The Effect of Backpack Carriage on the Biomechanics of Walking: A Systematic Review and Preliminary Meta-Analysis

Abstract

The aim of this systematic review was to evaluate the impact of bilaterally symmetrical backpack systems borne on the posterior trunk on walking biomechanics, as backpacks represent the most prevalent method of load carriage in the military and civilian population. A search of electronic databases was performed for studies that only investigated posteriorly-borne backpack carriage during level-grade walking (treadmill and over ground). Methodology of studies was assessed, and both meta-analysis and qualitative synthesis were completed. Fifty-four studies were included in this review. In summary, the available literature showed that backpack carriage in walking was associated with an increased trunk flexion angle, increased hip and ankle range of motion, increased vertical and horizontal ground reaction force, increased cadence, and reduced stride length. Several variations in backpack carriage protocols could explain between-study variations in results, including: walking speed, backpack carriage skill level, the use of a hip belt, and posterior displacement of the load away from the trunk. The findings of this systematic review will inform backpack carriage practices in the area of injury risk assessment and physical performance enhancement.

Full text

614
REVIEW AND META-ANALYSIS
Journal of Applied Biomechanics, 2016, 32, 614 -629
© 2016 Human Kinetics, Inc.
Liew, Morris, and Netto are with the School of Physiotherapy and Exercise
Sciences, Curtin University, Perth, WA, Australia. Address author corre-
spondence to Bernard Liew at b.liew@postgrad.curtin.edu.au.
http://dx.doi.org/10.1123/jab.2015-0339
The Effect of Backpack Carriage on the Biomechanics of Walking:
A Systematic Review and Preliminary Meta-Analysis
Bernard Liew, Susan Morris, and Kevin Netto
Curtin University
The aim of this systematic review was to evaluate the impact of bilaterally symmetrical backpack systems borne on the posterior
trunk on walking biomechanics, as backpacks represent the most prevalent method of load carriage in the military and civilian
population. A search of electronic databases was performed for studies that only investigated posteriorly-borne backpack car-
riage during level-grade walking (treadmill and over ground). Methodology of studies was assessed, and both meta-analysis
and qualitative synthesis were completed. Fifty-four studies were included in this review. In summary, the available literature
showed that backpack carriage in walking was associated with an increased trunk exion angle, increased hip and ankle range of
motion, increased vertical and horizontal ground reaction force, increased cadence, and reduced stride length. Several variations
in backpack carriage protocols could explain between-study variations in results, including: walking speed, backpack carriage
skill level, the use of a hip belt, and posterior displacement of the load away from the trunk. The ndings of this systematic
review will inform backpack carriage practices in the area of injury risk assessment and physical performance enhancement.
Keywords: backpack, walking, biomechanics, systematic review
Despite advances in modern transportation, backpack carriage
remains an essential mode of object transportation in occupations.1
In addition, the rising popularity of adventure sport2 and ultra-
endurance events among recreational and elite athletes3 broadens
the utilization of backpack carriage systems. The varied spectrum
of individuals, and type of occupational and sporting task involved
in backpack carriage, has seen a wealth of studies investigating the
impact of symmetrically-borne backpack systems (ie, backpack
borne equally on bilateral shoulders on the posterior trunk) on the
biomechanics of gait, especially in walking.
The study of backpack carriage biomechanics is important
given that suboptimal gait mechanics may undermine injury risk4
and physical performance.5 In the adult population, the increased
overuse injury rate associated with backpack carriage has largely
been investigated within the military context. A previous study
reported that 8% of the 5000 injuries reported in the Australian
Defense Force from January 2009 to December 2010 were related to
heavy backpack carriage.6 Of these injuries, 56% affected the lower
limb, while 26% affected the spinal region, and these injuries were
attributed to muscular stress.6 In addition, greater increases in back-
pack load magnitude results in a reduction in physical performance
in soldiers by approximately 1% per kilogram of load borne.1 The
increased lower limb and spinal injury risk associated with backpack
carriage, and the importance of lower limb muscles in performing
mechanical work in walking, means that a signicant number of
primary biomechanical studies have focused on the mechanical
effects of backpack carriage on these anatomical regions.
The biomechanical outcome variables investigated in the area
of backpack carriage can be broadly categorized into 4 groups: (1)
kinetics (eg, joint moments), (2) kinematics (eg, joint angle), (3)
spatiotemporal (eg, cadence), and (4) electromyography (EMG)
(eg, integrated EMG [iEMG]). Collectively, these biomechanical
outcome variables increase the understanding of potential injury
causative and physical performance deterioration mechanisms.
For example, ground reaction force (GRF) analysis of backpack
carriage walking has been performed in both the civilian and
military population,7,8 as increased rst vertical GRF peak has
been linked to increased incidence of lower limb stress fractures.9
In addition, joint kinetic analyses coupled with EMG analysis
has been routinely undertaken,8,10,11 as these parameters provide
information about the increased metabolic cost during backpack
walking.11 When a backpack is carried in walking, a signicant
increase in the metabolic cost arises due to an increased energy
demand associated with generating muscle force to support the
increased total weight.12 A thorough understanding of the effects of
backpack walking biomechanics could inuence the development
of injury prevention and performance enhancing interventions. For
example, recent understandings on the increased mechanical work
demand of the ankle during backpack carriage walking has led to
the development of energy-saving ankle exoskeletons.13 In addition,
an increased understanding of alterations in GRF patterns during
backpack carriage has led to the development of pressure-relief
shoe insoles with the aim of reducing the heightened risk of lower
limb overuse injuries.14
Although previous reviews have reported about the impact of
load carriage on walking biomechanics,1,15 these reviews did not
parse out the relative contributions of anterior and posterior load
systems on walking biomechanics, since it was not the primary aim
of the reviews. Loads carried symmetrically on the posterior trunk
induce distinct biomechanical changes to walking compared with
loads carried on the anterior trunk.16 Furthermore, these reviews1,15
are more accurately labeled as “narrative” reviews, where explicit
methods on identication, inclusion, quality appraisal, and synthesis
of current studies have not been used. Given that backpacks remain
the most prevalent means of load carriage in the military and civilian

Biomechanics of Backpack Carriage 615
JAB Vol. 32, No. 6, 2016
setting, the aim of this systematic review was to evaluate the impact
of bilaterally symmetrical backpack systems borne on the posterior
trunk on walking biomechanics in a general adult population.
Methods
Literature Search Strategy
A systematic review was performed in line with the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses
(PRISMA Statement17). The protocol of this review was not
registered. A single author (BL) conducted a literature search on
electronic databases PubMed (Medline), Ovid (AMED, EMBASE,
GlobalHealth, Medline), EBSCO (SportDiscus, CINAHL), and
Current Contents Connect, from inception until April 30, 2014. The
search was updated from April 2014– February 25, 2016. The gray
literature was searched using OpenGrey and The Gray Literature
Report.18,19 The search strategy was designed to identify studies
that investigated the biomechanical effect of backpack carriage on
walking. The specic search strategy used for each database was
adapted individually, consisting of a combination of key words
and subject heading search, and can be found in the electronic
supplementary material (ESM [available online]) (Appendix 1).
Variations in search strategy between databases were necessary
to exploit each database’s usage of Medical Subject Headings
(MeSH). No limits on language and date were initially imposed.
All titles and abstracts initially identied through the searches
were downloaded into Endnote version 7 (Thomson, Reuters,
Carlsbad, CA), cross-referenced, and any duplicate references
were deleted. As a preliminary screen, each title was evaluated by
a single author (BL) for potential inclusion, while a subsequent
evaluation of abstract and/or full text for inclusion was performed
by 2 authors (BL and SM). The reference lists of all included stud-
ies were searched recursively until no additional eligible publica-
tions were identied. Forward citation of included articles using
SCOPUS was performed. Unpublished work was not sought in
this review.
Selection Criteria
Articles were included into this review if they met the following
criteria:
1. Full-text papers published in English language
2. Evaluated backpack carriage on gait in skeletally-mature adults
(18 to 65-years-old), or if skeletally-mature adults were evalu-
ated as a separate cohort that enabled data extraction
3. Investigated the biomechanics of the trunk and/or lower limb
during level walking when carrying a backpack relative to body
weight (BW) walking (walking with no backpack)
4. Backpacks were worn symmetrically on both shoulders on the
posterior surface of the trunk
Articles were excluded from this review if they met the fol-
lowing criteria:
1. Conference proceedings, abstracts, thesis, technical reports
2. Participants had any existing diseases that could alter gait
mechanics and cause pain on testing
3. Studies that evaluated only unilateral and crossed-sling carriage
systems, anterior-posterior carriage systems, military webbing
systems, and ries
4. Studies that evaluated a composite set of load carriage systems
(eg, rie plus vest plus backpack), unless data on backpack
carriage only could be extracted
5. Studies that evaluated static postural alignment, postural stabil-
ity
6. Biomechanical simulation studies
Quality of Reporting and Risk of Bias
Assessment
Publications that met the inclusion criteria were assessed for quality
of reporting and risk of bias by 2 independent reviewers (BL and
SM). Quality of reporting was assessed using a modied version
of the STrengthening the Reporting of OBservational studies in
Epidemiology (STROBE) checklist20 (see Appendix 2 in ESM).
Modications were made to the STROBE checklist to only identify
reporting criteria essential for the judgment of risk of bias, judg-
ment of external generalizability of results, and replicability of the
study’s methods. The risk of bias tool (see Appendix 3 in ESM)
was developed to detect 5 potential biases which could confound
contemporary biomechanical studies.21,22 First, bias could arise from
the order of condition administered. Second, bias could arise when
a signicant time has elapsed between repeated testing, such that a
participant’s baseline physiological status is no longer maintained.
Third, bias can arise when participants drop out within a testing
session. Fourth, bias can arise when only signicant ndings were
reported. Lastly, bias can arise when methods used to derive the
dependent variables are not valid and reliable. Overall, risk of bias
within a study was judged based on the criteria recommended in the
updated Cochrane Handbook for Systematic Reviews of Interven-
tions.21 A study was judged as an overall “low risk of bias” when
all domains were judged as “low risk”. It was judged as an overall
“high risk of bias” if 1 or more domains were judged as “high risk”.
It was judged an overall “unclear risk of bias” if 1 or more domains
were judged as “unclear risk”. Between-rater agreement for the risk
of bias assessment was measured using the percentage agreement.
Disagreements were discussed at a consensus meeting.
Data Extraction and Analysis
The following data were extracted from all included articles: authors,
population sources, sample size, demographics (age, sex, weight,
and height), backpack type, locomotion protocol, load magnitude,
and dependent variables studied. Load magnitude was stratied into
“light” (< 20% BW), “medium” (20% to 30% BW), and “heavy”
(> 30% BW). The following criteria were met to be considered for
data-pooling across studies: (1) a dependent variable was evaluated
in more than 1 study and (2) when a variable was assessed within
the same phase of gait cycle (where applicable). Data pooling was
performed irrespective of the risk of bias score23 to increase the
number of studies eligible for data pooling. Subgroup analysis of
gait data were performed based on the load magnitude category.
Data pooling was performed in Cochrane Review Manager (v 5.3),
using the standardized mean difference (SMD) in a random-effects
model. Reported SMDs (95%CI) were categorized as small (<
0.60), medium (0.60 to < 1.20) or large (≥ 1.20),23 with statistical
signicance set at P < .05. Heterogeneity between studies was
assessed using the I2 test. Thresholds for the interpretation of I2
were determined as follows: (1) not important (< 40%), (2) moderate
heterogeneity (30% to 60%), (3) substantial heterogeneity (50% to
90%), (4) considerable heterogeneity (75% to 100%).21 Reported
I2 value importance will be interpreted relative to the direction and
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

