023 .000 -1.946 -1.817
6 -.324* .026 .000 -.395 -.253
8 .586* .054 .000 .435 .736
8
dimension2
5 -2.467* .034 .000 -2.561 -2.374
6 -.910* .032 .000 -.999 -.822
7 -.586* .054 .000 -.736 -.435
191
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Placement
Estimates
Measure:MEASURE_1
placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 5.114 .032 5.024 5.204
L 4.793 .041 4.679 4.908
Pairwise Comparisons
Measure:MEASURE_1
(I) placement (J) placement
Mean
Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
H dimension2 L .321* .010 .000 .292 .349
L dimension2 H -.321* .010 .000 -.349 -.292
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Strap*placement
3. strap * placement
Measure:MEASURE_1
Strap
(cm)
placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
192
5 dimension2
H 6.754 .031 6.669 6.839
L 6.106 .052 5.963 6.249
6 dimension2
H 5.128 .028 5.051 5.205
L 4.618 .044 4.497 4.739
7 dimension2
H 4.645 .056 4.490 4.800
L 4.452 .067 4.265 4.639
8 dimension2
H 3.928 .022 3.867 3.989
L 3.997 .038 3.892 4.102
Estimates
placement Strap
(cm) Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 5 6.754 .031 6.669 6.839
6 5.128 .028 5.051 5.205
7 4.645 .056 4.490 4.800
8 3.928 .022 3.867 3.989
L 5 6.106 .052 5.963 6.249
6 4.618 .044 4.497 4.739
7 4.452 .067 4.265 4.639
8 3.997 .038 3.892 4.102
193
Pairwise Comparisons
Measure:MEASURE_1
Strap
(cm)
(I) placement (J) placement
Mean
Difference (I-
J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
5
dimension2
H dimension3 L .648* .028 .000 .571 .725
L dimension3 H -.648* .028 .000 -.725 -.571
6
dimension2
H dimension3 L .510* .017 .000 .463 .557
L dimension3 H -.510* .017 .000 -.557 -.463
7
dimension2
H dimension3 L .193* .026 .002 .121 .265
L dimension3 H -.193* .026 .002 -.265 -.121
8
dimension2
H dimension3 L -.069 .044 .190 -.190 .052
L dimension3 H .069 .044 .190 -.052 .190
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
194
Pairwise Comparisons
Measure:MEASURE_1
placement (I) strap
(cm)
(J) strap
(cm) Mean
Difference (I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
H
dimension2
5
dimension3
6 2.412* .031 .000 2.327 2.497
7 4.094* .047 .000 3.964 4.223
8 5.706* .026 .000 5.635 5.778
6
dimension3
5 -2.412* .031 .000 -2.497 -2.327
7 1.682* .071 .000 1.485 1.878
8 3.294* .040 .000 3.183 3.406
7
dimension3
5 -4.094* .047 .000 -4.223 -3.964
6 -1.682* .071 .000 -1.878 -1.485
8 1.613* .035 .000 1.515 1.710
8
dimension3
5 -5.706* .026 .000 -5.778 -5.635
6 -3.294* .040 .000 -3.406 -3.183
7 -1.613* .035 .000 -1.710 -1.515
L
dimension2
5
dimension3
6 3.143* .035 .000 3.046 3.241
7 15.699* .128 .000 15.344 16.054
8 16.471* .072 .000 16.271 16.672
6
dimension3
5 -3.143* .035 .000 -3.241 -3.046
7 12.556* .114 .000 12.240 12.872
8 13.328* .056 .000 13.173 13.483
195
7
dimension3
5 -15.699* .128 .000 -16.054 -15.344
6 -12.556* .114 .000 -12.872 -12.240
8 .772* .067 .000 .587 .957
8
dimension3
5 -16.471* .072 .000 -16.672 -16.271
6 -13.328* .056 .000 -13.483 -13.173
7 -.772* .067 .000 -.957 -.587
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Average axillary area interface pressure
Descriptive Statistics
Mean Std. Deviation N
avearmpit5H 6.4091 .06932 5
avearmpit5L 5.7490 .16378 5
avearmpit6H 4.8380 .05799 5
avearmpit6L 4.2217 .15946 5
avearmpit7H 4.3280 .17209 5
avearmpit7L 4.0066 .25180 5
avearmpit8H 3.6111 .06148 5
avearmpit8L 3.6497 .11404 5
196
Tests of Within-Subjects Effects (Measure:MEASURE_1)
Source Type III
Sum of
Squares df
Mean
Square F Sig.