616 Liew et al
JAB Vol. 32, No. 6, 2016
magnitude of the effect sizes.21 Parameters that were not pooled
were qualitatively synthesized.
Results
Search Results
From a total of 1261 results originating from the search strategy, 54
studies were nally included in this review (Figure 1).
Characteristics of Included Studies
Recruited participants ranged from university students,24 rec-
reational hikers,10 to military personnel (Table 1).25 Twenty-six
studies did not explicitly state the population from which par-
ticipants were recruited. The number of participants included
ranged from 426 to 60.7 Eighteen studies assessed backpack car-
riage in a male cohort25–42 and 7 studies in a female cohort,10,43–47
while 29 studies assessed a mixed-sex cohort,7,8,11,14,24,48–71 and
1 study72 did not mention the sex prole of the included par-
ticipants (Table 1). Only 1 study included both healthy individuals
and individuals with pes planus.63 Of the 54 studies, 20 studies
used a treadmill,11,24,26,31,33,37,39–41,48,51,54–56,59,62,67,69,71,72 22 studies
assessed gait using participants’ self-paced speed,7,10,28–30,34–38,42,45–
48,50,52,63,67,68,70,72, with 1 study not reporting the method of speed
selection,53 and 19 studies explicitly mentioning the use of a hip
belt.10,11,24,26,27,33,41,43–46,54–58,62,65,66
Quality of Reporting and Risk of Bias
Experimental setting (criterion 5) and drop outs (criterion 9) were
the 2 most poorly-reported criteria (Table 2).The between-rater
percentage agreement for the risk of bias assessment ranged from
61.1% to 88.9% (Table 3). A low risk of bias was scored for all stud-
Figure 1 — PRISMA ow diagram of included studies.
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

617JAB Vol. 32, No. 6, 2016
Table 1 Characteristics of included studies
No. Author(s) Population Sex Age
(Years) Height (m) Weight (kg) Backpack Type Dependent
Variables Hip Belt
Mentioned? O/T? F/S?
Prepost Fatigue/
Prolonged
Protocols
1 Abaraogu et al70 Students 13 M, 12 F 22.1 (2.7),
21.8 (3.1) Unknown Unknown Standard nylon
backpack with
internal frame
3 N O S N
2 Bobet &
Norman27
Unknown 11 M 19–22 1.66–1.90 53–85 Custom backpack 4 Y O F N
3 Caron et al48 Unknown 9 M, 8 F 25.4 (5.2) 1.70 (0.70) 70.6 (11.0) Custom frame
with water tank 1 N T S(b) N
4 Caron et al67 Unknown 9 M, 8 F 25.4 (5.2) 1.70 (0.70) 70.6 (11.0) Custom frame
with water tank 1 N T S(b) N
5 Castro et al7Students 30 M, 30 F 23.0 (3.7) 1.68 (0.10) 67.8 (11.2) Unknown type 2,3 N O S N
6 Castro et al68 Students 30 M, 30 F 22.8 (3.8) 1.688 (0.088) 65.5 (9.8) Unknown type 2 N O S N
7 Charteris28 Unknown 45 M 21.6 (4.18) 1.78 (0.08) Unknown H-frame hiker’s
backpack 3 N O S N
8 Cook &
Neumann49
Unknown 12 M, 12 F 26.8 (22–
36), 25.1
(21–35)
1.64, 1.37 69.8, 65.7 Rigid frame
backpack 4 N O F N
9 Dames & Smith71 Unknown 7 M, 5 F 24 (2) 1.73 (0.13) 71.1 (16.9) Backpack 1,3 N T F N
10 Peduzzi de Castro
et al14
Unknown 10 M, 11 F 25.81
(2.47) 1.68 (0.07) 63.62 (6.96) Unknown type 2 N O F N
11 Devroey et al50 Students 12 M, 8 F 23.9 (2.59) 1.76 (0.07) 69.41 (7.68) Backpack 1,3,4 N O S N
12 Fiolkowski et al51 Unknown 5 M, 8 F 24.6 (2.9),
25 (3.3) 1.75 (0.09),
1.68 (0.12) 74.1 (13.3),
61.7 (12.6) Backpack 1 N T F(b) N
13 Ghori &
Luckwill72
Unknown 6 individuals 17–35 Unknown Unknown Frameless
rucksack 1,3,4 N T S(b) N
14 Gillet et al29 Unknown 13 M 26 (3) Unknown 79 (12) Custom backpack 1,2 N O S N
15 Goh et al30 Soldiers 10 M 19.9 (1.1) 1.70 (0.06) 57.1 (2.2) Backpack 1,2,3 N O S N
16 Hageman et al52 University
cohort 9 M, 8 F 26 (3) 1.72 (0.09) 68.5(9.7) Custom internal
frame backpack 1,2 N O S N
17 Hall et al53 Unknown 9 M, 7 F 26 (3) 1.73 (0.08) 69.3 (9.4) Backpack 2 N O U N
18 Holewijn26 Unknown 4 M 24 (range
23–26) Unknown 75.1 (range
69–81.5) Military high
pack & custom
pack
4 Y T F N
19 Holt et al54 Unknown 5 M, 6 F 26 (7.1) Unknown Unknown Custom rigid
framed backpack 1,2 Y T F(a) N
20 Holt et al55 Unknown 5 M, 7 F 26 (7.1) Unknown Unknown Custom rigid
framed backpack 2 Y T F(a) N
21 Hsiang &
Chang31
Unknown 15 M 32.8 (9.4) 1.80 (0.07) 79.0 (17.1) Canvas bag 2 N T F(a) N
(continued)
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

618 JAB Vol. 32, No. 6, 2016
No. Author(s) Population Sex Age
(Years) Height (m) Weight (kg) Backpack Type Dependent
Variables Hip Belt
Mentioned? O/T? F/S?
Prepost Fatigue/
Prolonged
Protocols
22 Huang & Kuo11 Unknown 6 M, 2 F 19–26 Unknown 71.1 (12.0) External frame
backpack 1,2,3 Y T F N
23 Kinoshita32 Unknown 10 M Unknown 1.69 (0.05)
cm 64.0 (5.1) Rigid frame
backpack 1,2,3 N O F N
24 Krupenevich et
al8
Mix of students
and Army
Reserve Ofcer
Trainees
11 M, 11 F 20 (2.3),
20(1.8) 1.79 (0.09),
1.71 (0.08) 79.1 (13.3),
72.9 (15.1) MOLLE 1,2,3 N O F N
25 LaFiandra et al33 Soldiers 11 M 22.45
(3.83) 1.79 (0.11) 85.87 (17.36) MOLLE 2 Y T F N
26 LaFiandra et al24 University
cohort 5 M, 7 F 26 (SEM 2) Unknown Unknown Rigid framed
backpack 2 Y T F(a) N
27 LaFiandra et al56 Students 5 M, 7 F 26 (7.1
SEM) Unknown Unknown Rigid framed
backpack 1,3 Y T F(a) N
28 Lee et al34 Unknown 7 M 29.28
(2.14) 1.75 (0.04) 75.1 (8.6) Backpack 2,3 N O S(a) N
29 Ling et al43 Unknown 7 F 24.5 (3.4) 1.64 (0.05) 55.7 (6.2) MOLLE 1,3 Y O F N
30 Lloyd & Cooke57 Unknown 4 M, 5 F 24.7 (4.3) 1.73 (0.11) 73.4 (16.4) Double pack &
backpack 1,3 Y O F N
31 Lloyd & Cooke58 Unknown 4 M, 5 F 24.7 (4.3) 1.73 (0.11) 73.4 (16.4) Double pack &
backpack 1,2,3 Y O F N
32 Lloyd et al44 Xhosa students 16 F 21.3 (2.2) 1.57 (0.05) 62.7 (9.6) Backpack 2,3 Y O F N
33 Lucas-Cuevas
et al59
Students 16 M, 13 F 24.67
(4.38),
24.28
(2.06)
1.79 (0.05),
1.72 (0.07) 76.09(6.76),
62.18 (8.14) Hiker’s backpack 2 N T F N
34 Majumdar et al35 Soldiers 10 M 23.3 (2.6) 1.72 (0.04) 64.3 (7.4) Haversack &
backpack 1,3 N O S N
35 Majumdar et al36 Soldiers 10 M 23.3 (2.6) 1.72 (0.04) 64.3 (7.4) Haversack &
backpack 2 N O S N
36 Neumann &
Cook60
Unknown 12 M, 12 F 26.8 (22–
36), 25.1
(21–35)
1.73 68.7 Rigid framed
backpack 4 N O F N
37 Qu37 Current students
(post soldiers) 12 M 25.6 (2.4) 1.72 (0.05) 69.2 (11.2) Backpack 1 N T S(b) Y
38 Quesada et al25 Soldiers 12 M 22.4 (2.3) 1.79 (0.08) 78.6(11.8) ALICE 1,2 N O F Y
39 Rose et al61 Unknown 8 M, 8 F 23.1 (19–
32) 1.74 (0.10) 72.9 (13.9) Custom backpack 2,4 N O F N
Table 1 (continued)
(continued)
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

619JAB Vol. 32, No. 6, 2016
No. Author(s) Population Sex Age
(Years) Height (m) Weight (kg) Backpack Type Dependent
Variables Hip Belt
Mentioned? O/T? F/S?
Prepost Fatigue/
Prolonged
Protocols
40 Sharpe et ale62 Unknown 7 M, 5 F 23.0 (3.8),
27.2 (8.8) 1.74 (0.07),
1.70 (0.11) 74.5 (5.2),
65.6 (8.6) Modied
MOLLE 1 Y T F(a) N
41 Simpson et al45 Recreational
hikers 15 F 22.3 (3.9) 1.69 (0.10) 61.2 (5.3) Hiking backpack 1 Y O S Y
42 Simpson et al10 Recreational
hikers 15 F 22.3 (3.9) 1.69 (0.10) 61.2 (5.3) Hiking backpack 4 Y O S Y
43 Simpson et al46 Recreational
hikers 15 F 22.3 (3.9) 1.69 (0.10) 61.2 (5.3) Hiking backpack 1,2,3 Y O S Y
44 Smith et al47 College students 30 F 22.4 (2.2) Unknown 65.3(7.5) Unframed
backpack 1 N O S N
45 Son63 Unknown Flat foot: 8 M,
6 F Control: 7
M, 5 F
Flatfoot:
22 (0.7)
Control:
23.1 (0.7)
Flatfoot:
1.63 (0.02)
Control: 1.69
(0.02)
Flatfoot: 57.1
(3.3) Control:
64.2 (2.1)
Backpack 2,4 N O S N
46 Tilbury-Davis &
Hooper38
Soldiers 10 M 24.5 (3.5) 1.77 (0.07) 78.03 (6.66) Backpack 1,2 N O S N
47 Wang et al39 Students 18 M 21 (2) 1.81 (0.04) 77.6 (9.6) MOLLE 1,2 N T F Y
48 Wang et al40 Students 18 M 21 (2) 1.81 (0.04) 77.6 (9.6) MOLLE 2 N T F Y
49 Wang et al64 Students 15 M, 15 F 21.93
(2.73) 1.74 (0.08) 68.84 (12.83) Backpack 2,3 N O F(b) N
50 Watenabe &
Wang69
Unknown 5 M, 4 F 27.4 (5.0) 1.72 (0.081) 69.4 (11.8) Backpack 2,3 N T F N
51 Xu et al41 Students 9 M 26.3 (1.5) 1.77 (0.04) 70.8 (12.0) ALICE with
suspension plate 2,3 Y T F(a) N
52 Yang et al42 Unknown 10 M 24.5 (1.5) 1.712(0.042) 64.1(15.9) Unknown 2 N O S N
53 Yen et al65 University
cohort 5 M, 5 F 30.1 (3.7) 1.70 (0.13) 67.7 (14.2) MOLLE 1,3 Y O F N
54 Yen et al66 University
cohort 5 M, 5 F 30.1 (3.7) 1.70 (0.13) 67.7 (14.2) MOLLE 1 Y O F N
Abbreviations: M = male; F = female; SEM = standard error of measurement; MOLLE = modular lightweight load-carrying equipment; ALICE = all-purpose lightweight individual carrying equipment; N = no; Y =
yes; O = over ground; T = treadmill; F = xed; S = self-paced; (a) = multiple speed conditions; F(b) = xed cadence; S(b) = xed at self-paced speed.
Note. Dependent variables: 1 = kinematics, 2 = kinetics, 3 = spatiotemporal, 4 = electromyography.
Table 1 (continued)
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

620 Liew et al
JAB Vol. 32, No. 6, 2016
ies on baseline factors (criterion 2) and selective outcome reporting
(criterion 4) (Table 3). Order of presentation (criterion 1) had the
highest number of studies judged to have a high risk of bias (Table
3). For incomplete data reporting (criterion 3), 39 studies scored an
unclear risk of bias, as they did not report explicitly if all partici-
pants completed all testing conditions, which made it impossible
to judge if drop-outs occurred (Table 3). For validity and reliability
of outcome measures (criterion 5), 33 studies scored a low risk of
bias, with 21 studies scoring an unclear risk of bias (Table 3). Four
studies scored an overall low risk of bias, with 7 studies scoring an
overall high risk of bias (Table 3).
Impact of Backpack Carriage on Walking
Kinematics
Results of Meta-Analysis. The results of the meta-analysis dem-
onstrated that backpack carriage was associated with an increased
hip sagittal plane range of motion (ROM) (SMD = 2.94; 95% CI
1.88–4.00; I2 = 88%; P < .001), unchanged knee sagittal plane ROM
(SMD = –0.22; 95% CI –0.57 - 0.14; I2 = 34%; P = .13), an increased
ankle sagittal plane ROM (SMD = 0.80; 95% CI 0.51–1.09; I2 = 0%;
P = .83), and unchanged trunk sagittal plane ROM (SMD = –0.18;
95% CI –0.63 - 0.26; I2 = 52%; P = .05) (Table 4).
Results of Qualitative Synthesis. Although meta-analysis could
not be performed on trunk segment exion angle, an increase in
trunk segment exion angle was consistently found when a backpack
was carried in walking (Appendix 4 in ESM).8,29,30,32,35,45,48,50,57,58
Backpack carriage had variable inuences on hip, knee, and ankle
angles at specic gait phases.
Between initial contact to loading response, studies have
reported an increased39 and unchanged35 hip exion angle (Appendix
4 in ESM). For knee angle, studies have reported increased32,39,46
and unchanged knee exion angle.35 For ankle angle, carrying a
backpack reduced ankle dorsiexion (DF),32 although 2 studies
reported unchanged ankle DF angle.35,39,46 For segment angles,
Caron et al48 reported signicantly more extended thigh, leg, and
foot segment angles with respect to the vertical only at a load of 40%
BW compared with BW walking. However, Kinoshita32 reported a
signicant decrease in thigh segment angle, a signicant increase
in leg segment angle, and unchanged foot segment angle with a
20% and 40% BW load compared with BW walking (Appendix
4 in ESM).
Between early to midstance, one study reported no change in
hip extension angle (Appendix 4 in ESM).35 Three studies reported
increased knee exion angle during backpack carriage,25,39,46 and
three studies reported no effect.35,72 One study reported increased
ankle DF angle,35 and one reported unchanged ankle DF angle.25 For
segment angles, Caron et al48 reported a signicantly less extended
thigh segment, a greater exed leg segment, and unchanged foot
segment at 40% BW compared with BW walking. In contrast,
Kinoshita32 reported unchanged thigh and foot segment angles,
but increased leg segment angle with a 20% and 40% BW load
compared with BW walking (Appendix 4 in ESM).
During terminal stance to preswing, backpack carriage was
associated with increased hip extension,32,35 and reduced48 and
unchanged hip extension angle (Appendix 4 in ESM).25 One study
reported reduced knee exion angle,32 while 2 studies reported
no effect.35,46 At the ankle, 5 studies reported unchanged plantar
exion (PF) angle.25,32,35,46,48 For segment angles, Caron et al48
reported a less extended thigh and leg segment, but unchanged foot
segment angle when carrying a 40%BW backpack. Kinoshita32
reported an increased thigh segment angle with a 20% and 40%
BW backpack, unchanged leg segment angle, and an increased
foot segment angle only with a 40% BW backpack (Appendix 4
in ESM).
Impact of Backpack Carriage on Walking Kinetics
Results of Meta-Analysis. Backpack carriage was associated
with increased rst vertical GRF peak (SMD = 2.29; 95% CI =
1.69–2.90; I2 = 82%; P < .001), increased second vertical GRF peak
(SMD = 2.16; 95% CI = 1.51–2.80; I2 = 80%; P < .001), increased
vertical GRF minima (SMD = 2.55; 95% CI 0.98–4.13; I2 = 90%;
P < .001), and increased vertical impulse (SMD = 3.08; 95%CI =
2.10–4.05; I2 = 80%; P < .001). Backpack carriage was also asso-
ciated with increased braking GRF peak (SMD = 2.08; 95% CI =
1.67–2.49; I2 = 8%; P = .37), increased braking impulse (SMD =
2.52; 95% CI = 1.87–3.17; I2 = 0%; P = .44), increased propulsive
GRF peak (SMD = 1.69; 95% CI 0.76–2.61; I2 = 79%; P < .001),
and increased propulsive impulse (SMD = 2.34; 95% CI 1.25–3.43;
I2 = 64%; P = .06) (Table 4).
Results of Qualitative Synthesis. The impact of backpack car-
riage on mediolateral (ML) GRF parameters was inconsistent
and depended on the type of GRF normalization. Backpack car-
riage increased absolute ML GRF7 and when normalized to body
mass,44 but this effect was not observed by Peduzzi de Castro
et al,14 who derived absolute ML GRF. In addition, backpack
carriage reduced ML GRF normalized to total weight7 and body
weight,38 and unchanged ML GRF normalized to body weight
(Appendix 5 in ESM).36 Five studies evaluated the impact of
backpack carriage on sagittal plane lower limb joint moment
and/or power variables,8,11,25,39,42 although Yang et al42 did not
perform a statistical analysis. Between initial contact and loading
response, backpack carriage was associated with increased hip
and knee extension internal moment,39 hip and knee power,39 and
unchanged ankle internal moment,39 but increased ankle absorp-
tive power.39 In contrast, descriptively no alterations in all 3 joint
internal moments and powers were observed during the rst 10%
of stance and during walking when load was added, although
no statistical analysis was performed in this phase (Appendix 5
in ESM).8
From early to midstance, knee extension internal moment
increased when a backpack was carried in walking (Appendix 5
in ESM).8,25 Knee negative power increased when a backpack was
carried during the early portion of this phase,8 while knee positive
power increased with a backpack during the latter portion of this
phase.11 From late stance to toe-off, backpack carriage was associ-
ated with increased hip extension internal moment in one study,25
but increased hip exor internal moment in another study.8 Two
studies reported an increased hip power generation during this
phase when a backpack was carried during walking.8,11 Increased
ankle PF internal moment was reported during backpack carriage
during this phase,8,25 and 2 studies reported increased ankle power
generation during backpack carriage compared with BW walking
(Appendix 5 in ESM).8,11
One study11 documented the effects of backpack carriage
on the cumulative joint work performed within a complete stride
(Appendix 5 in ESM). Backpack carriage increased positive work
performed at the ankle and knee, with no change at the hip,11
compared with BW walking. Backpack carriage also increased
negative work performed at the ankle and hip, but not at the knee,11
compared with BW walking. Backpack carriage was associated
with increased knee joint stiffness compared with BW walking
(Appendix 5 in ESM).54
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