Partial Eta
Squared
strap Sphericity
Assumed
33.198 3 11.066 513.458 .000 .992
Greenhouse-
Geisser
33.198 1.797 18.475 513.458 .000 .992
Huynh-Feldt 33.198 3.000 11.066 513.458 .000 .992
Lower-bound 33.198 1.000 33.198 513.458 .000 .992
Error(strap) Sphericity
Assumed
.259 12 .022
Greenhouse-
Geisser
.259 7.188 .036
Huynh-Feldt .259 12.000 .022
Lower-bound .259 4.000 .065
placement Sphericity
Assumed
1.519 1 1.519 571.042 .000 .993
Greenhouse-
Geisser
1.519 1.000 1.519 571.042 .000 .993
Huynh-Feldt 1.519 1.000 1.519 571.042 .000 .993
Lower-bound 1.519 1.000 1.519 571.042 .000 .993
Error(placement) Sphericity
Assumed
.011 4 .003
Greenhouse-
Geisser
.011 4.000 .003
197
Huynh-Feldt .011 4.000 .003
Lower-bound .011 4.000 .003
strap * placement Sphericity
Assumed
.781 3 .260 19.518 .000 .830
Greenhouse-
Geisser
.781 2.268 .344 19.518 .000 .830
Huynh-Feldt .781 3.000 .260 19.518 .000 .830
Lower-bound .781 1.000 .781 19.518 .012 .830
Error(strap*placement) Sphericity
Assumed
.160 12 .013
Greenhouse-
Geisser
.160 9.074 .018
Huynh-Feldt .160 12.000 .013
Lower-bound .160 4.000 .040
198
Strap
Estimates
Measure:MEASURE_1
Strap
(cm) Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
5 6.079 .051 5.936 6.222
6 4.530 .045 4.406 4.654
7 4.167 .087 3.925 4.410
8 3.630 .022 3.570 3.691
Pairwise Comparisons
Measure:MEASURE_1
(I) strap
(cm)
(J) strap
(cm) Mean Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
5
dimension2
6 1.549* .041 .000 1.436 1.663
7 1.912* .067 .000 1.726 2.097
8 2.449* .052 .000 2.304 2.593
6
dimension2
5 -1.549* .041 .000 -1.663 -1.436
7 .363* .067 .006 .175 .550
8 .899* .058 .000 .739 1.060
7
dimension2
5 -1.912* .067 .000 -2.097 -1.726
6 -.363* .067 .006 -.550 -.175
8 .537* .096 .005 .272 .802
8
dimension2
5 -2.449* .052 .000 -2.593 -2.304
6 -.899* .058 .000 -1.060 -.739
199
7 -.537* .096 .005 -.802 -.272
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Placement
Estimates
Measure:MEASURE_1
placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 4.797 .039 4.688 4.906
L 4.407 .042 4.291 4.522
Pairwise Comparisons
Measure:MEASURE_1
(I) placement (J) placement
Mean Difference
(I-J) Std. Error Sig.a
95% Confidence Interval
for Differencea
Lower Bound
Upper
Bound
200
dimension1
H dimension2 L .390* .016 .000 .345 .435
L dimension2 H -.390* .016 .000 -.435 -.345
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Strap * placement
3. strap * placement
Measure:MEASURE_1
strap placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
5 dimension2
H 6.409 .031 6.323 6.495
L 4.273 .062 4.101 4.444
6 dimension2
H 4.838 .026 4.766 4.910
L 5.789 .049 5.653 5.925
7 dimension2
H 4.328 .077 4.114 4.542
L 4.094 .083 3.863 4.324
8 dimension2
H 3.611 .027 3.535 3.687
L 3.694 .035 3.596 3.792
Estimates
Measure:MEASURE_1
placement strap
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 5 6.409 .031 6.323 6.495
6 4.838 .026 4.766 4.910
7 4.328 .077 4.114 4.542
8 3.611 .027 3.535 3.687
201
L 5 5.749 .073 5.546 5.952
6 4.222 .071 4.024 4.420
7 4.007 .113 3.694 4.319
8 3.650 .051 3.508 3.791
Pairwise Comparisons
Measure:MEASURE_1
strap (I) placement (J) placement
Mean
Difference
(I-J) Std. Error Sig.a
95% Confidence Interval
for Differencea
Lower
Bound Upper Bound
5 dimension2
H dimension3 L .660* .045 .000 .534 .786
L dimension3 H -.660* .045 .000 -.786 -.534