621JAB Vol. 32, No. 6, 2016
Table 2 Quality of reporting
No. Author(s) 1 2 3 4 5 6 7 8 9 10
1 Abaraogu et al70 Y Y Y Y/NA Y N Y Y N Y/Y
2 Bobet & Norman27 Y Y Y Y/NA N Y Y Y N Y/Y
3 Caron et al48 Y Y N Y/NA Y Y Y Y Y Y/N
4 Caron et al67 Y Y N Y/NA Y Y Y Y Y Y/N
5 Castro et al7Y Y N N/NA N Y Y Y N Y/Y
6 Castro et al68 Y Y Y N/Y N Y Y Y N Y/Y
7 Charteris28 Y Y Y Y/NA N Y Y Y N Y/Y
8 Cook & Neumann49 Y Y Y Y/NA Y Y Y Y Y Y/Y
9 Dames & Smith71 Y Y Y Y/NA Y Y Y Y Y Y/Y
10 Peduzzi de Castro et al14 Y Y Y Y/NA N Y Y Y N Y/Y
11 Devroey et al50 Y Y Y Y/NA Y Y N Y Y Y/Y
12 Fiolkowski et al51 Y Y Y Y/NA N Y N Y N Y/Y
13 Ghori & Luckwill72 Y Y Y N/NA Y Y N Y N Y/Y
14 Gillet et al29 Y N N N/NA N N N Y N Y/Y
15 Goh et al30 Y Y Y N/NA N Y N Y Y Y/Y
16 Hageman et al52 Y Y N Y/Y N Y Y Y N Y/Y
17 Hall et al53 Y Y N Y/Y N N Y Y N Y/Y
18 Holewijn26 Y Y Y Y/NA Y Y Y Y N Y/N
19 Holt et al54 Y Y Y Y/N N Y Y Y N Y/Y
20 Holt et al55 Y Y Y Y/N N Y Y Y N Y/Y
21 Hsiang & Chang31 Y Y Y Y/Y N Y Y N Y Y/N
22 Huang & Kuo11 Y Y Y N/NA N Y Y Y N Y/Y
23 Kinoshita32 Y Y Y N/NA N Y N Y N Y/Y
24 Krupenevich et al8Y Y Y Y/NA Y Y Y Y Y Y/Y
25 LaFiandra et al33 Y Y Y Y/NA N Y Y Y Y Y/Y
26 LaFiandra et al24 Y Y Y Y/N N Y Y Y Y Y/Y
27 LaFiandra et al56 Y Y Y Y/N N Y Y Y Y Y/Y
28 Lee et al34 Y Y N N/N N Y Y Y N Y/Y
29 Ling et al43 Y Y Y N/NA N N N Y Y Y/Y
30 Lloyd & Cooke57 Y Y Y Y/N Y Y Y Y N Y/Y
31 Lloyd & Cooke58 Y Y Y Y/NA N N Y Y N Y/Y
32 Lloyd et al44 Y Y Y Y/NA N Y Y Y N Y/Y
33 Lucas-Cuevas et al59 Y Y Y Y/NA N Y N Y N Y/Y
34 Majumdar et al35 Y Y Y Y/NA Y Y N Y N Y/Y
35 Majumdar et al36 Y Y Y Y/NA N Y Y Y N Y/Y
36 Neumann & Cook60 Y Y Y Y/NA Y Y N Y Y Y/Y
37 Qu37 Y Y N Y/NA N Y N Y N Y/Y
38 Quesada et al25 Y Y Y Y/NA Y Y N N Y Y/N
39 Rose et al61 Y Y Y Y/NA Y Y Y Y N Y/Y
40 Sharpe et al62 Y Y Y Y/Y Y Y Y Y N Y/Y
41 Simpson et al45 Y Y Y Y/NA Y Y Y Y Y Y/Y
42 Simpson et al10 Y Y N Y/NA N Y Y Y Y Y/Y
43 Simpson et al46 Y Y Y Y/NA N Y Y Y Y Y/Y
44 Smith et al47 Y Y N N/NA N Y N Y Y Y/Y
45 Son63 Y N N N/NA N Y N Y N Y/Y
46 Tilbury-Davis & Hooper38 Y Y N Y/NA N Y N Y N Y/Y
47 Wang et al39 Y Y Y Y/NA Y Y Y Y N Y/Y
48 Wang et al40 Y Y Y Y/NA Y Y Y Y N Y/Y
49 Wang et al64 Y Y Y Y/Y N Y N Y Y Y/Y
50 Watanabe & Weng69 Y Y Y Y/Y N N Y Y Y Y/Y
51 Xu et al41 Y Y Y Y/Y Y Y Y Y N Y/Y
52 Yang et al42 Y Y Y N/N N Y Y N N Y/N
53 Yen et al65 Y Y Y Y/NA Y Y Y Y N Y/Y
54 Yen et al66 Y Y Y Y/NA Y Y Y Y N Y/Y
Abbreviations: Y = yes; N = no; NA = not applicable.
Note. Criteria 1: aims description; 2: dependent variables description; 3: interventions description; 4: load/speed order randomized; 5: setting description; 6: date
collection description; 7: data analysis description; 8: statistical description; 9: drop outs; 10: point estimate/ variability.
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

622 JAB Vol. 32, No. 6, 2016
Table 3 Risk of bias assessment
No. Author(s)
Order
Randomized or
Counterbalanced
Similar Baseline
Prognostic
Level
Incomplete
Outcome
Data
Selective Outcome
Measures
Reporting
Valid and
Reliable Outcome
Measured
Overall
Risk of
Bias
1 Abaraogu et al70 Low Low Unclear Low Unclear Unclear
2 Bobet & Norman27 Low Low Unclear Low Low Unclear
3 Caron et al48 High Low Low Low Low High
4 Caron et al67 High Low Low Low Low High
5 Castro et al7Unclear Low Unclear Low Low Unclear
6 Castro et al68 Unclear Low Unclear Low Low Unclear
7 Charteris28 Low Low Unclear Low Low Unclear
8 Cook & Neumann49 Low Low Low Low Low Low
9 Dames & Smith71 Low Low High Low Low High
10 Peduzzi de Castro et al14 Low Low Unclear Low Low Unclear
11 Devroey et al50 Low Low Low Low Unclear Unclear
12 Fiolkowski et al51 Low Low Unclear Low Unclear Unclear
13 Ghori & Luckwill72 Unclear Low Unclear Low Unclear Unclear
14 Gillet et al29 Unclear Low Unclear Low Unclear Unclear
15 Goh et al30 Unclear Low Low Low Unclear Unclear
16 Hageman et al52 Low Low Unclear Low Low Unclear
17 Hall et al53 Low Low Unclear Low Unclear Unclear
18 Holewijn26 Low Low Unclear Low Low Unclear
19 Holt et al54 Unclear Low Unclear Low Low Unclear
20 Holt et al55 Unclear Low Unclear Low Low Unclear
21 Hsiang & Chang31 Low Low Low Low Low Low
22 Huang & Kuo11 Unclear Low Unclear Low Unclear Unclear
23 Kinoshita32 Unclear Low Unclear Low Unclear Unclear
24 Krupenevich et al8Low Low Low Low Low Low
25 LaFiandra et al33 Low Low Low Low Low Low
26 LaFiandra et al24 Unclear Low Low Low Low Unclear
27 LaFiandra et al56 Unclear Low Low Low Low Unclear
28 Lee et al34 Unclear Low Unclear Low Low Unclear
29 Ling et al43 Unclear Low Unclear Low Unclear Unclear
30 Lloyd & Cooke57 Unclear Low Unclear Low Low Unclear
31 Lloyd & Cooke58 Low Low Unclear Low Unclear Unclear
32 Lloyd et al44 Low Low Unclear Low Low Unclear
33 Lucas-Cuevas et al59 Low Low Unclear Low Unclear Unclear
34 Majumdar et al35 Low Low Unclear Low Unclear Unclear
35 Majumdar et al36 Low Low Unclear Low Low Unclear
36 Neumann & Cook60 Low Low Low Low Unclear Unclear
37 Qu37 Low Low Unclear Low Unclear Unclear
38 Quesada et al25 Low Low Low Low Unclear Unclear
39 Rose et al61 Low Low Unclear Low Low Unclear
40 Sharpe et al62 Low Low Unclear Low Low Unclear
41 Simpson et al45 Low Low Unclear Low Low Unclear
42 Simpson et al10 Low Low Unclear Low Low Unclear
43 Simpson et al46 Low Low Unclear Low Low Unclear
44 Smith et al47 Unclear Low Low Low Unclear Unclear
45 Son63 Unclear Low Unclear Low Unclear Unclear
46 Tilbury-Davis & Hooper38 Low Low Unclear Low Unclear Unclear
47 Wang et al39 High Low Unclear Low Low High
48 Wang et al40 High Low Unclear Low Low High
49 Wang et al64 Low Low Low Low Unclear Unclear
50 Watanabe & Weng69 Low Low Low Low Unclear Unclear
51 Xu et al41 Low Low Unclear Low Low Unclear
52 Yang et al42 Unclear Low Unclear Low Low Unclear
53 Yen et al65 High Low Unclear Low Low High
54 Yen et al66 High Low Unclear Low Low High
68.5 77.8 63.0 88.9 61.1 NA
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

623JAB Vol. 32, No. 6, 2016
Table 4 Meta-analysis—Summary effect size of load carriage magnitude on biomechanical variables during walking
Biomechanics
Variable Dependent Variable Load
Category Subgroup Effect Size
(SMD, IV, Random, 95%CI) Summary Effect Size (SMD,
IV, Random, 95%CI) Random Order
Presentation Incomplete
Outcome Data Valid and Reliable
Outcome
Kinematics Hip ROM Low 0.64 (0.16,1.13) 1L;1U;1H 2U;1H 1L;2U
Medium 1.78 (0.48,3.09) 1U 1U 1U
High 4.28 (3.52,5.04) 2U 1L;1U 1L;1U
2.94 (1.88,4.00)
Knee ROM Low 0.21 (–0.26,0.0.68) 1L;1U;1H 2U;1H 1L;2U
Medium –0.14 (–0.60,0.32) 1L;1U 2U 1L;1U
High –0.83 (–1.43,–0.22) 1L;1U 2U 1L;1U
–0.22 (–0.57,0.14)
Ankle ROM Low 0.90 (0.40,1.40) 1L;1U;1H 2U;1H 1L;2U
Medium 0.67 (0.20,1.14) 1L;1U 2U 1L;1U
High 0.84 (0.29,1.38) 1L;1U 2U 1L;1U
0.80 (0.51,1.09)
Trunk ROM Low –0.08 (–0.59,0.43) 1L;1U 1L;1U 2U
Medium –0.31 (–1.36,0.73) 1L;1U 1L;1U 1L;1U
High 0.00 (–0.72,0.72) 1L 1U 1L
–0.18 (–0.63,0.26)
Kinetics First vGRF peak Low 0.92 (0.58,1.25) 3L 2L;1U 1L;2U
Medium 2.00 (1.59,2.41) 2L;1U 2L;1U 1L;2U
High 4.73 (3.67,5.80) 3L;1U;1H 1L;4U 3L;2U
2.29 (1.69,2.90)
Second vGRF peak Low 0.94 (0.49,1.40) 2L 1L;1U 1L;1U
Medium 1.90 (1.19,2.60) 2L;1U 2L;1U 1L;2U
High 4.51 (3.68,5.34) 3L;1U 1L;3U 2L;2U
2.16 (1.51,2.80)
vGRF minima Low 0.42 (–0.21,1.05) 1L 1U 1L
Medium 1.84 (0.76,2.93) 1U 1U 1U
High 4.41 (2.65,6.16) 2L;1U 3U 1L;2U
2.55 (0.98,4.13)
vGRF impulse Low 1.27 (0.57,1.97) 1L 1U 1L
Medium 3.46 (2.48,4.44) 1L;1U 2U 1L;1U
High 3.99 (2.41,5.56) 2L;1U 3U 2L;1U
3.08 (2.10,4.05)
Braking GRF peak Low 1.68 (0.93,2.43) 1L 1U 1L
Medium 1.78 (0.71,2.85) 1U 1U 1U
High 2.34 (1.81,2.88) 2L;1U;1H 4U 2L;2U
2.08 (1.67,2.49)
(continued)
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

624 JAB Vol. 32, No. 6, 2016
Biomechanics
Variable Dependent Variable Load
Category Subgroup Effect Size
(SMD, IV, Random, 95%CI) Summary Effect Size (SMD,
IV, Random, 95%CI) Random Order
Presentation Incomplete
Outcome Data Valid and Reliable
Outcome
Propulsive GRF peak Low 1.00 (0.33,1.67) 1L 1U 1L
Medium 2.07 (0.94,3.21) 1U 1U 1U
High 2.20 (0.03,4.37) 2L;1U 3U 1L;2U
1.69 (0.76,2.61)
Braking GRF impulse Low –
Medium 2.11 (0.97,3.25) 1U 1U 1U
High 2.72 (1.93,3.51) 1L;1U 2U 1L;1U
2.25 (1.87,3.17)
Propulsive GRF impulse Low –
Medium 1.53 (0.51,2.56) 1U 1U 1U
High 2.84 (1.23,4.45) 1L;1U 2U 1L;1U
2.34 (1.25,3.43)
Spatiotemporal Stride/step length Low 0.00 (–0.27,0.28) 5L;2U 3L;3U;1H 1L;6U
Medium –0.30 (–0.50,-0.11) 4L;3U 3L;4U 3L;4U
High –0.32 (–0.51,–0.13) 3L;3U 2L;4U 4L;2U
–0.24 (–0.36,–0.12)
Cadence Low 0.08 (–0.20,0.36) 3L;2U;1H 2L;4U 1L;5U
Medium 0.31 (0.10,0.53) 2L;2U;1H 2L;3U 2L;3U
High 0.37 (0.15,0.58) 1L;2U 1L;2U 2L;1U
0.28 (0.15,0.41)
Percentage rst double
support phase Low 0.27 (–0.42,0.97) 1L 1U 1U
Medium 0.65 (0.22,1.08) 1L;2U 3U 2L;1U
High 1.66 (1.07,2.26) 1L;1U 2U 1L;1U
0.99 (0.53,1.45)
Percentage second double
support phase Low –
Medium 0.27 (0.05,0.49) 1L;2U 3U 2L;1U
High 1.04 (0.45,1.63) 1L;1U 2U 1L;1U
0.67 (0.30,1.40)
Percentage single support
phase Low –0.18 (–0.80,0.44) 1L;1U 2U 2U
Medium –0.25 (–0.52,0.02) 1L;2U 3U 1L;2U
High –0.73 (–1.11,–0.36) 1L;2U 3U 1L;2U
–0.49 (–0.73,–0.25)
Abbreviations: SMD = standardized mean difference; IV = inverse variance; CI = condence interval; ROM = range of motion; vGRF = vertical ground reaction force; GRF = ground reaction force; U = unclear risk of bias; L
= low risk of bias; H = high risk of bias.
Table 4 (continued)
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

Biomechanics of Backpack Carriage 625
JAB Vol. 32, No. 6, 2016
Impact of Backpack Carriage on Spatiotemporal
Parameters
Results of Meta-Analyses. There was a small effect of load on
stride length (SMD = –0.24; 95% CI –0.36 - –0.12; I2 = 0%; P = .59).
There was again a small effect of load on walking cadence (SMD =
0.28; 95% CI 0.15–0.41; I2 = 0%; P = .94). Backpack carriage was
associated with a moderate effect on the percentage of gait spent
during rst double support (SMD = 0.99; 95% CI 0.53–1.45; I2 =
85%; P < .001), on second double support (SMD = 0.67; 95% CI
0.30–1.04; I2 =78%; P < .001), and on single support phase (SMD
= –0.49; 95% CI –0.73 - –0.25; I2 = 42%; P = .05) (Table 4).
Results of Qualitative Synthesis. Backpack carriage of up to
60% BW largely had no effect on swing time,28,35,50,71 although 1
study found an increase in swing duration with a 20% to 50% BW
load relative to BW walking (Appendix 6 in ESM).72 One study
reported step width did not change with backpack carriage although
the same study reported that step width variability increased linearly
as a function of load magnitude.11
Impact of Backpack Carriage on Walking
Neuromuscular Activation
Results of Meta-Analyses. No studies were suitable for meta-
analysis (see Discussion).
Results of Qualitative Synthesis. The impact of backpack car-
riage was evaluated on the following muscles: upper trapezius,26,27,50
sternocleidomastoid,50 latissimus dorsi,61 rectus abdominis,50,61,63
external oblique,50,61 internal oblique,61 erector spinae,27,49,50,61,63
gluteus medius,60,72 quadriceps complex,10,50,63,72 hamstring com-
plex,10,50,63,72 triceps surae,10,63,72 and tibialis anterior (Appendix 7
in ESM).10,63,72 Current studies have used different EMG indices
including: linear envelope,27,50 iEMG,10 percentage of maximal
voluntary contraction,60,63 burst duration,10,72 median power fre-
quency,10,26,72 and muscle onset timing (Appendix 7 in ESM).10
The spectrum of muscles investigated and different indices used
preclude quantitative synthesis.
From the studies, backpack carriage (> 15% BW) increased the
EMG amplitude of the triceps surae (Appendix 7 in ESM).10,63 The
effect of load on tibialis anterior EMG amplitude was equivocal,
with 2 studies reporting opposite ndings.10,63 Two studies which
investigated loads of up to 20% BW reported increased quadriceps
complex EMG amplitude,10,63 although 1 study which only inves-
tigated loads of up to 15% BW reported no change (Appendix 7
in ESM).50 All studies reported no changes to hamstring complex
EMG amplitude with backpack carriage.10,50,63 There was incon-
sistent evidence demonstrating the effect of load on increased
upper trapezius (23% to 36% BW),27 increased rectus abdominis
(> 10% BW),50 and decreased erector spinae EMG amplitude (<
10% BW).50 Cook and Neumann49 did not nd a signicant effect
of load on erector spinae amplitude when investigating loads of up
to 20% BW, while Bobet and Norman27 found that erector spinae
amplitude increased with loads ranging from 23% BW to 36% BW.
Devroey et al50 did not nd a signicant effect of load on upper
trapezius amplitude with loads of up to 15% BW. Son63 did not nd
a signicant effect of load of up to 20% BW on rectus abdominis
amplitude (Appendix 7 in ESM). There was no effect of load on
the burst duration of tibialis anterior and medial gasctronemius.10
Backpack carriage increased quadriceps complex burst duration
in 1 study (up to 50% BW),72 but this was not supported in a later
study (up to 40% BW).10 The effect of load on hamstring muscle
complex varied with muscle group. The lateral hamstring complex
increased burst duration,10 but no change was observed with the
medial complex (Appendix 7 in ESM).10,72
Discussion
Symmetrically-worn backpacks on the posterior trunk remains a
universal means of load transport for the general adult (military and
civilian) population. Understanding the biomechanical impact of
backpack carriage on walking could inform future injury prevention
and physical conditioning programs of the neuromuscular demands
involved. Although many studies on backpack carriage have been
performed, this is the rst systematic review to our knowledge to
have quantitatively synthesized the effects of backpack carriage on
walking biomechanics.
Integrating Risk of Bias and Effect Size
Estimation
Between-study variations in risk of bias scores occurred for order
of random presentation, incomplete data reporting, and validity
and reliability of outcome measures used. For example, inclusion
of 1 high risk of bias study71 into the meta-analysis could have
overestimated the effect size observed for hip and ankle ROM “low
load” subgroup analyses (Table 4). The high risk of bias study could
have resulted in an overestimate of the effect size observed in these
subgroup analyses. For rst vertical and braking GRF peak, the
inclusion of 1 high risk of bias study40 each would likely mean that
the effect size observed of high load (> 30% BW) may be smaller
than what was reported in this study, but potentially still remain-
ing signicantly large (Table 4). On the contrary, inclusion of one
high-risk study71 into the analysis of step length and 1 high-risk
study65 into the analysis of cadence may not inuence the pooled
respective results. This is so as the direction and magnitude of
effect of the high-risk studies appeared similar to that of studies
with low and unclear risk of bias. Of all the pooled biomechanical
variables, results of backpack carriage on step length and cadence
are supported by the most proportion of low risk of bias studies,
and would likely represent the closest estimate of a ‘true’ effect of
increasing load magnitude (Table 4).
Assessing Importance of Reported I2
Several outcome variables in the meta-analysis had substantial
to considerable heterogeneity (I2) despite subgroup analysis (eg,
trunk sagittal plane ROM, second vertical GRF peak, vertical GRF
minima, and vertical impulse). High I2 magnitude for second vertical
GRF peak, vertical GRF minima, and vertical impulse represents a
lower level of importance. This is so as the studies included for each
of these dependent variables demonstrated consistent direction of
effect, and the reported magnitude of effect size was large. The I2
magnitude in representing heterogeneity is more relevant for trunk
sagittal plane ROM due to less consistent between-study direction
of effect and a small effect size reported.
Goh et al30 reported that backpack carriage of 30% BW resulted
in greater trunk sagittal plane ROM compared with BW walking,
while Simpson et al45 reported smaller trunk sagittal plane ROM
using the same load magnitude. Although the study by Goh et al30
and Simpson et al45 independently evaluated only male and female
participants, respectively, the existing literature does not support
sex differences in load carriage walking biomechanics.8,73 In con-
trast, differences in trunk ROM could be attributed to the level of
fatigue experienced by the participants within each study’s protocol.
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