6 dimension2
H dimension3 L .616* .059 .000 .451 .781
L dimension3 H -.616* .059 .000 -.781 -.451
7 dimension2
H dimension3 L .321* .081 .017 .095 .548
L dimension3 H -.321* .081 .017 -.548 -.095
8 dimension2
H dimension3 L -.039 .070 .609 -.232 .155
L dimension3 H .039 .070 .609 -.155 .232
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
202
Pairwise Comparisons
Measure:MEASURE_1
placement (I)
strap
(cm)
(J)
strap
(cm)
Mean Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
H
dimension2
5
dimension3
6 1.571* .014 .000 1.533 1.610
7 2.081* .051 .000 1.940 2.222
8 2.798* .017 .000 2.751 2.845
6
dimension3
5 -1.571* .014 .000 -1.610 -1.533
7 .510* .052 .001 .367 .653
8 1.227* .014 .000 1.189 1.265
7
dimension3
5 -2.081* .051 .000 -2.222 -1.940
6 -.510* .052 .001 -.653 -.367
8 .717* .052 .000 .573 .861
8
dimension3
5 -2.798* .017 .000 -2.845 -2.751
6 -1.227* .014 .000 -1.265 -1.189
7 -.717* .052 .000 -.861 -.573
L
dimension2
5
dimension3
6 1.527* .071 .000 1.331 1.724
7 1.742* .108 .000 1.443 2.042
8 2.099* .098 .000 1.826 2.373
6
dimension3
5 -1.527* .071 .000 -1.724 -1.331
7 .215 .116 .138 -.108 .538
8 .572* .118 .008 .244 .900
7 dimension3 5 -1.742* .108 .000 -2.042 -1.443
203
6 -.215 .116 .138 -.538 .108
8 .357 .146 .071 -.049 .763
8
n3
5 -2.099* .098 .000 -2.373 -1.826
6 -.572* .118 .008 -.900 -.244
7 -.357 .146 .071 -.763 .049
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Peak axillary area interface pressure
Descriptive Statistics
Mean Std. Deviation N
maxarmpit5H 6.7538 .06865 5
maxarmpit5L 6.1059 .11535 5
maxarmpit6H 5.1278 .06200 5
maxarmpit6L 4.6178 .09756 5
maxarmpit7H 4.6450 .12489 5
maxarmpit7L 4.4520 .15068 5
maxarmpit8H 3.9284 .04901 5
maxarmpit8L 3.9971 .08435 5
204
Tests of Within-Subjects Effects
Measure:MEASURE_1
Source Type III
Sum of
Squares df
Mean
Square F Sig.
Partial Eta
Squared
strap Sphericity
Assumed
33.318 3 11.106 2066.914 .000 .998
Greenhouse-
Geisser
33.318 1.162 28.668 2066.914 .000 .998
Huynh-Feldt 33.318 1.343 24.810 2066.914 .000 .998
Lower-bound 33.318 1.000 33.318 2066.914 .000 .998
Error(strap) Sphericity
Assumed
.064 12 .005
Greenhouse-
Geisser
.064 4.649 .014
Huynh-Feldt .064 5.372 .012
Lower-bound .064 4.000 .016
placement Sphericity
Assumed
1.027 1 1.027 969.826 .000 .996
Greenhouse-
Geisser
1.027 1.000 1.027 969.826 .000 .996
Huynh-Feldt 1.027 1.000 1.027 969.826 .000 .996
Lower-bound 1.027 1.000 1.027 969.826 .000 .996
Error(placement) Sphericity
Assumed
.004 4 .001
Greenhouse-
Geisser
.004 4.000 .001
Huynh-Feldt .004 4.000 .001
Lower-bound .004 4.000 .001
strap * placement Sphericity
Assumed
.777 3 .259 97.353 .000 .961
Greenhouse-
Geisser
.777 1.148 .677 97.353 .000 .961
Huynh-Feldt .777 1.312 .592 97.353 .000 .961
205
Lower-bound .777 1.000 .777 97.353 .001 .961
Error(strap*placement) Sphericity
Assumed
.032 12 .003
Greenhouse-
Geisser
.032 4.594 .007
Huynh-Feldt .032 5.249 .006
Lower-bound .032 4.000 .008
206
Strap
Estimates
Measure:MEASURE_1
Strap
(cm) Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
5 6.430 .040 6.318 6.541
6 4.873 .036 4.774 4.972
7 4.548 .061 4.380 4.717
8 3.963 .022 3.902 4.023
Pairwise Comparisons
Measure:MEASURE_1
(I) strap
(cm)
(J) strap
(cm) Mean Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
5
dimension2
6 1.557* .013 .000 1.522 1.593
7 1.881* .023 .000 1.817 1.946
8 2.467* .034 .000 2.374 2.561
6
dimension2
5 -1.557* .013 .000 -1.593 -1.522
7 .324* .026 .000 .253 .395
8 .910* .032 .000 .822 .999
7
dimension2
5 -1.881* .023 .000 -1.946 -1.817
6 -.324* .026 .000 -.395 -.253
8 .586* .054 .000 .435 .736
8
dimension2
5 -2.467* .034 .000 -2.561 -2.374
6 -.910* .032 .000 -.999 -.822
7 -.586* .054 .000 -.736 -.435
207
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Placement
Estimates
Measure:MEASURE_1
placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 5.114 .032 5.024 5.204
L 4.793 .041 4.679 4.908
Pairwise Comparisons
Measure:MEASURE_1
(I) placement (J)
placement
Mean Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
H dimension2 L .321* .010 .000 .292 .349
L dimension2 H -.321* .010 .000 -.349 -.292
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
208
Strap* placement
3. strap * placement
Measure:MEASURE_1
Strap
(cm)
placement
Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
5 dimension2
H 6.754 .031 6.669 6.839
L 6.106 .052 5.963 6.249
6 dimension2
H 5.128 .028 5.051 5.205
L 4.618 .044 4.497 4.739
7 dimension2
H 4.645 .056 4.490 4.800
L 4.452 .067 4.265 4.639
8 dimension2
H 3.928 .022 3.867 3.989
L 3.997 .038 3.892 4.102
Pairwise Comparisons
Measure:MEASURE_1
strap (I)
placeme
nt
(J) placement
Mean
Differenc
e (I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
5
dimension2
H dimension3 L .648* .028 .000 .571 .725
L dimension3 H -.648* .028 .000 -.725 -.571
6
dimension2
H dimension3 L .510* .017 .000 .463 .557
L dimension3 H -.510* .017 .000 -.557 -.463
7
dimension2
H dimension3 L .193* .026 .002 .121 .265
L dimension3 H -.193* .026 .002 -.265 -.121
209
8
dimension2
H dimension3 L -.069 .044 .190 -.190 .052
L dimension3 H .069 .044 .190 -.052 .190
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).
Estimates
Measure:MEASURE_1
placement Strap
(cm) Mean Std. Error
95% Confidence Interval
Lower Bound Upper Bound
dimension1
H 5 6.754 .031 6.669 6.839
6 5.128 .028 5.051 5.205
7 4.645 .056 4.490 4.800
8 3.928 .022 3.867 3.989
L 5 6.106 .052 5.963 6.249
6 4.618 .044 4.497 4.739
7 4.452 .067 4.265 4.639
8 3.997 .038 3.892 4.102
210
Pairwise Comparisons
Measure:MEASURE_1
placement (I) strap
(cm)
(J) strap
(cm)
Mean
Difference
(I-J) Std. Error Sig.a
95% Confidence Interval for
Differencea
Lower Bound Upper Bound
dimension1
H
dimension2
5
dimension3
6 1.626* .017 .000 1.579 1.673
7 2.109* .027 .000 2.032 2.185
8 2.825* .022 .000 2.765 2.886
6
dimension3
5 -1.626* .017 .000 -1.673 -1.579
7 .483* .031 .000 .398 .568
8 1.199* .017 .000 1.152 1.247
7
dimension3
5 -2.109* .027 .000 -2.185 -2.032
6 -.483* .031 .000 -.568 -.398
8 .717* .041 .000 .602 .831
8
dimension3
5 -2.825* .022 .000 -2.886 -2.765
6 -1.199* .017 .000 -1.247 -1.152
7 -.717* .041 .000 -.831 -.602
L
dimension2
5
dimension3
6 1.488* .017 .000 1.441 1.536
7 1.654* .022 .000 1.594 1.714
8 2.109* .064 .000 1.931 2.286
6
dimension3
5 -1.488* .017 .000 -1.536 -1.441
7 .166* .028 .004 .089 .242
8 .621* .053 .000 .473 .768
7
dimension3
5 -1.654* .022 .000 -1.714 -1.594
6 -.166* .028 .004 -.242 -.089
211
8 .455* .080 .005 .232 .677
8
dimension3
5 -2.109* .064 .000 -2.286 -1.931
6 -.621* .053 .000 -.768 -.473
7 -.455* .080 .005 -.677 -.232
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons: Least Significant Difference (equivalent to no adjustments).