626 Liew et al
JAB Vol. 32, No. 6, 2016
Simpson et al45 measured trunk mechanics after participants had
walked at least 2 km in a prolonged 8-km task. In contrast, Goh et
al30 involved 3 successful walking trials within a standard biome-
chanics laboratory.
Biomechanical Changes During Backpack
Carriage
Global Biomechanical Changes. An increase in force applica-
tion needed to support an increased total weight, and accelerate
the COM, is evident from the large and consistent effect of load
magnitude on vertical GRF and horizontal GRF parameters (Table
4). However, when the vertical GRF was normalized to total mass
carried (ie, body plus load mass), backpack carriage reduced the
rst and second vertical GRF peak compared with BW walking,7,68
or resulted in no signicant differences between BW and backpack
walking.44 For horizontal GRF, participants in 1 study44 did not
increase the magnitude of braking GRF per unit total mass, but that
of 2 studies reported increased braking GRF per unit total mass,7,68
during backpack walking compared with BW walking.
Differences on the effect of total mass normalization between
studies could be due to differences in the population sampled.
Lloyd et al44 recruited Xhosa women with at least 10 years’ of load
carriage experience, while the population of Castro et al7,68 was
adult students. A previous study of native African woman during
load carriage found that these individuals were better able to har-
ness pendulum exchange of potential and kinetic energy during
walking than their European counterparts.74 This could imply that
many years of routine load carriage would have resulted in specic
neuromuscular adaptations which avoided increased braking GRF
peak per unit total mass during backpack carriage.
The effect of total mass normalization on the effects of back-
pack carriage on GRF parameters may depend on interactions with
spatiotemporal parameters and kinematics. These latter factors
could vary due to variations in load carriage skill level or imposed
experimental constraints. Experimentally-imposed restrictions on
cadence (and step length) could explain the mechanisms behind an
increased braking GRF per unit total mass in backpack carriage.
Castro et al68 reported that normalized braking GRF peak was
greater during backpack walking, compared with BW walking,
only at a relatively slow cadence of 70 steps⋅min–1. A relatively slow
xed cadence would enable longer strides to be maintained during
backpack carriage, which could result in higher normalized braking
GRF peak.75 However, even in protocols which allowed participants
to self-select their walking speed and cadence, participants in Lloyd
et al44 did not increase their per unit total mass braking GRF peak,
but participants in Castro et al7 did increase. Important distinctions
in the studies by Lloyd et al44 and Castro et al7 could be that the
former study investigated backpack carriage on native load carriers
and used a backpack with a hip belt, but the latter study investigated
adult students and did not report the use of a hip belt. The use of a
hip belt could act as a physical constraint for reducing step length.
Based on the ndings of this meta-analysis with regard to the
effect of high backpack load magnitude on step length (Figure A4a
in ESM), it is difcult to conclude if variations in population and
the use of a hip belt could inuence step length. This was because
the inherent magnitude of differences in step length between studies
was small to subjectively determine the pattern of effect, population
type, and effect hip belt has on step length during backpack walking.
Regional Biomechanical Changes in Specific Phases of Walk-
ing. The need to support an increased total weight and generate
larger horizontal forces during backpack carriage requires adaptive
responses from the neuromuscular system. The ndings from this
review found several biomechanical alterations in walking that could
be inferred as adaptive neuromuscular responses. Hip and knee
extensor internal moments increased during initial contact to loading
response,39 which could provide increased total weight support and
shock absorption.76 This response to backpack carriage may only
be apparent when carrying signicantly heavier weight (eg, > 40%
BW) coupled with a faster walking speed (eg, >1.6 m⋅s–1), as studies
using lighter loads and slower speeds did not identify signicant
changes to the hip and knee joint internal moments.8,11 It is unlikely
that these differences on joint moments could be due to variations in
the use of over ground walking compared8 to treadmill walking.11,39
The differences between over ground and treadmill BW walking
as it pertains to joint moment has been reported to be minimally
different (< 0.1 N⋅m⋅kg–1) at the same speed.77
Variations in initial contact joint moments could be due to the
different population sampled. Wang et al39 recruited adult students and
Krupenevich et al8 recruited a mixture of students and military person-
nel. Huang and Kuo11 did not report their sample type. As previously
discussed, a relatively novice load carrier applies greater braking GRF
per unit total mass during backpack relative to BW walking, and a
skilled carrier does not apply greater normalized braking GRF. This
could explain why participants in Wang et al,39 and not those from
Krupenevich et al,8 increased knee and hip joint internal moments
during initial contact where braking forces were high.
Ankle extensor internal moment and power generation
increased in late stance, which could provide increased total weight
support and propulsion forces.8,11,25 Our review also reported consis-
tently increased gastrocnemius EMG amplitude from 2 studies,10,63
reecting the increased importance of the ankle extensors in gen-
erating supportive and propulsive forces in backpack carriage. An
increased ankle power generation during backpack carriage in late
stance was not related to an increased ankle PF angle,32,35,46 or PF
velocity,71 but due to an increased ankle PF internal moment.8,11,25
In military personnel, there is qualitative evidence (visual
inspection of gures) that backpack carriage not only increased
ankle push-off power, but peak power occurred slightly earlier in
late stance, compared with BW walking.8 Although the population
of Huang and Kuo11 was not reported, backpack carriage resulted
in a slight delay in peak ankle power in late stance, compared with
BW walking. A previous study found that the temporal coupling
between the trailing and the lead stance limb force generation plays
an important role in the energetics of BW walking.78 An earlier
push-off by the trailing limb by skilled load carriers, relative to
novice load carriers, could minimize increased braking GRF as a
result of backpack carriage in the lead limb, and minimize potential
increases in the lead limb’s joint internal moments.
The effects of backpack carriage on hip joint internal moment
in late stance was inconsistent, with 1 study reporting increased hip
extensor internal moment,25 1 study reporting increased hip exor
internal moment,8and 1 study descriptively reporting increased hip
exor internal moment.11 Variations of backpack carriage on late
stance hip joint moment may be related to the magnitude of forward
trunk exion angle. Participants in Krupenevich et al8 walked with
an average of 11° to 14° trunk exion when a backpack was car-
ried. Trunk exion angle was not reported in Quesada et al25 and
Huang and Kuo.11 A previous study on trunk exion in running
documented an increased hip extensor internal moment in late stance
with increased trunk exion angle.79
Surprisingly, backpack carriage did not consistently increase
the EMG amplitude of the quadriceps complex. The iEMG of the
vastus lateralis increased with backpack carriage starting from 20%
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

Biomechanics of Backpack Carriage 627
JAB Vol. 32, No. 6, 2016
BW,10 but did not increase the linear envelope of rectus femoris
even at a load of 15% BW.50 This highlights the importance of
considering intermuscular functional differences. Importantly, the
vastus lateralis has been shown to be important for energy absorption
and BW support in early stance, while the rectus femoris transfers
energy to the trunk in swing.80 This review found that the role of the
knee joint in backpack carriage may lie in increasing intersegmental
energy transfer81 by increasing knee joint stiffness.54
Optimal Mechanical Strategies in Backpack Carriage Walking.
Reduced stride length and increased cadence were only observed
when load magnitude exceeded 30% BW (Table 4). The energetic
cost of walking is proportional to the fourth power of stride/step
length and to the third power of cadence.82 It is likely that to maintain
walking speed at heavy backpack loads, it is energetically advanta-
geous to increase cadence and reduce stride length.82 Walking with a
backpack, as compared with BW walking, would increase the trans-
verse plane moment of inertia.24 The inherent mechanical response
to load would be to reduce transverse plane pelvic rotation,56,62
which reduces counter-rotation torque of the trunk–backpack seg-
ment.56 To maintain step length, a compensatory increase in sagittal
plane hip joint ROM would need to occur (Table 4). Applying a
hip belt during backpack carriage increases pelvic rotation62 and
may reduce the need to increase sagittal plane hip excursion. A hip
belt in backpack walking could reduce reliance on sagittal plane
lower limb biomechanical changes to maintain stride length. This
could explain why studies which reported a hip belt usage resulted
in smaller propulsive GRF and impulse,44,58 compared with studies
which did not report a hip belt usage.32
A very consistent nding in this review was that backpack
carriage increased trunk exion angle, without increasing ROM.
A relatively restricted trunk ROM could represent a mechani-
cal strategy to tightly regulate the kinematic coordination of the
trunk and lower limb in the sagittal plane.48,66,67 The spatial and
temporal coordination between the trunk and leg segments could
inuence both postural control,66 and potentially the energetics of
backpack carriage, by inuencing the spatiotemporal and kinematic
parameters of walking.48,67 A previous study by Krupenevich et al8
reported that a reduction in trunk exion could be associated with
an increased second vertical GRF peak in backpack walking. An
increased trunk exion segment angle could result in more optimal
alignment of applied GRF in the direction of the trailing limb.
Skilled load carriers may increase their trunk exion angle during
backpack walking, relative to their baseline trunk angle during
BW walking, compared with novice load carriers. Studies which
involved military personnel8,30 increased trunk exion angle from
BW walking by 13° to 15°, while a study which included recre-
ational hikers increased trunk exion angle by only 9°.
Although an increased trunk exion angle was associated with
an increased load magnitude, there is no consistent evidence that the
vertical position of the backpack on the trunk mediates trunk exion
angle.50 However, posterior displacement of the backpack’s COM
away from the trunk may be involved in mediating the magnitude
of trunk exion angle, due to alterations in externally imposed
extensor torque. This could explain why a front–back loaded pack
resulted in less forward trunk exion compared with a backpack
of similar load mass.58
Unlike clinical trials, there are no current standards to judge the
quality of reporting and risk of bias of contemporary biomechani-
cal studies. Biomechanical studies are often of a cross-sectional
repeated-measures design with distinct requirements for handling
recruitment, allocation, and drop-out issues, unlike clinical trials.
It is for this reason that the authors developed and used a tailored
appraisal checklist for the purpose of this study. The lack of valida-
tion on the appropriateness of these checklists may be deemed as
a limitation of this study. Further research is needed to establish
reporting and risk of bias standards that are relevant for contempo-
rary biomechanical studies.
In summary, the available literature showed that backpack
carriage was associated with several consistent biomechanical
alterations in walking: increased trunk exion angle, increased hip
and ankle ROM, increased vertical and horizontal GRF parameters,
increased cadence, and reduced stride length. Several variations in
backpack carriage protocols could explain between-study variations
in results, including: walking speed, backpack carriage skill level,
the use of a hip belt, and posterior displacement of the load away
from the trunk. The ndings of this systematic review would inform
backpack carriage practices in the area of injury.
Acknowledgment
No funds were received in support of this work. No benets in any form
have been or will be received from a commercial party related directly or
indirectly to the subject of this manuscript. Mr. Bernard Liew is currently
under a postgraduate scholarship: “Curtin Strategic International Research
Scholarship (CSIRS)”.
References
1. Knapik JJ, Reynolds KL, Harman E. Soldier load carriage: his-
torical, physiological, biomechanical, and medical aspects. Mil Med.
2004;169(1):45–56. PubMed doi:10.7205/MILMED.169.1.45
2. Bentley TA, Page S, Edwards J. Monitoring injury in the New Zea-
land adventure tourism sector: an operator survey. J Travel Med.
2008;15(6):395–403. PubMed doi:10.1111/j.1708-8305.2008.00234.x
3. Cejka N, Rust CA, Lepers R, Onywera V, Rosemann T, Knechtle
B. Participation and performance trends in 100-km ultra-marathons
worldwide. J Sports Sci. 2014;32(4):354–366. PubMed doi:10.1080
/02640414.2013.825729
4. Orr RM, Johnston V, Coyle J, Pope R. Reported load carriage injuries
of the Australian army soldier. J Occup Rehabil. 2015;25(2):316–322.
PubMed doi:10.1007/s10926-014-9540-7
5. Lazzer S, Taboga P, Salvadego D, et al. Factors affecting metabolic
cost of transport during a multi-stage running race. J Exp Biol.
2014;217(5):787–795. PubMed doi:10.1242/jeb.091645
6. Orr RM, Johnston V, Coyle J, Pope R. Reported load carriage injuries
of the Australian army soldier. J Occup Rehabil. 2015;25(2):316–322.
PubMed
7. Castro M, Abreu S, Sousa H, Machado L, Santos R, Vilas-Boas
JP. Ground reaction forces and plantar pressure distribution during
occasional loaded gait. Appl Ergon. 2013;44(3):503–509. PubMed
doi:10.1016/j.apergo.2012.10.016
8. Krupenevich R, Rider P, Domire Z, DeVita P. Males and females
respond similarly to walking with a standardized, heavy load. Mil Med.
2015;180(9):994–1000. PubMed doi:10.7205/MILMED-D-14-00499
9. Zadpoor AA, Nikooyan AA. The relationship between lower-extremity
stress fractures and the ground reaction force: a systematic review. Clin
Biomech (Bristol, Avon). 2011;26(1):23–28. PubMed doi:10.1016/j.
clinbiomech.2010.08.005
10. Simpson KM, Munro BJ, Steele JR. Backpack load affects lower
limb muscle activity patterns of female hikers during prolonged load
carriage. J Electromyogr Kinesiol. 2011;21(5):782–788. PubMed
doi:10.1016/j.jelekin.2011.05.012
11. Huang T-WP, Kuo AD. Mechanics and energetics of load carriage
during human walking. J Exp Biol. 2014;217(Pt 4):605–613. PubMed
doi:10.1242/jeb.091587
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

628 Liew et al
JAB Vol. 32, No. 6, 2016
12. Grifn TM, Roberts TJ, Kram R. Metabolic cost of generating mus-
cular force in human walking: insights from load-carrying and speed
experiments. J Appl Physiol (1985). 2003;95(1):172–183.
13. Mooney LM, Rouse EJ, Herr HM. Autonomous exoskeleton reduces
metabolic cost of human walking during load carriage. J Neuroeng
Rehabil. 2014;11:80. PubMed doi:10.1186/1743-0003-11-80
14. Peduzzi de Castro MP, Abreu S, Pinto V, et al. Influence of
pressure-relief insoles developed for loaded gait (backpackers and
obese people) on plantar pressure distribution and ground reaction
forces. Appl Ergon. 2014;45(4):1028–1034. PubMed doi:10.1016/j.
apergo.2014.01.005
15. Knapik J, Harman E, Reynolds K. Load carriage using packs: A review
of physiological, biomechanical and medical aspects. Appl Ergon.
1996;27(3):207–216. PubMed doi:10.1016/0003-6870(96)00013-0
16. Gill SV, Ogamba M, Lewis CL. Effects of additional anterior body
mass on gait. BMC Pregnancy Childbirth. 2016;16:109. PubMed
doi:10.1186/s12884-016-0893-0
17. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting
items for systematic reviews and meta-analyses: the PRISMA state-
ment. PLoS Med. 2009;6(7):e1000097. PubMed doi:10.1371/journal.
pmed.1000097
18. Open Grey. http://www.opengrey.eu/. Accessed Jun 31, 2014.
19. Grey Literature Report. http://greylit.org/. Accessed Jun 31, 2014.
20. von Elm E, Altman DG, Egger M, et al. The Strengthening the
Reporting of Observational Studies in Epidemiology (STROBE)
Statement: guidelines for reporting observational studies. PLoS Med.
2007;4(10):e296. PubMed doi:10.1371/journal.pmed.0040296
21. Higgins J. Cochrane Handbook for Systematic Reviews of Interven-
tion. In: Higgins J, Green S. (Eds.). The Cochrane Collaboration;
2011: www.cochrane-handbook.org.
22. Gravetter FJ, Forzano L-AB. Experimental designs: Within-subjects
design. In: Gravetter FJ, Forzano L-AB, eds. Research methods for
the behavioral sciences. 4th ed. Belmont, CA: Wadsworth Cengage
Learning; 2012:253–280.
23. Hall JP, Barton C, Jones PR, Morrissey D. The biomechanical differ-
ences between barefoot and shod distance running: a systematic review
and preliminary meta-analysis. Sports Med. 2013;43(12):1335–1353.
PubMed doi:10.1007/s40279-013-0084-3
24. LaFiandra M, Holt KG, Wagenaar RC, Obusek JP. Transverse plane
kinetics during treadmill walking with and without a load. Clin Bio-
mech (Bristol, Avon). 2002;17(2):116–122. PubMed doi:10.1016/
S0268-0033(01)00113-9
25. Quesada PM, Mengelkoch LJ, Hale RC, Simon SR. Biome-
chanical and metabolic effects of varying backpack loading on
simulated marching. Ergonomics. 2000;43(3):293–309. PubMed
doi:10.1080/001401300184413
26. Holewijn M. Physiological strain due to load carrying. Eur J Appl
Physiol Occup Physiol. 1990;61(3-4):237–245. PubMed doi:10.1007/
BF00357606
27. Bobet J, Norman RW. Effects of load placement on back muscle activ-
ity in load carriage. Eur J Appl Physiol Occup Physiol. 1984;53(1):71–
75. PubMed doi:10.1007/BF00964693
28. Charteris J. Comparison of the effects of backpack loading and
of walking speed on foot-floor contact patterns. Ergonomics.
1998;41(12):1792–1809. PubMed doi:10.1080/001401398185974
29. Gillet C, Foissac M, Leteneur S, Freychat P, Barbier F. Inuence of a
new backpack design on kinematics and dynamics of walking. Sports
Eng (Shefeld, England). 2006;9(4):246.
30. Goh JH, Thambyah A, Bose K. Effects of varying backpack loads on
peak forces in the lumbosacral spine during walking. Clin Biomech
(Bristol, Avon). 1998;13(Suppl 1):S26–S31. PubMed doi:10.1016/
S0268-0033(97)00071-5
31. Hsiang SM, Chang C. The effect of gait speed and load carrying on the
reliability of ground reaction forces. Saf Sci. 2002;40(7–8):639–657.
doi:10.1016/S0925-7535(01)00064-9
32. Kinoshita H. Effects of different loads and carrying systems on selected
biomechanical parameters describing walking gait. Ergonomics.
1985;28(9):1347–1362. PubMed doi:10.1080/00140138508963251
33. Laandra M, Harman E. The distribution of forces between the upper and
lower back during load carriage. Med Sci Sports Exerc. 2004;36(3):460–
467. PubMed doi:10.1249/01.MSS.0000117113.77904.46
34. Lee M, Kim S, Park S. Resonance-based oscillations could describe
human gait mechanics under various loading conditions. J Biomech.
2014;47(1):319–322. PubMed doi:10.1016/j.jbiomech.2013.09.011
35. Majumdar D, Pal MS, Majumdar D. Effects of military load carriage
on kinematics of gait. Ergonomics. 2010;53(6):782–791. PubMed
doi:10.1080/00140131003672015
36. Majumdar D, Pal MS, Pramanik A, Majumdar D. Kinetic changes
in gait during low magnitude military load carriage. Ergonomics.
2013;56(12):1917–1927. PubMed doi:10.1080/00140139.2013.835
871
37. Qu X. Effects of cognitive and physical loads on local dynamic stability
during gait. Appl Ergon. 2013;44(3):455–458. PubMed doi:10.1016/j.
apergo.2012.10.018
38. Tilbury-Davis DC, Hooper RH. The kinetic and kinematic effects
of increasing load carriage upon the lower limb. Hum Mov Sci.
1999;18(5):693–700. doi:10.1016/S0167-9457(99)00026-3
39. Wang H, Frame J, Ozimek E, Leib D, Dugan E. The effects of load
carriage and muscle fatigue on lower-extremity joint mechanics. Res
Q Exerc Sport. 2013;84(3):305–312. PubMed doi:10.1080/0270136
7.2013.814097
40. Wang H, Frame J, Ozimek E, Leib D, Dugan EL. Inuence of fatigue
and load carriage on mechanical loading during walking. Mil Med.
2012;177(2):152–156. PubMed doi:10.7205/MILMED-D-11-00210
41. Xu X, Hsiang SM, Mirka GA. The effects of a suspended-load
backpack on gait. Gait Posture. 2009;29(1):151–153. PubMed
doi:10.1016/j.gaitpost.2008.06.008
42. Yang X, Zhao G, Liu D, Zhou W, Zhao H. Biomechanics analy-
sis of human walking with load carriage. Technol Health Care.
2015;23(Suppl 2):S567–S575. PubMed doi:10.3233/THC-150995
43. Ling W, Houston V, Tsai YS, Chui K, Kirk J. Women’s load car-
riage performance using modular lightweight load-carrying equip-
ment. Mil Med. 2004;169(11):914–919. PubMed doi:10.7205/
MILMED.169.11.914
44. Lloyd R, Parr B, Davies S, Cooke C. A kinetic comparison of
back-loading and head-loading in Xhosa women. Ergonomics.
2011;54(4):380–391. PubMed doi:10.1080/00140139.2011.558636
45. Simpson K, Munro B, Steele J. Effect of load mass on posture,
heart rate and subjective responses of recreational female hikers to
prolonged load carriage. Appl Ergon. 2011;42(3):403–410. PubMed
doi:10.1016/j.apergo.2010.08.018
46. Simpson KM, Munro BJ, Steele JR. Effects of prolonged load car-
riage on ground reaction forces, lower limb kinematics and spatio-
temporal parameters in female recreational hikers. Ergonomics.
2012;55(3):316–326. PubMed doi:10.1080/00140139.2011.642004
47. Smith B, Ashton KM, Bohl D, Clark RC, Metheny JB, Klassen S. Inu-
ence of carrying a backpack on pelvic tilt, rotation, and obliquity in
female college students. Gait Posture. 2006;23(3):263–267. PubMed
doi:10.1016/j.gaitpost.2005.02.009
48. Caron RR, Wagenaar RC, Lewis CL, Saltzman E, Holt KG. Center
of mass trajectory and orientation to ankle and knee in sagittal
plane is maintained with forward lean when backpack load changes
during treadmill walking. J Biomech. 2013;46(1):70–76. PubMed
doi:10.1016/j.jbiomech.2012.10.004
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}

Biomechanics of Backpack Carriage 629
JAB Vol. 32, No. 6, 2016
49. Cook TM, Neumann DA. The effects of load placement on the
EMG activity of the low back muscles during load carrying by
men and women. Ergonomics. 1987;30(10):1413–1423. PubMed
doi:10.1080/00140138708966035
50. Devroey C, Jonkers I, de Becker A, Lenaerts G, Spaepen A. Evaluation
of the effect of backpack load and position during standing and walking
using biomechanical, physiological and subjective measures. Ergonom-
ics. 2007;50(5):728–742. PubMed doi:10.1080/00140130701194850
51. Fiolkowski P, Horodyski M, Bishop M, Williams M, Styli-
anou L. Changes in gait kinematics and posture with the use
of a front pack. Ergonomics. 2006;49(9):885–894. PubMed
doi:10.1080/00140130600667444
52. Hageman E, Hall M, Sterner E, Mirk G. Medial longitudinal arch
deformation during walking and stair navigation while carrying
loads. Foot Ankle Int. 2011;32(6):623–629. PubMed doi:10.3113/
FAI.2011.0623
53. Hall M, Boyer ER, Gillette JC, Mirka GA. Medial knee joint loading
during stair ambulation and walking while carrying loads. Gait Posture.
2013;37(3):460–462. PubMed doi:10.1016/j.gaitpost.2012.08.008
54. Holt K, Wagenaar R, LaFiandra M, Kubo M, Obusek J. Increased
musculoskeletal stiffness during load carriage at increasing walking
speeds maintains constant vertical excursion of the body center of
mass. J Biomech. 2003;36(4):465–471. PubMed doi:10.1016/S0021-
9290(02)00457-8
55. Holt KG, Wagenaar RC, Kubo M, LaFiandra ME, Obusek JP. Modu-
lation of force transmission to the head while carrying a backpack
load at different walking speeds. J Biomech. 2005;38(8):1621–1628.
PubMed doi:10.1016/j.jbiomech.2004.07.033
56. LaFiandra M, Wagenaar RC, Holt KG, Obusek JP. How do load
carriage and walking speed inuence trunk coordination and stride
parameters? J Biomech. 2003;36(1):87–95. PubMed doi:10.1016/
S0021-9290(02)00243-9
57. Lloyd R, Cooke C. Biomechanical differences associated with two
different load carriage systems and their relationship to economy.
Hum Mov Sci. 2011;12(1):65–74.
58. Lloyd R, Cooke CB. Kinetic changes associated with load carriage
using two rucksack designs. Ergonomics. 2000;43(9):1331–1341.
PubMed doi:10.1080/001401300421770
59. Lucas-Cuevas AG, Perez-Soriano P, Bush M, et al. Effects of differ-
ent backpack loads in acceleration transmission during recreational
distance walking. J Hum Kinet. 2013;37:81–89. PubMed doi:10.2478/
hukin-2013-0028
60. Neumann DA, Cook TM. Effect of load and carrying position on
the electromyographic activity of the gluteus medius muscle during
walking. Phys Ther. 1985;65(3):305–311. PubMed
61. Rose JD, Mendel E, Marras WS. Carrying and spine loading. Ergo-
nomics. 2013;56(11):1722–1732. PubMed doi:10.1080/00140139.2
013.835870
62. Sharpe SR, Holt KG, Saltzman E, Wagenaar RC. Effects of a hip
belt on transverse plane trunk coordination and stability during load
carriage. J Biomech. 2008;41(5):968–976. PubMed doi:10.1016/j.
jbiomech.2007.12.018
63. Son H. The effect of backpack load on muscle activities of the trunk
and lower extremities and plantar foot pressure in atfoot. J Phys Ther
Sci. 2013;25(11):1383–1386. PubMed doi:10.1589/jpts.25.1383
64. Wang YT, Pascoe DD, Weimar W. Evaluation of book backpack
load during walking. Ergonomics. 2001;44(9):858–869. PubMed
doi:10.1080/00140130118572
65. Yen SC, Gutierrez GM, Ling W, Magill R, McDonough A. Coor-
dination variability during load carriage walking: Can it contribute
to low back pain? Hum Mov Sci. 2012;31(5):1286–1301. PubMed
doi:10.1016/j.humov.2012.02.006
66. Yen SC, Ling W, Magill R, McDonough A, Gutierrez GM. Temporal
relationship between trunk and thigh contributes to balance control
in load carriage walking. Gait Posture. 2011;34(3):402–408. PubMed
doi:10.1016/j.gaitpost.2011.06.010
67. Caron RR, Lewis CL, Saltzman E, Wagenaar RC, Holt KG. Muscu-
loskeletal stiffness changes linearly in response to increasing load
during walking gait. J Biomech. 2015;48(6):1165–1171. PubMed
doi:10.1016/j.jbiomech.2014.12.046
68. Castro M, Figueiredo M, Abreu S, et al. The influence of gait
cadence on the ground reaction forces and plantar pressures during
load carriage of young adults. Appl Ergon. 2015;49:41–46. PubMed
doi:10.1016/j.apergo.2015.01.004
69. Watanabe K, Wang Y. Inuence of backpack load and gait speed on
plantar forces during walking. Res Sports Med. 2013;21(4):395–401.
PubMed
70. Abaraogu UO, Ugwa WO, Onwuka E, Orji E. Effect of single and
double backpack strap loading on gait and perceived exertion of young
adults. J Back Musculoskelet Rehabil. 2016;29(1):109–115. PubMed
doi:10.3233/BMR-150605
71. Dames KD, Smith JD. Effects of load carriage and footwear on
spatiotemporal parameters, kinematics, and metabolic cost of walk-
ing. Gait Posture. 2015;42(2):122–126. PubMed doi:10.1016/j.
gaitpost.2015.04.017
72. Ghori GM, Luckwill RG. Responses of the lower limb to load carrying
in walking man. Eur J Appl Physiol Occup Physiol. 1985;54(2):145–
150. PubMed doi:10.1007/BF02335921
73. Silder A, Delp SL, Besier T. Men and women adopt similar walk-
ing mechanics and muscle activation patterns during load carriage.
J Biomech. 2013;46(14):2522–2528. PubMed doi:10.1016/j.jbio-
mech.2013.06.020
74. Cavagna GA, Willems PA, Legramandi MA, Heglund NC. Pendular
energy transduction within the step in human walking. J Exp Biol.
2002;205(Pt 21):3413–3422. PubMed
75. Martin PE, Marsh AP. Step length and frequency effects on ground
reaction forces during walking. J Biomech. 1992;25(10):1237–1239.
PubMed doi:10.1016/0021-9290(92)90081-B
76. Liu MQ, Anderson FC, Pandy MG, Delp SL. Muscles that sup-
port the body also modulate forward progression during walking.
J Biomech. 2006;39(14):2623–2630. PubMed doi:10.1016/j.jbio-
mech.2005.08.017
77. Riley PO, Paolini G, Della Croce U, Paylo KW, Kerrigan DC. A
kinematic and kinetic comparison of overground and treadmill walk-
ing in healthy subjects. Gait Posture. 2007;26(1):17–24. PubMed
doi:10.1016/j.gaitpost.2006.07.003
78. Soo CH, Donelan JM. Coordination of push-off and collision deter-
mine the mechanical work of step-to-step transitions when isolated
from human walking. Gait Posture. 2012;35(2):292–297. PubMed
doi:10.1016/j.gaitpost.2011.09.102
79. Teng HL, Powers CM. Inuence of trunk posture on lower extremity
energetics during running. Med Sci Sports Exerc. 2015;47(3):625–630.
PubMed doi:10.1249/MSS.0000000000000436
80. Zajac FE, Neptune RR, Kautz SA. Biomechanics and muscle coordi-
nation of human walking: part II: lessons from dynamical simulations
and clinical implications. Gait Posture. 2003;17(1):1–17. PubMed
doi:10.1016/S0966-6362(02)00069-3
81. Prilutsky BI, Zatsiorsky VM. Tendon action of two-joint muscles:
Transfer of mechanical energy between joints during jumping,
landing, and running. J Biomech. 1994;27(1):25–34. PubMed
doi:10.1016/0021-9290(94)90029-9
82. Kuo AD, Donelan JM, Ruina A. Energetic consequences of walking
like an inverted pendulum: step-to-step transitions. Exerc Sport Sci Rev.
2005;33(2):88–97. PubMed doi:10.1097/00003677-200504000-00006
Downloaded by CURTIN UNIV on 02/27/18, Volume ${article.issue.volume}, Article Number ${article.issue.issue}